Comprehensive Evolutionary Analysis of the SMXL Gene Family in Rosaceae: Further Insights into Its Origin, Expansion, Diversification, and Role in Regulating Pear Branching

SMXL genes constitute a conserved gene family that is ubiquitous in angiosperms and involved in regulating various plant processes, including branching, leaf elongation, and anthocyanin biosynthesis, but little is known about their molecular functions in pear branching. Here, we performed genome-wide identification and investigation of the SMXL genes in 16 angiosperms and analyzed their phylogenetics, structural features, conserved motifs, and expression patterns. In total, 121 SMXLs genes were identified and were classified into four groups. The number of non-redundant SMXL genes in each species varied from 3 (Amborella trichopoda Baill.) to 18 (Glycine max Merr.) and revealed clear gene expansion events over evolutionary history. All the SMXL genes showed conserved structures, containing no more than two introns. Three-dimensional protein structure prediction revealed distinct structures between but similar structures within groups. A quantitative real-time PCR analysis revealed different expressions of 10 SMXL genes from pear branching induced by fruit-thinning treatment. Overall, our study provides a comprehensive investigation of SMXL genes in the Rosaceae family, especially pear. The results offer a reference for understanding the evolutionary history of SMXL genes and provide excellent candidates for studying fruit tree branching regulation, and in facilitating pear pruning and planting strategies.


Introduction
Plants possess remarkable developmental plasticity and the adaptive ability to reshape their architecture in response to changes in light and other environmental signals or other artificial measures [1].For example, branch development affects plant shape and, ultimately, crop yield and quality [2].In fruit crops, tree pruning and shaping are essential to the promotion of branching, and understanding the molecular basis of branching is critical to fruit production [3].Plant growth and development are regulated by many factors including hormone signaling, gene transcription, and others [4].Strigolactones (SLs) are a family of terpenoid lactone hormones that control multiple developmental events in plants [4][5][6][7][8].SLs were first identified in cotton root exudates as compounds that stimulate the germination of parasitic weeds [9] and play a critical negative role in regulating the growth and development of plant branches [10].Later, SLs were shown to promote the symbiotic relationship between arbuscular mycorrhizal fungi and plant roots [11].SLs are recognized as important plant endogenous hormones that are able to inhibit branch growth [4,5].

Identification and Classification of SMXL Genes
To explore and increase our understanding of the evolution of SL signaling in the regulation of branching, we examined target repressor SMXLs in 16 angiosperm species, including 12 dicots (A.thaliana, G. max, Ficus macrocarpa L.f., Solanum lycopersicum L., Vitis vinifera L., and seven members of Rosaceae: P. pyrifolia, Eriobotrya japonica Lindl., Gillenia trifoliate Moench., Prunus persica Batsch., Malus sylvestris Mill., Rosa chinensis Jacq., and Fragaria vesca L.), two monocots (Dioscorea alata L. and Oryza sativa L.), and two basal angiosperms (Nymphaea colorata Verdc.and A. trichopoda).In total, 121 non-redundant SMXL gene family members were identified using a local BLAST method employing 8 SMXLs from A. thaliana as a query protein set.Among the 16 species investigated, the number of SMXL genes in each ranged from 3 to 18. G. max possessed the greatest number of SMXL genes (18), which was six-fold more than that in A. trichopoda (3).Each of the remaining 14 species contained 5 to 11 SMXL members.The open reading frame lengths of the SMXL proteins ranged from 722 to 1132 amino acids, with projected pIs of 5.76-8.82and MWs of 80.67-112.53kDa for the resulting proteins (Figure 1 and Table S1).
provide information for their use in future studies on the functional characterization of SMXL genes in branch regulation.The findings will also be useful in facilitating pear planting and pruning strategies.

Identification and Classification of SMXL Genes
To explore and increase our understanding of the evolution of SL signaling in the regulation of branching, we examined target repressor SMXLs in 16 angiosperm species, including 12 dicots (A.thaliana, G. max, Ficus macrocarpa L.f., Solanum lycopersicum L., Vitis vinifera L., and seven members of Rosaceae: P. pyrifolia, Eriobotrya japonica Lindl., Gillenia trifoliate Moench., Prunus persica Batsch., Malus sylvestris Mill., Rosa chinensis Jacq., and Fragaria vesca L.), two monocots (Dioscorea alata L. and Oryza sativa L.), and two basal angiosperms (Nymphaea colorata Verdc.and A. trichopoda).In total, 121 non-redundant SMXL gene family members were identified using a local BLAST method employing 8 SMXLs from A. thaliana as a query protein set.Among the 16 species investigated, the number of SMXL genes in each ranged from 3 to 18. G. max possessed the greatest number of SMXL genes (18), which was six-fold more than that in A. trichopoda (3).Each of the remaining 14 species contained 5 to 11 SMXL members.The open reading frame lengths of the SMXL proteins ranged from 722 to 1132 amino acids, with projected pIs of 5.76-8.82and MWs of 80.67-112.53kDa for the resulting proteins (Figure 1 and Table S1).
The robust classification and phylogenetic relationships between the 121 SMXL genes in the 16 angiosperms (Figure 2a) clearly showed that the SMXL family genes were clustered into four branches comprising SMXL1,2 (homologs of SMAX1, SUPPRESSOR OF MAX2 1, and SMXL2) as group I, SMXL6,7,8 as group II, SMXL3 as group III, and SMXL4,5 as group IV (Figure 2a).These findings are consistent with previous classifications based on eight SMXL homologs in A. thaliana.Overall, group II contained the most SMXL family members (43), and group III contained the second highest number of SMXL genes (33), while groups I (22) and IV (23) contained the fewest and approximately equal numbers of SMXLs.To some extent, these results reflect SMXL gene duplication or loss events during evolution.P. pyrifolia contained two, four, two, and two SMXL genes in groups I, II, III, and IV, respectively, and we renamed them PpySMXLs (Table S1) in accordance with their classification on the phylogenetic tree.Group IV was absent in F. microcarpa and S. lycopersicum, and group III was not found in A. trichopoda.The other 13 species contained all four groups of SMXL members (Figures 1 and 2a).

