Report

A Mammalian Telomerase-Associated Protein

See allHide authors and affiliations

Science  14 Feb 1997:
Vol. 275, Issue 5302, pp. 973-977
DOI: 10.1126/science.275.5302.973

Abstract

The telomerase ribonucleoprotein catalyzes the addition of new telomeres onto chromosome ends. A gene encoding a mammalian telomerase homolog called TP1 (telomerase-associated protein 1) was identified and cloned. TP1 exhibited extensive amino acid similarity to the Tetrahymena telomerase protein p80 and was shown to interact specifically with mammalian telomerase RNA. Antiserum to TP1 immunoprecipitated telomerase activity from cell extracts, suggesting that TP1 is associated with telomerase in vivo. The identification of TP1 suggests that telomerase-associated proteins are conserved from ciliates to humans.

Telomerase is an unusual RNA-dependent DNA polymerase that uses an RNA component to specify the addition of telomeric repeat sequences to chromosome ends (1). In humans the telomeric repeat is 5′-TTAGGG-3′, and the telomerase RNA contains a sequence complementary to this telomeric repeat (2, 3). The telomerase RNA template is required for telomere repeat synthesis in vitro and in vivo (46). Telomerase activity is differentially regulated in normal and immortalized cells. In germline cells telomeres are maintained, whereas several somatic tissues lack telomerase activity and undergo progressive telomere shortening with increasing age (1). In immortalized cells telomere length is stabilized and telomerase activity is often reactivated (79). Telomerase activity has also been detected in many cancers (1, 9).

Telomerase activities have been identified in several organisms, and their RNA components have been cloned from mouse, human, yeast, and several ciliates (1). Putative telomerase or telomere-associated proteins have been identified in yeast (10). The ribonucleoprotein complex responsible for telomerase activity, however, has been purified only in ciliates (1113). Purified Tetrahymena telomerase contains an RNA and two protein components, p80 and p95 (4, 13). The p80 component can be specifically cross-linked to telomerase RNA, whereas the p95 component binds and cross-links to single-stranded, telomeric DNA (11, 13). Antiserum to Tetrahymena p80 is also able to specifically immunoprecipitate telomerase activity from Tetrahymena cell extracts (13).

A cDNA encoding a Tetrahymena p80 homolog was identified from a murine colonic crypt expressed sequence tag (EST) database. Overlapping clones were subsequently identified from both a colonic crypt library and a bone marrow stromal cell library. The resultant contiguous cDNA sequence was 8160 nucleotides long and contained a single open reading frame (ORF) of 2629 amino acids. On the basis of functional criteria described below, the sequence was termed telomerase-associated protein 1 (TP1). The mouse sequence was used as a probe to identify contiguous human cDNA clones from a library prepared from a human colon carcinoma LIM1863 cell line. The mouse and human ORFs are 75% identical at the amino acid level (Fig. 1).

Fig. 1.

(A) Amino acid sequence of human TP1. Four NH2-terminal repeats of unknown significance are indicated by solid lines over the sequence (R1 to R4). Boxed regions show homology to Tetrahymena p80. Amino acids corresponding to the WD40 consensus sequence of the putative WD40 repeats are indicated as follows: Conserved “WD motif” amino acid pairs are in bold- face and underlined; conserved D residues are boxed; conserved “GH motif” residues are overlined. Abbreviations for the amino acid residues are as follows: A, Ala; C, Cys; D, Asp; E, Glu; F, Phe; G, Gly; H, His; I, Ile; K, Lys; L, Leu; M, Met; N, Asn; P, Pro; Q, Gln; R, Arg; S, Ser; T, Thr; V, Val; W, Trp; and Y, Tyr. The complete human and murine TP1 sequences have been deposited with GenBank (accession numbers U86136 and U86137, respectively). (B) (Top) The TP1 amino acid sequence is depicted schematically, with boxed regions as follows: NH2-terminal repeats 1 to 4 (R1-R4); gray boxes (1 to 3) show regions of homology between TP1 and Tetrahymena p80; and 16 putative WD40 repeats are boxed. (Bottom) Human (Hu) and murine (Mu) TP1 amino acid sequences were aligned to Tetrahymena p80 by means of a PAM250 matrix and then adjusted manually; Tetrahymena (Te) sequences identical to either human or mouse at each position are shown in boldface. The underlined, 90-amino acid segment in region 2 contains 46% identity between p80 and human TP1. Seventy amino acids of human TP1 (amino acids 484 to 554) also show 18% identity to the human Ro/SS ribonucleoprotein (the corresponding p80 sequence shows 29% identity) (31).

