Association of 15q14 and 15q25 with High Myopia in Japanese

Abstract

Purpose.: To investigate whether there are associations of genetic variations in chromosome 15q14 and 15q25, recently shown to confer risk of refractive error and myopia in Caucasians, with high myopia in Japanese.

Methods.: A total of 1125 unrelated Japanese patients with high myopia and two independent control groups were evaluated (366 cataract patients without high myopia and 929 healthy Japanese patients). The four single nucleotide polymorphisms (SNPs) rs634990 and rs524952 at 15q14 and rs8027411 and rs17175798 at 15q25 were genotyped.

Results.: A significant association with high myopia was observed in 15q14 (P = 0.0035 for rs634990 and P = 0.0017 for rs524952 when evaluated with cataract controls and P = 1.91 × 10^-6 for rs634990 and P = 8.78 × 10^-7 for rs524952 with healthy Japanese controls). When evaluated with cataract controls, the odds ratios (95% confidence intervals) were 1.30 (1.10–1.53) for rs634990 C allele and 1.32 (1.11–1.56) for rs524952 A allele. The population attributable risks were 0.29 and 0.30, respectively. The SNPs in 15q25 did not show a significant association with high myopia when evaluated with cataract control (P > 0.42), while it showed a weak association when evaluated with healthy Japanese controls (P = 0.031 for rs8027411 and P = 0.047 for rs17175798) with odds ratios of 1.17 (1.03–1.33) for rs8027411 T allele and 1.15 (1.02–1.31) for rs17175798 C allele.

Conclusions.: These findings suggest that a region in 15q14 is susceptibility loci for high myopia. This locus harbor susceptibility genes for not only common myopia but also for high myopia. The 15q25 locus might also have association to myopia.

Myopia is the most common visual disorder in the world, with the prevalence estimated to be 25% in the United States and Western Europe and to be much higher (40%–70%) in Asians, including Chinese and Japanese. ^{1 –5} Myopic eyes with very long axial lengths (≥26 mm) or a high degree of myopic refractive error (≤ −6 D) are classified as high myopia. ⁶ Recently, the prevalence of high myopia has been increasing worldwide, especially in the younger East Asian population. ^7,8 It is well known that high myopia is associated with various ocular complications, ⁹ and it is one of the leading causes of legal blindness in developed countries. ^{10 –12} Therefore, it is important to develop methods for preventing or delaying the onset of high myopia or for limiting its progression. 

The cause of myopia is not fully understood, and, in fact, it is not yet clear that common myopia and high myopia share the same background; high myopia may have a unique background that distinguishes it from common myopia. It has been shown that environmental factors (in particular near work, a higher level of education, and less time spent in outdoor activities) and genetic factors contribute to the development of myopia. ¹³ Twin studies provide compelling evidence that myopia is inherited, ^14,15 and multiple family-based whole genome linkage analysis indicate several heritable myopia susceptibility loci, such as the MYP1–18 loci. ^{16 –31} Some of these loci are reported to be associated with common myopia, high myopia, or both. However, several studies could not replicate these associations. ^{32 –34} In addition, many candidate genes associated with high myopia have been reported. ^{35 –43} However, most of these associated candidate genes have been negated by subsequent studies, and no genes have yet been identified that are consistently responsible for either common or high myopia. 

Genome-wide association studies (GWAS) are expected to reveal the susceptibility genes of many complex diseases, as shown in age-related macular degeneration (AMD). We performed the first genome-wide association study for high myopia and showed that several single nucleotide polymorphisms (SNPs) at 11q24.1 are associated significantly with high myopia in Japanese. ⁴⁴ However, a study in Han Chinese did not show the association of rs577948, one of the SNPs, with high myopia. ⁴⁵  

Recently, two other groups have performed GWAS for refractive error and showed that SNPs in 15q14 and 15q25 are associated with refractive error and myopia, ^46,47 but the cohorts used in these two studies were population-based, and only a limited number (1.7%–4.0%) of patients with high myopia were included. Although it is not clear if common and high myopia share the same genetic background, some MYP loci are reported to be associated with both common myopia and high myopia. In this study, we evaluated the associations of reported SNPs with high myopia in Japanese. 

