The morning glory family (Convolvulaceae) includes 50–60 genera and 1600–1700 species, which are circumscribed within twelve tribes (Stefanović et al., 2002). More than half of the species are included in tribe Ipomoeeae Hallier f (Manos et al., 2001; Stefanović et al., 2003), which contains Ipomoea and nine other genera. Morning glories have served as important model systems for diverse evolutionary and molecular genetic investigations (Miller et al., 1999; Muñoz-Rodríguez et al., 2019). Resolving their phylogenetic relationships is essential for advancing these research efforts. Phylogenetic analyses incorporating both morphological and molecular data have consistently shown that Ipomoea is not monophyletic, with the nine other genera of Ipomoeeae nested within it (Miller et al., 1999, 2002; Manos et al., 2001; Stefanović et al., 2002). Accordingly, the option was proposed that Ipomoea in its broad sense can be defined as the sole genus of Ipomoeeae based on morphological evidence (Wilkin, 1999).

Phylogenetic analyses of four plastid DNA markers classify Ipomoeeae into two principal clades: Argyreiinae and Astripomoeinae (Stefanović et al., 2003). Subsequent whole plastome sequencing further resolved Astripomoeinae into five smaller clades (Batatas, Murucoides, Pes-caprae, Quamoclit, and Cairica) and Argyreiinae into two (Pes-tigridis and Obscura) (Eserman et al., 2014). Each clade is named after its oldest constituent species identified in that study. Despite their phylogenetic distinction, no clear morphological synapomorphies separate the two major clades (Argyreiinae and Astripomoeinae) (Wilkin, 1999; Wood et al., 2020). In fact, the members of both the Astripomoeinae and Argyreiinae were often grouped together in historical classifications. For instance, Ipomoea purpurea (L.) Roth (Astripomoeinae) and I. pes-tigridis L. (Argyreiinae), which share remarkably similar gross morphology, were both classified within Ipomoea section Pharbitis (Roberty, 1952). Furthermore, the taxonomy and phylogenetic relationships for Ipomoeeae species are often incongruent in previous studies (Miller et al., 1999, 2002; Manos et al., 2001; Eserman et al., 2014). For instance, Eserman et al. (2014) recovered the Batatas, Murucoides, Pes-caprae, and Quamoclit clades as a closely related group, with Cairica positioned as the basal lineage of Astripomoeinae. In contrast, earlier studies consistently resolved Cairica and Quamoclit as sister clades (Miller et al., 1999, 2002; Manos et al., 2001), highlighting persistent phylogenetic discordance in Ipomoeeae.

Chromosomal changes have been demonstrated to serve as highly informative markers for resolving phylogenetic relationships among species (Braz et al., 2018; Meng et al., 2024). Chromosome painting (CP), using DNA probes prepared from flow-sorted or microdissected chromosomes, has been shown to be a powerful tool for tracing inter-chromosomal rearrangements in evolution among related species through cross-species chromosome painting (CCP) (Thomas et al., 1998). CCP has been widely used in animal and human cytogenetic studies, revealing that species with more similar chromosomal staining patterns generally exhibit closer phylogenetic relationships (Muller et al., 2000; Ferguson-Smith and Trifonov, 2007). However, the application of CCP in plants were not successful due to the cross-hybridization of repetitive DNA sequences in the probes that cannot be efficiently blocked (Fuchs et al., 1996).

Advances in DNA synthesis technology have made it possible to simultaneously synthesize thousands of independent oligonucleotide (oligo) fragments, providing a novel approach for developing chromosome-specific painting probes in plants (Han et al., 2015). Oligo painting probes can be designed for any plant species with a sequenced genome. Since the single-copy sequences of oligos are primarily derived from conserved gene regions, oligo probes designed from one species can be used in other plants related to the target species (Jiang, 2019). To date, oligo-based CCP has been applied in chromosome evolution studies across multiple plant genera, including Populus (Xin et al., 2020), Citrus (He et al., 2020), Fragaria (Qu et al., 2021), Cucumis (Bi et al., 2020; Zhao et al., 2021), Triticeae (Li et al., 2021; Chen et al., 2024), Saccharum (Yu et al., 2022), Aegilops (Shi et al., 2022), Thinopyrum (Chen et al., 2023), and Glycyrrhiza (Meng et al., 2024).

We have previously developed chromosome-specific oligo probes for two I. nil chromosomes, revealing genomic architecture and interspecific relationships among three polyploid Ipomoea species (Sun et al., 2022). Here, 16 representative diploid Ipomoea species were analyzed by CCP using these I. nil-derived probes to evaluate the congruence between cytogenetic data and existing molecular phylogenies of Ipomoeeae, which provides new insight into the phylogenetic relationships among these species.

