Fed-batch and continuous fermentation ofSelenomonas ruminantium for natural propionic,acetic and succinic acids

An anaerobic fermentation process was developed for production of natural propionic, acetic and succinic acids froml-lactic acid usingSelenomonas ruminantium. Thel-lactic acid was quickly converted to a racemic mixture and there was no enantiomeric preference for further metabolism. The lactic acid was metabolized to propionic, acetic and succinic acids typically in a molar ratio of about 531. However, the ratio of propionate: succinate started high (as much as 221), before declining to as low as 51 after the first 48 h. Nutrients in corn steep liquor and yeast extract were necessary for optimal production of propionic acid. The corn steep liquor and yeast extract were heat stable at neutral pH, but some nutritional qualities were lost when heated at pH 2.4. In fed-batch fermentation on lactic acid 2.0% propionic acid was produced in 48 h and 2.3% in 68 h. A continuous culture operated at a dilution rate of 0.055 h^–1 and a lactic acid feed concentration of 30 gL^–1 had a propionic acid productivity of 0.59 gL^–1h^–1. The steady state results were: lactic acid 0.6%, propionic acid 1.1%, acetic acid 0.50%, and succinic acid 0.33%.     

Growth of Thermus aquaticus and its TaqI endonuclease production

The culture behaviour of Thermus aquaticus was characterized. The response of the bacterium to various carbon (tryptone, glucose, glycerol) and nitrogen sources (yeast extract, NaNO₃, (NH₄)₂SO₄, leucine, thymine, thiamine, glutamic acid) was studied. Amino acids did not support growth, but CASTENHOLZ salt medium supplemented with yeast extract and glucose or tryptone resulted in good growth and production. A suitable medium composition giving the highest biomass concentration and enzyme yield was developed. The simple medium containing TYE-NaCl resulted in the highest biomass concentration, whereas CASTENHOLZ mineral medium supplemented with tryptone and yeast extract gave the highest specific activity and enzyme yield. The effect of inoculum age and size on growth was also investigated in order to improve the yield and process consistency. The use of shake flasks inoculated with precultures at their early or late stationary phase resulted in the same biomass concentration (0.56 ± 0.015 g/l) and similar maximum specific growth rates (0.258 ± 0.003 h^?1). Inoculum sizes between 1 and 2.5 per cent were optimal for cell growth. As the other papers on thermophilic microorganisms, including the T. aquaticus YT-1 strain, gave qualitative information on growth, the results presented here cannot be compared with others on a quantitative basis. TaqI endonuclease was purified using a 5 step protocol including cell disruption, adsorption, precipitation, column chromatography and final dialysis. The enriched fraction had a specific activity of 33,600 U TaqI endonuclease per mg protein.     

The growth behaviour of Thermus aquaticus in continuous cultivation

Summary The growth of the yellow pigmented non-sporulating caldoactive bacterium Thermus aquaticus was investigated in different culture vessels and using differnt culture techniques. Each combination of these two variables led to very specific characteristic behaviour of the culture. A synthetic medium for a white cell type of T. aquaticus was optimized by means of pulse and medium-shift techniques. The main kinetic parameters of the organism have been determined to be =1.62h^–1 and Y (glucose)=0.4 g g^–1 at T=68 °C and pH=7.3. In complex medium only a mixed population of white and yellow cells could be cultivated. The cell yield was shown to be very low (Y=0.02 g g^–1) due to incomplete substrate utilisation, but a very high maximal specific growth rate was determined ( _max=3.5h^–1) at 75 °C and pH=7.3. The maintenance coefficient for oxygen uptake was approximately Mo=16 mMol g^–1 h^–1. A discussion of the problems arising in the cultivation of thermophilic microorganisms with respect to their physiology and stability is given.     

