Hydrogen Sulfide Alleviates Dark-promoted Senescence in Postharvest Broccoli

Broccoli (Brassica oleracea var. italica) is an important vegetable crop rich in vitamins and sulforaphane. However, the floral heads of broccoli experience rapid postharvest senescence. Here we found that hydrogen sulfide (H2S) treatment alleviated dark-promoted senescence in broccoli florets. H2S delayed the symptoms of senescence and maintained higher levels of chlorophyll and Rubisco and lower protease activity compared with water control. Gene expression analysis showed that H2S down-regulated the expression of chlorophyll degradation-related genes BoSGR, BoNYC, BoCLH1, BoPPH, and BoRCCR. Expression of lipoxygenase gene BoLOX1 and the genes involved in the ethylene synthesis pathway, BoACS2 and BoACS3, were also down-regulated by H2S. The reduced expression level in cysteine protease gene BoCP3 and aspartic protease gene BoLSC807 suggested the role of H2S in alleviating protein degradation during broccoli senescence. H2S up-regulated the expression of sulfur metabolism genes BoSR and BoOASTL, and the antioxidant gene BoCAT. These results show that H2S plays a vital role in alleviating broccoli senescence through a broad regulation on gene expression of reactive oxygen species (ROS) metabolism genes, ethylene synthesis genes, and protease genes. Broccoli (Brassica oleracea var. italica) is an important vegetable of high nutritional value and a common component of the human diet. Floral heads of broccoli are harvested during the immature stage when florets are composed of male and female reproductive structures surrounded by immature petals and enclosed with chlorophyll-containing sepals (Chen et al., 2008). Postharvest broccoli florets experience senescence similar to those seen in developmental leaf senescence at biochemical and transcriptional levels, involving decline in photosynthesis and degradation of macromolecules such as proteins, lipids, and chlorophyll (Buchanan-Wollaston, 1997; Smart, 1994). Production of ethylene plays an important role in regulating the yellowing of florets after harvest, because chlorophyll loss is associated with an increase in ethylene synthesis in florets (Tian et al., 1994). Dark treatment has been shown to accelerate leaf senescence (Quirino et al., 2000). In many plant species, dark treatment causes chlorophyll and protein degradation, which also occurs in developmental senescent leaves (Biswal and Biswal, 1988). When stored in darkness, broccoli florets experienced more rapid deterioration in quality compared with those under light (Li et al., 2014). To extend the shelf life of broccoli, various methods have been developed, including modified atmosphere packaging, ultraviolet-C and light irradiance, and appropriate refrigerated storage (Chen et al., 2008). H2S has been shown to play various physiological roles in animal cells (Wang, 2002). Recently, more evidence has demonstrated that H2S plays multiple roles in plants, including seed germination, root organogenesis, abiotic stress response, and photosynthesis (Jin et al., 2011, 2013; Lisjak et al., 2013; Zhang et al., 2008, 2009). More recently, H2S was found to prolong postharvest shelf life of strawberry, kiwifruit, and broccoli by acting as an antioxidant (Gao et al., 2013; Hu et al., 2012; Li et al., 2014). Previously we reported that H2S fumigation effectively alleviates postharvest senescence of broccoli through modulation of the endogenous antioxidant system. In this article we investigate accelerated senescence of broccoli florets in darkness using biochemical and molecular analysis and the interplay of H2S with sulfur metabolism, the ethylene pathway, and protein degradation. Materials and Methods Plant materials and treatment. Fresh broccoli (Brassica oleracea var. italica) heads were kindly supplied by Anhui Academy of Agricultural Sciences, Hefei, Anhui province, China. Broccoli heads having similar size and maturity and florets without physical damage or microbial infection were used in this study. The H2S donor, sodium hydrosulfide (NaHS·3H2O; Sigma, St. Louis, MO) was used in aqueous solutions (150 mL at concentrations of 0, 0.24, 0.48, 0.72, 0.96, and 1.2 