SEARCH

SEARCH BY CITATION

Keywords:

  • Arabidopsis thaliana;
  • cation transport;
  • cGMP;
  • signal transduction;
  • transcriptomics;
  • microarray

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. References
  8. Supporting Information

The occurrence of the second messenger 3′,5′-cyclic guanyl monophosphate (cGMP) has been shown in a number of plant species, including barley, tobacco and Arabidopsis. Physiological processes where cGMP signalling has been observed, or has been inferred, to play a role include chloroplast development, α-amylase production in aleurone tissue, NO-dependent expression of defence-related genes and salt/osmotic stress. In most cases, it is unknown how cGMP exerts its effects and what the downstream targets are. A transcriptomics approach was therefore used to identify putative targets for cGMP signalling. Root exposure to 10 μm membrane permeable cGMP induced changes in abundance for many transcripts involved in metabolism, gene transcription, signalling and defence. In particular, monovalent cation transporters such as non-selective ion channels and cation:proton antiporters were found to be affected in cGMP exposed roots. In addition, exposure to cGMP was found to modulate influx and efflux of the monovalent cations Na+ and K+.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. References
  8. Supporting Information

The basic mechanisms employed by all organisms to transduce information encoded in external or internal stimuli into a response show large degrees of similarity. In many cases, intracellular soluble signalling molecules both amplify and transduce signals from extracellular stimuli. The second messenger, 3′,5′-cyclic guanyl monophosphate (cGMP), is well established as signalling intermediary in bacteria, fungi, animals and algae, and in the last decade a growing number of reports has described both the occurrence and function of cGMP in higher plants. Using a range of different techniques, GMP has been detected in a number of plant species, including barley, tobacco and Arabidopsis (Donaldson et al., 2004; Durner et al., 1998; Penson et al., 1996) and in different tissues, Although the data vary considerably, amounts are typically 1–10 pmol per gram Fresh Weight (gFW) which roughly translates into cytoplasmic levels of cGMP in the range of 10–100 nm. Furthermore, during perception of stimuli, these cellular cGMP concentrations have been shown to increase 5- to 10-fold (Donaldson et al., 2004; Durner et al., 1998).

Additional evidence for a role of cGMP in cellular physiology stems from the recent discoveries that all components of cGMP metabolism have been identified in plants: cyclic nucleotide phosphodiesterase activity to convert cGMP to GMP has been demonstrated in plants using biochemical assays (Newton and Smith, 2004), whereas an Arabidopsis protein (AtGC1) with guanylyl cyclase activity was recently cloned (Ludidi and Gehring, 2003) showing plants have the capacity to synthesise cGMP. Several genes with high homology to cyclic nucleotide cyclases and to cyclic nucleotide phosphodiesterases have been found in the genomes of Arabidopsis and other plant species.

Physiological processes where cGMP signalling has been shown, or has been inferred, to play a role are legion. cGMP has been demonstrated to be involved in the phytochrome-mediated induction of the gene encoding chalcone synthase (CHS), a key enzyme in the biosynthesis pathway for anthocyanins and chloroplast development (Bowler et al., 1994). Both cGMP and Ca2+ based signalling is required for full chloroplast development with the two pathways interacting in a negative manner. The involvement of a G protein downstream of phytochrome A and upstream of the cGMP signal was also suggested (Bowler et al., 1994; Wu et al., 1996). Exposure of barley aleurone layers to giberrellic acid (GA) generates a transient increase in cGMP levels and induction of gene transcription, which leads ultimately to α-amylase production (Penson et al., 1996). In tobacco, nitric oxide (NO) exposure caused a transient increase in cGMP levels (Durner et al., 1998). cGMP functions as a second messenger for NO signalling in animals by inducing the expression of defence-related genes. In plants too, cGMP has been implicated in NO-dependent induction of gene expression, for example of defence genes such as PAL and PR-1. (Durner et al., 1998).

The use of membrane permeable cyclic nucleotide derivatives also suggests the participation of cGMP in the modulation of ionic fluxes. In protoplasts, exposure to membrane permeable cGMP leads to a rapid influx of Ca2+, presumably through Ca2+ channels (Volotovski et al., 1998). During stomatal functioning, cGMP may directly affect K+ fluxes that generate turgor changes (Pharmawati et al., 2001) although no evidence for a direct effect on channels or transporters is available in this regard. In contrast, a direct effect of cGMP on non-selective ion channels has been shown in Arabidopsis root plasma membranes (Maathuis and Sanders, 2001), which could explain the inhibitory effect of externally applied cGMP on net Na+ uptake and Na+ accumulation (Essah et al., 2003; Maathuis and Sanders, 2001; Rubio et al., 2003). More recent work showed that a rapid increase in cellular cGMP occurs after the onset of salt and osmotic stress (Donaldson et al., 2004).

In all these examples, external and internal stimuli have either been shown or are believed to cause a rise in cytoplasmic cGMP, which subsequently modulates the activity of downstream components. The latter may therefore form targets for cGMP based signalling, either through direct binding of the nucleotide or indirectly through other signalling components. However, specific cGMP targets in plants are largely unknown. A classic set of downstream effectors of cyclic nucleotide signalling in both animals and fungi are cyclic Adenine Mono Phosphate (cAMP)-dependent and cGMP-dependent protein kinases (PKA and PKG, respectively, Finn et al., 1996). Although there is some biochemical evidence supporting the existence of plant cyclic nucleotide-responsive kinases (Newton and Smith, 2004), little molecular evidence of bona fide cyclic nucleotide-dependent kinases in plants is available. However, the Arabidopsis genome does contain sequences that encode gene products with both a cyclic nucleotide binding domain and a protein kinase domain indicating that PKA- and PKG-type kinases may function in plants.

Interestingly, most of the known putative targets, i.e. gene products containing a cyclic nucleotide monophosphate (cNMP) binding domain, consist of ion transporters. Prevalent amongst these are (monovalent) cation channels from the families of cyclic nucleotide gated channels (CNGCs) and potassium selective Shaker-type channels, and sodium-proton antiport mechanisms. For representatives of both the CNGC and Shaker families, it has been shown that their activity is modulated by cNMPs (Balague et al., 2003; Hoshi, 1995; Leng et al., 1999) although it is not in all cases clear whether this is a direct effect.

