Abstract Hypericum perforatum, also known as “natural fluoxetine,” is a commonly used herbal remedy for treating depression. It is unclear whether melatonin in plants regulated by the endogenous circadian clock system is like in vertebrates. In this work, we found that the melatonin signal and melatonin biosynthesis gene, serotonin N-acetyltransferase HpSNAT1, oscillates in a 24-hour cycle in H. perforatum. First, we constructed a yeast complementary DNA library of H. perforatum and found a clock protein HpLHY that can directly bind to the HpSNAT1 promoter. Second, it was confirmed that HpLHY inhibits the expression of HpSNAT1 by targeting the Evening Element. Last, it indicated that HpLHY-overexpressing plants had reduced levels of melatonin in 12-hour light/12-hour dark cycle photoperiod, while loss-of-function mutants exhibited high levels, but this rhythm seems to disappear as well. The results revealed the regulatory role of LHY in melatonin biosynthesis, which may make an important contribution to the field of melatonin synthesis regulation. __________________________________________________________________ .The core biological clock factor LHY regulates the biosynthesis of melatonin by targeting SNAT1 in Hypericum perforatum. INTRODUCTION Hypericum perforatum is a perennial herbaceous plant commonly known as St. John’s wort. In traditional Chinese medicine, the herb H. perforatum (Guan Ye Lian Qiao) belongs to the group of plants that has the effects of clearing heat and detoxifying, astringent and hemostatic, and diuretic ([34]1–[35]22), exhibiting good pharmacological activities such as anti-inflammatory, antiviral, antitumor, antidepressant, antibacterial, liver protection, and nerve cell protection ([36]4–[37]52). H. perforatum is widely used in European and American countries to treat depression and is known as “natural fluoxetine” ([38]7, [39]8). In October 2008, “Shugan Jieyu Capsule” was approved as a traditional Chinese medicine for the treatment of mild to moderate depression in China. Its main raw materials are H. perforatum and Eleutherococcus senticosus. Melatonin has been found in many plants and in different parts of plants, and it has been reported to be in extremely high levels in several medical herbs including feverfew and St. John’ s wort and can reach a dry weight of 1 to 4 μg/g ([40]9, [41]10), which has the functions of regulating animal circadian rhythm, seasonal reproductive development, enhancing immunity, and improving sleep ([42]11, [43]12). Melatonin was initially detected in land plants in 1995 ([44]13, [45]14) and participates in the entire growth and development stage of plants ([46]15, [47]16) and also acts as an antioxidant to enhance the ability to cope with abiotic stresses ([48]17, [49]18). However, research on the regulation of plant melatonin biosynthesis is still in its early stage; in particular, the reference literature on the regulatory mechanisms of upstream transcription factors is almost blank. Melatonin is sensitive to light, and its synthesis in humans and animals exhibits a circadian rhythm. Kolár et al. ([50]19) investigated the changes in melatonin levels in the short-day plant Chenopodium rubrum and found that melatonin is almost undetectable in C. rubrum during light exposure, while its content increases in the dark and reaches its peak at 4 to 6 hours, which is very similar to the circadian synthesis rhythm of melatonin animals. However, many studies have shown the opposite results. Li et al. ([51]20) found that the melatonin signal exhibited rhythmic changes, but unlike patterns in animals, the melatonin signal in Arabidopsis reached its peak in the morning. Byeon et al. ([52]21) exposed senescent detached rice leaves to continuous light and found that the content of melatonin and its synthesis intermediates, tryptamine, 5-hydroxytryptamine, and N-acetyl-5-hydroxytryptamine, were significantly higher than those in the dark. The central oscillator of the biological clock is composed of three core factors, among which CIRCADIAN CLOCK-ASSOCIATED 1 (CCA1) and LATE ELONGATED HYPOCOTYL (LHY) are morning genes, encoding two homologous MYB transcription factors ([53]22, [54]23). CCA1 and LHY can inhibit the expression of another core gene TIMING OF CAB EXPRESSION 1 (TOC1), forming the core feedback inhibition loop ([55]24). This inhibitory loop regulates the rhythmic movement and physiological activities of different plant tissues by regulating other genes ([56]25, [57]26). Adams et al. ([58]27, [59]28) revealed the complex regulatory relationship between the biological clock and abscisic acid (ABA) signaling pathway and found that LHY can directly inhibit the expression level of ABA biosynthesis enzyme gene NCED3, indicating that LHY regulates the accumulation of ABA in wild-type (WT) Arabidopsis with a circadian rhythm. Wang et al. ([60]29) found that the loss of four GmLHYs genes enhances stomatal response to drought, reduces leaf water loss rate, and enhances plant drought resistance, indicating that they play a negative regulatory role in soybean drought resistance. Our research group has completed the analysis of the biosynthetic pathway of melatonin in H. perforatum ([61]30, [62]31). In addition, we found that melatonin signaling and serotonin N-acetyltransferase HpSNAT1 gene expression in H. perforatum exhibited a diurnal rhythm during the 12-hour light/12-hour dark cycle (12 L/12D) photoperiod. Therefore, we speculate that this regular change in melatonin signaling may be related to the some biological clock proteins. Herein, a cDNA yeast library of H. perforatum was established, which was screened using the HpSNAT1 promoter proHpSNAT1 as a bait. A total of 96 positive yeast clones