Association of SMUG1 SNPs in Intron Region and Linkage Disequilibrium with Occurrence of Cervical Carcinoma and HPV Infection in Chinese Population

Abstract

Background and Aims: This study was aim to investigate the relationship between the four intron SNPs (rs3087404, rs2029167, rs2029166 and rs7296239) of SMUG1 and the susceptibility of cervical squamous cell carcinoma.

Methods: Four SMUG1 intron SNPs (rs3087404, rs2029167, rs2029166 and rs7296239) were genotyped by MA-PCR in 400 CSCCs, 400 CIN III and 1200 controls. qRT-PCR and Western blot were used to detect the SMUG1 mRNA and protein expression.

Results: Interestingly, we found that the homozygous GG of rs3087404 had a significantly increased risk of CIN III [OR=1.78(1.27-2.51), P= 0.001] and CSCCs [OR=4.04(2.94-5.55), P=0.000]. The individuals with G allele or G carrier (AG +GG) at rs3087404 were at higher risk for CSCCs [OR=1.34 (1.04-1.71), P= 0.022]. Similarly, the homozygous GG of rs2029167 also had an increased risk of CIN III [OR=2.56 (1.91-3.43), P= 0.000] and CSCCs [OR=4.05(3.02-5.44), P=0.000]. The individuals with G allele or G carrier (AG +GG) at rs2029167 were at higher risk for CINIII [OR=1.41(1.10-1.80), P= 0.006] and CSCCs [OR=1.91 (1.48-2.47), P= 0.000]. In HR-HPV positive group, both the homozygous GG of rs3087404 and the homozygous GG of rs2029167 had an increased risk to CIN III and CSCC. Stratified analysis of the number of sexual partners and the age of first sexual intercourse found that the rs3087404 (A/G) had a particularly high level of enrichment in the CIN III or CSCCs groups. About the rs2029167 (A/G), we only found a particularly high level of enrichment grouping by the number of sexual partners in the CIN III and CSCCs groups. Meanwhile, we also found that there is a correlation between the SNPs of SMUG1 rs3087404 (A/G) and rs2029167 (A/G) with tumor cell differentiation and family heredity. But we didn't find that there was an association between the deferent genotypes of SMUG1 rs2029166 and rs7296239 with SMUG1 gene mRNA or protein expression. During the linkage disequilibrium analysis between rs3087404 (A/G) and rs2029167 (A/G), the genotype with AA-GG [OR=3.14(1.95-5.05)], AG-GG [OR=2.45(1.58-3.89)], GG-AA [OR=2.24(1.28-3.90)] and GG-AG [OR=2.58(1.54-4.32)] significantly increased the risk of CIN III. More notably, this risk is much greater in CSCCs: AA-GG [OR=7.13(4.03-12.61)], AG-GG [OR=7.22(4.21-12.38)], GG-AA [OR=8.60(4.73-15.63)], GG-AG [OR=9.64(5.43-17.13)]. Additionally, most GG (rs3087404) genotypes were linkage GG-AG (44/77, 80/140) in the CIN III and CSCCs, while most GG (rs2029167) genotypes were linkage genotype AG-GG (79/145, 112/184) in the CIN III and CSCCs, respectively.

Conclusions: These findings suggested that there was association between the two genetic polymorphisms of SMUG1 rs3087404(A/G) and rs2029167(A/G) with the susceptibility of CIN III and CSCCs, and there was a linkage disequilibrium between the rs3087404 with the rs2029167 in CIN III and CSCCs. This particular linkage disequilibrium can be used as predictive biomarkers of CIN III and CSCC.

Introduction

Around the world, cervical cancer (CC) is the fourth most common cancer among women, accounting for an estimated 529,572 diagnosed new cases and 274,967 deaths per year ¹. This cancer is the 3rd-leading cause of death in women' neoplasis worldwide and the morbidity of cervical cancer has increased recently ². In China, the cervical cancer has become the first major female cancer (98.9 per 100000) in addition to breast cancer. The mortality rate is up to 30.5 per 100000. And the incidence rate has the increasing trends ³.

Although several factors that contribute to cervical cancer development have been identified—mainly intrinsic factors (genetic), and extrinsic factors belonging to the high risk Human Papillomavirus (HR-HPV)—genetic factors show great potential as susceptibility or prognosis indicators^4,5. Only a small fraction (~1%) of cervical HR-HPV infection outcomes to cervical neoplasia, and the factors determining risk of progression are not entirely understood ⁶. Many genetic variants were associated with the risk of cervical cancer as supported by the epidemiological evidence ⁷.

Genomic instability due to DNA damage by carcinogens has been implicated in the initiation and development of cancer. DNA damage response and repair counteract the threats to genomic integrity, and variations in DNA repair capacity resulting from genetic variants could correlate with cancer predisposition ^8-11. The base excision repair (BER) pathway is the major DNA genetic damage repair pathway involved in genomic instability and tumorigenesis. Previous candidate gene studies showed that selected functional single nucleotide polymorphisms (SNP) in BER genes are associated with higher risk of several solid cancers ^12-17.

Single-strand selective mono-functional uracil-DNA glycosylase (SMUG1) is one of the BER genes which remove uracil from double-stranded and single-stranded DNA to maintain genomic stability following oxidative attacks ¹⁸. SNPs in this gene could have an effect on its enzyme capability of repairing DNA damage.

Xie et al evaluated the associations of 167 SNPs from 19 genes of the BER pathway with the risk of bladder cancer. 13 SNPs in 10 BER pathway genes were significantly associated with bladder cancer risk. The most significant SNP was rs2029167 in the SMUG1 gene ¹⁹. Similar studies also found a correlation between SNP of SMUG1 with breast cancer ²⁰ and colorectal cancer (CRC) ²¹. Until now, there is no report of SNP of SMUG1 in cervical cancer.

In this large scale case-control study, the aim was to investigate the relationship between the four intron SNPs (rs3087404, rs2029167, rs2029166 and rs7296239) of SMUG1 and the susceptibility of cervical squamous cell carcinoma (CSCC). Genotyping analyses of the four SMUG1 SNPs were performed in 400 CSCCs, 400 precursor lesion CIN III and 1200 normal controls.

