MOLECULAR GENETIC MANIPULATION OF WINE YEASTS

Transcription

1 1,n. 1. L2_n> tr"ffirrlìí{t 'ir1. ;1.,:.,.: Ì i.uiri^ YO té \.,q MOLECULAR GENETIC MANIPULATION OF WINE YEASTS BY JENNY E. PETERING B.Sc.(Hns) Mnash University A thesis submitted fr the Degree f Dctr f Philsphy in the Faculty f Agricultural and Natural Resurce Sciences at The University f Adelaide. Deparrnent f Plant Science Waite Agricultural Research Institute The University f Adelaide December 1991

2 -11- DECLARATION The wrk presented in this thesis is my wn unless therwise acknwledged, and has nt previusly been submitted t any university fr the award f any degree f diplma. This thesis may be made available fr lan rphtcpying prvided that an acknwledgement is made in the instance f any reference t this wrk. Jenny E. Petering December 1991

3 111 ACKNOWLEDGMENTS I sincerely thank my supervisrs Dr Peter Langridge and Dr Paul Henschke fr their invaluable guidance, supprt and friendship thrughut my candidature. Their cntinual encuragement and advice was essential t my prgress, and was very much appreciated. I wuld like t acknwledge The Australian V/ine Research Institute fr the prvisin f an Australian Grape and Wine Research Cuncil schlarship, and I thank the staff and students f the institute fr their cpedin and technical advice. Varius staff members f the V/aite and V/ine Institutes have prvided me with assistance during the curse f this study. I specifically thank Andrew Dunbar (phtgraphy), Lynne Giles (statistical analyses), Jan Nield and Nancy Davis (labratry rganisatin). I am als very grateful t my fellw pstgraduate students and friends, especially Carle Smith, Graham Due, Michael Burnet, Anne Tassie, David Hein (Zeke), Phil Murphy, Jack Christpher and Dene Cuthbertsn fr their camaraderie and humur ver the past few years. Finally, and mst imprtantly, I thank my family - especially my parents- fr abslutely everything but particularly fr their cntinual supprt and encuragement thrughut my studies.

4 -tv- Publicatins arising frm this thesis Petering, J., P. Langridge and P. Henschke (1988). Fingerprinting wine yeasts. Aust. NZ V/ine Ind. J. 3Q): a8-52. Petering, J.E., P.A. Henschke and P. Langridge (1991). The Escherichia cli þ- glucurnidase gene as a ma ker fr Sacclnrmyces yeast strain identificatin. Am. J. Enl. Vitic. 42:6-12. Petering, J.E., M.R. Symns, P. Langridge and P.A. Henschke (1991). Determinatin f killer txin activity in fermenting grape juice by using a marked Saccharmyc s wine yeast strain. Appl. Env. Micrbil. 57: Cnference prceedings Petering, J.E., P.A. Henschke and P. Langridge (1989). Genetic engineering f wine yeasts. Prc. 7th Aust. rü/ine Ind. Tech. Cnf. Adelaide, Australia. pp Petering, J.E., P. Langridge and P.A. Henschke (1990). Identificatin f yeast strains by mlecular bilgy techniques. Prc. 9th Int. Oenl. Symp. Cascais, Prtugal. pp

5 -v- Summary A system has been established fr the transfrmatin f wine yeast strains. This system utilizes a mutant Sacclarmyces cerevisíae acetlacøte synthase gene which cnfers resistance t the herbicide sulfmeturn methyl and acts as a dminant selectin ma ker. Satisfactry transfrmatin efficiencies were achieved with bth a self-replicating plasmid and an integrating vectr in several wine yeast strains. The integrating vectr was successfully targeted t the ILV2lcus n chrmsme XIII f the yeast genme, and was stably maintained thrughut fermentatin. Fermentatin trials indicated that the transfrmatin system des nt adversely affect the grwth kinetics r fermentatin rate f the wine yeast strain. Similarly, there were n significant differences in the ph measurements r the alchl cntent f wines prduced by the parent and transfrmed stríuns. This transfrmatin system was used t develp a prcedure fr the genetic marking f wine yeast strains. Marking was achieved by the intrductin int wine yeasts f the Escheríchia cli p-glucurnidase (GUS) gene. The GUS gene was adapted fr expressin in yeast by ligating the Sacclarmyces cerevisiae alchl dehydrgenase prmter and terminatr sequences t the cding regin f the Escherichia cli uida gene. The GUS cnstruct was intrduced int the genme f a wine yeast strain by integratin int chrmsme XIII. The marked strain did nt shw any significant differences in fermentatin perfrmance when cmpared with the riginal parent strain. Stability f the marker was cnfirmed by the bservatin that the GUS cnstruct was maintained in > 99V f the ttal yeast ppulatin at the end f the fermentatin. A simple assay prcedure was develped t detect GUS activity in the marked yeast cells r clnies. This methd prvides a means fr genetic marking and subsequent rapid identificatin f wine yeast strains f chice.

6 -vl- The applicatin f the marked strain t studies in enlgy was demnstrated in tw separate investigatins. First, the efficiency f killer txin in fermenting grape juice was determined. The ma ked killer strain was cured f its M-dsRNA genme t enable direct assessment f the efficiency f killer txin under fermentatin cnditins. Killer activity was clearly evident in fermenting Riesling grape juice f ph 3.1 at 18C but depended n the prprtin f killer t sensitive cells at the time f inculatin. Killer activity was detected nly when the rati f killer t sensitive cells exceedd l:2. At the highest rati f killer t sensitive cells tested (2:1), cmplete eliminatin f sensitive cells was nt achieved. Secndly, inculatin efficiency f the marked strain under varius fermentatin cnditins was analysed. Va iables studied were time f inculatin, additin t the must f SO2 (100 mgll),prductin f killer txin and fermentatin temperature. Results indicated that dminance f the ferment by the inculated strain culd be ensured by an early inculatin regardless f any ther variable. A cmbinatin f SO2 treatment and lw fermentatin temperature (10C) was als effective in ensuring inculatin success. This thesis describes, therefre, the establishment f wine yeast transfrmatin prcedure and subsequent use in the develpment f a yeast genetic marking system; and demnstrates the applicatin f ma ked yeast strains t the wine industry.

7 -v11- Table f Cntents Chapter I General intrductin and Prject aims. Page 1 Chapter 2 Literature Review.. 2.1, 2.2 ORIGIN AND CLASSIFICATION OF WINE YEASTS.. GENETIC FEATURES OF WINE YEASTS Life cycle and sprulatin Chrmsmes and plidy Extrachrmsmal elements GENETTC TECHNIQUES FOR YEAST STRAIN IMPROVEMENT Clnal selectin Clnal selectin after mutagenesis Hybridizatin Rare mating Spherplast fusin Recmbinant DNA technlgy TARGETS FOR WINE YEAST STRAIN DEVELOPMENT Sedimentatin and flcculatin Nn-faming mutants Sulfite and Sulfide Prductin CONCLUSIONS Ethanl tlerance. Higher alchls.. Killer factr. Mallactic fermentatin Genetic marking.. Chapter 3 Materials and Methds 3.1 STRAINS AND MEDIA 3.2 YEAST TRANSFORMATION METHODS Alkali catin transfrmatin Spherplast fusin methd Electrpratin f intact yeast cells RECOMBTNANT DNA TECHNIQUES I Electrphresis f DNA Bacterial transfrmatin. Restrictin digests, fragment islatin and 1igatin Plasmid islatin Yeast DNA islatin..

8 -v Suthern hybridizatin PROTEIN SYNTHESIS 4N4LYSIS L Incrpratin f radiactivity int yeast prteins Electrphresis f prteins PULSED FIELD GEL ELECTROPHORESIS p-glucuronidase (GUS) ASSAYS. 3.6.L Enzyme activity assays Agar plate assays Micrscpic visualizatin KILLER YEAST MANIPULATIONS L Curing f killer strain 34M Assay fr cured strain dsrna islatin MICROVINIFICATION TRIALS Fermentatin prcedure Wine analysis LARGE SCALE LABORATORY FERMENTATIONS Chapter 4 Establishing a system fr the genetic manipulatin f wine yeasts.. 4,1 INTRODUCTION... 4.t.t 4.t.2 Selectin systems fr identifying transfrmants... Gene transfer vectrs... Tranfrmatin prcedures... 4.r.3 4.t.4 Expressin and secretin f freign prteins Chapter aims 4.2 RESULTS r Chice f a selectable marker Chice f a transfrmatin prcedure Analysis f transfrmants Fermentatin trials Gene regulatin during fermentatin DISCUS SION. Chapter 5 Develpment f a system fr wine yeast strain marking and identificatin 5.1 INTRODUCTION Yeast chrmsmal fingerprints Marked strains RESULTS &

9 -D( V/ine yeast chrmsme fingerprinting GUS-Vectr cnstruct. Transfrmatin and Suthern analysis Develpment f a GUS assay fr yeasts p-glucurnidase enzyme activity Fermentatin triais Stability f GUS cnstruct. 5.3 DIS CUS SION. Chapter 6 Determinatin f killer txin activity in fermenting grape juice using a marked S ac charmy c e s strain INTRODUCTION RESULTS Curing f Strain 34M Analysis f killer activity during fermentatin. 6.3 DISCUSSION. Chapter 7 Cmparisn f fermentatin cnditins by use f a marked strain 7.1 INTRODUCTION L Killer yeast inculatin... 7.I.2 Additin f sulfur dixide t grape must Temperature f fermentatin. 7.2 RESULTS Fermentatins cnducted at 20C Fermentatins cnducted at l0c Killer activity in the indigenus yeast ppulatin 7.3 DISCUS SION r02 r t1l tt2 tl4 lt7 It9 119 Chapter I References General Cnclusins r25 r27 Appendix 1 Pubticatins 156

10 -1- W/\ITE INSTITUÏE!IFRÀRY chapter 1.General intrductin and Prject aims Yeasts were emplyed in the prductin f wine fr several thusand years befre their existence was recgnised by Antnie van læeuwenhek in During the secnd half f the nineteenth century, Luis Pasteur demnstrated that living yeast cells were respnsible fr fermentatin, r the cnversin f sugar t ethanl and ca bn and Slmn, 1981). Originally, yeasts present n grape skins and equipment were respnsible fr the "spntaneus" fermentatin invlved in wine-making. Fr the last century, hwever, the availability f pure yeast cultures has imprved reprducibility in fermentatins and the quality f the prduct (Tubb and Hammnd, 1987). It is nw pssible t select specific yeast str ins n the basis f their fermentatin perfrmance and the characteristics f their prduct. Althugh the wine indusüry rapidly adpted pure culture inculatin technlgy, it has taken little active interest in yeast genetics and strain develpment prgrams (Thrntn, 19S3). Recent trends in the wine market, hwever, demand the mdificatin f traditinal wine yeast strains in the develpment f mre cst-effective wine making practices. Twa d this end, it is imprtant t define the requirements f the wine indusury in genetic terms - t select specific targets fr yeast breeding prgrams. In rder t identify these targets, it is necessary t cnsider the mst desirable characteristics f a wine yeast (Thrntn,1983): - the efficient cnversin f gape sug lr t alchl; - the rapid initiatin f fermenøtin immediately upn inculatin; - the ability t ferment at lw temperatures such as 10-14C; - tlerance t sulfur dixide (used in wine making as a sterilizing agent and as an anti-xidant); - tlerance t ethanl (in rder t ferment t dryness);

11 -2- - lw faming activity; - lw prductin f vlatile acids, acetaldehyde and sulphite; - lw hydrgen sulfide r mercaptan prductin; - relatively lw higher alchl prductin; - effective flcculatin at the end f fermentatin t aid clarificatin; - relatively high glycerl prductin t cntribute t the sensry qualities f the wine; - the prductin f desirable fermentatin buqueq - resistance t killer txins and ther zymcidal cmpunds. T date, n wine yeast in cmmercial use has all the cha acteristics listed abve, and it is well established that wine yeasts vary in their wine making abilities. The majr surce f this va iatin is the genetic cnstitutin f the wine yeasts (Thrntn, 1983). Althugh sme f the requirements listed abve are cmplex and difficult t define genetically withut a better understanding f the bichemistry invlved, they cmprise specific targets fr wine yeast mdificatin prgrams. - Other pssible targets fr strain mdificatin are prcesses invlved in yeast management and in the synthesis f new prducts. Fr example, genetic manipulatin can be used t insert specific markers int wine yeast strains as an aid t strain identificatin. This wuld be particularly useful fr wineries using mre than ne straid, and in the ptimizatin f must prcessing strategies. An area which has nt yet been explited is the use f wine yeasts t prduce valuable by-prducts. Spent yeast frm industrial prcesses is already used t prduce yeast extracts and as an ingredient in fds and flavurings. Pssibilities exist t prduce higher value materials such as vitamins, enzymes, carbhydrates and lipids frm wine yeasts. Genetic manipulatin culd facilitate maximum frmatin f the required prduct.

12 -3- These and ther ptential applicatins f genetic manipulatin t industrial yeast strains have been recgnised fr sme time (Spencer and Spencer, 1983). Since the pineering genetic studies f V/inge (1935), in which the basic life-cycle f Sacclnrmyces was established, yeast strains have been interbred t prduce new hybids. Classical genetic methds which have been invlved in yeast manipulatin prgrams t date include mutatin and selectin, hybridizatin, rare mating and spherplast fusin (Tubb and Hammnd, 1987). These techniques have enjyed nly limited success in the past, essentially fr tw reasns. First, methds such as crss breeding invlve mating which requires spre frmatin; industrial yeasts, hwever, tend t be plyplid, which means they have a reduced ability t frm spres. Secndly, althugh ther techniques d nt require spre frmatin n the part f yeast strains, they d invlve grss and nn-specific exchanges f genetic material. Therefre, while it is pssible t intduce favurable cha acteristics int yeast strains using these techniques, it is quite prbable that deleterius prperties will be intrduced simultaneusly. The relatively new methds f recmbinant DNA technlgy vercme sme f the prblems inherent in the mre traditinal techniques. Recmbinant DNA technlgy des nt require spre frmatin by the parental yeast strains. It is, by nature, highly specific in its actin as it allws the transfer f single genes frm ne yeast strain t anther. Anther advantage f this technlgy is that it ffers the ptential t intrduce genes frm any rganism (plant, animal, bacterial r fungal) int yeast strains. The applicatin f recmbinant DNA technlgy t industrial yeast strains is still a relatively yung area f research. Althugh the brewing indusç has invested sme effrt in this field, the technlgy is still nt yet well enugh established fr cmmercial r industrial use. There have been n reprts t date f the successful applicatin f recmbinant DNA technlgy t wine yeast strain imprvement. The aim f this prject is t establish a system fr genetic engineering f wine yeast strains which will be suitable fr industrial applicatin. The fllwing criteria a e cnsidered mst imprtant in achieving

13 -4- this aim: an efficient transfrmatin methd fr the intrductin f freign DNA int wine yeasts; demnstratin that the intrductin f freign DNA des nt adversely affect the fermentatin perfrmance f the yeast strain; stability f the freign DNA in the yeast ppulatin thrughut the fermentatin; and apprpriate expressin f the freign gene in the yeast cell. A further bjective is t utilise this sysæm in the develpment f a strain which will have cmmercial applicatin. The targeted area fr develpment will be the genetic marking f wine yeast strains. The aims here a e twfld. First, t intrduce a nvel prperty int wine yeasts t enable rapid and unequivcal strain identificatin. Secndly, an attempt will be made t demnstrate the ptential applicatin f such a ma ked strain in enlgical studies.

14 -5- Chapter 2 Literature Review 2.T ORIGIN AND CI/.SSIFICATION OF WINE YEASTS The riginal wine yeasts strains were derived frm the natural ppulatin f yeasts that ccur n the skins f grapes. Reprts f the islatin f yeast species frm grapes, wines and winery equipment frm the different wine regins a und the wrld have been reviewed by Kunkee and Amerine (1970), Kunkee and Gswell (1977), Benda (1982), Lafn-Lafurcade (1983) and Farkas (1983). Apprximately 200 wine yeast species are listed in the table by Kunkee and Gswell (1977), hwever, many f these species are nly ccasinally assciated with grapes and wines. Accrding t Lafn-Lafurcade (1983), there a e nly abut 14 yeast species that are frequently islated frm musts and wines and a further 4O species which are smetimes present. The general cnclusins that have been drawn frm these reprts are that a prgressin f yeast species are invlved in natural fermentatin. Grapes and freshly extracted grape juice have a dminant flra f apiculate yeasts such as Kleckera apiculata and Hanseniaspra uvarurn, as well as species f Candida, Hansenula, Metschníkwia and Pichia. These species exhibit limited grwth during the early stages f alchlic fermentatin but are inhibited as the ethanl cncentratin rises t 3-4V (vvvl). The mre alchl tlerant species f. Søcclnrmyces, ften initially present in smaller numbers, then prliferate and dminate the remainder f the fermentatin. Fermented wines may be spiled by the grwth f alchl tlerant species such as Zygsacclnrmyces bailii, species f Sacclnrmyces and film-frming yeasts such as Pichiamentbranaefaciens. V/ine yeasts f the genus Sacclørmyc s have been catagnzed taxnmically t at least 29 different species r varieties (Ldder, 1970; Kunkee and Gswell, 1977). Mst f these species are separated primarily n the basis f their sugar fermentatin and

15 -6- assimilatin patterns. In many cases, hwever, a mutatin in a single gene can result in the lss f capacity t ferment a sugar - few wuld cnsider such a mutant strain as a new species. A mre fundamental means f classificatin is based n DNA sequence hmlgy. Vaughan and Martini (1980) have reprted ttrat ttre G+C cntent f the DNA f S. cerevisiae, S. bayanus, S. chevalieri, S. italicils, and S. uvarum ranges frm 38.5 t 39.5V. Furthermre the DNA sequence hmlgy is greater than9d%. They cncluded that it is best t cnsider all these strains as belnging t the ne species; S. cerevisiae having taxnmic pririty. Because f its dminance in alchlic fermentatin,.s. cerevisiae has emerged, almst universally, as the single mst imprtant species assciated with the winemaking prcess. As a cnsequence, this species is nw widely recgnised as 'the wine yeast'. In sme established vineyards f Eurpe, grape juice is still allwed t ferment naturally with the yeasts riginating frm the grapes and the flra established n winery equipment. Hwever, in cuntries such as Australia, USA and Suth Africa, many winemakers use pure yeast starter cultures (usually a strain f S. cerevisiae) t inculate the must. Pure culture inculatin was first described by Hansen (1886, 1888), wh intrduced its use int brewery practice. The advantages f inculating with a pure culture f S. cerevisiae are that it prmtes a rapid and even nset f fermentatin; allws the cnduct f a cntrlled fermentatin; and inhibits the grwth f indigenus yeasts which may detract frm wine quality and even lead t wine spilage (Rankine and Llyd, 1963; Rankine, 1977; Kunkee, 1984). Althugh the term pure culture is still in use, is shuld be nted that this des nt necessarily mean that the cultwe is genetically unifrm. The cultures re pure in the sense that they were derived frm a single cell. Hwever, after years f mass prpagatin f these cultures, mutatins are likely t ccur and may be expressed (even thugh they are recessive) thrugh mittic crssing-ver r gene cnversin. An example f the hetergeneity f a'pure' culture is given by Zimmerman (1978) wh islated a strain with

16 -7- cnsiderably imprved characteristics frm successive single-cell cultures f an Epernay yeast. 2.2 GENETC FEATURES OF WINE YEASTS Genetic studies with S. cerevisiae were pineered at the Carlsberg Labratries in the 1930's by Winge wh first bserved haplid and diplid phases in the life cycle f Sacchnrmyces (Winge, 1935). The early literature n yeast genetics has been reviewed by Lindegren, L949;V/inge and Rberts, 1958; and Mrtimer and Hawthrne, A brief review f the genetics f S. cerevisiae will be described here, particularly as it relates t wine yeast strains and methds fr strain mdificatin. 2.2.t Life c] cle and sprulatin S. cerevísiae can exist in either the haplid r diplid state. Strains in which the haplid frm is stable and can be maintained fr many generatins, are termed heterthallic. The haplids frm such strains exist as ne f the tw mating types, MATa r MATcr, and mate t frm diplids when a cell f ne mating type cmes int cntact with a cell f the ther mating type. Strains in which cell fusin and diplid frmatin ccur amng cells derived frm a single spre a e termed hmthallic. This behaviur is caused by the allele HO (Harashima et al.,197 ;Hicks and Herskwitz, t976), such strains being gentypically HO/HO. The presence f the HO gene brings abut a high frequency f switching between mating types during vegetative grwth. Under the influence f this gene, the mating type lcus, MAT, readily changes frm MATa t MATø r vice versa. The MAT gene is fund n chrmsme III f the yeast genme tgether with tw silent genes HMLcr and HMRa which prvide the infrmatin t allw the switch f mating type at the MAT lcus. Cells f hmthallic yeasts have t bud at least nce befre they are cmpetent t switch mating

17 -8- type, but thereafter a high frequency f switching ccurs at each budding fr many generatins (flerskwitz and Oshima, 1981). The surveys that have been made f wine yeasrs indicate that they are t pically hmthallic (Thrntn and Eschenbruch,1976; Snw, L979; Kusewicz and Jhnstn, 1980). Because f pr spre viability it has in many cases been impssible t islate cmplete tetrads, and therefre it has nt been pssible t determine whether the riginal strain is gentypically HOÆIO r HO/h. The verall infrmatin available hwever, indicates that the majrity f wine yeasts are gentypically HO/HO. In bth hmthallic and heterthallic strains, mating takes place when the cells f ppsite mating type cme int clse prximity. Cells f a-mating type prduce an ligpeptide (12 r 13 amin acid residues) calted ct-factr which arrests a-mating type cells in the Gl phase and causes a-cells and c-cells t adhere t each ther. Cells f a-mating type prduce a-factr which has simila effects n cr<ells. In the presence f these factrs the cells adhere and cytplasmic fusin takes place t frm a heterkaryn. Nuclear fusin fllws rapidly t give î zy5te (Lindegren and Lindegren, 1943). By subsequent cell divisin this frms the diplid phase f the yeast life cycle which can be stably maintained fr many generatins. Meisis and sprulatin f diplid cells is triggered by nitrgen deprivatin in the presence f a nn-fermentable ca bn surce, and will nly ccur if MATa and MATa genes a e bth present. Fllwing entry int meisis the chrmsmes in the yeast nucleus underg premeitic DNA synthesis, pairing, recmbinatin and segregatin. Spre walls grw and envelpe fur haplid genmes (tw each f a and d mating types), frming the characteristic fur-spred ascus. V/hen placed in suitable nutrient media, the spres germinate t frm haplids and begin the cycle nce mre. Hwever, wine yeasts behave very differently frm labratry strains: they generally sprulate inefficiently, they prduce few viable spres f which mst are unable

18 -9- t mate, their chrmsmal cnstitutins are unknwn, they shw a grert deal f genetic hetergeniety and generally lack selectable genetic ma kers (Snw, 1983; Spencer and Spencer, 1983; Beckerich et al.,1984; Subden, 1987; Rank et a1.,1988) Chrmsmes and lidy The chrmsmes f S. cerevisíae a e lcated in the cell nucleus and accunt fr f the ttal yeast DNA (Petes, 1980). In haplid strains, chrmsmal DNA has a mlecular weight f l0l0da, which is equivalent t kilbase pairs (kbp). T date, apprximately 750 lci have been mapped t 17 chrmsmes (Mrtimer et a1.,1989). Each chrmsme is a single DNA mlecule f between 150 and 2500 kbp. As is the case with higher eukarytes, yeast chrmsmes als cntain basic histne mlecules. In cntrast with higher rganisms, hwever, S. cerevisiae DNA cntains a relatively small fractin f repeated andzaktan, 1981). Mbile genetic elements are fund in Saccharmyces strains. These elements (called Ty) cnsist f a 5.1 kbp DNA sequence flanked by a 250 bp repeated sequence (Beke et a1.,1985). As many as 35 cpies f Ty can be present per haplid genme and their ability t ranspse frm ne chrmsmal lcatin t anther can result in substantial rearrangements f the genme (Scherer et a1.,1982). The randm excisin and insertin f Ty elements int the genmes f wine yeasts can therefre inactivate genes encding desirable prteins and cause genetic instability f selected strains. The reverse can als ccur, s that imprved wine yeast strains evlve. In mst labratry studies, the strains f. S. cerevisiøe used are either haplid r diplid. Industrial strains, hwever, are predminantly diplid r plyplid. The precise determinatin f the chrmsme number f yeast strains is difficult, since they are t small fr direct chrmsme cunts. Methds including the determinatin f DNA-cntent per cell, measurement f cell vlume, and irradiatin and death rate have been used t

19 -10- esrimate the ptidy f yeast strains (Gunge and Nakatmi, L97I; Lewis et al.,1976; Aigle et a1.,1983; Takagi et al., 1985). These prcedures, hwever, are prblematic. The determinatin f DNA cntent is dependent n very specific cell cncentratins. Va iatin in the chrmsmal sizes f industrial strains culd als affect the precisin f the test. Althugh the cell size is clearly a functin f plidy, mst individual strains f the same plidy have sizes significantly different frm ther strains in the same plidy grup. Furthermre, aneuplidy cannt be determined by these methds. The majrity f attempts t estimate the plidy f brewing and distitling yeasts have relied n measuring the DNAcntent per cell and cmparing this with the value btained frm defined haplid strains. Results frm these studies suggest that many brewing and distilling yeasts are plyplid, particularly riplid, tetraplid r aneuplid (Tubb and Hammnd, 1987) Where strains have been crssed t labratry haplids, segregatins f genetic ma kers can prvide insight int the plidy f industrial yeast strains. Using this apprach, Cummings and Fgel (1978) were able t shw that tw wine yeasts were almst certainly nrmal diplids since matings f ascspres with cells f labratry strains gave regulat 2;2 segregatins fr markers n 13 f the 16 knwn yeast chrmsmes. One f the wine yeast strains studied by Thrntn and Eschenbruch (1976) was als prbably diplid, as it shwed 2:2 segregatins fr mst markers n six different chrmsmes. On the ther hand, Takahashi (1978), in a study f a widely used cmmercial German wine yeast (Hefix 1000), cncluded that it was abut, if nt exactly, tetraplid and had an alalala mating lcus gentype. Circumstantial evidence supprting the wide-spread ccurrence f aneuplidy and/r plyplidy amng wine yeast strains includes bservatins f pr spre viability, great variability in grwth rates amng spre prgeny, and a very lw frequency f matingcmpetent meitic segregants. It is nt yet clea whether plyplidy n wine yeasts is advantageus. Sme resea chers claim that the plyplid state might enable industrial yeasts t harbur a high dsage f genes imprtant fr efficient fermentatin (Mwshwitz,

20 -11- L979; Stewart et a1.,1981). It is knwn, hwever, that plyplid/anueplid state makes analyses f imprtant enlgical traits mre difficult and cmplicates the genetic imprvement f wine yeast strains ExEachrmsmal elements A number f extrachrmsmal genetic elements have been described in yeast, and are discussed belw. 2 um DNA Mst labratry strains f S.cerevisiae cntain a class f small extrachrmsmal DNA mlecules that are abut 2pm in length (Sinclair et al.,1967). The mlecules are generally referred t as the 2pm plasmids. There are usually 50 t 100 cpies f 2pm DNA per cell and they represent apprximately 5V f the ttal yeast DNA. These circular DNA mlecules cnsist f tw identical repeats f 599 bp separated by tw unique regins f 2774 and 2346 bp (Brach, 1981). Prtins f the DNA a e transcribed int three different plyadenylated mrna mlecules which can direct prtein synthesis ínvitr. One f these genes (FI-P) prduces a prtein which is actively invlved in 2pm recmbinatin (Cx, 1983), while ttre thers (REPI and REP2) are required fr stable replicatin (Brach, 1982). Their functin in the yeast cell has nt been established and since phentypically nrmal yeast strains have been identified that lack the 2 pm plasmid (Livingstn,1977), they are nt required fr cell viability. Other than its wn maintenance, the 2 rm plasmid appears t cnfer n advantage n the hst cell. Althugh mst f this infrmatin has been btained with labratry strains, a2 tm DNA f similar structure is fund in wine yeast strains. The 2pm DNA seryes an imprtant tl in the genetic manipulatin f wine yeasts, as many plasmid vectrs are based n the 2pm rigin f replicatin.

21 -t2- Mitchndrial DNA In labratry strains, mitchndrial DNA has a mlecular weight f.abut 50 x 106 Da and cnsists f a 75 kbp circular mlecule. It is present at mlecules per haplid cell and represents between 5 and 207 f the ttal cell DNA. It is very A-T rich cmpared with chrmsmal DNA resulting in a lwer buyant density (Fangman andzakran, 1981). Mitchndrial DNA shws typical cytplasmic inheriønce and its replicatin is independent f nuclear cntrl, taking place thrughut the cell cycle (Newln and Fangman, L975). The mitchndrial genme caries the genetic infrmatin fr nly a few essential mitchndrial cmpnents; mre thang}v f mitchndrial prteins are cded by nuclear genes (Dujn, 1981). Mutatins in mitchndrial DNA prduce petite strains which are unable t utilize nn-fermentable substrates. Such respiratry-deficient mutatins can range frm pint mutatins (mir) thrugh deletin mutatins (p-) t cmplete eliminatin f mitchndrial DNA (p). The generatin f petite mutants f wine yeasts ccurs spntaneusly at quite high rates. It is imprtant t nte, hwever, that yeasts with different mtdnas can differ in their flcculatin characteristics, lipid metablism, higher alchl prductin and the frmatin f flavur cmpunds (Lewis et al., 7976; Hammmnd and Eckersly, 1984). This indicates the imprtance f mtdna encded functins. Fr this reasn, petite strains are nt used fr wine making. Killer factr Many strains f Saccharmyces yeast cntain cytplasmic duble-stranded RNA mlecules (dsrna), encapsidated in virus-like particles (Tipper and Bstian, 1984). There are tw main varieties f dsrna mlecules with characteristic prperties. L dsrna is present in mst yeast strains and encdes the capsid prtein f the virus-like particles. M dsrna is present nly in killer strains f. Saccharmyces (Wickner, 1983) and has been

22 -13- shwn t encde bth a killer txin, which is lethal t sensitive strains, and the immunity factr which prevents self-killing. Bth L and M dsrna mlecules are linear but they have little sequence hmlgy. In killer strains there are nrmally abut 12 cpies f M dsrna per cell and abut 100 cpies f L dsrna per cell. Killer strains can be cured f M dsrna by grwth at elevated temperature r by treatment with cyclhexamide (Fink and Styles, 1973); such cured strains ften prduce mre L dsrna. The maintenance, regulatin and expressin f M dsrna a e all regulated by nuclear genes, many f which have been mapped (Wickner, 1983). Killer activity has been detected in yeasts islated frm estabtished vineyards and wineries in va ius regins f the wrld including Eurpe and Russia (Barre, 1984; Gaia, 1984: Naumv and Naumva, 1973), Suth Africa (Tredux et a1.,1986) and Australia (Heard and Fleet, 1987a,b). Killer factrs are f interest t the wine industry fr tw reasns. First, van Vuuren and Wingfield (1986) shwed that stuck r sluggish wine fermentatins can be caused by cntaminating killer yeasts. Secndly, in thery, selected killer yeasts culd be used as the inculated strain t suppress grwth f undesirable wild strains f Saccharmyces cerevísiac during grape juice fermentatin. Other senetic elements Several minr genetic elements have been described ver the years. The y factr (Cx, 1965) is a ptentiatr f nnsense suppressin. It is inherited cytplasmically and has been shwn nt t be assciated with dsrna, 2pm DNA r the mitchndrial genme. ture3l allws cells grwing n ammnia r glutamate t use ureidsuccinate and s bypass ura2 mutants. It is inherited cytplasmically but is nt distributed t all prgeny at meisis (Aigle and Lacrute, 1975). 20S RNA is a sprulatin-specifrc RNA mlecule whse synthesis is cntrlled by a cytplasmic genetic element. It is nt present in vegetative cells and is nly prduced by cells under sprulatin cnditins (Kadwaki and Halvrsn, l97 L).

23 -14- Due t the lack f apprpriate investigatins, nne f these minr genetic elements have been described in wine yeast strains. 2.3 GENETIC TECHNIQUES FOR YEAST STRAIN IMPROVEMENT Several genetic techniques can be used in yeast imprvement prgrams. Sme f these techniques are used t recmbine r rearrange the entire genme, whereas thers alter specific regins f the genme. Techniques having the greatest ptential in the genetic imprvement f wine yeast strains include clnal selectin; clnal selectin after mutagenesis; hybridizatin; spherplast fusin; ra e-mating; and recmbinant DNA technlgy. All f these methds have been used with industrial strains (either brewing, distilling r wine yeasts) and will be described belw Clnal selectin This methd takes advantage f the natural genetic va iatin present in wine yeast strains. Sme degree f heterzygsity is almst certain t be present in all yeast strains and new substrains may arise, pssibly thrugh mutatin, but mre likely thrugh mittic recmbinatin during vegetative grwth. The selectin prcedure requires the testing f large numbers f clnes derived frm single cells f the parental strain. An example f the successful use f clnal selectin in which va iants f an Epernay yeast with imprved fermentatin characteristics were islated has already been cited (Zimmerman, 1978). It has als been used t select nn-faming wine yeast mutants (Ouchi and Akiyama, 1971; Eschenbruch and Rassell, 1975); variants with imprved ethanl tlerance (Brwn and Oliver, 1982); and strains with reduced H2S prductin (Rupela and Taura, 1984).

24 Clnal selectin after mutagenesis Selectin after mutagen treatment has been used in the imprvement f wine, brewing, baking and distiller's yeasts. The cmmn mutagens used with wine yeasts have been UV r X-rays, ethyl methane sulfnate (EMS), N-Methyl-N-nitr-N-ninsguanidine (NTG), N-nitrsurea, r diethylstilbestrl (Tubb and Hammnd, 1987). Since many f the desirable mutatins re recessive, their expressin in diplid r plyplid wine strains has resulted frm hmzygsity brught abut by gene cnversin r mittic crssing ver. Mlzahn (1977) emplyed mutagenesis t successfully islate brewing yeast mutants with increased flcculatin, and with mdified abilities t prduce diacetyl and hydrgen sulfide. Ingraham and Guymn (1960) used ultravilet light t generate isleucine and valine requiring mutants that prduced nly trace amunts f isamyl alchl and isbutyl alchl respectively. Frm EMS treated wine yeasts, Rus ef al. (1983) islated leucine auxtrphic recessive mutants that als prduced reduced cncentratins f higher alchls. Mutagenesis has the ptential t disrupt r eliminate undesirable characteristics and t enhance favurable prperties in wine yeasts. Hwever, the heavy mutagen treatment frequently used culd prduce mutatins in additin t the ne f interest, which is a pssible disadvantage f this methd Hybridizatin Hybridizatin invlves the mating f haplids f ppsite mating-types t yield a heterzygus diplid. Recmbinant prgeny are recvered by sprulating the diplid, recvering individual haplid ascspres and repeating the mating/sprulatin cycle. The fact that wine yeasts are generally hmthallic cmplicates the use f this prcedure.

25 -16- Hwever, this prblem can be vercme by direct spre-cell mating, as mating type switching des nt ccur until the third r furth generatin f grwth after spre germinatin Clakan and Oshima, 1970). Thrntn (1982) used selective hybridizatin f pure culture wine yeasts t significantly imprve fermentatin efficiency (frm 84 t 937). Als, useful killer wine yeast strains have been bred by hybridizatin f a killer sake yeast with mesphilic (Hara er a/., 1980) and cryphitic (Hara et a1.,1981) wine yeass. Hybridizatin prgrams are als hampered by the pr sprulatin and lw spre viability f wine yeast strains. Cnsequently, this breeding methd has nt been widely emplyed in the imprvement f wine yeasts Rare mating Industrial yeast strains which fait t shw a mating-type can be frce-mated with haplid a r str ins. In this prcedure, knwn as rare mating, a large number f cells f the parental strains are mixed tgether and a strng psitive selective pressure is applied t identify the ra e hybrids. Hybrids are usually selected as respiratry-suffrcient prttrphs frm crsses between a respiratry-deficient mutant f the industrial strain and an auxrrphic haplid strain (Gunge and Nakatmi, 1971). Hybrids prduced by rare mating are ften able t sprulate (Spencer and Spencer,1977) prviding anther pssible rute fr the genetic analysis f industrially imprtant cha acteristics. By rare mating,tubb et al. (1981) cnstructed brewing strains with the ability t ferment dextrins. An imprtant develpment has been the use f r rre mating t generate prgeny ('cytductants' r 'heterplasmns') which receive cytplasmic cntributins frm bth parents but retain the nuclear genme f nly ne f them. This frm f strain cnstructin has been termed cytductin. The technique has becme highly effrcient thrugh the use f haplid strains which carry the karl mutatin and are, therefre, defective in nuclear fusin

26 -17 - (Cnde and Fink, 1976). Such strains can be used as dnrs f cytplasmic genetic material t industrial strains. Fr example, transfer f the duble stranded (ds) RNA determinants fr the Kl zymcin (and assciated immunity) has been used t prduce brewing strains with anti-cntaminatin prperties (Yung, 1981,1983; Hammnd and Eckersley, 1984) Spherplast fusin Spherplast fusin prvides anther direct asexual technique fr manipulating industrial yeasts genetically and, like rare mating, can be used t prduce either hybrids r cytductants. The prcedure was first described by van Slingen and van der Platt (1977). Spherplasts are frmed by remval f the cell wall with an apprpriate lytic er^zyme preparatin such as Zymlyase (a glucanase frm Arthrbacter luteus) in a medium cntaining an smtic stabilizer (usuatly 1.0 M srbitl) t prevent cell lysis. Spherplasts frm different strains are fused tgether in the presence f plyetþlene glycl and calcium ins, and then allwed t regenerate their cell walls in an smtically stabilized agar medium. Spherplast fusin can als be btained by electrpratin in a weak inhmgeneus alternating electric field. Fusin f the aligned cells can then be induced by applying a higher-intensity elecric field (Tubb and Hammnd, 1987). Spherplast fusin f nn-sprulating yeast strains serves t remve the natural bariers t hybridisatin. Cells f different species, r levels f plidy, can be fused. The use f spherplast fusin t mdify an industrial yeast is illustrated by the cnstructin f brewing strains able t ferment dextrins (Freeman, 1981; Russell et a1.,1983). Ykmri et at. (1989) prduced cytductants f a sake wine yeast by spherplast fusin that exhibited gd fermentatin perfrmances and prduced quality wine with lw vlatile acids. Spherplast fusin has als been used t create yeast strains with imprved tlerance t ethanl (Seki er a1.,1983).

