The Telomere Function And Synthesis Essay

00:00:03.20 Hello, my name's Elizabeth Blackburn. I'm in the
00:00:06.25 Department of Biochemistry and Biophysics at the
00:00:09.19 University of California, San Francisco. And in this set of
00:00:13.19 lectures, I'm going to talk about telomeres and
00:00:16.08 telomerase. And I'll get to their implications for human
00:00:20.19 health and disease. The first part of this series of three
00:00:26.07 lectures is going to introduce you to the roles of telomeres
00:00:30.18 and telomerase. So, let's begin by focusing in on what
00:00:36.27 goes on at the very heart of a cell. Now if you looked at a
00:00:42.10 cell which is just about to divide, you looked into the
00:00:46.06 microscope and you stained the chromosomes, this is
00:00:48.14 what you would see. Those blue, double sausage-like
00:00:51.25 objects that you can see all over this microscope slide are
00:00:57.17 the human chromosomes, and the DNA has just
00:00:59.24 duplicated, which is why they look double. Now if you
00:01:03.11 look closely, you can see red spots, two red spots at the
00:01:06.15 ends of every double chromosome pair. And these red
00:01:11.15 spots are a molecular probe that's lighting up the telomeric
00:01:15.05 DNA that's found in common at all of the chromosome
00:01:18.25 ends. And I'm going to tell you a lot about that telomeric
00:01:23.04 DNA. So why are telomeres important? Their role is to cap
00:01:30.14 off the ends of chromosomes. So that's a simple concept,
00:01:35.26 but we can dive down into it further and think a little bit
00:01:38.17 more about what that actually means. So when we think
00:01:42.07 of the end of the DNA, there can be two kinds: There
00:01:45.09 can be the natural end of the DNA, such as the ones we
00:01:48.07 see here, the ends of the chromosomal DNA; and there
00:01:52.12 can also be DNA breaks. Now, the job of a cell is to seal
00:02:01.05 up any DNA breaks that happen by accident. And so a
00:02:05.03 major part of this capping function, which is one of the
00:02:09.04 aspects of telomere functions, is to prevent the telomeres
00:02:13.23 from undergoing those very DNA transactions of very
00:02:17.11 DNA reactions that are undergone by a broken DNA end.
00:02:24.00 So if you have a break in the DNA, it can be sutured
00:02:27.14 together by, for example, recombination or just simply the
00:02:31.21 two broken ends can be ligated right back together by
00:02:34.25 end-to-end fusions. Also, such DNA breaks are subject to
00:02:40.15 degradation. Now the telomere protects, it caps, the end
00:02:45.05 of the chromosome and protects against all of these kinds
00:02:48.13 of things that would normally happen to a broken DNA
00:02:52.12 end. How does it do that? The DNA sequences that you
00:02:58.08 find at the ends of chromosomes, repeated over and
00:03:01.15 over, are fairly similar in nature to each other. They're
00:03:06.11 relatively simple sequences, and they're too simple to
00:03:09.24 code for any proteins. They're not genes in the sense of
00:03:15.09 coding for any proteins or RNAs. In humans for example,
00:03:21.20 this repeated sequence is found up to a few thousand
00:03:24.14 times, tandemly repeated over and over at the ends of the
00:03:29.08 chromosomes. Another feature of the chromosomal DNAs
00:03:34.06 is that, of course, unlike most of the DNA of a
00:03:38.00 chromosome, which is duplex DNA, double-stranded
00:03:40.11 DNA, the very end is single-stranded. And in fact the
00:03:44.26 DNA strand is oriented 5' to 3', going toward the end of
00:03:50.03 the chromosome. And that turns out to be important. So,
00:03:58.12 we have now the telomere structure to begin with: We
00:04:01.27 have again, if we blow up the end of a chromosome, we
00:04:05.23 have these highly repeated sequences made up of a G-
00:04:10.12 rich sequence that's the repeat unit repeated over and
00:04:13.13 over again. As I said, it doesn't encode any protein
00:04:17.13 sequences, but each of these repeated sequences is like
00:04:23.05 a little attractive magnet for specific proteins that bind
00:04:27.25 sequence specifically to the telomeric DNA. They bind to
00:04:31.22 the telomeric repeats for the double-stranded portion, and
00:04:35.27 some of them bind the single-stranded portion, and that's
00:04:39.06 actually the G-rich strand that's forming the overhanging
00:04:45.11 end here. These together make some form of higher order
00:04:53.02 architecture we don't understand. We understand a lot
00:04:56.13 about the protein-DNA interactions, some of the molecular
00:04:59.11 details of that. Some of the details of the protein-protein
00:05:02.19 interactions in this complex, but we don't really
00:05:05.07 understand the higher order structure, so that's still a
00:05:07.22 challenge in the field. So I've just shown you a functional
00:05:14.04 telomere, but if telomeres cease to function... and we use
00:05:18.10 the term "telomere dysfunction" to just describe that
00:05:21.21 general state of the telomere that is not carrying out those
00:05:25.04 capping functions and other functions that I will get to...
00:05:30.07 there are a couple of different ways this can happen. The
00:05:33.09 first is if the tract of telomeric repeat is simply too short,
00:05:38.27 there's just not enough of the length of the repeats to form
00:05:44.27 a nice, long array that can form this higher order structure
00:05:48.27 that's necessary. This kind of dysfunction caused by the
00:05:54.27 shortening of the telomere, that can happen naturally, and
00:05:59.28 it does, and we'll get back to that in a moment, because
00:06:02.24 this is going to be an important part of these lectures, and
00:06:05.28 in fact it'll be really the focus of the third lecture in this
00:06:10.00 series. The other way that telomeres can become
00:06:15.07 dysfunctional: If for one reason or another, experimentally
00:06:19.12 induced, most commonly, one of these proteins cannot
00:06:23.27 bind correctly to the telomeric DNA; if its binding is
00:06:27.19 disrupted through some molecular intervention or other...
00:06:32.22 in both cases, cells sense and respond to this state of
00:06:40.03 telomere dysfunction. Now indeed, cells have a lot of very
00:06:46.22 strict regulatory reactions to the lack of proper telomeric
00:06:53.03 DNA. And the consequences for the cell is that, usually
00:06:59.06 this state of the cell's telomere dysfunction, through one
00:07:05.02 reason or another, will mean that the cell will cease to
00:07:10.20 divide. So this limits cell renewal capability if this happens
00:07:15.16 to one or more of its telomeres in the cell. If by chance the
00:07:20.20 cell does continue to multiply, now those telomeres
00:07:26.21 become subject to the very kinds of fusions (the DNA-
00:07:31.28 joining events that I told you telomeres shouldn't allow to
00:07:35.11 happen), and that can lead to genomic instability,
00:07:39.21 because the end-to-end joining of telomeres to
00:07:42.17 themselves (other telomeres, that is) or to broken DNAs,
00:07:47.00 that can cause the chromosomes, which fuse to each
00:07:52.07 other, to tear themselves apart as the cells divide, leading
00:07:56.09 to genomic instability. So clearly telomere function is very
00:08:02.20 important for cells. And in fact, one of the consequences
00:08:07.10 of genomic instability in human cells is that the cells can
00:08:11.13 become cancerous. I'm just going to show you a picture.
00:08:17.26 If you look under a microscope of some cells in which
00:08:22.20 we've disrupted one of the telomeric proteins, what you
00:08:25.22 can see is... remember I told you the blue, double things
00:08:30.01 are the chromosomal DNAs... and look, here's a
00:08:32.29 chromosome here in which there's been a telomere
00:08:35.19 fusion. So here's the two telomeres at the end, here's the
00:08:38.25 other two all the way here, but there are fused telomeres
00:08:43.27 here. So this is now a chromosome that has two
00:08:47.15 centromeres, it's got a centromere here, it's a got a
00:08:50.02 centromere here, and if those two centromeres try to pull
00:08:54.05 apart in a cell that is dividing, the chromosome will get
00:08:59.15 ripped apart. Here's another example of such an end-to-
00:09:04.22 end fused chromosome. This kind of change can happen
00:09:09.05 if you disrupt the telomeric integrity by, for example,
00:09:13.05 disrupting the binding of proteins. The other kind of
00:09:18.23 function, it's related, but we can distinguish it, the other
00:09:23.02 kind of function of telomeres is that they have to allow for
00:09:27.13 the complete replication of the telomeric DNA. So what's
00:09:33.07 the issue here? Well, the mechanism of DNA replication,
00:09:40.12 the machinery of DNA replication, has a particular quirk to
00:09:44.23 it. It's very good at faithfully copying almost all the way
00:09:49.19 along the length of the chromosomal DNA (or any linear
00:09:53.17 DNA), but the make-up of the DNA replication machinery
00:09:59.00 is such that is cannot copy the very, very end of the linear
00:10:04.28 DNA, such as a eukaryotic chromosomal DNA. Now the
00:10:10.11 predicted and observed consequence of that inability is
00:10:14.22 that, each time the DNA replicates, which is has to do as
00:10:20.01 the cell divides, and then the cell divides, the daughter
00:10:24.00 DNAs are predicted to become shorter and shorter and
00:10:27.13 shorter. And this is just a simple consequence of the
00:10:32.15 nature of the DNA replication machinery that is otherwise
00:10:37.00 so good at replicating all the way along all the length of
00:10:41.07 the chromosomal DNA. But the very ends cannot be
00:10:45.09 completely replicated without some form of compensatory
00:10:49.20 mechanism. So what's the consequence of this loss?
00:10:53.20 Well, obviously, something has to compensate in the long
00:10:56.06 run, otherwise, we wouldn't be here. But even on a
00:10:59.29 shorter timeframe, as cells divide and divide, the DNA
00:11:03.08 gets shorter and shorter, one might predict that there
00:11:05.24 would come a point when there wouldn't be enough of
00:11:10.10 something or other at the end that then the chromosomes
00:11:14.19 would no longer be able to support cell division, and the
00:11:18.00 cells might eventually undergo what's called senescence.
00:11:22.00 And indeed James Watson in 1972, just considering the
00:11:26.02 mechanism of DNA replication, proposed this constant
00:11:30.15 shortening problem, and Olovnikov around the same time
00:11:36.20 proposed that in fact perhaps such loss of terminal DNA,
00:11:41.22 without knowing at that stage what the molecular nature
00:11:44.07 of the terminal DNA was, Olovnikov proposed that
00:11:47.12 perhaps such gradual loss could be something that
00:11:50.20 underlies the eventual senescence of cells that is seen
00:11:54.14 sometimes when, for example, human cells are grown in
00:11:57.08 culture. This was a prescient idea because, in fact, this
00:12:02.04 indeed has been found to be one of the causes of why
00:12:07.17 human cells cannot replicate themselves, the cells cannot
00:12:12.18 proliferate, indefinitely in culture. So, shortening of
00:12:19.26 telomeric DNA is something that will be problematic for
00:12:25.12 cells. How is this problem solved? Well, it's solved by an
