Autoclavable low cost serum-free cell culture media. The growth of L Cells and BHK cells on peptones

The growth of L-929 cells on a series of peptones, and protein hydrolysates was examined, and it was found that when MEM was supplemented with any of a series of peptones, cell growth was about as good as when serum was used as a supplement. Protein hydrolysates did not support cell growth very well and at higher concentrations actually reduced cell growth. L-929 and L-60TM cells were grown both as monolayers or stationary suspensions and in agitated systems in MEM supplemented with 0.5—1% bactopeptone. The addition of macromolecular compounds, insulin or oleic acid had no effect on cell growth. BHK cells were also grown on media supplemented with bactopeptone but richer media (MEM-alpha, F-12, or RMPI1640) gave higher cell yields. The cells did not form the monolayers observed with fetal calf serum, but a partial suspension system. Addition of a detergent Darvan #2 gave a totally suspension culture in both stationary and agitated systems. The production of Sindbis virus in BHK cells grown in serum-free media was examined and the yield of virus was found to be about the same as that produced in serum-supplemented systems. It is estimated that the cost of cell production media could be reduced by about 90% by the replacement of serum supplement by peptones.     

Development of a serum-free culture medium for the large scale production of recombinant protein from a Chinese hamster ovary cell line

A serum-free medium, WCM5, has been developed for the large scale propagation of CHO (Chinese hamster ovary) cells which express recombinant protein using dihydrofolate reductase as a selectable marker. WCM5 was prepared by supplementing Iscoves medium without lecithin, albumin or transferrin with a number of components which were shown to benefit growth. WCM5 medium contained 5 mg l^–1 human recombinant insulin (Nucellin) but was otherwise protein-free. CHO 3D11^* cells which had been engineered to express a humanised antibody, CAMPATH^*-1H, were routinely grown using serum-containing medium. From a seeding density of 10⁵ cells ml^–1, cells grown in static culture with serum reached a maximal cell density of 6.5×10⁵ cells ml^–1 after 6 days in culture and produced a maximal antibody concentration of 69 mg l^–1 after 11 days in culture. CHO 3D11^* cells grown with serum were washed in serum-free medium then cultured in WCM5 medium. Following a period of adaptation the cell growth and product yield was superior to that achieved with serum-containing medium. CHO cells producing CAMPATH-1H grown in an 8000 l stirred bioreactor seeded with 2×10⁵ cells ml^–1 reached a maximal viable cell density of 2.16×10⁶ cells ml^–1 after 108 h in culture and a maximal antibody concentration of 131.1 mg l^–1 after 122 h in culture.Abbreviations CHO Chinese hamster ovary - dhfr dihydrofolate reductase - dhfr dihydrofolate reductase deficient - MTX methotrexate - H hypoxanthine - T thymidine - T/V trypsin versene - F12 Hams F12 medium - NEAA non essential amino acids     

Epidermal growth factor-dependent growth of human KB cells in a defined medium and altered growth factor requirements of KB mutants resistant to EGF-Pseudomonas exotoxin conjugates

A serum-free culture system was established for human KB carcinoma (HeLa) cells that consisted of a chemically defined medium and several growth factors including epidermal growth factor (EGF), insulin, transferrin, hydrocortisone, and ethanolamine. EGF and insulin showed the greatest effects on the growth rate of KB cells. Insulin-like growth factor I (IGF-I) at the same concentration as insulin stimulated cell growth less than insulin. Transferrin, hydrocortisone, or ethanolamine had no growth-stimulatory effects alone but were stimulatory when combined with EGF and/or insulin. Transforming growth factor-beta inhibited growth and triiodothyronine stimulated growth. The growth factor requirements were established for several KB mutants with low EGF receptor levels that had been selected for resistance to a conjugate of EGF with Pseudomonas exotoxin (EGF-PE). Three of five KB mutants did not respond to EGF; two other mutants responded to a lesser extent than the parental KB cells. Four mutants had a reduced response to insulin and responded to T3; one mutant (ET-30) responded to neither. These results indicate that KB cells selected for EGF-PE resistance have lost their growth response to EGF and illustrate the usefulness of serum-free medium for studying the growth factor requirements of mutants with altered receptor levels.     

Effect of cyclophosphamide in vitro and on vaccinia virus replication in tissue culture.