Thus, with clear evidence of the presence and function of cGMP in plant tissues, it is important to identify components of cGMP-dependent signalling events. In particular, very little is known regarding cGMP-dependent regulation of gene expression (Trewavas et al., 2002). An in silico and transcriptomics approach was therefore employed to discover which gene products may be affected by changes in cGMP. The obtained data suggest very few gene products are likely to be directly regulated by interaction with cGMP, but a large amount of transcripts, involved in a variety of cellular processes, is affected in abundance when plants are exposed to membrane permeable cGMP.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. References
  8. Supporting Information

Many monovalent cation transporters contain putative cNMP binding domains

Binding domains for cNMP typically consist of around 120 residues and are highly structure conserved (Shabb and Corbin, 1992). X-ray crystallography showed that the cNMP binding domain is composed of two subdomains: four pairs of antiparallel β strands and two short α helices (A and B helix) form the roll subdomain, and a single, long α-helix forms the C-helix subdomain (Weber et al., 1987). The cyclic nucleotide binds within a pocket between the C helix and the β roll. This well-conserved domain can be used to query protein databases to identify putative cGMP targets in higher plants. Across green plants, around 90 hits [e.g. European Molecular Biology Laboratory-European Bioinformatics Institute (EMBL-EBI) at http://www.ebi.ac.uk/interpro] are obtained (Table 1 and supplementary data ‘CNBdomainSupplement.xls’) almost exclusively consisting of ion transporters. The approximately 30 hits for Arabidopsis proteins include 20 CNGC isoforms and nine members of the Shaker-type inward and outward rectifying K+ channel (KAT/AKT/SKOR) families. Both these ion channel families are represented across many green plant species. In addition, (putative) Na+/H+ antiporters (NHX-type) were identified in several plant species. Members of the acyl Co-A thioesterase family, involved in peroxisomal lipid metabolism, of both Arabidopsis and rice also contain cNMP binding domains. Interestingly, four putative proteins were identified that contain both a cyclic nucleotide binding domain and a ser/thre protein kinase domain. In Arabidopsis, this putative cyclic nucleotide dependent kinase, At4g00510, shows a high degree of homology (P: 2e−6) to the bovine, human and other type I cGMP-dependent protein kinases with around 55% similarity. The protein has four matching expressed sequence tags (ESTs) and its transcript has been detected on microarrays (MIPS: http://mips.gsf.de) providing the exciting prospect that this may be a genuine plant PKG.

Table 1.  Occurrence of putative cyclic nucleotide monophosphate (cNMP) binding domains in green plant species. The EBI InterPro (http://www.ebi.ac.uk) database was used to survey plant genomes for the presence of IPR000595, a consensus domain for cyclic nucleotide binding. KAT/AKT: Shaker-type, K+ selective ion channels; CNGC: cyclic nucleotide gated ion channels; PKA/PKG: putative cAMP- and cGMP-dependent protein kinase
SpeciesKAT/AKTCNGCNa/H antiportPKA/PKGAcylCoA
Arabidopsis920111
Chlamydomonas   2 
Cymodocea nodosa  1  
Daucus carota2    
Egeria densa1    
Hordeum vulgare16   
Lycopersicon esculentum1    
Mesembryanthemum crystallinum3    
Nicotiana tabacum13   
Oryza sativa411111
Phaseolus vulgaris2    
Physcomitrella patens  1  
Populus tremula3    
Samanea saman3    
Solanum tuberosum4    
Triticum aestivum1    
Vicia faba1    
Vitis vinifera3    
Zea mays5    
All green plants4242441

cGMP affects transcription of many genes

The number of gene products with cyclic nucleotide binding domains in plants is comparable to many other eukaryotic species (e.g. 67 in humans, 57 in mouse), but far less than found in bacteria (around 900) and metazoa (around 400). Nevertheless, the turnover and activity of many proteins may be affected by cNMP signalling: for example, through the action of cNMP-dependent kinases or via transcriptional modulation. To investigate whether a rise in cGMP alters transcript composition in Arabidopsis roots, mature plants were exposed to 10 μm membrane permeable Bromide (Br)-cGMP for 30 min. RNA from these and control plants was subsequently harvested after 2 and 5 h to survey alterations in transcript levels by using a microarray approach.

After data analysis and application of the significance criteria (see methods section), approximately 16 000 spots out of 27 000 were found to produce a significant level in the analysed root tissue representing around 15 000 unique transcripts. Of this population, approximately 1000 transcripts were affected in their levels by cGMP (Supplementary Table 4 in cGMPSupplementData), i.e. they showed a two-fold or more change in transcript level in either the 2 h, the 5 h, or both time points. To verify the obtained microarray data, changes in abundance for 11 transcripts, spanning ratios of 0.2 to 14, were determined through RT-PCR . (Table 2), using ubiquitin transcripts as a normalisation signal. Although absolute transcript level ratios obtained for specific genes using RT-PCR are somewhat different, there is an overall good agreement between the two methods.

Table 2.  Comparison of fold changes in transcript abundance (standard deviations are in brackets) as observed with either a microarray hybridization approach or the use of RT-PCR. RT-PCR was carried out using two sets of ubiquitin primers to standardize signals, which were semi-quantified using densitometry
GeneAnnotationFold change
MicroarrayRT-PCR
2 h5 h2 h5 h
At5g46360KCO31.0 (0.15)1.4 (0.45)0.8 (0.32)1.2 (0.61)
At3g17690CNGC190.9 (0.41) 1.2 (0.42) 
At1g62690Unknown1.3 (0.42) 2.1 (0.92) 
At5g05930Putative cyclase1.5 (0.48) 1.2 (0.81) 
At1g69860PTR141.0 (0.02)1.4 (006)0.8 (0.39)1.2 (0.56)
At2g31910CHX222.0 (1.0)14 (8.8)2.8 (1.2)6.8 (3.9)
At4g38920VHA-c30.2 (0.14) 0.1 (0.06) 
At5g24270SOS32.0 (0.25)1.5 (0.66)3.4 (1.6) 
At1g09640EF 1B 0.5 (0.11) 1.3 (0.99)
At4g03560TPC11.0 (0.33) 0.6 (0.45) 
At1g33440NTL14.7 (2.7) 2.9 (1.8) 

To establish whether specific functional transcript categories were over- or under-represented in the various collections of transcripts, the gene ontology functional categorisation tool available at The Arabidopsis Information Resource (TAIR; (http://arabidopsis.org/tools/bulk/go/index.jsp) was used. Figure 1(a) shows the functional categorisation of all transcripts that are represented on the microarray (approximately 25 000) across 15 biological functions. In Figure 1(b), the same functional classes were applied to the approximately 15 000 transcripts found to be expressed in root tissue. Comparing Figures 1(a) and (b) shows that the relative proportions of functional classes are very similar, suggesting that there is no functional bias in the total complement of root expressed transcripts. In Figure 1(c), the functional categorisation for the subpopulation of 1000 significantly regulated transcripts is depicted. Two classes appear to be differentially represented: transport occurs with a percentage of 8.2% in the regulated subset and 5.0% in the background population whereas in contrast, the class ‘DNA or RNA metabolism’ forms 0.5% of the group of regulated transcripts and 1.4% in the background. These disproportionalities become even more pronounced if a larger fold change criterion is applied to the set of regulated transcripts, i.e. 9.2 and 0.2% for ‘transport’ and ‘DNA or RNA metabolism’ respectively when a fold change of >3 is applied.

image

Figure 1. cGMP affects gene transcription. Functional classification as provided by TAIR (http://arabidopsis.org/tools/bulk/index.jsp) for all genes (a) represented on the microarrays used in this study, for all genes that gave a significant fluorescent signal in root tissue (b) and for genes that showed a 2-fold or larger change in transcript abundance after cyclic guanyl monophosphate (cGMP) treatment (c).

