Genome-wide annotation, expression profiling, and protein interaction studies of the core cell-cycle

24 Pages·2013·2.78 MB·English

PlantMolBiol(2014)84:203–226
DOI10.1007/s11103-013-0128-y
Genome-wide annotation, expression profiling, and protein
interaction studies of the core cell-cycle genes in Phalaenopsis
aphrodite
Hsiang-Yin Lin • Jhun-Chen Chen •
Miao-Ju Wei • Yi-Chen Lien • Huang-Hsien Li •
Swee-Suak Ko • Zin-Huang Liu • Su-Chiung Fang
Received:23April2013/Accepted:3September2013/Publishedonline:25September2013
(cid:2)TheAuthor(s)2013.ThisarticleispublishedwithopenaccessatSpringerlink.com
Abstract Orchidaceae is one of the most abundant and activity. In addition, subcellular localization and pairwise
diverse families in the plant kingdom and its unique interaction analyses of various combinations of CDKs and
developmental patterns have drawn the attention of many cyclins, and of E2 promoter-binding factors and dimer-
evolutionary biologists. Particular areas of interest have ization partners confirmed interactions of the functional
included the co-evolution of pollinators and distinct floral units. Furthermore, our data showed that expression of the
structures, and symbiotic relationships with mycorrhizal core cell-cycle genes was coordinately regulated during
flora. However, comprehensive studies to decipher the pollination-induced reproductive development. The data
molecular basis of growth and development in orchids obtained establish a fundamental framework for study of
remain scarce. Cell proliferation governed by cell-cycle the cell-cycle machinery in Phalaenopsis orchids.
regulation is fundamental to growth and development of
the plant body. We took advantage of recently released Keywords Cell cycle (cid:2) Phalaenopsis aphrodite (cid:2)
transcriptome information to systematically isolate and Orchid (cid:2) Protocorm (cid:2) Meristematic tissues
annotate the core cell-cycle regulators in the moth orchid
Phalaenopsis aphrodite. Our data verified that Phalaen-
opsis cyclin-dependent kinase A (CDKA) is an evolution- Introduction
arily conserved CDK. Expression profiling studies
suggestedthatcorecell-cyclegenesfunctioningduringthe Cell reproduction is a basic biological process that is
G1/S, S, and G2/M stages were preferentially enriched in essentialforthegrowthofallorganisms(Halletal.2004).
the meristematic tissues that have high proliferation Cell-cycle regulation plays a pivotal role in the cell pro-
liferation required for plant growth. The core cell-cycle
genes are central regulators of cell division and hence a
Electronicsupplementarymaterial Theonlineversionofthis
common convergingpointforinternalandexternalsignals
article(doi:10.1007/s11103-013-0128-y)containssupplementary
material,whichisavailabletoauthorizedusers. during growth and development. In plants, growth derives
from meristematic tissues. Cell proliferation regulated by
H.-Y.Lin(cid:2)J.-C.Chen(cid:2)M.-J.Wei(cid:2)Y.-C.Lien(cid:2)H.-H.Li(cid:2)
intrinsic developmental signals and extrinsic environmen-
S.-S.Ko(cid:2)S.-C.Fang(&)
tal cues is required for meristem replenishment and orga-
BiotechnologyCenterinSouthernTaiwan,AcademiaSinica,
No.59,SirayaBlvd.,XinshiDistrict,Tainan741,Taiwan nization. In addition, cell division machinery governing
e-mail:[email protected] cell proliferation has to be carefully regulated in order to
meet differentiation and fate specification requirements
H.-Y.Lin(cid:2)J.-C.Chen(cid:2)M.-J.Wei(cid:2)Y.-C.Lien(cid:2)H.-H.Li(cid:2)
during plant development.
S.-S.Ko(cid:2)S.-C.Fang
AgriculturalBiotechnologyResearchCenter,AcademiaSinica, The eukaryotic cell cycle is controlled by an evolu-
Taipei115,Taiwan tionarily conserved set of cell-cycle proteins (Mironov
et al. 1999). Despite the universal principle of cell-cycle
H.-Y.Lin(cid:2)Z.-H.Liu
regulation, the core cell-cycle genes in plants are dupli-
InstituteofTropicalPlantSciences,NationalChengKung
University,Tainan701,Taiwan cated and have diverged to accommodate complex
123
204 PlantMolBiol(2014)84:203–226
developmental requirements. Based on a homology-based functional role of individual CYCs may be oversimplified
annotation strategy, more than 90 cell-cycle genes have and their roles during cell-cycle progression need to be
been classified in Arabidopsis (Vandepoele et al. 2002; tested experimentally.
Menges et al. 2005). However, how these cell-cycle pro- Cyclin-dependent kinase activity is also modulated by
teins are coordinated to drive cell division and regulate CDK-activatingkinases(CAKs).TwotypesofCAKshave
differentiation programs in plants remains largely unclear. beenfoundinhigherplantsthathavebeendesignatedasD-
Cyclin-dependent kinases (CDKs) are central cell-cycle and F-type. CDKDs are functionally related to vertebrate
regulatorswhoseactivitiesaresubjecttomultiplelevelsof CAKs, whereas CDKFs are plant specific. D- and F-type
regulation and whose expression oscillates in a periodic CDKs have been shown to activate A-type CDKs via
manner to drive cell-cycle phase transitions (Morgan phosphorylation (Umeda et al. 1998; Yamaguchi et al.
