ABSTRACT
Capsid assembly during virus replication is a potential target for antiviral therapy. The Gag polyprotein is the main structural component of retroviral particles, and in human immunodeficiency virus type 1 (HIV-1), it contains the sequences for the matrix, capsid, nucleocapsid, and several small polypeptides. Here, we report that at a concentration of 100 μM, 7 of 83 tripeptide amides from the carboxyl-terminal sequence of the HIV-1 capsid protein p24 suppressed HIV-1 replication (>80%). The three most potent tripeptides, glycyl-prolyl-glycine-amide (GPG-NH2), alanyl-leucyl-glycine-amide (ALG-NH2), and arginyl-glutaminyl-glycine-amide (RQG-NH2), were found to interact with p24. With electron microscopy, disarranged core structures of HIV-1 progeny were extensively observed when the cells were treated with GPG-NH2 and ALG-NH2. Furthermore, nodular structures of approximately the same size as the broad end of HIV-1 conical capsids were observed at the plasma membranes of treated cells only, possibly indicating an arrest of the budding process. Corresponding tripeptides with nonamidated carboxyl termini were not biologically active and did not interact with p24.
In all virus particles, the genome is encapsulated in a protein shell which is formed by the condensation of protein subunits in an ordered fashion to form either a helical or icosahedral capsid or, as in the case of human immunodeficiency virus type 1 (HIV-1), a conical capsid structure (21, 28). The assembly of the capsid is thus a potential target for antiviral therapy. HIV-1 assembles at the plasma membrane of infected cells and buds off as immature spherical particles containing an electron-dense shell composed of the Gag and Gag-Pol polyproteins surrounded by the envelope membrane. These particles must undergo maturation to become infectious. This is accomplished by the virus protease present in the immature virus particles, which cleaves the Gag and Gag-Pol polyproteins into individual protein components. The Gag polyprotein is the main structural component of retroviral particles, and in HIV-1, it contains the sequences for the matrix (p17, MA), capsid (p24, CA), nucleocapsid (p7, NC), and p6 proteins and two spacer peptides called spacer peptide 1 (SP1) and spacer peptide 2 (SP2). Upon maturation, p24 is released from the matrix protein and condenses into a conical capsid that encases and stabilizes the ribonucleoprotein complex (24, 27), which also includes the enzymes needed for virus replication (28). The mature conical capsid of HIV-1 contains 1,500 to 2,000 copies of p24. Capsid protein p24 is composed of approximately 230 amino acids, is highly hydrophobic, and contains two distinct α-helical domains that are connected by a flexible linker (5, 28, 30). The amino-terminal domain (residues 1 to 146) binds cyclophilin A (CyPA) (4, 13, 14) and participates in capsid formation (8, 33, 43). The second carboxyl-terminal domain (residues 148 to 231) is involved in p55 Gag polyprotein assembly and the budding of virus particles (3, 6, 8, 41, 44). Within the carboxyl-terminal domain is a region conserved among lentiviruses known as the major homology region. Deletions in this region of the carboxyl terminal abolish the ability of HIV-1 to form infectious viral particles (8). Furthermore, an interaction between SP1 and the carboxyl-terminal domain of p24 is probably necessary, at least transiently, in order to weaken CA-CA interactions and allow the rearrangement of the capsid shell during virus maturation (20, 46). It has also been shown that extensions at the N terminal of the capsid protein with sequences corresponding to the matrix protein redirect in vitro protein assembly from cylinders to spheres (9, 19, 42). These results together with those showing that capsid cones can be formed in vitro from CA-NC fusion proteins (16) indicate that release of the capsid from the matrix protein is necessary for capsid cone formation. Moreover, the introduction of mutations in the amino-terminal domain of p24 has produced viruses with retained ability to assemble and bud but which developed aberrant capsid morphologies (8, 33, 43). The formation of a conical capsid appears to be essential, because gag mutations that disrupt proper core formation invariably lead to a reduction or loss of viral infectivity (8, 17, 33, 43, 46).
Immature capsids are stable and can be isolated easily, whereas mature capsids are less stable and have only recently been successfully purified from infectious particles by mild detergent treatment and centrifugation (45). Therefore, it appears that the maturation process prepares the virus particle for infection by facilitating the disassembly of the capsid structure upon virus entry into the target cell. Substances that stabilize capsid structures and inhibit uncoating are well known inhibitors of picornavirus replication (2, 18, 32, 34).