Compared with DNA probes prepared from flow-sorted or microdissected chromosomes, the synthetic oligo probes designed from single copy DNA sequences have such advantages as superior resolution and versatility, customized design, labor-saving and cost-efficient, which greatly expand the application of CCP among genetically related plant species (Jiang, 2019). In this study, we found that each probe designed from two chromosomes of I. nil produced specific FISH signals exclusively on one chromosome pair in all analyzed species that diverged ca. 25 million years (Eserman et al., 2014), suggesting that the chromosomal synteny has been maintained among these species. Similar examples of syntenic maintenance within the same genus have been described in multiple plant genera (Xin et al., 2020; He et al., 2020; Qu et al., 2021; Chen et al., 2024; Meng et al., 2024).

Despite conserved synteny, distinct chromosome painting patterns were observed among species. In all analyzed 2n=30 species except for I. nil and I. hederacea, oligo probes exhibited specific hybridization to the terminal regions of the chromosomes, with notably weak or absent FISH signals in interstitial regions. This could be caused by sequence divergence and/or structural expansion of these chromosomal regions. In contrast, in all analyzed 2n=28 species, two probes from the same chromosome showed co-localization to the same chromosomal arm, indicating that inter-chromosomal translocations or rearrangements have involved these two chromosomes. Building upon the correlation between phylogenetic proximity and chromosomal painting pattern similarity (Muller et al., 2000), our CCP analysis revealed a clear cytogenetic divergence between species with chromosome numbers of 2n=28 and 2n=30, challenging the current molecular-based classification that groups them within the same small clade (Eserman et al., 2014).

While CCP analysis revealed significant cytogenetic divergence between 2n=28 and 2n=30 species, all examined species from the Batatas, Murucoides, Pes-caprae, and Cairica clades (Astripomoeinae) displayed indistinguishable painting patterns for the 7-1/7–2 and 15-1/15–2 probes. The number and position of the 5S and 45S rDNA loci are important species characteristics, and closely related species usually have more similar rDNA FISH patterns than those of distantly related ones (Garcia et al., 2017). Therefore, physical mapping of rDNA loci using the FISH technique is widely used to identify species and clarify phylogenetic relationships among related species (Han et al., 2018; Qu et al., 2021; Yucel et al., 2022; Hu et al., 2025). In the genus Ipomoea, pioneering work by Srisuwan et al. (2006) employed physical mapping of 5S and 18S rDNA to investigate the genome organization and evolution of sweetpotato (Ipomoea batatas (L.) Lam.) and its seven wild relatives, and Wu et al. (2024) subsequently determined genomic distribution of 18S and 5S rDNA sites across 13 Ipomoea species. Our previous research (Sun et al., 2024) determined the rDNA distribution in 17 Ipomoea species from the Batatas clade, yielding critical insights into sweetpotato’s origin. In this study, rDNA-FISH analysis revealed that 5S rDNA loci were consistently localized to 1–2 chromosome pairs while 45S rDNA loci occupied 2–3 chromosome pairs in all analyzed species except those in the Quamoclit clade. Consequently, while rDNA distribution patterns can effectively distinguish species within same clade, they cannot be utilized for distinguishing species across different clades.

Notably, previous studies indicated close relationships between I. hederacea and I. nil with difficult morphological distinction (Austin, 1975; Austin et al., 2001), we identified diagnostic cytogenetic differences: I. nil possesses 5S rDNA on a single chromosome pair, whereas I. hederacea exhibits two 5S rDNA-bearing chromosome pairs, with one syntenic to 45S rDNA loci. Therefore, the 5S rDNA distribution pattern serves as a reliable cytogenetic marker for distinguishing between these two morphologically similar species.

Taken together, this study presents the first phylogenetic analysis of Ipomoeeae species utilizing the CCP technique, revealing significant cytogenetic divergence between 2n=28 and 2n=30 species that challenges existing molecular-based classifications which cluster them within the same small clade. Our study not only provides new cytogenetic insights into the phylogenetic relationships for 16 representative species but also highlights the critical need for integrating molecular and cytogenetic data to accurately resolve complex phylogenetic relationships within this tribe. Building on these findings, future research will develop chromosome-specific oligos probes for each I. nil chromosome to enable detailed characterization of inter-chromosomal rearrangements and karyotype evolution across Ipomoeeae species through CCP analysis.