Cross-reactivity of a normal human cell surface antigen with primate retrovirus glycoproteins

Polymorphic expression of CD46 protein isoforms due to tissue-specific RNA splicing

CD46 is a member of the regulators of complement activation (RCA) family and serves to protect autologous cells from complement mediated lysis. The CD46 gene consists of 14 exons and extensive RNA splicing produces protein isoforms of different molecular weight. Predominant protein isoforms of 66 and 56 kDa arise from splicing in or out of exon 8 which encodes a region rich in serine, threonine and proline residues known to be heavily O-glycosylated. An inherited allelic polymorphism controls the relative expression of these isoforms in PBL and other tissues. This study has analysed an independent and overriding tissue specific regulation of CD46 splicing. Salivary gland and kidney produce RNA transcripts that preferentially include exon 8, giving rise to the 66 kDa protein species, while exon 8 is spliced out in brain tissue to give the 56 kDa protein. The cytoplasmic tail of CD46 is encoded by either exon 13 (CYT 1) or exon 14 (CYT 2). There is a preferential deletion of exon 13 from transcripts in salivary gland, kidney and brain to encode a protein containing cytoplasmic tail CYT 2. This preferential production of the CYT 2 tail is contrary to that seen on peripheral blood lymphocytes where equivalent expression of both CYT 1 and CYT 2 is observed. Our results suggest that while the splicing of exons within most cells is controlled by nucleotide sequences within or close to the CD46 gene (i.e. cis-regulation), splicing in tissues such as salivary gland, kidney and brain is regulated by trans-splicing factors encoded by another gene(s).

Membrane cofactor protein (MCP, CD46) in seminal plasma and on spermatozoa in normal and "sterile" subjects

A sperm protein of molecular mass 43 kDa (the spermatozoa membrane cofactor protein, smMCP) and a seminal plasma protein of 60 kDa (ssMCP) were identified by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) followed by immunoblotting with four monoclonal antibodies (mAb) against membrane cofactor protein (MCP, CD46). These proteins served as factor I cofactors for the cleavage of methylamine-treated C3 (C3ma), the activity of which was blocked by M75, an MCP cofactor-activity-blocking mAb. Thus, these semen proteins are antigenic and functional homologous of MCP. On SDS-PAGE analysis these MCP migrated as single-band proteins which differed from the two-band forms of MCP expressed on other cells. smMCP was N-glycosylated but not O-glycosylated, while ssMCP was O-glycosylated: after deglycosylation of these proteins bands were detected at 38-40 kDa and 43 kDa on SDS-PAGE, respectively. These semen MCP are therefore, structurally different from the conventional MCP. ssMCP in both normal and "sterile" subject groups was determined by sandwich enzyme-linked immunosorbent assay. Seminal plasma in the two groups contained 250-700 ng/ml ssMCP. The difference between the two groups was marginal, although samples from normal subjects tended to show higher concentrations of ssMCP than samples from "sterile" subjects. No molecular difference was observed with ssMCP and smMCP in the two groups by SDS-PAGE/immunoblotting analysis. Immunohistochemical analysis suggested that MCP was positive in glandular epithelial cells and the lumen of the prostate, and in most intra-lumen cells of the testis. Using antibody M177, solubilized prostate and testis were analyzed by immunoblotting and compared with other cell MCP. The major band of MCP in the testis, but not in the prostate, was of 60 kDa, which aligned with ssMCP. No band of testis or prostate MCP, however, aligned with smMCP. ssMCP may be produced in the testis, while the origin of smMCP remains unknown. We hypothesize that ssMCP is important in the survival of spermatozoa, protecting them against local secretion of immunoglobulin and complement in the female genital tract, and that smMCP, which is expressed on acrosome-reacted spermatozoa, plays an essential role in the interaction of spermatozoa with oocytes.

