• Open Access

Microarray analysis of nitric oxide responsive transcripts in Arabidopsis

Authors


Correspondence (fax +1 419 530 1539; e-mail Sgoldma@utnet.utoledo.edu)

Summary

Nitric oxide (NO) is emerging as an important signalling molecule with diverse physiological functions in plants. In the current study, changes in gene expression in response to 0.1 mm and 1.0 mm sodium nitroprusside (SNP), a donor of NO, were studied in Arabidopsis using the whole genome ATH1 microarray, representing over 24 000 genes. We observed 342 up-regulated and 80 down-regulated genes in response to NO treatments. These included 126 novel genes with unknown functions. Most of these changes were specific to NO treatment, as we observed a reverse trend when the plants were treated with NO scavenger, 2-[4-carboxyphenyl]-4,4,5,5-tetramethylimidazoline-1-oxy-3-oxide (c-PTIO). Hierarchical clustering revealed 162 genes showing a dose-dependent increase in signal from 0.1 mm SNP to 1.0 mm SNP treatment. We observed the up-regulation of several genes encoding disease-resistance proteins, WRKY proteins, transcription factors, zinc finger proteins, glutathione S-transferases, ABC transporters, kinases and biosynthetic genes of ethylene, jasmonic acid, lignin and alkaloids. This report provides an insight into the molecular basis for the seemingly diverse biological functions of NO in plants. Interestingly, about 2.0% of the genes in Arabidopsis responded to NO treatment, about 10% of which were transcription factors. NO may also influence the plant's signal transduction network as indicated by the transcriptional activation of several protein kinases, including a mitogen-activated protein (MAP) kinase. We identified many genes previously not shown to be associated with NO responses in plants, and this is the first report of NO responsive genes based on a whole genome microarray.

Introduction

Nitric oxide (NO) was first discovered as an endothelium-derived relaxing factor in smooth muscle cells (Ignarro et al., 1987). Subsequently, intensive research on the biological functions of NO in humans and animals have recognized NO as a critical signalling molecule with life-saving properties (for a review, see Lamattina et al., 2003). Recently, the first mammalian type of response to NO in plants was convincingly shown (Delledonne et al., 1998). Since then, NO has been shown to be involved in several plant functions, including defence response (Delledonne et al., 1998), stress response (Garcia-Mata and Lamattina, 2001), growth and development (Beligni and Lamattina, 2000; Pagnussat et al., 2002), senescence (Leshem and Pinchasov, 2000), iron homeostasis (Murgia et al., 2002) and cell death (Zottini et al., 2002). NO has also been shown to interact with the signalling pathways of plant hormones, such as indole acetic acid (IAA) and abscisic acid (ABA) (Garcia-Mata and Lamattina, 2002; Pagnussat et al., 2003).

NO in plants can be enzymatically produced by nitrate reductase (Yamasaki et al., 1999) and nitric oxide synthase (Guo et al., 2003) or by non-enzymatic reduction of apoplastic nitrite under acidic conditions (Bethke et al., 2004). As an alternative to endogenous production, NO may enter the plant cell from the atmosphere or soil (Beligni and Lamattina, 2001). NO induces direct and indirect changes in gene expression and protein functions in animals (Mayer and Hemmens, 1997; Xu et al., 1998). The analysis of such molecular changes in plants is necessary to understand its effect on metabolism and physiological functions. We investigated the transcript changes in Arabidopsis in response to NO using the whole genome microarray. Several NO-induced and NO-repressed genes were identified, and the results are discussed in the context of the diverse biological functions of NO in plants.

