RESEARCH ARTICLE da Silva-Malta et al., Journal of Medical Microbiology 2017;66: DOI /jmm

Transcription

1 RESEARCH ARTICLE da Silva-Malta et al., Journal of Medical Microbiology 217;66: DOI 1.199/jmm..539 The Duffy null genotype is associated with a lower level of CCL2, leukocytes and neutrophil count but not with the clinical outcome of HTLV-1 infection Maria Clara Fernandes da Silva-Malta, 1,2 Camila Campos Sales, 3 Jacqueline Cronemberger Guimar~aes, 4 Poliane de Cassia Gonçalves, 1 Daniel Gonçalves Chaves, 1,2 Hadassa Campos Santos, 5 Alexandre da Costa Pereira, 5 Jo~ao Gabriel Ribas, 2 Anna Barbara de Freitas Carneiro-Proietti 1,2 and Marina Lobato Martins 1,2, * Abstract Purpose. Chemokines are important in the immune response against viral infections, and may play a role in human T- lymphotropic virus 1 (HTLV-1)-associated myelopathy/tropical spastic paraparesis (/TSP) pathogenesis. Polymorphisms in the Duffy antigen receptor for chemokines (DARC), such as rs1275 (A>G; FY*B>FY*A) and rs ( 46T>C; GATA-1 box) may influence circulating concentrations of proinflammatory chemokines. We investigate whether Duffy genotypes influence the HTLV-1 proviral load (PVL) level, HTLV-1 infection outcome and chemokine concentrations in HTLV-1 asymptomatic carriers (=162), /TSP patients (=135) and seronegative individuals (SN=71). Methodology. Quantification of plasmatic IL8, CCL2 and CCL5 were performed by flow cytometry and Duffy genotypes were investigated by real-time PCR. HTLV-1 PVL was quantified in peripheral blood. To control for spurious association, individual ancestry profiles in and groups were investigated. Results/Key findings. PVL and IL8 level were significantly higher in the group than in the group, but were not associated with Duffy genotypes. The highest CCL2 and CCL5 levels were seen in the SN group, and there was no difference when comparing the infected groups. The level of CCL5 was not associated with Duffy genotypes. The polymorphism 46 C/C that abrogates the DARC expression on the erythrocytes was significantly associated with lower levels of CCL2, neutrophil and white blood cell (WBC) counts in HTLV-1-infected individuals. Conclusion. We conclude that although the Duffy null genotype was associated with leukopenia, neutropenia and lower levels of CCL2, the data do not suggest the influence of Duffy genotypes on the neurologic outcome of HTLV-1 infection, but may be a confounding factor in comparison HTLV-1-infected populations with different ancestries, especially when defining inflammatory biomarkers. INTRODUCTION Human T-lymphotropic virus 1 (HTLV-1) infects 15 2 million individuals worldwide and may cause, in 5 1 % of infected persons, a malignant disease, known as adult T- cell leukemia (ATL) and inflammatory conditions, including HTLV-1-associated myelopathy/tropical spastic paraparesis (/TSP) [1]. The pathogenesis of /TSP involves the migration of infected cells and HTLV-1-specific CD8+ cells from blood to the central nervous system (CNS), leading to a chronic inflammatory response that impairs nervous cells [2, 3]. Several studies have shown that high levels of HTLV-1 proviral load (PVL) in blood and in the CNS is a risk marker for /TSP development [4 7]. Chemokines are small molecules that play a role in the migration of immune cells from blood to specific tissues. Received 17 February 217; Accepted 19 June 217 Author affiliations: 1 Serviço de Pesquisa, Fundaç~ao Hemominas, Belo Horizonte, Minas Gerais, Brazil; 2 Interdisciplinary HTLV Research Group (GIPH), Brazil; 3 Faculdade de Farmacia, Universidade Federal de Minas Gerais, Belo Horizonte, Minas, Gerais, Brazil; 4 Instituto de Ci^encias Biológicas, Universidade Federal de Minas, Gerais, Belo Horizonte, Brazil; 5 Laboratório de Genetica e Cardiologia Molecular, Instituto do Coraç~ao, Faculdade de Medicina da Universidade de S~ao Paulo, S~ao Paulo, Brazil. *Correspondence: Marina Lobato Martins, marina.martins@hemominas.mg.gov.br Keywords: HTLV-1; /TSP; proviral load; chemokines; Duffy; DARC. Abbreviations:, asymptomatic carriers; AIM, ancestry informative markers; ATL, adult T-cell leukemia; CNS, central nervous system; DARC, Duffy antigen receptor for chemokines;, /TSP patients; /TSP, HTLV-1-associated myelopathy/tropical spastic paraparesis; HLA, human leukocyte antigens; HTLV-1, human T-lymphotropic virus 1; HWE, Hardy Weinberg equilibrium; PVL, proviral load; RBC, red blood cells; SN, seronegative individuals; SNP, single nucleotide polymorphism; WBC, white blood cells; WHO, World Health Organization. 539 ã 217 The Authors 127

2 da Silva-Malta et al., Journal of Medical Microbiology 217;66: They regulate the cell traffic of various types of leukocytes through interactions with a subset of 7-transmembrane, G protein-coupled receptors on the surface of circulating and endothelial cells [8]. Since /TSP is characterized by an inflammatory condition in the CNS, the level of chemokines in blood and the cerebrospinal fluid may be a risk factor to /TSP outcome [9, 1]. The Duffy antigen receptor for chemokines (DARC) is a high affinity receptor for pro-inflammatory chemokines, such as CCL2 (MCP-1, monocyte chemoattractant protein-1), CCL5 (RANTES, regulated on activation, normal T-cell expressed and secreted) and IL8 (Interleukin 8, CXCL8), but do not bind CCL3 (MIP-1a, macrophage inflammatory protein) and CCL4 (MIP-1b) [11]. However, the chemokine ligation on DARC does not transmit a signal via G proteins. DARC