Three RNA Proteins Found to Control How Telomerase Reaches and Maintains Chromosome Ends
A new Nature Communications study reveals NONO, SFPQ, and PSPC1 proteins guide telomerase from Cajal bodies to telomeres, with their loss causing progressive telomere shortening.
Summary
Researchers at the Children's Medical Research Institute, University of Sydney, have identified three RNA/DNA binding proteins — NONO, SFPQ, and PSPC1 — as critical regulators of telomerase trafficking. These proteins, collectively called the DBHS family, physically bind to the RNA component of telomerase (hTR) and help escort the enzyme from nuclear Cajal bodies to chromosome ends during cell division. When these proteins are depleted, telomerase gets stuck in Cajal bodies and fails to reach telomeres, ultimately causing progressive telomere shortening across multiple cancer cell lines. The findings add a new layer to our understanding of how telomere length is maintained and open potential avenues for targeting telomerase regulation in aging and cancer.
Detailed Summary
Telomere length maintenance is fundamental to both healthy aging and cancer biology. Telomerase — the enzyme that rebuilds chromosome-end repeats — must be correctly assembled, stabilized, and physically transported to telomeres during S-phase replication. Despite decades of research, the full cast of proteins orchestrating this trafficking process remained incomplete. This study in Nature Communications identifies the DBHS protein family (NONO, SFPQ, PSPC1) as essential components of that machinery, using a combination of bioinformatics, biochemistry, cell biology, and telomere-length tracking across multiple human cell lines.
The investigation began with a bioinformatics screen of the Cancer Dependency Map (DepMap, 23Q4), examining linear regression between expression of 402 RNA binding proteins and telomere length across 325 cancer cell lines. NONO emerged as the second-highest ranked RNA binding protein positively correlated with telomere length — just below hTERT itself — with SFPQ and PSPC1 also trending positive. This computational signal prompted mechanistic follow-up in five telomerase-positive cell lines: 293T, 293, HT1080, HCT116, and HeLa.
Immunoprecipitation experiments confirmed that all three DBHS proteins physically associate with hTR, the RNA template component of telomerase, in both 293T and HT1080 cells. Electrophoretic mobility shift assays (EMSAs) using immunopurified, exogenously expressed DBHS proteins demonstrated concentration-dependent binding to radiolabelled hTR, with super-shift patterns confirming specificity. TRAP (Telomere Repeat Amplification Protocol) assays on immunopurified NONO-Myc/FLAG from 293 and HT1080 cells, and on SFPQ- and PSPC1-Myc/FLAG from 293T cells, produced positive telomerase activity signals — demonstrating association with catalytically active telomerase, not merely the RNA alone. Critically, in ALT-positive U-2 OS cells (which lack endogenous hTERT and hTR), DBHS proteins only co-immunoprecipitated with hTERT when hTR was also present, pinpointing hTR as the interaction bridge.
The trafficking consequences of DBHS loss were measured by combined FISH for hTR and telomeres alongside immunostaining for the Cajal body marker coilin in S-phase cells. Depletion of any individual DBHS protein caused a statistically significant reduction in hTR/telomere co-localizations (p < 0.0001, Kruskal-Wallis test; n = 204 cells per condition across three experiments). Loss of NONO or SFPQ additionally caused a substantial increase in hTR-positive Cajal bodies, indicating telomerase retention at that station. PSPC1 depletion did not increase hTR-positive Cajal bodies but instead reduced coilin protein and mRNA levels, collapsing Cajal body structure — explaining the distinct phenotype without changing the net outcome of impaired telomerase delivery to telomeres.
Long-term depletion experiments confirmed that NONO and PSPC1 loss produces progressive telomere shortening across HCT116, HeLa, and HT1080 cells over multiple passages, while 293 and 293T cells showed a notable exception — suggesting cell-line-specific compensatory mechanisms exist. Domain mapping demonstrated that the conserved DBHS region governs telomerase recruitment and Cajal body exit, while SFPQ and PSPC1 polymerization capacity proved essential for cell viability. Taken together, the study establishes the DBHS family as a previously unrecognized layer of the telomerase trafficking machinery, with implications for understanding telomere-driven aging diseases and cancer.
Key Findings
- NONO was the second-highest RNA binding protein positively correlated with telomere length across 325 cancer cell lines in the DepMap dataset, ranking just below hTERT itself.
- All three DBHS proteins (NONO, SFPQ, PSPC1) co-immunoprecipitated with hTR and were present in active telomerase complexes, confirmed by positive TRAP signal from immunopurified NONO-Myc/FLAG in 293 and HT1080 cells.
- DBHS protein depletion caused a statistically significant reduction in telomere/hTR co-localizations in both 293T and HT1080 cells (p < 0.0001, Kruskal-Wallis test; n = 204 cells per condition across 3 experiments).
- NONO and SFPQ depletion caused telomerase retention in Cajal bodies (significant increase in hTR-positive Cajal body foci, p < 0.0001), while PSPC1 depletion reduced coilin protein levels and collapsed Cajal body structure instead.
- DBHS proteins interact with active telomerase exclusively via hTR: in U-2 OS cells lacking both components, DBHS proteins only co-immunoprecipitated with hTERT when hTR was co-expressed.
- Long-term NONO and PSPC1 depletion caused progressive telomere shortening across HCT116, HeLa, and HT1080 cell lines, with 293 and 293T cells displaying a notable exception suggesting cell-line-specific compensation.
- The conserved DBHS domain region (not the N-terminal G-quadruplex-binding region) was identified as necessary for telomerase recruitment and Cajal body exit, and SFPQ/PSPC1 polymerization was found essential for cell viability.
Methodology
The study used five telomerase-positive human cancer cell lines (293T, 293, HT1080, HCT116, HeLa) and ALT-positive U-2 OS cells as controls. Key techniques included immunoprecipitation with Western and northern dot blot, EMSA with radiolabelled hTR, TRAP assays on immunopurified protein complexes, combined telomere/hTR FISH with coilin immunostaining in S-phase cells (n = 150–204 cells per condition, 3 independent experiments), and long-term telomere length monitoring via TRF Southern blotting. Statistical comparisons used Kruskal-Wallis tests with appropriate post-hoc corrections; siRNA-mediated knockdown was used for short-term depletion and stable doxycycline-inducible shRNA for long-term studies.
Study Limitations
The study was conducted exclusively in cancer cell lines, which may not fully recapitulate normal cell biology or the telomere dynamics relevant to aging in primary somatic cells. A notable inconsistency was observed in 293 and 293T cells, which did not show progressive telomere shortening with long-term NONO depletion, indicating cell-line-specific compensatory mechanisms that are not yet understood. The authors declare no conflicts of interest, and the work was funded by the NHMRC, Cancer Institute NSW, and Tour De Cure.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.