Advertisement
JBC

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M412325200 on February 25, 2005

J. Biol. Chem., Vol. 280, Issue 18, 17664-17670, May 6, 2005
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
280/18/17664    most recent
M412325200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Huseby, D.
Right arrow Articles by Barklis, E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Huseby, D.
Right arrow Articles by Barklis, E.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Assembly of Human Immunodeficiency Virus Precursor Gag Proteins*

Doug Huseby, Robin Lid Barklis, Ayna Alfadhli, and Eric Barklis{ddagger}

From the Vollum Institute and Department of Microbiology, Oregon Health and Science University, Portland, Oregon 97201-3098

Received for publication, November 1, 2004 , and in revised form, February 15, 2005.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
To investigate the mechanism by which human immunodeficiency virus (HIV) precursor Gag (PrGag) proteins assemble to form immature virus particles, we examined the in vitro assembly of MACANC proteins, composed of the PrGag matrix, capsid, and nucleocapsid domains. In the absence of other components, MACANC proteins assembled efficiently at physiological temperature but inefficiently at lower temperatures. However, the addition of RNA reduced the temperature sensitivity of assembly reactions. Assembly of MACANC proteins also was affected by pH because the proteins preferentially formed tubes at pH 6.0, whereas spheres were obtained at pH 8.0. Because neither tubes nor spheres were amenable to analysis of protein-protein contacts, we also examined the membrane-bound assemblies of MACANC proteins. Interestingly, MACANC proteins organized on membranes in tightly packed hexameric rings. The observed hexamer spacing of 79.7 Å is consistent with the notion that more PrGag proteins assemble into virions than are needed to provide capsid proteins for mature virus cores. Our data are also consistent with a model for PrGag contacts in immature virions where capsid hexamers are tightly packed, where nucleocapsid domains align beneath capsid C-terminal domains, and where matrix domains form trimers at the nexus of three neighbor hexamers.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
During human immunodeficiency virus 1 (HIV-1)1 assembly, the viral precursor Gag (PrGag) protein oligomerizes on cellular membranes and directs the budding of immature virus particles. Normally, during or after budding, PrGag proteins are processed into the mature Gag proteins: matrix (MA), capsid, nucleocapsid (NC), and p6 (1). PrGag processing is accompanied by a dramatic change in HIV-1 particle morphology. By conventional thin section electron microscopy, the electrondense protein shell of immature virions appears to reorganize into centrally located, conical, or cylindrical mature virus cores (1).

Recently, we demonstrated that the major HIV-1 Gag protein capsid could assemble in vitro into two distinct arrangements (2). Both of these arrangements showed hexameric rings of capsid N-terminal domains (NTDs) linked via C-terminal domain (CTD) contacts but differed in that one showed tight packing of hexamers with adjacent NTD rings in apparent contact, whereas the other featured clearly separated NTD rings (2). Interestingly, the two in vitro arrangements appear to have in vivo counterparts. The tightly packed arrangement is similar to that observed in immature virions, where hexamer-to-hexamer spacing is in the range of 65 to 80 Å (2–5). In contrast, the loosely packed arrangement appears to correspond to the organization of capsid proteins in mature cores assembled in vitro and in vivo with a hexamer ring-to-ring spacing of 95–110 Å (6, 7). The similarity between our results and those observed in vivo prompted us to propose a model in which viral morphogenesis was accompanied by a shift from tightly packed hexamer rings to loosely packed rings, and we predicted that virions would have more capsid protein than was needed for mature core formation (2). This prediction was borne out by the observation of frequent virions with multiple cores and the determination that HIV-1 appears to contain on the order of 5000 Gag proteins/particle, higher than previously expected (5).

Despite the apparent agreement between studies on particles and our investigations on HIV-1 capsid proteins in vitro, our capsid proteins lacked other PrGag domains that might have an effect on Gag protein oligomerization, particularly as the proteins are organized in immature assemblies. To address this issue, we have undertaken the in vitro analysis of multidomain Gag proteins, carrying matrix, capsid, spacer 1 (SP1), and nucleocapsid domains. Interestingly, these MACANC proteins assembled efficiently in the absence of other components at 37 °C but not at lower temperatures, suggesting a temperature-sensitive assembly switch. We also observed that the proteins preferentially assembled into tube forms at low pH but into spheres at high pH. Because these tube and sphere forms were not amenable to structural analysis, we additionally examined membrane-bound assemblies of the proteins. Image analysis of membrane-bound protein arrays indicated that the MACANC proteins organized into hexamer rings with a spacing of 79.7 Å. Our results indicate that the presence of MA and NC domains is compatible with a tightly packed arrangement of PrGag proteins and support a model where PrGag processing yields a loose capsid core from a more crowded precursor arrangement.


   EXPERIMENTAL PROCEDURES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
RESULTS
 DISCUSSION
 REFERENCES
 
Protein Purification—MACANC and MACANCexact proteins were expressed in the Escherichia coli strain BL21(DE3)/pLysS (Novagen) from the bacterial expression plasmids pET15B-HIVMACANC and pET15B-HIVMACANCexact. Both constructs encode an N-terminal His tag of MGSSH HHHHH SSGLV PRGSH MLEDPP fused to the second codon of HIV-1 HXB2 gag. The pET15B-HIVMACANCexact plasmid expresses protein, which terminates precisely at the HIV-1 NC C terminus; the pET15B-HIVMACANC plasmid encodes the vector-derived residues KIRAA NKARK EAELA AATAEQ at the C terminus of NC. Bacterial expression and purification of proteins by nickel chelate chromatography followed methods described previously (2, 8–12). Purified protein fractions were desalted by buffer exchange in Sephadex G25 spin columns in 10 mM sodium phosphate (pH 7.8), 500 mM NaCl, and stored under nitrogen at –80 °C. Purified (>90%) proteins at 0.5–1.5 mg/ml as well as protein purification fractions were evaluated after SDS-PAGE (13) by Coomassie Blue staining (2, 8–12) and immunoblot detection. As described previously (13), immunoblot detection employed an anti-HIV capsid monoclonal antibody (Hy183) as the primary antibody (13), an alkaline phosphatase-conjugated anti-mouse secondary antibody (Promega), and visualization of bands using a color reaction of nitro blue tetrazolium plus 5-bromo-4-chloro-3-indolyl phosphate in 100 mM Tris-hydrochloride (pH 9.5), 100 mM NaCl, and 5 mM MgCl2.

