Abstract Background Melon is an important horticultural crop cultivated extensively worldwide. The size of the fruit navel, the terminal region of melon fruits, significantly influences the appearance quality of the fruit. However, the regulatory factors and molecular mechanisms governing the fruit navel development remain poorly understood in melon. Results In this study, the regulators and mechanisms underlying fruit navel development were investigated through phenotypic analysis, RNA sequencing (RNA-seq) and RT-qPCR methods. The inbred line ‘T03’ and a big fruit navel mutant ‘BFN’ of melon were used as experimental materials. RNA-seq analysis identified 116 differentially expressed genes (DEGs), including 54 up-regulated and 62 down-regulated genes, in both the green bud (GB) and ovary at anthesis (OA) stages of the ‘BFN’ melon compared to the ‘T03’ melon. Functional enrichment analysis revealed that these 116 DEGs were significantly associated with “Sesquiterpenoid and triterpenoid biosynthesis”, “Circadian rhythm—plant”, “Galactose metabolism” and “Biosynthesis of various alkaloids” pathways. There were three (AP2/ERF, MYB and C2H2 types) and eight (AP2/ERF, MADS-box, homeobox domain and bZIP types) transcription factors presented in up-regulated and down-regulated DEGs, and their putative target genes were predicted. Based on KEGG and expression analyses, two terpene cyclase/mutase genes (MELO3 C001812 and MELO3 C004329) were identified as being involved in the “Sesquiterpenoid and triterpenoid biosynthesis” pathway, and their transcripts were significantly downregulated in all detected development stages (EGB, GB, GYB, YB and OA) of ‘BFN’ fruits compared with ‘T03’ fruits. Conclusions The findings of this study elucidate a fundamental regulatory mechanism underlying fruit navel formation, and identify two key negative regulators, MELO3C001812 and MELO3C004329, involved in the development of the fruit navel in melon. Supplementary Information The online version contains supplementary material available at 10.1186/s12870-025-06444-7. Keywords: Melon, Fruit navel development, Transcriptome analysis, Expression analysis Background Melon is an important horticultural crop, highly valued by consumers for its sweet and flavorful fruit. The flowers of melon are classified into male flowers, female flowers and bisexual flowers [[36]1]. Most melon cultivars exhibit andromonoecy in which bisexual flowers develop into fruits [[37]2]. Under optimal growth conditions, bisexual flowers typically take approximately 5 days to progress from visible buds to full bloom [[38]3]. During anthesis, pollen from male flowers lands on the stigma of the pistil and germinates, the ovule is fertilized in virtue of the growth of pollen tubes [[39]4]. Subsequently, the fruit undergoes rapid growth driven by cell division and expansion. In melons, the morphology of ovary and mature fruit is closely correlated, and the fruit shape is predominantly determined prior to flowering [[40]3]. To date, several quantitative trait loci (QTLs) have been identified as regulators of fruit length (flqs2.1, flqs3b.1, flqs6a.1, flqs8.1, flqs10b.1), fruit diameter (fdqs3a.1, fdqs12.1) and fruit shape (fsqs2.1, fsqs8.1, fsqs12.1) in melon [[41]5]. The fruit navel is also a critical component of the exterior quality in melon fruit, and the size of the fruit navel greatly influences consumer preference and its associated economic value. However, the molecular mechanisms underlying fruit navel formation in melon remain largely unexplored. The fruit navel is the residual trace left after the flower abscises, a prominent scar often appears due to the interplay of genotype, environmental factors and hormonal regulation, and compromises the aesthetic quality of the fruit [[42]6–[43]8]. In tomato, low temperature and growth regulators are important factors contributing to the formation of navel-like scars at the blossom end of fruits [[44]9]. Under low temperature condition, abnormal development of