VIROLOGY
185, 195-205 (1991)
Naturally Occurring Mutations within HIV-1 V3 Genomic RNA Lead to Antigenic Variation Dependent on a Single Amino Acid Substitution TOM F . W . WOLFS, GABRIEL ZWART, MARGREET BAKKER, MARGREETH VALK, CARLA L . KUIKEN, AND JAAP GOUDSMIT Human Retrovirus Laboratory, Academic Medical Centre, Amsterdam, The Netherlands Received April 26, 1991 ; accepted July 22, 1991
In a study on the evolution of genomic diversity of HIV-1, genornic RNA was isolated from serum of two in Starting at the time of primary infection we collected six samples of serum from each patient over aa period Ninety-four cDNA clones (50 of patient 1 and 44 of patient 496) of part of the envelope coding region inckicirkthe principal neutralization domain (PND) were sequenced . Aroundd the time of antibody seroconversion, gahemlo RNA levels reached a peak and the population of sequences was highly homogeneous . In the course of the infection ; the number of amino acid substitutions accumulated, which led to a higher genomic diversity within successive, opmples and a drift in the consensus sequence, progressively differing from the first found consensus sequence . Fixation of a substitution at glycoprotein 120 amino acid 308 was observed in both patients between two time points ?ipppientt,H- • P; patient 495, P --* H) . With the use of 16-meric synthetic peptides . differing only at the 308 position (f1 ° versusPon), antibody binding specificity was found to be dependent on this difference . In patient 495, the nonconserv*Rive (P0a - • H) substitution reduced the binding affinity with the patient's antibodies . Furthermore, antibody oompolildon assays showed that the observed substitution at position 308 elicited a new antibody populatiot .N indicating a t$gsntc variaLion. After the decline of V3-specific antibodies, the simultaneous increase in genornic RNA levels and pry greISsion to AIDS in patient 495, a new variant with major changes in the PND emerged, again forming a homogeneous population of sequences . 01991 Academk Presa, Inc. INTRODUCTION
highly antigenic (Devash et al., 1990 ; Zwart at al ., 1991) . The precise function of V3 in virus infection is not known . Callahan et al (1991) suggested that V3 plays an assisting role in membrane fusion after gpt 20 binding to CD4 . Moreover, V3 appears to contain a determinant for syncytium-inducing capacity in CD4+ HeLa cells (Freed et al., 1991), infectivity in Sup TI cells (de Jong, manuscript in preparation), and cell tropism (Takeuchi at al., 1991) . Mutations in the V3 region, as have been described to occur in vitro (McKeating et al., 1989; Masuda et al., 1990), enable variants to escape neutralization by antiV3 loop antibodies . Furthermore, even changes outside the V3 region conferred neutralization resistance to V3-specific neutralizing antibodies (Nare at al., 1990) . Analysis of sequences of the V3 domain from 245 different HIV-1 isolates revealed that the PND is less variable than was originally thought and varies at particular sites (LaRosa et al., 1990), Some of these more variable residues have been found earlier to be critical for antibody binding specificity (306: (Looney at at, 1988), 308 : (McKeating at at, 1989), 315 : Langedijk, manuscript in preparation) . Variation of the essential binding residues on top of the V3 loop (Meloen et at, 1989), Langedijk, in preparation) was extremely limited (LaRosa et al, 1990) .
HIV-1, like other lentiviruses and RNA viruses, has a high degree of variability, particularly in the envelope gene (Fisher et al., 1988 ; Hahn et al., 1986 ; Saag et al ., 1988) . In some of these viruses, such as equine infectious anemia virus (EIAV) and visna virus, genomic diversity leading to antigenic change allows the virus to escape surveillance by an intact immune system (Kono et at, 1973 ; Salinovich et al., 1986 ; Clements et al., 1980 ; Narayan et al., 1977, 1981) . The degree of sequence variation is not uniform throughout the HIV envelope gene . Five highly variable regions (VI-V5) have been defined where variability between a number of published sequences exceeds 75% on the amino acid level (Starcich et al., 1986 ; Modrow at at, 1987), The majority of the predicted epitopes in the envelope protein are found in these variable regions (Modrow et al., 1987) . The third variable region (V3) contains a neutralization site and has been shown to bind and elicit isolate-specific neutralizing antibodies (Goudsmit at at, 1988 ; Matsushita et at, 1988 ; Packer et al., 1988 ; Rusche at al., 1988) . This segment of the envelope, which forms a disulfide loop between Cys residues at gpl20 amino acids 296 and 330 (Leonard at al., 1990), has therefore been called the principal neutralization domain (PND) . The PND is 195
0042-6822/91 $3 .00 Copyright 0 1991 .by Academic Preas, Inc . . All fights of reproduclcn in any form reserved
1 96
WOLFS ET AL .
We (Wolfs etal., 1990) and others (Balfe et al., 1990 ; Simmonds etal., 1990) found genomic changes in vivo within the PND . The question remained whether antigenic drift is operational in vivo and whether the immune system recognizes antigenic variants as new targets and is able to mount an effective and lasting response against them . The present report describes the change in viral diversity of part of the HIV-1 envelope including the PND in the course of natural HIV infection and presents evidence of a related antigenic drift . Because in vitro propagation of virus is likely to select for viruses (Meyerhans et al., 1989), sequences were derived by PCR from genomic RNA out of serum without cultivation . Genomic RNA might be the best reflection of viruses actively replicating in vivo . Two individuals were studied during a follow-up period of 5 years after primary HIV infection . An observed substitution in the PND at a critical position for antibody specificity was mimicked by synthetic peptides containing the relevant substitution . Serological follow-up provided evidence that a nonconservative amino acid replacement is responsible for the modified interaction with the patient's own serum and is able to elicit a new antibody population .
MATERIAL AND METHODS Sera Sequential sera were collected from two homosexual men (patients 1 and 495) participating in a Dutch study reported by de Wolf et al. (1988) . Both men acquired HIV infection in 1985 and were followed for a period of more than five years . Individual 1 remained healthy although CD4` cell counts declined gradually and p24-antigenemia reappeared 33 months after infection . Only nonsyncytium-inducing (NSI) viruses were isolated during the follow-up period . In contrast, individual 495 was diagnosed as having AIDS (CDC IV C1) in 1989, 55 months after primary infection . Both a decline in CD4* cell counts and a reappearance of p24 antigenemia preceded the clinical deterioration . In addition, a switch from NSI to SI viruses was found between 1988 and 1989 . At the time AIDS was diagnosed, Zidovudine (AZT) treatment was started . In both cases, the first serum sample was taken in 1985 at the moment of p24 antigen conversion, which coincided with the detection of HIV antibodies . Serum samples taken 3 months earlier were negative both for HIV-antibody and p24 antigen . In these samples genomic RNA could not be detected either by PCR . Successive samples were taken 13, 22, 33, 46, and 59 months laterfor
patient 1, and 11, 23, 35, 44, and 56 months later for patient 495 . Molecular cloning Isolation of virus particle (genomic) RNA from 50 pl serum was performed according to Boom et al. (1990) . Viral RNA was converted to cDNA with the use of AMV RT (5 U) (Boehringer, Mannheim, BRD) and 10 pmol of the 3'V3 NOT primer (described below) in an RT buffer containing 75 mM KCI, 50 mMTris-HCI (pH 8 .3), 3 mM MgC1 2 , 10 mM DTT, and 0 .25 mM each of dATP, dCTP, dGTP, and dTTP (Pharmacia, Woerden, The Netherlands) in a total volume of 20 pl . A sample without AMV RT was used as negative control . cDNA was subjected to a nested polymerase chain reaction (PCR) . The primers used for the first PCR were 5'V3 NOT (5' ATAAGCTT CAATGTACACATGGAATT 3', HXB2 position 6506-6523, Los Alamos, 1990) and 3'V3 NOT (5' ATGAATTCA TTACAGTAGAAAAATTCCC 3', HXB2 position 6909-6928) bracketing the primers J-5'-2-KSI (5' ATAAGCTT GCAGTCTAGCAGAAGAAGA 3', HXB2 position 6558-6576), and J-3'-2KSI-2 (5' ATGAATTC TGGGTCCCTCCTGAGGA 3', HXB2 position 6860-6877) which were used for the second PCR . To allow subsequent cloning, a HindIll (AAGCTT) restriction site was incorporated into the J-5'-2-KSI primer and an EcoRl (GAATTC) restriction site into the J-3'-2-KSI-2 primer . The final PCR reaction mixture (100 pl) consisted of 53 mM KCI, 25 mMTrisHCI (pH 8 .4), 3 .3 MM MgCl 2 , 75 pg/ml BSA, 0 .24 mM each of dNTP, 10 pmol each of oligonucleotide primer, and 3 U Taq polymerase (a gift from Perkin-Elmer Cetus, Emeryville, CA, USA) . The reaction was performed for 35 cycles . Each cycle comprised a 1-min denaturation step at 95°, a 1-min annealing step at 55°, and a 2-min elongation step at 72° . Reamplification of 10 pl of the first PCR was performed in a second PCR making use of the KSI-primers under the same conditions . PCR products were purified by preparative agarose gel electrophoresis and digested with EcoRl and Hindlll . The digested fragment was cloned into EcoRl and Hindlll digested plasmid pGEM-7 (Promega Biotec, Madison, WI, USA) and transformed into Escherichia coli strain HB 101 . Sequencing Plasmid DNA from 50-m1 cultures was extracted with use of the Qiagen plasmid kit according to the manufacturer's recommendations (Qiagen, Inc ., CA) . Sequencing was performed with an automatic sequencer (Applied Biosystems, CA) using sequenase (USB, OH) .
HIV-1 ANTIGENIC DRIFT
Nucleotide sequence accession number
197
108
RKSIHI . .GPGRAFYTTG
109 RKSINI . .GPGRAFYTTG
All sequences used in this study have been submitted to GenBank (accession numbers M74591M74684) .
110 RKSIPI . .GPGRAFYTTG 111 RRRITM . .GPGRVLYTTG 112 SRGIRI . .GPGRAILATE
Consensus amino acid sequence A consensus sequence of each sample was made by assigning the most frequently found deduced amino acid in the individual clones to each position . Consensus sequences were based on 6 to 11 clones (part of sequence 3' adjacent to the V3 region based on 3 to 8 clones) .
