Subtype-Specific Conformational Differences ... - Journal of Virology

0 downloads 0 Views 485KB Size Report
Jul 3, 2007 - quent (1, 4, 7, 8, 13, 15, 41, 54, 75), although disease progres- .... a count is kept of the number of random regroupings of subtype B ... possible pairs among the 22 positions; thus, the significance cutoff for ... subtype B and subtype C V3 sequences was calculated using Fisher's exact test ...... J. 27:379–423.
JOURNAL OF VIROLOGY, Jan. 2008, p. 903–916 0022-538X/08/$08.00⫹0 doi:10.1128/JVI.01444-07 Copyright © 2008, American Society for Microbiology. All Rights Reserved.

Vol. 82, No. 2

Subtype-Specific Conformational Differences within the V3 Region of Subtype B and Subtype C Human Immunodeficiency Virus Type 1 Env Proteins䌤 Milloni B. Patel, Noah G. Hoffman,† and Ronald Swanstrom* Department of Microbiology and Immunology and Center for AIDS Research, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina 27599-7295 Received 3 July 2007/Accepted 31 October 2007

The V3 region of the human immunodeficiency virus type 1 gp120 Env protein is a key domain in Env due to its role in interacting with the coreceptors CCR5 and CXCR4. We examined potential subtype-specific V3 region differences by comparing patterns of amino acid variability and probing for subtype-specific structures using 11 anti-V3 monoclonal antibodies (V3 MAbs). Differences between the subtypes in patterns of variability were most evident in the stem and turn regions of V3 (positions 9 to 24), with the two subtypes being very similar in the base region. The characteristics of the binding of V3 MAbs to Env proteins of the subtype B virus JR-FL and the subtype C virus BR025 suggested three patterns, as each group of MAbs recognized a specific conformation- or sequence-based epitope. Viruses pseudotyped with Env from JR-FL and BR025 were resistant to neutralization by the V3 MAbs, although the replacement of the Env V3 region of the SF162 virus with the JR-FL V3 created a pseudotyped virus that was hypersensitive to neutralization. A single mutation in V3 (H13R) made this chimeric Env selectively resistant to one group of V3 MAbs, consistent with the mAb binding properties. We hypothesize that there are intrinsic differences in V3 conformation between subtype B and subtype C that are localized to the stem and turn regions and that these differences have two important biological consequences: first, subtype B and subtype C V3 regions can have subtype-specific epitopes that will inherently limit antibody cross-reactivity, and second, V3 conformational differences may potentiate the frequent evolution of R5- into X4-tropic variants of subtype B but limit subtype C virus from using the same mechanism to evolve X4-tropic variants as efficiently. changes within V3 are often associated with a change in coreceptor usage (37, 40, 49, 58). During the transmission of HIV-1, the virus that is predominantly transmitted is a CCR5using (R5-tropic) virus (63, 76, 83). Replacements within the V3 region with basic amino acids at V3 positions 11 and 25 (positions 306 and 322, respectively, according to strain HXB2 numbering) are associated with CXCR4 usage (X4 tropism), as are other less well defined changes elsewhere in Env (37, 49, 53). In approximately 50% of subjects, the virus switches to CXCR4 usage, and this coreceptor switch is associated with more rapid progression to AIDS (17, 59, 64, 65). This phenotype is observed in subtype B virus, which is found predominantly in Western Europe and the United States. However, in subtype C virus, which predominates in Asia and sub-Saharan Africa, the switch from R5 tropism to X4 tropism is less frequent (1, 4, 7, 8, 13, 15, 41, 54, 75), although disease progression to AIDS is still evident. Several approaches have been used to identify how V3 interacts with its coreceptor, since a gp120-coreceptor complex has not been crystallized. Most studies suggest that extracellular loop 2 of the chemokine receptor binds to residues in the stem and turn domains of V3, whereas the N terminus of the chemokine receptor binds to residues at the base of V3 and/or other residues in the bridging sheet (18–20, 60). Farzan et al. first identified the N terminus of CCR5 as being responsible for making contacts with the base of V3 (R298) through sulfated tyrosines (24). In the crystal structure of a V3-containing gp120 protein bound to a CD4 molecule, the V3 is in a con-

The binding of the human immunodeficiency virus type 1 (HIV-1) to its target cell is mediated by the viral envelope glycoprotein Env. The glycoprotein is expressed as a gp160 precursor that is proteolytically cleaved into two mature subunits, the gp120 surface protein and the gp41 transmembrane protein. On the surface of the virus, Env is present as a trimer of the gp120 and gp41 heterodimers (2, 62). When the virus binds to its primary receptor, CD4, Env undergoes conformational changes, exposing previously hidden regions and creating new structural elements. Env then binds to a coreceptor, usually the chemokine receptor CCR5 or CXCR4 (23, 26, 80), which triggers further structural changes that promote fusion to the target cell (27, 35). The Env sequence can be viewed as being composed of regions that are relatively constant (C1 to C5) and regions that on a population basis are much more variable (V1 to V5) (70). As the only viral protein expressed on the surface of the virus, Env is the sole target of neutralizing antibodies that supply selective pressure to favor mutated variants capable of evading the immune response (47, 79). The V3 region is associated with coreceptor preference and physically contacts the chemokine receptor as part of the fusion process (23, 26, 80). Sequence * Corresponding author. Mailing address: Lineberger Comprehensive Cancer Center, University of North Carolina at Chapel Hill, CB #7295, Chapel Hill, NC 27599-7295. Phone: (919) 966 5710. Fax: (919) 966 8212. E-mail: [email protected]. † Present address: Department of Laboratory Medicine, University of Washington, Seattle, WA 98195-7110. 䌤 Published ahead of print on 14 November 2007. 903

904

PATEL ET AL.

formation found immediately before binding to the coreceptor (38). The CCR5-using envelope (in JR-FL virus) contains a V3 that may act as a molecular hook, extending away from the core of gp120 to engage in coreceptor interactions (38). The V3 region is a target for antibodies, and for some virus isolates, these antibodies can be neutralizing (30). Although there is evidence of the neutralization of subtype C virus by some subtype B anti-V3 monoclonal antibodies (V3 MAbs), the majority of subtype B V3 MAbs do not share this ability (84). The limited ability of V3 MAbs to neutralize different subtypes may be the result of differences in the structures of V3 between the two subtypes. In this study, we examined potential conformational differences between subtype B and subtype C V3 regions to inform our understanding of the differences in biological properties between these lineages of HIV-1. We used a set of V3 sequences to compare the patterns of amino acid variability between subtypes B and C, and we used recombinant gp120 to compare the binding properties of V3 MAbs. Because we were specifically trying to understand differences between the V3 regions of subtypes B and C, we created chimeric Envs with reciprocal V3 regions and also mutated individual amino acids within V3 that differ between subtypes B and C. These mutated Envs were used in binding studies and in the pseudotyping of HIV-1 to test for neutralization by the V3 MAbs. We demonstrate that there are intrinsic differences in the V3 sequences of subtypes B and C within the stem and turn regions (positions 9 to 24), and we suggest that the conformation of the turn region may underlie the mechanism by which coreceptor interactions and antibody recognition differ between the two subtypes.

MATERIALS AND METHODS Data set. The subtype B and C nucleotide sequences corresponding to the V3 region of gp120 (HXB2 nucleotides 7111 to 7217) were downloaded from the Los Alamos HIV Sequence Database (http://hiv-web.lanl.gov). Additional sequences collected in Malawi (GenBank accession numbers AF153129 to AF153190) were included in the subtype C data set (54). Two methods were used serially to create sets of epidemiologically and phylogenetically unrelated subtype B and C sequences. First, from groups of sequences derived from the same person, a single sequence was randomly selected; multiple representatives of closely linked chains of transmission were treated as originating from the same individual. Sequences for which no such epidemiological information was available were discarded. This selection was accomplished using a database classifying sequences by the patient of origin based mainly on annotations of GenBank references from the Los Alamos HIV Sequence Database (B. Foley, personal communication). Next, a neighbor-joining tree with jackknife analysis was created for each set with phylogenetic analysis using parsimony (PAUP). The general time-reversible model was used for the distance measure, with a gamma distribution describing the rates across sites (other parameters were set to default values). A total of 100 jackknife replicates were performed. A small number of sequences fell on branches supported by at least 50 jackknife replicates: all but a single sequence from each such cluster was discarded. The final data sets resulting from these two selection steps contained 391 and 351 subtype B and C sequences, respectively. The data set is available upon request. One ambiguity to note is amino acid 22 in our subtype B V3 data set. The Los Alamos National Laboratory consensus amino acid for subtype B V3 position 22 is Thr; however, in our data set Ala was present in 53% of sequences and Thr was present in 46%, further demonstrating the polymorphism at this position. Our subtype C V3 data set was the same as the Los Alamos National Library consensus sequence for V3. V3 variability. The variability at each amino acid position in alignments of each of the two sets of sequences was characterized using Shannon entropy (H) (66), which is calculated as follows:

