Introduction

The circadian rhythm is an approximately 24 h cell-autonomous biological rhythm observed in virtually all living organisms. The mammalian circadian clock involves a transcription-translation feedback loop in which CLOCK and BMAL1 heterodimers activate the PERIOD (PER) and CRYPTOCHROME (CRY) genes, which then provide feedback and repress their transcriptions1,2. It has been reported that chronic circadian disruption is associated with many diseases, such as cardiovascular and cerebrovascular diseases3, cancer4, immune5, and metabolic disorders6.

The development of new pharmaceuticals is time-consuming and expensive. Therefore, identifying new functions for existing drugs, known as the drug repositioning approach, is a popular and powerful approach7. Crude drugs used in traditional Japanese Kampo medicine are the major sources of new chemical entities for drug discovery8. In this study, we screened 137 crude drug extracts to identify circadian clock modulators using a human osteosarcoma U2OS cell line stably expressing the clock reporter Bmal1-dLuc and identified 17 hit crude drugs. We noticed that more than half of the major ingredients of these crude drugs targeted AKT and relevant signaling pathways. Therefore, we examined the effects of the major ingredients targeted AKT. We then examined the effects of AKT activator and inhibitor, as well as the triple knockdown of AKT1/2/3 using siRNA.

Results

Identification of crude drug extracts that modulate the circadian clock

To identify circadian clock modulators, 137 crude drug extracts were screened in Bmal1-dLuc U2OS for primary screening7. Since crude drug extract is a mixture of many active ingredients and their effects could be highly variable, all crude drug extracts were tested in triplicates at two different concentrations (20 µg/mL and 0.20 mg/mL) in a 384-well plate format, and the experiments were repeated twice (Trials 1 and 2) to reduce the number of false positives. Two crude drug (Polygalae Radix and Allii Chinense Bulbus) extracts at 20 µg/mL and four crude drug (Pueraiae Radix, Platycodi Radix, Chrysanthemi Flos, and Paeoniae Radix) extracts at 0.20 mg/mL were identified to show period lengthening effects by one or more hours, respectively (Fig. 1, Table 1). Five crude drug (Artemisiae Capillaris Flos, Polygalae Radix, Perillae Herba, Caryophylli Flos, and Alpiniae Officinari Rhizoma) extracts at 20 µg/mL and eight crude drug (Chrysanthemi Flos, Schizonepetae Spica, Cicadae Periosrtacum, Uncariae Uncis Cum Ramulus, Lonicerae Folium Cum Caulis, Perillae Fruitus, Acorus Graminei Rhizoma, and Mume Fructus Praeparatus) extracts at 0.20 mg/mL were identified to show phase changing effects by two or more hours, respectively (Fig. 1, Table 1). To validate the 17 potential hit crude drug extracts, we performed secondary screening for dose-dependency and confirmed the period and/or phase-changing effects for all 17 crude drug extracts (Fig. 2). To determine whether the effects of crude drugs on the circadian clock were reporter gene-, cell type-, or tissue-specific, the effects of three representative hit crude drugs (Artemisaie Capillaris Flos, Perillae Herba, and Allii Chinense Bulbus) were examined in Rat-1 Per2-luc fibroblasts9 (Fig. 3A). The period lengthening and phase changing effects of the extracts of Artemisiae Capillaris Flos and Perillae Herba were consistent with the results obtained for U2OS cells. Interestingly, however, Allii Chinense Bulbus extract showed period-shortening effects. When we examined the effects of these three crude drug extracts in lung explant cultures from Per2::Luciferase knockin mice7, all three crude drug extracts showed period-lengthening effects (Fig. 3B). We also examined the effect of Allii Chinense Bulbus extract on the locomotor activity rhythm of zebrafish larvae in vivo under constant darkness and observed a period shortening effect and a tendency to phase change (Fig. 4). As we are aware that some of the compounds affect the bioluminescence amplitude through the activation or inhibition of luciferase activity but not through the circadian clock, we did not focus on the circadian amplitude in the present study.

