N-terminal domains of the human telomerase catalytic subunit required for enzyme activity in vivo

Abstract

Most tumor cells depend upon activation of the ribonucleoprotein enzyme telomerase for telomere maintenance and continual proliferation. The catalytic activity of this enzyme can be reconstituted in vitro with the RNA (hTR) and catalytic (hTERT) subunits. However, catalytic activity alone is insufficient for the full in vivo function of the enzyme. In addition, the enzyme must localize to the nucleus, recognize chromosome ends, and orchestrate telomere elongation in a highly regulated fashion. To identify domains of hTERT involved in these biological functions, we introduced a panel of 90 N-terminal hTERT substitution mutants into telomerase-negative cells and assayed the resulting cells for catalytic activity and, as a marker of in vivo function, for cellular proliferation. We found four domains to be essential for in vitro and in vivo enzyme activity, two of which were required for hTR binding. These domains map to regions defined by sequence alignments and mutational analysis in yeast, indicating that the N terminus has also been functionally conserved throughout evolution. Additionally, we discovered a novel domain, DAT, that "dissociates activities of telomerase," where mutations left the enzyme catalytically active, but was unable to function in vivo. Since mutations in this domain had no measurable effect on hTERT homomultimerization, hTR binding, or nuclear targeting, we propose that this domain is involved in other aspects of in vivo telomere elongation. The discovery of these domains provides the first step in dissecting the biological functions of human telomerase, with the ultimate goal of targeting this enzyme for the treatment of human cancers.

FIG. 1

Expression and telomerase activity of N-terminal hTERT mutants. (A) Total RNA was isolated from HA5 cell lines stably infected with vector (▵), FLAG-hTERT (▪, or FLAG-hTERT NAAIRS substitution mutants representative of nonessential (+212: ○), essential (+158, □), slow-growth (+110, ⧫), and biologically essential (+128, ●) and RT-PCR amplified with primers specific for hTERT by quantitative, real-time RT-PCR. The amount of transcript detected by fluorescence with FRET probes is plotted in arbitrary units against each PCR cycle (top panel). The housekeeping PBGD transcript was similarly measured to verify equivalent RNA addition per reaction (bottom panel), while H₂O (◊) was assayed in both reactions as a negative control. (B) A total of 0.2 μg of lysate prepared from the described HA5 cell lines was assayed for telomerase activity by TRAP assay. As a control, a portion of the lysate was heat treated (HT) to inactivate telomerase prior to assaying. The internal standard (IS) served as a positive control for PCR amplification. Catalytic activity for each sample was normalized with the internal standard and is expressed as a percentage of wild-type FLAG-hTERT activity, indicated as follows: ++ (>60%), + (60 to 15%), +/− (<15%), and − (extremely low or no detectable activity). Domain refers to the location of the mutant, as described in the text. Life span (M, mortal; I, immortal; S, slow growth) as defined in the text. (C) Biologic activity of hTERT mutants was measured by serially passaging HA5 cell lines to determine whther cells entered crisis like vector or immortalized like wild-type hTERT. Representative clones are shown: vector (▴), FLAG-hTERT (▪), +212 (○), +50 (□), +14 (⋄), and +128 (▵). (D) Telomere length of representative HA5 cells infected with NAAIRS mutants that result in an immortal, slow-growth, or finite life span was determined by releasing the terminal restriction fragments of genomic DNA isolated from the described cell lines at early passage (pd 2 to 3) with the restriction enzymes HinfI and RsaI. These fragments were resolved and detected by Southern hybridization with a telomeric probe. ✽, Sample +212 was underloaded. Domain refers to the location of the mutant, as described in the text.

FIG. 2

Absence of endogenous hTERT expression in telomerase-positive HA5 cell lines. Total RNA collected from HA5 cells at either early passage (pd 2 to 3) (A) or at late passage (pd >39) (B) was analyzed by RT-PCR with primers specific for either endogenous or ectopic hTERT or GAPDH (control for RNA content). Results with nonessential (+212, +326, +422, and +524), slow-growth (+14, +110, and +398), and biologically essential (+128) mutants are shown. CWR and LNCaP are prostate cancer cell lines expressing endogenous hTERT and serve as positive controls for endogenous and a negative control for ectopic hTERT expression. A water sample is used to control for contaminating DNA in the reaction mix.

