Light controls phospholipase A2alpha and beta gene expression in Citrus sinensis.

The low-molecular weight secretory phospholipase A2alpha (CssPLA2alpha) and beta (CsPLA2beta) cloned in this study exhibited diurnal rhythmicity in leaf tissue of Citrus sinensis. Only CssPLA2alpha displayed distinct diurnal patterns in fruit tissues. CssPLA2alpha and CsPLA2beta diurnal expression exhibited periods of approximately 24 h; CssPLA2alpha amplitude averaged 990-fold in the leaf blades from field-grown trees, whereas CsPLA2beta amplitude averaged 6.4-fold. Diurnal oscillation of CssPLA2alpha and CsPLA2beta gene expression in the growth chamber experiments was markedly dampened 24 h after transfer to continuous light or dark conditions. CssPLA2alpha and CsPLA2beta expressions were redundantly mediated by blue, green, red and red/far-red light, but blue light was a major factor affecting CssPLA2alpha and CsPLA2beta expression. Total and low molecular weight CsPLA2 enzyme activity closely followed diurnal changes in CssPLA2alpha transcript expression in leaf blades of seedlings treated with low intensity blue light (24 micromol m(-2) s(-1)). Compared with CssPLA2alpha basal expression, CsPLA2beta expression was at least 10-fold higher. Diurnal fluctuation and light regulation of PLA2 gene expression and enzyme activity in citrus leaf and fruit tissues suggests that accompanying diurnal changes in lipophilic second messengers participate in the regulation of physiological processes associated with phospholipase A2 action.

Introduction

Phospholipases are diverse enzymes that cleave phospholipids into fatty acids and other lipophilic substances. Three major forms, known as phospholipases A, C, and D, have been characterized in plants based upon the positional specificity of catalysis on phospholipids (Wang, 2001). Phospholipase A2 (PLA2) hydrolyses the 2-acyl ester linkage of 1,2-diacyl-sn-3-phosphoglycerides to free fatty acids and 1-acyl-2-lysophospholipids (Six and Dennis, 2000; Wang, 2001). Within four major biological subfamilies of PLA2 identified in animals, only low molecular weight secretory PLA2 (sPLA2) and intracellular Ca2+-independent PLA2 (iPLA2), but not cytosolic Ca2+-dependent PLA2 (cPLA2) or PAF acetyl hydrolase/oxidized lipid Lp PLA2 have been identified in plants (Wang, 2001; Burke and Dennis, 2008).

The action of PLA2 has been implicated in various cellular processes including lipid signalling and metabolism (Chapman, 1998; May ; Munnik ), wounding responses (Narváez-Vásquez ), plant–pathogen interactions, defence signalling (Munnik ), fruit abscission (Alferez et al., 2005), auxin-regulated responses (Andre and Scherer, 1991; Ryu and Palta, 1999), and stomatal movement (Seo ). Such diverse physiological actions begin with receptor(s) that receive the released lipid signals, such as lysophosphatidylcholine, lysophosphatidylethanolamine, and fatty acids, and initiate a cascade of events leading to a physiological response (Wang ; Meijer and Munnik, 2003; Wang, 2004). Signals required to activate PLA2 have not been understood clearly.

Light was shown to mediate stomatal movement (Suh ; Seo ), and sPLA2 was shown to play a role in regulating stomatal opening via action on the proton pump (Palmgren ; Lee ; Seo ). Evidence suggests that PLA2-derived lipid products and downstream physiological processes are diurnally regulated by a biological clock that responds to 24 h rhythms. Transcript accumulation of lipoxygenase was found to be under circadian control in maize (Nemchenko ), and linoleic acid content fluctuated diurnally in sunflower seeds (Pleite ). Diurnal oscillations can be controlled by internal elements and exogenous environmental signals, such as temperature, humidity, and light/dark (LD) conditions (McClung, 2006). Several photoreceptors, such as phytochromes and cryptochromes, provide light information to the biological clock (Devlin and Kay, 2000; Salomé ) leading to a wide range of physiological responses. The plant growth hormone auxin is influenced by circadian rhythms in tobacco leaves (Nováková ). Notably, many auxin response genes, response factors, and efflux carriers were found to be diurnally regulated (Harmer ; Covington and Harmer, 2007; Lau ). Auxin-mediated hypocotyl elongation was found to be regulated by PLA2β (Lee ). PLA2-derived lysophospholipids regulate auxin-related physiological processes (Scherer, 2002). Such evidence suggests a link between PLA2 involved auxin-mediated responses and the biological clock. Thus, PLA2-mediated signalling and metabolites may mediate many physiological processes that follow diurnal cycles controled by an oscillator. However, no direct evidence has connected PLA2 gene expression, enzyme activity, and diurnal rhythms. Here, it is reported that two PLA2s isolated from Citrus sinensis, CssPLA and CsPLA, were modulated diurnally by exogenous stimuli. It is demonstrated that select wavelengths of lights, especially blue light, control these gene expressions and total PLA2 activity.

