large-scale analyses of angiosperm nucleotide-binding site … · large-scale analyses of...

Large-Scale Analyses of Angiosperm Nucleotide-BindingSite-Leucine-Rich Repeat Genes Reveal Three AncientlyDiverged Classes with Distinct Evolutionary Patterns1

Zhu-Qing Shao2, Jia-Yu Xue2, Ping Wu, Yan-Mei Zhang, Yue Wu, Yue-Yu Hang, Bin Wang*, andJian-Qun Chen*

State Key Laboratory of Pharmaceutical Biotechnology, School of Life Sciences, Nanjing University, Nanjing,210023, China (Z.-Q.S., P.W., Y.-M.Z., Y.W., B.W., J.-Q.C.); and Institute of Botany, Jiangsu Province andChinese Academy of Sciences, Nanjing, 210014, China (J.-Y.X., Y.-M.Z., Y.-Y.H.)

Nucleotide-binding site-leucine-rich repeat (NBS-LRR) genes make up the largest plant disease resistance gene family (R genes),with hundreds of copies occurring in individual angiosperm genomes. However, the expansion history of NBS-LRR genesduring angiosperm evolution is largely unknown. By identifying more than 6,000 NBS-LRR genes in 22 representativeangiosperms and reconstructing their phylogenies, we present a potential framework of NBS-LRR gene evolution in theangiosperm. Three anciently diverged NBS-LRR classes (TNLs, CNLs, and RNLs) were distinguished with unique exon-intron structures and DNA motif sequences. A total of seven ancient TNL, 14 CNL, and two RNL lineages were discoveredin the ancestral angiosperm, from which all current NBS-LRR gene repertoires were evolved. A pattern of gradual expansionduring the first 100 million years of evolution of the angiosperm clade was observed for CNLs. TNL numbers remained stableduring this period but were eventually deleted in three divergent angiosperm lineages. We inferred that an intense expansion ofboth TNL and CNL genes started from the Cretaceous-Paleogene boundary. Because dramatic environmental changes and anexplosion in fungal diversity occurred during this period, the observed expansions of R genes probably reflect convergentadaptive responses of various angiosperm families. An ancient whole-genome duplication event that occurred in anangiosperm ancestor resulted in two RNL lineages, which were conservatively evolved and acted as scaffold proteins fordefense signal transduction. Overall, the reconstructed framework of angiosperm NBS-LRR gene evolution in this study mayserve as a fundamental reference for better understanding angiosperm NBS-LRR genes.

Plant disease resistance (R) genes are a set of genes thatconfer resistance to various pathogens. Among all knowntypes of R genes, nucleotide-binding site-leucine-rich re-peat (NBS-LRR,NBS for short) genes represent the largestclass, encompassing more than 80% of the characterizedR genes (Meyers et al., 1999; Meyers et al., 2005; McHale

et al., 2006; Friedman and Baker, 2007). Since the firstcomprehensive study on NBS genes in Arabidopsis(Arabidopsis thaliana; Meyers et al., 2003), NBS genes havebeen surveyed across various plant genomes, most ofwhich belong to the rosid lineage of eudicots and thePoaceae of monocots (Monosi et al., 2004; Zhou et al.,2004; Yang et al., 2006; Ameline-Torregrosa et al., 2008;Yang et al., 2008; Mun et al., 2009; Porter et al., 2009; Liet al., 2010; Zhang et al., 2011; Lozano et al., 2012; Luoet al., 2012; Andolfo et al., 2013; Shao et al., 2014).

NBS genes are historically divided into two classes,namely, TIR-NBS-LRR (TNL) and the non-TIR-NBS-LRR (nTNL), which are differentiated by the presenceof a Toll/IL-1 Receptor-like (TIR) domain in the proteinamino terminus (Meyers et al., 1999; Bai et al., 2002;Zhou et al., 2004). Because most nTNL genes encode acoiled-coil (CC) domain at the N terminus, the nTNLgenes often are called CC-NBS-LRR (CNL) genes(Meyers et al., 2003; Ameline-Torregrosa et al., 2008).However, recent studies have revealed that apart fromCNL genes, a small group of nTNL genes that possess aspecial N-terminal domain, RPW8 (resistance to pow-dery mildew8) domain, likely represent a distinct classof NBS genes (RPW8-NBS-LRR [RNL]; Xiao et al., 2001;Cannon et al., 2004; Bonardi et al., 2011; Collier et al.,2011; Shao et al., 2014; Zhang et al., 2016). Investiga-tions of NBS genes in the Fabaceae (rosid I lineage) and

1 This work was supported by the National Natural Science Foun-dation of China (30930008, 31170210, 91231102, 31300190, 31400201,31470327, 31500191, and 31570217), the Natural Science Foundationof Jiangsu Province (BK20130565), the Fundamental Research Fundsfor the Central Universities (2062140215, 20620140546, and20620140558), and the National Postdoctoral Science Foundation ofChina (2013M540435 and 2014T70503).

2 These authors contributed equally to the article.* Address correspondence to [email protected] or binwang@nju.

edu.cn.The author responsible for distribution of materials integral to the

findings presented in this article in accordance with the policy de-scribed in the Instructions for Authors (www.plantphysiol.org) is:Jian-Qun Chen ([email protected]).

J.-Q.C., B.W., J.-Y.X., and Z.-Q.S. conceived and designed the pro-ject; J.-Y.X., Z.-Q.S., and B.W. obtained and analyzed the data; P.W.,Y.-M.Z., Y.W., and Y.-Y.H. participated in the data analysis and dis-cussion; Z.-Q.S. and J.-Y.X. drafted the initial manuscript; B.W. com-plemented the writing; J.-Q.C. supervised the analysis and writing;all authors contributed to discussion of the results, reviewed the man-uscript, and approved the final article.

www.plantphysiol.org/cgi/doi/10.1104/pp.15.01487

Plant Physiology�, April 2016, Vol. 170, pp. 2095–2109, www.plantphysiol.org � 2016 American Society of Plant Biologists. All Rights Reserved. 2095 www.plantphysiol.orgon June 12, 2020 - Published by Downloaded from

Copyright © 2016 American Society of Plant Biologists. All rights reserved.

mailto:[email protected]



http://www.plantphysiol.org


http://www.plantphysiol.org/cgi/doi/10.1104/pp.15.01487


Brassicaceae (rosid II lineage) have shown that RNLsand CNLs diverged prior to the divergence of these twofamilies (Shao et al., 2014; Zhang et al., 2016). Interest-ingly, although both CNL and RNL genes are present inmonocots and dicots, TNL genes only are present indicots (Meyers et al., 2003; Tarr and Alexander, 2009;Andolfo et al., 2013; Shao et al., 2014). Considering theabundance of TNLs in various dicot species, their ab-sence in monocots is perplexing.

For several years, our knowledge of NBS gene evo-lution has been based on investigations of individualangiosperm genomes (Meyers et al., 2003; Ameline-Torregrosa et al., 2008; Kohler et al., 2008). The num-ber of NBS genes varied dramatically among differentspecies, even among different accessions of the samespecies (Zhang et al., 2010). However, the relativecontributions of recent and ancient expansions to theabundance of NBS genes in modern plant genomesremain elusive. Recent studies report the evolutionaryanalyses of NBS genes across multiple species in a plantfamily (Cannon et al., 2002; Guo et al., 2011; Luo et al.,2012; Lin et al., 2013; Shao et al., 2014; Zhang et al.,2016). While studies in Brassicaceae and Poaceae bothrevealed abundant and recent losses of NBS genes invarious species after the divergence of these two fami-lies, a study involving Fabaceae has suggested the op-posite scenario in which all surveyed species recentlyunderwent a multitude of NBS expansions (Luo et al.,2012; Shao et al., 2014; Zhang et al., 2016). Althoughdistinct evolutionary patterns of NBS genes have beenidentified at the family level, hundreds of NBS genelineages arose in the ancestors of Fabaceae, Brassicaceae,and Poaceae, indicating that NBS genes may haveprobably undergone ancient expansions prior to fa-milial divergence. Therefore, a phylogenetic analysisat a larger scale is necessary to better understand theexpansion history of NBS genes in angiosperms.

Several functional NBS genes have been identified inangiosperm species (Liu et al., 2007); however, their timeof emergence and relationship with other NBS geneshave not been established. For example, several CNLgenes have been identified in both soybean (Glycine max;Rpg1b and Rpg1r) and Arabidopsis (RPM1 and RPS2)against the bacterial pathogen Pseudomonas syringae byguarding the host protein RIN4 (Bent et al., 1994;Mackeyet al., 2002; Mackey et al., 2003; Ashfield et al., 2014).While the Rpg1b and Rpg1r genes were deduced toemerge after the divergence of Fabaceae species, theevolutionary time point of the origin of RPM1 and RPS2genes in Arabidopsis is largely unknown (Ashfield et al.,2014; Shao et al., 2014). Because all these R genes havelikely adopted a similar strategy to defend against thesame pathogen, investigating their evolution may eluci-date the mechanism underlying the establishment andmaintenance of resistance in various plant species.