is a glycoprotein present on red blood cells (RBCs) and is expressed by erythroid and non-erythroid cells, such as neurons, endothelial cells from brain, kidney and spleen, and lung epithelial cells [12]. The two major codominant alleles in the Duffy blood group are FY*A and FY*B, which differ by a single nucleotide polymorphism (SNP) 125G>A with an amino acid substitution Gly42Asp and codify for Fy a and Fy b antigens, respectively, originating four phenotypes: Fy (a+b+), Fy(a+b ), Fy(a b+) and Fy(a b ) [13]. Since DARC is also a receptor for Plasmodium vivax, individuals with the null phenotype present more resistance against malaria, which leads to it being present in >95 % of the West African population and in about 68 % of Afro-Americans, whereas it is rare in Caucasians [12, 14, 15]. The null phenotype is associated with a recessive FY*B ES allele, that shows an SNP in the GATA-1 box on the promotor region of the Duffy gene ( 46T>C), which disrupts the erythrocyte-specific transcription binding site, abrogating the DARC expression only on RBCs [16, 17]. Considering that chemokines bound to erythrocyte DARC are not accessible to leukocytes, some studies have proposed that DARC functions as a sink for circulating chemokines, binding intravascular chemokines that become unable to activate leukocytes [18, 19]. Thus, it is hypothesized that expression of DARC on the surface of RBCs may regulate inflammatory responses leading to circulating chemokine levels being changed in blood and tissues. In addition, the absence of DARC on RBCs is associated with lower leukocytes and neutrophil levels [2, 21]. As a result, functional polymorphisms in DARC may contribute to individual variability in inflammatory responses. We investigated whether polymorphisms rs1275 (A>G; FY*B>FY*A) and rs ( 46T>C; FY*B>FY*B ES ), that change DARC expression on RBCs, influence the plasmatic level of IL8, CCL2 and CCL5 and whether they are associated with the HTLV-1 PVL level and /TSP. To the best of our knowledge, this is the first study to analyse the influence of Duffy genotypes on HTLV-1 infection outcome. METHODS Population study We analysed 162 HTLV-1 asymptomatic carriers (), 135 /TSP patients () and 71 healthy HTLV seronegative individuals (SN) as the control group. and SN subjects were blood donors enrolled in the GIPH (Interdisciplinary HTLV Research Group) cohort study, which began in 1997 at a public blood centre (Fundaç~ao Hemominas, Belo Horizonte, Brazil). were from a rehabilitation hospital (Rede Sarah, Belo Horizonte, Brazil). and groups were classified according to the World Health Organization s (WHO) criteria [22] and HTLV-1 infection was diagnosed by ELISA or chemiluminescence tests and was confirmed by Western blot (WB HTLV 2.4, Genelabs Diagnostics, Singapore) or real-time PCR [23]. All subjects were seronegative for other blood-borne pathogens tested at Fundaç~ao Hemominas (HIV, HCV, HBV, Chagas disease and syphilis) whose screening tests are mandatory in donated blood in Brazil. Total WBC count and neutrophils were obtained from an automatic blood count test (Beckman Coulter T89, CA, USA) close to when the sample used for PVL and chemokine analysis was collected. Duffy genotyping by real-time PCR Duffy genotyping was performed for all participants by realtime PCR, using genomic DNA from peripheral blood. Two distinct tests were standardized for the detection of FY*B and FY*A alleles, and the wild and mutated GATA-1 box (FY*B and FY*B ES alleles, respectively) in a 2 µl reaction using the primers described previously [24] and probes designed using the Primer Express Software (Applied Biosystems, Foster City, CA, USA). The primers and probes used for genotyping the FY*A and FY*B alleles were as follows: FY1 (5 -CTGAGA TCAAGTCAGCTG-3 ); FY2 (5 -AGGATGAAGAAGGG- CAGTGC-3 ); FYA-P (5 -VIC-CAGATGGAGTATGG TG-3 ); FYB-P (5 -FAM-ATGGAGTATGATGCCA-3 ). The primers and probes used for genotyping the GATA region were as follows: GATA-1 (5 -CGTGGGGTAAGGCTTCC TGA-3 ); GATA-2 (5 -CTGTGCAGAGTTCCCCAT-3 ); GATA-W (5 -VIC-TTGGCTCTTATCTTGGA-3 ); GATA- M (5 -FAM-TTGGCTCTTCTTGG-3 ). The tests used 1X PCR Genotyping Master Mix Kit (Applied Biosystems, Foster City, CA, USA), 1 ng of DNA, 6 nm of primers and 2 nm of each of the fluorescent MGB probes specific for each Duffy polymorphism. Both reactions were performed on a real-time PCR system model ABI 75 Fast (Applied Biosystems, Foster City, CA, USA) under the following conditions: a pre-amplification step of 6 C for 3 min and 95 C for 1 min, followed by 5 cycles at 92 C for 15 s and 6 C for 1.5 min, and a post-amplification step of 6 C for 3 s. The analysis was performed using the allelic discrimination plot. Quantification of HTLV-1 PVL DNA isolated from peripheral blood by column extraction kit (QIAamp DNA Blood kit; Qiagen GmbH, Hilden, Germany) was used to quantify HTLV-1 PVL by a real-time 128

