Genetic association of ANGPT2 with primary open-angle glaucoma.

BACKGROUND: To determine the association of the ANGPT2 gene with primary open-angle glaucoma (POAG) in Chinese.
METHODS: Six single-nucleotide polymorphisms (SNPs) in ANGPT2 (rs2515487, rs2922869, rs13255574, rs4455855, rs13269021, and rs11775442) were genotyped in a total of 2601 study subjects from two cohorts. One is a Hong Kong Chinese cohort of 484 high tension glaucoma (HTG) and 537 normal tension glaucoma (NTG) patients, and 496 non-glaucoma control subjects. Another cohort is a Shantou Chinese cohort of 403 HTG and 135 NTG patients, and 543 non-glaucoma control subjects. Subgroup analysis by sex was conducted. Outcomes from different cohorts were combined for meta-analysis.
RESULTS: The association of SNP rs11775442 with NTG in the Hong Kong cohort [P = 0.0498, OR = 1.24, 95% confidence interval (CI) 1.00-1.55] after adjusting for age and sex did not reach statistical significance after Bonferroni correction. Other SNPs were not significantly associated with NTG, HTG and POAG in individual cohort or in the combined analyses (P > 0.05). In the subgroup analysis by sex, SNP rs13269021 in the Shantou cohort, but not in the Hong Kong cohort, was significantly associated with NTG in males (P = 0.0081, OR = 1.67, 95% CI: 1.14-2.43) but not in females (P = 0.874). In the combined analyses by sex, no SNPs were significantly associated with NTG, HTG and POAG.
CONCLUSIONS: In the subgroup analysis by sex, a significant association was shown in SNP rs13269021 with NTG in Shantou males, but not in Hong Kong males. Further studies are needed to verify the association between ANGPT2 locus (rs13269021) and NTG in Chinese males.

Background

Glaucoma is a leading cause of irreversible blindness worldwide [1]. In 2010, glaucoma caused 2.1 million (6.5%) out of a total 32.4 million blindness globally [2]. Primary open-angle glaucoma (POAG) is the most common type of glaucoma. According to the highest detected intraocular pressure (IOP), POAG can be classified into high tension POAG (IOP > 21 mmHg) and normal tension POAG (IOP ≤ 21 mmHg).

POAG is a multifactorial progressive optic neuropathy. Both genetic and environmental factors are related to the etiology of POAG. Until now, four genes have been reported to be causative or probably causative of POAG: myocilin (MYOC, also called GLC1A) [3, 4], optineurin (OPTN, also called GLC1E) [5, 6], neurotrophin 4 (NTF4) [7, 8] and tank-binding kinase 1 (TBK1, also called GLC1P) [9-11].

IOP is a major risk factor for POAG and lowering IOP is currently the only intervention that has been proven to be able to delay the progression of glaucoma in human [12, 13]. Trabecular meshwork/Schlemm’s canal (TM/SC) conventional aqueous humor (AH) outflow pathway plays a key role in the regulation of IOP [14]. Previous studies found that the SC tissues show features of lymphatic vessels including expressing lymphatic regulator PROX1, lymphangiogenic growth factor VEGF-C and its receptor VEGFR-3, suggesting lymphangiogenesis induction could be developed as a novel treatment of glaucoma [15-18]. Tie receptors and their angiopoietin (Angpt) ligands are key regulators of vascular morphogenesis and homeostasis [19]. There are six members in the Tie-Angpt system: TIE1, TIE2, ANGPT1, ANGPT2, ANGPT3 and ANGPT4. ANGPT2 acts as an agonist and an antagonist to TIE2 in a context-dependent manner [20]. Deletion of Angpt1/Angpt2 or Tie2 in adult mice severely impaired SC integrity and transcytosis, leading to hallmarks of POAG including elevated IOP, retinal neuron damages, and impairment of retinal ganglion cell functions [21]. A POAG genome-wide meta-analysis in 2021 identified the association between an intronic variant rs2515437 in ANGPT2 and POAG [22]. Another ANGPT2 SNP, rs76020419, has been reported to be associated with IOP [23, 24]. Therefore, we hypothesize that ANGPT2 is a potential gene associated with POAG. However, both rs2515437 and rs76020419 are monomorphic in East Asians. Therefore, we aim to study other POAG-associated variants in this gene in East Asians, especially in our local population. Previously, we reported six ANGPT2 SNPs that were associated with neovascular age-related macular degeneration (nAMD) and polypoidal choroidal vasculopathy (PCV) [25]. As nAMD and PCV are related to retinal vessel abnormalities, we hypothesize these six ANGPT2 SNPs may also associate with POAG, in which SC is also a lymphatic-like vessel. In this study, we performed a SNP association analysis to determine the association of these six ANGPT2 SNPs with POAG.

