Rm

CAa

b

c

d

e

a

ARA

KPMTGL

1

tncPfstSlDl2

K

h0

Postharvest Biology and Technology 97 (2014) 68–76

Contents lists available at ScienceDirect

Postharvest Biology and Technology

journa l h om epa ge : www.elsev ier .com/ locate /postharvbio

egulation of lignin biosynthesis in fruit pericarp hardening ofangosteen (Garcinia mangostana L.) after impact

hanattika Kamdeea, Wachiraya Imsabaib, Rebecca Kirkc,ndrew C. Allanc,d, Ian B. Fergusonc, Saichol Ketsaa,e,∗

Department of Horticulture, Faculty of Agriculture, Kasetsart University, Bangkok 10900, ThailandDepartment of Horticulture, Faculty of Agriculture at Kamphaeng Saen, Kasetsart University, Kamphaeng Saen Campus, Nakhon Pathom 73140, ThailandThe New Zealand Institute for Plant & Food Research Ltd., Private Bag 92 169, Auckland, New ZealandSchool of Biological Sciences, University of Auckland, Private Bag 92019, Auckland, New ZealandPostharvest Technology Innovation Center, Commission on Higher Education, Bangkok 10400, Thailand

r t i c l e i n f o

rticle history:eceived 22 April 2014ccepted 3 June 2014

eywords:ericarp hardeningangosteen

ranscription factor

a b s t r a c t

Pericarp hardening in fresh mangosteen (Garcinia mangostana L.) fruit is a rapid response to mishandlingduring and after harvest. Firmness, lignin content and lignin composition (G and S lignin) increasedrapidly, while total free phenolic content decreased in damaged mangosteen pericarp following impact.Application of nitrogen to the fruit after impact reduced these effects, compared with fruit kept in ambientair. The majority of the genes encoding the mangosteen lignin biosynthetic pathway, and a full lengthMYB transcription factor (R2R3 MYB), were isolated. Expression analysis using qPCR showed that of thegenes encoding enzymes in lignin biosynthesis, only GmCCoAMT and GmF5H increased after impact and

ene expressionignin biosynthesis

correlated with increases in firmness and lignin content. The transcript level of a stress-related R2R3 MYBtranscription factor was significantly increased by impact, and delayed by elevated nitrogen. These resultssuggest that pericarp hardening of mangosteen after impact is due to rapid transcriptional activation oflate steps of the lignin biosynthetic pathway, potentially via up-regulation of transcription factors suchas R2R3 MYBs.

© 2014 Elsevier B.V. All rights reserved.

. Introduction

Mangosteen (Garcinia mangostana L.) is an increasingly impor-ant economic fruit crop. However, its value is often affected by aumber of internal postharvest disorders including flesh translu-ence and pericarp cracking or hardening (Jarimopas et al., 2009).ericarp hardening is one of the most important and occurs duringruit handling, elicited by low storage temperature, or compres-ion and physical impact. This negatively affects fruit quality, andhus acceptability (Ketsa and Koolpluksee, 1993; Tongdee anduwanagul, 1998). This disorder is associated with an increase inignin biosynthesis (Ketsa and Atantee, 1998; Bunsiri et al., 2003;angcham et al., 2008), paralleled by a decrease in phenolic acid

evels, principally in p-coumaric and sinapic acids (Bunsiri et al.,003).

∗ Corresponding author at: Department of Horticulture, Faculty of Agriculture,asetsart University, Bangkok 10900, Thailand.

E-mail address: agrsck@ku.ac.th (S. Ketsa).

ttp://dx.doi.org/10.1016/j.postharvbio.2014.06.004925-5214/© 2014 Elsevier B.V. All rights reserved.

Lignin is a product of the phenylpropanoid metabolic pathwayand its abundance in plants is only exceeded by cellulose (Whettenet al., 1998). It is an important component of the cell wall, impartingstrength and rigidity during plant growth and development, and inplant response to biotic and abiotic stresses (Boerjan et al., 2003;Boudet et al., 2003; Li and Chapple, 2010). Lignification in fruit andvegetables occurs in response to different abiotic stresses, such asphysical impact in mangosteen (Ketsa and Atantee, 1998; Bunsiriet al., 2003), wounding in bamboo shoots (Luo et al., 2007), andlong-term low temperature storage in apple, custard apple, loquat,asparagus, and mangosteen (Cai et al., 2006; Liu and Jiang, 2006;Dangcham et al., 2008; Li and Chapple, 2010). Lignification has beenalso observed in response to biotic stress.

The lignification process involves monolignol formation, trans-portation and polymerization. Monolignols (alcohol monomers),which via polymerization produce lignin, consist of p-coumaryl,coniferyl, and sinapyl alcohols (abbreviated as H,- G- and S-

monolignols, respectively) (Donaldson, 2001; Boerjan et al., 2003;Bonawitz and Chapple, 2010; Li and Chapple, 2010). Lignin synthe-sis involves the coordinated expression of many genes as well asthe activity of at least ten enzymes (Fig. 1), required to synthesize

C. Kamdee et al. / Postharvest Biology and Technology 97 (2014) 68–76 69

Fig. 1. Monolignol biosynthesis by the phenylpropanoid pathway. Abbreviations: phenylalanine ammonia lyase (PAL), cytochrome P450-dependent monooxygenases cinna-mate 4-hydroxylase (C4H), p-coumaroyl shikimate 3-hydroxylase (C3H), and ferulate (coniferaldehyde) 5-hydroxylase (F5H), caffeoyl CoA O-methyltransferase (CCoAMT),c oA rel sferas

S

prlms

tsptPIt2tsaasbtTwg2

daKfi(ga1b

(

affeic acid (5-hydroxyconiferaldehyde) O-methyltransferase (COMT), cinnamoyl Cigase (4CL) and hydroxycinnamoyl coenzyme A: shikimate hydroxycinnamoyl tran

ource: Bonawitz and Chapple (2010).

athway intermediates, which serve as substrates for subsequenteactions (Bonawitz and Chapple, 2010). The last major step inignin synthesis involves monolignol dehydrogenation and poly-

erization, and different classes of oxidative enzymes have beenhown to be implicated.

