Truncated gp120 envelope glycoprotein of human ...

Journal of General Virology (2002), 83, 2723–2732. Printed in Great Britain ...................................................................................................................................................................................................................................................................................

Truncated gp120 envelope glycoprotein of human immunodeficiency virus 1 elicits a broadly reactive neutralizing immune response S. A. Jeffs,1 C. Shotton,2 P. Balfe3 and J. A. McKeating4 1

Division of Retrovirology, NIBSC, Blanche Lane, South Mimms, Potters Bar, Herts EN6 3QG, UK School of Biological Sciences, University of Surrey, Guildford, Surrey GU2 7XH, UK 3 Division of Virology, University College of London Medical School, Windeyer Building, 46 Cleveland Street, London W1P 6DB, UK 4 School of Animal and Microbial Sciences, University of Reading, Whiteknights, PO Box 228, Reading, Berks RG6 2AJ, UK 2

Removal of the V1–V3 loops from IIIB gp120 results in a protein, PR12, with altered immunogenicity compared to the full-length protein. Polyclonal immune sera raised in rats using PR12 as immunogen recognizes envelope glycoproteins of clades A, B, C, E, F and G and can neutralize chimeric human immunodeficiency virus type 1 (HIV-1) HXB2 viruses expressing envelopes from primary HIV-1 clades B, C, E and F. These data suggest that the immune response to PR12 is directed toward conserved epitopes expressed by viral glycoproteins of diverse genotypes. Five monoclonal antibodies (mAb) derived from PR12-immunized rats were unable to neutralize virus infectivity ; hence the epitopes responsible for the induction of this cross-clade neutralizing activity remain to be elucidated. However, PR12 immune sera were able to compete with the human neutralizing mAb 2G12 for gp120 binding, implying that this epitope may be immunogenic when expressed in the context of this truncated protein.

Introduction The surface glycoprotein (gp120\41) of human immunodeficiency virus type 1 (HIV-1) is the major target of the host antibody response. Gp120 is divided into five conserved (C1–C5) and five hypervariable regions (V1–V5). In monomeric recombinant gp120, the V2, V3, C1, C5 and discontinuous CD4 binding domain (CD4bd) regions are immunodominant whereas the V1, V4, V5 and C2 regions are immunosilent ; three sets of epitopes are located within gp41 (‘ Clusters I–III ’) (Burton & Montefiori, 1997). However, the native viral surface glycoprotein appears to be trimeric, which reduces the accessibility of these important epitopes. One study which examined the antigenic conservation of epitopes of primary isolates from clades A, B, D, F, G and H concluded that only the V3 and C5 structures are shared and wellexposed, whereas CD4bd, V2 and gp41 are immunosilent (Nyambi et al., 1998). The C1 and C5 epitopes appear to be buried, probably due to interaction with gp41 (Burton & Author for correspondence : Simon Jeffs. Fax j44 1707 649865. e-mail sjeffs!nibsc.ac.uk

0001-8551 # 2002 SGM

Montefiori, 1997). If envelope-based vaccines are to be efficient then it is believed that they will need to elicit antibodies capable of neutralizing virus infectivity. The conformationdependent CD4 and chemokine receptor-binding domains of gp120 represent important targets for neutralization, but the 3dimensional conformation of gp120 suggests that these regions are located in the inner core of gp120 rendering them poorly accessible to antibody. The presence of carbohydrate glycans on the outer domain of gp120 and the possibility of subunit–subunit interactions within the native trimeric complex further occlude this region. This detailed structural knowledge may explain partially why the humoral response to HIV is fairly ineffective and why (to date) it has not proved possible to generate a potent neutralizing (and protective) response in human vaccinees against primary isolates by immunization with monomeric gp120 (Connor et al., 1998). Despite these caveats, at least three human monoclonal antibodies (mAb) capable of neutralizing divergent primary isolates have been isolated. Two of these (b12 and 2G12) interact with gp120 epitopes, while a third (2F5) recognizes a C-terminal epitope in gp41 (Roben et al., 1994 ; Trkola et al., 1996 ; Muster et al., 1993). Furthermore, passive transfer of

Downloaded from www.microbiologyresearch.org by IP: 94.0.120.204 On: Wed, 04 Nov 2015 23:52:40

CHCD

S. A. Jeffs and others

combinations of these antibodies have proved effective at controlling infection in model systems (Baba et al., 2000 ; Mascola et al., 1999, 2000). Thus the challenge is to improve the immunogenicity of recombinant gp120\41 to generate potent neutralizing responses. There are a number of ways by which this may be achieved, but most involve alteration of the molecular conformation of the molecule in order to expose these cryptic epitopes. These include (a) the removal of selected N-linked glycosylation sites (Back et al., 1994 ; Chackerian et al., 1997 ; Chakrabarti et al., 2002 ; Reitter et al., 1998) ; (b) the use of multivalent envelope glycoprotein (Montefiori & Evans, 1999) ; (c) a consideration of the role of monomeric versus oligomeric recombinant glycoprotein (Burton & Moore, 1998 ; Montefiori & Evans, 1999) ; (d) the use of fusioncompetent immunogens (LaCasse et al., 1999) ; and (e) the production of novel stabilized trimers of envelope glycoprotein (Chakrabarti et al., 2002 ; Srivastava et al., 2002 ; Binley et al., 2000 ; Yang et al., 2000). In previous studies, we (Jeffs et al., 1996) and others (Pollard et al., 1992 ; Wyatt et al., 1993) have reported that removal of the V1, V2 and V3 hypervariable loops results in a molecule which is capable of binding sCD4 with an affinity comparable to that of the full-length protein. Furthermore, removal of these loops increases accessibility of the C1 and C4 regions to mAbs (Jeffs et al., 1996). We therefore investigated the immunogenicity of this truncated protein (‘ PR12 ’) with the intention of re-directing the immune response from the variable loops to the more conserved regions critical for gp120 function.

