Genome-Wide Identification and Characterization of the SWEET Gene Family in Phoebe bournei with an Emphasis on Hormonal Responses and Plant Physiological Changes

Abstract

The Sugars Will Eventually be Exported Transporters (SWEET) family plays a crucial role in the carbohydrate distribution, phloem loading, and stress response of plants, yet the evolutionary characteristics and functional diversification of SWEET genes in the endangered timber species Phoebe bournei (Hemsl.) Yen C. Yang remain largely unexplored. In this study, 21 PbSWEET genes were identified and classified into four subfamilies (A–D). Subfamily A exhibited a unique lineage expansion, mainly driven by tandem and segmental duplications. The nonsynonymous-to-synonymous substitution ratio (Ka/Ks) values of all duplicate gene pairs were all less than 1, indicating a strong selective suppression effect; consistent with this evolutionary constraint, the majority of PbSWEET proteins harbor the conserved Medicago truncatula Nodulin 3/saliva (MtN3_slv) domain, with only a few exceptions lacking a complete version. Promoter and hormone response analyses revealed that under abscisic acid (ABA) stress, PbSWEET4 exhibited an immediate burst, whereas PbSWEET10 showed a delayed burst. Physiological data indicated that soluble sugars may be more dominant osmolytes than proline (Pro), a pattern that points to a potential carbon-centric regulatory strategy. PbSWEET4 showed an early burst before sugar/oxidative peaks, suggesting a possible non-canonical signaling role, whereas PbSWEET10 exhibited a late increase coinciding with sugar/malondialdehyde (MDA) peaks, suggesting potential involvement in sugar redistribution. Under methyl jasmonate (MeJA) treatment, PbSWEET10 was rapidly induced, yet sugar accumulation occurred only at 24 h, a temporal decoupling that suggests a possible transcription–metabolism decoupling. Collectively, these correlative patterns point to a possible dual-wave transcriptional mechanism and nominate PbSWEET10 as a candidate for stress response, though these inferences require functional validation.

1. Introduction

Sugars serve as the energy foundation and structural components of plant life activities, and their transmembrane transport is crucial for cellular functions, growth, development, and environmental adaptation [1,2,3,4,5]. In plants, sugar allocation relies on the precise regulation of a series of sugar transporters, which mediate long-distance transport between source and sink organs as well as intracellular compartmentalization, thereby supporting key physiological processes, including photoassimilate partitioning, seed filling, and fruit development [1,6,7]. In recent years, the Sugars Will Eventually be Exported Transporters (SWEET) family, a new type of transport sugar protein, has received considerable attention because of its unique role. They belong to the Medicago truncatula Nodulin 3/saliva (MtN3_slv) superfamily. This class of proteins typically contains two conserved M. truncatula Nodulin 3 (MtN3) domains, forming a membrane topology with seven transmembrane helices, and these are low-affinity, high-capacity bidirectional transport proteins that help sugars pass through the cell membrane [8]. SWEET proteins play significant roles in phloem loading, nectar secretion, nutrient supply to pollen tubes, and seed development, and have also been suggested to participate in cytokinin transport [7,8,9]. Particularly important is the pivotal role of the SWEET family in plant interactions with biotic and abiotic environments. Research has demonstrated that SWEET-mediated sugar outflow is a crucial interface in carbon allocation and plays a major part in plant–microbe interactions; for example, pathogenic microbes can use them to obtain host sugars and alter plant susceptibility or resistance [10,11,12,13]. Meanwhile, by controlling the buildup and distribution of soluble sugars, SWEET members actively participate in plant responses to environmental challenges, including drought and low temperature, to assist in the maintenance of cellular osmotic balance and energy homeostasis [14,15]. Therefore, the evolutionary characteristics and functional changes of the SWEET gene family can help us better understand how plants transport sugar and the molecular mechanism of sugar transport in the plant’s adaptation environment [1,7,8].

Growing evidence confirms the integral role of these transporters in key physiological processes, including plant growth and development, host–pathogen interactions, and responses to abiotic stresses [10,16,17]. For example, in the classical field of sugar transport, in Oryza sativa L. (rice), OsSWEET11 and OsSWEET15 have been confirmed as key regulatory factors in rice grain filling, mediating efficient sucrose unloading from maternal tissues to the endosperm via the apoplastic pathway, thereby ensuring normal seed development [18].

In addition to these well-established roles in sugar allocation, an increasing number of SWEET proteins have been found to transport non-sugar substrates, including phytohormones. In Arabidopsis thaliana (L.) Heynh, AtSWEET13 and AtSWEET14 have been discovered to transport gibberellic acid (GA), and they can also be used to help GA-mediated processes such as anther dehiscence and seed development [19,20]. This pioneering discovery reveals a new role of SWEET proteins as hormone transport carriers, greatly expanding the boundaries of their functional research. Furthermore, the function of SWEET genes in response to hormonal signals has increasingly attracted attention. Among monocotyledonous herbaceous plants, studies on maize (Zea mays L.) and Lilium ‘Sorbonne’ (Oriental hybrid lily) have revealed the effect of the precise regulation of abscisic acid (ABA) signaling on the expression of SWEET genes: In maize, members of the maize SWEET gene family are rich in ABA-responsive elements, their expression being induced by exogenous ABA, and genes such as ZmSWEET1a respond simultaneously to drought, saline–alkaline stress and ABA treatment, suggesting that they may participate in abiotic stress responses through ABA-dependent signaling pathways [21]; LoSWEET14 in L. ‘Sorbonne’ reveals a direct regulatory pathway from hormone signaling to gene expression, where the transcription factor LoABF2 can directly bind to the ABA response element in its promoter, activating LoSWEET14 expression and subsequently responding to various abiotic stresses by promoting soluble sugar accumulation [22]. In dicotyledonous herbaceous plants, such as Nicotiana tabacum L., the overexpression of NtSWEET12i can promote the transport of soluble sugars and significantly enhance plant tolerance to drought and saline–alkaline stress by activating the ABA signaling pathway [23]. Similarly, in dicotyledonous herbaceous vine plants, in Passiflora edulis Sims, PeSWEET3 has been proven to have ABA-responsive elements in its promoter, and exogenous ABA treatment can significantly upregulate its expression and promote the accumulation of soluble sugars, directly establishing the link between ABA signaling and SWEET-mediated sugar accumulation [24]. Research on a dicotyledonous woody plant, Camellia oleifera Abel, is more in-depth; CoSWEET2a not only maintains cellular sugar homeostasis to enhance drought resistance but its expression is also significantly induced by both ABA and GA, demonstrating the fine regulation of SWEET function by hormonal signals at the transcriptional level [25]. Furthermore, recent studies have found that ABA-induced seed germination can coordinately regulate the biosynthesis of cuticular wax and cuticle by activating endogenous ABA and jasmonic acid (JA) signaling pathways to enhance drought tolerance. This supports the idea that hormone signaling can reshape carbon allocation under stress, raising the possibility that ABA or methyl jasmonate (MeJA) may directly regulate SWEET expression [26]. In summary, the functions of the SWEET family have expanded from mere sugar transport to active interplay with hormone signaling networks (especially ABA, MeJA, and GA). Some SWEETs directly transport hormones, whereas others serve as downstream effectors of hormone signaling; both roles allow SWEETs to finely modulate plant growth, development, and stress adaptation by controlling sugar distribution and accumulation.

Despite these advances in model plants and major crops, research on the SWEET gene family in forest trees remains remarkably limited. Phoebe bournei (Hemsl.) Yen C. Yang is a near-threatened timber tree species of significant ecological and commercial value in China. Its excellent wood quality and high decay resistance make it popular in forestry and landscape building [27]. However, environmental pressures, particularly frequent droughts and soil microplastic contamination caused by global climate change, have seriously threatened the natural distribution and long-term management of P. bournei plants [28,29]. Recent studies have shown that P. bournei can respond to such stresses through complex physiological and molecular processes; drought stress, for example, significantly reduces photosynthesis, activates antioxidant systems, and modifies hormone signaling pathways such as ABA [29,30,31]. These adaptive responses are frequently accompanied by the accumulation and redistribution of photoassimilates [30], such as soluble sugars, implying that sugar transporters may play an important role in these processes, and their potential role, especially in the ABA and MeJA signaling pathways, has not been explored yet. As a major subfamily of sugar transporters, SWEETs are prime candidates to explore this role. However, the distribution, evolutionary features, and functional roles under stress conditions of the SWEET family remain systematically unexplored in P. bournei.