Calculations The sequence population diversity was calculated by analysis of the amino acid sequences . The Hamming distance between two sequences is defined as the number of amino acid differences between them . The mean Hamming distance is the average value taken for all sequence pairs of one sample . This approach was only applied to fully sequenced clones (92 amino acids) . Quantifying virus load cDNA was prepared essentially as described above . A series of two-step dilutions of cDNA was made in a solution identical to the one used for the RT reaction . Dilutions were assayed by amplification in the nested PCR . This method does not allow calculation of the actual number of virus particles, because the efficiency of the method that was used to extract RNA from the serum and to transcribe the RNA into cDNA was not established . Peptides The peptides used had a length of 16 (or 17) residues and covered the entire neutralization domain in gp 120 V3 . Sequences of these peptides were derived from natural variants from the United States and The Netherlands . These variants were selected on the basis of (1) variant frequency (judged by the frequency of occurrence in the V3 sequence data base from Repligen (LaRosa et al. 1990) and our own sequence data (Goudsmit et at 1990) and (2) physicochemical differences between residues . The following peptides were used :
113
RRSIYT . .GPGRAFHTTD
114
RKSISI . .GPGRAIF FTTG
115 RKRITM . .GPGRVYYTTG 116 RRSIHV . .GPGQAFYATG
117
RKGIFI . .GPGRNIYTTG
51
.KSIRIQRGPGRAFVTIG
To check sera for aspecific reactivity a peptide was used with a sequence from bovine leukemia virus : EALTKVQVVQSKGSGPVCFNCKKPG . The peptides were produced with the use of Merrifield solid phase synthesis by the European Veterinary Laboratory (EVL, Amsterdam, The Netherlands) . ELISA The peptides were coated in microtiter plates in 100 gl phosphate-buffered saline (PBS) overnight at room temperature . Nonspecific binding sites were blocked by incubation with PBS, 2% (w/v) bovine serum albumin (BSA), 0 .1% tween-20 for 1 hr at 37° . Sera were diluted in PBS/2% BSA/0 .1% tween-20 and incubated for 1 hr at 37° . Next, wells were incubated with horseradish peroxidase-labeled conjugate (goat anti-human IgG, KPL, Gaithersburg, MD) diluted 1 :500 in PBS/2% BSAN0 .1% tween-20, for 1 hr at 37° . All incubations were followed by washing with PBS/0 .1% tween-20 . Bound antibodies were visualized with o-phenylene diamine and the reaction was stopped with 1 N HzSO, . Optical density was read at an absorbance of 450 nm . The midpoint titer of a serum for a particular peptide was defined as the reciprocal serum dilution at which the optical density was half the maximal optical density for that peptide . Antibody competition experiments A serum dilution was preincubated overnight with the homologous or a heterologous peptide at 1 .0 uM, at room temperature . This antibody peptide mixture was then incubated on a microtiter plate coated with the competing peptide and the further ELISA procedure was performed as described. The optical density was compared with the optical density of the serum
1 98
WOLFS ET AL .
which was subjected to the same procedure without the addition of peptide in the preincubation step .
patient 1 titer
RESULTS Level of HIV-1 genomic RNA present in serum during the course of infection To rule out the risk of amplification of single molecules and the subsequent cloning and sequencing of identical copies, we started by determining the level of HIV-1 genomic RNA in the serum . Genomic RNA was isolated from the first p24-antigen conversion serum (which coincided with the first positive RNA PCR sample) and five successive sera taken at 1-yr intervals . Complementary DNA prepared from 25 Al serum was assayed by nested PCR after dilution . The amount of genomic RNA present in serum collected at different time points of one individual varied significantly, as illustrated by the ultimate cDNA dilution that gave a positive PCR reaction . For patient 1, this highest positive dilution ranged from 1024-fold (first time point) to 32fold, showing an initial virus replication peak at the moment of p24-antigen conversion followed by a period of relatively low replication (Fig . 1) . Remarkably, the rise in p24-antigen at month 33 is not reflected by a higher level of genomic RNA . Likewise, the highest positive dilution was found for the first serum sample from patient 495 (>8000-fold) . In contrast to patient 1, after titer decline (to 32 ; third and fourth time point), genomic RNA titers increased in the infection (to 1024 ; last time point) coinciding with a secondary p24-antigenemia . In the last sample, both p24-antigenemia and the level of genomic RNA may well have decreased due to Zidovudine (AZT) treatment, which had been started 1 month earlier . Distribution and diversity of sequences during infection Nucleotide sequences of part of the envelope coding region (HXB2 nucleotide position 6577-6859, Los Alamos, 1990), which included the PND were obtained for each time point . The extent of genomic diversity within the viral population varied during the course of the infection . A measure of the diversity is given by the mean Hamming distance . The population of viral sequences obtained at the time of initial virus replication was found to be the most homogeneous (Fig . 2) . Successive samples showed a more heterogeneous population of sequences . However, as AIDS developed in patient 495 the virus population again seemed to be more homogeneous . Since genetic diversity showed inverse correlation with the levels of genomic RNA in
100
patient 495 titer
2600 0g/m1
10000 AZT
2000 1000 1600
1000 100 500,
10 4
4
20
30
a
40 A
50 A
60 4
- 10 70 months
Fic. 1 . Sequential changes in p24-antigen concentrations (lines) and genomic RNA titers (bars) in the serum of the two patients studied . RNA isolated from 25 MI serum was diluted after conversion into cDNA and PCR-ed . The highest dilution that resulted in a positive PCR reaction represents the reciprocal genomic RNA titer . "AZT denotes the start of anti-retroviral treatment . Arrows indicate samples used for sequence analyses . Note that the Y-axis scale is different for the two patients .
the serum, it is highly unlikely that the observed low variation is an artefact attributable to amplification of single or very few molecules . A consensus amino acid sequence was made for each time point by assigning to each position the deduced amino acid most commonly found in the individual clones . In successive samples an increase of the number of differences of both clonal and consensus sequences to the first or founder sequence was observed in both individuals, leading to a progressive drift in the course of the infection (Fig . 2) . Five years after seroconversion, consensus sequences differed by as much as 13% from the founder sequence . The rate at
iso
HIV-1 ANTIGENIC DRIFT
patient 1
A
patient 495
B
no . of dllferent amino acids 20 i
no . of different amino acids 20r
16
to
5
0 . -10
10
20
30
40
50
60
0
70
70
20
30
40
50
60
70
month&
months
hamming distance
hamming dislanoe 6 ;
a
z,l
0
10
20
30
40
50
60
70
0
10
20
30
40
Fic . 2 . (A) Number of amino acid differences in clonal sequences (dots) and consensus sequences (open squares) from six time points compared to the consensus sequence of the first time point . A total of 92 amino acids (position 270-363 of the HXB22 envelope gp120, Los Alamos, 1990) was compared, (B) Amino acid diversity within a population of sequences derived from one time point expressed by the mean Hamming distance .
which sequence changes from the founder strain occurred, varied, but was not correlated with any of the known variables (p24 antigen, virus titer, CD4' cell count, disease state) . Figure 3 shows the deduced consensus amino acid sequence of each time point . As is evident from this figure, not all amino acid substitutions were preserved in later samples . Amino acid substitutions in the central portion of the PND were found in both patients . At position 308, which had been reported earlier to change in vitro in the presence of neutralizing antibodies (McKeating at al., 1989), an amino acid replacement occurred (patient 1 H - P and patient 495 P --• H) and was preserved in later samples . Before the amino acid substitution, none of the individual clonal sequences harbored the later found amino acid at the 308 position, after the substitution all of the clones yielded the substituted amino acid . Position 313 also changed simultaneously in all clones in patient 1 (K R). In patient 495 an additional change was found at a later time point on position 310 (G -. Q) . The PND of the last sequence of patient 495, obtained at the time
AIDS was diagnosed, differed greatly from the former sequences (position 306 S -+ G, position 308 H -+ Y, position 310 reverted to the original amino acid G, and position 315 F - V), Interestingly, the biological phenotype too changed from nonsyncytium inducing to syncytium inducing in the latter year (personal communication, M . Tersmette) . Antibody reactivity to V3-derived peptides To find out whether the observed sequence changes had indeed led to an antigenic distinctness and affected the binding of the host's antibodies to the PND, we used a set of synthetic peptides with PN{) sequences derived from field isolates (see under Material and Methods) . From preliminary results {Zwart, in preparation) we learned that early antibodies produced by an infected individual had the highest reactivity with the peptide that had the best match to the corresponding sequence of the infecting variants . This finding supported our assumption that the PRO peptides mimick the corresponding sequence on the native viral envelope sufficiently to reflect antigenic differences .
200
WOLFS ET AL .
1 .00 :
1 .13 :
x 308 IVIRSDNITD NAKTIIVQLK EAVQINCTRP NNNTRKSIHI GPGKAFYATG EIIGDIRQAH
1 .22 :
---------- ---------- ----------"' T __________ ----------
1 .33 :
V --------- T K---- I P- ---R ----------
.46 :
----------
N KT___I P- ---R ----------
.59 :
__________
N KT-E-1 P_ ___R __________
.00 :
x CNLSRVDWED TLKQIAERLR EQFRNKTIVF NQ
1 .13 :
1 .22 :
---------N ""-""- -GN__K __E
1 .33 :
---- K_N__I __E_-_K A
1 .46 :
---- KEN--I __H_V
1 .59 :
RN__I __H_V A_ __
x 308 495 .00 : VVIRSDNFTD NAKIIIVQLN ESVEINCTRP NNNTRESIPI GPGRAFYTTG EIIGDIRQAH
495 .11 : ---------- ---------- ---------- __________ _
495 .23 : ---------- __________ ----------
R D
495 .35 : ---------- ---------- ----------
g_ __________ D_T
495 .44 :
N ----------
495 .56 :
N ----------
E_ Q K A 4i_y_
V___K K
x 495 .00 : CNISRAKNNN TLGQIVEELR EQFRN .KTII FNR
495 .11 : ----------
495 .23 : __________
G K Q
495 .35 : __________ -- EK__K_-C Q 495 .44 : ---------- -- EX--K--G Q 495 .56 : ---------- _- EK-- K G-N Q FiG . 3 . Alignment of deduced amino acid consensus sequences of different time points with the consensus sequence of the first time point . The entire sequenced region (amino acid 270-363) of the envelope is shown . Dashes represent amino acid identity . Asterisks mark the cysteine residues which create the V3 loop by forming a disulfide bond. The number after the patient's number indicates the number of months after p24-antigen conversion when the first sample was taken . The central portion of the PND is underlined and residue 308 is indicated .
Early serum (13 months) of patient 1 tested on the set of peptides showed the highest reactivity with peptide p108 (RKSIHIGPGRAFYTTG) (Fig . 4) . This peptide had the greatest homology with the genomic RNA derived amino acid sequence . The peptide differed from the founder sequence at positions 10 (R) and 14 (T) corresponding to positions 313 (K) and 317 (A), respectively, in the envelope gene . Early serum (11 months) of patient 495 had the highest reactivity to peptide p110 (RKSIPIGPGRAFYTTG), which is completely homologous to this individual's founder sequence .
