Genes differentially expressed in broccoli as an early and late response to wounding stress

Postharvest Biology and Technology 145 (2018) 172–182

Contents lists available at ScienceDirect

Postharvest Biology and Technology journal homepage: www.elsevier.com/locate/postharvbio

Genes differentially expressed in broccoli as an early and late response to wounding stress

T

Ana M. Torres-Contrerasa, Carolina Senés-Guerreroa,b, Adriana Pachecoa, Mauricio González-Agüeroc, Perla A. Ramos-Parraa, Luis Cisneros-Zevallosd, ⁎ Daniel A. Jacobo-Velázqueza,b, a

Tecnologico de Monterrey, Escuela de Ingenieria y Ciencias, Av. Eugenio Garza Sada 2501 Sur, C.P. 64849, Monterrey, NL, Mexico Tecnologico de Monterrey, Escuela de Ingenieria y Ciencias, Av. General Ramon Corona 2514, Nuevo México, C.P. 45138, Zapopan, Jal., Mexico c Institute for Agricultural Research, INIA-La Platina, Postharvest Unit. Santa Rosa 11610, Santiago, Chile d Texas A&M University, Department of Horticultural Sciences, College Station, TX, 77843-2133, United States b

A R T I C LE I N FO

A B S T R A C T

Keywords: Wounding stress Early and late response RNA-Seq Transcriptome Broccoli Primary and secondary metabolism

The plant wound-response is a complex process that generates changes in physiological, biochemical, and genetic mechanisms. The objective of the present study was to increase our understanding of the genetic woundresponse of broccoli (Brassica oleracea L.) as an early (1 h) and late response (9 h) to two different wounding intensities (florets and chops) through transcriptome analysis by RNA-Seq. Chops generated the highest differential expression at both, early and late response; in the early response, genes that showed the highest upregulation were those involved in jasmonic acid biosynthesis and phenylpropanoid pathway, whereas in the late response those involved in amino acid and indolyl glucosinolate biosynthesis were upregulated. Likewise, in florets, only a few genes involved in the phenylpropanoid pathway were induced, mainly in the early response. The information generated will help to elucidate effective strategies leading to the enhancement of nutraceutical characteristics and shelf-life stability of fresh-cut broccoli products.

1. Introduction Wounding by cutting is a common practice in the postharvest handling of fresh fruits and vegetables. This activity has increased in the last decades due to the growth of the ready to eat fresh-cut and frozen vegetables sectors (Cisneros-Zevallos et al., 2014). Wounding causes quality changes in the food product, as well as physiological, biochemical, and genetic alterations corresponding to defense mechanisms, such as those involved in signaling molecules and secondary metabolite production (Cisneros-Zevallos et al., 2014). Furthermore, wounding alters various aspects of the primary metabolism, like cellular respiration, photosynthesis, and sink/source relationships (Schwachtje and Baldwin, 2008; Jacobo-Velázquez et al., 2015). Also, it is known that the integration of different signals induced by wounding results in a complex cross-talk between the primary and secondary metabolism (Jacobo-Velázquez et al., 2015). Moreover, it has been established that the accumulation of secondary metabolites induced by wounding stress is partially due to the activation of the primary metabolism, because primary metabolites can work as defense compounds,

signaling molecules, and as carbon source for the biosynthesis of secondary metabolites (Jacobo-Velázquez and Cisneros-Zevallos, 2002; Jacobo-Velázquez et al., 2015). Recently, it was reported that the application of wounding stress induces the accumulation of important bioactive compounds in broccoli (Brassica oleracea L.), such as phenolic compounds and glucosinolates (Villarreal-García et al., 2016; TorresContreras et al., 2017). The dynamics and regulation of genes that contribute to wounding defense have not been deeply studied. In Arabidopsis, the highest gene expression response after wounding was observed few minutes (early response, between the first 30 and 90 min) and hours (late response, after 6 h and until 12 h) after wounding (Reymond et al., 2000; Cheong et al., 2002). Some genes involved in the phenylpropanoid pathway, and jasmonate and glucosinolate biosynthesis were reported to be wound-inducible and having different expression patterns after wounding (Cheong et al., 2002). The authors suggested that the time of induction of each gene reflects the position of the gene product in the response pathway. Thus, a transcriptome evaluation at an early and late response to wounding could provide the full panorama about the

⁎ Corresponding author at: Tecnologico de Monterrey, Campus Guadalajara, Av. General Ramon Corona 2514, Edificio MED 4to Piso, C.P. 45138, Zapopan, Jalisco, Mexico. E-mail address: [email protected] (D.A. Jacobo-Velázquez).

https://doi.org/10.1016/j.postharvbio.2018.07.010 Received 31 March 2018; Received in revised form 18 June 2018; Accepted 15 July 2018 0925-5214/ © 2018 Elsevier B.V. All rights reserved.

Postharvest Biology and Technology 145 (2018) 172–182

A.M. Torres-Contreras et al.

the genome of Brassica oleracea (version 2.1, from http://plants. ensembl.org/Brassica_oleracea/Info/Index?db=core) using the default settings of Bowtie2 (Langmead and Salzberg, 2012). The concordantly paired reads that mapped to the genome were used for quantification of the gene level with HTSeq-count using the default settings (Anders et al., 2014). All these steps were performed with the Galaxy platform (Afgan et al., 2016). To determine differentially expressed genes (DEGs), the package DESeq2 version 1.14.1 (Love et al., 2014) was used in R version 3.3.2. DEGs were defined as those having a p-adjusted value ≤0.01, upregulated genes were defined as having a log2 fold change (log2FC) ≥ 1 and downregulated genes were defined as those having a log2FC ≤ −1. To visualize the overall effect of the treatments, principal component analysis (PCA) of counts normalized with a variance stabilizing transformation and a heat map showing gene expression according to treatment were constructed using DESeq2 and different packages available in R (R Development Core Team, 2015).

wound response that leads to the production of secondary metabolites in broccoli. Therefore, the objective of the present study was to increase our understanding of the genetic wound-response of broccoli as an early (1 h) and late response (9 h) to two different wounding intensities (florets and chops) through transcriptome analysis by RNA-Seq. Increasing the scientific knowledge in this area is essential to design effective strategies leading to the enhancement of nutraceutical characteristics and shelf-life stability of fresh-cut broccoli products. 2. Materials and methods 2.1. Plant material, postharvest processing, and storage Broccoli var. Heritage was harvested in Aguascalientes, in May 2016, and obtained in Monterrey (Nuevo León, México) from a local distributor. Broccoli heads were washed and disinfected with chlorinated water (0.02% of sodium hypochlorite solution, pH 6.5), and treated with two different wounding intensities: florets and chops. Whole broccoli heads were used as the control. Florets were obtained using a commercial straight-edged knife, whereas chops were done from broccoli florets using a commercial food processor (Waring Commercial, WFP11, Torrington, CT, USA). Samples were stored inside hermetic plastic containers with periodic ventilation (every 12 h) to avoid CO2 accumulation higher than 0.5% (v/v). Two biological replicates were performed for each treatment. All samples (whole broccoli heads, florets, and chops) were stored at 20 °C in an incubator (VWR, Radnor, PA, USA) under dark conditions. Samples of each treatment were taken at 1 h (early response) and 9 h (late response) of storage time, immediately frozen with liquid nitrogen and stored at −80 °C until needed.