Methods

Study participants

This study involved 2601 unrelated participants from two independent Chinese cohorts. One cohort is a Hong Kong Chinese cohort of 484 high tension glaucoma (HTG) and 537 normal tension glaucoma (NTG) patients, and 496 non-glaucoma control subjects. Another cohort is a Shantou Chinese cohort of 403 HTG and 135 NTG patients, and 543 non-glaucoma control subjects. All participants are Han Chinese, recruited from the eye clinics of the Prince of Wales Hospital, the Hong Kong Eye Hospital and the Joint Shantou International Eye Center. The study protocol was approved by the institutional ethics committee at the respective institutions (KC/KE-18-0045/ER-3). Written informed consent was obtained from each participant. The study procedures were performed in accordance with the tenets of the Declaration of Helsinki.

Diagnosis was based on meeting all the following criteria: exclusion of secondary causes of glaucoma (steroid-induced glaucoma, neovascular glaucoma, uveitis or trauma); anterior chamber angle open (grade III or IV on gonioscopy); characteristic optic disc changes (vertical cup-to-disc ratio > 0.5, disc hemorrhage, or thin/notched neuroretinal rim); and characteristic visual field changes with reference to the Anderson’s criteria for minimal abnormality in glaucoma [26]. IOP was measured by using applanation tonometry and visual field was evaluated by a perimeter (Humphrey Field Analyzer; Carl Zeiss Meditec, Dublin, CA) using the Glaucoma Hemifield Test. Patients whose highest recorded IOP lower than 21 mmHg without medication were regarded as NTG patients. Patients whose highest recorded IOP higher than or equal to 21 mmHg without medication were regarded as HTG patients. Patients with congenital glaucoma were excluded. People who attended the clinic for conditions other than glaucoma, including mild senile cataract, floaters, mild refractive errors and itchy eyes were recruited as non-glaucoma control subjects.

Single-nucleotide polymorphisms (SNPs) selection and genotyping

Six SNPs in ANGPT2 (rs2515487, rs2922869, rs13255574, rs4455855, rs13269021, and rs11775442) were reported as suggestive disease-association in nAMD and PCV [25]. These six SNPs are independent of each other. They were selected for genotyping assays in POAG patients from the Hong Kong and Shantou cohorts. Genomic DNA from peripheral blood was extracted using a QIAamp Blood Kit (Qiagen, Hilden, Germany) according to the manufacturer’s protocol. SNPs were genotyped using TaqMan genotyping assays (Applied Biosystems, Foster City, CA, USA) with a Roche LightCycle 480 Real-Time PCR System (Roche Diagnostics; Basel, Switzerland) following the manufacturer’s instructions.

Statistical analysis

Hardy-Weinberg equilibrium (HWE) of each SNP in the non-glaucoma controls was assessed using the χ2 test in PLINK (v.1.07; available in the public domain at http://pngu.mgh.harvard.edu/purcell/plink/). Allelic and genotypic distributions were compared by the χ2 test between cases and controls among different study cohorts. The odds ratio (OR) and 95% confidence intervals (CI) for each SNP were calculated. Logistic regression analysis was used to evaluate the genetic effects of the SNPs adjusted for age and sex.

