To identify and characterize the KCS genes in the passion fruit (Passiflora edulis) genome, passion fruit proteins, CDS, gff3, genome files were downloaded from the passion fruit genome¹ (Ma et al., 2021). We performed two methods, which are Basic Local Alignment Search Tool for proteinsT (BLASTp) and Hidden Markov Models (HMMER) search tool, to identify the KCS genes in the passion fruit genome. All known KCS family protein sequences of Arabidopsis thaliana (AtKCS) (Joubès et al., 2008) were downloaded from The Arabidopsis Information Resource (TAIR) database², M. domestica (Apple) MdKCS (Lian et al., 2020) from apple genome³, and A. hypogaea (Peanut) AhKCS (Huai et al., 2020) from peanut genome⁴. The BLASTp was performed in the passion fruit genome based on protein homology alignments with default mode using the known KCS protein sequences via the TBTools software package version 1.0986858 (Chen C. et al., 2018). The KCS domains were downloaded from the Pfam database⁵ [PF08392 (FAE1/Type III polyketide synthase-like protein) and PF08541 (3-Oxoacyl-[acyl-carrier protein (ACP)] synthase III C terminal)]. hese domains were used for HMMER in passion fruit genome via TBTools with default mode. The results of the two methods were merged and further analyzed by Web CD-Search Tool⁶ and SMART⁷ to confirm the domains in passion fruit KCS protein.

The PeKCS genes’ physical and chemical properties include molecular weight (M.W), amino acid (A.A) length, coding sequence (CDS) length, isoelectric point (pl), instability index (II), aliphatic index (Ai), number of an exon, and intron (E/I) were evaluated by using the ExPASY-Prot⁸. The online CELLO version 0.2.5⁹ program was used to predict the subcellular localization of PeKCS genes. The phylogenetic analyses were performed by aligning the full-length KCS protein sequences of P. edulis (PeKCS), A. thaliana (AtKCS), M. domestica (MdKCS), and A. hypogaea (AhKCS) by Molecular Evolutionary Genetics Analyses (MEGA) software version 10.1.8¹⁰. The aligned sequences were subjected to neighbor-joining (NJ) tree construction using the MEGA software with 1,000 bootstrap replicates and all other parameters were set to default. The online web tool iTOL¹¹ was used to visualize the phylogenetic tree and divided it into four main groups and eight subgroups according to AtKCS (Joubès et al., 2008). The PeKCS genes distribution and mapping on all 9 assembled and 1 unassembled scaffold passion fruit chromosomes were investigated by using the passion fruit genomic and annotation files via Tbtools software (Chen C. et al., 2018). Moreover, the online Sequence Identity and Similarity (SIAS)¹² tool was used to perform a pairwise sequences identity between passion fruit KCS genes.

The PeKCS genes structural organizations and chromosome locations were retrieved from passion fruit genomic files. The conserved motifs of PeKCS protein sequences were predicted by using the Multiple Expectation Maximization for Motif Elicitation (MEME) online tools¹³ and the numbers of motifs were set to 10 motifs. The PeKCS gene structures and conserved motifs from the passion fruit were visualized by TBTools software.

To predict the putative cis-regulatory elements in promoter regions of PeKCS genes, the upstream 2,000 bp sequences of all the PeKCS genes were extracted from the genomic DNA sequences. Moreover, the promoter sequence of each gene was subjected to the PlantCARE database¹⁴ and the cis-regulatory elements figure was drawn by using TBTools software (Chen C. et al., 2018). Additionally, the numbers, functions, and sequences of putative cis-regulatory elements of PeKCS genes were summarized and highlighted into plant growth and development, phytohormones responses, and stress-responsive categories.

The tandem and segmental or whole-genome duplication (WGD) duplication provides a new understanding of family genes development and genome progression. The homologous PeKCS genes with only one intervening gene on the same passion fruit chromosome were considered to be tandem duplicated, while homologs on other chromosomes were considered to be segmental duplications. PeKCS duplicated genes were identified and visualized on their physical location on the chromosomes in the passion fruit genome. Gene duplication, synteny analysis, and Ka (non-synonymous)/Ks (synonymous) values calculation of PeKCS genes were performed in TBTools. The syntenic relationships of KCS genes between P. edulis, A. thaliana, and M. domestica were performed using the MCScanX toolkit, and KCS relationships between the target species were visualized by the Advance Circos package in the TBTools software (Chen C. et al., 2018). Furthermore, the multiple collinearity analysis of KCS genes in P. edulis, A. thaliana, M. domestica, and A. hypogaea whole genomes were also performed by Multiple synteny Plot packages in TBTools software. The Ka), Ks nucleotide substitution rates, and the Ka/Ks ratios of duplicated PeKCS genes were annotated using TBTools and divergence time (T, MYA; million years ago) was calculated by the following formula; T = Ks/2x (x = 6.38 × 10^–9) (Ma et al., 2021).

The online tool STRING version 11¹⁵ was used to predict and construct the protein–protein interaction network of PeKCS proteins based on Arabidopsis homologous proteins. The STRING tool parameters were set as follows; network type-full STRING network; the meaning of network edges-evidence; the minimum required interaction score was set to medium confidence parameter (0.4) and max number of interaction display was no more than 10. The PeKCS proteins’ secondary structures were predicted by SOPMA SECONDARY STRUCTURE PREDICTION METHOD¹⁶. Three-dimensional modeling (3D) of PeKCS protein was performed using the Phyre2 online tool¹⁷ with default settings.

To predict putative miRNA target sites in the PeKCS genes, at first, the mature sequences of passion fruit miRNAs were downloaded from published data (Paul et al., 2020). Moreover, the PeKCS CDS sequences were submitted to the online psRNATarget Server¹⁸ with default parameters for prediction of potential miRNAs in PeKCS genes. The interaction network between the predicted miRNAs and PeKCS target genes was constructed and visualized by Cytoscape software version 3.9¹⁹ (Kohl et al., 2011). Furthermore, the GO and KEGG annotation analyses were performed by submitting PeKCS protein sequences to the online database eggNOG²⁰ and their enrichment analysis were performed in TBTools software (Chen C. et al., 2018).

The TFs’ prediction and regulatory network analysis were performed according to Zheng et al. (2020) with slight modifications. The 1000 bp nucleotide sequences from upstream regions of PeKCS genes were extracted and were submitted to the Plant Transcriptional Regulatory Map (PTRM)²¹ with p-value ≤ 1e^–6 (Tian et al., 2020) for the prediction of TFs. Cytoscape version 3.9 software (Kohl et al., 2011) was used to construct and visualized the TF regulatory network.