27 Recmbinant DNA technle.v The varius mating prcedures described abve result in parental strains cntributing majr prtins f their respective genmes. This hybridizatin is essentially empirical and a lengthy selectin prcedure is ften required t btain a suitable strain frm the large number f recmbinant types prduced. Recmbinant DNA technlgy, hwever, prvides the pprtunity t specifically alter single characteristics in wine yeast strains. Furthermre, genetic engineering permits intrductin f genes frm any surce int wine yeasts. The additin f exgenus DNA t yeast and its subsequent incrpratin int the genetic framewrk f the cell, resulting in the acquisitin f a nvel cha acteristic, is termed transfrmatin (described in Sectin 4.1). Transfrmatin has nt yet been used in the imprvement f wine yeast strains but has been successfully applied t brewing yeasts. Fr example, enhanced degradatin f sta ch and sugar utilizatin by brewing yeasts has been achieved by clning starch-degrading enzymes int brewing yeasts (Stewart, 1981; Meaden et a1.,1985). A p-glucanase gene frm Bacil/ s has als been intrduced and expressed in brewing yeasts (Cantwell et a1.,1936). The transfrmed yeasts effectively secrete substantial levels f p-glucanase enzyme which degrades p-glucan in wrt. In additin t the inrductin f specific genes int wine yeasts, recmbinant DNA appraches ffer wider applicability. Fr example, the ptential exists t eliminate specific undesirable strain characteristics by gene inactivatin. Als, it may be pssible t develp new gene prducts with mdified characteristics by site directed mutagenesis. 2.4 TARGETS FOR WINE YEAST STRAIN DEVELOPMENT Althugh the techniques invlved in the genetic manipulatin f industial yeasts are quite well established, the wine industry has been slw t engage in yeast genetics and

28 -19- srrain develpment prgrammes (Thrntn, 1983). This is due in part t the technical difficulties invlved in such prgrammes, and partly t the fact that the requirements f the wine industry have nt been defined in genetic terms. Despite these bstacles, the genetic basis fr several characteristics f enlgical imprtance has been determined and there a e a number f examples f wine yeast strain imprvement. These examples are included belw in a list f specific targets fr yeast genetics in wine-making Sedimentatin and flcculatin The aggregatin f dispersed yeast cells int flcs twards the end f fermentatin is calted flcculatin. Nn-flcculent yeasts settle slwly fllwing fermentatin and frm a fine sediment which is easily disturbed n racking. This prperty may necessitate lnger settling times, centrifugatin r the use f fining agents t clarify the wine with a cnsequent increase in prductin csts. Flcculent yeasts, hwever, frm a heavy sediment and a clear wine which can be easily racked clse t the lees. The nature f the interactins amng flcculent yeast cells is prly understd, and basically tw mdels fr the mechanism exist. Thse based n physicchemical principles prpse cperative bnding between cell surface plysaccharides (Mill, 1964). The bservatin that prtease treatment leads t an irreversible lss f flcculatin supprts this thery (Miki et a1.,1930). Alternatively, flcculatin interactins may be mediated by a specific cell surface recgnitin mechanism, invlving lectin-like binding f surface prteins t plysaccharides n adjacent cells (Iaylr and Ortn, 1978). Genetic studies f yeast flcculatin were first reprted by Gilliland (1951) and Thrne (1951). Since then, a number f genes have been reprted fr the flcculence phentype in Sacclnrmyces species: FLOl, fl3, FLO5, fl6, fl7, FLO8, fsu1, fsu2 and tupl (Jhnstn and Reader, 1983; Yamashita and Fukui, 1983; Lipke and Hull-Pilsbury, 1984). Of these genes, the dminant flcculatin gene, FLOl, which is 37cM distal t

29 -20- adel. n chrmsme 1, has been mst extensively analysed bth genetically and bichemically (Jhnstn and Reader, 1983; Miki er a1.,1982 a,b). The prperty f flcculatin was intrduced int a pwdery wine yeast strain lvd26 by mating spres f MD26 with a haplid labratry yeast strain which ca:ried the dminant FLO1 gene (Thrntn, 1985). Fllwing the hybridizatin step, a series f backcrsses t the riginal wine yeast parent were perfrmed - the utcme f this genetic mdificatin prgram was the cnversin f the pwdery yeast MD26 t a flcculent yeast while retaining its psitive winemaking prperties. Mre recently, mlecula clning f the FLO1 gene (Watari et a1.,1989) has prvided the pprtunity fr specific intrductin f the flcculatin char cteristic int industrial yeast strains by transfrmatin Nn-faming mutants The prductin f frth-head during fermentatin is an undesirable trait f wine yeasts as up t SV f the capacity f the fermentatin vessel may have t be reserved t prevent the frth frm spilling ut. Selectin f nn-faming mutants f the widely used Kykai N. 7 srain f sake yeast was achieved by Ouchi and Akiyama (1971) by either a cell agglutinatin r a frth fltatin methd. The cell aggluúnatin methd is based n the bservatin that a nrmal sake yeast was agglutinated when mixed in acid slutin with certain species f lactbacilli, whereas a nn-faming strain was nt (Mmse et a1.,1968). Ouchi and Akiyama (1971) were able t enrich fr nn-faming mutants by a series f repeated selectin cycles. \ù/ashed cells frm a culture f nrmal Kykai N. 7 were mixed with cells f Inctbacillw plantarurnrn dilute citric acid. After mixing and allwing t settle, an aliqut f the upper part f the suspensin was used as the inculum fr anther rund f yeast grwth. After the ninth selectin step mst f the clnes sampled were nn-faming. When cells f the parental

30 -21- strain \ilere mutaganized by expsure t UV light prir t selectin cycles, 1007 f the sampled clnes at the seventh cycle were nn-famers. Selectin by the frth fltatin methd was described by Akiyama et al. (1971), wh fund that cells frm the nrmal faming strain adhered t the CO2 gas bubbles frmed during fermentatin, while cells f a nn-faming strain did nt. Nn-faming mutants were prgressively enriched in repeated selectin cycles by bubbling air thrugh the culture. Aliquts f the culture, nw enriched fr nn-faming mutants, were used as the inculum fr the next rund. After nine selectin cycles, apprximately 507 f the tested clnes were nn-famers; when the series was started with UV-iradiated cells, abut 807 f.the sampled clnes were nn-famers after seven cycles. Using the frth fltatin technique, Eschenbruch and Rassell (1975) were able t select nn-faming mutants frm tw strains f New Zealandwine yeasts. In studies f the nature f cell walls, Ouchi and Nunkawa (1973) fund that nnfaming mutants have fewer hydrphbic grups n their surface, prbably because f masking by phsphmannan. The genetic basis fr faming has been investigated by Kasahara et at. (1974), wh fund that the nn-faming sake mutatins were recessive. Tetrad analysis revealed that the faming character was under the cntrl f at least tw genes. Using wine yeasts, Thrntn (1978 a,b) als shwed that tw dminant genes (designated FROI and FRO2) cntrl faming, that they are allelic t the sake yeast genes and that they are linked n chrmsme VII, 2l cmfrm ne anther and near ade Sulfite and Sulf,rde Prductin The frmatin f SO2 and H2S by wine yeasts geatly affects the quality f wine. Sulfur dixide is regularly added t disinfect fermentatin equipment, t cnûl rganisms that wuld cmpete with the yeast fermentatin and t prevent excessive xidatin f the wine. Health cncerns have led t effrts t restrict its use as an additive. Hence the

31 - 22- prductin f SO2 itself has becme a matter f sme imprtance. Althugh SO2, when prperly used, has beneficial effects, the same cannt be said f H2S. Frm an enlgical perspective, HzS is ne f the mst undesirable yeast metablites affecting the smell and taste f wines. The frmatin f these tw substances during wine making has been well reviewed by Eschenbruch (1974). Bth a e cmplex prcesses that a e nt yet fully understd. Frm a genetic viewpint, hwever, it is imprtant t nte that wide va iatin has been fund between yeast strains. Mst S. cerevisia strains prduce between 10 a d 30 mglliter f SO2 when tested under cmparable cnditins but sme frm as little as 10 mg while thers frm in excess f ; Eschenbruch and Bnish, 1976b). Strain variatin in H2S prducún has als been revealed (Zambnelli, 1964 a,b; Rankine, 1968; and Eschenbruch et a1.,1978). Studies int the sulfur metablism f high and lw sulfîte-prducing strains have revealed cnsiderable differences in the levels f activity f sulfate petmease (Dtt et al., tg77), ATP-sulphurylase (Heinzel and Truper, 1978) and sulfite reductase (Dtt and Truper, 1976). These differences suggest that it may be pssible t intrduce specific prperties frm lw sulfite prducing yeasts int selected wine yeast strains. Sulfide can be frmed frm sulfate r sulfite, elemental sulfur applied t grapes as a fungicide r frm cysteine (Eschenbruch,l974). Its frmatin can be indirectly influenced by the amunt f yeast grwth, pantthenate r pyridxine deficiencies r excess levels f certain amin acids, metal ins and yeast cell autlysis (Snw, 1983). The reductin f elemental sulfur may ccur bth enrymaticatly r by reactin with thil grups. The variatin and genetic cntrl f H2S prductin has been investigated by Zambnelli Qge a,b; 1965 a,b,c). In screening 100 srains fr H2S prductin frm sulfate r sulfite, he fund fur strains that prduced nne under all cnditins tested; the

32 -23- rest frmed varying amunts, frm traces t ver 200 pg/50m1culture medium. When H2S psitive strains were crssed with negative strains, the hybrid prduced H2S and spres segregated fr H2S prductin at a rati f 2:2. Hwever, since the amunt f H2S prduced by the psitive clnes varied cnsiderably, segregatin f mdifying genes was indicated. In tw later papers, Zambnelli et al. (1975) and Rman et al. (1976) reprted genetic results with prttrphic mutants f sulf,rte reductase. This enzyme is essential fr the bisynthesis f the sulfur amin acids and the prttrphy f the reductase-negative strains was ascribed t leakiness f the reductase mutatin. Frm a survey f several s ains, it was cncluded that a number f factrs influence the prductin f sulfide frm sulfate: inhibitin f sulfite reductase by endgenus factrs, reduced functin f the reductase caused by mutatin, enzymatic blckage after the reductase step causing methinine auxtrphy and the state f genetic heterzygsity f the cell. Given these varius surces f sulfide and the number f influences n its frmatin, ne culd nt expect t find a single gene that wuld eliminate it. Hwever, the fact that strains d vary in their capacity t prduce it suggests that there is cnsiderable nan ral genetic hetergeneity that culd be explited. The specific intrductin f mutatins in certain enzymes f the sulfur, sulfur amin acids and pantthenate and pyridxine pathways may result in reduced prductin f sulfide (Snw, 1983) Ethanl tlerance The inhibitry actin f ethanl prduced in the curse f fermentatin r added externally, is cmplex. A number f parameters have been used as indicatrs f the relative sensitivity r tlerance f yeast strains t the alchl. These include fermentatin rate; bimass yield; g wth rate and cell viability (Oliver, 1987).

33 -24- One f the majr target sites f ethanl in the yeast cell is the plasma membrane, as well as the membrane f the va ius cellula rganelles (Thmas and Rse, 1979). The damage caused by ethanl t the cell membrane results in altered membrane rganizatin and permeability. It has been shwn that ethanl causes the leakage f essential cfactrs and cenzymes frm Zymmnas rnbíiis (Osman and Ingram, 1985). The leakage f these cmpnents, which a e essential fr the activity f enzymes invlved in glyclysis and alchl prductin, \ilas sufficient t explain the inhibitry effect f ethanl n fermentatininz. ntbilß as well as in yeasts (Ingram and Buttke' 1984). There have been many ther mechanisms prpsed fr the inhibitry effects f ethanl. These include the inhibitin and denaturatin f varius intacellular prteins and glyclytic enzymes (Nagdawithana et a1.,1977), inhibitin slute transprt systems (van Uden, 1985), inhibitin f glucse-induced prtn fluxes (Jurszek et ai., 1987), accelerated passive re-entry f prtns in a manner resembling the actin f an uncupler (Lea and van Uden, 1984, Cartwright et al. 1986), derepressin f the ptimum and maximum temperature fr grwth (Sa-Crreia and van Uden, 1983) and the enhancement f thermal death ([æa and van Uden, L982) and petite mutatins in yeast (Cabeca-Silva et a1.,1982). Furthermre, the inhibitry effects f alchls were bserved t increase with increasing carbn number, suggesting that the ptency f alchls is related t lipid slubility (van Uden, 1984). In shrt, ethanl has a cmplex inhibitry actin. This has been reviewed by van Uden (1985). Given the pleitrpic nature f the effect f ethanl n yeast, it is mst unlikely that any single gene will be respnsible fr the sensitivity r tlerance f the rganism t the alchl. A number f mutant nuclear genes have been fund t cnfer an ethanl-sensitive phentype n yeast (Jnes, 1977; Sugden and Oliver, 1983). Hwever, attempts t islate ethanl-tlerant mutants by cnventinal agar plate screening methds, have failed (Ismail and Ali, 1971a,b). This failure is nt surprising given the nature f ethanl txicity - it is likely that the mutatin f a number f genes will be required t imprve the ethanl

34 -25- tlerance tlerance f yeast and, furthennre, such imprvements are likely t be small, since S. cerevisiac is already a highly tlerant rganism. In this situatin, when nly small quantitative increases are likely t be btained as a result f multiple mutatins, the use f cntinuus selectin is preferred. Brwn and Oliver (1932) adpted a system in which the intensity f selectin was determined by the yeast culture itself via a feedback cntrl circuit. Using this system, they successfully islated yeast mutants with increased ethanl tlerance Hieher alchls Higher alchls a e alchls with carbn numbers greatü then that f ethanl, such as isbutyl and isamyl alchl. They are frmed frm either sugar metablism r intermediates in the branched chain amin acids pathway leading t leucine, isleucine and valine by transaminatin, decarbxylatin and reductin (Webb and Ingraham, 1963). Alttrugh they have undesirable flavr and dr characteristics, they are usually present in wines belw the flavr threshld and may, in sme cases, cntribute t wine quality (Kunkee and Amerine, 1970). Hwever, their reductin in wines that are t be distilled (fr example, fr brandy prductin) culd be f cnsiderable imprtance, since they are cncentrated by the distilling prccess (Snw, 1983). Ingraham and Guymn (1960) and Ingraharn et al. (1961) were able t prduce unusually lw levels f isbutanl and is-amyl alchls in fermentatins carried ut with valine, isleucine, and leucine mutants. Hwever, these mutants \ /ere f n cmmercial use as their grwth rate and fermentatin rate were cmprmised. A leu- mutant derived frm the widely used Mntrachet wine yeast (UCD, Enlgy 522) was reprted t prduce mre than 507 less isamyl alchl during fermentatin than the prttrphic parent (Snw, 1933). Taste panel trials indicated n difference between wines prduced with the mutant and the Mntrachet parent strain.

35 Killer factr Killer strains f yeast were first recgnised by Bevan and Makwer (1963). Killer yeasts secrete plypeptide txins which kill sensitive strains f the same genus and less frequently, strains f different genera (Philliskirk and Yung, 797 5; Tipper and Bstian, 1984). The enlgical significance f killer yeasts is still largely speculative. It is cnsidered that such strains culd be used t resrict the grwth f undesirable wild strains f S. cerevisiae and ther clsely related Saccharmyces species during and after alchlic fermentatin. Ouchi and Akiyama (I97 6) intrduced the killer plasmid int sake and wine yeast by crssing them with a wild killer strain islated as a cntaminant frm a sake mash. Repeated backcrssing with a selectin at each generatin fr particular characteristics f the sake strain gave killer hybrids that prduced sake r grape wine f cmparable quality t ttre parent. This backcrssing prgram f Ouich and Akiyama was hampered by pr sprulatin and spre viability. T vercme these prblems, Ouchi et al. (L979) emplyed a dnr f killer character that was deficient in nuclea fusin, mated this with a haplid (derived frm a sake yeast), and selected fr sake strains cntaining cytplasmic elements f bth strains. This strain gave results in trial fermentatins that were better than the parental strain with regard t rapidity f fermentatin, vlatile flavur cmpnents and acidity Mallactic fermentatin Mallactic fermentatin invlves the decarbxylatin f L-malate t L-lactate and CO2 and is caried ut by several species f lactic acid bacteria. These species all belng t ne f three genera: Lactbacillus, Leucnstc r Pedicccus (Snw, 1983). This prcess perfrms three significant rles fr the winemaker: i) reductin in acidity; ii) micrbilgical stability fllwing $wth f the bacteria; and iii) changes in wine flavur

36 -27 - caused by prducts f the bacterial fermentatin (Kunkee and Gswell, 1977). The fermentatin can be brught abut by hlding the wine under cnditins that are favurable fr the grwth f the bacteria already present r by inculatin with the apprpriate bacterial species. An advantage t the winemaker wuld be t have the mallactic fermentatin ccur during r shrtly after the alchlic fermentatin s the wine can be adjusted fr cella strage withut risk f becming spiled. This culd be achieved if the wine yeast were able t carry ut the mallactic fermentatin. T this end, attempts have been made t transfer the genetic infrmatin necessary fr the mallactic fermentatin frm lactic acid bacteria t a wine yeast. Fusins between S. bailli, S. rwii and Schiz. pmbe with S' cerevisíae have been made, but the resultant hybrids had less ability t ferment malate than the parent strains (Subden and Osthsilp, 1987). The clning and expressin f the mallactic gene (L-malate:NAD carbxylase) frm l ctbacillus delbrucc,tii (Williams er al.,1984) andl ucnnstc enas (Lautensach and SuMen, 1984) in Escherichia clí and ins. cerevisi has been reprted. Hwever, in bth cases, the level f cnversin f malate t lactate by the engineered yeast strain was insufficient t be f practical benefit. This lack f success may be due t prblems with expressin f the clned genes, r t the limited ability f the yeast hst t take up the malate. In an attempt t vercme these prblems, curent research is directed twa ds the clning and intrductin f genes encding malate permease and the malic eîzyme frm Schiz. pmbe int S. cerevisiae (Subden and Osthsilp, 1987) Genetic marking As an aid t yeast management, particularly fr wineries using mre than ne yeast strain, the genmes f cmmercial wine yeasts can be tagged. A marking system assists in mnitring yeast strains used in fermentatins and discurages the illegal use f cmmercial

37 -28- wine yeast smins. A deliberately marked enlgical strain was develped by Vezinhet and clleagues (Vezinhet and Lacrix, 1984; Vezinhet, 1985) by selecting fr natural mutants in a ppulatin f the Lalvin V yeast. The strain, which is nw cmmercialized as Kl, is duble ma ked with tw antibitic markers, diurn and erythrmycin. An extensive survey f yeasts fr resistance t these antibitics demnstrated that few strains are naturally resistant t bth drugs simultaneusly. 2.5 CONCLUSIONS Limited success has been achieved t date in the genetic imprvement f wine yeast strains. Fr example, Thrntn (1983) was able t intrduce the flcculatin character frm a labratry strain int a wine yeast by hybridizatin and a series f subsequent backcrsses t the wine yeast parent. Als, a leucine auxtrph derived frm a widely used Mntrachet wine yeast was reprted t prduce at least SOVI less isamyl alchl during fermentatin than the prttrphic parent (Snw, 1983). Seki el c/. (1985) were able t cnstruct a killer wine yeast by spherplast fusin and shwed that the grwth f sensitive cells in grape juice was inhibited by the killer fusant. Success was achieved in the selectin f a genetically marked wine yeast strain (Vezinhet, 1985), which has prvided an insight int the kinetics f yeast ppulatins during fermentatin (Delteil and Aizac, 1988). Mst f the attempts t develp imprved wine yeast strains have relied n traditinal genetic techniques such as mutatin and selectin, hybridizatin, rare mating and spherplast fusin. These techniques are prblematic because they invlve unspecific alteratins r exchanges f genetic material in yeast srains. These prblems can be avided by the use f recminant DNA technlgy in yeast imprvement prgrams. Recmbinant DNA technlgy has nt yet been emplyed in the prductin f an imprved wine yeast strain. Hwever, classical genetic studies have prvided backgrund infrmatin which culd be used in genetic engineering prgrams. Fr example, the dminant flcculatin gene, FLOl, which has been extensively analysed bth genetically and bichemically

38 -29- (Jhnstn and Reader, 1983; Miki er al.,1982 a,b), culd be intrduced int wine yeasts t prduce flcculent strains. An alternative t the use f hybridizatin and cytductin t intrduce the killer cha acter int wine yeasts wuld be t clne the txin and immunity genes int wine yeast strains. Bth the txin and immunity genes reside n the same M- dsrna mlecule, and reverse transcriptin has already been used t prduce a cdna mlecule f these tw genes (Bstian et a1.,1984). Genetic engineering culd als be used t eliminate r reduce undesirable characteristics by gene disruptin. Snw (1983) suggested that the deliberate intrductin f mutatins in certain enzymes f the sulfur, sulfur amin acids, pantthenate and pyridxine pathways might enable stepwise eliminatin f these characteristics and hence a reductin in sulfide prductin in wine yeasts. Integrative disruptin f specific ILE, LEU amd VAL genes f wine yeasts may result in lwer cncentratins f higher alchls. Recmbinant DNA techniques culd als be used t eliminate the faming charateristics f wine yeast stains by specific disruptin f the FROI and FRO2 genes n chrmsme VII. The extent t which recmbinant DNA technlgy can influence the breeding f imprved wine yeast stains will largely depend upn the requirements f the wine industry and advances in the field f wine bichemisury. The true ptential f this technlgy will becme apparent nce the research is under way.

39 -30- Chapter 3 Materials and Methds 3,1 STRAINS AND MEDIA The Escherichia cli strain used fr bacterial transfrmatins was DH5a [F-, endal, hsdllt, supe44, thi-l, )u-, rec{l, gyra96, relal, L (argf-laczya) U169, ø$}tacz/lmrl5l. Haplid yeast strain Oll (MAT- a, his 3-11, 3-15, leu 2-3,2-112, ura 3-251,3-373) was btained frm Dr. H.B. Lukins, Department f Bichemistry, Mnash University. V/ine yeast strains used in this study were btained frm The Australian Wine Resea ch Institute Cllectin and a e listed in Table 3.1. Bacterial grwth media was LB Íl% bact-tryptne (Difc), 0.5V yeast extract (Difc), 17 NaCll. Ampicillin (100 pg/ml) was added t mlten media at 50C. Yeast grwrh media was YPD llv yeastextr rct (Difc), 2%bacttpeptne (Difc),Z% glucsel, SD I0.67V Bact yeast nitrgen base withut amin glucsel, r YEPG fl% yeast extract (Difc), 27 ZV glycerl). Chlramphenicl (disslved in ethanl) and cyclheximide were added t mlten media (YEPG and YPD respectively) just prir t puring plates. Sulfmeturn methyl (SM) (btained frm DuPnt denemurs and C.) was disslved in acetne and added t mlten SD medium just prir t puring plates. As SM is light sensitive, plates cntaining SM were incubated in the dark. 3.2 YEAST TRANSFORMATION METHODS Alkali catin transfrmatin Alkali catin transfrmatin f yeast was perfrmed accrding t the methd f It et al. (1983) with slight mdificatin. Yeast cells were grwn t late lgarithmic phase (OD

40 Table 3.1. Wine yeasts strains btained frm The Australian Wine Research Institute Culmre Cllectin. Yeast Strain AWRI IA (3s0) AWRI2A (729\ A\ryRI3A 096',) AWRI5A (138) AWRI6A (348) AWRITA (833) AWRI9A (81) Surce Thmas Ha dy & Sn, Suth Australia Lindeman's winery, Suth Ausralia R. Eschenbruch (1975) Rsewrthy Agricultural Cllese (1945) Pasteur Institute, Tunis (19s0) Penfld's Winery, Ba ssa Valley (1979) N.M. Bretez, Victria Descriptin Gd flcculatin prperties and prduces n HzS. The yeast prduces a significant level f higher alchls and esters which give rise t a flral character. Tendancy t accumulate acetic acid when fermenting sme musts, particularly at lw temperatures and in the presence f high sugar cncentratin. Very active yeast with a fast rate f sugíu attenuatin at lw cell numbers. Lw frmatin f higher alchls and esters. Used fr red and white wine Originally islated in Suth Africa, used fr preparing red and white wine. It is a lw SOz and and has the killer A flcculent yeast prducing esters which gives a fruit arma t wine which decreases n Generall used fr it has a lw tlerance t Reprted t prduce lw levels f higher alchls. Limited use fr the prductin f frtified wines and tlerates fermentatin Islated frm a stafier culture in 1979 where it exhibited rapid fermentatin at lw temperature. l,abratry fermentatins shwed lw vlatile acidity and SOz frmatin. It has been examined cmmercially n a small scale with acceptable results. N infrmatin available...cnt'd

41 AWRI SA (834) AWRI loa (83s) AWRI lla (R2) Unknwn Dept. f Agriculture, Vy'estern Australia (JnD Petaluma winery, Suth Australia Selected as suitable fr secndary fermentatin in sparkling wine prductin fllwing cmparisn with several ther yeasts in small-scale cmmercial triats. A killer (KÐ yeast. Simila prperties t strain A\ryRI 2A (729), except that it accumulates less acetic acid. It has killer Frmerly Saccharmyces byantn It is a vigrus fennenter when prpagated crrectly. P duces a highly armatic buquet and is mst suited t white wine prductin. It has killer AWRI l2a (143) Te Kawata Frmerly Saccharmyces rseü. Demnsnated tlerance fr high baume musts and prduces a lw level f acetic acid under thse cnditins. It ferments relatively slwly and prduces neutral arma cmpnents. This yerist can ferment t l5v alchl.

42 ) ar 28C in liquid YPD. Cells were han ested, washed nce in 10 ml TE buffer (lgml,{ Tris, ph7.5, LmM EDTA), suspended ín 20 ml LiOAcÆE buffer (10mM Tris- HCl, ph7.5, 0.lM LiOac, lmm EDTA) and mixed gently by shaking at 30C fr t hur. Apprximately 8 x 107 cells were ha vested and suspended in 100 pl f LiOAcÆE buffer. Plasmid DNA (10pg) [and smetimes ca rier DNA (salmn spenn, Sigma) was added t 5 pgl was added in a ttal vlume <10 pl and the suspensin was incubated at 28C fr 30 mins. Seven times the vlume f filter -sterilized PEG reagent (40V plyethylene glycl 4000, 0.1M LiOAc, 10mM Tris-HCl, ph7.5, lmm EDTA) was added and the mixture was vrtexed befre incubatin at 28rc fr I hur. Cells were then heat shcked by incubatin at 42C fr 5 mins, ha vested, and resuspended in I ml sterile water. Fr the selectin f auxtrphic ma kers in the transfrmatin f strain Ol1, 100 pl f cell suspensin was spread n SD media cntaining 50 pg f amin acid per ml (Lhistidine and L-leucine fr plasmid pcy2-4-10; L-histidine and uracil fr plasmids paw219 and prim-c3). Selectin f the CAT gene n plasmid paw119 was perfrmed by spreading 100 pl f the cell suspensin n YEPG media cntaining 5 mg/ml chlramphenicl. The RIM-C gene f plasmid prm-c3 was selected by plating 1@ pl f the cell suspensin n YPD cntaining 2þglmlf cyclheximide. In rder t select fr the SMRL gene n plasmids pcp and pwx509, 100 pl aliquts f the cell suspensin was plated n slid SD media cntaining 10 pg/ml sulfmeturn methyl (and 50 tglml f L-histidine, L-leucine and uracil when selecting Ol1 transfrmants) Sherplast fusin methd The sherplast fusin methd was perfrmed essentially as described in Burgers and Percival (1987). Cells were grwn vernight with vigrus aeratin in 50 ml YPD t apprximately 3 x 107 cells/ml. After harvesting, cells were washed successively with 20 ml f sterile water and 20 ml f 1.2M srbitl, fllwed by 5 min spins. They were resuspended in 20 ml SCEM (1.2M srbitl, 0.1M sdium citrate, ph 5.8, 10mM EDTA,

43 -32-30mÀ/t p-mercaptethanl), I mg/ml zymlyase (Seikagaku Kgu C. LTD) was added, and the cells were incubated at 30C with ccasinal inversins. Spherplasting was mnitred by measuring the decrease in turbidity at 800 nm f a lo-fld dilutin f spherplasts in water. 'When spherplasting had prceeded t 90V (abut mins) the spherplasts were harvested by centrifugatin at 3009 fr 5 mins. They were gently resuspended in 20 ml f 1.2M srbitl and pelleted fr 5 mins at The spherplasts were gently resuspended in 20 ml f STC (1.2M srbitl, 10mM Tris-HCl, ph 7.5, lomm CaClz) and pelleted again fr 5 mins. This pellet was resuspended in 2 ml f STC. Aliquts (100 pl) were mixed with plasmid DNA (10 pg) land smetimes carrier DNA (salmn spenn, Sigma) was added t 5 rg I in a ttal vlume <10 pl in a 10ml plastic tube. After 10 minutes at rm temperature, I ml f PEG (10mM Tris-HCl, ph 7.5, 10mM CaC!2,20V plyethylene glycl 8000; filær særilized) was added and mixed gently. After a further 10 min rm temperature incubatin spherplasts were harvested by a 4 min centrifugatin. The pellet was gently resuspended in 150 rl f SOS (1.2M srbitl, 6.5mM CaCI2,0.257 yeast extract,o.sv bactpeptne; filter sterilized) and left at 30C fr mins. Eight milliliters f TOP (1.2M srbitl, 2.57 agar in SD medium) kept at 45C was added. The tube was inverted quickly several times t mix and plated immediately n SORB plates [SD plates cntaining 0.9M srbitl, 37 glucse and sulfmeturn methyl (10pdml) Electrpratin f intact yeast cells Transfrmatin f yeast cells by electrpratin was perfrmed accrding t the methd f Hashimt et ai., Yeast cells were grwn vernight in 100 ml YPD at 28C. Lgarithmic phase cells were harvested, washed in sterile distilled water and resuspended in 2 ml f the water. A 50 rl aliqut f this cell suspensin was transferred t a 1.5 ml Eppendrf tube. Plasmid DNA (10 ttg) and 60 tl f TOV PEG4000 were added and mixed thrughly by vrtexing. After standing fr t hur at rm temperature, 40 pl f the cell suspensin was placed in a parallel-electrde chamber. This electrfusin

44 -33- chamber was equipped with tw platinum plate electrdes 2 mm apart. Electric field pulses were applied by an electric capacitr discharge methd using a BiRad Gene Pulseril unit. Th ee successive pulses f an initial intensity f 5 KV/cm were applied with a capacitance f I rf. After applicatin f the pulses, the cell-dna mixture was left t stand fr I hur. The cell suspensin was then transferred t an Eppendrf tube. One milliliter f sterile water was added t the tube and cells were sedimented by centrifugatin. The sedimented cells were resuspended in 200 pl sterile water, and 100 pl aliquts were plated nt SD media cntaining 10 pglml sulfmenun methyl. 3.3 RECOMBTNANT DNA TECHNIQIJES Resrictin digests. fraement islatin and ligatin Restrictin enzymes were btained frm Behringer Mannheim and digests were perfrmed in buffers supplied by the manufacturer. All digests were perfrmed fr l-2 hurs at 37C. DNA fragments were islated with a GENE CLEAN (BIO 101) kit fllwing supplier's instructins. Dephsphrylatin f fragment ends was achieved by incubatin f DNA with 1 unit f calf-intestinal alkaline phsphatase (Behringer- Mannheim) fr 30 minutes Lt37C in buffer as recnìmended by supplier. DNA ligatins were carried ut using T4 DNA ligase (Behringer-Mannheim) in the recmmended buffer Electrhresis f DNA Agarse gel elctrphresis (described in Maniatis et al., t982) was used fr the separatin f DNA mlecules up t 20 kbp in size. Samples were electrphresed in I.OV agarse gels in TAE buffer (40mI4 Tris-acetate, ph 8.0, lmm EDTA) at a cnstant curent f apprximately 5OmA. One part f gel-lading buffer (0.257 brmphenl blue, xylene cyanl, l1%ficll [type 400]) was added t five parts f DNA sample prir t gel lading. After electrphresis, gels were stained with ethidium brmide and

45 -34- visualized na254nm UV light bx. DNA fragments required fr clning were bserved n a 360 nm UV lighrbx Bacterial transfrmatin Transfrmatin f the E. cli strain DH5cr was perfrmed accrding t the methd f Hanahan (1983) Plasmidislatin (1e81). Rapid small-scale plasmid islatin was as described by Ish-Hrwicz and Burke Yeast DNA islatin Ttal DNA was islated by a prcedure based n that described by Davis et al. (1980). The yeast culture was grwn in 10ml YPD t late lg phase. Cells were pelleted by centrifugatin at fr 5 mins, washed nce in TE buffer and resuspended in 0.8 ml f 1.2M srbitl, 0.1M EDTA, pín.4, and 14mM p-mercaptethanl. Zymlyase (Seikagaku Kgy C. LTD) was added t a ttal cncentratin f 1 mg/ml and the suspensin was incubated fr 20 mins at 30C. Spherplasts were pelleted by centrifugatin at fr 5 mins, then resuspended in 0.8 ml TE buffer. T this suspensin 80 pl f 1 M Tris-HCl, ph 7.4,80 pl f 0.5 M EDTA, and 40 rl f a 1'0V SDS slutin were added. After mixing, the suspensin was incubated at 65C fr 30 mins. A 4M ptassium acetate slutin (0.25 ml) was then added and the suspensin incubated n ice fr t hur. After centrifugatin at fr 25 mins, the supernatant was carefully decanted int a fresh tube. An equal vlume f 99.5V ethanl was added t this supernant, and after mixing the slutin was centrifuged at fr 10 mins. The pellet was then dried and resuspended in 500 pt TE buffer, and RNase A (Behringer

46 -35- Mannheim) was added t a final cncentratin f 100pg/ml. The suspensin was incubated at37c fr 30 mins, then extracted twice with and equal vlume f phenvchlrfrm. The salt cncentratin f the aqueus phase was adjusted t 0.3M by the additin f 3M sdium acetate, and 2 vlumes f cld 99.5V ethanl were added. After mixing the slutin was placed at -20C fr 2 hurs r lnger. DNA was pelleted by centrifugatin at fr 10 mins. The pellet was washed twice in7sv ethanl, dried and resuspended in pl TE buffer Suthern h)rbridizatin Suthern bltting and hybridizatins were ca ied ut by the prcedure f Suthern (1975) with minr mdificatins. The membrane used was Hybnd N+ (Amersham), and DNA fixatin was achieved by expsing the membrane t UV light fr 5 mins. Hybridizatins were perfrmed ú 42C in slutins cnsistin g f. 4V plyethylene glycl 4000,2 x SSPE, l7 SDS,507 frmamide,o.s% bltt and carrierdna (0.5 mg/ml final cncentratin). DNA prbes were prepared with an liglabeling kit frm Amersham. Autradigraphy was perfrmed by expsure f the membrane t X-ray film (Fuji RX) between tw intensifying screens vernight at -80C. 3.4 PROTEIN SYNTFIESIS ANALYSIS Incrpratin f radiactivitv int ]reast prteins A sample f yeast cells remved frm the ferment (500 pl) was incubated with 3H- Leucine (5 pci) fr 2 hurs at 30C. Cetls were then pelleted by centrifugatin in a micrfuge and resuspended in a2% SDS slutin (500 pl). Glass beads were added t U2 vlume and the suspensin was vrtexed fr 20 minutes. At five minute intervals the suspensin was placed n ice t avid ver heating. The slutin was then centrifuged fr 5 mins in a micrfuge, the supernatant was remved and placed in a fresh tube. Prteins

47 -36- were precipitated by the additin f 5 vlumes f cld acetne and incubatin at -20pCfr2 hurs r lnger Electrphresis f prteins Prtein prfiles were btained by SDS-pylacrylamide gel electrphresis n l07 plyacrylamide separating gels accrding t the prcedure f Laemmli (1970). Flurgraphy f the gel was achieved by saking the gel in glacial acetic acid fr 5 mins, and then in a slutin f 2OV PPO (2,5 diphenylxazle) in glacial acetic acid fr I hur. The gel was then rinsed in water fr t hur and dried under vaccuum. Autradigraphy was perfrmed by expsing the dried gel t X-ray film (Fuji RX) between tw intensifying screens at -80C fr 4-6 weeks. 3.5 PULSED FIELD GEL ELECTROPHORESIS Transverse alternating field electrphresis (TAFE) was carried ut using a unit. Yeast chrmsmes were prepared in agarse plugs essentially accrding t supplier's recmmendatins. Yeast cells were glwn t early statinary phase in YPD media and apprximately I x 10e cells were ha vested by centrifugatin at fr 5 mins. T minimize prblems with degradatin f DNA, cells were prcessed as quickly as pssible after harvesting. Cells were washed twice in ET buffer (50 mm EDTA, 1 mm Tris-HCl, ph 8.0) and resuspended in 2 ml ET buffer. Tw milliliters f lw melting pint agarse mixture (IV lw melting pint agarse (Adelab Scientific) in 0.lM EDTA, 10 mm Tris, ph 8.0, 20 mm NaCl) at 50C was added, alng with 0.2 ml f a zymlyase slutin. The suspensin was mixed by vrtexing and immediately laded int a beckman plug mld by pasteur pipette. The plugs were slidifred by refrigeratin fr 30 mins, then carefully pushed frm the mld int 10 ml f 10X ET buffer. An additinal 0.15 ml zymlyase slutin (10 mg/ml) was added, and the plugs we e incubated fr 48 hurs at 37C with ne change f buffer. Plugs were then transfened t 10 ml f 10X ET

48 -37 - buffer cntaining 17 sdium lauryl sa csinate and 1 mg/ml Prteinase K. The plugs were then incubated at 50C fr 48 hurs, with ne change f buffer and enzyme replacement. The resulting plugs were then washed 4 times in ET buffer, and stred in ET buffer at 4C. Chrmsmal DNA mlecules were separated by elecrphresis in a L% l-w endsmsis (LE) (Adel-ab Scientific) agarse gel in TAFE buffer (l0ml\'l Tris,4.35mM acetic acid, 0.5mM EDTA). Electrphresis was carried ut at l2c with a cnstant crurent f 150 ma; pulse times were 60 secnds fr 18 hurs, ttren 35 secnds fr 6 hurs. 3.6 p-glucuronidase (GUS) ASSAYS Enzvme activitv assavs Yeast cells were grwn t lg phase, then 1 ml f the culture was pelleted by centrifugatin, and the cells resuspended in 1 ml GUS extractin buffer (50mM Na2HPO4, ph7.0, 1OmM beta-mercaptethanl, 10mM Na2EDTA,O.I7 sarcsyl, 0.17 tritn X- 100). Glass beads (Sigma, micrns) were then added t apprximately half the vlume, and the suspensin was vrtexed fr 10 mins at 1000 r.p.m. After centrifugatin, the supernatant was remved and used as the cell extract pl extract was added t 0.5 ml assay buffer (lmm MUG in extractin buffer) and incubated at 37C. At varius time intervals, 100 pl samples were remved frm the assay mix and added t 900p1 stp buffer (0.2M Na2CO3). Slutins were then assayed in a spectrphtflurmeter with xenn lamp (Aminc SPF-125 TM), excitatin 365 nm, emissin 455 nm. Standard slutins f 4-methylumbelliferne (I,fÐ (Sigma) in the range f 100 nm t I ttm were prepared fr reference values. lpm MU crrespnded t 100 relative units. Prtein cntent f samples was determined using a bicinchninic acid prtein assay kit (Sigma C.).

