An oriental melon variety (longqing) was chosen as experimental material based on its susceptibility to biotic and abiotic stresses and weaker plant growth during its lifespan. The varietal seeds were offered by the Heilongjiang Academy of Agricultural Sciences (HAAS), Daqing Branch, China, and entire experiment was carried out in a controlled condition in the aseptic laboratory at HAAS. An exogenous melatonin treatment was checked on seed germination traits and morphological and physiological parameters of seedlings grown under heavy metal-induced stress. To sort out the treatments of melatonin and PD stress, a pre-experiment was conducted at small scale in lab and the actual experiment was conducted.

For evaluating the seed germination status and related parameters (sprout length, average diameter, and surface area), the seeds were gently sterilized in a solution of potassium permanganate (KMnO₄, 10%) for one minute, then washed with double distilled water (ddH₂O), and allowed to dry. Later, a total of 50 melon seeds were spread on a 9-cm-wide autoclaved petri dish having 4 layers of wet qualitative slow filter paper, and moisture was regularly retained with a total of 12 mL of test solution. Finally, a total of four treatment groups along with 3 replications of each treatment were used [e.g., (i) control with no stress (CK, just ddH₂O), (ii) 100 μM of melatonin (M) application, (iii) 40 mg L⁻¹ of potassium dichromate (PD) application as a stress, and (iv) melatonin treatment (M) + potassium dichromate stress (PD)]. The respective concentrations of potassium dichromate and melatonin treatment were referenced from previous research [29,30,31]. The treated melon seeds in petri dishes were incubated for 60 h at 28°C, and seed germination status as well as linked indices [sprout length (cm), average diameter (mm), and surface area (cm²)] were tested. The germination rate of 50 melon seeds of comparative treatments was calculated to determine the efficiency of melatonin using the equation as follows:

To cultivate proper melon seedlings and test the effect of exogenous melatonin treatment to mitigate the PD stress, the treated seeds were allowed to grow in sterilized plastic pots (diameter 6 cm and depth 7.5 cm) in controlled conditions using completely randomized design (CRD) geometry with three repetitions of each treatment. Briefly, seeds were put in muslin bags and allowed to germinate for two days at 28°C before being transferred to plastic pots filled with 3% organic matter soil and potting compost with a pH of 7–7.5. The characteristics of the compost are as follows: organic carbon (OC) = 20% (weight percentage on a dry product basis), organic nitrogen (ON) = 1%, organic matter (OM) = 35%, and pH = 6–7.

The emerging seedlings at the 4-true-leaf stage were divided into four groups. (i) spraying ddH₂O as a control (without any stress), (ii) spraying 100 μM of melatonin (M), (iii) irrigating 40 mg L⁻¹ of potassium dichromate solution as PD stress, and (iv) melatonin treatment (M) + potassium dichromate (PD) stress, each treatment was carried out with equal volume (40 mL) daily. The leaf samples from each treatment group were collected on the 1st day, 3rd day, and 5th day to determine the variation in physio-chemical indicators, as well as for transcriptional regulation through qPCR analysis; however, the morphological and ultrastructural observations were recorded on the 5th day.

The leaf tissues of raised seedlings sampled from each treatment were checked for cytological observations using a transmission electron microscopy analysis at different magnification levels (1.5k×, 8k×, and 20k×), as reported in a previous study of Liu et al. [32]. Briefly, the true leaf tissues from normal (control, without stress) and stress-induced obvious symptoms were sampled from melon seedlings cultivated under four different treatments (above-mentioned). The sampled leaves were sliced into 2 mm-wide pieces and fixed with a 2.5% (v/v) solution containing 1% osmium tetroxide. Then, all samples were embedded in epoxy resin embedding medium (EMBed-812), sliced in thin sections, dried using acetone solutions, viewed with a high-resolution transmission electron microscope, and photographed.