J. VIROL.

冘 m

H共x兲 ⫽ ⫺

P共xi兲logP共xi兲

i⫽1

Here, m is the number of different amino acids represented at position x, and P(xi) is the frequency of amino acid i at position x. The maximum theoretical entropy of ⬇3 at a position would be reached in the event of equal representation of all 20 characters. The minimum entropy of zero is reached for positions with 100% conservation of a single amino acid. The significance of the difference in entropy among the positions in the alignments of subtype B and C V3 sequences was calculated using a permutation test with 10,000 iterations as described in reference 37. Briefly, the permutation test is a Monte Carlo simulation in which a count is kept of the number of random regroupings of subtype B and C sequences that result in an entropy difference equal to or more extreme than the difference resulting from the true classification of sequences. An uncorrected P value is calculated as this count divided by the number of iterations. If no regrouping of sequences resulted in an entropy difference as great as the true difference, a value of P of ⬍0.001 was assigned. We considered positions achieving P of ⬍0.001 as significantly different. At this cutoff, the significance level of the entire analysis (or the probability that at least one of the 35 positions achieved P of ⬍0.001 by chance) is ⬍0.001 ⫻ 35, or less than 0.05 (5). Covariation. The extent of covariation of each possible pair of variable positions in the alignments of subtype B and C V3 sequences was measured using mutual information (also called joint entropy) as described by Hoffman et al. (36). A position was considered variable if at least 5% of either subtype B or C sequences contained a nonconsensus amino acid; 22 positions met this criterion, including positions 2, 5, 10, 11, 12, 13, 14, 15, 16, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 29, 32, and 34. Positions in V3 are numbered according to the 35-amino-acid V3 consensus sequence, with the first cysteine (HXB2 position 296) set at position 1. Mutual information was calculated as described previously (5, 43, 72). The significance of mutual information values was calculated using a permutation test with 105 iterations as described by Hoffman et al. (36). There are 231 possible pairs among the 22 positions; thus, the significance cutoff for the covariation of a single pair was set at 0.05/231, or P of ⬍0.0002. Amino acid composition. The difference in amino acid composition between subtype B and subtype C V3 sequences was calculated using Fisher’s exact test for the frequency of the consensus amino acid versus those of the most common substitutions. The significance cutoff for the difference in amino acid composition was set at 0.05/22 for the 22 variable positions analyzed, or P of ⬍0.002. Plasmids. Molecular clones of env from subtype B and C were obtained from the following. The pSV plasmid containing JR-FL gp160 (21) was kindly provided by Nathaniel Landau. The pCR2.1 plasmid containing BR025 gp160 was obtained through the NIH AIDS Research and Reference Reagent Program (NIH ARRRP; catalog no. 2430) from Beatrice Hahn. The BR025 gp160 gene in the pSV plasmid was kindly provided by Feng Gao. Both the JR-FL and BR025 isolates have been shown to use the CCR5 coreceptor (i.e., to be R5 tropic) and fail to form syncytia in MT-2 cell culture (21, 28). Cloning and expression of mutant and wild-type gp120. env genes containing modifications in V3 were generated by introducing mutations by site-directed mutagenesis using QuikChange (Stratagene, Inc.). A Venezuelan equine encephalitis virus replicon vector (pERK; kindly provided by Alphavax, Inc.) was used to express gp120 with a C-terminal six-His tag (57). The pERK vector contains the portion of the Venezuelan equine encephalitis genome encoding the nonstructural proteins that are responsible for replication. The structural proteins were deleted and replaced by the HIV-1 gp120 protein of choice. The resulting pERK plasmid was linearized by NotI digestion and used as a template for the synthesis of mRNA in vitro. This mRNA was electroporated into mammalian cells, and the gp120 protein was secreted into the culture supernatant. Production of recombinant MBP-V3 fusion peptide. The pc2x plasmid (New England Biolabs) was used to generate the maltose binding protein (MBP)-V3 fusion peptide. The JR-FL and BR025 V3 inserts each encompassed the V3 codons along with five codons upstream and downstream of V3, followed by codons for a six-histidine tag. Both upstream and downstream primers were generated with BamHI restriction sites to clone into the multiple cloning site of pc2x. Inserts were screened using M13 reverse primers to identify V3 inserts in the correct orientation. Purification of gp120 and V3 peptides. Both recombinant gp120 and V3 peptides were purified using nickel beads (Ni-nitrilotriacetic acid superflow; QIAGEN) that bind the six-histidine tags on the C termini of gp120 and the MBP-V3 peptides. The proteins were purified from culture supernatant (gp120) or bacterial cell lysates (V3 peptides) according to the manufacturer’s protocol for purification. Eluted protein was dialyzed into 1⫻ phosphate-buffered saline (PBS) for storage, purity was determined by denaturing gel electrophoresis and

VOL. 82, 2008

DIFFERENCES IN SUBTYPE B AND C V3 CONFORMATIONS

silver staining, and specificity was determined by Western blotting using an anti-His MAb. Purified protein was quantified using the Bradford assay. MAbs. V3 MAbs 19b, F3.9F, 2.1E, 2.10H, CO11, and H211 were kindly provided by James Robinson. All remaining V3 MAbs were obtained through the NIH ARRRP: 447-52D (catalog no. 4030), F425 B4e8 (catalog no. 7626), F425 B4a1 (catalog no. 7625), 257-DI V (catalog no. 1510), and 286-DI V (catalog no. 1511). The following anti-Env antibodies were also obtained through the NIH ARRRP: 2G12 (catalog no. 1476) and 2F5 (catalog no. 1475). ELISAs. Purified protein in PBS was bound to Ni-nitrilotriacetic acid-coated HisSorb enzyme-linked immunosorbent assay (ELISA) plates (QIAGEN) for 2 h at room temperature. The primary V3 MAb was added to duplicate wells for 2 h at room temperature, followed by washing and then incubation with anti-human immunoglobulin G–horseradish peroxidase at a dilution of 1:2,000. For the avidity assays, different concentrations of NaSCN were added prior to the addition of anti-human immunoglobulin G–horseradish peroxidase for 15 min at room temperature. The assay was developed using 5,5⬘-tetramethylbenzidine substrate, and the optical density (OD) was read on a SpectraMAX 340 ELISA reader at 450 nm every 30 s over 15 min and then analyzed using SoftMax Pro software. The slope of binding (OD/time) was linear for approximately 10 min, and data collected after this point were not used. The slope was used to infer binding efficiency. Pseudotyping assay. To generate pseudotyped luciferase reporter viruses capable of a single cycle of replication, we used the pNL4-3.Luc.R-E⫺ (Luc3) plasmid, obtained from the NIH ARRRP (catalog no. 3418). The Luc3 plasmid contains a frameshift within the env gene and has a firefly luciferase gene in place of the nef gene. To generate viral stocks, we cotransfected 293T cells with plasmids expressing the HIV-1 gp160 and the Luc3 plasmid after mixing with Fugene 6 transfection reagent (Roche Applied Science). Supernatants were collected and assessed by p24 ELISA to normalize for infection. Infection and neutralization assays. Pseudotyped single-cycle reporter virus (100 ␮l) was mixed with 100 ␮l of CCR5 MAGI cells (catalog no. 3522; NIH ARRRP) at 5 ⫻ 105/ml for 48 h at 37°C in a 48-well plate. To confirm that luciferase readings were in the linear range, extra wells were infected with dilutions of the virus. Cells were washed with 200 ␮l of 1⫻ PBS and then lysed with 100 ␮l of 1⫻ reporter lysis buffer (Promega). A 50-␮l aliquot of lysate was used to measure luciferase activity with the FLUOstar luminometer, and results were analyzed with the FLUOstar software (BMG LABTECH). To measure neutralization sensitivity, 50 ␮l of virus was preincubated with 50 ␮l of antibody for 1 h at 37°C in a 48-well plate (amphotropic Env-pseudotyped murine leukemia virus was used to control for nonspecific neutralization by the V3 MAbs). The V3 MAbs were used at the following concentrations: 19b at 4.0 ␮g/ml, F3.9F at 4.0 ␮g/ml, F425 B4a1 at 4.0 ␮g/ml, F425 B4e8 at 4.0 ␮g/ml, 447-52D at 1.0 ␮g/ml, 2.1E at 4.0 ␮g/ml, 2.10H at 4.0 ␮g/ml, H211 at 4.0 ␮g/ml, 257-DI V at 1.0 ␮g/ml, 268-DI V at 1.0 ␮g/ml, and CO11 at 4.0 ␮g/ml. Phage display. The Ph.D.-12 phage display peptide library was purchased from New England Biolabs. The library contains a 12-mer library expressed in the minor coat protein, with a complexity of ⬃2 ⫻ 109 sequences. Four rounds of panning were carried out by incubating phage with a plate coated with V3 MAb. The eluted phage was characterized by DNA sequencing.

RESULTS Construction of subtype B and subtype C V3 data sets. We first created a data set to explore sequence variation between the subtype B and subtype C V3 regions at the amino acid level. Sequences containing insertions or deletions (with respect to the 35-amino-acid V3 consensus sequence) and those with more than one point of ambiguity were discarded. All but one representative sequence from a group of identical sequences were removed from the data set. In addition, X4-like subtype B sequences were eliminated from the data set to reduce the effects of variation associated with CXCR4 coreceptor usage by removing all sequences with Arg or Lys at V3 position 11 and/or position 25 (positions 306 and 322 according to HXB2 numbering) (39, 58). These steps resulted in data sets of 391 subtype B V3 sequences and 351 subtype C V3 sequences; the consensus V3 sequence for each subtype is shown in Fig. 1A.

905

Differences in V3 variability between subtype B and C sequences. The subtype B V3 data set and the subtype C V3 data set were used to characterize the patterns of amino acid variability in V3 by using two separate analyses. First, the variability of amino acids in the V3 regions of subtype B and C sequences was characterized using Shannon entropy (Fig. 1A). For convenience, the V3 region can be discussed as containing three distinct domains: the base domain (the short beta sheet adjacent to the cysteines at the start and end of the V3 region, along with the flanking region that is conserved between subtypes B and C [residues 1 to 8 and 25 to 35; see below]), the GPG turn (positions 15 to 17) at the tip of the V3 loop, and the connector or stem domain joining the base and the turn. We examined the concordance of the amounts of variability at each position in subtype B and subtype C V3 sequences (Fig. 1A). The residues within the base of V3 had virtually identical entropy patterns, an indication both of similarity in the selective pressures on this domain in the two subtypes and of the quality of the data sets for this analysis to be able to reveal such strong concordance. In contrast, residues 9 to 24 (stem and turn) in the center of the V3 region had very different intersubtype entropy patterns. We found significant differences in entropy at positions 2, 10, 11, 12, 13, 14, 15, 16, 18, 19, 20, 22, and 24, all but one of which are present in the stem and turn regions from residues 9 to 24 (Fig. 1A). To quantify these differences, we calculated the average per-position difference in entropy between the subtype B and C sequences, comparing the base with the stem and turn domains (Fig. 1B). There was significantly greater per-position difference in entropy in the stem and turn domains of the subtype B and C data sets than in the base domains. Between positions 9 and 24 of both the subtype B and C V3 regions, there is an alternating high-low pattern of variability among adjacent amino acids within two tracts of the sequence (positions 11 to 13 and 18 to 22). These regions of high-low variability immediately flank the V3 GPG turn, and the pattern is shifted by one amino acid position between the subtype B and C sequences. For example, at position 19 of V3, variability is relatively low in subtype B compared to that in subtype C. Conversely, at the adjacent position 20, variability is relatively high in subtype B compared to that in subtype C. The high-low pattern is further evidenced by the significant differences in entropy at positions 11 to 13 and 18 to 20 (Fig. 1A). We propose that the tracts of high-low entropy contained within the stem and turn regions (positions 9 to 24) that are shifted by a single residue position indicate a feature of V3 structure that is divergent between the two subtypes, resulting in two distinct conformations of the V3 region. These results indicate that different regions within V3 vary in the extent of sequence variability, highlighting differences in the nature of the selective pressures, and potentially the conformations, among these different regions. In the second analysis, we considered the composition of the substituting amino acids at the positions where variability is present (Table 1). The most obvious differences are the changes in the consensus amino acid sequences at five positions: in subtype B, H13, R18, A19, T22, and E25, and in subtype C, R13, Q18, T19, A22, and D25 (Fig. 1A). There are other differences in the pattern of amino acid variability in these two subtypes beyond the differences in the consensus

906

PATEL ET AL.

