Multifacets in silico study of Plant Defensin 1.1 protein (PDF1.1) in Brassica rapa (turnip)

The amino-acid sequences of PDF1.1 protein were retrieved from the Ensembl Plant Database (https://plants.ensembl.org/index.html) (Cao et al. 2020) That contains the genomic sequences of various plant species. Blastp was done by using a genomic portal where we analyze the evolutionary history at sequence, structure, and gene family level, Phytozome v.13 (https://phytozome-next.jgi.doe.gov/blast-search) (Li et al. 2023) used to find orthologues in B. rapa FPsc v1.3, Prunus persica v2.1, Triticum aestivum v2.2, and in Arabidopsis thaliana TAIR10. The specific transcript, genomic sequence, and CDS files of PDF1.1 from B. rapa were downloaded. The genomic IDs of B. rapa PDF genes are given in Table 1. Named the B. rapa PDF1.1 proteins as BRa.PDF1.1 for convenience.

To analyze the phylogenetic relationship between the PDF1.1 proteins of B. rapa FPsc v1.3, Prunus persica v2.1, Triticum aestivum v2.2, and Arabidopsis thaliana TAIR10, the Mega-X software was used where we upload the FASTA file that carries the amino-acid sequences of all these proteins and aligned by selecting option, i.e., multiple sequence alignment (Stecher et al. 2020). By using the results of alignment, a phylogenetic tree was created with Maximum likelihood with 300 Bootstrap values, and the Jones–Taylor–Thornton (JTT) model. The radial tree was constructed to view the clades separately in different colors using the itol (https://itol.embl.de/) (Letunic and Bork 2021).

The physicochemical properties of BRa.PDF1.1 proteins, including the amino acid chain length, molecular weight (kDa), number of negatively and positively charged residues, isoelectric point (pI), instability index, GRAVY, and aliphatic index, are predicted by using the ExPASy-ProtParam tool (https://web.expasy.org/protparam/) (Garg et al. 2016). This particular tool takes the amino-acid sequence of the protein to be analyzed in FASTA format and then processes it accordingly. However, the results are presented in the form of different percentages and scores in tabular form.

To predict the location of BRa.PDF1.1 proteins within the cell, WoLF PSORT (https://wolfpsort.hgc.jp/) (Jiang et al. 2021) were used and the heatmap that contains the combined results was generated by applying the TBtool (https://github.com/CJ-Chen/TBtools) (Chen and Xia 2022).

The chromosomal mapping of BRa.PDF1.1 genes in B. rapa were predicated by using the Mapgene2chromosome (MG2C) web v2.1 database (http://mg2c.iask.in/mg2c_v2.0/) (Chao et al. 2021). This particular tool enables us to analyze that at which location and at which arm of chromosome the required gene would be located.

NCBI Conserved Domain database (https://www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi) (Yang et al. 2020) was used to predict the protein family and conserved domain of the BRa.PDF1.1 protein which are the conserved sequences associated with the specific function. To find motifs that are the small sequence of amino acids in the BRa.PDF1.1 protein sequences, MEME suite version 5.5.1 (https://meme-suite.org/meme/) (Nystrom and McKay 2021) was used by selecting the maximum 20 motifs limit that plays an essential role in protein structure and function. However, the gene structure display server 2.0 (http://gsds.gao-lab.org/) (Hu et al. 2015) was used to generate the intron/exon schematic diagram of BRa.PDF1.1 genes or to determine the exact number of coding and noncoding regions within the gene.

The PANZZER2 web server (http://ekhidna2.biocenter.helsinki.fi/sanspanz/) (Törönen and Holm 2022) was used to predict the gene ontology or functions of the BRa.PDF1.1 protein at biological, molecular, and cellular levels with positive predicted values (PPVs). The KEGG analysis was done by using the BlastKOALA web server (https://www.kegg.jp/blastkoala/) (Kanehisa and Sato 2020). By this analysis, we can predict and visualize the cascade like how the protein is activated in which stress and involved in which pathway.

The exploration of functional sites that include phosphorylation and glycosylation sites in PDF1.1 proteins (Table 2) was achieved by using the NetPhos 3.1 server (https://services.healthtech.dtu.dk/services/NetPhos-3.1/) (Li et al. 2018), selecting the score >0.5, which shows the phosphorylation with related kinases and NetNGlyc 1.0 server (https://services.healthtech.dtu.dk/services/NetNGlyc-1.0/) (Joshi and Gupta 2015), respectively.

