APPLIED MICROBIOLOGY, Dec. 1973, p. 982-990Copyright © 1973 American Society for Microbiology

Vol. 26, No. 6Printed in U.S.A.

Isolation and Characterization of a

Thermotolerant Methanol-Utilizing YeastD. W. LEVINE AND C. L. COONEY

Department of Nutrition and Food Science, Massachusetts Institute of Technology, Cambridge, Massachusetts02139

Received for publication 2 August 1973

A yeast capable of growth on methanol as its sole carbon-energy source was

isolated from soil samples and identified as a strain of Hansenula polymorpha. Acontinuous enrichment culture at 37 C with a simple mineral salts medium was

used to select this organism. The isolate, designated DL-1, has a maximalspecific growth rate of 0.22 per h, at pH 4.5 to 5.5 and temperatures of 37 to42 C, in simple mineral salts medium with methanol (0.5%), biotin, and thi-amine. Growth occurred in a chemostat at temperatures up to 50 C, withstrong growth at 45 C. The maximal growth yield of the yeast on methanol was

0.36 g of dry cell weight per g of methanol, and the yield on oxygen was 0.37 g ofdry cell weight per g of 02 Protein content of the isolate is 46%, and total nucleicacid content varies from 5.0 to 7.0% with increasing growth rate from 0.08 to 0.20per h. The amino acid profile of this yeast protein indicates that it could serve

as a good source of food protein. Feeding studies with rats show the yeast tohave no toxic effects.

The ability of microorganisms to utilizemethanol as their sole carbon source is wellestablished (6). Most of the work in this areahas been concerned with bacteria; however, inrecent years, studies by Ogata et al. (11-13),Asthana et al. (3), Oki et al. (14), Sahm andWagner (16), and Hazeu et al. (9) have dealtwith the growth of yeast on methanol. Anexamination of this literature, however, showsthat quantative data relating to methanol utili-zation by yeasts are still quite sparse. Thepurpose of this study was to isolate and examinethe growth properties of a yeast able to usemethanol as its sole carbon-energy source in adefined mineral salts medium at temperaturesgreater than 35 C. This system was chosen forthe following reasons.

(i) Recovery of yeast by either centrifugationor filtration is significantly easier and cheaperthan recovery of bacteria due to the larger celldiameter.

(ii) Yeast, as single-cell protein, is psycholog-ically more palatable for human consumption.In addition, a precedent for yeast as a foodsupplement has already been set.

(iii) This system with high temperature (35to 45 C) and low pH (4.0 to 5.0) has a lowertendency toward contamination.

(iv) Higher growth temperatures would meanless expense for fermentor cooling.

These criteria evolve from the desire to assessmethanol utilization of yeast as it applies tosingle-cell protein production and, more gener-ally, to any methanol-based fermentation.

MATERIALS AND METHODSIsolation of organism. The organism used in this

study was isolated from soil by means of a continuousenrichment technique. Soil samples were collectedand incubated at 37 C as a slurry in a methanol-watersolution. After approximately 1 week, these slurrieswere used as inocula for a nonseptic, continuousculture. The medium was the simple mineral saltsmedium described below, except without vitamins.Methanol concentration in the feed was 10 ml/liter.The temperature was maintained at 37 C, and the pHwas set initially to 4.5 but was allowed to settle withculture growth to 3.5. The pH was set low to selectpreferentially for yeast over bacteria. The dilutionrate of the continuous enrichment was set at 0.07 perh to insure selection of an organism capable of dou-bling its mass in at least 10 h.

After 2 to 3 weeks, this selection technique yieldedan unstable, mixed population growing on methanol.From this mixed culture, a yeast (DL-1) was isolatedby streaking on agar plates. The plates were madewith Difco yeast nitrogen base medium, containing 15ml of methanol/liter. The isolate was maintained onagar slants of either the Difco yeast nitrogen base withmethanol or the mineral salts-methanol medium withbiotin and thiamine.

Media. Two growth media were used during this982

on January 7, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

THERMOTOLERANT METHANOL-UTILIZING YEAST

investigation: a mineral salts medium and the pre-pared Difco yeast nitrogen base. The mineral saltsmedium has the following composition (grams perliter): (NH4)2.SO4, 5.0; CaCl2 2H2O, 0.1; NaCl, 0.1;MgSo4, 0.5; biotin, 2.0 x 10-6; thiamine-hydrochlo-ride, 400 x 10-6; and trace salts solution, 10 ml/literstock solution. Stock trace salts solution used has thefollowing composition (grams per liter): MgSO4, 6.0;CaCl2 2H20, 0.015; FeSO4 7H2O, 0.028;ZnSO4 7H20, 0.140; CuSO4-5H20, 0.025;Na2MoO4-2H20, 0.024; CoCl2*6H20, 0.024; andMnSO4 1 20, 0.084. The phosphate concentrationwas maintained at 0.07 M for all experiments. Forshake-flask experiments, pH was set by varying theratio of KH2PO4 to Na2HPO4. In chemostat experi-ments, 0.07 M KH2PO4 was used to give the broth a

slight buffering capacity to minimize the "overshoot"effect of the pH control system. The concentration ofmethanol in the chemostat media was 5 ml/liter andin most shake flask experiments it was 10 ml/liter.

