The Plant Journal (2000) 23(5), 1±11
Analysis of the G-overhang structures on plant telomeres: evidence for two distinct telomere architectures Karel Riha1, Thomas D. McKnight2, Jiri Fajkus3, Boris Vyskot4 and Dorothy E. Shippen1,* 1 Department of Biochemistry and Biophysics, Texas A & M University, College Station, Texas 77843-2128, USA, 2 Department of Biology, Texas A & M University, College Station, Texas 77843-3258, USA, 3 Department of Analysis of Biologically Important Molecular Complexes, Masaryk University, and Institute of Biophysics of the Czech Academy of Sciences, Kralovopolska 135, 612 65 Brno, Czech Republic, and 4 Institute of Biophysics, Czech Academy of Sciences, Kralovopolska 135, 612 65 Brno, Czech Republic Received 10 April 2000; revised 9 June 2000; accepted 9 June 2000 *For correspondence (fax: +1 979 845 9274; E-mail: [email protected]
Summary Telomeres are highly conserved structures essential for maintaining the integrity of eukaryotic genomes. In yeast, ciliates and mammals, the G-rich strand of the telomere forms a 3¢ overhang on the chromosome terminus. Here we investigate the architecture of telomeres in the dicot plants Silene latifolia and Arabidopsis thaliana using the PENT (primer extension/nick translation) assay. We show that both Arabidopsis and Silene telomeres carry G-overhangs longer than 20±30 nucleotides. However, in contrast to yeast and ciliate telomeres, only half of the telomeres in Silene seedlings possess detectable G-overhangs. PENT reactions using a variety of primers and reaction conditions revealed that the remaining fraction of Silene telomeres either carries no overhangs or overhangs less than 12 nucleotides in length. G-overhangs were observed in Silene seeds and leaves, tissues that lack telomerase activity. These ®ndings suggest that incomplete DNA replication of the lagging strand, rather than synthesis by telomerase, is the primary mechanism for G-overhang synthesis in plants. Unexpectedly, we found that the fraction of telomeres with detectable G-overhangs decreased from 50% in seedlings to 35% in leaves. The difference may re¯ect increased susceptibility of the G-overhangs to nuclease attack in adult leaves, an event that could act as a precursor for the catabolic processes accompanying leaf senescence. Q1
Introduction McClintock (1939, 1941) demonstrated that telomeres provide a protective cap for the ends of eukaryotic chromosomes, and we now know that telomeres play essential roles in many cellular processes including DNA replication, cell cycle progression, meiosis and mitosis (Cooper et al., 1998; Kirk et al., 1997; Zakian, 1996). In plants, as in animals, telomeric DNA consists of short G-rich repeat sequences that are synthesized and maintained by the ribonucleoprotein enzyme telomerase (reviewed in Shippen and McKnight, 1998). The structure of the chromosome terminus is remarkably conserved, and in all organisms examined the guanine-rich strand (G-strand) of the telomere extends beyond the complementary cytosine-rich strand (C-strand) to create a 3¢ protrusion (Henderson and Blackburn, 1989; Klobutcher et al., 1981; McElligott and Wellinger, 1997; Makarov et al., ã 2000 Blackwell Science Ltd
1997; Venkatesan and Price, 1998; Wellinger et al., 1993; Wright et al., 1997). A proper terminal structure is critical for the capping function performed by natural telomeres (van Steensel et al., 1998), and several proteins that in¯uence the processing or exposure of the chromosome terminus have been described (Gravel et al., 1998; Nugent et al., 1996; Parenteau and Wellinger, 1999). For example, single-stranded termini of yeast chromosomes are capped by Cdc13p (Garvik et al., 1995), while in human cells, T-loops are created by tucking G-overhangs back into the duplex region of telomeres through interactions with TRF2 (Grif®th et al., 1999). Despite the array of G-strand binding proteins, it remains unclear whether the G-overhang structure per se is essential for proper telomere function, or is simply a by-product of the telomere replication machinery. 1
Karel Riha et al.
Three mechanisms have been proposed for the generation of G-overhangs: incomplete synthesis of the telomeric C-strand by lagging strand DNA replication, synthesis of G-overhangs by telomerase, and nucleolytic degradation of the C-strand (Lingner et al., 1995; Wellinger et al., 1996). In yeast, G-overhangs longer than 30 nucleotides are present on both ends of linear DNA molecules, even in cells that lack telomerase activity (Dionne and Wellinger, 1996; Wellinger et al., 1996). These structures appear transiently at S-phase, and their synthesis is predicated upon the passage of replication forks (Dionne and Wellinger, 1998). Since the semi-conservative DNA replication machinery generates molecules that contain a 3¢ overhang on one end and a blunt end on the other, it is likely that a C-strand-speci®c nuclease is involved in creating symmetric telomeres on yeast chromosomes after passage of the replication machinery (Wellinger et al., 1996). G-overhangs may also be found in vertebrate cells that lack telomerase, but these structures persist throughout the cell cycle (McElligott and Wellinger, 1997; Makarov et al., 1997; Venkatesan and Price, 1998; Wright et al., 1997). In addition, it is unclear whether long G-overhangs exist on both ends of mammalian chromosomes. Using the primer extension/nick translation (PENT) assay, Makarov et al. (1997) observed G-overhangs of approximately 200 nucleotides on more than 80% of human chromosome ends. In contrast, Wright et al. (1997), using an af®nity puri®cation approach, found that human chromosomes are asymmetric, having a long G-overhang at one end and no overhang or a very short protrusion (< 12 nucleotides) on the opposite end. Recent studies indicate that these asymmetric G-overhangs appear immediately after the telomere is replicated (Wright et al., 1999). Thus, the timing of telomere replication and processing in human cells is clearly distinct from the situation in yeast. Despite their more indeterminate developmental programmes, plants share many common features in the regulation of telomere length and telomerase activity with animals. As in mammals, telomerase expression is strongly correlated with cell proliferation and de-differentiation (Fajkus et al., 1998; Fitzgerald et al., 1996; Fitzgerald et al., 1999; Heller et al., 1996; Kilian et al., 1998; Riha et al., 1998), and telomere length is precisely regulated during development (Kilian et al., 1995; Riha et al., 1998). Here we examine the architecture of chromosome ends in the higher plants Silene latifolia and Arabidopsis thaliana using the PENT method. Our results show that G-overhangs longer than 20±30 nucleotides are present in both organisms. However, in contrast to the previous PENT study of mammalian DNA, we detected G-overhangs on only half or fewer chromosome ends, implying that two distinct telomere architectures exist in plants.
