The polymerase chain reaction in the diagnosis of vertically transmitted HIV infection

Diagnosis of HIV-1 infection in infants using dried blood spots in Tamil Nadu, South India

INTRODUCTION

Diagnosis of HIV infection in infants is difficult due to the presence of maternal antibodies; only nucleic acid assays are very helpful in early detection. Filter papers are especially useful for blood collection in resource-poor settings with limited access to diagnostic facilities. MATERIALS #ENTITYSTARTX00026;

METHODS

DBS samples were collected from the infants born to HIV seropositive mothers who had received single dose nevirapine at onset of labor. The samples were directly spotted onto the Whatman 903 cards from heel, big toe or finger prick depending on the age of the infants. A total of 766 infant samples were collected on dried blood spots (DBS) and transported to the Department of Experimental Medicine (DEM), Chennai, for testing from different government hospitals of rural and urban parts of Tamil Nadu, South India. According to National AIDS Control Organization's (NACO) protocol DNA was extracted from all these DBS and PCR was performed using the Roche kit version 1.5.

RESULTS

Fifteen infants were found to be HIV positive and 751 were HIV negative; all these 15 positive infants and 49 negative infants who were in the age group between 10 and 18 months were repeated with another DBS and compared with whole blood. The DBS results were concordant with the whole blood method and the sensitivity and specificity were 100%.

Evaluation of the prototype Roche DNA amplification kit incorporating the new SSK145 and SKCC1B primers in detection of human immunodeficiency virus type 1 DNA in Zimbabwe

We assessed the sensitivity and specificity of a newly developed DNA PCR kit (Roche Diagnostic Corporation, Indianapolis, Ind.) that incorporates primers for all the group M viruses for the detection of human immunodeficiency virus (HIV) type 1 (HIV-1) infection in Zimbabwe. A total of 202 whole-blood samples from adults whose HIV status was known were studied. This included 100 HIV-1-positive and 102 HIV-1-negative samples selected on the basis of concordant results obtained with two enzyme-linked immunosorbent assay kits. The prototype Roche DNA PCR assay had a 100% sensitivity for the detection of HIV-1 DNA and a specificity of 100%. We conclude that the new Roche DNA PCR kit is accurate for the detection of HIV DNA in Zimbabwean samples, in which HIV-1 subtype C dominates.

The placental barrier to maternal HIV infection

Infection with HIV destroys the immune system and causes acquired immunodeficiency syndrome (AIDS). Death results from common bacterial and opportunistic infections that are rare in persons with a healthy immune system. HIV infection frequently is a fatal sexually transmitted disease that can also be transmitted from an infected mother to her offspring.

HIV-1 detection by nested PCR and viral culture in fresh or cryopreserved postmortem skin: potential implications for skin handling and allografting

AIMS

To date, the risk relating to the handling or allografting of human immunodeficiency virus type 1 (HIV-1) infected postmortem skin remains hypothetical. While blood screening for HIV antibodies is still the key safety procedure to detect HIV infected cadavers, false negative results are a concern. Conversely, false positive results may hamper the collection of skin allografts. Accordingly, viral culture was used to clarify skin infectivity and the nested polymerase chain reaction (PCR) was used to assess the reliability of skin PCR testing.

METHODS

Viral culture and nested PCR performed with gag and pol specific primers were investigated in cadaveric skin and blood from 12 HIV-1 infected patients. Samples were collected repeatedly between one and five days in seven patients. In most cases, analyses were performed on triplicate skin samples: fresh (n = 26); cryopreserved in 5% dimethylsulphoxide (n = 21), or cryopreserved in 30% glycerol (n = 26).

RESULTS

HIV was isolated in two of 26 cultures of fresh skin specimens (8%), seven of 47 cryopreserved skin specimens (15%), and eight of 26 blood specimens (31%). The nested PCR detected HIV-1 in all skin samples (n = 73), regardless of the postmortem interval or cryopreservation. In blood, a positive signal was found in eight of 12 patients but two of them had discordant results on successive samples.

CONCLUSIONS

These data suggest that nested PCR on postmortem skin samples can detect HIV more reliably than on blood. They also demonstrate the potential viral infectivity of fresh or stored skin postmortem samples in HIV infected patients. They underscore the need for caution during the handling of skin tissue from HIV infected cadavers and confirm the potential risk related to accidental allografting of HIV contaminated skin.

