Regression of established murine carcinoma metastases following vaccination with tumour-associated antigen peptides.

The cure of micrometastases following surgery is the major goal of cancer immunotherapy. We have recently isolated tumour-associated antigen (TAA) peptides, MUT 1 and MUT 2, derived from a mutated connexin 37 gap-junction protein, from the malignant 3LL-D122 murine lung carcinoma. We now report that synthetic MUT 1 or MUT 2 induces effective antitumour cytoxic T lymphocytes. Peptide vaccines protect mice from spontaneous metastases of 3LL-D122 tumours. Moreover, peptide vaccines reduce metastatic loads in mice carrying pre-established micrometastases. Tumour-specific immunity was primarily mediated by CD8+ T cells. This is the first evidence that peptide therapy may be effective in treatment of residual tumours and provides a rationale for the development of peptide vaccines as a modality for cancer therapy.

KBF1 (p50 NF-kappa B homodimer) acts as a repressor of H-2Kb gene expression in metastatic tumor cells.

Downregulation of major histocompatibility complex class I expression is causally related to high malignancy and low immunogenicity of certain murine tumors. In this study, we have analyzed the roles of the nuclear factors KBF1/p50 and p65 in regulation of class I expression in high and low metastatic tumor cells. Low class I-expressing cells show at higher levels of KBF1/p50 and NF-kappa B (p50/p65) binding activity than high class I-expressing cells. However, an excess of KBF1 over NF-kappa B is observed in low expressing cells, while an excess of NF-kappa B over KBF1 is observed in high expressing cells. Stable transfection of a p65 expression vector into low class I-expressing cells activated H-2 transcription and cell surface expression, while stable transfection of p50 expression vector into high expressing cells suppressed H-2Kb transcription and cell surface expression. Our studies suggest that KBF1 has the potential of downregulating class I gene expression, whereas dimers containing the p65 subunit are activators of class I gene expression.

Biochemical studies of the excitable membrane of paramecium tetraurelia. IX. Antibodies against ciliary membrane proteins.

The excitable ciliary membrane of Paramecium regulates the direction of the ciliary beat, and thereby the swimming behavior of this organism. One approach to the problem of identifying the molecular components of the excitable membrane is to use antibodies as probes of function. We produced rabbit antisera against isolated ciliary membranes and against partially purified immobilization antigens derived from three serotypes (A, B, and H), and used these antisera as reagents to explore the role of specific membrane proteins in the immobilization reaction and in behavior. The immobilization characteristics and serotype cross-reactivities of the antisera were examined. We identified the antigens recognized by these sera using immunodiffusion and immunoprecipitation with 35S-labeled ciliary membranes. The major antigen recognized in homologous combinations of antigen-antiserum is the immobilization antigen (i-antigen), approximately 250,000 mol wt. Several secondary antigens, including a family of polypeptides of 42,000-45,000 mol wt, are common to the membranes of serotypes A, B, and H, and antibodies against these secondary antigens can apparently immobilize cells. This characterization of antiserum specificity has provided the basis for our studies on the effects of the antibodies on electrophysiological properties of cells and electron microscopic localization studies, which are reported in the accompanying paper. We have also used these antibodies to study the mechanism of cell immobilization by antibodies against the i-antigen. Monovalent fragments (Fab) against purified i-antigens bound to, but did not immobilize, living cells. Subsequent addition of goat anti-Fab antibodies caused immediate immobilization, presumably by cross-linking Fab fragments already bound to the surface. We conclude that antigen-antibody interaction per se is not sufficient for immobilization, and that antibody bivalency, which allows antigen cross-linking, is essential.

Proteolytic enzymes in tumor metastasis. I. Plasminogen activator in clones of Lewis lung carcinoma and T10 sarcoma.