Materials and methods

Study samples selection

400 CSCC cases, 400 CIN III cases and 1200 normal controls were recruited from Zhejiang Province, China. The diagnosis was determined by two pathologists. All subjects were unrelated ethnic Chinese women and recruited between 2004 Jun to 2008 Dec. Normal controls were randomly selected from healthy women volunteers during gynecologic examinations. The inclusion criteria for healthy volunteers were without gynecological neoplasm, cytological findings, endometriosis, other solid cancer and immune disorders.

Patients with pathological diagnosis of CINIII and cervical squamous cell carcinoma (CSCC) were included in the study. Considering that patients with CINI and CINII have unstable disease progression, we excluded these patients and only selected patients with CINIII. All cases of CSCC were FIGO stage Ia-IIb, histologically confirmed primary cervical carcinoma, treated radical hysterectomy with pelvic lymph node dissection, and did not receive any anticancer therapy prior to their surgery. Patients who are eligible for any of the following criteria are excluded: over 70 years of age, with other serious complications, or previous malignant disease.

Of these, 201 CSCC patients, 357 CIN III patients and 609 normal controls agreed to provide cervical brush-off samples for detecting HR-HPV. This study was approved by the Medical Ethical Committee of Women's Hospital, School of Medicine, Zhejiang University (No.2004002). All patients signed informed consent.

DNA Extraction and Genotyping

Genomic DNA was extracted from anticoagulant peripheral blood using a DNA extraction kit according to the manufactor's guideline (Sangon Bioengineering Co., Shanghai, China). All DNA samples were dissolved in water and hypothermic preservation ready to use.

The four intron SNPs (rs3087404[A/G], rs2029167 [A/G], rs2029166 [C/T] and rs7296239 [C/T]) of SMUG1was detected by Modified polymerase chain reaction-mismatch amplification (MA-PCR) (As described in detail previously ²²). The PCR forward and reverse primers and product length were showed in following table 1.

Table 1.

The PCR forward and reverse primers

SNP No.	Forward primer		Reverse primer	Product length
rs3087404 [A/G]	For “A”	5'-CTCATCAAGAGACTGCTGGA-3'	5'-ACTTTCATTGTTCCATAACT-3'	240bp
	For “G”	5'-CTCATCAAGAGACTGCTGGG-3'
rs2029167 [A/G]	For “A”	5'-GGGTGGTCCTCAGCTTGGCA-3'	5'-GCAGTGACTGGCAGGAGGCG-3'	184bp
	For “G”	5'-GGGTGGTCCTCAGCTTGGCG-3'
rs2029166 [C/T]	For “C”	5'-GCCATCTCTCATGGATTAAC-3'	5'-TTATGAGATAGCAGTGACTG-3'	228bp
	For “T”	5'-GCCATCTCTCATGGATTAAT-3'
rs7296239 [C/T]	For “C”	5'-CAGCCTCAACCCCAAAAGAC-3'	5'-TGGCTAATGTTGAGCGAAAT-3'	128bp
	For “T”	5'-CAGCCTCAACCCCAAAAGAT-3'

The PCR was performed in a 25ul reaction mixture, containing 50 ng of genomic DNA, 5.0 pmol of each primer, 0.2 mM of each dNTP and 1.0U of Taq DNA polymerase (TAKARA, Dalian, China). PCR undertook the following conditions: an initial denaturation at 94°C for 5 min, followed by 35 cycles of 94°C for 30s, 57°C for 30s, and 72°C for 1min, and a final step of 72°C for 10min. The PCR products were developed by 1.5% agarose gel electrophoresis, stained with ethidium bromide and visualized with a TyphoonTM 9410 Imaging System (GE Healthcare, USA). All samples were tested twice in double blind by two different technicians, and the reproducibility of the experiment was 100%.

HR-HPV detection

HR-HPV infection was identified using the Hybrid Capture II(HC II) assay (Digene Diagnostics Inc., Gaitherburg, MD, USA) using probe B, which includes a pool of RNA probes for HR-HPV 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, and 68. Cervical sampling for HR-HPV DNA was performed with the Digene Cervical Sampler.

SMUG1 mRNA expression by qRT-PCR

Freshly frozen tumor tissues of eighty seven CSCCs were selected. Total RNA was extracted using Trizol reagent according to the manufacturer's protocol (Invitrogen, USA). Total RNA was treated with RNase-free DNase I. cDNA was reversed transcription and used as a template for qPCR detection. The following PCR primer pairs were used for quantitative amplification; 95°C 30s, 40 cycles at 95°C 5s followed with 60°C 30s. The primers of SMUG1 (mRNA: NM_001243787.1) were 5'-CGCAACTACGTGACTCGCTA-3'; 5'-GTCCCAGCACTGGTCGTTTA- 3'. GAPDH was used as internal control. The primers of GAPDH (mRNA: NM_001256799.2) were 5'-GAGAAGGCTGGGGCTCATTT-3'; 5'-AGTGATGGCATGGACTGTGG-3'. The PCR product length of SMUG1 and GAPDH were 190bp and 231 bp, respectively. All reactions were performed with a ViiA 7 Dx System (ABI). The cutoff point (Ct) was defined as the value when the fluorescent signal increased above the background threshold. The ΔCt for gene-specific mRNA expression was calculated relative to the Ct of GAPDH. Relative mRNA expression was calculated with the formula: 2^-ΔCt.

SMUG1 protein expression by Western blotting

Eighty seven freshly CSCCs tissues were used to detect SMUG1 protein expression. Briefly, the tissue samples were minced on ice, and then the tissue was homogenized in the RIPA protein lysis buffer. The homogenized mixtures were rotated in the tubes at 4℃ for 1 h, and after centrifugation at 12,000 rpm at 4℃, the supernatant was collected and the protein concentrations were quantified.

10μl protein lysates were loaded into an 8% PAGE gel. Subsequently, electrophoretic separated proteins were transferred onto a 0.45µm PVDF membrane. After blocking with 5% non-fat milk for 1h, PVDF membrane was incubated with primary mouse monoclonal antibodies: SMUG1 (1:2000) purchased from NOVUS Biologicals (Cat No. H00023583-M07) and GAPDH (1:5000) purchased from Proteintech(Cat No. Cat.60004-1-Ig) for 4°C overnight, then were washed with TBS containing 0.05%Tween-20 for three times, followed by a 1h incubation with an HRP-conjugated secondary antibody (1:5000). After washing with TBS, the membranes were imaged with ECL using an Image Quant LAS 4000 mini (GE Healthcare).