2.5. Quantitative RT-PCR To confirm the differential expression of a set of genes a qRT-PCR approach was followed. The same RNA used for RNA-Seq analysis, as well as another independent RNA extraction of all samples, were used to synthesize cDNA with the AffinityScript QPCR cDNA Synthesis kit (Agilent Technologies, Santa Clara, CA, USA) using random nonamers primers as described by the manufacturer’s protocol. Quantification of transcripts, generated from cDNA, was performed in a Gene 3000 Rotor System (Corbett Life Science, San Francisco, CA, USA) with a 36-well rotor using the Brilliant III Ultra-Fast SYBR Green qPCR Master Mix (Agilent Technologies, Santa Clara, CA, USA) and primers for selected genes (Table 1). Conditions, procedures, and analysis of qRT-PCR data was performed as described by Salzman et al. (2005) using two biological replicates and three technical replicates for each gene validated (n = 6). Amplification specificity of each set of primers was determined by analysis of the cleavage curve and amplicon size on agarose gel electrophoresis, to ensure the absence of non-specific PCR products. Calibration curves of genes are shown in Supplementary Material (Fig. S1). Differential gene expression was calculated using the 2−ΔΔCt method following the protocol of Livak and Schmittgen (2001).

2.2. RNA extraction Two independent RNA isolations of each sample were carried out. RNA extraction was performed following the hot borate method (Wan and Wilkins, 1994). RNA quality was determined using a NanoDrop 1000 spectrophotometer (Thermo Scientific, Waltham, MA, USA), and RNA integrity was evaluated in 1% (w/v) agarose gel. Likewise, RNA integrity number (RIN) was determined using an Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). Total RNA was treated with DNAse using RNAse Free DNAse (Qiagen, Hilden, NRW, Germany) and cleaned using the RNeasy Plant Mini kit (Qiagen, Hilden, NRW, Germany) following manufacturer's recommendations. Total RNA was evaluated with the following quality parameters: r26S/ 18S > 1, RIN > 8, OD 260/280 > 1.9 and OD 260/230 > 1.5.

3. Results and discussion 3.1. Transcriptome gene expression analysis To elucidate a comprehensive overview of postharvest wound-response in broccoli at the gene level, RNA-Seq libraries were designed including whole broccoli head (control), florets, and chops after 1 h (early response) and 9 h (late response) of storage at 20 °C. Sequencing of all treatments together produced a total of 28,495,468 high-quality reads. Paired-end reads from all treatments were mapped to the genome of Brassica oleracea version 2.1, obtaining a 79.81% of overall alignment rate. A global view of the transcriptomes of broccoli under the different treatments is shown in Fig. 1. by means of a PCA (Fig. 1A) and a heat map (Fig. 1B). PCA showed that chops treatment grouped together and differed from the cluster generated by the florets treatment and the control (Fig. 1A). Also, the storage time explained some of the variances, except in the control where both storage times combine in the PC2-axis. In identified DEGs, no significant differences were found when comparing both storage time in control samples, thus from here and along the study whole broccoli heads at 1 h of storage was used as the control sample. Accordingly, chops exhibited the major difference in gene expression when comparing expression levels of the 50 genes with the highest variance in a heat map (Fig. 1B). The early response (1 h) in chops consisted of 3,007 DEGs (83% upregulated and 17% downregulated), while florets showed a lower response with 397 DEGs (94% upregulated and 6% downregulated). In chops, the late response

2.3. Library preparation and RNA-Seq analysis DNase treated RNA was used to prepare 12 separate Illumina sequencing libraries using TruSeq™ mRNA LT sample preparation kit (Illumina, San Diego, CA, USA), corresponding to two independent replicates of each wounding treatment (florets, chops), the control (broccoli head), and sampling times (1 h and 9 h). Before normalization and pooling, libraries were validated with a Bioanalyzer 2100 (Agilent Technologies, Santa Clara, CA, USA) using a DNA High Sensitivity chip. Libraries were sequenced (2 × 81 bp paired-end) at the facilities of Tecnologico de Monterrey (Monterrey, NL, Mexico) with the MiSeq Reagent kit v3 (150 cycles) using the MiSeq system (Illumina, San Diego, CA, USA). 2.4. Bioinformatic analysis Sequenced reads from each sample were first assessed for quality using FastQC (Q ≥ 30; Andrews, 2017), adapters were trimmed with Trimmomatic (Bolger et al., 2014), and paired reads were mapped to 173

Postharvest Biology and Technology 145 (2018) 172–182

A.M. Torres-Contreras et al.

Table 1 Primers used in qRT-PCR for validation of 16 DEGs related to phenolic, glucosinolate, and signaling molecule biosynthesis in broccoli and 2 housekeeping genes. Gene

Description according to GenBank

Forward primer (5´-3´)

Reverse primer (5´-3´)

Amplicon size (bp)

BoPAL1 BoPAL2 BoOPR3 BoAOS BoAOC2 BoGPX2 BoACO4 BoAHA2 BoSTc17 BoCYP79B1 BoIGMT1 BoSTa BoSTb BoMYB122 BoCYP79F2 BoISYNA BoEF1a BoACT2

Phenylalanine ammonia-lyase 1 Phenylalanine ammonia-lyase 2 12-oxophytodienoate reductase 3 allene oxide synthase allene oxide cyclase 2 glutathione peroxidase 2 1-aminocyclopropane-1-carboxylate oxidase ATP phosphohydrolase cytosolic sulfotransferase 17 cytochrome P450 79B1 Indole glucosinolate O- methyltransferase 1 Cytosolic sulfotransferase 16 Cytosolic sulfotransferase 18 Transcription factor MYB122 quercetin 3-O- methyltransferase 1-like inositol-3-phosphate synthase Elongation factor 1α Actin 2

TGGCAGCAATCTCGACCCTTG CATGGCGTCGATGGTTCTATTC CGATAGGAGCGAGTAAAGTTGG GAATCCGTAAATACAACTCCACAG CTCGTCCCATTCACCAACAAAC CGTTGCCTCCAAATGTGGTCT TTGAGGTGATAACCAATGGGAAG AAAACGGGAACGCTGACTCTTA GAGGGTCGTGAAGCTTTGTAG GGAATGGTCCCAACGATGCTAA CGACAGCGTCTTCCTCAACACTTG TCATCCAACACGGCGGACACT CGACCGTACCGAACCAAGACAAGA TAAGCTCATCGCCTACGTCCAA AGAAGGCGAAAGCCGAGATAGA AAGTTTCATTCTTTCCACCCAGT TGCCAACTTCACATCCCAG GTCGCTATTCAAGCTGTTCTCT

CCATAACTATCGGTGCCTTTGC TGGTGCTTCAGCCTGTGCGTTA TTGAGCAAGTCAACCACGGCTA CTCGACCTTATCAACATCGAACAA TGACCGTAGTCGCCGAAGTAGA ATTGTTTCCTGGTTCTTGTCCC TCCAGGGTTGTAGAATGATGCA TCAATTCTCGAAGCCATAGCAG GCGTCAAATAATTCTCCCAGTC AGAGCGTCTTGTTGCTTGAGTA GAGCCTTCTTCACGACAGCGATGG CGAAGCGAGAACGGTTGACGAT AGTGACCACCGTACTCGATGAAGG GCTCCTCTTCGCTAAACTCACC GAGAACGTGGGACAAGGAGTG ATGTTCTCCAGCATAGCCCTC ACCAGCATCACCATTCTTC GAGAGCTTCTCCTTGATGTCTC

124 132 109 134 152 136 104 119 149 167 176 171 144 155 157 122 190 251

previous studies with carrots, this results in a complex cross-talk between the primary and secondary metabolism regulated by stress signaling molecules (Jacobo-Velázquez et al., 2015).

