Novel p53 splice site mutations in three families with Li-Fraumeni syndrome.

Germline mutations in the p53 tumor suppressor gene predispose to a variety of cancers in families with Li-Fraumeni syndrome. Most germline p53 mutations observed to date cause amino acid substitutions in the protein's central sequence-specific DNA binding domain. Outside this conserved core region, however, we found novel alterations in sequences that regulate precursor mRNA splicing in three Li-Fraumeni syndrome families. Two splice site mutations affected the consensus sequence at the splice donor sites of introns 1 and 9, and produced unstable variant transcripts in normal cells. A third mutation at the splice acceptor site of intron 9 generated splicing at a cryptic acceptor site in intron 9. These splice site alterations emphasize the need to examine both noncoding and untranslated regions of the p53 gene for germline mutations in Li-Fraumeni syndrome families. Oncogene (2000) 19, 4230 - 4235

Human keratinocytes that express hTERT and also bypass a p16(INK4a)-enforced mechanism that limits life span become immortal yet retain normal growth and differentiation characteristics.

Normal human cells exhibit a limited replicative life span in culture, eventually arresting growth by a process termed senescence. Progressive telomere shortening appears to trigger senescence in normal human fibroblasts and retinal pigment epithelial cells, as ectopic expression of the telomerase catalytic subunit, hTERT, immortalizes these cell types directly. Telomerase expression alone is insufficient to enable certain other cell types to evade senescence, however. Such cells, including keratinocytes and mammary epithelial cells, appear to require loss of the pRB/p16(INK4a) cell cycle control mechanism in addition to hTERT expression to achieve immortality. To investigate the relationships among telomerase activity, cell cycle control, senescence, and differentiation, we expressed hTERT in two epithelial cell types, keratinocytes and mesothelial cells, and determined the effect on proliferation potential and on the function of cell-type-specific growth control and differentiation systems. Ectopic hTERT expression immortalized normal mesothelial cells and a premalignant, p16(INK4a)-negative keratinocyte line. In contrast, when four keratinocyte strains cultured from normal tissue were transduced to express hTERT, they were incompletely rescued from senescence. After reaching the population doubling limit of their parent cell strains, hTERT(+) keratinocytes entered a slow growth phase of indefinite length, from which rare, rapidly dividing immortal cells emerged. These immortal cell lines frequently had sustained deletions of the CDK2NA/INK4A locus or otherwise were deficient in p16(INK4a) expression. They nevertheless typically retained other keratinocyte growth controls and differentiated normally in culture and in xenografts. Thus, keratinocyte replicative potential is limited by a p16(INK4a)-dependent mechanism, the activation of which can occur independent of telomere length. Abrogation of this mechanism together with telomerase expression immortalizes keratinocytes without affecting other major growth control or differentiation systems.

Human sweat gland myoepithelial cells express a unique set of cytokeratins and reveal the potential for alternative epithelial and mesenchymal differentiation states in culture.

We have characterized precisely the cytokeratin expression pattern of sweat gland myoepithelial cells and have identified conditions for propagating this cell type and modulating its differentiation in culture. Rare, unstratified epithelioid colonies were identified in cultures initiated from several specimens of full-thickness human skin. These cells divided rapidly in medium containing serum, epidermal growth factor (EGF), and hydrocortisone, and maintained a closely packed, epithelioid morphology when co-cultured with 3T3 feeder cells. Immunocytochemical and immunoblot analysis disclosed that the cells differed from keratinocytes in that they were E-cadherin-negative, vimentin-positive, and expressed an unusual set of cytokeratins, K5, K7, K14, and K17. When subcultured without feeder cells, they converted reversibly to a spindle morphology and ceased K5 and K14 expression. Under these conditions, EGF deprivation induced flattening, growth arrest, and expression of alpha-smooth muscle actin ((&agr;)-sma). Coexpression of keratins and alpha-sma is a hallmark of myoepithelial cells, a constituent of secretory glands. Immunostaining of skin sections revealed that only sweat gland myoepithelial cells expressed the same pattern of keratins and alpha-sma and lack of E-cadherin as the cell type we had cultured. Interestingly, our immunocytochemical analysis of ndk, a skin-derived cell line of uncertain identity, suggests that this line is of myoepithelial origin. Earlier immunohistochemical studies by others had found myoepithelial cells to be K7-negative. We tested five K7-specific antibodies that can recognize this protein in western blots and in the assembled keratin filaments of mesothelial cells. Three of these antibodies did not recognize the K7 present in myoepithelial cell filaments or in HeLa cell filaments, indicating that some K7 epitopes are masked when K7 pairs with K17 instead of with its usual keratin filament partner, K19.

