Reference-aided full-length transcript assembly, cDNA cloning, and molecular characterization of coronatine-insensitive 1b (COI1b) gene in coconut (Cocos nucifera L.)

In the Philippines, 26% of the total agricultural land is devoted to coconut production making coconut one of the most valuable industrial crop in the country. However, the country’s multimillion-dollar coconut industry is threatened by the outbreak of coconut scale insect (CSI) and other re-emerging insect pests promoting national research institutes to work jointly on developing new tolerant coconut varieties. Here, we report the cloning and characterization of coronatine-insensitive 1 (COI1) gene, one of the candidate insect defense genes, using ‘Catigan Green Dwarf’ (CATD) genome sequence assembly as reference. Two (2) splicing variants were identified and annotated—CnCOI1b-1 and CnCOI1b-2. The full-length cDNA of CnCOI1b-1 was 7919 bp with an ORF of 1176 bp encoding for a deduced protein of 391 amino acids while CnCOI1b-2 has 2360 bp full-length cDNA with an ORF of 1743 bp encoding a deduced protein of 580 amino acids. The 3D structural model for the two (2) isoforms were generated through homology modelling. Functional analysis revealed that both isoforms are involved in various physiological and developmental plant processes including defense response of plants to insects and pathogens. Phylogenetic analysis confirms high degree of COI1 protein conservation during evolution, especially among monocot species. Differential gene expression via qRT-PCR analysis revealed a seven-fold increase of COI1 gene expression in coconut post introduction of CSI relative to base levels. This study provided the groundwork for further research on the actual role of COI1 in coconut in response to insect damage. The findings of this study are also vital to facilitate the development of improved insect-resistant coconut varieties for vibrant coconut industry.


Introduction
More than 90 countries produce over 60 million metric tons of coconut annually [1]. Currently, Philippines is the third largest producer of coconut around the world but remains as the top global producer of coconut oil [2]. However, the outbreak of coconut-scale insect (CSI) in 2014 which affected more than 2.6 million nut-bearing coconut trees threatened the coconut production in the country [3]. Various insect pest control methods were utilized in order to address the outbreak including physical, chemical, and biological control approaches.
Recombinant DNA technologies allow the effective cloning and characterization of candidate insect defense genes. However, most of the genes involved for insect defense in coconut have not been fully sequenced and characterized. One of the candidate insect defense response genes in coconut is coronatine-insensitive 1 (COI1) gene which is reported to be involved in the jasmonate (JA) signaling pathway [4]. Jasmonate is known as an immunity hormone that promotes plant growth-and defense-related processes, such as, wounding and necrophitic pathogens [5]. In the JA signaling pathway, COI1 protein forms a complex with SCF (Skp/Cullin/F-box) to form an E3 ubiquitin ligase known as SCFCOI1 complex [6]. This protein complex is necessary for the ubiquitination and subsequent degradation of JAZ proteins, transcription repressors in JA signaling pathway, causing the de-repression of transcription factors and activation of JA response genes. To date, the only identified jasmonate receptor is the COI1 protein [7]. In Arabidopsis, COI1 mutants were found to be insensitive to JA resulting in defects in pest and pathogen resistance [8,9]. COI1 plays a vital role in wound-and JA-signaling since it was found to be required for the activation and repression of genes associated with wounding and JA signaling [10]. Similarly, COI1 in rice was reported to be indispensable in signaling component of rice defense response to chewing insects [11].
With the advent of next-generation sequencing (NGS) and bioinformatics tools, the first whole genome sequence assembly of catigan green dwarf" (CATD) variety of coconut has been generated using PacBio SMRT sequencing platform and 50 × Illumina Miseq, and improved through Dovetail Chicago® [17]. With the aid of the genome assembly and information from transcriptome data, full-length complementary deoxyribonucleic acid (cDNA) can be obtained. Furthermore, defense and host-response genes such as COI1 can be characterized at the genome level with the aid of the genome assembly.
The main objective of this study is to clone and elucidate the molecular characteristics of COI1 gene in coconut. Specifically, this study aims to (1) generate the full-length cDNA sequence of COI1b from transcriptome data in reference to 'catigan green dwarf' (CATD) genome sequence assembly, (2) clone partial COI1b cDNA sequence to validate the predicted full-length cDNA sequence of COI1b, (3) to characterize COI1b gene and obtain phylogenetic relationships of coconut COI1b against known COI1 protein sequences, and (4) to determine relative expression levels of COI1 gene in pre-and pos-CSI insfetation.