Reactivity of these early sera from both individuals was markedly lower with a peptide with a single amino acid difference at position 5 (envelope position 308) (p108-H ; p109-N ; p110-P) . Although we cannot determine the contribution of each individual residue to antibody binding, apparently the 308 residue is important for the specificity of the host's antibodies . It can be expected that an amino acid c hange . at the 308 position will lead to diminished affinity of the preexisting antibodies for the viral glycoprotein . This made it interesting to investigate the antibody response of these
201
HIV-1 ANTIGENIC DRIFT
serum 1.13
serum 495.11
Re . 4 . IgG antibody reactivity in ELISA of early sera (13 and 11 months, respectively) of both patients to a panel of V3 peptides . We used the coating at which the greatest distinction between the different peptides was seen, 100 ng peptide per well for patient 1 and 10 ng per well for patient 495 .
individuals after the virus had undergone a change at the 308 position . Antibody response to amino acid changes in the PND To assess the effect of the observed genomic amino acid 308 substitutions on the immune response to the PND, sequential sera of both men were examined in ELISA for antibodies binding to p108 and p110 . Figure 5 shows the development of antibody reactivity against the two peptides after primary HIV-1 infection . Patient 1 showed antibody reactivity with p108 (H), soon after p24-antigen conversion, followed a year later by the appearance of antibodies reactive with p110 (P) . Both reactivities increased over time and persisted throughout the follow-up period . Absorbance in ELISA was expressed in optical density because titers of antibodies
to these peptides were relatively low . For patient 495, the antibody profiles over time were different . Antibodies reactive with p110 (P) appeared shortly after p24antigen conversion and increased in titer, followed by antibody reactivity against p108 (H) . Up to 24 months after infection (1 month after the first time we found the P - H substitution on position 308), titers remained stable and were twofold higher to p110 than to p108 . After that, however, titers against p110 and p108 showed a secondary parallel increase, followed at 33 months by a sudden decrease in titer against p110, whereas titers against p108 continued to rise until 49 months after infection . From this time point onward, both antibody reactivities decreased rapidly, paralleling the appearance of a novel sequence variant (GIYI), a further loss of CD4` cells, and the reappearance of high genomic RNA levels and viral antigen . In an antibody blocking experiment in which we used sera taken 9 and 48 months after infection, we wanted to find out whether the PND antibodies of patient 495 indeed consisted of two different populations . i n the early serum sample (P308) only 20% of reactivity to p110 (P) could be blocked by preincubation with p108 (H), whereas over 90% of lower reactivity to p108 could be blocked by p110 (Fig . 6) . In contrast, in the late serum sample (H308) reactivity to p108 could only be blocked for 36% by p110 and lower reactivity to p110 could be blocked by both p108 and p1 10 . These findings indicate that the antibody population of patient 495 directed against the PND was heterogeneous and consisted of at least two different populations . In the early serum samples the majority of anti-V3 antibodies was specific for P308 , whereas in the late serum the bulk of antibodies were specific for H 3°8 . For patient 1, antibody blocking experiments with p108 and p110 revealed that up to 37 months reactivity against p1 10 was mainly due to cross-reactive antibodies (reactivity could best be blocked by the heterologous peptide) . After this timepoint a specific antibody response to p110 was observed (reactivity could best be blocked by the homologous peptide) (data not shown) .
DISCUSSION In this study part of the envelope coding region from genomic RNA recovered from sequential serum samples of two individuals who seroconverted for HIV antibodies was sequenced . The extent of genomic diversity within the viral population varied during the course of infection . The viral population was highly homoge-
202
WOLFS ET AL .
patient 495
patient 1 Liar
I
H
I
I
P
I
I
-'- pin(X)
-°- Pilo (P)
P
I
I
M
I
a
I
FIG . 5 . Sequential changes in IgG antibody reactivity to peptides p108 and p110 . Reactivity was measured in ELISA and expressed in optical density (patient 1) or in midpoint titer (patient 495) . Peptides p108 and p110 only differ atone position from each other (H versus P, as indicated) . Indicated in the open boxes under the X-axis is the corresponding envelope 308 residue found in the genomic RNA from the patient at that time-point .
neous at the time of initial viral replication, probably indicating a selective outgrowth of the most replication-competent virus . In accordance with two recent studies in which primary HIV infection was associated with high levels of viremia (Clark et at, 1991 ; Daar et al., 1991), we found high levels of genomic RNA in both patients at this stage of infection . At later time points, the population became more heterogeneous . Genetic changes accumulated with time of infection, as has recently been described for simian immunodeficiency virus in an experimental infection (Burns et at, 1991) . After 5 years of follow-up, the amino acid variation of the consensus sequence of this envelope region to the consensus sequence of the founder virus amounted to 13-1496, a percentage which approximates the amino acid difference between the first consensus sequences of these two epidemiologically unrelated individuals (20%) . If variant viruses indeed do not replace the infecting strain of virus (Clements etal., 1988) and integrate into the host genome, the number of virus variants within one individual will be enormous . According to a model recently put forward by Nowak at at increase in the number of virus variants may be the most important factor leading to the breakdown of the immune system (Nowak at al., 1990) .
Not all substitutions were preserved in consecutive sequences, a finding which provides evidence for at least a partially occurring parallel evolution of sequential isolates (Saag etal., 1988) . On the other hand, fixation of a number of substitutions indicates that genetic evolution of HIV simultaneously follows a major evolutionary line, which is probably directed by the host's immune system . Between two time-points (1 year apart) a total replacement in all sequenced clones of gp120 amino acid 308 was found in both patients . Fixation of the replacing amino acid occurred in later samples . This amino acid has been reported to be critical for antibody-binding specificity (McKeating at al., 1989) . To measure the effect of the substitutions on antigenic distinctness, we made use of a set of V3-derived peptides . From the specificity of early sera for the PND peptide with the greatest homology with the corresponding domain in the infecting viruses, as observed in patients 1 and 495 and nine other individuals (Zwart, in preparation), it may be deduced that the synthetic peptides used mimick the viral domains sufficiently for comparison of antigenicity . Both the Pica -* H change in patient 495 and the H3oe - P change in patient 1 decreased the affinity of the respective preexisting anti-
HIV-1
ANTIGENIC DRIFT
203
p108
p110
100
so
A 40
0 lob
early
l01*
ShoMolopoM Mbrologaus I '
Fro. 6 . Percentage blocking of IgG antibody reactivity in an early (9 months) serum diluted 1 :150 and aa late (48 months) serum diluted 1 :200 of patient 495 to peptides p108 (A) and p110 (B) with the homologous and the heterologous peptide . Black bars represent blocking with the homologous peptide ; shaded bars represent blocking with the heterologous peptide .
bodies . Neutralization of the homologous HIV-strains, which appears to be critically dependent on high-affinity antibodies (Langedijk, in preparation) will probably be impaired by such a substitution . In visna and EIAV, variants with altered neutralization properties are rapidly recognized and neutralized by the host's immune system . For HIV, however, Albert et al. (1990) showed that infected humans did not develop antibodies capable of neutralizing newvirus variants . Longitudinal serological follow-up of our patients and additional antibody competition studies have shown that a specific humoral'immune response was mounted against the new mutant . Whether this antibody population is able to neutralize the newly arisen variant selectively has not yet been established . The emergence of these variants does not seem to be correlated with an increase of viral titer or a rapid deterioration of the patient's clinical status . However, when V3-specific antibodies declined in patient 495, higher levels of p24-antigenemia and genomic RNA were found and AIDS developed, again a new variant was found with major changes in the antibody binding site of the V3 domain . Furthermore, the viral population converged to a homogeneous one . Unlike what would be expected if the initial subtypes or clones persisted throughout the duration of the infection, we did not see a reactivation of the initially controlled virus populations at the time of humoral immune
function failure . Either the new variant was the fastest replicating one that was around when the immune system broke down, or this variant was a highly virulent one that directly caused the immurlo ogical breakdownThe large size and rapid evolution of the viral RNA genome pools will constantly produce get ome mutations that never before existed . We showed that naturally occurring mutations in genomic RNA leads to antigenic variation that is followed by an immunological response in the host . However, immunological dysfunction will eventually disable the immune system such that it is no longer capable of reacting to these new variants .
ACKNOWLEDGMENTS We thank John Sninsky for the generous gift of Taq polyrnerase, Frank de Wolf for providing the seraa from thee cohort studies, and Lie van der Hoek for technical assistance . This study was supported by Grant 28-1258 of the Netherlands Foundation of Preventive Medicine .
REFERENCES ALBERT, J ., ABRAHAMssoN, B ., NAGY, K ., AuRELius, E„ GAINES, H ., NVsTROM, G., and FENVO, E. M . (1990) . Rapid development of isolate-
204
WOLFS ET AL .
specific neutralizing antibodies after primary HIV-1 infection and consequent emergence of virus variants which resist neutralization by autologous sera . AIDS 4, 107-112 . BALFE, P ., SIMMONDS, P ., LUDLAM, A ., BISHOP, J . 0 ., and LEIGH BROWN, A . 1 . (1990) . Concurrent evolution of human immunodeficiency virus type 1 in patients infected from the same source : rate of sequence change and low frequency of inactivating mutations . J. Virol. 64(12), 6221-6233 . BOOM, R ., SOL, C . J . A ., SALIMANS, M . M . M ., JANSEN, C . L., WERTHEIMVAN ELLEN, P. M . E ., and NOORDAA VAN DER, J . (1990) . A rapid and simple method for purification of nucleic acids, J . Clin. Microbial. 28(3), 495-503 . BURNS, D . P. W . and DESROSIERS, R . C . (1991) . Selection of genetic variants of simian immunodeficiency virus in persistently infected rhesus monkeys . J. Virol. 65(4), 1843-1854. CALLAHAN, L . N ., PHELAN, M ., MALLINSON . M ., and NORCROSS, M . A. (1991) . Dextran sulfate blocks antibody binding to the principal neutralizing domain of human immunodeficiency virus type 1 without interfering with gpl 20-CD4 interactions . J. Virol. 65(3),15431550 . CLARK, S . J ., SHAG, M . S ., DECKER, W . D ., CAMPBELL-HILL, S ., ROBERSON, J . L ., VELDKAMP, P . J ., KAPPES, 1 . C ., HAHN, B . H ., and SHAW, G . M . (1991). High titers of cytopathic virus in plasma of patients with symptomatic primary HIV-1 infection . New Engl. J. Med. 324, 954-960 . CLEMENTS, J . E ., PEDERSON, F . S ., NARAYAN, O ., and HASELTINE, W . A. (1980) . Genomic changes associated with antigenic variation of visna virus during persistent infection . Proc. Nat!. Aced. Sci. USA 77,4454-4458 .
