Cleveland Clinic offers a comprehensive and competitive total rewards program that recognizes the needs of a diverse workforce, provides individuals and families with meaningful choices and lets employees change work locations without experiencing interruptions in benefit coverage. The latest Tweets from Cleveland Clinic MD (@CleClinicMD). News and insights for physicians and clinical professionals from @ClevelandClinic. Research, case studies and CME opportunities. Cleveland, Ohio, United States.
doi: 10.1016/j.exger.2011.11.012
PMID: 22178633
This article has been cited by other articles in PMC.
Abstract
Telomeres are regions of repetitive DNA at the end of eukaryotic chromosomes, which prevent chromosomal instability. Telomere shortening is linked to age-related disease including Alzheimer's disease (AD) and has been reported to be reduced in leukocytes of AD patients. The aim of the present study was to measure telomere length in monocytes of patients with AD or mild cognitive impairment (MCI) compared to healthy subjects. Our data show significant shorter telomere length in AD patients (6.6 ± 0.2 kb; p = 0.05) compared to controls (7.3 ± 0.2 kb). Telomere length of MCI patients did not differ compared to healthy subjects (7.0 ± 0.2 kb). We observe a strong correlation between telomere length and age (p = 0.01, r = − 0.38), but no association between telomere length and Mini-Mental State Examination score. In conclusion, the telomere length is age-dependent in monocytes and decreased in AD patients, which could mean that the AD pathology may contribute to telomere length shortening. The high variability of telomere lengths in individuals suggests that it will not be useful as a general biomarker for AD. However, it could become a biomarker in personalized long-term monitoring of an individuals’ health.
Keywords: Monocyte, Alzheimer, Blood, Diagnose, Telomere
Highlights
► Alzheimer‘s disease is a severe neurodegenerative disorder of the brain. ► Laboratory diagnosis of AD is difficult and not well established. ► Monocytes play a role in inflammation and may reflect AD pathology. ► Telomer length is reduced in monocytes of AD patients. ► Telomer length could be a putative (personalized) biomarker for diagnosis.
1. Introduction
Alzheimer's disease (AD) is a progressive neurodegenerative disorder which is characterised by cognitive impairment, memory loss and characteristic pathological changes in the brain, like senile plaques and neurofibrillary tangles (Burns et al., 2002). To complement diagnosis an intense search is underway to identify disease-specific biomarkers in the cerebrospinal fluid (CSF), blood plasma, and blood cells. To date, three biomarkers have been established in CFS: beta-amyloid1–42 (Aβ), total tau, and phospho-tau-181 (Humpel, 2011). So far no specific blood biomarkers have been established, despite an intense research on proteins and genes of blood cells (Humpel, 2011).
Telomeres are short and highly conserved hexanucleotide repeats (TTAGGG) found at the end of eukaryotic chromosomes, which prevent end-to-end fusions and other structural and functional cell abnormalities. During aging 50–150 bp of telomeric DNA is lost with each proliferation cycle (Allsopp et al., 1992). Shorter telomere length of leukocytes has been linked to age-related diabetes, cardiovascular and heart disease and also to an elevated risk of neurodegenerative disease including dementia (Honig et al., 2006; Panossian et al., 2003; Tentolouris et al., 2007; von Zglinicki et al., 2000). In particular, telomere shortening in white blood cells and altered immune function as a possible result has been linked to AD (Honig et al., 2006; Panossian et al., 2003; Thomas et al., 2008). Immune cells, like monocytes are further associated with Aβ depositions and are capable of phagocytosing Aβ (Fiala et al., 2007).
The objective of this study was to investigate, if telomere length in monocytes is altered in patients with AD or MCI compared to healthy subjects. If so, these results will provide a basis to further investigate monocytic involvement in the pathology of AD, which could help to use telomere length to distinguish between healthy subjects and MCI, or AD patients.
2. Methods
2.1. Selection of patients
Healthy subjects and patients suffering from AD or MCI were recruited from the Department of Psychiatry in Innsbruck or Klagenfurt, Austria. All groups were assessed by the same diagnostic procedure. Psychiatrists clinically examined all subjects, performed a standardized neurological examination, neuropsychological tests (Mini-Mental State Examination, MMSE), reviewed medical records, and conferred with referring physicians for all patients. MCI was diagnosed according to the Petersen criteria (Petersen et al., 2001). Probable AD was diagnosed according to the National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer's Disease and Related Disorders Association criteria (McKhann et al., 1984). The geriatric depression scale (GDS) was applied to all participants. Magnetic resonance imaging was performed for all participants. Subjects were excluded when they suffered from another mental disease, any kind of metabolic decompensation or had any signs of inflammation. The study was approved by the ethical committee of Innsbruck Medical University.
2.2. Monocyte collection
Monocytes were isolated as described recently in detail (Hochstrasser et al., 2010). Briefly, EDTA blood (10 ml) was collected during normal routine clinical assessments and processed within 3 h. Plasma and peripheral mononuclear cells (PBMCs) were separated from whole blood on a continuous Biocoll gradient (1.077 g/ml, Biochrom, Germany) after centrifugation (400 ×g, 30 min, room temperature). Two-thirds of the upper plasma phase and the interphase with the PBMCs, which is visible as a white stratum between plasma phase and Biocoll, were carefully removed. Plasma was directly frozen at − 80 °C until use. PBMCs were washed in 50 ml phosphate-buffered saline (PBS), centrifuged (250 ×g, 6 min) and the pellet was dissolved in PBS with 1% bovine serum albumine (BSA). Monoyctes were isolated by negative magnetic isolation as described by the manufacturer (Miltenyi Biotech, Germany). Briefly, PBMCs were incubated with a cocktail of different biotinylated antibodies (CD3, CD7, CD16, CD19, CD56, CD123, CD235a) for 10 min on ice. Then anti-biotin magnetic beads were added, incubated for further 15 min on ice, washed and the cells applied onto MACS MS columns (Miltenyi Biotech, Germany) on a strong magnet. The non-labelled monocytes were eluted and collected. Finally, cells were frozen at − 80 °C until use.
2.3. Telomere length assay
Telomere length analysis was performed by the TeloTAGGG telomere length assay kit (Roche, Austria) as described by the manufacturer. Briefly, genomic DNA was extracted from monocytes using the QIAamp DNA mini kit (Qiagen, Austria) according to the manufacturer's instructions. DNA was digested by the restriction endonucleases HinfI and RsaI for 2 h at 37 °C. Following DNA digestion, the DNA fragments were separated by gel electrophoreses (0.8% agarose gel, 50 V, 3 h). Then, DNA was transferred to a positively charged nylon membrane (Roche, Austria) by capillary Southern blotting with 20 × saline-sodium citrate (SSC) buffer (overnight, room temperature). DNA was fixed to the membrane by UV light for 5 min, washed with 2 × SSC buffer and air dried. DNA fragments were hybridized with telomeric specific digoxygenin (DIG)-labeled hybridization probe (3 h, 42 °C), incubated with anti-DIG-alkaline phospahatase for 30 min at room temperature and detected with CDP-Star chemiluminescent substrate. The signals were analyzed using a CCD imager. Telomere length is given as the average terminal restriction fragment (TRF) length and the signal intensity is plotted in function of migrating distance for each sample. Mean TRF length was quantified by integrating the signal intensity of the TRF bands on the blot as a function of its mean molecular weight, which is determined based on the molecular weight standard (Fig. 1A).
Telomere length analysis. (A) A representative Southern Blot shows telomere lengths (TRF) from monocytes of control subjects (Co), Alzheimer's disease (AD) and mild cognitive impaired (MCI) patients. A size marker on the left gives the size of DNA fragments in kilo bases (kb) (B) Scatter plot shows the telomere length of controls, MCI, and AD patients. Mean telomere length from AD patients is shorter (*p = 0.05) than from controls. Value in parenthesis gives the number of samples.
2.4. SearchLight Multiplex ELISA
SearchLight Multiplex ELISAs (Aushon Biosystems) for plasma pro-inflammatory cytokines tumor necrosis factor-alpha (TNFα), RANTES, interleukin-1 alpha (IL-1α), and monocyte chemotactic protein-3 (MCP3) were performed as described by us in detail (Marksteiner et al., 2011).
2.5. Statistical analysis
Sample size considerations for our study are based on the study of Panossian et al. (2003) who found a fairly large difference in monocyte telomere length between AD patients and healthy controls compared to the standard deviation (SD) in each of the two groups (mean difference = 0.91 kb, SD < 0.5 kb, giving rise to an effect size d > 1.8; d = mean difference/SD). Our sample size of 19 patients with AD, 18 with MCI and 14 healthy controls was chosen such that in a two sample t-test (AD vs controls, MCI vs. controls) effect sizes d > 1 can be detected with 80% power at a 5% level of significance. Considerations for ANCOVA are similar (detectable effect size d ≈ 1.1, 1.2, 1.3 for one, two and three covariates, respectively).
Statistical analysis was performed with analysis of variance (ANOVA) followed up by post-hoc pairwise comparisons of groups using Fisher's least significant difference (LSD) method. The ability of the telomere length to discriminate between diagnostic groups was tested by analysis of covariance (ANCOVA). ANCOVA was conducted in order to adjust for age, sex, and GDS. The correlation of telomere length to age or MMSE was assessed by ANCOVA, where p < 0.05 and was considered as statistically significant.
3. Results
Subjects’ characteristics are presented in Table 1. Healthy controls did not show a difference in sex, age, and GDS compared to MCI and AD patients (Table 1). Controls had an MMSE score of 28.4 ± 0.4, which was significantly different from AD but not from MCI patients (Table 1). Pro-inflammatory markers were measured by SearchLight ELISA in plasma, but no significant difference was observed between controls, MCI and AD patients (Table 2).
Table 1
n | m/f | Age | MMSE | GDS | |
---|---|---|---|---|---|
Control | 14 | 6/7 | 71.9 ± 1.9 | 28.4 ± 0.4 | 6.1 ± 1.2 |
MCI | 19 | 6/13 ns | 71.2 ± 1.6 ns | 27.4 ± 0.4 ns | 9.5 ± 1.4 ns |
AD | 18 | 4/14 ns | 74.7 ± 1.7 ns | 19.6 ± 1.1 *** | 10.1 ± 1.4 ns |
Monocytes were isolated from healthy controls, mild cognitive impaired (MCI), and Alzheimer's disease (AD) patients. Values are given as mean ± SEM. n gives the number of samples and the number of respective male/female patients (m/f) is given. The table gives the age of the patients in years, the Minimental State Examination (MMSE) and the geriatric depression scale (GDS). Statistical analysis was performed by one-way ANOVA with a Fisher LSD post-hoc test. *** p < 0.001; ns not significant.
Table 2
Inflammatory marker in plasma of controls, MCI, and AD patients.
TNFα [pg/ml] | RANTES [ng/ml] | IL-1α [pg/ml] | MCP3 [pg/ml] | |
---|---|---|---|---|
Control | 8.2 ± 1.1 (7)– | 10.5 ± 3.2 (14)– | 3.1 ± 0.9 (7)– | 3.4 ± 1.5 (7)– |
MCI | 11.4 ± 2.4 (9) ns | 10.4 ± 2.8 (18) ns | 2.3 ± 0.4 (9) ns | 1.5 ± 0.1 (9) ns |
AD | 10.7 ± 1.8 (12) ns | 12.8 ± 4.0 (18) ns | 2.5 ± 0.3 (12) ns | 2.2 ± 0.3 (12) ns |
Plasma was taken from healthy controls, mild cognitive impaired (MCI), and Alzheimer's disease (AD) patients. Plasma was analyzed for detection of inflammatory markers (tumor necrosis factor-alpha, TNFα; RANTES; interleukin-1 alpha, IL-1α; monocyte chemotactic protein-3, MCP-3) by SearchLight Multiplex ELISA. Values are given as mean ± SEM pg/ml or ng/ml. The number of analyzed samples is given in parenthesis. Statistical analysis was performed by one-way ANOVA. ns, not significant.
The telomere length in monocytes was determined by Southern blotting and was found to be between 6 and 7 kilobases (Fig. 1A). In order to measure size differences, the blots were scanned and the DNA size was blotted against the running distance related to 21.2 kb, which yielded in a formula with a very high regression coefficient (R2 = 0.99). The size of the monocyte telomere length was calculated accordingly, and the monocyte telomere length was found to be 7.3 ± 0.2 kb in healthy subjects (n = 14; Fig. 1B). For MCI patients, the telomere length was 7.0 ± 0.2 kb (n = 19) and AD patients had a mean monocyte telomere length of 6.6 ± 0.2 kb (n = 18) (Fig. 1B). When telomere length was adjusted for age, the analysis of covariance revealed an all over trend towards shorter telomeres when controls were compared to MCI and AD patients (p = 0.08). Telomere length of MCI patients did not significantly differ from telomere lengths of healthy controls (p = 0.15; Bonferroni-corrected: p = 0.30). AD patients showed a significance towards shorter telomeres (p = 0.03; Bonferroni-corrected: p = 0.05) compared to healthy controls. A significant negative correlation was seen between telomere length and age in monocytes of all groups (Fig. 2A; r = − 0.38, p = 0.01), while there was no correlation between telomere length and MMSE score (Fig. 2B; r = 0.21, p = 0.13). There was no significant association between the telomere length and GDS (r = − 0.08; p = 0.6), or sex (p = 0.34) adjusted for age.
Correlation of telomere length to age or to cognitive function (MMSE). The telomere length of monocytes is plotted against (A) age in years (r = − 0.376, p = 0.007, analysis of covariance), or (B) MMSE score (r = 0.213, p = 0.134, analysis of covariance). r = correlation coefficient.
4. Discussion
Our findings indicate a significant reduction in telomere length in monocytes of AD patients compared to healthy subjects.