A total of 1–2 g (g) of the fresh leaves from the top of the grown seedlings were randomly sampled from three biological repetitions of different treatment groups, and the physiological and biochemical indices [e.g., peroxidase (POD), superoxide dismutase (SOD), and ascorbate peroxidase (APX)] were estimated based on the following instructions of commercially available enzyme-linked immunosorbent assay (ELISA) kits, Michy Biology Company (Suzhou, China). The level of PD-induced changes in membrane lipid peroxidation was also deliberated by quantifying the malondialdehyde (MDA) and hydrogen peroxide (H₂O₂) contents using ELISA test kits according to the briefed protocol. The lipid composition in the cell membrane of seedling leaves was checked through the electrospray ionization (ESI) mass spectrometry method with a slight modification to the earlier reported protocol as Narayanan et al. [33].

The lignin content was determined using the same method as Shao et al. [34]. In short, the collected leaf tissues from various treatments and replications were standardized in a mortar with 10 mL of 95% ethanol (v/v) at 4°C, then centrifuged at a relative force of 5000× g for 4 min. The formed supernatant were separated and washed double with ethanol (80% v/v) and one time with acetone before being thoroughly dried in an oven at 60°C. The dry samples were incubated at 80°C for 2 h in a 4:1 (v/v) acetyl bromide (CH₃COBr) and acetic acid (CH₃COOH) before cooling to 25°C and being transferred to 50 mL glass bottles containing 5 mL of acetic acid and 2 M sodium hydroxide (NaOH). The reaction was ended through adding the 7.5 M hydroxylamine hydrochloride (NH₂OH·HCl), and the final volume was adjusted with acetic acid. In the end, the lignin was measured using the absorbance rate at 480 nm. Metallothionein (Met) was measured through ELISA test kits, following the same technique of Zhang et al. [35].

Chlorophyll of the leaf samples taken for each treatment and replication were also measured using the ethanol-acetone extraction technique as Dong et al. [36]. Briefly, a sharp cutting razor was used to cut a total of 0.30 g of fresh weight (FW) leaf samples from the main vein of leaves joined at the same node. All samples were submerged in a 10 mL solution of ethanol and acetone (1:1) for 24 h. The NanoDrop™ One/OneC Microvolume UV Spectrophotometer (Thermo Fisher Scientific, Delaware, USA) was used to measure the extract’s absorbance rate at 440, 663, and 645 nm. Total chlorophyll was deliberated based on the formula:

Total chlorophyll (mg·g⁻¹ FW) = (8.04 × A663 + 20.29 × A645) × V/W × 1000; W = weight, V = volume, and A = absorbance rate, individually.

The chlorophyll fluorescence emission from the aerial surfaces of the intact plant leaves were monitored as described in a previous study [36]. The samples were collected from different treatment groups in slightly cloudy conditions, and the leaves were kept in fully dark conditions for 20 min prior to measuring fluorescence indices [e.g., light-adapted minimum fluorescence (F0), photochemical quenching coefficient (qP), maximum photochemical efficiency of PSII (Fv/Fm), and non-photochemical quenching coefficient (NPQ)].

The samples of melon root, stalk, and leaf from different treatments were collected and completely dried. A 0.50 g (g) of every sample was homogenized with 7.50 mL of nitric acid (HNO₃) and 2.50 mL of Perchloric Acid (HClO₄) and kept overnight. Afterward, 5 mL of HNO₃ was added to the mixture for full digestion and endogenous Cr was quantified by the same method of Zaheer et al. [37].

Translocation ratios of Cr in different organs were estimated based on the formulas of previous Luo et al. [38]: e.g., T roots to stalk (%) = C_stalk/C_roots × 100 (1); T stalk to leaves (%) = C_leaves/C_stalk × 100 (2); T roots to leaves (%) = C_leaves/C_roots × 100 (3), where C_leaves is concentration of Cr (μg/g) in leaves; C_roots is in roots (μg/g); C_stalk is in stalk (μg/g).