Moreover, we investigated their associations with the occurrence of choroidal neovascularization (CNV) in high myopia. Some highly myopic eyes develop CNV, while other highly myopic eyes do not, and because CNV is one of the most vision-threatening complications in highly myopic eyes, the investigation of the mechanisms of how it occurs is important. We evaluated the influence of susceptibility genes for neovascular AMD on the occurrence of CNV in myopic eyes and found that these genes do not affect the development of myopic CNV. ⁴⁸ Axial elongation of highly myopic eyes results in thinning of the retina and choroid, patchy chorioretinal atrophy, and lacquer cracks, all of which are important predisposing conditions for the development of CNV. ^49,50 It could be hypothesized that CNV occurs when the eye is strongly affected by susceptibility genes for myopia, or it may be that specific susceptibility genes exist for myopic CNV—genes that are in addition to the susceptibility genes for myopia. 

Materials and Methods

All procedures in this study adhered to the tenets of the Declaration of Helsinki. The institutional review boards and the ethics committees of each institution involved approved the protocols of this study. All patients were fully informed of the purpose and procedures of this study, and written consent was obtained from each patients. 

Patients

One thousand one hundred twenty-five unrelated highly myopic Japanese patients were recruited from Kyoto University Hospital, Tokyo Medical and Dental University Hospital, Ozaki Eye Hospital, and Otsu Red-cross Hospital (mean age ± SD, 57.6 ± 14.8 years; male/female ratio, 33.5%/66.5%). All patients had a comprehensive ophthalmic examination, including dilated indirect and contact lens slit lamp biomicroscopy, automatic objective refraction, and measurements of the axial length by applanation A-scan ultrasonography or partial coherence interferometry (IOLMaster; Carl Zeiss Meditec, Dublin, CA). The difference between these two devices would be around 0.1 mm. ⁵² When a patient who visited the aforementioned hospitals had an axial length of >26.1 mm in both eyes, peripheral blood was obtained after informed consent and used as high-myopia patients group. To evaluate the contribution of the SNPs between high myopic patients with CNV and high myopic patients without CNV, the high myopic patients group was separated into a CNV group (600 patients) and a no CNV group (450 patients). Inclusion criteria of the CNV group were clinical presentation and angiographic manifestations of macular CNV or Fuchs' spot in at least one eye. 

As the control subjects, we used 366 cataract patients with axial lengths <25.0 mm in both eyes (control 1). These patients were recruited from the Department of Ophthalmology at Kyoto University Hospital, the Ozaki Eye Hospital, the Japanese Red Cross Otsu Hospital, and Nagahama City Hospital; their mean age (± SD) was 74.4 ± 8.4 years, and there were 39.9% men and 60.1% women. The axial length was measured with applanation A-scan ultrasonography or partial coherence interferometry before cataract surgery and dilated fundus examination was performed after surgery. If the fundus examination revealed myopic change such as lacquer cracks/peripapillary atrophy, staphyloma or CNV, the subject was eliminated from the control 1 group. 

We also used DNA samples from 929 subjects who were randomly selected from the Pharma SNP Consortium (PSC); this constituted control group 2. This group has been used for previous genomic studies and is regarded as being representative of the general Japanese population (mean age ± SD, 38.8 ± 11.8 years; male/female ratio, 61.7%/38.3%). ⁵¹ All were Japanese and none had any history of ocular disease. 

Genotyping and Statistical Analyses

Genomic DNAs were prepared from peripheral blood using a DNA extraction kit (QuickGene-610L; Fujifilm, Minato, Tokyo, Japan). Four SNPs (rs634990, rs524952, rs8027411, and rs17175798) were selected based on their specific presence in two previous GWAS; they were genotyped using a commercially available assay (TaqMan SNP assay with the ABI PRISM 7700 system; Applied Biosystems, Foster City, CA). Deviations in genotype distributions from the Hardy–Weinberg equilibrium (HWE) were assessed for each group with the HWE exact test. The χ² test for trend or its exact counterpart was used to compare the genotype distributions of two groups. To adjust age and sex, we performed multiple regression and logistic regression analysis. These statistical analyses were performed with R software R (R Foundation for Statistical Computing, Vienna, Austria, available at http://www.rproject.org/). P < 0.05 was considered statistically significant. 

