Association of the rs1333048 polymorphism in the ANRIL gene with different clinical forms of periodontitis: A meta-analysis

Abstract

BACKGROUND

Periodontitis is a chronic inflammatory disorder influenced by both behavioral and genetic factors. Among epigenetic regulators, the ANRIL gene has been proposed as a risk factor for periodontitis; however, findings on the association between ANRIL polymorphisms and disease susceptibility remain inconsistent.

AIM

To analyze the association between the rs1333048 genetic polymorphism in the ANRIL gene and periodontitis via meta-analysis.

METHODS

A literature search was performed for studies published before May 2, 2025. The Review Manager statistical program was used in analyses with calculations of heterogeneity index (I²) and odds ratio (OR) with 95% of confidence intervals (CI). Begg’s test and the Egger’s linear regression test were used for publication bias evaluation using Comprehensive meta-analysis software. P < 0.05 was considered significant.

RESULTS

From 12 studies including 5489 participants across multiple ethnic groups, we observed a statistically significant association between ANRIL gene polymorphisms and periodontitis in the allelic contrast model (OR = 1.24 95%CI: 1.15-1.34, P < 0.00001). Conversely, the wild-type allele was significantly associated with the control group (OR = 0.80 95%CI: 0.75-0.87, P < 0.00001). Heterogeneity was low (I² = 28%, P_{heterogeneity} = 0.17), and no significant risk of publication bias was detected (P > 0.05).

CONCLUSION

In conclusion, this meta-analysis demonstrated a significant association between the rs1333048 polymorphism and periodontitis in the overall analysis and in stratified analyses of Caucasian populations, but in for mixed-race populations.

Key Words: Periodontal diseases; RNA; Alleles; Odds ratio; DNA methylation

Core Tip: Many case-control studies have identified a relationship between the rs1333048 polymorphism and periodontitis. However, the impact of this genetic variation on the clinical aspect of periodontitis is still unknown.

INTRODUCTION

Periodontal diseases arise from dysbiotic shifts in the oral microbiota, with the host immune response serving as the central etiological factor. Hyper-inflammation drives tissue damage in the periodontium, ultimately leading to tooth loss[1]. Periodontal diseases are classified in several ways, with chronic periodontitis (CP) and aggressive periodontitis (AgP) being the most common forms.

The epidemiology of periodontitis is substantial, with recent data indicating a high global prevalence. Severe forms affect approximately 10% of the global population[2]. In the United States, periodontitis has been reported in 42% of adults aged 30 years and older[3].

Multiple factors contribute to the development and progression of periodontitis, including poor oral hygiene, specific subgingival bacterial species[4], and genetic variations in inflammatory mediators[5]. Epigenetic events also play an important role by altering local gene expression linked to inflammation[6].

Epigenetic events are complex but potentially reversible processes that influence gene expression without altering DNA sequence. They involve mechanisms such as DNA methylation, histone modifications and regulation by long non-coding RNAs (lnRNAs)[7]. It has been hypothesized that lnRNAs interact with histone-modifying machinery and thereby contribute to diverse biological conditions, ranging from cancer to inflammatory diseases.

The antisense non-coding RNA in the INK4 locus (ANRIL) is one such lnRNA, identified as a candidate molecule in periodontitis. ANRIL regulates the expression of multiple genes through both cis- and trans-acting mechanisms, interacting with DNA methyltransferases and other regulatory proteins[8]. It has been shown to contribute significantly to cancer and has therefore garnered considerable attention in recent years.

Schaefer et al[9] first reported significant associations between two ANRIL polymorphisms (rs1333042 and rs1333048) and generalized AgP. However, a subsequent study in a Han Chinese population did not find a significant association between these polymorphisms and the disease[10].

Other studies have reported on the associations between single nucleotide polymorphisms (SNPs) in the ANRIL gene and periodontitis[11,12]. However, the results remained unclear and are challenging to reconcile. Contradictory findings may be better evaluated through meta-analyses, as demonstrated by many previous studies on polymorphisms and inflammatory conditions[13-15].

