Inherited variant in NFκB‐1 promoter is associated with increased risk of IBD in an Algerian population and modulates SOX9 binding

Abstract

Background

The link between inflammation and cancer development was intensively studied in the last decade. To date, few studies explored the association between inflammatory genes and colorectal cancer (CRC) development.

Aim

The present study aimed to evaluate the implication of three single nucleotide polymorphisms (SNPs), rs28362491 ins/del −94 ATTG in NFκB1, rs6920220 (G/A) in TNFAIP3, and rs419598 (C/T) in IL1RN, which play a role in inflammation regulation in CRC development.

Methods and results

A case‐control study was conducted on an Algerian cohort of 358 subjects (147 healthy people, 89 individuals affected by inflammatory bowel disease [IBD], and 122 CRC patients enrolled at the University Hospital Center Ben Badis of Constantine). SNPs genotyping was performed by allelic discrimination TaqMan assay. The rs28362491 ins/del heterozygous genotype in NFκB1 conferred an increased risk of IBD compared with ins/ins homozygous genotype, with an increase of twofold (OR = 2.34 [1.29‐4.21]; 95% CI, 1.29‐4.21, P value = 0.004). No significant association was detected for the other two variants. Dual‐Luciferase Reporter Assay System performed in LoVo cells showed a significantly higher activity of the construct with ins allele of rs28362491 compared with the one harboring the del allele. Computational analysis nominated SOX9 as putative transcription factor (TF) with higher probability to bind the NFκB1 promoter at the SNP site, and we demonstrated in the in vitro assay that its overexpression modulates NFκB1 promoter activity in allele‐specific manner.

Conclusion

We speculate that SOX9 may modulate the NFκB1 activity by binding its promoter at the SNP site in allelic specific manner.

1. INTRODUCTION

Inflammatory bowel disease (IBD) involves chronic inflammation of all or part of the digestive tract and includes ulcerative colitis (UC) and Crohn's disease (CD). The yearly European prevalence of UC and CD is 505/100 000 and 322/100 000 for CD, while in North America is 249/100 000 and 319/100 000, respectively.1 The pathogenesis of IBD is a combination of intestinal flora, environmental factors, and genetic susceptibility.2, 3 The concordance rate of IBD in dizygotic twins is less than 10% in contrast to 20% to 50% in monozygotic twins.4 The incidence of colorectal cancer (CRC) in patients with IBD is 60% higher than in the general population, supporting that IBD is associated with an increased risk of CRC.5 CRC can also be influenced by the microenvironment such as stromal cells, inflammatory cells, and cytokines release leading to tumor heterogeneity.6, 7 Increasing evidence shows that persistence of inflammatory microenvironment may initiate tumorigenesis of CRC. It has been reported that inflammatory and cancer cells release cytokines and chemokines to promote cancer development8, 9 and that inherited polymorphisms in inflammation‐related genes (including IL10, TNFα, and TLR4) could modify cancer susceptibility.10, 11, 12, 13

Nuclear factor‐kappa B (NFκB) is a pleiotropic transcription factor,14 key regulator of inflammation,15 and a controller of several genes implicated in tumor progression.16, 17 In mammals, the NFκB family consists of five members: NFκB1 (p105/p50), NFκB2 (p100/p52), RelA (p65), RelB, and c‐Rel, that can give rise to homodimeric or heterodimeric complexes.18 NFκB complexes are inactivated by NFκB inhibitor family (IkB), and while transiently activated in normal cells, they exhibit sustained activation in cancer cells. Subsequently, to its persistent activation, NFκB potentiates genes, which are implicated in tumor development.19

Furthermore, NFκB signaling is under control of tumor necrosis factor alpha‐induced protein 3 (TNFAIP3), which acts as a negative‐feedback regulator of NFκB activation in response to multiple stimuli, including tumor necrosis factor (TNF) and interleukin‐1 (IL‐1). The physiological importance of TNFAIP3 as an anti‐inflammatory protein is clearly demonstrated by the phenotype of TNFAIP3‐deficient mice, which develop severe multiorgan inflammation causing premature lethality.20

