DC-SIGN (CD209) gene promoter polymorphisms in a Brazilian population and their association with human T-cell lymphotropic virus type 1 infection

This study evaluated four polymorphisms located in the DC-SIGN (CD209) gene promoter region (positions −336, −332 −201 and −139) in DNA samples from four Brazilian ethnic groups (Caucasians, Afro-Brazilian, Asians and Amerindians) to establish the population distribution of these single-nucleotide polymorphisms (SNPs) and correlated DC-SIGN polymorphisms and infection in samples from human T-cell lymphotropic virus type 1 (HTLV-1)-infected individuals. To identify CD209 SNPs, 452 bp of the CD209 promoter region were sequenced and the genotype and allelic frequencies were evaluated. This is the first study to show genetic polymorphism in the CD209 gene in distinct Brazilian ethnic groups with the distribution of allelic and genotypic frequency. The results showed that −336A and −139A SNPs were quite common in Asians and that the −201T allele was not observed in Caucasians, Asians or Amerindians. No significant differences were observed between individuals with HTLV-1 disease and asymptomatic patients. However, the −336A variant was more frequent in HTLV-1-infected patients [HTLV-1-associated myelopathy/tropical spastic paraparesis (HAM/TSP), 80 %; healthy asymptomatic HTLV-1 carriers, 90 %] than in the control group (70 %) [P=0.0197, odds ratio (OR)=2.511, 95 % confidence interval (CI)=1.218–5.179). In addition, the −139A allele was found to be associated with protection against HTLV-1 infection (P=0.0037, OR=0.3758, 95 % CI=0.1954–0.7229) when the HTLV-1-infected patients as a whole were compared with the healthy-control group. These observations suggest that the −139A allele may be associated with HTLV-1 infection, although no significant association was observed among asymptomatic and HAM/TSP patients. In conclusion, the variation observed in SNPs −336 and −139 indicates that this lectin may be of crucial importance in the susceptibility/transmission of HTLV-1 infections.


INTRODUCTION
Dendritic cells (DCs) are professional antigen-processing cells that act against invading pathogens (Banchereau & Steinman, 1998), being important as initiators and modulators of the immune response. DCs are efficient stimulators of T and B cells. As invading pathogens appear, DCs are produced in bone marrow and migrate to the tissues, where they capture and process the antigens. DC-SIGN [DC-specific intercellular adhesion molecule 3 (ICAM-3)-grabbing non-integrin], encoded by a member of the CD209 gene family, is a receptor on DCs that binds to ICAM-3, which is expressed on T cells to facilitate the initial interaction between DCs and T cells (Soilleux et al., has been shown to be an important receptor of HTLV-1, Ebola virus, cytomegalovirus, dengue virus and other agents of infectious disease (Alvarez et al., 2002;Halary et al., 2002;Tailleux et al., 2003;Tassaneetrithep et al., 2003). In addition, Sakuntabhai et al. (2005) demonstrated that polymorphisms in the CD209 gene promoter region (position 2336) affect its activity by modifying gene transcription significantly. Therefore, variations in this region may have an influence on the pathogenesis of human infectious diseases.
Human T-cell lymphotropic virus type 1 (HTLV-1) was the first retrovirus isolated from humans (Poiesz et al., 1980). The virus has tropism for T cells and is related to clinical disorders such as adult T-cell leukaemia/lymphoma (Hinuma et al., 1981;Yoshida et al., 1982) and HTLV-1associated myelopathy/tropical spastic paraparesis (HAM/ TSP) (Gessain et al., 1985;Osame et al., 1986). Other inflammatory syndromes are also related to HTLV-1, such as polymyositis (Morgan et al., 1989), uveitis (Mochizuki et al., 1992) and infective dermatitis (LaGrenade et al., 1990). Nevertheless, only 2-3 % of infected individuals develop these disorders. These distinct clinical disorders may depend on the type and magnitude of the host immune response to HTLV-1 antigens and the site/organ where the inflammatory reactions occurred. Several studies have demonstrated that genetic heterogeneity may determine the molecular basis of predisposition to HTLV-1 infection (Plancoulaine et al., 2006;Pociot et al., 1992;Turner et al., 1997). Thus, is important to point out that host genetic factors may contribute to disease susceptibility, although the precise mechanisms involved are not completely understood.