Download figure to PowerPoint

Of the 1000 regulated transcripts, around 570 could be functionally annotated into specific functional classes (Supplementary Tables 3 and 5 in cGMPSupplementData). The largest group of affected transcripts (around 150) is formed by those encoding a myriad of cellular functions in general processes such as C and N metabolism and lipid biochemistry. The second largest group (around 125) consists of the many transcription factors whose transcript abundance was found to be altered after cGMP treatment. Most major classes of transcription factors are represented in this list but some classes, particularly MCTH, Agamous, Deficiens, SRF (MADS)-box and U-box transcription factors appear in disproportionately low numbers suggesting specific types of transcription factor may be more prone than others to regulation by cGMP.

Many genes involved in cellular signalling are affected by cGMP including 14-3-3 proteins, calmodulins and several G proteins. A putative cyclic nucleotide phosphodiesterase (At4g18940) was also found to be upregulated, possibly in response to the constitutively high levels of cGMP imposed by the treatment.

A large number of genes involved in the perception of biotic and abiotic stress was identified. Around 20 ‘disease resistance’ type transcripts, or approximately 9% of all ‘disease resistance’ annotated genes in the Arabidopsis genome, are present in the pathogen/disease class. The role of cGMP in pathogen response signalling has been reported as part of NO-induced signalling in tobacco leaves leading to transcription of pathogenesis related genes (Durner et al., 1998) and during NO-induced cell death in Arabidopsis (Clarke et al., 2000).

Monovalent cation transporter transcripts are over-represented amongst cGMP regulated transcripts

The previous analysis suggests that membrane transporters in particular form downstream targets of cGMP signalling. Table 4 lists all membrane transporters that were identified within the group of significantly regulated transcripts. The three major classes of transporter, i.e. primary pumps, carriers and ion channels, are not evenly distributed amongst this group. A disproportionally large number (around 20) represents transporters involved in monovalent cation transport and includes CNGCs, K+ channels and monovalent antiport systems. A large section of the CNGC family (seven out of 20, or 35%) appears in this class and three out of eight (38%) of the NHX isoforms are included. In contrast, only eight ABC-type transporter isoforms out of around 160 (5%) were affected by cGMP at the transcript level. Members of other (sometimes extensive) transporter families such as the plasma membrane H+ ATPases (12 members) occurred only once, and sugar transporters and phosphate transporters with respectively 67 and 16 members were not represented.

Table 4.  Membrane transporter transcripts whose abundance changed two-fold or more after cyclic guanyl monophosphate (cGMP) treatment. Transcripts are categorized according to (putative) function. The third and fourth columns depict average fold-change data for 2- and 5-h treatments, respectively, with standard deviations in brackets. The fifth column indicates published or publicly available data on transcript regulation by salt, osmotic, or K+ depletion treatments. References: 1, Maathuis et al., 2003; 2, Genevestigator website http://www.genevestigator.ethz.ch; 3, Armengaud et al., 2004
Gene codeGene annotation2 h5 hRef.
At5g08680ATP synthase beta chain, mitochondrial2.1 (0.7)  
At5g55290VHA-e1 vacuolar ATPase subunit2.44 (0.66)  
At4g34720VHAc1 vacuolar H+ ATPase0.37 (.18)1.44 (0.35)1
At4g38920VHA-c3 vacuolar H+-transporting ATPase0.24 (0.14)1.30 (0.15) 
At1g75630VHA-c4 vacuolar ATPase0.44 (0.20)0.87 (0.11) 
At1g03905ABC transporter family protein0.08 (0.04)0.96 (0.35) 
At1g70610TAP1 ABC transporter family protein0.34 (0.11)0.88 (0.21)2
At3g16340PDR1 ABC transporter family protein0.46 (0.14)0.91 (0.11) 
At4g15230PDR2 ABC transporter like protein2.38 (0.55)0.91 (0.14) 
At2g36380PDR6 ABC transporter family protein 0.37 (0.16)2
At5g13580WBC6 ABC transporter family protein1.5 (0.82)0.25 (0.16) 
At5g06530WBC23 ABC transporter family protein2.2 (1.26)  
At1g27770ACA1 calcium-transporting ATPase0.49 (0.32)4.5 (2.62)2,3
At3g63380ACA12 calcium-transporting ATPase0.48 (0.07)0.99 (0.1)1,2
At5g51050Calcium channel (putative)2.1 (0.74) 2
At1g55720CAX6 H+/Ca2+antiporter0.45 (0.13)  
At5g17860CAX7 calcium-H+ exchanger0.86 (0.12)2.1 (0.77)2
At5g17850CAX8 calcium-H+ exchanger0.40 (0.21)0.87 (0.14)2
At1g54115CAX10 calcium-H+ exchanger0.49 (0.13)0.89 (0.32) 
At2g28180CHX8 cation/H+ antiporter2.0 (1.2)  
At5g22910CHX9 cation/H+ antiporter2.54 (1.5) 1,2
At2g31910CHX22 cation/H+ antiporter2.5 (1.1)14.2 (8.4)1
At5g27150NHX1 Na+/H+ exchanger2.1 (0.49)0.97 (0.19)1
At2g01980NHX7 Na+/H+ antiporter (SOS1)1.45 (0.38)2.2 (0.74)1
At1g14660NHX8 Na+ H+ antiporter2.0 (0.52)1.4 (0.34)2
At5g44790RAN1 ATP-dependent copper transporter0.49 (0.23)  
At3g46900COPT3 copper transport2.7 (1.6)  
At2g32270ZIP3 Fe (II)transporter1.5 (0.66)0.41 (0.33)2
At4g37270HMA1 Cu2+-transporting ATPase2.1 (0.63)2.1 (0.53)2
At4g30110HMA2 cadmium-transporting ATPase0.48 (0.19) 2
At4g30120HMA3 cadmium-transporting ATPase4.0 (2.5) 2
At2g46800MTP1 zinc transporter0.43 (0.12)0.85 (0.21) 
At4g18790nRAMP5 metal transporter0.39 (0.12)0.87 (0.14) 
At5g15410CNGC2 cyclic nucleotide-regulated ion channel0.48 (0.18)1.2 (0.43)2
At2g23980CNGC6 cyclic nucleotide-regulated ion channel2.6 (1.38)  
At2g46450CNGC12 cyclic nucleotide-regulated ion channel2.1 (0.32) 2
At2g28260CNGC15 cyclic nucleotide-regulated ion channel0.83 (0.05)2.7 (0.93)2
At3g48010CNGC16 cyclic nucleotide-regulated ion channel 2.0 (0.06) 
At4g30360CNGC17 cyclic nucleotide-regulated ion channel0.72 (0.41)2.3 (0.87)1
At3g17700CNGC20 cyclic nucleotide-regulated ion channel0.92 (0.14)2.9 (1.79)1,2
At5g11210GLR2.5 glutamate receptor family2.3 (1.4) 2
At1g05200GLR3.4 glutamate receptor family2.9 (1.9) 2
At1g04690K+ channel, beta subunit0.50 (0.23)0.93 (0.07)2
At1g01790KEA1 K efflux antiporter0.40 (0.21)1.2 (0.83)2
At4g18160KCO6 outward rectifying potassium channel2.2 (1.3)1.4 (0.39) 
At4g18290KAT2 potassium channel2.2 (0.43) 2
At2g25600AKT6 potassium channel0.47 (0.03)0.83 (0.14) 
At4g19960KUP/KT9 potassium transporter0.44 (0.17)1.0 (0.18) 
At4g13510AMT1 ammonium transport0.27 (0.16)  
At3g62270Anion exchange family protein2.1 (0.50)1.1 (0.41) 
At1g69850NRT1.2 nitrate transporter (NTL1)0.42 (0.27)0.95 (0.45)1
At1g08090NRT2.1 high-affinity nitrate transporter0.94 (0.19)0.47 (0.27)1,2
At1g08100NRT2.2 high-affinity nitrate transporter (ACH2)0.91 (0.09)0.34 (0.07) 
At5g09220AAP2 amino acid transport0.49 (0.29)0.86 (0.31) 
At5g49630AAP6 amino acid permease 0.44 (0.3)2
At3g10600CAT3 amino acid transporter2.1 (0.64) 2
At4g38250Amino acid transport protein0.47 (0.01)1.0 (0.37)2
At3g25410Bile acid:sodium symporter family0.40 (0.18)1.5 (0.93) 
At5g01180PTR6 oligopeptide transporter2.1 (0.81) 2
At1g33440PTR30 peptide transporter4.7 (2.7)  
At3g53960PTR34 peptide transporter1.2 (0.26)5.2 (2.6) 
At3g45650PTR51 proton-dependent oligopeptide transport10.31 (0.10)  
At2g16850PIP2.8 plasmamembrane aquaporin0.46 (0.20) 1
At3g53420PIP2.1 plasmamembrane aquaporin0.49 (0.23) 1
At3g26520TIP1.2 gamma tonoplast aquaporin0.38 (0.24)  
At1g66760MATE efflux family protein, putative1.1 (0.32)2.1 (0.13)2
At3g01340Transport protein SEC13, putative0.10 (0.02) 2
At2g05760Membrane transporter2.0 (0.47)  
At2g13100Membrane transporter0.50 (0.24)  