Statistical Analysis

For the association between the genotypes and risk of cervical carcinoma, the odds ratio (OR), 95% confidence intervals (CIs) and P-values were obtained by binary logistic regression analysis. The control was set as the reference group for analysis. Stratified analysis of life style habits and genotype frequencies were evaluated with Kruskal-Wallis H test. The differences of quantitative mRNA and protein expression were calculated by ANOVA with a post hoc analysis (Fisher least significant difference). All reported values are two-tailed. The level of statistically significant difference was set at P≤0.05. All statistical analysis was done with SPSS software 18.0 ver for Windows.

Results

Clinical Features of Cases and Controls

40 years old/40 years old individuals were 602/598, 258/ 142 and 160/240 in the control, CIN III and carcinoma respectively. The carcinoma group had significantly more individuals >40 years old, but the CIN III group had more <40 years old individuals (P<0.001) compared to the control. There was no significant difference beside the increase of the proportion of individuals with number of parities more than 3 in the CIN III and carcinoma groups. The HR- HPV infection rate was 31.4%, 86.8% and 88.6% in the control, the CIN III and the carcinoma, respectively. HR-HPV infection in CIN III and cervical carcinoma cases were more than in controls. The Table 2 and data are quoted from our previously published work ²³.

Table 2.

Frequency distribution of select features by case control status

Variable		Control	CIN III	χ2*	P	Carcinoma	χ2*	P
		N=1200, N(%)	N=400, N(%)			N=400, N(%)
Age	≤40	602 (50.2)	258(64.5)	24.793	<0.001	160(40.0)	12.431	<0.001
	>40	598(49.8)	142(35.5)			240(60.0)
Age at the first intercourse	≤20 years	359(29.9)	130(32.5)	0.943	0.331	125(31.3)	0.253	0.615
	>20 years	841(70.1)	270(67.5)			275(68.8)
Number of sexual partners	≤ 1	963(80.3)	316(79.0)	0.292	0.589	309(77.3)	1.657	0.198
	>1	237(19.8)	84(21.0)			91(22.8)
Age at the first birth	≤22years	235(19.6)	91(22.8)	1.854	0.173	89(22.3)	1.321	0.25
	>22 years	965(80.4)	309(77.3)			311(77.8)
Number of parities**	≤ 3	548(45.7)	158(39.5)	4.627	0.031	131(32.8)	20.49	<0.001
	>3	652(54.3)	242(60.5)			269(67.3)
Smoking status	smoker	4(0.3)	2(0.5)	0.223	0.637	2(0.5)	0.223	0.637
	nonsmoker	1196(99.7)	398(99.5)			398(99.5)
HR-HPV infection	Positive	191(31.4)	310(86.8)	277.107	<0.001	178(88.6)	199.315	<0.001
	Negative	418(68.6)	47(13.2)			23(11.4)
	total	609	357			201

Bold values show statistical data with significant difference.^* Two-sided χ2 test. ^** Parities including full-term pregnancy and abortion at or after 28 weeks

Correlation analysis of SMUG1 SNPs Genotypes with risk of CSCC

Table 3 represents the genotypic and allelic frequencies of SMUG1 rs3087404, rs2029166, rs2029167 and rs7296239. Genotype distributions were in Hardy-Weinberg equilibrium. The CC, CT, and TT frequency of SMUG1 rs2029166 was 42.1%, 46.1% and 11.6% in the controls, 40.8%, 45.3% and 14.0% in the CIN III, 43.5%, 39.8% and 16.8% in CSCCs. The TT, TC, and CC frequency of SMUG1 rs7296239 was 33.0%, 51.9% and 15.1% in the controls, 36.0%, 51.8% and 12.3% in the CIN III, 34.8%, 49.3% and 16.0% in CSCCs, respectively. These results indicated that the SMUG1 rs2029166 and rs7296239 polymorphism were not associated with the risk for CIN III or CSCCs.

Table 3.

Association between SMUG1 polymorphisms and the risk of CIN III and cervical carcinoma

SMUG1 Genotypes			All patients and controls
			Control N=1200		CIN III N=400		adjusted OR* (95% CI)	P	Carcinoma N=400		adjusted OR* (95% CI)	P
			N	%	N	%			N	%
rs3087404
	AA		414	34.5	141	35.3	1.00(ref)		113	28.3	1.00(ref)
	AG		659	54.9	182	45.5	0.81(0.63-1.04)	0.103	147	36.8	0.82(0.62-1.08)	0.149
	GG		127	10.6	77	19.3	1.78(1.27-2.51)	0.001	140	35.0	4.04(2.94-5.55)	0.000
	AG+GG		786	65.5	259	64.8	0.97(0.76-1.23)	0.785	287	71.8	1.34(1.04-1.71)	0.022
	Allelic frequency
		Allele A	1487	62.0	464	58.0	1.00(ref)		373	46.6	1.00(ref)
		Allele G	913	38.0	336	42.0	1.78(1.00-1.39)	0.047	427	53.4	1.86(1.59-2.19)	0.000
rs2029166
	CC		504	42.0	163	40.8	1.00(ref)		174	43.5	1.00(ref)
	CT		557	46.4	181	45.3	1.01(0.79-1.28)	0.969	159	39.8	0.83(0.65-1.06)	0.131
	TT		139	11.6	56	14.0	1.25(0.87-1.78)	0.228	67	16.8	1.40(1.00-1.96)	0.053
	CT+TT		696	58.0	237	59.3	1.05(0.84-1.33)	0.661	226	56.5	0.94(0.75-1.18)	0.599
	Allelic frequency
		Allele C	1565	65.2	507	63.4	1.00(ref)		507	63.4	1.00(ref)
		Allele T	835	34.8	293	36.6	1.08(0.92-1.28)	0.347	293	36.6	1.08(0.92-1.28)	0.347
rs2029167
	AA		455	37.9	121	30.3	1.00(ref)		97	24.3	1.00(ref)
	AG		532	44.3	134	33.5	0.95(0.72-1.25)	0.700	119	29.8	1.05(0.78-1.41)	0.750
	GG		213	17.8	145	36.3	2.56(1.91-3.43)	0.000	184	46.0	4.05(3.02-5.44)	0.000
	AG+GG		745	62.1	279	69.8	1.41(1.10-1.80)	0.006	303	75.8	1.91(1.48-2.47)	0.000
	Allelic frequency
		Allele A	1442	60.1	376	47.0	1.00(ref)		313	39.1	1.00(ref)
		Allele G	958	39.9	424	53.0	1.70(1.45-1.99)	0.000	487	60.9	2.34(1.99-2.76)	0.000
rs7296239
	TT		396	33.0	144	36.0	1.00(ref)		139	34.8	1.00(ref)
	TC		623	51.9	207	51.8	0.91(0.71-1.17)	0.474	197	49.3	0.90(0.70-1.16)	0.415
	CC		181	15.1	49	12.3	0.74(0.52-1.08)	0.117	64	16.0	1.00(0.71-1.42)	0.967
	TC+CC		804	67.0	256	64.0	0.88(0.69-1.11)	0.272	261	65.3	0.93(0.73-1.17)	0.521
	Allelic frequency
		Allele T	1415	59.0	495	61.9	1.00(ref)		475	59.4	1.00(ref)
		Allele C	985	41.0	305	38.1	0.89(0.75-1.04)	0.145	325	40.6	0.98(0.84-1.16)	0.836

