|Home | About | Journals | Submit | Contact Us | Français|
The fusion of enveloped viruses with the host cell is driven by specialized fusion proteins to initiate infection. The “class I” fusion proteins harbor two regions, typically two heptad repeat (HR) domains, which are central to the complex conformational changes leading to fusion: the first heptad repeat (HRN) is adjacent to the fusion peptide, while the second (HRC) immediately precedes the transmembrane domain. Peptides derived from the HR regions can inhibit fusion, and one HR peptide, T20 (enfuvirtide), is in clinical use for HIV-1. For paramyxoviruses, the activities of two membrane proteins, the receptor-binding protein (hemagglutinin-neuraminidase [HN] or G) and the fusion protein (F), initiate viral entry. The binding of HN or G to its receptor on a target cell triggers the activation of F, which then inserts into the target cell and mediates the membrane fusion that initiates infection. We have shown that for paramyxoviruses, the inhibitory efficacy of HR peptides is inversely proportional to the rate of F activation. For HIV-1, the antiviral potency of an HRC-derived peptide can be dramatically increased by targeting it to the membrane microdomains where fusion occurs, via the addition of a cholesterol group. We report here that for three paramyxoviruses—human parainfluenza virus type 3 (HPIV3), a major cause of lower respiratory tract diseases in infants, and the emerging zoonotic viruses Hendra virus (HeV) and Nipah virus (NiV), which cause lethal central nervous system diseases—the addition of cholesterol to a paramyxovirus HRC-derived peptide increased antiviral potency by 2 log units. Our data suggest that this enhanced activity is indeed the result of the targeting of the peptide to the plasma membrane, where fusion occurs. The cholesterol-tagged peptides on the cell surface create a protective antiviral shield, target the F protein directly at its site of action, and expand the potential utility of inhibitory peptides for paramyxoviruses.
Fusion of enveloped viruses with the host cell is a key step in viral infectivity, and interference with this process can lead to highly effective antivirals. Viral fusion is driven by specialized proteins that undergo an ordered series of conformational changes. These changes facilitate the initial, close apposition of the viral and host membranes, and they ultimately result in the formation of a fusion pore (reviewed in reference 12). The “class I” fusion proteins harbor two regions, typically two heptad repeat (HR) domains: the first one (HRN) adjacent to the fusion peptide and the second one (HRC) immediately preceding the transmembrane domain. Peptides derived from the HR regions can inhibit fusion, and one of them, T20 (enfuvirtide), is in clinical use for HIV-1 (19). Peptides derived from the HRN and HRC regions of paramyxovirus fusion (F) proteins can interact with fusion intermediates of F (3, 20, 22, 37, 46, 49) and provide a promising antiviral strategy.
The current model for class I-driven fusion postulates the existence of a so-called prehairpin intermediate, a high-energy structure that bridges the viral and cell membranes, where the HRN and the HRC are separated. The prehairpin intermediate spontaneously collapses into the postfusion structure—a six-helical bundle (6HB), with an inner trimeric coiled-coil formed by the HRN onto which the HRC folds (12, 14, 30, 40). The key to these events is the initial activation step, whereby HN triggers F to initiate the process. Structural and biophysical analyses of the paramyxovirus 6HB (30, 50, 51) suggest that inhibitors bind to the prehairpin intermediate and prevent its transition to the 6HB, thus inhibiting viral entry. The peptides bind to their complementary HR region and thereby prevent HRN and HRC from refolding into the stable 6HB structure required for fusion (3, 10, 40). The efficiency of F triggering by HN critically influences the degree of fusion mediated by F and thus the extent of viral entry (35). In addition, differences in the efficiency of triggering of the fusion process impact the efficacy of potential antiviral molecules that target intermediate states of the fusion protein (36).
Paramyxoviruses cause important human illnesses, significantly contributing to global disease and mortality, ranging from lower-respiratory-tract diseases in infants caused by human parainfluenza virus types 1, 2, and 3 (HPIV1, -2, and -3) (9, 48), to highly lethal central nervous system diseases caused by the emerging paramyxoviruses HeV and NiV. No antiviral therapies or vaccines yet exist for these paramyxoviruses, and vaccines would be unlikely to protect the youngest infants. Antiviral agents, therefore, would be particularly beneficial. All paramyxoviruses possess two envelope glycoproteins directly involved in viral entry and pathogenesis: a fusion protein (F) and a receptor-binding protein (HN, H, or G). The paramyxovirus F proteins belong to the group of “class I” fusion proteins (44, 45), which also include the influenza virus hemagglutinin protein and the HIV-1 fusion protein gp120. The F protein is synthesized as a precursor protein (F0) that is proteolytically processed posttranslationally to form a trimer of disulfide-linked heterodimers (F1-F2). This cleavage event places the fusion peptide at the F1 terminus in the mature F protein and is essential for membrane fusion activity. The exact triggers that initiate a series of conformational changes in F leading to membrane fusion differ depending on the pathway the virus uses to enter the cell. In the case of HPIV, HeV, and NiV, the receptor-binding protein, hemagglutinin-neuraminidase (HN) (in HPIV3) or G (in HeV and NiV), binds to cellular surface receptors, brings the viral envelope into proximity with the plasma membrane, and activates the viral F protein. This receptor-ligand interaction is required for the F protein to mediate the fusion of the viral envelope with the host cell membrane (23, 33, 35).
