Adenoviruses belong to the most common human pathogens. The severity of infection varies greatly, from subclinical to lethal, depending on the virus type and immune status of the infected host. The 51 known human adenovirus serotypes are divided into six species (A-F) based on characteristics such as tropism. Species B adenoviruses, which are the subjects of this thesis, are further divided into subspecies B:1 that contains Ad3, Ad7, Ad16, Ad21 and Ad50 and subspecies B:2 that contains Ad11, Ad14, Ad34 and Ad35. Species B adenoviruses primarily cause ocular and respiratory tract infections, but certain serotypes (Ad11, Ad34 and Ad35) are also associated with renal disease. The main aim of this thesis was to identify and characterize cellular receptors for species B adenoviruses. This will ultimately help to understand the diverse tropism shown by different adenoviruses and perhaps contribute to development of antivirals. Also, since adenoviruses are among the most commonly used vector for gene therapy it is of importance to characterize the initial steps of adenovirus life cycle. Members of species B adenoviruses have been shown to utilize both the complement regulating membrane cofactor protein (MCP), i.e. CD46, and a still unknown receptor. CD80 and CD86, usually found on antigen-presenting cells, have also been suggested as receptors We found first that Ad11 used CD46 as a cellular receptor on respiratory A549 cells, and subsequently that CD46 is a cellular receptor for all species B adenovirus serotypes, except for adenovirus types 3 and 7, using cells that represent the tropism of species B adenoviruses, i.e. respiratory, conjunctival and renal epithelial cells. We further compared the relative roles of CD46 with CD80 and CD86 using cells that represent species B adenovirus tropism. Using soluble candidate receptors and antibodies against corresponding receptors to challenge virus binding to and infection of cells, we found that on these cells, CD46 is a cellular receptor for all species B adenoviruses except Ad3 and Ad7, and that CD80 and CD86 do not play an important role. We have further pinpointed the interaction site for Ad11 on CD46 by X-ray crystallography. The extracellular region of CD46 contains four short consensus repeats (SCR1-4) of which the outermost N-terminal SCR1 and SCR2 mediate binding to Ad11. This interaction was confirmed by inhibiting infection and binding of Ad11 to A549 cells using soluble SCR1-2 fragments. Surprisingly the conformation of bound CD46 differs profoundly from its unbound state, with the bent surface structure straightened into an elongated rod. Viral proteins can sometimes undergo large conformational changes upon receptor binding, but this is, to the best of our knowledge, the first example of a virus protein dramatically changing the overall structure of its receptor. CD46 serves as a receptor for a large number of viral and bacterial pathogens and it is structurally and functionally related to other viral receptors such as CD21 and CD55. The mode of interaction presented here may serve as a conceptual framework for studies of many other receptors that are constructed from SCR domains.

At present, the human adenoviruses (Ads) comprise 51 members, which have been classified into six species (A to F). In general, adenovirus (Ad) tissue tropism or disease patterns vary according to species, although adenoviruses from different species can sometimes cause the same symptoms. The current interest in adenoviruses is partly due to the aim of using them as vectors for gene therapy. Hematopoietic cells are attractive targets for gene therapy and the transductions can be performed ex vivo. However, the most commonly used adenovirus vectors, based on Ad2 or Ad5, are inefficient in their transduction of hematopoietic cells since they attach poorly to these cells. Most Ads, including Ad2 and Ad5, appear to use the coxsackie-adenovirus receptor (CAR) (a component of tight junctions), for attachment to host cells. However, species B Ads do not bind to CAR and several studies have indicated that species B-based vectors would be more suitable for hematopoietic cells. Species B Ads can be further divided into species B1 and B2, which display different tissue tropisms. Species B1 Ads mostly cause acute respiratory infections whereas species B2 Ads have been associated with persistent infections of the kidney and urinary tract. One of the key determinants of tropism is believed to be the initial high-affinity attachment of the virion to host cell fiber receptors. By reciprocal blocking experiments and different ways of characterizing the species B attachment receptors, we have shown that the species B2 serotypes Ad11p and Ad35 and the species B1 serotypes Ad3p and Ad7p also differ in receptor usage. There are at least two different Ad species B receptors. Since one of these receptors appeared to be used by all four serotypes, we designated this receptor sBAR (species B adenovirus receptor). The other receptor appeared to be used exclusively by the two species B2 serotypes and was therefore designated sB2AR (species B2 adenovirus receptor). Binding to sBAR can be abolished by EDTA and restored with Mn2+ or Ca2+, whereas binding (of Ad11p and Ad35) to sB2AR is independent of divalent cations. Furthermore, sBAR appears to be trypsin sensitive whereas sB2AR is not. We also identified CD46 as a receptor for Ad11p. Even so, CD46 does not appear to be a functional receptor for Ad7p. Both Ad7p and Ad11p attached to CD46-transfected Chinese hamster ovary (CHO) cells more efficiently than to control CHO cells. However, only Ad11p (selectively) infected CD46-transfected CHO cells. Anti-CD46 antibodies inhibited Ad7p and Ad11p from binding to, and Ad11p from infecting, CD46-transfected CHO cells. However, in human cells, anti-CD46 antibodies had an inhibitory effect only on Ad11p binding (~30%) but did not affect Ad7p binding. In binding experiments with EDTA, divalent cations and pretrypsinized cells, Ad11p and Ad7p showed the same pattern in their binding to CHO-CD46 cells as in the previous study. Since Ad7p interacted almost as efficiently with control CHO cells as with CHO-CD46 cells after addition of Mn2+, it seems that Ad7p mainly addressed an endogenously expressed hamster receptor on CHO-CD46, the properties of which resemble sBAR. In addition, Ad3p and Ad7p attach poorly to PBMCs and CD46 is expressed on all nucleated cells. Thus, CD46 appears to correspond to sB2AR rather than to sBAR. With these differences in receptor usage in mind, we studied the binding and infectious capacity of these species B Ads in various hematopoietic cells. We found that all species B serotypes bound efficiently to CD34+ hematopoietic stem cells (HSCs) and also productively infected HSCs. However, only the sB2AR binding Ad serotypes Ad11p and Ad35 could attach primary PBMCs efficiently. Our results regarding the subsequent steps in infection of PBMCs suggest that both Ad11p and Ad35 enter PBMCs and deliver viral DNA to the nuclei of most PBMC cell types. However, productive infections were only clearly detected in stimulated T-cells (most frequently) and monocytes, whereas Ad infection seemed eclipsed in unstimulated lymphocytes. Replication of Ad DNA seemed seriously impaired in at least T-cells, indicating limited production of infectious particles in PBMCs. The capacity of species C Ads to establish persistent infections in lymphatic tissues has been described previously. These Ads also persistently infect various transformed hematopoietic cell lines in vitro. Our studies indicate that replication of the species B2 Ads is also restricted in cells of hematopoietic origin (both in primary and transformed cells). Taken together, the results indicate that species B2 Ads (as compared to other Ads) seem to enter and infect most hematopoietic cells efficiently, which is in line with the persistent nature of these Ads. They would presumably act as suitable vectors for efficient transduction of most cells of hematopoietic origin, as has already been shown for e.g. HSCs and dendritic cells. The finding that replication of Ads in T-cells appears to depend on the level of T-cell activation, strengthens the hypothesis that T-cells may serve as a reservoir for human Ads and raises possible safety issues for usage of species B-based vectors in hematopoietic cells.