Influence of chemical allograft preservation procedures on the human immunodeficiency virus

Since chemically preserved allogenic transplants have an established place in reconstructive procedures, the possibility of transferring the human immunodeficiency virus (HIV) with these transplants has been intensively discussed. In this study the authors obtained brain and spleen samples from six HIV-infected cadavers and preserved them with Merthiolate, Cialit, and formaldehyde. After preservation, the tissues were examined for proviral HIV-1 DNA (gag, pol, env) using the polymerase chain reaction. Proviral sequences were clearly demonstrated after the preservation procedure. The results of this study indicate that HIV remains in tissues that have been treated with Merthiolate, formaldehyde, or Cialit. Further investigations are necessary to determine if the virus is in an inactivated or activated form. It can be concluded that, because of the possible transmission of HIV by chemically preserved homografts, serologic screening of donors should be mandatory.

Use of polymerase chain reaction and quantitative antibody tests in children born to human immunodeficiency virus-1-infected mothers

The diagnosis of human immunodeficiency virus (HIV) infection in children born to HIV-infected mothers is complicated by the presence of passively acquired maternal antibodies, and exclusion of infection in these infants remains problematic. The use of genome detection by polymerase chain reaction (PCR) amplification and the quantification of anti-HIV-1 antibodies were examined as methods for early diagnosis. Blood samples were taken from 84 non-breast-fed infants of HIV-infected mothers in five Italian and Spanish centres, a subgroup of children enrolled in the European Collaborative Study (ECS) for whom clinical and immunological information has been documented from birth. Whole blood was added to glycigel cryopreservative, stored, and tested in the United Kingdom by a nested PCR method. Antibody to HIV-1 was detected and quantified by titration using a gelatin particle agglutination test. PCR sensitivity and specificity were assessed. Twenty-one of the 84 children tested were infected. The estimated PCR sensitivity ranged from 0% (95% CI 0-26%) on day 1, 57% (19-85) on day 7, to 63% (33-92) on day 30. The negative predictive value of PCR ranged from 85% (83-88) on day 0 to 98% (94-100) at 3 months of age. On average, the level of maternal antibody halved every 33 days (31-36.5) in uninfected children. Between 6 and 9 months of age, increases in antibody titres in infected children were not more informative than absolute levels. These findings suggest that antibody measurement may supplement genomic diagnosis and that this collection method provides an alternative to the use of dried blood spots.

HIV Infection in Children Born Before and After Immigration to Belgium

Methods: The records of 119 children, either infected with HIV or born to HIV-infected mothers were reviewed. This group was mainly represented by children whose families came to Belgium from Central Africa. Seven children were infected through blood or blood products transfusion, whereas the remaining 112 were born to infected mothers. Most of the latter were infected through heterosexual contacts. Results: The overall rate of perinatal transmission was 44%. The transmission rate observed in our population is higher than in other European studies and closer to African studies. Lymphadenopathy was observed in 86% of infected children versus 16% in noninfected children during the first year of life (p =.0002). Serum IgG level was higher in infected children than in noninfected children. Conclusions: HIV transmission is firmly related to breast feeding.

Detection of HIV-1 proviral sequences in lymphocytes using a qualitative polymerase chain reaction assay

The performance and clinical relevance of a qualitative PCR-based assay for the detection of HIV-1 DNA sequences in peripheral blood mononuclear cells (PBMCs) was evaluated by two different laboratories. Four hundred and one samples were obtained from 397 individuals from different risk populations. All blood donors tested had negative results; positive signals were obtained from all infected patients. HIV-1 DNA was detected in 3 of 17 infants born to seropositive mothers; Western blot indeterminate blood donors and exposed health-care workers had negative results. Our results demonstrate that this PCR assay provides both sensitive and specific results and is suitable for testing large numbers of samples and for rapid identification of HIV-1 infection.

Diagnosis and prediction of pediatric HIV-1 infection and AIDS: current status

The increase in the incidence of HIV-1 infection in women of child bearing age has resulted in a surge in the number of cases of pediatric AIDS. The World Health Organization (WHO) has estimated that the number of cases of pediatric AIDS worldwide will be at least 10 million by the year 2000. This alarming statistic underscores the need for accurate prediction and diagnosis of pediatric HIV-1 infection which is of paramount importance for the initiation of effective therapeutic interventions. Since circulating maternal anti-HIV-1 antibody persists in the baby for up to 21 months, early conventional serological diagnosis of infection is not possible. Other methods for diagnosis of HIV-1 infection in a child less than 2 years of age have been utilized including the polymerase chain reaction (PCR), measurements of the HIV-1 p24 core protein and anti-HIV-1 IgA, as well in vitro measurements of antibody producing cells. In addition, the ability to predict HIV-1 infection in the child based upon maternal humoral immune responses to the envelope glycoprotein has also been suggested. This review summarizes the recent serological, biological and molecular methodologies used to predict and diagnose pediatric HIV-1 infection and AIDS.

