untitled ACTA BOT. CROAT. 75 (1), 2016 1 Acta Bot. Croat. 75 (1), 1–10, 2016 CODEN: ABCRA 25 DOI: 10.1515/botcro-2016-0003 ISSN 0365-0588 eISSN 1847-8476 Cloning, characterization and expression analysis of NBS-LRR-type resistance gene analogues (RGAs) in coconut Kaitheri Edathil Rachana1, Sudalaimuthu Asari Naganeeswaran1, Thayale Purayil Fayas1, Regi Jacob Thomas2, Muliyar Krishna Rajesh1* 1 Division of Crop Improvement, ICAR-Central Plantation Crops Research Institute, Kasaragod 671124, Kerala, India 2 ICAR-Central Plantation Crops Research Institute (RS), Kayamkulam, Kerala, India Abstract – Coconut palms are highly susceptible to diseases caused by different pathogens, and replanting with resistant varieties is the best way to manage them. Obtaining a collection of resistance gene analogues (RGAs) is an effective strategy to identify genomic regions linked to disease resistance. We have successfully used a comparative genomics approach to amplify putative RGAs from the coconut root (wilt) disease resis- tant cultivar Chowghat Green Dwarf (CGD) by using primers designed based on conserved motifs of the NBS-LRR domain of the date palm. The amplifi ed sequences were cloned, sequenced and characterized. The coconut RGAs had high identity to monocot NBS-LRRs. A complete structural analysis and 3-D modeling of the NBS domain of coconut RGA was also undertaken. Real-time quantitative polymerase chain reaction analysis indicated that the isolated coconut NBS-LRR class RGAs was expressed more in root (wilt) disease resistant genotypes than in susceptible ones. This study would provide a base for future efforts to map disease resistant traits in coconut. Key words: coconut, homology modeling, nucleotide binding site-leucine rich repeat type, real time PCR, resistant gene analogue * Corresponding author, e-mail: mkraju.cpcri@gmail.com Introduction The coconut palm (Cocos nucifera L.) is one of the most important and useful palms in the world and has been a part of human history for over 4000 years. It is an impor- tant crop in the agrarian economy of many tropical coun- tries of the world, providing food, drink, shelter and raw materials for industry. The coconut palm is a member of the palm family Arecaceae and is the only accepted species in the genus Cocos. Coconut is confronted with an array of diverse pathogens (bacterial, fungal, and phytoplasma), which contributes substantially to yield loss. Thus, breeding in coconut has to aim at the development of high-yielding hybrids with resistance to pathogens. During the course of evolution, plants have developed complicated defense mechanisms to resist phytopathogens (Staskawicz et al. 1995). The identifi cation of many plant resistance (R) genes has helped to understand how plants prevent or slow down infection and disease progression (Hammond-Kosack and Jones 1997). The plant R genes en- code proteins that can recognize and bind to products en- coded by the avirulence genes of the invading pathogens (Scofi eld et al. 1996). As a result of such a molecular recog- nition, a signal transduction cascade is initiated, and plant defenses are expressed. These defense responses include hypersensitive response that results in localized cell death, and other general responses such as strengthening of the cell wall, formation of phytoalexins etc. (Dangl et al. 1996). Plant R genes have been cloned and characterized from a number of species, including both mono- and dicotyle- donous plants. These R genes can be broadly divided into four structurally distinct classes (Ellis et al. 2000). The fi rst class of resistance genes belongs to the serine-threonine ki- nases (Ritter and Dangl 1996). These protein kinases phos- phorylate serine/threonine residues and thus control certain signaling networks during the resistance response. The sec- ond class of resistance genes encodes putative trans-mem- brane receptors with extra-cellular leucine rich repeat (LRR) domains (Jones et al. 1994). The third class encodes for a receptor-like kinase and combines qualities of both the previous classes. Both the LRR domain and the protein ki- nase regions are encoded in the same protein. The fourth RACHANA K. E., NAGANEESWARAN S. A., FAYAS T. P., THOMAS R. J., RAJESH M. K. 2 ACTA BOT. CROAT. 75 (1), 2016 class, which represents the majority of plant disease resis- tance genes cloned so far, is the nucleotide binding site-leu- cine rich repeat (NBS-LRR) resistance genes. The leucine- rich repeat (LRR) domains provide pathogen recognition specifi city by participating in protein-protein interactions and ligand binding (Ellis et al. 2000). NBS-LRR proteins are composed of three domains: a variable N-terminal domain (~200 amino acids), a nucleo- tide binding site domain (~300 amino acids) and a variable LRR domain (~10–40 short Leucine-Rich-Repeat motif) (Cannon et al. 2002). The NBS domain consists of several conserved motifs including the P-loop and the kinase-2 do- mains, which are ATP- and GTP-binding sites (Meyers et al. 1999); the kinase-3a domain and the putative membrane spanning hydrophobic GLPL domain (Baldi et al. 2004). The conserved motifs offer opportunities for designing de- generate primers for polymerase chain reaction (PCR)- based strategies to recover similar sequences called resis- tance gene analogues (RGAs) from other plant species (Kanazin et al. 1996). In the present study, our aim was to clone and character- ize RGAs from coconut, whose genome has not been se- quenced and for which the genomic resources available in public databases are scarce. We used oligonucleotide prim- ers designed to the conserved motifs of NBS-LRR domain of resistance genes of date palm, the genome of which has been sequenced, to recover RGAs from coconut. The sec- ondary structure of partial R-protein and the 3-D model of NB-ARC domain