Results and discussion

NO in plants has been reported to be involved in several physiological and developmental functions, which are very diverse in nature. The literature contains numerous physiological studies reporting the effect of exogenously applied NO on plant functions. However, recently, considerable effort has been made to understand the response to NO at the molecular level (Huang et al., 2002; Kumar and Klessig, 2000; Murgia et al., 2002; Polverari et al., 2003). Given the diverse nature of NO functions, it is more informative to conduct large-scale gene expression studies in order to understand overlapping functions and signalling networks. This is evident from the report by Polverari et al. (2003) that NO in Arabidopsis induces several transcripts involved in signal transduction and basic metabolism. However, in this study, only 2500 transcripts were surveyed using the cDNA-amplified fragment length polymorphism (AFLP) technique, which is qualitative and depends on the sequencing of the differentially displayed fragments for transcript identification. We took advantage of the recently released ATH1 Genome Array and microarray hybridization for direct quantitative measurement of changes in several thousand transcripts simultaneously. The ATH1 array includes 22 500 probe sets representing approximately 24 000 gene sequences of Arabidopsis, and thus can provide genome-scale transcript changes in response to NO treatment.

To assess the effects of different NO concentrations on cellular transcriptional response, Arabidopsis roots were treated with 0.1 mm and 1.0 mm of sodium nitroprusside (SNP), a donor of NO. Plants treated with water were used as controls. To ensure that the observed effects were NO specific, a set of plants treated with 0.1 mm SNP was subsequently treated with 0.1 mm NO scavenger, 2-[4-carboxyphenyl]-4,4,5,5-tetramethylimidazoline-1-oxy-3-oxide (c-PTIO). Total RNA from the leaf tissue of control and treated plants was labelled, and hybridized to Arabidopsis ATH1 microarrays. Based on the hybridization data, NO-induced and NO-repressed transcripts were identified by following standard selection criteria and statistical methods as described in ‘Experimental procedures’. Reverse transcriptase polymerase chain reaction (RT-PCR) was carried out for 16 putative NO responsive genes to confirm the microarray data (Figure 1). The expression patterns obtained by RT-PCR were consistent with those obtained by microarray analysis. For example, the relative induction of glucosyltransferase and squalene monooxygenase was higher with 0.1 mm SNP treatment, whereas the induction of ABC transporter and glutathione S-transferase (GST) was higher with 1.0 mm SNP treatment. Auxin-regulated protein transcript was repressed with 0.1 mm SNP treatment, and proline-rich protein transcript was repressed with 1.0 mm SNP treatment. Furthermore, for select genes, the fold-change in expression was quantified using real-time PCR. The tested genes exhibited approximately 2–50-fold changes in expression. For all the transcripts tested, the average fold-change estimated from real-time PCR data was higher than that from microarray data (Figure 2).

Figure 1.

Reverse transcriptase polymerase chain reaction (RT-PCR) analysis of changes in gene expression in response to nitric oxide (NO). Arabidopsis plants were treated with 0.1 mm and 1.0 mm solutions of an NO donor, sodium nitroprusside (SNP), for 3 h. R1, R2 and R3 are three biological replicates in each treatment. The control plants were treated with distilled water for 3 h. For 2-[4-carboxyphenyl]-4,4,5,5-tetramethylimidazoline-1-oxy-3-oxide (c-PTIO) treatment, the plants previously treated with 0.1 mm SNP for 3 h were treated for another 3 h with 0.1 mm c-PTIO. The RT-PCR products were separated in 1.5% agarose gels, and a 1 kb DNA ladder (Invitrogen) was used as a marker.

Figure 2.

Comparison of gene expression data from microarray hybridization and real-time polymerase chain reaction (PCR). Changes in gene expression were estimated as the fold-change over the control. The genes tested and the source of RNA are indicated below each group of bars. The bars represent the average fold-change and standard error in transcript changes estimated from three biological replications for both real-time PCR and microarray hybridization.

