CsPLDalpha1 and CsPLDgamma1 are differentially induced during leaf and fruit abscission and diurnally regulated in Citrus sinensis.

Understanding leaf and fruit abscission is essential in order to develop strategies for controlling the process in fruit crops. Mechanisms involved in signalling leaf and fruit abscission upon induction by abscission agents were investigated in Citrus sinensis cv. 'Valencia'. Previous studies have suggested a role for phospholipid signalling; hence, two phospholipase D cDNA sequences, CsPLDalpha1 and CsPLDgamma1, were isolated and their role was examined. CsPLDalpha1 expression was reduced in leaves but unaltered in fruit peel tissue treated with an ethylene-releasing compound (ethephon), or a fruit-specific abscission agent, 5-chloro-3-methyl-4-nitro-1H-pyrazole (CMNP). By contrast, CsPLDgamma1 expression was up-regulated within 6 h (leaves) and 24 h (fruit peel) after treatment with ethephon or CMNP, respectively. CsPLDalpha1 expression was diurnally regulated in leaf blade but not fruit peel. CsPLDgamma1 exhibited strong diurnal oscillation in expression in leaves and fruit peel with peak expression around midday. While diurnal fluctuation in CsPLDalpha1 expression appeared to be light-entrained in leaves, CsPLDgamma1 expression was regulated by light and the circadian clock. The diurnal expression of both genes was modulated by ethylene-signalling. The ethephon-induced leaf abscission and the ethephon- and CMNP-induced decrease in fruit detachment force were enhanced by application during rising diurnal expression of CsPLDgamma1. The results indicate differential regulation of CsPLDalpha1 and CsPLDgamma1 in leaves and fruit, and suggest possible roles for PLD-dependent signalling in regulating abscission responses in citrus.

Introduction

Abscission is a tightly regulated process that ultimately results in organ detachment from the parent plant. Abscission occurs primarily at developmentally pre-determined and anatomically distinct regions termed abscission zones (AZ). Knowledge of mechanisms involved in leaf and fruit abscission is essential to develop strategies to control them and to improve harvesting practices or unwanted crop loss in fruit crops such as citrus. Ethylene plays a primary role in signalling and accelerating abscission responses (Jackson and Osborne, 1970; Bleecker and Patterson, 1997). Increased ethylene biosynthesis through over-expression of aminocyclopropane-1-carboxylate (ACC) synthase, an enzyme involved in ethylene biosynthesis, leads to premature flower abscission, while a block in ethylene perception in the never ripe (nr) mutant, delays petal abscission in tomato (Lanahan ; Wilkinson ). Interplay between ethylene and other plant hormones also play a critical role in regulating the progression of organ abscission. A balance of ethylene and auxin levels at the AZ is considered a key factor determining cell separation (Sexton and Roberts, 1982; Brown, 1997; Patterson, 2001; Taylor and Whitelaw, 2001). While ethylene promotes cell separation, auxin inhibits it and also reduces sensitivity of the AZ to ethylene. In addition, factors such as wounding and stress regulate organ abscission (Taylor and Whitelaw, 2001). However, the mechanisms involved in signalling abscission responses are still unclear.

In citrus, ethylene-releasing compounds such as ethephon (2-chloroethane-phosphonic acid) or chemical wounding agents such as CMNP (5-chloro-3-methyl-4-nitro-1H-pyrazole), reduce mature fruit detachment force (Goren, 1993; Burns, 2002; Yuan and Burns, 2004). In addition, ethephon application results in extensive leaf abscission. Blocking ethylene perception through the application of 1-MCP (1-methylcyclopropene) prevents ethephon-induced leaf abscission, although it has a minimal effect on mature citrus fruit detachment force (Pozo ). The application of two putative heterotrimeric guanine nucleotide–binding protein (G-protein) receptor agonists, clonidine and guanfacine, mitigates ethephon-induced leaf abscission in citrus (Burns ; Yuan ). These agonists reduce ethephon-induced expression of ACS1 (ACC synthase) and ACO (ACC oxidase) and markedly decrease ethylene production in citrus leaves (Yuan ). Hence, a role for heterotrimeric G-protein signalling in mediating ethylene-regulated abscission responses seems plausible.