of isolated RGAs were predicted. Gene expression studies were carried out using qRT-PCR to con- fi rm the up-regulated expression patterns of the coconut RGAs in disease resistant genotypes. Materials and methods Retrieving date palm genome sequence and creation of a standalone database Whole genome sequences of date palm (Phoenix dacty- lifera L. ‘Khalas’ variety), comprising a total of 1, 42, 304 whole genome shotgun sequence (which includes 28,889 annotated genes), were retrieved (http://qatar-weill.cornell. edu/research/datepalmGenome/index.html) and a standalone date palm sequence database was constructed using format- db program available in standalone BLAST tool for com- parative analysis. PCR amplifi cation of RGAs and sequencing Nucleotide sequences of date palm NBS-LRR-type R genes were translated using EMBOSS TRANSEQ program and searched against Pfam (http://pfam.sanger.ac.uk) and PRINTS (http://www.bioinf.man.ac.uk/dbbrowser/PRINTS/ index.php) databases. Motifs identifi ed with Pfam and PRINTS were confi rmed using PROSITE (http:/www.ex- pasy.ch/prosite/). Based on sequence structural motif prediction by PROSITE/PRINTS database, major disease resistant amino acid motifs characteristic of the NBS-LRR class were iden- tifi ed in NBS-LRR proteins of date palm and these were used to design specifi c oligonucleotide primers for poly- merase chain reaction (PCR) amplifi cation of coconut ge- nomic DNA. Three pairs of forward and reverse primers were designed from different consensus motifs. A fourth primer was designed in the internal region of the fi rst prim- er. The oligo parameters were optimized by PCR oligonu- cleotide resources such as OLIGOANALYZER (http:// www.idtdna.com/analyzer/applications/oligoanalyzer/) and designed primers for effi cient amplifi cation of coconut RGA by FASTPCR primer design software (http://primer- digital.com/fastpcr.html). DNA was extracted from the spindle leaf samples of the root (wilt) disease resistant cultivar Chowghat Green Dwarf (CGD) palm using a modifi ed SDS-procedure (Rajesh et al. 2013). PCR was carried out in a 20 μL reaction mixture con- taining 25 ng template DNA, 200 nM each of forward and reverse primer, 10 mM of each dATP, dCTP, dGTP and dTTP, 10× PCR buffer, 15 mM MgCl2 and 1 U Taq DNA polymerase (M/s Bangalore Genie, India). The cycling pro- fi le consisted of an initial denaturation at 94 °C for 2 min followed by 34 cycles each consisting of DNA denaturation at 94 °C for 30 s, primer annealing at the standardized an- nealing temperature for 1 min and primer extension at 72 °C for 1 min, and a fi nal extension for 10 min in a thermal cycler (BIORAD). PCR products were further analyzed by electrophoresis on 1.2% agarose gel (1× TBE containing 0.1 μg mL–1 ethidium bromide), under 80 V for 1 h. After electrophoresis, the gel images were taken using a gel docu- mentation system (BIORAD) for later analysis. The PCR-amplifi ed products were purifi ed with the use of a PCR purifi cation kit (QIAGEN). Longer amplicons were sequenced by primer walking strategy using the de- signed internal primers. All amplicons were cloned and se- quenced. Sequencing was done using a BigDye® Termina- tor v3.1 sequencing kit and analyzed on ABI PRISM® 377 Genetic Analyzer. Sequence analysis Raw sequences were trimmed to remove the vector con- tamination using VECSCREEN. The trimmed sequences were analyzed to identify potential open reading frames (ORFs). Amino acid identities of coconut RGAs to well characterized gene sequences were determined by perform- ing translated BLAST (BLASTX) with non-redundant ami- no acid sequences (http://blast.ncbi.nlm.nih.gov/blast.cgi/). Multiple sequence alignment was carried out using the soft- ware CLUSTALX. This alignment was used in the MEGA software to build neighbor- joining tree under the Jones- Thornton-Taylor (JTT) model (Tamura et al. 2011) with 1000 bootstraps. The amino acid sequences of cloned known disease resistance genes L6 (LUU27081) from fl ax (Linum usitatissimum), RGC4A (AF017754) from lettuce (Lactuca sativa) and RPP1 (AF098962), RPS2 (ATU14158) and RPM1 (NM111584) from Arabidopsis and Rx (AJ 011801) from potato (Solanum tuberosum) were retrieved from Genbank. RGAs from date palm (gi_672130146) and oil palm (gi_6690745) were also used as references. The NBS-LRR domains were extracted and aligned with coco- NBS-LRR RESISTANT GENE ANALOGUES IN COCONUT ACTA BOT. CROAT. 75 (1), 2016 3 nut RGAs. The comparison of coconut sequence with date palm sequence was also done with the use of mVISTA (MLAGAN algorithm) server (Frazer et al. 2004). The nucleotide sequences obtained were translated in silico into protein sequences using EMBOSS TRANSEQ program and motifs present in the sequences were con- fi rmed using Pfam, PROSITE, PRINTS and SMART data- base search. The conserved motifs within the NBS domain were identifi ed using MEME server (multiple expectation maximization for motif elicitation) (Bailey et al. 2006). Additionally, protein sequences obtained by in silico translation of isolated coconut RGAs were subjected to BLASTP analysis. Highly signifi cant hits obtained by BLASTP analysis were subjected to the SMART tool (http://smart.embl-heidelberg.de) for predicting the NB- ARC domain. The selected sequences were aligned using MAFFT with jalview (http://mafft.cbrc.jp/alignment/soft- ware/), for the multiple sequence alignment and phyloge- netic tree with bootstrap method was constructed. Homology modeling For further characterization of the translated proteins, we submitted the amino acid sequences to a secondary structure prediction server, PSIPRED (http://bioinf.cs.ucl. ac.uk/psipred/). The two feed forward neural networks of PSIPRED perform an analysis on output obtained from PSI-BLAST (position specifi c iterated BLAST) homology search algorithm (Altschul et