Report

Pot1, the Putative Telomere End-Binding Protein in Fission Yeast and Humans

See allHide authors and affiliations

Science  11 May 2001:
Vol. 292, Issue 5519, pp. 1171-1175
DOI: 10.1126/science.1060036

This article has a correction. Please see:

Abstract

Telomere proteins from ciliated protozoa bind to the single-stranded G-rich DNA extensions at the ends of macronuclear chromosomes. We have now identified homologous proteins in fission yeast and in humans. These Pot1 (protection oftelomeres) proteins each bind the G-rich strand of their own telomeric repeat sequence, consistent with a direct role in protecting chromosome ends. Deletion of the fission yeastpot1 + gene has an immediate effect on chromosome stability, causing rapid loss of telomeric DNA and chromosome circularization. It now appears that the protein that caps the ends of chromosomes is widely dispersed throughout the eukaryotic kingdom.

Telomeres, the protein-DNA complexes at chromosome ends, protect chromosomes from degradation and end-to-end fusion, and they serve as substrates for extension by telomerase. The telomeric DNA terminates with a single-stranded overhang of the G-rich strand in ciliated protozoa (1), yeast (2, 3), and mammalian cells (4–6). In budding yeast, the Cdc13 protein binds to this single-stranded DNA, protecting the chromosome end (7, 8) and recruiting telomerase (9). In the hypotrichous ciliate Oxytricha nova, an α-β protein heterodimer binds specifically to the single-stranded telomeric DNA (10–12) to form a ternary complex (13), the crystal structure of which has been solved (14). Euplotes crassus, another hypotrich, uses an α subunit but apparently no β subunit (15). These yeast and ciliate end-binding proteins have no obvious sequence similarity to each other, and no homologs have been reported in distant species such as mammals. Indeed, the t-loop DNA structure that can form at the ends of mammalian chromosomes (16) might have been thought to obviate the need for an end-binding protein.

Database searching has now revealed that theSchizosaccharomyces pombe open reading frame (ORF) SPAC26H5.06 contains a region of limited similarity to the α subunits of telomere proteins from Oxytricha and other ciliates (Fig. 1A). Conservation is most apparent in a region of about 120 amino acids near the NH2-termini of the proteins (Fig. 1B), where the S. pombe and O. nova sequences share 19% identity and 40% similarity. This region coincides with the most highly conserved domain within the ciliate sequences (42% amino acid identity and 61% similarity betweenO. nova and E. crassus). Because the ciliate telomere proteins are thought to act as protective caps at the ends of macronuclear chromosomes (10, 14), we named theS. pombe gene pot1 +(protection of telomeres).

Figure 1

Sequence comparison and morphological phenotype associated with deletion of pot1 +. (A) Multiple sequence alignment of the NH2-terminal regions of the α subunits of ciliate telomere proteins [Ec, Euplotes crassus (15); Sm, Stylonychia mytilis (39); Ot, Oxytricha trifallax (40); On, O. nova(11)] and yeast and human Pot1p (Hs, Homo sapiens; Sp, S. pombe). Starting and ending amino acid numbers are shown for each sequence. Sequences were aligned in ClustalW using the Blosum35 score table followed by minor manual adjustment. Shaded amino acids are conserved in four or more sequences. Single-letter abbreviations for 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. (B) Domain structure of the O. nova telomere protein and yeast and human Pot1p. Positions of OB folds (14) and functional domains (23) are depicted for the O. nova telomere protein. The positions of the regions aligned in (A) are indicated by open boxes. (C) Colony morphology of pot1 +,pot1 , trt1 +, andtrt1 after tetrad dissection and germination. (D) Phase-contrast micrographs ofpot1 + and pot1 cells 5 to 10 generations after germination. Scale bar, 5 μm. (E) Cells as in (D) but stained with 4′,6′-diamidino-2-phenylindole (DAPI) to reveal chromosome segregation defect inpot1 . Scale bar, 5 μm.