Short telomeres have been found as a marker for biological aging and several studies have shown an association between short telomeres and age-related disease such as dementia (Honig et al., 2006; Panossian et al., 2003; Thomas et al., 2008; von Zglinicki et al., 2000). Telomere shortening was first experimentally demonstrated in fibroblasts and has been shown to be associated with aging (Baird, 2006). Here, we report that monocytes of AD patients show shorter telomeres, which are marginally significant when adjusted for age. MCI patients displayed no difference in monocyte telomere length compared to control subjects. A correlation between peripheral blood leukocytes and cerebellum telomere length, a correlation between leukocyte telomere length and cognitive performance, and increased telomerase activity and decreased proliferation activity in lymphocytes has been demonstrated in AD patients (Lukens et al., 2009; Valdes et al., 2010). We did not observe a correlation between cognitive performance (MMSE) and monocyte telomere length, which is in agreement with others (Valdes et al., 2010). However, they show lower MMSE scores in T cell telomere length of AD patients compared to controls. In order to measure the pro-inflammatory status, we analyzed four well established markers in plasma but did not find a change in inflammation in AD and MCI patients.
We observed a significant negative correlation between telomere length and age, which points to an age-related telomere shortening, rather than to a disease-specific event. In consistence, cerebellum telomere length correlates inversely to age in AD patients (Lukens et al., 2009). It is suggested, that telomere length is not the major determinant of AD, since individuals would develop AD as soon as telomeres shorten to a certain size (Lukens et al., 2009). However, it is still not completely clear if and how telomere shortening contributes to AD. It cannot be determined whether these changes in telomere length are the result in response to AD brain pathology, such as tissue damage, or if reduced telomere length somehow contributes to AD. We found a correlation between age and telomere length and a reduced telomere length in AD patients suggesting that age is the major contributor to telomere shortening. However, AD and other neurodegenerative diseases, such as vascular dementia and Parkinson's disease are associated with telomere length shortening (von Zglinicki et al., 2000; Guan et al., 2008). It seems likely that long-term chronic inflammation and/or oxidative stress contributes to telomere shortening in monocytes (von Zglinicki et al., 2000; Guan et al., 2008). In addition telomere length highly correlates to aging, and telomeres might be more vulnerable in old age and then contribute to AD development. Further research regarding these findings in large, longitudinal prospective studies is needed. Interestingly, microglia also display shorter telomeres in the AD brain, suggesting that these cells undergo early replicative senescence, which could be induced by the severe amyloid plaque overload in AD (Flanary et al., 2007). Monocytes migrate through the blood-brain barrier in AD and convert into microglia cells in the brain and microglial activation has been reported in association with amyloid-plaques in the AD brain (Floris et al., 2002; Hickman and El Khoury, 2010; Malm et al., 2005). Additionally, increased expression of chemokine receptors and cytokines in peripheral blood mononuclear cell of AD patients has been shown (Reale et al., 2008).
In conclusion, we observed a strong correlation between age and telomere length and we found a significant shorter telomere length in monocytes of AD patients. This may suggest that the AD pathology further enhances telomere shortening. Regardless of how telomere length is associated to AD, we assume that telomere length of monocytes could serve as a biomarker for AD in a personalized long-term monitoring of an individuals’ health. However, there are limitations for a general biomarker, because telomere length is strongly correlated to age and the variability of the telomere length between individuals is very high.
Acknowledgments
This study was supported by Austrian Science Funds (L429-B05). TH was supported by the European Community (Moodinflame, Early diagnosis, treatment and prevention of mood disorders targeting the activated inflammatory response system. FP7-Health-2007 Project no: 222963). We thank Dr. Georg Kemmler for help with statistical analysis and Ursula Kirzenberger-Winkler for excellent technical assistance. All authors reported NO biomedical financial interests or potential conflict of interest.
References
Allsopp R.C., Vaziri H., Patterson C., Goldstein S., Younglai E.V., Futcher A.B., Greider C.W., Harley C.B. Telomere length predicts replicative capacity of human fibroblasts. Proc. Natl. Acad. Sci. U. S. A. 1992;89:10114–10118.[PMC free article] [PubMed] [Google Scholar]
Baird D.M. Telomeres. Exp. Gerontol. 2006;41:1223–1227. [PubMed] [Google Scholar]
Burns A., Byrne E.J., Maurer K. Alzheimer's disease. Lancet. 2002;360:163–165. [PubMed] [Google Scholar]
Fiala M., Cribbs D.H., Rosenthal M., Bernard G. Phagocytosis of amyloid-beta and inflammation: two faces of innate immunity in Alzheimer's disease. J. Alzheimers Dis. 2007;11:457–463. [PubMed] [Google Scholar]
Flanary B.E., Sammons N.W., Nguyen C., Walker D., Streit W.J. Evidence that aging and amyloid promote microglial cell senescence. Rejuvenation Res. 2007;10:61–74. [PubMed] [Google Scholar]
Floris S., Ruuls S.R., Wierinckx A., van der Pol S.M., Döpp E., van der Meide P.H., Dijkstra C.D., De Vries H.E. Interferon-beta directly influences monocyte infiltration into the central nervous system. J. Neuroimmunol. 2002;127:69–79. [PubMed] [Google Scholar]
Guan J.Z., Maeda T., Sugano M., Oyama J., Higuchi Y., Suzuki T., Makino N. A percentage analysis of the telomere length in Parkinson's disease patients. J. Gerontol. A Biol. Sci. Med. Sci. 2008;63:467–473. [PubMed] [Google Scholar]
Hickman S.E., El Khoury J. Mechanisms of mononuclear phagocyte recruitment in Alzheimer's disease. CNS Neurol. Disord. Drug Targets. 2010;9:168–173.[PMC free article] [PubMed] [Google Scholar]
Hochstrasser T., Weiss E., Marksteiner J., Humpel C. Soluble cell adhesion molecules in monocytes of Alzheimer's disease and mild cognitive impairment. Exp. Gerontol. 2010;45:70–74.[PMC free article] [PubMed] [Google Scholar]
Honig L.S., Schupf N., Lee J.H., Tang M.X., Mayeux R. Shorter telomeres are associated with mortality in those with APOE epsilon4 and dementia. Ann. Neurol. 2006;60:181–187. [PubMed] [Google Scholar]
Humpel C. Identifying and validating biomarkers for Alzheimer's disease. Trends Biotechnol. 2011;29:26–32.[PMC free article] [PubMed] [Google Scholar]
Lukens J.N., Van Deerlin V., Clark C.M., Xie S.X., Johnson F.B. Comparisons of telomere lengths in peripheral blood and cerebellum in Alzheimer's disease. Alzheimers Dement. 2009;5:463–469.[PMC free article] [PubMed] [Google Scholar]
Malm T.M., Koistinaho M., Pärepalo M., Vatanen T., Ooka A., Karlsson S., Koistinaho J. Bone-marrow-derived cells contribute to the recruitment of microglial cells in response to beta-amyloid deposition in APP/PS1 double transgenic Alzheimer mice. Neurobiol. Dis. 2005;18:134–142. [PubMed] [Google Scholar]
Marksteiner J., Kemmler G., Weiss E.M., Knaus G., Ullrich C., Mechtcheriakov S., Oberbauer H., Auffinger S., Hinterhölzl J., Hinterhuber H., Humpel C. Five out of 16 plasma signaling proteins are enhanced in plasma of patients with mild cognitive impairment and Alzheimer's disease. Neurobiol. Aging. 2011;32:539–540.[PMC free article] [PubMed] [Google Scholar]
McKhann G., Drachman D., Folstein M., Katzman R., Price D., Stadlan E.M. Clinical diagnosis of Alzheimer's disease: report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer's Disease. Neurology. 1984;34:939–944. [PubMed] [Google Scholar]
Panossian L.A., Porter V.R., Valenzuela H.F., Zhu X., Reback E., Masterman D., Cummings J.L., Effros R.B. Telomere shortening in T cells correlates with Alzheimer's disease status. Neurobiol. Aging. 2003;24:77–84. [PubMed] [Google Scholar]
Petersen R.C., Doody R., Kurz A., Mohs R.C., Morris J.C., Rabins P.V., Ritchie K., Rossor M., Thal L., Winblad B. Current concepts in mild cognitive impairment. Arch. Neurol. 2001;58:1985–1992. [PubMed] [Google Scholar]
Reale M., Iarlori C., Feliciani C., Gambi D. Peripheral chemokine receptors, their ligands, cytokines and Alzheimer's disease. J. Alzheimers Dis. 2008;14:147–159. [PubMed] [Google Scholar]
Tentolouris N., Nzietchueng R., Cattan V., Poitevin G., Lacolley P., Papazafiropoulou A., Perrea D., Katsilambros N., Benetos A. White blood cells telomere length is shorter in males with type 2 diabetes and microalbuminuria. Diabetes Care. 2007;11:2909–29015. [PubMed] [Google Scholar]
Thomas P., O' Callaghan N.J., Fenech M. Telomere length in white blood cells, buccal cells and brain tissue and its variation with ageing and Alzheimer's disease. Mech. Ageing Dev. 2008;129:183–190. [PubMed] [Google Scholar]
Valdes A.M., Deary I.J., Gardner J., Kimura M., Lu X., Spector T.D., Aviv A., Cherkas L.F. Leukocyte telomere length is associated with cognitive performance in healthy women. Neurobiol. Aging. 2010;31:986–992.[PMC free article] [PubMed] [Google Scholar]
von Zglinicki T., Serra V., Lorenz M., Saretzki G., Lenzen-Grossimlighaus R., Gessner R., Risch A., Steinhagen-Thiessen E. Short telomeres in patients with vascular dementia: an indicator of low antioxidative capacity and a possible risk factor? Lab. Invest. 2000;80:1739–1747. [PubMed] [Google Scholar]
1528–1535 Nucleic Acids Research, 1998, Vol. 26, No. 6
1998 Oxford University Press
The yeast telomere length regulator TEL2 encodes a protein that binds to telomeric DNA Rama S. Kota and Kurt W. Runge* The Lerner Research Institute, Cleveland Clinic Foundation, Department of Molecular Biology, NC 20, 9500 Euclid Avenue, Cleveland, OH 44195, USA Received September 12, 1997; Revised and Accepted January 26, 1998
ABSTRACT
TEL2 is required for telomere length regulation and viability in Saccharomyces cerevisiae. To investigate the mechanism by which Tel2p regulates telomere length, the majority (65%) of the TEL2 ORF was fused to the 3′-end of the gene for maltose binding protein, expressed in bacteria and the purified protein used in DNA binding studies. Rap1p, the major yeast telomere binding protein, recognizes a 13 bp duplex site 5′-GGTGTGTGGGTGT-3′ in yeast telomeric DNA with high affinity. Gel shift experiments revealed that the MBP–Tel2p fusion binds the double-stranded yeast telomeric Rap1p site in a sequence-specific manner. Analysis of mutated sites showed that MBP–Tel2p could bind 5′-GTGTGTGG-3′ within this 13 bp site. Methylation interference analysis revealed that Tel2p contacts the 5′-terminal guanine in the major groove. MBP–Tel2p did not bind duplex telomeric DNA repeats from vertebrates, Tetrahymena or Oxytricha. These results suggest that Tel2p is a DNA binding protein that recognizes yeast telomeric DNA. INTRODUCTION Telomere length regulation, regulation of the number of repeated DNA sequences that make up the chromosome ends in many organisms, has recently been recognized as a critical function in human cells. Telomere DNA repeats are lost from somatic cell telomeres as cells divide (1), leading to the proposal that telomeres both activate cellular senescence when telomeres become too short and eliminate many transformed cells by causing chromosome destruction by loss of telomere function. The telomeric DNA repeats are the only sequences required for telomere function in yeast and humans (2–6), so loss of these repeats is expected to cause chromosome ends to behave as double-strand (ds)DNA breaks (7,8). Experiments with transformed cells grown in culture show a correlation between the number of cell divisions, decreased telomere length, the onset of ‘crisis’ and end-to-end chromosome fusions consistent with loss of telomere function (9,10). Thus regulation of telomere length is critical in humans for determining the lifespan of individual cells. While some of the factors that lengthen and shorten telomeres have been
identified (11–14), how telomere length is maintained at an appropriate length in human germ cells, where lengthening mechanisms are active, is still unknown. Recent evidence suggests that telomere length is measured by monitoring the number of telomere binding proteins (14,15; A.Ray and K.W.Runge, submitted for publication), but how these protein molecules are counted has not been determined. One approach to this problem is to identify regulators of telomere length and determine how they function. The yeast Saccharomyces cerevisiae has been an ideal model organism for investigating the structure and function of telomeres. Both humans and yeast have telomeres composed of short repeated sequences that are elongated by telomerase, a reverse transcriptase containing its own RNA template (12,16), and both contain telomere binding proteins whose cellular concentration alters telomere length (14,17). Yeast molecular genetics has allowed identification of a number of genes that alter telomere length and identification of two yeast telomere length regulators, TEL1 and TEL2 (18–20). TEL1 shows strong sequence similarity to human ATM (ataxia telangiectasia mutated) and DNA-PKcs (the catalytic subunit of the DNA-dependent serine/threonine kinase) (19–22) and inactivating mutations of TEL1 yield viable cells with short telomeres. In contrast, TEL2 has no strong sequence homology to any known genes and is required for cell viability (20). Yeast telomeres consist of 325 ± 75 bp of irregular TG1–3 repeats in many laboratory strains (reviewed in 23). Telomere length is probably maintained by balancing lengthening (mediated by telomerase) and shortening (due to incomplete replication, rapid deletion and nuclease action) processes (23–25). Wild-type yeast cells respond to different perturbations that cause telomeres to be maintained at different equilibrium lengths. Cells double the length of their telomeres if a large excess of yeast telomeric sequences are introduced on a high copy plasmid (26). A similar increase occurs when the C-terminus of the yeast telomere binding protein Rap1p is overproduced in cells (17). Unlike wild-type cells, cells bearing viable tel1-1 and tel2-1 mutations maintain short telomeres (∼50 and 100 bp respectively) (18) and these telomeres do not increase in length when cells are subjected to the above perturbations (20). These results have classified the TEL1 and TEL2 gene products as part of the telomere length regulatory system because these mutations disrupt the feedback loops that cause telomere lengthening. In addition, the double