To combine the data from the two study cohorts, we adopted the Mantel-Haenszel χ2 test to obtain the combined ORs and 95% CIs for all SNPs. The test was performed using Review Manager (RevMan, version 5.2; The Cochrane Collaboration, Copenhagen, Denmark).

In this study, we adopted the Bonferroni method to correct the P values in multiple testing. A final P value of less than 0.0083 (0.05/6, where 6 was the number of SNPs included in data analysis) would be required to conclude a significant disease association.

Results

Demographics and quality control outcomes

Table 1 shows the demographics of the POAG patients and non-glaucoma controls. Mean ages and sex ratios were statistically different between POAG cases and non-glaucoma control group in the two cohorts (P < 0.05; Table 1). Therefore, age and sex were adjusted in logistic regressions. In the Shantou cohort, all SNPs conformed to HWE (P > 0.05) in the non-glaucoma control group with call rates ≥ 98%. In the Hong Kong group, the rs2922869 showed deviation from HWE (P < 0.05) and thus was excluded from further analysis.

Table 1

Demographics of study subjects

	POAG	HTG	NTG	Non-glaucoma control	aP
Hong Kong
N	1021	484	537	496
Age (years)	62.0 ± 14.0	62.0 ± 15.3	63.2 ± 12.8	70.2 ± 10.8	< 0.05
Sex (male %)	56.10%	61.00%	51.80%	37.10%	< 0.05
Shantou
N	541	403	135	543
Age (years)	55.9 ± 18.6	54.0 ± 19.4	61.6 ± 14.6	74.4 ± 6.9	< 0.05
Sex (male %)	68.40%	71.70%	58.50%	52.10%	< 0.05

HTG = high tension glaucoma; NTG = normal tension glaucoma; POAG = primary open-angle glaucoma

aP value: mean age between cases and controls was compared using independent t-test; sex proportion between cases and controls was compared with the χ test

Association of ANGPT2 with POAG

There was no significant association of all the six SNPs with NTG, HTG and POAG in individual cohorts or in the combined analyses after Bonferroni correction (P > 0.0083, Table 2). Only one SNP, rs11775442, showed a borderline association with NTG in the Hong Kong cohort at the level of P value < 0.05 (P = 0.0498, OR = 1.24, 95% CI 1.00–1.55) after adjusting for age and sex (Table 2).