al. 1997, Jones 1999). The pre liminary 3-D framework of our amino acid sequences based on homology molecular modeling was initiated by the search for a suitable template for 3-D structure predic- tion of the NBS and LRR region of coconut RGA in Protein data bank (PDB). Templates with signifi cant scores were selected. In fold and function assignment (FFSA03) (Jaro- szewski et al. 2005), an alignment profi le was created with template and the target protein sequence was then deposited as an input for ProtMod (http://ffas.burnham.org/protmod- cgi/protModHome.pl) structure modelling server for 3-D structure, with SCWRL as default modelling method, and MODELLER (https://salilab.org/modeller/). Desired PDB fi le of predicted model was generated by ProtMod, subse- quently viewed by Pymol software version 1.0. The accu- racy of predicted 3-D-protein model was further verifi ed by SAVES (structural analysis and veri fi cation server, http://ni- hserver.mbi.ucla.edu/SAVES), where the PDB fi le of the protein model was subjected separately for the analysis. Further, the protein model validation was also done based on ϕ – ψ ratio of Ramachandran plot obtained from PRO- CHECK analysis (Wiederstein and Sippl 2007), which provides detailed view on the stereo-chemistry. Finally, PROSA-Web (https://prosa.services.came.sbg.ac.at/prosa. php) search was applied to assess the energy criteria of the constructed protein structure. qRT-PCR analysis Total RNA from root (wilt) disease resistant and suscep- tible CGD genotypes was isolated using plant RNA kit (M/s QIAGEN) and integrity, concentration, quality checks and DNase l treatment were performed. Reverse transcription was performed with High Capacity RNA-to-cDNA kit (M/s Life Technologies) using aliquots of total RNA extracted. The cDNA samples were diluted to 20 ng μL–1. CN-NBS- LRR gene specifi c primers were designed by using PRIM- ER 3.0 (http://primer3.ut.ee/) software and selected for PCR amplifi cation. Amplifi cation was carried out with a re- action mixture containing 2 μL of 10× reaction buffer, 2 μL of 25 mM MgCl2, 0.2 μL of 10 mM dNTP mix, 1 μM of each forward and reverse primer, 1 μL of synthesized cDNA and 1 U of Taq polymerase. RT-PCR reactions were carried out using BIORAD thermal cycler. The cycling pro- gram was as follows: 1 min at 94 °C, 30 cycles of 30 s at 94 °C, 1 min at 55 °C, 1 min at 72 °C and a 10 min extension at 72 °C, with a negative control. All real-time PCR reactions were performed using the StepOne™ real time PCR system (Applied Biosystems) and the amplifi cations were done using the SYBR green PCR master mix (Applied Biosystems). The thermal condi- tions were as follows: initial holding stage 52 °C for 2 min, 95 °C for 10 min, followed by 40 cycles at 95 °C for 15 s and a fi nal step at 60 °C for 1 min. The experiments were carried out in triplicate for each data point. The relative quantifi cation in gene expression was determined using the 2-ΔΔCt method (Livak and Schmittgen 2001). Using this method, we obtained fold change gene expression in resis- tant and susceptible genotypes, normalized to an internal control gene, α-tubulin (α-TUB). Results Identifi cation of motifs of R gene in date palm Many R genes have been cloned from different plant species and these genes have been found to share some structural domains, which include P-loop (kinase-1a), the kinase-2, the GLPLAL and the MHD. Characteristic disease resistant motifs of date palm NBS-LRR RGA sequences (PDK_30s665281g007, PDK_30s742701g004 and PDK_ 30s816511g003) were identifi ed from the motif prediction tools such as PROSITE/ PRINTS/ Pfam. Based on sequence structural motif prediction by the PROSITE/PRINTS data- base, four major disease resistant amino acid motifs charac- teristic of the NBS-LRR class were identifi ed in the three NBS-LRR proteins of date palm and these were used to de- sign specifi c oligonucleotide primers for polymerase chain reaction (PCR) amplifi cation of coconut genomic DNA (Fig. 1a). PCR amplifi cation of targeted RGA fragments in coconut Primers were designed in such a way that around 2200, 1200, 900 and 600 bp PCR products were obtained upon amplifi cation respectively using the designed primers (Fig. 1a, Tab. 1). All four primer pairs, designed according to the consensus motifs of R genes in date palm, were used for PCR amplifi cation of template DNA of coconut root (wilt) disease resistant CGD genotype. These primer pairs pro- duced expected size bands according to the date palm spe- cies (source) from which the primers were designed (Fig. 1b). Primer CN1F1/R1 amplifi ed the largest band of ap- RACHANA K. E., NAGANEESWARAN S. A., FAYAS T. P., THOMAS R. J., RAJESH M. K. 4 ACTA BOT. CROAT. 75 (1), 2016 proximately 2200 bp size and CN3 F1/R1 amplifi ed a target of around 1200 bp. In addition, a 600 bp sized fragment was amplifi ed by CN4 F1/R1. Primer mCN1F1/R1, de- signed from the middle portion of the date palm RGA se- quence, gave an additional amplifi ed product of around 900 bp in coconut, which is in between the product amplifi ed by CN1 F1/R1. Target specifi c large amplicon (2200 bp) were purifi ed, cloned and sequenced by primer walking. These sequences were deposited in the GenBank database and their accession numbers are KC465244 (2211 bp; designat- ed as CN1- NBS-LRR), KF002584 (1165 bp; designated as CN3- NBS-LRR) and KM983337 (616 bp; designated as CN4- NBS-LRR). Analysis of sequence homology and characterization Like other RGA sequences, coconut RGA sequences with high identity to each other at the amino acid level may belong to the same gene families or subfamilies. The three putative RGAs from coconut, which showed ≤ 90% identi- ty to each other, were selected for further characterization (On-line Suppl. Tab. 1). Coconut RGAs were highly similar to RGAs cloned from different