Pelleting Assays—Incubations of 10 µM MACANC proteins were performed in 20-µl reactions at 23, 30, or 37 °C for 16 h. Incubation buffers consisted of 375 mM NaCl, 5 mM dithiothreitol, 100 mM Hepes (pH 8.0 or 6.0) in the presence or absence of 0.2 mg/ml E. coli rRNA (Roche Applied Science). Following incubations, reactions were centrifuged for 10 min at 4 °C at 14,000 x g. Supernatant and pellet fractions were collected for fractionation by SDS-PAGE and visualized by Coomassie Blue staining.

Two-dimensional Crystallization—Two-dimensional crystallization incubations were essentially as described previously (2, 8–12). Briefly, 10-µl drops of protein (0.25–0.75 mg/ml) in buffer (25 mM sodium phosphate (pH 8.3), 500 mM NaCl, 5 mM sodium acetate, 20% glycerol) were incubated beneath 1 µl of 200 ng/µl phosphatidylcholine (Avanti) plus 50 ng/µl nickel-charged 1,2-dioleoyl-sn-glycero-3-[N-((5-amino-1-carboxypentyl)iminodiacetic acid)-succinyl] (Avanti) in 1:1 chloroform: hexane. Incubations were performed for 16–24 h at 22–25 °C in sealed plastic dishes humidified with 6.25 mM sodium phosphate (pH 8.3) and 125 mM NaCl. Following the incubations, monolayers and associated proteins were lifted onto lacey carbon grids (Ted Pella 01883) for 30–60 s, washed on drops of distilled water for 15–30 s, wicked, stained with 1.33% uranyl acetate, wicked again, and allowed to dry for at least 15 min.

Samples were viewed on a Philips CM120 transmission electron microscope (EM) under low dose conditions. All images were taken at 100 kV and at defocus values between 200 and 1500 nm. Images were collected on either a Gatan 794 charge-coupled device multiscan camera or on Kodak SO163 negative films. For processing, charge-coupled device images were converted from the Gatan Digital Micrograph 3 (DM3) format to 8-bit gray scale TIFF format using Gatan software. Negatives were scanned and digitized using a Nikon Coolscan 8000 film and slide scanner in super fine scan (4000 dots per inch) mode with the multi-sampling mode at 8x. Images were saved in the TIFF format using the Nikon Scan 3.1 software.

Sphere and Tube Incubations—Proteins were incubated at pH 6.0 (for tubes) or pH 8.0 (for spheres) beneath lipid monolayers and then lifted onto lacey EM grids prior to staining, EM viewing, and image collection as described above.

Unit Cell Statistics—Untilted two-dimensional crystal images converted from the TIFF to the MRC format as described previously (2, 8–12) were boxed using the MRC BOXMRC program (2, 14, 15), and Fourier-transformed using the 2DFFT function of the ICE program (2, 16, 17). Transform TNF files were viewed as power spectra on SPECTRA (18), hand-indexed, and converted to amplitudes and phases files using the LATREF, UNBENDA, and MNBOX functions of SPECTRA (18). Unit cell dimensions were obtained using the SPECTRA file information interface from indexed transforms, and axis lengths were corrected by dividing raw values by sin{gamma}* (2). Space group analysis was performed using ALLSPACE in 3° phase origin search steps (2, 14, 15).

Three-dimensional Reconstructions—Tilt series images of membrane-bound protein arrays were collected on negatives at tilt angles of –55°, –45°, –30°, –15°, 0°, +15°, +30°, +45°, and +50° and scanned as described above to produce TIFF files. From these TIFF files, 30 separate crystalline patches were tracked through the entire range of tilt angles, and 512 x 512 pixel areas for each angle of each of the 30 crystalline patches were windowed using the Gatan Digital Micrograph software. The first steps in preparation of 0° reference images were performed using the digital micrograph software; the 512 x 512 pixel areas were Fourier-transformed, hand-indexed from power spectra, masked, low pass-filtered to include only the three lowest resolution reflections (1,0; 1,1; 2,0), back-transformed, and windowed to obtain 128 x 128 pixel (262.4 x 262.4 Å) raw reference images. Raw 0° reference images along with all 512 x 512 pixel areas from the 30 tilt series were converted from the TIFF format to the MRC format and then to the SPIDER SPI format (19), and subsequent processing steps were performed using the SPIDER suite (19) of programs.

To generate final 0° reference images, raw 128 x 128 pixel images were thresholded with the SPIDER TH command so that pixels with gray scale values below the image average were raised to the image average. After this, images were padded to 512 x 512 pixels using the PD command and the original average gray scale value as background to create the final 0° tilt reference images. At this point, references were used in cross-correlations (CC command) versus their original 512 x 512-pixel 0° tilt images to generate cross-correlation maps. Peaks in cross-correlation maps were hand-picked using the WEB interface (19) after rotating the maps by 180° using the SPIDER RT 90 command to account for the fact that SPIDER designates the origin as the lower right corner of the image, whereas the WEB origin is at the upper left. From 20 to100 hand-picked cross-correlation peak locations per 0° tilt image, corresponding 148 x 148 pixel windows in the original images were cut and averaged using the WV command. Final 128 x 128 pixel 0° averages centered on hexameric protein-free centers were cut using the WV command, and an image center location was hand-picked from the180° rotated image on WEB. After obtaining 30 0° averages, the steps described above were used to generate averages for each tilt angle in each tilt series. During this process, thresholded and padded 0° averages were used as references in –15 and +15° cross-correlation searches, and output averages were used recursively in obtaining higher tilt averages.