Materials and methods

Plant material and growth conditions

To examine the circadian rhythm in CssPLA and CsPLA gene expression in citrus, 17-year-old Citrus sinensis cv. ‘Valencia’ citrus trees on ‘Swingle’ rootstock were used. Field samples of leaf blades (LB), leaf abscission zones (LAZ), fruit flavedo (FF), and fruit abscission zones (FAZ) were collected as described by Malladi and Burns (2008). Four biological replicates of each tissue were harvested from random canopy locations of four trees at 4 h intervals over 48 h from 19 June to 21 June in 2007. To test diurnal changes under controlled conditions, leaf blades were collected from light-entrained Citrus sinensis cv. ‘Valencia’ potted citrus trees as described by Malladi and Burns (2008). Briefly, 7-year-old trees were entrained to test conditions by transferring to growth rooms set at 25/19 °C in light/dark (LD) for 12:12 h photoperiod length for 7 d prior to sampling. The plants were then transferred into either constant light (LL), constant dark (DD) or kept in LD. Metal halide lamps (1000 W Clear BT37, Philips, Somerset, NJ) with a fluence rate of 245 μmol m−2 s−1 were used as the light source in the growth chambers. Temperature was maintained at 24 °C during the experiments. Mature leaf blades were collected at 4 h or 8 h intervals for 48 h. At least four biological replications from eight trees were collected for each time period.

To determine the effects of different spectra of light on CssPLA, and CsPLA gene expression, and CsPLA2 enzyme activity, 1-year-old seedlings were transferred to light boxes (see below) set to LD photoperiods with various light spectra and intensity for at least 1 week prior to sampling. Leaf blades were harvested at 4 h intervals over 60 h. Temperature and humidity in all light boxes ranged from 23.5 °C to 24.5 °C and 60–65% RH during the sampling period. At least four biological replications from 16 seedlings were collected for each time period.

Light boxes and light resources

Light boxes were constructed with blue, green, red, and red/far-red spectra LED lights as described by Folta . The light boxes containing blue, green, and red lights were supplied with 36 Luxdrive 7007 Endor Star LED light modules whereas the red/far-red lights were supplied by Snap-Lite l (Quantum Devices, Barneveld, WI). Light emission range and fluence rate were measured using a spectroradiometer (SpectraWiz PS-100R, Apogee, Roseville, CA). Blue, green, red, and red/far-red LED lights emitted light from 410–540, 470–620, 580–670, and 600–780 nm, respectively, with emission peaks of blue, green, and red light of 456, 530, and 630 nm, respectively. Two emission peaks of 655 nm and 725 nm were measured with red/far-red light. Light intensity (μmol m−2 s−1) in each box was adjusted to the appropriate illuminating rate as indicated below for experiments.

Nucleotide extraction and gene expression analysis

For CssPLA promoter analysis and gene structure characterization, total DNA was extracted from leaf blades of Citrus sinensis cv. ‘Valencia’ using the Wizard Genomic DNA Purification kit (Promega, Madison, WI). For gene cloning and expression analysis using quantitative RT- PCR, LB, LAZ, FF, and FAZ were frozen in liquid N2, and stored at –80 °C as described by Malladi and Burns (2008). Total RNA was extracted and purified using the RNeasy Mini Kit (Qiagen, Valencia, CA), and DNA contamination was removed using DNase treatment (Qiagen). Total RNA (1 μg) was reverse transcribed using ‘Superscript III’ reverse transcriptase (Invitrogen, Carlsbad, CA) according to the manufacturer's instructions. The cDNA was diluted 1.5-fold and stored at –20 °C until quantitative RT-PCR analysis. Gene expression analysis was performed using quantitative RT-PCR. Citrus glyceraldehyde-3-phosphate-dehydrogenase (CsGAPDH) was used as the constitutively expressed internal calibrator. Optimal primer concentration used in quantitative RT-PCR analysis for CssPLA (PLA2af1: 5′-GTGGGCTGCTGTACAGTGGAT-3′; PLA2ar1: 5′-CAAGCATCAAGGCCATCACA-3′) was 0.25 μM, and was 0.125 μM for CsPLA (PLA2bf1: 5′-TCGGATTTTCTCGGGAATGT-3′; PLA2br1: 5′-CATGTCCATACCCTGTACCATAGTG-3′). Gene expression analysis was performed on each biological replicate in duplicate and results averaged for each replicate. The lowest expression value was used to compare all others within the experiment. To compare expression levels between CssPLA and CsPLA, relative expressions were compared within an experiment to the lowest expression sample of CssPLA.

