Skip to main content
NCBI home page
Proceedings of the National Academy of Sciences of the United States of America logoLink to Proceedings of the National Academy of Sciences of the United States of America
. 2007 Mar 5;104(11):4353–4358. doi: 10.1073/pnas.0609477104

Domain-swapped dimerization of the HIV-1 capsid C-terminal domain

Dmitri Ivanov *,†,‡, Oleg V Tsodikov †,§, Jeremy Kasanov *, Tom Ellenberger , Gerhard Wagner †, Tucker Collins *
PMCID: PMC1838606  PMID: 17360528

Abstract

Assembly of the HIV and other retroviruses is primarily driven by the oligomerization of the Gag polyprotein, the major viral structural protein capable of forming virus-like particles even in the absence of all other virally encoded components. Several critical determinants of Gag oligomerization are located in the C-terminal domain of the capsid protein (CA-CTD), which encompasses the most conserved segment in the highly variable Gag protein called the major homology region (MHR). The CA-CTD is thought to function as a dimerization module, although the existing model of CA-CTD dimerization does not readily explain why the conserved residues of the MHR are essential for retroviral assembly. Here we describe an x-ray structure of a distinct domain-swapped variant of the HIV-1 CA-CTD dimer stabilized by a single amino acid deletion. In the domain-swapped structure, the MHR-containing segment forms a major part of the dimerization interface, providing a structural mechanism for the enigmatic function of the MHR in HIV assembly. Our observations suggest that swapping of the MHR segments of adjacent Gag molecules may be a critical intermediate in retroviral assembly.

Keywords: Gag, major homology region, viral assembly

The main structural protein encoded by the HIV-1 virus is a 55-kDa multidomain protein, Gag, that oligomerizes to form a protein shell around the viral genome during assembly and budding of the immature viral particle (1, 2). The released immature virus subsequently undergoes a maturation process that is mediated by proteolytic processing of Gag by the viral protease and involves a major structural rearrangement within the virion. The most remarkable change is the formation of the characteristic conical core structure formed by the proteolytically released capsid protein (CA) around the viral genome. Only ≈1,000–1,500 CAs are estimated to form the mature viral core, whereas ≈5,000 Gag subunits form the shell of the immature viral particle, suggesting that the assembly of the immature virus particle and the assembly of the mature viral core are two distinct assembly events in the viral life cycle rather than a structural rearrangement of the same protein assembly (3). The assembly of the immature particle is a complex multistep process mediated by numerous distinct interactions of Gag proteins with each other and other viral and cellular components. No high-resolution structural information is currently available for the arrangement of Gag units within the shell of the immature retroviral particle, and existing models of retroviral assembly are derived from high-resolution studies of isolated Gag fragments. Reconstruction of the protein arrangement in the immature particle from such partial structural data is challenging, because it is not immediately apparent whether the protein contacts observed in the crystals of the isolated domains faithfully represent the arrangement of the protein subunits in the viral shell. Important clues on the architecture of the viral shell come from the electron microscopy (EM) studies of the virus-like structures formed in vivo or in vitro (see refs. 4 and 5 and references therein), although the resolution of these EM reconstructions is still not sufficient for reliable fitting of the high-resolution x-ray or NMR structures into the EM density.

A general consensus from numerous mutagenesis studies is that some of the key determinants of Gag oligomerization are contained in the C-terminal domain of the CA (CA-CTD) (1). Viral assembly is very sensitive to mutations or deletions in this region of Gag. In particular, the CA-CTD encompasses the most conserved segment in the otherwise highly variable Gag polyprotein known as the major homology region (MHR) (6). Disruptive mutations of the conserved residues in this region of ≈20 aa result in severe defects in viral assembly, whereas some of the more conservative MHR substitutions are still assembly-competent but seem to interfere with viral infectivity (7–10). The mechanism of the critical MHR function in the assembly of the immature particle is not known. The current model of the CA-CTD function is based on its ability to act as a dimerization domain (1). The functional dimer interface of the CA-CTD is inferred from the crystal structure of the isolated CA-CTD domain of HIV-1 (11). In the CA-CTD crystal, the dimer is formed through the packing of the helix 2 against the helix 2 of the monomer related by the crystal symmetry, and mutations of the interface residues completely abolish CA dimerization in solution. Nevertheless, it is still not clear how this interaction can account for the critical role played by the CA-CTD in the assembly of the immature viral particle. First, mutations of the crystal dimer interface residues have a pronounced effect on viral assembly but do not completely block it (12). Second, the interface residues are not conserved in different retroviruses, and the capsids of the Rous sarcoma virus, human T cell leukemia virus type 1, and equine infectious anemia virus all seem to be monomeric in solution at concentrations of up to 1 mM (13–15). Finally, the critical MHR segment is not located at the dimer interface in the HIV-1 CA-CTD crystal, and the reason for its remarkable conservation in retroviruses is not apparent from the existing CA-CTD dimerization model.