49 Agarplate assa],s Yeast clnies were grrvn n slid YPD media cntaining pglml X- GLUC. After apprximately 36 hurs grwth, a slutin cntaining 0.1M Na2HPO4, ph7.0, 17 sarcsll, 50 pglml X-GLUC and O.7V agarse was pured as a thin verlay n the plaæ and allwed t set. After 4-5 hun incubatin at37c a blue precipitate culd be detected in the transfrmed clnies Micrscic visualizatin Yeast cells were grwn vernight in YPD media, harvested and resuspended in a slutin f 0.1M Na2HPO4, ph 7.0, 17 sarksyl and pgml X-GLUC. This suspensin was incubated at 37C fr 2-6 hurs r until a deep blue clur had frmed. Aliquts f the suspensin were then visualized under a micrscpe. 3.7 KILLER YEAST MANIPULATIONS Curins f killer strain 3AM A culture f strain 3AM was grwn vernight in YPD at 28C. Serial dilutins were made in O.9V NaCl and 0.1 ml aliquts (cntaining apprximately 100 cells) were spread n YPD plates and incubated at 37C. After 48 hurs incubatin, single clnies were selected at randm and assayed fr killer activity as described belw Assa]r fr cured strain YPD (cntaining l7 agar) was autclaved at 120C fr 20 mins. After cling t 49C, the medium was buffered t ph 4.2withal07 ta trate slutin. Methylene blue (t

50 -39- g.ffi3v 11l7v) and killer sensitive strain 5A (t lf cells per ml) were added t the medium prir t puring the plates. Clnies islated after heat treatment were then patched nt these assay plates and incubated at 18C fr apprximately 72 hurs. Curing was recgnised by the absence f grwth inhibitin (clear znes) and lack f blue stained cells arund the patched clny dsrna islatin The dsrna extractin prcedure was essentially that described by Fried and Fink (8). Samples f RNA were analysed by electrphresis n L.SV agarse slab gels at a cnstant current f 100 ma. Gels werc stained with ethidium brmide and phtgraphed n a shrt wave UV light bx. 3.8 MCROVINIFICATIONTRIALS 3.8. I Fermentatin prcedure Starter cultures were prepared by inculating 10 ml f YPD medium cntained in a cnical flask with a lpful f yeast and incubated with vigrus aeratin at 28C. After reaching statinary phase, cells were subcultured int YPD and incubated with vigrus aeratin at 28C. After 24 hun, the cell density was determined by micrscpic cunts. Samples were used t inculate Riesling must (200 ml) t a density f 5 x 1S cells per ml. The must cntained 220 gper litre ttal sugü and had a ph f 3.1. Fermentatins were carried ut in 250 ml cnical flasks fitted with airlcks. The juice was sterilized by membrane filtratin (0.a5 pm) prir t inculatin and fermentatins were carried ut at 18C with agitatin (apprximately 100.p.m). Samples were remved anaerbically and aseptically during fermentatin by needle and syringe thrugh prts cvered with rubber sepra. They were analysed fr the prgress f fermentatin by refractmeter readings and yeast grwth was measured spectrphtmetrically at 650nm.

51 Wine anah sis Residual sugars and acetic acid cncentratins were determined with apprpriate kits frm Behringer-Mannheim. Measurement f sulfur dixide in wine was as accrding t Rankine and Pcck (1972). The alchl cntent was determined by near infra-red reflectance spectrscpy accrding t 'Instnrctins fr the Use f the Technicn260Infraanalyser', Bran & Luebbe, Australia. The instrument was calibrated accrding t Sneyd er a1., I.A,RGE SCALE LABORATORY FERMENTATIONS Grape juice was btained frm St. Hallet's winery, Ba ssa Valley, during the 1991 harvest. PedrXímines grapes were machine harvested and crushed n the same mming. Immediately after crushing, grape must was pumped int the winery press. Samples f twenty litres were cllected in plastic cntainers (25 litre capacity) which had been fitted with airlcks. The grape must samples were then transprted t The Australian Wine Research Institute fr fermentatin triats. Diammnium rthphsphate (DAP) was added t all samples t 0.3gll-. Sdium metabisulhte (150 mg/l) additins were then made t sme samples t achieve a ttal sulfur dixide cncentratin f 100 mg/l. Grape must was then allwed t stand fr 4 hurs prir t inculatin. Sta ter cultures were prepared by inculating 100 ml f YPD medium cntained in a cnical flask with a lpful f yeast and incubated with vigrus aeratin vernight at 28C. This culture was used t inculate 2L f Rhine Riesling grape juice which was incubated fr 24 hurs at 28C with vigrus aeratin. The cell density f this starter culture was determined by micrscpic cunts, and grape must samples were inculated t a density f 4 x 106 cells/ml.

52 -41- Samples (10 ml) were taken frm the centre f the vessel at regular times thrughut the fermentatin. Prgress f fermentatin was mnitred by refractmeter readings and yeast grwth was measured spectrphtmetrically at 650nm. Duplicate aliquts f each sample (0.1 ml) were plated nt slid YPD media and incubated fr 48 hurs at 28C. Clnies were then assayed fr GUS activity by the methd described in Sectin

53 - 42- Chapter 4 Establishing a system fr the genetic manipulatin f wine Yeasts 4.1 INTRODUCTION The cmpnents f a transfrmatin system which need t be cnsidered include the selectin prcedure fr islating transfrmants, the type f vectr used t intrduce the freign DNA, and cnditins fr cellula uptake f DNA selectin systems fr identif) ing transfrmants The infductin f recmbinant DNA int yeast strains requires sme way f detecting the very small prprtin f cells (the transfrmants) which actually receive and express the freign DNA. Yeast strains were first transfrmed with vectrs which cntained auxtrphic markers such as LE(12 fr selectin (Hinnen et al., 1978; Beggs, 1978). Auxtrphic markers are limited in that they require a yeast hst strain that is mutated in the crrespnding marker gene, and therefre usually has t be a labratry bred haplid strain. Fr example, a leu2 mutant (deficient in the synthesis f leucine) is required as the recipient strain when the wild type LEU2 gene is included as the selectable marker. Transfrmants can then be selected fr their ability t grw withut leucine. Recessive auxtrphic mutatins can nt be readily selected in amphiplid r plyplid industrial strains, nr are they likely t allw nrmal prductin yields. Therefre 'dminant' ma kers are required that enable transfrmants f wild-type wine yeast snains t be selected withut prir genetic manipulatin. A number f dminant

54 Table 4.1. Dminant selectable ma ken used in yeast transfrmatins Gene Surce Prduct Selectin References adt bacterial (Tn903, plasmid Tn5) kanamycin phsphtransferase G418 (geneticin) resistance Jiminez and Davies,1980; Webster and Dicksn, 1983; Hadfield et al, ah bacterial plasmid (pck203) hygrmycin B phsphtransferase hygrmycin B resistance Gritz and Davies, 1983; Kaster et al, 1984 at bacterial plasmid (tn 9) chlramphenicl acetyltansferase chlramphenicl resistance Chen et al., 1980 Hadfreld et al, 1986 dhfr (cdna) muse dihydrfl ate reductase methtrexate resistance Miyajima et al., 1984 Zhuetal,1985 CUP1 yeast cpper chelatin cpper resistance Fgel and Welch,1982 Butt et al., 1984 Hendersn et al, 1985; Hinchliffe and Daubney, 1986 MGR yeast nt defined mettrylglyxal resistance Murata et al, 1985 Kil-K1 (cdna) yeast killer txin resistance t killer txin Bussey and Meaden, 1985 DEX1 yeast amylglucsidase grwth n dextrin Meaden et al, 1985 ble nt defined rcsrstance t Gastignl et. al, 1987 phlemycin bacterial plasmid (In 5) cnt'd

55 CRY 1 yeast ribsmal prtein 59 resistance t cryptpleurine Larkin and Wlfrd, 1983 TunR CmpR yeast yeast UDP-N-acetylglucs-amine resistance t tunicamycin l-p-tansferase HMG-CA redustase resistance t cmpactin Rine et al., 1983 Rine et al., 1983 tcrnl yeast ribsmal prtein L3 resistance t Schultz and Friesen, 1983 tichdermin RM-c yeast nt defined resistance t cyclheximide Takagi et al, 1986 SMR1 yeast acetlactate synthase resistance t sulfmeturn methyl Casey et al, 1988 Falc, 1986 Yadav et al., 1986

56 -43- selectable markers have been reprted fr use with industrial yeast strains and are listed in Table 4.1. Genes fr resistance t chlramphenicl (chlramphenicl acetyltransferase) G418/geneticin (aminglycsidase phsph-transferase), hygrmycin B (hyg mycin B phsphtransferase) and scissin) all riginate frm genes calried by E. cli transpsns. The bacterial cntrl sequences assciated with genes invlved in chlramphenicl and hygrmycin B resistances perate inefficiently in yeast. All rranspsn genes [chlramphenicl acetlyltransferase (Hadfield et â1., 1986); aminglycsidase phsphtransferase (Ycum, 1986; Hadfield et a1.,1990); hygrmycin B phphtransferase (Gritz and Davies, 1983); the ble gene cded resistance t the DNA scissin agenr phlemycin (Gastign l et al., 1987) were made t functin efhciently as selectable markers by replacement f bacterial DNA with yeast 5' prmter and 3' terminatin signals. Tw f the three bacterial genes, chlramphenicl acetyltransferase (Hadfîeld et al., 1936) and aminglycsidase phphtransferase (Sakai and Yamamt, 1986; Ycum, 1986), were shwn t functin as selectable markers in industrial strains f Sacclnrmyces. Resistance t methtrexate was accmplished by mdif,rcatin f the muse dihydrflate reductase gene (dhlr) fr expressin in yeast (Miyajima et a1.,1984; Zhu et /., 1986); a disadvantage f this resistance ma ker is that multiple cpies are required fr selectin. Experiments perfrmed by Fgel and Welch (1982) defined the CUPI site n chrmsme VIII as the cding sequence fr methallthinein; hwever, resistance t cpper required the presence f up t 15 tandem nuclear cpies f CUP I. Meaden and Tubb (1985) usedcupl as a selectable ma ker n a multicpy plasmid in the cnstructin f a dextrin fermenting strain f brewing yeast. Multicpy plasmids have als been utilized in prcedures aimed t identify enzymes inactivated by inhibitrs. By using minimal grwth limiting cncentratins f tunicamycin,

57 -44- cmpactin and ethine in a haplid labratry strain, Rine er ai. (1983) were able t islate frm a YEp library wild type genes cding fr three different enzymes inhibited by these cmpunds. Multicpy plasmids have als been used t islate wild type sensitive and recessive resistance alleles f three ribsmal prteins. The recessive alleles tcml (Fned and Warner, 1981), cyh2 (Fn $and Warner,1982) and cryl (La kin and Wlfrd' 1983) result in resistance t tricdermin, cyclheximide and cryptpleurine respectively. Althugh nne f the abve six genes were used in industrial strains, the methdlgy used was shwn t be applicable t the selectin f methylglyxal resistance that was subsequently expressed in bakers, brewery and sake yeasts (Murata et a1.,1985)' Glucamylase (amylglucsidase) genes have been clned frm S ccharmyces diastaticus: STAI and STA3 (Yamasita and Fukui, 1983X and DEX1 (Meaden et al., 1985). Meaden et ai. (1985) reprted thatde)u was used as a selectable ma ker in haplid labratry strains by allwing plasmid treated prtplasts t regenerate n glucse-srbitl medium prir t selectin n dextrin medium. Hwever, this prtcl was nt successful with brewer's yeast (Tubb, 1987), and there have been n reprts f successful transfrmant selectin in industrial yeasts using the STAI r S?"Ai genes. Bussey and Meaden (1985) develped a transfrmatin prtcl fr expressin f killer K1-cDNA based n medium cntaining killer txin. Use f the alkali catin transfrmatin prcedure (It et a1.,1983), 7-9 hurs f preincubatin prir t selectin, and a final screening fr the killer phentype resulted in a transfrmatin efficiency f that bserved fr selectin f a prttrphic marker. Tw industrial yeasts transfrmed by this prcedure (Bussey and Meaden, 1985) were shwn t have stable inheritance f the killer and immunity phentype. Tw f the mst prmising dminant markers described t date are the RIM-C gene encding resistance t cyclheximide (Takagi et a1.,1986) and SMRI which cnfers resistance t the herbicide sulfmeturn methyl (Falc and Dumas, 1985; Casey et al.,

58 ). Advantages f these tw markers are: i) the genes invlved are derived frm yeasts; ii) bth markers result in transfrmatin efficiencies simila t thse btained with prttrphic markers; and iii) bth genes are functinal selective markers when present as single cpies. The RIM-C gene was clned frm Candida maltsa (Takagi et al., 1986) and functins in such a way as t mdify ribsmes s that prtein synthesis in the cells is n lnger inhibited by cyclheximide. The exact mechanism fr this mdificatin is nt understd. The SMRI gene f S. cerevisiae prvides resistance t sulfmeturn methyl (SM) (N-t(4,6 dimethylpyrimidin-2-yl) amincarbnyll-2-methxycarbnyl-benzenesulfnamide). The target site f SM in S. cerevisí e is the enzyme acetlactate synthase, which is invlved in the bisynthetic pathway f isleucine and valine, and is encded by the ILV2 gene (Falc and Dumas, 1985; Falc et al., 1985). Sequence analysis has revealed thatilv2 and SMRI are identical alleles, except fr a C t T transitin mutatin at nucletide 574 f the pen reading frame (Falc et a1.,1985; Yadav et a1.,1986). This mutatin results in a prline t serine change in amin acid sequence f acetlactate synthase and cnfers resistance t inhibitin by SM Gene transfer vectrs Freign genes t be expressed in yeast hst cells need t becme incrprated int a transprt vehicle, r vectr. The fur different types f yeast vectrs available are distinguished by their interactin with recipient cells: they are either integrating vectrs, designated YIp, replicating vectrs (YRp), epismal vectrs (YEp) r artificial chrmsmes (YAC). Figure 4.1 illustrates the different classes f vectrs used in yeast transfrmatins.

59 Designattn Structr e Mde f transfrrnatin -E aúf seþctable ma ke Yea-gt ilægrating 9 Ctr Yeast prtrphic selecuble marker E l üi ORI - lw frequncy - inægratiue - \ ery stable - ne cpy per cell Z tn E att' selectable ma ker Yeast episrnnì Yecl0r YeaJt prtcüplic selec able marker E aüt' ORI - high freqræncy - aubnrdì.ft - very unshbh cpþs per cell E tút' selectable markcr Yeast replicating 9ectr Yeeet ARS Yeait pmbtrphic selectable ma k r E aút' ORI - h:gh frequncy - aubnmui, - mderaþ slabiliry cpies percell E CLTI selecuble ma ker Yeast a tificial gþ mn36mp E öhlr ORI Yeast ARS and centrlere - lw frequency - replicales as chtrnsrne - stablê - 1 cpy per cell Yeasl Elcmere Bam Hl Barn Hl Yeasl Þlmere Yeast prrphic selecuble marker Figure 4.1. Classes f clning vectrs available fr use in S ac c hnr my c es c erevisiae.

60 -46- Yeast replicatin g vectrs The YRp (Yeast Replicating Plasmid) vectrs which are based n pbr322 and cntain the TRP I gene and the adjacent chrmsmal ørs (autnmusly replicating sequence). These vectrs transfrm yeast at high frequency (1000 t transfrmants per pg DNA) and are present at 3-30 cpies per cell (Struhl, 1983). The transfrmants, hwever, tend t be unstable; abut 90V f the cells lse the plasmid in the absence f selectin after l0 generatins (Struhl et al., t979). The intrductin f chrmsmal centrmere (CEN) sequences t these vectrs stabilizes them cnsiderably, hwever their cpy number is reduced t abut 1-2 cpies (Cla ke and Carbn, 1980). Yeast eismal vectrs The YEp (Yeast Epismal Plasmid) vectrs cntain the rigin f replicatin frm the naturally ccurring 2pm yeast plasmid (Beggs, 1978). 2pm-based plasmids have the advanrage that they are usually maintained at high cpy number ( cpies per cell). Therefre the high dsage f a gene lcated n such a plasmid can lead t elevated levels f the gene prduct (Mellr et a1.,1985). Althugh relatively stable, 2pm-based plasmids are usually lst frm yeast cells in the absence f cntinuus selectin fr plasmid- encded characteristics. The frequency f plasmid lss (the failure f plasmid mlecules t segregate t a daughter cell) is extremely variable, ranging frm less tha l7 f cells per generatin with the mre stable cnstructs (Beggs, 1981) t as high as 307 in ther cases (Struhl, 1983). Mst available data, hwever, relates t haplid strains and it is interesting t nte that in cntinuus culture 2 pm DNA is mre stable in diplid cells than haplids (Mead et a1.,1986).

61 DBR322 A a YIpl fitñ B a.&- y.3 :üütt n B t Inægratin ffi a i nræ i* l l t----.l.t.l.l Excisin pbr322 a.ã y'3 P:(m R c D 4.Ely.3 TffÑ a R a P Figure 4.2. Illustratin f site specific integratin f vecrr DNA int chrmsmal DNA. (A) Hmlgus recmbinatin ar rhe HIS3 regin between YIpl and yeast chrmsmal DNA. (B) Chrmsmal structure f the integrated transfrmant. (C) Pssible structures after excisin f the transfrmed DNA.

62 -47 - Inteerative vectrs The YIp (Yeast Integrative Plasmid) vectrs can nly achieve clnal expressin thrugh integratin int the genme by hmlgus recmbinatin. They must carry at least ne regin hmlgus t a yeast chrmsmal sequence; a selectable marker; and they must lack any drs sequence whse presence wuld cnfer autnmus replicatin. Site specific integratin f vectr DNA int chrmsmal DNA may tte mediated by hmlgus recmbinatin between ch msmal and vectr DNA. As shwn in Figure 4.2 fr the example f the HIS3 gene regin n chrmsme XV, hmlgus recmbinatin results in a duplicatin f the clned yeast gene (Struhl et al., L979). Transfrmed cells ften cntain nly ne cpy (smetimes a few cpies) f the vectr, which replicates with and under the cntrl f chrmsmal DNA. The nntandem duplicated structure is nt cmpletely stable. After 15 generatins f grwth in nnselective medium, apprximately l% f the clnies a e His-. This segregatin is accmpanied by the cmplete lss f transfrming DNA and almst certainly results frm a reversal f the riginal transfrmatin event - ie., excisin by hmlgus recmbinatin. V/ith yields f apprximately l-10 transfrmants per pg DNA, transfrmatin frequencies btained with circular integrating vectrs a e several rders f magnitude lwer than thse achieved with self-replicating plasmids. Orr-Weaver et al.. (1981) demnstrated that the efficiency f transfrmatin with intregrating plasmids is increased l0- t fld when the plasmid DNA is fust made linea by restrictin enzyme digestin within the regin f hmlgy t the yeast genme. They als shwed that plasmids cnsistently integrated at the chrmsmal lcatin crrespnding t the regin f the plasmid that had been cleaved by the restrictin endnuclease. It was cncluded that duble-strand breaks in DNA are highly recmbingenic and interact directly with hmlgus chrmsmal sequences by strand invasin and repair synthesis during recmbinatin. This lcalizatin

63 -48- effect has been ærmed targetting and is the basis fr many f the replacement and disruptin techniques used t manipulate genes n yeast chrmsmes. Yeast artificial chrmsmes The islatin and clning f yeast telmeric DNA pened the way fr cmbining these individual functinal elements int an a tificial chrmsme. Muray and Szstak (1983) cnstructed a series f linea plasmids, all f which cntained a selectable ma ker (eg. TRPI, HIS3, URA3), ars replicatr, CEN3 DNA, and telmeres frm Tetrahymena ribsmal DNA. The plasmids varied in length frm 9.8 t 55kb. The very large (>50kb) linea plasmids cntaining rs, CEN and telmeric DNA are present at ne r tw cpies per cell and are mre mittically stable than their circula ørs,cen cunterparts. This søbility cntrasts with the behaviur f shrt (<16kb) linea plasmids in which the additin f telmeric DNA reduces the mittic stability and increases cpy number. The generatin f Yeast artificial chrmsme (YAC) vectrs has prvided technlgy fr clning freign DNA fragments f several hund ed kilbase pairs. This system ffers a tenfld increase in the size f DNA mlecules than can be clned in a micrbial hst, and therefre prvides a system fr the analysis f cmplex DNA genmes Tranfrmatin prcedures DNA can be intrduced int yeast by remving the cell wall enzymatically t prduce spherplasts. Hinnen et al. (1978) first succeeded in the transfrmatin f spherplasts by using a plyethylene glycl-cacl2 medium which was riginally develped fr prtplast fusin f plant cells (van Slingen and van der Plaat, L977). Ha ashima er al.. (1984) have shwn that transfrmants btained under these cnditins were diplids r plyplids in cell size, shape and segregatin patterns f genetic ma kers after crssing wittr a standard haplid strain, althugh the strains used as recipients in the transfrmatin were

64 -49- haplids. It is therefre cnsidered that transfrmatin f spherplasted yeast cells is directly assciated with cell fusin. An alternative transfrmatin prcedure, which avids cell wall disslutin, invlves the treatment f whle cells with alkali catin salts r sulftrydryl reagents and plyethylene glycl (lt et a1.,1983). This is generally less eff,rcient (by 1-2 rders f magnitude) but the verall simplicity f the prcedure makes it the prefered methd fr many applicatins. Furthermre transfrmants can be detected after nly 36 hurs incubatin, whereas regeneratin f yeast cell walls takes 3-7 days after spherplast fusin. Recently, the methd f electrpratin has becme available fr yeast transfrmatin. This prcedu e invlves expsure f a sqspensin f cells and clned DNA t a high vltage electric discharge (Zimmerman and Vienken, 1982; Ptter et al., 1984). In essence, electrpratin makes use f the fact that the cell membrane acts as an electical capacitr which is unable (except thrugh in channels) t pass cwrent. Subjecting membranes t a high vltage electric field results in their temprary breakdwn and the frmatin f pres that a e large enugh t allw macrmlecules t enter r leave ttre cell. The recvery f membrane integrity is a natural decay prcess which can be delayed at 0C. In 1985 Karube et al. succeeded in using electric f,reld pulses t intrduce plasmid DNA int yeast spherplasts with an efficiency f 103 transfrrnants per pg DNA and Hashimt et al. (L985) used electrpratin t transfrm intact cells, althugh nly a lw efficiency, f apprximately 90 transfrmants per rg DNA, was achieved. Since then, varius mdificatins f the electrpratin methd have been described (Simn and McEntee, 1989; Rech er al. 1990) and an efficiency f as high as 107 Eansfrmants per pg DNA has been reprted (Meilhc et al. l99o). Varius ther yeast transfrmatin prcedures have been described including a spherplast treatment which des nt invlve cell fusin (Bergers and Percival, 1987);

65 -50- agitatin f whle cells with glass beads (Csta rz and Fx, 1988); and the direct treaünent f clnies frm agar plates (Keszenman-Pereyra and Hieda, 1988). Hwever, the alkali catin methd f lt et al. (1983), remains the favured prtcl fr mst yeast manipulatins. 4.I.4 Expressin and secretin f freign rteins Fr a transfrmed yeast strain t express a freign gene, the vectr must cntain varius DNA sequences and signals which regualte the prcessing f a freign DNA sequence int a prtein. These cllectively are termed the "expressin cmplex". The expressin cmplex cmprises a prmter and terminatr srrrunding the multiple clning sites fr freign gene insertin, and it may als cntain signals fr prtein secretin. Prmters riginating frm yeast genes are used since prmters ate fairly hst-specific and thse frm higher eukarytes tend t have lwer efficiency f initiatin f transcriptins when used in yeast (Beg gs et ai., 1980; Rthstein et ai., 1984). Table 4.2 gives sme cnìmn examples f prmter elements derived frm yeast. Prmters fr glyclytic enzymes are ppular as these enzymes are amng the best represented prteins in a yeast cell. Each can represent l-57 f ttal cell prtein - and it is assumed that this high-level expressin is is due t the assciatin f strng prmters with these genes (Gdey et a1.,1987). Sme f these prmters a e inducible r repressible (Brent, 1985) such that by the additin f an inducer r repressr, prtein prductin is either switched n r ff. This trait is used t achieve regulated expressin f heterlgus prteins in yeast. As an example, the transcriptin f the acid phsphatase gene is tightly repressed when inrganic phsphate is present in the gwth medium and induced by depletin f inrganic phsphate. Other examples have been discussed by Kingsman et al. (1985) and Gdey et al. (1987).

66 -51- Prmter element Abbreviatin Reference acid phsphatase PHO5 alchl dehydrgenase ADH1 galactkinase GALl glyceraldehyde-3- GAP3 phsphate dehydgrgenase Mating factr-cr MFa Phsphglycerate kinase PGK trise phphate ismerase TPI Kramer et al., 1984 Meyhack etai., L982 Hitzeman et al., 1981 Bennetzen and Hall, 1982 Stepien et al., 1983 Gff et al., 1984 Bitter and Egan, 1884 Hlland and Hlland, 1980 Bitter et al., 1984 Kurjan and Herskwitz, 1982 Dbsn et ai.,1982 Tuite et a1., 1982 Alber and Kawas akt, 1982 Table 4.2. Sme f the cnìmn prmter elements used fr freign gene expressin vectrs in yeast. The bilgical activity f a prtein als depends n its crrect prcessing which includes prtelytic cleavage f signal sequences, glycsylating, and tertiary structure flding. Several reviews have cvered heterlgus prtein prcessing frm yeast (Kingsman et al.l988; Smith et a1.,1985; Gdey et a1.,1987). Generally, it is useful t have the prtein prduct f freign genes secreted frm the yeast cell. The mechanism f secretin appears t be similar t that f higher eukarytes (Schekman and Nvmic, 1982). Hwever, heterlgus gene prducts derived frm higher eukarytes and prduced in yeast are nt secreted in high amunts indicating that the secretin signals fr these genes are nt crrectly recgnised in the ycast cell (Kingsman ef ø/., 1985). Therefre signal sequences frm prteins naturally secreted frm yeast are

67 - 52- usually used, such as that fr acid phsphatase (Smith et al., 1985), the a-mating type factr (Julius et a1.,1984), and and Halvrsen, 1983) Chapter aims The questin arises as t which f these systems is mst useful t the wine indusüry, and what are the limitatins r advantages f particular vectrs. A desirable vectr shuld have the fllwing attributes: i) gd transfrmatin efficiency, ii) stable inheritance f nansferred genes; and iü) usefulness in a wide range f industrial strains. It is als imperative that the vectr r the transfrmatin prcedure des nt adversely affect the characteristics f the yeast strain under fermentatin cnditins. A current limiting factr in the cmmerciat applicatin f recmbinant DNA technlgy t industrial fd and beverage yeasts is the presence f nn-fd yeast nucletide sequences in the transfrmants. In light f regulatry and market requirements then, it is imprtant t cnsider the pssibility f prducing transfrmants cntaining nly yeast sequences when assessing the advantages f vectr systems. The aim f this chapter is t investigate a number f selectable markers, vectr systems and transfrmatin prtcls t establish a system (which meets the criteria discussed abve) fr the applicatin f recmbinant DNA technlgy t wine yeasts. 4,2 RESULTS Chice f a selectable ma ker In rder t chse an efficient dminant selectable marker fr use in wine yeast transfrmatins, an experiment was perfrmed t test three available markers.

68 Ec! R1 111 Hind 111 Sal I Bam Ht Sma I Xba I H nd llt Cla I Hind lll Ec Rl Xba pcp bp pwxs bp Ec Pvu 11 Kpn 1 Ec Rf Kpn I Pvu 1f Nru I Sac I Ec Fl Pvu ll Xh I Sal 1 H nd Ec R1 11f SIU I Xh I Nar I Ec Rl 8am Ht Ec Rf Bam Hl sph I H nd lll Ec R1 Alu 1 Bam Hl paw bp H nd 1f l Sph 1 8am Hl Sph I Sal I prtm-c3 1f700 bp Ec Rl Ec Rl 4rm Ec Rl Ec Rt Xh l Figure 4.3. Clning vectrs used in transfrrnatin experiments.

69 -53- Ð Chlramphenicl Resistance Chlramphenicl (Cm) inhibits prtein synthesis ca ried ut by 70S ribsmes, as fund in prcarytes and mitchndria, but nt by 80S ribsmes, fund in the cytplasm f eukarytic cells. Eukarytic cells may be inhibited by Cm if their grwth depends upn utilising a ca bn surce that can nly be assimilated via mitchndrial-dependent aerbic metablism. Resistance t Cm by certain bacteria is mediated by plasmid encded chlramphenicl acetyl transferase GAD (Shaw, 1933). This enzyme causes inactivatin f the antibitic by acetylatin derived frm acetyl-ca. The CAT cding sequence frm Tn9 has been shwn t cnfer resistance t chlramphenicl in yeast when present as a single cpy per cell (Hadfield et al., 1986), The sequence, hwever requires yeast prmter and terminatr signals fr expressin. Therefre, the CAT cding regin was clned int the HindIII site f the vectr AAH5 (Ammerer, 1933) which cntains the yeast alchl dehydrgenase 1 (ADCI) prmter and terminatr sequences. A clne with the cding sequence in crrect rientatin with respect t the prmrer (paw 221) (Figure 4.3) was selected fr transfrmatin experiments. ü) Clclheximide The cyclheximide resistance gene frm Candida maltsa (RIM-C) has already been described. The plasmid prim-c3 (Figure 4.3) which cntains the resistance gene was btained frmdr. M. Takagi fr transfrmatin trials. üi) Sulfmeturn methyl Plasmid pcp which cntains the SMR1-410 gene was btained frm Dr. G Casey fr transfrmatin studies.

70 -54- In rder t test the suitability f each f these selectable ma kers fr use in the transfrmatin f wine yeasts, it was first necessary t determine the minimum inhibitry levels f the relevant chemicals n a range f yeast strains. Yeast grwth (clny size) was assessed n slid media cntaining the inhibitrs. Table 4.3 depicts the grwth respnses f yeast strains t a rìange f cncentratins f chlramphenicl, sulfmeturn methyl and cyclheximide. It shuld be nted that the minimal inhibitry cncentratin f cyclheximide was nt accurately determined, as cmplete grwth inhibitin f all strains was achieved with the lwest cncentratin tested (2 ttglml). Hwever, as successful selectin f primc-3 transfrmants had been reprted using 10 pglml cyclheximide (Takagi et al., 1986), it was decided that 2 pgml wuld be a suitable cncentratin fr transfrmatin experiments. Cmplete grwth inhibitin f all strains tested was achieved at a cncentratin f 10 pg/mt sulfmeturn methyl and 5 mg/ml chlramphenicl. These cncentratins were used in transfrmatin studies. The effîciency f each marker was tested in transfrmatin experiments using each respective plasmid n the haplid labratry yeast strain Ol1. This strain is auxtrphic fr bth uracil and leucine. Apprximately 5 lrg f plasmid DNA was used t transfrm OLl using ttre lithium acetate methd (It et al.,1983). Aliquts f the Eansfrmatin mix were then plated ut n different selectin media fr direct cmparisn f the auxtrphic and dminant ma kers n each plasmid. Table 4.4 shws the numbers f transfrmants btained in each case. These results indicate that under cnditins described here transfrmants f plasmids paw2zl and prim-c3 can be selected directly fr their auxtrphic ma ker (leucine), but nt fr their dminant ma kers (chlramphenicl and cyclheximide, respectively). PlamsidpCP2-4-10, hwever, resulted in an equal number f transfrmants when selected by either the auxrphic (uracil) r the dminant (sulfmeturn methyl) ma kers.

71 -55- chemical cncentatin YEAST STRAIN GROW'TH sulfbmeturn methyl (pelml) l0 20 Ll IA ? A A cyclheximide (ue/ml) zrj chlramphenicl (mslml) 0 I z Table 4.3. Respnses f a number f yeast strains t a range f cncentratins f chemicals t be used in the selectin prcess fr yeast transfrmants. + designates ne arbitrary unit f grwth after 4 days incubatin.

72 -56- Plasmid Transfrmants per 10Pg DNA Auxtrphic ma ker Dminant marker pl+w prim-c pcp Table 4.4. Efficlencies f auxtrphic and dminant selectable markers used in transfrmatin f strain O11. 'lïanstrmatrn methd pcp2-4-r0 transfrmants P WX509 Incu time transfrmans (hurs) Alkali catin (It er al.,1983) Alkali catin + ca rierdna Spherplast fusin (Burgers and Percival, 1987) Spherplast fusin + ca rierdna Electrpratin (Hashimt et al., 1985) Table 4.5. Cmparisn f yeast transfrmatin methds'

73 -57 - In rder t cnfirm apprpriate levels f expressin f the dminant ma kers n plasmids paw22l and prim-c3, transfrmants which had been selected by their auxtrphic markers were streaked nt the relevant dminant selectin media. Bth pavf221 and prm-c3 transfnnants shwed psitive grwth n the dminant selectin media, thus they displayed an indirect resistance phentype. One can cnclude that apprpriaæ levels f the RIM-C andcat genes are achieved in S. cerevísiae, but that direct expsure f transfrmants t either chlramphenicl r cyclheximide is nt feasible. It is pssible that pst-transfrmatin incubatin prir t selectin wuld result in islatin f transfrmants based n cyclheximide r chlramphenicl resistance. In fact, the transfrmatin methd described by Takagi et al. (L986) includes an vernight incubatin f transfrmants prir t expsure t cyclheximide. Transfrmatin f Oll using prim- C3 and pltw22l was therefre repeated with an altered selectin prcedure. Fllwing transfrmatin, cells were plated directly nt nn-selective media, incubated vernight at 28C, and then verlayed with selective media. This prcedure, hwever, again failed t yield tansfrmants with either plasmid. Due t ease f selectin and a transfrmatin cmparable efficiency with theura3 marker, the SMRI gene was used in further transfrmatin experiments Chice f a transfrmatin prcedure A number f transfrmatin methds were tested t find ne mst suitable fr manipulatin f wine yeast strains: Ð ttre alkali catin methd f It et ai. (1983): ä) the alkali catin methd f It et al. (7983) with the additin f carrier DNA t the plasmid DNA; üi) the spherplast fusin methd f Burgers and Percival (1987);

74 -58- iv) the spherplast fusin methd f Burgers and Percival (1987) with the additin f ca rierdna t ttre plasmid DNA; and v) the electrpratin f intact cells as described by Hashimt et al. (1e85). Wine yeast strain 5A was transfrmed with plasmids pcp2-4-l0 and linearised pv/x509 (digested with PvuII). Transfrmants were selected by resistance t sulfmeturn methyl. Each transfrmatin experiment was perfrmed in triplicate - the results f these transfrmatin experiments are presented in Table 4.5. Success f the methd is determined by the average number f transfrmants, and the expediency with which transfrmans a e btained. The use f carrier DNA in alkali catin transfrmatins and in the spherplast fusin methd increased efficiencies apprximately three fld. This bservatin may be explained by the fact that freign DNA is susceptible t nuclease attack in the hst cell, and that the carrier DNA prvides an alternative substate fr these nucleases thus'prtecting', t sme extent, the plasmid DNA. Direct cmparisn f all methds tested shws that the highest yields f transfrmants were btained with the spherplast fusirvca rier DNA prcedure, and that electrpratin prved t be the least efficient. Electrpratin, hwever is the quickest and easiest methd t perfrm (althugh expensive equipment is required) and ransftmants are visible after apprximately 36 hurs incubatin. The alkali catin methd is relatively inexpensive and expedient, and again, transfrmants can be detected after appprximately 36 hurs incubatin. The spherplast fusin methd is the mst time-cnsuming t perfrm, and2-4 days are required fr detectin f transfrmants.

75 -59- AWRI strain Transfrmants per 10pg DNA pcp linea p 7A &q. 10A 114 3A 6A 5A 2A 50u) W )ó(, Table 4.6. Transfrmants btained using the alkali catin/ca rier DNA methd n a range f cmmercially ppula wine yeast strains.