The HRC peptide regions of a number of paramyxoviruses, including Sendai virus, measles virus, Newcastle disease virus (NDV), respiratory syncytial virus (RSV), simian virus 5 (SV5), Hendra virus (HeV), and Nipah virus (NiV), can inhibit the infectivity of the homologous virus (17, 20, 31, 37, 47, 49, 52, 53). Recently, we showed that peptides derived from the HRC region of the F protein of HPIV3 are effective inhibitors of both HPIV and HeV/NiV fusion (31) and that, for HeV, the strength of HRC peptide binding to the corresponding HRN region correlates with the potency of fusion and infection inhibition (30). However, peptides derived from the HPIV3 F protein HRC region are more effective at inhibiting HeV/NiV fusion than HPIV3 fusion, despite a stronger homotypic HRN-HRC interaction for HPIV3 (30, 31). We showed (36) that the kinetics of fusion (kinetics of F activation) impacts sensitivity to inhibition by peptides, as is the case for HIV (39). Alterations in HPIV3 HN′s property of F activation affect the kinetics of F's progression through its conformational changes, thus altering inhibitor efficacy. Once the extended intermediate stage of F has passed, and fusion proceeds, peptide inhibitors are ineffective. We have proposed that the design of effective inhibitors may require either targeting an earlier stage of F activation or increasing the concentration of inhibitor at the location of receptor binding, in order to enhance the access and association of the inhibitor with the intermediate-stage fusion protein (36).
A substantial body of evidence supports the notion that viral fusion occurs in confined areas of the interacting viral and host membranes (26). For HIV-1, the lipid composition of the viral membrane is strikingly different from that of the host cell membrane; the former is particularly enriched in cholesterol and sphingomyelin (4, 5, 7, 8). Cholesterol and sphingolipids are often laterally segregated in membrane microdomains or “lipid rafts” (7, 11). In fact, the antiviral potency of the HIV-inhibitory HRC peptide C34 is dramatically increased by targeting it to the “lipid rafts” via the addition of a cholesterol group (16).
We applied the targeting strategy based on cholesterol derivatization to paramyxoviruses, and we show here that by adding a cholesterol tag to HPIV3-derived HRC E459V (30) inhibitory peptides, we increased antiviral potency by 2 log units (50% inhibitory concentrations [IC50], <2 nM). We chose to use the HPIV3-derived peptides for HeV/NiV, because we have previously shown that they are far more effective inhibitors of HeV and NiV than the homotypic peptides (30, 31). We propose that the enhanced activity resulting from the addition of a cholesterol tag is a result of the targeting of the peptide to the plasma membrane, where fusion occurs.
Anti-HRC antibodies were custom generated in rabbits by Invitrogen using the previously described HPIV HRC sequence (31).
Transfections were performed according to the manufacturer's protocol for Lipofectamine and Plus or for Lipofectamine 2000 (Invitrogen).
293T (human kidney epithelial), Vero (African green monkey kidney), and CV1 cells were grown in Dulbecco's modified Eagle's medium (DMEM) (Cellgro; Mediatech) supplemented with 10% fetal bovine serum (FBS) and antibiotics under 5% CO2. The effect of peptides on the number of HPIV3 plaques was assessed by a plaque reduction assay performed as described previously (21). Briefly, CV1 cell monolayers were inoculated with 100 to 200 PFU of HPIV3 in the presence of various concentrations of peptides. After 90 min, 2× minimal essential medium containing 10% FBS was mixed with 1% methylcellulose and added to the dishes. The plates were then incubated at 37°C for 24 h. After removal of the medium overlay, the cells were immunostained for plaque detection. The number of plaques in the control (no peptide, or a scrambled peptide) and experimental wells were counted under a dissecting stereoscope.
CV1 cells were infected with wild-type SV5 (ATCC VR-288 [strain 21005-2WR]) at a multiplicity of infection (MOI) of 3 × 10−4 in the presence of increasing concentrations of the C-terminally tagged HPIV3 HRN peptide. After a 180-min incubation at 37°C, cells were overlaid with methylcellulose. At 36 h postinfection, the overlay was removed, and the plaques were immunostained and counted.