Thus, substances that interfere with the proper assembly of p24 could be potential antiretroviral agents. In this regard, it has been shown that a 10-amino-acid-long peptide derived from the p24 amino acid sequence can block virus replication by interfering with capsid formation (31).
Here, we report that tripeptides derived from the carboxyl-terminal sequence of the HIV-1 capsid protein p24 were found to interfere with HIV-1 morphogenesis and to suppress viral replication.
MATERIALS AND METHODS
Peptide synthesis.To manufacture screening peptides, solid-phase peptide synthesis was performed as described previously (25). The peptides were carboxyl-terminal amides (CONH2), i.e., the hydroxyl group was replaced by an amide group. For all experiments except the initial screening, peptides GPG-NH2, ALG-NH2, CQG-NH2, RQG-NH2, and ALGPG-NH2 were obtained by custom order from Bachem Feinchemikalien AG (Bubendorf, Switzerland), as were peptides GPG-OH and ALG-OH, which have normal carboxyl termini (COOH). [1-14C]glycyl-prolyl-glycine-amide (2.5 mCi/ml with a specific activity of 56 mCi/mmol) was custom ordered from Amersham Pharmacia Biotech (Uppsala, Sweden).
Viruses, cells, and infections.HIV-1 SF-2 stock was prepared from HUT78 cells, and two clinical isolates were prepared from donors' peripheral blood mononuclear cells (PBMC). Fifty percent tissue culture infectious doses (TCID50) were prepared as described previously (38). All T-cell lines were kindly provided by the NIBSC AIDS Reagent Project, National Institute for Biological Standards and Control, Potters Bar, United Kingdom. HUT78, H9, and ACH-2 (an HIV-1 chronically infected human T-cell line [12]) cells were propagated and maintained in RPMI 1640 medium (GIBCO Laboratories) supplemented with 10% (vol/vol) heat-inactivated fetal bovine serum (GIBCO Laboratories), penicillin, and streptomycin (100 U/ml; Sigma, St. Louis, Mo.). PBMC were purified by Ficoll-Hypaque density gradient centrifugation and stimulated with phytohemagglutinin (KEBO Lab, Stockholm, Sweden) for 3 days in RPMI 1640 medium supplemented as described above before use.
For peptide screening, 105 H9 cells were infected with HIV-1 SF-2 (25 TCID50) with or without 100 μmol of the peptides. After virus adsorption for 1 h at 37°C, the cells were washed three times in RPMI medium and then resuspended in culture medium with or without the peptides. Supernatants were collected on days 4, 7, and 11 postinfection, and the medium was refreshed in the presence (100 μM) or absence of the peptides. HIV-1 p24 antigen (ELISA kit; Abbott Laboratories, North Chicago, Ill.) and reverse transcriptase activity (Lenti RT kit; Cavidi AB, Uppsala, Sweden) in growth medium were assayed on the collected supernatants (36). Infectivity assays of the HIV-1 clinical isolates were performed as described above in PBMC stimulated for 3 days with 2.5 μg of phytohemagglutinin (Difco, Franklin Lakes, New Jersey)/ml and infected at 25 TCID50/2 × 105 cells with or without different concentrations of peptides GPG-NH2, ALG-NH2, and ALGPG-NH2 in the medium. Medium containing different concentrations of the peptides was refreshed on days 4 and 7 postinfection. The production of p24(37) in the culture supernatants collected on days 7 and 11 was measured to monitor viral replication by use of the Abbott ELISA kit.
Transmission electron microscopy (TEM) of HIV-1 assembly.Virus-infected cells were fixed by freshly made 2.5% glutaraldehyde in phosphate buffer and postfixed in 1% osmium tetroxide. The cells were embedded in Epon and poststained with 1% uranyl acetate. Sections were made to be approximately 60 nm thick to accommodate the volume of the core structure parallel to the section plane. Duplicate sample preparations were done so that possible volume changes during polymerization and the capacity to withstand electron beam exposure could both be estimated. Specimens were analyzed with a Zeiss CEM 902 electron microscope, equipped with a spectrometer to enhance image contrast, at an accelerating voltage of 80 kV. A liquid nitrogen-cooling trap for the specimen holder was used throughout. Statistical evaluation of morphology was done with a series of electron micrographs to depict different categories of virus morphology, specifically focusing on the packing of the virus core structure. A test for the difference between two population proportions was also included.