were screened and amplified, resulting in 39 different sequences. Rotation validation was performed on the LHY sequence annotated as a biological clock protein, and it was found that this protein can directly bind to the Evening Element (EE) motif (AAATATCT) on the proHpSNAT1 to inhibit the expression of HpSNAT1. The HpLHY-overexpressing plants had reduced levels of melatonin in 12 L/12D photoperiod, whereas loss-of-function mutants exhibited an altered rhythm of melatonin accumulation. RESULTS The melatonin signaling shows day/night rhythms in H. perforatum One-month-old H. perforatum seedlings were grown under 2 cycles of 12 L/12D and 2 cycles of continuous light conditions. The endogenous melatonin showed notable rhythmicity during the 12 L/12D photoperiod, with the main peak appearing in the morning at zeitgeber time 4 hours (ZT4; [63]Fig. 1A). As shown in [64]Fig. 1B, it was found that the expression of HpSNAT1 gene in H. perforatum also exhibits a clear rhythm in the morning, reaching its peak in ZT4, similar to changes in endogenous melatonin concentration. The endogenous concentrations of melatonin and expression of HpSNAT1 related to biosynthesis of melatonin persist under constant light (ZT48-ZT96), indicating that the melatonin signaling might be regulated by the internal biological clock. On the basis of RNA sequencing (RNA-seq) analysis of differentially expressed genes (DEGs) between WT and overexpression of HpSNAT1 (OE-HpSNAT1) lines, the gene expression patterns were similar between the biological replicates but differed obviously between the WT and OE-HpSNAT1 as expected ([65]Fig. 1C). A large number of transcripts related to circadian rhythm were enriched, indicating a close association between the melatonin signaling pathway and the endogenous circadian clock system by Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis ([66]Fig. 1D). Fig. 1. Melatonin and HpSNAT1 signaling in H. perforatum display day/night rhythms. [67]Fig. 1. [68]Open in a new tab (A and B) Rhythmicity in endogenous concentration of melatonin and gene expression of HpSNAT1 in H. perforatum WT seedlings. (ng/g FW) means the content of melatonin in 1 g of fresh weight H. perforatum plant. Actual and subjective days are denoted by white bars. Actual and subjective nights are denoted by black and gray bars, respectively. ZT48-ZT96 refers to consistent light condition. Data are presented as the means ± SD (n = 5). (C) Heatmap of DEGs in WT and OE-HpSNAT1 (n = 3) for each group in Arabidopsis. Relative transcript level is indicated on a color scale from red (high) to blue (low). (D) KEGG pathway enrichment analysis between WT and OE-HpSNAT1 comparison in Arabidopsis (n = 3). Rich factor refers to the ratio of the number of transcripts in the pathway entry in the differentially expressed transcript to the total number of transcripts in the transcript that are located in the pathway entry. The dot size indicates the number of DEGs of the pathway. Construction of H. perforatum yeast cDNA library As shown in fig. S1A, the bands of 28S and 18S are bright and nondegradable, indicating good extraction quality. In the step of isolating mRNA from total RNA, the diffuse band of mRNA are shown in fig. S1B, with the main concentrated regions ranging from 1000 to 2000 bp. The double stranded cDNA band obtained by long distance polymerase chain reaction (LD-PCR) amplification of mRNA are shown in fig. S1C, which are basically consistent with the length of mRNA bands, with a range of 100 to 3000 bp. After removing the small fragments, as shown in fig. S1D, the bands are concentrated in the region around 750 bp, indicating good cDNA integrity, and can be used for the next step of the experiment. The cDNA library of H. perforatum contains approximately 290 monoclonal clones on a 100-mm plate with a 1:1000 dilution gradient, as shown in fig. S2A. The storage capacity is 5.8 × 10^6, and among the 24 randomly selected monoclonal clones, most of the inserted fragments are around 1000 to 1500 bp (fig. S2B). Genome-wide identification of HpSNAT1 interacting proteins As shown in fig. S3, from the results of HpSNAT1 self-activation detection, it can be seen that there was no colony growth in the transformed cells on the plate with 20 mM 3-amino-1,2,4-triazole (3AT) added, indicating that the HIS3 reporter gene was not activated. Therefore, it was decided to conduct yeast heterozygous screening on the basis of 20 mM 3AT. The initial positive transformants grown on the above plate were marked onto a SD-TLH + 20 mM 3AT selection plate. As shown in fig. S4, 96 monoclonal colonies were grown on the plate, which were amplified for DNA sequencing. Last, 39 different sequences were obtained after BLAST alignment analysis. Next, rotational validation was performed on these 39 yeast positive clones, and all positive clones obtained through screening can grow on SD-TL–, SD-TLH–, and SD-TLH + 20 mM 3AT–deficient plates. Among them, number 9 is a MYB-related transcription factor LHY, annotated as a biological clock protein (table S1). As a core factor in the plant biological clock system, LHY is involved in regulating multiple aspects of plant growth and development. Therefore, it is speculated that the rhythmic signal of melatonin is inevitably related to it. Characterization and expression profile of HpLHY The No. 9 MYB-related transcription factor LHY screened by the yeast system is named HpLHY and shares homology with six proteins with LHY/CCA1 function ([69]Fig. 2A). The coding region of the HpLHY gene is 1308 bp, encoding 435 amino acids, with an isoelectric point of 6.05 and a molecular weight of 48.53 kDa. Through confocal microscopy observation, it was found that the HpLHY gene may