Gene Structures and Conserved Motif Composition of SMXL Genes
The intron-exon arrangement is often regarded as an important parameter in gene phylogenies.To confirm the classification of the 121 SMXL genes, we analyzed the predicted intron-exon organizations of their coding sequences.The results indicated that the genes had very conserved structures (Figure 2b).The majority of the SMXL family members in the four groups had two introns, with the exception only of Ej00071280, PpySMXL8a, Gt00023314, Glyma.11G230700.1,Solyc06g051460.4.1, and AmTr_v1.0_scaffold00007.53 from group IV, AtSMXL5 from group I, and Glyma.10G121542.1 from group II, which had three introns (Figure 2b).
To further investigate the structural features of SMXL proteins, the MEME online tool was applied to explore the divergence of conserved motifs (Figure 2c).As a result, 40 conserved motifs in total were identified and renamed motifs 1-40 (Figure S1).As in previous reports, all SMXLs were characterized by a domain architecture of a double Clp-N motif and a P-loop containing a nucleoside triphosphate hydrolase [12,13].Motif analysis revealed that almost all SMXL proteins had a highly conserved C-terminus and a variable N-terminus.Overall, the motifs in each group showed conserved organization, but some motifs were only contained in certain groups or SMXL members.For example, motifs 1-7, 10-12, 14-15, 17-18, and 22-23 were highly conserved in almost all of the SMXL family members (Figure 2c).Some motifs were conserved in one group, such as motifs 26, 31, 38, and 40, which were specific to group II; motifs 8, 16, 20, and 21, which were observed in groups I and II; motif 28, which was observed in groups II and IV; and motifs 34 and 36, which were contained in groups III and IV.Moreover, motifs 9 and 19 were absent from group IV, while motifs 13 and 19 were absent from group III, and motif 24 was absent from group II.Furthermore, some motifs dominated only in one group, such as motifs 27, 33, and 35 in group II; motifs 32 and 39 in group IV, and motif 25 in groups II and III.These different motif arrangements might indicate different gene functions.
Taken together, our phylogenetic and intron-exon structure analyses of the SMXL genes in angiosperms clearly showed that there were four ancestral groups predating the divergence of monocots and dicots.