Figure 1. The PENT reaction on a hypothetical Silene telomere. A new telomeric C-strand (Cs) is extended from a primer annealed to the G-overhang by Taq polymerase or DNA polymerase I. At the same time, an original Co strand is trimmed by the 5¢ ® 3¢ exonuclease activity of the polymerase. In the absence of dGTP, the reaction stops across from the ®rst cytosine in the TAS. A Taq I restriction site in the TAS is indicated.
Results Silene latifolia telomeres have G-overhangs To study plant telomeres, we modi®ed the primer extension/nick translation (PENT) assay originally developed to study human telomeres (Makarov et al., 1997). In the PENT assay, a telomeric oligonucleotide, in our case (C3TA3)3, is hybridized to available single-stranded G-overhangs (T3AG3)n under non-denaturing conditions, and is elongated by a DNA polymerase with 5¢ to 3¢ exonuclease activity, such as Taq polymerase or DNA polymerase I (Figure 1). In the presence of dATP, dCTP and dTTP (but not dGTP), a newly synthesized Cs strand will replace the original Co telomeric strand until the reaction stops across from the ®rst cytosine in the telomere-associated sequences (TAS). The nick between Cs and Co strands permits separation of Cs strands from bulk DNA by alkaline electrophoresis. DNA products are detected by hybridization with a (T3AG3)4 probe. The telomeres in Silene range from 2.5 to 4.5 kb (Riha et al., 1998), making this plant suitable for G-overhang analysis by PENT. Furthermore, since the processivity of Taq polymerase is signi®cantly greater than 4.5 kb, the original Co telomeric strand should be completely replaced by a Cs strand. To test this prediction, we performed PENT reactions with DNA extracted from leaves. The (C3TA3)3 oligonucleotide was elongated by Taq polymerase at 55°C in the presence of all four dNTPs. A time-course analysis of PENT reactions containing all four dNTPs showed gradual elongation of Cs strands by approximately 300 nucleotides per minute (Figure 2a, lanes 1±5), a rate similar to what was observed in PENT reactions with human DNA (Makarov et al., 1997). After 20 min, the Cs extended beyond the size ã Blackwell Science Ltd, The Plant Journal, (2000), 23, 1±11
Telomere structure in Silene and Arabidopsis
synthesis of the Cs product (Figure 2c, lane 2). The absence of any low molecular weight products ruled out the possibility that products originated from extension of an endogenous nick in the C-rich telomeric strand. One half of the telomeres in Silene seedlings possess long G-overhangs
Figure 2. Synthesis of the Cs extension product is G-overhang-speci®c. (a) Time-course analysis of PENT products synthesized on Silene telomeres. Reactions with DNA extracted from leaves were performed either for 0, 5, 10, 20 or 30 min in the presence of all four dNTPs (lanes 1±5). DNA cut with Taq I was run in the same gel as a control (lane 6). The DNA was transferred to a membrane and hybridized with a [32P]labelled telomeric probe. (b) A PENT reaction was carried at 55°C with seedling DNA for 1 h without dGTP (lane 1). TRFs generated with Taq I were run in lane 2 as a size control. (c) DNA from seedlings was treated with exonuclease I and PENT reactions were performed with (C3TA3)3 and dATP, dCTP and dTTP at 55°C (lane 2). PENT products were detected by in-gel hybridization. A control reaction in the absence of exonuclease I is shown in lane 1.
