Identification and characterization of histone modification genes in T. aestivum, H. vulgare, S. bicolor, S. viridis, S. italic and Z. mays
In Arabidopsis and rice, there were respectively 102 and 92 HMs, including 48 and 42 HMTs, 24 and 24 HDMs, 12 and 8 HATs, and 18 and 18 HDACs (Fig. 1a). In total, 245, 72, 84, 93, 90 and 90 HMs were respectively identified in T. aestivum, H. vulgare, S. bicolor, S. viridis, S. italic and Z. mays (Fig. 1a and 1b). The number of HMTs, HDMs, HATs and HDACs were broadly equal among gramineae species, except for T. aestivum (Fig. 1a). Wheat HMs (HMTs, HDMs, HATs and HDACs) was severally about 2.4- and 2.7-folds as many as that of Arabidopsis and rice ones (Fig. 1a). HMTs, HDMs, HATs and HDACs could be separately classified into SDGs-PRMTs, HDMAs-JMJs, HAG-HAM-HAC-HAF, and HAD-SRT-HDT. There were 30 to 117 SDGs, 1 to 7 PRMTs, 3 to 12 HDMAs, 11 to 48 JMJs, 1 to 6 HAGs, 1 to 3 HAMs, 3–10 HACs, 1–6 HAFs, 11–32 HDAs, 1–6 SRTs and 1–5 HDTs among all species (Fig. 1b). Hexaploid wheat containing A, B and D sub-genomes, the HMs on each wheat chromosome were counted to identify their distribution in each sub-genome. There were severally 3–8 T. aestivum SDGs (TaSDGs), 0–1 TaPRMT-TaHAG-TaHAM-TaSRT-TaHDT, 0–2 TaHDMAs-TaHACs-TaHAFs, 1–4 TaJMJs, and 0–3 TaHDAs on chromosome 1A-7D (Fig. 1c). However, one TaHAG and TaSRT were located in unknown chromosome (Fig. 1c).
These gramineae HMs were respectively named based on their chromosomal location (Fig. S1). Wheat chromosome 5A (Ta5A) contained the most of HMs, followed by Ta2D (Fig. <link rid="fig1">1</link>c and S1-1). Most of barley HMs (HvHMs) were found in the longest chromosome 2 (chr2H), and HvSDG29, HvSDG30 and HvHDA14 were on unknown chromosome (Fig. S1-2). Sorghum SDGs (SbSDGs) were most numerous among all HMs genes, 38 SbSDGs were distributed throughout nine chromosomes, and chromosome 2 contained most of SbHMs genes (Fig. S1-3). S. viridis HMs (SvHMs) were unevenly distributed on chromosomes 1–9; For example, there were 17 SvHMs genes on chromosome 1, while only two SvHMs (SvSDG36 and SvJMJ13) on chromosome 8 (Fig. S1-4). S. italic HMs genes were found from chromosome 1 to 9, and chromosome 1 and 9 respectively shared the highest and smallest gene density (Fig. S1-5). Maize chromosome 1–10 respectively contained 4–13 maize HMs (ZmHMs) genes (Fig. S1-6). The detail information of gramineae HMs genes was list in Table S1. Their coding regions (CDS) lengths were from 195 (HvHDT3) to 7008 (AtSDG2) bp, with the deduced polypeptides ranged from 64 to 2335 amino acids (aa).
Conserved domain and phylogenetic analysis of HMs genes
Conserved domains of HMs genes were investigated, various domains were presented in different kinds of HMs genes (Fig. S2). A total of 35 conserved motifs were identified in Arabidopsis and rice HMs proteins (Fig. S2-1). There were one to seven domains to be found in AtSDGs and OsSDGs proteins. For example, AtSDG7 only contained one SET_SETD2-like domain, three motifs (PHA03247, DUF4689 and SRA) were uncovered in OsSDG704, and OsSDG723 shared seven conserved domains, including SET_SETD1-like, ePHD_ATX1_2_like, PHD_ATX1_2_like, PWWP, FYRC, FYRN and TUDOR. Fifty-one elements were identified in TaSDGs, and most of these elements were the same as the ones discovered in AtSDGs and OsSDGs (Fig. S2-2). However, 15 specific elements, such as DUF5585, Jas, NupH_GANP, PostSET and so on, were identified in several TaSDGs. One to six conserved motifs were uncovered in each HvSDG protein, while we only found one distinctive motif (AMN1) in HvSDG26 compared with reference SDGs (AtSDGs and OsSDGs) (Fig. S2-3). Totally, 40 domains were identified in SbSDGs, and almost all of them were as the same as the ones identified in Arabidopsis and rice homologous proteins, while jas, CAF-1_p150, zf-TRM13_CCCH, GYF and PLN02983 were the exceptions (Fig. S2-4). Most of SvSDGs included more than two domains, and Jas, PRK12678, zf-TRM13_CCCH, Nucleo_P87, RSRP and GYF were characterized as peculiar motifs of SvSDGs (Fig. S2-5). All of SiSDGs contained two conserved domains, except for SiSDG1-2, SiSDG6-8, SiSDG10, SiSDG15, SiSDG20-21, SiSDG23, SiSDG26-29 and SiSDG33-34; And there were specific domains in SiSDG6 and SiSDG33 (Fig. S2-6). In maize SDGs, one to six conserved motifs were found, and eight distinctive (PRK07003, p450, jas, PHA03132, LbR-like, RSRP, F-box-like and HCP_like) domains were respectively identified in ZmSDG5, ZmSDG11, ZmSDG12, ZmSDG21, ZmSDG27 and ZmSDG39 (Fig. S2-7). About half of PRMTs proteins included PRMT_TIM, PRMT5 and PRMT_C domains, and all gramineaes PRMTs, except for ZmPRMT1, shared same structures with AtPRMT15 and OsPRMT708 (Fig. S2-8).