Immune recognition at the maternal-fetal interface: overview

Trophoblast antigens at the maternal-fetal interface that are capable of stimulating maternal immune responses have been studied. Candidates are blood group I and P, HLA, Fc gamma-receptors, TLX, and phospholipids. Antigens I and P on trophoblast have been implicated in pregnancy loss but incompatible i,p mothers are rare. HLA-G is expressed on cytotrophoblast; however, no evidence for HLA-G allotypy or maternal responses to these molecules exists, although HLA-G has been implicated in recruitment of suppressor T cells. Receptors for IgG (Fc gamma-RI, Fc gamma-RII and Fc gamma-III) are present on trophoblast but allotypy is limited to the NA1-NA2 antigen system associated with Fc gamma-RIII on neutrophils. Maternal Fc-gamma R blocking antibodies have been linked to pregnancy success. The TLX alloantigen system was described by using xenogeneic antisera. Idiotype-antiidiotype regulated maternal responses to TLX are proposed as necessary for successful pregnancy. Several putative TLX monoclonal antibodies (Mab) recognize a regulator of complement activation called MCP (membrane cofactor protein, or CD46). Mab to MCP do not exhibit allotypy. Syncytial and cytotrophoblastic membranes are rich sources of MCP. Preliminary data suggest that a conformational site induced by C3b (iC3) binding to MCP may be responsible for TLX allotypy. Certain pregnancy loss patients produce antiphospholipid antibodies (aPA). Some investigators believe that aPA recognize a plasma protein cofactor, beta 2 GPI and not phospholipid per se. We produced three Mab specific for beta 2 GPI, one of which fails to recognize beta 2 GPI bound to phospholipid [corrected].(ABSTRACT TRUNCATED AT 250 WORDS)

Tissue-specific and allelic expression of the complement regulator CD46 is controlled by alternative splicing

CD46 (membrane cofactor protein) is a human cell surface glycoprotein with cofactor activity for factor I-mediated cleavage of complement components C3b and C4b. The CD46 protein from normal lymphocytes resolves on sodium dodecyl sulfate-polyacrylamide gel electrophoresis as two major bands of 66 and 56 kDa. CD46 cDNA encodes four extracellular short consensus repeat domains, a Ser/Thr/Pro (STP)-rich region, a transmembrane region and a cytoplasmic tail. We now show that exquisite control of mRNA splicing is responsible for the heterogeneous expression of CD46 isoforms. Differential splicing of 5 exons generates at least 14 CD46 mRNA variants whose expression is stringently regulated by allelic, tissue-specific and malignancy-related factors, as: (a) leukemic cells and Epstein-Barr virus-transformed B cells preferentially incorporate the first of three STP exons (exon 7) into mRNA, and produce a larger CD46 isoform of 74 kDa, (b) an allelic difference in the proportion of 66- and 56-kDa CD46 isoforms on lymphocytes corresponds to the preferential inclusion or exclusion of the second STP exon (exon 8), (c) the third STP exon (exon 9) is specifically deleted in some placentae, (d) spermatozoa delete both exons 12 and 13, encoding a shorter transmembrane region and a unique cytoplasmic tail and (e) all tissues tested differentially splice exon 13, resulting in two alternative cytoplasmic tails. The distribution of the 14 alternatively spliced RNA transcripts correlated with the presence of protein isoforms of the predicted size, indicating that alternative splicing leads to heterogeneity of CD46 glycoproteins.

Strong associations between RFLP and protein polymorphisms for CD46

Human CD46 (membrane cofactor protein) is a cell surface glycoprotein with cofactor activity for the factor I mediated cleavage of components C3b and C4b. Using a CD46 cDNA clone, three restriction enzymes give simple two allele restriction fragment length polymorphisms (RFLPs) in samples of over 300 Caucasians. For Pvu II, P1 with a 16.5 kilobase (kb) fragment and P2 with 14.8 kb + 1.9 kb fragments have frequencies of .40 and .60. For Hin dIII, H1 with a 4.3 kb fragment and H2 with a 2.3 kb fragment have similar frequencies. For Bgl. II, B1 with a 10 kb fragment and B2 with 8.3 kb + 1.8 kb fragments have frequencies of 0.08 and 0.92. There is strong linkage disequilibrium between these polymorphic sites. Designating haplotypes by Hin dIII, Pvu II, Bgl II alleles, there are two common haplotypes P2, H2, B2 and P1, H1, B2, expected at frequencies of .6 and .32, one less common haplotype P1, H1, B1 expected at a frequency .08. The two major protein isoforms of CD46, as detected on peripheral blood lymphocytes by western blot, of Mr 66,000 (alpha) and 56,000 (beta) are determined by differential splicing in production of the mRNA. A strong association between protein isoform and RFLP haplotypes in 30 unrelated subjects suggests that the splicing preference site is in linkage disequilibrium with the RFLPs. The results are consistent with haplotypes P2, H2, B2 and P1, H1, B1 producing predominantly alpha; P1, H1, B2 producing predominantly beta in about 72% of cases and alpha in 28% of cases.

