Research Article Hydrogen sulfide regulates abiotic stress tolerance and biotic stress resistance in Arabidopsis Running title: Involvement of H2S in abiotic and biotic stresses Haitao Shi1, Tiantian Ye1,2, Ning Han3, Hongwu Bian3, Xiaodong Liu4 and Zhulong Chan1* 1
Key Laboratory of Plant Germplasm Enhancement and Specialty Agriculture, Wuhan Botanical Garden, Chinese
Academy of Sciences, Wuhan 430074, China 2
University of Chinese Academy of Sciences, Beijing 100039, China
3
Institute of Genetics, State Key Laboratory of Plant Physiology and Biochemistry, College of Life Sciences,
Zhejiang University, Hangzhou 310058, China 4
College of Agronomy, Xinjiang Agricultural University, Urumqi 830052, China
* Correspondence: [email protected]
Edited By: Ildoo Hwang, POSTECH Biotech Center, Korea
This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version and the Version of Record. Please cite this article as doi: [10.1111/jipb.12302] This article is protected by copyright. All rights reserved. Received: August 26, 2014; Accepted: October 17, 2014
1
Abstract Hydrogen sulfide (H2S) is an important gaseous molecule in various plant developmental processes and plant stress responses. In this study, the transgenic plants with modulation expressions of two cysteine desulfhydrases and exogenous H2S donor (sodium hydrosulfide, NaHS) and H2S scavenger (hypotaurine, HT) pre-treated plants were used to dissect the involvement of H2S in plant stress responses. The cysteine desulfhydrases overexpressing plants and NaHS pre-treated plants exhibited higher endogenous H2S level and improved abiotic stress tolerance and biotic stress resistance, while cysteine desulfhydrases knockdown plants and HT pre-treated plants displayed lower endogenous H2S level and decreased stress resistance. Moreover, H2S up-regulated the transcripts of multiple abiotic and biotic stress-related genes, and inhibited reactive oxygen species (ROS) accumulation. Interestingly, MIR393-mediated auxin signaling including MIR393a/b and their target genes (TIR1, AFB1, AFB2, and AFB3) was transcriptionally regulated by H2S, and was related with H2S-induced antibacterial resistance. Moreover, H2S regulated 50 carbon metabolites including amino acids, organic acids, sugars, sugar alcohols and aromatic amines. Taken together, these results indicated that cysteine desulfhydrase and H2S conferred abiotic stress tolerance and biotic stress resistance, via affecting the stress-related gene expressions, ROS metabolism, metabolic homeostasis, and MIR393-targeted auxin receptors. Keywords: Hydrogen sulfide; cysteine desulfhydrase; abiotic stress; biotic stress; MIR393; auxin receptor.
2
INTRODUCTION Along with nitric oxide (NO) and carbon monoxide (CO), hydrogen sulfide (H2S) has emerged as the third important gaseous molecule in various plant developmental processes and stress responses (Rausch and Wachter 2005; Wang 2010; Lisjak et al. 2013; Calderwood and Kopriva 2014; Hancock and Whiteman 2014). H2S promotes root organogenesis (Zhang et al. 2009), seed germination (Zhang et al. 2010), lateral root formation (Fang et al. 2014), and enhances photosynthesis (Chen et al. 2011a). Moreover, H2S confers protective roles in responses to heat, drought, salt, osmotic, and freezing stresses (Wang et al. 2010; Zhang et al. 2010b, 2011; Jin et al. 2011, 2013; Shen et al. 2012, 2013; Xie et al. 2014). Understanding the complex effects of H2S in plants requires further detailed analyses at the physiological, biochemiscal and molecular levels. In recent years, most of the in vivo roles of H2S in plants were obtained from pharmacology experiment through exogenous application of H2S donor (sodium hydrosulfide, NaHS), H2S scavenger (hypotaurine, HT) and H2S inhibitors (potassium pyruvate, PP; and hydroxylamine, HA) (Li et al. 2013; Shi et al. 2013, 2014a). However, the treatments with H2S modulating chemicals may have limitation in exactly reflecting the in vivo roles of H2S in plants. So far besides exogenous application of H2S in pharmacological experiments, it would be better to use plants with altered in planta H2S production as candidates for such an approach. To date, several proteins have been reported to be directly responsible for H2S generation in Arabidopsis, including L-cysteine desulfhydrase (LCD), D-cysteine desulfhydrase 1 (DCD1), D-cysteine desulfhydrase 2 (DCD2), and DES1 (Bloem et al. 2004; Riemenschneider et al. 2005; Papenbrock et al. 2007; Álvarez et al. 2010, 2012a, 2012b; Shen et al. 2013). The knockout mutant of AtLCD (lcd) with decreased endogenous H2S level has been used to investigate the role of H2S in plant response to drought stress (Jin et al. 2013), and expression of AtLCD and AtDCD1 in Escherichia coli showed increased cadmium resistance (Shen et al. 2012). Álvarez et al. (2012a, 2012b) found that cysteine-generated sulfide in the cytosol affected autophagy and transcriptional profile of Arabidopsis, and knockout mutant of the DES1 shows high resistance to biotrophic and necrotrophic pathogens. However, mechanisms of H2S mediated plant development and abiotic stress response remain obscure (Lisjak et al. 2013). Additionally, the connection between H2S and plant biotic stress resistance is largely unknown (Bloem et al. 2004, 2007; Lisjak et al. 2013). In this study, the transgenic plants with modulated expressions of two cysteine desulfhydrases (AtLCD and AtDCD1) together with exogenous pre-treatments with H2S donor (NaHS) and scavenger (HT), were used to dissect the involvement of H2S in plant abiotic and biotic stress responses. Further physiological, molecular, and metabolic studies were performed to investigate the underlying mechanisms of H2S-mediated abiotic and biotic stress responses, and the involvement of MIR393-targeted auxin receptors were highlighted in H2S-mediated disease resistance.
