Modulation of Circadian Gene Expression and Metabolic Compensation by the RCO-1 Corepressor of Neurospora crassa

Abstract

Neurospora crassa is a model organism for the study of circadian clocks, molecular machineries that confer ∼24-hr rhythms to different processes at the cellular and organismal levels. The FREQUENCY (FRQ) protein is a central component of the Neurospora core clock, a transcription/translation negative feedback loop that controls genome-wide rhythmic gene expression. A genetic screen aimed at determining new components involved in the latter process identified regulation of conidiation 1 (rco-1), the ortholog of the Saccharomyces cerevisiae Tup1 corepressor, as affecting period length. By employing bioluminescent transcriptional and translational fusion reporters, we evaluated frq and FRQ expression levels in the rco-1 mutant background observing that, in contrast to prior reports, frq and FRQ expression are robustly rhythmic in the absence of RCO-1, although both amplitude and period length of the core clock are affected. Moreover, we detected a defect in metabolic compensation, such that high-glucose concentrations in the medium result in a significant decrease in period when RCO-1 is absent. Proteins physically interacting with RCO-1 were identified through co-immunoprecipitation and mass spectrometry; these include several components involved in chromatin remodeling and transcription, some of which, when absent, lead to a slight change in period. In the aggregate, these results indicate a dual role for RCO-1: although it is not essential for core-clock function, it regulates proper period and amplitude of core-clock dynamics and is also required for the rhythmic regulation of several clock-controlled genes.

Keywords: Neurospora crassa; circadian clocks; core-clock mechanism; corepressor; frequency.

Figure 1

The absence of SUB-1 or RCO-1 affects the rhythmic expression of a con-10^luc circadian output reporter. A luciferase reporter (con-10^luc) was used as a proxy for circadian output regulation, and the impact of eliminating selected TFs was evaluated. The absence of SUB-1 (∆sub-1) leads to decreased amplitude and a phase change of CON-10^LUC expression, while in ∆rco-1 expression of CON-10^LUC appears high and arrhythmic. The upper row depicts graphs containing raw luminescence data (A), while in the graphs in the lower row LUC levels were normalized using Spectrum resampling (B) (Costa et al. 2013).

Figure 2

The absence of RCO-1 abrogates rhythms in conidiation and leads to a functional circadian oscillator with altered period and amplitude. (A) Race tube assays of WT (rco-1⁺, ras-1^bd) and ∆rco-1 (Δrco-1, ras-1^bd) strains in DD confirms the absence of overt circadian rhythms in conidiation in ∆rco-1. (B) WT (rco-1⁺, ras-1^bd) and ∆rco-1, ras-1^bd strains carrying the transcriptional reporter frq_c-box-luc were analyzed, confirming the presence of a functional core circadian oscillator in ∆rco-1, which exhibits a long period and decreased amplitude. (C) The absence of RCO-1 was analyzed in combination with the frq⁷ allele, observing a further increase in period consistent with a FWO-based rhythm. The luminescence data were analyzed using Spectrum resampling (Costa et al. 2013). Period ± SEM, WT: 20.89 ± 0.5; Δrco-1: 25.88 ± 0.36; frq⁷ WT: 28.66 ± 0.16; frq⁷∆rco-1: 33.74 ± 0.26. (D) Real-time qPCR analyses show rhythms in frq mRNA levels in time courses conducted in solid media conditions. Three biological replicates were measured and frq expression was normalized using two reference genes: tbp (NCU04770) and suc (NCU08336). For plotting, DD4 of WT strain was set to 1, and other time points were normalized accordingly. To evaluate frq rhythmicity, data were analyzed with ARSER (Yang and Su 2010), revealing that its expression was rhythmic (P < 0.05) in both strains. (E) Representative Western blots of samples derived from time courses as described in D, reveal oscillating levels of FRQ in WT and ∆rco-1 strains. A densitometric analysis (right) was conducted for three biological replicates of each strain. Standard error is shown. For plotting, the LL time point was set to 1 and the other points were normalized accordingly.

Figure 3

rco-1 transcription and translation are rhythmic although its mRNA and protein levels do not appear to oscillate. (A) Reporter construct containing the rco-1 promoter region (3000 bp) fused to luciferase was integrated at the his-3 locus and the luminescence data were measured as previously described. (B) A translational fusion between RCO-1 and LUC was engineered at the endogenous rco-1 locus by homologous recombination, and the resulting strain was analyzed under the same experimental conditions as A. (C) rco-1 and frq mRNA levels were measured over a time course experiment conducted under solid media conditions, utilizing tbp (NCU04770) and suc (NCU08336) as reference genes. (D) Quantification of FRQ and RCO-1 protein levels from time courses as described in C. As V5-tagged strains of FRQ or RCO-1 were employed to facilitate detection, Western blots were performed with α-V5 antibody, followed by densitometric analyses (levels at DD4 was set as 1 for normalization), confirming rhythmic expression of FRQ and constant nonrhythmic levels of RCO-1. The figure is representative of three biological replicates. (E) Cycloheximide (CHX) was used to determine RCO-1^V5 stability. Samples were grown in liquid culture media (LCM) 0.03% glucose for 48 hr after which CHX 10 μg/ml was added. Cultures were transferred to DD and harvested every 2 hr. Western blot with αV5 (left) and the corresponding densitometric analysis (right) reveal that RCO-1 exhibits high stability. The data correspond to three biological replicates. FRQ^V5 was used as a control of an unstable protein.

Figure 4

RCO-1 regulates the expression of several genes involved in a variety of processes. (A) The expression of several putative RCO-1 target genes was evaluated by RT-qPCR in WT and ∆rco-1 strains under LL conditions. The plots indicate the results for a set of genes from each biological process described in Table 1 and also for some genes previously described as targets of RCO-1. Almost all analyzed genes present higher expression levels in the ∆rco-1 mutant. The data corresponded to three biological replicates. A Mann–Whitney test was performed; significance is indicated with an asterisk (* P < 0.05). (B) Heat map representation of gene expression (measured through RT-qPCR) in absence of RCO-1 in three independent biological replicates (Nb); only 58% of the evaluated genes changed their expression levels in absence of the corepressor.

Figure 5

RCO-1 is involved in proper circadian control of several clock-controlled genes. Transcriptional luciferase reporters were used to evaluate the circadian expression of RCO-1 target genes in WT and ∆rco-1 strains, revealing the loss of proper circadian control in ∆rco-1. Bioluminescence was measured as previously described.

Figure 6

RCO-1 is required for proper metabolic compensation of the clock. WT and Δrco-1 strains carrying the transcriptional reporter frq_c-box-luc were inoculated in 96-well plates containing LN-CDD QA 0.01 M media with 0% glucose (low glucose, LG) or 0.5% glucose (high glucose, HG), and kept under LD 12:12 conditions for 72 hr prior to CCD analysis. Bioluminescence was measured as previously described. The absence of RCO-1 leads to reduced period in high-glucose conditions, indicating a loss of metabolic compensation of the clock. Period ± SEM. WT LG: 21.08 ± 0.23, HG: 20.94 ± 0.04; Δrco-1 LG: 24.52 ± 0.07, HG: 22.41 ± 0.095.