Genome-wide investigation of glycoside hydrolase 9 (GH9) gene family unveils implications in orchestrating the mastication trait of Citrus sinensis fruits

Abstract

Mastication trait of citrus significantly influences the fruit’s overall quality and consumer preference. The accumulation of cellulose in fruits significantly impacts the mastication trait of citrus fruits, and the glycoside hydrolase 9 (GH9) family plays a crucial role in cellulose metabolism. In this study, we successfully identified 32 GH9 genes from the Citrus sinensis genome and subsequently conducted detailed bioinformatics analyses of the GH9 family. Additionally, we profiled the spatiotemporal expression patterns of CsGH9 genes across four distinct fruit tissue types and six crucial developmental stages of citrus fruits, leveraging transcriptome data. Parallel to this, we undertook a comparative analysis of transcriptome profiles and cellulose content among diverse fruit tissues spanning six developmental stages. Furthermore, to identify the pivotal genes involved in cellulose metabolism within the GH9 family during fruit maturity, we employed correlation analysis between cellulose content and gene expression in varying tissues across diverse citrus varieties. This analysis highlighted key genes such as CsGH9A2/6 and CsGH9B12/13/14/22. Collectively, this study provides an in-depth analysis of the GH9 gene family in citrus and offers novel molecular insights into the underlying mechanisms governing the mastication trait formation in citrus fruits.

Introduction

Cellulose, a ubiquitous polysaccharide found in the cell walls of plants, serves as a fundamental structural component in the plant [1]. It affects the mechanical strength of cell walls and further plays an important role in the formation of fruit quality [2]. In recent years, there has been significant attention on the study of cellulose-degrading enzymes and their roles in processes like fruit hardness, mastication traits, and stone cell formation in the field of the formation and regulation of fruit quality [2,3,4]. Among these enzymes, the GH9 family of cellulases has emerged as a critical player in cellulose metabolism [5].

In the initial phases of fruit development, the cellulose content within the primary cell wall assumes a significant role, offering structural reinforcement while facilitating cellular expansion and growth [6, 7]. Specifically, apples exhibit a high concentration of cellulose during their early stages of maturation, which contributes significantly to their structural integrity and brittleness [8]. As fruits progress towards ripening, the cellulose in their cell walls undergoes degradation, a process catalyzed by enzymes like cellulase [9]. In the case of pears, cellulose deposition can lead to the development of stone cells, thereby influencing the overall taste of the fruit [10]. GH9 family comprises endo-β-1,4-glucanases that are pivotal in various biological processes, including cellulose metabolism, cell wall formation, organ abscission, and fruit development [11, 12]. Notably, cellulose degradation is a crucial step in rendering fruits softer and more palatable. In Arabidopsis, KOR1, a member of the GH9 family, plays a fundamental role in cellulose biosynthesis for both primary and secondary cell walls [13, 14]. Considerable experimental evidence suggests that KOR1 and its homologs, such as PttCel9A1, have a direct impact on plant growth and modulate cellulose content in the stem [15, 16]. Despite their prevalence, the specific functions and regulatory mechanisms of GH9 family genes in fruit development and mastication traits remain relatively understudied.

Citrus is cultivated globally, and its mastication trait plays a crucial role in determining the texture and taste of the fruit, serving as a key indicator of its overall quality. The mastication trait of citrus fruits is closely linked to the cell wall components of both the segment membrane and juice sacs, primarily encompassing pectin, cellulose, and lignin [17,18,19]. Among these components, cellulose content has been identified as a key factor influencing the mastication trait of citrus fruit; lower cellulose content corresponds to improved mastication characteristics [20]. Comparing citrus varieties with distinct mastication traits reveals significant variations in cellulose content [21]. Efficient cellulose degradation in citrus fruit is pivotal for enhancing fruit quality. Exploring the role of GH9 family genes in cellulose degradation will further our understanding of the regulatory mechanisms governing the mastication trait in citrus fruit.

In this study, we aim to analyze the GH9 family genes in the citrus genome, examine the cellulose content in the fruit, and clarify their roles in cellulose metabolism during fruit development. Overall, the investigation of cellulose-degrading GH9 family genes in citrus fruits will provide insights into GH9 genes contribute to cellulose metabolism and their impact on the mastication traits of citrus fruits.