Failure to detect silent HIV infection by polymerase chain reaction in subjects at risk for heterosexually transmitted HIV type 1 infection

The prevalence of silent HIV-1 infections in subjects at risk of acquiring HIV infection by heterosexual transmission was assessed using a nested polymerase chain reaction (PCR) assay. The two groups of risk subjects consisted of 92 female prostitutes and 43 heterosexual partners of infected individuals. Appropriate positive and negative control persons were included in the study. Serum samples were also tested for antibody to HIV-1 by an enzyme immunoassay (EIA) and positive results confirmed by Western blot. PCR results in the two risk groups and the positive and negative controls were in full agreement with serological results. It is concluded that silent infection with HIV-1 is infrequent in persons at risk for heterosexual transmission.

Diagnosis of HIV-1 infection by PCR with two primer pairs

We have used two different primer pairs to assess HIV-1 infection of peripheral blood mononuclear cells (PBMC) by polymerase chain reaction (PCR). The study was carried out on 150 individuals: 50 seronegative individuals without risk behaviours for HIV-1 infection, 50 individuals with risk-behaviours but seronegative for HIV-1 and 50 patients with risk-behaviours who were HIV-1 seropositive. Discordances were found between the two primer pairs (SK38/39; SK68/69) in 3 cases. In the non-risk seronegative group, one specimen was scored positive with only one primer pair (SK38/SK39); all the samples belonging to seropositive individuals were found to be positive for HIV-1 DNA using both primer pairs; in the seronegative at risk group 2 samples were positive with only one primer pair (SK38/SK39), and 4 samples were found positive by both primer pairs (SK38/39 & SK68/69). Our study demonstrates that discrepant results take place with relatively high frequency; we propose that all specimens should be tested twice using at least two different primer pairs.

Primer extension analysis provides a sensitive tool for the identification of PCR-amplified DNA from HIV-1

Primer extension analysis was evaluated as a means to identify PCR-amplified DNA from HIV-1. Solution hybridization with radioactive labeled oligonucleotides followed by an extension reaction with Klenow fragment of Escherichia coli DNA polymerase I and subsequent separation on denaturing polyacrylamide gels reveals single stranded DNA products of the predicted size. The specificity of the reaction is further demonstrated by specific endonuclease digestion. The analysis is more sensitive than Southern blotting and about equally sensitive as Slot blot analysis when double-stranded DNA probes are used for hybridization. With end-labeled oligonucleotide probes, primer extension analysis proved an order of magnitude more sensitive than membrane hybridization. The analysis also allows quantitation of amplified DNA from 1 pg to about 1 ng of DNA product. Under the conditions described for amplification, primer extension analysis is capable of detecting a single HIV-1 plasmid DNA molecule in the presence of 1 microgram of total DNA. 3'-end mismatching of the oligonucleotide probe does not result in a significantly altered detection limit. Primer extension analysis can also be carried out with at least three different PCR-amplified DNAs in the same reaction tube.

Rapid serologic testing with immune-complex-dissociated HIV p24 antigen for early detection of HIV infection in neonates. Southern California Pediatric AIDS Consortium

BACKGROUND

Serologic detection of human immunodeficiency virus (HIV) infection in neonates is complicated by the presence of immune complexes, consisting of passively transferred maternal antibodies and HIV antigens. A new, rapid assay has been designed to disrupt these immune complexes in order to permit the detection of a specific HIV antigen. We evaluated the efficacy of this assay in detecting HIV infection in neonates.

METHODS

We measured p24 antigen in blood samples from both infected and uninfected children of HIV-infected mothers. The samples were treated with glycine hydrochloride to dissociate the immune complexes, followed by neutralization with TRIS-hydrochloric acid. A commercial HIV p24 antigen assay was then used, with an optical density greater than 0.120 at a wavelength of 450 nm defined as indicating a positive result.