3
RESULTS Effect of abiotic and biotic stresses on cysteine desulfhydrases and the endogenous H2S content in Arabidopsis Through real-time quantitative PCR analysis, we found that the transcript levels of LCD and DCD1 increased upon all stress treatments tested (cold, dehydration, salt, hydrogen peroxide (H2O2) and bacterial pathogen) except abscisic acid (ABA) which had no effect on DCD1 transcript amount (Figure 1A). All treatments for 6 hours increased transcript amounts of LCD and DCD1 than those of treatments for 3 hours except pathogen infection, which induced highest transcript levels of LCD and DCD1 at 3 hours (Figure 1A). Consistently, the enzyme activity of LCD significantly increased after all these stress treatments, and the enzyme activity of DCD was significantly induced after dehydration, salt, H2O2 and bacterial pathogen treatments (Figure 1B). In addition, the endogenous H2S content significantly increased with 1.2-3 fold changes after stress treatments (Figure 1C). Modulation of cysteine desulfhydrase expression affects endogenous H2S level To further investigate the in vivo role of LCD and DCD1, we isolated the T-DNA mutant of LCD (lcd mutant, SAIL_739_C08/CS835466), and constructed LCD and DCD1 overexpressing plants and DCD1 RNAi knockdown plants. Consistently, the expression of LCD and DCD1 and the activities of LCD and DCD showed higher levels in the overexpressing plants, but exhibited lower levels in the knockdown plants (Figure S1A-D). LCD and DCD1 overexpressing plants exhibited significantly higher concentration of H2S, while LCD and DCD1 knockdown plants displayed significantly lower concentration of H2S (Figure 2). Moreover, 100 μM NaHS and 100 μM HT treatments significantly changed endogenous H2S level as expected (Figures 2, S2). Additionally, the accumulation of cysteine was negatively regulated by LCD and DCD1 expressions, but not affected by NaHS and HT treatments (Figure S3A). However, modulation of H2S content had no significant effect on reduced glutathione (GSH) content (Figure S3B). Modulation of cysteine desulfhydrase expression and endogenous H2S level affect abiotic stress tolerance and reactive oxygen species (ROS) metabolism Under stress conditions, although the endogenous H2S levels were up-regulated in all lines, LCD and DCD1 overexpressing plants and NaHS-treated plants showed higher concentrations of H2S compared with WT plants, while LCD and DCD1 knockdown plants and HT-treated plants displayed lower levels of them than WT plants (Figure S4). When subjected to drought, salt, and freezing, LCD and DCD1 overexpressing plants and NaHS-treated plants showed better growth and higher survival rate than those of wild type (WT) plants, while LCD and DCD1 knockdown plants and HT-treated plants exhibited worse growth and lower survival rate in comparison to WT plants (Figure 3A-B). Moreover, the increased abiotic stress tolerance of LCD and DCD1 overexpressing plants were restored by HT treatment, and the descreased abiotic stress tolerance of LCD and DCD1 knockdown plants were restored by NaHS treatment (Figure S5), indicating that H2S might be largely related with the affection of cysteine desulfhydrase on abiotic stress tolerance. Additionally, the transcripts of several abiotic stress-related genes (including CBF1, CBF3, CBF4, DREB2A, 4
DREB2B, RAB18, RD22, RD29A, and RD29B) exhibited higher levels in the LCD and DCD1 overexpressing plants and NaHS-treated plants (Figure 4 and Table S1), while lower in the LCD and DCD1 knockdown plants and HT-treated plants (Figure 4 and Table S1). This result indicated the significant effect of H2S on transcript levels of stress-related genes. Under control condition, no significant differences of ROS level and several antioxidants were observed among wild type and various lines with different endogenous H2S levels (Figure 5A-I). When abiotic stresses were applied, LCD and DCD1 overexpressing plants and NaHS-treated plants displayed lower levels of H2O2, O2•-, and oxidized glutathione (GSSG), but higher activities of superoxide dismutase (SOD), catalase (CAT), peroxidase (POD), and glutathione reductase (GR) and higher levels of GSH and GSH redox state than those of WT plants (Figure 5A-I). On the contrary, LCD and DCD1 knockdown plants and HT-treated plants exhibited the opposite effects on the ROS accumulation (Figure 5A-I). Cysteine desulfhydrase expression and H2S regulate defense resistance against bacterial pathogen As expected, NaHS and HT treatments significantly increased and decreased endogenous H2S level, respectively (Figure S6). Through quantification of bacteria number in the pathogen-infected Arabidopsis leaves, we found that the LCD and DCD1 overexpressing plants showed significantly less bacterial than WT at 3 day post infection (dpi), while LCD and DCD1 knockdown plants exhibited more bacterial than WT (Figure 6A). Consistently, H2S donor (NaHS)-treated plants exhibited improved defense resistance, while H2S scavenger (HT)-treated plants showed decreased defense resistance (Figure 6A). Additionally, the transcript levels of ENHANCED DISEASE SUSCEPTIBILITY 1 (EDS1), PHYTOALEXIN DEFICIENT 4 (PAD4), PR1, PR2, PR3, PR4, and PR5 were higher in the LCD and DCD1 overexpressing plants and NaHS-treated plants, but were lower in the LCD and DCD1 knockdown plants and HT-treated plants (Figure 6B and Table S1). Therefore, these results indicated that cysteine desulfhydrase expression and H2S regulated defense resistance against Pseudomonas, partially via modulating SA signaling-related genes. The involvement of MIR393-mediated auxin signaling in H2S-mediated antibacterial resistance The transcript levels of MIR393a and MIR393b were up-regulated by H2S effect, while the transcript levels of TRANSPORT INHIBITOR RESPONSE 1 (TIR1), AUXIN SIGNALING F BOX PROTEIN 1 (AFB1), AFB2, and AFB3 were negatively regulated by H2S effect (Fig. 7A and Table S1). In accordance with the real-time assay, the GUS activities of pMIR393a:GUS and pMIR393b:GUS transgenic plants were significantly activated by exogenous NaHS treatment, but were significantly decreased by exogenous HT treatment (Figure 7B). Under control condition, 35S:MIR393a-6, 35S:MIR393b-1, and tir1afb1afb2afb3 plants exhibited enhanced antibacterial resistance with less bacterial numbers in plant leaves at 3 dpi, while 35S:TIR1-5 and 35S:mTIR1-9 transgenic plant leaves showed more bacterial numbers at 3 dpi in comparison to WT plants (Figure 7C). To further investigate the involvement of MIR393-targeted auxin receptors in H2S-mediated antibacterial resistance, we determined the antibacterial resistance of 35S:MIR393a, 35S:MIR393b, tir1afb1afb2afb3, 35S:TIR1, and 35S:mTIR1 plants in 5
comparison to WT after the treatments of NaHS and HT. Unlike the WT plants, the improved antibacterial resistance of 35S:MIR393a-6, 35S:MIR393b-1, and tir1afb1afb2afb3 plants and the decreased antibacterial resistance of 35S:TIR1-5 and 35S:mTIR1-9 transgenic plants were not significantly affected by either NaHS or HT treatment (Figure 7C). However, the endogenous H2S level in these transgenic plants exhibited no significant difference compared with that in WT plants under both control and NaHS/HT-treated conditions (Figure 7D). These results suggested that H2S induced resistance might be mainly mediated by MIR393-dependent auxin signaling. Modulation of metabolic homeostasis by H2S effect To gain more insight into the metabolic homeostasis which might be affected by H2S, multiple carbon metabolites were assayed through gas chromatography time-of-flight mass spectrometry (GC-TOF-MS) analysis. Totally, 54 metabolites including 16 amino acids, 13 organic acids, 18 sugars, 5 sugar alcohols and 2 aromatic amines were reproducibly examined in WT (without treatment, and with NaHS and HT treatments), LCD and DCD1 overexpressing plants and knockdown mutants (Figure 8 and Table S2). Generally, 50 of 54 metabolites (except aspartic acid, lysine, hexadecanoic acid, and melibose) were significantly affected by NaHS/HT treatment and in the plants with altered LCD or DCD activity. Interestingly, most of the metabolite concentrations were higher in NaHS-treated and LCD-OX/DCD1-OX plants and lower in HT-treated and LCD/DCD1 knockdown plants (Figure 8 and Table S2), indicating the modulation of H2S in several carbon metabolites including amino acids, organic acids, sugars, sugar alcohols and aromatic amines.