00:12:33.01 enzyme called telomerase. So I'd like to introduce you to
00:12:37.12 telomerase now, and how it was found. Telomerase was
00:12:46.06 sought because there was a set of accumulating
00:12:49.24 observations on telomeric DNA in cells, the telomeric
00:12:54.24 DNA as it was in cells in vivo, that couldn't be readily
00:12:58.22 explained by what was currently known about DNA
00:13:02.12 replication or DNA recombination or other kinds of DNA
00:13:06.22 reactions at the time, which was the late 1970s, early
00:13:10.21 1980s. And let me give you some examples of such
00:13:15.14 puzzling observations. Well, the first one was that in the
00:13:21.26 ciliated protozoan Tetrahymena, which has a lot of very
00:13:24.26 small mini-chromosomes and is therefore amenable to
00:13:28.02 molecular analyses of its telomeric DNA by direct
00:13:31.09 methods, the telomeric repeat sequences (remember I told
00:13:36.05 you about the repeated sequences repeated over and
00:13:38.16 over at the ends of chromosomes)... the repeat sequence
00:13:41.28 in this organism, G4T2 repeats, were heterogeneous in
00:13:48.05 their number, in different molecules in a population of
00:13:53.01 otherwise homogeneous cells. And heterogeneous in this
00:13:57.20 setting here is meaning, these were different in number,
00:14:04.04 so some copies of the mini-chromosome would have
00:14:07.07 twenty repeats on the end, some would have 50-some,
00:14:09.09 49-some, 82-some, 53... they all had different numbers of
00:14:13.16 repeats in this sort of more-or-less normal distribution. So
00:14:18.18 that was very surprising, because if you look across at the
00:14:21.15 internal region of a DNA, such as a chromosomal DNA, if
00:14:25.20 you look at one cell compared with another in a
00:14:29.05 population of cells from, say, one organism, it should
00:14:32.08 always be an identical sequence. So here were different
00:14:35.21 numbers of repeats at the ends of different molecules in a
00:14:40.15 population of cells that should have otherwise been
00:14:43.07 homogeneous, and were homogeneous in their internal
00:14:46.09 regions of the chromosome. Now, a second kind of
00:14:50.17 observation was a somewhat more complicated one, and
00:14:55.14 it means that I have to just take a moment to tell you
00:14:59.12 about the lifecycle of a particular group of ciliated
00:15:02.23 protozoans, which the species Tetrahymena in which
00:15:07.08 these G4T2 repeat tracts were found to be the telomeric
00:15:10.26 tracts. The Tetrahymena cells go through a lifecycle
00:15:15.24 stage where they have a somatic nucleus which
00:15:20.09 undergoes developmentally controlled fragmentation. And
00:15:24.16 fascinatingly, telomeric DNA sequences were added
00:15:27.29 directly to those freshly formed DNA ends, making from
00:15:33.27 longer chromosomes a series of shorter mini-chromosomes.
00:15:38.25 How did that telomeric DNA get added to
00:15:41.15 the ends? It wasn't clear. The third observation came
00:15:48.11 from observing the cells of an organism which causes
00:15:52.09 sleeping sickness, a single-celled parasitic organism
00:15:55.24 called a trypanosome. And these were being propagated
00:15:59.07 in the laboratory setting, and what was found was that the
00:16:04.05 telomeric DNA restriction fragments, the end fragments of
00:16:08.05 the chromosomes in these organisms, were gradually
00:16:11.09 getting longer and longer and longer. And this didn't look
00:16:14.10 like, say, recombination, and certainly was not expected
00:16:18.01 for normal, as one thought about it, DNA replication. The
00:16:23.27 fourth observation was again something that came out an
00:16:29.00 experiment that I'll have to explain. Now one could put
00:16:33.07 circular plasmids into yeast cells, and those plasmids are
00:16:37.01 linearized, essentially they're unstable, they get gobbled
00:16:40.16 up or, rarely, will recombine into the chromosomes and
00:16:44.12 thus be preserved. But what was found was that if one
00:16:48.10 simply grafted onto the ends of such a linearized and
00:16:52.24 therefore normally very unstable yeast plasmid, if one
00:16:56.24 grafted onto its ends Tetrahymena telomeric DNA
00:17:01.15 fragments (the telomeres of Tetrahymena mini-
00:17:03.24 chromosomes) and introduced those into cells... so the
00:17:08.01 grafting was done in vitro with enzymes and the purified
00:17:11.09 DNA molecules, and then the resulting hybrid of the yeast
00:17:17.05 linearized plasmid and the telomeres added onto its end,
00:17:21.29 that was introduced into the yeast cells, those were
00:17:25.00 maintained in yeast cells as linear mini-chromosomes, and
00:17:28.12 yeast telomeric DNA repeat were grafted on somehow
00:17:34.11 inside the yeast cells, to the ends of the Tetrahymena
00:17:39.02 telomeric repeats. And that didn't look like a reaction one
00:17:43.11 would expect from standard known models of DNA
00:17:46.26 replication or recombination. All of these suggested that
00:17:52.03 perhaps there was some capability of cells to add
00:17:57.16 telomeres. And that idea, in a conceptual way, was given
00:18:04.22 some force by an observation made by the noted
00:18:09.25 geneticist, Barbara McClintock, who worked with maize
00:18:13.06 (corn), and she noted that a particular maize mutant stock
00:18:17.28 lost the capacity to do something that is normally found in
00:18:23.24 normal, wild-type maize. In such wild-type maize, if a
00:18:28.28 chromosome is broken by, for example, exo-radiation or
00:18:32.15 some mechanical rupture at a particular stage in
00:18:37.07 development, then that broken end can be, as she
00:18:40.18 described it, "healed." It becomes a normal, stable
00:18:43.29 telomere. Nobody knew the molecular basis of that. And
00:18:47.29 she found a mutant that had lost that capacity. And when
00:18:51.04 you see a mutant where something is changed, you have
00:18:53.17 a feeling that that's reflective of a cellular process that
00:18:57.21 can take place, and that it's not just by chance that this
00:19:01.08 healing event was taking place. All of which, then,
00:19:07.23 focused in on the question of: Was there a new enzyme
00:19:10.02 that worked in cells that could extend telomeric DNA? So
00:19:14.17 we sought such an enzyme in the early to mid-1980s. And
00:19:19.24 we used for that purpose the single-celled, ciliated
00:19:24.18 protozoan Tetrahymena thermophila, shown in this
00:19:27.18 scanning electron micrograph here. And you can see the
00:19:31.02 cilia are on the cell surface. This experimental system was
00:19:34.26 chosen because the organism contains large numbers of
00:19:40.00 very short mini-chromosome, therefore, large numbers of
00:19:44.15 telomeres, and therefore, one would reason, perhaps if
00:19:49.03 there were such an enzyme that existed that could add
00:19:53.03 telomeric DNA to the end of chromosomes or to
00:19:56.25 preexisting telomeres, then this would be a potentially
00:20:00.05 good source for it because there's a lot of telomeres in
00:20:03.28 this organism. And indeed that proved to be the case.
00:20:08.18 And Carol Greider, my then-graduate student, joined the
00:20:12.11 lab in 1984, and here's a picture of Carol, freshly from
00:20:16.10 visiting Southern California, and looking at an
00:20:20.29 autoradiogram shown with this x-ray film here in the lab.
00:20:26.17 And we together found this enzyme telomerase. Now let
00:20:32.15 me show you briefly what we did. We took a mimic of that
00:20:37.18 overhanging G-rich strand DNA that's normally found at
00:20:41.11 the ends of chromosomes, in the form of an oligonucleotide,
00:20:45.18 and here it is shown here. And I've just
00:20:47.13 colored the bases different colors so that we can see
00:20:50.22 them easily. And here's its 3' hydroxyl end. We mixed it
00:20:55.28 with an extract of Tetrahymena cells, and this worked
00:20:59.13 particularly well at the stage in development that I
00:21:03.22 mentioned to you earlier... that stage at which telomeric
00:21:07.20 DNA is added to freshly broken ends of chromosomes.
00:21:12.09 We reasoned that that would be a particularly good time
00:21:15.29 to find such an activity, thinking it might be likely to be
00:21:20.16 induced or present in perhaps larger-than-normal
00:21:23.25 amounts, because this would be a time when there'd be a
00:21:27.02 demand for such an enzyme. All of this was hypothetical
00:21:30.00 at the time, but by putting in an appropriate mix of things
00:21:33.17 that are often added to make polymerases happy (such
00:21:39.02 as magnesium ion), and just by using two nucleoside
00:21:43.17 triphosphate precursors, dGTP and TTP, we were able to
00:21:48.21 find that the telomeric DNA indeed was added to the end
00:21:53.02 of such an oligonucleotide. In fact, in the test tube, a
00:21:56.09 large number of repeats could be added. So we get a lot
00:22:00.22 of repeats added, through eventually something limits the
00:22:06.18 addition. So this is what telomerase did. Now how is it
00:22:14.16 doing that? It was adding a given sequence. The first
00:22:18.09 clue came from the observation... and here I'm just
00:22:22.01 showing you a nice picture of the products run on a DNA
00:22:27.10 sequencing gel, just to give you feel for what it looks like.
00:22:30.14 We'll just look at one of these groups, so here are
00:22:33.13 different time points, increasing number of minutes after
00:22:37.09 incubating this input primer with dGTP and TTP. The
00:22:41.06 dGTP was radiolabeled, so we're just looking at an
00:22:43.26 autoradiogram. Every time we see a labeled band, this is
00:22:47.22 a product that's been added, so the input would run here.
00:22:50.20 If you added one, two, three, four nucleotides, you could
00:22:53.11 see you'd get longer and longer bands. And what you can
00:22:56.20 see is that there is kind of a striped pattern. Every six
00:23:01.05 nucleotides, there was a pause in the addition, and so
00:23:04.16 you could see a pattern of six nucleotides being added,
00:23:09.03 and you could see more and more repeats, as shown by
00:23:11.20 the bands getting higher and higher and higher, were
00:23:14.18 being added with time. Now, certain clues came. Not any
00:23:21.27 nucleotide could work. So, two examples of the telomeric
00:23:27.01 DNA, where we've got a permutation ending with four Gs
00:23:30.12 here, or a permutation ending with two Ts here. These
00:23:34.26 were each competent for addition by this activity.
00:23:41.04 However, the complementary strand, and again, I've just
00:23:44.07 colored the bases distinctively and given you two
00:23:47.01 examples... the complementary strand oligonucleotides
00:23:50.25 were not competent for such addition. Another interesting
00:23:59.04 clue came when we looked in more detail at different
00:24:02.16 oligonucleotides that could act as primers. Now, we had
00:24:10.01 found, as I said to you, that when you put Tetrahymena
00:24:13.20 telomeric repeats into yeast cells, live cells, then yeast
00:24:18.12 repeats were added somehow, it was unknown then, to
00:24:22.15 the Tetrahymena telomeres. And the yeast repeats have
00:24:27.06 a different sequence, and I've shown you a little segment
00:24:29.06 of it. It's T, G, sometimes one G, sometimes two,
00:24:34.13 sometimes three Gs. So we wondered if the converse