Table 2

Association of SNPs from ANGPT2 in NTG, HTG and POAG adjusted for age and sex

SNP	Nucleotide change	Risk allele	RAF				HTG vs. Control				NTG vs. Control				POAG vs. Control
			HTG (n = 484)	NTG (n = 537)	POAG (n = 1021)	Control (n = 496)	Pnon-adjusted	Non-adjusted OR (95% CI)	Padjusted	Adjusted OR (95% CI)	Pnon-adjusted	Non-adjusted OR (95% CI)	Padjusted	Adjusted OR (95% CI)	Pnon-adjusted	Non-adjusted OR (95% CI)	Padjusted	Adjusted OR (95% CI)
Hong Kong
rs2515487	c.289-10440C>A	A	0.240	0.238	0.239	0.239	0.950	1.01 (0.82–1.24)	0.815	1.03 (0.82–1.28)	0.977	1.00 (0.81–1.22)	0.761	1.03 (0.84–1.28)	0.986	1.00 (0.84–1.20)	0.704	1.04 (0.86–1.25)
rs13255574	c.289-8669C>T	C	0.789	0.789	0.789	0.762	0.148	1.17 (0.95–1.45)	0.367	1.11 (0.89–1.38)	0.148	1.16 (0.95–1.43)	0.118	1.19 (0.96–1.47)	0.093	1.17 (0.97–1.40)	0.173	1.14 (0.95–1.37)
rs4455855	c.289-7422T>A	A	0.300	0.29	0.295	0.293	0.740	1.03 (0.85–1.26)	0.833	1.02 (0.83–1.26)	0.850	0.98 (0.81–1.19)	0.739	1.03 (0.85–1.26)	0.943	1.01 (0.85–1.19)	0.882	1.01 (0.85–1.21)
rs13269021	c.289-6755G>T	G	0.543	0.54	0.541	0.523	0.389	1.08 (0.90–1.29)	0.715	1.04 (0.86–1.25)	0.441	1.07 (0.90–1.27)	0.446	1.07 (0.90–1.29)	0.348	1.08 (0.92–1.25)	0.542	1.05 (0.90–1.23)
rs11775442	c.289-6350A>G	A	0.782	0.796	0.789	0.764	0.354	1.11 (0.89–1.37)	0.558	1.07 (0.86–1.33)	0.083	1.20 (0.98–1.48)	0.0498*	1.24 (1.00–1.55)	0.119	1.16 (0.96–1.38)	0.149	1.15 (0.95–1.39)
Shantou
rs2515487	c.289-10440C>A	A	0.286	0.282	0.285	0.286	0.981	1.00 (0.82–1.23)	0.995	1.00 (0.76–1.31)	0.908	0.98 (0.73–1.32)	0.971	1.01 (0.71–1.43)	0.989	1.00 (0.83–1.21)	0.916	0.99 (0.78–1.26)
rs2922869	c.289-9785T>C	T	0.729	0.727	0.727	0.704	0.251	1.13 (0.92–1.39)	0.439	1.12 (0.85–1.47)	0.458	1.12 (0.83–1.52)	0.529	1.12 (0.78–1.60)	0.247	1.12 (0.92–1.36)	0.407	1.11 (0.87–1.42)
rs13255574	c.289-8669C>T	C	0.785	0.782	0.782	0.763	0.276	1.13 (0.91–1.42)	0.320	1.16 (0.86–1.57)	0.530	1.11 (0.80–1.54)	0.607	1.11 (0.75–1.63)	0.297	1.12 (0.91–1.37)	0.380	1.13 (0.86–1.46)
rs4455855	c.289-7422T>A	A	0.323	0.308	0.319	0.330	0.735	0.97 (0.79–1.18)	0.776	0.96 (0.74–1.25)	0.503	0.91 (0.68–1.21)	0.633	0.92 (0.66–1.29)	0.576	0.95 (0.79–1.14)	0.558	0.93 (0.74–1.18)
rs13269021	c.289-6755G>T	G	0.584	0.626	0.594	0.500	0.354	1.09 (0.91–1.32)	0.217	1.17 (0.91–1.49)	0.059	1.31 (0.99–1.73)	0.183	1.25 (0.90–1.72)	0.137	1.14 (0.96–1.36)	0.142	1.18 (0.95–1.47)
rs11775442	c.289-6350A>G	A	0.808	0.788	0.802	0.796	0.535	1.07 (0.86–1.35)	0.357	1.15 (0.85–1.55)	0.803	0.96 (0.70–1.32)	0.915	1.02 (0.70–1.49)	0.734	1.04 (0.84–1.28)	0.586	1.08 (0.83–1.40)

CI = confidence interval; HTG = high tension glaucoma; NTG = normal tension glaucoma; OR = odds ratio; POAG = primary open-angle glaucoma; RAF = risk allele frequency

Pnon-adjusted = P values derived from the χ2 test

Padjusted = P values derived from logistic regression models after adjusting for age and sex

*P value < 0.05

In the subgroup analysis by sex, the top association was shown in SNP rs13269021 with NTG in Shantou males (P = 0.0081, OR = 1.67, 95% CI: 1.14–2.43; Table 3) but not in females (P = 0.874). With regards to the glaucoma subtypes, P values of this SNP were 0.040 in POAG (OR = 1.27) and 0.197 in HTG (OR = 1.17), which could not withstand the Bonferroni correction. Other associations were also non-significant after Bonferroni correction (P > 0.0083). Some of the SNP associations showed differences between male and female subgroups. In the Hong Kong cohort, the associations of rs13255574 in HTG (P = 0.010) and POAG (P = 0.017) in males could not withstand the Bonferroni correction, while P values of the associations were high in females i.e., HTG and POAG (P = 0.605 and 0.824, respectively). In the Shantou cohort, no SNP was significantly associated with NTG, HTG and POAG in both sexes. The meta-analysis of Hong Kong and Shantou subjects showed a slight difference between the sexes. The association of rs13269021 in POAG was still non-significant in meta-analysis (P = 0.020 in males and 0.94 in females, Fig. 1). In the combined analyses by sex, no SNP was significantly associated with NTG, HTG and POAG (P > 0.05).

