Comparative proteomic analysis of cauliflower under high temperature and flooding stresses.

High-temperature and waterlogging are major abiotic stresses that affect the yield and quality of cauliflower. Cauliflower cultivars 'H41' and 'H69' are tolerant to high temperature and flooding, respectively; however, 'H71' is sensitive to both stresses. The objectives of this study were to identify the proteins that were differentially regulated and the physiological changes that occurred during different time periods in 'H41', 'H69', and 'H71' when responding to treatments of flooding, 40 °C, and both stresses combined. Changes in the leaf proteome were analyzed by matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF-MS) and identified by Mascot peptide mass fingerprint (PMF) and database searching. Stress treatments caused significant reductions in electrolyte leakage, chlorophyll fluorescence Fv/Fm, chlorophyll content, and water potential as stress times were prolonged. By the comparative proteomic analysis, 85 protein peaks that were differentially expressed in response to combination treatments at 0, 6, and 24 h, 69 (33 in 'H41', 29 in 'H69', and 9 in 'H71') were identified, of which were cultivar specific. Differentially regulated proteins predominantly functioned in photosynthesis and to a lesser extent in energy metabolism, cellular homeostasis, transcription and translation, signal transduction, and protein biosynthesis. This is the first report that utilizes proteomics to discover changes in the protein expression profile of cauliflower in response to heat and flooding.

Introduction

Cauliflower (Brassica oleracea var. botrytis), a member of the Brassicaceae family, is an economically and nutritionally important cole crop. The optimum mean temperature range for growing cauliflower is 18–25 °C. High temperatures cause cauliflower to form uneven and loose heads, puffy buds, yellow eyes and leaves in the head, narrow leaves, reduced leaf growth, delayed initiation of heading, and increased petiole-to-blade ratio, all of which lower the quality of or even make the heads unmarketable (Wurr et al., 1996, Nowbuth and Pearson, 1998). However, heat-tolerant cultivars are able to form heads at mean temperatures higher than 25 °C. To expand the area of food production, crops are grown under stressful environments that likely lead to lower yields. The main contributing factors in the reduction of yield and quality in these areas are variations in climatic conditions such as flooding caused by rains (Lin et al., 2010). Heavy rainstorms and standing water can leave soils saturated for days before draining, making waterlogging a problem in many parts of the world. Air pockets in the soil become filled with water during saturation, thus creating hypoxic conditions followed by anoxia. Flooding causes oxygen starvation, which is a consequence of the relatively slow diffusion of gases in water and from oxygen consumption by plant roots (Takeshi and Julia, 2004). Rapid leaf chlorosis is seen when cauliflower is subjected to waterlogging (Shih et al., 2013). Flooding from large rainfall events is a major risk to fresh-market cauliflower production in Taiwan, and most cauliflower cultivars are unable to tolerate flooding during typhoon-caused heavy summer rains. Waterlogging and increasing temperatures associated with global warming are a growing concern, as they limit plant growth and productivity, especially in temperate species. Over the past decade in the tropics and subtropics, there have been considerable increases in production because of the availability of new, tropically adapted cultivars, resulting in increased farmer incomes.

In the field, the co-occurrence of several abiotic stresses rather than individual stresses are most damaging to crop production (Mittler, 2006). Many physiological changes occur during high temperature and flood stressing that result in increased heat and flood (HF) tolerance. The physiological mechanisms of HF tolerance are extensively studied in various plant species; however, our current understanding of the molecular biology of acquired tolerances to high temperature and waterlogging are still limited, and relatively little is known about proteins that are critical for controlling these dual mechanisms (Wahid et al., 2007). Understanding these processes requires the identification and analysis of major proteins that underlie stress-regulatory networks. Plant adaptations to environmental stresses depend on the activation of cascades of molecular networks involved in stress perception, signal transduction, and the expression of stress-related proteins. Knowledge of HF-responsive proteins is critical for further understanding of the molecular mechanisms of stress tolerance. Plants respond to stress in part by modulating protein regulation, which eventually leads to restoration of cellular homeostasis, neutralization of toxins, and recovery of growth. However, there are no reports on the effect of HF stresses on the functioning of cauliflower. Proteins are the direct effectors of the response of plants to stress. They not only include enzymes that mediate changes in the levels of metabolites but also serve as components of the transcription and translation machinery.

In recent years, methods for analyzing the proteome have advanced considerably, and together with emerging sequence information in crops, plant proteomics has become increasingly useful for understanding gene functioning and networks in response to environmental stimuli. Proteomics is a powerful tool for the quantitative analyses of different biochemical pathways including plant stress-related responses. Comparative proteomic analyses of plants subjected to specific stress conditions allow the exploration of various defense-related mechanisms. For instance, proteomics approaches for the comparative analysis of protein abundance between untreated and stress-treated or tolerant and intolerant rice plants have greatly facilitated the study of plant cellular stress responses (Komatsu et al., 2003). Therefore, identifying novel proteins and studying their differential display patterns in response to HF stresses will provide the molecular and physiological bases for improving tolerance to HF by cauliflowers. Understanding the basis of HF stress signaling and tolerance mechanisms in cauliflowers is required for engineering local cauliflower genotypes that are more tolerant to heat and flood stressing. This goal can be achieved by deciphering the physiological and proteomic responses of cauliflower genotypes, heat-tolerant ‘H41’, flood-tolerant ‘H69’, and HF-sensitive ‘H71’ to flood stress at 40 °C. The objective of this study was to identify the leaf proteins that are differentially regulated in response to HF stress using a ProteomeLab PF-2D aided by improved databases. Our hypothesis was that both HF stresses trigger plant responses that result in quantifiable changes in the proteins of ‘H41’, ‘H69’, and ‘H71’. Proteomic characterization and functional analysis should facilitate a better understanding of the HF response mechanisms in Brassica so that effective strategies for the genetic improvement of HF-tolerant plant cultivars can be established. To the best of our knowledge, there are no published reports addressing the identification of HF-responsive proteins in Brassica using a proteomic approach.

Materials and methods

Plant materials, culturing, and heat- and flood-stress treatments

Seeds of cauliflower (B. oleracea var. botrytis) ‘H41’, ‘H69’, and ‘H71’ were obtained from Chin-Long Seed Co. (Tainan, Taiwan). ‘H41’ is a heat-tolerant cultivar used especially in warm-subtropical regions such as southern Taiwan, where average day temperatures reach as high as 40 °C during the summer (June–August). ‘H69’ is a popular cultivar grown in Taiwan and is flood-tolerant during the rainy season, particularly in rain-fed lowlands. ‘H71’ is a heat- and flood-sensitive cultivar, and mostly grown during winter in Taiwan due to its cold tolerance. Seeds were sterilized with 1.5% (v/v) sodium hypochlorite, rinsed with distilled-deionized (dd) H2O, and sown in a commercial potting soil mixture, and germinated seedlings were transplanted into 12.7-cm diameter plastic pots and raised in a growth chamber under 300 μmol m−2  s−1 light with a 14 h photoperiod and 22/18 °C day/night temperatures at a relative humidity (RH) of 80%. Plants were watered three times a week and fertilizer (17:36:39, N:P:K) applied once a week to maintain optimal growth for 30 days before the imposition of stress treatments. Pots containing ‘H41’ and ‘H69’ plants were subjected to four treatments: non-flooding at 20 °C (NFC, as control), flooding at 20 °C (FC, control temperature), non-flooding at 40 °C (NFH, high temperature), and flooding at 40 °C (FH), for periods of 0, 6, 12, 24, 48, 72, and 96 h in four growth chambers having a 14 h photoperiod at 300 μmol m−2  s−1 radiation and 80% RH (Lin et al., 2010). In the case of flooding treatments, pots were randomly placed in 28 cm × 14 cm × 14 cm plastic buckets and subjected to flooding by filling the buckets with tap water to 5 cm above the soil surface. Pots were removed from the buckets at different times following flooding, and plants were removed and their leaves from each plant clipped, frozen in liquid nitrogen, and stored at −80 °C in an ultra-freezer until used. Three replicates of each time period for the four treatments were randomly placed in a growth chamber. The experiment was performed twice independently in a randomized design for growth environment, sampling day, and physiological analyses.

