Amino acid sequences of the model plant A. thaliana P5CS were obtained from the Arabidopsis Information Resource database (https://www.arabidopsis.org/). Using these sequences as queries, the pear genome database was screened by BLASTp (E-value <1e^-5). The complete genome assembly of pear (Pyrus betulifolia Bunge.) and the complete proteome sequence file were obtained from the Genome Database for Rosaceae (https://www.rosaceae.org/). Putative P5CS genes were confirmed by BLASTp searches of the National Center for Biotechnology Information database (https://www.ncbi.nlm.nih.gov/). Other information obtained from this database included chromosome number, gene accession numbers, predicted masses of proteins encoded by P5CS genes, and genomic information. The isoelectric point, grand average of hydropathicity (GRAVY), and molecular weight of P5CS proteins were calculated via the ExPasy website (Duvaud et al., 2021). Subcellular locations of pear P5CS proteins were predicted using WoLF PSORT II (Horton et al., 2007).

P5CS amino acid sequences from apple (Malus domestica), peach (Prunus persica), poplar (Populus trichocarpa), and A. thaliana were downloaded from Ensembl (https://plants.ensembl.org/index.html). ClustalW v1.83 (Hung et al., 2015) was used for multiple sequence alignments of P5CS proteins. The Muscle module within MEGA 7.0 (Kumar et al., 2016) was used to align the sequences of full-length proteins. Construction of phylogenetic trees based on PbP5CS protein sequences was performed using the neighbor-joining approach with Poisson model, pairwise deletion, and 1000 bootstrap replicates.

Conserved motifs of all P5CS proteins were identified using the online MEME analysis tool (Bailey et al., 2015) with the maximum number of motifs set at 10, and default values for all other parameters. The NCBI CDD database (https://www.ncbi.nlm.nih.gov/Structure/bwrpsb/bwrpsb.cgi/) was used to analyze the conserved domains of 11 PbP5CS protein sequences, and P5CS genes’ domain information data were retained. The information of exon (coding sequence), intron, and untranslated region for 11 PbP5CS genes was determined according to the alignments of their sequences in the P. betulifolia genome database (https://www.rosaceae.org/species/pyrus_betulifolia/genome_v1.0/). To compare conserved motifs, conserved domains, and gene structures of different groups, TBtools software (Chen et al., 2020) was used for clustering, drawing phylogenetic trees, and mapping conserved motifs of PbP5CS. Exon-intron structures were visualized using Gene Structure Display Server 2.0 (Hu et al., 2014).

The chromosomal localization of each PbP5CS gene was determined based on physical location information obtained from the pear genome database (https://www.rosaceae.org/species/pyrus_betulifolia/genome_v1.0/). Then, a gene localization and distribution map was drawn using TBtools. Pear genome proteins data were self-compared by BLASTp, and fragment replication type and tandem repeat replication type of PbP5CSs were analyzed by MCScanX (Wang et al., 2012). Tandem duplicated genes were identified by analyzing physical locations on specific chromosomes. MCScanX was used to assess syntenic blocks for PbP5CS genes, as well as those between pear and A. thaliana, between pear and apple, between pear and peach, and between pear and poplar

The promoter sequences of 2000 bp regions upstream of each PbP5CS gene-coding region were retrieved from the pear genome database. PlantCARE (Lescot et al., 2002) was then used to annotate elements, and elements related to hormones, stress, growth, and development were selected for location distribution mapping.

Expression patterns of PbP5CS genes in various tissues were acquired from RNA sequencing (RNA-seq) data available in the NCBI database (https://www.ncbi.nlm.nih.gov/sra/?term=SRP230672/; SRA accession no.: SRP230672). Fragments per kilobase of exon per million mapped reads (FPKM) values were used to estimate gene expression levels. Multi Experiment Viewer (Saeed et al., 2006) was used to evaluate and graphically characterize means of expression values for each gene in all tissues. A heatmap of PbP5CS genes was generated using the OmicShare Tool (https://www.omicshare.com/tools/).