(9 h) increased to 4,646 DEGs (67% upregulated and 33% downregulated), while florets DEGs slightly decreased to 362 DEGs (32% upregulated and 68% downregulated). Functional classification of all DEGs was done using the Plant Reactome Pathway Database (Naithani et al., 2016). The number of pathways and genes affected increased with the wounding-intensity, chops being the most severe treatment (Table 2). Most genes were involved in metabolism and regulation and; in second place, those involved in hormone biosynthesis and signaling. These results agree with previous reports where the gene response of Arabidopsis to different types of preharvest stresses, such as mechanical wounding, pathogen, oxidative stress, and hormonal response was characterized (Reymond et al., 2000; Cheong et al., 2002; Schenk et al., 2000; Mahalingam et al., 2003). In these studies, the largest group of genes that responded to different stresses were those related to metabolism categories whereas genes involved in signal transduction were the second largest category. Cheong et al. (2002) reported that after 30 min of mechanical wounding at least 20% of all altered gene expression in Arabidopsis were involved in recognition and signal transduction. Likewise, Schenk et al. (2000) reported that from diverse signaling molecules, methyl jasmonate and ethylene were particularly associated with enhanced transcription of several regulatory genes. The results show that different signaling pathways are induced by wounding, and as suggested in

3.2. Primary metabolism related genes differentially expressed in broccoli as an early and late response to different wounding intensities It is known that wounding stress alters essential aspects of the primary metabolism, such as cellular respiration, photosynthesis, and sink/source relationships to supply cellular energy for defense response (Schwachtje and Baldwin, 2008). Herein, most of the genes involved in glycolysis, pentose phosphate pathway (PPP), tricarboxylic acid (TCA) cycle, and shikimate pathway were upregulated in response to wounding stress (Fig. 2). These pathways are of great importance since they produce specific and essential precursors for amino acid, secondary metabolite, and signaling molecules biosynthesis (Fig. 2A). The early response in gene expression in the treatment florets was low, since only six putative genes were up-regulated: fructose-bisphosphate aldolase (FBPA), glyceraldehyde 3-phosphate dehydrogenase (G3PD) and pyruvate kinase (PK) from glycolysis; glucose 6-phosphate dehydrogenase (G6PD) and 6-phosphogluconolactonase (6 PG L) from PPP; and 3-deoxy-d-arabinoheptulosonate 7-phosphate synthase (DHAPS), from the shikimate pathway. On the other hand, in chops, the early

Fig. 1. Global view of transcriptome analysis. (A) Principal component analysis (PCA) of normalized read counts after variance stabilizing transformation. Each dot corresponds to a sample replicate and the color to a specific wounding intensity and storage time. (B) Heat map based on the expression level of 50 genes with the highest variance among broccoli samples. Abbreviations: Whole broccoli head (control) stored for 1 h (W1) and 9 h (W9), broccoli florets stored for 1 h (F1) and 9 h (F9), broccoli chops stored for 1 h (C1) and 9 h (C9) at 20 °C. 174

Postharvest Biology and Technology 145 (2018) 172–182

A.M. Torres-Contreras et al.

Table 2 Principal pathways of DEGs in broccoli subjected to different wounding intensities and storage times. Pathway ID

R-BOL-2894885 R-BOL-2744345 R-BOL-2744341 R-BOL-2883407 R-BOL-2744344 R-BOL-2744343 R-BOL-5655122 R-BOL-2867929 R-BOL-1119582 R-BOL-6788019 R-BOL-2961031 R-BOL-1119533 R-BOL-6787011 R-BOL-5608118 R-BOL-1119418 R-BOL-1119501 R-BOL-1119494 R-BOL-5632095 R-BOL-5225808 R-BOL-1119486 R-BOL-3906998

Pathway name

Plant pathways Metabolism and regulation Hormone biosynthesis, signaling, and transport Carbohydrate metabolism Secondary metabolite biosynthesis Amino acid metabolism Amino acid biosynthesis Cofactor biosynthesis Phenylpropanoid biosynthesis, initial reactions Salicylic acid signaling Generation of precursor metabolites and energy TCA cycle (plant) Jasmonic acid biosynthesis and signaling Auxin signalling Suberin biosynthesis S-adenosyl-L-methionine cycle Tryptophan biosynthesis Brassinosteroid biosynthesis and signaling Ethylene biosynthesis and signaling IAA biosynthesis I Fatty acid and lipid metabolism

Total entities in pathway

1608 1520 538 249 197 306 267 194 12 61 39 39 57 72 25 16 20 92 35 44 62

Treatments F1

C1

F9

C9

52 47 27 4 9 4 4 3 8 3 – – 16 2 7 – – 2 1 – 1

236 215 88 38 34 36 31 16 16 18 10 10 24 10 12 8 2 9 10 6 6

27 23 6 5 5 3 2 2 1 2 – – 1 – – – – – – 3 –

306 281 95 31 38 85 72 25 17 16 15 15 14 12 12 11 11 10 10 10 9

The number of DEGs was determined by comparing with whole broccoli heads stored for 1 h at 20 °C. Total entities indicate the number of all genes in each pathway. (–) No DEGs found. Abbreviations: Broccoli florets stored for 1 h (F1) and 9 h (F9), broccoli chops stored for 1 h (C1) and 9 h (C9) at 20 °C.