be mainly localized in the nucleus and expressed in small amounts in the cytoplasm or cell membrane ([70]Fig. 2B). The HpLHY protein may be influenced by the recruitment of other proteins, causing it to shuttle from the nucleus to the cytoplasm or cell membrane region and exert its function ([71]32). To avoid the cleaving of the green fluorescent protein (GFP) tag of the fusion protein affecting the results, we also performed Western blot on the transfected protoplasts. As shown in fig. S5, the HpLHY-GFP lane only had a band that matched the size of the fusion protein (~80 kDa), and no bands of GFP tag protein size were observed (~30 kDa). The results confirm that the fluorescence is the localization of the target protein HpLHY. Through reverse transcription quantitative PCR (RT-qPCR) analysis of the expression level of HpLHY gene in different tissues, it was found that its expression level is relatively low in flowers and highest in stems ([72]Fig. 2C). In the 12 L/12D photoperiod, the expression trend of HpLHY is completely opposite to HpSNAT1, with the lowest expression level observed in ZT8-ZT12 and a sudden increase from ZT12 ([73]Fig. 2D). According to the transcriptional activation of HpLHY, as shown in [74]Fig. 2E, compared to the positive control, the pGBKT7-HpLHY group cannot appear blue on SD-TAH + X-α-gal; therefore, the HpLHY transcription factor does not have self-activation activity. Fig. 2. Bio-informative analysis, expression profiles, and subcellular localization of HpLHY. [75]Fig. 2. [76]Open in a new tab (A) Phylogenetic analysis and multiple sequence alignment of the HpLHY protein in H. perforatum, Arabidopsis, and Glycine max. The MYB-DNA binding domain is represented with red lines. (B) 35S-HpLHY-GFP were transiently expressed in Arabidopsis mesophyll protoplasts. 35S-GFP was used as a control. From left to right, we observed subcellular localization of the target genes (GFP), chloroplast autofluorescence (Auto), bright-field (Bright), and merged (Merged) images. Scale bars, 5 μm. (C) Expression profiles of HpLHY in different organs (F, flowers; L, leaves; R, roots; S, stems) of H. perforatum. (D) Expression profiles of HpLHY and HpSNAT1 in leaves during 12 L/12D photoperiod of H. perforatum. Data were normalized to HpACT2 and presented as the means ± SD (n = 3). Different letters indicate significant differences (P < 0.05) between each group tested by one-way analysis of variance (ANOVA). (E) Transactivation activity test of HpLHY in the yeast GAL4 system. HpLHY directly binds the promoter of HpSNAT1 to suppress its expression EE (AAATATCT or AGATATTT) as the most highly represented motif has been reported as a binding region of LHY transcription factors by chromatin immunoprecipitation sequencing analyses ([77]27). However, there has been no other experimental validation. There were four MYB-related binding sites in the 1.5-kb region of the proHpSNAT1 ([78]Fig. 3A). To examine whether HpLHY can bind to any of these sites in the proHpSNAT1, YIH assays were first performed. The result indicates that HpLHY only binds to the EE motif (AAATATCT) of P1 fragment ([79]Fig. 3B). Fig. 3. HpLHY inhibits the expression of HpSNAT1 by directly binding to EE motif. [80]Fig. 3. [81]Open in a new tab (A) The P1-P4 segmentation and predicted binding sites of proHpSNAT1 by PlantPan 4.0. The black box represents the binding site, and the red box represents the mutation site. (B) Yeast one-hybrid assays showing the interaction of HpLHY with the proHpSNAT1 P1-P4. (C and D) EMSA analysis of the directive binding of HpLHY to the EE motif in P1 of proHpSNAT1. (E and F) Transient dual-luciferase assays in N. benthamiana leaves showed that HpLHY repressed the expression of HpSNAT1. Data are presented as the means ± SD (n = 3). Different letters indicate significant differences (P < 0.05) between each group tested by one-way ANOVA. To confirm that the site were the real binding site of HpLHY, we performed an electrophoretic mobility shift assay (EMSA) and designed a biotin-labeled probe containing an EE site with a length of approximately 45 bp in P1, as well as corresponding competitive and mutant probes. As shown in [82]Fig. 3C, as the concentration of MBP–HpLHY protein increases, the color of the band of bound probe gradually deepens, and the free probe gradually becomes lighter, indicating an increase in the number of HpLHY protein–bound biotin-labeled probes containing EE motif. As shown in [83]Fig. 3D, when the MBP–HpLHY protein was incubated with cold competition and WT probes containing EE elements, as the concentration of cold probes gradually increased, the color of the bound band gradually became lighter because of competition. When reacting MBP–HpLHY protein with mutant and WT probes, it was found that as competition increased, the color of the bound band did not change notably, indicating that competition could not occur. To further investigate how HpLHY regulated HpSNAT1 expression, we performed a luciferase reporter gene expression assay. We prepared proHpSNAT1-driven LUC plasmids as reporter constructs, CaMV-35S promoter–driven GFP or HpLHY-GFP plasmids as effectors, and pGreenII 0800-LUC and pGreenII62-SK empty vectors as negative control. As shown in [84]Fig. 3 (E and F), the luciferase activity of the cotransformed tobacco with pGREEN II-62SK and proHpSNAT1-driven LUC reporters was higher than that of other parts, indicating that the HpSNAT1 promoter has strong self-activation activity. The luciferase activity of HpLHY-GFP effectors and proHpSNAT1-driven LUC reporter region was slightly reduced, but still higher than the negative control group, indicating that HpLHY has a negative regulatory effect on the expression of