Downloaded from www.microbiologyresearch.org by IP: 50.16.49.105 On: Sun, 08 May 2016 06:29:27 Evolution of human immunodeficiency virus type 2 coreceptor usage, autologous neutralization, envelope sequence and glycosylation Yu Shi, 1 Eleonor Brandin, 1 Elzbieta Vincic, 2 Marianne Jansson, 2 Anders Blaxhult, 3 Katarina Gyllensten, 3 Lars Moberg, 4 Christina Brostro ¨ m, 3 Eva Maria Fenyo ¨ 2 and Jan Albert 1 Correspondence Jan Albert [email protected]1 Department of Virology, Swedish Institute for Infectious Disease Control and Microbiology and Tumorbiology Center, Karolinska Institutet, SE-171 82 Solna, Sweden 2 Unit of Virology, Division of Medical Microbiology, Department of Laboratory Medicine, Lund University, SE-223 62 Lund, Sweden 3 Department of Infectious Diseases/Solna, Karolinska University Hospital, Karolinska Institutet, SE-171 76 Stockholm, Sweden 4 Department of Infectious Diseases/Huddinge, Karolinska University Hospital, Karolinska Institutet, SE-141 86 Stockholm, Sweden Received 13 June 2005 Accepted 12 August 2005 To investigate why human immunodeficiency virus type 2 (HIV-2) is less virulent than HIV-1, the evolution of coreceptor usage, autologous neutralization, envelope sequence and glycosylation was studied in sequentially obtained virus isolates and sera from four HIV-2-infected individuals. Neutralization of primary HIV-2 isolates was tested by a cell line-based assay and IgG purified from patients’ sera. Significant autologous neutralization was observed for the majority (39 of 54) of the HIV-2 serum–virus combinations tested, indicating that neutralization escape is rare in HIV-2 infection. Furthermore, sera from 18 HIV-2 patients displayed extensive heterologous cross-neutralization when tested against a panel of six primary HIV-2 isolates. This indicates that HIV-2 is intrinsically more sensitive to antibody neutralization than HIV-1. In line with earlier reports, HIV-2 isolates could use several alternative receptors in addition to the major coreceptors CCR5 and CXCR4. Intrapatient evolution from CCR5 use to CXCR4 use was documented for the first time. Furthermore, CXCR4 use was linked to the immunological status of the patients. Thus, all CXCR4-using isolates, except one, were obtained from patients with CD4 counts below 200 cells ml ”1 . Sequence analysis revealed an association between coreceptor usage and charge of the V3 loop of the HIV-2 envelope, as well as an association between the rate of disease progression and the glycosylation pattern of the envelope protein. Furthermore, HIV-2 isolates had fewer glycosylation sites in the V3 domain than HIV-1 (two to three versus four to five). It is proposed here that HIV-2 has a more open and accessible V3 domain than HIV-1, due to differences in glycan packing, and that this may explain its broader coreceptor usage and greater sensitivity to neutralizing antibodies. INTRODUCTION The global human immunodeficiency virus (HIV) epidemic is dominated by HIV type 1 (HIV-1), but HIV-2 infections are also common in certain countries, such as Guinea-Bissau and Portugal (Reeves & Doms, 2002). There are many similarities, but also important differences, between HIV-1 and HIV-2. Thus, HIV-2 is less virulent than HIV-1 and is associated with slower rates of disease progression, mortality and transmission (Andreasson et al., 1993; Kanki et al., 1994; Marlink et al., 1994; Pepin et al., 1991; Reeves & Doms, 2002). Furthermore, HIV-2 infection is associated with lower plasma-virus levels than HIV-1 infection (Andersson, 2001; Andersson et al., 2000; Berry et al., 1998; Reeves & Doms, 2002). The reasons for the differences between HIV-1 and HIV-2 infection remain unclear, but several factors have been An amino acid sequence alignment of the V1, V2 and V3 domains of the HIV-2 envelope protein is available as supplementary material in JGV Online. The GenBank/EMBL/DDBJ accession numbers for the sequences determined in this study are DQ213026–DQ213040. 0008-1259 G 2005 SGM Printed in Great Britain 3385 Journal of General Virology (2005), 86, 3385–3396 DOI 10.1099/vir.0.81259-0
Transcript
Downloaded from www.microbiologyresearch.org by
IP: 50.16.49.105
On: Sun, 08 May 2016 06:29:27
Evolution of human immunodeficiency virus type 2coreceptor usage, autologous neutralization,envelope sequence and glycosylation
1Department of Virology, Swedish Institute for Infectious Disease Control and Microbiology andTumorbiology Center, Karolinska Institutet, SE-171 82 Solna, Sweden
2Unit of Virology, Division of Medical Microbiology, Department of Laboratory Medicine, LundUniversity, SE-223 62 Lund, Sweden
3Department of Infectious Diseases/Solna, Karolinska University Hospital, Karolinska Institutet,SE-171 76 Stockholm, Sweden
4Department of Infectious Diseases/Huddinge, Karolinska University Hospital, KarolinskaInstitutet, SE-141 86 Stockholm, Sweden
Received 13 June 2005
Accepted 12 August 2005
To investigate why human immunodeficiency virus type 2 (HIV-2) is less virulent than HIV-1, the
evolution of coreceptor usage, autologous neutralization, envelope sequence and glycosylation was
studied in sequentially obtained virus isolates and sera from four HIV-2-infected individuals.
Neutralization of primary HIV-2 isolates was tested by a cell line-based assay and IgG purified from
patients’ sera. Significant autologous neutralization was observed for the majority (39 of 54) of
the HIV-2 serum–virus combinations tested, indicating that neutralization escape is rare in
HIV-2 infection. Furthermore, sera from 18 HIV-2 patients displayed extensive heterologous
cross-neutralization when tested against a panel of six primary HIV-2 isolates. This indicates that
HIV-2 is intrinsically more sensitive to antibody neutralization than HIV-1. In line with earlier reports,
HIV-2 isolates could use several alternative receptors in addition to the major coreceptors
CCR5 and CXCR4. Intrapatient evolution from CCR5 use to CXCR4 use was documented for the
first time. Furthermore, CXCR4 use was linked to the immunological status of the patients.
Thus, all CXCR4-using isolates, except one, were obtained from patients with CD4 counts below
200 cells ml”1. Sequence analysis revealed an association between coreceptor usage and
charge of the V3 loop of the HIV-2 envelope, as well as an association between the rate of disease
progression and the glycosylation pattern of the envelope protein. Furthermore, HIV-2 isolates
had fewer glycosylation sites in the V3 domain than HIV-1 (two to three versus four to five). It is
proposed here that HIV-2 has a more open and accessible V3 domain than HIV-1, due to
differences in glycan packing, and that this may explain its broader coreceptor usage and greater
sensitivity to neutralizing antibodies.
INTRODUCTION
The global human immunodeficiency virus (HIV) epidemicis dominated by HIV type 1 (HIV-1), but HIV-2 infectionsare also common in certain countries, such as Guinea-Bissauand Portugal (Reeves & Doms, 2002). There are many
similarities, but also important differences, between HIV-1and HIV-2. Thus, HIV-2 is less virulent than HIV-1 and isassociated with slower rates of disease progression, mortalityand transmission (Andreasson et al., 1993; Kanki et al., 1994;Marlink et al., 1994; Pepin et al., 1991; Reeves & Doms,2002). Furthermore, HIV-2 infection is associated withlower plasma-virus levels than HIV-1 infection (Andersson,2001; Andersson et al., 2000; Berry et al., 1998; Reeves &Doms, 2002).