HDMAs contained one PLN03000, PLN02529, PLN02328, or PLN02976, respectively, and PLN02976 was the longest one (Fig. S2-9). Almost all of JMJs proteins included JmjC or JmjN, and 12–29 conserved structures were identified in both model and gramineae plants, totally (Fig. S2-10, 2–11, 2–12, 2–13, 2–14, 2–15 and 2–16). Specific motifs were found in gramineae JMJs. For example, there were respectively 14 distinctive domains in TaJMJ3, TaJMJ8, TaJMJ12, TaJMJ17, TaJMJ21, TaJMJ23, TaJMJ28, TaJMJ34, TaJMJ37, TaJMJ39, TaJMJ42, TaJMJ44 and TaJMJ48 (Fig. S2-11); zf-C5HC2 was characterized as barley peculiar motif in HvJMJ2 and HvJMJ3 (Fig. S2-12); Eight characteristic elements were identified in SbJMJ1, SbJMJ4, SbJMJ5, SbJMJ8, SbJMJ9 and SbJMJ17 (Fig. S2-13); CDS45 in SvJMJ1, ARID-TNG2-zf-RING_2 in SvJMJ8, cf-H2C2_2 in SvJMJ16 and ZZ in almost all of SvJMJs were identified as specific S. viridis motifs (Fig. S2-14); SiJMJ1 and SiJMJ9 contained several specific motifs (Fig. S2-15); And zf-H2C2 was identified as characteristic motif in ZmJMJ15 (Fig. S2-16).
There was Bromo_gcn5_like, COG5076, or ELP3 motif in all HAGs, except for AtHAG2, OsHAG704, SvHAG2, SvHAG4 and SiHAG1; And Hat1_N, NAT_SF, PLN02706 and acetyltransf_1 were also found in HAGs (Fig. S2-17). Only one domain (PLN00104) was found in all HAMs proteins (Fig. S2-18). Twelve conserved motifs, such as HAT_KAT11, PHD_HAC_like and ZnF_TAZ, were in HACs proteins (Fig. S2-19). All HAFs were made up of three to five conserved motifs, and they generally shared similar structures (Fig. S2-20).
Almost all of Arabidopsis and rice HDAs proteins contained one conserved domain (HDACs, Arginase_HDAC or Smc), while Smc was only found in AtHDA18 (Fig. S2-21). Like model plant HDAs proteins, all gramineae HDAs contained one HDACs or Arginase_HDAC domain, while LRAT was also in ZmHDA10 and ZmHDA11 (Fig. S2-22, 2–23, 2–24, 2–25, 2–26 and 2–27). One SIRT4, SIRT7 or SIR2 domain was respectively uncovered in each SRT protein (Fig. S2-28). All HDTs proteins, except for HvHDT1, HvHDT3 and ZmHDT1, included NPL domain; While other domains (PRK13108, ZnF_U1, lambda-1, zf-C2H2_jaz, TFⅡF_alpha, PRK13808 and zf-C2H2_6) were also found in several HDTs proteins (Fig. S2-29).