flower organs leads to the production of navel-like scars at the fruit apex [[45]10]. Exogenous application of gibberellic acid (GA[3]), N-methatolylphthalamic acid, or 2,3,5-triiodobenzoic acid (TIBA) promotes the formation of large fruit navels in tomato, and the severity of large fruit navel is connected with the plant's developmental stage at the time of induction [[46]9, [47]11–[48]13]. Specifically, auxin plays a direct role in the development of pointed-tip fruits in tomato. Knockdown of auxin response factor 7 (ARF7) leads to the production of heart-shaped fruits [[49]14]. The POINTED TIP (PT) gene, which encodes a C2H2-type zinc finger transcription factor, has been demonstrated to suppress the formation of pointed-tip fruits by downregulating the transcription of FRUTFULL 2 (FUL2) and altering auxin transport. Consequently, knockout of PT results in the development of pointed-tip fruits in tomato [[50]15]. In melon, low temperature alters the ultrastructure of the phloem, which is responsible for the photoassimilate transport, and inhibits the phloem loading [[51]16, [52]17]. Additionally, low temperature also results in decreased carbohydrate content in fruits, an altered plant metabolism and a significant loss of productivity [[53]18, [54]19]. At present, the effects of environmental factors and hormones on the formation of fruit navel in melon remain poorly understood, and the key regulatory genes involved in this process have yet to be identified. Currently, RNA-seq analysis is widely utilized to identify candidate genes and regulatory pathways associated with key developmental traits in plants. For instance, RNA-seq has been employed to investigate the mechanism underlying the formation of bisexual flowers in melon, revealing that the ethylene signaling pathway plays a crucial role in the initiation of bisexual buds [[55]2]. In this study, the dynamics of fruit navel development were examined using the melon inbred line ‘T03’ and a big fruit navel mutant ‘BFN’. Transcriptome analysis was conducted at the green bud (GB) and ovary at anthesis (OA) stages of ‘T03’ and ‘BFN’ fruit navels to explore the molecular mechanisms governing fruit navel development in melon. The results of this study revealed several potential regulators and regulatory networks involved in fruit navel development, and providing valuable insights for improving the exterior quality of melon as well as its consumer appeal. Methods Plant materials and growth conditions This study was conducted from March 2023 to August 2024. The melon inbred line ‘T03’ (Cucumis melo ssp. agrestis Jeffrey) and a spontaneous mutant ‘BFN’ (Cucumis melo ssp. agrestis Jeffrey) were used for morphological observation, transcriptome sequencing and reverse transcription quantitative polymerase chain reaction (RT-qPCR) analyses. The melon inbred line ‘T03’ is a widely cultivated variety in Hebei province of China, while ‘BFN’ is a spontaneous mutant derived from ‘T03’. The primary difference between ‘T03’ and ‘BFN’ is the size of fruit navel. The melon seedlings were cultivated in the greenhouse of Hebei Agricultural University at 25–28℃/16–19℃ of light/dark of 16/8 h. Standard irrigation and pest management practices were followed throughout the experiment. RNA extraction and sequencing Total RNA was extracted from the terminal region of ‘T03’ and ‘BFN’ melon fruits at five developmental stages (early green bud (EGB), green bud (GB), green yellow bud (GYB), yellow bud (YB) and ovary at anthesis (OA) stages) using the TianGen Quick RNA Isolation Kit (TianGen Biotech, China). The quality of the isolated total RNA was evaluated by the NanoDrop2000 spectrophotometer (IMPLEN, CA, USA) and 1% RNase-free agarose gel electrophoresis. Samples meeting the quality criteria were selected for further analyses, and the samples at GB and OA stages were chosen to perform RNA-sequencing. Independent samples of melon fruits collected before flowering were used for RNA-seq data validation and expression