Results

The demographics of the study population are shown in Table 1. The axial length of the 2258 eyes of the 1129 high myopia cases ranged from 26.11 to 39.73 mm, with a mean ± SD of 29.09 ± 2.04 mm. Among the 2250 eyes enrolled, 1753 (77.9%) were phakic, 427 (19.0%) were pseudophakic, and 70 (3.1%) were aphakic. The mean refraction of the 1753 phakic eyes was −10.52 ± 6.71 D. The axial length of the 732 eyes of the cataract patients ranged from 18.67 to 24.89 mm, with a mean ± SD of 23.02 ± 1.28 mm. Mean refraction of the phakic eyes in control group 1 was −0.14 ± 2.2 D. 

Table 1.

 

Characteristics of the Study Population

	Patients	Controls
	High Myopia*	Cataract†	PSC‡
Patients, n	1125	366	929
Age in years, Mean ± SD	57.57 ± 14.75	74.40 ± 8.37	38.81 ± 11.83
Sex, n (%)
Male	377 (33.5%)	146 (39.9%)	573 (61.7%)
Female	748 (66.5%)	220 (60.1%)	356 (38.3%)
Axial length, mm ± SD
Right eyes	29.18 ± 1.95	23.05 ± 1.00	NA
Left eyes	29.01 ± 2.16	23.01 ± 1.66	NA
Refraction of the phakic eyes, D§
Right eyes	−10.79 ± 6.96	−0.23 ± 2.56	NA
Left eyes	−10.36 ± 5.82	−0.11 ± 2.52	NA

*  Axial length of >26.1 mm in both eyes.

†  Individuals who underwent cataract surgery and who had an axial length <25.00 mm in both eyes.

‡  DNA samples randomly selected from the Pharma SNP Consortium (PSC).

§  For calculations of refraction, eyes that had undergone cataract surgery or corneal refractive surgery were excluded.

Genotype counts, associations examined with χ² test for trend, and odds ratios for the four SNPs between the high myopia patients and the controls are shown in Tables 2 and 3. The distributions of the genotypes for the four SNPs were all in HWE (P > 0.1), as assessed with the exact test. Rs634990 and rs524952 were in almost complete linkage disequilibrium, as were rs8027411 and rs17175798 (D′ > 0.95). 

Table 2.

 

Genotype Counts, Associations, and Odds Ratios in the High Myopia Patients and Cataract Controls

Locus	SNP ID	Genotype			P *	Adjusted P †	OR (95% CI)	PAR
			High Myopia	Controls
15q14	rs634990 (C/T)	CC	304	84	0.0026	0.0035	1.65 (1.19–2.29)	0.29
		CT	571	165			1.58 (1.19–2.09)
		TT	246	112			1.00 (ref)
	rs524952 (A/T)	AA	303	81	0.0015	0.0017	1.70 (1.22–2.37)	0.30
		AT	572	164			1.59 (1.19–2.11)
		TT	244	111			1.00 (ref)
15q25	rs8027411 (G/T)	TT	428	116	0.17	0.42	0.87 (0.60–1.26)	0.03
		GT	525	193			1.17 (0.82–1.67)
		GG	166	52			1.00 (ref)
	rs17175798 (C/T)	CC	422	117	0.25	0.60	1.12 (0.77–1.62)	0.04
		CT	528	193			0.85 (0.60–1.20)
		TT	171	53			1.00 (ref)

 SNP, single nucleotide polymorphism; OR, odds ratio; CI, confidence interval; PAR, population attributable risk.

*  Differences in the observed genotypic distribution were examined by χ² test for trend.

†  Age and sex adjustment was performed based on a logistic regression model.

Table 3.