Although a previous meta-analysis has attempted to clarify the relationship between the rs1333048 polymorphism and periodontitis[16], the authors’ approach did not allow a complete evaluation on this genetic variation and periodontitis risk, as stratified which analyses by different clinical forms of periodontitis are lacking.

Therefore, this study aimed to perform a meta-analysis of the association between the rs1333048 polymorphism in the ANRIL gene and periodontitis risk, including an evaluation of stratified data.

MATERIALS AND METHODS

Protocol registration

This meta-analysis was registered in the PROSPERO database with the following assigned number: CRD42018116487.

For the design and conduct of this study, the authors followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines[17] and completed the corresponding PRISMA Checklist (Supplementary Table 1).

Question formulation

Is periodontitis directly influenced by a genetic variation in the ANRIL gene? What is the association between the rs1333048 polymorphism in ANRIL gene and periodontitis? What is the clinical significance of this polymorphism in relation to periodontitis?

Eligibility criteria

The following inclusion criteria were applied in the literature search: (1) Case-control studies or replication genetic studies; (2) Diagnosis of periodontitis (in any of its clinical manifestations) confirmed through rigorous clinical evaluation[18]; (3) Control patients with clinically healthy periodontium; (4) Complete documentation of genotypic frequencies; (5) Studies conducted in human subjects; and (6) Participants (cases and controls) without cardiovascular diseases, autoimmune conditions, or pregnancy.

Strategy search

The literature search strategy was independently conducted by two authors in the Cochrane Library, EMBASE, Google Scholar, MEDLINE, PubMed and Web of Science. Studies published before May 2, 2025, were considered. The following combination of keywords was used: “periodontitis OR periodontal disease OR chronic periodontitis OR aggressive periodontitis” and “polymorphism OR genetic variation” and “ANRIL OR human chromosome region 9p21.3” and “rs1333048 polymorphism”.

No language restrictions were applied, and references from the included studies were evaluated for inclusion.

Data collection process

Three investigators collected the data using a standardized form, which was used to compile the table of study characteristics presented in the Results. The data were extracted by author, year of publication, ethnicity/country of the included participants, study design, clinical diagnosis, subject type evaluated, sample size, mean age of the participants, respect to the Hardy-Weinberg Equilibrium, genotyping method, risk of bias assessment, described below.

Risk of bias assessment

To assess methodological quality, the scale proposed by Nibali[19] for improving the quality of meta-analyses in periodontal genetic studies was applied. Studies that received fewer than 10 points on this binary scale were excluded.

Statistical analysis

The Review Manager software version 5.3 (RevMan, Nordic Cochrane Centre, The Cochrane Collaboration, 2012) for systematic reviews and meta-analyses was used for statistical analysis. Publication bias was evaluated using the Comprehensive Meta-analysis statistical software (Software version 3.3.070 (2014) available free online as a trial.

The presence or absence of true heterogeneity (I²) was calculated by Cochran’s Q test or the χ²Q-based statistical test. I² was analyzed through visualization of the Funnel plot for heterogeneity. When the observed value of I² presented a non-statistical significance and was defined as mild or moderate (I² < 50%, P > 0.05), the authors used the Fixed-effect model for the pooled odds ratio (OR) calculation. When I² presented a statistically significant value and was defined as elevated (I² > 50%, P < 0.05), the Random-effects statistical model was used for the OR calculations. A P value < 0.05 was considered statistically significant.