2007). Based on the nomenclature derived from genome- 2000; Shimotohno et al. 2003). In Arabidopsis, cyclin
wide classification of the Arabidopsis cell-cycle genes, H-activated CDKDs are capable of phosphorylating CDKs
thereareeightclassesofCDKs,designatedasAtoGtype as well as the C-terminal domain of the largest subunit of
and CDK-like (L-type) CDKs (Vandepoele et al. 2002; RNA polymerase II (Shimotohno et al. 2004). Conversely,
Menges et al. 2005). CDK activity is activated by binding F-typeCDKsdonotrequireaCYCinteractingpartnerand
to cyclins (CYCs), inhibited by CDK inhibitors, and reg- are capable of phosphorylating CDKs (Shimotohno et al.
ulated by phosphorylation. Plant cell-cycle transitions are 2004, 2006). Arabidopsis CDKF has been shown to phos-
controlledbytwotypesofCDKs—A-typeCDKs(CDKAs) phorylate and activate CDKD (Shimotohno et al. 2004).
and B-type CDKs (CDKBs). Plant CDKA contains a Cyclin-dependent kinase inhibitors (CKIs) negatively
canonical cyclin-binding motif PSTAIRE in the alpha-1 regulate the CDK activity by direct binding. Plants contain
helixthatisconservedacrossdifferentkingdoms(Inzeand CKIsthatshowweaksimilaritytotheN-terminallylocated
DeVeylder2006).CDKAhasbeenshowntoplayapivotal CKIdomainofthemammalianCip/Kipproteins(Wangetal.
role at both the G1/S and G2/M transitions and can func- 1997;Luietal.2000)andare,therefore,commonlyreferredto
tionallyrescueCDK-deficientyeastmutants(Ferreiraetal. asKip-relatedproteins(KRPs).TheArabidopsisKRPprotein
1991; Hirayama et al. 1991). Members of the CDKB gene familyhassevenmembers.WorkwithArabidopsishasshown
family are plant specific (Burssens et al. 1998; Mironov thattheKRPfamilyproteinsareexpressedinbothmitotically
et al. 1999). They have been reported to play roles during dividingandendoreduplicatingcells(Ormeneseetal.2004).
the G2-M and S-G2-M phases (Fobertet al. 1996; Magyar The involvement of some KRP genes in regulating endore-
et al. 1997; Umeda et al. 1999; Meszaros et al. 2000; duplication (part of the plant differentiation program) has
Porceddu et al. 2001; Sorrell et al. 2001; Breyne et al. beenconfirmedbyoverexpressionapproaches.Forinstance,
2002; Menges et al. 2002; Corellou et al. 2005). overexpression of either Arabidopsis KRP1 or KRP2, pref-
Cyclins activate the kinase activity of the CDKs and erentiallyexpressed inendoreduplicating cells, reducescell
control timely entry into the cell cycle. In plants and ani- divisionandaffectstheswitchtotheendoreduplicationcycle
mals, three main classes of CYCs are required for cell- (Wang et al. 2000; De Veylder et al. 2001; Verkest et al.
cycle progression. Generally, but not exclusively, D-type 2005b;Gutierrez2005;DeVeylderetal.2007;Verkestetal.
CYCsfunctionintheG1stageandarerequiredtoregulate 2005a).The defects causedbyoverexpressionofKRP1are
G1/Stransition,A-typeCYCsfunctionduringSphase,and overcomebysimultaneousoverexpressionofD-typeCYCs
B-type CYCs are important in regulation of S phase and (Jasinskietal.2002;Schnittgeretal.2003;Zhouetal.2003).
mitosis. Transcripts of CDKB and CYCB gene family Additionally, KRP6 whose expression is present in both
members are preferentially accumulated during S, G2, and mitotically dividing and endoreduplicating cells (Ormenese
M phases and are often used as markers for dividing cells et al. 2004) is required for timely regulation of cell-cycle
(Inze and De Veylder 2006). Plant A- and B-type CYCs progressionduringgametogenesis(Liuetal.2008).