*To whom correspondence should be addressed. Tel: +1 216 445 9771; Fax: +1 216 444 0512; Email: [email protected]
1529 Nucleic Acids Acids Research, Research,1994, 1998,Vol. Vol.22, 26,No. No.16 Nucleic tel1-1 tel2-1 mutant maintains telomeres at the same length as tel1-1 cells (18), indicating that both TEL1 and TEL2 act in the same genetic pathway. Mutations in a number of other genes encoding proteins that bind telomeric DNA in vitro also alter telomere length. RAP1 encodes the major yeast telomere binding protein and different mutations can cause an increase (27,28) or decrease (29) in telomere length. Mutations in the genes encoding two proteins that associate with the Rap1p C-terminus, rif1 (30) and rif2 (31), cause telomeres to be maintained at longer equilibrium lengths. Mutations in either of two single-stranded yeast telomere binding proteins encoded by EST1 (32) and CDC13/EST4 (33) can cause telomeres to gradually shorten until cells die. Thus, yeast telomere binding proteins play a crucial role in regulating yeast telomere length. How the actions of these proteins are coordinated to maintain telomeres within a defined range has yet to be determined. The DNA binding and cellular functions of Rap1p have been characterized both genetically and biochemically. In vitro binding studies have shown that Rap1p binds a 13 bp telomeric site 5′-GGTGTGTGGGTGT-3′ (35,36). Besides telomeres, Rap1p also binds to various upstream activating sequences and silencer elements with a minimal consensus binding site 5′-RRT/ GGNNTGGGTGY-3′ (35,37). The Rap1p site from the yeast telomere is one of the highest affinity Rap1p sites known (35). As described above, TEL2 plays an unknown role in yeast telomere length regulation and cell viability. TEL2 has no significant sequence similarity to any known gene in the database and the deduced amino acid sequence of Tel2p has no known functional motifs. To understand the role of Tel2p in telomere length regulation, we constructed an MBP–Tel2p fusion and tested its ability to bind telomeric dsDNA. The MBP–Tel2p fusion binds to duplex yeast telomeric DNA in a sequence-specific manner, requiring only 8 bp of the 13 bp Rap1p site. Methylation interference and competition experiments indicate that this binding is specific. These data suggest that Tel2p function involves interactions with DNA. MATERIALS and METHODS Construction of RAP1 and TEL2 fusions A 446 amino acid portion of the 688 amino acid TEL2 ORF (amino acids 181–626) was fused to the 3′-end of the maltose binding protein (MBP) gene. The MBP gene is present in plasmid pMAL-c2 under control of the pTac promoter (New England Biolabs). Expression of full-length Tel2p is toxic to bacteria and repeated attempts to generate a full-length fusion resulted in no bacterial transformants or transformants containing rearranged plasmids. The TEL2 (Genbank accession no. U38538) NheI fragment from nt 1159 to 2499 was cloned into pMal-c2 at the XbaI site. The NheI fragment was brought in-frame with the MBP ORF by sequential filling in of the upstream BamHI and EcoRI sites with T4 DNA polymerase. This fusion could be propagated and produced in bacteria. This fusion contains 65% of the entire TEL2 ORF, where only nt 1441–2683 are required for cell viability (20). This plasmid construct is designated pMT-1. A MBP–Rap1p fusion was also constructed (857–3245, as numbered in 38; Genbank accession no. M18068) as a positive control, since Rap1p is known to bind yeast duplex telomeric DNA with high affinity (35). Plasmid D123 (RAP1 in YCp50; 38), which
1529
contained the RAP1 gene, was first digested with EcoRI and the ends rendered blunt with the Klenow fragment of DNA polymerase I and then digested with PstI. Plasmid pMAL-c2 was cut with HindIII, the ends rendered blunt with the Klenow fragment of DNA polymerase I and then digested with PstI. The cleaved plasmid was gel purified. The PstI–blunt RAP1 fragment was also gel purified and ligated to the PstI–blunt pMAL-c2 vector. The RAP1 ORF was then brought in-frame with the MBP ORF by filling in the upstream SalI site with T4 DNA polymerase. This plasmid construct is designated pMR-1. All constructs were verified by DNA sequencing of the MBP–TEL2 and MBP–RAP1 fusion junctions. Expression and purification of Tel2p and Rap1p in bacteria Both pMT-1 and pMr-1 were transformed into BL21(DE3) bacterial cells. The bacterial cells were grown at 37°C in 500 ml LB broth with 50 mg/l carbenicillin. The cultures were induced at an OD600 of O.3–0.4 by addition of 0.1 M IPTG to a final concentration of 0.3 mM and grown for 6–12 h at 22°C. Cells were harvested and cell pellets dissolved in 10 ml column buffer consisting of 100 mM Tris–Cl, pH 8, 200 mM NaCl, 0.005 mM DTT and stored at –80°C until further use. The bacterial cell pellet was thawed and sonicated for 6 0001 10 s with 10 s intervals on ice for 2 min. The sonicated slurry was spun at 14 000 g for 10 min and the supernatant chromatographed on a 15 ml amylose resin column. The fusion proteins were purified at 4°C according to the manufacturer’s specifications (New England Biolabs). Protein quantitation was performed using the BioRad Bradford assay using BSA to construct the standard curve. The purified proteins were analyzed by SDS–PAGE and found to be 30–50% pure, with the major contaminant being a band that co-migrated with pure MBP (New England Biolabs). Western blotting used polyclonal anti-MBP Ab (New England Biolabs) and the BioRad Chemiluminescent kit. Preparation of DNA substrates Desalted purified oligomers of different lengths of telomeric and non-telomeric single-stranded (ss) DNA sequences were purchased from Ransom Hill Biosciences Inc. (Ramona, CA). The G-rich strand oligonucleotide was end-labeled using T4 DNA polynucleotide kinase (39). Duplexes were formed by mixing the end-labeled oligonucleotides with a 5-fold molar excess of the complementary strand for all oligonucleotides except R3-ST, where a 6-fold molar excess was used (40). Thus, in addition to the competitor DNA present in the binding reaction, each reaction also contains a 5 (or 6)-fold molar excess of competitor ssDNA from the C-rich strand. The DNAs were denatured at 950001C for 5 min and re-annealed by allowing cooling slowly to room temperature over a period of 6–12 h. The duplex DNAs were ethanol precipitated and stored at –800001C until further use. Non-radioactive competitor DNAs were also prepared with a 5-fold molar excess of the C-rich strand in the same manner. Gel retardation assays The 32P-labeled probe, 8 ng dsDNA (0.023 µM final concentration for the 26 bp duplex probes) and 32 ng unlabeled C-rich oligonucleotide were incubated with either bacterial cell lysate or partially purified protein in 20 µl binding buffer consisting of 25 mM HEPES, pH 7.5, 150 mM NaCl, 50 mM KCl, 1 mM DTT
1530 Nucleic Acids Research, 1998, Vol. 26, No. 6 Table 1. Oligonucleotides used in the gel shift assay
Table 2. Mutant oligonucleotides to define the Tel2p binding site in R1-ST
Rap1p consensus sites are either underlined or in bold. Gel shifts were performed with partially purified MBP–Tel2p as described in Materials and Methods. The relative amount of shifted probe is indicated by the number of + signs, where each + indicates a visual increase in the amount of probe shifted relative to other probes; – indicates that no probe was shifted.
and 10% glycerol. To this reaction buffer were added 0.5 µg sonicated salmon sperm ssDNA (described in 41) and 0.1 µg GAL1-707 oligonucleotide (18 nt, 0.842 µM final concentration; sequence in 42) as non-specific competitors. End-labeled duplex oligonucleotide (50 000–80 000 c.p.m.) and 100 ng of either purified protein or 0.5 µg cell lysates were then added. After incubation for 20 min at room temperature, the mixture was run on a 5% native acrylamide gel (Protogel; National Diagnostics) in 1× TBE buffer at 10 V/cm for 2 h at room temperature. For the supershift reactions (Fig. 2C) 1.6 µg anti-MBP Ab (New England Biolabs) was added to the completed binding reaction and incubated on ice for 30 min prior to electrophoresis. Following electrophoresis, gels were dried under vacuum at 800001C. Protein–DNA complexes were visualized either by autoradiography (Table 1) or in a PhosphorImager (Molecular Dynamics) (Table 2). Quantitation of the amount of probe shifted was performed using the PhosphorImager by quantitating the amount of free and shifted probe region in reactions containing and lacking MBP–Tel2p or MBP–Rap1p, subtracting the c.p.m. for the region of the shifted probe in reactions lacking protein from the c.p.m. of shifted probe in reactions containing protein and dividing it by the total number of c.p.m. in the free probe in the reaction containing protein. The R1-ST value in each gel was then used as a standard for the mutant oligonucleotides (Table 2) on the same gel. PhosphorImager gel files were imported into Adobe Photoshop 3.0 and the levels of black and white (maximum and minimum levels) of the gray scale file were adjusted so the shifted band was visible. These manipulations changed the sensitivity so that low signals could be detected. The gray scale levels between the maximum and minimum were not altered (i.e. they remained linear between the maximum and minimum values).
Site truncation mutations were made by swapping the bold portions of the M13 sequence with the R1-ST sequence. The number of the bases in each strand of the Rap1p site is shown (after 46). Owing to some sequence similarities, some of the substitutions (bold bases) do not change all of the Rap1p consensus nucleotides (underlined bases). The lower portion of the table are mutations which eliminate G residues identified by methylation interference or that are in the minimal Tel2p recognition site. The mut3 oligonucleotide appears in both parts of the table. Gel shift analysis was performed using partially purified MBP–Tel2p and analyzed for amount of binding relative to R1-ST.
Methylation interference Dimethyl sulfate (DMS) methylation of 32P-labeled dsDNA (100 ng, 106 c.p.m.) was carried out in a 200 µl reaction containing 50 mM sodium cacodylate, pH 8.0, 1 mM EDTA, 1 µg salmon sperm DNA and 1 µl 1 M DMS. The reactions were incubated for 5 min at room temperature and quenched by addition of 50 µl stop buffer consisting of 1 M 2-mercaptoethanol, 1.5 M NaOAc and 1 µg glycogen as a carrier (43). The DNAs were ethanol precipitated twice and dissolved in sterilized milliQ-purified water and stored at –800001C. Five MBP–Tel2p gel retardation assay reactions (described above) were loaded and, following electrophoresis, the shifted bands were cut out and cast in a 0.7% agarose gel with DE81 paper (Whatman) next to the acrylamide gel slice. The labeled DNAs were run into the DE81 paper and the paper removed from the gel. The DNAs were eluted in 1.0 M NaCl, 10 mM Tris–HCl, pH 8.0, 1 mM EDTA, ethanol precipitated as above and subjected to 0.1 M piperidine hydrolysis. Other methods of elution included significant amounts of
1531 Nucleic Acids Acids Research, Research,1994, 1998,Vol. Vol.22, 26,No. No.16 Nucleic
1531
acrylamide with the DNA that interfered with electrophoresis of the hydrolyzed DNA. The hydrolyzed DNAs were lyophilized under vacuum to remove trace amounts of piperidine. The purified DNAs were then dissolved in 80% formamide and resolved on a 7 M urea–18% acrylamide gel (39). Other methods Total RNA was isolated from exponentially growing KR95-2A cells (20) as described in McKinney et al. (44) and quantitated by OD260. RESULTS Expression of fusion proteins Recombinant MBP–Rap1p and MBP–Tel2p were expressed in BL21(DE3) cells at various temperatures. Fusion proteins expressed at 370001C fractionated with the insoluble pellet. However, active soluble proteins were obtained by inducing expression at 220001C for 6–12 h. The predicted length of the Tel2p ORF is 688 amino acids. Fusion protein constructs containing full-length Tel2p were toxic to bacterial growth, even when cells were grown under repressing conditions. A 446 amino acid portion of the central region of Tel2p (amino acids 181–626) could be expressed in BL21 cells. The purified MBP–Rap1p and MBP–Tel2p proteins were 30–50% pure as judged by SDS–PAGE, with the major contaminating band being free MBP (as determined by Western blotting using anti-MBP as primary antibody; data not shown). Both the MBP–Rap1p and MBP–Tel2p purified fusion proteins migrated at the expected molecular weights of ∼140 and ∼100 kDa respectively (data not shown). The yields of final purified product were 0.3 mg for MBP–Tel2p and 0.5 mg for MBP–Rap1p per 500 ml culture. Tel2p binds yeast telomeric duplex DNA in a sequencespecific manner The binding specificity of Tel2p was initially determined in vitro using bacterial cell lysates and various duplex sequences (binding to single-stranded substrates will be considered elsewhere). These gel retardation assays all contained 0.5 µg sonicated salmon sperm DNA and 0.1 µg single-stranded oligonucleotide (∼40-fold molar excess over radiolabeled probes) as competitor DNAs. Using telomeric dsDNA probe R3-ST (Fig. 1 and Table 1) we observed a DNA–protein complex in the MBP–Tel2p lysate as well as the MBP–Rap1p lysate, but not in the MBP lysate. The MBP–Tel2p complex in lysates was specifically competed by a 20-fold molar excess of unlabeled probe (Fig. 1). Thus MBP– Tel2p can bind to the same telomeric DNA repeats as the known yeast telomere binding protein Rap1p. The specificity of Tel2p binding was further tested with various DNA probes using partially purified MBP–Tel2p. The major yeast telomere binding protein is Rap1p, which has a strongly conserved consensus binding site (45). Because MBP–Tel2p can bind to TG1–3 repeats containing three Rap1p sites, Tel2p binding to sequences containing one (R1-ST), two (R2-ST), three (R3-ST) or four (YTCA-1X) Rap1p sites was examined (Table 1). MBP-Tel2p bound all of the yeast telomeric dsDNA sequences to give a single shifted band (Fig. 2 and Table 1). No binding was observed with purified MBP alone (data not shown). Competition analysis showed that binding to the R2-ST probe by
Figure 1. Gel shift analysis of cell lysates from E.coli expressing MBP–Tel2p, MBP–Rap1p, and MBP used labeled R3-ST as probe. Increasing molar amounts (2×, 5×, 10× and 20×) of unlabeled R3-ST (Table 1) were used in competition with the MBP–Tel2p/R3-ST complex. Lane 1, R3-ST probe with no lysate; lane 2, R3-ST probe with MBP–Tel2p lysate; lanes 3–6, as lane 2 with increasing amounts of unlabeled R3-ST DNA added to the binding reaction; lane 7, R3-ST probe with MBP lysate; lane 8, R3-ST probe with MBP-Rap1p lysate. All lanes are from the same gel.