The NMR structure of the SCAN domain, a dimerization module found in mammalian zinc-finger proteins that is structurally and evolutionary related to the retroviral CA-CTD, suggests an entirely different model of CA-CTD dimerization (16). By analogy to the SCAN dimer, dimerization of the CA-CTD could occur via the so-called three-dimensional domain-swapping mechanism (Fig. 1 A and B). Three-dimensional domain swapping occurs by exchanging a structural element of one protein monomer with an equivalent element of the other monomer, resulting in a dimeric structure in which only the linker region, which connects the swapped element with the rest of the protein, adopts a distinct conformation (17, 18). The striking feature of the SCAN-like model of CA-CTD dimerization is that the swapped structural element contains the MHR region of the retroviral CA-CTD, providing a straightforward model of the critical role played by the MHR in the CA-CTD function. Here we report that the HIV-1 CA-CTD is capable of domain-swapped dimerization and discuss the implications for retroviral assembly.

Fig. 1.

Fig. 1.

Open in a new tab

Production of the Δ177 CA-CTD mutant. (A and B) The SCAN dimer is a domain-swapped homolog of the HIV-1 CA-CTD. (A) X-ray structure of the HIV-1 CA-CTD (11). Location of the Ala-177 residue in the linker region is shown in red. (B) NMR structure of the SCAN domain dimer (16). Helices are numbered. The SCAN dimer is formed by swapping of the N-terminal protein segment containing helix 1. The corresponding segment in the HIV-1 CA-CTD contains the conserved MHR sequence. (C and D) Dimerization properties of the Δ177 CA-CTD mutant. (C) The Δ177 CA-CTD elutes as two distinct peaks after size-exclusion chromatography. The two peaks interconvert slowly at 4°C. (D) Analytical ultracentrifugation results. The sedimentation equilibrium distributions of the Δ177 CA-CTD peak 1 (triangles) and Δ177 CA-CTD peak 2 (circles) are close to the theoretical lines of the dimer and the monomer, respectively.

Results

Dimerization Properties of the Δ177 CA-CTD Mutant.

The dimer–monomer equilibrium in domain-swapped dimerization is known to be sensitive to the composition and the length of the linker region (19). To perturb the dimerization equilibrium of the isolated CA-CTD, we used site-directed mutagenesis to prepare CA-CTD mutant with a single amino acid deletion of Ala-177 in the linker region (Δ177 CA-CTD). The mutant CA-CTD was purified as a fusion with a GST affinity tag and subsequently cleaved off by using thrombin digestion. Size-exclusion chromatography of the cleaved Δ177 CA-CTD revealed that the construct exists as a mixture of two distinct species that can be readily separated by gel filtration (Fig. 1C). The sedimentation equilibrium analysis revealed that the two chromatographic peaks correspond to the monomer and the dimer of Δ177 CA-CTD (Fig. 1D) (see Materials and Methods). Exchange between the monomer and dimer is very slow at 4°C and does not reach an equilibrium distribution even after 1 week (Fig. 1C). Such slow exchange is characteristic of the domain-swapped dimerization (18).

Crystal Structure of the Δ177 CA-CTD Dimer.

To characterize the structure of the dimer and to check whether it indeed forms a domain-swapped structure similar to the one observed in the SCAN domain, we pursued a crystallographic approach. We were able to crystallize the purified Δ177 CA-CTD dimer at 4°C and determine its structure at 2.4-Å resolution (Table 1). The x-ray structure confirms that Δ177 CA-CTD forms a domain-swapped dimer (Fig. 2A). The architecture of the dimer is essentially identical to the architecture of the SCAN domain dimer: the dimer is formed by exchanging the protein segment that contains the N-terminal strand, a turn, and helix 1. This segment encompasses the MHR, and the conserved MHR residues are essential for formation and stability of the dimer (Fig. 2B). In particular, conserved Gln-155 and Tyr-164 form an important part of the dimer interface. Residue 164, which only occurs either as a phenylalanine or a tyrosine in retroviruses, is located at the center of the largely hydrophobic dimer interface flanked by Phe-161 and Phe-168 and packed against Leu-190 of the other monomer. Consistent with this observation, mutation of Leu-190 to alanine in HIV-1 results in a pronounced assembly defect (20). The orientation of Gln-155 is quite unusual because this polar residue is not surface-exposed but, rather, projects into the interior of the protein where it bridges across the dimer interface by forming an intramolecular hydrogen bond with the carbonyl oxygen of Glu-159 of one monomer and an intermolecular hydrogen bond with the Asn-195 amide proton of the other monomer (Fig. 2B). This unusual structural feature explains the remarkable conservation of Gln-155 in diverse retroviruses and the observation that mutations of this residue essentially block viral assembly.

Table 1.