HpSNAT1. Overexpression of HpLHY in Arabidopsis By overexpression of HpLHY in the Arabidopsis mutant lhy cca1, the overexpression (OE) lines contained the expected 926-bp fragments of the CaMV35S promoter were confirmed by PCR (fig. S6). Five transgenic lines of OE-HpLHY were obtained at ZT4 for gene expression, Western blot, and melatonin detection. Through semiquantitative results, it can be seen that HpLHY has been detected in all five OE lines, and the expression level of HpLHY in OE1 to OE3 was higher than that in OE4 and OE5 ([85]Fig. 4A). We further examined the protein levels of Flag tag on overexpression vectors in these transgenic plants by Western blot analysis. We observed that no Flag tags were detected in both Ws and cca1 lhy, and relative protein level of Flag in OE1 to OE3 lines was higher than that in OE4 and OE5 ([86]Fig. 4, B and C). Therefore, the three lines OE1 to OE3 were ultimately chosen for subsequent experiments. Fig. 4. Overexpression of HpLHY in lhy cca1 attenuated lhy cca1 phenotype and reduced melatonin accumulation in Arabidopsis. [87]Fig. 4. [88]Open in a new tab (A) HpLHY expression levels in Ws, lhy cca1, and five OE transgenic lines analyzed by RT-PCR. (B and C) Flag tag protein accumulation in Ws, lhy cca1, and five OE transgenic lines analyzed by Western blot analyses using an anti-Flag antibody and the anti–Plant Actin antibody as protein loading control. (D and E) The overexpression of HpLHY in lhy cca1 attenuated the lhy cca1 phenotype of hypocotyl shortening. Data are presented as the means ± SD (n = 20). Different letters indicate significant differences (P < 0.05) between each group tested by one-way ANOVA. (F) HpLHY and AtSNAT expression levels in Ws, lhy cca1, and OE lines at ZT4 and ZT16 analyzed by RT-qPCR. Data were normalized to AtACT and different letters indicate significant differences (P < 0.05) between each group according to the Tukey post hoc multiple range test. (G) Melatonin accumulation in Ws, lhy cca1, and OE lines at ZT4 and ZT16 analyzed by liquid chromatography–mass spectrometry (LC-MS). Different letters indicate significant differences (P < 0.05) between each group according to the Tukey post hoc multiple range test. As is well known, CCA1 or LHY functional mutants exhibit shortened hypocotyl phenotypes in Arabidopsis ([89]22, [90]26). HpLHY, as a homologous gene, should also have the same function. As shown in [91]Fig. 4 (D and E), the hypocotyl length of the OE-HpLHY lines have nearly returned to the level of the WT. Subsequently, plant samples were taken from ZT4 and ZT16 during the 12 L/12D photoperiod, and the expression levels of HpLHY and AtSNAT, as well as the content of endogenous melatonin, were measured. As shown in [92]Fig. 4F, the expression level of HpLHY in the OE lines is significantly higher than that in the WT and lhy cca1, and it is higher at ZT16 than at ZT4. This result once again proves that HpLHY is a dawn gene. Compared to the cca1 lhy, the expression level of AtSNAT in OE lines was inhibited, while at ZT16, the level of AtSNAT in the OE lines is more significantly inhibited. By detecting endogenous melatonin content, all three Arabidopsis lines showed a higher melatonin content at ZT4 than at ZT16, and the melatonin content in the cca1 lhy significantly increased, while the melatonin content in the OE returned to the WT level ([93]Fig. 4G). Overexpression and knockout of HpLHY in H. perforatum By overexpressing HpLHY in the wild-type H. perforatum, the 926-bp fragment of the CaMV35S promoter was detected by PCR in different hairy root (HR) lines (fig. S7A). The relative expression level of HpLHY in empty vector (EV) and five OE HR lines was shown in fig. S7B. The expression level of HpLHY in OE was significantly higher than that in EV, especially in OE-3 and OE-5. To detect whether the sequence of HpLHY in the knockout strain (KO) has changed (whether there are base deletions/additions near the target), the genomic DNA of the different knockout lines were extracted. The DNA sequencing results of the KO lines revealed a mutation in the sequence near the target. The KO-1 had five missing bases in the HpLHY sequence, one missing base in KO-3, and eight missing bases in KO-4. These base deletions caused a change in the amino acid encoded by HpLHY, resulting in a frameshift mutation (fig. S7C). As shown in [94]Fig. 5A, in the first month of growing HR on the explants, no differences were observed between different transgenic lines. However, after cutting HR from the explants and growing in liquid medium for the second month, the OE HR line showed remarkable growth retardation, while the growth of KO HR lines appeared to be almost identical to that of EV. By exploring the transcription and translation levels of HpLHY in different HR, it was found that the expression and protein levels of HpLHY in OE lines were increased, while KO lines showed the opposite results ([95]Fig. 5, B and C). The OE HR lines significantly reduced the content of melatonin, while the KO showed an increase in melatonin content. The expression and protein levels of HpSNAT1 in different HR lines are also positively correlated with the content of melatonin ([96]Fig. 5, D to F). Fig. 5. Overexpression and knockout of HpLHY in H. perforatum. [97]Fig. 5. [98]Open in a new tab (A) The phenotypes of different transgenic hairy roots in H. perforatum. The hairy roots from the explants that have grown for 1 month were cut off and continued growing in liquid MS medium for another month. (B) The HpLHY expression levels in EV, OE, and KO transgenic lines analyzed by RT-qPCR. (C) HpLHY protein accumulation in EV, OE, and KO transgenic lines analyzed by Western blot analyses using an anti-HpLHY antibody