Gene cloning and promoter and amino acid analysis

Genes and promoters were obtained from total RNA and genomic DNA extracted from leaf blades in Citrus sinensis cv. ‘Valencia’. Gene specific primers designed based on the partial sequences of PLA (5′-TCTGATATGAGAATACTCTGGTGC-3′) and β (5′-GCCACGAAAAGTTTAAGAGATGC-3′) obtained from citrus HarvEST database (http://harvest.ucr.edu/) were used for CssPLA and CsPLA cloning. 5′-RACE (Rapid Amplification of cDNA End, Invitrogen), and 3′-RACE were performed for specific gene amplification via Polymerase Chain Reaction (PCR) using Taq DNA polymerase (Qiagen). Inverse PCR was performed to obtain CssPLA and CsPLA promoter sequences by restriction enzyme digestion of genomic DNA, self-ligation, followed by PCR amplification using Elongase enzyme mix (Invitrogen) with inverse primers. The amplified DNA fragments were cloned into pGEM-T easy vector (Promega). Transformation of Escherichia coli JM109 cells (Promega) was performed, and the resulting plasmid DNA was purified using a Wizard DNA purification kit (Promega). Sequencing was performed at the Interdisciplinary Center for Biotechnology Research, University of Florida (Gainesville, FL). Nucleotide sequences of CssPLA (accession no. GU075396), CsPLA (accession no. GU075398), and 1000 bp promoter regions of CssPLA (accession no. GU075397) from Citrus sinensis cv. ‘Valencia’ were used for this study. Alignment of amino acid sequences were performed using ClustalW2 software (http://www.ebi.ac.uk/Tools/clustalw2/index.html). Signal peptide prediction was performed using SIG-Pred: Signal Peptide Prediction web server (http://bmbpcu36.leeds.ac.uk/prot_analysis/Signal.html).

CssPLA2 enzyme extraction and activity analysis

Citrus leaves (1 g) were ground in liquid N2, dissolved in 2.5 ml of cold extraction buffer (1 mM EDTA, 100 mM TRIS-HCl, 2% (w/v) PVPP, 0.15 M sorbitol, pH 7) with 1 mM PMSF added, and the resulting homogenate centrifuged at 12 000 rpm at 4 °C for 15 min. Supernatant proteins were desalted using a PD-10 column (GE Healthcare, Buckinghamshire, UK) equilibrated and eluted with column buffer (10 mM HEPES, 1 mM MgCl2, pH 7). Total proteins in the eluate were precipitated overnight at 4 °C using ammonium sulphate to saturation, centrifuged 14 000 rpm at 4 °C, washed with acetone, and resuspended in extraction buffer. Concentrated proteins were quantified and stored at –20 °C for the total PLA2 enzyme activity assay. Total PLA2 activity assay (Cayman Chemical Co, Ann Arbor, MI, USA) was performed as described (Alferez et al., 2005). sPLA2 was fractionated from total protein samples using centrifugal filter (Microcon, Bedford, MA, USA) with a molecular weight cut-off of 30 kDa, and enzyme activity measured.

