A Re-Evaluation Of The Homoploid Hybrid Origin Of Aegilops Tauschii, The Donor Of The Wheat D-Subgenome
Lin-Feng Li,Bao Liu,Kenneth M Olsen,Jonathan F Wendel
DOI: https://doi.org/10.1111/nph.13294
IF: 9.4
2015-01-01
New Phytologist
Abstract:Hybridization is a prominent evolutionary force promoting plant diversification, either with or without subsequent genome doubling (Abbott et al., 2013; Soltis et al., 2014; Yakimowski & Rieseberg, 2014). The Aegilops–Triticum complex is an ideal system to investigate how natural hybridization and allopolyploidization have caused species diversification (Matsuoka, 2011). Recently, Marcussen et al. (2014) proposed the tantalizing scenario that the ancestral D lineage originated via homoploid hybridization between ancient A and B lineages some five million years ago (Mya) (the definition of A, B and D lineages shown in Fig. 1). Evidence for this mode of origin was derived from phylogenomic and population genetic analyses of nuclear genes, but without taking into account the evolutionary history and chloroplast topology of this species complex. Meanwhile, in a recent issue of New Phytologist, Gornicki et al. (2014) reported the chloroplast phylogeny of the Triticum–Aegilops complex based on 25 chloroplast genomes of eight modern A, S and D genome diploid species and four polyploid wheat species, but they did not address the origin of the D genome. Here, by re-analyzing critical data used by both studies and additional data, we present evidence for a more complex hybrid origin of the D genome of A. tauschii. To date, 13 diploid species of the Triticum–Aegilops complex, which belong to eight distinct but related genome groups (A, D, S, M, C, U, N and T), have been identified (Table 1) and which are variously and sometimes sympatrically distributed in the Middle East (Lilienfeld, 1951; Gill & Friebe, 2002; Huang et al., 2002). The A, B and D genomes, harbored by diploid species T. urartu, A. speltoides (or a closely related species) and A. tauschii, respectively, are established as the diploid genome donors of the A-, B- and D-subgenomes of hexaploid bread wheat, T. aestivum (Cox, 1998; Huang et al., 2002; Petersen et al., 2006). The remaining genomes harbored by the diploid species (except the T genome) are found in polyploid Aegilops species (Gill & Friebe, 2002). Eighteen naturally occurring allopolyploid species have been described (Table 1), which also are widely distributed across the Near East. On the basis of plant habit, spike morphology and cytological data, Zohary & Feldman (1962) classified these allopolyploid species into three major genome groups (called cytological clusters), which are A, U and D (Table 1). Variation patterns of the three cytological clusters led to the hypothesis that the current allopolyploid species within each cluster probably evolved from only a few initial amphidiploids (Zohary & Feldman, 1962; Feldman, 1965; Pazy & Zohary, 1965). This hypothesis is further supported by cytological and phylogenetic analyses showing that the A-, U- and D-genomes have indeed donated the maternal genome to most of the current polyploid species (Kimber & Tsunewaki, 1988; Meimberg et al., 2009; Tsunewaki, 2009). Together, these attributes suggest that hybridization, either at the homoploid level or followed by polyploidization, has occurred frequently within the Triticum–Aegilops complex. Marcussen et al. (2014) proposed that all extant diploid species of the Aegilops–Triticum complex are derived from A, B and D lineages (Fig. S6 in Marcussen et al., 2014). Phylogenomic analyses based on nuclear genome sequences revealed that the phylogenetic positions of A (T. monococcum/T. urartu), B (A. speltoides) and D (A. tauschii) genome species varied among nuclear genes, with topologies A (B, D) and B (A, D) each being about twice as common as D (A, B) (Table 1 in Marcussen et al., 2014). However, evidence for the homoploid hybrid origin of A. tauschii derives from phylogenomic analyses of modern S, A and D genomes, without taking into account the other genomes (e.g. M, N, T, U and C) within this species complex. Given that the breadth of taxonomic sampling could affect the identification of hybridization events, we re-analyzed the topologies of the four focal genomes (A, S, D and Ssh) for the 275 nuclear genes used in Marcussen et al. (2014), who proposed a homoploid hybrid origin of the D lineage (Table S4 in Marcussen et al., 2014). If the D lineage species were formed from a single homoploid hybridization event, as proposed by Marcussen et al. (2014), then all species derived from this event would be expected to cluster together at most nuclear genes. However, our results showed that the two D lineage species, A. tauschii and A. sharonensis, are separated in 40% of the 209 gene trees (Supporting Information Fig. S1). Marcussen et al. (2014) also reported a distinct hybrid pattern for A. sharonensis, with only c. 25% of the 275 gene trees reflecting B-lineage ancestry. Indeed, Waines & Johnson (1972) have documented that A. sharonensis is a hybrid between A. longissima and A. bicornis. In addition, cytogenetic analyses revealed that modern S* genome species (D lineage) are closer to A. speltoides (B lineage) than to A. tauschii (D lineage) (Kihara, 1954). Based on these observations, we hypothesize that the origin of A. tauschii may be more complicated than envisioned by Marcussen et al. (2014). Additional information bearing on the history of the Triticum–Aegilops complex may derive from analysis of chloroplast genomes, which are maternally inherited in this species complex (Fukasawa, 1959; Kihara, 1959). The cpDNA topology of this group has been investigated in previous studies using selected genes (Hirai & Tsunewaki, 1981; Terachi et al., 1987; Wang et al., 1997; Yamane & Kawahara, 2005; Meimberg et al., 2009), and more recently whole genome sequences (Gornicki et al., 2014; Middleton et al., 2014). These studies focused on the origin of domesticated wheat and phylogenetic relationships within the Triticum–Aegilops complex. For example, Gornicki