An integrated strategy to identify key genes in almond adventitious shoot regeneration

15 Pages·2009·0.76 MB·English

JournalofExperimentalBotany,Vol.60,No.14,pp.4159–4173,2009
doi:10.1093/jxb/erp250 AdvanceAccesspublication11August,2009
RESEARCH PAPER
An integrated strategy to identify key genes in almond
adventitious shoot regeneration
AnaMargaridaSantos1,MelvinJohnOliver2,AnaMariaSa´nchez1,PaxtonRobertPayton3,Joa˜oPauloGomes4,
Ce´lia Miguel1 and M. Margarida Oliveira1,*
D
o
w
1 InstitutodeTecnologiaQuı´micaeBiolo´gica/InstitutodeBiologiaExperimentaleTecnolo´gica,QuintadoMarqueˆs,2784-505Oeiras, n
lo
Portugal a
d
2 UnitedStatesDepartmentofAgriculture-ARS,PlantGeneticsResearchUnit,205CurtisHall,UniversityofMissouri,Columbia,MO ed
65211,USA fro
m
3 UnitedStatesDepartmentofAgriculture-ARS,38104thSt,Lubbock,TX79415,USA h
4 InstitutoNacionaldeSau´de,AvenidaPadreCruz,1649-016Lisbon,Portugal ttp
s
://a
c
Received12May2009;Revised13July2009;Accepted29July2009 a
d
e
m
ic
.o
u
Abstract p
.c
o
m
Plant genetic transformation usually depends on efficient adventitious regeneration systems. In almond (Prunus
/jx
dulcis Mill.), regeneration of transgenic adventitious shoots was achieved but with low efficiency. Histological b
/a
studies identified two main stages of organogenesis in almond explants that were induced for adventitious shoot rtic
regeneration; a dedifferentiation stage (early) and a shoot initiation stage (late). Histological observation revealed le
-a
that the limitation in the recovery of transformed shoots is primarily a function of the low organogenic competence bs
of the transformed tissues rather than transformation efficiency. To identify key genes involved in organogenesis, tra
c
shoot-induced leaves and suppression-subtractive hybridization were used, to build a cDNA library from each t/6
0
organogenic stage. cDNA clones from both libraries were randomly picked, PCR-amplified, and arrayed on glass /1
4
slides. For transcript profiling, microarray hybridization was performed using cDNA pools from both the early and /4
1
5
thelatestages.Statisticallysignificantdifferentialexpressionwasfoundfor128cDNAclones(58early,and70late), 9
/6
representing 92 unique gene functions. Genes encoding proteins related to protein synthesis and processing and 5
6
1
nitrogen and carbon metabolism were differentially expressed in the early stage, whilst genes encoding proteins 4
7
involved in plant cell rescue and defence and interaction with the environment were mostly found in the late stage. b
y
The LTP/a-amylase inhibitor/trypsin gene was more strongly expressed at an early stage, as confirmed by g
u
e
quantitative RT-PCR, while a gibberellic acid stimulated protein geneseems to be agood marker forthe latestage. s
t o
These results are discussed on the basis of the putative roles of the annotated differentially regulated genes in n
1
almondorganogenesis. 9
N
o
v
Keywords: Adventitiousshoot,almond,earlystage(ES),latestage(LS),quantitativePCR,suppressionsubtractivehybridization em
(SSH),transcriptprofiling. be
r 2
0
1
8
Introduction
Upon the application of appropriate exogenous signals, revert, re-enter the cell cycle, and express its pluripotency
meristematic and differentiated plant cells in culture are resulting in the regeneration of a new whole plant. For the
sometimes able to modify their fate, behaving as true stem almond industry to profit from genetic transformation
cells. This plasticity has given rise to the concept that plant technologies, efficient and reproducible adventitious regen-
cells are totipotent. Some of the signals perceived by the eration systems are essential. An efficient protocol for
plant are so strong that they can induce a mature cell to somatic embryogenesis has not been established for almond
*Towhomcorrespondenceshouldbeaddressed:E-mail:[email protected]
ªTheAuthor[2009].PublishedbyOxfordUniversityPress[onbehalfoftheSocietyforExperimentalBiology].Allrightsreserved.
ForPermissions,pleasee-mail:[email protected]
4160 | Santos et al.
mature tissues, however, adventitious shoot regeneration gene (histidine kinase domain-containing) CKI1 (cytokinin-
has been accomplished using leaf explants of mature and independent1) from a mutant tagged line, that when
juvenile ex vitro or in vitro shoots (Miguel et al., reintroduced and over-expressed, promotes green callus
1996; Ainsley et al., 2000, 2001). At present, only two andshootformation intheabsence ofexogenouscytokinin.