Recent progress in G-protein signalling research in plants suggests their involvement in mediating physiological responses to plant growth regulators (Ashikari ; Ueguchi-Tanaka ; Romanov ; Coursol ). Transduction of signals perceived by the G-protein signalling pathway is achieved in part through interaction with phospholipases, enzymes that catalyse the hydrolysis of membrane phospholipids (Assmann, 2002; Wang, 2005). D-type phospholipases (PLDs) can bind to and interact with Gα, a heterotrimeric G-protein subunit, through a motif analogous to the ‘DRY’ motif in animal G-protein coupled receptors (GPCRs) (Zhao and Wang, 2004). Activation of heterotrimeric G-proteins through the release of the Gα subunit removes inhibition of PLDs (Lein and Saalbach, 2001). Through direct or indirect enzymatic action of PLDs, a variety of lipid signalling molecules such as phosphatidic acid (PA) are produced. PLD activity and associated signalling molecules are regulated by various types of stress, wounding, and ABA and ethylene, plant growth regulators often associated with senescence and/or abscission (Lee ; Ritchie and Gilroy, 1998; Frank ; Taylor and Whitelaw, 2001; Wang, 2002; Welti ; Hong ).

Several classes of plant PLDs have been identified (Pappan , b; Qin ; Qin and Wang, 2002; Wang, 2004). The Arabidopsis PLD family has 12 PLD-encoding genes classified into six types: PLDα, PLDβ, PLDγ, PLDδ, PLDϵ, and PLDζ (Wang, 2005). PLDα is involved in mediating hyperosmotic stress responses, and ABA- and ethylene-dependent senescence of detached leaves in Arabidopsis (Fan ; Hong ). PLDβ regulates active oxygen species production and polyphenol oxidase activity in tomato (Laxalt ; Bargmann ). PLDδ may regulate freezing responses as its over-expression enhances freezing tolerance (Li ). By contrast, PLDα negatively regulates freezing tolerance in Arabidopsis as PLDα anti-sense plants exhibit increased survival at low temperature (Welti ). PLDζ2 is involved in mediating auxin responses in Arabidopsis (Li and Xue, 2007). These data indicate diverse functions for PLDs in plant growth and development, and also suggest specific roles for individual isoforms. As with Arabidopsis, citrus probably has multiple PLD genes that participate in a variety of growth- and development-related processes. However, little is known about the citrus PLD family.

In previous studies, expression of a PLD was found to be up-regulated in Arabidopsis following application of CMNP (Alferez ). CMNP also induces phospholipase A2 and lipoxygenase activities in citrus fruit flavedo (peel), suggesting modulation of lipid-signalling during abscission and a role for fruit flavedo in the response (Alferez ). In addition, expression of several genes involved in phospholipid metabolism and signalling was altered in various citrus tissues treated with abscission-inducing agents (J Burns et al., unpublished results). Hence, it is hypothesized that PLDs and phospholipid signalling played a role in mediating leaf and mature fruit abscission responses in Citrus sinensis cv. ‘Valencia’ sweet orange. In this study, the isolation and characterization of two abscission agent-regulated PLDs are reported. Evidence is presented for diurnal, light- and ethylene signalling-dependent regulation of PLD expression. A relationship is suggested between PLD-dependent signalling and the regulation of AZ sensitivity.

Materials and methods

Leaf abscission in whole trees

Seventeen-year-old Citrus sinensis cv. ‘Valencia’ citrus trees on ‘Swingle’ rootstock located at the Citrus Research and Education Center, Lake Alfred, FL, USA, were used for field abscission experiments. Leaf abscission was studied using the ethylene-releasing agent, ethephon (Ethrel®). Ethephon concentrations were selected based on previous experiments and forecasted temperatures at application, as high temperatures are known to increase efficacy (Yuan and Burns, 2004). Canopy sections on 10 trees were tagged and randomly assigned to water (control) or ethephon treatments (n=5). Water (control) and ethephon (600 mg l−1) were applied to ‘run-off’ at 10.00 h using a backpack sprayer. Temperature at the time of application was 26 °C. A branch was tagged in each section and leaf number was counted daily for 6 d, and then at 10, 14, and 20 d after application. Leaf samples were collected from water- and ethephon-treated trees for analysis of PLD gene expression at 0, 3, 6, 24, and 48 h after treatment. To determine the effect of time of ethephon application on abscission response, four separate trees were sprayed either with water or ethephon (400 mg l−1) at 09.00 h and 13.00 h and leaf number on tagged branches was counted at various times up to 10 d after application. Temperature at the time of ethephon application was 30 °C (09.00 h) and 35°C (13.00 h).