The reasons for the differences between HIV-1 and HIV-2infection remain unclear, but several factors have been
An amino acid sequence alignment of the V1, V2 and V3 domains ofthe HIV-2 envelope protein is available as supplementary material inJGV Online.
The GenBank/EMBL/DDBJ accession numbers for the sequencesdetermined in this study are DQ213026–DQ213040.
0008-1259 G 2005 SGM Printed in Great Britain 3385
Journal of General Virology (2005), 86, 3385–3396 DOI 10.1099/vir.0.81259-0
Downloaded from www.microbiologyresearch.org by
IP: 50.16.49.105
On: Sun, 08 May 2016 06:29:27
suggested to contribute. Better immune control of HIV-2than HIV-1 is an obvious possibility. Indeed, cross-sectionalstudies have indicated that autologous neutralizing-antibody responses are more common in HIV-2 infectionthan in HIV-1 infection (Bjorling et al., 1993; Fenyo &Putkonen, 1996; Tamalet et al., 1995). Differences have alsobeen reported in cytotoxic T-lymphocyte (CTL) responses,levels of apoptosis, production of b-chemokines and cap-ability of CD4+ and CD8+ lymphocytes to produce inter-leukins 2 and 4, as well as gamma interferon (Andersson,2001; Reeves & Doms, 2002). However, a recent studyreported no differences in cellular immune responsesbetween HIV-1- and HIV-2-infected Gambian patients(Jaye et al., 2004).
Prior to this study, there have been no studies on theevolution of autologous neutralizing-antibody responses inHIV-2 infection. However, for HIV-1, it has been shownthat escape from neutralization by antibodies is frequentand rapid in early infection (Albert et al., 1990; Richmanet al., 2003; von Gegerfelt et al., 1991; Wei et al., 2003) andthat autologous neutralizing-antibody responses are fre-quently low or absent later in infection (Ariyoshi et al., 1992;Fenyo & Putkonen, 1996; Homsy et al., 1990; Scarlatti et al.,1993; von Gegerfelt et al., 1991). Similarly, there is verylimited information on heterologous neutralization of pri-mary HIV-2 isolates, whereas it is known that heterologousneutralization of primary HIV-1 isolates is often absent orof low titre, except in some long-term non-progressors(Carotenuto et al., 1998; Dreyer et al., 1999; Scarlatti et al.,1993; Weber et al., 1996).
For effective infection of target cells, most HIV-1 and HIV-2isolates require binding to a chemokine coreceptor, in addi-tion to the CD4 receptor. The CCR5 receptor is used by themajority of primary HIV-1 isolates (R5 viruses), but someisolates from patients with more advanced immunodefici-ency use CXCR4 instead of (X4 viruses) or in addition to(X4R5 viruses) CCR5 (Asjo et al., 1986; Bjorndal et al., 1997;Tersmette et al., 1988; Zhang et al., 1996). The presence ofX4 virus, as well as a switch from CCR5 to CXCR4 usage,is associated with accelerated rate of disease progression(Connor et al., 1997; Koot et al., 1993). CCR5 and CXCR4are the major coreceptors for HIV-1, but a minor pro-portion of primary HIV-1 isolates can also utilize otheralternative coreceptors (CCR1, CCR2, CCR3, CXCR6 orBOB) in vitro (Bjorndal et al., 1997; Rucker et al., 1997). Incontrast to HIV-1, HIV-2 isolates can frequently use alter-native coreceptors in vitro (Blaak et al., 2005;McKnight et al.,1998; Morner et al., 1999b). However, CCR5 and CXCR4also appear to be the major coreceptors for HIV-2 (Blaaket al., 2005; Morner et al., 2002). Some primary HIV-2isolates can infect coreceptor-positive cells in the absence ofCD4 (Clapham et al., 1992; Reeves et al., 1999) and suchisolates are highly sensitive to neutralizing antibodies(Thomas et al., 2003).
In this study, we provide the first data on evolution ofautologous neutralizing-antibody responses and coreceptor
usage of viruses isolated sequentially from four HIV-2-infected individuals. Moreover, we examined the env genesequences of 15 HIV-2 isolates. We propose that the HIV-2V3 domain has a more open and accessible configurationthan that of HIV-1. This may explain the higher sensitivityto neutralizing antibodies and the broader use of alternativecoreceptors.
METHODS
Study subjects and virus isolates. Four HIV-2-infected Swedishindividuals were selected from a cohort of approximately 20 SwedishHIV-2 patients (Brandin et al., 2003) based on availability ofsequential virus isolates and serum samples. All four patients werefemale immigrants from different West African countries. The treat-ment history and HIV-2 pol gene evolution in three of the fourpatients has been described in detail elsewhere (Brandin et al.,2003). Basic clinical, immunological and virological characteristicsare shown in Table 1. Plasma HIV-2 RNA levels were determined byusing an experimental assay that was kindly provided by KarenYoung at Roche Molecular Systems, Alameda, CA, USA. HIV-2 wasisolated by cocultivation of peripheral blood mononuclear cells(PBMCs) from the patients and phytohaemagglutinin-activatedblood-donor PBMCs as described previously (Albert et al., 1990,1996).
Thirteen additional sera from 11 other Swedish HIV-2-infectedpatients were used for heterologous-neutralization experiments(Brandin et al., 2003). Four of these patients were infected withHIV-2 of genetic subtype A, whereas the genetic subtype was unknownfor the other patients. The heterologous-neutralization experimentsincluded one isolate from each of the four principal study subjects, aswell as the primary HIV-2 isolates 6669 (subtype A) and 1653 (subtypeB) (Albert et al., 1987, 1996).
The study was approved by the medical ethics committee of theKarolinska Institute (nos 96-189 and 99-462).
Virus stocks and coreceptor usage. Virus stocks were preparedas described previously (Shi et al., 2002). The infectivity of the HIV-2 isolates was determined by triplicate titration on U87.CD4 cellsexpressing CXCR4 or CCR5 (Shi et al., 2002). The coreceptor usageof the HIV-2 isolates was determined by using U87.CD4 cellsexpressing CCR1, CCR2, CCR3, CCR5 or CXCR4 and GHOST(3)cells expressing CCR3, CCR5, CXCR4, CXCR6 or BOB (Karlssonet al., 2003). Parental U87.CD4 and GHOST(3) cells without co-receptors were also included. The GHOST(3) cells were tested withand without the CXCR4 antagonist AMD3100. Presence of manysyncytia and high HIV-2 p27 antigen levels were scored as strongusage of a specific coreceptor. Presence of low HIV-2 p27 levels inthe absence of syncytia was scored as weak coreceptor usage.Absence of both syncytia and HIV-2 p27 antigen was scored as nousage of a specific coreceptor. All tests were repeated at least twice.HIV-2 p27 antigen levels were tested by using Murex HIVag mAb(Murex Biotech).