Methods

Antigen expression, purification and immunization. During this study, cDNA from the BH10 clone of the HIV-1 IIIB isolate was used. Full details of the construction of the expression plasmids, the expression of recombinant gp120 from stably engineered Chinese hamster ovary (CHO) cell lines and its purification are provided in Jeffs et al. (1996). The truncated mutant gp120 PR12 comprises three separate fragments : F1 l aa 75–123, linked by an XbaI (Ser–Arg) linkage to F2 l aa 205–299, linked by a KpnI (Gly–Thr) linkage to F3 l aa 329–510. All samples of full-length (FL) and PR12 gp120 were characterized by SDS–PAGE and immunoblotting using polyclonal sera [ADP421 – Centralised Facility for AIDS Reagents (CFAR)] raised against CHO gp120 IIIBBH as described "! previously. Elucidation of sCD4 binding and recognition of envelope glycoprotein by IIIB-specific mAbs recognizing linear and conformationdependent epitopes were performed by ELISA-based assays as described previously. Anti-FL\PR12 gp120 sera were raised in pairs of naı$ ve CBH\Cbi rats by three immunizations, at fortnightly intervals, with 10 µg of CHOexpressed IIIBBH FL or PR12 gp120 emulsified in Freundhs complete "! (first immunization) or incomplete (second and third immunizations) adjuvant (McKeating et al., 1996). Anti-PR12 mAbs were generated in rats using the method of Shotton et al. (1995).

Transient expression and quantification of gp120 ; antibody-binding ELISAs VTF7.3-infected human embryonal kidney 293 cells were transfected, using lipofectamine (Gibco BRL), with pCDNA.3 CHCE

plasmids encoding divergent gp120 proteins. Cell supernatants were collected after 72 h, clarified by centrifugation at 1500 g and assayed for gp120 levels using a quantitative gp120 capture ELISA. This assay involves coating plates with the ovine antisera D7324 (Aalto Bioreagents), raised against a peptide conforming to the 15 C-terminal amino acids of IIIBBH gp120, to bind HIV-1 gp120 present in the sample. The "! presence of gp120 is detected by a rabbit CHO-derived IIIBBH gp120 "! antisera (ARP421, CFAR). All of the divergent proteins possess the conserved C-terminal ‘ D7324 epitope ’ encoded by oligonucleotide 4141 used for PCR amplification. Bound ligands were visualized with antirabbit IgG–HRP. Recombinant IIIBBH gp120 derived from CHO cells "! (EVA657, CFAR) was used for calibration. For antibody-binding ELISAs, gp120 was allowed to bind to D7324 at an input concentration of 500 ng\ml, followed by addition of serially diluted mAb solutions. Detection was by the appropriate anti-species IgG–HRP and the titre of mAb giving half-maximal binding was determined graphically.

PCR amplification of gp120 sequences both for expression of soluble proteins and for generation of chimeric constructs. The PCR followed the protocols previously reported (McKeating et al., 1993). The gp120-encoding region was amplified from a panel of pBluescript (Stratagene) clones containing diverse envelope genes (the kind gift of F. McCutchan, Henry Jackson Foundation, Rockville, MD, USA) using primers 626 (sense, restriction enzyme site underlined) 5h GTG GGTCACCGT CTA TTA TTG GG and 524 (antisense) 5h CAC CACGCGTCT CTT TGC CTT GGT GGG. PCR products were purified using ‘ Geneclean ’ (Bio 101), restriction enzyme digested with either BstEII or MluI, ligated into pHXB2-MCS ∆ env (McKeating et al., 1993) and transformed into competent TG2 E. coli. Individual colonies were screened for inserts by PCR using primers 626 and 524. Plasmid preparations of clones containing inserts were prepared using Wizard miniprep kits (Promega). The gp120 open reading frame was also amplified using primers 4142 and 4141, uracilated derivatives of the primers 626 and 524 described above, with 100 ng of plasmid template and using the conditions described above. The PCR product was gel-purified and 100 ng annealed to 25 ng of the vector pCDNA3.tpa, which contains complementary uracilated sequences, in the presence of 0n25 U of uracil deglycosylase (CloneAmp, Gibco BRL). Colonies resulting from transformation with the annealed vector were screened for the presence of insert and plasmid DNA prepared from them.

Transfection of HXB2 chimeric clones and neutralization assays. Plasmid DNAs (5 µg) were transfected into HeLa cells using lipofectamine. At 72 h post-transfection the cells were co-cultured with 2i10' PHA-stimulated peripheral blood mononuclear cells (PBMC) for 24 h. PBMC were recovered from the HeLa monolayer, washed and cultured in RPMI–10 % FCS–IL-2 (5 U\ml). The extracellular fluid was tested for the presence of soluble p24 antigen as described previously (McKeating et al., 1993). Cell-free supernatants were collected from PBMC cultures and assessed for the levels of infectious virus by determining the median infectious dose on human phytohaemagglutinin (PHA) stimulation. Infection was measured by the detection of p24 antigen and the TCID values determined by the Karber formula (Lewis &! et al., 1998). Neutralization was assessed using PHA-stimulated PBMC as target cells and virus replication was assessed by measuring p24 antigen production as reported previously (Von Gegerfelt et al., 1991 ; Scarlatti et al., 1993 ; Albert et al., 1990). A 75 µl sample from each diluted rat antisera was incubated for 1 h at 37 mC with an equal volume of virus (in triplicate, at dilutions of 1 : 5, 1 : 25 and 1 : 125). Following virus–antibody


Truncated gp120 elicits neutralizing responses

Table 1. Comparative antigenicity of gp120 and PR12 proteins expressed in CHO and 293 cells All mAbs are McKeating & Shotton (unpublished data) ; see footnotes for exceptions. , Not tested ; , No saturation observed ; , No detectable binding. Concentration of mAb resulting in half-maximal binding to proteins (µg/ml) 293-produced mAb 33\46b 55\1c\1g 55\1b\2d 35\45b\6j 11\65a* 55\58\7h 38.1a† 55\45b\6e 39.13g† 10\46c 35\25\2b 35\75\4k 33\90 33\50a 8\19b\3e 2G12‡