genotypes can be transformed, a seedling obtained from the Functional screening of an Arabidopsis cDNA library
Portuguese ‘Boa Casta’ (Miguel and Oliveira, 1999; Costa allowed the identification of a novel gene, Enhancer of
et al., 2006) and the North American ‘Ne Plus Ultra’ Shoot Regeneration (ESR1), that encodes a putative tran-
(Ramesh et al.,2006). scription factor whose over-expression leads to cytokinin-
Withtheprogressinplantgeneticsandmolecularbiology independent shoot formation in root explants (Banno et al.,
(often using Arabidopsis mutants and tagged probes to 2001). Cary et al. (2001) described Arabidopsis mutants
mark events), our understanding of in vivo shoot meristem with an increased capacity for shoot formation in tissue
development and plant cell cycle regulation has progressed culture. One of these mutants established a new locus
significantly. In shoot meristems, several genes regulating named increased organ regeneration1 (ire1). IRE may down-
D
development have been identified including the Zea mays regulate the competence of vegetative tissue to respond to o
w
(maize) KNOTTED-1 (KN1) (Volbrecht et al., 1991), hormonal signals involved in shoot and root organogenesis. nlo
Arabidopsis thaliana SHOOTMERISTEMLESS (STM) More recently, an Arabidopsis mutant named hoc (High ad
e
C(LLoAnVgAetTaAl.1, 1(C99la6r)k, WetUaSl.,C1H9E93L, 1(W99U7)S,)C(LLAauVxAeTtAa2l.,(K19a9y6e)s, OcarpgaacniotygefnoirciCnavpiatrcoitys)hwooatsrdeegteenrmerianteiodnt.oHhoacveroaontinecxrpelaasnetds d from
and Clark, 1998), and CLAVATA3 (Fletcher et al., 1999). developshootsinvitrowithouttheexogenousapplicationof h
ttp
Also, the Arabidopsis genes CUP-SHAPED COTYLE- growth regulators (Catterou et al., 2002). Prakash and s
DON1 and 2 (CUC1 and CUC2), have become useful Kumar (2002) isolated a novel MADS box cDNA, ://a
c
a
markers for following de novo shoot development. These PkMADS1, from a Paulownia kawakamii (a woody-tree d
e
genes, requiredforthepropershootapical meristem (SAM) species) cDNA library constructed from leaf explants m
ic
formation, are expressed during post-embryonic develop- induced for adventitious shoot formation. The expression .o
u
ment(Aidaetal.,1997;Takadaetal.,2001).Theexpression of PkMADS1 was restricted to shoot apices in planta and p.c
o
of KN1 in maize, and its homologue in barley has been shoot-forming in vitro cultures, thus suggesting an essential m
chronicled during in vitro axillary shoot meristem pro- role in shoot formation. /jxb
liferation and adventitious shoot meristem (ADM) forma- A genomic approach to analyse in vitro shoot organogen- /artic
tion (Zhang et al., 1998). The ADMs appeared to derive esis from roots, in Arabidopsis, has also been successful. le
directly from KN1-expressing shoot meristematic cells. In Using the Affymetrix GeneChip(cid:1), which consists of a pop- -ab
s
Brassica, expression of Brostm (a KN1 homologue) marked ulation of oligonucleotides that represent the complete tra
c
the switch from developmentally fated to shoot meriste- genome, a gene expression profile for in vitro shoot t/6
0
maticcells(Teoetal.,2001).Cellcycle-relatedgenessuchas morphogenesis identified the expression of Arabidopsis Re- /1
4
thecyclin-dependentkinasetypeA(CDKA:1)havealsobeen sponse Regulator5 (ARR5), type A response regulator gene. /4
1
extensivelystudiedasdevelopmentalmarkersinArabidopsis It was up-regulated by 7-fold at the time of shoot 5
9
(Ferreira et al., 1991) and in several other species such as commitment, and its expression was localized to sites of /6
5
6
Pisum sativum (pea) (Feiler and Jacobs, 1991), and maize presumptive shoot formation. Two hybrid His Kinases 1
4
(Colasanti et al., 1991). The expression of CDKA:1 was involved in cytokinin responses, CRE1, which encodes 7 b
y
observed mainly in actively proliferating cells such as those a cytokinin receptor, and CKI1, a gene that is capable of g
u
from shoot meristems and young leaves (Colasanti et al., confering cytokinin-independent shoot development. Mes- e
s
1991; Martinez et al., 1992) leading to the suggestion that sage levels of CKI1 and of the cytokinin receptor encoded t o
n
this gene is a useful molecular marker for the in situ by CRE/AMKA/WOL were enhanced 3-fold upon the 1
9
identification of quiescent cells that are able to re-enter the transfer of explants to shoot induction conditions (Che N
o
cell cycle and follow the morphogenetic pathway (Zhang et al., 2002). Also in Arabidopsis, following the same ve
m
et al., 1998; Teo et al., 2001). It is clear from these studies approach, STM, CUC1, CUC2, and WUS staged expres- b
e
that the identification of a single gene that is central to sion marked the cryptic developmental events of adventi- r 2
0
a particular developmental pathway is a critical step in tious shootinduction (Cary etal.,2002). 1
8
understanding and following any important developmental To understand the sequence of events that take place
event in vivo or in vitro. This is true even though it is also during adventitious regeneration in almond, histology
clear that a single gene is insufficient to support the initial studies were performed. The work described here is not
stages of de novo organogenesis, which is complex and well an exhaustive histology study, but rather a tissue level
co-ordinated. search for the early developmental details that are rele-
Mutants impaired at different stages of organogenesis vant to shoot recovery. Based on these studies, a cDNA
have also been useful in dissecting morphogenic processes. microarray-based expression profiling strategy was com-
In Arabidopsis, temperature-sensitive mutants (srd1, srd2, bined with the construction of suppression subtractive
srd3)thataredefectiveforshootredifferentiationhavebeen hybridization (SSH) libraries (Diatchencko et al., 1996), to
isolated and characterized (Yasutani et al., 1994; Osawa identify differentially expressed transcripts that are specifi-
et al., 1998). Kakimoto (1996) identified an Arabidopsis cally regulated during either the early or late stages of
Uncovering adventitious shoot regeneration in almond | 4161
organogenesis. Selected genes were characterized in real- One microgram of total RNA from each experimental
timeRT-PCRstudiesthroughoutthewholeshootinduction sample (each of the ‘testers’ and the ‘driver’) was used to
period. Almond is a difficult species with regard to the in synthesize cDNA using Power Script(cid:1) reverse transcriptase