Human astrocytes express membrane cofactor protein (CD46), a regulator of complement activation

Expression of membrane cofactor protein (CD46) on cultured human astrocytes was demonstrated by indirect immunofluorescence microscopy and flow cytometry following staining with a monoclonal antibody specific for CD46. Western transfer and immunoblotting detected a doublet of Mr 66,000 and 56,000. Analysis of astrocyte mRNA revealed the presence of multiple alternatively spliced transcripts encoding different extracellular regions or cytoplasmic tails of CD46. Astrocytes were also shown to express decay accelerating factor, but not the type 1 complement receptor. Upregulation of astrocyte CD46 occurred following cytomegalovirus infection. These results indicate that astrocytes express proteins involved in regulation of complement activation and protection against autologous complement.

Alternatively spliced RNAs encode several isoforms of CD46 (MCP), a regulator of complement activation

Five alternative cDNA clones were isolated for CD46, also known as the membrane cofactor protein (MCP) for the factor I-mediated cleavage of the complement convertases. One of these cDNA clones (a) was identical to an earlier MCP clone. The other four CD46 clones contained the four NH2-terminal short consensus repeat (SCR) units of MCP, but differed at the region encoding the carboxyl-terminal of the protein which includes an extracellular segment rich in Ser, Thr, and Pro residues, a hydrophobic membrane-spanning domain, and a 33 amino acid cytoplasmic tail. The different CD46 cDNAs have variously: (b) inserted a 93 base pair (bp) exon resulting in a new cytoplasmic tail of 26 amino acids; (c) deleted a 42 bp exon from the extracellular Ser/Thr rich region: (d) used a cryptic splice acceptor sequence to delete 37 bp from an exon encoding transmembrane sequence; or (e) failed to splice the intron after the four SCR units. These were shown by northern blot and polymerase chain reaction to arise by alternative splicing of CD46 RNA. Forms (a), (b), and (c) of CD46 RNA are common in placental RNA, but (d) was rare, and (e) was incompletely processed and therefore aberrant. The polymerase chain reaction (PCR) was used to map the sites of the intron/exon junctions and demonstrate further possible splice variants of CD46. The alternative RNAs for CD46 may correlate to the different isoforms of CD46 found in different tissues, tumors, and in serum.

Complement alternative pathway activation and control on membranes of human lymphoid B cell lines

Membrane regulatory molecules normally prevent complement activation by autologous cells, therefore we compared the membrane control system of human lymphoid cell lines which activate or not human complement through the alternative pathway (AP). Membrane expression of decay-accelerating factor (DAF), membrane cofactor protein (MCP), complement receptors (CR)1, CR2 and H was measured either by radioimmunoassay or enzyme-linked immunosorbent assay on cell lysates. Soluble extracts of isolated membranes were tested functionally for their ability to accelerate the decay of C3bBb C3-convertase and allow the cleavage of C3b by factor I. Both regulatory functions were detected in solubilized membranes of Ramos cells, which do not activate the AP, as well as on the potent AP activator, Raji. Raji cells were found to express CR2, DAF and MCP molecules, while MCP was the only known regulatory protein detected on Ramos cells which expressed neither CR1, nor CR2, H or DAF. The I-cofactor activity of both Raji and Ramos cells was immunoprecipitated by anti-MCP, but the decay-accelerating activity was not adsorbed by anti-DAF nor by any of the available antibodies. Two EBV genome-negative cell lines (BJAB, BL41) were tested before and after in vitro conversion by EBV. As previously shown, EBV-converted cell lines activate the AP more efficiently than EBV- cell lines. At the same time, EBV superinfection induces an increase of both AP regulatory functions of cell membranes and enhances the expression of DAF, MCP and CR2. The results of this study show that complement activation by lymphoid cell lines is not related to an impaired autologous control of these cells, but that the expression of regulatory molecules increases together with the appearance of activating structures on the cell surface. Our results also suggest the occurrence of a new factor involved in the decay-accelerating activity on BL lines.