NiV was isolated in Vero cells from the brain of a human fatally infected in the 1998-to-1999 Malaysian outbreak and was passaged three times in Vero cells, double plaque purified, and passaged a further three times in Vero cells as previously described (41). HeV was isolated in Vero cells from the lung of a horse infected in the Brisbane, Australia, outbreak in October 1994 and was passaged five times in Vero cells, followed by triple plaque purification and a further five passages in Vero cells as previously described (15). HeV and NiV stock titers were adjusted to 1 × 106 50% tissue culture infective doses (TCID50)/ml. For titrations, serial 10-fold dilutions of samples were made in Eagle's minimum essential medium (EMEM), and 25 ml was transferred to five wells of a 96-well microtiter plate. Vero E6 cells in EMEM containing 10% fetal calf serum (EMEM-10) were added (2 × 104 cells/well). Plates were incubated at 37°C for 5 to 7 days, and wells displaying cytopathic effect were scored as infected. The virus titer was calculated using the Reed-Muench method (38), and the limit of detection was 126 TCID50/ml virus. All work with live virus was carried out under biosafety level 4 (BSL-4) conditions.
VSV-ΔG-RFP is a recombinant vesicular stomatitis virus (VSV) derived from the cDNA of VSV Indiana, in which the G gene is replaced with the Ds-Red (red fluorescent protein [RFP]) gene. Pseudotypes with HeV or NiV G and F were generated as described elsewhere (25, 43). Briefly, 293T cells were transfected with plasmids encoding either VSV-G, HeV-GCT32/F, or NiV G/F. Twenty-four hours posttransfection, the dishes were washed and infected (MOI, 1) with VSV-ΔG-RFP complemented with VSV G. Supernatant fluids containing pseudotyped virus (HeV F/CT32-G, NiV G/F, or VSV-G) were collected 24 h postinfection and were stored at −80°C. For infection assays, HeV F/CT32-G and NiV G/F or VSV-G pseudotypes (controls) (data not shown) were used at an MOI of 0.25 to infect 293T cells. Peptides were added at various concentrations. RFP production at 36 h was analyzed by fluorescent microscopy (34) and fluorescence-activated cell sorting (FACS) (FACSCalibur; Becton Dickinson).
The effects of the peptides on plaque numbers were assessed by a plaque reduction test. Briefly, Vero cells (African green monkey kidney cells) were seeded into 24-well plates at 2 × 104 cells/1 ml and were grown to 90% confluence in EMEM-10 at 37°C under a humidified 5% CO2 atmosphere. For the experiment for which results are shown in Fig. Fig.1,1, virus dilutions were chosen to generate 50 to 80 plaques. An MOI of 0.05 was used for Fig. Fig.3.3. Peptides were diluted 10-fold in EMEM-10 under biohazard level 4 conditions and were added either at the time of infection (Fig. (Fig.1)1) or 60 min prior to infection (Fig. (Fig.3);3); they were then removed with the virus 30 min after infection. For Fig. Fig.1,1, cells were then overlaid with a 0.5% agarose solution in EMEM-10 and were incubated at 37°C for 18 h. For Fig. Fig.3,3, the medium was removed, and cells were washed with fresh medium prior to replacement with fresh EMEM-10 and incubation at 37°C for 18 h. The culture medium was discarded, and plates were immersed in ice-cold absolute methanol for at least 20 min prior to air drying outside the biohazard level 4 facility. Fixed plates were immunolabeled with monospecific anti-N antisera (18). Briefly, plates were washed in 0.01 M phosphate-buffered saline (PBS), pH 7.2, containing 1% bovine serum albumin (BSA) for 5 min. A 100-μl volume of anti-N antiserum (1:1,000 in PBS-BSA) was applied to each well and incubated at 37°C for 30 min. Plates were rinsed with PBS containing 0.05% Tween 20 (PBS-T) and were washed for 5 min in PBS-BSA. A 100-μl volume of Alexa Fluor 488-labeled goat anti-rabbit antiserum (Invitrogen, Carlsbad, CA) diluted 1:1,000 in PBS-BSA (Fig. (Fig.1)1) or 100 μl of horseradish peroxidase (HRP)-labeled goat anti-rabbit antiserum (Sigma, St. Louis, MO) diluted 1:2,000 in PBS-BSA (Fig. (Fig.3)3) was then applied to each well and incubated at 37°C for 30 min. Plates were rinsed again with PBS-T and were washed for 5 min in PBS-BSA. For Fig. Fig.1,1, wells were overlaid with glycerol-PBS (1:1) containing 1,4-diazabicyclo(2,2,2)octane (DABCO) (25 μg/ml) and were stored in the dark prior to imaging, while for Fig. Fig.3,3, 100 ml of chemiluminescent peroxidase substrate-3 (CPS-3; Sigma) diluted 1:10 in chemiluminescent assay buffer (20 mM Tris-HCl, 1 mM MgCl2 [pH 9.6]) was added. Plates were incubated at room temperature for approximately 15 min and were then read with a Luminoskan Ascent luminometer (Thermo Fisher Scientific, Waltham, MA) using 100-ms integration per well. Fluorescein isothiocyanate (FITC) immunofluorescence was visualized using an Olympus IX71 inverted microscope (Olympus Australia, Mt. Waverley, Australia) coupled to an Olympus DP70 high-resolution color camera. Image analysis was performed using AnalySIS software (Soft Imaging System GmbH, Munster, Germany) and enabled the numbers of plaques in the control and experimental wells to be counted.