Results

Cloning and sequence analysis

Gene specific primers designed based on the EST sequences of PLA and β from the citrus HarvEST database were used for CssPLA and CsPLA cloning. 5′- and 3′-RACE was performed to obtain the full-length sequences. Full-length CssPLA and CsPLA cDNA sequences encoded 17.1 and 31.6 kDa proteins of 156 and 279 amino acids, respectively, and shared 31% identity (Fig. 1). CssPLA2α and CsPLA2β contained key phospholipase catalytic sites of a Ca2+-binding loop, the active site motif with a conserved his/asp dyad (HD), and 12 conserved Cys residues that form six disulphide bonds for conformational integrity. A eukaryotic signal peptide sequence was present in the N-terminus of CssPLA2α but not CsPLA2β. The KxEL endoplasmic reticulum (ER) retention sequence (Pagny ; Seo ) was found in the 3′-terminus of CsPLA2β but not CssPLA2α. The presence of catalytic Ca2+-binding loop and active site motifs suggests that CssPLA2α and CsPLA2β encoded PLA proteins that catalyse hydrolysis of the phospholipase sn-2 bond in a Ca2+-dependent manner (White ; Yu and Dennis, 1991; Six and Dennis, 2000; Lee ; Hsu ). The molecular weight of CssPLA suggests that it encoded a secretory phospholipase A2 (Six and Dennis, 2000).

Fig. 1.

Amino acid sequence alignment of CssPLA2α and CsPLA2β. The KxEL endoplasmic reticulum (ER) retention sequence in the 3′-terminus is only present in CsPLA2β. Eukaryotic signal peptide cleavage site present only in CssPLA2α is underlined. ‘*’, identical amino acids; ‘:’, conserved substitutions; ‘.’, semi-conserved substitutions; ‘–’, sequence gaps; ↓, conserved Cys residues. Ca2+-binding loop, PLA2 catalytic motifs, and endoplasmic reticulum (ER) retention sequence are enclosed in boxes.

Sequences (1 kb) upstream from the putative ATG translational start cordon of CssPLA and β were examined for promoter regulatory elements. Core regulatory TATA box (TATAAA) and CAAT/CCAAT elements, light regulatory elements, and some of clock regulated elements, such as G-box and CCA1 binding sites, were found in CssPLA and CsPLA promoter sequences (Fig. 2; Table 1). Clock-regulated elements associated with blue light and evening element were present in CssPLA but absent in CsPLA, whereas a phytochrome A element was present in CsPLA but not CssPLA. The abundance of light/clock and evening regulatory elements in the promoter region of CssPLAα suggests CssPLA may function as a light-activated gene diurnally governed with an evening-specific expression pattern. CsPLA may be regulated by light but less responsive.

Fig. 2.

Nucleotide sequences of CssPLA (A) and CsPLA (B) promoter regions. One kb 5′ upstream sequences from the transcription initiation start sites (ATG) are listed. The light- and clock-regulated binding elements described in Table 1 are underlined. SRE, shade response element; CCA1B, CCA1 binding element; CBE, clock and blue-light-regulated element; EE, evening element; all other elements presented are listed in Table 1.

Table 1.

Promoter analysis of 1 kb sequences upstream of translational start codon (ATG) of CssPLA and CsPLA

Cis-regulatory element	Sequencea	No. of elements		Reference
		CssPLA2α	CsPLA2β
Core regulatory elements
TATA box	TATAAA	2	1	Breathnach and Chambon, 1981
CAAT	CAAT	13	5	Terzaghi and Cashmore, 1995
CCAAT	CCAAT	2	1	Terzaghi and Cashmore, 1995
Light-regulatory element
WC-1/WC-2 binding element	GATA	4	2	Giuliano et al., 1988
GT-1	GRWAAW	4	3	Terzaghi and Cashmore, 1995; Zhou 1999
I-box	GATAAG	1	1	Chatterjee et al., 2006
Shade response element	TAATTA	1	1	Devlin et al., 2003
Clock regulated element
G-box	CACG	3	1	Hudson and Quail., 2003
CCA1 binding element	ATCT	7	5	Wang and Tobin, 1998
CCA1ATLHCB1	AA N1-2AATCT	1	0	Chatterjee et al., 2006
CIACADIANLELHC	CAAN2-4ATC	1	1	Chatterjee et al., 2006
Clock and blue light-regulated element	ACTN1-2CCAAT	2	0	Folta and Kaufman, 1999
Evening element	AATATHT	4	0	Harmer et al., 2000; Xu and Johnson, 2001
Evening element	TTAATATCT	1	0	Harmer et al., 2000; Xu and Johnson, 2001
Phytochrome A-induced element
SORLIP1	GCCAC	0	1	Hudson and Quail, 2003
SORLIP2	GGGCC	0	0
SORLIP3	CTCAAGTGA	0	0
SORLIP4	GTATGATGG	0	0
SORLIP5	GAGTGAG	0	0

Cis-regulatory elements, sequence and references are listed.