Of the original 30 tilt series, 226 image averages from a total of 8455 windows were used for three-dimensional reconstructions using the SPIDER suite (19). Average pixel intensity values for each image average were normalized and inverted (to account for negative staining) and rotated +223° to align the tilt axes with the image y axes. Each set of tilt series averages then was used to create a three-dimensional volume by back-projection using the BP three-dimensional command and input phi, theta, psi Eulerian angles of 0.0, tilt angle, 0.0. Determination of the relative rotations between tilt series was accomplished using the AP RA command on two-dimensional projections from 3-fold symmetrized and calculated tilt series volumes. The output of these operations included rotation angles, which corresponded to the reverse of the psi angles needed for final reconstructions. Thus, the 226 tilt series image averages were employed as input for final three-dimensional reconstructions using the SPIDER three-dimensional BP command. The quality of three-dimensional reconstructions was assessed by splitting data sets in half, calculating three-dimensional volumes for each half-data set, and comparing the two half-data sets using the RF 3 command. The resolution of the reconstruction was estimated to be at 27.5 Å, based on the resolution at which a phase residual value of 45° was reached (2, 8–12, 19). Final filtered volumes were viewed using WEB (19) or a visualization tool kit-based ISO_VIEW interface, and were converted to the TIFF or JPEG format for presentation.

Rotational Correlations—0° tilt average images from tilt series were compared with rotated versions by use of the SPIDER CC command. Maximum peak values (1.0 is defined as a perfect match) were recorded and plotted for comparisons made at 1° intervals.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
DISCUSSION
 REFERENCES
 
In Vitro Assembly of MACANC Proteins—To investigate the influences that the HIV-1 Gag MA and NC domains might have on the assembly of capsid proteins, His-tagged MACANC proteins were expressed in bacteria (see "Experimental Procedures"). Two proteins, MACANC and MACANCexact, were made. These differed by the presence of 21 vector-derived codons at the C terminus of MACANC, but because the additional encoded residues did not appear to have an effect in our studies, we will refer to them interchangeably as MACANC proteins. As shown in Fig. 1, both of the MACANC proteins carried N-terminal His tags in place of the PrGag myristate (20) group, and neither encoded the PrGag p6 domain, which does not appear to play a structural role but does hamper purification of bacterially expressed proteins (21, 22). Following previous methods (see Refs. 2 and 8–12 and "Experimental Procedures"), the proteins were purified using nickel chelate chromatography to concentrations of 0.5–1.5 mg/ml and estimated purities of >90% (Fig. 2, lane I).



View larger version (6K):
[in this window]
[in a new window]
  		FIG. 1.
HIV Gag proteins. The HIV-1 Gag precursor protein Pr55Gag is an N-terminally myristoylated precursor protein composed of the matrix, capsid (CA), NC, and p6 domains and the spacer peptides SP1 and SP2. For bacterial expression and purification, the MACANC and MACANCexact proteins carry an N-terminal histidine tag fused to the second codon of HIV-1 HXB2 gag and are truncated either exactly at the NC C terminus (MACANCexact) or after 21 additional vector-derived residues (MACANC). Because the additional vector-encoded sequences did not appear to have an effect in our studies, we refer to them collectively as MACANC proteins. As shown, both proteins carry the HIV-1 MA, capsid, SP1, and NC domains.

 



View larger version (81K):
[in this window]
[in a new window]
  		FIG. 2.
Protein purification. Nickel chelate chromatography column fractions of the bacterially expressed MACANC protein were fractionated by SDS-PAGE and visualized by Coomassie Blue staining. Column fractions consisted of starting bacterial lysate material (A), column-loading buffer washes (B and C), 10 mM imidazole buffer washes (D and E), and elutions (F–I). Final elutions yielded a protein with a mobility similar to the 52.2-kDa marker in the protein standards lane (J), which was detected by anti-HIV capsid antibodies in immunoblots performed in parallel.

 
For monitoring the assembly properties of the MACANC proteins, we employed pelleting assays in which purified proteins were incubated under varying conditions. Following incubations, centrifugation was used to separate pelletable assembled products from monomers and small oligomers in the supernatant, and protein levels in each fraction were visualized by staining after electrophoresis. Results from these studies demonstrated a marked temperature dependence in the assembly process. As illustrated in Fig. 3, lanes 1 and 2, during 23 °C incubations, very little MACANC protein assembled into a pelletable form at either pH 6.0 (A) or pH 8.0 (B). In contrast, incubations performed at 30 and 37 °C yielded increasing amounts of assembled protein (Fig. 3, lanes 5 and 6 and 9 and 10).



View larger version (60K):
[in this window]
[in a new window]
  		FIG. 3.
Assembly properties of His-MACANC proteins. Incubations containing 10 µM MACANC proteins were performed at the indicated pHs (A, pH 6.0; B, pH 8.0) and temperatures in the absence or presence of 0.2 mg/ml E. coli rRNA. After incubations, supernatant (S) and pellet (P) fractions were separated by centrifugation. Samples were subjected to SDS-PAGE and visualized by Coomassie Blue staining. Arrowheads indicate the full-length MACANC proteins.

