Search

The structure and genetic organization of the transcription unit of the c-sis proto-oncogene was determined. Comparative nucleotide sequence analysis of the exon sequences of feline and human c-sis revealed a very high degree of homology. The cap site as well as the poly(A)-addition site of the sis transcript of each species was identified and found in similar positions. An insert of 4 amino acids was found In the deduced translational product of feline c-sis and It was located at the amino-terminus of the region that constituted the platelet-derived growth factor domain. An insert of 149 bp present at the 5′ end of exon 7 of human c-sis but absent In the simian sarcoma virus v-sis oncogene was also present In the feline c-sis proto-oncogene.

In previous studies, the nucleotide sequence of a 3′-portion of the human fur gene was determined (1,2). Here, the nucleotide sequence of the region immediately upstream of the described sequence (1) is presented. The sequence data of both DNA strands were obtained from overlapping clones. Based upon preliminary cDNA sequence and nuclease S1 analysis, the cap site of the fur gene could be localized at position 4597. A putative “TATA” box is located 32 bp upstream of the cap site. The repetitive sequences beloning to the Alu family (3) are underlined.

The chemical shifts of 1H resonances of non exchangeable protons (except H5′, H5″ and adenine H2) of over six hundred nucleotides have been collected. The influence which the base of the nucleotide itself as well as the bases on its 5′ and 3′ side exert on the chemical shifts of the various resonances has been investigated. Most of the resonances appear to be predominantly influenced by only one base. For H2′, H2″, H3′, H4′ and H6/H8 this is the base of the central nucleotide, for H5(C) and CH3 it is the one on the 5′ side and for H1′ it is the one on the 3′ side. Chemical shift distribution profiles are presented which allow an estimation of the probability of finding a particular resonance at a particular position in the spectrum.

Comparative analysis of coamid clones containing the human and feline c-sis genetic regions revealed the similar structural organization of these areas in the two species. The areas shared seven different genetic regions in and around the c-sis locus and one of these was related to v-sis. Another region, 1.9 kbp in size and located about 8 kbp upstream of the v-sis homologous region in the human genome, also hybridized to the main c-sis transcriptional product of 3.5 kb. Comparison with a recently described c-sis cDNA clone (Collins et al., Nature 316, 748–750 (1985)) revealed that the 1.9 kbp DNA region contained a large 5′ c-sis exon of at least 1050 bp. In this exon, the presumed initiation site of the predicted PDGF-2 containing precursor protein was located and appeared to be preceded by a large untranslated region. In the region Immediately upstream of this exon, a TATA box and a consensus sequence for a potential Sp1 binding site were found at similar positions in both species. This region also exhibited promoter activity when tested in an assay in which coding sequences of bacterial chloramphenicol acetyltransferase (CAT; acetyl-CoA: chloramphenicol 3-0-acetyltransferase, EC 2.3.1.28) were placed under its control. The five other DNA regions were found upstream and downstream of the human c-sis transcription unit and also in an intron. Four of them contained repetitive sequences. Hybridization analysis of human and feline c-sis containing cosmid clones with a mixed synthetic nucleotide probe, which corresponded to sequences encoding amino acid residues 2–7 of chain 1 of platelet-derived growth factor (PDGF-1), suggested that the c-sis cosmld clones did not include PDGF-1-specific genetic sequences.

The monoclonal antibody MOC-32 detected a 40 kDa protein in Western blot analysis. Immunological screening of an expression library of human SCLC cells with MOC-32 led to the isolation of overlapping cDNA clones. One of these, cHD4, was 1.0 kbp long and of about the same size as its corresponding mRNA. Preceded by an in phase stop codon, an open reading frame of 885 bp was present in cHD4 and a translational product of only 33 kDa could be calculated. Biochemical and immunological analysis established the relationship between the 40 kDa antigen and the isolated coding sequences and resolved the apparent discrepancy between the calculated molecular weight and the observed electrophoretic mobility. Nucleotide sequence comparison of cHD4 to the EMBL database revealed that cHD4 was nearly identical to a sequence claimed to encode a laminin binding protein. Southern blot and nucleotide sequence analysis indicated the presence of multiple copies of the gene in the human genome. At least five of these appeared to represent processed pseudogenes.