MBP–Tel2p was specifically competed by a 5-fold molar excess of R2-ST (Fig. 2B, lanes 3–5), but not by a 50-fold molar excess of duplex M13 oligonucleotide. Similar results were obtained with a 100-fold mass excess of pUC19 DNA sonicated to an average size of 0.7 kb (data not shown). The amount of R1-ST probe (Table 1) shifted in the presence of a 100- or 1000-fold mass excess of total S.cerevisiae RNA was also monitored (not shown). When normalized to the amount of R1-ST shifted in the absence of RNA, 0.5× R1-ST probe was shifted in the 100-fold RNA excess reaction and 0.3× was shifted in the 1000-fold RNA excess reaction. We note that the reactions containing a 50-fold molar excess of the non-specific duplex competitor M13 (Fig. 2B, lanes 6–8) also contained a 200-fold molar excess of one M13 single strand (Materials and Methods). This 200-fold molar excess of a single-stranded non-telomeric oligonucleotide did not inhibit MBP–Tel2p binding to R2-ST. As there was little or no competition by RNA and non-telomeric DNA, Tel2p is not a general DNA or RNA binding protein. Addition of anti-MBP Ab to the completed MBP–Tel2p R1-ST binding reaction caused the appearance of a supershifted band (Fig. 2C), consistent with purified MBP–Tel2p causing the R1-ST band shift. R1-ST, R2-ST, R3-ST and YTCA-1X were bound by Rap1p, as expected, and multiple Rap1p–DNA retarded complexes were observed for DNAs with multiple Rap1p sites (data not shown), in agreement with previous studies (46). MBP–Tel2p and MBP–Rap1p did not bind non-telomeric duplex sequences (M13, GAL4 and Y′ Telo) and no binding was observed with heterologous human, Tetrahymena or Oxytricha telomeric DNA sequence in our in vitro analyses (Table 1). These data show that Tel2p is a double-stranded TG1–3 binding protein in vitro.
1532 Nucleic Acids Research, 1998, Vol. 26, No. 6
Figure 3. MBP–Tel2p requires 8 bp of the 13 bp telomeric Rap1p site in R1-ST to bind DNA. The probe used in each binding reaction is indicated above each pair of lanes and the presence of MBP–Tel2p in the reaction is indicated by +. Mut10 shows that the specific G·C bp that shows the strongest methylation interference is required for binding. The MBP–Tel2p band is indicated by a line. (A) A scanned autoradiograph of representative binding reactions quantitated in Table 2. (B) Comparison of MBP–Tel2p and MBP–Rap1p binding to R1-ST and mut13 probes. The amount of probe shifted relative to MBP–Tel2p/R1-ST shift (as in Table 2) for MBP–Rap1p is 32× for mut3, 14× for mut 6, 0.0× for mut13 and 18× for R1-ST. Two shifted bands caused by a single Rap1p binding event have been observed previously (46,48).
Figure 2. Partially purified MBP-Tel2p binds duplex DNAs that contain telomeric sequences. (A) The probe is listed above the pair of lanes. The lane containing the binding reaction with MBP–Tel2p is indicated and the MBP–Tel2p shifted band is indicated by an arrow. All lanes are from the same gel. The sequences of the probes are given in Table 1. Lane 1, R1-ST probe alone; lane 2, R1-ST + MBP–Tel2p; lane 3, R2-ST; lane 4, R2-ST + MBP–Tel2p; lane 5, R3-ST; lane 6, R3-ST + MBP–Tel2p; lane 7, M13; lane 8, M13 + MBP-Tel2p. (B) Partially purified MBP–Tel2p specifically binds telomeric DNA. Partially purified MBP– Tel2p was used in the same experiments as in Figure 1A, performed using R2-ST as probe. The fold molar excesses of specific competitor, R2-ST, and non-specific competitor, M13 (Table 1), were 5×, 25× and 50×. Lane 1, R2-ST probe with no protein; lane 2, R2-ST probe with MBP–Tel2p protein; lanes 3–5, as lane 2 with 5× (lane 3), 25× (lane 4) and 50× (lane 5) molar excess of unlabeled R2-ST added to the binding reaction; lanes 6–8, as lane 2 with 5× (lane 6), 25× (lane 7) and 50× (lane 8) molar excess of unlabeled M13 duplex added to the binding reaction. (C) Partially purified MBP–Tel2p was used to shift the R1-ST probe in the absence or presence of 1.6 µg polyclonal anti-MBP Ab. Arrows indicate the shifted band and supershifted band.
The Tel2p–DNA complex migrated similarly to the fastest migrating Rap1p–DNA complex (see for example Figs 1 and 3B) and the intensity of the MBP–Tel2p shifted complex increased with increasing number of Rap1p binding sites and GT base pairs present in the DNA substrate (Fig. 2A, compare lane 2 with lanes 4 and 6). However, only a single MBP–Tel2p shifted band that increased in intensity with increasing protein levels irrespective of the number of Rap1p sites present on the DNA substrate (data not shown) was observed. While Tel2p bound to the telomeric duplex sequences containing Rap1p sites, Tel2p did not bind the non-telomeric TEF DNA probe (Table 1). The TEF probe contains a non-telomeric Rap1p site from the TEF2 UAS, which is the second highest affinity Rap1p site known (35). In addition, Tel2p did not bind the M13, GAL4 or Y′ Telo junction or the heterologous telomeric DNA sequences from humans, Oxytricha and Tetrahymena (Table 1). Thus Tel2p has the in vitro properties of a novel yeast telomere-specific binding protein. The results with the TEF probe indicate that Tel2p DNA binding is distinct from Rap1p binding.
1533 Nucleic Acids Acids Research, Research,1994, 1998,Vol. Vol.22, 26,No. No.16 Nucleic
1533
A minimum duplex telomeric sequence of 8 bp is required for Tel2p binding To delineate the minimum Tel2p binding site within the telomeric Rap1p site, a series of mutations were introduced into the single Rap1p site in the R1-ST probe and tested for Tel2p binding. The mutations were constructed by taking two 26 bp duplexes, R1-ST and M13 (Table 1), and replacing portions of the R1-ST sequence with M13 sequences. The M13 duplex was not bound by Tel2p, so as residues critical for binding are altered, the gel shift should be greatly reduced. To comply with the previous nomenclature for the Rap1p site (46) and to relate these data to the methylation interference analysis (below), the bases in the Rap1p site in the CA-rich strand are numbered 1–13 in the 5′!3′ direction and in the TG-rich strand are numbered 13′–1′ in the 5′!3′ direction (Table 2). The fraction of probe shifted by a constant amount of MBP–Tel2p was measured in a PhosphorImager and then normalized to R1-ST for a quantitative comparison of the different probes (Table 2). Gel shift analysis of duplex oligomers with bases mutated at various positions revealed that bases 1′, 3′ and 4′ of the Rap1p binding site were not necessary for MBP–Tel2p binding, because binding was observed in mut5 (Fig. 3 and Table 2). Methylation interference data below and TEF probe binding (Table 1) suggested that the G at 2′ is also not required for MBP–Tel2p binding (confirmed below). Individual base pairs in positions 4–6 (CCC), which are essential for Rap1p–DNA interaction (37,47), were dispensible for Tel2p binding because the mut8 and mut9 probes were shifted (Fig. 3 and Table 2). Similar mutations abolished binding by Rap1p (47). Mutant duplexes with single base substitutions (mut8–mut12 and mut3 in Table 2) revealed that Tel2p binding was undetectable only when a G!C base substitution at position 12 was made (mut10). Other single base substitutions at G13 (mut3), G10 (mut11), G8 (mut12), G6 (mut8) or G5 (mut9) did not eliminate Tel2p–DNA interaction. However, Tel2p binding was abolished with multiple base substitutions in the combinations G13′ and G10′ (mut2), G12′ and G10′–G8′ (mut7) or G13’ and G10′–G8′ (mut1). The shortest telomeric sequence that could be bound by Tel2p in this sequence context was mut5 (Fig. 3A and Table 2), which contains a 9 bp contiguous tract identical to the Rap1p site and G2′. Because G2′ was not identified as an important base by methylation interference (below) and was present in the TEF duplex that was not bound by Tel2p (Table 1) and G13′ was not required for binding, this analysis suggested that the 8 bp GT tract from bp 5′ to 12′ of the Rap1p site was required for Tel2p binding. To directly test the importance of G2′ in Tel2p binding, the mut13 probe bearing the 8 bp GT tract was made and tested. Tel2p shifted the mut13 probe as well as mut8, mut9 and several other probes (Fig. 3B and Table 2). Thus, G2′ was not required for Tel2p binding. The relative binding of equal amounts of partially purified MBP–Tel2p and MBP–Rap1p was compared using R1-ST, mut3, mut6 and mut13 substrates. Rap1p bound R1-ST, mut6 and mut3 efficiently as expected (47), but no binding of the mut13 probe was detected (Fig. 3B). An equal amount of MBP–Rap1p shifted 10- to 18-fold more R1-ST probe than MBP–Tel2p in independent experiments. These data further distinguish Rap1p and Tel2p binding of yeast telomeric DNA.
Figure 4. DMS interference with MBP–Tel2p and MBP–Rap1p bound to the duplex R1-ST probe. (A) Methylation interference of MBP–Rap1p binding to R1-ST. Free (total) DNA and Rap1p bound DNA are indicated. G5′, G6′, G12′ and G13′ show strong interference. (B) Methylation interference with MBP–Tel2p binding to R1-ST. G12′ shows strong interference. The equivalent of a longer exposure of the Tel2p shifted lane is shown to the left of the gel. In both panels arrows indicate bases that show significant interference.
Tel2p contacts duplex telomeric DNA in the major groove To further examine the specificity of MBP–Tel2p binding on the R1-ST probe, the DNA–protein contacts were studied by DMS interference analysis. DMS methylates the N7 position of guanines in the major groove of the DNA (39). Partially methylated DNA was used to form DNA–protein complexes with both MBP–Tel2p and MBP–Rap1p. The DNA from both complexes and the free probe were isolated and purified (Materials and Methods). As the free probe was in large excess over the amount of bound material, the free probe is equivalent to the input material. After piperidine cleavage the products were resolved on an 18% urea–acrylamide gel (Fig. 4). In agreement with previous reports (36,47,48), strong interference was observed at bases G5′, G6′, G12′ and G13′ in the Rap1p bound complex. The enhancement of bands near the 3′-end of the probe in the Rap1p bound material over the free probe has been observed previously by others (48) and did not alter the identification of bases critical for Rap1p binding. MBP–Tel2p showed a different pattern of DMS interference. Strong interference was observed at G12′ (Fig. 4). A comparison of protein–DNA contacts between the two proteins is shown in Figure 5. The strong interference at G12′ is consistent with the binding site mutations, as changing the G·C bp at position 12 to C·G (mut10, Table 2) eliminated Tel2p binding (Fig. 3). Thus the methylation interference data combined with the mutant oligo-
1534 Nucleic Acids Research, 1998, Vol. 26, No. 6
Figure 5. Bases identified by methylation interference of MBP–Rap1p and MBP–Tel2p binding. The ovals indicate bases with strong interference. Underlined bases make up the Rap1p binding site. The base numbering on each strand is shown after the convention of Gilson et al. (46).