Result

Identification of glycosyl hydrolase family 9 (GH9) members in Citrus sinensis

Through genomic analysis, we identified 32 GH9s from Citrus sinensis, in comparison to the previously reported 25 GH9s in Arabidopsis thaliana and Oryza sativa [22, 23]. CsGH9 proteins were defined and verified using CDD and SMART, chosen for their established relevance in characterizing GH9 proteins. This classification includes 7 Type-A and 25 Type-B CsGH9 proteins. Basic information on CsGH9 genes, proteins in CsGH9 range in length from 110 to 792 amino acids, have an estimated molecular weight range of 11,789.15 to 86,881.3 Da, and have a theoretical pI between 4.57 and 9.84. The results of CsGH9s reported in this paper are shown in Table S1.

Phylogenetic analysis, chromosomal localization, inter-genomic relationships, and covariate analysis of CsGH9 genes in Citrus sinensis

To investigate the evolutionary relationship among different GH9 gene family members among citrus, rice, and Arabidopsis, the full-length protein sequences were used to generate a phylogenetic tree constructed using the Maximum Likelihood Method with MEGA11 (Fig. 1A). 32 genes in the citrus genome were assigned names CsGH9A1 through CsGH9A7 and CsGH9B1 through CsGH9B25 based on their clustering relationship with GH9 genes in Arabidopsis and rice. Further, the GH9s gene exhibited uneven distribution across various chromosomes in the citrus genome, except chr7 (Fig. 1B), suggesting potential genomic dynamics or selective pressures. Among them, the observed collinear relationship between CsGH9B22 and CsGH9B17 implies a conserved gene structure and potentially similar biological functions, underscoring the evolutionary link between these two gene groups.

Fig. 1

Phylogenetic relationships among GH9 genes in Oryza sativa, Arabidopsis thaliana, and Citrus sinensis (A), and chromosomal localization, gene duplication, and colinearity analyses of CsPME genes (B). Numbers at each branch node indicate bootstrap probabilities (> 80% based on 1000 bootstrap replicates). Proteins from different species are distinguished by varying colors. The green curved lines depict collinear blocks within the entire citrus genome

Structural and motif identification of CsGH9 genes in Citrus sinensis

In order to deepen our comprehension of the structural conservation and divergence of GH9 genes across evolutionary epochs, we acquired the UTR-CDS arrangement within the coding sequences of individual GH9 genes for each group’s members (Fig. 2A). Moreover, eight distinct motifs were identified within the CsGH9s. While the majority of genes exhibited all eight motifs, certain genes displayed a reduced motif count (Fig. 2B, C). The motifs varied in size, with motif 8 measuring 21 amino acids, and motifs 2, 3, 6, and 7 spanning 50 amino acids. Furthermore, members closely related within the GH9 clade demonstrated comparable conserved motifs. When the motifs were sorted by their frequency within the GH9 families, it became evident that Motif 2 and Motif 5 were the most prevalent among genes, appearing in 26 genes, whereas Motif 7 was the least common, being found in only 12 genes (Fig. 2B, C).

Fig. 2

Structural features (A), motif identification (B), and conserved amino acids (C) of CsGH9 genes. The coding sequence sizes (CDS) and untranslated regions (UTR) are highlighted in yellow and green boxes, respectively. The prediction identified eight conserved motifs within CsGH9 proteins, with each small box denoting a specific motif. The scale bar represents the length of deduced protein sequences, maintaining proportionality

