Similar Structural Models of the Transmembrane Proteins of Ebola and Avian Sarcoma Viruses
1996; Cell Press; Volume: 85; Issue: 4 Linguagem: Inglês
10.1016/s0092-8674(00)81248-9
ISSN1097-4172
Autores Tópico(s)Viral Infections and Vectors
ResumoEpidemics of African hemorrhagic fever caused by viruses of the Filoviridae, Ebola and Marburg, have recently appeared with increasing frequency (Centers for Disease Control and Prevention, 1995). These agents have enormous potential for human morbidity and mortality. This virus family has a genomic structure and replication strategy similar to that of the Paramyxoviridae, but very little direct homology to this or other virus families at the RNA or protein levels (13Sanchez A Kiley M.P Holloway B.P Auperin D.D Virus Res. 1993; 29: 215-240Crossref PubMed Scopus (263) Google Scholar). In 1992, Volchkov et al. noted that the carboxy-terminal region of the Ebola glycoprotein (GP) bore some overall resemblance to the transmembrane (TM) protein of oncogenic Retroviruses, particularly over a 26 amino acid region that defines a conserved immunosuppressive peptide (4Cianciolo G.J Copeland T.D Oroszlan S Snyderman R Science. 1985; 230: 453-455Crossref PubMed Scopus (331) Google Scholar). Subsequent discussion has been limited to speculation as to the significance of this region in Ebola infection. We have shown that Retroviruses generally dissimilar in sequence can have limited sequence identities and similar structural propensities that can generate an overall model for homologous proteins that may otherwise not be apparent (9Gallaher W.R Ball J.M Garry R.F Griffin M.C Montelaro R.C AIDS Res. Human Retroviruses. 1989; 5: 431-440Crossref PubMed Scopus (308) Google Scholar). The model of the gp41 TM protein of human immunodeficiency virus (HIV) thus generated has since been supported in a number of laboratories (5Decroly E Cornet B Martin I Ruysschaert J.-M Vandenbranden M J. Virol. 1993; 67: 3552-3560PubMed Google Scholar, 16Wild C Dubay J.W Greenwell T Baird Jr., T Oas T.G McDanal C Hunter E Matthews T Proc. Natl. Acad. Sci. USA. 1995; 91: 12676-12680Crossref Scopus (194) Google Scholar, 1Blacklow S.C Lu M Kim P.S Biochemistry. 1995; 34: 14955-14962Crossref PubMed Scopus (126) Google Scholar). Applying similar methods to a comparison of Ebola and avian Retroviruses, it has been found that models for the carboxy-terminal 181 amino acids of Ebola GP and Rous sarcoma virus (RSV) TM are very similar (Figure 1). A description of the similarities between the two proteins is given below, beginning from the N-terminal side of the models in Figure 1. At 152-155 amino acids prior to membrane insertion, there is a polybasic region in both viruses that in RSV serves as the site of precursor cleavage yielding a new amino-terminus (14Schwartz D.E Tizard R Gilbert W Cell. 1983; 32: 853-869Abstract Full Text PDF PubMed Scopus (475) Google Scholar). Since cleavage of Ebola GP has not been reported, the functional significance of this similarity is unknown. Within 8-9 amino acids, there is in both viruses a 37-46 amino acid region bounded by cysteines, which could form a disulfide-defined loop structure. This region has at its center a sequence of 13-16 uncharged and hydrophobic amino acids in the same relative position as the fusion peptides critical for viral entry and cell fusion in the Retroviridae, most highly detailed for HIV (7Freed E Myers D Risser R Proc. Natl. Acad. Sci. USA. 1990; 87: 4650-4654Crossref PubMed Scopus (301) Google Scholar). Marburg contains within this region a canonical fusion tripeptide FFG, conservatively substituted with YFG in Ebola, and with FLG in HIV (12Richardson C Scheid A Choppin P.W Virology. 1980; 105: 205-222Crossref PubMed Scopus (227) Google Scholar, 8Gallaher W.R Cell. 1987; 50: 327-328Abstract Full Text PDF PubMed Scopus (347) Google Scholar). Further, this hydrophobic region exhibits the highest degree of identity within this loop between Ebola and Marburg (13Sanchez A Kiley M.P Holloway B.P Auperin D.D Virus Res. 1993; 29: 215-240Crossref PubMed Scopus (263) Google Scholar). Next, proceeding for 42 amino acids, there is in Ebola and RSV a region with high propensity to form an amphipathic helix similar to other Retroviridae (9Gallaher W.R Ball J.M Garry R.F Griffin M.C Montelaro R.C AIDS Res. Human Retroviruses. 