LARK activates posttranscriptional expression of an essential mammalian clock protein, PERIOD1

Abstract

The mammalian molecular clock is composed of feedback loops to keep circadian 24-h rhythms. Although much focus has been on transcriptional regulation, it is clear that posttranscriptional controls also play important roles in molecular circadian clocks. In this study, we found that mouse LARK (mLARK), an RNA binding protein, activates the posttranscriptional expression of the mouse Period1 (mPer1) mRNA. A strong circadian cycling of the mLARK protein is observed in the suprachiasmatic nuclei with a phase similar to that of mPER1, although the level of the Lark transcripts are not rhythmic. We demonstrate that LARK causes increased mPER1 protein levels, most likely through translational regulation and that the LARK1 protein binds directly to a cis element in the 3' UTR of the mPer1 mRNA. Alterations of mLark expression in cycling cells caused significant changes in circadian period, with mLark knockdown by siRNA resulting in a shorter circadian period, and the overexpression of mLARK1 resulting in a lengthened period. These data indicate that mLARKs are novel posttranscriptional regulators of mammalian circadian clocks.

Fig. 1.

Expression profiles of mLarks in mouse SCN. (A) (Left) Representative in situ hybridization results for mLark1 and mLark2 mRNAs in mouse SCN harvested at different times of day from mice housed in either LD or DD conditions are shown. (Right) The graph shows mean quantification of mLark1 (□, LD; ■, DD) and mLark2 (□, LD; ●, DD) mRNAs. (B) (Left) Representative in situ hybridization results for mLark1 and mLark2 mRNAs in SCN harvested from mice after 600 lux of light exposure for 30 min are shown. (Right) The graph shows mean quantitation of mLark1 (□) and mLark2 (■) mRNAs. (C) Representative immunohistochemical analysis of mLARKs in mouse SCN (Left) in DD and quantitation of mLARK-IR cells (Right) are shown. One-way ANOVA revealed a highly significant (P < 0.0001) time effect. (D) Western blot analysis of mLARK (Left) in extracts of microdissected mouse SCNs harvested in DD and quantification of mLARK levels (Right) are shown. One-way ANOVA revealed a highly significant effect of time (P < 0.001). All data are representative of at least three independent experiments.

Fig. 2.

Posttranscriptional regulation of mPer1 chimeric reporter genes by mLARKs. (A) Structures of chimeric luciferase genes are shown. Both constructs contain a 6.8-kb mPer1 promoter driving a luciferase reporter gene followed by either the simian virus 40 poly(A) signal (pPLS) or the mPer1 3′ UTR (pPL3). (B) Relative luciferase activities of pPLS and pPL3 either with or without the presence of mLARK expression plasmids (165 ng) with the transcriptional activators CLOCK and BMAL1 are shown. (C) mLARK1 increases reporter gene expression in a dose-dependent manner. The experiment was done as in B except with varying levels of mLARK1 plasmid as indicated. (D) Relative luciferase mRNA levels are not altered by mLARK1. The experiment was done as in B, except total RNA was isolated and the level of luciferase mRNA was measured by real-time RT-PCR. Data consist of two or three independent experiments.

Fig. 3.

Posttranscriptional regulation of endogenous mPER1 by mLARK1. Measurement of endogenous mPER1, TUBULIN, and transfected mLARK1 by Western blot (Left) and measurement of endogenous mPer1 and β-actin by Northern blot and transfected mLark1 and β-actin by RT-PCR (Right) from NIH 3T3 cells with or without transfected mLARK1 are shown.

Fig. 4.

Direct interaction between mLark1 and mPer1 3′ UTR. (A) RMSA of His-mLARK1 and RNAs derived from mPer1 3′ UTR is shown. The numbers correspond to RNA oligonucleotides that span the entire mPer1 3′ UTR as diagrammed at the top (the sequences are listed in SI Table 2). The position of the previously identified ARE and LOX-DICE elements are marked (18). Arrowhead denotes the LARK-specific retarded band. (B) Dose–response relationship between His-mLARK1 and RNA 19 is shown. Shown is an RMSA as in A, but using only RNA 19 as probe and with increasing doses of His-mLARK1 as indicated. (C) Competitive analyses verify specificity of interaction between mLARK1 and RNA. RMSAs were done as in B but with the addition of increasing amounts of antibody or cold RNA competitors as indicated. (D) RMSA using normal and mutated versions of RNA 19 is shown. Underlined nucleotides are mutated nucleotides. The arrowhead indicates the RNA bands retarded by mLARK1, and the arrow indicates a nonspecific band. (E) LARK1 only activates reporter gene expression when the LARK binding site is intact. Reporter analyses were done as in Fig. 2, except an additional reporter construct was included in which the mPer1 3′ UTR contains the mutations shown in mut4 (D) [pPL3(m4)]. (F) The predicted RNA secondary structure is shown. Bold type indicates nucleotides corresponding to RNA fragment 19. (G) The direct interaction between LARK1 protein and Per1 mRNA is shown. Extracts from NIH 3T3 cells with either overexpressed mLARK1 or control plasmid were immunoprecipitated with either anti-LARK or control antibody. Immunoprecipitates were analyzed by Western blot using LARK antibody (Upper) or RT-PCR for mPer1 (Lower). Asterisk indicates the IgG heavy chain from the immunoprecipitating antibody.

Fig. 5.

mLARK1 affects mPER1 translation. (A) Schematic representation of bicistronic vectors. The promoter is minimal promoter containing five times ecdysone/glucocorticoid (E/GRE) responsive element (18). See Results for further details. (B) Luciferase assay of bicistronic vector. The y axis represents the ratio of R-luc/F-luc activity. Data consist of three independent experiments.

Fig. 6.

Circadian rhythms of bioluminescence in NIH-PL cells are affected by LARK levels. Shown are representative recordings of NIH-PL cells over 4 days in culture, measuring bioluminescence as an output of the mPer1::luc reporter gene. (A and B) Oscillations are not observed in these cells when treated with DMSO (A) but are induced after synchronization with dexamethasone (Dex) (B). (C–H) Representative recordings of NIH-PL cells as in B, but with the addition of siRNAs against EGFP (C), Lark-B (D), Bmal11 (F), Cry1 (G), and Cry2 (H) or with overexpression of mLARK1 (E). (I) The effect of siRNAs on the expression of mLARKs was confirmed by immunoblotting. Extracts of cells containing siRNAs against EGFP, Lark-A, and Lark-B were analyzed on Western blots using antibodies to LARK and TUBULIN.