Promoter characteristics of CsGH9 genes in Citrus sinensis

To elucidate the regulatory functions of the promoters of the CsGH9 genes in Citrus sinensis, we extracted the 2000 bp upstream promoter sequences. As illustrated in Fig. 3A, these sequences contain numerous cis-acting elements associated with hormone responsiveness and environmental stress response. Our analysis of hormone-related cis-acting elements revealed that all CsGH9 genes include at least one element responsive to phytohormones. Regarding abiotic stress-related elements, we identified several key motifs: MBS and TCCC-motif for drought response, STRE and TC-rich repeats for defense/stress response, WUN-motif and WRE3 for wound response, and LTR for cold response. Among these, MBS is the most widely distributed cis-element within the GH9 gene family, whereas LTR elements are present in 11 members of the GH9 family (Fig. 3A). Furthermore, we characterized the expression response of GH9 family genes to various plant hormones (ABA, IAA, SA, and GA) and different environmental stresses (drought, defense/stress, wound, and cold). The results indicate differential gene expression responses to these factors. Notably, CsGH9B5 exhibited a significant induction (3-fold) by IAA, while CsGH9B3 and CsGH9B19 were significantly induced (5-fold and 2-fold, respectively) by SA. Conversely, CsGH9A6 was inhibited (2-fold) by cold stress, whereas CsGH9B22 showed a substantial induction (4-fold) under the same condition (Fig. 3B).

Fig. 3

Analysis of cis-acting regulatory elements (A) and their response (B) in the promoters of CsGH9 genes in Citrus sinensis. A 2000 bp nucleotide sequence upstream of the initiation codon was considered for each CsGH9 gene to predict cis-regulatory elements

Bioinformatic analysis for the GH9 family members from different citrus types

To understand the variations within the GH9 gene family across diverse citrus varieties, we screened 23, 24, and 26 GH9 genes in Citrus grandis, Citrus clementina, and Citrus reticulata, respectively (Table S1). The distribution of the GH9 gene across chromosomes was found to be uneven among different citrus cultivars (Fig. S1). The subcellular localizations of GH9 proteins, which are closely related to their functions, were predicted, with most GH9 proteins localized to the cytoskeleton and chloroplast (Table S2). A phylogenetic tree was constructed to explore the evolutionary relationships among the different members of the GH9 gene family in these citrus types (Fig. 4A). In this analysis, we identified 11, 12, and 10 genes exhibiting collinear relationships with Citrus sinensis, respectively (Fig. 4B; Table S3). We calculated the Ka/Ks values of gene pairs with replication relationships, which ranged from 0 to 0.7 across different citrus varieties, all values being less than 1 (Table S4). This indicates that these gene pairs underwent strong purifying selection. We also compared cis-elements in the promoter regions of GH9 genes from different citrus varieties in response to plant hormones and environmental stresses. The results revealed similar compositions and quantities of cis-elements. Among plant hormones, ABA and JA elicited the most pronounced responses, while for environmental stresses, drought and cold stresses triggered the most significant responses (Fig. 4C). Our motif identification and structural analysis showed a high degree of similarity among these genes (Fig. S2).

These findings suggest that while there are distinct differences in the GH9 gene family among various citrus varieties, most characteristics exhibit remarkable similarity.

Fig. 4

Differentiation of the GH9 family in diverse citrus varieties. Phylogenetic relationships (A), colinearity (B), and promoter analyses (C). Numbers at each branch node indicate bootstrap probabilities (> 80% based on 1000 bootstrap replicates)

Spatiotemporal expression patterns of CsGH9 genes in citrus fruits

The expression patterns of CsGH9 genes in citrus fruits were investigated in a spatial-temporal context using global gene expression transcriptome data. This data encompassed four distinct fruit tissues (EP, epicarp; AL, albedo; SM, segment membrane; JS, juice sacs) across six developmental stages (50 DAF, 80 DAF, 120 DAF, 155 DAF, 180 DAF, and 220 DAF) as detailed in our previous study (Feng et al., 2021b). Expressions were considered significant when the FPKM (fragments per kilobase of transcript sequence per million base pairs sequenced) value exceeded 0.3 (Kang et al., 2013), leading to the identification of 10 expressed genes from the 32 CsGH9 gene families. Our focus centered on analyzing the expression characteristics of six highly expressed genes in citrus fruit (Table S5). Figure 5A illustrates that CsGH9 genes predominantly exhibit low, constitutive, and tissue-specific expression patterns across different citrus fruit tissues during various developmental stages. Notably, CsGH9B12’s expression increased consistently in all four fruit tissues, while CsGH9B14 showed a continuous decrease. CsGH9B13 exhibited specific expression in the later stage of SM, and CsGH9A2 was specifically expressed in fruit peel tissue (EP and AL), with expression levels steadily increasing. Additionally, CsGH9A6 and CsGH9B22 displayed variable expression patterns in different tissues, maintaining high expression in the early stages. Upon examining the overall expression of CsGH9 family genes in citrus fruits, distinctions emerged among the four tissues. Figure 4B highlights a decreasing expression trend in EP, while AL, SM, and JS demonstrated elevated expression during the initial stage, followed by a continuous increase throughout fruit development and maturation.