Synteny Analysis and Evolution of SMXL Genes
Whole-genome duplication (WGD) contributes to the generation of single-gene duplicates and has played important roles in plant genome function and evolution [33,34].To explore the evolutionary mechanisms of the SMXL family, comparative genome mapping was performed to investigate the gene duplication events of SMXL genes in four Rosaceae members, including P. Pyrifolia, E. japonica, G. trifoliate, and P. persica.The results demonstrated a high level of conserved synteny among these species (Figure 3 and Table S2).According to our results, more than six speculative SMXL genes were expected from the common ancestor of these four species.Previous studies have detected no recent WGD events in P. persica and G. trifoliate genomes [35,36].Plant genomes are constantly in dynamic change and have suffered many gene losses since their two WGDs.There were six SMXL members in the P. persica genome but only five in G. trifoliate, so one SMXL gene in P. persica (Prupe.2G071700.1)had no corresponding link in G. trifoliate.Interestingly, one P. persica and one G. trifoliate gene recognized two E. japonica and P. Pyrifolia genes because of a recent WGD in the apple (Malus) tribe [37].The number of SMXL genes in E. japonica, M. sylvestris, and P. Pyrifolia expanded rapidly, but through a long process of evolution and domestication, and after natural or artificial selection, there are now 10 SMXL genes in pear, 10 in E. japonica (including an incomplete short sequence), and 11 in M. sylvestris.and domestication, and after natural or artificial selection, there are now 10 SMXL genes in pear, 10 in E. japonica (including an incomplete short sequence), and 11 in M. sylvestris.To further trace the phylogenetic relationships between SMXL genes in angiosperms, we constructed an unrooted phylogenetic tree of the 121 SMXL genes from the 16 species (Figure 2a).This allowed us to obtain a clearer picture of the evolution of the SMXL genes.In the common ancestor of angiosperms, there are presumed to be four SMXL genes, with one member in each group.The A. trichopoda genome lost one member of group III during evolution, leaving three members.There was an ancient single WGD event that was probably shared among basal angiosperm (N.colorata) gene members [38]; now six SMXL genes remain and none of the four groups are absent (Figures 1 and 2a).
Studies have shown that a WGD event was shared by all grasses [39], and the O. sativa genome underwent an additional two WGD events accompanied by large-scale gene loss, with nine SMXL genes remaining.Another basal monocot, D. alata, contained eight SMXL genes.In eudicots, there was a shared γ paleohexaploidy event, noted as a whole-genome triplication (WGT) event [40].The S. lycopersicum genome experienced one more WGT event, while A. thaliana and G. max experienced α and β WGD events, respectively, but they had different rates of gene loss [41][42][43], which contributed to the present number of SMXL genes in A. thaliana (8), S. lycopersicum (7), and G. max (18) (Figures 1 and 2a).After Rosaceae split from Moraceae, a recent WGD event occurred in the apple tribe.Therefore, the number in the Maloideae is double that in the G. trifoliate, P. persica, F. vesca, and R. chinensis genomes.
To further investigate the phylogeny and evolution of the SMXL gene family, an ML phylogenetic tree of pear SMXL genes was constructed and the 10 SMXL members were clearly classified into four groups (Figure 4).Furthermore, the spectrum of synonymous substitutions per synonymous site (Ks) of the 10 SMXLs were confirmed to have two WGD peaks, one with a Ks of ~0.13-0.21 and the other with a Ks of ~1.66-1.93(Table S3).Additionally, the syntenic relationships between the 10 pear SMXL genes were constructed, and there were eight pair links (PpySMXL1 vs. PpySMXL2, PpySMXL4 vs. PpySMXL5, PpySMXL6 vs. PpySMXL7, PpySMXL6 vs. PpySMXL8a, PpySMXL6 vs. PpySMXL8b, PpySMXL7 vs. PpySMXL8a, PpySMXL7 vs. PpySMXL8b, and PpySMXL8a vs. PpySMXL8b).Moreover, PpySMXL3a vs. PpySMXL3b might be a syntenic pair link but PpySMXL3b was unmapped on the chromosome (Figure 5).To further trace the phylogenetic relationships between SMXL genes in angiosperms, we constructed an unrooted phylogenetic tree of the 121 SMXL genes from the 16 species (Figure 2a).This allowed us to obtain a clearer picture of the evolution of the SMXL genes.In the common ancestor of angiosperms, there are presumed to be four SMXL genes, with one member in each group.The A. trichopoda genome lost one member of group III during evolution, leaving three members.There was an ancient single WGD event that was probably shared among basal angiosperm (N.colorata) gene members [38]; now six SMXL genes remain and none of the four groups are absent (Figures 1 and 2a).
Studies have shown that a WGD event was shared by all grasses [39], and the O. sativa genome underwent an additional two WGD events accompanied by large-scale gene loss, with nine SMXL genes remaining.Another basal monocot, D. alata, contained eight SMXL genes.In eudicots, there was a shared γ paleohexaploidy event, noted as a whole-genome triplication (WGT) event [40].The S. lycopersicum genome experienced one more WGT event, while A. thaliana and G. max experienced α and β WGD events, respectively, but they had different rates of gene loss [41][42][43], which contributed to the present number of SMXL genes in A. thaliana (8), S. lycopersicum (7), and G. max (18) (Figures 1 and 2a).After Rosaceae split from Moraceae, a recent WGD event occurred in the apple tribe.Therefore, the number in the Maloideae is double that in the G. trifoliate, P. persica, F. vesca, and R. chinensis genomes.
To further investigate the phylogeny and evolution of the SMXL gene family, an ML phylogenetic tree of pear SMXL genes was constructed and the 10 SMXL members were clearly classified into four groups (Figure 4).Furthermore, the spectrum of synonymous substitutions per synonymous site (Ks) of the 10 SMXLs were confirmed to have two WGD peaks, one with a Ks of ~0.13-0.21 and the other with a Ks of ~1.66-1.93(Table S3).Additionally, the syntenic relationships between the 10 pear SMXL genes were constructed, and there were eight pair links (PpySMXL1 vs. PpySMXL2, PpySMXL4 vs. PpySMXL5, PpySMXL6 vs. PpySMXL7, PpySMXL6 vs. PpySMXL8a, PpySMXL6 vs. PpySMXL8b, PpySMXL7 vs. PpySMXL8a, PpySMXL7 vs. PpySMXL8b, and PpySMXL8a vs. PpySMXL8b).Moreover, PpySMXL3a vs. PpySMXL3b might be a syntenic pair link but PpySMXL3b was unmapped on the chromosome (Figure 5).S3), and then interpreted the replication nodes based on the Ks value.The common ancestor nodes marked by red stars represent the duplication event during the recent pear lineage-specific WGD.The common ancestor nodes marked by blue stars represent the duplication event during the ancient legume WGD.The blue lines indicate a gene pair that is duplicated with SMXL genes.The chromosome number is indicated in each chromosome bar.The lengths of the chromosomes are marked using scales, with each small scale indicating 500,000 amino acids, and every 10th scale is marked using a number.S3), and then interpreted the replication nodes based on the Ks value.The common ancestor nodes marked by red stars represent the duplication event during the recent pear lineage-specific WGD.The common ancestor nodes marked by blue stars represent the duplication event during the ancient legume WGD.The blue lines indicate a gene pair that is duplicated with SMXL genes.The chromosome number is indicated in each chromosome bar.The lengths of the chromosomes are marked using scales, with each small scale indicating 500,000 amino acids, and every 10th scale is marked using a number.The blue lines indicate a gene pair that is duplicated with SMXL genes.The chromosome number is indicated in each chromosome bar.The lengths of the chromosomes are marked using scales, with each small scale indicating 500,000 amino acids, and every 10th scale is marked using a number.
The broad-scale phylogenetic analyses suggested some subtree topologies that are consistent with the occurrence of ancient gene duplications.We conducted syntenic network analyses for SMXL genes using the collection of 16 available plant genomes (Figure 1).Syntenic SMXL genes were observed in all the selected plants.We visualized this subnetwork using Gephi [44] and color-coded the clusters using the k-clique percolation clustering method with k = 4 (Figure 6a).To reveal syntenic relationships between distant gene clades, we displayed pairwise syntenic relationships between the SMXL genes in a gene tree constructed for the entire gene family in the 16 genomes (Figure 6b).The results are consistent with previous classification and evolution analyses of the SMXL genes.The broad-scale phylogenetic analyses suggested some subtree topologies that are consistent with the occurrence of ancient gene duplications.We conducted syntenic network analyses for SMXL genes using the collection of 16 available plant genomes (Figure 1).Syntenic SMXL genes were observed in all the selected plants.We visualized this subnetwork using Gephi [44] and color-coded the clusters using the k-clique percolation clustering method with k = 4 (Figure 6a).To reveal syntenic relationships between distant gene clades, we displayed pairwise syntenic relationships between the SMXL genes in a gene tree constructed for the entire gene family in the 16 genomes (Figure 6b).The results are consistent with previous classification and evolution analyses of the SMXL genes.