To close this knowledge gap, this project identified and analyzed the SWEET gene family across the genome in P. bournei. Drawing on the strategies used in previous studies, we conducted a multi-dimensional analysis of the SWEET family, covering its evolutionary dynamics, structural features, and transcriptional responses to hormones [32,33,34,35]. These analyses, together with physiological indicator measurements, are expected to clarify the evolutionary relationships of PbSWEET members and reveal their potential roles in abiotic stress adaptation, including how they may contribute to key physiological processes such as osmotic regulation and oxidative stress mitigation, particularly within hormone-mediated signaling pathways [36]. The results of this study may also provide a theoretical basis for further exploration of the potential role of this family in plant–microbe interactions [37] and sugar allocation [38,39]. This study aims to provide key genetic resources for clarifying the molecular basis of sugar signaling and transport in P. bournei, thereby offering candidate genes and a theoretical foundation for the genetic improvement of stress resistance in forest trees.

2. Results

2.1. Identification and Physicochemical Property Analysis of PbSWEET Genes in P. bournei

P. bournei was found to have 21 SWEET genes, which were renamed PbSWEET1 through PbSWEET21 (

Table 1. Detailed information on 21 PbSWEET genes in P. bournei and their encoded proteins.

Gene Accession	Gene ID	Size/aa 1	MW 2/Da	Theoretical pI	Instability Index	Aliphatic Index	GRAVY	Subcellular Localization
OF15105-RA	PbSWEET1	239	26,550.74	8.77	31.26	118.2	0.819	Plasma membrane
OF18582-RA	PbSWEET2	239	26,550.74	8.77	31.26	118.2	0.819	Plasma membrane
OF29277-RA	PbSWEET3	145	15,589.40	9.60	43.26	102.28	0.21	Chloroplast
OF29278-RA	PbSWEET4	68	7869.20	8.03	39.55	90.29	0.419	Chloroplast
OF29276-RA	PbSWEET5	243	26,574.74	9.02	43.1	110.78	0.675	Plasma membrane
OF00547-RA	PbSWEET6	236	26,111.14	9.54	38.35	121.44	0.894	Vacuole
OF16634-RA	PbSWEET7	238	26,330.59	9.64	43.31	127.69	0.987	Plasma membrane
OF02601-RA	PbSWEET8	172	19,574.05	8.91	39.37	113.37	0.691	Plasma membrane
OF07621-RA	PbSWEET9	196	21,808.12	8.82	39.27	120.31	0.837	Vacuole
OF07626-RA	PbSWEET10	244	27,405.63	9.32	38.09	112.62	0.566	Plasma membrane
OF18660-RA	PbSWEET11	439	50,020.08	8.76	45.61	90.84	−0.178	Plasma membrane
OF24651-RA	PbSWEET12	259	28,611.42	9.34	30.24	127.07	0.816	Plasma membrane
OF20979-RA	PbSWEET13	241	26,761.05	8.77	43.12	116.85	0.85	Chloroplast
OF04050-RA	PbSWEET14	274	30,573.99	7.75	34.43	107.81	0.501	Plasma membrane
OF05885-RA	PbSWEET15	274	30,561.08	7.70	30.05	110.66	0.583	Plasma membrane
OF11473-RA	PbSWEET16	278	31,404.66	9.20	46.45	119.24	0.727	Plasma membrane
OF24996-RA	PbSWEET17	294	33,112.08	8.48	41.87	116.67	0.6	Plasma membrane
OF05961-RA	PbSWEET18	249	27,458.71	9.18	39.37	119.72	0.7	Plasma membrane
OF05963-RA	PbSWEET19	255	27,612.82	7.58	40.63	126	0.844	Plasma membrane
OF05962-RA	PbSWEET20	242	26,327.45	9.20	31.48	122.44	0.766	Plasma membrane
OF05964-RA	PbSWEET21	297	32,377.34	9.64	36.58	114.88	0.354	Plasma membrane

¹ aa: amino acid number; ² MW: molecular weight.

). The average length of the encoded proteins is roughly 244 aa, with all lengths ranging from 68 aa to 439 aa. The molecular weights of the PbSWEET proteins show significant variation in protein size within this family, ranging from 7.87 kDa (PbSWEET4) to 50.02 kDa (PbSWEET11), with an average of roughly 27.104 kDa. With theoretical isoelectric points (pIs) ranging from 7.58 to 9.64, all 21 PbSWEET proteins are categorized as basic proteins. In terms of protein stability, it is predicted that 13 proteins (PbSWEET1, PbSWEET2, PbSWEET4, and others with an instability index < 40) will be stable, while the remaining eight are unstable (instability index > 40), with all values ranging from 30.05 to 46.45. The aliphatic index of the PbSWEET proteins has a mean value of 114.64, and its range is from 90.29 to 127.69, which suggests the proteins generally have strong heat stability. Also, all grand average of hydropathicity (GRAVY) values are positive with the exception of PbSWEET11 (−0.178), with the maximum GRAVY value being 0.987 (PbSWEET7), which means the PbSWEET proteins are hydrophobic for the most part. Meanwhile, subcellular localization was predicted. Most PbSWEET proteins (16) are located in the plasma membrane, supporting their core role in transmembrane sugar transport. Three proteins (PbSWEET3, PbSWEET4, and PbSWEET13) are found in the chloroplast, and two proteins (PbSWEET6 and PbSWEET9) are located in the vacuole.

2.2. Chromosome Localization and Evolutionary Analysis of PbSWEET Genes in P. bournei

To ascertain the chromosomal locations of 21 PbSWEET genes in P. bournei, we conducted a location map that shows they were located unevenly across 10 chromosomes (Figure 1). We found that chromosome (Chr) 08 contains the largest number of this gene family (a total of five), and these genes are arranged closely together. Similarly, three genes on Chr04 were also densely distributed. The clustered distribution pattern of these genes conforms to the typical pattern of tandem repeats. Six out of the ten chromosomes carried only one PbSWEET gene each, and this single-gene distribution pattern could conform to the characteristics of whole-genome or large-scale replication events. Chr04, Chr06, and Chr12 harbored four, three, and three PbSWEET genes, respectively. On these chromosomes, the genes were found in both tight clusters and as single, scattered genes. This coupling distribution pattern also conforms to the characteristic of multiple evolutionary models acting simultaneously.

Furthermore, we constructed an evolutionary tree using 21 proteins from P. bournei, 17 from A. thaliana, and 21 from O. sativa, totaling 59 proteins (Figure 2). These proteins were grouped into four subfamilies, named A through D. Subfamilies A, B, and C had an approximately similar number of members, while the D subfamily is relatively small, containing only seven members of the SWEET protein. As can be seen from the figure, among the PbSWEET genes, subfamily A was the largest, containing 11 members, accounting for 52.38% of all PbSWEET genes; subfamily B was the smallest, containing only two members, PbSWEET12 and PbSWEET13, accounting for 9.52%. This pattern shows the PbSWEET family expanded in an unbalanced way during evolution; subfamily A appears to be the core amplified group, while subfamily B seems to have followed a more conserved evolutionary path. Moreover, the figure also shows that the number of SWEET genes in different species varies significantly across subfamilies.

2.3. Gene Structure, Conserved Motifs, and Conserved Domain Analysis of PbSWEET Genes in P. bournei

To clarify the structural diversity and evolutionary conservation of the PbSWEET gene family, we analyzed the exon–intron structures of 21 PbSWEET genes (Figure 3D). This analysis showed great variety in gene length and implied great diversity in the arrangement configuration of the exons and introns, simultaneously reflecting the evolutionary flexibility of this gene family. Nearly half of the genes (PbSWEET1, PbSWEET2, PbSWEET3, PbSWEET4 and PbSWEET9) in the A subfamily lacked untranslated regions (UTRs), perhaps suggesting a simpler regulatory network in this subfamily’s work. Notably, PbSWEET4 lacks UTRs but encodes a truncated protein (68 aa) that is unlikely to function as a canonical sugar transporter; its biological significance remains to be determined. The B subfamily contained only two genes (PbSWEET12 and PbSWEET13), both of which showed a classic structure with dispersed exons, which suggests they evolved with high conservation and their functions are also likely very similar. Additionally, the presence of UTRs in these genes could indicate that they can reach precise regulation in work. The C and D subfamilies both showed notable differences in their internal structures, demonstrating that they both have a pattern of separating evolution in their internal structures, and such divergence supports the idea that they have developed diverse functions.