Fig. 5

Spatiotemporal expression of CsGH9 genes (A) and global expression patterns (B) in four fruit tissues at six developmental stages in Citrus sinensis

Four distinct fruit tissues were examined: EP (epicarp), AL (albedo), SM (segment membrane), and JS (juice sacs), spanning six developmental stages from young fruits to ripe fruits (50 days after flower (DAF), 80 DAF, 120 DAF, 155 DAF, 180 DAF, and 220 DAF). The hierarchical clustering of gene sets was determined using one minus Pearson correlation as the distance metric and average as the linkage method, based on the log₂ (FPKM + 1) values.

Investigation of spatiotemporal cellulose accumulation patterns in citrus fruits

Previous studies have established a correlation between cellulose content and the mastication trait in citrus fruits [20]. To investigate this correlation, we collected samples from three distinct fruit tissues (AL, SM, and JS) at six developmental stages (50 DAF, 80 DAF, 120 DAF, 155 DAF, 180 DAF, and 220 DAF) of the ‘Fengjie’ navel orange (FJ). Subsequently, we measured the contents of cellulose, and hemicellulose in these citrus fruits (Fig. 6A, Table S6). The cellulose content exhibited a declining trend in all three fruit tissues with the progression of fruit development, and noteworthy variations were observed among these tissues (Fig. 6B). Specifically, cellulose content in the JS tissue increased during the veraison period (155–180 DAF). The hemicellulose content, illustrated by inverse V-shaped curves, displayed distinctive patterns in the AL and SM tissues. Contrarily, in the JS tissue, hemicellulose content exhibited a lower initial level and gradually decreased with fruit development and maturity (Fig. 6B).

Furthermore, for a more comprehensive understanding of the role of cellulose and hemicellulose in the mastication trait of citrus fruits, we collected samples from four major citrus varieties - red tangerine (Clementine Mandarin, CL), pummelo (HB pomelo, HB), ponkan (Huagan No. 2, HG), and nanfeng tangerine (Nanfeng Mandarin, NF) - during the ripening period (Fig. 6C). As shown in Fig. 6D, the hierarchy of cellulose content across various tissues in three citrus varieties is as follows: AL > SM > JS, with the exception of HB. Conversely, the hemicellulose content follows the order: SM > AL > JS, except for HG. Similarly, the JS tissue consistently exhibited the lowest content in both cellulose and hemicellulose among the five varieties.

Fig. 6

Cellulose accumulation patterns in citrus fruits. (A) and four major citrus varieties including red tangerine (CL), pummelo (HB), ponkan (HG), and nanfeng tangerine (NF) during the ripening period (C), collected for cellulose determination. Scale bar = 3 cm. The content of cellulose in C. sinensis fruit (B) and four major citrus varieties fruit (D). Each sample was characterized based on six replicates. Significance was determined by one-way ANOVA in SPSS software, with bars labeled with distinct lowercase letters indicating statistically significant differences (p < 0.05)

Identification of key genes involved in cellulose metabolism within the GH9 family of citrus fruits

The significance of the CsGH9 family of cellulases in cellulose metabolism has been underscored as pivotal [5]. Our investigation delved into the expression patterns across stages and tissue of CsGH9 genes in citrus fruits, extending our focus to their expression in diverse tissues of different citrus varieties during the ripening stage. As shown in Fig. 7, the expression profiles of the six genes exhibited notable variations among different varieties, with a predominant higher expression observed in AL and SM compared to JS. Among the CsGH9 family members, the expression levels of CsGH9A6, CsGH9A2, and CsGH9B22 genes in AL across different varieties were significantly higher than those in the pulp tissues (SM and JS). Furthermore, the expression of CsGH9B12 and CsGH9B13 in SM surpassed that in other tissues, signifying distinct roles within the cellulose metabolism pathway. Finally, we systematically analyzed the correlations between the expression levels of GH9 family genes and cellulose content across various developmental stages, fruit tissues, and citrus varieties. Our findings, detailed in Table S7, reveal a significant correlation between the expression of these six GH9 genes and cellulose accumulation in the fruits. This suggests a potential role for these genes in the regulation of cellulose biosynthesis during fruit development.