The effect of cyclophosphamide on the growth of Vero, BSC-1, and HeLa cells in monolayer cultures was studied. By using hemocytometer counts and tritiated thymidine uptake as indicators of growth, it was found that cyclophosphamide significantly interfered with the metabolism of Vero and BSC-1 cells when sustained in Leibovitz medium. Vero cells and HeLa cells grown in Eagle medium were not affected by exposure to cyclophosphamide. Vaccinia virus replication in Vero cell monolayer cultures incubated with cyclophosphamide was markedly augmented, and this enhanced growth was reflected by virus quantitation techniques and metabolic studies using tritiated thymidine uptake. No difference in the distribution of infectious particles was found when cyclophosphamide-treated and control infected cultures were compared. Pathways other than through hepatic enzymes appear available to activate cyclophosphamide in vitro. These effects are dependent on both the cell type and the medium in which the cells are grown. Cyclophosphamide can facilitate vaccinia virus replication in vitro through metabolic interactions at the cellular level. The precise mechanisms underlying this effect require further study.     

Effect of iron on adherence and cytotoxicity of Entamoeba histolytica to CHO cell monolayers

Iron is an essential element for almost all living organisms. The possible role of iron for growth, adherence and cytotoxicity of Entamoeba histolytica was evaluated in this study. The absence of iron from TYI-S-33 medium stopped amebic growth in vitro. However, iron concentrations in the culture media of 21.4-285.6 microM did not affect the growth of the amebae. Although growth was not retarded at these concentrations, the adhesive abilities of E. histolytica and their cytotoxicities to CHO cell monolayer were correlated with iron concentration. Amebic adhesion to CHO cell monolayers was significantly reduced by low-iron (24.6 +/- 2.1%) compared with 62.7 +/- 2.8 and 63.1 +/- 1.4% of amebae grown in a normal-iron and high-iron media, respectively. E. histolytica cultured in the normal- and high-iron media destroyed 69.1 +/- 4.3% and 72.6 +/- 5.7% of cultured CHO cell monolayers, but amebae grown in the low-iron medium showed a significantly reduced level of cytotoxicity to CHO cells (2.8 +/- 0.2%). Addition of divalent cations other than iron to amebic trophozoites grown in the low-iron medium failed to restore levels of the cytotoxicity. However, when E. histolytica grown in low-iron medium were transferred to normal-iron medium, the amebae showed completely restored cytotoxicity within 7 days. The result suggests that iron is an important factor in the adherence and cytotoxicity of E. histolytica to CHO cell monolayer.     

Change of insulin-like growth factor gene expression in Chinese hamster ovary cells cultured in serum-free media

Although the sera used in animal cell culture media provide the macromolecules, nutrients, hormones, and growth factors necessary to support cell growth, it could also be an obstacle to the production of recombinant proteins in animal cell culture systems used in many sectors of the biotechnology industry. For this reason, many research groups, including our laboratory, have been trying to develop serum-free media (SFM) or serum-supplemented media (SSM) for special or multi-purpose cell lines. The Chinese hamster ovary (CHO) cell, for example, is frequently used to produce proteins and is especially valuable in the large-scale production of pharmaceutically important proteins, yet information about its genome is lacking. Also, SFMs have only been evaluated by comparing growth patterns for cells grown in SFMs with those grown in SSM or by measuring the titer of the target protein obtained from cells grown in each type of medium. These are not reliable methods of obtaining the type of information needed to determine whether an SFM should be replaced with an SSM. We carried out a cDNA microarray analysis to evaluate MED-3, an SFM developed in our laboratory, as a CHO culture medium. When CHO cells were cultured in MED-3 instead of an SSM, several genes associated with cell growth were down-regulated, although this change diminished over time. We found that the insulin-like growth factor (IGF) gene was representative of the proteins that were down-regulated in cells cultured in MED-3. When several key supplements-including insulin, transferrin, ethanolamine, and selenium-were removed from MED-3, theIGF expression was consistently down-regulated and cell growth decreased proportionately. Based on these results, we concluded that when an SFM is used as a culture medium, it is important to supplement it with substances that can help the cells maintain a high level ofIGF expression. The data presented in this study, therefore, might provide useful information for the design and development of SFM or SSM, as well as for the design of genome-based studies of CHO cells to determine how they can be used optimally for protein production in pharmaceutical and biomedical research.     