In order to identify evolutionary relationships of HMs genes, unrooted phylogenetic trees were constructed (Fig. S3). AtHMTs (except AtSDG41), OsHMTs and TaHMTs were classed into groups A-E which can be further subdivided (Fig. S3-1). For example, AtPRMTs, OsPRMTs and TaPRMTs were in class B; TaPRMTs, AtPRMT15 and OsPRMT708 were clustered in subclass b1, and other ones were found in subclass b2. All Arabidopsis, rice and barley HMTs genes were classified into eight groups. All SDGs were clustered together in class A-D and F-H, and PRMTs were divided into subgroup e1 and e2 in group E (Fig. S3-2). In Fig. S3-3, AtPRMTs, OsPRMTs and SbPRMTs were clustered together in group A which could be divided into subgroup a1 and a2, and SDGs (except for AtSDG41) were identified in groups B-G. AtPRMTs, OsPRMTs and SvRMTs were grouped in either class A or B, and the other HMTs were in class C-H (Fig. S3-4). All Arabidopsis, rice and S. italic SDGs got together in group B-E, with exceptions of OsSDG738, SiSDG17 and SiSDG34, and PRMTs were all in group A (Fig. S3-5). AtPRMTs_OsPRMTs_ZmPRMTs and AtSDGs_OsSDGs_ZmSDGs were severally clustered together in group A and B-G (Fig. S3-6). In class A, B and D, model and gramineae JMJs closely got together, and HDMAs were divided into subclass c1 and c2 in group C (Fig. S3-7). The evolutionary relationship of HATs was investigated in Figure S3-8, HAGs were in group B, C and E, and HAFs, HAMs and HACs were separately divided into group A, D and F. HDTs and SRTs were severally clustered into group A and B, while HDAs were found in group C and D (Fig. S3-9).
Synteny analysis among HMs genes
In order to identify expansion patterns of HMs genes, duplicated blocks within each gramineae genome were investigated (Fig. S4). Totally, 144 pairs of TaHMs were identified from 21 chromosomes (Fig. 2a and Fig. S4-1). There were respectively 72 TaSDGs, 2 TaPRMTs, 8 TaHDMAs, 30 TaJMJ, 1 TaHAG, 2 TaHAM, 3 TaHAC, 3 TaHAF, 20 TaHDA and 3 TaSRT gene pairs (Fig. 2b). Only 4 SbHMs gene pairs, including SbSDG16-SbSDG37, SbSDG22-SbSDG26, SbJMJ1-SbJMJ10 and SbHDA11-SbHDA5, were identified in S. bicolor genome (Fig. 2a, Fig. 2b and Fig. S4-2). A total of four types of SvHMs genes (4 SvSDGs, 2 SvJMJs, 1 SvHACs and 2 SvHDAs) pairs were found (Fig. 2a, Fig. 2b and Fig. S4-3). Four pairs of SiSDGs, two pairs of SiJMJs and one pair of SiHACs were characterized (Fig. 2a, Fig. 2b and Fig. S4-4). In 10 chromosomes, we found fourteen ZmHMs gene (6 ZmSDGs, 2 ZmJMJs, 1 ZmHAGs, 1 ZmHAMs, 2 ZmHACs and 2 ZmHDAs) pairs (Fig. 2a, Fig. 2b and Fig. S4-5). However, no HvHMs gene pairs were identified (Fig. 2a and Fig. 2b).
We investigated syntenic relationship of gramineae and Arabidopsis HMs (Fig. 2c and Fig. S5). One HMTs (AtSDG24 and TaSDG97), four HDMs (AtJMJ13 and TaJMJ3, AtJMJ13 and TaJMJ7, AtJMJ13 and TaJMJ11, AtJMJ13 and TaJMJ42) and one HDACs (AtHDA9 and TaHDA12) gene pairs were respectively identified between Arabidopsis and wheat (Fig. 2c and Fig. S5-1). Only AtSDG24 and SbSDG19 was found in the same gene pair (Fig. 2c and Fig. S5-2). Three HMs gene pairs were respectively characterized in Arabidopsis-S. viridis (AtSDG24 and SvSDG2, AtJMJ13 and SvJMJ2, AtHDA5 and SvHDA4) and Arabidopsis-S. italic (AtSDG31 and SiSDG4, AtJMJ13 and SiJMJ12, AtHDA10 and SiHDA10) (Fig. 2c, Fig. S5-3 and Fig. S5-4). While no HMs gene pairs were identified in Arabidopsis-barley and Arabidopsis-maize.
