Introduction

The pear, belonging to the Rosaceae family, has a rich history of global cultivation (Wu et al. 2013). Stone cells in pears significantly affect fruit quality due to their lignified cell walls (Gong et al. 2020a), characterized by sclerenchyma cells with hardened secondary cell walls (SCWs) due to lignin deposition (Tao et al. 2009).

Plant cell walls are complex structures mainly composed of lignin, cellulose, hemicellulose and pectin (Fry 2004). Pectin methylesterase (PME, EC 3.1.1.11) is an enzyme modifying pectin by removing methyl groups, leaving carboxyl groups on pectin chains (Röckel et al. 2008). PME can be classified based on the presence or absence of the PME inhibitor (PMEI) domain. Most PME proteins have both PME and PMEI domains, though some have only the PME domain (Jolie et al. 2010). Moreover, PME influences the hardness of plant cell walls, with increased activity decreasing hardness and vice versa (Pelletier et al. 2010; Levesque-Tremblay et al. 2015; Wen et al. 2020). Furthermore, PME plays a crucial role in pectin modifications during the deposition of secondary walls in xylem cells (Hertzberg et al. 2001; Geisler-Lee et al. 2006). The significance of these modifications lies in the role of de-esterified homogalacturonan, a type of pectic polysaccharide domain, in xylem lignification (Pelloux et al. 2007). When bridged by Ca2+, de-esterified homogalacturonan binds peroxidases, potentially initiating lignin polymerization (Carpin et al. 2001; Dunand et al. 2002). Supporting this, colocalization of PME, de-esterified homogalacturonan, peroxidase, and lignification initiation occurs at cell junctions in woody tissues (Guglielmino et al. 1997; Wi et al. 2005). Furthermore, the interaction between pectin and lignin monomers influences lignin polymerization (Lairez et al. 2005). In summary, these insights provide novel perspectives on the roles of PME and pectin in lignification.

PMEs are associated with different developmental processes, including lignin biosynthesis (Sexton et al. 2012), fruit ripening and softening (Wen et al. 2020), hypocotyl growth (Pelletier et al. 2010), embryo development, seed germination (Levesque-Tremblay et al. 2015), and pollen tube elongation (Tang et al. 2020). Eucalyptus pilularis pectin methylesterase 7 (EpPME7) can negatively regulate cellulose and lignin contents (Sexton et al. 2012). Studies also indicate a close relationship between pectin metabolism and post-harvest loquat fruit cellular lignification (Huang et al. 2022). In strawberries and tomatoes, specific PME gene family members play essential roles in regulating fruit firmness (Wen et al. 2020; Xue et al. 2020). However, the relationships between lignification, PME, and the role of PME in stone cell formation in pear fruits remain unclear. Elucidating this relationship could offer a novel perspective on stone cell formation mechanisms.

Functional characterization of PMEs has been limited to a subset, including those in Arabidopsis (Louvet et al. 2006), tomato (Wen et al. 2020), and strawberry (Xue et al. 2020). The distribution of PMEs in pears and many other Rosaceae species is not fully understood. Additionally, limited research on the pear PME family leaves the relationship between these genes and pear fruit lignification unclear. This study identified PME family genes from five Rosaceae species, evaluating their gene structure, duplication modes, location, collinearity, and the ratio of the nonsynonymous substitutions per nonsynonymous site (Ka) to the synonymous substitutions per synonymous site (Ks) (Ka/Ks values). Transcriptome analysis revealed that PbPME35 is correlated with stone cell formation and lignin biosynthesis, with its biological function of suppressing lignification confirmed through homologous and allogenic transformation. These findings provide new insights into the mechanism of lignification in pear fruit stone cell formation.

Materials and methods

Plant materials

Pear fruit samples from two cultivars, Pyrus bretschneideri ‘Dangshan Su’ (DS) and Pyrus pyrifolia ‘Cuiguan’ (CG), were collected at three developmental stages (15, 35, and 55 days after flowering [DAF]) from Gaoyou Orchard in Yangzhou, Jiangsu, China. Ten trees (five for each cultivar), grown under healthy and uniform conditions without disease or insect infection, were randomly selected. Upon returning to the laboratory, the pear fruit samples were immediately frozen in liquid nitrogen and stored in an ultralow-temperature refrigerator at –80°C for subsequent experiments.