Three regions in the NH2-terminal one-third of TP1 exhibited amino acid similarity to Tetrahymena p80 (Fig. 1, A and B). Region 2 showed the most homology to p80 with 46% identity over 90 amino acids (Fig. 1B). Other common motifs were also identified in the mouse and human TP1 protein. A Procite search identified an adenosine triphosphate/guanosine triphosphate (ATP/GTP) binding motif, beginning at amino acid 1179 of the mouse sequence (Fig. 1B). A 30-amino acid repeated sequence at the NH2-terminus of TP1 (R1 to R4) showed weak homology to a protein of unknown function in the ononis yellow mosaic tymovirus (14) (Fig. 1A). The COOH-terminal third of TP1 contained 16 putative WD40 repeats (Fig. 1A) (15).

To determine the tissue distribution of the TP1 transcripts, we performed Northern (RNA) blot analysis on RNAs from adult mouse tissues. An ∼8-kilobase pair (kbp) transcript was detected at varying levels in multiple tissues with a murine TP1 RNA probe (Fig. 2A). Human tissues also showed widespread expression of TP1 mRNA and often contained two transcripts of 8 and 9.5 kbp (Fig. 2B). Expression of TP1 mRNA was also observed in all human and murine cell lines examined, including nontransformed primary cells such as human foreskin fibroblasts (16).

Fig. 2.

Expression of murine and human TP1 mRNA in tissues and cell lines. (A) Northern blots from the indicated murine tissues (Clontech, San Diego, California) were probed with 32P-labeled murine TP1 RNA probe (top) and with random-primed β-actin probe (bottom). (B) Northern blot of the indicated human RNA samples (Clontech), probed with human TP1 RNA probe as in (A). (C) Northern blot of mRNA from immortalized human and mouse cell lines. (Left) RNA from the indicated immortalized human cell lines (Clontech) was probed with human TP1 and β-actin as in (B); (right) RNA from the indicated murine cell lines was probed with murine TP1 as in (A). Molecular sizes are indicated on the left (in kilobase pairs).

To determine whether TP1 interacts with telomerase RNA, we used a three-hybrid screen, a version of the two-hybrid screen in which an RNA fusion to the hairpin sequence of phage MS2 is used to activate transcription through an interaction with the MS2 coat protein (17). The NH2-terminus of murine TP1, which contained the region with Tetrahymena p80 homology, was tested for interaction with mouse telomerase RNA (mTR) (Fig. 3). Coexpression of TP1 and mTR resulted in strong activation of the reporter genes HIS3 and lacZ and allowed growth on 3-aminotriazole (3-AT) medium and detection of β-galactosidase (β-Gal) activity (Fig. 3, A and B). TP1 also interacted with a mutant mTR, mTR-1, but not with several control RNAs such as yeast U2, U4, and U6 small nuclear RNAs (snRNAs), the yeast telomerase RNA TLC1, and antisense mTR (“RTm”) (Fig. 3, A and B). Another negative control, in which the MS2 tag was placed in an antisense direction relative to mTR (“2SM”), also did not allow growth on 3-AT medium (Fig. 3). Coexpression of TP1 and the iron response element, IRE, resulted in growth on 3-AT medium (18); however, β-Gal expression was not significantly above background levels (Fig. 3B). Only mTR and TP1 together resulted in the strong activation of both reporter genes in the three-hybrid assay, suggesting that TP1 is not a general RNA binding protein, but specifically binds telomerase RNA.

Fig. 3.

Three-hybrid analysis of the mTP1 and mTR interaction. (A) Various MS2-tagged RNAs as indicated were cotransformed with TP1 into the three-hybrid yeast strain L40-coat and patched on synthetic drop-out plates lacking uracil and leucine (−AT) and the same plates also lacking histidine and containing 5 mM 3-aminotriazole (+AT) (17). An interaction between the tagged RNA and TP1 allows growth on 3-AT. mTR is wild-type mouse telomerase RNA (19); RTm indicates the mTR sequence cloned in the antisense direction relative to the MS2 hairpins; 2SM refers to the full-length mTR in the sense direction and the MS2 hairpins in the antisense direction; mTR-1 is a mutant containing three nucleotide substitutions: C142→T, G202→C, and G227→A [relative to the transcription start site (19)]; U2, U4, and U6 are yeast snRNAs (U6 is not MS2-tagged) (32); and TLC1 is the full-length yeast telomerase RNA (16). (B) β-Galactosidase assays on transformants shown in (A). Liquid β-Gal assays were performed in triplicate and quantified as in (29), with 10−3 standard β-Gal units (420 nm) shown on the y axis. Error bars indicate standard deviation for each sample average. Another control RNA, the iron response element hairpin IRE (17), is shown for comparison to mTR and mTR-1. Although cotransformation of TP1 and IRE resulted in growth on 3-AT medium, β-Gal levels were not significantly above those of other negative controls (18). A positive control for the three-hybrid interaction, consisting of IRE and its binding protein IRP (IRE/IRP) (17), is shown at right.

