Interstitial Arabidopsis-Type Telomeric Repeats in Asteraceae

Tandem repeats of telomeric-like motifs at intra-chromosomal regions, known as interstitial telomeric repeats (ITR), have drawn attention as potential markers of structural changes, which might convey information about evolutionary relationships if preserved through time. Building on our previous work that reported outstanding ITR polymorphisms in the genus Anacyclus, we undertook a survey across 132 Asteraceae species, focusing on the six most speciose subfamilies and considering all the ITR data published to date. The goal was to assess whether the presence, site number, and chromosomal location of ITRs convey any phylogenetic signal. We conducted fluorescent in situ hybridization (FISH) using an Arabidopsis-type telomeric sequence as a probe on karyotypes obtained from mitotic chromosomes. FISH signals of ITR sites were detected in species of subfamilies Asteroideae, Carduoideae, Cichorioideae, Gymnarhenoideae, and Mutisioideae, but not in Barnadesioideae. Although six small subfamilies have not yet been sampled, altogether, our results suggest that the dynamics of ITR formation in Asteraceae cannot accurately trace the complex karyological evolution that occurred since the early diversification of this family. Thus, ITRs do not convey a reliable signal at deep or shallow phylogenetic levels and cannot help to delimitate taxonomic categories, a conclusion that might also hold for other important families such as Fabaceae.