Growth behavior of Chinese hamster ovary cells in a compact loop bioreactor: 1. Effects of physical and chemical environments.

Chinese hamster ovary (CHO) cells were cultivated in a compact loop bioreactor using MEM-alpha medium supplemented with 10% fetal calf serum. Effects of physical and chemical environments, i.e., pH in the medium, stirring speed of impellers, temperature and partial pressure of oxygen (pO2) upon growth of suspended cells in the bioreactor were determined in batch cultures. Growth behavior was characterized by specific rates of growth (mu), glucose consumption (qG) and lactate production (qL), and the yield coefficients (cell yield from glucose, YX/G, and lactate yield from glucose, YL/G). An effect of medium osmolality was also evaluated with T-flask monolayer cultivation. The best growth was observed at pH 7.6, 37 degrees C, 400 rpm, 50-100% saturation with oxygen and 320 mOsmol kg-1. Corresponding to the previous work with a human melanoma cell line, the sophisticated cultivation and process control systems have been improved for CHO cells.     

Microfilament bundles in cultured cells : Correlation with anchorage independence and tumorigenicity in nude mice

The distribution of actin microfilament bundles in cell lines 3T3B, CHO, HeLa and CLID extracted with 0.1% Triton X-100 was examined by indirect immunofluorescence using human actin antibodies and by electron microscopy of whole cells grown directly on support grids. Anchorage dependence as determined by growth in soft agar and tumorigenicity in nude mice was also investigated. Immunofluorescent staining showed that CHO and HeLa cells have normal numbers and distributions of actin microfilament bundles as compared with similarly spread control 3T3B cells. A significant fraction of the mouse CLID cells showed comparable numbers of microfilament bundles as 3T3B cells but their distribution was markedly different. In many cases the bundles radiated from a region close to the cell's centre or near its projections and usually penetrated the projections. The presence of diffuse staining in 4% of the cell population also indicated the existence in these cells of disorganized actin. Electron microscope studies of well spread regions of negatively-stained, Triton-extracted cells corroborated the observations made with the immunofluorescence technique. In 3T3B, CHO and CLID cells abundant microtubules were found, colinearly arranged with actin filaments in the thin cytoplasmic extensions. While CLID, CHO and HeLa cells showed the capacity to grow in soft agar, only CLID and HeLa cells produced tumours in athymic nude mice. The observations suggest that a reduction or disorganisation of the actin microfilament bundles may not in itself be essential at least for the non-virally transformed cells studied to show anchorage independence or to produce tumours in nude mice.     

Evaluation of various serum and animal protein free media for the production of a veterinary rabies vaccine in BHK-21 cells

We have carried out the adaptation of BHK-21 cells to two serum free (Ex Cell 520 and HyQ PF CHO) and three animal protein free media: Ex Cell 302, HyQ PF CHO MPS and Rencyte BHK. After a direct switch or a gradual adaptation, we have achieved BHK-21 cells growth in the following media: HyQ PF CHO, HyQ PF CHO MPS, Rencyte BHK and Ex Cell 302. The most suitable media for BHK-21 cells growth, with respect to cell density and specific growth rate, were HyQ PF CHO and HyQ PF CHO MPS. Hence we have selected these media to study cell growth and the production of rabies virus. Kinetic studies of cell growth in spinner flasks using the selected media have shown that a maximal cell density of 2x10(6) cells x ml(-1) was reached in both media. For rabies virus production, the viral titer obtained was 1.7x10(6) FFU x ml(-1) in HyQ PF CHO as well as in HyQ PF CHO MPS medium. The optimization of rabies virus production by BHK-21 cells grown in a 2 l bioreactor using the selected media, pointed to the following parameters: culture mode, perfusion rate and multiplicity of infection (MOI), as being the critical factors for achieving a good virus yield. When tested in mice, the activity of the experimental vaccines prepared on HyQ PF CHO MPS medium has shown a protective activity that meets WHO requirements.     

