Cell mass synthesis and degradation by immobilized Escherichia coli

Escherichia coli K-12 cells were grown in a confined volume using microporous hollow fiber membranes. The local cell concentrations in the reactors were above 400 g dry mass/L, in excess of the predicted limit based on the specific volume of free cells determined by tracer exclusion. Cell mass synthesis and degradation rates in these reactors were measured using radioisotope labeling with (35)S. Net accumulation of cell material persisted at these high cell densities. The rates of substrate uptake and cell growth were predicted from the theory of reaction and diffusion assuming that kinetics of cell metabolism are identical for free-living and immobilized cells. This theory was tested by comparison of overall rates and by the size of the region in which cell growth occurred, measured by autoradiography. A yield coefficient of 4 +/- 1 mol sulfur/mol glucose was measured, in agreement with the value determined for free-living cells in similar conditions. Cell growth occurs in a thin layer (10-30 mum), at a rate similar to the growth rate for free cells. Volume expansion by the cells as a consequence of proliferation induces convection of cell mass out of the growth region into a region of the reactor filled with starving cells, which then accumulate in the reactor. The combination of mass-balance and spatial distribution measurements made possible by the use of radioisotope labeling enables a direct test for mass transfer limitations, the determination of the intrinsic cell kinetics, and noninvasive measurements of cell growth in immobilized cell reactors.     

Autoradiographic determination of mass-transfer limitations in immobilized cell reactors

Pseudomonas putida cells were grown in confined volumes in dual-membrane immobilized cell reactors constructed from microporous polyethylene hollow fibers and silicone rubber tubules as a model system for the study of mass transport in microbial aggregates. Local cell concentrations in the reactors reached 300 g dry mass/L. Pulse-chase radioisotope labeling with (35)SO(4) (2-) was used to estimate the rates of cell mass synthesis and degradation. Sulfur incorporation consistently exceeded sulfur release, implying that the cell mass concentration continually increases. The location and size of the cell growth region was determined using liquid emulsion autoradiography of thin sections prepared from labeled reactors. Cell growth occurs in a region less than 25 mum in depth adjacent to the oxygen supply, and the expansion of the cells caused by cell growth promotes convection of the cell mass into regions of the reactor where starving cells accumulate. The combination of mass-balance and spatial distribution measurements that can be made using radioisotope tracers provides a versatile method for determining metabolic rates and limitations caused by mass transfer in immobilized cell reactors.     

Product inhibition of immobilized Escherichia coli arising from mass transfer limitation.

Mass transfer-limited removal of metabolic products led to product-inhibited growth of Escherichia coli that was immobilized in a model system. Comparison of the growth kinetics of immobilized and free-living cells revealed no further physiological differences between cells in these two modes of existence beyond those manifested in the local concentrations of substrate and product. Bacteria were retained on a microporous membrane in a dense, planar aggregate and were grown anaerobically on a glucose-based minimal medium. Radioisotope labeling of the immobilized cell mass with 35S was used to determine growth kinetic parameters. Growth rates in the immobilized cell layer were measured by an autoradiographic technique which allowed comparison of the size of the growing region with the rate of cell convection caused by growth. Immobilized cell growth rates and growth yields ranged from near maximal (0.56 h-1 and 39 g of dry cell weight/mol of glucose, respectively) to substantially reduced (0.15 h-1 and 15 g/mol). The depression of these kinetic parameters was attributed to product inhibition arising from mass transfer-limited removal of acidic waste products from the cell mass. A simple one-dimensional reaction-diffusion model, which incorporated data on the product-inhibited growth kinetics of free-living cells collected in a product-limited chemostat, satisfactorily predicted product inhibition of immobilized cell growth.     

Product inhibition of immobilized Escherichia coli arising from mass transfer limitation

Mass transfer-limited removal of metabolic products led to product-inhibited growth of Escherichia coli that was immobilized in a model system. Comparison of the growth kinetics of immobilized and free-living cells revealed no further physiological differences between cells in these two modes of existence beyond those manifested in the local concentrations of substrate and product. Bacteria were retained on a microporous membrane in a dense, planar aggregate and were grown anaerobically on a glucose-based minimal medium. Radioisotope labeling of the immobilized cell mass with 35S was used to determine growth kinetic parameters. Growth rates in the immobilized cell layer were measured by an autoradiographic technique which allowed comparison of the size of the growing region with the rate of cell convection caused by growth. Immobilized cell growth rates and growth yields ranged from near maximal (0.56 h-1 and 39 g of dry cell weight/mol of glucose, respectively) to substantially reduced (0.15 h-1 and 15 g/mol). The depression of these kinetic parameters was attributed to product inhibition arising from mass transfer-limited removal of acidic waste products from the cell mass. A simple one-dimensional reaction-diffusion model, which incorporated data on the product-inhibited growth kinetics of free-living cells collected in a product-limited chemostat, satisfactorily predicted product inhibition of immobilized cell growth.     