Neutralization assay. The levels of autologous neutralizing anti-bodies were determined by using a recently developed assay basedon plaque formation in U87.CD4 cells (Shi et al., 2002). Briefly, thevirus stocks were diluted to contain a final concentration of 20–50 p.f.u. per well. In all neutralization experiments, we used IgGthat had been purified from patients’ sera to avoid false-positiveneutralization due to the presence of antiretroviral drugs. IgG waspurified by using protein G–sepharose 4 Fast Flow (PharmaciaBiotech) according to the instructions of the manufacturers andquantified by using an in-house ELISA. The purified IgG was used
3386 Journal of General Virology 86
Y. Shi and others
Downloaded from www.microbiologyresearch.org by
IP: 50.16.49.105
On: Sun, 08 May 2016 06:29:27
at a concentration that corresponded to a 1 : 30 dilution of theoriginal serum. Virus–IgG mixtures were kept at 37 uC for 1 h andthen further diluted in two to three fivefold-dilution steps. Eachvirus–IgG dilution (200 ml) was distributed into triplicate wells (orduplicate wells when the volume of serum was limited) containingthe U87.CD4-CXCR4 or U87.CD4-CCR5 cells. Results are presentedas means of parallel determinations. Positive virus controls consistedof wells with cells and virus, but no serum; negative virus controlsconsisted of wells with virus only; cell controls consisted of wellswith cells only. The experiment was terminated on day 3 or day 4 byfixation with methanol : acetone (1 : 1). The number of p.f.u. wasdetermined following haematoxylin staining. The neutralizing capa-city of the serum was calculated by the formula [12(p.f.u. withserum/p.f.u. without serum) 6100] and thus expresses the degree ofreduction in p.f.u. in the presence of serum relative to wells withoutserum. The cut-off for neutralization for this assay has been derivedstatistically and was determined to be a 30% reduction in thenumber of plaques. The assay qualitatively compares well with tradi-tional PBMC-based HIV-neutralization assays and has good repro-ducibility (Shi et al., 2002). However, the percentage plaquereduction in our assay does not translate directly into traditionalneutralizing titres and therefore we focused on qualitative, ratherthan quantitative, aspects of the neutralization results.
HIV-2 env gene sequencing. HIV-2 RNA was extracted from200 ml virus-infected supernatant from PBMC cultures by using aNucliSense Isolation kit (Organon Teknika) according to the manu-facturer’s recommendations. A 1588 bp fragment encompassing theentire gp125 coding sequence was amplified by nested RT-PCR.
The PCR was carried out in a 50 ml reaction mixture containing
16 PCR buffer (Applied Biosystems), 2?5 mM MgCl2 (Applied
Biosystems), 50 mM each dNTP (Amersham Pharmacia), 0?1 mM
each primer (Table 2), 1 U AmpliTaq DNA polymerase (Applied
Biosystems), 5 U reverse transcriptase (M-MuLV; Roche Diagnostics)
and 28 U rRNasin RNase inhibitor (Promega). The RT-PCR profile
consisted of reverse transcription at 37 uC for 60 min, denaturation
at 92 uC for 5 min, 30 amplification cycles of 20 s at 92 uC, 20 s at
55 uC and 1 min at 72 uC, and final elongation at 72 uC for 5 min. A
portion of the PCR product (2?5 ml) was used as the template for
nested PCR; the mixture for the nested PCR contained the same
reagents as the mixture for the first PCR, except for reverse tran-
scriptase and RNasin. The amplification profile of the nested PCR
was identical to that the first PCR, except that the initial reverse-
transcription step at 37 uC was omitted.
The PCR amplicons were purified with a QIAquick PCR purification
kit (Qiagen). Sequencing reactions were carried out with an ABI Prism
for these phylogenetic trees were obtained from the Los Alamos
National Laboratory HIV sequence database (www.hiv.lanl.gov).
Potential N-linked glycosylation sites were identified by using the N-
GLYCOSITE software available at the website of the Los Alamos National
Laboratory HIV sequence database. The isolates were not clonal in
sequence and some of the intrasample sequence polymorphisms
involved potential N-linked glycosylation sites. In our calculations of
the number of potential N-linked glycosylation sites, we have not
distinguished between glycosylation sites that were present in part or
all of the virus population. Changes in glycosylation sites over time
were calculated by comparing each sequence to the closest previous
sequence from the same patient. Glycosylation sites were considered to
have moved (counted as one event) if a glycosylation site was lost and
gained within three amino acid positions, whereas losses and gains
of glycosylation sites at longer distances were counted as two
events.
Statistical analyses. Statistical analyses were complicated by the
fact that observations on our 15 HIV-2 isolates could not be consid-
ered to be independent because the isolates were obtained sequen-
tially from four individuals. For this reason, we deliberately avoided
extensive statistical testing. When statistical analyses were performed,
we used the Statistica v6.1 software (Statsoft Inc.).
RESULTS
Characteristics of the study subjects
The study involved 15HIV-2 isolates that had been obtainedsequentially from four HIV-2-infected Swedish individualsover periods that ranged from 4 to 11 years (Table 1).Patients 1 and 2 had moderately progressive HIV-2 disease,moderately suppressed CD4+ T-lymphocyte (CD4) countsand no clinical symptoms. In contrast, patient 4 and, inparticular, patient 8 had advanced immunodeficiency,severe clinical symptoms and high HIV-2 RNA levels inplasma, despite combination antiretroviral therapy. Patient1 had not received antiretroviral treatment, whereas patients2, 4 and 8 underwent therapy with some interruptions.
HIV-2 coreceptor usage associated with thedegree of immunodeficiency
The coreceptor usage of the HIV-2 isolates was tested onU87.CD4 and GHOST(3) cells expressing different chemo-kine coreceptors (Table 1). In accordance with earlierstudies of HIV-2 coreceptor usage, most isolates could useseveral alternative coreceptors (CCR1, CCR2, CCR3,CXCR6 and BOB) in addition to one or both of the twomajor coreceptors (CCR5 and CXCR4). However, noneof the isolates replicated in the parental U87.CD4 orGHOST(3) cells that lacked coreceptors.
Table 2. PCR and sequencing primers used for amplification and sequencing of the HIV-2gp125 region
Y in primer JA255 denotes a position that was synthesized with a C/T wobble.
Primer Position* Direction Composition
Outer PCR primers
JA207 6021–6041 Sense 59-CTGTTACCATTGCCAGCTGTGTT-39
DInner PCR primers were also used as sequencing primers.
dAdditional sequencing primers were used for a few samples that were difficult to sequence through the
V1/V2 region.
3388 Journal of General Virology 86
Y. Shi and others
Downloaded from www.microbiologyresearch.org by
IP: 50.16.49.105
On: Sun, 08 May 2016 06:29:27
In two patients (patients 1 and 4), we observed an acquisi-tion of the ability to use the CXCR4 coreceptor. Interest-ingly, the ability of the isolates to infect through CXCR4appeared to be associated with the immunological status ofthe patients. Thus, all isolates, except one, that could infectthrough the CXCR4 receptor were obtained at time pointswhen the CD4 counts of the respective patients were below200 cells ml21. Conversely, R5 isolates were obtained at timepoints when the patients had higher CD4 counts.
Autologous-neutralization escape is rare inHIV-2 infection
The evolution of autologous neutralization was tested in achequerboard fashion with purified IgG from sequentiallycollected sera by using a recently described method basedon plaque formation in U87.CD4 cells (Shi et al., 2002).Significant autologous neutralization (>30% plaquereduction) was observed for the majority (39 of 54) of thetested serum–virus combinations (Fig. 1). Importantly, weobserved no consistent pattern of neutralization escape inour four study subjects. Thus, many sera could neutralizeautologous viruses that had been isolated many years later.
The clinical and immunological status of the patients wasassociated with the ability of their sera to neutralize auto-logous virus. Thus, patient 8, who had the most advanceddisease, showed the weakest neutralization. However, excep-tions to this pattern also existed, e.g. the 2001 isolate frompatient 4, which was sensitive to neutralization even thoughthe CD4 counts were low at the time of virus isolation. Weobserved no clear association between neutralization andcoreceptor usage. Thus, some X4-using viruses were difficultto neutralize, e.g. the isolates from patient 8, whereas otherswere sensitive to neutralization, e.g. the 2000 and 2001
isolates from patient 4. However, most R5 isolates weresensitive to autologous neutralization.