Super-CHO—A cell line capable of autocrine growth under fully defined protein-free conditions

Chinese Hamster Ovary (CHO) cells are widely used for the large scale production of recombinant biopharmaceuticals. Growth of the CHO-K1 cell line has been demonstrated in serum-free medium containing insulin, transferrin and selenium. In an attempt to get autocrine growth in protein-free medium, DNA coding for insulin and transferrin production was transfected into CHO-K1 cells. Transferrin was expressed well, with clones secreting approximately 1000 ng/10⁶ cells/24h. Insulin was poorly expressed, with rates peaking at 5 ng/10⁶ cells/24h. Characterisation of the secreted insulin indicated that the CHO cells were incompletely processing the insulin molecule. Site-directed mutagenesis was used to introduce a furin (prohormone converting enzyme) recognition sequence into the insulin molecule, allowing the production of active insulin. However, the levels were still too low to support autocrine growth. Further investigations revealed insulin degrading activity (presumably due to the presence of insulin degrading enzymes) in the cytoplasm of CHO cells. To overcome these problems insulin-like growth factor I (instead of insulin) was transfected into the cells. IGF-1 was completely processed and expressed at rates greater than 500 ng/10⁶cells/24h. In this paper we report autonomous growth of the transfected CHO-K1 cell line expressing transferrin and IGF-1 in protein-free medium without the addition of exogenous growth factors. Growth rates and final cell densities of these cells were identical to that of the parent cell line CHO-K1 growing in insulin, transferrin, and selenium supplemented serum-free media.     

Effects of insulin and LongR(3) on serum-free Chinese hamster ovary cell cultures expressing two recombinant proteins

Insulin is the most commonly used growth factor for sustaining cell growth and viability in serum-free Chinese hamster ovary (CHO) cell cultures. In the present study insulin and IGF-1 analogue (LongR(3)) were compared for their ability to support growth, viability, and production of two serum-free CHO cell lines expressing recombinant protein. The first cell line, VA12, expresses protein B, and the second cell line, CL23, expresses protein C. Both molecules are recombinant cytokine receptors. VA12 will grow in serum-free media lacking growth factor, while CL23 requires either insulin or LongR(3) for cell growth. Both cell lines, however, require a growth factor for optimal performance under production conditions. In this study, LongR(3) was better able to sustain the viability of both cell lines under production conditions than insulin. These data indicate that while insulin and LongR(3) can both serve as growth and viability factors for CHO cells, LongR(3) is the preferred growth factor for cell lines VA12 and CL23.     

Chlamydia trachomatis in cell culture. II. Susceptibility of seven established mammalian cell types in vitro. Adaptation of trachoma organisms to McCoy and BHK-21 cells

Trachoma organisms of serotype B were grown serially in irradiated cells (McCoy, BHK-21, Microbiological Associates, and BHK-21, Lister) and tested for infectivity in monolayers of five mammalian cell lines (BHK-21, CHO, HeLa S3, McCoy and OWMK) and two diploid strains (ST/BTL and WI-38). All cell types had low susceptibility to chlamydial infection but the number of inclusions increased when the inoculum was centrifuged onto the monolayers, or when the cells were irradiated. Infection was higher in non-irradiated CHO than in irradiated CHO in three out of a total of six experiments. Inclusion number was increased 300 times in HeLa S3 and up to three times in the other cell types after treatment with diethylaminoethyl-dextran (DEAE-D). Serial passage of Chlamydia trachomatis serotype B (strain Har-36) in CO60 McCoy and CO60 BHK-21 Lister resulted in partial adaptation of the strain to the host cell. The phenomenon of adaptation of serotype B to McCoy compensated for the lower susceptibility of this cell revealed when McCoy cells were inoculated with trachoma elementary bodies grown in BHK-21 Lister or in chick embryo yolk sac. Trachoma organisms of immunotypes A, B and C prepared in yolk sac produced more inclusion-forming units per ml in CO60 BHK-21 Lister than in CO60 McCoy.     

