Polymeric immunoglobulin receptor polymorphisms and risk of nasopharyngeal cancer

Background Epstein-Barr virus (EBV) associated nasopharyngeal cancer (NPC) is an important squamous cell cancer endemic in Southeast Asia and the Far East and can be considered a multifactorial genetic disease. This research explores potential associations between nasopharyngeal epithelial EBV receptor and NPC susceptibility. To prove the hypothesis, we evaluated two candidate genes, complement receptor 2 (CR2) and polymeric immunoglobulin receptor (PIGR) by using 4 SNPs, CR2IVS2-848C→T, PIGRIVS3-156G→T, PIGR1093G→A and PIGR1739C→T, to genotype 175 cases and 317 controls, divided into Thai, Chinese and Thai-Chinese based on their respective ethnic origins. Results The results obtained indicated that PIGR is an NPC susceptibility gene. The risk association pertaining to each ethnic group was detected for homozygous PIGR1739C with a significant ethnic group adjusted OR (95%CI) of 2.71(1.72–4.23) and p < 0.00001. Haplotype of the two missense PIGR SNPs, 1093G→A and 1739C→T, and sequence analyses have confirmed the role of the nucleotide PIGR1739 and excluded possibility of an additional significant nonsynonymous NPC susceptibility SNP. Conclusions We present genetic evidence leading to hypothesize a possibility of PIGR to function as the EBV nasopharyngeal epithelium receptor via IgA-EBV complex transcytosis failure. The PIGR1739C→T is a missense mutation changing alanine to valine near endoproteolytic cleavage site. This variant could alter the efficiency of PIGR to release IgA-EBV complex and consequently increase the susceptibility of populations in endemic areas to develop NPC.

frequencies in Southern China, Southeast Asia, and among the Greenland Inuit. The highest aged-adjusted incidence has been reported in South China to be 30-50/ 100,000. Intermediate incidence has been noted in South East Asia with a ratio of 3/100,000 among Thais and 10/ 100,000 among Thais of Chinese extraction, respectively [2]. Regarding environmental carcinogenesis, both viral and chemical carcinogens contribute to NPC with Epstein Barr virus (EBV) as the most important etiological factor. The single clonally derived viral genome can be found in all endemic NPC cells [3][4][5]. On the other hand, several reports indicate that consuming salty fish or preserved food and thus concentrating chemical carcinogens such as nitrosamine, can promote the development of NPC [6,7]. Interestingly, two NPC susceptibility genes, HLA and cytochrome P450 2E1 (CYP2E1), were discovered based on their hypothesized relationship with environmental factors contributing to NPC etiology. HLA was first studied in Singapore as this gene would be responsible for patients' immune response to cancer or EBV infected cells [8]. In particular HLA A2 and B46 conferred a high relative risk as to NPC development according to various reports in Asia [9]. In addition, NPC development was proposed to correlate with patients' history of consuming preserved food and hence CYP2E1, which metabolises nitrosamine, was studied revealing a risk association in Taiwan [10] and Thailand [11]. This research aims at exploring the genetic aspect of EBV entry into the nasopharyngeal epithelium (NE). This process not only requires specific host factors but also differentiates between the mechanisms of EBV carcinogenesis originating in epithelial and/or lymphoid cells. Whereas EBV-associated Burkitt's lymphoma is prevalent in Africa, populations demonstrating a higher risk for developing NPC do not show any significant increase in the lymphoma incidence [12].
The mechanism of EBV entry into the NE has not yet been conclusively elucidated but at least two receptors, complement receptor type 2 (CR2) and polymeric immunoglobulin receptor (PIGR), have been proposed. CR2 is an integral membrane glycoprotein to which EBV can bind and thus infect B-lymphocytes [13]. EBV can infect recombinant epithelial cells expressing CR2 especially when in contact with virus-producing lymphocytes [14,15]. In addition, there has been a study reporting CR2 expression in embryonic NE cell by RT-PCR [16]. As for PIGR, the protein can be discovered on NE, interestingly as a pneumoccocus receptor, [17] where it mediates endocytosis and transcytosis of IgA-EBV complexes to deliver EBV into the luminal surface [18]. The viral translocation process can fail and consequently cause EBV infection if the epithelium loses its polarity or has mutated PIGR [19,20]. In addition to NE, EBV was reported to enter NPC in vitro depending on the presence of viral specific IgA [21]. Interestingly, serology studies indicate indirectly that PIGR is involved in NPC development. High EBV-IgA titers can be detected in a significant portion of NPC patients and the test has significant predictive role in diagnosing the disease [22]. Based on specific endemic distribution, NPC susceptibility genes can be hypothesized to originate from ancestor alleles. Hence, this study has been designed candidate-gene approaches for studying complex genetic traits [23]. The specific aim has been to explore whether CR2 or PIGR might be NPC susceptibility genes by determining the risk association between their DNA polymorphisms and haplotypes among NPC patients and controls from several individuals of high-risk ethnic origin.