nucleotide probes (Table 2) indicate that MBP–Tel2p specifically binds to bp 5–12 of the telomeric Rap1p site. DISCUSSION The data presented here show that Tel2p can specifically bind to telomeric dsDNA sequences in vitro. TEL2 and TEL1 are two yeast genes that behave as genetic regulators of telomere length (20). By sequence homology, Tel1p is a large serine/threonine kinase similar to DNA-PKcs (19,49). In contrast, Tel2p is a novel protein that has no significant homology to any protein in the database (20) or ‘telobox’ proteins that bind to telomeric DNA through Myb motifs (50). These data provide the first step for constructing testable models of Tel2p function in telomere length regulation and cell viability. Presently, it is unknown if Tel2p binds DNA in vivo and, if so, alone or in conjunction with another protein that alters or increases its affinity for specific sites. Purification of Tel2p from yeast cells would help answer these questions. However, since the TEL2 sequence and GAL1–TEL2 and TEL2–lacZ fusions suggest that Tel2p is a very low abundance protein (20), this biochemical approach will be challenging. The information that Tel2p can bind yeast telomeric dsDNA in vitro combined with yeast molecular genetics should allow identification of Tel2p interacting proteins. Production and analysis of these proteins in combination with Tel2p in DNA binding studies in vitro will define more clearly the potential Tel2p–DNA interactions in vivo. The MBP–Tel2p fusion protein could gel shift duplex telomeric oligonucleotides in the presence of a large excess of salmon sperm DNA and a single-stranded oligonucleotide present in each binding reaction (Materials and Methods) and a non-specific competitor (Figs 1 and 2 and Table 1). Competition for Tel2p binding of telomeric DNA was specific (Figs 1 and 2). In addition, not all duplex probes tested were bound by MBP–Tel2p (Table 1). These data and the methylation interference results indicate that MBP–Tel2p shows sequence specificity in its dsDNA binding. At least 8 bp of the 13 bp Rap1p telomeric site were required for Tel2p binding and the G·C bp at position 12 (Fig. 5) was essential (Fig. 3 and Table 2). In contrast, Rap1p did not bind to this 8 bp sequence (mut13, Fig. 3B). Because Tel2p could bind duplexes with point mutations at positions 6, 8 and 10 (mut8, mut12 and mut11, Table 2), Tel2p may make other
contacts outside the major groove that compensate for these single base changes. Tel2p has not been previously identified as part of a telomeric DNA binding activity in yeast extracts, most likely for two reasons. First, Rap1p is an abundant yeast telomere binding activity in yeast extracts (35), making detection of additional yeast telomere binding proteins difficult. Partial fractionation of yeast nuclear extracts has revealed the presence of a second non-Rap1p telomere binding activity called TBFβ (34), but this activity has not been further characterized. It is not yet known if Tel2p is TBFβ or part of TBFβ. Second, expression of Tel2p is toxic to bacteria. Indeed, even placing the TEL2 ORF behind strong yeast promoters, which are active in bacteria makes plasmids unstable in Escherichia coli such that transformants grow poorly and rearranged plasmids are frequently isolated (K.W.Runge, unpublished data). Only by constructing a truncated bacterial fusion that contains 65% of the TEL2 ORF were we able to obtain unrearranged plasmids in bacteria and produce protein. Thus, previous searches for yeast telomeric binding proteins with oligonucleotide probes (51) might not have found TEL2 because high level expression would inhibit bacterial colony formation. The known homology of Tel1p with DNA-PKcs, the fact that TEL1 and TEL2 act in the same genetic pathway and these data showing that Tel2p can bind to telomeric DNA suggest a model for Tel1p and Tel2p function in telomere length regulation. DNA-PK is a DNA-dependent protein kinase in which the catalytic subunit, DNA-PKcs, is activated by binding to two DNA-bound subunits, Ku-70 and Ku-80, and can phosphorylate a number of proteins critical to cell cycle progression and DNA repair, including p53 (52). By analogy, a speculative model for Tel2p function is that Tel2p binds to telomeric DNA and, in conjunction with another protein bound to the chromosome terminus, activates Tel1p to phosphorylate telomere-associated proteins such as Rap1p, Rif1p (30), Rif2p (31), Sir2p, 3p and 4p (53), Est1p (32), Cdc13p/Est4p (33) or Stn1p (54) in the course of telomere length regulation. It is interesting to note that the point mutation in tel2-1 that causes cells to maintain short telomeres occurs in a part of the protein that is not present in the MBP–Tel2p fusion. In our model the tel2-1 protein would bind DNA but fail to bind to or activate the Tel1p kinase. Deletion of TEL2 is lethal to cells, while deletion of TEL1 results in viable cells with short telomeres (19,20). Thus, Tel2p must have an essential function besides its activities in the TEL1 pathway. The physical similarity between telomeres and doublestrand breaks, the evolutionary conservation of the telomere-related protein Tel1p and the double-strand break-activated kinase DNA-PKcs and the data presented here showing that Tel2p can bind telomere DNA suggest that the essential function of Tel2p involves metabolism of DNA ends. Another Tel1p/DNA-PKcs homolog, Mec1p, controls cell cycle progression in response to DNA damage (55) and is essential for viability in some yeast strains (21). The present data on tel2 mutations and DNA binding are consistent with an essential role for Tel2p in this process. If so, TEL2 should show genetic interactions with components of the DNA damage signalling pathway. Thus, these data on Tel2p DNA binding suggest specific testable models for the function of this novel protein with no known homologs.
1535 Nucleic Acids Acids Research, Research,1994, 1998,Vol. Vol.22, 26,No. No.16 Nucleic ACKNOWLEDGEMENTS This work was supported by NIH grant GM50752 and a American Cancer Society Junior Faculty Research Award to K.W.R. We thank Drs Donal Luse for helpful advice during this project and Richard Gronostajski for critical reading of the manuscript. REFERENCES 1 Harley,C.B. and Villeponteau,B. (1995) Curr. Opin. Genet. Dev., 5, 249–255. 2 Wilkie,A.O.M., Lamb,J., Harris,P.C., Finney,R.D. and Higgs,D.R. (1990) Nature, 346, 868–871. 3 Wellinger,R.J. and Zakian,V. A. (1989) Proc. Natl. Acad. Sci. USA, 86, 973–977. 4 Zakian,V.A., Blanton,J.M. ,Wetzel,L. and Dani,G.M. (1986) Mol. Cell. Biol., 6, 925–932. 5 Murray,A.W., Schultes,N.P. and Szostak,J.W. (1986) Cell, 45, 529–536. 6 Sandell,L.L., Gottschling,D.E. and Zakian,V.A. (1994) Proc. Natl. Acad. Sci. USA, 91, 12061–12065. 7 Sandell,L.L. and Zakian,V.A. (1993) Cell, 75, 729–739. 8 Zakian,V.A. (1989) Annu. Rev. Genet., 23, 579–604. 9 Harley,C.B. (1995) In Blackburn,E.H. and Greider,C.W. (ed.), Telomeres. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pp. 247–263. 10 de Lange,T. (1995) In Blackburn,E.H. and Greider,C.W. (ed.), Telomeres. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pp. 265–293 11 Blackburn,E.H. (1992) Annu. Rev. Biochem., 61, 113–129. 12 Singer,M.S. and Gottschling,D.E. (1994) Science, 266, 404–409. 13 Wellinger,R.J., Ethier,K., Labrecque,P. and Zakian,V.A. (1996) Cell, 85, 423–433. 14 van Steensel,B. and de Lange,T. (1997) Nature, 385, 740–743. 15 Marcand,S., Gilson,E. and Shore,D. (1997) Science, 275, 986–990. 16 Lingner,J., Hughes,T.R., Shevchenko,A., Mann,M., Lundblad,V. and Cech,T.R. (1997) Science, 276, 561–567. 17 Conrad,M.N., Wright,J.H., Wolf,A.J. and Zakian,V.A. (1990) Cell, 63, 739–750. 18 Lustig,A.J. and Petes,T.D. (1986) Proc. Natl. Acad. Sci. USA, 83, 1398–1402. 19 Greenwell,P.A., Kronmal,S.L., Porter,S.E., Gassenhuber,J., Obermaier,B. and Petes,T.D. (1995) Cell, 82, 823–829. 20 Runge,K.W. and Zakian,V.A. (1996) Mol. Cell. Biol., 16, 3094–3105. 21 Morrow,D.M., Tagle,D.A., Shiloh,Y., Collins,F.S. and Heiter,P. (1995) Cell, 82, 831–840. 22 Savitsky,K., Bar-Shira,A., Gilad,S., Rotman,G., Ziv,Y., Vanagaite,L., Tagle,D.A., Smith,S., Uziel,T., Sfez,S., Ashkenazi,M., Pecker,I., Frydman,M., Harnik,R., Patanjali,S.R., Simmons,A., Clines,G.A., Sartiel,A., Gatti,R.A., Chessa,L., Sanal,O., Lavin,M.F., Jaspers,N.G.J., Taylor,A.M.R., Arlett,C.F., Miki,T., Weissman,S.M., Lovett,M., Collins,F.S. and Shiloh,Y. (1995) Science, 268, 1749–1753. 23 Zakian,V.A. (1996) Annu. Rev. Genet., 30, 141–172.
24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56
1535
Li,B. and Lustig,A. J. (1996) Genes Dev., 10, 1310–1326. Garvik,B., Carson,M. and Hartwell,L. (1995) Mol. Cell. Biol., 15, 6128–6138. Runge,K.W. and Zakian,V.A. (1989) Mol. Cell. Biol., 9, 1488–1497. Kyrion,G., Boakye,K.A. and Lustig,A.J. (1992) Mol. Cell. Biol., 12, 5159–5173. Liu,C., Mao,X. and Lustig,A.J. (1994) Genetics, 138, 1025–1040. Lustig,A.J., Kurtz,S. and Shore,D. (1990) Science, 250, 549–553. Hardy,C.F.J., Sussel,L. and Shore,D. (1992) Genes Dev., 6, 801–814. Wotton,D. and Shore,D. (1997) Genes Dev., 11, 748–760. Virta-Pearlman,V., Morris,D.K. and Lundblad,V. (1996) Genes Dev., 10, 3094–3104. Nugent,C.I., Hughes, T.R., Lue,N.F. and Lundblad,V. (1996) Science, 274, 249–252. Liu,Z.P. and Tye,B.K. (1991) Genes Dev., 5, 49–59. Buchman,A.R., Lue,N.F. and Kornberg,R.D. (1988) Mol. Cell. Biol., 8, 5086–5099. Longtine,M.S., Wilson, N.M., Petracek,M.E. and Berman.,J. (1989) Curr. Genet., 16, 225–240. Graham,I.R. and Chambers,A. (1994) Nucleic Acids Res., 22, 124–130. Shore,D. and Nasmyth,K. (1987) Cell, 51, 721–732. Sambrook,J., Fritsch,E.F. and Maniatis,T. (1989) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor LaboratoryPress, Cold Spring Harbor, NY. Sheng,H., Hou,Z., Schierer,T., Dobbs,D.L. and Henderson,E. (1995) Mol. Cell. Biol., 15, 1144–1153. Schiestl,R.H. and Gietz,R.D. (1989) Curr. Genet., 16, 339–346. Liu,H., Krizek,J. and Bretscher,A. (1992) Genetics, 132, 665–673. Baldwin,A.S.,Jr (1992) In Ausubel,F.M., Brent,R., Kingston,R.E., Seidman,J.G., Smith,J.A. and Struhl,K. (eds), Short Protocols in Molecular Biology. John Wiley & Sons, New York, NY, pp. 12-8–12-10 McKinney,J.D., Chang,F., Heintz,N. and Cross,F.R. (1993) Genes Dev., 7, 833–843. Shore,D. (1994) Trends Genet., 10, 408–412. Gilson,E., Roberge,M., Giraldo,R., Rhodes,D. and Gasser,S.M. (1993) J. Mol. Biol., 231, 293–310. Vignais,M.L., Huet,J., Buhler,J.M. and Sentenac,A. (1990) J. Biol. Chem., 265, 14669–14674. Chambers,A., Tsang, J.S., Stanway,C., Kingsman,A.J. and Kingsman,S.M. (1989) Mol. Cell. Biol., 9, 5516–5524. Hartley,K.O., Gell,D., Smith,G.C.M., Zhang,H., Divecha,N., Connelly,M.A., Admon,A., Lees-Miller,S.P., Anderson,C.W. and Jackson,S.P. (1995) Cell, 82, 849–856. Brun,C., Marcand,S. and Gilson,E. (1997) Trends Cell Biol., 7, 317–324. Lin,J.-J. and Zakian,V. A. (1994) Nucleic Acids Res., 22, 4906–4913. Anderson,C.W. (1993) Trends Biochem. Sci., 18, 433–437. Grunstein,M. (1997) Curr. Opin. Cell Biol., 9, 383–387. Grandin,N., Reed,S.I. and Charbonneau,M. (1997) Genes Dev., 11, 512–527. Weinert,T.A., Kiser,G.L. and Hartwell,L.H. (1994) Genes Dev., 8, 652–665. Runge,K.W. and Zakian,V.A. (1990) Nucleic Acids Res., 18, 1783–1787.