gene response was higher than in florets, since it included the upregulation of 25 genes from glycolysis, TCA cycle, shikimate pathway, and PPP, as well as other 12 genes related to amino acid biosynthesis (Fig. 2B). Different induced expression patterns were observed among genes that code for the same enzyme, for instance, those that code for PFK, FBPA, DHAPS (Fig. 2B). This result indicates that paralog genes differ in their sensitivity of being wound-inducible. The phosphoglycerate mutase (PGM), pyruvate decarboxylase (PDC), and PK were the highest expressed genes reaching around 4 log2FC (Fig. 2B). Additionally, chorismate mutase (ChM), prephenate dehydratase (PrDH), adenosylhomocisteinase (AHC), and methionine synthase (MetS), which are involved in phenylalanine and methionine biosynthesis, precursors of indole and aliphatic glucosinolates, respectively, were among the highest expressed genes in this response (Yi et al., 2015). For the late response in florets, only 3 DEGs were upregulated: FBPA, 6-PGL and phosphoglycerate kinase (PGK), while chops showed the highest response, with 70 upregulated genes. Glycolysis, PPP, TCA cycle, shikimate pathway, and amino acid biosynthesis were all highly induced in chops after 9 h of storage at 20 °C. For the late response, MetS and AHC, both involved in methionine biosynthesis, were the highest upregulated genes (> 8 log2FC, Fig. 2B). In general, at higher wounding intensity and storage time, more genes were induced in primary metabolism. These results agree with previous reports where primary metabolism was highly activated by wounding stress in carrot (Jacobo-Velázquez et al., 2015; Roitsch, 1999). Additionally, PFK and PK are well known as major regulatory enzymes in plant glycolysis. The gene coding for PK enzyme was induced around 4, 3 and 2 log2FC in chops 1 h, chops 9 h, and floret 1 h, respectively; while PFK was equally induced (3 log2FC) in chops at 1 and 9 h, but not in florets (Fig. 2B). Thus, these results indicate that chops would have a higher supply of sugars and/or carbon skeletons for anabolic pathways, such as amino acid and secondary metabolites biosynthesis. Also, sugars have been recognized as signaling molecules for the expression of defense genes, particularly in the source-sink relationship (Roitsch, 1999; Bolton, 2009). Pentose phosphate pathway (PPP) has been previously reported to be wound-inducible in carrot (Jacobo-Velázquez et al., 2015). This pathway produces erythrose 4-phosphate (E4P), which is required to

synthesize aromatic amino acids. The gene that code for G6PD, which catalyzes the key step in E4P biosynthesis, was induced (2 log2FC) in both wounding intensities (florets and chops), but only as an early response (Fig. 2B). Thus, results suggest that the production of E4P was induced regardless the wounding intensity and only lasted a short time, showing that its accumulation is important as an early wound response. In the shikimate pathway, 2 DEGs that code for DAHPS were induced in the early response, in both floret and chops (2 and 4 log2FC, respectively), while in the late response the induction remained only in chops (4 log2FC). Other genes that also code for DAHPS were induced in chops as a late response (3 log2FC) (Fig. 2B); showing that at higher wounding intensity the gene response was higher and durable, likely because there were substrates available and the products of the pathway are indeed necessary in the defense response (PEP and E4P and amino acids, respectively, in this case). This result agreed with previous reports where DAHPS was induced by wounding in Arabidopsis (Devoto and Turner, 2005) and carrot (Jacobo-Velázquez et al., 2015). Finally, the last step in this pathway catalyzed by chorismate synthase (ChS) was induced in chops (2 log2FC), as an early and late response, indicating the formation of chorismate (Fig. 2). Our results suggest that genes were induced to supply shikimate pathway intermediates for both primary and secondary metabolism in order to increase the concentration of phenylpropanoid precursors needed for lignin production, which should be accelerated according to the wounding intensity level (Becerra-Moreno et al., 2015). Aromatic amino acids are the major precursors for the synthesis of secondary metabolites in plants (Aharoni and Galili, 2011). Different DEGs involved in phenylalanine and tryptophan biosynthesis were induced in chops, predominantly in the late response. Some DEGs that code for chorismate mutase (ChM) and prephenate dehydratase (PrD), involved in phenylalanine biosynthesis, were induced from the early to the late response, while others showed different expression patterns (Fig. 2B). PrD was the highest induced gene (6 log2FC) at the early response in chops, while its expression was reduced at the late response (2 log2FC), while the other genes coding for this enzyme were induced either at the late or early response. These results indicate that paralog genes along the genome respond in a different manner to wounding stress. DEGs that code to anthranilate synthase (AnS), anthranilate phosphoribosyltransferase (AnPRT), and tryptophan synthase (TrS), 175

Postharvest Biology and Technology 145 (2018) 172–182

A.M. Torres-Contreras et al.

attributed to the various biosynthetic processes in which this molecule is involved. As precursor of S-Adenosyl methionine (SAM), methionine is an essential intermediate in plant cellular metabolism, that participates in diverse processes in both primary and secondary metabolism, including the biosynthesis of proteins, lipids, phytosterols, lignin,

involved in tryptophan biosynthesis were all induced, practically only as a late response in chops. On the other hand, DEGs related to the biosynthesis of methionine, AHC, and MetS were among the highest expressed set of primary metabolism related genes, as an early and late response in chops. Methionine biosynthetic induction must be

(caption on next page) 176

Postharvest Biology and Technology 145 (2018) 172–182

A.M. Torres-Contreras et al.

Fig. 2. Primary metabolism related genes differentially expressed in broccoli as an early (1 h) and late (9 h) response to different wounding intensities (florets and chops). (A) Primary metabolism pathway using KEGG as reference to show the genes induced. (B) Heat map of differentially expressed genes. Color intensity represents change in log2FC from negative (light orange) to positive (dark orange). Data represent the log2FC as compared with whole broccoli heads (control) at 1 h of storage at 20 °C. Blue and red letters represent amino acids and signaling molecules, respectively. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article). Abbreviations: pentose phosphate pathway (PPP), tricarboxylic acid cycle (TCA), glucose 6-phosphate isomerase (G6PI), transketolase (TKL), phosphofructokinase (PFK), fructose-bisphosphate aldolase (FBPA), Glyceraldehyde 3-phosphate dehydrogenase (G3PD), phosphoglycerate kinase (PGK), phosphoglycerate mutase (PGM), phosphopyruvate hydratase (PPH), pyruvate kinase (PK), pyruvate decarboxylase (PDC), pyruvate dehydrogenase (PD), dihydrolipoyllysine-residue acetyltransferase (DHL-AT), dihyrdolipoyl dehydrogenase (DHL-DH), glucose 6-phosphate dehydrogenase (G6PD), 6-phosphogluconolactonase (6PGL), 6-phosphogluconate dehydrogenase (6PGD), ribose-5-phosphate pyrophosphokinase (R5PP), 5-Phospho-alpha-D-ribose 1-diphosphate (PRPP), phosphoenol pyruvate carboxylase (PEC), aconitate hydratase (AH), isocitrate dehydrogenase (IDH), oxoglutarate dehydrogenase (OXD), succinyl-CoA synthase (SCoAS), succinate dehydrogenase (SDH), malate dehydrogenase (MDH), 3-phosphoshikimate 1-carboxyvinyltransferase (3PS1C), chorismate synthase (ChS), chorismate mutase (ChM), prephenate dehydratase (PrDH), anthranilate synthase (AnS), anthranilate phosphoribosyltransferase (AnPRT), tryptophan synthase (TrS), adenosylhomocisteinase (AHC), methionine synthase (MetS).

suberin, polyamines, flavonoids, glucosinolates, among others (Ravanel et al., 1998). However, wounding stress can redirect plant metabolism to activate diverse defense mechanisms to heal the tissue: build a physical barrier through lignin biosynthesis, synthesize compounds that act as a deterrent for herbivores such as glucosinolates, and/or the activation of signaling pathways, as phytohormone biosynthesis. Methionine is an essential metabolite in the wound-response since it participates in these pathways, being in some way a branch point in wounding stress response.