We identified 124 genes up-regulated in response to 0.1 mm SNP treatment, and 261 genes up-regulated in response to 1.0 mm SNP treatment. These included 43 genes up-regulated in both treatments (available as ‘Supplementary material’). In total, SNP treatments up-regulated 342 unique genes. The expression data for these 342 genes were compared between the SNP treatments and the c-PTIO treatment. For approximately 96% of the genes, a decline in average fold-change was observed with the c-PTIO treatment compared with the SNP treatment. For 72%, the decline was more than twofold when compared with the average fold-change with either 0.1 or 1.0 mm SNP treatment. These data suggest that the observed effect on the transcripts is specific to NO, because c-PTIO reacts rapidly with NO to yield NO2/NO3 (Akaike et al., 1993). The same approach has been used in several studies to demonstrate the NO specificity of physiological and molecular responses (Garcia-Mata and Lamattina, 2001; Murgia et al., 2002; Pagnussat et al., 2003). NO down-regulated 50 genes with 0.1 mm SNP treatment, 34 genes with 1.0 mm SNP treatment and four genes with both treatments (available as ‘Supplementary material’). In total, SNP treatments down-regulated 80 unique genes. Hierarchical clustering showed 162 genes with a dose-dependent increase in transcript abundance from 0.1 mm to 1.0 mm SNP treatment, followed by a decrease on c-PTIO treatment (Figure 3). The current data demonstrate that NO affects transcript levels in numerous genes from diverse pathways. These transcripts could be related to plant defence response, protection against oxidative stress, iron homeostasis, signal transduction and gene expression control through transcription factors. However, 126 genes (98 NO-induced and 28 NO-repressed) code for proteins with unknown functions. A selective list of genes up-regulated by NO is given in Table 1, and complete lists are available as ‘Supplementary material’.

Figure 3.

Hierarchical clustering. Hierarchical clustering was performed using microarray hybridization signal data with the correlation measure-based distance and average linkage clustering method. Data from the control, three biological replicates of 0.1 mm sodium nitroprusside (SNP) treatment, 1.0 mm SNP treatment and 0.1 mm 2-[4-carboxyphenyl]-4,4,5,5-tetramethylimidazoline-1-oxy-3-oxide (c-PTIO) treatment were used. The c-PTIO treatment was given for 3 h to plants previously treated with 0.1 mm SNP for 3 h. A subcluster (N = 162) showing a dose-dependent increase in transcript abundance from 0.1 mm to 1.0 mm SNP treatment, followed by a decrease on c-PTIO treatment, is indicated.

Table 1.  Select list of genes up-regulated in response to the NO donor, sodium nitroprusside (SNP), at 0.1 mm and 1.0 mm concentrations. The transcript response to the NO scavenger, 2-[4-carboxyphenyl]-4,4,5,5-tetramethylimidazoline-1-oxy-3-oxide (c-PTIO), at 0.1 mm concentration is given for comparison. The NO-induced transcripts in this list had change P value < 0.05 and coefficient of variation < 30%
Probe/gene IDFunctionAnnotationFold-change
SNP treatment (average ± SD)c-PTIO treatment
AffymetrixAGI0.1 mm1.0 mm0.1 mm
258452_atAt3g22370Pathogen inducedAlternative oxidase 1a  2.79 ± 0.37−1.24
254447_atAt4g20860 Berberine bridge enzyme  2.59 ± 0.11−3.46
259879_atAt1g76650 Calmodulin 2.87 ± 0.46  3.42 ± 0.12−6.63
251895_atAt3g54420 Chitinase class IV  4.01 ± 1.15−1.45
262381_atAt1g72900 Virus-resistance protein 2.47 ± 0.46  4.50 ± 0.11−1.25
259297_atAt3g05360R genesDisease-resistance protein Cf-2  6.41 ± 1.70−1.32
260296_atAt1g63750 Disease-resistance protein RPP1-WsC  2.01 ± 0.24−8.75
249125_atAt5g43450Ethylene biosynthesisACC oxidase  2.22 ± 0.16−2.45
254926_atAt4g11280 ACC synthase 6 3.22 ± 0.45  2.70 ± 0.06−2.64
255787_atAt2g33590Lignin biosynthesisCinnamoyl-CoA reductase  1.99 ± 0.06−1.06
259911_atAt1g72680 Cinnamyl-alcohol dehydrogenase  2.95 ± 0.38−1.13
251603_atAt3g57760KinasesWall-associated kinase 1 10.84 ± 1.59−1.22
245731_atAt1g73500 MAP kinase kinase 5 2.11 ± 0.43−1.97
251479_atAt3g59700 Lectin receptor-like kinase  2.52 ± 0.16−1.46
254408_atAt4g21390 Serine/threonine kinase  2.89 ± 0.33−2.50
246858_atAt5g25930 Receptor-like protein kinase  2.59 ± 0.51−1.89
266821_atAt2g44840Transcription factorsEthylene responsive element-binding protein13.27 ± 1.19 4.32
253259_atAt4g34410 Ethylene responsive element-binding protein  4.12 ± 0.73−3.12
257919_atAt3g23250 Myb-related transcription factor 15.49 ± 2.91−1.93
259705_atAt1g77450 GRAB1-like protein, NAC domain protein 10.49 ± 2.25 1.52
263783_atAt2g46400 WRKY-type transcription factor WRKY46 3.10 ± 0.52−1.38
266010_atAt2g37430 C2H2-type zinc finger protein 16.87 ± 4.31−2.69
254066_atAt4g25480 Transcriptional activator CBF1/DREB1B 1.92 ± 0.23 1.13
250781_atAt5g05410 Transcriptional activator DREB2A  2.68 ± 0.34 1.16
262518_atAt1g17170Cellular detoxificationGlutathione S-transferase 7.92 ± 1.03180.28 ± 49.22 6.36
267319_atAt2g34660 ABC transporter (AtMRP2)  3.13 ± 0.30 1.16
261443_atAt1g28480 Glutaredoxin  3.72 ± 0.68−1.10
246464_atAt5g16980 Quinone oxidoreductase  6.95 ± 0.62 1.43
258979_atAt3g09440Heat shock proteinHeat shock protein AtHsc70–3 2.26 ± 0.50  3.52 ± 0.54 1.80
262883_atAt1g64780OthersAmmonium transporter 3.21 ± 0.92  2.84 ± 0.43 1.06
265199_s_atAt2g36770 Glucosyltransferase 17.65 ± 4.45−1.14
249774_atAt5g2415 Squalene monooxygenase 2.44 ± 0.43−1.01
251109_atAt5g01600 Ferritin 1 3.44 ± 0.70  5.41 ± 0.40 2.36
264758_atAt1g61340 LEA protein 2.41 ± 0.31  3.97 ± 0.98−1.82
257644_atAt3g25780UnknownUnknown protein 2.44 ± 0.54  3.29 ± 0.28−1.64
257670_atAt3g20340 Unknown protein 49.53 ± 10.85 4.38
266800_atAt2g22880 Hypothetical protein 8.32 ± 0.93−3.92