To obtain a more comprehensive understanding ofNBS gene evolution, it is essential to analyze morespecies/lineages to determine how the current NBSgenes have evolved and diverged from ancestral NBSlineages in the ancestral angiosperm. This also will

provide when and how resistance loci emerged or dis-appeared in certain lineages. In this study, a total of 22plant species (Fig. 1) from key systematic clades ineudicots, monocots, and the early diverging basal an-giosperm were chosen to depict the ancient evolution-ary history of NBS genes in angiosperms. More than6,000 NBS genes were identified; family level phylo-genetic trees were constructed first and then expandedto higher systematic levels to eventually analyze all ofthe angiosperms. Ancient taxonomies of NBS geneswere traced to the origin of angiosperms, and a total of23 ancestral NBS genes were recovered in the lastcommon ancestor of angiosperms. This angiospermNBS frame provides a fundamental reference for un-derstanding the evolutionary history of NBS genes aswell as insight into the finer details of particular NBSgene classes.

RESULTS

Identification and Classification of NBS Genes from22 Angiosperm Species

A total of 6,135 NBS genes were identified in 22 an-giosperm species using previously described proce-dures (Shao et al., 2014; Zhang et al., 2016). Five of the22 genomes were surveyed in this study, Amborellatrichopoda, Musa acuminata (banana), Phyllostachys het-erocycla (bamboo), Capsicum annuum (hot pepper), andSesamum indicum (sesame), whereas the others havebeen previously reported (Meyers et al., 2003; Luo et al.,2012; Malacarne et al., 2012; Andolfo et al., 2013; Shaoet al., 2014; Zhang et al., 2016). A. trichopoda, a basalangiosperm, had a total of 105 NBS genes, including 15TNLs, one RNL, and 89 CNLs (Fig. 1; SupplementalTable S1). TheNBS genes varied among the 22 surveyedspecies (Fig. 1). Of all species surveyed, Medicago trun-catula possessed the highest number of NBS genes (571),more than 6-fold of that observed in Thellungiella sal-suginea (88). RNL genes were detected in all of thesurveyed species, although at lower numbers (1–18).A large proportion of NBS genes were either TNLs orCNLs, although TNLs were not detected in eight of 22species, including seven monocots and S. indicum. Be-cause the basal angiosperm A. trichopoda harbored TNLgenes, these genes were most likely also present in theancestor of angiosperms but independently lost in cer-tain lineages. In species with both TNL and CNL genes,CNLs often outnumbered TNLs, except for the fourBrassicaceae genomes analyzed (Zhang et al., 2016).The asterid genomes showed a lower ratio of TNL toCNL genes, with the C. annuum exhibiting the highestlevel of disparity (TNL:CNL = 1:18).

Phylogenetic Analysis of NBS Genes in MajorAngiosperm Clades

The 22 angiosperm species surveyed in this studyoccupy important phylogenetic positions (Fig. 1). To

2096 Plant Physiol. Vol. 170, 2016

Shao et al.

www.plantphysiol.orgon June 12, 2020 - Published by Downloaded from Copyright © 2016 American Society of Plant Biologists. All rights reserved.

http://www.plantphysiol.org/cgi/content/full/pp.15.01487/DC1



trace the evolutionary history of NBS genes, we exam-ined multiple angiosperm clades, including rosids (Ro),asterids (As), eudicots (Ed), monocots (Mo), mesan-giospermae (Ma), and all angiosperms (An; Fig. 1). It iswidely recognized that the NBS domain effectivelydistinguishes NBS genes into TNL and nTNL classes(Bai et al., 2002; Meyers et al., 2003). Although a sisterrelationship between CNLs and RNLs has been estab-lished in rosids, it has not been examined at angiospermlevel. Therefore, phylogenetic analyses of TNLs andnTNLs (including both CNLs and RNLs) were sepa-rately performed in this study. We first conductedphylogenetic analysis of three major angiosperm line-ages, namely, monocots, asterids, and rosids, whichcover seven, four, and 10 genomes, respectively.The M. acuminata genome, together with six Poaceae

genomes, includingOryza sativa, Brachypodium distachyon,P. heterocycla, Zea mays, Sorghum bicolor, and Setaria italica,were used to elucidate the evolution of NBS genes inmonocots. Because these species did not harbor TNLgenes, phylogenetic analysis of only nTNL genes wasperformed. The reconstructed monocot nTNL phylogeny(Supplemental Fig. S1) revealed that nTNLgenes from thesix Poaceae species formed 456 monophyletic lineages,indicating the presence of at least 456 ancestral nTNLgenes in the Poaceae common ancestor (;62million yearsago [MYA]; Peng et al., 2013, Zhang et al., 2012). ThesePoaceae nTNL genes formed 57 monophyletic lineageswhen grouped withM. acuminata nTNL genes (one RNL

lineage designated as Mo-R1, and 56 CNL lineagesidentified as Mo-C1 to Mo-C56), suggesting that nTNLgenes from the seven monocot genomes were actuallyderived from57 ancestralMo-NBS genes;110MYA (Fig.2; Supplemental Fig. S1; D’Hont et al., 2012). From theircommonancestor, the Poaceae andM. acuminata inherited51 and 32 ancestral Mo-NBS genes, respectively. Amongthese, 26 were shared by both Poaceae andM. acuminata.Several ancestral genes in the Poaceae underwent exten-sive duplications (from 51 to 456) during the time intervalof 48 million years. For example, the NBS gene Mo-C56duplicated into 197 genes during this time period(Supplemental Fig. S1). Asmultiple species radiated fromthe Poaceae common ancestor, in some species NBS genenumbers contracted drastically (Z. mays, 139 from 456)and in others only slightly (S. italica and P. heterocycla, 424and 344 from 456, respectively), whereas expansion wasonly found in the O. sativa genome (from 456 to 498).

Within the asterids four species were analyzed, in-cluding three Solanaceae species, namely, Solanumlycopersicum, Solanum tuberosum, andC. annuum, as wellas S. indicum from Pedaliaceae. NBS genes from thethree Solanaceae genomes formed 19 TNL monophy-letic lineages (Supplemental Fig. S2) and 147 nTNLmonophyletic lineages (Supplemental Fig. S3), sug-gesting that at least 166 ancestral NBS genes were pre-sent in their last common ancestor. The SolanaceaenTNL genes formed 55 monophyletic lineages withnTNL genes from S. indicum (two RNL lineages

Figure 1. Identification and classificationof NBS genes in 22 angiosperm genomes.The phylogenetic relationship was con-structed according to the APG III system(Bremer et al., 2009). The divergence timesat different nodes of angiosperms werecombined from previous studies (Lavinet al., 2005; Stefanovic et al., 2009;D’Hont et al., 2012; Zhang et al., 2012;Peng et al., 2013; Yang et al., 2013; Kimet al., 2014; Wang et al., 2014; Zeng et al.,2014). The total number of NBS genes andtheir classification in each species areshown. The number of ancestral NBSgenes recovered by phylogenetic analysisat each node is shown in a green circle,and an asterisk after the number indicatesthat it is not directly unraveled by phylo-genetic analysis.

Plant Physiol. Vol. 170, 2016 2097

Evolution of NBS-LRR Genes in Angiosperms








designated as As-R1 and As-R2, and 53 CNL lineagesidentified as As-C1 to As-C53), indicating a minimum of55 ancestral genes in the asterid common ancestor (Fig. 2;Supplemental Fig. S2). Solanales inheritedmoreAs-nTNLlineages than S. indicum, which represents Lamiales (47versus 30). No TNL genes were detected in the S. indicumgenome; therefore, the accurate number of ancestral TNLgenes at this asterid node was not estimated.

Rosids are a major lineage of eudicots that consists ofthree sublineages, rosid I (fabids), rosid II (malvids),and the basal rosids. Previous work analyzed NBSgenes in Fabaceae and Brassicaceae (Shao et al., 2014;Zhang et al., 2016), which belong to the rosid I sub-lineage and the rosid II sublineage, respectively. A totalof 64 nTNL and 55 TNL ancestral lineages were dis-covered in the last common ancestor of the four

Figure 2. Cladograms depicting reconciled phylogenetic relationships of ancestral NBS genes in the divergence nodes ofmonocots, rosids, and asterids. The presence of a NBS gene lineage in different species and families are indicatedwith species orlineage-specific icons. The maximum likelihood trees of each cladogram are presented in Supplemental Figures S1 to S5.


Shao et al.