All kinds of HMs gene pairs were found between rice and wheat genome, including 62 pairs of HMTs (59 pairs of SDGs and 3 pairs of PRMTs), 25 pairs of HDMs (9 pairs of HDMAs and 16 pairs of JMJs), 8 pairs of HATs (1 pairs of HAGs, 3 pairs of HAMs, 3 pairs of HACs and 3 pairs of HAFs) and 16 pairs of HDACs (12 pairs of HDAs, 2 pairs of SRTs and 2 pairs of HDTs) (Fig. 2d, Fig. 2e and Fig. S6-1). A total of 27 pairs of OsHMs-HvHMs were identified, such as OsSDG713-HvSDG3, OsPRMT708-HvPRMT1, OsHDMA701-HvHDMA1, OsJMJ718-HvJMJ8, OsHAM701-HvHAM1, OsHAC703-HvHAC3, OsHAF701-HvHAF1, OsHDA712-HvHDA2, OsSRT702-HvSRT2 and OsHDT701-HvHDT1 (Fig. 2d, Fig. 2e and Fig. S6-2). Nine of eleven kinds of HMs gene pairs were found between S. bicolor and rice, including twenty-one pairs of SDGs, three pairs of HDMAs, ten pairs of JMJs, two pairs of HAGs, one pair of HACs, PRMTs as well as HAFs, 6 pairs of HDAs and 2 pairs of HDTs (Fig. 2d, Fig. 2e and Fig. S6-3). There were 41 pairs of HMs between S. viridis and rice, including most kinds of HMs (Fig. 2d, Fig. 2e and Fig. S6-4). Orthologous pairs were identified in all HMs between S. italic and rice, except for HAGs (Fig. 2d, Fig. 2e and Fig. S6-5). More than 50 pairs of maize and rice HMs were found through comparison of their genomes (Fig. 2d, Fig. 2e and Fig. S6-6).
In order to evaluate selection pressure during duplication of above gene pairs, their Ka, Ks and Ka/Ks values were calculated. The data showed that Ka/Ks values were all less than or generally equal than 1 (Table S2-4). However, several gene pairs, such as SiJMJ5-SiJMJ19, AtJMJ13-TaJMJ3, and AtJMJ13-TaJMJ7, shared no nonsynonymous mutation according to their Ks values.
Promoter and structure analyses of HMs genes
HMs genes played important roles in plant stress and defense responses [24, 25]. Therefore, we identified stress-related elements in gramineae HMs genes (Fig. S7). In TaHMTs, TaHDMs and TaHDACs genes, at least one abscisic acid-, MeJA-, defense-, drought-, low temperature- or salt-related elements were uncovered (Fig. S7-1, 2 and 4). Abscisic acid, MeJA, and defense responsiveness elements were also identified in TaHATs genes (Fig. S7-3). There were two to thirteen stress-related motifs (defense and stress, abscisic acid and MeJA-responsiveness elements) in HvHMTs; But no defense and stress responsiveness cis-elements were uncovered in all HvHDMs, except for HvJMJ10 and HvJMJ11 (Fig. S7-5). SbSDG3, SbSDG13, SbPRMT1 and SbJMJ16 only contained one defense and stress, abscisic acid or MeJA-responsiveness motif, and other SbHMs included more than two stress-related elements (Fig. S7-6). Like SbHMs, numbers of stress, abscisic acid or MeJA-related elements were identified in S. viridis and S. italic HMs (Fig. S7-7 and S7-8). In ZmHMs, stress-related (MeJA-responsive, drough-inducibility, abscisic acid responsive, salicylic acid responsive, low-temperature responsive, defense and stress responsive, and wound-responsive), gibberellin-responsive and seed-specific regulation elements were identified (Fig. S7-9).
We identified HMs gene structures. In general, homologous HMs genes, especially genes in same pairs, shared similar structures, but genes lengths in different groups were various (Fig. S8). Regarding TaHMTs, most of homologous genes were made up of more than one CDS, and most of TaHMTs gene lengths were more than 3000 bp (Fig. S8-1). TaJMJs genes can be broadly divided into two classes, all TaHDMAs were in the same group; there were one to nine CDSs in TaHDMAs, and TaJMJs generally shared more CDSs than TaHDMAs; And all TaHDMs were about 15000 bp length, with the exceptions of TaJMJ31 and TaJMJ34 (Fig. S8-2). In terms of TaHATs, genes in each cluster contained relatively consistent structures, and TaHAG2 were longer about 7000 bp than other genes for its long non-coding sequence (Fig. S8-3). All TaHATs, except for TaHDA18 and TaHDA23, were segmented by non-coding region, and they were shorter than other TaHMs (Fig. S8-4). All HvHMTs, except for