Table 3

Association of SNPs from ANGPT2 in females and males with NTG, HTG and POAG

SNP	Nucleotide change	Risk allele	RAF								HTG vs. Control				NTG vs. Control				POAG vs. Control
			HTG	HTG	NTG	NTG	POAG	POAG	Control	Control	Pmale	Male OR (95% CI)	Pfemale	Female OR (95% CI)	Pmale	ORmale (95% CI)	Pfemale	ORfemale (95% CI)	Pmale	ORmale (95% CI)	Pfemale	ORfemale (95% CI)
			Male	Female	Male	Female	Male	Female	Male	Female
			(n = 295)	(n = 189)	(n = 278)	(n = 259)	(n = 573)	(n = 448)	(n = 184)	(n = 312)
Hong Kong
rs2515487	c.289-10440C>A	A	0.238	0.243	0.212	0.266	0.226	0.257	0.223	0.248	0.586	1.09 (0.80–1.49)	0.857	0.97 (0.72–1.31)	0.702	0.94 (0.68–1.29)	0.488	1.10 (0.84–1.43)	0.914	1.02 (0.77–1.35)	0.714	1.05 (0.83–1.32)
rs13255574	c.289-8669C>T	C	0.805	0.765	0.781	0.797	0.793	0.784	0.734	0.779	0.010*	1.50 (1.10–2.04)	0.605	0.92 (0.68–1.25)	0.101	1.29 (0.95–1.75)	0.448	1.12 (0.84–1.49)	0.017*	1.39 (1.06–1.83)	0.824	1.03 (0.80–1.32)
rs4455855	c.289-7422T>A	A	0.307	0.290	0.29	0.290	0.298	0.29	0.272	0.306	0.247	1.19 (0.89–1.58)	0.588	0.93 (0.70–1.23)	0.556	1.09 (0.81–1.47)	0.544	0.92 (0.72–1.19)	0.327	1.14 (0.88–1.48)	0.492	0.92 (0.74–1.16)
rs13269021	c.289-6755G>T	G	0.553	0.527	0.545	0.535	0.549	0.531	0.511	0.530	0.207	1.18 (0.91–1.54)	0.910	0.99 (0.76–1.27)	0.307	1.15 (0.88–1.49)	0.885	1.02 (0.81–1.29)	0.201	1.17 (0.92–1.47)	0.971	1.00 (0.82–1.23)
rs11775442	c.289-6350A>G	A	0.781	0.782	0.784	0.808	0.783	0.797	0.767	0.766	0.461	1.12 (0.82–1.53)	0.555	1.10 (0.81–1.49)	0.406	1.14 (0.83–1.56)	0.085	1.29 (0.97–1.71)	0.381	1.13 (0.86–1.49)	0.146	1.20 (0.94–1.54)
Shantou
rs2515487	c.289-10440C>A	A	0.295	0.263	0.295	0.264	0.296	0.262	0.298	0.272	0.909	0.99 (0.76–1.27)	0.808	0.96 (0.66–1.38)	0.936	0.98 (0.67–1.45)	0.864	0.96 (0.61–1.52)	0.937	0.99 (0.78–1.26)	0.746	0.95 (0.69–1.30)
rs2922869	c.289-9785T>C	T	0.712	0.770	0.721	0.736	0.713	0.757	0.713	0.695	0.980	1.00 (0.77–1.30)	0.037	1.49 (1.02–2.16)	0.849	1.04 (0.70–1.55)	0.383	1.23 (0.77–1.98)	0.999	1.00 (0.78–1.29)	0.051	1.37 (1.00–1.88)
rs13255574	c.289-8669C>T	C	0.771	0.821	0.799	0.759	0.775	0.798	0.765	0.760	0.809	1.04 (0.78–1.37)	0.082	1.44 (0.95–2.16)	0.394	1.22 (0.77–1.91)	0.98	0.99 (0.62–1.59)	0.668	1.06 (0.81–1.39)	0.226	1.23 (0.88–1.72)
rs4455855	c.289-7422T>A	A	0.339	0.282	0.346	0.255	0.341	0.271	0.340	0.319	0.960	0.99 (0.78–1.27)	0.300	0.83 (0.58–1.18)	0.889	1.03 (0.71–1.49)	0.193	0.74 (0.46–1.17)	0.981	1.00 (0.79–1.27)	0.128	0.78 (0.57–1.07)
rs13269021	c.289-6755G>T	G	0.577	0.602	0.658	0.580	0.596	0.591	0.539	0.443	0.197	1.17 (0.92–1.49)	0.733	1.06 (0.77–1.46)	0.008*	1.67 (1.14–2.43)	0.874	0.97 (0.63–1.48)	0.040	1.27 (1.01–1.60)	0.934	1.01 (0.76–1.34)
rs11775442	c.289-6350A>G	A	0.798	0.833	0.796	0.778	0.796	0.813	0.793	0.798	0.870	1.02 (0.77–1.36)	0.272	1.26 (0.84–1.89)	0.946	1.02 (0.66–1.57)	0.652	0.90 (0.55–1.45)	0.900	1.02 (0.78–1.33)	0.597	1.10 (0.78–1.54)