NO potentiates the induction of hypersensitive cell death, regulates some aspects of systemic acquired resistance and also induces defence-related proteins (Delledonne et al., 1998). Huang et al. (2002) have identified several NO-induced, defence-related genes in Arabidopsis suspension cells using cDNA microarrays. Recently, Polverari et al. (2003) have identified a few NO-induced genes in Arabidopsis leaves that are specifically involved in disease resistance. In the current study, by using whole genome Arabidopsis arrays, NO treatment was observed to increase the transcript level of several typical pathogen-induced genes, including R genes with nucleotide-binding site leucine-rich repeats (NBS-LRRs), NDR1 (non-race-specific disease resistance), RPP1-WsA-C and genes for disease-resistance proteins. The identification of these and other genes (see below) induced by NO supports its role in plant defence response, as reported previously (Delledonne et al., 1998; Durner et al., 1998). In addition, NO up-regulated several alkaloid biosynthetic genes, such as berberine bridge enzyme, steroid sulphotransferase and strictosidine synthase, which may also contribute to disease resistance in plants. NO also up-regulated several transcription factors that moderate plant defence responses. Of particular interest are WRKY transcription factors, which bind to the W box present in the promoters of many plant defence genes (Maleck et al., 2000). We observed an approximate two- to fourfold induction of WRKY33, 40 and 46 in response to NO treatments. Previous studies have shown that salicylic acid and pathogen infection induce WRKY46 (Kalde et al., 2003), and wounding induces WRKY33 and WRKY40 (Cheong et al., 2002). The other class of defence-related transcription factors induced by NO was the ethylene responsive element-binding proteins (EREBPs). SNP treatment induced transcripts coding for several EREBPs, by 2–13-fold over control expression. Interestingly, NO also up-regulated the genes for ACC synthase and ACC oxidase, the two enzymes required for ethylene biosynthesis. We have not determined whether this actually resulted in increased ethylene synthesis. However, it should be noted that indirect evidence suggests that NO may have an antagonistic effect on ethylene content in fruits (Leshem and Pinchasov, 2000).

