In Arabidopsis thaliana Cd differentially impacts on hormone genetic pathways in the methylation defective ddc mutant compared to wild type

Plant growth

Primary root length and rosette size were estimated. Control root length, measured until 21 DAG, was lightly minor in ddc vs WT (Fig. 1A). Cd differentially inhibited root growth in the two samples: at 21 DAG, 25 and 50 µM Cd-treated roots were 1.2 and 2.2 fold shorter than control roots in ddc, while in the WT Cd-treated roots were 1.8 and 2.8 fold shorter than control ones (Fig. 1A). Consequently, at 21 DAG root of Cd-treated samples was longer in ddc vs WT, particularly at the lowest Cd concentration.

Rosette size was estimated at 21 DAG, corresponding to the period necessary for its full development¹¹, by evaluating leaf number and area. Control plants of both ddc and WT exhibited a complete leaf series, although most leaves resulted smaller in ddc (Fig. 1B,C). Cd affected rosette development reducing leaf number and area, less in ddc than WT, resulting into a higher leaf area and/or number in ddc under both Cd concentrations (Fig. 1B,C).

Gene expression profile

RNA-Seq analysis provided an overview of gene expression profile of Cd-treated and control plants of both ddc and WT. The following comparisons were performed: ddc vs WT under control (Ctrl) conditions (ddc vs WT-Ctrl) and 25 and 50 µM Cd treatment (ddc vs WT-25 µM Cd; ddc vs WT-50 µM Cd); 25/50 µM Cd-treated vs Ctrl in ddc (25 µM Cd vs Ctrl-ddc; 50 µM Cd vs Ctrl-ddc); 25/50 µM Cd-treated vs Ctrl in the WT (25 µM Cd vs Ctrl-WT; 50 µM Cd vs Ctrl-WT).

After DEGs identification (see Supplementary Fig. S1 online) 14 of them were analysed through qRT-PCR to validate transcriptomic analysis (see Supplementary Fig. S2 online). Results were fully consistent with RNA-seq data. Gene Enrichment analysis was also performed, evidencing that Cd strongly impacted on transcriptome in both ddc and WT, but in a largely different way (see Supplementary Figs. S3–S9 online). Notwithstanding, a common aspect was that in both ddc and WT the genetic pathways (GPs) more impacted by Cd dealt with photosynthesis, stress responses and hormone biosynthesis and signalling.

Expression pattern of genetic pathways related to hormones

In view of hormones pivotal role in plant development and stress response and considering the assessed epigenetic control on their action and signalling¹², in this work we analysed in depth how the expression pattern (EP) of hormone-related GPs was modulated in ddc vs WT under Cd stress. The most relevant differences are discussed.

Auxins

Under control conditions, GPs related to auxin biosynthesis showed comparable EP in ddc and WT and no DEGs were detected (Fig. 2A). 25 µM Cd induced significant changes only in ddc resulting into: i) TAA1 and YUC5 downregulation along indole-3-pyruvic acid (IPA) pathway; ii) CYP71A13 and NIT2 overexpression along indole-3-acetaldoxime (IAOX) auxiliary pathway, while CYP79B3 was downexpressed (Fig. 2A). Differently, 50 µM Cd induced similar changes in ddc and WT consisting in: (i) a downexpression of YUC2 along IPA pathway in both samples and YUC5 and YUC9 in ddc and WT, respectively; (ii) overexpression of CYP71A12, CYP71A13, NIT2, NIT4 and downexpression of CYP71A16 along IAOX pathway in ddc and WT (Fig. 2A).

