Anti-CD81 Antibody [M38] (A86719)

BACKGROUND:

Hepatitis C virus (HCV) could induce chronic liver diseases and hepatocellular carcinoma in human. The use of primary human hepatocyte as a viral host is restrained with the scarcity of tissue supply. A culture model restricted to HCV genotype 2a (JFH-1) has been established using Huh7-derived hepatocyte. Other genotypes including the wild-type virus could not propagate in Huh7, Huh7.5 and Huh7.5.1 cells.

METHODS:

Functional hepatocyte-like cells (HLCs) were developed from normal human iPS cells as a host for HCV infection. Mature HLCs were identified for selective hepatocyte markers, CYP450s, HCV associated receptors and HCV essential host factors. HLCs were either transfected with JFH-1 HCV RNA or infected with HCV particles derived from patient serum. The enhancing effect of α-tocopherol and the inhibitory effects of INF-α, ribavirin and sofosbuvir to HCV infection were studied. The HCV viral load and HCV RNA were assayed for the infection efficiency.

RESULTS:

The fully-developed HLCs expressed phase I, II, and III drug-metabolizing enzymes, HCV associated receptors (claudin-1, occludin, CD81, ApoE, ApoB, LDL-R) and HCV essential host factors (miR-122 and SEC14L2) comparable to the primary human hepatocyte. SEC14L2, an α-tocopherol transfer protein, was expressed in HLCs, but not in Huh7 cell, had been implicated in effective HCVser infection. The HLCs permitted not only the replication of HCV RNA, but also the production of HCV particles (HCVcc) released to the culture media. HLCs drove higher propagation of HCVcc derived from JFH-1 than did the classical host Huh7 cells. HLCs infected with either JFH-1 or wild-type HCV expressed HCV core antigen, NS5A, NS5B, NS3 and HCV negative-stand RNA. HLCs allowed entire HCV life cycle derived from either JFH-1, HCVcc or wild-type HCV (genotype 1a, 1b, 3a, 3b, 6f and 6n). Further increasing the HCVser infection in HLCs was achieved by incubating cell with α-tocopherol. The supernatant from infected HLCs could infect both naïve HLC and Huh7 cell. Treating infected HLC with INF-α and ribavirin decreased HCV RNA in both the cellular fraction and the culture medium. The HLCs reacted to HCVcc or wild-type HCV infection by upregulating TNF-α, IL-28B and IL-29.

CONCLUSIONS:

This robust cell culture model for serum-derived HCV using HLCs as host cells provides a remarkable system for investigating HCV life cycle, HCV-associated hepatocellular carcinoma development and the screening for new anti HCV drugs.

Association of major histocompatibility complex (MHC) class II molecules with peptides occurs in a series of endocytic vacuoles, termed MHC class II-enriched compartments (MIICs). Morphological criteria have defined several types of MIICs, including multivesicular MIICs, which are composed of 50-60-nm vesicles surrounded by a limiting membrane. Multivesicular MIICs can fuse with the plasma membrane, thereby releasing their internal vesicles into the extracellular space. The externalized vesicles, termed exosomes, carry MHC class II and can stimulate T-cells in vitro. In this study, we show that exosomes are enriched in the co-stimulatory molecule CD86 and in several tetraspan proteins, including CD37, CD53, CD63, CD81, and CD82. Interestingly, subcellular localization of these molecules revealed that they were concentrated on the internal membranes of multivesicular MIICs. In contrast to the tetraspans, other membrane proteins of MIICs, such as HLA-DM, Lamp-1, and Lamp-2, were mainly localized to the limiting membrane and were hardly detectable on the internal membranes of MIICs nor on exosomes. Because internal vesicles of multivesicular MIICs are thought to originate from inward budding of the limiting membrane, the differential distribution of membrane proteins on the internal and limiting membranes of MIICs has to be driven by active protein sorting.