Fruit abscission in whole trees

Citrus sinensis cv. ‘Valencia’ sweet orange trees were used for fruit abscission experiments. Ethephon (600 mg l−1), CMNP (250 mg l−1) and water were applied at 09.00 h to canopy sections until runoff (n=3). Ethephon and CMNP concentrations were selected based on previous experiments on fruit abscission. The temperature at the time of application in this study was 29 °C. Fruit detachment force (FDF) was measured using a digital force gauge (Force One, Wagner Instruments, Greenwich, CT, USA) at 0, 2, 4, and 7 d after application. Fruit were sampled from treated canopies at various intervals up to 96 h after application for gene expression analysis. Fruit peel (flavedo) was removed from the equatorial region of sampled fruit, frozen in liquid nitrogen, and stored at –80 °C for gene expression analysis. To determine the effect of time-of-application of abscission agents on FDF, ethephon (600 mg l−1) and CMNP (250 mg l−1) were applied at 09.00 h and 13.00 h to canopy sections (n ≥4). Temperatures at the time of application were 25 °C and 28 °C, respectively. FDF was measured 4 d after application.

RNA isolation and RT-PCR

For leaf tissue, approximately 4 cm of leaf blade from fully expanded leaves was excised from the mid-section of the leaf blade and snap-frozen in liquid N2. Laminar abscission zones (LAZs) were removed by excising approximately 5 mm of tissue proximal and distal to the abscission zone plane. Fruit flavedo was removed using a kitchen-type potato peeler and snap-frozen. Tissues were ground in liquid N2 and RNA from all tissues was extracted using the guanidine isothiocyanate method. RNA was precipitated using a salt solution (1.4 M sodium chloride and 0.8 M sodium citrate) and isopropanol, washed in 70% ethanol, and precipitated overnight in ethanol. The resulting RNA was pelleted, washed in 70% ethanol, dried, and dissolved in diethyl-pyrocarbonate (DEPC) treated water. Total RNA (0.5 μg) was treated with DNase I (Promega) to remove genomic DNA contamination and reverse transcribed using ‘Superscript III’ reverse transcriptase (Invitrogen) according to the manufacturer's instructions. The cDNA was diluted 5-fold and stored at –20 °C until further analysis.

Isolation of full-length CsPLDα1 and CsPLDγ1

Phospholipase genes were isolated from fruit flavedo cDNA using degenerate primers, RT-PCR, and 3′ and 5′ rapid amplification of cDNA ends (RACE). At each step, the amplified fragments were cloned into pGEM-T-Easy and sequenced at the University of Florida Core Sequencing Facility. Degenerate primers P1 and P2, designed based on amino acid sequence similarity among several plant PLDs, amplified a 537 base pair fragment with similarity to plant PLDα from citrus flavedo cDNA. Another degenerate primer (P3) similarly designed and used with a gene-specific primer (P4) amplified a fragment from the 5′ region of citrus PLDα. The 5′ RACE system (Invitrogen, Carlsbad, CA, USA) was used along with a gene specific primer (P5) to isolate the 5′ region of PLDα. A 3′ RACE strategy was used with primers P6 and P7 to amplify the 3′ region. Similarly, degenerate primers (P8 and P9) were used to amplify a 1047 bp fragment with similarity to PLDγ. The 5′ region was amplified using 5′ RACE with P10 as a gene specific primer. Extension of the 3′ region was achieved with a degenerate primer (P11) and a gene specific primer (P12). Finally, 3′ RACE was performed with P13 and P7. Primer sequences used for PLD gene isolation are presented in the Supplementary data (Table S1) that can be found at JXB online.

Gene expression analysis

An Applied Biosystems (Foster City, CA, USA) 7500 Fast Real-Time PCR system was utilized for quantitative real-time RT-PCR analyses. Analysis was performed on 1 μl of diluted cDNA in a final reaction volume of 20 μl using the SYBR® Green PCR Master Mix (Applied Biosystems). PCR conditions were 50 °C 2 min; 95 °C 10 min followed by 40 cycles of 95 °C 15 s; 60 °C 1 min. Melting curve analysis was performed to confirm target-specific amplification. Primer concentration was optimized and primer validation was performed to enable relative gene expression analysis using the ΔΔCt method. Citrus glyceraldehyde-3-phosphate-dehydrogenase (CsGAPDH) was used as the calibrator. A citrus actin sequence was also used for confirmation in some experiments with similar results. Data were normalized to a control replicate (time = 0 h). At least three biological replicates were utilized in every experiment and all analyses were performed in duplicate. Primers used for real-time analyses are listed in the Supplementary data (Table S2) that can be found at JXB online.