and the anti–Plant Actin antibody as protein loading control. (D) The HpSNAT1 expression levels in EV, OE, and KO transgenic lines analyzed by RT-qPCR. Data were normalized to HpACT2 and presented as the means ± SD (n = 3). (E) HpSNAT1 protein accumulation in EV, OE, and KO transgenic lines analyzed by Western blot analyses using an anti-HpSNAT1 antibody and the anti–Plant Actin antibody as protein loading control. (F) Melatonin accumulation in EV, OE, and KO transgenic lines analyzed by LC-MS. Different letters indicate significant differences (P < 0.05) between each group tested by one-way ANOVA. Induction of callus, adventitious buds, and roots was performed on hairy root lines OE-2, OE-3, and OE-5 and KO-1, KO-3, and KO-4 to obtain complete transgenic plants of H. perforatum, as shown in [99]Fig. 6A. Similar to the growth of HR, OE plants exhibit delayed growth during callus induction and differentiation of buds and roots. The melatonin content of different transgenic H. perforatum plants was detected at different times under 12 L/12D photoperiod conditions. In EV H. perforatum, melatonin accumulation was still rhythmic and peaked in the ZT4 ([100]Fig. 6B). The phase of this rhythm appeared lost in the KO lines, as there is no statistically significant difference at ZT4 compared to other time points. The melatonin levels of OE plants substantially decrease and exhibit abnormal rhythms. These results propose a circadian rhythm control model for melatonin accumulation under 12 L/12D photoperiod, where HpLHY inhibits the expression of HpSNAT1 gene, leading to a decrease melatonin accumulation. Fig. 6. The phenotypes and melatonin accumulation of different transgenic H. perforatum plants. [101]Fig. 6. [102]Open in a new tab (A) The induction process of regenerated plants from hairy roots of H. perforatum. 1M, inducing hairy roots from explants; 2M, differentiation of hairy roots to form callus tissue; 3M, differentiation of callus tissue into buds; 4M, roots induction to form complete transgenic H. perforatum plants. (B) Overexpression of HpLHY results in reduced melatonin levels. EV, OE, and KO transgenic seedlings were grown in MS medium and entrained to 12 L/12D photoperiod. Leaves were harvested at 2-hour intervals across a 24-hour period for melatonin quantification. Data represent the mean from technical triplicates and three biological replicates. White and black bars above the chart indicate days and nights, respectively. Error bars presented as the means ± SD (n = 3). DISCUSSION Melatonin not only plays a role as a growth regulator and broad-spectrum antioxidant but also participates in plant growth and development and response to stress. The genes involved in melatonin biosynthesis have been identified in various plants ([103]33, [104]34), but their extremely low plant expression levels and transcription factors regulated upstream are still unclear in plants, resulting in the molecular regulatory mechanism of melatonin being in its early stages. This study constructed a yeast cDNA library of H. perforatum and used the proHpSNAT1 as a bait to screen for upstream regulatory proteins by using yeast one-hybrid (Y1H) technology, providing a reliable molecular basis for revealing the regulatory mechanism of melatonin biosynthesis. Last, 96 monoclonal colonies were amplified, and 39 different sequences were obtained after BLAST alignment analysis. Among them, there are 12 transcription factors screened that have regulatory relationships with HpSNAT1. The first type of circadian clock–related proteins, such as the LHY that forms the feedback regulatory loop of the biological clock. The second type is photoresponse related proteins, such as GATA binding protein (GATA1) that is involved in the light morphogenesis ([105]35); chlorophyll A/B binding protein (Lhcb4.2), which can bind chlorophyll and carotenoids and capture light energy to funnel them to the photosystem II reaction center ([106]36); G-box binding factor (GBF1), which plays a key role in the photomorphogenesis and chloroplast function ([107]37, [108]38); and photosensitive pigment interaction factor (PIF7), which regulates responses to prolonged red light by modulating phyB levels ([109]39). The third type of proteins responds to biotic and abiotic stress, such as WRKY11 ([110]40, [111]41). The fourth type is transcription factors related to plant growth and development, aging, hormones, and biological responses, such as the auxin response factor ARF8 that regulates the development of immature flower pistils ([112]42). There are various types of transcription factors that HpSNAT1 responds to, all of which may be involved in regulating the synthesis of endogenous melatonin in plants. Therefore, from the results, it can be seen that HpSNAT1 is regulated by various types of transcription factors, so these proteins may be involved in regulating the synthesis of endogenous melatonin in plants. Circadian clock genes CCA1 and LHY in controlling circadian rhythms have been widely researched, and their molecular mechanisms have also been well studied ([113]22, [114]43, [115]44). However, the other functions of CCA1 and LHY are still unclear, especially in the regulation of secondary metabolites in medicinal plants. CCA1 has high sequence homology with LHY. Although this gene duplication event is thought to have occurred during the evolution of the Brassicaceae family, there is only a single CCA1/LHY-like gene present in some terrestrial plant clades with similar duplications ([116]45), and H. perforatum is one of them. BLAST was performed by using AtLHY and AtCCA1 with the genome database, while anchoring to the same gene (Scaf 346.283), with similarity of 64 and 59%, respectively. This