To quantify the influence of genetic variation on disease risk, six genetic models were calculated, using “M” to denote the major allele and “m” the minor allele. Therefore, the calculations included: (1) Allelic comparisons (M vs m; m vs M); (2) Genotypic comparisons (MM vs mm; mm vs MM); and (3) Combined genotypic models (MM vs mm + Mm; and Mm vs MM + mm). In addition, a sensitivity analysis was performed by omitting one study at a time to assess any significant changes in the OR value. For the assessment of publication bias, Begg’s test and Egger’s linear regression test were used (with P < 0.05). Funnel plot asymmetry for publication bias was also considered to validate Begg’s test and Egger’s test results. All included data in the studies have been dichotomous data expressed as OR with 95% of confidence intervals (CI) to determine the possible association between the genetic variations and periodontitis.

RESULTS

Characteristics of included studies

The initial literature search identified 102 records across the selected databases. After screening and evaluation of eligibility, 95 studies were excluded, leaving seven for inclusion in the qualitative synthesis (Figure 1). In total, the analyses enrolled 5489 participants genotyped for the ANRIL gene polymorphism. Further details are provided in Table 1.

Figure 1  Flow diagram for identification, screening, eligibility and analysis of included studies in the meta-analysis.

Table 1 Characteristics of included studies on rs1333048 polymorphisms in ANRIL gene and periodontitis.

Reference	Ethnicity/country	Study design	Diagnosis (clinical subtypes)	Subject type	Sample size (Cc/Co)	Age (years) (Cc/Co)	HWE (P value)	Methods of genotyping	Score
Chen et al[10], 2012	Asian/China	Case/Control	CP	-	221/172	41.6 ± 8.8/40.0 ± 9.2	Yes	Real time PCR	14
Ernst et al[11], 2010	Caucasian/German and Northern Ireland	Case/Control	gAgP	-	130/339	15-35/40-80	Yes	Real time PCR	13
Schaefer et al[9], 2009	Caucasian/Germany	Verification/Replication	gAgP	-	151/736	38 ± 6.6/63 ± 7.4	Yes	SNPlex/TaqMan	14
			LAgP	-	137/368	37 ± 7.0/37 ± 9.6
Schaefer et al[12], 2011	Caucasian/Germany and Netherlands	Case/Control	CP	-	154/421	45.2 ± 9.5/32.3 ± 9.0	Yes	SNPlex/TaqMan	15
			AgP		159/421	33.8 ± 4.7/32.3 ± 9.0
Schaefer et al[20], 2013	Caucasian/Turkey, Italy and Germany	Exploration	AgP	Turkey	89/72	-	Yes	Array Immunochip	18
				Italy	44/99		Yes
				Germany	594/1441		Yes
Taiete et al[21], 2018	Mixed/Brazil	Case/Control	AgP	-	200/196	34.0 ± 4.6/30.5 ± 5.8	No	PCR-RFLP	17
			CP	-	190/196	50.0 ± 7.2/30.5 ± 5.8	Yes		16
Teeuw et al[22], 2015	Caucasian/Netherlands	Case/Control	Periodontitis	-	115/56	45.5 ± 9.9/43.9 ± 13.2	Yes	TaqMan

Cc: Case; Co: Control; HWE: Hardy-Weinberg Equilibrium, CP: Chronic periodontitis; gAgP: Generalized aggressive periodontitis; LAgP: Localized aggressive periodontitis; AgP: Aggressive periodontitis; PCR: Polymerase chain reaction.

The included articles, published between 2009 and 2018, encompassed several ethnic groups and comprised seven articles with a total of 12 studies on the ANRIL gene polymorphism[9,10,11,12,20-22]. The studies were performed in Caucasians from different countries[9,11,12,20,22], Asian from China[10] and mixed-race population from Brazil[21]. It is important to note that Schaefer[9] performed genetic evaluation on Caucasians from Germany, Schaefer[12] on Caucasians from Germany and Netherlands and Schaefer[20] on Caucasians from Turkey, Italy and Germany. Five articles were performed by case/control design[10,11,12,21,22], one by exploration design[9] and one by verification/replication design[20].

In the included studies, periodontitis was diagnosed as CP[10,12,20,21] generalized AgP[9,11] and localized AgP[9]. All included studies reached more than 10 points on the scale proposed by Nibali[19] as shown in Table 1.