DISCUSSION Previous studies dissecting the in vivo roles of H2S in plants were mainly obtained from pharmacology experiment using H2S donor, scavenger and inhibitors (Li et al. 2013; Shi et al. 2013, 2014a). Therefore, it is difficult to distinguish the effects of H2S from its upstream (cysteine) and other downstream metabolites of cysteine (such as glutathione) (Wu et al. 2010; Cao et al. 2013, 2014; Calderwood and Kopriva 2014). In this study, modulation of cysteine desulfhydrase expression affected endogenous H2S and cysteine levels, but had no significant effect on the accumulation of GSH under control condition (Figures 2, S1, S2). However, exogenous treatments of H2S donor (NaHS) and scavenger (HT) specifically affected the endogenous H2S, and did not affect the accumulations of cysteine levels and GSH under control condition (Figures 2, S3). Since the endogenous sulfate metabolism is very complex, the upstream and downstream metabolites of H2S might form part of a feedback mechanism to reset its homeostasis. In this study, using the transgenic plants with modulated endogenous H2S level together with the exogenous treatments of H2S donor (NaHS) and scavenger (HT), we assigned the protective role of H2S in both abiotic stress tolerance and biotic stress resistance (Figures 3, 6). This result is in accordance with the enhanced abiotic stress tolerance by H2S donor (NaHS) in various plant species, and this is the first report to show the protective role of H2S in antibacterial resistance. H2S and abiotic stress tolerance in Arabidopsis 6
When subjected to abiotic stress conditions, plant ROS level (mainly including H2O2 and O2•-) was largely induced, along with induced unstabilization of plasma membrane and decreased cell turgor (Mittler et al. 2004, 2011; Miller et al. 2010). Consistently, plants have evolved complex ROS defence system, including several enzymatic antioxidant enzymes (SOD, CAT, POD, GR, etc) and non-enzymatic antioxidant molecules (such as glutathione) to regulate endogenous ROS level. In this study, increased endogenous H2S and exogenous application of NaHS significantly activated ROS detoxification system, including enzymatic antioxidant enzymes and non-enzymatic glutathione to make cellular ROS at the relatively low level under abiotic stress conditions, thus conferring improved abiotic stress tolerance (Figure 5). When the endogenous H2S level decreased, significantly increased level of ROS and decreased levels of antioxidants resulted in decreased abiotic stress tolerance (Figure 5). Thus, the modulation of endogenous and exogenous H2S in ROS detoxification system might be directly connected with the effect of H2S on abiotic stress tolerance. In Arabidopsis, CBF1, CBF2, CBF3 (also known as DREB1b, DREB1c, DREB1a, respectively) and CBF4 have
important
roles in
cold, salt
and drought stress
responses via binding to
the C-repeat
(CRT)/dehydration-responsive element (DRE) cis-acting element of several stress-responsive genes (Seki et al. 2001; Haake et al. 2002; Achard et al. 2008; Novillo et al. 2012). When constitutively overexpressing AtCBF1/2/3/4, the expression of several downstream genes such as COR (cold related), LTI (low-temperature induced), KIN (cold inducible), RD (responsive to dehydration) and ERD (early responsive to dehydration) were largely induced, and the transgenic plants conferred improved abiotic stress tolerance (Seki et al. 2001; Haake et al. 2002; Achard et al. 2008; Novillo et al. 2012). Moreover, ABA2 and NCED3 are responsible for ABA generation in Arabidopsis (Chan 2012). Abiotic stress responses in plants are initiated via both ABA-dependent and ABA-independent signal transduction pathways (Jin et al. 2011; Novillo et al. 2012). CBF4 is ABA-independent, and RAB18, RD22, RD29A, and RD29B are ABA dependent stress-responsive genes (Jin et al. 2011; Novillo et al. 2012). In this study, the modulation of H2S in the transcripts of multiple abiotic stress-related genes (CBF1, CBF3, CBF4, DREB2A, DREB2B, RAB18, RD22, RD29A, and RD29B) might be directly related with the abiotic stress tolerance. H2S and biotic stress resistance in Arabidopsis The transcripts of EDS1 and PAD4, which are responsible for SA biosynthesis in Arabidopsis, were constitutively activated in the plants with higher H2S levels, but were significantly reduced in the plants with lower H2S levels (Figure 6B). SA accumulation is necessary for the activation of SA downstream gene expression such as PRs (Wang et al. 2007; Kazan and Manners 2009; Fu and Dong 2013). Consistently, the plants with higher H2S levels showed increased expression of SA-dependent PR genes, resulting in improved immunity resistance, while the plants with lower H2S levels exhibited decreased immunity resistance (Figure 6). Álvarez et al. (2012a) found that knockout mutant of the DES1 showed improved resistance to biotrophic and necrotrophic pathogens, SA accumulation and PR1 induction. The controversy might be attributed to at least three aspects as described following. Firstly, the effect of LCD1 and DES1 on sulfate metabolism is different. Although they have the same 7
effects on cysteine content, they modulate endogenous GSH and H2S levels differently (Figures 2, S3; Álvarez et al. 2012a). Modulation of LCD1 expression affected H2S levels (Figures 2, S4) but had no significant effects on endogenous GSH level (Figure S3) and DES1 transcript (Figure S7), while DES1 knockout mutant displayed higher concentration of GSH, but no report about its effect on H2S level (Álvarez et al. 2012a). Secondly, 28-day-old plants were used for disease resistance assay in this study, whereas 6-7-week-old plants were used by Álvarez et al. (2012a), and the different growth stage might have differences in sulfate metabolism including cysteine, GSH, and H2S levels. Additionally, Álvarez et al. (2012a) found that DES1 was localized in cytoplasm. LCD is predicted to be localized in cytoplasm, but DCD1 is predicted to be localized in chloroplast, cytosol, and mitochondrion. The differential subcellular localization might also contribute to the contradictory results. Wang et al. (2007) showed that SA and auxin act individually or through antagonistic crosstalk, and the finely tuned balance between them is critical for plant-pathogen interaction. Shen et al. (2013) showed that the transcripts of multiple miRNAs including MIR167, MIR393, MIR396, and MIR398 were significantly up-regulated by H2S. Among these miRNAs, MIR393 has been shown to be involved in plant-pathogen interaction (Navarro et al. 2006; Robert-Seilaniantz et al. 2011). Overexpression of MIR393, a microRNA that targets auxin receptors (TIR1, AFB1, AFB2, and AFB3), renders plants more resistant to bacterial pathogens, while over-expression of AFB1 results in plants more susceptible to bacterial pathogens (Navarro et al. 2006; Chen et al. 2011b; Robert-Seilaniantz et al. 2011; Bian et al. 2012; Liu et al. 2013). Besides SA-related genes, the transcript levels of MIR393a/b and their target genes (TIR1, AFB1, AFB2, and AFB3) were up-regulated and down-regulated by H2S, respectively (Figure 7A, B). Since the transcript of MIR393 is induced by IAA (Chen et al. 2011b) and H2S (Shen et al. 2013 and this study), the connection between auxin and H2S is further investigated in this study. DR5:GUS is a widely used auxin-related marker line under the control of the auxin-responsive DR5 promoter (Ulmasov et al. 1997), and NaHS and HT treatments had no significant effects on the GUS activities of DR5:GUS line (Figure S8). These studies suggested that H2S affected only the transcripts of MIR393 and its targeted auxin receptors, but not auxin responses. As shown in Figure 7C, MIR393a/b and their target auxin receptors mediated antibacterial resistance was not affected by NaHS and HT treatments, suggesting that MIR393-repressing auxin signaling might be largely contributed to H2S-mediated antibacterial resistance. Thus, H2S conferred plant immunity resistance via modulating both the transcript levels of SA-related genes and MIR393-mediated auxin signaling. H2S and metabolic homeostasis in Arabidopsis Not only stress-related genes, but also several metabolites were significantly affected by H2S effect (Figure 8 and Table S2). Notably, higher endogenous H2S level largely activated the accumulation of compatible solutes such as proline, arabinose, sucrose, lactulose, allose, fructose, lactose, tagatofuranose, talose, mannose, galactinol, dulcitol (Figure 8 and Table S2). Proline and some carbohydrates are important compatible solutes to respond to abiotic stress for osmotic adaptation (Krasensky and Jonak 2012). Thus, higher levels of proline and the carbohydrates might provide beneficial effect in response to stress conditions. On the contrary, lower endogenous H2S level exhibited lower concentrations of these compatible solutes (Figure 8 and Table S2), which might be related with 8
the decreased stress resistance. Among these amino acids, the mutant and HT-treated plants displayed significantly higher concentration of alanine, but lower concentrations of other amino acids, indicating the complex role of H2S in regulating amino acid pools. Asparagine accumulation shows that nitrogen re-distribution and mobilization are important features of the salt stress response (Maaroufi-Dguimi et al. 2011). In addition, asparagine was an amino group donor for the synthesis of the photorespiratory intermediate glycine, and Nagy et al. (2013) found this was also a good indicator of drought stress in drought tolerant and sensitive wheat cultivars. In the meanwhile, the modulation of other amino acids and organic acids by H2S treatment indicated common changes of H2S in carbon metabolism, and might contribute to the affected stress resistance. Based on the above results, a model for H2S-mediated stress responses in Arabidopsis was proposed (Figure 9). In response to various abiotic and biotic stresses, the H2S synthetic pathway was largely activated and H2S level was significantly increased. The increased endogenous H2S and exogenous application of H2S donor, on one hand constitutively activated the expression of several abiotic stress-related genes and SA signaling genes, and accumulate of compatible solutes such as proline and soluble sugars; on the other hand, changes of antioxidant enzyme activities and glutathione redox state under stress conditions, conferred improved both abiotic stress tolerance and biotic stress resistance (Figure 9). When the endogenous H2S was down-regulated, the opposite results were observed. Moreover, MIR393-targeted auxin receptors were regulated by H2S, and were largely contributed to H2S-mediated antibacterial resistance (Figure 9). Taken together, this study assigned the protective role of H2S in Arabidopsis responses to abiotic and biotic stresses, via regulating multiple stress-related gene transcripts, ROS metabolism, metabolic homeostasis and MIR393-targeted auxin signaling pathways.
