Abstract:Ornamental plants are important horticultural crops， and their traits such as plant shape， flower color， flower form， flower fragrance， flowering habitat， prickle directly determine their ornamental and market value. Modern molecular breeding techniques are effective ways for rapidly developing new varieties. However， the lack of localization and mining of key genes for important traits greatly limits the application of these techniques in ornamental plants. Genetic map is a linear arrangement map of the relative positions of chromosomal or intra-genomic loci based on the recombination frequency of genetic markers. It is a crucial tool for studying the genetic laws and molecular mechanisms of complex traits in organisms， providing a foundation for the localization， identification and genetic mechanism analysis of key genes for traits， and offering theoretical and practical guidance for molecular breeding. In recent years， with the development of sequencing and molecular biology techniques， some progress has been made in genetic map construction and quantitative trait loci （QTL） mapping of ornamental plants， but it lacks systematic induction and summarization. This paper reviews the research progress on the construction of genetic maps and QTL mapping for important ornamental traits including plant shape， flower color， flower fragrance， petal number， and resistance-related traits in ornamental plants such as roses， orchids， and chrysanthemums， etc. It also discusses the existing problems， solutions and future development in the construction and application of genetic maps in ornamental plants， aiming to provide references and theoretical guidance for molecular breeding of ornamental plants.

Abstract:Bakanae disease， caused by Fusarium fujikuroi， is a globally important fungal disease of rice. The pathogen produces gibberellins， cell wall-degrading enzymes， and effector proteins that interfere with host metabolic processes and suppress immune responses， ultimately leading to elongate seedling growth and reduce yield. This article systematically reviews the application of artificial inoculation methods and molecular marker-assisted genotyping techniques in the evaluation of rice resistance to bakanae disease. It compares existing screening results and evaluation criteria for resistant germplasm resources， laying an important foundation for in-depth exploration of key disease-resistant genes and the development of resistant rice varieties. Furthermore， in response to challenges such as the scarcity of resistant germplasm， this paper proposes that future efforts should integrate high-throughput screening with multi-environment field validation to establish a standardized， reproducible， and widely applicable resistance evaluation platform for bakanae disease across diverse genetic backgrounds. This will provide a solid basis for further mining of critical resistance genes and the assessment of resistant resources.

Abstract:Throughout their growth and development， plants are exposed to numerous adverse environmental conditions. In response to these stresses， plants synthesize secondary metabolites such as lignin and lignans， which contribute to cell wall reinforcement and defense. Dirigent （DIR） proteins were characterized as a family of auxiliary proteins in plants， typically containing a conserved dirigent domain. They have been implicated in lignin polymerization and lignan biosynthesis by mediating the stereo- and regioselectivity of oxidative coupling reactions. In recent years， DIR family in various plant species have been extensively studied for their roles in plant growth， development， and responses to biotic and abiotic stresses. This review comprehensively focuses on recent advances in the structural characteristics， subfamily classification， and biological functions of the DIR family members in plants. It particularly emphasizes elucidating the mechanisms by which DIR genes respond to stresses such as drought， low temperature， and pathogen infection. The information synthesized in this review provides the theoretical basis for elucidating how DIR genes regulate plant growth， development， stress tolerance， and quality formation. It will ultimately guide the development of elite germplasm characterized by high yield， superior quality， and broad-spectrum stress tolerance.