The SOPMA online server (https://npsa.lyon.inserm.fr/cgi-bin/npsa_automat.pl?page=/NPSA/npsa_sopma.html) (Angamuthu and Piramanayagam 2017) was used for the comprehensive secondary structure analysis of BRa.PDF1.1 protein which includes alpha helices, beta turns, beta strands, random coils, etc. This server also processes the FASTA format amino-acid sequence of the given protein and presents the results in different forms (peaks and bars). However, for homology modeling or three-dimensional (3D) structure prediction the BRa.PDF1.1 proteins, the Phyre (Paul et al. 2019) web server (http://www.sbg.bio.ic.ac.uk/phyre2) was used. After the prediction of 3D model of a protein, it is always necessary to validate its quality. For this purpose, Ramachandran plot was employed which is available on PROCHECK server (https://saves.mbi.ucla.edu/) (Laskowski et al. 2013). This tool analyzes the given protein on the basis of the stereochemical properties (analyzing the torsional angles phi and psi) and demonstrates the results in the form of a map, comprises of four different quadrants. Each quadrant provides the details regarding the most favorable, additionally allowed, generously allowed and unfavored regions, depending on the steric clashes, the respective amino acids of the given protein possess. However, the active protein pocket sites were predicted by using the CASTp tool (http://sts.bioe.uic.edu/castp/calculation.html) (Tian et al. 2018). This specific tool predicts different active sites or pockets of varied area and volume sizes that are presented in different colors. In context of a molecular interaction, these sites are responsible for interaction with other proteins or any other interacting molecules.

The following figure illustrates the whole workflow which is followed in this work.

The phylogenetic analysis of PDF 1.1 proteins was constructed by employing the maximum-likelihood method which exhibited thought-provoking results. The selection of Prunus persica and Triticum aestivum as variants for constructing the phylogenetic tree (Fig. 1) was based on the orthology aspect of PDF1.1 proteins. These two species exhibited close orthology than the rest of the any other specie that is why P. persica and T. aestivum were specifically chosen. In this particular phylogenetic tree, a total of seven different clades were produced. This tree was further annotated to be converted into radial form to present in a bit more tangible manner. Each clade was presented with their respective colors to discriminate them from each other. This tree was constructed to infer the phylogenetic relationship of PDF1.1 proteins of B. rapa with those of Arabidopsis thaliana, Prunus persica, and Triticum aestivum. Clade 1 is designated with a light blue color, comprised of just two members, i.e., AT.PDF.1h and Bra.PDF1.1.f; Clade 2 was the largest among the rest of the others. It comprised various variants of PDF1.1 proteins of B. rapa and Arabidopsis thaliana. Clade 3 contained just one AT.PDF1.1f protein which exhibited its evolutionary association with PDF1.1.h and PDF1.1.c proteins of B. rapa. However, Clades 4 and 5 represented unique phylogeny behaviors. Clade 4 comprised four entities and all were different variants of Bra.PDF1.1 proteins. All of them were found in evolutionary association with each other. Clade 5 had only one entity i.e., BRaPDF1.1.j. The second largest clade in the phylogenetic tree was Clade 6, which is represented with a mild brown color in Fig. 2. This group displayed a diverse evolutionary linkage among PDF1.1 proteins of Arabidopsis thaliana, B. rapa, and Triticum aestivum. Clade 7 also showed a similar relatedness fashion of PDF1.1 proteins which belong to B. rapa, Arabidopsis thaliana, and Prunus persica. In conclusion, B. rapa PDF1.1 demonstrates greater similarity to Arabidopsis thaliana PDF1.1 than the PDF1.1 proteins of T. aestivum because multiple evolutionary linkages were detected at various clades of the phylogenetic tree.

The results of physicochemical characterization exhibited that the BRa.PDF1.1 proteins possessed mostly positively charged amino acids with protein lengths ranging between 57 and 130, and molecular weights varying from 6162.26 to 13867.87 kDa, while their instability index (II) ranges between 14.82 and 51.39. Furthermore, the BRa.PDF1.1 proteins GRAVY ranged from 0.060 to 0.72, their isoelectric point (pI) ranged from 5.05 to 9.57, and their aliphatic index ranged from 57.00 to 99.47. The compiled results of the physicochemical properties of all these proteins are summarized in Table 1.

The subcellular analysis of the proteins under study was performed by WoLF PSORT. It presented the results in the form of a heat map. The different colors in the heat map indicated the relevant scores concerning the localities of these proteins in various cellular and extracellular compartments. These results suggested that the BRa.PDF1.1 proteins are primarily located in the extracellular spaces but some of them are also located in chloroplast, cytosol, endoplasmic reticulum, nucleus, vacuoles, and golgi bodies. The graphical demonstration of this analysis is illustrated in Fig. 3. The names of the proteins and the cellular compartments are mentioned along the X-axis and Y-axis, respectively. A colored scheme is also given which contains different confidence scores for correspondence to infer the proteins’ respective localization.