Diffusional limitations of immobilized Escherichia coli in hollow-fiber reactors: influence on 31P NMR spectroscopy

Escherichia coli cells were immobilized and grown in hollow-fiber reactors allowing simultaneous NMR spectroscopy and perfusion with nutrient medium. The extent to which the cells were starved due to inadequate mass transfer was predicted using a mathematical model of reaction and diffusion. Reactors were experimentally characterized using (35)S autoradiography to visualize spatial variations in protein synthesis rates and transmission electron microscopy to indicate spatial variations in cell morphology. Mass transfer limitations in reactors operated at 37 degrees C were shown to be severe, with regions of starved cells occupying up to 80% of the cell-containing region. Phosphorus-31 nuclear magnetic resonance (NMR) spectra of the immobilized, perfused cells revealed abnormally low volume-averaged concentrations of sugar phosphates, NTP, and ratios of NTP/NDP in these reactors. Intracellular pH was also depressed in the cells. In order to overcome mass transfer limitations in the cell layer, the reactor growth temperature was decreased. Sulfur-35 autoradiographs of a reactor operated at 16 degrees C did not indicate the presence of starved cells. The NMR spectra obtained from this reactor showed near-normal intracellular pH, metabolite concentrations, and NTP/NDP ratios. The presence of significant mass transfer limitations in a perfused cell sample during NMR spectroscopy is generally undesirable since the resulting spectra can be ambiguous and difficult to interpret. The strategy adopted in this work, namely estimation of the relative rates of reaction and diffusion in the cell mass and appropriate changes in reactor design and operating parameters, should prove generally applicable for the design of perfused cell samples for NMR spectroscopic experiments.     

Kinetic study of hybridoma cell growth in continuous culture. I. A model for non-producing cells

The kinetic behavior of a nonproducing hybridoma clone AFP-27-NP was investigated in continuous culture under glucose-limited conditions. A total of more than 21, 000 h of cultures were operated at dilution rates ranging from 0.01 to 0.06 h(-1). The viable cell concentrations, dead cell concentrations, and cell volumes all varied with the dilution rate. A steady-state model was developed based on the biomass concentration and the glucose concentration. The specific growth rate as a function of glucose concentration is described by a model similar to the Monod model with a threshold glucose concentration and a minimum specific growth rate incorporated; the model is meaningful only at glucose concentrations and specific growth rates above these levels. A death rate is included in the model which is described by an inverted Monod-type function of glucose concentration. The yield coefficient based on glucose is constant in the lower range of specific growth rates and changes to a new constant value in the upper region of specific growth rates. No maintenance term for glucose consumption was needed; in the plot of specific glucose consumption rate vs. specific growth rate, the line intercepted the specific growth rate axis at a value close to the minimum growth rate. The values for the model parameters were determined from regression analysis of the steady-state data. The model predictions and experimental results fit very well.     

Intrinsic kinetic parameters of substrate utilization by immobilized anaerobic sludge

This article presents a method for evaluating the intrinsic kinetic parameters of the specific substrate utilization rate (r) equation and discusses the results obtained for anaerobic sludge-bed samples taken from a horizontal-flow anaerobic immobilized sludge (HAIS) reactor. This method utilizes a differential reactor filled with polyurethane foam matrices containing immobilized anaerobic sludge which is subjected to a range of feeding substrate flow rates. The range of liquid superficial velocities thus obtained are used for generating data of observed specific substrate utilization rates (r(obs)) under a diversity of external mass transfer resistance conditions. The r(obs) curves are then adjusted to permit their extrapolation for the condition of no external mass transfer resistance, and the values determined are used as a test for the condition of absence of limitation of internal mass transfer. The intrinsic parameters r(max), the maximum specific substrate utilization rate, and K(s), the half-velocity coefficient, are evaluated from the r values under no external mass transfer resistance and no internal mass transfer limitation. The application of such a method for anaerobic sludge immobilized in polyurethane foam particles treating a glucose substrate at 30 degrees C resulted in intrinsic r(max) and K(s), respectively, of 0.330 mg chemical oxygen demand (COD) . mg(-1) volatile suspended solids (VSS) . h(-1) and 72 mg COD . L(-1). In comparison with the values found in the literature, intrinsic r(max) is significantly high and intrinsic K(s) is relatively low. (c) 1997 John Wiley & Sons, Inc.     