et al. (2014) revealed a basal clade of A. speltoides and A. tauschii grouped together with four S* genome species (A. longissima, A. searsii, A. sharonensis and A. bicornis). However, the other diploid genome (e.g. M genome) and polyploid species within Triticum–Aegilops complex were not included in their study. Middleton et al. (2014) also reported a similar chloroplast topology with five A, S, D diploid species and three polyploid species (including M genome), but the S* genome species were not included in their study. To encompass all diploid and polyploid genomes in a single framework, we retrieved 20 whole chloroplast genome sequences from GenBank deposited by Gornicki et al. (2014) and Middleton et al. (2014) and conducted neighbor-joining phylogenetic analysis using MEGA 6 (Tamura et al., 2013) (Fig. 1). The resulting topology showed that the chloroplast genome of A. tauschii (D genome) is more closely related to that of other D and all S* genome species (including A. sharonensis) than it is to either the A. speltoides (S genome) or T. monococcum/T. urartu (A genome) chloroplast genomes. This result would not be expected under the scenario of a single homoploid hybrid origin of A. tauschii, in which case the cpDNA of A. tauschii would be expected to be more similar to that of either the A or B genome lineages. Marcussen et al. (2014) proposed that A. tauschii originated from homoploid hybridization between A. speltoides and the ancestor of modern T. urartu/T. monococcum. If this were the case, A. tauschii would be expected to share a chloroplast genome with one of these two putative progenitor lineages. However, the chloroplast topology reveals that A. tauschii is cladistically nested between the A and remaining S* and M genomes (Fig. 1). One possible explanation for the apparent discrepancy between the cpDNA-derived phylogeny and the interpretation of Marcussen et al. (2014) is that A. speltoides itself underwent later hybridization(s), in which it captured the chloroplast genome from some other species in the B genome lineage. As shown in the chloroplast topology, however, A. speltoides is basal in the clade, differing phylogenetically and quantitatively from the remaining species in the Triticum–Aegilops complex. The latter is evidenced in an analysis of single nucleotide polymorphisms (SNPs), in which A. speltoides shares fewer SNPs with the remaining genomes than they do with each other (Table 2). These attributes indicate that A. speltoides possesses a rather distinct chloroplast genome. In addition, the number of shared SNPs between D and A genomes is only slightly higher than between D and M or between D and S* genomes (Table 2), suggesting that the D genome is equidistant from the A, S* and M genomes. An alternative scenario is that the origin of the D genome lineage that gave rise to A. tauschii is more complex than that proposed by Marcussen et al. (2014). For example, A. speltoides and T. urartu/T. monococcum could have evolved from the ancient B and A genomes, respectively, with the remaining species (including the modern D, S* and M genomes in Fig. 1) of Triticum–Aegilops complex derived from the ancient D lineage (Fig. 2a). If the ancient D lineage were to have evolved through a single homoploid hybridization between A and B lineages, it would be expected to phylogenetically cluster with either of its two donors and have a cpDNA genome that is much more like one parent than the other. The chloroplast topology reveals that the ancient D lineage clusters together with A lineage (Fig. 1). Indeed, the number of shared SNPs between the A and D lineages is higher than that between the B and D lineages (Table 2). Under this hypothesis, recent hybridization might be responsible for the distinct hybrid pattern of modern A. tauschii and A. sharonensis in the phylogenomic analyses. Another possibility for the origin of modern A. tauschii is a chloroplast capture model (Fig. 2b). Under this scenario A and B lineages were involved in homoploid hybridization, leading to the formation of an A. tauschii-like nuclear genome (ancestry of modern A. tauschii), which was followed by hybridization(s) with a third species in the S*/M genomes. In this case, modern A. tauschii captured the chloroplast genome from S*/M genome species without much nuclear introgression. Alternatively, modern A. tauschii might have originated from homoploid hybridization between the A and B lineages, but the hybridization event was preceded by an earlier hybridization with a third species that donated the S*/M lineage chloroplast genome to the maternal parent of the later homoploid hybridization event (Fig. 2c). If this were the case, modern A. tauschii would have evolved with A and B lineage nuclear genomes, but with a S*/M type chloroplast genome. Taken together, our integrated re-evaluation, while confirming the hybrid nature of A. tauschii, points clearly to a more complex history of the species than that proposed by Marcussen et al. (2014), one that may have involved multiple rounds of both recent and ancient hybridizations. All chloroplast genome sequences deposited by Gornicki et al. (2014) and Middleton et al. (2014) were retrieved from GenBank. Aegilops speltoides, JQ740834, NC_022135; A. sharonensis, KJ614419; A. bicornis, KJ614417; A. longissima, KJ614416; A. searsii, KJ614415; A. tauschii, JQ754651, NC_022133; Triticum monococcum, KC912690, NC_021760; T. urartu, KC912693, NC_021762; T. aestivum, KC912694; A. cylindrica, KF534489, NC_023096; A. geniculata, KF534490, NC_023097; Hordeum vulgare, KC912687; Secale cereale, KC912691, NC_021761. The authors thank Xutong Wang, Fengxue Shi and Cui Zhang for assistance with data analyses. This study was supported by the National Natural Science Foundation of China (31290210, 31470010) and the Program for Introducing Talents to Universities (B07017). Please note: Wiley Blackwell are not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing material) should be directed to the New Phytologist Central Office. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.