3 da Silva-Malta et al., Journal of Medical Microbiology 217;66: SYBR Green PCR method, as previously described [7]. A 186 bp fragment of the pol gene was amplified with primers SK11 (5 -CCCTAATCCACAGCTCAG-3 ) and SK111 (5 -GTGGTGAAGCTGCCATCGGGTTTT-3 ), in parallel with the amplification of the albumin gene (141 bp), using the ALB-S (5 -GCTGTCATCTCTTGTGGGCTGT- 3 ) and ALB-AS (5 -AATCATGGGAGCTGCTGGTT- 3 ) primers to determine the total cell number. The amplifications were performed in separate reactions using 24 ng DNA (equivalent to 39 cells), 1 SYBR Green PCR Master Mix (Applied Biosystems, Foster City, CA, USA) and 2 nm of each primer in an ABI Prism 73 Sequence Detector System (Applied Biosystems), with the following cycle conditions: 2 min at 5 C and 1 min at 95 C followed by 45 cycles of 15 s at 95 C and 1 min at 65 C. Melting curves were performed after the end of the amplification cycles to validate the specificity of the amplified products. Standard curves were generated using 1-fold serial dilutions of DNA from MT2 cells ( ), and normalized to three copies of the HTLV-1 pol gene and two copies of the cellular albumin gene per MT2 cell. All standard dilutions as well as control and individual samples were run in duplicate for both HTLV-1 and albumin DNA quantification. Standard curves were accepted when slopes were between 3.1 (efficiency of 11 %) and 3.74 (efficiency of 85 %) and the r 2 was >.99. The value for the HTLV-1 PVL was reported as [(pol average copy number)/(albumin average copy number/2)]1 4 and expressed as the number of HTLV-1 copies 1 4 cells -1. Quantification of plasmatic chemokines Chemokines were measured from plasma samples that were stored at 7 C. Plasma samples were centrifuged at 1 g for 1 min and the supernatants were used for chemokine quantification. CCL2 (MCP-1 or monocyte chemoattractant protein-1), CCL5 (regulated on activation, normal T-cell expressed and secreted or RANTES) and IL8 (CXCL8) were quantified using the Human Chemokine CBA kit (Becton Dickinson, San Jose, CA, USA), according to the manufacturer s instructions. The fluorescence from the labelled beads was captured by a flow cytometer (Accuri C6, Becton Dickinson, BD, Franklin Lakes, NJ, USA), using 18 events/sample, 3 beads/chemokine. The data were analysed using the FCAP Array Software Version 3. (Becton Dickinson, BD, San Jose, CA, USA). Ancestry analysis A total of 192 autosomal ancestry informative markers (AIM) developed for the Brazilian population [25] were genotyped on each individual to estimate an admixture and control its potential effect on the association between the frequency of Duffy polymorphisms (rs1275 and rs281477) and the phenotypes of interest. AIM genotyping was performed by OpenArray Real-Time PCR (Applied Biosystems). The evaluation of genomic ancestry was conducted using the Admixture program [26]. We used 2 bootstrap replicates (default) and k=3 (number of reference populations assumed for the analysis). As reference ancestral populations, individuals from the Human Genome Diversity Project and from the HapMap project were assumed as follows: Pima, Maya as Amerindians; YRI (Yoruba in Ibadan, Nigeria), LWK (Luhya in Webuye, Kenya), ASW (Americans of African Ancestry in SW, USA) as Africans; CEU (Utah Residents (CEPH) with Northern and Western European ancestry) and TSI (Tuscan in Italia) as European. Statistical analysis Statistical analysis was performed using the GraphPad Prism 5 software package (GraphPad Software, San Diego, CA, USA). The proportion analyses of gender and Duffy genotype frequency were performed by Fisher s exact test. Normal distribution of age, PVL and chemokine level were assessed using Kolgomorov Smirnov s test. Since these variables were non-parametric, statistical differences between the groups were determined by Mann Whitney s U-test. The Hardy Weinberg equilibrium (HWE) of Duffy polymorphisms was analysed using the exact test implemented in the web tool SNPstats [27]. Linear regression models were performed by the SNPstats tool to assess the proportion of variation in the HTLV-1 PVL level and chemokine concentrations explained by the Duffy polymorphisms, rs1275 and rs A codominant genetic model was assumed for the rs1275 polymorphism (genotypes A/A vs A/G vs G/G), whereas an additive model was assumed for the rs polymorphism (meaning that each allele copy modifies the risk in an additive form). When necessary, age, gender, and African, European or Native American ancestry were included as covariates in the analysis. The level of significance was considered to be P<.5. RESULTS We analysed 162, 135 and 71 SN to investigate whether Duffy genotypes influence the level of HTLV-1 PVL and chemokines and are associated with /TSP. As shown in Table 1, the proportion of male and female, and age, were statistically different among the groups, with a higher frequency of female and older individuals in the infected groups. The Duffy genotypes showed a similar distribution among the groups, with a higher frequency of the wild FY*B allele (polymorphisms 46T and 125A). The infected groups do not fit the HWE (: P=.6; : P=.26), with an excess of the homozygous at the locus rs (GATA box-1). HTLV-1 PVL was significantly higher in the group (median of 292 proviral copies 1 cells 1 ) than in the group (63 proviral copies 1 cells 1 ) (P<.1). In comparison to the group, the group also showed significantly higher median levels of IL8 (3.21 vs 2.48 pg ml 1 ; P<.1), but not CCL2 (4.84 vs 4.34 pg ml 1 ; P=.38) and CCL5 (378 vs 3643 pg ml 1 ; P=.443) levels. The highest levels of CCL2 (6.56 pg ml 1 ; SN vs : P=.1; SN vs : P=.15) and CCL5 (413 pg ml 1 ; SN vs : P=.36; SN vs : P=.87) were seen in the 129