Furthermore, we examined 10 conserved motifs in the 21 PbSWEET proteins (Figure 3B). Most members shared similar sets and orders of these motifs, strongly indicating there could be evolutionary conservation and consistent function within the family. However, Motif 7 was found only in the C subfamily. This exclusive presence could suggest a specialized role for this subfamily. Analysis of protein domains showed clear patterns (Figure 3C). The A subfamily primarily contained domains from the MtN3_slv superfamily, while the B, C, and D subfamilies mainly consist of MtN3_slv domains. Detailed domain coordinates and E-values for each protein are provided (Table S1).

It is worth noting that among the 21 PbSWEET proteins, PbSWEET4 is particularly short, containing only 68 amino acid residues (Table 1). This length is much shorter than the typical length of the SWEET protein. To assess whether this short peptide can fold into a functional transporter protein, we performed homology modeling (Figure S1). The obtained model shows a local structural similarity (with an identity of 48.33%), but the coverage is extremely low (approximately 20–30%), and there are discontinuous alignments, making it impossible to construct a complete MtN3_slv domain or a typical seven-span membrane topology structure. These observations indicate that although PbSWEET4 belongs to the MtN3_slv superfamily based on sequence analysis, it is unlikely to function as an independent sugar transporter.

2.4. Cis-Acting Elements Analysis of PbSWEET Genes in P. bournei

In order to explore the potential regulatory mechanisms underlying the expression and function of the PbSWEET genes, we predicted and analyzed the region that is located 2000 bp upstream of each PbSWEET gene, identifying a total of 26 types of cis-acting elements (Figure 4). These elements are related to mechanisms such as light response, hormone signal transduction, stress adaptation, and developmental regulation. In the category of light response, the abundance of G-Box (CACGTG core motif) elements is the highest, accounting for 0.28; in terms of hormone regulation, the ABA-responsive element (ABRE) is the most abundant, with an abundance of 0.37, and the study also found two MeJA response motifs (CGTCA and TGACG), with a total abundance of 0.42; among the stress-related elements, the anaerobic induction element (ARE) sequence involved in anaerobic reactions is the most significant, with an abundance of 0.55; in the developmental regulatory elements, the abundance of the CAT-box motif, which is related to the activity of the meristem, is the highest, at 0.27.

Among different element categories, photoreceptor elements are the most numerous, followed by ABA elements and MeJA elements. The analysis could suggest the gene family is primarily involved in light response and also plays a key role in hormone signaling. Notably, some elements (seed-specific and endosperm expression elements) had very low abundance.

This predictive distribution could suggest that hormone signal transduction and stress response are the most critical mechanisms driving the expression regulation of this gene family.

2.5. Intraspecific Collinearity Analysis and Pairwise Evolutionary Selection Pressure Analysis of PbSWEET Genes in P. bournei

To investigate the duplication events contributing to maintenance and evolutionary optimization of the PbSWEET gene family, we analyzed duplication events in the PbSWEET genes of P. bournei (Figure 5). We specifically identified four pairs of segmental duplication events, occurring between Chr01 and Chr03, Chr05 and Chr10, Chr05 and Chr12, and within Chr06. The finding could indicate that segmental duplication serves as a significant driving force for the emergence of new genes and the functional evolution of this gene family.

To further investigate the evolutionary constraints imposed on these segment-copying genes, we conducted a pairwise nonsynonymous-to-synonymous substitution ratio (Ka/Ks) analysis on them (

Table 2. Pairwise evolutionary selection pressure analysis of PbSWEET genes.

Sequence 1	Sequence 2	Ka 1	Ks 2	Ka/Ks 3	Effective Length	S-Sites 4	N-Sites 5	cN 6	cS 7	pN 8	pS 9
PbSWEET1	PbSWEET2	0	0	NaN 10	717	175	542	0	0	0	0
PbSWEET6	PbSWEET8	0.27	1.03	0.261	516	124	392	88.75	69.25	0.226	0.560
PbSWEET6	PbSWEET7	0.11	0.43	0.259	705	175	530	54.75	57.25	0.103	0.327
PbSWEET17	PbSWEET16	0.25	0.82	0.307	828	196	632	135.17	97.83	0.214	0.499

¹ Ka: nonsynonymous substitution rate; ² Ks: synonymous substitution rate; ³ Ka/Ks: nonsynonymous-to-synonymous substitution ratio; ⁴ S-sites: average number of synonymous sites; ⁵ N-sites: average number of nonsynonymous sites; ⁶ cN: corrected number of nonsynonymous substitutions; ⁷ cS: corrected number of synonymous substitutions; ⁸ pN: proportion of nonsynonymous differences; ⁹ pS: proportion of synonymous differences; ¹⁰ NaN: not a number.

). All examined gene pairs had a Ka/Ks ratio that was less than 1, which conforms to the evolutionary characteristics of purifying selection. For example, the ratio for PbSWEET6/PbSWEET8 was 0.261, the ratio for PbSWEET6/PbSWEET7 was 0.259, and the ratio for PbSWEET17/PbSWEET16 was 0.307. Notably, although the nonsynonymous substitution rate (Ka) and synonymous substitution rate (Ks) values of PbSWEET1 and PbSWEET2 are both zero, which might imply genome annotation redundancy, the chromosome localization results clearly show that these two genes are located at different positions (Figure 1), thereby confirming that they are two independent gene duplication copies rather than a redundant annotation. A likely explanation is that these two genes originated from a very recent gene duplication event and have not yet accumulated any nucleotide substitutions.

2.6. Interspecific Collinearity Analysis of SWEET Genes in P. bournei, A. thaliana, and O. sativa

To investigate the evolutionary relationships of SWEET genes among different species, we compared SWEET genes from P. bournei with those from A. thaliana (typical dicot) and O. sativa (representative monocot) (Figure 6). The results showed that seven PbSWEET genes showed considerable collinearity with genes in O. sativa, while four PbSWEET genes showed less collinearity with genes in A. thaliana. We also revealed specific numbers of collinear gene pairs, namely, 10 collinear gene pairs between P. bournei and O. sativa and five collinear gene pairs between P. bournei and A. thaliana.

The above results could imply that the SWEET genes in P. bournei and O. sativa share higher structural and historical similarity, but only within a certain level of confidence. The above analysis is reasonable because the level of collinearity is not directly equivalent to the phylogenetic relationship or the genetic evolutionary distance; the phenomena we have observed may be influenced by a variety of factors, including genomic rearrangement, gene loss, and gene duplication events [40].

2.7. Tissue-Specific Expression of PbSWEET Genes in P. bournei

To reveal the expression specificity of the PbSWEET gene family in different tissues of P. bournei and to explore its potential division of labor in the process of sugar transport and distribution, we detected the expression levels of the members of this family in five major tissues: root bark, stem bark, root xylem, stem xylem, and leaves (Figure 7). Most genes were underexpressed or not expressed in all five tissues. The A subfamily had multiple genes that showed highly tissue-specific expression patterns; for example, PbSWEET4 and PbSWEET3 had very high expression in plant xylem; meanwhile, some genes in the A subfamily showed moderate expression; PbSWEET7 and PbSWEET6 were expressed across multiple tissues. In subfamily B, PbSWEET12 was mainly expressed in the root bark, while PbSWEET13 was underexpressed or not expressed in any tissue. The C subfamily showed weak expression or no expression in all tested tissues, which implies strict regulatory constraints or condition-specific induction. In contrast, the D subfamily had a more diverse expression profile.

2.8. GO Enrichment Analysis and Prediction of PPI Network of PbSWEET Genes and Their Encoded Proteins in P. bournei

To further clarify the biological functions and molecular roles of the PbSWEET gene family, we conducted a Gene Ontology (GO) enrichment analysis (Figure 8). The results showed this family is mainly linked to carbohydrate transport across membranes, of which the strongest signals were for carbohydrate transmembrane transporter activity, whose term had a false discovery rate (FDR) value of 1.0 × 10⁻²⁷. Another strong signal was for sugar transmembrane transporter activity with an FDR of 1.0 × 10⁻²². Meanwhile, some genes were also enriched for glycosyltransferase activity, namely beta-1,4-mannosylglycoprotein 4-beta-N-acetylglucosaminyltransferase activity with an FDR of 1.0 × 10⁻¹⁷, which could imply a potential role in modifying glycoproteins. Additionally, we also identified functions related to proton-coupled sugar transport, and these include sucrose–proton cotransport activity and hexose transmembrane transporter activity, which match the already known functions of the SWEET gene family [8,41,42].

