Frequency of the CCR 5 ∆ 32 allele in Brazilians : a study in colorectal cancer and in HTLV-I infection

The cc-chemokine receptor-5 gene protein (CCR5) is a seven-transmembrane-domain G protein coupled receptor that has as its natural ligands RANTES (regulated on activation normal T-cell expressed and secreted), MIP-alpha (macrophage inflammatory protein) and MIP-beta, which are members of the CC subfamily of chemokines (Samson et al., 1996a). The CCR5 protein, in concert with CD4, is also the receptor that mediates the internalization of HIV in macrophages and monocytes (Dragic et al., 1996). Studies of individuals who have been multiply exposed to HIV and yet remain uninfected led to the identification of the CCR5∆32 allele, that exhibits a 32-bp deletion in the CCR5 gene (Liu et al., 1996). Because of a frameshift at amino acid 185 the deleted allele codes for a truncated protein that is consequently not integrated into the cell membrane (Liu et al., 1996). The homozygous state for the CCR5∆32 allele has been shown to be associated with a high degree of protection against HIV infection in vivo (Dean et al., 1996; Huang et al., 1996; Samson et al., 1996b). This protection is not absolute as demonstrated by identification of one HIV-1-infected CCR5∆32 homozygote (Biti et al., 1997; O’Brien et al., 1997). Although there is no evidence for a strong protective effect in heterozygotes (Dean et al., 1996; Huang et al., 1996; Michael et al., 1997), a modest reduction in HIV incidence has been reported (Samson et al., 1996b). Also, the observation that cohorts of long-term survivors are enriched for CCR5∆32 heterozygotes suggests the existence of some protection against the AIDS progression (Dean et al., 1996; Huang et al., 1996; Michael et al., 1997). The CCR5∆32 allele is common in North America and Europe, with heterozygote frequencies varying from 10 to 20% in Caucasians (Dean et al., 1996; Huang et al., 1996; Liu et al., 1996; Samson et al., 1996b; Martinson et al., 1997). In Eurasia CCR5∆32 frequencies show a north-south gradient with the highest allele frequencies in Finnish and Mordvinian populations (16%) and the lowest in Sardinia (4%) (Libert et al., 1998). The CCR5∆32 allele is not found in aboriginal populations outside Eurasia (Martinson et al., 1997; Leboute et al., 1999; Lu et al., 1999). The presence of CCR5∆32 allele elsewhere throughout the world probably represents recent European gene flow into local populations (Martinson et al., 1997). The origin of the CCR5∆32 has been speculatively dated to approximately 700 years Frequency of the CCR5∆32 allele in Brazilians: a study in colorectal cancer and in HTLV-I infection


INTRODUCTION
The cc-chemokine receptor-5 gene protein (CCR5) is a seven-transmembrane-domain G protein coupled receptor that has as its natural ligands RANTES (regulated on activation normal T-cell expressed and secreted), MIP-alpha (macrophage inflammatory protein) and MIP-beta, which are members of the CC subfamily of chemokines (Samson et al., 1996a).The CCR5 protein, in concert with CD4, is also the receptor that mediates the internalization of HIV in macrophages and monocytes (Dragic et al., 1996).Studies of individuals who have been multiply exposed to HIV and yet remain uninfected led to the identification of the CCR5∆32 allele, that exhibits a 32-bp deletion in the CCR5 gene (Liu et al., 1996).Because of a frameshift at amino acid 185 the deleted allele codes for a truncated protein that is consequently not integrated into the cell membrane (Liu et al., 1996).The homozygous state for the CCR5∆32 allele has been shown to be associated with a high degree of protection against HIV infection in vivo (Dean et al., 1996;Huang et al., 1996;Samson et al., 1996b).This protection is not absolute as demonstrated by identification of one HIV-1-infected CCR5∆32 homozygote (Biti et al., 1997;O'Brien et al., 1997).Although there is no evidence for a strong protective effect in heterozygotes (Dean et al., 1996;Huang et al., 1996;Michael et al., 1997), a modest reduction in HIV incidence has been reported (Samson et al., 1996b).Also, the observation that cohorts of long-term survivors are enriched for CCR5∆32 heterozygotes suggests the existence of some protection against the AIDS progression (Dean et al., 1996;Huang et al., 1996;Michael et al., 1997).