Secondary Structure Analysis and Prediction of Three-Dimensional Structure of SMXL Proteins
Three-dimensional structure prediction is critical to understanding protein function, so all of the 10 PpySMXLs proteins were included in the structure prediction analysis (Figure 7).The three-dimensional (3D) structure of the protein was predicted using Swiss-Model, and the quality of the model was evaluated via the global model quality estimate (GMQE).The SMXL protein is mainly composed of α-helices and random coils.We found that the GMQEs of the 10 selected PpySMXL proteins were all >0.57, indicating that the predicted 3D structures were reliable (Figure 7).Diverse 3D structures among SMXLs in different groups were observed, but the structures were shown to be similar between some members of the same group, suggesting that these proteins have various functions and conserved roles between duplicated genes.

Secondary Structure Analysis and Prediction of Three-Dimensional Structure of SMXL Proteins
Three-dimensional structure prediction is critical to understanding protein function, so all of the 10 PpySMXLs proteins were included in the structure prediction analysis (Figure 7).The three-dimensional (3D) structure of the protein was predicted using Swiss-Model, and the quality of the model was evaluated via the global model quality estimate (GMQE).The SMXL protein is mainly composed of α-helices and random coils.We found that the GMQEs of the 10 selected PpySMXL proteins were all >0.57, indicating that the predicted 3D structures were reliable (Figure 7).Diverse 3D structures among SMXLs in different groups were observed, but the structures were shown to be similar between some members of the same group, suggesting that these proteins have various functions and conserved roles between duplicated genes.
(Ppy09g2030.1),TCP20 (Ppy09g0071.1),TCP20 (Ppy17g0069.1),SPL9(Ppy14g0522.1),etc. (Figure 8).We found the buds fully sprouted at 14 d after fruit-thinning treatment (Figure 9a), so we selected samples from 1 d, 5 d, and 14 d to verify the expression patterns.The expression of the PpySMXLs and the other 25 selected candidate DEGs was further verified via quantitative real-time (qRT)-PCR (Table S4).The results were basically consistent with the RNA sequencing data.Most PpySMXL genes, including PpySMXL1, PpySMXL3a/b, PpySMXL4, PpySMXL5, PpySMXL6, and PpySMXL7 were highly expressed after fruitthinning treatment, particularly at 1 day after treatment, after which expression declined.The expression of PpySMXL1, PpySMXL3a/b, and PpySMXL4 was higher than that in the controls until 14 days after treatment (Figure 9b).In particular, PpySMXL3a/b expression levels were 68 times higher than those in controls at 1 day after fruit-thinning treatment; however, the expression of PpySMXL8a/b was suppressed after fruit-thinning treatment (Figure 9b).The results indicated that members of the SMXL gene family play important roles in the regulation of pear branching after fruit thinning.The negative regulatory genes including TCP and SPL genes exhibited repressed expression after fruit-thinning treatment, which represented the start of branching (Figure 10).We found the buds fully sprouted at 14 d after fruit-thinning treatment (Figure 9a), so we selected samples from 1 d, 5 d, and 14 d to verify the expression patterns.The expression of the PpySMXLs and the other 25 selected candidate DEGs was further verified via quantitative real-time (qRT)-PCR (Table S4).The results were basically consistent with the RNA sequencing data.Most PpySMXL genes, including PpySMXL1, PpySMXL3a/b, PpySMXL4, PpySMXL5, PpySMXL6, and PpySMXL7 were highly expressed after fruitthinning treatment, particularly at 1 day after treatment, after which expression declined.The expression of PpySMXL1, PpySMXL3a/b, and PpySMXL4 was higher than that in the controls until 14 days after treatment (Figure 9b).In particular, PpySMXL3a/b expression levels were 68 times higher than those in controls at 1 day after fruit-thinning treatment; however, the expression of PpySMXL8a/b was suppressed after fruit-thinning treatment (Figure 9b).The results indicated that members of the SMXL gene family play important roles in the regulation of pear branching after fruit thinning.The negative regulatory genes including TCP and SPL genes exhibited repressed expression after fruit-thinning treatment, which represented the start of branching (Figure 10).   .Gene expressions of selected genes under fruit-thinning (FT) treatment determined via qRT-PCR, with blank treatments as controls (CK).The mean expression values were calculated from three independent replicates.The X-ray indicates the number of days after treatment.Mean values and standard errors were calculated from three replicates.A t-test was used for statistical analysis.The asterisk and double asterisks represent significant differences at the levels of 0.05 and 0.01, respectively.R.e.l indicates the relative expression level.

Discussion
Since the discovery of D53/SMXL proteins in rice [12] and Arabidopsis [45], SMXL family members have gradually been characterized in many additional plant species.These proteins play important roles in various aspects of plant physiology, and evidence suggests distinct physiological roles for different SMXL orthologs [17].However, the basis of their functional diversification has been largely unknown to date.We performed a genome-wide survey of SMXL genes in 16 plant species, including basal angiosperms, monocots, and dicots, and reconstructed the evolutionary history of this gene family.The .Gene expressions of selected genes under fruit-thinning (FT) treatment determined via qRT-PCR, with blank treatments as controls (CK).The mean expression values were calculated from three independent replicates.The X-ray indicates the number of days after treatment.Mean values and standard errors were calculated from three replicates.A t-test was used for statistical analysis.The asterisk and double asterisks represent significant differences at the levels of 0.05 and 0.01, respectively.R.e.l indicates the relative expression level.