Pear seeds after vernalization were sown in nutritive soil (65% fertile garden soil, 25% burning soil, 10% fine sand, 0.4% calcium-magnesium-phosphate fertilizer). All materials were kept in a plant incubator. When seedlings grew to the five-leaf stage, they were transplanted into wet vermiculite in pots (7 cm × 7 cm × 10 cm) and kept in an artificial climate room. Pear seedlings received Hoagland solution every 3 days and were grown at 23 ± 2°C with a light intensity of 800 µmol m^-2·s^-1. Two-month-old seedlings were used to measure mRNA expression levels of PbP5CSs under biotic and abiotic stress conditions. Drought stress of potted pear plants was inflicted by withholding water for 20 days (Yang S. et al., 2021); waterlogging stress was inflicted by submerging plants in water (Yu et al., 2019); salinity-alkalinity stress was performed at a ratio of 1:1.4 NaCl and NaHCO₃ (Zhang et al., 2020); cold stress was simulated at 4°C (Xi et al., 2011); heat stress was simulated at 40°C (Liu et al., 2013). Pear rust was applied by infecting leaves with Gymnosporangium haraeanum Syd. (Li et al., 2006). Pear leaves were taken at 0, 1, 3, and 6 days after abiotic stresses, and at 0, 6, 12, and 24 hours after biotic stress. The samples (with three independent biological replications) were immediately frozen in liquid nitrogen and stored at -80°C until analysis. To explore the gene expression profiles of PbP5CSs in response to salt, cold, drought, and Alternaria alternate infection, pear RNA-seq datasets were retrieved from published supplemental datasets (SRA accession nos.: SRP077703, SRP287704, SRP148620, and SRP276846).

P5CS enzyme activities were measured using a commercial kit (Suzhou Geruisi Biotechnology, Suzhou, China) following the manufacturer’s instructions. Each experiment was independently repeated three times.

Total RNA extraction from leaf samples was performed using the method of Ma et al. (2022). First-strand cDNA was prepared using PrimeScript RTase (TaKaRa Biotechnology, Dalian, China). Primers for qPCR were designed using Primer Premier 5.0 (Premier Biosoft International, Silicon Valley, CA, USA). Primer sequences are detailed in Supplementary Table S1 . A LightCycler R 480 SYBR Green Master (Roche, Mannheim, Germany) was used for qPCR assays with a LightCycler R 480 II system (Roche, Rotkreuz, Switzerland). Relative expression levels of the target genes were calculated using the 2^–ΔΔCT method (Livak and Schmittgen, 2001) and normalized against the Actin gene (GenBank: AB190176).

Based on the conserved domains of protein sequences, 11 P5CS protein sequences were screened and named PbP5CS1−PbP5CS11 according to their chromosomal sequences and positions. Detailed physical and chemical characterizations of PbP5CS proteins are listed in Table 1 . The 11 PbP5CS proteins have different numbers of amino acids; PbP5CS7 is the shortest (276 amino acids), while PbP5CS4 is the longest (755 amino acids). The molecular weight of PbP5CSs ranged from 31.16 kDa (PbP5CS7) to 82.67 kDa (PbP5CS4). The isoelectric point values of PbP5CS proteins ranged from 6.06 (PbP5CS9) to 8.95 (PbP5CS2). Except for PbP5CS9 and PbP5CS11, the GRAVY values of other PbP5CSs were less than zero. The predicted subcellular localizations were cytoplasm for PbP5CS2, PbP5CS5, and PbP5CS9, endoplasmic reticulum for PbP5CS1 and PbP5CS4, chloroplast for PbP5CS3 and PbP5CS11, and nucleus for the other PbP5CSs.

To explore the evolutionary relationships of PbP5CS members, a phylogenetic tree was built using 50 conserved domain sequences of P5CS proteins from pear (11), apple (8), peach (14), poplar (13), and A. thaliana (4; Figure 1 ). In the tree, PbP5CS, MdP5CS, PpP5CS, PtP5CS, and AtP5CS were classified into five groups (Groups I, II, III, IV, and V). Notably, P5CS genes of woody plants (pear, apple, peach, and poplar) clustered together. Most of the pear P5CSs also clustered together with proteins from A. thaliana, consistent with the closer relationship of pear to eudicots. The distribution of pear P5CSs was uneven in these groups. Group I was the largest with 24 members, including nearly half of all pear P5CSs (PbP5CS1, PbP5CS3, PbP5CS4, PbP5CS5, and PbP5CS8). PbP5CS9 and PbP5CS11 were in both Groups II and IV, PbP5CS6 and PbP5CS7 were in Group III, and PbP5CS2 and PbP5CS10 were in Group V.

A total of 10 conserved motifs were predicted in PbP5CSs ( Figure 2B ; Supplementary Table S2 ), ranging from 21 to 100 amino acids in length. Interestingly, we observed that Motifs 1, 2, 3, 6, 7, 9, and 10 were present only in Group I members, which might contribute to the functional divergence of P5CS genes. Motif 5 was found not only in all members of Group I, but also in PbP5CS6 and PbP5CS7 in Group III, this suggests that PbP5CS6 and PbP5CS7 may have evolved from Group I. Group II only comprised Motif 4, Group IV only contained Motif 8. As Motifs 4 and 8 were found in PbP5CS2 and PbP5CS10 in Group V, PbP5CS2 and PbP5CS10 may have evolved from Groups II and IV ( Figures 2A, B ). We also found that four genes in Group I (PbP5CS1, PbP5CS3, PbP5CS4, and PbP5CS5) had more than 20 exons, while all others carried between six and 11 exons ( Figure 2C ).