mmol·L) and applied to florets in sealed containers (volume 3 L). Six pieces of fresh-cut broccoli florets (16 ± 2 g per piece) from six different broccoli heads allocated on board with pores that separated florets and NaHS solution were fumigated with H2S in the sealed containers at 25 C with 85% to 90% relative humidity and stored in darkness. NaHS solutions were renewed every 24 h. Florets were sampled from different broccoli for RNA extraction and other experiments at the time points indicated in figure legends. Each experiment was repeated three times. Visual evaluation of postharvest broccoli. The appearance of fresh-cut broccoli florets fumigated with H2S gas was evaluated subjectively by 20 people using five indices, including freshness, yellowness, texture, flavor, and rotting rate (Takamiya et al., 2000). Each index was divided into four ranks and the score was 0, 1.0, 1.5, and 2.0 with lower numbers indicating lower quality. A sum of the five indices ranging from 0 to 10 was used for visual evaluation. Determination of chlorophyll content. Chlorophyll content of broccoli florets was determined according to the method of Lichtenthaler and Wellburn (1983) with minor modifications. Broccoli florets (0.5 ± 0.01 g) were homogenized using a pestle and mortar in the presence of ice and then incubated in Received for publication 18 Nov. 2014. Accepted for publication 30 Dec. 2014. Funding for this work was provided by Natural Science Foundation of China (31470013, 31301820, 31300133, 31271803), the Scientific Research Foundation for Returned Overseas Chinese Scholars (SRF for ROCS, MOE), the Natural Science Foundations of Anhui Province (11040606M85), and the Anhui Provincial Education Department (2012AJZR0028, ZD200910). We thank Russell Jones, University of California at Berkeley, for the revision of the manuscript. These authors contributed equally to this work. To whom reprint requests should be addressed; e-mail hzhanglab@gmail.com. 416 HORTSCIENCE VOL. 50(3) MARCH 2015 10 mL of 80% acetone for 24 h in darkness at 4 C. The supernatant was measured at 663 and 645 nm, respectively. Chlorophyll content was calculated using the following formulas: total chlorophyll (Chl) (mg·g) = (20.2A645 + 8.02A633) V/W; Chl a (mg·g ) = (12.7A633 – 2.69A645) V/W; Chl b (mg·g ) = (22.9A645 – 4.68A663) V/W. Each analysis was repeated three times and the results were expressed as mg·g fresh weight. Determination of endoprotease and Rubisco. Protease activity was assayed according to the method of Reimerdes and Klostermeyer (1976). Frozen broccoli florets (0.50 ± 0.01 g) were homogenized with a cold mortar and pestle in 5 mL of 50 mmol·L Tris-HCl buffer (pH 7.5) containing 1 mmol·L EDTA, 15 mmol·L b-mercaptoethanol, and 1% (w/v) polyvinyl pyrrolidone. The homogenate was centrifuged at 10,000 g at 4 C for 30 min and the supernatant was used for protease activity and ribulose-1,5bisphosphate carboxylase (Rubisco) determination. One unit of protease activity was defined as an increase of 0.1 OD at 278 nm·h. Rubisco was analyzed by sodium dodecyl sulphate– polyacrylamide gel with Coomassie R-250 staining according to the method of Ono et al. (2013). RNA extraction and reverse transcription– polymerase chain reaction. Broccoli florets were ground in liquid nitrogen, and RNA was extracted by the RNAiso Plus according to the manufacturer’s instructions (Takara, Kyoto, Japan). The cDNA was synthesized from total RNA using a reverse transcription kit (Prime Script RT Master Mix; Takara) and was used as template for polymerase chain reaction (PCR) with Easy Taq DNA Polymerase (TransGen, Beijing, China). The actin (AF044573) gene was used as a control. All primers used for reverse transcription–PCR are listed in Table 1. The primer sequences of Chl degrading genes were designed according to Hasperue et al. (2013). Three independent extractions of RNA were carried out and similar results were obtained. Statistics. Statistical significance was tested by one-way analysis of variance, and the results are expressed as the mean values ± SD of three independent experiments. Fisher’s least significant differences were calculated following a significant (P < 0.01 or P < 0.05) t test.