Determination of electrolyte leakage (EL), chlorophyll fluorescence (CF), chlorophyll content (CC), and water potential (WP)

Cell membrane stability was estimated by measuring leaf ion leakage according to the method of Huang and Guo (2005). Leaves were excised and immersed in 15 ml of distilled water in test tubes overnight at room temperature. The initial conductivity of the water was determined using a conductivity meter (model CDM 210, Radiometer, Cedex, France). Tubes were placed in boiling water for 15 min and then cooled to room temperature, and conductivity was again determined. The relative EL (%) was calculated as the ratio of conductivity before boiling to that after boiling. CF components were quantified with a portable modulated fluorometer (Mini-Pam Photosynthesis Yield Analyzer, Walz, Effeltrich, Germany). The measurement of variable fluorescence to (Fv)/maximum fluorescence level in light-adapted leaves (Fv/Fm) was previously described (Lin et al., 2007). Relative CC per unit leaf area was determined using a SPAD (Soil Plant Analysis Development) analyzer (SPAD-502 Chlorophyll Meter, Konica Minolta, Tokyo, Japan). WP (in bars) was measured on the third leaf from the top of each plant using a pressure chamber (Plant Water System, Skye SKPM 1400, Tokyo, Japan) (Sairam et al., 1998).

The data shown in Table 1, Table 2, Table 3, Table 4 represent the means of at least two independent sets of experiments with similar results. Measurements of physiological parameters were analyzed by a three-factor completely randomized ANOVA that compared cultivars, treatments, and time periods. For significant values, means were separated by a least significant difference (LSD) test at p  ≤ 0.05 using PC SAS 8.2 (SAS Institute, Cary, NC, USA).

Table 1

Effect of stress treatments on Fv/Fm value in cauliflower ‘H41’, ‘H69’, and ‘H71’.

Cultivar	Treatment	Duration (h)
		0	6	12	24	48	72	96
H41	NFC	0.835Aa	0.851Aa	0.846Aa	0.831Aa	0.829Aa	0.810Aa	0.808Aa
	FC	0.835Aa	0.753Bb	0.756Bb	0.745Bb	0.654Cc	0.587Cbc	0.304Db
	NFH	0.835Aa	0.852Aa	0.851Aa	0.821Aa	0.797Bab	0.725Ba	0.701Ba
	FH	0.835Aa	0.788Bb	0.781Bb	0.775Bb	0.521Dc	0.463Ec	0.002Hc
H69	NFC	0.880Aa	0.850Aa	0.842Aa	0.839Aa	0.805Ab	0.822Aa	0.826Aa
	FC	0.880Aa	0.845Aa	0.850Aa	0.825Aa	0.758Bb	0.747Ba	0.713Ba
	NFH	0.880Aa	0.775Bb	0.763Bb	0.750Bb	0.668Cc	0.655Cb	0.243Eb
	FH	0.880Aa	0.778Bb	0.770Bb	0.597CDc	0.487Ed	0.176Gd	0.000Hc
H71	NFC	0.864Aa	0.852Aa	0.844Aa	0.846Aa	0.831Aa	0.813Aa	0.833Aa
	FC	0.864Aa	0.750Bb	0.734Bb	0.714Bb	0.704BCb	0.558Cb	0.230Eb
	NFH	0.864Aa	0.768Bb	0.775Bb	0.701Bb	0.692Cb	0.576Db	0.001Fc
	FH	0.864Aa	0.718BCb	0.670Cc	0.557Dc	0.347Fd	0.128Gd	0.000Hc

Among seven time periods for each treatment (row), means with the same capital letter do not significantly differ. Among four treatments for each time-period (column), means with the same lowercase letter do not significantly differ. NFC, non-flooding at control temperature 20 °C. FC, flooding at 20 °C. NFH, non-flooding at high temperature 40 °C. FH, flooding at 40 °C.

Table 2

Effect of stress treatments on chlorophyll content SPAD value in cauliflower ‘H41’, ‘H69’, and ‘H71’.

Variety	Treatment	Duration (h)
		0	6	12	24	48	72	96
H41	NFC	49.3Cb	48.1Cc	47.2Cc	47.4Cc	47.4Ccd	47.2Cbc	47.5Cb
	FC	49.3Cb	45.8Dd	45.1Dd	42.7Ed	40.4Ee	40.8Ed	34.2Fd
	NFH	49.3Cb	48.7Cc	47.6CDc	46.9Dc	45.9Dd	44.3Ec	41.4Ec
	FH	49.3Cb	46.9Dc	41.53Ee	41.4Ed	35.3Ff	33.4Fe	25.8Ge
H69	NFC	61.0Aa	61.6Aa	61.9Aa	59.6ABa	59.1ABa	56.2Ba	56.4Ba
	FC	61.0Aa	62.3Aa	62.6Aa	62.4Aa	56.5Bab	56.8Ba	53.9BCa
	NFH	61.0Aa	55.4Bb	54.7Bb	54.6Bb	51.7BCb	47.2CDbc	32.3Fd
	FH	61.0Aa	53.5Bb	53.3Bbc	49.2Cc	49.3Cc	39.1Ed	28.6Ge
H71	NFC	56.4Bab	56.0Bab	57.7Bb	58.2Ba	55.7Bb	55.3Ba	56.1Ba
	FC	56.4Bab	56.1Bab	50.1Cc	46.6Dcd	46.8Dd	44.4DEc	34.1Fd
	NFH	56.4Bab	50.2Cbc	49.7Cc	42.7Ed	41.9Ee	32.4Fe	28.7Ge
	FH	56.4Bab	46.5Dc	43.2DEde	35.0Fe	26.2Gg	20.4Hf	16.5If

Among seven time periods for each treatment (row), means with the same capital letter do not significantly differ. Among four treatments for each time period (column), means with the same lowercase letter do not significantly differ. NFC, non-flooding at control temperature 20 °C. FC, flooding at 20 °C. NFH, non-flooding at high temperature 40 °C. FH, flooding at 40 °C.

Table 3

Effect of stress treatments on electrolyte leakage (%) in cauliflower ‘H41’, ‘H69’, and ‘H71’.

Variety	Treatment	Duration (h)
		0	6	12	24	48	72	96
H41	NFC	10.1Aa	10.4Ade	11.6Ae	12.1Ae	13.6Af	13.7Af	14.7Ad
	FC	10.1Ia	19.1FGb	22.6Fc	35.2Ec	61.1Cc	70.6Bd	89.9Aab
	NFH	10.1KLa	11.4Kd	20.5IJc	21.4Id	33.2He	43.7Ge	60.0Ec
	FH	10.1Ia	18.8Hb	29.3FGa	57.3Da	74.7Ca	86.1Bab	96.2Aa
H69	NFC	7.8Aa	8.1Ae	8.5Af	10.0Ae	10.4Af	12.4Af	12.0Ad
	FC	7.8Ia	11.7HId	13.3Hde	20.9Fd	32.9Ee	46.0De	65.7Cc
	NFH	7.8La	15.8Jc	16.0Jd	21.6Id	70.3Dab	82.5BCb	91.5Aa
	FH	7.8Ia	25.4Ga	32.4Fa	50.3Eb	68.7Cb	90.8Ba	95.7Aa
H71	NFC	7.8Aa	10.4Ade	11.0Ae	11.1Ae	12.1Af	13.9Af	12.3Ad
	FC	7.8Ia	14.7Hc	16.6Gd	21.5Fd	60.4Cc	72.3Bcd	83.3Ab
	NFH	7.8La	13.6Kc	25.5Ib	35.1Hc	54.0Fd	76.4Cc	87.0Ab
	FH	7.8Ia	17.1Hbc	26.9Gb	48.5Eb	64.1CDc	96.5Aa	97.4Aa

Among seven time periods for each treatment (row), means with the same capital letter do not significantly differ. Among four treatments for each time period (column), means with the same lowercase letter do not significantly differ. NFC, non-flooding at control temperature 20 °C. FC, flooding at 20 °C. NFH, non-flooding at high temperature 40 °C. FH, flooding at 40 °C.

Table 4

Effect of stress treatments on water potential (-MPa) in cauliflower ‘H41’, ‘H69’, and ‘H71’.