have orthologs in animals. Plant D-type CYCs show less In yeast and animal systems, the kinase activity of the
similarity to D-type CYCs from animals and are therefore CDKsisnegativelyregulatedbytheWEE1familykinases
referredtoasplantspecific(Wangetal.2004).Thereareat (Kellogg 2003; Perry and Kornbluth 2007). WEE1 protein
least 30 CYCs with potential roles in cell-cycle regulation kinases phosphorylate a conserved Tyr residue of the
in Arabidopsis (Vandepoele et al. 2002). Because D-type CDKsandnegativelyregulateCDKactivity.Suchnegative
CYCs regulate G1-S transition and the commitment to regulation is necessary to coordinate transition between
enter the cell cycle, they are thought to be primary targets DNA replication and mitosis (Russell and Nurse 1987;
forexternalandinternalsignalsleadingtocellproliferation Gould and Nurse 1989; Jin et al. 1996). In Arabidopsis,
(Nieuwlandetal.2009).However,manyexceptionstothis WEE1 transcripts are induced under DNA damage condi-
simplefunctionalassignmenthavebeenreported(Inzeand tions. WEE1 deficient plants grow normally without
De Veylder 2006; Vanneste et al. 2011). Therefore, the obvious cell division or endoreduplication phenotype but
123
PlantMolBiol(2014)84:203–226 205
display hypersensitivity to DNA-damaging agents (De DNA-bindingdomains that allow them tobind as a mono-
Schutter et al. 2007). mer in a DP-independent manner to E2F target genes
The retinoblastoma (RB)/E2 promoter-binding factors (Maricontietal.2002;Lammensetal.2009).Theseatypical
(E2F)/dimerization partners (DP) pathway plays a pivotal E2F proteins do not contain a transactivation domain and
role in control of G1/S transition in eukaryotic organisms playversatilerolesduringplantdevelopment.
(Weinberg 1995; Gutierrez 1998, 2005). Within this para- TheOrchidaceaefamilyisoneofthelargestfamiliesof
digm, RB protein represses transcription of E2F-regulated flowering plants. Orchid species inhabit a wide range of
genes by physically interacting with the heterodimeric ecological environments and possess highly specialized
E2F/DP transcription factor complex (Harbour and Dean morphological, physiological, and developmental charac-
2000). The mitogenic signal induces cyclin-activated teristics (Dresser 1990). For example, ovule development
CDKA activity, which in turn phosphorylates RB and does not initiate until pollination occurs (Nadeau et al.
releases RB from E2F/DP promoter complexes. Active 1996). Fertilization only occurs after ovules and female
transcriptional activity of E2F/DP protein then allows gametophytes are fully developed, which normally takes
expression of S-phase genes and cell cycle entry. Like approximately 60–70 days. Additionally, orchid embryos
animal Rb, plant retinoblastoma-related (RBR) protein is donothavetheobviousorganizeddomainsforspecification
phosphorylated by CYCD-activated CDKA activity in a ofthefutureorgans.Duringgermination,orchidseedspass
cellcycle-dependentmanner(BoniottiandGutierrez2001; throughatransitionalstageduringwhichaspecializedsmall
Nakagamietal.2002;Korolevaetal.2004).However,the spherical tuber-like structure termed the protocorm is pro-
biological consequences of RBR phosphorylation/depho- duced. The protocorm has a meristematic domain at the
sphorylation in regulating cell proliferation and fate spec- anterior end where new leaves and roots are derived
ification during plant development remains elusive. In (Nishimura 1981). Given such unique developmental pro-
Arabidopsis,RBRknock-outmutantsaresterilebecauseof grams,cell-cycleregulatorsinorchidsmayhaveevolvedand
excess mitotic divisions in the mature female megagame- be regulated differently to those of other plants to accom-
tophyte(Ebeletal.2004).RBRisalsorequiredtoregulate modatespecializedanduniquedevelopmentaldecisions.
differentiation in root stem cells (Wildwater et al. 2005). Although many studies have reported the isolation and
Reduction of RBR protein negatively influences the molecular functions of the core cell-cycle genes in Arabi-
establishment of cell differentiation. For instance, meri- dopsis and rice, similar effort has not been paid to non-
stematic stem cells such as shoot apical meristem cells, model plant species with specialized developmental pro-
meristemoid mother cells, and procambial cells fail to grams. Despite the general principle of cell-cycle regula-
produce differentiated cells subsequently resulting in tion, there are important variations in how cell-cycle
defects in lateral organ formation (Borghi et al. 2010). In programs are modified to deal with environmental cues or
addition, RBR is required for developmental phase transi- developmentaldecisionsacrossplantspecies.Forinstance,
tion(Gutzatetal.2011)andasymmetricstemcelldivision endoreduplication (the modified cell-cycle program) is
in roots (Cruz-Ramirez et al. 2012). In summary, RBR is incorporated into DNA stress adaptation in Arabidopsis
important in the regulation of cell division and differenti- (Adachi et al. 2011). Rice plants, on the other hand, deal
ation during different stages of plant development. with DNA stress by reducing endoreduplication (Endo
Plants encode multiple E2F and DP family members et al. 2011). Hence, biological consequences derived from
(Gutierrezet al.2002;De Veylder et al.2002;DeVeylder functional studies of the cell-cycle regulators in Arabi-
et al. 2003; Dewitte and Murray 2003). E2F/DP protein dopsis might not directly translate to the functions of
complexescanserveastranscriptionactivatorsorrepressors orthologous counterparts in other plants. Identification and
(Muller and Helin 2000; Bracken et al. 2004; Gutierrez expression profiling analysis of the cell-cycle genes in
2005). In Arabidopsis, overexpression of AtE2Fa and other species might provide novel insights into diversifi-
AtE2Fb with AtDPa induced cell proliferation in differen- cation of the cell-cycle program across the plant kingdom.
