Digestive Diseases and Sciences

, Volume 61, Issue 1, pp 107–116 | Cite as

Population, Epidemiological, and Functional Genetics of Gastric Cancer Candidate Genes in Peruvians with Predominant Amerindian Ancestry

  • Roxana Zamudio
  • Latife Pereira
  • Carolina D. Rocha
  • Douglas E. Berg
  • Thaís Muniz-Queiroz
  • Hanaisa P. Sant Anna
  • Lilia Cabrera
  • Juan M. Combe
  • Phabiola Herrera
  • Martha H. Jahuira
  • Felipe B. Leão
  • Fernanda Lyon
  • William A. Prado
  • Maíra R. Rodrigues
  • Fernanda Rodrigues-Soares
  • Meddly L. Santolalla
  • Camila Zolini
  • Aristóbolo M. Silva
  • Robert H. Gilman
  • Eduardo Tarazona-Santos
  • Fernanda S. G. Kehdy
Original Article

Abstract

Background

Gastric adenocarcinoma is associated with chronic infection by Helicobacter pylori and with the host inflammatory response triggered by it, with substantial inter-person variation in the immune response profile due to host genetic factors.

Aim

To investigate the diversity of the proinflammatory genes IL8, its receptors and PTGS2 in Amerindians; to test whether candidate SNPs in these genes are associated with gastric cancer in an admixed population with high Amerindian ancestry from Lima, Peru; and to assess whether an IL8RB promoter-derived haplotype affects gene expression.

Methods

We performed a Sanger-resequencing population survey, a candidate-gene association study (220 cases, 288 controls) and meta-analyses. We also performed an in vitro validation by a reporter gene assay of IL8RB promoter.

Results

The diversity of the promoter of studied genes in Native Americans is similar to Europeans. Although an association between candidate SNPs and gastric cancer was not found in Peruvians, trend in our data is consistent with meta-analyses results that suggest PTGS2-rs689466-A is associated with H. pylori-associated gastric cancer in East Asia. IL8RB promoter-derived haplotype (rs3890158-A/rs4674258-T), common in Peruvians, was up-regulated by TNF-α unlike the ancestral haplotype (rs3890158-G/rs4674258-C). Bioinformatics analysis suggests that this effect stemmed from creation of a binding site for the FOXO3 transcription factor by rs3890158G>A.

Conclusions

Our updated meta-analysis reinforces the role of PTGS2-rs689466-A in gastric cancer in Asians, although more studies that control for ancestry are necessary to clarify its role in Latin Americans. Finally, we suggest that IL8RB-rs3890158G>A is a cis-regulatory SNP.

Keywords

Amerindians Ancestry Association studies Gastric cancer Meta-analyses Proinflammatory genes 

Introduction

Gastric adenocarcinoma, the third most frequent cause of cancer-related deaths in Latin America [1], is associated with the high prevalence of chronic infection by Helicobacter pylori, a gastric pathogen that infects billions of people worldwide [2] and that is classified as a type I carcinogen [3]. Most infections are asymptomatic; however, host responses and environmental factors also affect infection outcomes. In particular, several immunological events associated with gastric inflammatory responses to infection are likely to provoke gastric carcinogenesis [4, 5]. This inflammatory response is mediated by pro-inflammatory cytokines such as interleukin-8 (IL8), its receptors IL8RA and IL8RB, IL-1β and TNF-α, and the pro-inflammatory enzyme COX-2 (encoded by PTGS2 gene) [6, 7, 8]. Recavarren-Arce et al. [9] and Correa [10] proposed a progression model for the development of intestinal-type gastric adenocarcinoma, based on histopathology observations, and consistent with subsequent analyses of host immune responses to H. pylori infection. Importantly, there is a substantial inter-person variation in intensity of immune responses, progression of gastric pathologies, and in symptoms and clinical manifestations of carcinogenesis. Part of this variability must be due to host genetic factors [11, 12], which are the focus of this article.

The incidence of gastric cancer is particularly high in the Andean region of South America, where individuals with high Native American ancestry predominate. Therefore, in a previous study mostly based on the same individuals as in this article [13], to infer whether susceptibility alleles specific to Andean populations may exist, we tested whether individual Native American ancestry is a risk factor for gastric cancer, controlling for the effect of nongenetic factors (i.e., socioeconomic, nutritional, and clinical). We showed that: (1) Native American ancestry, which is >70 % in the target population from Lima (Peru), is associated with gastric cancer, but socioeconomic and nutritional factors explain this association. (2) Low socioeconomic status and nutritional variables were associated with both Native American ancestry and gastric cancer. We concluded [13] that the high incidence of gastric cancer in Peru does not seem to be related to common susceptibility alleles, related to ancestry, common in this population. Instead, our results suggested a predominant role for ethnic-associated socioeconomic factors and disparities in access to health services. Complementing our study by Pereira et al. [13], here we further test the presence of susceptibility alleles for gastric cancer. The goals of the present study are: (1) to investigate in Native Americans from Peru the pattern of nucleotide diversity of the promoter regions of the gastric cancer candidate genes IL8, IL8RA, IL8RB, and PTGS2, selecting tag-SNPs and identifying SNPs/haplotypes characteristics of these populations. We focus on promoter regions of different genes because there are several examples of the role of cis-regulatory mutations located in these regions that confer susceptibility to complex phenotypes [14]. (2) To test the hypothesis that variants with the following characteristics are susceptibility alleles for gastric cancer: (a) to be common in admixed populations with high Native American ancestry and (b) to be mapped in the promoter region of IL8, IL8RA, IL8RB, and PTGS2, as well as in the candidate-gene and genome-wide association studies cancer hits MSMB and FGFR2. (3) To test whether an IL8RB promoter-derived haplotype, common in Native Americans, affects gene expression, in order to provide evidence that rs3890158G>A is a cis-regulatory SNP that could be involved in the expression of IL8RB, thus modifying the risk of gastric cancer.