1998 Oxford University Press
The yeast telomere length regulator TEL2 encodes a protein that binds to telomeric DNA Rama S. Kota and Kurt W. Runge* The Lerner Research Institute, Cleveland Clinic Foundation, Department of Molecular Biology, NC 20, 9500 Euclid Avenue, Cleveland, OH 44195, USA Received September 12, 1997; Revised and Accepted January 26, 1998
ABSTRACT
TEL2 is required for telomere length regulation and viability in Saccharomyces cerevisiae. To investigate the mechanism by which Tel2p regulates telomere length, the majority (65%) of the TEL2 ORF was fused to the 3′-end of the gene for maltose binding protein, expressed in bacteria and the purified protein used in DNA binding studies. Rap1p, the major yeast telomere binding protein, recognizes a 13 bp duplex site 5′-GGTGTGTGGGTGT-3′ in yeast telomeric DNA with high affinity. Gel shift experiments revealed that the MBP–Tel2p fusion binds the double-stranded yeast telomeric Rap1p site in a sequence-specific manner. Analysis of mutated sites showed that MBP–Tel2p could bind 5′-GTGTGTGG-3′ within this 13 bp site. Methylation interference analysis revealed that Tel2p contacts the 5′-terminal guanine in the major groove. MBP–Tel2p did not bind duplex telomeric DNA repeats from vertebrates, Tetrahymena or Oxytricha. These results suggest that Tel2p is a DNA binding protein that recognizes yeast telomeric DNA. INTRODUCTION Telomere length regulation, regulation of the number of repeated DNA sequences that make up the chromosome ends in many organisms, has recently been recognized as a critical function in human cells. Telomere DNA repeats are lost from somatic cell telomeres as cells divide (1), leading to the proposal that telomeres both activate cellular senescence when telomeres become too short and eliminate many transformed cells by causing chromosome destruction by loss of telomere function. The telomeric DNA repeats are the only sequences required for telomere function in yeast and humans (2–6), so loss of these repeats is expected to cause chromosome ends to behave as double-strand (ds)DNA breaks (7,8). Experiments with transformed cells grown in culture show a correlation between the number of cell divisions, decreased telomere length, the onset of ‘crisis’ and end-to-end chromosome fusions consistent with loss of telomere function (9,10). Thus regulation of telomere length is critical in humans for determining the lifespan of individual cells. While some of the factors that lengthen and shorten telomeres have been
identified (11–14), how telomere length is maintained at an appropriate length in human germ cells, where lengthening mechanisms are active, is still unknown. Recent evidence suggests that telomere length is measured by monitoring the number of telomere binding proteins (14,15; A.Ray and K.W.Runge, submitted for publication), but how these protein molecules are counted has not been determined. One approach to this problem is to identify regulators of telomere length and determine how they function. The yeast Saccharomyces cerevisiae has been an ideal model organism for investigating the structure and function of telomeres. Both humans and yeast have telomeres composed of short repeated sequences that are elongated by telomerase, a reverse transcriptase containing its own RNA template (12,16), and both contain telomere binding proteins whose cellular concentration alters telomere length (14,17). Yeast molecular genetics has allowed identification of a number of genes that alter telomere length and identification of two yeast telomere length regulators, TEL1 and TEL2 (18–20). TEL1 shows strong sequence similarity to human ATM (ataxia telangiectasia mutated) and DNA-PKcs (the catalytic subunit of the DNA-dependent serine/threonine kinase) (19–22) and inactivating mutations of TEL1 yield viable cells with short telomeres. In contrast, TEL2 has no strong sequence homology to any known genes and is required for cell viability (20). Yeast telomeres consist of 325 ± 75 bp of irregular TG1–3 repeats in many laboratory strains (reviewed in 23). Telomere length is probably maintained by balancing lengthening (mediated by telomerase) and shortening (due to incomplete replication, rapid deletion and nuclease action) processes (23–25). Wild-type yeast cells respond to different perturbations that cause telomeres to be maintained at different equilibrium lengths. Cells double the length of their telomeres if a large excess of yeast telomeric sequences are introduced on a high copy plasmid (26). A similar increase occurs when the C-terminus of the yeast telomere binding protein Rap1p is overproduced in cells (17). Unlike wild-type cells, cells bearing viable tel1-1 and tel2-1 mutations maintain short telomeres (∼50 and 100 bp respectively) (18) and these telomeres do not increase in length when cells are subjected to the above perturbations (20). These results have classified the TEL1 and TEL2 gene products as part of the telomere length regulatory system because these mutations disrupt the feedback loops that cause telomere lengthening. In addition, the double
*To whom correspondence should be addressed. Tel: +1 216 445 9771; Fax: +1 216 444 0512; Email: [email protected]
1529 Nucleic Acids Acids Research, Research,1994, 1998,Vol. Vol.22, 26,No. No.16 Nucleic tel1-1 tel2-1 mutant maintains telomeres at the same length as tel1-1 cells (18), indicating that both TEL1 and TEL2 act in the same genetic pathway. Mutations in a number of other genes encoding proteins that bind telomeric DNA in vitro also alter telomere length. RAP1 encodes the major yeast telomere binding protein and different mutations can cause an increase (27,28) or decrease (29) in telomere length. Mutations in the genes encoding two proteins that associate with the Rap1p C-terminus, rif1 (30) and rif2 (31), cause telomeres to be maintained at longer equilibrium lengths. Mutations in either of two single-stranded yeast telomere binding proteins encoded by EST1 (32) and CDC13/EST4 (33) can cause telomeres to gradually shorten until cells die. Thus, yeast telomere binding proteins play a crucial role in regulating yeast telomere length. How the actions of these proteins are coordinated to maintain telomeres within a defined range has yet to be determined. The DNA binding and cellular functions of Rap1p have been characterized both genetically and biochemically. In vitro binding studies have shown that Rap1p binds a 13 bp telomeric site 5′-GGTGTGTGGGTGT-3′ (35,36). Besides telomeres, Rap1p also binds to various upstream activating sequences and silencer elements with a minimal consensus binding site 5′-RRT/ GGNNTGGGTGY-3′ (35,37). The Rap1p site from the yeast telomere is one of the highest affinity Rap1p sites known (35). As described above, TEL2 plays an unknown role in yeast telomere length regulation and cell viability. TEL2 has no significant sequence similarity to any known gene in the database and the deduced amino acid sequence of Tel2p has no known functional motifs. To understand the role of Tel2p in telomere length regulation, we constructed an MBP–Tel2p fusion and tested its ability to bind telomeric dsDNA. The MBP–Tel2p fusion binds to duplex yeast telomeric DNA in a sequence-specific manner, requiring only 8 bp of the 13 bp Rap1p site. Methylation interference and competition experiments indicate that this binding is specific. These data suggest that Tel2p function involves interactions with DNA. MATERIALS and METHODS Construction of RAP1 and TEL2 fusions A 446 amino acid portion of the 688 amino acid TEL2 ORF (amino acids 181–626) was fused to the 3′-end of the maltose binding protein (MBP) gene. The MBP gene is present in plasmid pMAL-c2 under control of the pTac promoter (New England Biolabs). Expression of full-length Tel2p is toxic to bacteria and repeated attempts to generate a full-length fusion resulted in no bacterial transformants or transformants containing rearranged plasmids. The TEL2 (Genbank accession no. U38538) NheI fragment from nt 1159 to 2499 was cloned into pMal-c2 at the XbaI site. The NheI fragment was brought in-frame with the MBP ORF by sequential filling in of the upstream BamHI and EcoRI sites with T4 DNA polymerase. This fusion could be propagated and produced in bacteria. This fusion contains 65% of the entire TEL2 ORF, where only nt 1441–2683 are required for cell viability (20). This plasmid construct is designated pMT-1. A MBP–Rap1p fusion was also constructed (857–3245, as numbered in 38; Genbank accession no. M18068) as a positive control, since Rap1p is known to bind yeast duplex telomeric DNA with high affinity (35). Plasmid D123 (RAP1 in YCp50; 38), which
1529
contained the RAP1 gene, was first digested with EcoRI and the ends rendered blunt with the Klenow fragment of DNA polymerase I and then digested with PstI. Plasmid pMAL-c2 was cut with HindIII, the ends rendered blunt with the Klenow fragment of DNA polymerase I and then digested with PstI. The cleaved plasmid was gel purified. The PstI–blunt RAP1 fragment was also gel purified and ligated to the PstI–blunt pMAL-c2 vector. The RAP1 ORF was then brought in-frame with the MBP ORF by filling in the upstream SalI site with T4 DNA polymerase. This plasmid construct is designated pMR-1. All constructs were verified by DNA sequencing of the MBP–TEL2 and MBP–RAP1 fusion junctions. Expression and purification of Tel2p and Rap1p in bacteria Both pMT-1 and pMr-1 were transformed into BL21(DE3) bacterial cells. The bacterial cells were grown at 37°C in 500 ml LB broth with 50 mg/l carbenicillin. The cultures were induced at an OD600 of O.3–0.4 by addition of 0.1 M IPTG to a final concentration of 0.3 mM and grown for 6–12 h at 22°C. Cells were harvested and cell pellets dissolved in 10 ml column buffer consisting of 100 mM Tris–Cl, pH 8, 200 mM NaCl, 0.005 mM DTT and stored at –80°C until further use. The bacterial cell pellet was thawed and sonicated for 6 0001 10 s with 10 s intervals on ice for 2 min. The sonicated slurry was spun at 14 000 g for 10 min and the supernatant chromatographed on a 15 ml amylose resin column. The fusion proteins were purified at 4°C according to the manufacturer’s specifications (New England Biolabs). Protein quantitation was performed using the BioRad Bradford assay using BSA to construct the standard curve. The purified proteins were analyzed by SDS–PAGE and found to be 30–50% pure, with the major contaminant being a band that co-migrated with pure MBP (New England Biolabs). Western blotting used polyclonal anti-MBP Ab (New England Biolabs) and the BioRad Chemiluminescent kit. Preparation of DNA substrates Desalted purified oligomers of different lengths of telomeric and non-telomeric single-stranded (ss) DNA sequences were purchased from Ransom Hill Biosciences Inc. (Ramona, CA). The G-rich strand oligonucleotide was end-labeled using T4 DNA polynucleotide kinase (39). Duplexes were formed by mixing the end-labeled oligonucleotides with a 5-fold molar excess of the complementary strand for all oligonucleotides except R3-ST, where a 6-fold molar excess was used (40). Thus, in addition to the competitor DNA present in the binding reaction, each reaction also contains a 5 (or 6)-fold molar excess of competitor ssDNA from the C-rich strand. The DNAs were denatured at 950001C for 5 min and re-annealed by allowing cooling slowly to room temperature over a period of 6–12 h. The duplex DNAs were ethanol precipitated and stored at –800001C until further use. Non-radioactive competitor DNAs were also prepared with a 5-fold molar excess of the C-rich strand in the same manner. Gel retardation assays The 32P-labeled probe, 8 ng dsDNA (0.023 µM final concentration for the 26 bp duplex probes) and 32 ng unlabeled C-rich oligonucleotide were incubated with either bacterial cell lysate or partially purified protein in 20 µl binding buffer consisting of 25 mM HEPES, pH 7.5, 150 mM NaCl, 50 mM KCl, 1 mM DTT
1530 Nucleic Acids Research, 1998, Vol. 26, No. 6 Table 1. Oligonucleotides used in the gel shift assay
Table 2. Mutant oligonucleotides to define the Tel2p binding site in R1-ST
Rap1p consensus sites are either underlined or in bold. Gel shifts were performed with partially purified MBP–Tel2p as described in Materials and Methods. The relative amount of shifted probe is indicated by the number of + signs, where each + indicates a visual increase in the amount of probe shifted relative to other probes; – indicates that no probe was shifted.
and 10% glycerol. To this reaction buffer were added 0.5 µg sonicated salmon sperm ssDNA (described in 41) and 0.1 µg GAL1-707 oligonucleotide (18 nt, 0.842 µM final concentration; sequence in 42) as non-specific competitors. End-labeled duplex oligonucleotide (50 000–80 000 c.p.m.) and 100 ng of either purified protein or 0.5 µg cell lysates were then added. After incubation for 20 min at room temperature, the mixture was run on a 5% native acrylamide gel (Protogel; National Diagnostics) in 1× TBE buffer at 10 V/cm for 2 h at room temperature. For the supershift reactions (Fig. 2C) 1.6 µg anti-MBP Ab (New England Biolabs) was added to the completed binding reaction and incubated on ice for 30 min prior to electrophoresis. Following electrophoresis, gels were dried under vacuum at 800001C. Protein–DNA complexes were visualized either by autoradiography (Table 1) or in a PhosphorImager (Molecular Dynamics) (Table 2). Quantitation of the amount of probe shifted was performed using the PhosphorImager by quantitating the amount of free and shifted probe region in reactions containing and lacking MBP–Tel2p or MBP–Rap1p, subtracting the c.p.m. for the region of the shifted probe in reactions lacking protein from the c.p.m. of shifted probe in reactions containing protein and dividing it by the total number of c.p.m. in the free probe in the reaction containing protein. The R1-ST value in each gel was then used as a standard for the mutant oligonucleotides (Table 2) on the same gel. PhosphorImager gel files were imported into Adobe Photoshop 3.0 and the levels of black and white (maximum and minimum levels) of the gray scale file were adjusted so the shifted band was visible. These manipulations changed the sensitivity so that low signals could be detected. The gray scale levels between the maximum and minimum were not altered (i.e. they remained linear between the maximum and minimum values).
Site truncation mutations were made by swapping the bold portions of the M13 sequence with the R1-ST sequence. The number of the bases in each strand of the Rap1p site is shown (after 46). Owing to some sequence similarities, some of the substitutions (bold bases) do not change all of the Rap1p consensus nucleotides (underlined bases). The lower portion of the table are mutations which eliminate G residues identified by methylation interference or that are in the minimal Tel2p recognition site. The mut3 oligonucleotide appears in both parts of the table. Gel shift analysis was performed using partially purified MBP–Tel2p and analyzed for amount of binding relative to R1-ST.
