![[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, September 2007, p. 9152-9161, Vol. 81, No. 17
0022-538X/07/$08.00+0 doi:10.1128/JVI.00888-07
Copyright © 2007, American Society for Microbiology. All Rights Reserved.
Matteo Porotto,1,
Olga Greengard,1,2 and
Anne Moscona1,*
Departments of Pediatrics and of Microbiology and Immunology, Weill Medical College of Cornell University, 515 East 71st St., Box 309, New York, New York 10021,1 Department of Pediatrics, Mount Sinai School of Medicine, New York, New York 100292
Received 25 April 2007/ Accepted 6 June 2007
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The HN molecule carries out its three different critical activities—receptor binding, activation of fusion, and receptor cleaving—at specific points in the process of viral entry, and understanding the regulation of these activities is crucial for the design of strategies that block viral entry (14). Once HN and F are synthesized in the infected cell, the three activities of HN need to be tightly regulated in time and space. It would be, for example, detrimental for survival of the virus if HN were to bind to receptor moieties during the process of budding and thereby to activate the fusion protein during viral packaging.
One bifunctional site (site I) on the HN of HPIV3 possesses both binding and neuraminidase activities (11), and we recently obtained experimental evidence for a second receptor binding site (site II) on HPIV3 HN (21) that has both binding and F-activation properties. Site I binds the sialic acid-containing receptor with high avidity, and site II contributes to F activation upon receptor binding, although the mechanism remains unclear. Mutation of HN at specific residues at this site, which is next to the dimer interface, confers enhanced fusion properties without affecting neuraminidase activity or receptor binding at neutral pH (21).
Mutations in both site I and site II have evolved in culture under selective pressure. When infection was carried out in cells that were treated with neuraminidase to achieve partial depletion of cell surface receptors, the selective pressure of decreased receptor availability favored those variants with enhanced entry. Variants emerged in site I, with alterations at T193 (17) and at D216 (9), and at site II, with alterations at H552 and N551 (21). The variants with mutations in site I at T193 have been studied extensively and have elucidated functions of the primary binding site (9, 14, 17, 19, 23, 24). D216 is adjacent to T193 in the three-dimensional structure, and mutations that confer neuraminidase deficiency have arisen under this selective pressure (9, 22). The mutations in site II that we have recently characterized, H552Q and N551D, also arose under the same selective pressure for HNs with enhanced promotion of viral entry, and the second of these, N551D, is specifically enhanced in its F-activation function (21).
Using HN molecules bearing specific mutations in the residues located at site I and II, we found that the two functional sites are each critical to the life cycle of the virus and can counterbalance each other under specific conditions, allowing the virus to maintain key functions in the requisite equilibrium, even under extreme conditions. We demonstrate here that mutations at HN receptor binding site I, as well as at site II, alter receptor avidity at low pH. These mutations allow pH to modulate the binding and subsequent fusion processes, and we discuss the implications of this result for the life cycle of the virus.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Assay of neuraminidase activity. Neuraminidase assays were performed in transiently transfected 293T cell monolayers as previously described (17).
Mutagenesis. HN genes carrying sequence alterations in the regions coding for amino acid residue D216 were generated through the use of the PCR-based approach of overlap extension (7, 8). In order to introduce mutations in a site-directed manner, three PCRs were used with primers designed against the wild-type (WT) HPIV3 HN. Two external primers carrying restriction enzyme sites for cloning into the pEGFP-C3 vector (BD Biosciences/Clontech) were used, along with two pairs of internal primers designed to introduce the desired mutation. Two reactions using one external primer and one internal mutagenesis primer were performed; the resulting products were gel purified and used as a template for the third PCR. The success of the third reaction relied on sequence overlap between the products of the first two reactions and resulted in a full-length product carrying the desired mutation.
HN constructs. Mutagenized HPIV3 HN cDNAs were digested with SacI or EcoRI and BamHI and ligated into digested pCAGGS and pEGFP mammalian expression vectors as previously described (19). Positive clones were sent for sequencing to verify the mutations and to ensure that no additional alterations had been introduced, as described previously (22).
Transient expression of HN and F genes. Transfections were performed according to the Lipofectamine and Plus or Lipofectamine 2000 manufacturer's protocols (Invitrogen).