Most characterized transcriptional regulators of lignin biosyn-hetic genes have been members of the MYB family. Severalubgroups of R2R3-MYB factors have been reported to eitherositively or negatively control lignin biosynthesis. Examples ofhese regulators include Pinus taeda PtMYB4, poplar (Populus sp.)ttMYB21a and Eucalyptus gunnii EgMYB2 (Zhao and Dixon, 2011).n Arabidopsis, over expression of MYB85 leads to ectopic deposi-ion of lignin in epidermal and cortical cells of stems (Zhong et al.,008). In addition, MYB46 functions as a transcriptional switchhat turns on the genes necessary for secondary wall biosynthe-is, while in the same clade, MYB83 over-expression is able toctivate a number of the biosynthetic genes of cellulose, xylemnd lignin and concomitantly induce ectopic secondary wall depo-ition (Maldonado et al., 2002; Ko et al., 2009). It has recentlyeen shown that MYB63 (and MYB58) are transcriptional regula-ors specifically activating lignin biosynthetic genes in Arabidopsis.hey are expressed in fibers and vessels undergoing secondaryall thickening and are able to activate directly lignin biosynthetic

enes and a secondary wall-associated laccase (LAC4) (Zhou et al.,009).

Peroxidase and laccase use H2O2 and molecular oxygen to oxi-ize monolignols. Therefore, by decreasing oxygen levels, modifiedtmosphere (MA) treatments of fruit can avoid these reactions.eeping persimmon fruit in high nitrogen can maintain quality andrmness (Dorria et al., 2011), while MA treatment of bamboo shoots2% O2 and 5% CO2) inhibits lignification (Shen et al., 2006). In man-osteen after impact, fruit held under N2 have greater firmness,nd lower lignin content than fruit held in air (Ketsa and Atantee,

998; Bunsiri et al., 2003). Low O2 storage is therefore likely to beeneficial in reducing these postharvest disorders.

In mangosteen, only two genes [PAL and lignin peroxidaseLgPOD)] have been identified during pericarp hardening resulting

ductase (CCR), cinnamyl alcohol dehydrogenase (CAD), 4-coumarate: coenzyme Ae (HCT).

from low storage temperature (Dangcham et al., 2008). Conse-quently, the genes involved in lignification in this species remainlargely unexplored. In the present study, fruit firmness, total freephenolic contents, and lignin content in relation to pericarp hard-ening were studied. All the genes encoding enzymes of the ligninbiosynthesis pathway were isolated as well as a potential stress-related R2R3 MYB transcription factor. The expression of thesegenes was examined using real-time quantitative PCR (qPCR) dur-ing pericarp hardening. We also report the effect of nitrogenatmosphere on pericarp hardening and gene expression in termsof the mangosteen lignin biosynthetic pathway.

2. Materials and methods

2.1. Fruit treatments

Mangosteen fruit (Garcinia mangostana L.) were purchased froma commercial grower, located in Chanthaburi, Central East Thailand.The fruit were carefully collected and packed into 10 kg plasticboxes to avoid physical damage, and then transported to the lab-oratory within 6 h. Upon arrival, the fruit were selected based ontheir size (75–90 g) and color (dark purple). This color, which is amaturity indicator, corresponds to stage 5 according to the scaledefined by Palapol et al. (2009). For impact treatments, the fruitwere dropped from a height of 100 cm onto a smooth concretefloor. Before impact, the fruit were orientated in such a way thatthe calyx was in the horizontal direction (i.e. parallel to the floor).White powder (talc) was spread on the floor, to clearly indicate theimpact area on the fruit. After impact, the fruit were held at 25 ◦C(78–80% RH). The fruit were then sampled at times 0, 5, 10, 15, 20,25, 30, 60, 120, 180 min and 1 day.

For the experiment involving nitrogen treatment, the fruit weredropped from a height 100 cm and then held in the following two

atmospheric compositions: (1) ambient air (21% O2; control), and(2) nitrogen (very low oxygen atmosphere (0.01% O2)). Ten fruit(constituting one replication) were placed in an 11 L plastic jarunder the ambient air and nitrogen atmospheres. Atmospheric

7 logy a

cc

2tppser

2

fief

2

((mamTwttNg1waaNw

c2dpeommwvwp8dodauwwf0idatS

0 C. Kamdee et al. / Postharvest Bio

onditions were set up using a flow system with air and N2 inompressed tanks (OFN grade, 99.99% N2).

Following the impact treatment, fruit were measured at 0, 1,, 3, 4, 5, 6, 12, 18, and 24 h after treatment for fruit firmness,otal free phenolic content, lignin content, lignin monomer com-osition and gene expression. Fruit pericarp was cut into smallieces and immediately frozen in liquid nitrogen. The samples weretored at −80 ◦C for biochemical and molecular analyses. During thexperiments, 30 fruit per treatment were sampled, 10 fruit for eacheplication.

.2. Pericarp firmness

Fruit firmness, without peeling, was determined using an Effegirmness meter (FT011, Italy). A cylinder plunger head with a diam-ter of 0.2 cm, pressed 0.5 cm deep into the pericarp, was used. Theorce was recorded in Newtons (N).

.3. Lignin analysis

Lignin was assayed by derivatization with thioglycolic acidmodified from the method of Bruce and West (1989), Lange et al.1995)). Five grams of pericarp was homogenized with 100 mL

ethanol for 1 min. The powder was transferred to aluminum cups,nd dried at 60 ◦C for 24 h. Twenty milligrams of dried powder wasixed with 1 mL of 2 N HCl and 0.2 mL of 98% thioglycolic acid.

he solutions were heated at 100 ◦C for 4 h in a boiling water bathith gentle agitation (GFL 3017 shaker, Germany). After cooling,

he mixture was centrifuged at 15,700 × g for 20 min then washedhree times with H2O. The pellet was suspended in 1 mL of 0.5 NaOH, agitated gently at 25 ◦C for 18 h to extract the lignin thio-lycolate. Subsequently, the samples were centrifuged again at5,700 × g for 20 min. The supernatants were transferred to tubes,here 1 mL of concentrated HCl was added and lignin thioglycolic

cid was allowed to precipitate at 4 ◦C for 4 h. After centrifugationt 15,700 × g for 20 min, the pellet was dissolved in 1 mL of 0.5 NaOH and diluted to 40 times (100 �L: 3900 �L). The absorbanceas subsequently measured at 280 nm.