Generation of COI1b full-length cDNA sequences
The full-length assembly of COI1b transcript was generated using genome-guided transcript assembly approach. The gene sequence of COI1 in Elaeis guineensis (NW_011551140.1) was downloaded from the database and used as input for BLASTn [18]) search for gene homolog in the existing build of coconut 'catigan green dwarf' genome [17]. The COI1-harboring genomic scaffold of coconut was extracted using a local perl script (https:// github. com/ solge nomics/ sgn-bioto ols/ blob/ master/ bin/ fasta_ extra ct. pl) and utilized as reference in the genome-guided assembly of COI1b full-length transcript. Sequence read archive (SRA) files from 'chowghat green dwarf' (CGD; SRR1173229), 'taro green dwarf' (TGD; SRR1273070), 'aromatic green dwarf' (AGD; SRR1273070), and 'west coast tall' (WCT; SRR1137438) coconut varieties were retrieved from NCBI and preprocessed using Trimmomatic v0.36 (SLIDING WINDOW: 5:30 LEADING: 5 TRAILING:5 MINLEN:85) [19] to remove low-quality base calls. The preprocessed RNA-seq reads were aligned to the COI1b-containing coconut genomic scaffold using TopHat2 [20] aligner with default parameters. The resulting binary map alignment (BAM) files from each data set were used as input to the genome-guided Trinity software module [21] to assemble the full length COI1b transcript expressed across the four (4) coconut varieties.

Plant material
Healthy and coconut scale insect (CSI)-infested coconut leaf samples were collected from CATD variety of coconut at the Institute of Plant Breeding, University of the Philippines Los Baños. Sterile distilled water and RNAse AWAY™ (Invitrogen Corporation, USA) was used to wash sample surface and remove possible RNA-degrading contaminants.

Extraction of total RNA
Approximately 250 mg of leaf tissue was used for total RNA extraction using PureLink™ Plant RNA reagent (Invitrogen Corporation, USA) following the manufacturer's protocol. A total of 20 μL diethyl pyrocarbonate (DEPC) water was used to resuspend each sample. In order to ensure the purity of the samples, the isolated total RNAs were treated with DNase (Invitrogen Corporation, USA) followed by 1% agarose gel electrophoresis in 1 × TBE buffer (90 mM tris-borate, 2 nM EDTA) to check the RNA integrity. The quantity of the RNA was further determined using Qubit 2.0 Fluorometer (Life Technologies Corporation, USA). The first-strand cDNA was synthesized using Superscript III First-Strand Synthesis System (Invitrogen Corporation, USA) following the kit's protocol.

PCR amplification
The mRNA sequence of COI1 among different palm species was mined in the NCBI public repository (https:// ncbi. nlm. nih. gov) and Elaeis guineensis coronatine-insensitve 1 homolog b (XM_010909322.2) cDNA sequence was retrieved. With the retrieved cDNA sequence, the genespecific primers (GSPs) were designed using Primer3Plus software [22] The designed GSPs were subjected to in silico PCR using the CATD genome sequence assembly to determine complementation and specificity prior to primer synthesis. The primer pair exhibiting in silico amplification product with specificity to the target gene of interest was sent for outsourced primer synthesis (Invitrogen Corporation, USA).
The thermal cycler profile to amplify the mid-transcript of the target gene is as follows: 94 °C for 2 min; 35 cycles of 94 °C for 15 s, 62 °C for 30 s, and 68 °C for 1 min. The PCR reaction cocktail has a total reaction volume of 25 μL composed of 10X HiFi Buffer (1 ×), MgSO 4 (2 mM), dNTPs (0.2 mM), 0.2 μM each of forward (5′-GCA TTG GAA GAG TTT GGT GGG GGC TCA -3′) and reverse primer (5′-TCC AGC TTC TGC AGG CTT GGG CAA C-3′), Platinum® Taq DNA Polymerase HiFi (1U), and cDNA (30 ng). The PCR products were subjected to 1% agarose gel electrophoresis in 1 × TBE buffer (90 mM tris-borate, 2 nM EDTA) for 40 min at 100 V to validate fragment size, and specificity and quality of the amplification product. Once validated, amplified cDNA fragments were excised from the agarose gel and further purified using the PureLink® Quick Gel Extraction Kit (Life Technologies Corporation, USA). The gel extracted cDNA fragments were eluted with DEPC water to a final volume of 35 μL.