Fabaceae species, and 80 nTNL and 148 TNL ancestrallineages were identified in the last common ancestor ofthe five surveyed Brassicaceae species. In this study,one representative sequence from each of the 119Fabaceae ancestral NBS lineages and the 228 Brassica-ceae ancestral NBS lineages were used to represent thediversity of ancient NBS genes in these two families,and then combined with NBS genes identified in Vitisvinifera (basal rosid) to construct NBS phylogenies forthe whole rosid lineage. A total of 61 ancestral Ro-NBSgenes were identified (four RNLs, 49 CNLs, and eightTNLs) in the common ancestor of the rosids (;108MYA;Fig. 2; Supplemental Figs. S4 and S5), among which theFabaceae of rosid I lineage inherited 35 Ro-NBS genes(two RNLs, 27 CNLs, and six TNLs), Brassicaceae ofrosid II lineage inherited 19 (two RNLs, 12 CNLs, andfive TNLs), and basal rosids inherited 33 (two RNLs,29 CNLs, and two TNLs). Less than half of the Ro-NBSgenes (26/61) were detected in two of the lineages, and35/61 were observed in one descending lineage (Fig. 2),thereby reflecting a pattern of frequent NBS gene lossafter their split from the common rosid ancestor. The35 and 19 Ro-NBS genes inherited by Fabaceae andBrassicaceae lineages expanded to 119 (55 TNLs and 64nTNLs) and 228 (148 TNLs and 80 nTNLs) NBS genes inthe Fabaceae ancestor (;54 MYA) and the Brassicaceaeancestor (;42 MYA), respectively (Shao et al., 2014;Zhang et al., 2016). The NBS genes continued to expandin Fabaceae (Shao et al., 2014), whereas these genescontracted as the Brassicaceae diverged into multiplespecies (Zhang et al., 2016). Although the Brassicaceaeinherited fewer ancestral TNLs than CNLs (five versus12) from the rosid ancestor, this lineage vigorously ex-panded its TNLs during the time interval of;108MYAto ;42 MYA. A 30-fold expansion was observed in theBrassicaceae TNLs (from five to 148), far beyond that oflegume TNLs (;9-fold, from six to 55; Shao et al., 2014;Zhang et al., 2016).

Reconstruction of NBS Gene Phylogenies at theWhole-Angiosperm Scale

Representative NBS sequences from ancestral lineagesof asterids, rosids, and monocots were used to constructancestral NBS gene lineages at the divergence nodes ofeudicots and mesangiospermae. A total of eight Ed-TNLand 49 Ed-nTNL genes (47 Ed-CNLs and two Ed-RNLs)were recovered at the eudicot node (Supplemental Figs.S6 and S7),whereas 37Ma-nTNLgenes (35Ma-CNLs andtwo Ma-RNLs) were detected at the mesangiospermaenode (Supplemental Fig. S8). Independent loss and dif-ferential inheritance of ancestral NBS genes were alsoobserved at these two nodes. Representative genes fromeach ancestral Ma-nTNL lineage were compiled withA. trichopoda nTNL genes to reconstruct angiospermnTNL phylogeny (Supplemental Fig. S9). Because amesangiospermae TNL phylogeny could not be con-structed due to the absence of TNL genes inmonocots, aTNL phylogeny was generated at the angiosperm levelusing eudicot representative TNLs and A. trichopoda

TNLs (Supplemental Fig. S10). Figure 3 shows that allTNL genes clustered into seven ancestral angiospermlineages (designated as An-T1 to An-T7), and all nTNLgenes clustered into 16 ancestral lineages (two RNL lin-eages designated as An-R1 and An-R2, and 14 CNL lin-eages identified as An-C1 to An-C14), indicating at least23 NBS genes were present in the last angiosperm an-cestor (;225 MYA). The basal angiosperm A. trichopodainherited only eight of the 23 An-NBS genes, whereas themesangiospermae inherited 19 (Fig. 3). The angiospermnTNL tree was composed of two monophyletic groups,namely, RNL and CNL (Supplemental Fig. S9). There-fore, despite the extensive discrepancy in gene numbersbetween the two groups, this study revealed that theRNL is another ancient class ofNBS genes that has a sisterrelationship with the CNL class in angiosperms.

The angiosperm NBS gene phylogenies haverevealed the evolutionary state of NBS genes at criticalangiosperm nodes, which indicated that differentialinheritance of ancestral NBS genes and lineage-specificexpansions are major themes of NBS gene evolution.

Characterization of Class-Specific Signatures among ThreeAncient NBS Gene Classes

Because the results of phylogenetic analysis suggestthat RNLs represent an ancient NBS gene class inde-pendent of the CNL and TNL classes, characterizationof distinctive features of each class was performed.

Distinctive Signatures in Conserved NBS Motifs Differentiatethe Three Classes

The NBS domain is composed of a number of char-acteristic motifs that are conserved and shared by allNBS genes, whereas other motifs are more variable(DeYoung and Innes, 2006). All three classes of NBSgenes share five functionally important motifs: P-loop,Kinase-2, RNBS-B, GLPL, and RNBS-D. The amino acidsequence conservation of these five motifs was esti-mated using sequence logos using MEME software(Fig. 4; Bailey and Elkan, 1995). All motifs were highlyconserved at most sites among all classes, except for theRNBS-Dmotif, which was poorly conserved in the TNLand CNL classes. The class-specific features withinthese motifs also were identified. The Kinase-2 motifhad a conserved “DDVW” sequence in the RNL andCNL genes, but frequently appeared as “DDVD” in theTNL genes. A “TSR” sequence in the RNBS-B motif ofRNLs was conserved, but frequently appeared as“TTR” and “TTRD” in the CNL and TNL genes, re-spectively. CNL did not have specific highly conservedmotifs, but could be distinguished by a RNL likeKinase-2 motif and a TNL like RNBS-B motif.

Intron-Exon Structures Also Distinguish NBS Classes

The absence/presence, position, and phase of intronsoften are regarded as important parameters in phy-logeny (Qiu et al., 1998; Sánchez et al., 2003; Qian et al.,












2015). In a previous study, the intron positions andphases were determined to be distinctive among theclasses of NBS genes in bryophytes (Xue et al., 2012). Inthis study, three classes of NBS genes in angiospermsalso differed in intron characteristics (Fig. 4). Three in-trons with conserved positions and phases were sharedby most TNL genes, with the first intron separating theTIR domain and the NBS domain (phase 2), the secondintron separating the NBS domain and the whole LRRdomain (phase 0), and the third intron separating thefirst and the remaining LRR domains (phase 0). An-giosperm RNLs share four class-specific introns, withthe exception of RNLs in monocots, which did notpossess the second intron (Fig. 4). Because the secondintron is present in both A. trichopoda and dicot RNLgenes, intron loss likely occurred in the monocot line-age. For the CNL class, conserved introns were notdetected across all CNL lineages, especially in the basallineages. In fact, a number of CNL genes were intron-less, and were not concentrated in a specific lineage butdistributed across all ancestral angiosperm nTNL line-ages. Although some CNL genes had introns, thesewere more likely to have been gained at later evolu-tionary stages. This finding is in agreement with theintronless structure reported in bryophytes (Xue et al.,2012).

The Three NBS Classes Show Distinct EvolutionaryPatterns in Angiosperms

Our step-by-step phylogenetic analysis not onlysupports the presence of three distinct NBS classesin the common ancestor of angiosperms, but also

unraveled the ancient NBS gene lineages at multipleangiosperm evolutionary nodes (Fig. 5). Due to the lossof TNL genes in S. indicum and the monocot lineage, theTNL ancestral states at the asterid andmesangiospermaenodes could not be directly obtained from phylogeneticanalyses. However, based on the recovered ancestralNBS lineages at the divergence nodes of Solanaceae(Supplemental Fig. S2), dicots (Supplemental Fig. S6),and all angiosperms (Supplemental Fig. S10), bothasterid and mesangiospermae nodes were deduced tohave at least five TNL lineages.

Above the family divergence nodes (An, Ma, Mo, Ed,As, Ro, and Er), CNL genes were abundant but onlyslightly more than those of earlier ancestral nodes (Fig.5). This finding indicates that CNL genes adopted agradual expansion strategy in the early stage of angio-sperm evolution (from ;225 MYA as angiosperms di-verged to around 100 MYA as asterids, rosids, andmonocots diverged; Fig. 1; D’Hont et al., 2012; Wanget al., 2014; Zeng et al., 2014). Intensive expansion ofCNL genes seems to have occurred in the familial an-cestors of Solanaceae and Poaceae, reaching hundredsof copies. TNL genes apparently followed a differentevolutionary path, as these were completely lost inmonocots and in S. indicum. Although inherited by di-cots, TNL numbers remained relatively constant for along evolutionary period, with similar numbers of an-cestral TNL genes identified at various ancestral nodes(An, Ma, Ed, As, Ro, and Er; Fig. 5). The expansion ofTNLs mainly occurred prior to the divergence of fam-ilies, such as in the Fabaceae ancestor, 55 TNLs wereexpanded from six of eight Ro-TNLs. A similar patternwas observed in the Brassicaceae ancestor, in which

Figure 3. Cladograms depicting reconciled phylogenetic relationships of ancestral NBS genes in the common ancestor of an-giosperms. The maximum likelihood trees of each cladogram are provided in Supplemental Figures S9 and S10.