Interleukin‐1 receptor antagonist (IL1RN) regulates the biological activity of the proinflammatory cytokines. IL1RN has anti‐inflammatory effects by binding to the IL‐1 receptor 1 (IL1R1) and specifically inhibiting IL‐1 signaling. Cytokine gene polymorphisms affect transcript and protein levels leading thereby to inflammatory abnormalities that increase susceptibility to multiple cancers, including CRC.21, 22

On the basis of this evidence, we aimed at specifically investigating the role of three SNPs in genes involved in inflammation, namely, rs28362491 (−94 ATTG) in NFκB1 promoter, rs6920220 (G/A) in TNFAIP3, and rs419598 (C/T) in IL1RN. We first tested them in a case‐control cohort with respect to CRC and/or IBD phenotypes and next investigated in vitro the functional role of the NFκB1 variant.

2. MATERIAL AND METHODS

2.1. Population description and genotyping

All study subjects were recruited at the University Hospital of Ben Badis between 2013 and 2015 and were subjected to clinical tests including radiography, endoscopy, C‐reactive protein (CRP), and erythrocyte sedimentation rate (ESR) evaluation. Subjects who were negative for all the tests were classified as “healthy subjects.” Subjects designated as IBD patients were those with inflammation of the colon but with noncancerous features, described in details in Table 1. For CRC patients, the diagnosis was confirmed by histological and endoscopy examinations of the colon or rectal biopsies collected during endoscopy or specimen from colonic surgery, following the normal clinical routine. A total of 147 healthy individuals (no IBDs), 89 individuals affected by IBD, and 122 CRC patients were collected.

Table 1.

Clinical data of patients study cohort

IBD patients (n = 89)	n (%)
Age, y
<50	71 (79.78)
>50	18 (20.22)
Gender
Male	41 (46.07)
Female	48 (53.93)
Family history of cancer
Yes	3 (3.37)
No	86 (96.63)
Smoking status
Ever	15 (16.85)
Never	74 (83.15)
Montreal classification of Crohn's disease and ulcerative colitis
Age of onset (A), y
A1: 16 or younger	5 (5.62)
A2: 17‐40	67 (75.28)
A3: older than 40	17 (19.10)
Location (L)
L1: ileal	39 (43.82)
L2: colonic	4 (4.49)
L3: ileocolonic	23 (25.84)
L4: upper	0 (0.00)
Disease extent (E)
E1: proctitis	0 (0.00)
E2: left‐sided UC; proximal extent of inflammation is distal to the rectosigmoid.	11 (12.36)
E3: extensive UC; involvement extends proximal to the splenic flexure.	7 (7.87)
Behavior (B)
B1: nonstricturing, nonpenetrating	40 (44.94)
B2: stricturing	24 (26.97)
B3: penetrating	24 (26.97)
CRC patients (n = 122)
Age
<50	40 (32.79)
>50	82 (67.21)
Gender
Male	68 (55.73)
Female	54 (44.26)
Family history of cancer
Yes	16 (13.11)
No	106 (86.89)
Smoking status
Ever	30 (24.59)
Never	92 (75.41)
Tumor site
Right colon	33 (27.05)
Left colon	58 (47.54)
Rectum	28 (22.95)
Indetermined	2 (1.64)
Tumor stages
I	11 (9.02)
II	33 (27.05)
III	30 (24.59)
IV	44 (36.07)
XX	3 (2.46)

Abbreviations: CRC, colorectal cancer; IBD, inflammatory bowel disease; UC, ulcerative colitis.

The volunteers gave their written informed consent to participate in the study and to allow the genetic analyses on biological samples. The interview of all study individuals was carried out via self‐administered questionnaires, at time of blood collection. For each subject, data collected included gender, age, body mass index (BMI), and smoking habits. The study was approved by the local Ethical Committee of Dr. Benbadis Hospital, University Center of Constantine, Algeria.

2.2. Sample preparation and genotyping

DNA was extracted from whole blood samples with the salting out method.23 Genotyping was performed using TaqMan SNP genotyping assays (Applied Biosystem, Life Technology) at the University of Trento, Italy. For TNFAIP3 G > A (rs6920220) and for IL1RN T > C (rs419598), the assays C_29431952_10 and C_8737990_10 were used, respectively. The genotype of NFκB −94 −/ATTG polymorphism was determined by using primers and probes as previously described.24 TaqMan Universal PCR Master Mix was used according to the manufacturer's instructions. The amplification step was performed using the CFX384 or CFX96 Detection Systems (BioRad).