After clarifying the functional characteristics of the PbSWEET gene family through GO enrichment analysis, we further explored the regulatory interactions among these proteins using protein–protein interaction (PPI) network analysis (Figure 9).

Network analysis showed that SWEET family members exhibited high connectivity and interactions with a variety of chaperone-related proteins, including glycosyltransferases like AT2G13290, sucrose transporters like SUC2 and SUC4, cytoskeleton-associated components like ARPC2B, and pentatricopeptide repeat (PPR) family members like PCMP-H43. It suggests that PbSWEET may cooperate with sucrose transporters in interactions or use the cytoskeleton to play a major role in sugar transfer. Meanwhile, combined with the analysis of interaction breadth and functional span, we speculate that AtSWEET3 (PbSWEET9/10) is the hub protein of the interaction network, but this speculation still needs to be verified by quantitative indices and functional experiments.

Sucrose transporter (SUC) family transporters directly interact only with SWEET family proteins in the network, suggesting that PbSWEET may be involved in the transmembrane transport of sucrose in synergy with the SUC family. Meanwhile, it suggests that PbSWEET may act as an intermediary node to indirectly connect the SUC sucrose transporter with cytoskeletal transport and glycosylation modification. Furthermore, ARPC2B interacts with multiple SWEET proteins as a connecting node, suggesting that SWEET may complete intracellular transport through ARPC2B with the assistance of the cytoskeleton to achieve transport function [43]. Although glycosyltransferase is located at the edge of the network, it is still connected to a variety of SWEET proteins. We hypothesize that PbSWEET may form a protein complex with glycosyltransferase, which is synergistic and participates in glycosylation modification-related processes, and then participates in the maintenance of plant homeostasis. Additionally, PCMP-H43, a PPR protein implicated in mitochondrial/chloroplasts and participating in RNA processing and energy metabolism regulation [44], appeared in this network, indicating that PbSWEET-mediated sugar transport may be functionally coupled with intracellular energy metabolism [45], but the specific mechanism between the two still needs to be further verified experimentally.

2.9. Hormone-Induced Expression Analysis of PbSWEET Genes in Phoebe bournei

To verify the expression patterns of the four candidate genes (PbSWEET3, PbSWEET4, PbSWEET10, and PbSWEET18) (screened based on cis-regulatory elements and tissue expression profiles) in the leaf tissues under ABA and MeJA treatments, we conducted qRT-PCR analysis (Figure 10).

Overall, this result supports the existence of distinct member-specific and time-dependent response patterns in the expression of the PbSWEET genes under ABA and MeJA treatment conditions. PbSWEET4 exhibited the most rapid positive response, peaking at 4 h under both ABA and MeJA treatments. At this time point, its expression was induced approximately 38-fold by ABA and 23-fold by MeJA, after which expression levels decreased. PbSWEET10 exhibited distinct hormone-specific delayed response characteristics under ABA treatment: It rapidly increased by approximately 155 times at 4 h after MeJA treatment, while under ABA treatment, its expression peak was delayed until 12 h, and the expression level skyrocketed to approximately 281 times that of the control. In contrast, PbSWEET18 showed only a weak early response to ABA treatment (upregulated by approximately 2.65 times at 4 h), while its expression level was generally lower than that of the control (except for 12 h) in the presence of MeJA treatment. The expression of PbSWEET3 was generally inhibited by the two types of hormones; at all the tested time points, the expression level was lower than the control level (except for a slight and weak increase in ABA at 4 h). Among them, PbSWEET10 exhibited extremely significant delayed burst-like induction expression during the long-term ABA stress exposure, together with its high expression in leaves (Figure 7), thus becoming the primary candidate gene for further exploration of ABA-mediated sugar transport mechanisms under abiotic stress in P. bournei.

Furthermore, based on the results of the one-way ANOVA, both the ABA treatment and the MeJA treatment significantly affected the expression levels of the PbSWEETs. It is worth noting that both PbSWEET4 and PbSWEET10 exhibited sustained upregulation under both ABA and MeJA treatments, with a more pronounced and longer-lasting effect under MeJA treatment, suggesting that these two PbSWEET genes respond more persistently to MeJA signaling than to ABA signaling.

2.10. Time-Dependent Physiological Responses of P. bournei Leaves to ABA and MeJA Treatments

To investigate the dynamic physiological regulation of P. bournei leaves under the induction of ABA and MeJA, we measured the leaf water loss rate, proline (Pro) (a compatible osmolyte involved in osmotic adjustment) content, malondialdehyde (MDA) (a biomarker of membrane lipid peroxidation) content, soluble sugar content and soluble protein content at 0, 4, 8, 12 and 24 h after treatment with ABA (2 mmol/L) or MeJA (2 mmol/L), with three biological replicates at each time point (each with three technical replicates) (Figure 11).

Under ABA treatment, the overall leaf water loss rate showed a continuous upward trend, reaching approximately 20% within just 4 h, but the rate of increase slowed after 12 h (Figure 11E). The contents of soluble sugars, soluble proteins, and MDA all increased first and then decreased, but the timing of the peaks differed: Soluble sugars peaked at 12 h and then dropped rapidly, suggesting dynamically regulated accumulation, possibly as a means of preventing excessive carbon consumption (Figure 11C); MDA also rapidly recovered after reaching its peak at 12 h, indicating that although ABA treatment temporarily induced membrane lipid peroxidation, it did not cause sustained oxidative damage (Figure 11A); soluble protein peaked at 4 h (about 1.50 times that of the control), a pattern consistent with the possibility that early ABA treatment may have rapidly induced the synthesis of stress-related proteins, while the subsequent continued decline suggested these proteins may be rapidly turned over or degraded (Figure 11B). Notably, the trend of Pro content was reversed, reaching its lowest point at 4 h (about 70% of the control) and then slowly recovering (Figure 11D), contrasting with the significant increase in soluble sugars, suggesting that under ABA stress, soluble sugars may serve as the main osmolyte rather than Pro.

Under MeJA treatment, the increase in leaf water loss rate was lower than the ABA level at 4 h, but then continued to rise rapidly, reaching approximately 55% in 24 h (Figure 11E), surpassing the ABA group. The responses of soluble sugars and MDA were significantly delayed. Soluble sugar content peaked at 24 h (about 1.50 times that of the control), while the change was minimal between 4 and 12 h (Figure 11C), indicating that MeJA-induced sugar accumulation initiated more slowly than ABA-induced accumulation. MDA content remained low until 24 h and only increased significantly at 24 h (about 1.50 times that of the control) (Figure 11A), indicating that early MeJA treatment did not cause significant membrane lipid peroxidation. Similar to ABA treatment, Pro slowly rebounded after reaching its lowest point at 4 h (about 50% of the control), but remained below the control (Figure 11D), further supporting the hypothesis that soluble sugars may serve as the main osmolyte in P. bournei in response to hormone stress. In contrast, soluble protein content fluctuated only slightly throughout the treatment period, with no early dramatic changes as observed under ABA treatment (Figure 11B).

These results suggest that ABA and MeJA activate distinct temporal programs of physiological adjustment in P. bournei leaves.

3. Discussion

The SWEET sugar transporter family has been shown to be involved in basic physiological processes, such as the distribution of plant photosynthetic products and reproductive development in herbaceous plants [1,6,7,46], and it also plays an important role in biotic stress [10,11,12,13] and abiotic stress [18,19]. However, there are differences in the life cycle, source and reservoir dynamics, and sugar transport distance between herbaceous and woody plants, so the functional understanding of the SWEET family in herbaceous plants cannot be directly applied to woody plants. In particular, as an endangered and high-value timber species, P. bournei has faced increasing environmental pressures in recent years, which have seriously threatened its natural distribution and long-term management. Although genome-wide analysis of the SWEET gene family has been conducted in other plants, the evolutionary characteristics of the SWEET gene family in P. bournei, as well as the expression mode and biological function in the hormone regulatory network, are still unknown. To this end, we identified the SWEET gene family in P. bournei using the whole-genome approach and conducted a systematic analysis of its evolutionary characteristics, hormone-responsive expression patterns, and related plant physiological responses under ABA and MeJA treatments. These analyses further reveal the functions of these genes and lay a solid foundation for more complex applications such as forest tree breeding in practice.