4 da Silva-Malta et al., Journal of Medical Microbiology 217;66: Table 1. Demographic characteristic and Duffy genotype frequency in HTLV-1, and SN Duffy +, Duffy antigen expression on erythrocytes; Duffy, absence of Duffy antigen expression on erythrocytes. Comparisons of gender and genotype frequency between each group were analysed using Fisher s exact test, whereas age was analysed by Mann Whitney s U-test. Characteristic (n=162) (n=135) SN (n=71) P vs P vs SN P vs SN Male 64 (39.5 %) 38 (28.1 %) 43 (6.6 %).49.4 <.1 Female 98 (6.5 %) 97 (71.9 %) 28 (39.4 %) Mean age (range) 45 (2 73) 54 (15 87) 37 (2 65) <.1 <.1 <.1 GATA-1 box genotype (rs ) 46 T/T (Duffy +/+) 85 (52.5 %) 71 (52.6 %) 35 (49.3 %) T/C (Duffy +/ ) 54 (33.3 %) 46 (34.1 %) 26 (36.6 %) C/C (Duffy / ) 23 (14.2 %) 18 (13.3 %) 1 (14.1 %) FY*A and FY*B genotype (rs1275) 125A/A (FY*B/F*B) 87 (53.7 %) 79 (58.5 %) 39 (54.9 %) A/G (FY*B/F*A) 65 (4.1 %) 45 (33.4 %) 26 (36.6 %) G/G (FY*A/F*A) 1 (6.2 %) 11 (8.1 %) 6 (8.5 %) SN group, whereas IL8 was significantly lower in comparison to both infected groups (P<.1). The Duffy null phenotype was associated with a higher level of CCL2 in HTLV-1-infected individuals We evaluated whether Duffy polymorphisms influence the level of PVL in and groups and the level of chemokines in infected and non-infected individuals. patients showed significantly higher PVL than the group, independent of the Duffy genotype. Significant intragroup differences were not observed for polymorphisms rs1275 and rs (Table 2 and Fig. 1). On the other hand, the level of CCL2 was associated with the GATA-1 box polymorphism, being significantly lower in, and SN who did not express Duffy antigen on erythrocytes (Table 3 and Fig. 2). For the other chemokines, there were no plasma-level associations with Duffy genotypes in any of the three groups analysed (Table 3). Association of the Duffy null phenotype with the total WBC count and percentage of neutrophils Since it has been described that individuals with the Duffy null phenotype ( 46CC) have significantly lower total WBC counts [2] and percentage of neutrophils [21] when compared to Duffy-positive ( 46T) individuals, we analysed the level of these cells in the three groups according to Duffy genotypes (Table 4 and Fig. 3). A significantly lower count of total WBC was associated with the absence of Duffy antigen on erythrocytes in the group and in all infected individuals (+). In turn, individuals who did not express DARC on RBCs had a significantly decreased percentage of neutrophils when compared to subjects with Duffy-positive genotypes in all of the groups analysed (Table 4 and Fig. 3). Analysis of the association between the Duffy null phenotype, CCL2 and leukocytes, adjusted for ancestry Considering that polymorphisms in the Duffy gene are not homogenously distributed in world populations, we investigated the admixture profile of infected individuals to confirm whether the observed associations were true or spurious. This analysis could not be performed for all individuals due to samples being unavailable for testing. Using genetic markers informative for European, African and Amerindian ancestry, the individual ancestry Table 2. HTLV-1 PVL in and according to Duffy polymorphisms HTLV-1 PVL: proviral copies 1 4 cells -1. The P-value indicates the significance of the association test between the level of each chemokine and gene polymorphism, adjusted for age and gender. Group Median (log) HTLV-1 PVL FY*A and FY*B genotype (rs1275) GATA-1 box genotype (rs ) A/A A/G G/G P T/T T/C C/C P 6 (1.78) 79 (1.9) 7 (.78) (1.79) 71 (1.86) 63 (1.8) (2.5) 372 (2.57) 174 (2.24)* (2.37) 396 (2.6) 282 (2.44) (2.24) 13 (2.12) 111 (2.5) (2.12) 165 (2.22) 173 (2.24).97 *Excluding one outlier (.1 proviral copies 1 cells -1 ). 121

5 da Silva-Malta et al., Journal of Medical Microbiology 217;66: rs rs log (proviral copies/1 cells) log (proviral copies/1 cells) AA AG GG AA AG 2 GG TT CT CC TT CT CC Fig. 1. HTLV-1 PVL in and according to Duffy genotypes. Polymorphism rs1275 (A>G) defines alleles FY*B and FY*A, respectively. Polymorphism rs (T>C) defines the presence or absence of Duffy antigen on erythrocytes, respectively. There was no significant difference among the three genotypes in the same clinical group. A linear regression-based test under a codominant and additive model of the allelic effect was used for statistical analysis of rs1275 and rs , respectively. The line shows the median value. estimated for these three populations was not different between the (n=123) and (n=5) groups (Fig. 4). Since the admixture difference among the groups was not significant, the analysis considering this parameter is not presented. DISCUSSION It is well described that HTLV-1 may trigger a neuroinflammatory disease, but the reason why only some infected carriers are susceptible to the pathogenic effects of the virus is still unknown. Host factors such as genetic polymorphisms are involved in the ability of a carrier to establish an efficient innate and adaptive immune response against HTLV-1. For example, it was previously described that the variability observed in human leukocyte antigens (HLA) are relevant to the risk of developing /TSP [28, 29]. Considering that DARC on the surface of RBCs is a high affinity chemokine receptor that does not trigger activation signals, some studies suggest that DARC function as a sink to bind and internalize chemokines, clear them from sites of inflammation and systemic circulation, and prevent WBC activation [19, 3]. On the other hand, DARC expressed on endothelial cells has a role in enhancing leukocyte trafficking [18]. Thus, we hypothesized that Duffy gene polymorphisms that Table 3. Plasma levels of IL8, CCL2 and CCL5 in and according to Duffy polymorphisms The data are median values of chemokines in pg ml 1. The data were analysed using a linear regression-based test under an additive model of the allelic effect. The P-value indicates the significance of the association test between the level of each chemokine and Duffy polymorphism, adjusted for age and gender. The boldfaced text shows significant P values. Group Chemokine FY*A and FY*B genotype (rs1275) GATA-1 box genotype (rs ) A/A A/G G/G P T/T T/C C/C P IL CCL CCL IL CCL CCL IL CCL <.1 CCL SN IL CCL CCL