Monolignol contents were analyzed by GC–MS using a modifi-ation of the previous methods (Meyer et al., 1998; Zhang et al.,007). The pericarps of mangosteen fruit were ground to a pow-er in liquid nitrogen and extracted with 20 mL of 0.1 M sodiumhosphate buffer (pH 7.2) for 30 min at 37 ◦C followed by threextractions with 80% ethanol at 80 ◦C. The tissue was then extractednce with acetone and dried at 70 ◦C in an oven. Samples wereixed with 4 mL of 2 M NaOH and 40 �L of nitrobenzene. Thisixture was incubated at 160 ◦C for 3 h. The reaction productsere cooled and 10 �L of 1,4-dioxane (containing 100 �g aceto-

anillone) was added as an internal standard before the mixtureas extracted twice with 2 mL of dichloromethane. The aqueoushase was acidified with 6 M HCl to pH 2 and extracted twice with

mL of ether. The combined ether phases were dried with anhy-rous sodium sulfate and the ether was evaporated in a streamf nitrogen. The dried residue was re-suspended in 50 �L of pyri-ine, with 10 �L of BSA [N,O-bis-(trimethylsilyl)-trifluoracetamide]dded and 2 �L aliquots of the silylated products were analyzedsing an GC–MS QP 2010 Shimadzu at spilt ratio 1:70 equippedith DB-1 column (30 mm × 0.32 mm). The operation conditionsere as follows: initial temperature 40 ◦C for 2 min, 40–200 ◦C

or 4 min ramped at 0.67 ◦C s−1, 200–230 ◦C for 3 min ramped at.17 ◦C s−1. Lignin monomer composition was calculated from the

ntegrated areas of the peaks representing the trimethylsilylated

erivatives of vanillin, syringaldehyde, vanillic acid and syringiccid. Total nitrobenzene oxidation susceptible guaiacyl units orotal G lignin (vanillin and vanillic acid) and syringyl units or total

lignin (syringaldehyde and syringic acid) were determined after

nd Technology 97 (2014) 68–76

correction for recovery efficiencies for each of the products duringthe extraction procedure relative to the internal standard. The iden-tity of each of the peaks used for quantification of lignin monomercomposition was confirmed using GC-electron impact MS by com-parison to authentic compounds. Authentic compounds and all theorganic solvents were of GC purity.

2.4. Total free phenolic content

Total free phenolic contents in the pericarp were deter-mined using the Folin–Ciocalteau method, where the results areexpressed as gallic acid equivalents (GAE) per gram of fresh weight(Waterhouse, 2002). Frozen pericarp (3 g) tissue was extracted with20 mL of methanol. The extraction sample was homogenized usinga Polytron PT 2100 (Kinematica, Luzeen, Switzerland) then cen-trifuged at 18,000 × g for 20 min. The supernatant was used foranalyzing total free phenolic content. The reaction mixture wasprepared by mixing 40 �L of methanol solution of extract, with3.16 mL of distilled water, 0.2 mL of Folin-Ciocalteau’s reagent, and0.6 mL Na2CO3 (20:80, w/v). The mixture was incubated in a waterbath at 40 ◦C for 30 min, the reduction of the Folin–Ciocalteaureagent by phenolic compounds measured as the development of ablue color. Absorbance was assessed by a spectrophotometer (1700UV–visible, Shimadzu, Japan) at 765 nm.

2.5. RNA extraction and cDNA synthesis

Approximately 5 g of frozen samples were ground in a RetschMM301 mixing mill (MM 301, Retsch, Düsseldorf, Germany). TotalRNA was isolated from 5 g of the mangosteen pericarp tissue asdescribed by López-Gómez and Gómez-Lim (1992). RNA was DNaseI treated using Turbo DNAfreeTM Kit (Ambion, TX, USA) follow-ing the manufacturer’s protocol. The RNA was quantified using theNano Drop nd-1000 (Thermo Scientific, MA, USA). The first standcDNA was synthesized from 4 �g of total RNA following the pro-tocol of the SuperScript (SuperScriptTM III First-Strand SynthesisSystem for RT-PCR, Invitrogen, USA) as a template for quantitativereal-time PCR.

2.6. Isolation of candidate lignin biosynthesis genes andregulatory genes

Lignin biosynthesis genes and a MYB transcription factor wereisolated from dark purple mangosteen pericarp tissue (stage 5),using forward and reverse degenerate primers (Table S1; Sup-plementary material), which were designed based on conservedregions of similar genes to plant sequences available from the pub-lic database NCBI (National Center of Biotechnology Informationhttp://www.ncbi.nlm.nih.gov/). The 3′ UTR region of these genefragments were amplified following the protocol of GeneRacerTM

kit (Invitrogen, CA, USA) using specific primers based on initialsequence information (Supplementary Table 1; Supplementarymaterial S2). The 5′ UTR region was obtained using SMARTTM RACEcDNA Amplification Kit SMART (Clontech, CA, USA). The ampli-fied PCR fragments from each gene were purified by using a gelextraction kit (QIAprep Gel Extraction, QIAGEN, Hilden, Germany)and cloned into pGEM-T Vector (pGEM®-T Easy Vector System,Promega, WI, USA). Transformation of the vector with inserted DNAwas conducted using Escherichia coli strain DH5�. The sequences

were examined using ABI PRISM 377 DNA sequencer (AppliedBiosystems, Foster City, CA, USA). The sequences of amplified frag-ments were compared with genes in the GenBank database usingthe BLAST program.

logy and Technology 97 (2014) 68–76 71

2

dudgisw

ptGwptlcpfnp5nt<sc

2

crfic

3

3i

aaafifi

tipcto

3

upsut

C. Kamdee et al. / Postharvest Bio

.7. Real-time quantitative PCR analysis

The oligonucelotide primer sets used for qPCR analysis wereesigned on the basis of 3′-untranslated regions (UTR) of individ-al genes (Table S1; Supplementary material). The primers wereesigned using Primer3 (http://bioinfo.ut.ee/primer3-0.4.0/). Theene-specificity of these primer sets were tested using the follow-ng: (i) individual PCR products were separated on 1% agarose gelstained with SYBR-safe to examine the size; (ii) the PCR productsere cloned into pGEM-T Vector and sequence analysis confirmed.