To examine whether pot1 + is indeed involved in telomere maintenance, we constructed a heterozygous diploidpot1 +/pot1 strain (17). Tetrad dissections revealed that thepot1 spores formed very small colonies compared with their pot1 + sisters (Fig. 1C). This immediate phenotype is in stark contrast to that observed for strains lacking the catalytic subunit of telomerase (trt1 ), which form normal-sized colonies upon sporulation (Fig. 1C) and only begin to show a growth defect after ∼75 generations, when telomeres have shortened considerably (18). For ∼10 generations after sporulation,pot1 colonies contained a large number of elongated cells (Fig. 1D), most of which failed to undergo further division. DNA staining revealed a high incidence of chromosome missegregation, often leading to daughter cells without any chromosomal DNA (Fig. 1E). These phenotypes diminished during successive restreaks; after ∼75 generations, the colony and cell morphology appeared to be wild type, a development reminiscent of the emergence of survivors in strains lacking functional telomerase (18).

Deletion of pot1 + had a marked effect on telomere stability. When genomic DNA from pot1 strains was analyzed by Southern blotting, telomeric sequences could not be detected (Fig. 2A). Using three probes that recognize distinct subregions of the telomere-associated sequence (TAS), we observed hybridization signals with the telomere-distal TAS3 probe (Fig. 2C), but not with the more telomere-proximal TAS1 and TAS2 probes (Fig. 2B) (19). These results indicate that ∼5 kb of terminal sequence had been lost.

Figure 2

Telomere phenotype inpot1 strains. (A) Colonies from the indicated strains were used to inoculate 10 ml of YES (yeast extract supplemented with amino acids) medium at 32°C. Cells were grown to late log phase, and genomic DNA was prepared. After digestion of DNA (∼20 μg) with Eco RI, samples were subjected to 1.1% agarose gel electrophoresis, transferred to a nylon membrane, and hybridized to a telomeric probe. As a loading control, a probe against the single-copy pol α gene was used. (B) Genomic DNA (∼20 μg) was digested with Nsi I, fractionated by 0.8% agarose gel electrophoresis, transferred to a nylon membrane, and hybridized to a TAS2 probe (18). (C) The blot in (B) was stripped (34) and hybridized to a TAS3 probe.

One way the pot1 cells might survive without telomeres is through chromosome circularization (18, 20). When DNA from pot1 strains was digested with Not I (Fig. 3A) and analyzed by pulsed-field gel electrophoresis, the terminal C, I, L, and M fragments of chromosomes I and II were missing, whereas internal fragments were unperturbed (Fig. 3B). Hybridization with gene-specific probes revealed the presence of two new bands corresponding to C+M and I+L, the products of chromosome circularization (Fig. 3C). Circularization of all three chromosomes has been found only in S. pombestrains that fail to maintain chromosome ends, supporting the conclusion that pot1 + is required for telomere maintenance.

Figure 3

Pulsed-field gel electrophoresis and detection of terminal fragments. (A) Schematic showing the location of Not I restriction sites on S. pombe chromosomes (41). Only terminal fragments on chromosomes I and II are labeled. (B) Ethidium bromide–stained pulsed-field gel containing Not I–digested DNA from twopot1 + strains and six independentpot1 isolates. Genomic DNA was prepared, Not I–digested, and fractionated by pulsed-field gel electrophoresis as described (34). (C) DNA from the gel shown in (B) was transferred to a nylon membrane and hybridized to gene-specific probes on the C, I, L, and M fragments.

The rapid loss of terminal DNA in pot1 strains and the sequence similarity of Pot1p with telomere end-binding proteins from ciliated protozoa suggested that Pot1p has a direct role in the protection of telomeres. To investigate whether Pot1p binds telomeric DNA, we expressed the protein as a His6-fusion protein inEscherichia coli (21). In an electrophoretic mobility shift assay, the purified Pot1p fusion bound specifically to the G-rich strand of S. pombe telomeric DNA, but not to the complementary C-rich strand or double-stranded telomeric DNA (Fig. 4A) (22). The binding affinity of Pot1p for telomeric repeats at the 3′ end of an oligonucleotide was greater by a factor of ∼5 relative to its affinity for the same telomeric repeats flanked by nontelomeric sequences (19).