Broccoli (Brassica oleracea var. italica)
is an important vegetable of high nutritional value and a common component of the human diet. Floral heads of broccoli are harvested during the immature stage when florets are composed of male and female reproductive structures surrounded by immature petals and enclosed with chlorophyll-containing sepals (Chen et al., 2008). Postharvest broccoli florets experience senescence similar to those seen in developmental leaf senescence at biochemical and transcriptional levels, involving decline in photosynthesis and degradation of macromolecules such as proteins, lipids, and chlorophyll (Buchanan-Wollaston, 1997;Smart, 1994). Production of ethylene plays an important role in regulating the yellowing of florets after harvest, because chlorophyll loss is associated with an increase in ethylene synthesis in florets (Tian et al., 1994). Dark treatment has been shown to accelerate leaf senescence (Quirino et al., 2000). In many plant species, dark treatment causes chlorophyll and protein degradation, which also occurs in developmental senescent leaves (Biswal and Biswal, 1988). When stored in darkness, broccoli florets experienced more rapid deterioration in quality compared with those under light . To extend the shelf life of broccoli, various methods have been developed, including modified atmosphere packaging, ultraviolet-C and light irradiance, and appropriate refrigerated storage (Chen et al., 2008).
H 2 S has been shown to play various physiological roles in animal cells (Wang, 2002). Recently, more evidence has demonstrated that H 2 S plays multiple roles in plants, including seed germination, root organogenesis, abiotic stress response, and photosynthesis (Jin et al., 2011(Jin et al., , 2013Lisjak et al., 2013;Zhang et al., 2008Zhang et al., , 2009. More recently, H 2 S was found to prolong postharvest shelf life of strawberry, kiwifruit, and broccoli by acting as an antioxidant (Gao et al., 2013;Hu et al., 2012;Li et al., 2014). Previously we reported that H 2 S fumigation effectively alleviates postharvest senescence of broccoli through modulation of the endogenous antioxidant system. In this article we investigate accelerated senescence of broccoli florets in darkness using biochemical and molecular analysis and the interplay of H 2 S with sulfur metabolism, the ethylene pathway, and protein degradation.

Materials and Methods
Plant materials and treatment. Fresh broccoli (Brassica oleracea var. italica) heads were kindly supplied by Anhui Academy of Agricultural Sciences, Hefei, Anhui province, China. Broccoli heads having similar size and maturity and florets without physical damage or microbial infection were used in this study. The H 2 S donor, sodium hydrosulfide (NaHS · 3H 2 O; Sigma, St. Louis, MO) was used in aqueous solutions (150 mL at concentrations of 0, 0.24, 0.48, 0.72, 0.96, and 1.2 mmol · L -1 ) and applied to florets in sealed containers (volume 3 L). Six pieces of fresh-cut broccoli florets (16 ± 2 g per piece) from six different broccoli heads allocated on board with pores that separated florets and NaHS solution were fumigated with H 2 S in the sealed containers at 25°C with 85% to 90% relative humidity and stored in darkness. NaHS solutions were renewed every 24 h. Florets were sampled from different broccoli for RNA extraction and other experiments at the time points indicated in figure legends. Each experiment was repeated three times.
Visual evaluation of postharvest broccoli. The appearance of fresh-cut broccoli florets fumigated with H 2 S gas was evaluated subjectively by 20 people using five indices, including freshness, yellowness, texture, flavor, and rotting rate (Takamiya et al., 2000). Each index was divided into four ranks and the score was 0, 1.0, 1.5, and 2.0 with lower numbers indicating lower quality. A sum of the five indices ranging from 0 to 10 was used for visual evaluation.
Determination of chlorophyll content. Chlorophyll content of broccoli florets was determined according to the method of Lichtenthaler and Wellburn (1983) with minor modifications. Broccoli florets (0.5 ± 0.01 g) were homogenized using a pestle and mortar in the presence of ice and then incubated in 10 mL of 80% acetone for 24 h in darkness at 4°C. The supernatant was measured at 663 and 645 nm, respectively. Chlorophyll content was calculated using the following formulas: total chlorophyll (Chl) (mg · g -1 ) = (20.2A 645 + 8.02A 633 ) V/W; Chl a (mg · g -1 ) = (12.7A 633 -2.69A 645 ) V/W; Chl b (mg · g -1 ) = (22.9A 645 -4.68A 663 ) V/W. Each analysis was repeated three times and the results were expressed as mg · g -1 fresh weight.
