A specific missense mutation in GTF2I occurs at high frequency in thymic epithelial tumors

Abstract

We analyzed 28 thymic epithelial tumors (TETs) using next-generation sequencing and identified a missense mutation (chromosome 7 c.74146970T>A) in GTF2I at high frequency in type A thymomas, a relatively indolent subtype. In a series of 274 TETs, we detected the GTF2I mutation in 82% of type A and 74% of type AB thymomas but rarely in the aggressive subtypes, where recurrent mutations of known cancer genes have been identified. Therefore, GTF2I mutation correlated with better survival. GTF2I β and δ isoforms were expressed in TETs, and both mutant isoforms were able to stimulate cell proliferation in vitro. Thymic carcinomas carried a higher number of mutations than thymomas (average of 43.5 and 18.4, respectively). Notably, we identified recurrent mutations of known cancer genes, including TP53, CYLD, CDKN2A, BAP1 and PBRM1, in thymic carcinomas. These findings will complement the diagnostic assessment of these tumors and also facilitate development of a molecular classification and assessment of prognosis and treatment strategies.

Figure 1

Overview of somatic mutations in TETs. (a) Average number of mutations (SNVs and indels) by WHO histotypes as detected by exome sequencing (n = 28). More mutations were observed in thymic carcinomas than in thymomas (Mann-Whitney U P = 0.001). Type B3 thymomas had more mutations than types A and B2, but this difference was not significant. (b) Mutations detected by exome sequencing or 197-gene assay (n = 54). At the bottom, the two tracks report the CGH clusters to which the cases belong and whether the samples have been sequenced by exome sequencing and/or 197-gene assay (Seq).

Figure 2

GTF2I mutation in TETs. (a) Frequency of GTF2I mutation by WHO histotype (combining the results of all analytic platforms). The number of patients sequenced is given in parenthesis next to each WHO histotype. (b) Kaplan-Meier survival curve demonstrating a more favorable outcome in GTF2I-mutated (MUT, blue line) than in wild-type (WT, red line) TETs (log-rank test P < 0.001; 83 and 121 evaluable patients, respectively).

Figure 3

Expression of GTF2I isoforms and their role in cell proliferation. (a) There are five known isoforms of TFII-I (α, β, γ, δ and 5) that differ by alternative splicing of exons 10 and 12. In a case of type A thymoma with GTF2I mutation (MUT), the estimated gene expression status is depicted at the exon level. Exon 10 was silent, and exon 12 was expressed at about half the level of the other exons (gray bars in the read count). The absence of exon 10 limits the expression to the β and δ isoforms. Chr7, chromosome 7. (b) TFII-I expression was tested in 21 frozen tumors using antibodies to TFII-I by western blot analysis. The tumors were divided into three separate blots (Blot-1, Blot-2 and Blot-3). Every blot contains wild-type (WT) and mutant (MUT) cases in order to facilitate the comparison. TFII-I protein expression was higher in MUT than WT tumors. For each tumor, the histotype is indicated, and a loading control with β-actin is included. Normalized quantification of TFII-I protein expression is reported at the bottom of the blots. (c) The average of FPKM (fragments per kilobase of exon per million fragments mapped) values, an estimator of mRNA expression, was higher in wild-type (n = 7) than mutated (n = 5) tumors for both the β and δ GTF2I isoforms for tumors sequenced with HiSeq2000. (d) Using a lentiviral vector, WT and MUT β and δ isoforms were ectopically expressed in NIH-3T3 cells. Transfected β and δ isoforms of TFII-I were visualized using antibody to V5 (V5-Ab). The loading control was α-tubulin (α-tub.). In all clones tested, the expression of TFII-I–mutated isoforms was consistently higher than that of the wild-type isoforms. The extra lower molecular weight band indicated by the orange asterisk might be the result of TFII-I degradation during sample preparation. (e) NIH-3T3 cells carrying the mutated β and δ isoforms (solid lines) exhibit a higher proliferation rate than those carrying the wild-type isoforms (dashed lines). Experiments were repeated three times, and all error bars indicate the s.d.

Figure 4

Structure of TFII-I protein and distributions of genomic alterations in TETs histotypes. (a) Schematic illustration of TFII-I domains, the I-repeat region and the position of the leucine-to-histidine change. The TFII-I repeats are as follows: six helix-loop-helix–like domains (purple boxes); the DNA binding domain basic region (BR, orange bar); the nuclear localization signal (NLS, green bar); the leucine zipper domain (LZ, gray bars) and the mutation locus (M, red arrow). (b) Model of TFII-I with p.Leu404His (blue) obtained after a 3-ns molecular dynamics simulation superimposed with wild-type TFII-I (tan). The residue Glu400 and the mutated residue p.Leu404His on TFII-I are highlighted. (c) Relative aberration frequencies according to WHO histotypes: GTF2I mutations (blue), arm-level copy number aberrations (orange) and number of mutations excluding GTF2I (purple).