Search

Genomic and proteomic based global gene expression profiling has altered the landscape of biological research in the past few years. Its potential impact on cell culture bioprocessing has only begun to emanate, partly due to the lack of genomic sequence information for the most widely used industrial cells, Chinese hamster ovary (CHO) cells. Transcriptome and proteome profiling work for species lacking extensive genomic resources must rely on information for other related species or on data obtained from expressed sequence tag (EST) sequencing projects, for which burgeoning efforts have only recently begun. This article discusses the aspects of EST sequencing in those industrially important, genomic resources-poor cell lines, articulates some of the unique features in employing microarray in the study of cultured cells, and highlights the infrastructural needs in establishing a platform for genomics based cell culture research. Recent experience has revealed that generally, most changes in culture conditions only elicit a moderate level of alteration in gene expression. Nevertheless, by broadening the conventional scope of microarray analysis to consider estimated levels of transcript abundance, much physiological insight can be gained. Examples of the application of microarray in cell culture are discussed, and the utility of pattern identification and process diagnosis are highlighted. As genomic resources continue to expand, the power of genomic tools in cell culture processing research will be amply evident. The key to harnessing the immense benefit of these genomic resources resides in the development of physiological understanding from their application.

Hepatocytes self-assemble in culture to form compacted spherical aggregates, or spheroids, that mimic the structure of the liver by forming tight junctions and bile canalicular channels. Hepatocyte spheroids thus resemble the liver to a great extent. However, liver tissue contains other cell types and has bile ducts and sinusoids formed by endothelial cells. Reproducing 3-D co-culture in vitro could provide a means to develop a more complex tissue-like structure. Stellate cells participate in revascularization after liver injury by excreting between hepatocytes a laminin trail that endothelial cells follow to form sinusoids. In this study we investigated co-culture of rat hepatocytes and a rat hepatic stellate cell line, HSC-T6. HSC-T6, which does not grow in serum-free spheroid medium, was able to grow under co-culture conditions. Using a three-dimensional cell tracking technique, the interactions of HSC-T6 and hepatocyte spheroids were visualized. The two cell types formed heterospheroids in culture, and HSC-T6 cell invasion into hepatocyte spheroids and subsequent retraction was observed. RT-PCR revealed that albumin and cytochrome P450 2B1/2 expression were better maintained in co-culture conditions. These three-dimensional heterospheroids provide an attractive system for in vitro studies of hepatocyte-stellate cell interactions.

Continuous culture is frequently used in the cultivation of mammalian cells for the manufacturing of recombinant protein pharmaceuticals. In such operations a large volume of medium is turned over each day, especially in the case where cell recycle, or perfusion cultivation, is practiced. In principle, the volumetric throughput of medium can be reduced by using a more concentrated feed while maintaining the same nutrient provision rate. Overall, the medium components are divided into two categories: ‘consumable nutrients' and ‘unconsumable inorganic bulk salts’. In such fortified medium, the concentrations of consumable nutrients, but not bulk salts, are increased. With a stoichiometrically-balanced medium, the large amount of nutrients fed into the culture is largely consumed by cells to give rise to residual concentrations of these nutrients in their optimal range. However, unless care is taken to initiate the continuous culture, overshoot of nutrients may occur during the transient period. The high nutrient concentration during overshoot may be inhibitory by itself, or the resulting high osmolality may retard the growth. Using a mathematical model that incorporates the growth inhibitory effect of high osmolality we demonstrate such a potentially catastrophic effect of nutrient and osmolality overshoot by simulation. To avoid overshoot a controlled nutrient feeding scheme should be devised at the initiation of continuous culture.

The metabolism of hybridoma cells was controlled to reduce metabolic formation in fed-batch cultures by dynamically feeding a salt-free nutrient concentrate. For this purpose, on-line oxygen uptake rate (OUR) measurement was used to estimate the metabolic demand of hybridoma cells and to determine the feeding rate of a concentrated solution of salt-free DMEM/F12 medium supplemented with other medium components. The ratios among glucose, glutamine and other medium components in the feeding nutrient concentrate were adjusted stoichiometrically to provide balanced nutrient conditions for cell growth. Through on-line control of the feeding rate of the nutrient concentrate, both glucose and glutamine concentrations were maintained at low levels of 0.5 and 0.2 mM respectively during the growth stage. The concentrations of the other essential amino acids were also maintained without large fluctuations. The cell metabolism was altered from that observed in batch cultures resulting in a significant reduction of lactate, ammonia and alanine production. Compared to a previously reported fed-batch culture in which only glucose was maintained at a low level and only a reduced lactate production was observed, this culture has also reduced the production of other metabolites, such as ammonium and alanine. As a result, a high viable cell concentration of more than 1.0 × 107 cells/mL was achieved and sustained over an extended period. The results demonstrate an efficient nutrient feeding strategy for controlling cell metabolism to achieve and sustain a high viable cell concentration in fed-batch mammalian cell cultures in order to enhance the productivity.