Determination of endoprotease and Rubisco. Protease activity was assayed according to the method of Reimerdes and Klostermeyer (1976). Frozen broccoli florets (0.50 ± 0.01 g) were homogenized with a cold mortar and pestle in 5 mL of 50 mmol · L -1 Tris-HCl buffer (pH 7.5) containing 1 mmol · L -1 EDTA, 15 mmol · L -1 b-mercaptoethanol, and 1% (w/v) polyvinyl pyrrolidone. The homogenate was centrifuged at 10,000 g at 4°C for 30 min and the supernatant was used for protease activity and ribulose-1,5bisphosphate carboxylase (Rubisco) determination. One unit of protease activity was defined as an increase of 0.1 OD at 278 nm · h -1 . Rubisco was analyzed by sodium dodecyl sulphatepolyacrylamide gel with Coomassie R-250 staining according to the method of Ono et al. (2013).
RNA extraction and reverse transcriptionpolymerase chain reaction. Broccoli florets were ground in liquid nitrogen, and RNA was extracted by the RNAiso Plus according to the manufacturer's instructions (Takara, Kyoto, Japan). The cDNA was synthesized from total RNA using a reverse transcription kit (Prime Script TM RT Master Mix; Takara) and was used as template for polymerase chain reaction (PCR) with Easy Taq TM DNA Polymerase (TransGen, Beijing, China). The actin (AF044573) gene was used as a control. All primers used for reverse transcription-PCR are listed in Table 1. The primer sequences of Chl degrading genes were designed according to Hasperue et al. (2013). Three independent extractions of RNA were carried out and similar results were obtained.
Statistics. Statistical significance was tested by one-way analysis of variance, and the results are expressed as the mean values ± SD of three independent experiments. Fisher's least significant differences were calculated following a significant (P < 0.01 or P < 0.05) t test.

Results
H 2 S alleviates dark-induced senescence in broccoli. Broccoli florets stored in darkness were fumigated with different concentrations of NaHS. As shown in Figure 1A, H 2 S effectively alleviated dark-induced senescence in broccoli florets in a dose-dependent manner. Control broccoli yellowed quickly after 3 d of storage, whereas NaHS at an optimal concentration of 0.96 mmol · L -1 postponed yellowing process until Day 5. Florets exposed to 0.96 mmol · L -1 NaHS showed the highest visual value according to the five indices of freshness, yellowness, texture, flavor, and rotting rate (Fig. 1B).
Effects of H 2 S fumigation on content of chlorophyll, activity of protease, and degradation of Rubisco. We determined the effect of H 2 S fumigation on Chl degradation, which is the principle symptom of broccoli senescence. As shown in Figure 2A-C, total Chl, Chl a, and Chl b in both control and H 2 S-fumigated broccoli decreased gradually during the whole storage time. However, H 2 S fumigation maintained significantly higher Chl level than a water control ( Fig. 2A-C).
Senescence-associated proteases, which are accumulated in the vacuole, play a vital role in protein degradation during plant senescence . As expected, protease activity in control samples increased steadily during the first 4 d of storage followed by a decrease on Day 5 (Fig. 2D). In contrast, H 2 S sustained stable and lower levels of protease activity in broccoli (Fig.  2D). During leaf senescence, Rubisco is gradually degraded for recycling within the plant. As shown in Figure 2E, Rubisco was degraded gradually during dark senescence in control broccoli florets, whereas H 2 S fumigation alleviated the degradation of Rubisco, especially on Days 4, 5, and 6.