To investigate the genetic regulation mechanism of photosynthetic and physiological indices in melon, the annotated putative genes were obtained from previously reported studies, e.g., 4CL [34], HEMA [39], GOGAT [40], POR [41], PAO [42], SPS [43], PEPC [44], ruBisCo [45], PSB [46], Hcf136 [47], APX [48], POD [49], SOD [50], Met [51], and the housekeeping gene “Actin” [32], as shown in the Supplementary File. The functional annotation of selected genes was predicted in the referenced genome of melon [(v3.6.1, DHL92); Cucurbit Genomics Database, http://cucurbitgenomics.org].

The transcription expression of putative genes contributing in key metabolic pathways in melon plants were analyzed in response to melatonin treatment under PD-induced stress using qPCR verification. For the molecular experiment, total RNA was firstly extracted by RNAprep Pure Plant Kit (Tiangen Biotech, Beijing, China) according to the instructions, the integrity was checked through NanoDrop 1000 (Thermo Fisher Scientific, Delaware, USA), and the concentration index of A260/A230 ratio >2.0 was used for complementary DNA (cDNA) synthesis. The first strand cDNA was synthesized by the PrimeScript™ RT reagent Kit with gDNA Eraser (Takara Bio, Kusatsu, Japan) using 2 μL of total RNA. The cDNA products were diluted 5-fold with ddH₂O before experimental use. For the qPCR analysis, a total of 10 µL of reaction mixture containing 1 µL of cDNA, 0.2 µL of primer, 5 µL of 2 × SYBR Green PCR Master Mixture, and 3.8 µL of ddH₂O was used to examine expression level in the Real-Time PCR Detection System (Bio-Rad Laboratories, Inc., California, USA) [32]. The detailed information about the molecular sequences (forward and reverse) of exported primers of predicted annotated genes in the melon genome and the referenced actin gene is available in the Supplementary File (Table S1).

All the experimental data was regularly collected, and the numerical values were computed using the Microsoft Excel Worksheet (v2019) [32]. The statistical differences among treatments within each measurement of time series data including physiological parameters were evaluated by two-factor repeated measures ANOVA (Analysis of Variance) and Tukey’s HSD (p < 0.05) using GraphPad Prism (v10.2.0), while one-way ANOVA with Tukey’s HSD (p < 0.05) was employed for the morphological parameters. The endogenous Cr quantification and translocation rates were compared using an independent samples t-test. The data were assessed for normality and homogeneity of variances prior to analysis.

We noticed that seed germination rate followed an S-shaped pattern under stress treatment, with an early slow phase followed by a rapid increase and a downward tendency at the end (Table 1). However, the observed S-shaped pattern varied between treatments. At the initial stage, the melatonin-treated seeds showed a quicker and higher germination rate (about 13.94%) during the first 36 h compared to the control (CK) group. However, as the time progressed, the germination rate differences decreased between the CK and melatonin-treated groups. It is noticeable that the saturation in germination under melatonin treatment is 12 h earlier than the control (CK), indicating that exogenous melatonin has the beneficial function of promoting the germination of seeds.

The germination rate in PD-stress treated groups exhibited a sluggish trend, with the highest rate reaching only 38.73% at 60 h, which represents a 59.66% decrease relative to the CK group. However, exogenous melatonin application significantly alleviated the detrimental impact of potassium dichromate (PD)-stress on germination rate, which increased by 84.17%. PD stress significantly impaired the development condition of melon seed sprouts, as demonstrated by lower sprout length (27.59%), average diameter (19.54%), and surface area (37.50%) compared to the CK group (Fig. 1). However, exogenous melatonin treatment effectively reversed the negative effects of PD stress on these phenotypic parameters and even enhanced more when compared to CK treatment.

The inhibitory effect of PD stress and the enhancing effect of melatonin on the morphological characteristics of melon were evaluated. The treated seedlings exhibited typical stress symptoms, but melatonin showed the improved leaf morphology, as illustrated in the figure (Fig. 2A). Statistical analysis consistently showed that seedling growth and development were most severely affected under PD stress; however, extensive dehydration, chlorosis, and necrosis were observed on the leaf surfaces, indicating that PD stress induced significant osmotic stress and chlorophyll degradation in melon leaves compared to the melatonin and control (CK) treatment.

