Cytogenetic characterization of telomeres in the holocentric chromosomes of the lepidopteran Mamestra brassicae

Telomeres of the Mamestra brassica holocentric chromosomes were studied by Southern blotting, in-situ hybridization and Bal31 assay evidencing the presence of the telomeric (TTAGG)n repeat. Successively, molecular analysis of telomeres showed that TRAS1 transposable elements were present at the subtelomeric regions of autosomes but not in the NOR-bearing telomeres of the Z and W sex chromosomes. TRAS1 appeared to be transcriptionally active and non-methylated, as evaluated by RT-PCR and digestion with MspI and HpaII. Finally, dot-blotting experiments showed that the 2.8 ± 0.5% of the M. brassicae genome consists of TRAS1.


Introduction
Telomeres, the physical chromosomal termini, are essential chromosomal structures. They are involved both in preventing end-to-end chromosome fusion and in the protection of the chromosomal ends from the gradual erosion occurring during successive rounds of replication (Blackburn 1991, Biessmann & Mason 1994. Telomeric DNA consists frequently of a simple sequence: one strand is G-rich and its 3 0 end points to the chromosome end (Sahara et al. 1999). Although different telomeric sequences were reported in eukaryotes, some are more frequently represented. The (TTAGGG) n sequence, for example, constitutes telomeres of all vertebrates (Meyne et al. 1989) and the (TTTAGGGA) n repeat is conserved in the plant kingdom (Cox et al. 1993), with few exceptions (Pich et al. 1996).
Insects represent the most numerous group of organisms in the animal kingdom. Three different telomere types were reported in this group: the pentanucleotide (TTAGGA) n repeat in Bombyx mori (Okazaki et al. 1993, Sahara et al. 1999, the Het-A and TART transposable elements in Drosophila melanogaster (Biessmann et al. 1990) and the complex tandem repeat in the Chiranomus genus (Zhang et al. 1994).
In this paper, the organization of telomere in the holocentric chromosomes of the lepidopteran Mamestra brassicae is described. In view of the suitability for both in-vivo and in-vitro analysis, M. brassicae could represent an important model for the study of the architecture of holocentric chromosomes.

Materials and methods
Chromosome preparations were made by spreading CRL-8003 hemocyte cells of Mamestra brassicae following a modi¢ed protocol of the method described by Odierna et al. (1993). In brief, cells were kept in a 0.8% hypotonic solution of sodium citrate for about 45 min and then transferred to minitubes and centrifuged at 350g for 3 min. Methano^acetic acid 3 : 1 was successively added to the pellet, which was made to £ow up and down for 1 min through a needle of a 1-ml hypodermic syringe. Finally, the pellet was resuspended in 200 ml of fresh ¢xative and 20 ml of cellular suspension was dropped onto clean slides and air dried.
The (TTAGG) n , (TTAGGG) n and (TTAGGGG) n telomeric probes were obtained by PCR ampli¢cation according to Ijdo et al. (1991). The 197-bp Chironomus telomeric probe has been obtained by PCR using the primers F (5 0 -TGAGGAAATTGAAGGGCAAA) and R (5 0 -CTTCCTCATCGGCTGGAGTT) designed on the basis of the GenBank sequence AF026061 from Chironomus pallividittatus larvae DNA. TRAS1 probe has been obtained by PCR using the primers F (5 0 -ATGAAGGGTGGGTGAGAGTG) and R (AGGGTTACGGGCACTCTTT) designed on the TRAS1 sequence from B. mori (BMOTRAS1). The 28S rDNA probe has been ampli¢ed according to Mandrioli et al. (1999).
Probes for digoxigenin-and biotin-labeling were performed by random priming following Roche protocols. Fluorescent in-situ hybridization (FISH) was carried out as described by Mandrioli et al. (1999) making two stringency washes in 0.1ÂSSC for 10 min at room temperature and two in 0.1ÂSSC at 45 ‡C for 10min after hybridization.
DNA extraction from cultured hemocytes has been performed using a standard protocol as described in Mandrioli et al. (1999), whereas Southern blotting, restriction enzyme digestion and Ba131 assay have been carried out following Okazaki et al. (1993).