 

Genotype Counts, Associations, and Odds Ratios in the High Myopia Patients and Pharma SNP Consortium Controls

Locus	SNP ID	Genotype			P *	Adjusted P †	OR (95% CI)	PAR
			High Myopia	Controls
15q14	rs634990 (C/T)	CC	304	191	1.1 × 10−6	1.91 × 10−6	1.84 (1.44–2.36)	0.29
		CT	571	442			1.50 (1.21–1.85)
		TT	246	285			1.00 (ref)
	rs524952 (A/T)	AA	303	191	7.60 × 10−7	8.78 × 10−7	1.86 (1.45–2.39)	0.30
		AT	572	444			1.51 (1.22–1.86)
		TT	244	286			1.00 (ref)
15q25	rs8027411 (G/T)	TT	428	310	0.013	0.031	1.37 (1.06–1.78)	0.17
		GT	525	445			1.17 (0.91–1.51)
		GG	166	165			1.00 (ref)
	rs17175798 (C/T)	CC	422	309	0.026	0.047	1.32 (1.02–1.71)	0.14
		CT	528	453			1.12 (0.88–1.44)
		TT	171	165			1.00 (ref)

 SNP, single nucleotide polymorphism; OR, odds ratio; CI, confidence interval; PAR, population attributable risk.

*  Differences in the observed genotypic distribution were examined by χ² test for trend.

†  Age and sex adjustment was performed based on a logistic regression model.

When evaluated with control group 1 (cataract patient controls, n = 366), the SNPs rs634990 and rs524952 in 15q14 showed a significant association with high myopia (P = 0.0035 and 0.0017, respectively). The odds ratios (95% confidence intervals [CIs]) were 1.30 (1.10–1.53) for rs634990 C allele and 1.32 (1.11–1.56) for rs524952 A allele. When evaluated with control group 2 (healthy Japanese controls, n = 929), these two SNPs showed strong association with high myopia (P = 2.21 × 10⁻⁶ and 1.62 × 10⁻⁶, respectively). The odds ratios were 1.36 (1.20–1.54) for rs634990 C allele and 1.37 (1.21–1.55) for rs524952 A allele. The population attributable risk (PAR) was 0.29 for rs634990 and 0.30 for rs524952 when evaluated with both controls, and the risk alleles were the same as those reported previously. 

The variants in 15q25 (rs8027411 and rs17175798) showed no association to high myopia when evaluated with control group 1 (n = 366), while these two SNPs showed marginal associations (P = 0.031 and 0.047, respectively) when evaluated with control group 2 (n = 929). The minor allele frequencies in these two controls were very similar; 0.41 for both SNPs in control group 1 and 0.42 for both SNPs in control group 2, and the risk alleles seemed to be the same as those reported previously. The odds ratios were 1.13 (0.95–1.34) for rs8027411 T allele and 1.10 (0.93–1.31) for rs17175798 C allele when evaluated with control group 1 and 1.17 (1.03–1.33) and 1.15 (1.02–1.31), respectively, when evaluated with control group 2. The PARs were 0.03 for rs8027411 and 0.04 for rs17175798 when evaluated with control group 1 and 0.17 and 0.14, respectively, when evaluated with control group 2. 

Associations for the four SNPs with CNV in highly myopic eyes of the Japanese population in this study are shown in Table 4. We analyzed the genotype distributions for rs634990, rs524952, rs8027411, and rs17175798 in high myopia patients with CNV and compared them with high myopia patients without CNV to check for a possible difference between these two groups in the genetic background of the four myopia-susceptibility variants. Distributions of these four genotypes were all in HWE (P > 0.2), and there was no significant difference in the allele frequencies of SNPs between high myopia with CNV and high myopia without CNV—even after age and sex adjustment. 

Table 4.