Diurnal oscillation in PLD gene expression

To examine diurnal oscillation in PLD gene expression in citrus leaves, potted ‘Valencia’ trees were placed in growth rooms maintained at 25/19 °C (12/12 h; light/dark) for at least 7 d prior to sampling. Halogen lamps were used as light source (230–260 μmol m−2 s−1). Mature leaves were sampled at 4 h or 8 h intervals for 48 h, RNA was extracted, and gene expression analysis was performed (n=4). Diurnal oscillation in PLD gene expression in fruit flavedo was studied using citrus trees in the field. Fruit flavedo was collected from citrus fruit harvested from the exterior region of the canopy at 4 h or 8 h intervals, RNA was extracted and gene expression analysis was performed (n=4).

To determine the effect of continuous light and continuous dark exposure on PLD gene expression, potted citrus trees were entrained in the growth room under the above described light/dark conditions (LD) for 1 week and subsequently transferred to either constant light (LL) or constant dark (DD). A constant temperature of 24 °C was maintained during entrainment and during continuous light and dark treatments. Leaves were sampled at 4 h or 6 h intervals and gene expression analysis was performed (n=4).

Effect of 1-MCP on diurnal PLD gene expression

‘Valencia’ field trees from the above-mentioned field block were treated with water (control) or 1-MCP (5 mM SmartFresh®; according to Pozo ) to inhibit ethylene perception (n=4). Applications were made at 10.00 h on a clear and calm day. Leaf samples were collected from treated trees at 4 h intervals for 48 h and PLD gene expression analysis was performed. 1-MCP applications were followed by ethephon (400 mg l−1) application on selected branches. 1-MCP decreased the extent of ethephon induced leaf abscission on these branches, indicating that it reduced ethylene perception.

Results

Isolation of phospholipase D genes from citrus

Full-length citrus PLDα and PLDγ sequences, designated CsPLDα1 and CsPLDγ1, were isolated and characterized. The predicted amino acid sequence of CsPLDα1 has 86% and 81% identity with PLDα from castor bean and Arabidopsis, respectively. CsPLDα1 possesses an N-terminal C2 domain thought to be involved in calcium and phospholipid binding and two HKD domains essential for catalytic activity. A motif containing ‘ERF’ residues followed by a hydrophobic chain (VYIVV), and analogous to the ‘DRY’ motif of GPCRs was identified within the catalytic region. The ERF motif, a modified form of the DRY motif, has been implicated in direct interactions of PLD with the heterotrimeric protein Gα subunit (Zhao and Wang, 2004). CsPLDγ1 shares 71% identity with Arabidopsis PLDγ1 and has a calcium-binding C2 and two catalytic HKD domains. An ‘ERF’ motif was also identified within its catalytic region along with a modified hydrophobic region immediately downstream of this site. CsPLDα1 and CsPLDγ1 sequences were deposited in GenBank and assigned accession numbers, EU340031 and EU340032, respectively.

Ethephon induces CsPLDγ1 and decreases CsPLDα1 expression in citrus leaves and LAZ

Ethephon application to whole citrus tree canopy sections induced rapid defoliation. Leaf abscission began 24 h after application (Fig. 1A). The rate of abscission increased from 24 h to day 3 and resulted in >70% leaf drop by day 5 after treatment. Ethephon induced a significantly greater leaf drop when compared with controls, beginning from day 1 to the end of the experiment.

Fig. 1.

(A) Effect of ethephon on leaf abscission in citrus. Cumulative leaf drop (%) induced by 600 mg l−1 ethephon in Citrus sinensis cv. ‘Valencia’. Asterisk indicates significant difference between ethephon treatment and controls on the same day after treatment (Students t test P <0.01). Vertical lines through markers depict SE mean (n=5). Absence of SE lines indicates marker larger than SE. (B–E) Changes in expression within the leaf blade and LAZ of CsPLDα1 (B, D, respectively) and CsPLDγ1 (C, E, respectively), following ethephon application (600 mg l−1) were measured using quantitative RT-PCR. Vertical lines through bars depict SE mean (n ≥4). In each graph, significant differences between control and ethephon-induced expression within each sampled time after application are depicted by **or * (P <0.01 or P <0.05, respectively; Student's t test). NS, not significant; ND, not determined.