Three-dimensional (3D) visualization of the internal structure of HIV was elicited from several TEM projections taken at evenly spaced tilt angles and subjected to computer processing. Duplicate sample preparations were done so that possible volume changes during polymerization and the capacity to withstand electron beam exposure could both be estimated. Furthermore, the Epon sections were covered by an approximately 1-nm-thick carbon layer for protection. Approximately 60-nm-thick sections were cut and mounted on 400-mesh copper grids. The specimens were poststained with lead citrate. Ten-nanometer average-diameter gold particles were applied as fiduciary marks to align the images for computer analysis. The specimens were analyzed in a Zeiss CEM 902 system with a goniometer stage at 80 kV. The spectrometer unit was used to improve image quality, especially at high tilt angles. The minimal beam dose technique was employed throughout. Forty projections were obtained in each tilt series taken from −60° to +60°. All tilt series were carefully scrutinized for visible beam damage by comparison of images taken before and after the tilt series. Series of micrographs demonstrating no visible changes were further digitized with a photoscanner. Image alignment and 3D reconstruction were done essentially as described before (24; A. Höglund, personal communication). The reconstructed virus models were viewed on a computer-controlled display (Silicon Graphics O2 workstation) as vector models.
Volume images that are obtained by 3D reconstructions are important for the evaluation of virus structure. In two-dimensional (2D) micrographs, it is difficult to determine whether two overlapping objects are present, something which is clearly discernible in 3D images. Another topological feature that can be revealed in volume images, but not in 2D micrographs, is the presence of cavities in the particles, i.e., it can be determined whether the particles are solid or hollow. The 3D reconstructions also reveal surface structures of the particles that cannot be observed easily with 2D micrographs. These may be visualized by common rendering techniques (26).
The samples used for immunocytochemical analysis were fixed in 4% formaldehyde and 0.2% glutaraldehyde in phosphate-buffered saline. The rabbit anti-HIV p24 polyclonal antibody used for primary immunolabeling was obtained from an Abbott p24 ELISA kit (at dilutions of 1:5 and 1:10). For secondary immunolabeling, a goat anti-rabbit gold (5 nm diameter)-conjugated antibody was used (dilutions of 1:500 and 1:1,000; BBI International, Cardiff, United Kingdom). Low image contrast is obtained by this method irrespective of poststaining.
Affinity capillary electrophoresis (ACE).An interaction between the tripeptides and p24 is likely to be attended by a change of the charge of the peptide. Therefore, electrophoresis ought to be the method of choice for the study of these possible interactions, particularly in the capillary mode, since only minute amounts of material are required for a rapid high-resolving analysis. Capillary electrophoresis permits calculation of the electrophoretic mobility of a charged species. The electrophoretic mobility of a given species, which is related to its net surface charge density, can be used to characterize and identify a substance (7). In this study, we used the partial-filling technique, employing a non-UV-absorbing buffer and injecting a zone of the slow UV-absorbing constituent first and then a zone of the fast UV-absorbing constituent (1, 22, 40), with the HIV-1 capsid protein p24 being the slow component and five tripeptides being the fast components.
Fused silica tubing (inner diameter, 50 μm) was purchased from MicroQuartz (Munich, Germany) and cut to a length of 23 cm (effective length, 18.5 cm). A BioFocus 3000 capillary electrophoresis system (Bio-Rad Laboratories, Hercules, Calif.) was used in all experiments. The internal wall of the capillary was coated prior to use with 5% (wt/vol) linear polyacrylamide (23) in order to suppress the electroendosmotic flow and to prevent the adsorption of the proteins onto the capillary wall. Sodium phosphate solutions (0.01 M) at pH values between 6.8 and 8.2 were used as buffers. The polarity was negative (the detection point was closer to the cathode) for GPG-NH2, ALG-NH2, and RQG-NH2 at pH values between 6.8 and 7.5 (see Fig. 2a) but positive (the detection point was closer to the anode) at pH 8.2, which was also the case for GPG-OH and ALG-OH (see Fig. 2b). The peptides have low UV absorbance and were therefore dissolved in the buffer at a relatively high concentration (0.5 mg/ml) to permit online monitoring by a standard UV detector. The stock solution of p24 was diluted 10-fold with the running buffer to a final concentration of 50 μg/ml. Hemoglobin was prepared from human red blood cells from a healthy donor following the method described by Molteni et al. (29). The normal hemoglobin variants in blood have isoelectric points (pI) between 6.7 and 7.4, with the major component having a pI of 6.9. The concentration of the hemoglobin solution was much higher than that of p24.