2.3. Transcription factor DNA binding sites analysis

We considered 4920 unique transcription factor DNA binding sites (TFBSs) consensus motifs from JASPAR,25 HOMER,26 and HOCOMOCO27 (publicly available databases) and from TRANSFAC Professional database.28 We run the Transcription Factor Search System (TESS) on the variant locus against the compiled consensus motif collection to search for TFBSs.29 Both ancestral (−) and minor (ATTG) alleles for rs28362491 variant were tested considering 30‐bp flanking regions (length of collected consensus motifs ranges between 5 and 30). For each TFBS consensus motif, TESS provides a set of log‐likelihood‐ratio–based scores. Specifically, we used the La score, which represents the log‐odds ratio of the match, and the Lm score, which represents the maximum possible log‐odds ratio for a match of the given TFBS consensus motif. To select high confident results, we restricted the TFBS matches to La scores with false discovery rate (FDR) less than 5% when compared with a distribution of score matches computed from random regions of the genome. Next, we retained transcription factors with La/Lm score greater than 0.75 for at least one allele.

2.4. Cell lines

The colon cancer cell line LoVo was kindly donated by professor Alberto Inga (University of Trento, Italy) and maintained in Dulbecco's modified Eagle's medium (DMEM) (Gibco, Life Technologies) with 10% FBS, 100‐unit/mL penicillin, 100‐μg/mL streptomycin, and 2mM L‐glutamine. Cells were grown in humidified atmosphere at 37°C with 5% CO₂ in a cell culture incubator. The colon cancer cell line LoVo was tested and authenticated by PCR‐single‐locus‐technology (16 independent polymerase chain reaction [PCR] systems, Eurofins Genomics Europe Applied Genomics GmbH).

2.5. Plasmids and dual‐luciferase assay

The genomic sequence corresponding to the region of NFκB1 promoter (hg19 chr4: 103421416‐103,422,206) was cloned upstream the luciferase gene into the pGL4.14 vector (Promega, Madison, WI) using CloneEZ PCR Cloning Kit (GenScript) and KpnI and XhoI restriction enzymes. The plasmid harboring the allele “−” was created using the QuikChange II Site‐Directed Mutagenesis Kits (Agilent, Santa Clara, CA). The sequences of the mutagenic primers were gcctgcgttccccgaccattgggcccggcaggcg (forward) and cgcctgccgggcccaatggtcggggaacgcaggc (reverse).

These plasmids will be referred to as “pGL4.14_NFκB1ins” and “pGL4.14_NFκB1del.” The day before transfection, LoVo cells (8 × 10⁴ cells) were seeded in 24‐well plates. Cells were transfected with one of the reporter plasmids pGL4.14_NFκB1ins and pGL4.14_NFκB1del or with pGL4.14_Empty (400 ng), along with pRL‐SV40 control (50 ng) using FugeneHD reagent (Promega).

For KLF6 overexpression experiments, each reporter plasmid was cotransfected with either pCMV6‐XL5_KLF6 or pCMV6‐XL5_Empty (100 ng) using FugeneHD reagent (Promega). For SOX9 silencing, cells were transfected with siRNA against SOX9 (20nM) (FlexiTube GeneSolution for SOX9, Qiagen) and RNAmax transfection reagent (Invitrogen). All Stars Hs Cell Death siRNA and All Stars Negative Control siRNA (Qiagen) were used as positive and negative control, respectively. The day after the silencing, cells were transfected with reporter plasmids using FugeneHD (Promega). Forty‐eight hours after overexpression or 72 hours after silencing, cells were lysed using Passive Lysis Buffer 1X (Promega), and Firefly and Renilla luciferase activities were measured with Dual‐Luciferase Reporter Assay (Promega) using the Infinite M200 multiplate reader (Tecan).