Broad neutralization of heterologous primaryHIV-2 isolates
To investigate whether HIV-2 sera also can neutralizeheterologous primary HIV-2 isolates, we tested the ability ofIgG purified from 18HIV-2 sera from 15 Swedish patients toneutralize six heterologous HIV-2 isolates (Table 3). Theisolates consisted of one isolate from each of our four studysubjects, the reference HIV-2 subtype A isolate 6669 (Albertet al., 1987, 1996) and the HIV-2 subtype B isolate 1653(Albert et al., 1996). The main observation was that signifi-cant neutralization was observed for the majority of thevirus–IgG combinations, i.e. 84 of 97 tests (87%). Further-more, there was no obvious association between the neutra-lization results and the CD4 count of the patient fromwhomthe serum was drawn. Thus, serum P4-2001, which wasobtained when the patient had a CD4 count of 30 cells ml21,could neutralize all five heterologous HIV-2 primary iso-lates. Similarly, there was no obvious association betweenthe neutralization results and the coreceptor usage of thevirus. A few interesting details from the heterologous-neutralization experiments should be noted. The X4 isolateP4-V2000 was comparably resistant to heterologous neutra-lization, as significant neutralization was observed with onlysix of 15 sera, even though this isolate was neutralized by allfour autologous sera (Fig. 1). Furthermore, virus P8-V1998,which was neutralized poorly by autologous sera (Fig. 1),was neutralized efficiently by all heterologous HIV-2 sera.This indicates that the poor autologous neutralization thatwe observed in patient 8 was due to the immunologicalstatus of this patient, rather than the characteristics of thevirus. Finally, isolate 1653, which is of HIV-2 subtype B, wasneutralized by 13 of 15 sera, even though they, when known,
Fig. 1. Autologous HIV-2 neutralization. Virusisolates (V) and sera (S) are named by time ofsampling. The assay was based on reductionof plaque formation in U87.CD4-CCR5 orU87.CD4-CXCR4 cells; R5 viruses (seeTable 1) were tested on U87.CD4-CCR5cells and X4 and X4R5 viruses (see Table 1)were tested on U87.CD4-CXCR4 cells. Thecut-off for neutralization was plaque reductionby 30% and is indicated by a dashed line.
http://vir.sgmjournals.org 3389
HIV-2 evolution
Downloaded from www.microbiologyresearch.org by
IP: 50.16.49.105
On: Sun, 08 May 2016 06:29:27
were obtained from patients infected with subtype A virus(Brandin et al., 2003). Thus, the ability of HIV-2 sera tocross-neutralize heterologous HIV-2 isolates appears toextend to other HIV-2 subtypes.
V1/V2 length is associated with clinicalprogression
The intrapatient evolution of the HIV-2 envelope genewas investigated by direct sequencing of the virus isolates(see Supplementary Figure, available in JGV Online). Thesesequences represent bulk population sequences of the virusisolates, rather than sequences of individual viral clones.Phylogenetic-tree analysis showed that all of our newlygenerated HIV-2 env sequences belonged to subtype A ofHIV-2 (data not shown). Furthermore, the sequences showeda patient-specific clustering, indicating that the sequenceswere authentic and that no PCR contamination or sampleconfusion had occurred. The HIV-2 env gene sequencesdisplayed no stop codons or other clearly inactivatingrearrangements. The lengths of the translated HIV-2 SUproteins varied between 492 and 522 aa (Table 4). Lengthvariations were concentrated in the V1 and V2 regions(Fig. 2).
The length of the V1/V2 region of the SU protein wasassociated with the rate of disease progression andimmunological status of the patients. Thus, the V1/V2region increased in length over time in patients 1 and 2, whodisplayed slower disease progression and less advancedimmunodeficiency. The most striking example was seen in
patient 1, where the virus acquired 16 new amino acids inthe V1 region between 1986 and 1991, mainly as a result of
Table 3. Heterologous neutralization of six primary HIV-2 isolates
ND, Not done; results shown in bold represent neutralization below the cut-off of 30% plaque reduction.
Serum Subtype CD4 count
(cells ml”1)
Virus isolate [neutralization (%)]
P1-V1997
(R5)
P2-V1996
(R5)
P4-V2000
(X4)
P8-V1998
(X4)
6669
(X4)
1653
(R5)
P1-1986 A 650 ND 47 40 67 58 ND
P2-1997 A 317 53 ND 4 54 51 45
P3 A 1170 61 40 27 70 76 54
P4-2000 A 140 ND 35 ND 65 57 ND
P4-2001 A 30 49 53 ND 63 36 49
P7-1994 ND 655 69 69 67 85 97 83
P7-2003 ND 660 66 87 81 90 100 52
P8-2002 A 315 33 38 17 ND 55 54
P9 ND 1440 45 50 53 65 57 62
P10-1998 A 250 ND 55 ND 56 55 ND
P10-2004 A 380 43 31 26 54 43 71
P12 ND 870 23 41 34 44 32 48
P14 ND 891 29 41 16 58 38 42
P16 A 588 53 52 22 63 44 25
P21 ND 425 37 49 40 53 42 51
P23 ND 705 32 57 37 61 53 31
P24 A 170 50 59 27 64 74 63
P26 ND 140 39 56 16 49 39 28
Table 4. Characteristics of the translated gp125 (SU) pro-teins of the HIV-2 isolates
Isolate Length of
SU protein
(aa)
No. potential
glycosylation
sites
No. changes
of potential
glycosylation
sites*
P1-1986 508 24 NA
P1-1991 527 26 12
P1-1997 532 25 11
P2-1994 527 26 NA
P2-1996 525 29 7
P2-1998 536 27 9
P2-2002 526 28 9
P4-1992 509 26 NA
P4-2000 507 26 8
P4-2001 505 27 3
P4-2002 502 25 3
P8-1998 510 24 NA
P8-2001 March 511 25 2
P8-2001 June 511 27 2
P8-2002 512 27 0
NA, Not applicable.
*Calculated from Supplementary Figure (available in JGV Online) as
described in Methods.
3390 Journal of General Virology 86
Y. Shi and others
Downloaded from www.microbiologyresearch.org by
IP: 50.16.49.105
On: Sun, 08 May 2016 06:29:27
introduction of threonine residues. In contrast, the length ofthe V1/V2 region was stable over time in patients 4 and 8,who had more advanced HIV-2 disease. There was no clearassociation between V1/V2 length and chemokine corecep-tor usage or neutralization sensitivity.
V3 sequence is associated with coreceptor usage
We found an association between the overall charge of theHIV-2 V3 loop and coreceptor usage (Fig. 2). All isolatesthat could use CCR5 efficiently had V3 loops with a netcharge of+5 or+6, whereas all isolates that were unable touse this receptor efficiently had a charge of +7. A reversedependence was observed between V3 charge and CXCR4usage. Thus, all isolates that could use CXCR4 efficiently hadV3 loop charges of+7, except for the dual-tropic isolate P4-2001, which had a charge of+6. Isolates that were unable touse CXCR4 efficiently had V3 loop charges of +5 or +6.
We also found that all six isolates that lacked the ability toinfect efficiently through the CCR5 receptor displayedpositively charged amino acids (arginine or lysine) atposition 323 in the V3 loop (corresponding to position 19in Fig. 2), whereas all nine isolates that could use CCR5displayed neutral amino acids (valine or isoleucine) at thisposition. Furthermore, positively charged amino acids atposition 323 characterized six of seven isolates that couldinfect efficiently through the CXCR4 receptor, but none ofthe CXCR4-negative isolates.