HAHN, B . H ., SHAW, G . M ., TAYLOR, M . E ., REDFIELD, R . R ., MARKHAM, P . D ., SALAHUDDIN, S . Z., WONG-STAAL, F ., GALLO, R. C ., PARKS, E . S ., and PARKS, W. P. (1986) . Genetic variation in HTLV-III/LAV over time in patients with AIDS or at risk for AIDS . Science 232, 1548-1553 . KoNo, Y ., KOBAYASHI, K ., and FIJKUNAGA, Y . (1973) . Antigenic drift of equine infectious anemia virus in chronically infected horses . Arch . Gas. Urusforsch . 41, 1-10 . IAROSA, G . J ., DAVIDE, J . P ., WEINHOLD, K . . WATERBURY, J . A ., PROFY, A . T ., LEWIS, J . A ., LANGLOIS, A . J ., DREESMAN, G . R ., BOSWELL, R . N ., SHADDUCK, P ., HOLLEY, L . H ., KARPLUS, M ., BOLOGNESI, D . P ., MATTHEWS, T . J ., EMINI, E . A., and PUTNEY, S . D . (1990) . Conserved sequence and structural elements in the HIV-1 principal neutralizing determinant . Science 249, 932-935 . LEONARD, C . K., SPELLMAN, M . W ., RIDDLE, L ., HARRIS, R . J ., THOMAS, J . N ., and GREGORY, T.1 . (1990) . Assignment of intrachain disulfide bonds and characterization of potential glycosylation sites of the type 1 recombinant human immunodeficiency virus envelope glycoprotein (gp120) expressed in Chinese hamster ovary cells . J. Biol. Chem . 265(18), 10373-10382 . LOONEY, D . J ., FISHER, A . G ., PUTNEY, S . D ., RUSCHE, J . R ., REDFIELD, R . R ., BURKE, S . D ., GALLO, R . C ., and WONG-STAAL, F . (1988) . Type restricted neutralization of molecular clones of human imm unodeficiency virus . Science 241, 357-359 . MASUDA, T ., MATSUSHITA, S ., KURODA, M . J ., KANNAGI, M ., TAKATSUKI, K., and HARADA, S . (1990) . Generation of neutralization-resistant HIV-1 in vitro due to amino acid interchanges of third hypervariable env region . J. lmmunol. 145, 3240-3246 .
CLEMENTS, J . E ., GDOVIN, S . L ., MONTELARO, R . C ., and NARAYAN, 0 . (1988) . Antigenic variation in lentiviral diseases . Annu . Rev. Immunol. 6, 139-159 .
MATSUSHITA, S., ROBERT-GUROFF, M ., RUSCHE, J ., Korro, A ., HATTORI, T ., HOSHINO, H ., JAVAHERIAN, K ., TAKATSUKI, K., and PUTNEY, S. (1988) . Characterization of a human immunodeficiency virus neutralizing monoclonal antibody and mapping of the neutralizing epilope . l. Virol. 62, 2107-2114 .
DAAR, E. S ., MOUDGIL, T ., MEYER, R . D ., and Ho, D . D . (1991) . Transient high levels of viremia in patients with primary human immunodeficiency virus type 1 infection . New Engl . J. Med. 324,961-964 .
McKEATING, 1 . A ., GOW, J ., GOUDSMIT, J ., PEARL, L . H ., MULDER, C ., and WEISS, R . A . (1989). Characterization of HIV-1 neutralization escape mutants . AIDS 3(12), 777-784 .
DEVASH, Y ., MATTHEWS, T . J ., DRUMMOND, 1 . E ., JAVAHERLAN, K ., WATERS, D. J ., ARTHUR, L . 0 ., BLATTNER, W . A„ and RUSCHE, J . R . (1990) . C-terminal fragments of gpl20 and synthetic peptides from five HTLV-111 strains : prevalence of antibodies to the HTLV-111MN isolate in infected individuals . AIDS Res . Human Retrov. 6(3),
MELOEN, R . H ., LISKAMP, R . M ., and GOUDSMIT, J . (1989) . Specificity and function of the individual amino acids of an important determinant of human immunodeficiency virus type 1 that induces neutralizing activity . J. Gen. Virol. 70, 1505-1512 .
307-316 . FISHER, A . G ., ENSOLI, B ., LOONEY, D ., ROSE, A ., GALLO, R . C ., SHAG, M, S ., SHAW, G . M ., HAHN, B . H., and WONG-STAAL, F . (1988) . Biologically diverse molecular variants within a single HIV-1 isolate . Nature (London) 334, 444-447 . FREED, E . 0 ., MYERS, D . J ., and RISSER, R . (1991) . Identification of the principal neutralizing determinant of human immunodeficiency virus type 1 as a fusion domain . J. Virol . 65, 190-194 . GOUDSMIT, J ., DEBOUCK, C ., MELOEN, R . H ., SMIT, L ., BAKKER, M ., ASHER, D . M ., WOLFF, A . V ., GIBBS, C . J ., JR ., and GAIDUSEK, D, C . (1988) . Human immunodeficiency virus type 1 neutralization epilope with conserved architecture elicits early type-specific antibodies in experimentally infected chimpanzees . Proc. Nail. Aced . Sci . USA 85, 4478-4482 . GoUDSMrr, J ., JONG DE, J ., EPSTEIN, L., TERSMETTE, M ., MIEDEMA, F., CLEMENT, M ., MELOEN, R ., ZwART, G ., and KRONE, W . (1990). Genomic and antigenic variation in the gpl 20 neutralization domain V3 : Challenge to a HIV vaccin . In "UCLA Symposia on Molecular and Cellular Biology New Series, Human Retroviruses" (J . E . Groopman, I . S . Y, Chen, M . Essex, and R . A. Weiss, Eds .), pp . 283-292 . Wiley-Liss, New York .
MEYERHANS, A., CHEYNIER, R ., ALBERT, J ., SETH, M ., KwOK, S ., SNINSKY, J ., MORFELDT-MANSON, L ., AsiO, B ., and WAIN-HOBSON, S . (1989) . Temporal fluctuations in HIV quasispecies in vivo are not reflected by sequential HIV isolations . Cell 58, 901-910 . MODROW, S ., HAHN, B . H ., SHAW, G . M ., GALLO, R . C ., WONG-STAAL, F ., and WOLF, H . (1987) . Computer-assisted analysis of envelope protein sequences of seven human immunodeficiency virus isolates : prediction of antigenic epitopes in conserved and variable regions .!. Virol. 61(2), 570-578 . NARA, P . L ., SMrr, L ., DUNLOP, N ., HATCH, W ., MERGES, M ., WATERS, D ., KELLIHER, J ., GALLO, R . C ., FISCHINGER, P . J ., and GOUDSMIT, J . (1990) . Emergence of viruses resistant to neutralization by V3 specific antibodies in experimental HIV-1 IIIB infection of chimpanzees . J. Virol. 64(8), 3779-3791 . NARAYAN, 0 ., GRIFFIN, D . E ., and CHASE, J . (1977), Antigenic shift of visna virus in persistently infected sheep . Science 197, 376-378 . NARAYAN, 0 ., CLEMENTS, J . E ., GRIFFIN, D . E ., and WOUNSKY, J . S . (1981) . Neutralizing antibody spectrum determines the antigenic profiles of emerging mutants of vans virus . Infect. Immunol. 32, 1045-1050 . NOWAK, A., MAY, R . M ., and ANDERSON, R . M . (1990). The evolutionary dynamics of HIV-1 quasispecies and the development of immunodeficiency disease . AIDS 4, 1095-1103 .
HIV-1 ANTIGENIC DRIFT
205
PALKER, T . J ., CLARK, M . E ., LANGLOIS, A . J ., MATTHEWS, T . J ., WEINHOLD, K . J ., RANDALL, R . R ., BOLOGNESI, D . P ., and HAYNES, B . F . (1988) . Type-specific neutralization of the human immunodefi-
S ., WOLF, H ., PARKS, E . S ., PARKS, W . P ., JOSEPHS, S . F ., GALLO, R . C ., and WONG-STAAL . F. (1986) . Identification and characterization of conserved and variable regions in the envelope gene of
ciency virus with antibodies to env-encoded synthetic peptides . Proc . Nat?. Acad. Sci. USA 85, 1932-1936 . RUSCHE, J . R ., JAVAHERIAN, K ., MCDANAL, C ., PETRO, J ., LYNN, D . L ., GRIMAILA, R., LANGLOIS, A ., GALLO, R . C ., ARTHUR, L . 0 ., Fis-
HTLV-III/LAV, the retrovirus of AIDS . Cell 45, 637-648TAKEUCHI . Y ., AKUTSU, M ., MURAYAMA, K ., SHIMIZU .. N ., and HOSHINO, H . (1991) . Host range mutant of human immunodeficiency virus type 1 : modification of cell tropism by a single point mutation at
CHINGER, P . J ., BOLOGNESI, D . P ., PUTNEY, S . D ., and MATTHEWS, T. J . (1988) . Antibodies that inhibit fusion of human immunodeficiency virus-infected cells bind a 24-amino acid sequence of the viral envelope, gpl20 . Proc . Natl . Aced. Sci . USA 85, 3198-3202 . SHAG, M . S ., HAHN, B . H ., GIesoNs, J ., Li, Y ., PARKS, E . S ., PARKS, W . P ., and SHAW, G. M . (1988) . Extensive variation of human immunodeficiency virus type 1 in vivo . Nature (London) 334, 440-444 . SALINOVICH, 0 ., PAYNE, S . L., MONTELERO, R. C ., HUSSAM, K . A ., ISSEL, C . J ., and ScHNORR, K . L . (1986). Rapid emergence of novel antigenic and genetic variants of EIAV during persistent infection . J. Virol. 57, 71-80 . SIMMONDS, P ., BALFE, P ., LUDLAM, C . A ., BISHOP, J . 0 ., and LEIGH BROWN, A . J . (1990) . Analysis of sequence diversity in hypervari-
the neutralization epitope in the env gene . l. Virol. 65, 1710-1718 . WOLF DE, F ., LANGE, J . M . A ., HOUWELING, J . T . M, COUTINHO, R . A., SCHELLEKENS, P . T ., NOORDAA VAN DER, .J . . and GOUDSMIT, 1 . (1988) . Numbers of CD4+ cells and the levels of core antigens of and antibodies to the human immunodeficiency virus as predictors of AIDS among seropositive homosexual men . J. Infect. Dis . 158(3), 615-622 . WOLFS, T . F . W ., JONG DE, J ., BERG VAN DEN, H ., TUNAGEL, J . M . G . H ., KRONE, W. J . A ., and GGUDSMIT, J . (1990) . Evolution of sequences encoding the principal neutralization epitope of HIV -1 is host-dependent, rapid and continuous . Proc. Natl Aced. Sci. LISA 87, 9938-9942 . ZWART, G ., LANGEDIJK, H ., HOEK VAN DER, L., JONG DE, J . J ., WOLFS,
able regions of the external glycoprotein of human immunodeficiency virus type 1 .1. Virol. 64(12), 5840-5850 . STARCICH, B . R ., HAHN, B . H ., SHAW, G . M ., McNEELY, P . D ., MODROW,
T . F . W ., RAMAUTARSING, C ., BAKKER, M ., RONDE DE, A ., and GOUDSMIT, J . (1991). Immunodominance and antigenic variation of the principal neutralization domain of HIV 1 . Virology 181, 481-489,
185, 195-205 (1991)
Naturally Occurring Mutations within HIV-1 V3 Genomic RNA Lead to Antigenic Variation Dependent on a Single Amino Acid Substitution TOM F . W . WOLFS, GABRIEL ZWART, MARGREET BAKKER, MARGREETH VALK, CARLA L . KUIKEN, AND JAAP GOUDSMIT Human Retrovirus Laboratory, Academic Medical Centre, Amsterdam, The Netherlands Received April 26, 1991 ; accepted July 22, 1991
In a study on the evolution of genomic diversity of HIV-1, genornic RNA was isolated from serum of two in Starting at the time of primary infection we collected six samples of serum from each patient over aa period Ninety-four cDNA clones (50 of patient 1 and 44 of patient 496) of part of the envelope coding region inckicirkthe principal neutralization domain (PND) were sequenced . Aroundd the time of antibody seroconversion, gahemlo RNA levels reached a peak and the population of sequences was highly homogeneous . In the course of the infection ; the number of amino acid substitutions accumulated, which led to a higher genomic diversity within successive, opmples and a drift in the consensus sequence, progressively differing from the first found consensus sequence . Fixation of a substitution at glycoprotein 120 amino acid 308 was observed in both patients between two time points ?ipppientt,H- • P; patient 495, P --* H) . With the use of 16-meric synthetic peptides . differing only at the 308 position (f1 ° versusPon), antibody binding specificity was found to be dependent on this difference . In patient 495, the nonconserv*Rive (P0a - • H) substitution reduced the binding affinity with the patient's antibodies . Furthermore, antibody oompolildon assays showed that the observed substitution at position 308 elicited a new antibody populatiot .N indicating a t$gsntc variaLion. After the decline of V3-specific antibodies, the simultaneous increase in genornic RNA levels and pry greISsion to AIDS in patient 495, a new variant with major changes in the PND emerged, again forming a homogeneous population of sequences . 01991 Academk Presa, Inc. INTRODUCTION
highly antigenic (Devash et al., 1990 ; Zwart at al ., 1991) . The precise function of V3 in virus infection is not known . Callahan et al (1991) suggested that V3 plays an assisting role in membrane fusion after gpt 20 binding to CD4 . Moreover, V3 appears to contain a determinant for syncytium-inducing capacity in CD4+ HeLa cells (Freed et al., 1991), infectivity in Sup TI cells (de Jong, manuscript in preparation), and cell tropism (Takeuchi at al., 1991) . Mutations in the V3 region, as have been described to occur in vitro (McKeating et al., 1989; Masuda et al., 1990), enable variants to escape neutralization by antiV3 loop antibodies . Furthermore, even changes outside the V3 region conferred neutralization resistance to V3-specific neutralizing antibodies (Nare at al., 1990) . Analysis of sequences of the V3 domain from 245 different HIV-1 isolates revealed that the PND is less variable than was originally thought and varies at particular sites (LaRosa et al., 1990), Some of these more variable residues have been found earlier to be critical for antibody binding specificity (306: (Looney at at, 1988), 308 : (McKeating at at, 1989), 315 : Langedijk, manuscript in preparation) . Variation of the essential binding residues on top of the V3 loop (Meloen et at, 1989), Langedijk, in preparation) was extremely limited (LaRosa et al, 1990) .