Figure 1
figure 1

Screening of crude drug extracts for circadian clock modulators. A total of 137 crude drug extracts were screened in Bmal1-dluc U2OS cells at 20 µg/mL (top) and 0.20 mg/mL (bottom) for circadian period (left) and phase (right) changes. (Left) Two independent trials identified two crude drugs at 20 µg/mL and four crude drugs at 0.20 mg/mL that show consistent period lengthening effects by 1 or more hours. (Right) This screening also identified five crude drugs at 20 µg/mL and eight crude drugs at 0.20 mg/mL that show consistent phase changing effects by 2 or more hours.

Full size image
Table 1 List of hit crude drugs.
Full size table
Figure 2
figure 2

Dose-dependent effect of 17 hit crude drug extracts in U2OS cells expressing Bmal1-dLuc. Luminescent traces (left) and dose-dependent effects on the circadian period (middle) and phase (right) in U2OS cells. Values are averages of six replicates ± SD and were analyzed using one way ANOVA, followed by Dunnett’s multiple comparisons test (*p < 0.05, **p < 0.01, ***p < 0.001, **** p < 0.0001).

Full size image
Figure 3
figure 3

Effect of three crude drug extracts on Rat-1 fibroblasts expressing Per2-luc and lung explants from Per2::Luciferase knockin mice. (A) Luminescent traces (left) and dose-dependent effects on the circadian period (middle) and phase (right) in Rat-1 cells. Values are averages of three replicates ± SD and were analyzed using one-way ANOVA, followed by Dunnett’s multiple comparisons test (*p < 0.05, **p < 0.01, ***p < 0.001, **** p < 0.0001). (B) Effects of the extracts of Artemisiae Capillaris Flos (top), Perillae Herba (middle), and Allii Chinense Bulbus (bottom) on lung explants of Per2::Luciferase knockin mice. Luminescent traces (left) and effects on period (middle) and phase (right). Values are averages of three replicates ± SD and were analyzed using Student's t-test. (*p < 0.05).

Full size image
Figure 4
figure 4

Effects of Allii Chinense Bulbus on locomotor activity rhythms of zebrafish larvae. (A) Locomotor activity rhythms under Allii Chinense Bulbus extract (10 µg/mL) and MeOH control. The moving distance (mm) per 10 min were averaged and plotted (n = 11, 12). (B,C) Effects on circadian period (B) and phase (C) are shown. Values are average ± SD and were analyzed by Student's t-test (n = 11, 12, *p < 0.05).

Full size image

Ingredients of hit crude drugs modulate the circadian rhythm

As previously mentioned, the crude drug extract is a mixture containing many active ingredients. In addition, because the administration of crude lysates sometimes affects the experimental results owing to the formation of macromolecular complexes and/or medium pH change10, validation using known active ingredients is important. The well-known active ingredients for hit crude drugs are summarized in Table S1. To further understand the possible mode of action of these crude drugs, we searched for known targets of these active ingredients from previous literature and found that AKT, NFκB, and mTOR are the most common targets for these active ingredients (Table S1). To test whether the major ingredients that target AKT, NF-κB, and mTOR can modulate the circadian rhythm, the effects of eight major ingredients were examined in U2OS cells, and it was observed that treatment with these ingredients leads to significant changes in the circadian period and/or phase (Fig. 5).

Figure 5
figure 5

Effects of 8 major ingredients of hit crude drugs. Luminescent traces (top) and dose-dependent effects on the circadian period (middle) and phase (bottom) in U2OS cells. Values are the mean of six replicates ± SD and were analyzed using one way ANOVA, followed by Dunnett’s multiple comparisons test (*p < 0.05, **p < 0.01, ***p < 0.001, **** p < 0.0001).