Bold values show statistical data with significant difference. *All P-values are adjusted for age, number of sexual partners, age at first intercourse, parities (including full-term pregnancy and abortion at or after 28 weeks) and age at first full-term pregnancy.

The AA, AG, and GG frequency of SMUG1 rs3087404 was 34.5%, 54.9% and 10.6% in the controls, 35.3%, 45.5% and 19.3% in the CIN III, 28.3%, 36.8% and 35.0% in CSCCs, respectively. These results revealed that women with the homozygous GG of rs3087404 had a significantly increased risk of CIN III [OR=1.78(1.27-2.51), P= 0.001] and CSCC [OR=4.04 (2.94-5.55), P=0.000]. We observed “A” allele is the major form at rs3087404 in controls (62.0%, 1487/ 2400), but “G” allele is the major form in CSCCs (53.4%, 427/800). The increased risk of “G” allele in CIN III and CSCCs were 1.78(1.00-1.39) and 1.86 (1.59-2.19) respectively. Data also indicated that individuals with “G” allele or “G” carrier (AG +GG) at rs3087404 were at higher risk for CSCCs [OR=1.34 (1.04-1.71), P= 0.022].

The AA, AG, and GG frequency of SMUG1 rs2029167 was 37.9%, 44.3% and 17.8% in the controls, 30.3%, 33.5% and 36.3% in the CIN III, 24.3%, 29.8% and 46.0% in CSCCs, respectively. These results revealed that women with the homozygous GG of rs2029167 had an increased risk of CIN III [OR=2.56 (1.91-3.43), P= 0.000] and CSCCs [OR=4.05(3.02-5.44), P=0.000]. “A” allele is the major form at rs2029167 in controls (60.1%, 1442/2400), but “G” allele is the major form in CIN III (53.0%, 424/800) and in CSCCs (60.9%, 487/800). “G” allele at rs2029167 was significantly higher in CIN III and CSCCs compared with normal controls. The increased risk of “G” allele in CIN III and CSCCs were OR=1.70(1.45-1.99) and OR=2.34(1.99-2.76) respectively. “G” allele or “G” carrier (AG +GG) at rs2029167 were at higher risk for CINIII [OR=1.41(1.10-1.80), P= 0.006] and CSCCs [OR=1.91 (1.48-2.47), P= 0.000].

As show in Table 4, in the HR-HPV positive group, though the homozygous GG of rs3087404 have not increased the risk of CIN III [OR=1.43(0.80-2.53, P= 0.226], it significantly increased the risk of CSCCs [OR=3.91(2.15-7.15), P=0.000], the increased risk of “G” allele in CSCCs were OR=1.94(1.45-2.60). The homozygous GG of rs2029167 had an increased risk of CIN III [OR=2.40 (1.48-3.89), P= 0.000] and CSCCs [OR=3.88(2.26-6.88), P=0.000]. Meanwhile, the increased risk of “G” allele in CIN III and CSCCs were OR=1.66(1.29-2.15) and OR=2.35(1.75-3.15).

Table 4.

Association between SMUG1 rs3087404 and rs2029167 polymorphisms with the risk of HR-HPV positive cervical carcinoma and CIN III

SMUG1 Genotypes			HPV-positive patients and controls
			Control N=191		CIN III N=310		adjusted OR*(95% CI)	P	carcinoma N=178		adjusted OR*(95% CI)	P
			N	%	N	%			n	%
rs3087404
	AA		65	34.0	111	35.8	1.00(ref)		49	27.5	1.00(ref)
	AG		103	53.9	143	46.1	0.81(0.55-1.21)	0.307	61	34.3	0.79(0.48-1.28)	0.332
	GG		23	12.0	56	18.1	1.43(0.80-2.53)	0.226	68	38.2	3.91(2.15-7.15)	0.000
	AG+GG		126	66.0	199	64.2	0.93(0.63-1.35)	0.686	129	72.5	1.36(0.87-2.12)	0.177
	Allelic frequency
		Allele A	233	61.0	365	58.9	1.00(ref)		159	44.7	1.00(ref)
		Allele G	149	39.0	255	41.1	1.09(0.84-1.42)	0.506	197	55.3	1.94(1.45-2.60)	0.000
rs2029167
	AA		71	37.2	92	29.7	1.00(ref)		42	23.6	1.00(ref)
	AG		83	43.5	103	33.2	0.96(0.63-1.46)	0.841	51	28.7	1.04(0.62-1.74)	0.885
	GG		37	19.4	115	37.1	2.40(1.48-3.89)	0.000	85	47.8	3.88(2.26-6.88)	0.000
	AG+GG		120	62.8	218	70.3	1.40(0.96-2.05)	0.083	136	76.4	1.92(1.22-3.02)	0.005
	Allelic frequency
		Allele A	225	58.9	287	46.3	1.00(ref)		135	37.9	1.00(ref)
		Allele G	157	41.1	333	53.7	1.66(1.29-2.15)	0.000	221	62.1	2.35(1.75-3.15)	0.000

Bold values show statistical data with significant difference. *All P-values are adjusted for age, number of sexual partners, age at first intercourse, parities (including full-term pregnancy and abortion at or after 28 weeks) and age at first full-term pregnancy.