3.1. Evolutionary Conservation and Lineage-Specific Expansion of the PbSWEET Gene Family

This study identified 21 PbSWEET genes in the genome of P. bournei, which were classified into 4 subfamilies (Figure 2). This quantity is comparable to that of O. sativa (21) [47] and is slightly higher than that of A. thaliana (17) [48]. In studies on woody plants, there are 36 Hevea brasiliensis (Willd. ex A. Juss.) Müll. Arg. and an astonishing 52 Eucalyptus grandis W. Hill ex Maiden [49,50]. Generally, woody plants tend to possess more SWEET genes than A. thaliana, which may be related to the increased demand for phloem loading during wood development [51]. However, compared with other woody species, the SWEET gene family in P. bournei follows a relatively conservative expansion pattern. From the evolutionary tree, it can be seen that the SWEET gene family of P. bournei shows a non-equilibrium expansion characteristic during the evolutionary process, especially in subfamily A, where a significant expansion has occurred. Similar to this pattern, the SWEET gene family in E. grandis has also undergone a significant expansion in the B1 subgroup; this differential expansion of E. grandis is consistent with the need of woody plants for more sugar to be transported from the phloem to support tree growth [50], but the lineage-specific expansion mechanism and reasons in P. bournei are not yet clear. Notably, the expansion of the A subfamily of P. bournei may be closely related to its unique growth characteristics and stress adaptation strategies [52].

Chromosome localization analysis revealed that the 21 PbSWEET genes were unevenly distributed across 10 chromosomes (Figure 1). Among them, Chr08 contains five genes, showing a distinct clustered distribution pattern, which is highly similar to tandem repeat clusters observed in E. grandis, Populus trichocarpa Torr. & A. Gray ex Hook., H. brasiliensis, and Glycine max (L.) Merr. (both woody plants and herbaceous plants) [49,50,51,53]. It indicates that tandem duplication is an important and conserved mechanism for the expansion of the SWEET family. The further identified four pairs of segment duplication events (occurring between Chr01 and Chr03, Chr05 and Chr10, Chr05 and Chr12, and within Chr06) further confirmed that segment duplication was also an important driving force for the expansion of this family (Figure 5). This pattern has also been clearly mentioned in P. trichocarpa with five pairs of repetitive segments [51] and G. max with 95 genes derived from repetitive segments (both woody plants and herbaceous plants) [53]. All the detected segmentally duplicated PbSWEET genes had Ka/Ks < 1 (Table 2), indicating that they had undergone purifying selection, and the core sugar transport function was strictly protected during evolution. This observation is consistent with previous reports demonstrating that purifying selection is the dominant evolutionary force shaping the SWEET family across diverse plant lineages [49,51,53,54]. It is worth noting that the Ka/Ks value between PbSWEET1 and PbSWEET2 is NaN, indicating that they are the result of a recent duplication event and have not yet accumulated significant sequence differences. Similarly, in H. brasiliensis, HbSWEET2e/HbSWEET2f (Ks = 0.0085) and HbSWEET2c/HbSWEET2d (Ks = 0.1065) also exhibit very low Ks values for recent replicating sequences [49]. These recent replication events may provide raw materials for new functionalization or sub-functionalization and are closely related to the adaptation of species to specific habitats [55,56]. The interspecific collinearity analysis revealed that seven PbSWEET genes have collinearity relationships with O. sativa, while only four have collinearity relationships with A. thaliana (Figure 6). This suggests that the SWEET family in P. bournei may have retained more ancient collinearity modules with monocotyledonous plants like O. sativa than with A. thaliana. This result contrasts with the conclusion drawn in previous studies that the correlation between Mangifera indica L. and A. thaliana was more significant [57]. It may be related to the specific genomic evolutionary history experienced by different dicotyledonous woody plants. P. bournei’s phylogenetic position as a Laurales, an early-diverging angiosperm lineage, is coupled with a slow rate of genomic structural evolution and the absence of additional lineage-specific whole-genome duplications, which together preserved ancestral syntenic blocks largely lost in most eudicots.

Most of the 21 PbSWEET proteins contain the typical MtN3_slv domain (Figure 3C), and the composition of conserved motifs within the same subfamily is highly similar (Figure 3B), which is consistent with the findings in both woody plants and herbaceous plants [23,50,51,53,57]. This further confirms the high structural conservation of the SWEET family during evolution, and its core sugar transport function has a definite molecular basis [8,58]. It is worth noting that nearly half of the members in the A subfamily of P. bournei lack the UTR region (Figure 3D), suggesting that part of them may perform constitutive expression functions as their primary role, which is in agreement with the observations regarding the expression analysis made in the study of Populus alba L. × Populus glandulosa Uyeki (clone 84K; hereafter Pag) [51].

Based on the above analysis, the SWEET gene family exhibits characteristics of both conservation and plasticity during evolution. Its core sugar transport function is strictly protected by purification selection (manifested as the conserved MtN3 domain), but the subfamily expansion pattern and replication type (dominated by tandem repeats and segmental duplications) exhibit significant lineage specificity. This evolutionary pattern enables the SWEET family to not only maintain the basic sugar transport function but also to adapt to the specific physiological requirements and environmental challenges of different species.

3.2. Functional Prediction of PbSWEET Proteins

Subcellular localization prediction (Table 1) indicates that the majority of PbSWEET proteins are located in the plasma membrane (16), and this finding aligns with the observations in E. grandis (nearly 40), collectively supporting the core function of transmembrane sugar transport of this family [50,59]. GO enrichment analysis further confirmed that the PbSWEET family was significantly enriched in functions such as sugar transmembrane transport activity and carbohydrate transport (Figure 8). In addition to the predominant plasma membrane localization, some PbSWEETs are targeted to chloroplasts or vacuoles, which has also been mentioned in woody plants; for instance, PtoSWEET17b in Populus tomentosa Carrière is a vacuolar membrane sucrose transporter [13,46,60], and EgSWEET16a/16b in E. grandis are also located on the vacuolar membrane [50]. Overall, the diverse subcellular localization of the PbSWEET protein is consistent with its multiple functions in sugar transport and homeostasis regulation. Most of the members are located on the cell membrane to facilitate the transport of sugar between cells, while some subgroups targeting chloroplasts and vacuoles may regulate the intracellular sugar dynamic balance. This distribution pattern is highly consistent with the functional specialization characteristics of SWEET proteins in other woody plants, highlighting the conservation of the sugar allocation mechanism in perennial plants.

Consistent with their predicted roles in transmembrane sugar transport, the tissue-specific expression profiles of PbSWEET genes revealed distinct organ-specific preferences that align with known physiological demands (Figure 7). The highly expressed PbSWEET5 in the xylem, PbSWEET12/21 enriched in root/stem parenchyma, and PbSWEET10/19 specifically expressed in leaves, respectively, suggest that they play a transmembrane transport role during long-distance transport in xylem tissues and the formation of secondary walls [61], sugar unloading in storage organs, and the export of photosynthetic products from source tissues. To further explore the regulatory mechanisms underlying the sugar transport function of PbSWEET proteins, PPI network prediction was performed. It indicates that PbSWEETs may interact with glycosyltransferases, sucrose transporters SUC2/SUC4, cytoskeletal component ARPC2B, PPR protein PCMP-H43, etc. (Figure 9). This suggests that it not only participates in sugar transport but may also act as an intermediary node connecting sucrose transporters to cytoskeletal transport (via ARPC2B) and glycosylation pathways; it also potentially couples sugar transport with intracellular energy metabolism via PCMP-H43. Collectively, PbSWEETs, through dynamic PPIs, act as a central hub that integrates sugar transport with intracellular transport, energy metabolism, and a broader range of physiological processes.