Variety	Treatment	Duration (h)
		0	6	12	24	48	72	96
H41	NFC	0.69Bb	0.68Bd	0.66Be	0.70ABe	0.71ABd	0.73ABc	0.72ABc
	FC	0.69Fb	0.93Eb	1.12Dc	1.33Cc	1.76Bbc	2.17Aa	–
	NFH	0.69Gb	0.77FGd	0.84Fd	1.34Dc	1.52Cc	1.87Bb	1.92Bb
	FH	0.69Eb	1.51BCa	1.63Bb	1.76Bb	2.15Aa	–	–
H69	NFC	0.80Aab	0.81Ac	0.82Ad	0.79Ae	0.83Ad	0.80Ac	0.78Ac
	FC	0.80Eab	1.09Dab	1.38Cb	1.40Cc	1.76Bbc	2.15Aa	2.26Aa
	NFH	0.80Fab	0.96EFb	1.14Ec	1.57Cbc	2.13Aa	–	–
	FH	0.80DEab	1.34Ca	1.47Cb	2.17Aa	2.25Aa	–	–
H71	NFC	0.95Aa	0.95Ab	0.96Acd	0.92Ad	0.92Acd	0.96Abc	0.91Ac
	FC	0.95Ea	1.09DEab	1.48BCb	1.74Bb	1.90Bb	2.20Aa	2.41Aa
	NFH	0.95EFa	1.24DEa	1.38Db	1.75BCb	2.27Fa	–	–
	FH	0.95Da	1.51BCa	2.23Aa	–	–	–	–

Among seven time periods for each treatment (row), means with the same capital letter do not significantly differ. Among four treatments for each time period (column), means with the same lowercase letter do not significantly differ. NFC, non-flooding at control temperature 20 °C. FC, flooding at 20 °C. NFH, non-flooding at high temperature 40 °C. FH, flooding at 40 °C. –, not measureable due to dry leaves.

Protein extraction and purification for 2-D liquid chromatography

Because the negative effects of flood stressing on HF-sensitive ‘H71’ at 40 °C were observed after 48 h of stress treatments (see Section 4), ‘H41’, ‘H69’, and ‘H71’ plant leaves subjected to HF conditions for 0, 6, and 24 h were used for subsequent proteomics analysis. Proteins were extracted according to a previously published method (Yan et al., 2013) with some modifications. Briefly, two grams of plant leaves were ground in liquid nitrogen and crude protein extracts solubilized in 7 ml of extraction buffer containing 0.5% SDS (sodium dodecyl sulfate), 5% β-mercaptoethanol, 10% glycerol, 0.2% polyvinylpyrrolidone, and 50 mM Tris–HCl, pH 8. After 30 min of incubation at 4 °C, samples were centrifuged for 30 min at 15,000 ×  g at 4 °C; these two steps were done twice. The extraction buffer was further replaced by 25 ml of start buffer (50 mM Tris–HCl, pH 8) with a PD-10 column (GE Healthcare) according to manufacturer (ProteomeLab PF-2D kit, Beckman Coulter) protocols. Protein concentrations in the samples were determined using the Bradford assay (Bio-Rad, Hercules, CA, USA) and bovine serum albumin (BSA) was used to generate a standard curve.

A 200 μg aliquot of the total protein sample was passed through a 0.45 μm filter before injection into the chromatography column. The High Performance ChromatoFocusing (HPCF) column was treated according to manufacturer instructions (ProteomeLab PF-2D Protein Fractionation System, Beckman Coulter, Fullerton, CA, USA). Briefly, the column was washed with 10 volumes of water at a flow rate of 0.2 ml/min for 45 min and then equilibrated with 30 volumes of start buffer for 130 min at 0.2 ml/min. After equilibration, each sample was introduced with a manual injector into the column and absorbance of the column effluent was monitored at 280 nm. In the first dimension, proteins were bound to a strong anion exchanger and eluted with a continuously decreasing pH from 8.0 to 4.0. Fractions were collected at pH intervals of 0.3 in a 96 deepwell plate. Proteins eluted in the gradient were then separated in the second dimension using High Performance Reversed Phase (HPRP) chromatography. Fractions were separated from the second dimension with an RP-HPLC column using two solvents: 0.1% trifluroracetic acid (TFA) in HPLC water (Solvent A) and 0.08% TFA in acetonitrile (ACN) (Solvent B) (Lee et al., 2008). Separation was performed at 50 °C with a flow rate of 0.75 ml/min and protein fractions were detected by UV absorbance at 214 nm. Equilibration was achieved with Solvent A for 10 min followed by Solvent B for 5 min prior to each injection. From the selected first dimension fractions, 0.2 ml were injected, run for 2 min, and the column eluted with a linear gradient of 0–100% Solvent B for 25 min. Thereafter, Solvent B was continued for 5 min, followed by re-equilibration with 100% Solvent A for 10 min (Irar et al., 2010). Fractions from the second dimension were analyzed with Karat™ Version 7.0 (Beckman Coulter). Protein extraction and two-dimensional gel electrophoresis (2-DE) were carried out on three independent technical replicas of the bulk samples.

Protein identification by MALDI-TOF mass spectrometry (MS) and database search

Proteins were in-gel digested with trypsin in the automatic Investigator ProGest robot of Genomic Solutions. Briefly, excised gel bands were washed sequentially with 50 mM ammonium bicarbonate (NH4HCO3) buffer and ACN. Proteins were reduced and alkylated, respectively, by treatment with 10 mM dithiothreitol (DTT) solution for 1 h at 60 °C and treatment with a 50 mM solution of iodine acetamide. After sequential washings with buffer and ACN, proteins were digested overnight at 37 °C with 120 μg/mL of trypsin. Tryptic peptides were extracted from the gel matrix with 10% formic acid and ACN, and extracts were then pooled and dried in a vacuum centrifuge. Digested peptide solutions were desalted with Zip-Tip pipet tips (Nikkyo Technos, Tokyo, Japan) and re-dissolved in 10 μl of 0.1% TFA in 50% ACN according to manufacturer instructions.

Proteins excised from two-dimensional gels were analyzed in matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF, 4700 Proteomics Analyzer, Applied Biosystems) mass spectrometers. A 1 μl aliquot was mixed with the same volume of a matrix solution (5 mg/ml α-cyano-4-hydroxycinnamic acid in 0.1% TFA in 50% ACN) and spotted on a MALDI plate. MS spectra were acquired in the positive reflector mode, with 500 shots per spectrum being accumulated. Three major peaks were selected for further characterization by MS/MS analysis. MS spectra were acquired using collision-induced dissociation (CID) with atmospheric air as the collision gas, the MS 1 kV positive mode being used. The peak lists for MS/MS spectra from same-replicate samples were merged into a single file prior to database searching. The pkl and mgf files were searched against the Swiss-Prot database using MASCOT (http://www.matrixscience.com; London, UK). Search parameters were set as follows: 1 missed cleavage, fixed modification, peptide charge ±1, and variable modifications were carbamidomethyl of cysteine and oxidation of methionine. Trypsin was specified as the proteolytic enzyme. Peptide tolerance and MS/MS mass tolerance were 50 ppm and 0.25 Da, respectively. Peptide mass fingerprinting (PMF) match confidence was based on the MOWSE score and confirmed by accurate overlapping of matched peptides with mass spectrum major peaks. Scores greater than 67 (p  < 0.05) were considered positive. Peaks with multiple proteins detected by MS were not considered. Only significant hits, as defined by MASCOT probability analysis (p  < 0.05), were accepted.

Results

Physiological characteristics of cauliflower

Characteristic physiological responses of plants to different temperature and waterlogging treatments were evaluated. Table 1 illustrates the comparison of Fv/Fm values under four treatments at seven different times in leaves of three genotypes of cauliflower. Levels of Fv/Fm in all plants progressively decreased as flood stress and heat stress (NFH and FH) durations were extended. Fv/Fm in ‘H41’ plants showed significantly lower values after 6 h with flooding treatments (FC and FH) compared to 0 h, and the lowest values (0.304 and 0.002) were found at 96 h under FC and FH treatments, respectively. However, significant decreases in the Fv/Fm ratio of ‘H41’ plants were observed as late as 48 h under heat treatment (NFH). In addition, when treatments were compared across stressing times of 6–96 h, both control (NFC) and NFH treatments exhibited significantly higher Fv/Fm values than FC and FH treatments, indicating that ‘H41’ is a heat-tolerant cultivar, and this increase in Fv/Fm value was clearly evident with the 40 °C treatment (NFH). Slight decreases in Fv/Fm values were noted in ‘H69’ plants with FC treatment as flood stress time was extended, but high temperature treatments (NFH and FH) rapidly decreased Fv/Fm values in ‘H69’ over time. ‘H69’ Fv/Fm values over time under NFC and FC treatments were both significantly higher compared to NFH and FH treatments, suggesting that ‘H69’ is a flood-tolerant cultivar, and this increase in Fv/Fm value was obvious in the flooding-alone treatment (FC). Fv/Fm values in ‘H71’ plants under stressed treatments (FC, NFC, and FH) appeared to be significantly lower than under NFC after 6 h of treatment, indicating that ‘H71’ is not a flood- or heat-tolerant cultivar.