tiatedtissues(DeVeylderetal.2002;Rossignoletal.2002; In this study, we conducted a genome-wide study of the
Magyar et al. 2005; Sozzani et al. 2006). Recent work corecell-cyclegenesofthemothorchidPhalaenopsisaph-
revealedthatAtE2Fastimulatescellproliferationbyform- rodite and profiled their expression patterns. We chose P.
ing a stable complex with AtRBR1 protein to inhibit aphrodite because it is an important parent plant for com-
endoreduplication and premature differentiation (Magyar mercialbreedingprogramsinTaiwan,anditstranscriptome
etal.2012).E2FcandDPb,ontheotherhand,worktogether databaseispubliclyavailable(Suetal.2011;Anetal.2011;
toexertanegativeeffectoncellproliferation(delPozoetal. Fu et al. 2011). We present a comprehensive analysis of
2002). In addition to the typical E2F transcription factors, transcriptionalregulationofthecorecell-cyclegenesduring
plants have evolved atypical E2F or DP-E2F-like (DEL) orchid development. We used yeast two-hybrid (Y2H) and
factors that lack an RBR-binding motif and possess two bimolecularfluorescencecomplementation(BiFC)assaysto
123
206 PlantMolBiol(2014)84:203–226
confirm the interaction network of the selected cell-cycle Phylogenetic tree construction
genes. Furthermore, a protein–protein interaction map pro-
videsmolecularevidenceofthefunctionalunitsofcell-cycle ProteinsequenceswerealignedbyClustalW.Theresulting
proteincomplexes.Takentogether,ourdatarepresentthefirst alignments were used to construct phylogenetic trees in
comprehensivecharacterizationofthecorecell-cyclegenes Mega5.05(Tamuraetal.2011).TheMaximum-likelihood
in the Phalaenopsis orchid. In addition, distinct expression method was used to generate phylogenetic trees and 1,000
patterns of cell-cycle regulators during reproductive devel- replicateswereusedforbootstrapping.Bootstrapvaluesof
opmenthavebeendocumented. 50 % or higher were shown for each clade. The evolu-
tionary distances were computed using the Poisson cor-
rection method and the rate variation among sites was
Materials and methods modeled with a gamma distribution. Gene identification
numbers used to generate phylogenetic trees are listed in
Plant materials Supplementary Table S3.
Phalaenopsis aphrodite Subsp. formosana (v1656, v1644 Yeast two-hybrid assay
or v1642) seedlings in 2.5 in. pots were purchased from
ChainPortOrchidNursery(PingTung,Taiwan).Allplants cDNA of PaDP1 or PaDP2 gene was introduced into
were grown in alternating 12 h light (23 (cid:3)C)/12 h dark pDEST32vectorandanin-framefusiontotheGAL4DNA
(18 (cid:3)C) cyclesin agrowthchamber with regular irrigation binding domain was generated as bait. cDNA of PaE2F3
and fertilization. genewasintroducedintopDEST22vectorandanin-frame
fusion to the GAL4 activation domain was generated as a
Annotation strategy prey (Invitrogen, USA). cDNAs of the PaE2F1 and
PaE2F2 failed to be cloned into the Gateway system.
Comprehensivesearchesforcorecell-cyclegenesincluding Alternatively,theywereintroducedintoapGADT7vector
CDKs,CYCs,RB-related,E2FsandDPs,WEE1,andCKIs (Clontech, USA) to make an in-frame fusion to the GAL4
were conducted on the Orchidstra Phalaenopsis Genome activationdomain.cDNAofPaDP1orPaDP2wascloned
Annotation Database (Su et al. 2011). We used the tran- intopGBKT7vector(Clontech,USA)tomakeanin-frame
scripts of core cell-cycle genes from either Arabidopsis or fusion to the GAL4 DNA binding domain. The pairs of
rice (Supplementary Table S1) to run genome-wide Basic constructs to be tested were co-transformed into AH109
Local Alignment Search Tool (BLAST) searches against yeast competent cells using a lithium acetate method
Orchidstra Phalaenopsis Genome Annotation Database according to the manufacturer’s instructions (Invitrogen,
http://orchidstra.abrc.sinica.edu.tw/none/.Theinitialsearch USA). Co-transformed yeast cells were selected on plates
yieldedafewhigh-scoringpositives(E-valueof1e-005was with Leu (for pDEST32 and pGADT7 plasmids) and Trp
set as the cutoff). The sequences of tentative EST contigs (for pDEST22 and pGBKT7 plasmids) drop-out medium
were retrieved and blasted against the National Center for for 3–8 days at 30 (cid:3)C. Transformants were tested for spe-
Biotechnology Information (NCBI) non-redundant protein cific interactions by growing on SC–Leu–Trp–His plates
database to confirm their identification, and against the with 10 or 20 mM 3-amino-1,2,4 triazole (3AT).