Taken together, the PbSWEET protein has a conserved and multifunctional role. Among them, the members located on the cell membrane are mainly responsible for the exchange of sugars between cells, while the subtypes containing cell-targeting domains can precisely regulate the sugar homeostasis within the cells. Furthermore, it is known that several interacting partners (such as SUC2/SUC4 and ARPC2B) are involved in hormone-mediated regulatory pathways, including ABA and auxin signal transduction [62,63,64], which indicates that PbSWEETs may also be regulated or participate in hormone-dependent processes. To further explore this possibility and these mechanisms, we predicted the cis-acting elements of the promoter regions of the PbSWEET genes and analyzed their expression patterns under different hormone treatment conditions.

3.3. Temporal Coordination of PbSWEET Expression and Plant Physiological Adjustments Under ABA and MeJA Treatments

To investigate the functional correlation of the PbSWEET gene under hormonal stress, we analyzed the temporal expression profiles of four candidate genes by qRT-PCR and simultaneously measured the key physiological parameters (leaf water loss rate, Pro content, soluble sugar content, soluble protein content, and MDA content) in the leaves of P. bournei treated with ABA or MeJA. The results show that although all four genes contain hormone-responsive cis-regulatory elements (Figure 4), they all exhibit significant member specificity, hormone specificity, and time-dependent expression patterns (Figure 10). This transcriptional-level difference suggests that different members of PbSWEET have undergone functional specialization in response to abiotic stress. It is worth noting that the dynamic changes in physiological regulation (particularly the accumulation of soluble sugars and the inhibition by Pro) (Figure 11) show a temporal coordination with the expression of specific PbSWEET genes, suggesting a possible functional association between the transcriptional activation of sugar transporters and the realization of the osmotic regulation mechanism.

PbSWEET10, a plasma membrane-localized bidirectional sugar transporter, exhibits high transcript abundance in leaf tissues. Under ABA treatment, the expression level gradually increased at the beginning, and then experienced a delayed explosive increase at 12 h, which coincided with the peak of soluble sugar content and the sharp increase in MDA. The gradual accumulation of soluble sugars alongside declining Pro levels is consistent with the hypothesis that early osmotic adjustment might rely primarily on sugar import. The delayed burst of PbSWEET10 coincided with the timing of soluble sugar accumulation, and it may be speculated that this sugar influx into mesophyll cells could lower water potential and help maintain turgor under stress. However, excessive sugar influx, if it occurs, might be associated with enhanced metabolic activity, potentially contributing to reactive oxygen species (ROS) accumulation and subsequent membrane lipid peroxidation, as suggested by the MDA peak. Subsequently, PbSWEET10 expression declines sharply, accompanied by a concurrent decrease in soluble sugar content, which is coordinated downregulation that coincides with alleviation of oxidative damage. A working hypothesis arising from this correlation is that rapid suppression of the gene could prevent excessive carbon consumption and restore cellular redox homeostasis. However, this remains a correlative inference that requires direct functional validation. The consistently higher soluble sugar levels relative to Pro further support the hypothesis that sugars function as the primary osmotic regulators under ABA-induced stress. Under the treatment of MeJA, PbSWEET10 exhibited an explosive increase immediately at 4 h, and although it rapidly declined from the peak level afterwards, it still maintained a relatively high expression level. The soluble sugar levels rose sharply at 24 h, suggesting a possible hypothesis that the early activation of PbSWEET10 might temporarily pump the sugar out of the cell into the periplasm to initiate defense signals, while the 24 h sugar surge might result from other transporters or metabolic changes. The bidirectional transport ability of PbSWEET10 is speculated to enable it to switch between different functions, such as osmotic regulation and defense initiation in response to stress conditions.

Although PbSWEET10 exhibited the most intense hormone-induced expression, the second-largest member in terms of transcriptional amplitude was PbSWEET4, a short peptide enriched in the xylem and located in the chloroplast, lacking the classical sugar transport function. Under ABA and MeJA treatments, PbSWEET4 exhibited an immediate and explosive transcriptional burst, followed by a sharp decline, preceding in time the subsequent oxidative burst and the late increase in soluble sugars observed under MeJA treatment. This temporal sequence raises the possibility that PbSWEET4 does not function as a sugar transporter but rather as a stress-induced signal peptide. A plausible hypothesis is that PbSWEET4 acts as a retrograde signal from the chloroplast to the nucleus in the early stage, specifically speaking, after sensing ABA or MeJA, its rapid upregulation in the parenchyma cells of the xylem could trigger changes in the redox state of the chloroplasts or the release of secondary messengers, potentially contributing to downstream defense and antioxidant responses, including the observed delayed accumulation of MDA and (under MeJA conditions) the accumulation of soluble sugars. The rapid decrease in PbSWEET4 after its peak level might help limit the duration of signal amplification, potentially contributing to a brief and controlled stress response. Therefore, PbSWEET4 emerges as a possible atypical member of the SWEET family. It could evolve from the ancestral sugar transporter and become a signal peptide localized in the chloroplast, coordinating the early perception of stress with subsequent physiological adaptation.

PbSWEET3 is a protein that is enriched in the xylem and localized in the chloroplast, and it lacks sugar transport activity. Unlike PbSWEET4, which undergoes a transcriptional burst, PbSWEET3 shows only a slight and brief increase under ABA before returning to low expression and is rapidly suppressed under MeJA. This pattern raises the possibility that PbSWEET3 is not a signal amplifier, but rather a negative regulator of chloroplast homeostasis. A testable hypothesis is that, under normal conditions, the basic expression of PbSWEET3 maintains the quiescence of the chloroplasts; after ABA perception, its slight increase provides a brief signal window, but quickly returns to low expression to avoid oxidative stress; under MeJA, the continuous downregulation of PbSWEET3 can relieve the inhibition of the chloroplast defense pathway, potentially allowing for the outbreak of late MDA and a significant increase in soluble sugars. PbSWEET18 is a sugar transporter located on the plasma membrane and is expressed at low levels in all tissues. It showed a persistent and slightly upward trend under ABA, but was rapidly suppressed under MeJA (with only a brief small peak observed at 12 h). One possible interpretation is that it functions as a homeostatic transport protein, maintaining the basal glucose flux. Under ABA stress, a mild increase in activity might facilitate sugar influx, supporting the hypothesis of soluble sugar accumulation for osmotic regulation without triggering oxidative bursts. Under MeJA, it immediately reduced, potentially coinciding with restricted sugar excretion and redirected carbon allocation; the 12 h peak could be tentatively associated with short-term sugar supply for defense before the 24 h oxidative burst. In summary, PbSWEET18 could be a constitutively low-expressing transporter protein that is regulated bidirectionally through ABA induction and MeJA inhibition.

It must be emphasized that all these inferences are based on temporal concordance between gene expression and physiological changes, and should therefore be considered testable hypotheses rather than definitive conclusions. Direct functional validation—for example, through gain-of-function or loss-of-function experiments—is essential to establish causality.

The ANOVA analysis indicated that the induction duration of the two genes under MeJA treatment was longer, which is consistent with the persistent role of JA signaling in leaf senescence and stress adaptation. The transcriptional reprogramming activated by MeJA through the COI1-JAZ-MYC2 pathway usually involves extensive changes in metabolic flux, and sugar excretion—the core link in the redistribution of the carbon skeleton—requires the continuous expression of SWEET genes [65]. In contrast, the SWEET expression induced by ABA (especially PbSWEET10) has a much higher peak value, which is in line with the characteristic of ABA as a stress switch that triggers acute osmotic adjustment and transient oxidative bursts. To illustrate this regulatory divergence, we propose a working model [66,67] in which the duration and amplitude of hormonal signals are decoded into distinct transcriptional outputs of the PbSWEETs (Figure 12).

P. bournei is hypothesized to employ a carbon-centric osmotic strategy, where proline is actively inhibited. PbSWEET4 could trigger an early response as a non-translocated small peptide signal, while PbSWEET10 could perform sugar redistribution in the later stage, potentially forming a dual-wave transcriptional program. Meanwhile, a transcription–metabolism uncoupling is suggested under MeJA conditions (PbSWEET10), and the oxidative damage is transient, which is consistent with the interpretation that sugar transport might represent an adaptive regulation rather than passive damage. PbSWEET10 also exhibited a specific burst-like induction under ABA treatment, which mimics certain aspects of osmotic stress signaling. Based on these observations, PbSWEET10 emerges as a promising candidate for future functional characterization. Furthermore, we speculate that, if confirmed under actual other stress conditions, the sugar transport mediated by PbSWEET10 might contribute to the stress resistance. This contribution could occur in the later stage of the stress via two hypothesized pathways: isolating sugar into vacuoles or cell walls to reduce water potential and minimize water loss; exporting sugar outside the mesophyll cells to transmit the acquired adaptive signals of the system. Subsequent studies will further verify the biochemical function and stress resistance contribution of PbSWEET10 through transgenic heterologous expression, subcellular localization, and sugaromics analysis under other typical stress conditions.