00:24:38.23 would be the case; if we added a yeast primer to a
00:24:43.14 Tetrahymena extract, now in vitro, would we see
00:24:48.04 Tetrahymena sequences added to the yeast telomeric
00:24:52.19 oligonucleotide. And indeed, we did see such addition.
00:24:58.26 And again, large numbers of repeats could be added to
00:25:02.14 such a primer quite efficiently. Now we also noticed
00:25:08.19 another interesting thing about the reaction. When we
00:25:12.21 had a primer that ended with four Gs, what we found was
00:25:17.15 that first two Ts were added, and then four Gs, two Ts,
00:25:20.15 and so on. When we had a primer that ended with three
00:25:23.06 Gs, first a G was added, and then the two Ts, in other
00:25:27.19 words completing this run of four Gs. And I haven't shown
00:25:30.26 it, but if you had a primer that ended in two Gs, then GG,
00:25:35.07 and then TT. So this suggested that perhaps something
00:25:39.22 was aligning this very first portion, the portion of the
00:25:46.29 reaction that's taking place, adding to the very 3' end of
00:25:50.09 the primer, that perhaps something was aligning this, in
00:25:53.29 the enzyme somehow. Because in other words, if you
00:25:58.10 lined it up like this, then everything was lined up. What
00:26:01.08 could be doing this? So was there something that aligned
00:26:04.28 the product-forming part of this enzyme? Remember this
00:26:10.17 was very mysterious at the time. So in fact we looked and
00:26:15.07 found that, indeed, as I said, you had four Gs, and you
00:26:19.25 added two Ts. If you had three Gs, one G was added.
00:26:23.08 Two Gs, and you'd have two Gs added, and then the two
00:26:25.14 Ts. And if you had one G, then it would be three Gs, and
00:26:29.08 then the two Ts. So in fact, there was something that was
00:26:31.27 aligning how the next repeats were added, dependent on
00:26:36.10 the 3' end of the primer. The result of all of this kind of
00:26:44.02 analysis led us to the idea that there was in fact a
00:26:47.15 template within the telomerase, and to our surprise, this
00:26:52.07 template turned out to be made of RNA, an RNA that is
00:26:56.25 actually built in to the telomerase complex. And this
00:27:01.04 template is a short portion of this RNA, which otherwise
00:27:05.26 has a lot of other structure which is built into the enzyme
00:27:09.20 particle, which contains also a protein, which is called
00:27:15.18 TERT. So what telomerase does is it takes that single-
00:27:20.18 stranded G-rich strand, it aligns the 3' end nucleotides by
00:27:27.07 Watson-Crick base pairing onto the template sequence,
00:27:31.29 so here's the example where we have two Gs at the end
00:27:35.05 of the primer. It aligns it on this part of the sequence, and
00:27:39.20 then it polymerizes, one at a time, each of these
00:27:44.29 nucleotides onto the DNA end, extending the DNA end
00:27:53.02 by copying this template. And so now, you end up with
00:27:58.23 longer DNA. So telomerase is a unique polymerase, it's a
00:28:06.22 reverse transcriptase by the definition, that is, copies RNA
00:28:12.18 into DNA. It's unique because the RNA component is
00:28:20.07 actually intrinsically built into the telomerase ribonucleoprotein
00:28:24.13 particle. It has to be built in for that
00:28:29.13 templating to take place. And indeed the enzyme is truly a
00:28:34.07 ribonucleoprotein enzyme, we believe. And unlike, for
00:28:39.26 example, the reverse transcriptases that copy, say, the
00:28:43.03 HIV viral genome, which is a genome thousands of
00:28:47.27 nucleotides long, a very complicated sequence which
00:28:51.14 encodes proteins, this enzyme copies very short
00:28:56.25 sequences over and over again. And it does it by this
00:29:00.26 process of aligning the DNA end on the template, and so
00:29:08.19 just to complete the thought here, here's the template. If
00:29:11.28 we polymerize all these together onto the DNA end, now
00:29:15.14 you'll have a new DNA end that ends with TTG, and that
00:29:19.26 will realign in the next round back in this AAC, for another
00:29:24.12 round of synthesis, and this can go on in the test tube for
00:29:28.07 many repeats. I've told you about some of the enzymatic
00:29:32.13 properties of telomerase, but what good is telomerase for
00:29:37.03 cells? Well, the answer came by manipulating telomerase
00:29:42.25 in Tetrahymena, the organism in which it was first
00:29:46.14 discovered. So if you remember, telomerase is adding
00:29:52.25 telomeric DNA to the ends of chromosomes, and so the
00:29:59.19 question was: What happened if it couldn't do that? So
00:30:03.17 we looked in Tetrahymena, which, as I said, was a good
00:30:07.03 source of telomerase, and another feature of these
00:30:10.23 organisms that I didn't tell you was that, if you grow them
00:30:14.00 in culture in the laboratory, they're effectively immortal, so
00:30:17.15 long as you keep them fed and under good conditions,
00:30:20.12 they can just propagate and propagate and propagate,
00:30:22.27 seemingly forever. And they have plenty of telomerase.
00:30:29.19 We manipulated the telomerase RNA. Now, we made
00:30:33.14 some changes in the RNA, and I won't go into the details
00:30:36.20 of it, but the effect of such small changes in the RNA I've
00:30:42.17 shown diagrammatically here. The telomeric DNA repeat
00:30:46.25 sequences, which as I said consist of multiple repeats at
00:30:51.17 the ends of chromosomes normally, as the cells went
00:30:56.17 through cell divisions, the telomeres progressively got
00:30:59.11 shorter and shorter, and after about 20 or 25 cell
00:31:02.20 divisions, the cells ceased to divide. So in other words,
00:31:07.07 when we inactivated telomerase, over the succeeding
00:31:11.16 cell divisions the telomeres progressively shortened, so in
00:31:15.18 effect when the cells ceased dividing, they'd become
00:31:19.07 mortal. From being immortal, they'd become mortal. And all
00:31:23.22 we had done was to inactivate telomerase by this small
00:31:28.23 change, so like a little stiletto stuck at the heart of
00:31:31.25 telomerase, we killed the enzyme very surgically, and we
00:31:35.08 were able to make immortal cells become mortal. So, the
00:31:40.17 conclusion is that the telomerase maintains the ends of
00:31:43.24 chromosomes. Telomeres are replenished by telomerase
00:31:47.07 as they keep dividing. And in fact, that continuing
00:31:51.22 replenishment, even in the face of the continuing
00:31:55.01 shortening processes that take place, can compensate
00:31:58.20 for those shortening processes and allow the cells to
00:32:01.15 keep on dividing. So that's what telomerase does for
00:32:05.28 cells. And that's the important message: Telomeres are
00:32:10.18 replenished by telomerase. Now, let's go into a more
00:32:16.03 detailed experiment in yeast cells. If we look at a culture
00:32:20.07 of yeast cells and we plate them out on a plate in a
00:32:24.03 laboratory... all those little white spots are cells that have
00:32:28.04 grown up from individual cells distributed on the plate,
00:32:30.24 they've been distributed on kind of a V-shape pattern
00:32:32.26 here, a triangular pattern. So, the cells are growing well
00:32:37.11 normally. Now if we delete one of the genes for
00:32:40.16 telomerase, either the RNA structural gene or the protein
00:32:43.26 structural gene for the core of the enzyme, now just as I
00:32:48.20 described for you in Tetrahymena, the cells divide and
00:32:54.16 divide, and the telomeres progressively get shorter and
00:32:57.03 shorter, until they get to a point where most of the cells
00:33:00.03 completely cease to divide. And so now they're no longer
00:33:04.27 capable of producing a rich growth of colonies on such a
00:33:09.29 plate. See, there's very little growth here. And we call that
00:33:13.27 senescence. So taking away telomerase is a bit of a
00:33:17.23 delay as the cells go through about 50 or so cell
00:33:20.06 generations, 50 to 80, and then the cells' telomeres get
00:33:24.05 too short, and they undergo senescence. We've
00:33:27.12 scrutinized more closely what happens to cells at this
00:33:31.09 point. Interestingly, a few cells do survive, and in fact,
00:33:42.09 those cells eventually become quite capable of producing
00:33:46.26 good growth on a plate, but they're different. These cells
00:33:51.05 now have very heterogeneous and long telomeres. I'll
00:33:55.01 show you these in a more graphical form in a moment. A
00:34:01.17 gene required for this to happen, for such "survivors," as
00:34:06.03 we call them, to appear, is the gene RAD52. For at least
00:34:13.05 one of these pathways, the RAD50 gene is also required.
00:34:17.11 What are RAD52 and RAD50? These are in the
00:34:22.07 homologous recombination pathway. They're needed for
00:34:27.02 recombination and therefore, to get survivors required
00:34:30.28 recombination. Similar phenomena have been seen in
00:34:35.16 mammalian cells. They've been less well characterized
00:34:38.16 genetically, but similar survivor cells have been seen, and
00:34:42.23 they're called ALT cells in mammals, for "alternative
00:34:47.18 lengthening of telomeres"... ALT cells. So, without
00:34:54.10 telomerase, most cells go through senescence, but rare
00:34:57.28 cells can survive. So, that is another way that
00:35:04.21 chromosomes can survive, but interestingly enough, ALT
00:35:10.04 is not normally seen in most normal cells. Now we
00:35:16.17 wondered why, because on the face of it, it looks as if
00:35:20.05 these cells are growing perfectly well. Although if you look
00:35:24.06 more closely, you will see that some of the cells are not
00:35:27.04 living well, but the great majority can survive with these
00:35:30.14 very long telomeres, that they can keep recombining, and
00:35:34.03 keep the population up that way. We asked what's
00:35:38.01 different. So first of all, we quantified very carefully the cell
00:35:43.19 growth, so this is just one way of showing it. The colony-
00:35:46.26 forming units, which is a measure of viability, as the cells
00:35:51.21 went through the progression of shortening telomeres,
00:35:54.17 getting to this point when they get to senescence, and
00:35:56.14 then those rare survivors now overgrew, and now you
00:36:00.29 can see them taking over and becoming the population.
00:36:03.23 So they're growing again quite well. And we looked at the
00:36:07.23 gene expression profile in such cells. First of all we looked
00:36:11.19 at the telomeres, just to make sure that the telomeres
00:36:14.20 were doing what I showed you diagrammatically before,
00:36:17.28 yes they are, so here's the telomeres. And this is what's
00:36:20.20 called a Southern blot, where we're probing with a
00:36:23.00 telomeric DNA sequence. And the most important are
00:36:26.13 these bands down here, that's the easiest to look at
00:36:28.21 initially, because this is a lot of the telomeres. You can
00:36:31.19 see they're of a certain length, and as they get shorter
00:36:34.11 they run faster and faster in the gel, so you can see
00:36:37.07 they're getting shorter and shorter with these successive
00:36:39.21 days of passaging of cell divisions without any telomerase
00:36:44.13 present. Now, there are very few cells here actually, but
00:36:49.09 we loaded enough DNA so you can see what little