Abstract
The identification of a 32-bp deletion in the cc-chemokine receptor-5 gene (CCR5∆32 allele) that renders homozygous individuals highly resistant to HIV infection has prompted worldwide investigations of the frequency of the CCR5∆32 allele in regional populations.It is important to ascertain if CCR5∆32 is a factor to be considered in the overall epidemiology of HIV in individual populations.With this in mind we determined the CCR5∆32 allele frequency in a large sample (907 individuals) of the southeastern Brazilian urban population, stratified as follows: 322 healthy unrelated individuals, 354 unselected colorectal cancer patients, and 229 blood donors.The three groups displayed essentially identical allelic frequencies of CCR5∆32 and pairwise comparisons did not show significant differences.Thus, our results can be pooled to provide a reliable estimate of the CCR5∆32 allele frequency in the southeastern Brazil of 0.053 ± 0.005.The blood donors comprised 50 HTLV-I serologically negative individuals, 115 non-symptomatic individuals HTLV-I positive by ELISA but with indeterminate Western blot results, 49 healthy blood donors HTLV-I positive both at ELISA and Western blot and 15 patients with clinical spinal cord disease (HAM).A suggestive trend was observed, with the CCR5∆32 frequencies decreasing progressively in these four categories.However, when we applied Fischer's exact test no significant differences emerged.We believe that further studies in larger cohorts should be performed to ascertain whether the CCR5∆32 allele influences the chance of becoming infected or developing clinical symptoms of HTLV-I infection.
ago, with its frequency increasing rapidly as consequence of a strong selective pressure, possibly an ancient plague (Stephens et al., 1998).A possible candidate is the bubonic plague that claimed the lives of 25-33% of Europeans in the 14th century (Stephens et al., 1998).
HTLV-I is a retrovirus from the oncoretroviridae family with variable prevalence rate in human populations.The vast majority of HTLV-I-infected individuals have no associated symptoms.However, 1 in every 1000-2000 individuals with positive serology per year develop adult T-cell leukemia/lymphoma (ATLL) (Poiesz et al., 1980;Tajima, 1990) and other patients may present with myelopathy (HAM) (Gessain et al., 1985).The prevalence of the latter has been estimated to be 2.4% in the United States (Murphy et al., 1997) but only about 0.07% among the Japanese (Osame et al., 1990).Several genetic aspects, viral or human, have been tested as possible determinants of clinical course of the HTLV-I infection.Among the viral markers, tax mutations (Renjifo et al., 1995) and the proportion of synonymous and non-synonymous mutations on the tax have been tested (Niewiesk and Bangham, 1996).Regarding host genetic markers, the HLA polymorphisms are the main class tested until now, mostly with weak associations (LaGrenade et al., 1996;Sonoda et al., 1996;Jeffery et al., 1999).Recently we have tested polymorphisms of endogenous retroviral sequences (Pereira, R.W., Proietti, A.B. and Pena, S.D.J., unpublished results) but did not see any association with one of the clinical outcome of HTLV-I infection.
Inter-population differences in CCR5∆32 frequency may influence the pattern of HIV transmission and will need to be incorporated into future predictions of HIV levels (Martinson et al., 1997).We thus considered it important to establish the frequency of CCR5∆32 allele in the Brazilian population to support future studies in groups at risk of HIV infection.Moreover, we evaluated two special groups: patients with colorectal cancer and patients with HTLV-I infection.

