Ustilago maydis telomere protein Pot1 harbors an extra N-terminal OB fold and regulates homology-directed DNA repair factors in a dichotomous and context-dependent manner

Abstract

The telomere G-strand binding protein Pot1 plays multifaceted roles in telomere maintenance and protection. We examined the structure and activities of Pot1 in Ustilago maydis, a fungal model that recapitulates key features of mammalian telomere regulation. Compared to the well-characterized primate and fission yeast Pot1 orthologs, UmPot1 harbors an extra N-terminal OB-fold domain (OB-N), which was recently shown to be present in most metazoans. UmPot1 binds directly to Rad51 and regulates the latter's strand exchange activity. Deleting the OB-N domain, which is implicated in Rad51-binding, caused telomere shortening, suggesting that Pot1-Rad51 interaction facilitates telomere maintenance. Depleting Pot1 through transcriptional repression triggered growth arrest as well as rampant recombination, leading to multiple telomere aberrations. In addition, telomere repeat RNAs transcribed from both the G- and C-strand were dramatically up-regulated, and this was accompanied by elevated levels of telomere RNA-DNA hybrids. Telomere abnormalities of pot1-deficient cells were suppressed, and cell viability was restored by the deletion of genes encoding Rad51 or Brh2 (the BRCA2 ortholog), indicating that homology-directed repair (HDR) proteins are key mediators of telomere aberrations and cellular toxicity. Together, these observations underscore the complex physical and functional interactions between Pot1 and DNA repair factors, leading to context-dependent and dichotomous effects of HDR proteins on telomere maintenance and protection.

Fig 1. U maydis Pot1 and other basidiomycetes harbor an extra N-terminal OB fold domain.

A. Domain organization of Pot1 homologs from major fungal branches. B. Multiple sequence alignment of the N-terminal region of Basidiomycete Pot1 homologs. The alignment and secondary structure predictions were generated using PROMALS3D and displayed using MView. C. A homology model of the putative OB-N domain from Paxillus ammoniavirescens (PaPot1) (generated from SWISS-MODEL) is superimposed on the OB-A domain of human RPA70. Note that the five beta strands (labeled as S1 to S5) in the two structures are spatially well aligned. D. EMSA analysis of the DNA-binding activities of UmPot1 truncation variants. The domains included in each variant are illustrated at the top and the binding assays shown at the bottom. 5’-labeled G4 oligonucleotide was used as the probe.

Fig 2. U maydis Pot1 binds to Rad51 and regulate the strand-exchange activity of Rad51.

A. Anti-FLAG agarose resin was pre-bound with Pot1-FLAG and tested for interaction with purified Rad51 or MBP-Brh2/Dss1 complex. The bound proteins were eluted with FLAG₃ peptides and the levels of Rad51, Brh2 and Pot1 in the eluates analyzed using anti-Rad51, anti-MBP, and anti-FLAG antibodies. B. The interaction between Pot1-FLAG and Rad51 was analyzed in the absence or presence of Benzonase. C. The ability of Rad51 to mediate strand exchange between a 70-mer ssDNA and a labeled 39-mer dsDNA was analyzed in the absence or presence of ATP. The substrates and products are marked by an open and a closed arrowhead, respectively. Asterisks designate the positions of the 5’ ³²P label. D. The effects of Pot1 and Tay1 on strand exchange were analyzed in the absence or presence of Rad51. The substrates and products are marked by an open and a closed arrowhead, respectively. Asterisks designate the positions of the 5’ ³²P label. An additional band is detected in Pot1-containing reactions (marked by a filled circle). While the identity of this is unclear, it is not Rad51-dependent and unlikely to be relevant to the strand-exchange reaction.

Fig 3. Pot1 N-terminal deletion impairs telomere maintenance.