R=A/G; W=A/T; M=A/C, H=A/C/T, N=A/C/T/G.

Tissue-specific and diurnal expression of CssPLA and CsPLA

CssPLA transcript accumulation displayed a distinct diurnal rhythm in leaf blades (LB), leaf abscission zones (LAZ), fruit flavedo (FF), and fruit abscission zones (FAZ) in field samples (Fig. 3). Minimum and maximum CssPLA transcript accumulation in all tissues occurred at approximately 10.00 h and 18.00 h, respectively. The order of maximum amplitude of CssPLA relative expression was LB>LAZ>FAZ>FF (Fig. 3A). At its maximum, relative expression in LB was 1337-fold and 643-fold higher than its minimum over two diurnal cycles. Accumulation of CsPLA RNA oscillated in LB and much less in LAZ. CsPLA transcript accumulation maxima and minima occurred at 18.00 h and 10.00 h, respectively, however, the decline in expression during the dark period was gradual compared with CssPLA. Relative CsPLA expression in LB was 7.7-fold and 5.0-fold higher at its maxima than its minima over two diurnal cycles. Closer inspection of LAZ, FF, and FAZ expression revealed the order of maximum rhythmic amplitude of CsPLA was LB>LAZ, but no oscillations occurred in FAZ or FF (Fig. 3B). The basal expression of CsPLA was over 50-fold higher in LB compared to LB CssPLA expression.

Fig. 3.

Diurnal oscillation in CssPLA (upper graph) and CsPLA (lower graph) gene expression under field conditions. (A) Oscillation of gene expression in leaf blades (LB, closed circles), leaf abscission zones (LAZ, open circles), fruit flavedo (FF, closed triangles), and fruit abscission zones (FAZ, open triangles) during natural daylight (white area) and darkness (grey area) over a 48 h period. (B) Enlargement of LAZ, FF, and FAZ gene expression showing diurnal pattern. Actual time of days is indicated on the x-axis. Vertical bars represent standard error of mean (n=4).

Light regulates oscillation of LB CssPLA and CsPLA expression

CssPLA and CsPLA expression in LB from potted citrus plants held in light/dark (LD) and constant temperature in the growth room demonstrated diurnal oscillation (Fig. 4). Average maximum amplitude was similar to that from field samples. Transfer of plants from LD to constant dark (DD) resulted in the cessation of CssPLA expression oscillation within 24 h, whereas transfer of plants from LD to constant light (LL) markedly dampened expression. The period of CsPLA expression shifted when transferred from LD to LL or DD conditions. Thus, light was a major factor regulating CssPLA and CsPLA expression.

Fig. 4.

Diurnal regulation of CssPLA (upper graph) and CsPLA (lower graph) gene expression in leaf blade under light/dark (open triangles), constant dark (closed circles) or constant light (open circles). Subjective light and dark photoperiods are indicated by open and closed boxes, respectively. Double arrow lines indicate the samples taken from the last 10 h (light period) of the 7 d entrainment under light/dark condition. Vertical bars represent standard error of mean (n=4).

CssPLA and CsPLA gene expression in seedling LB oscillated in blue and green light (Fig. 5). The amplitude of CssPLA expression was much higher than CsPLA under all light spectra examined. Rhythmicity of CsPLA gene expression in blue and green light was erratic, but oscillations were more pronounced with blue light treatment. CssPLA expression was lowest 1 h before the end of dark period through 3 h after the light period was initiated. Expression increased 7–11 h after the light treatment commenced and peaked 11 h after illumination. Expressions of both genes in blue light were greater at 24 μmol m−2 s−1 intensity compared with 3 μmol m−2 s−1. CssPLA gene expression in seedling LB also oscillated in red, red/far-red, and the combination of blue, green, and red light held in LD (see Supplementary Fig. S1 at JXB online). The amplitudes of CssPLA expression were 18-fold and 8-fold lower in red and red/far-red compared to blue light at 24 μmol m−2 s−1 intensity, respectively. Blue light was as effective as the combination light treatment in inducing maximum CssPLA gene expression. Rhythmicity of CsPLA gene expression in red/far-red light was also erratic and the amplitude was much lower than blue light. Increasing green, red or red/far-red light intensity from 3 to 24 μmol m−2 s−1 had minor effects on expression of both genes (see Supplementary Fig. S1 at JXB online).

Fig. 5.