Crystallographic data collection and structure refinement statistics for Δ177 HIV-1 CA-CTD

Measurement Value
Data collection
    Space group P43212
    Cell dimensions
        a/b/c, Å 40.13/40.13/105.3
        α/β/γ, ° 90/90/90
    Resolution, Å 50.00–2.30 (2.38–2.30)
    Rmerge 0.046 (0.031)
    I/σI 40.6 (4.7)
    Completeness 0.98 (0.90)
    Redundancy 6.9 (5.3)
Refinement
    Resolution, Å 30.0–2.40 (2.46–2.40)
    No. of reflections 3,540 (175)
    Rwork/Rfree 0.24/0.29 (0.34/0.51)
    No. of atoms
        Protein 561
        Water 19
    B factors, Å2
        Protein 15.6
        Water 7.3
    rms deviations
        Bond lengths, Å 0.018
        Bond angles, ° 1.76
Open in a new tab

The highest-resolution-shell values are shown in parentheses.

Fig. 2.

Fig. 2.

Open in a new tab

X-ray structure of the Δ177 CA-CTD dimer. (A) The Δ177 CA-CTD domain-swapped dimer is formed by exchanging the N-terminal strand and helix 1 between the two monomers. Side chains of the conserved MHR residues are shown as rods. Location of the Δ177 deletion is marked by two parallel lines. (B) Magnified view of the dimer interface formed by the MHR segment. Highly conserved Y164, Q155, E159, and R167 are shown as thick rods, other important interface residues are shown as wire frames, and the hydrogen-bonding network is shown as dashed lines. (C and D) Comparison of the wild-type CA-CTD monomer (11) (C) and half of the domain-swapped Δ177 CA-CTD dimer (D). The kink in helix 2 of the monomer is relieved after formation of the domain-swapped dimer. Deletion of Ala-177 (shown in red) is likely to increase the kink and affect the dimerization equilibrium.

Other highly conserved residues of the MHR (Gly-156, Glu-159, and Arg-167; Fig. 2B) do not form dimer contacts, but instead participate in an extensive intramolecular hydrogen-bonding network on the surface of the MHR fold. This network of interacting residues might contribute to the folded stability of the MHR and/or facilitate the domain-swapping process by preventing the complete unfolding of the MHR during the transient opening of the CA-CTD required for formation of the domain-swapped dimer. Conformational rigidity of the swapped MHR element could facilitate domain swapping by avoiding the kinetically unfavorable, partially unfolded intermediates that otherwise could arise as protein–protein interactions are lost during the process of domain swapping (19).

A comparison of our domain-swapped structure with previously reported crystal structures of the HIV-1 CA-CTD (11, 21) reveals that most of the CA-CTD structure is unchanged in the domain-swapped dimer relative to the side-by-side dimers. However, the MHR fold mediates intermolecular contacts in the domain-swapped dimer instead of the intramolecular contacts observed in side-by-side dimers of the CA-CTD. As expected, the altered protein topology in the swapped dimer results in a different conformation of the linker region (amino acids 174–178) between helices 1 and 2, but the most striking structural change occurs in helix 2 of the monomer after dimerization (Fig. 2 C and D). Helix 2 is markedly kinked in earlier x-ray structures of the HIV-1 CA-CTD (11, 21) (Fig. 2C), whereas helix 2 is an ideal, straight α-helix in the domain-swapped dimer (Fig. 2D). A release of strain accompanying the unkinking of helix 2 could provide a driving force for the formation of a domain-swapped dimer. Inspection of other known retroviral CA-CTD structures (13–15) reveals notable deviations from the ideal helical geometry in helix 2 in all of these structures, which suggests a “loaded spring” mechanism that may be conserved in many retroviruses. Comparison of the structures also reveals the most likely effect of the Δ177 mutation in the CA-CTD monomer: shortening of the linker region is most likely to destabilize the packing interactions of helix 2 and shift the conformational equilibrium toward the domain-swapped dimer (Fig. 2C).

In Vitro Assembly of the Δ177 Mutants of the HIV-1 Capsid and Gag.

To check whether the Δ177 mutation is compatible with viral assembly, we investigated the assembly properties of the Δ177 mutants in vitro. Various recombinant HIV-1 Gag constructs were shown to assemble in vitro into three-dimensional structures similar to the ones observed in the viral particles isolated from cell culture. Purified CA can assemble at high salt conditions into tubes and cones, mimicking the architecture of mature viral cores isolated from the live virus (22–26). On the other hand, longer Gag constructs, encompassing various portions of the matrix domain, CA, spacer peptide 1 (SP1), and nucleocapsid, were shown to assemble into spheres that are similar to the immature viral particles (5, 27–29). To probe the effects of the Δ177 mutation on assembly, we introduced this deletion into the full-length CA protein (30) and the p6-deleted construct of Gag (5). The proteins were expressed in Escherichia coli and purified as described in Materials and Methods. The purified proteins were then incubated at conditions that promote assembly of the wild-type proteins. Fig. 3 shows EM images of the particles formed by the proteins containing the Δ177 mutation. The Δ177 CA was observed to form cylinders of ≈50–60 nm in diameter, very similar to the structures formed by the wild-type CA (Fig. 3B). The EM images of the Δ177 p6-deleted Gag assemblies reveal collapsed spherical particles, indicating that the Δ177 Gag is also assembly-competent (Fig. 3A). However, the sizes of the spheres vary significantly (from 50 to >300 nm), with the most common size being ≈200 nm. These spheres are not as uniform and are significantly larger than the particles reported for the wild-type p6-deleted Gag constructs (5, 27, 29). Intriguingly, the size distribution of the particles formed by the Δ177 Gag in vitro is similar to the one observed in vivo for a Gag mutant with impaired CA-CTD dimerization (31). Moreover, the size of the virus-like particles formed by the p6-deleted Gag was shown to be influenced by inositol phosphates (32), and the C-terminal domain of the CA seems to be involved in the inositol phosphate binding (33). Therefore, it is possible that the CA-CTD-mediated interactions may affect the curvature and size of the assembled particle. A careful comparison of the assembly properties of the wild-type Gag constructs and Gag mutants with altered CA-CTD dimerization will be required to establish any such connection.