Quantification of cell surface expression of HN and F using ELISA. Surface expression of proteins was determined on the preparations of cells expressing HN, so as to assure that the expression levels of the WT and mutant HN proteins in these preparations were within 10% of one another. To quantify the amount of HN or F expressed on the cell surface of 293T cells, an enzyme-linked immunosorbent assay (ELISA) was performed as described previously (19). Transfected 293T cells were washed with phosphate-buffered saline (PBS) after incubation at 37°C, fixed for 10 min with 4% formaldehyde PBS, and reacted with a mixture of anti-HPIV3 HN or anti-F monoclonal antibodies supplied by Judy Beeler from the World Health Organization repository. The cells were left at room temperature for 30 min and then washed three times with PBS. Horseradish peroxidase-conjugated anti-mouse immunoglobulin G (Bio-Rad) was then added to the cells in PBS-1% bovine serum albumin (1:10,000 dilution); the cells were then incubated for 30 min at room temperature. The cells were washed three times with PBS before incubation with TMB substrate (Pierce) according to the manufacturer's protocol. Absorbance measurements were taken at 450 nm with a Spectramax M5 ELISA reader.
HAD assays. Hemadsorption (HAD) was performed and quantified as previously described (22). After aspiration of the medium from transfected 293T cell monolayers in 24- or 48-well Biocoat plates (Becton Dickinson Labware), the medium was replaced with 300 or 150 µl of 1% red blood cells (RBCs) in serum-free, CO2-independent medium (pH 7.3; Gibco) and placed at 4°C for 30 min. The wells were then washed three times with 300 or 150 µl of cold CO2-independent medium. The bound RBCs were lysed with 200 µl of RBC lysis solution (0.145 M NH4Cl, 17 mM Tris-HCl), and the absorbance was read at 405 nm with a Spectramax M5 ELISA reader.
Partial removal of sialic acid receptors from RBCs. Partial receptor depletion of RBCs was achieved by treatment of 2 ml of a 10% RBC solution in serum-free medium for 2 h at 37°C with 0 to 200 mU of C. perfringens neuraminidase (type X from C. perfringens [catalog no. N-2876; Sigma Scientific]) as described previously (19). Neuraminidase was then removed by washing the RBCs three times with serum-free medium. Each set of RBCs was then resuspended in serum-free, CO2-independent medium to make final 2% RBC stocks.
Use of receptor-depleted RBCs to assess HN receptor-binding avidity. RBCs partially depleted of their surface sialic acid receptors (described above) were used to determine the relative receptor binding avidities of variant HN molecules as described previously (19). In each experiment, all of the RBCs were from the same preparation of depleted stocks (described above). The RBCs were overlaid onto 48-well plates of 293T cell monolayers transiently transfected 24 h prior with WT or variant HN expressed as described above. The plates were placed at 4°C for 30 min to allow RBC binding. The cell monolayers were then washed at 4°C with cold CO2-independent medium to remove unbound RBCs, the bound RBCs were lysed (0.145 M NH4Cl, 17 mM Tris-HCl), and the absorbance was read at 405 nm on the Spectramax M5 ELISA reader. The results are presented as percent retention of RBCs relative to control (undepleted RBCs) versus degree of depletion expressed as milliunits of bacterial neuraminidase. For pretreatment of HN-expressing cells with neuraminidase, cells in monolayers were treated with 25 mU per well in 48-well plates (5 x 105 cells) for 3 h at 37°C, transferred to 4°C, and washed.
Assays for F activation, receptor retention, and receptor release. Monolayers of 293T cells transiently expressing WT F or F K108G (cleavage-site mutant F) and WT or mutant HNs were washed three times and incubated with 1% RBC suspensions at pH 7.3 for 30 min at 4°C. After a rinse to remove unbound RBCs, the plates were transferred to 37°C with simultaneous addition of media buffered to pH 8.0 or 5.0. At the indicated times of incubation, the RBCs that (i) were released to the medium, (ii) remained bound, or (iii) had undergone membrane fusion were quantified as follows. The liquid phase was collected in V-bottom tubes for measurement of released RBCs. The monolayers were incubated at 4°C with 200 µl of RBC lysis solution (0.145 M NH4Cl, 17 mM Tris-HCl), which lyses only the unfused RBCs, leaving the cell monolayers intact, and the liquid phase was collected for measurement of reversibly bound RBCs. The cell monolayers were then lysed in 200 µl 0.2% Triton X-100-PBS and transferred to flat-bottom 96-well plates for quantification of RBCs that have undergone fusion or F insertion. The amount of RBCs in each of the three compartments was determined by absorption measurement at 405 nm.