 
Because NC domains on PrGag proteins have been postulated to facilitate assembly after binding to RNA (23–25), we repeated the pelleting assays in the presence of RNA (Fig. 3). As illustrated in Fig. 3, lanes 3 and 4, 7 and 8, and 11 and 12, although assembly efficiency still appeared affected by temperature at pH 6.0 (Fig. 3A), RNA addition modulated this effect. The influence of RNA was even more pronounced at pH 8.0 (Fig. 3B, lanes 3 and 4, 7 and 8, and 11 and 12). These results suggest that putative NC-RNA interactions decrease the temperature sensitivity of MACANC assembly reactions, an observation that is discussed below (see "Discussion").

What are the morphologies of the structures assembled by MACANC proteins? To examine this issue, assembly products were lifted onto EM grids, negatively stained, and viewed by transmission electron microscopy. Of note was the fact that pH 8.0 MACANC incubations yielded spherical particles with diameters on the order of 100 nm. In some cases, these spheres clustered into paracrystalline arrays (Fig. 4). In contrast, pH 6.0 incubation of MACANC proteins gave large tubes, which at low magnification appeared as dark rods of up to 750 nm in length against the grid substrate (Fig. 5, A and B). At higher magnification, the MACANC rods appeared to be 60–70-nm-wide cylinders with hollow centers of 20–25 nm (Fig. 5, C and D). Our observation of different particle morphologies at different pHs is in agreement with the observations of others (22, 26) and is consistent with the notion that pH regulates the HIV-1 Gag protein assembly pathway, at least in vitro. Additionally, the cylinder wall widths in Fig. 5, C and D (20–25 nm) are comparable with the distances observed for the MA plus capsid plus NC domains observed in radial density profiles of immature HIV-1 virus-like particles (27, 28), suggesting a similar MA-capsid-NC radial arrangement in our tubes. Despite these correlations, the MACANC sphere and tube forms were not amenable to higher resolution analysis of how the proteins associate in higher order assemblies. Consequently, we opted to examine membrane-bound assemblies of MACANC proteins.



View larger version (105K):
[in this window]
[in a new window]
  		FIG. 4.
Morphology of MACANC spheres. Spheres assembled during pH 8.0 incubations of MACANC were lifted onto lacey EM grids, negatively stained with uranyl acetate, and imaged by transmission electron microscopy. As compared with the 200-nm size standard, sphere diameters appeared to be ~100 nm.

 



View larger version (136K):
[in this window]
[in a new window]
  		FIG. 5.
Morphology of MACANC tubes. After protein incubation at pH 6.0, tubes of up to 750 nm in length were observed by electron microscopy at low magnification as dark rods on gray carbon lace-lipid monolayer substrates (A and B). At higher magnification (C and D), the tubes appeared as hollow cylinders of 60–70 nm in diameter with white (protein) walls and central pores containing accumulated uranyl acetate negative stain (dark). Size standards are as indicated.

 
MACANC Membrane-bound Arrays—As noted above, MACANC spheres and tubes were not amenable to image analysis techniques that might help clarify how protein units associate to form higher order assemblies. This was because the MACANC tubes were not regular enough to give clear diffraction patterns, and the spheres were not compatible with icosahedral reconstruction methods. As an alternative approach, we examined the assembly of MACANC proteins on membrane monolayers, as we have done previously (2, 8–12). To do so, the N-terminally His-tagged MACANC proteins were incubated beneath nickel-chelating lipid monolayers, lifted along with the membranes onto EM grids, and imaged by a transmission EM. Images of such membranes showed darkly staining protein patches on lightly staining lipid layers (Fig. 6A). Although the MACANC arrays were only 50–200 nm wide, Fourier transforms of boxed arrays yielded calculated diffraction patterns with hexagonal sets of reflections (Fig. 6B). The lowest resolution 1,0 reflections were observed at 1/69 Å, corresponding to a real space hexagonal unit cell spacing of 79.7 Å, consistent with the tightly packed arrangement observed for HIV-1 capsid proteins. Indeed, back-transformation of filtered reflections gave a two-dimensional projection image (Fig. 6C) of proteins (in white) viewed perpendicular to membranes, surrounding protein-free holes spaced at 8-nm distances.



View larger version (148K):
[in this window]
[in a new window]
  		FIG. 6.
Membrane-bound assemblies of MACANC proteins. When incubated beneath membrane monolayers containing nickel-chelating lipids, MACANC proteins assembled two-dimensional crystalline patches, which appear as dark areas on negatively stained membranes (A). Fourier transformation of crystalline patch areas yielded diffraction patterns (power spectra) with a set of six reflections at 1/69 Å, one of which is boxed in B. After indexing and masking hexagonal reflections from diffraction patterns, back-transformation yielded Fourier-filtered, real space images as shown in C, where protein areas are light, and protein-free zones are dark. Note that for a hexagonal unit cell, unit cell dimensions of ~8 nm, indicative of the spacing between the large protein-free holes, correspond to the 1/69 Å 1,0 reflections observed in diffraction patterns.

 
Compilation of data from multiple crystalline patches supported the results shown in Fig. 6. Our average membrane-bound MACANC unit cell had edges of a = b = 79.7 ± 2.0 Å and a reciprocal space {gamma}* angle of 60.9° (Table I). Consistent with the hexagonal appearance of the two-dimensional projection (Fig. 6C), phase comparisons of Fourier transform reflections gave low phase residuals when unit cells were assumed to be p1, p2, p3, or p6 symmetry (Table I). Based on the observation that MACANC crystalline patches obeyed p3 or p6 symmetry, we modified a single-particle reconstruction approach to obtain a three-dimensional model of membrane-bound MACANC from low dose tilt series of the crystalline patches. To do so, crystal unit cell centers were identified in untilted images by cross-correlation with Fourier-filtered projections such as that in Fig. 6C. The highest peaks in cross-correlation maps such as those shown in Fig. 7 (upper panel, 0° map) were used to pick real space windows to sum in image averages (Fig. 7, lower panel). For each series, after obtaining a 0° tilt image average (Fig. 7, lower panel, center image), the average was used in successive cross-correlation, windowing, and averaging steps to obtain image averages for tilt angles in the range of –45 to +45° (Fig. 7).