Identification and phylogenetic analysis of PME genes

PME genes from different species, including Arabidopsis thaliana, Pyrus bretschneideri, Pyrus pyrifolia, Malus domestica, Fragaria vesca, Prunus mume, and Pyrus communis, were identified using the methods outlined by Qiao et al. (2015). An evolutionary tree was constructed using Orthofinder (v2.2.7), MAFFT (v7.475), TrimAI (v1.4.1), Raxml (v8.2.12), Astral (v5.7.8), and EasySpeciesTree (v1.0). The TimeTree database (http://timetree.org/home) was referenced for mapping species divergence time (Fig. S1a). Genome files from various databases reported previously were collected (Table 1). HMMER (v3.3.2) and BLAST (v2.11.0) were used to identify candidate genes of each species. The Hidden Markov Model profile (PF01095) was obtained from the Pfam database, and CDD/SPARCLE was used to confirm PME domain accuracy in the candidate proteins (Lu et al. 2020). Multiple sequence alignments were performed using MAFFT (v7.475), and maximum likelihood (ML) phylogenetic trees were constructed using IQ-TREE (v2.1.2) (Katoh et al. 2019; Minh et al. 2020). Nomenclature guidelines and details for candidate PME genes from five Rosaceae species are provided in Table S1.

Table 1 Genomic information
Full size table

Chromosomal location and gene structure analysis of PME genes

Chromosomal location information for PME genes was extracted using genomic annotation data from various species, with details presented in Table S1. MEME (v5.0.5) was used to identify conserved protein motifs (Bailey et al. 2010) with the maximum value set to 10. The motif length was adjusted within the range of 6 to 50 during motif analysis.

Gene duplication modes, collinearity analysis, and Ka/Ks calculation

To ascertain gene duplication modes in the five Rosaceae species, we employed DupGen Finder, a specialized software package (Qiao et al. 2019). TBtools (v1.108) was used to determine the localization and synteny of the PME genes and calculate the Ka/Ks value (Chen et al. 2020). MCscan (Python version) was used to analyze the collinearity relationships among the five Rosaceae species (Tang et al. 2008).

Expression pattern of PME genes in pear

RNA-seq data from two regions, one rich in stone cells content and the other with few stone cells, were obtained for three stages (15, 35, and 55 DAF) of CG and DS fruits from the NCBI bioProject PRJNA825067. The expression patterns of PbPME genes were extracted from these data, and the results are shown in Table S6.

Arabidopsis plant and pear fruit callus transformation

Following the methods outlined by Gong et al. (2022), PbPME35 was overexpressed in the Arabidopsis Columbia ecotype through the floral dip assay. T1-generation seeds were screened on media supplemented with sucrose (30 g/L), agar (0.75%), hygromycin (20 mg/L), timentin (100 mg/L), and carbenicillin (100 mg/L). Transgenic lines were identified using PCR. Positive transgenic Arabidopsis plants from the T3 generation and wild-type (WT) plants were grown on MS media without growth regulators for functional verification.

For pear callus transformation, callus tissues were added to LB cultures under aseptic conditions and placed on a shaker at 120 r·min−1 for 30 min. Infected calli were evenly spread on acetosyringone (AS) plates for 24 h after draining the liquid. Subsequently, the calli were transferred to hygromycin-containing MS solid media to screen for transgenic calli, which were then cultured at 24°C under continuous dark conditions (Bai et al. 2019).