HvSDG4, shared short non-coding sequence, and most of them were about 2000 to 5000 bp (Fig. S8-5). Four HvHDMs (HvJMJ4, HvHDMA1, HvHDMA2 and HvHDMA4) only contained one CDS, while other ones were separated, and most of them (except for HvJMJ6) were about 9000 bp length (Fig. S8-6). Short CDSs were identified in HvHAT and HvHDAC genes, most of HvHAT and HvHDAC genes were about 1000 bp, while HvHAG1, HvHAF1, HvHDA3, HvHDA8 and HvHDT1 were obviously long than other genes (Fig. S8-7 and Fig. S8-8). Lots of SbHMTs (SbSDGs and SbPRMTs), SbHDMs (SbHDMAs and SbJMJs), SbHATs (SbHAGs, SbHAMs, SbHACs and SbHAFs) and SbHDACs (SbHDAs, SbSRTs and SbHDTs) were made up of short CDSs, while several genes contained one to two long CDSs (Fig. S8-9, 8–10, 8–11 and 8–12). About a third of SbSDGs included one long CDS; And all SbHMTs were less than 20000 bp, except for SbSDG7 (Fig. S8-9). SbHDMs were about from 2900 to 23000 bp long, and their CDSs and UTRs were generally short (Fig. S8-10). SbHATs and SbHDACs were respectively shorter than SbHMTs and SbHDMs, they were separated by short non-coding sequences, while several SbHDACs (SbHDA7-8 and SbSRT1) contained long non-coding sequences (Fig. S8-11 and 8–12). Numbers of short CDSs were found in many SvHMTs and SvHDMs; And several genes, especially SvSDG7-8, SvSDG16, SvSDG18, SvSDG30, SvHDMA3-4, SvJMJ2, SvHAC4 and SvHDA3, shared long CDS and UTR (Fig. S8-13 and 8–14). SvHATs and SvHDACs were generally short in gene and CDS length, while UTRs in several genes were long (Fig. <link rid="fig8">8</link>–15 and 8–16). Most of SiHMs contained numbers of CDSs, while only one to two long CDSs were in SiSDG18, SiSDG31, SiSDG41, SiSDG5, SiSDG11, SiSDG35, SiSDG27, SiHDMA1 and SiHDMA3 (Fig. S8-17, 8–18, 8–19 and 8–20). ZmHMs (except for ZmSDG1, ZmSDG3, ZmSDG5, ZmSDG16, ZmSDG18, ZmSDG28-29, ZmSDG36, ZmSDG41, ZmHDMA2 and ZmHDA3) were made up of either lots of short CDSs or one to four long ones, and one long 5’ or 3’ UTR were in ZmSDG31, ZmJMJ9, ZmHDMA1, ZmHAC1, ZmHAC3-4, ZmHDA3 and ZmHDA5 (Fig. S8-21, 8–22, 8–23 and 8–24).
The expression patterns of TaHMs in developing wheat grain, response to brassinosteroid and activated charcoal
In order to investigate potential roles of HMs in wheat grain growth and development, their expression profiles were generated from endosperm, inner pericarp and outer pericarp (Fig. 3). In cluster 1, TaSDG53, TaSDG29, TaSDG56 and TaSDG61 shared high expression levels in all tissues, especially in inner pericarp; Several TaSDGs in cluster 2, such as TaSDG15, TaSDG103 and TaSDG21¸ also highly expressed in inner pericarp; In cluster 3, seven TaSDGs showed relatively low level in outer pericarp; TaSDGs in cluster 4 and 5 shared lower expressions than genes in other clusters; In cluster 6, most of genes were highly expressed in inner or outer pericarp (Fig. 3a). The expression level of TaHDMAs and TaJMJs were generally low than other type genes (Fig. 3b). Genes in cluster 1 mainly expressed in endosperm or outer pericarp, while expression of several genes, particularly TaJMJ5, TaJMJ9, TaJMJ21 and TaJMJ39, were very low; Levels of TaHDMs were higher in cluster 2 and 3 than 1, transcription levels of about half of TaHDMs (especially TaHDMA3, TaJMJ47, TaJMJ27, TaJMJ35 and TaJMJ41) were high in inner or outer pericarp, and TaJMJ12 and TaJMJ47 were mainly detected in endosperm (Fig. 3b). TaHATs can be classified into two classes according to their expression patterns; And genes in cluster 1 were more highly expressed in all tissues than that in cluster 2 (Fig. 3c). In terms of TaHDACs, TaHDA22, TaHDA31, TaHDA28, TaHDA10 and TaHDA20 in cluster 1 shared lower expression levels in all grain tissue layers than other TaHDACs genes in cluster 2–4; in cluster 2, TaHDA4 and TaHDA6 were mainly present in endosperm, TaHDA16-17 showed high level in outer pericarp, and TaHDA29, TaHDA19 and TaSRT1 were mainly detected in inner pericarp; In cluster 3 and 4, TaHDAs, TaSRT3 and TaSRT5 were highly expressed in pericarps (Fig. 3d).
