Association of TP53 Polymorphisms with Primary Open-Angle Glaucoma: A Meta-Analysis

Abstract

Purpose.: To offer a comprehensive evaluation of the potential associations of TP53 polymorphisms with primary open-angle glaucoma (POAG) through a systematic review and meta-analysis of candidate genetic association study.

Methods.: Medline, Embase, Science Citation Index, the Cochrane Library, and other databases (up to January 20, 2012) were searched by two investigators independently. Pooled odd ratios (ORs) and 95% confidence interval (CI) were used to assess the strength of the associations between two TP53 polymorphisms (codon 72 in exon 4 and 16 base-pair [bp] insertion in intron 3) and POAG. Statistical analysis was performed with a commercial statistical and data analysis software package.

Results.: Nine independent studies on TP codon 72 (1930 cases and 1463 controls) and four articles on TP intron 3 16-bp insertion (858 cases and 683 controls) were identified. The overall results showed that there was significant association between TP53 codon 72 genotype and POAG risk in the recessive model (OR = 1.31, 95% CI 1.05−1.64, P = 0.017). Also, our analysis suggested that TP53 intron 3 16-bp insertion polymorphism was associated with decreased POAG risk in overall population when examining the contrast of Ins versus Del (OR 0.75, 95% CI = 0.57–0.97, P = 0.031). In subgroup analyses for ethnicity (Caucasian, Asian), we detected the association between codon 72 polymorphism and risk for POAG in Asian populations (recessive model: OR = 1.36, 95% CI 1.03−1.80, P = 0.026) but not in Caucasian populations. However, no significant finding was noted between P53Arg72Pro and risk for open-angle glaucoma either in high tension glaucoma or in normal tension glaucoma. Because of insufficient studies on TP53 16-bp insertion polymorphism, no subgroup analyses were conducted according to ethnicity and glaucoma subtype to detect the effect of this polymorphism on the susceptibility to POAG.

Conclusions.: This meta-analysis showed the evidence that TP53 codon 72 (CC versus CG+GG) and intron 3 16-bp insertion (Ins versus Del) polymorphisms may affect individual susceptibility to POAG. Moreover, stratified analyses that detected the effect of TP53 codon 72 polymorphism seemed to be varied by ethnicity. Given the limited sample size, further investigations are needed to validate the association.

Introduction

As the second leading cause of blindness in the world, glaucoma is characterized by visual field defects, retinal ganglion cell death, and progressive degeneration of the optic nerve. ^1,2 Primary open-angle glaucoma (POAG) is the major type of primary glaucoma in most populations, and affects 67 million individuals worldwide. ^3–5 It is recognized that POAG is a multifactorial disorder involving the role of multiple genes as well as environmental factors. ^6–14 Family studies and mutations in three genes (myocilin [MYOC], optineurin [OPTN], and WD repeat domain 36 [WDR36]) support a high heritability of POAG, thus suggesting a definite genetic basis for POAG. ^15–20 Quite a number of POAG susceptibility genes have been identified. ^21–28 The majority of findings are conflicting, including those for the Tumor Protein 53 (TP53) gene located on the short arm of chromosome 17. ^29,30  

The TP53 (OMIM/191170) is a prototypical tumor suppressor gene encoding a 53-kDa protein (p53) with important functions in cell cycle control, apoptosis, and maintenance of DNA integrity (4–6). ^31,32 The importance of p53 in cell cycle regulation (via gene transcription) and DNA integrity is such that it has been called the “guardian of the genome.” ^33,34 It has been considered as the key regulatory gene for apoptosis, which may be involved in the death of retinal ganglion cells in glaucoma. ^35,36  

In the past years, as a good potential candidate susceptibility gene for glaucoma, TP53 polymorphisms have attracted widespread attention. Although several TP53 polymorphisms have been investigated as risk factors for POAG, by far the most extensively investigated would be two polymorphisms: one is a single nucleotide polymorphism (SNP) in exon 4 of p53 at codon 72, where a cytosine (C; variant allele) for guanine (G) substitution (Arg72Pro); another is a 16 base-pair (bp) insertion/deletion polymorphism located within TP53 intron 3. However, whether TP53 gene polymorphisms may contribute to the pathogenesis of POAG is still strongly debated. Lin et al. ³⁷ first reported the p53 codon 72 polymorphism to be associated with POAG. However, the association between such polymorphisms in p53 and POAG remains controversial according to previous reports. ^38–44  

To date, no meta-analysis has been conducted to elevate the association of the polymorphisms of TP53 with POAG. Thus, it stimulated us to conduct a meta-analysis based on a total of nine independent studies, which may provide the evidence for the association of p53 polymorphisms with POAG susceptibility. 