Cellular selenoproteins and the effects of selenite on cell proliferation

The incorporation of radioactive selenium into cellular proteins and the effect of selenite on proliferation were examined in human (HeLa, HT-29, and IMR-90) and mouse (3T3 and CMT-93) cell lines. Proteins incorporating selenium were detected by one-dimensional sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Major polypeptide subunits at 60, 23, 21, 19, and 16 kD were detected in the two tumorigenic and one normal human cell lines. The 23 kD polypeptide migrated to the same position on the gel as the major subunit of human erythrocyte glutathione peroxidase. In the mouse cells, the 60 kD polypeptide was almost entirely absent; four other major selenoproteins were detected, with molecular weights similar to those in the human cells. In both mouse and human cells, the same pattern of selenoproteins was observed irrespective of whether the cells were grown in medium containing 10% fetal bovine serum or in defined medium supplemented with 0.1 or 1 microM selenite, or with 1% serum. The effect of selenite on proliferation of HeLa, HT-29, and CMT-93 cells in medium supplemented with 10% fetal bovine serum and in serum-free medium was examined. At concentrations up to about 1 microM, selenite stimulated proliferation of the human cells slightly in serum-free medium but not in serum-supplemented medium. At concentrations of about 5 microM and higher selenite significantly inhibited proliferation of all cells in both types of media. In CMT-93 cells, this inhibition was greater in serum-free medium, but there were no significant differences in this regard in the human cells. These results demonstrate that selenium is stably incorporated into several polypeptides in human and mouse cells, that there are no significant differences in this regard among several cell lines, and slight differences between human and mouse cells. They further confirm that selenium can have a slight stimulatory effect on cell growth, and a much larger inhibitory effect, depending on its concentration.     

Studies of alpha-protein in human cell cultures

alpha-Protein growth fraction (AGF) eliminates the 60- to 90-day adaptive phase required to establish actively growing cultures of HeLa (Gey), human heart (Girardi), KB (Eagle) and other established cell lines in serum-free chemically defined medium A3. AGF is effective at less than 0.4 microgram per ml. By using the procedures described in the text, it is possible to culture HeLa cells in very simple media such as Eagle's basal medium. The properties of AGF are such that it may be adsorbed on glass or plastic flasks. Glass flasks treated with AGF retain full activity after washing with acetone, and treatment with ethyl ether and chemically defined medium. Adsorbed AGF is destroyed by trypsin. AGF can detoxify protamines, polylysines or histones. It will reverse the aggregation response induced by adding complexes composed of carboxymethylcellulose (CMC) and basic proteins. The results support the contention that highly adsorptive AGF functions at the cell surface and is capable of modifying the response of the cell to its environment.     

Effect of iron on the virulence of Trichomonas vaginalis

The role of iron was evaluated with respect to the virulence of Trichomonas vaginalis in mice. Iron-supplemented and iron-depleted Diamond's trypticase-yeast extract-maltose (TYM) media were prepared by adding 360 microM of ferrous sulfate and 100 microM of 2,2'-dipyridyl. Trophozoites cultivated from normal TYM and iron-supplemented TYM media produced subcutaneous abscesses; however, trichomonads grown in an iron-deficient TYM medium failed to produce any pathology. In addition to the increased virulence of trophozoites in mice, iron affects the level of adherence and the cytotoxicity of trichomonads to HeLa cells, which are significantly reduced in trophozoites grown in iron-deficient medium. In conclusion, it is suggested that under iron-depleted conditions such as that induced by 2,2'-dipyridyl the virulence of T. vaginalis is reduced.     

The sensitivity to growth factors of NIH 3T3 cells transformed by the v-myc oncogene

Transformation of NIH 3T3 cells, induced by v-myc oncogene, activates a proliferative potential of the cells cultivated in the serum-free medium, and reduces the ratio of 3H-Tdr incorporation into the cells grown in the presence of 10% fetal serum in comparison to those grown in the serum-free medium. The v-myc transformed cells (NIH 3T3-v-myc) as well as the untransformed ones are very responsive to insulin. On the other hand, the epidermal growth factor, able to stimulate proliferation of NIH 3T3 cells, exert no effects on the NIH 3T3-v-myc cells. The NIH 3T3-v-myc cells cultivated in the medium, containing 2.5% human plasma enriched with thrombocytes, have the same proliferative characteristics as cells grown in the thrombocyte-free plasma. It is concluded that transformation of NIH 3T3 cells induced by v-myc oncogene may reduce a requirement for thrombocyte-released growth factors and EGF but not for insulin.     