of the terminal restriction fragments (TRFs) generated by Taq I restriction endonuclease (Figure 2a, lanes 4 and 6), indicating that polymerization continued into the TAS. In contrast, when the PENT reaction was performed in the absence of dGTP, synthesis stopped when the Cs reached the size of the TRFs, although the reactions were carried out for 60 min (Figure 2b, lane 1). It is likely that all available G-overhangs were extended in these reactions, because we used a large excess of oligonucleotide and the product pro®le did not change after incubation for another 60 min with fresh enzyme (data not shown). In the absence of dGTP, the distribution of Cs PENT products resembled the pattern of TRFs generated by Taq I (Figure 2b, lanes 1 and 2), indicating that polymerization of Cs initiated at the end of a chromosome and covered the entire telomeric region, stopping at a telomere/sub-telomere junction. To verify that Cs strand synthesis was dependent on the presence of terminal G-overhangs, we pre-treated genomic DNA from seedlings with exonuclease I, which speci®cally degrades single-stranded DNA from 3¢ ends. Pre-treatment with exonuclease I completely abolished ã Blackwell Science Ltd, The Plant Journal, (2000), 23, 1±11
In addition to the Cs signal, we observed strong hybridization to high molecular weight DNA in all PENT reactions (Figures 2±6). It is formally possible that this signal was derived from interstitial telomeric repeats. In Silene, interstitial telomeric repeats are evident after cleavage with Taq I as several discrete bands ranging from 0.5 to 1.5 kb (Figure 3a, lane 1; Riha et al., 1998). However, this hybridization signal is completely removed by the highstringency washing conditions used in our PENT assays (Figure 3a, lane 2; see Experimental procedures). No low molecular weight bands were ever observed in Taq I restriction reactions run alongside PENT samples. Hence, the signal in bulk DNA is derived from original Co strands. The model presented in Figure 1 shows that when a primer anneals to a terminal G-overhang, its elongation proceeds by nick translation until the Co strand is completely degraded. Another possible explanation for the Co fraction is that priming occurred on an internal stretch of single-stranded G-rich DNA. The T-loops on human DNA are generated by tucking the 3¢ singlestranded overhang into the duplex region of the telomere (Grif®th et al., 1999; Figure 3d). Such a con®guration would cause melting of an internal region to create a binding site for the PENT primer. In this case, the 5¢ end of the chromosome would not be subject to nuclease degradation ahead of Cs synthesis and primer elongation would continue by strand displacement rather than nick translation. To investigate whether priming occurred on an internal stretch of G-rich telomeric DNA, we compared the extension products generated with DNA polymerase I and Klenow fragment. Because Klenow fragment does not have 5¢ ® 3¢ exonuclease activity, primer elongation must occur via strand displacement. PhosphorImager analysis revealed that Klenow fragment and DNA polymerase I produced the same amount of Cs (Figure 3b,c). However, the fraction of Co was reduced by one half in the reaction catalysed by DNA polymerase I (Figure 3b,c), consistent with degradation of the Co strand. In contrast, Klenow fragment produced Co in an amount similar to the total telomeric signal for undigested DNA (Figure 3b, lanes 2 and 4). This result is also re¯ected in the Co:Cs ratios, which are 1:1 (49:51) for the DNA polymerase I reaction and 2:1 (64:36) for the Klenow reaction (Figure 3c). These data provide strong evidence that the Co strand is degraded and not displaced during the course of the
Karel Riha et al.
PENT reaction. Therefore, we conclude that the Co signal represents telomeres with no or very short G-overhangs. To quantify the fraction of telomere carrying detectable G-overhangs, we calculated the ratio Is/(Is + Io), where Is and Io represent the intensities of corresponding Co and Cs signals in a single reaction. The amount of Cs synthesized was measured in PENT reactions with several different oligonucleotides (Figure 4). The average amount of Cs
product at 55°C from the (C3TA3)3 and (C3TA3)4CCC reactions was 49.2% 6 1.66% (n = 6; two reactions with each of three different DNA samples) (Figure 4, lanes 5 and 6). These data imply that the remaining telomeres have either no G-overhangs or overhangs that are too short to be detected by this method (< 18 bases).
Figure 4. Quanti®cation of telomeres with G-overhangs in telomerasepositive seedlings. DNA from Silene seedlings was subjected to the PENT reaction with different primers at various temperatures. Reactions were with DNA polymerase I (lanes 1±3) or Taq polymerase (lanes 4±8). PENT products were detected by in-gel hybridization. The intensity of the Cs signal relative to the total signal in each lane was determined by PhosphorImager analysis and is reported below the lane.
Figure 3. The Co signal is derived from telomeres without detectable G-overhangs. (a) Silene DNA was cut with Taq I, separated in 1% native agarose gel and analysed by Southern hybridization with a telomeric probe. The membrane was ®rst washed at low stringency (2 3 SSC, 55°C) and exposed to ®lm (lane 2). The membrane was then washed at high stringency (0.2 3 SSC, 55°C) (lane 2). The signals from interstitial sequences are indicated by arrowheads. (b) Primer extension reactions were performed with seedling DNA in the absence of dGTP, using either DNA polymerase I or Klenow fragment (lanes 1 and 2). Reactions were carried out at 16°C for 8 h. Untreated DNA and DNA cut with Taq I were run in the same gel as controls (lanes 4 and 3). Reaction products were detected by in-gel hybridization. (c) PhosphorImager analysis of products from DNA polymerase I and Klenow fragment reactions. (d) Model for primer elongation on internal telomeric DNA in a T-loop. Synthesis of the Cs strand is indicated by the dashed arrow.