Plasminogen activator (PA; urokinase) levels were studied in metastatic and nonmetastatic clones of the Lewis lung carcinoma (3LL) and of the T10 sarcoma. Tests of clones grown in vitro revealed that the cell content and secretion of PA correlated positively with the metastatic capacity of the clones of both tumors. When cell-associated activities were examined in cloned cell populations grown subcutaneously in vivo, the apparent activities in the solid tumors produced by low-metastatic clones were equal to or even higher than those in solid tumors produced by high-metastatic clones. This finding was attributed to the observation that solid tumors produced by low-metastatic clones, but not those produced by high-metastatic clones, were highly infiltrated with macrophages. Subsequent tests indicated that the ip inoculation of X-irradiated or mitomycin-treated tumor cells of low-metastatic clones elicited a significantly greater peritoneal infiltration of macrophages than did tumor cells of high-metastatic clones. Such "tumor-associated" macrophages manifested high levels of PA, whereas resident (nonactivated) peritoneal macrophages did not. These findings suggest that although PA may cause the initial detachment from the local tumor of both nonmetastatic (via the macrophage PA) and metastatic cells (via their own PA), the PA secreted by the metastatic cells may enable them to complete subsequent stages of the metastatic process that may be PA-dependent.

Proteolytic enzymes in tumor metastasis. II. Collagenase type IV activity in subcellular fractions of cloned tumor cell populations.

Collagenase type IV degradation activity was examined in metastatic and nonmetastatic clones of the Lewis lung carcinoma (3LL) and of the T10 sarcoma. Conditioned media prepared from cells of both tumors grown in vitro contained low degradation activities, whereas conditioned media from organ cultures of the same clones grown as solid tumors in animals exhibited higher degradation activities. Analysis of subcellular fractions of tumor cells showed that collagenase type IV activity was localized mainly in the cytoplasmic fraction. Crude homogenates or detergent lysates manifested low degradation activities. Little activity was associated with purified plasma membrane preparations and endoplasmic reticular fractions. However, addition of plasma membrane to conditioned media and to cytoplasmic fractions reduced the degradation activities of the cytoplasmic fractions. Possibly a factor that inhibits collagenase type IV exists in the cells in a vesicular form. No correlation between degradation activity and metastatic capacity was demonstrated in the models used in this study. Both metastatic and nonmetastatic clones of the same tumor similarly could degrade basement membrane components.

T-cell subset analysis of Lewis lung carcinoma tumor rejection: heterogeneity of effectors and evidence for negative regulatory lymphocytes correlating with metastasis.

We have analyzed the phenotypes of the T-cell subsets generated in response to Lewis lung carcinoma clones in C57BL/6J recipients. The metastatic derivative, which expresses low levels of H-2Kb gene, predominantly elicited CD8, V beta 8, and V beta 9+ T-cells. The nonmetastatic clone expressing high levels of H-Kb gene triggered a more heterogeneous response of V beta-5, -6, -8, -9, and -11 CD8+ T-cells. Comparison of the T-cell receptor (TCR) expression of the T-cells infiltrating the tumor site with the lymphocytes in the periphery of tumor-bearing animals revealed a pattern of homing of CD4+ T-cells bearing V beta-5, -6, and -11 TCR chains and CD8+ T-cells bearing V beta-5, -6, -9, and -11. Depletion of V beta 5 or V beta 6+ T-cells correlated with accelerated tumor growth, implying their protective role as tumor-specific effectors and consistent with the cytotoxicity of T-cells with this TCR phenotype. V beta 11 TCR expression in the tumor-infiltrating lymphocytes increased with the tumor size. Depletion of V beta 11+ T-cells enhanced resistance to primary tumor growth and conferred protection from metastasis in recipients cleared of V beta 5 and V beta 6 T-cell subsets. Those results suggest that tumor-specific effectors as well as negative regulator T-cells home, infiltrate, and coexist in the tumor site.

Combined vaccination with major histocompatibility class I and interleukin 2 gene-transduced melanoma cells synergizes the cure of postsurgical established lung metastases.

We have analyzed and compared in detail the malignant phenotypes of, the immune mechanisms induced by, and the immunotherapeutic potentials of B16-F10.9 melanoma cells manipulated by gene transfer to express syngeneic H-2Kb molecules or to secrete the cytokines interleukin 2 (IL-2) or IL-6. Local tumor growth in the footpad of transduced cells is mainly retarded by expression of H-2Kb and IL-2 genes and less by expression of IL-6. Mice given injections intrafootpad of tumorigenic doses of transduced clones manifested significantly reduced postsurgical spontaneous metastasis. After i.v. inoculation, mice given injections of F10.9-Kb expressors did not develop experimental lung metastases; mice given injections of F10.9-IL-6 secretors developed reduced metastatic loads; whereas mice given injections of F10.9-IL-2 secretors developed high loads of lung metastases. On the basis of injections into nude mice, in vivo depletions of CD4+, CD8+, and NK1.1+ cells, and in vitro CTL and natural killer (NK) assays, we show that all F10.9-modified cells induce CD8+ tumor-specific CTL activity and that F10.9-IL-2 secretors also induce nonspecific NK/lymphokine-activated killer cell activity. Vaccinations with F10.9-modified cells were capable of significantly reducing metastatic spread from small established F10.9 footpad tumors. However, in mice carrying preestablished lung metastases, a highly therapeutic effect was achieved only when H-2Kb expressors and IL-2 secretors were combined in vaccination, whereas individual vaccines or other combinations had marginal effects. This higher efficiency of the combined vaccine is due to the combined effect of efficient CTL induction and NK/lymphokine-activated killer cell activity as concluded from depletion of CD8+ and NK1.1 cells during immunotherapy. Thus, the cure of established metastasis can be achieved by the synergistic effects of vaccination with class I and IL-2-transduced tumor cells.