Abstract:Based on The Third National Survey and Collection of Crop Germplasm Resources， along with subsequent supplementary collections， a total of 216 glutinous rice landrace germplasm resources were obtained from Jiangxi province. The distribution patterns and germplasm types of these resources were analyzed，and their morphological characteristics， agronomic traits， and quality characteristics were evaluated. A comprehensive assessment and selection of superior local glutinous rice germplasms were carried out. From the perspective of geographical distribution， the glutinous rice landrace germplasm resources in Jiangxi province are mainly concentrated in five less developed regions： Shangrao， Ganzhou， Fuzhou， Ji'an and Yichun. These landraces exhibited high genetic diversity and harbored two subspecies： indica and japonica， with indica being the predominant type. Additionally， these resources can be categorized into early， medium， and late varieties according to their maturity， with medium and late rice varieties predominating. Moreover， the seed coat color varies widely， and eight colored glutinous rice resources were identified. Most of the spikelets are yellow， with either no or short awns， suitable panicle， moderate blade and stem angles， resistance to lodging， and suitable seed holding. These characteristics indicate great potential for further utilization. Through two years repeated evaluation of the agronomic and quality traits of local rice resources in Jiangxi province， ten glutinous rice germplasm resources were selected， which showed large panicles， multiple grains， high seed setting rate and high 1000-grain weight. Among them， the effective number of panicles per plant ranged from 8.2 to 10.3， the number of grains per panicle ranged from 183.4 to 236.6， the 1000-grain weight ranged from 23.2 to 27.2 g and the seed setting rate ranged from 84.3% to 94.0%. These germplasm resources have the characteristics of high yield and good quality， which have considerable values for breeding application. These elite germplasm resources selected in this study have great potential for further utilization， which could provide a material basis for the genetic improvement of high yield and quality in glutinous rice.

Abstract:To develop high-yield and high-quality maize varieties and alleviate the supply-demand pressure arising from the difficulty in balancing high yield and high quality， this study evaluated protein， starch， and oil content， three key quality traits， in 588 elite maize inbred lines （across four locations） and 939 landraces （over two consecutive years at a single location） using a near-infrared analyzer. By integrating multi-year and multi-location data， we analyzed phenotypic variation， compared trait differences between the two populations， and examined trait correlations. The results indicate that： （1） Landraces exhibited high genetic diversity， with shannon-wiener index H′ exceeding 2.000 for all three quality traits. Oil content displayed the highest phenotypic variation （CV=8.66%） and starch content showed the lowest （CV=2.46%）. Protein content was significantly higher in landraces from the southwest region than those from the Huanghuai region （P=1.4×10^-8）， whereas starch content was significantly lower （P=0.00164）. （2） Elite inbred lines also exhibited rich diversity with H′ values of 2.075， 2.049， and 1.980 for protein， starch， and oil， respectively. The coefficient of variation for protein and oil content exceeded 10% in certain environments， and both traits generally decreased in content from south to north regions. （3） Landraces possessed significantly higher protein （P=1.08×10^-64） and oil content （P=4.4×10^-136） than elite inbred lines， through starch content did not differ significantly between the populations. （4） Significant correlations （P<0.01） were observed among the three traits： starch content was negatively correlated with both protein and oil content， whereas protein and oil content were positively correlated. （5） We identified four accessions with high starch and protein content， three with high oil and starch， and six with high oil and protein， providing valuable genetic resources for analyzing the synergistic improvement mechanism of yield and quality. Additionally， extreme-quality materials ranking in the top 10 for protein， starch， and oil content， respectively， including widely used core inbred lines such as HZ32， PH4CN， and Su37， were selected. In summary， both landraces and elite inbred lines harbor considerable variation in quality-related traits. This study provides a foundation resource for developing high-quality maize varieties and offers practical insights for quality-oriented breeding breeding.

Abstract:An evaluation of key agronomic traits and quality indicators was conducted on Perilla germplasm resources obtained from inventory and the Third National General Survey and Collection Action of Crop Germplasm Resources. Analysis of 22 agronomic and quality traits across 450 perilla accessions revealed coefficient of variation （CV） ranging from 6.1% to 206.4%. Perilla aldehyde content exhibited the highest CV， whereas α-linolenic acid content exhibited the lowest. The Shannon diversity index ranged from 0.90 to 4.00， with the lowest and highest values corresponding to perilla aldehyde and rosmarinic acid contents， respectively. Assessment of the adaptability of perilla accessions from diverse geographical origins， along with the effects of spatiotemporal variation on phenotypic traits， indicated that southward introduction of high-latitude germplasm carries a high risk of premature senescence. In contrast， transfer between similar environments proved more reliable. Despite no significant differences were observed in rosmarinic and caffeic acid content across differently colored perilla leaves， extreme values were detected in double-sided green varieties. Furthermore， breeding efforts over the past three decades appear to have prioritized increased leaf size and single-plant yield， with less focus on the directed selection for grain size. A total of 40 elite perilla germplasm resources were identified and classified into three groups via K-means clustering. Group I was characterized by small grain size and low oil content， and included most accessions with high rosmarinic acid levels. Group II featured large grain size and high oil content， whereas Group III exhibited medium grain size， high leaf weight per plant， high grain yield， and elevated levels of α-linolenic and caffeic acids. This study underscores the rich genetic diversity and environmental adaptability of perilla germplasm resources in China. The findings provide valuable data for germplasm identification and a theoretical basis for the strategic distribution and utilization of perilla germplasm across different ecological regions.