Wild-type (wt) HeV G and wt HeV F in pCAGGS were gifts from Lin-Fa Wang. To generate the shortened cytoplasmic tail variant of HeV G (HeV G-CT32), an internal primer containing an EcoRI site and initiating at position 32 of the open reading frame was used for nested PCR. The primer sequence was 5′-GGAATTCGGCACAATGGACATCAAG-3′. The genes of NiV wt G and wt F were codon optimized and synthesized by GeneArt (Germany) and were subcloned into the mammalian expression vector pCAGGS by using EcoRI or XhoI and BglII.
The effect of cholesterol-tagged peptides on the HPIV3 plaque number was measured by a plaque reduction test. CV1 cell monolayers grown in 24-well plates were incubated for 60 min with various/increasing concentrations of the untagged or N- or C-terminally tagged HPIV3 HRC peptide. The cells were washed and infected with HPIV3 for 90 min at an MOI of 5 × 10−4. After 90 min, 2× minimal essential medium containing 10% fetal bovine serum was mixed with methylcellulose and added to the plates, which were then incubated at 37°C for 24 h. After removal of the methylcellulose overlay, the cells were immunostained for plaque detection. The numbers of plaques in the control and experimental wells were counted under a dissecting stereoscope. The IC50 (concentrations required for a 50% decrease in the plaque number) were determined from the graphs by plotting the percentage of decrease in the plaque number versus the log inhibitor concentration.
CV1 cell monolayers grown in 24-well plates were incubated for 60 min with various concentrations of the N- or C-terminally tagged or untagged HPIV3 HRC peptide. The monolayers were washed with Opti-MEM (source), fixed for 10 min with 4% formaldehyde in PBS, blocked with PBS-3% bovine serum albumin, and incubated with an anti-HRC polyclonal antibody (dilution, 1:500). After a 60-min incubation at room temperature, the monolayers were washed three times with PBS. A horseradish peroxidase-conjugated protein G antibody (Abcam) diluted 1:5,000 in PBS-1% bovine serum albumin was added, and the cells were incubated for 60 min at room temperature. The cells were washed three times with PBS and were then incubated with tetramethylbenzidine (Promega) according to the manufacturer's protocol. Absorbance measurements were taken with a Spectramax M5 ELISA reader.
CV1 cell monolayers preincubated with the N- or C-terminally tagged or untagged HPIV3 HRC peptide as described above were washed with Opti-MEM and were then incubated with 10 mM methyl-β-cyclodextrin (MBCD; Sigma). For cholesterol replenishment, after 30 min at 37°C, the cells were washed with Opti-MEM; they were then incubated with 10 mM water-soluble cholesterol (Sigma) for 15 min at 37°C. The concentration of cell surface cholesterol-tagged peptides and their effect on HPIV3 infection were measured as described above.
All peptides were produced by standard 9-fluorenylmethoxy carbonyl (Fmoc) solid-phase methods. The cholesterol moiety was attached to the peptide via a chemoselective reaction between the thiol group of an extra cysteine residue, added C- or N-terminally to the sequence, and a bromoacetyl derivative of cholesterol, as previously described (16).
We have previously shown that a 36-amino-acid peptide corresponding to the HRC region of HPIV3 F effectively inhibits infection by henipaviruses (HeV and NiV). The inhibitory activity of the HPIV3 peptide for the heterotypic viruses HeV and NiV is far higher than that against the homotypic HPIV3 (30, 31). We have shown that for each individual virus, the inhibitory efficiency of HRC-derived peptides correlates directly with the strength of interaction between the HRC peptide and the target HRN domain (30). However, comparison between viruses revealed that although the interaction of HPIV3-derived HRC peptides is strongest with the HPIV3 HRN target, these peptides are relatively poor inhibitors of HPIV3 infection (30). We recently showed that the kinetics of HN-mediated activation of F to its final fusion-ready state modulates the efficacy of fusion inhibitors. The high rate of HPIV3 F activation by HN is a key determinant of the decreased efficacy of HRC peptides for HPIV3 (36) and renders these peptides relatively ineffective for HPIV3.