The passion fruit RNA-Seq raw reads were downloaded from National Center for Biotechnology Information (NCBI) Sequence Read Archive (SRA) database²² with the following accession numbers (SRP150688) and (PRJNA634206). The peel samples were from yellow (P. edulis. Flavicarpa cv Huangjin) and purple (P. edulis. Sims cv Tainong) passion fruit cultivars at the ripening stage, while pulp samples were from both cultivars at different fruit developmental stages. The root samples were from purple passion fruit Pingtan-1 cold-tolerant cultivar under two cultivation areas as limestone (L) and sandy dolomite (D) rocky desertification areas. The leave samples were from yellow Huangjinguo (HJG) cold-sensitive and purple Tainong-1 cold-tolerant cultivars under normal temperature (NT) and chilling stress (CS) conditions. The RNA-Seq raw reads were quality controlled and filtered by fastp package (Chen S. et al., 2018) and mapped to the passion fruit reference genome (Ma et al., 2021) with HISAT2 package (Kim and Langmead, 2015) in Ubuntu wsl v 20.04.3²³. The sequence alignment map (SAM) files were transformed to binary alignment map (BAM) and sorted with Samtools packages (Li et al., 2009). Fragments per kilobase million (FPKM) values were calculated using the limma and edgeR (Law et al., 2016). Due to large differences in FPKM values among different tissues of passion fruit, the FPKM values were transformed to log². Heatmaps were constructed using TB-Tools software (Chen C. et al., 2018) to visualize the expression profiles of PeKCS genes in different tissues.

The KCS genes play an important role in VLCFA biosynthesis and plant resistance to biotic and abiotic stresses. The PeKCS genes expression profiles were studied by applying the drought stress (abiotic) and Fusarium kyushuense fungal pathogen (biotic) stressed conditions compared to controls of yellow and purple passion fruit cultivars. To prepare drought-stressed and normal condition samples for PeKCS genes expression analysis, seeds of two commercial passion fruit cultivars yellow (P. edulis. Flavicarpa cv Huangjin) and purple (P. edulis. Sims cv Tainong) were planted in plastic pots filled with peat moss and soil (2:1 ratio). Greenhouse conditions were set as follows; temperature 25 ± 2°C, photoperiod (16-h), and 75% relative humidity. One-month-old passion fruit plants were subjected to dehydration for 10 days and rewatered to study the qRT-PCR (Quantitative real-time polymerase chain reaction) expression profile of PeKCS selected genes under drought conditions. Leaves, root, and stem samples of both cultivars with three biological replicates were collected and quickly frozen in liquid nitrogen and were stored at −80°C for further uses. The samples from normally watered plants were taken as controls. In order to prepare the fungal stress samples, fruits of both passion fruit cultivars were collected from a commercial orchard located in Fujian province, China (23°48035.200 N and 117°07008.100 E). Fruits were surface sterilized and infected with the F. kyushuense pathogenic fungus following the protocol mentioned in our previous publication (Rizwan et al., 2021). Peels from infected areas were collected after the 9th and 12th days post-inoculation (dpi), while uninfected fruit at 9th and 12th days were used controls.

RNA was extracted from frozen samples using the Tiangen mini-RNA extraction kit (Tiangen, China) according to the manufacturer’s instructions, and RNA was quantified with a Thermo Scientific NanoDrop 2000 UV-Vis Spectrophotometer (Thermo Scientific, United States). One μg of total RNA was used to synthesize the complementary DNA (cDNA) by Takara PrimeScript™ RT Kit with gDNA eraser (TAKARA, China) and cDNAs were diluted to 5x with deionized distilled water. qRT-PCR was performed on a LightCycler^® 96 (Roche Applied Science, Penzberg, Germany) in 20 μL reaction mix consisting of 10 μL TB Green master mix solution (TAKARA), 1 μL of each forward and reverse primer (100 μM), 1 μL of cDNA and 7 μL of ddH₂O. Using Pe60s (Munhoz et al., 2015), As an internal reference gene, the qRT-PCR reactions were performed with the conditions as; preincubation at 95°C for 30 s, followed by 45 cycles at 95°C for 10 s, and 60°C for 30 s. Three biological replicates were used for each reaction and the relative gene expression levels were calculated using the 2^–ΔΔCT method (Schmittgen and Livak, 2008). The primer used in this study has been provided in Supplementary Table 1.

The sub-cellular localization of PeKCS genes was hypothetically predicted by CELLO v.2.5 and for the validation, the transient expression assay was performed in onion epidermal cells by selecting a PeKCS2 candidate gene. To construct a vector for transient expression assays, the pCAMBIA1302 vector with cauliflower mosaic virus 35S (CaMV35S) promoter and green fluorescent protein (GFP) marker in the upstream region of the multiple cloning site (MCS) was used. The PeKCS full-length coding DNA sequences (CDS) without stop codon was amplified by PCR using the following primers: 5′-ACGGGGGACTCTTGACCATGGATGGATAGAGAAA GAGA TCTTTTGTCCACG-3′ (NcoI) and 5′-TCTCCTTTACTAGTC AGATCTCAGAGTCGCA GGAGGATATCTGT-3′ (BglII) (underlined are the restriction sites), and cloned into the pMD 19-T vector (Cat# 6013, TAKARA, Shiga, Japan) following the manufactures instructions. The positive clones were confirmed by sequencing (Sangon Biotech Co., Ltd, Shanghai, China), the gene fragment was digested with NcoI and BglII restriction enzymes and ligated into the NcoI-BglII digested pCAMBIA1302 vector using ClonExpress II One Step Cloning Kit (Cat# C112, Vazyme Biotech Co., Ltd., Nanjing, China) and then transferred to Agrobacterium tumefaciens EHA105 strain for the following infection. The resulting plasmids were named CaMV35S-PeKCS2-GFP and empty vector CaMV35S-GFP as control. These vectors were successfully transformed into onion epidermal cells by agroinfiltration method and GFP expressions were examined by laser scanning confocal microscopy (Olympus, Tokyo, Japan; FV1200) after 24–72 h of agroinfiltration (Xu et al., 2014).

One-way ANOVA was performed for statistical analyses and Figures were generated by GraphPad Prism version 8.0.1²⁴. The comparisons between treated and controlled samples were performed using Student’s t-test and were considered statistically significant if p < 0.05.