Materials and Methods

Polymorphism Characterization in Native American Population and SNPs Selection

In the following paragraphs, we outline the Materials and Methods, while details are in the Supplementary Material. To study the pattern of nucleotide diversity in Native Americans, we performed bidirectional Sanger sequencing of parts of the genes IL8, IL8RA, IL8RB, and PTGS2 (Table S1) in 25 Quechuas from the Peruvian Central Andes [15, 16]. For IL8 and its receptors, we also sequenced 67 gastric cancer patients and 58 of their controls from Lima. For comparison of population genetic diversity, we used the re-sequencing data for the same genomic regions in 24 Africans and 23 Europeans from the SeattleSNPs database. We treated the sequences using the pipeline by Machado et al. [17]. The following population genetic analyses were performed using the method and software indicated in parentheses: Hardy–Weinberg equilibrium (http://code.google.com/p/glu-genetics/, GLU), genetic diversity and neutrality tests ([18], DNAsp), haplotypes inferences ([19], PHASE), and linkage disequilibrium ([20], Haploview).

SNP Genotyping

We capitalized the information on the nucleotide diversity of the promoter region of the studied genes to select SNPs to be genotyped in 660 individuals (285 cases and 375 controls) and in 198 healthy Native Americans. All these genotyped cases and controls were used to perform the meta-analyses (see below). For the association study, we selected a subset of 508 genotyped individuals (220 cases and 288 controls) that also have ancestry results for which we have socioeconomic information (see below). Besides the selected four SNPs in IL8, IL8RB, and PTGS2, we also genotyped two SNPs associated with prostate (MSMB-rs10993994) and breast cancer (FGFR2-rs1219648) from other cancer GWAS, because they were common SNPs, and considering their OR and allele frequency (See Supplementary Material), SNPs were genotyped using TaqMan assays (Life Technologies, CA, USA). The total number of samples included in the experiments varied from the analysis because some samples failed to amplify due to the low DNA quality. More details about the sample size are described in Figure S1.

Samples

We recruited individuals attending Gastroenterology Divisions and prescribed for an endoscopy in three hospitals in Lima (2006–2009). The recruited individuals were classified as cases or controls as described in Pereira et al. [13]. Each participant answered a questionnaire to record age, gender, civil status, place of birth and residence, and other socioeconomic, dietary, and clinical information that were organized in 43 variables (Table S4 from Pereira et al. [13]). To reduce the dimensionality of the set of 43 personal, socioeconomic, nutritional, and clinical nongenetic variables, we performed a multivariate factor analysis [21]. The same analysis was performed in our previous study [13], but we re-conducted the analysis by adding 160 new individuals, leading to a total of 701 subjects for this analysis. To estimate Native American, European, and African ancestry for each individual, we genotyped a validated set of 87–103 ancestry informative markers [22] for 508 of the 701 individuals recruited for this study. The tri-hybrid continental ancestry was inferred with the software ADMIXTURE v.1.2 [23], using as parental populations Africans and Europeans from the HapMap database and Native Americans from our research group [13].

Association Analysis of Candidate SNPs

To test the association between candidate SNPs and gastric cancer, we used the logistic regression analysis implemented in the SNPassoc package (version 1.9-2, [24]), assuming additive, dominant, and recessive models of inheritance. This method allows the determination of odds ratio (OR) for each SNP and its confidence interval, as well as permitting controlling for covariates.

Systematic Review and Meta-Analyses

We included all case–control studies published until September 2014 on the association of one polymorphism (rs4073) of IL8 and two polymorphisms (rs689465 and rs689466) of PTGS2 with gastric cancer risk. Eligible studies were identified by a systematic literature search on PubMed database. We assessed heterogeneity between studies by Cochran’s Q test [25] and I2 statistic [26]. The meta-analyses were performed under the random-effects model, and publication bias was tested by the Egger’s test and inverted funnel plots [27].

Cloning, Transfection and Reporter Gene Studies

All promoter haplotype sequences of IL8, IL8RA, IL8RB, and PTGS2 identified in our study were submitted to the MATCH™ software program from the TRANSFAC database [28] to identify binding sites of transcription factors in these promoters. We focused on performing functional studies on IL8RB promoter because in the bioinformatics analysis we identified binding site for an important transcription factor (FOXO3) linked to tumorigenesis. To verify whether the SNPs that defined the haplotypes influence transcriptional activity of IL8RB, we synthesized two DNA fragments: One corresponds to the ancestral haplotype (defined by allele rs3890158-G and rs4674258-C) and the other corresponds to the derived common haplotype (defined by allele rs3890158-A and rs4674258-T). Both fragments were inserted into BglII and MluI sites present in the multiple cloning site of the pGL3-basic luciferase reporter vector (Promega, Madison, WI, USA) using T4 DNA ligase (Invitrogen, USA) following manufacturer’s instructions. These plasmids were sequenced to confirm the cloned fragments. For reporter gene studies, human embryonic kidney 293 cells (HEK293—ATCC, Manassas, VA) cultured at 1 × 105 cells/well in 24-well plates were co-transfected with 400 ng of constructed luciferase reporter plasmids plus 100 ng of pRL-TK plasmid, an internal control. For the transfection, we used Lipofectamine® 2000 Transfection Reagent (Life Technologies, USA) in a 2:1 DNA/Lipofectamine ratio. After 20 h, cells were left either untreated or treated with 10 ng/ml of recombinant TNF-α (Sigma, St. Louis, MO, USA) for 8 h, and cell lysates were harvested for further measurements of luciferase activity using the dual luciferase assay system (Promega, Madison, WI, USA).

