FISH Mapping of Telomeric and Non-Telomeric (AG3T3)3 Reveal the Chromosome Numbers and Chromosome Rearrangements of 41 Woody Plants

Data for the chromosomal FISH mapping localization of (AG3T3)3 are compiled for 37 species belonging 27 families; for 24 species and 14 families, this is the first such report. The chromosome number and length ranged from 14–136 and 0.56–14.48 μm, respectively. A total of 23 woody plants presented chromosome length less than 3 μm, thus belonging to the small chromosome group. Telomeric signals were observed at each chromosome terminus in 38 plants (90.5%) and were absent at several chromosome termini in only four woody plants (9.5%). Non-telomeric signals were observed in the chromosomes of 23 plants (54.8%); in particular, abundant non-telomeric (AG3T3)3 was obviously observed in Chimonanthus campanulatus. Telomeric signals outside of the chromosome were observed in 11 woody plants (26.2%). Overall, ten (AG3T3)3 signal pattern types were determined, indicating the complex genome architecture of the 37 considered species. The variation in signal pattern was likely due to chromosome deletion, duplication, inversion, and translocation. In addition, large primary constriction was observed in some species, probably due to or leading to chromosome breakage and the formation of new chromosomes. The presented results will guide further research focused on determining the chromosome number and disclosing chromosome rearrangements of woody plants.

Telomeric repeats are not often localized at chromosomal termini and have also been found in multiple intercalary sites of chromosomes in many species [3,[23][24][25]. Interstitial telomeric sequences may represent a significant part of the telomeric DNA [3]. Cytogenetic analysis has found two major types of interstitial telomeric sequences: one is heterochromatic and large, found in centromeric or pericentromeric regions, while the other type is short and distributed at various positions in chromosomes [3]. Unlike telomeric repeats located at termini, interstitial telomeric sequences confer karyotype plasticity. Their unstable and high-length polymorphisms may increase chromosomal fragility and contribute to chromosomal reorganization [3,23,24]. Interstitial telomeric sequences have been shown to cause a change in the chromosome number in Cardamine cordifolia A. Gray [26], as well

Materials and Methods
The species chosen for these experiments were firstly considered due to the occurrence of karyotype rearrangements [11,[47][48][49][50][51][52][53], and secondly for investigation of species in which (AG 3 T 3 ) 3 has not yet been explored. Zea mays L. was chosen as it possesses the telomeric repetitive unit AGGGTTT conserved in plant chromosome telomeres. It is contained in the sequence named M8-2D, a B chromosome-specific sequence in Z. mays, which has low homology to clones from Z. mays chromosome 4 centromere. M8-2D is localized in B chromosome centromeric and telomeric regions [53]. Hence, we used Z. mays to test the used probe and as a control.
Details of the seeds or seedlings of Z. mays and the 41 woody plants (belonging to 37 species, 27 genera, 18 families) used in the present work are provided in Table 1. All 42 plants were collected from 12 Counties or Districts of Sichuan Province, China.
Collected seeds were germinated in culture dishes with wet filter paper and kept at 25 • C in the daytime and at 18 • C in the night until the roots were~2 cm in length, which were then cut. The collected seedlings were cultured in soil at room temperature (15-25 • C) until many new roots grew out, which were then cut again. The cut roots were treated with nitrous oxide (N 2 O) gas for 2-6 h, with treatment time depending on chromosome length and cell wall lignification. Next, the samples were fixed in glacial acetic acid for 5-10 min, then kept in 75% ethyl alcohol. Chromosome preparation was carried out according to the procedure described by Luo et al. [51]. As these techniques have been described elsewhere, they will be detailed briefly here. Approximately 1 mm of the meristematic zone of the root tip was enzymolyzed at 37 • C for 45 min by using cellulase and pectinase (1 mL buffer + 0.04 g cellulase + 0.02 g pectinase, the buffer 50 mL was included 0.5707 g trisodium citrate + 0.4324 g citric acid), which were produced by Yakult Pharmaceutical Ind. Co., Ltd. (Tokyo, Japan) and Kyowa Chemical Products Co., Ltd. (Osaka, Japan), and then mixed into suspension for dropping onto slides. These slides were air dried then examined using an Olympus CX23 microscope (Olympus, Tokyo, Japan).