6 da Silva-Malta et al., Journal of Medical Microbiology 217;66: (a) IL8 (pg ml 1 ) (b) IL8 (pg ml 1 ) A/A A/G G/G A/A A/G G/G A/A A/G G/G A/A A/G G/G A/A A/G G/G A/A A/G G/G A/A A/G G/G A/A A/G G/G A/A A/G G/G SN SN SN CCL2 (pg ml 1 ) CCL2 (pg ml 1 ) CCL5 (pg ml 1 ) CCL5 (pg ml 1 ) TT TC SN CC TT TC CC TT TC CC TT TC CC TT TC CC TT TC CC TT TC SN SN CC TT TC CC TT TC CC Fig. 2. Levels of CCL2 according to the Duffy polymorphisms, rs1275 and rs Significant intra-group differences were not seen among the three genotypes for the polymorphism rs1275. On the other hand, a higher level of CCL2 was associated with 46T (polymorphism rs ) that defines the presence of Duffy antigen on erythrocytes in all of the groups analysed. A linear regression-based test under an additive model of the allelic effect was used for statistical analysis. When a significant difference was shown by the linear regression test, we used the Kruskal Wallis test followed by the Dunn multiple comparison test to compare each Duffy genotype with the others (TT vs TC; TT vs CC; TC vs CC) in the same group; * represents a significant difference (P<.5). The line shows the median value. alter their expression on the surface of RBCs may be associated with HTLV-1 infection outcome. In the current study, we showed that the level of PVL was not associated with the Duffy antigen polymorphism (rs1275) nor with presence or absence of Duffy antigen on RBCs (rs281477) in HTLV-1-infected individuals, either asymptomatic or with neurologic disease. In addition, the frequency of Duffy genotypes was not significantly different among, and the seronegative control group. We also analysed whether the levels of chemokines were associated with DARC polymorphisms in HTLV-1 carriers and uninfected individuals. The levels of CCL2, IL8 and CCL5 chemokines were not influenced by the polymorphism rs1275 (125A>G Asp42Gly) in any of the groups analysed, whereas rs ( 46C) was associated with lower plasma level of CCL2 in infected ( and groups) and uninfected individuals. It is interesting to observe that in spite of Duffy genotypes, significant increases in the levels of IL8 and lower levels of CCL2 were seen in HTLV-1-infected groups, when compared to the seronegative group. The control group also showed higher levels of CCL5, but this difference was significant only in comparison to. Some studies have reported increased levels of CCL2 and CCL5 associated with HTLV-1 infection or with /TSP [9, 31]. We have previously discussed that the divergent chemokine data observed among different studies in the context of HTLV-1 infection may be due to the heterogeneity of the individuals that had composed the analysed groups regarding age, gender, time of infection, duration of disease, PVL level and other possible confounding factors [32]. The present results may point to the inclusion of a Duffy null genotype among them. Previous studies analysed the association of Duffy genotypes with chemokines, but not in the context of virus infection. 1212

7 da Silva-Malta et al., Journal of Medical Microbiology 217;66: Table 4. Total WBC and percentage of neutrophils in and according to GATA-1 box polymorphism The data are median values of total WBC and percentage of neutrophils and were analysed using the linear regression-based test under an additive model of the allelic effect. The P-value indicates the significance of the association test between each parameter and GATA-1 box polymorphism, adjusted for age and gender. The boldfaced text shows significant P values. Group Parameter GATA-1 box genotype (rs ) (n=148) (n=8) T/T T/C C/C P Total WBC <.1 % neutrophils <.1 Total WBC % neutrophils <.1 + Total WBC <.1 SN (n=68) % neutrophils <.1 Total WBC % neutrophils Schnabel and colleagues [33] showed an association between rs1275 and lower levels of CCL2, IL8 and CCL5 in serum from Caucasians, but not in plasma. They reported that approximately 2 % of variation in CCL2 serum concentration could be explained by rs1275, and they considered that the discrepancy between serum and plasma was due to the additional CCL2 that is liberated from Duffy antigen on erythrocyte by clotting. Voruganti and coworkers [34] also showed that rs1275 was associated with elevated levels of CCL2 in serum samples from Hispanic children, being responsible for approximately 1 % of the variability in CCL2 levels. They did not analyse plasma samples. Regrettably, these studies did not investigate the effect of GATA-1 polymorphism due to the rarity of 46C mutation in the ethnic groups analysed. Similar to our results, decreased plasma levels of CCL2 were previously described in individuals with the Duffy null genotype ( 46 C/C) compared to persons who express DARC on the surface of RBCs [35, 36]. On the other hand, Lee and colleagues [37] showed that individuals homozygous for the Duffy null genotype had elevated levels of CCL2 and IL8 in plasma following lipopolysaccharide stimulation of whole