Methylation interference Dimethyl sulfate (DMS) methylation of 32P-labeled dsDNA (100 ng, 106 c.p.m.) was carried out in a 200 µl reaction containing 50 mM sodium cacodylate, pH 8.0, 1 mM EDTA, 1 µg salmon sperm DNA and 1 µl 1 M DMS. The reactions were incubated for 5 min at room temperature and quenched by addition of 50 µl stop buffer consisting of 1 M 2-mercaptoethanol, 1.5 M NaOAc and 1 µg glycogen as a carrier (43). The DNAs were ethanol precipitated twice and dissolved in sterilized milliQ-purified water and stored at –800001C. Five MBP–Tel2p gel retardation assay reactions (described above) were loaded and, following electrophoresis, the shifted bands were cut out and cast in a 0.7% agarose gel with DE81 paper (Whatman) next to the acrylamide gel slice. The labeled DNAs were run into the DE81 paper and the paper removed from the gel. The DNAs were eluted in 1.0 M NaCl, 10 mM Tris–HCl, pH 8.0, 1 mM EDTA, ethanol precipitated as above and subjected to 0.1 M piperidine hydrolysis. Other methods of elution included significant amounts of
1531 Nucleic Acids Acids Research, Research,1994, 1998,Vol. Vol.22, 26,No. No.16 Nucleic
1531
acrylamide with the DNA that interfered with electrophoresis of the hydrolyzed DNA. The hydrolyzed DNAs were lyophilized under vacuum to remove trace amounts of piperidine. The purified DNAs were then dissolved in 80% formamide and resolved on a 7 M urea–18% acrylamide gel (39). Other methods Total RNA was isolated from exponentially growing KR95-2A cells (20) as described in McKinney et al. (44) and quantitated by OD260. RESULTS Expression of fusion proteins Recombinant MBP–Rap1p and MBP–Tel2p were expressed in BL21(DE3) cells at various temperatures. Fusion proteins expressed at 370001C fractionated with the insoluble pellet. However, active soluble proteins were obtained by inducing expression at 220001C for 6–12 h. The predicted length of the Tel2p ORF is 688 amino acids. Fusion protein constructs containing full-length Tel2p were toxic to bacterial growth, even when cells were grown under repressing conditions. A 446 amino acid portion of the central region of Tel2p (amino acids 181–626) could be expressed in BL21 cells. The purified MBP–Rap1p and MBP–Tel2p proteins were 30–50% pure as judged by SDS–PAGE, with the major contaminating band being free MBP (as determined by Western blotting using anti-MBP as primary antibody; data not shown). Both the MBP–Rap1p and MBP–Tel2p purified fusion proteins migrated at the expected molecular weights of ∼140 and ∼100 kDa respectively (data not shown). The yields of final purified product were 0.3 mg for MBP–Tel2p and 0.5 mg for MBP–Rap1p per 500 ml culture. Tel2p binds yeast telomeric duplex DNA in a sequencespecific manner The binding specificity of Tel2p was initially determined in vitro using bacterial cell lysates and various duplex sequences (binding to single-stranded substrates will be considered elsewhere). These gel retardation assays all contained 0.5 µg sonicated salmon sperm DNA and 0.1 µg single-stranded oligonucleotide (∼40-fold molar excess over radiolabeled probes) as competitor DNAs. Using telomeric dsDNA probe R3-ST (Fig. 1 and Table 1) we observed a DNA–protein complex in the MBP–Tel2p lysate as well as the MBP–Rap1p lysate, but not in the MBP lysate. The MBP–Tel2p complex in lysates was specifically competed by a 20-fold molar excess of unlabeled probe (Fig. 1). Thus MBP– Tel2p can bind to the same telomeric DNA repeats as the known yeast telomere binding protein Rap1p. The specificity of Tel2p binding was further tested with various DNA probes using partially purified MBP–Tel2p. The major yeast telomere binding protein is Rap1p, which has a strongly conserved consensus binding site (45). Because MBP–Tel2p can bind to TG1–3 repeats containing three Rap1p sites, Tel2p binding to sequences containing one (R1-ST), two (R2-ST), three (R3-ST) or four (YTCA-1X) Rap1p sites was examined (Table 1). MBP-Tel2p bound all of the yeast telomeric dsDNA sequences to give a single shifted band (Fig. 2 and Table 1). No binding was observed with purified MBP alone (data not shown). Competition analysis showed that binding to the R2-ST probe by
Figure 1. Gel shift analysis of cell lysates from E.coli expressing MBP–Tel2p, MBP–Rap1p, and MBP used labeled R3-ST as probe. Increasing molar amounts (2×, 5×, 10× and 20×) of unlabeled R3-ST (Table 1) were used in competition with the MBP–Tel2p/R3-ST complex. Lane 1, R3-ST probe with no lysate; lane 2, R3-ST probe with MBP–Tel2p lysate; lanes 3–6, as lane 2 with increasing amounts of unlabeled R3-ST DNA added to the binding reaction; lane 7, R3-ST probe with MBP lysate; lane 8, R3-ST probe with MBP-Rap1p lysate. All lanes are from the same gel.
MBP–Tel2p was specifically competed by a 5-fold molar excess of R2-ST (Fig. 2B, lanes 3–5), but not by a 50-fold molar excess of duplex M13 oligonucleotide. Similar results were obtained with a 100-fold mass excess of pUC19 DNA sonicated to an average size of 0.7 kb (data not shown). The amount of R1-ST probe (Table 1) shifted in the presence of a 100- or 1000-fold mass excess of total S.cerevisiae RNA was also monitored (not shown). When normalized to the amount of R1-ST shifted in the absence of RNA, 0.5× R1-ST probe was shifted in the 100-fold RNA excess reaction and 0.3× was shifted in the 1000-fold RNA excess reaction. We note that the reactions containing a 50-fold molar excess of the non-specific duplex competitor M13 (Fig. 2B, lanes 6–8) also contained a 200-fold molar excess of one M13 single strand (Materials and Methods). This 200-fold molar excess of a single-stranded non-telomeric oligonucleotide did not inhibit MBP–Tel2p binding to R2-ST. As there was little or no competition by RNA and non-telomeric DNA, Tel2p is not a general DNA or RNA binding protein. Addition of anti-MBP Ab to the completed MBP–Tel2p R1-ST binding reaction caused the appearance of a supershifted band (Fig. 2C), consistent with purified MBP–Tel2p causing the R1-ST band shift. R1-ST, R2-ST, R3-ST and YTCA-1X were bound by Rap1p, as expected, and multiple Rap1p–DNA retarded complexes were observed for DNAs with multiple Rap1p sites (data not shown), in agreement with previous studies (46). MBP–Tel2p and MBP–Rap1p did not bind non-telomeric duplex sequences (M13, GAL4 and Y′ Telo) and no binding was observed with heterologous human, Tetrahymena or Oxytricha telomeric DNA sequence in our in vitro analyses (Table 1). These data show that Tel2p is a double-stranded TG1–3 binding protein in vitro.
1532 Nucleic Acids Research, 1998, Vol. 26, No. 6
Figure 3. MBP–Tel2p requires 8 bp of the 13 bp telomeric Rap1p site in R1-ST to bind DNA. The probe used in each binding reaction is indicated above each pair of lanes and the presence of MBP–Tel2p in the reaction is indicated by +. Mut10 shows that the specific G·C bp that shows the strongest methylation interference is required for binding. The MBP–Tel2p band is indicated by a line. (A) A scanned autoradiograph of representative binding reactions quantitated in Table 2. (B) Comparison of MBP–Tel2p and MBP–Rap1p binding to R1-ST and mut13 probes. The amount of probe shifted relative to MBP–Tel2p/R1-ST shift (as in Table 2) for MBP–Rap1p is 32× for mut3, 14× for mut 6, 0.0× for mut13 and 18× for R1-ST. Two shifted bands caused by a single Rap1p binding event have been observed previously (46,48).
Figure 2. Partially purified MBP-Tel2p binds duplex DNAs that contain telomeric sequences. (A) The probe is listed above the pair of lanes. The lane containing the binding reaction with MBP–Tel2p is indicated and the MBP–Tel2p shifted band is indicated by an arrow. All lanes are from the same gel. The sequences of the probes are given in Table 1. Lane 1, R1-ST probe alone; lane 2, R1-ST + MBP–Tel2p; lane 3, R2-ST; lane 4, R2-ST + MBP–Tel2p; lane 5, R3-ST; lane 6, R3-ST + MBP–Tel2p; lane 7, M13; lane 8, M13 + MBP-Tel2p. (B) Partially purified MBP–Tel2p specifically binds telomeric DNA. Partially purified MBP– Tel2p was used in the same experiments as in Figure 1A, performed using R2-ST as probe. The fold molar excesses of specific competitor, R2-ST, and non-specific competitor, M13 (Table 1), were 5×, 25× and 50×. Lane 1, R2-ST probe with no protein; lane 2, R2-ST probe with MBP–Tel2p protein; lanes 3–5, as lane 2 with 5× (lane 3), 25× (lane 4) and 50× (lane 5) molar excess of unlabeled R2-ST added to the binding reaction; lanes 6–8, as lane 2 with 5× (lane 6), 25× (lane 7) and 50× (lane 8) molar excess of unlabeled M13 duplex added to the binding reaction. (C) Partially purified MBP–Tel2p was used to shift the R1-ST probe in the absence or presence of 1.6 µg polyclonal anti-MBP Ab. Arrows indicate the shifted band and supershifted band.
The Tel2p–DNA complex migrated similarly to the fastest migrating Rap1p–DNA complex (see for example Figs 1 and 3B) and the intensity of the MBP–Tel2p shifted complex increased with increasing number of Rap1p binding sites and GT base pairs present in the DNA substrate (Fig. 2A, compare lane 2 with lanes 4 and 6). However, only a single MBP–Tel2p shifted band that increased in intensity with increasing protein levels irrespective of the number of Rap1p sites present on the DNA substrate (data not shown) was observed. While Tel2p bound to the telomeric duplex sequences containing Rap1p sites, Tel2p did not bind the non-telomeric TEF DNA probe (Table 1). The TEF probe contains a non-telomeric Rap1p site from the TEF2 UAS, which is the second highest affinity Rap1p site known (35). In addition, Tel2p did not bind the M13, GAL4 or Y′ Telo junction or the heterologous telomeric DNA sequences from humans, Oxytricha and Tetrahymena (Table 1). Thus Tel2p has the in vitro properties of a novel yeast telomere-specific binding protein. The results with the TEF probe indicate that Tel2p DNA binding is distinct from Rap1p binding.
1533 Nucleic Acids Acids Research, Research,1994, 1998,Vol. Vol.22, 26,No. No.16 Nucleic
1533
A minimum duplex telomeric sequence of 8 bp is required for Tel2p binding To delineate the minimum Tel2p binding site within the telomeric Rap1p site, a series of mutations were introduced into the single Rap1p site in the R1-ST probe and tested for Tel2p binding. The mutations were constructed by taking two 26 bp duplexes, R1-ST and M13 (Table 1), and replacing portions of the R1-ST sequence with M13 sequences. The M13 duplex was not bound by Tel2p, so as residues critical for binding are altered, the gel shift should be greatly reduced. To comply with the previous nomenclature for the Rap1p site (46) and to relate these data to the methylation interference analysis (below), the bases in the Rap1p site in the CA-rich strand are numbered 1–13 in the 5′!3′ direction and in the TG-rich strand are numbered 13′–1′ in the 5′!3′ direction (Table 2). The fraction of probe shifted by a constant amount of MBP–Tel2p was measured in a PhosphorImager and then normalized to R1-ST for a quantitative comparison of the different probes (Table 2). Gel shift analysis of duplex oligomers with bases mutated at various positions revealed that bases 1′, 3′ and 4′ of the Rap1p binding site were not necessary for MBP–Tel2p binding, because binding was observed in mut5 (Fig. 3 and Table 2). Methylation interference data below and TEF probe binding (Table 1) suggested that the G at 2′ is also not required for MBP–Tel2p binding (confirmed below). Individual base pairs in positions 4–6 (CCC), which are essential for Rap1p–DNA interaction (37,47), were dispensible for Tel2p binding because the mut8 and mut9 probes were shifted (Fig. 3 and Table 2). Similar mutations abolished binding by Rap1p (47). Mutant duplexes with single base substitutions (mut8–mut12 and mut3 in Table 2) revealed that Tel2p binding was undetectable only when a G!C base substitution at position 12 was made (mut10). Other single base substitutions at G13 (mut3), G10 (mut11), G8 (mut12), G6 (mut8) or G5 (mut9) did not eliminate Tel2p–DNA interaction. However, Tel2p binding was abolished with multiple base substitutions in the combinations G13′ and G10′ (mut2), G12′ and G10′–G8′ (mut7) or G13’ and G10′–G8′ (mut1). The shortest telomeric sequence that could be bound by Tel2p in this sequence context was mut5 (Fig. 3A and Table 2), which contains a 9 bp contiguous tract identical to the Rap1p site and G2′. Because G2′ was not identified as an important base by methylation interference (below) and was present in the TEF duplex that was not bound by Tel2p (Table 1) and G13′ was not required for binding, this analysis suggested that the 8 bp GT tract from bp 5′ to 12′ of the Rap1p site was required for Tel2p binding. To directly test the importance of G2′ in Tel2p binding, the mut13 probe bearing the 8 bp GT tract was made and tested. Tel2p shifted the mut13 probe as well as mut8, mut9 and several other probes (Fig. 3B and Table 2). Thus, G2′ was not required for Tel2p binding. The relative binding of equal amounts of partially purified MBP–Tel2p and MBP–Rap1p was compared using R1-ST, mut3, mut6 and mut13 substrates. Rap1p bound R1-ST, mut6 and mut3 efficiently as expected (47), but no binding of the mut13 probe was detected (Fig. 3B). An equal amount of MBP–Rap1p shifted 10- to 18-fold more R1-ST probe than MBP–Tel2p in independent experiments. These data further distinguish Rap1p and Tel2p binding of yeast telomeric DNA.
Figure 4. DMS interference with MBP–Tel2p and MBP–Rap1p bound to the duplex R1-ST probe. (A) Methylation interference of MBP–Rap1p binding to R1-ST. Free (total) DNA and Rap1p bound DNA are indicated. G5′, G6′, G12′ and G13′ show strong interference. (B) Methylation interference with MBP–Tel2p binding to R1-ST. G12′ shows strong interference. The equivalent of a longer exposure of the Tel2p shifted lane is shown to the left of the gel. In both panels arrows indicate bases that show significant interference.