Changes in overall glycosylation patternassociated with disease progression
Recent data indicate that neutralization escape in HIV-1infection may involve changes in glycosylation pattern (Weiet al., 2003). N-linked glycosylation may also influencecoreceptor usage in HIV-1 (Nabatov et al., 2004; Ogert et al.,
Fig. 2. Amino acid sequence of the V1, V2and V3 domains of the HIV-2 envelope pro-tein of sequential HIV-2 isolates from thefour study subjects (P1, P2, P4 and P8). Allsequences were obtained from PBMC-passaged stocks of primary HIV-2 isolates.Dashes indicate identity with the P1-1986isolate, letters represent differences relativeto the P1-1986 isolate and dots indicategaps introduced to align the sequences. X,Polymorphic amino acid. Potential N-linkedglycosylation sites are shaded in grey. Theboxed area in the V3 panel denotes position19, at which the presence of positivelycharged amino acids (K, lysine; R, arginine)correlated with X4 coreceptor usage. Thecharge of the V3 loop (which spans posi-tions 1–35) was calculated according to theformula {[(no. arginine and lysine residues)61]+[(no. glutamic acid and aspartic acidacid residues) 6”1]}.
http://vir.sgmjournals.org 3391
HIV-2 evolution
Downloaded from www.microbiologyresearch.org by
IP: 50.16.49.105
On: Sun, 08 May 2016 06:29:27
2001; Pollakis et al., 2001; Polzer et al., 2002). For this reason,we examined changes in potential N-linked glycosylationsites in our HIV-2 isolates in relation to neutralization,coreceptor usage and clinical progression.
The number of potential N-linked glycosylation sites variedbetween 24 and 29 (Table 4) and there was a tendency forthe glycosylation sites to increase in number over time.Furthermore, viruses with longer V1/V2 regions tended tohave a higher number of potential N-linked glycosylationsites.
We also examined changes in the positioning and presenceof glycosylation sites over time (Table 4; SupplementaryFigure). There were substantial differences in the number ofchanges of N-linked glycosylation sites between isolates andpatients. The highest number of changes was observed in the1991 isolate of patient 1 (compared with the preceding 1986isolate). The lowest number, i.e. no change at all, wasobserved for the March 2001 sample and the 2002 samplefrom patient 8.
The number of changes in N-linked glycosylation sites waslinked to the clinical and immunological status of thepatients. Thus, the highest number of changes was observedin patients 1 and 2, who had moderate immunodeficiency,an intermediate number was observed in patient 4 and thelowest number was observed in patient 8, who had the mostadvanced immunodeficiency.
Glycosylation of the V3 region differed betweenHIV-1 and HIV-2
The glycosylation pattern of our HIV-2 isolates was alsoinspected for possible correlation with coreceptor usage andneutralization. One glycosylation site just downstream ofthe HIV-2 V3 loop, i.e. at position 40 in Fig. 2, appeared tobe associated with the coreceptor usage of our HIV-2isolates. However, this association was probably coinciden-tal, because it does not hold true for previously publishedHIV-2 isolates with known env sequence and coreceptorusage (data not shown). Similarly, the neutralizationsensitivity of our isolates did not correlate with the overallnumber of N-linked glycosylation sites or the absence orpresence of specific glycosylation sites. The most importantfinding was that the HIV-2 isolates had fewer potentialglycosylation sites in the V3 domain than have beenreported for HIV-1, i.e. two to three for HIV-2 versus four tofive for HIV-1 (Figs 2 and 3). This is interesting because theglycosylation pattern in the HIV-1 V3 domain has beenreported to influence the neutralization sensitivity andcoreceptor usage of HIV-1 isolates (McCaffrey et al., 2004;Nabatov et al., 2004; Polzer et al., 2002; Schønning et al.,1996).
DISCUSSION
In this study, we have provided the first data on theevolution of autologous neutralizing-antibody responses,
coreceptor usage and env gene sequence in HIV-2 infec-tion. We show that neutralization escape is rare in HIV-2infection and that HIV-2 sera are capable of broadlyneutralizing heterologous primary HIV-2 isolates. Further-more, we have documented for the first time that HIV-2 canswitch within the same infected individual from CCR5 toCXCR4 coreceptor usage and that the coreceptor usage ofHIV-2 appears to be linked intimately to the immunologicalstatus of the patients. Finally, we have found that the chargeof the HIV-2 V3 loop appears to determine the coreceptorusage of HIV-2.
Our study is the first investigation of the evolution ofautologous neutralizing-antibody responses in HIV-2 infec-tion. The main finding is that neutralization escape appearsrare in HIV-2 infection, even though some primary HIV-2isolates are more difficult to neutralize than others. This is incontrast to HIV-1 infection, where neutralization escape iscommon and patients’ sera are rarely capable of neutralizingcontemporaneous autologous virus isolates (Albert et al.,1990; Ariyoshi et al., 1992; Fenyo & Putkonen, 1996; Homsyet al., 1990; Richman et al., 2003; Scarlatti et al., 1993; vonGegerfelt et al., 1991;Wahlberg et al., 1991; Wei et al., 2003).However, it should be pointed out that our patients werenot followed from the time of infection, thus it is possiblethat neutralization escape early after infection may havebeen missed. Furthermore, sera from HIV-2-infectedpatients with varying severity of disease were capable ofcross-neutralizing a panel of six heterologous primary HIV-2 isolates. This also differs from the findings in HIV-1infection (Carotenuto et al., 1998; Dreyer et al., 1999;Scarlatti et al., 1993; Weber et al., 1996). Dreyer et al. (1999)found that antiretroviral drugs contributed to neutralizationin some sera. Such an influence from antiretroviral therapycan be ruled out in our study, because we used IgG that hadbeen purified from the sera of the HIV-2-infected indivi-duals. In our study, all HIV-2 sera neutralized the majorityof the six heterologous primary HIV-2 isolates, includingthe HIV-2 subtype B isolate 1653. Some HIV-2 sera wereeven able to cross-neutralize HIV-1 (Weiss et al., 1988).Taken together, our and previous studies show that there arefundamental differences between HIV-1 and HIV-2 in theinduction of and sensitivity to neutralizing antibodies.Natural HIV-2 infection, but not HIV-1 infection, appearsto induce broadly neutralizing antibody responses. Further-more, HIV-2 appears intrinsically less able than HIV-1 toevade these neutralizing-antibody responses. Whether theseimportant differences are a cause or a consequence of thelower virulence of HIV-2 remains to be elucidated.
In this study, we have shown for the first time that HIV-2can evolve fromCCR5use toCXCR4use in infected patients.The acquisition of CXCR4 usage was linked closely to theimmunological status of the patients. Thus, all CXCR4-using isolates, except one, were isolated when the CD4counts of the patients were lower than 200 cells ml21. This isin agreement with a large number of studies on HIV-1,which have shown that CXCR4-using viruses are rare in
3392 Journal of General Virology 86
Y. Shi and others
Downloaded from www.microbiologyresearch.org by
IP: 50.16.49.105
On: Sun, 08 May 2016 06:29:27
early HIV-1 infection, but can be isolated from approxi-mately 50% of patients with AIDS (Asjo et al., 1986;Bjorndal et al., 1997; Tersmette et al., 1988; Zhang et al.,1996). In HIV-1 infection, the emergence of CXCR4-usingvirus variants is associated directly with an accelerated rateof decline of CD4 counts (Connor et al., 1997; Koot et al.,1993). It is possible that the same is true for HIV-2 infection,but larger studies are needed to explore this.
Our study shows that many primary HIV-2 isolates can useone or several alternative coreceptors (CCR1, CCR2, CCR3,CXCR6 or BOB) in addition to CCR5 and/or CXCR4. Thisconfirms the findings of earlier studies (Blaak et al., 2005;McKnight et al., 1998; Morner et al., 1999b). We havechosen to refer to these receptors as ‘alternative coreceptors’,because it is unclear whether they are utilized in vivo (Blaaket al., 2005; Morner et al., 2002). Some HIV-2 isolatesdisplayed an R5X4 phenotype, but we did not investigatewhether individual viral clones were dually tropic orwhether the isolates contained a mixture of R5 and X4clones.