 

Associations of the Four Single Nucleotide Polymorphisms with Choroidal Neovascularization Development in Highly Myopic Eyes

	15q14						15q25
	rs634990 C			rs524952 A			rs8027411 T			rs17175798 C
	Frq	OR (95% CI)	P *	Frq	OR (95% CI)	P *	Frq	OR (95% CI)	P *	Frq	OR (95% CI)	P *
High myopia with no CNV (n = 600)	0.53			0.53			0.61			0.61
High myopia with CNV (n = 450)	0.51	0.93 (0.78–1.11)	0.41	0.51	0.93 (0.78–1.10)	0.38	0.62	1.04 (0.87–1.25)	0.64	0.60	0.95 (0.79–1.13)	0.54

 Frq, allele frequency.

*  Age and sex adjustment was performed based on a logistic regression model.

The average age was 54.9 ± 14.9 years in the patients without CNV, while the average age was 60.7 ± 13.3 years in the patients with CNV (P < 0.0001). Because the average age was significantly higher in the group with CNV, we performed subanalysis dividing the cohort into 40- to 49-year-old, 50- to 59-year-old, 60- to 69-year-old, and 70- to 79-year-old subgroups (Table 5). Our subanalysis revealed that there were no associations between CNV occurrence and the genotype variation in rs634990, rs524952, rs8027411, and rs17175798. 

Table 5.

 

Subanalysis of the Associations of the Four Single Nucleotide Polymorphisms with Choroidal Neovascularization Development in Highly Myopic Eyes

Age, y	n		P Value*
	High Myopia with CNV	High Myopia with No CNV	rs634990	rs524952	rs8027411	rs17175798
<40	37	99	0.55	0.42	0.63	0.63
40–49	47	92	0.13	0.14	0.18	0.28
50–59	91	143	0.88	0.89	0.74	0.79
60–69	145	149	0.79	0.73	0.98	0.66
70–79	103	84	0.91	0.90	0.48	0.61
≥80	27	33	0.58	0.61	0.54	0.65

Discussion

Myopia has been thought to be a multifactorial disease, and for more than a decade many researchers have sought the susceptibility genes associated with myopia. Several chromosome loci have been reported to be associated with common myopia, high myopia, or both ^{16 –31} ; however, some other investigators could not replicate these associations. ^{32 –34} To date, no genes have been identified that are consistently responsible for either common or high myopia. Furthermore, it has not been clear if common and high myopia share the same genetic background or if high myopia has a unique genetic background that distinguishes it from common myopia. In the present study, we have shown that SNPs in 15q14 are associated significantly with high myopia in Japanese and that SNPs in 15q25 might also be associated. These same SNPs have been reported recently to be associated with myopia in Caucasians, ^46,47 although the Caucasian cohort was population-based and patients with high myopia were extremely rare (1.7%–4.0%). Common and high myopia may well share the same genetic background—at least in part. 

In the present study, we used two distinct control groups: cataract patients with axial lengths <25.00 mm in both eyes (control group 1, n = 366) and representative of the general Japanese population (control group 2, n = 929). Because control group 2 is representative of the general Japanese population, high myopia patients may well be included as control subjects. When one considers the high rate of high myopia in Japanese compared to Caucasians, in whom it is estimated to be present in approximately 5% of the general population, this could weaken the detection of any genetic association with high myopia. However, we believe that using the general population as a control is certainly acceptable for a case control study of high myopia. ⁴⁴  

When evaluated with both controls, 15q14 showed a significant association with high myopia. The odds ratio of rs634990 CC genotype to TT genotype was 1.65 (95% CI, 1.19–2.29) when compared with control group 1 and 1.84 (95% CI, 1.44–2.36) when compared with control group 2, findings that are compatible with the reported odds ratio of 1.83 (1.42–2.36) for refractive error and myopia in a general population of Caucasians. The genetic variation at 15q14 seems to contribute to common myopia and to high myopia in a similar manner. 