Fed-batch production of baker's yeast using millet (Pennisetum typhoides) flour hydrolysate as the carbon source

A fermentation medium based on millet (Pennisetum typhoides) flour hydrolysate and a four-phase feeding strategy for fed-batch production of baker's yeast,Saccharomyces cerevisiae, are presented. Millet flour was prepared by dry-milling and sieving of whole grain. A 25% (w/v) flour mash was liquefied with a thermostable 1,4--d-glucanohydrolase (EC 3.2.1.1) in the presence of 100 ppm Ca²⁺, at 80°C, pH 6.1–6.3, for 1 h. The liquefied mash was saccharified with 1,4--d-glucan glucohydrolase (EC 3.2.1.3) at 55°C, pH 5.5, for 2 h. An average of 75% of the flour was hydrolysed and about 82% of the hydrolysate was glucose. The feeding profile, which was based on a model with desired specific growth rate range of 0.18–0.23 h^–1, biomass yield coefficient of 0.5 g g^–1 and feed substrate concentration of 200 g L^–1, was implemented manually using the millet flour hydrolysate in test experiments and glucose feed in control experiments. The fermentation off-gas was analyzed on-line by mass spectrometry for the calculation of carbon dioxide production rate, oxygen up-take rate and the respiratory quotient. Off-line determination of biomass, ethanol and glucose were done, respectively, by dry weight, gas chromatography and spectrophotometry. Cell mass concentrations of 49.9–51.9 g L^–1 were achieved in all experiments within 27 h of which the last 15 h were in the fedbatch mode. The average biomass yields for the millet flour and glucose media were 0.48 and 0.49 g g^–1, respectively. No significant differences were observed between the dough-leavening activities of the products of the test and the control media and a commercial preparation of instant active dry yeast. Millet flour hydrolysate was established to be a satisfactory low cost replacement for glucose in the production of baking quality yeast.Nomenclature C _ox Dissolved oxygen concentration (mg L^–1) - CPR Carbon dioxide production rate (mmol h^–1) - C _s0 Glucose concentration in the feed (g L^–1) - C _s Substrate concentration in the fermenter (g L^–1) - C _s.crit Critical substrate concentration (g L^–1) - E Ethanol concentration (g L^–1) - F _s Substrate flow rate (g h^–1) - i Sample number (–) - K _e Constant in Equation 6 (g L^–1) - K _o Constant in Equation 7 (mg L^–1) - K _s Constant in Equation 5 (g L^–1) - m Specific maintenance term (h^–1) - OUR Oxygen up-take rate (mmol h^–1) - q _ox Specific oxygen up-take rate (h^–1) - q _ox.max Maximum specific oxygen up-take rate (h^–1) - q _p Specific product formation rate (h^–1) - q _s Specific substrate up-take rate (g g^–1 h^–1) - q _s.max Maximum specific substrate up-take rate (g g^–1 h^–1) - RQ Respiratory quotient (–) - S Total substrate in the fermenter at timet (g) - S ₀ Substrate mass fraction in the feed (g g^–1) - t Fermentation time (h) - V Instantaneous volume of the broth in the fermenter (L) - V ₀ Starting volume in the fermenter (L) - V _si Volume of samplei (L) - x Biomass concentration in the fermenter (g L^–1) - X ₀ Total amount of initial biomass (g) - X _t Total amount of biomass at timet (g) - Y _p/s Product yield coefficient on substrate (–) - Y _x/e Biomass yield coefficient on ethanol (–) - Y _x/s Biomass yield coefficient on substrate (–) Greek letters Moles of carbon per mole of yeast (–) - Moles of hydrogen atom per mole of yeast (–) - Moles of oxygen atom per mole of yeast (–) - Moles of nitrogen atom per mole of yeast (–) - Specific growth rate (h^–1) - _crit Critical specific growth rate (h^–1) - _E Specific ethanol up-take rate (h^–1) - _max.E Maximum specific ethanol up-take rate (h^–1)     

Development of medium components for the production of glucosyl-transferring enzyme by Aureobasidium

Aureobasidium sp. ATCC 20524 produced a glucosyl-transferring enzyme which produced panose (O--D-glucopyranosyl-(1»6)-O--D-glucopyranosyl-(1»4)-d-glucose) from maltose. Optimum production for the enzyme was with maltose at 2% (w/v) and yeast extract at 1.5% (w/v). Enzymatic activity reached 0.7×10³ U/g dry cells after 48 h.     

Growth of Thermoanaerobium brockii in batch and continuous culture at supraoptimal temperatures

Summary Thermoanaerobium brockii was grown in batch and continuous culture at supraoptimal temperatures (>65° C). Specific growth rates were lower in batch (_max>1.0 h^-1) than in continuous cultures (_max1.2–1.4 h^-1). Acetone addition to the medium did not increase critical dilution rate significantly. The media used contained significantly less organic material and sulfide than previously reported media; however, yeast extract requirements were shown to be exceptionally high (60% of the glucose concentration used). Organic substrates inhibited growth and product formation in chemostat cultures whereas the slow formation of acetic acid was observed in batch cultures, but also with virtually no growth. The inhibiting concentration was found to be approximately 15 g organic carbon·l^-1. The maintenance requirements of T. brockii were in the same range as expected of aerobic extreme thermophiles (m_s0.5 g·g^-1·h^-1) and could be met only by glucose and not by yeast extract. Maintenance was obviously not independent of specific growth rate. Production of the stereospecific alcohol-aldehyde/ketone oxidore-ductase was strictly growth associated and its formation was not affected by acetone added to medium.     