Methods

Publication Search

A systematic literature search was performed for articles regarding TP53 polymorphisms with POAG. The MEDLINE, EMBASE, Science Citation Index, the Cochrane Library, and Chinese National Knowledge Infrastructure were simultaneously used with key terms “TP53” or “tumor suppressor protein gene”; “P53”; “polymorphism,” “SNP,” or “single nucleotide polymorphism”; “variation” or “mutation”; “glaucoma,” “primary open-angle glaucoma,” or “POAG”; or “high tension glaucoma” or “normal tension glaucoma.” Relevant publications were examined for references until no further studies were found. The electronic databases were last searched on January 20th, 2012. 

Inclusion Criteria

The following inclusion criteria were used to select literatures for the meta-analysis: (1) case-control, nested case-control, or cohort studies; (2) description of the association of P53 polymorphisms with POAG; and (3) the number of cases and controls, number of different genotypes for p53 codon 72, intron 3 16-bp insertion in cases and controls, as well as the information that can help infer the results in the published studies. 

Data Extraction

Two observers (YTG, HTZ) independently abstracted data from all eligible publications onto paper data collection forms. Two reviewers were blinded to the details (title, author, and academic address) of these studies during assessment. Disagreements were resolved by discussion or consensus involving a third reviewer (XC) when required. The following items were collected from each study: first author's surname, year of publication, statistical data, ethnicity, total number of cases and controls, as well as numbers of cases and controls for each TP53 genotypes, respectively. 

Statistical Analysis

All statistical analyses were performed using a commercial statistical software package (Stata Statistics Software, Version 10.0; StataCorp LP, College Station, TX). Two-sided P values < 0.05 were considered statistically significant. 

Hardy–Weinberg equilibrium (HWE) in controls was calculated again in our meta-analysis. The χ ² goodness of fit was used to test deviation from HWE (significant at the 0.05 level). 

The effect measure of choice was a pooled odds ratio (OR) with its corresponding 95% confidence interval (CI). Heterogeneity assumption was checked by Q-test. A value of P < 0.10 for the Q-test indicated lack of heterogeneity among the studies. Based on Q-test value, two models of meta-analysis were applied for dichotomous outcomes: A fixed-effects model, using the Mantel–Haenszel (M-H) method, was used to calculate the pooled ORs when the Q-test value was ≥0.1. By contrast, a random-effects (DerSimonian and Laird [D+L]) model was utilized if the Q-test value was <0.1. First, we compared allele frequencies (Allele C versus Allele G; Ins versus Del) between cases and controls. Then, we examined TP53 genotypes using additive (CC versus GG; Ins/Ins versus Del/Del), recessive (CC versus GC+ GG; Ins/Ins versus Del/Ins+Del/Del), and dominant (GC + CC versus GG; Del/Ins+Ins/Ins versus Del/Del) genetic models for C and Ins alleles. Furthermore, subgroup analyses were performed by ethnicity (Asian, Caucasian) and glaucoma subtypes (high tension glaucoma [HTG]; normal tension glaucoma [NTG]). 

Finally, publication bias was assessed by performing Begg's funnel plots qualitatively and evaluated by Egger's test quantitatively (P < 0.05 was considered representative of statistically significant publication bias). 

Results

Literature Search and Characteristics

The initial search yielded 228 articles. Based on the title, the content of the abstract, and key words, 218 studies were excluded. Ten articles were reviewed in their entirety. One arm had to be excluded because of no useful data available. Finally, nine studies that met our inclusion criteria were included in this review. ^37–45 The flow chart of literature search is shown in Figure 1. Nine studies ^37–45 were included in the meta-analysis of p53 codon 72 genotype (1930 cases, 1463 controls), of which, in two studies, ^43,44 the distribution of the genotypes in the control group were not in HWE (Fisher's exact test, P < 0.05). Four case-control studies ^38–40,42 were included in the meta-analysis of intron 3 16-bp insertion genotype (858 cases, 683 controls). Similarly, one study's distribution of the genotypes in the control group was not in HWE. ⁴² For the meta-analysis of TP53 codon 72, four studies on Caucasians, ^40,41,43,44 four studies on Asians, ^37,38,42,45 and one study on Brazilians were included. ⁴⁴ Also four studies ^40–42,45 on HTG and NTG were included in the meta-analysis of p53 codon 72. For the TP53 intron 3 16-bp insertion, subgroup analyses included that two Caucasian studies ^39,40 and two Asian studies ^38,42 for ethnicity as well as two studies ^40,42 for subtype of POAG. Detailed study characteristics are summarized in Table 1. 