Auxin level and homeostasis also depend on its oxidative degradation, conjugation and methylation¹³. Under control conditions, GPs related to auxin conjugation and methylation showed comparable EPs in ddc and WT and no DEGs were detected (Fig. 2B,C), but were differentially impacted by Cd, mainly at 25 µM concentration. Namely, at 25 µM Cd several genes related to auxin conjugation (GH3.3, GH3.17, YDK1) and methylation (MES7, MES17) were downregulated in ddc, whereas in WT only MES18, involved in methyl-indole-3-acetate production, was downregulated (Fig. 2B,C). At 25 µM Cd most of the above genes were downregulated in ddc vs WT (Fig. 2B,C), while at 50 µM Cd only two genes, working in auxin methylation were differentially modulated in ddc and WT (Fig. 2B,C).

Under control conditions, GP related to auxin signalling exhibited similar EP in ddc and WT (Fig. 2D). Differences were induced by Cd. Namely, at 25 µM Cd, AUX/IAA family genes, which acts in signalling repression, were globally downregulated in ddc, while in WT the repressor IAA34 was overexpressed (Fig. 2D). Differently, 50 µM Cd effects on ddc and WT were quite similar, dealing with AUX/IAA family genes downregulation (Fig. 2D). In ddc vs WT comparisons, at 25 µM Cd above genes were downregulated, while no differences were found at 50 µM Cd (Fig. 2D). Moreover, and somehow unexpectedly, following 50 µM Cd it was observed in both ddc and WT a downregulation of several SAURs members, belonging to a large family of auxin responsive genes, which in turn can also have an impact on auxin pathway¹⁴ (Fig. 2D). Interestingly, such effect was more pronounced in ddc than WT. However, it must be mentioned that, although most of them are induced by auxin, several other hormones and co-factors acts upstream SAUR genes, regulating their activity in response to both endogenous stimuli and environmental cues¹⁴.

In summary, in both ddc and WT, Cd induced: i) a downregulation of IPA pathway, which is the main auxin biosynthetic pathway¹⁵ and a simultaneous upregulation of IAOX auxiliary biosynthetic pathway; ii) an enhancement of hormone signalling. However, in the WT such effects occurred only at 50 μM Cd. Moreover, in ddc Cd also induced a downregulation of GPs related to auxin conjugation.

Cytokinins

In all comparisons, GPs related to CKs biosynthesis showed similar EPs, unless for the downregulation in 50 µM Cd-treated WT vs Ctrl of IPT5, encoding rate-limiting enzyme along the pathway¹⁶ (Fig. 3A).

Major differences were observed for GPs related to CKs catabolism and conjugation, occurring through cleavage by oxidation and glycosylation, respectively¹⁶. Under control conditions, these GPs also exhibited similar EPs in ddc and WT (Fig. 3B-D). At both 25 and 50 µM Cd a downregulation of CKX5 and CKX6, encoding cytokinin-oxidases, occurred only in ddc (Fig. 3B). Differently, Cd impact on GPs related to CKs N-glycosylation was almost comparable in ddc and WT, resulting into the overexpression of two different genes working in N⁷– and N⁹–glycosylation pathways at 25 µM Cd, and one gene at the higher concentration (Fig. 3C).

Note that at 25 µM Cd the above genes were both overexpressed in ddc vs WT, while no differences were found at 50 µM Cd (Fig. 3C). Cd impact on GP related to cytokinin O-glycosylation was major, especially in ddc, involving at 25 µM Cd the overexpression of seven genes along this pathway compared to Ctrl (Fig. 3D), and one gene in WT 25 µM Cd vs Ctrl (Fig. 3D). By contrast, at 50 μM Cd the expression pattern along this pathway was similar in ddc and WT, being characterised by the upregulation of the same seven genes above mentioned and the downregulation of AT5G38010 (Fig. 3D). Finally, at 25 μM Cd, five genes along these pathways resulted upregulated in ddc vs WT while a similar EP occurred in ddc and WT under 50 μM Cd treatment (Fig. 3D).

