Molecular analysis of three FUT 3 gene single nucleotide polymorphisms and their relationship with the lewis erythrocytary phenotype in a human population of japanese-ancestry living in Tomé Açu , a town in the Brazilian Amazon

The Lewis blood group system involves two major antigens, Le and Le. Their antigenic determinants are not primary gene products but are synthesized by the transfer of sugar subunits to a precursory chain by a specific enzyme which is the product of the FUT3 gene (Lewis gene). The presence of three FUT3 gene single nucleotide polymorphisms (SNPs) (59T > G; 508G > A and 1067T > A) was related to the Lewis phenotype of erythrocytes from 185 individuals of Japanese ancestry living in the town of Tomé-Açu in the Brazilian Amazon region. This relationship was detected using a serological hemagglutination test and the Dot-ELISA assay along with the molecular technique PCR-RFLP. We found that the three SNPs investigated in this study only accounted for a proportion of the Lewis-negative phenotype of the erythrocytes.


Introduction
The Lewis (FUT3 or Le) gene is located on the short arm of chromosome 19 (Mollicone et al., 1990(Mollicone et al., , 1992) ) and codes for α1-3/4-L-fucosyltransferase III (FucT III, also called the Lewis enzime), an enzyme with 361 amino acids which catalyzes the transfer of a fucose unit to the subterminal GlcNAc unit of type 1 precursory chains (De Vries et al., 2001).The Lewis blood group system is comprised of two major antigens, one being the monofucosylated Le a antigen formed by the action of FucT III on a type 1 precursor and the other the difucosylated Le b antigen produced by the epistatic interaction between the FUT3 and secretor (FUT2) gene products on a type 1 precursor in tissues such as salivary glands, digestive mucosa and respiratory mucosa (Watkins, 1980;Oriol et al., 1986).
The Lewis antigens are not synthesized in erythrocyte progenitors, erythrocytes acquiring their Le a and Le b epitopes by adsorbing Lewis antigenic glycosphingolipids from plasma.Accordingly, typing of Lewis phenotypes is difficult and is sometimes misjudged because of weak hemagglutination as a result of the low specificities of anti-Le a and anti-Le b antibodies and the low number of antigens on red cells, resulting in low titers.(Nishihara et al., 1993;Liu et al., 1996).
The situation is even more complex due to the presence of FUT3 single nucleotide polymorphisms (SNPs) which can result in nucleotide mutations that generate enzymes which may have different catalytic activities or even be inactive (Nishihara et al., 1993(Nishihara et al., , 1994;;Kudo et al., 1996;Cakir et al., 2002;Cooling and Gu, 2003;Soejima et al., 2004;Jost et al., 2005).The most frequent FUT3 SNPs are the 202T > C, 314C > T, 508G > A and 1067T > A polymorphisms (le genes) which affect and inactivate the catalytic domain of FucT III enzyme (Nishihara et al., 1993(Nishihara et al., , 1994;;Kudo et al., 1996), while a further SNP, 59T > G, is known to cause the substitution of an amino acid in the transmembrane region but its effect on enzymatic activity has not yet been defined (Elmgren et al., 1996).
In 1929, the first Japanese colonists to immigrate to the Northern region of Brazil arrived in the state of Pará and settled predominantly in areas near the town of Tomé Açu, there initially being 189 people divided into 43 families which settled permanently in the area supported by incentives provided by the state government for developing agricultural activities.As of 2006, about two percent of the inhabitants of Tomé Açu are of Japanese ancestry (SEPOF, 2006).
The aim of the present study was to establish the frequency of three FUT3 SNPs and investigate the expression of red blood cell Lewis antigens in a population of Japanese ancestry in the Northern region of Brazil.

Participants
Blood samples were obtained from 185 apparently healthy individuals randomly selected (134 females, 51 males; median age 46, range 20 to 88 years) of Japanese ancestry living in the town of Tomé-Açu (02°25'0" S; 48°09'09" W) in the northeastern region of the Brazilian state of Pará.The participants in this study were either born in Japan or had both parents and all grandparents born in Japan.There were no kinship ties between the participants and their paternal origin was from 22 different Japanese cities situated on the three main japanese islands of Hokkaido, Honshu and Kyushu.
This study was approved by the NMT -Federal University of Pará Ethics Research Committee (protocol number 005/2000) and all individuals gave their informed consent to participate in the study.

Peripheral blood samples
From each participant we collected 5 mL of peripheral blood in tubes containing EDTA.Erythrocytes were subjected to Lewis blood group typing, and genomic DNA was extracted from white blood cells using the QIAamp DNA Blood Mini kit (QIAGEN, Inc.Valencia, CA, USA).

