Search

The genome of a recently identified virus, hepatitis G virus (HGV), shows considerable homology to hepatitis C virus (HCV). Two HGV proteases similar to nonstructural proteins NS2 and NS3 of HCV were identified, and their cleavage site specificity was investigated. Amino acids essential for the protease activities were determined by mutation analysis. NS4A of HGV was demonstrated to be a cofactor for NS3-mediated proteolysis, with a region critical for activity residing between Leu1561, and Ala1598.

Background: Shc and Grb2 form a complex in cells in response to growth factor stimulation and link tyrosine kinases to Ras during the resulting signaling process. Shc and Grb2 each contain domains that mediate interactions with other unidentified intracellular proteins. For example, the Shc PTB domain binds to 130 kDa and 145 kDa tyrosine-phosphorylated proteins in response to stimulation of cells by growth factors, cytokines and crosslinking of antigen receptors. The Grb2 SH3 domains bind to an unidentified 116 kDa protein in T cells. We have identified three proteins, of 110 kDa, 130 kDa and 145 kDa, as a new family of molecules encoded by the same gene. In vivo studies show that these proteins form signal transduction complexes with Shc and with Grb2., Results: The 130 kDa and 145 kDa tyrosine-phosphorylated proteins that associate with the Shc PTB domain were purified by conventional chromatographic methods. Partial peptide and cDNA sequences corresponding to these proteins, termed SIP-145 and SIP-130 (SIP for signaling inositol polyphosphate 5-phosphatase), identified them as SH2 domain-containing products of a single gene and as members of the inositol polyphosphate 5-phosphatase family. The SIP-130 and SIP-145 proteins and inositol polyphosphate 5-phosphatase activity associated with Shc in vivo in response to B-cell activation. By using an independent approach, expression cloning, we found that the Grb2 SH3 domains bind specifically to SIP-110, a 110 kDa splice variant of SIP-145 and SIP-130, which lacks the SH2 domain. The SIP proteins hydrolyzed phosphatidylinositol (3,4,5)-trisphosphate (PtdIns (3,4,5)-P3) and Ins (1,3,4,5)-P4, but not PtdIns (4,5)-P2 or Ins (1,4,5)-P3., Conclusions: These findings strongly implicate the inositol polyphosphate 5-phosphatases in Shc- and Grb2-mediated signal transduction. Furthermore, SIP-110, SIP-130 and SIP-145 prefer 3-phosphorylated substrates, suggesting a link to the phosphatidylinositol 3-kinase signaling pathway.

The expression of the core gene of two different hepatitis C virus (HCV) isolates was analyzed. In the presence of its downstream E1 envelope protein sequence, two major core protein products with molecular masses of 21 kDa (P21) and 19 kDa (P19) and a minor protein product with molecular mass of 16 kDa (P16) were detected. In the absence of its downstream E1 envelope protein sequence, P21 and P19 remained the major protein products expressed from the core gene of the HCV-RH isolate, whereas P16 became the major protein product of the core gene of the HCV-1 isolate. Analysis of the amino-terminal sequences of P21 and P16 expressed in Escherichia coli revealed that P21 and P16 were co-amino terminal. Deletion-mapping analysis indicated that P16 lacked the carboxy-terminal sequence of P21. Immunofluorescence analysis of the subcellular localization of different HCV core proteins indicated that P21 and P19 displayed a reticular and punctate staining pattern typical of endoplasmic reticulum-associated proteins, while P16 was localized to the nucleus. The distinct subcellular localization of P16 raises the possibility that P16 may have a biological function very different from those of P21 and P19.