To address this problem, we took advantage of an advance recently made for HIV entry inhibitors (16). The addition of a cholesterol moiety to an HIV gp41-derived HRC peptide was shown to greatly increase the peptide's inhibitory potency against HIV (16). Starting with our most potent HPIV3 HRC peptide (36 residues in length with E459V ), we synthesized peptides bearing a cholesterol tag at either the C terminus or the N terminus. We assessed the inhibitory efficacies of these peptides compared to that of the untagged peptide against wt HPIV3, HeV and NiV pseudotyped virions, and live HeV and NiV (Fig. (Fig.1),1), and we found that viral entry is inhibited more effectively by cholesterol-tagged peptides than by untagged peptides. Viral entry is assessed here by a plaque reduction assay, in which each plaque (or, in the case of pseudotyped viruses, each single infected cell) represents the entry of a single infectious particle (30). For wt HPIV3, the presence of cholesterol at either the C terminus or the N terminus increases peptide efficiency by 2 log units (Fig. (Fig.1A).1A). For the cholesterol-tagged peptides, the IC50 is ~7 nM, while for the untagged peptide, the IC50 is approximately 100-fold higher at ~700 nM. Furthermore, the cholesterol-tagged HRC peptides achieve complete inhibition of HPIV3 viral entry, while the untagged peptides do not fully block viral infection, even at the highest concentrations tested.
The inhibitory activities of these three peptides were next assessed for HeV (Fig. (Fig.1B)1B) and NiV (Fig. (Fig.1C)1C) pseudotyped virions. Interestingly, for both pseudotyped viruses, the IC50 of the C-terminally tagged peptide dropped to the picomolar range, while the inhibitory activity of the N-terminally tagged peptide was virtually indistinguishable from that of the untagged peptide. We confirmed these results in plaque reduction assays with live HeV (Fig. (Fig.1D)1D) and NiV (Fig. (Fig.1E)1E) and found that the C-terminally tagged peptides outperformed the N-terminally tagged peptides over a range of concentrations. The IC50 of the C-terminally tagged peptide for HeV and NiV are similar to each other at ~10 nM, 15 to 20 times lower than the previously reported IC50 for untagged peptides (179 nM for HeV and 208 nM for NiV ). Taken together, the data in Fig. Fig.11 indicate that while the addition of cholesterol to the C terminus of this HPIV3-derived HRC peptide increased the overall efficacy of this entry inhibitor, the addition of the cholesterol moiety to the N terminus increased the peptide efficacy only for HPIV3, not for the henipaviruses. The henipavirus result is similar to that reported for HIV, where N-terminal tagging of the peptide with cholesterol led to a 50-fold decrease in antiviral activity, consistent with the idea that insertion into the membrane in that orientation drastically decreases the ability of the peptide to interact with its HRN target (16).
We hypothesized that the effectiveness of the N-terminally tagged peptide against HPIV3 but not the henipaviruses might be due to the difference in the kinetics of F activation between the two viruses; the kinetics of F activation might influence efficacy, even for the tagged peptides. To test this hypothesis, we used HPIV3 HN variants with mutations affecting their F-triggering abilities: an HN with a N551D mutation in receptor binding site II, more efficient at F triggering than the wild type (28, 32), and an HN variant with P111S and D216N mutations (P111S in the HN stalk and D216N in the primary binding site), which has impaired F triggering (35). We assessed the inhibitory activities of the wt and N- and C-terminally tagged peptides against these HPIV3 HN variants and found that while the cholesterol-tagged peptides outperform the untagged peptides for every virus (Table (Table1),1), the faster kinetics conferred by the HN N551D mutant causes partial resistance to peptide inhibition, and the slower kinetics of the P111S D216N mutant allows for better inhibition. In particular, the slower kinetics of the P111S D216N HN variant renders it susceptible to inhibition by the N-terminally tagged peptide. This suggests that the different effectiveness of tagging at the N versus the C terminus is influenced by the kinetics of F activation and that cholesterol tagging enhances inhibition in every case.