Discussion
Since the discovery of D53/SMXL proteins in rice [12] and Arabidopsis [45], SMXL family members have gradually been characterized in many additional plant species.These proteins play important roles in various aspects of plant physiology, and evidence suggests distinct physiological roles for different SMXL orthologs [17].However, the basis of their functional diversification has been largely unknown to date.We performed a genome-wide survey of SMXL genes in 16 plant species, including basal angiosperms, monocots, and dicots, and reconstructed the evolutionary history of this gene family.The SMXL gene family was conserved during evolution, and the genes can be grouped into four distinct clades in almost all of the 16 species studied, apart from A. trichopoda, S. lycopersicum, and F. macrocarpa, which were found not to contain group III, IV, and IV members, respectively (Figure 1).SMXLs are likely to have first appeared in liverworts and mosses, where they are involved in developmental responses [8].The SMXL family likely arose from a single ancestral SMXL group through two whole-genome duplication events, at the point of seed plant and angiosperm origin [17,20].Functionally, at present, the SMXL family can be divided neatly into a clade that mediates KAR/KL responses (SMAX1,2) and a clade that mediates SL responses (SMXL6,7,8) [46].
Phylogenetic and conserved motif analyses offered deep insights into the possible evolution and functional diversification of the SMXL homologs from ancient SMAX1/SMXL1.Previous studies indicated that SMXL homologs were completely absent in the lower plant groups [12,17].In our study, SMXL genes were identified in all the 16 selected angiosperms, although some members were absent in certain species (Figure 1).Our evidence confirmed that the origin of the SMXL gene family predates the emergence of angiosperms, possibly in a common ancestor containing SMAX1/SMXL1.Previous studies based on genomic and RNA-seq data from several species belonging to bryophytes, lycophytes, and monilophytes concluded that these plants possess only one ancestral SMXL clade and, most often, a single SMXL copy [20].Ancestral SMXL genes are the most similar to the SMAX1/SMXL1 clade and are thought to be involved in the ancient KAR/KL response to SL [20,47].In this study, a group that was the SMAX1/SMXL1 clade of the SMXL gene family was present in all of the 16 species studied here (Figures 1 and 2), indicating that these genes were conserved during evolution.
WGD or polyploidy contributes to the generation of single-gene duplicates or whole gene families and has played crucial roles in plant genome function and evolution [33,34].Instead of fungi and animals, the most frequent occurrence of paleo-polyploidization has been detected in angiosperms [34].In the present study, we found an obvious expansion of the SMXL gene family in angiosperms, starting from 3-4 members in basal angiosperms to 9-11 members in the Maleae (Figure 1).This expansion of SMXL genes occurred thanks to a WGD in the common ancestor of P. persica, M. sylvestris, and G. trifoliate.After several rounds of WGD events, the number of genes in the SMXL family in pear reached its current number.
The common ancestor of mesangiosperms underwent WGD event(s), while basal angiosperms (N.colorata and A. trichopoda) did not.Consequently, the number of SMXL genes doubled, but then homologs were lost.Interestingly, the gene loss events occurred selectively in groups I, III, and IV (Figure 1).When the common ancestor of pear, apple, and loquat underwent another recent WGD event [37], the numbers of SMXL genes doubled again (Figure 1).We constructed a phylogenetic tree of the 10 pear SMXL genes, with the results clearly showing expansion of these genes (Figure 4).Furthermore, the spectrum of the Ks values of the 10 SMXLs confirmed two WGD peaks, one with a Ks ~0.13-0.21 that occurred approximately ~7.0-11.3 million years ago (MYA) and the other with a Ks ~1.66-1.93 that took place ~89.6-104.2MYA (Table S3).These findings demonstrated the evolution and expansion of SMXL genes.
SMXL genes regulate diverse aspects of plant development and responses to various environmental signals in the SL signaling pathway [10,24].In Arabidopsis, AtSMAX1 responds to KAR signals to regulate seed germination and hypocotyl length [17,48].AtMAX2 is responsible for the poly-ubiquitination of target SMXL proteins, which are consequently degraded by the 26S proteasome, resulting in downstream signaling.AtSMXL3,4,5 do not respond to KARs or SLs [23], and their involvement in primary phloem formation has been discovered relatively recently [29].AtSMXL6,7,8 respond to SL signals to inhibit the expression of transcription factors BRC1, TCP1, and PAP1 by binding to the transcriptional corepressor protein TPL and TPL-related protein (TPR), thereby regulating plant branching, leaf morphology, and lateral root growth [27,30].In addition, it has been found that SMXL6,7,8 are not only repressors but also transcription factors, which negatively regulate their own transcriptional expression to maintain the dynamic balance of SMXL6,7,8 proteins and SL signal responses [24].There are studies showing that the expression levels of SMXL6,7 are also regulated by the light environment [49].
Our previous study revealed that fruit-thinning treatment could induce bud germination and that SMXL gene members respond to the germination process after fruit-thinning treatment.In this study, we detected transcripts of all 10 PpySMXL genes in pear after fruitthinning treatment and observed their distinct expression levels (Figure 8).The expression of PpySMXL6,7,8a/b were the highest, followed by PpySMXL1 and PpySMXL2 (with no difference between the experimental group and control), while the expression levels of PpyS-MXL3,4,5 were the lowest (Figure 8).Using qRT-PCR, we detected the relative expression levels of PpySMXL1, PpySMXL3a/b, PpySMXL4, PpySMXL5, PpySMXL6, PpySMXL7, and PpySMXL8a/b, which were consistent with those found in previous studies of model plants but showed a unique regulation of the pear bud after fruit-thinning treatment (Figure 9).This regulatory function might be related to the expansion, functional redundancy, and functional differentiation of SMXL genes in pear.With respect to evolution, in the short term, the functions of homologous SMXL genes in pear are relatively undifferentiated, and redundant homologous genes may play similar roles, but they may have more functions in the future [50].However, the genes have their own characteristics and functional responses to fruit-thinning treatment.For example, PpySMXL5,6,7 were found to respond quickly at an early stage (1 day) after treatment before being downregulated (Figure 9b), suggesting an important role in the early fruit thinning response.Previous studies have reported that PpySMXL6,7,8 are involved in regulating shoot branching, leaf elongation, and anthocyanin biosynthesis [24].In this study, the expression of PpySMXL6 and PpySMXL7 was significantly upregulated after fruit-thinning treatment (Figure 9b), but PpySMXL8 was inhibited (Figure 9b), indicating the probable functional differentiation of these SMXL genes in pear.Other new functions in other signaling pathways involved in regulating pear growth and development may also be present, although to date, these have not been identified.In summary, SMXL genes involved in the regulation of branching in pear are candidates for the study of branching in plants.
Besides the SMXL genes, there are also some other genes associated with cell division, differentiation, and SAM activity, the responses of which to branching were identified, including transcription factors belonging to the ERF, SPL, TCP, WRKY, and NAC families, among others (Figure 8).We selected and detected the expression of 25 genes in response to fruit-thinning treatment here (Figure 10).Transcription factors play a vital role in plant development and the regulation of gene expression, forming a complex gene regulatory network [51].Most of the selected transcription factors were highly expressed under fruitthinning treatment in the regulation of branching traits.The TCP transcription factors are involved in the growth of lateral meristems, cell proliferation, and the regulation of hormones [52,53].In peach, PpTCP18 can reduce secondary branches through the brassinolide pathway [54].However, we did not find similar results in this study might because we missed the closest homologs of TCP18.The SPL genes are mainly expressed in meristems and are critical to regulating the branching and vegetative growth of alfalfa plants [55].Previous studies indicated that overexpression of SPL13 inhibits the growth of axillary buds and reduces the number of lateral branches [56].We found in this study that the SPL4 and SPL9 genes were significantly downregulated after the fruit-thinning treatment (Figure 10).The results show that SPL genes were indeed involved in tree branching, and that this is a complex regulatory process involving a large number of genes.Furthermore, there are many other transcription factor genes including ERF, NAC, WRKY, MYB, WOX, CYP, and so on that were induced at more than one time point via fruit-thinning treatment (Figure 10).The present research not only provides good candidate genes for fruit tree branching research, but also provides a new perspective for researchers studying plant branching.

