![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
Previous Article | Next Article 
Journal of Virology, December 2002, p. 13034-13038, Vol. 76, No. 24
0022-538X/02/$04.00+0 DOI: 10.1128/JVI.76.24.13034-13038.2002
Copyright © 2002, American Society for Microbiology. All Rights Reserved.
Inserm U404, Immunité et vaccination, CERVI, IFR 74, 69365 Lyon, Cedex 07, France,1 Institute for Animal Health, Pirbright Laboratory, Pirbright, Woking, Surrey GU24 0NF, United Kingdom,2 Laboratoire National de Santé, L-1011 Luxembourg, Luxembourg3
Received 19 June 2002/ Accepted 12 September 2002
 |
ABSTRACT |
|---|
 Top Abstract TextReferences |
|---|
|
TEXT |
|---|
|
TopAbstractTextReferences |
|---|
MV possesses two glycoproteins in its envelope: the hemagglutinin (H) protein (MVH) and the fusion protein (MVF). MVH is responsible for attachment to the cellular receptor, whereas MVF mediates the fusion of the viral and host cell membranes (34). MV fusion has been shown to depend upon the coexpression of the two glycoproteins (36) It is believed that the fusion helper function of the H protein depends upon a specific physical interaction with the MVF which is mediated by the cysteine-rich domain of the latter protein (35). Nearly 10 years ago it was shown that two laboratory strains of MV use the protein CD46, a member of the regulators of complement activation superfamily, as a cellular receptor (5, 23). Although the MVH-CD46 interaction is probably conformational and many amino acids on MVH contribute, it was found that a tyrosine residue at position 481 of MVH plays a crucial role: mutation of this residue leads to abrogation of the interaction with CD46 (1, 11, 17). The observation that residue 481 is asparagine in most if not all MVH proteins from wt MV strains and the finding that anti-CD46 antibodies did not block wt MV infection led to the less than universally accepted hypothesis that wt MV use an alternative cellular receptor (2, 3, 4, 11). The ensuing controversy was resolved recently when SLAM (or CD150) was shown to act as a receptor for wt MV (6, 12, 30), but even if it is generally agreed now that laboratory strains use both CD46 and SLAM and wt strains use exclusively SLAM (7), wt MV receptor usage in vivo remains enigmatic. This is because it is difficult to relate MV pathology to the known tissue specificity of SLAM. Although SLAM is constitutively expressed on dendritic cells, certain memory T cells, and some B cells and is upregulated on activated B and T cells (32) and this corresponds to the known lymphotropism of MV, many of the tissues infected during MV infection are believed to be SLAM negative.
MV is spread by aerosols, and the initial infection is established in the trachial and bronchial epithelial cells. The infection may then spread to local lymphatic tissues, where the virus replicates before being dispersed via the blood to cause a systemic infection affecting many organs of the body (9). Although certain lymphocytic cells are positive for SLAM and the spread of the virus from the respiratory tract to the lymph nodes could be due to the migration of MV-infected dendritic cells which are also SLAM positive, trachial and bronchial epithelial cells and indeed the organs which succumb to the systemic infection are thought to be SLAM negative. Since all human cells (excepting erythrocytes) express CD46, this molecule would appear to be a better candidate as a cellular receptor for MV than SLAM, but several studies have established that wt MV strains do not appear to interact with CD46 (7, 11, 17). However, it has been shown that in vitro, wt MV strains can infect CD46+ SLAM- Vero cells, albeit with the absence of cytopathic effects (33). Furthermore, no differences were found in the primary sequence of the wt H protein following adaptation to Vero cells. In order to explain such results, the existence of a third MV receptor in addition to CD46 and SLAM has been invoked (10, 20, 33). An alternative plausible explanation is that wt MV strains enter CD46+ SLAM- cells via low-affinity interaction with CD46. Prior to the identification of SLAM as a receptor for wt MV, Manchester et al. (19) made this hypothesis following their finding that splenocytes from CD46+ transgenic mice could be infected by wt MV strains and that infection could be blocked by anti-CD46 antibodies. A possible interpretation of these various studies is that in natural infection, wt MV uses CD46 as a low-affinity receptor and SLAM as a high-affinity receptor. If wt MV strains have the capacity to bind with low affinity to CD46, we conjectured that this would be facilitated by an as-yet-unidentified domain in MVH common to both wt and vaccine strains.