Phalaenopsis Genome Annotation Database to identify
additional family members. To categorize gene members Complementation of a S. cerevisiae cdc28-as1 mutant
fromthesamefamily,theBLASTsearchwasrepeatedwith
the Phalaenopsis candidate genes against the protein data- PaCDKA1cDNAwasclonedintoapYES-DEST52vector
base at The Arabidopsis Information Resource and The (Invitrogen, USA). Theconstructwas transformed into the
MichiganStateUniversityRiceGenomeAnnotationProject cdc28-as1 mutant by the lithium acetate method described
Database and Resource. The signature motifs conserved above. The transformants were allowed to grow in the
during evolution were taken into consideration during presence of 1,000 nM 4-Amino-1-tert-butyl-3-(10-naph-
annotation.TocorrecterrorsofassembledESTcontigsand thylmethyl) pyrazolo [3,4-d] pyrimidine (1NM-PP1),
obtainfull-lengthgenemodels,rapidamplificationofcDNA MerckMilliporeChemicals,USA)andscoredforviability.
ends (RACE)-PCR and reverse transcription were carried
outtoconfirmandcompletethegenemodels.Basedonthis Sample collection and RNA extraction
strategy, we were able to identify and isolate most of the
core cell-cycle genes in P. aphrodite. The results from For gametophytic and embryonic tissues, orchid flowers
individual gene families are discussed below. The annota- were hand pollinated and developing ovaries were har-
tionresultsaresummarizedinSupplementaryTableS2. vested at the specified day. Only the interior tissues from
123
PlantMolBiol(2014)84:203–226 207
developing ovaries were scooped and pooled for RNA Labstation (Milestone, USA). Tissues (10 lm thick) were
extraction. For root samples, 2-cm tip tissue containing sectioned using a MICROM 315R microtome (Thermo
root apical meristems was collected. For stalk samples, Scientific, USA) and mounted onto a poly-L-lysine-coated
5–10 cm long stalks were collected. For protocorm-like slide (Matsunami, Japan). Sections were then de-paraffi-
body(PLB)samples,one-month-oldtissueswerecollected. nizedinxylene,rehydratedindecreasingconcentrationsof
For protocorm samples, 20- and 30-day-old tissues were ethanol, and digested with 2 mg/ml proteinase K at 37 (cid:3)C
pooled and collected. The collected samples were flash for 30 min. In situ hybridization was performed as previ-
frozeninliquidnitrogenandstoredinafreezerat-80 (cid:3)C. ously described (Bi et al. 2005) with slight modification.
RNA was isolated using MaestroZol RNA Plus extraction Briefly, hybridization was performed at 59 (cid:3)C in the pre-
reagent (Maestrogen, USA) according to the manufac- sence of 40 ng of DIG-labeled RNA probe. Sense and
turer’s instructions. The isolated total RNA was treated antisense probes were synthesized using a SP6/T7 digoxi-
with RNase-free DNase (Qiagen, USA) followed by genin RNA labeling kit according to the manufacturer’s
RNeasy mini-column purification according to the manu- instructions (Roche, USA). Hybridization signals were
facturer’s instructions (Qiagen, USA). detectedbyNBT/BCIPdetectionkit(Roche,USA).Tissue
sections and in situ hybridization were photographed on a
Real-time quantitative RT-PCR and RACE-PCR Zeiss Axio Scope A1 microscope equipped with an Axio-
Cam HRc camera (Zeiss, Germany).
Three micrograms (for experiments shown in Fig. 5) or
5 lg(forexperimentsshowninFig. 9)totalRNAwasused BiFC assay and microscopy
for cDNA synthesis. DNA-free RNA was reverse tran-
scribed in the presence of a mixture of oligo dT and ran- cDNAs encoding the CDK genes were introduced into
dom primers (9:1 ratio) using the GoScript Reverse pE3134[fordetails, please seehttp://www.bio.purdue.edu/
Transcription System (Promega, USA) according to the people/faculty/gelvin/nsf/protocols_vectors.htm (Tzfira
manufacturer’s instructions. Ten microliters of RT-PCR et al. 2005; Citovsky et al. 2006)]. The N-terminal (amino
reaction contained 2.5 ll of 1/20 diluted cDNA, 0.2 mM acid residues 1–174) half of YFP was C-terminally in-
(for experiments shown in Fig. 5) or 0.25 mM (for exper- framefusedtotheCDKprotein.cDNAsencodingtheCYC
iments showninFig. 9)ofprimers, and5 llof29KAPA geneswereintroducedintothepE3130orpE3132vectors.