corresponding to aliphatic and indolyl glucosinolates, respectively, were upregulated. The highest DEGs (4 log2FC) in this response were sulfotransferases (ST): ST5a and ST5c in chops. In florets, ST5b and ST5c were less overexpressed (2 log2FC). Different STs catalyze the last step of glucosinolate core structure biosynthesis. In previous reports, it was established that STs are differentially expressed and have different substrate specificities under various conditions in Arabidopsis, resulting in differences in individual glucosinolate profiles (Klein et al., 2006; Klein and Papenbrock, 2009). Thus, since STs were differentially overexpressed, wounding intensity could modulate the accumulation of individual glucosinolate, mainly depending on the specific STs induced. These results suggest that florets would show accumulation of aliphatic glucosinolates; whereas chops would show accumulation of indolyl glucosinolates. These observations agree with a previous report where broccoli florets showed high accumulation of glucoraphanin (324%) and glucoerucin (440%) after 24 h of storage (Torres-Contreras et al., 2017). Regarding the activation of genes related with glucosinolate biosynthesis as a late response, it was null in florets, while for chops the response was the highest observed in this study, especially those related with the indolyl glucosinolate biosynthetic pathway, reaching differential expression values of up to 8 log2FC (Fig. 4B and D). The highest expressed gene was CYP79B2 (8 log2FC), which catalyzes the first step in the indolyl glucosinolate biosynthesis pathway. CYP79B2 was previously reported to be wound-inducible, and its overexpression was related to an increase in indolyl glucosinolates in Arabidopsis (Mikkelsen et al., 2000). Each step in the indolyl glucosinolate biosynthetic pathway was induced in chops including the transcription factors MYB51 and MYB122, with differential expression values of 2 and 4 log2FC, respectively (Fig. 4D). This agree with a previous report in which mechanical stimuli produced a transient induction of MYB51 resulting in the accumulation of indolyl glucosinolates in Arabidopsis (Gigolashvili et al., 2007). MYB family transcription factors have been reported as one of the major regulators in glucosinolate biosynthesis (Frerigmann and Gigolashvili, 2014). Therefore, indolyl glucosinolates would be the type of glucosinolates with the highest accumulation in wounded broccoli, since induced MYB detected were those involved in indolyl and not in aliphatic glucosinolate. Our results correlate well with the previously reported glucosinolate accumulation in chops by Verkerk et al. (2001) and Torres-Contreras et al. (2017), since indolyl glucosinolate accumulation was higher than aliphatic. Induced genes in indolyl pathway must be the responsible for the increment of 4-hydroxy-glucobrassicin and neoglucobrassicin, 250% and 70%, respectively, and although genes in aliphatic pathway were also induced, lower accumulation was reported (Torres-Contreras et al., 2017). In addition, these findings suggest that methionine was utilized in others pathways rather than aliphatic glucosinolate biosynthesis, thus limiting its accumulation.

3.3. Secondary metabolism related genes differentially expressed in broccoli as an early and late response to different wounding intensities It has been recently reported that wounding stress in broccoli induces the accumulation of secondary metabolites, mainly phenolic compounds, and glucosinolates (Villarreal-García et al., 2016; TorresContreras et al., 2017). However, the wound-induced activation of biosynthetic genes involved in the accumulation of these metabolites has not yet been elucidated. Herein, in the phenylpropanoid pathway, 13 different DEGs were found upregulated among treatments (Fig. 3A). The early response consisted in 9 and 12 DEGs in florets and chops, respectively (Fig. 3B). In chops, 5 genes coding for phenylalanine ammonia-lyase (PAL) were up-regulated, while florets showed an increase in expression of only 1 of them. The phenylpropanoid gene that showed the highest upregulation in florets was the cinnamate 4-hydroxylase (up to 6 log2FC), while in chops this gene was not differentially expressed (Fig. 3B). This early response was significantly different when comparing the wounding intensities evaluated, suggesting that diverse steps in the phenylpropanoid pathway were affected selectively and differentially by the wounding intensity applied after the first hours of incubation at 20 °C. These results indicate that the right selection of wounding intensity would selectively induce specific genes in the phenylpropanoid pathway. In the late response, only CCoAMT of the induced genes in the early response showed slight overexpressed condition in florets, indicating that the gene induction in florets is mainly temporal and not sustained, while in chops 10 of the 12 genes maintained their high overexpression, including cinnamic acid 4-hydroxylase (C4H), 4-coumarate:CoA ligase AMP-forming (4CL), hydroxycinnamoyl CoA quinate hydroxycinnamoyl transferase (HQT), cinnamoyl CoA reductase (CCR), cinnamyl alcohol dehydrogenase (CAD), and the 5 genes coding for PAL. These results suggest that at higher wounding intensity the response of the wound-inducible DEGs related with the phenylpropanoid metabolism is higher and durable. On the other hand, both aliphatic (Fig. 4A) and indolyl (Fig. 4C) glucosinolate pathways were induced by wounding. In general, indolyl glucosinolates showed higher expression than aliphatic, since most of the genes in the indolyl pathway were overexpressed. The early response in florets was practically null, while in chops 3 and 4 DEGs,

177

Postharvest Biology and Technology 145 (2018) 172–182

A.M. Torres-Contreras et al.

Fig. 3. Secondary metabolism related to phenylpropanoid pathway genes differentially expressed in broccoli as an early and late response to different wounding intensities. (A) Phenylpropanoid pathway using KEGG as reference to show the genes induced. (B) Heat map of differentially expressed broccoli genes at 1 h and 9 h after being subjected to two different wounding intensities (florets and chops). The color bar is at the bottom right. Data represent the log2FC as compared with whole broccoli heads at 1 h of storage at 20 °C. Abbreviations: phenylalanine ammonia-lyase (PAL), cinnamic acid 4-hydroxylase (C4H), 4-coumarate:CoA ligase AMP forming (4CL), hydroxycinnamoyl CoA shikimate/quinate hydroxycinnamoyl transferase (HCT), p-coumarate 3´-hydroxylase (C3H), hydroxycinnamoyl CoA quinate hydroxycinnamoyl transferase (HQT); cinnamoyl CoA reductase (CCR), cinnamyl alcohol dehydrogenase (CAD).

content in acerola (Badejo et al., 2008). However, although the gene coding for GMP was highly activated by wounding stress, previous reports indicated that wounding stress does not induce the accumulation of ascorbic acid (Torres-Contreras et al., 2017). This observation could be related with the balance between the rate of synthesis and the rate of utilization of this metabolite. Ascorbic acid is the substrate of ascorbate peroxidase enzyme and, in the present study, it was observed that wounding highly induced the gene encoding for this enzyme. Thus, it is likely that ascorbic acid being produced is rapidly oxidized in order to neutralize free radicals produced in wounded-tissue (Jacobo-Velázquez et al., 2011).