RT-qPCR analysis

To validate the expression levels of the selected PbPME genes, RNA was extracted from pear pulp samples using a Fuji total RNA purification kit. The extracted RNA was then reverse transcribed using a TransGen Biotech reverse transcription kit. Primers were designed using Primer (v5.0) software for specificity (Table S8). The PCR mixture comprised premix (water:fluorescent = 1:2; 15 µL), template cDNA (200 ng), and forward (2.5 µL) and reverse (2.5 µL) primers. The PCR procedure involved preincubation for 5 min at 95°C, followed by 55 cycles of denaturation at 95°C for 3 s, annealing at 60°C for 10 s, and extension at 72°C for 30 s. P. bretschneideri ubiquitin (PbUBQ) served as the internal reference gene, and the relative gene expression level was calculated using the 2−ΔΔct method.

Subcellular localization of PbPME35

The transient transformation of tobacco leaves was conducted following the protocol from a previous study (Yang et al. 2000). An Agrobacterium suspension transfected with 35S:PbPME35-GFP, cultured in LB liquid media supplemented with kanamycin and rifampicin, was injected into the lower surface of 5-week-old Nicotiana benthamiana leaves. After 48 h of growth, live cell imaging of the transformed leaves was conducted using an inverted confocal microscope (Zeiss LSM 780, Carl Zeiss AG, Oberkochen, Germany).

Lignin and stone cell content determination

Lignin content was assessed using the acetyl bromide method outlined by Tao et al. (2009), utilizing a standard sample curve (Sigma-Aldrich, St. Louis, MO, USA) for calculation. Stone cell content was quantified using the Frozen-HCL method outlined by Gong et al. (2020).

Anatomical structure observation

To examine the anatomical structures of Arabidopsis stems and pear pulp, paraffin sectioning was performed (Zhang et al. 2020). Sections were cut using a microtome (Leica, Biosystems, Wetzlar, Germany), stained with safranin and fast green or toluidine blue, and observed under a microscope (Zeiss LSM 780).

Statistical analysis

The data were subjected to analysis of variance using IBM SPSS (v19.0) for significance assessment. Microsoft Office Excel 2013 was used for data processing. Plots were generated using Origin (v9.0), RStudio, Hiplot (https://hiplot.com.cn), and TBtools (Chen et al. 2020).

Results

Identification, phylogenetic analysis, and conserved motif patterns of PME family members

In this study, we identified 396 putative PME family candidate genes distributed across various species, including 81 in P. bretschneideri, 92 in M. domestica, 62 in F. vesca, 65 in P. mume, 15 in P. communis, and 81 in P. pyrifolia (Table S1). Additionally, 66 PME family genes were identified in A. thaliana, consistent with the TAIR database and previous research (Louvet et al. 2006).

A phylogenetic tree, constructed using PME proteins from P. bretschneideri, P. pyrifolia, and A. thaliana through the ML method, revealed four clades (I, II, III, and IV) indicating an evolutionary divergence approximately 3.54 million years ago (MYA) (Fig. 1a, Fig. S1a). Further, a phylogenetic tree encompassing PME genes from Arabidopsis, along with those from P. bretschneideri, M. domestica, F. vesca, P. mume, and P. communis, showcased the same four clades (I, II, III, and IV), indicating the conserved evolutionary relationships within the PME family (Fig. S1b).

Fig. 1
figure 1

Phylogenetic and conserved motif analyses of Pectin methylesterase (PME) family members. (a) Phylogenetic analysis depicting the relationship among PMEs from Arabidopsis, Pyrus bretschneideri, and P. pyrifolia. (b) Phylogeny, conserved motif distribution, and coding sequence (CDS)–untranslated region (UTR) structure of PME genes in pear

Full size image

Conserved motif analysis revealed variability in motif composition among P. bretschneideri PME genes, with motif 6 being predominant in most clades, except for PbPME17 in clade I and PbPME74 in clade IV (Fig. 1b, Fig, S2a). Exon–intron analysis revealed notable patterns, such as 64.10% of clade I PbPME genes having only two exons, consistent with similar trends observed in clades II and III. There were seven PbPME genes with two exons in clades II and III, accounting for 77.78% and 87.50%, respectively. Clade IV exhibited a higher proportion (19 PbPME genes) with more than three exons, with only two genes having two exons (Fig. 1b).