With 15q25, however, the association significance was only marginal when evaluated with control group 2, and the analysis with control group 1 showed no association to high myopia. In accordance with the original report, the risk allele was T in rs8027411. Because control group 2 consisted of the general population, and because patients with high myopia may be included as control subjects, this could lower the power of this study to detect a genetic association with high myopia. However, two SNPs in 15q25 were significantly associated with high myopia evaluated with control group 2. Considering that the minor allele frequencies in 15q25 were very similar between our control groups 1 (n = 366) and 2 (n = 929) and the original study showed that the odds ratio was very low as 1.16 (1.02–1.28) for the rs8027411 T allele, the size of control group 1 might be insufficient to detect the association. Even for the analysis with control group 2, the association becomes negative after Bonferroni correction. We might have to negate the association of 15q25 locus. Although our present study could be regarded as a replication study for the association of previously reported loci to myopia, and because the Bonferroni correction might not be applicable for a replication study, we would have to interpret the association of 15q25 locus in the present study with caution. Because the association of 15q25 to myopia was reported to be very low and the minor allele frequencies in 15q25 were very similar between control groups 1 and 2, additional study with a larger cohort might reveal the true association of 15q25 to myopia. 

We also evaluated the genetic difference between high myopia patients with CNV and high myopia patients without CNV, and found that the genotype distribution of the SNPs evaluated was not significantly different. These high myopia patients without CNV might develop CNV later. Because CNV will develop only in 5% to 10% of high myopia patients and because it typically starts to develop in the fourth or fifth decades of life, the number of patients who will develop CNV later in these groups would be limited. However, the individuals who will develop CNV later in the control cohort would weaken the power to detect the associations to CNV development. Our negative result might be partly related to this weakened power. To eliminate the influence of such individuals, we performed subanalysis dividing the cohort into 40- to 49-year-old, 50- to 59-year-old, 60- to 69-year-old, and 70- to 79-year-old subgroups. Our subanalysis showed no associations of these loci to the CNV development. Although our findings should be interpreted with caution, factors other than 15q14 and 15q25 might affect the development of CNV in highly myopic eyes. Lacquer cracks and peripapillary atrophy can be the basis of CNV development. Although lacquer cracks and peripapillary atrophy do not always lead to CNV development, genomic studies paying attention to these features might give us loci associated with CNV development. The occurrence of CNV beneath the fovea is one of the most vision-threatening complications of highly myopic eyes, so it is important to study the mechanisms of CNV occurrence in such eyes even after susceptibility genes for myopia are known. 

Because genetic associations have ethnic differences, our findings could not be compared directly to the previous studies that were performed in Caucasians. Furthermore, these previous reports of associations of 15q14 and 15q25 with myopia used a general population, while we performed a case control study using high myopia patients as the cases and the general population as controls, so our control cohort is almost the same as the entire cohort used in the previous studies. Although the associations of 15q14 and 15q25 to refractive error (myopia) and/or high myopia need to be evaluated in various ethnicities, our findings of similar contribution of these loci to common myopia in Caucasians and to high myopia in Japanese suggests that 15q14 and 15q25 contribute to these two conditions in a similar manner, although additional studies might reveal genetic differences that differentiate high myopia from common myopia. Moreover, additional genetic background might also affect the occurrence of myopia/high myopia, given at most 30% of PAR. 

In conclusion, we have shown that genetic variations at 15q14 are associated significantly with high myopia in Japanese. Based on our findings and those of previous studies, this might be a susceptibility locus for both myopia and high myopia. The association of 15q25 should be evaluated in additional studies with larger cohorts. Our findings also suggest that CNV occurs independent of genetic variations at these loci, and that other factors affect the occurrence of CNV in highly myopic eyes. 

Footnotes

 Supported in part by Grants-in-aid for scientific research (nos. 21249084 and 22791653) from the Japan Society for the Promotion of Science, Tokyo, Japan. The funding organization had no role in the design or conduct of this research.

Footnotes

 Disclosure: H. Hayashi, None; K. Yamashiro, None; H. Nakanishi, None; I. Nakata, None; Y. Kurashige, None; A. Tsujikawa, None; M. Moriyama, None; K. Ohno-Matsui, None; M. Mochizuki, None; M. Ozaki, None; R. Yamada, None; F. Matsuda, None; N. Yoshimura, None.

The authors thank Shoji Kuriyama and Yoshiki Ueda for their assistance in the recruitment of patients.