Diurnal expression of CssPLA (upper graph) and CsPLA (lower graph) under blue and green 12/12 h light/dark cycles. Seedlings were exposed to 12/12 h light/dark cycle of blue (24 μmol m−2 s−1, closed circles; and 3 μmol m−2 s−1, open circles), green (3 μmol m−2 s−1, open triangles) or dark conditions (0 μmol m−2 s−1 , closed squares). Subjective light and dark photoperiods are indicated by open and closed boxes and white or grey background. Vertical bars represent standard error of mean (n=4).

PLA2 enzyme activities

CssPLA relative gene expression in LB of seedlings treated with blue light under LD conditions and total CsPLA2 enzyme activity showed similar rhythmicity, with the peak occurring 11 h after the light period commenced (Fig. 6A, B). Like CssPLA expression, total PLA2 activity remained low and constant under DD conditions. The peak activity of the low-molecular-weight PLA2 fraction corresponded with the maximum amplitude of CssPLA expression.

Fig. 6.

(A) Diurnal CssPLA (open symbols) and CsPLA (closed symbols) expression, and (B) CsPLA2 total (open symbols) and low molecular weight (sPLA2) (closed triangles) enzyme activity under 12/12 h blue/dark cycle, or 24 h constant dark. Blue light was applied at 24 μmol m−2 s−1. Subjective light and dark photoperiods are indicated by open and closed boxes and white or grey background, respectively. Vertical bars represent standard error of mean (n=4).

Discussion

Phospholipase A2 are enzymes that hydrolyse phospholipids to free fatty acids and 1-acyl-2-lysophospholipids; such hydrolysis products have downstream signalling functions (Schaffer ; Wang, 2004). Thus, PLAs represent an important control point for phospholipid signalling. This study demonstrated diurnal fluctuation of two PLA2 transcripts, PLA2 enzyme activity, and their regulation by light. These data suggest that increased production of PLA2-derived signals during the light period may have significant influence over physiological processes influenced by PLA2 action.

Sequence analysis of CssPLA indicated that CssPLA encoded a low molecular-weight secretory Ca2+-dependent lipase. The presence of an N-terminal eukaryotic secretory signal peptide in CssPLA suggests that the mature peptide functions as a secretory protein, whereas the C-terminal ER retention-like signal KxEL in CsPLA2β indicates the mature peptide may be localized to ER (Pagny ; Seo ). The higher molecular weight of CsPLA, the presence of the Ca2+-binding loop, and its predicted target in the ER indicates it may represent a cPLA2 or related sequence.

The presence of diurnal and circadian promoter elements in 5′-upstream sequences of CssPLA and CsPLA suggests their regulatory control over oscillating expressions. A tight resetting pattern at ‘dusk’ was evident with CssPLA expression in all tissues examined but more prominent in leaf tissue; however, CsPLA diurnal expression was leaf tissue-specific and more slowly declined during the dark period. The clock-regulated evening element present only in the promoter region of CssPLA assumed necessary for evening-specific transcription (Harmer ; Michael and McClung, 2002) may regulate its expression. Light was an important external factor controlling diurnal expression pattern. The failure of oscillation in CssPLA and CsPLA gene expression in the absence of the light/dark transition indicated that the rhythmicity of CssPLA and CsPLA gene expressions is controlled by the LD cycle and transiently regulated via an internal circadian clock. Citrus patatins were also cloned and examined to determine if diurnal expression patterns existed. None showed a diurnal profile over a 48 h period in any tissue examined (Y Lluch and JK Burns, unpublished data).

Low intensity blue light regulated diurnal rhythmicity of CssPLA and CsPLA gene expression in a dose-dependent manner. Diurnal oscillation occurred with green, red and red/far-red light treatment, but blue light induced much higher amplitude. Increasing blue light intensity from 0.18 to 24 μmol m−2 s−1 was sufficient to increase amplitude of CssPLA gene expression 60-fold, whereas amplitude of CsPLA gene expression increased only 2-fold over this same intensity range (data not shown). High light intensities over a wide range of wavelengths in field conditions and the growth room induced comparatively larger CssPLA and CsPLA gene expression amplitudes. Some members of the light harvesting chlorophyll binding (Lhcb) gene family strongly respond to increasing blue light fluence by increasing transcription rate (Folta and Kaufman, 1999; Harmer ). The CCA1 binding element implicated in circadian blue light responses found in the promoter region of Lhcp (Wang and Tobin, 1998) is present in CssPLA and CsPLA. Thus, CCA1 may function as a core oscillator and connect blue-light and clock signals with diurnal control of CsPLA gene expression.