Chlorophyll content (CC) was assessed over time under various treatments, and its patterns and trends were similar to those of Fv/Fm. When NFC treatment over time was compared, all ‘H69’ and ‘H71’ plants exhibited significantly higher CC levels (61.0–56.1) than ‘H41’ plants (49.3–47.5) at all times (Table 2); thus, the CC of different genotypes responded totally differently to non-flooding and heat stresses. CC also gradually decreased in all plants during all time intervals under stressing treatments. The trend in CC under NFC treatment was similar to that of NFH treatment in ‘H41’ and FC treatment in ‘H69’ at all time intervals. A maximal decrease from 56.4 (0 h) to 16.5 (96 h) was found with FH treatment over time for ‘H71’ plants.

As shown in Table 3, the levels of EL (%) in all plants increased at different rates as stress duration was extended. ‘H41’ plants under NFH treatment exhibited lower levels of EL compared to ‘H69’ and ‘H71’ from 48 to 96 h. However, EL in ‘H41’ showed significantly higher values after 6 h with FC treatments compared to ‘H69’ plants, and the highest value (89.9%) was found with FC treatment at 96 h. Furthermore, EL showed no significant difference for all plants over time under the NFC condition. Overall, the study showed different EL values in the different cultivars under different treatment durations.

Table 4 lists the different points in time that waterlogging and high temperature treatments were monitored by measuring the changes in WP. When different treatments across time were compared, WP with stress treatments was significantly lower (more negative) than those with NFC treatment from 12 to 96 h for all genotypes, indicating that high temperature and/or flooding induced a decrease in leaf water level, subsequently affecting leaf WP. The trends in WP differed in all stressing treatments from 12 to 96 h in all plants. Thus, the WP of the different genotypes responded differently to specific stresses. The lowest value in WP (−2.27 MPa) was detected at 48 h with exposure to NFH treatment in ‘H71’.

Comparison of overall changes in protein levels among stress treatments

Proteome alterations in the three genotypes in response to short-term exposures to flood and high-temperature conditions were compared relative to treatments and controls. The ProteomeLab system uses two-dimensional liquid chromatography based on high-performance chromatofocusing in the first dimension followed by high-resolution reversed-phase chromatography in the second dimension, and has become available for sample fractionation and more resolution at extreme pH values (Soldi et al., 2005, Pirondini et al., 2006). Protein peaks were identified and their expression patterns analyzed. Protein fractions were separated according to isoelectric point (pI) and hydrophobicity and detected in a pH range of 8.0–4.0 by ProteomeLab PF-2D. Proteins were analyzed by MALDI-TOF-MS and identified via PMF and database searching by their calculated molecular weight, pI score, and percent coverage, and their database accession numbers and alteration values among stress treatments are represented in Table 5, Table 6, Table 7, Table 8 .

Table 5

Identification of differentially expressed proteins found in cauliflower ‘H41’ plants by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) in comparison to combination treatments for 0 and 6 h.

Comparison	Peak (change of regulation)	Homologous protein	Species	pI	Scorea	Covb (%)	Accession no.
FC and NFC at 6 h	291 (increased)	Phosphoglucosamine mutase	Magnetococcus sp.	6.07	70	25	GLMM_MAGSM
	292 (increased)	Orotidine 5′-phosphate decarboxylase	Lactobacillus salivarius	5.90	62	29	PYRF_LACS1
NFH and NFC at 6 h	294 (increased)	S-adenosylmethionine synthetase	Methanococcoides burtonii	4.84	66	42	METK_METBU
FC 6-h and NFC 0-h	301 (increased)	16 kDa calcium-binding protein	Schistosoma mansoni	4.96	69	71	SM16_SCHMA
	302 (increased)	Acetyl-coenzyme A carboxylase carboxyl transferase subunit beta	Flavobacterium johnsoniae	6.54	71	26	ACCD_FLAJ1

Probability-based molecular weight search (Mowse) score.

Sequence coverage percentage.

Table 6

Identification of differentially expressed proteins found in cauliflower ‘H71’ plants by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) in comparison to combination treatments for 0 and 6 h.

Comparison	Peak (change of regulation)	Homologous protein	Species	pI	Scorea	Covb (%)	Accession no.
NFH 6-h and NFC 0-h	271 (increased)	Phosphoserine aminotransferase	Photobacterium profundum	5.40	72	21	SERC_PHOPR
	272 (increased)	Imidazole glycerol phosphate synthase subunit hisF	Clostridium thermocellum	5.32	64	35	HIS6_CLOTH
	273 (increased)	Phosphoribosylformyl glycinamidine synthase 1	Leptospira interrogans	6.90	72	41	PURQ_LEPIC
	295 (increased)	Probable hydrogenase nickel incorporation protein hypA	Rhodopseudomonas palustris	5.0	74	74	HYPA_RHOPS
FC 6-h and NFC 0-h	296 (increased)	Putative Holliday junction resolvase	Bacillus halodurans	6.14	60	61	RUVX_BACHD
FH and NFC at 6-h	281 (increased)	tRNA (mo5U34)-methyltransferase	Pseudomonas stutzeri	5.8	70	43	CMOB_PSEU5
	282 (increased)	Arginyl-tRNA synthetase	Chlamydophila felis	5.8	63	26	SYR_CHLFF
	297 (increased)	Elongation factor G	Thiomicrospira crunogena	4.84	68	18	EFG_THICR
	304 (increased)	Methionyl-tRNA formyltransferase	Clostridium perfringens	5.34	67	31	FMT_CLOPE

Probability-based molecular weight search (Mowse) score.

Sequence coverage percentage.

Table 7

Identification of differentially expressed proteins found in cauliflower ‘H41’ plants by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) in comparison to combination treatments for 0 and 24 h.