genera viz., Oryza, Triticum, Aegilopus, Musa and Brachypodium (On-line Suppl. Tab. 1). The maximum amino acid identity between coconut RGAs and other plants’ RGAs ranged from 45–100% (On- line Suppl. Tab. 1).The multiple sequence alignment analy- sis of coconut RGAs with known resista nt genes of Toll and Interleukin receptor-NBS-LRR (TNL) and coiled-coiled- NBS-LRR (CNL) classes showed that coconut RGA se- quences were highly clustered with CNL- NBS sequences (Fig. 2a) and contain typical motifs (P-Loop, RNBS-A-non TIR, Kinase- 2, RNBS-B, RNBS-C and GLPL) of the CNL class R proteins within the NBS-domain (Fig. 2b) including RPM1 (gi_440572080) and RPS2 from Arabidopsis (gi_549979) and Rx from potato (gi_5524754), NBS-LRR sequences from date palm (gi_672130146, gi_672112993) and oil palm (gi_6690745). Coconut RGAs were found to cluster with date palm and oil palm RGAs. TNL sequences formed a separate group which included L6 from fl ax (gi_862905), RPP1 from Arabidopsis (gi_332644385), RGC4A from lettuce (gi_2852690) and contained typical motifs within the NBS domain (P-Loop, RNBS-A-TIR, Ki- nase-2, RNBS-B, RNBS-C and GLPL). The amino acid identity among the three coconut RGAs was found to be 15–26%. Fig. 1. a) Diagram showing the NBS domain motifs used in primer design. (CN)nF and (CN)nR are general NBS primer sets (CN1F1/R1, CN3F2/R2, CN4F3/R3; see Tab. 1) designed from P-loop and GLPL motifs of three different date palm resistance gene analogue se- quences; b) Gel profi le of PCR products amplifi ed from four R gene of coconut by specifi c motif-based primers: CN1F1/R1 (around 2200 bp), mCN1F1/R1 (around 900 bp), CN3F2/R2 (around 1200 bp), and CN4F3/R3 (around 600 bp). First line on each gel presents 1 Kb DNA marker. Tab. 1. Specifi c primers of disease resistant motifs used for screening for NBS-LRR resistance gene analogues in coconut. S. No. – sam- ple number. S. No. Primer name Forward (5’–3’) Tm (°C) Conserved motif and R–Gene Amplifi ed product size (bp) Annealing temperature (°C) 1 CN1F1 CTCTGGACAATCAGTGATG 56 ALLIVGMAGLGKTTLA 2200 58 CN1R1 GGGCTTGGGAAAACAACT 58 HNLRSLEKLTITDCPEL 2 CN3F2 GAGTCCTAGATACCGAAGC 53.1 GLGKTTL 1200 58 CN3R2 GTTCTCCCAATAGTTGGC 50.1 GNLIQLRYLGLSDAKTK 3 CN4F3 GATGGGTGGGGTTGGTAAG 55.4 GIGIVGMGGVGKTLLA 600 58 CN4R3 CTACAGTCTTCGCAGCTAATG 53.2 CGGLPLAAKTVGEIL 4 mCN1F1 GATGGGTGGGGTTGGTAAG 60 – 900 58 mCN1R1 CTACAGTCTTCGCAGCTAATG 62 – NBS-LRR RESISTANT GENE ANALOGUES IN COCONUT ACTA BOT. CROAT. 75 (1), 2016 5 Alignment and the region of high conservation of R genes across two palms, coconut (CN1-NBS-LRR, CN3- NBS-LRR and CN4-NBS-LRR) and date palm (PDK_ 30s665281g007, PDK_30s742701g004, PDK_30s816511g 003) could easily be identifi ed and analyzed using mVISTA server. High percentage of conservation of coconut NBS- LRR sequences with date palm NBS-LRR sequences were observed (data not shown). The BLASTP analysis of the extracted NB-ARC domain of amino acid sequences of co- conut RGAs was carried out and the hits obtained were aligned using MAFFT multiple alignment program. Phylo- genetic tree was constructed using a neighbour-joining al- gorithm with 1000 bootstrap trials (On-line Suppl. Fig. 1). Analysis of conserved motifs in coconut RGA The three isolated coconut NBS-LRR nucleotide se- quences were translated into amino acids using EMBOSS TRANSEQ and subjected to PRINTS/ Pfam databases for Fig. 2. Relationship among three putative coconut resistance gene analogues (RGAs) and nine known plant NBS-LRR class R-genes based on amino acid sequences: a) Phylogenetic tree of three putative coconut resistance gene candidates and RGAs from date palm, oil palm, RPM1, RPP1, RPS2 from Arabidopsis, Rx of potato and L6 of fl ax. Numbers by the branch joints are bootstrap values (out of 1000) that support the clustering. The scale bar represents substitutions per site. b) Amino acid sequence alignment of three coconut RGAs and nine known NBS-LRR class R-genes from date palm, RPS2, RPM1 from Arabidopsis, Rx of potato of CC-NBS-LRR (CNL) class. Spe- cifi c motifs of P-Loop, RNBS-A-non TIR, Kinase-2, RNBS-B, RNBS-C and GLPL hydrophobic domain within the NBS-domain are shown. L6 of fl ax, RPP1 of Arabidopsis and RGC4A of lettuce were clustered together and the P-loop, RNBS-A-TIR, Kinase-2, RNBS- B, RNBS-C and GLPL motifs of TIR-NB-LRR class genes (TNL) within the NBS-domain are shown. Only amino acids in the above mentioned motifs are shown in the fi gure while amino acids between the motifs were replaced with ‘----’ to reduce the space needed. RACHANA K. E., NAGANEESWARAN S. A., FAYAS T. P., THOMAS R. J., RAJESH M. K. 6 ACTA BOT. CROAT. 75 (1), 2016 the prediction of conserved resistance motifs. NB-ARC and LRR domains were identifi ed from Pfam database (On-line Suppl. Tab. 2a), while seven signature motifs were identi- fi ed from PRINTS database (On-line Suppl. Tab. 2b). New- ly identifi ed coconut NBS-LRR type RGAs showed conser- vation in the nucleotide-binding domain. MEME analyses revealed several conserved motifs like P-Loop, RNBS-A- nonTIR, Kinase-2, Kinase-3 and RNBS- C motifs within the NBS-region of all the three coconut NBS-LRR type RGAs. The position and logos of the motifs within the NBS domain are shown in Fig. 3. Exact organization in each pro- tein is obtained by the length and position of each motif. Secondary structure of CNRGAs Secondary structure of proteins, comprising constantly repeating local structures like α- helix and β- sheet stabi- lized by hydrogen bonds, were predicted using the PSI- PRED algorithm. CN1-NBS-LRR was found to possess 20 helices, 50 random coils and 18 strands present at various positions in the protein structure (On-line Suppl. Fig. 2a), while CN3-NBS-LRR possessed 14 helices, 8 strands and 26 random coils (On-line Suppl. Fig. 2b) and CN4-NBS- LRR was made up of 10 helices, 4 strands and 16 random coils (On-line Suppl. Fig. 2c). 