J. VIROL.

FIG. 1. Differences in variability between subtype B and subtype C V3 regions. (A) Entropy differences at each amino acid position in subtype B and subtype C V3 regions. An asterisk indicates significantly different entropies in subtype B and subtype C at this position. The consensus amino acid sequences of the subtype B and subtype C V3 regions are indicated on the right. Dots in the subtype C sequence represent amino acids identical to those in the subtype B sequence. (B) Average difference in entropy per amino acid in regions from positions 1 to 8, 9 to 24, and 25 to 35 of V3. Significance was measured using a two-tailed, unpaired t test. (C) Covarying pairs of amino acids within subtype B and subtype C V3 regions. Arrows indicate similar covarying pairs.

sequences, in which patterns of substitution differ dramatically even at those positions with the same consensus amino acid (adding detail to the entropy differences). In this analysis, we found that 13 individual positions in subtype B and C sequences differed significantly in patterns of substitution (positions 5, 11 to 16, 18 to 20, 22, 24, and 25), with only positions 5 and 25 occurring outside of the stem and turn regions from positions 9 to 24 (Table 1). We interpret the differences between the subtypes as falling into three patterns, reflecting differences in the side chain environment. First, consistent with the entropy differences, for nine positions, the differences between subtype B and C appear to be driven by relative conservation in one subtype but detectable diversity in the other subtype (positions 11 to 16, 18, 20, and 24). However, even with the differences in conservation, positions 13 and 24 also appear to differ in the composition of substituting amino acids. Second, two positions, 5 and 25, display significant diversity in both subtypes, but the compositions of the diverse groups of

amino acids differ. Finally, positions 19 and 22 display reciprocal patterns in the two subtypes. Since the side chains for these two positions are near each other in the subtype B gp120 structure (38), this finding suggests that these positions may be flipped but are otherwise serving analogous roles in the two subtypes. Overall, the majority of differences in polymorphism are contained in the stem and turn regions, again suggesting that structural differences in the V3 region between subtypes B and C are located here while more minor differences exist in the base residues 1 to 8 and 25 to 35. Differences in covarying pairs within V3 by subtype. The same subtype B and C V3 data sets were used to determine if covarying pairs were present within the V3 region (where covariation is defined as the likelihood of one amino acid in V3 being substituted at a higher-than-expected frequency in the presence of a second substitution). Covarying pairs were measured by mutual information (Table 2) (36). There are more covarying pairs in subtype B V3 than in subtype C V3 (Fig. 1C),

VOL. 82, 2008

DIFFERENCES IN SUBTYPE B AND C V3 CONFORMATIONS

907

TABLE 1. Percentages of sequences with the indicated amino acids present at each V3 position in subtypes B and Ca V3 position

% of sequences with amino acid in subtype:

Consensus amino acid(s)b

1 2 3 4 5* 6 7 8 9 10 11* 12* 13* 14* 15* 16* 17 18* 19* 20* 21 22* 23 24* 25* 26 27 28 29 30 31 32 33 34 35

C T R P N N N T R K S I H, R I G P G R, Q A, T F Y T, A T G E, D I I G D I R Q A H C

B

C

⬎99 84 ⬎99 96 77 ⬎97 ⬎99 ⬎99 ⬎98 83 78 97 63 (H) 68 95 88 98 73 (R) 90 (A) 75 90 46 (T) ⬎95 97 38 (E) 95 90 ⬎99 85 ⬎98 ⬎98 88 ⬎99 87 ⬎99

⬎99 77 ⬎99 99 64 ⬎97 ⬎99 ⬎99 ⬎98 73 96 54 94 (R) 99 99 100 100 99 (Q) 68 (T) 97 88 98 (A) ⬎95 79 54 (D) 95 92 ⬎99 79 ⬎98 ⬎98 83 ⬎99 83 ⬎99

Alternative amino acid(s) in sequences (% of sequences with amino acid) in subtype: B

C

I (11)

I (10), V (5), A (4)

L (2), H (⬍1), L (⬍1), G (⬍1) S (11), G (7), H (3)

R (⬍1), T (⬍1) G (19), S (11)

R (16) G (21) V (3) P (16), N (11), S (4), T (3) M (15), L (14), V (2) A (4) W (5), L (2) R (⬍1), E (⬍1), W (⬍1), V (⬍1) K (11), S (6), Q (6), G (3) T (6), V (2) W (11), L (8), I (4) F (6), H (3) A (53)

R (11), E (8), Q (4), T (3) G (3) V (29), M (15) G (5), K (⬍1), T (⬍1)

E (2) D (30), Q (14), A (7), G (5), N (3) V (4) T (4), V (4)

N (13), D (3), K (2) E (17), G (14), A (7), N (3) V (5) T (4), V (3)

N (13)

N (20)

K (7), R (4), E (1)

K (6), E (5), R (3)

Y (12)

Y (15)

R (⬍1) A (28), V (2) L (1) F (11) T (1)

a Positions correspond to amino acids 296 to 331 of the HXB2 V3 region. Asterisks indicate positions that have significantly different patterns of substitution according to Fisher’s exact test (P ⬍ 0.002). All numbers are rounded to the nearest percentage. b Where two amino acids are listed, the one on the left is the subtype B amino acid and the one on the right is the subtype C amino acid.

TABLE 2. Mutual information scores of each covariation pairing Subtype

Position 1

Position 2

Mutual information score

P valuea

B

5 11 13 14 14 14 14 18 19 20 22 24

27 13 18 16 18 20 25 20 20 25 25 27

0.067 0.078 0.096 0.108 0.093 0.158 0.120 0.091 0.063 0.171 0.077 0.047

0.00001 * * * * * * 0.00001 0.00009 * * *

C

10 12 12

12 19 24

0.093 0.086 0.079

0.00001 0.00001 0.00001

a

Asterisks denote P values of ⬍0.00001.

which is consistent with the higher overall entropy in subtype B V3 than in subtype C V3. All of the covarying pairs in subtype C V3 appear to correspond to a pair in subtype B V3 that is shifted by one amino acid position (Fig. 1C). For example, in subtype B V3, amino acids 13 and 18 form a covarying pair, and in subtype C V3, amino acids 12 and 19 form a corresponding pairing. Similarly, an 11-13 pair in subtype B mirrors a 10-12 pair in subtype C. This apparent shift in the positions of covarying pairs reinforces the idea of intersubtype differences and is consistent with the shift in alternating entropy values. The chemical interactions implied by these covarying pairs (either directly or indirectly) in the two subtypes are very different. This finding suggests either that the potential for variability is creating apparent linkage between these positions (but with restricted specificity) or that these chemically disparate pairs are having an indirect effect on structure rather than recreating compatible chemical interactions. Differences in the binding of V3 MAbs to subtype B and C V3 protein sequences. We next tested the hypothesis that conformation, driven by the sequence differences, differs between subtype B and C V3 regions. We assembled a panel of 11 V3

908

PATEL ET AL.

J. VIROL. TABLE 3. Previously characterized epitopes of V3 MAbs

Group

1

2

3 a b

No. of phage displaydetermined epitope sequences/total no. of phage display sequences

V3 MAb

Previously reported epitopea

Epitope determined by phage displayb

19b 257-DI V F3.9F F425 B4a1 F425 B4e8

Nonlinear epitope -I----G--FY-T KRIHI Unknown Base of V3 loop I-----R-F

IH-GP--A K-I--GP ND ND (F/L)H(K/Q)P

5/20

447-52D 2.1E 2.10H H211

Crown of V3 (GP-R) Unknown Unknown Unknown

ND PGR(A/T)F H-GPGR GR-FY

2/10 1/10 3/20

268-DI V CO11

HIGPGR Unknown

(F/L)H(K/Q)P H-GPGR

1/10 2/20

7/20 11/20

These epitopes are identified based on previously published data (10, 11, 33, 34, 50, 52, 68, 84). ND, no data were generated for this V3 MAb epitope by phage display.

MAbs that had been isolated from patients naturally infected with subtype B HIV-1 (Table 3) and tested them for binding to the subtype B and subtype C V3 protein sequences to assess differences in binding. In our initial assessment of the V3 MAbs, we measured binding to subtype B and C V3 sequences in two contexts: recombinant gp120 containing the V3 sequence, and a V3 peptide. The JR-FL Env was used to represent subtype B as it contains the consensus V3 amino acid sequence for a subtype B R5-tropic virus. To represent subtype C, we used the BR025 Env, which contains three amino acids changes compared to the JR-FL V3 region: H13R, R18Q, and T22A (this Env protein does not contain the subtype C consensus amino acids at the polymorphic positions T19 and D25). Using a V3 MAb concentration that saturated the subtype B gp120 protein or V3 peptide, we determined the total amount of each V3 MAb that was able to bind to the same amount of the subtype C gp120 protein or V3 peptide. As shown in Fig. 2A, all of the V3 MAbs bound to JR-FL gp120; however, when binding to BR025 gp120 occurred (with 4 of 11 MAbs), it was decreased 40 to 55% compared to binding to JR-FL gp120. These results suggest either that all of the antibodies recognize epitopes that are disrupted by the amino acid differences between subtype B and C V3 sequences or that these amino acid differences affect the overall conformation of V3 in such a way as to impact the binding of all of the V3 MAbs. In an effort to determine the effect of the backbone Env structure on the binding of MAbs to the subtype B and C V3 sequences, we created V3 peptides as a heterologous fusion protein to use in the same binding assay. As shown in Fig. 2B, all of the V3 MAbs were able to bind to the JR-FL V3 peptide and 9 of the 11 V3 MAbs bound to the BR025 V3 peptide. This result suggests that these antibodies can bind the sequence in the BR025 Env protein but that V3 in gp120 is conformationally constrained relative to V3 in the peptide and in a way that creates a distinction from the subtype B Env protein. The remaining two antibodies (268-DI V and CO11) did not bind to either BR025 gp120 or the BR025 V3 peptide, suggesting that for these V3 MAbs, the sequence differences in the subtype B and subtype C V3 regions disrupt the epitope. In comparing the patterns of binding to the gp120 proteins