CHO-produced

Epitope

gp120

PR12

gp120

PR12

C1 aa 81–101 C1 aa 102–121 C1 aa 102–121 C1 aa 102–121 C1 aa 102–121 C1 aa 73–92 C4 aa 430–446 C4 aa 393–412 conf CD4 bs conf CD4 bs conf gp120 conf gp120 conf gp120 conf gp120 conf gp120 conf gp120

0n12 0n20 0n30 0n16 1n12 2n50 0n002 0n28 0n01 0n10 0n09 0n36 0n03 0n05 0n31 0n03

0n01 0n01 0n04 0n03 0n05 0n35 0n0003 0n06 0n0006 0n0004 0n08 0n41 0n03 0n05 2n56 0n06

      0n08  0n025 0n50 0n07  0n05 0n06 0n60 0n04

0n01 0n02 0n03 0n05 0n03 0n30 0n0008 0n05 0n001 0n0007 0n06  0n06 0n07 3n25 0n05

* Shotton et al. (1995). † Cordell et al. (1991). ‡ Buchacher et al. (1994). incubation, 1i10& PBMC in a final volume of 75 µl RPMI–10 % FCS–IL-2 (5 U\ml) were added. PBMC were washed on days 1 and 3 by centrifugation and the medium changed (Albert et al., 1993). The TCID &! of the virus stock was determined in parallel and the neutralization assay was evaluated for virus dilutions containing 10–75 TCID . The neutral&! izing titre of a serum is defined as the highest serum dilution which completely inhibits virus replication as assessed by p24 antigen production, that is an absorbance value below the cutoff value for the antigen ELISA of 3 pg (the meanj2 SD of negative controls containing cells only). Controls included in each assay were virus alone and cells alone. Positive virus replication controls contained either IL-2 medium or rat pre-bleed sera diluted 1 : 10. Positive anti-HIV-1 serum control utilized serum from an asymptomatic homosexual man with a high-titre of neutralizing antibodies against HIV-1 IIIB diluted 1 : 10 in medium.

Phylogenetic analysis of divergent envelope sequences. In order to assess the genetic variation present in the epitopes of the PR12 construct, we constructed an alignment of representative sequences from the major subtypes of HIV-1 (A–F). The sequence for PR12 was added to this alignment and the deleted regions removed from all sequences (assigned weights of zero). Maximum-likelihood trees were constructed from the reduced data set and a tree constructed (using the ‘  ’ suite of programs, version 3.5c, J. Felsenstein, University of Washington). Pairwise distances were calculated from the alignment (Kimura 2parameter method) and comparisons made of intra-subtype and intersubtype distances. For complete sequences, such comparisons show that

members of a subtype are more closely related to one another than to unrelated subtypes (analysis not shown). Finally, bootstrap resampling of the data set by neighbour joining was used to assess the robustness of the tree obtained.

Results Antigenicity of the truncated gp120 protein PR12

We (Jeffs et al., 1996) and others (Pollard et al., 1992 ; Wyatt et al., 1993) reported previously that mAbs to the C1 and CD4bd were able to bind with higher affinity to PR12, a truncated gp120 which lacks the V1, V2 and V3 regions. To characterize the antigenicity of the PR12 protein further, we evaluated the ability of a number of mAbs that recognize discontinuous epitopes, distinct from the CD4bd, to bind FL and PR12 gp120. With the exception of antibody 8\19b\3e, mAbs bound equally well to both full-length and loop-deleted gp120, suggesting that deletion of the variable loops did not lead to gross conformational changes in the molecule (Table 1). All of the mAbs specific for epitopes within the C1, C4 and CD4bd regions showed improved binding to the PR12 protein. It is interesting to note that mAb 55\45b\6e, which bound well to both PR12 and peptide 740.43, bound poorly to FL gp120


CHCF


Fig. 2. Immunogenicity of CHO-produced PR12 and FL gp120. Reciprocal dilutions of antisera from PR12 gp120-immunized rats 7 and 8 (A) or FL gp120-immunized rats 7 and 8 (B) binding to recombinant CHO PR12 or FL gp120.

Fig. 1. Comparative recognition of CHO and 293 cell-produced antigen by mAbs specific for linear epitopes within the C1 and CV4 regions of gp120. A, mAb 55/58/7h ; B, mAb 55/45b/6e.

(Table 1 and Fig. 1A), suggesting that this epitope within the C4 region is masked by the presence of the variable loops. This mAb also bound with high affinity to another truncated IIIBBH gp120, PR6, in which the V1 and V2 loops were "! deleted, implying that the masking of this region within C4 is independent of the V3 region (data not shown). Furthermore, mAb 55\45b\6e failed to inhibit sCD4 recognition of the PR12 protein (data not shown), whereas mAb 38.1a, binding to the adjacent C4 amino acids 427–436, inhibited the CD4-PR12 interaction (data not shown). These data are consistent with the occlusion of this region by the presence of glycosylated variable loops on the native molecule. All of the above experiments analysed FL and PR12 gp120 proteins produced in CHO cells. In order to address any possible effects of cell-specific post-translational modifications, such as glycosylation, we compared the antigenicity of FL and PR12 gp120 proteins produced in a human embryonal kidney cell line 293, with those from CHO cells. Proteins were CHCG

expressed transiently in both 293 and CHO cells and levels of soluble protein measured using a quantitative ELISA. Equivalent amounts of protein were compared for their ability to bind the panel of mAbs listed in Table 1. All of the mAbs specific for the C1 region bound with higher affinity to both FL and PR12 gp120 produced in 293 cells than CHO cells (Table 1). This is exemplified by the C1 mAb 55\58\7h which fails to recognize CHO-produced FL gp120 but binds, albeit with low affinity, to 293-produced FL gp120 (Fig. 1A). Similarly, the C4 mAb 55\45b\6e bound to 293-produced FL gp120 despite showing negligible reactivity to CHO-produced FL gp120 (Table 1 and Fig. 1B). The altered conformation(s) within the C1 and C4 regions of FL gp120 may result from different post-translational modifications such that care should be taken in the interpretation of differential mAb reactivity for full-length and truncated proteins. Immunogenicity of the truncated PR12 gp120 protein

Analysis of the antibody response from rats immunized with PR12 demonstrated that serum antibodies reacted preferentially with PR12 compared to FL gp120 (Fig. 2A).