As much as 11–13.4% of the Arabidopsis transcriptome is diurnally regulated with gene changes more than 2-fold (Schaffer ; Bläsing ). The amplitude of most diurnally regulated genes treated with low intensity light (130 μmol m−2 s−1) was less than 2-fold. Only 2.4% reached amplitudes of 4 or more and only 0.001% were grouped in the lipid metabolism functional category (Bläsing ). The 990-fold and 300-fold amplitude change in CssPLA gene expression under high light intensity in the field and low intensity blue light-LD cycle, respectively, suggests important functions for CssPLA in citrus and perhaps other plants. However, the remarkable changes in expression resulted in only a 2-fold increase in total PLA2 enzyme activity. The striking reduction in efficiency of output implies post-transcriptional regulation, loss of mRNA stability, or suppression of translation (Lidder ).

Diurnal fluctuation in CssPLA gene expression and resulting enzyme activity may allow plants to respond to daily changes in environment. Abscission, stomatal conductance, and hypocotyl elongation respond to light and diurnal cues (Decoteau and Craker, 1983; Somers ; Chatterjee ; Xu ; Malladi and Burns, 2008). It was previously demonstrated that inhibiting sPLA2 enzyme activity markedly reduced lipid hydroperoxides and efficacy of a citrus abscission agent (Alferez ), suggesting that production of lipid-derived signals was associated with the acceleration of abscission. Phospholipase D gene expression diurnally fluctuated in citrus leaf and fruit tissues, with peak amplitude occurring at the end of the light period (Malladi and Burns, 2008). Natural fruit attachment force fluctuated diurnally, with maximum loosening occurring mid-day (Pozo ). Similarly, the response to abscission agents was greater when applied at midday, even when temperatures were held constant (Pozo ; Malladi and Burns, 2008).

sPLA2 action was shown to play a role in regulating stomatal opening during the light hours in C4 plants. Inhibition of sPLA2 inhibited light-induced stomatal opening (Suh ). Application of lysophospholipids and free fatty acids activated the stomatal proton pump, enhanced blue light-induced stomatal opening, and reversed the effect of PLA2 gene silencing (Palmgren ; Lee ; Seo ). sPLA2 activity is enhanced by auxin application, and PLA2-derived lysophospholipids can regulate auxin-related physiological processes (Scherer, 2002). Diurnal hypocotyl enlongation and leaf expansion that occurred with blue light treatment (Cashmore, 1997) appeared to be regulated via auxin signalling (Lee ; Wilmoth ). Cell elongation and auxin content diurnally fluctuated (Harmer ; Nováková ), and peak auxin content occurred near the end of the light cycle. Our work demonstrated that low intensity blue light increased total CsPLA2 enzyme activity and activity associated with the low molecular weight fraction, presumably sPLA2, and activity oscillations corresponded with gene expression in citrus seedling leaves. Taken together, these results indicate the potential for production of diurnally oscillating lipid signals that may regulate organ loosening, stomatal conductance, cell-elongation and perhaps other light-regulated diurnal physiological processes.

Although blue light appears to play key roles in these processes, red and red/far-red light interact with blue light (Folta and Maruhnich, 2007), and in our study promoted diurnal expression, albeit at lower amplitude. Red and red/far-red light regulates K+ accumulation and stomatal opening in guard cells (Göring ; Talbott ; Doi and Shimazaki, 2008), auxin-regulated cell and hypocotyl elongation (Gotô and Suzuki, 1980; Reed ; Takase ), and influences abscission (Craker ). CssPLA and CsPLA redundantly oscillated with low amplitude under low intensity red/far red light (see Supplementary Fig. S1 at JXB online), implying such light interactions participate in PLA2 regulation. Low-intensity green light also promoted CssPLA and CsPLA diurnal oscillation. Green light reversed blue-induced stomatal opening (Frechilla ; Talbott ), but accelerated stem elongation (Folta and Maruhnich, 2007). How green, red, and red/far-red light interact with blue light and contribute to the regulation of CssPLA and CsPLA gene expression and downstream biological activities are unclear.

Supplementary data

Supplementary data are available at JXB online.

Supplementary Fig. S1. Diurnal expression of CssPLA (A) and CsPLA (B) under 12 h different light/12 h dark cycles.