2.6. Real‐time quantitative PCR

Total RNA was extracted from LoVo cells (seeded as 2.5‐3 × 10⁵ in a 6‐well plate) transfected with pCMV6‐XL5_KLF6 or empty vector (2.5 μg per well) or with siRNA against SOX9 (20nM) using the RNeasy kit (Qiagen) according to the manufacturer's instructions. A 2 μg of total RNA was converted into cDNA using the RevertAid First Strand cDNA Synthesis Kit (Thermo Fisher Scientific), and quantitative PCR reactions in real time were performed using KAPA SYBR kit (Kapa Biosystems, Resnova). ΔΔCt method was used to calculate the mRNA expression levels of each gene, and GAPDH (glyceraldehyde 3‐phosphate dehydrogenase) was used as reference gene. The endogenous expression of SOX9 and KLF6 was determined using the following primers: forward primer 5′‐GTAATCCGGGTGGTCCTTCT‐3′ and reverse primer 5′‐ACGCTGGGCAAGCTCT‐3′ for SOX9 and forward primer 5′‐ATTTGATGCATTCAGGGAGG‐3′ and reverse primer 5′‐GTAAGAAGCGGCATAGCACC‐3′ for KLF6.

2.7. Statistical analysis

To verify whether the genotypes were in Hardy‐Weinberg equilibrium, we used the chi‐square test (one degree of freedom) with the type‐I alpha error of 10e‐05 conventionally used in genome‐wide studies. Cases and controls were analyzed for association between the three genotypes and risk for IBD or CRC applying the dominant, recessive, or codominant models. The odds ratio (OR) and its 95% confidence intervals (CIs) were calculated. For the luciferase assay data, statistical significance was determined using Student's t test. P values of 0.05 were considered significant. Statistical analysis was performed using Prism 6.0 (GraphPad Software, La Jolla, CA).

3. RESULTS

3.1. Association between NFκB1, TNFAIP3, and IL1RN polymorphisms and IBD

The genotype and the allele frequencies of the studied polymorphisms (rs28362491, rs6920220, and rs419598) in the Algerian cohort are shown in Tables 2 and 3. Minor allele frequencies (MAFs) resulted in 0.372, 0.184, and 0.173, respectively, in line with the MAF observed in the general populations of the 1000 Genome Project data (0.419, 0.094, and 0.192, respectively), and each variant resulted in Hardy‐Weinberg equilibrium in the control set. Given the modest number of homozygotes, we applied a dominant model, where homozygotes individuals are pooled with heterozygotes. The association between the rs28362491 polymorphism of the NFκB1 gene and the risk of IBD was statistically significant (P = 0.012), with OR of 2.034 (95% CI, 1.16‐3.56), suggesting that the del allele confers an increased risk of developing IBD. No statistically significant associations were detected for the other tested SNPs with IBD and/or CRC.

Table 2.

Genotypes of study variants in the Algerian study cohort

Polymorphisms	CRC, n (%)	IBD, n (%)	CRC + IBD, n (%)	Control, n (%)
NFκB −94 ATTG ins/del (rs28362491)
ins/ins	54 (42.5)	26 (29.2)	80 (37)	68 (45.6)
ins/del	56 (44.1)	51 (57.3)	107 (49.5)	57 (38.3)
del/del	17 (13.4)	12 (13.5)	29 (13.4)	24 (16.6)
ins/del + del/del	73 (57.5)	63 (70.8)	136 (63)	79 (53)
TNFAIP3 (rs6920220)
GG	51 (66.2)	55 (65.5)	106 (65.8)	76 (66.7)
GA	25 (32.5)	27 (32.1)	52 (32.3)	34 (29.8)
AA	1 (1.3)	2 (2.4)	3 (1.9)	4 (3.5)
GA + AA	26 (57.5)	29 (34.5)	55 (34.2)	38 (33.3)
IL1RN (rs419598)
TT	59 (75.6)	63 (70.8)	122 (73.1)	81 (68.6)
CT	17 (21.8)	25 (28.1)	42 (25.1)	33 (28)
CC	2 (2.6)	1 (1.1)	3 (1.8)	4 (3.4)
CC + TT	19 (24.4)	26 (29.2)	45 (26.9)	37 (31.4)

Abbreviations: CRC, colorectal cancer; IBD, inflammatory bowel disease.

Table 3.