ß-Gal complementation-based fusion assay. We adapted an assay (13) based on alpha complementation of ß-galactosidase (ß-Gal); the ß-Gal protein lacking the N-terminal 85 residues (omega peptide) is expressed from one plasmid, and the N-terminal 85 residues (alpha peptide) is expressed from a second plasmid. Cell fusion leads to complementation, and ß-Gal is quantitated by using the Galacto-Star (Applied Biosystems) chemiluminescent reporter gene assay system. Receptor-bearing cells expressing the omega peptide are mixed with HPIV3 HN/F-coexpressing cells that also express the alpha peptide at various pHs. Fusion is stopped by lysing the cells with lysis buffer and, after addition of the substrate, the levels of fusion are measured by using a Spectramax M5 ELISA reader.
Infectivity assay. CV1 cell monolayers were treated with different amounts of neuraminidase to achieve graded receptor depletion in serum-free medium for 3 h at 37°C with 0, 30, 45, or 90 mU of C. perfringens neuraminidase. Intact viruses—WT and N551D HN—were bound to the cell monolayers for 90 min at 4°C in CO2-independent medium at pH 5.0 or 8.0, and then viral entry at 37°C was quantified by plaque assay as described previously (18).
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
50% loss of neuraminidase activity, without a change in receptor binding and fusion-promotion potential (22). This residue is highly conserved in the paramyxovirus HN proteins. For Newcastle disease virus (NDV), substitution of the corresponding residue (D198) with arginine resulted in an HN that is neuraminidase deficient, with impaired binding and fusion promotion (4, 10). We introduced the same substitution in the corresponding residue (D216) on HPIV3-HN's active site I, hypothesizing that this alteration may impact the molecule's receptor binding and fusion promotion function. In monolayer cultures of 293T cells transfected with WT HN or D216R HN, with similar surface expression of HN, no neuraminidase activity could be detected for D216R HN, using a sensitive fluorimetric assay (6; data not shown). D216R HN had minimal receptor-binding capacity at 4°C as measured by HAD, unless pretreated with exogenous neuraminidase (Fig. 1). In this HAD assay we quantitated RBC adherence to cells expressing WT or variant HN at 4°C, a temperature at which neuraminidase activity is negligible but WT HN's binding function is intact (6). At 4°C, HAD for the D216R HN was minimal but reached the WT level if the HN-expressing cells were pretreated with exogenous neuraminidase (Fig. 1A). This requirement for neuraminidase pretreatment in order to permit HN receptor binding at 4°C was similar to that observed for the only previously available neuraminidase-dead HPIV3 HN, the doubly mutated variant D216N/P111S HN (22), and indicates that neuraminidase pretreatment serves to prevent the occupation of binding sites on the neuraminidase-dead HN by adjacent sialic acid moieties on the cell surface (21, 22), sialic acid moieties that would normally be cleaved by HN's neuraminidase. At 37°C, however, higher dissociation rates between HN and its receptor on the cell surface ("decoy receptors") should allow sialic moieties on erythrocytes to successfully compete with the adjacent cell surface receptors for HN. Accordingly, Fig. 1B shows that, at 37°C, D216R HN, like D216N/P111S HN (22), exhibited significant receptor-binding activity even without pretreatment with exogenous neuraminidase.
|
We had previously shown that lowering the pH toward the optimum of HN's neuraminidase activity (pH 4.7[9]) reduces HN's receptor retention, and thus reduces F-triggering potential, for WT but not for neuraminidase-deficient HN (23). We have also previously reported that the HN bearing the N551D mutation has a WT neuraminidase activity level (21).
For measurement of avidity under various conditions, RBC preparations with different degrees of receptor depletions were used to quantify HAD on WT and variant HN-expressing cell monolayers (19, 21). HN molecules with higher avidities can bind to RBCs that have lower receptor densities. The results of this quantitative receptor avidity assay for WT HN, D216R HN, and N551D HN at pH 8.0 and 5.0 are presented in Fig. 2. The RBC binding, expressed as a percentage of that obtained with nondepleted RBCs (y axis), is plotted against the degree of RBC receptor depletion (x axis); thus, the less steeply the curve declines, the higher the HN's receptor avidity.