Concerning GP involved in CKs signalling, under control conditions A-ARRs, encoding negative regulators of CKs signalling¹⁷, were downregulated and signalling was likely enhanced in ddc vs WT (Fig. 3E). 25 µM Cd induced a downregulation of ARR11 A-type ARRs and B-ARR family ARR10 transcription factors, which control primary plant response to CKs, only in ddc (Fig. 3E). Whereas, at 50 µM Cd both ddc and WT showed A-ARR downregulation, supposedly leading to pathway upregulation (Fig. 3E). At 25 µM Cd AHP1, encoding positive regulators of CKs signalling¹⁸, was downregulated in ddc vs WT, while ARR17 was overexpressed, suggesting that signalling was downregulated also in ddc vs WT (Fig. 3E). No differences occurred between ddc and WT at 50 µM Cd (Fig. 3E).

In summary, transcriptomic analysis evidenced that GP related to the biosynthesis of trans-zeatin, the most relevant CK, was negatively affected by Cd only in the WT at 50 μM Cd. In response to Cd, GPs related to CKs inactivation were enhanced in both ddc and WT, but in ddc a downregulation of GP related to CKs cleavage also occurred. Finally, hormone signalling was differentially modulated by Cd in relation to both the sample (ddc vs WT) and heavy metal concentration, resulting into a downregulation at 25 µM Cd only in ddc and an enhancement in ddc and WT at 50 µM Cd.

Gibberellins

Under control conditions, GPs related GAs biosynthesis showed similar EPs in ddc vs WT (Fig. 4A). 25 μM Cd induced in ddc: i) a downregulation of GA2 encoding the ent-kaurene synthase, a pivotal enzyme along the early GAs biosynthetic pathways to synthetize GA₁₂; ii) a downregulation of GA4, a key gene of GAs biosynthesis, along which bioactive GAs are synthetized¹⁹ (Fig. 4A). No Cd-induced modulation was observed in the WT (Fig. 4A). On the contrary, at 50 μM Cd both ddc and WT showed a downregulation of GA5 (Fig. 4A). Finally, in ddc vs WT the only difference dealt with GA2 downregulation at 25 µM Cd (Fig. 4A).

GPs controlling GAs inactivation also showed a comparable transcriptional pattern in ddc and WT under control conditions, and no DEGs were detected (Fig. 4B). 25 μM Cd induced a downregulation of DAO2 and AOP1, encoding GA2ox enzymes, only in ddc (Fig. 4B). Accordingly, in ddc vs WT these genes were downregulated only at the lowest Cd concentration (Fig. 4B).

Under control conditions, also GP related to GAs signalling was not differentially modulated in ddc vs WT (Fig. 4C). A downregulation of genes encoding DELLA proteins, which act as repressors²⁰, was induced only in ddc by 25 µM Cd (Fig. 4C) and in both ddc and WT at 50 µM Cd (Fig. 4C), suggesting an enhancement of hormone signalling. Consequently, DELLA-codifying genes resulted downregulated also in ddc vs WT only at the lowest Cd concentration (Fig. 4C).

In summary, in ddc the lowest Cd treatment negatively affected GPs related to GAs biosynthesis but, at the same time, hormone signalling resulted enhanced. In the WT similar effects were observed only at 50 µM Cd.

Jasmonic acid

Under control conditions, six genes along the GP related to JA biosynthesis were downregulated in ddc vs WT (Fig. 5A). 25 µM Cd induced in ddc a downregulation of this GP, except for LOX4 upregulation (Fig. 5A) and a downregulation involving eight genes in WT (Fig. 5A). At 50 μM, Cd effects were limited to LOX5 downregulation and LOX4 and OPR1 upregulation in ddc and LOX5 and AOS downregulation in WT (Fig. 5A). No Cd–induced differences were found in ddc vs WT (Fig. 5A).

Concerning the JA signalling-related GP, in control conditions JAZ5 gene, encoding a protein acting as repressor²¹, was downregulated in ddc vs WT, highlighting a signalling enhancement (Fig. 5B). Interestingly, JAZ10 and JAZ9 were differentially impacted by 25 µM Cd in ddc and WT resulting upregulated and downregulated, respectively. (Fig. 5B). At 50 µM Cd, JAZs were overexpressed in both ddc and WT (Fig. 5B). No differences were detected in ddc vs WT exposed to Cd (Fig. 5B).