In comparison to the untreated control, ethephon decreased CsPLDα1 expression in the leaf blade and LAZ by 30% and 36%, respectively, 6 h after application (Fig. 1B, D). Reduction in expression was also observed at 24 h and 48 h after treatment. In contrast, CsPLDγ1 expression was rapidly induced (Fig. 1C, E). CsPLDγ1 expression in leaf blade tissue increased by 3-fold within 3 h and >5-fold within 6 h of ethephon application (Fig. 1C). Increased expression was also observed 24 h and 48 h after application. A 4-fold increase in CsPLDγ1 expression occurred in the LAZ 6 h after application but no significant difference was noted at 48 h after application (Fig. 1E). These data indicate a rapid induction of CsPLDγ1 and suppression of CsPLDα1 expression in the leaf blade and LAZ by ethephon, and that trends in expression were similar in leaf blade and LAZ.

CsPLDγ1 expression in citrus fruit flavedo increases during abscission-agent induced fruit abscission

The abscission agents ethephon and CMNP significantly decreased mature fruit detachment force within 4 d after application (Fig. 2A). Ethephon decreased fruit detachment force to 59% and 80% of the control at 4 d and 7 d after application, respectively. CMNP application was more effective in loosening mature fruit. FDF was reduced to 42% that of the control by 4 d after application and did not change thereafter.

Fig. 2.

(A) Effect of abscission agents on fruit detachment force (FDF, in kg force [KgF]). Ethephon (600 mg l−1) and CMNP (250 mg l−1) were applied to mature citrus trees and FDF was measured at the indicated times (n=3). (B, C) Changes in CsPLDα1 (B) and CsPLDγ1 (C) expression in the flavedo, following ethephon (600 mg l−1) and CMNP (250 mg l−1) application were measured using quantitative RT-PCR. Vertical lines through bars depict SE mean (n=3). The same letter over the bars indicates that means are not significantly different. Mean separation analysis was performed using Fisher's LSD (P <0.05) after ANOVA. NS, not significant.

Ethephon and CMNP did not alter CsPLDα1 expression in fruit flavedo (Fig. 2B). Ethephon application transiently increased CsPLDγ1 expression by 60% at 24 h after application, but the increase was not statistically significant (Fig. 2C; Fisher's LSD 0.05). Ethephon did not have any effect on CsPLDγ1 expression at the later stages (48 h and 96 h). By contrast, CMNP application increased CsPLDγ1 expression by almost 2-fold within 24 h after application. CMNP application also resulted in a sustained increase in CsPLDγ1 expression by ∼2.5-fold at 48 h and at 96 h after application. The above data indicate differential regulation of PLD gene expression by abscission agents in the fruit flavedo.

Diurnal oscillation of PLD gene expression in citrus leaves and fruit

During our analysis of PLD gene expression in the leaf blade, an increase in expression within control samples was noted at 6 h after application (sampling time: 16.00 h; Fig. 1B, C). Hence, diurnal oscillation in PLD gene expression was investigated. CsPLDα1 and CsPLDγ1 exhibited diurnal oscillation in gene expression within the leaf blade (Fig. 3A). CsPLDα1 and CsPLDγ1 expression were low early in the light period (09.00 h), and increased as light duration increased, reaching a maximum around 17.00 h, and then declined during the dark period. The maximum/minimum change in CsPLDα1 expression was about 1.7-fold. CsPLDγ1 exhibited a similar phase but greater amplitude in the diurnal rhythm of expression with maximum/minimum change of almost 5-fold. These data indicate that PLD gene expression is diurnally regulated in citrus leaves.

Fig. 3.

Diurnal rhythms in CsPLDα1 and CsPLDγ1 expression. Natural diurnal fluctuations in CsPLDα1 and CsPLDγ1 gene expression were determined in mature citrus leaf blade (A) and fruit flavedo (B) at the times indicated using quantitative RT-PCR. Vertical lines through mean markers depict SE mean (n=4). Absence of SE lines indicates marker larger than SE. White and black bars under the graph depict the timing of the light/dark cycles.