SYBR FAST master mix (KAPA Biosystems, USA). The TheC-terminal(aminoacidresidues175–239)halfofYFP
following program was used for amplification: 95 (cid:3)C for was N-terminally or C-terminally in-frame fused to the
1 min,40cycles of95 (cid:3)Cfor 5 sand58 (cid:3)Cfor20 s.PCR CYC proteins (N(cEYFP)-PaCYC or C(cEYFP)-PaCYC).
was performed in triplicate, and the experiments were For PACYCB1;2 and PaCYCD2;1, only the N-terminally
repeated twice with RNA isolated from independent sam- tagged version, N(cEYFP)-PaCYCB1;2 and N(cEYFP)-
ples. For PaCYCA3;1 and PaCYCD5;4, annealing and PaCYCD2;1, was obtained. Ballistic bombardment-medi-
extension temperature was 60 (cid:3)C. Real-time PCR was ated transient transformation was carried out as previously
performed using a CFX96 Real-Time PCR detection sys- described (Hsu et al. 2011a) with slight modification.
tem (Bio-Rad, USA). Quantification analysis was carried Briefly, 1 lg DNA coated on gold particles (1 lm in
out by CFX Manager Software following manufacturer’s diameter) was bombarded into white petals of Phalaen-
instructions (BioRad, USA). Primers used for qPCR are opsis Soga Yukidian ‘V3’ using a Biolistic PDS-1000-He
listed in Supplementary Table S4. 50 and 30 RACE PCR particle delivery system (Bio-Rad, USA) at the following
were carried out using a SMARTerTM RACE cDNA settings: helium pressure of projectile, 1,100 psi;
amplification kit according to manufacturer’s instructions 27 mmHg partial vacuum; and target-distance, 9 cm. Flo-
(Clontech,USA).PrimersusedforRACE-PCRarelistedin rescence images were photographed on a LSM 780 Plus
Supplementary Table S5. ELYRA S.1 Confocal Microscope with Plan-Apochromat
409/1.4 oil objective lens (Zeiss, Germany).
In situ hybridization
Twenty-eight-day old protocorms, female gametophytes, Results
and developing embryos were collected and fixed imme-
diately in 4 % paraformaldehyde, 4 % dimethylsulfoxide, Annotation ofcore cell-cycle genes inthe P. aphrodite
0.25 %glutaraldehyde,0.1 %Tween20,and0.1 %Triton database
X-100 in diethylpyrocarbonate-treated H O at 4 (cid:3)C over-
2
night. Tissues were then dehydrated and infiltrated with BLASTsearcheswereconductedtoidentifythemajorcell-
Paraplast (Leica, USA) using a KOS Rapid Microwave cycle genes in a P. aphrodite transcriptome database.
123
208 PlantMolBiol(2014)84:203–226
Sixty-five potential genes encoding the core cell-cycle alsocarriesthePITAIREmotif(SupplementaryTableS2).
proteins important for cell-cycle control were identified. The G-type CDK class is homologous to the human p58
The isolated genes belonged to the CDK, CYC, Rb, galactosyltransferase protein whose role is important for
E2F/DP, Wee1, and CKI gene families. cytokinesis (Menges et al. 2005). Co-purification of
CDKGs with CYCL1 suggests that CYCL1 may be an
CDK family interacting partner of CDKG (Van Leene et al. 2010).
CYCL was also identified from the P. aphrodite tran-
AtleastonememberofeachclassofCDKwasidentifiedin scriptome database (Supplementary Table S1). The func-
the P. aphrodite genome (Supplementary Table S1). tions of plant CKL protein kinases remain to be clarified.
A-andB-typeCDKsarethemajorCDKsthatcontrolcell- Because A- and B-type CDKs are the only CDKs that
cycle transitions in plant cells. P. aphrodite contains at directly regulate cell-cycle progression in plants (Vande-
least one ortholog of the CDKA genes, which was desig- poele et al. 2002; Menges et al. 2005), we chose to focus
nated as PaCDKA1. Like other plant CDKA proteins, the remainder of our study on A- and B-type CDKs.
PaCDKA1 protein has a conserved PSTAIRE signature in
thealpha-1helix,ahallmarkCYC-bindingmotif(Fig. 1a). Cyclin family
In addition, the T-loop and the phosphorylating threonine
residue required for activation are also conserved in PaC- There are 10 types of CYCs in Arabidopsis (Wang et al.