Association between SMUG1 rs3087404, rs2029167 polymorphisms and the Sexual, Reproductive History in CSCCs and CIN III

As show in Table 5, the participants were divided into two groups according to Age, age of first sexual intercourse, number of sexual partners, age at first parity, number of parities and HR-HPV infection, then stratified analysis was done with the SMUG1 rs3087404 (A/G) and rs2029167 (A/G) genotype. Stratified analysis of age, number of parities, and age at first parity showed no correlation with rs3087404 (A/G) polymorphisms. However, we find a particularly high level of enrichment between groups with stratified analysis of the number of sexual partners in the CIN III (χ²=15.610, P=0.000) and CSCCs (χ²=13.468, P=0.000), and the age of first sexual intercourse in the CIN III (χ²=18.453, P=0.000) and CSCCs (χ²=15.528, P=0.000). We also did not find a high level of enrichment between HR-HPV positive and negative group of CIN III (χ²=0.176, P=0.675) and CSCCs (χ²=0.017, P=0.895).

Table 5.

Association between SMUG1 rs3087404 polymorphisms and the risk for CIN III and cervical carcinoma stratified by the sexual, reproductive history

High risk exposure		Controls						χ2	P	CIN III						χ2	P	Carcinoma						χ2	P
		AA		AG		GG				AA		AG		GG				AA		AG		GG
		N	%	N	%	N	%			N	%	N	%	N	%			N	%	N	%	N	%
Age
	≤ 40	212	35.2	323	53.7	67	11.1	0.036	8.490	93	36.0	120	46.5	45	17.4	0.837	0.360	44	27.5	61	38.1	55	34.4	0.000	0.991
	>40	202	33.8	336	56.2	60	10.0			48	33.8	62	43.7	32	22.5			69	28.8	86	35.8	85	35.4
Number of sexual partners
	≤ 1	321	33.3	538	55.9	104	10.8	2.590	0.108	120	38.0	151	47.8	45	14.2	15.610	0.000	93	30.1	126	40.8	90	29.1	13.468	0.000
	>1	93	39.2	121	51.1	23	9.7			21	25.0	31	36.9	32	38.1			20	22.0	21	23.1	50	54.9
Age at the first intercourse
	≤20	122	34.0	187	52.1	50	13.9	1.364	0.243	32	24.6	57	43.8	41	31.5	18.453	0.000	26	20.8	36	28.8	63	50.4	15.528	0.000
	>20	292	34.7	472	56.1	77	9.2			109	40.4	125	46.3	36	13.3			87	31.6	111	40.4	77	28.0
Number of parities
	≤ 3	184	33.6	299	54.6	65	11.9	1.055	0.304	51	32.3	74	46.8	33	20.9	1.094	0.296	34	26.0	52	39.7	45	34.4	0.069	0.792
	>3	230	35.3	360	55.2	62	9.5			90	37.2	108	44.6	44	18.2			79	29.4	95	35.3	95	35.3
Age at the first parity
	≤22	84	35.7	131	55.7	20	8.5	0.675	0.411	31	34.1	43	47.3	17	18.7	0.014	0.905	22	24.7	36	40.4	31	34.8	0.183	0.669
	>22	330	34.2	528	54.7	107	11.1			110	35.6	139	45.0	60	19.4			91	29.3	111	35.7	109	35.0
HR-HPV infection status
	Positive	65	34.0	103	53.9	23	12.0	0.595	0.440	111	35.8	143	46.1	56	18.1	0.176	0.675	49	27.5	61	34.3	68	38.2	0.017	0.895
	Negative	151	36.1	227	54.3	40	9.6			16	34.0	21	44.7	10	21.3			6	26.1	9	39.1	8	34.8

Bold values show statistical data with significant difference. Stratified analysis were applied by the Kruskal-Wallis H. A p value less than 0.05 was considered significant.

Data display of rs2029167 (A/G) as show in Table 6, we did not find a particularly high level of enrichment between groups, except for the number of sexual partners in the CIN III (χ²=10.214, P=0.001) and CSCCs (χ²=12.366, P=0.000), there was a particularly high level of enrichment.

Table 6.

Association between SMUG1 rs2029167 polymorphisms and the risk for CIN and cervical carcinoma stratified by the sexual, reproductive history

High risk exposure		Controls						χ2	P	CIN III						χ2	P	Carcinoma						χ2	P
		AA		AG		GG				AA		AG		GG				AA		AG		GG
		N	%	N	%	N	%			N	%	N	%	N	%			N	%	N	%	N	%
Age
	≤ 40	234	38.9	278	46.2	90	15.0	2.702	0.100	75	29.1	88	34.1	95	36.8	0.325	0.569	38	23.8	44	27.5	78	48.8	0.513	0.474
	>40	221	37.0	254	42.5	123	20.6			46	32.4	46	32.4	50	35.2			59	24.6	75	31.3	106	44.2
Number of sexual partners
	≤ 1	362	37.6	431	44.8	170	17.7	0.074	0.786	103	32.6	112	35.4	101	32.0	10.214	0.001	83	26.9	99	32.0	127	41.1	12.366	0.000
	>1	93	39.2	101	42.6	43	18.1			18	21.4	22	26.2	44	52.4			14	15.4	20	22.0	57	62.6
Age at the first intercourse
	≤20	132	36.8	164	45.7	63	17.5	0.121	0.728	38	29.2	42	32.3	50	38.5	0.312	0.576	29	23.2	36	28.8	60	48.0	0.274	0.601
	>20	323	38.4	368	43.8	150	17.8			83	30.7	92	34.1	95	35.2			68	24.7	83	30.2	124	45.1
Number of parities
	≤ 3	202	36.9	238	43.4	108	19.7	1.539	0.215	46	29.1	51	32.3	61	38.6	0.494	0.482	31	23.7	36	27.5	64	48.9	0.416	0.519
	>3	253	38.8	294	45.1	105	16.1			75	31.0	83	34.3	84	34.7			66	24.5	83	30.9	120	44.6
Age at the first parity
	≤22	88	37.4	101	43.0	46	19.6	0.237	0.626	25	27.5	29	31.9	37	40.7	0.933	0.334	19	21.3	23	25.8	47	52.8	1.797	0.180
	>22	367	38.0	431	44.7	167	17.3			96	31.1	105	34.0	108	35.0			78	25.1	96	30.9	137	44.1
HR-HPV infection status
	Positive	71	37.2	83	43.5	37	19.4	0.028	0.867	92	29.7	103	33.2	115	37.1	0.330	0.566	42	23.6	51	28.7	85	47.8	0.135	0.713
	Negative	149	35.6	191	45.7	78	18.7			16	34.0	15	31.9	16	34.0			5	21.7	6	26.1	12	52.2

Bold values show statistical data with significant difference. Stratified analysis were applied by the Kruskal-Wallis H. A p value less than 0.05 was considered significant.