Cell dry weight analysis. The cell pellet from a

centrifuged volume of fermentation broth was washedtwice and dried in a predried, pretared aluminumweighing dish for 8 h at 110 C. The weight of the driedpellet determined the cell dry weight per liter of broth.

Total nucleic acids analysis. The total nucleicacids content of actively growing cells was estimatedby extracting a cell sample with perchloric acid(PCA). A 2-ml volume of cell suspension (about 1 g ofdry cell weight/liter) was centrifuged, washed oncewith 3 ml of cold, distilled water, and centrifugedagain. The supernatant from the wash was saved, andthe pellet was extracted with 3 ml of 0.5 N PCA at 0 Cfor 30 min. This suspension was centrifuged and thesupernatant was saved. The pellet was then extractedwith 3 ml of 0.5 N PCA at 70 C for 20 min. Again thesuspension was centrifuged and the supernatant was

saved. The 260-nm absorbance of the wash and twoPCS extractions was measured on a Gilford 240spectrophotometer, and total nucleic acids contentwas determined by calculation, assuming a gramextinction coefficient of 32.

Cellular protein analysis. The protein content ofisolate DL-1 was estimated by the biuret method. A1-ml volume of washed cell suspension was contactedwith 4 ml of 1 N NaOH at 100 C for 5 min. Thesuspension was then cooled to room temperature, and0.15 ml of 25% CUSO4 5H2O was added. The precipi-tate was broken up with a Vortex mixer, and thesuspension was allowed to stand at room temperaturefor 30 min. After being centrifuged, the supernatantwas examined for 540-nm absorbance on a Gilfordspectrophotometer. Protein concentration was deter-mined by comparison with a standard curve preparedfrom known concentrations of bovine albumin, frac-tion V (Sigma).Amino acids analysis. The amino acid content of

isolate DL-1 was analyzed for two duplicate samples:samples I and II taken from steady state continuousculture, and III and IV taken from a batch, 14-literfermentation. All samples were acid hydrolyzed priorto analysis on a Beckman model 121 amino acidanalyzer (samples I, II) and a Beckman 120C aminoacid analyzer (samples III, IV). The procedure usedfor acid hydrolysis is the following. A known amount

of washed cells was suspended in cold 6 N HCl. Thissuspension was then sealed under vacuum and incu-bated for 20 h at 110 C. After incubation, the vial wascooled to room temperature and opened, and thesuspension was flash evaporated'at 50 C. The residuewas dissolved in sodium citrate buffer, pH 2.2, andfiltered through a membrane filter (Millipore Corp.)to remove solids. The initial cell concentrations variedwith the individual requirements of each amino acidanalyzer. Model 121 required protein concentrationsof less than 500 mg/ml and model 120C requiredconcentrates less than 150 mg/ml. Dilutions were

made before hydrolysis.Methanol analysis. Methanol concentration was

analyzed in a Varian Aerograph 1200 gas chromato-graph with a stainless-steel column (3 ft by 1'18 inch[about 91.44 by 0.32 cm]) packed with 80- to 100-meshPorapak T. The detector was of the flame ionizationtype. An injector temperature of 215 C, column tem-perature of 110 C, and a carrier gas (N2) flow rate of 25ml/min were the operating conditions. An internalstandard of ethanol was used, and the sample size was1 Aliter. When the methanol concentration in thefermentor broth was to be analyzed, the cells were

removed when chilled broth was passed through a

membrane filter (Millipore Corp.) under positivepressure. All samples were held frozen until analyzedand were kept at 0 C during analysis.

Identification of the isolate. The yeast isolate wasidentified by the typing service of Centraalbureauvoor Schimmelcultures (CBS), Delft, the Nether-lands. Confirming tests were carried out in our

laboratory by following the procedures and identifica-tion keys of Lodder (10).Shake flask experiments. Growth characteristics

of isolate DL-1 were found to be sensitive to cultureconditions prior to inoculation; therefore, a standardprocedure was followed for all shake flask experi-ments. All growth experiments were performed withmineral salts-methanol medium with vitamins, at a

methanol concentration of 10 ml/liter. Methanol andvitamins were added aseptically to autoclaved growthmedium. Baffled, 500-ml side arm flasks containing50 ml of media were inoculated either from a slant,or serially with a 1% log-phase inoculum. All flaskswere brought to growth temperatures before inocula-tion. Reproducibility of data was examined over atleast two serially inoculated flasks at identicalconditions. Growth rate was determined by measuringbroth optical density on a Klett-Summerson colo-rimetor with a red filter over a short interval oflog-phase growth. This was done to insure constantpH of the medium. For pH 4.5, broth pH was found tobe highly unstable with increasing cell growth; conse-

quently, the more sensitive Beckman DU spectropho-tometer was used to follow cell growth over a smallerrange of cell densities, before significant pH changesoccurred.