Table 2. Cross-reactive antibody response induced in rats immunized with gp120 (rats 10, 11) or PR12 (rats 7, 8) Reciprocal dilution of serum resulting in half-maximal gp120 binding from rats immunized with : Gp120

Clade designation

PR12.7

PR12.8

gp120.10

gp120.11

BH10 UG273 UG266 UG268 SF-2 BK132 US3 US4 ZAS29 SE364 CM235 CM239 CM243 BZ126 G3

B A A A B B B B C C E E E F G

1300 800 1300 1650 1000 1100 1400 1500 900 1200 900 1600 750 1300 1700

1800 1400 1800 1950 1200 2800 1700 1600 1300 1650 1300 2700 1900 2500 2100

5000 350 280 240 340 460 280 300 550 400 550 5700 300 310 670

2500 460 450 400 340 500 320 310 200 280 200 4100 310 380 1300

Table 3. Neutralizing antibody response induced in rats immunized with gp120 (rats 10, 11) or PR12 (rats 7, 8) Neutralization titre (reciprocal dilution) of sera from rats immunized with : Virus HXB2 HXB2.BK132 HXB2.US3 HXB2.US4 HXB2.SF2 HXB2.SE364 HXB2.SE365 HXB2.CM243 HXB2.BZ126

Clade designation

Phenotype

PR12.7

PR12.8

B B B B B C C E F

SI SI NSI NSI SI NSI SI NSI SI

20 20 40 20 80 40 40 40 20

40 20 40 20 40 80 20 40 20

Similarly, serum from FL gp120-immunized rats bound with higher affinity to FL gp120 than to PR12 (Fig. 2B). However, none of the sera were able to inhibit sCD4\FL or PR12 gp120 interactions by competition ELISA (data not shown), suggesting that this region remains immunosilent in PR12 immunized rats. Deletion of the variable loops may result in an immune response which is directed towards the more conserved regions of the glycoprotein. To examine this, we compared the ability of sera from FL and PR12 gp120-immunized rats to recognize gp120 proteins derived from other primary isolates of HIV-1.

gp120.10 40 20 20 20 20 20 20 20 20

gp120.11 20 20 20 20 20 20 20 20 20

Sera from PR12-immunized rats bound to both FL gp120 and the divergent proteins with comparable relative affinities (Table 2). In contrast, sera from FL gp120-immunized rats bound with high affinity to the homologous protein but demonstrated an approximate one log reduction in affinity for the divergent proteins (Table 2). These data are consistent with the immunodominance of the variable loops and the typespecific nature of such antibodies. A central aim of this project is the generation of antibodies able to neutralize HIV-1 from divergent clades. Therefore we measured the ability of sera to neutralize both HXB2 and a


CHCH


Fig. 4. Percentage inhibition of mAb 2G12 or 39.13g binding to FL gp120 in the presence of serum from PR12 gp120- (rats 7, 8) or FL gp120- (rats 10, 11) immunized rats.

Fig. 3. Unrooted maximum-likelihood analysis of divergent gp120 sequences, ignoring variable loops (program DANML). Clade A, DJ264 and UG273 ; Clade B, US-3, HXB2, BK132 and US-4 ; Clade C, UG268, ZAM18 and SE364 ; Clade D, SE365 and UG266 ; Clade E, CM235, CM239 and CM243 ; Clade F, BZ126. Comparison showed no significant differences between intra-clade and inter-clade distances (Kimura distance, Wilcoxon U statistic, NS). Bootstrap re-sampling showed no groupings with greater than 70 % bootstrap values, except for the three related CM samples which constitute subtype E (PHYLIP programs).

number of HXB2 chimeric viruses expressing gp120 open reading frames (ORFs) cloned from viruses of diverse clades. All gp120 sequences were confirmed to encode a full ORF both by sequence analysis and by the ability to encode expression of a soluble protein. Sera from PR12-immunized rats were able to cross-neutralize chimeric viruses expressing gp120 proteins of different clades, albeit with low titres. In contrast, sera from FL gp120-immunized rats neutralized the homologous virus only (Table 3). Pre-bleed sera had no neutralization activity. This greater degree of cross-recognition of serum antibodies from PR12-immunized rats is consistent with phylogenetic analyses of the divergent gp120 peptides (Fig. 3). Inter- and intra-clade distances calculated from truncated sequences, based on the PR12 deletions, are less distinct than those calculated for the full-length sequences (Fig. 3). The branches distinguishing subtypes are not significant for the truncated proteins, where only three clades (E, B\D and A\C\F) are weakly supported by bootstrap resampling. The inability of the rat immune sera to block sCD4–gp120 interaction suggested that antibodies to CD4bd were not contributing to the cross-neutralizing activity of the PR12 immune sera. The human neutralizing mAb 2G12, specific for a glycan-dependent epitope in the C4\V4 region of gp120, is capable of binding PR12 (Table 1), raising the possibility that the ‘ 2G12 epitope ’ may be immunogenic within the context of the PR12 protein. To address whether antibodies to epitopes similar to or overlapping with that of 2G12 are present in the immune sera, we measured the ability of sera to compete with CHCI