Gene cloning
The mid-transcript PCR product was ligated to pGEM®-T Easy vector (Promega Corporation, USA), following the manufacturer's instructions and transformed into Escherichia coli JM109 competent cells. The transformants were screened through blue-white screening and T7/SP6 colony PCR. The positive transformants were grown overnight in Luria-Bertani (LB) broth containing ampicillin (100 µg/ mL). The plasmids were isolated using the PureLink® Quick Plasmid Miniprep (Life Technologies Corporation, USA) following manufacturer's instructions. Then, 3 μL of the purified plasmid DNA recovered was subjected to plasmid PCR using T7/SP6 sequencing primers to a total volume of 10 μL in order to detect the integrity of the recovered DNA plasmids. High quality plasmid extracts were sent for double-pass outsourced capillary sequencing (AIT Biotech, Singapore).
Raw sequence files (.abl and.seq file) that were generated from the capillary sequencing were pre-processed based on the sequence alignment of the designed primer pairs using the Vector NTI software (Invitrogen, USA). The trimmed sequences were then loaded to the CAP3 Sequence Assembly Program [23] in order to generate the contig assembly from the double-pass paired reads.

Characterization of reference-aided full-length COI1b cDNA sequences and deduced amino acid sequences
Multiple sequence alignment (MSA) was performed on the assembled COI1b full-length assembled variants from CGD, TGD, AGD, and WCT and the corresponding CATD gene sequence of COI1b gene in CATD using Clustal Omega [24]. The structural gene annotations were manually done using SnapGene software (GSL Biotech; www. snapg ene. com). Similarly, the cloned mid-transcript sequence was also aligned to the consensus sequences to further validate the COI1b gene identity. The cDNA sequences of CnCOI1 were also compared with sequences found in the database using BLASTn [18].
The deduced amino acid sequences of coconut COI1b were derived by translating the mRNA sequences at different reading frame using the ORF Finder (https:// www. ncbi. nlm. nih. gov/ orffi nder/) and conserved domains were identified using NCBI Conserved Domain Database (https:// www. ncbi. nlm. nih. gov/ Struc ture/ cdd/ wrpsb. cgi). The composition of the amino acid sequences, molecular weight, and isoelectric point (pI) were predicted and analyzed using ProtParam (https:// web. expasy. org/ protp aram/). The 3D molecular models were constructed through homology modelling using I-TASSER [25][26][27] and functional analysis using COFAC-TOR and COACH (https:// zhang lab. ccmb. med. umich. edu). The coconut COI1b amino acid sequences were also used for BLASTp search [18] and retrieval of homologous COI1 proteins from other species in the NCBI database (https:// ncbi. nlm. nih. gov). Multiple sequence alignment using Clustal Omega [24] was also performed followed by phylogenetic tree construction using FastTree [28] to establish protein divergence and relationships among the COI1 amino acid sequences.

Differential gene expression of COI1
Differential gene expression by qRT-PCR was conducted to validate response to insect damage stimuli. A separate set of qRT-PCR primers targeting the COI1 gene in coconut were specifically designed based on the MIQE guidelines [29] and were outsourced for synthesis (Forward: 5′-GAG GTG AGG AAT GTC ATA GG-3′; Reverse: 5′-ACC CTA CCC TGT TCA TCC -3′). Three (3) biological replicates each of 'catigan green dwarf' (CATD) pre-and post-artificial CSI infestation was initiated in glasshouse conditions. CSI was introduced to seedlings based on the procedure described by Cortaga et al. (2019) [30]. Post-infestation leaf samples were gathered 14 days after CSI introduction (approximately when visual symptoms appear). Healthy and infested leaf samples were stored in − 80 °C prior to RNA extraction.
RNA extraction was carried out using Macherey-Nagel™ NucleoSpin™ RNA Plant Kit (Fisher Scientific, UK) following the manufacturer's instructions. The isolated RNA from each sample were immediately converted into cDNA using iScript™ cDNA synthesis kit (Bio-Rad Laboratories, California, USA) and stored in − 20 °C. qRT-PCR reactions were performed using SsoAdvanced Universal SYBR Green Supermix (Bio-Rad Laboratories, California, USA) following manufacturer's protocol in CFX96 Touch Real-Time PCR Detection System (Bio-Rad Laboratories, California, USA). UBC10 was used as internal control gene as previously established in coconut [31]. C q values of the three technical replicates per sample were recorded and analyzed using the 2 −ΔΔCq method of relative gene expression [32].