Chemostat studies. Continuous culture studieswere carried out aseptically in a 1-liter fermentor. Amagnetic stirrer provided agitation. Temperature was

controlled by a water bath, and pH was monitoredand controlled by an Ingold pH probe connected to a

Leeds and Northrup controller. The controlled addi-tion of 1 M KOH maintained a constant pH. Fermen-

VOL. 26, 1973 983

on January 7, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

LEVINE AND COONEY

tor volume was set at 375 ml by means of a levelcontroller, rather than by the conventional overflowtube, since discrepancies were found between exit cellconcentration and reactor volume cell concentrationwith the overflow system. A vibrator pump providedaeration by pumping room air at approximately 1liter/min. All continuous culture experiments werecarried out with the mineral salts-methanol mediumwith vitamins. Feed methanol concentration was setat 5 ml/liter. In all cases, vitamins and methanol wereadded aseptically to the medium reservoir after it wasautoclaved. The cellular yield on methanol (YM.OH,grams of dry cell weight per grams of methanolconsumed) of isolate DL-1 was determined for anumber of steady states at various reactor conditions.Steady states were defined by stability of cell concen-tration over a time period equal to at least fourturnover volumes of the fermentor. YMoH is deter-mined by the following equation: YM.OH = x/SO - S,where x is cell concentration (grams of dry cell weightper liter), SO is methanol concentration in the feed(grams per liter), and S is methanol concentration inthe fermentation broth (grams per liter). Methanol va-por in the exit air stream was monitored by gas chro-matography and found to be negligible.The maximal specific growth rate of isolate DL-1

was determined by the washout technique at temper-atures of 32, 37, 42, 45, and 50 C. In these experi-ments, once the chemostat came to steady state atsome known dilution rate, the dilution rate wasincreased to an arbitrary value greater than themaximal growth rate of the organism. A non-steadystate balance on the cells in the chemostat yields thefollowing equation: Alnx = (A - D)At, where x is cellconcentration (grams of dry cell weight per liter), A isthe specific growth rate of the organism (generationper hour), D is dilution rate (per hour), and t is time(hours). For a culture growing at its maximal specificgrowth rate (Amax), Alnx is a linear function of time,with a slope equal to Ama. - D. Therefore, maximalspecific growth rate can be determined by followingcell concentration as a function of time at a known di-lution rate such that D > Ama.. The slope of the rela-tionship is determined graphically.

Studies with 14-liter fermentor. A number of14-liter, semi-batch fermentations were run for thepurpose of determining (i) the cellular yield on oxygen(Yo2, grams of dry cell weight per gram of 02 con-sumed), and (ii) the highest attainable cell density. A14-liter New Brunswick fermentor was charged with10 liters of mineral salts-methanol medium, at amethanol concentration of 5 ml/liter. The fermenta-tion was run aseptically at 37 C and pH 4.5. Bothtemperature and pH were controlled. During thecourse of the fermentation, additional nutrients wereadded discontinuously. Dissolved oxygen in the brothwas monitored by means of a galvanic probe (4), andaeration was adjusted to maintain >50% saturationdissolved oxygen. To avoid a foam problem, aerationrate was kept low (<5 liters/min) and the inlet airstream was supplemented with pure oxygen as neededto meet the oxygen demand of the culture. Cellulargrowth was followed by measuring optical densitywith a Klett-Summerson calorimeter, and periodicsampling for dry weight analysis.

The oxygen yield (Yo,) was determined by Yo0 =

jux/Na, where A is specific growth rate (per hour); x iscell concentration (grams of dry cell weight per liter),and Na is oxygen uptake rate (grams of 02 per literper hour). The oxygen uptake rates were determinedby making a mass balance for oxygen on the fermen-tor. These balances were performed only during thoseparts of the fermentation when inlet air was notsupplemented by pure oxygen. Oxygen input wastherefore taken to be 20.9% of total air fed. Oxygencontent of the exit gas was measured directly with aLeeds and Northrup magnetic oxygen analyzer. Otherrelevant parameters, such as volumetric gas flow andfermentor volume, were determined by direct mea-surement.

Feeding study. The feeding study comprised twosections, the single-dose acute toxicity test and thesub-acute toxicity test. Both tests were carried outunder the direction of Ronald Shank (Department ofNutrition and Food Science, Massachusetts Instituteof Technology).