The effect of carbon dioxide on the growth kinetics of fructose-limited chemostat cultures of Acetobacterium woodii DSM 1030

The growth of the anaerobic acetogenic bacterium Acetobacterium woodii DSM 1030 was investigated in fructose-limited chemostat cultures. A defined medium was developed which contained fructose, mineral salts, cysteine · HCl and Ca pantothenate (1 mg · 1^–1) supplied in a vitamin supplement. Growth at high dilution rates was dependent on the presence of CO₂ in the gas phase. The ^max was found to be 0.16 h^–1 and the fructose maintenance requirement was 0.1 to 0.13 mmol fructose · (g dry wt)^–1 · h^–1. A growth yield of 61 g dry wt · (mol fructose)^–1, corrected for the cell maintenance requirement and for incorporation of fructose carbon into cell biomass, was determined from the fructose consumption. A corresponding growth yield of 69 g dry wt · (mol fructose)^–1 was calculated from the acetate production assuming that fructose fermentation was homoacetogenic. A Y_ATP of 12.2 to 13.8 g dry wt · (mol ATP)^–1 was calculated from these growth yields using a value of 5 mol ATP · (mol fructose)^–1 as an estimate of the amount of ATP synthesised from fructose fermentation. The addition of yeast extract (0.5 g · 1^–1) to the medium did not influence the ^max or cell yield. After prolonged growth under fructose-limited conditions the requirement of the culture for CO₂ in the gas phase was reduced.Abbreviations YE yeast extract - IC inorganic carbon - D fermenter dilution rate : h^–1 - M_X maintenance requirement for X: mmol X · (g dry wt)^–1 · h^–1 - X may be fructose (Fruct), fructose consumed in energy metabolism (Fruct [E]), acetate (Ac) - ATP CO₂, NH _inf4 ^sup+ or Pi - qX specific rate of utilisation or consumption of X: mmol X · (g dry wt)^–1 · h^–1 - V fermenter volume: litre - rC · Cell, fermenter cell carbon production: mmol C · h^–1 - Y_X yield of cells on X: g dry wt · (mol X)^–1 - Y _infx ^supmax the yield corrected for cell maintenance: g dry wt · (mol X)^–1 - S_ATP stoichiometry of ATP synthesis from fructose: mol ATP · (mol frucose)^–1 - x cell concentration: g dry wt · 1^–1 - specific growth rate : h^–1 - ^max maximum specific growth rate: h^–1     

The fermentation of hexose and pentose sugars by Candida shehatae and Pichia stipitis

Summary The fermentation by Candida shehatae and Pichia stipitis of xylitol and the various sugars which are liberated upon hydrolysis of lignocellulosic biomass was investigated. Both yeasts produced ethanol from d-glucose, d-mannose, d-galactose and d-xylose. Only P. stipitis fermented d-cellobiose, producing 6.5 g·l^-1 ethanol from 20 g·l^-1 cellobiose within 48 h. No ethanol was produced from l-arabinose, l-rhamnose or xylitol. Diauxie was evident during the fermentation of a sugar mixture. Following the depletion of glucose, P. stipitis fermented galactose, mannose, xylose and cellobiose simultaneously with no noticeable preceding lag period. A similar fermentation pattern was observed with C. shehatae, except that it failed to utilize cellobiose even though it grew on cellobiose when supplied as the sole sugar. P. stipitis produced considerably more ethanol from the sugar mixture than C. shehatae, primarily due to its ability to ferment cellobiose. In general P. stipitis exhibited a higher volumetric rate and yield of ethanol production. This yeast fermented glucose 30–50% more rapidly than xylose, whereas the rates of ethanol production from these two sugars by C. shehatae were similar. P. stipitis had no absolute vitamin requirement for xylose fermentation, but biotin and thiamine enhanced the rate and yield of ethanol production significantly.Nomenclature _max Maximum specific growth rate, h^-1 - Q _p Maximum volumetric rate of ethanol production, calculated from the slope of the ethanol vs. time curve, g·(l·h)^-1 - q _p Maximum specific rate of ethanol production, g·(g cells·h) - Y _p/s Ethanol yield coefficient, g ethanol·(g substrate utilized)^-1 - Y _x/s Cell yield coefficient, g biomass·(g substrate utilized)^-1 - E Efficiency of substrate utilization, g substrate consumed·(g initial substrate)^-1·100     