Abolishment of metastasis formation by murine tumor cells transfected with "foreign" H-2K genes.

High metastatic, low immunogenic Lewis lung carcinoma clones express low levels of H-2Kb major histocompatibility complex antigens. These cells metastasize spontaneously in mice with C57BL/6 genetic background possessing the H-2Db locus. Transfection of different H-2K genes abrogates metastasis in H-2K, H-2D compatible mice and in C57BL/6 recipients. The transfected cells are potent inducers of H-2K-restricted and alloreactive cytotoxic lymphocytes that kill H-2K-positive cells and cross-react with parental nontransfected cells.

Expression of HLA-A,B,C antigens on primary and metastatic tumor cell populations of human carcinomas.

The expression of monomorphic determinants of the histocompatibility leukocyte antigens (HLA) class I antigens by human malignant tumor cells was studied in tissue specimens of 70 primary tumor lesions obtained from patients with carcinoma of the breast (41 patients), colon (8 patients), urinary bladder (8 patients), and kidney (13 patients), and in samples of either synchronous or metachronous lymph node, lung, or liver metastases available in 44 of the patients. The frequencies of HLA class I expressor and nonexpressor tumor cells were determined by immunohistochemical staining of histological sections of fresh frozen tissue samples with the W6/32 monoclonal antibody. The tumor cell populations in the majority of the primary lesions consisted predominantly of HLA-immunoreactive cells (observed in 38 of 70 patients; 54%), especially in those patients who did not have clinical evidence of metastatic disease (8 of 11 patients; 73%). Various degrees of loss of reactivity were observed in other primary lesions, although in only 8 (12%) tumors (7 of which were obtained from patients with metastatic disease), the neoplastic cells were nearly exclusively HLA-nonreactive. In contrast, the majority of metastatic lesions consisted of either predominantly HLA-negative cells (33 of 44 specimens; 75%) or mixed populations (10 of 44 specimens; 23%), whereas only one metastatic lesion manifested HLA class I antigen staining in more than 70% of its tumor cells (P = 0.0005). Intravascular clusters of tumor cells consisted predominantly of HLA class I nonexpressors. The observed patterns of distribution of HLA expressors and nonexpressor tumor cells are compatible with the notion that HLA-negative cells in human carcinomas manifest a selective advantage with regard to metastatic progression and growth. The suppressed expression of major histocompatibility complex class I antigens on metastatic cells may lead to failure of presentation of cell surface tumor specific epitopes to host cytotoxic T-lymphocytes. Such a process would enable tumor cells to evade host immune responses and would promote and enhance cell dissemination and metastatic growth.

Interleukin 6 gene transfection into Lewis lung carcinoma tumor cells suppresses the malignant phenotype and confers immunotherapeutic competence against parental metastatic cells.