Full size image

Involvement of AKT signaling in the circadian clock

It is well established that the signaling pathways for NF-κB, mTOR, and AKT are closely related, and AKT is a major signaling hub (Fig. 6A). Involvement of Akt in the circadian clock has been reported in Drosophila11. There are three isoforms of AKT in mammals. However, analysis of the circadian behavior in Akt1-/- mice revealed a normal behavioral rhythm12. We examined the effects of AKT activator and inhibitor to explore the role of AKTs in the mammalian circadian clock. The AKT activator SC79 shortened the circadian period and advanced the phase at the highest dose (Fig. 6B). In contrast, the AKT inhibitor A-443654 shortened the circadian period at lower doses and lengthened the period at the highest dose (Fig. 6C). A-443654 also delayed the circadian phase in a dose-dependent manner (Fig. 6C). Furthermore, an siRNA-mediated triple knockdown of AKT1/2/3 shortened the circadian period (Fig. 6D).

Figure 6
figure 6

Involvement of AKT signaling in the circadian rhythm. (A) Schematic of interactions of the AKT pathway. (B,C) Effects of AKT activator SC79 (B) and AKT inhibitor A-443654 (C) at different concentration. Luminescent traces (left) and dose-dependent effects on period (middle) and phase (right) in U2OS cells. (D) Effects of the knockdown of AKT1/2/3 by siRNA. U2OS cells were transfected with negative control (NC) siRNA, AKT1 siRNA, AKT2 siRNA, or AKT3 siRNA. Luminescent traces (left) and effects on circadian period (middle) in U2OS cells. Efficiency of knockdown of AKT was confirmed using RT-qPCR (right). Values are the averages of six replicates ± SD and were analyzed using one way ANOVA, followed by Dunnett’s multiple comparisons test (*p < 0.05, **p < 0.01, ***p < 0.001, **** p < 0.0001).

Full size image

Discussion

Chronic disruption of the circadian clock due to shift work or travel across time zones has long-term consequences on human health, resulting in an increased risk of many diseases such as diabetes, cardiovascular disease, and cancer3,4,5,6. The development of new pharmaceuticals is costly and time-consuming, with a high rate of failure. Identifying new functions for existing drugs, known as the drug repositioning approach, is a popular and powerful approach, and crude drugs used in traditional Japanese Kampo medicine are major sources of new chemical entities for drug discovery7,8. Indeed, Motohashi et al. succeeded in identifying circadian clock modulators after screening 40 crude drugs13.

In this study, we screened 137 crude drug extracts, of which approximately 12% modulated circadian rhythms (Figs. 1 and 2, Table 1). In our previous study, using the US Food and Drug Administration (FDA)-approved drug library and the International Drug Collection (IDC) library, we found that approximately 5% of the existing drugs alter circadian rhythms7. The hit rate of Kampo medicines was much higher than that of other existing drug libraries. The reason for the high hit rate is probably because the crude drug extract is a mixture containing many active ingredients. Indeed, the effects of the major ingredients were not as strong as those of the original crude drugs (Fig. 5).

To further test whether these effects are specific to the reporter gene (Bmal1-dLuc) or cell type (U2OS cells), we examined the effects of three representative hit crude drugs (Artemisiae Capillaris Flos, Perillae Herba, and Allii Chinense Bulbus) in Rat-1 fibroblasts expressing Per2-luc and lung explants from Per2::Luciferase knockin mice (Fig. 3). Although the effects of the extracts of Artemisiae Capillaris Flos and Perillae Herba were consistent with those observed in U2OS cells, Allii Chinense Bulbus extract showed the opposite period-shortening effects in Rat-1 cells (Fig. 3). When we examined the effects of Allii Chinense Bulbus extract on the locomotor rhythm of zebrafish larvae in vivo, we observed period shortening effects, as in the case of Rat-1 cells (Fig. 4). Discrepancies in the effects of crude drugs (i.e., period shortening or lengthening) between different reporters and cell types could be a result of the cell type-specific function of the circadian clock gene14 and/or the dissociation of clock gene rhythmicity within the cells/tissues15. However, as crude drug extracts are a mixture containing many active ingredients, it is also possible that the expression level of the drug targets could be different between different cells and tissues. Notably, tissue-dependent effects of crude drug extracts have also been observed13. This study clearly demonstrated the circadian clock modulating effects of crude drugs at the cellular, tissue, and organism levels.