According to their annotated genomic locations, we found that the 11 PbP5CSs were widely distributed among the pear chromosomes. Chromosome 12 contained four PbP5CS genes, whereas Chromosomes 1, 2, 4, 7, 13, 14, and 15 had only one gene ( Figure 3 ). PbP5CS1, PbP5CS3, PbP5CS4, and PbP5CS5 (six pairs) were segmental ( Figure 4A ). In order to further explore the homologous gene relationships of PbP5CSs, we compared the physical locations of P5CS genes among the genomes of pear, apple, peach, poplar, and A. thaliana ( Figure 4B ). P5CS genes showed intimate collinear relationships in these five species. P5CSs in pear and apple showed the closest collinear relationship, with seven P5CS genes in apple sharing a close evolutionary relationship with PbP5CS genes. Moreover, six P5CS genes in each of peach and poplar, as well as two P5CS genes in A. thaliana, shared close evolutionary relationships with PbP5CS genes ( Supplementary Table S3 ).

To better understand the gene functions and transcriptional regulation of PbP5CSs, we analyzed cis-elements in the promoter regions of PbP5CSs ( Figure 5 ). The conventional promoter element CAAT-box was found in all PbP5CS promoters. Various cis-elements related to plant growth, development, and responses to stresses and phytohormones were also identified ( Supplementary Table S4 ). Additionally, the conventional promoter element GC-motif was present in PbP5CS2, the seed-specific regulation element (RY-element) was found in the promoter of PbP5CS7, and the endosperm expression regulation element (GCN4_motif) was observed in the promoters of PbP5CS1 and PbP5CS8. The zein metabolism regulation element (O₂-site) was identified in the promoters of seven PbP5CS genes.

Among the cis-elements that respond to plant hormones, the abscisic acid responsive element (ABRE) was present in the promoter of all PbP5CS genes, methyl jasmonate responsive elements (TGACG-motif and CGTCA-motif) were observed in the promoters of all PbP5CS genes excluding PbP5CS8, and salicylic acid-responsive element (TCA-element) was found in eight PbP5CS genes. Gibberellin responsive elements (GARE-motif, P-box, and TATC-box) and auxin responsive elements (AUXRR-core and TGA) were identified in six PbP5CS genes. We also found the stress-related cis-acting element (ARE) involved in anaerobic induction in all PbP5CS genes. Furthermore, defense and stress response element (TC-rich repeat), drought-inducible response element (MBS), low temperature responsiveness (LTR), and circadian features were identified in the promoters of PbP5CS genes ( Figure 5 ; Supplementary Table S4 ).

To investigate the putative roles of the PbP5CS genes in pear development, we analyzed organic-specific expression patterns of PbP5CSs. Expression patterns of PbP5CS genes in five different tissues (petal, stigma, leaf, ovary, and shoot) were analyzed using publicly available gene expression data (SRP230672). Some PbP5CSs were expressed at considerably high levels in specific tissues ( Figure 6 ). For example, three PbP5CSs (PbP5CS1, PbP5CS5, and PbP5CS8) displayed higher expression levels in the petal than other tissues, implying that they play important roles in pear petal growth and development.

PbP5CS1, PbP5CS2, and PbP5CS4 exhibited high levels of expression in the stigma, which suggests that they may play roles in stigma growth and development. PbP5CS4, PbP5CS6, and PbP5CS9 were preferentially expressed in the ovary, which suggests that they may take part in in ovary growth and development. PbP5CS3, PbP5CS5, PbP5CS10, and PbP5CS11 showed relatively high expression levels in the leaf, which means that they could participate in leaf development. PbP5CS7 showed higher expression in the shoot than other tissues, which reflects its possible role in shoot growth ( Supplementary Table S5 ). These results suggest that PbP5CS genes have different expression patterns, and may play diverse roles in pear during growth and development of different tissues.

