COMMUNICATIONS DETERMINATION OF YEAST VIABILITY. By R. B. Gilliland, B.A., B.Sc, A.R.I.C.

Transcription

1 424 [J. Inst. Brew. COMMUNICATIONS DETERMINATION OF YEAST VIABILITY By R. B. Gilliland, B.A., B.Sc, A.R.I.C. (Arthur Guinness Son & Co. (Dublin) Ltd., St. James's Gate, Dublin) Received 3th April, 59 A new technique for the determination of yeast viability, by cultivation on haemacytometer slides followed by counting the microcolonies, gave a demonstrably true estimate of the percentage of cells which were able to reproduce. By com parison with this method, staining techniques correctly estimated the percentage of viable cells In fresh yeasts, but overestimated the percentage of viable cells in old yeasts. Gelatin plate counts always underestimated the percentage viability. With some brewery yeasts it was possible to demonstrate by slide culture that the dilution necessary for a plate count was lethal to a large proportion of the cells. The gelatin plate method therefore gave very low and erroneous results for the viability of these yeasts. Slide culture was also found to be useful for the enumeration of chain-forming yeasts in a mixed culture. Introduction In the routine examination of brewery yeast, it is usual to make an estimate of viability. This is normally determined either by microscopical examination of the yeast on a plain slide, or by examination of a suitable suspension of the yeast on a haemacytometer slide. In the latter case the numbers of presumed living and dead cells may be counted and the percentage viability esti mated. A dye such as methylene blue, methyl green, eosin, acridine orange, rhodamine B, neutral red, or erythrosine may be added to the yeast suspension to make the distinction between living and dead cells easier to detect.6'8'8''2-4 By these methods, the viability of a brewery yeast is normally found to lie in the range of 9-99%. Yeast viability may also be determined by plate counts on wort-gelatin or wort-agar. The total number of cells added to a plate can be estimated from haemacytometer counts on the yeast suspension used for inoculating the plate, and the numbers of these cells which are viable can be obtained by counting the colonies produced after incubation. This method is seldom used for brewery routine determination of viability as it is too slow, taking four days to complete. Another method of determining viability is by isolating single cells into droplets of wort by means of a micromanipulator and later examining the droplets for growth of yeasts. Both the plate method and the droplet method may give viability percentages much lower than are given by direct microscopical examination. Mills2 noted that the plate viability on wort-agar was about 5% lower than that yielded by staining, and he com ments that the former method indicates only repro cells whereas the latter differ entiates all viable cells from those that are dead. Hall,6 using the droplet culture method, found that the viability thus obtained was in most cases markedly lower than the viability obtained by methylene blue staining. Some brewery yeasts which she received at the Brewing Industry Research Foundation had very low viabilities as determined by single cell isolations and, as storage was found to have relatively little effect on viability under controlled conditions, it was inferred that the yeasts probably had low initial viabilities, although deterioration during transit to the laboratory could not be completely ruled out. Oppenoorth4 found that of the cells unstained by methylene blue 34% would not grow when isolated by micromanipulator into wort droplets; using rhodamine B, 9% of the unstained cells did not grow. Other writers have mentioned low viabilities obtained by the gelatin plate method, and from personal enquiry it has been learned that in some English breweries