When we searched for the potential targets of the identified ingredients in previous literature, we noticed that most of the ingredients targeted the NF-κB, mTOR, or AKT signaling pathways (Table S1). Among them, AKT is known to be a major signaling hub (Fig. 6A). It is noteworthy that Akt signaling has been shown to affect the circadian period in Drosophila11. However, there are three isoforms of AKT in mammals, and Akt1-/- mice show normal behavioral rhythms12. In the present study, we found period and phase changes induced by AKT activator and inhibitor. We also confirmed a significant change in the circadian period in the triple knockdown of AKT isoforms, suggesting the involvement of AKT signaling in the mammalian circadian clock (Fig. 6). Notably, PI3K, one of the signaling molecules upstream of the AKT signaling pathway, regulates feeding-mediated entrainment of the peripheral clock16. Furthermore, the CLOCK protein has recently been reported to be the substrate of AKT17. Further analysis of the precise mode of action of Kampo medicines may uncover the role of the AKT signaling pathway in the circadian clock.

In conclusion, the drug repositioning approach is a useful approach for understanding the underlying mechanisms of the circadian clock. Furthermore, tenuifolin, one of the main active ingredients in Polygalae Radix, has been reported to enhance sleep in mice18. Thus, a drug repositioning approach using Kampo medicines is a useful approach for identifying potential therapeutic treatments for circadian misalignment.

Materials and methods

Crude drugs

We selected 137 crude drugs that are frequently used as ingredients in traditional Japanese Kampo formulations, and purchased crude drugs that met the grade standards of the Japanese Pharmacopoeia 17th Edition or of non-pharmacopoeial crude drugs from Daiko Shoyaku (Nagoya, Japan), Tsumura (Tokyo, Japan), Uchida Wakanyaku (Tokyo, Japan), and Tochimoto Tenkaido (Osaka, Japan). The preparation of methanol extracts was described in our previous report8. Each extract was suspended in methanol at a concentration of 100 mg/mL, and stored at − 20 °C. The origins and the distributors of the hit crude drugs are listed in Table 1.

Chemicals

Platycodin D (cat no. CFN98134), luteolin (cat no. CFN98784), luteoloside (cat no. CFN98565), galangin (cat no. CFN98918), kaempferide (cat no. CFN98782), and chrysoeriol (cat no. CFN98785) were purchased from ChemFaces (Wuhan, China). Amygdalin (cat no. A0443-1G) was purchased from Tokyo Chemical Industry (Tokyo, Japan). Apigenin (cat no. 010-18914) was obtained from Fujifilm Wako Pure Chemicals (Osaka, Japan). The AKT inhibitor, A-443654 (cat no. HY-10425), and the activator, SC79 (cat no. HY-18749) were purchased from MedChemExpress (Monmouth Junction, NJ, USA). Each chemical was suspended in DMSO at a concentration of 5 mM, stored at − 20 °C.

Cell culture

Human U2OS cell (HTB-96) and Rat-1 cell were obtained from ATCC (American Type Culture Collection). Human U2OS cells containing a Bmal1‐dLuc reporter gene and Rat-1 cells containing Per2-Luc reporter gene were cultured in Dulbecco's minimum essential medium (DMEM; D2902, Sigma-Aldrich, St. Louis, MI, USA) supplemented with 10% fetal calf serum (FBS,

FB-1290/500; Biosera, Rue Lacaille, Nuaillen, France), 2 mM L‐glutamine, 100 U/mL penicillin, and 100 μg/mL streptomycin (Pen Strep, 15070063, Thermo Fisher Scientific, Waltham, MA, USA) as previously described7.