Fig. 3.

Fig. 3.

Open in a new tab

In vitro assembly of the Δ177 mutants. (A) The Δ177 p6-deleted construct of HIV-1 Gag assembles into spherical virus-like particles after addition of 0.2 mg/ml E. coli rRNA. (B) The Δ177 HIV-1 CA assembles into tubular structures in the high-salt buffer. Particles were analyzed by EM as described in Materials and Methods.

Discussion

Our understanding of retroviral assembly is limited by the lack of high-resolution structural information about the arrangement of Gag subunits within the shell of the immature viral particle. In the absence of sufficiently high-resolution EM reconstructions of the virus-like structures, it is difficult to piece together the architecture of the immature virus particle from high-resolution structures of isolated Gag domains. For example, the molecular basis for the critical assembly function of the MHR, the most conserved segment of Gag, has remained elusive for over a decade because the existing structural information does not readily explain the remarkable conservation of the MHR and the extensive mutagenesis data. The domain-swapped CA-CTD structure described here suggests that the MHR may be directly involved in CA-CTD dimerization.

Patterns of amino acid conservation within retroviral proteins can be informative because of the strong selective pressure to change amino acid composition to evade the host's immune response while maintaining the essential functions of viral proteins. Therefore, the conserved patterns in the retroviral genomes reveal critical primary sequence elements for which alternatives are not easily found. In general, one observes that the conservation within the viral enzymes is considerably higher than within the structural proteins, probably reflecting the fact that the productive dynamic modes required for enzymatic catalysis are more sensitive to amino acid substitutions than the purely structural elements of the virus. Indeed, a remarkably low level of amino acid identity is observed for the structural Gag polyprotein in different retroviruses, given that the three-dimensional structures of its domains seem to be essentially the same across different retroviral families. The conservation of the MHR within the otherwise highly variable Gag is exceptional, and this conserved motif is present in all known retroviruses and in some retrotransposons. An unexpected model of the enigmatic MHR function was suggested by the NMR structure of the evolutionary and structurally related SCAN domain (16). The domain-swapped SCAN structure suggested that this dimerization mode may represent a previously unrecognized conformation of CA-CTD that is critical for viral assembly. Here we report direct evidence in support of this model. By introducing a single alanine deletion into CA-CTD, we were able to perturb the dimerization equilibrium of the isolated CA-CTD domain. The x-ray structure of the purified Δ177 CA-CTD dimer revealed that, indeed, it does adopt a domain-swapped architecture in which the MHR forms a major part of the dimer interface. This structural role of the MHR suggests why mutations in this region are detrimental to viral assembly. Furthermore, the molecular motions inferred for the domain-swapping process provide an additional explanation for the remarkable conservation of the MHR during retroviral evolution. A conserved network of hydrogen bonds (Fig. 2B) could stabilize the MHR and prevent unfolding as helix 1 relinquishes its cis packing interactions in the monomer to repack on a partner subunit in the domain-swapped dimer.

Another notable structural feature of the CA-CTD may activate the transition from the CA-CTD monomer to the domain-swapped dimer. Because dimerization by domain swapping requires a transient opening of the monomer, an inherent instability of the native monomer could lower the activation barrier for domain swapping. Notably, the structures of HIV-1 CA-CTD monomers and side-by-side dimers reveal a pronounced kink in helix 2. A similar kinking of helix 2 is observed in the x-ray structure of the CA-CTD of the equine infectious anemia virus (13). NMR structures of the CA-CTDs of the Rous sarcoma virus and human T cell leukemia virus type 1 also display notable deviations from the ideal backbone geometry within the helix 2 segment (14, 15). In contrast, helix 2 straightens and adopts ideal helical geometry when packed against a neighboring CA-CTD subunit of the domain-swapped dimer. The unbending of helix 2 slightly changes the overall shape of the CA-CTD; therefore, it may be possible to distinguish between the swapped and unswapped conformations of the CA-CTD in cryo-EM reconstructions of virus-like particles determined at sufficiently high resolution.