3-D structure prediction and validation of NBS-domain PDB search result for suitable template detection to pre- dict the 3-D structure of CN1-NBS-LRR showed maximum identity to chain A of APAF-1 (PDB ID: 1z6t:A) with an E < 0.0 and with signifi cant identity (61%). FFAS03 pair wise alignment showed that NBS region from 1 to 249 of CN1- NBS-LRR had the expected homology with 147 to 398 of template chain A. ProtMod generated the desired PDB fi le of the predicted 3-D model of NBS region of CN1-NBS- LRR, which was analyzed in PyMol software, version 1.0. Stereochemical quality, the different parameters, static in- teractions between the non-bonded atoms, stability etc. of the predicted protein model were verifi ed by SAVES. The 3-D structure of NBS region of CN1-NBS-LRR provides the complete details of different spatial arrangements and function of conserved motifs. The modeled cartoon struc- ture of NBS region is shown in Fig. 4a. NBS domain con- sists of motifs from N-terminus P-Loop (pink), RNBS-A (light blue), Kinase-2 (wheat), RNBS-B (blue), RNBS-C (yellow), RNBS-D (violet) and GLPL (cyan). The nucleo- tide binding cleft and relative arrangements of all the motifs of NBS are well conserved in the homology model of CN1- NBS-LRR. The catalytic cleft at the central region was found to be surrounded by other conserved motifs: P-loop, GLPL etc. The P-loop was made up of β-sheets fl anked by α-helix, kinase-2 of β-sheets and other motifs of α-helices. Fig. 3. Different conserved motifs span within nucleotide binding domain identifi ed using MEME. The name of each sequence and com- bined P-value are shown here. Different motifs are shown by different coloured boxes. NBS-LRR RESISTANT GENE ANALOGUES IN COCONUT ACTA BOT. CROAT. 75 (1), 2016 7 The stereo-chemical quality and accuracy of the pre- dicted model was assessed by Ramachandran map calcula- tion by PROCHECK program (Fig. 4c). In the Ramachan- dran plot, residue classifi cations were done according to its regions in the quadrangle. The red regions in the graph show most allowed regions and the yellow represent al- lowed regions. Glycine is represented by triangles and oth- ers are represented by squares. The Ramachandran plot de- rived by PROCHECK showed 91% residues in the most favorable region, 7.5% in additionally allowed region, 0.5% in generously allowed region and only 1.0% in disallowed region as compared to the 1z6t:A template profi le 93.1%, 6.1%, 0.0%, 0.0% respectively (Fig. 4c). Most of the amino acids were clustered in a ϕ – ψ distribution, consistent with right handed α-helix, indicating a good quality model. PDB search result for suitable template detection to pre- dict the 3-D structure of CN3-NBS-LRR and CN4-NBS- LRR showed maximum identity to chain A of Lycopersicon esculentum NBS domain (PDB ID: 2ft4:A) with an E < 0.0 and with signifi cant identity (58%). FFAS03 pair wise alignment showed that NBS region from 1 to 263 of CN3- NBS-LRR had the expected homology with 63 to 320 of template chain A and 1 to 205 of CN4-NBS-LRR with 51 to 339 of template. ProtMod generated the desired PDB fi le of the predicted 3-D model of NBS region of CN3-NBS-LRR and CN4-NBS-LRR, which were analyzed in PyMol and subjected to SAVES for quality check. The 3-D structure of NBS region of CN3-NBS-LRR provides the complete de- tails of different spatial arrangements and function of con- served motifs and the modeled cartoon structure of NBS region is shown in (On-line Suppl. Fig. 3a). The NBS do- main consists of the same motifs identifi ed in NBS domain of CN1-NBS-LRR (from N-terminus P-Loop, RNBS-A, ki- nase-2, RNBS-B, RNBS-C, RNBS-D and GLPL. The nu- cleotide binding cleft and relative arrangements of all the motifs of NBS were well conserved in homology model of CN3-NBS-LRR. Also, the catalytic cleft at the central re- gion was found to be surrounded by conserved motifs like P-loop and GLPL. The P-loop consisted of β-sheets fl anked by α-helix, kinase-2 of β-sheets and other motifs of α-he- lices. The stereo-chemical quality and accuracy of the pre- dicted model, assessed by Ramachandran map calculation derived by PROCHECK program (On-line Suppl. Fig. 3b) showed 87.3% residues in the most favorable region, 8.1% in the additionally allowed region, 3.8% in the generously allowed region and only 0.8% in the disallowed region (On-line Suppl. Fig. 3b). Most of the amino acids were clustered in ϕ – ψ distribution, consistent with right handed α-helix, indicating a good quality model. The modeled cartoon structure of NBS region of CN4- NBS-LRR is shown in On-line Suppl. Fig. 3c. Stereo- chemical quality and accuracy of the predicted model was Fig. 4. 3D models of CN1-NBS-LRR domains: (a) NBS domain (PDB ID- 1z6t:A); 3-D cartoon view revealed the relative position of different conserved motifs from N-terminus: P-Loop (pink), RNBS-A (light blue), Kinase-2 (wheat), RNBS-B (blue), RNBS-C (yellow), RNBS-D (violet), GLPL (cyan); (b) LRR domain (PDB ID-2z64:A); Cartoon shows a horseshoe shaped receptor like structure. The par- allel β-sheets forming the concave face of the ‘horseshoe’ are represented as yellow arrows; (c) Ramachandran plot for CN1-NBS model; (d) Ramachandran plot for CN1-LRR model. RACHANA K. E., NAGANEESWARAN S. A., FAYAS T. P., THOMAS R. J., RAJESH M. K. 8 ACTA BOT. CROAT. 75 (1), 2016 assessed by Ramachandran map calculation by PRO- CHECK program (On-line Suppl. Fig. 3d) showed that 81.6% of the residues were in the most favorable region, 12.3% in the additionally allowed region, 4.5% in the gen- erously allowed region and only 1.7% in the disallowed re- gion as compared to the 2ft4:A template profi le (On-line Suppl. Fig. 3d). Most of the amino acids were clustered in ϕ – ψ distribution, consistent with right handed α-helix, in- dicating a good quality model. The complete energy profi le and Z-score of the predict- ed models of the three CN-NBS–LRR RGAs were calcu- lated by PROSA-WEB search. The calculated values of the models ranged from –8.3 to –7.1, indicating the models to be acceptable. 