Shao et al.







four Ro-TNL lineages expanding to 148 genes. The TNLgenes were probably maintained at a lower number formore than 100 million years (from ;225 MYA as an-giosperm diverged to around 100 MYA as rosids andasterids diverged) and only began to expand thereafter.This may explain why TNL genes were easily lost inmultiple independent angiosperm lineages.RNL genes are differentiated from the other two

classes by their relatively conserved evolutionary pat-tern. Nearly all species from basal angiosperm, mono-cots, and asterids have one or two RNL genes, whereasa slight expansion of RNL genes was observed in cer-tain rosid genomes (Fig. 1). Tracing the evolutionaryhistory of RNL genes revealed two anciently divergedlineages in the common ancestor of angiosperms (Fig.6). One RNL lineage contained the Arabidopsis ADR1gene, whereas the other included the tobacco (Nicotianabenthamiana) NRG1 orthologs. Both lineages were in-herited by dicot species, except for S. indicum, whichlost the NRG1 lineage, similar to all monocot genomes.Several whole-genome duplications (WGDs) occurredthroughout angiosperm evolution, causing gene du-plications and functional divergence (Jiao et al., 2011;Vanneste et al., 2014). The separation of the ADR1 andNRG1 lineages prior to the divergence of angiospermsprompted us to investigate the involvement of an an-cient WGD event. The syntenic relationships between

chromosome fragments containing ADR1 or NRG1orthologs were conserved in angiosperm genomes. Fig-ure 6 (detailed in Supplemental Table S2) shows thesynteny between ADR1- and NRG1-containing chromo-somal regions between the dicot genomes of Phaseolusvulgaris (rosid lineage) and S. lycopersicum (asterid line-age). The syntenic blocks were also detected in themonocot genome of S. bicolor, in the absence of theNRG1gene. Moreover, most RNL flanking gene pairs con-served in the duplicated blocks within the three specieshave high Ks values (.1; Supplemental Table S2), indi-cating that these genes may have resulted from one an-cient WGD event. Because the most recent WGD eventshared by monocots and dicots occurred in the commonancestor of all angiosperms (Jiao et al., 2011; Vannesteet al., 2014), the syntenic data and phylogenetic resultssupport that RNL genes separated into two lineagessometime before this time point.

Deciphering the Evolution of Arabidopsis NBS Genes

By performing phylogenetic analyses at differentevolutionary nodes, we have established a frameworkfor the evolution of angiosperm NBS genes, whichcould help decipher the evolution of NBS genes inspecific species as well as gain a much deeper view ofthe evolution of functionally characterized NBS genes.

Figure 4. Motif and exon-intron structure differences among the three NBS classes. Top, The conservation of amino acid se-quences at five motifs in the NBS domain is shown. Bottom, Proposed ancient exon-intron structure of TNL, CNL, and RNLclasses.







In total, five out of 14 An-CNL genes were inheritedby the Arabidopsis genome (Fig. 7). Gradual duplica-tions occurred in four of these An-CNLs (An-C2, -C5,-C6, and C8), with An-C5 and An-C6 expanding to afew copies. Two An-CNLs (An-C2 and An-C8) pro-duced more than 80% of Arabidopsis CNL genes,mainly occurring prior to the divergence of the Brassi-caceae family. One CNL gene (At3g07040) representedan ancient gene (An-C14) that was conserved from theangiosperm ancestor to Arabidopsis without duplica-tion. Together, the evolutionary paths of the five an-cestral CNL genes resulted in the current CNL profile ofArabidopsis genome (Fig. 7).

The structural changes were also traced using theevolutionary framework of angiospermNBS genes. Forexample, nearly all Arabidopsis CNL genes showed anintronless structure regardless of their origin (Meyerset al., 2003), which is in accordance with our proposedintronless structure for the ancestral angiosperm CNLgene. However, a group of Arabidopsis CNL geneswere determined to possess two shared introns. Allthese genes originated from a lineage that divergedfrom An-C8 (Fig. 7) in the common ancestor ofmesangiospermae, with the sister lineage remainingintronless. Based on the phylogenies of CNLs at dif-ferent taxonomic levels, the two intronswere frequently

detected in orthologs of Brassicaceae species. The firstintron also was found in orthologous genes of the le-gume family, grape, and the asterid species, but theywere absent in monocots and A. trichopoda. Therefore,the exon-intron structure detected in these ArabidopsisCNL genes gradually formed from two independentintron gain events, one in the common ancestor ofeudicots and the other acquired in the common ances-tor of Brassicaceae (Fig. 7).

Furthermore, the angiosperm evolutionary frame-work of NBS genes enabled us to determine when afunctional NBS gene likely arose during angiospermevolution. Five functional CNL genes against differentpathogens have been identified in Arabidopsis (Liuet al., 2007). According to the recovered evolutionaryhistory of Arabidopsis CNL genes in Figure 7, RPM1was the most ancient one among the five genes, possi-bly originating from the common ancestor of angio-sperms, whereas the RPP8/HRT was most recent,possibly originating after Arabidopsis speciation. Theorigin of the other three genes could be traced to thedivergence nodes of rosids, Brassicaceae, and Arabi-dopsis. Among the five genes, RPS2, RPS5, and RPM1were R genes against P. syringae, whereas RPP8/HRTand RPP13 were both against Peronospora parasitica,indicating functional R genes continuously arose

Figure 5. Comparison of the evolutionarypatterns of CNL and TNL classes in angio-sperms. A, Reconstruction of the NBS geneloss and gain events during angiospermevolution. The number of ancestral NBSgenes at each node is depicted in a solidcircle. The thickness of each terminalbranch is roughly proportional to thenumber of NBS genes in the genome or theancestral genes in the family. Gene loss andgain events at different evolutionary statesare indicated by minus and plus symbols,respectively. An asterisk after the numberindicates that the number is not directlyrecovered from phylogenetic analysis. B,Extensive expansion of CNL and TNLclasses at the K-P boundary. The numbersof CNL (red circles) and TNL (blue circles)genes at different divergence nodes above/at family level, including all angiosperms(An), mesangiospermae (Ma), monocots(Mo), eudicots (Ed), rosids (Ro), asterids(As), Poaceae (P), Solanaceae (S), Fabaceae(F), and Brassicaceae (B), were plotted toreveal their expansion history.


Shao et al.



Figure 6. Evolutionary history of RNL genes inferred from phylogenetic and syntenic analyses. Top, The inheritance of the twoRNL lineages in different angiosperm lineages. The inferred duplication is markedwith a green circle, whereas inferred losses areindicated by black circles. Bottom, Intra- and intergenome syntenic blocks containing RNL genes detected within and amongP. vulgaris, S. lycopersicum, and S. bicolor genomes.





against the same pathogen during plant evolution.Actually, apart from these genes, many other functionalgenes may have appeared in the history and then beenlost during plant evolution. However, such cases oftencould not be traced due to their absence in model plantgenomes.

The adoption of the angiosperm NBS evolutionaryframework in Arabidopsis provides new insights into

the evolutionary history of the NBS gene in this species.To facilitate the use of this framework in the analysis ofNBS genes in other plants that were not included in thisstudy, we generated hidden Markov model (HMM)profiles (Supplemental Dataset S1) for the 23 ancestralNBS lineages of angiosperms. Taken together, the an-giosperm NBS evolutionary framework established inthis study provides a fundamental reference for the

Figure 7. A few ancestral angiosperm NBS genes that are responsible in generating the current Arabidopsis CNL repertoire viagradual expansion. The ancient states of Arabidopsis at different evolutionary nodes were retrieved from the constructed NBSphylogenies. Offspring generated from the same ancestral gene at different divergence nodes are indicated by boxes of the samecolor. Black triangles indicate the birth of known functional genes. Gray arrows indicate intron gain events. The exon-intronstructure of each ancestral lineage is shown.


Shao et al.




evolutionary paths, duplication patterns, and possiblefunctionalization time frames for a deeper understand-ing of NBS genes in various plant species.

DISCUSSION

Three Ancient NBS Gene Classes in the AncestralAngiosperm Genome

In this study, we performed genome-wide analysesof NBS genes in 22 species that belonged to major an-giosperm clades. We identified 105 NBS genes from theA. trichopoda genome, and detected all threeNBS classes(TNL, CNL, and RNL) in this basal angiosperm species.Unlike the well-separated TNL and CNL classes, thephylogenetic place of RNL genes has not been fullyestablished. For a long time, RNL genes were treated asa special lineage of CNLs, but with their specific func-tion and distinct domain composition, these genes haveonly been recently differentiated (Meyers et al., 2003;Bonardi et al., 2011; Collier et al., 2011). Recent studiesinvolving legumes and the Brassicaceae have revealed asister relationship between RNLs andCNLs (Shao et al.,2014; Zhang et al., 2016). However, whether this NBSclass was derived from an ancestral CNL gene beforethe divergence of the rosid lineage or if these geneswere of a more ancient origin remained elusive. If theformer case were true, then we would expect RNLsnested as a sublineage of angiosperm CNLs; otherwise,a sister relationship would be consistently detectedwith different datasets. To clarify this issue, CNLs andRNLs were combined in phylogenetic analysis in thisstudy that supports the sister relationship stemmingfrom an ancient origin, prior to the divergence of theancestral angiosperm. Class-specific signatures werefurther detected among the CNL, RNL, and TNL clas-ses at several motifs in the conserved NBS domain (Fig.3), which was suggestive of an ancient separation of thethree NBS classes. Furthermore, distinct intron-exonstructures among the three NBS classes provide addi-tional evidence that supports the presence of three an-cient classes of NBS genes in the ancestral angiosperm.