Phylogenetic Tree, Gene, and Three-Dimensional Protein Structures, Conserved Domains, and Motifs
Multiple-SMXL-sequences alignment was performed using MEGA11.0[59], and the resulting sequence was trimmed using the Simple MSA Trimmer assembled in TBtools V1.113 [58].The phylogenetic tree of species was constructed in accordance with the AGP IV system [60], and polyploidization events described in previous studies [37,40,[61][62][63] are mapped onto the tree.The phylogenetic tree with 121 SMXL genes in 16 angiosperms was created based on the maximum likelihood method and Poisson's correction model using 1000 bootstrap replicates [64], and then edited and visualized using FigTree software (http: //tree.bio.ed.ac.uk, accessed on 12 September 2023).SMXL gene structure information was parsed from the GFF files of each genome.Gene Structure View (Advanced) within TBtools V1.113 was used to create diagrams of domain locations and gene structures.MEME online tools (https://meme-suite.org/meme/tools/meme, accessed on accessed on 21 September 2023) were used to predict conserved motifs using the following parameters: repeats per sequence (any); motif width range (6-50 amino acids); and maximum number of motifs (50).The tertiary structures of SMXL proteins were modeled and displayed using the Swiss-Model interactive tool (https://swissmodel.expasy.org/interactive,accessed on 2 September 2023).

Genome Synteny and Gene Duplication Analysis
Genome assembly and annotation files were obtained from Genome Database for Rosaceae (P.pyrifolia); Phytozome (P.persica) and China National GeneBank (E.japonica and G. trifoliate).Genome synteny information for P. pyrifolia, P. persica, E. japonica, and G. trifoliate was calculated using MCScanX [65].MCScanX was also used to recognize whole-genome duplication events in SMXL genes from pear (P.pyrifolia).
SMXL syntenic networks were analyzed with CFinder v2.0.6 [66] using the unweighted CPM algorithm and no time limit.All possible k-clique communities for the SMXL gene syntenic networks were identified using k = 4 as the clique community threshold-in this scenario, one SMXL gene (node) involved in a subnetwork community needed to have at least two connections (edges) with other nodes in the community.Increasing k values make the communities smaller and more disintegrated but also more connected.TBtools V1.113 was applied to perform a collinearity analysis of SMXL genes with default parameters.The Ks values of syntenic genes were calculated using KaKs_Calculator 3 (https://ngdc.cncb.ac.cn, accessed on 15 September 2023) [49].

Plant Materials and Treatment
Cultivar 'Qiuyue' (P.pyrifolia), a widely planted cultivar imported from Japan, was used for fruit-thinning treatments in this study.It was grown in the farm.Samples of buds were collected at 0, 1, 3, 5, 7, 14, and 21 days after fruit-thinning treatment and immediately frozen using liquid nitrogen for future experiments.Each treatment had three biological replicates.The frozen samples were stored at −80 • C until further use.