Chemical screening

U2OS cells were seeded at a density of 4,000 cells/well in 384‐well plates and incubated for 2 days until confluent. Using a Caliper Life Sciences ALH 3000 Workstation, crude drug extracts were diluted to 20 µg/mL and 0.20 mg/mL respectively, in assay media for measuring luminescence. This resulted in a final MeOH concentration of 0.2% in the control and drug‐treated samples. Multiple doses (0.63, 1.3, 2.5, 5.0, 10, 20, 40, and 80 μM) of ingredients and AKT inhibitor A-443654 were analyzed, while AKT activator SC79 was analyzed at eight different doses (0.063, 0.13, 0.25, 0.5, 1.0, 2.0, 4.0, 8.0 μg/mL). Luminescent assay media (also called “air” or recording media) was composed of DMEM, 10 mM HEPES, 3.5 g/L D‐glucose, 0.35 g/L sodium bicarbonate, 100 U/mL penicillin, 100 μg/mL streptomycin, 2% B27 (Life Technologies, Carlsbad, CA, USA), 0.1 mM luciferin (Fujifilm), and 100 nM dexamethasone (Sigma). Diluted extracts were pipetted into triplicate wells of a 384‐well plate containing U2OS cells (prepared above), and bioluminescence was monitored for 1 week using a Churitsu CL384 Series luminometer (Churitsu Electric Corporation, Nagoya, Japan). The circadian period and phase were determined using NINJA SL00‐01 software for time series analysis (Churitsu Electric Corporation), and potential hit compounds were identified based on changes in the circadian period and/or phase.

Lung explant culture

Lungs were dissected from male Per2::Luciferase knockin mice and cultured in 35 mm dishes containing Millicell cell culture inserts (Millipore) in assay media to measure luminescence, as previously described7. The dishes were sealed with silicon grease and parafilm and bioluminescence was measured using a LumiCycle 32 (Actimetrics). The circadian period and phase were determined using LumiCycle Analysis software (Actimetrics). Note that phase-delays and -advances are indicated by positive and negative values, respectively.

Zebrafish behavior assay

Zebrafish (Danio rerio) larvae (3.5 dpf) were placed in each well of a 96 deep-well plate (96 Well Clear Assay Plate Pyramid Flat Bottom); 800 μL E3 medium (5 mM NaCl, 0.17 mM KCl, 0.33 mM CaCl2· 2H2O, 0.33 mM MgSO4· 7H2O, 10 mM HEPES, pH 7.2) containing the compound (10 μg/mL) was added to each well. The plate was set on DanioVision (Noldus) and the behavior of the fish was tracked under constant darkness with a CCD camera with infrared light. The amount of activity (i.e., distance moved) per minute was calculated using the image processing function of EthoVision XT11 (Noldus). The free-running period and phase were calculated using Actogram J and MultiCycle (Actimetrics), respectively19. All animal studies were carried out in accordance with ARRIVE guidelines and all methods were in compliant with relevant guidelines and regulations and were approved by the Animal Experiment Committee of Nagoya University.

siRNA mediated knockdown

For gene knockdown analysis, small interfering RNA (siRNA) targeting the AKT1, AKT2, and AKT3 sequences and non-targeting siRNA were obtained from Qiagen (FlexiTube GeneSolution siRNA, Hilden, Germany). These siRNAs were introduced singly, or as a mixture of two or three siRNAs into U2OS cells using a reverse transfection protocol in 35 mm dishes. Bioluminescence was monitored in a LumiCycle 32 (Actimetrix, Wilmette, IL, USA) for 1 week. The knockdown effect was verified using quantitative PCR. Reverse transcription was performed on total RNA (200 ng) using ReverTra Ace (Toyobo, Osaka, Japan) and oligo-dT primers. Samples contained SYBR Premix Ex Taq II (Takara, Kusatsu, Japan), 0.4 μM gene-specific primers (Table S2) and 2 µL synthesized cDNA in 20 µL. qPCR was performed on an Applied Biosystems QuantStudio 3 Real-Time PCR System (Foster City, CA, USA) as follows: 95 °C for 30 s, followed by 40 cycles of 95 °C for 5 s and 60 °C for 30 s. ΔCt was determined using GAPDH as a housekeeping gene, and relative expression was calculated using the ΔΔCt method by comparing gene‐specific siRNA samples to the negative siRNA control.

Statistical analysis

Statistical analysis was conducted using GraphPad Prism 9. Significance of differences was analyzed using one way analysis of variance (ANOVA) with Dunnett’s multiple comparisons test or Student's t-test. Values are expressed as mean ± SD.