Gene duplication, evolution, and synteny analysis of PME family

Using DupGen Finder to evaluate gene duplication modes within the PME family, we identified five modes—whole-genome duplication (WGD), tandem duplication (TD), proximal duplication (PD), transpose duplication (TRD), and dispersed duplication (DSD) as potential drivers (Qiao et al. 2019). All PME genes across the studied species showed involvement in WGD, TD, PD, TRD, or DSD. DSD played a substantial role in evolution, with proportions of 51.41%, 56.49%, 65.12%, 75.68%, and 57.89% in P. bretschneideri, M. domestica, F. vesca, P. mume, and P. communis, respectively. However, the second significant duplication mode differed among these species, with WGD being the second most significant mode in P. bretschneideri and M. domestica, TRD in F. vesca and P. communis, and PD in P. mume (Fig. 2a, Table S2).

Fig. 2
figure 2

Gene-duplication modes, Ka/Ks values, gene locations, and collinearity analysis of the Pectin methylesterase (PME) gene family. (a) Various modes of gene duplication, with the x-axis representing the number of duplicated gene pairs and the y-axis denoting the species. (b) Ka/Ks values across five Rosaceae species. (c) Ka/Ks values corresponding to different duplication modes in Pyrus bretschneideri. (d) Gene locations in P. bretschneideri and collinearity analysis, displaying genes distributed across different chromosomes. Red lines represent syntenic gene pairs. (e) Collinearity analysis comparing P. bretschneideri with four Rosaceae species. DSD, dispersed duplication; TRD, transpose duplication; PD, proximal duplication; TD, tandem duplication; WGD, whole genome duplication; Ka, nonsynonymous substitutions per nonsynonymous site; Ks, synonymous substitutions per synonymous site

Full size image

To evaluate selection pressure, we calculated Ka/Ks values for PME genes in the five species using TBtools (Table S3). Results indicated Ka/Ks values below 1 for almost all gene pairs in P. bretschneideri and all gene pairs in other species, signifying purifying selection. However, two gene pairs (PbPME67–PbPME68 and PbPME20–PbPME21) in P. bretschneideri underwent positive selection (Ka/Ks > 1; Fig. 2b). These gene pairs evolved through PD (PbPME67–PbPME68) and DSD (PbPME20–PbPME21) (Table S1, Fig. 2c and d).

In P. bretschneideri, 62 PbPME genes were unevenly distributed across 16 chromosomes, except for chromosome 14, and 19 PbPME genes were located on unanchored scaffolds (Fig. 2d). Chromosome 2 had the most PME genes, accounting for 17.28% of all PME genes (Fig. 2d and Table S1). Additionally, intragenomic synteny blocks identified 69 syntenic gene pairs, with P. bretschneideri and M. domestica having the highest number (37 and 26 pairs, respectively). F. vesca, P. mume, and P. communis had fewer syntenic pairs (3, 1, and 2 pairs, respectively; Fig. 2d and Table S4). Furthermore, a comparative syntenic graph revealed orthologous relationships between PME genes in P. bretschneideri and other species (M. domestica, F. vesca, P. mume, and P. communis), showing 81 syntenic gene pairs with M. domestica, 20 with P. communis, 32 with F. vesca, and 45 with P. mume (Fig. 2e and Table S5).

Expression analysis of PME genes

Transcriptome data from two pear varieties with varying stone cell contents identified 19 PbPME genes expressed in pear fruits (Fig. 3a and Table S6). RT-qPCR analysis was performed to validate expression patterns for four randomly selected genes (PbPME13, PbPME29, PbPME35, and PbPME55), highly consistent with RNA-seq data (Fig. 3b). Critical stone cell formation periods in pear fruits were identified as 15, 35, and 55 DAF. Additionally, although stone cells are distributed near the fruit cores, they can be categorized into different regions with few (L) or rich (R) stone cells (Gong et al. 2022).