cGMP modulates both Na+ and K+ fluxes

The transcriptomics analyses suggest that ion transport and particularly monovalent cation transport respond to changes in cGMP. Because many of the affected transcripts encode proteins that are involved, or believed to be involved, with Na+ and/or K+ transport, it was investigated whether the cGMP regulated transporters listed in Table 4 are also transcriptionally regulated by salinity and/or K+ stress. When the same cut-off criterion (>2-fold) was applied, 38 out of the 71 transcripts in Table 4 were found to have been described as affected at the transcriptional level by salt stress, osmotic stress or K+ deficiency (see Table 4 for references). The effect of cGMP on K+ and Na+ movement was therefore studied in more detail by determining the uptake, efflux, and root and shoot tissue concentrations of these ions in the presence and absence of cGMP.

When measured over 5 or 24 h in the presence of 100 mm NaCl, cGMP significantly (P < 0.05, unpaired t-test) reduced net Na+ uptake in mature Arabidopsis plants (Figure 2a) as was previously established for seedlings (Maathuis and Sanders, 2001). However, the effect of cGMP is less significant after a 24-h period and virtually absent after 3 days (Figure 2a). When plants are pre-loaded with Na+ (80 mm) over a 3-day period, net Na+ efflux from pre-loaded plants in the presence of membrane permeable cGMP was significantly higher than for the control plants (Figure 2c). This suggests that in addition to modulating Na+ uptake pathways, cGMP may also act on Na+ efflux systems such as Na:H antiporters.

image

Figure 2. Modulation of Na+ and K+ fluxes by cyclic guanyl monophosphate (cGMP). Membrane permeable cGMP was added to a final concentration of 10 μm to uptake and efflux buffers. (a) Net Na+ influx measured in the absence and presence of cGMP over 5, 24 and 72 h. (b) Net K+ influx measured in the absence and presence of cGMP over 5 and 24 h for plants grown on standard medium or plants starved (st) for K+ over a 3-day period. (c) Net Efflux measured over 5 h for Na+ from plants grown in the presence of 80 mm NaCl for 3 days, for K+ from plants grown on standard medium transferred to medium without K+ and for K+ from plants grown on standard medium transferred to medium without K+ plus 80 mm NaCl. Fluxes were calculated from total accumulated Na+ divided by time (Na+ uptake), depletion of K+ in the uptake buffer (K+ influx) or appearance of K+ or Na+ in the efflux buffer (K+ and Na+ efflux). Error bars denote standard deviations, stars denote significance at the P < 0.05 level, using an unpaired t-test.

Download figure to PowerPoint

Similar experiments were carried out to test whether cGMP modulates transport of K+. Plants grown on standard growth medium did not show significantly different K+ influx, irrespective of the presence of cGMP (Figure 2b). However, when plants were K+ starved for 3 days on medium that lacked K+, the presence of cGMP significantly augmented K+ uptake (P < 0.05, unpaired t-test). cGMP did not affect K+ efflux from plants that were grown on standard medium and transferred to zero K+ medium (Figure 2c). Because the presence of large quantities of Na+ during salinity stress typically leads to exchange of cellular K+ for Na+, K+ efflux in the presence of 80 mm NaCl was also measured but not found to be significantly different in the presence of cGMP.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. References
  8. Supporting Information

With the biochemical machinery in place, a confirmation of its presence in plant cells and an increasing number of potential functions, it becomes imperative to determine putative targets of cGMP mediated signalling. A simple query of plant genomic sequences shows that only a relatively small number of gene products is likely to be modulated through direct binding of cGMP. Only around 30 sequences in the Arabidopsis genome contain putative cNMP binding domains. The majority of these genes consists of membrane cation transporters, providing an initial clue that cGMP mediated signalling may modulate the uptake and translocation of nutrients such as K+ and Ca2+ and of harmful ions such as Na+ through direct interaction at the protein level. To test whether these and other gene products may also be regulated at the transcript level, we carried out a transcriptomics study using cGMP exposed root tissue.