Abstract:Glycine tabacina （Labill.） Benth.， a perennial species of the subgenus Glycine， is endemic to Fujian province， China， and was listed as a National Grade Ⅱ Key Protected Wild Plant. Rational population sampling to maximize the preservation of genetic diversity is crucial for germplasm collection of this species. In this study， species-specific SSR markers were employed to analyze both inter-population genetic structures across 13 natural populations and intra-population genetic structures in a single-population. The results show that geographically there were extremely low gene differentiation （F_st=0.056）， high expected heterozygosity （H_e=0.741）， high polymorphism information content （PIC=0.70）， and substantial gene flow （N_m=4.445）； intra-population variation （H_s accounting for 78.0% of the total variation） exceeded inter-population differentiation （D_st accounting for 22.0% of the total variation）. The results of this study clearly support that： although G. tabacina is a self-pollinating plant， its geospatial population genetic variation characteristic resembles that of outcrossing plants， which may be attributed to the species' perennial nature facilitating high genetic heterogeneity and extensive long-distance dispersal enabling high geographical gene flow. The fine-scale spatial genetic structure analysis within the single population （Longhai） with higher-level disturbance revealed that the size range of spatial genetic patches among plants was about 5.0 m. When the sampling size was 20-25 plants， 95% of the population's genetic diversity was captured. Based on the findings of this study， we propose the following strategies for collecting G. tabacina germplasm resources： at the population level， at least 20-25 individual plants should be sampled within a population， with an interval of at least 9-15 m between sampled plants. At the geographical spatial level of populations， since G. tabacina in China is only distributed in a narrow coastal area of Fujian province， it is recommended to collect as many populations as possible.

Abstract:Based on 129 hyacinth bean landraces collected from Zhejiang province during "The Third National Census and Collection Campaign of Crop Germplasm Resources"， this study systematically investigated their agronomic and quality traits. Combined with single nucleotide polymorphisms （SNPs） obtained from resequencing datasets， we conducted a comprehensive assessment of their genetic diversity of these landraces. The results revealed abundant genetic diversity in phenotypic traits among the hyacinth bean resources. Among the 13 qualitative traits analyzed， 48 variant types were identified， with Shannon diversity index （H’） ranging from 0.091 to 1.486. Pod shape exhibited the highest H’， while young leaf color showed the lowest. Leaf color was predominantly green （98%）， main stem color was mainly purple （49%）， standard petal and wing petal colors were primarily light purple （69%） and purplish-red（69%）， respectively. Pod color was mostly light green （35%）， pod shape was mainly falcate （42%）， seed shape was predominantly elliptical （57%）， and seed color was mainly black （49%）. For the 23 quantitative traits， the coefficients of variation （CV） ranged from 2.88% to 43.26%， with grain yield per plant having the highest CV， and water content the lowest. The Shannon diversity index （H’） for quantitative traits ranged from 1.83 to 2.25， where soluble sugar content displayed the highest diversity and maturity duration the lowest. Principal component analysis （PCA） based on SNPs classified the 129 landraces into four clusters. Cluster I had the lowest cellulose content （6.48%） and the highest crude protein content （20.34%）； Cluster III exhibited the highest water content （87.53%）， soluble sugar content （21.88%）， crude fat content （5.99%）， and free amino acid content （7.98 mg/g）； Cluster Ⅳ consisted of medicinal hyacinth beans with white flowers and white seeds. This study systematically elucidated the genetic diversity of lablab germplasm resources in Zhejiang province at the phenotypic， quality， and genotypic levels， providing a scientific basis for germplasm innovation and the breeding of new varieties.