We have identified and characterized a novel human tissue inhibitor of metalloproteinase (TIMP). It is found exclusively in the extracellular matrix of a large number of cultured human cells, including: primary embryonal kidney (293), neuroblastoma (SK-N-SH), normal whole embryo (FHs 173We), cervical carcinoma (HeLa S3), colon adenocarcinoma (Caco-2), ileocecal adenocarcinoma (HCT-8), fibrosarcomas (SW 684 and Hs 913T) and normal gingival fibroblasts (GF11 and 1292). It was not detected in the conditioned media from any of these cell lines. Its apparent molecular mass of 24-25 kDa, as determined by its migration on protease-substrate gels, is intermediate between TIMP-1 (28.5 kDa) and TIMP-2 (21 kDa). Like the latter two proteins, human TIMP-3 contains intrachain disulfide bonds and displays altered electrophoretic mobility in the presence of beta-mercaptoethanol. The N-terminal, amino acid sequence of the protein is identical to that of chicken TIMP-3 (ChIMP-3), and its amino acid composition is similar. The protein is not N-glycosylated, as determined by treatment with N-glycosidase-F. Finally, it is recognized by antisera raised against pure ChIMP-3 but not by anti-human TIMP-1 or anti-human TIMP-2 antibodies. Based on these properties, we propose that this protein is TIMP-3 and is the human counterpart of ChIMP-3 (Pavloff et al., J. Biol. Chem. 267: 17321-17326, 1992). Two additional inhibitors detected in the matrix of human cell lines, designated inhibitor of metalloproteinase (IMP)-a and IMP-b, migrate with apparent masses of 29 kDa and 30 kDa. Both are N-glycosylated. A fourth inhibitor activity, which is smaller in mass than TIMP-3 and is also pecifically located in the matrix, is detectable in some cell lines.

Hepatitis C virus (HCV), the principal cause of parenteral non-A, non-B hepatitis, is an RNA virus and a member of the Flaviviridae family. Its genome is translated into a single polyprotein that is processed co- and post-translationally into both structural and nonstructural (NS) proteins. There are three putative structural proteins, consisting of the nucleocapsid protein and two envelope glycoproteins, E1 and E2. Analysis of transient transfections of serially extended templates covering the E2/NS2 region provided evidence for three E2 species with distinct C-termini. One form is E2 terminating at amino acid 729, while the larger two species represent fusions with the downstream NS2A and NS2A/NS2B proteins terminating at amino acids 809 and 1026, respectively. Using the same E2 templates, we defined a region of E2 important for co-immunoprecipitation of E1 and observed that this region also prevents E2 secretion. The N-terminus of NS2B was determined by radiosequencing and a novel association of NS2B and probable NS4B with E2 was observed; the regions of NS2B and E2 important for this association have been mapped. These data indicate that complex processing and protein:protein interactions occur during HCV morphogenesis.

Herpes simplex viruses (HSV) use multiple and sequential receptors to enter host cells. HSV glycoprotein D (gD) has been implicated in binding to cellular receptors that facilitate virus penetration into cells. We used soluble forms of gD that were expressed in Chinese hamster ovary cells to characterize and identify a putative cellular receptor for HSV as the 275-kDa mannose 6-phosphate/insulin-like growth factor II receptor. Soluble gD also bound to the 46-kDa cation-dependent mannose 6-phosphate (Man-6-P) receptor and was extensively modified with Man-6-P residues on its Asn-linked oligosaccharides. Additionally, soluble gD was a high affinity substrate for N-acetylglucosamine-1-phosphotransferase, the first enzyme in the biosynthetic pathway for the addition of Man-6-P residues to lysosomal enzymes. The membrane form of gD immunoprecipitated from HSV-infected cells also bound to the 275-kDa mannose 6-phosphate/insulin-like growth factor II receptor, albeit poorly, and only a small fraction of the membrane gD was modified with Man-6-P. Notwithstanding this low level of mannose phosphorylation, the interaction between gD and Man-6-P receptors may play a role in some aspect of virus entry or egress.

Mutations in the human APC gene are linked to familial adenomatous polyposis and to the progression of sporadic colorectal and gastric tumors. To gain insight into APC function, APC-associated proteins were identified by immunoprecipitation experiments. Antibodies to APC precipitated a 95-kilodalton protein that was purified and identified by sequencing as beta-catenin, a protein that binds to the cell adhesion molecule E-cadherin. An antibody specific to beta-catenin also recognized the 95-kilodalton protein in the immunoprecipitates. These results suggest that APC is involved in cell adhesion.

We have recently cloned cDNAs encoding human insulin-like growth factor binding proteins (IGFBP)-4, -5 and -6 and have now expressed these BPs in yeast as ubiquitin (Ub)-IGFBP fusion proteins. Western ligand blotting with 125I-IGF II under nonreducing conditions of recombinant human (rh) IGFBP-containing yeast lysates revealed specific binding bands for IGFBP-4, -5, and -6 at apparent molecular masses of 24-26, 30-32, and 24-26 kDa, respectively, indicating expression and processing of the fusion proteins. HPLC purified rhIGFBPs had virtually the same amino acid composition, amino acid number, and NH2-terminal sequences as the native BPs. Rabbit antiserum directed against each rhIGFBP-4, -5 and -6 reacted specifically with the respective rhIGFBP as well as with the native human counterpart and displayed very low cross-reactivity with other IGFBPs. Except for the affinity of rhIGFBP-6 for IGF I (Ka = 8.5 x 10(8) M-1), the affinity constants of the three IGFBPs for IGF I and II lie between 1.7 and 3.3 x 10(10) M-1. When present in excess, rhIGFBP-4, -5, and -6 inhibited IGF I- and II-stimulated DNA and glycogen synthesis in human osteoblastic cells, although rh-IGFBP-6 had only a weak inhibitory effect on IGF I in agreement with its relatively lower IGF I affinity constant.

