Main

In adults infected with HIV-1, an association between the in vitro phenotypic properties and replicative capacity of virus isolates and progression of disease has been described(1–4). Also, changes in phenotype, from NSI to SI HIV-1 strains are associated with a rapid decline in CD4+ lymphocyte counts and the progression to symptomatic disease(5). Furthermore, the biologic phenotype of HIV-1 isolates, as determined by the ability to replicate in MT-2 cells, is an independent marker for a poor prognosis(6).

Perinatally acquired HIV-1 infection follows a different clinical course than HIV-1 infection in adults, and the age at infection of HIV-1-infected children appears to be an important risk factor for disease progression(7–9). However, the described immunologic defects of HIV-1-infected children do not differ qualitatively from those described in adults(8, 10). Very little information exists concerning the relationship among in vitro phenotypic properties and the replicative capacity of HIV-1 isolates, the viral load, and the clinical manifestations of HIV-1 infection in pediatric patients(11–15). We have performed a prospective 28-mo follow-up of the evolution of HIV infection in a cohort of 13 HIV-infected children who were all less than 1 y of age. We have investigated the relationship among the HIV-1 biologic phenotype, replicative capacity of virus isolates, HIV-RNA copy number in plasma, p24 antigenemia, CD4+ T lymphocyte counts in peripheral blood, and the clinical status in this cohort of HIV-infected children.

METHODS

Patients. The patients included in the study were 64 infants admitted to our Division of Pediatrics at the University General Hospital of the Community of Madrid (ages ranging from 2 to 12 mo) who were born to HIV-1-infected mothers and had no history of hemophilia or blood transfusion. We performed a prospective follow-up lasting from 1 to 28 mo of the evolution of HIV infection in a cohort of 13 HIV-infected infants. The infants were classified according to the new Center for Disease Control and Prevention classification system for HIV infection in infants and children under 13 y of age(16).

Serologic analysis. Serum samples were tested for HIV-1-specific antibodies by Western blot (Pasteur-Sanofi)(17).

Extraction of RNA from plasma. RNA was extracted from plasma by a previously described method(18–19). Briefly, 200 μL of clarified plasma to which 200 μL of 5 M guanidinium thiocyanate had previously been added were extracted with phenol/chloroform and precipitated with isopropanol. The resulting pellet was then washed in 75% ethanol, dried, and dissolved in diethylpyrocarbonate-treated glass distilled water.

Reverse transcription and amplification of cDNA. HIV RNA was transcribed to cDNA using avian myeloblastosis virus reverse transcriptase(Boehringer Mannheim, Indianapolis, IN) by a previously described method(19, 20). Oligomers used for amplification included SK38, SK39, and SK19, all of whose sequences have been published previously(21). Biotinylation of SK38 and horseradish peroxidase labeling of probe SK19 were carried out as described previously(22). Amplification of HIV cDNA was carried out in 100μL of a reaction mixture containing 50 pmol of primers SK38 and SK39, 10 mM dNTP, 10 mM Tris (pH 8.3), 2.5 mM MgCl2, 50 mM KCl, and 2.5 U of recombinant Taq DNA polymerase (Perkin-Elmer, Norwalk, CT). The mixture was then overlaid with 50 μL of mineral oil. Tubes were placed in a DNA thermal cycler (Perkin-Elmer) for 30 cycles of amplification with the following program: 95 °C, 30 s; 55 °C, 30 s; and 72 °C, 60 s for denaturation, annealing, and extension, respectively. Negative and positive controls which included a standard curve were prepared by adding different amounts of a stock of HIV S61 virus (0, 1, 10, 50, 100, 500, 1000, 5000, and 10000 viral particles) to HIV seronegative control plasma.

Detection and quantitation of PCR products. To detect and quantitate the PCR products, 96-well microplates were coated with 100 μL of a 0.1 mg/mL solution of avidin (Sigma Chemical Co., St. Louis, MO) in 50 mM Na2CO3 (pH 9.6) overnight at room temperature. Wells were washed twice with PBS and then filled with 300 μL of a blocking solution containing 5 × Denhardt's solution, 1% gelatin (Sigma Chemical Co.), 250μg/mL sheared herring sperma DNA (Promega Biotech Inc., Madison, WI) and incubated at least overnight at 4 °C. Immediately before use, the blocking solution was aspirated from each well, and 5 μL of PCR product, 65 μL of a hybridization solution, containing 5 × saline sodium phosphate EDTA, 5× Denhardt's solution, and 1 pmol horseradish peroxidase-labeled SK19 HIV gag specific probe were added to each well. A capture and hybridization reaction was then carried out for 1 h at 42 °C. The 96-well microplates were washed five times with PBS containing 0.05% Tween-20. The horseradish peroxidase substrate O-phenylenediamine (Sigma Chemical Co.) was prepared at 0.6 mg/mL in 0.1 M citrate buffer (pH 5.5) containing 0.03% hydrogen peroxide. One hundred and fifty microliters of this substrate solution were added to each well. After 10 min, the reaction was stopped with 1 N H2SO4, and OD of each well was measured at 490 nm.