Transcripts of many functional categories are affected by cGMP

The transcriptomics study showed that around 650 functionally annotated transcripts are affected by cGMP. It has been suggested that, in analogy to mammalian cells, plants also contain proteins that, after phosphorylation, bind to promoters containing cAMP and cGMP responsive elements (Katagiri et al., 1989; Newton and Smith, 2004). For example, in mammalian cells, cAMP responsive element binding (CREB) type proteins that are phosphorylated by PKA were found to affect transcription of over 900 genes across many functional classes (Bailey et al., 2005). The mammalian cAMP responsive element (CRE) core motif of TGACGTCA was not significantly over-represented in the cGMP regulated transcripts. However, two significantly over-represented motifs were identified in upstream regions of cGMP regulated transcripts: CGTGGGGA|TCGCCACG and GTAGTTTA|TAAACTAC with respective Z scores of 7.4 and 5.3. Thus, it is feasible that these and other motifs are instrumental in binding cGMP activated transcription factors that modulate transcription of many genes.

Amongst regulated genes, there are many that are involved in signalling such as G proteins and calmodulins. In mammalian cells, many cGMP-dependent signalling pathways incorporate the action of G proteins, e.g. during photo perception or NO signalling. In plants, there is some indication that the cAMP concentration in pollen tubes is modulated by G proteins (Tsuruhara and Tezuka, 2001) and the suggestion that G proteins couple phytochrome light perception to cGMP signalling has also been made (Bowler et al., 1994). These scenarios imply G-protein activity, which affects cNMP levels by interacting with cyclase and/or phosphodiesterase activity. However, results from Table 3 and supplementary data indicate that transcription and therefore potentially activity of a number of G proteins may depend on upstream cNMP signalling.

Table 3.  Summary of transcripts significantly altered in abundance after cyclic guanyl monophosphate (cGMP) treatment. The total number of transcripts for each class is given in brackets with some of the major subclasses listed. A full list is available in the supplementary data file in Table 4
General metabolism (148)
 Sugar metabolism, nitrogen metabolism, peptide synthesis, lipid metabolism
Transcription factors (123)
 bHLH, bZIP, F-box, GATA, Myb, RRM, zinc finger, WRKY
Protein phosphorylation (78)
 CDPK, MAPK, Ser-Thre, receptor kinases, WAK, PP2C phosphatases
Transporters (71)
 P-pumps, ABC pumps, cation exchange, ion channels, aquaporins
Protein synthesis/targeting (35)
 Proteases, elongation, translation, syntaxins
Disease/pathogen response (33)
 Disease resistance genes, lectins
Cell wall (29)
 Cellulose synthases, pectinesterases, expansins
Signalling (26)
 14-3-3, calmodulins, G proteins
Hormone function (19)
 Auxin response, ethylene response, IAA induction
Stress perception (15)
 Dehydration response, salt induced, LEA

Interestingly, the class of signalling components also includes the calcium sensor homolog Salt Oven Sensitive 3 (SOS3), which is implicated in regulating both salinity stress and high affinity K+ uptake (Qi and Spalding, 2004; Zhu, 2002). SOS3 is believed to act through SOS2, a protein kinase, which in turn regulates the activity of the plasma membrane localised antiporter SOS1 whose transcription is increased during salinity stress (Zhu, 2002). Initiation of the SOS pathway is believed to depend on an elevation of cytoplasmic Ca2+ (Zhu, 2002). Recent work shows evidence that a salt-induced rise in cytoplasmic cGMP is likely to form an upstream component of the salt-induced Ca2+ signal (Donaldson et al., 2004). In addition, several studies have shown cNMP mediated Ca2+ influx into plant cells (e.g. Volotovski et al., 1998). Thus, cGMP may be involved in upregulation of this pathway through the generation of a Ca2+ signal.

cGMP is likely to play an important role in cation homeostasis

When plants were exposed to cGMP in standard conditions, the category of membrane transporters was the only class of transcripts found to be over-represented amongst significantly regulated transcripts. Nevertheless, within this group, marked variation was observed. Very few transcripts encoding primary transporters were present. Relatively few ABC transporters and no sugar or phosphate transporters are included. Indeed, very few anion transporters appear in this list. In contrast, two groups are heavily represented: first, a total of seven Ca2+ transporters in the form of P-type pumps (ACA/ECA family), exchangers (CAX family) and a putative Ca2+ channel was identified representing around 20% of all annotated Ca2+ transporters. The second group contains 21 monovalent cation transporters in the form of both ion channels (AKT/KAT/KCO/CNGC and GLR families) and H+-coupled exchangers (CHX and NHX families), equivalent to around 18% of all annotated monovalent cation transporters. This bias towards Ca2+ and monovalent cation transporters could suggest that the homeostasis of Ca2+ and monovalent cations forms a primary target for cNMP signal transduction pathways.

There are several additional observations that add weight to the notion that cGMP is involved in monovalent cation homeostasis. Around 50% (Table 4) of the cGMP regulated transporters have been previously reported to be also transcriptionally regulated in response to abiotic stresses such as salinity, drought and K+ deprivation. Furthermore, when cationic stress conditions such as salinity or K+ deficiency are applied, exposure to cGMP affects the magnitude of Na+ and K+ fluxes (Figure 2, Essah et al., 2003; Maathuis and Sanders, 2001; Rubio et al., 2003). In conjunction with an early and rapid cGMP signal at the onset of salt or osmotic stress (Donaldson et al., 2004), all these observations suggest that cGMP may play important roles in the transcriptional and post-transcriptional regulation of monovalent cation fluxes during turgor adjustment.

cGMP may act on specific transporters involved in cation homeostasis

The exact mechanisms by which cGMP may affect cation fluxes are likely to be various. Addition of membrane permeable cGMP has no significant effect on root cell membrane potential (results not shown) and it is thus unlikely that cGMP has a discernible influence on the driving forces that are responsible for cation movement across the membrane. Earlier studies on Arabidopsis seedlings (Essah et al., 2003; Maathuis and Sanders, 2001) and on mature pepper plants (Rubio et al., 2003) have shown that unidirectional Na+ influx is directly affected by cyclic nucleotides. A possible mechanism for the observed reduction in Na+ uptake constitutes the decrease in Na+ permeability through non-selective cation channels that was observed in the presence of cGMP (Maathuis and Sanders, 2001). The molecular identity of these channels remains to be revealed but is unlikely to be a CNGC because characterisation of a number of these (Balague et al., 2003; Leng et al., 1999) suggests that binding of cNMP to CNGCs leads to channel activation rather than deactivation. Nevertheless, CNGCs may participate in Na+ uptake and thus their transcriptional regulation could impact on Na+ influx. The only CNGC isoform that was found to be significantly downregulated was CNGC2 (Table 4). However, this particular isoform was shown not to be permeable to Na+ (Hua et al., 2003). Na+ influx may also occur through K uptake transporter (KUP)-like K+ transporters (Santa-Maria et al., 1997) such as KUP9. Thus, KUP9 downregulation (Table 4) would restrict Na+ entry into the root symplast.