To determine whether TP1 associated with telomerase RNA and telomerase activity in vivo, we transfected a murine TP1 cDNA containing two copies of an NH2- terminal Myc-epitope tag into murine neuroblastoma (N2A) cells. We then examined the cell lysates for immunoprecipitation of TP1 and telomerase activity using a monoclonal antibody to the Myc epitope (anti-Myc) (Fig. 4, A and B). Immunoblot analysis of the immunoprecipitates with anti-Myc confirmed that tagged TP1 was expressed in the transfected N2A cells (Fig. 4A). Telomerase activity was immunoprecipitated from Myc-TP1-transfected cells, but not from mock-transfected lysates, or Myc-TP1 lysates incubated with excess Myc peptide (Fig. 4B). Addition of a nonspecific peptide did not compete for the ability of anti-Myc to immunoprecipitate TP1 and telomerase activity (Fig. 4A, lane 6, and Fig. 4B, lane 11). We also examined whether anti-peptide antibodies against murine TP1 could immunoprecipitate telomerase activity from untransfected, immortalized mouse fibroblast (NIH 3T3) cell lysates (Fig. 4C). Telomerase activity was immunoprecipitated with an antipeptide antibody to the COOH-terminus of TP1 (Fig. 4C, lane 8) but was not immunoprecipitated by protein A resin alone or by preimmune serum (Fig. 4C, lanes 6 and 7). The immunoprecipitation of telomerase activity was specifically competed by incubation of the antipeptide antiserum with excess peptide to which the antibodies were raised (P3), but was not competed by another TP1 peptide, P1 (Fig. 4C, lanes 9 and 10). Reverse transcription and polymerase chain reaction (RT-PCR) analysis was used to establish that mTR was present when TP1, and telomerase activity, were immunoprecipitated (Fig. 4D).

Fig. 4.

Telomerase activity in TP1 immunoprecipitates. (A) Immunoblot developed with an antibody to Myc. S100 lysates (lanes 1 and 2) and immunoprecipitates (lanes 3 to 6) of N2A cells transfected with a Myc-epitope fusion gene containing the entire murine TP1 coding sequence (TP1; lanes 2, and 4 to 6) and mock-transfected cells (lanes 1 and 3) were resolved by SDS-polyacrylamide gel electrophoresis, transferred to nylon membrane, and probed with anti-Myc (33). The Myc-TP1 in the TP1 cell lysate and in the TP1 immunoprecipitate can be detected above the 220-kD protein marker indicated at right. Lanes 5 and 6: 10 μg of Myc peptide [sequence: 408-AEEQKLISEEDLLRKRREQLKHKLEQLRNSCA] (lane 5), or the nonspecific peptide P3 (lane 6, see also below) were incubated with anti-Myc before the immunoprecipitation to demonstrate the antibody specificity. No TP1 degradation products were detected below the region shown. (B) Telomerase activity in cells transfected with Myc-TP1 and mock-transfected cells as in (A). Protein G beads were assayed for telomerase activity through use of the telomere repeat amplification protocol (TRAP) (30), in the absence (odd-numbered lanes) and presence (even-numbered lanes) of 5 μg of ribonuclease A (RNase) (Sigma). Lanes 9 and 11: Myc peptide and P3 peptide (10 μg each) were added as competitor to the immunoprecipitation of Myc-TP1 lysates in lane 7. (C) Immunoprecipitation of telomerase activity by antipeptide antibodies against murine TP1. DEAE-purified NIH 3T3 S100 cell lysates were precipitated onto protein A-Sepharose beads adsorbed with the following: no antibody (lanes 1 and 6); normal rabbit preimmune sera (lanes 2 and 7); rabbit antisera raised against a COOH-terminal TP1 peptide, P3 [sequence: 1536-DPDASGTFRSCPPEALKDL] (lanes 3 and 8); antibodies to P3 peptide in the presence of 60 μg of P3 peptide (lanes 4 and 9); and antibodies to P3 peptide plus 60 μg of an NH2-terminal TP1 peptide, P1 [sequence: 390-RSKRRSRQPPRPQKTERPFSERGK] (lanes 5 and 10) (33). Both the supernatants after incubation with 3T3 S100 lysate (Sup; lanes 1 to 5) and the washed beads (IP; lanes 6 to 10) were assayed by TRAP and resolved on a 10% acrylamide nondenaturing gel. Lane 11: RNase treatment of the immunoprecipitate in lane 8. Similar results were obtained with both Myc-TP1-transfected N2A cell lysates and NIH 3T3 cell lysates (n = 6). (D) Analysis of mTR in the TP1 immunoprecipitates. Approximately 5 μl of protein A beads were phenol-extracted and subjected to RT-PCR with mTR-specific primers as in (19). Lanes 1 and 2: positive control lysate (NIH 3T3) and “no RNA” negative controls; lanes 3 to 6: no antibody, preimmune, immune, and immune plus P3 peptide [as in (C)].