Effects of H 2 S on expression of chlorophyll degrading genes. To understand the role of H 2 S in alleviating Chl degradation, we assayed the expression of Chl degradingrelated genes in broccoli. Stay-Green (SGR) is required for destabilization of light harvesting complex to Chls (Hörtensteiner and Kräutler, 2011). The enzymes named, nonyellow coloring 1 (NYC1), chlorophyllase (CLH), pheophytinase (PPH), pheophorbide a oxygenase (PaO), and RCC reductase (RCCR), serially degrade Chl b to ''primary'' fluorescent Chl catabolite (Hörtensteiner and Kräutler, 2011). As shown in Figure 3, the expression of BoSGR, BoCLH1, BoCLH2, BoPPH, BoPaO, and BoRCCR was activated after 1 d of storage, suggesting the ongoing Chl degradation. H 2 S obviously repressed the expression of BoSGR, BoCLH1, BoCLH2, BoPPH, and BoRCCR especially on Day 1 of storage, whereas the expression of BoPaO and BoNYC was not significantly changed by H 2 S fumigation.
Effects of H 2 S on expression of ethylenerelated genes. Because ethylene plays a crucial role in floret yellowing of broccoli (Tian et al., 1994), we studied the effect of H 2 S on ethylene biosynthesis and signal transduction genes. As shown in Figure 4, BoACS2, BoACS3, BoERS, and BoETR1 increased markedly in control broccoli during storage time, whereas BoACS1, BoACO1, and BoACO2 remained stable, and BoACO3 and BoETR2 were not detectable. In contrast, H 2 S fumigation repressed the expression of BoACS2 and BoACS3, suggesting the antagonistic role of H 2 S in ethylene biosynthesis.
Effects of H 2 S on expression of sulfur metabolism genes. We determined the effect of H 2 S on sulfur metabolism through the expression assay of O-acetylersine(thiol)lyase (OASTL) gene BoOASTL, BoSR (sulfite reductase), and BoLCD (L-cysteine desulfhydrase). As shown in Figure 5A, H 2 S fumigation induced higher expression of BoOASTL and BoSR on Day 3 of storage compared with the water control, suggesting accelerated sulfur metabolism induced by exogenous H 2 S application.
Effects of H 2 S on expression of antioxidant enzyme genes. The expression of genes encoding enzymes involved in redox metabolism such as lipoxygenase (LOX), catalase (CAT), ascorbate peroxidase (APX), and glutathione reductase (GR) were studied in broccoli florets. As shown in Figure 5B, the expression of BoLOX was induced in water controls on Day 3, whereas H 2 S fumigation sustained lower expression level. Besides, H 2 S application induced higher expression of BoCAT on Day 3 compared with water control. In contrast, the expression levels of BoAPX and BoGR were not significantly changed in H 2 S-fumigated broccoli.
Effects of H 2 S on the transcription of protein degrading genes. Genes involved in protein degradation and in amino acids recycling from degraded protein are thought to be up-regulated in senescencing tissues (Chen et al., 2008). Expression of senescence related protease genes BoCP1, BoCP3, and BoLSC807 were assayed in broccoli (Fig. 5C). BoCP3 was activated on Days 1 and 3 of storage, whereas H 2 S fumigation significantly decreased BoCP3 expression (Fig. 5C). Meanwhile, expression of BoLSC807 (aspartic protease encoding gene) was also repressed by H 2 S fumigation on Day 3 of storage, suggesting the role of H 2 S in inhibiting protein degradation (Fig. 5C).

Discussion
Dark exposure is used as a conventional method to induce plant senescence (Smart, 1994). Our results suggest that dark treatment induces broccoli senescence by Chl and protein degradation, and H 2 S effectively alleviates the senescence symptom (Fig. 1). Chlorophyll content decreases dramatically in control broccoli during storage, whereas H 2 S application sustains higher Chl content ( Fig.  2A-C). Consistently, H 2 S application was found to delay Chl degradation under normal light storage conditions . However, we find that broccoli stored in darkness becomes yellow quickly compare with broccoli under normal light conditions, suggesting that dark treatment is effective in triggering broccoli senescence (data not shown).