Globally, Cd negatively impacted on the GP related to JA biosynthesis especially in WT. Under Cd treatment hormone signalling was downregulated more in ddc than in WT, whatever concentration was applied.

Abscisic acid

Under control conditions, ABA biosynthesis-related GP showed comparable EP in ddc and WT, and no DEGs were detected (Fig. 5C). The only significant Cd effect dealt with NCED3 downregulation both in ddc and WT, regardless of applied concentration (Fig. 5C). Under control conditions, also the GPs related to ABA catabolism showed a comparable EP in ddc and WT and no DEGs were detected (Fig. 5D), but Cd differentially impacted on CYP genes, involved in phaseic acid degradative production²². Namely, at 25 μM Cd, CYP707A3 was downregulated only in ddc (Fig. 5D). Moreover, also CYP707A2 appeared downregulated in ddc vs WT (Fig. 5D). At 50 µM Cd it was observed an upregulation of both CYP707A2 and CYP707A4 in ddc, and of only CYP707A4 in WT (Fig. 5D).

Under control conditions, GP related to ABA inactivation through glucose conjugation showed similar EP in ddc vs WT (Fig. 5E). 25 μM Cd determined AT4G15260 upregulation and UGT71C3 downregulation only in ddc (Fig. 5E). At 50 μM, Cd equally impacted on ddc and WT, resulting into UGT71C1 and UGT2 downregulation, AT5G49690, UGT71B5, UGT71B6 upregulation and, limited to WT, AT4G15260 upregulation (Fig. 5E). No differences were highlighted in ddc vs WT exposed to Cd (Fig. 5E).

Under control conditions, the GP related to hormone signalling also presented a comparable EP in ddc and WT (Fig. 5F). In ddc, 25 µM Cd impact on this GP appeared rather complex, resulting in an upregulation of PYL3, encoding ABA receptor, and a downregulation of PP2Cs (PP2CA and HAI1) encoding negative regulators of ABA signalling²³. Moreover, ABI5, codifying a key transcription factor in ABA signalling²⁴ belonging to AREBs/ABFs family, was upregulated. However, SnRK2.7 gene, codifying a protein which activate the AREBs/ABFs transcription factors²⁴, was downregulated. Based on the prominent role of SnRK2s in plant response to ABA, it is likely that at 25 µM Cd ABA signalling was downregulated in ddc (Fig. 5F). Instead, 25 µM Cd determined in WT the upregulation of PYL6 and the downregulation of PP2Cs, suggesting an enhancement of ABA signalling (Fig. 5F). At 50 µM Cd, PYL6 was upregulated and SnRK2.7 downregulated in both ddc and WT, while HAI1 was downregulated only in ddc (Fig. 5F). When comparing ddc vs WT, at 25 µM Cd only SnRK2.7 was downregulated, while no differences occurred at 50 μM Cd (Fig. 5F).

In summary, Cd determined a slight downregulation of GP related to ABA biosynthesis in both ddc and WT regardless of its concentration. ABA catabolic pathway was lightly downregulated in ddc at 25 µM Cd but upregulated in both samples at 50 μM Cd. At the transcriptomic level, ABA signalling featured as enhanced in WT and downregulated in ddc regardless of Cd concentration.

Ethylene

Along GP related to ethylene biosynthesis, under control conditions ACS8 and ACS11 were upregulated in ddc vs WT (Fig. 6A). 25 µM Cd determined ACS8 and ACO5 downregulation in ddc and ACS4 upregulation in the WT (Fig. 6A). 50 μM Cd induced ACS7 upregulation and ACO5 downregulation in both ddc and WT and ACS2 and ACS11 overexpression only in WT (Fig. 6A). Finally, the only Cd-induced difference in ddc vs WT dealt with ACO1 downregulation at 25 µM Cd (Fig. 6A).