Fifty-fold diluted cDNA samples were used for qPCR. qPCR waserformed on the Light Cycle 480 system (384-well plates) andhe Light Cycler® 480 SYBR Green I Master kit (Roche Diagnostics,ermany) following the manufacturer’s instructions. All reactionsere performed in four replications using 3 �L of the dilute tem-late (50×), 2 �L of each primer (2.5 �M) and 5 �L of 2× Mastermixo final volume of 10 �L. PCR was initiated by 5 min at 95 ◦C, fol-owed by 50 cycles of 95 ◦C for 5 s, 60 ◦C for 5 s, 72 ◦C for 10 s andompleted by melt curve analysis. No-template controls for eachrimer pair were included in each run. The mangosteen elongationactor 1 alpha (Gm ELF, EU274578) was used as an internal control toormalize small differences in template amounts with the forwardrimer 5′-GCC CAA AAG ACC ATC AGA CAA GC-3′ and reverse primer′-CGG AAG GAC CAA AAG TGA CAA CC-3′. Gm ELF was selected forormalization because of its consistent transcript level throughouthe fruit samples with crossing threshold (Ct) values changing by2. The standard curve was generated for each gene using cDNAerial dilution (at least 5 dilutions) and the resulting PCR efficiencyalculations were imported into relative expression data analysis.

.8. Statistics

Ten mangosteen fruit comprised one replicate, with three repli-ates being used in each treatment. Five mangosteen fruit wereandomly sampled for determination of firmness. The remainingruit were pooled together for determination of other character-stics. Data were compared by t-test. Differences at P < 0.05 wereonsidered as significant.

. Results

.1. Changes in fruit firmness and phenolic/lignin content aftermpact

Impacted mangosteen fruit showed a color change in the dam-ged pericarp within 10 min. This developed to almost black 60 minfter impact (Fig. 2A), confined to the damaged area. Non-impactednd impacted pericarp differed in firmness; impacted pericarprmness rapidly increased over the period of 30 min to 1 day, whilermness of non-impacted pericarp remained stable (Fig. 2B).

The impacted pericarp had lower total free phenolic contentshan non-impacted pericarp. Total free phenolic contents of thempacted pericarp gradually decreased throughout the studyeriod, unlike that of the non-impacted pericarp (Fig. 2C). Ligninontent of impacted pericarp increased continuously throughouthe period after impact, showing significant differences above thatf the non-impacted pericarp (Fig. 2D).

.2. Monolignol content

GC–MS was used to measure the nitrobenzene oxidation prod-cts of cell walls after extraction from non-impacted and impacted

ericarp. Nitrobenzene oxidation degrades the phenylpropanetructure from a C6–C3 to a C6–C1 unit. The degraded prod-cts are p-hydroxybenzaldehyde, vanillin, syringaldehyde andhe corresponding acids p-hydroxybenzoic acid, vanillic acid, and

Fig. 2. Changes in color (A), firmness (B), total free phenolics (C) and lignin content(D) of mangosteen pericarp with (�) and without impact (♦). Data are means ± SEof three replicates.

syringic acid. We confirmed the formation of nitrobenzene oxi-dation products which showed the retention times of authenticstandards p-hydroxybenzaldehyde, vanillin, acetovanillone (inter-nal standard), p-hydroxybenzoic acid, syringaldehyde, vanillic acid,syringic acid in mangosteen pericarp tissue with and withoutimpact (Fig. S1; Supplementary material). The products were iden-tified by comparison of their retention times and mass spectra withstandards (Table S2; Supplementary material).

GC–MS result showed higher G (vanillin and vanillic acid) and S(syringaldehyde and syringic acid) lignin in the impacted pericarp

than the non-impacted pericarp (Fig. 3A and B). The vanillin levelswere 5-fold higher than vanillic acid (Fig. S2A and B Supplemen-tary material). Similarly the syringaldehyde levels were higher thansyringic acid in both non-impacted and impacted pericarp (Fig. S3A

72 C. Kamdee et al. / Postharvest Biology and Technology 97 (2014) 68–76

Time after impact (min)

To

tal G

lig

nin

mm

ol kg

-1

To

tal S

lig

nin

mm

ol kg

-1

**

**

0

50

100

150

200

1801206030150

0

50

100

150

200 A

B

F(i

ata

3l

tsoispTa

tnsa

3t

ct(F(pIt

Fir

mn

ess

(N)

Lig

nin

co

nte

nt

(Ab

s280)

To

tal fr

ee p

hen

olics

(mg

g-1

)

To

tal G

lig

nin

(mm

ol

kg

-1)

To

tal S

lig

nin

Time after impact (h)

(mm

ol

kg

-1)

0

10

20

30

40

0

0.2

0.4

0.60

3

6

9

0

40

80

120

160

200

0

40

80

120

160

200

0 1 2 3 4 5 6 12 18 24

B

C

**

**** *

***

**

**

D

**ns

**

**

**

****

*** ns

E

* *

A

E

Fig. 4. Changes after impact in firmness (A), total free phenolic (B) and total lignin

ig. 3. Changes in lignin monomer composition as total G (vanillin + vanillic acid)A) and S (syringaldehyde + syringic acid) (B) lignin in mangosteen pericarp aftermpact (�) or control (♦). Data are means ± SE of three replicates.

nd B; Supplementary material). Interestingly, G lignin concentra-ions were twice as high as S lignin in the impacted pericarp (Fig. 3And B).