76 -60- Cnsidering all factrs, the alkali catiry' ca rier DNA methd was selected. This methd was used t transfrm a number f wine yeasts frm the AWRI cllectin with plasmids pcp and prilx509, selecting directly fr resistance t sulfmeturn methyl. Results frm this experiment (Table 4.6) verify that the selectable marker and transfrmatin methd are suitable fr a wide range f wine yeasts strains. Figure 4.4 shws the results f the transfrmatin f ne f these wine yeast strains (74) with pcp2-4-10, circular pv/x509 and linea pwx509. Plate A shws the transfrmants btained with pcp2-4-lo; an efficiency f 5000 transfrmants per 10 pg DNA was btained. Plate B is the cntrl ransfrmatin in which n DNA was present in the transfrmatin mixture. Five clnies are present n this plate and represent the backgrund spntaneus mutants which are resistant t sulfmeturn methyl. Plate C shws the transfrmants btained with linearised pv/x509 - an efficiency f 560 transfrmants per 10pg DNA was achieved. Plate D shws the results f transfrmatin with circula p\ü/x509, and in this case the efficiency was similar t that btained in the cntrl transfrmatin (Plate B). One can cnclude, therefre, that transfrmatin with circular pwx509 was nt successful Analysis f transfrmants Transfrmants f strain 5A were analysed fr the presence f plasmids pcp r pwx509. Ttal DNA was islated frm 3 transfrmants f pcp (5A-pCP2-al0), 3 transfrmants f pwx509 (54-pWX509) and a cntrl, unüansfrmed clny f 54. The DNA was digested with PvuII, electrphresed n a 0.8V agíuse gel and transferred by Suthern bltting t a nyln membrane. Bth plasmids cntain sequences f pbr322. Therefre, the presence f each plasmid culd be detected by hybridizatin f the membrane with radilabeled pbr322. Figure 4.5 depicts the result f this hybridizatin experiment. As expected, n hybridizatin signal is apparent in the lanes cntaining DNA frm the untransfrmed clny

77 Figure 4.4. Transfrmatin f wine yeast srain AWRI TA with the fllwing plasmids: Plate A - pcp2-4-10: Plate B - n DNA cntrl; Plate C - pwx509 linearised at the Pvu II site; Plate D - circular pwx509'

78 Figure 4.5. Suthern hybridizatin detecting plasmids pcp and plwx509 in transfrmants f strain 54. Panel A: Ttal yeast DNA samples electrphresed n a l% agarse gel, stained with ethidium brmide and visualised n a UV light bx. Panel B: Autradigram f gel depicted in Panel A, after hybridizatin with a radilabeled sample f pbr322. Lanes: t0-t L î. hind III markers 5A 5A -pcp A-pWX509 5A digested with hu tr 5A-pCP digested with Pvu II 5A-pWX509 digested with Pvu tr pcp pcp digested with Pvu II prwx509 pwx509 digested with Pvu tr

79 A L I910 11'L r.rr rhlldldlqtlil llhrlf Ë -a ra B t L3141s l - OO- (t < 10kb (D < 6kb < 3kb

80 -61- (lanes 2 andg). The undigested DNA frm transfrmants btained with plasmid pcp (lanes 3-5) shw a strng signal at the psitin f the chrmsmal DNA, and a weaker signal at apprximately 9.8kb, which crrelates t the size f the undigested plasmid pcp (lane 16). The signal at the psitin f the chrmsmal DNA is unexpected as the plasmid is a self-replicating vectr and is nt expected t be assciated with genmic DNA. The PvuII digested DNA samples (lanes 10-12) shw strng hybridizatin signals t fragments f apprximaæly 10kb and 3kb. These crrespnd t the fragments generated by a PvuII digest f pcp (lane 17). Three ther weaker signals are apparent at apprximately 4.0, 3.8 and 2.6kb respectively. The undigested DNA samples frm transfrmants f the pwx509 vectr (lanes 6-8) shw weak signals at the chrmsmal DNA psitin. As pwx509 is an integrating vectr, the hybridizatin signal is expected t be assciated with genmic DNA. The strength f the signal is indicative f the lw cpy number (pssibly nly ne cpy per genme) f integrating plasmids. The PvuII digested samples (lanes 13-15) give rise t signals f 6.0kb, which crrespnds t the size f the plasmid after excisin frm the genme. Anther band f apprximately equal intensity is als present at a psitin which represents a fragment size f 2.6kb. This signal is cmmn t bth pcp and pwx509 transfrmants. As plasmid pwx509 was linearised at the PvuII site prir t transfrmatin, the highty recmbingenic ends f the plasmid mlecule will target the vectr t integrate directly int the PvuII site f the ILV2 gene n chrmsme XIII (Figure 4.6). The psitin f plasmid pwx509 in the genme was analysed by pulsed field gel electrphresis and Suthern hybridizatin. Chrmsmes were islated frm a cntrl untra sfrïned 5A clny and a pv/x509 transfrmant. They were then separated by electrphresis using the Transverse Alternating Field Electrphresis (TAFE) system (Gardiner et at. 1986) and transferred by Suthern bltting t a nyln membrane. A radilabeled EcRl fragment f the SMRL gene wírs hybridized t the membrane t lcate

81 - 62- pwxs09 smrf-410 Pvull {-llv 24 Pvull a- llv 2 I Pvull I nteg rat n chrmsme Xlll <-llv 24, Pvull chrmsme Xlll Figure 4.6. Integratin pattern f pwx509 int yeast chrmsme XIIL

82 -63- the ILV2 gene. The signal resulting frm this hybridizatin (Figure 4.7) is present in bth the untransfrmed an'd transfrmed strains and, as the ILV2lcus has previusly been mapped t chrmsme XIII (Mrtimet et al.,1989), this band represents chrmsme XIII. The membrane was then stripped f the SMRI prbe and rehybridized with a labeled pbr322 plasmid t identify the psitin f the integrated plasmid. Figure 4.7 shws that the pbr322 prbe des nt hybridize t the untransfrmed strain but, in the tranfrmant, has hybridized t the same band as the SMRI prbe, indicating that plasmid pwx509 has been successfully targetted t chrmsme )üii. An experiment uras perfrmed t determine whether plasmid pv/x509 had integrated int every cpy f the ILV2 gene in the transfrmant. Given that wine yeast strains are diplid r plyplid, there will be mre than ne cpy f the ILV2 gene present in the nucleus f strain 54. One applicatin f the transfrmatin system described in this chapter culd be t disrupt (and therefre inactivate) undesirable genes in wine yeasts. In rder t achieve gene inactivatin, it is necessary t disrupt every cpy f that particular gene in the cell. It is therefre imprtant t knw whether this transfrmatin prcedure targets plasmid pwx509 t every cpy f the ILV2 sequence. Ttal DNA was islated frm cntrl strain 5A and frm srain 5A transfrmed with pwx509 (5A-pWX509). Bth DNA samples were digested with EcRV. There are n EcRV sites in plasmid pwx509. Therefre EcRV generated fragments f chrmsme XtrI in which pwx509 has integrated in the ILV2 gene will be 6kb lnger than fragments which cntain the intact ILV2 gene. The DNA samples were electphresed n a t% agarse gel and transferred t a nyln membrane. The membrane was hybridized with a radiactively tabeled sequence f the ILV2 gene. Results f this hybridizatin are presented in Figure 4.8. A fragment f apprximately 5.9kb is detected in the untansfrmed strain 5A DNA (Panel B). This band represents EcRV fragments cntaining intact ILV2 genes. In the transfrmant DNA (54-pWX509) tw fragments were highlighted - ne at 5.9kb and the ther at apprximately 12kb. The 12kb fragment represents EcRV fragments which

83 Figure 4.7. Suthern hybridizatin depicting ttre chrmsmal lcatin f the pwx509 vectr in tansfrmed strain 5A-pWX509. Panel A: Yeast chrmsmes sepafated n al% agarse gel by transverse altemating freld electrphresis. Panel B: Autradigram f gel depicted in Panel A prbed with a radilabeled sequence f he ILV2 gene. Panel C: Autradigram f gel depicted in Panel A prbed with a radilabeled sequence fpbr322.

84 I I I A ō = = J = 5A 5A'pWX509 5A 5A-pWX509 5A 5A-pWX509

85 Figure 4.8. Suthern hybridizatin t detect cpy number f integrated plasmid pwx509 in transfrmant 5A-pìWX509. Panel A: Ttal cellular DNA digested with Ec RV, electrphred n a l7 agarse gel, stained with ethidium brmide and visualised n a UV light bx. Panel B: Autradigram f gel depicted in Panel A prbed with a radilabeled sequence f the ILV2 gene.

86 (,r æ I b,j \ t, úú tl \J 5O\\O(, (,r \ O\ O\ Fxxx úúúú I l, Hind III markers 5A 5A-pWX509 fw I t 5A 5A-pWX509

87 -64- cntain pwx509 integrated int the ILV2 gene. The 5.9kb fragment detected in the transfrmant DNA sample indicate that there are sme intact cpies f the ILV2 gene present in this strain. Therefre plasmid pv/x509 has nt integrated int every available cpy f the ILV2 gene. It is interesring t nte that the hybridizatin signal f fragment 12kb is apprximately twice as intense as the signat f the 5.9kb fragment in the 5A-pWX509 DNA. There are tw pssible explanatins fr this bservatin. First, there may be twice as many cpies f the ILV2 gene which have integrated cpies f pwx509 as there are f intact ILV2 genes. If this is the case, it may reflect the plidy level f chrmsme XtrI f strain 5A; that is, the strain may be triplid fr chrmsme XIII. Secndly, there may be mre than ne cpy f the plasmid pwx509 integrated int ne particularlzv2 síte. If this event has ccurred, it is nt pssible t draw any cnclusins abut the plidy level f chrmsme XIII Fermentatin trials Transfrmants f strain 5A were subjected t fermentatin trials in rder t determine whether the intrductin f freign DNA had a deleterius effect n yeast perfrmance. Trials were cnducted n: Ð a cntrl untransfrmed 5A; ü) 5A transfrmed wittr the self-replicating plasmid (54- pcp2-4-10); and üi) 5A transfrmed with the integrating plasmid (54-pWX509). Three clnies f each type f strain were inculated in duplicate int Riesling grape juice and fermented at 25C. Samples were taken at regular intervals and assayed fr sugar cntent (by refractmeter) and yeast grwth (by ptical density at 650nm). Average readings were pltted n graphs (Figure 4.9). N significant differences were fund between the three strains.

88 -65- The ph and percent alchl (vvvl) were als measured fr each f the ferments. Average readings are presented in Table 4.7. Agun, there are n significant differences in data beween the three strains. On cmpletin f each ferment, aliquts f the yeast ppulatin were apprpriately diluted and equal vlumes were plated nt selective (SD cntaining 10 flg/nl sulfmeturn methyl) and nn-selective (YPD) media in rder t analyse stability f the intrduced plasmids. The number f clnies per plate after 48 hurs incubatin at 28C are recrded in Table 4.8 The number f clnies n the YPD media represents the ttal clny frming units present at the end f the ferment The clnies grwing n selecúve media represent clny frming units which have retained the ptasmid. These results indicate that apprximately 85V f the cells f the strain Eansfrmed with the self-replicating plasmid (pcp2-a-10) have retained the plasmid thrughut fermentatin. N plasmid lss is evident frm the strain transfrmed with the integrative vectr (pwx509).

89 X ) E ( ) I 6l tr C- r q) ú 2W 160 r Time (h) r Figure 4.9. Fermentatin curyes f the cntrl untransfrmed (AWRI 5A) (tr) strain, the strain transfrmed with the replicating vectr (54-pCP2-4-10) (A) and the strain transfrmed with the integrating vectr (54-pWX509) (O). Strain ph Alchl (V vuvl) AWRI5A A (pcp2-4-10) sa (pwx509) Table 4.7. phvalues and alchl cntent f wines prduced by transfrmed strains.

90 -67 - STRAIN VIABIE COI-ONIES per ml IIERBICIDE RESISTANT CTOLONIES per ml 7 FIERBICIDE RESISTANT CELI.S AWRISA x A (pcp ) x 107 r x A (pwx509) LZ x x Table 4.8. Søbitity f intrduced plasmids in transfrmed yeast strains during fermentatin.

91 Gene regulatin during fermentatin The pssibility f islating prmters which a e 'switched n' at specific stages f fermentatin was investigated. This investigatin invlved the analysis f prtein synthesis in a wine yeast snain at different times f a fermentatin. Variatins in the prfile f prteins synthesized at different fermentatin stages wuld eflect specific prmter regulatin. Fr example, the presence f a prtein late in the ferment that had been absent in early stages wuld indicate that the prmter assciated with this prtein had been 'switched n' late in the fermentatin. Riesling grape juice was inculated with an vernight culture f strain AWRI 54. Fermentatin was cnducted at 18C under anaerbic cnditins with gentle agitatin. Samples were remved aseptically and analysed fr yeast grwth (by ptical density) and sugar utilizatin (by refractive index). At th ee stages f the fermentatin - early mid phase (M), and late phase (L) - samples were remved and incubated with radiactively labeled leucine fr tw hurs. These three stages f fermentatin are depicted in Figure 4.10 (Panel A). Fllwing each incubatin, ttal cellula prteins were islated frm the suspensin and stred fr analysis. Prtein samples were later electrphresed n a plyacrylamide gel, stained with cmassie blue and treated by flurgraphy. The fluragraphic treatment is perfrmed prir t autradigraphy, and enhances the detectin f radiactivity. Results f the cmassie blue staining and autradigraphy a e shwn in Figure 4.10 (Panel B). The autradigram indicates that the highest level f incrprtatin f radiactive leucine ccurred in the early phase f fermentain (and yeast grwth). The level f incrpratin decreased as the fermentatin prgressed and by late phase is difficult t detect by autradigraphy. Althugh definitin f the bands is pr, it appears that sme bands my be specific t the mid r late phases f fermentatin. An example f such a band

92 Figure Prtein synthesis during fermentatin. Tp panel: Fermentatin prgrcss f strain AWRI5A depicted by yeast grwth (O) and sugar utilizatin (tr). Samples were remved frm the ferment at early (E)' middle (M) and late (L) stages as indicated. Lwer panel Prtein prfiles f each different sample after cmassie blue staining and flurgraphy.

93 30 E I M f L 300 E lf ú t( h c a,! F c rf It È 20 t x t,! c at I nl b /, TiDE (ùu s) clnîstie ùluc stain E M L E M L i.*å.f* i I :1,'K. õ-fratc Çt*} I { 1 4d tã -^-í *, t llt a l I l fl ;x;'- { fuà+1r} a ' {>l*,. a I I I l ì l

94 -69- is indicated by an arrw. This band may r present a prtein which is selectively expressed in the later stages f fermentatin. 4.3 DISCUSSION A system has been described fr the stable intrductin f freign DNA t a range f wine yeast strains. This system utilizes the SMRI gene (Casey et al., 1988) as a dminant selectable ma ker. Advantages f this marker ver thers that have previusly been reprted include the fact ttrat the gene is derived frm a natural ILV2 yeast gene, and therefre shares sufficient hmlgy with sequences n the yeast genme t target direct integratin. Only ne cpy f the gene is required t cnfer resistance t the herbicide sulfmeturn methyl upn transfrmed cells. Mrever, the use f genetically engineered rganisms in the wine industry is likely t be mre acceptable if the intrduced DNA cmprises naturally ccurring yeast sequences. The mst efficient transfrmatin methd tested in this study, with respect t yield f ûansfrmants, cst and expediency was a mdificatin f the alkati catin methd (It et at.,1983). This mdi rcatin included the incubatin f cmpetent yeast cells with carrier DNA and plasmid DNA at the time f cellula uptake. This methd was shwn t be successful with a range f cmmercially ppular wine yeast strains. Suthern hybridizatin analysis f transfrmants cnfirmed the presence f bth the self-replicating (pcp2-4-10) and integrating (pwx509) vectrs. Plasmid pwx509 was linea ised at the },lutr site f the SMRI gene prir t transfrmatin. This was designed t targer the plasmid t the ILV2 gene n chrmsme XIII. TAFE separatin and Suthern hybridizatin f chrmsmes frm the pwx509 transfrmant revealed that the plasmid had been successfully targetted t chrmsme XtrI.

95 -70- Fermentatin trials indicaæd that there were n significant changes in the grwth r fermenatin rates between the riginal strain 5A and the strains transfrmed with either the self-replicating plasmid (pcp2-4-10) r the integrating vectr (pwx509). Similarly, there were n significant differences in the ph measurements r the alchl cntent f the resultant wines. Stability f the intrduced plasmids thrughut fermentatin was measured by cmparing clny frming units (cfu) in aliquts f the finished ferment n bth selective and nn-selective media. These results indicated that the self-replicating plasmid displayed instability, with apprximately 857 f the ttal yeast cells reøining the plasmid thrughut fermentatin. N plasmid lss was detected in the transfrmed strain which cntained the integrating vectr. Suthern analysis f the pwx509 transfrmants shwed that the vectr did nt integmte int every cpy f the ILV2 gene. It shuld be nted that nly ne transfrmant was analysed in this experiment, s it is pssible that all cpies f the gene are disrupted in sme cases. Als, different transfrmatin cnditins may encurage integratin int all pssible sites. Fr example, increasing the amunt f plasmid DNA in the transfrmatin mixture may result in disruptin f all hmlgus sites. Nevertheless ne can cnclude frm this result that under the transfrmatin cnditins described in this chapter, integratin will nt necessarily ccur in alt cpies f the targeted gene. This is an imprtant bservatin cnsidering that ne pssible applicatin f this transfrmatin system may be t inactivate undesirable genes. Fr example, Suizu et al. (1990) recently inacrivated the CARI gene f a labratry haplid S. cerevisiae strain by hmlgus integratin. The disruptin f this gene, which cdes fr arginase, resulted in a yeast strain which did nt prduce urea. The chemical reactin f urea and ethyl alchl in alchlic beverages prduces ethyl carbamate, which is a suspected carcingen and, therefre, an undesirable cmpund in wine. The CARI gene disruptin f a wine yeast

96 -7 l- strain may be desirable in terms f reducing ptential levels f ethyl ca bamate in wine. Hwever, t achieve gene disruptin, integratin wuld be required int every cpy f the CARI gene f the plyplid wine yeast cells. Results f this chapter shw that this may nt ccgr in a ne-step transfrmatin prcedure. Multiple transfrmatin events, each ne requiring a different selectable marker, may prvide a means fr gene inactivatin in wine yeast strains. Cnclusive results abut regulated gene expressin during fermentatin were nt btained in this study. Atthugh prtein synthesis investigatins indicated that sme genes may be transcribed at specific stages f fermentatin, the search fr regulated prmters will require a mre thrugh and detailed analysis. Fr example, mrna species shuld be islated frm different stages f fermentatin and used t prduce cdna libraries. Differential screening f these libraries may reveal stage-specific clnes. These clnes culd then be used t screen a genmic library in rder t identify the assciated prmter elements The identificatin f prmters in this type f study wuld increase the scpe fr designing efficient strategies fr freign gene expressin during fermentatin. Fr example, the prducts f sme intrduced genes may interfere with grwth f the yeast strain. Switching n freign gene expressin twards the end f a fermentatin after a high yeast bimass has been reached may result in mre ecnmical yields f the freign prduct. It can be cncluded that the system described in this chapter can be used t target freign DNA t specific regins f the yeast genme such that it will be maintained stably in the strain thrughut fermentatin. The transfrmatin prcedure is applicable t a wide range f wine yeast strains and des nt adversely affect the fermentatin perfrmance f the yeast.

97 -72- Chapter 5 Develpment f a system fr wine yeast strain marking and identificatin 5.1 INTRODUCTION Winemakers have recgnized the imprtance f the yeast strain in determining the flavur and quality f wine, and the ecnmics f its prductin (Rankine, 1968). Pure culture inculatin nw dminates the industry, particularly in the newer wine prducing regins. This practice prvides greater reliablility and cntrl f the fermentatin thrugh mre rapid nset and cmpletin resulting in wines with fewer flavur defects (Rankine, lg77). Furthermre, this technlgy enables strain differences t be utilized in the prductin f a wide variety f wines. The selecún and cha acterizatin f wine yeasts by The Australian V/ine Resea ch Institute supprted and enhanced the utility f pure culture inculatin technlgy (Rankine, 1968). Mre recently, the availability f active dried yeasts have given the winemaker even greater scpe fr expliting the enlgical prperties f different strains. These factrs have led t the use f several yeast cultures in the winery; a practice which addresses the need fr a suitable yeast identificatin scheme. The prcess f enlgy emplys a raw material cntaining an unknwn lad f indigenus yeasts, sme f which are capable f prducing ff-flavurs and spiling wine. Natural grape must will underg a spntaneus fermentatin by the indigenus yeasts t prduce wine. In the case f wines prduced by pure culture inculatin, it is cnsidered that the additin f S02 will suppress the indigenus yeast ppulatin and thereby encr rage fermentatin by the inculated strain (Kunkee and Amerine, 1970; Benda, 1982; Kunkee, 1984). Only recently have quantitative kinetic studies f the eclgy f grape juice fermentatin been made (Fleet et al. 1984; Hea d and Fleet, 1985, 1986a). These

98 -73- investigatins, made with bth inculated and uninculated grape juices under a range f enlgical cnditins, prvide evidence which challenges the cnceptthat Saccharmyces necessarily suppresses nn-saccharmyces yeasts t establish itself as the dminant rganism. Fermentatin cnducted at lw temperature, fr example 1@C, and t a lesser extent at high ph, fr example ph 3.5, resulted in greater grwth and survival f Kleckcra apiculata and several Candida species than previusly believed, and in sme cases apprached that f Saccharmyc J yeasts (Hea d and Fleet, 1938). Furthermre, it was shwn that SO2 did nt necessarily suppress the grwth f nn-s ccharmyces yeasts (Heard and Fleet, 1988a). These findings, therefre, call int questin the effectiveness f sme enlgical pracrices fr cntrlling the grwth and survivat f indigenus yeasts. Relevant studies have been hindered by the lack f suitable techniques fr the differential quantitatin f the inculated strain and wild yeast. While methds fr differentiating between Sacclnrmyces and many nn-saccharmyces yeasts have recently been develped (Heard and Fleet, 1986b), few methds arc yet available t distinguish between the inculated srain and wild strains f Sacclnrrnyces. The reasn fr this is simply that enlgical strains are merely wild strain s f Sacclnrmyces selected fr desirable prperties. Therefre, n classical taxnmic differences exist which wuld permit strain differentiatin. Research presented in this chapter invlves tw different appraches t mnitr yeast in the prductin f wine. The first investigates a methd fr identifying yeasts t the strain level by prducing a chrmsmal fingerprint. The secnd apprach invlves tagging a wine yeast with a genetic marker enabling the efficiency f enlgical practices t be determined.

99 Yeast chrmsmal fingerprints Saccharmyces cerevisiae may be identifîed by classical methds including cell shape and mde f reprductin; sugars fermented; ca bn and nitrgen cmpunds assimilated; vitamin requirements; sensitivity t inhibitrs; c-enzyme Q system and DNA base cmpsitin (Kreger-van Rij, 1984; Kreger-van Rij, 1987), r by the pattern f respnses t sme 80 physilgical tests (Barnet. et at., t983). A cnsiderable input f expertise and time precludes the applicatin f these methds t utine identificatin. Furthermre, these methds are nt useful in distinguishing strains f S. cerevisiae. Tw fundamentally different appraches t identifying strains f yeast have been pursued, namely phentypic and gentypic analysis. In the fîrst, a yeast characteristic r family f cmpnents is measured r prfiled. Giant clny mrphlgy has prved useful but may require at least 2-4 weeks fr cnclusive results (Heard and Fleet, 1987). Serlgy (Nishikawa et al.,lg7g) is rapid but generally lacks specificity and reslutin. Fatty acid cmpsitin as measured by gas chrmatgraphy is relatively rapid and a generally accessible technique (Osthuizen et a1.,1987) but its ptential fr strain identificatin still needs t be cnfirmed. Yeast prtein fingerprinting in the frm f electrphretic patterns f selected enzymes (Subden et a1.,1982), ttal sluble cell prtein (van Vuuren and van der Meer, 1987) and exrracellular yeast prteins (Cilfi, 1938) indicate high specificity and sensitivity. All tt ee prtein methds generated unique fingerprints enabling identificatin f clsely related süains. An imprtant disadvantage f all phentypic methds is the rigrus requirement fr cntrl f cultural cnditins in rder t minimize the variatin f results. In the gentypic apprach, DNA cmpsitin and structure is analysed. The principal advantage f DNA analysis is the cmplete independence f results n yeast cultural cnditins. Tw early methds which have been widely applied t yeast systematics a e DNA base cmpsitin and DNA hmlgy (Price et a1.,1978; Kurtzman

100 -75- and phaff, 1987). DNA base cmpsitin is measured by either thermal denaturatin prcedures r by buyant density measurements and gives nly a tentative indicatin f strain identity. DNA hmlgy, which measures the degree f affinity r assciatin between fragments f single-stranded DNA frm tw yeasts, is a measure f strain relatedness but generally is nt sufficiently sensitive t reveal strain identity. Of greater interest, hwever, is characterizatin f DNA by the techniques f restrictin analysis and pulsed freld gel electphresis. Restrictin endnucleases cut DNA at specific pints. This actin generates a unique range f DNA fragments whse size is readily determined by agarse gel electrphresis resulting in a characteristic pattern f bands in the gel. This technique has nly recently been applied t yeast and is becming a critical technique in yeast systematics (Kreger-van Rij, 1934). Bth nuclear (Prffrn et al., 984;Panchall et a1.,19s7) and mitchndrial DNA (mtdna) (l-ee et al.,1985) have been cha acterised. Restrictin f nuclear DNA yields a large number f prly reslved fragments. Therefre, specific prbes which nly hybridize t sequences fund in ne r a few f the fragments are used t generate simpler patterns. Ciriacy and Grssman (1988) screened eight wine yeasts with three prbes cnstn cted frm the alchl dehydrgenase gene (ADH), 2 pm DNA and Ty sequence. The ADH and2 pm DNA prbes revealed little difference between strains, while Ty, a dispersed repetitive sequence, prduced unique fingerprints. Analysis f mtdna purifîed frm srains f yeast islated frm wine has been reprted by Duburdieu et al. (1987). Digestin f mtdna with the restrictin enzyme EcRl yielded frm 4 t 10 fragments in the 23 strains assayed. All but tw strains shwed unique frngerprints. The technique, therefre, prvides unequivcal identificatin f wine yeast but the cmplexity f mtdna islatin precludes applicatin t rutine analysis.

101 -76- A recently develped technique, pulsed field electrphresis, has extended the size range f DNA mlecules which can be fractinated by the electrphretic technique (Shwartz and Cantr, 1984). Utilizing this technique, it is nw pssible t electrphretically separate intact yeast chrmsmes n a gel. Chrmsmes frm S.cerevisiae have been shwn t range in size frm 150kbp (chrmsme I) t apprximately 2,500 kbp (chrmsme XII) (Carle and Olsn, 1985; DeJnge et al., 1936). The methd invlves the applicatin f an altemating electric field t the slab gel which frces the chrmsmes t peridically change their directin r mbility; it is thught that the smaller chrmsmes migrate mre rapidly thrugh the gel because they can rerientate themselves mre rapidly t the changing electric field thrugh the gel and becme stuck in the gel matrix less frequently. A variety f electrde a rangements have been used and many a e available as cmmercial units. Field inversin gel elctrphresis (FIGE) is the least cmplex arangement (Carle er a1.,1986). A standard hrizntal submarine gel is used but the electric field is alternatively inverted and the frequency f inversin is cntrlled. Unequal perids f field inversn prduces a net migratin in ne directin, while the duratin and frequency f the electric pulses affects migratin rates f different sized macrmlecules. This methd allws the reslutin f DNA mlecules f several hundred thusand base pairs. Hwever, this is nt adequate fr yeast chrmsmal DNA which extends int the megabase (mbp) range. Orthgnal field alternatin gel electrphresis (OFAGE) differs frm FIGE essentially by the angle betweeen the tw alternating electric fields. In FIGE, the field is inverred by 180 whereas the inversin in OFAGE is 12@ in the plane f the gel. The rthgnal field causes the DNA mlecules t zigzag acrss the plane f the gel. The reslutin range, which als depends n the pulse frequency, is extended t the 9,000 kbp range (Carle and Olsn, 1984). A disadvantage f OFAGE is that the lanes f DNA migratin are nt straight, making cmparisns between samples difficult.

102 -77 - In 1986, tw new cncepts were published: Transverse Alternating Field Electrphresis (TAFE) and Cntur-clamped Hmgenus Electric Field (CHEF) electrphresis. In TAFE, the gel is riented vertically with a simple fur electrde a:ray placed n either side f the gel (Gardiner and Pattersn, 1986). Sample mlecules are frced t zi1zagthrugh the thickness f the gel. As all lanes experience the same electrical field effects the bands remain straight. As the mlecules mve dwn the gel, they are subjected t cntinual variatins in field strength and rerientatin angle - but t all lanes equally. Als in 1986, Chu er /. demnstrated that the critical factr in reslving large DNA is keeping the angle between the tw directins f alternating fields grcater than 90 degrees. Field gradients were nt necessary fr high reslutin. T avid field distrtins arising frm the finite size f an electrphresis chamber, the researchers designed a hexagnal a aay that placed pairs f electrdes at a 12C0. angle. In their CI{EF system, rather than use lng electrdes, the resea chers brke the electrdes up int shrt segments. Then, using resistrs (and later transistrs), they clamped and frced the elcrdes t their ideal vltages. The samples in the CHEF system experience a large, highly unifrm electric field with n edge effects. The chrmsme banding patterns f S. cerevisiae (Carle and Olsn, 1985; De Jnge et a1.,1986), Candida albicans (Snell and Wilkens, 1986) and Schizsaccharmyces pmbe (Smith et al., 1987; Vll ath and Davis, 1937) have been determined using pulsed freld gel electrphresis. The karytypes f the va ius yeasts shw great va iatin in the size and number f chrmsmes. Such variants are nted even amng strains f the ne species (De Jnge et al., 1936). The technique f pulsed field gel electrphresis, therefre, ffers an imprved system fr the identificatin f wine yeast strains.

103 -78-5.L.2 Marked strains Determining the efficiency f winemaking practices (such as mnitring yeast prpagatin and fermentatin fr cntaminatin) is hampered by the absence f suitable methds t differentially quantify the inculated wine yeast strain and indigenus yeasts. Sme studies have explited the naturally ccurring killer r zymcidal ma ker present in sme strains as a tl fr such analysis (Heard and Fleet, 1988). This ma ker can be used in either f tw ways: (i) samples can be plated nt nutrient media fr a ttal cunt and replica plated nt media cntaining txin t determine the prprtin f marked strain; r (ii) the prprtin f the marked strain can be determined by replica plating nt a lawn f sensitive yeast. Hwever, a deliberately marked enlgical strain as develped by Vezinhet and clleagues (Vezinhet and Lacrix, 1984; Vezinhet, 1985) prvides a mre pwerful technique fr investigatins f ttris type. The strain, which is nw cmmercialized as Kl, is duble marked with tw antibitic markers, diurn and erythrmycin. An extensive survey f yeasts fr resistance t these antibitics demnstrated that few strains are naturally resistant t bth drugs simultaneusly. Kl was develped by selecting fr natural mutants in a ppulatin f the Lalvin V yeast. The results f extensive winery trials using this marked strain have already been and Aizac, 1988). The main emphasis f these trials was t investigate the efficiency f varius inculatin techniques, including methd f yeast prpagatin, direct inculatin, inculum level and timing f inculatin. The marked strain enabled data t be accumulated n the effect f winery prcedures by ttre extent f K1 dminatin during fermentatin. Recmmendatins culd then be given fr imprving the efficiency f fermentatin cntrl.

104 -79- The ma ked strain, Kl, prvides an insight int the kinetics f yeast ppulatins during fermentatin and therefre represents a majr advance in the cntrl f wine prductin. Hwever, a majr limitatin exists in the fact that wine yeast strains f chice cannt be easily marked. A prcedure is described here which emplys recmbinant DNA technlgy t intrduce the Eschcrichia clí þglucurnidase (GUS) gene as a ma ker int any desired yeast strain. The GUS gene was develped as a reprter gene system fr use in nematdes and mre recently in the study f plant gene expressin (Jeffersen et al., 1986, 1987). The advantages f the GUS system as a ma ker in yeast strains include its lw backgrund levels in S.cerevisíae andther yeasts assciaæd with grape must fennentatin and its ease f assay by flurimetry, spectrphtmetry and agar plate tests. 5.2 RESULTS Wine )reast chrmsme fingerprinting The TAFE gel banding patterns f twelve winemaking yeasts were cmpa ed with that f a standa d strain f S. cerevisl (Beckman334). Seven srains are shwn in Figure 5.1. The standa d strain prduced 13 bands (numbered frm bttm f the gel), althugh bands 10 and 11 were usually fused. Bands 9 and 12, which stained mre intensely, are dublets. On the basis f data frm geneúc maps f S. cerevfuiae (Mrtimer and Schild, 1985) and chrmsme physical size (Anand, 1986), Beckman have assigned the fllwing chrmsme numbers t the bands starting frm the bttm f the gel: chrmsme number I, VI,I[,IX,V[I,V,XI,X, dublet f II and XIV, XI[, XVI, dublet f VII and XV, and IV. Chrmsmes )fli and XVtr were nt detected. The largest, chrmsme fv (band l3), was estimated t be in the range f 1500 t 2500 kbp and the smallest detected, chrmsme I, at 245 kbp. The tw pairs f c-migrating chrmsmes, tr and XIV (band 9), and YII and XV (band 12) arc apparent as mre intensively stained bands.

105 Figure 5.1. Chrmsme banding pattern f AWRI winemaking yeasts prduced by transverse alternating field electrphresis. The chrmsmes fr Saccharmyces cerevisiae strain 334 are labeled accrding t data prvided by Beckman Instruments, Califrnia. Lanes: I Sacclnrrnyces cera isíae A (3s0) 3 2A (729) 4 8A (834) s 9A (J7) (R2) 7 Trulaspra delbrueckü I2A(L43) 8 Lalvin ECl118 9 Saccharmyces cerevisiae334

106 IV vii, XV xvr XIII II, XIV x XI v VIII IX III VI I-

107 S. cerevislae K. aplalata C. celliculsa Beckman 3U n I IT I I C. stellata C. krusel H. anmla S. cerevisiae WnlY12 Beckman 334 II IT T I I - === =- Ir -: r= :- =E r r r,rlr =- = Irlrrr r - == I = Figure 5.2. Diagramatic representatin f chrmsmal patterns btained frm yeast strains after transverse alternatin g field elecrphresis.

108 -80- The twelve wine strains investigated revealed banding patterns different frm that f the standard str in. Only ne strain, 124, shwed a pattern that was dramatically different frm the standard. This strain will be discussed separately belw. Three variatins are app ìfent: psitin, number and intensity f bands. Band psitin n the gel prvides an estimate f chrmsme physical size. All eleven strains shwed varius degrees f chrmsme plymrphism except fr chrmsmes IV (band 13), XI (band 7) and IX (band 4). Cnsiderable chrmsme plymrphism has been reprted ín Saccharmyces strains frm ther Jnge et a1.,1986; Casey et ai., 1988b)' The number f bands des nt necessarily crrelate with the number f chrmsmes as already indicated by the standard srrain. Mst f the wine yeasts exhibit less than 13 bands indicating that several grups f chrmsmes were nt reslved under these cnditins. Such grups f chrmsmes may be identifred by brad r mfe intensely stained bands' The banding pattern fr strain l2a Trulspra delbrucckü was fundamentally different when cmpared t thse f Sacclørmyces yeasts, suggesting a gleatly reduced number f chrmsmes. Several nn-saccharmyces yeasts islated frm fermenting grape must and wine were analysed and a e depicted diagramatically in Figure 5.2. The chrmsme fingerprints f these yeasts are generally quite different t thse f Saccharmyces strains, in particular, strains f Kleckera apiculata, Candida stellata, Candida kruseí and Hanseniaspra anmala did nt shw small chrmsmes. These results suggest that this methd may readily distinguish between Sacclarmyces and nn- S ac c lnrmyces strains GUS-Vectr cnstruct The efficiency f the E. cli ß-glucurnidase gene as a marker gene in yeasts is dependent upn maintenance f the gene in a grwing ppulatin and efficient gene expressin.

109 Figure 5.3. Cnstructin f plasmids paw218 and paw219. The 1.8kb Hind III GUS fragment was ligated with a Hind III digested AAH5 ptasmid. This ligatin gave rise t the AAH5 plasmid cntaining the GUS gene in either the wrng rientatin with respect t the prmter (paw218), r the crrect rientatin with respect t the prmter (paw219).

110 Bam Hl AAH bp a Hind lll Sph I Bam Hl Sph I Ec! Rl 2um Ec Rl Digest viltr Hind III Hitrd III GUS 1800 bp HiDl III \-,-- Bam Hl Bam Hl 4ôc, 1 Sal /Htnóttt + AOc, 1 Sal I Hind lll Ec BV Ec Rl paw bP Fv Ec Rl paw bp ^f nò.,ò Hind lll Þé p"" Hind lll Sal I Sal I Ec Rl Bam Hl Ec Rl Bam Hl Sal I Sal I ^f

111 Figure 5.4. Cnstructin f plasmid pav/220. The 3.7kb Bam Hl fragment f paw219 was ligated with the Bam Hl digested pwx509 t give rise t paw220.

112 Bam Hl 4ô, + Sal I H nd lll Ec RV Ec Rl paw bp Sal I Bam Hl Sma I xba I Hlnd lll Cla I Hind lll Ec Rl nø p" ^tê Hind lll Sal I Bam Hl Sal I pwx bp Xba I Ec Bl Ban HI \- iligest 7 Bam HI fragment Kpn I Nru I Sac I Ec ßl Xh I Stu I Xh I Nar I Ec Rt Pvu ll Bgl ll Ec RV Cla I I Bam H Sal I Ec Rl H nd ll Hind lll Xba I Tdm Ëc Rl Sal I Pst I Sph I Hind lll paw bp Pvu ll Bam Hl Sma I Xba I Hind lll Cla I Ec Rl Kpn I Nru I Sac I Ec Rl Xh I Xh I Nar I

113 -81- Expressin f the GUS cding regin was achieved by use f the yeast alchl dehydrgenase (ADCI) prmter and terminatr sequences. The GUS gene was islated as a Hind Itr fragment frm the plasmid pklg4. This was ligated int the Hind Itr site f the vectr AAH5 (Ammerer, 1983). Clnes with the GUS cding regin, in bth rientatins with respect t the ADCL prmter, were btained (Figure 5.3). A clne with the GUS cding regin in the crrect rientatin with respect t the ADCI prmter was identified (ptasmid paw2l9). Plasmid parw219 was then digested with Bam Hl. This digestin results in excisin f the GUS gene flanked by the ADCI expressin signals. The GUS cding regin with attachd ADCL signals cassette was clned int the Bam Hl site f vecrr pwx509 (Casey et a1.,1988a) - grving rise t plasmid paw220 (Figure 5.4) Transfrmatin and Suthern analysis The SMR -410 gene n plasmid paw220 is almst identical in sequence t the ILV2 gene - a single base pint mutatin leads t resistance t the herbicide sulfmeturn methyl. Therefre, sufficient hmlgy exists between the tw sequences t target integratin t the ILV2 gene via hmlgus recmbinatin with SMRl-410. Recmbinatin is enhanced by digesting plasmid paw220 with Pvu II prir t transfrmatin. This gives rise t a linea mlecule with SMRI -410 se4uences at either end. The DNA ends are highly recmbingenic and integratin is mst likely t ccur at the Pvu tr site in the ILV 2 gene. The result f this event wilt be tw ILV2 genes flanking the GUS cassette, ne f which will cntain the SMR I-410 mutatin cnferring herbicide resistance upn the transfrmed cell (see Figure 5.5).

114 -82- GUS paw220 smrl -410 Pvull {-llv 2 Pvull lnteg ratln chrmsme Xlll <-llv 2+ GUS Pvull - IEffiI {-llv 2+, Pvull chrmsme Xlll - Figure 5.5. Integratin pattern f pav/220 int yeast chrmsme XIII.

115 Figure 5.6. Suthern hybridizatin detecting the 9.7kb GUS vectr in transfrmed strain 3AM. Panel A: Hind III mlecular weight markers. Panel B: Ttal yeast DNA digested with Pvu II and electrphresed n a l% agarsegel. The gel was stained with ethidium brmide and visualized n a UV light bx. Panel C: Suthern hybridizatin perfrmed n samples depicted in Panel B, prbed with a radilabeled sequence f the GUS gene.

116 eð = c!) c) = c 23.6kb- 9.6 kb kb - 4.5kb- 2.3kb- -9.7kb A B c

117 Figure 5.7. Suthern hybridizatin depicting the chrmsmal lcatin f the GUS vectr in strain 3AM. Panel A: Yeast chrmsmes sepafated n a l% agarse gel by transverse alternating field electrphresis. Panel B: Suthern hybridizatin f chrmsme patterns depicted in Panel A prbed with a sequence f the ILyz gene. Panel C: Chrmsme patterns depicted in Panel A prbed with a radilabeled sequence f the GUS gene.