Transmission electron microscopy (TEM) of the internal leaf organelles in melon seedlings subjected to various treatments revealed significant structural variations, as shown in the figure (Fig. 2B). PD-stressed samples exhibited swollen, spherical chloroplasts with indistinct and irregular thylakoid structures, supplemented by a reduced number of starch granules. In contrast, exogenous melatonin treatment effectively preserved the developmental integrity of melon leaves by stabilizing the chloroplast ultrastructure, maintaining normal thylakoid organization, and supporting regulated metabolic processes under stress.

Regarding the physiological indices, PD-stressed samples exhibited substantially higher levels of MDA and H₂O₂ compared to the control (CK) on the 5th day, as shown in the figure (Fig. 3A,B). However, exogenous melatonin treatment effectively reduced MDA and H₂O₂ levels by 25.10% and 32.29% on the 3rd day, and by 35.72% and 21.98% on the 5th day, compared to PD-stressed samples. Additionally, SOD, POD, and APX activities increased by 88.94%, 105.26%, and 410.00% on the first day of PD stress compared to CK at the same time point, as illustrated in the figure (Fig. 3C–E).

As stress duration increased, the activity of these antioxidants significantly declined during anaphase, decreasing by 23.17%, 7.02%, and 9.43% on the 5th day compared to the 3rd day. In contrast, melatonin-treated samples exhibited continuous improvement, with antioxidant levels rising by 46.13%, 37.74%, and 25.00% relative to PD-stressed samples on the 5th day. Analysis of membrane lipid composition showed that the molar percentages of DGDG and MGDG accounted for over 60% of total leaf glycolipids. PD stress reduced MGDG and DGDG concentrations by 8.09% and 11.03%, respectively, while increasing the MGDG to DGDG ratio by 3.15%, indicating that PD stress adversely affected plastid membrane structure and composition, as illustrated in the figure (Fig. 3F).

In terms of photosynthetic-linked activity, total chlorophyll (Chl) contents were noticed to be decreased significantly by 29.82% on the 3rd day and 33.77% on the 5th day as shown in the figure (Fig. 3G), revealing the significant upsurge in superoxidative photosynthesis apparatus, implying that PD stress negatively affected the photosynthesis apparatus, so that caused the significant degradation of chlorophyll. On the other hand, exogenous melatonin application efficiently retained the Chl contents and homeostasis of MGDG and DGDG levels, which increased by 1.78% and 11.45% compared to PD-stressed samples on the 5th day, and lowered the MGDG to DGDG ratio by 9.17% at the same time period. It was also discovered that the protective mechanism of exogenous melatonin to chlorophyll synthesis is not limited to stabilizing the relative content under PD-stressed conditions, which increased by 47.71% on the third day and 67.11% on the 5th day; the chlorophyll content in melatonin-treated samples showed an obvious increasing trend, with improvements of 26.83% on the 3rd day and 27.02% on the 5th day, respectively.

Chlorophyll-linked photosynthesis fluorescence parameters [F0 (minimal fluorescence yield of chlorophyll), Fv (variable fluorescence), and Fm (maximum fluorescence)] were checked to evaluate the PD stress-induced photoinhibition and the photoprotective function of melatonin. Chlf results showed that PD-stress induced a 17.38% reduction in F0 and 18.29% in Fv/Fm as compared to CK on the 5th day as shown in the figure (Fig. 4A,B), denoting the inhibited photosynthetic electron transport rate. Likewise, PD-stress caused a significant reduction in qP (18.01%) and NPQ (26.79%) (Fig. 4C,D), indicating the comprehensively restrained photochemical electron transfer under PD-induced stress treatment. However, the beneficial effect of melatonin on photosynthesis was further confirmed by the enhanced Chlf parameters, which increased 13.73% in F0, 18.65% in Fv/Fm, 12.81% in qP, and 30.26% in NPQ on the 5th day compared to that of the PD-stressed samples, respectively.