Expression Analysis of Pear SMXL Genes
Expression data for SMXL genes were obtained from unpublished transcriptome data collected following a fruit-thinning treatment to detect the ability of pear branches to germinate.HeatMap Illustrator within TBtools V1.113 [58] was used to draw a graphic representation of expression patterns.Gene expression levels at the different time points were calculated in accordance with the log2 FPKM values.Based on heatmap results, the 10 PpySMXLs and 25 candidate genes for fruit-thinning treatment were selected to verify the RNA-seq data.
Total RNA was isolated using a plant RNA purification kit (Takara, Japan) from leaf tissues in accordance with the manufacturer's instructions.Then, gDNA digestion was carried out using 3 µL 5 × gDNA Digester Mix and RNase-free ddH 2 O.The PCR products were sequenced to verify the absence of gDNA in the RNA fraction.The pure RNA was reverse-transcribed into cDNA using the reverse transcription kit (Vazyme, Nanjing, China).The specific primers were designed based on the selected gene sequences using TBtools V1.113 [58], and then, all the qRT-PCR products were sequenced to select the specific primers for follow-up detection.The expression of the selected genes was analyzed using a BIOER, FQD-96A real-time PCR system with 2×HS Taq Universal SYBR Green qPCR Master Mix, SAIPUBIO.Gene-specific primers were designed based on the selected gene sequences using TBtools V1.113 [58].Actin from P. pyrifolia (Ppy01g0117.1)was used as the reference gene.The amplification parameters were as follows: 95 • C hold for 10 min, followed by 40 cycles at 95 • C for 15 s, 58 • C for 15 s, and 72 • C for 15 s.For the melting curve stage, the default settings were chosen.Monospecific products were identified by inspecting the melting curves (Figure S2).The relative expression level for each gene was calculated with the 2 −∆∆Ct method.Each sample analysis was repeated three times.A T-test was used for statistical analysis.All the primers used in this paper are listed in Table S4.

Conclusions
In summary, this study provides important insights into the SMXL gene family in Rosaceae species and their roles in regulating pear branching.In conclusion, 10 PpySMXL genes were identified in a pear genome, and 121 SMXL genes were comprehensively

Figure 1 .
Figure 1.Identification and classification of SMXL genes in 16 angiosperm species.The phylogenetic tree was constructed in accordance with the AGP IV system.The estimated time is from GSA Geologic Time Scale (https://www.geosociety.org/GSA/GSA/timescale/home.aspx,accessed on 10 August 2023).Polyploidization events described in previous studies mentioned in the method are mapped onto the tree (orange ovals and red diamonds).The total number of SMXL genes and their classification in each plant are shown in the figure.WGD: whole-genome duplication; WGT: whole-genome triplication; Mya: million years ago.

Figure 2 .
Figure 2. Phylogenetic tree analysis, exon/intron structures, and motif distribution of SMXL genes.(a) The phylogenetic tree of 121 SMXL genes in 16 angiosperm species constructed using the maximum likelihood (ML) method and Poisson's correction model.Four groups are represented in variously colored branches.(b) The distribution of the intron and exon organization of SMXL genes shown using TBtools V1.113 software.The green boxes indicate the CDSs (coding sequences) and the gray lines indicate the introns.The scale bar at the bottom allows for the estimation of the length and relative position of each box.(c) The conserved motifs of SMXL proteins.Differently colored boxes indicate various motifs.

Figure 3 .
Figure 3. Synteny analysis of SMXL genes among P. pyrifolia, E. japonica, G. trifoliate, and P. persica.The gray lines in the background indicate all synteny blocks within the P. pyrifolia genome and the other three genomes, and blue lines indicate the duplicated SMXL gene pairs among them.

Figure 3 .
Figure 3. Synteny analysis of SMXL genes among P. pyrifolia, E. japonica, G. trifoliate, and P. persica.The gray lines in the background indicate all synteny blocks within the P. pyrifolia genome and the other three genomes, and blue lines indicate the duplicated SMXL gene pairs among them.

Figure 4 .
Figure 4. Phylogenetic tree of SMXL genes in pear.The ML tree was constructed with MEGA 11.0 software using amino acid sequences based on the WAG model; the numbers on the branches represent the bootstrap supports.SMXL genes are from P. pyrifolia.We calculated the Ks value of homologous gene pairs in each subfamily (TableS3), and then interpreted the replication nodes based on the Ks value.The common ancestor nodes marked by red stars represent the duplication event during the recent pear lineage-specific WGD.The common ancestor nodes marked by blue stars represent the duplication event during the ancient legume WGD.

Figure 5 .
Figure 5. Schematic representations for chromosomal distribution and interchromosomal relationships of pear SMXL genes.Differently colored lines indicate all synteny blocks in the pear genome.The blue lines indicate a gene pair that is duplicated with SMXL genes.The chromosome number is indicated in each chromosome bar.The lengths of the chromosomes are marked using scales, with each small scale indicating 500,000 amino acids, and every 10th scale is marked using a number.

Figure 4 .
Figure 4. Phylogenetic tree of SMXL genes in pear.The ML tree was constructed with MEGA 11.0 software using amino acid sequences based on the WAG model; the numbers on the branches represent the bootstrap supports.SMXL genes are from P. pyrifolia.We calculated the Ks value of homologous gene pairs in each subfamily (TableS3), and then interpreted the replication nodes based on the Ks value.The common ancestor nodes marked by red stars represent the duplication event during the recent pear lineage-specific WGD.The common ancestor nodes marked by blue stars represent the duplication event during the ancient legume WGD.