|
Table 1 quantitates the receptor avidity comparison between the WT and the two HN variants. The amount of neuraminidase treatment required for 50% binding for each HN is listed at both pH values: 5.0 and 8.0. At pH 8.0 the avidity of D216R HN for sialic acid is far below that of the WT, whereas (as we reported recently) that of N551D HN is identical to that of the WT (21). At pH 5.0, on the other hand, both variants show appreciably lower avidity than the WT, with the variants requiring less neuraminidase treatment than the WT in order to reduce binding by 50% (although for D216R HN this deficit is less striking than at pH 8.0).
|
To compare the F activation function of WT HN, D216R HN, and N551D HN at different pHs, cells coexpressing WT F and either WT or variant HN were allowed to bind RBCs at 4°C. The cells were then transferred to 37°C at various pH conditions and, at various times after transfer, we determined the amounts of target RBCs that (i) were released into the medium, (ii) were bound but not fused, or (iii) had undergone fusion. The graphs in Fig. 3 show the amount of RBCs in these three pools as a function of incubation time for cells coexpressing F with WT or variant HN at each pH. Since F activation and/or triggering precedes F insertion and fusion, pool iii reflects the effects of pH on the F-triggering potential of the different HNs.
|
At pH 5.0, close to the pH optimum of the HPIV3 neuraminidase, expression of the inherent triggering potential of both WT HN and N551D HN is limited by their high neuraminidase activity, which depletes the pool of reversibly bound target cells before F can be activated. For WT HN and N551D HN (which possesses WT level of neuraminidase (21), F triggering at pH 5.0 is well below that at pH 8.0. For N551D HN, F triggering is further limited by the lower receptor avidity of this variant at pH 5.0. Consequently, at pH 5.0 N551D HN triggers even less effectively than WT HN: after 60 min of incubation at pH 5.0, WT HN induces fusion of 40% of the RBCs, whereas N551D HN induces fusion of only 20% of the RBCs, an 80% reduction of its triggering activity at pH 8.0.
In striking contrast, acidification enhances the F-triggering function of D216R HN. At pH 5.0, this variant HN exhibits increased receptor avidity (see Fig. 2), and it does trigger F more effectively than at pH 8.0. After 30 min, the fusion in the presence of D216R HN at pH 5.0 was double that observed at neutral pH, and the maximum (80% fusion) was reached 1 h earlier than at pH 8.0. For WT HN at low pH, F triggering is severely limited by release from receptor, so that the neuraminidase-dead D216R HN promotes fusion more effectively than the WT at pH 5.0. In fact, D216R HN is as effective at pH 5.0 as is the WT HN at a pH (8.0) that is not conducive to neuraminidase activity, indicating that the D216R mutation did not alter HN's inherent triggering efficacy.
Receptor binding and release by D216R HN and N551D HN is modulated in opposite directions by pH. We next addressed the effect of pH on receptor binding and release that occurs independently of fusion. For this purpose the experiments of Fig. 3 were repeated using cells that coexpress each HN with an F protein mutated at the F1-F2 cleavage site (K108G) that is expressed on the cell surface but is incapable of mediating fusion (R. Iorio, unpublished data). In the absence of functional F, the release of bound RBCs upon transfer to 37°C should depend primarily on the HN's neuraminidase activity and, as seen in Fig. 4, receptor release from the D216R neuraminidase-dead HN was indeed far below that from the N551D HN with its WT neuraminidase activity. At pH 8.0, by 120 min both WT HN and N551D HN released 40% of bound RBCs. D216R HN showed negligible receptor release over this period of time.
|
Activation of fusion requires a threshold level of HN receptor binding. In order to assess whether pH modulation of F triggering can operate independently of its effect on receptor binding, the fusion assay described in Fig. 3 was carried out using a series of RBC preparations with various degrees of receptor depletion (19, 25).
The basis for this strategy for assessing F triggering is that HN molecules with higher avidities can bind to and fuse with RBCs that have lower receptor densities (19). Cells coexpressing WT F and WT or variant HNs were allowed to bind to a panel of receptor-depleted RBCs at 4°C at pH 5.0 and 8.0. The cells were then transferred to 37°C and assessed as in Fig. 3. It was important in this experiment to distinguish between an elevated RBC fusion level that results from enhanced HN binding and an elevated RBC fusion level that results from enhanced HN F triggering. Therefore, we divided the fusion value by the percentage of RBCs that were bound to each set of HN/F-coexpressing cells at time zero (i.e., before initiating fusion by transfer to 37°C). In this way we eliminated the variable of the different amounts of RBCs with graded levels of depletion that bound to HN/F-coexpressing cells at 4°C.