result is consistent with the evolutionary analysis in [117]Fig. 2A. Harmer et al. were the first to study the roles for LHY in gene regulation and found that LHY binds to the promoter region of 519 genes in the genome of Arabidopsis and these promoters almost all contain the circadian-regulated EE ([118]46). Our in vitro experiments have once again confirmed this result, and it serves as an inhibitor. Melatonin exhibits a pronounced daily (circadian) variation in its levels in animals. Melatonin levels typically increase sharply at night after the onset of darkness. This surge signals to the body that it is time to initiate nocturnal activities, such as sleep. During the daytime, melatonin levels drop evidently, reaching their lowest point. This low level helps the body remain awake and alert ([119]47, [120]48). However, although we found rhythmic changes in the melatonin signal in H. perforatum, unlike in animals, the melatonin signal in H. perforatum reaches its peak in the morning, similar to Arabidopsis ([121]20). Therefore, we constructed overexpression and knockout vectors of HpLHY to investigate the regulatory function of HpLHY in the biosynthesis of melatonin. By detecting the expression levels of related genes and melatonin content in different transgenic lines, overexpression of HpLHY significantly reduced the expression of HpSNAT1, thereby inhibiting melatonin synthesis, while knockout strains showed the opposite. In the process of constructing HpLHY-overexpressing H. perforatum and Arabidopsis, it was found that the growth and development of plants were inhibited, resulting in stunted growth, reduced biomass, and growth inhibition. It is speculated that overexpression of HpLHY may lead to incorrect perception of the light cycle in H. perforatum, causing the loss of normal circadian rhythm in the biological clock and hindering normal plant growth. This suggests a previously unidentified mechanism connecting circadian regulation with melatonin biosynthesis. In summary, our data indicate that endogenous concentrations of melatonin and the expression of its biosynthesis-related genes exhibit obvious rhythmic changes, reaching their peak in the morning during the day, whereas the rhythmic expression of HpLHY is exactly opposite to it, reaching its peak in the late night ([122]Fig. 7). In the dark cycle, as the expression level of HpLHY begins to increase, the negative regulatory effect gradually increases by binding to the upstream EE MOTIF (AAATATCT) of the snap gene, leading to a decrease in the synthesis of melatonin at night. However, the endogenous biological clock in plant is a mechanism through which a series of genes and protein networks regulate the time interval between life experiences, and an LHY cannot fully represent it. Further research is needed to explore the possible relationship between melatonin rhythms and the circadian clock genes. Fig. 7. Relationship between circadian oscillator component LHY and melatonin biosynthesis in H. perforatum. [123]Fig. 7. [124]Open in a new tab LHY binds to EE motif and suppresses transcription of SNAT, which encodes a serotonin N-acetyltransferase participated in melatonin biosynthesis. Genes, pink boxes; protein, blue circle. Melatonin indicates the relative concentration of melatonin in H. perforatum. Shaded boxes on the time course of melatonin indicate the dark period within the day:night cycle. MATERIALS AND METHODS Plant materials and growth conditions WT H. perforatum seeds preserved in our laboratory were sterilized with 12% chloros solution and transferred to Murashige and Skoog (MS) agar medium (Solarbio, Beijing, China). After incubation at 4°C in the dark for 3 days, the plates were placed in a growth chamber at 23°C with 60% relative humidity under cool light-emitting diode white light at light intensity of 108 μmol m^−2 s^−1. The seedlings were grown under 12 L/12D conditions for 1 month and then transferred into 12 L/12D cycles or continuous light conditions. Arabidopsis thaliana lhy-11 cca1-21 double mutant [Wassilewskija (Ws) ecotype, lhy cca1] seeds were from Q. Sun ([125]26), and WT Columbia-0 (Col) and Ws seeds were obtained from the European Arabidopsis Stock Centre. The seeds were sterilized with 50% bleach and grown on one-half-strength MS agar medium. The growth conditions are the same as above. RNA-seq analysis At ZT4 and ZT16, total RNA was extracted from Arabidopsis WT Col-0 and overexpressing OE-HpSNAT1 lines using the FastPure plant RNA kit (Vazyme, Nanjing, China). The quality assessment was conducted using an Agient2100/LabChip GX. RNA-Seq libraries were prepared using the VAHTS Stranded mRNA-seq Library Prep Kit (Vazyme, Nanjing, China) and sequenced on an Illumina 150-bp paired-end platform. The DEGs between the lines were conducted using Cufflinks v2.2.1 ([126]49). The DEGs were decided on the basis of a P value <0.05 and at least a twofold change between the two FPKM (fragments per kilobase per million mapped reads) values ([127]50). KEGG pathway enrichment analysis was performed using BMKCloud ([128]www.biocloud.net) ([129]51). Construction of the H. perforatum cDNA yeast library Using SMART technology to construct a H. perforatum yeast cDNA library, the specific method is as follows ([130]52). The mRNA samples were isolated by the mRNA Isolation Kit (Oligo-dT Cellulose; Zeye, Shanghai, China) according to the instructions. First-strand and second-strand cDNA were synthesized using the Maxima Double-Stranded cDNA Synthesis Kit (Thermo Fisher Scientific, Waltham, USA). The 8 μl of 2× hybridization buffer was added into the Double-Stranded cDNA and incubated at 98°C for 2 min and at 68°C for 5 hours. Then, 5 μl of 2× DSN (duplex-specific nuclease) Master