MATERIALS AND METHODS Plant materials and growth conditions Arabidopsis thaliana (ecotype Columbia-0) seeds were sterilized with 70% (v/v) ethyl alcohol, and 10% (w/v) NaClO and then washed with deionized water. The seeds were sown on Murashige and Skoog (MS) medium plate containing 1% (w/v) sucrose or in soil after stratification at 4 °C for 3 days in darkness. The plants were then kept in the growth chamber, which was controlled at 23 °C and at an irradiance of about 150 µmol quanta m-2 s-1, with 65% relative humidity under 16 hour light and 8 hour dark cycles. Nutrient solution of modified Hoagland solution (Tocquin et al. 2003) was watered twice in the bottom of the pots with plants every week. The lcd mutant (SAIL_739_C08/CS835466) was obtained from the Arabidopsis Biological Resource Center, and the transgenic lines of pMIR393a:GUS, pMIR393b:GUS, 35S:MIR393a-6, 35S:MIR393b-1, 35S:TIR1-5 and 35S:mTIR1-9 have been described in Chen et al. (2011b), and the tir1afb1afb2afb3 mutant through crossing tir1 (CS3798), afb1 (SALK_070172), afb2 (SALK_137151), and afb3 (SALK_068787) has been described in Liu et al. (2013). To determine the effects of different stresses on the expression and activities of AtLCD and AtDCD1, 28-day-old Arabidopsis Col-0 plants were treated by cold stress (4 °C), dehydration (put the plants in a 9
un-covering the plate in the growth chamber), or watered by 50 µM ABA, 200 mM NaCl or 20 mM H2O2 in the pots with plants for 3 and 6 hours, or the plant leaves were infected by Pst DC3000 as Shi et al. (2012) described at OD600 = 0.001 for 1, 3 and 6 hours. The plant leaves were then harvested at indicated timepoints for further analysis. Transgenic plasmid construction and plant transformation For the AtLCD and AtDCD1 overexpressing constructs, the coding regions of AtLCD and AtDCD1 were amplified and cloned into pBARN vector under the control of CaMV 35S promoter with basta resistance (LeClere and Bartel 2001). For the amiR-AtDCD1 knockdown transgenic construct, amiR-AtDCD1 fragment were amplified from the plasmid pRS300 by PCR using specific primers from WMD3 (http://wmd3.weigelworld.org/cgi-bin/webapp.cgi) (Schwab et al. 2006), and the amiR-AtDCD1 fragment was then cloned into pBARN vector. The primers for vector constructs were listed in Table S3. The above recombinant constructs were then transformed into Agrobacterium tumefaciens strain GV3101 and introduced into Columbia-0 plants using the floral dip method (Clough and Bent 1998). Homozygous transgenic plants were selected on MS medium using basta resistance and were confirmed by PCR analysis. RNA isolation and real-time quantitative PCR Total RNA was extracted from plant leaves using TRIzol reagent (Invitrogen, California, USA) and RQ1 RNase-free DNase (Promega, Wisconsin, USA) was added to avoid possible genomic DNA contamination. First-strand cDNA was then synthesized using reverse transcriptase (TOYOBO, Osaka city, Japan) from 2 µg of total RNA. Real-time quantitative PCR was carried out using CFX96TM Real Time System (BIO-RAD, California, USA) with iQTM SYBR® Green Super mix (BIO-RAD, California, USA) as Shi et al. (2012) described. Ubiquitin 10 (UBQ10) was chosen as the reference gene in the treatments according to geNorm software (Czechowski et al. 2005), and housekeeping genes of ACTIN2 and EIF4A were used as the control. The gene transcript levels were standardized with ubiquitin 10 (UBQ10) using comparative ΔΔCT method, and all experiments were repeated at least three times. The specific primers for real-time quantitative PCR were listed in Table S4. Determination of H2S concentration H2S concentration was assayed as Chen et al. (2011a) described that based on formation of methylene blue. Briefly, 1 g of plant leaves were ground in 10 mL of extraction buffer (50 mM phosphate buffer, pH 6.8, 0.2 M ascorbic acid, and 0.1 M EDTA). The homogenate was mixed with 1 ml of 1 M HCl to release H2S, and H2S was absorbed in a 1 mL of 1% (w/v) zinc acetate trap. After 30 min of reaction, 0.5 mL of 5 mM dimethyl-p-phenylenediamine dissolved in 5 mM H2SO4 was added to the trap, and then 0.5 ml of 50 mM ferric ammonium sulphate in 100 mM H2SO4 was added into the trap. The concentration of H2S in zinc acetate traps was examined at 667 nm of absorbance. Determination of abiotic and biotic stress resistances Plant abiotic stress tolerance was assayed as Shi et al. (2012) described, and survival rate was determined at 4 d after recovery from the abiotic stress treatment. For drought stress treatment, 14-day-old plants were withheld 10
water for 21 days and then re-watered for 4 days. For salt stress treatment, 14-day-old plants were watered with NaCl solution, and the NaCl concentration was increased stepwise by 50 mM every two days to 150 mM for another 21 days. For freezing stress treatment, 2-week-old seedlings were cold acclimated at 4 °C for 14 days, then the plants were placed at -8 °C for 8 hours, and then transferred to a growth chamber at 22 oC for another 4 days. For disease resistance assay, 28-d-old plant leaves were infected with the pathogen strain of Pst DC3000 at OD600 = 0.001 as described by Shi et al. (2012) described, and the bacterial growth from infected Arabidopsis leaves was monitored at 0 and 3 days post infection (dpi). Determination of ROS level and antioxidant activities 0.5 g of plant leave samples were harvest from ten independent plants in each experiment and used for the ROS assay. H2O2 and O2•- contents were quantified using the titanium sulfate method and the Plant O2•- ELISA Kit (10-40-488, Dingguo, Beijing, China), respectively, as Shi et al. (2013) described. Determination of enzyme activities LCD and DCD activities were determined by measuring the production rate of H2S from L-cysteine and D-cysteine as Jin et al. (2013) described. The activities of antioxidant enzymes [SOD (EC 1.15.1.1), CAT (EC 1.11.1.6), POD (EC 1.11.1.7), and GR (EC 1.6.4.2)] were assayed using Total SOD Assay Kit (S0102, Beyotime, Haimen city, China), CAT Assay Kit (S0051, Beyotime, Haimen city, China), Plant POD Assay Kit (A084-3, Nanjing Jiancheng, Nanjing city, China) and GR Assay Kit (S0055, Beyotime, Haimen city, China), respectively, according to previously described protocols (Shi et al. 2013). The GSH content, GSSG content, and GSH redox state [GSH/(GSH+GSSG)] were quantified using the GSH and GSSG Assay Kit (S0053, Haimen Beyotime, China) as previously described (Shi et al. 2013). Quantification of GUS activity Quantification of GUS activity was performed as Jefferson et al. (1987) described by detecting the amount of 4-methylumbelliferone (MU) produced from the substrate 4-methylumbelliferyl-β-glucuronide (MUG). Quantification of metabolites The metabolite extraction from 0.2 g of plant leaves and sample derivatization were carried out as Lisec et al. (2006) described. The metabolites were determined using GC-TOF-MS (Agilent 7890A/5975C, California, USA) with a DB-5MS capillary (30 m × 0.25 mm × 0.25µm, Agilent J&W GC column, California, USA) according to Shi et al. (2014b) described protocol. Various metabolites were identified by comparing retention time index specific masses with reference spectra in mass spectral libraries (NIST 2005, Wiley 7.0). Cluster analysis The hierarchical cluster analysis of differentially expressed physiological parameters was performed using CLUSTER