Production of ψ-linolenic acid by the fungus Mucor rouxii on cheap nitrogen and carbon sources

Summary The production of -linolenic acid (GLA) by the fungus Mucor rouxii CBS 416.77 was studied on low budget nitrogen and carbon sources, i.e. rape meal, cocos expeller and two types of yeast extract (nitrogen sources), and starch, starch hydrolysate, beet molasses and cocos expeller (carbon sources). As references, Difco yeast extract and glucose were used. In flask cultivations the three yeast extracts were fully interchangeable, while the Difco yeast extract (the most expensive of those tested) gave a higher productivity of GLA in fermentor cultures (14 mg·l^–1·h^–1). The yield of lipids and GLA were increased in the order yeast extract < rape meal < cocos expeller. Thus the amount of lipid increased from 0.56 to 2.8 g·l^–1, and that of GLA from 0.15 to 0.33 g·l^–1. Use of beet molasses or cocos expeller as carbon sources gave poor growth. Starch and starch hydrolysate resulted in better productivity of GLA than glucose (4.7 and 4.9 compared to 3.4 mg·l^–1·h^–1). Offsprint requests to: A.-M. Lindberg     

Development of a continuous system for the degradation of a cyanuric acid by adsorbed Pseudomonas sp. NRRL B-12228

Cyanuric acid in high concentrations (15.5 mm) was degraded completely by Pseudomonas sp. NRRL B-12228 independently of glucose concentration. In the batch fermentations there was a relation between the glucose concentration, on the one hand, and the liberation of ammonia or production of protein, on the other. The greater the supply of carbon, the more biomass was produced, and fewer NH _inf4 ^sup+ ions were released. Continuous fermentations using adsorbed cells could be performed to degrade cyanuric acid. In spite of different glucose feeding there was only a negligible difference in residues of s-triazine. In a one-step continuous system with dilution rates between 0.021 h^–1 and 0.035 h^–1, even a ratio of 0.65 between glucose and cyanuric acid was not sufficient to degrade the cyanuric acid supplied (320–540 mol l^–1 h^–1) completely. When a continuous two-step system was applied with dilution rates between 0.035 h^–1 and 0.056 h^–1, the consumption of carbon source could be minimized while s-triazine degradation up to 860 mol l^–1 h^–1 was complete. In this way the ratio between glucose and cyanuric acid could be increased to 0.25 (molar C:N ratio = 0.33:1). Thereby the process was made considerably more economic.     

Comparative study of the enzymatic synthesis of l-valine by purified enzymes,crude extract,intact or permeabilized cells from Bacillus megaterium

Summary A detailed study on the reductive amination of -ketoisovalerate to l-valine by l-valine dehydrogenase using glucose dehydrogenase as an NADH regeneration enzyme was performed. The presence of both enzyme activities in Bacillus megaterium ATCC 39 118 permitted a direct and systematic comparison of the performances (initial l-valine production rate, productivity, molar conversion yield) of different types of conversion systems: purified enzymes or crude extract and whole cells, intact or permeabilized. A maximal l-valine productivity of 8 mmol·l^–1 · h^–1 was obtained using purified enzymes which constituted the most efficient system with a maximal rate of 0.87 mol · ml^–1 · min^–1 and a molar conversion yield of 0.91. Permeabilized cells were also an attractive system because of their easy preparation and of the good performances attained.Offprint requests to: F. Monot     

Immobilization of β-galactosidases from Thermus aquaticus YT-1 for oligosaccharides synthesis

Summary -galactosidases of Thermus aquaticus YT-1, exhibiting a galactosyl transferase activity, were immobilized using different techniques. Entrapment in agarose or gellan gum beads was unsuitable for enzyme immobilization due to enzyme leakage. A technique that efficiently immobilized the enzymes was developed using glutaraldehyde co-crosslinking of -galactosidases with bovine serum albumin, followed by entrapment in agarose beads.     