Fig. 3
figure 3

Expression analysis of Pectin methylesterase (PME) genes in pear. (a) Heatmap analysis presenting the expression patterns of PME genes. (b) Correlation analysis between the expression levels and RNA-seq data of four PME genes. (c) Gene expression clustering trend analysis of PME genes at various fruit development stages. (d) Candidate PME genes were screened through gene expression trend analysis and Venn analysis. CG and DS indicate different cultivars (CG, Pyrus pyrifolia ‘Cuiguan’; DS, Pyrus bretschneideri ‘Dangshan Su’). L and R indicate different fruit pulp parts (L, Regions with few stone cells in pear fruits; R, Regions with rich stone cells in pear fruits). The numbers 15, 35, and 55 indicate different pear fruit development stages concerning the number of days after flowering (DAF). PbPME, Pyrus bretschneideri Pectin methylesterase

Full size image

To ensure consistency of PbPME gene expression in cultivars or stone cell regions, we categorized expression levels of the PbPME genes extracted from RNA-seq data into four groups: regions in CG with few stone cells (CG_L), regions in CG with rich stone cells (CG_R), and the same for DS (DS_L) and (DS_R) (Fig. 3c). Our results revealed two distinct expression patterns (Trend 1 and Trend 2) shared across all groups, each containing two common genes (Trend 1: PbPME34 and PbPME35; Trend 2: PbPME20 and PbPME75). In all groups, PbPME34 and PbPME35 exhibited an initial increase followed by a decrease, whereas PbPME20 and PbPME75 showed a decrease, followed by stabilization. RT-qPCR data for these four genes confirmed these patterns (Fig. 3c and d). Notably, PbPME34 and PbPME35 exhibited high expression levels in RNA-seq data, whereas PbPME20 and PbPME75 exhibited low expression (Fig. 3d). Moreover, the gene pair PbPME34–PbPME35, originating from WGD (Table S2) and displaying synteny (Table S4) with an identity of up to 99.19% (Table S7), led us to select PbPME35 as a candidate gene influencing lignin biosynthesis due to its high homology with PbPME34.

Analysis of phenotypic traits, PbPME35 expression in mature pear fruit and subcellular localization of PbPME35

Utilizing the phloroglucinol-HCl method to stain cross sections of mature fruits revealed a higher concentration of stone cells near the fruit core. Further observation of stone cell clusters in different pear fruit parts through paraffin sectioning indicated larger stone cell clusters and greater stone cell density near the core regions compared to those near the peel (Fig. 4a). Consequently, we categorized pear fruits into two regions: the stone cell enrichment area (near the core) and the non-enrichment area (near the peel). Results from fruit firmness, stone cell content, and lignin content determination demonstrated higher expression of these phenotypic traits in the stone cell enrichment area than in the non-enrichment area (Fig. 4b). However, the expression level of PbPME35 in the stone cell enrichment area was lower than that in non-enrichment area (Fig. 4c). Furthermore, the expression levels of lignin biosynthetic genes (P. bretschneideri 4-coumarate: CoA ligase 1 [Pb4CL1], P. bretschneideri p-coumarate 3-hydroxylase 1 [PbC3H1], PbC3H2, P. bretschneideri cinnamyl alcohol dehydrogenase 2 [PbCAD2], P. bretschneideri Caffeoyl coenzyme A O-methyltransferase [PbCCOAOMT], P. bretschneideri cinnamoyl-CoA reductase 1 [PbCCR1], P. bretschneideri caffeoyl shikimate esterase 1 [PbCSE1], P. bretschneideri hydroxycinnamoyl CoA:shikimate/quinate hydroxycinnamoyl transferase 17 [PbHCT17], PbPAL1: P. bretschneideri phenylalanine ammonia-lyase 1 [PbPAL1], P. bretschneideri peroxidase [PbPOD1], and PbPOD2) were significantly higher in regions with greater stone cell counts (Fig. 4d). Additionally, correlation analysis was used to evaluate the relationships between PbPME35 and other stone cell-related traits (or genes). The expression of PbPME35 was significantly negatively correlated with stone cell-related traits (including stone cell content, lignin content, and fruit firmness) and stone cell-related genes (all mentioned lignin biosynthetic genes), highlighting the inhibitory role of PbPME35 in stone cell formation (Fig. 4e).