ã Blackwell Science Ltd, The Plant Journal, (2000), 23, 1±11
Telomere structure in Silene and Arabidopsis
The Co signal may be under-estimated if the high molecular weight products are not ef®ciently transferred to a membrane for hybridization (Wright et al., 1997). Therefore, we used a second approach to detect PENT products, omitting the transfer step and performing the hybridization with the PENT products directly in the gel. If the original DNA samples were of high quality, we found no difference in the Is/(Is + Io) ratio using these two approaches (data not shown; see Figures 4 and 5). Most of our experiments were carried out using the in-gel hybridization approach. To learn more about the length of the G-overhangs, DNA from seedlings was subjected to PENT with (C3TA3)4CCC and (C3TA3)6 at different temperatures (Figures 4 and 6b). We detected elongation products in the 55, 65 and 72°C reactions, although the Cs fraction dropped to 32% in the 72°C reactions. Because both oligonucleotides were elongated very ef®ciently at 72°C when annealed to a telomeric sequence in a control construct (data not shown), we estimate the length of the G-overhangs to be at least 20±30 nucleotides. Since a G-overhang must be at least 18 nucleotides long for ef®cient annealing of the (C3TA3)4CCC oligonucleotide at 55°C (Table 1), we could not exclude the presence of G-overhangs shorter than 18 nucleotides in the Co fraction. Therefore, we increased the sensitivity of the method by annealing the oligonucleotides TelAAA12 and TelCCC13 (Table 1) to seedling genomic DNA at 16°C and conducting elongation reactions with DNA polymerase I (Figure 4, lanes 1±3). The higher temperature optimum of Taq polymerase precludes experiments in this temperature range. To reveal possible differences in the sequence permutation of the short G-tails, these oligonucleotides had different 3¢ termini (three C residues or three A residues). If G-overhangs between 12 and 18 nucleotides were present in the Co fraction of telomeres, we would expect a dramatic increase in the Cs signal (> 50%). Instead, we observed a decrease in the Cs signal (37±39% versus 50% in reactions with Taq polymerase in Figure 4). This ®nding is consistent with some chromosome ends carrying G-overhangs shorter than 12 nucleotides. We postulate that the decrease in Cs signal compared to reactions with Taq polymerase at 55°C is caused by decreased processivity of DNA polymerase I at 16°C. Even after 4 h of extension, the Cs strands did not completely reach the size of TRFs, as is apparent from the slight shift of the PENT products synthesized with DNA polymerase I (Figure 4, compare lanes 1±3 with 4±6). The G-overhang signal is reduced in Silene leaves We previously demonstrated that telomerase activity is elevated in tissues with high proliferation potential, such as young seedlings, but is dramatically down-regulated in ã Blackwell Science Ltd, The Plant Journal, (2000), 23, 1±11
Figure 5. Detection of G-overhangs in telomerase negative-tissues. (a) PENT reactions were performed at 55°C with DNA from Silene seedlings, seeds and leaves. (C3TA3)3 was used as a primer. PENT products were detected following transfer of the DNA to a membrane. (b) PhosphorImager pro®les of individual lanes. Relative Cs and Co values are indicated.
mature seeds and leaves (Riha et al., 1998). To investigate the role of telomerase in creating or sustaining G-overhangs in plants, we performed PENT assays at 55°C with DNA isolated from seedlings, leaves and seeds (Figure 5a). Surprisingly, the Is/(Is + Io) ratio was approximately 50% in both telomerase-positive seedlings and telomerasenegative seeds, but dropped to 36.8% in leaves (Figure 5b). Quanti®cation of PENT products resulting from 55°C extension reactions (Figure 6a,b) showed that the average intensity of the Cs signal from adult leaf DNA was
Karel Riha et al. Table 1. Oligonucleotides and their melting point temperature Oligonucleotide
TelAAA12(CTAAACCCTAAA) TelCCC13(AAACCCTAAACCC) (C3TA3)3 (C3TA3)4CCC (C3TA3)6
25.0 38.7 57.2 70.2 74.7
*As calculated using melting point temperature (Tm) prediction software (Virtual Genome Center Website; http://alces.med.umn. edu/rawtm.html
Figure 6. Quanti®cation of telomeres with G-overhangs in Silene leaves. (a) DNA from leaves was subjected to PENT with different primers at various temperatures. PENT products were detected by in-gel hybridization. (b) Quantitative analysis of Cs signals generated at various temperatures with DNA from seedlings and leaves. Data represent means and standard deviations of the intensity of Cs signal relative to total telomeric signal from six experiments for seedlings and leaves at 55°C (P < 0.001); otherwise n = 2 for seedlings and n = 1 for leaves.
33.5% 6 4.9% (n = 6; two reactions with each of three different DNA samples). G-overhangs were also detected in reactions with leaf DNA performed at 65 and 72°C (Figure 6a,b). We conclude that G-overhangs are present on telomeres from cells that lack telomerase. However, since the G-overhang signal is reduced from 50% in seeds to 35% in leaves, the terminal structure of chromosomes is not identical in all telomerase-negative tissues. Detection of G-overhangs in Arabidopsis To determine whether G-overhangs are a conserved feature of plant telomeres, we subjected genomic DNA isolated from entire Arabidopsis thaliana plants to the PENT assay. Elongation of telomeric oligonucleotides at temperatures from 16 to 72°C consistently resulted in Cs synthesis (Figure 7a), and the reaction was sensitive to the pre-treatment of genomic DNA with exonuclease I (Figure 7b). Similar to the telomeres of Silene, TRFs in Arabidopsis range in size from 2.5 to 5 kb. However, in contrast to Silene telomeres, the centromere-proximal
Figure 7. Detection of G-overhangs on Arabidopsis telomeres. (a) PENT reactions with different primers at different temperatures were performed on DNA from Arabidopsis. PENT products were detected by in-gel hybridization. (b) Sensitivity of Cs product to exonuclease degradation. Arabidopsis DNA was pre-treated with exonuclease I and subjected to PENT with (C3TA3)3 at 55°C. (c) Structure of a hypothetical Arabidopsis telomere based on the nucleotide sequences of three Arabidopsis telomere-associated sequences ± YpAtT1 and YpAtT7 (Richards et al., 1992) and pAtT51 (Genebank accession number AC007348).