C33 Ag and M38 Ag had been identified by mAb inhibitory to HTLV-1-induced syncytium formation. The cDNA encoding C33 Ag had revealed that it belongs to the newly defined transmembrane 4 superfamily (TM4SF). M38 Ag was detected on virtually all human cell lines and fresh leukocytes except for most granulocytes. It was also expressed on a mouse hybrid cell clone containing human chromosome 11q23-pter. Immunoprecipitation and immunoblot analyses identified a monomeric 26-kDa protein. The M38 epitope was dependent on S-S bonding. These characteristics were very similar to those reported for TAPA-1 (the target of antiproliferative antibody-1), which also belongs to TM4SF as C33 Ag. We therefore cloned the cDNA of human TAPA-1 and expressed it in COS cells. M38 indeed reacted with COS cells expressing human TAPA-1. We concluded that M38 Ag was identical to TAPA-1. To further investigate the biologic functions of C33 Ag and M38 Ag (TAPA-1) and their roles in HTLV-1-induced syncytium formation, molecules associated with these Ag were examined in T cells. Immunoprecipitation from surface-iodinated cell lysates revealed that proteins co-precipitated by C33 and M38 were mostly common including each other. Sequential immunoprecipitation-immunoblot experiments confirmed that C33 Ag and M38 Ag (TAPA-1) were associated with each other. The association was further confirmed in BHK cells doubly transfected with human cDNA for C33 Ag and TAPA-1. We extended similar analyses and found that C33 Ag and M38 Ag (TAPA-1) were regularly associated with CD4 or CD8. The association of these Ag on the cell surface was further supported by co-modulation of M38 Ag (TAPA-1), CD4 and CD8 with C33 Ag. This is the first time that a physical association between the members of TM4SF is demonstrated. Furthermore, the regular association of C33 Ag and M38 Ag (TAPA-1) with CD4 or CD8 might indicate that they play a role in expression and/or function of the CD4/CD8 co-receptor complex.

Previously, we have shown that CD81 and CD82, two members of the transmembrane 4 superfamily, form multimolecular membrane complexes by associating with each other and with CD4 or CD8 in T cells. In the present study, we further analyzed the molecular basis of the CD4 association with CD81 and CD82 by co-precipitation experiments. First, we examined the regions of CD4 involved in the association with CD81 and CD82 by employing chimeric proteins generated from CD4 and CD2. It was confirmed that CD4, but not CD2, was capable of binding with CD81 and CD82 in transfected cells. We found that the cytoplasmic region of CD4 was sufficient for the chimeric proteins to co-precipitate CD81, while both the cytoplasmic and extracellular regions of CD4 were required for them to efficiently co-precipitate CD82. We next found, by using truncated CD4 lacking the C-terminal 31 amino acids or mutated CD4 with the cysteine residues at 394 and 397 replaced by serine, that the p56lck binding site or the covalent modification with palmitic acid was not necessary for CD4 to associate with CD81 and CD82. Finally, we found that the binding of p56lck to CD4 strongly inhibited its association with CD81 and CD82. It is, therefore, suggested that CD4 exists at least in two physical states, one associated with p56lck and another associated with CD81 and CD82 in the absence or uncoupling of p56lck.

We isolated four monoclonal antibodies (MAbs), M38, M101, M104, and C33, which were capable of inhibiting syncytium formation induced in a human T-cell line, MOLT-4-#8, by coculture with human T-cell leukemia virus type 1 (HTLV-1)-positive human T-cell lines. The MAbs had, however, no inhibitory activity on syncytium formation induced in a human osteosarcoma line, HOS, by HTLV-1-positive T-cell lines. They also did not inhibit syncytium formation induced in MOLT-4-#8 by human immunodeficiency virus type 1-positive MOLT-4. All MAbs reacted with various human cell lines of lymphoid and nonlymphoid origins, including HTLV-1-positive T-cell lines. Furthermore, they all reacted with a murine A9 clone containing human chromosome 11 fragment q23-pter. Two MAbs, M104 and C33, immunoprecipitated a membrane antigen with the same molecular size. The antigen (henceforth called C33 antigen) was about 40 to 55 kDa in HTLV-1-negative Jurkat, CEM, MOLT-4, and normal peripheral blood CD4-positive human T cells and about 40 to 75 kDa in HTLV-1-positive C91/PL, TCL-Kan, MT-2, and in fresh HTLV-1-transformed CD4-positive human T-cell lines. Pulse-chase experiments revealed that C33 antigen was synthesized as a 35-kDa precursor that was then processed to 41 to 50 kDa in MOLT-4 and to 44 to 70 kDa in C91/PL. In the presence of tunicamycin, a 28-kDa protein was synthesized. The conversion from 35 kDa to 41 to 50 kDa in MOLT-4 and to 44 to 70 kDa in C91/PL was inhibited by monensin. Treatment with N-glycanase alone, but not with sialidase and O-glycanase in combination, completely removed the sugar moiety of C33 antigen from both HTLV-1-negative Jurkat and HTLV-1-positive C91/PL. Therefore, C33 antigen has only N-linked carbohydrates, the modification of which appears to be substantially altered in the presence of the HTLV-1 genome.