On this basis, the objective of the present study was to investigate the distribution of four single-nucleotide polymorphisms (SNPs) located in the CD209 gene promoter at positions 2336, 2332, 2201 and 2139 in four distinct Brazilian ethnic groups (Caucasians, Afro-Brazilians, Asians and Amerindians). We also compared SNP distribution between HTLV-1-infected individuals and a control group to determine whether the frequencies of CD209 polymorphisms influence the susceptibility of HTLV-1-infected individuals to developing HAM/TSP.

Study design
Brazilian ethnic groups. Blood samples were collected from four ethnically distinct Brazilian populations: 45 from Caucasian individuals (46.7 % males; mean age, 38.7 years), 25 from Afro-Brazilians (60.0 % males; mean age, 34.4 years), 28 from Asians (57.1 % males; mean age, 40.7 years) and 25 from Amerindians (69.6 % males; it was not possible to determine the age of the Amerindians). The Caucasian, Afro-Brazilian and Asian groups were selected at random from blood donors of the Regional Blood Center of Ribeirão Preto and Nú cleo de Haemoterapia of Franca, Brazil. Only individuals who reported the absence of racial admixture in their four grandparents were included in the study. The Amerindian samples were collected from Xavante Amerindians who live in Campinápolis city, Mato Grosso state, Brazil.
HTLV-1-seropositive individuals. The HTLV-1-seropositive patients studied here were detected among blood-donor candidates of the Regional Blood Center of Ribeirão Preto and among patients of the Neurology Department of the University Hospital, São Paulo, Brazil. Diagnosis of HTLV-1/2 infection was established by antibody screening of serum/plasma samples using an enzyme immunoassay (rp21e-enhanced EIA; Cambridge Biotech), followed by confirmation by PCR (tax and LTR regions). The total HTLV-1 group consisted of 66 patients (62.1 % females; mean age, 47.5 years), most of whom were Caucasian (77.3 %). The HTLV-1 group was classified into healthy asymptomatic HTLV-1 carriers (HAC) and patients with HAM/TSP. The HAC group consisted of 33 individuals (54.5 % females; mean age, 40.5 years) and the HAM/TSP group consisted of 33 patients (69.7 % females; mean age, 53.2 years). Both groups were mainly of Caucasian origin (75.8 and 78.8 %, respectively). The admixed control group consisted of 32 randomly selected blood donors from the same geographical region as the HTLV-1-positive individuals described above, with a mean age of 34.8 years (84.4 % males); the ethnic background was evaluated based on skin colour and individuals were mainly Caucasians (62.5 %) and African descendants (34.4 %). The study was approved by the Institutional Ethics Committee of the University Hospital, Faculty of Medicine of Ribeirão Preto, University of São Paulo (process number 7639/2005) and all subjects gave written, informed consent to participate. DNA extraction. A 5 ml blood sample was collected into EDTAcontaining tubes and plasma was separated under aseptic conditions by centrifugation at 900g for 10 min at 4 uC. The white-cell layer was aspirated together with the red blood cells and placed in sterile 15 ml polypropylene tubes. DNA was extracted from the buffy coat by using a Super Quick Gene DNA Isolation kit (Analytical Genetic Testing Center, CO, USA) and diluted in 200 ml ultrapure water. All samples were quantified and diluted to 0.1 mg DNA ml 21 .
Amplification reaction. For genotyping, the CD209 gene promoter was amplified in all samples under study by using the primers 59-CAAAAATGAGGACAGCAGCA-39 and 59-CTCCAAGGAACCAA-GACTGC-39, which delimit the promoter region between nt 2423 and 28 (defined from GenBank accession no. NC_000019) and amplify a fragment of 452 bp. Reactions were performed in 25 ml mixtures containing 100 ng DNA, 1.0 U Taq DNA polymerase (Invitrogen Life Technologies), 50 mM KCl, 20 mM Tris/HCl (pH 8.3), 1.5 mM MgCl 2 , 0.2 mM each dNTP and 0.3 pmol each primer. Thermocycling was performed in a GeneAmp PCR system 9700 (Applied Biosystems) under the following conditions: an initial cycle at 95 uC for 5 min and 35 cycles at 95 uC for 40 s, 62 uC for 40 s and 72 uC for 60 s, with a final extension step of 72 uC for 10 min. The amplified products were analysed by 1 % agarose gel electrophoresis followed by ethidium bromide staining. DNA sequencing. Sequencing reactions were performed using PCR primers and a DYEnamic ET dye terminator kit (GE Healthcare UK) according to the manufacturer's instructions. Electrophoresis was carried out using an automated apparatus MegaBace DNA sequencing system 1000 (Amersham Biosciences). The electropherograms were analysed by using Sequence Analyser software version 1 (Amersham Biosciences). Consensus sequences were obtained and compared by multiple alignments, using GenBank accession no. NC_000019 as the prototype sequence.