For the single-dose acute toxicity test, five SpragueDawley Charles River male rats (average body weight100 g) were given, by gastric intubation, 1 g (wetweight) of methanol yeast per kg of body weight.Matched control animals were given bakers' yeast atthe same dose. Both yeast preparations were sus-pended in distilled water (1:1) to permit intubation.The animals were fed rat chow ad libitum for 12 daysand were then weighed and decapitated; liver and kid-neys were fixed in buffered, neutral 10% Formalin,sectioned, and stained with hematoxylin and eosin.For the sub-acute toxicity test, five Sprague DawleyCharles River male rats (average body weight 100 g)were fed semisynthetic agar gel diets in which 16% ofthe protein was supplied by yeast, that is, the dietcontained 7% yeast on a dry weight basis. Feedingcontinued for a total of 34 days, except for days 26 to29, when these animals were fed rat chow because ofdepletion of the yeast diet. Matched control animalswere fed a similar diet made up with bakers' yeast.Growth was followed by body weight measurements.On day 34, the animals were decapitated and ex-amined histologically as above. In both tests, theyeast pastes were heat treated at 60 C for 1 min tostop metabolic activity before incorporation into thediets.

RESULTSIsolation. A few weeks after initiation of the

continuous enrichment culture, the apparatuscontained a mixed population of microorganisms(bacteria, yeast, mycelial forms, and protozoa).From this mixed culture, a methanol-utilizingyeast was isolated and examined.

Although isolated initially in simple mineralsalts-methanol medium without growth factors,the pure culture of the yeast would not grow inthis medium. An examination of the growthrequirements showed that the isolate had anabsolute requirement for biotin, whereas theaddition of thiamine facilitated growth signifi-cantly. Since a significant increase in growth

984 APPL. MICROBIOL.

on January 7, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

THERMOTOLERANT METHANOL-UTILIZING YEAST

rate was noted in the presence of thiamine, allfurther experiments were carried out in a min-eral salts-methanol medium with biotin andthiamine added.

Identification of isolate. A slant of ourisolate was sent to the CBS typing service foridentification. They identified the culture as astrain of Hansenula polymorpha (Table 1).Identification tests performed in our own labo-ratory agree with these observations except thatwe have not observed sporulation of this yeast.On the basis of our initial observations and

the keying guides provided in Lodder (10), wetentatively classified our isolate as a strain ofCandida silvicola (6), which is the haploid formof H. holstii. However, comparison of our isolateto culture collection strains of H. polymorphaNRRL-Y-1798 and H. holstii NRRL-Y-2154shows it to resemble H. polymorpha in assimila-tive pattern and colony morphology moreclosely than it resembles H. holstii. The isolateDL-1 has been assigned culture collection num-bers NRRL-Y-7560 and ATCC 26012 and isavailable from these agencies.Growth characteristics. At 37 C the isolate

has a pH optimum from 4.5 to 5.5 as determinedin shake flask experiments (Fig. 1). At pH 4.5the isolate showed optimal growth from 37 to42 C as determined by continuous culture wash-out experiments (Fig. 2). The maximal specificgrowth rate of the organism under these condi-tions was 0.22 per h.

At methanol levels less than 0.5% (vol/vol),the isolate follows the Monod growth model.Analysis of chemostat data indicates a K8constant of 120 mg of methanol per liter and amaximal specific growth rate of 0.23 per h.In shake flask experiments, greater than 1%methanol (vol/vol) inhibited growth, and nogrowth was observed with 10% methanol; how-ever, subsequent transfer to medium with lessmethanol showed that the cells were not killedby these concentrations.The growth yield on methanol as a function of

dilution rate is shown in Fig. 3. Maximal yieldof 0.36 g of dry cell weight per g of methanoloccurred at a dilution rate of 0.13 per h. Thegrowth yield on oxygen, as determined in a batchculture using a New Brunswick 14-liter fermen-tor, was 0.37 g of dry cell weight -per g of 02(Table 2).Our isolate has a protein content of 46%,

which is independent of the dilution rate, atotal nucleic acid content varying from 5.0 to7.0%, and a growth rate varying from 0.08 to0.20 per h. Figure 4 summarizes cell compo-sition data for the isolate growing at varyingdilution rates and temperatures. These data

TABLE 1. Strain characteristics of DL-1 asdetermined by CBSa

Determination

FermentationGlucose .........................Galactose ........................Sucrose ..........................Maltose .........................

Lactose ..........................Raffinose ........................

AssimilationGlucose .........................Galactose ........................Sucrose ..........................Maltose .........................Lactose ..........................L-Sorbose ........................Cellobiose .......................Trehalose ........................Melibiose ........................Raffinose ........................Melezitose .......................Inulin ............Soluble starch ..................D-Xylose .........................L-Arabinose ......................D-Arabinose .....................D-Ribose .........................Rhamnose .......................Glycerol .........................Erithritol ........................Adonitol .........................Dulcitol ............