region of the TRFs in Arabidopsis contains degenerate telomeric repeats, primarily (TCTAGGG)n (Richards et al., 1992; Figure 7c). Since approximately 75% of the telomeric ã Blackwell Science Ltd, The Plant Journal, (2000), 23, 1±11
Telomere structure in Silene and Arabidopsis signal was concentrated in high molecular weight DNA products (Figure 7a, lanes 1±5 and data not shown), this signal is likely to be derived from telomeres without detectable G-overhangs as well as from sub-terminal regions of the telomeres that have G-overhangs. The signal may also include recognition of interstitial repeats known to be located in the centromeric region of Arabidopsis chromosomes (Richards et al., 1991). Hybridization washing conditions were not optimized to exclude signal from Arabidopsis interstitial telomeric repeats. In contrast to our ®ndings with Silene, synthesis of the Cs strand did not reach the full size of TRFs generated with AluI restriction endonuclease. Instead, the reaction stopped at about 3 kb (Figure 7a). This observation is consistent with polymerase reaching a cytosine within the region of degenerate telomeric repeats. Nevertheless, we can conclude that G-overhangs are a feature of Arabidopsis telomeres. Discussion The structural and functional similarity of telomeres in eukaryotes indicates that this chromosome-capping mechanism evolved early and remains an essential feature of the cellular machinery. Using the PENT assay, we provide further evidence for a conserved telomere architecture by demonstrating that telomeres in the plants Silene and Arabidopsis carry single-stranded G-overhangs similar to those identi®ed on ciliate, yeast, chicken and human telomeres (Henderson and Blackburn, 1989; Klobutcher et al., 1981; McElligott and Wellinger, 1997; Makarov et al., 1997; Venkatesan and Price, 1998; Wellinger et al., 1993; Wright et al., 1997). Unexpectedly, we found that G-overhangs greater than 20±30 nucleotides are present on only 50% of the DNA molecules from Silene. The previous application of PENT to human DNA found long G-overhangs on approximately 80% of the ends and this result was interpreted to mean that essentially all human telomeres carry G-overhangs (Makarov et al., 1997). While we cannot rule out the possibility that a substantial fraction of G-overhangs in Silene are engaged in G±G base pairing interactions, we optimized our reaction conditions to ensure that all available G-overhangs were observed. Nevertheless, there remained a possibility that some of the G-overhangs were engaged in T-loops which could preclude their detection by PENT. T-loops are notoriously fragile and have only been seen following gentle chromatin extraction (Murti and Prescott, 1999) or cross-linking to stabilize binding of the TRF2 protein (Grif®th et al., 1999). Since our DNA samples were puri®ed by centrifugation in a caesium chloride gradient, it is very unlikely that any T-loops survived. Furthermore, we demonstrated that primer extension initiated at terminal ã Blackwell Science Ltd, The Plant Journal, (2000), 23, 1±11
G-overhangs and proceeded by nick translation and not by a strand displacement mechanism as expected for a primer annealed to an internal stretch of telomeric DNA in a T-loop (Figure 3d). It is interesting to note that only 15± 40% of the telomeres in human cells could be shown terminate in T-loops (Grif®th et al., 1999). While this number may re¯ect the technical dif®culties in preserving this structure during sample preparation, it could also indicate that long G-overhangs exist on only a subset of human telomeres. Current models for telomere function assume that G-overhangs are found on both chromosome ends and that their association with speci®c end binding proteins is critical in allowing cells to distinguish natural ends from double-strand breaks (Grif®th et al., 1999; Horvath et al., 1998; van Steensel et al., 1998). A different interpretation of the data is that G-overhangs are simply a by-product of the DNA replication mechanism that must be hidden to prevent chromosome instability or cell-cycle arrest. This idea is supported by a recent study showing that accumulation of single-stranded G-telomeric DNA triggers p53-dependent cell-cycle arrest in human cells (Saretzki et al., 1999). Our identi®cation of two distinct classes of telomere ends suggests that plant chromosome termini may be asymmetric, one end blunt and the other carrying a G-overhang in the con®guration predicted by semiconservative DNA replication. While direct evidence for plant chromosomes is lacking, two recent studies by Wright et al. (1997, 1999) provide intriguing data indicating that human chromosome termini may indeed be asymmetric. As with other eukaryotes (Dionne and Wellinger, 1996; Hemann and Greider, 1999; McElligott and Wellinger, 1997; Makarov et al., 1997; Wellinger et al., 1996), we found G-overhangs in cells from seeds and leaves which do not express telomerase or express it at low levels (Fitzgerald et al., 1996; Fitzgerald et al., 1999; Heller et al., 1996; Kilian et al., 1998; Riha et al., 1998). Although these ®ndings indicate that telomerase does not play a major role in maintaining G-overhangs in plants, it is possible that telomerase is involved in creating these structures. Telomerase expression is regulated in the plant cell cycle and is detectable only in S-phase (Tamura et al., 1999). Since the majority of cells in leaf and seed arrest in G0, telomerase activity in the previous cell cycle could account for the presence of G-overhangs in these tissues. One unexpected ®nding in our study was the observation that telomere architecture is not the same in all telomerase-negative tissues: the Cs signal was substantially reduced in leaves relative to seedlings and mature seeds. Whereas seedlings and seeds represent early stages of plant development, adult leaves do not develop any further and ultimately undergo senescence (Bleecker and Patterson, 1997). We extracted DNA from green, 1±2-
Karel Riha et al.