Detection of proviral DNA by PCR in PBMC. PBMC from all infants were analyzed for the presence of HIV-1 DNA by PCR with three different sets of nested primers specific for regions of the gag, pol, andenv genes (JA4 to JA7, JA17 to JA20, and JA13 to JA16, respectively) as previously described(23, 24). Briefly, the samples were first amplified for 24 cycles with the outer primers, then 0.1 (5μL) of the product from the first PCR reaction was amplified for 30 cycles with the inner primers. Preparation of samples was done in a biosafety laminar airflow hood, exclusively used for this purpose(25). Samples from all individuals gave positive signals with primers PCO3-PCO4 specific for the human β-globin gene(26) and were therefore suitable for amplification. Negative controls were included in each run and consisted of lysis buffer alone and samples from healthy blood donors. Positive controls consisted of samples from PBMC cultures infected with HIV-1 virus. The PCR product was analyzed by electrophoresis on 1.5% agarose gel stained by ethidium bromide. Samples were considered positive if at least two different primer sets repeatedly gave a specific amplification product.

Virus isolation. The HIV-1 virus was isolated by a microculture technique which is useful for pediatric patients, because it requires substantially fewer PBMC(27). PBMC were isolated from blood by Ficoll-Hypaque density gradient centrifugation (Pharmacia)(27, 28). Both viral culture and PCR assay were performed on the same sample; PBMC (2 × 106) from patients were cultured with an equal number of PBMC from healthy donors. Donor PBMS had been pretreated for 48 h with PHA (Difco Laboratories, Detroit, MI) in RPMI 1640(Life Technologies, Inc., Paisley, Scotland, UK) supplemented with 20% heat-inactivated fetal bovine serum, 5% IL-2 (Boehringer Mannheim), 100 IU/mL penicillin, and 100 μg/mL streptomycin (Life Technologies, Inc.). Fresh PBMC (2 × 106) were added to the culture once a week. Supernatants were collected twice a week and tested for the presence of p24 HIV-1 antigen with the use of a commercially available kit (Elavia-Ag I, Diagnostics Pasteur)(19). The criteria for a positive coculture and determination of the days to the first positive result were based on HIV p24 results obtained from biweekly culture samples(11). According to the lag phase before p24 antigen detection and to the level of p24 antigen production by PBMC, HIV-1 isolates from these patients were classified as rapid/high or slow/low. The HIV isolates were operatively defined as rapid replication output (rapid/high), because p24 antigen was detected in PBMC supernatants as early as 3 d after being cultured, and moreover they produced persistently high p24 levels. In contrast, HIV isolates that had a lag phase of ≥7 d and thereafter produced consistently much lower p24 levels were designated as slow/low.

Determination of the viral phenotype. PHA-stimulated PBMC were infected with cell-free supernatant collected from each virus isolate, and after 7-10 d, the infected PBMC were cocultured with the MT-2 cell line. Cultures were maintained in the same medium as described for the PBMC cultures but without IL-2. Half of the cultures were replaced with fresh culture medium two times a week. MT-2 cell and infected PBMC cultures were checked for syncytium formation under light microscopy(29).

HIV-p24 antigen assay. To improve detection of the levels of HIV-p24 antigen in serum, the circulating immune complexes were dissociated(ICD p24) by a method based on low pH treatment of the patient's sample before using an ELISA (ELAVIA-Ag I, diagnostics Pasteur)(19–24). Briefly, four volumes of serum were mixed with one volume of 0.5 N HCI. After incubation of the mixture for 90 min at room temperature, the treated patient's serum was neutralized with one volume of 0.5 NaOH, mixed, and immediately added in two wells of the reaction microplate. After this treatment on the patient's serum the ELISA procedure was performed according to the manufacturer's instructions.

Quantification of T cell subsets in peripheral blood. T lymphocyte subsets in peripheral blood were quantified by direct immunofluorescence using MAb of the T series and flow cytometry (FACScan, Becton Dickinson, Franklin Lakes, NJ) as previously described(17).