2G12 for gp120 binding. Sera from PR12-immunized rats competed with 2G12 for its interaction with FL gp120, whereas sera from the FL gp120-immunized animals showed minimal levels of competition (Fig. 4). In addition, all of the sera failed to compete with mAb 39.13g, specific for the discontinuous CD4 binding site. These data are consistent with the inability of the sera to block sCD4–gp120 interactions. In order to further understand which epitope(s) on PR12 may be responsible for such cross-neutralizing activity, we generated a number of mAbs from PR12-immunized rats (Table 4). All of the mAbs recognized conformation-dependent epitopes ; however, mAbs 58\1b\7i and 58\30 competed for FL gp120 binding with mAb 55\45b\6e (Table 4), specific for a linear epitope within the C4 region. MAbs 58\1b\7i and 58\30 showed enhanced recognition of PR12 compared to FL gp120. These mAbs bound with higher affinity to FL gp120 produced in 293 cells than that expressed in CHO cells (Table 1). Furthermore, these mAbs demonstrated broad crossreactivity for gp120 proteins of diverse clades but failed to neutralize virus infectivity (data not shown). Three additional mAbs, 62\59b, 62\41 and 62\51, bound only to PR12 and failed to recognize any of the full-length proteins tested (listed in Table 2). Not surprisingly, these mAbs were unable to neutralize any of the viruses tested (data not shown). Furthermore, none of the mAbs raised against PR12 were able to crosscompete with mAb 2G12 for FL gp120 binding (data not shown). In summary, these results support the view that truncated forms (PR12) of IIIB gp120 induce a different immune response from the full-length protein. The serum neutralizing response is cross-reactive with chimeric viruses expressing gp120 proteins derived from primary viruses of different genetic subtypes. However, the specificity of the neutralizing antibodies remains to be elucidated since the mAbs generated from these immunized animals failed to mimic the polyclonal response. Further studies will examine the neutralizing ability of the PR12 mAbs in combination.



Table 4. Characterization of mAbs raised from PR12 immunized rats Concentration of mAb resulting in half maximal binding to proteins (µg/ml) 293-produced

CHO-produced

mAb

Epitope

Neutralizing

gp120

PR12

gp120

58\1b\7i 58\30 62\59b 62\41 62\51

conf-dept* conf-dept conf-dept conf-dept conf-dept

k k k k k

0n16 0n04 Neg† Neg Neg

0n02 0n002 0n25 0n08 0n58

0n8 0n3 Neg Neg Neg

PR12 0n04 0n004 0n30 0n09 0n04

* MAbs 58\1b\7i and 58\30 compete with mAb 55\45b\6e which is specific for aa 393–412 within the C4 region. Conf-dept, conformation dependent. † Neg, negative.

Discussion The data presented here show that immunization of rats with a truncated IIIBBH gp120 protein, lacking the V1–V3 "! regions, induced a cross-reactive neutralizing response (Table 3), whereas immunization with the full-length gp120 protein induced a type-specific neutralizing response, as reported by others (Berman et al., 1992). It should be noted that the chimeric viruses used for the neutralization studies show similar patterns of neutralization resistance to those of the parental primary isolates (McKeating et al., 1996). Antibodies from PR12-immunized rodents reacted preferentially with the PR12 protein compared to the FL gp120, suggesting that they recognize epitopes which may be cryptic on the native molecule (Fig. 2). However, the ability of PR12 antisera to neutralize chimeric viruses expressing divergent gp120 sequences and to recognize divergent gp120 proteins with affinities comparable to that of the FL protein confirms that a number of these antibodies recognize epitopes exposed on the native molecule. In contrast, FL gp120 antisera only neutralized the homologous virus and reacted preferentially with the FL gp120 protein over the divergent proteins (Table 3). These results are consistent with the immunodominant nature of the V2 and V3 regions, which generally leads to the induction of antibodies which neutralize only homologous virus (Fenouillet et al., 1995 ; Moore et al., 1994). Removal of the V1–V3 loops may influence glycosylation and processing of the truncated molecule, such that increased mAb recognition may be due to an indirect effect of altered processing rather than a direct effect on the disruption of interdomain interactions. Wyatt et al. (1993) reported that processing and transport to the cell surface of a V1–V3-deleted gp160 protein was increased relative to the full-length molecule. It is therefore possible that cell type-dependent glycosylation patterns may explain the different binding

affinities of some mAbs for gp120 expressed in 293 and CHO cells (Table 1). Generally, only mAbs specific for epitopes within C1, C4 and the discontinuous CD4 binding site bound with greater affinity to FL gp120 expressed in 293 cells compared to that from CHO cells (Table 1). These mAbs were also able to bind gp120 proteins of diverse clades, expressed from 293 cells, suggesting that their ability to bind full-length envelope glycoprotein is not specific for the immunizing strain (data not shown). Interestingly, exposure of the same gp120 regions, as assessed by mAb binding, was increased on the PR12 molecule (Table 1 and Fig. 1A). We were interested to know which regions, if any, were immunodominant on the PR12 molecule. Several authors have suggested that the enhanced exposure of epitopes within C4 and CD4bd may correlate with enhanced immunogenicity of these regions (Pollard et al., 1992 ; Wyatt et al., 1993 ; Moore et al., 1994 ; Jeffs et al., 1996). Initially, we investigated the reactivity of the immune sera for a series of overlapping gp120 peptides. FL gp120 immune sera reacted with epitopes in the C1, V1, V2 and V3 regions, whereas PR12 immune sera failed to react to any of the tested gp120 peptides (data not shown). These data demonstrate that antibodies raised to PR12 were specific for conformation-dependent epitopes and that the immune response was not directed to ‘ new ’ linear epitope(s). Neither PR12 nor FL gp120 immune sera competed with human mAb 697, specific for the CD4 binding site, for binding to FL gp120 (data not shown). Furthermore, we were unable to detect antibodies capable of inhibiting the envelope gp–sCD4 interaction in either FL or PR12 gp120 immune sera (data not shown). Overall, these data suggest that CD4bd is not immunogenic in the PR12 molecule and that antibodies to this region do not contribute to the neutralizing activity of the immune sera. Moore et al. (1994) reported a lack of association between the ability of antibodies to inhibit the sCD4–gp120 interaction and their ability to neutralize, suggesting dif-