00:36:52.11 telomeric DNA there is. And then the cells start growing
00:36:55.00 well again, and what you see is now the telomeres, the
00:36:57.29 pattern has changed, they're very heterogeneous, and
00:37:00.19 there's much longer telomeres, you can see they're longer
00:37:03.18 than these ones here. So that's the telomeric profile. Now,
00:37:13.20 let's look at the gene expression profile. This is what's
00:37:18.01 called a microarray experiment, and in this experiment,
00:37:22.27 one compares the pattern of gene expression by looking
00:37:26.07 at the messenger RNA levels for all of the genes in the
00:37:30.05 genome. And one asks, does a particular gene,
00:37:34.11 compared with a reference, which is right at the
00:37:36.17 beginning, when everything is growing well, does a
00:37:38.26 particular gene's level of expression become higher or
00:37:42.09 lower. If it becomes lower, it gets greener and greener.
00:37:46.12 And so each of these going across is a time point, each
00:37:57.23 line in the column represent a single gene, and each of
00:38:01.05 these horizontal areas represents the day at which the
00:38:08.13 RNA was taken out of the cells and analyzed in this way.
00:38:11.26 And so what we find, and it's all compressed and you just
00:38:14.23 need to look at the pattern, was that about 650 genes
00:38:19.07 changed their expression, and they got either lower levels
00:38:23.22 of expression (and we'll talk about this in more detail, and
00:38:27.20 that's shown in the green) or higher levels of expression
00:38:31.09 (and that shows up as red). Now right away what you can
00:38:35.15 see is that the maximum changes occurred right when the
00:38:38.12 cells were approaching and into senescence. Six-
00:38:43.23 hundred and fifty genes is something like 10% of the
00:38:48.09 genes in the genome, which is about 6000 to 7000
00:38:52.22 genes. So about 10% of the genome shows a change in
00:38:58.03 expression, particularly at senescence, but also, as I'll
00:39:03.05 show you, some genes remain changed even when the
00:39:08.03 cells are growing well, maintaining their telomeres through
00:39:11.26 this recombination mechanism. So one can analyze all the
00:39:17.02 patterns of genes whose expression has changed and
00:39:21.00 compare them with known patterns of gene expression
00:39:24.20 changes that have been looked at in a variety of
00:39:28.00 experimental settings by the many people in the yeast
00:39:33.05 genomics and gene expression community, who've
00:39:36.17 looked at gene expression changes. So one can
00:39:38.28 compare what we see here with what's been seen in
00:39:42.04 other experiments done by many other groups looking at
00:39:46.00 many other conditions of changes to the cells in yeast. So
00:39:53.19 what we found was that in fact, this pattern we observed
00:39:57.04 was unique, and we gave it a name: the telomerase
00:40:00.19 deletion response, the TDR. Now we found a set of
00:40:05.02 genes uniquely upregulated in response to the deletion of
00:40:09.17 telomerase RNA, and we called that the "telomerase
00:40:12.04 deletion signature." It was a particular group that
00:40:15.08 behaved as a group only when you deleted the
00:40:17.18 telomerase RNA, so that would be a group of those that
00:40:20.19 showed up as red, because they were upregulated. And
00:40:24.01 that hadn't been seen by any other manipulation. So the
00:40:27.16 cells indeed were sensing that they were losing
00:40:30.16 telomerase, and they responded in a particular way, by
00:40:34.08 changing their physiology, which we read out as a
00:40:36.21 change in gene expression profile. Right at the stage of
00:40:41.27 senescence, when I showed you that there were a lot of
00:40:44.13 genes that particularly were turned up or down in their
00:40:48.03 activity, there was a particular known DNA damage
00:40:54.01 response profile that was very prominent. That actually
00:40:57.25 makes a lot of sense. If you remember, I told you that,
00:41:00.24 when telomeres become short, they will become prone to
00:41:04.05 fusions, and now when they fuse, then chromosomes will
00:41:07.21 get ripped apart as they try to go through the ensuing cell
00:41:11.23 divisions, because the chromosomes with fused telomeres
00:41:14.26 will now have two centromeres that will pull apart in
00:41:17.15 mitosis, breaking the chromosomes at random positions,
00:41:21.09 and causing essentially genomic havoc, which can
00:41:23.27 quickly lead to cell death if the cells try to keep dividing.
00:41:28.09 And there's a damage response that takes place when
00:41:31.23 DNA breaks appear, and this is what we see, so this
00:41:34.29 makes very good sense. At senescence, we see a
00:41:37.14 damage response. What was very intriguing was that we
00:41:41.03 also saw others things. We saw what's called an
00:41:42.25 "environmental stress" cellular response. This is a
00:41:46.17 common set of genes that are upregulated and
00:41:49.00 downregulated in response to various insults, chemical
00:41:52.28 insults of various kinds, but it's a common set of genes, so
00:41:57.09 this is called the environmental stress response, and we
00:41:59.19 saw that response occurring, particularly at senescence.
00:42:05.05 And there was a change to an aerobic metabolism
00:42:08.15 program, and indeed we found that the number of
00:42:10.05 mitochondria at senescence went up enormously. Very
00:42:13.11 intriguing observations, not expected. Now, the survivors
00:42:20.02 were fascinating because, even though I showed you
00:42:22.26 they appear to grow quite well, they had a gene
00:42:26.15 expression transcriptional profile which was distinct from
00:42:29.24 the wild-type cells. And what persisted was a subset of
00:42:34.27 the environmental stress response genes that stayed on,
00:42:39.27 so what's interesting is that those cells, which appear to
00:42:42.10 be growing well on the surface, "stoics," they're really
00:42:45.27 hurting; their physiology is different, and they sense
00:42:49.25 themselves as being under what they sense as a cellular
00:42:53.23 stress. So, you can have cells that grow well without
00:42:58.20 telomerase, but if you look at the cells, these cells are
00:43:02.05 behaving as though under an environmental stress, and
00:43:05.25 so they indicate by their gene expression profiles a cellular
00:43:09.21 stress response. So in fact, cells do grow better with
00:43:15.04 telomerase than without it, even though superficially they
00:43:18.20 look similar, cells growing without telomerase maintaining
00:43:22.00 their telomeres via this recombination type of mechanism,
00:43:25.24 which is presumably somewhat more haphazard, is putting
00:43:28.27 a continuous stress on the cells, even though they've
00:43:32.04 adapted to it by a stress response. So, what we learned
00:43:38.09 then is that yeast lacking telomerase, using recombination
00:43:42.09 to maintain telomeres, do grow well, but they're under
00:43:45.10 continuing cellular stress, and this is what these gene
00:43:48.29 expression profile analyses showed us. So, let's go back
00:43:55.28 to telomerase and consider again what it's doing. If you
00:44:00.19 have plenty of telomerase, as indeed yeast cells normally
00:44:05.00 do, unless we genetically or in other ways ablate the
00:44:08.13 action of telomerase... if you have plenty of telomerase,
00:44:11.25 then there's kind of a homeostasis of the telomeres. The
00:44:15.21 telomeres stay within a certain range of lengths, and
00:44:19.06 remember I said at the very beginning of this lecture that
00:44:22.23 telomeric repeats are different in different molecules of a
00:44:27.02 population of cells, in different DNA molecules. And in
00:44:31.03 fact, they distribute around an upper and a lower limit.
00:44:35.27 And telomerase is one of the things that's crucial for
00:44:39.06 keeping them within this limit, and if you don't have
00:44:41.25 telomerase, then because of the DNA replication
00:44:46.05 problems, they gradually fall below. But they don't get too
00:44:49.25 long either, and a great many factors limit the action of
00:44:54.01 telomerase on telomeres as well, so they don't get too
00:44:57.05 long, and telomerase is also regulated on telomeric ends,
00:45:03.22 so that as the telomeres get shorter, then telomerase has
00:45:07.07 a higher probability of acting on the telomeres. This is a
00:45:09.25 complex system, we call it a homeostasis type of system,
00:45:14.27 and it balances out the lengthening and shortening
00:45:17.23 processes, so that the net result is that the telomeres stay
00:45:21.02 some average length within limits, and therefore the cells
00:45:24.21 keep dividing. Now, everything I've said implied that, if
00:45:32.22 you took away telomerase, there would be quite a delay
00:45:37.11 before any response was seen, and indeed, at a gross
00:45:42.09 level that is true. But if one scrutinizes the cells very
00:45:46.21 carefully, one sees that actually, things are a little bit
00:45:50.12 different. So an experiment in yeast was done, which was
00:45:54.26 to remove telomerase, as I showed you, and if you look at
00:45:58.00 the cells, I showed you growing on the plate, lots of
00:46:01.10 colonies, right away things look pretty okay. But if one
00:46:07.12 used very sensitive, quantitative molecular probes to look
00:46:12.05 at the telomeres in those cells, still with their long
00:46:16.22 telomeres, one found that there were occasional... every
00:46:20.22 few thousand cells or so had very short telomeres, and
00:46:24.28 they would actually fuse with a DNA break that was
00:46:28.13 induced to occur in those cells, something a telomere
00:46:33.05 never should do if it's a properly kept, functional telomere.
00:46:37.24 So, one in a few thousand cells had dysfunctional
00:46:42.17 telomeres, as manifested by the fact that they underwent
00:46:46.09 these normally forbidden fusion events. So immediately,
00:46:52.23 there was a response, even if one made the telomeres all
00:46:55.25 longer and then took away telomerase, one still saw this,
00:46:59.16 even when the bulk telomeres were long, one could see
00:47:02.16 this kind of aberrant dysfunctional telomeres immediately
00:47:09.03 after loss of telomerase. So this is one piece of evidence
00:47:14.20 that telomerase actually is protecting the ends of
00:47:18.11 telomeres, even when they're plenty long enough,
00:47:20.25 telomerase itself seems to protect and sit at the ends of
00:47:24.21 telomeres, is one interpretation, and protects them from
00:47:29.14 catastrophic shortening and fusioning to a double-strand
00:47:32.10 break, even when telomeres are long. This doesn't
00:47:36.28 happen in a high number of the cells, but it happens
00:47:41.10 measurably, and so this protective function of telomerase
00:47:46.18 is important. So I've introduced you to telomerase, I've
00:47:52.15 shown you how it's important for the long-term growth of
00:47:54.21 cells because it is necessary for continuous replenishment
00:47:58.28 of telomeres, and I've introduced you to the first piece of
00:48:02.18 experimental evidence that telomerase has a protective
00:48:06.01 function in cells, even when telomeres are long. So in the
00:48:11.06 next lecture, we'll talk about telomerase and more of its
00:48:16.28 protective functions in different cellular settings.