In Fig. 5, the extent of HN-induced fusion (y axis) is shown as a function of the degree of RBC receptor depletion (x axis). In this experimental setting, fusion promoted by the two variant HNs again was affected in opposite directions (relative to WT HN) by pH conditions; fusion promoted by each HN decreased with increasing receptor depletion, but at different rates. For the N551D HN, at pH 8.0 F activation was enhanced compared to WT HN, confirming that this mutation augments HN's triggering efficacy. At pH 8.0 this HN promoted fusion effectively in the face of more extensive (10 mU or more) RBC receptor depletion. This result was expected, since the mutation had no effect on N551D HN's receptor avidity at pH 8.0 (see Fig. 2), and the use of receptor-depleted RBCs highlights this inherent F-triggering advantage. However, this HN, with its WT neuraminidase activity, was less effective under acidic than neutral conditions; a receptor depletion of <5 mU neuraminidase was sufficient to minimize fusion at pH 5.0, a pH at which the N551D HN shows inefficient F triggering. The neuraminidase-dead D216R HN, in contrast, activated F less efficiently at pH 8.0 than at pH 5.0. At this low pH, the D216R HN was more efficient at F triggering than WT HN and promoted fusion at 10 mU or even greater degrees of receptor depletion.
|
For the D216R HN, however, the lack of neuraminidase activity prevents receptor release (see Fig. 4). This feature allows comparison of the requirements for binding versus fusion, without the effect of receptor cleavage. Under these conditions, it was found that the threshold avidity level required for fusion is higher than the threshold avidity required for binding. Below this threshold avidity level required for fusion, while binding may occur, fusion cannot take place.
The D216R HN mutant provides specific support for the concept that the threshold avidity level required for fusion is higher than that required for attachment, since this variant allows for manipulation of avidity without altering neuraminidase, so that the effect of receptor binding on fusion promotion can be observed. The D216R HN at pH 8.0 binds efficiently, retaining undepleted RBCs (see Fig. 3) or minimally depleted RBCs (Fig. 5), but, with low avidity, does not promote fusion efficiently. However, at lower pH, when its avidity is higher, this HN promotes fusion efficiently. At pH 8.0, a minimal receptor depletion (5 mU) still allows for D216R HN to bind receptors as avidly as WT HN (see Fig. 2) but sharply reduces its fusion promotion, revealing the higher threshold avidity that is needed for the HN's fusion promotion activity.
Modulation of D216R HN and N551D HN's F activation by pH in epithelial cell fusion. To evaluate pH dependence for HN/F-mediated fusion between epithelial cells, we assessed fusion promotion activity of the expressed mutant HNs in 293T cells, coexpressed with WT F on the cell surface. The HN/F-coexpressing cells were pretreated with neuraminidase to prevent fusion from occurring before the start of the experiment. These cells were incubated with receptor-bearing Vero cells, and fusion was quantified by using a ß-Gal complementation-based fusion assay (13, 21). Figure 6 shows the extent of fusion for the two variant HNs at each pH, expressed as a percentage of the fusion in cells coexpressing WT HN. The N551D HN is more fusion promoting than WT HN when coexpressed with WT F at pH 8.0, due to its F-triggering advantage, but much less fusion promoting at pH 5.0. In contrast, the D216R HN showed a much higher fusion promotion capacity at pH 5.0 but virtually no fusion promotion whatsoever at pH 8.0. Thus, the study of epithelial cell fusion reveals the pH dependence of HN's fusion promotion function, modulated in the opposite direction for N551D and D216R HN, and does so even more strikingly than assays using RBCs as target cells.
|
Modulation of receptor binding by pH during viral entry. In order to assess the impact that pH modulation of avidity might have on infectivity mediated by the N551D HN, we used an approach based on partial receptor depletion from target cell surfaces. Cell monolayers were treated with different amounts of neuraminidase to achieve graded receptor depletion (0, 30, 45, or 90 mU of neuraminidase treatment). Intact viruses—WT and N551D HN—were bound to the cell monolayers at 4°C at pH 5.0, or 8.0, and viral entry at 37°C was quantified by plaque assay.