Statistical analysis

In the overall analysis, a statistically significant association was found between the mutant allele of the ANRIL gene polymorphism and increased risk of periodontitis (OR = 1.24, 95%CI: 1.15-1.34, P < 0.05). Conversely, the wild-type allele was significantly associated with the control group (OR = 0.80, 95%CI: 0.75-0.87, P < 0.05). A non-significant I² value (28%, P_{heterogeneity} = 0.17) was observed for both evaluations (Figure 2). Similar results were obtained for the mutant homozygous and wild-type homozygous genotypes. Due to the non-significant interference of heterogeneity, the Fixed-effect statistical model was applied. The OR values and data about I² are shown in Table 2. Sub-group evaluation demonstrated a significant association between the rs1333048 polymorphism in ANRIL gene and CP (P = 0.006), AgP (P < 0.05), and the Caucasian population (P < 0.05), but not in the mixed-race population (P = 0.55). The Fixed-effect statistical model was applied due to the decreased I² value. However, the calculations for the mixed-race population were affected by significant heterogeneity.

Figure 2 Forest plot. A and B: Comparison between the mutant allele vs the wild-type allele (A) and the wild-type allele vs the mutant allele (B) in the rs1333048 polymorphism in ANRIL gene and periodontitis risk.

Table 2 Meta-analysis calculation of the rs1333048 polymorphism in ANRIL gene and periodontitis risk (allelic and genotypic comparisons; subgroups analyses).

Comparison (n)	OR (95%CI)	P value (Z test)	I², %	Pheterogeneity	Statistical model used
Overall (n = 12)
M vs m	1.24 (1.15-1.34)	P < 0.05	28	P = 0.17	F
m vs M	0.80 (0.75-0.87)	P < 0.05	28	P = 0.17	F
MM vs mm	1.53 (1.32-1.78)	P < 0.05	30	P = 0.16	F
mm vs MM	0.65 (0.56-0.76)	P < 0.05	30	P = 0.16	F
MM vs Mm/mm	1.38 (1.22-1.56)	P < 0.05	27	P = 0.18	F
Mm vs MM/mm	0.97 (0.87-1.08)	P = 0.53	25	P = 0.20	F
CP (n = 3)
M vs m	1.25 (1.07-1.47)	P = 0.006	0	P = 0.81	F
m vs M	0.80 (0.68-0.94)	P = 0.006	0	P = 0.81	F
MM vs mm	1.58 (1.14-2.18)	P = 0.005	0	P = 0.77	F
mm vs MM	0.63 (0.46-0.87)	P = 0.005	0	P = 0.77	F
MM vs Mm/mm	1.36 (1.05-1.77)	P = 0.02	0	P = 0.80	F
Mm vs MM/mm	0.98 (0.79-1.23)	P = 0.89	0	P = 0.92	F
AgP (n = 8)
M vs m	1.23 (1.13-1.34)	P < 0.05	47	P = 0.07	F
m vs M	0.81 (0.75-0.89)	P < 0.05	47	P = 0.07	F
MM vs mm	1.49 (1.25-1.77)	P < 0.05	48	P = 0.06	F
mm vs MM	0.67 (0.57-0.80)	P < 0.05	48	P = 0.06	F
MM vs Mm/mm	1.37 (1.19-1.58)	P < 0.05	51	P = 0.05	F
Mm vs MM/mm	0.95 (0.84-1.08)	P = 0.44	49	P = 0.06	F
Caucasian (n = 9)
M vs m	1.27 (1.16-1.38)	P < 0.05	25	P = 0.22	F
m vs M	0.79 (0.72-0.86)	P < 0.05	25	P = 0.22	F
MM vs mm	1.59 (1.34-1.88)	P < 0.05	26	P = 0.21	F
mm vs MM	0.63 (0.53-0.74)	P < 0.05	26	P = 0.21	F
MM vs Mm/mm	1.39 (1.21-1.60)	P < 0.05	43	P = 0.08	F
Mm vs MM/mm	0.97 (0.87-1.08)	P = 0.94	33	P = 0.15	F
Mixed (n = 2)
M vs m	1.11 (0.91-1.36)	P = 0.55	68	P = 0.08	R
m vs M	0.90 (0.63-1.28)	P = 0.55	68	P = 0.08	R
MM vs mm	1.22 (0.83-1.81)	P = 0.55	70	P = 0.07	R
mm vs MM	0.82 (0.55-1.21)	P = 0.55	70	P = 0.07	R
MM vs Mm/mm	1.26 (0.93-1.72)	P = 0.14	0	P = 0.49	F
Mm vs MM/mm	0.84 (0.64-1.11)	P = 0.23	30	P = 0.23	F