Statistical analysis of single nucleotide polymorphism (SNP)/phenotype association

	−/wt versus wt/wt		−/− + −/wt versus wt/wt
	OR (95% CI)	P Value	OR (95% CI)	P Value
NFκB1 −94 ATTG ins/del (rs28362491)
Control vs CRC	1.24 (0.74‐2.06)	0.41	1.13 (0.70‐1.82)	0.603
Control vs IBD	2.34 (1.29‐4.21)	0.004***	2.03 (1.16‐3.55)	0.012*
Control vs IBD + CRC	1.59 (1.01‐2.51)	0.044*	1.42 (0.93‐2.12)	0.09
IBD vs CRC	0.52 (0.28‐0.96)	0.037*	0.55 (0.44‐2.19)	0.054
TNFAIP3 (rs6920220)
Control vs CRC	1.07 (0.63‐1.81)	0.72	0.51 (0.11‐2.34)	0.38
Control vs IBD	1.05 (0.57‐1.95)	0.85	0.94 (0.52‐1.69)	0.84
Control vs IBD + CRC	1.09 (0.58‐2.05)	0.77	1.01 (0.55‐1.88)	.95
IBD vs CRC	1.03 (0.53‐2.01)	0.91	1.002 (0.52‐1.92)	0.99
IL1RN C > T (rs419598)
Control vs CRC	0.7 (0.36‐1.38)	0.31	0.7 (0.36‐1.34)	0.28
Control vs IBD	0.97 (0.52‐1.8)	0.93	0.90 (0.49‐1.64)	0.74
Control vs IBD + CRC	0.84 (0.49‐1.44)	0.53	0.8 (0.48‐1.35)	0.41
IBD vs CRC	0.72 (0.35‐1.47)	0.37	0.78 (0.39‐1.55)	0.48

Abbreviations: CRC, colorectal cancer; IBD, inflammatory bowel disease.

^* p < 0.05.

^** p < 0.01.

^*** p < 0.005.

3.2. In silico transcription factor selection for downstream experiments

In order to study the impact of the rs28362491 polymorphism on NFκB‐1, we first nominated TFs that potentially bind the region surrounding the SNP. In silico analysis identified 15 putative TFs with high binding score (La/Lm > 0.75) based on TESS analysis.29

Allele‐specific binding data were used to group the TFs in four sets: (a) NF652, HLTF, and SOX3 that bind both alleles; (b) SOX9 and CTNNB1 that preferentially bind the ATTGATTG “ins” allele; (c) KLF6, SOX10, and TCF12 that preferentially bind the ATTG “del” allele; and (d) LHX8, DUX, ONECUT1, SRY, GFI1, GATA6, and ARIDA3 that selectively bind the ATTGATTG “ins” allele (see Figure 1A).

Figure 1.

Putative transcription factors binding site (TFBS) data in the context of the selected variant (rs28362491) at NFκB1 promoter. A, Transcription factors (TFs) may bind DNA sequence with the insertion or the deletion with different affinities. B, Gene expression levels for the in silico selected TFs in colon tissues as per the cBioPortal database. Data are expressed as mean and SD

To prioritize TFs with marked expression in colon tissues, we first verified the expression levels in the TCGA colon cancer dataset using cBioPortal (http://www.cbioportal.org). Colorectal adenocarcinoma (TCGA, Provisional) data were available for 382 individuals (RNA Seq V2); only the diploid group was considered in the analysis (see Figure 1B). Next, we considered the TF expression levels in the study cell line (LoVo). Finally, we selected one TF that preferentially binds the ins allele (SOX9 with a score La/Lm = 0.8) and one that preferentially binds the del allele (KLF6 with a score La/Lm = 0.75) for further experimental characterization (see Figure 1A).

3.3. NFκB1 promoter activity upon modulation of SOX9 and KLF6 in colon cancer cells

A clear induction of reporter plasmid harboring the NFκB1 promoter was observed when compared with the empty vector. Interestingly, higher activity was detected with the pGL4.14_NFκB1_del compared with the pGL4.14_NFκB1_ins (P = .014, Figure 2).

Figure 2.