Some genes involved in glucosinolate hydrolysis were also overexpressed as a response to wounding stress (Fig. 4E). One of the genes code to epithiospecifier protein (ESP), which presence during the hydrolysis of aliphatic glucosinolates leads to the formation of epithionitriles instead of isothiocyanates. Other two genes related with glucosinolate hydrolysis code for peroxisomal β-glucosidase and myrosinase enzyme (PEN2 and PEN3). It has been reported that breakdown glucosinolates products are part of the defense response, since they are toxic against bacteria and fungi, so can be the link between abiotic and biotic stress responses (Wittstock and Burow, 2010). Therefore, glucosinolate accumulation observed in previous reports (Villarreal-García et al., 2016; Torres-Contreras et al., 2017) must be the balance among biosynthesis and breakdown of glucosinolates. Regarding ascorbic acid biosynthesis and recycling genes, they were induced as an early response to wounding. In florets, one DEG that codes for GDP-D-mannose pyrophosphorylase (GMP) was induced (2 log2FC), while in chops 10 DEGs were overexpressed (Fig. 6). The highest upregulated DEGs were those that code for L-ascorbate peroxidase and UDP-glucose 6-dehydrogenase (4 log2FC, Fig. 5). Interestingly, in the late response, the activation of genes related with ascorbic acid biosynthesis in florets was higher than in the early response; while in chops the response slightly decreased as compared with the early response (Fig. 6). Furthermore, 3 and 9 DEGs were identified in the late response in florets and chops, respectively. The gene that code for Lascorbate oxidase was the one that showed the highest upregulation (4 log2FC, Fig. 5). GMP has been reported to play a major role in ascorbic acid biosynthesis and effectively used as a target gene to increase ascorbate acid

3.4. Signaling molecules biosynthesis related genes differentially expressed in broccoli as an early and late response to different wounding intensities Previous reports indicated that wounding induces the de novo synthesis of some signaling molecules such as reactive oxygen species (ROS), jasmonic acid (JA), and ethylene (ET), which are known to activate interconnected pathways in the defense response, including secondary metabolites biosynthesis (Jacobo-Velázquez et al., 2015; Rojo et al., 2003; Savatin et al., 2014). Additionally, the exogenous application of JA and ET has been reported to affect glucosinolate accumulation in broccoli (Villarreal-García et al., 2016; MoreiraRodríguez et al., 2017). Nevertheless, the activation of genes related with signaling molecules biosynthesis as a response to wounding stress in broccoli has not yet been studied. As an early response to wounding stress, JA biosynthesis related genes were induced in florets and chops (Fig. 6). In this context, 2 DEGs that code to allene oxide synthase,

178

Postharvest Biology and Technology 145 (2018) 172–182

A.M. Torres-Contreras et al.

Fig. 4. Secondary metabolism related to glucosinolate biosynthesis genes differentially expressed in broccoli as an early and late response to different wounding intensities. (A) Aliphatic glucosinolate biosynthetic pathway and (B) heat map of its differentially expressed broccoli genes. (C) Indolyl glucosinolate biosynthetic pathway and (D) heat map of its differentially expressed broccoli genes. (E) Heat map of differentially expressed genes involved in glucosinolate hydrolysis. All genes were evaluated at 1 and 9 h after being subjected to two different wounding intensities (florets and chops). The color bar is at the center right. Data represent the log2FC comparing with whole broccoli heads after 1 h of storage at 20 °C.

179

Postharvest Biology and Technology 145 (2018) 172–182

A.M. Torres-Contreras et al.

Fig. 5. Ascorbic acid biosynthesis related genes differentially expressed in broccoli as an early and late response to different wounding intensities. The color bar is at the bottom. Data represent the log2FC of each treatment compared to whole broccoli heads at 1 h of storage at 20 °C.

Fig. 6. Signaling molecules biosynthesis (jasmonic acid, ethylene and reactive oxygen species) related genes differentially expressed in broccoli as an early and late response to different wounding intensities. The color bar is at the bottom. Data represent the log2FC of each treatment compared to whole broccoli heads at 1 h of storage at 20 °C. 180

Postharvest Biology and Technology 145 (2018) 172–182

A.M. Torres-Contreras et al.

which catalyzes the rate-limiting step in JA biosynthesis were identified (León et al., 2001) with different expression pattern each one. The first gene coding for allene oxide synthase was induced in both, the florets and chops, showing 2 and 4 log2FC, respectively; while the second one, was only induced in florets. These results indicate that wounding intensity can modulate the expression of genes that control JA biosynthesis. The highest expressed (> 8 log2FC) gene was allene oxide cyclase (AOC) in chops. This gene was previously reported to be transiently upregulated upon wounding and was proposed to have the function of amplifying the wound response in order to produce a rapid defense response (Stenzel et al., 2003), further explaining why this gene showed higher expression in chops than in florets. The JA biosynthesis defense response is transient and short since, in the late response, the expression of the AOC gene was quite diminished. In the late response, this gene remained induced in the florets (3 log2FC), while in chops the induction decreased significantly. Genes involved in ethylene biosynthesis showed higher induction in the late than in the early response. In the early response, aminocyclopropane-1-carboxylic acid synthase (ACS) and ACC oxidase (ACO), 2 and 4 log2FC, respectively, were induce in chops, but not in florets, where no activation of the genes was detected (Fig. 6). The late response in florets was null, while in chops 4 genes that code for ACO and 2 genes coding for ACS were upregulated. These results suggest that ethylene biosynthesis was induced after other events needed for its induction to take place. Finally, it is known that ROS are essential in wounding stress response, since its production occurs immediately after wounding (Orozco-Cardenas and Ryan, 1999; Jacobo-Velázquez et al., 2011). In low concentrations ROS act as secondary messengers, while at a high concentration are part of direct defense (Maffei et al., 2007). In this context, the late response in chops included the induction of NADH dehydrogenase (2 log2FC), this enzyme reduces generation of ROS, helping to maintain the redox status in the cells.

Fig. 7. Correlation analysis between RNA-Seq and qRT-PCR data sets. The log2FC obtained by RNA-Seq (x-axis) was plotted against log2FC by qRT-PCR (y-axis). Values were derived from Table 3.

techniques (Fig. 7). The qRT-PCR results correlated with the RNA-Seq data with a high R2 coefficient of 0.869, supporting the RNA-Seq analysis. 4. Conclusions In the present study, it was demonstrated that the genetic response of broccoli to wounding stress is affected by wounding intensity. Likewise, the early and late response of wound-inducible genes was elucidated. Interestingly, genes not previously reported as wound-inducible were identified. Genes involved in jasmonic acid biosynthesis, phenylpropanoid pathway, amino acid, and glucosinolate biosynthesis were among the highest upregulated genes by wounding stress. These findings are of importance since they contribute to increase our understanding of the genetic response that governs wound-response in broccoli. The information generated in this study will help elucidate effective strategies leading to the enhancement of nutraceutical characteristics and shelf-life stability of fresh-cut broccoli products. One possible practical approach to these results, would be to apply wounding stress in fresh-cut produce and explore the possibility of using 1-methylcyclopropene together with wounding stress to

3.5. Validation of RNA-Seq data by qRT-PCR To determine the reliability of the RNA-Seq analysis, the expression of 16 genes involved in primary metabolism, glucosinolate biosynthesis, phenolic biosynthesis, redox reaction, jasmonic acid biosynthesis, ethylene biosynthesis, and one downregulated gene were selected for qRT-PCR analysis (Table 1). The differential gene expression obtained for these genes was according to the RNA-Seq results (Table 3). Additionally, a plot was generated to compare the log2FC results with both