DKA1 (Fig. 1a). The plant CDKB family has been classi- 2004), five of which were identified in the P. aphrodite
fied into two subgroups (Vandepoele et al. 2002; Dewitte transcriptomedatabase(SupplementaryTableS2).BLAST
andMurray2003;InzeandDeVeylder2006).TheCDKB1 analysis revealed that at least six A-type CYCs (CYCAs)
subgroup has PPTALRE as the CYC binding motif. The belonging to three different subgroups of the family
CDKB2 subgroup is characterized by a P[S/P]TTLRE (Chaboute et al. 2000) were present in the P. aphrodite
signature motif. Orthologs of B-type CDKs designated as transcriptome. They were designated as PaCYCA1;1,
PaCDKB1 and PaCDKB2 were annotated. PaCDKB1 and PaCYCA2;1, PaCYCA2;2, PaCYCA2;3, PaCYCA3;1, and
PaCDKB2 share 64 % identity and 81 % similarity at the PaCYCA3;2(Fig. 2a).AllPaCYCAgenesencodedproteins
amino acid level. Based on a previous classification containing a typical LVEVxEEY (x = any amino acid)
(Dewitte and Murray 2003; Inze and De Veylder 2006; signaturethatissharedbyA-typeCYCsinplantandanimal
Vandepoele et al. 2002), PaCDKB1 contains a PPTTLRE systems (Renaudin et al. 1996) and a conserved CYC box
motif and therefore belongs to the type 2 CDKB subgroup motif MRA/GILI/VDW (Morgan 2007) (Fig. 2a). Five
(Fig. 1b). However, phylogenetic analysis indicated that B-type CYCs were identified from our genome-wide ana-
PaCDKB1 was clustered with type 1 CDKB members that lysis(Fig. 2b).FourofthemhaveaCYCB-specificHXKF
contain the PPTALRE motif (Fig. 1c). PaCDKB2, on the motif and a conserved CYC box motif: MRAILVDW (for
other hand, carried a modified signature motif—PAT- the PaCYCB1 subgroup), and MRAILIDW (for the
TLRE—and formed a cluster with a group of monocot PaCYCB2 subgroup). Fifteen members from six different
CDKB family members(Fig. 1b, c). The PATTLRE motif subgroupsoftheD-typeCYCgenefamilywereannotated.
was also found in CDKB family members of the moss All but CYCD6;1 shared a conserved N-terminal motif
Physcomitrellapatens(unpublisheddata).OftheCAKs,at LXCXE that was found to bind to the pocket domain of
least one member each of the CDKD and CDKF family RB-related proteins (Fig. 2c, Supplementary Table S2).
was identified from the Phalaenopsis annotation database PaCYCD6;1,ontheotherhand,containedaLXCXEmotif
(SupplementaryTablesS1,S2).TheotherCDKclassesare attheC-terminalendoftheprotein(Fig. 2c).Thissuggests
C-type, E-type, G-type, and CDK-like (CKL) families. At that Phalaenopsis D-type CYCs are capable of interacting
leastonegenefromeachoftheCDKCandCDKEfamilies, with RB protein to regulate the cell-cycle progression. No
two CDKG and 13 CDK-like (CKL) genes were identified CYCD7 was found in available orchid transcriptome dat-
from the P. aphrodite transcriptome database (Supple- abases(Suetal.2011;Tsaietal.2013).Toaddressthelack
mentary Table S1). Plant C-type CDKs and E-type CDKs oftheevolutionarilyconservedCYCD7inthisstudy,wedid
are characterized by their similarity to mammalian CDK7 low-stringencySouthernblottingofthePhalaenopsisorchid
and CDK8, respectively (Joubes et al. 2000; Kitsios et al. genomic DNA and probed with the rice cDNA fragment
2008). They are assumed to regulate transcription in a corresponding to the conserved regions among monocot
similar manner to their counterparts in mammals (Inagaki CYCD7;1. No Phalaenopsis CYCD7 signal was detected
and Umeda 2011). Arabidopsis CDKC has been shown to (Supplementary Fig.S1). However,we cannot ruleout the
be involved in splicing-related transcriptional regulation possibility that the orchid CYCD7 sequence may be too
(Kitsios et al. 2008). Like other plant CDKC protein divergenttobedetectedbySouthernblotting.Phalaenopsis
kinases (Joubes et al. 2000), the CDKC1 of P. aphrodite D-,A-,andB-typeCYCswereclearlyseparatedandformed
123
PlantMolBiol(2014)84:203–226 209
Fig.1 ClustalW alignment of a CDKA1 and b CDKB proteins. [I/L]RE motif of B-type CDK proteins in various plant species is
Asterisk indicates a conserved residue. Colon indicates a residue markedinred.cUnrootedmaximum-likelihoodtreeofB-typeCDK
showing strong similarity among sequences. Dot indicates a residue proteins from different plant species. cdc28, cell division cycle 28;
showingweaksimilarityamongsequences.TheconservedPSTAIRE Pa,P.aphrodite;Os,Oryzasativa;Zm,Zeamays;Bd,Brachypodium
motifforcyclinbindingamongvariousCDKA1proteinsismarkedin distachyon; At, Arabidopsis thaliana; Pp, P. patens; Cr, Chlamydo-
red.TheconservedT-loopofA-typeCDKisindictedbyabluebar. monas reinhardtii; Vc, Volvox carteri; HS, Homo sapiens; Sl,
The conserved phosphorylating threonine residue is labeled in blue Solanumlycopersicum
and marked by an inverse triangle. The modified P[P/A/S]T[T/A]
cladeswithdifferentCYCgroupsinArabidopsis(Fig. 2d). RB and E2F/DP families
The remaining CYCs identified from the P. aphrodite
transcriptome were H- or L-type CYCs (Supplementary TwoRB-relatedgeneswereidentifiedfromtheP.aphrodite
TableS2). transcriptomedatabase(Fig. 3a,SupplementaryTableS2).