2 Vol. 65, 959] gilliland: determination of yeast viability 425 the gelatin plate viability of the brewing yeast may be very low but that this is not regarded as any cause for alarm. The strange anomaly between gelatin plate viability and apparent viability by micro scopical examination has been investigated in a brewery yeast in which the plate via bility was low but which otherwise behaved normally. A new method of determining yeast viability, which gave a demonstrably true estimate of the ability of yeast to reproduce, was devised for this work. Experimental Estimation of viability by standard methods. The viability of a brewery yeast was determined by the following methods: Microscopical observation: The unstained cells were examined microscopically on haemacytometer slides. The dead cells were judged by their darker outline and by their granulated appearance. The percentage of such cells was estimated and subtracted from to give the percentage viability. Methylene blue staining: A suspension of the yeast was mixed with an equal volume of Burns' methylene blue solution3 (2 g. sodium citrate, ml. % aqueous methylene blue, water to ml.) and the mixture was examined on haemacytometer slides. The percentage of unstained cells was estimated. Single cell isolation: With the micromanipulator, cells were isolated from a suspension of the yeast and these cells were inoculated singly into droplets of sterile wort in a moist chamber. The percentage of single cells which produced colonies in the droplets after incubation for two days at 8 C. was determined. Gelatin plate counts: The total number of cells per ml. in a suspension of the yeast was estimated from counts on haemacytometer slides. Dilutions of this suspension were added to melted wort-gelatin and petri dish cultures were made. The colonies which had appeared on the plates after four days' incubation at 8 C. were counted. The viability was estimated as the number of colonies produced from a total seeding of cells. Serial dilution: A series of -fold dilutions in wort of a counted suspension of the yeast was made up. Each dilution was used to inoculate ml. into each of ten -ml. bottles of sterile wort. From the number of bottles at each seeding rate which showed growth after four days' incubation at 25 C. the number of viable cells per ml. in the original suspension was calculated by means of Halvorson & Ziegler's probability Tables. Count of budding cells: The yeast was inoculated at the rate of g. per litre into wort. The inoculated wort was incubated for 4 hr. at 25 C. and then used to make up haemacytometer slides. The percentage of cells which had buds was estimated. The results of these determinations are given in Table I. TABLE Viability of a Brewery Yeast as Determined by Various Methods Method or estimation Hacmncytomcter count of apparently normal cells Hacmncytomctcr count or cells unstained by mcthylcnc blue.. Isolation or single cells in droplets of wort Wort-gclatln plate Dilution scries in wort.. Haemncytomctcr count of budding cells after 4-hr, growth in wort I Percentage viability OS From the point of view of the brewer the truest and most important estimation of viability is that obtained by counting the percentage of cells which produce buds in wort. This is, however, a difficult deter mination to carry out satisfactorily as the result depends on the time at which the count is done; if it is done too early some cells capable of reproduction have not yet produced a bud, but if it is done too late the count is influenced by the presence of second generation cells. The haemacytometer counts give an approximation to the result obtained by bud-counting, but it is obvious that gelatin plates, droplet isolations or serial dilutions have given grossly incorrect esimates of viability. Variation in plate viability during a brewery fermentation. In order to find whether the gelatin plate viability varied during a brewery fermentation, viabilities were determined at different stages. The results recorded in Table II show that plate viability is at a maximum during active fermentation. This confirms the findings of Hall,6 who estimated viability of yeasts during a brewing by single cell isolations and found that viability was at a maximum when the yeast was taken from the actively fermenting liquid. 5

3 426 gilliland: determination of yeast viability [J. Inst. Brew. TABLE Gelatin Plate Viabilities op Yeast at Various Stages of a Brewing Sample II pitching yeast Fermentation U hr. after pitching Fermentation 48 hr. after pitching.. Yeast ex skimmer Yeast ex storage vessel % Viability by gelatin putte Estimation of viability by slide culture. It would be a great advantage to be able to keep a large number of single cells under observation in order to correlate micro scopical appearance with reproductive ability. The isolation of cells by micromanipulator was too laborious for the numbers required, and a very poor viability was obtained in the cells thus isolated. It was discovered that when a weak wort-gelatin medium was used in haemacytometer slides the cells in the medium retained their position during incubation and reproduced to give discrete colonies. The following "slide culture" technique was therefore devised for estimating viability. A set of clean Thoma ruled haemacytometer slides - mm. deep was sterilized with alcohol. Two drops of a suitable suspension of the yeast in a medium consisting of wort (O.G. 46) containing 6% gelatin were put on each slide which was then covered with a flame-sterilized coverslip. The slides were placed in a 3 C. incubator for 5 min. to allow the yeast cells to fall onto the ruled base of the haema cytometer slide. The edges of the coverslips were then covered with melted vaseline to prevent the medium drying out and the slides were incubated overnight at 8 C. The living cells had by this time produced colonies of about 8-2 cells, whereas the dead cells were single and were generally heavily granulated in contrast to the clear appearance of the living cells. The numbers of microcolonies and of non-repro cells were counted on a series of slides and the per centage viabilities estimated. It was found that this method gave percentage viabilities which were very similar to those obtained by bud-counting. It also enabled a direct comparison to be made with the detection of dead cells by straight microscopical examina tion or by staining. Evaluation of dead-cell staining techniques by slide culture. Numerous stains and tech- niques have been suggested for the differ entiation between living and dead yeast cells.8-8'9'.2.8 Evaluation of these methods has generally depended on correlation with gelatin plate counts which are known to give an incorrect estimation of viability.2 Oppenoorth8 improved on this by first staining the yeast and then isolating single cells into droplets so that the reproductive capacity of the individual cell, either stained or unstained, could be determined. The laborious task of isolating hundreds of single cells, which is entailed in such a study, can be eliminated by the use of slide cultures, as in the following brief examination of the Burns' methylene blue test. A suitable suspension of the yeast in Burns' methylene blue solution was mixed with an equal volume of wort-gelatin and slide cultures were made up. The numbers and positions of stained cells were noted. The slides were incubated overnight and then examined microscopically. Under these con ditions no stained cell produced a colony. Of the unstained cells, 99% produced a colony when normal fresh yeast was under test, but when older yeast was used more of the unstained cells failed to reproduce. The correspondence between percentage viability by staining and by slide culture was good for fresh yeast, but with older yeast the methylene blue method overestimated the percentage of viable cells, as is shown in Table III. Enumeration of chain-formers by slide culture. It was incidentally observed that chain-forming yeasts under the conditions of slide culture gave loose colonies of branched chains, frequently in one plane only, whereas non-chain-forming yeasts gave tight, more or less spherical, colonies. It was thus possible to determine the percentage of chain-forming yeasts in a mixed culture by slide culture. This is much simpler to carry out than the estimation of chain-formers in liquid media as suggested by Curtis.4 A dilution effect. In using the slide culture technique, it was observed that with a brewery yeast the apparent viability de pended on the concentration of yeast in the suspension used to make up the slides. When a series of different concentrations was used, a "dilution effect" could be very strikingly demonstrated, as is shown in Table IV. In the slide containing 9 cells in the counting