Comparison	Peak (change of regulation)	Homologous protein	Species	pI	Scorea	Covb (%)	Accession no.
FH and NFC at 24 h	231 (decreased)	Uncharacterized ribulose bisphosphate carboxylase-like protein	Symbiodinium sp.	6.30	66	23	RBLL_SYMSP
	232 (decreased)	Fructose-1,6-bisphosphatase	Dictyostelium discoideum	6.60	68	44	F16P_DICDI
	241 (decreased)	Adenylosuccinate synthetase	Vibrio parahaemolyticus	5.38	73	34	PURA_VIBPA
	251 (decreased)	Nodulation protein nolK	Azorhizobium caulinodans	6.11	85	33	NOLK_AZOC5
	252 (decreased)	Uronate isomerase	Haemophilus influenzae	5.62	69	40	UXAC_HAEIG
	253 (decreased)	DNA mismatch repair protein mutS	Pseudomonas stutzeri	5.55	62	10	MUTS_PSEU5
	242 (increased)	Threonyl-tRNA synthetase	Desulfovibrio vulgaris	5.68	69	22	SYT_DESVH
	243 (increased)	DNA-directed RNA polymerase subunit beta	Agrobacterium tumefaciens	6.41	71	15	RPOC_AGRT5
	244 (increased)	Phosphoserine aminotransferase	Photobacterium profundum	5.40	72	21	SERC_PHOPR
	257 (increased)	Ribosomal RNA large subunit methyltransferase N	Anaeromyxobacter dehalogenans	9.09	65	34	RLMN_ANADE
	262 (increased)	Piwi-like protein 2	Schmidtea mediterranea	9.40	65	19	PIWI2_SCHMD
	263 (increased)	Phosphoglucosamine mutase	Actinobacillus succinogenes	5.69	67	27	GLMM_ACTSZ
	276 (increased)	Protein P	Hepatitis B virus	9.87	71	17	DPOL_HBVE1
	277 (increased)	Probable ribosome biogenesis protein NEP1-like	Thermoplasma volcanium	6.47	69	28	NEP1_THEVO
	278 (increased)	UPF0217 protein Sfri_1778	Shewanella frigidimarina	7.13	64	51	Y1778_SHEFN
	285 (increased)	Host range protein 1	Cowpox virus	5.33	75	27	VHR1_COWPX
	286 (increased)	tRNA pseudouridine synthase B	Aeromonas salmonicida	5.44	71	28	TRUB_AERS4
	287 (increased)	Fibrinogen beta chain	Petromyzon marinus	7.27	62	31	FIBB_PETMA
FH and NFH at 24 h	233 (increased)	GTP-binding protein lepA	Mycoplasma capricolum	5.79	72	23	LEPA_MYCCT
	235 (increased)	2-isopropylmalate synthase	Photobacterium profundum	5.18	70	34	LEU1_PHOPR
	236 (increased)	Ribulose bisphosphate carboxylase small chain, chloroplastic	Brassica napus	8.23	95	63	RBS1_BRANA
FC and NFC at 24 h	254 (increased)	DNA mismatch repair protein mutS	Chlamydia pneumoniae	7.84	64	19	MUTS_CHLPN
	255 (increased)	Ribulose bisphosphate carboxylase large chain	Brassica oleracea	5.88	64	20	RBL_BRAOL
	256 (increased)	Shaggy-related protein kinase delta	Arabidopsis thaliana	7.57	64	35	KSG4_ARATH
	274 (increased)	Formate-tetrahydrofolate ligase 1	Desulfitobacterium hafniense	5.67	66	30	FTHS1_DESHY
	275 (increased)	Molybdenum cofactor biosynthesis protein A	Rhizobium etli	6.76	68	30	MOAA_RHIEC
	283 (increased)	DNA polymerase IV	Clostridium kluyveri	9.51	72	33	DPO4_CLOK5
NFH and NFC at 24 h	261 (increased)	ATP-dependent Clp protease proteolytic subunit 1	Mesorhizobium sp.	5.57	66	30	CLPP1_MESSB
NFH and FC at 24 h	284 (increased)	Replicase polyprotein 1ab	Bovine coronavirus	6.44	83	8	R1AB_CVBQ
FH 24-h and NFC 0-h	237 (increased)	3-octaprenyl-4-hydroxybenzoate carboxylyase	Actinobacillus pleuropneumoniae serotype 7	5.75	70	23	UBID_ACTP7

Probability-based molecular weight search (Mowse) score.

Sequence coverage percentage.

Table 8

Identification of differentially expressed proteins found in cauliflower ‘H69’ plants by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) in comparison to combination treatments for 0 and 24 h.

Comparison	Peak (change of regulation)	Homologous protein	Species	pI	Scorea	Covb (%)	Accession no.
FH and NFC at 24 h	258 (decreased)	Ribose import ATP-binding protein RbsA	Haemophilus influenzae	5.52	72	28	RBSA_HAEI8
	259 (decreased)	Phenylalanyl-tRNA synthetase alpha chain	Chlamydia muridarum	5.55	64	38	SYFA_CHLMU
	2511 (decreased)	Ribulose bisphosphate carboxylase large chain	Draba nemorosa	6.04	66	16	RBL_DRANE
	2512 (decreased)	GMP synthase	Treponema denticola	8.31	62	25	GUAA_TREDE
	2513 (decreased)	ATP synthase subunit beta, chloroplastic	Brassica napus	5.21	62	31	ATPB_BRANA
	264 (decreased)	Ribulose bisphosphate carboxylase large chain	Arabis hirsuta	5.96	71	24	RBL_ARAHI
	265 (decreased)	Phosphoribosylformyl glycinamidine synthase 1	Leptospira interrogans	6.90	71	41	PURQ_LEPIC
	266 (decreased)	Ribulose bisphosphate carboxylase large chain	Brassica oleracea	5.88	75	26	RBL_BRAOL
	2312 (increased)	Ribulose bisphosphate carboxylase small chain, chloroplastic	Brassica napus	8.23	92	42	RBS1_BRANA
	2313 (increased)	Pentatricopeptide repeat-containing protein At1g63330	Arabidopsis thaliana	6.83	63	29	PP101_ARATH
	2314 (increased)	Ribulose bisphosphate carboxylase small chain chloroplasti	Brassica napus	8.23	110	66	RBS1_BRANA
	2813 (increased)	Keratin, type II cytoskeletal 1	Pan troglodytes	7.62	72	24	K2C1_PANTR
	2814 (increased)	Molybdenum cofactor biosynthesis protein A	Haemophilus influenzae	8.13	69	38	MOAA_HAEIE
	2815 (increased)	ATP-dependent DNA helicase recQ	Escherichia coli	6.90	69	24	RECQ_ECOLI
	2816 (increased)	Transcription elongation factor greA	Bartonella quintana	5.12	64	40	GREA_BARQU
	2817 (increased)	Phosphoserine aminotransferase	Photobacterium profundum	5.40	60	25	SERC_PHOPR
	2818 (increased)	Ribulose bisphosphate carboxylase large chain	Arabis hirsuta	5.96	65	18	RBL_ARAHI
	2819 (increased)	Non-capsid protein NS-1	Murine minute virus	6.25	77	28	VNCS_MUMIM
	2820 (increased)	Ryanodine receptor 44F	Drosophila melanogaster	5.40	63	8	RY44_DROME
	311 (increased)	16 kDa calcium-binding protein	Schistosoma mansoni	4.96	62	51	SM16_SCHMA
FC and NFC at 24 h	221 (increased)	Mitochondrial inner membrane protease subunit 1	Saccharomyces cerevisiae	7.00	71	47	IMP1_YEAST
	222 (increased)	Ribulose bisphosphate carboxylase small chain chloroplastic	Brassica napus	8.23	74	59	RBS1_BRANA
	223 (increased)	Ribulose bisphosphate carboxylase small chain chloroplastic	Brassica napus	8.23	109	49	RBS1_BRANA
NFH and NFC at 24 h	238 (increased)	Adenylate kinase	Desulfovibrio desulfuricans	8.32	66	60	KAD_DESDA
	239 (increased)	Peptidyl-tRNA hydrolase	Eubacterium eligens	7.63	74	55	PTH_EUBE2
	2310 (increased)	Ribulose bisphosphate carboxylase small chain chloroplastic	Brassica napus	8.23	101	49	RBS1_BRANA
	2311 (increased)	Ribulose bisphosphate carboxylase small chain chloroplastic	Brassica napus	8.23	102	71	RBS1_BRANA
	288 (increased)	Pentatricopeptide repeat-containing protein At3g61360	Arabidopsis thaliana	8.84	64	32	PP291_ARATH
	289 (increased)	Ribose import ATP-binding protein RbsA	Haemophilus influenzae	5.52	74	29	RBSA_HAEI8
	2810 (increased)	Probable O-sialoglycoprotein endopeptidase	Bacteroides fragilis	5.94	66	32	GCP_BACFN
	2811 (increased)	Adenomatous polyposis coli homolog	Xenopus laevis	8.44	60	13	APC_XENLA
	2812 (increased)	Arginyl-tRNA synthetase	Chlamydophila felis	5.80	64	23	SYR_CHLFF
	2912 (increased)	N(2),N(2)-dimethylguanosine tRNA methyltransferase	Methanosphaera stadtmanae	7.18	66	39	TRM1_METST
	2913 (increased)	Putative RNA-directed RNA polymerase	Red clover necrotic mosaic virus	9.03	66	17	RDRP_RCNMV
	2914 (increased)	50S ribosomal protein L11P	Methanocorpusculum labreanum	5.22	57	47	RL11_METLZ
FH and FC at 24 h	2514 (increased)	Uncharacterized 37.2 kDa protein	Orgyia pseudotsugata multicapsid polyhedrosis virus	8.09	71	32	Y011_NPVOP
	2515 (increased)	SEC14 cytosolic factor	Candida glabrata	5.29	70	38	SEC14_CANGA
NFH 24-h and NFC 0-h	245 (increased)	HTH-type protein slmA	Vibrio vulnificus	8.92	67	40	SLMA_VIBVU
	246 (increased)	Ribulose bisphosphate carboxylase small chain chloroplastic	Brassica napus	8.23	87	56	RBS1_BRANA
	247 (increased)	Ribulose bisphosphate carboxylase small chain chloroplastic	Brassica napus	8.23	83	67	RBS1_BRANA
FH 24-h and NFC 0-h	268 (increased)	Ribulose bisphosphate carboxylase small chain chloroplastic	Brassica napus	8.23	71	54	MYOM1_HUMAN

Probability-based molecular weight search (Mowse) score.

