Distribution of CCR5Δ32 in Human Immunodeficiency Virus-Infected Children and Its Relationship to Disease Course

Abstract

Homozygosity for a 32-bp deletion in the CCR5 gene (CCR5Δ32) has been shown to confer resistance to infection with the macrophage-tropic strain of human immunodeficiency virus (HIV) type 1. We examined the distribution of CCR5Δ32 in 47 children (age range, 1.5 to 19 years), of whom 43 were infected with HIV, by the perinatal route (n = 41) or by the intravenous route (n = 2). The infected patients were classified as rapid progressors (RP) (n = 7) (CDC category C3 or death by 2 years of age), non-rapid progressors (NRP) (n = 17) (survival for ≥8 years after infection), or intermediate (n = 19). CCR5Δ32 heterozygosity was found in two HIV-infected children, both NRP. None of the subjects were homozygous for CCR5Δ32, and the remaining children had no evidence of CCR5Δ32. The presence of CCR5Δ32 heterozygosity in 4.8% of this, predominantly non-Caucasian population is consistent with the published distribution of the mutation. The finding that CCR5Δ32 was present only in NRP and not in any RP is in agreement with previous reports suggesting that heterozygosity for CCR5Δ32 may confer limited protection from disease progression.

Viral, immunologic, and genetic host factors are known to influence the risk of human immunodeficiency virus (HIV) infection and the rate of disease progression. Although the role of the CD4 molecule as a high-affinity receptor which is necessary but not sufficient for HIV entry has been known for a long time, it is only recently that the identity and contribution of other coreceptors in HIV infection and disease pathogenesis have begun to be recognized. One such receptor is the β-chemokine receptor 5 (CCR5) to which macrophage-tropic (M-tropic) strains of HIV-1 must bind in order to enter the CD4 cell (1, 4, 8, 9). β Chemokines RANTES, MIP-1α, MIP-1β are the natural ligands of CCR5 and inhibit the entry of M-tropic viruses into the cells (5). M-tropic non-syncytium-inducing viruses are important for establishing infection by the mucosal route and through inoculation of blood and are believed to play a role in mother-infant transmission of HIV (16). They are the predominant viruses early in the disease course. In contrast, viruses in advanced disease are predominantly T cell tropic, syncytium-inducing viruses (6, 14, 17) and utilize the other coreceptor, CXCR-4 (previously known as fusin or LESTR), which like CCR5 is a seven-transmembrane protein whose ligand has been identified as the stromal cell-derived factor (10).

A genetic mutation in the CCR5 gene consisting of 32-bp deletion (CCR5Δ32 [Δ32]) results in a nonfunctional chemokine receptor. Homozygosity for this allele (Δ32/Δ32) confers strong resistance to infection by HIV (12, 13). The Δ32 allele has also been reported to influence the rate of disease progression, although the data is conflicting (7, 11).

The frequency of the Δ32 gene in the HIV-infected pediatric population and its influence on disease progression in HIV-infected children are relatively unknown. The objectives of this study were to assess the distribution of Δ32 in a population of HIV-infected children and to study the effect of the Δ32 mutation on the course of the disease. No child was found to be homozygous for Δ32. Two infected subjects, both non-rapid progressors, were heterozygous for Δ32 (wild type [wt]/Δ32), but the Δ32 allele was not identified in any child with rapid disease progression. Although the difference in the prevalence of wt/Δ32 in rapid disease progressors compared to non-rapid disease progressors did not achieve statistical significance (P = 0.076), analysis of disease severity in this cohort suggests that Δ32 heterozygosity may contribute to slowing disease progression.

MATERIALS AND METHODS

Patients.

Forty-seven pediatric patients (43 HIV infected, 2 exposed to HIV [mothers were HIV positive], and 2 not exposed to HIV [mothers were HIV negative]) were evaluated. The infected children had acquired disease through the perinatal route (n = 41) or through infected blood products (2 hemophiliacs). Children in the perinatal-infection group ranged in age from 1.7 to 18.8 years. There were 24 males and 17 females. The two hemophiliacs were 16.6 and 19.5 years of age. Ethnically there were 63% African-Americans, 15% Caucasians, 10% Latinos, 5% of mixed parentage, 5% Asians, and 2% of unknown ethnic background.