118 E ō 3 U' 3 ã I I I t 3A 3AM 3A 3AM 3A 3AM

119 -83- Pvu tr digested plasmid, p{w22},was intrduced int wine yeast smin 3A by the methd f It et at. (1983). Transfrmants were selected fr resistance t the herbicide sulfmeturn methyl (at 10 pglml). A transfrmed clny (designated 3AM) was then screened fr the presence f the GUS cnstruct. Ttal DNA was islated and digested with pvu II, thereby releasing the GUS-vectr cnstruct intact frm the chrmsmal DNA. A Suthern hybridizatin was then perfrmed using the Hind III fragment frm plasmid pklg4 t prbe fr the GUS sequence (Figure 5.6). A band f apprximately 9.7 kbp was evident in the transfrmed strain indicating the presence f the GUS cnstruct. The chrmsmal lcatin f the GUS cnstruct in the transfrmed strain was determined. Intact chrmsmes were prepared frm strain 3AM, separated by pulsed field gel elecrphresis n a TAFE unit and screened by Suthern hybridizatin with the GUS sequence. Results f this analysis (Figure 5.7) shw that the GUS cnstruct has integrated int chrmsme (XI[). This chrmsme als cntains the ILV2 gene; the expected site f integratin Develpment f a GUS assa]' fr yeasts A number f substrates are available cmmercially fr GUS detectin assays. Tw f these substrates were selected fr develpment f a GUS assay fr yeass: Ð 5-brm-4-chlr-3-indlyl-p-glucurnide(X-GLUC). Thissubstrateis available fr histchemical lcalizatin f p-glucurnidase activity in tissues and cells. The X-GLUC mlecule is cleaved by the p-glucurnidase enzyme t prduce an indxyl derivative which, upn xidatin, gives rise t an insluble and higtrly clured indig dye. ü) 4-methyl umbelliferyl glucurnide (MUG). This flurgenic substrate has been described in the literatr re fr assay f p-glucurnidase activity (Jeffersn, 1987). The cmpund is nt flurescent until cleaved by B-glucurnidase t release 4-methyl umbelliferne which is flurescent nly when the hydrxyl grup is inized. The PK" f

120 -84- the hydrxyl is between 8 and 9 but maximal flurescence will nly be btained if the prduct is in slutin at a ph greater than the PK". previus reprts f GUS activity in yeast had nt been dcumented, therefre an empirical investigatin was necessary t deærmine apprpriate cnditins fr detectin f the GUS cnstruct i Saccharmyces strains. GUS activity had been detected in filamentus fungi by grwth n agar media cntaining X-GLUC; transfrmed strains giving rise t blue clnies (Rberts et a1.,1989). An initial experiment was perfrmed t determine whether GUS activity culd be detecæd in yeast cells simply by incubatin with the substrate X-GLUC. plasmids pav/218 and pav/219 were transfrmed int the Saccharmyces lab strain O11. Bth types f transfrmants were grcwn t statinary phase in liquid culture and after harvesting, were resuspended in 0.1M Na2HPO4, ph 7.0. The substrate X- GLUC (spg/ml) was added and the suspensins were incubated at 30C. At regular intervals, the cells were checked fr develpment f blue clur. After a 24 hur incubatin, n clur was evident in either f the transfrmants. Pssible explanatins fr this negative respnse include: Ð the substrate is nt entering the cell and therefre is nt expsed t the p-glucurnidase enzymel' r ii) the ADCI-GUS cnstruct is nt directing suff,rcient expressin f the GUS cding regin and therefre, inadequate levels f the p-glucurnidase enryme exist in the yeast cell- T investigate the pssibility that the substrate is nt cming int cntact with the enzyme, the experiment was repeated but this time the yeast cell wall was physically disrupted by agitatin with glass beads. The result f this experiment is presented in Figure 5.8 (A). The psitive respnse btained after physical disruptin f the yeast strain cnraining paw219 indicates that sufficient expressin f the GUS gene is achieved with the ADH-GUS cnstruct. The psitive respnse was nly recrded with the transfrmant

121 -85- in which the GUS cding regin is riented crrectly with respect t the ADH prmter. This cnf, ms that frtuitus expressin f the GUS gene was nt sufficient t generate a detectable level f GUS activity. Furthermre, it can be cncluded that ver the time curse f this experiment, the X-GLUC was nt taken up by the cells nr culd it diffuse int intact cells. A number f ther treaenents were tested fr the ability t induce permeatin f the yeast cell wall and t facilitate detectin f GUS activity. These teatments included: Ð additin f the inic detergents sa ksyl r SDS (sdium ddecyl sulphate) t the cell suspensin (ttal cncentratins ranging frm 0 - SV); ü) additin f the nn-inic detergent tritn-x t the cell suspensin (ttal cncenfatin ranging frm 0-57); and üi) additin f the sulphydral reagent p-mercaptethanl t the cell suspensin (ttal cncentratin ranging frm mm). Cells f strain Oll cntaining plasmid pav/219 were suspended in 0.1M Na2HPO4 cntaining 50 mm X-GLUC prir t the additin f each cmpund. Each suspensin was then incubated at 37C fr 4 hurs. Results f these treatnents are presented in Figure 5.8 (B). Additin f the inic detergents sarksyl and SDS (at cncentratins f l7) resulted in the fastest clur develpment, and after 4 hurs these suspensins were the deepest blue. A decrease in rate f clur develpment was evident in samples which cntained the detergent in cncentratins higher than l%. This culd be due t a partial denaturatin f the p-glucurnidase enzyme, which wuld result in a decrease in enzyme acúvity. Additin f tritn-x was als successful in inducing leakiness in the yeast cells, hwever, clur develpment was slwer with this treatment. Treatment f the suspensins with p- mercaptethanl (up t 100mM) was nt succesful, in the time curse f this experiment, in facilitating detectin f GUS activity. It can be cncluded, therefre, that treaünent f cells

122 Figure 5.8. Visual assays t determine suitable cnditins fr the detectin f GUS activity in yeast cultures. Panel A: Saccharmyces cerevisiae srain OLl transfrmed withpaw2ls and parü/219. Cells were suspended in 0.1M Na2HPO4, ph 7.0 and vrtexed fr five minutes with glass beads. Panel B: Sacclørmyces cerdísiae strain Ol1 transfrmed with pav/219. Cells were suspended in O.lM Na2HPO4 and substances were added as indicated.

123 A lt pawlle pawr 19 B sarksyl ì p-mercaptethanl 0mM 10 mm 50 mm 100 mm

124 -86- with an inic detergent (sarksyl r SDS) at a cncentr tin f l7 induces permeability f the yeast cell and allws detectin f GUS activity in marked cells. The MUG substrate was then investigated as a ptential detectin system fr ma ked yeasr strains. Cell suspensins f 3A and 3AM were prepared in 0.1M Na2HPO4, ph 7.0' 17 sa ksyl and lmm MUG, and incubated with shaking at 37C. After 2 hurs incubatin, the suspensins were viewed n a shrt wave UV light bx (Figure 5.9). Flurescence is easily detected in the 3AM suspensin, while strain 3A gives a negative respnse. The ptential use f a marked strain in the wine indusury relies n the ability t quantify the marked cells in a mixed ppulatin f yeass. Therefre, a prcedure in which individual cells can be assayed fr GUS activity is required. In light f the bservatins described abve, ne premise is that such a prcedure will invlve a tificial inductin f permeability f the substrate t the yeast cells. The direct cunting f clnies n agal medium is an accurate methd fr quantifying yeast cells and was chsen as the basis f an assay prcedure. A methd was sught which disrupted the yeast clny permeability barrier, withut disturbing the distinct clny frmatins (and therefre the cunting f these clnies). This was achieved by allwing clnies t grw n nutrient agar media cntaining X-GLUC (sopg/ml). The clnies were then verlayed with a mlten 0.87 agarse slutin (at apprximately 45C) cntaining 0.1M Na2HPO4, ph 7.0 and l7 sarksyl. V/ith a thin verlay, the agarse slutin slidified within 2 minutes, and clnies remained intact and distinct. The sarksyl present in the verlay induced leakiness f the yeast cell walvmembrane and after 4-6 hurs incubatin at 37C, clnies f the ma ked strain appeared blue. Results f this assay are depicted in Figure Mixed ppulatins f yeast cells cntaining va ius prprtins f the marked strain 3AM were prepared as fllws. Srains 3AM and 3A were grwn separately in liquid culture t statinary phase where similar

125 3AM Figure 5.9. Flurescence assay fr detecting GUS activity. Yeast cell extracts cntaining lmm MUG were incubated at37c fr tw hurs and the visualized n a UV light bx.

126 Figure Agar plate assays t detect GUS activity in a yeast ppulatin cntaining 0,20,50 and l00v strain 3AM.

127 Figure Micrscpic visualisatin f GUS activity in cells f strain 3AM.

128 -87 - ptical density and cell number per ml were achieved. These snains were then mixed tgether in different ratis (vvvl), resulting in a range f mixed ppulatins in which the percentage f strain 3AM was 0, 20, 50 and L0O7. Samples frm each ppulatin were serially diluted in saline slutin (O.857 w/v) and plated nt YPD media cntaining X- GLUC (50 mg/ml) befre being assayed as described abve. In each case, the percentage f blue clnies per plate was apprximately equal t the percentage f the 3AM culture (vvvl) in the prepared mixed ppulatin. An attempt was als made t visualize GUS activity at the cellular level. A suspensin f 3AM cells cntaining 0.lM Na2HPO4pH 7.0, 17 sarksyl and X-GLUC (50 pglml), was incubated at 3Tcuntil a deep blue clur had develpeà Q-a hurs). A sample f this suspensin was then viewed under a micrscpe (Figure 5.11). The blue precipitate was detected inside the ma ked cells. This result indicates that mixed ppulatins f yeasts can be analysed fr presence f the marked smin within tw t fur hurs f sampling ÈGlucurnidase enzyme activity The p-glucurnidase activity in the marked strain was measured using the flurgenic substrate 4-methyl umbelliferyl glucurnide (MUG) (Jeffersn, 1987). The detectin f flurescent mlecules ffers a very high signal-t-nise rati because the incident excitatin light des nt impinge n the detectin apparatus, and has a spectrum distinct and separable frm that f emissin. The use f flurescence measurements t detect enzyme activity allws tw t fur rders f magnitude greater sensitivity than methds that rely n spectrphtmetric determinatin f prduct cncentratin. Preparatin f yeast cell extracts fr enzyme assays invlved vrtexing a suspensin f cells in GUS exrractin buffer (50 mm NazHPO, 10 mm p-mercaptethanl, 10 mm Na2EDTA,Q.lft sdium lauryl sarcsine, 0.17 tritn X-100) with glass beads (t half the

129 -88- vlume). Thrughut the vrtexing, the cell suspensin was peridically placed n ice t avid verheating. A time cgrse experiment was perfrmed t determine the minimum perid f vtexing required t release the mar<imum amunt f prtein frm the cells' At va ius intervals samples were remved frm the vrtexing suspensin and cell debris was pelleted by centrifugatin. The supernatant was then stred fr analysis f prtein cntent' Results f this experiment afe presented in figure The maximal release f prtein frm the cell suspensin was achieved after 2O minutes f vrtexing' 450! 6 I Ê õ À 300 TD tlmc (mlns) Figure Time curse analysis f the annunt f prtein released frm a suspensin f yeast cells after vrtexing with glass beads' In rder t analyse p-glucurnidase activity, strains 3AM and 3A were grwn t early statinary phase in liquid YPD media. Cell extracts were prepared as described abve by vrtexing with glass beads fr 20 minutes. Bth 5 and 50 pl samples f exeact were assayed using the MUG substrate as described in Sectin At the same time a standard curye f 4-methylumbelliferne (MU) was prepared, s that the relative flurescence readings frm the extract samples culd be interpreted as cncentratin f MU and then cnverted t nmles f MU prduced. This experiment was perfrmed twice and

130 -89- average readings are presented graphically in Figure The prtein cntent f each extract was fund t be 5 mg/mt fr strain 3AM and 4.4 mg/rnl fr the 3A extract' N enzyme activity was detected in strain 34, while the average p-glucurnidase activity f strain 3AM was calculated t be 0.65 nmles MU/mirVmg prtein' 25 E (Ð I! È À ttt ( ) E É time (mins) Figure p-glucurnidase enzyme assays: (O) 50 rl strain 3AM extract; (O)spl strain 3AM extract; (tr) 50 pl strain 3A extract Fermentatin trials The 3AM strain was used in a fermentatin trial t mnitr the effects f transfrmatin n the yeast enlgical prperties. Three different transfrmants were islated and used t inculate separate starter cultures. Three clnies f cntrl, un- Eansfrmed 3A yeast were als inculated int sta ter cultures. Each f these six cultures was inculated in duplicate int flasks f Riesling grape juice at a cncentratin f 4 x 106 cells/ml. Fermentatins were carried ut under anaerbic cnditins at 18C. Samples were taken at regular intervals and assayed fr yeast grwth (by measuring ptical density

131 -90- at 650 nm) and prgress f fermentatin (by refractive index). Refractive indices were averaged and the tw resulting fermentatin 200 X ( ) E (9 I ãl (Ð time (hurs) 3A (). Figure Fermentatin cuwes f the marked str in 3AM (EI) and parent strain Wine analysis Yeast strain Statistical analysis 3A 3AM V.R. (Ft,s # F. Prb.* ph Residual sugars (g/l) SO2 (ttal) (mg/l) SO2 (free) (mell) Acetic acid (s/l\ 3.32 r s # Va iance rati beween the tw strains. * F. Prb. indicates the prbability f the assciated variance rati. Table 5.1. Analyses f wines prduced frm the riginal yeast strain (34) and the ma ked strain (3AM).

132 -91- curves were pltted (Figure 5.14). A lg transfrmatin f the data indicated that the V.R. (Fl,99) is 1.01, and the assciated F prbability is Therefre, statistical analysis has revealed n significant difference between the fermentatin curves f seains 3A and 3AM. On cmpletin f fermentatitr, ph, residual sugars, sulfur dixide, alchl and acetic acid cncentratins were measured fr all Welve samples. Averages were calculated fr each par rmeter (table 5.1). Again, n significant differences were evident between the tw strains Stability f GUS cnstruct An experiment was perfrmed t measure the stability, r maintenance, f the GUS cnstruct ver successive generatins f strain 3AM. Under fermentatin cnditins there is an absence f selectin fr the GUS gene, therfre it is imprtant t determine the spntanus rate f lss f the gene in a dividing ppulatin. An vernight culture f strain 3AM was generated frm a single clny. This culture was then diluted by a factr f 10-3 in a sub-culturing prcess (by inculating 100 pl int 100 ml YPD) and grwn t statinary phase. This sub-culturing was repeated seven times. Samples were remved at statinary phase f each sub-culture and assayed fr GUS activity by the agar plate methd. The number f blue and white clnies were scred and recrded in Table 5.2. An instability f the gene in less than l7 f the ttal ppulatin was recrded at each sampling and did nt increase signifrcantly ver time.

133 -92- Subculn re Dilutin effect Blue clnies White clnies Ttal clnies 4 white clnies I ó 7& e I r l Table 5.2. Stability test f GUS cnstruct in strain 3AM after successive sub-culturing. The agar plate GUS assay was perfrmed at each sub-cultu e step t detect the number f negative (white) clnies.

134 DISCUSSION Clse examinatin f pulsed field gel electrphresis fingerprints f twelve wine yeasts indicate that nne f the strains have identical electrphretic chrmsmal prfiles. Similar results have recently been reprted by Vezinhet et al. (1990). In an analysis f 22 wine yeasts islated frm varius wine grwing regins, nly 3 strains (riginating frm the same vineyard) culd nt be differentiated. These differences in prfiles result frm chrmsme rearrangements which have taken place in yeast strains during the curse f evlutin. The high degree f plymrphism amng wine yeast strains is perhaps nt surprising in view f their diverse rigins. It is likely that the strains hetd in The Australian'Wine Research Institute Yeast Cllectin have undergne reprductive islatin fr sme time by virtue f their lcalizatin in different viticultural regins f the wrld. Pulsed field gel electrphresis can be used t gain a better understanding f the prcesses which generate chrmsme-length plymrphisms between strains. Fr example, the FIGE and OFAGE systems have been used t shw that certain chrmsmelength plymrphisms segregate in a2:2rai,indicating single structural alteratins f the chrmsmes (On and Ishin-Ara, 1983). Chrmsme-length plymrphisms, hwever, can als result frm tw r mre structural alteratins per chrmsme and are nt restricted t specific chrmsmes. The TAFE system has been used fr the analysis f chrmsmal segregants and inheritance (Bilinski and Casey, 1989), and Viljen et ai. (1989) used the OFAGE system t establish pssible amamrptr/telemrph relatins in yeasts. Anther applicatin f pulsed-field gel electrphresis is the lcalizatin f specif,rc genes and the distinctin f tw yeast strains that differ nly in the chrmsmal lcalisatin f specific genes. Pretrius and Marmur (1988) prbed a Suthern blt f an OFAGE gel cntaining the reslved chrmsmes f fur S. cerevisi e strains with the

135 -94- clned STA2 glucamylase gene and shwed that the strains differed nly in the presence and/r ch msmal lcatin f the gene. Results btained with chrmsme fingerprinting f wine yeasts reveal an immediate applicatin f TAFE technlgy t Resea ch and Develpment and Quality Cntrl in wine fermentatins. At fxed pulse and electrphresis times the pattern fr any strain is highly reprducible and results in a unique signature easily distinguished frm ther cntaminating strains. The relatively rapid technique f electnphretic karytyping f yeasts may therefre be applied t imprvement f yeasts by clnal selectin during successive runds f fermentatin. This methd wuld ensu e that cntaminant strains f S. cerevisiac were nt inadvertently selected. P cesses which cause chrmsme damage culd.als be identif,red using this techmique (Cntpulu et a1.,1987). The effect f envirnmental selectin pressures (such as strage cnditins r culture media) n the genmes f wine yeasts may als be determined. Such infrmatin culd have significant impact n the preparatin f starter cultures and the mde f strage f wine yeasts. In additin, in develping new strains fr industrial applicatins, TAFE analyses can be useful fr mnitring chrmsme prfiles f parent and prgeny strains emplyed in such research. Fingerprints that examine all chrmsmes within a yeast strain can als be viewed as genetic 'snapshts' f that yeast at a fixed pint in time and therefre can be used t mnitr chrmsme stability. This system is, therefre, useful fr rutinely mnitring the stability f a yeast strain used within a winery as well as fr distinguishing srains used in different wineries. Hwever, fr quantitative studies f yeast grwth kinetics during fermentatin where large numbers f cells need t be mnitred fr statistical analyses, the TAFE identificatin system is unsuitable. The time invlved in the preparatin f plugs, and the number f gels required fr the analysis f large numbers f clnies precludes the TAFE system frm this applicatin. Fr this reasn, the GUS marking system was develped. Results presented in this chapter shw that expressin f the E.cli GUS gene can be

136 -95- achieved in Sacclnrmyces strains by a cnstruct in which the GUS cding regin is linked t the yeast alchl dehydrgenase (ADCI) prmter and teminatr sequences. The cnstruct was successfully targeted t the site f the ILV2 gene n chrmsme XIII f wine yeast strain 3A. Fermentatin tials indicaæd that the transfrmatin prcedu e did nt significantly alter vinificatin prperties f wine yeast strain 3A. Cmmercial ferments wuld need t be ca ried ut t cnfirm that the üansfrmatin prcedure des nt prduce rganleptic changes in the wine. The transfrmatin system used t intrduce the cnstruct has been used successfully t mnsfrm a range f strains in The Australian Wine Research Institute yeast cllectin (see Sectin 4.2.2). The implicatins are, therefre, that the GUS marking system can be used t tag any wine yeast strain f chice. In this respect the GUS cnsntct ffers a unique marking system fr wine yeasts. The marked enlgical strain Kl (Vezinhet and Lacrix, 1984; Vezinhet, 1985) was develped by selecting natural mutants in a ppulatin f the Latvin V yeast. This prcedure is a time cnsuming and nn-specific methd f genetic manipulatin and, therefre, it is nt readily applied t wine yeasts f interest. Assaying fr the GUS ma ker can be achieved by flurimetry, spectrphtmetry, r by agar plate methd. Althugh natural transprt f X-GLUC r MUG substrates did nt ccur acrss yeast cell membranes in the time curse f experimens described here, this prblem was vercme by inducing artificial permeatin in assay prcedures. Analysis f a large sample f clnies during the successive sub-culturing f strain 3AM revealed an instability f the GUS cnstruct. This instability was detected by the presence f white clnies in the agar plate assay prcedure. The frequency f clnies which respnded negatively t the GUS plate assay was always less than l% f the ttal plate cunt and did nt increase significantly ver the perid f sub-culturing. Occassinally a sectred clny was detected, suggesting either excisin f the gene by

137 Figure Theretical steps invlved in the develpment f a GUS-SMR1 cnstruct fr gene replacement. Plylinkers culd be added t the 3.7kb GUS cassette (derived frm paw2l9) t prduce flanking Kpn 1 and Xh 1 ends. This cassette culd then replace ttre 2.3kb Xh 1-Kpn I fragment in the nn-essential regin f the SMRI gene (Xia and Rank, 1989). The resulting plasmid wuld cntain the GUS cassette flanked by SMRI sequences. A Cla 1 + Hind III digestin wuld then generate a GUS-SMRI cnstruct suitable fr gene replacement.

138 ,*; dl' g d' Etr Dc EY Xf I sërr ChI pbr.322 scqnênce5 Ec Pvr II Ec EI Ei.ra m E ri EI ûigest vittr Kpn I, Xh I aligest Yith Kpn I, Xh I Ð,I Cl I GUS cassetæ Elr Ec SI IRT abr322 teqlrêdees Pvr II Ec PI Ei f III t Yilh Clû I, HiDn III Cl I Xf I Ec t r I PI Pvr If Ec EI Ei,1 Itr GUS ca set'te SI.iRt

139 -96- hmlgus recmbinatin (Struhl et al.,1979) r lss f the gene after mittic crssing ver (Reder et a1.,1988). The implicatins f this instability f the GUS gene require cnsideratin. In mixed-ppulatin kinetic studies invlving a GUS-marked strain, the prprtin f the marked strain will be underestimated due t the 'false negative' cells which have lst the ma ker gene. This underestimatin can be crrected if the frequency f instability is calculated by apprpriate cntrl experiments' The instability detected in this chapter indicate that statistical analysis f herbicide resistant clnies perfrmed in sectin was nt sensitive enugh t detect a <17 lss f the integrative vectr. That experiment culd be imprved by analysing a larger number f clnies. Furthermre, determinatin f the prprtin f herbicide resistant clnies may be better achieved by replica-plating clnies frm nn-selective media directly nt selective media. Lss f the gene thrugh excisin by hmlgus recmbinatin may have been avided by adpting an alternative methd f targeting the GUS cnstuct t the ILV2 gene. Fr example, instead f intrducing the plasmid palv,i22l by hmlgus recmbinatin, the GUS cnstruct culd be integrated t the genme f yeast by the prcess f gene replacement (Rthstein, 1983). It wuld be pssible t clne the GUS cnstruct int a nnessential regin f the SMRI cding regin (Xia and Rank, 1989), thus prducing a chimaeric GUS-SMRI cassette where the GUS cnstruct is flanked by SMRI upstream sequences. The release f this cassette by restrictin enzyme digestin wuld generate a fragment with bth recmbingenic ends hmlgus t yeast chrmsme XtrI DNA at the ILV2lcus. Transfrmatin f yeast with this fragment wuld result in integratin by gene replacement at the ILV2 lcus (Figure 5. 15). The use f this methd t intrduce the GUS cnstruct t yeast strains ffers the prspect f greater stability. In determining suitable prcedures fr the intrductin f a stable GUS cnstruct in wine yeasts, it is als imprtant t cnsider the phenmenn f mittic recmbinatin and

140 -97 - gene cnversin (Zimmerman, 1990). This surce f instability in wine yeasts can be vercme by a prcedure which invlves sprulatin and islatin f clnes frm single ascspres. The haplid hmthallic spre diplidizes shrtly after germinatin giving rise t a cmpletely hmzygus diplid. A prblem with this technique, hwever lies in pr sprulatin ability f wine yeasts, and in the labrius selectin prcedures invlved in btaining a spf with identical prperties t the riginal parcnl The GUS system described in this chapter will enable a wide range f yeast strains t be marked, prviding the means fr unequivcal identificatin and mnitring during fermentatin. Recent investigatins by Heard and Fleet (1985,1986,1988) made with bth inculated and uninculated grape juices under a range f fermentatin cnditins suggest that Sacclarmyces strains are nt necessarily the dminant rganism during vinificatin. These studies, alng with a lack f knwledge regarding incidence and imprtance f wild strains f Saccharmyces in fermenting grape juice, call fr detailed mnitring f inculated strains under va ius enlgical cnditins. The agar plate methd will find applicatin in the mnitring f sta ter cultures, effect f prefermentatin prcesses n the micrbial ppulatins f must, and the efficiency f inculatin. Furthermre, the cell suspensin assay has the advantage f btaining results frm the starter culture r ferment ppulatins within tw t fur hurs f sampling. It is imprtant t nte, hwever, that the GUS system has been develped using recmbinant DNA technlgy and utilizes a gene derived frm E. cli. Cnsequently, yeast strains marked with the GUS gene are nt immediately available fr use in the Australian Wine Industry. It is envisaged, hwever, that public perceptin f genetically engineered rganisms and legislatins limiting their release will be less restricting ver time. The benefits f recmbinant DNA technlgy have been widely reprted and micrrganisms have been successfully explited fr the prductin f a range f pharmaceutical prducts such as interfern, insulin and human g wth hrmne (Blm, 1980; VÍdruff, 1980). Genetically engineered rganisms are being cnstructed fr a variety f envirnmental

141 -98- applicatins. These include their use in agriculture as pesticides (Bishp et al-,1988); r fr agrnmic crp prductin; pllutin cntrl f txic waste in land filled sites, wastewater trearment facilities r after accidental spillages; and in mining and the petrchemical industry fr enhancing it and mineral recvery (Keeler, 1988). A genetically engineered agrbacterium which cntrls crwn gall in plants has recently been released in Australia (Kerr, 1989). These advances will lead t a mfe infrmed general public n matters invlving genetically engineered rganisms and, perhaps, will smth the way fr thers. In cnclusin, tw systems fr the identificatin f yeast strains have been described. Advantages f TAFE system a e that it facilitates a better understanding f prcesses invlved in yeast genme rea Tangements, and can be used t mnitr chrmsme stability in wine yeasts. The GUS system enables the mnitring f specifrc wine yeast strains in cmmercial fermentatins which cntain an unknwn lad f indigenus yeasts.

142 -99- Chapter 6.Determinatin f kitler txin activity in fermenting grape juice using a marked Saccharmyces strain 6.1. INTRODUCTION Killer activity in yeasts was first reprted in strains f S. cerevisiae in 1963 by Bevan and Makwer. Killer yeasts secrete pl peptide txins which kill sensitive strains f the same genus and less frequently, strains f different genera (Philliskirk and Yung, 1975; Tipper and Bstian, 1984). Previus studies indicate that the txin f Sacclnrmyces is a prtein which binds t a receptr n the cell wall f the sensitive yeast, disrupting the electrchemical gradient acrss the cell membrane and hence the intracellular inic balance (De la Pena et a1.,1981; Skipper and Bussey, 1977). Prductin f the txin and immunity t it are determined by a cytplasmically inherited duble stranded (ds) RNA plasmid, therwise knwn as the M-genme (Bstian et al., 1980) which is fund nly in cells cntaining an additinal dsrna species designated the L-genme. Bth types f dsrna exist in virus-like particles and require a prtein encded by the L-dsRNA fr encapsidatin (Bstian et a1.,1980; Ha ris, 1978). Based upn prperties f the txin, killer yeasts have been classified int eleven grups (K1 thrugh Krr) (Naumv and Naumva,1973; Yung and Yagiu, 1978). Thse unique t Saccharmyces fall int the first three (K1, K2 and K ). The Saccharmyces txin is reversibly inactivated at lw ph (2.0) and irreversibly inactivated at ph in excess f 5.0 (Yung and Yagiu, 1978). Mre specifrcally, the bilgical activity f K1 is ptimal berween ph 4.6 and 4.8, while K2 shws ptimal activity between 4.2 and 4.7 (Shimazu er ø/., 1985). Cmpared with K1, the K2 txin is stable ver a wider ph range (2.8 t 4.8) (Rgers and Bevan, 1978) and is therefre mre relevant in wine fermentatin.

143 -100- Killer activity has been detected in yeasts islated frm established vineyards and wineries in varius regins f the wrld including Eurpe and Russia (Barre, 1984; Gaia, 1984; Naumv and Naumva,1973), Suth Africa (Tredux et a1.,1986) and Australia (Heard and Fleet, 1987a,b). This widespread ccu ence has prmpted interest in the enlgical significance f killer wine yeasts. In thery, selected kitler yeasts culd be used as the inculated strain t suppfess grwth f undesirable wild strains f s. cerevisiae during grape juice fermentatin. In additin, as killer interactins have been reprted t ccur between yeasts f different genera (Radler et al., 1985; Rsini, 1985a), the pssibility exists t genetically engineer brad spectrum killer strains f S. cerevisiae (Bne et a1.,1990). Studies have been cnducted t assess the efficiency f killer txin n sensitive yeast strains. Hwever, reprts have been cntradictry n the expressin f killer activity under fermentatin cnditins (Cuinier and Grs, 1983; Delteil and Aizac, 1988; Lafn- Lafurcade and Ribereau-Gayn, 1984). Attempts t determine the ppulatin kinetics f killer and sensitive strains during wine fermentatin have been restricted because f the difficulty invlved in identifying the tw types when grwn in mixed cultures. Appraches used t date include i) chice f killer and sensitive strains that can be distinguished by their grwth rates (Bare, 1984) r prductin f hydrgen sulf,rde (Rsini, 1985b); ii) use f auxtrphic and respiratry deficient mutants f killer strains and apprpriate plating cnditins under which they can be identified (tlara et a1.,1980, 1981; Seki er ai., 1985); iii) use f killer and sensitive strains which can be distinguished by differences in clny mrphlgy (Heard and Fleet, 1987a); and iv) assaying clnies directly fr killer activity (Lng et a1.,1990). All f these methds a e limited by the fact that the assays invlved are labrius and time-cnsudng, r that nly killer strains with specific characteristics can be studied.

144 -101- wat'te lnstttule ldn BY This chapter describes the use f a marked S. cerevísiø killer strain in a mixed culture inculum t quantify directly the effect f killer txin n a sensitive s- cerevisiae strain under fermentatin cnditins. As a wide range f yeast strains can be readily and stably marked, this sysæm f analysis is unlimited in applicatin and prvides a simple and unequivcal nrcans f quantifying killer yeast strains in mixed culture ferments. 6.2 RESULTS Curing f Strain 3AM In rder t specifically analyse the effect f killer txin in fermentatins, an experiment was designed t cmprire tw isgenic strains which differ nly in the presence f the M-dsRNA genme and therefre, in their ability t prduce killer txin. Killer strain 3AM has previusly been ma ked with the Escherichia cli þ- Glucurnidase (GUS) gene (see Chapter 5). This system allws the ma ked strain t be readily identified in a mixed ppulatin by a simple plate assay which esults in the frmatin f a blue precipitate in ma ked clnies. Strain 3AM was cured f its M-dsRNA pla.smid by heat treaünent (wickner, lg74), the cured r sensitive clnies being identifred by kilter plate assays. Figure 6.1 shws the respnse f snain 3AM and an islated cured derivative (designated 3AMC) t rhe killer plate assay. The zne f inhibitin clearly evident arund strain 3AM is absent arund 3AMC, indicating that strain 3AMC is nt prducing killer txin. As stain 3AMC is derived frm 3AM, it inherits the GUS gene and therefre is als a ma ked strain. Duble stranded RNA species were islated frm strains 3AM and 3AMC and analysed by standard electrphresis techniques (Figure 6.2). A band representing the M- dsrna genme is present in srain 3AM, and absent in strain 3AMC.

145 114 3AM 3AMC ph Figure 6.1. Agar plate assey fr killer actívity. The agar (pii4.2,a.ffi:3%a methylene blue) is seeded with an vrnight culture f strain 3AMC, and strains t be tested fr killer activity are patched rìt the slid media. 11A is a knwn killer strain, and 2A is a knwn sensitive stain.

146 e -.1 CD þ cn CÞ c)?. z 't) N) ', 7 '-l (. )? a '-ì A 7 -.'â z I AA äz F> z À Hind III markers 1tA 2A 3AM 3AMC

147 -102- Fermentatin trials were then perfrmed n strains 3AM and 3AMC t determine the effect f the curing prcedure n yeast grwth and fermentatin rates. Starter cultures f each strain were inculated in triplicate int flasks f Riesling grape juice at a cncentratin f 5 x 106 cell per ml. Samples were taken at regular intervals and asssayed fr yeast grwth and prgress f fermentatin. The average readings fr each strain were pltted ver time (Figure 6.3). There are n significant differences in the g wth r fermentatin rates between strains 3AM and 3AMC Anal) sis f killer activity during fermentatin Strains 3AM and 3AMC were analysed fr killer activity in Riesling juice by cinculating each strain with the sensitive Saccharmyc J strain 54. Cntrl ferments f each strain (34M, 3AMC and 5A) as pure inculums were als perfrmed. Each ferment was cnducted in duplicate at 18C with gentle agitatin under anaerbic cnditins. GUS plate assays were then perfrmed t identify the ma ked strain (3AM r 3AMC). Clnies f the marked strain turn a deep blue clur as a result f this assay, allwing simple identificatin. GUS plate assays were als perfrmed n the cntrl ferments t cnfirm the validity f the assay. Plate assays n the cntrl 5A ferment were cnsistently negative' highlighting the absence f backgrund GUS activity in natural yeast cells. Hwever, cntrl 3AM and 3AMC ferments gave values f between 99 - laùv f ttal clnies per plate fr the marked strain cunt. This level f instability crrespnds t that determined in Chapter 5. The backgrund reversin frequency was taken int accunt thrughut the analysis.

148 -103- C' r (D E ' ) Tlmc (hurs) x! s È E Tlme (hurs) Figure 6.3. Yeast grwth (Panel A) and sugar utilisatin (Panel B) curves f strains 3AM (O) and 3AMC ().

149 l r 15! ' Et Tlmc (hurs) 20 u) t 15 E ì g, Tlme (hurs) 300 Figure 6.4. Panel A: Grwth curves f cntrl single mnculture ferments. Symbls: O 3AM; O 3AMC; tr 54. Panel B: Grwth curves f mixed culture ferments. Symbls: a 3AM and 5A at an inculum rati f 2:l; O 3AMC and 5A at an inculatin rati f 2:1; I 3AM and 5A at an inculum rati f 1:1; A 3AMC and 5A at an inculum rati f 1:1.

150 -105- The fllwing mixed culture ferments were carried ut: Ð 3AM and 5A at an inculum rati f 1:1; ü) 3AMC and 5A at an inculum rati f 1:1; üi) 3AM and 5A at an inculum rati f 2:1; and iv) 3AMC and 5A at an inculumrati f 2:1. These mixed ferments exhibited nrmal grwth kinetics, as did the three cntrl ferments (Figure 6.4). The time curse f grwth (clny frming unis per ml) f each strain in the mixed culture ferments is pltted in Figure 6.5. At inculum ratis f 1:1, there was a ntable increase in the prprtin f killer strain 3AM, whereas the cured strain 3AMC fails t exert any dminance ver the sensitive strain under therwise identical cnditins. Statistical analysis was used t test the null hypthesis that the rati f killer : sensitive cells remains 1:1 thrughut the ferment. A gdness f fit test (nrmal test) rejected the null hypthesis, with p-value << Hwever, identical analysis f the cured : sensitive strain ferment accepted the null hypthesis that the rati f the tw smins remains at 1:1 thrughut the ferment. With an increased prprtin f strains 3AM and 3AMC in the inculum (rati 2:1), the dminating effect f strain 3AM was mre prnunced, whereas strain 3AMC again shwed n apparent change in prprtin ver time. The dminance f strain 3AM in mixed culture ferments is illustrated mre clearly when the percentage f each strain is pltted ver the time f the ferment (Figure 6.6). Fr an inculum rati f 1:1', 3AM increased t apprximately 807 after 3 days but fr a higher inculum rati f 2:1, 3AM eventually accunted fr 977 f the ttal yeast ppulatin. It is imprtant t nte that the süain 5A persisted, albeit at lw levels, thrughut the ferments. Experiments were cnducted t determine the lwest inculum rati f killer t sensitive cells at which significant killer activity can be bserved. Mixed ferments f strain 3AM and 5A ar inculum ratis f 1:2 and l:4 respectively were carried ut

151 Figure 6.5. Grwth curves f each srain in mixed-culture ferments expressed as cfu per ml. (A) Mixed ferment f 3AM (O) and 5A (tr) at an inculum rati f 1:1. (B) Mixed ferment f 3AMC (O) and 5A (tr) at an inculum rati f 1:1. (C) Mixed ferment f 3AM (O) and 5A (tr) at an inculum rati f 2:1. (D) Mixed ferment f 3AMC (O) and 5A (tr) at an inculum rati f 2:1.

152 â E f 108 ß7 A = E J B l 106 Et 106 a.9 6 a.!, 6 t 10" rc Tlme 200 (hurs) 300 Ê,.9 f CL. at, E 10" rc4 103 ß Time 200 (hurs) 300 = E f (, 108 ß7 c = E f (, 108 ß7 D (,) 9 c I 6.. a (! 106 E 10" ß Time 200 (hurs) t c,.9 (ú J - - al, (ú 106 t rc Time (hurs)

153 Figure 6.6. Prprtins f each strain in mixed-culture ferments expressed as percentage f the ttal yeast ppulatin. (A) Mixed ferment f 3AM (O) and 5A (D) at an inculum rati f 1:1. (B) Mixed ferment f 3AMC (O) and 5A (tr) at an inculum rati f 1:1. (C) Mixed ferment f 3AM (O) and 5A (tr) at an inculum rati f 2:1. (D) Mixed ferment f 3AMC (O) and 5A (tr) at an inculumrati f 2:L.