program
(http://bonsai.ims.u-tokyo.ac.jp/~mdehoon/software/cluster/)
and
Java
Treeview
(http://jtreeview.sourceforge.net/) as Shi et al. (2014a, b) described. Statistical analysis All experiments in this study were repeated at least three times, and one sample was harvest from ten independent 11
plants in each experiment. Means ± SEs of three independent experiments are shown in the results. In all results, Duncan’s range test was used to determine the significant difference among WT and other lines, and asterisk (*) indicates the significant difference of p
Key Laboratory of Plant Germplasm Enhancement and Specialty Agriculture, Wuhan Botanical Garden, Chinese
Academy of Sciences, Wuhan 430074, China 2
University of Chinese Academy of Sciences, Beijing 100039, China
3
Institute of Genetics, State Key Laboratory of Plant Physiology and Biochemistry, College of Life Sciences,
Zhejiang University, Hangzhou 310058, China 4
College of Agronomy, Xinjiang Agricultural University, Urumqi 830052, China
* Correspondence: [email protected]
Edited By: Ildoo Hwang, POSTECH Biotech Center, Korea
This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version and the Version of Record. Please cite this article as doi: [10.1111/jipb.12302] This article is protected by copyright. All rights reserved. Received: August 26, 2014; Accepted: October 17, 2014
1
Abstract Hydrogen sulfide (H2S) is an important gaseous molecule in various plant developmental processes and plant stress responses. In this study, the transgenic plants with modulation expressions of two cysteine desulfhydrases and exogenous H2S donor (sodium hydrosulfide, NaHS) and H2S scavenger (hypotaurine, HT) pre-treated plants were used to dissect the involvement of H2S in plant stress responses. The cysteine desulfhydrases overexpressing plants and NaHS pre-treated plants exhibited higher endogenous H2S level and improved abiotic stress tolerance and biotic stress resistance, while cysteine desulfhydrases knockdown plants and HT pre-treated plants displayed lower endogenous H2S level and decreased stress resistance. Moreover, H2S up-regulated the transcripts of multiple abiotic and biotic stress-related genes, and inhibited reactive oxygen species (ROS) accumulation. Interestingly, MIR393-mediated auxin signaling including MIR393a/b and their target genes (TIR1, AFB1, AFB2, and AFB3) was transcriptionally regulated by H2S, and was related with H2S-induced antibacterial resistance. Moreover, H2S regulated 50 carbon metabolites including amino acids, organic acids, sugars, sugar alcohols and aromatic amines. Taken together, these results indicated that cysteine desulfhydrase and H2S conferred abiotic stress tolerance and biotic stress resistance, via affecting the stress-related gene expressions, ROS metabolism, metabolic homeostasis, and MIR393-targeted auxin receptors. Keywords: Hydrogen sulfide; cysteine desulfhydrase; abiotic stress; biotic stress; MIR393; auxin receptor.
2
INTRODUCTION Along with nitric oxide (NO) and carbon monoxide (CO), hydrogen sulfide (H2S) has emerged as the third important gaseous molecule in various plant developmental processes and stress responses (Rausch and Wachter 2005; Wang 2010; Lisjak et al. 2013; Calderwood and Kopriva 2014; Hancock and Whiteman 2014). H2S promotes root organogenesis (Zhang et al. 2009), seed germination (Zhang et al. 2010), lateral root formation (Fang et al. 2014), and enhances photosynthesis (Chen et al. 2011a). Moreover, H2S confers protective roles in responses to heat, drought, salt, osmotic, and freezing stresses (Wang et al. 2010; Zhang et al. 2010b, 2011; Jin et al. 2011, 2013; Shen et al. 2012, 2013; Xie et al. 2014). Understanding the complex effects of H2S in plants requires further detailed analyses at the physiological, biochemiscal and molecular levels. In recent years, most of the in vivo roles of H2S in plants were obtained from pharmacology experiment through exogenous application of H2S donor (sodium hydrosulfide, NaHS), H2S scavenger (hypotaurine, HT) and H2S inhibitors (potassium pyruvate, PP; and hydroxylamine, HA) (Li et al. 2013; Shi et al. 2013, 2014a). However, the treatments with H2S modulating chemicals may have limitation in exactly reflecting the in vivo roles of H2S in plants. So far besides exogenous application of H2S in pharmacological experiments, it would be better to use plants with altered in planta H2S production as candidates for such an approach. To date, several proteins have been reported to be directly responsible for H2S generation in Arabidopsis, including L-cysteine desulfhydrase (LCD), D-cysteine desulfhydrase 1 (DCD1), D-cysteine desulfhydrase 2 (DCD2), and DES1 (Bloem et al. 2004; Riemenschneider et al. 2005; Papenbrock et al. 2007; Álvarez et al. 2010, 2012a, 2012b; Shen et al. 2013). The knockout mutant of AtLCD (lcd) with decreased endogenous H2S level has been used to investigate the role of H2S in plant response to drought stress (Jin et al. 2013), and expression of AtLCD and AtDCD1 in Escherichia coli showed increased cadmium resistance (Shen et al. 2012). Álvarez et al. (2012a, 2012b) found that cysteine-generated sulfide in the cytosol affected autophagy and transcriptional profile of Arabidopsis, and knockout mutant of the DES1 shows high resistance to biotrophic and necrotrophic pathogens. However, mechanisms of H2S mediated plant development and abiotic stress response remain obscure (Lisjak et al. 2013). Additionally, the connection between H2S and plant biotic stress resistance is largely unknown (Bloem et al. 2004, 2007; Lisjak et al. 2013). In this study, the transgenic plants with modulated expressions of two cysteine desulfhydrases (AtLCD and AtDCD1) together with exogenous pre-treatments with H2S donor (NaHS) and scavenger (HT), were used to dissect the involvement of H2S in plant abiotic and biotic stress responses. Further physiological, molecular, and metabolic studies were performed to investigate the underlying mechanisms of H2S-mediated abiotic and biotic stress responses, and the involvement of MIR393-targeted auxin receptors were highlighted in H2S-mediated disease resistance.