The insulin-like growth factor binding protein (IGFBP) family comprises six structurally distinct, but highly homologous proteins. They have been identified in serum and other biological fluids, tissue extracts, and cell culture media. We have recently cloned cDNAs encoding human IGFBP-4, -5, and -6 and have now expressed these BPs in yeast as ubiquitin (Ub)-IGFBP fusion proteins. Western ligand blotting with 125I-IGF II under nonreducing conditions of recombinant human (rh) IGFBP-containing yeast lysates revealed specific binding bands for IGFBP-4, -5, and -6 at apparent molecular masses of 24-26, 30-32, and 24-26 kDa, respectively, indicating processing of the fusion proteins. High-performance liquid chromatography-purified rhIGFBPs had virually the same amino acid composition, amino acid number, and NH2-terminal sequences as the native BPs. Except for the affinity of rhIGFBP-6 for IGF I (Ka = 8.5 x 10(8) M-1), the affinity constants of the three IGFBPs for IGF I and II lie between 1.7 and 3.3 x 10(10) M-1, i.e. 25-100 times higher than the IGF I and II affinities of the type I IGF receptor. When present in excess, rhIGFBP-4, -5, and -6 inhibited IGF I- and II-stimulated DNA and glycogen synthesis in human osteoblastic cells, but rhIGFBP-6 had only a weak inhibitory effect on IGF I in agreement with its relatively lower IGF I affinity constant. The results of this study show that the primary effect of the three rhIGFBPs is the attenuation of IGF activity and suggest that IGFBPs contribute to the control of IGF-mediated cell growth and metabolism.

The complete amino acid sequence of fibrolase, a fibrinolytic enzyme from southern copperhead (Agkistrodon contortrix contortrix) venom, has been determined. This is the first report of the sequence of a direct-acting, nonhemorrhagic fibrinolytic enzyme found in snake venom. The majority of the sequence was established by automated Edman degradation of overlapping peptides generated by a variety of selective cleavage procedures. The amino-terminus is blocked by a cyclized glutamine (pyroglutamic acid) residue, and the sequence of this region of the molecule was determined by mass spectrometry. Fibrolase is composed of 203 residues in a single polypeptide chain with a molecular weight of 22,891, as determined by the sequence. Its sequence is homologous to the sequence of the hemorrhagic toxin Ht-d of Crotalus atrox venom and with the sequences of two metalloproteinases from Trimeresurus flavoviridis venom. Microheterogeneity in the sequence was found at both the amino-terminus and at residues 189 and 192. All six cysteine residues in fibrolase are involved in disulfide bonds. A disulfide bond between cysteine-118 and cysteine-198 has been established and bonds between cysteines-158/165 and between cysteines-160/192 are inferred from the homology to Ht-d. Secondary structure prediction reveals a very low percentage of alpha-helix (4%), but much greater beta-structure (39.5%). Analysis of the sequence reveals the absence of asparagine-linked glycosylation sites defined by the consensus sequence: asparagine-X-serine/threonine.

Three insulin-like growth factor binding proteins (IGFBP) with apparent molecular masses of 24, 28-30, and 30 kDa, nonreduced, have been isolated from human serum. The 15 NH2-terminal amino acids of the 24-kDa binding protein are identical with those of the 30-kDa BP. The apparent molecular mass of the latter is reduced to 24 kDa by N-glycanase, suggesting that the 30-kDa BP is the glycosylated form of the isolated 24-kDa BP. The complete amino acid sequences derived from the cloned cDNAs represent two new IGFBPs. They are tentatively termed IGFBP-4 and -5. The prepeptide sequences of BP-4 and -5 contain 27 and 21, the mature proteins 213 and 237 amino acids, respectively (Mr = 22,610 and 25,980). The NH2- and COOH-terminal thirds of BP-4 and -5 display pronounced homology to the other three human BPs. 16 of the 16-20 cysteines and 37 of the 213-289 amino acids (12.8-17.1%) are conserved in all five mature BPs. 10 amino acid positions located in the NH2-terminal region and shared by BP-1, -2, -3, and -5 are different in BP-4. These differences may account for the preferential affinity of BP-4 for IGF II. A most intriguing homology exists between the COOH-terminal quarter of the five IGFBPs, 10 repetitive domains of human thyroglobulin, a gastrointestinal tumor-associated antigen, and the invariant chain of the class II histocompatibility antigen. The cDNAs of five human IGFBPs are now available. They will allow their expression and production in sufficient quantities for in vivo studies to unravel their role in growth and metabolism.