In contrast to the expression pattern in the leaf blade, expression of CsPLDα1 did not oscillate diurnally in the fruit flavedo (Fig. 3B). However, CsPLDγ1 expression oscillated with a diurnal pattern. CsPLDγ1 expression was lowest during the early part of the day (09.00 h), increased with the progression of the day, and reached peak expression during midday (13.00 h and 17.00 h). Peak expression of CsPLDγ1 was 3-fold higher than that early in the day.

PLD gene expression in citrus leaves is regulated by light and the circadian clock

The role of temperature, light, and the circadian clock in entraining PLD diurnal gene expression was studied in citrus leaves. Entrainment of potted citrus plants at constant temperature (24 °C) and 12/12 h light/dark (LD) for 1 week did not abolish diurnal rhythms in CsPLDα1 or CsPLDγ1 expression (Fig. 4A, B; the initial two time-of-day samples), but the magnitude of CsPLDγ1 expression increased (compare with Fig. 3A). The transfer of constant temperature and LD-entrained plants to continuous light (LL) resulted in the loss of oscillation in CsPLDα1 expression (Fig. 4A). By contrast, oscillation of CsPLDγ1 expression persisted after transfer to LL. However, peak CsPLDγ1 expression was lower and a phase change occurred, resulting in earlier peak expression at 14.00 h followed by a decline in expression.

Fig. 4.

Light dependence of CsPLDα1 and CsPLDγ1 expression in citrus leaves. Potted ‘Valencia’ citrus trees were entrained under 12/12 h: light/dark cycles (LD) and constant temperature (24 °C) conditions for 1 week. Trees were then transferred to either (A) continuous light (LL) or (B) continuous dark (DD), and CsPLDα1 and CsPLDγ1 expression in the leaf blade was analysed by quantitative RT-PCR at the indicated times. Vertical lines through markers depict SE mean (n=4). The absence of SE lines indicates a marker larger than the SE. Short white bars under the graph depict the end of the entrainment period. Long white and black bars depict the length of LL (A) and DD (B) treatment, respectively.

Transfer of constant temperature- and LD-entrained plants to continuous dark (DD) resulted in a loss of oscillation in CsPLDα1 expression (Fig. 4B). CsPLDγ1 expression in leaves exhibited diurnal oscillation on day 1 after transfer to DD (2–2.5-fold increase during midday), but remained at basal levels during prolonged exposure to DD. These data suggest that light strongly influences CsPLDα1 expression, and that CsPLDγ1 expression is regulated by light as well as the circadian clock.

Diurnal oscillation in PLD gene expression in citrus leaves is partially dependent on ethylene signalling

As in Arabidopsis seedlings (Thain ), ethylene emission from citrus leaves occurs in a diurnal pattern with peak emission during the day (A Malladi et al., unpublished data). It is hypothesized that diurnal changes in ethylene levels and signalling may modulate diurnal PLD expression. The ethylene perception inhibitor 1-MCP was utilized to reduce ethylene perception in citrus leaves. Application of 1-MCP did not prevent oscillation in CsPLDα1 and CsPLDγ1 gene expression (Fig. 5A and B, respectively); however, the amplitude of oscillation decreased. These data indicate that ethylene perception and signalling modulate the magnitude of diurnal oscillation in CsPLDα1 and CsPLDγ1 expression in citrus leaves. Diurnal oscillation in CsPLDα1 expression (Fig. 5A) was higher than in Fig. 3A, possibly as this experiment was performed in the field.

Fig. 5.

Effect of 1-MCP on diurnal fluctuation in PLD expression in citrus leaves. 1-MCP (5 mM) was applied to mature field-grown citrus trees at 10.00 h and its effect on diurnal CsPLDα1 (A) and CsPLDγ1 (B) expression in the leaf blade was measured using quantitative RT-PCR at the indicated times. Vertical lines through the markers depict the SE mean (n=4). Absence of SE lines indicates a marker larger than the SE.

Sensitivity to abscission agent-induced leaf and fruit abscission increases during midday

The relationship between diurnal oscillation in PLD gene expression and leaf and fruit abscission sensitivity was investigated by the application of abscission agents during minimal (early-day) and rising (midday) CsPLDγ1 expression. Early-day application of ethephon (09.00 h) induced 47% leaf abscission by day 10, with the majority of leaf drop occurring by day 4 (40%; Fig. 6A). Midday (13.00 h) application increased total leaf abscission to almost 60% by day 10, with the majority of leaf drop occurring by day 3 (48%). These data indicate greater sensitivity to ethephon when applied during rising CsPLDγ1 expression.