Sequence coverage percentage.

Changes in the protein levels of ‘H41’ exposed to FH stresses at 0 and 6 h and their differentially expressed protein peaks are listed in Table 5. Compared to NFC control treatment, two protein peaks (291 of phosphoglucosamine mutase and 292 of rotidine 5′-phosphate decarboxylase) under FC treatment for 6 h were up-regulated (supplementary Fig. S1a). Furthermore, the expression level of S-adenosylmethionine synthetase (peak 294) was increased during high-temperature stressing over 6 h (supplementary Fig. S1b). Two identified peaks, 301 of 16 kDa calcium-binding protein and 302 of acetyl-coenzyme A carboxylase carboxyl transferase subunit beta, were also up-regulated when ‘H41’ plants were treated with FC for 6 h in comparison to NFC treatment at 0 h (Supplementary Fig. S1c).

Nine protein expression patterns were detected in ‘H71’ plants under high-temperature and flood stress at 6 h, and the dynamic changes in the level of each protein are displayed in Table 6. Compared to NFC treatment at 0 h, NFH treatment for 6 h showed that the abundances of peaks 271 (phosphoserine aminotransferase), 272 (imidazole glycerol phosphate synthase subunit hisF), 273 (phosphoribosyl formyl glycin amidine synthase 1), and 295 (probably hydrogenase nickel incorporation protein hypA) were up-regulated (supplementary Fig. S2a). Moreover, the abundance of peak 296 (putative Holliday junction resolvase) in ‘H71’ plants was also up-regulated during 6 h of FC treatment as compared to NFC treatment at 0 h (supplementary Fig. S2b). The last comparison includes four proteins (peaks 281, 282, 297, and 304, namely tRNA-methyltransferase, arginyl-tRNA synthetase, elongation factor G, and methionyl-tRNA formyltransferase, respectively) whose abundances were consistently up-regulated during flood stressing for 6 h at 40 °C. Interestingly, all peaks were up-regulated after all treatments for 0 h and 6 h in both ‘H41’ (Table 5) and ‘H71’ (Table 6) plants.

The identified proteins showed altered expressions by up-regulation or down-regulation in ‘H41’ plants at the 0 h and 24 h time points when combination treatments were compared (Table 7). Only six protein peaks (231, 232, 241, 251, 252, and 253; Supplementary Fig. S3) from comparative proteomic analyses between FH and NFC for 24 h were down-regulated, all others being up-regulated. There were 18 peaks found in FH vs NFC for 24 h, with the remaining 12 peaks being among other comparisons, including FH vs NFH, FC vs NFC, NFH vs NFC, NFH vs FC, and FH at 24 h vs NFC at 0 h. The carbohydrate metabolism related protein, fructose-1,6-bisphosphatase (peak 232), was down-regulated under FH treatment compared to NFC treatment. Nevertheless, the photosynthetic-related proteins ribulose bisphosphate carboxylase (Rubisco) small chain (peak 236) and large chain (peak 255) were increased in abundance in ‘H41’ under flooding for 24 h. Additionally, DNA mismatch repair protein mutS was found in peak 253 with down-regulation (FH vs NFC) and peak 254 with up-regulation (FC vs NFC) in protein abundance through a homologue search.

Differentially expressed proteins identified in ‘H69’ plants at 0 h and 24 h under stressing are shown in Table 8. The protein abundances of eight peaks (258, 259, 2511, 2512, 2513, 264, 265, and 266 of ribose import ATP-binding protein, phenylalanyl-tRNA synthetase, Rubisco large subunit, GMP synthase, ATP synthase subunit, and phosphoribosylformyl glycinamidine synthase) were consistently down-regulated during the FH condition when compared to NFC treatment for 24 h, but the other 33 peaks were up-regulated in protein abundance. Mitochondrial inner membrane protease (peak 221) and Rubisco small subunit (peaks 222 and 223) were up-regulated in ‘H69’ upon FC treatment compared to NFC after 24 h (Supplementary Fig. S4a). Peaks 2514 and 2515, respectively identified as uncharacterized 37.2 kDa and SEC14 cytosolic factor, were increased under heat stress at 24 h (Supplementary Fig. S4b). Meanwhile, Rubisco small subunit (peak 268) was also increased in ‘H69’ under FH at 24 h as compared to NFC at 0 h (Supplementary Fig. S4c). Notably, the majority of peaks found in ‘H69’ at 24 h represented Rubisco down-regulation (peaks 2511, 264, and 266) and up-regulation (peaks 2312, 2314, 2318, 222, 223, 2310, 2311, 246, 247, and 268) in protein abundance.

Discussion

Effects of flooding and high temperature on Fv/Fm, CC, EL, and WP

Because flood and heat often co-occur in stress-prone environments, this study compared the combined effects of flood and heat with those of the single stresses on plant physiology. In general, the changes in plant physiology were greater under the combined treatment than under heating or flooding alone. The differences in responses to flood and heat in Fv/Fm, CC, EL, and WP seen between cultivars suggested that the three genotypes have unique mechanisms for coping with environmental stresses. The heat-tolerant ‘H41’ genotype showed significantly higher Fv/Fm, CC, and WP values (Table 1, Table 2, Table 4) and lower EL% (Table 3) than the sensitive ‘H71’ genotype when heated for 24 h under NFH treatment, which reflects its tolerant nature. The exposure to flooding for at least 6 h under FC treatment provoked greater reductions in Fv/Fm and CC values in ‘H71’ plants than ‘H69’ plants (Table 1, Table 2). In addition, flooding also significantly decreased the EL in ‘H69’ plants relative to ‘H71’ plants after 48–96 h (Table 3), indicating that ‘H69’ is flood-tolerant. Similar losses of Fv/Fm, CC, and WP and an increase in EL were observed in ‘H71’ under flooding and high temperature stresses (FH), but responded at different levels over time (Table 1, Table 2, Table 3, Table 4).

A soil water potential of approximately −2 MPa was used in greenhouse experiments to simulate field situations under severe stress conditions (Ashoub et al., 2013). The leaf WP of ‘H71’ plants under high-temperature treatment was significantly lower than in ‘H41’ and ‘H69’ plants after 12 h, indicating that ‘H71’ plants suffered from heat exposure. A lower osmotic potential is responsible for maintaining turgor when water loss occurs due to high temperature. Thus, high electrical conductivity due to a high concentration of electrolytes in the leaf sap of ‘H71’ plants should lead to alterations in membrane permeability and a reduced ability to retain solutes and water during high temperature and waterlogging. However, the ability of ‘H41’ and ‘H69’ plants to tolerate heat and flood stress, respectively, strongly depends on adjustments in water balance and ionic leakage. A decline in WP and an increase in EL during heat and flood stress were associated with leaf water deficits as well as increases in ionic leakage. At the physiological level, the many effects of flooding and heat stresses indicate the importance of protecting plants from oxidative damage caused by the overproduction of ROS that is elicited by increased ion leakage (Kangasjarvi et al., 2008). The chlorophyll fluorescence emission parameter, Fv/Fm, which is widely used as a proxy for the maximum quantum efficiency of PS II photochemistry, was correlated with flooding and high temperature tolerance. In healthy leaves, the Fv/Fm value is close to 0.8, which is a typical value for uninhibited plants. A lower value indicates that some proportion of the PSII reaction centers are damaged, which is often observed in plants under stress (Camejo et al., 2006). Flooding and high temperatures inhibit photosynthetic CO2 fixation and damages photosynthetic electron transport at the site of PSII where there exists a very sensitive photosynthesis apparatus. This reduction in the CO2 assimilation rate observed in ‘H71’ plants was generated by effects on the Calvin cycle and also on PSII functioning. Exposure of ‘H71’ plants to flooding and high temperature may lead to reductions in net photosynthetic rate, stomatal closure, and cell activities that damage photosynthetic membranes (Jiang and Huang, 2001). These results suggested that the chlorophyll fluorescence parameters were stress specific and were not expressed solely in response to an increasing excess of photon energy. Chloroplast development in ‘H71’ plants may be particularly sensitive to high temperature and flooding. Alternatively, the pigments of the sensitive cultivar might have been destroyed because of a high sensitivity to oxidative stress. ‘H71’ plants suffered greater CC losses than ‘H41’ and ‘H69’ plants after 24–96 h of exposure. Flooding and heat may induce stomatal closure and consequently reduce CC levels and WP as well. Typically, the amount of CC is reduced by stress. This is consistent with our observations that chlorophyll loss was more pronounced in waterlog- and heat-sensitive ‘H71’ plants in accordance with the more pronounced and increased visible symptoms of leaf injury. Leaf curling or folding were the initial and most obvious changes observed. Most ‘H71’ leaves progressively became necrotic, epinastic, or wilted over the course of time; however, most ‘H41’ and ‘H69’ leaves visually appeared to be green and healthy after 72 h of flooding and high temperature (photos not shown). Along with visual symptoms, reduced CC could be used to monitor flooding and heat damage in green or senescent leaves. Attempts have been made to breed for increased flooding tolerance and modify Brassica cultivation or management practices to avoid injury from flooding and heat.