In this study, after removing redundant, repetitive, and unrecognized sequences from BLASTp and HMMR results, finally, thirty-two KCS genes in the passion fruit genome by computational tools were identified. Passion fruit KCS genes were named from PeKCS1 to PeKCS32 according to their chromosomal positions. Except for chromosomes numbers five and nine, all PeKCS genes were unevenly distributed on seven of the nine passion fruit chromosomes and PeKCS32 was localized on an unassembled chromosome region. The largest number of PeKCS genes (14 genes) appeared on chromosome number one, followed by chromosome number four (5 genes), chromosome 2 (3 genes), chromosomes number three, seven, and eight (2 genes each), There is only one PeKCS gene on chromosome number six. PeKCS32 was found in the scaffold chromosome region, while no PeKCS gene was found on chromosomes 5 and 9 (Figure 1A and Table 1). Details of all 32 PeKCS genes can be found in Table 1 and the protein sequences have been provided in Supplementary Table 2. CD and SMART search tools for domains verification was used and found that PeKCS proteins contained two domains such as ACP_syn_III_C ((PF08541) 3-Oxoacyl-[acyl-carrier protein (ACP)] synthase III C terminal domain) and FAE1_CUT1_RppA ((PF08392) FAE1/Type III polyketide synthase-like protein domain) (Figure 1B).

Most of the PeKCS proteins contained both proteins 68% (22 genes), 19% (6 genes) contained only FAE1_CUT1_RppA domains and 9% (3 genes) contained only ACP_syn_III_C domain respectively (Figure 1B). Furthermore, the results of physiochemical properties showed that PeKCS genes varied in their properties such as; the protein length varied from 197 bp (PeKCS7) to 1052 bp (PeKCS9), the CDS length ranged from594 bp (PeKCS7) to 3159 bp (PeKCS9), as well as the protein molecular weight (M.W), ranged from 21.96 KDa (PeKCS7) to 118.05 KDa (PeKCS9). The isoelectric point (pI) of PeKCS proteins also varied ranged from 5.89 (PeKCS17) to 9.38 (PeKCS4 and PeKCS5), protein instability index (II) varied from 27.47 (PeKCS25) to 48.7 (PeKCS2) and aliphatic index (Ai) ranged from 81.34 (PeKCS30) to 101.11 (PeKCS11). The grand average of hydropathicity (GRAVY) ranged from −0.38 (PeCS7) to 0.08 (PeKCS1). PeKCS protein also varied in number of exons ranged from 1 (PeKCS2, PeKCS5, PeKCS6, PeKCS10, PeKCS11, PeKCS14, PeKCS15, PeKCS27, PeKCS31, and PeKCS32) to 9 (PeKCS1). The proteins subcellular localization prediction showed that all the PeKCS proteins were associated with the plasma membrane (Table 1).

Multiple sequence alignment and evolutionary tree analyses of identified passion fruit KCS genes were performed. The multiple sequences analysis between AtKCS and PeKCS genes exhibited two conserved domains FAE1_CUT1_RppA (Supplementary Figure 1, presented in red box) domain and the ACP_syn_III_C domain (Supplementary Figure 1, presented in the blue box). Both sequences were highly identical and conserved in both regions, indicating the importance of these domains for KCS gene functions (Supplementary Figure 1). Furthermore, the N-terminal deletion of the FAE1_CUT1_RppA domain in PeKCS7 was precise compared to other PeKCS genes, suggesting that PeKCS7 may have a unique function (Supplementary Figure 1). One hundred and eleven KCS protein sequences from A. thaliana (AtKCS), P. edulis (PeKCS), M. domestica (MdKCS), and A. hypogaea (AhKCS) species were used to assess the evolutionary relationship and unrooted phylogenetic tree (All the KCS protein sequences used in the phylogenetic tree have been provided in Supplementary Table 2. The KCS protein sequences were aligned and a neighbor-joining tree was constructed using MEGA software.

The phylogenetic tree was divided into four main groups (FAE-like, FDH-like, KCS1-like, andCER6-like) and eight subgroups including FAE-like (α), FAE-like (β), CER6-like (γ), KCS1-like (δ), KCS1-like (ζ), FDH-like (ε), FDH-like (η), and FDH-like (θ) (Figure 2) according to AtKCS (Joubès et al., 2008). Phylogenetic analysis showed that FDH-like (θ) subgroup consisted of 35 KCS genes (17 PeKCS, 3 AtKCS, 11 MdKCS, and 4 AhKCS) followed by KCS1-like (ζ) 26 genes (6 PeKCS, 3 AtKCS, 5 MdKCS, and 12 AhKCS), FAE-like (α) consists of 15 genes (4 PeKCS, 3 AtKCS, 4 MdKCS, and 4 AhKCS), KCS1-like (δ) consists of 10 genes (1 PeKCS, 3 AtKCS, 2 MdKCS, and 4 AhKCS), FDH-like (ε) consists of 9 genes (1 PeKCS, 2 AtKCS, 2 MdKCS, and 4 AhKCS), FDH-like (η) consists of 6 genes (1 PeKCS, 2 AtKCS, and 3 MdKCS), CER6-like (γ) consists of 6 genes (2 PeKCS, 2 AtKCS, and 2 MdKCS), and FAE-like (β) composed of only 3 Arabidopsis KCS genes (Figure 2). FDH-like (θ) was found to be the largest subgroup consisting of 35 KCS genes followed by KCS1-like (ζ) 26 genes and FAE-like (β) was the smallest subgroup containing only 3 Arabidopsis KCS genes. The number of KCS genes was similar in FDH-like (η) and CER6-like (γ). There were also differences in the uneven distribution of PeKCS genes between different subgroups, for example, the largest number of PeKCS genes (17 genes) was found in the FDH-like (θ) subgroup, followed by the KCS1-like (ζ) subgroup (6 genes), and only 1 PeKCS gene was found in the KCS1-like (δ), FDH-like (n), FDH-like (ε) subgroups and no any PeKCS gene was aligned in the FAE-like (β) subgroup (Figure 2). The results revealed that P. edulis KCS proteins shared greater homology with M. domestica (MdKCS) and A. thaliana (AtKCS) compared to A. hypogaea (AhKCS) (Figure 2). Furthermore, the pair-wise identity between PeKCS genes was performed to better understand the evolution and the result showed that the similarity varied from 22.71% (PeKCS1/PeKCS13) to 99.84% (PeKCS16/PeKCS17 and PeKCS17/PeKCS19) respectively (Supplementary Table 3). PeKCS17 showed the highest similarity of 99.84% with PeKCS16 and PeKCS19 genes. Pairwise similarity results were contrasted with phylogenetic tree results, where PeKCS16, PeKCS17, and PeKCS19 genes were grouped in the same FDH-like (θ) subgroup (Figure 2 and Supplementary Table 3).

