This thesis studied aspects of the interaction of hepatitis B virus (HBV) with haemopoietic cells and cell lines, to address the reported tropism of HBV for haemopoietic tissues. Emphasis was directed at demonstrating specific attachment of HBV to defined sub- populations of peripheral blood leucocytes (PBL) and bone malrow cells (BM), and the distribution of receptors for HBV on well-defined haemopoietic cell lines. Biochemical characterisation of the virus-cell interaction was also performed, and the question of infectivity of haemopoietic cell lines was addressed.

Firstly, a quantitative assay of HBV binding to liver plasma membranes (PM) adapted was to show that isolated PBL PM bound serum-derived FIBV particles to a similar degree, based on their protein content. Using synthetic peptides representative of various amino acid sequences of the preSl and preS2 regions of L HBsAg to inhibit HBV binding to the pM, it was found that peptide pres l(12-32) inhibited binding to PBL PM by 6O-80Vo and peptide preSl(21-47) inhibitedby O-30Vo (depending on the source of PM), while peptides preSl(32-49) and preS2(120-145) did not inhibit binding. This contrasts with results obtained using liverPM, where peptide preSl(12-32) did not inhibitbinding, while peptide preS1(21- 47) inhibited by '|OVo, andpreSl(32-49) inhibited by approximately IZVo. Peptide preS2(120- 145) had no effect on binding. Thus, different regions of the L surface protein appear to mediate attachment to PBL and hepatocytes.

HBV particles isolated from serum are complexed with serum proteins including IgG' To test the involvement of receptors for IgG and complement fragments (opsonins) in the HBV- PM interaction, a panel of ligand-blocking monoclonal antibodies (MAbs) to opsonin receptors was used, and it was shown that the three classes of receptors for IgG (Fc1RI, Fc^yRII and Fc1RIII) and CR3, are not major receptors for HBV on PBL or hepatocytes, as MAbs to these did not inhibit HBV binding. It was also shown that HBV does not utilise the receptor for IgA, FccrR, for attachment to PBL, despite reported sequence homology between the large envelope protein of HBV and the Fc portion of human IgA. In contrast to a published reportthat IL-6 mediates binding of HBV to hepatocytes, IL-6 was shown not to mediate attachment to either liver or PBL PM, by virtue of pre-incubation with a blocking polyclonal anti-serum to IL-6.

Glycosaminoglycans (GAGs) were found to influence HBV binding to PM:​ soluble heparin (HE) inhibited binding to liver PM by up to 807o, and to leucocyte PM by up to 40Vo;​ chondroitin sulphare C (CS-C) enhanced virus binding (approximately 1.5-fold) to leucocyte pM only. Chondroitin sulphate A and hyaluronate had no effect on binding to either PM, arguing that simple electrostatic properties of GAGs were not responsible for the observed effecrs. The incomplete inhibition by HE and enhancement by CS-C could indicate the presence of more than one class of binding site for HBV on the respective PM, and coupled with the differential pattern of inhibition in the presence of synthetic peptides, argues that receptors for HBV on pBL and hepatocytes may be either different, or altered forms of the same molecule(s).

To extend these studies, whole cell binding assays were developed in order to accurately define which subsets of pBL and BM cells could bind HBV. Using purified HBV particles as the first stage in an immunofluorescence-based detection system, followed by detection of bound HBV using anti-preS1 MAbs F35.25 or MAl8/7, and a FlTC-conjugated third-stage antibody, specific membrane staining of peripheral blood monocytes from 8/9 donors was observed. In addition, binding of HBV to the erythroleukaemia cell line K562 was observed, while other myeloid cell lines did not appear to bind virus. This assay was then adapted to a suspension cell assay with analysis by flow cytometry, using phycoerythrin as the detecting fluorochrome. The parameters of binding were optimised for K562 cells and these were then applied to analyse HBV binding to PBL and BM cells obtained from healthy volunteers, whose serum was free of HBV markers. Based on their light scattel characteristics, monocytes and neutrophils were the only cell types in the peripheral blood that bound HBV' while binding to lymphocytes was not observed. This was confirmed by two-colour immunofluorescence to simultaneously detect bound HBV and subset-specific leucocytestem markers. Similarly, in the BM, only monocytes bound HBV. Importantly, haemopoietic cells (cD34+) did not bind HBV. Binding was tested to 'activated' populations of (fMlP-treated). The lymphocytes (pHA-treated), monocytes (LPS-treated), and neutrophils pattern of HBV binding was not affected by these treatments. Monocytes cultured in vitro, bound significantly more virus than freshly isolated monocytes. Taken together, these results potential receptors for indicate that only monocytes, and to a lesser extent neutrophils, express HBV, and a differentiation-dependent upregulation of receptor sites for HBV is observed on monocytes.

The distribution of potential HBV receptors was determined on a number of cell lines, representative of various haemopoietic lineages. K562 (erythroid), and the bound HBV, while monocyte cell line THP-1, were the only haemopoietic cell lines which of other binding was also observed to the human hepatoma cell line HepG2. A number erythroid and monocyte cell lines, as well as T and B tymphoid, and a megakaryocytic line, all failed to bind HBv. A comparison of the surface immunophenotypes of all the cell lines tested excluded all known CD-classified molecules (including opsonin receptors), as candidate FIBV recePptors.

The biochemical characteristics of the interaction of HBV with all of these then examined. on K562 and rHp-l, HBV binding chymopapain but insensitive was sensitive to the protease to trypsin, indicating that the molecule was a glycosylated protein. pre-treatment of these cell lines with tunicamycin, to inhibit post-translational HBV binding, addition of N-linked carbohydrate to surface glycoproteins, did not influence removal of indicating that these moieties are not important for virus attachment' Enzymatic binding to K562 and cell surface sialic acids with neuraminidase significantly enhanced HBV of HBV to THp-1 cells but did not confer binding to otherwise 'negative' cell lines. Binding and was not cultured monocytes and HepG2 cells was trypsin and chymopapain sensitive' that HBV binding increased by neuraminidase pre-treatment. Cation chelation demonstrated that binding of to all cell types was ca2*nvlg'*-independent, and acid elution of cells showedwas not mediated by peripherally-bound molecules. Binding of HBV to monocytes and to HepG2 cells was significantly reduced by pre-treatment of the cells with PI-PLC, implying that the molecule responsible for binding to these cells is GPl-linked. In this case, a comparison with HBV binding to K562 was not informative due to the resistance of the GPI linkages on these cells, and possibly also on THP-I cells (based on CD59 cleavage), to hydrolysis by PI-PLC. Thus, cells expressing potential receptor(s) for HBV, whose characteristics do not correlate with any other proposed candidate, have been identified.

Immunoprecipitation analysis using HBV particles covalently cross-linked to the surface of ¹²⁵I-labelledK562 cells, and anti-S MAb coupled to goat anti-mouse lgG-Sepharose, resulted in the identification of a 50 kDa species as a putative HBV receptor.

Contrary to reports that HBV inhibits colony formation by myeloid cell lines in semi-solid media, no inhibitory effect by HBV was seen on clonal cell growth in liquid culture. K562 cells were found to be able to efficiently internalise HBV particles, which accumulated in a perinuclear compartment. In infection studies, K562 cells positive for HBsAg aftet 2-4 days post-infection became enlarged, and their numbers decreased steadily over an 11 day period. It is not clear whether these cells represent appears that the level of a transient or differentiated cell type. Similarly, it HBV DNA in these cells declines steadily during the infection course.