View this table:
[in this window]
[in a new window]
  		TABLE I
Unit cell statistics

Unit cell statistics were obtained from images of untilted membrane-bound crystalline arrays. The real space a axis length was calculated from Fourier transform a* axes and corrected for {gamma}* angles. Statistics for space group fitting were obtained from the program ALLSPACE, which compares the phases of Fourier transform reflections for internal consistency with two-dimensional space groups and outputs phase residuals for consistency comparisons. Using this algorithm, a perfect fit gives a phase residual of 0°, whereas a random fit yields a 90° phase residual. Note that because internal phase residual comparisons are not relevant with the primitive (p1) space group, the phase residual for p1 is on the basis of signal-to-noise ratios of observed amplitudes. n, indicates the number of transforms that contributed to the averages and standard deviations. Note that space group fitting statistics were obtained from transforms in which a minimum of seven comparisons could be made.

 



View larger version (70K):
[in this window]
[in a new window]
  		FIG. 7.
Projection images of membrane-bound MACANC proteins. Crystalline patches of membrane-bound MACANC proteins were identified as white peaks in cross-correlation maps (upper panel) as described under "Experimental Procedures." Major peaks in cross-correlation maps were used for picking windows of crystalline patches for real space averaging to generate a panel of projection image averages (lower panel) from micrographs taken at different degrees of tilt. Note that protein areas appear white in the image averages and that the size bar and tilt axis for this tilt series (number 19) are as indicated.

 
With the 0° tilt averages obtained from our 30 tilt series, it was of interest to examine whether the MACANC proteins assembled around protein-free holes in a trigonal (p3) or hexagonal (p6) fashion. For this analysis, individual 0° tilt averages from different tilt series were compared with rotated versions of themselves by cross-correlation. The rationale for this approach was that for a trigonal arrangement, one would predict a set of three roughly equivalent cross-correlation peaks at 0°, 120°, and 240°; whereas for a hexagonal pattern, six equal peaks spaced at 60° intervals were predicted. As shown in Fig. 8, the latter prediction was observed, supporting the interpretation that MACANC proteins align on membranes in a hexagonal fashion.



View larger version (28K):
[in this window]
[in a new window]
  		FIG. 8.
Rotational correlation of membrane-bound proteins. 0° tilt image averages from tilt series number 13 (A), number 17 (B), and number 30 (C) were compared with rotated versions to examine unit cell symmetry. Plots show maximum peak values obtained in cross-correlations performed with unrotated images versus the same images rotated at the indicated angles. Note that perfect matching corresponds to a peak value of 1.0.

 
To generate a three-dimensional model of membrane-bound MACANC proteins, tilt image averages from the 30 tilt series were used to build a three-dimensional volume by back-projection (see "Experimental Procedures"). The quality of the reconstruction was evaluated by splitting the data set in half, calculating three-dimensional volumes for each half-set, and comparing the two half-set volumes. Using this methodology, we estimated our three-dimensional reconstruction to be at 27.5 Å resolution. Not surprisingly, surface renderings of the three-dimensional reconstruction viewed perpendicular to the membrane from either membrane side up (Fig. 9A) or membrane side down (Fig. 9B) look similar to the two-dimensional projection shown in Fig. 6C; central protein-free holes are surrounded by interconnected dumbbell units (highlighted in white in Fig. 9B) that we interpret to be dimers of capsid NTDs.



View larger version (36K):
[in this window]
[in a new window]
  		FIG. 9.
Protein assemblies viewed from below and above membrane monolayers. Three-dimensional reconstruction of membrane-bound MACANC assemblies to a resolution of 27.5 Å was performed as described under "Experimental Procedures." A shows the arrangement of proteins viewed perpendicular to the membrane from below, and B shows the arrangement viewed from above. Proteins appear gray against a black background, and a putative symmetric capsid dimer is outlined in white. The volumes were depicted using WEB at a contour level of ~1.1 S.D.

 
At a lower contour level, a view of the bottom surface of the volume shows putative CTD connections as peanut shapes extending radially from hole centers (Fig. 10A). Successive images of the volume tilted gradually to the reader (Fig. 10, B–D) show additional views of the protein ring arrangement. In the final panel (Fig. 10D), the view is parallel to the membrane with the membrane side up. Although the knobs at the bottom of the panel could represent some portion of NC density, neither NC nor MA domains are resolved in the reconstruction. Nevertheless, the arrangements of putative capsid NTDs and CTDs and the relatively tight packing of the hexamer rings are similar to the tightly packed capsid pattern that we have observed previously (2). These results demonstrate that MA and NC domains do not interfere with HIV-1 capsid tight packing preferences and support the notion that immature virions feature closely associated capsid hexamer rings.



View larger version (40K):
[in this window]
[in a new window]
  		FIG. 10.
Organization of membrane-bound MACANC proteins. Three-dimensional reconstruction of membrane-bound MACANC assemblies to a resolution of 27.5 Å was performed as described under "Experimental Procedures." A is oriented so that the membrane is perpendicular to the viewer with the membrane side away from the viewer. B is tilted 30° toward the viewer; C is tilted 60° toward the viewer; and D is tilted 90° toward the viewer such that the view is parallel to the membrane with the membrane side to the top of the page. Proteins appear as gray or white against a black background and are depicted using WEB at a contour level of ~0.88 S.D.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
REFERENCES
 