Ethics Statement

This investigation was approved by the Ethics Committee from: Asociación Benéfica PRISMA, Universidad Peruana Cayetano Heredia, Johns Hopkins University, Universidade Federal de Minas Gerais, Hospital Arzobispo Loayza, Hospital Dos de Mayo, and the Instituto Nacional de Enfermedades Neoplásicas. All participants in the study signed a free informed consent form.

Results and Discussion

Patterns of Nucleotide and Haplotype Diversity

The resequencing of promoters and 5′UTRs of IL8, IL8RA, IL8RB, and PTGS2 genes in Native American or case–control Peruvians confirmed the presence of SNPs identified in African and European populations. We also identified six SNPs exclusive in all studied Peruvian samples, 4 of them not present in the dbSNP database (Table S5). The pattern of intra-population diversity and results of evolutionary neutrality tests (assessed by the statistics Tajima D, D and F of Fu and Li) for the 5′UTRs and promoter regions of these genes are detailed in Table 1. However, these results should be interpreted with caution because the summary statistics may be subjected to large errors due to the small regions considered. Overall, the diversity of these loci in these South Amerindians is similar to that observed in the European sample from the SeattleSNPs database, without a reduced diversity frequently observed in Native American populations for most genes [15]. Also, the neutrality tests based on the allelic spectrum (i.e., the distribution of frequency classes of alleles) suggest that the observed proportions of common and rare alleles are compatible with mutation-drift equilibrium (i.e., the null hypothesis) and do not provide evidence of natural selection operating on these sites. Specifically, the PTGS2 promoter region shows an important differentiation in Native American with a high frequency of the haplotype PTGS2-6 (defined by the SNP rs689466-G).
Table 1

Intra-population diversity indexes and results of neutrality tests in the studied populations for IL8, IL8RA, IL8RB, and PTGS2 genes

 

African

n = 24b

European

n = 23b

Amerindian

n = 25b

Case

n = 67b

Control

n = 58b

Number of chromosomes

 IL8

48

46

50

124c

112c

 IL8RA

46c

46

50

134

116

 IL8RB

48

46

50

118c

112c

 PTGS2

48

46

48c

ND

ND

Segregating sites/singletons

 IL8

6/2

1/0

1/0

1/0

2/0

 IL8RA

2/1

3/2

1/0

2/1

3/1

 IL8RB

2/1

1/0

3/0

5/1

4/1

 PTGS2

11/3

5/1

6/3

ND

ND

Haplotype structure

Number of inferred haplotypes

 IL8

6

2

2

2

3

 IL8RA

3

4

2

3

4

 IL8RB

3

2

4

8

7

 PTGS2

11

5

6

ND

ND

Haplotype diversity ± SD

 IL8

0.69 ± 0.05

0.45 ± 0.05

0.33 ± 0.07

0.29 ± 0.04

0.37 ± 0.05

 IL8RA

0.43 ± 0.07

0.21 ± 0.08

0.35 ± 0.07

0.35 ± 0.04

0.45 ± 0.04

 IL8RB

0.51 ± 0.04

0.50 ± 0.02

0.54 ± 0.03

0.60 ± 0.02

0.56 ± 0.02

 PTGS2

0.79 ± 0.05

0.59 ± 0.07

0.68 ± 0.04

ND

ND

θ estimatorsa

π × 103

     

 IL8

1.69

0.77

0.56

0.50

0.67

 IL8RA

0.41

0.20

0.33

0.33

0.43

 IL8RB

0.69

0.64

1.39

1.52

1.39

 PTGS2

1.15

0.60

0.68

ND

ND

θW × 103

 IL8

2.30

0.39

0.38

0.32

0.64

 IL8RA

0.43

0.64

0.21

0.34

0.53

 IL8RB

0.58

0.29

0.86

1.20

0.97

 PTGS2

1.44

0.66

0.79

ND

ND

Tajima’s D

 IL8

−0.68684

1.33852

0.62150

0.53323

0.06820

 IL8RA

−0.06958

−1.45374

0.76397

−0.06997

−0.31392

 IL8RB

0.34973

1.65929

1.29078

0.55002

0.85431

 PTGS2

−0.58526

−0.23357

−0.35354

ND

ND

Fu and Li’s D

 IL8

−0.52827

0.55053

0.54316

0.47114

0.67670

 IL8RA

−0.86913

−1.68049

0.54316

−1.13815

−0.61912

 IL8RB

−0.88132

0.55053

0.89169

−0.06104

−0.28955

 PTGS2

−0.26661

0.15473

−1.38132

ND

ND

Fu and Li’s F

 IL8

−0.67749

0.89679

0.65430

0.57368

0.57141

 IL8RA

−0.73617

−1.87650

0.70173

−0.94353

−0.61348

 IL8RB

−0.60235

1.00286

1.17861

0.16634

0.09506

 PTGS2

−0.43810

0.04052

−1.23990

ND

ND

ND no data

aπ: Tajima [46], θW: Watterson [47]

bn represents the number of individuals included in the analysis

cThis number of chromosomes is lower than 2n because some samples failed to amplify due to the low DNA quality

We also note that the IL8RB SNP rs3890158 was absent both in Africans and in Europeans in the SeattleSNPs database (even though the position was reported as sequenced by Sanger technology, which is more reliable than Next-Generation-Sequencing platforms) and from HapMap, suggesting that it might be specific to Native Americans (Table S8). A later survey in the 1000 Genomes database revealed that this SNP was sequenced with a low coverage (5.48×), and in 1000 Genomes, the allele A is reported as abundant: 86, 59, 51, and 34 % in Africans, South Asians, Europeans, and East Asians, respectively. Thus, it is not clear whether this SNP is common only in Native Americans.