The capillary was filled with the buffer. The protein was injected by pressure (10 to 60 lb/in2), and then the peptide was injected (1 lb/in2). Since the electrophoretic migration velocity of the peptides was higher than that of the protein, the peptide molecules moved through the protein zone.
Dialysis of [14C]GPG-NH2 versus p24.Fifty microliters of 10 μM solutions of recombinant protein p24 (NIBSC AIDS Reagent Project) or, as controls, recombinant gp120 (NIBSC AIDS Reagent Project) or bovine serum albumin (BSA; Sigma) was dialyzed, using a 10-kDa cutoff dialysis cassette (Slide-A-Lyzer; Pierce), against 27.5 μM [14C]GPG-NH2 in 500 ml of 150 mM NaCl and 50 mM Tris-HCl, pH 7.4, buffer at 4°C for 2 days. Radioactivity was quantified in a Rackbeta 1218 instrument (LKB-Wallac) after mixing 10 or 5 μl of the protein solutions with 3 ml of ReadySafe (Beckman).
RESULTS
Screening of the peptides corresponding to the C terminus of p24 for antiviral activity.The overlapping tripeptides corresponding to the C-terminal domain of HIV-1 p24 (residues 146 to 231) were tested at a concentration of 100 μM for their ability to inhibit viral replication in H9 cells. The results are shown in Fig. 1a. Besides GPG-NH2, 31 of the 82 tripeptides inhibited HIV-1 SF-2 replication by 50% or more. ALG-NH2 showed a strong inhibitory effect and reduced virus production by more than 95%. In fact, the reduction of viral replication obtained with ALG-NH2 at different concentrations was close to that obtained with GPG-NH2 (Fig. 1). The pentapeptide amide ALGPG-NH2, corresponding to residues 204 to 208, also inhibited HIV-1 replication (Fig. 1), albeit slightly less efficiently than the tripeptides ALG-NH2 and GPG-NH2. ALG-NH2 also inhibited the replication of two clinical isolates of HIV-1 in PBMC in a dose-dependent manner (Fig. 1).
Screening of the peptides corresponding to the C terminus of p24 for antiviral activity. (a) The HIV-1 capsid protein p24 amino acid sequence (15) secondary structure and the overlapping tripeptides were tested for antiviral activity. Peptides with inhibitory effects, as determined by p24 production and reverse transcriptase activity in the culture supernatants, of >90% are indicated in red, of 80 to 89% are indicated in blue, and of 50 to 79% are indicated in green. The regions in the sequence of p24 shown with waved shading represent helical structures. MHR, major homology region. (b) The antiviral activities of peptides ALG-NH2 and ALGPG-NH2 were compared with that of GPG-NH2 in HUT78 cells infected with HIV-1 SF-2 and in PBMC infected with two clinical isolates. All results presented are the means of triplicate cultures.
Binding of GPG-NH2 to p24.The amidated peptides are positively charged at and below pH 8.2, except for ALG-NH2, which is not charged and hence does not move at pH 8.2. The two tripeptides with nonamidated carboxyl termini (GPG-OH and ALG-OH) and p24 migrated very slowly in the capillaries at pH 7.0. The isoelectric points of these substances should, therefore, be around 7.