Abstract:This study investigated 57 superior Hevea brasiliensis clones collected from first-generation rubber plantations in China. A total of 25 quantitative traits were measured， encompassing stem circumference， leaf morphological characteristics， and anatomical structure. The variation extent， genetic diversity indices， and inter-trait correlations were analyzed. Systematic cluster analysis and principal component-based evaluation were subsequently conducted using this comprehensive phenotypic dataset. The results revealed highly significant differences among the 57 clones for all 25 traits. The coefficient of variation （CV） ranged from 7.26% to 45.29%. The largest CV was observed for stem circumference at the seedling stage （first year）， while the smallest was for the ratio of left vein width to leaf width. The phenotypic genetic diversity index ranged from 1.018 to 2.188， with leaf area showing the highest genetic diversity index and the left vein width/leaf width ratio exhibiting the lowest. Among the 25 traits， most showed highly significant or significant correlations with one another. Exceptions were the ratio of left vein width to leaf width， leaf vein number， midrib thickness， leaf vein angle， palisade to spongy tissue ratio， cell structural tightness ratio， and cell structural looseness ratio， which showed no significant correlation with most other traits. Notably， stem circumference was strongly positively correlated with leaf area， leaf length， and leaf thickness， with correlation coefficients all exceeding 0.8. Among leaf anatomical traits， indicators such as leaf thickness， cuticle thickness， palisade tissue thickness， spongy tissue thickness， and upper epidermis thickness exhibited highly significant or significant pairwise correlations， with the exception of palisade tissue thickness and spongy tissue thickness， which was not significantly correlated. Moreover， these anatomical traits were also significantly correlated with most leaf morphological traits. Systematic cluster analysis categorized the 57 clones into six distinct groups， with no apparent relationship to their geographic origins. Group I contained 29 clones. Groups IV and V exhibited desirable traits， such as larger stem circumference and leaf area， sparse vein distribution， and greater thickness across various leaf anatomical parameters. Based on cluster analysis and principal component comprehensive evaluation， six superior clones， with numbers 7， 10， 33， 27， 46， and 14， were identified as fast-growing and drought resistant. By the seventh year， all six clones attained a stem circumference exceeding 52 cm. Among them， clone 7 and 10 were triploids. This study highlights the phenotypic diversity present in scattered superior tree resources within China′s rubber planting areas， providing valuable genetic materials for parental selection and targeted breeding in H. brasiliensis.

Abstract:In this study， a multiplex PCR-based SNP genotyping panel was developed using genotyping-by-sequencing （GBS） data， which includes 170 core SNP markers. Based on this SNP panel and 30 phenotypic traits， analysis was conducted on 202 pepper germplasms to resove their genetic diversity and explore the associated loci for excellent phenotypic traits of pepper. Research showed that a total of 170 core SNP markers distributed across 12 chromosomes， with the average Shannon's information index （I）， polymorphic information content （PIC） and Nei's gene diversity （H′） being 0.616， 0.334 and 0.428， respectively. This indicates high genetic diversity among the tested materials. Results from cluster analysis， population structure analysis， and principal component analysis were consistent： all divided the tested materials into five groups， and each group showed a certain correlation with geographical origin and fruit morphology. The coefficient of variation （CV） for the 30 phenotypic traits ranged from 7.00% to 87.88%， with an average of 34.85%. The Shannon-Wiener diversity index varied from 0.03 to 2.07， with an average of 1.19. Among these traits， single fruit weight showed the highest CV （87.88%）， while fruit length had the highest Shannon-Wiener diversity index （2.07）. Corolla color exhibited the lowest values of both CV （7.00%） and Shannon-Wiener diversity index （0.03）. Most of the 30 phenotypic traits showed significant or highly significant correlations. Furthermore， association analysis was performed between phenotypic traits and SNP markers. The results showed that a total of 53 SNPs were detected by two methods （GLM and MLM）， which were significantly associated with 12 phenotypic traits. GLM and MLM could explain 4.83%-48.41% and 10.86%-19.19% of the phenotypic variance， respectively. Marker 980-003 located on chromosome 4 explains the highest phenotypic variance for corolla color. Four of the 53 SNPs were simultaneously detected by both methods. The SNP genotyping panel developed in this study is a genotyping method with high throughput， good accuracy， and low cost， which can be applied to genetic structure analysis， molecular marker-assisted breeding， and variety identification of peppers. In addition， this study established a database of phenotypic traits and genotypes for 202 pepper germplasms， effectively establishing the correspondence between phenotypes and genotypes， and providing theoretical guidance and a material basis for the exploration of excellent pepper genes， germplasm innovation and the genetic improvement of varieties.