Figure 4 .
Figure 4. Phylogenetic tree of SMXL genes in pear.The ML tree was constructed with MEGA 11.0 software using amino acid sequences based on the WAG model; the numbers on the branches represent the bootstrap supports.SMXL genes are from P. pyrifolia.We calculated the Ks value of homologous gene pairs in each subfamily (TableS3), and then interpreted the replication nodes based on the Ks value.The common ancestor nodes marked by red stars represent the duplication event during the recent pear lineage-specific WGD.The common ancestor nodes marked by blue stars represent the duplication event during the ancient legume WGD.

Figure 5 .
Figure 5. Schematic representations for chromosomal distribution and interchromosomal relationships of pear SMXL genes.Differently colored lines indicate all synteny blocks in the pear genome.The blue lines indicate a gene pair that is duplicated with SMXL genes.The chromosome number is indicated in each chromosome bar.The lengths of the chromosomes are marked using scales, with each small scale indicating 500,000 amino acids, and every 10th scale is marked using a number.

Figure 5 .
Figure 5. Schematic representations for chromosomal distribution and interchromosomal relationships of pear SMXL genes.Differently colored lines indicate all synteny blocks in the pear genome.The blue lines indicate a gene pair that is duplicated with SMXL genes.The chromosome number is indicated in each chromosome bar.The lengths of the chromosomes are marked using scales, with each small scale indicating 500,000 amino acids, and every 10th scale is marked using a number.

Figure 6 .
Figure 6.Synteny network of SMXL family genes and syntenic relationships within and between three groups of genes.(a) The synteny network of SMXL family genes.Communities were rendered based on the clique percolation method at k = 4.The size of each node indicates the number of connected edges (the node decree).The communities are denoted by the four groups (Group I-IV) involved.(b) Syntenic relationships among the SMXL genes within the phylogenetic tree.Each connecting line located inside the inverted circular gene tree indicates a syntenic relationship between two SMXL genes.Lineage information is contained in the branches.

Figure 6 .
Figure 6.Synteny network of SMXL family genes and syntenic relationships within and between three groups of genes.(a) The synteny network of SMXL family genes.Communities were rendered based on the clique percolation method at k = 4.The size of each node indicates the number of connected edges (the node decree).The communities are denoted by the four groups (Group I-IV) involved.(b) Syntenic relationships among the SMXL genes within the phylogenetic tree.Each connecting line located inside the inverted circular gene tree indicates a syntenic relationship between two SMXL genes.Lineage information is contained in the branches.

Figure 7 .
Figure 7.The predicted three-dimensional (3D) structure of nine PpySMXL proteins.Using the protein homology modeling method based on the SMXL structure of the Swiss-Model database, the structure with the highest score was chosen as the optimal structure for the PpySMXL protein.

Figure 7 .
Figure 7.The predicted three-dimensional (3D) structure of nine PpySMXL proteins.Using the protein homology modeling method based on the SMXL structure of the Swiss-Model database, the structure with the highest score was chosen as the optimal structure for the PpySMXL protein.

Figure 8 .
Figure 8. Expression pattern of PpySMXL genes under fruit-thinning treatment at different stages.The transcript levels of PpySMXL genes at 0 d, 1 d, 3 d, 5 d, 7 d, 14 d, and 21 d were investigated based on transcriptome data.The expression level of PpySMXL genes is shown as a heatmap of log2 (FPKM) values.The color scale shows increasing expression levels from blue to red.

Figure 8 .
Figure 8. Expression pattern of PpySMXL genes under fruit-thinning treatment at different stages.The transcript levels of PpySMXL genes at 0 d, 1 d, 3 d, 5 d, 7 d, 14 d, and 21 d were investigated based on transcriptome data.The expression level of PpySMXL genes is shown as a heatmap of log2 (FPKM) values.The color scale shows increasing expression levels from blue to red.

Figure 9 .
Figure 9. (a) Bud development after 14 days of fruit thinning (FT) treatment; blank treatments act as controls (CK).(b) Gene expressions of SMXL genes under fruit-thinning treatment determined via qRT-PCR.The mean expression values were calculated from three independent replicates.The Xray indicates the number of days after treatment.Mean values and standard errors were calculated from three replicates.A T-test was used for statistical analysis.The asterisk and double asterisks represent significant differences at the levels of 0.05 and 0.01, respectively.R.e.l indicates the relative expression level.

Figure 9 .
Figure 9. (a) Bud development after 14 days of fruit thinning (FT) treatment; blank treatments act as controls (CK).(b) Gene expressions of SMXL genes under fruit-thinning treatment determined via qRT-PCR.The mean expression values were calculated from three independent replicates.The X-ray indicates the number of days after treatment.Mean values and standard errors were calculated from three replicates.A T-test was used for statistical analysis.The asterisk and double asterisks represent significant differences at the levels of 0.05 and 0.01, respectively.R.e.l indicates the relative expression level.

Figure 10
Figure10.Gene expressions of selected genes under fruit-thinning (FT) treatment determined via qRT-PCR, with blank treatments as controls (CK).The mean expression values were calculated from three independent replicates.The X-ray indicates the number of days after treatment.Mean values and standard errors were calculated from three replicates.A t-test was used for statistical analysis.The asterisk and double asterisks represent significant differences at the levels of 0.05 and 0.01, respectively.R.e.l indicates the relative expression level.

Figure 10
Figure10.Gene expressions of selected genes under fruit-thinning (FT) treatment determined via qRT-PCR, with blank treatments as controls (CK).The mean expression values were calculated from three independent replicates.The X-ray indicates the number of days after treatment.Mean values and standard errors were calculated from three replicates.A t-test was used for statistical analysis.The asterisk and double asterisks represent significant differences at the levels of 0.05 and 0.01, respectively.R.e.l indicates the relative expression level.