Based on the pattern of linkage disequilibrium (Fig. 1), we selected the following tag-SNPs for genotyping in our case–control association study: rs4073 (IL8), rs4674258 (IL8RB), rs689465, and rs689466 (PTGS2). The rationale for this selection was as follows: (1) For IL8RB, the rs3890158 was firstly selected because its A allele is very common in Native Americans, cases, and controls (Table S8). This polymorphism and rs4674258 are in strong LD in Native Americans (r2 = 0.92) and in cases and controls (r2 = 0.86). We selected rs4674258 as tag-SNP because it was suitable for genotyping by TaqMan assays. (2) rs4073 (IL8) was the only common SNP not in Hardy–Weinberg equilibrium in this gene in Native American (Table S5), and it presents highly differentiated frequencies among continents (84 % in African, 42 % in European, 41 % in East Asian, 40 % in South Asian, and 1–18 % in Native American [Table 2]). (3) Two PTGS2 SNPs (rs689465 and rs689466) were selected for genotyping. Both are associated with gastric cancer in the Chinese population [29, 30, 31]. Of these SNPs, rs689465 is also a tag-SNP of rs20417 (r2 = 0.85) in Native Americans and rs689466 has very different frequencies in diverse continental groups (Table 2). No SNP was selected in the IL8RA gene for genotyping because the affinity of IL8 for IL8RB is two to five times greater than the affinity for IL8RA, suggesting that IL8RB has more active role in the inflammatory process [7, 32].
Fig. 1

Pattern of linkage disequilibrium for the promoter regions of: aIL8RB in Native Americans, cases and controls (top); and bPTGS2 in Africans, Europeans, and Native Americans (bottom)

Table 2

Risk allele frequency of candidate SNPs in African, European, and Asian populations from 1000 Genomes database, Native Americans (Ashaninka, Shimaa, Aymara, and Quechua), and case–control samples from our research group

Gene-SNP

Risk allele

African

European

East Asian

South Asian

Ashaninkas

Shimaa

Aymara

Quechua

Case

Control

MSMB-rs10993994

T

0.64

0.39

0.47

0.62

0.38

0.58

0.28

0.17

0.28

0.29

FGFR2-rs1219648

C

0.44

0.43

0.38

0.38

0.50

0.78

0.41

0.23

0.32

0.35

IL8-rs4073

A

0.84

0.42

0.42

0.40

0.13

0.01

ND

0.18

0.22

0.25

IL8RB-rs4674258

T

0.58

0.46

0.32

0.48

0.49

0.24

ND

0.52

0.56

0.56

PTGS2-rs689465

G

0.16

0.13

0.05

0.16

0.07

0.18

0.19

0.31

0.18

0.17

PTGS2-rs689466a

A

0.92

0.81

0.52

0.87

0.43

0.59

0.53

0.74

0.62

0.61

ND no data

aNot in HWE in Shimaa, p = 0.03

Testing Association of Candidate SNPs with Gastric Cancer Controlling Covariates

Before testing the association between gastric cancer and the studied candidate SNPs, we used 87–103 AIMs to estimate ancestry of the studied individuals. We also collected forty-three variables that contain socioeconomic, dietary, and symptomatic information that was appropriately summarized in three variables obtained by factor analysis: socioeconomics (first factor), symptoms (second factor), and nutritional (third factor) (Table S12 and Figures S4–S8). Using an additional 160 individuals relative to Pereira et al. [13], we reproduced the analysis and confirmed their conclusions. Namely, that socioeconomic and nutritional variables account for the association between gastric cancer and Native American ancestry, and that lower socioeconomic status and nutritional variables are associated with both gastric cancer and Native American ancestry.

Genotypes and alleles frequencies of MSMB-rs10993994, FGFR2-rs1219648, IL8-rs4073, IL8RB-rs4674258, PTGS2-rs689465, and PTGS2-rs689466 in controls and gastric cancer patients are summarized in Table 3. Logistic regression analysis both with and without the covariates (more details see Table S13) indicated no significant association of these six candidate SNPs with gastric cancer in Peruvians with high Native American ancestry.
Table 3

Genotype and allele frequencies, odds ratio, confidence interval, and p value of the logistic regression analysis with and without the covariates sex, age, Native American ancestry, socioeconomic (first factor), symptoms (second factor), and nutritional (third factor) information, using additive, dominant, and recessive model

Gene-SNP

Genotype/allele

Cases n (%)

Controls n (%)

OR [95 % CI]; p valuea

OR [95 % CI]; p valueb

MSMB-rs10993994

 

n = 220

n = 288

  

CC

112 (50.9 %)

148 (51.4 %)

0.96 [0.73–1.25]; 0.7496c

1.00 [0.74–1.36]; 0.9932c

CT

91 (41.4 %)

111 (38.5 %)

1.02 [0.72–1.45]; 0.9146d

1.02 [0.69–1.51]; 0.9041d

TT

17 (7.7 %)

29 (10.1 %)

0.75 [0.40–1.40]; 0.3589e

0.92 [0.45–1.87]; 0.8117e

T

125 (28.4 %)

169 (29.3 %)

  

FGFR2-rs1219648

 

n = 220

n = 288

  

TT

108 (49.1 %)

122 (42.4 %)

0.90 [0.70–1.16]; 0.4253c

0.98 [0.74–1.31]; 0.9029c

CT

82 (37.3 %)

132 (45.8 %)

0.76 [0.54–1.08]; 0.1311d

0.88 [0.60–1.30]; 0.5348d

CC

30 (13.6 %)

34 (11.8 %)

1.18 [0.70–2.00]; 0.5388e

1.23 [0.68–2.21]; 0.4916e

C

142 (32.3 %)

200 (34.7 %)

  

IL8-rs4073

 

n = 216

n = 280

  

TT

130 (60.2 %)

159 (56.8 %)

0.87 [0.64–1.16]; 0.3392c

0.85 [0.61–1.18]; 0.3285c

AT

76 (35.2 %)

103 (36.8 %)