To investigate the influence of interleukin 6 (IL-6) production on malignancy of tumor cells we transfected cells of the high-metastatic, low-immunogenic D122 clone of the Lewis lung carcinoma with a mammalian expression vector containing the human IL-6 complementary DNA. In vitro, IL-6 positive transfectants showed growth inhibition that was directly correlated with the levels of IL-6 production. The in vitro growth arrest did not seem to be a function of an autocrine system mediated via the secreted human IL-6 acting on the tumor cell surface receptors since neutralizing antibodies to human IL-6 did not prevent the growth inhibition. Neither did exogenous human recombinant IL-6 affect the growth of D122 cells. In vivo, IL-6 positive transfectants showed reduction of tumorigenicity and significant suppression of metastatic competence in syngeneic, immunocompetent mice. In mature T-cell deficient nude mice, the IL-6 transfectants showed some arrest of local growth but no suppression of lung metastasis. It seems therefore that the reduction of metastatic competence of IL-6 transfectants is primarily a function of stimulation by the transfectants of host T-cell immune responses. Immunization with inactivated high-positive IL-6 transfectants induced high levels of anti-tumor cytotoxic T-lymphocytes and protected mice against metastatic growth of a subsequent graft of parental tumor cells. Moreover, reduction of metastatic growth of parental highly metastatic D122 cells was also achieved when immunization of mice was begun after establishment of the primary parental tumors. Thus, inactivated IL-6 transfectants were effective when used as a cellular vaccine for experimental immunotherapy of metastasis.

c-fos and c-jun overexpression in malignant cells reduces their tumorigenic and metastatic potential, and affects their MHC class I gene expression.

Reduced co-expression of the c-fos and c-jun protooncogenes has been correlated with the down regulation of H-2K class I major histocompatibility antigens in high-metastatic cell lines from the Lewis lung carcinoma, B16 melanoma and the K1735 melanoma. Transfection of c-jun and c-fos genes into the high metastatic clones D122 (3LL) and F10.9 (B16 melanoma) resulted in activation of H-2 class I gene expression. D122 transfectants expressing high levels of c-jun and c-fos and F10.9 transfectants expressing high levels of c-fos exhibited markedly reduced tumorigenicity and were of low metastatic potential. In contrast, transfection of junB into the low metastatic, high H-2Kb, Db expressor clone A9 (3LL), reduced MHC class I gene expression, and converted the parental low, into high-metastatic cells. The data demonstrate the involvement of genes from the fos and jun family in regulation of MHC class I expression and consequently in regulation of immunogenicity and metastatic competence of tumor cells.

The c-fos proto-oncogene in murine 3LL carcinoma clones controls the expression of MHC genes.

The c-fos proto-oncogene and H-2K class I major histocompatibility antigens are differentially expressed in low-metastatic Lewis lung carcinoma clones, but not in high-metastatic clones. Interferons induce mRNA expression of fos and H-2 in non-expressor cells and elevate mRNA steady state levels of expressor cells. Transfection of non-expressor cells by v-fos or c-fos genes induces the transcription of H-2K mRNA and elevates the levels of H-2 proteins, but not of other gene products. These results, correlated with observations in other cell systems, suggest that the c-fos proto-oncogene controls the expression of MHC genes coding for class 1 antigens.

Modification of PDGFalpha receptor expression or function alters the metastatic phenotype of 3LL cells.

Functional PDGFalpha receptors are selectively expressed on highly lung-metastasizing clones of the 3LL Lewis lung carcinoma, but not on low-mestastatic clones. The highly metastatic clones are also growth induced in vitro by PDGF and lung conditioned medium. To investigate whether modification of PDGFalpha receptor expression or function can affect metastatic capability, we transfected cells of a low-metastatic 3LL clone with a full length PDGFalpha receptor gene and cells of a highly-metastatic clone with a truncated kinase domain PDGFalpha receptor gene. Introduction of the full length PDGFalpha receptor conferred upon low-metastatic cells the ability to grow in vitro in the presence of PDGF-AA and to colonize the lung in experimental and spontaneous metastases assays. Conversely, introduction of a truncated version of the PDGFalpha receptor into highly metastatic cells reduced their metastatic load to control levels. Accordingly, their responses to PDGF-AA, including growth stimulation and receptor autophosphorylation, were reduced. These results demonstrate that PDGFalpha receptor expression and function can control the capacity of tumor cells to generate metastases in the lung. The response of this receptor to lung-derived PDGF-like factors may define a paracrine mode of metastatic cell growth in the target organ.

Mouse platelet-derived growth factor alpha receptor: sequence, tissue-specific expression and correlation with metastatic phenotype.