3
RESULTS Effect of abiotic and biotic stresses on cysteine desulfhydrases and the endogenous H2S content in Arabidopsis Through real-time quantitative PCR analysis, we found that the transcript levels of LCD and DCD1 increased upon all stress treatments tested (cold, dehydration, salt, hydrogen peroxide (H2O2) and bacterial pathogen) except abscisic acid (ABA) which had no effect on DCD1 transcript amount (Figure 1A). All treatments for 6 hours increased transcript amounts of LCD and DCD1 than those of treatments for 3 hours except pathogen infection, which induced highest transcript levels of LCD and DCD1 at 3 hours (Figure 1A). Consistently, the enzyme activity of LCD significantly increased after all these stress treatments, and the enzyme activity of DCD was significantly induced after dehydration, salt, H2O2 and bacterial pathogen treatments (Figure 1B). In addition, the endogenous H2S content significantly increased with 1.2-3 fold changes after stress treatments (Figure 1C). Modulation of cysteine desulfhydrase expression affects endogenous H2S level To further investigate the in vivo role of LCD and DCD1, we isolated the T-DNA mutant of LCD (lcd mutant, SAIL_739_C08/CS835466), and constructed LCD and DCD1 overexpressing plants and DCD1 RNAi knockdown plants. Consistently, the expression of LCD and DCD1 and the activities of LCD and DCD showed higher levels in the overexpressing plants, but exhibited lower levels in the knockdown plants (Figure S1A-D). LCD and DCD1 overexpressing plants exhibited significantly higher concentration of H2S, while LCD and DCD1 knockdown plants displayed significantly lower concentration of H2S (Figure 2). Moreover, 100 μM NaHS and 100 μM HT treatments significantly changed endogenous H2S level as expected (Figures 2, S2). Additionally, the accumulation of cysteine was negatively regulated by LCD and DCD1 expressions, but not affected by NaHS and HT treatments (Figure S3A). However, modulation of H2S content had no significant effect on reduced glutathione (GSH) content (Figure S3B). Modulation of cysteine desulfhydrase expression and endogenous H2S level affect abiotic stress tolerance and reactive oxygen species (ROS) metabolism Under stress conditions, although the endogenous H2S levels were up-regulated in all lines, LCD and DCD1 overexpressing plants and NaHS-treated plants showed higher concentrations of H2S compared with WT plants, while LCD and DCD1 knockdown plants and HT-treated plants displayed lower levels of them than WT plants (Figure S4). When subjected to drought, salt, and freezing, LCD and DCD1 overexpressing plants and NaHS-treated plants showed better growth and higher survival rate than those of wild type (WT) plants, while LCD and DCD1 knockdown plants and HT-treated plants exhibited worse growth and lower survival rate in comparison to WT plants (Figure 3A-B). Moreover, the increased abiotic stress tolerance of LCD and DCD1 overexpressing plants were restored by HT treatment, and the descreased abiotic stress tolerance of LCD and DCD1 knockdown plants were restored by NaHS treatment (Figure S5), indicating that H2S might be largely related with the affection of cysteine desulfhydrase on abiotic stress tolerance. Additionally, the transcripts of several abiotic stress-related genes (including CBF1, CBF3, CBF4, DREB2A, 4
DREB2B, RAB18, RD22, RD29A, and RD29B) exhibited higher levels in the LCD and DCD1 overexpressing plants and NaHS-treated plants (Figure 4 and Table S1), while lower in the LCD and DCD1 knockdown plants and HT-treated plants (Figure 4 and Table S1). This result indicated the significant effect of H2S on transcript levels of stress-related genes. Under control condition, no significant differences of ROS level and several antioxidants were observed among wild type and various lines with different endogenous H2S levels (Figure 5A-I). When abiotic stresses were applied, LCD and DCD1 overexpressing plants and NaHS-treated plants displayed lower levels of H2O2, O2•-, and oxidized glutathione (GSSG), but higher activities of superoxide dismutase (SOD), catalase (CAT), peroxidase (POD), and glutathione reductase (GR) and higher levels of GSH and GSH redox state than those of WT plants (Figure 5A-I). On the contrary, LCD and DCD1 knockdown plants and HT-treated plants exhibited the opposite effects on the ROS accumulation (Figure 5A-I). Cysteine desulfhydrase expression and H2S regulate defense resistance against bacterial pathogen As expected, NaHS and HT treatments significantly increased and decreased endogenous H2S level, respectively (Figure S6). Through quantification of bacteria number in the pathogen-infected Arabidopsis leaves, we found that the LCD and DCD1 overexpressing plants showed significantly less bacterial than WT at 3 day post infection (dpi), while LCD and DCD1 knockdown plants exhibited more bacterial than WT (Figure 6A). Consistently, H2S donor (NaHS)-treated plants exhibited improved defense resistance, while H2S scavenger (HT)-treated plants showed decreased defense resistance (Figure 6A). Additionally, the transcript levels of ENHANCED DISEASE SUSCEPTIBILITY 1 (EDS1), PHYTOALEXIN DEFICIENT 4 (PAD4), PR1, PR2, PR3, PR4, and PR5 were higher in the LCD and DCD1 overexpressing plants and NaHS-treated plants, but were lower in the LCD and DCD1 knockdown plants and HT-treated plants (Figure 6B and Table S1). Therefore, these results indicated that cysteine desulfhydrase expression and H2S regulated defense resistance against Pseudomonas, partially via modulating SA signaling-related genes. The involvement of MIR393-mediated auxin signaling in H2S-mediated antibacterial resistance The transcript levels of MIR393a and MIR393b were up-regulated by H2S effect, while the transcript levels of TRANSPORT INHIBITOR RESPONSE 1 (TIR1), AUXIN SIGNALING F BOX PROTEIN 1 (AFB1), AFB2, and AFB3 were negatively regulated by H2S effect (Fig. 7A and Table S1). In accordance with the real-time assay, the GUS activities of pMIR393a:GUS and pMIR393b:GUS transgenic plants were significantly activated by exogenous NaHS treatment, but were significantly decreased by exogenous HT treatment (Figure 7B). Under control condition, 35S:MIR393a-6, 35S:MIR393b-1, and tir1afb1afb2afb3 plants exhibited enhanced antibacterial resistance with less bacterial numbers in plant leaves at 3 dpi, while 35S:TIR1-5 and 35S:mTIR1-9 transgenic plant leaves showed more bacterial numbers at 3 dpi in comparison to WT plants (Figure 7C). To further investigate the involvement of MIR393-targeted auxin receptors in H2S-mediated antibacterial resistance, we determined the antibacterial resistance of 35S:MIR393a, 35S:MIR393b, tir1afb1afb2afb3, 35S:TIR1, and 35S:mTIR1 plants in 5
comparison to WT after the treatments of NaHS and HT. Unlike the WT plants, the improved antibacterial resistance of 35S:MIR393a-6, 35S:MIR393b-1, and tir1afb1afb2afb3 plants and the decreased antibacterial resistance of 35S:TIR1-5 and 35S:mTIR1-9 transgenic plants were not significantly affected by either NaHS or HT treatment (Figure 7C). However, the endogenous H2S level in these transgenic plants exhibited no significant difference compared with that in WT plants under both control and NaHS/HT-treated conditions (Figure 7D). These results suggested that H2S induced resistance might be mainly mediated by MIR393-dependent auxin signaling. Modulation of metabolic homeostasis by H2S effect To gain more insight into the metabolic homeostasis which might be affected by H2S, multiple carbon metabolites were assayed through gas chromatography time-of-flight mass spectrometry (GC-TOF-MS) analysis. Totally, 54 metabolites including 16 amino acids, 13 organic acids, 18 sugars, 5 sugar alcohols and 2 aromatic amines were reproducibly examined in WT (without treatment, and with NaHS and HT treatments), LCD and DCD1 overexpressing plants and knockdown mutants (Figure 8 and Table S2). Generally, 50 of 54 metabolites (except aspartic acid, lysine, hexadecanoic acid, and melibose) were significantly affected by NaHS/HT treatment and in the plants with altered LCD or DCD activity. Interestingly, most of the metabolite concentrations were higher in NaHS-treated and LCD-OX/DCD1-OX plants and lower in HT-treated and LCD/DCD1 knockdown plants (Figure 8 and Table S2), indicating the modulation of H2S in several carbon metabolites including amino acids, organic acids, sugars, sugar alcohols and aromatic amines.