Although much is now known concerning the three-dimensional structures of the individual HIV-1 Gag protein structures (29–37), considerably less is understood about how the Gag proteins or PrGag protein domains associate in immature and mature HIV virions. Our recent demonstration that HIV-1 capsid proteins can assemble onto membranes in either tightly packed or loosely packed hexagonal arrangements (2) substantiated observations concerning capsid contacts in alternative in vivo and in vitro assemblies (3–7) and suggested a model for viral morphogenesis (2). However, the contributions of matrix and nucleocapsid domains were not tested in our studies, and thus we decided to examine the interactions of MACANC proteins in vitro. It should be noted that our proteins have excluded the C-terminal PrGag SP2 and p6 domains. Previous work suggested that the in vitro assembly properties of p6-deleted and partially pure full-length PrGag proteins are similar (21), and that p6 domain deletions appear to have no effects on radial density profiles of immature virions (28). However, it has been observed in fluorescence studies that full-length HIV-1 PrGag proteins are not induced to oligomerize by the addition of tRNA (43). Thus, although that study employed a different RNA and lower concentrations of protein (200 nM) than we did, our results might not be completely comparable with results from full-length PrGag proteins.

One observation (38) consistent with our SP2-p6-truncated MACANC proteins is that they preferentially assembled into higher order structures at 37 °C versus 23 or 30 °C (Fig. 3). One interpretation of this result is that the increased temperature may be necessary for the proteins to overcome an energy barrier between assembly impaired and assembly efficient conformations. If this is the case, then at least two models might explain how RNA (Fig. 3) might facilitate assembly. In one case, MACANC binding to RNA might substitute for the temperature effect by lowering the energy barrier for the conformation change. Alternatively, a two-step process might be envisioned, where MACANC-RNA binding might increase the levels of an assembly intermediate. Either way, it is interesting that in the absence of RNA and membranes, MACANC proteins assembled tubes at pH 6.0 and spheres at pH 8.0 as was observed previously (22). Because HIV-1 capsid proteins can assemble tubes in either a loose packing arrangement (6) or in a tight packing arrangement,2 we do not believe that MACANC tube or sphere morphologies are dictated by loose versus tight hexamer packing arrangements. Rather, we speculate that the tube and sphere morphologies employ the same type of hexamer packing but differ in the frequency and location of pentamer placement (39, 40).

If HIV-1 MACANC tubes or spheres were of high enough regularity, it might be possible to extract unit cell information from their Fourier transforms (2, 5–7). Alternatively, we opted to investigate the assembly properties of the His-tagged proteins on nickel-chelating membrane monolayers. As demonstrated in Table I and Figs. 6, 9, and 10, the membrane-bound proteins associated to form hexamer rings (Fig. 8, Table I) around protein-free holes, which may accommodate HIV envelope protein cytoplasmic tails in immature virions. The hole-to-hole distance that we observed is in the range of that observed for tightly packed hexamers in immature virus forms (2–5) and below that observed for mature virus cores assembled in vivo and in vitro (6, 7). Thus, the presence of MA and NC domains on MACANC proteins is wholly compatible with a tightly packed pattern. These and other results support a model where immature HIV-1 virions incorporate about 5000 PrGag proteins (5) in a tightly packed arrangement and that only a fraction of the PrGag capsid domains contribute to the formation of mature cores, composed of loosely packed hexamers. Although the excess capsid proteins may assemble an additional virus core within a mature virion (7), the fate of the unutilized capsid is unknown. Also unknown are the precise contacts made by the MA and NC domains, which are not resolved well, presumably because of their greater freedom of motion within arrays and/or the often observed loss of resolution perpendicular to the membrane plane (2, 10–12, 15). Because the assembly function of NC domains can be replaced by protein dimerization domains (41, 42), it seems probable that NC domains group in pairs in PrGag assemblies. In contrast, matrix domains have shown a tendency to trimerize (29, 30), suggesting that they will associate as trimers within PrGag arrays. A model to accommodate these observations and our current results is illustrated in Fig. 11. As shown, NTDs are grouped in hexamer rings using asymmetric, side-by-side contacts and are tightly packed by virtue of putative symmetric contacts between the NTDs of adjacent hexamers. Above the NTD hexamers are MA trimers, at the point where three hexamers come together. Aligned beneath putative NTD dimers are CTD dimers, which interconnect hexamer rings, and NC dimers, presumably associated with the viral HIV-1 RNA. It will be of interest to test this and other models of HIV assembly in vitro and in vivo.



View larger version (22K):
[in this window]
[in a new window]
  		FIG. 11.
Model for PrGag contacts. Shown is a model for how HIV-1 MACANC proteins associate at membranes in immature virions. In A, domains of PrGag proteins in hexameric arrays are viewed at increasing distances away from the membrane. In the MA panel, an MA trimer unit is outlined. In the capsid NTD panel, a hexamer ring of NTD units is circled in which asymmetric side-by-side contacts are employed. Also circled is a symmetric, head-to-head NTD dimer, and similar dimer units are outlined in the capsid CTD and NC panels. At the bottom is the composite of the MA, NTD, CTD, and NC layers (All). In B, a subset of contacts is depicted as viewed nearly parallel to the membrane. As shown, NTDs form hexamer rings that are tightly associated via symmetric contacts and interconnected via symmetric CTD dimers. NC dimers are proposed to align beneath CTDs, whereas MA domains are modeled as trimers at the nexus of three neighboring hexamer rings. Note that for clarity, only two or three CTDs and only the NC and MA domains are depicted.

 

   FOOTNOTES
 
* This work was supported by NIGMS, National Institutes of Health Grant GM060170 (to E. B.) and American Foundation for AIDS Research (amfAR) Postdoctoral Fellowship Grant 106523-35-RFNT (to A. A.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Back

{ddagger} To whom correspondence should be addressed: Mail Code L220, 3181 S. W. Sam Jackson Park Rd., Portland, OR 97201-3098. Tel.: 503-494-8098; Fax: 503-494-6862; E-mail: barklis{at}ohsu.edu.