Fig. 6.

Effect of time of application of abscission agents on citrus leaf abscission and fruit detachment force (FDF, in kg force [KgF]). (A) Cumulative leaf drop (%) as affected by time of day of ethephon application. Water and ethephon (400 mg l−1) were applied at either 09.00 h or 13.00 h. Vertical lines through markers depict the SE mean (n=4). Absence of SE lines indicates a marker larger than the SE. An asterisk indicates significant difference between the 09.00 h and 13.00 h applications of ethephon. Mean separation analysis was performed using Fisher's LSD (P <0.05) after ANOVA. (B) FDF as affected by ethephon (600 mg l−1) and CMNP (250 mg l−1) applications. Ethephon or CMNP were applied at 09.00 h and 13.00 h. FDF was measured 4 d after application. Different letters on the mean marker indicates significant difference between means within the group. Mean separation analysis was performed using Fisher's LSD (P <0.05) after ANOVA.

The application of ethephon at 09.00 h decreased FDF to 50% of the control while the midday (13.00 h) application decreased it to 42% of the control (Fig. 6B). Application of CMNP during midday (13.00 h) had a greater effect on reducing FDF. While early-day (09.00 h) application of CMNP decreased FDF to 42% of the control, midday application of CMNP decreased FDF to less than 9% of the control. The above data indicate that sensitivity to abscission-agent induced leaf and fruit abscission increased during midday.

Discussion

Initial signals during ethylene-induced leaf abscission are thought to be generated within the leaf blade, while the trigger for citrus mature fruit abscission is generated within the fruit flavedo (Beyer, 1975; Alferez ). Hence, our analysis of citrus PLD expression was largely focused in the above tissues. Expression of CsPLDγ1 was rapidly induced in ethephon-treated citrus leaf blades while that of CsPLDα1 decreased, indicating an important role for PLDs during the early responses to induced leaf abscission. The G-protein-related signalling pathway was implicated in modulating ethephon-induced leaf abscission in citrus as the application of G-protein receptor agonists blocked ethephon-induced leaf abscission (Yuan ). Ethylene may promote leaf abscission by modulating interactions within the heterotrimeric G-protein complex (Gα, Gβ, and Gγ), thereby affecting G-protein signalling. Rapid and differential changes in CsPLDα1 and CsPLDγ1 expression by ethephon may facilitate this process directly through interactions with the Gα subunit via the putative ‘DRY’ motif or indirectly through effects on Gβ and Gγ subunits of the heterotrimeric G-protein complex.

Differential regulation of PLD expression was associated with ethephon-induced, early leaf abscission responses. Diversity in products and/or physiological outcomes generated by different PLD isoforms may account for the seemingly opposing functions of CsPLDα1 and CsPLDγ1. Opposing roles for different PLDs have previously been reported in Arabidopsis (Wang ; Welti ; Li ; Wang, 2005).

Notwithstanding the numerical increase in expression, ethephon application did not significantly alter CsPLDα1 and CsPLDγ1 expression in fruit flavedo. Hence, a PLD-independent mechanism may regulate ethephon-induced reduction in FDF. These data indicate that diverse mechanisms regulate ethephon-induced citrus leaf and fruit abscission responses. Such mechanisms may include differential ethylene sensitivity between organs. The ethylene perception inhibitor, 1-MCP, modulates ethephon-induced leaf abscission responses but not FDF responses (Pozo ). Also, G-protein agonists modulate ethephon-induced leaf abscission responses but have little effect on fruit abscission (Burns ).

CMNP is a fruit-specific abscission agent that elicits unique physiological reactions when compared with ethephon or ethylene (Li ; Alferez ). Although mechanisms through which CMNP specifically induces citrus mature fruit abscission remain unclear, chemical wounding, transient alteration in membrane permeability, and reduction in ATP levels are important factors contributing to the acceleration of abscission. The application of CMNP rapidly increased and sustained CsPLDγ1 expression in citrus fruit flavedo and this was correlated with higher and consistent reduction in FDF. Rapid elevation of PA levels occurs in several plants upon wounding and this may be involved in wound-induced and phospholipid-based signal transduction mechanisms (Lee ). A similar, PA-dependent wound-induced signalling mechanism may operate in triggering abscission responses to CMNP. An increase in PA levels facilitated by CsPLDγ1 may trigger downstream signalling mechanisms leading to fruit abscission. In addition, modification of membrane phospholipid composition by CsPLDγ1 activity and PA may further contribute to the acceleration of fruit abscission responses. Mechanical wounding of flavedo induces fruit abscission in citrus (Kostenyuk and Burns, 2004). It may be that CsPLDγ1 is similarly involved in signalling mechanical wounding-induced fruit abscission responses. Together, the above data suggest that rapid and sustained increase in CsPLDγ1 expression may constitute a key mechanism that mediates abscission responses to diverse abscission-agents in different organs of citrus. CsPLDγ1 may have dual functions in abscission signalling: to facilitate ethylene and G-protein interaction-dependent modulation of leaf abscission, and to mediate wound-induced fruit abscission responses.