Sequences analysis. Electropherograms were processed by using the base-calling program Phred , resulting in the generation of files including information about base content and base quality of each sequence. The sequences were aligned pairwise to the specific genomic region by using the Cross Math program (http://www.phrap.org). Multiple alignments were produced by using the POLYBAYES multiple-alignment algorithms (Marth et al., 1999). The SNPs were detected in the highly similar regions of multiple anchored alignments by using the PolyPhred program (Nickerson et al., 1997). An automated pipeline was prepared to process the above steps, integrated with a number of in-house-developed Perl scripts.
Statistical analysis. Ethnic differences between patients and controls were evaluated by an exact test that employed the Metropolis algorithm to obtain an unbiased estimate of the exact P value and its standard error by using R6C software (http:// www.marksgeneticsoftware.net/rxc.htm). This same method was used to compare the frequencies of SNPs among the ethnic groups. Allelic frequencies and observed heterozygosity were computed by a direct counting method. Adherences of genotypic proportions to expectations under Hardy-Weinberg equilibrium were tested by the exact test of Guo & Thompson (1992), employing Arlequin software version 3.11 (Excoffier et al., 2005). GENEPOP version 3.4 software (Raymond & Rousset, 1995a) was used to perform a pairwise exact test of population differentiation (Raymond & Rousset, 1995b) based on allele and genotype frequencies. The Arlequin 3.11 software was also used to perform an exact test of population differentiation based on haplotype frequencies.
The presence of a significant association between loci was determined by means of a likelihood-ratio test of linkage disequilibrium (Excoffier & Slatkin, 1998) using the Arlequin program, version 3.1 (Excoffier et al., 2005). Given the positive association but unknown gametic phase, the PHASE (Stephens et al., 2001) and EM (Excoffier & Slatkin, 1995) algorithms were also used to reconstruct CD209 haplotypes.
The allele, genotype and haplotype frequencies were compared by using Fisher's exact test implemented in GraphPad InStat 3 software, which was also used to estimate the odds ratio (OR) and its 95 % confidence interval (CI).

CD209 promoter polymorphisms in the Brazilian population
We investigated four polymorphisms located in the CD209 promoter region at positions 2336, 2332, 2201 and 2139 that are known to be associated with susceptibility/ resistance to pathogens. The allelic frequencies of these SNPs for Brazilian ethnic groups and other populations previously described are presented in Table 1. All loci evaluated in the CD209 promoter region fitted the Hardy-Weinberg equilibrium expectations for all groups.
The 2336G allele showed higher frequencies in Afro-Brazilians (42.0 %) and Caucasians (26.7 %) than in Asians (5.4 %). In addition, this allele was not observed in the Amerindian group, and the admixed control group presented frequencies similar to those of the Caucasian group (Table 1). SNP 2332A was present at low frequencies in almost all ethnic groups from Brazil, except the Amerindians. Due to its low frequency among the groups, the control group did not present SNP 2332A DC-SIGN (CD209) polymorphisms in HTLV-1 infection (Table 1). The 2332A allele was identified at higher frequencies in the Asian group (10.7 %) compared with Afro-Brazilian (4.0 %) and Caucasian (4.4 %) individuals.
We also analysed the 2201T/G position, which was absent in all ethnic groups from Brazil except for the Afro-Brazilian population (6.0 %) and was present in the control group at a frequency of 1.6 %. Analysis of the 2139A/G position showed that the allelic frequency of SNP 2139A was higher in Asians (62.5 %) than in Afro-Brazilians (26.0 %), Caucasians (35.6 %) and Amerindians (30.0 %). The Brazilian control group showed frequencies similar to those of the Caucasian group (Table 1). The presence of a significant association between SNPs was estimated by a likelihood-ratio test of linkage disequilibrium considering all of the samples from the control admixed and ethnic groups. A positive association was detected between SNP 2336 and SNP 2201 (P50.0138±0.0005) and -139 (P50.0000±0.0000), and between SNP 2332 and SNP 2139 (P50.0028±0.0002). Given the positive association but unknown gametic phase, the PHASE and EM algorithms were used to reconstruct CD209 haplotypes ( Table 2). The same haplotype was estimated in all samples by the PHASE and EM algorithms, with mean probabilities of 0.9753 and 0.9837, respectively.

