D-Arabitol Production by Endomycopsis chodati

APPLIED MICROBIOLOGY Vol. 12, No. 1, p. 87-92 January, 1964 Copyright (D 1964 by the American Society for Microbiology Printed in U.S.A. D-Arabitol Production by Endomycopsis chodati Forest Products Laboratory,1 Forest Service, G. J. HAJNY U.S. Department of Agriculture, M1adison, Wisconsin Received for publication 21 October 1963 ABSTRACT HAJNY, G. J. (Forest Products Laboratory, Forest Service, U.S. Department of Agriculture, Madison, Wis.). D-Arabitol production by Endomycopsis chodati. Appl. Microbiol. 12:87-92. 1964.-Endomycopsis chodati in an aerated fermentation produced D-arabitol in yields of 35 to 40% of the sugar supplied. Glucose, mannose, and sucrose were suitable substrates. A synthetic medium was developed for the fermentation that showed that nitrogen in the medium must be limiting to obtain high yields of arabitol. Excess phosphate also tended to lower arabitol yields, although the effect was not so great as with nitrogen. Pilot plant-size fermentations were made in which all the nutrients were supplied by blackstrap molasses and urea. Arabitol yields in these fermentations were about 40 0 of the sugar supplied. D-Arabitol was until recently a rather rare chemical that was found in nature in many lichens (Lindberg, MIisiorny, and Wachmeister, 1953) and mushrooms (Frierejacque, 1939). The usual method for the chemical preparation of D-arabitol is the catalytic reduction of D-arabinose or D-lyxose. Binkley and Wolfrom (1950) identified D-arabitol as one of the residual compounds during chromatographic fractionation of cane blackstrap residue that remained after fermentation with baker's yeast. They stated that D-arabitol probably originated in the yeast. Spencer and Sallans (1956) worked with isolates of osmophilic yeasts, and found that many of their cultures, later identified as Saccharomyces rouxii or S. mwllis (Spencer, Roxburgh, and Sallans, 1957), would produce D-arabitol alone or in conjunction with glycerol or erythritol, or both. This was the first report that these polyols could be produced from yeasts in high yields as a normal metabolic product. An interesting feature of this work was the demonstration that the yields of polyhydric alcohols were increased with increasing oxygen tension and decreased with increasing phosphate concentration. Peterson, Hendershot, and Hajny (1958) investigated 11 species of Zygosaccharomyces in respect to their polyhydric alcohol-producing ability. All species produced D-arabitol, and most of them also produced glycerol. In general, changes in fermentation conditions, e.g., aeration, concentration of yeast extract, or inorganic phosphate, had a greater effect upon glycerol than on arabitol yield. 1 Maintained at Madison, Wis., in cooperation with the University of Wisconsin. Onishi (1960) found that osmophilic yeasts isolated from soy mashes or miso-pastes are generally excellent producers of polyhydric alcohols. He surveyed the polyolproducing ability of species of yeasts from the genera Saccharomnyces, Zygosaccharomnyces, Hansenula, Debaryomyces, and Pichia. In the Saccharomyces, glycerol is generally the only polyol produced; in Debaryomyces, D-arabitol is generally the sole polyol. In the other genera, arabitol and glycerol are generally produced together, although a few species also produce erythritol. The ability to produce these polyols is thus much more widespread than was formerly believed. The objectives of the present investigation were to determine the fermentation conditions necessary for an organism to produce high yields of D-arabitol as the sole polyhydric alcohol. MATERIALS AND MIETHODS Cultures. Endomycopsis chodati was previously isolated in a survey of osmophilic yeasts during a search for organisms capable of producing glycerol (Hajny, Hendershot, and Peterson, 1960). The organism used in this study was isolated from a clover blossom growing near a bee colony. Isolation