We found that HIV-2, like HIV-1 (Fouchier et al., 1992;Nabatov et al., 2004; Pollakis et al., 2001), displays anassociation between the charge of the V3 loop and co-receptor use. Thus, all of our HIV-2 isolates with low V3charge (+5 or+6) used CCR5, whereas isolates with higherV3 charge (+7) preferred CXCR4. These results are inagreement with our earlier finding that rapid/high HIV-2isolates had a higher V3 charge than slow/lowHIV-2 isolates(Albert et al., 1996), as well as with the results of Isaka et al.(1999), who also showed that theHIV-2 V3 domain containsdeterminants for coreceptor use, as exchange of the C-terminal half of the V3 loop between the laboratory HIV-2strains ROD and GH-1 altered the coreceptor use recipro-cally. At present, it is unclear whether the exact positioningof positively charged amino acids in the HIV-2 V3 loop alsoinfluences the coreceptor usage. However, we saw a prefer-ence for positively charged amino acids (lysine or arginine)at position 19 in the V3 loop in X4 HIV-2 isolates, similar tothe preference for positively charged amino acids at posi-tions 11 and 25 in X4 HIV-1 isolates (De Jong et al., 1992;Fouchier et al., 1992).
We believe that the high sensitivity to neutralizing anti-bodies, the inability to escape neutralization and broadcoreceptor use of HIV-2 may be due to differences in thestructure of the V3 domain between HIV-1 and HIV-2. Asdescribed above, the V3 domain is a determinant forcoreceptor use of both HIV-1 and HIV-2. Furthermore, theV3 domain of both HIV-1 and HIV-2 contains neutralizingepitopes (Goudsmit et al., 1988; Javaherian et al., 1989;McKnight et al., 1996; Morner et al., 1999a; Rusche et al.,1988). We observed that HIV-2 isolates only have two or, ina few cases, three potential N-linked glycosylation sites inand around the V3 loop, whereas HIV-1 isolates have four orfive (Fig. 3). In HIV-1, these glycosylation sites appear tobe utilized (Ogert et al., 2001; Polzer et al., 2002), but it isunknown whether the same is true for HIV-2. We propose
that these differences in glycan packing confer a more openand accessible V3 domain on HIV-2 compared with HIV-1.The more open envelope configuration may explain thebroader coreceptor usage and greater sensitivity to neu-tralizing antibodies of HIV-2. In support of this hypothesis,it has been shown for HIV-1 that deglycosylation of the V3loop may lead to a broadening of the coreceptor repertoireto include CCR3 (Pollakis et al., 2001). Furthermore, severalstudies have shown that removal of glycans in and aroundthe HIV-1 V3 loop may increase the sensitivity to neutraliz-ing antibodies (Benjouad et al., 1992; McCaffrey et al., 2004;Nabatov et al., 2004; Polzer et al., 2002; Schønning et al.,1996). An additional interesting observation is that HIV-2isolates that can infect coreceptor-positive cells in theabsence of CD4 are highly sensitive to neutralizing anti-bodies (Clapham et al., 1992; Thomas et al., 2003).
We studied four HIV-2-infected patients who had moderateto advanced HIV-2 disease. Clearly, it would have beeninteresting to also study patients with fully asymptomaticHIV-2 infection, but such patients from our cohort could
Fig. 3. Schematic figure showing the reduced glycan shieldingand greater accessibility of the HIV-2 V3 domain comparedwith the HIV-1 V3 domain. The HIV-1 figure is based on theconsensus sequence of HIV-1 subtype B and the HIV-2 figureis based on the consensus sequence of HIV-2 subtype A(www.hiv.lanl.gov). Numbering is according to the HIV-1HxB2and HIV-2ROD sequences (GenBank accession nos K03455and M15390, respectively).
http://vir.sgmjournals.org 3393
HIV-2 evolution
Downloaded from www.microbiologyresearch.org by
IP: 50.16.49.105
On: Sun, 08 May 2016 06:29:27
not be included in this study because attempts to isolate viruswere generally unsuccessful. This difficulty in isolating virusfrom asymptomatic HIV-2 carriers is in agreement withresults from other researchers (Simon et al., 1993). However,it has been reported that HIV-2 may be isolated morefrequently if CD8 cells are depleted from the cultures. Recentstudies on neutralization of HIV-1 have utilized plasma-derived SU proteins in recombinant virus-neutralizationassays (Richman et al., 2003; Wei et al., 2003). We aredeveloping such an assay for HIV-2, but we will still beunable to study asymptomatic HIV-2 carriers because theytypically have undetectable plasma HIV-2 levels (Anderssonet al., 2000; Berry et al., 1998; Brandin et al., 2003; Popperet al., 1999).
We noted that the immunological status of the patients wasassociated with the length of the V1/V2 domain and changesin glycosylation pattern. Thus, the virus carried by patients 1and 2, who had relatively high CD4 counts, displayed amarked elongation of the V1/V2 domain and a high numberof glycosylation changes. In contrast, patient 8, who hadadvanced immunodeficiency, displayed a stable V1/V2 lengthand few changes in glycosylation pattern. Interestingly, theelongation of the V1/V2 domain was primarily due to inser-tions of threonine residues and thereby increased the numberof potential O-linked glycosylation sites. Similar changesin the V1/V2 domain correlate with neutralizing-antibodyresponses and neutralization escape in experimental simianimmunodeficiency virus (SIVsm and SIVmne) infection(Chackerian et al., 1997; Rybarczyk et al., 2004). Both ofthese viruses are members of the HIV-2/SIVsm familyand thus are related closely to HIV-2. It is known that theV1/V2 domain of HIV-2 also contains neutralizing epitopes(McKnight et al., 1996). Many studies show that neutraliza-tion escape in HIV-1 and SIV frequently involves changes inglycosylation pattern (Benjouad et al., 1992; Chackerianet al., 1997; Derdeyn et al., 2004; McCaffrey et al., 2004;Nabatov et al., 2004; Polzer et al., 2002; Schønning et al.,1996; Wei et al., 2003). Thus, it is possible that the changesthat we observed in the V1/V2 region of HIV-2 could beimmunologically driven, but this possibility appears to becontradicted by the fact that we did not observe neutraliza-tion escape. However, it is possible that the putativeneutralizing epitopes in the HIV-2 V1/V2 region resemblesubdominant CTL epitopes (Goulder et al., 2001), so thatmutations in these epitopes do not lead to completeneutralization escape.
The hypothesis that we present concerning the mechanismsfor evolution and interdependence between HIV-2 neu-tralization escape, coreceptor usage and glycosylationpattern is testable. Thus, we aim to produce infectiousclones of HIV-2 with different biological characteristics.Through mutagenesis and construction of HIV-2 chimeras,we will try to verify formally whether the HIV-2 V3 domainhas a more open configuration than that of HIV-1, and alsoto identify the molecular determinants for the biologicalproperties listed above.
ACKNOWLEDGEMENTS
This work was supported by grants from the Swedish Physicians against
AIDS Research Foundation and the Swedish Medical Research
Council. Grants were also received from the Swedish International
Development Cooperation Agency/Department for ResearchCooperation (SIDA/SAREC). We thank Kajsa Aperia, Maj Westman
and Zhong He for excellent technical assistance.
REFERENCES
Albert, J., Bredberg, U., Chiodi, F., Bottiger, B., Fenyo, E. M., Norrby,E. & Biberfeld, G. (1987). A new human retrovirus isolate of West
African origin (SBL-6669) and its relationship to HTLV-IV, LAV-II,
and HTLV-IIIB. AIDS Res Hum Retroviruses 3, 3–10.