Mathematical modelling of microenvironment and growth in EMT6/Ro multicellular tumour spheroids

In order to determine the role of micromilieu in tumour spheroid growth, a mathematical model was developed to predict EMT6/Ro spheroid growth and microenvironment based upon numerical solution of the diffusion/reaction equation for oxygen, glucose, lactate ion, carbon dioxide, bicarbonate ion, chlorine ion and hydrogen ion along with the equation of electroneutrality. This model takes into account the effects of oxygen concentration, glucose concentration and extracellular pH on cell growth and metabolism. Since independent measurements of EMT6/Ro single cell growth and metabolic rates, spheroid diffusion constants, and spinner flask mass transfer coefficients are available, model predictions using these parameters were compared with published data on EMT6/Ro spheroid growth and micro-environment. The model predictions of reduced spheroid growth due to reduced cell growth rates and cell shedding fit experimental spheroid growth data below 700 microns, but overestimated the spheroid growth rate at larger diameters. Predicted viable rim thicknesses based on predicted near zero glucose concentrations fit published viable rim thickness data for 1000 microns spheroids grown at medium glucose concentrations of 5.5 mM or less. However, the model did not accurately predict the onset of necrosis. Moreover, the model could not predict the observed decreases in oxygen and glucose metabolism seen in spheroids with time, nor could it predict the observed growth plateau. This suggests that other unknown factors, such as inhibitors or cell-cell contact effects, must also be important in affecting spheroid growth and cellular metabolism.     

Real-time detection of 13C NMR labeling kinetics in perfused EMT6 mouse mammary tumor cells and betaHC9 mouse insulinomas

A method was developed for obtaining high signal-to-noise 13C NMR spectra of intracellular compounds in metabolically active cultured cells. The method allows TCA cycle labeling kinetics to be determined in real time without significant oxygen transport limitations. Cells were immobilized on the surface of nonporous microcarriers that were either uncoated or coated with polypeptides and used in a 12-cm3 packed bed. The methods were tested with two EMT6 mouse mammary tumor cell lines, one strongly adherent and the other moderately adherent, and a weakly adherent mouse insulinoma line (betaHC9). For both EMT6 lines, NTP and oxygen consumption measurements indicated that the number of cells in the spectrometer ranged from 6 x 10(8) to 1 x 10(9). During infusion of [1-13C]glucose, labeling in C-4 glutamate (indicative of flux into the first half of the TCA cycle) could be detected with 15-min resolution. However, labeling for C-3 and C-2 glutamate (indicative of complete TCA cycle activity) was fivefold lower and difficult to quantify. To increase TCA cycle labeling, cells were infused with medium containing [1,6-13C2]glucose. A 2.5-fold increase was observed in C-4 glutamate labeling and C-3 and C-2 glutamate labeling could be monitored with 30-min resolution. Citrate synthase activity was indirectly detected in real time, as [3,4-13C2]glutamate was formed from [2-13C]oxaloacetate and [2-13C]acetate (of acetyl-CoA). Cell mass levels observed with betaHC9 cells were somewhat lower. However, the 13C S/N was sufficient to allow real-time monitoring of the response of intracellular metabolite labeling to a step change in glucose and a combined glutamine/serum pulse.     

Growth and production characteristics of permeabilized Coleus blumei cells in immobilized fed-batch culture

Summary Permeabilized Coleus blumei cells were cultivated in an immobilized state to study the effect of dimethyl sulfoxide (DMSO) concentrations and growth regulators on cell growth and rosmarinic acid (RA) production characteristics. Luffa (the fibrous skeleton of mature fruit of Luffa cylindrica) was a good support matrix for cell immobilization because of its high void volume. Maximum cell loading capacity was 1.33 g dry cell weight (DCW)/g dry Luffa. The experiments were done in shake flasks with no free medium. The medium was supplied in a fed-batch mode to avoid the flotation of Luffa pieces. The sucrose in the medium was completely hydrolyzed to glucose and fructose without any sugar accumulation in the medium. The cell viability was slightly higher in the cells on top of the Luffa than those in the middle. Cell growth rate and rosmarinic acid (RA) production were approximately half that obtained in cell suspension cultures. Cell yield (g DCW/g glucose) was similar to that of cell suspension cultures. The absence of growth regulators did not promote an increase of RA production but did decrease the cell mass. The second step preconditioning with 0.5% DMSO did not improve the cell's adaptability to higher DMSO concentrations and the cell mass did not increase with 2.5% DMSO.     