154 Yeast ppulatln (1" ttal) õ Yeast ppulat n (/" ttal) tu50r@, :. -: 3 = c -9ru :f c *9p G) C ) Þ Yeast ppulatin (V" ttal) Yeast ppulatln (% ttal) rþq)5( ø>te õ tuþdr@ O -: 3 - -: 3 t c il ru t c i9ru G) ( )

155 Ê I ql f. - E s tlme 200 (hurs) 300 (."_ Figure 6.7. The time cuße in the prprtin f killer strain 3AM in the ttal ppulatin f a mixed culture ferment with strain 5A fr different inculum ratis. Symbls: rati 3AM t 5A tr 2:l; O 1:1; I t:2; ^ l:4-

156 under cnditins described abve. N change frm the initial prprtin f strain 3AM was detected in either f these ferments. The results f all mixed culture ferments invlving stain 3AM are summarised in Figure DISCUSSION The GUS marking system has enabled a direct cmparisn t be made between the inculatin efficiency f a killer strain (3AM) and a cured isgenic derivative (3AMC) in fermenting grape juice. At a rati f killer t sensitive cells f 1:1 the cured strain 3AMC remained at 50V f the ttal ppulatin while the killer strain increased t 807. The ability f strain 3AM t dminate 5A during fermentatin is likely t be due t the prductin f killer txin by strain 3AM and nt t a difference in respective gwth rates favuring the killer strain. We can cnclude, therefre, that the killer txin has displayed significant activity under these fermentatin cnditins. This result is f parricular interest t the enlgist since the K2 txin prduced by strain 3A is reprted t shw maximum activity at ph 4.2 (Rgers and Bevan, 1978), which is 0.5 t I ph unit higher than generally fund in grape musts. In cases where killer activity has been reprted in fermenting grape juice, a discrepancy exists as t whether effective killing actin ccurs when the prprtin f killer cells is less than 507 f a mixed culture ferment. Heard and Fleet (1987a) did nt bserve killer actin when the rati f killer t sensitive cells was apprximate y l:7 whereas thers have reprted killer activity with killer t sensitive cell ratis f 1:10 and lwer (Bane, 1984; Hara et al., 1980,1981). Our results shwed that an increase in rati f killer t sensitive cells t apprximaæly 2:l resulted in a prnunced dminance f the fermenøtin by strain 3AM t 97V f the ttal mixed ppulatin by the end f the fermentatin. Hwever, with killer t sensitive cell ratis f 1:2 r 1:4, n effective killer actin was evident. It is pssible that differences in either cmpsitin f medium, fermentatin

157 -108- cnditins r strain sensitivity may accunt fr discrepancies in reprts f killer txin efficiency. The relevance f killer strains in wine making has been the fcus f attentin in cuntries where selected yeast cultures are inculated int musts t induce fermentatin. This fcus has intensified since the bservatins that yeasts which a e naturally present in the must als play significant rles in suppsedly "pure" culture fermentatins (Heard and Fleet, 1985; Lafn-Lafurcade and Ribereau-Gayn, 1934). These natural yeasts include species frm the genera Kleckera, Candida, Hansenulaandsaccharmyces. Killer Saccharmyc s wine yeast strains may be effective in suppressing natural Saccharmyces yeasts during fermentatin and the pssibility exists t engineer brad range killer yeasts t cntrl strains frm ther genera. Fr these reasns, further study is needed t determine apprpriate fermentatin cnditins fr effective killer activity. The GUS marking system prvides a methd which allws a brad range f killer strains t be rapidly and unequivcally identified in a mixed culture. This system can be emplyed t gain a better understanding f killer activity during fermentatin.

158 -109- Chapter 7 Cmparisn f fermentatin cnditins by use f a marked strain 7.1 INTRODUCTION One f the mst significant technlgical advances in winemaking has been the cmmercial availabitity f selected yeasts, usually strains f S. cerevisiae, fr inculatin int the juice (Reed and Nagdawaithana, 1988). It is assumed that the inculated,s. cerevisiae will suppress and utgrw the indigenus yeasts and dminate the fermentatin. Althugh this assumptin is widely accepted, there is little dcumented evidence f its validity. In studies cnducted at several Austalian wineries, Hea d and Fleet (1985) shwed that grwth f Kleckera apiculata and Candida species was nt suppressed in fermentatins inculated with cmmercial strains f S. cerevisiae. Similar bservatins have been reprted by Martinez et al. (1989). Therefre, it can be cncluded that inculatin f grape juice with a high ppulatin f S. cerevisiae will nt necessarily prevent grwth f indigenus nn-s ccharmyce.r yeasts. The assumed dminance f inculated strains f S. cerevisiae ver indigenus strains f this species, has als been called t questin, highlighted by the fact that cnventinal cultural techniques fr identifying wine yeasts d nt distinguish between strain types. Using electrphretic methds t distinguish between stains f S. cerevisiae, Buix et al. (1931) shwed that the inculated strain did nt always dminate the fermentatin. Ma tinez et ai. (1989) used hydrgen sulfide prductin as a marker t differentiate between strains f S. cerevisiae anddemnstrated that the inculated strain tended t dminate but that indigenus strains remained in significant numbers thrughut the fermentatin. Liseau et al. (1987) and Delteil and Aizac (1988) cnducted

159 -110- investigatins using genetically ma ked strains f S. cerevísiae that culd be fllwed independently f indigenus srains. They cncluded that dminance f the inculated strain was nt always assured and depended n the specific cnditins f fermentatin (fr example, methd f inculatin). In light f these bservatins, ne must cnclude that inculatin f any particular strain f S. cerevisiae des nt necessarily guarantee its dminance r exclusive cntributin t the fermentatin. Clearly, further studies n the subject f strain dminance are required. An experiment was devised, therefre, t mnitr the gfwttr f an inculated strain under different cnditins f fermentatin. The use f a genetically marked strain prvided the means fr accurate identificatin and quantitatin f the inculated strain thrughut the fermentatin. Cnditins f fermentatin t be investigated included prductin f killer txin by the inculated strain, pre-treatment f the must with SO2, and temperature f fermentatin. These cnditins f fermentatin a e discussed briefly belw. 7.l.l Killer yeast inculatin The subject f killer yeasts has already been addressed in Chapter 6 and therefre will nt be discussed here in detail. Several studies have nw demnstrated that killerprducing and killer sensitive strains f S. cerevisiae may ccur as pa t f the natural flra f wine fermentatins (Heard and Fleet, 1987a,b). Labratry experiments have demnstrated that killer strains can inhibit sensitive strains and becme the dminant strain in mixed-culture wine fermentatins (Heard and Fleet, l987a,b; Lng et a1.,1990). Tw imprtant reasns exist fr killer yeasts t be f interest t winemakers. First, they may be respnsible fr a number f stuck r undesirable fermentatins. Inculated S. cerevisiae strains culd be destryed by indigenus killer strains f S. cerevisiae r nn- Saccharmyces species, leading t premature terminatin f the fermentatin, slw fermentatin r cmpletin f the fermentatin by a less desi able strain. Secndly, there may be sme advantage in cnducting the fermentatin with desired killer strains f S.

160 -111- cerevisiaeithe expectatin being that the grwth f less desired indigenus strains wuld be supressed. Kiler strains f S. cerevisiae are nw cmmercially available t winemakers but little evidence exists t assure their activity against indigenus yeasts in any particular wlnery Additin f sulfur dixide t Srape must Additin f SO2 t grape juice fr the purpses f cntrlling xidatin and restricting gfwth f indigenus micrflra is a well established practice in winemaking (Beech and Thmas, 1985). In the case f wines prduced by natural fermentatin, it is cnsidered that the additin f SO2 will suppress the grwth f indigenus nn- Sacclnrmyces yeasts and encurage dminance f fermentatin by the mre SO2 tlerant strains f S. cerevisiøe (Ribereau-Gayn et al., 197 5). Fr wines prduced by inculatin, it is assumed that added SO2 will cntrl indigenus nn-saccharmyces, as well as indigenus S. cerevisiae and encurage fermentatin by the inculated strain. Hwever, experimental evidence supprting these effects f SO2 n yeast eclgy in ferments is scatce. Grwth f indigenus yeasts in cmmercial wine ferments has been fund where the usual cncentratins f SOz ( mg/l) have been added t the juice (Fleet et al., 1984; Heard and Fleet, 1985, 1986a). Mre specifically, grwth f K. apiculata, ne f the main indigenus yeasts, was nt inhibited by ttal added SO2 cncentratins f mgll (Heard and Fleet, 1988a). These findings cntradict the assumptin that SO2 cnels indigenus yeasts and challenge ne f the reasns fr using SO2 in winemaking Temperanre f fermentatin Temperature cntrl has becme an imprtant practice in mdern winemaking. In recent years there has been a trend t ferment white wines in particulaí, al lwer temperatures (lg15c) t encurage ttre frmatin f vlatile flavun such as esters and t reduce lsses f prducts, including ethanl, by evapratin (Killian and Ough, 1979;

161 -112- Kunkee, 1984). The effect f temperatr re n the eclgy and kinetics f fermentatin, therefre requires cnsideratin. Heard and Fleet (1988b) demnstated that temperature can have a dramatic effect n the eclgy f the fermentatin. Decreasing the temperature belw 20C substantially increased the cntributin f the nn-s c clørmyces yeasts t ttre fermentatin. Kleckera apiculataand C. stellata,fr example, remained at high ppulatins ( cells/ml) thrughut the fermentatin and in fact, K. apiculata replaced S. cerevisiae as the dminant yeast. It appeared that sme f the indigenus nn-sacchartnyces species grew faster than S. cerevisiae at lwer temperatures and, in additin, have enhanced ability t tlerate ethanl (Ga and Fleet, 1988). This change in eclgy with fermentatin at lwer temperature culd lead t prductin f wines with altered chemical and sensry cmpstin. Hwever, research has nt yet been cnducted t examine this crrelatin - further studies f yeast ppulatin dynamics at lwer temperatures are required. 7.2 RESULTS Grape must was cllected frm St. Hallet's winery in the Barssa Valley. Grapes were crushed and pressed immediately after harvesting, and must was cllected directly frm the winery press. Details f the must are presented belw: Grape variety: Hanesting Mde: Date f harvesting: ph f must: Glucse + fructse: Free SO2: Ttal SO2: Indigenus yeast level: PedrXímines Mechanical 9th April, sll 0 me/l 0 mg/l 5 x ld cels/ml-

162 -113- Sample SO2 added Inculated yeast Temperature f Time f strain fermentatin (C) inculatin (hrs) 1 3AMC AMC AMC AM AN,I ANd I 100 mg/l 100 mg/l 3AN{ 3AM mg/l 3Al\d Alvl Al\4 r0 30 L2 3AN,I t3 100 mg/l 3AM 10 6 t4 100 mg/l 3AN,I l5 100 msll- 3AN,t Table 7.1. Cnditins under which 15 different fermentatins were perfrmed.

163 Figure 7.1. Grwth curves f the (O) inculated strain and (O) indigenus ppulatin in ferments cnducted at 20C (samples 1-9 in Table 7.r).

164 = E c 6 t 108 ß È f c. õf " 104 E f c õ f '12 rc I r0 12 â E f c.9 f t Time (days) :E l c!f 109 't " J 104 Time (days) E l c õ f " 't " 104 Time (days) 103 d 10" 103 :E f c õ f a " 107 A 10" 1OJ Time (days) = E l I c d l è õ 102 ô 10" 10" " " 246 Time (days) I Time (days) Time (days) E f c ñ l a ñ 10' 10" 107 t a 10" Time (days) Time (days)

165 Figure 7.2. Grwth curves f the (O) inculated strain and (O) indigenus ppulatin in ferments cnducted at 10C (samples in Table 7.r).

166 E l c _9 õl l4 10 f C.9 õl " 104 d 103 ñ 't È J c.9 õ l 6 't " ' Time (days) 11 Ē - _9 c.s d f a 10 " l0j Time (days) 't? Ē " Time (days) = E f " T me (days) 1 c = c d ā 107 A 10" õ ' ' Time (days) T me (days)

167 -114- Fifteen must samples (f 20litre vlume each) were fermented under the cnditins described in Table 7.1. Aliquts were remved frm the centre f the fermentatin vessel at regular intervals and plated in duplicate nt YPD media. Resultant clnies were assayed fr GUS activity by the agar plate methd. The number f blue clnies (representing the inculated strain) and the number f white clnies (representing the indigenus yeast strains) were cunted t deærmine the prprtin f the inculated strain Fermentatins cnducted at 2@e Results f the ferments cnducted at 20C, which cmprise three different sets f cnditins, are depicted in Figure 7.1. The effect f killer txin n the indigenus yeast ppulatin in untreated must was analysed. This analysis was achieved by cnducting identical ferments, ne with killer strain 3AM and ne with its cured, sensitive derivative 3AMC. These strains were shwn t have simila grwth kinetics during fermentatin (Chapter 6) and therefre differences bserved in these experiments can be attributed t the prductin f killer txin. Als, the effect f SO2 additin t grape must prir t inculatin was investigated. Under each f these fermentatin cnditins, different times f must inculatin were studied. These times were 6, 30, and 50 hurs after grape pressing. Results f each different set f fermentatin cnditins will be discussed separately. i) Inculatin with killer sensitive strain 3AMC (samples 1,2 and 3) At the inculatin time f 6 hurs after pressing (sample 1) indigenus yeasts were present in the must at a level f 5 x 105 cells/rnl, and the inculum was added t 4 x 106 cells/ml. Therefre, strain 3AMC had an eight-fld increase in ppulatin size ver the indigenus yeasts at the time f inculatin. Strain 3AMC clearly dminated this ferment,

168 -115- reaching 1.5 x 108 cells/ml and accunting fr 957 f the ppulatin by day three f the ferment. When inculatin was delayed until 30 hurs after pressing f the grapes (sample 2), the indigenus yeast ppulatin had reached a cell density f 5 x 106 cells/ml just prir t the additin f strain 3AMC. At this inculatin time, then, 3AMC represented nly 45V fthe ttal ppulatin. Under these cnditins, the indigenus yeasts reached a level f 1.9 x 108 cells/ml (95V f the ppulatin) n day 4 and clearly dminated thrughut the ferment. Snain 3AMC reached a peak cell density n day 3 f nly 6 x 106 cells/ml, and after this pint displayed a steady decrease in ppulatin size. By day 11 f the ferment it had decreased t 1 x lff cells/ml, and represented less than l7 f the ttal ppulatin. In the ferment in which inculatin was perfrmed 50 hurs after pressing (sample 3), the indigenus yeast ppulatin had entered lg phase f grwth and reached 3.9 x 107 cellvrnl by the time f inculatin. The indigenus yeasts therefre had a ten-fld higher ppulatin density than the inculated strain and maintained dminance ver the ferment. Strain 3AMC displayed a rapid eductin in ppulaún size (r death rate) and by day 4had drpped t 5 x los cells/ml and accunted fr less than l7 f the ttal ppulatin. ü) Inculatin with the killer strain 3AM (samples 4,5, and 6) The kinetics f killer yeast 3AM in untreated must at an inculatin time f 6 hurs after pressing (sample 4) was similar t that f strain 3AMC under identical cnditins (sample 1). Strain 3AM reached I x 108 cells by day 2 f the ferment, accunting fr 967 f the ttal ppulatin and clearly dminated thrughut the ferment. At inculatin times f 30 and 50 hurs after grape presssing, 3AM was dminated by the indigenus yeast ppulatin. Hwever it disptayed different grwth kinetics t 3AMC. In the 30 hur inculum (sample 5), 3AM reached a density f 4 x 107 cells/ml and

169 -116- represented 25V f the ttal ppulatin. A fairly cnstant ppulatin size was maintained by 3AM thrughut this ferment, and by day 11 it represented 35V f the ttal ppulatin, cmpared t less thanl% represented by strain 3AMC at the same time. In the ferment inculated 50 hurs after pressing (sample 6), strain 3AM accunted fr nly l57 fthe ttal ppulatin at the time f inculatin, and was strngly dminated by the indigenus yeasts thrughut the ferment. Hwever, it differed frm strain 3AMC in that a cnstant cell density (between 1 and 4 x lff cells/ml) was maintained after inculatin. N dramatic reductin in ppulatin size f strain 3AM was evident in this ferment, althugh grwth f the strain was clearly suppressed. iü) Inculatin with 3AM f must pre-treated with sulfur dixide (samples 7, 8, and e) Additin f SO2 t the grape must had the expected effect f reducing the indigenus yeast ppulatin. Fur hurs after the additin f SO2 (and 6 hurs after grape pressing), the indigenus yeast level had drpped frm 5 x 105 cells/ml t I x 105 cells/ml. At this time the f,rst inculatin f 3AM was perfrmed (sample 7) and a cncmitant reductin in the inculum ppulatin ccurred. The ppulatin density f 3AM drpped frm 4 x 106 cells/ml at the time f inculatin t I x 106 cells/ml n day ne f the ferment, at which pint the inculum represented 82V f the ttal ppulatin. After day ne, bth the inculated srain and the indigenus ppulatin had recvered frm the SO2 treatment, and entered lg phase f grwth. By day fur 3AM had reached 2 x 108 cells/ml (accunting fr 937 f the ttal ppulatin) and clearly dminated the ferment. At this time f inculatin, the additin f SO2 reduced the efficiency f the inculum in attaining dminance f the ferment when cmpared t the untreated must sample (sample 4). In ferments cnducted with an inculatin time f 30 hurs after pressing, the additin f SO2 t the must increased the inculatin efficiency. Under these cnditins

170 -117- (sample 8), the level f inculated strain was apprximately 10-fld higher than the indigenus yeasts at the time f inculatin. Hwever, the indigenus ppulatin entered lg phase f grwth befre the inculated strain and by day three had reached a cell density f 7 x 107 cells/ml - apprximately equal t that f strain 3AM. Frm this pint f the ferment nwards, strain 3AM was present as apprximaæly 55V f the ttal ppulatin and therefre nly narrwly dminated the fermentatin. This cntrasts t the ferment in untreated must with the same inculatin time (sample 5), in which the indigenus yeast ppulatin was dminant. There was little difference in the kinetics f strain 3AM in bth untreated (sample 6) and SO2 (sample 9) treated must with an inculatin time f 50 hurs after pressing. In bth cases the indigenus ppulatin had reached 2 x 108 cells/ml and accunæd ft 977 f the ferment by day 4 f the ferment. Strain 3AM remained at a steady cell density thrughut bth f these ferments, with a ppulatin f 2 x 106 cells/ml by day seven Fermentatins cnducted at l@e Tw cnditins were cmpared at a fermentatin temperature f l0c: n treatment f must prir t inculatin; and treatment with SOz (100 mgll) prir t inculatin. Bth cnditins were investigated with the marked strain 3AM and are presented graphically in Figure 7.2. Ð Inculatin f untreated must In the untreated must inculated 6 hurs after pressing (sample l0), the indigenus ppulatin and strain 3AM exhibited a tw day lag phase, during which 3AM was present as78v f the ttal ppulatin. By day 7, 3AM peaked at 1 x 108 cells/ml and accunted fr 967 f the ttal number f yeast cells. At this time f inculatin, strain 3AM exerted a similar dminance ver the ferment as it did in ttre 20C ferment (sample 4).

171 -118- \ù/ith an inculatin time f 30 hurs after pressing (sample 1l) bth indigenus yeasts and 3AM shwed a lag phase up t day 3, at which pint strain 3AM cmprised apprximately 50V f the ttal ppulatin. On day 7 f the ferment, 3AM reached a peak f 7 x 10e cells/ml and at this pint was 48V f the ttal ppulatin. Strain 3AM accunted fr apprximately 507 f the ttal ppulatin thrughut the ferment and therefre inculatin eff,rciency under these cnditins was grcater at 10C than it was in the ferment cnducted at 2@C (sample 5). The indigenus yeast ppulatin dminated the untreated must ferment when inculatin f 3AM was perfrmed 50 hurs after grape presssing. At the time f inculatin, the indigenus yeast ppulatin had reached 9.5 x 105 cells/ml and strain 3AM accunted fr 77V f the ttal yeast ppulatin. Hwever, the indigenus yeast ppulatin entered lg phase f gru'th befre the inculated strain and by day seven had a density f 4 x 107 cells/ml and represented 797 f the ttal ppulatin. The dminance f the indigenus yeast ppulatin was nt as great under these cnditins as it was in the zwc ferment (sample 6). ü) Inculatin f sulfur dixide treated must Srain 3AM clea ly dminated the ferment when the inculum was added t the SOz treated must 6 hurs after grape pressing (sample 13). In cntrast t the 20C ferment inculated at this time (sample 7), pre-reatnent f SO2 increased the dminance f 3AM. The lwer temperature slwed recvery f the indigenus ppulatin after additin f SO2, allwing strain 3AM t enter lg phase grwth withut cmpetitin. Frm day 1 nwards, strain 3AM accunted fr greater thang3% f the ttal ppulatin and reached a maximum cell density f 9 x 107 cells/ml (987 f the ttal ppulatin) by day seven.

172 Figure 7.3. Killer assays perfrmed n indigenus yeast clnies islated n day seven f the fermentatin.

173 -119- V/hen inculatin was perfrmed 30 hurs after pressing (sample l4), the level f indigenus yeasts had drpped belw lf cells/ml and 3AM dminated the ferment as it cmprised apprximately SOV f the ttal ppulatin fr the duratin f the fermentatin. At this time f inculatin, strain 3AM achieved greaær dminance f the SO2 treated must at 10C than it did in the 20C ferment (sample 8). When inculatin was delayed fr 50 hurs after pressing (sample 15), strain 3AM was still able t dminate the ferment in SO2 treated must at 10C. At the time f inculatin, strain 3AM accunted fr 977 f the ttal ppulatin. The indigenus ppulatin, hwever, recvered frm the SO2 treatment and entered lg phase f grwth after day 3. Frm day 5 nwards, strain 3AM represented apprximately 70V f the ttal ppulatin. The efficiency f inculatin under these cnditins is markedly imprved at a temperature f 10C, cmpared t simila cnditins at 20C (sample 9) in which the indigenus yeast ppulatin was dminant Killer activit), in the indigenus yeast ppulatin In rder t interpret the rle f killer activity in the zoc ferments, killer assays were perfrmed n a number f GUS negative clnies (indigenus yeasts) islated frm SO2 Eeated and nn-treated ferments. These indigenus yeasts were islated frm samples taken n day seven frm ferments N. 2 (nn-treated must) and 9 (SO2 treated must). Assays were perfrmed by testing their ability t kill strain 3AMC and results are presented in Figure 7.3. Twenty eight indigenus yeast clnies were assayed, and f these, twenty (7IV) displayed distinct killer activity against srain 3AMC. 7.3 DISCUSSION General trends have been bserved in inculatin efficiency under different cnditins. Results presented here clearly demnstrate the imprtance f the time f

174 -120- inculatin in ensuring dminance f the ferment by the inculated strain. Under all cnditins examined, the inculated strain clearly dminated the ferment (accunting fr > gov f the ttal ppulatin by day three) when it was added t the must within six hurs f pressing. These results are t be expected, as a delay in inculatin time allws the indigenus ppulatin t increase in cell number. Cnsequently, the indigenus ppulatin is mre cmpetitive at the time f inculatin. The killer (3AM) and killer sensitive (3AMC) strains bth achieved a similar dminance f the ferment with an inculatin time f six hurs after pressing. This result suggests that the greater ppulatin size f the selected strain at the time f inculatin is sufficient alne t ensure dminance f the fermentatin and that the prductin f killer txin des nt play an imprtant rle under these cnditins. It shuld be nted that the indigenus ppulatin did nt display a reductin in cell density in the presence f the killer strain. Therefre, it appears that the indigenus yeast cells are being suppressed rather than actively killed by strain 3AM. This bsen atin can be explained by the fact that 757 f the indigenus yeasts islated frm the ferments n day seven displayed killer activity (see Figure 7.3) and therefre are immune t the killer txin. Differences were bserved, hwever, between the grwth kinetics f the killer and sensitive strains when inculatin was perfrmed at 30 and 50 hurs after pressing. Under these cnditins, in which the indigenus ppulatin was dminant, the sensitive strain shwed a marked decrease in ppulatin density and appeared t be killed by the indigenus yeasts. The killer strain, hwever, displayed either an increase in grwth (after 30 hur delay in inculatin), r maintained a cnstant cell density (after a 50 hur delay in inculatin). In interpreting these bservatins, it is again imprtant t remember that757 f the indigenus yeasts n day seven displayed killer activity. This killer activity f the indigenus yeasts explains the cell death displayed by the sensitive strain 3AMC in the ferments inculated 30 and 50 hurs after pressing. The fact that killer activity was nt evident against strain 3AMC in the ferment inculated after six hurs (sample 1) can be

175 explained by results btained in Chapter 6. These results shwed that the efficiency f killer activity was dependent n the prprtin f killer t sensitive cells in the ferment. When killer cells a e present as less than 50V f the ppulatin, as they were at inculatin time and thrughut this ferment, killer activity was nt detected. The finding thattl{z f the indigenus yeast clnies tested displayed killer activity is smewhat surprising cnsidering that Hea d and Fleet (1987a) identif,red nly 9 killer yeast frm a ttal f 61 S. cerevisiae and 36 nn-saccharmyces strains islated frm Australian wineries. Hwever, when ne cnsiders the prbable species and rigin f the clnies tested, this result is perhaps nt unexpected. First, it is likely that the clnies tested were strains f S. cerevisíae. The clnies were islated n day seven f 20C ferments and, invariably, S. cerevisiae is the nly species islated frm fermenting wine after the first three r fur days (Kunkee and Amerine, l97o; Fleet, 1990). Secndly, it is likely that the rigin f S. cerevisiae strains in wine is the winery and its equipment (Ma tini and Martini, 1990). Killer strains f S. cerevßiae arecurrently ppular chices fr use as starter cultures in the Australian wine industry. Of 57 dried wine yeasts currently available t rhe Australian industry, 2l are killer strains (Henschke, 1990). It is likely therefre, that killer yeasts will be increasingly clnising equipment f Australian wineries and cntributing t the indigenus yeast ppulatin f freshly extracted grape juice. The additin f SO2 had the expected effect f reducing the indigenus yeast ppulatin in the grape must f ferments cnducted at 20C. Hwever, when inculatin was perfrmed within six hurs f pressing, and fur hurs after the additin f SO2, the inculated yeast strain was als reduced by the added preservative. The pre-treatnent f the must with SOz did nt increase the efficiency f inculatin in this case. When the inculum lryas added 30 hurs after pressing, the additin f SO2 was effective in imprving the effîciency f the inculated strain at 20C. In this case, the indigenus yeast ppulatin was decreased while the inculated srain did nt suffer an

176 -122- initial reductin in ppulatin. Hwever, by 50 hurs after pressing, the indigenus ppulatin had recvered frm the SO2 treaunent, entered lg phase f grwth and reached a high cell density (2 x L07 cells/ml). The inculated strain was then easily dminated by the indigenus yeasts. Therefre, ne can cnclude that the additin f SO2 was nt an effective sterilisatin prcess in this case. The different efficiencies f SO2 treatment with respect t time f inculatin can be understd by cnsidering the mechanism f SO2 actin n yeast cells. In slutin, SO2 underges dissciatin t frm bisulfite and sulfite ins. The antimicrbial activity is due t penetratin f yeast cell membranes by the free, undissciated frm f SO2 and its subsequent inactivatin f intracellula cnstituents (Schmiz, 1980). In the cell membrane, the SO2 activates ATPase causing a decrease in the ATP cncentratin. Als, nce the mlecula SO2 has entered the cell, it dissciates, because f the ph difference between the grape juice and the cell, and becmes trapped. The cmbined effects f ATP depletin and the activity f sulfite and bisulfite ins inside the cell lead t cell inactivatin and death (Schmiz, 1980; Beech and Thmas, 1985; Stratfrd and Rse, 1985). It shuld als be nted that SO2 can be bund t grape juice cnstituents r fermentatin prducts, thus reducing its efficacy as an antimicrbial agent. Binding f SO2 by grape juice cnstituents, principally sugars, has been reprted by Bluin (1966) and Rankine (1966). The main fermentatin prducts that bind with SO2 are acetaldehyde, pynrvic acid and 2-ketglutaric acid and t a lesser extent, galacturnic acid, sugars and anthcyanins. This phenmenn has been reviewed by Beech et al. (1979), Beech and Thmas (1985) and Lafn-Lafurcade (1985). The strngest bnding f SO2 is frmed with acetaldehyde. Fr 50 mg/l f free SO2, 99V is bund with acetaldehyde in the wine (Lafn-Lafurcade, 1985) and the resulting sulfnate has little inhibitry effect against yeasts (Ussegli-Tmasset et al., 1981; Beech and Thmas, 1985). Because the dissciatin cnstant (1.4 x 10-6 t 1.5 x l0-ó) f this reactin in wine is high (Rankine,

177 ; Burrughs and Sparks, 1973), the binding reactin is nt easily reversed and the SO2 bund by the acetaldehyde is effectively remved frm inhibitry actin (Rankine, 1966). Mlecula SO2 in the musts f these ferments then, wuld either enter the yeast cells and then dissciate, causing death; r becme effectively irreversibly bund t cnstituents in the juice. The result f these prcessses wuld be a reductin in the amunt f available SO2 in the ferment ver time. This reductin in available mlecula SO2 explains the decrease in efficacy f SO2 treatment ver the three different times f inculatin. The lwer fermentatin temperature f 10C substantially enhanced the efficiency f SO2 in its actin f suppressing indigenus yeast strains. Even at inculatin times f 30 and 50 hurs after pressing, the inculated strain was able t dminate the ferment althugh the degree f dminance decreased ver time. The increase in antimicrbial activity f SO2 at 10C can be explained by the fact that the lwer temperature decreases cell grwth and prductin rate f fermentatin prducts such as acetaldehyde. Therefre the level f available mlecular SO2 (fr diffusin int cells f the indigenus yeasts) wuld be higher ver the first few days f the fermentatin at 10C than at 20C. The direct signifîcance f results presented in this chapter t the wine industry is difficult t assess, as prcedures such as cld settling and juice cla ificatin were nt cnducted prir t inculatin. These prcedures are likely t have a majr impact n the ppulatins f indigenus yeasts in the juice. The extent f these influences a e unclear and require study. Sme grwth f indigenus yeasts may be expected during settling - the types f species that grw and their rate f grwth will depend n the temperature and ther cnditins. Hwever, cnclusins have beeen drawn frm these results regarding the efficiency f inculatin in the presence f different levels f indigenus yeast species. Mst imprtantly, perhaps, this chapter hightights the suitability f the genetically marked strain

178 -124- in cnducting investigatins f yeast eclgy during fermentatin. Very few studies f this type have been cnducted due t the diff,rculties with strain identificatin. The ma ked strain prvides the means fr further studies int, fr example, the impact f winery p cesses such as cld senling and juice cla ificatin n indigenus yeast ppulatins.

179 -125- Chapter I General Cnclusins Results btained during the curse f this prject have demnstrated the ptential applicatin f mlecular genetic yeast manipulatins t the wine industry. A prcedure which enables the intrductin f new genetic material t wine yeasts has been described and, subsequently, emplyed in the develpment f a genetic marking system fr wine yeast strains. This system utilises the E.cIí p-glucurnidase (GUS) gene as a marker and, unlike previus marking systems described t date, enables a wide range f wine yeast strains t be readily marked. Furthermre, methds have been develped whereby GUS activity can be detected in single cells r clnies within a tw t fur hur perid. The GUS ma king system, therefre, prvides a means fr rapid and unequivcal identificatin and mnitring f yeast strains during fermentatin. Marked strains have ptential applicatin in mnitring the efficiency f varius aspects f wine prductin n the indigenus yeast lad. These aspects include yeast prpagatin, the preparatin f pure starter cultures and prefermentatin prcesses. Ma ked strains can als play an imprtant rle in enlgical studies aimed t increase ur knwledge and understanding f yeast kinetics, ultimately allwing the ptimizatin f fermentatin cnditins. This ptential rle f marked stains has been demnstrated in tw separate enlgical studies cnducted during this prject. First, the activity f killer txin in fermenting grape juice was assessed. Secnd, an analysis f inculatin efficiency under varius fermentatin cnditins was perfrmed. In cnclusin, this prject has three distinct cmpnents: the establisment f a wine yeast transfrmatin prcedure; use f this transfrmatin prcedure in the develpment f a genetic marking system; and demnstratin f the practicality and applicatin f ma ked strains t the wine industry.

180 -126- Results described in this thesis als have brader implicatins. It has been dernnstrated that a freign gene can be stably intrduced and expressed at the ILV2 site in the genme f wine yeast strains withut adversely affecting fermentatin perfrmance. The SMRI410 gene is, therefre, bth an apprpriate selectable marker and target site fr the integratin f freign DNA int wine yeasts. As the SMRI-410 gene is derived frm yeast, it wuld be pssible t cnstruct integrating vectrs devid f bacterial sequences. This wuld be an imprtant develpment in facilitating the acceptance f recmbinant yeasts in fd and beverage prductin. The challenge nw is t define specific targets in the wine-making prcess t which this technlgy can be applied"

181 References Aigle, M. and F. Lacrute (1975). Genetic aspects f [URE3], a nn-mitchndrial, cytplasmically inherited mutatin in yeast. Ml. Gen. Genet. 136: Aigle, M., D. Erbs and M. Mll (1983). Determinatin f brewing yeast plidy by DNA measurement. J. Inst. Brew. Ð: Akiyama, H., C. Kumagai, Y. Tanaka and K. Ouchi (1971). Studies n nn-faming yeast. VII. Selectin f the practical nn-faming yeasts fr sake brewing frm varius stck cultures. J. Sc. Brew. Japan 66: Alber, T. and G. Kawasaki (1982). Nucletide sequence f the trise phsphate ismerase gene f Saccharmyces cerevisiae.j. Ml. Appl. Genet. l: Ammerer, G. (1983) Expressin f genes in yeast using the ADC1 prmter. In R. Wu, L. Grssman and K. Mldave (eds.), Methds in Enzymt. Vl 101. Academic Press Inc., New Yrk, pp Anand, R. (1986). Pulsed freld gel electrphresis: a technique fr fractinating large DNA mlecules. Trends in Genetics Nv: Barnett, J.4., R.W. Payne, and D. Yarrw (1933). Yeasts: characteristics and identificatin. Cambridge University Press, Cambridge. Barney, M.C., G.P. Jansen and J.R. Helbert (1980). Use f spherplast fusin and genetic Eansfrmatin t intrduce dexrin utilizatin int Sacclnrmyces uvarurn. J. Am. Sc. Brew. Chem.38: l-5. Barre, P. (1934). Le mecanisme killer dans la cncurrence entre suches de levures. Evaluatin et prise en cmpte. Bull. O.I.V. 64I-642: Beckerich, J.M., P. Furnier, C. Gaillardin, H. Heslt, M. Rchet, and B. Tretn (1984). Yeasts. In: Ball, C. (ed.), Genetics and breeding f industrial micrrganisms. CRC Press, Bca Ratn, Flrida, pp

182 -128- Beech, F.V/., and S. Thmas (1985). Actin antimicrbienne de I'anhydride sulfureux. Bull. O.I.V. 5E: Beech, F.W., L.F. Burrughs, C.F. Timberlake, and G.C. Whiting (1979). Prgres recents sur I'aspect chimique et I'actin antimicrbienne de I'anhydride sulfureux (SOz). Bull. O.I.V. 2: Beggs, J.D. (1978). Transfrmatin f yeast by plasmid. Nature 275: lo4-l09. a replicating hybrid Beggs, J.D. (1981). In D. vn'wettstein, J. Friis, M. Kielland-Brant and A. Stenderup (eds.), Mlecular Genetics in Yeast, Alfred Benzn Synpsium 16. Munksgaard, Cpenhagen. Beggs, J.D., J. Van den Berg, A. Van Oyen and C. Vy'eissmann (1980). Abnrmal expressin f chrmsmal rabbit p-glbin gene in Saccharmyces cerevisiae. Nature 283: Benda, I. (1932). Wine and Brandy. In: G. Reed (ed.), Presctt and Dunn's industrial micrbilgy. AVI technical bks. V/estprt, Cnnecticut, pp Bennetzen, J.L. and B.D. Hall (1982). The primary structure f Saccharmyces cerevisiae gene fr alchl dehydrgenase 1. J. Bil. Chem. 257:30t Bilinski, C.4., and G.P. Casey (1989). Develpments in sprulatin and breeding f brewer's yeast. Yeast 5: Bishp, D.H.L., P.F. Entwistle, I.R. Camern, C.J. Allen, and R.D. Pessee (19S8). Field trials f genetically engineered Baculvirus insecticides. In: M. Sussman, C.H. Cllins, F.A. Skinner and T.E. Stewa t-tull (eds.), The Release f Genetically Engineered Micrrganisms. Academic Press, Lndn; pp Bitter, G.A. and K.M. Egan (1984). Expressin f heterlgus genes in Saccharmyces cerevisiae frm vectrs utilizing the glyceraldehyde-3-phsphate dehydrgenase gene prmter. Gene 32:

183 Bitter, G.4., K.K. Chen, A.R. Banks and P.H. Lai (1984). Secretin f freign prteins frm S ccharmyces cerevisiae directed by cr-factr gene fusins. Prc. Natl. Acad. Sci. USA. 81: Blm, B.R. (1980). Interferns and the immune system. Nature 284: Bluin, J. (1966). Cntributin a I'etude des cmbinaisns de I'anhydride sulfureux dans les muts et les vins. Ann. Technl. Agnc. 15: Beke, J.D., D.J. Garfinkel, C.A. Styles, and G.R. Fink (1985). Ty elements transpse thrugh an RNA intermediate. Cell40: Bne, C., A.M. Sdicu, J. Wagner, R. Degre, C. Sanchez, and H. Bussey. (1990). Integratin f the yeast Kl killer txin gene int the genme f marked wine yeasts and its effect n vinificatin. Am. J. Enl. Vitic. 4l: Bstian, K.4., J.E. Hpper, D.J. Rgers, and D.J. Tipper. (1980). Translatinal analysis f the killer-assciated virus-like particle dsrna genme f Saccharmyces cerevísíae; M-dsRNA encdes txin. Cell 19: Buix, M., J.Y. Leveau, and C. Cuinier (1981). Determinatin de I'rigine des levures de vinificatin, par une methde de differenciatin fine des suches. Cnnaisance de la Vigne et du Vin 15: Brent, R. (1985). Repressin f transcriptin in yeast. CeIl42:3-4. Brach, J.R. (1981). The yeast plasmid2 tm circle. In: Strathern, J.N., E.V/. Jnes and J.R. Brach (eds.), The mlecular bilgy f the yeast.saccharmyces: life cycle and inheritance. Cld Spring Harbur, New Yrk. pp Brach, J.R. (1982). The rle f site-specific recmbinatin in expressin f the yeast plasmid 2 Vm circle. Rec. Adv. Yeast Ml. Bil. L:93-LL2 Brwn, S.W. and S.G. Oliver (1982).Islatin f ethanl-tlerant mutants f yeast by cntinuus selectin. Eu. J. Appl. Micrbil. Bitechnl. 16: Burgers, P.M.J. and K.J. Percival (1987). Transfrmatin f yeast spherplasts withut cell fusin. Anal. Bichem. 163:

184 -130- Burrughs, L.F. and A.H. Sparks (1973). Sulphite-binding pwer f wines and ciders. I. Equilibrium cnstants fr the dissciatin f carbnyl bisulphite cmpunds. J. Sci. Fd Agric. 24: Bussey, H. and P. Meaden (1985). Selectin and stability f yeast transfrmants expressing cdna f an Ml killer txin-immunity gene. Curr. Genet. 9: Butt, T.R., E. Sternberg, J. Herd and S.T. Crke (1984). Clning and expressin f a yeast cpper metallthinein gene. Gene 27: Cabeca-Silva, C., A. Madeira-Lpes and N. van Uden (1982). Temperature relating f ethanl-enhanced petite mutatin in Sacclarmyces cerevísi e: mitchndria as t ugets f thermal death. FEMS Micrbial. I-ett.15: Cantwell, B.A., G. Brazil, N. Murphy and D.J. McCnnell (1986). Cmparisn f expressin f the end-ß-1,3-1,4-glucanase gene frm Bacilhts subtilis in Saccharmyces cerevisiae frm the CYC-1 and ADH-1 prmters. Cur. Genet. 11: Carle, G.F. and M.V. Olsn (1985). An electrphretic karytype fr yeast. Prc. Natl Acad. Sci. USA þ: Carle, G.F. and M.V. Olsn (1984). Separatin f chrmsmal DNA mlecules frm yeast by rthgnal-field-alternatin gel electrphresis. Nucleic Acids Res. 12: Ca le, G.F., M. Frank and M.V. Olsn (1936). Electrphretic separatin f large DNA mlecules by peridic inversin f the electric field. Science232: Cartwright, C.P., J. Jurszek, M.J. Beaven, F.M.S. Ruby, S.M.F. De Mrais and A.H. Rse (1986). Ethanl dissipates the prtn mtive frce acrss the plasma membrane f Saccharmyces cerevisiae. J. Gen. Micrbil. 132: Casey, G.P., W. Xia and G.H. Rank (1988a). A cnvenient dminant selectin marker fr gene transfer in industrial strains f Saccharmyces yeast: SMRI encded resistance t the herbicide sulfmeturn methyl. J. Inst. Brew. 94:93-97.