1 The abbreviations used are: HIV, human immunodeficiency virus; PrGag, precursor Gag protein; MA, matrix; NC, nucleocapsid; EM, electron microscope; NTD, N-terminal domain; CTD, C-terminal domain; SP1, spacer 1; SP2, spacer 2. Back

2 D. Huseby and E. Barklis, unpublished observations. Back



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Swanstrom, R., and Wills, J. W. (1997) in Retroviruses (Coffin, J. M., Hughes, S. H., and Varmus, H. E., eds) pp. 263–334, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY
  2. Mayo, K., Huseby, D., McDermott, J., Arvidson, B., Finlay, L., and Barklis, E. (2003) J. Mol. Biol. 325, 225–237[CrossRef][Medline] [Order article via Infotrieve]
  3. Nermut, M., Hockley, D., Jowett, J., Jones, I., Garreau, M., and Thomas, D. (1994) Virology 198, 288–296[CrossRef][Medline] [Order article via Infotrieve]
  4. Nermut, M., Hockley, D., Bron, P., Thomas, D., Zhang, W., and Jones, I. (1998) J. Struct. Biol. 123, 143–149[CrossRef][Medline] [Order article via Infotrieve]
  5. Briggs, J., Simon, M., Gross, I., Krausslich, H.-G., Fuller, S., Vogt, V., and Johnson, M. (2004) Nat. Struct. Mol. Biol. 11, 672–675[CrossRef][Medline] [Order article via Infotrieve]
  6. Li, S., Hill, C., Sundquist, W., and Finch, J. (2000) Nature 407, 409–413[CrossRef][Medline] [Order article via Infotrieve]
  7. Briggs, J., Wilk, T., Welker, R., Krausslich, H.-G., and Fuller, S. (2003) EMBO J. 22, 1707–1715[CrossRef][Medline] [Order article via Infotrieve]
  8. Barklis, E., McDermott, J., Wilkens, S., Schabtach, E., Schmid, M. F., Fuller, S., Karanjia, S., Love, Z., Jones, R., Rui, Y., Zhao, Z., and Thompson, D. (1997) EMBO J. 16, 1199–1213[CrossRef][Medline] [Order article via Infotrieve]
  9. Barklis, E., McDermott, J., Wilkens, S., Fuller, S., and Thompson, D. (1998) J. Biol. Chem. 273, 7177–7180[Abstract/Free Full Text]
  10. McDermott, J., Mayo, K., and Barklis, E. (2000) J. Mol. Biol. 302, 121–133[CrossRef][Medline] [Order article via Infotrieve]
  11. Mayo, K., McDermott, J., and Barklis, E. (2002) Virology 298, 30–38[CrossRef][Medline] [Order article via Infotrieve]
  12. Mayo, K., Vana, M., McDermott, J., Huseby, D., Leis, J., and Barklis, E. (2002) J. Mol. Biol. 316, 667–678[CrossRef][Medline] [Order article via Infotrieve]
  13. Arvidson, B., Seeds, J., Webb, M., Finlay, L., and Barklis, E. (2003). Virology 308, 166–177[CrossRef][Medline] [Order article via Infotrieve]
  14. Amos, L.A., Henderson, R.H., and Unwin, P.N.T. (1982) Prog. Biophys. Mol. Biol. 39, 183–231[Medline] [Order article via Infotrieve]
  15. Crowther, R., Henderson, R., and Smith, J. (1996) J. Struct. Biol. 116, 9–16[CrossRef][Medline] [Order article via Infotrieve]
  16. Unwin, P.N.T., and Henderson, R. (1975) J. Mol. Biol. 94, 425–440[CrossRef][Medline] [Order article via Infotrieve]
  17. Hardt, S., Wang, B., and Schmid, M. F. (1996) J. Struct. Biol. 116, 68–70[CrossRef][Medline] [Order article via Infotrieve]
  18. Schmid, M. F., Dargahi, R., and Tam, M. (1993) Ultramicroscopy 48, 251–264[CrossRef][Medline] [Order article via Infotrieve]
  19. Frank, J., Radermacher, M., Penczek, P., Zhu, J., Li, Y., Ladjadj, M., and Leith, A. (1996) J. Struct. Biol. 116, 190–199[CrossRef][Medline] [Order article via Infotrieve]
  20. Bryant, M., and Ratner, L. (1990) Proc. Natl. Acad. Sci. U. S. A. 87, 523–527[Abstract/Free Full Text]
  21. Campbell, S., and Rein, A. (1999) J. Virol. 73, 2270–2279[Abstract/Free Full Text]
  22. Gross, I., Hohenberg, H., Wilk, T., Wiegers, K., Grattinger, M., Muller, B., Fuller, S., and Krausslich, H.-G. (2000) EMBO J. 19, 103–113[CrossRef][Medline] [Order article via Infotrieve]
  23. Campbell, S., and Vogt, V. (1995) J. Virol. 69, 6487–6497[Abstract]
  24. Ma, Y., and Vogt, V. (2002) J. Virol. 76, 5452–5462[Abstract/Free Full Text]
  25. Ma, Y., and Vogt, V. (2004) J. Virol. 78, 52–60[Abstract/Free Full Text]
  26. Ehrlich, L., Liu, T., Scarlata, S., Chu, B., and Carter, C. (2001) Biophys. J. 81, 586–594[Medline] [Order article via Infotrieve]
  27. Fuller, S., Wilk, T., Gowen, B., Krausslich, H.-G., and Vogt, V. (1997) Curr. Biol. 7, 729–738[CrossRef][Medline] [Order article via Infotrieve]
  28. Wilk T., Gross, I., Gowen, B., Rutten, T., de Haas, F., Welker, R., Krausslich, H.-G., Boulanger, P., and Fuller, S. D. (2001) J. Virol. 75, 759–771[Abstract/Free Full Text]
  29. Rao, Z., Belyaev, A., Fry, E., Roy, P., Jones, I., and Stuart, D. (1995) Nature 378, 743–747[CrossRef][Medline] [Order article via Infotrieve]
  30. Hill, C., Worthylake, D., Bancroft, D., Christensen, A., and Sundquist, W. (1996) Proc. Natl. Acad. Sci. U. S. A. 93, 3099–3104[Abstract/Free Full Text]
  31. Gitti, R., Lee, B., Walker, J., Summers, M., Yoo, S., and Sundquist, W. (1996) Science 273, 231–235[Abstract]
  32. Gamble, T., Vajdos, F., Yoo, S., Worthylake, D., Houseweart, M., Sundquist, W., and Hill, C. (1996) Cell 87, 1285–1294[CrossRef][Medline] [Order article via Infotrieve]
  33. Gamble, T., Yoo, S., Vajdos, F., von Schwedler, U., Worthylake, D., Wang, H., McCutcheon, J., Sundquist, W., and Hill, C. (1997) Science 278, 849–853[Abstract/Free Full Text]
  34. Momany, C., Kovari, L., Prongay, A., Keller, W., Gitti, R., Lee, B., Gorbalenya, A., Tong, L., McClure, J. Ehrlich, L., Summers, M., Carter, C., and Rossmann, M. (1996) Nat. Struct. Biol. 3, 763–770[CrossRef][Medline] [Order article via Infotrieve]
  35. Tang, C., Ndassa, Y., and Summers, M. F. (2002) Nat. Struct. Biol. 9, 537–543[Medline] [Order article via Infotrieve]
  36. DeGuzman, R., Wu, Z., Stalling, C., Pappalardo, L, Borer, P., and Summers, M. (1998) Science 279, 384–390[Abstract/Free Full Text]
  37. Amarasinghe, G., De Guzman, R., Turner, R., Chancellor, K., Wu, Z., and Summers, M. (2000) J. Mol. Biol. 301, 491–511[CrossRef][Medline] [Order article via Infotrieve]
  38. Morikawa, Y., Goto, T., and Momose, F. (2004) J. Biol. Chem. 279, 31964–31972[Abstract/Free Full Text]
  39. Ganser, B., Li, S., Klishko, V., Finch, J., and Sundquist, W. (1999) Science 283, 80–83[Abstract/Free Full Text]
  40. Ganser-Pornillos, B., von Schwedler, U., Stray, K., Aiken, C., and Sundquist, W. (2004) J. Virol. 78, 2545–2552[Abstract/Free Full Text]
  41. Zhang, Y., Qian, H., Love, Z., and Barklis, E. (1998) J. Virol. 72, 1782–1789[Abstract/Free Full Text]
  42. Johnson, M., Scobie, H., Ma, Y., and Vogt, V. (2002) J. Virol. 76, 11177–11185[Abstract/Free Full Text]
  43. Provitera, P., Goff, A., Harenberg, A., Bouamr, F., Carter, C., and Scarlata, S. (2001) Biochemistry 40, 5565–5572[CrossRef][Medline] [Order article via Infotrieve]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?