These experiments demonstrate that TP1 interacts with telomerase RNA and that TP1 is associated with telomerase activity in vivo. Further biochemical analysis is required to determine whether TP1 is essential for telomerase function. The NH2-terminus of TP1, which shows the most homology to Tetrahymena p80, is sufficient to bind telomerase RNA in vivo. These results support the suggestion that although the telomerase RNAs of ciliates and mammals differ considerably in sequence (3, 4, 19), they may share sufficient secondary structure to be recognized by a conserved RNA-binding motif (20).

The TP1 sequence has features distinct from that of p80. Unlike p80, TP1 does not contain a putative zinc finger (13). The significance of the potential ATP/GTP-binding domain is unknown, because ATP is apparently not required for human or mouse telomerase activity (2, 21). An intriguing feature of the TP1 sequence is the WD40 repeats in the COOH-terminus. The β subunit of the heterotrimeric GTP-binding protein is composed of seven WD40 repeats that form a structure resembling a seven-bladed propeller (22). Similar WD40 repeats are present in a number of proteins that form multiprotein complexes (15). TP1 may form one or more propeller structures that mediate interactions with other telomerase or telomere-binding proteins, such as the telomeric repeat binding factor TRF (23).

The expression pattern of TP1 mRNA was not restricted to tissues and cell lines that express telomerase activity, but showed varied expression levels in several tissues (1, 9). Mouse and human telomerase RNAs are also expressed in multiple tissues (1, 3, 19, 24). TP1 may not be a rate-limiting or essential subunit, or perhaps associated proteins inhibit telomerase activity in some cell types. Alternatively, TP1 protein expression may be restricted to only those cell types that have telomerase activity (1, 9). Given the reactivation of telomerase activity in several transformed cells and tumor tissues (9), inhibitors of the association of TP1 with telomerase may yield useful reagents for inhibition of certain cancers. Our identification of TP1 suggests that telomerases between organisms as divergent as ciliates and humans are similar and should facilitate the characterization of other components of mammalian telomerase.

REFERENCES AND NOTES

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
    For testing the mTR-TP1 interaction in this screen, the full-length mouse telomerase RNA gene, mTR (19), was tagged at its 3′ end with two MS2 hairpins and inserted into the RPR1 promoter-terminator of pIIIEx426 containing a URA3-selectable marker (25). The RNA molecules “RTm,” “2SM,” TLC1, and mTR-1 were constructed in a similar manner. Yeast U2 and U4 snRNAs were similarly tagged with the MS2 hairpins, U6 was not tagged, and all were inserted into the URA3-selectable plasmid, pRS316 (26). For the GAL4 activation domain (AD)-TP1 fusion protein, an Ssp I-Xba I fragment of plasmid pCR3MycTag2 (1 to 871 amino acids of murine TP1) was blunt-ended and cloned into Bam HI, blunt-ended pACTII containing a LEU2-selectable marker (27). The RNA and GAL4 fusion plasmids were transformed (28) into strain L40-coat, which carries an integrated copy of the MS2 coat protein fused to the lexA DNA binding domain. Cotransformants were selected by culturing the cells on yeast agar plates (SD) lacking leucine (−LEU) and uracil (−URA) at 30°C (29). Cotransformants were patched onto SD-URA-LEU plates, and SD-URA-LEU-HIS plates containing 5 to 20 mM 3-AT (Sigma), and incubated for 3 days at 30°C. Liquid β-Gal assays were performed as in (29).
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
View Abstract

Navigate This Article