The relationship between 32 PeKCS genes was investigated through the construction of a phylogenetic tree using the neighbor-joining (NJ) method and grouped from I-V11 (Figure 3A). The protein sequences of all PeKCS genes have been provided in Supplementary Table 2. According to the results, group VII was the largest group with 17 PeKCS genes followed by group 4 with 6 genes. The smallest groups were III, V, and VI with only one PeKCS gene (Figure 3A). To further understand the development and functions of PeKCS family genes, the conserved motifs and exon-intron configuration of PeKCS gene structure were analyzed (Figures 3B,C). Moreover, the conserved motifs among 32 PeKCS genes were predicted using the online MEME tool. The conserved motifs among PeKCS proteins vary ranged from 1 (PeKCS28) to 16 (PeKCS9) and overall, 10 conserved motifs were recognized in all PeKCS proteins (Supplementary Table 4). Most of the PeKCS genes (12 genes) have 8 motifs followed by 4 motifs (9 genes) and PeKCS28 have only one motif. PeKCS9 contained a maximum of 16 motifs. The motif distribution among group I-VII was also similar such as groups I, II, III, and V contained 8 conserved motifs except for PeKCS30. Furthermore, the 9 members of group VII contained 4 conserved motifs (Figures 3A,B). Genomic structural analysis of exons in 32 PeKCS genes revealed that the number of exons identified in PeKCS varied from 1 to 9 (Figure 3C). Most PeKCS genes (11 genes) consist of 2 exons and 1 exon (10 genes). In addition, PeKCS1 contains up to 9 exons (Figure 3C). Group II has only 1 exon and no intron, group 1, III, IV, and V have 1 to 4 introns, group VI have 3 introns and group VII contained 1 to 9 introns except PeKCS11, PeKCS14, PeKCS15, and PeKCS32 genes have no introns (Figure 3C). Overall, groups I, II, III, and V showed similar exon-intron patterns whereas groups IV and VII had different exon-intron association patterns (Figure 2C). These results suggest that PeKCS genes within a group have remarkably similar gene structures consistent with their phylogenetic relationships (Figure 3). In conclusion, after analyzing the composition of conserved motifs, gene structure, phylogenetic relationships, and group classification, the results indicated that PeKCS genes have highly conserved amino acids and genes within a group may have the same functions.

The cis-regulatory element analysis of the PeKCS genes from the 2000 bp upstream promoter region was conducted to better understand the role of PeKCS genes in biotic and abiotic responses (Figure 4 and Supplementary Table 5). Mainly four different categories of cis-regulatory elements were found in the PeKCS promoter regions including plant growth and development (8 different types of cis-elements), phytohormones (10 different types of cis-elements), light (15 different types of cis-elements), and stress (7 different types of cis-elements) responsiveness (Figure 4.1). Overall, a total of 969 cis-elements belonging to different categories were identified in 32 PeKCS genes, and when comparing between four different categories, the phytohormone-responsive category had the highest number of cis-elements at 40% (391/969), followed by light-responsive 37% (351/969), stress-responsive 15% (147/969) and the lowest number of cis-elements found in the plant growth and development response category 8.3% (80/969) (Figure 4.2). Furthermore, all the 32 PeKCS genes contained the phytohormone, light, and stress-responsive cis-elements, whereas only 25 PeKCS contained metabolism (plant growth and development) responsive cis-elements (Figure 4.2A).

The main four categories were further divided into subcategories based on different cis-elements. Light responsiveness category comprised of the 3-AF1 binding site, ACE, AE-box, AT1-motif, ATC-motif, ATCT-motif, TCT-motif, Box 4, chs-CMA1a/2a/2b/2c, GA-motif, GATA-motif, G-Box, GT1-motif, I-box, LAMP-element, MRE, Sp1, and TCCC-motif cis-elements (Figure 4.2 and Supplementary Table 5). The highest light-responsive cis-elements consisted of Box-4 22% followed by G-Box 21%, while the GA-motif consisted of only 1% cis-elements (Figure 4.2B). The phytohormone responsive category further includes ABRE (abscisic acid-responsive), AuxRR-core (auxin-responsive), ERE (ethylene-responsive), GARE-motif, P-box, TATC-box (gibberellin responsive); TCA-element (salicylic acid-responsive); TGA-element and TGACG-motif (MeJA responsive) cis-elements. The highest number of cis-elements from the phytohormones category was ABRE 31% followed by CGTCA-motif 18%, while the AuxRR-core consisted of only 1% cis-elements (Figure 4.2C and Supplementary Table 5).

The stress response category further included ARE, GC-motif, LTR, MBS, TC-rich repeats, CARE, and WUN-motif cis-element. ARE accounted for the highest number 35% followed by MBS with 19%, and CARE with only 3% (Figure 4.2D and Supplementary Table 5). The plant growth and development category further include subcategories of meristem, metabolism, seed related, and circadian. Meristem associated cis-elements included CAT-box (10%) and GCN4-motifs (10%). Metabolism-related cis-elements included O2-sit (25%), seed and circadian-related cis-element included RY-element (11%), and only 3% of MSA-like cis-elements were found (Figure 4.2E and Supplementary Table 5). The information on developmental, light, stress, and hormone-related cis-elements suggested that the transcriptional profiling of PeKCS genes may vary in developmental, hormonal, and stressful contexts and require further investigation. Details on the cis-regulatory elements in the passion fruit KCS genes have been provided in Supplementary Table 5.