vitroregenerationofshootsandwholeplantsfromexplants, according to the procedure provided in the SMART(cid:1)
and as such this study will provide valuable genomic data cDNA synthesis kit (Clontech, Palo Alto, CA, USA). The
for the elucidation of the steps involved in fruit tree suppressionsubtractivecDNAlibrariesoftheearlyandlate
adventitious organogenesis. stages of organogenesis were prepared using the PCR-
Select(cid:1) cDNA subtraction kit (Clontech) according to
manufacturer’s recommendations. The ‘driver’ RNA was
Materials and methods used to subtract both ‘testers’ (early and late stages)
separately and the resultant subtracted cDNAs were ligated
Biological material intopUCR19(O’MahonyandOliver,1999)andclonedinto
DH5a competent cells. The transformed bacteria were
Almondmicropropagatedshootswereobtainedfromaseed- D
grown on Luria Broth (LB) containing ampicillin (100 lg o
lingderivedfrom theopen-pollinationofcv.Boa Castaand w
subcultured every 3 weeks on micropropagation medium as ml(cid:1)1), 5-bromo-4-chloro-3-indolyl, b-D-galactoside (x-Gal, nlo
described by Miguel et al. (1996). For adventitious shoot at 40 lg ml(cid:1)1), and isopropyl-b-D-thiogalactopyranoside ade
ianlmduocntidonsh,otohteswfierrsetefxocuisredexapnadnsdeecdtiolneeadveascroofss3t-hweemekid-orlibd (fIoPrmTGed awthi0t.e1comloMni)esfowrerbelurae-nwdhoimtelyscpriecekneidng(4.5T00heoftreaanrlsy- d from
organogenesis colonies and 2500 of late) and grown as h
without fully separating the leaf pieces. Six to seven ttp
individual overnight cultures in LB liquid media with s
wounded leaves were placed with the adaxial surface facing ampicillin (100 lg ml(cid:1)1) at 37 (cid:2)C and 225 rpm in 96-well ://a
the induction medium, in a 9 mm Petri dish containing 20– ca
U-bottom plates. Glycerol (20% (v/v) final) was added for d
25 ml of culture medium. The explants were maintained in e
storage at –80(cid:2)C. m
darknessat2461 (cid:2)Cfor21 d (Miguelet al., 1996). ic
.o
Agrobacterium strain EHA105 carrying the plasmid u
p
p35SGUSINT was used for transformation of leaf explants .co
m
as described byMiguel and Oliveira (1999). Construction of the cDNA microarray /jx
b
Histological studies Bacteria carrying cDNA clones from the early and late /artic
stagelibrariesweregrownovernightinfreshmediumuntil le
To investigate regeneration patterns, an initial study was growth was confluent. Cells were sampled using a 96-well -ab
s
conducted using scanning electron microscopy (SEM), plate replicator and used directly to PCR-amplify the tra
c
which allows the preservation of large structures. This cDNA inserts using LacI and LacZ flanking primers of t/6
0
facilitates not only observations of the superficial organiza- pUCR19 plasmid vector. The PCR was performed in /1
4
tion of the explants but also, after sectioning, the internal a total of 50 ll of master mix, containing 13 Taq /4
1
organization as well. The areas inside the leaf that were polymerase buffer, 10 mM of MgCl2, 0.25 mM of 59
identified as organogenic were also prepared for light each dNTP, 1 lM of LacI primer (5#-CAgTCACTg- /65
microscopy, using either plastic or paraffin embedding for AgCATgCgCAATCg-3#) and 1 lM of LacZ primer 614
large sections (53332 mm). Procedures utilized for tissue (5#-AgATCCTTTTTATTTTTAATTTTCTTTC-3#) and 7 b
fixation and observation by SEM and light microscopy are 0.75 U ll(cid:1)1 Taq polymerase per 50 ll reaction mixture, in y g
u
describedintheSupplementarydata(ProtocolS1)available a 96-well PCR plate. The PCR conditions consisted of an es
at JXB online. initial denaturation step of 5 min at 94 (cid:2)C, followed by 45 t on
cycles of 94 (cid:2)C for 1 min, 60 (cid:2)C for 1 min, 72 (cid:2)C for 1 1
9
min, with a final step of 7 min at 72 (cid:2)C using an MJ N
RNA isolation and preparation of suppression o
v
Research PTC tetrad-225 thermocycler (Perkin Elmer, e
subtractive cDNA libraries m
Boston, MA, USA). The PCR amplified products were b
e
Leafsamples(100mg)forRNAextractionwerecollectedat analysed by gel electrophoreses (1.2% (w/v) agarose run in r 2
0
1, 2, 5, 8, 9, 13, 15, and 19 d during the 21 d dark 13 TAE). The PCR products were dried onto the surface 18
incubation period for induction of shoot organogenesis. of the wells by placement in a laminar flow hood,
Total RNA was extracted using the RNeasy extraction kit resuspended in printing buffer (33 SSC, 0.2% (wt/v)
(Qiagen, Valencia, CA, USA) according to the manufac- sarcosyl) to a concentration of 250 ng lg ll(cid:1)1, and
turer’s specifications. Two RNA pools were established by transferred to 384-well microtitre plates. The PCR prod-
combining1lgfromeach oftheRNA extracts from days1 ucts representing 3840 anonymous clones (1920 from each
to 8 (the early stages of organogenesis) and from days 9 to library) were arrayed in duplicate blocks onto poly
19 (the late stages of organogenesis) for construction of the L-lysine-treated glass microscope slides using a GMS 417
SSH ‘tester’ or treatment cDNA populations. The ‘driver’ microarrayer (Affymetrix, Santa Clara, CA, USA). The
or control cDNA population was derived from RNA slides were processed and the cDNAs cross-linked as
isolated from non-induced leaves of micropropagated described by Hegde et al. (2000) and stored in the dark at
shoots. room temperature.
4162 | Santos et al.
Fluorescent labelling of target RNA and microarray Reverse transcription for qRT-PCR
hybridization
For array validation, the total RNA from tissue taken for
Total RNA (10 lg) for each target transcript population each day of adventitious shoot induction was extracted and
from replicate RNA samples (three per induction period) combined in equal amounts into two main RNA popula-
were fluorescently labelled using the two-step protocol tions: the early stage of organogenesis from days 1 to 8 and
from the 3DNA Submicro EX Expression Array De- the late stage of organogenesis from days 9 to 20. The
tection Kit (Genisphere, Hatfield, PA, USA) according to cDNA was generated using SuperScript II RT reagents
the manufacturer’s protocol. The labelled target was (Invitrogen, Paisley, UK). Each RNA sample (4 lg) was
hybridized in a controlled water bath at 55 (cid:2)C for 4 h in combined with 1 mM of dNTP, 20 ng of random hexamers
the dark. as primers, and diluted to a final volume of 12 ll with
Arrays were washed at 55 (cid:2)C for 15 min in 23 SSC with distilledwater.Thereactionmixturewasincubatedat65(cid:2)C
0.1% SDS, 10 min at room temperature in 23 SSC, and for 5 min and added to 53 First Strand buffer, 100 mM