Growth model and metabolic activity of brewing yeast biofilm on the surface of spent grains: a biocatalyst for continuous beer fermentation

In the continuous systems, such as continuous beer fermentation, immobilized cells are kept inside the bioreactor for long periods of time. Thus an important factor in the design and performance of the immobilized yeast reactor is immobilized cell viability and physiology. Both the decreasing specific glucose consumption rate (q(im)) and intracellular redox potential of the cells immobilized to spent grains during continuous cultivation in bubble-column reactor implied alterations in cell physiology. It was hypothesized that the changes of the physiological state of the immobilized brewing yeast were due to the aging process to which the immobilized yeast are exposed in the continuous reactor. The amount of an actively growing fraction (X(im)act) of the total immobilized biomass (X(im)) was subsequently estimated at approximately X(im)act = 0.12 g(IB) g(C)(-1) (IB = dry immobilized biomass, C = dry carrier). A mathematical model of the immobilized yeast biofilm growth on the surface of spent grain particles based on cell deposition (cell-to-carrier adhesion and cell-to-cell attachment), immobilized cell growth, and immobilized biomass detachment (cell outgrowth, biofilm abrasion) was formulated. The concept of the active fraction of immobilized biomass (X(im)act) and the maximum attainable biomass load (X(im)max) was included into the model. Since the average biofilm thickness was estimated at ca. 10 microm, the limitation of the diffusion of substrates inside the yeast biofilm could be neglected. The model successfully predicted the dynamics of the immobilized cell growth, maximum biomass load, free cell growth, and glucose consumption under constant hydrodynamic conditions in a bubble-column reactor. Good agreement between model simulations and experimental data was achieved.     

Immobilized cell biocatalyst activation and pseudo-steady-state behavior: model and experiment

An intrinsic, unstructured model has been utilized to describethe startup dynamics of a continuous Caalginate-immobilized Zymomonas mobilis (ATCC 10988) fermentation. This model predicts, at least qualitatively, transients in the fermenter effluent glucose, ethanol, and biomass concentrations as well as radial gradients in immobilized-cell concentration and activity within the gelbiocatalysts. Predicted intrabiocatalyst gradients in immobilized-cell specific growth rate were used to calculate the corresponding gradients in intracellular RNA level based on a reported linear relationship between the two. Mathematical simulations of immobilized biomass concentration profiles and RNA content were verified using a novel, scanning microfluorimetry technique.     

Determination of the intrinsic kinetic constants of immobilized glucose oxidase and catalase

Models of membrane systems containing immobilized glucose oxidase and catalase operating together or independently have been developed. A rotated disk electrode apparatus was employed with novel electrochemical operating conditions to experimentally determine mass transfer and intrinsic kinetic parameters of enzyme-containing membranes. The value of a mass transfer parameter that describes internal and external diffusion was first determined under conditions that do not permit the enzyme reactions. In a subsequent experiment with the reaction allowed, kinetic parameters corresponding to the intrinsic maximal velocity and Michaelis constants of the immobilized enzymes were estimated by regression analysis of data based on an appropriate two- or three- parameter model. It was found that immobilization reduced the maximal intrinsic velocity but had no detectable effect on the Michaelis constants. In all but one case- these methods for membrane characterization are nondestructive and can be used repeatedly on a given membrane. These techniques provide the means for quantitative comparisons of immobilization methods and make possible temporal studies of immobilized enzyme inactivation.     