Synteny analysis was performed to further understand the evolution and expansion mechanism of the PeKCS gene family in the passion fruit genome and the genomes of other species. PeKCS gene duplications were assessed based on a tandem or segmental duplications. The results of gene duplication analysis indicate that there were 8 PeKCS gene pairs (Figure 5A and Supplementary Table 6), of which 3 gene pairs were segmentally duplicated on chromosomes 1, 2, 3, 6, and 8, whereas 5 gene pairs were tandem duplicated on chromosomes 1 and 4 in passion fruit genome. The duplicated genes belonged to different chromosomes, and chromosome 1 was found to have half of the duplicated genes (8 genes) (Figure 5A and Supplementary Table 6). These results suggest that gene duplication may play an important role in the development of the PeKCS gene family and the passion fruit genome. In addition, the Ka/Ks ratios were calculated in duplicated genes to assess the evolutionary rates and selection pressures (Supplementary Table 6). In general, a Ka/Ks ratio greater than 1 indicates that the gene was positively selected, a ratio less than 1 indicates a purifying selection, and a ratio equal to 1 indicates neutral selection. Details of the Ka, Ks values, and Ka/Ks ratios of the duplicated PeKCS gene pairs are provided in Supplementary Table 6. Overall, all duplicated PeKCS gene pairs showed a Ka/Ks ratio of less than 1, indicating that these genes had gone through purifying selection. Additionally, the divergence time between duplicated genes was measured as a substitution rate of 6.38 × 10^–9 substitutions per site per year (Ma et al., 2021). The results of the divergence indicated that the duplication process between the tandem and segmental PeKCS genes was estimated to be between 0.30 and 27.64 million years ago (Supplementary Table 6). It can be concluded that the evolutionary mechanism of the PeKCS gene showed maintenance during the domestication of passion fruit.

Comprehensive synteny analyses of KCS genes between P. edulis, A. thaliana, and M. domestica species were performed and 45 KCS gene pairs were identified within these three species. Among the 32 P. edulis KCS genes, 15 PeKCS genes were paired with 26 KCS genes in A. thaliana (10 AtKCS) and M. domestica (16 MdKCS) (Figure 5B and Supplementary Table 7). Among species, 15 KCS gene pairs were found between P. edulis (8 PeKCS) and A. thaliana (10 AtKCS) species (Supplementary Table 7). On the other hand, 30 KCS gene pairs were found between P. edulis (15 PeKCS) and M. domestica (16 MdKCS) species (Supplementary Table 7 and Figure 5B). Taken together, the P. edulis and M. domestica KCS genes showed a high degree of synteny compared with A. thaliana, suggesting that they may have originated from the same ancestors and performed the same functions, which required further study.

Multicollinearity analysis was performed to reveal robust orthologs of KCS genes between P. edulis A. thaliana, M. domestica, and A. hypogaea species (Figure 5C and Supplementary Table 8). The results showed that multiple collinear gene pairs between the four species were inferred as genetic copies with lineage-specific expansion. In general, 58 P. edulis KCS genes showed syntenic association with 31 M. domestica genes, 16 A. thaliana genes, and 11 A. hypogaea genes. Overall, the maximum collinear genes were found between P. edulis and M. domestica followed by P. edulis and A. thaliana, while the fewest were found between P. edulis and A. hypogaea species. P. edulis chromosome 1 shared the maximum collinear genes among M. domestica and A. thaliana (Figure 5C and Supplementary Table 8), suggesting that the KCS genes are conserved and may have the same ancestors besides with duplication or loss of KCS genes. Furthermore, many homologs from A. thaliana, M. domestica, and A. hypogaea species maintained a syntenic association with the P. edulis KCS gene family suggesting that, in addition to segmental duplications, the whole-genome duplications also play an important role in the evolution of the PeKCS gene family.

The protein–protein interaction network analysis of PeKCS proteins based on known Arabidopsis proteins was conducted. PeKCS portions having higher homologous similarity with Arabidopsis proteins were selected as STRING proteins. Among all, 31 PeKCS proteins were associated with known Arabidopsis proteins (Figure 6A and Supplementary Table 9). PeKCS proteins belonging to different groups may have diverse functions in Arabidopsis. PeKCS1 and PeKCS28 were homologous with AtKASIII protein and have a strong interaction among AtKASI, AtEMB3147, AT2G04540, and AtBCCP2 proteins. PeKCS2 showed homology with AtKCS1 protein and interacts with AtFAB1/KASI and AtKCS6 proteins. PeKCS5, PeKCS11, and PeKCS12 were homologous with AtKCS6 and showed an interaction with AtKCS6 and AtKCS10 proteins. PeKCS23 shows homology with AtKCS10 and interacts with AtBCCP2 and AtKCS6 proteins. PeKCS8 was homologous with AtKCS12 and interacts with AtBCCP2 protein. PeKCS3 was homologous with AtLAP6 and interacts with AtFLS1 and AtF3H proteins. PeKCS9, PeKCS20, PeKCS21, and PeKCS22 were homologous with AtKCS11 and have a strong interaction with AtKASI protein. PeKCS16, PeKCS17, PeKCS18, PeKCS19, PeKCS24, PeKCS25, PeKCS26, and PeKCS29 were homologous with AtTTP4 and have an interaction among AtFLS1 and AtTT5 proteins. The PeKCS proteins that have strong interaction with Arabidopsis proteins might have similar functions as in Arabidopsis. The higher the interaction coefficient, the thicker the line between proteins and vice versa (Figure 6A and Supplementary Table 9).

The protein secondary structure generally consists of alpha helices, extended strands, beta turns, and random coils. The PeKCS protein secondary structure analysis was performed and found that in all PeKCS proteins, the alpha helix accounted for the largest percentage of secondary structures 49.6% (PeKCS5) followed by random coils 41.96% (PeKCS1) and extended strands 22.89% (PeKCS28), whereas the beta turns accounts only for 9.2% (PeKCS28). PeKCS28 showed the highest percentage in extended strands and beta turns might have a specific function that needs further studies (Supplementary Table 10). Additionally, the 3D models of all the PeKCS proteins were predicted by using the Phyre2 server with default mode (Figure 6B). The following templates were used for the predicted PeKCS protein models including nineteen PeKCS (PeKCS2, PeKCS4, PeKCS5, PeKCS6, PeKCS7, PeKCS8, PeKCS9, PeKCS10, PeKCS12, PeKCS13, PeKCS14, PeKCS15, PeKCS20, PeKCS21, PeKCS22, PeKCS23, PeKCS27, PeKCS30, and PeKCS31) protein models were predicted with c3vs9F template. Eight PeKCS (PeKCS16, PeKCS17, PeKCS18, PeKCS19, PeKCS24, PeKCS25, PeKCS26, and PeKCS29) protein models were predicted with c1cmlA template. Template c1u0mA in PeKCS32 model, template c2d3mA in PeKCS3 model, template c3gwaA in PeKCS28 model, template c3tsyA in PeKCS11 model and template c4b3yB in PeKCS1 respectively. All PeKCS proteins showed a flexible structure due to the presence of coils (Figure 6B). Our finding suggested that PeKCS genes from individual genomes may be ancestrally similar with each other or initially differ maybe stabilized during long-term domestication, leading to changes in protein structure and functions.