.3. Changes in fruit firmness, total free phenolic contents andignin of impacted pericarp under enhanced nitrogen atmosphere

The firmness of impacted pericarp kept under nitrogen was lesshan that of impacted pericarp kept in air, and remained relativelytable from 0 to 6 h, then increased to almost the same extent as thatf fruit kept in air (Fig. 4A). The levels of total free phenolic contentsn impacted pericarp kept in ambient air and nitrogen declined inimilar patterns (Fig. 4B). However, lignin content of the impactedericarp kept in nitrogen showed only a slight increase (Fig. 4C).here was a large difference in lignin content between pericarp inir and nitrogen by the end of the experiment (Fig. 4C).

Analysis of lignin monomer composition by nitrobenzene oxida-ion indicated that G and S lignin in impacted pericarp kept underitrogen were lower than in air (Fig. 4D and E). The vanillin andyringaldehyde levels in impacted pericarp kept in nitrogen werelso lower than those kept in ambient air.

.4. Isolation of lignin biosynthesis genes and a R2R3 MYBranscription factor

Using mangosteen pericarp cDNA we isolated and partiallyharacterized genes involved in different steps of lignin biosyn-hesis; PAL (FJ197127), C4H (KJ671478), 4CL (KJ671474), HCTKJ671470), C3H (KJ671469), CCR (KJ671477), CCoAMT (KJ671473),5H (KJ671476), COMT (KJ671475), CAD (KJ671471) and POD

KJ671472). All genes showed considerable homology to otherlant lignin biosynthesis genes (Table S3; Supplementary material).

n order to isolate and characterize potential MYB transcription fac-ors involved in lignin biosynthesis, two degenerate primers were

contents (C), lignin monomer composition as total G (vanillin + vanillic acid) (D) andS (syringaldehyde + syringic acid) (E) lignin in mangosteen pericarp after impact andthen kept in ambient air (�) and nitrogen atmosphere (�). Data are means ± SE ofthree replicates.

designed to the R2R3 MYB domain. Putative R2R3MYB genes wereisolated by 3′ race and full-length obtained by 5′ race. The full-length cDNA was termed GmMYB30 (KJ671479) and was a 1377 bptranscript encoding a predicted protein of 381 amino acids whichshared high homology to Arabidopsis MYB30 and MYB94 (TAIR,http://www.arabidopsis.org/cgi-bin/wublast/wublast access 11April 2014).

3.5. Differential gene expression in fruit pericarp after impact

Expression of the genes encoding enzymes of the lignin biosyn-thesis pathway, as well as an R2R3 MYB transcription factor,was examined using quantitative real-time RT-PCR (Fig. 5). The

C. Kamdee et al. / Postharvest Biology and Technology 97 (2014) 68–76 73

0.0

0.5

1.0

1.5

2.0

0.0

1.0

2.0

3.0

0.0

0.2

0.4

0.6

0.0

0.5

1.0

1.5

0.0

0.5

1.0

1.5

2.0

0.0

1.0

2.0

3.0

0.0

0.1

0.2

0.3

0.0

0.3

0.6

0.9

1.2

0.0

1.0

2.0

3.0

0.0

0.3

0.6

0.9

0.0

0.4

0.8

1.2

0.0

1.0

2.0

3.0

Time after impact (min, d)

Re

lati

ve

ex

pre

ss

ion

(A) PAL

(B) C4H

(D) HCT

(C) 4CL

(E) C3 H

(F) CC R

(K) POD

(L) MYB3 0

(J) CAD

(G) CCoAMT

(H) F5H

(I) COMT

Non-impacted

Impacted

Non-impacted

Impacted

12050 25 201510 60 1d30 180 12050 25 201510 60 1d30 180

Fig. 5. Expression profiling of lignin biosynthesis and MYB transcription factors genes in mangosteen pericarp after impact (�), or control (�). Real-time PCR was used toanalyze GmPAL (A), GmC4H (B), Gm4CL (C), GmHCT (D), GmC3H (E), GmCCR (F), GmCCoAMT (G), GmF5H (H), GmCOMT (I), GmCAD (J), GmPOD (K) and MYB30 (L) expresspatterns. The transcription level was calculated with respect to ELF level.

7 logy a

tbGs

Fa(

4 C. Kamdee et al. / Postharvest Bio

ranscript levels of all genes were detectable in the pericarp ofoth impacted and non-impacted fruit. The expression patterns ofmPAL, GmHCT, GmC4H and GmC3H, GmCCR (Fig. 5) after impacthowed lower transcript levels than in the non–impacted fruit

0.0

0.5

1.0

1.5

0.0

1.0

2.0

3.0

0.0

0.2

0.4

0.6

0.0

0.5

1.0

1.5

2.0

0.0

0.5

1.0

1.5

2.0

0.0

1.0

2.0

3.0

Time af t

(A) PAL

(B) C4H

(C) 4CL

(D) HCT

(E) C3 H

(F) CC R

Nitroge n

Air

Re

lati

ve

ex

pre

ss

ion

61 50 43 422 18 21

ig. 6. Expression profiling of lignin biosynthesis and MYB transcription factors genes intmosphere (�). Real-time PCR was used to analyze GmPAL (A), GmC4H (B), Gm4CL (C), GmJ), GmPOD (K) and MYB30 (L) express patterns. The transcription level was calculated wi

nd Technology 97 (2014) 68–76

during the same time course. The expression of these genesdecreased between 60 min and 1 day after impact. This subsetof genes showed a negative correlation between the transcriptlevel and lignin content (r = 0.76, 0.70, 0.46 and 0.73, respectively)

0.0

0.2

0.4

0.6

0.0

1.0

2.0

3.0

0.0

1.0

2.0

3.0

0.0

1.0

2.0

3.0

0.0

0.5

1.0

1.5

0.0

0.5

1.0

1.5

2.0

er impact (h)

(G) CCoAMT

(H) F5H

(I) COMT

(J) CAD

(K) POD

(L) MYB30

Nitrogen

Air

61 50 43 422 1812

mangosteen pericarp after impact and then kept in ambient air (�) and nitrogenHCT (D), GmC3H (E), GmCCR (F), GmCCoAMT (G), GmF5H (H), GmCOMT (I), GmCAD

th respect to ELF level.

logy a

asssc

toiwRie

3n

haihs(

sgt

4

pttliw(iaK2

ucie2t(iaii

aitYpWPes

C. Kamdee et al. / Postharvest Bio

nd firmness (r = 0.77, 0.80, 0.60 and 0.84, respectively). The tran-cript levels of GmCCR, GmCCoAMT, GmPOD and GmCAD (Fig. 5)howed little correlation with lignin content and firmness, andhowed variable differences between impacted and control peri-arps (Figs. 4 and 5).