Statistics. The statistical differences in the values of HIV RNA viremia and in the percentages of CD4+ T lymphocytes between both groups of infants (group 1, six infants infected with rapid/high and SI HIV-1 isolates; group 2, seven infants infected with slow/low and NSI HIV-1 isolates) were analyzed by a nonparametric test (Mann-Whitney). To reassure the statistical significance of the differences in the percentages of CD4+ T cells between both groups of infants, an analysis of variance was used in which the age of infants was included as a covariate. Binomial distribution values were used to assure the significance of proportions of cases with AIDS in both groups of infants. Binomial distribution values were also used to assure the significance of proportions of cases with detectable p24 antigen in serum in both groups of infants. Association between variables was measured by means of the nonparametric Spearman's rank-correlation coefficients.

RESULTS

On the basis of nested PCR and viral culture findings, 13 (20%) of the 64 infants born from HIV-1 seropositive mothers were identified as infected with HIV-1. HIV-1 isolates with SI ability were isolated in 6 out of 13 (46%) infants, whereas NSI HIV-1 isolates were isolated in 7 out of 13 (54%)(Table 1). HIV-1 isolates were determined to have SI capacity by cocultivating the patient's PBMC either with healthy donor PBMC or MT-2 cell types.

Table 1 Virologic, immunologic, and clinical status in 13 HIV-infected children
Full size table

After coculturing the cells from the HIV+ infants with normal PHA-stimulated PBMC, two patterns of viral replication emerged. Six HIV isolates were operatively defined as rapid replication/high output(rapid/high) (Fig. 1A). By contrast, seven isolates were designated as slow/low (Fig. 1B). We have found a positive correlation between the replication rate of HIV isolates and their capacity to induce syncytia in vitro (Table 1).

Figure 1
figure 1

HIV-1 replication on primary PBMC cocultures. PBMCs from HIV+ infants were cocultured with PHA-stimulated-PBMCs from healthy donors, and culture supernatants were tested for p24 HIV antigen. According to lag phase and slopes of the p24 production curves, HIV-1 isolates were classified as rapid/high (panel A) and slow/low (panel B).

Full size image

Distinct differences were seen in the capacity of sequential isolates from nine infants to induce syncytium formation. Sequential isolates from patient nos. 8 to 11 who initially had NSI isolates remained unable to induce syncytium formation in MT-2 cell cultures irrespective of the time of isolation (Table 1). However, early HIV-1 isolates derived from patient nos. 6 and 13 who initially had NSI isolates were unable to induce syncytium formation, whereas later isolates induced syncytium formation in MT-2 cells (Table 1). Serial HIV-1 isolates derived from patient nos. 1, 2, and 4 who initially had SI isolates were uniformly able to induce syncytium formation irrespective of the time of isolation during follow up (Table 1). A direct association between syncytium formation in the MT-2 cell cultures with serial isolates from patient nos. 1, 2, 4, 6, and 13 and disease progression was observed.

HIV-1 ICD p24 antigen was detectable in the serum of 5 out of 13 (38%) HIV infected infants. In these infants four out of 5 (80%) had rapid/high and SI HIV-1 isolates. On the other hand, two out of 8 (28%) infants with negative p24 antigenemia had rapid/high and SI HIV-1 isolates (Table 1). These results indicate that there is a trend for the presence of detectable p24 antigen in serum to be associated with the SI capacity and the high replication rate of HIV-1 isolates, although this was not found to be statistically significant.

HIV-1 RNA in plasma was detectable in the 13 HIV-infected infants. At the initiation of the study, patients infected with rapid/high and SI HIV-1 isolates showed a significantly higher viral load than those with slow/low and NSI HIV-1 isolates (p < 0.001) (Table 2). We also analyzed the relationship among the percentage of CD4+ T lymphocytes, the viral phenotype, the viral replicative ability, and the viral load. Infants infected with rapid/high and SI HIV-1 isolates and with high levels of HIV RNA in plasma had a significantly lower percentage of CD4+ T cells and more severe disease than those with slow/low and NSI viral isolates and lower levels of HIV RNA (Table 2). The association of SI phenotype with lower CD4+ lymphocyte percentages was also statistically significant by using an analysis of variance in which the age of infants was included as a covariate (analysis of variance, p< 0.05). On the other hand, there was a strong negative correlation between the percentage of CD4+ T cells and the HIV RNA copy number/mL in plasma(r = -0.826).

Table 2 Relationship among the viral phenotype and replicative ability of HIV-1 isolates with the viral load and clinical status in 13 HIV-infected infants
Full size table

The results of quantitative RNA PCR studies on sequential plasma samples from infants nos. 1, 2, and 6 showed an association between high levels of HIV RNA, SI phenotype, CD4+ T cell depletion, and clinical progression(Table 1). In infant no. 6, an increase in the levels of HIV-1 RNA was seen concurrent with the switch from NSI-slow/low to SI-rapid/high phenotype and with the CD4+ T cell decline. In contrast, infants whose viruses exhibited an NSI phenotype throughout the follow-up showed lower levels of HIV RNA, stable CD4+ T cell counts, and mild symptomatic HIV infection (Table 1).