MicroRNAs are a class of small non-coding regulatory RNAs that control gene expression by directing target mRNA cleavage or translational repression (Sunkar and Zhu, 2004). Over the past few decades, various studies have been reported that miRNAs induce the regulation of stress, plant development, and single transduction (Witkos et al., 2011). Therefore, to better understand the regulatory mechanism of miRNA involved in the regulation of PeKCS genes, the identification of putative miRNA in PeKCS genes was performed by using the mature sequences of known passion fruit miRNAs as described in the material section. Details of the putative miRNA targeting sites and the PeKCS genes have been provided in Supplementary Table 11. Twenty putative ped-miRNAs belonging to 9 different families were identified, targeting the 13 PeKCS genes (Figure 7 and Supplementary Table 11). The results show that one member of the ped-miR157 family targeted one gene (PeKCS5); one member of the ped-miR162 family targeted one gene (PeKCS9); one member of ped-miR164 family targeted one gene (PeKCS5); two members of the ped-miR166 family targeted four genes (PeKCS2, PeKCS24, PeKCS25, and PeKCS26); one member of ped-miR171 family targeted one gene (PeKCS4); two members of ped-miR319 family targeted two genes (PeKCS5 and PeKCS10); one member of ped-miR397 family targeted one gene (PeKCS8); one member of ped-miR398 family targeted one gene (PeKCS23) and two members of ped-miR399 family targeted three genes (PeKCS12, PeKCS26, and PeKCS27) (Figure 7 and Supplementary Table 11). Taken together, the results indicated that PeKCS5 was most targeted by four putative ped-miRNAs (ped-miR157a-5p, ped-miR164b-5, ped-miR319b, and ped-miR319l) belonging to four distant families, whereas the ped-miR166 family that targeted up to four PeKCS genes (PeKCS2, PeKCS24, PeKCS5, and PeKCS26) (Figure 7A and Supplementary Table 11). Network and schematic diagram of putative miRNA targeting sites of PeKCS genes shown in Figures 7A,B.

The potential TFs in the upstream region of (1000bp) of PeKCS genes were predicted and constructed the TF regulatory networks (Figure 8). The results showed that, overall, 1701 TFs were identified in 32 PeKCS genes belonging to 38 different TF families including ERF, AP2, GRAS, ARF, MYB, Dof, C2H2, TCP, LBD, bHLH, MIKC_MADS, BBR-BPC, and NAC (Figure 8 and Supplementary Table 12). The most abundant TF families were ERF (719 members), MYB (138 members), Dof (102 members), C2H2 (91 members), TCP (75 members), LBD (64 members), and bHLH (61 members) (Supplementary Table 12). Whereas the least abundant TF families were also identified contained only 1 member including WOX, EIL SRS, HSF, BES1, and S1Fa-like (Supplementary Table 12). All 32 PeKCS genes were predicted to be targeted by numerous TFs belonging to different families, for example, PeKCS32 was the most targeted with 547 TFs followed by PeKCS30 with 113, PeKCS24 with 76, PeKCS2 with 72, and PeKCS27 with 64 TFs. Whereas PeKCS10 and PeKCS22 were minimally targeted genes with only 10 and 9 TFs respectively (Figure 8 and Supplementary Table 12).

Furthermore, a regulatory network of TFs for the top five PeKCS genes was constructed based on their degree of interaction, and the results showed that PeKCS32 was highly targeted, followed by PeKCS30, PeKCS24, PeKCS2, and PeKCS27 (Figure 8B and Supplementary Table 12). PeKCS genes were targeted with numerous members of different TF families such as PeKCS32 was enriched in ERF (412), LBD (40), and MYB (27) family members, PeKCS30 was enriched in ERF (38), TCP (20), and NAC (16) family members. PeKCS2 has enriched ERF (17), MYB (14), and MIKC_MADS (13) family members (Figure 8B and Supplementary Table 12). Overall, the ERF family was found to be dominant in all TF families. The TFs regulatory network of all the 32 PeKCS genes are shown in Figure 8A, and the networks of the top five highly enriched PeKCS genes are shown in Figure 8B, respectively and more details can be found in Supplementary Table 12. The different fatty acids and defense-related TFs belonging to different families were identified in the present study including ERF, AP2, bHLH, WRKY, and MYB. TFs involved in plant growth and development including TCP, bHLH, BBR-BPC, WRKY, LBD, and AP2 were also found in PeKCS genes. In addition, phytohormone-related TFs were also identified, including ERF and ARF. Interestingly, ERF, MYB, and AP2 TFs were shown to be universally distributed in most of the PeKCS genes (Figure 8 and Supplementary Table 12).

Gene ontology and KEGG annotation analysis of the PeKCS gene was performed to further understand the possible roles of PeKCS genes in molecular function (MF), cellular component (CC), and biological process (BP) at the molecular levels. The details of annotation results and numerous significantly enriched terms for MF, CC, and BP have been provided in Supplementary Tables 13a,b. Seventeen terms were identified in 32 PeKCS genes belonging to GO-MF class, among which four terms were highly enriched including fatty acid synthase activity (GO:0004312), fatty acid elongase activity (GO:0009922), and acyltransferase activity (GO:0016746, GO:001674747) (Figure 9A and Supplementary Table 13). The GO enrichment analysis showed that GO-CC class exhibited thirty-one terms in 32 PeKCS genes, and seven terms were highly enriched including endoplasmic reticulum (GO:0005783), endomembrane system (GO:0012505), obsolete cytoplasmic part (GO:0044444), membrane-bounded organelle (GO:0043227), and organelle membrane (GO:0031090) (Figure 9A and Supplementary Table 13).

For instance, the maximum number of sixty-seven terms were identified in GO-BP class and six terms were highly enriched including response to cold (GO:0009409), response to temperature stimulus (GO:0009266), response to light stimulus (GO:0009416), response to radiation (GO:0009314), response to abiotic stimulus (GO:0009628) and response to stress (GO:0006950) (Figure 9A and Supplementary Table 13). Taken together, the enrichment and annotation results revelated that biological class terms were highly enriched followed by the cellular class in PeKCS genes. In addition, KEGG pathway enrichment analysis revealed that 15 pathways were predicted to be involved in different functions among the 32 PeKCS genes. Among the predicted KEGG pathways, the highly enriched pathways are presented in Figure 9B, including metabolism (A09100), lipid metabolism (B09103), environmental adaptation (B09159), organismal systems (A09150), polyketide biosynthesis proteins (01008), plant-pathogen interaction (04626), fatty acid elongation (00062) and circadian rhythm (04712) respectively (Figure 9B and Supplementary Table 4). In conclusion, the GO and KEGG enrichment analysis suggests that PeKCS genes may play important roles in different biological, molecular, and cellular processes including metabolism, fatty acid biosynthesis, and responses to different biotic and abiotic stresses.