D-Mannitol ......................Sorbitol ........................a-Methyl glucoside ...............Salicin ................Lactic acid ......................Succinic acid ..................Citric acid ......................Inositol.....Arbutin .... .............

KNO3 ...........................

+ weak andlatent

+ latent++

++ ltn

+

+

a On yeast autolysate-peptone-glucose agar, thecells are oval or ellipsoidal, 3 to 4.5 by 2 to 4 pm. Thestreak is smooth, glistening, butyrous, with an entireborder. Neither septate nor pseudo-hyphae areformed in slide cultures. Sporulation is extremelypoor; a few hat-shaped ascospores were seen.

are plotted against a dilution rate normalizedagainst gmax at 37 C to allow easy comparisonamong data taken at different temperatures.The amino acid profiles of two samples of yeastprotein are shown in Table 3.

Finally, Fig. 5 shows cell concentration,methanol concentration in broth, and produc-

985VOL. 26, 1973

on January 7, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

LEVINE AND COONEY

0.20[

tJ 0.16H

m

H 0.1231.0cr0

o 0.08U.

Uwa- 0.04U)

ol3.0 4.0 5.0 6.0

pHFIG. 1. Specific growth rate of isolate DL

function of pH as determined in shake flasks.

0.20I

w

H

41X

H

0r

0wa.U)

0.161.

0.121

0.08

0.04[

a

2

coIJz0U

7.0 °70 z4

I-

,-1 as a w

0

I2!

CD-Jcm0-Jw

30 35 40 45 50TEMPERATURE (°C)

FIG. 2. Specific growth rate of isolate DL-1 as a

function of temperature as determined by chemostatwashout.

tivity as a function of dilution rate. The datashow good internal consistency and expectedbehavior.Feeding study. The isolate was found to have

no toxic effects on any of the experimental ratsused in these studies. In the single-dose acute

toxicity test, there were no significant differ-ences in body weight between the test group andthe control group. Examination of the animals'internal organs showed no evidence of toxicityof the yeast.

During the sub-acute toxicity test, there werealso neither external nor internal indications ofyeast toxicity. After the first 12 days of this test,there were no significant differences in bodyweight between the test animals and the controlanimals. At 34 days, the test group weighed 14%less than the control group; however, the rea-sons for the weight difference are not clearly dueto any inherent property of the methanol-grown

0.5

0.4j

0.3J

0.2L

0.1

0 0.1 0.2DILUTION RATE (hr-)

FIG. 3. Cellular yield of isolate DL-1 on methanolas a function of dilution rate.

TABLE 2. Data for determinations of yoa for isolateDL-1 grown in batch 14-liter fermentations

Respira-Specific Cell Air 02 tion rate Ye2growth

concnflow given (Na: g of (g of

rate (g/liter) (liters/ off 02 per cells/g(per h) g/ie) min) (%) liter of 02

per h)

0.126 1.71 1.15 14.30 0.62 0.360.115 3.69 1.50 12.6 1.15 0.3690.103 2.95 0.91 9.0 0.847 0.3580.145 1.32 1.16 15.45 0.537 0.3560.138 2.05 1.16 14.06 0.715 0.3960.102 4.44 2.13 14.74 1.189 0.382

a Y02 average, 0.37.

TEMP 37°C

I-

Lo

pH 4.5

0 0

T370C

TEMP. 37° C

I I I I I I I I0

986 APPL. MICROBIOL.

\

on January 7, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

THERMOTOLERANT METHANOL-UTILIZING YEAST

looL

80

z

6006crCL

a 'An^40v

20L

o0 0.1 0.2NORMALIZED DILUTION RATE ( hr'1)

Dmx3701FIG. 4. Protein and nucleic acid composition of

isolate DL-1 as a function of normalized dilution rate.

yeast. It is believed that the difference is theresult of an interruption of the diet and a 4-dayfeeding of rat chow.The results of the toxicity -tests as stated by

the experimenter are: "There is no evidence toindicate that the methanol yeast is more toxicthan Baker's yeast, and therefore would gener-ally be regarded as non-toxic."

DISCUSSIONWe used the technique of continuous culture

as a tool both for the isolation of H. polymorphaDL-1 and for studies on the characteristics ofthis yeast. Continuous enrichment culture per-mitted us to impose multiple selective pressureduring the isolation; by its very nature, continu-ous culture selects for the organism or orga-nisms best able to compete for the utilization ofmethanol.Agreement of morphological, fermentative,

and assimilative characteristics of isolate DL-1with the type strain of H. polymorpha (10) isgood. Isolate DL-1 differs from the type strainonly in its latent fermentation of maltose and inits extremely poor sporulation. Lodder describesH. polymorpha as fermentative negative formaltose and as an "abundantly" sporulatingspecies.The cellular yield of H. polymorpha DL-1 on

methanol (Fig. 3) progressed through a definitemaximum equal to 0.36 g of dry cell weight per gof methanol at a dilution rate of 0.13 per h. Therising part of the curve can be partially ex-plained in terms of the decreasing effect of

10 maintenance energy with higher growth rates.The falling part of the yield curve is probablyexplained partially by increased secretion of

8 a metabolites into the medium. An examination(, of the fermentation broth for 260-nm absorb-

ance on a per-gram-of-cell-dry-weight basis6 shows an increasing absorbance with increasingi dilution rate; however, no other data are availa-0 ble to indicate the nature of the product or evenZ if this excretion is able to account for the

4 observed decrease in cellular yield.