blood, in comparison to heterozygote subjects. This observation is in agreement with the initial idea of the erythrocyte Duffy antigen acting as a sink for circulating chemokines [3]. However, other studies suggested that DARC delays the disappearance of chemokines from the plasma, thus maintaining plasma chemokine concentrations, since the absence of DARC in knockout mice was associated with lower levels of CCL2 and more rapid clearance of this chemokine from the plasma [38]. The role of chemokines as scavenging receptors is not attributed to DARC anymore. Instead, DARC may play a role in the Total WBC ( 1 µl 1 ) * * * * Neutrophil count (%) * * * * * TT TC SN CC TT TC CC TT TC CC TT TC CC TT TC CC SN TT TC CC Fig. 3. Distribution of total WBC count and percentage of neutrophil in SN (n=68), (n=148) and (n=8) according to the presence or absence of Duffy antigen on erythrocytes. A linear regression-based test under an additive model of the allelic effect was used for statistical analysis. When a significant difference was shown by the linear regression test, we used the Kruskal Wallis test followed by the Dunn multiple comparison test to compare each Duffy genotype with the others (TT vs TC; TT vs CC; TC vs CC) in the same group; * represents a significant difference (P<.5). The line shows the median value and whiskers show minimum and maximum values. 1213

8 da Silva-Malta et al., Journal of Medical Microbiology 217;66: % Asymptomatic group 1 % group 8 % 8 % 6 % 6 % 4 % 4 % 2 % 2 % % % EUR AFR AMR EUR AFR AMR Fig. 4. Proportion of individual European, African and Native American admixture in (n=123) and (n=5). Each vertical bar represents one subject. transport of chemokines across barriers, allowing its concentration in tissues being difficult to reach [39]. It is described that the Duffy null genotype is linked with a decrease of total WBC count and percentage of neutrophils, as shown by genome-wide association studies [2, 21]. We observed an association between the Duffy null genotype and a decreased percentage of neutrophils in all groups analysed. The association between the Duffy null genotype and a decreased total WBC count was also seen for the group, and probably was not significant in the and SN groups because of the lower number of individuals analysed in these groups. It could be suggested that the association of lower plasma CCL2 levels with the 46 C/C genotype should be due to the decreased amount of WBC and neutrophils. However, there was no significant correlation between WBC count or percentage of neutrophils and CCL2 levels in any of the groups analysed (data not shown). Because of the assumption that Duffy genotypes are associated with decreased counts of immune cells and can regulate the amount of circulating chemokines, different studies investigated the relationship between Duffy genotypes and HIV infection. In 28, He and co-workers [4] originally described that HIV-1 binds to DARC on the surface of RBCs, enabling the virus to be transferred to target cells. They calculated that African Americans who did not express DARC on their RBCs are over 4 % more susceptible to acquiring HIV-1 infection. On the other hand, when HIV infection was already established, individuals possessing the Duffy null phenotype had a protective effect on HIV disease, including HIV-associated dementia. The association of the DARC 46 C/C genotype with susceptibility to acquire HIV and with survival advantage in HIV infection was also attributed to decreased neutrophils and WBC counts [41, 42]. After these studies were published, a number of other authors could not replicate these findings. Analysing HIV viral load, CD4+ T-cell count, rate of CD4+ T-cell decline, time to AIDS, time to death and other parameters, studies with different populations failed to find an association between the DARC genotype and progression to AIDS, or risk for HIV acquisition [43 46]. These discrepancies could be attributed to the failure of the first studies to perform an appropriate control for population stratification, since DARC 46C has a large stratification effect due to being a marker for African ancestry, and thus they contributed to the creaton of spurious genetic associations. Another work analysed the effect of the DARC polymorphism in the severity of HCV infection [47]. Studying three different cohorts of patients with HCV infection, the authors observed a significant association of serum CCL2 levels with fibrosis progression, but they did not detect the association between polymorphism rs1275 (125A>G, FY*B >FY*A) and HCV infection or liver fibrosis scores as being indicative of disease severity. Although the effect of DARC on the circulating chemokines could suggest a role in the chronic virus pathologies, these previous studies showed that DARC is not directly associated with the acquisition or disease severity in HIV or HCV infection. Our results provide evidence that HTLV-1 should be included on this list. In addition, our results draw attention to the fact that Duffy genotypes may be a confounding factor in studies to define inflammatory biomarkers in HTLV-1-infected populations with diverse ancestrally, since Duffy polymorphisms present different frequencies among worldwide populations. In summary, we showed, for the first time, that the frequency of the Duffy null genotype was not significantly different among, and SN, and that PVL level was not associated with DARC polymorphisms. Thus, our results support the absence of the effect of Duffy genotypes on the occurrence of HTLV infection or /TSP. Decreased CCL2 levels were associated with the 46C polymorphism, but although DARC influences circulating concentrations of chemokines, it does not seem to be a genetic factor that characterizes risk for /TSP. It is important to emphasize that we ensured an appropriate control for population stratification and we believe that our cohort is 1214