To explore whether the P5CS enzyme plays an essential role under biotic and abiotic stresses, we measured the enzyme activities of P5CS under drought, waterlogging, salinity-alkalinity, cold, heat, and G. haraeanum infection ( Supplementary Figure S1 ). The enzyme activities of P5CS gradually increased under the different stresses. We then investigated expression patterns of PbP5CS genes in response to different stresses. With regard to abiotic stresses, RNA-seq datasets for pear subjected to drought (SRP148620), salt (SRP077703), and cold (SRP287704) were explored. In general, expression levels of PbP5CS1, PbP5CS3, PbP5CS4, PbP5CS5, PbP5CS6, and PbP5CS11 were up-regulated by all abiotic stress treatments, indicating potential roles of these PbP5CSs in abiotic stress responses ( Figure 7 ; Supplementary Table S6 ). Under drought treatment, PbP5CS1, PbP5CS2, PbP5CS3, PbP5CS4, PbP5CS5, PbP5CS6, PbP5CS7, PbP5CS9, and PbP5CS11 were up-regulated, but PbP5CS8 and PbP5CS10 were repressed in response to short-term drought stress ( Figure 7A ). Furthermore, PbP5CS2, PbP5CS3, PbP5CS5, and PbP5CS8 were significantly up-regulated at 72 hours of NaCl treatment ( Figure 7B ). PbP5CS6 and PbP5CS9 were significantly induced at 50 days of cold treatment ( Figure 7C ). These results indicate that the responsive PbP5CSs may be involved in plant defense mechanisms under both short- and long-term abiotic stress.

To explore the potential roles of PbP5CS genes in responses to biotic stresses, we investigated RNA-seq datasets from the infection experiment of pear with A. alternate (SRP276846). PbP5CS1 and PbP5CS2 were significantly up-regulated at 24 and 8 hours post-infection, respectively. PbP5CS3, PbP5CS9, and PbP5CS11 showed significant up-regulation at 16 hours post-infection. PbP5CS4, PbP5CS5, and PbP5CS8 were markedly induced at 48 hours post-infection. However, PbP5CS6 and PbP5CS10 were distinctively repressed following A. alternate infection, whereas PbP5CS7 did not respond to infection ( Figure 7D ).

To validate previous RNA-seq data and reveal more details of PbP5CSs in stress responses, we investigated the transcription levels of PbP5CS genes under different environmental stresses, including drought, waterlogging, salinity-alkalinity, cold, heat, and G. haraeanum infection. Under drought stress, nine (PbP5CS1, PbP5CS2, PbP5CS3, PbP5CS4, PbP5CS5, PbP5CS6, PbP5CS7, PbP5CS9, and PbP5CS11) and two (PbP5CS8 and PbP5CS10) PbP5CSs were up-regulated and down-regulated, respectively ( Figure 8A ). Moreover, seven (PbP5CS2, PbP5CS3, PbP5CS4, PbP5CS5, PbP5CS7, PbP5CS9, and PbP5CS11) and four (PbP5CS3, PbP5CS8, PbP5CS9, and PbP5CS10) PbP5CS genes were respectively up-regulated and down-regulated under waterlogging treatment ( Figure 8B ). Similarly, eight PbP5CS genes (PbP5CS1, PbP5CS2, PbP5CS3, PbP5CS4, PbP5CS7, PbP5CS9, PbP5CS10, and PbP5CS11) showed increased expression levels and three PbP5CSs (PbP5CS5, PbP5CS6, and PbP5CS8) were down-regulated by salinity-alkalinity treatment ( Figure 8C ). Additionally, the transcription levels of seven PbP5CS genes (PbP5CS1, PbP5CS3, PbP5CS4, PbP5CS5, PbP5CS9, PbP5CS10, and PbP5CS11) were increased under cold stress ( Figure 8D ). Under heat stress, seven PbP5CS genes (PbP5CS1, PbP5CS2, PbP5CS4, PbP5CS5, PbP5CS6, PbP5CS7, and PbP5CS11) were up-regulated, and the remaining four PbP5CSs (PbP5CS3, PbP5CS8, PbP5CS9, and PbP5CS10) were down-regulated ( Figure 8E ).

Pear rust caused by G. haraeanum is one of the main diseases affecting pear production. To understand the potential functions of PbP5CS genes in response to biotic stress, transcript levels of 11 PbP5CSs were measured by qPCR in pear subjected to G. haraeanum infection. Seven (PbP5CS1, PbP5CS2, PbP5CS3, PbP5CS6, PbP5CS7, PbP5CS9, and PbP5CS11) and four (PbP5CS4, PbP5CS5, PbP5CS8, and PbP5CS10) PbP5CS genes were up-regulated and down-regulated, respectively ( Figure 8F ). Interestingly, PbP5CS4 and PbP5CS11 were also strongly induced by drought, waterlogging, salinity-alkalinity, cold, and heat, suggesting that these two genes might be candidate genes for mitigating abiotic stresses. In particular, PbP5CS11 was induced dramatically by various abiotic and biotic stresses ( Figure 8 ).