The major reverse transcriptase-incompetent splice variant of the human telomerase protein inhibits telomerase activity but protects from apoptosis

Abstract

Human telomerase reverse transcriptase (hTERT; the catalytic protein subunit of telomerase) is subjected to numerous alternative splicing events, but the regulation and function of these splice variants is obscure. Full-length hTERT includes conserved domains that encode reverse transcriptase activity, RNA binding, and other functions. The major splice variant termed α+β- or β-deletion is highly expressed in stem and cancer cells, where it codes for a truncated protein lacking most of the reverse transcriptase domain but retaining the known RNA-binding motifs. In a breast cancer cell panel, we found that β-deletion was the hTERT transcript that was most highly expressed. Splicing of this transcript was controlled by the splice regulators SRSF11, HNRNPH2, and HNRNPL, and the β-deletion transcript variant was associated with polyribosomes in cells. When ectopically overexpressed, β-deletion protein competed for binding to telomerase RNA (hTR/TERC), thereby inhibiting endogenous telomerase activity. Overexpressed β-deletion protein localized to the nucleus and mitochondria and protected breast cancer cells from cisplatin-induced apoptosis. Our results reveal that a major hTERT splice variant can confer a growth advantage to cancer cells independent of telomere maintenance, suggesting that hTERT makes multiple contributions to cancer pathophysiology.

Figure 1. Schematic of hTERT α/βsplice variants

Top: hTERT protein domain structure. Below: Full-length hTERT mRNA, α/β splice variants, drawn as open boxes and approximately to scale. Black arrow indicates open reading frame. Picture adapted from (6).

Figure 2. The β-deletion splice variant can escape NMD and is associated with polysomes

A, Top: hTERT α+β+ and β-deletion mRNA splice variant expression in copies/μg RNA, normalized to copies/μg RNA GAPDH mRNA in UM-UC-3, Jurkat and BT-549 cells. Below: Relative telomerase activity (RTA) in 2500 cells/μl. B, Accumulation of hTERT splice variant and SRSF3 PTC+ mRNAs upon UPF1 versus control shRNA knockdown. Error bars are SD from at least 3 experiments. C, Western Blot of UPF1 and GAPDH in cells treated with UPF1 or control shRNAs. D, Top: Representative absorbance profile for RNA separated by velocity sedimentation through a 10–50% sucrose gradient. Positions of 40S, 60S, 80S, and polysomal peaks are indicated. Below: Agarose gel electrophoresis of RNA extracted from each fraction. 28S and 18S rRNAs are indicated. E, Abundance of SRSF3 and hTERT variant mRNA were measured by RT-qPCR and visualized as % SRSF3 PTC+/total SRSF3 and % β-deletion/α+containing hTERT mRNA.

Figure 3. The β-deletion isoform localizes to the nucleus, nucleuolus and mitochondria

β-deletion FLAG/GFP constructs were transfected into HeLa cells and stained with Hoechst 33342 and mitotracker deep red (Molecular Probes).

Figure 4. SRSF11, hnRNPH2 and hnRNPL regulate hTERT β-deletion splicing

A, Top: Structure of the pSpliceExpress-hTERT reporter gene. SRSF11 and hnRNPH2 binding sites are indicated with asterisks or plus signs, respectively. Below: qPCR primers for α+β+/β-deletion hTERT variants. RT primer for reporter gene-derived hTERT mRNA anneals to rat insulin exon 3. B, qRT-PCR of hTERT variants from RNA extracted from HEK293T cells co-transfected with either empty plasmid, splicing factor proteins and pSpliceExpress-hTERT. Bars represent mean of % β site exclusion. Error bars are SEM from 3 biological replicates assayed in triplicate. C, Western blot of HEK293T cell extracts extracts confirms overexpression of splicing factors.

Figure 5. The β-deletion isoform is a dominant-negative inhibitor of telomerase by sequestering hTR

A, hTR RNA associated with FLAG-hTERT constructs in GM847 cells. RNA from a parallel anti-FLAG immunoprecipitates from B was analyzed for presence of hTR and GAPDH RNA by RT-qPCR and repesented as % recovered over input RNA. B, RTA associated with anti-FLAG immunoprecipitates from GM847 cells transduced with indicated lentivirual constructs. C, Western blot of GM847 telomerase RNP immunoprecipitates used in Figure 5B. D, RTA in UM-UC-3 bladder cancer cells transduced with indicated lentiviral vectors. E, RNA from a parallel sample from D was extracted and analyzed for hTR and hTERT variant RNA normalized to GAPDH and vector control. Error bars in B and D represent SD from at least 3 experiments.

Figure 6. Expression of α/β splice variants, telomerase activity and telomere length in 50 breast cancer cell lines

A–D, Levels of indicated splice variants, expressed as transcript numbers normalized to GAPDH transcript numbers x 104. E, Relative expression of individual α/β hTERT splice variants relative to total amount of hTERT transcripts. F, RTA was measured by RQ-TRAP. G, Modal telomere length was determined by Southern blot from telomere restriction fragments (TRF). H, Regression of log(RTA) and TRF reveals a linear relationship (r = 0.487, P < 0.0001). Results in A–D, F and G are the average of 2–4 biological replicates and assayed in triplicate reactions (A–D, F); error bars represent SD.

Figure 7. The β-deletion protein protects basal breast cancer cells from apoptosis

BT-549, HCC1806 and HCC3153 that stably overexpressed β-deletion or vector control for less than 2 weeks were treated with the indicated cisplatin concentration for 48 h. Bars represent averaged luciferase activity of caspase 3/7 reporter over cell confluency of 4 replicates; error bars represent SEM.