was accomplished by inoculating a 40 % glucose- 1 % yeast extract medium, contained in an Erlenmeyer flask, with the carrier material, and incubating with shaking at 30 C until good growth had taken place. Plates containing 20 % glucose, 1 % yeast extract, and 2 % agar were then streaked with the cultures, and individual colonies were picked and transferred to stock slants. The medium for the stock slants contained 20 % glucose, 1 % yeast extract, 0.1 % urea, and 2 % agar. A transfer of this culture was identified as Endomnycopsis chodati (Nechitch) Wickerham et Burton by Lynferd J. Wickerham, Northern Utilization and Development Division, Agricultural Research Service, U.S. Department of Agriculture, Peoria, Ill. Inoculum for the fermentations was prepared by transferring a loop of the culture from the stock slant to 50 ml of inoculum medium consisting of 20 % glucose, 1 % yeast extract, and 1 % urea contained in a 500-ml Erlenmeyer flask. This was incubated with shaking at 30 C for 48 to 72 hr. An inoculum size of 5 % by volume was used to inoculate the test fermentations. The standard medium consisted of 10% glucose, 0.5 % yeast extract, and 0.1 % urea in tap water. The ph of this 87

88 HAJNY\ APPL. AIICROBIOL. medium was approximately 6.3 before and after sterilization. Fermentationi results obtained by changing fermentation conditions or media were compared with results obtained with the standard medium, to gauge results. Departures from the standard medium are described in the text or tables. Fermentations were routinely made in 500-ml Erlenmeyer flasks, capped with a double layer of milk filter discs, to obtain a more uniform material for gas passage than cotton plugs would supply. 'Media were sterilized by autoclaving at 121 C for 15 min. lf'lasks were shaken on either a reciprocal or rotary shaker, depending on the degree of aeration desired. Oxygen availability was determined by the sulfite oxidation method of Cooper, Fernstrom, and M\Iiller (1944). The use of smooth and indented Erlenmeyer flasks and variation of speed of the rotary shaker allowed a wide range of oxygen availability to be obtained. Evaporation from the shake flasks was variable, depending on aeration rate, fermentation time, or temperature. Before any analytical determinations were made, the fermentation broth was again made up to the original volume. Analytical methods. Samples for analysis weere taken at 24-hr intervals until the fermentation was judged to be completed. Yeast cell volume was determined by centrifuging 10 ml of fermentation broth in 12-ml graduated centrifuge tubes for 10 min. The supernatant liquid from the centrifugation was used for further analyses. Reducing sugars were determined by the copper method of Shaffer and Somogyi (1933). Nitrogen in the media was determined by a micro-kjeldahl method. D-Arabitol was separated from other polyols and glucose by paper chromatography as described by Moore et al. (1960). The developing solvrent used was butanol-pyridinewater (10:3:3, by volume). After elution of the polyol from the paper, quantitative determination was made by the periodic acid oxidation method of Siggia (1959). RESULTS AND DISCUSSION Effect of temperature. The influence of temperature on the arabitol fermentation is shown in Table 1. At 25 C, the fermentation was relatively slow- and the yield of arabitol was very poor. Yeast growth at this temperature was comparable with that at 30 or 35 C. The fermenta- TABLE 1. Influence of temperatur-e on arabitol production Glucose concn Arabitol yield Tem.p Time ph Cell vol on glucose Initial Final consumed C hr % 25 96 4.6 8.0 10.2 0.2 10.5 30 72 4.8 8.0 11.1 0.2 30.3 35 72 5.1 7.0 10.6 0.2 29.3 40 96 5.8 2.0 10.3 7.6 14.3 Medium: 10%C glucose, 0.5%,- yeast extract, 0.1%, urea; aerations at 30 and 35 C were similar to each other in all respects. Arabitol yields were about 30 % for both temperatures. A temperature of 40 C was obviously too high, because