Abstract:Peanut （Arachis hypogaea L.） is an important economic and oilseed crop. The red testa in peanut is associated with higher anthocyanin content， which enhances its antioxidant and anti-inflammatory properties， highlighting the commercial potential of breeding new varieties with red testa. Here， we investigated the genetic basis of red testa color using three segregation populations generated from crosses between two red-testa varieties， Zhenzhuhong 1 （ZZH1） and Zhanhong 2 （ZH2）， and one pink-testa variety， Yuhua 15 （YH15）. The F1 progeny of YH15×ZZH1 and ZH2×ZZH1 displayed red testa， while those of ZH2×YH15 had pink testa， indicating that red testa is inherited as a dominant trait in ZZH1 but as a recessive trait in ZH2. Segregation analysis in the F2 populations showed that testa color in YH15×ZZH1 controlled by a single dominant gne， while that in ZH2×YH15 was governed by a single recessive gene. In contrast， the F2 population of ZH2×ZZH1 exhibited a segregation ratio consistent with control by two genes. These results suggest that red testa can be classified into dominant and recessive types， each regulated by distinct genes. Using bulked segregant analysis （BSA） with non-segregating red- and pink-testa lines from the YH15×ZZH1 population， we mapped a major quantitative trait locus （QTL） for dominant red testa to a 0.898 Mb region on chromosome 3. This QTL exhibited LOD scores ranging from 113.18-139.16 and accounted for 95.2%-97.6% of the phenotypic variance. Furthermore， by evaluating two newly designed markers according to previously reported markers and one newly developed marker across three populations， we found that the marker Tif2.A03.126107404 was completely linked to the dominant red testa trait， while Tif1.A03.125867405 was highly linked to it. In addition， the marker Tif1.A12.117190528 showed completely linkage to the recessive red testa. These markers provide valuable tools for molecular marker-assisted breeding of red testa in peanut.

Abstract:Metallopeptidase 24 （M24） is a type of protein， which shares a common structural-fold， the ‘pita bread’ conserved domain widely exists in animals， plants and microorganisms and plays an important role in individual development and organ sculpting as well as the environmental stress response.In this study， in order to systematically study the characteristics of OsM24 family members in rice， the bioinformatics analysis and the qRT-PCR methods were carried out. The results showed that 16 OsM24 members were identified from rice genome and found to be unevenly distributed on 9 of 12 rice chromosomes.There were significant differences among OsM24 family members in terms of basic physicochemical properties such as the number of amino acids， protein stability and hydrophobicity etc.Their phylogenetic relationship suggested that OsM24 family members could be classified into 3 subfamilies OsM24I， OsM24II and OsM24III and members within the same subfamily exhibited conserved gene structure and motif compositions.GO ontology （GO） and predicted protein-protein interaction networks （PPI） analysis suggested that OsM24 family members exerted their functions in protein synthesis， processing， and modification by relying on their hydrolase activity. Cis-acting elements analysis indicated that there were a number of diverse regulatory elements related to hormones， stress， light and plant growth and development in the promoter region of OsM24 family member genes. The expression patterns analysis showed the expression of OsM24 genes in rice was significantly tissue-specific.The qRT-PCR analysis suggested apart from OsM24-7， OsM24-12， OsM24-15 and OsM24-16， the remaining 12 OsM24 genes all are involved in salt and alkaline stress. However， their expression patterns differed under salt stress and alkali stress conditions.These suggests that the OsM24 proteins may respond to salt and alkali stress through different regulatory pathways.Collectively， this study provides an important theoretical reference for further research on the role of M24 in rice development and saline-alkali stress response.