Differential Expansion during Angiosperm EvolutionShaped TNLs and CNLs

Although both TNL and CNL genes could be datedback to at least the arrival of land plants (Akita andValkonen, 2002; Xue et al., 2012) and were inherited bythe common ancestor of angiosperms, the TNL genesare apparently outnumbered by CNLs in most angio-sperms, except for the Brassicaceae (Zhang et al., 2016).The phylogeny of TNLs also showed a lower level oftopological complexity and a relatively shorter branchlength compared to that in CNLs. All these discrep-ancies reflect different evolutionary patterns for TNLsand CNLs, and the results of this study suggest that thelong-term contraction of TNLs and gradual expansionof CNLs in the early stages of angiosperm evolution(from ;225 MYA as angiosperms diverged to around

100 MYA as asterids, rosids, and monocots diverged)may have been responsible for these differences. While,10 TNL genes were detected in rosids, eudicots, andprobably mesangiospermae and asterid nodes, thenumber of CNL genes rapidly expanded to 35 early inthe common ancestor of mesangiospermae and in-creased to 47 at the emergence of dicots (Fig. 5).

The long-term steady number of TNL genes also ex-plains their independent loss in several angiosperm lin-eages. The absence of TNLs was first reported in cerealsand major monocot lineages (Bai et al., 2002; Zhou et al.,2004; Yang et al., 2006; Li et al., 2010), then in two eudicotspecies, including an asterid, Mimulus guttatus, and abasal eudicot, Aquilegia coerulea (Collier et al., 2011). Be-cause both the basal angiosperm A. trichopoda and eudi-cots harbor TNLgenes, it is highly likely that themonocotlineage lost its TNLs near the dicot/monocot split.M. guttatus and S. indicum are from two different familiesof Lamiales, indicating that the Lamiales lineage mighthave lost its TNL genes soon after its separation with itssister lineage, Solanales. The presence of TNL genes inbasal angiosperms and in both the rosid and asterid lin-eages of eudicots suggests that the loss of TNLs in thebasal eudicot A. coerulea also stems from an independentloss event. Therefore, at least three independent gene lossevents occurred during angiosperm evolution. Further-more, the near singleton status of TNL genes at the earlystages of angiospermevolution suggests that these eventsmay have possibly occurred. Otherwise, deletion of anentire gene family consisting of dozens of members inthree independent lineages would be unlikely, especiallywhen these were dispersed in the genome.

Another interesting finding is that both CNLs andTNLs exhibit a recent expansion prior to the divergenceof angiosperm families, with more than 100 NBS genesindependently arising prior to the separation of each ofthe four families investigated in this study. Because thedivergence of these angiosperm families occurred in thePaleogene and the divergence of their higher leveltaxonomies all occurred in the Cretaceous (Fig. 1; Lavinet al., 2005; Stefanovic et al., 2009; D’Hont et al., 2012;Zhang et al., 2012; Peng et al., 2013; Yang et al., 2013;Kim et al., 2014; Wang et al., 2014; Zeng et al., 2014), theobserved intensive expansion of NBS genes in angio-sperms more likely coincided with the Cretaceous-Paleogene (K-P) boundary (;66 MYA). This findingsuggests that different lineages of angiosperms eitherunderwent a process to reinforce resistance or experi-enced increased selective pressure from environmentalpathogens during this period. The K-P boundary is animportant period when mass extinction greatly im-pacted the evolution of various species. Angiospermsunderwent drastic genomic and phenotypic changes tosurvive the climatic and environmental changes in thisperiod, including prevalently detected WGDs acrossdifferent angiosperm lineages and increased leaf veindensity (Blonder et al., 2014; Vanneste et al., 2014).A hypothesis has been proposed that the documentedbloom in fungal diversity during this period triggeredthe decline of reptiles, as well as promoted the expansion





of mammalian species, which have a more evolvedimmune system and body temperature (Vajda andMcLoughlin, 2004; Casadevall, 2012). Actually, mostfungal families diverged during the Cretaceous, andmost fungal genera diverged at the K-P boundary, suchas Botryosphaeriacea and Fusarium, which are twofungal groups with many species that cause severepathogenicity (Ma et al., 2013; Slippers et al., 2013;Hedges et al., 2015). Therefore, the fungal bloom duringthe K-P boundary may have enhanced the selectionpressure on plant R genes, which consequently accel-erated NBS gene expansions at this period, prior to thedivergence of angiosperm families.

An Ancient WGD Gave Rise to Two FunctionallyDivergent RNL Lineages

WGDs frequently occur during plant evolution (Jiaoet al., 2011; Li et al., 2015; Qian et al., 2015). However,evidence from various studies has revealed that mostNBS duplicons that originate from polyploidy arepreferentially lost during evolution, possibly to avoidfitness costs (Bergelson et al., 2001; Cannon et al., 2004;Nobuta et al., 2005; Ashfield et al., 2012; Shao et al.,2014). Therefore, NBS genes resulting from ancientWGD are rarely detected. Our phylogenetic analysishas revealed that RNL genes were derived from twoancient lineages, namely, ADR1 and NRG1, whichseparated prior to the divergence of angiosperms. Bysurveying the syntenic regions in P. vulgaris andS. lycopersicum genomes, we determined thatADR1 andNRG1 orthologous genes were located on syntenicblocks in both genomes, suggesting that the two genesresulted from an ancient WGD. Despite having lost theNRG1 orthologs, the two syntenic blocks were alsodetected in the monocot genome of S. bicolor, suggest-ing that the WGD event was shared by dicots andmonocots. Recent studies involving angiosperms haveindicated that the most recent WGD shared by mono-cots and dicots occurred in the common ancestor ofangiosperms (Jiao et al., 2011; Vanneste et al., 2014).Therefore, although syntenic blocks that contain ADR1and NRG1 orthologs were not detected in A. trichopoda,our data suggest that the WGD contributing to thegeneration of two RNL lineages occurred in the ances-tral angiosperm. This finding also is validated bythe results of phylogenetic analysis, which places theone RNL gene in A. trichopoda in the NRG1 lineage withhigh statistical support but not basal to all RNLgenes (Supplemental Fig. S9). The absence of an ADR1orthologous gene in A. trichopoda might be attributableto either the incomplete coverage of the genome se-quence or a species-specific gene loss. By searching thetranscriptome data of the 1,000 plants project (1KP;www.onekp.com), we determined that both ADR1 and NRG1orthologous genes are prevalently distributed in variousbasal angiosperms (data not shown). Taken together, boththe phylogenetic analysis and the synteny results indicatethat an ancient WGD involving the ancestral angiospermgave rise to the two RNL lineages.

Functional RNL genes have been identified in bothRNL lineages: ADR1 in Arabidopsis and NRG1 in to-bacco. However, interestingly, neither ADR1 norNRG1functions as a canonical R gene that detects specificpathogens (Collier et al., 2011). ADR1 and its paralogsare considered helper genes that transduce signalsdownstream from other R genes (Bonardi et al., 2011).Similarly, NRG1 assists the N gene in fighting againstthe Tobacco mosaic virus (Peart et al., 2005). Althoughneither ADR1 nor NRG1 acts as a pathogen detector,these are indispensable in resistance reactions. Thedistinct function, extremely conserved sequence, andgene structure suggest a downstream and crucial rolefor RNLs in disease resistance pathways. The simulta-neous loss of TNL and NRG1 orthologs in M. guttatus,A. coerulea, and the Poaceae family has been speculatedto be due to their close functional interdependence(Collier et al., 2011). It has been hypothesized that TNLgenes only transfer their signal toNRG1 genes, whereasCNL genes could transfer their signal through either theADR1 or NRG1 gene. The observed functional differ-entiation of ADR1 and NRG1 lineages suggests thatdivergence occurred after the two RNL copies emergedin the ancestral angiosperm WGD. As signal trans-duction protein-encoding genes, RNL genes are notinvolved in recognizing specific pathogens and thus arenot under the selection of pathogens. Our results showthat RNL lineages only kept a few copies in their ge-nomes, except in several rosid species. The differentevolutionary patterns among the three NBS classes re-flect their distinct functions as either pathogen detectorsor signal transducers, which further suggests that theNBS family diverged early to contribute to the plantimmune system from multiple aspects.