Figure 7 shows that the virus bearing the N551D HN is less effective at entry after receptor binding at pH 5.0 than the WT virus, as indicated by the steeper decline in the percentage of infection of the N551D HN-bearing virus with increasing receptor depletion. This inefficient infection reflects the lower affinity of the N551D HN-bearing virus at this acidic pH. Note that the difference between the larger pH effect noted for RBC fusion in Fig. 5 compared to the effect on virus-cell fusion here (Fig. 7) reflects the fact that more HN-receptor interactions are required for cell-cell fusion (including RBC fusion) than for virus-cell fusion (15), and therefore effects on virus-cell fusion are more difficult to detect. After receptor binding at pH 8.0, WT virus and the N551D HN-bearing variant viruses exhibit a similar decline in the percentage of infection with increasing receptor depletion. This finding indicates that at pH 8.0 at 4°C (a temperature at which fusogenic phenotypes are neutral) the two HNs function similarly, again reflecting the avidity results in Fig. 2. The experiment in Fig. 7 addresses the HN-receptor interaction during the binding phase that permits entry of viruses, and indicates that the pH modulation of the N551D HN's affinity has consequences for viral infectivity, highlighting the disadvantage conferred by the N551D mutation at a low pH.
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
To scrutinize the specific contribution of HN's neuraminidase to F triggering, we previously evaluated a doubly mutated variant HN that has a mutation at residue D216, replacing D with N and thereby partially reducing neuraminidase activity, and a second mutation in the HN stalk (P111S) that renders this variant neuraminidase dead and also confers a partial F-triggering defect. This triggering-defective variant (25), along with data from other laboratories (5, 12, 23, 25, 26, 28, 29, 32), established the stalk of HN as critical for fusion promotion and HN-F interaction. The doubly mutated (P111S/D216N) HN retains its F-triggering capacity and fusion promotion at conditions conducive to neuraminidase activity (acidic pH). However, at the same acidic pH, an HN with only the F-activation defect (P111S HN) that has residual neuraminidase fails to contact its receptor for long enough to trigger F (23). The difference between these two variants highlighted the potential impact of neuraminidase on F activation. The doubly mutated HN (D216N/P111S), in contrast to the mutant in the present study in which residue D216 is mutated to an R, was neuraminidase dead, but its receptor avidity was not modulated by pH. The fact that the D216R variant HN has an inactive neuraminidase alongside a normal inherent F-triggering function permits F activation to occur at low pH. Thus, it is the specific pH effect in the context of residue R216—and not the neuraminidase deficiency conferred by this mutation—that accounts for the pH sensitivity of the D216R HN's avidity.
For NDV, a mutation at D198 to R (corresponding to HPIV3 HN D216R) resulted in the abolition of receptor recognition, fusion, and neuraminidase activity (4). However, receptor recognition could be partially restored by exogenous neuraminidase treatment or coexpression of a heterotypic HN, suggesting that the defect in neuraminidase activity accounted for the observed effect on binding and perhaps also thereby for the effect on fusion. The HPIV3 D216R HN, when coexpressed with HPIV3 F, failed to induce syncytium formation (data not shown). However, neuraminidase treatment allowed these coexpressing cells to fuse with receptor-bearing cells, a finding consistent with the described rescue of NDV HN function by neuraminidase treatment.
One aspect of note in our data is the comparison between the function of the D216R HN in an F-triggering assay (Fig. 3 and 5) and in an epithelial cell fusion assay (Fig. 6). This comparison reveals the functional disadvantage of the receptor avidity defect conferred by the D216R mutation in HN. While in the F-activation assay shown in Fig. 2 the D216R HN activated F at pH 8.0 with 50% efficiency compared to pH 5.0, in the cell fusion experiments in Fig. 5 D216R HN shows no detectable promotion of cell fusion at pH 8.0. At pH 5.0, this variant's neuraminidase defect actually confers enhanced receptor interaction and F activation. The comparison between the D216R HN activity in these two different settings (i.e., F triggering and cell fusion) shows that the low receptor avidity of the D216R HN at pH 8.0 has its most significant impact when studied in an assay reflective of the viral life cycle and suggests that a virus bearing the D216R HN—or any variant HN with lower avidity at neutral pH—would not be infectious.