10 min at room temperature in 0.53 SSC. Arrays DTT, and 2 U ll(cid:1)1 of RNase Out(cid:1). Following 2 min of D
o
were immediately rinsed in 100% ethanol and dried by incubation at room temperature, 10 U ll(cid:1)1 of SuperScript wn
centrifugation. II Reverse Transcriptase was added to the mixture and loa
d
reverse transcription was allowed to proceed for 2 h at 42 e
d
(cid:2)C. The cDNA synthesis reaction was terminated by fro
m
Data analysis enzyme inactivation at70 (cid:2)C for15 min. h
Forthecandidategeneexpressionstudies,totalRNAwas ttp
s
For the microarray data analysis, image analysis and signal obtained from the leaves at each time point during the 20 ://a
quantification were performed using Imagene software v5.6 d of dark incubation for shoot organogenesis induction. ca
d
(BioDiscovery Inc., CA). Selection criteria for the elimina- cDNA was generated from 250 ng of each RNA sample em
tion of outlying data point spots was as follows: (i) spot using TaqMan RT reagents (Applied Biosystems) supplied ic.o
signal strength–processed signal (minus background) >33 by the manufacturer. The reaction mix contained 5.5 mM up
standard deviation of the mean local background; (ii) spot MgCl , 500 lM dNTPs, 2.5 lM random hexamers (pri- .co
2 m
signal uniformity–pixel to pixel signal variation <20%; (iii) mers), 13 RT buffer, 0.4 U ll(cid:1)1 RNAse inhibitor, and 1.25 /jx
replicate uniformity–mean processed signal (minus back- U ll(cid:1)1 Multiscribe RT in a final volume of 50 ll. The RT b/a
ground) must have a coefficient of variance <0.2; (iv) reaction proceeded for 10 min at 25 (cid:2)C, 30 min at 48 (cid:2)C, rtic
replicate mean signal strength–one channel must have and 5 min at 95 (cid:2)C. The cDNA concentration at the end- le-a
b
processedsignal >1.5%ofthequantifiablemaximum signal. point was measured by spectrophotometry. Each reaction s
Two technical replicates (duplicate spots on the array for yielded approximately 37.5 lg of cDNA. trac
each cDNA) and three biological RNA replicates were used A two-step qRT-PCR assay was employed in an effort to t/60
in the experiments. Data analysis was performed using avoiddifferences inRTreactionefficiencybetweendifferent /14
GeneSpring 5.1 (Silicon Genetics, Inc.). In the t test a P genes under study resulting from possible differences in /41
5
value <0.05 was used to select genes with >2-fold higher hybridization efficiencies of the 3# specific primers (used 9/6
expression inES versus LS. in one-step qRT-PCR). Random hexamers were employed 56
1
in an effort to ensure reproducibility and comparability 47
b
between the genes and time points throughout the 20 y
EST sequencing d period oforganogenesis. gu
e
s
DNA sequencing was performed by the use of a dRhod- t o
n
amine Terminator Cycle Sequencing kit (PE Applied 1
Quantitative PCR reaction (qRT-PCR) array validation 9
Biosystems, Foster City, CA) according to the man- N
o
v
ufacturer’s instructions. Sequence reactions were ana- Eight genes (see Supplementary Table S1 at JXB online) e
m
lysed using a Perkin Elmer/ABI Prism 310 automated that were differentially expressed in either the early or late b
e
sequencer. The cDNA clones were sequenced using a com- stages of organogenesis were selected for a qRT-PCR r 2
0
bination of plasmid and clone specific primers: the validationassaytoconfirmtheexpressionprofilesgenerated 1
8
plasmid primers were designed to match the LacZ by the microarray analysis. The primers used in the qRT-
(CGGGCTCTTCGCTATTAC) and LacI (TTCACACAG- PCRweredesignedusingthePrimerExpressSoftwarefrom
GAAACAGCTATGAC) regions of the pUC19r plasmid Applied Biosystems and are detailed in Supplementary
that are on opposite sides of the multi-cloning site. Table S1 at JXB online. All experiments were performed in
Sequences were manually edited to remove adaptor sequen- duplicate. Quantitative PCR optimization assays were
ces and assembled via Seqman II DNA analysis software performed to ensure the specificity and maximum efficiency
(DNASTAR Madison, WI, USA). The sequences were of the primers for the target and control sequences. Two
assessed and annotated by BLASTN and BLASTX from pairs of primers were tested for each sequence. The qRT-
GenBank Database resources. Matches and hits were PCR was performed using an ABI PRISM 7000 Sequence
considered when expectation value E was equal or inferior Detection System (Applied Biosystems) using SYBR Green
to 4e(cid:1)3. Technology and 96-micro-well optical plates (Applied
Uncovering adventitious shoot regeneration in almond | 4163
Biosystems). Each qRT-PCR reaction mix consisted of 13 the ABI7000SDS software by stepwise increase of the
SYBR Green PCR master mix (provided by the manufac- temperature from 60 (cid:2)C to 90 (cid:2)C.
turer),300nMofeachprimer,and5llofcDNAsample,in
a final volume of 50 ll. The thermal cycling profile was 10
min at 94 (cid:2)C followed by 40 cycles of 15 s at 94 (cid:2)C and 1 Results and discussion
min at 60 (cid:2)C. The specificity of the amplified products was
Histological studies
verified by analysis of the dissociation curves generated by
the ABI PRISM 7000 Software as well as by running the Inanattempttodetailsomeofthedevelopmentalstepsthat
products in 4%(w/v)agarose gel in13 TAE. distinguish organogenesis in regenerating almond leaves, an
exploratory histological study of almond adventitious re-
generation was conducted. After explanting leaves onto
culture media and throughout the 20 d dark incubation
Quantitative PCR reaction (qRT-PCR) for candidate
period, a compact and globular callus tissue developed at
marker gene expression analysis D
the wound surfaces (Fig. 1A–G). Shoot buds emerged from o
w
A total of five genes, based upon their putative identities the calli or from the globular structures (Fig. 1H, I) within nlo
a
and relevance to organogenesis, were chosen for a more 1–3 weeks after transfer to the shoot elongation medium d
e
dasetamilaedrkeexrsprefossriodnevaenlaolpymsisentotaldepterromceisnseest.heTirhepocsasinbdleiduastee unnigdhetr.Athtedsatyan3d,anrodchpahnogtoepsewreiordevciosinbdleitiaotntsheosfu1r6fa/8cehofdtahye/ d from
genes were: EF640654 a putative b-1,3-glucanase gene, leaf explants. By day 5, some cell division had occurred, h
ttp
EF640670 a putative glycine-rich protein gene, EF640686 particularly at the wound site, where many small and s
putative proline-rich protein gene, EF640687 a putative isodiametric cells with thin, irregular cell walls were ://a
c
a
