Identification and interpretation of TET2 noncanonical splicing site intronic variants in myeloid neoplasm patients

Abstract

Background: DNA hypermethylation and instability due to inactivation mutations in Ten-eleven translocation 2 (TET2) is a key biomarker of hematological malignancies. This study aims at characterizing two intronic noncanonical splice-site variants, c.3954+5_3954+8delGTTT and c.3954+5G>A. Methods: We used in silico prediction tools, reverse transcription (RT)-PCR, and Sanger sequencing on blood/bone marrow-derived RNA specimens to determine the aberrant splicing. Results: In silico prediction of both variants exhibited reduced splicing strength at the TET2 intron 7 splicing donor site. RT-PCR and Sanger sequencing identified a 62-bp deletion at the exon 7, producing a frameshift mutation, p.Cys1298*. Conclusion: This study provides functional evidence for two intronic TET2 variants that cause alternative splicing and frameshift mutation.

Keywords: TET2; in silico prediction; myeloid neoplasm; next‐generation sequencing (NGS); noncanonical splicing site; ten‐eleven translocation 2; tumor suppressor.

FIGURE 1

(A and B) In silico splicing prediction of TET2 intronic variants of (A) c.3954+5_3954+8del, and (B) c.3954+5G>A. Variants are indicated by red lines. Four splicing predictors were used; the impact on gene splicing is indicated with vertical blue bars. The predicted strength of canonical splice donor signals at the splicing junctions is reduced in both variants compared to the wild‐type reference transcript. The gains of splicing signals were observed at an internal cryptic splice donor site.

FIGURE 2

(A) Graphic representation showing TET2 splicing products with or without TET2 intronic variants (c.3954+5_3954+8delGTTT, c.3954+5G>A). Breakpoints at the cryptic and original 5′ splice site junctions are shown with blue triangles. The primer binding sites and subsequent PCR products in WT and mutant TET2 spliced mRNA are also shown. (B) RT‐PCR analysis on RNA from a negative control patient and a patient with TET2 c.3954+5_3954+8del variant. Gel electrophoresis image showing PCR products amplified using specific primers binding to Ex6 (forward) and Ex8 (reverse). Fragment size of 246 bp indicates TET2 wild‐type sequences (ex6+ex7+ex8) and fragment of 185 bp size (ex6+ partial ex7+ex8) indicates the TET2 splicing variant. Only wild‐type product was identified in the negative (WT) control. No visible product is shown in no template control (NTC).

FIGURE 3

Sanger sequencing of the 185 bp size PCR products obtained from a polyacrylamide gel. Nucleotide sequences of wild‐type TET2 transcript are shown at the bottom of each chromatogram. The coding exon sequences were highlighted in blue. The splicing junctions connecting exon 7 and exon 8, leading to partial depletion of exon 7, are indicated with black arrows. (A) Chromatogram of patient 1 fragment of TET2 c.3954+5_3954+8del. (B) Chromatogram of patient 2 fragment of c.3954+5G>A substitution.

FIGURE 4

(A and B) Examples of in silico splicing prediction of TET2 canonical splice donor variants of (A) c.3954+1G>A, and (B) c.3954+2T>A substitutions. Four splicing predictors were used; the impact on gene splicing is indicated with vertical blue bars. The predicted strength of canonical splice donor signals at the splicing junctions is reduced in both variants compared to the wild‐type reference transcript. The gains of splicing signals were observed at the same internal cryptic splice donor site as shown in Figure 1.