This study shows that other mechanisms may also contribute to cGMP-dependent modulation of Na+ fluxes. In particular, the observed increase in net Na+ efflux in the presence of cGMP may be relevant in this respect. The latter can be due to reduced Na+ influx as discussed above, an increase in efflux, or both. Enhanced Na+ extrusion is likely to be mediated by H+ coupled exchangers. Modulation of exchange activity by cNMPs has been observed for many mammalian exchange systems (e.g. Brett et al., 2002; Moe, 1999) though this usually occurs through cNMP-dependent protein kinases. In Arabidopsis, one antiport mechanism (NHX7 or SOS1) contains a putative cNMP binding domain and may therefore be activated directly when cytoplasmic cGMP levels rise after perception of osmotic/salinity stress. In addition, it is moderately upregulated at the transcriptional level by cGMP treatment. SOS1 has been shown to be upregulated and to confer tolerance during salt stress (Zhu, 2002), and either or both of the cGMP-dependent regulatory mechanisms would contribute to an increased Na+ efflux mediated by this mechanism.

With a value of around 14, the most substantially upregulated antiport mechanism was CHX22. It is only expressed at moderate levels in root tissue (Genevestigator, http://www.genevestigator.ethz.ch) and its membrane location is unknown. Furthermore, no functional data are available for this transporter but it is likely to mediate K+ and/or Na+ exchange (Sze et al., 2004). A putative Na+ transport capacity coupled to such an extensive increased level of CHX22 transcript could greatly enhance extrusion of Na+ into the apoplast and thus participate in the observed increase in net Na+ efflux.

In contrast to Na+, the presence of cGMP led to an increased K+ uptake particularly in K+ starved plants. cGMP-induced transcriptional and post-transcritional activation of CNGCs that are highly expressed in roots such as CNGC6, 17 and 20 (Tables 1 and 4) would provide an additional pathway for K+ uptake. Alternatively, the effects of cGMP observed during K+ uptake may primarily result from interactions of cGMP with K+ specific uptake pathways. cGMP induced an approximately two-fold increase in KAT2 transcript level (Table 4), which encodes a K+ selective ion channel. The latter could contribute to enlarged K+ influx, however, expression of KAT2 in roots is generally low. K+ uptake capacity may also be post-transcriptionally regulated. The main pathway for root low affinity K+ uptake is AKT1 (Very and Sentence, 2002). In common with many Shaker-type K+ selective ion channels, AKT1 contains a C-terminal putative cNMP binding domain (Table 1). Using a patch clamp, the activity of AKT1 has been shown to be modulated by cGMP (Very and Sentence 2002).

Conclusion

This study has shown that a large number of plant genes is potentially regulated by cGMP and thus forms putative cGMP targets. Genes include many transcription factors, signalling components, transporters and pathogenesis related transcripts. Amongst regulated transporter transcripts, monovalent cation transporters involved in the movement of Na+ and K+ were over-represented. This finding, together with the results showing that cGMP affects Na+ and K+ fluxes in planta and the recent evidence that a cGMP signal is evoked during salinity and osmotic stress, suggests that cGMP signalling forms an important constituent of monovalent cation homeostasis.

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. References
  8. Supporting Information

Plant growth and treatments

Arabidopsis thaliana (L) ecotype Columbia (0) seeds were surface sterilised and placed on top of water/agar (0.8% weight in volume, w/v) containing 1.5-ml tubes. After incubation at 4°C for 2 days, seeds were transferred to a growth cabinet and grown hydroponically in a standard medium containing 1.25 mm KNO3, 0.5 mm Ca(NO3)2, 0.5 mm MgSO4, 0.625 mm KH2PO4 and micronutrients as described previously (Maathuis et al., 2003). Growth medium was renewed every week. Plants were exposed to short day conditions [10 h 200 μmol m−1 sec−1 light, 24°C, Relative Humidity (RH) 70–80%] and harvested when they had developed a full rosette but had not yet started flowering. For microarray analysis, plant roots were exposed for 30 min to membrane permeable cGMP (8-Br-cGMP) at a concentration of 10 μm. After 30 min, both treated and control plants were transferred to fresh medium for either 1.5 (2 h time point) or 4.5 (5 h time point) before harvesting of root tissue.

Microarray hybridisation and analysis

Root RNA was isolated using RNeasy Plant Mini columns (Qiagen, London, UK) and pooled from two to four independent sets of 15–20 plants for each experiment. This procedure was repeated four times, i.e. a total of eight (four for each time point) microarrays was hybridised. For each hybridisation, approximately 100 μg of total RNA was primed with Random 15-mer primer (0.5 μμl−1) and reverse transcribed with Superscript II (Invitrogen). Fluorescent labelling was achieved by part replacement of d-cytosine triphosphate (dCTP) in the d-nucleotide triphosphate (dNTP) mix with Cyanine (Cy)3-dCTP and Cy5-dCTP (Amersham, Little Chalfont, UK). Labelled cDNA was cleaned on a QIAquick spin column (Qiagen). Arabidopsis Oligonucleotide Microarrays (http://ag.arizona.edu/microarray), using the Arabidopsis Qiagen-Operon Genome Oligo Set that represents around 26 000 coding sequences, were used for hybridisation. Array cross-linking, hybridisation and post-hybridisation washes were carried out as described by the manufacturer (http://ag.arizona.edu/microarray).

Arrays were scanned using an Axon (Axon Instruments, Braintree, UK) scanner and initial array analysis was carried out with ScanAlyze2 software (http://rana.lbl.gov/EisenSoftware.htm) for flagging of spots after visual inspection and to obtain spot and background fluorescence data. Background subtraction, global normalisation of fluorescence signals and lowess signal correction were performed using snomad software available at http://pevsnerlab.kennedykrieger.org/snomadinput.html. Global mean normalisation was carried out across microarray surfaces and local mean normalisation across element signal intensity. After normalisation and log2 transformation, signal ratio averages and the standard deviations for signal ratios of replica experiments were calculated. Transcripts were included for analysis and annotated as significantly regulated when the following criteria were met: (i) a signal minus background value of at least 100; (ii) a ‘present’ signal in at least three replicas; (iii) an average of signal ratios for treated and control transcripts of more than 2.0 or less than 0.5; and (iv) an average of the signal ratios that exceeded the standard deviation of the average signal ratio by more than 1.5-fold. The 2-fold-change cut-off criterion was based on the distribution of fold changes observed in normalised control data. These provide a measurement for inherent variability and thus an estimate for the proportion of false positives that can be expected. An example of an X–Y plot of such data is given in the supplementary file (Sheet1, cGMPSupplementData). The cut-off of a 2-fold change is such that that the number of treatment-induced false positives is 10% or lower. Although the background signal intensity can have a large impact on the interpretation of microarray data (Pan et al., 2005), a doubling or halving of the chosen value (100) did not significantly change the general analysis outcome.