Association between SMUG1 rs3087404, rs2029167 polymorphisms and the Clinical pathological characteristics in CSCCs

The correlation of SMUG1 rs3087404 and rs2029167 polymorphisms with CSCCs clinicopathological characteristics is shown in Table 7.The CSCCs were divided into two groups according to age, tumor family history, FIGO stage, tumor size, differentiation grade, lymph node metastasis, vascular involvement, stromal invasion, vaginal wall extension, parametrial extension, and endometrial extension, then stratified analysis was done with the SMUG1 rs3087404 (A/G) and rs2029167 (A/G) genotype.

Table 7.

Association between SMUG1 rs3087404 and rs2029167 polymorphisms and the risk for cervical carcinoma stratified by clinical pathological characteristics

Clinical pathological characteristics		SMUG1 rs3087404						χ2	P	SMUG1 rs2029167						χ2	P
		AA		AG		GG				AA		AG		GG
		N	%	N	%	N	%			N	%	N	%	N	%
Age
	≤ 40	42	26.3	66	41.3	52	32.5	0.022	0.882	36	22.5	52	32.5	72	45.0	0.003	0.955
	>40	71	29.6	81	33.8	88	36.7			61	25.4	67	27.9	112	46.7
Tumor family history
	Negative	108	29.4	139	37.9	120	32.7	8.792	0.003	87	23.7	113	30.8	167	45.5	0.01	0.919
	positive	5	15.2	8	24.2	20	60.6			10	30.3	6	18.2	17	51.5
FIGO stage
	I	96	28.4	123	36.4	119	35.2	0.002	0.966	84	24.9	103	30.5	151	44.7	1.337	0.248
	II	17	27.4	24	38.7	21	33.9			13	21.0	16	25.8	33	53.2
Tumor size
	<4cm	96	28.8	117	35.1	120	36.0	0.091	0.763	81	24.3	95	28.5	157	47.1	0.474	0.491
	≥4cm	17	25.4	30	44.8	20	29.9			16	23.9	24	35.8	27	40.3
Differentiation grade
	Grade I-II	104	30.1	131	37.9	111	32.1	9.265	0.002	89	25.7	108	31.2	149	43.1	8.112	0.004
	Grade III	9	16.7	16	29.6	29	53.7			8	14.8	11	20.4	35	64.8
Lymph node metastasis
	Negative	101	28.1	133	37.0	125	34.8	0.003	0.953	85	23.7	107	29.8	167	46.5	0.599	0.439
	positive	12	29.3	14	34.1	15	36.6			12	29.3	12	29.3	17	41.5
Vascular involvement
	Negative	96	28.1	127	37.1	119	34.8	0.001	0.979	82	24.0	104	30.4	156	45.6	0.017	0.896
	positive	17	29.3	20	34.5	21	36.2			15	25.9	15	25.9	28	48.3
Stromal invasion
	<2/3	84	29.8	106	37.6	92	32.6	2.363	0.124	74	26.2	84	29.8	124	44.0	2.263	0.132
	≥2/3	29	24.6	41	34.7	48	40.7			23	19.5	35	29.7	60	50.8
Vaginal wall extension
	Negative	92	29.0	121	38.2	104	32.8	2.224	0.136	76	24.0	92	29.0	149	47.0	0.438	0.508
	positive	21	25.3	26	31.3	36	43.4			21	25.3	27	32.5	35	42.2
Parametrail extension
	Negative	104	28.8	135	37.4	122	33.8	1.888	0.169	89	24.7	107	29.6	165	45.7	0.249	0.618
	positive	9	23.1	12	30.8	18	46.2			8	20.5	12	30.8	19	48.7
Endometrial extension
	Negative	106	28.7	139	37.7	124	33.6	2.767	0.096	91	24.7	111	30.1	167	45.3	1.017	0.313
	positive	7	22.6	8	25.8	16	51.6			6	19.4	8	25.8	17	54.8

Bold values show statistical data with significant difference. Stratified analysis were applied by the Kruskal-Wallis H. A P value less than 0.05 was considered significant.

Stratified analysis of age, FIGO stage, tumor size, lymph node metastasis, vascular involvement, stromal invasion, vaginal wall extension, parametrail extension, and endometrial extension showed no correlation with rs3087404 (A/G) or rs2029167 (A/G) polymorphism. However, we found a particularly high level of enrichment of rs3087404 (χ²=9.265, P= 0.002)) and rs2029167 (χ²=8.112, P=0.004) when stratified by differentiation grade. This means that GG homozygotes of rs3087404 and rs2029167 are significantly associated with the degree of malignancy of tumor differentiation. In addition, interestingly, we found that GG homozygote of rs3087404 is also associated with a family history of the tumor (χ²=8.792, P=0.003).

mRNA and protein expression from patients with different genotypes of SMUG1 rs3087404 (A/G) or rs2029167 (A/G)

As shown in Figure 1, among the 87 CSCCs patients, the genotypes of AA, AG, and GG at rs3087404 were 25(28.7%), 32(36.8%) and 30(34.5%) cases, while the genotypes of AA, AG, and GG at rs2029167 were 22(25.3%), 26(30.0%) and 39(44.8%) cases, respectively. There was no significant difference in the expression of SMUG1 mRNA with different genotypes at rs3087404 (F=1.022, P=0.364) or at rs2029167 (F=2.067, P=0.133).

Figure 1.

SMUG1 mRNA expression in CSCCs with different genotypes of rs3087404 and rs2029167 (qPCR).

As shown in Figure 2, Western Blot experiments confirmed that the polymorphism of rs3087404 did not affect the expression of SMUG1 protein (F=0.254, P=0.781). Similarly, the polymorphisms of rs2029167 is also independent of the expression level of SMUG1 protein (F=1.346, P=0.308).

Figure 2.