Generation of papillomavirus-immortalized cell lines from normal human ectocervical, endocervical, and vaginal epithelium that maintain expression of tissue-specific differentiation proteins.

We have established and characterized three cell lines from normal human vaginal, ectocervical, and endocervical epithelia immortalized by expression of human papillomavirus 16/E6E7. The lines (VK2/E6E7, Ect1/E6E7, and End1/E6E7) displayed distinctive morphologies at the level of light microscopy when cultured in calcium-supplemented (0.4 mM) keratinocyte serum-free medium and maintained a stable phenotype after more than 1 yr of continuous passage. They were compared to primary cell cultures and epithelial cells in sections of the respective native tissues for expression of epithelial differentiation proteins. All cell lines expressed cytokeratin (CK) 8, CK18, and CK19, and some cells in all three cell lines expressed CK16, involucrin, and the secretory component of the polymeric immunoglobulin receptor. The vaginal and ectocervical cell lines expressed CK10 and CK13, whereas the endocervical line did not. With the exception of CK8 and CK18 expression, the morphological and immunocytochemical characteristics of the immortalized lines closely resembled those of their respective tissues of origin and primary cultures, and all differed significantly from the HeLa cervical adenocarcinoma cell line. These new cell lines may provide the basis for valid, reproducible in vitro models for studies on cervicovaginal physiology and infections and for testing pharmacological agents for intravaginal application.

Intraperitoneal injection of genetically modified, human mesothelial cells for systemic gene therapy.

An ideal cell type for ex vivo gene therapy should be easy to biopsy, propagate, and genetically engineer in culture, should be transplantable using simple procedures, and should express therapeutic proteins at useful levels. The mesothelial cell appears to satisfy these criteria. Several thousand proliferative mesothelial cells were present in typical specimens of nonpathologic human peritoneal fluid obtained by needle aspiration. These divided rapidly in a specialized medium to yield pure cultures of approximately 10(7) cells within 2 weeks. The replicative lifespan of mesothelial cells cultured from adults was approximately 42-52 population doublings, permitting expansion and cryopreservation of a lifetime supply of autologous cells from one fluid sample. Cells transduced with a human growth hormone (hGH) adenoviral vector secreted 100-300 microg of hGH/10(6) cells per day for at least 6 weeks in culture when maintained at quiescence. Intraperitoneal injection of transduced cells into athymic mice resulted in rapid systemic delivery of hGH, with peak plasma levels of 0.1-1 microg/ml declining over 3 weeks to <1 ng/ml. Mice receiving a second injection of engineered cells displayed the same plasma hGH levels and duration as naive mice. Cells labeled with a beta-galactosidase vector were identifiable by in situ enzymatic staining as clusters attached to peritoneal surfaces at multiple sites for at least 19 days after injection. Cells serially passaged through about three-quarters of their lifespan before transduction and injection were as effective at hGH delivery as earlier-passage cells. These results indicate the clinical potential for ex vivo gene therapy using mesothelial cells.

Hyaluronic acid capsule and the role of streptococcal entry into keratinocytes in invasive skin infection.