Related effects of cell adaptation to serum-free conditions on murine EPO production and glycosylation by CHO cells

The necessity to perform serum-free cultures to produce recombinant glycoproteins generally requires an adaptation procedure of the cell line to new environmental conditions, which may therefore induce quantitative and qualitative effects on the product, particularly on its glycosylation. In previous studies, desialylation of EPO produced by CHO cells was shown to be dependent on the presence of serum in the medium. In this paper, to discriminate between the effects of the adaptation procedure to serum-free medium and the effects of the absence of serum on EPO production and glycosylation, adapted and non-adapted CHO cells were grown in serum-free and serum-containing media. The main kinetics of CHO cells were determined over batch processes as well as the glycosylation patterns of produced EPO by HPCE-LIF. A reversible decrease in EPO production was observed when cells were adapted to SFX-CHO^TM medium, as the same cells partially recovered their production capacity when cultivated in serum-containing medium or in the enriched SFM^TM serum-free medium. More interestingly, EPO desialylation that was not observed in both serum-free media was restored if the serum-independent cells were recultured in presence of serum. In the same way, while the serum-independent cells did not release a sialidase activity in both serum-free media, a significant activity was recovered when serum was added. In fact, the cell adaptation process to serum-free conditions did not specifically affect the sialidase release and the cellular mechanism of protein desialylation, which appeared to be mainly related to the presence of serum for both adapted and non-adapted cells.     

Adenovirus-dependent changes in cell membrane permeability: role of Na+, K+-ATPase.

Adenovirus-dependent release of choline phosphate from KB cells at pH 6.0 was partially blocked by ouabain. In K+-containing medium, maximum inhibition of release was obtained by 10(-5) M ouabain and half-maximal inhibition was achieved by about 0.5 X 10(-6)M ouabain. Ouabain did not block either the binding or the uptake of adenovirus by KB cells. Without K+, about 25% of cell-associated choline phosphate was released by adenovirus, whereas with 1 mM K+ about 50% was released. This activation by K+ was blocked by 0.1 mM ouabain. HeLa cells behaved like KB cells, but a mutant of HeLa cells resistant to ouabain (D98-OR) released much lower amounts of choline phosphate in response to human adenovirus type 2 (Ad2). Wild-type D98-OR cells bound nearly the same amount of adenovirus as did normal HeLa cells. Ad2 also increased the activity of Na+,K+-ATPase in KB cells, with maximum activation at 50 micrograms of Ad2 per ml. In D98-OR cells, Ad2 failed to activate Na+,K+-ATPase activity. Ad2-dependent lysis of endocytic vesicles (receptosomes) was assayed by measuring Ad2-dependent enhancement of epidermal growth factor-Pseudomonas exotoxin toxicity. This action of adenovirus was increased when K+ was present in the medium. Under the conditions used, K+ had no effect on the amount of Ad2 or epidermal growth factor taken up by the cells. On the basis of these results, it is suggested that Ad2-dependent cellular efflux of choline phosphate and adenovirus-dependent lysis of receptosomes may require Na+,K+-ATPase activity.     

Extracellular cAMP formation from host cell ATP by Bordetella pertussis adenylate cyclase

The effect of exogenously added adenylate cyclase from Bordetella pertussis (strain 114) has been investigated in Y-1 mouse adrenal tumor, chinese hamster ovary (CHO) and several other cells. A partially purified adenylate cyclase was found not to enter cells but, nevertheless, produced large amounts of cAMP in the medium. We could show that this resulted from release of ATP (and not larger molecules). The ATP released by the cells could be (1) directly measured and was replenished after each change of medium; (2) was reciprocally related to the cAMP produced; and (3) was competed for by ATPases present in added serum or by hexokinase and, less effectively, by exoenzymes on the cell surface. The extent of ATP leakage varied widely between different cell lines, being marked in CHO and Y-1 adrenal cells but negligible in transformed lymphocyte lines. The uncertainty of the origin of cAMP found in media of cultured cells requires separate analysis of cell and medium cAMP and an assessment of ATP leakage.     

The growth requirements of SV40 virus transformed Balb/c-3T3 cells in serum-free monolayer culture

The growth requirements of SV40 transformed Balb/c-3T3 cells have been studied in the absence of serum. For growth in serum-free medium, the cells require (i) insulin, (ii) transferrin, and (iii) cis-unsaturated fatty acids added in combination with fatty acid free bovine serum albumin. The growth rate, saturation density, and morphology of cells grown in this serum-free medium are the same as those of cells grown in serum supplemented medium. This mixture also supports the growth of SV40 transformed Swiss-3T3 cells and SV40 transformed primary mouse embryo cells, but does not support the growth of untransformed Balb/c-3T3 cells. The addition of fibronectin to this mixture allows routine subculture, repeated passage, and indefinite propagation of SV40 transformed Balb/c-3T3 cells. Cells grown in this medium for a period of two months retain their ability to induce tumors when injected into athymic nude mice.