ABSTRACT
We report the identification and cloning of the telomeres of the filamentous fungus,
Aspergillus nidulans
. We have identified three classes of cloned chromosomal ends based on the
telomere-associated sequences (TASs) and demonstrated that the telomeric repeat
sequence is TTAGGG, identical to that found in vertebrates, including humans, and some lower eukaryotes. One category of telomere clones was found to contain internal,
variant TAAGGG repeats. The
A.nidulans
telomeric tract length is strikingly short (4-22 repeats). We demonstrate that telomere length is remarkably stable in
different cell types and at altered growth temperatures, suggesting a highly
regulated mechanism for length control.
The study of telomeres has seen a renaissance since the early cytological
observations of eukaryotic chromosome termini in plants (
1
). The generally G/C-rich telomeric repeats found at chromosome ends are synthesized by the
enzyme telomerase (reviewed in
2
), providing a means for the complete replication of the chromosome. Telomeres
were originally shown to be involved in preventing end-to-end chromosome fusion thereby increasing chromosome stability (
1
,
3
,
4
). They have also been shown to silence nearby genes (
5
), and increase chromosome stability (
6
). More recent evidence seems to implicate telomeres in a role in nuclear
movement in pre-meiotic cells (
7
). However, the cellular roles of telomeres are still only poorly understood.
Telomeres also define the boundaries of the genetic and physical maps of
chromosomes and so are particularly important for the mapping of large genomes
that is now being attempted.
The multinucleate ascomycete fungus
Aspergillus nidulans
has been used extensively for the study of eukaryotic gene structure,
organization and regulation (
8
).
Aspergillus nidulans
has been especially valuable for investigating the genetic and molecular processes responsible for controlling development (
8
,
9
and references therein) and mitotic regulation (
10
and references therein).
Previous studies have shown that the
A.nidulans
telomeric repeat sequence is likely to be TTAGGG (
11
). Here we report the characterization of cloned
A.nidulans
telomeres, and demonstrate this organism has surprisingly short telomere tracts
of 4-22 TTAGGG repeats. These tracts appear to be rigorously maintained in
different cell types and under different growth temperatures.
Aspergillus nidulans
(FGSC26, ATCC #24758) were grown in solid YAG medium (5 g/l
yeast extract, 20 g/l
dextrose, 20 g/l
agar, 1 ml/l
Cove's Trace Elements, 1.2 g/l
MgSO
4
.7H
2
O) and in liquid YG medium (as described, excluding agar). Unless otherwise
specified
A.nidulans
cells were grown at 37oC; on solid medium, cells were incubated for 24-48 h. In liquid culture, sterile YG media was inoculated with
A.nidulans
spores to a final concentration of 10
6
spores/ml
and
grown
for typically 16-24 h with vigorous shaking/aeration (250-300 r.p.m.).
Escherichia coli
cells (XL-1Blue MRF') were grown in LB media containing 100 [mu]g/ml
ampicillin. Cells were grown for 12-16 h at 37oC, and in liquid culture shaken vigorously (250 r.p.m.) overnight.
Aspergillus nidulans
cells (mycelia) were grown in liquid culture by inoculating 100 ml YG medium with a liquid spore stock (1.2 * 10
9
spores/ml) to a final concentration of 1.2 * 10
6
spores/ml. Cells were harvested by filtration over two sheets of sterile
MiraCloth (CalBiochem), washed with 1 l sterile MilliQ water (MilliPore), and
the resulting mat of cells transferred to a sterile 50 ml Falcon tube. Cells
were lyophilized overnight, and subsequently ground to a fine powder with a
pestle and mortar using 425-600 micron glass beads (Sigma). Twenty milligrams of this dry cell
material was used to prepare DNA, using methods described by Raeder and Broda (
12
). Conidial spore genomic DNA was prepared as described in (
13
).
Plasmid DNA (Bluescript
TM
-based, Stratagene) was prepared by inoculating 5-10 ml LB containing 100 [mu]g/ml
ampicillin with a single bacterial colony. Cells were grown to saturation, harvested and plasmid DNA isolated using Qiagen DNA columns, according to manufacturer's instructions (Qiagen).