Figure 4

DNA-binding specificity of SpPot1p. (A) C-strand (CGTAACCGTAACCCTGTAACCTGTAACCTGTAACCGTGTAACC) and G-strand (GGTTACACGGTTACAGGTTACAGGTTACAGGGTTACGGTTACG) were 5′ 32P-labeled using T4 polynucleotide kinase and [γ-32P]adenosine triphosphate. Duplex was generated by annealing equimolar amounts of radiolabeled C-strand and unlabeled G-strand. Pot1p (50 ng) was incubated with the indicated DNA substrates (1 ng) for 15 min at 20°C in 10 μl of 25 mM Hepes (pH 7.5), 1 mM EDTA, 50 mM NaCl, 5% glycerol, and 2.5 μM PBoli109 (CCGTAAGCATTTCATTATTGGAATTCGAGCTCGTTTTCGA) as nonspecific competitor. Complexes were analyzed by electrophoresis at 4°C through a 4 to 20% tris-borate EDTA gel (Invitrogen) run at 150 V for 80 min. The Pot1p-DNA complex is indicated by an open arrow. (B) Same as (A) except that the added protein (100 ng) contained truncated Pot1p as well as full-length protein. Truncated Pot1p-DNA complex is indicated by a solid arrow. (C) Binding of hPot1p to human C-strand (CCCTAA)5, G-strand (TTAGGG)5, and duplex DNA.

Truncated forms of Pot1p, resulting from either premature termination or proteolytic degradation, copurified with the full-length protein. The affinity of COOH-terminally truncated Pot1p for DNA was about an order of magnitude greater than that of the full-length protein (apparent dissociation constant ∼10 nM versus ∼100 nM), but the protein retained the same specificity (Fig. 4B). Further purification and analysis by mass spectroscopy showed that the prominent complex (solid arrow in Fig. 4B, lane d) was attributable to the binding of a 22-kD NH2-terminal fragment of Pot1p. Increased DNA binding has also been observed with the COOH-terminally truncated α subunit of the O. nova telomere protein (23), further supporting a functional relationship to Pot1p. Intramolecular inhibitory sequences play a role in modulating the DNA binding characteristics of many transcription factors [e.g., (24)], and a similar situation may pertain to Pot1p.

The identification of related telomere proteins in ciliates and fission yeast prompted us to search for homologous proteins in other eukaryotes. A BLAST search with the S. pombe Pot1 protein sequence revealed the product of human cDNA FLJ11037 as the top-ranked match (P = 3 × 10−6). We refer to the protein encoded by this cDNA as hPot1p (human Pot1 protein). Sequence alignments of hPot1p with the fission yeast and ciliate proteins revealed the highest conservation near the NH2-terminus, where the S. pombe and human proteins share 26% identity and 48% similarity (Fig. 1A). Over the same region, the human and O. nova protein sequences are 23% identical and 39% similar.

Human POT1 mRNA was detected in all tissues examined (25). This finding is consistent with the idea thathPOT1 is a housekeeping gene required to ensure the integrity of chromosome ends in all cells. In contrast, human telomerase reverse transcriptase mRNA is detected primarily in immortalized and germ line cells, but not in most somatic cells (26–28).

The hPOT1 gene was cloned from ovary cDNA and was found to encode a 71-kD polypeptide. Recombinant hPot1p (with an NH2-terminal His6 tag) was expressed inE. coli and purified (29). As with the S. pombe Pot1 protein (SpPot1p), a fraction of the hPot1p was lacking COOH-terminal sequences as a result of degradation or premature termination. However, hPot1p produced in E. coli showed the same DNA binding specificity as full-length hPot1p from in vitro translation reactions (19). In bandshift assays, hPot1p bound to the G-rich strand of human telomeric DNA (Fig. 4C). In contrast, binding was not observed with the complementary C-rich strand or with double-stranded telomeric DNA.