Fig. 4
figure 4

Phenotypic trait analysis and Pyrus bretschneideri Pectin methylesterase 35 (PbPME35) expression levels in mature pear fruits. (a) Phloroglucinol staining and cross-sectional staining of mature pear fruit. (b) Analysis of fruit firmness, stone cell content, and lignin content in different regions of mature pear fruit. (c) PbPME35 expression levels in different regions of mature pear fruit. (d) Relative expression patterns of lignin biosynthetic genes in pear fruits enriched (or not) with stone cells. (e) Correlation analysis among phenotypic traits, expression levels of lignin biosynthetic genes, and PbPME35 expression levels. (f) Subcellular localization of PbPME35. * P < 0.05, ** P < 0.01. Pb4CL1, P. bretschneideri 4-coumarate: CoA ligase 1; PbC3H1, P. bretschneideri p-coumarate 3-hydroxylase 1; PbCAD2, P. bretschneideri cinnamyl alcohol dehydrogenase 2; PbCCOAOMT, P. bretschneideri Caffeoyl coenzyme A O-methyltransferase; PbCCR1, P. bretschneideri cinnamoyl-CoA reductase 1; PbCSE1, P. bretschneideri caffeoyl shikimate esterase 1; PbHCT17, P. bretschneideri hydroxycinnamoyl CoA:shikimate/quinate hydroxycinnamoyl transferase 17; PbPAL1, P. bretschneideri phenylalanine ammonia-lyase 1; PbPOD1, P. bretschneideri peroxidase 1; GFP, green fluorescence

Full size image

To investigate subcellular localization, we employed Agrobacterium-mediated transformation, introducing the 35S:PbPME35-GFP fusion construct (treatment) and 35S:GFP (control) into tobacco leaves. While the positive control expressed a green fluorescence signal in the cell membrane and nucleus, leaves injected with 35S:PbPME35-GFP expressed this signal solely in the cell membrane (Fig. 4f).

Overexpression of PbPME35 in Arabidopsis

Arabidopsis thaliana (Col-0) was infected with inflorescences to generate overexpression (OE) lines of PbPME35 (Fig. 5a and b). Three independent transgenic lines, exhibiting significantly higher PbPME35 expression levels, were selected for further analysis (Fig. 5c). After 10 d of growth, both root and hypocotyl lengths were significantly extended in the transgenic plants compared to the WT plants (Fig. 5a, d, and e). At 5 weeks of age, the transgenic plants displayed increased growth rates and taller heights compared to the controls (Fig. 5b). Additionally, lignin content in the stems of the transgenic Arabidopsis plants significantly decreased after PbPME35 OE (Fig. 5f). Paraffin sectioning further revealed less pronounced phenotypic traits in the interfascicular fibers and xylem cells of transgenic plants (Fig. 5i). Furthermore, RT-qPCR quantified the downregulation of six lignin biosynthetic genes (4CL1, CAD4, CAD9, CCOAOMT, ferulate 5-hydroxylase 1 [F5H1], and hydroxycinnamoyl CoA:shikimate/quinate hydroxycinnamoyl transferase [HCT]) in the transgenic plants (Fig. 5g). Additionally, six transcription factors involved in lignin biosynthesis (NST1, MYB85, MYB103, MYB58, MYB83, and MYB46) were also downregulated in the transgenic plants (Fig. 5h), indicating the negative regulatory effect of PbPME35 on lignin biosynthesis.

Fig. 5
figure 5

Overexpression (OE) of Pyrus bretschneideri Pectin methylesterase 35 (PbPME35) in Arabidopsis. (a) Ten-day-old plants. (b) Five-week-old plants. (c) Expression patterns of PbPME35. (d) Root length. (e) Hypocotyl length. (f) Lignin content. (g) Expression patterns of lignin biosynthetic genes. (h) Expression patterns of transcription factors regulating lignin biosynthesis and secondary wall synthesis. (i) Stem anatomy of wild-type (WT) and transgenic Arabidopsis plants. Toluidine blue-stained sections illustrate variations in cell walls in Arabidopsis plants overexpressing PbPME35. Black arrows highlight the lignification of the xylem and interfascicular fiber cells. Different lowercase letters in a column indicate significant differences at the 5% level. * P < 0.05, ** P < 0.01