We hypothesized that the enhanced efficacy observed after the addition of a cholesterol moiety to the HRC peptide results from an increase in the local concentration of the peptide at the site of viral fusion. For HIV, cholesterol-tagged inhibitory peptides are thought to insert into the target cell membrane (16). Here we tested the retention of the cholesterol-tagged peptides in the target cell membrane—and their abilities to inhibit subsequent viral entry—by preincubating cells with cholesterol-tagged peptides and washing the cells to remove unbound peptide prior to infection with wt HPIV3 (Fig. (Fig.2A).2A). We found that the HPIV3 HRC peptide tagged with cholesterol at the C terminus retained the same inhibitory efficiency shown in Fig. Fig.1A,1A, with an IC50 of ~7 nM, and completely inhibited infection. However, the N-terminally tagged peptide was much less effective under these conditions, with an IC50 10 times higher than that for unwashed cells shown in Fig. Fig.1A.1A. As expected, the untagged peptide would be removed by washing and therefore was not inhibitory.
To determine the extent to which the cholesterol-tagged peptides are retained in the target cell membrane, we assessed the levels of the HRC peptide on the cell surface after incubation by immunostaining (Fig. (Fig.2B).2B). Both the C- and N-terminally tagged peptides are retained on the cell surface. At lower peptide concentrations, the values for the C-terminally tagged peptide are higher than those for the N-terminally tagged peptide, but above a concentration of 100 nM, the staining plateaus and is similar for the two peptides, suggesting that saturation has been reached.
The differences in efficacy between the C-terminally tagged and N-terminally tagged peptides shown in Fig. Fig.2A2A do not correlate with differences in peptide concentration at the cell surface. For example, at 10 nM, the C-terminally tagged peptide inhibits 60% of infection (Fig. (Fig.2A),2A), with a surface immunostaining absorbance of 0.23 (Fig. (Fig.2B).2B). The N-terminally tagged peptide reaches the same inhibition (60%) only at 150 nM (Fig. (Fig.2A),2A), with a surface immunostaining absorbance at the plateau level of 0.4 (Fig. (Fig.2B).2B). Thus, the maximal level of inhibition reached by the N-terminally tagged peptide is achieved by the C-terminally tagged peptide at a markedly lower concentration at the cell surface, and at higher concentrations, the C-terminally tagged peptide is even more effective.
To assess the role of the cholesterol tag in the results observed, we determined whether cholesterol depletion removes the “protective shield” from cells. Cells preincubated with the N- or C-terminally tagged peptide or an untagged peptide were washed with cholesterol depletion medium and were then infected with HPIV3 or immunostained for surface concentrations. We found that, after cholesterol depletion, the HRC peptides were removed from the cell surface (Fig. (Fig.2D),2D), and all antiviral activity was lost (Fig. (Fig.2C).2C). This suggests that the cholesterol moiety is essential for the formation of the antiviral shield.
For clinical utility, the ability of an antiviral agent to prevent multiple rounds of infection is key. We assessed the antiviral activity of cholesterol-tagged peptides against the multicycle replication of live HeV (Fig. (Fig.3).3). At concentrations as low as 10 pM, the C-terminally tagged peptide achieved 30% inhibition of viral replication; at 100 pM, inhibition increased to more than 80%, and at 1 nM, no viral spread could be detected. The N-terminally tagged peptide, however, inhibited viral replication by only 40% at the highest concentration (1 nM). In the multicycle replication assay for which results are shown here, cell monolayers were not covered with an overlay, and the extent of viral spread through the monolayer was measured by a chemiluminescent assay (2). Whereas the experiments for which results are shown in Fig. Fig.11 measured inhibition of viral entry alone, these results suggest that the production of infective virions is impaired, and they thus demonstrate the inhibitory efficiency of the C-terminally tagged peptide for the ongoing replication process.
We next assessed the cholesterol-tagged HPIV3 peptides for broader antiviral efficacy. The N- and C-terminally tagged and untagged peptides were assessed for the ability to inhibit SV5 (HPIV5) (Fig. (Fig.4).4). While the N-terminally tagged and untagged peptides did not inhibit SV5 at the range of concentrations used, the HPIV3-derived C-terminally cholesterol-tagged peptides exhibited a dose-response inhibitory effect. For SV5, the IC50 of the C-terminally tagged peptide was ~5 nM, and 100% inhibition was reached at concentrations above 100 nM. Specificity was demonstrated by the failure of the HPIV3-derived C-terminally cholesterol-tagged peptide to inhibit influenza virus (A/H3N2), NDV-B1, and VSV. Even at the highest concentration tested, 10 μM, no inhibition was observed. These data suggest that cholesterol tagging expands the potential range for heterotypic virus inhibition.