Promoter cis-element analysis

For the detection of putative regulatory cis elements in the promoter regions of coregulated transcripts, 5′ upstream sequences of up to 800 bp (avoiding overlap with preceding coding sequences) were uploaded at the ‘Regulatory Sequences Analysis Tools’ service at http://rsat.ulb.ac.be/rsat/. Sequences were queried using algorithms to detect over-represented strings of four to eight nucleotides searching both DNA strands. The Z score represents the probability for the number of detected motifs to occur relative to the expected number of occurrences based on the motif distribution in the background dataset that contains all Arabidopsis 5′ upstream sequences. Identified putative promoter elements were used to query Arabidopsis cis-element databases such as PlantCARE (http://oberon.fvms.ugent.be:8080/PlantCARE/), Agris (http://arabidopsis.med.ohio-state.edu/AtcisDB/) and Atprobe (http://exon.cshl.org/cgi-bin/atprobe/atprobe.pl) for known functions.

Ion flux and ion contents analyses

For Na+ influx measurements, plants were exposed to growth medium supplemented with 100 mm NaCl. After indicated periods, plants were washed twice for 10 mins in ice cold 20 mm CaCl2 to remove cell wall Na+, weighed and dried at 80oC for 72 h. Ion analysis was carried out on acid extracted root and shoot tissue using Inductively Coupled Plasma Spectrometry (ICP) analysis. Fluxes were calculated by dividing tissue Na+ content by time. For Na+ efflux measurements, plants were pre-loaded with Na+ by growth on growth medium supplemented with 80 mm NaCl for a period of 3 days. After transfer to Na+-free medium, fluxes were calculated from the increased Na+ content in the efflux buffer over the indicated times. For K+ uptake measurements, plants were starved for 3 days on K+-free medium and subsequently transferred to normal growth medium containing 1.9 mm K+. Fluxes were calculated by measuring the decrease in K+ content in the uptake buffer. For K+ efflux measurements, plants grown on standard medium were transferred to K+-free medium or K+-free medium supplemented with 80 mm NaCl. Fluxes were calculated by measuring the increase in K+ content in the efflux buffer. For all experiments, 10–12 plants were used and at least three independent replicates were carried out.

Semi-quantitative RT-PCR

Microarray results were verified by carrying out semi-quantitative RT-PCR. Gene specific primers were designed spanning at least one intron sequence for the following genes: At5g46360 (KCO3) 5′TTTTGTTGTGAAAATCGGGTG3′ and 5′TGCGGTGGTTAAATCAACAAG3′; At3g17690 (CNGC19) 5′AGGTTGGGCTAAGCGATTTGC3′ and 5′ATCGATAAGATTTCTACGTTCATCCGCAG3′; At1g62690 (unknown expressed protein) 5′TCCAGAAGGTTGTGGTGTTCC3′ and 5′TCACCTGTCTTGAACCTCGGA3′, At5g05930 (putative guanylyl cyclase) 5′GGGATATTAGACGGAAATGGAGATT3′ and 5′TATCCGTTCTGGTTCCTCATATTCTC3′; At1g69860 (PTR14) 5′CGCAATCCCGACCGTTCTCA3′ and 5′CGGACCAAAACGGTTTCATAG3′; At2g31910 (CHX22) 5′TCCACCTAGTCGAGCTAACGG3′ and 5′TGTTGTTTTCTGCGGTTTGTC3′; At4g38920 (VHA-c3) 5′GCAAAGAGTGGTGTGGGAGT3′ and 5′CTCGGGAGGACAAAATGATCC3′; At5g24270 (SOS3) 5′AAGAAGAAGAAGAAGAATGCA3′ and 5′GAAACGAAACTTGGAAACGTCC3′; At1g09640 (EF 1B) 5′AAGGCACTCATTGCAGCAGAG3′ and 5′GTCCATACGCTGAAGGAACCC3′; At4g03560 (TPC1) 5′GCTCTATTGGCGTACAGGTCTTTG3′ and 5′GAAGAGTGTGACCATTCCATTGG3′; At1g33440 (NTL1) 5′TTTCCCTTTGTCAAAATCCGC3′ and 5′TAGGCTTCTCCGTCGTGGAAT3′.

