Three functional members of the 1-8 gene family have been isolated on a single human genomic DNA fragment of less than 18 kb. The 1-8U and 1-8D genes are extremely similar: each is contained within a <2-kb fragment, has in its S'flanking region two adjacent 14-base-pair sequences showing high similarity to interferon-stimulable response elements (ISREs) and has two highly related exons. The third gene (9-27) has a similar overall structure, shows substantial similarity to the 1-8s but has only one ISRE which is 3' of two CCAAT boxes not present in the 1-8U and D genes. The cDNAs corresponding to the three genes share 120 nucleotides of identical sequence and show > 90% identity over 70% of the coding sequence. For the 1-8U and D genes the high similarity extends into the 5' non-coding and flanking regions. The open reading frames encode polypeptides that are likely to be of very similar structure. Antiserum to a conserved peptide detects a polypeptide(s) of about 14 kDa on PAGE which separates into three components on isoelectric focussing. The 9-27 and 1-8U genes are highly interferoninducible, the 1-8D gene is much less so. These differences are mimicked by the activities of the corresponding ISREs placed 5' of a marker gene in expression constructs. They presumably reflect differences in the interaction of the ISREs with the various interferon-inducible and constitutive factors that govern the interferon response.The human 1-8 gene family is highly inducible by both type I(a/3) and 1I(y) interferons [l, 21. Three members of the family (1-8U, 1-8D and 9-27) are linked and have been isolated on a single fragment of human genomic DNA in a cosmid vector. One of these, 9-27, which has a unique 3' non-coding region, has previously been analysed in detail [3] (GenBank accession number 3004164). It was shown to belong to a group of interferon-inducible genes having, in the 5' flanking promoter/enhancer region, an interferon-stimulable response element (ISRE) including the sequence GGAAAN(N)-GAAAC) [4-81. ISREs of this type can, depending on context, confer responsiveness to either predominantly type I interferon, as is the case, for example, for the 6-16 gene [4], or to type I and I1 interferons, as is the case for the 9-27 gene [3]. In addition, initial analysis indicated a number of potentially interesting differences in the functional elements and organisation of the 5' flanking interferon-inducible promoter/enhancer regions of the 1-8U and D, 9-27 and 6-16 genes. It was of interest, therefore, to analyse the 1-8 genes in order to compare in detail the DNA elements governing the expression of these genes and the factors interacting with them, with those known to be of importance for other interferon-inducible genes [9 -121. The coding and immediate 5' and 3' flanking regions of the 1-8U and D genes have been sequenced, allowing comparison with each other and with 9-27. The generation of antisera to a conserved peptide has permitted an initial characterisation of the corresponding polypeptides. MATERIALS AND METHO...

A 14 bp interferon (IFN)‐stimulated response element (ISRE) from 6‐16, a human gene regulated by alpha‐IFN, confers IFN inducibility on a heterologous thymidine kinase promoter. A 39 bp double‐stranded oligonucleotide corresponding to a 5′ region of 6‐16 which includes the ISRE competes for factors required for gene expression by alpha‐IFN in transfected cells and a single base change (A‐11 to C) within the ISRE (GGGAAAATGAAACT) abolishes this competition. Band‐shift assays performed with whole‐cell extracts and the 39 bp oligonucleotide reveal specific complexes formed by rapidly induced and constitutive factors, both of which fail to bind to the A‐11 to C oligonucleotide. A detailed footprinting analysis reveals that these two types of factors bind to overlapping sites within the ISRE, but in very different ways. These data were used to design oligonucleotides which decreased the formation of the inducible complex without affecting the constitutive one. Changes at the 5′ margin of the ISRE and upstream of it markedly decrease formation of the induced but not the constitutive complex and also abolish the ability of the 39 bp sequence to function as an inducible enhancer with the thymidine kinase promoter. Thus, induction of 6‐16 transcription in IFN‐treated cells is likely to be stimulated by binding of the induced factor to the ISRE and upstream sequences, while the subsequent suppression of transcription may involve competition for the ISRE by the other class of factors.

9-27 mRNA is expressed to a high level in response to both alpha and gamma interferons. In contrast, 6-16 mRNA is expressed well in response to alpha but very poorly in response to gamma interferon in human cells. The factors governing these different levels of expression were investigated. For both genes the major effect of both interferons is on transcription. A transcriptional bias in the 6-16 promoter/enhancer accounts in large part for the differential response of 6-16 to the two interferons. No single DNA element appears responsible; the smaller the 5' region analysed the lower the absolute activity and the smaller the differential response to alpha and gamma interferons observed. Both the 6-16 and 9-27 mRNAs are very stable and no effect of the interferons on stability was detected. Nor was any direct evidence obtained for preferential processing of the 9-27 mRNA. Nevertheless, differentials between the transcription and accumulation of mature mRNAs, particularly for 6-16 mRNA in response to gamma interferon, suggest that post-transcriptional control(s) must additionally operate. The 9-27 5' promoter/enhancer is much less active than that from 6-16 when placed 5' of a marker gene, despite the similar response of the two genes to alpha interferon.

A 5' cDNA clone coding for human C4b-binding protein (C4bp) was isolated, characterised and sequenced to complete the cDNA sequence coding for residues 1-32 thus confirming the protein sequence data of Chung et al. [(1985) Biochem. J. 230, 133%1411. The sequence extended to allow derivation of the putative leader peptide sequence which was 32 residues in length and showed a high of hydrophobicity typical of other documented leader sequences. Cross hybridisation was detected between the human C4bp cDNA probes and genomic DNA isolated from various species on Southern blots suggesting that genomic sequence. homologous to that coding for C4bp has been conserved during evolution.