To clone the
A.nidulans
telomeres, 1 [mu]g genomic DNA was treated with 10 U T4 DNA polymerase for 10 min at 12oC in order to create blunt-ends. The genomic DNA was then ligated overnight at 15oC to an
Eco
RV-cut (blunt-ended) Bluescript
TM
vector (Stratagene). Subsequently, the genomic DNA was digested with
Eco
RI which cleaves the
A.nidulans
DNA and cuts once in the vector polylinker. The resulting vector joined to the
putative terminal DNA fragments were ligated at 15oC overnight with T4 DNA ligase (New England Biolabs, NEB) to re-circularize, thus producing a library of telomeric DNA ends. A proportion of the
library was transformed into
E.coli
(XL-1Blue MRF', Stratagene) and plated onto selective media. The plated library
was transferred to HyBondN+ (Amersham) membrane by performing colony lifts, and
screened for putative telomeric DNA using a radiolabeled (TTAGGG)
4
oligonucleotide probe. Positive bacterial colonies were identified and
Bluescript plasmids containing putative
A.nidulans
telomeric DNA were prepared. Unambiguous DNA sequencing data was obtained from
the clones containing telomeric repeat sequences using a Sequenase
TM
DNA sequencing kit (US Biochemicals) using [[alpha]-
35
S]ATP (1000 Ci/mmol) and primed by a 18 base T7 promoter oligonucleotide. All
DNA sequence data collected was, therefore, derived by copying the C-rich telomeric strand.
Bal31 nuclease (NEB) time course reactions were performed in a 35 [mu]l volume of buffer (12 mM CaCl
2
, 12 mM MgCl
2
, 0.2 M NaCl, 20 mM Tris-HCl pH 8.0, 1 mM EDTA) on ~14 [mu]g
A.nidulans
genomic DNA at 30oC. Aliquots of the reaction mix were taken at 5 min intervals, up to 65
min, and stopped by addition to 0.5 M EGTA to a final concentration of 20 mM in
the reaction aliquot. An initial 5 [mu]l aliquot of the reaction mix, prior to addition of nuclease, provided a
zero time point for the time course. Terminated reactions were kept frozen on
dry ice prior to digestion with
Eco
RI. Loading dyes (with RNAse A) were subsequently added and subject to agarose
gel electrophoresis.
Restriction enzyme digestion (NEB) reactions and enzyme reactions required for
cloning telomeric DNA, e.g. T4-derived DNA ligase and DNA polymerase (Boehringer), were carried out under
conditions specified by the manufacturer.
Agarose gels were typically electrophoresed at 1.2 V/cm for 12-16 h, and DNA visualized by staining in ethidium bromide solution (10 [mu]g/ml) and photographed (Polaroid) under UV illumination.
Agarose gels used for Southern blot analysis were subsequently depurinated,
denatured and transferred to HyBondN+ membrane (Amersham) in 0.4 M NaOH
typically for 6-12 h. Transferred DNA was UV crosslinked to the membrane (StrataLinker
1800, Stratagene) prior to hybridization to
32
P-radiolabeled probes. Hybridization was carried out in SDS-Na
2
PO
4
using the procedure of Church and Gilbert (
14
). Quantitation of radioactivity on Southern blots was performed using a PhosphorImager workstation (Molecular Dynamics, Sunnyvale, CA). The TAS probes used in Figure
4
were hybridized in 0.5 M phosphate buffer (
14
) at 37oC, overnight, and washed in 50 mM phosphate buffer 3 * 5 min at 25oC, followed by 2 * 5 min washes at 60oC.
DNA sequencing reactions (Sequenase, US Biochemicals) were analyzed on 0.4 mm thickness 6.0% polyacrylamide-7 M urea denaturing gels. Gels were dried down onto 3MM paper (Whatman) and subject to autoradiography.
Oligonucleotide DNA probes (Only DNA
TM
, TX) were generated by 5'-end-labeling with T4 polynucleotide kinase (NEB) and [[gamma]-
32
P]ATP (6000 Ci/mmol). The sequence of the 24 base 18S rDNA oligonucleotide probe
(Fig.
3
) used as a chromosome internal marker is 5'-AAGTCGTAACAAGGTTTCCGTAGG-3'.