An Angiosperm Framework for Understanding NBSGene Evolution

Most previous studies of NBS genes were mainlyperformed on individual species, which hindered inbetter understanding NBS gene evolution at the whole-angiosperm scale. The results reported in this study(covering.6,000 NBS genes from 22 genomes) providea fundamental phylogenetic framework of NBS genesin angiosperms. Because the whole dataset is too largefor phylogenetic analysis and subsequent gene loss/gain reconcile, we used a step-by-step method forphylogenetic analysis as described earlier. In suchprocedure, the phylogenetic relationships among allangiospermNBS genes could not be fully displayed in asingle tree, and supporting values for some internalnodes may decrease because only a few representativesequences were used in the construction of high taxo-nomic level NBS phylogenies. However, our data sug-gest that the results from this approach are generallyrobust, and the taxonomy of NBS genes often were notaffected by different datasets used. For example, one ofour major findings, the sister relationship of CNL andRNL, is well supported in nTNL phylogenies con-structed by different datasets (Supplemental Figs. S1,


Shao et al.



http://www.onekp.com

http://www.onekp.com



S3, S4, S6, and S9). Furthermore, during the procedure,with more data incorporated, the shared introns ina CNL lineage (An-C8; Fig. 7) also were in perfectagreement with its phylogenetic relationships. There-fore, to a large extent, the phylogenetic framework ofangiosperm NBS provided in this study is reliable forthe estimation of the evolutionary history of this largegene family. Using this framework as reference, NBSgene profiles in a particular plant genome could be fullyclassified to better understand the evolutionary dy-namics. For example, the 52 CNL genes present in theArabidopsis genome were traced to five ancestral CNLgenes in the common ancestor of angiosperms (Fig. 7).Gradual changes in gene content in these ancestrallineages were observed at five intermediate divergencenodes and the node to Arabidopsis speciation, whichnot only revealed different fates for the five ancestralgenes during their evolution but also uncovered theextent of gene expansion at different nodes.Furthermore, the five known Arabidopsis functional

CNL genes were traced to four ancient NBS genes andarose from different angiosperm nodes (Fig. 7). Nota-bly, two of the five R genes, RPS2 and RPM1, defendagainst P. syringae by guarding the Arabidopsis RIN4protein. The state change of RIN4 that is monitored byRPS2 and RPM1 also are modified by two different ef-fectors of P. syringae (Mackey et al., 2002; Mackey et al.,2003), suggesting a long-term coevolutionary arms raceamong Arabidopsis R genes and P. syringae (Fig. 7).Angiosperms may have first evolved the RPM1 gene asa mechanism against P. syringae infection by detectingeffector (AvrRpm1)-induced phosphorylation of RIN4.On the other hand, P. syringae evolved another effector(AvrRpt2) to cleave the RIN4 protein and evade theactivation of RPM1, and plants further evolved anotherCNL gene (RPS2, probably prior to the node of rosiddivergence) to monitor the cleavage of RIN4 and acti-vate the immune system. The long history of R genesguarding RIN4 also suggests that this protein is anancient target of pathogens, which in turn unravels itsimportant function in plant immune signal transduc-tion (Liu et al., 2009; Hou et al., 2011). It should be no-ticed that although we traced the emergence time ofthese genes in angiosperm lineages, we were not able toaccurately determine when they obtained specificfunctions. The resistance specificity may have existedprior to the duplication event but subsequently beenlost from some duplicated copy; alternatively, thespecificity may have arisen after the duplication event.In conclusion, by identifying more than 6,000 NBS

genes from 22 angiosperm species, this study has suc-cessfully constructed the phylogeny of angiospermNBS genes. A total of 23 ancestral NBS genes wereidentified in angiosperm ancestors, and three distinctNBS classes were revealed. Although exhibiting dif-ferent evolutionary patterns, both TNL and CNL genesunderwent extensive expansion at the K-P boundary. Incontrast, RNL genes conservatively evolved duringangiosperm evolution, which mirrors their distinctpathogen detection or signal transduction roles in plant

disease resistance. The angiosperm level NBS phylo-genetic framework constructed in this study providesa fundamental reference for further studies on NBSgenes.

MATERIALS AND METHODS

Database

The genomic sequences, annotations, and gene models for 13 out of 22 ge-nomes (except four Fabaceae and five Brassicaceae genomes) were used in thisstudy. Among these, data on Vitis vinifera, Solanum lycopersicum, Solanumtuberosum, Brachypodium distachyon, Setaria italica, Sorghum bicolor, Oryza sativa,and Zea mays were downloaded from the Phytozome database (http://www.phytozome.net/; Goodstein et al., 2012). Files of Amborella trichopoda, Capsicumannuum, Sesamum indicum, Phyllostachys heterocycla, and Musa acuminata weredownloaded from the Pepper Genome Database (http://peppersequence.genomics.cn/page/species/index.jsp; Qin et al., 2014), Sinbase (http://ocri-genomics.org/Sinbase/manual.htm; Wang et al., 2014), BambooGDB (http://www.bamboogdb.org/page/download.jsp; Peng et al., 2013), Banana GenomeHub (http://banana-genome.cirad.fr; D’Hont et al., 2012), and the Amborellagenome database (http://www.amborella.org; Amborella-Genome-Project,2013), respectively.

Identification of NBS-Encoding Genes

The NBS genes in the 13 genomes were identified as described in previousstudies (Shao et al., 2014; Shao et al., 2015; Zhang et al., 2016). Briefly, a HMMsearch was first performed for protein sequences in each genome usinghmmer3.0 (http://hmmer.org) using default parameter settings, with theNB-ARC domain (Pfam: PF00931) as query. The amino acid sequence of theNB-ARC domain was then used to run a BLASTp search against all proteinsequences in each genome, with the threshold expectation value set to 1.0. Allhits obtained using HMM or BLAST searches were then merged together, andthe redundant hits were removed. The remaining sequences were furthersubjected to Pfam analysis (http://pfam.sanger.ac.uk/) to further exclude se-quences that do not have a detectable NBS domain at an E value of 1024. Whentwo or more transcripts were annotated for a gene from alternative splicing, thelongest form with an NBS domain was selected. Finally, the identified NBSdomain-encoding genes were examined to see whether these encode the TIR,RPW8, CC, or LRR domains using Pfam and COILS analyses (Lupas et al.,1991). The exon position and intron phase of each gene were transformed fromthe gff3 file of the reference genome.

Sequence Alignment and Phylogenetic Analysis

Sequence alignment and phylogenetic analysis were performed as describedpreviously (Shao et al., 2014). Amino acid sequences of the NBS domain werealigned using ClustalW (Edgar, 2004) with default options (Thompson et al.,1994) and then manually corrected in MEGA 5.0 (Tamura et al., 2011). Theresulting amino acid sequence alignments were used to guide the alignments ofthe nucleotide coding sequences. Genes with very short NBS domains wereeliminated from the matrix because these interfered in fine alignment andsubsequent construction of phylogenetic trees. All alignments used for phylo-genetic analysis in the presented study are provided in Supplemental DatasetS2. Phylogenetic analysis was conducted using a maximum likelihood methodas provided in the Phyml program (Guindon and Gascuel, 2003), which wasintegrated in the Seaview package (Gouy et al., 2010). Branch support wasassessed with the aLRT statistic (Guindon et al., 2010). Gene losses and gains ateach divergence node were reconciled using Notung software (Stolzer et al.,2012). The number of ancestral NBS genes at different divergence nodes wasdetermined using monophyletic NBS lineages as described elsewhere (Shaoet al., 2014).

Because merging all data generates large datasets for both nTNL and TNLclasses,weadopteda step-by-step strategy to recover the full pictureofNBSgeneevolution in angiosperms. Generally, phylogenetic analysis was first performedat a low taxonomy level, mainly containing NBS genes from a few genomes.Then, representative NBS sequences from these angiosperm lineages werecompiled with NBS genes in more plant genomes to construct the NBS genephylogenies at higher taxonomy levels, finally reaching the whole-angiospermlevel.





http://www.phytozome.net/

http://www.phytozome.net/

http://peppersequence.genomics.cn/page/species/index.jsp

http://peppersequence.genomics.cn/page/species/index.jsp

http://ocri-genomics.org/Sinbase/manual.htm

http://ocri-genomics.org/Sinbase/manual.htm

http://www.bamboogdb.org/page/download.jsp

http://www.bamboogdb.org/page/download.jsp

http://banana-genome.cirad.fr

http://www.amborella.org

http://hmmer.org

http://pfam.sanger.ac.uk/




The conserved motifs within the NBS domain of three NBS classes wereidentified from the amino acid sequence using Multiple Expectation Maximi-zation forMotif Elicitation (Bailey and Elkan, 1995), and the generated sequencelogos were manually modified to compare their differences among differentclasses.

Synteny Analyses

Genome synteny between duplicate genepairswas examinedby searching thePlantGenomeDuplicationDatabase (http://chibba.agtec.uga.edu/duplication/;Tang et al., 2008).

Supplemental Data

The following supplemental materials are available.

Supplemental Figure S1. Phylogenetic tree of monocot nTNL genes.

Supplemental Figure S2. Phylogenetic tree of asterid nTNL genes.

Supplemental Figure S3. Phylogenetic tree of asterid TNL genes.

Supplemental Figure S4. Phylogenetic tree of rosid nTNL genes.

Supplemental Figure S5. Phylogenetic tree of rosid TNL genes.

Supplemental Figure S6. Phylogenetic tree of dicot nTNL genes.

Supplemental Figure S7. Phylogenetic tree of dicot TNL genes.

Supplemental Figure S8. Phylogenetic tree of mesangiospermae nTNLgenes.

Supplemental Figure S9. Phylogenetic tree of angiosperm nTNL genes.

Supplemental Figure S10. Phylogenetic tree of angiosperm TNL genes.

Supplemental Table S1. NBS-encoding genes detected in angiosperm ge-nomes examined in this study.

Supplemental Table S2. Inter- or intraspecies RNL-containing syntenicblocks.

Supplemental Dataset S1. HMM profiles for the 23 ancestral NBS lineages.

Supplemental Dataset S2. Alignments used for phylogenetic analysis inthis study.

Received September 18, 2015; accepted February 1, 2016; published February 2,2016.