Thioredoxin h has been purified to electrophoretic homogeneity from spinach roots using a procedure devised for leaves. The root thioredoxin (h2 form) differed from chloroplast and animal thioredoxins in showing an atypical active site (Cys-Ala-Pro-Cys) but otherwise resembled animal thioredoxin in structure. Sequence data for a total of 72 residues of spinach root thioredoxin h2 (about 69% of the primary structure) showed 43-44% identity with rabbit and rat thioredoxin. Analysis of cell fractions from the endosperm of germinating castor beans revealed that thioredoxin h occurs in the cytosol, endoplasmic reticulum, and mitochondria. The present findings demonstrate a similarity between plant thioredoxin h and animal thioredoxins in structure and intracellular location and raise the question of whether these proteins have similar functions.

A new method for characterizing O-linked glycopeptides without chemical degradation is presented. Collision-induced dissociation (CID) analysis of intact O-linked glycopeptides containing mono- and disaccharides was performed. For glycopeptides containing one hexose unit, both the peptide sequence and the site of attachment of the sugar moiety were obtained from a single high-energy CID spectrum. However, in a glycopeptide bearing multiple sugar residues per site, the CID spectrum was dominated by fragments resulting from cleavages of the carbohydrate substituents and the gas-phase deglycosylated peptide, thus obviating the concomitant observation of peptide sequence ions. Hence, information on the structures of the carbohydrate substituents was obtained, but not on the sites of attachment of these residues to the peptide. Subsequent CID analysis of the gas-phase deglycosylated peptide ion can be used to obtain the sequence of the peptide backbone from the same sample. This method holds promise for simultaneously determining the carbohydrate structure and the peptide sequence of intact O-linked glycopeptides without chemical degradation.

High-performance tandem mass spectrometry has been employed to structurally characterize intact O-linked glycopeptides and establish the complexity and extent of glycosylation for recombinant human platelet-derived growth factor B chain (rhPDGF-B) expressed in yeast. In addition, liquid secondary ion mass spectrometry (LSIMS) and Edman degradation have been employed to verify the protein sequence. LSIMS of high-performance liquid chromatographically fractionated proteolytic digests confirmed the complete amino acid sequence predicted by the human PDGF-B gene structure. Potential glycopeptides (as indicated by a mass shift of 162 or 324 Da from the mass of a predicted cleavage product) were sequenced using tandem mass spectrometry and Edman degradation. Ultraviolet matrix laser desorption mass spectrometry of rhPDGF-B dimer was used to determine the molecular weight distribution for the intact recombinant glycoprotein. In addition to the presence of unmodified peptides, corresponding peptides bearing monomannosyl moieties were found on serine 26 and threonines 20, 63, 88, 90 and 101. Further, dimannosyl moieties were found on threonines 6 and 63. These data reveal the presence of O-linked glycosylation at sites which do not fortify the concept of a consensus sequence involving proline residues, but which strengthen the concept of secondary and tertiary structure requirements. The advantages of high-energy collisionally induced dissociation analysis of O-linked glycopeptides over conventional base elimination and borohydride reduction and other mass spectrometric techniques are presented for the first time.