Typing of Lewis red blood group phenotypes
Lewis blood group phenotypes of erythrocytes were determined using anti-Le a and anti-Le b monoclonal antibodies (Ortho Diagnostic System, Raritan, NJ, USA) and hemagglutination tests according to the manufacturer's instructions.If a Lewis negative reaction had occurred in previous hemagglutination tests, the phenotypic expression of Lewis antigen on erythrocytes was reexamined using the dot-blot-ELISA assay (Pflug et al., 1989) and the same batches of monoclonal antibody used for the hemagglutination tests.The red cell suspensions were prepared at a concentration of 1:2, 1:4 and 1:10 in distilled water and 5 μL of appropriate dilution were spotted onto nitrocellulose membranes (Schleicher & Schuell, Dassel -Germany).Appropriate controls were used at all stages of testing.

SNP Screening
The FUT3 genotype was assigned based on the detection of the 59T > G, 508G > A and 1067T > A SNPs by the PCR restriction fragment length polymorphism (PCR-RFLP) method.The primers used (Invitrogen, São Paulo -Brasil) being listed in Table 1.
The DNA amplification were performed in a final volume of 25 μL of LA Taq buffer (Perkin Elmer, USA) containing 2.5 mM MgCl 2 , 20 mM of dNTPs, 25 pmol of each primer, 1 μL of Taq polymerase (Perkin Elmer, USA), and 40 ng of sample genomic DNA as template.The PCR cycles for the first, second and third set of primers in Table 1 consisted of initial denaturing at 94 °C for 5 min followed by 35 cycles of denaturing at 94 °C for 30 s, annealing at 65 °C for 30 s and extension at 72 °C for 1 min.
To detect the FUT3 1067T > A SNP, one μL of the product from the first PCR was used as the template for a nested PCR (denominated the 'second PCR' in Table 1) which was carried out under the same conditions the first PCR, except that the fourth primer set in Table 1 was used and the annealing temperature was 60 °C.
The FUT3 PCR products were digested by three different restriction endonucleases: MspI for the 59T > G SNP products; PvuII for the 508G > A SNP products; and HindIII for the 1067T > A SNP products.The FUT3 digestion products were separated on 7% (w/v) polyacrylamide gel (GibcoBRL, New York -USA) for the 59T/59G and 508G/508A SNP products and 10% (w/v) polyacrylamide gel for the FUT3 1067T/1067A SNP products.The electrophoresis was carried out using 10x TEB buffer (90 mM Tris; 2 mM EDTA, 90 mM boric acid pH 8.6), at 150 V, 25 mA, for 5 h and then silver stained (17% solution of silver nitrate).The primers, restriction enzymes and fragment sizes produced after PCR and digestion are listed in Table 1.

Statistical analysis
Data was analyzed by calculating the Hardy-Weinberg Equilibrium and applying the chi-squared (χ 2 ) heterogeneity test using the CLUMP computer program (Sham and Curtis, 1995) and BioEstat 3.0 (Ayres et al., 2003).The frequencies of the haplotypes were estimated using the ARLEQUIN version 2.000 software.

Results
We were successful in determining the Lewis genotype using PCR-RFLP detection of the 59T > G, 508G > A Francez et al.
and 1067T > A SNPs.In addition, we detected five different FUT3 haplotypes, estimated using the ARLEQUIN program (Table 2).
Our study of Lewis phenotype showed that 44 of the 185 individuals evaluated using the hemagglutination test and dot-ELISA assay were negative for Lewis antigen expression in erythrocytes (Table 3).
The complexity of the polymorphism profiles obtained in our study is supported by the observation that 20 out of 44 of the individuals with the Lewis negative phenotype had the FUT3 59G allele associated with the FUT3 508A or FUT3 1067A alleles in le 59, 508 /le 59, 508 , le 59, 508 /le 59,  1067 and le 59, 1067 /le 59, 1067 genotypes.The others 24/44 individuals with the Lewis negative phenotype presented the Le/Le, Le/ le 59, 508 and Le/ le 59, 1067 genotypes (Table 3).
We also compared the FUT3 Lewis genotype frequencies found in the current study with those reported by other authors and found that the frequencies obtained by us in the present study (Table 4) were similar to the frequency distributions (χ 2 4 = 2.603; 0.90 > p > 0.80) reported by Liu et al. (1996) in a Japanese population, although the authors did not cite the location of the population.But were significantly different to the frequencies reported by Liu et al. (1999) for two Chinese populations, one from Shenyang (χ 2 4 = 15.603;p < 0.05) and the other from Guangzhou (χ 2 4 = 15.993p < 0.05).
In our study the observed Le/Le genotype frequency was 39.46% (Table 4), significantly lower (p < 0.001) compared with the 61.72% (  (1999) for the same genotype in a population from Shenyang, China.