yeast growth was very poor, sugar utilization was very slow, and arabitol produced onl sugar used was very low. A temperature of 30 C was selected as the standard for this fermentation. Effect of yeast e.rtract concentration. In the standard medium, yeast extract supplied all the nutrieints neeessary for this organism with the exception of some of the nitrogen, which was supplied by urea. Fermentations were made in which the yeast extract coneentrationi was varied from 0.2a to 2.0%/. The yeast extract eoneentration exerted a tremendous influence on the fermentation (Table 2). At the highest level of yeast extract used, a very fast fermentation occurred, and 10 % glucose was used in 24 hr. The arabitol yield, however, was poor, being only.15.3 %/c based on glucose used. Cell growth was high, since packedcell volume w-as 11 %^. At the lowest level of yeast extract coneentration used, the fermentation was slow, with only 7.5 % glucose being used in 96 hr. The arabitol yield on glucose consumed was 29.6 %l, while the packed-cell volume was 6 %. The standard medium, containing 0.5 % yeast extract, was judged to be optimal since a high yield of arabitol (32.1 (%) was obtained in a reasonable fermeintation time. TABLE 2. Effect of yeast extract concentration on tyield of ar,abitot by Endomn ycopsis chodati Yeast Nitrogen Fermen- C v extract in tation ph Cell vol Glucose concn Arabitol yield concn medium time p Initial Final On total On gluglucose cose used o % r II % %h 0.25 0.066 96 4.4 6.0 10.2 2.7 21.8 29.6 0.5 0.086 72 4.7 8.0 10.1 0.2 31.5 32.1 1.0 0.126 48 5.8 10.0 10.0 0.3 10.5 10.9 2.0 I 0.206 24 6.1 11.0 10.1 0.3 5.1 5.3 Medium, 0.1c urea; temperature, 30 C; aerationi, mimaole of 02 per liter per hr. TABLE 3. Sobs/trates uised by Enldontycopsis chodati Substrate concn Arabitol Substrate Fermen- Cell tation time yield on dry wt initial Final substrate used h.r MOg/mi, (,4 D-XAlose.9.. 96 12.0 9.8 6.7 0.0 il-arabinose 96 12.0 10.4 7.6 0.0 D-Glucose. 96 16.0 10.8 0.2 31.0 D-Mannose.. 96 14.0 8.1 0.3 28.0 i)-g(alactose. 96 14.0 10.4 4.5 0.0 Sucrose... 96 14.0 10.9 1.6 35.1 Maltose.96.96 18.0 6.7 1.4 0.0 Lactose. 96 1.0 6.0 5.9 0.( Glycerol..96 16.0 10.3 4.5 0.0 Medium: substrate pltus 0.5%jf, y,east extract anid 0.1Xj; uirea; tion: mmole of 02 per liter per hr. teniperature, 30 C; aeratioti, niiiiole Of 02per liter pei- hr.

VOL. 12. 1964 D-ARABITOL FROM ENDOM-YCOPSIS CHODATI 89 Substr ates Jermented by Endoinycopsis chodati. Aerobic fermentation of various carbohydrate sources in media containing 0.5 % yeast extract and 0.1 % urea is shown in Table 3. Glucose, mannose, and sucrose were readily fermented, and arabitol was produced in comparable yields from these sugars. Xylose, arabinose, galactose, and maltose were slowly utilized with no production of arabitol. Lactose was the only sugar tried that was not used by the culture. In these fermentations, the quantity of cell material produced from all the sugars used was approximately the same. On the basis of substrate consumed, however, cell yields from glucose, mannose, or sucrose were approximately one-half those obtained from the other sugars. L'ffect of aeration. To obtain satisfactory yields of arabitol, this fermentation must be aerobic. If an anaerobic fermentation was made, the yield of arabitol decreased to the range of 5 to 10 % on the glucose consumed, and an ethyl alcohol yield of 20 to 30 % was obtained. In aerobic fermentations, ethyl alcohol yields of 5 % or less were usually obtained. Ethyl alcohol yields were not usually determined in the aerobic fermentations since, unless special precautions were taken, losses due to evaporationi were considered to be so great as to make the determination meaningless. Fermentations were made at aeration rates of, 200, and 360 mmole of 02 per liter per hr, and at 10, 20, and 30 % glucose concentrations. If the yeast extract concentration was held at 0.5 % in all the media, the fermentation time