Karyotype Analysis
Raw images were processed using the DP Manager (Olympus Corporation, Tokyo, Japan) and Photoshop CC 2015 (Adobe Systems Incorporated, San Jose, CA, USA) software. At least ten slides of each plant were observed, and at least fifteen cells with good chromosome spread were used for chromosome counting and length measurement. All chromosomes were aligned from longest to shortest. The chromosome ratio was determined by comparing the length of the longest chromosome to that of the shortest chromosome. Further karyotype analysis could not be carried out due to the small chromosome size and unclear centromere location of many of the species.

Karyotype Analysis Revealed Differences among 37 Species
For 24 species and 14 families, this is the first time that (AG 3 T 3 ) 3 testing has been reported. To visualize the chromosomal distribution of (AG 3            The chromosome number and length for the considered species were sorted in Table 2. The chromosome number in the 42 plants ranged from 14 (C. chinensis, A3) to 136 (Z. bungeanum, C15). A total of 14 woody plants possessed 24 chromosomes (one third), whereas seven woody plants possessed 22 chromosomes (one sixth). The longest chromosome length of each plant ranged from 1.12 µm (K. paniculata, C1) to 14.48 µm (C. revoluta, A10), while the shortest chromosome length of each plant ranged from 0.56 µm (K. paniculate, C1) to 8.06 µm (C. revoluta, A10). A total of 23 woody plants (nearly one third) had chromosome length less than 3 µm, thus falling into the small chromosome category. Due to the indistinct location of centromeres and small size of chromosomes in many of the considered woody plants, further karyotype analysis-such as long/short arm length and karyotype formula-was not carried out. Karyotype asymmetry was assessed using the ratio of longest to shortest chromosome length. The largest ratio was 4.28 in C. funebris (A7), while the smallest ratio was 1.12 in C. fortunei (A9). The ratio for 17 plants ranged from 1 to 2 (40.5%), while that of 19 plants ranged from 2 to 3 (45.2%). The ratio was greater than 3 for six plants: C. funebris (A7), E. lanceolata (B12), P. americana (B13), J. regia 'Chuanzao1' (C2), J. sigillata 'Muzhilinhe' (C5), and P. macrophyllus (C6). These results indicated that abundant differences exist among 37 of the considered species.