Figure 1. 

 

Flow chart of literature search and study selection.

Table 1.  

 

Main Characteristics of Studies Included in This Meta-Analysis

P53 Codon72
Author	Year	Ethnicity	HWE	Cases			Control
				GG	GC	CC	GG	GC	CC
Lin et al.	2002	Asian	0.96	12	26	20	25	26	8
Acharva et al.	2002	Asian	0.98	23	30	14	30	57	25
Dimasi et al.	2005	Caucasian	0.49	296	186	41	109	57	12
Mabuchi et al.	2009	Asian	0.95	177	197	51	83	83	23
Daugherty et al.	2009	Caucasian	0.87	124	55	12	82	72	13
Saglar et al.	2009	Caucasian	0.02	19	44	12	41	69	9
Silva et al.	2009	Brazilian	0.0003	24	78	2	18	40	0
Fan et al.	2010	Asian	0.51	100	176	121	55	108	38
Blanco-Marchite et al.	2011	Caucasian	0.57	148	106	14	252	111	17
HTG
Dimasi et al.	2005	Caucasian	0.49	158	101	24	109	57	12
Mabuchi et al.	2009	Asian	0.95	85	102	25	83	83	23
Daugherty et al.	2009	Caucasian	0.88	89	40	10	82	72	13
Fan et al.	2010	Asian	0.51	64	114	74	55	108	38
NTG
Dimasi et al.	2005	Caucasian	0.49	29	28	5	109	57	12
Mabuchi et al.	2009	Asian	0.95	92	95	26	83	83	23
Daugherty et al.	2009	Caucasian	0.88	36	15	1	82	72	13
Fan et al.	2010	Asian	0.51	22	43	34	55	108	38
P53 Intron 3 16-bp Insertion
				Del/Del	Del/Ins	Ins/Ins	Del/Del	Del/Ins	Ins/Ins
Acharva et al.	2002	Asian	0.14	56	11	0	67	33	0
Mabuchi et al.	2009	Asian	—	425	0	0	189	0	0
Daugherty et al.	2009	Caucasian	0.51	145	42	4	117	43	7
Blanco-Marchite et al.	2011	Caucasian	0.98	135	38	2	167	55	5
HTG
Mabuchi et al.	2009	Asian	—	212	0	0	189	0	0
Daugherty et al.	2009	Caucasian	0.51	107	29	3	117	43	7
NTG
Mabuchi et al.	2009	Asian	—	213	0	0	189	0	0
Daugherty et al.	2009	Caucasian	0.51	38	14	0	117	43	7

 HWE, Hardy–Weinberg equilibrium; HTG, high tension glaucoma; NTG, normal tension glaucoma; —, data not available.

Meta-Analysis Results

The overall analysis investigating the recessive model for the C allele (CC versus CG+GG) showed significant association between TP53 codon 72 polymorphism and increased open-angle glaucoma risk (OR = 1.31, 95% CI 1.05−1.64, P = 0.017; Fig. 2), although no evidence of associations was detected in the allelic, additive, and dominant models (allelic model: OR = 1.10, 95% CI 0.99–1.22, P = 0.054; additive model: OR = 1.23, 95% CI 0.97–1.56, P = 0.079; dominant model: OR = 1.06, 95% CI 0.94–1.19, P = 0.317) On the other hand, the association between the 16-bp insertion and decreased risk for POAG relative to the 16-bp deletion (Ins versus Del) revealed a significant association (OR 0.75, 95% CI = 0.57–0.97, P = 0.031; Fig. 3). No significant associations were observed between TP53 intron 3 16-bp insertion genotype and risk for POAG in the additive (OR = 0.49, 95% CI 0.18−1.32, P = 0.16), recessive (OR = 0.51, 95% CI 0.19−1.37, P = 0.18), and dominant models (OR = 1.01, 95% CI 0.87−1.17, P = 0.91). Because the Q-test of heterogeneity among studies was nonsignificant in all genetic models, a fixed-effects model was used (Table 2). 

Figure 2. 

 

Forest plot of the association between TP53 codon 72 polymorphism (CC versus CG+GG) and risk for POAG in order of publication year. The size of the square indicates the relative weight of each study. Bars, 95% CI.

Figure 3. 

 

Forest plot of the association between TP53 16-bp insertion polymorphism (Ins versus Del) and risk for POAG in order of publication year. The size of the square indicates the relative weight of each study. Bars, 95% CI.

Table 2.  