was extremely long at the higher glucose concentrations (Table 4). To obtain a more reasonable fermentation time, the yeast extract-to-glucose ratio was held constant in all the media (Table 4). For the 10% glucose concentration, aeration rates of and 200 mmole of 02 per liter per hr resulted in yields of arabitol of 32 %, with a sharp drop in yield to 22 %O at an aeration rate of 360 mmole of 02 per liter per hr. Results of fermentations made with 20 and TABLE 4. Effect of aeration, glucose, and yeast extract concentration on yields of arabitol by Endoisycopsis chodati Aeration Yeast ex- Fermenratet tract concn tation time 200 360 200 360 200 360 Cell vol X, AIr X,h 0.5 72 8.0 0.5 96 8.0 0.5 72 8.0 0.5 96 7.0 1.0 72 8.0 1.0 96 10.0 1.0 96 11.0 0.5 96 7.0 1.5 96 8.0 1.5 96 11.0 1.5 96 11.0 Glucose concn Initial /0 10.3 10.3 10.3 19.(6 20.6 20.5 20.5 29.7 31.9 31.9 31.1 Final 0.5 0.2 0.1 6(.5 0.4 0.4 0.3 13.9 9.3 3.5 8.0 Media, 0.1%'/' urea; temperature, 30 C. t Expressed as millimoles of 02 per liter per hour. Arabitol yield on glucose use 32.1 32.2 22.2 36.2 21.8 32.7 23.4 34.8 18.2 33.4 22.6 30 %/ glucose were similar. At the low aeration rate, yields of arabitol were low; at the intermediate rate, yields rose to a peak and then dropped off abruptly at the highest aeration level. An aeration rate of 200 mmole of 02 per liter per hr was satisfactory for all glucose concentrations used, although it was necessary to make adjustments in the yeast extract level to attain reasonably short fermentation times. Yeast extract substitutes. The yeast extract used as a source of nutrients in the fermentation was satisfactory, but it is too expensive to use in larger-sized fermentations. Corn steep liquor and blackstrap molasses were, therefore, used in fermentations as a substitute for yeast extract. Corn steep liquor (50 % solids) was used in concentrations of 0.25, 0.5, and 1.0%o with 0.1% urea and 10% glucose. At the 0.25 % level of corn steep, 10 % glucose was used in 96 hr with an arabitol yield of 27.7 %, which is somewhat lower than the yield from the standard fermentation. When 0.5 and 1.0 % corn steep was used, fermentation time for using 10O glucose was reduced to 72 and 48 hr, but the arabitol yield was reduced drastically to 16.2 and 11.5%, respectively. Blackstrap molasses was used at concentrations of 1, 5, and 10 'X. At the 1 and 5 %/ levels, the fermentations were slow, although arabitol yields on sugar used were good. At the 10 7 level, 13 % sugar was used in 96 hr with an arabitol yield of 32.1 %. Another fermentation was made by diluting blackstrap molasses to 10 % sugar and adding 0.1 '> urea. This resulted in a fast fermentation but produced an arabitol yield of only 17 %. Since 10 % blackstrap appeared to be a satisfactory substitute for yeast extract, a series of fermentations were made in which the nitrogeni level in the medium was varied by the use of various levels of urea while keeping the blackstrap level at 10 %. The results of these fermentations in which the nitrogen level was varied from 0.02 to 0.20/c are shown in Table 5. As the level of nitrogen in the medium increased, the yeast growth increased, leveling off at about 0.07 (Y> total nitrogen in the medium. The rate of sugar utilization increased with increase of nitrogen up to a level of 0.112 %c. Arabitol yields based on sugar consumed decreased with increasing nitrogen content of the medium. A nitrogen TABLE 5. Effect of urea concentration utpon the arabitol ferol.entation Urea Total Fermen- Cell vol Glucose concn yield Arabitol on concn nitrogen tation time Vo yil Initial Final ont glucose used hr C % % lyc 0.0 0.02 96 3.0 14.1 12.4 56.4 0.5 0.043 96 5.0 14.0 7.1 40.5 0.10 0.066 96 7.0 13.1 0.3 35.3 0.20 0.112 72 7.0 13.9 0.7 20.3 0.40 0.204 72 7.0 13.5 0.8 20.4 Conditions were as follows. Media: 10%,, glucose, 10%, blackstrap molasses, and urea as indicated; temperature, 30 C; aeration, mmole of 02 per liter per hr.