HTLV-1 individuals
All HTLV-1 individuals were evaluated for clinical status. Of these, 50 % were HAC and 50 % had HAM/TSP. None of the patients showed positive serology for other retroviruses, e.g. human immunodeficiency virus type 1 (HIV-1) or HTLV-2. The sex, age and ethnic background distribution of the blood donors are shown in Table 3. The HAM/TSP group showed a homogeneous distribution between females and males. A high frequency of infection among Caucasian individuals was detected in both the HAC and HAM/TSP groups (Table 3). Table 4 shows the genotype and allele frequencies for SNPs 2336, 2332, 2139 and 2201 among the HTLV-1-infected individuals (HAC and HAM/TSP) and the admixedcontrol group. It is important to note that the ethnic background of the HTLV-1 groups, including HAC and HAM/TSP, did not differ from that of the controls (P50.1125±0.0079, P50.2812±0.0089 and P50.0916± 0.0062, respectively).
An exact test of population differentiation based on the allele and genotype frequencies revealed that the total HTLV-1 group showed differences in the allele (P50.0035) and genotype (P50.0055) frequencies compared with the controls. However, no differences were found between the HAC and HAM/TSP groups. In order to detect a possible influence of specific alleles or genotypes, 262 contingency tables were created to compare the frequencies using Fisher's exact test.
The 2336A allele was found to be associated with susceptibility to HTLV-1 infection and the 2336G allele was found to be associated with protection against infection when the HTLV-1-infected patients as a whole were compared with the healthy-control group (P50.0197, OR52.511, 95 % CI51.218-5.179). Likewise, the 2336GG genotype was found to be associated with protection against HTLV-1 infection when the HTLV-1 group as a whole was compared with the healthy-control group (P50.0030, OR50.03759, 95 % CI50.002008-0.7038). On this basis, it is plausible to assume that the 2336G allele is the variant conferring resistance to HTLV-1 infection. When the asymptomatic HTLV-1-infected patients were compared with the HAM/TSP group, no difference was found regarding this polymorphism (P50.1353) (Table 4). However, the frequency of the 2336G allele was significantly lower in the asymptomatic HTLV-1-infected group than in healthy controls (P50.0036, OR50.2368, 95 % CI50.0874-0.6414) (Table  4). Likewise, the frequencies of both homozygous genotypes for the 2336 locus (AA and GG) in the asympto-  (Table 4). However, these results may also be a hitchhiking effect of the protective influence of the 2139A allele against HTLV-1 infection.
An exact linkage-disequilibrium test taking into account all samples revealed a high linkage disequilibrium between SNPs 2336 and 2139 (P50.0000±0.0000), which may indicate a possible additive or co-dependent effect for both polymorphisms in the resistance/susceptibility to HTLV-1. In order to determine such interactions, haplotypes for each sample were estimated by two distinct methods without taking into account any prior information. Table 5 shows the frequency of each haplotype found in HTLV-1infected patients and in the control group. An exact test of differentiation based on haplotype frequencies revealed a  difference between the HTLV-1 group as a whole and the healthy-control group (P50.0000±0.0000). As observed for the allele and genotype frequencies, there was no difference between the HAM/TSP and the asymptomatic group (P50.3328±0.0206), but both HTLV-1 groups did differ from the healthy-control group (P50.0005±0.0004 and P50.0000±0.0000, respectively).

DISCUSSION
We have described here for the first time the distribution of genetic polymorphisms in the CD209 gene promoter (2336A/G, 2332A/G, 2201T/G and 2139A/G positions) in distinct Brazilian ethnic groups and also their association with HTLV-1 infection.
In the Brazilian population studied here, the 2336G allele was more frequent in the Afro-Brazilian group than in the Caucasian and Asian groups. However, the GG genotype was not detected in the Asian group and most subjects (86 %) were AA homozygotes, with only 13 % being heterozygotes. These data agree with previous studies that reported a higher frequency of the 2336G allele in populations of African descent (Barreiro et al., 2006a;Boily-Larouche et al., 2007;Olesen et al., 2007).
An interesting result was the identification of the 2332A allele at high frequency in the Asian group (10.7 %). This variant has recently been described by Boily-Larouche et al. (2007) in a Zimbabwean population (2 %) as a probable position that can influence gene transcription.
The 2201T allele was exclusively detected in the Afro-Brazilian population and has been reported only in African ethnic groups (Boily-Larouche et al., 2007). This SNP was also observed here in the control group, demonstrating that, despite the miscegenation of the Brazilian population, the methodology used was efficient in separating the ethnic groups.