4 Vol. 65, 959] gilliland: determination of yeast viability 42 TABLE Comparison of Methylbne Blue Staining with Inability to Reproduce III Cells unstained by metbylene blue Cells stained by mothylcno blue % Viability Yeast Total number Number not repro %not repro Total number Number not repro % not repro By repro duction By staining test A A st test B B B st test 3rd test D C C C st test 2nd tost 3rd test D D D st test 3rd test E (very old). E (very old). st test area all except three cells produced healthy looking colonies, whereas in the slide con taining 9 cells only a single cell produced a colony and 96 cells were granulated and dead-looking, and further incubation did not cause any of these to grow. This effect has been repeatedly demonstrated on yeast from one brewery over a period of several years. Yeasts from some other breweries gave the same result, but yeasts from yet other breweries either did not show this dilution effect or showed it to a lesser extent. TABLE IV Viability of a Brewing Yeast at Different Dilutions as Determined by the Slide Culture Technique Number of cells per ml. 5 x «4- x «3-2 x " 2- x * -6 x " O-9 x * -3 x * Cells per 2 divisions of liacinacytoniotcr Total colonies 53 5 O % Viability by repro duction This observation that many of the yeast cells do not grow when the concentration of cells is below a certain limit may be the reason why poor viability percentages were obtained by plating out this brewery yeast on gelatin plates. The maximum number of cells which could be counted on a plate is of the order of 2, (2 cells per ml.), which is very much below the level at which the viability is impaired by dilution. Investiga tion of this phenomenon was hindered by the fact that once the brewery yeast was grown in the laboratory the "dilution effect" disappeared and the yeast would then give high percentage viabilities at any dilution. Further studies on the lethal effect of dilution on brewery yeast will be reported in another communication. Discussion It is difficult to give a satisfactory definition of yeast viability. It is possible that a cell may be still living, in that it is able to metabolize, and yet it may be unable to reproduce vegetatively. Barton suggested that a yeast cell can never produce a bud on the site at which a previous bud was formed. He observed a yeast cell budding twentythree times. Mortimer & Johnson3 report yeast cells budding up to forty-three times with a median of 24 buds. Bartholomew & Mittwer calculated from the area of a bud scar, and the surface area of a yeast cell, that the theoretical maximum number of buds possible on an average size yeast cell was of the order of. They did not, however, observe more than 2 bud scars on any cell examined, and they concluded that reproduction might cease long before the theoretical number of buds had been formed. They also observed that ascospore formation sometimes took place in old cells which had many bud scars, so that it is probable that in any yeast population there are old cells which are unable to bud, but yet are living as they can metabolize and can perhaps form ascospores. At first sight it might