3-D structure prediction and validation of LRR-domain The LRR region of CN1-NBS-LRR was found to con- tain a short stretch of amino acid residues with leucine at every second and third position, repeated to form fl exible, solvent exposed and parallel β sheets. The amino acid re- gion between residues 54 to 477 showed sequence homolo- gy with Toll-like receptor (PDB ID-2z64: A) with signifi - cant score and E-value < 0. ProtMod search predicted the 3-D structure of this LRR region by comparing with Toll- like receptor-4. The LRR region of CNRGA contains a se- ries of parallel β-sheets that form a concave face shaped like a horse shoe or banana (Fig. 4b). The number of re- peats in this LRR region was approximately 22 and each LRR contained series of about 20–26 amino acids having a consensus sequence. The Ramachandran plot showed 70.9% residues in the most favourable region, 27.1% in the allowed region and 1.4% in the generously allowed region and 0.6% in the dis- allowed region as compared to 2z64A template profi le as 71.4, 26.8, 1.6 and 0.1%, respectively. 3-D structure of LRR generated using PyMol is represented in Fig. 4d. Validation of expression of coconut RGAs To analyze the expression levels of isolated RGAs in the leaves of root (wilt) disease resistant and susceptible co- conut genotypes, two representatives (CN1-NBS-LRR and CN3-NBS-LRR) were used for expression analysis using RT-PCR. The primer pairs (Tab. 2) produced a single prod- uct (Fig. 5a). Real time profi ling of the endogenous control α-tubulin clearly indicated their stable expression in all the biological samples used for normalizing data (On-line Sup- pl. Figs. 4 a, b). Real time data indicated high expression levels of CN-NBS-LRR RGAs in resistant genotypes com- pared to susceptible genotypes (Fig. 5 b). Discussion Root (wilt) disease is one of the most devastating dis- eases of coconut palms in India. Breeding for disease resis- tance is one of the most important objectives in any crop breeding program, and is particularly relevant in the case of a long duration crop like coconut, where long generation time, low multiplication rate and requirement of a lot of land for experiments hinder conventional breeding efforts. To defend against diverse pathogens, all plants have de- veloped immunity by expressing R genes as receptors that recognize invading pathogens and mediate downstream sig- naling defense to either expel the pathogens or to prevent their spread (Staskawicz et al. 1995). New biotechnological approaches have provided much insight into the study of interaction between the plants and pathogens at different stages of infection. Identifying, screening and analyzing the diversity of R genes in various plant species are, therefore, vital to understanding the evolution of R genes and their functional acquisition. Several R genes against distinct pathogen types have been cloned and characterized (Shi- varamakrishnan and Seetharama 2001) and this has facili- tated our understanding on the mechanism of plant defenses and provided better insight in to effective management of plant diseases. In the present study, we report isolation and characterization of three putative resistant gene analogues from coconut, encoding NBS-LRR type proteins using a comparative genomics approach. Degenerate primers-based PCR strategies have been used for isolating and identifying putative R genes from dif- Fig. 5. qRT-PCR analysis of CN1-NBS-LRR and CN3-NBS-LRR in resistant (R) and susceptible (S) genotypes of coconut palm: a) RT-PCR gel image of α-Tubulin (1), CN1-NBS-LRR (2), CN3- NBS-LRR (3) and negative RT control (4); b) transcript level of CN1-NBS-LRR and CN3-NBS-LRR in resistant (R) and suscep- tible (S) genotypes. Data are average value ± SD, n = 3. R1–R5 and S1–S5 are biological replicates of resistant and susceptible Chowghat Green Dwarf leaves. Tab. 2. Coconut NBS-LRR gene-specifi c primer pairs for qRT- PCR. S. No. – sample number. S. No. Primer Name Sequence (5’–3’) Amplicon size (bp) 1 CN1F CCAGCGAGGTAGAGATGAGG 130 CN1R GGGAACTCACGACCGAAGTA 2 CN3F CAAGCTCCGTATGATCAGCA 125 CN3R AATGTCATCGGCATCGTACA 3 TUBF ATGTTCAGGCGCAAGGCT 101 TUBR CTGCAACCGGGTCATTCA NBS-LRR RESISTANT GENE ANALOGUES IN COCONUT ACTA BOT. CROAT. 75 (1), 2016 9 ferent plant species in earlier studies (Yu et al. 1996, Mago et al. 1999, Totad et al. 2005). Here we used an alternative, effi cient PCR method by aligning and comparing, and de- signed specifi c motif based primers of date palm, whose ge- nome has been sequenced. Four sets of primers designed from the conserved motifs (GLPL, P-loop and Kinase-2 etc.) were effi cient enough to produce target specifi c frag- ments. The designed primers amplifi ed fragments of 600– 2000 bp, which, on cloning and sequencing, showed signif- icant homology to known and well-characterized RGAs. Most NBS-LRR R genes implicated in plant disease re- sistance responses are known to share a common structure and are also presumed to possess a common evolutionary origin. Many studies in different plants have revealed new insights into the molecular evolution of NBS-LRR type R genes (McDowell and Simon 2006). Translated polypep- tides from the nucleotide sequences were subjected to ho- mology search by BLASTX algorithm for two reasons. Firstly, it is well-known that the protein level homology search and comparison have shown more homology with NBS-LRR region of many RGAs than at the nucleotide level due to genetic code degeneracy issue (Totad et al. 2005). Additionally, amino acid sequences around the structural motifs display a higher degree of conservation. Multiple sequence alignment of the NBS domain of the three coconut RGAs and known R genes showed signifi cant homology to the P-loop, kinase-2 and hydrophobic GLPL domain. All three coconut RGAs were classifi ed into the non-TIR-NBS-LRR subfamily of the R genes. The NBS- LRR classes of R genes are classifi ed into two subfamilies: the TIR-NBS-LRR subfamily, which is characterized by the presence of a highly conserved aspartic acid (D) or aspar- tate (N) as the last residue of the kinase-2 domain, and the non-TIR-NBS-LRR subfamily with a highly conserved tryptophan (W) as the last residue of the kinase-2 domain (Pan et al. 2000). N-terminal characteristic of only non-TIR class were present in coconut NBS-LRR RGAs isolated in the present study. 