Fig. 7

Expression profiles of CsGH9 genes analyzed by qRT-PCR

The endogenous reference gene utilized in this study was CsActin, and the experimental setup comprised four biological replicates. Significance was determined by one-way ANOVA in SPSS software, with bars labeled with distinct lowercase letters indicating statistically significant differences (p < 0.05). Citrus varieties examined in the study include navel orange (‘Fengjie 72 − 1’ navel orange, FJ), red tangerine (Clementine Mandarin, CL), pummelo (HB pomelo, HB), ponkan (Huagan No. 2, HG), and nanfeng tangerine (Nanfeng Mandarin, NF).

Discussion

Citrus fruits, being globally cultivated, exhibit a distinct mastication trait—a pivotal factor in determining fruit quality. Clarifying the regulatory mechanisms governing the mastication trait in citrus fruits is essential for advancing fruit quality. The current research in this domain is at an early stage, encountering several theoretical and technical challenges demanding thorough exploration. This study contributes valuable insights to the theoretical understanding of superior mastication trait fruit varieties, drawing from the comprehensive analysis of CsGH9 gene family expression and cellulose content in citrus fruits.

Insights into the role of cellulose in shaping the mastication trait of citrus fruit

The dynamic alterations in cell wall composition throughout fruit ripening play a pivotal role in shaping various horticultural traits [24]. Among the cell wall’s primary constituents, including pectin, cellulose, and lignin, cellulose has garnered significant attention due to its potential impact on the mastication properties of citrus fruits [17, 18]. While prior research has primarily concentrated on the overall accumulation of cell wall metabolites in citrus fruits, the specific role of cellulose in relation to mastication is gaining increasing recognition [25, 26]. Our study builds upon previous findings and underscores the importance of cellulose, particularly its spatial and temporal accumulation patterns, in determining the texture and overall quality of citrus fruits. A comparative analysis between the FJ72-1 navel orange and its bud mutant FJWC, which exhibits inferior mastication properties, revealed a higher content of cellulose and hemicellulose in FJWC during fruit ripening [25]. Extending these observations, our current research delves deeper into the spatiotemporal dynamics of cellulose accumulation across different fruit tissues, developmental stages, and citrus varieties. The findings indicate a gradual decrease in cellulose content during fruit development, with notable variations among the albedo (AL), segment membrane (SM), and juice sac (JS) tissues (Fig. 6B). Furthermore, a comparative analysis of cellulose content across different citrus varieties during maturity reveals a consistent pattern, with AL exhibiting the highest cellulose content and JS the lowest (Fig. 6D). Significantly, the HB pomelo, standing as the sole inedible segment membrane (SM) variety, exhibits the highest cellulose content in its SM, a noteworthy finding. This observation underscores the crucial role of cellulose as a determinant of the mastication properties in citrus fruits. The observed patterns of cellulose distribution and accumulation offer profound insights into its influence on fruit texture. To further elucidate the molecular mechanisms and regulatory pathways governing cellulose dynamics and their impact on citrus fruit quality, additional in-depth research is imperative.

Identification of key genes involved in cellulose metabolism of GH9 family of citrus fruit

Previously, research into cellulose and its degrading enzymes in horticultural crops has primarily concentrated on their influence on fruit softening, especially in strawberry [27], tomato [28], and avocado [29]. However, recent studies on Psidium guajava have revealed that the endoglucanase of the GH9 family plays a pivotal role in modulating cellulose content during guava fruit development [30]. This discovery suggests that identifying key genes within the GH9 family in citrus fruits may provide crucial insights into the molecular mechanisms underlying the mastication trait. In our investigation of the citrus genome, we successfully identified 32 CsGH9 family genes (Table S1). Notably, our study highlights specific genes within the GH9 family that may play a vital role in cellulose metabolism in citrus fruits. For instance, CsGH9A6 exhibited significantly higher expression levels in fruits compared to other GH9 family genes. Specifically, CsGH9A6 maintained high expression levels in SM and displayed a declining trend throughout fruit development (Fig. 5A). Remarkably, the expression of CsGH9A6 in the AL of mature citrus fruits surpassed that in other tissues (Fig. 7). These findings indicate that CsGH9A6 may be a key regulator of cellulose metabolism in citrus fruits, which is essential for maintaining fruit texture and quality.