The scrapie agent causes a progressive degeneration of the central nervous system of animals after a prolonged incubation period. Measurements of incubation period length, defined as the time from inoculation to the onset of clinical signs of neurological dysfunction, were related to the titer of the agent and the dilution of the inoculated sample. Equations defining the relationship provide a new assay for the agent requiring fewer animals than end point titrations. By use of this incubation period assay, the scrapie agent from hamster brain was found to have an s20,w of < 300 S but > 30 S assuming rho p = 1.2 g/cm3. A partially purified fraction P3 was obtained by differential centrifugation and sodium deoxycholate extraction. When P3 was extracted with phenol, virtually no infectivity was found in the aqueous phase even after examining such variables as pH, salt concentration, and predigestion of samples with proteinase K. Nonionic and nondenaturing, anionic detergents did not inactivate the scrapie agent; in contrast, denaturing detergents inactivated the agent. Sodium dodecyl sulfate (NaDodSO4) inactivated greater than 90% of the agent at a NaDodSO4 to protein ratio of 1.8 g/g. Inactivation by NaDodSO4 appears to be a cooperative process. Addition of a nonionic detergent to form mixed micelles with NaDodSO4 prevented inactivation of the agent by NaDodSO4. Weak chaotropic ions do not inactivate the scrapie agent while strong chaotropic ions like SCN- and Cl3CCOO- destroy infectivity at concentrations of 0.2 M. These data provide evidence in support of a protein component within the scrapie agent which is essential for maintenance of infectivity. Thus, it is unlikely that the scrapie agent is composed only of a "naked" nucleic acid as is the case for the plant viroids.

A DNA fragment containing the structural gene for the 5 S ribosomal RNA and intergenic regions before and after the 35 S ribosomal RNA precursor gene of Saccharomyces cerevisiae has been amplified in a bacterial plasmid and physically mapped by restriction endonuclease cleavage and hybridization to purified yeast 5 S ribosomal RNA. The nucleotide sequence of the DNA fragments carrying the 5 S ribosomal RNA gene and adjacent regions has been determined. The sequence unambiguously identifies the 5 S ribosomal RNA gene, determines its polarity within the ribosomal DNA repeating unit, and reveals the structure of its promoter and termination regions. Partial DNA sequence of the regions near the beginning and end of the 35 S ribosomal RNA gene has also been determined as a preliminary step in establishing the structure of promoter and termination regions for the 35 S ribosomal RNA gene.

The gene encoding the glycoprotein H (gH) homologue of CMV strain Towne was cloned, sequenced, and expressed. The predicted 742 amino acid gH protein had characteristics typical of a membrane glycoprotein including hydrophobic signal and transmembrane domains and six possible N-linked glycosylation sites. The CMV (Towne) gH gene had a 95% nucleotide identity and a 96.6% amino acid identity with the CMV (AD169) gH gene, as described by M. P. Cranage, G. L. Smith, S. E. Bell, H. Hart, C. Brown, A. T. Bankier, P. Tomlinson, B. G. Barrell, and T. C. Minson (1988, J. Virol. 62, 1416-1422). Transcriptional analysis of the gH gene revealed that the 2.9-kilobase (kb) gH transcript was not detected until late after CMV infection, indicating that the kinetics of gH expression were typical of the late class of CMV genes. The gH gene was expressed in COS cells using a vector in which transcription was driven by the SV40 early promoter. The expression of gH was detected by immunofluorescence using the virus neutralizing murine monoclonal antibody 1G6, which is specific for an 86-kilodalton (kDa) CMV virion membrane protein (p86). Amino acid sequence analysis of p86 tryptic peptides revealed sequence identity with peptides from the deduced gH amino acid sequence, confirming that the gH gene encodes p86. These results indicate that CMV gH can induce virus neutralizing antibodies and establishes gH as a candidate antigen for a subunit vaccine against CMV.

The scrapie agent causes a degenerative nervous system disease in sheep and goats. Studies with extensively purified preparations demonstrated that the agent contains a protein that is required for infectivity. Chemical modification of the scrapie agent by diethyl pyrocarbonate reduced the titer 1000-fold. Exposure of the inactivated agent to hydroxylamine, a strong nucleophile, resulted in complete restoration of infectivity. Presumably, nucleophilic residues within a scrapie agent protein undergo carbethoxylation on reaction with diethyl pyrocarbonate, and subsequent addition of hydroxylamine displaces these carbethoxy groups.

A cloned approximately 5 kb cDNA (human placenta) contains the coding sequences for the insulin receptor. The nucleotide sequence predicts a 1382 amino acid precursor. The alpha subunit comprises the N-terminal portion of the precursor and contains a striking cysteine-rich "cross-linking" domain. The beta-subunit (the C-terminal portion of the precursor) contains a transmembrane domain and, in the intracellular region, the elements of a tyrosine phosphokinase: an ATP-binding site and a possible tyrosine autophosphorylation site or sites. The overall structure is reminiscent of the EGF receptor; the cross-linking domain of the alpha subunit and several regions of the beta subunit exhibit sequence homology with the EGF receptor. The phosphokinase domain also exhibits homology with some oncogenic proteins that have tyrosine phosphokinase activity, in particular, a striking homology with v-ros. Southern blotting experiments suggest that the coding region spans more than 45 kb. The insulin receptor gene is located on chromosome 19.