90 HAJNY APPL. AIICROBIOL. contenit of about 0.07 'f appeared optimal from the standpoint of length of fermentation time and arabitol yield. Another series of fermentations was run that was similar to the one just described, with the exception that ammonium lactate was the source of added nitrogen rather than urea. The results were similar to those obtained using urea as the nitrogen source. At 0.07'7/ nitrogen in the medium, an arabitol yield of 32.7 7% was obtained in a fermentation time of 96 hr. At higher nitrogen levels, arabitol yields were low; at lower nitrogen levels, the fermentation time was excessive. The standard yeast extract medium has a nitrogen content of 0.086 cc, so that a nitrogen content of the medium of 0.07 to 0.086 ' is satisfactory. Synthetic miiediuml. The results of fermentations containing high levels of yeast extract, corn steep, or blackstrap molasses indicated that these substances contained a nutrient or nutrients which, if present in excessive amounts, led to decreased arabitol yields. The results indicate that the most probable nutrient to cause this is nitrogen, although phosphorus or growth factors might also contribute to the decreased yield. To answer this question unequivocally, a synthetic medium was designed, and fermentations were made at various levels of the different nutrients. The synthetic medium was patterned after that of Hendershot (19.59), which in turn was based on those of Atkin et al. (1944) and Olson and Johnson (1949). The complete synthetic medium is showni in Table 6. In the development of this medium, a number of nitrogen sources were considered. The nitrogen sources were used at such a concentration that a nitrogen level of 0.086/c; (the same as that in the yeast extract standard medium) was obtained in the media. Vitamin-free hydrolyzed casein was an excellent source of nitrogen; the fermentation was comparable with the yeast extract control in all respects. L-Glutamic acid was a poor nitrogen Glucose TABLE 6. Constituent Synthetic mediumt Concn wt per liter 100.00 g Urea 1.0 g Ammonium lactate... 3.0 g Salts KH2PO4... 0.3 g C 0.5 g MgSO4 7H20 0.25 g Trace minerals ZnS04 7H20... 2.0 mg FeSO4(NH4)2SO4 6H20... 1.0 mg CuS04 5H20... 0.1 mg MnSO4 H20... 0.005 mg Vitamins... Inositol... 10.0 mg Biotin... 0.01 mg Calcium pantothenate...... 0.50 mg Nicotinic acid... 2.0 mg Thiamine HCI... 1.0 mg Pyridoxamine HCl...... 1.0 rag source; the fermenitationi resulted in poor yeast growth, slow sugar utilization, and very poor arabitol yields oni sugar consumed. Asparagine was a better source of nitrogen, although arabitol yields were somewhat low aiid sugar utilization was slow compared with the standard fermentation. Ammoniium lactate and urea proved to be the best synthetic sources of nitrogen. In comparison with the yeast extract fermentationi, the only difference noted was the smaller amount of yeast growth. Ammoniium chloride and ammonium sulfate were not suitable as nitrogen sources, since the medium was unbuffered. Consequently, during the fermentation after some of the ammonium nitrogein had been used, the ph dropped so low as to prevent further fermentationi. The influence of the concentration of nitrogeni in the synthetic medium was next studied (Table 7). Nitrogen was supplied entirely by ammonium lactate. The results clearly show that excessive nitrogen in the medium causes a drastic decrease in the yields of arabitol. At the lower nitrogen levels, yeast growth was restricted and sugar utilization was slow. As the nitrogen level in the medium was increased, yeast growth increased, with a leveling off of growth at about 0.12' ( nitrogen in the medium. The rate of sugar utilization paralleled yeast growth, so that fast fermentations were obtained at the high nitrogen levels. Good yeast growth and fast fermentations are