GAST-like protein gene, and EF640716 a putative lipid observed. By day 7, explant size had greatly increased and d
e
transfer protein (LTP)/a-amylase inhibitor/trypsin gene. cell proliferation was clearly observable predominantly at m
ic
Gene-specific primers for real-time PCR were again the midrib along the cut edges. At day 10 (Fig. 1B, F), .o
u
p
designed using Primer Express Software (Applied Biosys- transverse sections close to the wounds revealed groups of .c
o
tems) (see Supplementary Table S2 at JXB online). Exhaus- meristematic cells arranged in cell layers immediately under m
tiveqRT-PCRoptimizationassays wereperformed foreach the adaxial and the abaxial surfaces, and close to the /jxb
gene, to deliver the maximum possible specificity and vasculartissue. These centresdeveloped globular structures, /artic
efficiency with the selected primer pairs. The uniformity of at first still immersed in the mesophyll (Fig. 1G) and later le
the 18SrRNA expression levels throughout the 20 time extruding out of the leaf surface (Fig. 1C, E, H). At day 21, -ab
s
points was confirmed before proceeding with subsequent the leaf structure was hardly recognizable. Strong cell tra
c
qRT-PCRcomparisons.Briefly,cDNAsamplesfromthe20 proliferation in the mesophyll formed a large callus close to t/6
0
time points from each of the three biological replicates were the midrib region and near the wounded surfaces. These /1
diluted to a concentration of 10 ng ll(cid:1)1 and subjected to calli had a loose structure deep inside, but contained 4/4
1
qRT-PCR analysis as described above. The obtained meristematic centres consisting of spherical masses of 5
9
threshold cycle (Ct) values varied only slightly (less than isodiametric meristematic cells with dense cytoplasm and /6
5
6
one Ct) between samples, validating the use of the largenuclei,occasionallyassociatedwithpre-existingvascu- 1
4
18SrRNA geneas an endogenous control. lar tissues. The simultaneous emergence of several shoots 7 b
y
cDNA standard curves (positive controls) were generated was observed upon rupture of the globular structure. g
u
using 2-fold dilutions of cDNA from sample nine (corre- Adventitious shoots did not develop synchronously and e
s
sponding to the ninth day of shoot induction from one of after 2 weeks under light conditions, several stages of shoot t o
n
the three replicates) ranging from 200 ng to 48 pg of total development could be observed. 1
9
cDNA. Each qRT-PCR run included replicates of each Microscopic studies of leaf samples after cocultivation N
o
v
cDNA from the adventitious shoot induction of each time with Agrobacterium, under SEM and light microscopy, e
m
point (using both the targetand the 18SrRNA gene primers showed extensive bacterial proliferation, particularly near b
e
in separate reactions), two standard curves (one for the the wounds and over the proliferating cells. Bacteria were r 2
0
target gene and one for the 18SrRNA gene), and negative also present in large numbers inside the mesophyll in the 1
8
controls. The latter included no template controls and intercellular spaces, entering through wounds and stomata
RNA-only controls to ensure the absence of contaminating and spreading to the inner tissues of the explants. Plastic-
genomic DNA. embedded leaves sectioned longitudinally through the
The levels of the 18S and target gene transcripts in each wounded edges revealed that GUS positive cells were
sample were quantitatively assessed from the standard located mainly in the inner tissues, in the proximity of
curves executed under the same experimental conditions wounds, and rarely in the epidermis (Fig. 1F). The trans-
(same plate) as the concomitant qRT-PCR measurement. formation assays and gus expression analyses revealed that
The quantity of the target gene was divided by the quantity the limiting step in almond transformation is not bacterial
ofcontrolgenetoobtain anormalized relative valueforthe access to the inner cell layers where regeneration-competent
target gene. The specificity of the amplified products was cellsarelocalizedbutitisthelowregenerationabilityofthe
verified by analysis of the dissociation curves generated by transformed tissues. The 40% regeneration success of
4164 | Santos et al.
D
o
w
n
lo
a
d
e
d
fro
m
h
ttp
s
://a
c
a
d
e
m
ic
.o
u
p
.c
o
m
/jx
b
/a
rtic
le
-a
b
s
tra
c
t/6
0
/1
4
/4
1
5
9
/6
5
6
1
4
7
b
y
Fig.1. Histologyoftheadventitiousalmondshootinductionalongthecultureperiodindarkness.(A,D)Aspectsofnon-inducedleaves g
u
(day3),(A)woundedleafseenfromtheadaxialsurface,(D)lightmicroscopyimageofacross-sectionshowingmesophyllorganization). e
s
(B)Abaxialleafsurfaceatday10ofinduction.(C)Abaxialleafsurfaceatday19clearlyshowingthetissueproliferationandglobular t o
n
structuresformed.(E)Scanningelectronmicroscopyofglobularstructuresformedattheadaxialsurface,fromwhichadventitiousshoots 1
9
arise(day20).(F)SectionacrossaleaftransformedwithAgrobacteriumEHA105/p35SGUSINT,atday10aftercultureinitiation. N
o
HistochemicalGUSexpressionisvisibleintheregiondelimitedbyaspottedline.(G)Cross-sectionofaleafatday13,showing ve
m
meristematiccentresatdifferentdevelopmentalstages(arrows).(H)Twostagesofemergence/developmentofbudinitialsvisibleatthe b
e
endofshootinduction(day20).Avascularconnectionisvisibleinthemoredevelopedbud(arrow).(I)Sectionacrossabudemerging r 2
0
fromaglobularstructureattheendoftheshootinductionperiod(day20).Bars(A–C)3mm;(D,I)1mm;(E,F)50lm;(G,H)200lm. 1
8
almond clone VII reported by Miguel et al. (1996), was results that were obtained allowed us to define the time
found to decline with time reaching values of 25% in our frame within which the organogenesis process occurs. This
experiments. This percentage reflects only the number of knowledge was crucial in defining the frontier between the
leaves able to regenerate at least one shoot (although, earlyandthelatestagesoforganogenesisascorresponding
frequently, numerous shoots could be recovered from each to days 8 and 9 of almond leaf culture: almond leaves
leaf). From this situation, it is obvious that the number of submitted to shoot induction collected at days 1, 2, 5, and
cells responding to the induction signals and successfully 8 gave rise to an early organogenesis RNA pool and at
contributingtothemRNApooliseffectivelymuchreduced. days 9, 13, 15, and 19 represents the RNA population of
Although the precise histology of the meristematic the late events of organogenesis, used for SSH libraries
centres in almond leaves could not be established, the elaboration.