 

Results of Meta-Analysis for TP Polymorphisms and Risk of Primary Open-Angle Glaucoma

Comparisons	Number of Studies	OR	95% CI	P Value	Heterogeneity		Effects Model	Egger's Test
					I 2	P Value		P > | t |
P53 Codon72
C vs. G	9	1.1	1.00–1.22	0.054	0.1264	0.125	Fixed	0.979
Caucasian	4	1.07	0.82–1.40	0.609	9.11%	0.028	Random	0.609
Asian	4	1.11	0.96–1.27	0.154	3.92%	0.27	Fixed	0.73
HTG	4	1.05	0.91–1.20	0.535	4.63%	0.201	Fixed	0.225
NTG	4	1.03	0.78–1.38	0.817	7.30%	0.063	Random	0.5
CC vs. GG	9	1.23	0.98–1.56	0.079	7.97%	0.437	Fixed	0.7
Caucasian	4	1.2	0.81–1.77	0.36	3.67%	0.299	Fixed	0.768
Asian	4	1.24	0.93–1.66	0.156	3.84%	0.279	Fixed	0.763
HTG	4	1.16	0.85–1.58	0.361	1.53%	0.675	Fixed	0.278
NTG	4	1.13	0.78–1.65	0.519	3.96%	0.266	Fixed	0.5
CC vs. CG+GG	9	1.31	1.05–1.63	0.017	7.99%	0.434	Fixed	0.96
Caucasian	4	1.19	0.82–1.74	0.365	2.42%	0.49	Fixed	0.56
Asian	4	1.37	1.03–1.80	0.026	5.06%	0.168	Fixed	0.964
HTG	4	1.26	0.94–1.69	0.125	2.14%	0.544	Fixed	0.22
NTG	4	1.25	0.88–1.78	0.22	4.89%	0.18	Fixed	0.252
CC+CG vs. GG	9	1.06	0.94–1.19	0.317	8.75%	0.364	Fixed	0.906
Caucasian	4	1.05	0.80–1.39	0.716	7.34%	0.062	Random	0.5
Asian	4	1.04	0.88–1.24	0.636	1.30%	0.73	Fixed	0.653
HTG	4	1	0.85–1.18	0.999	3.41%	0.332	Fixed	0.32
NTG	4	1.02	0.83–1.25	0.17	4.15%	0.246	Fixed	0.63
P53 Intron 3 16-bp Insertion
Ins vs. Del	4	0.75	0.57–0.97	0.031	1.55%	0.46	Fixed	0.169
Ins/Ins vs. Del/Del	4	0.49	0.18–1.32	0.156	0	0.96	Fixed	—
Ins/Ins vs. Del/Del+Del/Ins	4	0.51	0.19–1.37	0.181	0	0.97	Fixed	—
Ins/Ins+Del/Ins vs. Del/Del	4	1.01	0.87–1.17	0.91	0	1	Fixed	0.579

 —, data not available because of limited studies.

Next, stratified analyses by ethnicity were performed between TP53 codon 72 polymorphism and POAG risk. A significant association between TP53 Arg72Pro polymorphism and POAG susceptibility was found in Asian populations in the recessive model (OR = 1.36, 95% CI 1.03−1.80, P = 0.026), but not in Caucasian populations (Fig. 4A). Furthermore, we investigated the effect of the TP53 codon 72 genotype on the susceptibility to subtypes of POAG. Overall, no evidence of association was observed in any genetic model between TP53 Arg72Pro polymorphism and risk of NTG and HTG (Fig. 4B). Because each subgroup of TP53 16-bp insertion polymorphism studies included only two studies, the analysis was not performed and solid results need to be obtained from more large-sample and better-designed studies. 

Figure 4. 

 

Subgroup analyses of the association between TP53 codon 72 and risk for POAG in order of publication year. The size of the square indicate the relative weight of each study. Bars, 95% CI (A) Subgroup analysis stratified by ethnicity. (B) Subgroup analysis stratified by glaucoma subtype.

Publication Bias

Publication biases were assessed by Begg's funnel plot qualitatively and Egger's test quantitatively. Neither Begg's funnel plot nor Egger's test detected any obvious evidence of publication bias in the overall and subgroup analyses for all genetic models (Fig. 5; data available in Table 2). 

Figure 5. 

 

Begg's funnel plot of TP53 codon 72 polymorphism and POAG for overall CC versus CG+GG. Each circle represents a separate study for the indicated association, and its size is proportional to the sample size of each study. The study by Silva et al. was excluded because of no homozygous subjects (CC) in the control group.