To elucidate potential regulatory mechanisms, we conducted an in-depth analysis of the cis-acting elements located in the promoter regions of the candidate genes. This comprehensive analysis uncovered several potential regulatory motifs that may contribute to the observed patterns of upregulation and downregulation. Notably, the majority of the GH9 family gene promoters harbor at least one plant hormone responsive element (Fig. 3A), highlighting the pivotal role of plant hormones in regulating the expression of the GH9 family genes. Our previous research on Valencia oranges, which demonstrated that low temperatures can trigger lignin accumulation in the fruit, resulting in compromised mastication qualities [19]. Thus, the analysis identified the presence of low-temperature responsive elements in the promoter regions of CsGH9B12/13/14 (Fig. 3A). In contrast, CsGH9B22 was upregulated 4-fold under low-temperature conditions, despite the absence of low-temperature responsive cis-acting elements in its promoter region (Fig. 3). This discrepancy suggests that CsGH9B22 may be regulated through indirect mechanisms or influenced by chromatin structure and epigenetic modifications. Further investigation and verification are required to elucidate these regulatory pathways. Although these intriguing findings have piqued our interest, the precise functional roles of these genes in cellulose metabolism remain obscure. A more rigorous exploration of these genes is imperative to gain a comprehensive understanding of their influence on the textural characteristics of citrus fruits, specifically with regards to the mastication trait.

Conclusion

In conclusion, our study has identified a total of 32 CsGH9 genes in Citrus sinensis, and a comprehensive analysis of these genes has been conducted, providing valuable insights for further exploration of CsGH9 gene functions. Notably, cellulose has emerged as a pivotal factor in shaping the mastication trait of citrus fruits, with specific emphasis on key genes such as CsGH9A2/6, CsGH9B12/13/22, etc. These findings shed light on the intricate mechanisms underlying the regulation of the citrus fruit mastication trait.

Materials and methods

Plant materials, sample collection, and tissue isolation

Fruit samples of ‘Fengjie 72 − 1’ navel orange (Citrus sinensis L. Osbeck) were meticulously harvested at distinct developmental stages: 50, 80, 120, 150, 180, and 220 days after flowering (DAF). Additionally, mature stage fruit samples from Huagan No. 2 (Citrus reticulata Blanco, PK), HB pomelo (Citrus grandis L. Osbeck), Clementine Mandarin (Citrus clementina hort. ex Tanaka), and Nanfeng Mandarin (Citrus reticulata Blanco cv. Nanfeng tangerine) were collected. For each developmental stage, twelve representative fruits were carefully sampled, ensuring robust biological replicates (three trees per replicate). Employing precise manual dissection techniques, four distinct fruit tissues (Epicarp, Albedo, Segment membrane, and Juice sacs) were isolated and promptly frozen in liquid nitrogen at -80 °C to preserve their biological integrity.

Identification of GH9 gene family members from Citrus sinensis

To systematically identify members of the CsGH9 gene family, two complementary BLAST approaches were employed. Initially, the sweet orange (Citrus sinensis, v1.0) genome was queried using the GH9 Hidden Markov Model (HMM) in the HMMER software (version 3.0, http://hmmer.org/, E-value = 1e^− 20) [31]. The GH9 domain (PF00759), essential for cellulose hydrolysis, was sourced from the Pfam database (http://pfam.xfam.org/, version 33.1) and employed in the search [32]. Following this, a local BLAST analysis was conducted by comparing 25 AtGH9 protein sequences from Arabidopsis thaliana, obtained from the TAIR database (http://www.Arabidopsis.org/), against the Citrus sinensis genome (E-value = 1e^− 5).