DISCUSSION Previous studies dissecting the in vivo roles of H2S in plants were mainly obtained from pharmacology experiment using H2S donor, scavenger and inhibitors (Li et al. 2013; Shi et al. 2013, 2014a). Therefore, it is difficult to distinguish the effects of H2S from its upstream (cysteine) and other downstream metabolites of cysteine (such as glutathione) (Wu et al. 2010; Cao et al. 2013, 2014; Calderwood and Kopriva 2014). In this study, modulation of cysteine desulfhydrase expression affected endogenous H2S and cysteine levels, but had no significant effect on the accumulation of GSH under control condition (Figures 2, S1, S2). However, exogenous treatments of H2S donor (NaHS) and scavenger (HT) specifically affected the endogenous H2S, and did not affect the accumulations of cysteine levels and GSH under control condition (Figures 2, S3). Since the endogenous sulfate metabolism is very complex, the upstream and downstream metabolites of H2S might form part of a feedback mechanism to reset its homeostasis. In this study, using the transgenic plants with modulated endogenous H2S level together with the exogenous treatments of H2S donor (NaHS) and scavenger (HT), we assigned the protective role of H2S in both abiotic stress tolerance and biotic stress resistance (Figures 3, 6). This result is in accordance with the enhanced abiotic stress tolerance by H2S donor (NaHS) in various plant species, and this is the first report to show the protective role of H2S in antibacterial resistance. H2S and abiotic stress tolerance in Arabidopsis 6
When subjected to abiotic stress conditions, plant ROS level (mainly including H2O2 and O2•-) was largely induced, along with induced unstabilization of plasma membrane and decreased cell turgor (Mittler et al. 2004, 2011; Miller et al. 2010). Consistently, plants have evolved complex ROS defence system, including several enzymatic antioxidant enzymes (SOD, CAT, POD, GR, etc) and non-enzymatic antioxidant molecules (such as glutathione) to regulate endogenous ROS level. In this study, increased endogenous H2S and exogenous application of NaHS significantly activated ROS detoxification system, including enzymatic antioxidant enzymes and non-enzymatic glutathione to make cellular ROS at the relatively low level under abiotic stress conditions, thus conferring improved abiotic stress tolerance (Figure 5). When the endogenous H2S level decreased, significantly increased level of ROS and decreased levels of antioxidants resulted in decreased abiotic stress tolerance (Figure 5). Thus, the modulation of endogenous and exogenous H2S in ROS detoxification system might be directly connected with the effect of H2S on abiotic stress tolerance. In Arabidopsis, CBF1, CBF2, CBF3 (also known as DREB1b, DREB1c, DREB1a, respectively) and CBF4 have
important
roles in
cold, salt
and drought stress
responses via binding to
the C-repeat
(CRT)/dehydration-responsive element (DRE) cis-acting element of several stress-responsive genes (Seki et al. 2001; Haake et al. 2002; Achard et al. 2008; Novillo et al. 2012). When constitutively overexpressing AtCBF1/2/3/4, the expression of several downstream genes such as COR (cold related), LTI (low-temperature induced), KIN (cold inducible), RD (responsive to dehydration) and ERD (early responsive to dehydration) were largely induced, and the transgenic plants conferred improved abiotic stress tolerance (Seki et al. 2001; Haake et al. 2002; Achard et al. 2008; Novillo et al. 2012). Moreover, ABA2 and NCED3 are responsible for ABA generation in Arabidopsis (Chan 2012). Abiotic stress responses in plants are initiated via both ABA-dependent and ABA-independent signal transduction pathways (Jin et al. 2011; Novillo et al. 2012). CBF4 is ABA-independent, and RAB18, RD22, RD29A, and RD29B are ABA dependent stress-responsive genes (Jin et al. 2011; Novillo et al. 2012). In this study, the modulation of H2S in the transcripts of multiple abiotic stress-related genes (CBF1, CBF3, CBF4, DREB2A, DREB2B, RAB18, RD22, RD29A, and RD29B) might be directly related with the abiotic stress tolerance. H2S and biotic stress resistance in Arabidopsis The transcripts of EDS1 and PAD4, which are responsible for SA biosynthesis in Arabidopsis, were constitutively activated in the plants with higher H2S levels, but were significantly reduced in the plants with lower H2S levels (Figure 6B). SA accumulation is necessary for the activation of SA downstream gene expression such as PRs (Wang et al. 2007; Kazan and Manners 2009; Fu and Dong 2013). Consistently, the plants with higher H2S levels showed increased expression of SA-dependent PR genes, resulting in improved immunity resistance, while the plants with lower H2S levels exhibited decreased immunity resistance (Figure 6). Álvarez et al. (2012a) found that knockout mutant of the DES1 showed improved resistance to biotrophic and necrotrophic pathogens, SA accumulation and PR1 induction. The controversy might be attributed to at least three aspects as described following. Firstly, the effect of LCD1 and DES1 on sulfate metabolism is different. Although they have the same 7
effects on cysteine content, they modulate endogenous GSH and H2S levels differently (Figures 2, S3; Álvarez et al. 2012a). Modulation of LCD1 expression affected H2S levels (Figures 2, S4) but had no significant effects on endogenous GSH level (Figure S3) and DES1 transcript (Figure S7), while DES1 knockout mutant displayed higher concentration of GSH, but no report about its effect on H2S level (Álvarez et al. 2012a). Secondly, 28-day-old plants were used for disease resistance assay in this study, whereas 6-7-week-old plants were used by Álvarez et al. (2012a), and the different growth stage might have differences in sulfate metabolism including cysteine, GSH, and H2S levels. Additionally, Álvarez et al. (2012a) found that DES1 was localized in cytoplasm. LCD is predicted to be localized in cytoplasm, but DCD1 is predicted to be localized in chloroplast, cytosol, and mitochondrion. The differential subcellular localization might also contribute to the contradictory results. Wang et al. (2007) showed that SA and auxin act individually or through antagonistic crosstalk, and the finely tuned balance between them is critical for plant-pathogen interaction. Shen et al. (2013) showed that the transcripts of multiple miRNAs including MIR167, MIR393, MIR396, and MIR398 were significantly up-regulated by H2S. Among these miRNAs, MIR393 has been shown to be involved in plant-pathogen interaction (Navarro et al. 2006; Robert-Seilaniantz et al. 2011). Overexpression of MIR393, a microRNA that targets auxin receptors (TIR1, AFB1, AFB2, and AFB3), renders plants more resistant to bacterial pathogens, while over-expression of AFB1 results in plants more susceptible to bacterial pathogens (Navarro et al. 2006; Chen et al. 2011b; Robert-Seilaniantz et al. 2011; Bian et al. 2012; Liu et al. 2013). Besides SA-related genes, the transcript levels of MIR393a/b and their target genes (TIR1, AFB1, AFB2, and AFB3) were up-regulated and down-regulated by H2S, respectively (Figure 7A, B). Since the transcript of MIR393 is induced by IAA (Chen et al. 2011b) and H2S (Shen et al. 2013 and this study), the connection between auxin and H2S is further investigated in this study. DR5:GUS is a widely used auxin-related marker line under the control of the auxin-responsive DR5 promoter (Ulmasov et al. 1997), and NaHS and HT treatments had no significant effects on the GUS activities of DR5:GUS line (Figure S8). These studies suggested that H2S affected only the transcripts of MIR393 and its targeted auxin receptors, but not auxin responses. As shown in Figure 7C, MIR393a/b and their target auxin receptors mediated antibacterial resistance was not affected by NaHS and HT treatments, suggesting that MIR393-repressing auxin signaling might be largely contributed to H2S-mediated antibacterial resistance. Thus, H2S conferred plant immunity resistance via modulating both the transcript levels of SA-related genes and MIR393-mediated auxin signaling. H2S and metabolic homeostasis in Arabidopsis Not only stress-related genes, but also several metabolites were significantly affected by H2S effect (Figure 8 and Table S2). Notably, higher endogenous H2S level largely activated the accumulation of compatible solutes such as proline, arabinose, sucrose, lactulose, allose, fructose, lactose, tagatofuranose, talose, mannose, galactinol, dulcitol (Figure 8 and Table S2). Proline and some carbohydrates are important compatible solutes to respond to abiotic stress for osmotic adaptation (Krasensky and Jonak 2012). Thus, higher levels of proline and the carbohydrates might provide beneficial effect in response to stress conditions. On the contrary, lower endogenous H2S level exhibited lower concentrations of these compatible solutes (Figure 8 and Table S2), which might be related with 8
the decreased stress resistance. Among these amino acids, the mutant and HT-treated plants displayed significantly higher concentration of alanine, but lower concentrations of other amino acids, indicating the complex role of H2S in regulating amino acid pools. Asparagine accumulation shows that nitrogen re-distribution and mobilization are important features of the salt stress response (Maaroufi-Dguimi et al. 2011). In addition, asparagine was an amino group donor for the synthesis of the photorespiratory intermediate glycine, and Nagy et al. (2013) found this was also a good indicator of drought stress in drought tolerant and sensitive wheat cultivars. In the meanwhile, the modulation of other amino acids and organic acids by H2S treatment indicated common changes of H2S in carbon metabolism, and might contribute to the affected stress resistance. Based on the above results, a model for H2S-mediated stress responses in Arabidopsis was proposed (Figure 9). In response to various abiotic and biotic stresses, the H2S synthetic pathway was largely activated and H2S level was significantly increased. The increased endogenous H2S and exogenous application of H2S donor, on one hand constitutively activated the expression of several abiotic stress-related genes and SA signaling genes, and accumulate of compatible solutes such as proline and soluble sugars; on the other hand, changes of antioxidant enzyme activities and glutathione redox state under stress conditions, conferred improved both abiotic stress tolerance and biotic stress resistance (Figure 9). When the endogenous H2S was down-regulated, the opposite results were observed. Moreover, MIR393-targeted auxin receptors were regulated by H2S, and were largely contributed to H2S-mediated antibacterial resistance (Figure 9). Taken together, this study assigned the protective role of H2S in Arabidopsis responses to abiotic and biotic stresses, via regulating multiple stress-related gene transcripts, ROS metabolism, metabolic homeostasis and MIR393-targeted auxin signaling pathways.