Telomere-associated sequence (TAS) DNA fragments were produced using a GeneAmp kit
(Perkin-Elmer Cetus) using standard protocols. TAS PCR products were agarose gel
purified using GeneClean (Bio101, Inc.) and subsequently labeled to high
specific radioactivity using a MultiPrime kit (Amersham) in the presence of [[alpha]-
32
P]ATP and [[alpha]-
32
P]CTP (3000 Ci/mmol). The sequence of the TAS A PCR primers is described in
Figure
2
B. The sequence of the TAS B primers is 5'-GAAGTGCTAGGATAGCTTTAATC-3' and 5'-CGCAGGAATTCCTTTTAGAG-3'. The sequence of the TAS C
PCR primers is 5'-GACGTGCCGTACGTACGTAC-3' and 5'-TCCTTGGATAATGTCGCAG-3'. The sizes of TAS A, B and C
DNA fragments are ~240, ~200 and ~400 bp respectively.
32
P (and
35
S) radiolabeled DNA was visualized by autoradiography using Kodak X-OMat film at -80oC using fast-tungstate intensifier screens.
To further characterize the
A.nidulans
telomeres, we cloned the telomeres into a Bluescript vector using techniques
that enrich and preserve full telomere length (
15
-
17
). DNA sequence was obtained from 11 telomeric clones, and all clones were found
to contain a terminal tract of contiguous TTAGGG repeat sequence. The number of
repeats varied from 4 to 22 (Table
1
). We note that it is a possibility, though unclear, that the clone with the
smallest telomeric repeat number may have deleted some telomeric repeat
sequence due to possible sequence instability during cloning/propagation in
E.coli
.
Analyzing the frequency of clones by repeat number, the most frequently obtained
clones (7/11) contained 15-20 repeats (90-120 bp of repeat tract).
A mean telomere length of 84 bp of TTAGGG sequence was found in the clones
studied (Table
1
), with the G-rich repeat strand oriented as expected for a telomere 5' to 3' toward the chromosome terminus.
The telomere clones identified were classified according to the sequence of
their telomere-associated sequences (TASs; Fig.
2
A). Thus, TASs A, B and C defined telomere classes A, B and C respectively (Fig.
2
A).
Sequences of the cloned TAS and telomeric DNA tracts from these three classes of
clones have been entered into the GenBank database.
The class A telomere clones contained chromosome-internal variant (TAAGGG)
n
telomere-like DNA repeats, sometimes flanked by degenerate telomere repeat-like sequences, for example, 5'-..TAAGAA.. (Fig.
2
A and B).
The distal (TTAGGG)
n
repeats were segregated in a solely terminal domain. One telomere clone (class
A) lacked a 53 bp region containing variant repeat tracts within the TAS A
sequence (schematized in Fig.
2
B).
Table 1
To unequivocally demonstrate that we had indeed cloned the telomeric DNA
sequences from
A.nidulans
, we utilized the preferential sensitivity of telomeres to the exonuclease Bal31
(
15
and references therein). We performed a Bal31 digestion time course on uncut
A.nidulans
genomic DNA. The digestion products of the Bal31 reaction were digested with
Eco
RI, separated by agarose gel electrophoresis and analyzed by Southern blotting
using a (TTAGGG)
4
oligonucleotide probe (Fig.
3
). In contrast to the 18S rDNA chromosome-internal marker (inset, Fig.
3
), the signal from the TTAGGG hybridizing bands diminished in size and
eventually disappeared with time, indicating these sequences were indeed
telomeric. Moreover, there was very little hybridization to any chromosome-internal TTAGGG hybridizing tracts under these conditions. This suggests
there are few, if any, telomere-like repetitive sequences at internal positions in the
A.nidulans
genome.
The discrete nature of these TTAGGG hybridizing bands (Fig.
3
) suggests that rather than being heterogeneous in length,
A.nidulans
native telomeres are relatively uniform and defined in length. Furthermore, at
the time of near disappearance of the telomeric DNA bands there is only a small
decrease in size compared with the zero time point, indicating that
A.nidulans
native telomeres are relatively short. From these data, we estimate that the
average telomere length is ~30-100 bp. This correlates well to the mean length of 84 bp obtained
from the cloned telomere tracts.
In order to verify the telomeric position of the cloned TAS A sequence, we
constructed a hybridization probe containing no TTAGGG telomeric repeats using
PCR (Fig.