Tel2p contacts duplex telomeric DNA in the major groove To further examine the specificity of MBP–Tel2p binding on the R1-ST probe, the DNA–protein contacts were studied by DMS interference analysis. DMS methylates the N7 position of guanines in the major groove of the DNA (39). Partially methylated DNA was used to form DNA–protein complexes with both MBP–Tel2p and MBP–Rap1p. The DNA from both complexes and the free probe were isolated and purified (Materials and Methods). As the free probe was in large excess over the amount of bound material, the free probe is equivalent to the input material. After piperidine cleavage the products were resolved on an 18% urea–acrylamide gel (Fig. 4). In agreement with previous reports (36,47,48), strong interference was observed at bases G5′, G6′, G12′ and G13′ in the Rap1p bound complex. The enhancement of bands near the 3′-end of the probe in the Rap1p bound material over the free probe has been observed previously by others (48) and did not alter the identification of bases critical for Rap1p binding. MBP–Tel2p showed a different pattern of DMS interference. Strong interference was observed at G12′ (Fig. 4). A comparison of protein–DNA contacts between the two proteins is shown in Figure 5. The strong interference at G12′ is consistent with the binding site mutations, as changing the G·C bp at position 12 to C·G (mut10, Table 2) eliminated Tel2p binding (Fig. 3). Thus the methylation interference data combined with the mutant oligo-
1534 Nucleic Acids Research, 1998, Vol. 26, No. 6
Figure 5. Bases identified by methylation interference of MBP–Rap1p and MBP–Tel2p binding. The ovals indicate bases with strong interference. Underlined bases make up the Rap1p binding site. The base numbering on each strand is shown after the convention of Gilson et al. (46).
nucleotide probes (Table 2) indicate that MBP–Tel2p specifically binds to bp 5–12 of the telomeric Rap1p site. DISCUSSION The data presented here show that Tel2p can specifically bind to telomeric dsDNA sequences in vitro. TEL2 and TEL1 are two yeast genes that behave as genetic regulators of telomere length (20). By sequence homology, Tel1p is a large serine/threonine kinase similar to DNA-PKcs (19,49). In contrast, Tel2p is a novel protein that has no significant homology to any protein in the database (20) or ‘telobox’ proteins that bind to telomeric DNA through Myb motifs (50). These data provide the first step for constructing testable models of Tel2p function in telomere length regulation and cell viability. Presently, it is unknown if Tel2p binds DNA in vivo and, if so, alone or in conjunction with another protein that alters or increases its affinity for specific sites. Purification of Tel2p from yeast cells would help answer these questions. However, since the TEL2 sequence and GAL1–TEL2 and TEL2–lacZ fusions suggest that Tel2p is a very low abundance protein (20), this biochemical approach will be challenging. The information that Tel2p can bind yeast telomeric dsDNA in vitro combined with yeast molecular genetics should allow identification of Tel2p interacting proteins. Production and analysis of these proteins in combination with Tel2p in DNA binding studies in vitro will define more clearly the potential Tel2p–DNA interactions in vivo. The MBP–Tel2p fusion protein could gel shift duplex telomeric oligonucleotides in the presence of a large excess of salmon sperm DNA and a single-stranded oligonucleotide present in each binding reaction (Materials and Methods) and a non-specific competitor (Figs 1 and 2 and Table 1). Competition for Tel2p binding of telomeric DNA was specific (Figs 1 and 2). In addition, not all duplex probes tested were bound by MBP–Tel2p (Table 1). These data and the methylation interference results indicate that MBP–Tel2p shows sequence specificity in its dsDNA binding. At least 8 bp of the 13 bp Rap1p telomeric site were required for Tel2p binding and the G·C bp at position 12 (Fig. 5) was essential (Fig. 3 and Table 2). In contrast, Rap1p did not bind to this 8 bp sequence (mut13, Fig. 3B). Because Tel2p could bind duplexes with point mutations at positions 6, 8 and 10 (mut8, mut12 and mut11, Table 2), Tel2p may make other
contacts outside the major groove that compensate for these single base changes. Tel2p has not been previously identified as part of a telomeric DNA binding activity in yeast extracts, most likely for two reasons. First, Rap1p is an abundant yeast telomere binding activity in yeast extracts (35), making detection of additional yeast telomere binding proteins difficult. Partial fractionation of yeast nuclear extracts has revealed the presence of a second non-Rap1p telomere binding activity called TBFβ (34), but this activity has not been further characterized. It is not yet known if Tel2p is TBFβ or part of TBFβ. Second, expression of Tel2p is toxic to bacteria. Indeed, even placing the TEL2 ORF behind strong yeast promoters, which are active in bacteria makes plasmids unstable in Escherichia coli such that transformants grow poorly and rearranged plasmids are frequently isolated (K.W.Runge, unpublished data). Only by constructing a truncated bacterial fusion that contains 65% of the TEL2 ORF were we able to obtain unrearranged plasmids in bacteria and produce protein. Thus, previous searches for yeast telomeric binding proteins with oligonucleotide probes (51) might not have found TEL2 because high level expression would inhibit bacterial colony formation. The known homology of Tel1p with DNA-PKcs, the fact that TEL1 and TEL2 act in the same genetic pathway and these data showing that Tel2p can bind to telomeric DNA suggest a model for Tel1p and Tel2p function in telomere length regulation. DNA-PK is a DNA-dependent protein kinase in which the catalytic subunit, DNA-PKcs, is activated by binding to two DNA-bound subunits, Ku-70 and Ku-80, and can phosphorylate a number of proteins critical to cell cycle progression and DNA repair, including p53 (52). By analogy, a speculative model for Tel2p function is that Tel2p binds to telomeric DNA and, in conjunction with another protein bound to the chromosome terminus, activates Tel1p to phosphorylate telomere-associated proteins such as Rap1p, Rif1p (30), Rif2p (31), Sir2p, 3p and 4p (53), Est1p (32), Cdc13p/Est4p (33) or Stn1p (54) in the course of telomere length regulation. It is interesting to note that the point mutation in tel2-1 that causes cells to maintain short telomeres occurs in a part of the protein that is not present in the MBP–Tel2p fusion. In our model the tel2-1 protein would bind DNA but fail to bind to or activate the Tel1p kinase. Deletion of TEL2 is lethal to cells, while deletion of TEL1 results in viable cells with short telomeres (19,20). Thus, Tel2p must have an essential function besides its activities in the TEL1 pathway. The physical similarity between telomeres and doublestrand breaks, the evolutionary conservation of the telomere-related protein Tel1p and the double-strand break-activated kinase DNA-PKcs and the data presented here showing that Tel2p can bind telomere DNA suggest that the essential function of Tel2p involves metabolism of DNA ends. Another Tel1p/DNA-PKcs homolog, Mec1p, controls cell cycle progression in response to DNA damage (55) and is essential for viability in some yeast strains (21). The present data on tel2 mutations and DNA binding are consistent with an essential role for Tel2p in this process. If so, TEL2 should show genetic interactions with components of the DNA damage signalling pathway. Thus, these data on Tel2p DNA binding suggest specific testable models for the function of this novel protein with no known homologs.
1535 Nucleic Acids Acids Research, Research,1994, 1998,Vol. Vol.22, 26,No. No.16 Nucleic ACKNOWLEDGEMENTS This work was supported by NIH grant GM50752 and a American Cancer Society Junior Faculty Research Award to K.W.R. We thank Drs Donal Luse for helpful advice during this project and Richard Gronostajski for critical reading of the manuscript. REFERENCES 1 Harley,C.B. and Villeponteau,B. (1995) Curr. Opin. Genet. Dev., 5, 249–255. 2 Wilkie,A.O.M., Lamb,J., Harris,P.C., Finney,R.D. and Higgs,D.R. (1990) Nature, 346, 868–871. 3 Wellinger,R.J. and Zakian,V. A. (1989) Proc. Natl. Acad. Sci. USA, 86, 973–977. 4 Zakian,V.A., Blanton,J.M. ,Wetzel,L. and Dani,G.M. (1986) Mol. Cell. Biol., 6, 925–932. 5 Murray,A.W., Schultes,N.P. and Szostak,J.W. (1986) Cell, 45, 529–536. 6 Sandell,L.L., Gottschling,D.E. and Zakian,V.A. (1994) Proc. Natl. Acad. Sci. USA, 91, 12061–12065. 7 Sandell,L.L. and Zakian,V.A. (1993) Cell, 75, 729–739. 8 Zakian,V.A. (1989) Annu. Rev. Genet., 23, 579–604. 9 Harley,C.B. (1995) In Blackburn,E.H. and Greider,C.W. (ed.), Telomeres. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pp. 247–263. 10 de Lange,T. (1995) In Blackburn,E.H. and Greider,C.W. (ed.), Telomeres. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pp. 265–293 11 Blackburn,E.H. (1992) Annu. Rev. Biochem., 61, 113–129. 12 Singer,M.S. and Gottschling,D.E. (1994) Science, 266, 404–409. 13 Wellinger,R.J., Ethier,K., Labrecque,P. and Zakian,V.A. (1996) Cell, 85, 423–433. 14 van Steensel,B. and de Lange,T. (1997) Nature, 385, 740–743. 15 Marcand,S., Gilson,E. and Shore,D. (1997) Science, 275, 986–990. 16 Lingner,J., Hughes,T.R., Shevchenko,A., Mann,M., Lundblad,V. and Cech,T.R. (1997) Science, 276, 561–567. 17 Conrad,M.N., Wright,J.H., Wolf,A.J. and Zakian,V.A. (1990) Cell, 63, 739–750. 18 Lustig,A.J. and Petes,T.D. (1986) Proc. Natl. Acad. Sci. USA, 83, 1398–1402. 19 Greenwell,P.A., Kronmal,S.L., Porter,S.E., Gassenhuber,J., Obermaier,B. and Petes,T.D. (1995) Cell, 82, 823–829. 20 Runge,K.W. and Zakian,V.A. (1996) Mol. Cell. Biol., 16, 3094–3105. 21 Morrow,D.M., Tagle,D.A., Shiloh,Y., Collins,F.S. and Heiter,P. (1995) Cell, 82, 831–840. 22 Savitsky,K., Bar-Shira,A., Gilad,S., Rotman,G., Ziv,Y., Vanagaite,L., Tagle,D.A., Smith,S., Uziel,T., Sfez,S., Ashkenazi,M., Pecker,I., Frydman,M., Harnik,R., Patanjali,S.R., Simmons,A., Clines,G.A., Sartiel,A., Gatti,R.A., Chessa,L., Sanal,O., Lavin,M.F., Jaspers,N.G.J., Taylor,A.M.R., Arlett,C.F., Miki,T., Weissman,S.M., Lovett,M., Collins,F.S. and Shiloh,Y. (1995) Science, 268, 1749–1753. 23 Zakian,V.A. (1996) Annu. Rev. Genet., 30, 141–172.
24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56
1535
Li,B. and Lustig,A. J. (1996) Genes Dev., 10, 1310–1326. Garvik,B., Carson,M. and Hartwell,L. (1995) Mol. Cell. Biol., 15, 6128–6138. Runge,K.W. and Zakian,V.A. (1989) Mol. Cell. Biol., 9, 1488–1497. Kyrion,G., Boakye,K.A. and Lustig,A.J. (1992) Mol. Cell. Biol., 12, 5159–5173. Liu,C., Mao,X. and Lustig,A.J. (1994) Genetics, 138, 1025–1040. Lustig,A.J., Kurtz,S. and Shore,D. (1990) Science, 250, 549–553. Hardy,C.F.J., Sussel,L. and Shore,D. (1992) Genes Dev., 6, 801–814. Wotton,D. and Shore,D. (1997) Genes Dev., 11, 748–760. Virta-Pearlman,V., Morris,D.K. and Lundblad,V. (1996) Genes Dev., 10, 3094–3104. Nugent,C.I., Hughes, T.R., Lue,N.F. and Lundblad,V. (1996) Science, 274, 249–252. Liu,Z.P. and Tye,B.K. (1991) Genes Dev., 5, 49–59. Buchman,A.R., Lue,N.F. and Kornberg,R.D. (1988) Mol. Cell. Biol., 8, 5086–5099. Longtine,M.S., Wilson, N.M., Petracek,M.E. and Berman.,J. (1989) Curr. Genet., 16, 225–240. Graham,I.R. and Chambers,A. (1994) Nucleic Acids Res., 22, 124–130. Shore,D. and Nasmyth,K. (1987) Cell, 51, 721–732. Sambrook,J., Fritsch,E.F. and Maniatis,T. (1989) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor LaboratoryPress, Cold Spring Harbor, NY. Sheng,H., Hou,Z., Schierer,T., Dobbs,D.L. and Henderson,E. (1995) Mol. Cell. Biol., 15, 1144–1153. Schiestl,R.H. and Gietz,R.D. (1989) Curr. Genet., 16, 339–346. Liu,H., Krizek,J. and Bretscher,A. (1992) Genetics, 132, 665–673. Baldwin,A.S.,Jr (1992) In Ausubel,F.M., Brent,R., Kingston,R.E., Seidman,J.G., Smith,J.A. and Struhl,K. (eds), Short Protocols in Molecular Biology. John Wiley & Sons, New York, NY, pp. 12-8–12-10 McKinney,J.D., Chang,F., Heintz,N. and Cross,F.R. (1993) Genes Dev., 7, 833–843. Shore,D. (1994) Trends Genet., 10, 408–412. Gilson,E., Roberge,M., Giraldo,R., Rhodes,D. and Gasser,S.M. (1993) J. Mol. Biol., 231, 293–310. Vignais,M.L., Huet,J., Buhler,J.M. and Sentenac,A. (1990) J. Biol. Chem., 265, 14669–14674. Chambers,A., Tsang, J.S., Stanway,C., Kingsman,A.J. and Kingsman,S.M. (1989) Mol. Cell. Biol., 9, 5516–5524. Hartley,K.O., Gell,D., Smith,G.C.M., Zhang,H., Divecha,N., Connelly,M.A., Admon,A., Lees-Miller,S.P., Anderson,C.W. and Jackson,S.P. (1995) Cell, 82, 849–856. Brun,C., Marcand,S. and Gilson,E. (1997) Trends Cell Biol., 7, 317–324. Lin,J.-J. and Zakian,V. A. (1994) Nucleic Acids Res., 22, 4906–4913. Anderson,C.W. (1993) Trends Biochem. Sci., 18, 433–437. Grunstein,M. (1997) Curr. Opin. Cell Biol., 9, 383–387. Grandin,N., Reed,S.I. and Charbonneau,M. (1997) Genes Dev., 11, 512–527. Weinert,T.A., Kiser,G.L. and Hartwell,L.H. (1994) Genes Dev., 8, 652–665. Runge,K.W. and Zakian,V.A. (1990) Nucleic Acids Res., 18, 1783–1787.