buffer and 0.25 μl of DSN (1 U/μl) were added and incubated at 68°C for 25 min. Last, CHROMA SPIN + TE-1000 Columns (Clontech, USA) were used for small-fragment removal. The double-stranded cDNA of H. perforatum was ligated to the pGADT7 vector to construct pGADT7-HpcDNA by homologous recombination. The ligation system contained 2 μl of cDNA, 200 ng of linearized pGADT7 DNA, 2 μl of Exnase II enzyme (Vazyme, Nanjing, China), 4 μl of buffer, and double-distilled H[2]O to 20 μl ([131]53). Escherichia coli is transformed, and the parameters of the electroconvulator is set to 2.5 kV. After electroconvulation, the bacterial solution is coated on LB plates containing ampicillin antibiotics and incubated overnight at 37°C. Gene cloning and vector construction Genomic DNA and RNA were extracted using the Plant RNA/ DNA Kit (Vazyme, China). The cDNA was synthesized by using the SuperScript IV First-Standard cDNA Synthesis Kit (Invitrogen, Shanghai, China). Coding sequence (CDS) for H. perforatum Scaf 346.283 (HpLHY) and the promoter (the upstream about 1.5 kb of ATG) of Scaf 151.204 (HpSNAT1) were amplified by PCR using the gene-specific primers (table S2). To screen the yeast cDNA library, a 1522-bp promoter sequence of HpSNAT1 (proHpSNAT1) cloned into pHIS2 was used as the bait vector. The vector pEarleyGate202 was used for constitutive overexpression, and pYLCRISPR/Cas9-H was used as the vector for CRISPR-Cas9 constitutive system. The CDS of HpLHY was cloned into pBI221-GFP produce the 35S::HpLHY-GFP constructs for subcellular localization. To explore the regulatory relationship between HpLHY and HpSNAT1, CDS of HpLHY was cloned into pGADT7-AD, and the promoter fragments were cloned into pHIS2 vector for Y1H assays; CDS was cloned into pMAL-c5x vector to produce the HpLHY-MBP recombinant for EMSA; CDS was cloned into pGreenII62-SK and the promoter fragments cloned into pGreenII 0800-LUC vector for transient dual-luciferase reporter assay. HpSNAT1 self-activation detection and screening library The bait pHIS2-proHpSNAT1 recombinant plasmid was transformed into yeast strain Y187 (Weidi, Shanghai, China) for self-activation detection to screen for an appropriate concentration of 3AT. Three single bacterial colonies were randomly selected from each plate, diluted, and applied onto corresponding defective plates without the -His, but with different concentrations (0, 10, 20, 30, 40, 50, 75, and 100 mM) of 3AT added. The colonies were incubated at 28°C for 3 days. For Y1H screening library, pHIS2-p53 and pGAD53m are as positive control ([132]54). Then, the Y187 yeast transformant containing pHIS2-HpSNAT1 bait plasmid was used as the receptor strain to prepare the receptive cell. The library plasmid pGADT7-HpcDNA was transferred into above Y187 competent cells and coated on SD/-Trp/-Leu/-His–containing 3AT plates. The cells were incubated at a constant temperature of 28°C for 3 days, and the transformation results were observed. The initial positive transformants were scribed onto another SD-TLH + 3AT plate and continued to be incubated at 28°C for 3 days. Positive yeast clones were cloned using the Yeast Colony Rapid Detection Kit (Pronetbio, Nanjing, China), followed by DNA sequencing and BLAST alignment analysis with sequences in the H. perforatum GenBank database (PRJNA588586). Transcriptional activation and Y1H rotation verification Following the procedures previously described ([133]55), the CDS of HpLHY (1308 bp) obtained from the above library screening was cloned into pGBKT7 to experiment the transcriptional activity of HpLHY. The recombinant vectors pGBKT7-HpLHY, pGBKT7-VP16 (positive control), and the empty pGBKT7 (negative control) were transferred into the AH109 yeast-competent cells (Weidi, Shanghai, China), respectively. The transformants were grown on SD/−Trp (SD-T) and SD/−Trp/−Ade/-His/X-α-gal (SD-TAH + X-α-gal) deficiency media. Then, the transcriptional activities were tested according to their growth status at 28°C for 3 days in darkness. The promoter of HpSNAT1 was divided into four segments (P1, 370 bp; P2, 415 bp; P3, 207 bp; P4, 530 bp) based on the predicted binding sites, and they were cloned into the pHIS2 vector as the reporter. The two recombinant vectors described above were transformed into yeast Y187 strain and screened on SD-TL plates at 28°C for 3 days. The surviving colonies were then transferred to SD-TLH containing 3AT to observe the interaction between HpLHY protein and the elements in promoters. Subcellular localization Transient protoplast transformation was performed as previously described ([134]56). Twenty-five– to 30-day Arabidopsis leaves (without bolting) were cut into 1-mm-wide strips and completely immersed in enzymatic hydrolysates [1% cellulase R10, 0.2% mince protease R10, 0.4 M mannitol, 20 mM KCl, and 20 mM MES (pH 5.7)]. The samples were digested in the dark at 55°C for 10 min and at 25°C for 5 hours. The sample was washed with solution W5 [154 mM NaCl, 125 mM CaCl[2], 5 mM KCl, 5 mM glucose, and 0.03% MES (pH 5.8)], and the protoplast was suspended in MMG solution (15 mM MgCl[2], 0.1% MES, and 0.4 M mannitol, adjusting pH 5.6 with KOH) for microscopic examination (×40 magnification, 20 to 40 cells each field of view). Protoplast suspension and 35S: HpLHY GFP plasmid DNA were taken, mixed gently and evenly with polyethylene glycol, molecular weight 4000 solution, and placed at 25°C for 30 min. The mixture was diluted with W5 to terminate the reaction and centrifuged at 300 rpm for 3 min to collect protoplasts before washed with W5 one to two times and incubated in the dark at 28°C for 18 to 24 hours. The supernatant was removed and only about 100 μl of protoplasts was left for observation under a