analyses of several enzyme-coding genes. For RNA sequencing, ten and three fruit navel tissues (indicated by the white frame in Fig. [56]1A) were pooled to form one biological replicate for the GB and OA stages, respectively. Three biological replicates were performed for each set (GB-T03, GB-BFN, OA-T03, OA-BFN). RNA-Seq libraries were constructed by Majorbio Bio-Pharm Technology (Shanghai, China) using the Illumina Truseq™ RNA Library Prep Kit following the manufacturer’s instructions. Fig. 1. [57]Fig. 1 [58]Open in a new tab Phenotypic traits of fruit navel in ‘T03’ and ‘BFN’ fruits. A Comparison of fruit navels in ‘T03’ and ‘BFN’ fruits at the early green bud stage, green bud stage, green-yellow bud stage, yellow bud stage, ovary at anthesis, fruit at 3 days after pollination (DAP) and 7 DAP, fruit at veraison, and mature fruit stages. White frames represent the fruit navel tissues used for RNA-sequencing. B Diameter of fruit navel in ‘T03’ and ‘BFN’ fruits. Error bars represent ± SD. Asterisks indicate significant differences in the diameter of fruit navel in ‘BFN’ fruits relative to ‘T03’ fruits (Student t-test: *, P < 0.05; **, P < 0.01) Bioinformatics analysis of RNA-Seq data The raw paired end-reads were filtered and subjected to quality controlled using fastpusing default parameters [[59]20]. The clean reads were then aligned to the Melon Genome ([60]http://cucurbitgenomics.org/organism/18) in a directed mode using the HISAT2 software [[61]21]. The mapped reads were assembled and spliced by StringTie ([62]http://ccb.jhu.edu/software/stringtie/), and compared with known transcripts to functionally annotate the potential new transcripts [[63]22]. Subsequently, RSEM was employed to quantify the read counts for each gene [[64]23], and the Fragments Per Kilobase of transcript sequence per Millions reads (FPKM) were calculated. Differentially expressed genes (DEGs) between the ‘T03’ and ‘BFN’ fruit navels were identified using the DESeq2 R package [[65]24]. The P-values were adjusted to control the false discovery rate (FDR) using the Benjamini and Hochberg method. DEGs with a fold change > 2 and an FDR < 0.05 were considered to be differently expressed. KEGG enrichment analysis of DEGs The 54 up-regulated genes, 62 down-regulated genes in ‘BFN’ compared to ‘T03’ fruits, and all 116 DEGs were subjected to Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis. The KEGG pathway enrichment was performed by the ClusterProfiler R package [[66]25], with the FDR controlled by the Benjamini and Hochberg method. A KEGG term was considered significantly enriched if its P-value was less than 0.05. Heatmap of DEGs The heatmap of DEGs was generated by OriginPro 8.0 (OriginLab Corporation, Northampton, MA, USA) based on the read counts. For each gene, the highest and lowest transcript levels were normalized to 1.5 and − 1.5, respectively. The transcript levels in other samples were calculated as the relative values and visualized using a color gradient. RT-qPCR verification Total RNA extracted from the fruit navels of ‘T03’ and ‘BFN’ melon at the EGB, GB, GYB, YB and OA stages was used to synthesize the cDNA using the FastQuant cDNA Kit (Tiangen Biotech, China). RT-qPCR was performed on a CFX96 Real-time PCR System (Bio-Rad, USA) with SYBR Green PCR master Mix (Applied Biosystems, USA). The melon CmACTIN gene (MELO3C008032) was used as an internal control. The relative expression level of each gene was calculated by the 2^−ΔΔCt method [[67]26]. Gene-specific primers used for RT-qPCR were listed in Table S1. Statistical analysis Phenotypic data of the fruit navel in ‘T03’ and ‘BFN’ melons were measured from three biological replicates. The Student’s t-test (*, P < 0.05; **, P < 0.01) was used to compare the differences in fruit navel diameter between ‘BFN’ and ‘T03’ fruits (Fig. [68]1B). For RT-qPCR assays, three biological replicates and three technical replicates were performed. The Student’s t-test (*, P < 0.05; **, P < 0.01) was applied to assess significant