not synonymous with good fermentation, since the criterion TABLE 7. Influence of nitrogen concentration u?pon yields of arabitol by Endosnycopsis chodati Nitrogen in Fermentation Glucose concn Arabitol Cell yield medium time vol on glucose Initial Final used hr %7e % %r 0.020 96 4.0 9.4 5.0 32.5 0.040 96 5.5 9.6 1.1 30.1 0.080 72 6.0 10.2 0.2 34.4 0.120 48 7.0 9.0 0.1 13.6 0.160 48 7.0 9.1 0.2 14.2 Synthetic mediurm: nitrogen supplied entirely by ammoniumii lactate; temperatuire 30 C; aeration, mmole of 0(2 per liter per hr. TABLE 8. Phosphate requiremnent of Endonmycopsis chodati for gr7owth and arabitol produ ction Fermentation Glucose concn Arabitol yield KH2PO4 concn Fermentation Cell vol on glucose time ~~~Initial Final used g/liter Itr 7c 0.0 144 3.0 10.0 7.5 24.3 0.15 144 6.0 10.2 1.2 27.8 0.30 72 7.0 9.9 0.5 33.2 0.60 72 8.0 10.1 0.2 27.6 1.20 48 8.0 10.1 0.3 25.9 Complete synthetic medium with phosphate concentration as shown; temperature, 30 C; aeration, mmole of 02 per liter per hr.

VOL. 12, 1964 D-ARABITOL FROM ENDOlI YCOPSIS CHODA TI 91 for a good fermentation in this case is a good arabitol yield. Compromises must be made in choosing fermentation conditionis so that good yields of product may be obtainied in reasonable fermentation times. On this basis, nitrogeni content of the medium should be about 0.08' ;, which will result in a fermentation in which the arabitol yield will be about 30 to 35 T( and the fermentation time will be approximately 72 hr. In Table 8, the results of a study of the phosphorus requirements of E. chodati are shown. The cells used for the inioculum were washed twice in distilled water, so that the only phosphorus in the fermentation was that added, as showin in the table, and that contained in the cell inoculum. When no phosphorus was added to the medium, there was some cell growth and very slow sugar utilization with a somewhat low conversion of sugar to arabitol. As the concenitration of phosphorus was increased, cell volume increased, the rate of sugar utilization increased, and arabitol yield increased to a maximum at 0.3 % KH2PO4 and then decreased slightly. As in the case of the nitrogen concentration, a compromise must be made between speed of fermentation and arabitol yield. Thus, a concentration of moniopotassium phosphate of about 0.3 g per liter is indicated. IFermentations were made to determine whether the coiicentration of potassium or magnesium ions in the medium was critical. When studying the effect of potassium, the synthetic medium was prepared with NH4H2PO4 so that all the potassium was added as KCl. At levels below 0.1 g of KCl per liter, sugar utilization was slight aind arabitol yield on glucose used was low. In the range of KCI concentrations of 0.1 to 1.0 g of KCI per liter, the fermentations were normal with no indication that excess potassium was detrimental to the fermentation. Similar results were obtained in the magnesium study. Below a concentration of M\gSO4 7H20 of 0.05 g per liter, fermentationis were poor. Over the range of 0.05 to 0.5 g of M\IgSO4.7H20 per liter, fermentation results were constant. Both potassium and magnesium are essential to the fer- TABLI.E 9. Vitamin reqniirements of Endomycopsis chodati Vitamin omitted Vitamin Cell Glucose concn Arabitol yield on omittedvol Initial Final glucose used % % % % Biotini... 6.0 9.7 0.6 35.4 Thiamine -HCI... 6.0 10.0 0.2 36.4 Calciuin pantothenate...6.0 10.1 0.1 36.7 Inositol... 6; 0 10.1 0.1 35.1 Nicotinic acid...6.0 10.2 0.1 36.8 Pvridoxamine HCl...6.0 10.0 0.9 34.2 None...6.0 10.2 0.2 35.0 All... 6.0 10.0 0.3 36.7 Yeast extract control medium.. 