RESULTS

Of eight cord-blood samples from neonates with proved HIV infection, five were positive for immune-complex-dissociated p24 antigen. For two other neonates the first postnatal sample, obtained on days 12 and 18, was positive. There was no follow-up sample for the eighth neonate. Of 22 uninfected neonates, 20 were negative on the cord-blood assay. Two neonates had positive cord-blood samples, but the first postnatal sample was negative. Thus, the tests with early postnatal samples identified the HIV-infection status correctly for all 29 children who could be evaluated. In a separate group of 78 children (median age, 188 weeks), the specificity of the test was 100 percent and the sensitivity 81 percent.

CONCLUSIONS

The immune-complex-dissociated HIV p24 antigen assay is a rapid, simple serologic test that may be of value in diagnosing HIV infection in neonates born to HIV-infected women.

The use of viral culture and p24 antigen testing to diagnose human immunodeficiency virus infection in neonates. The HIV Infection in Newborns French Collaborative Study Group

BACKGROUND

Early diagnosis of human immunodeficiency virus (HIV) infection in infants born to infected mothers is important for the infants' medical care, but the presence of maternal antibodies makes serologic tests uninformative.

METHODS

In a cohort study of 181 infants born to HIV-infected mothers, we assessed the diagnostic value of HIV viral culture and testing for the presence of p24 antigen. The infants were tested at birth, again during the first 3 months, then followed and tested at the age of at least 18 months.

RESULTS

Of the 181 infants, 3 died of HIV infection and 37 were seropositive after the age of 18 months. Viral cultures at birth were positive in 19 of the 40 infected infants and in none of the uninfected infants, yielding a sensitivity of 48 percent (95 percent confidence interval, 32 to 63 percent) and a specificity of 100 percent (95 percent confidence interval, 97 to 100 percent). By the age of three months, 30 of the 40 infants (75 percent) had positive cultures; again, there were no false positive results among the infants who were tested a second time, of the 141 who remained uninfected. The sensitivity of testing for p24 antigen at birth was only 18 percent, with a specificity of 100 percent. The presence of p24 antigen at birth was associated with the development of early and severe HIV-related disease (P less than 0.04).

CONCLUSIONS

Viral culture at birth can correctly identify about half of newborns with HIV infection. The fact that this usually sensitive technique fails to identify about half the ultimately infected neonates suggests that vertical transmission of HIV may occur late in pregnancy or during delivery.

Infrequent detection of HIV-1 components in tears compared to blood of HIV-1-infected persons

Beside the risk of infection via HIV-1-contaminated blood, ophthalmologists are especially interested in the possibility of HIV-1 infection via tears. Therefore we tried to isolate HIV-1 from tears of 50 HIV-1-infected persons in different stages of disease by reverse transcriptase (RT) and by p24-antigen (p24-AG) in the cultures. Simultaneously we tried to isolate HIV-1 in the supernatant from peripheral blood lymphocytes (PBL), which was successful in 32 of the 50 examined specimens. HIV-1 could not be isolated from the tears of these persons. In addition, polymerasechain-reaction (PCR) was performed to detect proviral sequences (gag, pol, env) of HIV-1 in tears and blood of ten HIV-1-infected patients. While in all the examined patients gag, pol and env could be detected in the blood samples, only one tear sample was found positive for gag and pol DNA fragments. These results indicate that tears of HIV-1-positive contain extremely low quantities of tissue culture infectious doses (TCID) of HIV-1 in contrast to PBL. HIV-1 infection via tears therefore appears to be unlikely.

Laboratory methods for early detection of human immunodeficiency virus type 1 in newborns and infants

Cumulative data on serological testing of newborns and infants have shown that (i) maternal and newborn anti-HIV-1 IgG titers are high at delivery, which may explain the persistence of antibody in the infants of seropositive mothers; (ii) in some situations, serial HIV-1 antibody testing may identify infected infants; and (iii) detection of anti-HIV-1 IgA or IgM is specific for infection but the sensitivity of this assay may be compromised in certain situations, such as when infected infants are hypogammaglobulinemic or when the rise and fall of HIV-1-specific IgM synthesis following acute infection has been completed before delivery of the infant. Cumulative data on PCR, viral culture, and tests for antigen in newborns and infants have shown that (i) among all age groups, viral culture is probably the most specific test available for detection of HIV-1, as PCR and the p24 antigen test may (though rarely) give false-positive results; (ii) the sensitivity of these tests increases in the order of antigen, culture, and PCR, with relatively insensitive results in the first 3 months of life for all of these tests; (iii) the sensitivity of all of these tests improves and approximates 90 to 100% when infants over 6 months of age are tested; and (iv) data regarding the sensitivity, specificity, and usefulness of these virological assays in infants under 3 months of age are very scant and inconclusive.