and the V3 peptides, we were able to place the MAbs into three groups: group 1 MAbs were able to bind both the subtype B and C gp120 proteins and peptides, group 2 MAbs could bind subtype B but not subtype C gp120 and were able to bind both V3 peptides, and group 3 MAbs were able to bind only to the subtype B gp120 protein and peptide. It should be noted that these groupings do not necessarily define groups of distinct epitopes since even two antibodies binding the same region may favor interactions with different subsets of amino acids for the binding. We further investigated the interactions of V3 MAbs with gp120 and V3 peptides by measuring avidity by antibody elution with the chaotropic agent sodium thiocyanate (NaSCN) (Fig. 2D to E). We used six MAbs (two from each group) to test avidity and found various degrees even within each group. However, the patterns of antibody avidity were similar for both gp120 and the V3 peptide of JR-FL, with slightly tighter binding (i.e., a higher NaSCN elution concentration) consistently seen in the context of binding to gp120 than to the V3 peptide (Fig. 2D). Binding to the subtype C BR025 V3 sequence did reveal significantly different binding avidities for these antibodies (Fig. 2E). The group 1 antibody 19b bound to BR025 gp120 with low avidity, similar to that of its binding to subtype B gp120. However, the higher-avidity binding to subtype B gp120 seen with the other group 1 antibody, F3.9F, was lost with BR025 gp120, as this MAb showed much lower avidity in binding to this subtype C gp120. These two antibodies behaved similarly in the context of binding to the V3 peptides. A similar pattern of binding was seen with the two group 2 antibodies in binding to the V3 peptides: the antibody with low avidity for the subtype B sequence (2.1E) had similarly low avidity for the subtype C sequence, and the antibody that bound with high avidity to the subtype B sequence (447-52D) bound with low avidity to the subtype C V3 sequence. Thus, the detected binding to the subtype C V3 sequence was of lower avidity, although the detection of binding by the group 2 antibodies to the subtype C V3 peptides was in the context of a trend toward lower avidity of binding to the peptides than to gp120. Binding of chimeric Env to V3 MAbs. We next used gp120 protein but now made chimeras with respect to the V3 se-

VOL. 82, 2008

DIFFERENCES IN SUBTYPE B AND C V3 CONFORMATIONS

909

FIG. 2. Differences in binding of V3 MAbs to subtype B and C V3 protein sequences. (A) Binding to wild-type JR-FL and BR025 gp120. (B) Binding to wild-type JR-FL and BR025 V3 peptides. (C) Binding to chimeric JR-FL and BR025 gp120. The slope value indicates the ratio of the OD to the time (measured at 30-s intervals); thus, a slope of 0.1 indicates an OD of 0.4 at the 2-min time point (the background value is 0.03 ⫾ 0.009). There are no data available for V3 MAb H211. (D) Binding to wild-type JR-FL and BR025 gp120 in the presence of NaSCN. (E) Binding to wild-type JR-FL and BR025 V3 peptides in the presence of NaSCN. Group 1 MAbs are represented by squares, group 2 MAbs by triangles, and group 3 MAbs by circles. JR-FL symbols are closed and BR025 symbols are open; the black and gray identify MAbs in the same group binding to the same subtype.

quence to assess the contribution of the Env backbone to the V3 conformation. Chimeric Env proteins were created with BR025 V3 in the JR-FL backbone (JR-FL-V3BR025) and the reverse (BR025-V3JR-FL), and these gp120 proteins were analyzed for MAb binding. We again found distinct patterns of binding to the wild-type and chimeric Env proteins (Fig. 2C). The group 1 MAbs were able to bind both chimeric gp120 proteins, like the wild-type subtype B and C gp120 proteins, while the group 2 and 3 MAbs were able to bind only the chimeric protein with the subtype B V3 sequence. We conclude that the group 1 MAbs bind to V3 in a way that is not affected by the gp120 backbone but that the group 2 MAbs bind to a structure created by the subtype B V3 sequence that

is not impacted by the gp120 backbone and the group 3 MAbs bind an epitope that is defined by the subtype B V3 sequence. We next determined the impact of individual substitutions in V3 on MAb binding. JR-FL and BR025 V3 sequences contain three amino acid differences (H13, R18, and T22 in JR-FL and R13, Q18, and A22 in BR025). Positions 13 and 18 were singly and doubly mutated in the two gp120 proteins and the two V3 peptides, resulting in a panel of four JR-FL Envs (Fig. 3A). The binding of the MAbs to these mutant proteins divided the MAbs into the same three groups described above, with a representative member of each group shown in Fig. 3. The group 1 antibodies displayed a binding pattern that indicates that the subtype B backbone can accommodate changes in the

910

PATEL ET AL.

J. VIROL.

FIG. 3. Binding of anti-V3 MAbs with the progression of V3 from the JR-FL sequence to the BR025 sequence and the reverse. (A) Sequences of the mutant V3 gp120s generated. Residues at key positions are in bold. (B) Binding of JR-FL gp120 with the progression of V3 from the JR-FL sequence to the BR025 sequence. (C) Binding of the JR-FL V3 peptide with the progression of V3 from the JR-FL sequence to the BR025 sequence. (D) Binding of BR025 gp120 with the progression of V3 from the BR025 sequence to the JR-FL sequence. (E) Binding of the BR025 V3 peptide with the progression of V3 from the BR025 sequence to the JR-FL sequence. WT, wild type.

V3 region at position 13 (H13R) and/or position 18 (R18Q) (Fig. 3B). Binding to the peptides with one or both of these mutations showed no differences from binding to the wild-type sequence (Fig. 3C). Group 2 antibodies required an R18 residue in order to bind JR-FL gp120, but binding was not significantly affected by the H13R change (Fig. 3B). Similarly, binding to the V3 peptides was reduced for those peptides carrying the R18Q mutation (Fig. 3C). Finally, group 3 antibodies (268-DI V and CO11) could bind only the wild-type JR-FL gp120 protein and peptide (Fig. 3B and C), demonstrating again that H13 and R18 are both required for binding to this group of antibodies. Subtype C V3 mutant peptides also defined three distinct patterns of MAb binding. We repeated the above-described experiment using mutant forms of the subtype C gp120 protein and V3 peptide (Fig. 3A). For group 1 MAbs, the H13 residue enhanced binding to Env but the R18 mutation in the subtype C backbone reduced binding regardless of which amino acid was present at position 13 (Fig. 3D); however, none of these mutations affected binding to the V3 peptides (Fig. 3E). This

finding suggests that R18 has a specific interaction in the subtype C backbone that affects the conformation of the V3 region. For group 2 MAbs, the Q18R mutation did not restore significant binding but the R13H-Q18R double mutation did allow binding (Fig. 3D), again suggesting a specific interaction of R18 with the BR025 backbone that could be restored with H13. Peptide binding was affected only by the R13H mutation (Fig. 3E), supporting the idea that the poor binding of the Q18R mutant form of the subtype C gp120 protein had a structural basis. Finally, the group 3 antibodies behaved in the same way toward both the subtype B and the subtype C peptides and gp120 proteins; both H13 and R18 were required for binding to this group of antibodies, likely defining key aspects of the epitope (Fig. 3D and E). Avidity experiments with the pairs of antibodies from the different groups were also carried out using V3 peptides with individual mutations at positions 13 and 18 (data not shown). The binding patterns for these pairs of antibodies were complex, showing increases or decreases in avidity that appeared to be more antibody specific than group specific.

VOL. 82, 2008

DIFFERENCES IN SUBTYPE B AND C V3 CONFORMATIONS

911

FIG. 4. Neutralization of subtype B Envs by various MAbs; the same antibody concentration was used to neutralize each of the Envs. The wild-type JR-FL Env was neutralization insensitive compared to the wild-type SF162 Env (SF162 WT). When the JR-FL V3 was inserted into SF162 Env, this Env was neutralization hypersensitive. The addition of the mutation H13R to the hypersensitive SF162-V3JR-FL Env altered sensitivity to group 3 antibodies. There are no data available for the neutralization of JR-FL by V3 MAb 257-DI V.

V3 MAb epitope mapping. The patterns of V3 MAb binding to the chimeric subtype B and subtype C Env proteins suggest that antibody interactions with V3 can occur in distinct ways. Limited epitope data are available for the V3 MAbs used in this study (Table 3). Therefore, we used a phage display system to identify elements of the epitopes for the V3 MAbs. An M13 phage display library with 12-mer inserts (complexity of 2.7 ⫻ 109 sequences) was used in a panning strategy with a subset of the MAbs. After four rounds of panning, phage DNA was isolated and sequenced. For the most part, the selected phage did not give homogeneous sequences or even highly enriched V3 homologs; however, V3-like sequences were detected in the selected phage pools, and these are presented in Table 3. Sequences selected by the three group 1 antibodies all included a Pro residue and sequences in addition to the N terminus that were consistent with the H13 region’s being an important part of the epitope; this result was similar to the placing of the F425 B4e8 epitope in the crown region by scanning mutagenesis (52). The three group 2 antibodies all selected sequences that included Gly-Pro, similar to the region bound as seen in the structure of 447-52D (68) and consistent with R18’s playing a critical role in the binding of the group 2 MAbs. Finally, the two group 3 MAbs also selected sequences similar to the V3 turn that included a His followed by turn region-like amino acids, again consistent with H13 and R18’s being critical elements of their epitopes. Neutralization of subtype B virus by each group of antibodies. Using wild-type virus, we measured the abilities of all of the V3 MAbs to neutralize virus pseudotyped with the JR-FL Env protein at high antibody concentrations. Even though all the V3 MAbs were able to bind to JR-FL V3 (as V3 peptides or JR-FL gp120), we found that the JR-FL Env pseudotype

was resistant to neutralization by all of the V3 MAbs tested (Fig. 4). JR-FL is known to be a relatively difficult virus to neutralize (12, 31, 77); therefore, we also tested V3 MAb neutralization of SF162, a subtype B virus that is sensitive to neutralization (45). As demonstrated in Fig. 4, 6 of the 10 V3 MAbs tested were able to neutralize SF162 at least 50% at MAb concentrations that failed to neutralize JR-FL. In an attempt to account for the three amino acid differences between the SF162 and JR-FL V3 regions (H13T, T22A, and E25D), we created a chimeric SF162 variant with a JR-FL-like V3 sequence (SF162-V3JR-FL) to further differentiate the neutralization abilities of the V3 MAbs. Virus pseudotyped with the chimeric SF162-V3JR-FL Env protein was hypersensitive to neutralization by all of the V3 MAbs. One explanation for the apparent hypersensitivity of SF162-V3JR-FL is that, in the context of Env, the V3 region of SF162 Env is more accessible than that of JR-FL Env and the JR-FL V3 sequence is preferred for binding, making the SF162-V3JR-FL Env chimera especially sensitive to neutralization. The differential neutralizing activities of 19b, 2.10H, and the two group 3 MAbs against the SF162 and SF162-V3JR-FL Env proteins suggest an important role for one or more of the three V3 amino acids that differ between these proteins in defining the epitope and/or structure. We confirmed these differential sensitivities to neutralization by determining 50% inhibitory concentrations of six of the MAbs (two from each group) for JR-FL, SF162, and SF162-V3JR-FL; the hypersensitivity of SF162-V3JR-FL was seen, as was the resistance of JR-FL and the intermediate sensitivity of SF162 (data not shown). The hypersensitivity of SF162-V3JR-FL allowed us to compare the neutralization sensitivity of an infectious mutant, the H13R variant (Fig. 4). We found that the two group 3 anti-

912

PATEL ET AL.

