Telomerase-dependent 3' G-strand overhang maintenance facilitates GTBP1-mediated telomere protection from misplaced homologous recombination

Abstract

At the 3'-end of telomeres, single-stranded G-overhang telomeric repeats form a stable T-loop. Many studies have focused on the mechanisms that generate and regulate the length of telomere 3' G-strand overhangs, but the roles of G-strand overhang length control in proper T-loop formation and end protection remain unclear. Here, we examined functional relationships between the single-stranded telomere binding protein GTBP1 and G-strand overhang lengths maintained by telomerase in tobacco (Nicotiana tabacum). In tobacco plants, telomerase reverse transcriptase subunit (TERT) repression severely worsened the GTBP1 knockdown phenotypes, which were formally characterized as an outcome of telomere destabilization. TERT downregulation shortened the telomere 3' G-overhangs and increased telomere recombinational aberrations in GTBP1-suppressed plants. Correlatively, GTBP1-mediated inhibition of single-strand invasion into the double-strand telomeric sequences was impaired due to shorter single-stranded telomeres. Moreover, TERT/GTBP1 double knockdown amplified misplaced homologous recombination of G-strand overhangs into intertelomeric regions. Thus, proper G-overhang length maintenance is required to protect telomeres against intertelomeric recombination, which is achieved by the balanced functions of GTBP1 and telomerase activity.

Figure 1.

TERT Repression Intensifies GTBP1-Knockdown Phenotypes in Tobacco. (A) Schematic representation of hnRNPA1 and tobacco GTBP1. Amino acid sequence identities between hnRNPA1 and GTBP1 are indicated in each domain. RRM1 and RRM2, RNA recognition motifs. aa, amino acid residue. (B) Schematic structures of TERT and GTBP1 RNAi binary vector constructs. The 35S:RNAi-TERT vector includes the inverted-repeat sequence of two different regions (660 to 984 bp and 984 to 1478 bp) of TERT cDNA (see Supplemental Figure 1B online). The 35S:RNAi-GTBP1 vector contains the inverted-repeat sequence of the 726- to 1070-bp region of GTBP1 cDNA. LB, left border; OCS ter, octopine synthase terminator; NOS ter, nopaline synthase terminator; NPTII, neomycin phosphotransferase II; HPTII, hygromycin phosphotransferase II; RB, right border. (C) Suppression of TERT (left panel) and GTBP1 (right panel) mRNAs in transgenic tobacco plants. Total leaf RNA isolated from wild-type (WT), T0 35S:RNAi-GTBP1, and four independent T0 35S:RNAi-TERT and 35S:RNAi-GTBP1/35S:RNAi-TERT transgenic lines was used for qRT-PCR. Expression levels of TERT and GTBP1 were normalized to that of the EF1α gene. Error bars represent ± se from three independent experiments. (D) Telomerase activities in wild-type and T0 35S:RNAi-TERT, 35S:RNAi-GTBP1, and 35S:RNAi-GTBP1/35S:RNAi-TERT transgenic tobacco plants. The telomerase activities were examined in callus tissue by a TRAP assay. (E) Morphology of GTBP1- and TERT-repressed transgenic tobacco plants. Representative 2-month-old (left and middle panels) and 3-month-old (right panel) wild-type and T0 RNAi transgenic plants. All four independent 35S:RNAi-GTBP1/35S:RNAi-TERT transgenic lines died after 2 months, whereas the 35S:RNAi-GTBP1 lines remain alive after 3 months. One-month-old RNAi transgenic plants that displayed mild phenotypic defects are shown in Supplemental Figure 1E online. (F) Morphological comparison of leaves from 2-month-old wild-type and RNAi transgenic plants. Genotypes are a combination of those indicated in the row and column. Leaf positions are indicated at the bottom of the figure. (G) Leaf number and stem lengths of 2-month-old wild-type and T0 RNAi-suppressed transgenic tobacco plants. [See online article for color version of this figure.]

Figure 2.

