rDNA is highly repetitive and conserved in species. For plant species with a large number of chromosomes and small size, the appearance of rDNA probe provides a cytological approach for studying their genetic relationship. The distribution pattern of rDNA in different species is generally different. Observing the difference in the number and distribution position of rDNA sites can further analyze the chromosomal behavior of inter-genus species, which is a relatively reliable and stable molecular cytogenetic marker.
In this study, we carried out a comprehensive statistical analysis of the chromosome numbers and distribution patterns of 45S rDNA and 5S rDNA in all purple-fleshed sweet potato cultivars. We have found that only 62 varieties with 90 chromosomes, and the others were aneuploid with 88, 89, 91, 92 chromosomes. Wu et al.  implemented a statistical test to detect 6× +1 and 6× -2 aneuploidy in sweet potato cultivars based on read depth, indicating that the discovery of aneuploidy may present an extreme form of structural variation that clearly would affect transcript dosage and consequently, phenotypic variation. Analyses in potato also revealed extensive structural variation, including presence/absence variation of sequences up to 575 kb in length that impacts transcript dosage [39-41].
For hexaploid sweet potato cultivars, the number of 45S rDNA was variable. In the previous studies, researchers found the number of 45S rDNA varied from 12 to 22 and the number of 5S rDNA is 6. Over the past few decades, researchers have found that the intrarespecific variation in the number and intensity of 45S rDNA signals are common, and the distribution patterns of rDNA sites often vary between closely related species, the characteristics of highly variable and unstable have been found and confirmed in many species [42-45]. Compared with previous studies, our results confirmed the high variability of 45S rDNA in varieties, ranging from 16 to 21. Excluding the influence of objective factors, the intraspecific variation in the number and location of rDNA sites may be attributed to three mechanisms: the first is the unequal crossing over and transposition event. Due to the transposon activity, the intra-genome migration of rRNA genes has been widely reported in seed plants, and it is speculated that this is one of the main factors driving the evolution of rDNA sites. Schubert et al.  found that the entire 45S rDNA repeat sequence in the chromosomes of Allium and its subgenus can be freely transferred from one site to another, indicating that 45S rDNA may move as a transposable element. The second is the occurrence of chromosomal structure fracture and rearrangement. In recent years, the view that the 45S rDNA region is a fragile site (brittle site) prone to chromosomal damage has been confirmed in several species . The increase in the number of 45S rDNA sites above the critical threshold also increases the possibility of chromosomal breakage in the ribosome DNA region . Thomas et al.  found that 45S rDNA cleavage in rye grass led to the rearrangement of chromosome structure, and the position and number of 45S rDNA sites were different. The third is the change of different degree in the process of polyploidy [50-52]. Srisuwan et al. considered that the variation in the number of 45S rDNA in hexaploid I. batatas might be due to the fact that the hexaploid genome of batatas is always in the process of diploidization, and the genome is unstable, which leads to the interspecific and intraspecific variation in the number of 45S rDNA. In addition, 45S rDNA is mainly located at the terminal and satellite of chromosomes, which may also contribute to the variable number of 45S rDNA. As 45S rDNA often falls from the terminal to the satellite, which may be more prone to unequal crossing over or ectopic recombination. Mantovani et al.  also considered that a large number of hybridization events during cultivation may also be the cause of rDNA site changes.
In previous studies, the number of 5S rDNA sites in cultivated sweet potato was relatively stable, and 6 5S rDNA sites were detected. In this study, among the 76 purple-fleshed sweet potato varieties, the number of 5S rDNA was 6 in 74 varieties, and the other two were different, 7 sites of 5S rDNA was detected in Quanzishu 96 (Figure 1, a26) and 5 in Yuzixiang 10 (Figure 1, a19). However, Longshu 9 and Quanshu 10 showed 18 45S rDNA and 6 5S rDNA hybridization signals, which was the parent materials of Quanzishu 96, hadn’t found the same phenomenon. Some studies have suggested that the increase of the number of 5S rDNA may be caused by the amplification of the covert rDNA copy number during the crossing and transposable events or due to the translocation of rDNA gene to chromosomes without rDNA sites, and the loss may be caused by other DNA sequence fusion [54-55].
The signal size and intensity of 45S rDNA and 5S rDNA are different in different varieties or the same variety, the intensity of the signal is positively related to the copy number, and the weak signal may mean that the copy number is relatively low. Events such as amplification, deletion and unequal crossing over can also affect the copy number and result in signal differences.
In general, the evolution of 45S rDNA and 5S rDNA is independent and they tend to be distributed on different chromosomes due to physical distance . Roa and Guerra [57-58] believe that 5S-45S colocalization exists in at least one species of each genus, although the probability of colocalization is relatively low, which has been reported in the Hordeum, Cucumis and Brassica [59-60]. Among the varieties we studied, there was an adjacent situation of 5s-45s in two varieties, and in the study of Sun, there was a colocalization phenomenon in three cultivars of sweet potato, which may be the result of the interaction between the characteristics of rDNA instability and some other unknown factors.