differences in transcript levels relative to the GB-T03 sample (Fig. [69]3), while Duncan’s test (lowercase letters, P < 0.05) was used to compare the expression differences of five putative enzyme-coding genes in different fruit developmental stages of melon (Fig. [70]7). Fig. 3. [71]Fig. 3 [72]Open in a new tab RT-qPCR validation of RNA-seq data. Three upregulated and five downregulated DEGs identified by RNA-seq were randomly selected for RT-qPCR analyses. The expression trends of the selected DEGs detected by RT-qPCR were consistent with those obtained from RNA-seq. Error bars represent ± SD. Asterisks indicate significant differences in gene expression levels compared with the GB-T03 sample (Student t-test: *, P < 0.05; **, P < 0.01) Fig. 7. [73]Fig. 7 [74]Open in a new tab Expression analyses of five putative enzyme-coding genes identified in the “Sesquiterpenoid and triterpenoid biosynthesis”, “Galactose metabolism” and “Circadian rhythm—plant” pathways. RT-qPCR assays were conducted at the early green bud (EGB), green bud (GB), green yellow bud (GYB), yellow bud (YB) and ovary at anthesis (OA) stages. Error bars represent ± SD. Letters above the columns indicate significant differences in transcript levels at P < 0.05 (Duncan’s test) Results Phenotypic characteristics of fruit navel in ‘T03’ and ‘BFN’ melon Based on corolla color and fruit developmental characteristics, the morphology of the fruit navel was observed at the early green bud (EGB), green bud (GB), green yellow bud (GYB), yellow bud (YB), ovary at anthesis (OA), fruit at 3 days after pollination (DAP), fruit at 7 DAP, fruit at veraison and mature fruit stages (Fig. [75]1). The melon inbred line ‘BFN’ is a spontaneous mutant derived from the inbred line ‘T03’, and exhibits a significantly larger fruit navel compared with ‘T03’. During the early green bud stage, no obvious morphological difference was observed in the terminal region of the ovary between ‘T03’ and ‘BFN’ melons. However, the diameter of the fruit navel in ‘BFN’ was notably larger than that in ‘T03’ (Fig. [76]1A, B). Starting from the green bud stage, the ovary end of ‘BFN’ began to bulge, whereas the ovary end of ‘T03’ remained slightly sunken. As a result, a prominent navel formed in ‘BFN’ fruits, contrasting with the small and smooth navel of ‘T03’ fruits (Fig. [77]1A). At the mature fruit stage, the navel diameter of ‘T03’ fruits was 1.57 cm, while that of ‘BFN’ fruits reached 4.10 cm (Fig. [78]1B). Transcriptomic characteristics of the fruit navel from ‘T03’ and ‘BFN’ melons To investigate the mechanism underlying big fruit navel formation, RNA-seq was conducted on the terminal region of ovaries, specifically the fruit navel region, in ‘T03’ and ‘BFN’ melons. The ovary at anthesis (OA) was selected for sampling due to its ease of identification and ability to ensure sampling accuracy. Additionally, transcriptome profiling was performed on the fruit navel of ‘BFN’ and ‘T03’ at the green bud (GB) stage, in which the fruit navel exhibits distinct developmental characteristics. Three biological replicates were performed for each set (GB-T03, GB-BFN, OA-T03, OA-BFN), so that a total of 12 libraries were produced (Table. S2). In general, 41.56 to 60.85 million raw reads were generated per sample, and 41.13 to 60.29 million clean reads were mapped to the melon genome after initial data filtering (Table. S2). The square of Pearson's correlation coefficient (R^2) was closely associated with the developmental stage, and higher R^2 was observed among the three biological replicates within each set compared to those between different groups (Table. S3). Principal component analysis (PCA) revealed clear clustering of samples within the same set, with four distinct clusters corresponding to GB-T03, GB-BFN, OA-T03 and OA-BFN (Fig. [79]2A). These results indicated that distinct gene expression profiles existed in the fruit navel region of ‘BFN’ and ‘T03’ ovaries at the green bud and anthesis stages. Fig. 2. Fig. 2 [80]Open