In an attempt to identify such a domain, we have made chimeric H proteins between the MV Hallé laboratory strain H protein and that of the rinderpest virus (RPV) vaccine strain RBOK, which does not appear to interact with CD46. MVH has 60% homology with RPVH at the primary sequence level, and the number and location of the cysteine residues are conserved between the two proteins (15). We used CD46 downregulation as a phenotypic marker of the capacity of the chimeras to interact with CD46: whereas the Hallé H protein has the capacity to downregulate CD46, RBOKH does not. wt RPV strains have also been shown to use SLAM as a cellular receptor (31), and although CD46 downregulation has been reported to occur with RPV (8), this is not the case for the cloned H protein. According to the structural model proposed for the morbillivirus H protein (16), the globular head of the protein consists of a superbarrel in which six sheets (ß1 to ß6) are arranged cyclically around an axis like the blades of a propeller and loops protrude from the top and lower surfaces of each "blade." We found that when the primary sequence of the loop ß6L01, which protrudes from the top surface of the propeller-like globular head of the Hallé H protein, was converted to RBOKH sequence, the Hallé H protein lost the capacity to downregulate CD46. Since wt MVH proteins have an identical sequence in this domain, this could suggest that wt MV has the capacity to interact with CD46 cryptically at low affinity. An interaction of this type would provide an explanation for certain paradoxes regarding MV replication and would offer an alternative mode for wt MV cell entry.
The S548L F549S mutation in the MVHH protein results in a loss of binding to CD46. In contrast to the H protein from the Hallé laboratory strain of MV (MVHH), the H protein from the RBOK vaccine strain of rinderpest virus (RPV) does not appear to interact with CD46: when expressed from the eukaryotic expression plasmid pCopak in HeLa cells, RBOKH does not downregulate CD46 (Fig. 1A). To investigate whether MVHH harbors as-yet-unidentified domains necessary for the interaction with CD46, we constructed chimeras between the MVHH and RBOKH proteins, expressed them in HeLa cells (Fig. 1C), and tested (by flow-cytometric analysis) their capacity to downregulate CD46. We found that a chimera in which the primary sequence until residue 550 was derived from MVHH but the remainder of the C terminus was derived from RPVH retained the capacity to downregulate CD46 (Fig. 1B) and induced fusion (not shown). However, when the RPVH-derived primary sequence commenced at residue 532 (or indeed 521 or 511), CD46 downregulation was abolished (Fig. 1B) and fusion was abrogated (data not shown). This suggested that there is a domain which is important for CD46 binding between residues 532 and 550 of the MV Hallé H protein. Focusing upon this part of the H protein, we found that a chimera in which loop ß6L01 of the globular head derives from RBOKH, but with the totality of the primary sequence otherwise deriving from the MVHH protein, was also incapable of downregulating CD46 (Fig. 2A). This chimera differs from the native MVHH protein by just two amino acids: the serine at 548 is replaced by leucine, and the adjacent phenylalanine at 549 is replaced by serine (Fig. 3). When these changes are introduced singly into MVHH, the mutation S548L affects CD46 downregulation much more than F549S (compare Fig. 2B and C) but the double mutation is necessary to abrogate CD46 downregulation completely.
|
|
|
|
|
We report the identification of a novel CD46-binding domain that is present in the H proteins from both wt and laboratory strains of MV. Our rationale in searching for this domain was to throw some light on the enigma in terms of cell receptor usage that natural (wt) MV infection presents. Two cellular proteins have been identified as being receptors for MV: CD46 and SLAM (or CD150) (5, 6, 12, 23, 31). CD46 is ubiquitously expressed on all human nucleated cells, whereas the expression of SLAM appears to be limited to certain lymphocyte populations and dendritic cells. It is now generally accepted that whereas laboratory and vaccine strains of MV use both CD46 and SLAM as receptors, the wt strains use only SLAM (7). The use of SLAM as a receptor by wt strains fits well with the known lymphotropism of MV, but the initiation and termination of wt infection occurs in tissues which are believed to be SLAM negative. Such tissues are CD46 positive, but it has been shown that wt MV does not attach to CD46 unless an asparagine at position 481 in the wt MV attachment protein has been mutated to tyrosine (1, 11, 17).