Disease severity was determined by using the CDC (Centers for Disease Control and Prevention)-defined clinical and immunologic criteria for classification of disease in children with HIV infection (3). Rapid progressors were defined as children who died or developed severe clinical disease (CDC category C) or severe immune suppression (CDC immune category 3) by 2 years of age. Patients who survived for ≥8 years were considered non-rapid progressors; a subset of non-rapid progressors who had survived for 12 to 17 years after acquisition of perinatal infection were designated long-term survivors. Children who were neither rapid nor non-rapid progressors and ranged in age from 2 to 8 years were considered to have an intermediate disease course.

Based on disease severity, 7 (17%) perinatally infected children were classified as rapid disease progressors, and 15 (37%) were classified as non-rapid disease progressors; 7 of the non-rapid progressors were considered to be long-term survivors. There were 19 (44%) children who had an intermediate disease course. The two hemophiliacs who had survived for longer than 8 years after acquisition of infection were considered non-rapid progressors.

CCR5 genotyping.

(See reference 11.) DNA was isolated from stored peripheral blood mononuclear cells of 43 HIV infected, 2 uninfected HIV-exposed, and 2 uninfected unexposed children. To detect the presence or absence of the Δ32 allele, a PCR-based assay using primers flanking the 32-bp segment was used to amplify this region of the CCR5 gene. The presence or absence of the deletion was then determined by gel electrophoresis as follows.

Briefly, a portion of the CCR5 gene was amplified by PCR from genomic DNA and analyzed on a 4% Metaphore agarose gel (FMC BioProducts). Primers CCR5c, 5′-CAAAAAGAAGGTCTTCATTACACC-3′, and CCR5d, 5′-CCTGTGCCTCTTCTTCTCATTTCG-3′, which flank the 32-bp deletion were used to generate wild-type and deletion fragments of 189 and 157 bp, respectively. The PCR mixture contained 0.25 mM deoxynucleoside triphosphates, 20 pmol of each primer, and 0.5 U of Taq polymerase in 1× reaction buffer (Boehringer Mannheim). Each PCR amplification consisted of 40 cycles, with the first 5 cycles consisting of 94°C for 1 min, 55°C for 1 min, and 72°C for 1.5 min, followed by 35 cycles of 94°C for 30 s, 60°C for 30 s, and 72°C for 45 s.

HIV RNA was determined by Roche Amplicor assay.

RESULTS

None of the HIV-infected children had the homozygous gene for Δ32. The mutant CCR5 allele (wt/Δ32) was present in 2 (4.2%) of the 47 children—in none of the 7 rapid disease progressors; in 2 of the 17 non-rapid disease progressors (11.76%) and of the 36 children in the combined group of non-rapid progressors and those with intermediate disease course (5.5%) (Table 1). Differences in the prevalence of the mutant CCR5 gene between any of the groups were not significant by Fisher’s exact test, although there was a trend towards significance in the prevalence of wt/Δ32 in non-rapid progressors compared to the total study group (P = 0.076). There were no deaths on follow-up to average ages of 3.7 (1.7 to 5.7) years for the rapid progressors, 12.8 (8.4 to 19.4) years for non-rapid progressors, and 5.8 (3 to 7.9) years for the intermediate group. None of the uninfected HIV-exposed or uninfected unexposed children had the mutant allele. The frequency of the mutant CCR5 allele in this multiethnic population (African-Americans, Caucasians, Latinos, and Asians) is ∼4.8%.

TABLE 1.

Subject characteristics

Group (n)	No. of subjects
	Risk		CCR5 genotypinga
	Perinatal (%)	Transfusion	wt/wt	Δ32/wt
HIV infected (43)
Rapid progressors (7)	7 (17)		7
Non-rapid progressors (17b)	15 (37)	2	15	2c
Intermediate disease course (19)	19 (46)		19
Controls (4)
Exposed to HIV (2)			2
Not exposed to HIV (2)			2

^aNo patient was homozygous for Δ32. 

^bIncluding seven long-term survivors as defined in the text. 

^cOne subject had acquired HIV perinatally, and the other was a hemophiliac.  

Clinical details of patients with Δ32 heterozygosity are described below.

Case 1.

The first patient was an 8-year-old boy born to an HIV-positive African-American mother and an unknown father. He had fetal alcohol syndrome and thalassemia minor. The child was first diagnosed as having HIV infection at 3 years of age when he had septic arthritis and meningitis (Haemophilus influenza type B), which led to bilateral severe sensorineural hearing loss, mild microcephaly, and decreased visual motor coordination. At the time of the study, he had severe immunosuppression (CD4⁺, 1% [absolute count, 13]; CD4/CD8 ratio, 0.03) and viral load of 49,370 RNA copies (4.69 log₁₀)/ml.