CHCJ


ferential CD4bd exposure on soluble monomeric and viral oligomeric gp120. However, all neutralizing mAbs shown to recognize epitopes overlapping CD4bd have been reported to inhibit the in vitro sCD4–gp120 interaction (Cordell et al., 1991 ; McKeating et al., 1992). We noted that the neutralizing human mAb 2G12, specific for a glycan-dependent epitope within the carboxyl region of gp120, bound with equivalent affinities to FL and PR12 gp120. Furthermore, PR12 immune sera were able to compete with 2G12 for binding to FL gp120, whereas FL gp120 immune sera were unable to compete (Fig. 4). It is interesting to note that 2G12 is one of the few mAbs where an association between affinity for soluble gp120 and neutralization efficiency has been reported (Moore et al., 1995). These data suggest that ‘ 2G12like ’ antibodies may contribute to the neutralizing activity of the PR12 immune sera and this epitope may be immunogenic within PR12. In order to evaluate the immunogenicity of this region during natural infection, a panel of human HIV-positive sera were tested for their ability to compete with 2G12 binding to FL gp120. Our data are in agreement with those of Trkola et al. (1996) in that none of the sera were able to compete with 2G12 (data not shown), suggesting that this epitope is poorly immunogenic in naturally infected individuals. In order to further evaluate the PR12 immune response, we generated a number of mAbs from rats immunized with PR12. Five mAbs specific for conformation-dependent epitopes were obtained, three of which, 62\59b, 62\41 and 62\51, were only able to bind PR12 (Table 4). These mAbs failed to react with any of the gp120 proteins listed in Table 2 or to neutralize HXB2 (data not shown). The two additional mAbs, 58\1b\7i and 58\30, demonstrated broad cross-reactivity for gp120 proteins of different clades (data not shown), in that both mAbs were able to bind all proteins tested so far (total of 24 from clades A–G). Both mAbs showed increased binding affinity for PR12 and were able to compete with the C4 mAb 55\45b\6e (data not shown), suggesting that C4-dependent epitopes are immunogenic on PR12. We therefore tested the ability of the polyclonal immune sera to compete with mAb 55\45b\6e for FL gp120 binding ; however, no competition was observed with either FL gp120 or PR12 immune sera. Unfortunately, neither of these mAbs demonstrated neutralizing activity for HXB2 or any of the chimeric viruses (data not shown). In a parallel study, PR12 was used to select human mAbs from HIV-infected individuals (Jeffs et al., 2001). Five mAbs were generated with specificities overlapping the CD4 binding domain, one of which, mAb 1570, recognized a conserved epitope in recombinant envelope proteins derived from clade A, B, C, D and E viruses. Initial results suggest that this mAb may show cross-clade neutralizing activity (Zolla-Pazner and Jeffs, unpublished results). Clearly, one potential disadvantage of immunizing with PR12 is the generation of PR12-specific antibodies. However, mAbs 58\1b, 58\30 (this study) and 1570 (Jeffs et al., 2001) do appear to show broad cross-clade reactivity. Furthermore, the ability of PR12 serum antibodies CHDA

to neutralize and bind divergent gp120 proteins suggests that ‘ PR12-specific ’ antibodies may be a minor component of the polyclonal response (Table 2). Clearly, it is important to attempt to identify the epitopes responsible for inducing this neutralizing activity ; however, the polyclonal response may result from antibodies recognizing multiple epitopes and exerting synergistic effects. In summary, we have shown that removal of the V1–V3 loops from IIIB gp120 results in a protein, PR12, with altered immunogenicity compared to the full-length protein. Differences were found both in the ability of the immune sera to recognize glycoproteins of diverse origins and in their ability to neutralize viruses chimeric for primary gp120 proteins. These data lead us to conclude that the immune response to PR12 is directed toward conserved epitopes present on the majority of viral glycoproteins. None of the mAbs derived from the PR12-immunized rats were able to neutralize virus infectivity and hence the epitopes responsible for the induction of this cross-neutralizing activity remain to be elucidated. However, PR12 immune sera were able to compete with the human neutralizing mAb 2G12 for gp120 binding, implying that this epitope may be immunogenic within the truncated protein. In contrast, the epitopes recognized by PR12-selected human mAbs appear to be located within the CD4 binding domain and at least one of these shows cross-clade neutralizing activity (Jeffs et al., 2001). The challenge ahead is to induce potent cross-clade neutralizing antibody responses in experimental animals using glycoprotein immunogens based on primary isolates of HIV-1. To date, this has proved notoriously difficult to achieve. Indeed, one very promising approach using immunogens comprising formaldehyde-fixed co-cultures of cells expressing HIV-1 envelope glycoprotein, CD4 and CCR5 (LaCasse et al., 1999), has recently been retracted following reports that a major fraction of the reported neutralization was due to a specific cytotoxic effect of the antisera (Nunberg, 2002). We remain convinced that envelope-based immunogens do induce protective humoral responses. However, considerable effort is required to identify in vitro assays that correlate with protection observed in either model systems or clinical trials in vivo. Certainly at the moment it is important that promising results are demonstrated to be reproducible and evaluated using a wide variety of neutralization assays and HIV-1 isolates. In addition, a wide variety of projects are currently in progress in our laboratory, designed to evaluate the comparative immunogenicity of monomeric and oligomeric glycoproteins with modifications such as hypervariable loop-deletion, specific N-linked glycan removal and the addition of potentially adjuvanting molecules to the polypeptide backbone. It is hoped that these modifications will provide information on the structural components that will be required for an effective HIV envelope-based AIDS vaccine. Reagents ARP421, EVA657 and EVA3064 were provided by the EU Programme EVA\MRC Centralised Facility for AIDS Reagents, NIBSC,


Truncated gp120 elicits neutralizing responses UK (Grant Number QLK2-CT-1999-00609 and GP828102). J. A. M. was supported by the Lister Institute for Preventive Medicine. We thank Neil Almond (Division of Retrovirology, NIBSC, UK) for critical reading of this manuscript.