9 da Silva-Malta et al., Journal of Medical Microbiology 217;66: powered to detect the true effects of the DARC genotype, since we observed the association between 46 C/C with low WBC and neutrophils count, as expected. Funding information This study was funded by Fundaç~ao Hemominas, FAPEMIG (Fundaç~ao de Amparo a Pesquisa do Estado de Minas Gerais) and CNPq (Conselho Nacional de Desenvolvimento Científico e Tecnológico). M. C. F. Silva-Malta, J. C. Guimar~aes, C. C. Sales, P. C. Gonçalves, D. G. Chaves and M. L. Martins received fellowships from FAPEMIG and A. B. F. Carneiro-Proietti from CNPq. Acknowledgements The authors would like to thank all cohort participants and members of the Interdisciplinary HTLV Research Group for their ongoing collaboration. Conflicts of interest The authors declare that there are no conflicts of interest. Ethical statement This study was approved by the Ethical Committee of Fundaç~ao Hemominas and all subjects gave informed consent to the work. References 1. Uchiyama T. Human T cell leukemia virus type I (HTLV-I) and human diseases. Annu Rev Immunol 1997;15: Bangham CR. The immune response to HTLV-I. Curr Opin Immunol 2;12: Bangham CR. HTLV-1 infections. J Clin Pathol 2;53: Nagai M, Usuku K, Matsumoto W, Kodama D, Takenouchi N et al. Analysis of HTLV-I proviral load in 22 /TSP patients and 243 asymptomatic HTLV-I carriers: high proviral load strongly predisposes to /TSP. J Neurovirol 1998;4: Olindo S, Lezin A, Cabre P, Merle H, Saint-Vil M et al. HTLV-1 proviral load in peripheral blood mononuclear cells quantified in 1 /TSP patients: a marker of disease progression. J Neurol Sci 25;237: Lezin A, Olindo S, Oliere S, Varrin-Doyer M, Marlin R et al. Human T lymphotropic virus type I (HTLV-I) proviral load in cerebrospinal fluid: a new criterion for the diagnosis of HTLV-I-associated myelopathy/tropical spastic paraparesis? J Infect Dis 25;191: Furtado MS, Andrade RG, Romanelli LC, Ribeiro MA, Ribas JG et al. Monitoring the HTLV-1 proviral load in the peripheral blood of asymptomatic carriers and patients with HTLV-associated myelopathy/tropical spastic paraparesis from a Brazilian cohort: ROC curve analysis to establish the threshold for risk disease. J Med Virol 212;84: Zlotnik A, Yoshie O. Chemokines: a new classification system and their role in immunity. Immunity 2;12: Guerreiro JB, Santos SB, Morgan DJ, Porto AF, Muniz AL et al. Levels of serum chemokines discriminate clinical myelopathy associated with human T lymphotropic virus type 1 (HTLV-1)/tropical spastic paraparesis (/TSP) disease from HTLV-1 carrier state. Clin Exp Immunol 26;145: Sato T, Coler-Reilly A, Utsunomiya A, Araya N, Yagishita N et al. CSF CXCL1, CXCL9, and neopterin as candidate prognostic biomarkers for HTLV-1-associated myelopathy/tropical spastic paraparesis. PLoS Negl Trop Dis 213;7:e Neote K, Darbonne W, Ogez J, Horuk R, Schall TJ. Identification of a promiscuous inflammatory peptide receptor on the surface of red blood cells. J Biol Chem 1993;268: Hadley TJ, Peiper SC. From malaria to chemokine receptor: the emerging physiologic role of the duffy blood group antigen. Blood 1997;89: Chaudhuri A, Polyakova J, Zbrzezna V, Williams K, Gulati S et al. Cloning of glycoprotein D cdna, which encodes the major subunit of the Duffy blood group system and the receptor for the plasmodium vivax malaria parasite. Proc Natl Acad Sci USA 1993;9: Meny GM. The Duffy blood group system: a review. Immunohematology 21;26: Howes RE, Patil AP, Piel FB, Nyangiri OA, Kabaria CW et al. The global distribution of the Duffy blood group. Nat Commun 211;2: Tournamille C, Colin Y, Cartron JP, Le van Kim C. Disruption of a GATA motif in the Duffy gene promoter abolishes erythroid gene expression in Duffy-negative individuals. Nat Genet 1995;1: Iwamoto S, Li J, Sugimoto N, Okuda H, Kajii E. Characterization of the Duffy gene promoter: evidence for tissue-specific abolishment of expression in Fy(a-b-) of black individuals. Biochem Biophys Res Commun 1996;222: Lee JS, Frevert CW, Wurfel MM, Peiper SC, Wong VA et al. Duffy antigen facilitates movement of chemokine across the endothelium in vitro and promotes neutrophil transmigration in vitro and in vivo. J Immunol 23;17: Pruenster M, Rot A. Throwing light on DARC. Biochem Soc Trans 26;34: Nalls MA, Wilson JG, Patterson NJ, Tandon A, Zmuda JM et al. Admixture mapping of white cell count: genetic locus responsible for lower white blood cell count in the Health ABC and Jackson Heart studies. Am J Hum Genet 28;82: Reich D, Nalls MA, Kao WH, Akylbekova EL, Tandon A et al. Reduced neutrophil count in people of African descent is due to a regulatory variant in the Duffy antigen receptor for chemokines gene. PLoS Genet 29;5:e World Health Organization. Report from the scientific group on HTLV-1 infection and its associated diseases, convened by the regional office for the Western Pacific of the World Health Organization in Kagoshima, Japan, 1 15 December Wkly Epidemiol Rec 1989;49: Andrade RG, Ribeiro MA, Namen-Lopes