Uncovering adventitious shoot regeneration in almond | 4165
Transcript profiling of almond cDNA array than 2-fold were evaluated and those between 1- and 2-fold
were not considered for sequencing and further bioinfor-
To identify genes associated with the two main stages of
matic analysis.
almond adventitious shoot induction, two SSH cDNA
Of the 128 cDNAs that exhibit a 2-fold or greater
libraries were constructed from leaves induced to regener-
differential expression patter, 58 cDNA clones exhibited
ate. The shoot induction period was divided into two time
early stage-specific over-expression compared to 70 cDNA
frames, one representing the early stage of the shoot
clones over-expressed in late stage organogenesis (Fig. 2).
induction (ES) and the second representing the late stage
Sequenceassemblyanalysis revealedthatthese 128differen-
(LS). The separation of these two stages was to identify
tially expressed cDNAs represented 92 unique contiguous
genes differentially expressed (i) prior to and (ii) after
sequences (contigs). However, 27 of the 92 contigs are close
commitment (or the acquisition of competency) for shoot
to or smaller than 100 bp in length and thus gene
organogenesis. A total of 3840 cDNA clones derived from
annotation was difficult and IDs could not be obtained.
two SSH cDNA libraries, 1920 from each of the early and
Gene annotation for the remaining 65 contigs was based on
late almond adventitious root induction libraries, were similarity to known sequences recorded in the NCBI/ Do
w
evaluated for transcript abundance and developmental GenBank database, derived from BLASTN and BLASTX nlo
relevancy in a differential hybridization expression profile searchprotocols(Tables1,2).Genesencodingproteinsthat ad
e
aTnhaelysainsaulyssinisg ianvcoulvsteodm-amaddyee aslmwaopndstDraNteAgymwicirtoharrtawyo. bAerlaobnigdotposisdidffaetraebnatsfeafmoirlipesrowteeirnefcalmasisliyfiecdataulsoinguginthge(hMttIpP:/S/ d from
technical replicates (duplicate spots on the array for each mips.gsf.de/proj/funcatDB/search_main_frame.html). The h
cDNA) and three biological RNA replicates as the source functional groups within which the annotated genes clus- ttps
of cDNA targets for a total of six hybridizations. Of the tered include various protein families; proteins involved in ://a
c
3840 cDNAs, 3009 exhibited measurable fluorescence binding functions or cofactor requirement, cellular trans- ad
e
(passed flags) in at least two-thirds of the hybridizations port, transport facilitation and transport routes, cellular m
ic
with paired probes. 616 cDNA clones exhibited fluorescent communication/signal transduction mechanism, cell rescue .o
u
ratios that were statistically different from 1, but only 128 and defence, transcription, protein synthesis and fate, p.c
o
transcripts showed greater than 2-fold difference in abun- interaction with the environment, metabolism and sub- m
dance at P <0.05. On average, only 2.8% of the array cellular localization (see Supplementary Fig. S1 at JXB /jxb
edliefmfereenntscevsarhyigmhoerre tthhaann 22-f-ofoldldinbteetrwmesenof Creyp3rodauncdibiClityy5; onOlinfet)h.e 65 unique contigs, 37 represent transcripts that /article
fluorescence intensities in a co-hybridization experiment are exhibited differential expression profiles that suggest accu- -ab
s
believed to be significant (Ruan et al., 1998). Therefore, mulation and involvement in the early stage (ES) of tra
c
only cDNA elements with a relative signal intensity higher organogenesis (Table 1) compared to 28 transcripts, which t/6
0
/1
4
/4
1
5
9
/6
5
6
1
4
7
b
y
g
u
e
s
t o
n
1
9
N
o
v
e
m
b
e
r 2
0
1
8
Fig.2. ScatterplotofallgenesthathaveattestPvalue<0.05.Thespotsarecolouredbyexpressionratio.Thex-axisrepresentsthe
earlyversuslatecontrol.They-axisrepresentstheearlyversuslaterawdata.AbovetheupperdiagonallinearerepresentedallthecDNA
elementswithanexpressionratio>2-foldintheearlyconditionversuslatecondition.Underthelowerdiagonallinearerepresentedall
theelementswithanexpressionratio>2-foldinthelateconditionversustheearlycondition.
4166 | Santos et al.
Table1. Genesdifferentiallyexpressedintheearlystagesoforganogenesis
Accessionno. Annotation Length Foldchange
01.Metabolism
EF640666 PutativeS-adenosylmethioninedecarboxylase(SAMD) 402bp 2.39
EF640672 Putativeferredoxin–nitritereductase(NiR) 554bp 2.07
EF640674 PutativeS-adenosylmethioninecarboxypropyltransferase 684bp 2.09
EF640684 Putativepectinoesterase 458bp 2.14
EF640690 Putativeferredoxin–nitritereductase(NiR) 522bp 2.39
EF640695 PutativeNADdependentmalicenzyme 720bp 2.14
EF640698 Putativecarbonicanhydrase(CA) 889bp 2.30
11.Transcription
EF640665 HeatshockproteinHSP90 370bp 2.27
D
EF640688 PutativeoligouridylatebindingproteinhnRNP-likeproteinUBP1c 737bp 2.42 o
w
12.Proteinsynthesis n
lo
EF640659 Putativepurpleacidphosphatase1(PAP1) 450bp 2.14 a
d
EF640671 Putative40SribosomalproteinS26(RPS26C) 248bp 2.04 e
d
EF640678 Putative40SribosomalproteinS8(RPS8A) 446bp 2.08 fro
EF640701 PutativeribosomalproteinL36(RPL36) 765bp 2.00 m
h
EF640702 PutativeribosomalproteinS8(RPS8) 602bp 2.37 ttp
EEFF664400770145 PPuuttaattiivveer6ib0oSsroibmoaslopmroatlepinroLte2i2n(LR3P(LR2P2L)3A) 483796bbpp 22..3077 s://ac
a
13.Proteinfate(folding,modification,destination) d
e
EF640691 Chaperonin,60kDaGloEL 523bp 2.21 m
ic
EF640694 PutativeATPDILL-4electrontransporter 827bp 2.20 .o
u
16.Proteinwithbindingfunctionorcofactorrequirement(structuralorcatalytic) p