0.87 [0.61–1.25]; 0.4462d

0.85 [0.56–1.27]; 0.4157d

AA

10 (4.6 %)

18 (6.4 %)

0.71 [0.32–1.56]; 0.3853e

0.70 [0.29–1.68]; 0.4204e

A

96 (22.2 %)

139 (24.8 %)

  

IL8RB-rs4674258

 

n = 218

n = 281

  

CC

49 (22.5 %)

48 (17.1 %)

1.02 [0.79–1.30]; 0.9069c

1.05 [0.79–1.39]; 0.7326c

TC

92 (42.2 %)

151 (53.7 %)

0.71 [0.46–1.11]; 0.1321d

0.70 [0.43–1.15]; 0.1632d

TT

77 (35.3 %)

82 (29.2 %)

1.33 [0.91–1.94]; 0.1449e

1.43 [0.94–2.18]; 0.0906e

T

246 (56.4 %)

315 (56.0 %)

  

PTGS2-rs689465

 

n = 220

n = 288

  

AA

147 (66.8 %)

198 (68.8 %)

1.07 [0.77–1.49]; 0.6927c

1.12 [0.78–1.62]; 0.5287c

AG

67 (30.5 %)

82 (28.5 %)

1.09 [0.75–1.59]; 0.6442d

1.15 [0.76–1.74]; 0.5171d

GG

6 (2.7 %)

8 (2.8 %)

0.98 [0.34–2.87]; 0.9725e

1.13 [0.35–3.67]; 0.8407e

G

79 (18.0 %)

98 (17.0 %)

  

PTGS2-rs689466

 

n = 220

n = 288

  

GG

32 (14.6 %)

43 (14.9 %)

1.05 [0.81–1.35]; 0.7185c

1.17 [0.88–1.56]; 0.2736c

AG

103 (46.8 %)

139 (48.3 %)

1.03 [0.63–1.69]; 0.9035d

1.18 [0.68–2.02]; 0.5574d

AA

85 (38.6 %)

106 (36.8 %)

1.08 [0.75–1.55]; 0.6731e

1.26 [0.84–1.88]; 0.2622e

A

273 (62.0 %)

351 (60.9 %)

  

aLogistic regression analysis without covariates

bLogistic regression analysis with covariates (age, sex, Native American ancestry, and scores from the three factors)

cAdditive inheritance model

dDominant inheritance model

eRecessive inheritance model

Random-Effect Meta-Analyses

For IL8-rs4073 (23 studies with 5166 cases and 7903 controls, Figure S10), we found evidence of association of the AA genotype with gastric cancer under a dominant model only in Asia, although publication bias (test for funnel plot asymmetry, p = 0.02) suggests caution about this inference. Results in Europeans and Latin Americans are conflicting. Despite the diversity in the proportions of continental admixture among Latin American populations [33], the high differences in allele frequencies between continents for rs4073, and the potential confounder effect of ancestry, none of the previous studies conducted in Latin Americans controlled for ancestry, which limits the interpretation of these studies and the meta-analyses.

PTGS2-rs689465 has similar frequencies across continents, and therefore, in this case, continental ancestry in admixed populations is not a potential confounder. Our meta-analyses (2 studies for PTGS2 rs689465 with 608 cases and 1319 controls, Figure S12-A) revealed that, in addition to this study, there is only one additional association study on the Chinese population [31]. Altogether, meta-analysis does not provide a convincing evidence of association of rs689455 with gastric cancer.

In contrast to PTGS2-rs689465, PTGS2-rs689466 is highly differentiated between continents (Table 2). Of the five studies included in our meta-analysis (1252 cases and 2579 controls), four were carried out in Asians. Thus, our study is the first in non-Asians. Our meta-analysis showed association considering both a dominant model (OR 1.54, 95 % CI 1.24–1.93) and an additive model (OR 1.78, 95 % CI 1.33–2.37), and the trend of the result of our association study, although not significant, was consistent with the results of the meta-analysis. Remarkably, we do not have evidence of publication bias for this SNP (Figure S12-B and S12-C). This result is consistent with the association between PTGS2 expression and rs689466-AA genotype, as well as with the observation that the A allele creates a c-MYB transcription factor binding site [29]. Thus, in our meta-analyses, PTGS2-rs689466 seems to be the SNP for which its predicted effect on gastric cancer is more convincing.

Effect of Promoter Haplotype on IL8RB Transcriptional Activation

Bioinformatics analyses indicated that ch2:218990202, rs3890158, rs3890157, rs17844672, but not rs4674258, affected IL8RB transcription factor binding sites (Table S18). Specifically, a FOXO3 transcription factor binding site is created by rs3890158-A but not by the G allele. FOXO3, whose activity in gastric cells is affected by H. pylori infection [34], is also normally expressed in brain, blood, colon, kidney, ovary, skeletal muscle, T helper cells, heart, lung, liver, pancreas, and placenta [35, 36, 37]. Since FOXO3 has been linked to tumorigenesis and the progression of certain cancers [38], we consider the rs3890158-G>A mutation, which creates a binding site for FOXO3, as a candidate SNP for gastric cancer to follow up with a functional study.