Abstract:Gummy stem blight is one of the major diseases threatening watermelon production. Identifying and utilizing resistance genes to gummy stem blight is of great significance for the innovation of disease resistant watermelon germplasm and breeding of resistant watermelon varieties. In this study， a recombinant inbred line （RIL） population derived from the cross between the gummy stem blight resistant germplasm PI189225 and the susceptible germplasm K3 was used as the experimental materials. Bulked segregant analysis （BSA） combined with whole genome resequencing of the parents and resistant and susceptible bulks was performed to identify candidate genomic regions associated with gummy stem blight resistance. Additionally， transcriptome sequencing （RNA-seq） data were integrated to mine candidate genes. The results showed that BSA-seq analysis identified genome regions significantly associated with gummy stem blight resistance on chromosome 5 and 10， with a total length of 8.18 Mb containing 681 genes. Functional analysis revealed that these genes are primarily involved in biological processes such as plant-pathogen interaction and phenylpropanoid biosynthesis. Further analysis using InDel markers and recombinant individuals narrowed down the resistance interval to a 2.45 Mb region on chromosome 10， flanked by the markers Chr10_30103333 and Chr10_32554279. Combined with differentially expressed genes in watermelon responding to gummy stem blight infection， six candidate genes were identified on chromosome 10. qRT-PCR analysis indicated that the expression of all six candidate genes was induced by gummy stem blight pathogen. Among them， Cla97C10G200140 and Cla97C10G202140 were highly expressed in susceptible parent， whereas Cla97C10G200100， Cla97C10G201690， Cla97C10G202570， and Cla97C10G201940 were highly expressed in the resistant parent. The findings of this study provide important theoretical basis and genetic resources for marker assisted selection and breeding of gummy stem blight resistant watermelon varieties.

Abstract:Rhodiola crenulata is a medicinal plant with a long history， widely used in traditional Chinese medicine and Xizang medicine. However， due to its extremely special growth environment， collection and research are very difficult， and there are currently few reports on molecular biology. This study firstly cloned the RcNAC83 gene from the Xizang Rhodiola crenulata. Through bioinformatics and real-time quantitative PCR techniques， the physicochemical properties and expression patterns of this gene were preliminarily understood. Using chromosome walking technology， the promoter sequence was predicted to identify cis-acting elements. Yeast expression vectors were constructed to verify its transcriptional activation activity， and it was ectopically expressed in Salvia miltiorrhiza to analyze the changes in phenotypes and physiological indicators under salt stress. The results showed that the coding region of RcNAC83 gene was 918 bp， containing 3 exons， and encoded 248 amino acids. The protein had obvious hydrophilicity and no transmembrane domain. It was subcellularly located in the nucleus and contained multiple phosphorylation sites. The N-terminal of the RcNAC83 gene could be divided into 5 subdomains， among which D domain was the NARD region， which had a similar amino acid sequence and conserved domain to the homologous genes， and was closely related to the genus Kalanchoe genes. The RcNAC83 promoter sequence was 888 bp and contained multiple cis-acting elements for transcriptional activation， light response， anaerobic induction， plant hormone MeJA， and recognition by MYB， MYC. RcNAC83 was expressed in roots， mature stems， stems， leaves， apical buds and flowers， and was induced by various abiotic stresses and plant hormones. This gene had no toxicity to yeast cells and transcriptional activation activity. In overexpressing Salvia miltiorrhiza， the RcNAC83 gene did not affect the growth and metabolism of Salvia miltiorrhiza， but reduced the tolerance to salt stress. This study will further enrich the biological functions of members of the plant NAC gene family， and provide a theoretical basis for revealing the molecular mechanism of highland adaptation in Rhodiola crenulata.