The migration times of the peptides were determined at pH values of 6.8, 7.5, and 8.2 in the absence (t1 values in Table 1) and presence (t2 values in Table 1) of the protein. Figure 2 shows four typical electropherograms. As a measure of the interaction, we used the relative difference between the mobilities (RDM) in the presence and absence of protein {[(u1 − u2)/u1] × 100}. Note that the parameter RDM is not equivalent to the capacity factor since the peptides do not interact with p24 during the entire migration time. Accordingly, in terms of RDM, GPG-OH and ALG-OH showed almost no affinity to recombinant p24 whereas GPG-NH2, ALG-NH2, and RQG-NH2 interacted with it. Hemoglobin was chosen as a control protein because it has an isoelectric point similar to that of the capsid protein. In addition, hemoglobin (molecular weight, 68,000) has almost three times more amino acid residues than p24, which increases the possibility for unspecific interactions with tripeptides. In the series of experiments where p24 was replaced by hemoglobin, no differences in the migration times of the peptides were observed (Table 1). The three amidated tripeptides at all pH values tested (6.8 to 8.2) gave increased RDM values, indicating an interaction with p24 (Table 1). The two nonamidated tripeptides did not show any interaction with p24 at pH 8.2. We used this pH because the electrophoretic mobilities of these peptides at pH 6.8 and 7.5 are too low to permit interaction studies.
Capillary electrophoresis analysis of the tripeptides in the absence and presence of HIV-1 p24. The migration of GPG-NH2 in the presence of p24 (II) was retarded compared with the migration in the absence of the protein (I), whereas the migration time of GPG-OH was the same in the presence (IV) and in the absence (III) of p24. The positions of the starting zones in the capillary are indicated below each electropherogram.
ACE analysis of the interaction between tripeptides and HIV-1 p24 and human hemoglobin (used as control protein)a
The ability of GPG-NH2 to bind to the HIV-1 capsid protein was also tested by dialysis experiments. 14C-labeled tripeptide was added to the dialysis buffer outside the dialysis chamber containing p24 or, as controls, envelope glycoprotein gp120 or BSA. After 2 days of dialysis, the radioactivity of the buffer inside the dialysis chamber did not differ from that found outside in the control experiments (1.7 μCi/ml for both gp120 and BSA). However, the radioactivity of the buffer containing p24 was 13.7 μCi/ml, which was 7.6 times greater than that found in the buffer outside the dialysis chamber (1.8 μCi/ml). After consecutive dialysis against buffer without GPG-NH2 for 2 days, the p24 solution still contained 30% higher radioactivity than the dialysis buffer. These dialysis experiments, along with the capillary electrophoresis experiments, indicate that GPG-NH2 binds to p24.
Evaluation of virus assembly with TEM tilt series.In the initial experiments with HUT78 cells infected with HIV-1 SF-2, GPG-NH2 was added at the time of infection. However, too few virus particles were produced to permit statistical analysis. Therefore, TEM was performed on HUT78 cells to which the tripeptides were added 6 days postinfection. The cells were then cultured for an additional 4 days before being fixed for TEM. TEM was also performed on ACH-2 cells, which already carried 1 copy of HIV-1 proviral DNA and which were stimulated to produce virus by the addition of 100 nM phorbol-12-myristate-13-acetate (PMA) to the culture medium. After stimulation for 3 days, the ACH-2 cells were harvested. The results, i.e., changes in the morphology of the virus particles (Fig. 3) (see also below), obtained with ACH-2 cultures pretreated with 100 μM concentrations of the peptides for 4 days prior to stimulation to produce virus (Fig. 3f) were similar to those obtained with cultures to which a 1 mM concentration was added on the day of stimulation. Since HUT78 cells pretreated with GPG-NH2 or ALG-NH2 produce too few virus progeny to permit statistical analysis of the changes observed by TEM, the statistical results presented here were all from cultures treated at a peptide concentration of 1 mM. As an internal reference in every experiment, TEM was performed on virus particles and HIV-1-infected cells from infected cell cultures without tripeptides. Sixty-nanometer-thick slices of sample were made to allow the accommodation of viral core structure with a proper section plane.
Morphological study of the HIV-1 virions produced from the GPG-NH2-treated HUT78 cells from tilt series by TEM and 3D reconstruction. (a) Control virions from untreated cells showing a characteristic dense, conical capsid (middle) and a round, dense disk upon tangential sectioning. (b and c) Dense nodular material is shown protruding from the viral envelope. (c and e) Dislocation is shown, with dense material in the broad end and low density in the narrow part of the viral capsid. A nodule is also shown in panel e (arrow). (d) Round, dense material is shown accommodated outside of an empty conical capsid. (f) Virions from ACH-2 cells pretreated with 100 μM GPG-NH2 also showed dislocation of viral capsid material. (g) Evaluation by 3D reconstruction of the control virion from untreated cells. The volume rendering shows a tight, cone-shaped capsid which is attached by its narrow part to the viral envelope. (h) Evaluation by 3D reconstruction of HIV-1 particle from GPG-NH2-treated cells. This volume rendering shows tight viral capsid material irregularly condensed in the broad part of the capsid. Bar, 100 nm.