It has been suggested that entry of pathogenic bacteria, including streptococci, into epithelial cells may represent an early stage of invasive infections. We found that poorly encapsulated wild-type strains and unencapsulated mutants of group A Streptococcus entered cultured human keratinocytes with high efficiency, while strains that produced large amounts of hyaluronic acid capsule did not, regardless of M-protein type or clinical source of the isolate. However, encapsulated streptococci produced extensive local necrosis and systemic infection in a mouse model of skin infection, while an isogenic acapsular strain did not. The results implicate the hyaluronic acid capsule as a virulence factor in soft tissue infection. Entry of poorly encapsulated group A Streptococcus into human epithelial cells does not appear to represent an initial step in invasive disease; rather, the capacity of encapsulated strains to avoid uptake by epithelial cells is associated with enhanced virulence in skin and soft tissue infection.

In vitro propagation of human ocular surface epithelial cells for transplantation.

PURPOSE: To examine the possibility that ocular surface epithelial cells might be grown in culture for use as grafts. METHODS: The proliferative capacity of epithelial cells cultured from the conjunctiva, limbus, and central cornea of normal human eyes was compared. Single cells disaggregated from approximately 1 mm2 biopsy specimens were serially cocultured with lethally irradiated mouse 3T3 fibroblasts. To study the cells' ability to reform a stratified epithelium, confluent limbal cultures were released as an intact cell sheet with the enzyme Dispase and transplanted to a dermal connective tissue bed in nude mice. Attachment and differentiation properties of the reconstituted epithelium were examined immunohistochemically. RESULTS: Central corneal epithelial cells could not be propagated; they senesced in first or second passage. In contrast, limbal epithelial cells exhibited a substantial (i.e., mean of 23 population doublings) and conjunctival cells a moderate (i.e., mean of 11 population doublings) proliferative capacity. Within 4 days of transplantation to the nude mouse dermis, cultured limbal epithelial cells formed an epithelium 5-6 cell layers thick. The epithelium adhered firmly to the graft bed, and deposition of the basement membrane and anchoring fibril protein collagens IV and VII and laminin was detectable immunohistochemically. The transplanted epithelium displayed limbuslike compartmental expression of keratins K3, K13, and K19, and of the enzyme enolase. CONCLUSIONS: These results support the concept that corneal epithelial stem cells are located in the limbus and indicate that cultured autologous limbal cells may function as grafts to permanently restore the corneal epithelium after severe ocular surface injury.

Abnormal expression of retinoic acid receptors and keratin 19 by human oral and epidermal squamous cell carcinoma cell lines.

We have analyzed the expression of the three retinoic acid receptor (RAR) (alpha, beta, gamma) mRNAs and the intermediate filament protein keratin 19 (K19) mRNA in cell lines cultured from oral and epidermal human squamous cell carcinoma (SCC) and from benign, hyperplastic, and hyperkeratotic (leukoplakia) lesions arising in various regions of the oral cavity. Seven of the SCC lines were derived from tumors arising in regions of the oral cavity in which the normal epithelial cells (keratinocytes) express RAR beta transcripts. Seven of the nine SCC lines tested did not exhibit detectable RAR beta mRNA levels, even in response to addition of retinoic acid (RA). The RAR beta gene did not appear to be rearranged or deleted in the five nonexpressing SCC lines examined by Southern analysis. The steady-state RAR gamma mRNA levels were 2- to 4-fold lower in 6 of the 9 SCC lines than in their normal counterparts, whereas the RAR alpha message levels in SCC lines were similar to those of the normal cell strains. The expression of keratin 19 message, which is RA inducible in normal keratinocytes, was also abnormal in many of the SCC cell lines. Some SCC lines, e.g., those derived form tumors of the soft palate epithelium, did not express high levels of K19 message even though normal soft palate keratinocytes expressed high levels of K19 mRNA. Two of the nine SCC lines expressed higher than normal levels of K19 mRNA, and this expression was RA independent. Cells cultured from four oral leukoplakia lesions were also examined and found to express RAR beta mRNA at relatively normal levels, but they expressed RAR gamma message at half the level of epithelial cells cultured from normal tissue. These results show that the correlation between RAR beta gene expression and K19 gene expression that we have observed in the various normal keratinocyte subtypes of the oral cavity (D.L. Crowe et al., manuscript in preparation) is not present in transformed keratinocytes (SCC cells). The lack of apparent RA regulation of the K19 gene in SCC lines may be associated with other aberrations in differentiation which have been identified in SCC cells. Abnormally low expression of the RAR beta receptor may contribute to neoplastic progression in stratified squamous epithelia. It may also determine whether a tumor is responsive to RA as a chemotherapeutic agent.