HIV-1, like other lentiviruses and RNA viruses, has a high degree of variability, particularly in the envelope gene (Fisher et al., 1988 ; Hahn et al., 1986 ; Saag et al ., 1988) . In some of these viruses, such as equine infectious anemia virus (EIAV) and visna virus, genomic diversity leading to antigenic change allows the virus to escape surveillance by an intact immune system (Kono et at, 1973 ; Salinovich et al., 1986 ; Clements et al., 1980 ; Narayan et al., 1977, 1981) . The degree of sequence variation is not uniform throughout the HIV envelope gene . Five highly variable regions (VI-V5) have been defined where variability between a number of published sequences exceeds 75% on the amino acid level (Starcich et al., 1986 ; Modrow at at, 1987), The majority of the predicted epitopes in the envelope protein are found in these variable regions (Modrow et al., 1987) . The third variable region (V3) contains a neutralization site and has been shown to bind and elicit isolate-specific neutralizing antibodies (Goudsmit at at, 1988 ; Matsushita et at, 1988 ; Packer et al., 1988 ; Rusche at al., 1988) . This segment of the envelope, which forms a disulfide loop between Cys residues at gpl20 amino acids 296 and 330 (Leonard at al., 1990), has therefore been called the principal neutralization domain (PND) . The PND is 195
0042-6822/91 $3 .00 Copyright 0 1991 .by Academic Preas, Inc . . All fights of reproduclcn in any form reserved
1 96
WOLFS ET AL .
We (Wolfs etal., 1990) and others (Balfe et al., 1990 ; Simmonds etal., 1990) found genomic changes in vivo within the PND . The question remained whether antigenic drift is operational in vivo and whether the immune system recognizes antigenic variants as new targets and is able to mount an effective and lasting response against them . The present report describes the change in viral diversity of part of the HIV-1 envelope including the PND in the course of natural HIV infection and presents evidence of a related antigenic drift . Because in vitro propagation of virus is likely to select for viruses (Meyerhans et al., 1989), sequences were derived by PCR from genomic RNA out of serum without cultivation . Genomic RNA might be the best reflection of viruses actively replicating in vivo . Two individuals were studied during a follow-up period of 5 years after primary HIV infection . An observed substitution in the PND at a critical position for antibody specificity was mimicked by synthetic peptides containing the relevant substitution . Serological follow-up provided evidence that a nonconservative amino acid replacement is responsible for the modified interaction with the patient's own serum and is able to elicit a new antibody population .
MATERIAL AND METHODS Sera Sequential sera were collected from two homosexual men (patients 1 and 495) participating in a Dutch study reported by de Wolf et al. (1988) . Both men acquired HIV infection in 1985 and were followed for a period of more than five years . Individual 1 remained healthy although CD4` cell counts declined gradually and p24-antigenemia reappeared 33 months after infection . Only nonsyncytium-inducing (NSI) viruses were isolated during the follow-up period . In contrast, individual 495 was diagnosed as having AIDS (CDC IV C1) in 1989, 55 months after primary infection . Both a decline in CD4* cell counts and a reappearance of p24 antigenemia preceded the clinical deterioration . In addition, a switch from NSI to SI viruses was found between 1988 and 1989 . At the time AIDS was diagnosed, Zidovudine (AZT) treatment was started . In both cases, the first serum sample was taken in 1985 at the moment of p24 antigen conversion, which coincided with the detection of HIV antibodies . Serum samples taken 3 months earlier were negative both for HIV-antibody and p24 antigen . In these samples genomic RNA could not be detected either by PCR . Successive samples were taken 13, 22, 33, 46, and 59 months laterfor
patient 1, and 11, 23, 35, 44, and 56 months later for patient 495 . Molecular cloning Isolation of virus particle (genomic) RNA from 50 pl serum was performed according to Boom et al. (1990) . Viral RNA was converted to cDNA with the use of AMV RT (5 U) (Boehringer, Mannheim, BRD) and 10 pmol of the 3'V3 NOT primer (described below) in an RT buffer containing 75 mM KCI, 50 mMTris-HCI (pH 8 .3), 3 mM MgC1 2 , 10 mM DTT, and 0 .25 mM each of dATP, dCTP, dGTP, and dTTP (Pharmacia, Woerden, The Netherlands) in a total volume of 20 pl . A sample without AMV RT was used as negative control . cDNA was subjected to a nested polymerase chain reaction (PCR) . The primers used for the first PCR were 5'V3 NOT (5' ATAAGCTT CAATGTACACATGGAATT 3', HXB2 position 6506-6523, Los Alamos, 1990) and 3'V3 NOT (5' ATGAATTCA TTACAGTAGAAAAATTCCC 3', HXB2 position 6909-6928) bracketing the primers J-5'-2-KSI (5' ATAAGCTT GCAGTCTAGCAGAAGAAGA 3', HXB2 position 6558-6576), and J-3'-2KSI-2 (5' ATGAATTC TGGGTCCCTCCTGAGGA 3', HXB2 position 6860-6877) which were used for the second PCR . To allow subsequent cloning, a HindIll (AAGCTT) restriction site was incorporated into the J-5'-2-KSI primer and an EcoRl (GAATTC) restriction site into the J-3'-2-KSI-2 primer . The final PCR reaction mixture (100 pl) consisted of 53 mM KCI, 25 mMTrisHCI (pH 8 .4), 3 .3 MM MgCl 2 , 75 pg/ml BSA, 0 .24 mM each of dNTP, 10 pmol each of oligonucleotide primer, and 3 U Taq polymerase (a gift from Perkin-Elmer Cetus, Emeryville, CA, USA) . The reaction was performed for 35 cycles . Each cycle comprised a 1-min denaturation step at 95°, a 1-min annealing step at 55°, and a 2-min elongation step at 72° . Reamplification of 10 pl of the first PCR was performed in a second PCR making use of the KSI-primers under the same conditions . PCR products were purified by preparative agarose gel electrophoresis and digested with EcoRl and Hindlll . The digested fragment was cloned into EcoRl and Hindlll digested plasmid pGEM-7 (Promega Biotec, Madison, WI, USA) and transformed into Escherichia coli strain HB 101 . Sequencing Plasmid DNA from 50-m1 cultures was extracted with use of the Qiagen plasmid kit according to the manufacturer's recommendations (Qiagen, Inc ., CA) . Sequencing was performed with an automatic sequencer (Applied Biosystems, CA) using sequenase (USB, OH) .
HIV-1 ANTIGENIC DRIFT
Nucleotide sequence accession number
197
108
RKSIHI . .GPGRAFYTTG
109 RKSINI . .GPGRAFYTTG
All sequences used in this study have been submitted to GenBank (accession numbers M74591M74684) .
110 RKSIPI . .GPGRAFYTTG 111 RRRITM . .GPGRVLYTTG 112 SRGIRI . .GPGRAILATE
Consensus amino acid sequence A consensus sequence of each sample was made by assigning the most frequently found deduced amino acid in the individual clones to each position . Consensus sequences were based on 6 to 11 clones (part of sequence 3' adjacent to the V3 region based on 3 to 8 clones) .