2
B). The 238 bp TAS A PCR product was then used as a probe for the Bal31 Southern
blot employed in Figure
3
. Under these conditions, there was specific hybridization of the probe to the ~1 kb telomeric DNA fragments (Fig.
4
A and B). There was also a small amount of hybridization to the ~1.5 kb telomeric band which, however, washed off under more stringent wash
conditions (data not shown). As predicted for a telomere-located sequence, the 1 kb DNA fragment hybridizing to the TAS A probe was
shortened and then lost during the Bal31 time course (Fig.
4
B), indicating that this sequence does reside at a telomere. However, it is
possible that the TAS A probe is hybridizing to more than one chromosomal end
in the 1 kb class of telomeres. Additionally, we noted earlier that the TAS A
DNA sequence, now identified in the 1 kb telomeric band (Fig.
4
A), contained the variant TAAGGG repeats (Fig.
2
B). Cross hybridization of the (TTAGGG)
4
probe to the variant repeats could account for the slightly higher than expected signal obtained from quantitation of this band in Figure
1
B.
PCR probes were also made from TAS B and C (described in Materials and Methods)
and used to probe similar Southern blots. The TAS B probe hybridized to four
DNA fragments (~0.3, ~1 and ~2.2 kb telomeric bands, and ~0.5 kb DNA fragment) with varying signal intensities (data
not shown). The TAS C probe hybridized to >30 DNA fragments of generally higher
molecular weight than detected by TAS A and B probes (data not shown). The
class B and C telomere clones contained terminal telomeric arrays adjacent to
TAS B and C respectively, which each consisted of a complex DNA sequence, lacking any variant telomere-like or degenerate repeats as seen in TAS A (data not shown).
As discussed above, the Southern blot analysis of the
A.nidulans
telomeres (Figs
1
A and 3) showed the telomeric bands were strikingly discrete and uniform in
length. This is in contrast to the broadness of telomeric DNA bands observed
with some other eukaryotic telomeres, which is generally indicative of heterogeneous telomere repeat tract length (reviewed in
2
,
18
). This suggests that telomere length in
A.nidulans
may be tightly regulated. To search for conditions which might perturb the
telomere length or homogeneity, we first studied the length of the telomeres in
cells from a different developmental stage, asexual spores. Genomic DNA was
prepared from
A.nidulans
spores (described in Materials and Methods) and native telomere length was
analyzed by Southern blotting (Fig.
5
). No difference was seen in the native telomere lengths in the dormant spore
stage compared with the vegetatively grown mycelium (Fig.
5
, compare lane b with d). Furthermore, a Bal31 digestion time course performed
on spore genomic DNA also gave an estimated telomere length of 30-100 bp (data not shown).
We have cloned the telomeres of the multinucleate filamentous fungus
Aspergillus nidulans
. As previously predicted (
11
), the cloned telomeres comprise the repeated sequence TTAGGG. This is the same
repeat sequence found in vertebrates, including mammals (e.g. mice and humans),
Neurospora crassa
, slime molds (e.g.
Physarum
) and trypanosomes (
2
,
18
,
19
). The size, sequence and uniformity of the telomeric repeats of the filamentous fungi resemble those of higher eukaryotes rather than some budding yeasts,
which in
S.cerevisiae
contains short, degenerate repeats, while
Candida albicans
and
Kluyveramyces lactis
have very large (23 and 25 bp respectively) repeat units (
15
,
20
).
Although the distal
A.nidulans
telomeric repeats are uniform, some degeneracy was found centromere proximally.
In many organisms, variant telomeric repeats are found in subtelomeres, but the
number and degeneracy of the variant repeats depends on the species (
21
,
22
). In humans, several variant repeats are interspersed in the more inner repeats
(
23
). In one class of
A.nidulans
telomeric clones (TAS A) variant repeats were buried within a tract containing
degenerate repeats. The TAS sequence containing variant repeats was localized
to the ~1 kb telomeric DNA fragments shown to contain 6-7 chromosomal ends. It is thus possible that nearly half the
A.nidulans
telomeres contain the variant repeats. The other telomeres, and the remainder
of the
A.nidulans
genome is largely devoid of such variant telomere-like DNA sequences. Though the exact function of the internal variant
repeats is unclear, they likely arose from a mutation that occurred in a
telomeric tract and subsequent recombination, rather than through the action of
a second, variant-template telomerase (
21
and references therein).