J. VIROL.

bodies were completely dependent on H13 and that the mutant virus was no longer sensitive to neutralization in the presence of R13, the same pattern observed for the binding of group 3 antibodies to either the gp120 proteins or the V3 peptides (Fig. 3). In contrast, the other MAbs retained their neutralizing activity against this mutant, again consistent with the binding data (Fig. 3). DISCUSSION The V3 loop plays a central role in the life cycle of HIV-1 but, like other regions of the genome, displays subtype-specific sequence variation (42, 44, 51). We are interested in the possibility that these sequence differences can impact virus biology in a subtype-specific way. Using the analysis of sequence variability and the binding of V3-specific MAbs, we have found evidence for subtype-specific differences in the V3 region of subtype B and subtype C HIV-1 Env proteins. We believe these differences contribute to differences in the biology of subtype B and subtype C HIV-1 and that these differences need to be understood in greater detail, as this understanding will contribute to mechanistic insights into the role of the V3 region in virus replication and the host immune response. The V3 region of Env is responsible for interacting with the coreceptor at the initial stages of infection (23, 26, 80). This vital contact with an invariable host protein suggests that the sequence of V3 must be constrained in order to preserve the ability to interact. Consistent with this expectation, the V3 regions of the different HIV-1 subtypes are relatively similar, although subtype-specific differences exist (44, 68). The V3 base. Mutagenesis across the V3 region has been used to define functional constraints (3, 19, 22, 24, 25, 71, 73). Ala substitutions that affect binding to a peptide representing the N terminus of CCR5 showed that the positions critical for binding included positions near the Cys residues (up to position 8 on one side and starting at position 26 on the other side) (18). The CD4-induced class of antibodies as a group tend to be negatively charged, and these antibodies are thought to mimic the same N-terminal domain of CCR5 in binding to the bridging sheet of gp120 (60, 61, 81). These observations have led to a model in which part of the interaction with CCR5 involves the interaction of the N terminus of CCR5, negatively charged with sulfated tyrosines, with the bridging sheet and the base region of V3 (24, 60, 61). The expectation of sequence conservation is also consistent when considering the eight residues proximal to the terminal Cys residues in the base region; they are essentially identical in amino acid composition and entropy in subtype B and subtype C V3 regions (Fig. 1). This conservation can be seen over greater phylogenetic distances, as only one position on each side of the base is chemically distinct in HIV-1 and simian immunodeficiency virus SIVcpz. Thus, both the sequence conservation and the similarity in entropy between subtype B and subtype C V3 sequences in this base region (positions 1 to 8 and 25 to 35) are consistent with a V3 interaction with an invariable cellular protein. Entropy, covariation, and structure. We considered the relationship between entropy and side chain orientation (using the V3 region of the JR-FL V3 structure [38]), classifying residues as “up” (positions 2, 9, 13, 18, 20, 21, 26, 32, and 34), “down” (positions 3, 6, 12, 14, 19, 22, 30, 31, and 33), or located

in the plane of the V3 main chain (“flat”; positions 5, 7, 8, 10, 11, 23, 25, 27, and 29) (excluding Cys, Pro, and Gly residues). At 15 of 18 positions, there was concordance between residues either pointing up from the plane of V3 and having higher entropy or pointing down from the plane and having lower entropy. The exceptions were positions 21, 22 (an Ala-Thr mixture), and 26. In the plane of V3, eight residues point away from the main chain, and six of these have high entropy. Thus, the higher-entropy positions to a large extent make up one face and the sides of V3. Other evidence pointing to the importance of residues present on the faces of V3 comes from a study demonstrating the ability of another V3 MAb, 2219, to bind to three different V3 peptides in two distinct conformations, revealing binding that is dependent on the sets of residues present on the face of V3 (69). This study also showed that the base of the V3 region is conserved among the three peptides and that the dynamic portion of V3 is the flexible stem and turn regions (69). Given that the patterns of covariation differ between subtype B and subtype C V3 sequences, it is not surprising that the positions that show covariation are among the most variable and discordant positions, which lie in the stem and turn domains (Fig. 1; Table 2). Subtype B V3 has more covarying pairs than subtype C V3; however, the subtype C pairs appear to correspond to pairs in subtype B that are shifted by one amino acid position. This finding is similar to data presented in several other studies (5, 29, 43) that determined that covarying pairs in subtype B V3 did not generally covary in subtype C V3 and that there are fewer overall covarying pairs in subtype C V3. However, only one covarying pair overlapped between our study and the others (positions 19 and 20 in subtype B V3) (29). Other studies on variability within the V3 region have found that there is considerably less variability in subtype C V3 than in subtype B V3 (29, 54), also suggesting alternate structural characteristics of the two subtypes. A major difference in the previous studies of covariation was the inclusion of V3 sequences that used CCR5 and/or CXCR4 as the coreceptor. We have previously shown that much of V3 variability in subtype B is associated with sequences that have features indicative of X4 tropism (37, 49, 54, 58), prompting us to attempt to exclude X4 sequences from the present data set. Thus, at least some of the reported covariation may reflect the coreceptor switch or other methodological differences among the studies. Although we excluded X4 sequences, not all viruses capable of using X4 have sequence changes in V3 (39), raising the possibility that some of the excess variability in the subtype B data set may still reflect the presence of X4 viruses. Differences between subtype B and subtype C. The consensus sequence identity, the similarity in entropy patterns, and the extensive similarity in the amino acid substitution patterns of residues 1 to 8 and 25 to 35 in subtype B and subtype C sequences are in striking contrast to the values of these same measures for residues 9 to 24. Only positions G15 and G17 in the GPG turn, Y21, and T23 approximate the level of similarity between the subtype B and subtype C sequences that is seen in the base region. In the current model of V3 binding to CCR5, the stem and turn domains of V3 interact with the extracellular loops of CCR5 (24, 60, 61). Perhaps the strongest experimental evidence for this interaction comes from the occurrence of mutations affecting the coreceptor switch from CCR5 to

VOL. 82, 2008

DIFFERENCES IN SUBTYPE B AND C V3 CONFORMATIONS

CXCR4 within the stem and turn domains (20, 22, 56, 60, 78) and the appearance of mutations selectively in the stem domain that confer resistance to CCR5 binding inhibitors of HIV-1 that enhance affinity for CCR5 (46, 82). If subtype B and subtype C V3 sequences interact with the CCR5 extracellular loops in equivalent ways, then the GPG turn residues and positions Y21 and T23 may represent a conserved motif for this interaction. However, if other residues are involved, then the functionally equivalent interactions with CCR5 likely occur through different molecular interactions or at least start with conformationally distinct V3 regions. The idea that these sequence differences give rise to conformational differences is supported by the patterns of binding of 11 V3 MAbs used in this study, which fell into three distinctive groups. All of the V3 MAbs were able to bind both gp120 and the V3 peptide of JR-FL (as an example of subtype B); in addition, the group 1 and 2 antibodies were able to bind to the V3 peptide of BR025 (as an example of subtype C), indicating that for these two groups of MAbs, the sequence changes in V3 by themselves did not restrict the ability of the MAb to bind, although binding did generally occur with lower avidity. In contrast, the group 3 MAbs required H13 and R18 as both necessary and sufficient for binding to either gp120 or peptide forms of the V3 region in either the BR025 or the JR-FL background. All the group 1 MAbs bound JR-FL gp120, although the binding of some V3 MAbs to JR-FL-V3BR025 was reduced. In the same manner, only group 1 MAbs were able to bind to BR025 gp120 but both group 1 and 2 V3 MAbs were able to bind to BR025 V3 peptides. These results suggest that conformational constraints in gp120 prevent group 2 antibodies from binding to V3 at a detectable level, although the V3 sequence in the peptide form can be bound. The progression of V3 from a subtype B-like sequence to a subtype C-like sequence showed that group 1 MAbs are promiscuous in their ability to bind to gp120, suggesting that restrictions to binding are conformation based for this group. Consistent with the known 447-52D epitope, the other group 2 MAbs were very sensitive to the presence of an Arg at position 18. Finally, the group 3 MAbs were similar to the group 2 MAbs in being sensitive to a position 18 Arg but were also similar to the group 1 MAbs, which in at least some circumstances were sensitive to the amino acid at position 13. Thus, the three groups of MAbs appear to be dependent on different combinations of amino acids in and around the V3 stem and turn. We conclude that most V3 MAbs are likely targeted to the distal turn region of V3, which is responsible for making contact with the extracellular loops of the coreceptor. Our data demonstrate that different V3 MAbs are sensitive to different patterns of sequence change within the turn region. While such groupings are most easily explained by assuming a single epitope for a group, we cannot draw this conclusion from our data since subtle differences in antibody recognition and avidity likely play significant roles that our data do not fully explore. The five differences in the consensus sequences of the two subtypes occur at positions 13, 18, 19, 22, and 25. Positions 13 and 18 also differ in subtypes B and C in that they are variable in subtype B and conserved in subtype C. For positions 19, 22, and 25, composition is less variable: both positions 19 and 22 are polymorphic for Ala and Thr in subtypes B and C, and position 25 is composed of Asp or Glu in ⬃70% of se-