Introduction
Telomeres are the natural ends of eukaryotic linear chromosomes. They are constituted by ribonucleoprotein complexes that differ notably from other DNA sequences in both structure and function. Capping the chromosome ends by telomere structures facilitates the protection of genetic material against double-stranded breaks, degradation, and end-to-end fusion with other chromosomes, which lead to genome instability [1][2][3].
Telomeric regions are constituted by large stretches of DNA sequences that are GC rich and are usually tandemly arranged to attain up to several thousands of base pairs in length [4]. Overall, telomeric sequences are highly conserved across groups of unrelated organisms although lineage specific motifs are known too [5,6]. In land plants, the dominant consensus telomere repeat is composed of seven nucleotides known as the Arabidopsis-type (TTTAGGG) n [6][7][8], but divergent repeat motifs have been recognized [9][10][11][12][13][14][15][16][17]. It has been recently suggested that beyond the telomere motif exceptions detected, the real diversity in telomeric sequences in land plants has been probably underestimated [18].
Tandem repeats of telomeric-like motifs have been reported also at intrachromosomal regions, known as interstitial telomeric repeats (ITR) or interstitial telomeric sequences [19]. The use of molecular cytogenetic techniques has allowed the detection of ITRs in unrelated organisms [7,8,[20][21][22][23][24]. Overall, these studies have interpreted the occurrence of ITRs as stable karyological landmarks that they may be remnants of end-to-end fusion events between non-homologous chromosomes, translocations, inversions, and heterologous       In the studied accessions showing ITRs, the number of sites ranged from two (Anthemis cotula, Tanacetum vulgare, Leucanthemum graminifolium, Senecio vulgaris, Dymondia margaretae, Sonchus tenerrimus subsp. tenerrimus, and Leontodon tuberosus) to 52 in Cota nigellifolia, the highest number so far reported in Asteraceae. Intrageneric differences in the presence or absence of ITR signals were observed in Achillea (two absences, one presence), Anthemis (three absences, two presences), Cladanthus (one absence, one presence), Nassauvia (one absence, one presence), and Sonchus (one absence, one presence) accessions. Intraspecific variation was detected in a single species, Sonchus tenerrimus. Two ITR sites were present in subsp. tenerrimus whereas none were found in the closely related subsp. dianae.
ITR sites preferentially occur as homozygous loci (21 species). However, odd numbers of ITR sites reflecting a hemyzygous state were detected in Gonospermum fruticosum (29 sites), Achillea ligustica (17 sites), and Anthemis tinctoria (3 sites). The intensity of the ITR signals detected by FISH varied between site locations and samples, suggesting that a variable number of repeats are involved ( Figure 4). For instance, faint ITR signals (showing lower intensity than the telomeric sites) were detected in Stokesia laevis ( Figure 3A). In the studied accessions showing ITRs, the number of sites ranged from two (Anthemis cotula, Tanacetum vulgare, Leucanthemum graminifolium, Senecio vulgaris, Dymondia margaretae, Sonchus tenerrimus subsp. tenerrimus, and Leontodon tuberosus) to 52 in Cota nigellifolia, the highest number so far reported in Asteraceae. Intrageneric differences in the presence or absence of ITR signals were observed in Achillea (two absences, one presence), Anthemis (three absences, two presences), Cladanthus (one absence, one presence), Nassauvia (one absence, one presence), and Sonchus (one absence, one presence) accessions. Intraspecific variation was detected in a single species, Sonchus tenerrimus. Two ITR sites were present in subsp. tenerrimus whereas none were found in the closely related subsp. dianae.
ITR sites preferentially occur as homozygous loci (21 species). However, odd numbers of ITR sites reflecting a hemyzygous state were detected in Gonospermum fruticosum Plants 2021, 10, 2794 5 of 14 (29 sites), Achillea ligustica (17 sites), and Anthemis tinctoria (3 sites). The intensity of the ITR signals detected by FISH varied between site locations and samples, suggesting that a variable number of repeats are involved ( Figure 4). For instance, faint ITR signals (showing lower intensity than the telomeric sites) were detected in Stokesia laevis ( Figure 3A). The opposite was detected in Hyoseris taurina where the centromeric ITR sites showed a remarkably higher intensity than those present at the chromosome ends ( Figure 3C).
Altogether 176 species and three subspecies included in 115 genera have been analyzed for ITR. Of these, ITR signals have been detected in 36 species (20.11%) and 25 (21.74%) genera. The phylogenetic distribution of the species showing ITRs is shown in Figure 5 and Table 1.
Altogether 176 species and three subspecies included in 115 genera have been analyzed for ITR. Of these, ITR signals have been detected in 36 species (20.11%) and 25 (21.74%) genera. The phylogenetic distribution of the species showing ITRs is shown in Figure 5 and Table 1.  cluding the chromosome number, presence and number of ITR signals, and the somal distribution are indicated in Table S1. Altogether 176 species and three subspecies included in 115 genera have lyzed for ITR. Of these, ITR signals have been detected in 36 species (20.11 (21.74%) genera. The phylogenetic distribution of the species showing ITRs is Figure 5 and Table 1.  No association between the presence of ITRs and the evolutionary placement of the species has been evidenced. ITRs have been recorded from 11 tribes, distributed across the phylogenetic tree of the family, including both early-diverging (Nassauvieae) and more recently diverged lineages (Cardueae, Gymnarrheneae, Vernonieae, Cichorieae, Senecioneae, Anthemidae, Astereae, Tageteae, and Madieae).
ITR sites are present across a wide range of chromosome numbers, from species with 2n = 4 (Xanthisma gracile) to 2n = 68 (Porophyllum ruderale). No apparent relation was detected between the number of ITR sites and the chromosome number of the species (Pearson correlation value r = −0.05, p = 0.6213; Figure 6), even if the results are analyzed separately for each subfamily (data not shown). This is clearly observed in Asteroideae, where species showing the same chromosome number (2n = 18) exhibit the widest range of ITR sites (2-52; Figure 6). ITR sites are present across a wide range of chromosome numbers, from species with 2n = 4 (Xanthisma gracile) to 2n = 68 (Porophyllum ruderale). No apparent relation was detected between the number of ITR sites and the chromosome number of the species (Pearson correlation value r = −0.05, p = 0.6213; Figure 6), even if the results are analyzed separately for each subfamily (data not shown). This is clearly observed in Asteroideae, where species showing the same chromosome number (2n = 18) exhibit the widest range of ITR sites (2-52; Figure 6).  ITR signals have been recorded from all regions of the chromosomes (centromeric, proximal and interstitial). However, the longitudinal distribution of ITR sites along the chromosome arms is uneven and their frequency was biased towards proximal and interstitialproximal regions ( Figure 7A). Since a substantial number of records were contributed from the large sampling available from Anacyclus [49], the data were also reanalysed without the values from this genus ( Figure 7B). ITR signals have been recorded from all regions of the chromosomes (centromeric, proximal and interstitial). However, the longitudinal distribution of ITR sites along the chromosome arms is uneven and their frequency was biased towards proximal and interstitial-proximal regions ( Figure 7A). Since a substantial number of records were contributed from the large sampling available from Anacyclus [49], the data were also reanalysed without the values from this genus ( Figure 7B).  (Table S1). However, non-centromeric ITR signals were also noted to co-occur for most of the samples. Only Tanacetum parthenium and Galactites tomentosa showed exclusively centromeric ITR signals (14 and 10, respectively).

