Genetic Variation in TERF1 but Not TERF2 Is Associated with Aplastic Anemia Risk.

Approximately one-third of individuals with acquired aplastic anemia (AA) have abnormally short leukocyte telomere lengths. Mutations in genes responsible for maintaining telomere length and stability have been identified in constitutional AA (DKC1) and acquired AA (TERC and TERT). Telomerase, a reverse transcriptase, uses an RNA template (TERC) to extend nucleotide repeats. Telomeric repeat binding factor 1 (TERF1) binds to telomeric DNA, inhibits telomerase and induces bending, looping and pairing of duplex telomeric DNA. Telomeric repeat binding factor 2 (TERF2) binds to the telomere and protects it from degradation and fusion. There are few single nucleotide polymorphisms (SNPs) in TERF1 and TERF2 and nucleotide diversity in these genes is limited. We hypothesized that mutations and/or SNPs in TERF1 and/or TERF2 may be associated with acquired AA. Bi-directional sequence analysis was performed in 47 AA patients who failed to respond to immunosuppressive therapy, had family history of hematologic abnormalities without physical characteristics of dyskeratosis congenita, or short telomeres in leukocytes, on all exons and proximal promoter regions of TERF1 and TERF2. Regions with variation in the first 47 patients were sequenced in an additional 95 patients and 289 healthy controls. SNP frequencies were determined and case-control genotype data analyzed using additive and mutation dominant (MD) genetic models and SAS v8.02. Haplotypes were estimated and a case-control permutation test performed using PHASEv2.1. HaploStats was used to construct haplotypes, determine a global score p value, haplotype frequencies and odds ratios.

5244 base pairs (bp) were sequenced in TERF1 in 47 patients. SNPs were not present in exons 1–8. The proximal promoter, exon 9 and exon 10 were sequenced in the additional 95 patients and 289 controls. One patient had a mutation resulting in a conservative amino acid change (exon 9 Ala-Val). One SNP in intron 9 (C>T) was significant in the additive (p=0.049) and in the MD model, odds ratio 1.59 (p=0.033, 95% CI 1.06-2.39, 37.3% case and 31.8% control allele frequency [AF]). Two synonymous SNPs in exon 10 did not occur in patients but each had an AF of 1.4% in controls (p=0.044 for each). The MD model for these two SNPs suggested an association (p=0.044) but the low frequencies precluded additional calculations. Haplotype analysis showed statistically significant differences between cases and controls in global tests (HaploStats p=0.03 and PHASEv2.1 p=0.01). One haplotype had borderline significance in ordinal trait analysis and comparison to the most common haplotype (p= 0.057 and 0.079, respectively, 35.6 % case and 28.9% control frequency). These data suggest that genetic variation in TERF1 may be associated with aplastic anemia.

Sequence analysis of TERF2 consisted of 4515 bp. The proximal promoter, exons 1–2, 4–5 and 7–10 did not have SNPs or mutations. A mutation (Ala-Ser, exon 6) in TERF2 was seen in one patient and no controls. SNPs seen in exons 3 and 6 were also sequenced in 95 patients and 289 controls. SNPs and haplotypes in TERF2 were not associated with AA. This pilot study identified a SNP and haplotype in TERF1 that may be associated with increased risk for AA. Variation at this site in TERF1 in combination with variation at other sites in other genes of the telomere complex could be risk factors for aplastic anemia. Highly penetrant mutations in TERF1 and TERF2 were not identified. These results should be confirmed in a larger study.