in a new tab Transcriptomic characteristics of fruit navel in ‘T03’ and ‘BFN’ melon. A Principal component analysis (PCA) of GB-T03, GB-BFN, OA-T03 and OA-BFN samples. B, C Venn diagrams of significantly upregulated (B) and downregulated (C) genes in ‘BFN’ fruits compared with ‘T03’ fruits at GB and OA stages Distinct gene expression profiles contributing the formation of fruit navel Differential expression genes (DEGs) were identified using a FDR threshold of less than 0.05 and a fold change greater than 2 (Table. S4). During the green bud stage, 186 genes and 394 genes were significantly up-regulated and down-regulated in GB-BFN compared with GB-T03, respectively (Fig. [81]2B, C). At the anthesis stage, 1404 genes were significantly up-regulated and 478 genes were significantly down-regulated in the fruit navel of ‘BFN’ compared with that of ‘T03’ (Fig. [82]2B, C). Subsequently, independent samples were collected using the same way as for the RNA-seq analysis, and RT-qPCR was conducted to validate the DEGs results. Eight DEGs were randomly selected, and the expression trends of selected DEGs detected by RT-qPCR were consistent with that of RNA-seq data (Fig. [83]3). Compared with the ovary of ‘T03’, three genes were significantly up-regulated and five genes were significantly down-regulated in the ‘BFN’ melon. The Pearson’s correlation coefficient between the RNA-seq and RT-qPCR data was 0.91. These results indicated that the RNA-seq results were highly reliable. To investigate the mechanism underlying fruit navel formation, DEGs exhibiting consistent expression trends in both the OA and GB stages of ‘T03’ and ‘BFN’ melons were selected for further study. A total of 116 DEGs were identified, comprising 54 up-regulated and 62 down-regulated genes in the ‘BFN’ melon compared with the ‘T03’ melon during both GB and OA phases (Fig. [84]2B, C and Fig. [85]4). Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis revealed that the 54 up-regulated genes in ‘BFN’ were significantly enriched in “Biosynthesis of various alkaloids”, “Biosynthesis of unsaturated fatty acids”, “Isoquinoline alkaloid biosynthesis”, “Glycosaminoglycan degradation” pathways. The 62 down-regulated genes were significantly enriched in “Sesquiterpenoid and triterpenoid biosynthesis”, “Circadian rhythm—plant” and “Galactose metabolism” pathways (Fig. [86]5A, B, Table S5). KEGG analysis of the 116 DEGs identified four significantly enriched pathways including “Sesquiterpenoid and triterpenoid biosynthesis”, “Circadian rhythm—plant”, “Galactose metabolism” and “Biosynthesis of various alkaloids” (Fig. [87]5C, Table S5). These pathways were hypothesized to play a critical role in the fruit navel development of melon. Fig. 4. [88]Fig. 4 [89]Open in a new tab Expression pattern of 116 DEGs between ‘BFN’ and ‘T03’ fruit navels. A total of 54 and 62 genes were significantly upregulated and downregulated, respectively, in both the GB and OA phases of ‘BFN’ fruits compared with ‘T03’ fruits. The color blocks represent the relative expression levels of genes, with high and low expression shown in red and blue color, respectively Fig. 5. [90]Fig. 5 [91]Open in a new tab Significantly enriched (P < 0.05) KEGG terms in upregulated (A), downregulated (B), and all 116 DEGs (C) in the fruit navel of ‘BFN’ fruits compared with ‘T03’ fruits Key regulators are involved in the fruit navel formation Transcription factors play an important role in regulating gene transcription and expression by directly binding to cis-acting elements located in the promoter regions of target genes, thereby influencing various growth and developmental processes in plants [[92]27]. There were three (AP2/ERF, MYB and C2H2 types) and eight (AP2/ERF, MADS-box, homeobox domain and bZIP types) transcription factors presented in up-regulated and down-regulated DEGs, and their potential target genes were predicted from the remaining 51 up-regulated and 54 down-regulated DEGs. In the promoter regions of these genes, a total of 