The significant variations in metallothionein (Met) and lignin levels were observed across different tissues (leaf, stalk, and root) subjected to various treatments, with these differences becoming more evident as the duration of PD-stress treatment increased. In leaves, Met and lignin content surged by 26.98% and 4.64% on the 1st day and by 24.26% and 19.23% on the 5th day compared to CK, as shown in the figure (Fig. 5A,B). On the 1st day, Met and lignin contents increased by 23.33% and 2.52% under PD-stress, respectively, whereas there was a 37.84% and 14.03% increase on the 5th day, as shown in the figure (Fig. 5C,D). In roots, Met and lignin accumulation increased by 25.47% and 2.43% on the 1st day and by 29.90% and 23.80% on the 5th day, compared to CK treatment, as shown in the figure (Fig. 5E,F). Interestingly, Met content in lead, stalk, and root also increased by 8.57%, 1.69%, and 1.56% on the 5th day, compared to tissues under PD-stress; however, there was no obvious change at the initial stage. Additionally, the lignin content in the same tissues followed a similar increasing trend by 10.79%, 6.68%, and 2.54%, respectively.

We observed that exogenous melatonin treatment decreased endogenous Cr concentration (μg/g DW) in leaf, stalk, and root by 71.53%, 11.45%, and 2.43%, under PD-induced stress. Furthermore, the observed decreasing trend of Cr translocation rate (%) between various organs was consistent with the endogenous quantified Cr. However, exogenous melatonin treatment reduced Cr translocation rates by 9.28% and 67.71% from root to stalk and stalk to leaf, with the greatest inhibition rate seen in root to leaf, which dropped by 70.83% (Table 2).

To investigate the genetic effectiveness of melatonin in alleviating the inhibitory effect of Cr stress, the transcriptional regulation trends of the selected putative melon genes associated closely with chlorophyll and carbohydrate synthesis and a series of key metabolism pathways were examined by qPCR analysis as shown in the figure (Fig. 6). We noticed that the expression trend of the GOGAT gene (MELO3C008481.2) encoding ferredoxin-dependent glutamate synthase 1 and the HEMA gene (MELO3C006296.2) encoding glutamyl-tRNA reductase (GluTR), which catalyzes the biosynthesis of glutamate and tetrapyrrole in plastid, triggering the chlorophyll biosynthesis, reduced by 16.00% and 7.00% under PD stress as compared with CK, indicating that the limited accessibility of the substrate for chlorophyll synthesis. Meanwhile, the upregulated PAO gene (MELO3C023571.2) induced by PD stress, which increased by 202.00%, was consistent with the ever-increasing chlorophyll degradation to its final product.

Similarly, the severely downregulated expression of the SPS gene (MELO3C003715.2) (15.00%), the PEPC gene (MELO3C018724.2) (4.00%), and the Rubisco gene (MELO3C012180.2) (10.00%), which catalyzed the sucrose synthesis and carbon fixation, further confirmed the reduced starch content. However, exogenous melatonin treatment showed the genetic improvement by increasing the expression level of the antioxidant-related genes, e.g., the POD gene (MELO3C002242.2, by 3.89%) and the SOD gene (MELO3C017624.2, by 5.50%) as compared with PD-stressed samples, and the Rubisco (MELO3C012180.2), HEMA, and SPS genes were upregulated by 163.33%, 47.31%, and 257.65%, respectively. Likewise, the expression level of the Met synthesis-related gene in melon (CmMet, MELO3C016513.2) and the lignin synthesis-related gene (Cm4CL, MELO3C002346.2) were also seemed to be enhanced by 3.29% and 12.59% under melatonin treatment. These results significantly showed the genetic improvement of photosynthetic, physiological, and biochemical traits of melon in response to different combination of treatments.