Abstract:In order to elucidate the molecular mechanism of lettuce seed germination under heat stress， the transcriptome， metabolome and multi-omics combined analysis were performed on the seeds of two lettuce varieties， heat-tolerant cultivar Italian Lettuce and heat-sensitive cultivar Four-Season Fodder Lettuce. Untargeted metabolomics identified 391 differentially expressed metabolites （DEMs）， revealing specific activation of the tyrosine metabolism and flavonoid biosynthesis pathways. Metabolites including muconic acid， L-tyrosine， L-tyramine， and p-hydroxycinnamic acid were significantly enriched. Transcriptome analysis detected 3127 differentially expressed genes （DEGs）， indicating two distinct genetic regulatory mechanisms associated with early thermotolerance in lettuce germination： （1） Unique upregulation of heat shock protein （HSP）-related genes in Italian Lettuce； （2） Activation of β-glucanase-coding genes by AP2/ERF transcription factors， facilitating repair of heat-damaged cell walls. Key candidate genes were initially screened using Gene Set Enrichment Analysis （GSEA） combined with protein-protein interaction （PPI） network analysis. qRT-PCR validation corroborated the RNA-seq findings. Integrated multi-omics analysis demonstrated significant downregulation of ethylene biosynthesis genes （ACO， 1-Aminocyclopropane-1-Carboxylate Oxidase） in Four-Season Fodder Lettuce. Concurrent suppression occurred in core flavonoid pathway genes， β-glucosidase-encoding genes， and biosynthesis of aromatic amino acids （tyrosine）. These results indicate that heat stress inhibits seed germination through a dual mechanism involving metabolic flux blockage and inactivation of the antioxidant defense system. This study delineates the gene-metabolite regulatory network underlying the heat-stress response during lettuce seed germination， providing a theoretical foundation for the molecular interactions governing germination inhibition.

Abstract:This study systematically elucidated the molecular mechanisms by which plant hormones regulate the polygalacturonase （PG） gene family during lettuce seed germination， through an integrated approach involving microscopic morphology， hormone profiling， enzyme activity assays， and transcriptome sequencing. Scanning electron microscopy revealed that the radicle penetrated the endosperm cap to form a tearing surface after 15 h of imbibition， with penetration completed by 24 h. Germination assay showed nearly 100% germination by 36 h. Treatment with 10 μmol/L gibberellin （GA?） and 1 mmol/L ethylene （ETH） shifted the peak germination probability forward to 18-18.5 h， accompanied by radicle thickening and increased root hair density. In contrast， treatments with 1 μmol/L abscisic acid （ABA）， 25 μmol/L jasmonic acid （JA）， and 10 μmol/L indole-3-acetic acid （IAA） delayed peak germination probability to 21.5 h， 27.5 h and 29 h， respectively， while also inhibiting radicle elongation. Analysis of endogenous hormones showed ABA， abscisic acid glucose ester （ABA-GE）， JA， and salicylic acid （SA） levels progressively declined during germination， whereas IAA and 1-aminocyclopropane-1-carboxylic acid （ACC） exhibited sustained increases. The superoxide anion （O₂^·-） production rate in imbibed seeds from the control group initially increased then declined. Treatments with GA₃， SA， and ETH enhanced the O₂^·- generation rate， while treatments with ABA， IA， and JA had minimal effect. LsPG enzyme activity exhibited distinct dynamics across treatments： activity peaked at 30 h for water/ETH， at 36 h for GA₃/ABA/IAA， and at 24 h for SA/JA. Notably， ETH， GA₃， and SA treatments most significantly enhanced LsPG enzyme activity. Through genome-wide analysis， 54 LsPG genes were identified for the first time in lettuce， clustering into seven subfamilies and being unevenly distribution across nine chromosomes. Subcellular localization experiments confirmed that LsPG4 and LsPG38 proteins localize to the cell membrane. Conserved motif analysis revealed that Motif 3 was ubiquitously present， while Motif 2/4 were widely distributed. All LsPG genes contained 2 to 10 exons. Cis-acting element analysis revealed that LsPG1/7/24 were enriched in hormone-responsive elements， with ABA-responsive elements being the most abundant. qRT-PCR analysis revealed hormone-specific expression patterns of LsPG genes： GA₃ significantly upregulated LsPG4/20/47/53 （4.5/3.2/3.3/3.1 fold）； ETH specifically activated LsPG47 （3 fold）； IAA induced a 7.3 fold increase in LsPG20 expression， JA induced LsPG2/13/22； SA suppressed LsPG19/20/24， with LsPG19 expression being inhibited by ABA/SA/GA₃/IAA/ETH/JA. Overall， this study demonstrates that GA₃/ETH promote endosperm cell wall degradation via LsPG activation to accelerate germination， whereas ABA and SA delay germination through LsPG inhibition. The hormone-PG gene interaction network model provides a theoretical basis for deciphering the mechanism of lettuce seed germination and molecular breeding.