Three distinct keratinocyte subtypes identified in human oral epithelium by their patterns of keratin expression in culture and in xenografts.

We have characterized the cells that form the human oral epithelia by analyzing their patterns of keratin expression in culture and in transplants. Keratinocytes of all oral regions synthesized high levels of keratins K5/K14 and K6/K16,K17, as expressed by cells of all stratified squamous epithelia in culture. However, cells from different regions varied in their expression in culture of retinoid-inducible (K19 and K13) and simple epithelial (K7, K8 and K18) keratins. By these criteria, all oral cells could be classified as belonging to one of three intrinsically distinct subtypes: "keratinizing" (gingiva, hard palate), "typical nonkeratinizing" (inner cheek, floor of mouth, ventral tongue) and "special non-keratinizing" (soft palate), all of which differed from the epidermal keratinocyte subtype. Cells from fetal floor of mouth expressed a pattern of keratins in culture markedly different from that of adult floor of mouth cells but identical to that of the adult "special nonkeratinizing" subtype and similar to that of several oral squamous cell carcinoma lines. When cultures of oral keratinocytes were grafted to the dermis of nude mice, they formed stratified epithelial structures after 10 days. In some areas of the stratified structures, the basal layer recapitulated the K19 expression pattern of the oral region from which they had originated. Thus, regional differentiation of the oral epithelium is based on an intrinsic specialization of regional keratinocyte stem cells. Additionally, oral cell transformation either frequently involves reversion to the fetal keratin program or else oral cells that express this keratin program are especially susceptible to transformation.

Expression of ras oncogenes in cultured human cells alters the transcriptional and posttranscriptional regulation of cytokine genes.

Autonomous production of cytokines such as the hematopoietic colony-stimulating factors (CSFs), IL-1, or IL-6 has been demonstrated in numerous human and murine neoplasms, and may be involved in the pathogenesis of several paraneoplastic syndromes such as leukocytosis, fever, and hypercalcemia. Because of the high frequency with which mutations in ras protooncogenes have been detected in human tumors, as well as evidence linking ras gene products to activation of certain cellular functions, we investigated whether ras mutations might influence the regulation of cytokine genes. Normal human fibroblasts transfected with a mutant val12 H-ras oncogene expressed increased levels of mRNA transcripts encoding granulocyte-CSF (G-CSF), granulocyte-macrophage-CSF (GM-CSF), and IL-1 beta compared with controls. Human mesothelioma cells transfected with a mutant asp12 N-ras oncogene exhibited similar alterations in cytokine gene expression. Estimates of transcriptional activity by nuclear run-on analysis revealed a selective increase in transcription only for the IL-1 gene. Analysis of mRNA half-life demonstrated a marked increase in the stability of numerous cytokine transcripts, including G-CSF, GM-CSF, IL-1, and IL-6. The addition of anti-IL-1 neutralizing antibody to cultures of cells expressing ras mutants did not block the expression of any of the cytokines examined, suggesting that the baseline expression of GM-CSF, G-CSF, and IL-6 was not a secondary event due to the increased transcription of IL-1. These results indicate that mutations in ras genes may alter expression of several cytokine genes through both transcriptional and posttranscriptional mechanisms.

Specific effects of ras oncogene expression on the growth and histogenesis of human epidermal keratinocytes.