Calculations The sequence population diversity was calculated by analysis of the amino acid sequences . The Hamming distance between two sequences is defined as the number of amino acid differences between them . The mean Hamming distance is the average value taken for all sequence pairs of one sample . This approach was only applied to fully sequenced clones (92 amino acids) . Quantifying virus load cDNA was prepared essentially as described above . A series of two-step dilutions of cDNA was made in a solution identical to the one used for the RT reaction . Dilutions were assayed by amplification in the nested PCR . This method does not allow calculation of the actual number of virus particles, because the efficiency of the method that was used to extract RNA from the serum and to transcribe the RNA into cDNA was not established . Peptides The peptides used had a length of 16 (or 17) residues and covered the entire neutralization domain in gp 120 V3 . Sequences of these peptides were derived from natural variants from the United States and The Netherlands . These variants were selected on the basis of (1) variant frequency (judged by the frequency of occurrence in the V3 sequence data base from Repligen (LaRosa et al. 1990) and our own sequence data (Goudsmit et at 1990) and (2) physicochemical differences between residues . The following peptides were used :
113
RRSIYT . .GPGRAFHTTD
114
RKSISI . .GPGRAIF FTTG
115 RKRITM . .GPGRVYYTTG 116 RRSIHV . .GPGQAFYATG
117
RKGIFI . .GPGRNIYTTG
51
.KSIRIQRGPGRAFVTIG
To check sera for aspecific reactivity a peptide was used with a sequence from bovine leukemia virus : EALTKVQVVQSKGSGPVCFNCKKPG . The peptides were produced with the use of Merrifield solid phase synthesis by the European Veterinary Laboratory (EVL, Amsterdam, The Netherlands) . ELISA The peptides were coated in microtiter plates in 100 gl phosphate-buffered saline (PBS) overnight at room temperature . Nonspecific binding sites were blocked by incubation with PBS, 2% (w/v) bovine serum albumin (BSA), 0 .1% tween-20 for 1 hr at 37° . Sera were diluted in PBS/2% BSA/0 .1% tween-20 and incubated for 1 hr at 37° . Next, wells were incubated with horseradish peroxidase-labeled conjugate (goat anti-human IgG, KPL, Gaithersburg, MD) diluted 1 :500 in PBS/2% BSAN0 .1% tween-20, for 1 hr at 37° . All incubations were followed by washing with PBS/0 .1% tween-20 . Bound antibodies were visualized with o-phenylene diamine and the reaction was stopped with 1 N HzSO, . Optical density was read at an absorbance of 450 nm . The midpoint titer of a serum for a particular peptide was defined as the reciprocal serum dilution at which the optical density was half the maximal optical density for that peptide . Antibody competition experiments A serum dilution was preincubated overnight with the homologous or a heterologous peptide at 1 .0 uM, at room temperature . This antibody peptide mixture was then incubated on a microtiter plate coated with the competing peptide and the further ELISA procedure was performed as described. The optical density was compared with the optical density of the serum
1 98
WOLFS ET AL .
which was subjected to the same procedure without the addition of peptide in the preincubation step .
patient 1 titer
RESULTS Level of HIV-1 genomic RNA present in serum during the course of infection To rule out the risk of amplification of single molecules and the subsequent cloning and sequencing of identical copies, we started by determining the level of HIV-1 genomic RNA in the serum . Genomic RNA was isolated from the first p24-antigen conversion serum (which coincided with the first positive RNA PCR sample) and five successive sera taken at 1-yr intervals . Complementary DNA prepared from 25 Al serum was assayed by nested PCR after dilution . The amount of genomic RNA present in serum collected at different time points of one individual varied significantly, as illustrated by the ultimate cDNA dilution that gave a positive PCR reaction . For patient 1, this highest positive dilution ranged from 1024-fold (first time point) to 32fold, showing an initial virus replication peak at the moment of p24-antigen conversion followed by a period of relatively low replication (Fig . 1) . Remarkably, the rise in p24-antigen at month 33 is not reflected by a higher level of genomic RNA . Likewise, the highest positive dilution was found for the first serum sample from patient 495 (>8000-fold) . In contrast to patient 1, after titer decline (to 32 ; third and fourth time point), genomic RNA titers increased in the infection (to 1024 ; last time point) coinciding with a secondary p24-antigenemia . In the last sample, both p24-antigenemia and the level of genomic RNA may well have decreased due to Zidovudine (AZT) treatment, which had been started 1 month earlier . Distribution and diversity of sequences during infection Nucleotide sequences of part of the envelope coding region (HXB2 nucleotide position 6577-6859, Los Alamos, 1990), which included the PND were obtained for each time point . The extent of genomic diversity within the viral population varied during the course of the infection . A measure of the diversity is given by the mean Hamming distance . The population of viral sequences obtained at the time of initial virus replication was found to be the most homogeneous (Fig . 2) . Successive samples showed a more heterogeneous population of sequences . However, as AIDS developed in patient 495 the virus population again seemed to be more homogeneous . Since genetic diversity showed inverse correlation with the levels of genomic RNA in
100
patient 495 titer
2600 0g/m1
10000 AZT
2000 1000 1600
1000 100 500,
10 4
4
20
30
a
40 A
50 A
60 4
- 10 70 months
Fic. 1 . Sequential changes in p24-antigen concentrations (lines) and genomic RNA titers (bars) in the serum of the two patients studied . RNA isolated from 25 MI serum was diluted after conversion into cDNA and PCR-ed . The highest dilution that resulted in a positive PCR reaction represents the reciprocal genomic RNA titer . "AZT denotes the start of anti-retroviral treatment . Arrows indicate samples used for sequence analyses . Note that the Y-axis scale is different for the two patients .
the serum, it is highly unlikely that the observed low variation is an artefact attributable to amplification of single or very few molecules . A consensus amino acid sequence was made for each time point by assigning to each position the deduced amino acid most commonly found in the individual clones . In successive samples an increase of the number of differences of both clonal and consensus sequences to the first or founder sequence was observed in both individuals, leading to a progressive drift in the course of the infection (Fig . 2) . Five years after seroconversion, consensus sequences differed by as much as 13% from the founder sequence . The rate at
iso
HIV-1 ANTIGENIC DRIFT
patient 1
A
patient 495
B
no . of dllferent amino acids 20 i
no . of different amino acids 20r
16
to
5
0 . -10
10
20
30
40
50
60
0
70
70
20
30
40
50
60
70
month&
months
hamming distance
hamming dislanoe 6 ;
a
z,l
0
10
20
30
40
50
60
70
0
10
20
30
40
Fic . 2 . (A) Number of amino acid differences in clonal sequences (dots) and consensus sequences (open squares) from six time points compared to the consensus sequence of the first time point . A total of 92 amino acids (position 270-363 of the HXB22 envelope gp120, Los Alamos, 1990) was compared, (B) Amino acid diversity within a population of sequences derived from one time point expressed by the mean Hamming distance .
which sequence changes from the founder strain occurred, varied, but was not correlated with any of the known variables (p24 antigen, virus titer, CD4' cell count, disease state) . Figure 3 shows the deduced consensus amino acid sequence of each time point . As is evident from this figure, not all amino acid substitutions were preserved in later samples . Amino acid substitutions in the central portion of the PND were found in both patients . At position 308, which had been reported earlier to change in vitro in the presence of neutralizing antibodies (McKeating at al., 1989), an amino acid replacement occurred (patient 1 H - P and patient 495 P --• H) and was preserved in later samples . Before the amino acid substitution, none of the individual clonal sequences harbored the later found amino acid at the 308 position, after the substitution all of the clones yielded the substituted amino acid . Position 313 also changed simultaneously in all clones in patient 1 (K R). In patient 495 an additional change was found at a later time point on position 310 (G -. Q) . The PND of the last sequence of patient 495, obtained at the time
AIDS was diagnosed, differed greatly from the former sequences (position 306 S -+ G, position 308 H -+ Y, position 310 reverted to the original amino acid G, and position 315 F - V), Interestingly, the biological phenotype too changed from nonsyncytium inducing to syncytium inducing in the latter year (personal communication, M . Tersmette) . Antibody reactivity to V3-derived peptides To find out whether the observed sequence changes had indeed led to an antigenic distinctness and affected the binding of the host's antibodies to the PND, we used a set of synthetic peptides with PN{) sequences derived from field isolates (see under Material and Methods) . From preliminary results {Zwart, in preparation) we learned that early antibodies produced by an infected individual had the highest reactivity with the peptide that had the best match to the corresponding sequence of the infecting variants . This finding supported our assumption that the PRO peptides mimick the corresponding sequence on the native viral envelope sufficiently to reflect antigenic differences .
200
WOLFS ET AL .
1 .00 :
1 .13 :
x 308 IVIRSDNITD NAKTIIVQLK EAVQINCTRP NNNTRKSIHI GPGKAFYATG EIIGDIRQAH
1 .22 :
---------- ---------- ----------"' T __________ ----------
1 .33 :
V --------- T K---- I P- ---R ----------
.46 :
----------
N KT___I P- ---R ----------
.59 :
__________
N KT-E-1 P_ ___R __________
.00 :
x CNLSRVDWED TLKQIAERLR EQFRNKTIVF NQ
1 .13 :
1 .22 :
---------N ""-""- -GN__K __E
1 .33 :
---- K_N__I __E_-_K A
1 .46 :
---- KEN--I __H_V
1 .59 :
RN__I __H_V A_ __
x 308 495 .00 : VVIRSDNFTD NAKIIIVQLN ESVEINCTRP NNNTRESIPI GPGRAFYTTG EIIGDIRQAH
495 .11 : ---------- ---------- ---------- __________ _
495 .23 : ---------- __________ ----------
R D
495 .35 : ---------- ---------- ----------
g_ __________ D_T
495 .44 :
N ----------
495 .56 :
N ----------
E_ Q K A 4i_y_
V___K K
x 495 .00 : CNISRAKNNN TLGQIVEELR EQFRN .KTII FNR
495 .11 : ----------
495 .23 : __________
G K Q
495 .35 : __________ -- EK__K_-C Q 495 .44 : ---------- -- EX--K--G Q 495 .56 : ---------- _- EK-- K G-N Q FiG . 3 . Alignment of deduced amino acid consensus sequences of different time points with the consensus sequence of the first time point . The entire sequenced region (amino acid 270-363) of the envelope is shown . Dashes represent amino acid identity . Asterisks mark the cysteine residues which create the V3 loop by forming a disulfide bond. The number after the patient's number indicates the number of months after p24-antigen conversion when the first sample was taken . The central portion of the PND is underlined and residue 308 is indicated .
Early serum (13 months) of patient 1 tested on the set of peptides showed the highest reactivity with peptide p108 (RKSIHIGPGRAFYTTG) (Fig . 4) . This peptide had the greatest homology with the genomic RNA derived amino acid sequence . The peptide differed from the founder sequence at positions 10 (R) and 14 (T) corresponding to positions 313 (K) and 317 (A), respectively, in the envelope gene . Early serum (11 months) of patient 495 had the highest reactivity to peptide p110 (RKSIPIGPGRAFYTTG), which is completely homologous to this individual's founder sequence .