Additional functions for phospholipid-signalling in facilitating abscission cannot be excluded. PA has been shown to interact directly with and inhibit the activity of CTR1 (constitutive triple response 1) a negative regulator of ethylene signalling (Testerink ). Hence, PA generated by CsPLDγ1 activity may directly mediate ethylene-induced leaf abscission responses through the inhibition of CTR1 and the activation of ethylene signalling. In addition, PA may also modify auxin transport, another important regulator of ethylene-induced abscission (Patterson, 2001; Taylor and Whitelaw, 2001), through interaction with RCN1 (roots curl in NPA 1), a regulator of auxin transport (Testerink ; Muday ). Also, the Arabidopsis PLD isoform, PLDζ2, regulates auxin responses through the modification of auxin transport (Li and Xue, 2007). A decrease in auxin transport from the leaf blade or fruit flavedo to the AZ, facilitated by an increased expression of CsPLDγ1, may alter the critical auxin–ethylene balance at the AZ and promote abscission. The above possible mechanisms involved in phospholipid signalling-mediated abscission responses warrant further investigation.

Analysis of PLD expression in citrus leaves revealed diurnal oscillation in transcript abundance in this gene family. Expression of CsPLDα1 and CsPLDγ1 was highest at midday but declined later in the night, reaching the lowest levels early in the day. CsPLDα1 expression was not diurnally regulated in fruit flavedo; however, CsPLDγ1 expression followed a similar diurnal pattern to that seen in leaves, but with lower amplitude. Diurnal regulation of PLD gene expression in leaves may have implications for several physiological functions including diurnal regulation of stomatal opening and closure as PLDα plays a role in this process (Mishra ; Hong ).

Factors that contributed to diurnal oscillation in PLD gene expression in citrus leaves were investigated. While diurnal CsPLDα1 expression was regulated by light, oscillation of CsPLDγ1 expression was under light as well as circadian control. These data indicate that light- and circadian clock-dependent mechanisms modulate PLD expression in citrus leaves. Interestingly, studies in etiolated oat seedlings indicate that PLD activity is red-light dependent (Park ; Kabachevskaya ). Furthermore, application of 1-MCP reduced the magnitude of diurnal oscillation in CsPLDα1 and CsPLDγ1 gene expression. Hence, diurnal oscillation in PLD gene expression is, in part, dependent upon ethylene perception. Phospholipid-signal generation may therefore be regulated by light, the circadian clock, and ethylene-signalling and such diurnal regulation of phospholipid signals may mediate physiological responses including abscission.

Response to abscission agent-induced abscission appears to be modulated depending upon the time of day (Pozo ). Preliminary studies suggest that changes in leaf abscission response during different times of the day persist under constant temperature conditions (A Malladi et al., unpublished results). Fruit held under constant temperature probably respond in a similar manner, but this remains to be tested. These data indicate diurnal regulation of abscission in citrus. Peak sensitivity of abscission during the day corresponded with rising diurnal CsPLDγ1 expression in citrus leaf blade and flavedo. Assuming CsPLDγ1 enzyme activity closely follows diurnal changes in CsPLDγ1 gene expression, changes in PA and related lipid product levels in the leaf blade and flavedo during midday may more effectively trigger and propagate signalling mechanisms involved in facilitating abscission responses. Such mechanisms may also be diurnally regulated. Analysis of diurnal changes in PLD activity and/or PA levels should provide novel insights into their roles in mediating changes in abscission sensitivity.

Supplementary data

The following supplementary data for this article are available at JXB online.

Table S1. List of primers used for cloning CsPLDα1 and CsPLDγ1 from Citrus sinensis cv. ‘Valencia’.

Table S2. List of primers used for quantitative RT-PCR analysis (5′–3′).