SMUG1 protein expression in CSCCs with different genotypes of rs3087404 and rs2029167 (Western Blot) (A) AA: rs3087404 genotype is AA; AG: rs3087404 genotype is AG; GG: rs3087404 genotype is GG; (B) AA: rs2029167 genotype is AA; AG: rs2029167 genotype is AG; GG: rs2029167 genotype is GG

Haplotype Analysis between the Linkage Disequilibrium of the SMUG1 rs3087404 and rs2029167 Variants Genotypes and the Risk of CIN III and CSCCs

Since that the frequencies of both rs3087404 (A/G) and rs2029167 (A/G) genotypes change the risk of CIN III or CSCCs significantly, we further analyzed the linkage disequilibrium between rs3087404 (A/G) and rs2029167 (A/G). The frequencies of the nine haplotypes were shown in Table 8. GG (rs3087404)-GG (rs2029167) was not detected in normal control and CSCCs, except for 1 case detected in CIN III. Compared to AA (rs3087404)-AA (rs2029167), the genotype with AA-GG [OR=3.14 (1.95-5.05), P=0.000], AG-GG [OR=2.45(1.58-3.89), P= 0.000], GG-AA [OR=2.24(1.28-3.90), P=0.005], GG-AG [OR=2.58(1.54-4.32), P=0.027] significantly increased the risk of CIN III. More notably, this risk is much greater in CSCC: AA-GG [OR=7.13(4.03-12.61), P= 0.000], AG-GG [OR=7.22(4.21-12.38), P=0.000], GG- AA [OR=8.60(4.73-15.63), P=0.000], GG-AG [OR=9.64 (5.43-17.13, P=0.000]. This means that whether the rs3087404 or rs2029167 is GG homozygote, the linkage mode is at high risk. We also found that women with the AG-AG genotype had a decreased risk for CSCCs [OR=0.49(0.25-0.96), P=0.038].

Table 8.

Genotypes and the risk of all CIN III and cervical carcinoma subjects

SMUG1 Genotypesa	All patients and controls
	Control			CIN III		adjusted ORb(95% CI)	P	Carcinoma		adjusted ORb(95% CI)	P
	1200			400				400
	N	%		N	%			N	%
AA-AA	158	13.2		39	9.8	1.00(ref)		19	4.8	1.00(ref)
AA-AG	172	14.3		37	9.3	0.87(0.53-1.44)	0.589	22	5.5	1.06(0.56-2.04)	0.853
AA-GG	84	7.0		65	16.3	3.14(1.95-5.05)	0.000	72	18.0	7.13(4.03-12.61)	0.000
AG-AA	239	19.9		50	12.5	0.85(0.53-1.35)	0.485	18	4.5	0.63(0.32-1.23)	0.174
AG-AG	291	24.3		53	13.3	0.74(0.47-1.17)	0.192	17	4.3	0.49(0.25-0.96)	0.038
AG-GG	129	10.8		79	19.8	2.45(1.58-3.89)	0.000	112	28.0	7.22(4.21-12.38)	0.000
GG-AA	58	4.8		32	8.0	2.24(1.28-3.90)	0.005	60	15.0	8.60(4.73-15.63)	0.000
GG-AG	69	5.8		44	11.0	2.58(1.54-4.32)	0.027	80	20.0	9.64(5.43-17.13)	0.000
GG-GG	0	0.0		1	0.3	_	_	0	0.0	_	_

Bold values show statistical data with significant difference. ^agenotypes are composed of two polymorphic sites: rs3087404(A/G), rs2029167(A/G). ^bAll P-values are adjusted for age, number of sexual partners, age at first intercourse, parities (including full-term pregnancy and abortion at or after 28 weeks) and age at first full-term pregnancy.

Additionally, most GG (rs3087404) genotypes were linkage GG-AG (44/77, 80/140) in the CIN III and CSCC, while most GG (rs2029167) genotypes were linkage genotype AG-GG (79/145, 112/184) in the CIN III and CSCCs, respectively. These indicated that the majority of GG genotype distributions are caused by the linkage disequilibrium with the corresponding alleles.

Discussion

Uracil misincorporation into DNA arises spontaneously at low level as a result of cytosine deamination or misincorporation of dUMP during DNA replication ^{24, 25}. Under normal conditions, such lesions are rapidly recovered by the BER pathway initiated by uracil-DNA glycosylases (UDG) ^{26, 27}. In most organisms, including humans, uracil is generally an undesirable ingredient in the genome. Thus strategies are in place to remove uracil once occurring of the DNA damage. So, sophisticated mechanisms are essential for the removal of uracil from DNA and prevention of its misincorporation, and maintain genomic integrity and stability. The failure of removing misincorporated uracil from DNA will result in base abnormity during DNA replication, even lead to dsDNA breaks and chromosomal aberrations, these two events are the key genetic factors of tumorigenesis^{25, 28-30}.

BER is a highly conserved DNA repair system from bacteria to humans ^31-33. A great variety of DNA-damaging agents can cause genome instability, which would be a tremendous matter for cells if the damaged DNA is not recovered ¹¹. The most important role of BER is to remove DNA damage caused by various carcinogens, such as reactive oxygen species (ROS), ionizing radiation and so on³⁴. In humans, four UDGs have been identified, encoded by the UNG, SMUG1, MBD4 and TDG genes ^33,35,36. Most of these critical BER genes are highly polymorphic ³⁷. Genetic variations of these genes are likely to alter BER enzyme functional activity, and influence cancer risk ³⁶.

The human single-strand-selective monofunctional uracil-DNA glycosylase 1(SMUG1; also named: FDG, UNG3 and HMUDG) is located in the Chromosome 12q13.11-13.3 ^{26, 38}. This gene encodes for a uracil DNA glycosylase (UDG) of the BER pathway that removes uracil, from single stranded (ssDNA) as well as double stranded DNA (dsDNA) ³⁹. As SMUG1 removes uracil and 5-hmeU from ssDNA and dsDNA, this enzyme may take participate in the repair of deamination and oxidation damage. The SMUG1 is the major enzyme involved in the removal of 5-hmeU from damaged DNA ⁴⁰.