Reduction of 3′ G-Overhang Lengths and Enhanced t-Circle Formation in TERT/GTBP1 Double-Knockdown Telomeres. (A) Measurements of single-stranded 3′ G-overhangs and double-stranded telomere signals in wild-type (WT), 35S:RNAi-TERT, 35S:RNAi-GTBP1, and 35S:RNAi-GTBP1/35S:RNAi-TERT transgenic lines using pulse-field gel electrophoresis followed by in-gel hybridization. TaqI-restricted leaf genomic DNA from each transgenic plant was subjected to in-gel hybridization with a telomere (CCCTAAA)₄ probe under native pulse-field gel conditions to measure the single-stranded 3′ G-overhang telomere signal (left panel). Restricted leaf genomic DNA was rehybridized under denatured pulse-field gel conditions to measure double-stranded telomere signals (right panel). (B) Single-stranded 3′ G-overhang telomere signals in wild-type, 35S:RNAi-TERT, 35S:RNAi-GTBP1, and 35S:RNAi-GTBP1/35S:RNAi-TERT transgenic plants. Genomic DNA was purified and used for in-gel hybridization after native agarose gel electrophoresis. Quantified G-overhang signals were normalized to the levels of denatured total telomere signals determined by a telomere repeat fragment assay. (C) Pulse-field (left panel) and standard agarose (right panel) gel electrophoresis under native or denatured conditions followed by in-gel hybridization experiments in the presence (+) or absence (−) of Exo I. Quantified G-overhang signals were normalized to the levels of denatured total telomere signals. (D) Extrachromosomal t-circle formation in wild-type and RNAi-downregulated plants. Restricted leaf genomic DNA isolated from wild-type, 35S:RNAi-TERT, 35S:RNAi-GTBP1, and 35S:RNAi-GTBP1/35S:RNAi-TERT transgenic plants was subjected to 2-D pulse-field gel electrophoresis followed by in-gel hybridization with a ³²P-labeled (CCCTAAA)₄ probe. Linear telomeric DNA and extrachromosomal t-circles are indicated by arrows and arrowheads, respectively. (E) The TCA assay. Restricted tobacco leaf genomic DNA was treated with exonuclease V to remove linear DNAs. The TCA reaction was performed with enzyme-treated DNA mixtures and ϕ29 DNA polymerase. The reaction mixtures were separated on 1% agarose gels and subjected to in-gel hybridization with a ³²P-labeled (CCCTAAA)₄ probe. Quantified t-circle signals were normalized to the levels of denatured total telomere signals determined by a telomere repeat fragment assay. (F) Downregulation of TERT in cultured tobacco BY-2 suspension cells. BY-2 cells were transfected with the 35S:RNAi-TERT construct. The levels of TERT mRNA and telomerase activity in wild-type and two independent transfected BY-2 cell lines were determined by qRT-PCR and TRAP assays, respectively. Error bars represent ± se from three independent experiments. (G) Double-stranded telomere lengths in vector control and RNAi-transfected BY-2 cells during three rounds of subculture. Genomic DNA was purified and used for in-gel hybridization after denatured pulse-field gel electrophoresis. (H) Single-stranded 3′ G-overhang telomere signals in vector control and RNAi-transfected BY-2 cells during three rounds (1st, 2nd, and 3rd) of subculture. Genomic DNA was purified and used for in-gel hybridization after native agarose gel electrophoresis with (+) or without (−) Exo I treatment. Quantified G-overhang signals were normalized to the levels of denatured total telomere signals.

Figure 3.

Strand Invasion and ChIP Assays. (A) Schematic representation of GTBP1 and binding activities of GTBP1 to single-stranded telomere sequences. Different concentrations (0, 75, 150, and 300 nM) of bacterially expressed MBP-GTBP1 recombinant protein were subjected to gel mobility shift assays with radiolabeled, single-stranded (TTTAGGG)₃, (TTTAGGG)₄, (TTTAGGG)₅, (TTTAGGG)₆, or (TTTAGGG)₈ telomeric repeats. The “-” lanes contain single-stranded probes only. The relative binding activity was determined by the shifted band intensity. Asterisk indicates ³²P-labeled 5′ nucleotide end. aa, amino acids. (B) Strand invasion assay with different single-stranded telomeric repeats. GTBP1 (0, 75, 150, and 300 nM) was incubated with various ³²P-labeled single-stranded (TTTAGGG)n repeat probes (n = 3, 4, 5, 6, or 8) with T-vector plasmid containing a double-stranded (TTTAGGG)₇₀ telomere repeat. The relative level of invasion of the single-stranded telomeric probe into the plasmid was determined by the shifted band intensity. Asterisk indicates ³²P-labeled 5′ nucleotide end. (C) Binding activities of C-terminal deletion mutants of GTBP1 to single-stranded telomere sequences. MBP-GTBP1△C1^1-194 and MBP-GTBP1△C2^1-179 mutant proteins were subjected to gel mobility shift assays as described in (A). The relative binding activity was determined by the shifted band intensity. Asterisk indicates ³²P-labeled 5′ nucleotide end. (D) Telomere ChIP assay. The genomic DNA-protein complexes from wild-type (WT), 35S:HA-GTBP1, and 35S:HA-GTBP1/35S:RNAi-TERT BY-2 cells were fragmented by sonication and subjected to immunoprecipitation (IP) with an anti-HA antibody. The coimmunoprecipitated DNA was hybridized with ³²P-labeled (TTTAGGG)₇₀ or HRS60 repeated tobacco DNA sequences. The “-” lane indicates a negative control without anti-HA antibody. Quantified immunoprecipitation signals were normalized to 5% input signals. This experiment was independently repeated five times. Error bars represent ± se. [See online article for color version of this figure.]