Discussion

Primary open-angle glaucoma (POAG) is considered to be a multifactorial disease, and is estimated to have a significant heritable component. ^46,47 TP53 protein product is a core component to control the apoptosis and play a vital role in the regulation of the cell cycle. ⁴⁸ Thus, the TP53 polymorphisms, one of the most widely investigated polymorphisms in genetic epidemiology, have been thought to constitute a good candidate genetic risk factor for POAG. ³⁷ Since the identification of the TP53 polymorphisms, a number of studies have investigated the genetic effect of the TP53 polymorphisms on POAG susceptibility with conflicting results. Meta-analysis, as a powerful statistical method, can provide a quantitative approach for pooling the variant results on the same topic to estimate and explain their diversity. ^49,50 This led us to conduct this meta-analysis of nine published case-control studies, which may help us to distinguish the truth from the false, to explore a more robust estimate of the effects of these polymorphisms on POAG. 

The main finding of the pooled analyses showed that TP53 Arg72Pro polymorphism was a likely risk factor for POAG; meanwhile, intron 3 16-bp insertion polymorphism was detected to be associated with reduced POAG risk. However, it should be noted that genetic distributions of the controls in two studies ^43,44 of Arg72Pro group and in one study ⁴² of 16-bp insertion group deviated from Hardy–Weinberg equilibrium, which may have been the genotyping errors or selection bias in control and/or population stratification. Therefore, as recommended by Attia and colleagues, ⁵¹ we conducted the meta-analysis again when these studies were removed. The results indicated that estimates before or after the deletion of these studies were similar, suggesting high stability of the meta-analysis results with little effect of these studies. 

The results of many studies have represented that the ethnic differences may affect genetic predisposition to POAG. ^37,40 Subgroup analysis on different ethnicity was performed. Our data showed that the TP53 codon 72 polymorphisms from studies of Asian individuals were significantly associated with POAG. By contrast, it was not found in Caucasian populations. This indicates a possible role of ethnic difference in genetic background and the environment they live in. Moreover, the discrepancy may arise because studies with small sample sizes may be underpowered to detect a slight effect or may have generated a fluctuated risk estimate. 

Furthermore, the studies reported by Daugherty, ⁴⁰ Fan, ⁴⁵ Mabuchi, ⁴² and Dimasi ⁴¹ investigated the effect of this polymorphism on subtypes of open-angle glaucoma. Some of them ⁴⁰ reported that higher frequencies of the Arginine allele was observed in POAG cases with NTG compared with HTG, suggesting a genetic mechanism favoring apoptosis would have a greater role in NTG. However, no evidence was found that the association was significant in NTG or HTG, probably because of small size samples and limited trials. For an insufficient number of studies, subgroup analyses could not be conducted on different races and subtypes for TP53 intronic insertion polymorphisms. Further analysis should be performed in more large-scale cohorts or case-control studies. 

Although no obvious publication bias showed some limitations in our study should be addressed and the results should be interpreted with caution. First, controls were not uniformly defined. Thus, some inevitable selection bias might exist in the results and they may not be representative of the general population. Second, this meta-analysis was limited by the number of cases and controls as well as small sample size, especially in subgroup analysis. Third, our results were based on unadjusted estimates, whereas a more precise analysis of the various groups should be conducted according to other factors. Fourth, genotyping methods were different among these studies, which might have affected the results. This discrepancy between genotyping methods highlights the need for implementing rigorous quality control procedures in future studies. Fifth, the existing studies are the lack of information about potential gene–gene or gene–environment interactions. Given that the role of several environment factors in the pathogenesis of open-angle glaucoma is established, further research should be performed in this direction. 

In conclusion, the results of this meta-analysis suggest that TP53 codon 72 polymorphism is associated with increased risk for POAG in the recessive model and 16-bp insertion polymorphism has the association with a statistically significant decrease in POAG susceptibility under the allelic model but not in other models. Meanwhile, it is worthwhile to note that TP53 codon 72 polymorphism may be involved in the pathogenesis of POAG in Asians but not in Caucasians. Due to the limitations shown earlier, well-designed studies with large sample sizes are warranted to confirm our findings. 

Footnotes

 Supported in part by Major and Key Programs of Chinese National Natural Science Foundation Grant 30730099, and “Project 211,” The Innovation Fund for Graduate Students of Tianjin Medical University.

Footnotes

⁴  These authors contributed equally to the work presented here and should therefore be regarded as equivalent authors.

Footnotes

 Disclosure: Y. Guo, None; H. Zhang, None; X. Chen, None; X. Yang, None; W. Cheng, None; K. Zhao, None