OR: Odds ratio; CI: Confidence Intervals; I²: Heterogeneity index; M: Mutant allele; m: Wild type allele; F: Fixed-effect statistical model; R: Random-effects statistical model; CP: Chronic periodontitis; AgP: Aggressive periodontitis.

Sensitive analysis and publication bias

Sensitive analysis showed that no single study significantly altered the pooled OR values, supporting the robustness of the results (Figure 3). Furthermore, no obvious asymmetry was detected in the Funnel plot for the analyzed polymorphism (Figure 4). Egger’s linear regression test and Begg’s test provided no evidence of publication bias for the rs1333048 polymorphism (P = 0.477 and P = 0.243, respectively).

Figure 3  Sensitivity analysis for the mutant allele vs the wild-type allele in the rs1333048 polymorphism in ANRIL gene and periodontitis risk in the overall comparison.

Figure 4  Funnel plot for the detection of publication bias for the mutant allele vs the wild type.

DISCUSSION

To the best of our knowledge, this is the first systematic evaluation with meta-analysis study to investigate the association between the rs1333048 polymorphism in the ANRIL gene and periodontitis, with subgroup analyses and a larger number of included studies and participants than previous meta-analyses in this field[16]. The non-significant heterogeneity, results of sensitivity analysis, absence of publication bias, and quality scores of the included studies all support the methodological rigor and robustness of our findings.

The ANRIL gene consists of 19 exons spanning 126.3 kb within the p15/CDKN2B-p16/CDKN2A-p14/ARF gene cluster at the 9p21.3 locus[23]. At the genetic level, our results demonstrated a significant association between the C allele of the rs1333048 polymorphism and periodontitis (P < 0.05), with an increased OR (OR = 1.24). The C allele also showed a notable minor allele frequency (MAF) in the global population (C = 0.442)[24]. These findings corroborate previous reports indicating that the MAF of the C allele was significantly associated with patients diagnosed with periodontitis[22]. Moreover, this mutant allele has also been linked to prostate cancer risk and benign prostatic hyperplasia in Caucasian patients from an Iranian population[25].

Previous findings are inconsistent with these associations. The rs1333048 polymorphism was associated with cancer risk only at the haplotype level, when combined with the mutant allele of rs333045 and the wild-type alleles of rs3977574 and rs10757278[26]. Similar associations have also been reported between other ANRIL gene SNPs and multiple sclerosis[27].

A previous in vitro experiment with vascular smooth muscle cells demonstrated that the ANRIL transcript can affect physiological pathways in a splicing variant-dependent manner, with exon 1 and exon 19 influencing the expression of IL-1A and IL-1RN, respectively[28]. It is well established that IL-1 belongs to the pro-inflammatory cytokine family, and polymorphisms in the IL1A[29] and IL1B[30] genes have been associated with periodontal inflammation. These findings suggest a possible molecular link between ANRIL expression and genetic variations previously implicated in periodontitis.