7.0 13.1 0.3 33.9 Mledium: complete synthetic medium with individual vitamins omitted as indicated; temperature, 30 C; aeration, mmole of 02 per liter per hr. Fermentation time was 96 hr. mentation, but above a minimal concentration neither has any effect upon the fermentation. No attempt was made to purify the ingredients of the synthetic medium so that they would be free of trace minerals. Thus, the trace minerals in the synthetic medium could be omitted or used at double the concentration shown with no effect upon the fermentation. They were used routinely as a safety precaution. Since yeasts as a class of organisms have widely varying requirements for growth factors, the synthetic medium was made up with those growth factors commonly required by a wide variety of yeasts. To determine which of these growth factors were essential to E. chodati, fermentations were made in which the individual growth factors were omitted one by one (Table 9). To avoid carryover of the vitamins from the inoculum medium, the inoculum was centrifuged and the cells were washed twice in sterile distilled water and then made up to volume in distilled water before use as the inoculum. Fermentations were also made in which all the vitamins were omitted and in which none were omitted (Table 9). The results clearly show that none of the growth factors is essential. Since it is possible that the yeast inoculum was able to store enough of a growth factor for one fermentation, serial transfer fermentations were made with the synthetic medium in which all the vitamins were omitted. Parallel serial fermentations were made with the complete synthetic medium for comparison. Seven serial transfers were made, and arabitol yields were determined at each transfer. Yields of arabitol in the synthetic medium both with and without the growth factors ranged from 32 to 37 %. Thus, this culture of E. chodati was shown not to require any growth factors. Pilot-size fermnentations. The fermentations discussed thus far were all conducted on a small scale in 500-ml Erlenmeyer flasks, and aeration was achieved by shaking on either a reciprocal or rotary shaker. It was of interest to determine whether similar results could be obtained in pilot plant-sized equipment. A fermentation was therefore made in a stirred, jacketed, stainless-steel fermentor with a TAB LE 10. Pilot-size (20 gal) stirred aerated fermentation Fermentation ph Cell sol Glucose Arabitol Arabitol yield time in ph Cell vol Glucose concn on glucose used hr %c g/10o 111l g/10o ml nl 0 6.1 0.5 13.5 0.0 0.0 17 6.2 5.0 11.3 1.0 44.0 24 6.4 6.0 10.1 1.6 45.8 48 6.3 6.5 7.6 3.3 56.1 72 6.5 7.0 5.4 4.4 53.8 100 6.7 7.5 3.3 5.2 50.9 122 6.8 7.5 2.0 5.5 47.6 144 6.8 7.5 0.9 5.8 46.2 161 6.9 7.5 0.5 5.7 43.4 Medium: 10% glucose, 10% blackstrap molasses, 0.1% urea; temperature, 30 C; aeration, 46 mmole of (2 per liter per hr.

92 HAJNY APPL. MICROBIOL. 20-gal capacity. The stirrer was a flat-bladed turbine wheel, and aeration was accomplished by introducing air through 38-in. tubing, which contained h-in. perforations directly beneath the turbine wheel. Foaming during the fermentation was controlled by means of lard oil, which was added as needed by means of an automatic foam control unit. The air introduced into the fermentor was measured by means of a rotameter, sterilized by means of a cotton filter, and humidified by passage through a tower containing water. The medium was made up with 10% glucose, 10 %o blackstrap molasses, and 0.1 % urea. It was sterilized in the fermentor by indirect heating by use of a steam jacket. The air filter, piping, and water used for humidification were sterilized by direct steam heating at 15 psi. Since the blackstrap molasses used makes a sizable contribution to the sugar in the medium, all samples used for sugar analysis were first hydrolyzed, and the sugar was reported as glucose. The inoculum was grown up in a medium containing 10% glucose, 0.5% yeast extract, and 0.1 % urea. The amount of inoculum used was 5 % by volume of the medium used in the fermentation. The results of this larger fermentation were comparable with those of the shake flask fermentations, with the exception of time for complete sugar utilization (Table 10). The arabitol concentration in the beer was 5.7 % for a yield on total sugars in the fermentation of 41.9 or 43.4% of the sugar consumed. E. chodati thus was shown to be capable of producing D-arabitol in yields of approximately 35 % of the sugar consumed. 