TINF2 mutations in children with severe aplastic anemia

Abstract

To the Editor: Dyskeratosis congenita (DC), a bone marrow failure (BMF) syndrome associated with skin abnormalities, dystrophic nails, and leukoplakia, has been linked to mutations in genes that encode components of telomerase (dyskerin, TERC, TERT, NHP2, and NOP10). Recently, one of the components of the shelterin complex, TIN2, which is encoded by the TINF2 gene, was found to be mutated in some patients with autosomal dominant DC and in sporadic cases of DC [1,2]. Autosomal dominant forms of DC caused by mutations in TERT or TERC are often mild; many of these patients present with only BMF and are indistinguishable from individuals with idiopathic severe aplastic anemia (SAA) or myelodysplastic syndrome (MDS). In a recent issue of the journal, Pigullo et al. [3] hypothesized that mutations in NOLA1 (also known as GAR1), which encodes another component of the telomerase complex, might account for some cases of aplastic anemia (AA) in children. Of 108 patients with AA screened, none had a pathogenic mutation [3]. We postulated that TINF2 mutations might underlie some cases of idiopathic AA in childhood. To address this, we analyzed DNA samples from an anonymous registry [4] of pediatric patients who underwent unrelated donor hematopoietic stem cell transplantation (UD-HSCT) for SAA (n = 109) or MDS (n =137). We focused our mutational analysis on exon 6 of the gene, the site of all known TINF2 mutations linked to DC [1,2]. We identified potentially pathogenic mutations in 6 of 109 patients with SAA. A 2-year-old Caucasian male and a 10-year-old Caucasian male were heterozygous for TINF2 G845A (R282H), a mutation previously shown to cause DC with early onset AA [1,2]. A 6-year-old Caucasian female and a 17-year-old Caucasian male were heterozygous for C844T (R282C), another missense mutation linked to DC [2]. A 7-year-old Hispanic boy was heterozygous for a single nucleotide insertion that alters the TIN2 sequence beginning at residue 290 and introduces a stop codon (F290R291N292 → P290L291X). The remaining patient, a 4-year-old Hispanic girl, was heterozygous for the missense mutation T862C (F288L). The phenylalanine residue at position 288 of human TIN2 is conserved in chimpanzee, rat, and dog, whereas mouse and cow have a leucine residue at this position [1]. Therefore, it is unclear whether the F288L substitution is deleterious. No TINF2 mutations were detected in patients with MDS (P <0.01). Previously, 298 healthy control individuals of diverse ethnicities were screened for TINF2 mutations [1]; none had a TIN2 sequence change near amino acid 282 (P <0.001). We conclude that a significant proportion of patients who underwent UD-HSCT for SAA harbored mutations in TINF2 and may have had occult DC. Among pediatric patients with SAA in the registry, TINF2 mutations were more common than TERC mutations (5.5% vs. 0%, P <0.01) [4]. In pedigrees with DC due to mutation of TERC, signs and symptoms of the disease usually present in adulthood. Our findings, coupled with those of other investigators [1,2], suggest that mutation of TINF2 is more likely to manifest during childhood as SAA. In contrast, TERC and TERT gene mutations are associated with MDS and possibly acute myelogenous leukemia later in life.