Mass transfer effects on the reaction rate for heterogeneously distributed immobilized yeast cells

Here we examine the efficiency of different immobilized cell gradients applied to immobilized Saccharomyces cerevisiae fermenting glucose to ethanol. We developed a simulation model to fully study the competing effects of mass transfer hindrance and kinetics. It is based on a diffusion-reaction model and can be used to analyze the different cell concentration profiles inside an immobilized gel bead, in terms of effectiveness factors, productivity, and mass flux. The internal diffusion coefficient, which varies with the local cell concentration, as well as the external mass transfer, is taken into account when describing the efficiency. Although the diffusion hindrance is greater at higher cell concentrations, high cell concentration is still advantageous in the present case because the increase in reaction rate outweighs the diffusion hindrance. Thus, high cell concentrations contribute to increased productivity. The influence of the cell concentration gradient on the efficiency of the beads is negligible. Within the range of cell profiles studied it has been established that the location of the cells within the bead is of lesser importance. However, a steep cell gradient increases the importance of the external mass transfer.     

Synthesis and lysis of formate by immobilized cells of Escherichia coli

Formate hydrogenlyase (FHL) activity was induced in a strain of Escherichia coli S13 during anaerobic growth in yeast extract-tryptone medium containing 100 mM formate. The cells obtained at the optimum growth phase were immobilized in 2.5% (w/v) agar gel when 50-60% of the whole cell FHL activity was retained. The immobilized FHL system had good storage stability and recycling efficiency. In the lysis of formate, an increase of formate concentration to 1.18M increased QH(2) (initial) value of the immobilized cell, and subsequently cells, hydrogen evolution, in general, ceased after 6 to 8 of incubation, resulting in incomplete lysis of formate. Presence of small amount of glucose (28 mM) was more or less quantitatively lysed with concomitant disappearence of glucose from the medium. Synthesis of formate from hydrogen and bicarbonate solution by the immobilized cells was also characterized. Presence of glucose (10 mM) in 50 mM bicarbonate solution stimulated formate synthesis by immobilized cells. The pH optimum range, K(m), and specific activity of the immobilized cells for the lysis of formate were 6.8-7.2 0.4M, and 66 mL/g cell-h, respectively. The cells could fix hydrogen to the extent of 24.4% (w/w) of its own wet cell mass in a 72-h reaction cycle. Potentiality of the immobilized FHL system for biotechnological exploitation was discussed.     

Minimal nutritional requirements for immobilized yeast

The effect of reduced nutritional levels (particularly nitrogen source) for immobilized K. fragilis type yeast were studied using a trickle flow, "differential" plug flow type reactor with cells immobilized by adsorption onto an absorbant packing matrix. Minimizing nutrient levels in a feed stream to an immobilized cell reactor (ICR) might have the benefits of reducing cell growth and clogging problems in the ICR, reducing feed preparation costs, as well as reducing effluent disposal costs. In this study step changes in test feed medium nutrient compositions were introduced to the ICR, followed by a return to a basal medium. Gas evolution rates were monitored and logged on a continuous basis, and effluent cell density was used as an indicator of cell growth rate of the immobilized cell mass. Startup of the reactor using a YEP medium showed a rapid buildup of cells in the reactor during the initial 110 h operation. The population density then stabilized at 1.6 x 10(11) cells/g sponge. A defined medium containing a complex mix of essential nutrients with an inorganic nitrogen source (ammonium sulfate) was able to maintain 90% of the productivity in the ICR as compared to the YEP medium, but proved unable to promote growth of the immobilized cell mass during startup. Experiments on reduced ammonium sulfate in the defined medium, and reduced yeast extract and peptone in YEP medium indicated that stable productivity could be maintained for extended periods (80 h) in the complete absence of any nutrients besides a few salts (potassium phosphate and magnesium sulfate). It was found that productivity rates dropped by 35-65% from maximal values as nitrogenous nutrients were eliminated from the test mediums, while growth rates (as determined by shed cell density from the reactor) dropped by 75-95%. Thus, nutritional deficiencies largely decoupled growth and productivity of the immobilized yeast which suggests productivity is both growth- and non-growth-associated for the immobilized cells. A yeast extract concentration of 0.375 g/L with or without 1 g/L ammonium sulfate was determined to be the minimum level which gave a sustained increase in productivity rates as compared to the nutritionally deficient salt medium. This represents a 94% reduction in complex nitrogenous nutrient levels compared to standard YEP batch medium (3 g/L YE and 3.5 g/L peptone).     