MATERIALS AND METHODS Plant materials and growth conditions Arabidopsis thaliana (ecotype Columbia-0) seeds were sterilized with 70% (v/v) ethyl alcohol, and 10% (w/v) NaClO and then washed with deionized water. The seeds were sown on Murashige and Skoog (MS) medium plate containing 1% (w/v) sucrose or in soil after stratification at 4 °C for 3 days in darkness. The plants were then kept in the growth chamber, which was controlled at 23 °C and at an irradiance of about 150 µmol quanta m-2 s-1, with 65% relative humidity under 16 hour light and 8 hour dark cycles. Nutrient solution of modified Hoagland solution (Tocquin et al. 2003) was watered twice in the bottom of the pots with plants every week. The lcd mutant (SAIL_739_C08/CS835466) was obtained from the Arabidopsis Biological Resource Center, and the transgenic lines of pMIR393a:GUS, pMIR393b:GUS, 35S:MIR393a-6, 35S:MIR393b-1, 35S:TIR1-5 and 35S:mTIR1-9 have been described in Chen et al. (2011b), and the tir1afb1afb2afb3 mutant through crossing tir1 (CS3798), afb1 (SALK_070172), afb2 (SALK_137151), and afb3 (SALK_068787) has been described in Liu et al. (2013). To determine the effects of different stresses on the expression and activities of AtLCD and AtDCD1, 28-day-old Arabidopsis Col-0 plants were treated by cold stress (4 °C), dehydration (put the plants in a 9
un-covering the plate in the growth chamber), or watered by 50 µM ABA, 200 mM NaCl or 20 mM H2O2 in the pots with plants for 3 and 6 hours, or the plant leaves were infected by Pst DC3000 as Shi et al. (2012) described at OD600 = 0.001 for 1, 3 and 6 hours. The plant leaves were then harvested at indicated timepoints for further analysis. Transgenic plasmid construction and plant transformation For the AtLCD and AtDCD1 overexpressing constructs, the coding regions of AtLCD and AtDCD1 were amplified and cloned into pBARN vector under the control of CaMV 35S promoter with basta resistance (LeClere and Bartel 2001). For the amiR-AtDCD1 knockdown transgenic construct, amiR-AtDCD1 fragment were amplified from the plasmid pRS300 by PCR using specific primers from WMD3 (http://wmd3.weigelworld.org/cgi-bin/webapp.cgi) (Schwab et al. 2006), and the amiR-AtDCD1 fragment was then cloned into pBARN vector. The primers for vector constructs were listed in Table S3. The above recombinant constructs were then transformed into Agrobacterium tumefaciens strain GV3101 and introduced into Columbia-0 plants using the floral dip method (Clough and Bent 1998). Homozygous transgenic plants were selected on MS medium using basta resistance and were confirmed by PCR analysis. RNA isolation and real-time quantitative PCR Total RNA was extracted from plant leaves using TRIzol reagent (Invitrogen, California, USA) and RQ1 RNase-free DNase (Promega, Wisconsin, USA) was added to avoid possible genomic DNA contamination. First-strand cDNA was then synthesized using reverse transcriptase (TOYOBO, Osaka city, Japan) from 2 µg of total RNA. Real-time quantitative PCR was carried out using CFX96TM Real Time System (BIO-RAD, California, USA) with iQTM SYBR® Green Super mix (BIO-RAD, California, USA) as Shi et al. (2012) described. Ubiquitin 10 (UBQ10) was chosen as the reference gene in the treatments according to geNorm software (Czechowski et al. 2005), and housekeeping genes of ACTIN2 and EIF4A were used as the control. The gene transcript levels were standardized with ubiquitin 10 (UBQ10) using comparative ΔΔCT method, and all experiments were repeated at least three times. The specific primers for real-time quantitative PCR were listed in Table S4. Determination of H2S concentration H2S concentration was assayed as Chen et al. (2011a) described that based on formation of methylene blue. Briefly, 1 g of plant leaves were ground in 10 mL of extraction buffer (50 mM phosphate buffer, pH 6.8, 0.2 M ascorbic acid, and 0.1 M EDTA). The homogenate was mixed with 1 ml of 1 M HCl to release H2S, and H2S was absorbed in a 1 mL of 1% (w/v) zinc acetate trap. After 30 min of reaction, 0.5 mL of 5 mM dimethyl-p-phenylenediamine dissolved in 5 mM H2SO4 was added to the trap, and then 0.5 ml of 50 mM ferric ammonium sulphate in 100 mM H2SO4 was added into the trap. The concentration of H2S in zinc acetate traps was examined at 667 nm of absorbance. Determination of abiotic and biotic stress resistances Plant abiotic stress tolerance was assayed as Shi et al. (2012) described, and survival rate was determined at 4 d after recovery from the abiotic stress treatment. For drought stress treatment, 14-day-old plants were withheld 10
water for 21 days and then re-watered for 4 days. For salt stress treatment, 14-day-old plants were watered with NaCl solution, and the NaCl concentration was increased stepwise by 50 mM every two days to 150 mM for another 21 days. For freezing stress treatment, 2-week-old seedlings were cold acclimated at 4 °C for 14 days, then the plants were placed at -8 °C for 8 hours, and then transferred to a growth chamber at 22 oC for another 4 days. For disease resistance assay, 28-d-old plant leaves were infected with the pathogen strain of Pst DC3000 at OD600 = 0.001 as described by Shi et al. (2012) described, and the bacterial growth from infected Arabidopsis leaves was monitored at 0 and 3 days post infection (dpi). Determination of ROS level and antioxidant activities 0.5 g of plant leave samples were harvest from ten independent plants in each experiment and used for the ROS assay. H2O2 and O2•- contents were quantified using the titanium sulfate method and the Plant O2•- ELISA Kit (10-40-488, Dingguo, Beijing, China), respectively, as Shi et al. (2013) described. Determination of enzyme activities LCD and DCD activities were determined by measuring the production rate of H2S from L-cysteine and D-cysteine as Jin et al. (2013) described. The activities of antioxidant enzymes [SOD (EC 1.15.1.1), CAT (EC 1.11.1.6), POD (EC 1.11.1.7), and GR (EC 1.6.4.2)] were assayed using Total SOD Assay Kit (S0102, Beyotime, Haimen city, China), CAT Assay Kit (S0051, Beyotime, Haimen city, China), Plant POD Assay Kit (A084-3, Nanjing Jiancheng, Nanjing city, China) and GR Assay Kit (S0055, Beyotime, Haimen city, China), respectively, according to previously described protocols (Shi et al. 2013). The GSH content, GSSG content, and GSH redox state [GSH/(GSH+GSSG)] were quantified using the GSH and GSSG Assay Kit (S0053, Haimen Beyotime, China) as previously described (Shi et al. 2013). Quantification of GUS activity Quantification of GUS activity was performed as Jefferson et al. (1987) described by detecting the amount of 4-methylumbelliferone (MU) produced from the substrate 4-methylumbelliferyl-β-glucuronide (MUG). Quantification of metabolites The metabolite extraction from 0.2 g of plant leaves and sample derivatization were carried out as Lisec et al. (2006) described. The metabolites were determined using GC-TOF-MS (Agilent 7890A/5975C, California, USA) with a DB-5MS capillary (30 m × 0.25 mm × 0.25µm, Agilent J&W GC column, California, USA) according to Shi et al. (2014b) described protocol. Various metabolites were identified by comparing retention time index specific masses with reference spectra in mass spectral libraries (NIST 2005, Wiley 7.0). Cluster analysis The hierarchical cluster analysis of differentially expressed physiological parameters was performed using CLUSTER
program
(http://bonsai.ims.u-tokyo.ac.jp/~mdehoon/software/cluster/)
and
Java
Treeview
(http://jtreeview.sourceforge.net/) as Shi et al. (2014a, b) described. Statistical analysis All experiments in this study were repeated at least three times, and one sample was harvest from ten independent 11
plants in each experiment. Means ± SEs of three independent experiments are shown in the results. In all results, Duncan’s range test was used to determine the significant difference among WT and other lines, and asterisk (*) indicates the significant difference of p