Reactivity of these early sera from both individuals was markedly lower with a peptide with a single amino acid difference at position 5 (envelope position 308) (p108-H ; p109-N ; p110-P) . Although we cannot determine the contribution of each individual residue to antibody binding, apparently the 308 residue is important for the specificity of the host's antibodies . It can be expected that an amino acid c hange . at the 308 position will lead to diminished affinity of the preexisting antibodies for the viral glycoprotein . This made it interesting to investigate the antibody response of these
201
HIV-1 ANTIGENIC DRIFT
serum 1.13
serum 495.11
Re . 4 . IgG antibody reactivity in ELISA of early sera (13 and 11 months, respectively) of both patients to a panel of V3 peptides . We used the coating at which the greatest distinction between the different peptides was seen, 100 ng peptide per well for patient 1 and 10 ng per well for patient 495 .
individuals after the virus had undergone a change at the 308 position . Antibody response to amino acid changes in the PND To assess the effect of the observed genomic amino acid 308 substitutions on the immune response to the PND, sequential sera of both men were examined in ELISA for antibodies binding to p108 and p110 . Figure 5 shows the development of antibody reactivity against the two peptides after primary HIV-1 infection . Patient 1 showed antibody reactivity with p108 (H), soon after p24-antigen conversion, followed a year later by the appearance of antibodies reactive with p110 (P) . Both reactivities increased over time and persisted throughout the follow-up period . Absorbance in ELISA was expressed in optical density because titers of antibodies
to these peptides were relatively low . For patient 495, the antibody profiles over time were different . Antibodies reactive with p110 (P) appeared shortly after p24antigen conversion and increased in titer, followed by antibody reactivity against p108 (H) . Up to 24 months after infection (1 month after the first time we found the P - H substitution on position 308), titers remained stable and were twofold higher to p110 than to p108 . After that, however, titers against p110 and p108 showed a secondary parallel increase, followed at 33 months by a sudden decrease in titer against p110, whereas titers against p108 continued to rise until 49 months after infection . From this time point onward, both antibody reactivities decreased rapidly, paralleling the appearance of a novel sequence variant (GIYI), a further loss of CD4` cells, and the reappearance of high genomic RNA levels and viral antigen . In an antibody blocking experiment in which we used sera taken 9 and 48 months after infection, we wanted to find out whether the PND antibodies of patient 495 indeed consisted of two different populations . i n the early serum sample (P308) only 20% of reactivity to p110 (P) could be blocked by preincubation with p108 (H), whereas over 90% of lower reactivity to p108 could be blocked by p110 (Fig . 6) . In contrast, in the late serum sample (H308) reactivity to p108 could only be blocked for 36% by p110 and lower reactivity to p110 could be blocked by both p108 and p1 10 . These findings indicate that the antibody population of patient 495 directed against the PND was heterogeneous and consisted of at least two different populations . In the early serum samples the majority of anti-V3 antibodies was specific for P308 , whereas in the late serum the bulk of antibodies were specific for H 3°8 . For patient 1, antibody blocking experiments with p108 and p110 revealed that up to 37 months reactivity against p1 10 was mainly due to cross-reactive antibodies (reactivity could best be blocked by the heterologous peptide) . After this timepoint a specific antibody response to p110 was observed (reactivity could best be blocked by the homologous peptide) (data not shown) .
DISCUSSION In this study part of the envelope coding region from genomic RNA recovered from sequential serum samples of two individuals who seroconverted for HIV antibodies was sequenced . The extent of genomic diversity within the viral population varied during the course of infection . The viral population was highly homoge-
202
WOLFS ET AL .
patient 495
patient 1 Liar
I
H
I
I
P
I
I
-'- pin(X)
-°- Pilo (P)
P
I
I
M
I
a
I
FIG . 5 . Sequential changes in IgG antibody reactivity to peptides p108 and p110 . Reactivity was measured in ELISA and expressed in optical density (patient 1) or in midpoint titer (patient 495) . Peptides p108 and p110 only differ atone position from each other (H versus P, as indicated) . Indicated in the open boxes under the X-axis is the corresponding envelope 308 residue found in the genomic RNA from the patient at that time-point .
neous at the time of initial viral replication, probably indicating a selective outgrowth of the most replication-competent virus . In accordance with two recent studies in which primary HIV infection was associated with high levels of viremia (Clark et at, 1991 ; Daar et al., 1991), we found high levels of genomic RNA in both patients at this stage of infection . At later time points, the population became more heterogeneous . Genetic changes accumulated with time of infection, as has recently been described for simian immunodeficiency virus in an experimental infection (Burns et at, 1991) . After 5 years of follow-up, the amino acid variation of the consensus sequence of this envelope region to the consensus sequence of the founder virus amounted to 13-1496, a percentage which approximates the amino acid difference between the first consensus sequences of these two epidemiologically unrelated individuals (20%) . If variant viruses indeed do not replace the infecting strain of virus (Clements etal., 1988) and integrate into the host genome, the number of virus variants within one individual will be enormous . According to a model recently put forward by Nowak at at increase in the number of virus variants may be the most important factor leading to the breakdown of the immune system (Nowak at al., 1990) .
Not all substitutions were preserved in consecutive sequences, a finding which provides evidence for at least a partially occurring parallel evolution of sequential isolates (Saag etal., 1988) . On the other hand, fixation of a number of substitutions indicates that genetic evolution of HIV simultaneously follows a major evolutionary line, which is probably directed by the host's immune system . Between two time-points (1 year apart) a total replacement in all sequenced clones of gp120 amino acid 308 was found in both patients . Fixation of the replacing amino acid occurred in later samples . This amino acid has been reported to be critical for antibody-binding specificity (McKeating at al., 1989) . To measure the effect of the substitutions on antigenic distinctness, we made use of a set of V3-derived peptides . From the specificity of early sera for the PND peptide with the greatest homology with the corresponding domain in the infecting viruses, as observed in patients 1 and 495 and nine other individuals (Zwart, in preparation), it may be deduced that the synthetic peptides used mimick the viral domains sufficiently for comparison of antigenicity . Both the Pica -* H change in patient 495 and the H3oe - P change in patient 1 decreased the affinity of the respective preexisting anti-
HIV-1
ANTIGENIC DRIFT
203
p108
p110
100
so
A 40
0 lob
early
l01*
ShoMolopoM Mbrologaus I '
Fro. 6 . Percentage blocking of IgG antibody reactivity in an early (9 months) serum diluted 1 :150 and aa late (48 months) serum diluted 1 :200 of patient 495 to peptides p108 (A) and p110 (B) with the homologous and the heterologous peptide . Black bars represent blocking with the homologous peptide ; shaded bars represent blocking with the heterologous peptide .
bodies . Neutralization of the homologous HIV-strains, which appears to be critically dependent on high-affinity antibodies (Langedijk, in preparation) will probably be impaired by such a substitution . In visna and EIAV, variants with altered neutralization properties are rapidly recognized and neutralized by the host's immune system . For HIV, however, Albert et al. (1990) showed that infected humans did not develop antibodies capable of neutralizing newvirus variants . Longitudinal serological follow-up of our patients and additional antibody competition studies have shown that a specific humoral'immune response was mounted against the new mutant . Whether this antibody population is able to neutralize the newly arisen variant selectively has not yet been established . The emergence of these variants does not seem to be correlated with an increase of viral titer or a rapid deterioration of the patient's clinical status . However, when V3-specific antibodies declined in patient 495, higher levels of p24-antigenemia and genomic RNA were found and AIDS developed, again a new variant was found with major changes in the antibody binding site of the V3 domain . Furthermore, the viral population converged to a homogeneous one . Unlike what would be expected if the initial subtypes or clones persisted throughout the duration of the infection, we did not see a reactivation of the initially controlled virus populations at the time of humoral immune
function failure . Either the new variant was the fastest replicating one that was around when the immune system broke down, or this variant was a highly virulent one that directly caused the immurlo ogical breakdownThe large size and rapid evolution of the viral RNA genome pools will constantly produce get ome mutations that never before existed . We showed that naturally occurring mutations in genomic RNA leads to antigenic variation that is followed by an immunological response in the host . However, immunological dysfunction will eventually disable the immune system such that it is no longer capable of reacting to these new variants .
ACKNOWLEDGMENTS We thank John Sninsky for the generous gift of Taq polyrnerase, Frank de Wolf for providing the seraa from thee cohort studies, and Lie van der Hoek for technical assistance . This study was supported by Grant 28-1258 of the Netherlands Foundation of Preventive Medicine .
REFERENCES ALBERT, J ., ABRAHAMssoN, B ., NAGY, K ., AuRELius, E„ GAINES, H ., NVsTROM, G., and FENVO, E. M . (1990) . Rapid development of isolate-
204
WOLFS ET AL .
specific neutralizing antibodies after primary HIV-1 infection and consequent emergence of virus variants which resist neutralization by autologous sera . AIDS 4, 107-112 . BALFE, P ., SIMMONDS, P ., LUDLAM, A ., BISHOP, J . 0 ., and LEIGH BROWN, A . 1 . (1990) . Concurrent evolution of human immunodeficiency virus type 1 in patients infected from the same source : rate of sequence change and low frequency of inactivating mutations . J. Virol. 64(12), 6221-6233 . BOOM, R ., SOL, C . J . A ., SALIMANS, M . M . M ., JANSEN, C . L., WERTHEIMVAN ELLEN, P. M . E ., and NOORDAA VAN DER, J . (1990) . A rapid and simple method for purification of nucleic acids, J . Clin. Microbial. 28(3), 495-503 . BURNS, D . P. W . and DESROSIERS, R . C . (1991) . Selection of genetic variants of simian immunodeficiency virus in persistently infected rhesus monkeys . J. Virol. 65(4), 1843-1854. CALLAHAN, L . N ., PHELAN, M ., MALLINSON . M ., and NORCROSS, M . A. (1991) . Dextran sulfate blocks antibody binding to the principal neutralizing domain of human immunodeficiency virus type 1 without interfering with gpl 20-CD4 interactions . J. Virol. 65(3),15431550 . CLARK, S . J ., SHAG, M . S ., DECKER, W . D ., CAMPBELL-HILL, S ., ROBERSON, J . L ., VELDKAMP, P . J ., KAPPES, 1 . C ., HAHN, B . H ., and SHAW, G . M . (1991). High titers of cytopathic virus in plasma of patients with symptomatic primary HIV-1 infection . New Engl. J. Med. 324, 954-960 . CLEMENTS, J . E ., PEDERSON, F . S ., NARAYAN, O ., and HASELTINE, W . A. (1980) . Genomic changes associated with antigenic variation of visna virus during persistent infection . Proc. Nat!. Aced. Sci. USA 77,4454-4458 .