In contrast, the expression of GmF5H (Fig. 5H) in the pericarp ofhe non-impacted fruit remained at a relatively low level through-ut the time course, while in impacted fruit transcription levelsncreased from 30 min to 1 day after impact. This correlated well

ith increasing firmness and lignin content. The expression of the2R3 MYB GmMYB30 (Fig. 5) in impacted fruit showed a significant

ncrease in transcript level after 25 min. This increase in MYB30xpression was not seen in non-impacted fruit.

.6. Differential gene expression in impacted fruit under elevateditrogen atmosphere

The expression of GmC4H, GmC3H and GmCCR (Fig. 6) in fruiteld under nitrogen showed a negative correlation with firmnessnd lignin content. GmF5H (Fig. 6) expression gradually increasedn both atmosphere treatments. However, fruit held in air showedigher GmF5H transcript levels at 5, 18 and 24 h. This expressionhowed a positive correlation with the firmness and lignin contentFig. 4).

The expression of GmMYB30 (Fig. 6) in the fruit held in airhowed higher expression earlier in air than under elevated nitro-en. The increase in transcript levels of this MYB was delayed byhe nitrogen conditions.

. Discussion

In this study we examined the effect of impact on mangosteenericarp tissue. Data on firmness, lignin content, lignin monomers,otal free phenolics, and genes which are candidates for controllinghis rapid response, were integrated. Fruit firmness and pericarpignin content showed a rapid increase after impact. The first phys-ological change was a change in color of the pericarp (Fig. 2),

ith firmness and lignin content being altered 30 min after impactFig. 2B and C). Pericarp hardening is clearly associated with anncrease in firmness and lignin content and is concomitant with

decrease in total free phenolics (Ketsa and Koolpluksee, 1993;etsa and Atantee, 1998; Bunsiri et al., 2003; Dangcham et al.,008).

An increase in fruit firmness is an uncommon postharvest text-ral change, which has been reported in only relatively few fruitrops. This trait is caused by cell wall secondary lignification,nduced as a response to stress (Ketsa and Atantee, 1998; Boerjant al., 2003; Boudet et al., 2003; Bunsiri et al., 2003; Moura et al.,010). These lignins are synthesized via the oxidative coupling ofhree monolignols, p-hydroxyphenyl (H), guaiacyl (G), and syringylS). Our analysis of mangosteen monolignols revealed that thempacted fruit had higher levels of G (vanillin and vanillic acid)nd S (syringaldehyde and syringic acid) monolignols than the non-mpacted fruit (Fig. 4A and B), which accounted for most of thencrease in lignin content (Fig. 2D).

Lignin biosynthesis is not only developmentally regulated butlso induced in response to stresses such as wounding, UV lightrradiation, and pathogen attack. Less is known about the transcrip-ional activation of stress-induced lignin biosynthesis (Zhong ande, 2009). Plants exposed to different stresses stimulate phenyl-ropanoid metabolism to change lignin content and composition.

ounding induces genes related to lignin biosynthesis, such as

AL, C4H, F5H, CAD, CCR and 4CL (Delessert et al., 2004; Soltanit al., 2006; Moura et al., 2010), resulting in accumulation of ligninurrounding the wound site. Phenolics are the main substrate

nd Technology 97 (2014) 68–76 75

for lignin synthesis. The decline of p-coumaric and sinapic acidsin response to impact was associated with the increase in totallignin content. These changes may be linked with reduction in thesoluble phenolics pool as lignin is synthesized (Vanholme et al.,2010). Furthermore, p-coumaric acid decreased more rapidly inimpacted fruit than sinapic acid (Bunsiri et al., 2003).

Damaged mangosteen kept under elevated nitrogen had low-ered lignification and hence a lower firmness, compared to fruitin ambient air. In contrast there was a higher level of total freephenolics after impact. This confirms previous reports (Ketsa andAtantee, 1998; Bunsiri et al., 2003). It is understood that lignifica-tion is an oxidative process; the last major step in lignin synthesisinvolves monolignol polymerization, which uses many oxidativeenzymes, e.g. peroxidase, laccases or phenol oxidases. After theactivation of monolignols by these enzymes, oxidized monolignolradicals couple to form three dimensionally cross-linked structures.This polymerization constitutes the final step of lignin biosynthe-sis. These enzymes use oxygen and reactive oxygen species for thereaction, and therefore lowering O2 levels will affect monolignolpolymerization reducing lignification (Imberty et al., 1985). Theobservation that elevated nitrogen did not prevent changes in F5Hexpression (Fig. 6) also suggests nitrogen may be influence ligninvia enzymes such as peroxidises.

In previous research on fruit lignification, loquat provides themost comprehensive set of observations on lignin and monolig-nol precursors, and their associated enzymes and genes (Shanet al., 2008). The expression of genes encoding cinnamyl alcoholdehydrogenase (CAD) and peroxidase (POD), were most closelyassociated with loquat flesh lignification. EjCAD1 expression wasstimulated by low temperature, which may contribute to low tem-perature injury in fruit (Shan et al., 2008). In mangosteen, thepattern of gene expression of F5H, COMT, CAD and POD in impactedfruit showed some increases over non-impacted fruit. However,the expression of genes encoding early steps of lignin biosynthe-sis decreased following impact. F5H showed a strong increase inexpression in impacted fruit from 30 min to 1 day after impact. F5His a cytochrome P450 dependent monooxygenase that catalyses thehydroxylation of ferulic acid, coniferaldehyde and coniferyl alco-hol, leading to sinapic acid and syringyl lignin biosynthesis (Fig. 1).In leaves, F5H has been shown to be induced by wounding andsenescence (Ruegger et al., 1999; Kim et al., 2013).