month-old leaves that did not exhibit any indication of senescence. However, leaf cells survive for several weeks without replicating their DNA and this could place G-overhangs at risk for nuclease attack. A single-strand DNA nuclease activity has recently been detected during leaf senescence (Perez-Amador et al., 2000). Thus, erosion of G-overhangs in a fraction of cells could be an early event in the catabolic processes. Intriguingly, formation of a new telomere binding complex was recently detected in Arabidopsis leaves during the transition from pre-senescence to senescence (Zentgraf et al., 2000). Therefore, while the overall length of telomeric tracts is not reduced in Silene and Arabidopsis leaves (Fitzgerald et al., 1999; Riha et al., 1998; Zentgraf et al., 2000), remodelling of telomere architecture could be involved in plant senescence. Experimental procedures Plant material and DNA isolation Silene latifolia (syn Melandrium album; white campion) seeds were obtained from a large greenhouse population cultured at the Institute of Biophysics (Brno, Czech Republic). We prepared young seedlings by germinating seeds in water for 72 h. Silene plants grown in a greenhouse were used as the source of adult rosette leaves. Three-week-old Arabidopsis thaliana (Columbia ecotype) plants were cultivated at 21°C in an environmental growth chamber under a 24 h photoperiod. High molecular weight nuclear DNA was isolated by density gradient ultracentrifugation as described by Jofuku and Goldberg (1988).
Primer extension/nick translation (PENT) reactions PENT reactions (Makarov et al., 1997) were performed in a total volume of 100 ml either in Taq polymerase buffer (20 mM Tris±HCl, pH 8.3, 50 mM KCl and 2 mM MgCl2) or in DNA polymerase I buffer (50 mM Tris±HCl, pH 7.2, 10 mM MgSO4 and 0.1 mM DTT), containing 0.1 mM of each dATP, dCTP and dTTP and 100 nM oligonucleotide. For reactions with Taq polymerase, DNA from Silene (10 mg) or Arabidopsis (1 mg) was added to Taq polymerase buffer, the mixture was covered with mineral oil and incubated at 55, 65 or 72°C for 1 h to anneal the oligonucleotide to G-overhangs (Table 1). After the annealing step, oligonucleotides were elongated with 40 U ml±1 Taq polymerase (Promega) for 1 h at the temperature indicated. Reactions carried out at 16 or 37°C were performed with DNA polymerase. The 1 h annealing step was followed by primer extension with 100 U ml±1 DNA polymerase I (Promega) for 4 h. Samples were ethanol-precipitated, and PENT products were separated by alkaline gel electrophoresis (see below). A time-course analysis of PENT products was conducted at 55°C with Taq polymerase in the presence of all four dNTPs. Reactions were stopped at the desired times by addition of EDTA to a ®nal concentration of 10 mM followed by ethanol precipitation. To remove single-strand G-overhangs, the DNA was incubated in a buffer composed of 10 mM Tris±Cl (pH 8), 1 mM EDTA, 10 mM MgCl2, 20 mM KCl and 10 mM 2-mercaptoethanol, either with or without 20 U of exonuclease I (USB) for 2 h at 37°C. The enzyme was inactivated by heating the samples to 65°C for 10 min, followed by phenol extraction and ethanol precipitation.
Alkaline gel electrophoresis and in-gel hybridization Alkaline gel electrophoresis was performed with 1% agarose gels, and PENT products were separated at 1.2 V cm±1 for 18 h. Two different methods were used to detect PENT products: hybridization of a telomeric probe following capillary transfer of the PENT products to a nylon membrane (MSI) (Makarov et al., 1997) or hybridization of the probe directly to PENT products in the alkaline gel. For the in-gel hybridization, a modi®ed version of the Dionne and Wellinger (1996) protocol was used. Gels were soaked for 30 min in neutralization buffer (0.5 M Tris±HCl, pH 8, 150 mM NaCl) equilibrated for 1 h in 2 3 SSC (300 mM NaCl, 30 mM sodium citrate, pH 7) and dried for 30 min on a gel drier (Bio-Rad, model 583) at room temperature. Gels were placed in a plastic box and pre-hybridized for 1 h in hybridization buffer (5 3 SSC, 5 3 Denhardt's reagent, 0.1 M Na2HPO4, 10 mM Na4P2O7, 1 g l±1 BSA) at 55°C, and hybridized with a (T3AG3)4 probe at 55°C overnight. The oligonucleotide probe was end-labelled with [a-32P]ATP and puri®ed using a QIAgen Nucleotide Removal kit. After hybridization, gels were washed twice for 15 min in 2 3 SSC, 0.1% SDS at 55°C and once for 45 min in 0.2 3 SSC, 0.1% SDS at 55°C. The reaction products were scanned using a PhosphorImager (Molecular Dynamics) and the data were analysed by ImageQuant software (Molecular Dynamics).