The subcellular localization of all 32 PeKCS genes was hypothetically predicted using the online CELLO version 2.5 program. All 32 PeKCS genes were expected to localize to the plasma membrane (Table 1). To validate the hypothesized predicted results, PeKCS2 was used to perform a transient expression assay in onion epidermal cells. The CaMV35S-PeKCS2-GFP fusion construct was transformed into onion epidermal cells by the agroinfiltration method for transient expression assays (Xu et al., 2014). GFP expression was observed and results showed that GFP signals were highly expressed in the plasma membrane (Figures 10d–f). However, the empty vector CaMV35S-GFP was used as a control and the results showed that the GFP was dispersed throughout the cell (Figures 10a–c). The PeKCS2 subcellular localization results were consistent with the hypothesized predictions.

The expression profiles of all 32 PeKCS genes in the pulp tissue of yellow and purple cultivars at different fruit development stages (fruitlet stage, green stage, veraison stage, and ripening stage) based on FPKM values were evaluated. The FPKM expression values have been provided in Supplementary Table 15. The FPKM values were transformed to log² and a circular heatmap was generated by Tbtools software (Figure 11A and Supplementary Table 15). The expression profiles of PeKCS genes varied at different fruit developmental stages among both cultivars. Overall, 30 of the 32 PeKCS genes were expressed during different fruit developmental stages in both cultivars except PeKCS31 and PeKCS32. In the yellow cultivar, out of the 30 PeKCS expressed genes, 28 (87%), 27 (84%), 24 (75%), and 22 (68%) genes were expressed at fruitlet, green, veraison, and ripening stages. In terms of expression patterns, 14 (50%) genes were highly expressed (FPKM > 10) at the fruitlet stage, 5 (18%) genes at the green stage, 8 (20%) genes at the veraison stage, and 10 (18%) genes at ripening stage. In the comparison of different developmental stages and genes, the expression pattern of PeKCS30 (FPKM = 356.7) was highest at the green stage, followed by PeKCS10 (FPKM = 324.52) at the fruitlet stage, PeKCS30 (FPKM = 233.51) at veraison, and PeKCS30 (FPKM = 96.12) at the ripening stage. Only one PeKCS13 gene expression was constantly increased during fruitlet (FPKM = 4.71) to ripening stage (FPKM = 23.68), while the expression patterns of five genes (PeKCS17, PeKCS19, PeKCS21, PeKCS22, and PeKCS28) decreased during fruitlet (FPKM = 69.22, 24.43, 13.16, 14.11, and 20.59) to ripening stages (FPKM = 2.77, 0.6, 3.72, 3.6, and 3.68) (Figure 11A and Supplementary Table 15). Interestingly, three genes had a constant high expression pattern in the four fruit development stages of yellow cultivar, including PeKCS1 (FPKM = 17–55.62), PeKCS10 (FPKM = 13.73–324.52), and PeKCS30 (FPKM = 96.12–356.70) (Figure 11A and Supplementary Table 15).

In the purple cultivar, overall, 30 (93%) PeKCS genes were expressed at the fruitlet stage, 25 (78%) at the green stage, 27 (84%) at veraison and ripening stages. Among them, 8 (26%) genes were highly expressed (FPKM > 10) at fruitlet, 5 (20%) at the green, 9 (33%) at veraison, and 7 (26%) at ripening stages. PeKCS10 exhibited the highest expression pattern (FPKM = 231.18) at the fruitlet stage, followed by green stage (FPKM = 130.89), PeKCS29 (FPKM = 74.68) at the veraison stage, and PeKCS1 (FPKM = 43.24) at the ripening stage. The expression patterns of 6 genes (PeKCS13, PeKCS16, PeKCS17, PeKCS18, PeKCS19, and PeKCS19) showed an increased trend from small fruitlet (FPKM = 0.1–1.1) to ripening stage (FPKM = 7.0–74.68), while 3 genes (PeKCS10, PeKCS27, and PeKCS28) decreased expression pattern from fruitlet (FPKM = 23.07–231.18) to ripening stage (FPKM = 0.11–2.3). Remarkably, PeKCS1 had a constant high expression pattern in the four fruit development stages of the purple cultivar (FPKM = 25.96–42.53) (Figure 11A and Supplementary Table 15).

In the comparison of PeKCS genes between the tested stages, PeKCS1 and PeKCS30 showed similar higher expressions (FPKM > 10) (Figure 11A and Supplementary Table 15). In the yellow cultivar, the green stage exhibits to have the highest expressions (FPKM = 324.52), whereas in purple the fruitlet stage showed the highest expressions (FPKM = 231.18). Overall, the yellow cultivar had higher expression levels and maximum genes with higher expression patterns compared to the purple cultivar. In addition, the yellow cultivar had decreased an expression pattern (5 genes) from fruitlet to ripening, while purple had increased expression patterns (6 genes) from fruitlet to ripening stages. These findings suggest that PeKCS genes may have significant roles in fruit development and ripening stages in both cultivars and further research is still needed (Figure 11A and Supplementary Table 15).

The expression profiles of PeKCS genes in the root, stem, and peel tissues of yellow and purple passion fruit cultivars under different conditions were evaluated in FPKM values, and have been provided in Supplementary Table 16. The expression profiles of PeKCS genes in root tissue of the purple, leaf, and peel tissues of yellow cultivars were evaluated in FPKM values and visualized in circular heatmaps by converting to log2 values (Figure 11B and Supplementary Table 16). The expression profiles of PeKCS genes varied in different tissues among both cultivars. Overall, out of 32 PeKCS genes, 19 (59%) PeKCS genes were expressed in purple L and D root tissues, and 15 (46%) PeKCS genes were expressed in yellow and purple peel tissues. In contrast, 23 (72%) PeKCS genes were expressed in purple leaf NT condition, 25 (78%) in yellow leaf NT, 22 (69%) in purple leaf CS, and 24 (75%) genes were expressed in yellow leaf CS conditions respectively (Figure 11B and Supplementary Table 16). Among the expressed PeKCS genes in roots, 6 (31%) and 5 (26%) genes were highly expressed (FPKM > 10) in L and D roots. Roots under the L condition had the highest expression (PeKCS20, FPKM = 213.63) compared to roots in the L condition (FPKM = 82.58, PeKCS30) (Figure 11B and Supplementary Table 16).