Finally, RNA extraction and RT-PCR have been performed with the 'SV Total RNA Isolation System' (Promega) and with the 'Access RT-PCR System' (Promega), respectively, according to supplier's suggestions.
Pheromone-binding protein genes (PBP) were ampli¢ed using the primers F PBP (5 0 -ACTAGCATAGCAATTCGG) and R PBP (5 0 -GCCTTGACCATTACGGACG). This ampli-¢cation has been used as control of the RNA samples in RT-PCR experiments since these genes are expressed in the antennal tissues only and not in the M. brassicae cell line used for our experiments (Maibeche-Coisne et al. 1998).

Results
As, in insects, three different telomeric organizations have been observed, the presence of each telomere type was looked for in M. brassicae.
Southern blotting with the (TTAGG) n probe on M. brassicae DNA showed a diffuse smear of the hybridization (Figure 1). The hybridization of the same ¢lter with the (TTAGGG) n and (TTAGGGG) n probes did not show any hybridization signal even if weak stringency conditions were used. Analogous negative results have been obtained after Southern blotting with the Het-A, TART and the 197-bp Chironomus telomeric repeat.
These results, as a whole, suggested that the (TTAGG) n repeat could represent M. brassicae telomeres.
FISH experiments with the telomeric (TTAGG) n probe revealed bright £uorescent spots on telomeres of all chromosomes (Figure 2b, c, e), indicating that such sequences constituted M. brassicae telomeres.
In M. brassicae interphase nuclei, telomeric sequences were not clustered into a few foci but they appeared to be located near the nuclear periphery ( Figure 2a).
To determine whether the (TTAGGA) n repeats were located at the ends of M. brassicae chromosomes or below a true telomere, Mamestra DNA was digested with Bal31 for increasing amounts of time and then it has been digested with HindIII ( Figure 3a). After increasing time of Bal31 digestion, the diffuse smear of hybridization detected by the (TTAGG) n probe became progressively shorter and almost disappeared within 60 min of digestion. At high molecular weight, a strong-labeled band has been demonstrated even at 60 min digestion ( Figure 3a). This band did not appear to be heavily digested by Bal31, suggesting that it could represent non-telomeric (TTAGG) n sequence as previously reported in other organisms (Salvadori et al. 1995).
These results, as a whole, suggest that most of the (TAGGG) n repeats are located at telomeres.
The effects of the digestion with the exonuclease Bal31 have been quanti¢ed by a densitometric analysis and they have been compared to the effects of Bal31 digestion on a l DNA, according to Okasaki et al. (1993). In particular, the length of l DNA at each digestion time has been plotted against the results on the M. brassicae genome. In the Bal31 experiments, almost all the M. brassicae telomeric signals disappeared at the 2,3 rate of l DNA digestion (Figure 4). Since the digestion rate of the telomeric DNA with Bal31 is thought to be three to four times faster than that of non-telomeric DNA (Okazaki et al. 1993), the TTAGG terminal array of M. brassicae should be considered to be about 7^9 kb long.
In order to better analyze the structure of the M. brassicae telomeres, attention has been paid to the subtelorneric regions. In particular, the presence of the transposable element TRAS1 has been assessed by PCR ( Figure 6). The 300-bp TRAS1 fragment obtained by PCR has been labeled and was used as probes for both in-situ ( Figure  2d) and ¢lter hybridization (Figure 7). Southern blotting on M. brassicae DNA showed several positive bands indicating the presence of numerous TRAS1 copies. In particular, the comparison of the restriction patterns of MspI and HpaII did not reveal any difference in the hybridization signals, suggesting the absence of methylation in TRAS1.
FISH experiments with a digoxigenin-labeled TRAS1 (Figure 2d) probe and a biotinylated (TTAGG) n probe ( Figure 2e) showed bright £uorescent signals at both ends of all M. brassicae chromosomes with the (TTAGG) n probe, whereas just autosomes evidenced both £uorescent telomeres with TRAS1 probe. Only one telomere of each sex chromosome was in fact labeled with the TRAS1 probe. Silver staining of the same metaphases indicated that TRAS1 was absent from the NOR-bearing telomeres of the Z and W chromosomes (Figure 2f ).