The Diverse Signal Patterns of (AG 3 T 3 ) 3 Reveal the Complex Genome Architecture
To better investigate diversity of (AG 3       chromosome terminus in (B1,B3-B16), while telomeric signals were absent at several chromosome termini in (B2). Non-telomeric signals were observed at several chromosome termini in (B1-B5,B7-B11). Telomeric signals deviated from the chromosome in (B13,B14).  four in Oleaceae (orange), three in Fabaceae (pink), three in Lauraceae (light blue), tw Malvaceae (dark green), two in Rutaceae (grey), and one in each of Berberidaceae, C canthaceae, Cycadaceae, Euphorbiaceae, Fagaceae, Poaceae, Podocarpaceae, Salicac and Sapindaceae, respectively. Except for Lauraceae and Rutaceae, which each presen a single signal pattern type, the other families (Elaeagnaceae, Cupressaceae, Juglandac Oleaceae, Lauraceae, and Malvaceae) all presented at least two signal pattern types. As shown in Figure 10, there were ten (AG3T3)3 signal pattern types in total: Typ chromosome only includes signal at both ends; Type II, chromosome not only inclu As shown in Figure 10, there were ten (AG 3 T 3 ) 3 signal pattern types in total: Type I, chromosome only includes signal at both ends; Type II, chromosome not only includes signal at both ends but also includes a non-telomeric signal location; Type III, chromosome includes single end signal, and the other telomeric signals outside of the chromosome; Type IV, chromosome not only includes signal at both ends, but also includes telomeric signal deviating from chromosome; Type V, chromosome not only includes signal at both ends, but also includes a large primary constriction; Type VI: chromosome not only includes signal at both ends signal, but also includes a large primary constriction, as well as a nontelomeric signal location; Type VII, chromosome only includes telomeric signal outside of the chromosome; Type VIII, chromosome only includes single end signal; Type IX, chromosome only includes non-telomeric signal location; and Type X, chromosome includes no signals. These types of signal pattern indicate that there is an abundant diversity in (AG 3 T 3 ) 3 signal arrangement.
All 42 plants possessed the 12 signal pattern types or type combinations shown in Figure 10. Ten woody plants only possessed signal pattern type I; Z. mays and 11 woody plants possessed the combination of type I + type II; eight woody plants possessed the combination of type I + type III; P. macrophyllus possessed the combination of type I + type II + type III; C. campanulatus possessed the combination of type I + type II + type IV; J. sigillata 'Muzhilinhe' possessed the combination of type I + type II + type III + type IV; B. diaphana possessed the combination of type I + type V; T. media and T. yunnanensis possessed the combination of type I + type II + type V; T. chinensis, T. cuspidata, and T. wallichiana possessed the combination of type I + type II + type VI; K. paniculate possessed the combination of type I + type III + type VII + type VIII; C. revolute possessed the combination of type VIII + type IX; and C. funebris possessed the combination of type VIII + type IX + type X.
There were diverse signal patterns of (AG 3 T 3 ) 3 among 37 species, indicating a complex genome architecture. For example, considering (i) Elaeagnaceae, five plants of H. rhamnoides possessed type I + type II, but E. lanceolata possessed type I; (ii) in Taxaceae, T. media and T. yunnanensis possessed the combination type I + type II + type V, but T. chinensis, T. cuspidata, and T. wallichiana possessed the combination type I + type II + type VI; (iii) in Cupressaceae, C. fortune, C. japonica, and P. orientalis possessed type I, but C. funebris possessed the combination type VIII + type IX + type X; (iv) in Juglandaceae, J. regia 'Chuanzao1', J. regia 'Yanyuanzao', and J. sigillata 'Maerkang' possessed type I, but J. sigillata 'Muzhilinhe' possessed the combination type I + type II + type III + type IV; (v) in Oleaceae, L. lucidum possessed the combination type I + type II, L. × vicaryi and S. oblata possessed the combination type I + type III, and F. pennsylvanica possessed the combination type I + type II + type III; (vi) in Fabaceae, C. chinensis possessed type I, R. pseudoacacia possessed the combination type I + type II, and E. crista-galli possessed the combination type I + type III; and (vii) in Malvaceae, H. mutabilis possessed the combination type I + type II, but F. simplex possessed the combination type I + type III.
The number shown after the species name in Figure 10 represents the ratio of longest to shortest chromosome length, indicating karyotype asymmetry. Type I included ten plants with ratio ranging from 1.12-4.13, with variance of 0.94. Type I + Type II included 12 plants with ratio ranging from 1.39-3.65, with variance of 0.38. Type I + Type II + Type III included eight plants with ratio ranging from 1.34-3.46, with variance of 0.48. These results indicate that chromosomes with conserved telomeric signal (Type I), in fact, have a wider range of karyotype asymmetry (VAR 0.94), while chromosomes with non-telomeric signals (Type II, III) have relatively concentrated karyotype asymmetry (VAR 0.38 and 0.48, respectively). In addition, no correlation between non-telomeric signals and karyotype asymmetry was observed.