913

quences in both subtypes. Therefore, we argue that the substitution composition patterns indicate that positions 13 and 18 are more likely than the other positions to be responsible for defining structure in the turns of both subtypes B and C. Epitope accessibility and antibody binding. Neutralization is dependent on both epitope recognition and the accessibility of an antibody to the epitope. Differences in the neutralization sensitivities of primary isolates and lab-adapted strains have been well documented (6), in particular for both CD4-induced epitopes and V3 antibodies. Our use of JR-FL and SF162V3JR-FL with the H13R mutation demonstrated an extreme example of this phenomenon. All of the V3 MAbs were able to bind to JR-FL Env but were not able to neutralize JR-FL Env. The insertion of the JR-FL V3 into the SF162 Env backbone generated a virus that was hypersensitive to neutralization. Pinter et al. have shown that the JR-FL V1-V2 domain is at least one determinant of the occlusion of the V3 loop (55). It is also possible that the V1-V2 region of subtype C can obscure V3; this would be an alternative explanation of the inability of group 2 V3 MAbs to bind to BR025 gp120, even though they can bind BR025 V3 peptides. We were also able to reproduce the epitope sequence required for gp120 and peptide binding by the group 3 antibodies by showing that the H13R mutation in the context of the chimeric SF162-V3JR-FL pseudotyped virus conferred resistance to neutralization even at high antibody concentrations. Role of V3 conformation in coreceptor switching. Sharon et al. previously suggested that the V3 region adopts two alternate conformations (67), depending, for subtype B HIV-1, on whether the virus is R5 tropic or X4 tropic. However, a number of studies have reported that X4-tropic viruses are much less frequent in people infected with subtype C HIV-1 (1, 4, 7, 8, 13, 15, 41, 54, 75). We hypothesize that the V3 region in a subtype B R5-tropic virus is less constrained than that in a subtype C virus and is able to accommodate mutations that allow X4 tropism. Conversely, the subtype C virus V3 region is conformationally restricted, limiting the potential for evolution to usage of the CXCR4 coreceptor. This suggestion is supported in part by results from studies of X4-tropic subtype C virus that suggest that the appearance of Q18R (introducing a subtype B-like amino acid) is associated with the emergence of X4-tropic virus variants (16). Three other sequence differences are noteworthy in comparing subtypes B and C with regard to the coreceptor switch. First, a change at position H13, in some cases H13R, has been implicated as a contributing determinant of CXCR4 usage (14, 49); however, R13 is the consensus amino acid for subtype C as an R5 virus. Second, Env position R440, which is in the C4 domain and lies under the V3 region in the gp120 structure, frequently evolves into a polar or acidic side chain in subtype B X4 viruses (9, 37, 48, 49); in subtype C HIV-1, position 440 has Glu as the consensus amino acid. Finally, the consensus amino acid at V3 position 18 is Gln in subtypes A, C, D, F, and G and various circulating recombinant forms. One plausible explanation for these differences is that an early subtype C-like X4 variant (with R13, R18, and E440) was a progenitor of the subtype B lineage and involved a phenotypic reversion to CCR5 tropism, in part through changes at positions 13 and 440. Vaccine implications. We examined the binding of 11 human MAbs; all 11 appeared to bind to the distal turn domain

914

PATEL ET AL.

J. VIROL.

based on the sensitivity of variants with mutations in the stem and turn domains and on (incomplete) phage display selection (Table 3). While it is possible that other antibodies will be identified that bind to the conserved base region, our data would suggest that this is not a region of the protein that is readily immunogenic in the context of gp120, at least in comparison to the turn domain. Data that may help predict if the generation of anti-V3 neutralizing antibodies by a vaccine would be effective are still incomplete. It is uncertain if a cocktail of V3 MAbs used as prophylaxis would be beneficial in the face of an infection. Therapeutic administration of V3 MAbs (as was done with the neutralizing antibodies 2F5, 4E10, and 2G12 [74]) would be helpful to assess their potential to impact an infection. Antibodies raised against virus containing the GPGR turn sequence are also less likely to cross-react within a panel of viruses containing the GPGR or GPGQ turn sequence (44%) than antibodies raised against a virus containing the GPGQ turn sequence (71%) (32). There are a few V3 MAbs raised against subtype B that are cross-reactive with subtype C virus containing the GPGQ turn sequence, with some potency (30, 32). However, this cross-reactivity is the exception rather than the rule, as numerous other studies have shown that the GPGR turn sequence is required for subtype C virus neutralization by V3 MAbs raised against subtype B virus (6, 16, 68). To date, the only broadly neutralizing V3 MAb characterized requires a GPXR motif at the turn region and does not bind or neutralize subtype C virus (which usually contains a GPGQ turn sequence) (68, 84). The differences in antibody binding that we observed in this study suggest that the design of V3-based immunogens would benefit from a consideration of subtypespecific sequence differences; however, the occlusion of V3 in many primary isolates must be acknowledged as representing a significant limitation for this target.

7.

8.

9. 10. 11. 12.

13.

14.

15.

16. 17. 18. 19.

ACKNOWLEDGMENTS We thank LiHua Ping, Mark C. Johnson, and Adrienne Swanstrom for assistance in various aspects of this work. We also thank James Robinson for providing the anti-V3 MAbs for use in this study and Marcey Waters, Sharon Campbell, and Peter Kwong for helpful discussion. This research was supported in part by NIH grant R37-AI44667 and NIH training grants T32-AI07419 and T32-AI07001 to M.B.P. and N.G.H. and was performed under the auspices of the UNC Center for AIDS Research, supported by NIH grant P30-A150410.

20.

REFERENCES

23.

1. Abebe, A., D. Demissie, J. Goudsmit, M. Brouwer, C. L. Kuiken, G. Pollakis, H. Schuitemaker, A. L. Fontanet, and T. F. Rinke de Wit. 1999. HIV-1 subtype C syncytium- and non-syncytium-inducing phenotypes and coreceptor usage among Ethiopian patients with AIDS. AIDS 13:1305–1311. 2. Allan, J. S., J. E. Coligan, F. Barin, M. F. McLane, J. G. Sodroski, C. A. Rosen, W. A. Haseltine, T. H. Lee, and M. Essex. 1985. Major glycoprotein antigens that induce antibodies in AIDS patients are encoded by HTLV-III. Science 228:1091–1094. 3. Basmaciogullari, S., G. J. Babcock, D. Van Ryk, W. Wojtowicz, and J. Sodroski. 2002. Identification of conserved and variable structures in the human immunodeficiency virus gp120 glycoprotein of importance for CXCR4 binding. J. Virol. 76:10791–10800. 4. Batra, M., P. C. Tien, R. W. Shafer, C. H. Contag, and D. A. Katzenstein. 2000. HIV type 1 envelope subtype C sequences from recent seroconverters in Zimbabwe. AIDS Res. Hum. Retrovir. 16:973–979. 5. Bickel, P. J., P. C. Cosman, R. A. Olshen, P. C. Spector, A. G. Rodrigo, and J. I. Mullins. 1996. Covariability of V3 loop amino acids. AIDS Res. Hum. Retrovir. 12:1401–1411. 6. Binley, J. M., T. Wrin, B. Korber, M. B. Zwick, M. Wang, C. Chappey, G.

21.

22.

24.

25.

26. 27. 28.

Stiegler, R. Kunert, S. Zolla-Pazner, H. Katinger, C. J. Petropoulos, and D. R. Burton. 2004. Comprehensive cross-clade neutralization analysis of a panel of anti-human immunodeficiency virus type 1 monoclonal antibodies. J. Virol. 78:13232–13252. Bjorndal, A., H. Deng, M. Jansson, J. R. Fiore, C. Colognesi, A. Karlsson, J. Albert, G. Scarlatti, D. R. Littman, and E. M. Fenyo. 1997. Coreceptor usage of primary human immunodeficiency virus type 1 isolates varies according to biological phenotype. J. Virol. 71:7478–7487. Bjorndal, A., A. Sonnerborg, C. Tscherning, J. Albert, and E. M. Fenyo. 1999. Phenotypic characteristics of human immunodeficiency virus type 1 subtype C isolates of Ethiopian AIDS patients. AIDS Res. Hum. Retrovir. 15:647–653. Carrillo, A., and L. Ratner. 1996. Human immunodeficiency virus type 1 tropism for T-lymphoid cell lines: role of the V3 loop and C4 envelope determinants. J. Virol. 70:1301–1309. Cavacini, L., M. Duval, L. Song, R. Sangster, S. H. Xiang, J. Sodroski, and M. Posner. 2003. Conformational changes in env oligomer induced by an antibody dependent on the V3 loop base. AIDS 17:685–689. Cavacini, L. A., M. Duval, J. Robinson, and M. R. Posner. 2002. Interactions of human antibodies, epitope exposure, antibody binding and neutralization of primary isolate HIV-1 virions. AIDS 16:2409–2417. Cecilia, D., V. N. KewalRamani, J. O’Leary, B. Volsky, P. Nyambi, S. Burda, S. Xu, D. R. Littman, and S. Zolla-Pazner. 1998. Neutralization profiles of primary human immunodeficiency virus type 1 isolates in the context of coreceptor usage. J. Virol. 72:6988–6996. Cecilia, D., S. S. Kulkarni, S. P. Tripathy, R. R. Gangakhedkar, R. S. Paranjape, and D. A. Gadkari. 2000. Absence of coreceptor switch with disease progression in human immunodeficiency virus infections in India. Virology 271:253–258. Chesebro, B., K. Wehrly, J. Nishio, and S. Perryman. 1996. Mapping of independent V3 envelope determinants of human immunodeficiency virus type 1 macrophage tropism and syncytium formation in lymphocytes. J. Virol. 70:9055–9059. Cilliers, T., J. Nhlapo, M. Coetzer, D. Orlovic, T. Ketas, W. C. Olson, J. P. Moore, A. Trkola, and L. Morris. 2003. The CCR5 and CXCR4 coreceptors are both used by human immunodeficiency virus type 1 primary isolates from subtype C. J. Virol. 77:4449–4456. Coetzer, M., T. Cilliers, L. H. Ping, R. Swanstrom, and L. Morris. 2006. Genetic characteristics of the V3 region associated with CXCR4 usage in HIV-1 subtype C isolates. Virology 356:95–105. Connor, R. I., K. E. Sheridan, D. Ceradini, S. Choe, and N. R. Landau. 1997. Change in coreceptor use correlates with disease progression in HIV-1infected individuals. J. Exp. Med. 185:621–628. Cormier, E. G., and T. Dragic. 2002. The crown and stem of the V3 loop play distinct roles in human immunodeficiency virus type 1 envelope glycoprotein interactions with the CCR5 coreceptor. J. Virol. 76:8953–8957. Cormier, E. G., M. Persuh, D. A. Thompson, S. W. Lin, T. P. Sakmar, W. C. Olson, and T. Dragic. 2000. Specific interaction of CCR5 amino-terminal domain peptides containing sulfotyrosines with HIV-1 envelope glycoprotein gp120. Proc. Natl. Acad. Sci. USA 97:5762–5767. Cormier, E. G., D. N. Tran, L. Yukhayeva, W. C. Olson, and T. Dragic. 2001. Mapping the determinants of the CCR5 amino-terminal sulfopeptide interaction with soluble human immunodeficiency virus type 1 gp120-CD4 complexes. J. Virol. 75:5541–5549. Deng, H., R. Liu, W. Ellmeier, S. Choe, D. Unutmaz, M. Burkhart, P. Di Marzio, S. Marmon, R. E. Sutton, C. M. Hill, C. B. Davis, S. C. Peiper, T. J. Schall, D. R. Littman, and N. R. Landau. 1996. Identification of a major co-receptor for primary isolates of HIV-1. Nature 381:661–666. de Parseval, A., M. D. Bobardt, A. Chatterji, U. Chatterji, J. H. Elder, G. David, S. Zolla-Pazner, M. Farzan, T. H. Lee, and P. A. Gallay. 2005. A highly conserved arginine in gp120 governs HIV-1 binding to both syndecans and CCR5 via sulfated motifs. J. Biol. Chem. 280:39493–39504. Dragic, T., V. Litwin, G. P. Allaway, S. R. Martin, Y. Huang, K. A. Nagashima, C. Cayanan, P. J. Maddon, R. A. Koup, J. P. Moore, and W. A. Paxton. 1996. HIV-1 entry into CD4⫹ cells is mediated by the chemokine receptor CC-CKR-5. Nature 381:667–673. Farzan, M., H. Choe, L. Vaca, K. Martin, Y. Sun, E. Desjardins, N. Ruffing, L. Wu, R. Wyatt, N. Gerard, C. Gerard, and J. Sodroski. 1998. A tyrosinerich region in the N terminus of CCR5 is important for human immunodeficiency virus type 1 entry and mediates an association between gp120 and CCR5. J. Virol. 72:1160–1164. Farzan, M., S. Chung, W. Li, N. Vasilieva, P. L. Wright, C. E. Schnitzler, R. J. Marchione, C. Gerard, N. P. Gerard, J. Sodroski, and H. Choe. 2002. Tyrosine-sulfated peptides functionally reconstitute a CCR5 variant lacking a critical amino-terminal region. J. Biol. Chem. 277:40397–40402. Feng, Y., C. C. Broder, P. E. Kennedy, and E. A. Berger. 1996. HIV-1 entry cofactor: functional cDNA cloning of a seven-transmembrane, G proteincoupled receptor. Science 272:872–877. Furuta, R. A., C. T. Wild, Y. Weng, and C. D. Weiss. 1998. Capture of an early fusion-active conformation of HIV-1 gp41. Nat. Struct. Biol. 5:276–279. Gao, F., S. G. Morrison, D. L. Robertson, C. L. Thornton, S. Craig, G. Karlsson, J. Sodroski, M. Morgado, B. Galvao-Castro, H. von Briesen, S.