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.
Albert, J., Stalhandske, P., Marquina, S., Karis, J., Fouchier, R. A.,Norrby, E. & Chiodi, F. (1996). Biological phenotype of HIV type 2
isolates correlates with V3 genotype. AIDS Res Hum Retroviruses 12,
821–828.
Andersson, S. (2001). HIV-2 and the immune response. AIDS Rev 3,
11–23.
Andersson, S., Norrgren, H., da Silva, Z., Biague, A., Bamba, S.,Kwok, S., Christopherson, C., Biberfeld, G. & Albert, J. (2000).Plasma viral load in HIV-1 and HIV-2 singly and dually infected
individuals in Guinea-Bissau, West Africa: significantly lower plasma
virus set point in HIV-2 infection than in HIV-1 infection. Arch
Intern Med 160, 3286–3293.
Andreasson, P. A., Dias, F., Naucler, A., Andersson, S. & Biberfeld,G. (1993). A prospective study of vertical transmission of HIV-2 in
Bissau, Guinea-Bissau. AIDS 7, 989–993.
Ariyoshi, K., Harwood, E., Chiengsong-Popov, R. & Weber, J.(1992). Is clearance of HIV-1 viraemia at seroconversion mediated
by neutralising antibodies? Lancet 340, 1257–1258.
Asjo, B., Morfeldt-Mansson, L., Albert, J., Biberfeld, G., Karlsson, A.,Lidman, K. & Fenyo, E. M. (1986). Replicative capacity of human
immunodeficiency virus from patients with varying severity of HIVinfection. Lancet ii, 660–662.
Benjouad, A., Gluckman, J.-C., Rochat, H., Montagnier, L. &Bahraoui, E. (1992). Influence of carbohydrate moieties on the
immunogenicity of human immunodeficiency virus type 1 recombi-
nant gp160. J Virol 66, 2473–2483.
Berry, N., Ariyoshi, K., Jaffar, S., Sabally, S., Corrah, T., Tedder, R. &Whittle, H. (1998). Low peripheral blood viral HIV-2 RNA in
individuals with high CD4 percentage differentiates HIV-2 from
HIV-1 infection. J Hum Virol 1, 457–468.
Bjorling, E., Scarlatti, G., von Gegerfelt, A., Albert, J., Biberfeld, G.,Chiodi, F., Norrby, E. & Fenyo, E. M. (1993). Autologous neutralizingantibodies prevail in HIV-2 but not in HIV-1 infection. Virology 193,
528–530.
Bjorndal, A., Deng, H., Jansson, M. & 7 other authors (1997).Coreceptor usage of primary human immunodeficiency virus type 1
isolates varies according to biological phenotype. J Virol 71,
7478–7487.
Blaak, H., Boers, P. H. M., Gruters, R. A., Schuitemaker, H., van derEnde, M. E. & Osterhaus, A. D. M. E. (2005). CCR5, GPR15, andCXCR6 are major coreceptors of human immunodeficiency virus
3394 Journal of General Virology 86
Y. Shi and others
Downloaded from www.microbiologyresearch.org by
IP: 50.16.49.105
On: Sun, 08 May 2016 06:29:27
type 2 variants isolated from individuals with and without plasma
viremia. J Virol 79, 1686–1700.
Brandin, E., Lindborg, L., Gyllensten, K., Brostrom, C., Hagberg, L.,Gisslen, M., Tuvesson, B., Blaxhult, A. & Albert, J. (2003). pol genesequence variation in Swedish HIV-2 patients failing antiretroviral
therapy. AIDS Res Hum Retroviruses 19, 543–550.
Carotenuto, P., Looij, D., Keldermans, L., de Wolf, F. & Goudsmit, J.(1998). Neutralizing antibodies are positively associated with CD4+
T-cell counts and T-cell function in long-term AIDS-free infection.
AIDS 12, 1591–1600.
CDC (1992). 1993 revised classification system for HIV infection and
expanded surveillance case definition for AIDS among adolescents
and adults. MMWR Recomm Rep 41, 1–19.
Chackerian, B., Rudensey, L. M. & Overbaugh, J. (1997). SpecificN-linked and O-linked glycosylation modifications in the envelope
V1 domain of simian immunodeficiency virus variants that evolve in
the host alter recognition by neutralizing antibodies. J Virol 71,
7719–7727.
Clapham, P. R., McKnight, A. & Weiss, R. A. (1992). Human
immunodeficiency virus type 2 infection and fusion of CD4-negative
human cell lines: induction and enhancement by soluble CD4. J Virol
66, 3531–3537.
Connor, R. I., Sheridan, K. E., Ceradini, D., Choe, S. & Landau, N. R.(1997). Change in coreceptor use correlates with disease progression
in HIV-1-infected individuals. J Exp Med 185, 621–628.
De Jong, J.-J., De Ronde, A., Keulen, W., Tersmette, M. & Goudsmit,J. (1992). Minimal requirements for the human immunodeficiency
virus type 1 V3 domain to support the syncytium-inducing
phenotype: analysis by single amino acid substitution. J Virol 66,
6777–6780.
Derdeyn, C. A., Decker, J. M., Bibollet-Ruche, F. & 11 other authors(2004). Envelope-constrained neutralization-sensitive HIV-1 after
Dreyer, K., Kallas, E. G., Planelles, V., Montefiori, D., McDermott, M. P.,Hasan, M. S. & Evans, T. G. (1999). Primary isolate neutralization
by HIV type 1-infected patient sera in the era of highly active
antiretroviral therapy. AIDS Res Hum Retroviruses 15, 1563–1571.
Fenyo, E. M. & Putkonen, P. (1996). Broad cross-neutralizing activity
in serum is associated with slow progression and low risk of trans-
mission in primate lentivirus infections. Immunol Lett 51, 95–99.
Fouchier, R. A. M., Groenink, M., Kootstra, N. A., Tersmette, M.,Huisman, H. G., Miedema, F. & Schuitemaker, H. (1992).Phenotype-associated sequence variation in the third variable
domain of the human immunodeficiency virus type 1 gp120
molecule. J Virol 66, 3183–3187.
Goudsmit, J., Debouck, C., Meloen, R. H., Smit, L., Bakker, M.,Asher, D. M., Wolff, A. V., Gibbs, C. J., Jr & Gajdusek, D. C. (1988).Human immunodeficiency virus type 1 neutralization epitope with
conserved architecture elicits early type-specific antibodies in
experimentally infected chimpanzees. Proc Natl Acad Sci U S A
85, 4478–4482.
Goulder, P. J. R., Brander, C., Tang, Y. & 16 other authors (2001).Evolution and transmission of stable CTL escape mutations in HIV
infection. Nature 412, 334–338.
Hall, T. A. (1999). BioEdit: a user-friendly biological sequence
alignment editor and analysis program for Windows 95/98/NT.
Nucleic Acids Symp Ser 41, 95–98.
Homsy, J., Meyer, M. & Levy, J. A. (1990). Serum enhancement of
human immunodeficiency virus (HIV) infection correlates with
disease in HIV-infected individuals. J Virol 64, 1437–1440.
Isaka, Y., Sato, A., Miki, S. & 8 other authors (1999). Small amino
acid changes in the V3 loop of human immunodeficiency virus type
2 determines the coreceptor usage for CXCR4 and CCR5. Virology
264, 237–243.
Javaherian, K., Langlois, A. J., McDanal, C. & 8 other authors (1989).Principal neutralizing domain of the human immunodeficiency virus
type 1 envelope protein. Proc Natl Acad Sci U S A 86, 6768–6772.