MS, Silva SM, Basques FV et al. Evaluation of the use of real-time PCR for human T cell lymphotropic virus 1 and 2 as a confirmatory test in screening for blood donors. Rev Soc Bras Med Trop 21;43: Tanaka M, Takahahi J, Hirayama F, Tani Y. High-resolution melting analysis for genotyping Duffy, Kidd and Diego blood group antigens. Leg Med 211;13: Santos HC, Horimoto AV, Tarazona-Santos E, Rodrigues-Soares F, Barreto ML et al. A minimum set of ancestry informative markers for determining admixture proportions in a mixed American population: the Brazilian set. Eur J Hum Genet 216;24: Alexander DH, Novembre J, Lange K. Fast model-based estimation of ancestry in unrelated individuals. Genome Res 29;19: Sole X, Guinó E, Valls J, Iniesta R, Moreno V. SNPStats: a web tool for the analysis of association studies. Bioinformatics 26;22: Jeffery KJ, Usuku K, Hall SE, Matsumoto W, Taylor GP et al. HLA alleles determine human T-lymphotropic virus-i (HTLV-I) proviral load and the risk of HTLV-I-associated myelopathy. Proc Natl Acad Sci USA 1999;96: Jeffery KJ, Siddiqui AA, Bunce M, Lloyd AL, Vine AM et al. The influence of HLA class I alleles and heterozygosity on the outcome of human T cell lymphotropic virus type I infection. J Immunol 2;165: Darbonne WC, Rice GC, Mohler MA, Apple T, Hebert CA et al. Red blood cells are a sink for interleukin 8, a leukocyte chemotaxin. J Clin Invest 1991;88: Vine AM, Heaps AG, Kaftantzi L, Mosley A, Asquith B et al. The role of CTLs in persistent viral infection: cytolytic gene expression in CD8+ lymphocytes distinguishes between individuals with a high or low proviral load of human T cell lymphotropic virus type 1. J Immunol 24;173:

10 da Silva-Malta et al., Journal of Medical Microbiology 217;66: Chaves DG, Sales CC, de Cassia Gonçalves P, da Silva-Malta MC, Romanelli LC et al. Plasmatic proinflammatory chemokines levels are tricky markers to monitoring HTLV-1 carriers. J Med Virol 216;88: Schnabel RB, Baumert J, Barbalic M, Dupuis J, Ellinor PT et al. Duffy antigen receptor for chemokines (Darc) polymorphism regulates circulating concentrations of monocyte chemoattractant protein-1 and other inflammatory mediators. Blood 21;115: Voruganti VS, Laston S, Haack K, Mehta NR, Smith CW et al. Genome-wide association replicates the association of Duffy antigen receptor for chemokines (DARC) polymorphisms with serum monocyte chemoattractant protein-1 (MCP-1) levels in Hispanic children. Cytokine 212;6: Jilma-Stohlawetz P, Homoncik M, Drucker C, Marsik C, Rot A et al. Fy phenotype and gender determine plasma levels of monocyte chemotactic protein. Transfusion 21;41: Mayr FB, Spiel AO, Leitner JM, Firbas C, Kliegel T et al. Duffy antigen modifies the chemokine response in human endotoxemia. Crit Care Med 28;36: Lee JS, Wurfel MM, Matute-Bello G, Frevert CW, Rosengart MR et al. The Duffy antigen modifies systemic and local tissue chemokine responses following lipopolysaccharide stimulation. J Immunol 26;177: Fukuma N, Akimitsu N, Hamamoto H, Kusuhara H, Sugiyama Y et al. A role of the Duffy antigen for the maintenance of plasma chemokine concentrations. Biochem Biophys Res Commun 23; 33: Ulvmar MH, Hub E, Rot A. Atypical chemokine receptors. Exp Cell Res 211;317: He W, Neil S, Kulkarni H, Wright E, Agan BK et al. Duffy antigen receptor for chemokines mediates trans-infection of HIV-1 from red blood cells to target cells and affects HIV-AIDS susceptibility. Cell Host Microbe 28;4: Kulkarni H, Marconi VC, He W, Landrum ML, Okulicz JF et al. The Duffy-null state is associated with a survival advantage in leukopenic HIV-infected persons of African ancestry. Blood 29;114: Ramsuran V, Kulkarni H, He W, Mlisana K, Wright EJ et al. Duffynull-associated low neutrophil counts influence HIV-1 susceptibility in high-risk South African black women. Clin Infect Dis 211;52: Walley NM, Julg B, Dickson SP, Fellay J, Ge D et al. The Duffy antigen receptor for chemokines null promoter variant does not influence HIV-1 acquisition or disease progression. Cell Host Microbe 29;5: Julg B, Reddy S, van der Stok M, Kulkarni S, Qi Y et al. Lack of Duffy antigen receptor for chemokines: no influence on HIV disease progression in an African treatment-naive population. Cell Host Microbe 29;5: Horne KC, Li X, Jacobson LP, Palella F, Jamieson BD et al. Duffy antigen polymorphisms do not alter progression of HIV in African Americans in the MS cohort. Cell Host Microbe 29; 5: Winkler CA, An P, Johnson R, Nelson GW, Kirk G. Expression of Duffy antigen receptor for chemokines (DARC) has no effect on HIV-1 acquisition or progression to AIDS in African Americans. Cell Host Microbe 29;5: Lettow I, Berres ML, Schmitz P, Müller T, Berg T et al. A Duffy antigen receptor for chemokines (DARC) polymorphism that determines pro-fibrotic chemokine serum concentrations is not directly associated with severity of hepatitis C infection. Hum Immunol 211;72: Five reasons to publish your next article with a Microbiology Society journal 1. The Microbiology Society is a not-for-profit organization. 2. We offer fast and rigorous peer review average time to first decision is 4 6 weeks. 3. Our journals have a global readership with subscriptions held in research institutions around the world. 4. 8% of our authors rate our submission process as excellent or very good. 5. Your article will be published on an interactive journal platform with advanced metrics. Find out more and submit your article at microbiologyresearch.org. 1216