HIV-1 progeny produced in the presence of GPG-NH2 showed a completely altered assembly of viral capsids (Fig. 3). In the GPG-NH2-treated cultures, only 10% of the virus particles counted had a longitudinal sectioning of the capsid with a normal, mature filled appearance compared with 58% of the particles in untreated controls (Table 2). It should be kept in mind that virus in the treated HUT78 cells had been allowed to replicate without the peptides for 6 days before treatment. In the GPG-NH2-treated cultures, capsids with dislocation of material, with a dense assembly in the broad end and low density in the narrow part of the capsid structure (Fig. 3c and e), were observed. Another striking feature of viral cores from peptide amide-treated cells was irregular packaging in variable parts of the core. Occasionally, we also observed such a dense nodular structure attached to the outer part of the viral envelope (Fig. 3b) (7% of the number of virus particles with GPG-NH2 treatment). In addition, empty viral capsids were obtained with GPG-NH2 (Fig. 3d). Condensed circular projections located in the middle of the envelope or distributed at the edge of the envelope could represent both normal capsids and capsids that were aberrant in their narrow end and that were only slightly less abundant in the treated cells. The frequencies of misassembled virus cores (Table 2) were 90% for the GPG-NH2-treated cells (similar results were obtained with ALG-NH2 treatment) and 42% for the control cells. Further evaluation of the results was accomplished by a 3D computer-modeled reconstruction from the TEM tilt series of HIV-1 particles produced in the presence or absence of GPG-NH2 (Fig. 3g and h).
Statistical evaluation of electron micrographs of sectioned HIV-1 virions prepared from infected HUT78 cells
Viral nodules.A unique feature of nodular structures was observed to be associated with the outer membranes of virus-producing cells (Table 2) (Fig. 4). These structures were found only in cells treated with GPG-NH2 (Table 2) or ALG-NH2. Most of the treated cells (74% of GPG-NH2-treated ACH-2 cells; similar results were obtained with ALG-NH2-treated cells), which carried virus particles, had these nodular structures. ACH-2 cells that were treated with 1 mM GPG-NH2 but not stimulated with PMA to produce virus did not show such nodules. Combined treatment of PMA-stimulated ACH-2 cells with GPG-NH2 (1 mM) and the protease inhibitor ritonavir (2 μM) produced no such nodules. In the latter experiments, only budding virus particles of normal appearance and immature virus particles were seen (data not shown). With a tilt series, it was shown that the dense nodules were protruding from the outer cell membrane. The size of the nodules, i.e., of the dense material of the irregular viral core, was approximately 50 nm (Fig. 4a). Occasionally, such a dense nodular structure was also observed to be attached to the outer part of the viral envelope (Fig. 3b). Using immune electron microscopy analysis, it was shown that these viral nodules, the small particles, assembled on the outer membrane and bound gold-labeled anti-p24 antibody (Fig. 4b). Furthermore, evaluation of TEM results was accomplished by a 3D computer-modeled reconstruction from TEM tilt series of virus-infected cell cultures with GPG-NH2 (Fig. 4c).
The unique nodular structures on the cell outer membrane and viral envelope in HUT78 cells treated with GPG-NH2. Shown are dense viral nodules that are approximately 50 nm in size and an HIV-1 virion with a nonhomogenous capsid. (a) By use of a tilt series (+45°, 0°, and −45° [top, middle, and bottom panels, respectively]), three different nodules are shown (indicated by arrows), two of which are condensed as budding structures of the cell membrane while the third (right) is attached to the outside of the viral envelope. (b) Two different viral nodules, prepared by the immune electron microscopy technique, are shown. A dense protruding nodular unit is shown (left) along with another nodule at the cell membrane (right). Attached gold labels of anti-p24 antibodies (arrows) are clearly visible on the nodules. (c) 3D reconstruction of a tilt series of electron micrographs. The volume rendering shows the surface of a dense viral nodule, which is shown attached to part of the cell membrane with a narrow junction. Bars, 100 nm.