VOL. 82, 2008

29. 30.

31.

32.

33.

34. 35. 36. 37.

38.

39.

40. 41.

42. 43.

44. 45.

46.

47.

DIFFERENCES IN SUBTYPE B AND C V3 CONFORMATIONS

Beddows, J. Weber, P. M. Sharp, G. M. Shaw, B. H. Hahn, and The WHO and NIAID Networks for HIV Isolation and Characterization. 1996. Molecular cloning and analysis of functional envelope genes from human immunodeficiency virus type 1 sequence subtypes A through G. J. Virol. 70:1651– 1667. Gilbert, P. B., V. Novitsky, and M. Essex. 2005. Covariability of selected amino acid positions for HIV type 1 subtypes C and B. AIDS Res. Hum. Retrovir. 21:1016–1030. Gorny, M. K., A. J. Conley, S. Karwowska, A. Buchbinder, J. Y. Xu, E. A. Emini, S. Koenig, and S. Zolla-Pazner. 1992. Neutralization of diverse human immunodeficiency virus type 1 variants by an anti-V3 human monoclonal antibody. J. Virol. 66:7538–7542. Gorny, M. K., C. Williams, B. Volsky, K. Revesz, S. Cohen, V. R. Polonis, W. J. Honnen, S. C. Kayman, C. Krachmarov, A. Pinter, and S. ZollaPazner. 2002. Human monoclonal antibodies specific for conformation-sensitive epitopes of V3 neutralize human immunodeficiency virus type 1 primary isolates from various clades. J. Virol. 76:9035–9045. Gorny, M. K., C. Williams, B. Volsky, K. Revesz, X. H. Wang, S. Burda, T. Kimura, F. A. Konings, A. Nadas, C. A. Anyangwe, P. Nyambi, C. Krachmarov, A. Pinter, and S. Zolla-Pazner. 2006. Cross-clade neutralizing activity of human anti-V3 monoclonal antibodies derived from the cells of individuals infected with non-B clades of human immunodeficiency virus type 1. J. Virol. 80:6865–6872. Gorny, M. K., J. Y. Xu, V. Gianakakos, S. Karwowska, C. Williams, H. W. Sheppard, C. V. Hanson, and S. Zolla-Pazner. 1991. Production of siteselected neutralizing human monoclonal antibodies against the third variable domain of the human immunodeficiency virus type 1 envelope glycoprotein. Proc. Natl. Acad. Sci. USA 88:3238–3242. Gorny, M. K., J. Y. Xu, S. Karwowska, A. Buchbinder, and S. Zolla-Pazner. 1993. Repertoire of neutralizing human monoclonal antibodies specific for the V3 domain of HIV-1 gp120. J. Immunol. 150:635–643. He, Y., R. Vassell, M. Zaitseva, N. Nguyen, Z. Yang, Y. Weng, and C. D. Weiss. 2003. Peptides trap the human immunodeficiency virus type 1 envelope glycoprotein fusion intermediate at two sites. J. Virol. 77:1666–1671. Hoffman, N. G., C. A. Schiffer, and R. Swanstrom. 2003. Covariation of amino acid positions in HIV-1 protease. Virology 314:536–548. Hoffman, N. G., F. Seillier-Moiseiwitsch, J. Ahn, J. M. Walker, and R. Swanstrom. 2002. Variability in the human immunodeficiency virus type 1 gp120 Env protein linked to phenotype-associated changes in the V3 loop. J. Virol. 76:3852–3864. Huang, C. C., M. Tang, M. Y. Zhang, S. Majeed, E. Montabana, R. L. Stanfield, D. S. Dimitrov, B. Korber, J. Sodroski, I. A. Wilson, R. Wyatt, and P. D. Kwong. 2005. Structure of a V3-containing HIV-1 gp120 core. Science 310:1025–1028. Jensen, M. A., F. S. Li, A. B. van ’t Wout, D. C. Nickle, D. Shriner, H. X. He, S. McLaughlin, R. Shankarappa, J. B. Margolick, and J. I. Mullins. 2003. Improved coreceptor usage prediction and genotypic monitoring of R5to-X4 transition by motif analysis of human immunodeficiency virus type 1 env V3 loop sequences. J. Virol. 77:13376–13388. Jensen, M. A., and A. B. van ’t Wout. 2003. Predicting HIV-1 coreceptor usage with sequence analysis. AIDS Rev. 5:104–112. Johnston, E. R., L. S. Zijenah, S. Mutetwa, R. Kantor, C. Kittinunvorakoon, and D. A. Katzenstein. 2003. High frequency of syncytium-inducing and CXCR4-tropic viruses among human immunodeficiency virus type 1 subtype C-infected patients receiving antiretroviral treatment. J. Virol. 77:7682– 7688. Korber, B., S. Wolinsky, B. Haynes, K. Kunstman, R. Levy, M. Furtado, P. Otto, and G. Myers. 1992. HIV-1 intrapatient sequence diversity in the immunogenic V3 region. AIDS Res. Hum. Retrovir. 8:1461–1465. Korber, B. T., R. M. Farber, D. H. Wolpert, and A. S. Lapedes. 1993. Covariation of mutations in the V3 loop of human immunodeficiency virus type 1 envelope protein: an information theoretic analysis. Proc. Natl. Acad. Sci. USA 90:7176–7180. Korber, B. T., K. MacInnes, R. F. Smith, and G. Myers. 1994. Mutational trends in V3 loop protein sequences observed in different genetic lineages of human immunodeficiency virus type 1. J. Virol. 68:6730–6744. Krachmarov, C. P., W. J. Honnen, S. C. Kayman, M. K. Gorny, S. ZollaPazner, and A. Pinter. 2006. Factors determining the breadth and potency of neutralization by V3-specific human monoclonal antibodies derived from subjects infected with clade A or clade B strains of human immunodeficiency virus type 1. J. Virol. 80:7127–7135. Kuhmann, S. E., P. Pugach, K. J. Kunstman, J. Taylor, R. L. Stanfield, A. Snyder, J. M. Strizki, J. Riley, B. M. Baroudy, I. A. Wilson, B. T. Korber, S. M. Wolinsky, and J. P. Moore. 2004. Genetic and phenotypic analyses of human immunodeficiency virus type 1 escape from a small-molecule CCR5 inhibitor. J. Virol. 78:2790–2807. Kwong, P. D., M. L. Doyle, D. J. Casper, C. Cicala, S. A. Leavitt, S. Majeed, T. D. Steenbeke, M. Venturi, I. Chaiken, M. Fung, H. Katinger, P. W. Parren, J. Robinson, D. Van Ryk, L. Wang, D. R. Burton, E. Freire, R. Wyatt, J. Sodroski, W. A. Hendrickson, and J. Arthos. 2002. HIV-1 evades antibody-mediated neutralization through conformational masking of receptorbinding sites. Nature 420:678–682.