5 428 gilliland: determination of yeast viability [J. Inst. Brew. appear that there could be a significant percentage of such cells, but it has been calculated that if the maximum number of buds a cell can form is 2, then, in a contin uously growing yeast population, the per centage of cells which have already produced 2 buds is only -%. It is also clear from the very close correspondence between the percentage viability of fresh yeast by staining and by growth in slide culture (Table III) that such cells must be rare in fresh yeast. In old yeasts there are greater percentages of cells which do not stain but which cannot reproduce, and it appears that age alone (without production of buds) can produce cells of this type. Luzzio & Kereiakes found a similar condition in yeast cells subjected to X-radiation; at a certain dosage 9% of the cells failed to reproduce, but there was no corresponding increase of dye uptake at this dosage. It is debatable whether cells which are unable to reproduce are viable or not, but this question is not of great practical importance, as such cells can have little influence on a normal fermentation and are better regarded as dead cells in the estimation of seeding rates. Some brewery yeasts were found to give very low viabilities (of the order of %) when plated on wort-gelatin, owing to a "dilution effect." The selection which must take place when such a yeast is plated might invalidate tests which are designed to deter mine the percentages of different types of yeast present in a mixture, as the ratio of strains surviving may not be the same as the ratio in the original yeast. Unfortunately, it has not been found possible to subculture from the colonies in slide cultures where high viabilities are obtained. Hall6 has already considered this and has made the very good suggestion that samples of brewing yeast for test should be taken from active fermenta tions. Where this is not possible it might be advisable to seed the yeast sample into wort in the laboratory, and take the sample for test after a 24-hr, incubation at 8 C. In this way a high gelatin plate viability should be obtained. It is realized that the proportions of the various types of yeast present in the mixture would be disturbed by culture in wort if the different types had different growth rates, but serious errors seem less likely to be introduced in this way than in the selection which would occur if 9% of the cells failed to grow on the gelatin plate. The gelatin plate method for the determina tion of yeast viability, which has been the standard method for comparison with other more rapid methods, suffers from the following disadvantages. It frequently gives inexplicably low results. It depends on an indirect estimation of total cells and a direct count of colonies, and is therefore subject to large sampling errors. It is subject to errors due to clumping of cells during plating, or to the coalescing of colonies during growth. The slide culture method, on the other hand, depends on the direct observation and counting of repro and nonrepro cells so that sampling errors are reduced and all the other errors are eliminated; it has the further advantage that it can be completed in 8 hr. instead of the four days required for the plate test. The slide culture method is also superior to direct microscopical methods in that it gives a more accurate result, particularly for old yeasts. It has been shown, however, that for normal brewery yeasts direct micro scopical observation either of unstained cells or of cells suspended in buffered methylene blue gives results similar to those obtained by slide culture, and so these microscopical methods will probably remain the methods of choice for routine examination of brewery yeast as they are simple and require only a short time to carry out. The slide culture technique is of the greatest value as a method for checking the accuracy of the microscopical methods, for testing new methods, or for determining the viability of older yeasts. Acknowledgements. Thanks are due to Mr. J. P. Lacey for his technical assistance, to Dr. A. K. Mills for his interest and encouragement, and to the Directors of Arthur Guinness Son & Co. (Dublin) Ltd. for their permission to publish this work. References. Bartholomew, J. \V., & Mittwer, T., /. Bad., 9C3, 65, Barton, A. A., /. gen. Microbiol., 95, 4, Burns, J. A., Brewers' Guardian, Mar., 95,8. 4. Curtis, N. S., this Journal Fink, H., & Kuhles, R., Woch. Brau.. 33, 5, 85.. Hall, J. F., this Journal, 954, 6, 482.

6 Vol. 66,959] gilliland: determination of yeast viability 429. Halvorson, H. O., & Ziegler, N. R.. Quantitative Bacteriology. Minneapolis: Burgess Pub lishing Co., Heucke, R., & Henneberg, W., Stain Tech., 934,, 66. Q. Kcttcrer, A.. Brauwissensckaft.. 956, 4, 5.. Kutscher. V.. Braturei. Wissensch. Beil., 956, 9,2.. Luzzio. A. J., ft Kereiakes, J. G., Exp, Cell Res.. 968, 3, Mills, D. R., Food Res.. 94, Mortimer, M. K.. & Johnston, J. R., U.S. Atomic Energy Commission Document U.C.R.L (969). 4. Oppenoorth, W. F. F., BrauwissenschaJI, 968, 3.