Editor,—Mutations causing primary open angle glaucoma (POAG) have been identified in the chromosome 1 open angle glaucoma gene (GLCL1A), which encodes a 57 kD protein known as myocilin (MYOC).1-4 A number of mutations of the MYOC gene that are associated with glaucoma have been reported in different ethnic groups.1-3 The mutations occur in 2–4% of POAG patients, but mostly in the exons. Although five different such alleles have been identified among Japanese patients,5 6 they are rare and a promoter region has not received much attention. We conducted a study using the same methods1 2 in Japanese patients with POAG and normal tension glaucoma (NTG) who had a family history of the disease in first and second degree relatives (n=30, 53), and in control subjects (n=38, 63) who had no ophthalmological conditions.

The results of this screening for the Japanese patients with POAG disclosed no exon mutations. The introns at the exon/intron boundaries had no abnormalities. However, three abnormalities were found in the promoter region (Table 1). One was a polymorphism at –83 G→A (83 base pairs upstream from initiation codon), which had been reported previously in Westerners,1-3 5 7 and another was a new mutation of (GT) repeats at –339 to –314, which also seemed polymorphic. The last promoter mutation (–153 T→C) is newly observed, and was found in a substantial number of our patients. However, no significant difference was found in the first survey (Table1).

The –153 T→C creates tandem repeats of [CAGCCCCAC]. The CACCC motif, transcription regulator in β-like globin genes, bridges the two repeats [5′CAGCCCCAC]C[CAGCC(T→C)CAC]. Thus, this mutation seems to affect the expression of the myocilin gene. These results caused us to extend the examination in a second survey for the –153 T→C to other patients with POAG and NTG (n=53) and to new age matched control subjects (n=63). The result showed a significant difference between the patients and the control group (p<0.01, Fisher's exact probability). The age matched second survey provided results similar to those of the first survey. Even when we combine the results of both, the difference was also evident (Table 1). No homozygosity for the –153 C was detected.

Therefore, the –153 C mutation is very likely to be disease related. The mutation comprises nearly 20% of patients with POAG and NTG, but only 4.8% of the control subjects. An incidence as high as this has not been observed in any POAG related mutant reported previously.1-3 5-8 In fact, three control subjects carrying the –153C in the original survey were examined ophthalmologically, and to our surprise, two of them had POAG. These carriers in the control group, despite the absence of ophthalmological complaints, may have an insidious development of glaucoma.

Our preliminary study suggested that the promoter activity is different between the –153 T and C alleles, and it depends on the linkage with the number of GT dinucleotide repeats, about 160 nucleotides upstream from this mutation.

This –153 C allele may be found exclusively in Japanese patients and may be the major mutation associated with POAG and NTG in Japanese people. Clinical phenotypes differ among ethnic groups,8-10 which may be explained genetically.