In Arabidopsis, rice, wheat and maize, BR has been proved to play important roles in root growths, including lateral root (LR) initiation and hair formation [26–29]. BR treatment significantly increases wheat lateral root number, but inhibits root length and root diameter, and BR synthesis inhibitor BRZ showed opposite roles on lateral root number and root diameter [26]. HMs genes have been known to regulate various developmental processes, while their information in regulating wheat root is absent. In this study, we analyzed their expression profiles during BR- and BRZ-mediated root growth (Fig. 4). In cluster 2, TaSDG4, TaSDG23, TaSDG55 and TaSDG112 were obviously induced more than 2-fold by BRZ, but BRZ repressed TaSDG26, TaSDG68, TaSDG89, TaSDG92, TaSDG95, TaSDG103 and TaJMJ5 in cluster 1 (Fig. 4a). BR treatments increased more than ten TaHMs (especially TaJMJ5 and TaSDG28) expressions, but several TaHMs were repressed by BR1 or BR2 (Fig. 4b and c). For example, TaSDG26, TaSDG28 and TaJMJ5 were induced by both BR1 and BR2 treatments; TaSDG92 and TaSDG101 were specifically up-regulated by BR2; And transcription levels of TaJMJ21, TaSDG53 and TaHDA18 were repressed in both BR1 and BR2 groups.
In plant culture, AC is widely used to promote seedling growth, and its effect on growth of wheat seedlings has been identified [30]. AC treatment distinctly promotes wheat seedlings growth rate, accompanied by increasing soluble protein, root activity, total phenol and sugar content [30]. We found that 26 and 31 TaHMs genes were differentially expressed in root and leaf after AC treatment, respectively (Fig. 5). In root and leaf, about half of these TaHMs were down-regulated or up-regulated by AC. For example, TaSDG68 and TaSDG84 were respectively decreased about 4- and 8-folds in root and leaf after AC treatment, TaSDG55 in root was evidently increased in AC treatment group, and TaJMJ21 was up-regulated by AC in leaf (Fig. 5a and 5b).
The responses of TaHMs to abiotic and biotic stresses
In order to explore whether TaHMs respond to abiotic stresses, their expression analyses were analyzed after HS, DS and HD treatments using published RNA-seq [31]. We found 86 TaHMTs (83 TaSDGs and 3 TaPRMTs) genes were differentially expressed at 1 or 6 h after heat, drought or heat plus drought treatments (Fig. 6a). These TaHMTs genes were able to be divided into 6 clusters; In cluster 1, almost all TaSDGs were respectively induced and repressed by DS at 1 and 6 h, and up-regulated by HD at 6 h; About 20 TaSDGs in cluster 2 were obviously increased at 6 h after HS and HD treatments, while decreased at 1 h, and several TaSDGs were clearly induced or inhibited by DS; In cluster 3, TaSDGs were generally induced by both HS and HD at 6 h, while they were suppressed at 1 h in HS and HD groups, and DS increased these genes at 1 or 6 h; Eleven TaSDGs in cluster 4 were mainly increased by DS, but decreased by HD at 1 h; In cluster 5, TaSDG111 was respectively repressed at 6 h after HS treatment and induced by HD at 1 h, and other genes were obviously up-regulated at 6 h by DS; In cluster 6, TaSDG28 and TaSDG53 were mainly induced by HS at 1 h, while they were inhibited at 1 h after HD treatment, and DS upregulated other genes at 1 h. All TaHDMs were divided into four clusters (Fig. 6b). Expression of TaJMJ21 was high after HS and HD treatment, and it was increased by DS at 1 h; In cluster 2, TaJMJ7, TaJMJ11 and TaJMJ3 were generally up-regulated by DS and HD at 1 h, and other genes were generally up-regulated at 6 h after HS and HD treatments; TaHMTs in cluster 3 were clearly induced by DS, and several genes were induced by HS or HD at 6 h; In cluster 4, TaJMJ40 was respectively increased by HS at 6 h and HD at 1 h, and TaJMJ9 was increased at 6 h after HS treatment. TaHATs were clustered into 2 classes (Fig. 6c). TaHAG1, TaHAG2 and TaHAG5 were increased after HS and HD treatment in cluster 1; In cluster 2, TaHAM2 and TaHAM3 were obviously up-regulated at 6 h in HS and HD groups, and other genes were induced by HS, DS or HD at least one time point. As show in Fig. 6d, TaHDA4, TaHDA17 and TaSRT2 were induced by DS in cluster 1, and TaHDA4 was also increased in HS group; DS treatment increased 10 TaHDACs expression in cluster 2, and these genes were also affected by HS or HD at several time points; Nine TaHDACs were distinctly induced by HS, DS or HD in cluster 3; And TaHDACs (TaHDA10, TaHDA12, TaHDA16, TaHDA19 and TaHDA21) in cluster 4 were primarily expressed at 6 h in HS and HD groups.