Human non-lineage antigen, CD46 (HuLy-m5): purification and partial sequencing demonstrates structural homology with complement-regulating glycoproteins

CD46, until recently known as HuLy-m5, is a non-lineage restricted surface antigen ubiquitously expressed by almost all human cells except erythrocytes. The CD46 antigen is identified by the E4.3 monoclonal antibody (mAb) and exists at the surface of human peripheral blood lymphocytes (PBLs) as two acidic, non-disulfide bonded chains, alpha and beta, of Mr 66,000 and 56,000. Receptor density analysis showed that CD46 was of moderately low abundance on PBLs with 7.5 x 10(3) molecules present on each cell. The two chains of CD46 were purified (144,000-fold) by immunoaffinity-chromatography with E4.3 mAb from the plasma membranes of a human spleen infiltrated with chronic myelogenous leukemia cells. Amino acid sequence analysis of the NH2-terminal of both alpha and beta chains yielded the same sequence; XEEPPQ/TFEAMELIGKPKPYYEIGE. Peptide mapping studies confirmed that both CD46 chains were closely related, except for one peptide fragment. This amino acid sequence is identical to that of the NH2-terminal of the recently cloned membrane co-factor protein (MCP), a membrane protein that binds the C3b and C4b fragments of complement and acts as a co-factor for I protein-mediated decay of the complement convertases. CD46 shares a cross-reactive epitope with some primate retroviruses, and this may indicate that some retroviruses mimic the mechanisms used by autologous human cells to evade complement-mediated immune clearance.

The human non-lineage antigen CD46 (HuLy-m5) and primate retroviral gp70 molecules share protein-defined antigenic determinants

The CD46 lymphocyte surface antigen of man (until recently called HuLy-m5), and defined by the E4.3 monoclonal antibody (MoAb), shares cross-reactive antigenic epitopes with the envelope gp70 glycoproteins of gibbon ape leukaemia virus (GaLV) and Mason Pfizer monkey virus (MPMV) primate retroviruses. It is now shown that the cross-reactive antigenic epitope shared by these three molecules is determined solely by the protein portion of these glycoproteins, and that the N-linked and O-linked carbohydrate moieties of these glycoproteins do not directly or sterically contribute to the antigenic cross-reactivity. When CD46 molecules (mol.wt = 66 and 56 kDa) from human thymocytes were stripped of sialic acid with neuraminidase, or stripped of N-linked carbohydrate with endoglycosidase F, the E4.3 MoAb was still able to bind and immunoprecipitate the protein core of CD46 (mol.wt = 56 and 44 kDa). Similarly, polyclonal antisera to GaLV and MPMV precipitated deglycosylated CD46, although at a reduced efficiency. The cross-reacting E4.3 MoAb, anti-GaLV and anti-MPMV antisera also immunoprecipitated HuLy-m5 primary translation protein lacking N- or O-linked carbohydrate from the in vitro translation products of human thymocyte mRNA. Thus, the antigenic cross-reactivity of CD46 molecules with GaLV gp70 and MPMV gp70 is both specific and due to protein structure rather than to carbohydrate; the findings suggest that retroviruses may have acquired a functional epitope from human CD46 or that an endogenous retroviral sequence of human may partially or completely encode the CD46 antigen.

Analysis of circulating immune complexes from patients with ankylosing spondylitis by gel electrophoresis and immunoblotting using antiserum against a psoriasis associated retrovirus-like particle

Circulating immune complexes (CIC) were isolated from patients with ankylosing spondylitis (AS) and healthy blood donors by isopycnic ultracentrifugation in sucrose gradients. The CIC were analysed by sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblotting. The major components of the CIC were identified as albumin, immunoglobulins, and complement factors. A 70 kD component and several low molecular weight components (Mr 19 kD and 14 kD (doublet] were detectable only in CIC from patients with AS. An antiserum raised against the envelope glycoprotein, gp70, of a psoriasis associated retrovirus-like particle was applied to check for cross reacting activity. This antiserum reacted with both a 70 kD and a 40-45 kD component in CIC from three out of six patients but not with CIC from any of the blood donors.