1989; 5: 431-440Crossref PubMed Scopus (308) Google Scholar). Mutations in this region of HIV gp41 abolish infectivity and fusion (2Cao J Bergeron L Helseth E Thali M Repke H Sodroski J J. Virol. 1993; 67: 2747-2755Crossref PubMed Google Scholar). Homology of Ebola with Retroviral TM proteins is further reinforced by the finding of 18 of 42 amino acids identical between Ebola and RSV distributed throughout this region, including a site for N-linked glycosylation 7 amino acids after this second cysteine. Within 7-8 amino acids beyond this helical region, there is a set of cysteines separated by 6 amino acids that define a conserved motif (CX6CC) in Retroviral glycoproteins. In HIV this defines an antigenic site (10Gnann J.W.Jr Nelson J.A Oldstone M.A.B J. Virol. 1987; 61: 2639-2641Crossref PubMed Google Scholar). Overlying the latter third of the amphipathic helix and CX6CC is the immunosuppressive peptide similarity previously identified, 50% identical between Ebola and RSV. Within 2-9 residues beyond CX6CC there is a second N-linked glycosylation site that is conserved in the Retrovirus family and shared by Ebola and Marburg. This lies near or within a second region with high charge density and potential to form an amphipathic helix. Over the 41 amino acids between CX6CC and membrane insertion of Ebola GP, the identical amino acids are centered in the region of highest helical potential, with 9 of 24 positions identical between Ebola and RSV, and 10 of 24 between Ebola and Marburg. Beyond this region toward membrane insertion, the sequence diverges between Ebola and RSV. Even among the Retroviridae there is variation after the CX6CC motif, such that diversity between Ebola and RSV in this region does not detract from the overall similarity observed. Identification of regions with particular structural propensities align the highly charged helical region of Ebola GP with a region in HIV gp41 that is similar in character, if only 10% identical. A peptide analogue centered on this charged region in HIV-1 (17Wild C Greenwell T Matthews T AIDS Res. Human Retroviruses. 1993; 9: 1051-1053Crossref PubMed Scopus (363) Google Scholar), and based on our prior identification of this region (9Gallaher W.R Ball J.M Garry R.F Griffin M.C Montelaro R.C AIDS Res. Human Retroviruses. 1989; 5: 431-440Crossref PubMed Scopus (308) Google Scholar), has been found to inhibit HIV infection and fusion. Peptide analogues centered about these helical regions of Ebola and Marburg may have significant antiviral activity. Similar roles for the proposed helices of Filoviruses and Retroviruses are supported by the finding that both of these glycoproteins form trimers (11Hunter E Swanstrom R Curr. Top. Microbiol. Immunol. 1990; 157: 187-253PubMed Google Scholar, 6Feldmann H Will C Schikore M Slenczka W Klenk H.-D Virology. 1991; 182: 353-356Crossref PubMed Scopus (97) Google Scholar). In HIV-1 and SIV these helices play a role as coiled-coils in trimer formation (9Gallaher W.R Ball J.M Garry R.F Griffin M.C Montelaro R.C AIDS Res. Human Retroviruses. 1989; 5: 431-440Crossref PubMed Scopus (308) Google Scholar, 16Wild C Dubay J.W Greenwell T Baird Jr., T Oas T.G McDanal C Hunter E Matthews T Proc. Natl. Acad. Sci. USA. 1995; 91: 12676-12680Crossref Scopus (194) Google Scholar, 1Blacklow S.C Lu M Kim P.S Biochemistry. 1995; 34: 14955-14962Crossref PubMed Scopus (126) Google Scholar), as shown previously for HA2 of influenza virus (18Wilson I.A Skehel J.J Wiley D.C Nature. 1981; 289: 366-373Crossref PubMed Scopus (1919) Google Scholar). The noted similarities of Filovirus and Retrovirus transmembrane proteins thus immediately suggest investigations into Ebola and Marburg that could aid our understanding of molecular pathogenesis and medical intervention. A typical TM structure, including a fusion peptide motif, indicates that this region of the Filovirus GP is likely to play the same role in virus entry and cytopathogenicity as does the homologous TM protein superfamily in the Orthomyxoviridae, Paramyxoviridae, and Retroviridae. Once Filovirus GP genes are available in mammalian expression vectors, finely targeted site-directed mutagenesis should rapidly and safely elucidate the role of these peptide regions in Filovirus membrane biology and infectivity.
Referência(s)