Proposed Origin of (AG 3 T 3 ) 3 Signal Diversity
Based on Figures 7-10, the proposed origin of (AG 3 T 3 ) 3 signal diversity is illustrated in Figure 11. There are three major groups. (i) Signal number: Increase signal number is likely caused by chromosome duplication, inversion, translocation, and/or sequence changes; a constant signal number indicates chromosome conservation and a decreased signal number is likely caused by chromosome deletion and/or sequence changes. (ii) Signal location: End signals on chromosomes indicate chromosome conservation; non-end signals are likely caused by chromosome deletion, duplication, inversion, translocation, and/or sequence changes; end signals deviating from the chromosome are probably caused by chromosome satellites, while end signal loss is probably caused by chromosome end deletion and/or sequence changes. (iii) Primary constriction: Normal primary constriction indicates chromosome conservation, while primary constriction likely becomes large due to chromosome breakage and the formation of new chromosomes (e.g., in Figure 8, T. media chromosome 9 and B. diaphana chromosome 9, T. cuspidata chromosome 9 and T. wallichiana chromosome 9 possibly indicate the formation of new chromosomes, while T. cuspidata chromosome 12 and T. wallichiana chromosome 12 were possibly formed in a reverse manner). signals are likely caused by chromosome deletion, duplication, inversion, translocation, and/or sequence changes; end signals deviating from the chromosome are probably caused by chromosome satellites, while end signal loss is probably caused by chromosome end deletion and/or sequence changes. (iii) Primary constriction: Normal primary constriction indicates chromosome conservation, while primary constriction likely becomes large due to chromosome breakage and the formation of new chromosomes (e.g., in Figure 8, T. media chromosome 9 and B. diaphana chromosome 9, T. cuspidata chromosome 9 and T. wallichiana chromosome 9 possibly indicate the formation of new chromosomes, while T. cuspidata chromosome 12 and T. wallichiana chromosome 12 were possibly formed in a reverse manner).
In brief, the variations in signal number and signal location were probably caused by chromosome deletion, duplication, inversion, translocation, and sequence changes, as well as chromosome satellites. It is likely that large primary constriction was due to chromosome breakage and the formation of new chromosomes. Figure 11. Proposed origin of (AG3T3)3 signal diversity. The variations in signal number and signal location were likely caused by chromosome deletion, duplication, inversion, translocation, as Figure 11. Proposed origin of (AG 3 T 3 ) 3 signal diversity. The variations in signal number and signal location were likely caused by chromosome deletion, duplication, inversion, translocation, as sequence changes, as well as chromosome satellites, while primary constriction became large due to chromosome breakage and the formation of new chromosomes.
In brief, the variations in signal number and signal location were probably caused by chromosome deletion, duplication, inversion, translocation, and sequence changes, as well as chromosome satellites. It is likely that large primary constriction was due to chromosome breakage and the formation of new chromosomes.

Karyotype Analysis of Z. mays and 41 Woody Plants
Chromosome number, size, centromere location, long/short arm ratio, and satellites are basic characteristics of karyotype. The presence of telomeric (AG 3 T 3 ) 3 in both chromosome ends may guarantee the accuracy of chromosome counting. The chromosome number in the 42 plants ranged from 14 (C. chinensis) to 136 (Z. bungeanum). There were six woody plants that presented chromosome numbers different to that reported in previous studies: J. regia 'Chuanzao1', J. regia 'Yanyuanzao', J. sigillata 'Maerkang', and J. sigillata 'Muzhilinhe' presented 2n = 34, a result supported by the work of Luo and Chen [48] but contradicted by Mu and Xi [54] and Mu et al. [55] (who reported 2n = 32). The differences were probably caused by hybridization, aneuploidization/among-population variation, or inaccurate chromosome number counts. P. macrophyllus presented 2n = 36, a result supported by the work of Hizume et al. [56] but contradicted by Zhu et al. [57] for eight small chromosomes, which were treated as satellite chromosomes in the latter. Z. armatum 'Jinyang Qinghuajiao' presented 2n = 98 in this study, which is contradicted by the work of Luo et al. [58], who reported 2n =~128 for another variety, Z. armatum 'Hanyuan Putao Qingjiao'. After excluding the experimental count error, we infer this big difference to have been caused by the confusion and complexity between Z. armatum varieties. We tested 16 Z. armatum varieties by FISH and SSR in another study in order to address this issue. Another reason why the chromosome numbers differed from previous studies is because of the limited available research on the chromosome numbers of woody plants. Compared to herbaceous plants, it is more difficult to obtain their chromosome preparation due to root lignification.
A total of 23 woody plants (nearly one third) had chromosome lengths less than 3 µm, thus belonging to the small chromosome group. Due to the indistinct location of the centromere and the small size of chromosomes in many of the woody plants, as well as the chromosome length fluctuating according to the chromosome division phase and measuring tool, further karyotype analysis (e.g., long/short arm length and karyotype formula) was not carried out. Chromosome length is also not discussed further.