3-D structural modeling of R protein facilitates an un- derstanding of its molecular function in recognizing patho- gen effector protein. The crystal structure of NB-ARC do- main of mammalian APAF1 (Riedl et al. 2005) was used in templates for homology modeling. In our study, the predict- ed homology modeling based 3-D structure of the NBS do- main of CN-NBS-LRR revealed relative 3-D location of various conserved motifs present within this domain. CN- NBS-LRRs possessed conserved motifs that are located around the central catalytic cleft: the P-loop, kinase-2, hy- drophobic GLPL etc. all of which serve to orient the ADP molecule (Chattopadhyaya and Pal 2008). The LRR do- main is involved in protein–protein interactions, ligand binding (Jones and Jones 1996) and also determines the pathogen recognition specifi city of NBS-LRR proteins (Shen et al. 2003, Ellis et al. 2007). Up-regulation or higher expression level of CN-NBS- LRR genes in coconut root (wilt) resistant compared to sus- ceptible genotypes, as revealed by qRT-PCR, suggests that the sequence similarity-based targeted gene identifi cation approach has a high degree of accuracy. Breeding for resis- tance to coconut root (wilt) disease has been a long sought goal of coconut breeders. Identifi cation of genes that confer resistance to root (wilt) disease would pave the way for mo- lecular-based strategies to augment the effectiveness of breeding for disease resistant genotypes in coconut. Acknowledgements The work was supported by a grant from Indian Council of Agricultural Research (ICAR) and Kerala Biotechnology Commission (YIPB). References Altschul, S. F., Madden T. L., Schaffer, A., Zhang, J., Zhang, Z., Miller, W., Lipman, D. J., 1997: Gapped BLAST and PSI- BLAST: a new generation of protein data base search pro- grams. Nucleic Acids Research 25, 3389–3402. Bailey, T. L., Williams, N., Misleh, C., Li, W. W., 2006: MEME: Discovering and analyzing DNA and protein sequence motifs. Nucleic Acids Research 34, 369–373. Baldi, P., Patocchi, A., Zini, E., Toller, C., Velasco, R., Komjanc, M., 2004: Cloning and linkage mapping of resistance gene ho- mologues in apple. Theoretical and Applied Genetics 109, 231–239. Cannon, S. B., Zhu, H., Baumgarten, A. M., Spangler, R., May, G., Cook, D. R., Young, N. D., 2002: Diversity, distribution, and ancient taxonomic relationships within the TIR and non- TIR NBS-LRR resistance gene subfamilies. Journal of Mo- lecular Evolution 54, 548–562. Chattopadhyaya, R., Pal, A., 2008: Three-dimensional models of NB-ARC domains of disease resistance proteins in tomato, Arabidopsis, and fl ax. Journal of Bimolecular Structure and Dynamics 25, 357–71. Dangl, J. L., Dietrich, R. A., Richberg, M. H., 1996: Death don’t have no mercy: Cell death programs in plant-microbe interac- tions. Plant Cell 8, 1793–1807. Ellis, J. G., Lawrence, G. J., Dodds, P. N., 2007: Further analysis of gene-for-gene disease resistance specifi city in fl ax. Molecu- lar Plant Pathology 8, 103–109. Ellis, J., Dodds, P., Pyror, T., 2000: Structure, function and evolu- tion of plant disease resistance genes. Current Opinion in Plant Biology 3, 278–284. Frazer, K. A., Pachter, L., Poliakov, A., Rubin, E. M., Dubchak, I., 2004: VISTA: computational tools for comparative genomics. Nucleic Acids Research 32, W273–279. Hammond-Kosack, K., Jones, J. D. G., 1997: Plant disease resis- tance genes. Annual Review of Plant Biology 48, 575–607. Jaroszewski, L., Rychlewski, L., Li, Z., Li, W., Godzik, A., 2005: FFAS03: A server for profi le–profi le sequence alignments. Nucleic Acids Research 33, 284–288. Jones, D. T., 1999: Protein secondary structure prediction based on position-specifi c scoring matrices. Journal of Molecular Biology 292, 195–202. Jones, D. A., Jones, J. D. G., 1996: The role of leucine-rich repeat proteins in plant defences. Advances in Plant Pathology 24, 89–167. RACHANA K. E., NAGANEESWARAN S. A., FAYAS T. P., THOMAS R. J., RAJESH M. K. 10 ACTA BOT. CROAT. 75 (1), 2016 Jones, D. A., Thomas, C. M., Hammond-Kosack, K. E., Balint- Kurti, P. J., Jones, J. D. G., 1994: Isolation of the tomato Cf-9 gene for resistance to Cladosporium fulvum by transposon tagging. Science 266, 789–793. Kanazin, V., Marek, L. F., Shoemaker, R. C., 1996: Resistance gene analogs are conserved and clustered in soybean. Proceedings of the National Academy of Sciences 93, 11746–11750. Livak, K. J., Schmittgen, T. D., 2001: Analysis of relative gene expression data using real-time quantitative PCR and the 2-ΔΔCt method. Methods 25, 402–408. Mago, R., Nair, S., Mohan, M., 1999: Resistance gene analogues from rice: Cloning, sequencing and mapping. Theoretical and Applied Genetics 99, 50–57. McDowell, J. M., Simon, S. A., 2006: Recent insights into R gene evolution. Molecular Plant Pathology 7, 437–48. Meyers, B. C., Dickerman, A. W., Michelmore, R. W., Sivaramak- rishnan, S., Sobral, B. W., Young, N. D., 1999: Plant disease resistance genes encode members of an ancient and diverse protein family within the nucleotide-binding superfamily. Plant Journal 20, 