Normal human chromosomes have long G-rich telomeric overhangs at one end

  1. Woodring E. Wright1,3,
  2. Valerie M. Tesmer1,
  3. Kenneth E. Huffman2,
  4. Stephen D. Levene2, and
  5. Jerry W. Shay1
  1. 1Department of Cell Biology and Neuroscience, The University of Texas Southwestern Medical Center, Dallas, Texas 75235-9039 USA; 2Program in Molecular and Cell Biology, University of Texas at Dallas, Richardson, Texas 75083 USA


Telomeres protect the ends of linear chromosomes from degradation and abnormal recombination events, and in vertebrates may be important in cellular senescence and cancer. However, very little is known about the structure of human telomeres. In this report we purify telomeres and analyze their termini. We show that following replication the daughter telomeres have different terminal overhangs in normal diploid telomerase-negative human fibroblasts. Electron microscopy of those telomeres that have long overhangs yields 200 ± 75 nucleotides of single-stranded DNA. This overhang is four times greater than the amount of telomere shortening per division found in these cells. These results are consistent with models of telomere replication in which leading-strand synthesis generates a blunt end while lagging-strand synthesis produces a long G-rich 3′ overhang, and suggest that variations in lagging-strand synthesis may regulate the rate of telomere shortening in normal diploid human cells. Our results do not exclude the possibility that nuclease processing events following leading strand synthesis result in short overhangs on one end.

Telomeres are the specialized ends of linear chromosomes that are involved in a variety of functions, including meiotic chromosome segregation, chromatin silencing, and protecting the ends of the chromosomes from degradation or end-to-end fusion (for review, see Blackburn 1994; Zakian 1995; Greider 1996). In most organisms, telomeres are composed of repetitive sequences in which the strand with its 3′ end at the terminus is G-rich and may extend beyond the DNA duplex to form a single-stranded G-rich overhang. In humans, telomeres contain up to several thousand repeats of the sequence TTAGGG (Moyzis et al. 1988; Cross et al. 1989). Because of the requirement for an RNA primer, DNA polymerases are unable to replicate the extreme 3′ end of a parental DNA strand (Watson 1972; Olovnikov 1973) and, in the absence of compensatory mechanisms, telomeres shorten with each cell division. The ribonucleoprotein telomerase provides such a compensatory mechanism. Telomerase contains reverse transcriptase motifs (Lingner et al. 1997), and using its RNA component as a template (Greider and Blackburn 1989), it can add repetitive sequences to the 3′ end of the chromosomes. Eliminating the RNA component of telomerase prevents this activity and results in telomere shortening in organisms ranging from yeast to humans (Singer and Gottschling 1994;Blasco et al. 1995; Feng et al. 1995). Telomerase activity can be detected in the vertebrate testis (Prowse and Greider 1995; Wright et al. 1996), and telomere length is maintained in the germ line (Cooke and Smith 1986; Hastie et al. 1990; de Lange et al. 1990). However, telomerase activity is repressed in most human tissues during development (Wright et al. 1996) and progressive telomere shortening is then observed (Hastie et al. 1990; Lindsey et al. 1991). This shortening has been proposed to serve as a mitotic clock that counts cell divisions and ultimately results in cellular senescence (de Lange et al. 1990; Greider 1990; Harley et al. 1990; Harley 1991; Wright and Shay 1995). The ability to maintain telomere length may be important in cancer formation, as approximately 85% of all human primary tumors express telomerase activity (for review, see Shay and Bachetti 1997).

The detailed structure of telomeric ends has been determined in hypotrichous ciliates such as Oxytricha nova, where a double-stranded region of 28 bp of TTTTGGGG repeats is followed by 14 nucleotides of a G-rich single-stranded overhang (Klobutcher et al. 1981). In Saccharomyces cerevisiae, although a longer single-stranded region can be transiently observed in late S phase, during most of the cell cycle any G-rich overhangs that are present are shorter than a 30-nucleotide detection limit (Wellinger et al. 1993). The loss of ∼5 bp per division in yeast lacking telomerase RNA is consistent with a model in which both ends of the yeast telomere have an ∼10-nucleotide G-rich overhang (Zakian 1995). Recent models for the action of telomerase have emphasized the need for processing of the blunt end generated by leading strand synthesis so that it can be a substrate for telomerase, with subsequent processing events generating chromosomes with symmetrical telomeres containing short G-rich overhangs (Lingner et al. 1995; Lingner and Cech 1996; Wellinger et al. 1993, 1996). These models have a working assumption that there is a primase activity that can position an RNA primer at the extreme 3′ end of the chromosome. Such a primase activity has been found inO. nova (Zahler and Prescott 1988).