A high prevalence of the 2139A allele was observed in the Asian population compared with the other ethnic groups.
A similar result has been reported previously by Barreiro et al. (2006a) in a study assessing the frequency of this polymorphism in individuals of African, Asian and European descent. However, we have reported here for the first time the distribution of the 2129G allele among Amerindians, which was surprisingly high (70.3 %).
As recently demonstrated, DCs play an important role in HTLV-1 infection, promoting contagion of T cells in both the cis and trans forms (Jones et al., 2008). According to previous findings, HTLV-1-infected cells can use surfaceadhesion molecules to regulate fusion with target cells, with the involvement of DC-SIGN and ICAM ligands. This mechanism could have consequences for the regulation of both DC infection and HTLV-1 dissemination, but also for immune regulation (Ceccaldi et al., 2006). On this basis, in the present study we assessed SNP distribution in the promoter region of the CD209 gene in an attempt to associate these polymorphisms with HTLV-1 infection. Infected individuals were divided into asymptomatic carriers and HAM/TSP carriers according to clinical condition, and the distribution of the polymorphisms was analysed in each group.
Regarding SNP 2336, allele G was found to be less frequent among HAM/TSP (20 %) than HAC (9 %) individuals compared with controls (30 %). The absence of the GG genotype in the HTLV-1-infected group suggests that this 2336G allele is a variant that confers protection against HTLV-1 infection. Thus, the presence of the 2336G allele would result in lower DC susceptibility to HTLV-1 in the initial stages of infection, possibly protecting against HTLV-1 infectivity.
In agreement with the present results, the 2336G allele has been associated with reduced DC-SIGN expression on the DC surface and in macrophages, as position 2336 is located close to the major transcription site, possibly affecting binding of the transcription factor Sp1 and other factors that modulate transcriptional activity (Sakuntabhai et al., 2005). On this basis, the antigen-presenting capacity of individuals who carry the 2336G allele would be impaired, with a consequent alteration of the immune response. Regarding infectious diseases, position 2336 in the promoter region of the CD209 gene has been studied extensively. Vannberg et al. (2008) demonstrated that allele 2336G is associated with protection against tuberculosis in the population of sub-Saharan Africa. Conversely, the 2336G variant has been associated with susceptibility to HIV infection, haemorrhagic dengue and tuberculosis, suggesting that high levels of DC-SIGN expression permit better capture and processing of the antigens (Barreiro et al., 2006a;Martin et al., 2004;Sakuntabhai et al., 2005).
Similarly, in the present study, polymorphisms at positions 2332 and 2201 of the promoter region of CD209 were investigated and no significant association was found when the genotypes and alleles of infected individuals were compared with those of controls or when the HAC and HAM/TSP groups were compared.
For the SNP at position 2139, allele A was found to be associated with protection against HTLV-1 infection when the infected and control groups were compared. Thus, these results suggest that this SNP may be associated with HTLV-1 transmission and not with development of the associated clinical manifestations. Previous studies have already demonstrated a higher frequency of allele 2139A in individuals not infected with HIV compared with infected patients (Martin et al., 2004). In another study, allele 2139G was found to be associated with rapid progression of AIDS in a population of Japanese haemophiliacs. This SNP is located close to one of the binding sites of the transcription factor AP-1 in the promoter region of DC-SIGN and the substitution of one nucleotide close to this site can increase the level of expression of DC-SIGN, resulting in an accelerated progression of AIDS (Koizumi et al., 2007).
Comparison of the HTLV-1 group and the healthy-control group showed that haplotypes 2336G/2332G/2201G/ 2139G and 2336A/2332G/2201G/2139A were associated with protection against HTLV-1 infection, whilst haplotype 2336A/2332G/2201G/2139G was associated with susceptibility to HTLV-1 infection. These associations suggest that these polymorphisms may influence only the acquisition of HTLV-1 and not the associated diseases. Comparison of the HAM/TSP and asymptomatic groups with healthy controls revealed the presence of the same haplotypes conferring protection and susceptibility against HTLV-1, suggesting that these polymorphisms also have a hitchhiking effect on HTLV-1 acquisition.
In conclusion, the variations detected in the promoter region of the CD209 gene may not be involved directly in triggering the development of HAM/TSP. However, the differences in SNP distribution observed between HTLV-1infected individuals and controls suggest that these SNPs in the promoter region of CD209 may be associated with the risk of HTLV-1 infection. Thus, the variations observed at positions 2336 and 2139 in the lectin DC-SIGN may be of crucial importance in the susceptibility/transmission of HTLV-1 infection.