Dynamic optimization of hybridoma growth in a fed-batch bioreactor

This study addressed the problem of maximizing cell mass and monoclonal antibody production from a fed-batch hybridoma cell culture. We hypothesized that inaccuracies in the process model limited the mathematical optimization. On the basis of shaker flask data, we established a simple phenomenological model with cell mass and lactate production as the controlled variables. We then formulated an optimal control algorithm, which calculated the process-model mismatch at each sampling time, updated the model parameters, and re-optimized the substrate concentrations dynamically throughout the time course of the batch. Manipulated variables were feed rates of glucose and glutamine. Dynamic parameter adjustment was done using a fuzzy logic technique, while a heuristic random optimizer (HRO) optimized the feed rates. The parameters selected for updating were specific growth rate and the yield coefficient of lactate from glucose. These were chosen by a sensitivity analysis. The cell mass produced using dynamic optimization was compared to the cell mass produced for an unoptimized case, and for a one-time optimization at the beginning of the batch. Substantial improvements in reactor productivity resulted from dynamic re-optimization and parameter adjustment. We demonstrated first that a single offline optimization of substrate concentration at the start of the batch significantly increased the yield of cell mass by 27% over an unoptimized fermentation. Periodic optimization online increased yield of cell mass per batch by 44% over the single offline optimization. Concomitantly, the yield of monoclonal antibody increased by 31% over the off-line optimization case. For batch and fed-batch processes, this appears to be a suitable arrangement to account for inaccuracies in process models. This suggests that implementation of advanced yet inexpensive techniques can improve performance of fed-batch reactors employed in hybridoma cell culture.     

Cell culture media impact on drug product solution stability

To enable subcutaneous administration of monoclonal antibodies, drug product solutions are often needed at high concentrations. A significant risk associated with high drug product concentrations is an increase in aggregate level over the shelf‐life dating period. While much work has been done to understand the impact of drug product formulation on aggregation, there is limited understanding of the link between cell culture process conditions and soluble aggregate growth in drug product. During cell culture process development, soluble aggregates are often measured at harvest using cell‐free material purified by Protein A chromatography. In the work reported here, cell culture media components were evaluated with respect to their impact on aggregate levels in high concentration solution drug product during accelerated stability studies. Two components, cysteine and ferric ammonium citrate, were found to impact aggregate growth rates in our current media (version 1) leading to the development of new chemically defined media and concentrated feed formulations. The new version of media and associated concentrated feeds (version 2) were evaluated across four cell lines producing recombinant IgG₄ monoclonal antibodies and a bispecific antibody. In all four cell lines, the version 2 media reduced aggregate growth over the course of a 12 week accelerated stability study compared with the version 1 media, although the degree to which aggregate growth decreased was cell line dependent. © 2016 American Institute of Chemical Engineers Biotechnol. Prog., 32:998–1008, 2016     

Fermentation kinetics of free and immobilized Penicillium raistrickii able perform the 15α-hydroxylation of 13-ethyl-gon-4-ene-3,17-dione

Growing Penicillium raistrickii i 477 cells immobilized by microencapsulation, entrapment in calcium alginate beads and photopolymerization were used for the 15α-hydroxylation of 13-ethyl-gon-4-en-3,17-dione (I) to 15α-hydroxy-13-ethyl-gon-4-en-3,17-dione (II). The immobilized cells had lower maximum specific growth rates and yield coefficients when cultivated on the carbon source glucose than the non-immobilized cells, which leads to lower volumetric productivities than the use of nonimmobilized cells. However, the cells immobilized by microencapsulation and calcium alginate entrapment showed a specific productivity equal to that of the respective non-immobilized cells based on product formation per dry biomass and time. Photommobilized cells were not able to grow in the presence of the steroid because the substrate concentrations within the polymer reached inhibiting amounts for growth and product formation. In the absence of the steroid, the growing photoimmobilized cells showed a prolonged lag-phase in comparison with the free cells.     

Growth of Cephalotaxus harringtonia plant-cell cultures

Summary Cell cultures of Cephalotaxus harringtonia were examined to characterize growth kinetics. The requirement for an undefined medium supplement (coconut water) was eliminated by maintaining high cell concentrations in semicontinuous and batch growth. Sucrose fed to batch-cultured cells was completely hydrolyzed and a diauxic growth pattern was observed corresponding to first glucose and then fructose uptake. Examination of increases in cell concentrations on the basis of fresh and dry weight showed that a substantial lag period existed between the initiation of substrate uptake and increases in cell volume. Specific growth rates were highest during periods of glucose uptake, but cell yields were comparable for the two sugars. In contrast, studies with glucose or fructose as the sole carbon source indicated that cell yields were significantly lower with fructose but specific growth rates were comparable for the two sugars.Offprint requests to: P. J. Westgate