The expression of stress-induced genes in plants is largelyregulated by specific transcription factors. Transcription factorsbelonging to NAC, MYB and WRKY gene families, have been shownto regulate the lignin biosynthetic pathway in various plant species(Vanholme et al., 2010). MYB proteins have been shown to beinvolved in many significant physiological and biochemical pro-cesses, including regulation of primary and secondary metabolism,cell development and cell cycle, participation in defense andresponse to various biotic and abiotic stresses (Du et al., 2009).Some of these MYB transcription factors have been shown to regu-late the entire phenylpropanoid metabolism, and the others wereproposed to specifically regulate the lignin biosynthesis (Zhong andYe, 2009; Zhao and Dixon, 2011).

We isolated full length coding sequence of a MYB transcriptionfactor, GmMYB30 which is homologous to two genes in Arabidop-sis (AtMYB30 and AtMYB94) that are involved in stress response(Vailleau et al., 2002; Nikiforova et al., 2003). The expression ofGmMYB30 (Fig. 5) in impacted fruit showed a significant increaseafter 20–120 min. GmMYB30 and F5H showed low expression innon-impacted fruit, whereas GmMYB30 increased in 20 min, whilethe expression of F5H gradually increased 25 min after impact.

GmMYB30 showed higher expression in ambient air more than inN2 atmosphere at 1 h after impact (Fig. 6). The results show thatelevated N2 correlates with down-regulation of GmMYB30 whichcould then affect transcription of biosynthetic genes.

7 logy a

iwbtMbi

A

JKCbf

A

i2

R

B

B

B

B

B

C

D

D

D

D

D

I

J

K

K

6 C. Kamdee et al. / Postharvest Bio

In conclusion, pericarp hardening in mangosteen fruit aftermpact was associated with an increase in lignin biosynthesis,

hich in this fruit consists mainly of G and S lignin. GmF5H maye a key biosynthetic gene in mangosteen fruit. The isolated MYBranscription factor (GmMYB30) was homologous with Arabidopsis

YB transcription factors involved in cell wall thickening, ligniniosynthesis and stress response. Elevated N2 delayed an increase

n firmness and retarded lignin content increases.

cknowledgments

The research was financially supported by the Royal Goldenubilee PhD program, Thailand Research Fund (PHD/0317/2550),asetsart University, and the Postharvest Technology Innovationenter (KU.RD.7/2554). Technical assistance and advice providedy staff at The New Zealand Institute for Plant & Food Ltd. is grate-ully acknowledged.

ppendix A. Supplementary data

Supplementary data associated with this article can be found,n the online version, at http://dx.doi.org/10.1016/j.postharvbio.014.06.004.

eferences

oerjan, W., Ralph, J., Baucher, M., 2003. Lignin biosynthesis. Annu. Rev. Plant Biol.54, 519–546.

onawitz, N.D., Chapple, C., 2010. The genetics of lignin biosynthesis: connectinggenotype to phenotype. Annu. Rev. Genet. 44, 337–363.

oudet, A.M., Kajita, S., Grima-Pettenati, J., Goffner, D., 2003. Lignins and lignocel-lulosics: a better control of synthesis for new and improved uses. Trends PlantSci. 8, 576–581.

ruce, R.J., West, C.A., 1989. Elicitation of lignin biosynthesis and isoperoxidase activ-ity by pectic fragments in suspension cultures of castor bean. Plant Physiol. 91,889–897.

unsiri, A., Ketsa, S., Paull, R.E., 2003. Phenolic metabolism and lignin synthesis indamaged pericarp of mangosteen fruit after impact. Postharvest Biol. Technol.29, 61–71.

ai, C., Xu, C., Shan, L., Li, X., Zhou, C., Zhang, W., Ferguson, I., Chen, K., 2006. Lowtemperature conditioning reduces postharvest chilling injury in loquat fruit.Postharvest Biol. Technol. 41, 252–259.

angcham, S., Bowen, J., Ferguson, I.B., Ketsa, S., 2008. Effect of temperature and lowoxygen on pericarp hardening of mangosteen fruit stored at low temperature.Postharvest Biol. Technol. 50, 37–44.

elessert, C., Wilson, I.W., Van Der Straeten, D., Dennis, E.S., Dolferus, R., 2004. Spatialand temporal analysis of the local response to wounding in Arabidopsis leaves.Plant Mol. Biol. 55, 165–181.

onaldson, L.A., 2001. Lignification and lignin topochemistry an ultrastructuralview. Phytochemistry 57, 859–873.

orria, M.A., Yousef, A.R.M., Sarrwy, S.M.A., 2011. Modified atmosphere packagingfor maintaining quality and shelf life extension of persimmon fruits. Asian J.Agric. Sci. 3, 308–316.

u, H., Zhang, L., Liu, L., Tang, X.F., Yang, W.J., Wu, Y.M., Huang, Y.B., Tang, Y.X., 2009.Biochemical and molecular characterization of plant MYB transcription factorfamily. Biochemistry (Moscow) 74, 1–11.

mberty, A., Goldberg, R., Catesson, A.-M., 1985. Isolation and characterization ofPopulus isoperoxidases involved in the last step of lignin formation. Planta 164,221–226.

arimopas, B., Pushpariksha, P., Singh, S.P., 2009. Postharvest damage of mangosteenand quality grading using mechanical and optical properties as indicators. Int. J.Food Prop. 12, 414–426.

etsa, S., Atantee, S., 1998. Phenolics, lignin, peroxidase activity and increased firm-

ness of damaged pericarp of mangosteen fruit after impact. Postharvest Biol.Technol. 14, 117–124.

etsa, S., Koolpluksee, M., 1993. Some physical and biochemical characteristics ofdamaged pericarp of mangosteen fruit after impact. Postharvest Biol. Technol.2, 209–215.

nd Technology 97 (2014) 68–76

Kim, J., Choi, B., Park, Y.H., Cho, B.K., Lim, H.S., Natarajan, S., Park, S.U., Bae, H., 2013.Molecular characterization of ferulate 5-hydroxylase gene from Kenaf (Hibiscuscannabinus L.). Sci. World J. 2013, 1–11.