Bochenek[31] suggested a molecular linkage between ANRIL regulation and periodontitis pathways, identifying that ANRIL transcripts containing exon 13 were correlated with reduced expression of key genes involved in glucose and fatty acid metabolism, with polymorphisms in these genes associated with AgP. This relationship may be mediated by the interaction of ANRIL with polycomb repressive complexes, a group of proteins critical for chromatin remodeling[8] and inflammation[32]. In addition, ANRIL has been shown to interfere with Nuclear Factor kB (NF-kB) pathways[33], where NF-kB functions as a central mediator of cytokine expression and periodontal tissue disruption[34]. Recent findings further demonstrate that ANRIL overexpression can alter the transcription of genes involved in inflammation, NF-κB signaling, type I interferon-mediated signaling, and the innate immune response[35].

The significant association observed in the overall analysis was also evident when evaluating different forms of periodontitis. Although fewer studies focused on CP (n = 3) compared to AgP (n = 8), the results were similar across these clinical forms of the disease (Table 2), both showing a significant association with increased OR values.

The clinical classifications of CP and AgP were dissolved in the current periodontitis classification proposed by Caton et al[36], which unified these two forms under a single category of periodontitis. The current classification could introduce bias in our subgroup analysis evaluating the distinct clinical disease presentations. However, pooled overall analysis resolved this issue.

To further evaluate the influence of confounding factors in the overall analyses, this meta-analysis stratified the studies by participant ethnicity, distinguishing between Caucasian and mixed-race populations. We identified a significant association in the Caucasian group (Table 2). This result is understandable given the previously reported associations between ANRIL gene polymorphisms and periodontitis in Caucasian populations[11].

The rs333048 polymorphism was not significantly associated with periodontitis in the mixed-race population (P > 0.05). This may be explained by two important factors: The limited number of included studies (n = 2) and the increased I² value, which interferes with the statistical model applied on the data. Finally, the mixed-race population in the studies is from Brazil, and Brazilians are characterized by an elevated level of genetic admixture[37].

Although the results from this meta-analysis highlight an association between the rs1333048 polymorphism and periodontitis with complete ethnic evaluation, important limitations still exist. First, no studies were available in the literature assessing the impact on this polymorphism in ANRIL expression levels during periodontitis. Teeuw et al[22] analyzed the correlation between ANRIL rs1333048 polymorphism and C-reactive protein but did not evaluate ANRIL levels. Second, periodontitis is a complex disease whose pathogenesis involves several factors, such as age, sex, oral hygiene, alcohol use, smoking and other genetic variations, which must be considered. In this context, the present meta-analysis did not assess such factors due to the lack of data in the included studies. Third, a comprehensive evaluation of other ethnicities, such as African or Asian populations, could not be performed due to the lack of available studies. This also may explain the non-significant associations observed in the mixed-race population. Fourth, Begg’s and Egger’s linear regression tests revealed no publication bias. However, these tests are better used on meta-analyses with a larger range of included studies[38]. Therefore, the absence of obvious publication bias should be considered with caution. Fifth, the clinical classification of periodontitis has been revised, with a new model proposed that dissolves the previous distinction between CP and AgP[36]. Considering the new classification, both the case-patients included in previous meta-analyses and those in future studies on periodontitis should be interpreted in light of this updated framework. Finally, caution is warranted in interpreting the present results, as reliance on the P value as the gold standard of statistical validity may be misleading, given that statistical significance does not necessarily equate to clinical significance[39]. Therefore, further studies with specific approaches are required to validate these findings in silico and to evaluate this genetic variant as a potential molecular biomarker for the disease, as suggested in previous literature[40-42].

CONCLUSION

In conclusion, this systematic meta-analysis composed by 12 studies and 5489 participants demonstrated a significant association between the rs1333048 polymorphism in the ANRIL gene and periodontitis for both primary clinical forms and the Caucasian population.

ACKNOWLEDGEMENTS

We thank teacher Abilio Borgi for assistance with English language review.