Telomeric DNA binding by both the S. pombe and human Pot1 proteins was unaffected by the presence of a 60-fold excess of boiled herring sperm DNA and a 2000-fold excess of an oligonucleotide of nontelomeric sequence (19). To further investigate the sequence specificity, we tested whether the G-rich strand of telomeric DNA from different species could serve as a substrate in DNA-binding assays. In a side-by-side comparison, SpPot1p bound the human telomeric sequence less well than it bound the S. pombe sequence (Fig. 5A). In competition experiments, a 1000-fold excess of unlabeled S. pombe sequence abolished binding to the radiolabeled substrate, whereas unlabeled human andO. nova telomeric DNAs reduced binding by only ∼50% and <2%, respectively (Fig. 5B). Similarly, hPot1p showed only weak binding to the S. pombe sequence (Fig. 5C), which also was not an efficient competitor (Fig. 5D). In contrast, the presence of a 1000-fold excess of the O. nova sequence reduced binding to less than 25%. In summary, each protein shows specificity for binding its own telomeric DNA sequence.

Figure 5

Substrate specificity ofS. pombe and human Pot1p. (A) Binding of SpPot1p to S. pombe and human G-strands (DNA sequences as in Fig. 4). (B) Binding reactions (10 μl) contained SpPot1p (50 ng) and radiolabeled S. pombe G-strand (15 pg; GGTTACACGGTTAC- AGGTTACAGGTTAC- AG) in the presence of 10-, 100-, and 1000-fold excess of unlabeled S. pombe (G-strand sequence as in Fig. 4A), human (TTAGGG)5, orO. nova (GGGGTTT- TGGGGTTTTGGGGT) DNA. (C) Binding of hPot1p to S. pombe and human G-strands. (D) Binding of hPot1p to human G-strand under the same conditions as in (B).

Biochemical and structural data have long suggested a role for the Euplotes and Oxytricha telomere proteins in protecting the ends of chromosomes (12). However, as these organisms are not amenable to genetic studies, demonstration of such a capping function in vivo had been lacking. By deleting the S. pombe pot1 + gene, we have now provided evidence that this group of proteins plays a pivotal role in preventing rapid degradation of chromosome ends in vivo. Loss of Pot1p led to immediate chromosome instability, whereas the absence of functional telomerase causes gradual telomere shortening over many generations without an immediate effect on chromosome stability and cell viability (18, 26). It therefore appears that, at least in S. pombe, Pot1p is more important than telomerase for telomere maintenance in the short term.

In addition, Pot1p may be involved in regulating the access of telomerase and/or other enzymes to the chromosome terminus. Reconstitution of the Oxytricha α-β–telomeric DNA complex prevents extension by telomerase in vitro, consistent with a function for the α-β complex in the regulation of telomere length (30). (Note that we have not found an S. pombe or human counterpart to the β subunit by homology searching.) InSaccharomyces cerevisiae the single-stranded telomeric DNA binding protein Cdc13 recruits telomerase to the chromosome end via interactions with the telomerase component Est1p (9). Sequence alignments of Cdc13p with Pot1p and ciliate telomere proteins failed to detect obvious similarities. However, Cdc13p may nevertheless belong to the same family of proteins, because OB (oligonucleotide/oligosaccharide binding) folds, which are seen in the crystal structure of the Oxytricha α-β-DNA complex and are presumably present in Pot1p (Fig. 1B), are identified reliably only by structural analysis and not by sequence homology (31). It will hence be important to probe for interactions between Pot1p and telomerase in S. pombe and human cells and to determine whether these proteins fulfill analogous functions to Cdc13p.

It now appears that at least in mammalian cells, telomeres may exist in at least three interconvertible states: as t-loops, Pot1p-bound, and engaged with telomerase (32). Although these different states could correlate with particular stages of the cell cycle, they need not be mutually exclusive. As indicated above, Pot1p may be involved in actively recruiting telomerase. Alternatively or in addition, the 3′ end of telomeric DNA could be capped by Pot1p within the structure of a t-loop, which would prevent the chromosome end from being used as a primer for conventional DNA synthesis (33). Now that a key protein that binds at the chromosome end appears conserved across widely diverged eukaryotes, it will be an interesting challenge to determine how it contributes to the various structures and functions of the chromosome end.

  • * To whom correspondence should be addressed. E-mail: thomas.cech{at}colorado.edu

REFERENCES AND NOTES

View Abstract

Navigate This Article