Little progress has been made in identifying specific regulatory pathways that might be affected in cells by a mutationally activated p21ras when its expression does not lead to complete transformation. We wished to determine whether a normal, diploid human epithelial cell in which activation of ras had occurred could be identified in culture and, furthermore, whether expression of a mutant p21ras in such an otherwise normal cell would result in abnormal histogenic behavior in vivo. Thus, we introduced the v-Ha-ras gene into an early passage culture of normal human epidermal keratinocytes via a defective retrovirus. We examined these genetically engineered cells for changes in growth and differentiation, both in culture and in the epithelium formed when cultures were grafted to the skin of nude mice. We have found that keratinocytes expressing p21v-ras are independent of epidermal growth factor (EGF)--a factor which is normally essential for progressive colony growth, but that they are otherwise indistinguishable in culture from normal cells. v-ras keratinocytes also secrete a factor possessing some specific biological activities of members of the fibroblast growth factor (FGF) family, but which is distinct from acidic and basic FGF. In short-term dermal grafts the v-ras cells form a non-invasive and normally differentiating epidermis. However, the cells express elevated levels of keratin 19, which is a characteristic of fetal epidermis and of premalignant lesions of some stratified squamous epithelia.

Cloned cDNA sequence for the human mesothelial protein 'mesosecrin' discloses its identity as a plasminogen activator inhibitor (PAI-1) and a recent evolutionary change in transcript processing.

Mesosecrin, a Mr approximately 46 x 10(3) glycoprotein secreted in abundance by human mesothelial cells in culture, was recently described by this laboratory. We isolated partial cDNA clones for mesosecrin from a human mesothelial cell cDNA library in lambda gt11 using a specific antiserum. Comparison of mesosecrin cDNA sequences with the recently published sequence for plasminogen activator inhibitor-1 (PAI-1) cloned from cDNA libraries of endothelial and other cell types revealed that mesosecrin and PAI-1 are the same protein. Reverse fibrin autography of electrophoretically fractionated medium from mesothelial cell cultures confirmed that mesosecrin is functional as a plasminogen activator inhibitor. The mesosecrin/PAI-1 cDNA clones hybridized to abundant 3.6 and 2.6 kb (kb = 10(3) bases) mRNAs on Northern blots of cultured human mesothelial cell and endothelial cell RNA. These mRNA sizes correspond to those recently published for human endothelial and fibrosarcoma PAI-1 mRNA, which most likely result from alternate polyadenylation sites. Messages 3.6 and 2.6 kb long were also detected in cells cultured from orangutans and African green monkeys, but only an approximately 3.6 kb mRNA was detected in cells of lower primates and several other mammalian species. Thus the extra polyadenylation site in the PAI-1 gene, responsible for the shorter form of the RNA, apparently has been acquired recently during primate evolution. Because they are more easily propagated in culture than endothelial cells, human mesothelial cells offer a new and advantageous system for PAI-1 production and study of its regulation and function.

Expression of colony-stimulating factor genes by normal human mesothelial cells and human malignant mesothelioma cells lines in vitro.

We investigated normal human mesothelial cells and human malignant mesothelioma cell lines for the ability to produce hematopoietic colony-stimulating factors (CSFs) in culture. Early passage cultures of normal diploid human mesothelial cells spontaneously expressed detectable levels of M-CSF mRNA transcripts, but lacked detectable transcripts for GM-CSF or G-CSF. Exposure of normal mesothelial cells to epidermal growth factor (EGF), lipopolysaccharide (LPS), or tumor necrosis factor (TNF) induced expression of G-CSF mRNA. The combination of EGF and TNF induced threefold more G-CSF transcripts than did either factor alone. GM-CSF transcripts were induced only by the combination of TNF and EGF. Interleukin-1 beta (IL-1 beta) transcripts were induced by EGF, TNF, or LPS and were inhibited by hydrocortisone (HC). All malignant mesothelioma cell lines tested also spontaneously expressed M-CSF transcripts. However, in contrast to normal mesothelial cells, two of four malignant mesothelioma cell lines also autonomously expressed G-CSF and GM-CSF transcripts without TNF, EGF, or LPS stimulation. Secretion of biologically active CSFs was confirmed by testing media conditioned by the various cell types examined. The detection of biologically active CSFs correlated well with the presence of detectable CSF transcripts by Northern analysis. These data indicate that (a) normal human mesothelial cells spontaneously express detectable levels of M-CSF mRNA in culture; (b) EGF is an essential cofactor for optimal induction of G-CSF and GM-CSF expression; (c) exposure of normal mesothelial cells to inflammatory mediators such as LPS and TNF increases the levels of transcripts for CSFs and IL-1 beta; and (d) as compared with normal human mesothelial cells, some cell lines of human malignant mesothelioma exhibit aberrant gene expression for multiple cytokines, including G-CSF, GM-CSF, IL-1 beta, and IL-6.