References Albert, J., Abrahamsson, B., Nagy, K., Aurelius, E., Gaines, H., Nystrom, G. & Fenyo, E.-M. (1990). Rapid development of isolate-specific

neutralizing antibodies after primary HIV-1 infection and consequent emergence of virus variants which resist neutralization by autologous sera. AIDS 4, 107–112.

by human immunodeficiency virus type I while participating in trials of recombinant gp120 subunit vaccines. Journal of Virology 72, 1552–1576. Cordell, J., Moore, J. P., Dean, C. J., Klasse, P. J., Weiss, R. A. & McKeating, J. A. (1991). Rat monoclonal antibodies to nonoverlapping

epitopes of human immunodeficiency virus type 1 gp120 block CD4 binding in vitro. Virology 185, 72–79. Fenouillet, E., Blanes, N., Benjouad, A. & Gluckman, J. C. (1995). AntiV3 antibody reactivity correlates with clinical stage of HIV-1 infection and with serum neutralizing activity. Clinical and Experimental Immunology 99, 419–424. Jeffs, S. A., McKeating, J., Lewis, S., Craft, H., Biram, D., Stephens, P. E. & Brady, R. L. (1996). Antigenicity of truncated forms of the

Albert, J., Bjorling, E., Vongegerfelt, A., Scarlatti, G., Zhang, Y. J., Fenyo, E-M. & Thorstensson, R. (1993). Antigen-detection is a reliable

human immunodeficiency virus type 1 envelope glycoprotein. Journal of General Virology 77, 1403–1410.

method for evaluating HIV\SIV neutralization assays. AIDS Research and Human Retroviruses 9, 501–504.

Jeffs, S. A., Gorny, M. K., Williams, C., Revesz, K., Volsky, B., Burda, S., Wang, X.-H., Bandres, J., Zolla-Pazner, S. & Holmes, H. (2001).

Baba, T. W., Liska, V., Hofmann-Lehmann, R., Vlasak, J., Xu, W., Ayehunie, S., Cavacini, L. A., Posner, M. R., Katinger, H., Stiegler, G., Bernacky, B. J., Rizvi, T. A., Schmidt, R., Hill, L. R., Keeling, M. E., Lu, Y., Wright, J. E., Chou, T.-C. & Ruprecht, R. M. (2000). Human

Characterisation of human monoclonal antibodies selected with a hypervariable loop-deleted recombinant HIV-1IIIB gp120. Immunology Letters 79, 209–213.

neutralizing monoclonal antibodies of the IgG1 subtype protect against mucosal simian-human immunodeficiency virus interaction. Nature Medicine 6, 200–206. Back, N. K. T., Smith, L. & Dejog, J. J. (1994). An N-glycan within the human immunodeficiency virus type 1 gp120 V3 loop affects virus neutralization. Virology 199, 431–443.

LaCasse, R. A., Follis, K. E., Trahey, M., Scarborough, J. D., Littman, D. R. & Nunberg, J. H. (1999). Fusion-competent vaccines : broad

neutralisation of primary isolates of HIV. Science 283, 357–362. Lewis, J., Balfe, P., Arnold, C., Kaye, S., Tedder, R. S. & McKeating, J. M. (1998). Development of a neutralizing antibody response during

against gp120 from the MN isolate of HIV-1. Journal of Virology 66, 4464–4469.

acute primary human immunodeficiency virus type 1 infection and the emergence of antigenic variants. Journal of Virology 72, 8943–8951. McKeating, J. A., Cordell, J., Dean, C. J. & Balfe, P. (1992). Synergistic interaction between ligands binding to the CD4 binding site and V3 domain of human immunodeficiency virus type 1 gp120. Virology 191, 732–742.

Binley, J. M., Sanders, R. W., Las, B., Schuelke, N., Master, A., Guo, Y., Fajumo, F., Anselma, D. J., Maddon, P. J., Olson, W. C. & Moore, J. P. (2000). A recombinant human immunodeficiency virus type I envelope

McKeating, J. A., Shotton, C., Cordell, J., Graham, S., Balfe, P., Sullivan, N., Charles, M., Page, M., Bolmstedt, A., Olofsson, S., Wu, Z., Pinter, A., Dean, C., Sodroski, J. & Weiss, R. A. (1993). Characterization

glycoprotein complex stabilised by an intermolecular disulfide bond between the gp120 and gp41 subunits is an antigenic mimic of the trimeric virion-associated structure. Journal of Virology 74, 627–643.

of neutralizing monoclonal antibodies to linear and conformationdependent epitopes within the first and second variable domains of human immunodeficiency virus type 1 gp120. Journal of Virology 67, 4932–4944.

Berman, P. W., Matthews, T. J., Riddle, L., Champe, M., Hobbs, M. R., Nakamura, G. R., Mercer, J. & Eastman, D. J. (1992). Antisera raised

Buchacher, A., Predl, R., Strutzenberger, K., Steinfellner, W., Trkola, A., Purtscher, M., Gruber, G., Tauer, C., Steindl, F., Jungbauer, A. & Katinger, H. (1994). Generation of human monoclonal antibodies

against HIV-1 proteins ; electrofusion and Epstein–Barr virus transformation for peripheral blood lymphocyte immortalization. AIDS Research and Human Retroviruses 10, 359–369. Burton, D. R. & Montefiori, D. C. (1997). The antibody response in HIV-1 infection. AIDS 11, Supplement A, S87–S98. Burton, D. R. & Moore, J. P. (1998). Why do we not have an HIV vaccine and how can we make one? Nature Medicine 4, 495–498. Chackerian, B., Rudensey, L. M. & Overbaugh, J. (1997). Specific Nlinked and O-linked glycosylation modifications in the envelope V1 domain of simian immunodeficiency virus that evolve in the host alter recognition by neutralizing antibodies. Journal of Virology 71, 7719–7727. Chakrabarti, B. K., Kong, W.-P., Wu, B.-Y., Yang, Z.-Y., Friborg, J., Ling, X., King, S. R., Montefiori, D. C. & Nable, G. J. (2002).