Acute respiratory infection is now the leading cause of mortality among children under the age of 5 years, accounting for nearly one-fifth of childhood deaths worldwide. Human parainfluenza viruses (HPIV) and respiratory syncytial virus (RSV) cause the majority of childhood croup, bronchiolitis, and pneumonia cases. Despite the huge impact of these diseases on illness and hospitalization of infants worldwide, no drugs or vaccines are available, and the development of antiviral drugs for these viruses has been a great challenge.
Fusion-inhibitory peptides that correspond to a number of viral fusion (F) proteins can block fusion intermediates during viral entry and inhibit infection. For HIV, one such peptide is a clinically effective inhibitor of HIV-1 fusion (enfuvirtide). However, this approach has been hindered by the relatively low potencies and short half-lives of these peptides in vivo. We have shown that peptides derived from the fusion protein of HPIV3 are effective inhibitors of both parainfluenza virus and henipavirus infection (31, 36). We have shown that the efficacy of peptide inhibitors for paramyxoviruses depends on two variables: the strength of interaction of the peptide with the target fusion protein and the time window of access to the target sequence (i.e., the kinetics of triggering). Here we add a third important variable: location of the peptide in proximity to the target fusion protein. We discovered that if these peptides are targeted to the correct cellular compartment—the membrane where fusion/entry occurs—they become highly effective inhibitors. However, correct orientation of the peptide on the target membrane also plays a key role in inhibitory efficacy.
Our findings are in line with a recent study (1) showing that the antiviral activities of two broadly neutralizing anti-HIV antibodies (2F5 and 4E10) critically depend on the concomitant presence of a lipid binding epitope—harbored in an unusually long CDR3—and a gp41-binding epitope. These antibodies do not target the untriggered, prefusion state of gp41 and are thought to bind to the prehairpin intermediate (13). Mutations in the CDR3 that abolish lipid binding, although they have no effect or only a minimal effect on gp41 binding affinity, completely eliminate the ability to neutralize the virus. The authors propose a two-step mechanism, with the antibodies binding first to the membrane through their CDR3 loop and then to gp41 after triggering and formation of the prehairpin structure. The time window before this structure collapses to the postfusion 6HB is estimated to be only 15 min (24), and preconcentration of the antibody at the site of action becomes a key precondition for effective capture and neutralization. If this view is correct, cholesterol tagging could be the equivalent of the lipid-binding CDR3 structures. Its potency-enhancing effect, therefore, would be expected to correlate with the kinetics of fusion and hence to be variable for the different viruses.
While the C-terminally tagged peptide is highly effective against several paramyxovirus pathogens we tested, the N-terminally cholesterol-tagged HPIV3 HRC peptide is effective only against HPIV3. Both N- and C-terminally tagged peptides insert into the target cell and retain efficacy after washing; however, only the C-terminally tagged peptides—while membrane anchored—interact with heterotypic viruses to inhibit fusion. The basis for this distinct mechanism of action is of interest, in light of the activities of the C-terminally tagged HPIV3 peptide against several different viruses. One hypothesis to explain the efficacy of the C-terminally tagged peptide relative to that of the N-terminally tagged peptide assumes that the increase in local concentration due to the hydrophobic tail lowers the stringency of the requirement for a specific sequence, provided that the inhibitor retains sufficient interaction with the HRN domains. If a specific orientation of the peptide on the cell surface is required to block the refolding of F, as has been proposed for HIV (16), only the C-terminally tagged peptide would be properly oriented to maximize the interaction with the HRN. However, for this particular sequence, the N-terminally tagged peptide, in reverse orientation, would still retain sufficient affinity to bind to the homotypic HRN of HPIV3, but not to the heterotypic HRN of HeV/NiV.
Figure Figure5A5A illustrates our proposed explanation for the inhibitory activity of the N-terminally tagged HPIV3 peptide for HPIV3. This peptide, when driven by cholesterol into parallel alignment with the HRN—instead of the usual antiparallel orientation—can still establish some interactions with the HRN. Figure Figure5A5A shows the HPIV3 sequence in both the standard N- to C-terminal and the opposite C- to N-terminal direction. While parallel alignment would result in a considerably weaker interaction with HRN, the addition of cholesterol could still rescue the peptide's inhibitory activity, as shown in the case of HIV for inactive analogs of T20 (29) and C34 (P. Ingallinella et al., personal communication). Accordingly, in the more stringent washout experiment (Fig. (Fig.2),2), the N-terminally tagged peptide is 10-fold less potent than the C-terminally tagged peptide, despite equivalent binding to the target cells.