Ko, J.H., Kim, W.C., Han, K.H., 2009. Ectopic expression of MYB46 identifies transcrip-tional regulatory genes involved in secondary wall biosynthesis in Arabidopsis.Plant J. 60, 649–665.

Lange, B.M., Lapierre, C., Sandermann Jr., H., 1995. Elicitor-induced spruce stresslignin (structural similarity to early developmental lignins). Plant Physiol. 108,1277–1287.

Li, X., Chapple, C., 2010. Understanding lignification: challenges beyond monolignolbiosynthesis. Plant Physiol. 154, 449–452.

Liu, Z.Y., Jiang, W.B., 2006. Lignin deposition and effect of postharvest treatment onlignification of green asparagus (Asparagus officinalis L.). Plant Growth Regul. 48,187–193.

López-Gómez, R., Gómez-Lim, M.A., 1992. A method for extracting intact RNAfrom fruits rich in polysaccharides using ripe mango mesocarp. HortScience 27,440–442.

Luo, Z., Xu, X., Cai, Z., Yan, B., 2007. Effects of ethylene and 1-methylcyclopropene(1-MCP) on lignification of postharvest bamboo shoot. Food Chem. 105,521–527.

Maldonado, R., Molina-Garcia, A.D., Sanchez-Ballesta, M.T., Escribano, M.I., Merodio,C., 2002. High CO2 atmosphere modulating the phenolic response associatedwith cell adhesion and hardening of Annona cherimola fruit stored at chillingtemperature. J. Agric. Food Chem. 50, 7564–7569.

Meyer, K., Shirley, A.M., Cusumano, J.C., Bell-Lelong, D.A., Chapple, C., 1998. Ligninmonomer composition is determined by the expression of a cytochromeP450-dependent monooxygenase in Arabidopsis. Proc. Natl. Acad. Sci. USA 95,6619–6623.

Moura, J.C.M.S., Bonine, C.A.V., De Oliveira Fernandes Viana, J., Dornelas, M.C., Mazza-fera, P., 2010. Abiotic and biotic stresses and changes in the lignin content andcomposition in plants. J. Integr. Plant Biol. 52, 360–376.

Nikiforova, V., Freitag, J., Kempa, S., Adamik, M., Hesse, H., Hoefgen, R., 2003. Tran-scriptome analysis of sulfur depletion in Arabidopsis thaliana: interlacing ofbiosynthetic pathways provides response specificity. Plant J. 33, 633–650.

Palapol, Y., Ketsa, S., Stevenson, D., Cooney, J.M., Allan, A.C., Ferguson, I.B., 2009.Colour development and quality of mangosteen (Garcinia mangostana L.) fruitduring ripening and after harvest. Postharvest Biol. Technol. 51, 349–353.

Ruegger, M., Meyer, K., Cusumano, J.C., Chapple, C., 1999. Regulation of ferulate-5-hydroxylase expression in Arabidopsis in the context of sinapate esterbiosynthesis. Plant Physiol. 119, 101–110.

Shan, L.L., Li, X., Wang, P., Cai, C., Zhang, B., Sun, C.D., Zhang, W.S., Xu, C.J., Fergu-son, I., Chen, K.S., 2008. Characterization of cDNAs associated with lignificationand their expression profiles in loquat fruit with different lignin accumulation.Planta 227, 1243–1254.

Shen, Q., Kong, F., Wang, Q., 2006. Effect of modified atmosphere packaging on thebrowning and lignification of bamboo shoots. J. Food Eng. 77, 348–354.

Soltani, B.M., Ehlting, J., Hamberger, B., Douglas, C.J., 2006. Multiple cis-regulatoryelements regulate distinct and complex patterns of developmental and wound-induced expression of Arabidopsis thaliana 4CL gene family members. Planta 224,1226–1238.

Tongdee, S.C., Suwanagul, A., 1998. Postharvest mechanical damage in mangosteen.ASEAN Food J. 4, 151–155.

Vailleau, F., Daniel, X., Tronchet, M., Montillet, J.-L., Triantaphylidès, C., Roby, D.,2002. A R2R3-MYB gene, AtMYB30, acts as a positive regulator of the hypersen-sitive cell death program in plants in response to pathogen attack. Proc. Natl.Acad. Sci. USA 99, 10179–10184.

Vanholme, R., Demedts, B., Morreel, K., Ralph, J., Boerjan, W., 2010. Lignin biosyn-thesis and structure. Plant Physiol. 153, 895–905.

Waterhouse, A.L., 2002. Determination of total phenolics. Curr. Protoc. Food Anal.Chem., I1.1.1–I1.1.8.

Whetten, R.W., MacKay, J.J., Sederoff, R.R., 1998. Recent advances in understandinglignin biosynthesis. Annu. Rev. Plant Physiol. Plant Mol. Biol. 49, 585–609.

Zhang, S.H., Yang, Q., Ma, R.C., 2007. Erwinia carotovora ssp. carotovora infectioninduced defense lignin accumulation and lignin biosynthetic gene expressionin Chinese cabbage (Brassica rapa L. ssp. pekinensis). J. Integr. Plant Biol. 49,993–1002.

Zhao, Q., Dixon, R.A., 2011. Transcriptional networks for lignin biosynthesis: morecomplex than we thought? Trends Plant Sci. 16, 227–233.

Zhong, R., Lee, C., Zhou, J., McCarthy, R.L., Ye, Z.-H., 2008. A battery of transcrip-tion factors involved in the regulation of secondary cell wall biosynthesis inArabidopsis. Plant Cell 20, 2763–2782.

Zhong, R., Ye, Z.-H., 2009. Transcriptional regulation of lignin biosynthesis. PlantSignal. Behav. 4, 1028–1034.

Zhou, J., Lee, C., Zhong, R., Ye, Z.-H., 2009. MYB58 and MYB63 are transcriptional acti-vators of the lignin biosynthetic pathway during secondary cell wall formationin Arabidopsis. Plant Cell 21, 248–266.