Heterologous expression and secretion of sweet potato peroxidase isoenzyme A1 in recombinant Saccharomyces cerevisiae

A vector system has been developed to express isoenzyme A1 of sweet potato peroxidase (POD) and was introduced into Saccharomyces cerevisiae. The system contains the signal sequence of Aspergillus oryzae -amylase to facilitate the extracellular secretion of peroxidase under the control of constitutive glyceraldehyde-3-phosphate dehydrogenase (GPD) promoter. In a batch culture using YNBDCA medium (yeast nitrogen base without amino acids 6.7 g l^–1, Casamino acids 5 g l^–1 and glucose 20 g l^–1), the recombinant strain expressed the swpa1 gene giving a secretion yield of POD activity of ca. 90% of total expressed peroxidase. Supplementation with PMSF (0.05 mM) and Casamino acids (5 g/50 ml) increased extracellular POD activity to nearly 10 kU ml^–1, equivalent to 1.5 kU g^–1 cell dry wt. This is 9 fold higher than that obtained in medium without PMSF. From SDS-PAGE and native-PAGE analyses POD has an M _r of 53 kDa.     

Growth kinetics of Thermus thermophilus

Summary The nonsporulating extreme thermophile Thermus thermophilus was grown in continuous culture at dilution rates up to 2.65 h^–1 at 75°C and pH 6.9 on complex medium. Concomitantly very low yield (Y=0.12 g cell dry weight g^–1 utilized organic carbon) and incomplete substrate utilization (always less than 45%) were found. In batch cultures T. thermophilus could be grown with _max =h^–1, in shake flasks only with _max =h^–1 with the same low yield and incomplete substrate utilization. Stable steady states at 84C and 45°C were realized at a dilution rate of 0.3 h^–1 whereas at 86°C and 40°C no growth could be detected. Artefacts arising from wall growth (in bioreactors) or improper materials must be ruled out. Inhibition of growth by organic substrates was demonstrated at low concentrations: a decrease in the yield obtained was found when more than 0.7 gl^–1 of meat extract were supplied in the medium. The maintenance requirement for oxygen is potentially very high and was determined to be 10 to 15 mmol g^–1 h^–1.     

Therlt11 andraec1 mutants ofArabidopsis thaliana lack the activity of a basic-amino-acid transporter

The concentration dependence of the influx ofl-lysine in excised roots ofArabidopsis thaliana seedlings was analyzed for the wild-type (WT) and two mutants,rlt11 andraec1, which had been selected as resistant to lysine plus threonine, and to S-2-aminoethyl-l-cysteine, respectively. In the WT three components were resolved: (i) a high-affinity, low-capacity component [K _m = 2.2 M;V _max = 23 nmol·(g FW)^–1·h^–1]; (ii) a low-affinity, high-capacity component [K _m = 159 M;V _max = 742 nmol·(g FW)^–1·h^–1]; (iii) a component which is proportional to the external concentration, with a constant of proportionalityk = 104 nmol·(g FW)^–1 h^–1];·mM^–1. The influx ofl-lysine in the mutants was lower than in the WT, notably in the concentration range 0.1–0.4 mM, where it was only 7% of that in the WT. In both mutants the reduced influx could be fully attributed to the absence of the low-affinity (high-K _m ) component. This component most likely represents the activity of a specific basic-amino-acid transporter, since it was inhibited by several other basic amino acids (arginine, ornithine, hydroxylysine, aminoethylcysteine) but not byl-valine. The high-affinity uptake ofl-lysine may be due to the activity of at least two general amino acid transporters, as it was inhibitable byl-valine, and could be further dissected into two components with a high affinity (K _i = 1–5 M; and a low affinity (K _i = 0.5–1mM) forl-valine, respectively. Therlt11 andraecl mutant have the same phenotype and the corresponding loci were mapped on chromosome 1, but it is not yet clear whether they are allelic.Abbreviations AEC S-2-aminoethyl-l-cysteine - K _i equilibrium constant - WT wild-type     

Medium optimization for recombinant protein production by Bacillus subtilis

Summary Optimal concentrations of glucose, yeast extract and nutrient broth for the production of -lactamase by B. subtilis were predicted using a statistical experimental design and then tested. More yeast extract (13 vs. 1 g/L), less glucose (7.4 vs. 10 g/L), and less nutrient broth (12.6 vs. 15 g/L) were required to achieve high -lactamase activities (5700 U/L) instead of high cell growth rates (1.2 h^-1).     