4. Materials and Methods

4.1. Materials

4.1.1. Genomic Data

The genomic sequence data and annotation information of P. bournei were obtained from the China National GeneBank (CNGB) Sequence Archive (CNSA) database (https://db.cngb.org/cnsa/, accessed on 2 January 2026), and the corresponding accession number is CNP0002030 [27]. The genome sequence files of A. thaliana and O. sativa were obtained from EnsemblPlants (https://plants.ensembl.org/, accessed on 2 January 2026) and Phytozome v13 (https://phytozome-next.jgi.doe.gov/, accessed on 2 January 2026), respectively. In addition, the RNA-seq data of different tissues of P. bournei were obtained from BioProject (https://www.ncbi.nlm.nih.gov/bioproject/, accessed on 2 January 2026), and the corresponding accession number is PRJNA628065.

4.1.2. Plant Materials and Treatments

These plant materials were obtained from the Fujian Academy of Forestry Sciences (Fuzhou, China). One-year-old P. bournei seedlings were grown in an artificial climate chamber at 25 °C, 75% humidity, and a 16 h light/8 h dark photoperiod. Among these plants, we selected 27 young seedlings of P. bournei with basically the same growth potential and treated their leaves separately with ABA (2 mmol/L) and MeJA (2 mmol/L) for 0, 4, 8, 12, and 24 h. At the same time, photos were taken at the aforementioned time points to record the changes in the leaves of P. bournei over time under the treatment of ABA and MeJA in order to be used for phenotypic observation and analysis. Untreated leaves (0 h) from three biological replicates served as the control (CK). Each of the other treatments included three biological replicates. After processing, the leaf samples were collected. All the samples were immediately frozen in liquid nitrogen and then stored at −80 °C for subsequent qRT-PCR analysis and plant physiological index measurements.

4.2. Identification and Physicochemical Property Analysis

The conserved domains of the SWEET family in A. thaliana were downloaded from the Pfam database (https://www.ebi.ac.uk/interpro/entry/pfam/PF03083/, accessed on 3 January 2026). Through local BLASTp search, the conserved domains of SWEETs in P. bournei and A. thaliana were compared, and candidate SWEETs in P. bournei were screened out [68]. After removing the redundant results from the BLASTp search, the identified protein sequences of the P. bournei SWEETs were submitted to NCBI for further BLASTp search. To further identify the members of the SWEET gene family, the HMM model (PF03083) for the conserved domain of SWEET was downloaded from the Pfam database using the HMMER-3.2.1 software; the e-value was set to <10^−5, and the default parameters were used. Subsequently, the physicochemical properties of 21 PbSWEET protein sequences were analyzed using the ExPASy ProtParam online tool (https://web.expasy.org/protparam/, accessed on 4 January 2026), including predictions of parameters such as the number of amino acids, molecular weight, theoretical pI, instability index, lipid solubility index, and the GRAVY [69]. The protein subcellular localization was predicted using the WoLF PSORT online tool (https://wolfpsort.hgc.jp/, accessed on 4 January 2026) [70].

4.3. Chromosome Localization and Evolutionary Analysis

Using the TBtools (v2.468) software, based on the GFF of P. bournei, the 21 PbSWEET genes were located on the chromosomes, and a chromosome distribution map was drawn [71]. Using the Muscle program of MEGA11 software with default parameters, a multiple sequence alignment was conducted on the SWEET protein sequences of P. bournei (21), A. thaliana (17), and O. sativa (21) [72]. The complete list of A. thaliana and O. sativa SWEET genes, including their gene accession numbers, is provided (Table S2). An evolutionary tree was constructed using the maximum likelihood method with 1000 bootstrap repetitions, and the evolutionary tree was visualized and beautified using iTOL (https://itol.embl.de/, accessed on 5 January 2026) [73].

4.4. Gene Structure, Conserved Motifs, Conserved Domains, and Homology Modeling Analysis

Using the TBtools software, the information of gene structure, conserved domains, and conserved motifs was integrated to draw a three-in-one comprehensive analysis diagram. Among them, the gene structure was derived from the GFF annotation file of the P. bournei genome; the conserved domains were retrieved through the NCBI CDD database (https://www.ncbi.nlm.nih.gov/cdd/, accessed on 6 January 2026); the conserved motifs were predicted using MEME Suite (https://meme-suite.org/meme/, accessed on 6 January 2026), with the number of motifs set to 10 and all other parameters remaining at their default values [74].

The three-dimensional structure of the PbSWEET4 protein was modeled using the SWISS-MODEL workspace (https://swissmodel.expasy.org/, accessed on 2 June 2026). The template used was 5ctg.1.A, which is derived from the three-dimensional structure of the rice OsSWEET2b protein. The template was automatically screened as the optimal template based on sequence identity (48.33%) and modeling coverage. Modeling coverage was calculated as the ratio of the aligned region length to the total length of the template. After the model was constructed, the built-in tools of SWISS-MODEL were used for visualization and structural analysis.

4.5. Promoter Cis-Element Analysis

The 2000 bp promoter sequence upstream of the start codon of the PbSWEET gene was extracted, and the cis-acting regulatory elements were predicted and classified using the PlantCARE database (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/, accessed on 6 January 2026) [75]. TBtools was used to conduct a visual display based on functional categories (light response, hormone response, stress response, growth and development-related, etc.).

4.6. Collinearity and Selection Pressure Analysis

The MCScanX module in the TBtools software was used to conduct intraspecific collinearity analysis of the P. bournei genome, identify segmental duplication events of the PbSWEET gene family, and present the results in the form of a Circos diagram [76]. The interspecies collinearity analysis was conducted using the One Step MCScanX module in TBtools. Comparative genomic collinearity analyses were performed for P. bournei, A. thaliana, and O. sativa. The species-level homology relationship of SWEET genes was presented through a dual synteny plot.

Using the Simple Ka/Ks Calculator (NG) module in the TBtools software, the Ka and Ks values were calculated for the duplicated gene pairs within the PbSWEET gene family. Furthermore, the selection pressure type of the gene pair was evaluated using the Ka/Ks ratio (Ka/Ks < 1 indicates purifying selection, Ka/Ks = 1 indicates neutral evolution, and Ka/Ks > 1 indicates positive selection) [77].

4.7. Tissue-Specific Expression Analysis

Based on the transcriptome data of different tissues of P. bournei, the expression profiles of the PbSWEET gene family were analyzed. Using the Heatmap module of the TBtools software, a clustering heatmap was plotted based on the expression levels converted from log₂(FPKM + 1); both rows and columns were clustered using hierarchical clustering with Euclidean distance and the complete linkage method.

4.8. Functional Annotation and GO Enrichment Analysis

The functional annotation of the PbSWEET protein sequences was conducted using eggNOG-mapper (http://eggnog-mapper.embl.de/, accessed on 7 January 2026), and the GO annotation information was obtained [78]. The GO enrichment analysis of the annotation results was conducted using the TBtools software, and the results were classified and displayed according to the three major categories: biological process (BP), molecular function (MF), and cellular component (CC).

4.9. Prediction and Visualization of PPI Network

Based on the STRING database (https://string-db.org/, accessed on 8 January 2026), using the SWEET proteins of A. thaliana as a reference, the PPI network of PbSWEET proteins was predicted [79]. Using a medium confidence level parameter (0.4) as the minimum interaction score threshold, the interaction network was visualized and optimized using the Cytoscape (version 3.9) software.