The Oldstone group reported (19) that splenocytes from CD46+ transgenic mice could be infected by wt MV strains and that infection could be blocked by anti-CD46 antibodies, but it is well known that wt MV infection of human lymphocytes happens via SLAM even though they are CD46 positive too. It would appear that in the case of the murine CD46+ splenocytes, CD46 may be used as a receptor in the absence of SLAM (murine SLAM cannot be used as a receptor by MV), but in the case of the human lymphocytes, SLAM is preferred as a receptor even though CD46 is available. The interpretation that Manchester et al. make of their results is that clinical isolates of MV use CD46 as a cellular receptor at low affinity, but this makes it difficult to explain why CD46 is not used by these isolates for their infection of human lymphocytes. A possible interpretation which reconciles these observations is that wt MV uses CD46 as a low-affinity receptor and SLAM as a high-affinity receptor. Even if wt MV uses SLAM to infect human lymphocytes, this does not exclude the possibility that the initial contact made with these cells is via low-affinity binding to CD46. If wt MV uses both CD46 and SLAM as receptors, it could bind with low affinity to (abundant) CD46, insufficiently to trigger fusion, and then browse the host cell surface for (sparse) SLAM, to which it could bind with high affinity and set in motion the conformational changes in the H protein leading to fusion. Interestingly, Yanagi's group has shown that monocytes upregulate SLAM in the presence of wt MV, which is then used by the virus for entry (22). This introduces the possibility that if a cell does not express SLAM, low-affinity binding by wt MV to CD46 could be the means—via signaling—of upregulating this protein.
An alternative explanation that has been proposed for the mechanism of wt MV infection is the existence of a third receptor for MV (10, 20, 33). If a third MV receptor exists, it would have to be present on the respiratory epithelium and many tissues of the human body as well as on Vero monkey kidney epithelial cells (CD46+ SLAM-), which are infectible by wt MV in vitro. Infection of Vero cells by wt MV is silent, i.e., without cytopathic effects such as viral fusion, but wt MV infection of VeroSLAM cells (CD46+ SLAM+) leads to syncytium formation (33). Thus, if a third receptor for MV exists, its use does not lead to syncytium formation in vitro. Interestingly, a recent publication from the Yanagi group (10) reports that a recombinant wt MV virus expressing green fluorescent protein can infect SLAM-negative cells. However, the virus infection does not spread from the initially infected (green) cell and does not induce syncytia, and the infectivity for SLAM-negative cells is 2 to 3 logs lower than that for SLAM-positive cells.
In addition to the hypothesis of a third receptor, it has also been suggested that wt MV binds CD46 with low affinity (19). If so, it could be expected that a domain responsible for the binding would be conserved on H proteins from both laboratory and wt strains. The Oldstone group has in fact identified such a domain: when the residues 473 to 477 of a vaccine strain MVH were replaced by five contiguous alanine residues, CD46 interaction was abrogated (24). However, the 473-477 domain is adjacent to Y481, and although the domain's substitution by five contiguous alanine residues did not affect expression of the protein, it is not impossible that the interaction between Y481 and CD46 was perturbed.
In contrast, the domain that we have identified that appears to be crucial for CD46 interaction is situated some 65 residues C-terminal to Y481 on another loop protruding from the top surface of the globular head of MVH. Our results suggest that serine 548 and phenylalanine 549 play a crucial role in the interaction with CD46. Of the two residues, serine 548 would appear to play the major role in the interaction with CD46, since mutation of this residue alone drastically reduces CD46 downregulation although the S548L/F549S double mutation is necessary to block fusion. Fusion is not blocked completely, however—a few residual syncytia can be discerned, which presumably means that some attachment to CD46 is still possible via Y481. Importantly, S548 and F549 are conserved on wt H proteins, which makes them ideal candidates for allowing wt MV a low-affinity interaction with CD46. When we mutated other amino acids on the same loop, we found that the mutation G546S also abrogated the MVH-CD46 interaction. Although wt MVH does not appear to induce CD46 downregulation, we found that wt MVH carrying the S546G mutation provoked a slight but very reproducible shift when assayed by flow cytometry, suggesting that this mutation increases slightly the affinity for CD46.