The observed structural change of the CA-CTD suggests that the distortion of helix 2 in the monomer may act as a loaded spring that destabilizes the structure of the CA-CTD monomer. Close proximity of the CA-CTD modules brought about by interactions of the other Gag modules during assembly of the immature particle, therefore, may be sufficient to trigger formation of the domain-swapped dimer. The correct juxtaposition of the CA-CTD domains can most likely be accomplished by the adjacent nucleocapsid domain, which promotes Gag oligomerization through nonspecific interactions with the viral RNA (1). This hypothesis is supported by the observation that, after addition of RNA, the CA–SP1–nucleocapsid constructs assemble into tubular structures at lower salt and protein concentrations than the CA domain alone (23, 24). The putative domain-swapped dimerization may also be coupled to other molecular determinants of Gag assembly. For example, the integrity of the region spanning the CA–SP1 junction is critical for assembly (34, 35), whereas chemical compounds and mutations that prevent the CA–SP1 cleavage interfere with viral maturation (36). This region is disordered in all structural studies on isolated Gag fragments, including ours, but the mutational data indicate that it may be structured in the context of the full-length Gag within the immature viral shell (34, 35). Any interactions mediated by this polypeptide segment immediately adjacent to the CA-CTD structure, therefore, may be coupled to the CA-CTD dimerization equilibrium. In addition, the two cysteines located in the CA-CTD were suggested to participate in an oxidative capsid assembly mechanism, in which CA oligomerization or maturation is triggered by disulfide bond formation between the two cysteines as the budding virus enters the oxidizing environment of the bloodstream (14, 37). The cysteines are reduced in our crystal structure, but it is possible that formation of the intradomain disulfide bond flanking the domain-swapped dimer interface could alter the stability of the domain-swapped dimer.

In conjunction with the extensive data on the functional significance of the MHR in the assembly of the immature retroviral particle, our findings suggest that domain swapping may be a critical event in the immature particle assembly. CA-CTD dimerization is also involved in the subsequent assembly of the mature viral core; thus, it is formally possible that the domain-swapped conformation may be present in the mature core as well. However, there is little experimental evidence to argue in favor of the domain-swapped model over the conventional side-by-side model of CA-CTD dimerization in the mature core, and the resolution of the cryo-EM reconstructions of the core-like assemblies formed in vitro is not sufficient to distinguish between the two CA-CTD conformations (38, 39). Here we report that the Δ177 deletion mutants can assemble into immature-virus-like and core-like structures in vitro, but these results do not directly support involvement of domain swapping in these assembly events because the assemblies may be formed exclusively by the nondomain-swapped form of the proteins. Further investigation of capsid-mediated dimerization is required to test whether the domain-swapped conformation of the CA-CTD is present in the immature virions or the mature viral cores.

The interlocked arrangement of viral capsid subunits conferred by domain swapping may have several advantages over more conventional side-by-side interactions between structural modules. First, for a given protein size, the intertwined domain-swapped dimer is likely to create a much larger dimer interface than what can possibly be achieved in the side-by-side dimer. For example, the dimer interface of the HIV-1 CA-CTD domain-swapped dimer measures ≈1,620 Å2. This is a very large area for a protein of <100 aa, and it constitutes almost 30% of the total monomer surface. As a result, an extensive and robust interface between viral building blocks can be achieved without an increase in the size of the interacting protein and, therefore, in the size of the viral genome. In fact, domain swapping has been observed in the crystal structures of several icosahedral viruses (40–42). Second, although the domain-swapped protein dimer is mechanically robust, it is thermodynamically only marginally stable given the similar interactions of the swapped and unswapped configurations, which may be very important for the productive viral life cycle; although it is desirable for the viral coat to be strong enough to withstand environmental stress during cell-to-cell transmission, after cell entry the virus must efficiently uncoat to establish a productive infection. Finally, the metastable nature of the domain-swapped dimer provides a way to control the directionality of the assembly and the disassembly steps in the viral life cycle. For example, if within the full-length Gag the CA-CTD dimerization indeed is promoted by the adjacent Gag modules, once the Gag polyprotein is cleaved by the viral protease, the domain-swapped dimer would be only metastable, and its eventual disassembly would be irreversible. This delicate energetic balance of the domain-swapped dimerization may explain why the virus is so sensitive to mutations in the MHR region. Collectively, our findings strengthen the possibility that the swapping of the MHR segment between adjacent Gag molecules may represent a critical intermediate step in retroviral assembly. They also suggest that interfering with this process may be a viable strategy for inhibition of the critical assembly or disassembly events in the viral life cycle.

Materials and Methods

Molecular Cloning, Protein Purification, and Analytical Ultracentrifugation.