NO has been reported to affect the DNA-binding property of transcription factors with zinc finger motifs (Kroncke et al., 2001). However, we observed the up-regulation of transcripts coding for zinc finger proteins. The activation of some of these genes was strong, as indicated by an approximate 16- and 61-fold increase in the transcripts of zinc finger proteins, At2g37430 and At3g28210, after 1.0 mm SNP treatment (Table 1 and ‘Supplementary material’). The biological significance of the transcriptional activation of zinc finger proteins in response to NO in plants is not yet known. NO treatment also induced transcripts coding for dehydration responsive element-binding proteins (DREB1 and DREB2) and late embryogenesis abundant (LEA) proteins. DREB1 and DREB2 are transcriptional activators of genes, which in turn confer cold and drought tolerance (Shinwari et al., 1998). The overexpression of LEA proteins could provide drought and salinity tolerance in plants (Xu et al., 1996). These observations may be related to the previous report that drought tolerance in wheat was enhanced after NO treatment (Garcia-Mata and Lamattina, 2001). The identification of a large number of transcription factors induced by NO is significant in the light of its diverse physiological and developmental functions in plants. Surprisingly, no such transcription factors were identified in a similar experiment when the cDNA-AFLP technique was used (Polverari et al., 2003).

NO at higher concentrations can cause oxidative stress and cell death (Zottini et al., 2002). We observed a 2.79-fold induction of alternative oxidase with 1.0 mm SNP treatment (Table 1), which may help to counteract the inhibition of cytochrome oxidase and confer tolerance to NO toxicity (Huang et al., 2002). In addition, NO-induced GSTs and ABC transporters may scavenge toxic compounds generated during oxidative stress. NO up-regulated 10 of the 48 GSTs identified in Arabidopsis. Polverari et al. (2003) also identified one of these NO-induced GSTs (At2g47730) using the cDNA-AFLP technique. Indeed, the most highly induced gene in our experiment (At1g17170; 180-fold increase) encoded for a GST (Table 1). Seven of the full-molecule ABC transporters up-regulated in the current study encoded multiple resistance-associated proteins (AtMRPs) that function as glutathione S-conjugate pumps (Leier et al., 1996). GSTs and ABC transporters are considered as markers of environmental stress response; however, recent reports have indicated that at least some of them may have a specific role in plant physiology and development. For example, AtMRP5 has been shown to be uniquely involved in root development and stomatal movement (Gaedeke et al., 2001), and proteomic analysis revealed a role for GSTs in the development of early root epidermis in Arabidopsis (Mang et al., 2004). Therefore, it will be of interest to study the specific roles played by the NO-induced GSTs and ABC transporters in plants.

In mammals, NO is involved in iron homeostasis through iron regulatory proteins, which bind to iron responsive elements (IREs) in the ferritin mRNA (Gardner et al., 1997). There are no canonical IRE sequences in plant ferritin mRNAs. However, Murgia et al. (2002) reported that NO participates in iron homeostasis in plants through the iron-dependent regulatory sequence (IDRS) in the ferritin promoters. There are four ferritin genes in Arabidopsis, and NO increased the transcripts of AtFer1 (At5g01600, 5.4-fold), AtFer3 (At3g56090, 4.3-fold) and AtFer4 (At2g40300, 1.5-fold). Interestingly, these three ferritin genes were induced by iron in the same order of strength (Petit et al., 2001). Therefore, NO may regulate only iron responsive ferritins, although additional work is required to determine whether a common regulatory molecule is involved. The increased ferritin content may sequester free iron and protect the plant cells from oxidative damage (Deak et al., 1999).