Table 3 qRT-PCR validation of DEGs determined by RNA-Seq in broccoli subjected to different wounding intensities and storage times. Gene ID – Ensembl plant

Bo4g186590 Bo8g082620 Bo3g086880 Bo2g116210 Bo9g075840 Bo3g025320 Bo5g005140 Bol020978 Bo5g025610 Bo1g002970 Bo8g070650 Bo4g191120 Bo6g118360 Bo6g118359 Bo9g131960 Bo8g101260

Gene name

BoPAL1 BoPAL2 BoOPR3 BoAOS BoAOC2 BoGPX2 BoACO4 BoAHA2 BoSTc17 BoCYP79B1 BoIGMT1 BoSTa BoSTb BoMYB122 BoCYP79F2 BolSYNA

Gene description

phenylalanine ammonia-lyase 1 phenylalanine ammonia-lyase 2 12-oxophytodienoate reductase 3 allene oxide synthase allene oxide cyclase 2 glutathione peroxidase 2 1-aminocyclopropane-1-carboxylate oxidase ATP phosphohydrolase cytosolic sulfotransferase 17 cytochrome P450 79B1 indole glucosinolate O- methyltransferase 1 cytosolic sulfotransferase a cytosolic sulfotransferase b transcription factor MYB122 quercetin 3-O- methyltransferase 1-like inositol-3-phosphate synthase

F1 (Log2FC)

C1 (Log2FC)

F9 (Log2FC)

C9 (Log2FC)

RNAseq

qRT-PCR

RNAseq

qRT-PCR

RNAseq

qRT-PCR

RNAseq

qRT-PCR

3.4 – 2.1 2.2 3.2 – – – 2.6 – – – 2.3 – – –

2.9 – 2.3 2.4 2.8 – – – 2.1 – – – 1.5 – – –

6.7 4.7 4.5 4.3 – 3.5 3.5 2.6 4.1 – – – – – – –

5.8 4.1 4.8 3.7 – 2.9 3.6 2.2 3.5 – – – – – – –

– – – – – – – – – – – – – – – −5.7

– – – – – – – – – – – – – – – −8.7

5.8 4.1 1.7 2.5 5.9 2.5 3.6 2.9 6.3 8.7 3.5 7.2 2.6 4.4 5.5 –

5.6 4.5 1.5 2.8 5.1 2.1 3.9 2.3 5.4 7.7 5.2 2.2 2.1 5.6 7.6 –

(–) Not identified as DEGs. Abbreviations: Broccoli florets stored for 1 h (F1) and 9 h (F9), broccoli chops stored for 1 h (C1) and 9 h (C9) at 20 °C.

181

Postharvest Biology and Technology 145 (2018) 172–182

A.M. Torres-Contreras et al.

ameliorate the senescence effect of ethylene while eliciting the biosynthesis of active compounds; however, this would have to be explored in future studies. On the other hand, if the targeted markets are the emerging high value health markets (e.g., dietary supplements, functional foods, cosmetics, etc.), then wounded broccoli samples can be used as biofactories of active compounds with the purpose of obtaining enriched extracts, where quality and senescence are not an issue.

oxygen species on the accumulation of phenolic antioxidants in carrot tissue under wounding and hyperoxia stress. J. Agric. Food Chem. 59 (12), 6583–6593. Klein, M., Papenbrock, J., 2009. Kinetics and substrate specificities of desulfo‐glucosinolate sulfotransferases in Arabidopsis thaliana. Physiol. Plant. 135 (2), 140–149. Klein, M., Reichelt, M., Gershenzon, J., Papenbrock, J., 2006. The three desulfoglucosinolate sulfotransferase proteins in Arabidopsis have different substrate specificities and are differentially expressed. FEBS J. 273 (1), 122–136. Langmead, B., Salzberg, S.L., 2012. Fast gapped-read alignment with Bowtie 2. Nat. Methods 9 (4), 357–359. León, J., Rojo, E., Sánchez‐Serrano, J.J., 2001. Wound signalling in plants. J. Exp. Bot 52 (354), 1–9. Livak, K.J., Schmittgen, T., 2001. Analysis of relative gene expression data using real-time quantitative PCR and the 2− ΔΔCT method. Methods 25 (4), 402–408. Love, M.I., Huber, W., Anders, S., 2014. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15 (12), 550. Maffei, M.E., Mithöfer, A., Boland, W., 2007. Before gene expression: early events in plant–insect interaction. Trends Plant. Sci. 12 (7), 310–316. Mahalingam, R., Gomez-Buitrago, A., Eckardt, N., Shah, N., Guevara-Garcia, A., Day, P., Raina, R., Fedoroff, N.V., 2003. Characterizing the stress/defense transcriptome of Arabidopsis. Genome Biol. 4 (3), R20. Mikkelsen, M.D., Hansen, C.H., Wittstock, U., Halkier, B.A., 2000. Cytochrome P450 CYP79B2 from Arabidopsis catalyzes the conversion of tryptophan to indole-3-acetaldoxime, a precursor of indole glucosinolates and indole-3-acetic acid. J. Biol. Chem. 275 (43), 33712–33717. Moreira-Rodríguez, M., Nair, V., Benavides, J., Cisneros-Zevallos, L., Jacobo-Velázquez, D.A., 2017. UVA, UVB light doses and harvesting time differentially tailor glucosinolate and phenolic profiles in broccoli sprouts. Molecules 22 (7), 1065. Naithani, S., Preece, J., D’Eustachio, P., Gupta, P., Amarasinghe, V., Dharmawardhana, P.D., Wu, G., Fabregat, A., Elser, J.L., Weiser, J., Keays, M., Fuentes, A., Petryszak, R., Stein, L.D., Ware, D., Jaiswal, P., 2016. Plant Reactome: a resource for plant pathways and comparative analysis. Nucleic Acids Res. 45 (D1), D1029–D1039. Orozco-Cardenas, M., Ryan, C.A., 1999. Hydrogen peroxide is generated systemically in plant leaves by wounding and systemin via the octadecanoid pathway. Proc. Natl. Acad. Sci. 96 (11), 6553–6557. R Development Core Team, 2015. R: A Language and Environment for Statistical Computing. the R Foundation for Statistical Computing, Vienna, Austria. ISBN: 3900051-07-0. Available online at http://www.R-project.org/. Ravanel, S., Gakière, B., Job, D., Douce, R., 1998. The specific features of methionine biosynthesis and metabolism in plants. Proc. Natl. Acad. Sci. 95 (13), 7805–7812. Reymond, P., Weber, H., Damond, M., Farmer, E.E., 2000. Differential gene expression in response to mechanical wounding and insect feeding in Arabidopsis. Plant Cell 12 (5), 707–719. Roitsch, T., 1999. Source-sink regulation by sugar and stress. Curr. Opin. Plant Biol. 2 (3), 198–206. Rojo, E., Solano, R., Sánchez-Serrano, J.J., 2003. Interactions between signaling compounds involved in plant defense. J. Plant Growth Regul. 22 (1), 82–98. Salzman, R.A., Brady, J.A., Finlayson, S.A., Buchanan, C.D., Summer, E.J., Sun, F., Klein, P.E., Klein, R.R., Pratt, L.H., Cordonnier-Pratt, M.M., Mullet, J.E., 2005. Transcriptional profiling of sorghum induced by methyl jasmonate, salicylic acid, and aminocyclopropane carboxylic acid reveals cooperative regulation and novel gene responses. Plant Physiol. 138 (1), 352–368. Savatin, D.V., Gramegna, G., Modesti, V., Cervone, F., 2014. Wounding in the plant tissue: the defense of a dangerous passage. Front. Plant Sci. 5, 470. Schenk, P.M., Kazan, K., Wilson, I., Anderson, J.P., Richmond, T., Somerville, S.C., Manner, J.M., 2000. Coordinated plant defense responses in Arabidopsis revealed by microarray analysis. Proc. Natl. Acad. Sci. 97 (21), 11655–11660. Schwachtje, J., Baldwin, I.T., 2008. Why does herbivore attack reconfigure primary metabolism? Plant Physiol. 146 (3), 845–851. Stenzel, I., Hause, B., Miersch, O., Kurz, T., Maucher, H., Weichert, H., Ziegler, J., Feussner, I., Wasternack, C., 2003. Jasmonate biosynthesis and the allene oxide cyclase family of Arabidopsis thaliana. Plant Mol. Biol. 51 (6), 895–911. Torres-Contreras, A.M., Nair, V., Cisneros-Zevallos, L., Jacobo-Velázquez, D.A., 2017. Stability of bioactive compounds in broccoli as affected by cutting styles and storage time. Molecules. 22 (4), 636. Verkerk, R., Dekker, M., Jongen, W.M.F., 2001. Post-harvest increase of indolyl glucosinolates in response to chopping and storage of Brassica vegetables. J. Sci. Food Agric. 81 (9), 953–958. Villarreal-García, D., Nair, V., Cisneros-Zevallos, L., Jacobo-Velázquez, D.A., 2016. Plants as biofactories: postharvest stress-induced accumulation of phenolic compounds and glucosinolates in broccoli subjected to wounding stress and exogenous phytohormones. Front. Plant Sci. 7, 45. Wan, C.Y., Wilkins, T.A., 1994. A modified hot borate method significantly enhances the yield of high-quality RNA from cotton (Gossypium hirsutum L.). Anal. Biochem. 223 (1), 7–12. Wittstock, U., Burow, M., 2010. Glucosinolate breakdown in Arabidopsis: mechanism, regulation and biological significance. The Arabidopsis Book. pp. e0134. Yi, G.E., Robin, A.H.K., Yang, K., Park, J.I., Kang, J.G., Yang, T.J., Nou, I.S., 2015. Identification and expression analysis of glucosinolate biosynthetic genes and estimation of glucosinolate contents in edible organs of Brassica oleracea subspecies. Molecules 20 (7), 13089–13111.