4.10. qRT-PCR Analysis

Total RNA was extracted from both control and stress-treated samples using an RNA extraction Kit (Coolaber, Beijing, China). Then, cDNA was synthesized from the extracted RNA using the M-MLV 4 First-Strand cDNA Synthesis Kit (Biomed, Beijing, China). After that, the SYBR Green dye method was used for quantitative PCR detection. Each qRT-PCR reaction system was 20 μL, containing 10 μL of 2× SYBR Green Mix, 0.4 μL each of forward and reverse primers, 2 μL of cDNA template, and 7.2 μL of nuclease-free water. The reaction employs a three-step protocol: pre-denaturation at 95 °C for 30 s; 40 cycles (95 °C for 5 s, 60 °C for 30 s with fluorescence collection); and finally, melt curve analysis (95 °C for 15 s, 60 °C for 1 min, and 95 °C for 15 s). Each treatment and time point was set with 3 biological replicates, and each biological replicate was run with 3 technical replicates. The relative expression levels of PbSWEET genes were calculated using the 2^−ΔΔCT method, assuming that the amplification efficiencies of the target and reference genes were approximately equal and close to 100% [80,81]. The specific primers used for qRT-PCR were designed using TBtools and are listed (Table S3). The internal reference gene selected was PbEF1α (GenBank accession number KX682032). The stability of EF1α under various experimental conditions, including ABA and MeJA treatments as well as drought and osmotic stress, has been previously demonstrated in multiple plant species [82,83]. Therefore, PbEF1α was considered a reliable reference gene for qRT-PCR normalization in this study. The one-way ANOVA was conducted to determine whether there were significant differences among multiple time points, and Dunnett’s post hoc test was used for pairwise comparisons between each treatment time point and the 0 h control. This method was chosen because it can more accurately reflect statistical significance when dealing with data from multiple time points.

4.11. Physiological Index Determination

Physiological indices were determined from the same samples; for each treatment group and each time point, 3 biological replicates were set up, and 3 technical replicates were conducted for each biological replicate. A one-way ANOVA was used to test the overall differences among multiple time points, and Dunnett’s post hoc test was employed to conduct pairwise comparisons between each treatment time point (4, 8, 12, 24 h) and the 0 h control group. The measurements of all the following physiological indicators are all based on the aforementioned repetition and statistical methods.

4.11.1. Determination of MDA Content

The measurement was carried out using the MDA content kit (Sinobestbio, Shanghai, China) as follows: Weigh approximately 0.1 g of the hormonally treated and frozen leaves of P. bournei, add 1 mL of the extraction solution, and homogenize in an ice bath. Then, perform 8000× g centrifugation at 4 °C for 10 min. Collect the supernatant and place on ice for measurement. Take 0.3 mL of reagent 1 and add it to a 1.5 mL centrifuge tube. Then, add 0.1 mL of the sample and mix well. Incubate in a 95 °C water bath for 30 min; then, cool it in an ice bath. Centrifuge at 10,000× g, 25 °C for 10 min. Finally, transfer 200 μL of the supernatant to a 96-well plate and measure the absorbance at 532 nm and 600 nm. Calculate the MDA content according to the manufacturer’s instructions.

4.11.2. Determination of Soluble Protein Content

The protein content was determined using the BCA protein assay kit (Sinobestbio, Shanghai, China). Approximately 0.1 g of the hormonally treated and frozen leaves of P. bournei was weighed to which 1 mL of the extraction solution was added; the mixture was homogenized at 4 °C under an ice bath then centrifuged at 10,000 rpm for 10 min at 4 °C, and the supernatant was taken as the sample solution to be tested. The working solution was prepared by mixing reagent A and reagent B at a ratio of 50:1 immediately before use. Then, they were divided into three groups: blank tubes, standard tubes, and determination tubes. They were mixed in EP tubes with 200 μL of working solution and 4 μL of distilled water, 200 μL of working solution and 4 μL of standard sample, and 200 μL of working solution and 4 μL of sample to be tested. These were then placed in a 60 °C incubator for 30 min. The absorbance values were measured at 562 nm in a micro-96-well plate and recorded separately as the absorbance of the blank tube, standard tube, and sample tube. The soluble protein content was calculated according to the manufacturer’s instructions.

4.11.3. Determination of Soluble Sugar Content

The determination was carried out using a plant-soluble sugar kit (Sinobestbio, Shanghai, China). Approximately 0.1 g of the frozen and hormone-treated leaves of P. bournei was weighed, 1 mL of distilled water was added, and the mixture was homogenized. The homogenate was transferred to a capped centrifuge tube and placed in a boiling water bath for 10 min (with the cap tightly closed to prevent evaporation). After cooling to room temperature, the mixture was centrifuged at 8000× g for 10 min at room temperature. The supernatant was transferred to a 10 mL test tube. It was then diluted with distilled water to 10 mL, shaken well, and set aside for use. Then, the samples were divided into three groups: blank tubes, test tubes, and standard tubes. They were mixed with 200 μL of concentrated sulfuric acid, 20 μL of working solution, and 80 μL of distilled water; 200 μL of concentrated sulfuric acid, 20 μL of working solution, 40 μL of distilled water, and 40 μL of sample; and 200 μL of concentrated sulfuric acid, 20 μL of working solution, 40 μL of distilled water, and 40 μL of standard solution in EP tubes, respectively. Then, they were placed in a 95 °C water bath for 10 min. After cooling to room temperature, 200 μL was transferred to each well of a 96-well plate. The absorbance values of the blank tube and the test tube were read at 620 nm. A standard curve was generated using glucose standards (0.0125–0.3 mg/mL) according to the kit manual, and the soluble sugar content was calculated based on the standard curve following the manufacturer’s instructions.

4.11.4. Determination of Pro Content

The test was conducted using the Pro content detection kit (Sinobestbio, Shanghai, China) as follows: Weigh approximately 0.1 g of the frozen and hormone-treated leaves of P. bournei, and add 1 mL of the extraction solution for ice bath homogenization; then, place it in a boiling water bath for 10 min of shaking for extraction, centrifuge at 10,000× g at room temperature for 10 min, take the supernatant, cool it down, and wait for measurement. Take 0.25 mL of the supernatant, 0.25 mL of 4 °C glacial acetic acid, and 0.25 mL of reagent II, and mix them in a 2 mL capped EP tube. Place it in a boiling water bath and keep it there for 30 min. Shake it once every 10 min. After cooling, add 0.5 mL of 4 °C toluene, shake for 30 s, and then let it stand for a while; transfer 0.2 mL of the upper solution to a 96-well plate, measure the absorbance at a wavelength of 520 nm, and record the absorbance value. A standard curve was generated using proline standards (1–15 µg/mL) according to the kit manual, and the Pro content was calculated based on the standard curve following the manufacturer’s instructions.

4.11.5. Determination of Leaf Water Loss Rate

Leaf water loss rate was determined as follows. Leaves from the 0 h control group were collected, surface water was carefully wiped off, and their size and leaf age characteristics were recorded. They were weighed to obtain the initial fresh weight (W₀). At each subsequent time point (4, 8, 12, and 24 h), independent leaves of the same leaf age and similar size were collected and weighed (W_t). The leaf water loss rate was calculated as (W₀ − W_t)/W₀ × 100%, expressed as the percentage loss of fresh weight relative to the initial weight at 0 h.

5. Conclusions

This study systematically identified the SWEET gene family in P. bournei and revealed its evolutionary conservation and lineage-specific expansion patterns. The A subfamily was significantly expanded through tandem and segment duplication, while purifying selection strictly maintained the core sugar transport function. Most PbSWEET proteins retain the classic MtN3_slv domain, with only a few exceptions, demonstrating functional conservation.

By integrating transcriptome and physiological data analysis, this study suggests that this woody plant might possess a carbon-centered osmotic regulation strategy: Under ABA and MeJA stress, soluble sugars may replace Pro as the main osmotic regulatory substance. The time-dependent expression profile further points to a possible dual-wave transcriptional program: PbSWEET4 may act as an early stress-induced signal peptide (non-classical function), while PbSWEET10 might function as a late sugar redistributor. It is worth noting that PbSWEET10 exhibits delayed burst-like expression under ABA treatment, synchronizing with sugar accumulation and the transient oxidation peak, making it an important candidate gene for ABA-mediated stress adaptation. Under MeJA, the decoupling phenomenon between the transcription of PbSWEET10 and sugar accumulation suggests complex regulation at the post-transcriptional or metabolic level.

In conclusion, this study provides a comprehensive evolutionary framework for the PbSWEET family and nominates PbSWEET10 as a potential key target for future functional research and molecular breeding, with the ultimate goal of enhancing the stress resistance of this endangered timber tree species. Direct functional validation is required to confirm these inferred roles.