Discussion
Asteraceae, with nearly 180 analyzed species and subspecies, is the best sampled family of seed plants for the presence of ITR sites in chromosomes [53]. In this paper, we  (Table S1). However, non-centromeric ITR signals were also noted to co-occur for most of the samples. Only Tanacetum parthenium and Galactites tomentosa showed exclusively centromeric ITR signals (14 and 10, respectively).

The Long Evolutionary History of Asteraceae May Have Erased Phylogenetic Signals of ITR Sites
The perception that ITRs represent a labile chromosomal landmark in Asteraceae is evidenced in this study. Variation related to presence, site number, chromosomal location and copy number regarding the presence of ITRs is scattered across evolutionary lineages ( Figure 5). No association at high (subfamily, tribe, subtribe) or low taxonomic levels (genus, species) can be supported with our data. This is illustrated in Anthemideae, the most thoroughly sampled tribe to date (23 genera, 56 species and one subspecies) where ITR variation among congeneric species and between closely related genera is outstanding.
It should be emphasized that our survey was designed to prioritize the sampling of as many genera as possible at the expense of a low intrageneric coverage (ratio species/ genus = 1.22). This precluded the assessment of ITR variation at low taxonomic ranks and shallow evolutionary levels (species and populations). Accordingly, it is likely that this biased sampling may have even underestimated the values of ITR variation at the generic and intraspecific levels. Our data obtained in Sonchus tenerrimus, where two intraspecific variants have yielded contrasting results concerning ITRs, point in this direction. Thus, it should not be discarded that analyzing more accessions from poorly sampled higher groups, where ITRs have not been detected, may reveal additional presences and even intraand interspecific polymorphisms, as has been previously reported in Anacyclus [49] and Tanacetum [50]. However, we think that additional data would not contradict the present evidence that ITRs are not appropriate markers in tracing phylogenetic relationships and helping delimit taxonomic boundaries in Asteraceae.
Whether the presence of ITR sites is an ancestral state in Asteraceae remains an open question deserving further studies. We have refrained from exploring evolutionary trends in ITR site change by maximum-likelihood or parsimony-based reconstruction due to the scarce data available for the early-diverging nodes. Available comparative data for related families indicate absence of ITRs (Table 1), including Calyceraceae, the sister family of Asteraceae. However, sampling is anecdotal and may be hardly used for outgroup purposes.
It has been hypothesized that the origin of Asteraceae dates back to approximately 69.5 Ma in the late Cretaceous and subfamilies diverged between 64.75 and 43.25 Ma [75]. This long evolutionary history has been modulated by several episodes of genome du-Plants 2021, 10, 2794 9 of 14 plication (see below), adaptive radiation, and Cretaceous-Tertiary extinctions that have triggered diversification rate shifts [29,[75][76][77].
An accurate interpretation of karyotype changes can be compromised if chromosomal changes have occurred in ancient and complex evolutionary scenarios. The accumulation of old genomic changes (e.g., past genome duplications or paleopolyploidy, neopolyploidization, chromosome rearrangements leading to subsequent dysploidy and diploidization processes) in short periods of time can hinder elucidating the imprints left by each one of them in the evolutionary history of extant karyotypes [78]. In Asteraceae, hypotheses on chromosomal base number evolution were reported by [29]. In this work, hypothesized base numbers for each of the 36 main clades were superimposed on the backbone phylogeny provided by [79] and two main patterns of karyotype change emerged, (1) polyploidy is common in Asteraceae and occurs in most major clades and (2) descending dysploid events (in contrast to the rarer increasing dysploidy) are widespread across major lineages. Later on, [80], using explicit models of chromosome evolution, corroborated previous inferences and suggested that karyological evolution has been a very dynamic process in Asteraceae mainly shaped by polyploidy and descending dysploidy scenarios.
It can also be inferred from our previous work [49] that the deduced fast ITR turnover occurring at the population level in Anacyclus is not clearly associated to speciation events and it is likely that this finding also applies to other unrelated lineages. Thus, available evidence suggest that ITR features in Asteraceae do not convey a strong evolutionary signal to be used in deciphering phylogenetic relationships and are also of little, if any, value in helping to delimit taxonomic boundaries. On the contrary, patterns of hypervariability, such as those reported for Anacyclus, makes ITR a suitable marker for tracing genomic changes at micro-evolutionary scales using cytogenetic approaches.