Generation and validation of full-length COI1b cDNA
Multiple sequence alignment of the COI1b full-length assembled transcripts showed two (2) splicing variant detected in CGD and TGD while only one (1) splicing variant for WCT and AGD. CGD assembled transcript 1 and TGD assembled transcript 2 have 7590 bp and 7919 bp transcript length, respectively while all the remaining transcripts are around 2000 bp in length. All the varieties have 5′-3′ orientation except for WCT. Manual annotation from the transcript assembly showed that there are two (2) COI1b splicing variants in coconut (Fig. 1).
The cloned mid-transcript sequence (752 bp; Supplementary Material 1) of coconut COI1b gene was aligned to the two (2) CnCOI1b splicing variant in coconut generated using genome-guided transcript assembly approach (Fig. 1). It covers approximately 10% of CnCOI1b-1 cDNA (7919 bp) and 32% of CnCOIb-2 cDNA (2360 bp). The cloned mid-transcript sequence also showed 96% similarity with E. guineensis COI1b as revealed by BLASTn results. No sequence variation (Supplementary Material 2) on the cloned mid-transcript was detected for healthy and insectinfested leaf samples.

Characterization of COIb1 cDNA sequences
The full-length cDNA of the CnCOI1b-1 was 7919 bp, containing a 1176 bp ORF, with a 5′ UTR of 6477 bp upstream the start codon and a 3′UTR of 266 bp downstream the stop codon (Table 1). On the other hand, CnCOI1b-2 has 2360 bp full-length cDNA, containing a 1743 bp ORF, with a 5′ UTR of 351 bp upstream the start codon and a 3′UTR of 266 bp downstream the stop codon. Both COI1b transcript variants in coconut have more than 8000 bp gene length. However, CnCOI1b-1 has significantly longer full-length cDNA, but shorter ORF compared to CnCOI1b-2. This observation is likely caused by the non-splicing of the intron 2 in the transcript variant CnCOI1b-1 relative to the sequence of the CnCOI1b-2. Upon further investigation within the sequence of the intron 2, several stop codons were found in the nucleotide stretch of the expressed sequence. Thus, it is classified as the 5'UTR of the longest ORF predicted by ORF Finder. Moreover, CnCOI1b-2 and EgCOI1b have a very long 5561 bp intron which is absent in CnCOI1b-1, AtCOI, AsCOI1, and HbCOI1. Regardless, the cDNA of coconut COI1b sequences share 80.63% identical nucleotides (Supplementary Material 3) and more than 95% identical to the COI1b of cDNA sequence of E. guineensis. In terms of cDNA sequence characteristic, E. guineensis COI1 (XM_010909322.2) showed same number of exons (4) and introns (3), and same length of ORF region (1743 bp) with CnCOI1b-2. Of all the coconut and representative dicot cDNA sequences of COI1, CnCOI1b-1 had the longest gene length while AtCOI1 had the shortest. In terms of full-length cDNA sequence, CnCOI1b-1 was the longest and the shortest is HbCOI1. Although CnCOI1b-1 had the longest gene and full-length cDNA, it had the shortest ORF. COI from Aquilaria sinensis had the longest ORF compared to CnCOI1b, E. guineensis, and other dicot species. Furthermore, known COI1 sequences of representative dicot species (Table 1) displayed more similarity to CnCOI1b-2 compared to CnCOI1b-1.