123
210 PlantMolBiol(2014)84:203–226
123
PlantMolBiol(2014)84:203–226 211
bFig.2 ClustalWalignmentofaA-typeCYCs,bB-typeCYCs,and were also isolated andverified. Both PaDPs encodea pro-
c D-type CYCs. Asterisk indicates a conserved residue. Colon tein containingthe DPcanonical DNA binding and dimer-
indicates a residue showing strong similarity among sequences. Dot
ization domains (Fig. 3c). We were not able to obtain the
indicates a residue showing weak similarity among sequences. The
A-typeCYCshaveaconservedLVEVxEEYmotif(markedbyablue full-length cDNAs of PaE2F4 and PaDP3. Therefore,
bar) and a CYC box motif (green bar). B-type CYCs have a PaE2F4 and PaDP3 were not included in the following
conservedHXKFmotif(bluebar)andaCYCboxmotif(greenbar). analyses.
The conserved LXCXC motif is marked in red. d Unrooted maxi-
In addition to the typical E2F and DP gene family
mum-likelihoodtreeofD-,A-,andB-typeCYCsfromP.aphrodite
(Pa) and A. thaliana (At). Bootstrap values of 50% or higher are members,anatypicalDP-E2F-like(DEL)genewasidenti-
shownforeachclade.ClustalWalignmentofP.aphrodite fied in the P. aphrodite transcriptome database. Like other
plant atypical DEL proteins (Vandepoele et al. 2002;
Lammens et al. 2009), PaDEL1 contained a duplicated
DNA binding domain highly similar to that of the E2F
transcriptionfactor(Fig. 3d).Basedonpreviousstudies,the
atypical DEL proteins do not heterodimerize with DP and
insteadformhomodimerstoexerttheirfunction(DiStefano
et al. 2003). In plants, DEL proteins are involved in regu-
lation of cell wall biosynthesis and endoreduplication (Ra-
mirez-Parraetal.2004;Vliegheetal.2005;Lammensetal.
2008).
WEE1 and CKI gene families
One WEE1 ortholog was found in the Phalaenopsis
annotation database. The encoded protein displays 47 %
identity to AtWEE1 and 57 % identity to rice WEE1 pro-
teins at the amino acid sequence level. The Phalaenopsis
WEE1 protein has a conserved catalytic domain including
theATPbindingsiteoftheproteinkinasefamily(Fig. 4a).
It also contains a conserved E/DGD triplet motif that dis-
tinguishes WEE1-related kinases from other kinase fami-
lies (Booher et al. 1993; Sorrell et al. 2002).
Three potential CDK kinase inhibitor (CKI) genes were
annotatedandidentifiedinP.aphrodite.Allthreecarrieda
CKI domain (pfam02234) at the C-terminal region
(Fig. 4b).PaKRP1shares69 %aminoacidsequenceiden-
Fig.2 continued titywithPaKRP3.PaKRP2shares52and49 %aminoacid
sequenceidentitywithPaKRP1andPaKRP3,respectively.
BecausethesequenceofPaRBL2cDNAinthedatabasewas Cell-cycle gene expression profiles
incorrect, the full-length cDNA was identified and con-
firmed by reverse transcription PCR followed by sequenc- Toinvestigatetheexpressionpatternsofthecell-cyclegenes
ing.BothPaRBL1andPaRBL2containedcanonicalAand inPhalaenopsisorchid,weusedquantitativeRT-PCR(qRT-
B domains that showed high similarity to RB proteins of PCR) analysis to compare relative transcript abundance in
other plant species (Fig. 3a). The A and B domains are varioustissues.Meristematictissueswithahighrateofcell
important to form a binding pocket for E2F transcription division were marked by cytokinesis-specific syntaxin
factors.FourE2FandthreeDPgeneswereidentifiedfrom (PATC147993),KNOLLE(Lauberetal.1997),andamer-
the P. aphrodite transcriptome database (Supplementary istematic marker, KNOTTED-like (PATC127065) homeo-
TableS2).Thefull-lengthcDNAsofPaE2F1,PaE2F2,and box transcription factor (Fig. 5a). Our expression analysis
PaE2F3wereisolatedandtheencodedproteinscontaineda showed that PaCYCA1;1, PaCYCB1;1, PaCYCB1;2, Pa-
DNA binding domain, a dimerization domain, a marked CYCB2;1,PaCYCB2;2,PaCDKB1,andPaCDKB2mRNAs
box, and a conserved C-terminal region that potentially accumulated to relatively high levels in meristems of root
mediates their interactions with the PaRBL proteins tips,youngstalks,youngfloralbuds,PLBs,andprotocorm-
(Fig. 3b). The full-length cDNAs of PaDP1 and PaDP2 containingactivelydividingcells,butwerealmostabsentor
123
212 PlantMolBiol(2014)84:203–226
barelydetectableinthefullydifferentiatedleaves(Fig. 5a, the developing protocorms (Fig. 5b). Other distinct CYC
Supplementary Fig. S2). PaCDKB1 and PaCYCB1;1 geneexpressionpatternswerealsoobserved.Forexample,
mRNAswereconcentratedintheshootapicalmeristemsof transcripts of PaCYCA2;2, PaCYCA3;1, and PaCYCB3;1
123