13 up-regulated genes were found to contain the GCC (AGCCGCC) or CRT/DRE (A/GCCGAC) cis-acting elements, which are recognized by AP2/ERF transcription factors (Fig. [93]6A, Table S6) [[94]28, [95]29]. The promoters of 26 and 2 DEGs included the AC (ACCTACC, ACCAACC, ACCTAAC) and elicitor-responsive (AATTGACC) elements, which are recognized by MYB and C2H2 transcription factors, respectively (Fig. [96]6A, Table S6) [[97]30, [98]31]. The same way was conducted for the down-regulated DEGs. Specifically, the promoters of 21, 4, 7 and 46 DEGs contained cis-acting elements recognized by AP2/ERF, MADS-box (CArG, CC(A + T-rich)[6]GG), homeobox domain (HD-ZIP1, CAAT(A/T)ATTG) and bZIP (ACGT, CACGT, CACGTC, TACGTA) transcription factors, respectively (Fig. [99]6A, Table S6) [[100]27, [101]32, [102]33]. Fig. 6. [103]Fig. 6 [104]Open in a new tab Identification of key regulators involved in fruit navel formation. A Analysis of transcription factors in up-regulated and down-regulated DEGs. Orange and green color represent up-regulated and down-regulated transcription factors, respectively. B Promoter analyses of DEGs involved in the “Sesquiterpenoid and triterpenoid biosynthesis”, “Galactose metabolism” and “Circadian rhythm—plant” pathways Based on the results of KEGG analyses, two terpene cyclase/mutase family members (MELO3C001812 and MELO3C004329) were identified as participating in the “Sesquiterpenoid and triterpenoid biosynthesis” pathways (Table S5). The putative promoter regions of them contained cis-acting elements recognized by bZIP transcription factors. Additionally, the promoter of MELO3C004329 included cis-acting elements for AP2/ERF and homeobox domain transcription factors (Fig. [105]6B). Within the “Galactose metabolism” pathway, cis-acting elements for bZIP and MADS-box transcription factors were identified in putative promoters of MELO3C002287 (Raffinose synthase) and MELO3C025521 (Hexosyltransferase), respectively (Fig. [106]6B, Table S5). Furthermore, a bZIP transcription factor (MELO3C003686) and a chalcone synthase (MELO3C014767) were implicated in the “Circadian rhythm—plant” pathway (Table S5). The promoter of MELO3C014767 contained cis-acting elements for bZIP, AP2/ERF and homeobox domain transcription factors (Fig. [107]6B). However, no cis-acting elements for AP2/ERF, MYB or C2H2 transcription factors were detected in the promoter region of MELO3C022377, a cucumber peeling cupredoxin-like gene involved in the “Biosynthesis of various alkaloids” pathway. Subsequently, the expression patterns of five putative enzyme-coding genes were analyzed at different fruit developmental stages prior to fertilization. The transcripts of two terpene cyclase/mutase genes (MELO3C001812 and MELO3C004329) were significantly downregulated in all detected stages (EGB, GB, GYB, YB and OA) of ‘BFN’ fruits compared with ‘T03’ fruits (Fig. [108]7A, B). In contrast, MELO3C002287 exhibited significant upregulation at the YB stage in the fruit navel of ‘BFN’ compared with ‘T03’ (Fig. [109]7C). For MELO3C025521, no significant difference in expression levels was observed between ‘BFN’ and ‘T03’ fruit navels at the EGB, GB and YB stages (Fig. [110]7D). Similarly, MELO3C014767 showed a significant increase in expression at the GYB stage in the fruit navel of ‘BFN’ compared with ‘T03’, while no significant difference was detected at the EGB and YB stages (Fig. [111]7E). These results suggested that MELO3C001812 and MELO3C004329, involved in the “Sesquiterpenoid and triterpenoid biosynthesis” pathway, play a critical role in fruit navel development. Discussion Fruits shape and size exhibit significant variation among different melon varieties. Wild melons typically produce small and round fruits, whereas cultivated varieties display a remarkable diversity in fruits morphology due to continuous domestication [[112]3]. The fruit navel, an important component of melon appearance quality, significantly influences consumer preferences. However, the formation process and