To investigate responses of TaHMs to SS, their expression profiles were identified in salt sensitive wheat cultivar CS and insensitive cultivar QM after SS treatment. After SS treatment, almost all TaHMs were up-regulated at least one time point (Fig. 7). TaHMTs can be clustered into three clusters (Fig. 7a). In cluster 1, TaSDG17 and TaSDG21 were significantly induced at 12 h after SS treatment in QM. TaSDGs and TaPRMTs in cluster 2 were up-regulated at several points in either CS or QM group after SS treatment. For example, TaSDG76 was obviously induced at 6 and 12 h after SS treatment, and at 12 and 24 h, TaSDG2 and TaSDG11 were respectively increased by SS treatment in CS and QM. At most of time points, TaSDGs genes were visibly induced by SS in cluster 3. For example, expression level of TaSDG92 was increased at 6 and 12 h after SS treatment in CS, and it was induced by SS from 12 to 48 h in QM; SS up-regulated TaSDG28 at all time points in both of cultivars (Fig. 7a). In terms of TaHDMs, most of genes were up-regulated by SS from 6 to 24 h in cluster 1; Genes in cluster 2 were induced by SS treatment at most of time points; Transcript of TaJMJ18, TaJMJ27, TaJMJ23, TaJMJ25, TaJMJ38, TaJMJ44 and TaJMJ48 were increased from 12 to 48 h after SS treatment in cluster 3; In both CS and QM, there were six TaJMJs to be obviously induced by SS in cluster 4 (Fig. 7b). TaHATs were divided into two clusters (Fig. 7c). For example, in cluster 1, TaHAC8 and TaHAC10 were mainly regulated by SS in CS, and TaHAF4, TaHAG3 and TaHAC6 were induced by SS in both CS and QM; In cluster 2, SS treatment generally increased expression levels of TaHACs and TaHAGs, especially TaHAC1 and TaHAC2, at all time points (Fig. 7c). Expressions of TaHDACs, especially genes in cluster 3, were also clearly increased at least one time point after SS treatment (Fig. 7d).
Sitobion avenae (S. avenae) and Schizaphis graminum (S. graminum) are two main wheat insect pests which will lead to obvious yield losses [32]. Compared with non-phytotoxic aphid S. avenae, phytotoxic aphid S. graminum feeding causes more severe damage in wheat leaves [33]. Nitrogen (N) is an essential macronutrient for plant growth and development, and low N stress obviously repressed wheat leaf and root growth [33]. Cadmium (Cd) highly inhibits leaf photosynthesis, carbon and nitrogen metabolism, and wheat growth and yield [34]. In order to provides information for TaHMs in respond to biotic, nutrition and heavy metal stresses, expression patterns of TaHMs were identified in above transcriptome researches [32–34]. In wheat leaves, TaJMJ7 was increased about 3.4- and 4-fold after S. avenae and S. graminum feeding, respectively; Another TaJMJ gene which was induced by S. avenae infection is TaJMJ11; Transcription levels of TaJMJ40 and TaJMJ42 were higher in S. graminum feeding group than control one; While TaHDA17, TaSDG73, TaHDA20, TaSDG81, TaHDA22 and TaSDG89 were distinctly controlled by S. graminum feeding (Table 1). N stress obviously suppressed TaSDG73 and TaHDA20 expression in leaf, while it up-regulated TaJMJ11 and TaJMJ3 in root (Table 1). In roots, Cd induced 12 TaHMs, such as TaSDG13, TaJMJ28 and TaHDT1, about 2.2- to 6.4-folds, but decreased TaSDG102 expression (Table 1).
The diverse responses of TaHMs to growth and stress signals
In order to investigate multiple functions of TaHMs in wheat growth and stresses adaption, a Venn diagram was constructed with above identified DEGs (Fig. 8). DEGs can be clustered into six sets, including BR or BRZ (BR_BRZ) class, AC class, heat or drought (Heat_Drought) class, salt class, S. avenae or S. graminum (Sa_Sg) class, and N or Cd (N_Cd) class (Fig. 8 and Table S5). Two TaHMs (TaSDG68 and TaJMJ5) were commonly in response to BR_BRZ and AC; Expression patterns of TaSDG95 and TaSDG103 were altered by both BR_BRZ and salt treatments; Two DEGs (TaHDA4 and TaHAG5) were identified after AC and Heat_Drought treatments; The expressions of three TaSDGs (TaSDG27, TaSDG35 and TaSDG84) and TaJMJ25 were commonly induced or repressed by AC or salt; TaSDG102 was differently expressed in both AC and N_Cd groups; A total of 72 TaHMs simultaneously respond to heat, drought or salt treatments; TaSDG81 and TaJMJ7 were simultaneously uncovered as DEGs in Heat_Drought and Sa_Sg classes, when comparing with control group; TaSDG101 and TaJMJ21 were differently expressed after BR_BRZ, AC and Heat_Drought treatments; After BR_BRZ, AC and salt treatments, two genes (TaSDG55 and TaJMJ17) were identified as DEGs; There were 21 DEGs, such as TaSDG1, TaJMJ13, TaHAC2, TaHDA6 and so on, to be found in AC, Heat_Drought and salt classes at the same time; TaSDG2 and TaSDG4 expressions were affected by BR or BRZ, AC, heat or drought and salt stress; TaJMJ40 and TaJMJ42 were detected to respond to AC, heat or drought, salt, and S. avenae or S. graminum, commonly; We found that TaSDG13 and TaJMJ28 were common DEGs after AC, Heat_Drought, Salt and N_Cd treatments; Only one gene TaJMJ11 was in response to Heat_Drought, salt, Sa_Sg, and N_Cd, simultaneously; TaJMJ34 was commonly induced or repressed by five signals, including BR_BRZ, AC, Heat_Drought, Salt.and N_Cd (Fig. 8 and Table S5).