.c
EF640657 Putativecalmodulin-binding/translationelongation 548bp 2.25 o
m
EF640682 PutativetranslationelongationfactorEF-1alphachainA4-GTPbinding 322bp 2.02 /jx
b
EF640685 PutativetranslationelongationfactorEF-1alphachainA4-GTPbinding 175bp 2.14 /a
20.Cellulartransport,transportfacilitationandtransportroutes rtic
EF640656 Putativenon-photosynteticferredoxin/electron/hydrogentransport 396bp 2.57 le-a
EF640676 PutativePQ-looprepeatfamilyprotein 313bp 2.50 bs
32.Cellrescue,defenceandvirulence tra
c
EF640669 PutativeDnaK-typemolecularchaperonehsc70.1(HSP70) 597bp 2.16 t/6
0
EF640675 PutativeDnaK-typemolecularchaperonehsc70.1(HSP70) 378bp 2.07 /1
4
EF640696 Putativepolygalacturonase-inhibitingprotein 397bp 2.00 /4
1
EF640716 Lipidtransferprotein/a-amylaseinhibitor/trypsin 635bp 4.06 5
9
42.Biogenesisofcellularcomponents /6
5
EF640686 Cellwallproline-richprotein 690bp 2.06 6
1
4
70.Subcellularlocalization 7
b
EF640703 PutativephotosystemIassemblyprotein(Ycf3) 710bp 2.02 y
g
EF640706 PutativeATPsynthaseCfialphasubunit 640bp 2.10 u
e
EF640709 PutativeribosomalproteinL16(Rpl16) 320bp 2.56 st o
EF640711 PutativeATP-dependentproteaseproteolyticsubunit(ClpP) 644bp 2.07 n
1
EF640713 Putativechloroplasttransmembraneprotein(Ycf1) 315bp 2.27 9
N
Unknown o
v
EF640663 Unknownprotein 525bp 2.26 e
m
EF640664 Untranslatedleaderor3’endUTR 386bp 2.05 b
e
EF640673 Unknownprotein 473bp 2.32 r 2
0
EF640681 Unknownprotein 359bp 2.60 1
8
showed late stage (LS) organogenesis over-expression eventsthatshapeadventitiousshootinductioninalmond.A
(Table 2). Of the 65 organogenesis-related transcripts, 11 number of cellular processes are represented within these
represent novel genes or unknown functions (4 ES-specific collections and all appear to correlate well with what is
and 7 LS-specific). currently understood about the cellular events and changes
that occur as the tissues undergo dedifferentiation, rediffer-
entiation, and organogenesis. It is clear that the two SSH
The SSH EST collections
collections differ in their composition. The ES collection is
Gene expression profiles from two stage-specific SSH EST populated with cDNAs representing transcripts that appear
collections offer insight into the complex developmental tobemorecloselylinkedtoanincreaseinmetabolicactivity
Uncovering adventitious shoot regeneration in almond | 4167
Table2. Genesdifferentiallyexpressedinthelatestagesoforganogenesis
Accessionno. Annotation Length Foldchange
01.Metabolism
EF640693 Putativenitritereductase(NiR) 813bp 0.49
EF640714 Putativealcoholdehydrogenase(Adh-1) 326bp 0.49
11.Transcription
EF640677 Putativebromodomain-containingRNAbindingprotein(Brp1) 602bp 0.45
EF640705 PutativeMYBtranscriptfactor(MYB185) 335bp 0.44
20.Cellulartransport,transportfacilitationandtransportroutes
EF640670 Putativeglycinerich-protein 395bp 0.46
EF640683 Putativelipidtransportforcellwallsynthesis 361bp 0.44
32.Cellrescue,defenceandvirulence
D
EF640654 Putativeb-1,3-glucanase(Gns) 426bp 0.28 o
w
EF640658 Putativepathogenesis-relatedprotein(PR-4A) 396bp 0.35 n
lo
EF640662 Putativeacetone-cyanohydrinlyasedefence-relatedproteins 373bp 0.33 a
d
EF640667 Putativelipoxygenase 464bp 0.43 e
d
EF640680 Putativediseaseresistanceprotein(TIR-NBS-LRRclass) 585bp 0.38 fro
EF640689 PutativeclassIVchitinase 657bp 0.47 m
h
EF640699 PutativeperoxidaseclassIII 484bp 0.40 ttp
EEFF664400770172 PPuuttaattiivveeceltahsysleInIIep-reersopxoidnassiveeApTrPo3te5in(ETR) 332266bbpp 00..5204 s://ac
a
36.Interactionwiththeenvironment d
e
EF640661 Putative1-aminocyclopropane-1-carboxylateoxidase(ACCoxidase) 463bp 0.33 m
ic
EF640687 Putativegibberellicacidstimulated-protein(GAST-like) 204bp 0.48 .o
u
EF640717 PutativecytochromeP450mono-oxygenaselike-TBP 186bp 0.50 p
.c
42.Biogenesisofcellularcomponents o
m
EF640660 Extensin 293bp 0.32 /jx
b
Unclassified /a
EF640692 Metallothionein-likeprotein 373bp 0.45 rtic
EF640697 Wound-responsivefamilyprotein 539bp 0.48 le-a
Unknown bs
EF640655 Nohit 391bp 0.26 tra
c
EF640668 Unknownprotein 636bp 0.46 t/6
0
EF640679 Untranslatedleaderor3’endUTR 490bp 0.40 /1
4
EF640700 Unknownprotein 457bp 0.46 /4
1
EF640708 Unknownprotein 130bp 0.49 5
9
EF640710 Unknownprotein 551bp 0.48 /6
5
Noaccession Untranslatedleaderor3’endUTR 122bp 0.36 6
1
4
7
b
y
g
u
and biogenesis-related processes, whereas the LS collection duringtheinitialphasesofshootinduction(seebelow).The e
s
appears more related to stress responses and cell wall- contributionoftranscriptsinvolvedinproteinsynthesisand t o
n
related processes. Perhaps this difference is reflective of the related processes are, in all likelihood, associated with cell 1
9
change from one cell fate to another followed by a less division that occurs at about 2 d following culture N
o
v
activeperiodof cellular adjustmentand growth. initiation, apparently driven by a wounding response and e
m
The ES SSH collection contains a greater percentage of theactionofexogenoushormones included intheinduction b
e
differentially expressed transcripts involved in protein bio- media (Miguel, 1998). Thus, the differential expression of r 2
0
genesis than seen for the LS SSH collection. These tran- genescodingforribosomalproteins,elongationfactors,and 1
8
scripts not only encode a number of ribosomal proteins but chaperonins could be a reflection of cell division of the first
also transcripts for elongation factors (EF640682 and quiescent cellsand their initial proliferation.