Are Centromeric ITR Signals Remnants of Dysploidy in Asteraceae?
In plants, the presence of ITR in centromeric and peri-centromeric locations has been considered to represent cytological landmarks related to dysploidy events e.g., [81][82][83][84][85][86]. There are two types of chromosomal rearrangements involved in dysploidy: end-to end chromosome fusions (symmetrical translocations) and fusion-fission cycle (Robertsonian rearrangements) between non-homologous chromosomes. Thus, on theoretical grounds ITR sites could be remnants of these karyological events. However, it has been pointed out that to acquire stability after chromosomal fusion events they should be accompanied by loss or inactivation of one of the centromeres and the resulting interstitial telomere sequences [78]. This view was modulated by [83], who indicated that the interstitial telomere sequences may be preserved during the chromosomal rearrangements because the telomere region of one chromosome is the breakpoint and therefore partially participates in the reciprocal translocation between the two acrocentric or telocentric chromosomes involved in the rearrangement.
If ITR sites in Asteraceae were in fact cytogenetic signatures of disploidy, one question to be addressed is why ITR number contrasts between related species showing the same derived chromosome numbers. This is illustrated in the case of Brachyschome dichromosomatica and Xanthisma gracile, two species showing the highly reduced 2n = 4 chromosome number (the lowest chromosome number so far reported in seed plants). From the available evidence, it seems highly likely that the 2n = 4 karyotype was originated through many rounds of descendent disploidy events. Surprisingly, no ITRs have been documented in B. dichromosomatica [32], whereas in X. gracile two contrasting reports recording zero and four non-centromeric ITR sites are known [7,8,38].
Altogether, our data suggest that in Asteraceae the dynamics of ITR formation is too fast or labile to trace the complex karyological evolution that has been hypothesized to occur since the early diversification events in the family [75,80]. Further, part of such lability could be due to other types of karyological changes that erased previous ITR signs inconsistently across lineages.

Plant Materials
Seeds were collected in the field or were supplied by botanical gardens, plant breeding stations, research centres and commercial sources. The name and origin of the plant accessions used are provided in Table S2. Representative herbarium and seeds vouchers are deposited at the herbarium of the Botanical Garden of Valencia University (VAL) and the Germplasm banks of the Real Jardín Botánico de Madrid and the Botanical Garden of Valencia University. For comparative purposes, two species belonging to the related Calyceraceae and Goodeniaceae families (Asterales) were also studied.

FISH Analysis
Protocols for seed germination and the obtention of mitotic chromosomes follow the experimental procedures described in [87]. The telomeric sequence TTTAGGG was localised using the pAtT4 clone isolated from Arabidopsis thaliana [88]. The probe was labelled with biotin-16-dUTP through a nick translation procedure according to the manufacturer's protocols (Roche, Germany). Probe detection was conducted using the method of [89] with the modifications described by [90].

Karyotype Analysis
For each accession, well spread metaphase plates were selected for assessing the chromosome number and the presence of ITRs. Individuals showing ITRs were further analyzed and the number of sites and their location along the chromosome were recorded. The chromosome arm was divided into four domains of equal size according to [91], i.e., centromeric (c), proximal (p), interstitial-proximal (ip), and interstitial-terminal (it) and the ITR sites detected were mapped on these chromosomal regions. Since most karyotypes of Asteraceae show metacentric chromosomes a reliable distinction between both arms was usually not feasible. Accordingly, the locations of ITR sites on both chromosome arms were pooled. Chromosome measurements were made on digital images using the computer application MicroMeasure version 3.2 [92].
The overview of the phylogenetic relationships among tribes is based on [27]. The tree was redrawn using the iTOL application [93] and Inkscape software.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/plants10122794/s1, Table S1: Distribution of ITR sites in Asteraceae and related families of Asterales, Table S2: List of studied accessions and sources of analyzed material by FISH.