Discussion
In our study, the genotype frequencies of the FUT3 59T > G, 508G > A and 1067T > A SNPs was similar to data published for other Asiatic populations, especially Japanese populations (Table 4).This was clearly observable for the FUT3 59G and FUT3 508A alleles, predominant among oriental populations (Liu et al., 1996), which were present at a high frequency in our population sample (Table 3), reflecting the ethnic origin of the individuals sampled.The frequencies seen in our study were also different from the frequencies obtained for other populations, such as the European population studied by Pang et al. (1998) and the Chinese population investigated by Liu et al. (1999).
The FUT3 59G allele isolated (le 59 ), found in three individuals in our study, was not described by Liu et al., (1996) in a study involving Japanese populations but was observed at a low frequency in European (Pang et al., 1998) and Chinese (Liu et al., 1999) populations.Conversely, the FUT3 508A allele isolated (le 508 ) was not detected by us but was described by Liu et al. (1996) in an individual of Japanese origin.Because they are rare, both the FUT3 59G and FUT3 508A alleles, may be present in some populations but absent in others of the same ethnic group, due to the low frequency of these alleles, or alternatively, due to the phe-nomenon of population isolation and other processes that restrict gene flow.
Based on previous studies (Nishihara et al., 1993(Nishihara et al., , 1994;;Kudo et al., 1996, Liu et al., 1999), we inferred that individuals who presented the FUT3 59G allele associated with the FUT3 508A or FUT3 1067A allele would present a cis-type chromosomal disposition forming the le 59, 508 or le 59, 1067 haplotypes.This is because isolated FUT3 508A or FUT3 1067A alleles, although described, are uncommon, and their association with the FUT3 59G allele in a trans arrangement would be extremely rare.When estimating allele and genotype frequencies, this possibility can be discarded, but to eliminate totally this risk direct DNA sequencing is needed, which we did not carry out in our study.
The fact that in our study all the individuals carrying genotypes le 59, 508 /le 59, 508 , le 59, 508 /le 59, 1067 and le 59, 1067 /le 59,  1067 presented a negative Lewis phenotype, confirms that these mutations, especially those situated in the catalytic domain, eliminate enzymatic activity through changing the protein conformation of the enzyme, leading to the inactivity of the FUT3 gene product.The le 59 /le 59, 1067 genotype identified by us in a Lewis-positive individual suggests that the FUT3 59G allele alone is not capable of eliminating the enzymatic activity of the FUT3, although some studies (Elmgren et al., 1993(Elmgren et al., , 1996;;Kudo et al., 1996;Liu et al., 1996Liu et al., , 1999) ) suggest that the presence of this allele may result in an increased susceptibility to the occurrence of the false Lewis-negative phenotype.Comparison between the three FUT3 polymorphisms of individuals with Lewis-negative phenotype erythrocytes revealed that only 20 out of 44 presented the le 59, 508 and/or le 59, 1067 haplotypes related to FUT3 inactivation in both FUT3 alleles (Table 4).This implies that there were other FUT3 polymorphisms which were not detected in our study.However, on the other hand, such a discrepancy may be the result of the difficulty in characterizing the Lewis phenotype in erythrocytes, which may lead to false negatives because of specific biological conditions, as has been mentioned in this paper.
Our findings indicate that the frequency of the Lewis-negative trait, obtained by molecular analysis of the three FUT3 SNPs studied, was about 10.8% in the population sample studied, which is close to the serological phenotype frequencies previously described for Asiatic populations (Lin-Chu et al., 1988;Liu et al., 1996Liu et al., , 1999)).
Lewis antigens are particularly important in kidney transplants because rejection is lower in transplants involving Lewis-compatible donors and receptors (Oriol et al., 1978;Williams et al., 1978;Oriol and Danilovics, 1980;Myser et al., 1985;Blajchman et al., 1985;Delmotte et al., 2002), and the molecular test described in this paper may be helpful in avoiding the false negative results sometimes observed whit the serological methods available for typing Lewis antigens.

Table 1 -
FUT3 gene single nucleotide polymorphism (SNP), PCR primer sequences and fragment sizes in base pairs (bp) before and after digestion by each restriction enzyme.The 1067T > A SNP was treated with restriction enzyme only after the second PCR.

Table 2 -
FUT3 gene haplotypic frequencies and their standard deviations (SD) identified using the Arlequin program.

Table 3 -
Association between the FUT3 gene SNP genotype frequencies and the Lewis (Le) erythrocyte phenotypes for a sample population (n = 185) of Brazilians of Japanese ancestry.The le 59 genotype represents the 59G allele while le59, 508represents the association of the 59G and 508A alleles on the same chromosome and le59, 1067the association of the 59G and 1067A alleles on the same chromosome.

Table 4 -
Comparison of the genotype frequencies of the three FUT3 gene SNPs analyzed in the current study (n = 185) and in other populations described in the literature.The le 59 genotype represents the FUT3 59G allele while le59, 508represents the association of the FUT3 59G and FUT3 508A alleles on the same chromosome and le59, 1067the association of the FUT3 59G and FUT3 1067A alleles on the same chromosome.