Jaye, A., Sarge-Njie, R., van der Loeff, M. S., Todd, J., Alabi, A.,Sabally, S., Corrah, T. & Whittle, H. (2004). No differences in cellular
immune responses between asymptomatic HIV type 1- and type 2-
Kanki, P. J., Travers, K. U., Mboup, S. & 9 other authors (1994).Slower heterosexual spread of HIV-2 than HIV-1. Lancet 343,
943–946.
Karlsson, I., Antonsson, L., Shi, Y., Karlsson, A., Albert, J., Leitner,T., Olde, B., Owman, C. & Fenyo, E. M. (2003). HIV biological
variability unveiled: frequent isolations and chimeric receptors reveal
unprecedented variation of coreceptor use. AIDS 17, 2561–2569.
Kimura, M. (1980). A simple method for estimating evolutionary
rates of base substitutions through comparative studies of nucleotide
sequences. J Mol Evol 16, 111–120.
Koot, M., Keet, I. P. M., Vos, A. H. V., de Goede, R. E. Y., Roos,M. T. L., Coutinho, R. A., Miedema, F., Schellekens, P. T. A. &Tersmette, M. (1993). Prognostic value of HIV-1 syncytium-
inducing phenotype for rate of CD4+ cell depletion and progression
to AIDS. Ann Intern Med 118, 681–688.
Kumar, S., Tamura, K., Jakobsen, I. B. & Nei, M. (2001). MEGA2:
Marlink, R., Kanki, P., Thior, I. & 13 other authors (1994). Reducedrate of disease development after HIV-2 infection as compared to
HIV-1. Science 265, 1587–1590.
McCaffrey, R. A., Saunders, C., Hensel, M. & Stamatatos, L. (2004).N-linked glycosylation of the V3 loop and the immunologically silent
face of gp120 protects human immunodeficiency virus type 1 SF162
from neutralization by anti-gp120 and anti-gp41 antibodies. J Virol
78, 3279–3295.
McKnight, A., Shotton, C., Cordell, J., Jones, I., Simmons, G. &Clapham, P. R. (1996). Location, exposure, and conservation of
neutralizing and nonneutralizing epitopes on human immunodefi-
ciency virus type 2 SU glycoprotein. J Virol 70, 4598–4606.
McKnight, A., Dittmar, M. T., Moniz-Periera, J. & 8 other authors(1998). A broad range of chemokine receptors are used by primary
isolates of human immunodeficiency virus type 2 as coreceptors with
CD4. J Virol 72, 4065–4071.
Morner, A., Achour, A., Norin, M., Thorstensson, R. & Bjorling, E.(1999a). Fine characterization of a V3-region neutralizing epitope in
human immunodeficiency virus type 2. Virus Res 59, 49–60.
Morner, A., Bjorndal, A., Albert, J., KewalRamani, V. N., Littman,D. R., Inoue, R., Thorstensson, R., Fenyo, E. M. & Bjorling, E.(1999b). Primary human immunodeficiency virus type 2 (HIV-2)
isolates, like HIV-1 isolates, frequently use CCR5 but show
promiscuity in coreceptor usage. J Virol 73, 2343–2349.
Morner, A., Bjorndal, A., Leandersson, A.-C., Albert, J., Bjorling, E. &Jansson, M. (2002). CCR5 or CXCR4 is required for efficient
infection of peripheral blood mononuclear cells by promiscuous
human immunodeficiency virus type 2 primary isolates. AIDS Res
Hum Retroviruses 18, 193–200.
Nabatov, A. A., Pollakis, G., Linnemann, T., Kliphius, A., Chalaby,M. I. M. & Paxton, W. A. (2004). Intrapatient alterations in the
human immunodeficiency virus type 1 gp120 V1V2 and V3 regions
differentially modulate coreceptor usage, virus inhibition by CC/
CXC chemokines, soluble CD4, and the b12 and 2G12 monoclonal
antibodies. J Virol 78, 524–530.
http://vir.sgmjournals.org 3395
HIV-2 evolution
Downloaded from www.microbiologyresearch.org by
IP: 50.16.49.105
On: Sun, 08 May 2016 06:29:27
Ogert, R. A., Lee, M. K., Ross, W., Buckler-White, A., Martin, M. A. &Cho, M. W. (2001). N-linked glycosylation sites adjacent to andwithin the V1/V2 and the V3 loops of dualtropic humanimmunodeficiency virus type 1 isolate DH12 gp120 affect coreceptorusage and cellular tropism. J Virol 75, 5998–6006.
Pepin, J., Morgan, G., Dunn, D., Gevao, S., Mendy, M., Gaye, I.,Scollen, N., Tedder, R. & Whittle, H. (1991). HIV-2-inducedimmunosuppression among asymptomatic West African prostitutes:evidence that HIV-2 is pathogenic, but less so than HIV-1. AIDS 5,1165–1172.
Pollakis, G., Kang, S., Kliphuis, A., Chalaby, M. I. M., Goudsmit, J. &Paxton, W. A. (2001). N-linked glycosylation of the HIV type-1gp120 envelope glycoprotein as a major determinant of CCR5 andCXCR4 coreceptor utilization. J Biol Chem 276, 13433–13441.
Polzer, S., Dittmar, M. T., Schmitz, H. & Schreiber, M. (2002). TheN-linked glycan g15 within the V3 loop of the HIV-1 externalglycoprotein gp120 affects coreceptor usage, cellular tropism, andneutralization. Virology 304, 70–80.
Popper, S. J., Sarr, A. D., Travers, K. U., Gueye-Ndiaye, A., Mboup,S., Essex, M. E. & Kanki, P. J. (1999). Lower human immuno-deficiency virus (HIV) type 2 viral load reflects the difference inpathogenicity of HIV-1 and HIV-2. J Infect Dis 180, 1116–1121.
Reeves, J. D. & Doms, R. W. (2002). Human immunodeficiency virustype 2. J Gen Virol 83, 1253–1265.
Reeves, J. D., Hibbitts, S., Simmons, G., McKnight, A., Azevedo-Pereira, J. M., Moniz-Pereira, J. & Clapham, P. R. (1999). Primaryhuman immunodeficiency virus type 2 (HIV-2) isolates infect CD4-negative cells via CCR5 and CXCR4: comparison with HIV-1 andsimian immunodeficiency virus and relevance to cell tropism in vivo.J Virol 73, 7795–7804.
Richman, D. D., Wrin, T., Little, S. J. & Petropoulos, C. J. (2003).Rapid evolution of the neutralizing antibody response to HIV type 1infection. Proc Natl Acad Sci U S A 100, 4144–4149.
Rucker, J., Edinger, A. L., Sharron, M. & 9 other authors (1997).Utilization of chemokine receptors, orphan receptors, and herpes-virus-encoded receptors by diverse human and simian immuno-deficiency viruses. J Virol 71, 8999–9007.
Rusche, J. R., Javaherian, K., McDanal, C. & 10 other authors(1988). Antibodies that inhibit fusion of human immunodeficiencyvirus-infected cells bind a 24-amino acid sequence of the viralenvelope, gp120. Proc Natl Acad Sci U S A 85, 3198–3202.
Rybarczyk, B. J., Montefiori, D., Johnson, P. R., West, A., Johnston,R. E. & Swanstrom, R. (2004). Correlation between env V1/V2 regiondiversification and neutralizing antibodies during primary infectionby simian immunodeficiency virus sm in rhesus macaques. J Virol78, 3561–3571.
Saitou, N. & Nei, M. (1987). The neighbor-joining method: a newmethod for reconstructing phylogenetic trees. Mol Biol Evol 4,406–425.