LITERATURE CITED

Akita M, Valkonen JP (2002) A novel gene family in moss (Physcomitrellapatens) shows sequence homology and a phylogenetic relationship with theTIR-NBS class of plant disease resistance genes. J Mol Evol 55: 595–605

Amborella Genome Project (2013) The Amborella genome and the evolu-tion of flowering plants. Science 342: 1241089

Ameline-Torregrosa C, Wang BB, O’Bleness MS, Deshpande S, Zhu H,Roe B, Young ND, Cannon SB (2008) Identification and characteriza-tion of nucleotide-binding site-leucine-rich repeat genes in the modelplant Medicago truncatula. Plant Physiol 146: 5–21

Andolfo G, Sanseverino W, Rombauts S, Van de Peer Y, Bradeen JM,Carputo D, Frusciante L, Ercolano MR (2013) Overview of tomato(Solanum lycopersicum) candidate pathogen recognition genes revealsimportant Solanum R locus dynamics. New Phytol 197: 223–237

Ashfield T, Egan AN, Pfeil BE, Chen NW, Podicheti R, Ratnaparkhe MB,Ameline-Torregrosa C, Denny R, Cannon S, Doyle JJ, et al (2012)Evolution of a complex disease resistance gene cluster in diploid Phaseolusand tetraploid Glycine. Plant Physiol 159: 336–354

Ashfield T, Redditt T, Russell A, Kessens R, Rodibaugh N, Galloway L, KangQ, Podicheti R, Innes RW (2014) Evolutionary relationship of disease resis-tance genes in soybean and Arabidopsis specific for the Pseudomonas sy-ringae effectors AvrB and AvrRpm1. Plant Physiol 166: 235–251

Bai J, Pennill LA, Ning J, Lee SW, Ramalingam J, Webb CA, Zhao B, SunQ, Nelson JC, Leach JE, et al (2002) Diversity in nucleotide binding site-leucine-rich repeat genes in cereals. Genome Res 12: 1871–1884

Bailey TL, Elkan C (1995) The value of prior knowledge in discoveringmotifs with MEME. Proc Int Conf Intell Syst Mol Biol 3: 21–29

Bent AF, Kunkel BN, Dahlbeck D, Brown KL, Schmidt R, Giraudat J,Leung J, Staskawicz BJ (1994) RPS2 of Arabidopsis thaliana: a leucine-rich repeat class of plant disease resistance genes. Science 265: 1856–1860

Bergelson J, Kreitman M, Stahl EA, Tian D (2001) Evolutionary dynamicsof plant R-genes. Science 292: 2281–2285

Blonder B, Royer DL, Johnson KR, Miller I, Enquist BJ (2014) Plantecological strategies shift across the Cretaceous-Paleogene boundary.PLoS Biol 12: e1001949

Bonardi V, Tang S, Stallmann A, Roberts M, Cherkis K, Dangl JL (2011)Expanded functions for a family of plant intracellular immune receptorsbeyond specific recognition of pathogen effectors. Proc Natl Acad SciUSA 108: 16463–16468

Bremer B, Bremer K, Chase MW, Fay MF, Reveal JL, Soltis DE, Soltis PS,Stevens PF, Anderberg AA, Moore MJ, et al (2009) An update of theAngiosperm Phylogeny Group classification for the orders and familiesof flowering plants: APG III. Bot J Linn Soc 161: 105–121

Cannon SB, Mitra A, Baumgarten A, Young ND, May G (2004) The rolesof segmental and tandem gene duplication in the evolution of large genefamilies in Arabidopsis thaliana. BMC Plant Biol 4: 10

Cannon SB, Zhu H, Baumgarten AM, Spangler R, May G, Cook DR,Young ND (2002) Diversity, distribution, and ancient taxonomic rela-tionships within the TIR and non-TIR NBS-LRR resistance gene sub-families. J Mol Evol 54: 548–562

Casadevall A (2012) Fungi and the rise of mammals. PLoS Pathog 8:e1002808

Collier SM, Hamel LP, Moffett P (2011) Cell death mediated by the N-terminal domains of a unique and highly conserved class of NB-LRRprotein. Mol Plant Microbe Interact 24: 918–931

D’Hont A, Denoeud F, Aury JM, Baurens FC, Carreel F, Garsmeur O, Noel B,Bocs S, Droc G, Rouard M, et al (2012) The banana (Musa acuminata) ge-nome and the evolution of monocotyledonous plants. Nature 488: 213–217

DeYoung BJ, Innes RW (2006) Plant NBS-LRR proteins in pathogen sens-ing and host defense. Nat Immunol 7: 1243–1249

Friedman AR, Baker BJ (2007) The evolution of resistance genes in multi-protein plant resistance systems. Curr Opin Genet Dev 17: 493–499

Goodstein DM, Shu S, Howson R, Neupane R, Hayes RD, Fazo J, Mitros T,Dirks W, Hellsten U, Putnam N, et al (2012) Phytozome: a comparativeplatform for green plant genomics. Nucleic Acids Res 40: D1178–D1186

Gouy M, Guindon S, Gascuel O (2010) SeaView version 4: a multiplatformgraphical user interface for sequence alignment and phylogenetic treebuilding. Mol Biol Evol 27: 221–224

Guindon S, Dufayard JF, Lefort V, Anisimova M, Hordijk W, Gascuel O (2010)New algorithms and methods to estimate maximum-likelihood phylogenies:assessing the performance of PhyML 3.0. Syst Biol 59: 307–321

Guindon S, Gascuel O (2003) A simple, fast, and accurate algorithm to estimatelarge phylogenies by maximum likelihood. Syst Biol 52: 696–704

Guo YL, Fitz J, Schneeberger K, Ossowski S, Cao J, Weigel D (2011)Genome-wide comparison of nucleotide-binding site-leucine-richrepeat-encoding genes in Arabidopsis. Plant Physiol 157: 757–769

Hedges SB, Marin J, Suleski M, Paymer M, Kumar S (2015) Tree of lifereveals clock-like speciation and diversification. Mol Biol Evol 32: 835–845

Hou S, Yang Y, Wu D, Zhang C (2011) Plant immunity: evolutionary in-sights from PBS1, Pto, and RIN4. Plant Signal Behav 6: 794–799

Jiao Y, Wickett NJ, Ayyampalayam S, Chanderbali AS, Landherr L, RalphPE, Tomsho LP, Hu Y, Liang H, Soltis PS, et al (2011) Ancestral poly-ploidy in seed plants and angiosperms. Nature 473: 97–100

Kim S, Park M, Yeom SI, Kim YM, Lee JM, Lee HA, Seo E, Choi J, Cheong K,Kim KT, et al (2014) Genome sequence of the hot pepper provides insightsinto the evolution of pungency in Capsicum species. Nat Genet 46: 270–278

Kohler A, Rinaldi C, Duplessis S, Baucher M, Geelen D, Duchaussoy F,Meyers BC, Boerjan W, Martin F (2008) Genome-wide identification ofNBS resistance genes in Populus trichocarpa. Plant Mol Biol 66: 619–636

Lavin M, Herendeen PS, Wojciechowski MF (2005) Evolutionary ratesanalysis of Leguminosae implicates a rapid diversification of lineagesduring the tertiary. Syst Biol 54: 575–594

Li J, Ding J, Zhang W, Zhang Y, Tang P, Chen JQ, Tian D, Yang S (2010)Unique evolutionary pattern of numbers of gramineous NBS-LRR genes.Mol Genet Genomics 283: 427–438

Li Q, Zhang N, Zhang L, Ma H (2015) Differential evolution of members ofthe rhomboid gene family with conservative and divergent patterns.New Phytol 206: 368–380


Shao et al.


http://chibba.agtec.uga.edu/duplication/
















Lin X, Zhang Y, Kuang H, Chen J (2013) Frequent loss of lineages anddeficient duplications accounted for low copy number of disease resis-tance genes in Cucurbitaceae. BMC Genomics 14: 335

Liu J, Elmore JM, Fuglsang AT, Palmgren MG, Staskawicz BJ, Coaker G(2009) RIN4 functions with plasma membrane H+-ATPases to regulatestomatal apertures during pathogen attack. PLoS Biol 7: e1000139

Liu J, Liu X, Dai L, Wang G (2007) Recent progress in elucidating thestructure, function and evolution of disease resistance genes in plants.J Genet Genomics 34: 765–776

Lozano R, Ponce O, Ramirez M, Mostajo N, Orjeda G (2012) Genome-wide identification and mapping of NBS-encoding resistance genes inSolanum tuberosum group phureja. PLoS One 7: e34775

Luo S, Zhang Y, Hu Q, Chen J, Li K, Lu C, Liu H, Wang W, Kuang H(2012) Dynamic nucleotide-binding site and leucine-rich repeat-encodinggenes in the grass family. Plant Physiol 159: 197–210

Lupas A, Van Dyke M, Stock J (1991) Predicting coiled coils from proteinsequences. Science 252: 1162–1164

Ma LJ, Geiser DM, Proctor RH, Rooney AP, O’Donnell K, Trail F,Gardiner DM, Manners JM, Kazan K (2013) Fusarium pathogenomics.Annu Rev Microbiol 67: 399–416

Mackey D, Belkhadir Y, Alonso JM, Ecker JR, Dangl JL (2003) Arabi-dopsis RIN4 is a target of the type III virulence effector AvrRpt2 andmodulates RPS2-mediated resistance. Cell 112: 379–389