Structural and functional analysis of coconut COI1b protein
The information provided by structural and functional analysis of coconut COI1 can provide the essential and novel details about its biological and cellular function. Though COI1 protein has been extensively characterized in other crops [6,12,15,16], its specific function in coconut is yet to be described. The translated amino acid sequences from the two (2) coconut isoforms, CnCOI1b-1 and CnCOI1b-2, span 391 and 580 amino acid residues, respectively ( Table 1). The molecular weight was predicted to be higher in CnCOI1b-2 (65.5 kDa) as compared to CnCOI1b-1 (44.0 kDa). Both coconut COI1b isoforms are rich in leucine residues (> 14%). CnCOI1b-1 protein contains 31.7% helix, 11.8% strand, and 56.5% coiled. On the other hand, CnCOI1b-2 protein contains 35.5% helix, 11.6% strand, and 52.9% coiled. The two isoforms have AMN1 domain and leucinerich repeat domain (LRR)-8 LRRs for CnCOI1b-1 and 13 LRRs for CnCOI1b-2 (Fig. 2).
Compared to known COI1 protein sequences from other species, CnCOI1b-1 has the lowest molecular weight (44.0 kDa) and the least number of LRR domains (8) ( Table 1). In terms of secondary protein structure, no significant difference can be observed across COI1 proteins. Only CnCOI1b-1 showed different tertiary structure, having a horseshoe-like shape (Fig. 3). All the compared protein sequences showed AMN1 as the top domain hit. Threading templates used by I-TASSER are based on Z-score or the difference between the raw and average scores in the unit of standard deviation. Based on the result, the highest scoring threading template for both isoforms was the structure of COI1-ASK1 in complex with coronatine and incomplete JAZ1 degron (3ogkB) detected 5 × out of 10, from Arabidopsis thaliana, classified as protein binding, and expressed in Spodoptera frugiperda. The final model for CnCOI1b-1 and CnCOI1b-2 (Fig. 3) were chosen based on the C-score among the model predictions. The highest structural analog for both isoforms was the Transport Inhibitor Response 1 or TIR1 (2p1oB and 2p1nB), a signaling protein from A. thaliana and expressed in S. frugiperda.

Homologous alignment and phylogenetic analysis of coconut COI1b
Evolutionary analysis of the coconut COI1b isoforms and their orthologs revealed that these proteins form two (2) subfamilies, one composed of monocots and the other from dicots. The two (2) coconut COI1b isoforms are most related to E. guineensis COI1b, which indicate that both have almost similar structure and likely share some same gene function (Fig. 4). Two (2) subclades can be observed in the monocot group, one group composed of Arecaceae and the other composed of Musaceae, Bromeliaceae, and Poaceae. COI1 protein in Arecaceae family is more related to Poaceae compared to Musaceae and Bromeliaceae based on the phylogenetic tree.

COI1 Gene expression in coconut 'CATD'
The gene expression levels of COI1 in coconut var. 'catigan green dwarf' (CATD) in response to the introduction of coconut scale insect were assessed to verify its potential role in the cascade of defense mechanisms of coconut to the insect pest. qRT-PCR results revealed an approximately seven-fold increase in gene expression in the CSI-infested state relative to base levels, 14 days after CSI introduction. Figure 5 shows the average expression values (2 −∆∆Cq ) calculated from three (3) biological replicates pre-and post-CSI introduction at 1.88 and 7.76, respectively. The gene expression levels in coconut seedlings infested with CSI were highly variable (σ = 6.02) as compared to the healthy coconut (σ = 1.77), indicating that the COI1 expression was starting to decline as a component of initial basal defense response to insect.