Among the genes expressed in the peel, 8 (53%) and 3 (20%) PeKCS genes showed high expressions (FPKM > 10) in purple and yellow peels. Whereas, the purple peel was highly expressed (PeKCS29, FPKM = 1692.94) compared to yellow peel (PeKCS30, FPKM = 41.97). Under two different conditions (NT and CS), the expression profiles of PeKCS genes varied in leaf tissue of both cultivars such as 11 (48%) genes were highly expressed (FPKM > 10) in purple leaf-NT condition, 9 (36%) genes in yellow leaf-NT, 10 (45%) genes in purple leaf-CS and 9 (37%) genes in yellow leaf-CS conditions. Comparing both cultivars under NT and CS conditions, purple leaves under CS had the highest expression (FPKM = 359.34, PeKCS5), followed by yellow CS leaves (FPKM = 306.76, PeKCS5), purple NT leaves (FPKM = 303.78, PeKCS5), and yellow NT leaves (FPKM = 226.04, PeKCS5) (Figure 11B and Supplementary Table 16). Taken together, purple had the higher expression in all the tested tissues, whereas, in comparison between tissues, the peel was found to have the highest expression (FPKM = 1692.94, PeKCS29) followed by leave tissues (FPKM = 359.34, PeKCS5) and root (FPKM = 213.63, PeKCS20) (Figure 11B and Supplementary Table 16). Five genes (PeKCS16, PeKCS17, PeKCS18, PeKCS19, and PeKCS32) were not expressed in all tissues (leave, peel and root) indicating that they might not involve in passion fruit growth and development. These results indicate that PeKCS genes have specific expression patterns in all tested tissues and genes with the highest expressions (PeKCS5, PeKCS29, and PeKCS30) might play an important role in specific functions, but further research is required. These genotype-based tissue expression patterns provide ideas for further study of the PeKCS gene family in passion fruit.

The functional characterization of KCS genes has not been studied in passion fruit. Gene expression profiles can provide a reflection of gene function. In the current study, the expression profiles of ten candidate PeKCS genes were evaluated by qRT-PCR under drought stress conditions (Figure 12). In general, the expression profiles of all candidate PeKCS genes showed different expression levels in the stem, leave and root tissues of yellow and purple cultivars under drought conditions. Overall, the results showed that the expression levels of most genes were increased under drought conditions in leave and root tissues compared to controls. The relative expression level of the PeKCS20 gene was highest with a more than 29-fold increase in yellow roots under drought conditions (Figure 12G). However, under drought conditions, the relative expression levels of PeKCS13 and PeKCS27 in yellow and purple leaves increased by more than 10 and 8-folds, respectively (Figures 12F,H). Whereas, relative expressions of PeKCS1 decreased in most tissues under drought conditions (Figure 12A). On the contrary, the relative expression levels of PeKCS4 and PeKCS8 increased by more than 4 and 3-folds in yellow leaves, and all other tissues decreased. Interestingly, PeKCS2 showed an increased relative expression in yellow and purple leaves more than 2.5 folds under drought (Figure 12B). Except for PeKCS9 and PeKCS20, the relative expression levels of most genes decreased in the stem tissue of the yellow cultivar. The relative expression of PeKCS28 was increased by more than 4 and 3-folds in yellow leaves and purple roots under drought conditions (Figure 12I). All the yellow tissues showed decreased relative expression levels in PeKCS29, while increased in purple tissues under drought conditions (Figure 12J). The leave tissue showed the highest expression consistent with the FPKM expression values (Figure 12 and Supplementary Table 16). Taken together, the results indicated that PeKCS genes responded to external drought stress conditions by increasing or decreasing their expression levels in different tissues of the two cultivars, which provided evidence for further functional studies.

The functional characterization of KCS genes against fungal pathogens has not been studied in passion fruit. To gain insight into the potential functional role of PeKCS genes at expression levels in passion fruit defense against fungal pathogens as biotic stress, in the current study, the expression profiles of ten candidate PeKCS genes were assessed by qRT-PCR under F. kyushuense fungal stress (biotic stress) condition in peel tissues of yellow and purple cultivars at 9th and 12th days post-inoculation (dpi) (Figure 13). Overall, all PeKCS genes were expressed and showed increased or decreased expression levels under biotic stress conditions in yellow and purple cultivars compared to controls (Figure 13).

All the tested genes showed higher expression levels under biotic stress in purple 9th dpi (P-9.dpi) except PeKCS1 (Figure 13A) and PeKCS28 (Figure 13I) compared to yellow 9th dpi (Y-9.dpi) (Figure 13). Among them, PeKCS20 (Figure 13G) showed an increased expression level by more than 9-fold, followed by PeKCS2 more than 6-fold, PeKCS9, PeKCS27, and PeKCS29 more than 4-fold, PeKCS8 more than 2.5-fold, and PeKCS4 and PeKCS13 more than 1-fold at 9th dpi in purple cultivar (Figure 13). In yellow cultivar at 9th dpi, Only the expression levels of PeKCS8, PeKCS2, and PeKCS20 were found to be increased by 3-fold and more than 1-fold, while the expression level of all other genes decreased at Y-9.dpi (Figure 13). Interestingly, PeKCS1 and PeKCS28 expression levels were increased by 1.7- and 2.1-folds at 12th dpi in yellow cultivar (Figure 13). Overall, the results showed that PeKCS genes expression was increased under biotic stress conditions in purple at 9th dpi compared to controls (Figure 13). Taken together, the results showed that PeKCS genes have responded to fungal biotic stress by increasing or decreasing their expression levels at different time points of two cultivars providing evidence for further functional studies.

According to RNA-Seq FPKM expression data (Supplementary Tables 15, 16), the PeKCS1, PeKCS2, PeKCS4, PeKCS9, PeKCS10, PeKCS13, PeKCS20, PeKCS27, PeKCS28, and PeKCS29 have relative higher FPKM expressions in purple peels compared to yellow. The FPKM expression values were further validated by qRT-PCR analysis using PeKCS1, PeKCS2, PeKCS4, PeKCS8, PeKCS9, PeKCS13, PeKCS20, PeKCS27, PeKCS28, and PeKCS29 genes in yellow and purple passion fruit peel tissues (The primer details can be found in Supplementary Table 1). After normalization with Pe60s reference genes, all the tested PeKCS genes showed a trend line consistent with the RNA-Seq expression values (Figure 14). These results revealed that RNA-Seq expression values provided an appropriate expression result for all the tested tissues among both passion fruit cultivars.