The full-length CA was produced from the pWISP98–85 plasmid (30) obtained from the National Institutes of Health AIDS Research & Reference Reagent Program. The CA-CTD construct was prepared by subcloning the 146–231 fragment of the CA protein from the pWISP98–85 plasmid into the pGEX2T expression vector. The p6-deleted construct of HIV-1 Gag (MACANCexact), which contains a 6-His affinity tag preceding the N terminus of the matrix domain, has no deletions within the Gag coding sequence, and ends at the C terminus of the nucleocapsid (5), was kindly provided by Eric Barklis (Oregon Health and Sciences University, Portland, OR). The Δ177 mutants of the CA-CTD, CA, and MACANCexact constructs were prepared by performing QuikChange site-directed mutagenesis (Stratagene, LaJolla, CA) on the expression plasmids encoding wild-type proteins. The CA-CTD was expressed as a GST fusion and purified on the glutathione-conjugated Sepharose (HiTrap) by using the standard protocols. The GST affinity tag was subsequently cleaved off by using thrombin digestion. The resulting CA-CTD fragment had an additional glycine residue at the N terminus, followed by Ser-146 of the CA. The C terminus of the CA-CTD construct was exactly at the C terminus of the mature CA (Leu-231). The CA and the p6-deleted construct of Gag (MACANCexact) were expressed and purified following published procedures (5, 30). Sedimentation equilibrium experiments with the Δ177 CA-CTD mutant were performed by using a Beckman XLA analytical centrifuge in the 50 mM sodium phosphate (pH 6.5)/50 mM NaCl/5 mM DTT buffer (Beckman Coulter, Fullerton, CA). Protein concentration was ≈0.3 mg/ml, and the sedimentation equilibrium distributions were measured at rotor speeds of 25,000, 35,000, and 40,000 rpm. Fig. 1D shows the data acquired at 25,000 rpm. The solid and dashed lines show the expected theoretical equilibrium distributions for the Δ177 CA-CTD dimer and the monomer, respectively, assuming partial specific volumes, ν, of 0.72 ml/g for both samples.

Protein Crystallization and X-Ray Crystallography.

After cleavage of the GST affinity tag, the Δ177 CA-CTD dimer was separated from the monomer by size-exclusion chromatography on Superdex75 media (GE Healthcare, Chalfont St. Giles, U.K.) at 4°C in a buffer containing 50 mM sodium phosphate (pH 6.5), 50 mM NaCl, and 5 mM DTT. The Δ177 CA-CTD dimer sample was concentrated to 5 mg/ml protein, and 1 μl of the concentrated protein solution was mixed with 1 μl of reservoir solution [0.1 M sodium phosphate (pH 6.75)/28% (vol/vol) PEG 1500/15% (vol/vol) glycerol]. The resulting 2-μl drops were equilibrated by the hanging-drop vapor-diffusion method at 4°C. Protein crystals that were suitable for diffraction experimentation grew in 2–3 days. The crystals were flash-frozen in liquid nitrogen directly from the mother liquor. The x-ray diffraction data were collected at cryogenic conditions at the 19-ID beamline at the Advanced Photon Source (Argonne, IL). The x-ray diffraction phase problem was solved by a molecular-replacement approach by using the structure of the wild-type CA-CTD monomer (11) as the search model. In the molecular-replacement search model, the linker region and a part of helix 2 (residues 174–188 of the capsid) were removed to avoid introducing a bias in the possible mode of dimerization. These missing regions and other regions in which discrepancies with the search model were observed were subsequently rebuilt iteratively with structure refinement by using REFMAC (43) without any ambiguity solely on the basis of the observed difference electron density.

In Vitro Assembly Assays.

Purified CA was concentrated to 15 mg/ml by using Ultrafree 15 (10-kDa cutoff) concentrators (Millipore). The buffer of the concentrated CA sample was replaced with 0.1 mM Tris (pH 8.0) and 1 M NaCl by using BioSpin 6 desalting columns (Bio-Rad, Hercules, CA). The protein was incubated in the high-salt buffer for 12–15 h at 37°C, diluted 20-fold, and then analyzed by EM using negative staining. Purified p6-deleted construct of Gag (MACANCexact) was concentrated to 1.8 mg/ml by using Ultrafree 15 (10-kDa cutoff) spin concentrators (Millipore). The sample buffer was replaced with 0.1 M Tris (pH 8.0)/100 mM NaCl/2 mM DTT by using BioSpin 6 desalting columns (Bio-Rad). After the buffer exchange, 0.2 mg/ml E. coli rRNA (Roche Applied Science, Indianapolis, IN) was added to the samples, and the samples were incubated for 12–15 h at 37°C. After the incubation, the assembled particles were spun down for 1.5 h at 100,000 rpm, and the pellets were resuspended in 1/10th of the assembly buffer and analyzed by EM using negative staining. For EM analysis, 2–5 μl of the sample was absorbed on a carbon-coated grid that was glow-discharged for 2 min. Excess liquid was blotted away with filter paper, and the samples were stained with 1% uranyl acetate for 20 sec and dried immediately.