Expression analysis of ZmHMs in developing seed and response to GA treatment
To investigate functions of ZmHMs in maize growth and development, expression profiles of ZmHMs were analyzed in different seed growth stages of B73 and SWL01 (Fig. 9a and b). SWL01 is a mutant of B73 and contain higher viscosity than B73 [35]. From 0 to 24 d after pollination (DAP), 80 ZmHM genes were clustered into 5 classes (Fig. 9a). During whole experiment period (especially at 2 DAP), ZmSDG36 in cluster 1 shared higher expression level than genes in other classes; In cluster 2, ZmHMs shared higher expression at early stages (from 0 to 8 DAP) than late periods (from 16 to 24 DAP); Elven ZmHMs highly expressed at all stages, especially from 0 to 4 DAP, in cluster 3; Transcription level of ZmHMs in cluster 4 generally decreased over time, with the exceptions of ZmJMJ12, ZmSDG11, ZmSDG3 and ZmHDA13; Compared with genes in other clusters, ten ZmHM genes (especially ZmSDG16) were weekly expressed from 0 to 24 DAP in cluster 5. There were a total of 81 ZmHM genes to be detected during SWL01 seed development (Fig. 9b). Like genes in B73, these ZmHMs were distributed into 5 clusters in SWL01; ZmHMs in cluster 1 and 2 (especially in cluster 2) mainly expressed at 0, 2 and 4 DAP; ZmHMs, such as ZmSDG29, ZmSDG36, ZmSDG40 and ZmHDA1, shared higher expression level in cluster 3 than genes in other clusters, while transcription levels of genes in cluster 4, especially ZmSDG5, ZmSDG14, ZmSDG16, ZmSDG19, ZmJMJ4 and ZmHDT1, were generally low; In cluster 5, most of ZmHMs gradually decreased over time, but ZmJMJ12, ZmSDG11, ZmSDG41 and ZmSDG6 showed opposite trend. A total of 79 ZmHMs were commonly expressed in both B73 and SWL01 seeds, but most of them shared different expression patterns in two cultivars. For example, expression level of ZmSDG41 was higher in B73 than SWL01; ZmHAF1 were gradually decreased over time in B73, while its expression was almost no change in SWL01 (Fig. 9c). ZmSDG23 was specifically expressed in B73, and ZmJMJ4-ZmSDG14 were only detected in SWL01 (Fig. 9c). GA3 application significantly promoted leaf sheaths growth of D11 [36]. Seven ZmHM genes were differently expressed between GA and control groups (Fig. 9d). In cluster 1, ZmHDMA3, ZmHDA10, ZmJMJ10 and ZmSDG10 were down-regulated by GA, but ZmHDA12, ZmHDA3 and ZmSDG33 were up-regulated.
Expression analysis of ZmHMs in response to drought stress
To identify potential roles of ZmHMs in drought adaption, their expression patterns were analyzed in drought tolerant cultivars (ND476 and H082183), drought sensitive cultivars (ZX978 and Lv28), and C7-2 (Table 2). Totally, ten ZmHMs were identified as DEGs in response to drought stress. After drought treatment, transcription level of ZmJMJ2 was increased about 6-fold in ND476 than ZX978, but ZmHDA11 was repressed in ND476. ZmSDG5, ZmJMJ4 and ZmSDG24 were induced by drought treatment in C7-2, but ZmSDG33 and ZmJMJ17 were controlled. In Lv28 and H082183, ZmJMJ5 was up-regulated by both moderate and severe drought treatments. Expression level of ZmSDG1 was increased in H082183 after moderate drought treatment, while ZmHDA2 was obviously down-regulated after severe drought treatment.