Case 2.

The second patient was a 19 1/2-year-old Caucasian male with hemophilia A infected by contaminated blood products prior to March 1985. He was diagnosed as having HIV infection in 1988 and had received antiretroviral monotherapy in the past. His clinical course had been marked by drug reactions, episodes of pancreatitis, chronic herpes zoster, and spontaneous pneumothorax requiring blebectomy. At the time of the study, he had severe immunosuppression (no detectable CD4⁺ cells for 2 years) and a viral load of >750,000 RNA copies/ml.

DISCUSSION

The importance of chemokine receptors for HIV entry and disease pathogenesis is becoming increasingly apparent (7). A genetic defect consisting of a deletion of 32 bp in the coding sequence from position 794 to position 825 (Δ32) has been associated with resistance to HIV infection (13). Our study was undertaken to evaluate the frequency of Δ32 in children with HIV infection and to study its influence on disease pathogenesis.

Two patients heterozygous for Δ32 were identified among HIV-positive subjects in this study. In agreement with previous studies, these findings indicate that heterozygosity for Δ32 does not protect individuals from HIV infection. Several large studies have shown that the frequencies of wt/Δ32 in HIV-infected and HIV-negative individuals are similar (7).

The frequency of the heterozygous Δ32 in our study of HIV-exposed and -infected children predominantly of nonwhite background was 4.8%. This frequency is similar to the 1.7% frequency reported by Dean et al. for the African American population (7). Studies in large populations show that the mutant CCR5 gene is variably distributed in different racial groups. In the Western European Caucasian populations, the Δ32 allele is found at the high frequency of ∼0.20 but has not been found in any individuals of West African or Central African origin, in the Japanese, or in other ethnic groups (13).

None of the 43 HIV-infected children studied by us was homozygous for the Δ32 gene. The presence of two Δ32 alleles was initially described for two individuals who had remained uninfected despite repeated exposure to HIV (12). Peripheral blood mononuclear cells from these individuals did not transduce CCR5 signals and were highly resistant to infection in vitro with M-tropic HIV strains. Phenotypically these individuals were reported to be normal and without any immune defect. In 2,741 HIV-infected individuals from cohorts of homosexual men, hemophiliacs, and intravenous drug users from three different studies, there were no individuals with a homozygous Δ32 deletion (7). Recently however, an individual of European descent from among 265 HIV-infected Australian patients was found to be homozygous for Δ32 (2).

Because homozygous Δ32 confers strong protection from infection, it was surmised that the presence of this deletion may provide some survival advantage, and many studies have examined the relationship of the presence of a single mutant allele with the rate of disease progression (11). In our study cohort, the frequency of wt/Δ32 did not differ significantly between groups of children with different rates of disease progression, although there was a trend towards higher frequency in children who survived longer than 8 years after acquisition of infection and therefore were non-rapid progressors compared to those who died or were in disease category C3 by 2 years of age. Thus, in our study the presence of a heterozygous Δ32 allele appeared to provide a modest degree of protection from rapid disease progression. These results are similar to those of published studies in which Δ32 appears to provide some protection from development of severe HIV disease in adults (11). Interestingly, we did not find CCR5Δ in any of the long-term survivors in our population, indicating that the modification in the disease course achieved by a single allele is modest at best. In studies of homosexual cohorts, the frequency of wt/Δ32 in the long-term nonprogressors was found to be twice that in rapid progressors (7). In hemophiliac cohorts, on the other hand, the difference between the frequencies of the Δ32 allele in the two groups was not significant (7). This differential response may be related to differences in routes of transmission, exposure levels, or viral loads in different risk groups.

In summary, a relatively low frequency (4.8%) of the Δ32 mutation was observed in a small cohort of HIV-infected children and adolescents and was similar to the known frequency in the general nonwhite U.S. population. None of the infected children had the homozygous deletion. Children with wt/Δ32 had a relatively favorable disease course, but none of the children in our cohort who were truly long-term nonprogressors had this deletion, indicating that other genetic host or viral factors may be important in disease pathogenesis in this population. Study of polymorphism of CCR5 and other genes in large cohorts of long-term survivors and in HIV-exposed uninfected children are warranted to determine the roles of such genetic mutations in protection from vertical infection and/or disease progression in infants with perinatal HIV exposure.