Identi¢cation of M. brassicae sex chromosomes has been performed after karyotype reconstruction of DAPI-stained chromosomes ( Figure 5). In particular, the M. brassicae complement consists of 11 chromosome pairs, including a couple of heteromorphic sex chromosomes that were recognized as Z and W chromosomes on the basis of the data published by Traut & Marec (1996), indicating the presence of female heterogamety in M. brassicae.
DAPI staining of C-banded chromosomes demonstrated that most M. brassicae heterochromatin possesses AT-rich DNA, as evident after DAPI staining (Figure 2g), whereas only the C-band located on one telomere of each sex chromosome contained CMA 3 -positive GC-rich DNA ( Figure 2h). The bright CMA 3 £uorescence of one telomere of each sex chromosome was due to rDNA genes, as demonstrated by FISH with a 28S rDNA probe (Figure 2i).
Moreover, C banding (Figure 2g, h) showed that heterochromatin was not present on all chromosomes of the M. brassicae complement, but that it was located on the two sex chromosomes and on chromosomes 7 and 11 only. In particular, Z and W chromosomes appeared to be essentially heterochromatic with the exception of a central portion of both chromosomes. In view of these data, there was not any equilocal distribution of heterochromatin in the M. brassicae complement. Figure 2. Interphase nuclei (a) and metaphase chromosomes of M. brassicae (b^i) after FISH with the (TTAGG) n probe (a^c) showed that this repeat constitutes M. brassicae telomere. FISH on the same metaphase plate with the TRAS1 probe (d) and with the (TTAGG) n probe (e) indicated that TRAS1 is absent from NORs as deduced by silver staining (f). DAPI (g) and CMA 3 (h) staining of C-banded chromosomes indicated that most heterochromatin is AT rich with the exception of NORs that appeared GC rich and £uorescent after FISH with the 28S probe. Arrows indicate NORs. Bar corresponds to 10 mm.
Finally, only the two sex chromosomes and chromosomes 7 and 11 showed heterochromatic telomeres.
In view of such data, the presence of TRAS1 was not limited to heterochromatin since TRAS1 was present in almost all the telomeres whilst onlya few chromosomal ends were heterochromatic.
In order to evaluate if TRAS1 was located at telomeres or within the subtelomeric regions, Bal31 assays were performed (Figure 3b). Filter hybridization on Bal31-digested DNA showed that M. brassicae TRAS1 elements were partially digested only after long digestion times. TRAS1 was therefore located in the subtelomeric regions of M. brassicae chromosomes.   Furthermore, RT-PCR experiments were performed with the previously described primers evidencing a 300-bp portion of TRAS1 ( Figure  6). RT-PCR experiments suggested, therefore, that TRAS1 was transcribed and therefore it could be actively transposing in the M. brassicae genome. The presence of DNA contaminants in the RNA samples used for RT-PCR has been evaluated using the PBP primers. PBP genes are intronless coding sequences that are not expressed in the M. brassicae cell line used for these experiments. The absence of a PBP band after PCR strongly suggested that no DNA contaminants were present in the RNA sample. Therefore RT-PCR results were due exclusively to TRASl RNA.

Discussion
As three different telomere structures have been observed in insects (Biessmann et al. 1990, Okazaki et al. 1993, Zhang et al. 1994, Sahara et al. 1999, the presence of each telomeric type was investigated in M. brassicae. Southern blotting indicated that the (TTAGG) n telomeric sequence was present in the M. brassicae genome, whereas no signal has ever been observed after hybridization with the (TTAGGG) n and (TTAGGGG) n probes, with the Het-A, TART or Chironomus telomeric repeat probes.
FISH results with the (TTAGG) n repeat are consistent with those reported in numerous insects (Okazaki et al., 1993, Sahara et al. 1999, suggesting that the (TTAGG) n sequence is a component of the M. brassicae telomeres. The absence of interstitial hybridization signals might indicate that the (TTAGG) n repeat is present in a very low copy number in interstitial locations that are not detectable by FISH or, alternatively, that no chromosomal rearrangements (such as chromosome fusions) have occurred recently in the M. brassicae complement.