Full size image

Overexpression of PbPME35 in pear fruit callus

We generated a homologous stable OE system in pear fruit callus to elucidate the biological function of PbPME35. The expression level of PbPME35 significantly increased in OE calli compared to control calli (Fig. 6c). Phloroglucinol-HCl staining revealed a lighter red stain in control, whereas the transgenic line exhibited minimal red staining, suggesting a reduction in lignin accumulation upon PbPME35 OE (Fig. 6a). This observation aligned with the lignin determination results (Fig. 6b). Additionally, the expression levels of five lignin biosynthetic genes (Pb4CL1, PbC3H1, PbCCR1, PbPAL1, and PbPOD1) were significantly downregulated in the OE callus (Fig. 6d). These findings provide an explanation for the decreased lignin content and underscore the inhibitory role of PbPME35 in lignin biosynthesis.

Fig. 6
figure 6

Overexpression (OE) of Pyrus bretschneideri Pectin methylesterase 35 (PbPME35) in pear fruit callus. (a) Phloroglucinol staining. (b) Lignin content. (c) Expression level of PbPME35. (d) Expression patterns of lignin biosynthetic genes in pear fruit calli. * P < 0.05, ** P < 0.01

Full size image

Discussion

The PME family, recognized for its significant role in plant development, particularly in lignin biosynthesis (Sexton et al. 2012), was investigated in the context of stone cell formation, a process tightly linked to lignin accumulation. Five Rosaceae plants and pear species were analyzed, leading to the identification of 396 PME genes across six species, namely P. bretschneideri, M. domestica, F. vesca, P. mume, P. communis, and P. pyrifolia. The distribution of PME genes varied among species, with P. bretschneideri hosting 81, and M. domestica, F. vesca, P. mume, P. communis, and P. pyrifolia harboring 92, 62, 65, 15, and 81 PME genes, respectively. The classification of PME genes into four clades in P. bretschneideri and P. pyrifolia exhibited high consistency (Louvet et al. 2006).

Gene duplication, a prominent force in gene family evolution (Lynch and Force 2000), was evident in these Rosaceae species, indicating a shared WGD event from a common Rosaceae ancestor (Jiao et al. 2011). Despite a more recent WGD event estimated at approximately 35–45 MYA in pear (Huang et al. 2015), our analysis indicated that P. bretschneideri and P. pyrifolia diverged approximately 3.54 MYA (Fig. S1a). This discrepancy in divergence times may contribute to the high homology observed in the PME genes of these two species. The conservation of motifs within the same subfamilies supported the credibility of our classification, as closely related genes in the phylogenetic tree shared motifs, suggesting functional similarity among PbPME proteins within the same clade.

Previous studies have classified gene duplication patterns into five modes: WGD, PD, TD, DSD, and TRD, with each mode contributing uniquely to gene family expansion (Freeling 2009; Qiao et al. 2019). WGD, TD, and DSD are prominent features in eukaryotic genome evolution (Fawcett et al. 2009), a trend consistent with our findings, where DSD emerged as the primary force, aligning with previous research. Moreover, the Ka/Ks values for almost all PME genes were < 1, indicating robust purifying selection (Qiao et al. 2015). However, two gene pairs (PbPME67–PbPME68 and PbPME20–PbPME21) in P. bretschneideri exhibited Ka/Ks > 1.2, suggesting positive selection driving their evolution.

Mapping 62 PbPME genes to 16 chromosomes and locating 19 other PbPME genes on various scaffolds revealed an uneven distribution across chromosomes. To gain further insights into the evolutionary history and identify orthologous genes in the P. bretschneideri PME family, we constructed comparative syntenic maps (Fig. 2e). Interspecific collinear analysis highlighted a higher number of orthologous genes (81) between P. bretschneideri and M. domestica than between P. bretschneideri and P. communis (20). This suggests a significant divergence in the PME gene family post-split between P. bretschneideri and P. communis, with a relatively smaller divergence between P. bretschneideri and M. domestica after their divergence.