Figure 4.

TERT/GTBP1 Double Knockdown Amplified Aberrant Intertelomeric HR. (A) Schematic representation of the telomere (TTTAGGG)₇₀:DNA-tag construct. LB, left border; RB, right border; BAR, herbicide Basta (glufosinate ammonium) resistant gene. (B) DNA-tag intertelomeric integration assay. The telomere-repeat:DNA-tag construct was transformed into tobacco plants. Chromosomal integration of the construct was detected by in situ PCR followed by FISH in Telomere:DNA-tag #17 and two independent Telomere:DNA-tag/35S:RNAi-TERT (lines a and b) transgenic chromosomes. The DNA-tag signal was amplified by in situ PCR with DNA-tag-specific primers. Chromosomal DNA was denatured and incubated with an Alexa Fluor 488–labeled DNA-tag-specific probe and a Texas red-dUTP–incorporated (TTTAGGG)₇₀ telomeric probe. The chromosomes were counterstained with 4′,6-diamidino-2-phenylindole (DAPI) and observed using fluorescence microscopy. The green signals indicate the DNA-tag sequence, whereas red signals indicate internal telomere sequences. The DNA-tag sequences merged with telomere sequences were counted. At least 50 nuclei from each T0 transgenic plant were observed. Bars = 5 μm. (C) DNA-tag intertelomeric integration assays were conducted with Telomere:DNA-tag/35S:RNAi-GTBP1 and two independent Telomere:DNA-tag/35S:RNAi-GTBP1/35S:RNAi-TERT (lines a and b) transgenic chromosomes. At least 50 nuclei from independent T0 transgenic plants were observed. Arrowheads indicate the DNA-tag sequences that overlap with telomeric signals. Bars = 5 μm. (D) Schematic representation of possible HR between the telomere (TTTAGGG)₇₀:DNA-tag and internal telomere sequences in Telomere:DNA-tag/35S:RNAi-GTBP1/35S:RNAi-TERT transgenic chromosomes. Red bars indicate telomere repeat sequences, and green bars indicate a DNA-tag sequence. As a result of HR, chromosomal copy numbers of telomere:DNA-tag increased in the Telomere:DNA-tag/35S:RNAi-GTBP1/35S:RNAi-TERT chromosomes. (E) Genomic PCR analysis. Leaf genomic DNA was isolated from Telomere:DNA-tag #17, Telomere:DNA-tag/35S:RNAi-TERT (independent lines a and b), Telomere:DNA-tag/35S:RNAi-GTBP1, and Telomere:DNA-tag/35S:RNAi-GTBP1/35S:RNAi-TERT (independent lines a and b) transgenic plants. DNA was analyzed by PCR using DNA-tag-specific primers. BAR is a loading control. [See online article for color version of this figure.]

Figure 5.

A Working Model of Possible Roles of GTBP1 and Telomerase Activity against Aberrant Telomeric Recombination GTBP1 binds to single-stranded telomere 3′ G-overhang ends and plays a role in inhibiting abnormal telomeric HR. When GTBP1 was repressed, aberrant telomere recombination occurred and telomere stability was partially disrupted. TERT/GTBP1 double knockdown amplified misplaced HR of G-strand overhangs into telomeric regions, which is associated with genome instability and early senescence of tobacco plants. A reduction of G-overhang length, in conjunction with GTBP1 repression, caused G-overhang uncapping from GTBP1-mediated telomere protection, due to the G-overhang length-dependent binding of GTBP1. Open circles indicate GTBP1.