The localization of B. rapa PDF1.1 genes on chromosomes was predicted by using the Mapgene2choromosome web v2.1 database and is represented in Fig. 4. The results of this chromosomal mapping showed that the maximum number of genes are located on chromosome no. 7. Four genes were found to be located on chromosome no. 2, one on chromosome no. 5, chromosome no. 6, and chromosome no. 8. However, seven genes are located on chromosome no. 7.

Protein domains were identified by using the NCBI conserved domain database. Results showed that the BRa.PDF1.1 protein sequences possessed a common conserved domain, i.e., gamma thionin, and four of the query sequences showed no conserved domains. The function of the gamma-thionin family protein adopts the “knottin” fold, a stable cysteine-rich scaffold, in which one disulfide bridge crosses the macrocycle made by two other disulfide bridges and connects segments of the backbone. They are found in plant lectins/antimicrobial peptides, plant proteinase/amylase inhibitors, plant gamma-thionins, and arthropod defensins. Figure 5 represents the results of the domain analysis of BRa.PDF1.1s.

On the other hand, the protein motifs were identified by using the MEME suite version 5.5.1. The results of motifs prediction are presented in Fig. 4, which illustrates the motif locations along their corresponding p-values. Notably, 20 motifs were found in the amino-acid sequences of the proteins under study. Each motif displays a specific sequence and is represented with various colors which can be observed in Fig. 6 and Fig. 7. The intron/exon organization of BRa.PDF1.1 genes were performed by using the Gene Structure Display Server 2.0. All of the genes comprised two exons that are represented in yellow color and one intron which is represented in black lines as presented in Fig. 8.

To perform the gene ontology BRa.PDF1.1 proteins, PANZZER2 web server was employed. Nine of 14 proteins were found to be involved in various biological processes. Five proteins were predicted to be involved in different molecular functions and nine proteins were involved in the other cellular processes. The major biological processes found were the degeneration of cells of another organism, defense response to fungus, jasmonic acid, and ethylene-dependent systemic resistance and response to insects (GO:0031640, GO:0050832, GO:0009861, and GO:0009625). Similarly, the molecular function predicted was the protein binding (GO: 0005515), while the cellular components found were extracellular region, secretory vesicle, and membrane (GO: 0005576, GO: 0099503, and GO: 0016020). The KEGG analysis was also performed to explore more functions of Bra.PDF1.1 proteins by BlastKOALA. These results demonstrated that most of the BRa.PDF1.1 proteins were involved in the mitogen-activated protein kinase (MAPK) signaling pathway. The MAPK pathway is involved in the transduction of extracellular signals to the nucleus or cytoplasm for a specific cellular response. In plants, the MAPK signaling pathway is triggered by various biotic or abiotic stress stimuli such as pathogen infection, drought, salinity, high temperature, etc. The pathways in which these proteins were predicted to be involved by KEGG are displayed in Fig. 9.

The protein phosphorylation sites were predicted by using the NetPhos 3.1 server. These results suggested that the proteins had phosphorylation sites ranging from 2 sites to 15 sites with specific and nonspecific kinases. These phosphorylation sites were present only in three specific amino acids, i.e., serine, threonine, and tyrosine. Moreover, no N-Glycosylation sites were predicted in the protein sequences when subjected to the NetNGlyc 1.0 server.

The BRa.PDF1.1 protein’s secondary structure was predicted by using the SOPMA online server. Results showed that the alpha helix is the major component of BRa.PDF1.1 proteins. Other respective percentages of all the elements of the secondary structure including the beta sheets, extended strands, random coils, and beta turns are compiled in Table 3.

The 3D structures were predicted by using the Phyre (Paul et al. 2019) web server. All the selected protein models represented the maximum confidence (61.1% to 100.0%) and residue coverage (18% to 64%). All of the predicted models are displayed in Table 4. Furthermore, the model quality was evaluated by the Ramachandran plot. This analysis revealed that 59.1% to 100.0% of amino-acid residues in all the proteins under consideration were allocated in the favorable region which suggested the good quality of their predicted structures. The very next was to predict the active sites/pockets of BRa.PDF1.1 proteins. These pockets were predicted by using the CASTp tool. The pockets or active sites that were predicted during this particular analysis are shown in red color in Table 4. The amino-acid residues that were contained in these pockets were Alanine (Ala), Cytosine (Cys), Lysine (Lys), Glutamine (Gln), Isoleucine (Ile), Leucine (Leu), Arginine (Arg), Serine (Ser), Aspartic acid (Asp), Glycine (Gly), Valine (Val), Threonine (Thr), Glutamic acid (Gln), and Tryptophan (Trp). All of these pockets were of different area and volume sizes based on the number of residues they possessed.