Detection of human immunodeficiency virus type 1 infection in young pediatric patients by using polymerase chain reaction and biotinylated probes

Polymerase chain reaction (PCR) testing using up to four primer pairs and biotinylated probes was 97.9% sensitive (188 of 192 specimens positive) and 100% specific (267 of 267 specimens negative) for detecting the presence or absence of human immunodeficiency virus (HIV) DNA in peripheral blood mononuclear cells from pediatric patients whose HIV status has been confirmed. SK38/39 and SK145/150 were the most sensitive primer pairs, respectively detecting HIV DNA in 95.6 and 95.9% of peripheral blood mononuclear cell specimens from HIV-infected children and collectively detecting all adequately tested PCR-positive specimens. Primer pairs SK29/30 and SK68/69 respectively detected HIV DNA in only 76.4 and 76.6% of HIV-positive specimens. Among infants born to HIV-seropositive mothers, 30 who subsequently were confirmed to be infected were sampled when they were less than or equal to 6 months of age; in all but one infant, HIV DNA was found in the first specimen collected. Among the nine youngest infected infants tested, all were PCR positive by 38 days of age. PCR methods thus have reliably detected vertically transmitted HIV infection early in life.

In vivo persistence of a HIV-1-encoded HLA-B27-restricted cytotoxic T lymphocyte epitope despite specific in vitro reactivity

A large number of human immunodeficiency virus type 1 (HIV-1) specific HLA-restricted cytotoxic T cell (CTL) epitopes have been mapped, including an HLA-B27-restricted immunodominant epitope within p25gag. Accordingly, this segment of the HIV-1 provirus was amplified by the polymerase chain reaction from DNA derived from fresh uncultured peripheral blood mononuclear cells (PBMC) of four HLA-B27 HIV-1-infected individuals. In all cases the majority of infected PBMC bore sequences encoding the HLA-B27-restricted peptide. CTL escape mutants had not accumulated in vivo 8 and 14 months later despite demonstrable CTL activity in vitro. These data emphasize the importance of silently infected lymphocytes in evading immune surveillance.

Use of intravenous immunoglobulin in acquired immune deficiency syndrome

Patients infected with the human immunodeficiency virus (HIV) may have an antibody deficiency and a deficiency of cellular immunity. Intravenous immunoglobulin (IVIG) preparations may benefit HIV-infected children and adults with recurrent bacterial infections at doses of 200 to 400 mg/kg every 2 to 4 weeks. In addition, IVIG (1 to 2 g/kg) is effective at raising platelet counts to hemostatic levels in HIV-infected patients with idiopathic thrombocytopenic purpura and life-threatening bleeding. Indirect evidence also suggests that IVIG may be effective in preventing Pneumocystis carinii pneumonia. Finally, recent studies suggest that specific anti-HIV antibody preparations may have a therapeutic role, either as immunoglobulin concentrates or as immunoadhesions and immunotoxins. However, further investigations are needed to exclude antibody enhancement of HIV infection by the Fc receptor or the complement receptor.

Postnatal transmission of human immunodeficiency virus type 1 from mother to infant. A prospective cohort study in Kigali, Rwanda

BACKGROUND

Although transmission of human immunodeficiency virus type 1 (HIV-1) from mother to infant has been well documented during pregnancy and delivery, little is known about the possible transmission of HIV-1 during the postnatal period.

METHODS

We conducted a prospective cohort study in Kigali, Rwanda, of 212 mother-infant pairs who were seronegative for HIV-1 at delivery. All the infants were breast-fed. The subjects were followed at three-month intervals, with Western blot assays for antibodies to HIV-1 and testing of mononuclear cells by a double polymerase chain reaction (PCR) using three sets of primers. To evaluate potential risk factors, each mother who seroconverted was matched with three seronegative control women.