,, TABLE 3. Amino acid analyses of isolate DL-I

Amino FAO Amino acid100 g of protein (g)acid refer- -.

ence P I III IV Avg

Lys 5.5 8.10 8.06 8.10 8.15 8.10His 2.33 2.33 2.34 2.49 2.37Arg 5.44 5.44 5.56 5.88 5.58Asp 10.79 10.34 9.94 10.90 10.49Thr 4.0 4.97 4.90 5.14 5.67 5.17Ser 4.71 4.56 5.38 5.68 5.08Glu 13.98 14.34 13.58 14.66 14.14Gly 5.23 5.17 4.98 5.39 5.19Ala 5.79 5.84 6.06 6.54 6.06Cysb 0.68" 0.69b 2.39 0.69Metb 0.75 1.52 1.38" 0.39 1.45Val 5.0 6.43 6.39 5.70 6.33 6.21Ile 4.0 5.33 5.23 4.85 4.98 5.10Leu 7.0 8.27 8.09 8.37 8.64 8.34Tyr' 4.46 4.44 5.32 5.05 4.82Phec 4.73 4.68 4.85 5.58 4.96

a Sample.

bIn calculating these averages we have only usedexperimental values that tend to support each other.The FAO reference (1973) for combined methionineand cystine equals 3.5 g of amino acid per 100 g ofprotein.

"The FAO reference for combined tyrosine andphenylalanine equals 6.0 g of amino acid per 100 g ofprotein.

I-

r p H 45_.0 15a. TEMP 37'C 0 4

00

0-00-0-O-0O n 03

4.0 10. 0cO E

02.02AS>

;>20 0

-',/S_-0 _ e.)

X 0 0 10 0 20

DILUTION RATE (hr-1)

FIG. 5. Steady state behavior of isolate DL-) in amethanol-limited chemostat.

VOL. 26, 1973 987

A PROTEIN OF DL-1, 32A NA OF DL-I, 32-o PROTEIN OF DL-I, 37-* NA OF DL-I,37-* PROTEIN OF DL-I, 42-* NA OF DL-1,42-0 PROTEIN OF DL-I,45-

NA OF DL-I, 45A

0

0 0~~~~

'hi

ia -- L)

on January 7, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

LEVINE AND COONEY

The yield on methanol of 0.36 g of dry cellweight per g of methanol obtained with ourisolate is similar to average yields (0.3 to 0.4)reported for other bacteria and yeast utilizingmethanol. In comparison to other yeast isolates,our isolate does not have as high a yield as thatof Asthana et al. (3) at 45%, while beingsomewhat higher than the 29% obtained bySahm and Wagner (16).Using the methods described by Payne (15),

one may calculate a theoretical yield on metha-nol of 0.6 g of dry cell weight per g of 02 basedon the number of electrons available for transferto oxygen. An explanation of this discrepancybetween the theoretical yield and the actualyield reported for our organism and for otherswill be found probably upon the elucidation ofthe pathways of methanol oxidation and incor-poration. It has been shown that microbialmethanol dehydrogenation is carried out by anicotinamide adenine dinocleotide-independentenzyme (2, 7, 17), and it is quite probable thatdifferent cofactor requirements will greatly af-fect energy yield.A double-reciprocal plot (not shown) of the

specific growth rate versus the chemostat brothmethanol concentration demonstrates a surpris-ingly good adherence of our isolate's growth tothe Monod model y = [Mlma. (S/K8 + S)]. Theisolate's fit to the model is good not only interms of the near linearity of the data but also inthe closeness of the graphically determinedvalue of maximal specific growth rate, 0.23per hour, to the experimentally determinedvalue of 0.22 per h. The adherence is surprisingin that methanol inhibition of growth is de-scribed in the literature for other organismsand has been noted in our own work. Appar-ently no significant inhibition effect occursuntil higher concentrations (>0.5%) of meth-anol are present.A value for the Monod constant K8 equal to