Another important finding from our study was that several protein kinases were transcriptionally induced by NO treatment. Protein kinases are organized into signalling cascades that form the backbone of the signalling network within and between cells. In the present study, NO up-regulated the transcript level of 24 protein kinases of different classes. These included mitogen-activated protein (MAP) kinase kinase 5, which is part of the MAP kinase module consisting of MAP kinase kinase kinase 1, MAP kinase kinase 4/5 and MAP kinase 3/6. The induction of this MAP kinase in Arabidopsis conferred resistance to bacterial and fungal pathogens (Asai et al., 2002). NO induced a similar MAP kinase in tobacco, which also responds to salicylic acid (Kumar and Klessig, 2000). The other interesting class of kinases induced by NO belonged to the plant receptor kinases (PRKs) or receptor-like kinases (RLKs). There have been 417 PRKs identified in Arabidopsis (Shiu and Bleecker, 2001) and, for most of these, the extracellular ligands and intracellular downstream signalling molecules remain to be discovered. We observed transcriptional activation of several receptor kinases by NO, and the increase in transcript level was as high as 11.5-fold. Functional analysis is required to determine whether any of these PRKs actually function as extracellular ligands or downstream signalling molecules of NO.

In conclusion, the current study clearly shows that NO modulates the expression of a substantial number of genes at the transcriptional level. Several NO-induced and NO-repressed genes in Arabidopsis have been reported for the first time. The identification of a large number of NO-induced transcription factors related to disease resistance, drought tolerance and plant development is one of the significant outcomes of this study. NO may also have a profound effect on the plant signalling network, as revealed by the transcriptional activation of several kinases, including a MAP kinase. Transcriptional regulation is only a part of gene regulation, and the phenotypes, if any, associated with the observed transcript changes remain to be elucidated.

Experimental procedures

Plant material and RNA isolation

Arabidopsis thaliana ecotype Columbia was grown in Arabidopsis growing medium PM05 (Lehle Seeds, TX, USA) under 16 h light (100–150 µmol/m2/s) and 8 h dark cycles at a constant temperature of 22 °C and 85% relative humidity. For NO treatment, plants just after first bolting (28 days after planting) were irrigated with 0.1 mm or 1.0 mm SNP (Sigma, St. Louis, MO, USA) in water, and leaf tissues were harvested after 3 h. Plants were treated during the light period, and three replications with 12 plants in each were included. Plants irrigated with distilled water and those treated with 0.1 mm c-PTIO for 3 h after treatment with 0.1 mm SNP were used as controls. Total RNA from pooled leaf tissue was isolated and purified using Trizol reagent (Invitrogen, Carlsbad, CA, USA) and an RNeasy Mini Kit (Qiagen, Valencia, CA, USA).

Microarray hybridization

Total RNA samples were processed as recommended by the manufacturer (Affymetrix, Santa Clara, CA, USA). In brief, 10 µg of total RNA was reverse transcribed using SuperScript II RT (Invitrogen, Carlsbad, CA, USA) and T7-(dT)24 primer. All the first strand cDNA was used for double-strand cDNA synthesis. Double-strand cDNA was purified by phenol–chloroform extraction and ethanol precipitation. One-half of the purified double-strand cDNA was used to generate biotin-labelled cRNA from an in vitro transcription reaction (IVT) using the Bio-Array High-Yield RNA Transcript Labeling Kit (Enzo Diagnostics, Farmingdale, NY, USA). The reaction product of IVT was purified using an RNeasy Mini Kit (Qiagen, Valencia, CA, USA) and quantified with a Biophotometer (Brinkmann, Westbury, NY, USA). Fifteen micrograms of fragmented cRNA was used to make 300 µL of hybridization cocktail, and 225 µL of the cocktail was used for target hybridization. The biotin-labelled targets were hybridized to GeneChip Arabidopsis ATH1 Genome Array (Affymetrix, Inc, Santa Clara, CA, USA) for 16 h at 45 °C with rotation at 60 r.p.m. in an Affymetrix GeneChip Hybridization Oven 640. Washing and staining were carried out in an Affymetrix Fluidics Station 400, following the protocol for the standard format of antibody amplification for eukaryotic targets (EuKGEWSv4). The processed arrays were scanned in an Agilent GeneArray Scanner (Agilent, Palo Alto, CA, USA).