Acknowledgements We thank Jeff Kapler for critically reading the manuscript and for helpful discussions. This work was funded by grants from the National Science Foundation (MCB9982499) to D.E.S and T.D.M, the Advanced Texas Research Program to D.E.S. (999902-141) and T.D.M. (010366-141), and from the Grant Agency of the Czech Republic to B.V. (521/99/0696). K.R. is supported by a NATO-NSF postdoctoral fellowship.
References Bleecker, A.B. and Patterson, S.E. (1997) Last exit: senescence, abscission, and meristem arrest in Arabidopsis. Plant Cell, 9, 1169±1179. Cooper, J.P., Watanabe, Y. and Nurse, P. (1998) Fission yeast Taz1 protein is required for meiotic telomere clustering and recombination. Nature, 392, 828±831. Dionne, I. and Wellinger, R.J. (1996) Cell cycle-regulated generation of single-stranded G-rich DNA in the absence of telomerase. Proc. Natl Acad. Sci. USA, 93, 13902±13907. Dionne, I. and Wellinger, R.J. (1998) Processing of telomeric DNA ends requires the passage of a replication fork. Nucl. Acids Res. 26, 5365±5371. Donnelly, P.M., Bonetta, D., Tsukaya, H., Dengler, R.E. and Dengler, N.G. (1999) Cell cycling and cell enlargement in developing leaves of Arabidopsis. Dev. Biol. 215, 407±419. Fajkus, J., Fulneckova, J., Hulanova, M., Berkova, K., Riha, K. and Matyasek, R. (1998) Plant cells express telomerase activity upon transfer to callus culture without extensively changing telomere lengths. Mol. Gen. Genet. 260, 470±474. Fitzgerald, M.S., McKnight, T.D. and Shippen, D.E. (1996) Characterization and developmental patterns of telomerase expression in plants. Proc. Natl Acad. Sci. USA, 93, 14422± 14427. Fitzgerald, M.S., Riha, K., Gao, F., Ren, S., McKnight, T.D. and Shippen, D.E. (1999) Disruption of the telomerase catalytic subunit gene from Arabidopsis inactivates telomerase and ã Blackwell Science Ltd, The Plant Journal, (2000), 23, 1±11
Telomere structure in Silene and Arabidopsis
leads to a slow loss of telomeric DNA. Proc. Natl Acad. Sci. USA, 96, 14813±14818. Garvik, B., Carson, M. and Hartwell, L. (1995) Single-stranded DNA arising at telomeres in cdc13 mutants may constitute a speci®c signal for the RAD9 checkpoint. Mol. Cell Biol. 15, 6128± 6138. Gravel, S., Larrivee, M., Labrecque, P. and Wellinger, R.J. (1998) Yeast Ku as a regulator of chromosomal DNA end structure. Science, 280, 741±744. Grif®th, J.D., Comeau, L., Rosen®eld, S., Stansel, R.M., Bianchi, A., de Moss, H. and Lange, T. (1999) Mammalian telomeres end in a large duplex loop. Cell, 97, 503±514. Heller, K., Kilian, A., Piatyszek, M.A. and Kleinhofs, A. (1996) Telomerase activity in plant extracts. Mol. Gen. Genet. 252, 342±345. Hemann, M.T. and Greider, C.W. (1999) G-strand overhangs on telomeres in telomerase-de®cient mouse cells. Nucl. Acids Res. 27, 3964±3969. Henderson, E.R. and Blackburn, E.H. (1989) An overhanging 3¢ terminus is a conserved feature of telomeres. Mol. Cell Biol. 9, 345±348. Horvath, M.P., Schweiker, V.L., Bevilacqua, J.M., Ruggles, J.A. and Schultz, S.C. (1998) Crystal structure of the Oxytricha nova telomere end binding protein complexed with single strand DNA. Cell, 95, 963±974. Jofuku, K.D. and Goldberg, R.B. (1988) Analysis of plant genome structure. In: Plant Molecular Biology ± A Practical Approach (Shaw, C.H., ed.). Oxford: IRL Press, pp. 37±66. Kilian, A., Stiff, C. and Kleinhofs, A. (1995) Barley telomeres shorten during differentiation but grow in callus culture. Proc. Natl Acad. Sci. USA, 92, 9555±9559. Kilian, A., Heller, K. and Kleinhofs, A. (1998) Developmental patterns of telomerase activity in barley and maize. Plant Mol. Biol. 37, 621±628. Kirk, K.E., Harmon, B.P., Reichardt, I.K., Sedat, J.W. and Blackburn, E.H. (1997) Block in anaphase chromosome separation caused by a telomerase template mutation. Science, 275, 1478±1481. Klobutcher, L.A., Swanton, M.T., Donini, P. and Prescott, D.M. (1981) All gene-sized DNA molecules in four species of hypotrichs have the same terminal sequence and an unusual 3¢ terminus. Proc. Natl Acad. Sci. USA, 78, 3015±3019. Lingner, J., Cooper, J.P. and Cech, T.R. (1995) Telomerase and DNA end replication: no longer a lagging strand problem? Science, 269, 1533±1534. McClintock, B. (1939) The behavior in successive nuclear divisions of a chromosome broken at meiosis. Proc. Natl Acad. Sci. USA, 25, 405±416. McClintock, B. (1941) The stability of broken ends of chromosomes of Zea mays. Genetics, 26, 234±282. McElligott, R. and Wellinger, R.J. (1997) The terminal DNA structure of mammalian chromosomes. EMBO J. 16, 3705± 3714. Makarov, V.L., Hirose, Y. and Langmore, J.P. (1997) Long G tails at both ends of human chromosomes suggest a C strand degradation mechanism for telomere shortening. Cell, 88, 657±666. Murti, K.G. and Prescott, D.M. (1999) Telomeres of polytene chromosomes in a ciliated protozoan terminate in duplex DNA loops. Proc. Natl Acad. Sci. USA, 96, 14436±14439. Nugent, C.I., Hughes, T.R., Lue, N.F. and Lundblad, V. (1996)