1Miany of the common sugars support growth of the organism, but arabitol is produced only from glucose, mannose, and sucrose. A synthetic medium was developed for the fermentation. With this medium, it was shown that the organism does not require any growth factors and that no organic nitrogen is necessary for growth or fermentation. Nitrogen in the medium must be strictly controlled, since excess nitrogen leads to poor arabitol yields. Phosphate, too, needs to be limited, although the effect on arabitol yields is not so great as that of excess nitrogen. A medium suitable for pilot plant-size fermentations was developed, with only blackstrap molasses and urea as the source of inutrients. With this medium, 20-gal fermentatiolks were made in which arabitol yields of 40 % based on the sugar in the fermentation were obtained. ACKNOWLEDGMENT Appreciation is expressed to Lynferd J. Wickerham for identification of the organism used. LITERATURE CITED ATKIN, L., W. L. WILLIAMS, A. S. SCHULTZ, AND C. N. FREY. 1944. Yeast microbiological methods for determination of vitamnins. Pantothenic acid. Ind. Eng. Chem. Anal. Ed. 16:67-71. IIINKLEY, W. W., AND M. L. WOLFROM. 1950. Chromatographic fractionation of cane blackstrap molasses and its fermentation residue. J. Am. Chem. Soc. 72:4778-4782. COOPER, C. M., C. A. FERNSTROM, AND S. A. MILLER. 1944. Performance of gas-liquid contactors. Ind. Eng. Chem. 36:504-509. FREREJACQUE, M. 1939. Presence Of D-arabitol infistulina hepatica. Compt. Rend. 208:1123-1124. HAJNY, G. J., W. F. HENDERSHOT, AND W. H. PETERSON. 1960. Factors affecting glycerol production by a newly isolated osmophilic yeast. Appl. Microbiol. 8:5-11. HENDERSHIOT, W. F. 1959. Production of polyhydric alcothols from yeasts. Ph.D. Thesis. University of Wisconsin, Madison. LINDBERG, B., A. MISIORNY, AND C. A. WACHMEISTER. 1953. Chemistry of lichens. IV. Investigation of the low molecular weight carbohydrate constituents of different lichens. Acta Chem. Scad. 7:591-595. MOORE, W. E., M. J. EFFLAND, D. B. JOHNSON, M. A. DAUGHERTY, AND E. J. SCHWERDTFEGER. 1960. Chromatographic analyses of sugar alcohols and glycols. Appl. Microbiol. 8:169-173. OLSON, B. H., AND M. J. JOHNSON. 1949. Factors producing high yeast yields in synthetic media. J. Bacteriol. 57:235-246. ONISHI, H. 1960. Studies on osmophilic yeast. VIII. Polyalcohol production by various genera and species of yeast. Bull. Agr. Chem. Soc. Japan 24:131-140. PETERSON, W. H., W. F. HENDERSHOT, AND G. J. HAJNY. 1958. Factors affecting production of glycerol and D-arabitol by representative yeast of the genus Zygosaccharomyces. Appl. Microbiol. 6:349-357. SHAFFER, P. A., AND M. SOMOGYI. 1933. Copper-iodometric reagents for sugar determinations. J. Biol. Chem. 100:695-713. SIGGIA, S. 1959. Quantitative organic analysis via functional groups. John Wiley & Sons, Inc., New York. SPENCER, J. F. T., J. M. ROXBURGH, AND H. R. SALLANS. 1957. Production of glycerol and D-arabitol. U.S. Patent 2,793,981. SPENCER, J. F. T., AND H. R. SALLANS. 1956. Production of polyhydric alcohols by osmophilic yeasts. Can. J. Microbiol. 2: 72-79. ERRATUM Accuracy of a Plant-Infection Technique for Counting-Populations of Rhizobium trifolii J. BROCKWELL Division of Plant Industry, CSIRO, Canberra, Australia Volumn 11, no. 5, page 379, col. 1, Table 1. Concluded, last column: change 10.4-7.0 X 100 to 1.0-7.0 X 100.