The selected transcription factor SOX9 modulates NFκB1 promoter activity at the single nucleotide polymorphism (SNP) site in LoVo cells. LoVo cells were transfected with siRNA against SOX‐9 or scrambled siRNA. Cells were then transfected with different pGL4.14 reporter constructs containing one of the two alleles or empty vector. Mean ± SD of three technical replicates of a representative experiment was plotted. Biological replicates are shown in Figure S1. *P < 0.05; ***P < 0.005

In order to better investigate this allele‐specific activity, we decided to modulate the expression of the two selected TFs, SOX9 and KLF6. On the basis of their expression in LoVo cells (high and low, respectively; see Figure 3A,B), we silenced the first and overexpressed the second one. Upon KLF6 overexpression, a decrease of luciferase activity was observed for the plasmid pGL4.14_NFκB1_del (see Figure S2). In the same condition, no change was observed for the plasmid pGL4.14_NFκB1_ins.

Figure 3.

Real‐time polymerase chain reaction (PCR) validation results. SOX9 silencing (left) ***P < 0.0001 and KLF6 overexpression (right) ***P < 0.0004 according to their endogenous level expression in LoVo cells. Mean ± SD of three technical replicates was plotted

Furthermore, while silencing SOX9, a significant decrease of luciferase activity was observed for the plasmid pGL4.14_NFκB1_del (P value = .017, Figure 2). According to the in silico analysis, SOX9 binds with higher affinity (0.8) the ins allele compared with the del allele (0.4). This result suggests that lower levels of SOX9 are detrimental for regulation in the presence of the allele with TF lower binding affinity.

4. DISCUSSION

Chronic colonic inflammation is crucial for increasing the risk of colon cancer development, but the molecular mechanisms involved in this process have not been fully established. The classical NFκB signaling pathway, however, has been shown to influence both the severity of inflammation30 and colonic carcinogenesis in animal models.31 Moreover, TNFAIP3 has been described as a key player in the NFκB signaling and proinflammatory gene expression.32

In this study, we investigated the possible association of three inherited variants (rs28362491 −94 ATTG ins/del in the NFκB1 promoter, rs6920220 in TNFAIP3, and rs419598 in IL1RN) with CRC and/or IBD susceptibility.

A total of 147 healthy IBDs, 89 individuals affected by IBD, and 122 CRC patients were collected and genotyped. Among the studied variants, the rs28362491 del allele in NFκB1 conferred an increased risk of IBD (dominant model, OR = 2.034; 95% CI, 1.16‐3.55, P value = 0.012).

The rs28362491 polymorphism, located in the promoter region of the NFκB1 gene, may affect NFκB1 promoter activity leading to a lower amount of NFκB1 transcript levels.33, 34 A previous study on Danish population showed that the carriers of NFκB‐94 deletion were associated with high risk for developing CRC compared with the homozygous carriers of the insertion allele.35 In contrast, a Chinese population–based study reported that the ins allele might have an increased risk for sporadic CRC.36 No association was found between the rs28362491 variant and IBD in neither a Moroccan nor a Danish population.37, 38 The discrepancy between these previous results and the current study could be influenced by multiple aspects, including specific ethnic background of populations that might influence the distributions of the polymorphism genotypes, and environmental factors, diet, and lifestyle. Further, other genes may contribute to the pathogenesis of IBD.

Moreover, the functional and mechanistic roles of this variant were poorly explored. In our study, we have performed a functional assay in LoVo cell lines, where we have reported a decreased activity of the NFκB1 promoter harboring the ins allele compared with the plasmid with the del allele. On the contrary, it was reported by Karban et al in 2004 that the ins allele is associated with increased promoter activity of NFκB1 in HeLa or HT‐29 cell lines.33 These controversial results can be related to the differential expression of TFs among cell lines. Since some TFs can regulate the NFκB1 gene in an opposite way, activating or inhibiting its expression, the amount of its expression can affect the level of reporter gene in an allelic specific manner.

In order to investigate the role of some TFs, we modulated the expression of the pGL4.14‐derived plasmids by silencing SOX9, selected for its in silico prediction to bind the NFκB1 promoter. We observed decreased promoter activity of NFκB1 for the plasmid harboring the ATTG allele suggesting that a decrease in the amount of SOX9 affects only the allele that binds this TF with lower affinity and that SOX9 may act as activator transcription factor of NFκB1 by binding its promoter at the SNP site.