High- and low-metastatic cells derived from metastatic murine tumors were screened for the differential expression of proto-oncogenes which may code for cell-surface receptors to growth factors. We found that metastatic clones of 3LL carcinoma and T10 sarcoma but not non-metastatic clones of these tumors express a 6.5-kb mRNA that is recognized by a v-fms probe containing a tyrosine kinase domain. The cloning and sequence analysis of a full-length cDNA clone corresponding to the v-fms-related 6.5-kb transcript showed that this transcript is the murine homolog of platelet-derived growth factor alpha (PDGF-alpha) receptor. The cDNA contains an open reading frame that predicts a 1089 amino acid protein. Comparison with the human and rat PDGF-alpha receptor reveals an overall amino acid sequence identity of 91% and 94% respectively. Northern blot analysis shows that this gene is preferentially expressed in the high-metastatic clones and is also selectively expressed in normal mouse tissues. Immunoprecipitation using anti-PDGF-alpha receptor serum shows that 185-kDa and 170-kDa proteins were specifically precipitated from cells of the high-metastatic D122 but not from the low-metastatic A9 cells. The possibility that overexpression of PDGF-alpha receptor in high-metastatic clones may contribute to an increase in the capacity of tumor cells to generate metastases in the lung is discussed.

PACT: cloning and characterization of a cellular p53 binding protein that interacts with Rb.

Cellular functions of tumor suppressor proteins can be mediated by protein-protein interactions. Using p53 as a probe to screen an expression library, a cDNA encoding a 250 kDa protein was isolated. Recombinant forms of this protein, designated PACT, bind to wild type p53 while two different mutations abolish this interaction. PACT protein can also interfere with p53 specific DNA binding. PACT contains a serine/arginine (SR) rich region and a C' terminal lysine rich domain. The 250 kDa PACT protein can be precipitated from cell lysates by a method specific for SR proteins. snRNPs can be co-immunoprecipitated from cells with anti-PACT antibodies. These antibodies stain cell nuclei in a speckled pattern reminiscent of the distribution of known splicing factors. Recently, RBQ1, a truncated human homologue of PACT was identified by virtue of Rb binding. We show that RBQ1 is truncated as a result of a possible mutational event. PACT can interact with both cellular Rb and p53.

Unmasking by soluble IL-6 receptor of IL-6 effect on metastatic melanoma: growth inhibition and differentiation of B16-F10.9 tumor cells.

Interleukin-6 (IL-6) inhibits the growth of melanocytes and of early stage melanoma cells, but not that of advanced melanoma cells. The in vitro IL-6 response can be restored in the highly metastatic melanoma B16-F10.9 by addition of recombinant soluble IL-6 receptor alpha-chain (sIL-6R). The F10.9 cells then undergo irreversible growth-arrest and show increased adherence with changes from epithelioid to spindleoid morphology. The sIL-6R is required for IL-6 to induce a sustained activation of the various Stat transcription factors which bind to specific IL-6 inducible enhancers. The sIL-6R and IL-6 combination causes an increase in the level of the anti-oncogenic transcription factor IRF-1 protein and DNA-binding, which remain elevated for 24 h. The promoter activity of the anti-oncogenic p21/Waf-1/Cip-1 gene is induced and accumulation of the p21 protein is observed. These results illustrate the potent agonist activity of sIL-6R on molecular pathways which could mediate the growth-arrest and differentiation of the metastatic melanoma cells. Previously observed antimetastatic effects of IL-6 therapy in mice bearing F10.9 tumors may be at least partly due to direct growth inhibition and differentiation elicited by sIL-6R present in biological fluids.

Expression of major histocompatibility class I genes in differentiating leukemic cells is temporally related to activation of c-fos proto-oncogene.

The relationship between the expression of the c-fos proto-oncogene and the expression of the class I major histocompatibility (MHC) antigens during the early stages of induced differentiation in three different leukemic cell lines was examined. In the U937 histiocytic lymphoma line TPA induced an increase in mRNA and cell surface MHC expression which followed induction of c-fos. In contrast, in the murine erythro-leukemia cell line, DMSO induced declining constitutive c-fos levels that were accompanied by declining mRNA and cell surface MHC expression. In the pluripotent HL60 promyelocytic line induction of macrophage differentiation with TPA led to c-fos induction and rising MHC levels, whereas induction of granulocyte differentiation with DMSO did not induce c-fos expression and was followed by declining MHC levels. Taken together, the results suggest that the c-fos proto-oncogene might be involved in the control of class I MHC antigen expression during differentiation.

Surrogate markers of antitumor responses: in vitro activation of T cells by autologous tumor peptides.