HN, after its synthesis in the endoplasmic reticulum of the infected cell, is further processed during transit through the acidic environment of the Golgi network. It is possible that at this low pH, high receptor avidity can enhance HN's binding to intracellular sialic moieties, reduce its chances of reaching the membrane, and thereby reduce virion budding. For this reason pH modulation of receptor avidity would be advantageous, if it ensured that avidity for receptor is lower in the acidic environment. The N551D HN is highly fusion-promoting because of its enhanced inherent triggering efficacy, and it is possible that its low receptor avidity at pH 5.0 prevents this HN from binding to intracellular receptors during its transit to the cell surface. The lower avidity at acidic pH could thereby balance the hypertriggering property of this HN. For the D216R HN, in contrast, high receptor avidity at the acidic pH within the cell could result in attachment to sialic acid moieties on intracellular proteins and possibly to premature F activation by receptor-bound D216R-HN. This could deplete the prefusion F proteins that would normally be activated only upon contact with target cells. A virus with D216R HN as its binding protein would be unlikely to be infectious, since its entry into host cells would be handicapped by this HN's low receptor avidity and F activation at the neutral pH of the plasma membrane.
The residue at D216 is highly conserved among paramyxoviruses HN proteins, since it is a key catalytic residue for neuraminidase activity and serves a function critical to the viral life cycle. Only under the in vitro selective pressure of exogenous neuraminidase treatment, designed to reduce the density of cell surface sialic acid receptors, did variant viruses emerge with mutations at D216 that reduce neuraminidase activity (9, 22). The emergence of these variants, which cope successfully with conditions of receptor scarcity by cleaving less sialic acid, provided the first indication of the balance between receptor binding and cleaving activities that determines fusion for HPIV3 (23).
One of the variants that emerged under the selective pressure of receptor depletion, bearing the doubly mutated D216N/P111S HN mentioned above, provided a striking example of the balance between the three functions of HN. In light of the F-triggering defect of this doubly mutated HN, only the neuraminidase deficiency enables the variant virus to exist, by preserving target receptors for long enough to allow for adequate fusion (23). Evolution in the face of the selective pressures provided by host interactions has led HPIV3 to its existing sequence and structure, likely to be ideal for a virus whose infectious cycle involves different cellular compartments with significantly different pH conditions. It will thus be of interest to understand the consequences of the individual mutations described here that alter HN functions to pathogenesis in an animal model of infection and to assess the extent of allowable flexibility in vivo.
We have previously reported that the HN residue N551 (in site II of HN) is implicated in the F-triggering activity of HN and that the N551D mutation confers enhanced F triggering. As seen in Fig. 2, this enhancement is subtle at 37°C (although marked at lower temperatures [21]); however, upon the use of an assay that more precisely distinguishes differences in F activation using receptor-depleted RBCs (in Fig. 4), the inherent triggering advantage is revealed. This result suggests that the variant virus with the N551D HN may have an advantage in infecting cells that have a paucity of receptors. This advantage likely resulted in the emergence of this variant virus under selective pressure of partial receptor depletion (21).
Evidence for the presence of a second receptor-binding site on HN derives from the study of HNs with mutated residues at N551 and H552, and since these variants emerged under selective pressure favoring enhanced viral entry, the mutations at this site point to its critical importance in the WT virus life cycle (21). However, it has not thus far been possible to document the presence of this second site on the WT HN molecule (21). A speculative interpretation of some puzzling findings in the experiment shown in Fig. 7, where we tested pH effects on F activation during viral entry, may shed light on this problem. The experiment showed that the N551D HN-bearing virus is less effective at entering cells at low pH than the WT virus, a finding in line with its lower receptor affinity at this pH. However, the efficiency of WT virus entry at acidic pH—more than its efficiency at pH 8.0, indicating superior WT HN virus-receptor binding at 4°C at the acidic pH—was perplexing. We speculate that low pH may expose binding site II of WT HN on viruses, allowing the WT virus to bind avidly and enter cells at this pH; however, the N551D mutation alters site II, affecting its pH responsiveness so as to decrease avidity at acidic pH.
If in fact true, exposure of WT HN's site II at acidic pH could explain a confusing issue in the literature. In one recent report, fusion mediated by the paramyxovirus SER F protein coexpressed with SER HN was found to be activated at low pH (27). However, using somewhat different assays, another study did not detect activation of SER fusion at low pH (1). We have previously asked why the SER neuraminidase, when exposed to low pH in the first study, did not lead to receptor dissociation and thus decrease fusion, and we suggested that perhaps the short incubation times at low pH in the first study did not suffice for receptor depletion. In light of our present findings it is, however, conceivable that the short low-pH pulse, while not allowing adequate time for neuraminidase depletion of receptors, could have exposed the site II on SER HN.