Several researchers reported the SNPs of SMUG1were correlationship with bladder cancer, breast cancer and CRC susceptibility. In a matched study of 801 bladder cancer cases, Xie et al. found 13 SNPs in10 BER pathway genes significantly increased the risk of bladder cancer. The most significant variant was SMUG1 rs2029167 (A/G). The homozygous GG genotype increased a 1.42-fold risk of bladder cancer ¹⁹.In another 1,077 case-controls matched study of incident breast cancer, Marian et al suggested that there was increased risk of breast cancer among postmenopausal women who were heterozygous of two of SMUG1 SNPs which is thought to be the most active glycosylase in vivo, raises the possibility that subtle 'heterosis' effects on cancer risk might be produced by these SNPs ²⁰. In a study of CRC, Broderick et al reported that genetic variations in TDG, UNG and SMUG1 may play a role in the susceptibility of CRC ²¹. These reports remind us to make a hypothesis which there is an association between the genetic variants of SMUG1 gene with cancer risk. We carried out the correlation study of cervical cancer and SMUG1 SNPs.

The initiation and development of cervical carcinoma involves reversible transformation in the cervical squamous cells resulting in various cellular abnormalities and ultimately to cervical tumorigenesis. The development of cervical carcinoma usually requires multiple stages, eventually developing from precursor lesion cervical intraepithelial neoplasia (CIN) to cervical malignant carcinoma ²¹. In our results, the two of SMUG1 rs2029166 and rs7296239 polymorphisms were not associated with the risk for CIN III or CSCC. Interestingly, the homozygous GG of rs3087404 and rs2029167 had a significantly increased risk of CIN III and CSCC. We also observed the increased risk of G allele of these two SNP in CIN III and CSCC. The individuals with G allele or G carrier (AG +GG) at rs3087404 and rs2029167 were at higher risk for CSCC. These findings indicated that the SMUG1 rs2029166 and rs7296239 polymorphisms (G allele) maybe play a role in initiation and progression of precancerous lesions (CIN) and cervical carcinoma. So far, there is no study about the correlation between cervical carcinoma and the SMUG1 rs2029166 and rs7296239 polymorphisms (G allele). The present study is the first time to discover the association between the SMUG1 rs2029166 and rs7296239 polymorphisms (G allele) and cervical carcinoma or CIN III. In general, SNP loci that affect the structure and function of genes are located in the 5' UTR promoter, coding region, or 3' UTR region. Although our two variants both are located in the intron which cannot change the amino acid, it is possible that there is linkage disequilibrium with other functional genetic variants and serves as a genetic marker of susceptibility ⁴¹. Another possibility is that the SMUG1 rs2029166 and rs7296239 genetic variants maybe influence primary mRNA splicing and regulation, and affects SMUG1 protein expression or produce alternative spliceosome. To validate the SMUG1 expression change, we detected the mRNA and protein expression in fresh tumor tissues in the different genotype groups of SMUG1 rs2029166 and rs7296239, but we discovered that there was no association between the genotype of SMUG1 rs2029166 and rs7296239 with SMUG1 gene mRNA or protein expression. These indicate that the tumor susceptibility induced by the polymorphism of this locus was not achieved by altering gene expression. Bonnet et al. speculated that the introns take participate in maintaining genetic stability at certain locations, particularly in highly expressed genes ⁴², and repair genes are often high expression genes.

During the linkage disequilibrium analysis between rs3087404 (A/G) and rs2029167 (A/G), we found that whether the rs3087404 or rs2029167 is GG homozygote, the linkage mode is at high risk. Additionally, most GG (rs3087404) genotypes were linkage GG-AG (44/77, 80/140) in the CIN III and CSCCs, while most GG (rs2029167) genotypes were linkage genotype AG-GG (79/145, 112/184) in the CIN III and CSCCs, respectively. These indicated that the majority of GG genotype distributions are caused by the linkage disequilibrium with the corresponding alleles. These linkage modes can be used as genetic biomarker of early prediction of cervical carcinoma, as an indicator of primary prevention.

Stratified analysis of the number of sexual partners and the age of first sexual intercourse found that the rs3087404 (A/G) had a particularly high level of enrichment in the CIN III and CSCCs. About the rs2029167 (A/G), we only found a particularly high level of enrichment grouping by the number of sexual partners in the CIN III. This suggests that there may be a certain correlation between SMUG1 rs3087404 (A/G) and rs2029167 (A/G) variants with the female sexual behavior.

Among all of clinical parameters, we found that the genetic polymorphisms of rs3087404 (A/G) and rs2029167 (A/G) are significantly associated with the degree of malignancy of tumor differentiation; homozygous GG genotype increases the risk of malignant cell differentiation grade of tumors. In addition, interestingly, we found that GG homozygote of rs3087404 is also associated with a family history of the tumor. These indicate that there maybe a correlation between the SNPs of SMUG1 rs3087404 (A/G) and rs2029167 (A/G) with tumor cell differentiation and family heredity.

In HR-HPV positive group, we found that the homozygous GG of rs3087404 and rs2029167 both significantly increased the risk of CSCCs, only “G” allele or “G” carrier (AG +GG) at rs2029167 were at higher risk for CSCCs. But, in stratified analysis, we did not find a high level of enrichment between HR-HPV positive and negative groups of CIN III and CSCCs. These indicated that the rs3087404 and rs2029167 involved in the cervical tumorigenesis, but they maybe not affect the HR-HPVs infection at early onset of disease. In the process of affecting cervical tumorigenesis, rs2029167 variant may be more effective than those of rs3087404 variant.

These findings suggested that there was association between the two of SMUG1 rs3087404 (A/G) and rs2029167(A/G) genetic variant with the susceptibility of CIN III and CSCCs, but not HR-HPVs infection. Whether the rs3087404 or rs2029167 is GG homozygote, there was linkage disequilibrium between these two of polymorphism leading to increase the risk of CIN III and CSCC. These linkage modes can be used as genetic biomarker of early prediction of cervical carcinoma, as an indicator of primary prevention.

Abbreviations

CC

cervical cancer

HR-HPV

high-risk HPV

BER

base excision repair

SNP

single nucleotide polymorphisms

SMUG1

Single-strand selective mono-functional uracil-DNA glycosylase

CRC

colorectal cancer

CSCC

cervical squamous cell carcinoma

CIN

cervical intraepithelial neoplasia

MA-PCR

Modified polymerase chain reaction-mismatch amplification

HC II

Hybrid Capture II

OR

odds ratio

CIs

95% confidence intervals

UDG

uracil-DNA glycosylases

ROS

reactive oxygen species

ssDNA

single stranded DNA

dsDNA

double stranded DNA.