We constructed two luciferase reporter plasmids, spanning 510 bp of the IL8RB promoter, to evaluate the effect of the derived haplotype on IL8RB transcriptional activity. The nucleotide sequences of these constructs were identical except for the two SNP sites (i.e., rs3890158G>A and rs4674258C>T). They were used in the transfection of HEK293 cells, which were further stimulated with TNF-α, a cell-signaling protein important in normal physiology, inflammation, autoimmune diseases, and cancer-related inflammation [39, 40] that also affects IL8RB expression [41] and stimulates production of further inflammatory cytokines and chemokines [42]. TNF-α caused a 4.6-fold increase in reporter gene expression (p = 3.283e−5) only in cells containing the derived haplotype (rs3890158-A/rs4674258-T) (Fig. 2). This suggests a functional effect of the rs3890158 SNP, likely mediated by the FOXO3 transcription factor. Snoeks et al. [43] have shown that TNF-α induces the translocation of nuclear FOXO3 into the cytosol where it undergoes proteasomal degradation in human intestinal HT-29 cells. The interpretation of our results is consistent with the higher expression of IL8RB in gastric cancer tissues than in adjacent noncancerous tissues, given that IL8RB positively regulates the migration and invasion abilities of gastric cancer cells, which are characteristics of the malignant tumor [44].
Fig. 2

Luciferase reporter gene expression assays in HEK293A cells. IL8RB promoter activity is up-regulated by the derived haplotype (defined by allele rs3890158-A and rs4674258-T) when stimulated with TNF-α (p = 3.283e−5). For ancestral haplotype (defined by allele rs3890158-G and rs4674258-C), the up-regulation of IL8RB promoter activity was not observed. Each bar represents the mean relative activity ± SD of triplicate wells. The data shown are representative of two independent experiments

In conclusion, a combination of population genetics, genetic epidemiology, and functional approaches reveals that: (1) the diversity of the promoter regions of IL8, IL8RA, IL8RB, and PTGS2 in Native American is comparable with that of Europeans and seems to have evolved under mutation-drift equilibrium, without an evident effect of natural selection. (2) A population genetics approach contributes to the design of genetic epidemiology studies, particularly in populations such as admixed Latin Americans with high Native American ancestry that are under-represented in human genome initiatives. For instance, this approach revealed that the frequencies of IL8-4073 and PTGS2-rs689466 alleles in Native Americans are very different than in Europeans and Africans. Therefore, in Latin American admixed populations, when performing association studies with phenotypes such as gastric cancer, continental ancestry has to be controlled for. (3) Although controlling for ancestry is critical for variants that have different allele frequencies in Native Americans, Europeans, and Africans, ancestry (estimated with enough accuracy) is frequently ignored in candidate-gene association studies in Latin American populations [45]. (4) Although we did not find association between the tested SNPs and gastric cancer in the population from Lima with high Native American ancestry, meta-analyses suggest that IL8-rs4073-A and PTGS2-rs689466-A are susceptibility variants for gastric cancer in Asia. In the latter case, the trend of our result is consistent with the results of the meta-analysis, and our results are useful for future meta-analyses. (5) We provide functional evidence in vitro that stimulation by the pro-inflammatory cytokine TNF-α up-regulates the derived and common haplotype of the IL8RB promoter region (rs3890158-A/rs4674258-T), but not the ancestral haplotype (rs3890158-G/rs4674258-C). We used a bioinformatics approach to hypothesize that this effect is mediated by the creation of a binding site for the transcription factor FOXO3 by the mutation rs3890158G>A.

Notes

Acknowledgments

We thank Gifone Rocha, Denise Carmona, Carolina Gomes, Gilderlanio S Araujo, Giordano Soares Souza, Moara Machado, Mateus H Gouveia for discussions on different parts of the project and for technical help.

Funding

Fogarty International Center and National Cancer Institute (5R01TW007894) funded this study. The study and its participants also received funding and fellowships from the following Brazilian agencies: Brazilian National Research Council (CNPq), Ministry of Education (CAPES), Ministry of Health (PNPD-Saúde Program), and the Minas Gerais State Research Agency (FAPEMIG). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Compliance with ethical standards

Conflict of interest

The authors declare that they have no conflict of interest.

Supplementary material

10620_2015_3859_MOESM1_ESM.doc (3.2 mb)
Supplementary material 1 (DOC 3299 kb)