The present data, as a whole, suggested therefore that most of the TTAGG repeats are restric-  This hypothesis has been con¢rmed by Bal31 digestion that allowed us not only to de¢nitively locate the telomeric repeats, but also to estimate the telomere length (about 7^9 kb).
In the interphase nuclei of most organisms, the telomeric regions are situated in an ordered fashion with an association to the nuclear matrix and clustering at least at some stage of cell life (Palladino et al. 1993, Luderus et al. 1996, Pryde et al. 1997. In M. brassicae interphase nuclei, telomeres were not clustered into a few foci but they appeared to be located near the nuclear periphery as reported in D. melanogaster (Hochstrasser et al. 1986). In view of such data, it appears evident that, even if telomeric structure and function are conserved in eukaryotes, the positioning of telomeres in the interphase nuclei seems to vary among species.
Southern blotting, FISH and Bal31 assay performed on M. brassicae indicated the presence of the retrotransposon TRAS1. Whereas most retroelements are dispersed through the chromosome complement, TRAS1 appeared preferentially inserted into the telomeric regions of M. brassicae chromosomes. This peculiar chromosomal location, already reported in B. mori (Okazaki et al. 1995), strongly indicated that TRAS1 has a high speci¢city of insertion.
However, TRAS1 was not present in all M. brassicae telomeres since it was absent from the NOR-bearing telomeres of both sex chromosomes. Small subtelomeric regions in the NOR-bearing telomeres of sex chromosomes have already been reported in other insects having holocentric chromosomes but its functional meaning is still unclear (Spence et al. 1998). As a possible function of the subtelomeric repetitive DNA is to facilitate chromosome pairing in meiosis (Loidl 1990), the presence of a large subtelomeric region could make pairing between sex chromosomes more dif¢cult. This hypothesis would be particularly plausible if M. brassicae rDNA were involved in sex chromosome pairing as already reported in other insects (Mckee & Karpen 1990, Mandrioli et al. 2000. Moreover, Bal31 digestion revealed that TRAS1 was located in the subtelomeric region of all the chromosomes and not at the telomeres, as reported for the mobile elements Het-A and TART in D. melanogaster (Pardue & DeBaryshe 1999).
The RT-PCR experiments and the analysis of the methylation status indicated that TRAS1 may be actively transcribed and unmethylated. In this regard, several authors have reported that transposable elements are frequently methylated in order to prevent transposition that would otherwise be expected to damage the genome (Arnaud et al. 2000). The absence of methylation in TRAS1 could suggest a high frequency of transposition of this mobile DNA into the lepidopteran genome or that methylation is not involved in transposon control in Lepidoptera.
C-banding applied to M. brassicae mitotic metaphases revealed that heterochromatin is essentially linked to the sex chromosomes, whereas autosomes contain only faint bands. The tendency to accumulate heterochromatin on the sex chromosomes has been repeatedly described in insects with holocentric chromosomes (e.g. Mandrioli et al. 1999). This fact substantially differs from what is observed in monocentric chromosomes, where heterochromatic regions typically occupy speci¢c zones of all chromosomes, corresponding to centromeres and sometimes telomeres (Schweizer & Loidl 1987).
Finally, only telomeres of chromosomes Z, W, 7, and 11 appeared to be heterochromatic. These data suggest therefore that TRAS1 is associated with the M. brassicae telomere independently of the presence of heterochromatin at telomeres.
The interest in telomeres of holocentric chromosomes is not limited to the comparison of their architecture with that of monocentric ones, but it is emphasized by their peculiar structure which allows karyotype rearrangements due to chromosomal fragmentation (Okazaki et al. 1993). All fragments produced by X-ray irradiation will be in fact conserved in the chromosomal set in view of a centromeric activity in each fragment (Hughes-Schrader & Schrader 1961, White 1973. Therefore, on the basis of such properties and in view of the presence of both in-vivo and in-vitro analysis tools, M. brassicae holocentric chromosomes appear to be an intriguing model for the study of telomere dynamics and in particular of de-novo telomere formation at breakage sites.