120 mg of methanol per liter was calculatedfrom the data. This value is useful for estimat-ing residual substrate levels in a continuousculture. For continuous single-cell protein pro-duction (6) this value allows one to estimate theeconomic effect of substrate wastage and con-tamination of the recovered product. The valueobtained is somewhat higher than those ob-tained for microbial growth on a wide variety ofcarbon sources, which generally range from 1 to50 mg/liter.The variation of protein and nucleic acid

content with growth rate follows the expectedbehavior. Figure 4 presents data for protein andtotal nucleic acid content of isolate DL-1 as afunction of growth rate at 37 C. Data obtained

from single growth rates at 32, 42, and 45 C arealso shown. The plot is constructed using theformat of Alroy (1), in that the data are plottedagainst a normalized dilution rate to allow forcross-comparison among data taken at differenttemperatures. Protein content shows little vari-ation, being largely constant at 46%, whereasnucleic acid content shows a marked increasewith increasing growth rate. Total nucleic acidsvaried from 5.0 to 7.0%. Data taken at 32 and42 C show negligible variation from that takenat 37 C; at 45 C, however, both protein andnucleic acid contents of isolate DL-1 are signifi-cantly lower than the corresponding data at37 C. Alroy presents a correlation of nucleicacid/protein ratio on the basis of dilution ratesnormalized against Dmax at 30 C for manydifferent organisms grown under different con-ditions. Although Dmax at 30 C is not availablefor isolate DL-1 and so data presented herecannot be compared to Alroy's correlation, it ispossible to compare composition data for isolateDL-1 at 32, 42, and 45 C to corresponding dataat 37 C, on the basis of a dilution rate normal-ized against Dma. at 37 C. At 32 C the nucleicacid/protein ratio equals 0.13, and the corre-sponding value at 37 C equals 0.13. At 42 C theratio equals 0.11, while the value at 37 C is 0.12.At 45 C the ratio equals 0.12, and at 37 C the ra-tio is 0.13. All these cases show good agreementand demonstrate an internal consistency amongthe data not readily apparent from Fig. 4.

Actual protein contents of methanol-growncells have been reported from 35% for yeasts(16) to 71% in bacteria (8). Working with yeast,Ogata et al. (12) found their Kloeckera sp. no.2201 to have a protein content of 45.3%, andAsthana (Ph. D. thesis, University of Pennsyl-vania, Philadelphia, 1972) reports a similarfigure of 50%, whereas Sahm and Wagner (16)report a lower protein content of 35.4%. Verylittle nucleic acid data are available; however,Asthana reports a ribonucleic acid content of2.5%, and Ogata reports total nucleic acidcontent 5.4%. Our isolate's protein content of46% and nucleic acid content of 5 to 7% wereexpected values for yeast and are in line withthe data of other isolates.

Figure 5 shows typical variation of cell den-sity, substrate, and productivity with varyingdilution rate. The drop-off of cell density athigher dilution rates is not a matter of oxygenlimitation but of adherence to the Monodgrowth model. Although dissolved oxygen (DO)was not monitored during the fermentationsfrom which these data were obtained, previouschemostats were set up with a galvanic 02 probe(4), and DO was found to be greater than 50%

988 APPL. MICROBIOL.

on January 7, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

THERMOTOLERANT METHANOL-UTILIZING YEAST

saturation over a wide range of conditions.Furthermore, if one calculates a theoreticaloxygen demand for isolate DL-1 growing atAmax, at a cell concentration of 1 g/liter, it willequal 20 mM 02 per liter per h. This demand iswell within the mass transfer capabilities of thefermentor.Amino acid analyses (Table 3) were per-

formed on two pairs of duplicate samples ofisolate DL-1. Agreement is quite good amongthe four samples except in the cases of methio-nine and cystine. For all the essential aminoacids except methionine and cystine, the datacompare well with the Food and AgriculturalOrganization of the United Nations (FAO)reference levels. Isolate DL-1 is particularly richin lysine. Although methionine and cystine arebelow the FAO reference level, they are not solow as to make isolate DL-1 an unattractiveprotein source. Furthermore, values reportedhere should be considered as only minimalvalues since these amino acids are unstableunder acid hydrolysis. In general, isolate DL-1compares favorably to other methanol-grownprotein sources (5, 8, 12, 16; U.S. patent3,546,071, 1970) in lysine, methionine, and cys-tine content. No data were obtained for trypto-phane content. Calculation of cellular proteincontent, on the basis of total protein measuredas individual amino acids, gives a value of 42%for all samples. This value compares quite wellwith the direct measurement of protein contentequal to 46%. Thus, amino acid recovery is quitehigh.To determine growth yield on oxygen and the

highest possible cell density obtainable, a num-ber of fermentations were carried out in a14-liter New Brunswick fermentor. The highestcell density obtained was 24.4 g of dry cellweight per liter. At this point in the fermenta-tion cell growth stopped. The tapering off ofgrowth could be explained by the production ofan auto-inhibitor or by the build up of toxicmetabolites. However, these explanations donot appear to be adequate, since broth collectedat the end of the fermentation, filter sterilized,and inoculated with isolate DL-1 showed ex-ceptionally strong growth. Additional experi-ments are required to determine the reason forthe observed stoppage of growth at 24.4 g/liter.A series of oxygen balances was made during

the fermentation to be used in determining thecellular yield of isolate DL-1 on oxygen. All ofthese balances were performed during the earlyparts of the fermentations before the first addi-tion of methanol. The oxygen requirement ofthe culture was low enough at this stage of thefermentation to allow low aeration with room

air to meet the demand. Therefore, the oxygencontent of the inlet air was accurately known.Also, during this phase of the fermentation, cellgrowth was consistent and strong, and so majorchanges in growth rate during the course ofmeasurement were not a problem.Data used in determining the oxygen yield