Previous studies reported that SOX9 may play dual role in tumorigenesis; on one hand, SOX9 has been found to be overexpressed in CRC,39 and on the other hand, it has been reported to reduce tumorigenicity in HT‐29Cl.16E CRC cells.40 In addition, it has been shown that SOX9 is highly expressed in metastatic colon cancer cells (SW620) as compared with primary tumor cells (SW480), suggestive of a possible implication of SOX9 in the regulation of CRC cell plasticity.41 This finding was supported by a recent publication of Prevostel and Blache in 2017, showing that SOX9 changes its behavior in dependent‐level manner in CRC, acting as tumor suppressor or oncogene, and this could be relatively linked to the tumor stage (primary or metastatic).42, 43 In a recent study, it was demonstrated that high NF‐κB activity in bowel reflects high inflammatory burden.44 This is in accordance with our results that show (a) a higher activity of del allele in the reporter assay and (b) the association between the del allele and IBD.

It has been suggested that the alternative allele of the rs6920220 polymorphism is associated with decreased gene expression of TNFAIP3 gene (an inhibitor of the NFκB signaling pathway) in nonneoplastic intestinal tissue as compared with ancestral allele “G.”45 However, no significant association between this polymorphism and CRC risk has been reported in our study.

In addition, we have observed a lack of association between IL1RN polymorphism with IBD or CRC. The only study that investigated the association between IL1RN +2018 T > C (rs419598) polymorphism and the risk of CRC reported that the homozygous allele CC of the SNP rs419598 was associated with increased risk of sporadic CRC in a Romanian cohort.46

Collectively, this study is the first showing that SOX9 may modulate the NFκB1 activity by binding its promoter at the SNP site in allelic specific manner and that the del allele in NFκB1 conferred an increased risk of IBD. Lack of association for the two variants in TNFAIP3 and IL1RN could be due to the small sample size explored in our study. A larger study is recommended to better characterize the role of these SNPs in Algerian population.

CONFLICT OF INTEREST

No conflict of interest has been declared.

AUTHORS' CONTRIBUTIONS

All authors had full access to the study data and take responsibility for the integrity of the data and the accuracy of the data analysis. Conceptualization, I.H., S.G., A.R., D.N., T.H., F.D.; Methodology, I.H., S.G., A.R.; Investigation, I.H., S.G., A.R., F.D.; Formal Analysis, I.H., S.G., A.R., D.N., T.H., F.D.; Resources, F.D.; Writing—Original Draft, I.H., S.G., A.R., F.D.; Writing—Review & Editing, I.H., S.G., A.R., F.D.; Visualization, D.N., T.H.; Supervision, D.N., F.D.; Funding Acquisition, F.D.

Supporting information

Figure S1 Replication experiments to validate TFBS analysis selection for Sox9. A clear induction of plasmid harboring the ins allele and del was observed in both replicates (a) and (b), without stimulus, with a significant increase of luciferase activity in cells transfected with plasmid harboring ins allele. Although, while silencing Sox9 a significant decrease was observed in cells transfected with the plasmid harboring the del allele, confirmed in both experiments. Mean ± s.d. of three technical replicates of a representative experiment were plotted

Figure S2: Luciferase assays results in LoVo cells. Luciferase assays were performed in LoVo cells co‐transfected with pCMV6XL5_EMPTY vector or pCMV6XL5_KLF‐6 along with different pGL4.14 reporter constructs containing ATTGATTG “ins” allele or ATTG “del” allele or empty vector. A clear induction of plasmid harboring the ins and del allele was observed in both replicates (a) and (b) without stimulus with a significant increase of luciferase activity in cells transfected with plasmid harboring del allele. KLF‐6 overexpression shows a moderate decreased activity in case of plasmid harboring del allele.

Hamadou I, Garritano S, Romanel A, Naimi D, Hammada T, Demichelis F. Inherited variant in NFκB‐1 promoter is associated with increased risk of IBD in an Algerian population and modulates SOX9 binding. Cancer Reports. 2020;3:e1240. 10.1002/cnr2.1240

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon relevant request and in compliance with patient consent.