References

  1. 1.
    Ferlay J, Soerjomataram II, Dikshit R, et al. Cancer incidence and mortality worldwide: sources, methods and major patterns in GLOBOCAN 2012. Int J Cancer. 2014;136:E359–E386.CrossRefPubMedGoogle Scholar
  2. 2.
    Porras C, Nodora J, Sexton R, et al. Epidemiology of Helicobacter pylori infection in six Latin American countries (SWOG Trial S0701). Cancer Causes Control. 2013;24:209–215.CrossRefPubMedPubMedCentralGoogle Scholar
  3. 3.
    IARC. Monographs on the evaluation of carcinogenic risks to humans volume 61 schistosomes, liver flukes and Helicobacter pylori. World Heal Organ Int Agency Res Cancer. 1994;61:177.Google Scholar
  4. 4.
    Wang F, Meng W, Wang B, Qiao L. Helicobacter pylori-induced gastric inflammation and gastric cancer. Cancer Lett. 2014;345:196–202.CrossRefPubMedGoogle Scholar
  5. 5.
    Suzuki M, Mimuro H, Kiga K, et al. Helicobacter pylori CagA phosphorylation-independent function in epithelial proliferation and inflammation. Cell Host Microbe. 2009;5:23–34.CrossRefPubMedGoogle Scholar
  6. 6.
    Lamb A, Chen L-F. Role of the Helicobacter pylori-induced inflammatory response in the development of gastric cancer. J Cell Biochem. 2013;114:491–497.CrossRefPubMedPubMedCentralGoogle Scholar
  7. 7.
    Kitadai Y, Haruma K, Mukaida N, et al. Regulation of disease-progression genes in human gastric carcinoma cells by interleukin 8. Clin Cancer Res. 2000;6:2735–2740.PubMedGoogle Scholar
  8. 8.
    Sugimoto M, Yamaoka Y, Furuta T. Influence of interleukin polymorphisms on development of gastric cancer and peptic ulcer. World J Gastroenterol. 2010;16:1188–1200.CrossRefPubMedPubMedCentralGoogle Scholar
  9. 9.
    Recavarren-Arce S, León-Barúa R, Cok J, et al. Helicobacter pylori and progressive gastric pathology that predisposes to gastric cancer. Scand J Gastroenterol Suppl. 1991;181:51–57.CrossRefPubMedGoogle Scholar
  10. 10.
    Correa P. Human gastric carcinogenesis: a multistep and multifactorial process—First American Cancer Society Award Lecture on Cancer Epidemiology and Prevention. Cancer Res. 1992;52:6735–6740.PubMedGoogle Scholar
  11. 11.
    Shanks A, El-omar EM. Helicobacter pylori infection, host genetics and gastric cancer. J Dig Dis. 2009;10:157–164.CrossRefPubMedGoogle Scholar
  12. 12.
    Gehmert S, Velapatiño B, Herrera P, et al. Interleukin-1 beta single-nucleotide polymorphism’s C allele is associated with elevated risk of gastric cancer in Helicobacter pylori-infected Peruvians. Am J Trop Med Hyg. 2009;81:804–810.CrossRefPubMedPubMedCentralGoogle Scholar
  13. 13.
    Pereira L, Zamudio R, Soares-Souza G, et al. Socioeconomic and nutritional factors account for the association of gastric cancer with Amerindian ancestry in a Latin American admixed population. PLoS One. 2012;7:e41200.CrossRefPubMedPubMedCentralGoogle Scholar
  14. 14.
    Pastinen T. Genome-wide allele-specific analysis: insights into regulatory variation. Nat Rev Genet. 2010;11:533–538.CrossRefPubMedGoogle Scholar
  15. 15.
    Scliar MO, Soares-Souza GB, Chevitarese J, et al. The population genetics of Quechuas, the largest native South American group: autosomal sequences, SNPs, and microsatellites evidence high level of diversity. Am J Phys Anthropol. 2012;147:443–451.CrossRefPubMedGoogle Scholar
  16. 16.
    Tarazona-Santos E, Carvalho-Silva DR, Pettener D, et al. Genetic differentiation in South Amerindians is related to environmental and cultural diversity: evidence from the Y chromosome. Am J Hum Genet. 2001;68:1485–1496.CrossRefPubMedPubMedCentralGoogle Scholar
  17. 17.
    Machado M, Magalhães WC, Sene A, et al. Phred-Phrap package to analyses tools: a pipeline to facilitate population genetics re-sequencing studies. Investig Genet. 2011;2:1–7.CrossRefGoogle Scholar
  18. 18.
    Rozas J, Sánchez-DelBarrio JC, Messeguer X, Rozas R. DnaSP, DNA polymorphism analyses by the coalescent and other methods. Bioinformatics. 2003;19:2496–2497.CrossRefPubMedGoogle Scholar
  19. 19.
    Stephens M, Donnelly P. A comparison of bayesian methods for haplotype reconstruction from population genotype data. Am J Hum Genet. 2003;73:1162–1169.CrossRefPubMedPubMedCentralGoogle Scholar
  20. 20.
    Barrett JC, Fry B, Maller J, Daly MJ. Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics. 2005;21:263–265.CrossRefPubMedGoogle Scholar
  21. 21.
    Yong AG, Pearce S. A beginner’s guide to factor analysis : focusing on exploratory factor analysis. Tutor Quant Methods Psychol. 2013;9:79–94.Google Scholar
  22. 22.
    Yaeger R, Avila-Bront A, Abdul K, et al. Comparing genetic ancestry and self-described race in african americans born in the United States and in Africa. Cancer Epidemiol Biomark Prev. 2008;17:1329–1338.CrossRefGoogle Scholar
  23. 23.
    Alexander DH, Novembre J, Lange K. Fast model-based estimation of ancestry in unrelated individuals. Genome Res. 2009;19:1655–1664.CrossRefPubMedPubMedCentralGoogle Scholar
  24. 24.
    González JR, Armengol L, Solé X, et al. SNPassoc: an R package to perform whole genome association studies. Bioinformatics. 2007;23:644–645.PubMedGoogle Scholar
  25. 25.
    Cochran WG. The combination of estimates from different experiments. Int Biom Soc. 1954;10:101–129.Google Scholar
  26. 26.
    Higgins JPT, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med. 2002;21:1539–1558.CrossRefPubMedGoogle Scholar
  27. 27.
    Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ. 1997;315:629–634.CrossRefPubMedPubMedCentralGoogle Scholar
  28. 28.
    Matys V. TRANSFAC: transcriptional regulation, from patterns to profiles. Nucleic Acids Res. 2003;31:374–378.CrossRefPubMedPubMedCentralGoogle Scholar
  29. 29.
    Liu F, Pan K, Zhang X, et al. Genetic variants in cyclooxygenase-2: expression and risk of gastric cancer and its precursors in a Chinese population. Gastroenterology. 2006;130:1975–1984.CrossRefPubMedGoogle Scholar
  30. 30.
    Li Y, Dai L, Zhang J, et al. Cyclooxygenase-2 polymorphisms and the risk of gastric cancer in various degrees of relationship in the Chinese Han population. Oncol Lett. 2012;3:107–112.PubMedPubMedCentralGoogle Scholar
  31. 31.
    Zhang X, Zhong R, Zhang Z, et al. Interaction of cyclooxygenase-2 promoter polymorphisms with Helicobacter pylori infection and risk of gastric cancer. Mol Carcinog. 2011;50:876–883.CrossRefPubMedGoogle Scholar
  32. 32.
    Chuntharapai A, Kim KJ. Regulation of the expression of IL-8 receptor A/B by IL-8: possible functions of each receptor. J Immunol. 1995;155:2587–2594.PubMedGoogle Scholar
  33. 33.
    Galanter JM, Fernandez-Lopez JC, Gignoux CR, et al. Development of a panel of genome-wide ancestry informative markers to study admixture throughout the Americas. PLoS Genet. 2012;8:e1002554.CrossRefPubMedPubMedCentralGoogle Scholar
  34. 34.
    Tabassam FH, Graham DY, Yamaoka Y. Helicobacter pylori-associated regulation of forkhead transcription factors FoxO1/3a in human gastric cells. Helicobacter. 2012;17:193–202.CrossRefPubMedPubMedCentralGoogle Scholar
  35. 35.
    DaSilva L, Kirken RA, Taub DD, et al. Molecular cloning of FKHRL1P2, a member of the developmentally regulated fork head domain transcription factor family. Gene. 1998;221:135–142.CrossRefPubMedGoogle Scholar
  36. 36.
    Anderson MJ, Viars CS, Czekay S, Cavenee WK, Arden KC. Cloning and characterization of three human forkhead genes that comprise an FKHR-like gene subfamily. Genomics. 1998;47:187–199.CrossRefPubMedGoogle Scholar
  37. 37.
    Hillion J, Le Coniat M, Jonveaux P, Berger R, Bernard OA. AF6q21, a novel partner of the MLL gene in t(6;11)(q21;q23), defines a forkhead transcriptional factor subfamily. Blood. 1997;90:3714–3719.PubMedGoogle Scholar
  38. 38.
    Myatt SS. Lam EW-F. The emerging roles of forkhead box (Fox) proteins in cancer. Nat Rev Cancer. 2007;7:847–859.CrossRefPubMedGoogle Scholar
  39. 39.
    Balkwill F. Tumour necrosis factor and cancer. Nat Rev Cancer. 2009;9:361–371.CrossRefPubMedGoogle Scholar
  40. 40.
    Waters JP, Pober JS, Bradley JR. Tumour necrosis factor and cancer. J Pathol. 2013;230:241–248.CrossRefPubMedGoogle Scholar
  41. 41.
    Bhattacharya C, Samanta S, Gupta S, Samanta AK. A Ca2+-dependent autoregulation of lipopolysaccharide-induced IL-8 receptor expression in human polymorphonuclear neutrophils. J Immunol. 1997;158:1293–1301.PubMedGoogle Scholar
  42. 42.
    Chu W-M. Tumor necrosis factor. Cancer Lett. 2013;328:222–225.CrossRefPubMedPubMedCentralGoogle Scholar
  43. 43.
    Snoeks L, Weber CR, Wasland K, et al. Tumor suppressor FOXO3 participates in the regulation of intestinal inflammation. Lab Invest. 2009;89:1053–1062.CrossRefPubMedPubMedCentralGoogle Scholar
  44. 44.
    Cheng W-L, Wang C-S, Huang Y-H, Tsai M-M, Liang Y, Lin K-H. Overexpression of CXCL1 and its receptor CXCR2 promote tumor invasion in gastric cancer. Ann Oncol. 2011;22:2267–2276.CrossRefPubMedGoogle Scholar
  45. 45.
    De Oliveira JG, Rossi AFT, Nizato DM, Miyasaki K, Silva AE. Profiles of gene polymorphisms in cytokines and Toll-like receptors with higher risk for gastric cancer. Dig Dis Sci. 2013;58:978–988.CrossRefPubMedGoogle Scholar
  46. 46.
    Tajima F. DNA polymorphism in a subdivided population: the expected number of segregating sites in the two-subpopulation model. Genetics. 1989;123:229–240.PubMedPubMedCentralGoogle Scholar
  47. 47.
    Watterson GA. On the number of segregating sites in genetical models without recombination. Theor Popul Biol. 1975;7:256–276.CrossRefPubMedGoogle Scholar

