The in vitro real-time oscillation monitoring system identifies potential entrainment factors for circadian clocks

Abstract

Background: Circadian rhythms are endogenous, self-sustained oscillations with approximately 24-hr rhythmicity that are manifested in various physiological and metabolic processes. The circadian organization of these processes in mammals is governed by the master oscillator within the suprachiasmatic nuclei (SCN) of the hypothalamus. Recent findings revealed that circadian oscillators exist in most organs, tissues, and even in immortalized cells, and that the oscillators in peripheral tissues are likely to be coordinated by SCN, the master oscillator. Some candidates for endogenous entrainment factors have sporadically been reported, however, their details remain mainly obscure.

Results: We developed the in vitro real-time oscillation monitoring system (IV-ROMS) by measuring the activity of luciferase coupled to the oscillatory gene promoter using photomultiplier tubes and applied this system to screen and identify factors able to influence circadian rhythmicity. Using this IV-ROMS as the primary screening of entrainment factors for circadian clocks, we identified 12 candidates as the potential entrainment factor in a total of 299 peptides and bioactive lipids. Among them, four candidates (endothelin-1, all-trans retinoic acid, 9-cis retinoic acid, and 13-cis retinoic acid) have already been reported as the entrainment factors in vivo and in vitro. We demonstrated that one of the novel candidates, 15-deoxy-Delta12,14-prostaglandin J2 (15d-PGJ2), a natural ligand of the peroxisome proliferator-activated receptor-gamma (PPAR-gamma), triggers the rhythmic expression of endogenous clock genes in NIH3T3 cells. Furthermore, we showed that 15d-PGJ2 transiently induces Cry1, Cry2, and Roralpha mRNA expressions and that 15d-PGJ2-induced entrainment signaling pathway is PPAR-gamma--and MAPKs (ERK, JNK, p38MAPK)-independent.

Conclusion: Here, we identified 15d-PGJ2 as an entrainment factor in vitro. Using our developed IV-ROMS to screen 299 compounds, we found eight novel and four known molecules to be potential entrainment factors for circadian clocks, indicating that this assay system is a powerful and useful tool in initial screenings.

Figure 1

Screening of entrainment factors by using IV-ROMS. mPer2-luc/Rat1 cells were monitored the luciferase intensity for duration of 3 days using IV-ROMS. Figure 1A shows luciferase activity after 1 h treatment with 50% serum (positive control, left panel) or DMSO (negative control, right panel). Figure 1B shows luciferase activity after 1 h treatment with 1 or 10 μM of each of the 12 possible circadian entrainment factors. A representative result was chosen out of three independent experiments. Abscissa presents "day", ordinate "relative luciferase intensity", respectively. The first peak was set to 1.

Figure 2

Endogenous clock gene expression profiles stimulated by 15d-PGJ₂. NIT3T3 cells were stimulated by 50% serum (positive control; left column), DMSO (negative control; right column), or 10 μM 15d-PGJ₂ (middle panel) for 1 h. Total RNAs were isolated at each time point. Quantitative real-time RT-PCR was performed using mPer2 (upper panels), mBmal1 (bottom panels), and 18S rRNA primers. Abscissa presents "hour", ordinate "mRNA amount", respectively. mRNA amount at time 0 was set to 1. The relative levels of each mRNA were normalized to the corresponding 18S rRNA levels.

Figure 3

15d-PGJ₂ up-regulates Cry1, Cry2, and Rorα mRNA expression. NIH3T3 cells were stimulated by 10 μM 15d-PGJ₂ (solid line) or DMSO (dotted line) for 1 h. Total RNAs were isolated at each time point. Quantitative real-time RT-PCR was performed using mCry1, mCry2, mRorα, mPer1, mPer2, mBmal1, and 18S rRNA primers. Abscissa presents "hour", ordinate "mRNA amount", respectively. Each mRNA amount at time point 0 was set to 1. Data are shown as the mean ± SE from three to five independent experiments. The relative levels of each mRNA were normalized to the corresponding 18S rRNA levels. *p < 0.05, **p < 0.01 compared with relative control (Student's t-test).

Figure 4

Rhythmic clock gene expression triggered by 15d-PGJ₂ is independent of PPAR-γ-, MAPK-, JNK- and p38MAPK-signaling pathway. After 1 h pretreatment of DMSO (15d; upper left panel), 10 μM GW9662 (GW+15d; upper right panel), 10 μM U0126 (U+15d; middle left panel), 20 μM SP600125 (SP+15d; middle right panel), or 30 μM SB203580 (SB+15d; bottom left panel), NIH3T3 cells were stimulated by 10 μM 15d-PGJ₂ for 1 h. As a negative control, NIH3T3 cells were pretreated by DMSO and stimulated by DMSO for 1 h (DMSO; bottom right panel). Total RNAs were isolated at each time point. Quantitative real-time RT-PCR was performed using mPer2 (solid line), mBmal1 (dotted line), and 18S rRNA primers. Abscissa presents "hour", ordinate "mRNA amount", respectively. The maximum mRNA amount was set to 1. The relative levels of each mRNA were normalized to the corresponding 18S rRNA levels.

Figure 5

Schema of the entrainment mechanism by 15d-PGJ₂. 15d-PGJ₂ up-regulates transcription of Crys and Rorα (Fig. 5Bi). The translated RORα activates Bmal1 transcription, and translated BMAL1 then binds to CLOCK and their heterodimers activate Per/Cry and Rev-erb transcriptions via E/E' box (Fig. 5Bii). Translated PER/CRY and REV-ERB inhibit Per/Cry/Rev-erb and Bmal1 transcription, respectively (Fig. 5Biii). The inhibition of Rev-erb transcription reduces Bmal1 transcription (Fig. 5Biv). The reduced Per/Cry transcription and relatively increased RORα activity (by inhibition of REV-ERB) again up-regulate Bmal1 transcription (Fig. 5Biv -> 5Bii).