CI = confidence interval; HTG = high tension glaucoma; NTG = normal tension glaucoma; OR = odds ratio; POAG = primary open-angle glaucoma; RAF = risk allele frequency

*P value < 0.05

Fig. 1

Association of rs1326901 with POAG using meta‐analysis in males (a) and females (b) combining the Hong Kong and Shantou cohorts. The association of rs13269021 in POAG was non-significant after combination by meta-analysis (P = 0.020 in males and 0.94 in females). CI, confidence interval; IV, inverse variance

Discussion

In this study, all six SNPs in ANGPT2 were not significantly associated with NTG, HTG and POAG in individual cohorts or in the combined analyses. Only SNP rs11775442 showed a nominal association with NTG in the Hong Kong cohort (P = 0.0498, OR = 1.24, 95% CI: 1.00–1.55) after adjusting for age and sex. It has been reported that rs11775442 (nAMD: P = 0.016; OR = 1.43; 95% CI: 1.0–1.92; PCV: P = 0.044; OR = 1.34; 95% CI: 1.01–1.79) were nominally associated with nAMD and PCV in the Hong Kong cohort [25]. Currently, no significant association has been reported between rs11775442 and any other disease. Further studies are needed to verify the association between SNP 11775442 with NTG.

In the subgroup analysis by sex, SNP rs13269021 in the Shantou cohort was significantly associated with NTG in males but not in females. Therefore, further studies to confirm its sex‐specific effects are warranted. Inconsistent results have been reported on the association of sex in POAG. In the United States, no significant sex difference in the POAG prevalence could be detected [27-29]. In contrast, a study from Netherlands reported a three-times higher POAG risk in males, while POAG was found more prevalent in females in an Australian study [30, 31]. In another study in African Americans, males showed a higher prevalence of POAG than females [32]. In the Los Angeles Latino Eye Study, males were found to possess an OR of 1.73 of having glaucoma vs. females [33]. Mechanistically, female sex hormones have been suggested to be protective against POAG [34, 35]. Serum estradiol levels were found positively correlated with the blood flow in the ophthalmic artery [36]. Menopause women receiving estradiol 17-valerate hormone replacement therapy showed significantly lower blood flow resistance in their central retinal arteries [37]. In addition, 17β-estradiol was also shown to protect retinal ganglion cells in rat and mouse glaucoma models [38, 39]. Interestingly, 17β-estradiol and another female sex hormone medroxyprogesterone acetate were able to reduce the protein production of ANGPT2 in human endometrial stromal cells [40]. Apart from our current study reporting ANGPT2 SNP (rs13269021) associating with male NTG in the Shantou cohort, a previous study found that BDNF Val66Met was associated with slower progression of POAG in females [41]. To our knowledge, our results are the first to highlight a sex-specific association of ANGPT2 in POAG.