185 -131- Casey, G.P., Wei Xia, and G.H. Rank (19SSb). Applicatin f pulsed field chrmsme electrphresis in the study f chrmsme XIII and the electrphretic karytype f industrial strains f Saccharmyces yeasts. J. Inst. Brew. 94: Chu, G., D. Vllrath, and R.V/. Davis (1986). Separatin f large DNA mlecules by cntur-clamped hmgenus electric fields. Science 234: I Cilfi, G. (19SS). Ricnsciment e carattenzzazine dei lieviti. Vini d'italia 29:9-16. Ciriacy, M. and M. Grssman (1988). Characterisierung vn Saccharmyces cerevisiae weinhefen durch specifische DNA-markierungen. V/ein-Wiss. 43 : Clarke, L. and J. Carbn (1980). Islatin f yeast centrmere and cnstructin f functinal small chrmsmes. Nature 287 : Chen, J.D., T.R. Eccleshall, R.B. Needleman, H. Federfl B.A. Buchferer and J. Marmur (1980). Functinal expressin inyeast f the Esclrcrichia cli plasmid gene cding fr chlramphenicl acetyltransferase. Prc. Natl. Acad. Sci. USA. 77: Cnde, J. and G.R. Fink (L976). A mutant f Saccharmyces cerevisiae defective fr nuclear fusin. Prc. Natl. Acad. Sci. USA 73:365L Cntpulu, C.R., V.E. Ck, and R.K. Mrtimer. (1987). Analysis f DNA duble strand breakage and repair using rthgnal field alternatin gel electrphresis. Yeast 3: 7 l-7 6. Cstanz, M.C. and T.D. Fx (1938). Transfrmatin f yeast by agitatin with glass beads. Genetics 120: Cx, B.S. (1965). A cytplasmic suppressr f super-suppressr in yeast. Heredity 20: Cx, M.M. (1983). The FLP prtein f the yeast 2 ltm plasmid: expressin f a eukarytic genetic recmbinatin system in Escherichia cli (gene expressin/site specific recmbinatin/gene clning). Prc. Natl. Acad. Sci. USA 89: Cuinier, C. and C. Grs (1933). Enquete sur la repartitin des levures killers en France. Vignes et Vivs 38:25-27.

186 -132- Cummings, J. and S. Fgel (1987). Genetic hmlgy f wine yeasts with Saccharmyces cerevisiae. J.Inst. Brew. M: De Jnge, P., F.C.M. De Jngh, R. Meijers, H.Y. Steensma, and W.A. Scheffers (1986). Orthgnal-fietd alternatin gel elecrphresis banding patterns f DNA frm yeast. Yeast 2:193-2O4. De la Pena, P., F. Barrs, S. Gascn, P.S. Laz, and S. Rams. (1981). Effect f killer yeast txin n sensitive cells f Saccharmyces cerevisiæ.1. Bil. Chem.255: Delteil, D. and T. Aizac (19S8). Cmparisn f yeast inculatin techniques by the use f a 'ma ked strain'. Aust. NZ lvine Ind. J. 3(3): Demain, A.L. and N.A. Slmn (1981). Industrial micrbilgy: intrducing an issue n hw prducts useful t man a e manufactured by micr-rganisms. Sci. Am. 245: Dbsn, M.J., M.F. Tuite, N.A. Rberts, A.J. Kingsman, S.M. Kingsman, R.E. Perkins, S.C. Cnry, B. Dunbar and L.A. Fthergill (1982). Cnservatin f high efficiency prmter sequences in Saccharmyces cerevisiae. Nucleic Acids. Res. 10: Dtt, W. and H.G. Truper (1976). Sulfrte frmatin by wine yeasts. IV. Prperties f sulfite reductase. Arch. Micrbil. 108: Dtt, W., M. Heinzel and H.G. Truper (1977). Sulfite frmatin by wine yeasts. IV. Active uptake f sulfate by "lw" and "high" sulfite prducing wine yeasts. Arch. Micrbil. ll2: Duburdieu, D., A. Skl, J. Zucca, P. Thaluarn, A. Dattee, and M. Aigle (1987). Identificatin des suches de levures islees de vins par I'analyse de leur ADN mitchndrial. Cnnais Vigne Vin 21: Eschenbruch, R. (L974).Sulfite and sulfide frmatin during wine making - a review. Am. J. Enl. Vitic. 25:157-16L.

187 -133- Eschenbruch, R. and J.M. Rassell (1975). The develpment f nn-faming yeast strains fr wine-making. Vitis![: Eschenbruch, R. and P. Bnish (1976). Prductin f sulphite and sulphide by lw- and high-sulphite frming wine yeasts. Arch. Micrbil. 107: Eschenbruch, R., P. Bnish and B.M. Fisher (1978). The prductin f H2S by pure culture wine yeasts. Vitis!: Falc, S.C. (19S6). Selectable markers fr yeast transfrmatin. US patent 4, 626, 505. Falc, S.C. and K.S. Dumas (1985). Genetic analysis f mutants f Saccharmyces cerevisíae resistant t the herbicide sulfmeturn methyl. Genetics lo9: Falc, S.C., K.S. Dumas and K.J. Livak (1985). Nucletide sequence f the yeast LV2 gene which encdes acetlactate synthase. Nucleic Acids Res.!!: 4OLl Fangman, W.L., and V.A. Zakian (1981). Genme structure and replicatin. In: Strathern, J.N., E.W Jnes and J.R. Brach (eds.), The mlecular bilgy f the yeast Saccharmyces: life cycle and inheritance. Cld Spring Ha br Labratry, New Yrk. pp Fa kas, J. (1988). The technlgy and bichemistry f wine. Vl. 1. Grdn and Breach Sci. Pub. Mntreux. Fink, G.R. and C.A. Styles (1973). Curing f a killer factr in Saccharmyces cerevisiae. Prc. Natl. Acad. Sci. USA 69: Fleet, G.H. (1990). Grwth f yeasts during wine fermentatins. J. V/ine Res.!: Fleet, G.H., S. Lafn-Lafurcade, and P. Ribereau-Gayn. (1984). Evlutin f yeasts and lactic acid bacteria during fermentatin and strage by Brdeaux wines. Appl. Envirn. Micrbil. 48: Fgel, S. and J. Welch (1982). Tandem gene amplificatin mediates cpper resistance in yeast. Prc. Natl. Acad. Sci. USA 79:

188 -134- Freeman, R.F. (1981). Cnstructin f brewing yeasts fr prductin f lw carbhydrate beers. Eur. Brew. Cnv. Prc. Cng. 18th, Cpenhagen. pp M. Gaia, P. (1934). Lieviti in assciazine cntempranea. Vini Ital. þ: Ga, C. and G.H. Fleet (1988). The effects f temperature and ph n the ethanl tlerance f the wine yeast s Sacclørmyces ceraisiac, Candida Stellata and Kleclrcra apiculata. J. Appl. Bacteril. 6: Gardiner, K., W. Laas, and D. Pattersn (19S6). Fractinatin f large mammalian DNA restrictin fragments using vertical pulsed-field gradient gel electrphresis. Sm. Cell Ml. Genet. 12: Gastignl, 4., M. Ba n and G. Tiraby (1987). Phlemycin resistance encded by the ble gene transpsn Tn5 as a dminant selectable marker in Sacclørmyces cerevisiae.ml. Gen. Genet.M: Gilliland, R.B. (1951). The flcculatin characteristics f brewing yeasts during fermentatin. Prc. Eur. Brew. Cnv. Cngtr., Brightn, pp Gff, C.G., D.T. Mir, T. Khn, T.C. Gravius, R.A. Smithand A. Tauntn- Rigby (1984). Expressin f calf prchymsin in Saccharmyces cerevisiae. Gene 2l: Gdey, 4.R., S. Del, J.R. Piggtt, M.E.E. Watsn and B.L.A. Carter (1937). Expressin and secretin f freign plypeptides in yeast, In D.R. Berry, I. Russell and G.G. Stewa t (eds.), Yeast Bitechnlgy. Allen and Unwin, Lndn, UK, pp Gritz, L. and J. Davies (1983). Plasmid-encded hygrmysin B resistance: the sequence f hygrmycin B phsphtransferase gene and its expressin in Escherichia cli and Sacclnrmyces cerev iae. Gene 25: Gunge, N. and Y. Nakatmi (1971). Genetic mechanisms f rare matings f the yeast Sacclarmyces cerevisiaeheterzygus fr mating type. Genetics 4:

189 -135- Hadfield, C., A.M. Cashmre and P.A. Meacck (1986). An efficient chlrampheniclresistance ma ker fr Sacclnrmyces cerevisiae and Escherichia cli. Gene 45: Hadfield, C., B.E. Jrdan, R.C. Munt, G.H.J. Pretrius and E. Burak (1990). G418- resistance as a dminant marker and reprter fr gene expressin in Saccharmyces cerevisiae. Curr. Genet. 1E: Hammnd, J.R.M. and K.W. Eckersley (1984). Fermentatin prperties f brewing yeast with killer character. J. Inst. Brew. 99: L Hammnd, J.R.M. and K.W. Eckersley (1984). Fermentatin prperties f brewing yeast with killer cha acter. J. Inst. Brew. 99:167-L77. Hansen, E.C. (1886). Undersgelser ver alkhlsjaersvanpenes fysilgi g mrflgi. V. Methder til fremstilling af renkulturer af saccharmyceter g ligende mikrrganismer. Meddr. Carlsberg Lab. /: Hansen, E.C. (1388). Undersgelser fra gjaeringsindustriens praxis. Meddr. Carlsberg Lab. /: Hara, S., Y. Iimura, and K. Otsuka. (1980). Breeding f useful killer wine yeasts. Am. J. Enl. Vitic. p: Hara, S., Y. Iimura, H. Oyama, T. Kzeki, K. Kitan, and K. Otsuka (1981). Breeding f cryphilic killer wine yeasts. Agric. Bil. Chem. 45: Hara, S., Y. Iimura, H. Oyama, T. Kzeki, K. Kitna and K. Otsuka (1981). The breeding f cryphilic killer wine yeasts. Agric. Bil. Chem. 46: Harashima, S., A. Takagi and Y. Oshima (1984). Transfrmatin f prtplasted yeast cells is directly assciated with cell fusin. Ml. and Cell Bil. 4:77L-778. Harashima, S., Y. Ngi, and Y. Oshima. (1974). The genetic system cntrlling hmthallism in S ac c larmyc s yeasts. Genetics 77 : Haris, M.S. (1978). Virus-like particles and duble stranded RNA frm killer and nnkiller strains f Sacclarmyces cerevisiae. Micrbilgy Lt l6t'l74.

190 -136- Hashimt, H., H. Mrikawa, Y. Yamada and A. Kimura (1985). A nvel methd fr transfrmatin f intact yeast cells by electrinjectin f plasmid DNA. Appl. Micrbil. Bitechnl. 2t: Heard, G.M. and G.H. Fleet. (1985). Grwth f natural yeast flra during the fermentatin f inculated wines. Appl. Envirn. Micrbil. 59: Hea d, G.M. and G.H. Fleet. (1986a). Occurrence and grwth f yeast species during the fermentatin f sme Australian wines. Fd Technl. in Aust. 3E: Heard, G.M. and G.H. Fleet. (1986b). Evaluatin f selective media fr enumeratin f yeasts during wine fermentatin. J. Appl. Bateril. 69: Hea d, G.M. and G.H. Fleet. (1987a). Occunence and grwth f killer yeasts during wine fermentatin. Appl. Envirn. Micrbil. 5i: 217 t Hea d, G.M. and G.H. Fleet. (1987b). The ccurrence f killer character in yeasts during the fermentatin f Australian wines. Aust. and N.Z. Wine Ind. J. 1(4): Hea d, G.M. and G.H. Fleet. (1988a). The effect f sulphur dixide n yeast grwth during natural and inculated wine fermentatin. Aust. and N.Z. Wine Ind. J. 1: Hea d, G.M. and G.H. Fleet (1988b). The effects f temperature and ph n the grwth f yeast species during the fermentatin f grape juice. J. Appl. Bacteril. 65: Heinzel, M and H.G. Truper (1976). Sulfite frmatin by wine yeasts. II. Prperties f ATP-sulfurylase. Arch. Micrbil. 107 : Hendersn, R.C.A., B.S. Cx and R. Tubb (1935). The transfrmatin f brewing yeasts with a plasmid cntaining the gene fr cpper resistance. Curr. Genet. 9: Henschke, P.A. (1990). Yeast strains available t winemakers The Australian Wine Research Institute Technical Review N. 69: l2-i7. Herskwitz, I. and Y. Oshima (1981). In: J.N. Strathern, E.V/. Jnes & J.R. Brach, (eds.), The mlecular bilgy f the yeast Sacclørmyces, Life cycle and inheritance, Cld Spring Harbr Labratry, p.181.

191 Hicks, J. and I. Herskwitz(1976).Intercnversin f mating types in yeast. I. Direct bservatins f the actin f the hmthallism (HO) gene. Genetics E3: Hinchliffe, E. and C.J. Daubney (1986). The genetic mdificatin f brewing yeast with recmbinant DNA. Am. Sc. Brew. Chem.44: Hinnen, 4., J.B. Hicks and G.R. Fink (1978). Transfrmatin f yeast. Prc. Natl. Acad. Sci. USA 75: Hitzeman, R.4., F.A. Hagie, H.L. Levine, G.V. Geddel, G. Ammerer and B.D. Hall (1981). Expressin f human gene fr interfern in yeast. Nature 293: Hlland, J.P. and M.J. Hlland (1980). Structural cmparisn f tw nn-tandemly repeated yeast glyceraldehyde-3-phsphate dehydrgenase genes.j. Bil. Chem. 255: Ingraham, J.L. and J.F. Guymn (1960). The frmatin f higher aliphatic alchls by mutant strains f Sacclnrmyces cerevisiae. Arch. Bichem. Biphys. EE: Ingraham, J.L., J.F. Guymn and E.A. Crwell (1961). The pathway f frmatin f n- butyl and n-amyl alchls by a mutant strain f sacclarmyces cerevisiae. Arch. Bichem. Biphys.!l: Ingram, L.O. and M. Buttke (1984). Effects f alchls n micrrganisms. Adv Micrbil. Physil. 25: Ismail, A.A. and A.M.M. Ali (1971a). Selectin f high ethanl-yielding Sacclnrmyces. I. Ethanl tlerance and the effect f training in Saccharmyces cerevísíae Hansen. Flia Micrbil. t6: Ismail, A.A. and A.M.M. Ati (1971b). Selectin f high ethanl-yielding Saccha myces. II. Genetics f ethanl tlerance. Flia Micrbil. lé: It, H., Y. Fukuda, K. Murata and A. Kimura (1983). Transfrmatin f intact yeast cells treated with alkali catins. J. Bacteril. 153: Jeffersen, R.A., S.M Burgess and D. Hirsch (1986). p-glucurnidase fum Escherichia cli as a gene fusin marker Prc. Natl. Acad. Sci. USA 83:

192 -138- Jeffersen, R.4., T.A. Kavanagh, and M.V/. Beven (1937). Gus fusins: p-glucurnidase as a sensitive and versatile gene marker in higher plants. EMBO J 6:390L Jiminez, A. and J. Davies (19S0). Expressin f a transpsable antibitic resistance element in S accharmy ces. Nature 287 : l. Jhnstn, J.R. and H.P. Reader (1983). Genetic cntrl f flcculatin. In: J.F.T. Spencer, D.M. Spencer and A.R.V/. Smith (eds.), Yeast Genetics: Fundamental and Applied Aspects, Springer-Verlag, New Yrk, pp Jnes, E.V/. (1977). Prteinase mutants f Saccharmyces cerevisiae. Genetics 85: Julius, D., R. Schekman and J. Thrner (1984). Glycsylatin and prcessing f preprcr-factr thrugh the yeast secretry pathway. Cell36: Jurszek, J., M. Feuillat, and C. Charpentier. (1987). Effect f ethanl n glucseinduced mvements f prtns acrss the plasma membrane f Saccharmyces cerevisiae NCYC 431. Can. J. Micrbil. 33: Kadwaki, K. and H.O. Halvrsn (1971). Appearance f a new species f ribnucleic acid during sprulatin in Saccharmyces cerevisiae. J. Bacteril. 105: Karube, I., E. Tamiya and H. Matsuka (1985). Transfrmatin f Saccharmyces cerevisiae spherplasts by high electric pulse. FEBS Lett. 182: Kasahara, H., K. Ouchi, K. Kurata, T. Ishid and Y. Nunkawa (1974). Genetic characters f nn-faming mutants f sake yeast. J. Ferment. Technl. 52: Kaster, K.E., S.G. Burgett and T.D. Inglia (1984). Hygrmycin B resistance as a dminant selectable marker in yeast. Curr. Genet. 8: Keeler, K.H. (1988). Can rile guarantee the safety f genetically engineered rganisms in the envirnment? Crit. Rev. Bitechnl. S: Kerr, A. (1989). Cmmercial release f a genetically engineered bacterium fr the cntrl f crwn gall. Agricultural Science 2(6\:41-44.

193 -t39 - Keszenman-Pereyra, D. and K. Hieda (1938). A clny prcedure fr transfrmatin f S ac c harmy ces cerevi siae. Curr. Genet. 13 : 2l-23. Killian, R.E. and C.S. Ough (1979). Fermentatin esters - frmatin and retentin as affected by fermentatin temperature. Am. J. Enl. Vitic. 30: Kingsman, 4.J., C. Stanway and S.M. Kingsman (1988). The expressin f hmlgus and heterlgus genes in yeast. Antnie van Leeuwenhek J. Micrbil. 53: Kingsman, S.M., A.J. Kingsman, M.J. Dbsn, J. Mellr and N.A. Rberts (1935). Heterlgus gene expressin in Saccharmyces cerevisiae. In G.E. Russell (ed.), Bitechnl. Genet. Eng. Rev., V1.3. Intercept, Pnteland, Newcastle upn Tyne, pp Kramer, R.4., T.M. DeChiara, M.D. Schaber and S. Hilliker (1984). Regulated expressin f a human interfern gene in yeast: Cntrl by phphate cncentratin r temperature. Prc. Natl. Acad. Sci. USA 81: Kreger-van Rij, N.J.W. (1987). Classificatin f yeasts. In: A.H. Rse and J.S Harrisn (eds), The Yeasts: Bilgy f Yeasts. Lndn, Academic press, pp Kreger-van Rij, N.J.W. ed. (1984). The Yeasts: A taxnmic study. Elsevier Science Publishing C., Amsterdam. Kunkee, R.E. (1984). Selectin and mdificatin f yeasts and lactic acid bacteria fr wine fermentatin. Fd Micrbil. l: Kunkee, R.E. and M.A. Amerine, M.A. (1970). Yeast in winemaking. In A:H. Rse and J.S. Harrisn (eds.) Yeast Technlgy. Vl. 3. Academic Press, Lndn, pp Kunkee, R.E. and R.W. Gswell(1977).Table wines. In: A.H Rse and J.S Ha risn (eds.), The Yeasts. Vl. 3. Yeast technlgy. Academic Press, Lndn. pp Kurjan, J. and I. Herskwitz(1982). Structure f yeast phermne gene (MFcr): a putative g-factr precursr cntains fur tandem cpies f mature ü,-factr. Cell 30:

194 -140- Kurtzman, C.P. and H.J. Phaff (1937). Mlecular taxnmy. In A.H. Rse and J.S. Ha risn (eds.), The Yeasts: Bilgy f Yeasts. Academic press, Lndn, pp Kusewicz, D., and J. Jhnstn (1980). Genetic analysis f cryphilic and mesphilic wine yeasts. J. Inst. Brew. W: Lafn-Lafurcade, S. (1985). Rle des micrrganisms dans la frmatin de substances cmbinant le SOz. Bull. I.O.V.5E: Lafn-Lafurcade, S. (1983). Wine and brandy. In: H.J. Rehm and G. Reed (eds.), Bitechnlgy Vl. 5. Fd and feed prductin with micrrganisms. Verlag Chemie, Weinheim. pp Lafn-Lafurcade, S. and P. Ribereau-Gayn (1984). Develpments in the micrbilgy f wine prductin.prg. Ind. micrbil. 19: La kin, J.C and J.L. Wlfrd (19S3). Mlecula clniing and analysis f the CRYl gene: a yeast ribsmal prtein gene. Nucleic Acids Res. 1l: Lautensach, A. and R.E. Subden (198a). Clning f malic acid assimilating activity frm Leucnstc ens in E. cli. Micrbis. þ: Lea, C. and N. van Uden (1984). Effects f ethanl and ther alkanls n passive prtn flux in the yeast Sacclnrmyces cerevisiae.bichim. Biphys. Acta. 774: I-ea, C. and N. van Uden (1982). Effects f ethanl and ther alkanls n the kinetics and the activatin parameters f thermal death in Saccharmyces cerevísíae. Bitechnl. Bieng. 24: l58l Lee, S.Y., F.B. Kundsen, and R.O. Pyntn (1985). Differentiatin f brewery yeast strains by restrictin endnuclease analysis f their mitchndrial DNA. J. Inst. Brew.9l: Lewis, C.W., J.R. Jhnstn and P.A. Martin (1976). The genetics f yeast flcculatin. J. Inst. Brew. 82:

195 -L4l- Lindegren, C.C. (1949).The Yeast Cell, Its Genetics and Cytlgy. Educatinal Publ., St. Luis. Lindegren, C.C., and G. Lindegren (1943). A new methd fr hybridizing yeast. Prc. Natl. Acad. Sci. USA 29: Lipke, D.N. and C. Hull-Pillsbury (1984). Flcculatin f Saccharrnyces cerevísiae tttpl mutants. J. Bacteril. 159: Livingstn, D.M. (1977).Inheritance f the 2 pm DNA plasmid frm S accharmyces. Genetics 86: Ldder, J. (1970). The Yeasts (secnd editin). Nrth Hlland Publ. C., Amsterdam. Liseau, G., F. Vezinhet, M. Valade, A. Vertes, C. Cuinier, and D. Detteil (1937). Cntrle d'efficacite de levurage par la mise en euvre de suches de levures enlgiques marquees. Revue Francais d'oenlgie 106: Lng, E., J.B. Yelazqaez, J. Casad, P. Cal, and T.G. Villa. (1990). Rle f killer effect in fermentatins cnducted by mixed cultures f Saccharmyces cerevisíae. FEMS Micrbil. Lett. 7 I : Martinez, J., C. Millan, and J.M. Orega (1989). Grwth f natural flra during fermentatin f inculated musts frm 'Pedr Ximenez' grapes. S. Afr. J. Enl. Vitic. 10: Martini, A. and A.V. Martini (1990). Grape must fermentatin: past and present. In: Spencer, J.F.T. and D.M. Spencer (eds.), Yeast Technlgy, Berlin: Springer Verlag, pp t Mead, D.J., D.C.I. Gardner and S.G. Oliver (1985). Enhanced stability f a 2pm-based recmbinant plasmid in diplid yeast. Bitechnl. Letters 8:39L-396. Meaden, P., K. Ogden, H. Bussey and R.S. Tubb (1985) A DEX gene cnferring prductin f extracellular amylglucsidase n yeast. Gene 34:

196 -142- Meilhc, E., J. Massn and J. Teissie (1990). High efficiency transfrmatin f intact yeast cells by elecric f,reld pulses. Bitechnl. E: Mellr, J., M.J. Dbsn, N.A. Rberts, A.J. Kingsman and S.M. Kingsman (1985). Gene ll: Meyhack, 8., V/. Bajwa, H. Rudlph and A. Hinnen (1932). Tw yeast acid phphatase structural genes are the result f a tandem duplicatin and shw different degrees f hmlgy in their prmter and cding sequences. EMBO J.1: Miki, B.L.A., N.H. Pn, A.P. James and V.L. Seligy (1980). Flcculatin in Sacclarmyces cerevisiae: mechanism f cell-cell interactins. In: Stewart, G.G. and I. Russell (eds.), Current develpments in yeast resea ch. Pergamn Press, Canada Ltd., Trnt, pp Miki, B.L.A., N.H. Pn, A.P. James and V.L. Seligy (19S2b). Pssible mechanism fr flcculatin interactins gverned by gene FLOI in Saccharmyces cerevisiae.j. Bacteril. 150: Miki, B.L.A., N.H. Pn and V.L. Seligy (1982a). Repressin and inductin f flcculatin interactins in Sacclnrmyces cerevisíae. J. Bacteril. 150: Mill, P.J. (1964). The nature f the interactins between flcculent cells in the flcculatin f Saccharmyces cerevisiae. J. Gen. Micrbil.35: Miyajima, 4., I. Miyajima, K. Arai and N. Arai (1984). Expressin f plasmid R388- encded type II dihydrflate reductase as a dminant selective marker in S. cerevisiae. Ml. Cell. Bil. 4: Mlzahn, S.W. (1977).A new apprach t the applicatin f genetics t brewing yeast. J. Am. Sc. Brew. Chem.!L: Mmse, H., N. Okazaki and R. Tnike (1968). Studies n the aggegatin f yeasts caused by lactic acid bacteria. I. Aggregatin f yeast in mixture f pure-cultured yeast cells and lactic acid bacterial cells. J. Sc. Brew. Japan 63: Mrtimer, R.K. and D.E. Hawthrne (1969). Yeast Genetics. In: A.H. Rse and J.S Ha risn (eds.), The Yeasts, Vlume 1. Academic Press, New Yrk, pp

197 -143- Mrtimer, R.K. and D. Schild (1985). Genetic map f Saccharmyces cerevisiae, editin 9. Micrbil. Rev. 49: I8l-2L2. Mrtimer, R.K., D. Schild, C.R. Cnpulu and J.A. Kans (1989). Genetic map f S ac c larmy c e s c er ev i s iac, Editrn 1 0. Yeas t 5: 321' 403. Mwshwitz, D.B. (1979). Gene dsage effects n the synthesis f maltse in yeast. J. Bacteril. 137 : Murata, K., Y. Fukuda, M. Shimsaka, K. Watanabe, T. Saikusa and A. Kimura (1985). Phentypic character f the methylglyxal resistance gene in Sacclnrmyces cerevisíae: expressin in Esclvrichia clí andapplicatin t breeding wildtype yeast strains. Appl. Envirn. Micrbil. Ð: L2N Murray, A.V/. and J.V/. Szstak (1983). Cnstructin f artificial chrmsmes in yeast. Nature 305: Nagdawithana, T.W., T. Whitt and A.J. Cutaia (1977). Study f the effect f feedback f ethanl n selected enzymes f the glyclytic pathway. J. Am. Sc. Brew. Chem.35: Naumv, G.I. and T.I. Naumva. (1973). Cmparative genetics f yeast cmmunicatin. XI[. Cmparative study f killer strains f Saccharmyces frm different cllectins. Genetika 9: Newln, C.S. and W.L. Fangman (1975). Mitchndrial DNA synthesis in cell cycle mutants f Sacclnrmyces cerevisiae. Cell5: Nishikawa, N., M. Khg, and T. Ka akawa (1979). Serlgical classificatin f brewers yeast and related yeasts with cmmercial factr sera. J. Ferment. Technl. fl: Oliver, S.G. (1987). Physilgy and genetics f ethanl tlerrance in yeast. In: Stewart, G.G., I. Russell, R.D. Klein and R.R. Hiebsch (eds.), Bilgical research n industrial yeasts, Vl 1. CRC Press Inc. Bca Ratn, Flrida, pp

198 -t 44- On, B-f., and Y. Ishin-Ara (1988). Inheritance f chrmsme length plymrphisms in Saccharmyces cerevísiae. Cur. Genet 14: Osthuizen, 4., J.L.F. Kch, B.C. Viljen, H.B. Muller and P.M. Lategan (1987). The value f lng chain fatty acid cmpsitin in the identificatin f sme brewery yeasts. J. Inst. Brew. 93: I7 4-I76. Orr-Weaver, T.L., J.W. Szstak and R.J. Rthstein (1981). Yeast transfrmatin: A mdel system fr the study f recmbinatin. Prc. Natl. Acad. Sci. USA 78: Osman, Y.A. and L.O. Ingram (1985). Mechanism f ethanl inhibitin f fermentatin in Zymmnas mbilis CP4. J. Bacteril. 164:173-L80. Ouchi, K. and H. Akiyama(1971). Nn-faming mutants f sake yeasts. Selectin by cell agglutinatin methd and by frth fltatin methd. Agric. Bil. Chem.35: lo24- r32. Ouchi, K. and Y. Nunkawa (1973). Nn-faming mutants f sake yeasts. Their physicchemical characteristics. J. Ferment. Technl. 51: Ouchi, K., R.B. Vy'ickner, A. Th-e and H. Akiyama (L979). Breeding f killer yeast fr sake brewing by cytductin. J. Ferment. Technl.5T: Panchal, C.J., L. Bast, T. Dwhanick, and G.C. Stewart (1987). A rapid, simple and reliable methd fr differentiating brewing yeast strains based n DNA restrictin patterns. J. Inst. Brew. 93: Perlman, D. and H.O. Halvrsen (1983). A putative signal peptidase recgnitin site and sequence in eukarytic and prkarytic signal peptides. J. Ml. Bil. 167: 39I-409. Petes, T.D. (1980). Mlecular genetics f yeast. Ann. Rev. Bichem. 49: Philliskirk, G. and T.W. Yung. (1975). The ccurrence f killer chamcter in yeasts f varius genera. Antnie van I eeuwenhek J. Micrbil.4l: t47-15i.

199 -r45 - Ptter, H., L. V/ier and P. Leder (1984). Enhancer-dependent expressin f human r immunglbulin genes inrduced int muse pre-b lymphcytes by electrpratin. Prc. Natl. Acad. Sci. USA 8l: Pretrius, I.S. and J. Marmur (1988). Lcalizatin f yeast glucamylase genes by PFGE and OFAGE. Curr. Genet. 14:9-I3. Price, C.W., G.B. Fusn, and H.J. Phaff (1978). Genme cmparisn in yeast systematics: Deliminatin f species within the genera Sclanannimyces, Saccharmyces, Debarymyc s and Pichia. Micrbil. Rev. 42: L6l-I93. Prffitt, J.H., J.R. Davie. D. Swintn, S. Hattman (1984). 5'-Methyl cytsine is nt detecable in Saccharmyces cerevisiae DNA. Ml. Cell Bil. 4: Radler, F., P. Pfeiffer, and M. Dennert. (1985). Killer txins in new islates f the yeasts Hanseniaspra uvarurn and Pichia kluyveri. FEMS Micrbil. I-ett.29: Rank, G.H., G. Casey and W. Xia (1988). Gene uansfer in industrial Saccharmyces yeasts. Fd Bitech. 2: l-41. Rankin, B.C. (1966). Sulphur dixide in wines. Fd Tech. in Aust. 18: L34-I41. Rankine, B.C. (1968). The imprtance f yeasts in determining the cmpsitin and quality f wines. Vitis 7: Rankine, B.C. (1977). Mdern develpments in selectin and use f pure yeast cultures fr winemaking Aust. Wine Brew. Spirit Rev. 96: 3I-34. Rankine, B.C. and B. Llyd (1963). Quantitative assessment f dminance f added yeast in wine fermentatins. J. Sci. Fd Agric. 14: Rankine, B.C. and K.F. Pcck (1972). Alkalimetric determinatin f sulfur dixide in wine. Austral. V/ine Brew. Spirit Rev.91: 62,64,66. Rech, E.L., M.J. Dbsn, M.R. Davey and B.J. Mulligan (1990). Inrductin f yeast artificial chrmsme vectr int Saccharmyces cerevisiae cells by electrpratin. Nucleic Acids Res. 18: 1313.

200 -146- Reed, G. and T.W. Nagdawithana (1988). Technlgy f yeast usage in winemaking.am. J. Enl. Vitic.39: Ribereau-Gayn, J., E. Peynaud, P. Ribereau-Gayn, and P. Sudraud (1975). Sciences et techniques du vin. Tme tr. Dund, Paris. Rine, J., W. Hansen, E. Hardeman and R.V/. Davis (1933). Targeted selectin f recmbinant DNA clnes thrugh dsage effects. Prc. Natl. Acad. Sci. USA. &: Rberts, I.N., R.P. Oliver, P.J. Punt, and C.A.M.J.J. van den Hndel (1989). Expressin f the Escherichia cli p-glucurnidase gene in industrial and phytphathgenic filamentus fungi. Curr. Genet. 15: Reder, G.S., and S.E. Stewart (1988). Mittic recmbinatin in yeast. Trends in Genetics 4: Rgers, D. and E.A. Bevan. (1978). Grup classif,rcatin f killer yeasts based n crss reacúns between strains f different species and rigin. J. Gen. Micrbil. 105: Rman, P., C. Zambnelli and M.G. Sli (1976). Bisynthesis f sulphur amin acids in Saccharmyces cerevisae. II. Analysis f sulphur dixide prducing strains. Arch. Micrbil. 108: 21 l-215. Rsini, G. (1985a). Interactin between killer strains f Hansenula anmalavar. anmala and Saccharmyces cerevisiae yeast species. Can. J. Micrbil. 3!: Rsini, G. (1985b). Effet d'une levure "killer" de Saccharmyces cerevisiae sur une suche de levure sensible de la meme espece nn prducrice de H2S et selectinee pur la vinificatin dans un mileu de culture mixte. Bull. O.I.V : Rthstein, R. (1986). In: D.M. Glver (ed.), DNA clning: a practical apprach. IRL Press, Washingtn DC, pp Rthstein, S.J., C.M. Lazarus, W.E. Smith D.C. Baulcmbe and A.A. Gatenby (1984). Secretin f a wheat cr-amylase expressed in yeast. Nature 308:

201 Rus, C.V., R. Snw and R.E. Kunkee (1983). Reductin f higher alchls by fermentatin with a leucine-auxtrphic mutant f wine yeast. J. Inst. Brew. 89: Rupela, O.P. and P. Taur (1984). Islatin and characterizatin f lw hydrgen sulphide prducing wine yeasts. Enz. Micrb. Technl. 6: Russell, I., I.F. Hancck and G.G. Stewa t (1983). Cnstructin f dextrin fermentative yeast strains that d nt prduce phenlic ff-flavurs in beer. J. Am. Sc. Brew. Chem. 4l Sa- Crreia, I. and N. van Uden (1983). Temperature prfiles f ethanl tlerance - effects f ethanl n the minimum and the maximum temperature fr grwth f the yeasts Saccharmyces cerevisiae and Kluyvermyces fragilis. Bitechnl. Bieng. þ: Schekman, R., and P. Nvick (1982). The secretry prcess in yeast cell surface assembly. In: J. Strathern, E.W. Jnes, and J.R. Brach (eds.), The mlecular bilgy f the yeast Saccharmyc s, Vl. 2. Cld Spring Harbr, New Yrk, pp Scherer, S., C. Mann, and R.\il. Davis (1932). Reversin f a prmter deletin in yeast. Nature 289: Schmiz, K.L. (1980). The effect f sulphite n the yeast Saccharmyces cerevisiae. Arch. Micrbil. 125: Schultz, L.D. and J.D. Friesen (1983). Nucletide sequence f the tcml gene (ribsmal prtein M3) f Sacclnrmyces cerevisiae.j. Bacteril. 155: Schwa tz, D.C. and C.R. Cantr (198a). Separatin f yeast chrmsme-sized DNAs by pulsed field gradient gel electrphresis. Cell 37: Seki, T., E.H. Chi, and D. Ryu (1985). Cnstructin f killer wine yeast strains. Appl. Envirn. Micrbil. 49: l2ll-l215.

202 -148- Seki, T., S. Myga, S. Limtng, S. Uedn, J. Kumnuata and H' Taguchi (1933). Genetic cnstructin f a yeast fr high ethanl prductin. Bitechnl. Lett. l: Shaw, W.V. (1983). Chlramphenicl acetyltransferase: enzymlgy and mlecular bilgy. CRC Crit. Rev. Bichem.!l: Shimazu, K., T. Adachi, K. Kitan, T. Shimazaki, A. Ttsuku. S. Hara and H.H. Dittrich. (1985). Killer prperties f wine yeasts and cha acterisatin f killer wine yeasts. J. Ferment. Technl.63: Simn, J.R. and K. McEntee (1939). A rapid and efficient prcedure fr transfrmatin f intact Saccharmyces cerevisiae by electrpratin. Bichem. Biphys. Res. Cmm. 164: Ll Sinclair, J. H., B.J. Stevens, P. Sanghavi and M. Rabinwitz(1967). Mitchndrialsatellite and circula DNA filaments in yeast. Science 156: Skipper, N. and H. Bussey. (1977). Mde f actin f yeast txins: energy requirement fr Sacclørmyces cerevisiae killer txin. J. Bacteril. 129: Smith, C.L., T. Matsumt, O. Niwa, S. Klc, J.B Fan, M. Yanagida, and C.R. Cantr. (1937). An electrphretic karytype fr Schizsacclørmyces pmbe by pulsed field gel electrphresis. Nucleic Acids Res. É: Smith, R.4., M.J. Duncan and D.T. Mir (1985). Heterlgus prtein secretin frm yeast. Science 229: L2l Snell, R.G. and R.J. Wilkins (1986). Separatin f chrmsmal DNA mlecules frm C. albicans by pulsed field gel electrphresis. Nucleic Acids Res.![: 44Ol-M06. Sneyd, T.N., N.G.C. Bruer and T.H. Iæe (1990). A survey f five methds fr analyzing the alchlic strength f wine. Prc. Tth Aust. Wine Ind. Tech. Cnf. Adelaide, Ausralia. pp Snw, R. (1979). Tward genetic imprvement f wine yeast. Am. J. Vitic. Ð: Enl.