Occurrence of (AG 3 T 3 ) 3 in Woody Plants
AG 3 T 3 is a telomeric repetitive tandem that is conserved in higher plant chromosome telomeres. This oligos is included in a B chromosome-specific sequence, named M8-2D, in Z. mays [53]. M8-2D has shown low homology to sequences from the chromosome 4 centromere in Z. mays but has been detected in B chromosome centromeric and telomeric (2n = 90), Z. armatum (2n = 98), and Z. bungeanum (2n = 136), presented interstitial telomeric signals, all having a chromosome number 2n > 42. It is possible that high chromosome number species have not been used as frequently in previous studies. The third class (26.2% of plants) presented telomeric signal deviating from the chromosome. C. campanulatus, F. pennsylvanica, L. × vicaryi, and S. oblata showed this type of signal, as supported by the results of Luo and Liu [49] and Luo and Chen [52]. J. sigillata 'Muzhilinhe' also presented this type of signal, in contrast to the results of Luo and Chen [48]. The probable reasons for this discrepancy are similar to those for the second class.
Unusual telomere sequences described by non-telomeric signals are, in many cases, connected with high C-values [63]; for example, in species of Cestrum and Allium [64]. In the present study, Z. mays, Z. armatum, and Z. bungeanum presented non-telomeric signals along with giant C-values (Zea, 3.8 pg; Zanthoxylum, 4.57 pg) [65] (https://cvalues.science. kew.org/, accessed on 17 May 2022). Nevertheless, the small C-values of Juglans (0.64 pg), Robinia (0.74 pg), and Chimonanthus (0.86 pg) were also accompanied by non-telomeric signals (C. campanulatus, J. sigillata 'Muzhilinhe', R. pseudoacacia); especially for C. campanulatus, which presented highly diverse non-telomeric signals. Hence, to the best of our knowledge, there is no correlation between the presence of non-telomeric signals and the C-value in woody plants; this result agrees with the conclusion of Gorelick et al. [66].

Proposed Origin of (AG 3 T 3 ) 3 Diversity in Woody Plants
Telomeric sequences are not found exclusively at chromosome termini, but also in nonterminal sites of chromosomes in many species [23][24][25]. Interstitial telomeric sequences may represent a significant part of telomeric DNA. Bolzán [3] has revealed two major types of interstitial telomeric sequences: one is heterochromatic and is largely observed in centromeric or pericentromeric regions (e.g., in this study, C. campanulatus, five species of Taxus, and five plants of H. rhamnoides). The other type is short and distributed at various sites throughout chromosomes, such as in J. sigillata 'Muzhilinhe' and R. pseudoacacia.
Interstitial telomeric sequences could be considered the result of chromosomal rearrangements [23,24,67,68]. Messier et al. [69] explained interstitial telomeric sequences through the creation of a small number of repeats by random mutations followed by repeat expansion towards two flanks. In the present work, the non-terminal (AG 3 T 3 ) 3 signals were likely caused by chromosome deletion, duplication, inversion, translocation, and sequence changes, all of which are chromosomal reorganizations. Previous researchers have hypothesized that interstitial telomeric sequences in the heterochromatic region could be the trace of the chromosome end fusion, causing descendent hypoploidy (a decrease in the chromosome number) [28][29][30]. In the present work, we observed large primary constriction in chromosome 9 of B. diaphana and five species of Taxus. We may infer that these chromosomes with large primary constriction will possibly lead to ascent hyperploidyc. Thus, in the case of large primary constriction with non-end (AG 3 T 3 ) 3 signal, such as T. cuspidate chromosome 9 or T. wallichiana chromosome 9, ascent hyperploidy may occur. Similarly, we may also infer that the telocentric chromosome 12 in five species of Taxus was possibly formed in a reverse manner. However, in the case of large primary constriction with no signal, such as that observed in T. media chromosome 9 and B. diaphana chromosome 9, chromosome breakage is likely unrelated to interstitial (AG 3 T 3 ) 3 .

Conclusions
In this paper, we examined Z. mays and 41 woody plants, established FISH physical mapping, described the diverse distribution of (AG 3 T 3 ) 3 , and disclosed the complex genome architecture of woody plants. We inferred that the observed non-telomeric signals were probably caused by chromosome arrangements. We intend to continue our research by testing more woody plants, such as species of Calycanthaceae, to explore the abundant non-telomeric (AG 3 T 3 ) 3 , as well as Rutaceae, to determine the chromosome number.