Copyright information

© Springer Science+Business Media New York 2015

Authors and Affiliations

  • Roxana Zamudio
    • 1
    • 2
  • Latife Pereira
    • 1
  • Carolina D. Rocha
    • 4
  • Douglas E. Berg
    • 5
  • Thaís Muniz-Queiroz
    • 1
  • Hanaisa P. Sant Anna
    • 1
  • Lilia Cabrera
    • 2
    • 3
  • Juan M. Combe
    • 6
  • Phabiola Herrera
    • 2
    • 3
  • Martha H. Jahuira
    • 2
    • 3
  • Felipe B. Leão
    • 4
  • Fernanda Lyon
    • 1
  • William A. Prado
    • 7
  • Maíra R. Rodrigues
    • 1
  • Fernanda Rodrigues-Soares
    • 1
  • Meddly L. Santolalla
    • 1
  • Camila Zolini
    • 1
  • Aristóbolo M. Silva
    • 4
  • Robert H. Gilman
    • 2
    • 3
    • 8
  • Eduardo Tarazona-Santos
    • 1
  • Fernanda S. G. Kehdy
    • 1
    • 9
  1. 1.Department of General Biology, Institute of Biological SciencesFederal University of Minas GeraisBelo HorizonteBrazil
  2. 2.Laboratorios de Investigación y Desarrollo, Facultad de Ciencias y FilosofíaUniversidad Peruana Cayetano HerediaLimaPeru
  3. 3.Asociación Benéfica PRISMALimaPeru
  4. 4.Department of Morphology, Institute of Biological SciencesFederal University of Minas GeraisBelo HorizonteBrazil
  5. 5.Department of MedicineUniversity California San DiegoLa JollaUSA
  6. 6.Departamento de GastroenterologiaInstituto Nacional de Enfermedades NeoplásicasLimaPeru
  7. 7.Servicio de GastroenterologiaHospital Dos de MayoLimaPeru
  8. 8.Department of International Health, Bloomberg School of Public HealthJohns Hopkins UniversityBaltimoreUSA
  9. 9.Laboratório de Hanseníase, Instituto Oswaldo CruzFundação Oswaldo Cruz, FIOCRUZ-RJRio de JaneiroBrazil

Personalised recommendations