Acknowledgments

We are grateful to Laura Reither, Dr. David Gohara, and the staff of the 19-ID beamline at the Advanced Photon Source (Argonne, IL) for collecting the x-ray diffraction data; Dr. Eric Barklis for providing the expression vector for the MACANCexact construct of Gag; Dr. Judit Villén for analyzing the mutants by mass spectrometry; Dr. Howard L. Mulhern for assistance with EM; and Dr. Heinrich Göttlinger for sharing expertise in HIV assembly. The HIV-1 capsid expression vector was obtained from the National Institutes of Health AIDS Research and Reference Reagent Program. This research was supported in part by National Institutes of Health Grants GM066516 (to T.C.), HL035716 (to T.C.), GM52504 (to T.E.), and GM47467 (to G.W.). D.I. is a recipient of the Scholar Award from the Harvard Center for AIDS Research. J.K. is supported by National Institutes of Health Training Grant T32 HL076115.

Abbreviations

CA

capsid protein

CA-CTD

C-terminal domain of the CA

MHR

major homology region

SP1

spacer peptide 1.

Footnotes

The authors declare no conflict of interest.

This article is a PNAS direct submission.

Data deposition: The atomic coordinates and structure factors have been deposited in the Protein Data Bank, www.pdb.org (PDB ID code 2ONT).