Environment-dependent growth inhibition of human epidermal keratinocytes by recombinant human transforming growth factor-beta.

Transforming growth factor-beta (TGF-beta) purified from platelets is a potent growth inhibitor of several normal epithelial cell types in culture. In contrast, some carcinoma cell lines derived from tumors of these same tissues are resistant to this factor. Using recombinant human TGF-beta, the authors have confirmed these results with six normal human epidermal keratinocyte strains and four human epidermal squamous carcinoma cell lines. However, the sensitivity of normal cells to TGF-beta was found to depend on the culture conditions. When grown in a specialized nutrient medium supplemented with pituitary extract, keratinocytes were completely inhibited by the addition of 0.3 ng/ml TGF-beta. In contrast, when their growth was supported by cocultivation with 3T3 fibroblast feeder cells, 30- to 100-fold higher concentrations of TGF-beta were required to achieve comparable growth inhibition. This differential sensitivity occurred despite the fact that in both culture systems TGF-beta in the culture medium had a half-life of about 50 minutes, becoming tightly bound to the surface of the culture dish. Bound TGF-beta proved to be biologically active and stable for about a week in the absence of 3T3 feeder cells. Incubating 3T3 cells on TGF-beta-coated dishes, however, resulted in nearly quantitative removal and degradation of the TGF-beta within 2 days, permitting normal rates of keratinocyte growth. The binding of TGF-beta to surfaces and the ability of fibroblasts to attenuate its inhibitory activity for epithelial cells must be considered when evaluating in vitro models and in planning strategies for the use of this factor in vivo.

Suprabasal 40 kd keratin (K19) expression as an immunohistologic marker of premalignancy in oral epithelium.

The authors have studied the expression of keratin 19 in normal oral mucosa and in oral lesions exhibiting a range of histopathologic changes that are thought to precede squamous cell carcinoma. Formalin-fixed, paraffin-embedded sections were pretreated with pronase and stained with a K19-specific antibody by the avidin-biotin immunoperoxidase method. In nonkeratinized mucosa, whether normal or benign hyperplastic, K19 was detectable in the basal cell layer. In keratinized mucosa, whether normal or benign hyperplastic, there was no detectable K19. All lesions from any oral site that exhibited atypia diagnosed from hematoxylin and eosin stained sections as moderate-to-severe dysplasia or carcinoma in situ, whether hyperkeratotic or not, stained strongly for K19 in the basal and suprabasal cell layers. The number of cell layers that were K19-positive correlated with the level in the epithelium to which dysplasia persisted. Suprabasal K19 staining tended to occur in regions of the epithelium in which expression of the terminal differentiation protein involucrin was delayed or absent. Thus, K19 expression may be linked to the retention of stem cell character or a state otherwise uncommitted to terminal squamous differentiation. Suprabasal K19 staining is clearly correlated with premalignant change in oral epithelium and therefore promises to be a useful tool in oral histopathologic diagnosis.

Normal human mesothelial cells and fibroblasts transfected with the EJras oncogene become EGF-independent, but are not malignantly transformed.

