Alternative splicing and nonsense-mediated decay regulate telomerase reverse transcriptase (TERT) expression during virus-induced lymphomagenesis in vivo

Abstract

Background: Telomerase activation, a critical step in cell immortalization and oncogenesis, is partly regulated by alternative splicing. In this study, we aimed to use the Marek's disease virus (MDV) T-cell lymphoma model to evaluate TERT regulation by splicing during lymphomagenesis in vivo, from the start point to tumor establishment.

Results: We first screened cDNA libraries from the chicken MDV lymphoma-derived MSB-1 T- cell line, which we compared with B (DT40) and hepatocyte (LMH) cell lines. The chTERT splicing pattern was cell line-specific, despite similar high levels of telomerase activity. We identified 27 alternative transcripts of chicken TERT (chTERT). Five were in-frame alternative transcripts without in vitro telomerase activity in the presence of viral or chicken telomerase RNA (vTR or chTR), unlike the full-length transcript. Nineteen of the 22 transcripts with a premature termination codon (PTC) harbored a PTC more than 50 nucleotides upstream from the 3' splice junction, and were therefore predicted targets for nonsense-mediated decay (NMD). The major PTC-containing alternatively spliced form identified in MSB1 (ie10) was targeted to the NMD pathway, as demonstrated by UPF1 silencing. We then studied three splicing events separately, and the balance between in-frame alternative splice variants (d5f and d10f) plus the NMD target i10ec and constitutively spliced chTERT transcripts during lymphomagenesis induced by MDV indicated that basal telomerase activity in normal T cells was associated with a high proportion of in-frame non functional isoforms and a low proportion of constitutively spliced chTERT. Telomerase upregulation depended on an increase in active constitutively spliced chTERT levels and coincided with a switch in alternative splicing from an in-frame variant to NMD-targeted variants.

Conclusions: TERT regulation by splicing plays a key role in telomerase upregulation during lymphomagenesis, through the sophisticated control of constitutive and alternative splicing. Using the MDV T-cell lymphoma model, we identified a chTERT splice variant as a new NMD target.

Figure 1

Alternative transcripts of chTERT identified in three avian cell line. At the top, we show a schematic diagram of telomerase protein, showing major conserved protein motifs, including reverse transcriptase domains 1 and 2 and the A, B, C, D and E motifs, and of the chTERT gene, with its 16 exons shown as gray boxes. The PCR primers are shown as arrows below the diagram. Each alternative transcript is shown on the right, with the splicing event depicted as a clear gray box for insertion of the exon cassette; dark lines indicate deletion, black boxes indicate intron retention and white boxes indicate the deletion of part of an exon. Positions of premature stop codons (PTCs) are indicated by black triangles. On the left, we show the name of the spliced transcripts, the presence or absence of a PTC with its position relative to the 3' exon-exon junction and represented as a function of cell line. The name of the transcript is indicated on the left and is coded as follow: iXec for insertion of exon cassette X, dXf for full deletion of exon X, dXp for partial deletion of exon X, iXp for retention of part of intron X and iXf for insertion of full intron X.

Figure 2

chTERT in-frame isoforms are non functional in the in vitro telomerase assay. (A) Schematic diagram of the 3 reconstituted in-frame isoforms of chTERT inserted into pcDNA under the control of the T7 promoter. Amino acids deleted or inserted in each isoform are indicated and those belonging to the RT motif are underlined. Plasmids were used to produce recombinant in vitro-translated proteins that were incubated in the presence of in vitro-transcribed vTR or chTERT before the TRAP assay. (B) The histogram shows telomerase activity relative to the value obtained with vTR in the presence of the full-length constitutively spliced chTERT transcript arbitrary set at 100%. (C) Histograms show telomerase activities for a mixture (1:1) of in vitro-translated full-length chTERT and the indicated isoform in the presence of vTR or chTR. The values obtained with the mixture of full-length chTERT and isoform are expressed relative to the value for the full-length chTERT, arbitrarily set at 100%. The histogram shows the mean and standard deviation obtained from 3 biological analyses.

Figure 3

Downregulation of the i10ce variant by NMD. (A) Upf1 depletion increases levels of the chTERT splice variant i10ce. Levels of i10ce are expressed relative to total chTERT levels. Peak areas of i10ec were obtained by capillary electrophoresis analysis of the PCR products targeting exon 10, generated from cDNA from 1 well of P6 transfected with 25 pmol or 50 pmol of siUPF-1. The histogram shows the mean and standard deviation obtained from 4 biological analyses. All values are expressed as a fold difference with respect to "no silencing" (NS), for which the value was fixed at 1. (B) siUpf1 knockdown of Upf1. The histogram shows Upf1 mRNA levels, calculated by RT-PCR analysis. Upf1 levels in each sample were normalized with respect to GAPDH levels for the corresponding sample. The values obtained for "no silencing" (NS) LMH cells were set at 100%.

Figure 4

Regulation of chTERT splicing involving variants 5 and 10 during lymphomagenesis in vivo. Proportions of constitutively spliced chTERT transcript (in blue) and alternatively spliced variants d10f+i10ec (in yellow) and the d5f variant (in green), as shown in the panels on the right and left, respectively. The proportions correspond to the peak area obtained by capillary electrophoresis analysis of PCR targeting exon 5 (A) or 10 (B), performed on cDNA extracted from sorted CD4⁺ T cells sampled from GaHV-2-infected chickens at 5 different time points after infection, as indicated on the x-axis. The curve shows telomerase activity, which was obtained by summing the peak areas corresponding to the elongation products (right y-axis) (C) Proportions of d10f (green) and i10ec (orange) variants, as indicated above the graph and corresponding to 100% of alternative splice variants10 at each time point after infection, as shown in (B).

Figure 5

Model of the splicing regulation of chTERT during lymphomagenesis.