The increasing ability to augment antitumor immunity in model systems has led to increased numbers of clinical trials. However, progress in detecting immune responses by patients against autologous tumors has been slow. Although a considerable number of tumor antigens, as well as peptides derived from them, and the MHC determinants together with which they are presented have been identified for melanoma, this is not so for the majority of solid tumors. Furthermore, tumor cells themselves are poor stimulators of immunity. Thus, approaches that do not depend upon defined antigens or using tumor cells as stimulators would be desirable. To attempt to measure immune responses in these situations, we tested whether total peptides, prepared from autologous tumor tissue, stimulated cytokine release by T cells. Peripheral blood mononuclear cells (PBMCs) were mixed with antigen-presenting cells (APCs), pulsed with tumor peptides, and tested in the ELISPOT assay for IFN-gamma secretion. Few spots were obtained when PBMCs were cultured with unpulsed APCs or in wells with peptide-pulsed APC alone. In contrast, a strong response was seen when PBMCs were cultured with APCs that had been pulsed with autologous total tumor peptides. This system should help to identify those immunotherapeutic approaches that induce responses against tumor cells in vivo. Because different cytokine profiles are associated with distinct arms of the immune response, testing in the ELISPOT assay may also help us understand the mechanisms responsible.

Adenosine 3',5'-monophosphate suppresses metastatic spread in nude mice of steroidogenic rat granulosa cells transformed by simian virus-40 and Ha-ras oncogene.

8-Bromo-cAMP and substances elevating cAMP levels within cells, such as forskolin, cholera toxin, and Bordetella pertussis-invasive adenylate cyclase (BPAC), suppress the growth of cultured granulosa cells cotransfected by simian virus-40 (SV40) DNA and Ha-ras oncogene concomitantly with the induction of steroidogenesis and without affecting oncogene expression. We, therefore, tested the hypothesis that cAMP can modulate tumorigenesis and metastatic spread of these cells in vivo. The cotransfected cells induced rapid development of tumors when injected sc in nude mice. Tumor development was faster in less differentiated cotransfected cells originating from preantral ovarian follicles than in those obtained from highly differentiated transformed cells originating from preovulatory follicles. Cells transfected by SV40 DNA alone produced only slow-growing small tumors. Metastatic lesions of cotransfected cells were most abundant in lung and less frequent in ovaries, kidney, and spleen. No metastatic lesions were found in the liver. However, metastatic spread was dramatically suppressed when cotransfected cells injected into nude mice were pretreated with the invasive BPAC. In contrast, no suppression of metastases was observed when the cells were pretreated with 8-bromo-cAMP, forskolin, or cholera toxin. Removal of forskolin in cultured cotransfected cells yielded a rapid decrease in cAMP levels. In contrast, high levels of cAMP persist in cell cultures even several hours after 1-h pretreatment and subsequent removal of BPAC from the medium of culture cotransfected cells. It is suggested that the inhibitory effect of BPAC on the metastatic spread of these cells is due to prolonged elevation of cAMP in vivo. The newly established granulosa cell lines transformed by SV40 and the Ha-ras oncogene can serve as a model for further studies of cAMP modulation of carcinogenesis in ovarian malignancies.

The expression of PDGF-alpha but not PDGF-beta receptors is suppressed in Swiss/3T3 fibroblasts over-expressing protein kinase C-alpha.

The generation and characterization of Swiss/3T3 cells which stably over-express protein kinase C (PKC)-alpha were previously described by us. In these cells over-expression of PKC-alpha reduced the expression of epidermal growth factor (EGF) receptor molecules [(1990) J. Biol. Chem. 265, 13290-13296]. Here we show that the expression of PDGF-alpha receptors, but not PDGF-beta receptors, was specifically decreased in these cells. Not only were the levels of PDGF-alpha receptor mRNA transcript and protein significantly diminished in the PKC-alpha over-producing cells, but their ability to respond to short- and long-term growth factor signals was appropriately compromised. This was reflected in a reduced tyrosine autophosphorylation signal in response to PDGF-AA, as well as in decreased growth rates of PKC-alpha over-expressing cells when supplied with external PDGF-AA. A similar decrease in PDGF-alpha receptors was also demonstrated in parental Swiss/3T3 cells treated with phorbol esters. Our studies imply that PKC-alpha is involved in a cellular mechanism suppressing the expression of PDGF-alpha receptors in Swiss/3T3 cells. Hence, activation of PKC-alpha or alterations in its cellular levels may affect, in turn, the expression of a specific set of cell surface receptors and their responses to external growth factors.