Environmental stresses represent the most limiting conditions for horticultural productivity and plant exploitation worldwide. Important factors among them are water and temperature. Flooding and high temperature are major abiotic stresses resulting in serious problems for the growth and yield of flood- and heat-sensitive plants. It is necessary to identify the physiological characteristics that reflect their complex underlying genetic make-up. These easily measured physiological biomarkers could be used as generic tools to develop a reliable method for selecting cauliflower cultivars having flood- and heat-stress resistance. Table 1, Table 2, Table 3, Table 4 demonstrate that the chlorophyll losses in ‘H71’ plants were in concert with increasing EL accumulations after 6 h of flooding and heat treatments, which is an index of oxidative damages to cell constituents in general. This suggests that Fv/Fm and CC can be effectively used as reporter signals in screening for flooding and high temperature tolerances. These easily measured physiological biomarkers could be used as generic tools for developing a reliable method to select ‘H41’ and ‘H69’ cultivars having flooding and heat-stress resistances. These findings are important for farming in high-temperature areas and wetlands or other areas subject to short and intense rainfall events.

Biological or biochemical functions of flood- and heat-responsive proteins in genotypes

Proteomics techniques are becoming indispensable for understanding how plants adapt to abiotic stresses, and have promoted the identification of numerous stress-related proteins and the pathways in which they function (Kosová et al., 2011). Changes in protein accumulation under stress are directly related to the physiological phenotypic responses of plants to stress. The proteomic responses to high-temperature or flood stresses are known for several plants, including rice (Lin et al., 2005), wheat (Majoul et al., 2004), barley (Süle et al., 2004), soybean (Komatsu et al., 2013), radish (Zhang et al., 2013), Arabidopsis thaliana (Palmblad et al., 2008), Agave americana (Shakeel et al., 2013), and Populus euphratica (Ferreira et al., 2006). In the present study, we analyzed physiological and proteomics data in different genotypes of cauliflower under conditions of high-temperature and waterlogging stresses over different time periods. Approximately 250 peaks were detected at the tested time points, whereas 85 peaks were successfully observed to be differentially regulated (14 down-regulated, 71 up-regulated) and expressed at a minimum of one time point. Among them, 33, 29, and 9 proteins were identified in ‘H41’, ‘H69’, and ‘H71’ plants, respectively, indicating that they were regulated in a genotype-specific manner (Supplementary Fig. S5). Rubisco small and large chains, 16 kDa calcium-binding protein, molybdenum cofactor biosynthesis protein A, and phosphoserine aminotransferase were observed in both ‘H41’ and ‘H69’ plants. Phosphoribosylformyl glycinamidine synthase and arginyl-tRNA synthetase were both found in ‘H71’ and ‘H69’ plants (Supplementary Fig. S5). Moreover, in ‘H41’ plants, peaks 253 and 254 were both identified as DNA mismatch repair protein mutS, while 291 and 263 were identified as phosphoglucosamine mutase (Table 5, Table 7). These sister peaks may represent close homologues, but this could not be resolved based on mass spectrometric data and might instead be isoforms resulting from differential post-translational modifications (Rollins et al., 2013). As stress times increased, additional responsive proteins continued to be identified. For instance, 5 (Table 5) and 30 (Table 7) peaks were observed in ‘H41’ under stress treatments for 6 h and 24 h, respectively. In this study, some of the proteins identified were well characterized in terms of response to stressing, while others were not. The biological functions of differentially regulated proteins include roles in photosynthesis, energy metabolism, cellular homeostasis, response to stimuli, transcription and translation, protein biosynthesis, mediation of signal transduction, and other functions.

For survival, plants must respond to flood and heat stresses in a different manner from regulating protein expressions for biochemical and physiological adaptations. Photosynthesis is one of the systems that is most sensitive to high-temperature stress. Changes in environmental temperatures are primarily reflected in photosynthesis, triggering a response that is aimed toward the best possible performance under new conditions. For this, a balance is sought between the energy of absorbed light, carbon assimilation, and consumption by metabolic sinks. Several studies have shown that high-temperature stress can significantly inhibit the rate of photosynthesis (Yan et al., 2013, Salvucci and Crafts-Brandner, 2004). In our study, Rubisco units (small and large chain) related to photosynthesis were differentially expressed and regulated by combination treatments and among genotypes. For example, the up-regulation of Rubisco was detected in ‘H41’ (peaks 236 and 255, Table 7) and ‘H69’ (peaks 2312, 2314, 2818, 222,223, 2310, 2311, 246, 247, and 268, Table 8). However, the level of expression of the Rubisco large subunit was down-regulated in ‘H69’ during FH treatment for 24 h (peaks 2511, 264, and 266, Table 8). Rubisco proteins resist high-temperature and flood stresses by inhibiting photosynthesis and other nonessential metabolic processes. In contrast, to increase energy metabolism, the expression of proteins related to redox homeostasis and response to stimuli were up-regulated, thereby maintaining physiological balance during stress. Rubisco catalyzes the first major step of carbon fixation in photosynthesis. Affinity of Rubisco for CO2 decreases with increasing temperature (Yan et al., 2013). The diminished capacity of Rubisco and its low affinity for CO2 suggest that carbon fixation or assimilation was highly inhibited when ‘H69’ plants were exposed to stress, especially under the high temperature and flooding stresses (Table 8) that lead to reduced rates of photosynthetic CO2 assimilation. Photorespiration serves as an energy sink, preventing the over-reduction of the photosynthetic electron transport chain and photoinhibition. Lee et al. (2007) reported that high temperature over 2 h significantly inhibits the activity and quantity of Rubisco subunits in rice seedlings. The effects of drought stress on Rubisco represent reductions in Rubisco activity (Parry et al., 2002). Exposure to salt stress brings about a reduction in the overall growth and productivity of plants by disturbing the function of vital components of photosynthesis, such as PSII and Rubisco (Ghaffaria et al., 2014). Consequently, a strong reduction in Fv/Fm value (Table 1) and CC content (Table 2) while maintaining photosynthesis under stress after 24 h was seen in all plants. Some researchers report its up-regulation (Ge et al., 2012, Budak et al., 2013), whereas others find down-regulation (Gaoa et al., 2011) or even both (Guoa et al., 2012) in response to abiotic stresses. As a result, an up-regulation in Rubisco levels may also indicate an increase in the photorespiration rate.

Molecular chaperones, including heat shock proteins (HSPs), are key components of innate immunity in plants. They have major roles in protein folding and signal-transduction networks, cell-cycle control, protein degradation, and protein trafficking (Wang et al., 2004). HSP chaperone pathways require energy in the form of ATP hydrolysis in order to function. The ATP-dependent Clp proteases (HSP100) represent a class of HSP. In addition to their function as molecular chaperones, they function in protein disaggregation and protein degradation when removal of potentially harmful polypeptides arising from misfolding, denaturation, or aggregation are important for the maintenance of cellular homeostasis. The mechanism for rescuing proteins from aggregation is proposed to include the cooperation HSP100 (Kregel, 2002, Sarkar et al., 2011). HSPs are typically induced when cells are exposed to various types of environmental stresses. The synthesis of HSPs under high temperatures is reported to be related to thermo-tolerance in plants, and plants with a decreased expression of HSPs show compromised tolerance to acquired thermo-tolerance (Charng et al., 2006). HSPs are closely related to the acquisition of heat resistance in plants. Choi et al. (2012) identify several high molecular weight HSPs whose expression was up-regulated significantly in Pyropia tenera under high-temperature stress for 1 h and 3 h. Furthermore, Lee et al. (2007) report that several HSPs are expressed in rice leaves during heat stress. High temperature might increase the potential risk of protein misfolding, and an active protein quality control system inside cells plays an important role in plant tolerance to heat stress. HSPs are also increased by drought stress in the sugar beet (Hajheidari et al., 2005), wheat (Demirevska et al., 2008), and sugarcane (Jangpromma et al., 2010). In our study, peak 261 of ATP-dependent Clp protease proteolytic subunit 1 was up-regulated in response to heat stress for 24 h in ‘H41’ plants (Table 7). This suggests that the increased level of HSP100 under high-temperature stress allowed the heat-tolerant ‘H41’ to produce more energy through ATP hydrolysis and the degradation or reactivation of damaged proteins and prevented protein misfolding, resulting in reestablishing normal protein conformations and cellular homeostasis.