The nature of the lesion in growth control exerted by the cancer-derived c-H-ras mutation, EJ-ras, and its transforming potential in diploid cells are both poorly understood. We introduced EJ-ras into normal, diploid human mesothelial cells and fibroblasts and obtained transfectants expressing p21EJ-ras. All clones examined were independent of EGF for rapid growth, and all secreted an EGF-like mitogen into the medium at levels sufficient to satisfy the EGF requirement of normal cells. The EJ-ras transfectants were not altered with respect to any other growth requirement, and they were not transformed. Eleven clones tested all retained a finite replicative lifespan which, in most cases, was the same as that of the parent cell strain. Three transfectants tested were not tumorigenic in nude mice. Thus p21EJ-ras can circumvent an important mitogenic signal pathway in human cells. Nevertheless, neither the secretion of an autocrine growth factor nor any other effect of p21EJ-ras serves to malignantly transform normal human cells, in contrast to the susceptibility of some established rodent cell lines to transformation by these mechanisms.

Anchorage-independent growth of normal human mesothelial cells: a sensitive bioassay for EGF which discloses the absence of this factor in fetal calf serum.

This laboratory recently reported that normal human mesothelial cells require epidermal growth factor (EGF) and hydrocortisone (HC), in addition to fetal calf serum and a complex defined medium component, in order to grow optimally in surface culture. We report here that this normal cell type also forms large colonies at high efficiency in semi-solid medium, but exhibits more stringent serum and EGF requirements for anchorage-independent than for surface growth. Mesothelial cells are unable to divide at all in semi-solid medium without added EGF or with less than 2% serum, whereas they grow slowly but progressively in surface culture under such conditions. In semi-solid medium containing 20% serum and HC, mesothelial cells are stimulated to divide by the addition of as little as 30 pg/ml purified EGF. Human urine or male mouse plasma could substitute for purified EGF, yielding growth commensurate with the levels of EGF in these biological fluids previously measured by others using radioreceptor and radioimmune assays. Thus growth of mesothelial cells in semi-solid medium can serve as a highly sensitive assay of EGF biological activity which is unaffected by the presence of serum proteins. In addition, our results demonstrate that fetal calf serum does not provide mitogenic levels of EGF to cultured cells, raising the question of the identity of plasma and serum mitogens.

Polycyclic aromatic hydrocarbon mutagenesis of human epidermal keratinocytes in culture.

We have developed a culture system for detecting and isolating rare hypoxanthine phosphoribosyltransferase-deficient mutants of human epidermal keratinocytes. A thioguanine-resistant variant, 3T3M1, of the Swiss mouse fibroblast line 3T3 was used as a feeder layer to support clonal growth of mutant keratinocytes. A near-diploid, epidermal squamous cell carcinoma line, SCC-13Y, was used as a prototype to determine mutagen treatment conditions, plating density, and phenotypic expression time for maximum mutant recovery. To extend this system to normal keratinocytes, we improved the culture conditions by adding insulin, adenine, and Ham's nutrient mixture F-12, which increased colony-forming efficiencies to 30% in early passage and made feasible the detection of rare mutants in normal epidermal keratinocyte populations. We have quantitated mutation in SCC-13Y and three strains of normal human epidermal keratinocytes after exposure to polycyclic aromatic hydrocarbons, which are activated to their mutagenic forms by cellular mixed-function oxidases. 7,12-Dimethylbenz[a]anthracene and benzo[a]pyrene caused almost no cytotoxicity, but induced thioguanine-resistant mutants at frequencies as much as 50-fold higher than the spontaneous frequency of approximately 10(-6). The mutants were aminopterin-sensitive and possessed no measurable hypoxanthine phosphoribosyltransferase activity; their behavior was indistinguishable from that of keratinocytes cultured from individuals with Lesch-Nyhan syndrome. This mutagenesis assay system should also be applicable to other feeder layer-dependent human epithelial cell types, such as urothelial, mammary, and tracheal epithelial cells.