Acknowledgments This study was supported by funds from Consejo Nacional de Ciencia y Tecnologia (CONACYT, México) Grant (177012) and Tecnológico de Monterrey (Bioprocess, Bioengineering and Synthetic Biology Research Group). Author A.M.T.-C. also acknowledges the scholarship (279532) from CONACYT. The authors would like to thank Wilfrido Ocejo and Mr. Lucky for kindly supplying broccoli samples used in this study. Appendix A. Supplementary data Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.postharvbio.2018.07. 010. References Andrews, S., 2017. FastQC: A Quality Control Tool for High Throughput Sequence Data. (Accessed May 2017). http://www.bioinformatics.babraham.ac.uk/projects/fastqc/. Afgan, E., Baker, D., Van den Beek, M., Blankenberg, D., Bouvier, D., Čech, M., Chilton, J., Clements, D., Coraor, N., Eberhard, C., Grüning, B., Guerler, A., Hillman-Jackson, J., Kuster, G.V., Rasche, E., Soranzo, N., Turaga, N., Taylor, J., Nekrutenko, A., Goecks, J., 2016. The galaxy platform for accessible, reproducible and collaborative biomedical analyses: 2016 update. Nucleic Acids Res. 44, W3–W10. Aharoni, A., Galili, G., 2011. Metabolic engineering of the plant primary–secondary metabolism interface. Curr. Opin. Biotechnol. 22 (2), 239–244. Anders, S., Pyl, P.T.A., Huber, W., 2014. HTSeq-a Python framework to work with highthroughput sequencing data. BMC Bioinf. 31 (2), 166–169. Badejo, A.A., Tanaka, N., Esaka, M., 2008. Analysis of GDP-D-mannose pyrophosphorylase gene promoter from acerola (Malpighia glabra) and increase in ascorbate content of transgenic tobacco expressing the acerola gene. Plant Cell physiol. 49 (1), 126–132. Becerra-Moreno, A., Redondo-Gil, M., Benavides, J., Nair, V., Cisneros-Zevallos, L., Jacobo-Velázquez, D.A., 2015. Combined effect of water loss and wounding stress on gene activation of metabolic pathways associated with phenolic biosynthesis in carrot. Front. Plant Sci. 6, 837. Bolger, A.M., Lohse, M., Usadel, B., 2014. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics 30 (15), 2114–2120. Bolton, M.D., 2009. Primary metabolism and plant defense—fuel for the fire. Mol. Plant Microbe Interact. 22 (5), 487–497. Cheong, Y.H., Chang, H.S., Gupta, R., Wang, X., Zhu, T., Luan, S., 2002. Transcriptional profiling reveals novel interactions between wounding, pathogen, abiotic stress, and hormonal responses in Arabidopsis. Plant Physiol. 129 (2), 661–677. Cisneros-Zevallos, L., Jacobo-Velázquez, D.A., Pech, J.C., Koiwa, H., 2014. Signaling molecules involved in the postharvest stress response of plants. In: Pessarakli, M. (Ed.), Handbook of Plant and Crop Physiology, 3rd ed. CRC Press, Boca Raton, FL, USA, pp. 259–276. Devoto, A., Turner, J.G., 2005. Jasmonate-regulated Arabidopsis stress signalling network. Physiol. Plant. 123, 161–172. Frerigmann, H., Gigolashvili, T., 2014. MYB34, MYB51, and MYB122 distinctly regulate indolic glucosinolate biosynthesis in Arabidopsis thaliana. Mol. Plant. 7 (5), 814–828. Gigolashvili, T., Berger, B., Mock, H.P., Müller, C., Weisshaar, B., Flügge, U.I., 2007. The transcription factor HIG1/MYB51 regulates indolic glucosinolate biosynthesis in Arabidopsis thaliana. Plant J. 50, 886–901. Jacobo-Velázquez, D.A., Cisneros-Zevallos, L., 2002. An alternative use of horticultural crops: stressed plants as biofactories of bioactive phenolic compounds. Agriculture 2 (3), 259–271. Jacobo-Velázquez, D.A., González-Agüero, M., Cisneros-Zevallos, L., 2015. Cross-talk between signaling pathways: the link between plant secondary metabolite production and wounding stress response. Sci. Rep. 5, 8608. Jacobo-Velázquez, D.A., Martínez-Hernández, G.B., del C. Rodríguez, S., Cao, C.M., Cisneros-Zevallos, L., 2011. Plants as biofactories: physiological role of reactive

182