FIG. 3

Domain structure of the N terminus of hTERT. Secondary structure of the N terminus of hTERT as predicted by the Jprep2 program (http://jura.ebi.ac.uk:8888/) is shown, cylinders represent α-helices or β-sheets. Essential domains I-A, I-B, II, and III, as well as the T-motif, are denoted above structure prediction. Shaded regions denote the DAT domain; L1 and L2 define the nonessential linker regions. Two structured regions, outside of defined domains, are indicated by dashed lines. Essential domains I, II, and III found in the N terminus of Est2p (19) or conserved regions GQ, CP, and QFP identified in TERT proteins by alignment (63) are shown below structure prediction.

FIG. 4

Protein stability and hTR binding of mutants within essential domains of hTERT. (A) Lysates from 293T cells transiently transfected with FLAG-hTERT-FLAG, wild type, or the indicated NAAIRS mutants were resolved by SDS-PAGE and examined by anti-FLAG Western blotting. An anti-actin Western blot was used to ensure equal protein loading. (B) Binding of hTR with hTERT was examined in vitro by coimmunoprecipitating ³⁵S-labeled FLAG-hTERT-FLAG (F-hTERT-F) with purified ³²P-labeled hTR by using anti-FLAG antibodies. Immunoprecipitates were separated by SDS-PAGE and exposed to autoradiograph. Input hTR was diluted 1/1,000 and hTERT 1/10 for visualization. (C) Binding of hTR with FLAG-hTERT-FLAG protein containing NAAIRS substitutions (+50, +152, +386, and +512) in essential domains I-A, I-B, II, and III, respectively, was similarly examined. As a control for nonspecific interactions, HDAC1-FLAG (HDAC1-F) was immunoprecipitated in the presence of labeled hTR. The positions of F-hTERT-F, HDAC1-F, and hTR are indicated left of gel.

FIG. 5

Expression and cell viability of slow-growth hTERT mutants. (A) Anti-FLAG Western blot of slow-growth (+14 and +110) and nonessential (+212 and +422) hTERT mutants transiently expressed in 293T cells. Equal loading is shown by the anti-actin blot. (B) Viability of slow-growth and immortal HA5 cells was determined by flow cytometry of annexin V and propidium iodide double-stained cells. The percentages shown are averages for three independent experiments.

FIG. 6

Protein stability and nuclear localization of hTERT with mutations in the DAT domain. (A) Anti-FLAG Western blot of lysates from 293T cells transiently transfected with biologically essential hTERT mutants +92, +122, wild-type hTERT, or control vector. The anti-actin blot shows equal protein loading. (B) Subcellular localization of DAT domain mutants transiently expressed in U2OS cells by indirect immunofluorescence. Localization of FLAG-hTERT-FLAG was visualized with an anti-FLAG antibody recognized by a fluorescein isothiocyanate-conjugated secondary antibody (green). Hoechst was used to stain nuclei (blue).

FIG. 7

Homomeric complex formation of essential and DAT domain hTERT mutants. (A) Immunoprecipitation of ³⁵S-labeled FLAG-hTERT-FLAG (F-hTERT-F) with either ³⁵S-labeled GST-hTERT or GST in the presence or absence of hTR and Ts oligonucleotide substrate with anti-GST or anti-FLAG antibodies as indicated. (B) ³⁵S-labeled GST-hTERT and F-hTERT-F, wild-type, or NAAIRS substitution in domains I-A, DAT, I-B, II, and III (+50, +92, +152, +386, and +512 mutants, respectively) were incubated together and immunoprecipitated with an anti-FLAG antibody to monitor protein association. As a control, an irrelevant FLAG-tagged protein (HDAC1-F) failed to coimmunoprecipitate GST-hTERT.