Data analysis

The hybridization signals were quantified and analysed using MicroArray Suite 5.0 (Affymetrix, Inc). From the original hybridization data (‘Supplementary material’), the data for all the probe sets that had increase ‘I’ and decrease ‘D’ calls were extracted into separate Excel spreadsheets. Only those transcripts that had detection call ‘P’, signal value ≥ 25 and change P value < 0.05 in all the three replicates of each treatment were included in the identification of NO-induced genes. In the next step, the transcripts with a minimum twofold increase in signal over the control in at least two of the three biological replicates, and a coefficient of variation (CV) < 30%, were identified as NO-induced genes. Only those transcripts that had detection call ‘P’ in the control (however, they may be ‘present’ or ‘absent’ after treatment) and change P value > 0.99 were included in the identification of NO-repressed genes. The transcripts showing a minimum twofold decrease in signal over the control in at least two of the three biological replicates, and CV < 30%, were identified as NO-repressed genes. The signal log ratio is the change in the expression level of a transcript between the control and experimental samples, expressed as the log2 ratio. The fold-change was calculated as 2(signal log ratio) when the signal log ratio was ≥ 0 and (−1) × 2−(signal log ratio) when the signal log ratio was < 0. Hierarchical clustering was performed using epclust (http://ep.ebi.ac.uk/EP/EPCLUST), with the correlation measure-based distance and average linkage clustering method.

Reverse transcriptase polymerase chain reaction

Gene-specific primers were synthesized for 16 selected probe sets, and RT-PCR was carried out to verify the microarray results. Two micrograms of purified total RNA was reverse transcribed using Superscript II RT and poly (T) primer at 42 °C for 1 h. The reaction product was diluted to a concentration of 4 ng/µL, and 20 ng per reaction was used for PCR. The reaction mix (25 µL) contained 0.4 µm of each primer, 200 µm dNTP, 1 × reaction buffer and 1 unit Taq DNA polymerase (Qiagen, Valencia, CA, USA). PCR included 28 cycles of 94 °C for 30 s, 58 °C for 30 s and 72 °C for 45 s in a PTC 200 thermal cycler (MJ Research, Reno, NV, USA).

Real-time polymerase chain reaction

Real-time PCR was performed using the SYBR Core Reagent Kit (Applied Biosystems, Foster City, CA, USA) in a real-time PCR machine (iCycler, Bio-Rad, CA, USA). First strand cDNA for real-time PCR was prepared from DNase-treated total RNA as described for RT-PCR. The reaction mix (25 µL) contained cDNA from 20 ng total RNA, 0.2 µm of each primer, 1 × SYBR reaction buffer, 3 mm MgCl2, 200 µm dATP, dCTP and dGTP, 400 µm dUTP, 1 unit AmpliTaq Gold DNA polymerase and 0.25 unit Amperase. The PCR included denaturation at 95 °C for 5 min, and 40 cycles of 94 °C for 30 s, 58 °C for 30 s and 72 °C for 45 s. A melting curve was run after the PCR cycles, followed by a cooling step. Fold-change in the expression of RNA was estimated using threshold cycles.

Supplementary material

The original hybridization data files and the complete lists of genes up-regulated and down-regulated in response to NO treatments are available as supplementary tables from http://www.blackwellpublishing.com/products/journals/PBI/PBI085/PBI085sm.htm. Table S1 Original hybridization data. Table S2 List of genes up-regulated in 0.1 mm SNP treatment. Table S3 List of genes up-regulated in 1.0 mm SNP treatment. Table S4 List of genes up-regulated in 0.1 mm and 1.0 mm SNP treatment. Table S5 List of genes down-regulated in 1.0 mm SNP treatment. Table S6 List of genes down-regulated in 0.1 mm SNP treatment. Table S7 List of genes down-regulated in 0.1 mm and 1.0 mm SNP treatment.

Acknowledgements

We thank Dr Rangasamy Elumalai (University of Arizona), Drs Scott Leisner and Deborah Neher (University of Toledo) and Drs Raju Datla, Pierre Fobert and Jeff Pylatuik (National Research Council, Canada) for valuable discussions. This work was supported by USDA-ARS cooperative agreement grant #3607-2100-008-01S. We would like to dedicate this paper to Congresswoman Marcy Kaptur.

Ancillary