Chlorophyll degradation is the first visible symptom of senescence exhibited by a loss of green color and shortening of shelf life (Chen et al., 2008). SGR, a chloroplast-located protein, destabilizes light harvesting complex subunits to release Chls. During Chl degradation, transformation from Chl b to Chl a is catalyzed by a Chl(ide) b reductase [non-yellow coloring 1 (NYC1)] (Hörtensteiner and Kräutler, 2011). The enzyme, PPH, is involved in dephytilation of pheophytine to generate pheophorbide (Hörtensteiner and Kräutler, 2011). PaO catalyzes the opening of the porphyrin ring of pheophorbide and generates red chlorophyll catabolytes (RCC) followed by the reduction of RCC by RCC reductase (RCCR) (Hasperue et al., 2013). In this work, we analyze the expression of Chl degradation-related genes (BoSGR, BoCLH1, BoCLH2, BoPPH, BoRCCR, BoPaO, and BoNYC), and H 2 S is able to sustain a lower expression level of BoSGR, BoCLH1, BoPPH, and BoRCCR in broccoli. BoCLH1 and BoCLH2 are involved in ethylene-induced Chl degradation (Chen et al., 2008). The activation of BoCLH1 and BoCLH2 expression in control broccoli suggests that their expression might be induced by endogenous ethylene generated during senescence (Fig. 3).
Ethylene plays an important role in floret yellowing of broccoli (Tian et al., 1994). 1-aminocyclopropane-1-carboxylic acid (ACC) synthase (ACS) and ACC oxidase (ACO) are key enzymes for ethylene production, which catalyze the conversion of S-adenosylemethionine to ACC and ACC to ethylene, respectively (Hamilton et al., 1991;Lin et al., 2009). In our experiment, increased expression of ethylene synthesis and signaling related genes BoACS2, BoACS3, BoERS, and BoETR1 during storage suggests the key role of the ethylene pathway in inducing broccoli senescence (Fig. 4). However, we first present evidence that H 2 S application effectively blocks the ethylene synthesis pathway through reducing the transcripts of BoACS2 and BoACS3 (Fig. 4), suggesting the role of H 2 S in counteracting ethylene synthesis. The lowered ethylene synthesis in H 2 S treatment will also help to reduce Chl and protein degradation (Chen et al., 2008). Similarly, nitric oxide has been shown to transcriptionally repress the expression of genes of ethylene biosynthesis enzymes, suggesting a similar mechanism in decreasing ethylene synthesis (Manjunatha et al., 2012).
Previous study shows that exogenous H 2 S affects sulfur metabolism on enzymatic activity and gene expression levels (Lai et al., 2014). In the present study, the expression of BoOASTL and BoSR is up-regulated by H 2 S fumigation, implicating the accelerated sulfur metabolism induced by exogenous H 2 S.
Here we find that H 2 S can up-regulate the expression of BoCAT and down-regulate BoLOX expression, which might help to reduce ROS accumulation and alleviate broccoli senescence. Similar to the antioxidant role of H 2 S in plants, H 2 S was found to protect cells from oxidative stress and premature senescence in animal cells through S-sulfhydration of Keap1, antioxidant gene transcription, and glutathione production (Yang et al., 2013). Protein degradation processes rapidly during postharvest senescence of broccoli through the activities of cysteine proteases or aspartic proteases (Buchanan-Wollaston, 1997). In the present study, H 2 S inhibits the activity of protease during broccoli senescence (Fig. 2D). Consistently, reverse transcription-PCR results show that H 2 S reduces gene expression of BoCP3 (cysteine protease) and BoLSC807 (aspartic protease) (Fig. 5C). Cysteine proteases have been found to degrade the large subunit of ribulose-1,5-bisphosphate carboxylase (Rubisco) (Mae et al., 1984). Consistent with the lower protease activity and gene expression, Rubisco degradation is partially alleviated by H 2 S treatment (Fig. 2E).
In conclusion, we provide evidence that H 2 S delays dark-promoted senescence in broccoli by inhibiting Chl and protein degradation, and this observation offers the prospect of H 2 S application in managing postharvest storage of horticultural products. The gene expression assays confirms the multiple roles of H 2 S in regulating Chl and protein degrading genes and ethylene synthesis pathway. Senescence is a complex, highly regulated process in plants (Buchanan-Wollaston, 1997;Smart, 1994), which involved ethylene, H 2 O 2 , cytokinins, jasmonic acid, etc. Thus, more works regarding the interplay between H 2 S and other signals are required to elucidate the senescence regulator role of H 2 S.