Although it has been known for some time that the mutation S546G sometimes occurs when wt MV is adapted to Vero cells (18, 26, 27) and that G546 plays a role in the interaction with CD46 (2, 4), the significance of this residue has not been clear. As Rima et al. pointed out (26), in the Edmonston wt H protein residue 546 is glycine, whereas in the Edmonston vaccine H protein it is serine, and then again in the Hallé H protein it is glycine. However, our results now make it clear that this residue is part of a major CD46-binding domain on MVH. The first intimation that residue 546 of MVH plays a role in CD46 binding came from the pioneering study on the adaptation of wt MV to Vero cells made by Shibahara et al. (27). These authors reported that MV strains isolated on B95a cells did not (unlike Vero-isolated strains) hemadsorb African green monkey erythrocytes but gained the capacity to do so after 20 passages in Vero cells. Nucleotide sequencing of the H protein from these strains showed that they carried either the N481Y mutation or the S546G mutation. This change also was reported to have occurred in another study on the adaptation of wt MV to Vero cells (29), although these authors preferred to highlight changes that occurred in internal proteins to explain attenuation.
Interestingly, Shibahara et al. (27) found that although the N481Y mutation exhibited markedly increased binding activity for African green monkey red blood cells, Vero MV isolates carrying the S546G mutation showed only marginal degrees of binding, which parallels our finding that the wt MVH S546G mutation results in a slight (but reproducible) downregulation of CD46, suggesting that although there is a CD46-binding domain between residues 546 and 549 of the MVH protein, which is necessary for MVH's attachment to CD46, it is of low affinity. Thus, we concur with the suggestion made by Manchester et al. (19) that wt MV has a low natural affinity for CD46. Our results suggest that S548 and F549 on the ß6L01 loop are responsible for this low-affinity binding, which is increased to an observable and measurable level by the N481Y mutation (and to a lesser degree by the S546G mutation) upon adaptation to Vero cells. Interestingly, Sugiyama et al. (28) have recently identified residues 548 to 551 in the RPVH as a neutralizing epitope, which suggests that the ß6L01 loop could also play a role in receptor binding by the RPVH protein.
It has been proposed that the interaction between MV and CD46 contributes to the immunosuppression induced by the virus (13, 21). The revelation that only vaccinal and laboratory MV strains use CD46 as a cellular receptor threw some doubt on this hypothesis, but if wt MV possesses the capacity to make low-affinity contacts with CD46 during infection, this would explain the original findings.
If wt MV enters Vero cells via low-affinity binding to CD46, why is there a lack of fusion? Viral fusion is the result of cell-cell fusion, and the physical requirements could be quite different from virus-cell fusion, so that it is possible that low-affinity binding to CD46 suffices for virus-cell fusion but not for cell-cell fusion. Alternatively, it has been suggested that lower levels of viral fusion exhibited by wt MV strains could be due to the physical interaction between the two MV glycoproteins being stronger in wt MV than in laboratory strains (25). This hypothesis introduces the possibility that low-affinity binding to CD46 by wt MV would not be sufficient to induce the conformational changes in the MVH that lead to fusion, which could mean that the virus enters Vero cells by some other mechanism subsequent to low-affinity CD46 binding.
|
ACKNOWLEDGMENTS |
|---|
We thank J. Fayolle for advice and discussions. R.B. is a CNRS scientist.
|
FOOTNOTES |
|---|
|
REFERENCES |
|---|
|
TopAbstractTextReferences |
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| J. Bacteriol. | Mol. Cell. Biol. | Microbiol. Mol. Biol. Rev. |
|---|
| Clin. Vaccine Immunol. | ALL ASM JOURNALS |
|---|