MATERIAL AND METHODS
We analyzed 322 unrelated Brazilian individuals randomly chosen among unrelated participants in paternity testing studies.Most of the patients were self-classified as "white" and originated from the southeast region of Brazil.We also studied 354 unselected colorectal cancer patients from the Hospital do Câncer A.C. Camargo in São Paulo.Finally, 229 individuals participating in a cohort study at Fundação Hemominas were distributed as follows: 50 HTLV-I seronegative individuals, 115 non-symptomatic individuals HTLV-I positive at the ELISA test but with indeterminate Western blot results, 49 healthy blood donors HTLV-I positive both at ELISA and Western blot and 15 HTLV-I patients with clinical spinal cord disease (HAM).Consent was obtained from all participants and all analyses were performed anonymously.
The deletion in CCR5 was assayed by means of the polymerase chain reaction (PCR) with previously described primers (Samson et al., 1996b).PCR was carried out with 20 pmol of each primer, 0.1 mM dNTPs, 1.5 mM MgCl 2 , 75 mM KCl, 10 mM Tris-HCl, pH 8.3, and 1 unit of Taq polymerase in a final volume of 20 µl.Cycle conditions were: initial denaturation of 94°C for 3 min and 30 cycles of 94°C for 1 min, 62°C for 2 min, 72°C for 1 min, followed by a final extension step of 72°C for 5 min.A sample of the reaction product (5 µl) was run on 5% polyacrylamide gels and visualized by silver staining (Santos et al., 1993).Typing was unambiguous, the wild type allele and the CCR5∆32 allele generating fragments of 735 bp and 703 bp, respectively.
The allele frequencies were calculated by gene counting.The Hardy-Weinberg equilibrium of the groups and the significance level of the genotypic and allelic frequency differences between groups were calculated using Fisher's exact test.

RESULTS
Our three groups (healthy individuals, colorectal cancer patients and blood donors) displayed essentially identical allelic frequencies of CCR5∆32 (Table I) and pairwise comparisons did not show significant differences.In all the groups the genotype frequencies did not deviate significantly from the Hardy-Weinberg equilibrium.Thus, our results can be pooled to provide a reliable estimate of the CCR5∆32 allele frequency in southeastern Brazil of 0.053 ± 0.005, which agrees with previous estimates (Martinson et al., 1997).
Within the blood donor patients, we compared the allelic and genotypic frequency of CCR5∆32 in four groups: HTLV-I negative healthy blood donors, HTLV-I ELISA positive and Western blot negative healthy individuals, ELISA and Western blot positive non-symptomatic individuals, and ELISA and Western blot positive symptomatic patients with HAM (Table II).A suggestive trend was observed, with the CCR5∆32 frequencies decreasing progressively in these four categories.However, as expected from the small number of symptomatic patients and the broad confidence limits of the CCR5∆32 allele frequencies, no significant differences were obtained when we applied the Fischer's exact test.

DISCUSSION
The Brazilian population is predominantly of European ancestry with significant contributions from African and Amerindian gene pools (Pena et al., 2000).Since the CCR5∆32 allele has not been identified in Africans (Martinson et al., 1997) or Amerindians (Martinson et al., 1997;Leboute et al., 1999), the presence of the CCR5∆32 allele in Brazilians can be attributed to European immigration.The 0.053 frequency of CCR5∆32 allele that we ob-served in our sample is lower than the average frequency in Europe (0.10), reflecting the African and Amerindian contributions.A previous estimate of the CCR5∆32 allele frequency among Brazilian individuals was 0.035, but only 100 individuals were studied (Passos and Picanco, 1998).At any rate, the CCR5∆32 allele frequency will fluctuate according to the proportion of European versus African or Amerindian ancestry in the sample.
The CCR5∆32 allele frequency in patients with colorectal cancer was exactly the same as in healthy individuals.When we compared the frequencies in relation to HTLV-I infection a suggestive trend was observed with the frequency decreasing progressively among the four categories of blood donors (Table II).However, when we applied Fischer's exact test no significant differences emerged.We believe that further studies in larger cohorts should be performed to ascertain whether the CCR5∆32 allele influences the chance of becoming infected or developing clinical symptoms of HTLV-I infection.

ACKNOWLEDGMENTS
Research supported by grants from FAPEMIG and CNPq.R.W.P. is the recipient of a PhD studentship from CAPES.Publication supported by FAPESP.

Table II -
Determination of the CCR5∆32 allele frequency in blood donors with different status in reference to HTLV-I infection.

Table I -
Determination of the CCR5∆32 allele frequency in the urban population of southeastern Brazil.