EF640685) that influence elongation and protein folding or During the transition from the differentiated to a non-
transport; EF640657 a calmodulin-binding elongation fac- differentiated and proliferating state, in the early stages of
tor, EF640691 chaperonin-60, and EF640694 a protein shoot induction, chloroplasts undergo pivotal changes that
disulphide isomerase involved in oxidative protein folding ultimately lead to a reduction in photosynthesis in the cells
intheendoplasmicreticulum(WilkinsonandGilbert,2004). from which new shoots arise (competent cells) (Mazari and
EF640691 encodes a 60 kDa chaperonin (Cpn60) that is Camm, 2005). One of the more obvious changes in
essential for the entry of proteins into the chloroplast chloroplast structure that occurs in competent cells during
through the Toc and Tic complex (Vothknecht and Soll, shoot induction is the chloroplastidial conversion to pro-
2005) indicating an alteration in chloroplast morphogenesis plastids as the cells undergo dedifferentiation. Consistent
4168 | Santos et al.
with these alterations in chloroplast structure, the ES SSH a process mediated by the release of hydrogen peroxide and
collection is relatively rich in cDNAS representing up- involving molecular cross-linking by a wall peroxidase
regulated plastid-associated transcripts, the majority classi- (Cassab, 1998). Cell wall modifications appear to be more
fied in the Subcellular localization category in Table 1. In prevalent in the late stages of adventitious shoot induction
addition, EF640656 represents a non-photosynthetic ferre- if the inclusion of ESTs representing cell wall-related
doxin (Fd) protein transcript associated with plastid struc- transcripts in the LS SSH collection is any indication. The
ture and function. In citrus, Fd transcripts accumulate accumulation of a class IV chitinase transcript (EF640689)
preferentially in tissues containing undifferentiated plastids suggests the cleavage of cell wall arabinogalactans during
rather than chloroplasts (Alonso et al., 1995). The Fd the induction process, which has been linked to signalling
expression patterns in citrus suggest that the induction of processes that induce cell division in embryogenic cultures
non-photosynthetic Fd could be related to the demand for (Van Hengel et al., 1998). The presence of a class IV
reducing power in non-green but biosynthetically active chitinase may also indicate an important role for cell wall-
tissues. strengthening processes, in that chitinases coupled with an
D
Duringinvitroadventitiousshootformation,thedemand elevation in type III peroxidases (EF640699 and EF640707) o
w
for nutrients by the processes involved in cell dedifferentia- suggests diferulic acid-dependent cross-linking activity that nlo
tion and cell division becomes a priority. This demand is would result in strengthened cell walls (Andrews et al., ad
e
reenflceocdteedniitnritteheredduifcftearseenstia(Nl ieRx)prienssbioonthotfhetreaanrslcyriapntsdtlhaatet 2(E0F026)4.0I6n7a0d)daitniodn,exthteenEsiSnTs(EeFnc6o4d0i6n6g0)agalryecinime-proicrhtapnrtotceeinll d from
stages of adventitious shoot induction. The LS collection, wall constituents that can be highly cross-linked to other h
ttp
however, contains a contig that encodes a different NiR cell wall components. Glycine-rich proteins are often linked s
transcript than that seen in the ES collection, perhaps to the aromatic residues of lignin polymers (Cassab, 1998) ://a
c
a
indicatingthatdifferentNiRisoformsareactiveatdifferent and can be a marker for phase change in development (Gil d
e
stages of shoot induction. Several studies have demon- etal.,2003),whichwouldbeconsistentwithanadventitious m
ic
strated that nitrate assimilation is an important process in shoot induction process. Extensin is a hydroxyproline-rich .o
u
regeneration (Christianson and Warnick, 1984; Selby and glycoprotein (HRGP) that is highly cross-linked to the cell p.c
o
Harvey, 1990; Kintzos et al., 2004), and that high NiR wall and plays a major role in cell adhesion during plant m
activity or expression is crucial for regeneration ability, at development (Cassab, 1998) and the strengthening of the /jxb
least for rice (Nishimura et al., 2005). One would expect cell wall after cell division (Ito et al., 1998). The LS SSH /artic
perhaps that nitrate reductase is also up-regulated during EST collection also contains a cDNA encoding a lipid le
this process, but that remains to be determined as the EST transfer protein (EF640683) that has been suggested to play -ab
s
mapping and in-depth analysesprogress. a role in cell wall biosynthesis in response to environmental tra
c
The ES SSH collection contains two ESTs, EF640666 and stimuli (Kader,1997). t/6
0
EF640674, that encode S-adenosylmethionine decarboxylase Apart from the suggestion that processes involved in cell /1
4
(SAMD)andS-adenosylmethionine3-amino-3-carboxypropyl wall strengthening are important in the induction of /4
1
transferase (spermine synthase) respectively. The elevation of adventitious shoots, in almond there also appears to be an 5
9
these two transcripts in the early stages of shoot initiation emphasis on a reduction in cell wall adherence and /6
5
6
suggests the synthesis of polyamines in the explant tissue. elongationofcells.Ab-1,3-glucanase(EF640654)transcript 1
4
Several studies have shown correlation between increase in accumulates in the late stages of adventitious shoot in- 7 b
y
polyamine levels with cell division (Bais and Ravinshankar, duction and, although often attributed to be important in g
u
2002), which could explain the up-regulation of these defence-related mechanism, the ability of the enzyme to e
s
transcripts in the early stages of shoot initiation. Poly- cleave 1,3;1,6-b-D-glucans present in cell walls has led to t on
amines are also involved in interactions of cell wall suggestions that it is an important enzyme in several 1
9
components that contribute to cell wall rigidity and cell-to- developmental processes (Akiyama et al., 2004). The b-1,3- N
o
v
cell adhesion (Bais and Ravinshankar, 2002) and it was glucanase has also been described as being involved in e
m
recently suggested that the production of putrescine may spatial and temporal patterns in cell wall loosening that b
e
play a specific role in wound repair, a process that is an facilitate cell elongation during peach fruitlets development r 2
0
integral part of the explant system for shoot initiation (Ko et al., 2003) that may reflect what is occurring during 1
8
(Kuehn and Philips,2005). the late stages of adventitious shoot induction. The almond
The importance of cell wall processes is also reflected in GAST-like protein transcript (EF640687), a member of
the inclusion of pectinesterase (EF640684) and a cell wall a family of genes that encode small cell wall proteins that
proline-rich protein (EF640686) in the ES SSH collection. haveacysteine-richdomainandaputativesignalpeptide,is
Pectinoesterases have been linked to cell wall degradation also up-regulated late in organogenesis. GAST proteins are
during fruit ripening (Eriksson et al., 2004). Proline-rich thought to play an important role in phytohormone-
proteins represent one of five families of structural cell wall mediated cell expansion, as seen during flower development
proteins that have been identified in higher plants and are inGerbera(Kotilainen et al.,1999).
temporarily up-regulated during plant development(Fowler The ordered deposition of cell wall material and its
et al., 1999). Proline-rich proteins are presumed to be fixed composition are important factors for the co-ordination of
within the cell wall in response to physical damage by cell division and expansion. It is hypothesized that the