Liver failure is a major cause of mortality. A bioartificial liver (BAL) employing isolated hepatocytes can potentially provide temporary support for liver failure patients. We have developed a bioartificial liver by entrapping hepatocytes in collagen loaded in the luminal side of a hollow fiber bioreactor. In the first phase of development, liver-specific metabolic activities of biosynthesis, biotransformation and conjugation were demonstrated. Subsequently anhepatic rabbits were used to show that rat hepatocytes continued to function after the BAL was linked to the test animal. For scale-up studies, a canine liver failure model was developed using D-galactosamine overdose. In order to secure a sufficient number of hepatocytes for large animal treatment, a collagenase perfusion protocol was established for harvesting porcine hepatocytes at high yield and viability. An instrumented bioreactor system, which included dissolved oxygen measurement, pH control, flow rate control, an oxygenator and two hollow fiber bioreactors in series, was used for these studies. An improved survival of dogs treated with the BAL was shown over the controls. In anticipated clinical applications, it is desirable to have the liver-specific activities in the BAL as high as possible. To that end, the possibility of employing hepatocyte spheroids was explored. These self-assembled spheroids formed from monolayer culture exhibited higher liver-specific functions and remained viable longer than hepatocytes in a monolayer. To ease the surface requirement for large-scale preparation of hepatocyte spheroids, we succeeded in inducing spheroid formation in stirred tank bioreactors for both rat and porcine hepatocytes. These spheroids formed in stirred tanks were shown to be morphologically and functionally indistinguishable from those formed from a monolayer. Collagen entrapment of these spheroids resulted in sustaining their liver-specific functions at higher levels even longer than those of spheroids maintained in suspension. For use in the BAL, a mixture of spheroids and dispersed hepatocytes was used to ensure a proper degree of collagen gel contraction. This mixture of spheroids and dispersed cells entrapped in the BAL was shown to sustain the high level of liver-specific functions. The possibility of employing such a BAL for improved clinical performance warrants further investigations.

In a continuous culture with cell retention the perfusion rate must be adjusted dynamically to meet the cellular demand. An automated mechanism of adjusting the perfusion rate based on real-time measurement of the metabolic load of the bioreactor is important in achieving a high cell concentration and maintaining high viability. We employed oxygen uptake rate (OUR) measurement as an on-line metabolic indicator of the physiological state of the cells in the bioreactor and adjusted the perfusion rate accordingly. Using an internal hollow fiber microfiltration system for total cell retention, a cell concentration of almost 10(8) cells/mL was achieved. Although some aggregates were formed during the cultivation, the viability remained high as examined with confocal microscopy after fluorescent vital staining. The results demonstrate that on-line OUR measurement facilitates automated dynamic perfusion and allows a high cell concentration to be achieved.

A static culture model employing cylindrical collagen-hepatocyte gels is reported for large scale testing of conditions relevant to the three compartment hollow fiber bioartificial liver. High density hepatocyte cultivation was achieved by cell entrapment within the collagen-hepatocyte gel. Hepatocyte viability was assessed by vital staining, gel contraction, and insulin utilization. Measures of hepatocyte-specific function included albumin synthesis, ureagenesis, lidocaine biotransformation, and cholate conjugation. Although hepatocyte viability remained stable through the seven day incubation period, hepatocyte functions were not uniformly preserved. Albumin synthesis remained stable, while representative P-450 and conjugation activities decreased with time. This static culture system will facilitate the development of a hollow fiber bioartificial liver which utilizes cylindrical collagen-hepatocyte gels.

A new bioreactor for animal cell cultivation employs two compartments for cells and medium respectively. The two chambers are separated by an ultrafiltration membrane. Cells and solution of collagen or collagen/chitosan mixture were loaded to the cell chamber and were allowed to form gel inside. Contraction of the cell-laden gel occurred subsequently to create a new zone in the cell chamber. In such a bioreactor cells are retained in the reactor, the high molecular product(s) accumulate in the cell chamber, while the small molecular weight nutrients and metabolites are replenished and removed from the medium chamber. By adjusting the flow rates for cell and medium chambers, the resident time for cells, high and low molecular weight components of the system can be manipulated separately. The new bioreactor, in both flat-bed and hollow-fiber configurations, was used to cultivate recombinant human cell, 293, for Protein C production over 60 to 90 days.

Recombinant human kidney cells, 293, were cultivated in serum-free fed-batch cultures for the production of Protein C. By coupling the feeding of concentrated medium to pH control based on established stoichiometric relations, the titer of Protein C increased by more than ten fold as compared to batch culture, even though the total cell concentration increased only by less than two fold. Such a fed-batch culture is a simple system for enhancing the productivity of mammalian cells in culture.

Many microcarriers used for the cultivation of animal cells do not allow for convenient microscopic observation of cell morphology and viability due to their optical properties. Using fluorescent viable stain combining fluorescein diacetate and ethidium bromide, we observed the distribution, morphology and viability of cells on various microcarriers.

Light scattering properties of hybridoma cells were examined with flow cytometry. Viable and dead cells form two distinct populations. The distribution of the two populations changes during a batch culture. the concentration of dead cells measured by flow cytometry correlates well to that measured by hemacytometer. The distribution based on small-angle light scattering is similar to the distribution based on volume as measured by Elzone particle counter. It thus appears that viable cells form the population with a larger mean cell volume. The results also indicate that the volume of viable cells decreases during the cultivation while that of dead cells remains relatively constant.

Hepatoma cells, HepG2, grew normally on microcarriers even at a relatively high agitation rate if sufficient time was allowed for cell attachment and adhesion. However, if a high agitation rate was applied shortly after initial cell attachment, the growth rate was retarded. This sensitivity to mechanical agitation appears to be dependent on the inoculation cell density.