In summary, the infants with more severe clinical symptoms carried the viruses with more cytopathic biologic characteristics, had higher viral load in blood, and had lower CD4+ T cell counts.

DISCUSSION

Recently, it has become increasingly important to know which are the virologic and immunologic factors that relate to the rapid or slow progression of HIV infection in pediatric patients(8, 30–36). However, correlation of viral burden, phenotype and replication rate, p24 antigen, CD4+ count, and clinical status has not been comprehensively analyzed in an HIV-infected pediatric population in the past. In HIV-infected children younger than 12 mo of age we have found a complete positive correlation between the replication rate of HIV isolates and their capacity to induce syncytia in vitro. Our studies in sequential HIV-1 isolates derived from HIV-infected infants showed that HIV-1 isolates with SI phenotype had increased levels of viral replication in PHA-stimulated PBMC cultures (rapid/high) compared with HIV-1 isolates with NSI phenotype that were slow/low. The ability of HIV-1 isolates to induce syncytia and their pattern of replication in vitro also correlated with the stage of the disease. Most of serial HIV-1 isolates obtained from infants with AIDS had the rapid/high phenotype and induced syncytia, whereas only two out of 23 HIV-1 isolates obtained from infants without AIDS showed these properties. This datum is in agreement with other recent studies in HIV-infected infants(11, 29) and in HIV-infected adults(4–6, 37, 38). However, two other studies in HIV-infected children(12, 13) have concluded that HIV-1 isolates from children with or without AIDS have similar in vitro biologic properties (viral phenotype and replication rate). This is a controversial subject that due to the limited sizes of the studies of the above-mentioned reports, requires further investigation in large scale prospective studies.

Virus load and virus phenotype have both been indicated as potential markers of disease progression in HIV infection(11, 14, 39, 40). In our study there was observed a significant association between the viral growth patternin vitro, the HIV-1 copy number in vivo, and the severity of HIV-1 infection. In sequential analysis of HIV-1 isolates from infants with AIDS, the presence of viral isolates with rapid/high and SI phenotype was associated with higher levels of HIV-1 RNA in plasma. In agreement with these results other authors have observed a direct association between clinical progression, increase in viral burden, and emergence of a more cytopathic viral phenotype(14, 39). Another study has shown that both rapid progression to AIDS and asymptomatic clinical courses were observed in infants with persistently high levels of virus(41). In three of our cases of infants with AIDS and with HIV-1 isolates with SI phenotype the substantial increase in the levels of HIV-1 RNA was observed associated with a profound CD4+ T cell decline. This datum is in agreement with another report in a HIV-infected adult population(39). It has been suggested that an increase in viral load linked with increased virulence of the HIV isolate may have a synergistic effect resulting in a rapid decline of CD4+T lymphocytes(39). Cumulatively, the findings in this study indicate that children who carried viruses with more cytopathic biologic phenotype and who had higher viral RNA copy numbers in their blood were more likely to have lower CD4+ cell counts and were more likely to have AIDS.

Using PCR for detection of HIV RNA in plasma we have found that all of the HIV-infected infants had an active replicative infection. However, p24 antigenemia was not always detected in the AIDS and non-AIDS children. Serum p24 antigen was positive in four out of six infants with rapid/high-type HIV-1 isolates and with high HIV-1 copy number, whereas the seven infants with slow/low-type HIV-1 isolates and a significantly lower HIV-1 copy number had undetectable or low levels of serum p24 antigen. There was not a good correlation between the p24 antigenemia and the number of HIV-1 copiesin vivo. Our results indicate that HIV detection by PCR in plasma is more sensitive than p24 antigenemia for the assessment of viral replicationin vivo in HIV-infected infants(40, 41).

The factors that influence the evolution of HIV infection in children are currently unknown but are likely to include both viral factors and host immunologic factors(42). The link between viral load, virus phenotype, and clinical course does not necessarily imply a cause an effect relationship. It is conceivable that, in certain children vertically infected with HIV, during the first year of life, the existence of an immature or deficient immune system could account for the intensive viral replication and dissemination(43).

This study has been limited to the evaluation of children ≤12 mo of age and suggests that in some patients the viral phenotype is fixed very early in life. Our investigation addresses multiple prognostic features of HIV-1 infection which may be helpful in characterizing prognostic groups in HIV-1-infected children.