Cell, Vol. 33, 939-947, July 1983, Copyright Q 1983 by MIT 0092.8674/83/070939-09 $OZ.CO/O Molecular Cloning of Gene Sequences Regulated by Platelet-derived Growth Factor Brent H. Cochran,‘Angela C. Reffel, and Charles D. Stiles Department of Microbiology and Molecular Genetics Harvard Medical School and Dana-Farber Cancer Institute Boston, Massachusetts 02115 Summary We have screened a cDNA library for gene se- quences that are regulated by platelet-derived growth factor (PDGF) in BALB/c-3T3 cells. Of 8000 clones screened, less than 14 independent PDGF- inducible sequences were found. Two of these (KC and JE) were studied in detail. By hybrid-selection and translation the KC and JE mRNAs encode 10,000 and 19,000 dalton polypeptides, respectively. In the absence of PDGF, the JE and KC sequences corre- spond to low abundance mRNAs. One hour after addition of PDGF their abundance level can be in- creased lo- to 20-fold. Within 4 hr, a 80-fold induc- tion of JE can be attained. Nanogram per ml quan- tities of pure PDGF regulate these sequences whereas pg/ml quantities of chemically unrelated mitogens (EGF, insulin, or platelet-poor plasma) have either a weak or an undetectable effect. Inhib- itors of protein synthesis block the progression of quiescent 3T3 cells through Gl into S phase; how- ever these drugs do not block the induction of KC and JE by PDGF. This result indicates that these sequences correspond to “early genes” which are not induced as a consequence of cell growth, but rather are directly regulated by PDGF. Introduction Platelet-derived growth factor (PDGF) is a heat-stable cat- ionic polypeptide that can be purified to homogeneity from human platelet lysates (Antoniades et al., 1979; Heldin et al., 1979; Deuel et al., 1981; Raines and Ross, 1982). PDGF is contained within the alpha-granules of circulating blood platelets (Kaplan, D. R., et al., 1979; Kaplan, K. L., et al., 1979; Gerrard et al., 1980) and is normally released only when blood clots (Ross et al., 1974; Kohler and Lipton, 1974). Thus clotted blood serum contains about 15 rig/ml of PDGF whereas platelet-poor plasma (the product of unclotted blood) contains less than 1 .O rig/ml (Singh et al., 1982). Smooth muscle cells (Ross et al., 1974) fibroblasts (Kohler and Lipton, 1974; Scher et al., 1978) and a variety of other mesenchymal cells require PDGF for optimum growth in vitro whereas epithelioid cells and hematopoietic cells do not (Stiles et al., 1979). This target specificity of PDGF is reflected in the presence or absence of PDGF- specific receptors in cultured cell lines (Heldin et al., 1981; Huang et al., 1982; Singh et al., 1982; Bowen-Pope and Ross, 1982). The observations that one, PDGF is normally released only during clot formation and two, only connec- tive tissue cells can respond to PDGF have led to the attractive (though unproven) hypothesis that PDGF plays a central role in maintenance of the vascular lining (Ross and Glomset, 1976) or perhaps in wound healing re- sponses (Scher et al., 1979) in vivo. Overshadowing all of these findings is the demonstration that both RNA and DNA transforming viruses contain gene sequences which directly override the growth requirement for PDGF (Scher et al., 1978). PDGF, by itself, is not an efficient mitogen for 3T3 cells, A second set of growth factors contained in the platelet- poor plasma fraction of serum functions synergistically with PDGF to promote the optimum mitogenic response (Pledger et al., 1977; Vogel et al., 1978). The active components of plasma appear similar to, and can be replaced by, epidermal growth factor and insulin-like growth factors (somatomedins) (Stiles et al., 1979; Leof et al., 1982). In the absence of these plasma growth factors, PDGF-treated 3T3 cells remain growth arrested. Thus bio- chemical changes observed following PDGF-treatment do not occur merely as a consequence of growth and pro- gression through the cell cycle. The ability to uncouple the immediate biochemical responses to PDGF from cell growth per se facilitates the analysis of PDGF mitogenic action. In this article, we describe the isolation and char- acterization of rare gene sequences that are induced promptly (within 1 hr) by rig/ml concentrations of pure PDGF. Results Preparation and Screening of cDNA Clones from PDGF-inducible Genes A cDNA library was prepared from cells treated with partially purified PDGF. This cDNA library was screened for PDGF-inducible gene sequences by differential colony hybridization as described in Experimental Procedures. Colonies which gave positive hybridization signals when probed with PDGF-cDNA, but not when probed with quies- cent cell-cDNA, were picked as presumptive PDGF-induc- ible genes (see Figure 1). Because the mechanics of colony transfer to nitrocellulose filters generate some spu- rious signals in the differential hybridization assay, all clones scored as presumptively inducible after one round of screening were twice rescreened. Of approximately 8000 clones screened in this manner, 55 were scored as clearly inducible after three rounds of screening. Ideally, the number of times a given gene sequence is found in a cDNA library is proportional to the abundance of that particular gene in the mRNA population. Therefore, we reasoned that some PDGF-inducible sequences might have been cloned and isolated multiple times. To assess this possibility, cDNA inserts were excised from individual plasmids, nick-translated, and hybridized against the other PDGF-inducible clones. The results of this analysis are shown in Table 1. Forty-six of the clones represented only five independent gene sequences. The identification of multiple copies of individual genes in the library served as