317–332. Pan, Q., Wendel, J., Fluhr, R., 2000: Divergent evolution of plant NBS-LRR resistance gene homologues in dicot and cereal ge- nomes. Journal of Molecular Evolution 50, 203–213. Rajesh, M. K., Jerard, B. A., Preethi, P., Thomas, R. J., Fayas, T. P., Rachana, K. E., Karun, A., 2013: Development of a RAPD- derived SCAR marker associated with tall-type palm trait in coconut. Scientia Horticulturae 150, 312–316. Riedl, S. J., Li, W., Chao, Y., Schwarzenbacher, R., Shi, Y., 2005: The structure of the apoptotic-protease factor 1 bound to ADP. Nature 434, 926–934. Ritter, C., Dangl, J. L., 1996: Interface between two specifi c pathogen recognition events mediated by distinct plant disease resistance genes. Plant Cell 8, 351–257. Scofi eld, S. R., Tobias, C. M., Rathjen, J. P., Chang, J. H., Lavelle, D. T., Michelmore, R. W., Staskawicz, B. J., 1996: Molecular basis of gene-for-gene specificity in bacterial speck disease of tomato. Science 274, 2063–2065. Shen, Q. H., Zhou, F. S., Bieri, S., Haizel, T., Shirasu, K., Schulze- Lefert, P., 2003: Recognition specifi city and RAR1/SGT1 de- pendence in barley MLA disease resistance genes to the pow- dery mildew fungus. Plant Cell 15, 732–744. Shivaramakrishnan, S., Seetharama, N., 2001: Identifi cation and isolation of disease resistance genes in crop plants. Journal of Plant Biology 28, 25–38. Staskawicz, B. J., Ausubel, F. M., Baker, B. J., Ellis, J. G., Jones, J. D. G., 1995: Molecular genetics of plant disease resistance. Science 268, 661–667. Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M., Ku- mar, S., 2011: MEGA5: molecular evolutionary genetics anal- ysis using maximum likelihood, evolutionary distance, and maximum parsimony methods. Molecular Biology and Evolu- tion 28, 2731–2739. Totad, A. S., Fakruddin, B., Kuruvinashetty, M. S., 2005: Isolation and characterization of resistance gene analogs (RGAs) from sorghum (Sorghum bicolor L. Moench). Euphytica 143, 179– 188. Wiederstein, M., Sippl, M. J., 2007: ProSA-web: Interactive web service for the recognition of errors in three-dimensional structures of proteins. Nucleic Acids Research 35, W407– W410. Yu, Y., Buss, G., Maroof, M. A., 1996: Isolation of a super family of candidate disease resistance genes in soybean based on a conserved nucleotide binding site. Proceedings of the Nation- al Academy of Sciences 93, 11751–11756. NBS-LRR RESISTANT GENE ANALOGUES IN COCONUT ACTA BOT. CROAT. 75 (1), 2016 On-line Suppl. Tab. 1. E-values and amino acid identities of three coconut resistance gene analogues (CN RGAs) to the top hit Gene- bank accessions revealed by BLASTX searches. Top BLASTX hit, organism with Genebank accession number E-value Amino acid identity (%) Coconut NBS-LRR CN1 NBS-LRR resistance protein partial, Cocos nucifera (AGI96386.1) 0.0 100 NBS-LRR like protein, Oryza sativa Japonica group (AANO3742.1) 0.0 46 Putative disease resistance protein RGA4, Triticum urartu (EMS58751.1) 0.0 45 Coconut NBS-LRR CN3 NBS-LRR type resistance protein, partial, Oryza sativa Japonica Group (AB96985.1) 6e-167 66 Putative disease resistance protein RGA3-like, Brachypodium distachyon (XP_003566098.1) 3e-166 67 Putative disease resistance RPP13-like protein 1, Aegilops tauschii (EMT33711.1) 4e-161 66 Coconut NBS-LRR CN4 NBS-LRR disease resistance protein, Musa balbisiana x Musa textilis (CAP66367.1) 6e-91 67 Disease resistance RPP13-like protein 4-like, Brachypodium distachyon (XP_003564401.1) 7e-54 51 On-line Suppl. Tab. 2. Disease resistant motifs viewed in (a) Pfam and (b) PRINTS. a) Pfam Sequence Description Entry type clan Envelope Alignment HMM Bit-Score E-value start end start end from to CN1-NBS-LRR NB-ARC domain Domain CL0023 1 256 1 254 29 285 231.8 6.6e-69 CN3-NBS- LRR CN4-NBS-LRR LRR domain NB-ARC domain NB-ARC domain Family Domain Domain CL0022 CL0023 CL0023 365 1 1 405 263 205 366 1 1 404 261 205 2 3 2 40 254 170 40.6 201.5 200.3 1.2e-10 5.4e-29 4.6e-007 b) PRINTS Finger Print ID score P VAL PF Score Motif No Sequence Position Length Low High Disease Resistant Motifs CN1 50.83 46.00 36.32 1.08-e07 2.39-e06 2.9-e04 386 296 166 1 of 4 2 of 4 3 of 4 QLQGKRYLLVLLDW IKGLPLAAKTLGGLL IKIKCLRVLDLSHNDIK 64 167 362 15 15 17 0 0 0 0 0 0 Coconut RGA CN3 51.3 42.8 2.31e-08 8.19e-06 402 206 2 of 4 3 of 4 KLKGKRFLLVLDDIW LKGLPLAAKTLASLL 76 173 15 15 0 0 0 0 Coconut RGA CN4 44.67 33.00 3.67e-07 2.98e-05 342 236 2 of 4 3 of 4 QLMGKRYLIVFDDVW CSGLPLAAKTVD 84 195 15 15 0 0 0 0 On-line Suppl. Fig. 1. Phylogenetic tree constructed based on the selected resistance gene analogues. Numbers by the branch joints are bootstrap values (out of 1000) that support the clustering. The scale bar represents substitutions per site. 1 RACHANA K. E., NAGANEESWARAN S. A., FAYAS T. P., THOMAS R. J., RAJESH M. K. 2 ACTA BOT. CROAT. 75 (1), 2016 On-line Suppl. Fig. 2. Secondary structure prediction of (a) CN1-NBS-LRR; (b) CN3-NBS-LRR; (c) CN4-NBS-LRR. The prediction line (Pred) denotes the predicted conformation for each residue such as helical (H), random coil (C) and extended (E). The confi dence (Conf) indicates the reliability of the prediction for each position. NBS-LRR RESISTANT GENE ANALOGUES IN COCONUT ACTA BOT. CROAT. 75 (1), 2016 On-line Suppl. Fig. 3. (a) 3-D model of NBS domain of CN3-NBS-LRR. 3-D cartoon view revealed the relative position of different conserved motifs from N-terminus: P-Loop, RNBS-A,Kinase-2, RNBS-B, RNBS-C, RNBS-D, GLPL; (b) Ramachandran plot for CN3- NBS model; (c) Molecular 3-D- model of NBS domain of CN4-NBS-LRR. 3-D cartoon view revealed the relative position of different conserved motifs from N-terminus: P-Loop, RNBS-A,Kinase-2, RNBS-B, RNBS-C, RNBS-D, GLPL; (d) Ramachandran plot for CN4- LRR model. On-line Suppl. Fig. 4. (a) Real time profi ling of α-tubulin in resistant (R1-R5) and susceptible (S1-S5) samples; (b) Melt curve analysis. 3