GP related to ethylene signalling showed similar EP in ddc and WT both under control conditions (Fig. 6B) and at 25 µM Cd, except for the upregulation of ETR2, encoding ethylene receptor²⁵ in WT (Fig. 6B). At 50 µM Cd, both ddc and WT exhibited ETR2 and ERF1 overexpression suggesting an upregulation of ethylene signalling (Fig. 6B); no differences occurred in ddc vs WT (Fig. 6B).

In summary, in control conditions GP related to ethylene biosynthesis was upregulated in ddc vs WT. Cd determined a downregulation and upregulation of this GP in ddc and WT, respectively. Concerning hormone signalling, at the highest Cd concentration in both ddc and WT an upregulation of this GP occurred.

Salicylic acid

Regarding SA, only the GP related to signalling resulted differentially expressed. Under control conditions, the GP related to SA signalling showed similar EP in ddc and WT. However, PR1, a useful molecular marker for the systemic acquired resistance (SAR) in response to pathogens²⁶, was downregulated in ddc vs WT (Fig. 6C). 25 μM Cd induced a downregulation of genes codifying TGA10 transcription factor only in the ddc (Fig. 6C). Whereas, 50 μM Cd induced a downregulation of PRB1 in both ddc and WT and of TGA8 only in ddc (Fig. 6C). In ddc vs WT, differences were found only at 25 µM Cd, with the downregulation of TGA10 and CAPE3 (Fig. 6C).

Altogether, these results evidenced a Cd-induced downregulation of this GP, likely resulting in an impairment of hormone signalling in both WT and ddc, but in the latter this effect already occurred at the lowest Cd concentration.

Phytohormone level

Based on the major effects induced by 25 µM Cd treatment, hormone quantification was carried out on plants exposed to this concentration, compared to untreated control plants.

Under control conditions, IAA amount was higher in ddc than WT, although not significantly. After Cd treatment, a decreasing trend was observed only in WT, resulting into a significant lower level as compared to ddc (Fig. 7A).

Concerning CKs, both biological active (tZ) and inactive conjugate (tZR, cZR, tZOG, cZOG, iPR) forms were analysed (Fig. 7B–G). Under control conditions, all analysed CKs were present in ddc, but CKs conjugate forms and above all O-glycosylated exhibited the highest levels (Fig. 7B–G). By contrast, in WT tZ was not detectable and all the other CKs forms exhibited a lower level compared to the mutant, which appeared significant for tZR and cZOG (Fig. 7B–G). Following Cd treatment, CKs levels increased in ddc, except for tZ decrease. In WT, the unique Cd effect dealt with tZR increase and tZOG decrease. Consequently, under Cd treatment the level of all CKs forms remained higher (from 0.25 to 3 times) in ddc than in WT (Fig. 7B–G).

Concerning GAs, precursors (GA₉, GA₁₉, GA₂₀), biologically active forms (GA₁, GA₃, GA₄, GA₇) and catabolites (GA₈, GA₃₄, GA₂₉, GA₅₁) were analysed (Fig. 7H–R). Under control conditions, GA₁₉ amount was significantly higher in ddc vs WT, while the amount of GA_20, the other in serie precursor of hydroxylated forms, was comparable between the samples (Fig. 7H,I). Following Cd treatment, GA₁₉ amount significantly increased in WT while a slight downtrend occurred in ddc, leading to comparable values in the two samples. The same trend was observed, but at less extent, for GA₂₀ (Fig. 7H,I). Consistently, under control conditions also the amount of the active hydroxylated forms GA₁ and GA₃ was higher in WT than in ddc (Fig. 7J,K). Following Cd treatment, a decrease of their amount was detected only in WT, globally leading to a higher level of these GAs in ddc mutant compared to the WT (Fig. 7J,K). In addition, in ddc mutant also the related catabolites GA₈ and GA₂₉ were globally lower than in WT, under both control conditions and Cd treatment (Fig. 7L,M).