ã Blackwell Science Ltd, The Plant Journal, (2000), 23, 1±11
Cdc13p: a single-strand telomeric DNA-binding protein with a dual role in yeast telomere maintenance. Science, 274, 249±252. Parenteau, J. and Wellinger, R.J. (1999) Accumulation of singlestranded DNA and destabilization of telomeric repeats in yeast mutant strains carrying a deletion of RAD27. Mol. Cell Biol. 19, 4143±4152. Perez-Amador, M.A., Abler, M.L., De Rocher, E.J., Thompson, D.M., van Hoof, A., LeBrasseur, N.D., Lers, A. and Green, P.J. (2000) Identi®cation of BFN1, a bifunctional nuclease induced during leaf and stem senescence in Arabidopsis. Plant Physiol. 122, 169±180. Richards, E.J. (1995) Plant telomeres. In: Telomeres (Blackburn, E.H. and Greider, C.W., eds). Cold Spring Harbor, New York: Cold Spring Harbor Laboratory Press, pp. 371±378. Richards, E.J., Goodman, H.M. and Ausubel, F.M. (1991) The centromere region of Arabidopsis thaliana chromosome 1 contains telomere-similar sequences. Nucl. Acids Res. 19, 3351±3357. Richards, E.J., Chao, S., Vongs, A. and Yang, J. (1992) Characterization of Arabidopsis thaliana telomeres isolated in yeast. Nucl. Acids Res. 20, 4039±4046. Riha, K., Fajkus, J., Siroky, J. and Vyskot, B. (1998) Developmental control of telomere lengths and telomerase activity in plants. Plant Cell, 10, 1691±1698. Saretzki, G., Sitte, N., Merkel, U., Wurm, R.E. and von Zglinicki, T. (1999) Telomere shortening triggers a p53-dependent cell cycle arrest via accumulation of G-rich single stranded DNA fragments. Oncogene, 18, 5148±5158. Shippen, D.E. and McKnight, T.D. (1998) Telomeres, telomerase and plant development. Trends Plant Sci. 3, 129±130. van Steensel, B., de Smogorzewska, A. and Lange, T. (1998) TRF2 protects human telomeres from end-to-end fusions. Cell, 92, 401±413. Tamura, K., Liu, H. and Takahashi, H. (1999) Auxin induction of cell cycle regulated activity of tobacco telomerase. J. Biol. Chem. 274, 20997±21002. Venkatesan, R.N. and Price, C. (1998) Telomerase expression in chickens: constitutive activity in somatic tissues and downregulation in culture. Proc. Natl Acad. Sci. USA, 95, 14763± 14768. Wellinger, R.J., Wolf, A.J. and Zakian, V.A. (1993) Saccharomyces telomeres acquire single-stranded TG1±3 tails late in S phase. Cell, 72, 51±60. Wellinger, R.J., Ethier, K., Labrecque, P. and Zakian, V.A. (1996) Evidence for a new step in telomere maintenance. Cell, 85, 423± 433. Wright, W.E., Tesmer, V.M., Huffman, K.E., Levene, S.D. and Shay, J.W. (1997) Normal human chromosomes have long Grich telomeric overhangs at one end. Genes Dev. 11, 2801±2809. Wright, W.E., Tesmer, V.M., Liao, M.L. and Shay, J.W. (1999) Normal human telomeres are not late replicating. Exp. Cell Res. 251, 492±499. Zakian, V.A. (1996) Structure, function, and replication of Saccharomyces cerevisiae telomeres. Annu. Rev. Genet. 30, 141±172. Zentgraf, U., Hinderhofer, K. and Kolb, D. (2000) Speci®c association of a small protein with the telomeric DNA-protein complex during the onset of leaf senescence in Arabidopsis thaliana. Plant Mol. Biol. 42, 429±438.
Karel Riha et al.
Author Query Form Journal: Plant Journal Article: tpj831 Dear Author,
During the preparation of your manuscript for publication, the questions listed below have arisen. The numbers pertain to the numbers in the margin of the proof. Please attend to these matters and return the form with this proof. Many thanks for your assistance.
Query Refs. Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11 Q12 Q13 Q14
Please supply keywords Does the following text make sense: predicated upon? Does the following text make sense: processivity? Does the following text make sense: formally? Is the text OK: PhosphorImager If data are not shown, what do Figures 4 and 5 show? Does the following text make sense: processivity? Is the text OK: QIAgen Donnelly etal. 1999 has not been found in the text Is the text OK: Zea mays Richards etal. 1995 has not been found in the text Virtual Genome Center ref has been deleted and web address given in table instead. OK? Please check that there are no errors in the address Please de®ne: TAS Is the text OK: Genebank
ã Blackwell Science Ltd, The Plant Journal, (2000), 23, 1±11
Telomere structure in Silene and Arabidopsis
ã Blackwell Science Ltd, The Plant Journal, (2000), 23, 1±11