In contrast to yeast telomeres which lose only a few base pairs per division in the absence of telomerase, telomeres from normal diploid human cells have been found to shorten at rates varying between 40 and 200 bp per division (Harley et al. 1990; Counter et al. 1992; Shay et al. 1993; Vaziri et al. 1993). There are at least three hypotheses to explain the much greater losses in human cells. Exposure to oxygen levels >20% causes premature senescence in human fibroblasts, and it has been proposed that unrepaired oxidative damage causes the one-step loss of long stretches of telomeric repeats (von Zglinicki et al. 1995). This hypothesis predicts that the rate of loss of telomeric DNA under normoxic conditions would represent the average between slow rates of shortening on most chromosomes and rapid losses on some damaged chromosomes. A second hypothesis is that processing events involving the nucleolytic degradation of one or both strands would cause increased rates of shortening in human telomeres (Makarov et al. 1997). There is good evidence for a variety of processing mechanisms at telomeres. Different mutations in the yeast single-stranded telomeric binding protein cdc13p can cause the massive nucleolytic degradation of the C-rich strand (Garvik et al. 1995) or a failure of yeast telomerase to maintain telomere length (Nugent et al. 1996). The appearance of transient ⩾30-nucleotide overhangs on both ends of yeast chromosomes does not require yeast telomerase (Wellinger et al. 1996), and a nuclease able to digest G4 tetrastrand structures has been identified (Liu and Gilbert 1994). These observations suggest that specific nucleolytic processing of telomeres occurs in yeast. Nucleolytic processing is also seen in ciliates. The G-rich strand added by telomerase to the newly fragmented macronuclear DNA in hypotrichous ciliates is initially longer than in mature telomeres (Roth and Prescott 1985; Vermeesch and Price 1994), and the preferential pause site used by telomerase in vitro is not found at the end of ciliate telomeres synthesized in vivo (Klobutcher et al. 1981;Henderson et al. 1988; Shippen-Lentz and Blackburn 1989; Greider 1991). A third hypothesis is that human cells lack the ability to position the final RNA priming event at the very end of the chromosome. RNA priming events are thought to occur about every 100–600 bp during lagging strand synthesis in mammals (Anderson and DePamphilis 1979; DePamphilis 1993; Waga and Stillman 1994). This is roughly consistent with the rates of telomere shortening of 40–200 bp per cell division that has been observed in cultured human cells. The length of the single-stranded G-rich overhang might thus represent the distance between the last priming event during lagging strand synthesis and the end of the chromosome.

As a first step in distinguishing between these models, we have developed techniques for purifying human telomeres and examining their structure. Our results demonstrate that the telomeres generated by leading versus lagging strand DNA synthesis are different and suggest that each chromosome has one telomere with a long G-rich overhang and one that is either blunt or has a short G-rich overhang. We provide the first direct electron microscopic measurement of the single-stranded region in telomeres from normal diploid human cells and find a 200 ± 75-nucleotide overhang. The rate of telomere shortening of 50 bp per division in these cells is consistent with models in which shortening results from overhangs produced by lagging strand synthesis. Our results do not support models of telomere shortening in which the primary mechanism is either oxidative damage or nucleolytic processing.

Previous SectionNext Section


Purification of human telomeres

Human telomeres were purified based on the ability of biotinylated oligonucleotides complementary to the G-rich telomeric repeat to anneal to the G-rich overhang in otherwise double-stranded DNA (Shay et al. 1994). Following annealing, DNA/oligonucleotide complexes were bound to streptavidin-coated magnetic beads and washed, and the telomeres were eluted and analyzed on agarose gels. Figure 1A demonstrates the sequence specificity of this purification. Although the telomeres in human placental DNA can be retrieved using biotinylated oligonucleotides containing four or six C-rich telomeric repeats (CTR4 and CTR6), neither six copies of the G-rich repeat (GTR6) nor a non-telomeric oligonucleotide (ClaHin) were able to bind telomeres. The failure of the G-rich repeat to purify telomeres suggests that the binding of telomeres by the C-rich oligonucleotide is not due to strand invasion or gaps in the double-stranded DNA but, rather, is dependent on the presence of the G-rich 3′ overhang. Digestion of the DNA with exonuclease I resulted in a fourfold reduction in recovery and confirmed that most of this purification required a single-stranded overhang (Fig. 1B). We suspect that the presence of noncanonical structures such as G-quartets involving some of the overhangs may block exonuclease I activity, leaving some exonuclease-resistant single-stranded regions intact and thus available for hybridization to the biotinylated C-rich oligonucleotides. The minimal overhang that could be recovered with this technique was determined using an artificial telomere constructed by ligating a linearized 5-kbp plasmid to short double-stranded fragments containing variable numbers of TTAGGG repeats as single-stranded 3′ extensions. Overhangs containing as few as 12 nucleotides could be recovered (Fig. 1C).

Figure 1.

 Purification of telomeres. (A) Sequence specificity of the purification of telomeres.HinfI-digested human placental DNA was annealed to various biotinylated oligonucleotides, and telomere/oligonucleotide complexes were captured with streptavidin-coated magnetic beads. The DNA remaining in the supernatant vs. that bound to the beads was then analyzed on agarose gels and probed with a 32P-labeled (TTAGGG)4oligonucleotide. CTR4 and CTR6 contain four and six copies of the C-rich terminal repeat (CCCTAA), GTR6 contains six copies of the G-rich terminal repeat (TTAGGG), and ClaHin is a nontelomeric biotinylated oligonucleotide. Only the C-rich oligonucleotides complementary to the G-rich telomeric overhang were able to retrieve the double-stranded placental telomeres. (B) Purification requires single-stranded overhangs. Treatment of the DNA with the single-stranded exonuclease Exo 1 (1 U/μg) largely abolished the ability to retrieve telomeres. Noncanonical G structures (Henderson 1995; Kipling 1995) may make a small fraction of the overhangs resistant to complete digestion. (C) Purification requires ⩾12 bases of overhang. A 5-kbp artificial telomere containing single-stranded G-rich overhangs of variable lengths was annealed to a biotinylated C-rich oligonucleotide and purified using streptavidin-coated magnetic beads. The material bound to the magnetic beads or remaining in the supernatant is large and does not enter a denaturing polyacrylamide gel (uncut lanes). The radioactive telomeric repeats were released by digestion with an enzyme that cuts the plasmid just before the start of the telomeric repeats (ClaI). Sequences containing as few as 12 nucleotides of G-rich overhangs can be recovered even if they are part of a 5-kbp-long artificial telomere.

The average efficiency of purification of telomeres [bound ÷ (bound + unbound)] using DNA from normal diploid BJ human foreskin fibroblasts was 33% ± 15% (18 experiments). If fresh biotinylated C-rich oligonucleotide was annealed to the unbound fraction, an additional 10% ± 5% (8 experiments) of the original telomeres could be recovered. Only 2% ± 2% (7 experiments) of the original telomeres were bound following a third cycle of purification. The total recovery following three cycles of purification (33% + 10% + 2% = 45%) suggested that only half of the telomeres might have long G-rich overhangs.

Daughter telomeres do not have similar overhangs

The ability to purify telomeres containing overhangs allowed us to test the hypothesis that the G-rich overhang results from the gap between the last Okazaki priming event of lagging strand synthesis and the end of the chromosome. This model predicts that a blunt end is produced by leading strand synthesis, and thus the newly synthesized G-rich daughter strand would be present on a blunt-ended telomere and would not be purified by techniques that require overhangs (Fig.2A). The daughter strands in normal diploid BJ human foreskin fibroblasts were labeled by growing cells for zero, one, or four divisions in the presence of 5-bromodeoxyuridine (BrdU). This generated unsubstituted (Thy:Thy), hemisubsituted (Thy:BrdU), and fully substituted (BrdU:BrdU) DNA. The telomeres with long overhangs were then purified, melted, and the BrdU-containing strands recovered using anti-BrdU antibodies. Figure 2B demonstrates that C-rich and G-rich daughter strands were not uniformly distributed among telomeres that have long overhangs. Unsubstituted telomeres (Thy:Thy, containing only thymidine in both strands after zero divisions in BrdU) were not retrieved by the anti-BrdU antibodies, whereas both C-rich and G-rich strands were recovered with equal efficiency from telomeres that had incorporated BrdU into both strands (BrdU:BrdU, after four divisions in BrdU). Mostly C-rich strands were bound by anti-BrdU antibodies in the hemisubstituted telomeres with long overhangs (Thy:BrdU, containing BrdU in only the daughter strand after one division in BrdU), indicating that the G-rich strand was the parental strand (and thus lacked BrdU after one round of DNA synthesis).

Figure 2.