(Y(2) are presented in Table 2. As can be seen,agreement among the various data points isquite good. No consideration was given to theeffect on Yo2 that the maintenance coefficient ofoxygen would have. At the time of the experi-ments, the effect of maintenance was assumedto be small, but no data are available to allowthis maintenance coefficient to be estimated.A theoretical yield on oxygen can be calcu-

lated by assuming complete reaction of metha-nol according to the equation: CH3OH + NH3 +02 , cells + CO2 + H20. When YN1,0j, equals0.35 g cells per g of methanol and a cellularcomposition of 50% carbon, 8% nitrogen, 7%hydrogen, and 20% oxygen, a material balanceon the above equation results in a theoreticalYo2 of 0.37. This value is gratifyingly close to themeasured Yo2.

ACKNOWLEDGMENTS

We acknowledge the help of V. Young and R. Lees of theDepartment of Nutrition and Food Science, MassachusettsInstitute of Technology, for their aid in the amino acidanalyses of the methanol-grown protein; and R. Shank for hispart in evaluating yeast toxicity in an animal feeding study.We further acknowledge the financial support of the Lewis

and Rosa Strauss Foundation, and Public Health Servicegrant no. 2TOlES00063-ESTC from the Division of Envi-ronmental Health Services.

LITERATURE CITED

1. Alroy, Y., and S. R. Tannenbaum. 1973. The influence ofenvironmental conditions on the macromolecular com-position of Candida utilis. Biotechnol. Bioeng.15:239-256.

2. Anthony, C., and L. J. Zatman. 1967. Purification andproperties of the alcohol dehydrogenase of Pseudo-monas sp. M27. Biochem. J. 104:953-959.

3. Asthana, H., A. E. Humphrey, and V. Moritz. 1971.Growth of yeast on methanol as the sole carbonsubstrate. Biotechnol. Bioeng. 13:923.

4. Borkowski, J. D., and M. J. Johnson. 1967. Long-livedsteam-sterilizable membrane probes for dissolved oxy-gen measurement. Biotechnol. Bioeng. 9:635-639.

5. Chalfan, Y., and R. I. Mateles. 1972. New pseudomonadutilizing methanol for growth. Appl. Microbiol. 23:135.

6. Cooney, C. L., and D. W. Levine. 1972. Microbialutilization of methanol. Advan. Appl. Microbiol.15:337-365.

7. Fugii, T., and K. Tonomura. 1972. Oxidation of metha-nol, formaldehyde and formate by Candida sp. Agr.Biol. Chem. 36:2297-2306.

8. Hhggstrdm, L. 1969. Studies on methanol oxidizingbacteria. Biotechnol. Bioeng. 11:1043.

9. Hazeu, W., J. C. de Bruyn, and P. Bos. 1972. Methanolassimilation by yeast. Arch. Mikrobiol. 87:185-188.

10. Lodder, J., ed. 1970. The yeasts. North-Holland Pub-lishing Co., Amsterdam.

989VOL. 26, 1973

on January 7, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

990 LEVINE AND COONEY APPL. MICROBIOL.

11. Ogata, K., H. Nishikawa, and M. Ohsugi. 1969. A yeast J. Ferment. Technol. 48:470.capable of utilizing methanol. Agr. Biol. Chem. 14. Oki, T., K. Kouno, A. Kitai, and A. Ozaki. 1972. New33:1519. yeasts capable of assimilating methanol. J. Gen. Appl.

12. Ogata, K., H. Nishikawa, and M. Ohsugi. 1970. Studies Microbiol. 18:295-305.on the production of yeast. I. A yeast utilizing metha- 15. Payne, W. J. 1970. Energy yields and growth of hetero-nol as a sole carbon source. J. Ferment. Technol. trophs. Annu. Rev. Microbiol. 29:17.48:389. 16. Sahm, H., and F. Wagner. 1972. Microbial production

13. Ogata, K., H. Nishikawa, and M. Ohsugi. 1970. Studies from methanol. Arch. Mikrobiol. 84:29-42.on the production of yeast. II. The cultural conditions 17. Sahm, H., and F. Wagner. 1973. Microbial productionof methanol assimilating yeast, Kloekera sp. no. 2201. from methanol. Arch. Mikrobiol. 90:263-268.

on January 7, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from