PIGR and CR2 SNPs NPC case-control study
The study explored the risk of association between CR2 or PIGR DNA polymorphisms and NPC phenotype. Both the 175 patients and 317 controls were separated into Thai, Chinese and Thai-Chinese, according to their grandparents' ethnic origin so that each group would have a higher probability of shared ancestors. The genotype of four SNPs, CR2IVS2-848C→T, PIGRIVS3-156G→T,PIGR1093G→A and PIGR1739C→T, were investigated. CR2IVS2-848C→T located near exon 1, 2 [24], and PIGRIVS3-156G→T in intron 3 [25], have previously been published as RFLP polymorphisms. PIGR1093G→A and PIGR1739C→T, on exon 5 and 7, respectively, were selected from a SNP database http://www.ncbi.nlm.nih.gov/SNP/. PIGR1093G→A displays an amino acid alteration from glycine to serine and PIGR1739C→T from alanine to valine, respectively. No missense SNP from CR2 has recently been reported by genbank. The amplicons of CR2IVS2-848C→T, PIGRIVS3-156G→T and PIGR1739C→T were genotyped by PCR-RFLP, and PIGR1093G→A by ARMS (fig 1).
Upon comparison between the frequency of these alleles in patients and controls of identical ethnic origin, Thai or Chinese, a significant risk association to PIGR1739C→T was detected but neither to PIGRIVS3-156G→T nor PIGR1093G→A (  (table 1). This data suggests that the role of CR2 as an NPC susceptibility gene need further evaluation while the role of PIGR is crucial.
frequency of each haplotype based on its ethnic group was estimated by the EH program and distribution was compared between groups. Among the controls, there were no relevant difference between the Thai and the Chinese, yet the p values for the cases and the controls of both subgroups and the total were below 0.05 and 0.005, respectively (table 3).
Comparison among haplotypes confirmed 1739C→T as NPC susceptible mutation and excluded any other particular PIGR allele (table 3). First, whereas 1739C→T did, no haplotype was associated with significant OR in all Thai, Chinese and Thai-Chinese populations. Second, there were haplotypes with 1739C as susceptible alleles such as GC in the Chinese and total and 1739T as protective alleles such as AT in the Thai, GT in the Chinese and both haplotypes in the total population. In addition, there was no significant haplotype with 1739C as a protective allele or 1739T associated with significant higher relative risk. Finally, the relationship between each haplo- Total is Thai, Chinese and Thai-Chinese cases and controls. CR2IVS2-848C→T and PIGRIVS3-156G→T allele+ and allele-are digested and not digested with the restriction enzyme, respectively. G and A are nucleotide at PIGR1093, and C and T are nucleotide at PIGR1739, respectively. OR (95%CI) = odd ratios and 95% confidence interval of allele CR2+, PIGRIVS3-156G, PIGR1093G, PIGR1739C when compared with the other alleles of the same SNPs. 1 p < 0.05, 2 p < 0.001, 3 p < 0.0001

Discussion
The SNPs and haplotypes case control association study implicated that the probability to develop NPC might altered by DNA variation of PIGR, with nucleotide 1739 playing this crucial part. PIGR has been assigned to 1q31 and comprises of 11 exons [18]. The 1093G→A and 1739C→T are located on exon 5 and 7, respectively. The physical distance between nucleotide PIGRIVS3-156, 1093 and 1739 were 25 and 2.5 kb, respectively. Thus the three SNPs were closely linked but they distinctly contributed to NPC development. Whereas PIGR1739C→T exhibited a remarkably significant relative risk of NPC development among all Thai and Chinese populations, association of PIGRIVS3-156G→T and 1093G→A was not shown. Our data have provided an example that conclusions drawn from association studies aimed at identifying a susceptibility gene without related functional SNP should be interpreted with extreme caution [23]. If 1739C→T had not been chosen for this study, the importance of PIGR as an NPC susceptibility gene would have been misconstrued.
The manner, in which this PIGR1739C→T alters cellular movement of the IgA complex and consequently increases the possibility of EBV associate NPC development, remains to be elucidated. The polymorphism is a missense mutation altering the amino acid alanine to valine. The codon is located on exon 7, which is adjacent to the endoproteolytic cleavage site of the PIGR extracellular domain [29,30]. As a result, homozygous 1739C of the PIGR of individuals from high-risk ethnic origin may alter efficiency to release IgA-EBV complex and consequently, their NEs would have higher possibility to be infected by EBV. Future functional study is needed to elucidate how the polymorphisms in relationship with EBV biology alter the NPC susceptibility.
It is important to note that the 1739C→T mutation can also be found in other lower risk ethnic groups, such as Caucasian [30]. Hence, this PIGR nucleotide, despite a remarkably association with NPC, cannot account for its unique endemic distribution. On the other hand, this gene may be critical only for members of high-risk populations, to develop NPC as a consequence of other endemic genetic and/or environmental risk factors. Additionally, it is also interesting to find out whether the SNP on PIGR is a risk-associated factor affecting the disease susceptibility in the Caucasian population where NPC is rare and not 100% EBV related.
PIGR plays a crucial role in mucosal immunity not only against EBV, but also pneumococcal infection by facilitating transporting polymeric immunoglobulin transport across the mucosal epithelium. Human PIGR can bind to a major pneumococcal adhesion and enhanced pneumococcal adherence and invasion [17]. In other words, two important human pathogens, EBV and pneumococcus, employ the same human antibody transport protein, PIGR, to cause two common human diseases at the same tissue type, NE. The manner in which these two organisms interact with host (DNA) variation in vivo and what are the consequences might be on a global evolutionary scale will be very interesting subject for further investigation. Ethnic group adjusted odd ratios were calculated from the number of cases and controls of each haplotype in the first column by comparing with haplotype listed in the upper row. Numbers before () are odd ratios and within () are 95% confidence interval. GC, AC, GT, and AT are 1093G-1739C, 1093A-1739C, 1093G-1739T, and 1093A-1739T haplotypes, respectively.