915

48. Milich, L., B. Margolin, and R. Swanstrom. 1993. V3 loop of the human immunodeficiency virus type 1 Env protein: interpreting sequence variability. J. Virol. 67:5623–5634. 49. Milich, L., B. H. Margolin, and R. Swanstrom. 1997. Patterns of amino acid variability in NSI-like and SI-like V3 sequences and a linked change in the CD4-binding domain of the HIV-1 Env protein. Virology 239:108–118. 50. Moore, J. P., A. Trkola, B. Korber, L. J. Boots, J. A. Kessler, II, F. E. McCutchan, J. Mascola, D. D. Ho, J. Robinson, and A. J. Conley. 1995. A human monoclonal antibody to a complex epitope in the V3 region of gp120 of human immunodeficiency virus type 1 has broad reactivity within and outside clade B. J. Virol. 69:122–130. 51. Murphy, E., B. Korber, M. C. Georges-Courbot, B. You, A. Pinter, D. Cook, M. P. Kieny, A. Georges, C. Mathiot, F. Barre-Sinoussi, et al. 1993. Diversity of V3 region sequences of human immunodeficiency viruses type 1 from the central African Republic. AIDS Res. Hum. Retrovir. 9:997–1006. 52. Pantophlet, R., R. O. Aguilar-Sino, T. Wrin, L. A. Cavacini, and D. R. Burton. 2007. Analysis of the neutralization breadth of the anti-V3 antibody F425-B4e8 and re-assessment of its epitope fine specificity by scanning mutagenesis. Virology 364:441–453. 53. Pastore, C., R. Nedellec, A. Ramos, S. Pontow, L. Ratner, and D. E. Mosier. 2006. Human immunodeficiency virus type 1 coreceptor switching: V1/V2 gain-of-fitness mutations compensate for V3 loss-of-fitness mutations. J. Virol. 80:750–758. 54. Ping, L. H., J. A. Nelson, I. F. Hoffman, J. Schock, S. L. Lamers, M. Goodman, P. Vernazza, P. Kazembe, M. Maida, D. Zimba, M. M. Goodenow, J. J. Eron, Jr., S. A. Fiscus, M. S. Cohen, and R. Swanstrom. 1999. Characterization of V3 sequence heterogeneity in subtype C human immunodeficiency virus type 1 isolates from Malawi: underrepresentation of X4 variants. J. Virol. 73:6271–6281. 55. Pinter, A., W. J. Honnen, Y. He, M. K. Gorny, S. Zolla-Pazner, and S. C. Kayman. 2004. The V1/V2 domain of gp120 is a global regulator of the sensitivity of primary human immunodeficiency virus type 1 isolates to neutralization by antibodies commonly induced upon infection. J. Virol. 78: 5205–5215. 56. Pollakis, G., S. Kang, A. Kliphuis, M. I. Chalaby, J. Goudsmit, and W. A. Paxton. 2001. N-linked glycosylation of the HIV type-1 gp120 envelope glycoprotein as a major determinant of CCR5 and CXCR4 coreceptor utilization. J. Biol. Chem. 276:13433–13441. 57. Pushko, P., M. Parker, G. V. Ludwig, N. L. Davis, R. E. Johnston, and J. F. Smith. 1997. Replicon-helper systems from attenuated Venezuelan equine encephalitis virus: expression of heterologous genes in vitro and immunization against heterologous pathogens in vivo. Virology 239:389–401. 58. Resch, W., N. Hoffman, and R. Swanstrom. 2001. Improved success of phenotype prediction of the human immunodeficiency virus type 1 from envelope variable loop 3 sequence using neural networks. Virology 288: 51–62. 59. Richman, D. D., and S. A. Bozzette. 1994. The impact of the syncytiuminducing phenotype of human immunodeficiency virus on disease progression. J. Infect. Dis. 169:968–974. 60. Rizzuto, C., and J. Sodroski. 2000. Fine definition of a conserved CCR5binding region on the human immunodeficiency virus type 1 glycoprotein 120. AIDS Res. Hum. Retrovir. 16:741–749. 61. Rizzuto, C. D., R. Wyatt, N. Hernandez-Ramos, Y. Sun, P. D. Kwong, W. A. Hendrickson, and J. Sodroski. 1998. A conserved HIV gp120 glycoprotein structure involved in chemokine receptor binding. Science 280:1949–1953. 62. Robey, W. G., B. Safai, S. Oroszlan, L. O. Arthur, M. A. Gonda, R. C. Gallo, and P. J. Fischinger. 1985. Characterization of envelope and core structural gene products of HTLV-III with sera from AIDS patients. Science 228:593– 595. 63. Roos, M. T., J. M. Lange, R. E. de Goede, R. A. Coutinho, P. T. Schellekens, F. Miedema, and M. Tersmette. 1992. Viral phenotype and immune response in primary human immunodeficiency virus type 1 infection. J. Infect. Dis. 165:427–432. 64. Scarlatti, G., E. Tresoldi, A. Bjorndal, R. Fredriksson, C. Colognesi, H. K. Deng, M. S. Malnati, A. Plebani, A. G. Siccardi, D. R. Littman, E. M. Fenyo, and P. Lusso. 1997. In vivo evolution of HIV-1 co-receptor usage and sensitivity to chemokine-mediated suppression. Nat. Med. 3:1259–1265. 65. Schramm, B., M. L. Penn, R. F. Speck, S. Y. Chan, E. De Clercq, D. Schols, R. I. Connor, and M. A. Goldsmith. 2000. Viral entry through CXCR4 is a pathogenic factor and therapeutic target in human immunodeficiency virus type 1 disease. J. Virol. 74:184–192. 66. Shannon, C. E. 1948. A mathematical theory of communication. Bell Syst. Tech. J. 27:379–423. 67. Sharon, M., N. Kessler, R. Levy, S. Zolla-Pazner, M. Gorlach, and J. Anglister. 2003. Alternative conformations of HIV-1 V3 loops mimic beta hairpins in chemokines, suggesting a mechanism for coreceptor selectivity. Structure 11:225–236. 68. Stanfield, R. L., M. K. Gorny, C. Williams, S. Zolla-Pazner, and I. A. Wilson. 2004. Structural rationale for the broad neutralization of HIV-1 by human monoclonal antibody 447-52D. Structure 12:193–204. 69. Stanfield, R. L., M. K. Gorny, S. Zolla-Pazner, and I. A. Wilson. 2006. Crystal structures of human immunodeficiency virus type 1 (HIV-1) neutral-

916

70.

71.

72. 73. 74.

75. 76.

77.

PATEL ET AL. izing antibody 2219 in complex with three different V3 peptides reveal a new binding mode for HIV-1 cross-reactivity. J. Virol. 80:6093–6105. Starcich, B. R., B. H. Hahn, G. M. Shaw, P. D. McNeely, S. Modrow, H. Wolf, E. S. Parks, W. P. Parks, S. F. Josephs, R. C. Gallo, et al. 1986. Identification and characterization of conserved and variable regions in the envelope gene of HTLV-III/LAV, the retrovirus of AIDS. Cell 45:637–648. Suphaphiphat, P., A. Thitithanyanont, S. Paca-Uccaralertkun, M. Essex, and T. H. Lee. 2003. Effect of amino acid substitution of the V3 and bridging sheet residues in human immunodeficiency virus type 1 subtype C gp120 on CCR5 utilization. J. Virol. 77:3832–3837. Theil, H. 1972. Statistical decomposition analysis. With applications in the social and administrative sciences. North-Holland Publishing Co., Amsterdam, The Netherlands. Thompson, D. A., E. G. Cormier, and T. Dragic. 2002. CCR5 and CXCR4 usage by non-clade B human immunodeficiency virus type 1 primary isolates. J. Virol. 76:3059–3064. Trkola, A., H. Kuster, P. Rusert, B. Joos, M. Fischer, C. Leemann, A. Manrique, M. Huber, M. Rehr, A. Oxenius, R. Weber, G. Stiegler, B. Vcelar, H. Katinger, L. Aceto, and H. F. Gunthard. 2005. Delay of HIV-1 rebound after cessation of antiretroviral therapy through passive transfer of human neutralizing antibodies. Nat. Med. 11:615–622. Tscherning, C., A. Alaeus, R. Fredriksson, A. Bjorndal, H. Deng, D. R. Littman, E. M. Fenyo, and J. Albert. 1998. Differences in chemokine coreceptor usage between genetic subtypes of HIV-1. Virology 241:181–188. van’t Wout, A. B., N. A. Kootstra, G. A. Mulder-Kampinga, N. Albrecht-van Lent, H. J. Scherpbier, J. Veenstra, K. Boer, R. A. Coutinho, F. Miedema, and H. Schuitemaker. 1994. Macrophage-tropic variants initiate human immunodeficiency virus type 1 infection after sexual, parenteral, and vertical transmission. J. Clin. Investig. 94:2060–2067. Verrier, F., S. Burda, R. Belshe, A. M. Duliege, J. L. Excler, M. Klein, and

J. VIROL.

78.

79.

80.

81.

82. 83. 84.

S. Zolla-Pazner. 2000. A human immunodeficiency virus prime-boost immunization regimen in humans induces antibodies that show interclade crossreactivity and neutralize several X4-, R5-, and dualtropic clade B and C primary isolates. J. Virol. 74:10025–10033. Wang, W. K., T. Dudek, M. Essex, and T. H. Lee. 1999. Hypervariable region 3 residues of HIV type 1 gp120 involved in CCR5 coreceptor utilization: therapeutic and prophylactic implications. Proc. Natl. Acad. Sci. USA 96: 4558–4562. Wei, X., J. M. Decker, S. Wang, H. Hui, J. C. Kappes, X. Wu, J. F. SalazarGonzalez, M. G. Salazar, J. M. Kilby, M. S. Saag, N. L. Komarova, M. A. Nowak, B. H. Hahn, P. D. Kwong, and G. M. Shaw. 2003. Antibody neutralization and escape by HIV-1. Nature 422:307–312. Wu, L., N. P. Gerard, R. Wyatt, H. Choe, C. Parolin, N. Ruffing, A. Borsetti, A. A. Cardoso, E. Desjardin, W. Newman, C. Gerard, and J. Sodroski. 1996. CD4-induced interaction of primary HIV-1 gp120 glycoproteins with the chemokine receptor CCR-5. Nature 384:179–183. Xiang, S. H., M. Farzan, Z. Si, N. Madani, L. Wang, E. Rosenberg, J. Robinson, and J. Sodroski. 2005. Functional mimicry of a human immunodeficiency virus type 1 coreceptor by a neutralizing monoclonal antibody. J. Virol. 79:6068–6077. Yusa, K., Y. Maeda, A. Fujioka, K. Monde, and S. Harada. 2005. Isolation of TAK-779-resistant HIV-1 from an R5 HIV-1 GP120 V3 loop library. J. Biol. Chem. 280:30083–30090. Zhu, T., H. Mo, N. Wang, D. S. Nam, Y. Cao, R. A. Koup, and D. D. Ho. 1993. Genotypic and phenotypic characterization of HIV-1 patients with primary infection. Science 261:1179–1181. Zolla-Pazner, S., P. Zhong, K. Revesz, B. Volsky, C. Williams, P. Nyambi, and M. K. Gorny. 2004. The cross-clade neutralizing activity of a human monoclonal antibody is determined by the GPGR V3 motif of HIV type 1. AIDS Res. Hum. Retrovir. 20:1254–1258.