Although it is a formal possibility that the N551D mutation introduces a second site to HN that is not present in the WT HN, it would be highly unlikely for H552Q and N551D to arise during natural infection—each conferring distinct functional properties of a second receptor interaction site—were this site not biologically relevant. The facts that the site II mutations H552Q and N551D in HPIV3 HN arose naturally under selective pressure for enhancement of entry and that the two mutations are adjacent to one another support the notion that this position represents a bona fide binding/F-activation site on HN. Experiments have been performed in other laboratories in which HN mutations were introduced on the basis of primary sequence comparisons and affected in vitro binding but not fusogenicity or infectivity (2). Mutation of an HPIV1 HN residue based on the site II sequence of Sendai virus led to reduced elution of RBCs from HPIV1 HN (2); however, this mutation may not reflect an HN residue relevant for HPIV1 biology, and the result may not eliminate the possibility of a genuine second binding site. For HPIV3, further analysis of the contributions of H552 and N551 to a second binding site will depend in part on crystal structure analysis.
We have previously compared the NDV and HPIV3 HNs and shown that while HPIV3 HN's bifunctional site I is sensitive to inhibition by zanamivir (20, 24)—and only a mutation that excludes zanamivir from the HPIV3 HN active site I confers resistance (24)—NDV is resistant to zanamivir due to its resistant site II. Structural analysis had suggested (24) that the HPIV3 HN residue H552 forms part of the dimeric interface both in the HPIV3 HN structure (11) and in the NDV HN structure at pH 6.5 (3) and that for HPIV3 HN this residue is not involved in the formation of the primary ligand-binding site, which is compatible with the mutation having no effect on zanamivir binding affinity or on neuraminidase activity (24). We had speculated that the increased receptor-binding avidity of the H552Q variant could be explained by either an indirect alteration in conformation of the primary binding site, possibly via residue E549, or by a putative second receptor binding site. We noted, however, that should such a second receptor-binding site occur in the same vicinity in HPIV3 HN as that of NDV HN, it would have to be different in several respects from that in NDV HN (24), and in fact this is the case. In contrast to the high-avidity second binding site in NDV HN that is activated upon receptor binding to the primary site (20), the HPIV3 HN site II has lower receptor avidity than the primary binding site but is involved in F activation. These intriguing differences will merit further exploration, as the mechanistic details are dissected.
The HPIV3 virus variant bearing the N551D HN is a viable virus that emerged under selection pressure as described above. HN's receptor binding sites I and II can both affect the balance between receptor avidity, neuraminidase, and F activation, and we have shown that site II is involved in the interaction of HN with F that leads to F activation. Therefore, the N551D HN variant virus must compensate for its enhanced F triggering with some other feature that preserves the requisite balance between functions. Interpreted in light of the possible exposure of site II at acidic pH, it could be disadvantageous if F triggering were enhanced in an acidic milieu (i.e., intracellular compartments) and perhaps the compensation that permits this variant virus viability is its lower receptor avidity at acidic pH. In this way, F triggering is enhanced in the proper location for execution of the viral life cycle but not enhanced in inappropriate compartments.
The emergence of variants with mutations at sites I and II under selective pressure suggests that the design of inhibitory compounds must contend with the apparent flexibility of this virus. This flexibility allows the virus to modulate the balance of key functions that are distributed among several sites on the HN molecule. Plasticity may mean that the use of compounds that target only one site on the molecule will readily lead to the emergence of escape variants.
The nature of the interaction between the two binding sites on HN and the relationship of these sites to the stalk region, which is also important to F activation, are fundamental questions. Not only can a balance of key functions of HN be preserved via compensatory changes in other sites of HN, but these key functions of HN have evolved to respond in specific ways to pH conditions. The various options available to this HN molecule reveal the complex requirements for a virus that must travel through a variety of cellular compartments during its life cycle and must temper specific HN functions in one compartment while activating these functions in another cellular location.
|
ACKNOWLEDGMENTS |
|---|
We thank J. Beeler for providing needed antibodies, Richard Peluso for helpful discussions, and S. Rusconi for the ß-Gal complementation reagents.
|
FOOTNOTES |
|---|
Published ahead of print on 13 June 2007.
L.M.P. and M.P. contributed equally to this study.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||
) and after (
) neuraminidase pretreatment, were carried out as described in Materials and Methods. The bars (absorbance at 405 nm) represent means of results ± the standard deviation (SD) on triplicate cell monolayers from a representative experiment.
