We report the isolation of human members of a sub-family of structurally related finger protein genes. These potentially encode polypeptides containing finger motifs of the Krüppel type at the C-terminus, and a conserved amino acid module at the N-terminus; because of its invariant location the latter is referred to as finger preceding box (FPB). The FPB, detected also in previously described finger proteins from human, mouse and Xenopus, extends over approximately 65 amino acids and appears to be composed of two contiguous modules: FPB-A (residues 1-42) and FPB-B (residues 43-65). The latter is absent in some of the members analyzed. Elements A and B and the zinc finger domain are encoded by separate exons in the ZNF2 gene, a human member of this sub-family. The positioning of introns within this gene is remarkable. One intron flanks and a second interrupts the first codon of the FPB-A and FPB-B modules, respectively. A third intron occurs a few nucleotides downstream of FPB-B marking its separation from the remainder of the coding sequences. This organization, together with the absence of FPB-B in some cDNAs, supports the hypothesis that mRNAs encoding polypeptides that include one, both or none of the FPB-A and FPB-B modules may be assembled through alternative splicing pathways. Northern analyses showed that members of this sub-family are expressed as multiple transcripts in several cell lines. The sequences of distinct cDNAs homologous to the ZNF2 gene indicate that alternative splicing events adjoin either coding or non coding exons to the FPB sequences.

The hepatitis B virus (HBV), 1 a member of the hepadnaviruses family, is able to infect only humans or higher primates and has a strong organ tropism for hepatocytes (1). The mechanism of hepatocyte infection induced by this noncytolytic virusis not yet clear. The infectious unit of HBV is an enveloped virion of 42-nm diameter that contains an icosahedral nucleocapsid that encompasses a circular, partly double-stranded DNA molecule with a single strand region of variable length, a DNA-linked protein with functions critical for packaging and DNA replication, including priming, RNA-and DNA-dependent DNA polymerase, and RNAH activities. The HBV envelope that determines the targeting to the host cells and the early entry steps is composed of three proteins anchored in a lipid bilayer, occurring in mature virions in both glycosylated and nonglycosylated forms. The three proteins are called small (S), middle (M), and large (L) surface HBV antigens (HBsAg). They are the translational products of three overlapping open reading frames that start from different initiation codons localized at the 5Ј end of the preS1, preS2, and S regions of the env gene (2). The role of the preS1 region of the L protein appears to be important in cell attachment and consequently in viral infectivity, since preS1 synthetic peptides and corresponding antibodies inhibit virus binding to HepG2 cells (3,4).A cellular receptor is required for HBV binding and penetration in liver cells. Studies in rat hepatoma cells transfected with the HBV genome (5) and transgenic mice with the genome integrated (6) indicate that HBV is able to replicate in rodent cells once it bypasses the attachment and entry steps of infection, so that the absence of specific cellular receptor(s) constitutes the barrier of HBV infection and replication in nonhuman hepatocytes. A number of putative human cellular receptors for HBV have been proposed, including the receptor for immunoglobulin A (IgA), which shows a partial sequence similarity between the Fc region of the IgA ␣ chain and the preS1 domain of the virus (7,8), and the receptor for interleukin-6, due to the presence of a recognition site for the preS1 domain on interleukin-6

In situ hybridization of radiolabeled fibrinogen cDNAs to human and rat metaphase chromosomes has shown that the genes encoding the Aα, Bβ, and γ fibrinogen subunits are syntenic in both species. Our data localize the human fibrinogen gene cluster to band q31 on chromosome 4, thereby confirming and extending previous map assignments of these genes in man. We have also assigned these genes to the q31→q34 region of rat chromosome 2. This is the first map assignment of these genes in the rat and also the first report to clearly establish linkage of the Bβ subunit gene to the Aα and γ genes in this species.

The structure and the activity of urinary soluble TNF receptor type 1 (sTNF-R1), isolated from the urine of normal individuals, has been characterized and compared with that of recombinant sTNF-R1 expressed in CHO cells and with that of a nonglycosylated form expressed in Escherichia coli. Urinary sTNF-R1 was resolved in a major band of 31-33 kD and in a 48 kD band (less than 5% of total) by reducing SDS-PAGE; CHO sTNF-R1 was resolved in two bands of 29 and 31 kD. All bands were recognized by various anti-sTNF-R1 antibodies as well as by TNF-alpha in western and ligand blotting assays. No cross-reaction was observed with anti-TNF-R2 antibodies. N- and O-glycosylation studies indicated that (1) the 29-31 kD recombinant form as well as the 31-33 kD urinary form are N-glycosylated; (2) the differences between the 29-31 and 31-33 kD recombinant and natural products are mainly related to differences in the N-linked sugar content; and (3) the 48 kD sTNF-R1 isolated from urine also contains O-linked sugars. The urinary sTNF-R1 antigen mixture was able to inhibit TNF-alpha cytotoxicity with a potency comparable to that of nonglycosylated E. coli sTNF-R1. At variance, urinary sTNF-R1 was able to inhibit TNF-beta sevenfold more efficiently than E. coli sTNF-R1. In conclusion, two subtypes of sTNF-R1 have been isolated from urine: a main N-glycosylated form of 31-33 kD and a N- and O-glycosylated form of 48 kD that appears to be a minor constituent of the urinary sTNF-R1 antigen.

No abstract