The
A.nidulans
telomeres are dramatically different from vertebrate telomeres in one respect:
the length of the telomere tract. The telomeric TTAGGG repeat tracts of human
somatic tissues are ~10 kb, and those of some mice are up to ~150 kb (
24
). Thus, the telomeres constitute ~0.1% of a human chromosome. In contrast, we demonstrate that the
A.nidulans
telomeres are ~100 bp (0.003% of an average
A.nidulans
chromosome), only <= 3% of the vertebrate length. The only shorter known telomeres are found on
the subchromosomal fragments of some ciliated protozoa. For example, the
hypotrichous ciliates
Oxytricha
and
Stylonychia
contain 20 bp of telomeric duplex DNA, with 16 bases of 3' single-stranded overhang (
25
). Similarly,
Euplotes
telomeres comprise 28 bp of duplex telomeric sequence with a 14 base single-stranded 3' overhang (
25
). Although short relative to other organisms, the telomeric DNA of the
hypotrichous ciliates constitutes 1% of the chromosome, since the total length
of the hypotrichous ciliate subchromosome is only ~2 kb (
25
). Interestingly, the hypotrichous ciliate telomere length is extremely tightly regulated.
Perhaps the short telomeres in these organisms and in
A.nidulans
define the minimum length required to maintain function, and therefore must be rigorously controlled. Conversely, perhaps
A.nidulans
exhibits a strong selection against longer telomeres due to an unknown deleterious nature of longer telomeres.
Variation in telomere length in eukaryotes is common (
26
). In fact, one of the characteristic identifying features of telomeres of most
organisms is that they are usually heterogeneous in length, producing `fuzzy'
bands on Southern blots. However, the telomeres of
A.nidulans
appeared as more discrete bands on Southern analysis, indicating a greater
degree of uniformity in native telomere length. These results suggest that
telomere length is tightly controlled in
A.nidulans
. This is further supported by the observation that
A.nidulans
telomere length remains the same irrespective of cell types studied and changes
in growth temperature. By comparison, studies in the yeast,
Candida albicans
, revealed that telomere length increased with increase in growth temperature (
15
). One possible explanation is that disruption of telomeric chromatin at higher
growth temperatures would likely render the 3' chromosomal end accessible to extension by telomerase. In fact, recent
studies in the yeast,
Kluyveramyces lactis
,
indicate that this temperature effect on telomere length may indeed result as a
consequence of disruption of telomeric chromatin structure, in particular,
involving the protein Rap1p (
27
). Krauskopf and Blackburn (
27
) propose Rap1p participates in telomere length control in
K.lactis
via formation of a DNA tertiary structure mediated by protein-DNA and protein-protein interactions involving Rap1p molecules. Such interactions are likely strongly temperature dependent (
28
,
29
). However, in
A.nidulans
the invariance in telomere length suggests there may be a different or less
easily perturbed mechanism for telomere length control.
The studies presented in this paper will provide better landmarks for the
physical and genetic characterization of the
A.nidulans
genome. We also envision that further investigation into the tenacious control
of
A.nidulans
telomeres will reveal new insights into telomere homeostasis.
We are especially grateful to Karen Kirk for providing technical comments and
critical reading of the manuscript, and Mike McEachern and Marita Cohn (Lund, Sweden) for useful discussion. In addition, we thank Ron Morris (UMDNJ) for helpful advice. This work was
supported by a Lucille P. Markey Charitable Trust Visiting Postdoctoral
Fellowship (A.B.) and an American Cancer Society (California Division) Senior Fellowship (A.B.). Laboratory support was also provided in part by the Lucille P. Markey Charitable Trust,
and by NIH grant GM 26259 to E.H.B.
*To whom correspondence should be addressed. Tel: +1 415 476 7284; Fax: +1 415
476 8201; Email: anamitra@cgl.ucsf.edu
Clone and
No. of
No. of T
2
AG
3
Telomeric DNA
telomere class
clones
repeats
tract length (bp)
A3
1
4.7
28
A4
1
20.3
122
A5
1
20.2
121
A6
1
15.2
91
A8
1
18.0
108
A9
1
19.7
118
A10
1
18.2
109
B5.1
2
22.3
134
C8.1
2
15.8
95
REFERENCES
Return