Calmodulin was purified from Saccharomyces cerevisiae based on its characteristic properties. Like other calmodulins, the yeast protein is small, heat-stable, acidic, retained by hydrophobic matrices in a Ca2+-dependent manner, exhibits a pronounced Ca2+-induced shift in electrophoretic mobility, and binds 45Ca2+. Using synthetic oligonucleotide probes designed from the sequences of two tryptic peptides derived from the purified protein, the gene encoding yeast calmodulin was isolated. The gene (designated CMD1) is a unique, single-copy locus, contains no introns, and resides on chromosome II. The amino acid sequence of yeast calmodulin shares 60% identity with other calmodulins. Disruption or deletion of the yeast calmodulin gene results in a recessive-lethal mutation; thus, calmodulin is essential for the growth of yeast cells.

Extracts of rat intestine contain nine soluble lactose-binding lectins with subunit molecular weights ranging from 14,500 to 19,000 that were purified by affinity chromatography and ion-exchange chromatography. Two of them are either identical with or closely related to other known rat lectins. A third appears to be the isolated carbohydrate-binding C-terminal domain of a known lectin but lacks the N-terminal domain presumed to mediate a different function. The others have not been described previously. Among them, the major rat intestinal lectin, RI-H, and a related protein, RI-G, have N-terminal amino acid sequences with similarities to sequences found in other known rat lectins. Therefore, these results introduce new members of a growing family of these structurally homologous soluble lactose-binding proteins.

Transcription factor Sp1 is a protein present in mammalian cells that binds to GC box promoter elements and selectively activates mRNA synthesis from genes that contain functional recognition sites. We have isolated a cDNA that encodes the 696 C-terminal amino acid residues of human Sp1. By expression of truncated fragments of Sp1 in E. coli, we have localized the DNA binding activity to the C-terminal 168 amino acid residues. In this region, Sp1 has three contiguous Zn(II) finger motifs, which are believed to be metalloprotein structures that interact with DNA. We have found that purified Sp1 requires Zn(II) for sequence-specific binding to DNA. Thus, it is likely that Sp1 interacts with DNA by binding of the Zn(II) fingers. To facilitate the identification of mutant variants of Sp1 that are defective in DNA binding, we have also devised a bacterial colony assay for detection of Sp1 binding to DNA.

The gene encoding glycoprotein B of human cytomegalovirus (CMV) strain Towne was cloned, sequenced, and expressed in order to study potential targets for viral neutralization. Secondary structure analysis of the 907 amino acid protein predicted a 24 amino acid N-terminal signal sequence and a potential transmembrane region composed of two domains, 34 and 21 amino acids. The CMV (Towne) gB gene had a 94% nucleotide similarity and a 95% amino acid similarity to the CMV (AD169) gB gene [as described by M.P. Cranage et al. (1986, EMBO J. 5, 3057-3063)]. Transcriptional analysis of the CMV (Towne) gB coding strand revealed that the gB message (3.9 kb), was transcribed from this region as early as 4 hr postinfection, and well in advance of gB protein synthesis. Full-length and truncated versions of the gB gene were expressed in COS cells using expression vectors where transcription was driven by the SV40 early promoter or the CMV major immediate early promoter. Expression was detected by immunofluorescence and ELISA using the virus neutralizing murine monoclonal antibody 15D8 (L. Rasmussen, J. Mullenax, R. Nelson, and T.C. Merigan, 1985, J. Virol. 55, 274-280). This antibody had been shown previously to recognize a 55-kDa CMV virion protein and a related 130-kDa intracellular precursor. Amino acid sequence analysis of the N-terminus of the 55-kDa viral glycoprotein (gp55) showed that gp55 is derived from gB (gp130) by proteolytic cleavage and represents the C-terminal region of gp130. The truncated version of gB expressed in COS and CHO cells was also processed by proteolytic cleavage as demonstrated by Western blotting. Our study localizes the epitope recognized by 15D8 to within a 186 amino acid fragment of the gp55 protein. These results indicate that CMV gB is a target for neutralization and establishes gp55 as a candidate component for use in a subunit vaccine.