Discussion
Previously, COI1 gene have been cloned and characterized from Arabidopsis, Aquilaria, rubber, and a few other plant species [6,12,15,16] but not from Cocos nucifera. The present study is the first report on the cloning and characterization of the gene encoding COI1b protein in coconut. COI1, the first identified F-box protein, is one of the components of the SCF complex that mediates the activation of JA response genes [12]. Mutations in COI1 in other plant species resulted to JA insensitivity to defense response and physiological [12,15,16] which suggest that COI1 is an essential conserved component of JA signaling pathway in plant secondary metabolism. The multiple sequence alignment of the assembled COI1b full-length assembled transcripts against the coconut CATD gene scaffold showed a different orientation for WCT. Possible explanation to this observations are RNA-seq library preparation method (stranded and non-stranded), the treatment (infested or diseased) of the raw read sequence source, and varietal differences. Furthermore, the coconut COI1b was found to have two (2) splicing variant-low MW (CnCOI1b-1) and high MW (CnCOI1b-2). Both splicing variants share high similarity with the coronatine-insensitive protein homolog 1b in E. guineensis. Long full-length cDNA sequence in CnCOI1b-1 and short number of amino acid residue suggest that the two mRNAs might be generated by an alternative splicing from a single gene. The occurrence of this kind of post-transcriptional modification has also been reported in human proteosomal modulator subunit, p27 (PSMD9) wherein the longer cDNA sequence encodes a shorter polypeptide sequence [33]. Based on transcript data, both splicing variants can be expressed in coconut.
Based on striking similarity between COI1b and TIR1, a good structural model (C-score = 0.05, 1.08) for CnCOI1b-1 and CnCOI1b-2 were presented. The observed resemblance between the structures of COI1b and TIR1 provided indirect evidence to explain the similarity of the signaling pathways for JA and auxin as previously described [13]. The conserved LRR domains in CnCOI1b-1 assemble into a solenoid fold which resembles horseshoe shape while CnCOI1b-2 has a spiral structure. The characteristics difference of the LRR domains among the two (2) isoforms might imply that the integrity of the LRR domains is necessary to the structural framework of COI1. Ja-Ile, ligand for the COI-JAZ protein interaction, was reported to bind at the LRR domain of COI1 [12]. The LRR domain of TIR1, highest structural analog for both isoforms, was known to contain inositol hexakisphosphate co-factor and recognizes auxin [34].
The identified ligand OGK (Coronatine) of the CnCOI1 was known to have functions associated to plant fertility and defense response, regulation of genes induced by wounding, regulation of jasmonate, among many others [12,35,36]. Some GO terms are unique for each isoform while some are in consensus with one another. Both isoforms almost have the same molecular function except that CnCOI1b-1 has GTPase activator activity which binds to and increases the activity of a GTPase, an enzyme that catalyzes the hydrolysis of GTP. CnCOI1b-1 have biological process associated to auxin mediated signaling pathway while response to insect was only detected in CnCOI1b-2. Both have SCF ubiquitin Multiple alignment analysis showed that CnCOI1b had more than 80% sequence identity with COI1 protein of monocots and more than 65% in dicots, which suggest that COI1 is highly conserved specially in monocots. Phylogenetic analysis confirms the high degree of COI1 conservation during the evolution, which reflects the selective pressure imposed by the vital functions of COI1 in plants.
It is evident that COI1 is upregulated during CSI infestation based on qRT-PCR validation, thus suggesting that it has a vital role in host response against insect damage. However, to further investigate potential breeding strategies, the next course of action is to perform a time-course analysis of gene expression alongside genes in the JA signaling pathway. JA biosynthesis, along with other conjugated and/ or oxidized products, begins within minutes of damage as previously demonstrated by other research studies [37][38][39]. Furthermore, expression analysis of putative COI1 transcript variants could provide insights on potential multiple gene functions and its role in endogenous plant host defense mechanisms.

Conclusion
The coronatine-insensitive 1 gene (COI1) was cloned from coconut using 'catigan green dwarf' (CATD) genome sequence assembly as reference in this paper. Two (2) splicing variants were identified-CnCOI1b-1 (low MW) and CnCOI1b-2 (high MW). Although both isoforms revealed different 3D structural models, various physiological and developmental plant processes including defense response to insect and pathogens were conserved. However, differential gene expression response of these coconut COI1b isoforms against insect damage at various time points is still a subject of further research. This work may provide the foundation of future research on understanding the role of COI1 in coconut. Results of this study are also expected to assist in the development of new resistant coconut varieties as one of the strategies to address threats in coconut production.
Acknowledgements This research was conducted through funds provided by Department of Science and Technology -Philippine Council of Agriculture, Aquatic and Natural Resources Research and Development (DOST -PCAARRD) through the Coconut Project 8 entitled, "Development of web-based breeding resource and Eco-TILLING towards insect resistance breeding in coconut". We gratefully acknowledge the Institute of Plant Breeding for allowing us to use the facilities for the conduct of this study. Also, the authors extend their sincerest gratitude to Dr. Susan R. Strickler and Dr. Lukas A. Mueller of the Boyce Thompson Institute for Plant Research (BTI), Ithaca, New York, USA for sharing their expertise in generating the high-quality reference genome assembly. The authors would also like to thank Ms. Kristal L. Lanceta of Philippine Coconut Authority IV-A for the CATD seedlings, Mr. Joseph P. Lagman for the rearing of coconut scale insect, and Ms. Cynthia R. Gulay for her technical assistance during the conduct of the laboratory experiments. The authors extend their warmest thanks to Academician Evelyn Mae Tecson-Mendoza for her support and guidance.