References

  • 1.Adamson CS, Jones IM. Rev Med Virol. 2004;14:107–121. doi: 10.1002/rmv.418. [DOI] [PubMed] [Google Scholar]
  • 2.Scarlata S, Carter C. Biochim Biophys Acta. 2003;1614:62–72. doi: 10.1016/s0005-2736(03)00163-9. [DOI] [PubMed] [Google Scholar]
  • 3.Briggs JA, Simon MN, Gross I, Krausslich HG, Fuller SD, Vogt VM, Johnson MC. Nat Struct Mol Biol. 2004;11:672–675. doi: 10.1038/nsmb785. [DOI] [PubMed] [Google Scholar]
  • 4.Briggs JA, Johnson MC, Simon MN, Fuller SD, Vogt VM. J Mol Biol. 2006;355:157–168. doi: 10.1016/j.jmb.2005.10.025. [DOI] [PubMed] [Google Scholar]
  • 5.Huseby D, Barklis RL, Alfadhli A, Barklis E. J Biol Chem. 2005;280:17664–17670. doi: 10.1074/jbc.M412325200. [DOI] [PubMed] [Google Scholar]
  • 6.Wills JW, Craven RC. AIDS. 1991;5:639–654. doi: 10.1097/00002030-199106000-00002. [DOI] [PubMed] [Google Scholar]
  • 7.Craven RC, Leure-duPree AE, Weldon RA, Jr, Wills JW. J Virol. 1995;69:4213–4227. doi: 10.1128/jvi.69.7.4213-4227.1995. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Mammano F, Ohagen A, Hoglund S, Gottlinger HG. J Virol. 1994;68:4927–4936. doi: 10.1128/jvi.68.8.4927-4936.1994. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Orlinsky KJ, Gu J, Hoyt M, Sandmeyer S, Menees TM. J Virol. 1996;70:3440–3448. doi: 10.1128/jvi.70.6.3440-3448.1996. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Strambio-de-Castillia C, Hunter E. J Virol. 1992;66:7021–7032. doi: 10.1128/jvi.66.12.7021-7032.1992. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Gamble TR, Yoo S, Vajdos FF, von Schwedler UK, Worthylake DK, Wang H, McCutcheon JP, Sundquist WI, Hill CP. Science. 1997;278:849–853. doi: 10.1126/science.278.5339.849. [DOI] [PubMed] [Google Scholar]
  • 12.von Schwedler UK, Stray KM, Garrus JE, Sundquist WI. J Virol. 2003;77:5439–5450. doi: 10.1128/JVI.77.9.5439-5450.2003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Jin Z, Jin L, Peterson DL, Lawson CL. J Mol Biol. 1999;286:83–93. doi: 10.1006/jmbi.1998.2443. [DOI] [PubMed] [Google Scholar]
  • 14.Khorasanizadeh S, Campos-Olivas R, Summers MF. J Mol Biol. 1999;291:491–505. doi: 10.1006/jmbi.1999.2986. [DOI] [PubMed] [Google Scholar]
  • 15.Campos-Olivas R, Newman JL, Summers MF. J Mol Biol. 2000;296:633–649. doi: 10.1006/jmbi.1999.3475. [DOI] [PubMed] [Google Scholar]
  • 16.Ivanov D, Stone JR, Maki JL, Collins T, Wagner G. Mol Cell. 2005;17:137–143. doi: 10.1016/j.molcel.2004.12.015. [DOI] [PubMed] [Google Scholar]
  • 17.Bennett MJ, Choe S, Eisenberg D. Proc Natl Acad Sci USA. 1994;91:3127–3131. doi: 10.1073/pnas.91.8.3127. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Rousseau F, Schymkowitz JW, Itzhaki LS. Structure (London) 2003;11:243–251. doi: 10.1016/s0969-2126(03)00029-7. [DOI] [PubMed] [Google Scholar]
  • 19.Rousseau F, Schymkowitz JW, Wilkinson HR, Itzhaki LS. Proc Natl Acad Sci USA. 2001;98:5596–5601. doi: 10.1073/pnas.101542098. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Joshi A, Nagashima K, Freed EO. J Virol. 2006;80:7939–7951. doi: 10.1128/JVI.00355-06. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Ternois F, Sticht J, Duquerroy S, Krausslich HG, Rey FA. Nat Struct Mol Biol. 2005;12:678–682. doi: 10.1038/nsmb967. [DOI] [PubMed] [Google Scholar]
  • 22.Ehrlich LS, Agresta BE, Carter CA. J Virol. 1992;66:4874–4883. doi: 10.1128/jvi.66.8.4874-4883.1992. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Campbell S, Vogt VM. J Virol. 1995;69:6487–6497. doi: 10.1128/jvi.69.10.6487-6497.1995. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Gross I, Hohenberg H, Krausslich HG. Eur J Biochem. 1997;249:592–600. doi: 10.1111/j.1432-1033.1997.t01-1-00592.x. [DOI] [PubMed] [Google Scholar]
  • 25.Ganser BK, Li S, Klishko VY, Finch JT, Sundquist WI. Science. 1999;283:80–83. doi: 10.1126/science.283.5398.80. [DOI] [PubMed] [Google Scholar]
  • 26.Briggs JA, Wilk T, Welker R, Krausslich HG, Fuller SD. EMBO J. 2003;22:1707–1715. doi: 10.1093/emboj/cdg143. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Campbell S, Rein A. J Virol. 1999;73:2270–2279. doi: 10.1128/jvi.73.3.2270-2279.1999. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Gross I, Hohenberg H, Wilk T, Wiegers K, Grattinger M, Muller B, Fuller S, Krausslich HG. EMBO J. 2000;19:103–113. doi: 10.1093/emboj/19.1.103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Morikawa Y, Goto T, Sano K. J Biol Chem. 1999;274:27997–28002. doi: 10.1074/jbc.274.39.27997. [DOI] [PubMed] [Google Scholar]
  • 30.Yoo S, Myszka DG, Yeh C, McMurray M, Hill CP, Sundquist WI. J Mol Biol. 1997;269:780–795. doi: 10.1006/jmbi.1997.1051. [DOI] [PubMed] [Google Scholar]
  • 31.Datta SA, Curtis JE, Ratcliff W, Clark PK, Crist RM, Lebowitz J, Krueger S, Rein A. J Mol Biol. 2007;365:812–824. doi: 10.1016/j.jmb.2006.10.073. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Campbell S, Fisher RJ, Towler EM, Fox S, Issaq HJ, Wolfe T, Phillips LR, Rein A. Proc Natl Acad Sci USA. 2001;98:10875–10879. doi: 10.1073/pnas.191224698. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Datta SA, Zhao Z, Clark PK, Tarasov S, Alexandratos JN, Campbell SJ, Kvaratskhelia M, Lebowitz J, Rein A. J Mol Biol. 2007;365:799–811. doi: 10.1016/j.jmb.2006.10.072. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Accola MA, Hoglund S, Gottlinger HG. J Virol. 1998;72:2072–2078. doi: 10.1128/jvi.72.3.2072-2078.1998. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Liang C, Hu J, Whitney JB, Kleiman L, Wainberg MA. J Virol. 2003;77:1772–1783. doi: 10.1128/JVI.77.3.1772-1783.2003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Aiken C, Chen CH. Trends Mol Med. 2005;11:31–36. doi: 10.1016/j.molmed.2004.11.001. [DOI] [PubMed] [Google Scholar]
  • 37.McDermott J, Farrell L, Ross R, Barklis E. J Virol. 1996;70:5106–5114. doi: 10.1128/jvi.70.8.5106-5114.1996. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Ganser BK, Cheng A, Sundquist WI, Yeager M. EMBO J. 2003;22:2886–2892. doi: 10.1093/emboj/cdg276. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Li S, Hill CP, Sundquist WI, Finch JT. Nature. 2000;407:409–413. doi: 10.1038/35030177. [DOI] [PubMed] [Google Scholar]
  • 40.Qu C, Liljas L, Opalka N, Brugidou C, Yeager M, Beachy RN, Fauquet CM, Johnson JE, Lin T. Structure (London) 2000;8:1095–1103. doi: 10.1016/s0969-2126(00)00508-6. [DOI] [PubMed] [Google Scholar]
  • 41.Garriga D, Querol-Audi J, Abaitua F, Saugar I, Pous J, Verdaguer N, Caston JR, Rodriguez JF. J Virol. 2006;80:6895–6905. doi: 10.1128/JVI.00368-06. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Chen R, Neill JD, Estes MK, Prasad BV. Proc Natl Acad Sci USA. 2006;103:8048–8053. doi: 10.1073/pnas.0600421103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Murshudov GN, Vagin AA, Dodson EJ. Acta Crystallogr D. 1997;53:240–255. doi: 10.1107/S0907444996012255. [DOI] [PubMed] [Google Scholar]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences

RESOURCES