RESULTS

After a mean follow-up of 16.6 months, 16 of the 212 mothers became seropositive for HIV-1. Of their 16 infants, 9 became seropositive. One infant was excluded from the analysis because of a positive test by PCR on the blood sample obtained at birth. Postnatal seroconversion to HIV-1 occurred in four of the five infants born to the mothers who seroconverted during the first 3 months post partum, and in four infants of the 10 mothers who seroconverted between month 4 and month 21. In all cases, the infant seroconverted during the same three-month period as the mother. The main risk factor for maternal seroconversion was being single.

CONCLUSIONS

HIV-1 infection can be transmitted from mothers to infants during the postnatal period. Colostrum and breast milk may be efficient routes for the transmission of HIV-1 from recently infected mothers to their infants.

A rapid chemiluminescence detection method for PCR-amplified HIV-1 DNA

A solution hybridization assay using acridinium ester labelled probes is described for detection of amplified HIV-1 DNA segments. Amplification was achieved by 30 cycles of the polymerase chain reaction using SK38/SK39 primers specific for a constant region of the HIV-1 gag region together with HLA DQ alpha primers as internal control. Discrimination between hybridized and non-hybridized probes by differential hydrolysis resulted in a three log reduction of the chemiluminescence signal of the non-hybridized probe within 6 min without significant changes in the hybridized probes due to protection of the acridinium ester to hydrolysis by intercalation formation. Chemiluminescence was measured by a two-step-injection method with hydrogen-peroxide and NaOH. About 50 attomols of HIV-1 gag DNA could be detected. Chemiluminescence results, given in relative light units (RLU) of 159 HIV-1-infected patients (range 4013-458319) showed clear discrimination from 64 noninfected control samples (range 838-1477) (cut off 2000 RLU). Comparison with parallel detection of amplified products with autoradiography (32P) and ethidium bromide-stained agarose gels or a p24 antigen ELISA demonstrated better sensitivity and reproducibility of the chemiluminescence assay. The time required for the assay, including measurements, is less than 30 min, which allows reporting 'PCR results' on the same day.

Use of dried blood spot specimens in the detection of human immunodeficiency virus type 1 by the polymerase chain reaction

Dried blood spots (DBSs) constitute a potentially valuable source of material for human immunodeficiency virus (HIV) serologic and molecular testing. To facilitate molecular testing, we have adapted the polymerase chain reaction (PCR) to the detection of HIV proviral DNA in DBS samples. The method is highly reproducible, with 75 microliters of whole dried blood providing sufficient DNA for duplicate testing with three primer sets. By using DBS PCR, 66 of 69 (95.6%) seropositive at-risk individuals tested positive by at least two primer sets and 85 of 85 (100%) low-risk seronegative blood donors tested negative by all three sets of primers. The frequency of HIV DNA detection in seronegative at-risk individuals was low, with only 1 of 58 (1.7%) individuals testing positive. These results show that in a clinical environment, HIV PCR analysis of DBS specimens is specific and sensitive. The method is cost effective and presents a useful alternative to the isolation of HIV from seropositive babies with an undefined infection status.

Analysis of sequence diversity in hypervariable regions of the external glycoprotein of human immunodeficiency virus type 1

Nucleotide sequences in three hypervariable regions of the human immunodeficiency virus type 1 (HIV-1) env gene were obtained by sequencing provirus present in peripheral blood mononuclear cells of HIV-infected individuals. Single molecules of target sequences were isolated by limiting dilution and amplified in two stages by the polymerase chain reaction, using nested primers. The product was directly sequenced to avoid errors introduced by Taq polymerase during the amplification process. There was extensive variation between sequences from the same individual as well as between sequences from different individuals. Interpatient variability was markedly less in individuals infected from a common source. A high proportion of amino acid substitutions in the hypervariable regions altered the number and positions of potential N-linked glycosylation sites. Sequences in two hypervariable regions frequently contained short (3- to 15-bp) duplications or deletions, and by amplifying peripheral blood mononuclear cell DNA containing 10(2) or 10(3) proviral molecules and analyzing the product by high-resolution electrophoresis, the total number and abundance of distinct length variants within an individual could be estimated, providing a more comprehensive analysis of the variants present than would be obtained by sequencing alone. Sequences from many individuals showed frequent amino acid substitutions at certain key positions for neutralizing-antibody and cytotoxic T-cell recognition in the immunodominant loop. The rates of synonymous and nonsynonymous nucleotide substitution in the region of this and flanking regions indicate that strong positive selection for amino acid change is operating in the generation of antigenic diversity.