DISCUSSION
Proper capsid assembly is essential for the infectivity of progeny virus particles. In the present study, we have shown that the tripeptides GPG and ALG in their amide form can interfere with the formation of HIV-1 particles with a normal conical core structure. The antiviral effect of tripeptide GPG-NH2 could possibly involve interference with the assembly of viral core structures in particles that have budded off from the infected cells. This might be the case if the tripeptide competed with Gag polyprotein and/or p24 for protein-protein interaction sites. In support of this possibility, it was found in the ACE experiments that only the amidated tripeptides interacted with p24. These results were corroborated with the results of the dialysis experiments with [14C]GPG-NH2 and p24. However, it should be emphasized that the binding of a peptide to a protein per se does not necessarily mean that the peptide interferes with a biological function of the protein. Further studies with other tripeptides and other possible HIV-1 target proteins, as well as estimations of binding constants, will be of interest. Also, the internal pH of the virus particle should be determined since it very likely differs from that of the surrounding medium. In the present study, it was not known whether the CA protein p24 in the various experiments was in the form of monomers or dimers, which are formed under certain conditions (10, 15). It should also be pointed out that whether the binding of the amidated tripeptides to the capsid protein is in fact affecting p24-p24 interactions, and thus interfering with capsid assembly, cannot definitely be deduced from the present study. Indeed, it remains to be shown that the disarranged virus particles contain the antiviral tripeptide.
It is possible that the tripeptides bind to another protein contained in the virus particle, such as CyPA, or that binding of tripeptides to p24 interferes with its binding to another protein necessary for capsid formation. Recently, it was reported that CyPA binds with high affinity to a peptide corresponding to p24 amino acids 214 to 228 (10) that contains a GP motif (residues 223 and 224), which is a prerequisite for CyPA binding (11, 14, 39, 47, 48), that coincides with the site of the carboxyl-terminal GPG (residues 223 to 224) sequence of p24. However, it has been shown previously that the dipeptide GP-NH2, in contrast to the dipeptide PG-NH2, does not inhibit HIV-1 replication (35). Furthermore, preliminary ACE studies on CyPA-GPG-NH2 interactions have shown that this protein binds GPG-NH2 less well than does p24 (our unpublished results). Further studies on the affinity of the tripeptides for both cellular and viral structural proteins might contribute to our understanding of virion morphogenesis. The observations in the present study, however, confirm our earlier findings that the effect of GPG-NH2 on HIV-1 probably occurs at a late stage of the viral replication cycle. Additionally, in vitro studies using electron microscopy and dynamic light scattering have shown that GPG-NH2 can hinder the self-assembly of tubular structures of p24 (D. van der Spoel, C. Hetényi, A. McKenzie-Hose, L. Ågren, L. Goobar-Larsson, S. Höglund, and A. Vahlne, submitted for publication).
A striking finding in the present study was the observation of dense nodular structures at the plasma membrane of HIV-1-infected cells treated with GPG-NH2 or ALG-NH2. No such structures were seen in untreated infected cells or in GPG-NH2-treated ACH-2 cells, which were not stimulated to produce virus. The nodules contain the capsid protein p24, as determined by immune electron microscopy. Their size and density were similar, if not identical, to those of the broad end of the mature HIV-1 core structure. Possibly, these nodules represent virus particles that were arrested in their budding. If so, activation of the HIV-1 protease in such virus structures arrested at the plasma membrane might have cleaved off the p17 matrix protein, allowing the remaining polyprotein to condense, generating incomplete capsid structures trapped in the plasma membrane. Notably, no such nodular structures were observed after combined treatment with GPG-NH2 and an inhibitor of the viral protease.
New classes of antiviral pharmaceuticals are warranted in the fight against the AIDS epidemic. In the future, small peptides or peptidomimetic molecules derived from the peptides described in the present paper might prove useful for the treatment of HIV-infected individuals.
ACKNOWLEDGMENTS
We thank Marianne Ljungkvist, Anders Höglund, and Joel Reneby for valuable contributions to this study.
This study was supported by grants from the Swedish Medical Foundation (grant no. K2000-06X-09501-10B), Tripep AB, and the Swedish Research Council for Engineering Sciences.
FOOTNOTES
- Received 22 January 2002.
- Returned for modification 23 April 2002.
- Accepted 6 August 2002.
- American Society for Microbiology