The angiopoietin-Tie2 (Angpt-Tie2) pathway has been reported to control vascular maturation and stabilization [42]. Despite much research over the past decade, the roles of Angpt2 in the Angpt-Tie2 signaling system and vascular biology are still not fully understood. It is unclear whether Angpt2 is an antagonist or agonist of Tie2 in terms of vascular remodeling [42]. Better understanding of Angpt2 is very important to translate the use of its specific inhibitors alone or in combination with other anti-angiogenic agents into potential clinical treatments.

Regarding the SC, it has been reported that disruption of the Angpt-Tie2 signaling pathway results in high IOP, buphthalmos, and classical features of glaucoma, including retinal ganglion cell degeneration and vision loss in mice [21, 43]. Impairment of this system could be involved in the adult-onset POAG, as well as in the primary congenital glaucoma [44]. In our study, all six SNPs in ANGPT2 have not been found to be significantly associated with NTG, HTG and POAG in individual cohorts or in the combined analyses. Although Angpt-Tie2 signaling pathway is essential in the SC outflow pathway, Angpt2 was reported to play an indirect role as an agonist and an antagonist to Tie2 depending on the expression of an endothelial-specific phosphatase VEPTP [20]. Therefore, the genetic association between Angpt2 and POAG could also be influenced by the expression of other genes. For example, as a binding partner of Angpt2, a SVEP1 missense allele (rs61751937) has been reported to associate with POAG [22, 45]. According to the HapMap Genome Browser release #27 dataset, 34 SNPs could be selected in the locus of ANGPT2 [25]. In this study six SNPs were selected as they were reported as suggestive disease-associations in nAMD and PCV [25]. All these six SNPs are located in the introns of ANGPT2 [22]. Interestingly, instead of these six SNPs, SNP rs76020419 and rs2515437 have been reported to associate with POAG and elevated IOP [22-24]. rs76020419 is located in the 3ʹ untranslated region of ANGPT2, as well as in an intron of MCPH1. rs2515437 is located in the intron of ANGPT2 and MCPH1. MCPH1 encodes a microcephalin protein that is important for brain development [46]. Haplotype‐tagging SNP analysis of ANGPT2 and MCPH1 should be further studied in POAG patients.

In the subgroup analysis, our results indicated a male-specific association in rs13269021 in NTG from the Shantou cohort. In this cohort, rs13269021 conferred approximately 1.67-fold increased risk of NTG in males. But the OR of the SNP in females was 0.97, indicating a slightly decreased risk of NTG. The same effect has not been found in the Hong Kong cohort although the sample size of NTG patients from the Hong Kong cohort was larger than the Shantou cohort. Therefore, the sex-specific effects of this locus remained to be determined in more NTG samples.

Limitations

There are several limitations in this study. First, these six SNPs chosen according to our previous nAMD and PCV study could not cover the genetic linkage disequilibrium regions in the ANGPT2 gene. It may have narrowed the scope of the findings. Second, the key finding of rs13269021 in the Shantou cohort significantly associated with NTG in males but not in females could not be reproduced in the Hong Kong cohort. Our cohorts were limited to ethnic Chinese. Therefore, our results may not be translatable to other ethnicities. Third, all six SNPs are located in introns of ANGPT2. The functions of these SNPs are unclear. Indeed, no publication have reported that these six SNPs could cause any observable phenotypes.

Conclusions

In the subgroup analysis by sex, a significant association was shown in SNP rs13269021 with NTG in Shantou males, but not in Hong Kong males. Further studies are needed to verify the association between ANGPT2 locus (rs13269021) and NTG in Chinese males.