Mackey D, Holt BF III, Wiig A, Dangl JL (2002) RIN4 interacts withPseudomonas syringae type III effector molecules and is required forRPM1-mediated resistance in Arabidopsis. Cell 108: 743–754

Malacarne G, Perazzolli M, Cestaro A, Sterck L, Fontana P, Van de PeerY, Viola R, Velasco R, Salamini F (2012) Deconstruction of the (paleo)polyploid grapevine genome based on the analysis of transpositionevents involving NBS resistance genes. PLoS One 7: e29762

McHale L, Tan X, Koehl P, Michelmore RW (2006) Plant NBS-LRR pro-teins: adaptable guards. Genome Biol 7: 212

Meyers BC, Dickerman AW, Michelmore RW, Sivaramakrishnan S,Sobral BW, Young ND (1999) Plant disease resistance genes encodemembers of an ancient and diverse protein family within the nucleotide-binding superfamily. Plant J 20: 317–332

Meyers BC, Kaushik S, Nandety RS (2005) Evolving disease resistancegenes. Curr Opin Plant Biol 8: 129–134

Meyers BC, Kozik A, Griego A, Kuang H, Michelmore RW (2003)Genome-wide analysis of NBS-LRR-encoding genes in Arabidopsis.Plant Cell 15: 809–834

Monosi B, Wisser RJ, Pennill L, Hulbert SH (2004) Full-genome analysisof resistance gene homologues in rice. Theor Appl Genet 109: 1434–1447

Mun JH, Yu HJ, Park S, Park BS (2009) Genome-wide identification ofNBS-encoding resistance genes in Brassica rapa. Mol Genet Genomics282: 617–631

Nobuta K, Ashfield T, Kim S, Innes RW (2005) Diversification of non-TIRclass NB-LRR genes in relation to whole-genome duplication events inArabidopsis. Mol Plant Microbe Interact 18: 103–109

Peart JR, Mestre P, Lu R, Malcuit I, Baulcombe DC (2005) NRG1, a CC-NB-LRR protein, together with N, a TIR-NB-LRR protein, mediates re-sistance against tobacco mosaic virus. Curr Biol 15: 968–973

Peng Z, Lu Y, Li L, Zhao Q, Feng Q, Gao Z, Lu H, Hu T, Yao N, Liu K, et al(2013) The draft genome of the fast-growing non-timber forest speciesmoso bamboo (Phyllostachys heterocycla). Nat Genet 45: 456–461

Porter BW, Paidi M, Ming R, Alam M, Nishijima WT, Zhu YJ (2009)Genome-wide analysis of Carica papaya reveals a small NBS resistancegene family. Mol Genet Genomics 281: 609–626

Qian S, Wang Y, Ma H, Zhang L (2015) Expansion and functional divergence ofJumonji C-containing histone demethylases: significance of duplications inancestral angiosperms and vertebrates. Plant Physiol 168: 1321–1337

Qin C, Yu C, Shen Y, Fang X, Chen L, Min J, Cheng J, Zhao S, Xu M, LuoY, et al (2014) Whole-genome sequencing of cultivated and wild peppersprovides insights into Capsicum domestication and specialization. ProcNatl Acad Sci USA 111: 5135–5140

Qiu YL, Cho Y, Cox JC, Palmer JD (1998) The gain of three mitochondrial intronsidentifies liverworts as the earliest land plants. Nature 394: 671–674

Sánchez D, Ganfornina MD, Gutiérrez G, Marín A (2003) Exon-intronstructure and evolution of the Lipocalin gene family. Mol Biol Evol 20:775–783

Shao ZQ, Zhang YM, Hang YY, Xue JY, Zhou GC, Wu P, Wu XY, Wu XZ,Wang Q, Wang B, et al (2014) Long-term evolution of nucleotide-

binding site-leucine-rich repeat genes: understanding gained from andbeyond the legume family. Plant Physiol 166: 217–234

Shao ZQ, Zhang YM, Wang B, Chen JQ (2015) Computational identifica-tion of microRNA-targeted nucleotide-binding site-leucine-rich repeatgenes in plants. Bio Protoc 5: e1637

Slippers B, Boissin E, Phillips AJ, Groenewald JZ, Lombard L, WingfieldMJ, Postma A, Burgess T, Crous PW (2013) Phylogenetic lineages in theBotryosphaeriales: a systematic and evolutionary framework. StudMycol 76: 31–49

Stefanovic S, Pfeil BE, Palmer JD, Doyle JJ (2009) Relationships amongphaseoloid legumes based on sequences from eight chloroplast regions.Syst Bot 34: 115–128

Stolzer M, Lai H, Xu M, Sathaye D, Vernot B, Durand D (2012) Inferringduplications, losses, transfers and incomplete lineage sorting withnonbinary species trees. Bioinformatics 28: i409–i415

Tamura K, Peterson D, Peterson N, Stecher G, Nei M, Kumar S (2011)MEGA5: molecular evolutionary genetics analysis using maximumlikelihood, evolutionary distance, and maximum parsimony methods.Mol Biol Evol 28: 2731–2739

Tang H, Bowers JE, Wang X, Ming R, Alam M, Paterson AH (2008) Syn-teny and collinearity in plant genomes. Science 320: 486–488

Tarr DE, Alexander HM (2009) TIR-NBS-LRR genes are rare in monocots:evidence from diverse monocot orders. BMC Res Notes 2: 197

Thompson JD, Higgins DG, Gibson TJ (1994) CLUSTAL W: improving thesensitivity of progressive multiple sequence alignment through se-quence weighting, position-specific gap penalties and weight matrixchoice. Nucleic Acids Res 22: 4673–4680

Vajda V, McLoughlin S (2004) Fungal proliferation at the Cretaceous-Tertiary boundary. Science 303: 1489

Vanneste K, Baele G, Maere S, Van de Peer Y (2014) Analysis of 41 plant ge-nomes supports a wave of successful genome duplications in association withthe Cretaceous-Paleogene boundary. Genome Res 24: 1334–1347

Wang L, Yu S, Tong C, Zhao Y, Liu Y, Song C, Zhang Y, Zhang X, Wang Y,Hua W, et al (2014) Genome sequencing of the high oil crop sesameprovides insight into oil biosynthesis. Genome Biol 15: R39

Xiao S, Ellwood S, Calis O, Patrick E, Li T, Coleman M, Turner JG (2001)Broad-spectrum mildew resistance in Arabidopsis thaliana mediated byRPW8. Science 291: 118–120

Xue JY, Wang Y, Wu P, Wang Q, Yang LT, Pan XH, Wang B, Chen JQ(2012) A primary survey on bryophyte species reveals two novel classesof nucleotide-binding site (NBS) genes. PLoS One 7: e36700

Yang R, Jarvis DE, Chen H, Beilstein MA, Grimwood J, Jenkins J, Shu S,Prochnik S, Xin M, Ma C, et al (2013) The reference genome of thehalophytic plant Eutrema salsugineum. Front Plant Sci 4: 46

Yang S, Feng Z, Zhang X, Jiang K, Jin X, Hang Y, Chen JQ, Tian D (2006)Genome-wide investigation on the genetic variations of rice disease re-sistance genes. Plant Mol Biol 62: 181–193

Yang S, Zhang X, Yue JX, Tian D, Chen JQ (2008) Recent duplicationsdominate NBS-encoding gene expansion in two woody species. MolGenet Genomics 280: 187–198

Zeng L, Zhang Q, Sun R, Kong H, Zhang N, Ma H (2014) Resolution ofdeep angiosperm phylogeny using conserved nuclear genes and esti-mates of early divergence times. Nat Commun 5: 4956

Zhang G, Liu X, Quan Z, Cheng S, Xu X, Pan S, Xie M, Zeng P, Yue Z,Wang W, et al (2012) Genome sequence of foxtail millet (Setaria italica)provides insights into grass evolution and biofuel potential. Nat Bio-technol 30: 549–554

Zhang M, Wu YH, Lee MK, Liu YH, Rong Y, Santos TS, Wu C, Xie F,Nelson RL, Zhang HB (2010) Numbers of genes in the NBS and RLKfamilies vary by more than four-fold within a plant species and areregulated by multiple factors. Nucleic Acids Res 38: 6513–6525

Zhang X, Feng Y, Cheng H, Tian D, Yang S, Chen JQ (2011) Relativeevolutionary rates of NBS-encoding genes revealed by soybean seg-mental duplication. Mol Genet Genomics 285: 79–90

Zhang YM, Shao ZQ, Wang Q, Hang YY, Xue JY, Wang B, Chen JQ (2016)Uncovering the dynamic evolution of nucleotide-binding site-leucine-rich repeat (NBS-LRR) genes in Brassicaceae. J Integr Plant Biol 58:165–177

Zhou T, Wang Y, Chen JQ, Araki H, Jing Z, Jiang K, Shen J, Tian D (2004)Genome-wide identification of NBS genes in japonica rice reveals sig-nificant expansion of divergent non-TIR NBS-LRR genes. Mol GenetGenomics 271: 402–415





large-scale analyses of angiosperm nucleotide-binding site … · large-scale analyses of...

Documents