 Telomeres with long overhangs contain newly synthesized C-rich daughter strands. (A) Schematic model for telomere replication. This model postulates that lagging-strand synthesis leaves a 3′ overhang of the parental G-rich strand. Following one round of replication in BrdU, labeled G-rich strands are present in blunt-ended telomeres, whereas labeled C-rich strands are present in telomeres that have overhangs. (B) Retrieval of BrdU-labeled daughter strands. Telomeres from BJ fibroblasts, in which both strands contained thymidine (Thy:Thy), only one strand contained BrdU after a single round of replication (Thy:BrdU), or both strands contained BrdU after four rounds of replication (BrdU:BrdU), were purified using biotinylated C-rich oligonucleotides, melted, and then precipitated with anti-BrdU antibodies. The antibody-bound DNA was then released by boiling in SDS, analyzed on agarose gels, and probed with oligonucleotides specific for each strand. The amount of purified telomeres used in each sample was not identical, as the efficiency of magnetic bead purification varied between experiments. The exposure of each lane has been adjusted to represent equivalent amounts of input telomeres (antibody bound + unbound for each strand). The newly synthesized (BrdU-containing) strand on those telomeres that contained long overhangs was the C-rich and not the G-rich strand.

Table 1 presents the results from six experiments involving nine different samples of hemisubstituted DNA. On average, 5.4(±2.8) times as much C-rich as G-rich strands were bound by the anti-BrdU antibodies. This bias for BrdU incorporation into the C-strand in telomeres with long overhangs shows that these telomeres represent a population from one and not both ends of the chromosome.

Electron microscopic examination of telomeric overhangs

We measured the length of the telomeric overhang with electron microscopy by visualizing bacteriophage T4 gene 32 protein (gp 32) bound to single-stranded DNA in tungsten-shadowed preparations. Size standards for the quantitation of the length of single-stranded telomeric overhangs were prepared that contained cloned telomeric repeats as either terminal overhangs or internal single-stranded gaps. Figure 3A shows an example of the images obtained with an internal 450-nucleotide gap, and Figure 3B shows the linear relationship between the length of the gp32-coated region and the number of nucleotides for both internal and terminal single-stranded regions. Tracings of the protein-coated region yielded a value of 0.42 ± 0.02 nm/nucleotide, very close to previously published reports of 0.46 nm/nucleotide (Delius et al. 1972; Wu and Davidson 1975). Analysis of overhang-containing BJ fibroblast telomeres coated with T4 gp32 (Fig. 3C) yielded an average single-stranded overhang length of 200 ± 75 nucleotides (Fig. 3D). This is an order of magnitude greater than that found in organisms such as ciliates and yeast (Klobutcher et al. 1981; Wellinger et al. 1993) and consistent with recent estimates for human telomeres based on indirect biochemical techniques (Makarov et al. 1997; McEllingott and Wellinger 1997).

Figure 3.

 T4 gp32 decoration of single-stranded DNA and telomeric overhangs. (A) The single-stranded region of linearized plasmid DNA containing a 450-nucleotide gap of single-stranded telomeric repeats was decorated with the single-strand binding T4 gp32. Initial magnification, 100,000×. (B) Linear relationship between size standards and measured lengths. Different lengths of single-stranded gaps (•) or overhangs (○) coated with gp32 were examined. The 48- and 450-nucleotide single-stranded regions contained G-rich telomeric repeats; the 200- and 1000-nucleotide gaps contained plasmid sequences. Except for the 48-nucleotide overhang, 40–70 molecules of each type were examined. It was difficult to distinguish the very short decorated region from background for the 48-nucleotide overhang sample, and we consider 50 nucleotides of overhang to be the limit of detection for this technique. Error bars indicate 1 standard deviation (S.D.). (C) Purified BJ fibroblast telomere decorated with gp32. Initial magnification, 25,000×. (D) Histogram of the length of the gp32-decorated regions of purified telomeres from BJ fibroblast DNA. A total of 108 and 69 molecules from population doubling level (PDL) 20 (solid bar) and 87 (shaded bar) were examined. Between 70% and 80% of the molecules purified on the basis of having G-rich overhangs had one decorated end. The remaining undecorated molecules may represent fragments of broken telomeres. None of the telomeres was decorated at both ends. Average overhang lengths (±S.D.) were 157 ± 69 nucleotides for PDL20 and 226 ± 88 nucleotides for PDL87 fibroblasts. A higher background of free T4 gp32 in the PDL87 preparation may have compromised our ability to detect the shortest overhangs in that sample.

Telomere shortening in fibroblasts

Models of chromosome replication in the absence of telomerase in which only one telomere contains a long G-rich overhang predict that the size of the overhang should be four times the amount of shortening per doubling (Fig. 4, step 1). BJ fibroblast telomeres shorten by ∼50 bp per division in culture (Fig.5), which is one-fourth of the 200-nucleotide overhang we observed. Models for telomere shortening in which nuclease processing generates symmetrical overhangs at both ends of the chromosome predict that the size of the overhang should be twice the rate of shortening per division (Fig. 4, step 2), which is not supported by our data.

Figure 4.

 Telomere size changes after chromosome replication in the absence of telomerase. A replication fork is shown proceeding into a telomere of length L with a 3′ G-rich overhang 1 unit long. The size of the telomere before replication is the average of the two strands, which is [(L + 1) + L] ÷ 2 = L + 0.5 units. Lagging strand synthesis is illustrated as a series of discrete Okazaki fragments that would be joined together to form a continuous strand. Following replication, lagging strand synthesis would leave a long 3′ overhang; leading strand synthesis would generate a blunt end. After replication is complete (step 1) the average size of the four strands would be [(L + 1) + L + L + L] ÷ 4 = L + 0.25 units. The net shortening after replication would be (L + 0.5) − (L + 0.25) = 0.25, implying that the rate of telomere shortening should be one-quarter of the length of the G-rich overhang. A recent model (Makarov et al. 1997) has been postulated, in which extensive nuclease processing produces symmetrical long overhangs (step 2). If this were to occur, the average size of the four strands would be [(L + 1) + L + L + (L − 1)] ÷ 4 = L. The net shortening after replication and processing would thus be (L + 0.5) − L = 0.5, suggesting that the rate of telomere shortening in the absence of telomerase should be one-half the length of the G-rich overhang. Our data indicate that the rate of shortening (50 bp/division) is one quarter the length of the overhang (200 ± 75 nucleotides) in BJ fibroblasts.

Figure 5.

 Telomere shortening in BJ fibroblasts. (A) Terminal restriction fragment (TRF) gel of DNA from BJ foreskin fibroblasts at different mean population doubling levels. (B) Rate of telomere shortening. Data from three different TRF gels using DNA prepared from two different lifespan studies are shown. The average rate of shortening was 49 bp/population doubling.

Previous SectionNext Section


To our knowledge, the only rigorous evidence for symmetrical overhangs on both ends of a chromosome comes from studies of hypotrichous ciliates, in which the small and well-defined telomere length has permitted the direct sequencing of both telomeric strands of end-labeled total DNA (Klobutcher et al. 1981). The larger size and variable length of telomeres in other organisms has prevented a comparable analysis of their telomeres. The data in yeast showing the temporary circularization of linear plasmids owing to the presence of transient overhangs demonstrates that overhangs might be present on both ends of some of the chromosomes some of the time but does not directly address whether or not the overhangs are symmetrical or uniform. Our data show that the presence of long overhangs on only one end of the chromosome in normal human fibroblasts differs from the symmetrical ends found in the hypotrichous ciliates.

Anti-BrdU antibodies recognized the C-rich but not the G-rich strand in telomeres purified based on their long overhangs from cells after a single cycle of DNA replication. This demonstrates that these telomeres contain a daughter strand synthesized from the G-rich telomeric template. Yeast origins of replication are internal to the telomere (Raghuraman et al. 1997). Assuming that the origins for telomere replication are also internal in vertebrates, our results show that long overhangs are present in telomeres that are produced by lagging-strand synthesis.

The 200 ± 75-nucleotide overhang we observed in BJ fibroblasts is sufficient to fully explain the rate of telomere shortening of 50 bp per division found in these cells using calculations based on models in which the telomere at one end of the chromosome is blunt and the one at the other end has a long G-rich overhang. These observations do not support the hypothesis that the rate of shortening represents the average of a few telomeres, which suffer oxidative damage and lose kilobases of telomeric sequences, and the majority of telomeres, which lose only very few nucleotides because of short unreplicated overhangs (von Zglinicki et al. 1995).

The biotinylated C-rich oliogonucleotides we have used are unable to bind single-stranded G-rich overhangs of less than ∼12 nucleotides with sufficient stability to recover artificial telomeres 5 kbp long. We are thus unable to determine whether the telomeres resistant to purification are blunt ended or have very short overhangs. The presence of short overhangs would be consistent with the concept that proteins recognizing a single-stranded overhang might be required to cap the telomeres and prevent them from being degraded (Gottschling and Cech 1984; Gottschling and Zakian 1986). The BrdU labeling data does not exclude the possibility that leading and lagging strands might be processed differently, so that only one end is packaged into a G-quartet-like structure that is stable to the DNA isolation procedure and resistant to hybridization to the biotinylated C-rich oligonucleotides. Under this circumstance, one end might have a long but inaccessible G-rich overhang. However, the relationship between the measured length of the overhang and the observed rate of shortening (Figs. 3, 4, 5) argues against long overhangs on both daughter telomeres.