Nikon C2-ER laser confocal microscope ([135]57). RT-qPCR analysis GenScript ([136]www.genscript.com/) was used to design the quantitative primers (table S2). RT-qPCR was performed using the Roche LightCycler 96 system according to the manufacturer’s instructions. The reaction was performed in a volume of 20 ml, containing 5 ml of 30-fold diluted synthesized cDNA, 10 ml of SYBR Master Mix (Vazyme, China), 0.4 ml of 10 mM forward primer, 0.4 ml of 10 mM reverse primer, and 4.2 ml of deoxyribonuclease/ribonuclease-free deionized water. The cycling conditions were 95°C for 30 s, 40 cycles of 95°C for 10 s, and 60°C for 30 s, and a final melting curve analysis. No-template controls were included, and each of three biological samples was performed in triplicate. The relative expression level of each gene was calculated with the 2^−ΔΔCt method, and HpACT2 (GenBank: [137]MK054303) was used as an internal control as described previously ([138]58). Luciferase assay Agrobacterium tumefaciens strain GV3101 harboring targeted fragments was grown in LB medium [tryptone (10 g/liter), yeast extract (5 g/liter), and NaCl (10 g/liter)] to an optical density at 600 nm (OD[600]) of 1.5. Bacterial cells were resuspended in MES buffer (10 mM MgCl[2], 10 mM MES, and 100 μM acetosyringone) to an OD[600] of 1.0 and then introduced via a syringe into the leaves of about 4-week-old Nicotiana benthamiana plant ([139]59). After 48 hours, the leaves were coated with 0.1 M luciferase substrate (Chemstan, Wuhan, China) and subjected to dark treatment for 10 min before being placed in a charge-coupled device camera (Spectral, Colorado, USA) to observe luciferase activity detected by the Caliper IVIS Lumina II (Xenogen IVIS 200, USA). Then, the LUC/REN ratios (firefly luciferase/Renilla luciferase) were measured with a LUC kit (Promega, USA) according to the manufacturer’s instructions. Electrophoretic mobility shift assay The fusion proteins of HpLHY were generated through prokaryotic expression in vitro. Expression of MBP fusion proteins in Rosetta cells was induced by adding 0.1 mM isopropyl-β-D-1-thiogalactopyranoside to the culture medium and incubating the cells at 120 rpm, 16°C for 20 hours. The PurKine MBP-Tag Protein Purification Kit (Dextrin, Wuhan, China) was used for protein purification. The sequence of approximately 40 bp including the predicted binding site was used as a probe, and biotin labeling was performed at the 5′ end of the primer, followed by high-performance liquid chromatography (HPLC) purification. The P1 WT probe and P1-mutant probe (P1m) from the proHpSNAT1 were synthesized by Sangon (Shanghai, China). EMSA was performed as previously described ([140]60). Incubation of proteins and probes was performed following the instruction of the EMSA/Gel-Shift Kit (Beyotime, Shanghai, China). After performing electroblotting and ultraviolet cross-linking (1000 J, 10 min) on a nylon (Millipore, Germany), the membrane was incubated in blocking buffer for 30 min and rinsed in washing buffer. The chemiluminescence imaging system (Enfan, Shaanxi, China) was used to visualize the signal. Agrobacterium-mediated transformation The pEarleyGate202-HpLHY (688-HpLHY) and pCas9-HpLHY plasmid with the correct insertion was introduced into Agrobacterium rhizogenes strain K599 (Weidi, Shanghai, China). The roots of 3-month-old aseptic H. perforatum seedlings were used as explants to induce hairy root. The positive and high-expression hairy root lines were cut into 1.0- to 1.5-cm root segments and put on solid MS medium with phytohormones for plant induction ([141]31). The 688-HpLHY plasmid was transferred into A. tumefaciens strain GV3101 (Weidi, Shanghai, China) and transformed into the Arabidopsis lhy-11 cca1-21 using the flower dipping method ([142]61). Positive lines were selected on the basis of their resistance to BASTA. For specific steps, refer to the previous detailed description ([143]30). Western blot analysis By using the Total Protein Extraction Kit (Solarbio, Beijing), 100 mg of whole Arabidopsis seedlings and 100 mg of H. perforatum hairy roots were harvested for protein extraction. For the detection of proteins by immunoblotting in Arabidopsis, anti–Plant Actin antibody (Abbkine, Wuhan) and anti-Flag antibody (Yeasen, Shanghai, China) were used. Antibodies for HpLHY protein in H. perforatum were prepared by Nanjing Jinsrui Biological Co., with amino acid sequences of CESHRPSSPTEENSS. Melatonin extraction and detection The plant tissues from three independent biological samples were quickly frozen in liquid nitrogen and dried in a vacuum freeze-drying machine. The extraction and detection of melatonin was performed as previously described ([144]31). Dry powder (150 mg) was accurately weighed into 500 μl of 80% methanol, 80 HZ, 45°C, and sonicated for 40 min. The supernatant filtered through a 0.22-μm filter was collected for subsequent experimental detection. Melatonin content was detected on an Agilent 1260 HPLC system coupled to an Agilent 6460 QQQ LC/MS system. Chromatography was performed on a Welch UItimate XB-C18 column (150 mm by 2.1 mm, 3-μm particle size) with a flow rate of 0.3 ml/min and a 5-μl injection volume. The mobile phase included solvent A (acetonitrile) and solvent B (0.1% formic acid in deionized water) and followed the following gradient profile: 0 to 10 min, A 2 to 55%, and B 98 to 45%. Analytes were detected in positive ionization mode, and the drying gas flow was 10 liters/min. Statistics All experiments were repeated using three biological replicates and three technical replicates unless otherwise specified. The data were calculated as the means ± SEM. The statistical analyses were calculated by GraphPad Prism 8.0.2. Acknowledgments