To confirm the validity of candidate CsGH9 family genes, each sequence was subjected to verification through the Simple Modular Architecture Research Tool (SMART, http://smart.embl-heidelberg.de/) and the Conserved Domain Database (CCD, http://www.ncbi.nlm.nih.gov/cdd/). This ensured the presence of a GH9 domain, a characteristic feature of GH9 genes [33, 34]. Lastly, the physical and chemical properties of the identified CsGH9 family proteins were analyzed using ExPASy (https: //web.expasy.org/protparam/) [35]. Subcellular localization predictions were performed using Cell-PLoc (http://www.csbio.sjtu.edu.cn/bioinf/Cell-PLoc-2/) subsequent to gene identification [36].

Phylogenetic analysis, chromosome localization, gene density, and colinear analysis

Phylogenetic analysis aimed to unravel the evolutionary relationships among GH9s utilizing the Maximum Likelihood Method, employing MEGA software. The protein sequences of GH9s from citrus (http://citrus.hzau.edu.cn/index.php), Arabidopsis (https://phytozome-next.jgi.doe.gov/info/Athaliana_TAIR10), and rice (https://phytozome-next.jgi.doe.gov/info/Osativa_v7_0) were utilized for this purpose. Phylogenetic tree was constructed with 1000 bootstraps, and the final visualization and refinement were conducted using tvBOT (Tree Visualization By One Table, https://www.chiplot.online/#) [37]. Chromosome localization, gene density, and colinear analysis were subsequently executed utilizing TBtools software (v1.098769, Advanced Circos) [38].

Gene promoter, motif, domain, and structure analysis

The promoter sequence of CsGH9s was obtained by cloning the 2 kb upstream region and subsequently analyzed using the PlantCARE online tool (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/) [39]. To identify conservation motifs, the Multiple Em for Motif Elicitation (MEME) online tool (MEME, version 5.1.1, http://meme-suite.org/tools/meme) was employed [40]. The translated amino acid sequences of each CsGH9s were subjected to conservation domain identification using the NCBI CD Search tool [41]. Finally, a comprehensive illustration depicting the gene promoter, motifs, domains, and structures of GH9 family genes was created using the TBtools software [38].

Analysis of gene spatial-temporal expression

Transcriptomic data encompassing 72 samples from four distinct tissues across six developmental stages of sweet orange (Citrus sinensis L. Osbeck) were meticulously curated from our prior investigation [42]. The raw transcriptomic data for response analysis were downloaded from the NCBI Sequence Read Archive (SRA) (accession number: SRP325502, SRP155673, SRP345972, SRP404051, SRP483818, and ERP022936), and the analysis was based on previous publication [43]. To unveil the intricate patterns within the gene sets, TBtools, a bioinformatics toolkit, was employed. Hierarchical clustering was executed utilizing the one minus Pearson correlation as the distance metric, coupled with the average linkage method. The gene expression values were processed using log₂ (FPKM + 1) [38]. For a comprehensive visualization of the gene expression patterns, a boxplot was meticulously constructed. ChiPlot, a robust tool for visualizing RNA-seq data, was utilized for generating the boxplot, employing log₂ (FPKM + 1) as the standardized metric [37].

Determination of cellulose and hemicellulose contents

The quantification of cellulose and hemicellulose concentrations was accomplished through the application of their respective absorbance values, employing the standard curve method. Approximately 5.0 g of citrus flesh tissue underwent dehydration to a constant weight at 80 °C and was subsequently sieved through a 40-mesh sieve. The determination of total cellulose and hemicellulose content adhered to the established procedures outlined in the Plant Cellulose Content Kit (COMIN, CLL-2-Y) and the Plant Hemicellulose Content Kit (COMIN, BXW-2-G).

Quantitative analysis of gene expression

Total RNA extraction was conducted following established protocols [43]. Relative gene expression was assessed in four biological replicates using the QuantStudio™ 6 Flex real-time PCR system (Applied Biosystems), according to the manual of PowerUp™ SYBR™ Green Master Mix (A25742, Thermo Scientific). The housekeeping gene CsActin and CsGAPDH served as an endogenous reference [44, 45]. Relative abundance of GH9 genes was calculated with the 2^−ΔΔCt method. Detailed information on the primers employed for each gene in this study can be found in Supplementary Table S8.