Differences were observed also for GA₉, precursor of non-hydroxylated GAs: under Cd treatment its amount decreased in the WT and was instead induced in ddc mutant, resulting in a quite comparable value between the two samples (Fig. 7N). Consistently, the amount of active non-hydroxylated forms, GA₄ and GA₇, increased under Cd treatment only in ddc mutant; also in this case, at the end of heavy metal treatment, comparable values were detected in ddc and WT (Fig. 7O,P) In agreement with these results, following Cd treatment, the amount of catabolites GA₅₁ and GA₃₄ did not change in the WT, whereas in ddc it increased and decreased, respectively (Fig. 7Q,R).

As evident in Fig. 7S–V, differences were reported also for JA, ABA, SA and its predominant inactive conjugate, SA 2-O-β-D-glucoside (SAG). Under control conditions both JA and ABA amount was significantly lower in ddc vs WT and significantly decreased following Cd treatment only in the WT (Fig. 7S,T). Notwithstanding, under such condition the ABA amount remained lower in ddc than in WT while JA values were comparable in the two samples due its light, but not significant, increase in ddc (Fig. 7S,T). By contrast, under control conditions both SA and SAG amounts were significantly higher in ddc than in WT (Fig. 7U,V). Following Cd treatment, their amounts significantly decreased more in ddc than in WT, leading to an opposite condition (Fig. 7U,V).

Testing of the involvement of SUPPRESSOR OF DRM1 DRM2 CMT3 (SDC) gene in ddc response to Cd

Finally, we planned to inquire on the possible involvement of SDC gene in the response of ddc triple mutant to Cd exposure. Indeed, it has been reported that in ddc mutant the misexpression of such gene, which encodes a F-Box protein, is ultimately responsible of the developmental phenotypes of ddc, such as curled leaves and reduced growth, as evidenced by its reversion in the drm1 drm2 cmt3 sdc quadruple mutant²⁷. Note that in the WT SDC is silenced, being methylated in all its sequence contexts because of the redundant action of DRM2 and CMT3 enzymes. By contrast, in ddc, where DRM2 and CMT3 expression is silenced, the loss of non-CG methylation in the promoter region of SDC F-box gene determines its overexpression²⁷.

According to the above mentioned data²⁷, we firstly verified that under control conditions SDC resulted silent in the WT and overexpressed in ddc also in our transcriptomic analysis (data not shown, complete raw transcriptomic data are available at NCBI SRA under the BioProject accessionPRJNA641242;https://www.ncbi.nlm.nih.gov/Traces/study/?acc=PRJNA641242). Moreover, our data also showed that at the transcriptomic level SDC is not modulated by Cd since its expression level did not significantly change in ddc nor in WT whatever heavy metal concentration was applied.

Thereafter, we tested the involvement of SDC in the growth response of ddc mutant under Cd exposure, by monitoring primary root length of Arabidopsis thaliana WT, ddc and sdc plants grown under the following conditions: (i) on growth medium without Cd as control (Ctrl) (ii) on a medium supplemented with 25/50 µM Cd; (iii) limited to the WT and sdc mutant plants, on a medium supplemented with 25/50 µM Cd plus 15 µM 5-Azacytidine (5-Aza), an inhibitor of DNA methylation applied in order to mimic the hypomethylated state of ddc mutant.

Under control conditions, all three samples showed a similar root length. However, at 21 DAG, root was lightly shorter in ddc vs WT, while sdc displayed an intermediate length (Fig. 8A). Under both Cd treatments, roots were averagely longer in ddc than in WT. Again, sdc roots exhibited an intermediate length, more similar to WT than ddc (Fig. 8 B,C). Interestingly, WT and sdc plants treated with Cd plus 5-Aza had longer roots than the plants treated only with Cd, and quite comparable to ddc roots exposed to Cd (Fig. 8 D,E).

Source: Ecology - nature.com