HAHN, B . H ., SHAW, G . M ., TAYLOR, M . E ., REDFIELD, R . R ., MARKHAM, P . D ., SALAHUDDIN, S . Z., WONG-STAAL, F ., GALLO, R. C ., PARKS, E . S ., and PARKS, W. P. (1986) . Genetic variation in HTLV-III/LAV over time in patients with AIDS or at risk for AIDS . Science 232, 1548-1553 . KoNo, Y ., KOBAYASHI, K ., and FIJKUNAGA, Y . (1973) . Antigenic drift of equine infectious anemia virus in chronically infected horses . Arch . Gas. Urusforsch . 41, 1-10 . IAROSA, G . J ., DAVIDE, J . P ., WEINHOLD, K . . WATERBURY, J . A ., PROFY, A . T ., LEWIS, J . A ., LANGLOIS, A . J ., DREESMAN, G . R ., BOSWELL, R . N ., SHADDUCK, P ., HOLLEY, L . H ., KARPLUS, M ., BOLOGNESI, D . P ., MATTHEWS, T . J ., EMINI, E . A., and PUTNEY, S . D . (1990) . Conserved sequence and structural elements in the HIV-1 principal neutralizing determinant . Science 249, 932-935 . LEONARD, C . K., SPELLMAN, M . W ., RIDDLE, L ., HARRIS, R . J ., THOMAS, J . N ., and GREGORY, T.1 . (1990) . Assignment of intrachain disulfide bonds and characterization of potential glycosylation sites of the type 1 recombinant human immunodeficiency virus envelope glycoprotein (gp120) expressed in Chinese hamster ovary cells . J. Biol. Chem . 265(18), 10373-10382 . LOONEY, D . J ., FISHER, A . G ., PUTNEY, S . D ., RUSCHE, J . R ., REDFIELD, R . R ., BURKE, S . D ., GALLO, R . C ., and WONG-STAAL, F . (1988) . Type restricted neutralization of molecular clones of human imm unodeficiency virus . Science 241, 357-359 . MASUDA, T ., MATSUSHITA, S ., KURODA, M . J ., KANNAGI, M ., TAKATSUKI, K., and HARADA, S . (1990) . Generation of neutralization-resistant HIV-1 in vitro due to amino acid interchanges of third hypervariable env region . J. lmmunol. 145, 3240-3246 .
CLEMENTS, J . E ., GDOVIN, S . L ., MONTELARO, R . C ., and NARAYAN, 0 . (1988) . Antigenic variation in lentiviral diseases . Annu . Rev. Immunol. 6, 139-159 .
MATSUSHITA, S., ROBERT-GUROFF, M ., RUSCHE, J ., Korro, A ., HATTORI, T ., HOSHINO, H ., JAVAHERIAN, K ., TAKATSUKI, K., and PUTNEY, S. (1988) . Characterization of a human immunodeficiency virus neutralizing monoclonal antibody and mapping of the neutralizing epilope . l. Virol. 62, 2107-2114 .
DAAR, E. S ., MOUDGIL, T ., MEYER, R . D ., and Ho, D . D . (1991) . Transient high levels of viremia in patients with primary human immunodeficiency virus type 1 infection . New Engl . J. Med. 324,961-964 .
McKEATING, 1 . A ., GOW, J ., GOUDSMIT, J ., PEARL, L . H ., MULDER, C ., and WEISS, R . A . (1989). Characterization of HIV-1 neutralization escape mutants . AIDS 3(12), 777-784 .
DEVASH, Y ., MATTHEWS, T . J ., DRUMMOND, 1 . E ., JAVAHERLAN, K ., WATERS, D. J ., ARTHUR, L . 0 ., BLATTNER, W . A„ and RUSCHE, J . R . (1990) . C-terminal fragments of gpl20 and synthetic peptides from five HTLV-111 strains : prevalence of antibodies to the HTLV-111MN isolate in infected individuals . AIDS Res . Human Retrov. 6(3),
MELOEN, R . H ., LISKAMP, R . M ., and GOUDSMIT, J . (1989) . Specificity and function of the individual amino acids of an important determinant of human immunodeficiency virus type 1 that induces neutralizing activity . J. Gen. Virol. 70, 1505-1512 .
307-316 . FISHER, A . G ., ENSOLI, B ., LOONEY, D ., ROSE, A ., GALLO, R . C ., SHAG, M, S ., SHAW, G . M ., HAHN, B . H., and WONG-STAAL, F . (1988) . Biologically diverse molecular variants within a single HIV-1 isolate . Nature (London) 334, 444-447 . FREED, E . 0 ., MYERS, D . J ., and RISSER, R . (1991) . Identification of the principal neutralizing determinant of human immunodeficiency virus type 1 as a fusion domain . J. Virol . 65, 190-194 . GOUDSMIT, J ., DEBOUCK, C ., MELOEN, R . H ., SMIT, L ., BAKKER, M ., ASHER, D . M ., WOLFF, A . V ., GIBBS, C . J ., JR ., and GAIDUSEK, D, C . (1988) . Human immunodeficiency virus type 1 neutralization epilope with conserved architecture elicits early type-specific antibodies in experimentally infected chimpanzees . Proc. Nail. Aced . Sci . USA 85, 4478-4482 . GoUDSMrr, J ., JONG DE, J ., EPSTEIN, L., TERSMETTE, M ., MIEDEMA, F., CLEMENT, M ., MELOEN, R ., ZwART, G ., and KRONE, W . (1990). Genomic and antigenic variation in the gpl 20 neutralization domain V3 : Challenge to a HIV vaccin . In "UCLA Symposia on Molecular and Cellular Biology New Series, Human Retroviruses" (J . E . Groopman, I . S . Y, Chen, M . Essex, and R . A. Weiss, Eds .), pp . 283-292 . Wiley-Liss, New York .
MEYERHANS, A., CHEYNIER, R ., ALBERT, J ., SETH, M ., KwOK, S ., SNINSKY, J ., MORFELDT-MANSON, L ., AsiO, B ., and WAIN-HOBSON, S . (1989) . Temporal fluctuations in HIV quasispecies in vivo are not reflected by sequential HIV isolations . Cell 58, 901-910 . MODROW, S ., HAHN, B . H ., SHAW, G . M ., GALLO, R . C ., WONG-STAAL, F ., and WOLF, H . (1987) . Computer-assisted analysis of envelope protein sequences of seven human immunodeficiency virus isolates : prediction of antigenic epitopes in conserved and variable regions .!. Virol. 61(2), 570-578 . NARA, P . L ., SMrr, L ., DUNLOP, N ., HATCH, W ., MERGES, M ., WATERS, D ., KELLIHER, J ., GALLO, R . C ., FISCHINGER, P . J ., and GOUDSMIT, J . (1990) . Emergence of viruses resistant to neutralization by V3 specific antibodies in experimental HIV-1 IIIB infection of chimpanzees . J. Virol. 64(8), 3779-3791 . NARAYAN, 0 ., GRIFFIN, D . E ., and CHASE, J . (1977), Antigenic shift of visna virus in persistently infected sheep . Science 197, 376-378 . NARAYAN, 0 ., CLEMENTS, J . E ., GRIFFIN, D . E ., and WOUNSKY, J . S . (1981) . Neutralizing antibody spectrum determines the antigenic profiles of emerging mutants of vans virus . Infect. Immunol. 32, 1045-1050 . NOWAK, A., MAY, R . M ., and ANDERSON, R . M . (1990). The evolutionary dynamics of HIV-1 quasispecies and the development of immunodeficiency disease . AIDS 4, 1095-1103 .
HIV-1 ANTIGENIC DRIFT
205
PALKER, T . J ., CLARK, M . E ., LANGLOIS, A . J ., MATTHEWS, T . J ., WEINHOLD, K . J ., RANDALL, R . R ., BOLOGNESI, D . P ., and HAYNES, B . F . (1988) . Type-specific neutralization of the human immunodefi-
S ., WOLF, H ., PARKS, E . S ., PARKS, W . P ., JOSEPHS, S . F ., GALLO, R . C ., and WONG-STAAL . F. (1986) . Identification and characterization of conserved and variable regions in the envelope gene of
ciency virus with antibodies to env-encoded synthetic peptides . Proc . Nat?. Acad. Sci. USA 85, 1932-1936 . RUSCHE, J . R ., JAVAHERIAN, K ., MCDANAL, C ., PETRO, J ., LYNN, D . L ., GRIMAILA, R., LANGLOIS, A ., GALLO, R . C ., ARTHUR, L . 0 ., Fis-
HTLV-III/LAV, the retrovirus of AIDS . Cell 45, 637-648TAKEUCHI . Y ., AKUTSU, M ., MURAYAMA, K ., SHIMIZU .. N ., and HOSHINO, H . (1991) . Host range mutant of human immunodeficiency virus type 1 : modification of cell tropism by a single point mutation at
CHINGER, P . J ., BOLOGNESI, D . P ., PUTNEY, S . D ., and MATTHEWS, T. J . (1988) . Antibodies that inhibit fusion of human immunodeficiency virus-infected cells bind a 24-amino acid sequence of the viral envelope, gpl20 . Proc . Natl . Aced. Sci . USA 85, 3198-3202 . SHAG, M . S ., HAHN, B . H ., GIesoNs, J ., Li, Y ., PARKS, E . S ., PARKS, W . P ., and SHAW, G. M . (1988) . Extensive variation of human immunodeficiency virus type 1 in vivo . Nature (London) 334, 440-444 . SALINOVICH, 0 ., PAYNE, S . L., MONTELERO, R. C ., HUSSAM, K . A ., ISSEL, C . J ., and ScHNORR, K . L . (1986). Rapid emergence of novel antigenic and genetic variants of EIAV during persistent infection . J. Virol. 57, 71-80 . SIMMONDS, P ., BALFE, P ., LUDLAM, C . A ., BISHOP, J . 0 ., and LEIGH BROWN, A . J . (1990) . Analysis of sequence diversity in hypervari-
the neutralization epitope in the env gene . l. Virol. 65, 1710-1718 . WOLF DE, F ., LANGE, J . M . A ., HOUWELING, J . T . M, COUTINHO, R . A., SCHELLEKENS, P . T ., NOORDAA VAN DER, .J . . and GOUDSMIT, 1 . (1988) . Numbers of CD4+ cells and the levels of core antigens of and antibodies to the human immunodeficiency virus as predictors of AIDS among seropositive homosexual men . J. Infect. Dis . 158(3), 615-622 . WOLFS, T . F . W ., JONG DE, J ., BERG VAN DEN, H ., TUNAGEL, J . M . G . H ., KRONE, W. J . A ., and GGUDSMIT, J . (1990) . Evolution of sequences encoding the principal neutralization epitope of HIV -1 is host-dependent, rapid and continuous . Proc. Natl Aced. Sci. LISA 87, 9938-9942 . ZWART, G ., LANGEDIJK, H ., HOEK VAN DER, L., JONG DE, J . J ., WOLFS,
able regions of the external glycoprotein of human immunodeficiency virus type 1 .1. Virol. 64(12), 5840-5850 . STARCICH, B . R ., HAHN, B . H ., SHAW, G . M ., McNEELY, P . D ., MODROW,
T . F . W ., RAMAUTARSING, C ., BAKKER, M ., RONDE DE, A ., and GOUDSMIT, J . (1991). Immunodominance and antigenic variation of the principal neutralization domain of HIV 1 . Virology 181, 481-489,