This article has been cited by other articles:


Home page
 J. Virol.Home page
I. B. Hogue, A. Hoppe, and A. Ono
Quantitative Fluorescence Resonance Energy Transfer Microscopy Analysis of the Human Immunodeficiency Virus Type 1 Gag-Gag Interaction: Relative Contributions of the CA and NC Domains and Membrane Binding
J. Virol., July 15, 2009; 83(14): 7322 - 7336.
[Abstract] [Full Text] [PDF]

Home page
 Proc. Natl. Acad. Sci. USAHome page
D. Ivanov, O. V. Tsodikov, J. Kasanov, T. Ellenberger, G. Wagner, and T. Collins
Domain-swapped dimerization of the HIV-1 capsid C-terminal domain
PNAS, March 13, 2007; 104(11): 4353 - 4358.
[Abstract] [Full Text] [PDF]

Home page
 J. Virol.Home page
A. Alfadhli, D. Huseby, E. Kapit, D. Colman, and E. Barklis
Human Immunodeficiency Virus Type 1 Matrix Protein Assembles on Membranes as a Hexamer
J. Virol., February 1, 2007; 81(3): 1472 - 1478.
[Abstract] [Full Text] [PDF]

Home page
 J. Virol.Home page
M. R. Auerbach, K. R. Brown, A. Kaplan, D. de Las Nueces, and I. R. Singh
A Small Loop in the Capsid Protein of Moloney Murine Leukemia Virus Controls Assembly of Spherical Cores
J. Virol., March 15, 2006; 80(6): 2884 - 2893.
[Abstract] [Full Text] [PDF]

Home page
 J. Virol.Home page
A. Alfadhli, T. C. Dhenub, A. Still, and E. Barklis
Analysis of Human Immunodeficiency Virus Type 1 Gag Dimerization-Induced Assembly
J. Virol., December 1, 2005; 79(23): 14498 - 14506.
[Abstract] [Full Text] [PDF]

Home page
 J. Virol.Home page
A. Ono, A. A. Waheed, A. Joshi, and E. O. Freed
Association of Human Immunodeficiency Virus Type 1 Gag with Membrane Does Not Require Highly Basic Sequences in the Nucleocapsid: Use of a Novel Gag Multimerization Assay
J. Virol., November 15, 2005; 79(22): 14131 - 14140.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
280/18/17664    most recent
M412325200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Huseby, D.
Right arrow Articles by Barklis, E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Huseby, D.
Right arrow Articles by Barklis, E.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
 All ASBMB Journals   Molecular and Cellular Proteomics 
 Journal of Lipid Research   ASBMB Today 
Copyright © 2005 by the American Society for Biochemistry and Molecular Biology.
Advertisement
spacer
Advertisement
Advertisement