The ability of cholesterol moieties to rescue the inhibitory activity of weakly interacting peptides can also explain the efficacy of these peptides for PIV5, as shown in Fig. Fig.4.4. Figure Figure5B5B shows an alignment of the HPIV3 and PIV5 sequences, revealing the high degree of sequence homology, and in particular the almost perfect conservation of the critical amino acids in the “a” and “d” positions. In this case also, cholesterol tagging is expected to boost the inhibitory efficacy of the peptide as a result of heterotypic binding with the PIV5 HRN.
The inhibitory efficacies of peptides with C-terminal cholesterol were 15- to 20-fold higher than those of untagged peptides for HeV and NiV but 100 times higher for HPIV3. As a possible explanation, we propose that the potency of fusion inhibitors, like that of other inhibitors that bind to transiently exposed targets, can be kinetically rather than thermodynamically driven. The level of inhibition is strongly influenced by kinetic parameters such as the lifetime of the sensitive state and the rate of inhibitor association (42) in such a way that, beyond a certain level of potency, the equilibrium binding strength plays a diminished role. Since we propose that the mechanism of action of cholesterol tagging is to increase the rate of inhibitor association with the prehairpin intermediate, via preconcentration in the right membrane compartment, its beneficial effects are expected to be greater for weakly binding than for tightly binding inhibitors.
It is possible that peptide-resistant paramyxovirus variants with faster F-triggering kinetics may emerge under the selective pressure of peptide inhibitors. For HIV, clinical use of the T20 inhibitory peptide resulted in the emergence of drug-resistant HIV variants (39, 42). The resistance was due either to decreased interaction between the HRN region and T20 or to increased interaction between viral HRN and HRC. Increased kinetics of fusion led to resistance but also led to viruses whose growth depended on the presence of drug. While we anticipate that anti-HPIV therapy will be of shorter duration than anti-HIV therapy, resistance would be important clinically, as for influenza. For HPIV3, enhanced fusion kinetics led to partial resistance to HRC peptide inhibition in fusion assays in vitro (36). However, we have shown that mutations that affect fusion kinetics have a significant impact on the phenotype in human airway tissues, indicating that these differences are likely to impact fitness significantly in vivo (27). The variant viruses with faster fusion kinetics had a severe growth disadvantage in natural host tissues (cultured human airway epithelium [HAE]) and in an animal model (cotton rat) (27). For HPIV3, therefore, we propose that viral evolution toward faster fusion kinetics in the presence of peptide inhibitors would likely not yield fit, transmissible resistant viruses, and this hypothesis will be tested. Our experiments suggest that in HAE, the specific balance between viral receptor binding, receptor destruction (neuraminidase), and fusion can be accurately assessed. This will be important for future studies that address the potential for the emergence of peptide-resistant variants.
We hypothesize that the strategy of peptide insertion into the cell membrane may enable the use of fusion-inhibitory peptides for those viruses that fuse in the cell interior. Future studies will test the idea that cholesterol tagging of peptides allows for the inhibition of viruses that fuse at intracellular locations, including influenza virus. The major significant challenge in developing fusion inhibitors for influenza virus has been that by the time the influenza virus fusion protein (HA) acts, it is already sequestered in an intracellular compartment. We propose that lipid-tagged peptides may accompany the membrane-attached HA from the cell surface to the site of fusion activation, thus overcoming this problem.
Future studies will address the in vivo efficacy of membrane-targeted inhibitory peptides. Membrane targeting of antiviral peptides offers a new approach to the development of highly effective peptide fusion antivirals that inhibit important pediatric respiratory pathogens, as well as emerging paramyxoviruses. Peptides that prevent viral entry are likely to be safe antivirals for infants and young children, and since the paramyxovirus family encompasses prevalent and particularly infectious viruses for children, including (besides HPIV) RSV, measles virus, and mumps virus, this research has broader implications for pediatric antiviral development.
This work was supported by a March of Dimes research grant to A.M., Public Health Service grants AI076335 and AI31971 from the National Institutes of Health (NIAID) to A.M., and NIH (NIAID) Northeast Center of Excellence for Bio-defense and Emerging Infectious Disease Research grant U54AI057158 to A.M. and M.P. (principal investigator, W. I. Lipkin). We are grateful to Ashton Kutcher and Jonathan Ledecky for their support, to Dan and Nancy Paduano for support of innovative research projects, and to the Friedman Family Foundation for the renovation of our laboratories at Weill Cornell Medical College.
We acknowledge flow cytometry support from Stanka Semova and Sergei Rudchenko in the Flow Cytometry Facility of the Hospital for Special Surgery/Weill Cornell Medical College and technical support from Elisabetta Bianchi and Gennaco Ciliberto. We thank GTx, Inc. for the kind gift of VSV-ΔG-RFP VSV-G. HeV wt G and wt F in pCAGGS were gifts from Lin-Fa Wang.
Published ahead of print on 31 March 2010.