S-adenosylmethionine (SAM) synthetase catalyzes the synthesis of SAM that serves as a methyl group donor in transmethylation of proteins, nucleic acids, and a precursor in the biosynthesis of polyamines, biotin, and nicotianamine in plants (Roeder et al., 2009). SAM in shoots underwent increased protein synthesis in a heat-tolerant barley cultivar and showed a reduced trend in an abiotic stress-susceptible cultivar (Süle et al., 2004). In our study, the expression of SAM synthetase was significantly up-regulated (peak 294, Table 5) in heat stressed ‘H41’ over 6 h. Therefore, it is most likely that an increased amount of SAM synthetase (as a result of the increased formation of SAM methyl donors) in ‘H41’ contributes to its stronger heat tolerance compared to controls. In addition, peak 238 of adenylate kinase (ADK) was observed in ‘H69’ under heat stress for 24 h (Table 8). ADK catalyzes the salvage synthesis of adenine monophosphate from adenosine and ATP (Wang et al., 2010).

Heat and flood treatments markedly increased the abundance of two elongation factors (EFs): (1) elongation factor G was increased in ‘H71’ (peak 297) under flooding at 6 h (Table 6) and (2) transcription elongation factor greA was increased in ‘H69’ (peak 2816) under FH at 24 h (Table 8). The elongation of polypeptide chains during translation is a conserved process among prokaryotes and eukaryotes. EFs are the critical regulators of protein synthesis, which is influenced by high temperature and drought stress in Arabidopsis (Rizhsky et al., 2004) and rapeseed (Mohammadi et al., 2012). Regulation of the transcriptional and translational machinery is considered to be an important component of cellular stress response. The increased expression of EF greA in ‘H69’ and EF G in ‘H71’ likely led to increased protein synthesis, which served to reduce the effects of treatment stress.

Plants under stress can respond by sensing and transferring stress signals through signal transduction networks. The response to stress is likely mediated through signaling pathways that regulate the expression levels of a range of proteins (Hirayama and Shinozaki, 2010). In our study, GTP-binding protein lepA (peak 233, Table 7) involved in signal transduction was identified. The up-regulation of the lepA protein in response to flooding at 40 °C might reflect the role of this protein in cell signaling under stress.

Ca2+-binding protein is mainly resident in the endoplasmic reticulum where it serves as a calcium modulator and chaperone of newly synthesized glycoproteins (Nam et al., 2012). The latter have functions in plant growth and development as well as biotic and abiotic stress responses. Aghaei et al. (2008) indicated that Ca2+-binding protein was up-regulated by salt stress in potato leaves. Salinity induces Ca2+ accumulation in the cytosol that starts the signaling pathway. The modulation of intracellular Ca2+ levels is regulated by calcium-binding proteins, which, after activation, induce specific kinases (Capriotti et al., 2014). A 16 kDa calcium-binding protein was identified as an up-regulated peak (301) under flooding (Table 1), suggesting the involvement of intracellular calcium homeostasis and signal transduction in ‘H41’ during waterlogging.

ATP synthase is a large multi-subunit transmembrane enzyme complex. The catalytic site for ATP synthesis is localized primarily on the beta subunit. ATP synthase is known to be involved in energy metabolism and was found in ‘H69’ plants upon FH stressing for 24 h (Table 8). Down-regulation of the catalytic subunit of ATP synthase subunit beta (peak 2513) suggested that the energy production systems of ‘H69’ were highly affected by heat and flood stresses. Huseynova et al. (2007) also found that the ATP synthase complex was less accumulated in drought-sensitive wheat ‘Giymatli-2/17’ than in drought-tolerant Azamatli-95 under water stress. In addition, impaired ATP synthesis under drought is regarded as a major constraint to photosynthesis (Flexas and Medrano, 2002).

Fructose-1,6-bisphosphatase (FBPase) is involved in carbohydrate metabolism during gluconeogenesis, and also an important enzyme in the Calvin cycle that plays a crucial role in plant responses to some abiotic stimuli (Fan et al., 2009). The down-regulation of FBPase in ‘H41’ under FH at 24 h (peak 232, Table 7) indicated that gluconeogenesis was inhibited under stressing.

An increase in ribosomal protein is a candidate for initiating a translation site, and the phosphorylation of ribosomal proteins is important in the regulation of protein synthesis (Fatehi et al., 2012). The 50S ribosomal subunit catalyzes the peptidyl transfer reaction of mRNA-directed protein biosynthesis (Sobhanian et al., 2010). Up-regulation of the 50S ribosomal protein L11P (# 2914) was observed under high temperature stress at 24 h in ‘H69’ (Table 8), indicating the resistance of ‘H69’ plants to the inhibitory effect of high temperature on protein biosynthesis. In addition, pentatricopeptide repeat (PPR) proteins are characterized by tandem arrays of degenerate 35 amino acid motifs (O’Toole et al., 2008) and are thought to be RNA binding proteins involved in posttranscriptional processes in mitochondria and chloroplasts (Schmitz-Linneweber et al., 2006). Meanwhile, PPR proteins are also essential for RNA editing in chloroplasts (Emi et al., 2005). Peak 2313 (pentatricopeptide repeat containing protein) of ‘H69’ was up-regulated under FH within 24 h (Table 8), suggesting that ‘H69’ was able to increase RNA maturation and protein function in chloroplasts under FH.

Taken together, genetic differences are supported by the high number of proteins differentially regulated between genotypes. Our data suggest that the early response of cauliflower to flooding and high temperature might be an important stress adaptation for survival following not only hypoxia and heat, but also direct damage to cells by flooding and high temperature. All of the identified proteins likely work cooperatively to reestablish cellular homeostasis under stress. The long term goal of our work is to help breed a competitively higher flood- and heat-tolerant cauliflower to be grown in lowlands during summer. The identification of the unique stress-responsive proteins will allow further dissection of the genetic basis of this transgressive performance in offspring. Our results not only provide information for selecting lines having better tolerance to waterlogging and heat stresses, but also provide a basis for understanding cauliflower metabolic pathways and their cross-talk under stress. Further verification of the correlative stress-responsive protein by gene expression analyses of these stressed-responsive genes, such as Rubisco, EFs, HSP100, FBPase, ADK, ATP synthase, SAM synthetase, PPR protein, GTP-binding protein lepA, Ca+2-binding protein, and ribosomal protein L11P, would facilitate our understanding of the heat- and flood-response mechanism in cauliflowers.

Conclusions

This study provides information on mechanisms underlying the contrasting responses to HF stresses. Different genotypes showed differing physiological responses to combinations of stress treatments. All dynamics of the peaks were analyzed across all treatments, and a comparative analysis identified 85 stress-responsive proteins in cauliflower under short-term stressing. The majority of these stressed-responsive proteins were genotype specific. These identified proteins emerged as key participants in stress tolerance. Most of the changes in the expression levels of these proteins in response to stressing were involved in photosynthesis. ‘H69’ plants resisted high-temperature stressing by inhibiting photosynthesis and other unnecessary metabolic processes. Energy metabolism was increased and up-regulation of the expression of proteins related to cellular homeostasis, as well as those that respond to stimuli, served to maintain physiological balance during stress. Concurrently, the expression levels of proteins related to transcription, translation, and signal transduction were also up-regulated to promote survival under HF stressing. The genetic variations identified in the proteomes, plant growth, and photosynthetic performance in response to flood and heat represent stress adaption mechanisms to be exploited in future cauliflower breeding efforts.