Makarov et al. (1997) recently reported that human telomeres contain long G tails at both ends of the chromosome. The strand-replacement technique (PENT) that these workers used to detect G-rich overhangs does not distinguish between long and short overhangs, and thus their conclusion that human cells contain symmetrical overhangs is not justified. In addition, their claim that 85% of human telomeres have G-rich overhangs may be in error. This assertion is entirely dependent on PhosphorImager scans of the relative intensities of three bands: full-length C-rich strands (Co), newly synthesized replacement strands primed from the G-rich overhang (Cs), and the original C-strand being trimmed back as replacement synthesis progresses (Ct). Their quantitation showed that the intensity of Co was ∼15% of the sum of Cs + Ct, and they thus concluded that 85% of the telomeres had overhangs. However, their data (Fig. 5 in Makarov et al. 1997) indicate that they underestimated the amount of large DNA present. For example, even assuming that a 12.5-kb telomere digested with HinfI contains 2.5 kb of subtelomeric DNA that lacks repeats (Levy et al. 1992), the number of telomeric repeats in a 12.5-kb Co and a 10-kb Ct strand should be approximately three times the number in a 2.5-kb Cs strand. Rather than finding a threefold greater signal intensity in the large DNA probed with the G-rich telomeric repeat, their data show an approximately equal intensity of the small 2.5-kb Cs strand. Their quantitation of the fraction of telomeres that lacked available overhangs and were resistant to replacement synthesis (the fraction that was the Co strand) could thus be off by a factor of three or more. We suspect that these results may be explained by an inefficient transfer of large DNA to the membrane in the vacuum blotting procedure they employed.

Much recent evidence implicates single-strand nucleases in telomere processing in telomerase-expressing model organisms (Wellinger et al. 1993, 1996; Garvik et al. 1995; Linger et al. 1995; Nugent et al. 1996), raising the possibility that human telomeres might be processed to have uniform G-rich overhangs. The data in the present report provide direct evidence that this is not the case in a telomerase-negative cell strain. Leading and lagging strand synthesis results in distinct telomeric structures in normal diploid human fibroblasts. Our results do not formally exclude the possibility that asymmetric overhangs could be generated by nuclease activity. We consider this unlikely, as an additional mechanism to restrict the digestion to the newly synthesized C-rich strand and not the parental C-rich strand would be required.

The presence of telomerase activity in most human tumors has generated much excitement concerning the potential efficacy of anti-telomerase therapies for the treatment of cancer. Understanding what regulates the rate of telomere shortening and how to manipulate it may provide the tools to increase the effectiveness of telomerase inhibitors in preventing the regrowth of tumor cells.

Previous SectionNext Section

Materials and methods

Purification of telomeres

Double-stranded genomic DNA was digested overnight in 0.25 U/μg of HinfI to free the telomeric repeats from most of the subtelomeric sequences. In a typical experiment, 30 μg of DNA in a final volume of 30 μl was then adjusted to 1× SSC/1% Triton X-100, mixed with 1 pmole of a biotinylated oligonucleotide, annealed for 15 min each at 65°C, 55°C, 45°C, 35°C, and room temperature, and combined with 3 μl of washed streptavidin-coated magnetic beads (10 mg/ml suspension, Dynal Inc.) that had been preincubated for at least 1 hr in 5× Denhardt’s solution. The DNA–bead suspension was rotated end over end at 2 rpm and 4°C overnight. The magnetic beads were drawn to the side of the tubes using a rare earth magnet (Edmund Scientific), and the supernatant removed and saved. The beads were resuspended and washed twice with 100 μl of 1× SSC/1% Triton X-100. The bound telomeres were eluted from the beads by melting the oligonucleotide/telomere interaction at 65°C for 10 min in 30 μl of 0.1× SSC/1% Triton X-100.

The fraction of telomeres recovered [bound ÷ (bound + unbound)] was quantitated from PhosphorImager scans of agarose gels probed with labeled (CCCTAA)4 oligonucleotide. In some cases, the unbound fraction was subjected to additional cycles of purification. The percent of telomeres recovered during each cycle of binding was normalized to the original amount of input telomeres. For example, if 35% of the telomeres were recovered during cycle one, then only 65% of the telomeres would be left as the input to cycle two. A recovery of 20% of these telomeres would then represent 13% (0.2 × 0.65 = 0.13) of the original telomeres.

Technical points in this protocol include the following considerations. Forty micrograms of human DNA contains approximately 1 fmole of telomeres, so that 1 pmole oligonucleotide is in vast excess for even relatively long overhangs. One microliter of magnetic beads can bind at least 1 pmole of biotinylated oligonucleotide as measured by its ability to retrieve radioactive (TTAGGG)4 oligonucleotide annealed to the biotinylated CCCTAA oligonucleotides. Although the rate of formation of magnetic bead/small oligonucleotide complexes is very rapid (1-μl beads can clear 1 pmole of oligonucleotide from 1 ml in ∼15 min), the rate of binding to the large telomere/oligonucleotide hybrids is very slow. The amount of telomeres bound to 1 pmole of biotinylated oligonucleotide and 3 μl of beads after an overnight rotation in 30 μl was about twice as great as when 1 μl of beads was used, with a minimal increase observed using 10 μl of beads. Three microliters of magnetic beads was used, as larger amounts increased background binding slightly.

Artificial telomeres

Large DNA fragments with single-stranded G-rich overhangs were constructed by ligating variable numbers of telomeric sequences to a linearized plasmid. Single-stranded ladders of TTAGGG repeats were produced by the asymmetric PCR amplification of a cloned 450-bp telomeric insert using [α-32P]dGTP and ddG. The ∼100 bp of plasmid sequences between the forward primer and the start of the telomeric repeats was then made double-stranded, digested with SalI, and ligated to a 5-kbp plasmid digested withXhoI. These enzymes have compatible ends. The resulting artificial telomeres were then gel purified and tested for their ability to bind biotinylated C-rich sequences.

Anti-BrdU precipitation of telomeres

Fibroblasts were grown in 30 μm BrdU for 20 hr (less than one doubling) to produce hemisubstituted DNA and for 1 week in 8 μm BrdU to produce DNA labeled in both strands. Deoxycytidine (200 μm) was included under both conditions to reduce toxic effects of BrdU. Purified telomeres from 10 μg of total genomic DNA were melted at 99°C for 3 min, quick chilled, and incubated with 5 ng of anti-BrdU antibody (Becton-Dickinson) for 60 min at room temperature in 25 μl of PBS containing 1% Triton X-100. Antibody–DNA complexes were then recovered following a 1-hr incubation with 15 μl of protein–G agarose beads (Boehringer Mannheim) and washed twice for 15 min each in PBS/1% Triton X-100. The DNA was then eluted from the beads at 99°C for 3 min in TE buffer (10 mm Tris, 1 mm EDTA at pH 8) containing 1% SDS. Input, bound, and unbound fractions were analyzed on duplicate agarose gels probed with either labeled C-rich or G-rich telomeric probes. Approximately 75% of the input telomeres were recovered (anti-BrdU bound + unbound) for each strand. The fraction of each strand bound by anti-BrdU antibodies was determined from PhosphorImager scans of each lane using the formula % Retrieved = {anti-BrdU bound ÷ (bound + unbound)} × 100.

Electron microscopy

Gapped linear DNA (100 ng) or telomeric DNA (140 ng) was incubated with 200 ng of T4 gene 32 protein (Delius et al. 1972; Wu and Davidson 1975) for 5 min at room temperature in 50 μl of 10 mm HEPES (pH 7.5), 100 mm NaCl, and 2.5 mm MgCl2. Glutaraldehyde (Sigma) was added to a final concentration of 0.1% and incubated for 5 min at room temperature. The cross-linking reaction was quenched by the addition of an equal volume of 10 mm Tris-HCl and 1 mm EDTA. DNA was suspended in a buffer containing 2.5 mm spermidine and applied to glow-discharged, thin carbon films supported on copper grids (Griffith and Christiansen 1978), rinsed twice in double-distilled water, dehydrated in a graded series of ethanol solutions, briefly stained with 0.1 mm uranyl acetate in 90% ethanol, and air-dried. Samples were rotary shadowcast at an angle of 7° using evaporated tungsten wire in a vacuum evaporator cryopumped to <10−6 Torr. Size standards were prepared by annealing single-stranded phagemid containing different numbers of telomeric repeats to the complementary strand of the plasmid vector backbone. This hybrid DNA was then restriction digested to produce linear DNA molecules containing known lengths of either internal or terminal regions of single-stranded telomeric repeats.

Terminal restriction fragment analysis

DNAs isolated from cells at different population doublings throughout their cultured lifespan were digested with a mixture of six restriction enzymes (AluI, CfoI, HaeIII,HinfI, MspI, and RsaI; Rogalla et al. 1994), analyzed on 0.5% agarose gels, and probed with a labeled (CCCTAA)4 probe. Signal intensity is proportional to the number of telomeric repeats using this probe. Average telomere length was determined from PhosphorImager scans after normalizing for this effect using the formula Σ(Intensity) ÷ Σ(Intensity ÷ Length) (Harley et al. 1990).

Previous SectionNext Section


This work was supported by grants from the National Institutes of Health (NIH) (to W.E.W. and S.D.L), the AlliedSignal corporation (W.E.W. and J.W.S.), and the Welch Foundation (W.E.W.). V.M.T. was funded in part by an NIH Oncology Training Grant. We thank Jennifer Cuthbert for valuable scientific discussions, and Ia Dac-Korytko and Martha Liao for technical support.

The publication costs of this article were defrayed in part by payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 USC section 1734 solely to indicate this fact.

Previous SectionNext Section


  • ↵3 Corresponding author.

  • E-MAIL wright{at}; FAX (214) 648-8694.

    • Received July 31, 1997.
    • Accepted September 5, 1997.
  • Cold Spring Harbor Laboratory Press


  1. (1991) Telomerase is processive.Mol. Cell. Biol.11:4572–4580, .

0 Replies to “The Telomere Function And Synthesis Essay”

Lascia un Commento

L'indirizzo email non verrà pubblicato. I campi obbligatori sono contrassegnati *