PCR reactions were optimised for each pair of gene specific and ubiquitin (At4g05320) primers (5′CACACTCCACTTGGTCTTGCGT3′ and 5′TGGTCTTTCCGGTGAGAGTCTTCA3′) and products were obtained in the linear amplification range (typically 25–30 cycles). At least three reactions were carried out for each gene. PCR products were run on a 2% (w/v) agarose gel and quantified by densitometry using IS-1000 Digital Imaging System (Alpha Innotech, Cannock, UK) software.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. References
  8. Supporting Information
  • Armengaud, P., Breitling, R. and Amtmann, A. (2004) The potassium-dependent transcriptome of Arabidopsis reveals a prominent role of jasmonic acid in nutrient signaling. Plant Physiol. 136(1), 25562576.
  • Bailey, J., Tyson-Capper, A.J., Gilmore, K., Robson, S.C. and Europe-Finner, G.N. (2005) Identification of human myometrial target genes of the cAMP pathway: the role of cAMP-response element binding (CREB) and modulator (CREM alpha and CREM tau(2)alpha) proteins. J. Mol. Endocrinol. 34(1), 117.
  • Balague, C., Lin, B.Q., Alcon, C., Flottes, G., Malmstrom, S., Kohler, C., Neuhaus, G., Pelletier, G., Gaymard, F. and Roby, D. (2003) HLM1, an essential signaling component in the hypersensitive response, is a member of the cyclic nucleotide-gated channel ion channel family. Plant Cell. 15(2), 365379.
  • Bowler, C., Neuhaus, G., Yamagata, H. and Chua, N.H. (1994) Cyclic GMP and calcium mediate phytochrome phototransduction. Cell, 77, 7381.
  • Brett, C.L., Kelly, T., Sheldon, C. and Church, J. (2002) Regulation of Cl–HCO3- exchangers by cAMP-dependent protein kinase in adult rat hippocampal CA1 neurons. J. Physiol. Lond. 545(3), 837853.
  • Clarke, A., Desikan, R., Hurst, R.D., Hancock, J.T. and Neill, S.J. (2000) NO way back: nitric oxide and programmed cell death in Arabidopsis thaliana suspension cultures. Plant J. 24(5), 667677.
  • Donaldson, L., Ludidi, N., Knight, M.R., Gehring, C. and Denby, K. (2004) Salt and osmotic stress cause rapid increases in Arabidopsis thaliana cGMP levels. FEBS Let. 569(1–3), 317320.
  • Durner, J., Wendehenne, D. and Klessig, D.F. (1998) Defense gene induction in tobacco by nitric oxide, cyclic GMP, and cyclic ADP-ribose. Proc. Natl Acad. Sci. USA, 95, 1032810333.
  • Essah, P.A., Davenport, R. and Tester, M. (2003) Sodium influx and accumulation in Arabidopsis. Plant. Physiol. 133X(1), 307318.
  • Finn, J.T., Grunwald, M.E. and Yau, K.W. (1996) Cyclic nucleotide-gated ion channels: an extended family with diverse functions. Annu. Rev. Physiol. 58, 395426.
  • Hoshi, T. (1995) Regulation of voltage dependence of the KAT1 channel by intracellular factors. J. Gen. Physiol. 105, 309328.
  • Hua, B.G., Mercier, R.W., Leng, Q. and Berkowitz, G.A. (2003) Plants do it differently. A new basis for potassium/sodium selectivity in the pore of an ion channel. Plant. Physiol. 132, 13531361.
  • Katagiri, F., Lam, E. and Chua, N.H. (1989) Two tobacco DNA-binding proteins with homology to the nuclear factor CREB. Nature, 340, 727730.
  • Leng, Q., Mercier, R.W., Yao, W.Z. and Berkowitz, G.A. (1999) Cloning and first functional characterization of a plant cyclic nucleotide-gated cation channel. Plant Physiol. 121, 753761.
  • Ludidi, N. and Gehring, C. (2003) Identification of a novel protein with guanylyl cyclase activity in Arabidopsis thaliana. J. Biol. Chem. 278(8), 64906494.
  • Maathuis, F.J.M. and Sanders, D. (2001) Sodium uptake in Arabidopsis thaliana roots is regulated by cyclic nucleotides. Plant Physiol. 127, 16171625.
  • Maathuis, F.J.M., Filatov, V., Herzyk, P. et al. (2003) Transcriptome analysis of root transporters reveals participation of multiple gene families in the response to cation stress. Plant J. 35(6), 675692.
  • Moe, O.W. (1999) Acute regulation of proximal tubule apical membrane Na/H exchanger NHE-3: role of phosphorylation, protein trafficking, and regulatory factors. J. Am. Soc. Nephrol. 10(11), 24122425.
  • Newton, R.P. and Smith, C.J. (2004) Cyclic nucleotides. Phytochemistry, 65(17), 24232437.
  • Pan, K.H., Lih, C.J. and Cohen, S.N. (2005) Effects of threshold choice on biological conclusions reached during analysis of gene expression by DNA microarrays. Proc. Natl Acad. Sci. USA, 102, 89618965.
  • Penson, S.P., Schuurink, R.C., Fath, A., Gubler, F., Jacobsen, J.V. and Jones, R.L. (1996) cGMP is required for gibberellic acid-induced gene expression in barley aleurone. Plant Cell. 8, 23252333.
  • Pharmawati, M., Maryani, M.M., Nikolakopoulos, T., Gehring, C.A. and Irving, H.R. (2001) Cyclic GMP modulatesstomatal opening induced by natriuretic peptides and immunoreactive analogues. Plant Physiol. Biochem. 39, 385394.
  • Qi, Z. and Spalding, E.P. (2004) Protection of plasma membrane K+ transport by the salt overly sensitive1 Na+-H+ antiporter during salinity stress. Plant Physiol. 136(1), 25482555.
  • Rubio, F., Flores, P., Navarro, J.M. and Martinez, V. (2003) Effects of Ca2+, K+ and cGMP on Na+ uptake in pepper plants. Plant Sci. 165(5), 10431049.
  • Santa-Maria, G., Rubio, F., Dubcovsky, J. and Rodriguez-Navarro, A. (1997) The HAK1 gene of barley is a member of a large gene family and encodes a high-affinity potassium transporter. Plant Cell, 9, 22812289
  • Shabb, J.B. and Corbin, J.D. (1992) Cyclic nucleotide-binding domains in proteins having diverse functions. J. Biol. Chem. 267(9), 57235726.
  • Sze, H., Padmanaban, S., Cellier, F. et al. (2004) Expression patterns of a novel AtCHX gene family highlight potential roles in osmotic adjustment and K+ homeostasis in pollen development. Plant Physiol. 136(1), 25322547.
  • Trewavas, A.J., Rodrigues, C., Rato, C. and Malho, R. (2002) Cyclic nucleotides: the current dilemma! Curr. Opin. Plant Biol. 5(5), 425429.
  • Tsuruhara, A. and Tezuka, T. (2001) Relationship between the self-incompatibility and cAMP level in Lilium longiflorum. Plant Cell. Physiol. 42(11), 12341238.
  • Volotovski, I.D., Sokolovski, S.G., Molchan, O.V. and Knight, M.R. (1998) Second messengers mediate increases in cytosolic calcium in tobacco protoplasts. Plant Physiol. 117, 10231030.
  • Weber, I.T., Steitz, T.A., Bubis, J. and Taylor, S.S. (1987) Predicted structures of camp binding domains of type-I and type-II regulatory subunits of camp-dependent protein-kinase. Biochemistry, 26(2), 343351.
  • Wu, Y., Hiratsuka, K., Neuhaus, G. and Chua, N.H. (1996) Calcium and cGMP target distinct phytochrome-responsive elements. Plant J. 10, 11491154.
  • Zhu, J.K. (2002) Salt and drought stress signal transduction in plants. Ann. Rev. Plant Biol. 53, 247273.

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. References
  8. Supporting Information

Two spreadsheet supplementary files are available. TPJ2616Supplementarydata1 contains Supplementary Tables 1–5.

Supplementary Table 1 provides gene annotations, AGI codes and all raw signal and background fluorescence data for all 2-h (in red) and 5-h (in blue) replica arrays.

Supplementary Table 2 provides fold changes (FC) for each transcript of normalised data and present (p) or absent (a) calls for each array feature.

Supplementary Table 3 lists averages and standard deviations of the ratios for all genes that were analysed and fulfilled analysis criteria as described in the methods section.

Supplementary Table 4 provides ratios for genes as described for Supplementary Table 2 but with ratio values that exceed either 2.0-, 2.5- or 3-fold.

Supplementary Table 5 provides a functional classification of all annotated transcripts that showed a 2-fold or larger change in transcript level.

A second file TPJ2616Supplementarydata2 lists all hits found after searching the EMBL-EBI protein database to identify putative cGMP targets in higher plants.

Supplementary Table 6 provides occurrence of putative cyclic nucleotide (cNMP) binding domains in green plant species. The EBI InterPro (http://www.ebi.ac.uk) database was used to survey plant genomes for the presence of IPR000595, a consensus domain for cyclic nucleotide binding. KAT/AKT Shaker type, K+ selective ion channels; CNGC: cyclic nucleotide gated ion channels PKA/PKG: putative cAMP and cGMP dependent protein kinase.

Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

FilenameFormatSizeDescription
TPJ_2616_sm_Supplementarydata1.xls27980KSupporting info item
TPJ_2616_sm_Supplementarydata2.xls21KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.