PMEs are associated with cell lignification and fruit softening (Sexton et al. 2012; Wen et al. 2020). Therefore, we hypothesized that PbPME35 plays a critical role in regulating fruit firmness, lignin biosynthesis, and stone cell formation in pears. Our findings revealed a significant negative correlation between PbPME35 expression and stone cell content, lignin content, fruit firmness, and lignin biosynthetic genes. Furthermore, the overexpressing of PbPME35 in transgenic Arabidopsis and transgenic pear fruit calli exhibited decreased lignin content and tissue lignification. This reduction may arise from the downregulation of genes related to lignin biosynthesis, aligning with previous research (Sexton et al. 2012). Our results suggest that PbPME35 reduces lignin content by suppressing the expression of lignin biosynthesis-related genes, contributing to a decline in stone cell content and fruit firmness, similar to mechanisms identified in previous studies on stone cells (Gong et al. 2020b).

To explore whether PbPME35 influences transcription factor expression, we investigated six transcription factors associated with lignin biosynthesis and secondary cell wall biosynthesis (NST1, MYB85, MYB103, MYB58, MYB83, and MYB46). These transcription factors reportedly promote the expression of lignin-related genes and improve lignin biosynthesis (Mitsuda et al. 2005; Zhong et al. 2008; McCarthy et al. 2009; Zhou et al. 2009; Öhman et al. 2013). Notably, P. bretschneideri NAC stone cell promoting factor (PbrNSC) and P. bretschneideri myeloblastosis 169 (PbrMYB169), homologous to A. thaliana NAC secondary wall thickening promoting factor 1 (NST1) and A. thaliana myeloblastosis 85 (MYB85), respectively, regulate lignin biosynthesis genes and influence lignin formation (Xue et al. 2019; Wang et al. 2021). The expression levels of all investigated transcription factors were lower in transgenic Arabidopsis plants than in WT plants, indicating that PbPME35 could modulate lignification during stone cell development by regulating transcription factor expression. Therefore, we speculate that PbPME35 might interact with lignin biosynthetic proteins or critical switches, leading to the downregulation of lignin biosynthetic genes and, consequently, a reduction in lignin content and inhibition of cell lignification.

The quantity, distribution, and cell wall characteristics of stone cells significantly influence pear fruit texture and quality (Tao et al. 2009; Zhang et al. 2021). Lignin is a major component of stone cells (Gong et al. 2022). Our exploration of the function of PbPME35 provides new insights into lignin biosynthesis and stone cell formation in pear fruit, laying a theoretical groundwork for cultivating pear cultivars with reduced stone cell counts. Notably, we identified high homology between PbPME35 and MdPME42, MdPME43, and FvPME14 (Table S7 and Fig. S3). Consequently, we propose that MdPME42 and MdPME43 may regulate anthocyanin biosynthesis and fruit color in apples through substrate competition with lignin biosynthesis (Hu et al. 2021). Similarly, inhibiting the expression of FvPME14, assuming it functions similarly to PbPME35, may increase the lignin content and firmness of strawberry fruits, potentially extending their shelf life (He et al. 2018). However, these genes functions require further study to clarify their effects on lignin deposition or fruit firmness. Moreover, lignin deposition and stone cell formation in pear fruit is a complex biological process that is regulated by various developmental or environmental signals (Wang et al. 2021). Although this study clarifies the role of PbPME35 in suppressing fruit lignification, the lignin biosynthetic proteins or critical switches that might interact with PbPME35 have not been revealed. The molecular hubs connecting PME and lignification needs further study.

Conclusions

In this study, we identified 396 PME genes across six species (P. bretschneideri, M. domestica, F. vesca, P. mume, P. communis, and P. pyrifolia), with 81 genes belonging to P. bretschneideri. Notably, PbPME35 was identified as a key regulator, suppressing lignification in pear fruit. Specifically, our findings demonstrate that PbPME35 reduces pear fruit lignin content by downregulating the expression of transcription factors and lignin biosynthesis genes.