Conclusions
The aim of this research was to explore risk associations between polymorphisms of two hypothesized nasopharyngeal epithelial EBV receptors, CR2 and PIGR, and NPC susceptibility The results obtained indicated that PIGR is an NPC susceptibility gene. The risk association pertaining to each ethnic group was detected for homozygous PIGR1739C with a significant ethnic group adjusted OR (95%CI) of 2.71(1.72-4.23) and p < 0.00001. This suggested genetic evidence leading to hypothesize a possibility of PIGR as the epithelial receptor for EBV via IgA-EBV complex transcytosis failure. More importantly, the PIGR1739C→T is a missense mutation changing alanine to valine near endoproteolytic cleavage site. This variant could alter the efficiency of PIGR to release IgA-EBV complex and consequently increase the susceptibility of populations in endemic areas to develop NPC.

Sample collection
After having obtained the subjects' informed consent as to the purpose of the study blood samples were collected by venipuncture from NPC patients and controls. The patients and controls were interviewed and then separated into three groups, Thai, Chinese, and Thai-Chinese, respectively, based on the ethnic origins of their grandparents.

SNP's information
Four SNPs, CR2IVS2-848C→T, PIGRIVS3-156G→T,PIGR1093G→A and PIGR1739C→T, were chosen. CR2IVS2-848C→T located in intron 2 [24] and PIGRIVS3-156G→T in intron3, [25] had previously been published as RFLP polymorphisms, whereas PIGR1093 and PIGR1739 were selected from a SNPs database. http:/ /www.ncbi.nlm.nih.gov/SNP/ The PCR condition for CR2IVS2-848C→T comprised an initiation denaturation step at 94°C for 4 minutes, followed by 35 cycles of 95°C for 1 minute, 55°C for 1 minute, 72°C for 2 minutes and final extension step at 72°C for 7 minutes. PIGRIVS3-156G→T was amplified 40 cycles at an annealing temperature of 60°C for 1 minute, and extension at 72°C for 3 minutes. The conditions for PIGR1739C→T was 35 cycles at 60°C annealing temperature for 1 minute and extension at 72°C for 1 minute. Subsequently, RFLP analysis was performed on 20 microliters each of the respective PCR products by subjecting them to the following restriction enzymes (New England Biolabs) at a 10U concentration: TaqI for CR2IVS2-848C→T, PvuII for PIGRIVS3-156G→T, and HgaI for PIGR1739C→T with overnight incubation at 65°C for TaqI and at 37°C for both of PvuII and HgaI. The resulting products were further analyzed by 2% agarose gel electrophoresis.

ARMS of PIGR1093G→A
The Amplification Refractory Mutation System (ARMS) [26] was used to detect SNP PIGR1093G→A. The primers can be divided into 2 sets, A and B. Set A primers comprised of 1093Fa and 1093R (GCCCCACTGTGGTGAAG-GGGGTGGCAGGTG and ACTGGGCCTTAACCCACCCC), whereas 1093Fb (GCCCCACTGTGGTGAAGGGGGTGG CAGGTA) and 1093R were mixed in set B. For each sample, the PCR reaction was performed in duplicate, set A and B, in a total volume of 50 µl containing 100 ng of genomic DNA, 1.5 mM MgCl 2 , 5 µl of 10X buffer (20 µM Tris-HCl pH 8.4, 50 mM KCl), 4U Taq Gold DNA polymerase (Perkin Elmer) and 0.3 µM of primer set A or set B. PCR conditions in-cluded initial denaturation at 95°C for 10 minutes followed by 30 cycles of 95°C for 1 minute, 65°C for 1 minute, 72°C for 1 minutes and final extension for 72°C for 7 minutes in a Perkin-Elmer/DNA thermal Cycle 480. Each SNP was investigated by 2% agarose gel electrophoresis for complementarity of the PCR product with the 3' nucleotide of primers set A and/or set B.