Expression of glf Z.m. increases D-mannitol formation in whole cell biotransformation with resting cells of Corynebacterium glutamicum

A recombinant oxidation/reduction cycle for the conversion of D-fructose to D-mannitol was established in resting cells of Corynebacterium glutamicum. Whole cells were used as biocatalysts, supplied with 250 mM sodium formate and 500 mM D-fructose at pH 6.5. The mannitol dehydrogenase gene (mdh) from Leuconostoc pseudomesenteroides was overexpressed in strain C. glutamicum ATCC 13032. To ensure sufficient cofactor [nicotinamide adenine dinucleotide (reduced form, NADH)] supply, the fdh gene encoding formate dehydrogenase from Mycobacterium vaccae N10 was coexpressed. The recombinant C. glutamicum cells produced D-mannitol at a constant production rate of 0.22 g (g cdw)⁻¹ h⁻¹. Expression of the glucose/fructose facilitator gene glf from Zymomonas mobilis in C. glutamicum led to a 5.5-fold increased productivity of 1.25 g (g cdw)⁻¹ h⁻¹, yielding 87 g l⁻¹ D-mannitol from 93.7 g l⁻¹ D-fructose. Determination of intracellular NAD(H) concentration during biotransformation showed a constant NAD(H) pool size and a NADH/NAD⁺ ratio of approximately 1. In repetitive fed-batch biotransformation, 285 g l⁻¹ D-mannitol over a time period of 96 h with an average productivity of 1.0 g (g cdw)⁻¹ h⁻¹ was formed. These results show that C. glutamicum is a favorable biocatalyst for long-term biotransformation with resting cells. Dedicated to Prof. Hermann Sahm on the occasion of his 65th birthday.     

Expression of a thermostable restriction endonuclease in recombinant Escherichia coli cells and optimization of fermentation conditions

Taq I restriction endonuclease gene of the thermophilic eubacterium Thermus aquaticus YT-1 (ATCC 25104) was successfully cloned and expressed in recombinant Escherichia coli cells under the control of the lac promoter/operator system. Higher Taq I endonuclease specific activities and biomass yields were obtained from E. coli ER2508(pUCTaq) cells when they were induced at the late-exponential phase of their growth. Taq I endonuclease expression was found to be host strain-dependent such that, among the three different strains examined, E. coli XL1(pUCTaq) produced the highest specific Taq I endonuclease activities for longer induction periods. Decreasing the inducer concentration from 1 to 0.1 mM not only improved the specific enzyme activity yields but also is more economical, considering the high cost of isopropyl--D-thiogalactopyranoside (IPTG). The optimum culture temperature was found to be 37 °C. Taq I endonuclease specific activity recovered from E. coli XL1(pUCTaq) cells was 935 U/mg under optimum conditions.     

Improvement of the ethanol productivity in a high gravity brewing at pilot plant scale

A 2³ full factorial design was used to study the influence of different experimental variables, namely wort gravity, fermentation temperature and nutrient supplementation, on ethanol productivity from high gravity wort fermentation by Saccharomyces cerevisiae (lager strain), under pilot plant conditions. The highest ethanol productivity (0.69 g l^–1 h^–1) was obtained at 20°P [°P is the weight of extract (sugar) equivalent to the weight of sucrose in a 100 g solution at 20°C], 15°C, with the addition of 0.8% (w/v) yeast extract, 24 mg l^–1 ergosterol and 0.24% (v/v) Tween 80.     

Proliferation and metabolic significance of peroxisomes in Candida boidinii during growth on d-alanine or oleic acid as the sole carbon source

We have studied the induction of peroxisomes in the methylotrophic yeast Candida boidinii by d-alanine and oleic acid. The organism was able to utilize each of these compounds as the sole carbon source and grew with growth rates of =0.20 h^-1 (on d-alanine) or =0.43 h^-1 (on oleic acid). Growth was associated with the development of many peroxisomes in the cells. On d-alanine a cluster of tightly interwoven organelles was observed which made up 6.3% of the cytoplasmic volume and were characterized by the presence of d-amino acid oxidase and catalase. On oleic acid rounded to elongated peroxisomes were dominant which were scattered throughout the cytoplasm. These organelles contained increased levels of -oxidation enzymes; their relative volume fraction amounted 12.8% of the cytoplasmic volume.