A. Serial dilutions of the indicated U. maydis strains were directly spotted onto a YPD plate or first subjected to UV irradiation prior to spotting. The plates were incubated for 2 days (-UV) or 3 days (+UV) and the results photographed. B. Serial dilutions of cultures of UCM350 and pot1ΔN that had been passaged for the indicated numbers of streaks were spotted onto a YPD plate and incubated for 2 days. C. Chromosomal DNAs from two independently propagated clonal cultures of UCM350 and pot1ΔN were subjected to TRF Southern analysis at the indicated passages (i.e., Streak 3, 6 and 10). Mean TRF lengths from the pairs of assays were calculated and plotted at the bottom. P-values were calculated using two-tailed Student’s t tests and displayed. D. Chromosomal DNAs from two independently propagated clonal cultures of UCM350 and pot1ΔN strains were subjected to STELA analysis at Streak 10. Mean STELA lengths were calculated from 3 independent assays of each DNA sample and shown at the bottom.

Fig 4. pot1 deficiency triggers growth arrest and multiple telomere aberrations.

A. Serial dilutions of the indicated strains were grown on YPA (crg1 promoter on) and YPD (crg1 promoter off) plates for 2 and 3 days, and the results photographed. B. Chromosomal DNAs from three independently generated pot1^crg1 and pot1ΔN^crg1 strains grown in YPA and YPD were subjected to telomere restriction fragment analysis. The overall telomere contents of the samples were quantified and plotted on the right. P-values were calculated using two-tailed Student’s t tests. C. Chromosomal DNAs from three independently generated pot1^crg1 and pot1ΔN^crg1 strains grown in YPA and YPD were subjected to C-circle and G-circle assays. The relative levels of C- and G-circles from the three mutants were quantified and plotted at the bottom. P-values were calculated using two-tailed Student’s t tests. D. Chromosomal DNAs from three independently generated pot1^crg1 and pot1ΔN^crg1 strains grown in YPA and YPD were subjected to in-gel hybridization analysis of ssDNA on both the G- and the C-strand. E. The levels of ssDNA from the in-gel hybridization assays were quantified (N = 3) and plotted. P-values were calculated using two-tailed Student’s t test.

Fig 5. Telomere aberrations in pot1-deficient U. maydis are greatly suppressed by rad51Δ and brh2Δ.

A. Chromosomal DNAs from the indicated wild type and mutant strains grown in YPA and YPD were subjected to telomere restriction fragment analysis. Two independently generated pot1^crg1 strains were analyzed in all the assays in this figure. B. Chromosomal DNAs from the indicated wild type and mutant strains grown in YPA and YPD were subjected to in-gel hybridization analysis of ssDNA on both the G- and the C-strand (left). The levels of ssDNA from two independent sets of assays were quantified and plotted (right). P-values were calculated using two-tailed Student’s t tests and displayed. C. Chromosomal DNAs from the indicated wild type and mutant strains grown in YPA and YPD were subjected to C-circle assays (left). The relative levels of C-circle signals from these strains were quantified and plotted (right). Data are from two independent sets of assays. P-values were calculated using two-tailed Student’s t tests and displayed.

Fig 6. Telomere repeat-containing RNAs are elevated by pot1-deficiency and suppressed by rad51Δ and brh2Δ.

A. RNAs from the indicated wild type and mutant strains grown in YPD were isolated by acid phenol extraction and subjected to Northern analysis using sequentially probes that detect G-strand repeats, C-strand repeats, and 26S rRNA. B. The ratios of G- and C-strand RNA signals to the 26 rRNA signals from Northern analysis were calculated, and then normalized against the ratio of the pot1^crg1 mutant and plotted. The normalized values for the pot1^crg1 G- and C-strand RNA levels were each set to 100. C. RNAs from the indicated wild type and mutant strains grown in YPD were isolated by acid phenol extraction, treated with Turbo DNase (Thermo Fisher Inc.), and subjected to dot blot analysis using sequentially probes that detect the G-strand repeats, C-strand repeats, and 26S rRNA. D. The ratios of G- and C-strand signals to the 26 rRNA signals from dot blot analysis (n = 2 or 3) were calculated, and then normalized against the ratio of the pot1^crg1 mutant and plotted. P-values were calculated using two-tailed Student’s t tests and displayed.