203 -149- Snw, R. (1983). Genetic imprvement f wine yeast. In: Spencer, J.F.T., D.M. Spencer and A.R.W. Smith (eds.), Yeast genetics, fundamental and applied aspects. Springer-Verlag, New Yrk, pp Spencer, J.F.T., and D.M. Spencer (1983). Genetic imprvement f industrial yeasts. Ann. Rev. Micrbil. 37: l2l-142. Spencer, J.F.T. and D.M. Spencer (1977). Hybridizatin f nn-sprulating and weakly sprulating strains f brewer's and distiller's yeasts. J. Inst Brew. E3:287'289. Stepien, P.P., R. Brusseau, R. 'Wu, S. Narang and D.Y. Thmas (1983). Synthesis f human insulin gene. VI. Expressin f the synthetic prinsulin gene in yeast. Gene 24: Stewart, G.G. (19S1). The genetic manipulatin f industrial yeast strains. Can. J. Micrbil. 27 : Stewart, G.G., I. Russell and C. Panchal (1981). The genetic manipulatin f industrial yeast strains. In: Stewart, G.G. and I. Russell (eds.), Current develpments in yeast resea ch. Pergamn Press, Trnt. pp Stratfrd, M. and A.H. Rse (19S5). Transprt f sulphur dixide by Saccharmyces cerevisiae. J. Gen. Micrbil. 132: I-6. Struhl. K. (19S3). The new yeast genetics. Nature 305: Struhl, K., D.T. Stinchcmb, S. Scherer and R.W. Davis (1979). High-frequency transfrmatin f yeast: Autnmus replicatin f hybrid DNA mlecules. Prc. Natl. Acad. Sci. USA 76: Subden, R.E. (1987). Current develpments in wine yeasts. CRC Crit. Rev. Bitech. t: Subden, R.E. and C. Osthsilp (1987). Malic acid metablism in wine yeasts. In: Stewarr, G.G., I. Russell, R.D. Ktein and R.R. Hiebsch (eds.). Bilgical research n industrial yeasts, Vl 2. CRC Press inc., Bca Ratn, Flrida, pp

204 -150- Subden, R.E., D. Irwin, J.D. Cunningham, and A.G. Meiering. (1982).'Wine yeast iszymes. I. Genetic differences in 18 stck cultures. Can. J. Micrbil. 28: lo47-loso. Sugden, D.A. and S.G. Oliver (1933). Reduced ethanl tlerance: ne f the pleitrpic effects f the pep4.3 mutatin in Saccharmyces cerevisiae. Bitechnl. Len. 5: Suizu, T., Y. Iimura, K. Gmi, K. Takahashi, S. Ilara, K. Yshizawa and G. Tamura (1990). Cnstmctin f urea nn-prducing yeast Saccharmyces cerevisiae by disruptin f the CAR1 gene. Agric. Bil. Chem. 54: Takagi, 4., S. Harashima and Y. Oshima (1985). Hybridizatin and plyplidizatin f Sacclnrmyces cerevisi strains by transfrmatin-assciated cell fusin. Appl. Envirn. Micrbil. 49: Takagi, M., S. Kawai, I. Shibuya, M. Miyazaki and K. Yan (1986). Clning in Saccharmyces cerevisiae f a cyclheximide resistance gene frm the Candida maltsa genme which mdifies ribsmes. J. Bacteril. 168: Takahashi, T. (1978). Genetic analysis f a German wine yeast. Bull. Brew. Sci. 24: Takan, I. and Y. Oshima (1970). Allelism tests amng varius hmthallism cntlling genes and gene systems in Saccharmyces. Genetics 64: Taylr, N.V/. and W.L. Ortn (1978). Armatic cmpunds and sugars in flcculatin f Saccharmyces cerevisiae.l.inst. Brew. 84: Thmas, D.S. and A.H. Rse (1979).Inhibitry effect f ethanl n grwth and slute accumulatin by Saccharmyces cerevisiae as effected by plasma membrane lipid cmpsitin. Arch. Micrbil. 122: Thrne, R.S.W. (1951). Sme aspects f yeast flcculence. Prc. Eur. Brew. Cnv. Cngr., brightn, pp.2i-34. Thrntn, R.J. (1982). Selective hybridisatin f pure culture wine yeasts. II. Imprvement f fermentatin efficiency and inheritance f SO2 tlerance. Eur. J. Appl. Micrbil. Bitechnl. 14: L59-I64.

205 -151- Thrntn, R.J. (1983). New yeast strains frm ld - the applicatin f genetics t wine yeasts. Fd Technl. Australia S: Thrntn, R.J. (1985). The intrductin f flcculatin int a hmthallic wine yeast. A practical example f the mdificatin f winemaking prperties by the use f genetic techniques. Am. J. Enl. Vitic. X: Thrntn, R.J. (197Sa). The mapping f tw dminant genes fr faming in wine yeasts. FEMS L-ett. L: Thrntn, R.J. (l97sb). Investigatin n the genetics f faming in wine yeasts. Eur. J. Appl. Micrbil. Bitechnl. 5: Thrntn, R.J., and R. Eschenbruch (197í).Hmthallism in wine yeasts. Antnie van Leeuwenhek J. Micrbil. 42: Tipper, D.J. and K.A. Bstian. (1984). Duble stranded ribnucleic acid killer systems in yeasts. Micrbil. Rev. $: Tredux, H.G., R.P. Tracey, a A. Trmp. (1986). Killer factr in wine yeasts and its effect n fermentatin. S. Afr. J. Enl. Vitic. 7: Tubb, R.S. and J.R.M. Hammnd (1987). Yeast genetics. In: Priest, F.G. and I. Campbell (eds.), Brewing micrbilgy. Elsevier Applied Science Publishers, Essex, England, pp Tubb, R.S., B.A. Searle, A.R. Gdey and A.J.P. Brwn (1981). Rare mating and transfrmatin fr cnstructin f nvel brewing yeasts. Eur. Brew. Cnv. Prc. Cng. l8th, Cpenhagen. pp Tuite, M.R., M.J. Dbsn, N.A. Rberts, R.M. King, D.C. Burke, S.M. Kingsman and A.J. Kingsman (1982). Regulated high effrciency expressin f human interfern-alpha in Saccharmyces cerevisae.emb J. 1: Ussegli-Tmasset, L., G. Cilfi, and A. Pagliara. (1981). Estimating sulphur dixide resistance f yeasts. I: The delaying actin n fermentatin start. Vini d'ital. 23:78-90.

206 -152- van Slingen, P. and J.B. van der Plaat (L977). Fusin f yeast spherplasts. J Bacteril. 130: van Uden, N. (1985). Ethanl txicity and ethanl tlerance in yeasts. Ann. Rep. Ferment. Prc. S: van Uden, N. (1984). Temperature prfiles f yeast. Adv. Micrb. Physil.2S:195-25r. van Vuuren, H.J.J. and B.D. V/ingfield (1986). Killer yeasts - cause f stuck fermentatins in a wine cellar. S. Afr. J. Enl. Vitic. 7: van Vuuren, H.J.J. and L. van der Meer (1987). Fingerprinting f yeasts by prtein electrphresis. Am. J. Enl. Vitic. ü: Vaughan, A.E. and A. Ma tini. DNA base cmpsitin and DNA-DNA reassciatin f the "wine" species f the genus Saccharmyces Hansen. VI Int. ferm. Symp./V Int. Symp. Yeast, Lndn (Canada). Vezinhet, F. (1935). Le marquage genetique de suches de levures enlgiques. Rev. Fr. Oenl. 97: Vezinhet, F. and S. Lacrix (1934). Marquage genetique de levures:util de cntrle des fermentatins en suche pure. Bull. OIV 57: Vezinhet, F., B. Blndin and J. Hallet (1990). Chrmsmal DNA patterns and mitchndriat DNA plymrphism as tls fr identificatin f enlgical strains f S ac c harmy c es cerev isiae. Appl. Micrbil. Bitechnl. 32: l. Viljen, 8.C., J.L.F. Kck, M. Miller, and D.J. Cetzee (1989). The value f rthgnal-field-alternatin gel electrphresis and ther criteria in the deliminatin f anmrphic/telmrphic relatins. System. Appl. Micrbil. 1 1: Vllrath, D. and R.\il. Davis (1987). Reslutin f DNA mlecules greater thatn 5 megabases by cntur-clamped hmgenus electric fields. Nucleic Acids Res.!l:

207 -153- Watari, J., Y. Takata, M. Ogawa, N. Nishikawa and M. Kamimura (1989). Mlecular clning f a flcculatin gene in Saccharmyces cerevisiae. Agnc. Bil. Chem.53: webb, A.D. and J.L. Ingraham (1963). Fusel il: a review. Adv. Appl Micrbil. 5: 'Webster, T.D. and R.C. Dicksn (1983). Direct selectin f Saccharmyces cerevisiae resistant t the antibitic G418 fllwing transfrmatin with a DNA vectr carrying the kanamycin resistance gene f Tn903. Gene X: Vy'ickner, R.B. (1974). "Killer chatacter" f Saccharm]'ces: curing by grwth at elevated temperatures. J. Bacteril. 717 : 'Wickner, R.B. (1983). Genetic cntrl f replicatin f the duble-stranded RNA segments f the killer systems in Saccharmyces cerevisiae. Arch. Bichem. Bihvs. 222: l-ll. Williams, S.4., R.A. Hdges, T.L. Strike, R. Snw and R.E. Kunkee (1984). Clning the gene fr the mallactic fermentatin f wine frm Inctbacilltts delbruckii in Escherichia cli and yeasts. Appl. Envirn. Micrbil. 47: V/inge, O. (1935). On haplphase and diplphase in sme Saccharmycetes. C.R. Lab. Carlsberg Ser. Physil. 22: 77-1I2. Winge, O. and C. Rberts (1958). In: H. Ck (ed.), The Chemistry and Bilgy f Yeasts, Academic, New Yrk, pp Wdruff, H.B. (1980). Natural prducts frm micrrganisms. Science 208: Xia, W. and G.H. Rank (1989) The cnstructin f recmbinant industrial yeasts free f bacterial sequences by directed gene replacement int a nn-essential regin f the genme. Gene 76: Yadav, N., R.E. McDevitt, S. Benard and S.C. Falc (1986). Single amin acid substitutins in the enzyme acetlactate synthase cnfer resistance t the herbicide sulfmeturn methyl. Prc. Natl. Acad. Sci. USA 83:44L

208 -154- Yamashita, I. and S. Fukui (1983). Mating signals cntrl expressin f bth starch fermentatin genes and a nvel flcculatin gene FLOS in the yeast Sacccharmyces. Agric. Bil. Chem. 47: Ykmri, Y., H. Akiyama and K. Shimizu (1989). Breeding f wine yeasts thrugh prtplast fusin. Yeast (Spec. Iss.) pp. S Yung, T.W. (1981). The genetic manipulatin f killer character int brewing yeasts. J. Inst. Brew. 87 : Yung, T.W. (1983). The prperties and brewing perfrmance f brewing yeasts pssessing killer cha acter. J. Am. Sc. Brew. Chem. 4l: l-4. Yung, T.W. and M. Yagiu. (1978). A cmparisn f the killer character in different yeasts and its classificatin. Antnie van I-eewenhek J. Micrbil.44: Zambnelli, C. (1964a). Ricerche bimetriche sulla prduzine di idrgen slfrat da slfati e slfiti in Saccharmyces cerevísiae var. ellipsideus. Ann. Micrbil. 14: 129- t4r. Zambnelli, C. (1964b). Ricerche genetiche sullla prduzine di idrgen slfrat in Saccharmyces cerevisiae var. ellipsideu.s. Ann. Micrbil. 14: Zambnelli, C. (1965a). Richerche genetiche sulla prduzine di idrgen slfrat in Saccharmyces cerevisiae var. ellipsidew. II. Ereditarieta del carattere dal punt di vista quantitativ. Ann. Micrbil. 15: Zambnelli, C. (1965b). Ricerche genetiche sulla prduzine di idrgen slfrat in Saccarmyces cerevísiae vw. ellípsidew. tri. Ulteriri studi sulla eredita ieta del carattere quantitativ. Ann. Micrbil. l5: Zambnelli, C. (1965c). Stabilita ed eredita ieta delle variazine nella prduzine de H2S indtte dal nitrat fenil-mercuric in Saccharmyces cerevisae var. ellipsideus. Ann. Micrbil. 15:

209 -155- Zambnelli, C., P. Mutinelli and G. Pachetti (1975). Bisynthesis f sulphur amin acids \n Saccharmyces cerevisiae. I. Genetic analysis f leaky mutnts f sulphite reductase. Arch. micrbil. 102: Zhu, J., R. Cntreras, D. Gheysn, J. Ernst and W. Friers (1985). A system fr dminant transfrmatin and plasmid amplificatin in Saccharmyces cerevisiae. Bitechnl. 3: Zimmerman, F.K. (1978). On the use f systematically selected yeasts in large wineries. Prc. 5th Int. Oenl. Symp., Aukland. Zimmerman, F.K. (1990). Genetics and stabilizatin f yeast fr still and sparkling wine prductin. Prc. 9th Int. Oenlgical sympsium, Prtugal. pp Zimmermann, U. and J. Vienken (1982). Electric field-induced cell-t-cell fusin. J. Membr. Bil. fl:

210 -156- Appendix 1 Publicatins

211 Petering J., Langridge, P. & Henshke, P. (1988). Fingerprinting wine yeasts: the applicatin f chrmsme electrphresis. Australian & New Zealand Wine Industry Jurnal, 3(3), NOTE: This publicatin is included in the print cpy f the thesis held in the University f Adelaide Library.

212 Petering, J. E., Henschke, P. A. & Langridge, P. (1991). The Escherichia cli ß-glucurnidase gene as a marker fr Saccharmyces yeast strain identificatin, American Jurnal f Enlgy and Viticulture, 42(1), NOTE: This publicatin is included in the print cpy f the thesis held in the University f Adelaide Library.

213

214 Apprr e.r Er.rv rnr.r M ENTAL Mr cnsr r-cv, Nv. 1991, p I 9I I II $02, 00/0 Cpyright 1991, American Sciety fr Micrbilgy Vl. 57, N. 11 Determinatin f Killer Yeast Activity in Fermenting Grape Juice by Using a Marked Saccharmyces Wine Yeast Strain JENNY E. PETERING,I MICHAEL R. SYMONS,II PETBN LANGRIDGE,'NNO PAUL A. HENSCHKE'* Department f Plant Science, Waite Agricultural Research Institute, The University f Adelaide, Suth Australia 5000,r and The Australian Wine Research Institute, P.O. Bx 197, Gten Osmnd, Suth Australia 5064,2 Australia Received 7 June 199l/Accepted 26 August 1991 'lhe Escherichia cli þ-ghrcrtrnidase gene has been used as a marker gene t mnitr a killer Saccharmyces cerevisiøe strain in mixed-culture ferments. The marked killer strain was cured f its M-dsRNA genme t enable direct assessment f the efficiency f killer txin under fermentatin cnditins. Killer activity was clearly evident in fermenting Rhine Riesling grape juice f ph 3.1 at l8'c, but the extent f killing depended n the prprtin f killer t sensitive cells at the time f inculatin. Killer activity was detected nly when the rati fkiller t sensitive cells exceeded lz2. At the highest rati fkiller t sensitive cells tested (2:l), cmplete eliminatin f sensitive cells was nt achieved. Killer activity in yeasts was first reprted fr strains f Saccharmyces cerevisiae in 1963 by Bevan and Makwer (2). Killer yeasts secrete plypeptide txins which kill sensitive strains f the same genus and, less frequently, strains fdifferent genera (20, 30). Previus studies indicate that the txin f Saccharmyces cerevisiae is a prtein which binds t a receptr n the wall f the sensitive yeast cell, disrupting the electrchemical gradient acrss the cell membrane and hence the intracellular inic balance (6, 28). Prductin f the txin and immunity t it are determined by a cytplasmically inherited duble-stranded (ds) RNA plasmid, therwise knwn as the M genme (4). The M-dsRNA killer plasmids are dependent satellites f L-AdsRNA, and L-BC-dsRNA exists as a species unrelated t the first r t M. All types f dsrna exist in virus-like particles and require a prtein encded by the L-A-dsRNA fr encapsidatin (4, 12, 30). On the basis f the prperties f the txin, killer yeasts have been classified int 11 grups (K, thrugh K11) (18, 33). Thse unique t Saccharmyces strains fall int the first three grups (Kr, Kr, and K.). The Saccharmyces txin is reversibly inactivated at lw ph (2.0) and irreversibly inactivated at ph in excess f 5.0 (33). Mre specifically, the bilgical activity f K, is ptimal between ph 4.6 and 4.8, while K, shws ptimal activity between ph 4.2 and 4.7 (27). The K, txin is stable ver a wider ph range than the K, txin (2.8 t 4.8) (23) and is therefre mre relevant in wine fermentatin. Killer activity has been detected in yeasts islated frm established vineyards and wineries in varius regins f the wrld, including Eurpe and Russia (1, 9, 18), Suth Africa (31), and Australia (14, 15). This widespread ccurrence has prmpted interest in the enlgical significance f killer wine yeasts. In thery, selected killer yeast strains culd be used as the inculated strain t suppress the grwth f undesirable wild strains f S. cerevisiae during grape juice fermentatin. In additin, as killer interactins have been reprted t ccur between yeasts f different genera (21, 25), the * Crrespnding authr. T Present address: Petaluma Pty Ltd., Piccadilly, Suth Australia 5151, Australia. pssibility f genetically engineering brad-spectrum killer strains f S. cerevisia exists (3). Studies have been cnducted t assess the efficiency f killer txin n sensitive yeast strains. Hwever, reprts n the expressin f kilter activity under fermentatin cnditins have been cntradictry (5, 7, 16). Attempts t determine the ppulatin kinetics f killer and sensitive strains during wine fermentatin have been restricted because f the difficulty invlved in identifying the tw types when they are grwn in mixed cultures. Appraches used t date include (i) chice f killer and sensitive strains that can be distinguished by their grwth rates (1) r their prductin f hydrgen sulfide (24), (ii) use f auxtrphic and respiratry-deficient mutants f killer strains and apprpriate plating cnditins under which they can be identified (10, 11,26), (iii) use f killer and sensitive strains which can be distinguished by differences in clny mrphlgy (14), and (iv) assaying clnies directly fr killer activity (17). All f these methds are limited by the fact that the assays invlved are labrius and time-cnsuming r that nly killer strains with specific characteristics can be studied. We describe here the use f a marked S. cerevisiae killer strain in a mixed-culture inculum t quantify directly the effect f killer txin n a sensitive S. cerevisiae strain under fermentatin cnditins. As a wide range f yeast strains can be readily and stably marked, this system f analysis is unlimited in applicatin and prvides a simple and unequivcal means f quantifying killer yeast strains in mixedculture ferments. MATERIALS AND METHODS Strains and media. Sensitive S. cerevisiae strains 5A (AV/RI 138) and 2A (AWRI 729) and killer (Kr) strain 114 (AWRI 92F) were btained frm the Australian Wine Research Institute. Generatin f the marked killer (Kr) strain 3AM (AWRI 796) has been previusly described (19). Yeast grwth medium was YPD (I% yeast extract [Difc], 2% Bact Peptne [Difc] and2v glucse). Curing f killer strain 3AM. A culture f strain 3AM was grwn vernight in YPD al28"c. Serial dilutins were made in0.9% NaCl, and 0.1-ml aliquts (cntaining apprximately 100 cells) were spread n YPD plates and incubated at3'7"c. 3232

215 Vr. 57, 1991 YEAST KILLER ACTIVITY IN FERMENTING GRAPE JUICE 3233 After 48 h f incubatin, single clnies were selected at randm and assayed fr killer activity as described belw. Assay fr cured strain. YPD (cntaining l% agar) was autclaved at 120'C fr 20 min. After cling t 49"C, the medium was buffered t ph 4.2 with a l0% tartrate slutin. Methylene blue (t 0.003% [wt/vl]) and killer-sensitive strain 5A (t 105 cells per ml) were added t the medium befre the medium was pured int the plates. Clnies islated after heat treatment were then transferred t these assay plates and incubated at 18"C fr apprximately 72 h. Curing was recgnized by the absence f grwth inhibitin (clear znes) and the lack f blue-stained cells arund the clny. dsrna islatin. The dsrna extractin prcedure was essentially that deseribed by Fried and Fink (8). Samples f RNA were analyzed by electrphresis n 1.5Ø agarse slab gels at a cnstant current f 100 ma. Gels were stained with ethidium bl-mide and phtgraphed n a shrtwave UV light bx. Fermentatin trials. Starter culiures were prepared by inculating 10 ml f YPD medium cntained in a cnical flask with a ipiul í yeasi and incubate<i with vigrus aeratin at 28"C. After 24 h, the cell density was determined by micrscpic cunts. Samples were used t inculate Rhine Riesling must (200 ml) t a density f 5 x 106 cells per' ml. The must cntained220 g f reducing sugars per liter and had a ph f 3.1. Fermentatins were carried ut in 250-ml cnical flasks fitted with airlcks. The juice was sterilized by membrane fiitratin (0.45-f M pre size) prir i inculatin, and fermentatins were carried ut at 18'C with agitatin (apprximately 100 scillatins per min). Samples were remved anaerbically and aseptically during fermentatin by needle and syringe thrugh prts cvered with rubber septa. Samples were analyzed fr the prgress f fermentatin by refractmeter readings, and yeast grwth was measured spectrphtmetrically at 650 nm. Fr analysis f the prprtin f malked strain in the yeast ppulatin, serial dilutins f the samples were made in sterile 0.9% NaCl, and 0.1-ml aliquts (cntaining 200 t 500 cells) were plated n YPD media. The plates were then assayed as described belw. B"Glucurnidase (GUS) platc assays. Yeast clnies were grwn n slid YPD medium fr apprximately 36 h at 28"C. A slutin cntaining 0.1 M NarHPO (ph 7.0), l% sarcsyl, 5-brm-4-chlr-3-indlyl glucurnide (100 t 150 pg/ml), and 0.77 agarse was then pured as a thin verlay n the plate and allwed t set. After4 t 6 h f incubatin ai 37"C, a blue precipitate culd be detected in the marked clnies. RESULTS Curing f strain 3AM. In rder t specifically analyze Íhe effect f killer txin in fermentatins, an experiment was designed t cmpare tw isgenic strains which differ nly in the presence f the M-dsRNA genme and therefre in their ability t prduce killer txin. Killer strain 3AM has previusly been marked with the Escherichia cli GUS gene (19). This system allws the marked strain t be readily identified in a mixed ppulatin by a simple plate assay, which results in the frmatin f a blue precipitate in marked clnies. Strain 3AM was cured f its M-dSRNA plasmid by heat treatment (32), the cured r' sensitive clnies being identified by killer activity plate assays. Figure 1 shws the respnse f strain 3AM and an islated cured derivative (designated 3AMC) t the killer ph 4.2!-F FIG. t. Agal plate assay fr killer activity. The agar'(ph 4.2, 0.003% methylene blue) is seeded with an vernight culture f strain 3AMC, and strains t be tested fr killer activity are patched nt the slid media. 114 is a knwn killer strain, and 2A is a knwn sensitive strain. Strain 3AM displays a respnse identical t that f kitler 11,{, with a clear zne and methylene blue-stained brder arund thc patch f gr'wth. activity plate assay. The zne f inhibitin evident arund strain 3AM is absent arund 3AMC, indicating that strain 3AMC is nt prducing killer txin. GUS activity was detected in stl'ain 3AMC by the agar plate methd (results nt shwn), indicating that the cured strain is an authentic derivative f stlain 3AM. Finally, cisrna species were isiated frm stlains 3AM and 3AMC and analyzed by standard electrphresis techniques (Fig. 2). A band representing the M-dsRNA genme is present in strain 3AM and absent in strain 3AMC. Fermentatin trials were then perfrmed with strains 3AM and 3AMC t determine the effect f the curing prcedure n yeast grwth and fermentatin rates. Starter cultures feach strain were inculated in triplicate int flasks f Rhine L.51 L õ ú ÊU,F> :1{< 114 3AMC DNA LdsRNA < MdsRNA < trna FIG. 2. Electrphresis f dsrna species frm killer strains 114 and 3AM and sensitive strains 2A and 3AMC. Cntaminating DNA and trna species ale als present.

216 J234 PETERING ET AL. Appl. ENvlnr.. Mrcnsrr A 20 '15 A E 10 E! 'l 6 5 c Tlme (hurs) Tlme (hurs) 300 x! s 210 ' B 20 't5 B cé c È B Tlme (hurs) FIG. 3. Yeast grwth (A) and sugar utilizatin (B) curves f strains 3AM (O) and 3AMC (O). Riesling grape juice at a cncentratin f 5 x 106 cells per ml. Samples were taken at regular intervals and assayed fr yeast grwth and prgress f fermentatin. The average readings fr each strain were pltted ver time (Fig. 3). There are n significant differences in the grwth r fermentatin rates between strains 3AM and 3AMC. Analysis f killer activity during fermentatin. Strains 3AM and 3AMC were analyzed fr killer activity in Rhine Riesling juice by cinculating each strain with the sensitive.!. cerevisiae strain 5A. Cntrl ferments f each strain (34M, 3AMC, and 5A) as pure incula were als perfrmed. Each ferment was cnducted in duplicate at 18'C with gentle agitatin under anaerbic cnditins. GUS plate assays were then perfrmed t identify the marked strain (3AM r 3AMC). Clnies f the marked strain turn a deep blue clr as a result f this assay, allwing simple identificatin. GUS plate assays were als perfrmed n the cntrl ferments t cnfirm the validity fthe assay. Plate assays n the cntrl 5A ferment were cnsistently negative, highlighting the absence f backgrund GUS activity in natural yeast cells. Hwever, cntrl 3AM and 3AMC ferments gave values fbetween 99 and I00V fttal clnies per plate fr the marked strain cunt. This bservatin represents a reversin frequency f less than IV fr the GUS gene. The fllwing mixed-culture ferments were carried ut: (i) 3AM and 5A at an inculum rati f 1:1, (ii) 3AMC and 5A at an inculum rati f 1:1, (iii) 3AM and 5A at an inculum rati f 2:1, and (iv) 3AMC and 5A at an inculum rati f 2:1. These mixed ferments exhibited nrmal grwth kinetics, as did the three cntrl ferments (Fig. a). The time curse f grwth (CFU per milliliter) f each strain in the mixed-culture ferments is pltted in Fig. 5. At an inculum rati f 1:1, there was a ntable increase in the prprtin f killer strain 3AM, whereas the cured strain TlmE (hurs) FIG, 4. (A) Grwth curves f cntrl single mnculture ferments. Symbls: O, 3AM; O, 3AMC;!, 54. (B) Grwth curves f mixed-culture ferments. Symbls: a, 3AM and 5A at an inculum rati f 2:1; O,3AMC and 5A at an inculum rati f 2:1; l,3am and 5A at an inculum rati f 1:1; A, 3AMC and 5A at an inculum rati f 1:1. 3AMC failed t exert any dminance ver the sensitive strain under therwise identical cnditins. Statistical analysis was used t test the null hypthesis that the rati f killer t sensitive cells remains 1:1 thrughut the ferment. A gdness ffit test (nrmal test) rejected the null hypthesis, with P < Hwever, identical analysis f the cured t sensitive strain ferment rati accepted the null hypthesis that the rati f the tw strains remains at 1:1 thrughut the ferment. With an increased prprtin f strain 3AM in the inculum (rati 2:1), the dminating effect f strain 3AM was mre prnunced. It is imprtant t nte that strain 5A persisted, albeit at lw levels, thrughut the ferments. Experiments were cnducted t determine the lwest inculum rati f killer t sensitive cells at which significant killer activity can be bserved. Mixed ferments f strain 3AM and 5A at inculum ratis f 1:2 and 1:4, respectively, were carried ut under the cnditins described abve. N change frm the initial prprtin f strain 3AM was detected in either f these ferments. The results f all mixedculture ferments invlving strain 3AM are summarized in Fig. 6. DISCUSSION Previus studies have indicatedl00% stability f the GUS marker gene in strain 3A thrughut fermentatin (19). Hwever, analysis f a larger sample f clnies in these experiments has revealed an instability f the cnstruct. This instability was detected in the cntrl fermentatins which were inculated with mncultures f either the marked

217 Vr. 57, 1991 YEAST KILLER ACTIVITY IN FERMENTING GRAPE JUICE 3235 I E f c 6 f e È 6 â E J q I tr I ga Ê Â Ø 6 'tb 'r J r4 r3 't2 'l I r7 r6 10" r4 r 'I Tlm (hurs) Tlme (hurs) A c E a c I!5 È 6 E f 6 a Â Õ 6 rb 'tt 106 r5 104 'l3 102 ' Tlm6 (hurs) t8 l7 r6 105 r Tlm (hurs) FIG. 5. Grwth curves f each strain in mixed-culture ferments expresseci as CFU per miliiliter. (A) Ìvíixeci íermeni i3âìr{ (O) ai-rd 5A (D) ar ar-r inculum rati f 1:1. (B) Mixed ferment f 3AMC (O) and 5A (n) at an inculum rati f 1:1. (C) Mir-erl fermenr f 3AM () and j,f ( ) ut an inculum ati f 2:1. (D) Mixed ferment f 3AMC (O) and 5A (!) at an inculum rati f 2:1. B D 300 strain 3AM r 3AMC. Samples frm these fermentatins gave rise t clnies which respnded negatively t the GUS plate assay at a frequency f less than lv f the ttal plate ðunt. Occasinally, a clny which was sectred in its respnse t the assay was detected, suggesting either excisin f the gene by hmlgus recmbinatin (29) r lss f the gene after mittic crssing-ver (22). The frequency f instability did nt increase ver time during fermentatin and culd be directly quantified in the cntrl 3AM and 3AMC f'erments. This marking system has enabled a direct cmparisn t be made between the inculatin efficiency f a killer strain (3AM) and an isgenic cured derivative (strain 3AMC) in fermenting grape juice. At a l'ati f killer t sensitive cells f 1:1, the cured strain, 3AMC, remained ã1 50% f the ttal ppulatin, while the killer strain increased t The ability f strain 3AM t dminate strain 5A during fermentatin is likely t be due t the prductin f kilter txin by strain 3AM and nt t a difference in respective grwth rates favring the killer strain. We can cnclude, therefre, that the killer txin has displayed significant activity under these fermentatin cnditins. This result is f particular interest t the enlgist, since the K, txin prduced by strain 3A is reprted t shw maximum activity at ph 4.2 (23), which is 0.5 t 1 ph unit higher than generally fund in grape musts. In cases in which killel activity in fermenting grape juice has been reprted, a discrepaney as t whether effective killing actin ccurs when the prprtin f killer cells is less than507 f a mixed-culture ferment exists. Heard and Fleet (14) did nt bserve killer actin when the rati f killer t sensitive cells was apprximately 1:7, whereas thers have reprted killer activity with killer-t-sensitive-cell ratis f 1:10 and lwer (1, 10, i1). Our results shwed that an increase in the rati f killer t sensitive cells t apprximately 2:1 resulted in a prnunced dminance f the fermentatin by strain 3AM t 97% f the ttal mixed ppulatin by the end f the fermentatin. Hwever, with killer-t-sensitive-cell ratis f l'.2 r l'.4, n effective killer actiu was evident. It is pssible that differences in either cmpsitin f medium, fermentatin cnditins, r strain sensitivity may accunt fr discrepancies in reprts f killer txin efficiency. The relevance f killer strains in wine making has been the fcus f attentin in cuntries where selected yeast cultures are inculated int musts t induce fermentatin. 'l'his tbcus has intensified since the bservatin that yeasts which are naturally present in the must als play significant rles in suppsedly "pure" culture fermentatins (13, 16). These natural yeasts include species frm the genera Klecker, C ct nd i du, H a ns e n u I a, and S ac c har my c e s. Killer S ac c h arc g ) È ã s 100 BO tlme (hurs) FIG. 6. Time curse in the prprtin f killer strain 3AM t strain 5A in the ttal ppulatin f a mixed-culture fet'ment fr inculum ratis f 2:1 (n), 1:1 (O), 1:2 (l), and 1:4 (A).

218 3236 PETERING ET AL. myces wine yeast strains may be effective in suppressing natural Saccharmyce.r yeast strains during fermentatin, and the pssibility f engineering brad-range killer yeast strains t cntrl strains frm ther genera exists. Fr these reasns, further study is needed t determine apprpriate fermentatin cnditins fr effective killer activity. The GUS marking system prvides a methd which allws a brad range f killer strains t be rapidly and unequivcally identified in a mixed culture. This system can be emplyed t gain a better understanding f killer activity during fermentatin. ACKNOWLEDGMENTS J.E.P. gratefully acknwledges the receipt f a schlarship frm The Australian Wine Research lnstitute. This wrk was funded by the Grape and Wine Research Cuncil f Australia. We thank Brian Crser, Petaluma Pty Ltd., fr supplying the Rhine Riesling grape juice. REFERENCES 1. Barre, P Le mecanisme killer dans la cncurrence entre suches de levures. Evaluatin et prise en cmpte. Butl. O.l.V Bevan, E. 4., and M. Makwer The physilgical basis f the killer character in yeasts, p ^ln S. J. Gert (ed.), Genetics tday. Prceedings fthe 11th lnternatinal Cngress f Genetics, vl Bne, C., A.-M. Sdicu, J. Wagner, R. Degré, C. Sanchez, and H. Bussey. 1990, Integratin f the yeast Kl killer txin gene int the genme f marked wine yeasts and its effect n vinificatin. Am. J. Enl. Yitic. 413' Bstian, K. A.., J. E, Hpper, D. J. Rgers, and D. J. Tipper Translatinal aalysis f the killer-assciated virus-like particle dsrna genme f Saccharmyces cerevisiae: M-dsRNA encdes txin. Cell 19': Cuinier, C., and C. Grs, Enquete sur la repartitin des levures "killer" en France. Vignes Vìns 318: De la Peña, P., F. Barrs, S. Gascón, P. S. Laz, and S. Rams Effect f yeast killer txin n sensitive cells f Saccharmyces cerevisi r. J. Bil. Chem. 256:1042U Delteil, D., and T, Aizac, Cmparisn f yeast inculatin techniques by the use f a'marked' yeast strain. Aust. N.Z. Wine lnd. J. 3(3): Fried, H. M., and G. Fink Electrn micrscpe heterduplex aalysis f "killer" duble-stranded RNA species frm yeast. Prc. Natl. Acad. Sci. USA Ç Gaia, P Lieviti in assciazine cntempranea. Vini d'ltalia 26t Hara, S., Y. Iimura, and K. Otsuka Breeding f useful killer wine yeasts. Am. J. Enl. Vitic. 32: Hara, S., Y. Iimura, H. Oyama, T. Kzeki, K. Kitan, and K. Otsuka Breeding f cryphilic killer wine yeasts. Agric. Bil. Chem. 45t1327-\ Harris, M. S Virus-like particles and duble-stranded RNA frm killer and nn-killer strains f Saccharmyces cerevlslae. Micrbi lgy 2lz16l Heard, G. M., and G, H. Fleet Grwth f atural yeast flra during the fermentatin f inculated wines. Appl, Envirn. Micrbil '7J28. Appl. ENvlnr.. MlcnnlL. i 14. Heard, G. M., and G. H. Fleet Occurrence and grwth f killer yeasts during wine fermentatin. Appl' Envirn. Micrbil. 53z2I7I2Il Heard, G. M., and G. H. Fleet The ccurrece f killer cbaracter in yeasts during the fermentatin f Australian wines. Aust. N.Z. Wine Ind. J. f(4): Lafn-Lafurcade, S., and P. Ribereau-Gayn Develpments in the micrbilgy f wine prductin' Prg. Ind' Micrbil. 19: Lng, E., J. B. Velazquez, J. Cansad, P. Cal, and T. G. Villa Rle f killer effect in fermentatins cnducted by mixed cultures f Saccharmyces cerevisiae. FEMS Micrbil' Lett. 7l: Naumv, G. I., and T. I. Naumva, Cmpalative genetics f yeast cmmunicatin. XIII. Cmparative study f killer strains f Saccharmvces frm different cllectins. Genetika 9: Petering, J. E., P. A. Henschke, and P' Langridge The Escherichia cll p-glucurnidase gene as a marker fr Saccharmyces yeast strain identificatin. Am. J. Enl. Yilic' 42:Ç Philliskirk, G., and T. W. Yung' The ccurrence f killer character in yeasts f varius genera. Antnie van Leeuwenhek J. Micrbil. 4l: Radler, F., P. Pfeiffer, and M. Dennert Kitler txins in new islates f the yeasts Hanseniaspra uvúrum and Pichia kluyveri. FEMS Micrbil. Lett. 29: Reder, G. S., and S. E. Stewart Mittic recmbinatin in yeast. Trends Gene. 4: Rgers, D., and E. A, Bevan Grup classificatin f killer yeasts based n crss reactins between strains f different species and rigin. J. Gen. Micrbil, 105: Rsini, G Effet d'une levure "killer" de Saccharmyces cerevisiue sur une suche de levure sensìble de la même espèce, nn prductrice de HrS et sélectinée pur la vinificatin dans un milieu de culture mixte. Bull. O.l.Y. 58: Rsini, G Interactin between killer strains f Hansenula anmala var. unmala and Sncchurtnyces cerevisiae yeàsf species. Can, J. Micrbil. 3l: Seki, T., E.-H. Chi, and D. Ryu Cnstructin f killer wine yeast strain. Appl. Envirn. Micrbil' 49:I2ll-I2I5' 27. Shimazu, K., T. Adachi, K. Kitan, T. Shimazaki, A. Ttsuku, S. Hara, and H. H. Dittrich Killer prperties f wine yeasts and characterisatin f killer wine yeasts. J. Ferment. Technl' 63:' Skipper, N., and H. Bussey Mde f actin f yeast txins: energy requirement fr Succharmyces cerevisiae killer txin. J. Bacteril. 129: Struhl, K., D, T, Stinchcmb, S. Scherer, and R. W. Davis. 1979' High-frequency transfrmatin f yeast: autnmus replicatin f hybrid DNA mlecules. Prc. Natl' Acad' Sci. USA 76: Tipper, D. J,, and K. A. Bstian Duble-stranded ribnucleic acid killer systems in yeasts. Micrbil. Rev' 48: Tredux, H. G., R. P. Tracey, and A. Trmp' Killerfactr in wine yeasts and its effect n fermentatin. S. Afr' J. Enl. Vitic. 7: Wickner, R, B "Killer character" f Saccharmyces cerevisiue', curing by grwth at elevated temperature. J' Bacteril. l17:135g Yung, T, W., and M. Yagiu A cmparisn f the killer character in different yeasts and its classificatin. Antnie van Leeuwenhek J. Micrbil. 44t59-77.