Modifications of the human immunodeficiency virus envelope glycoprotein enhance immunogenicity for genetic immunization. Journal of Virology 76, 5357–5368. Connor, R. I., Korber, B. T. M., Graham, B. S., Hahn, B. H., Ho, D. D., Walker, B. D., Neuman, A. U., Vermund, S. H., Mestecky, J., Jackson, S., Fenamore, E., Cao, Y., Gao, F., Kalams, S., Kuntsman, K. J., McDonald, D., McWilliams, N., Trkola, A., Moore, J. P. & Wolinksky, S. M. (1998). Immunological and virological analyses of persons infected

McKeating, J. A., Shotton, C., Jeffs, S., Palmer, C., Hammond, A., Lewis, J., Oliver, K., May, J & Balfe, P. (1996). Immunogenicity of full-

length and truncated forms of the human immunodeficiency virus type I envelope glycoprotein. Immunology Letters 51, 101–105. Mascola, J. R., Lewis, M. G., Stiegler, G., Harris, D., VanCott, T. C., Hayes, D., Louder, M. L., Brown, C. R., Sapan, C. V., Frankel., S. S., Lu, Y., Robb, M. L., Katinger, H. & Birx, D. L. (1999). Protection of

macaques against pathogenic simian\human immunodeficiency virus 89.6PD by passive transfer of neutralizing antibodies. Journal of Virology 73, 4009–4018. Mascola, J. R., Stiegler, G, VanCott, T. C., Katinger, H., Carpenter, C. B., Hanson, C. E., Beary, H., Hayes, D., Frankel, S. S., Birx, D. L. & Lewis, M. G. (2000). Protection of macaques against vaginal trans-

mission of a pathogenic HIV-1\SIV chimeric virus by passive infusion of neutralizing antibodies. Nature Medicine 6, 207–210. Montefiori, D. C. & Evans, T. G. (1999). Towards an HIV type1 vaccine that generates potent, broadly cross-reactive neutralizing antibodies. AIDS Research and Human Retroviruses 15, 689–698. Moore, J. P., Sattentau, Q., Wyatt, R. & Sodroski, J. (1994). Probing the structure of the human immunodeficiency virus surface glycoprotein gp120 with a panel of monoclonal antibodies. Journal of Virology 68, 469–484. Moore, J. P., Trkola, A., Korber, B., Boots, L. J., Kessler, J., McCutchan,


CHDB

S. A. Jeffs and others F. E., Mascola, J., Ho, D. D., Robinson, J. & Conley, A. J. (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. Journal of Virology 69, 122–30. Muster, T., Steindl, F., Purtscher, M., Trkola, A., Klima, A., Himmler, G., Ruker, F. & Katinger, H. (1993). A conserved neutralizing epitope on

gp41 of human immunodeficiency virus type 1. Journal of Virology 67, 6642–6647. Nunberg, J. (2002). Retraction (letter). Science 296, 1025. Nyambi, P. N., Gorny, M. K., Bastiani, L., Van Der Groen, G., Williams, C. & Zolla-Pazner, S. (1998). Mapping of epitopes exposed on intact

human immunodeficiency virus type 1 (HIV-1) virions : a new strategy for studying the immunologic relatedness of HIV-1. Journal of Virology 72, 9384–9391. Pollard, S. R., Rosa, M. D., Rosa, J. J. & Wiley, D. C. (1992). Truncated variants of gp120 bind CD4 with high affinity and suggest a minimum CD4 binding region. EMBO Journal 11, 585–591. Reitter, J. N., Means, R. E. & Desrosiers, R. C. (1998). A role for carbohydrates in immune evasion in AIDS. Nature Medicine 4, 679–684. Roben, P., Moore, J. P., Thali, M., Sodroski, J. & Barbas, C. F., III (1994). Recognition properties of a panel of human recombinant Fab

fragments to the CD4 binding site of gp120 that show differing abilities to neutralize human immunodeficiency virus type 1. Journal of Virology 68, 4821–4828. Scarlatti, G. E., Albert, J., Rossi, P., Hodara, V., Biraghi, P., Muggiasca, L. & Fenyo, E.-M. (1993). Mother-to-child transmission of human

immunodeficiency virus type 1 – correlation with neutralising antibodies against primary isolates. Journal of Infectious Diseases 168, 207–210.

CHDC

Shotton, C., Arnold, C., Sattentau, Q., Sodroski, J. & McKeating, J. M. (1995). Identification and characterisation of monoclonal antibodies

specific for polymorphic antigenic determinants within the V2 region of the HIV-1 envelope glycoprotein. Journal of Virology 69, 222–230. Srivastava, I. K., Stamatatos, L., Legg, H., Kan, E., Fong, A., Coates, S. R., Leung, L., Wininger, M., Donnolly, J. L., Ulmer, J. B. & Barnett, S. W. (2002). Purification and characterization of oligomeric envelope

glycoprotein from a primary R5 subtype B human immunodeficiency virus. Journal of Virology 76, 2835–2847. Trkola, A., Purtscher, M., Muster, T., Ballaun, C., Buchacher, A., Sullivan, N., Srinivasan, K., Sodroski, J., Moore, J. P. & Katinger, H. (1996). Human monoclonal antibody 2G12 defines a distinctive

neutralization epitope on the gp120 glycoprotein of human immunodeficiency virus type 1. Journal of Virology 70, 1100–1108. Von Gegerfelt, A., Albert, J., Morfeldt-Manson, L., Broliden, K. & Fenyo, E.-M. (1991). Isolate-specific neutralizing antibodies in patients

with progressive HIV-1-related disease. Virology 185, 162–168. Wyatt, R., Sullivan, N., Thali, M., Repke, H., Ho, D., Robinson, J., Posner, M. & Sodroski, J. (1993). Functional and immunologic

characterization of human immunodeficiency virus type 1 envelope glycoproteins containing deletions of the major variable regions. Journal of Virology 67, 4557–4565. Yang, X., Florin, L., Farzan, M., Kolchinsky, P., Kwong, P. D., Sodroski, J. & Wyatt, R. (2000). Modifications that stabilize human immuno-

deficiency virus envelope glycoprotein trimers in solution. Journal of Virology 74, 4746–4754. Received 30 April 2002 ; Accepted 5 July 2002
