Transcriptional repression of frequency by the IEC-1-INO80 complex is required for normal Neurospora circadian clock function

Abstract

Rhythmic activation and repression of the frequency (frq) gene are essential for normal function of the Neurospora circadian clock. WHITE COLLAR (WC) complex, the positive element of the Neurospora circadian system, is responsible for stimulation of frq transcription. We report that a C2H2 finger domain-containing protein IEC-1 and its associated chromatin remodeling complex INO80 play important roles in normal Neurospora circadian clock function. In iec-1KO strains, circadian rhythms are abolished, and the frq transcript levels are increased compared to that of the wild-type strain. Similar results are observed in mutant strains of the INO80 subunits. Furthermore, ChIP data show that recruitment of the INO80 complex to the frq promoter is IEC-1-dependent. WC-mediated transcription of frq contributes to the rhythmic binding of the INO80 complex at the frq promoter. As demonstrated by ChIP analysis, the INO80 complex is required for the re-establishment of the dense chromatin environment at the frq promoter. In addition, WC-independent frq transcription is present in ino80 mutants. Altogether, our data indicate that the INO80 complex suppresses frq transcription by re-assembling the suppressive mechanisms at the frq promoter after transcription of frq.

Fig 1. IEC-1 is required for normal circadian clock function.

(A) Race tube assays of the wild-type and iec-1^KO strains. (B) Amino acid sequence alignment of the zf-C2H2 domains of IEC-1 from Neurospora crassa, Penicillium brasilianum, Aspergillus fumigates and Nectria haematococca. (C) Race tube assays of the wild-type strain, iec-1^KO strain, and iec-1^KO, qa-Myc-IEC-1 transformants in a race tube with or without QA. Growth media on the race tubes did not consist of glucose. (D) Luciferase reporter assay showing the frq promoter activity in the wt, frq-luc and iec-1^KO, frq-luc strains grown in DD for several days. Raw data were normalized to subtract the baseline calculated by the LumiCycle analysis software.

Fig 2. IEC-1 suppresses frq transcription and rhythmically binds to the frq promoter.

(A) Western blot analysis showing the circadian oscillation of FRQ proteins in the wild-type and iec-1^KO strains. The strains were grown in 2% glucose liquid media. The asterisk indicates a nonspecific cross-reacted protein band recognized by our FRQ antiserum. The Coomassie Brilliant Blue-stained membranes (mem) represent the total protein in each sample and were used as a loading control. (B) Northern blot analysis of frq transcription in the wild-type and iec-1^KO strains. rRNA was used as a loading control. The strains were grown in 2% glucose liquid media. (C) Immunodetection of IEC-1 protein in the wild-type strain and the iec-1^KO mutant using antiserum that specifically recognizes the IEC-1 protein in the wild-type strain. The arrow notes the specific IEC-1 protein band detected by our IEC-1 antibody. The strains were grown in 2% glucose liquid media. (D) ChIP analysis showing the recruitment of IEC-1 at different regions of the frq locus in the wild-type and iec-1^KO strains at DD18. The strains were grown in 2% glucose liquid media. C-box, clock box; PLRE, proximal light-regulated element; TSS, transcription start site; ORF, open reading frame; UTR, untranslated region. (E) ChIP analysis showing the enrichment of IEC-1 at the C-box of the frq promoter in the wild-type and iec-1^KO strains at the indicated time points. The strains were grown in 2% glucose liquid media. Significance was assessed by two-tailed t-test. *P<0.05, **P<0.01. Error bars show the mean ±S.D. (n = 3).

Fig 3. INO80 is rhythmically recruited at the C-box by IEC-1 and WCC-driven frq expression.

(A) Immunodetection of INO80 in the wild-type strain and the ino80^KO mutant using antiserum that specifically recognizes INO80 protein in the wild-type strain. The strains were grown in 2% glucose liquid media. The arrow indicates the INO80 specific band in the wild-type strains. (B) ChIP analysis showing the recruitment of INO80 at different regions of the frq locus in the wild-type and ino80^KO strains at DD18. The ino80^KO strain was used as a negative control in the ChIP assay. The strains were grown in 2% glucose liquid media. Significance was assessed by two-tailed t-test. *P<0.05, **P<0.01. Error bars show the mean ±S.D. (n = 3). (C) ChIP analysis showing the recruitment of INO80 at the C-box in the wild-type, ino80^KO and iec-1^KO strains at the indicated time points. The strains were grown in 2% glucose liquid media. Significance was assessed by two-tailed t-test. *P<0.05, **P<0.01. Error bars show the mean ±S.D. (n = 3). (D) ChIP analysis showing the enrichment of histone H3 at the C-box, PLRE or TSS of the frq promoter region in the wild-type, wc-2^KO (bd) and frq⁹ (bd) mutant strains. The strains were grown in 2% glucose liquid media. Significance was assessed by two-tailed t-test. *P<0.05, **P<0.01. Error bars show the mean ±S.D. (n = 3). (E) ChIP analysis showing the recruitment of INO80 at the C-box of the frq promoter in the wild-type, ino80^KO, wc-2^KO(bd), frq⁹(bd), and wc-2^KO frq⁹ (bd) strains at the indicated time points. The strains were grown in 2% glucose liquid media. Significance was assessed by two-tailed t-test. *P<0.05, **P<0.01. Error bars show the mean ±S.D. (n = 3). (F) Western blot analysis showing the INO80 protein levels in the wild-type, ino80^KO, wc-2^KO (bd), frq⁹ (bd) and wc-2^KO frq⁹ (bd) strains. The strains were grown in 2% glucose liquid media.

Fig 4. The INO80 complex is required for the suppression of WC-independent frq transcription.

(A) ChIP analysis showing WC-2 enrichment at the C-box in the wild-type, ino80^KO, iec-1^KO, and wc-2^KO (bd) strains. The strains were grown in 2% glucose liquid media. Significance was assessed by two-tailed t-test. *P<0.05, **P<0.01. Error bars show the mean ±S.D. (n = 3). (B) Western blot analysis showing the phosphorylation of WC-1 in the wild-type, ies-1^KO, ino80^KO and iec-1^KO strains. The numbers indicate the ratio of acrylamide/bisacrylamide used in the SDS-PAGE gel. The strains were grown in 2% glucose liquid media. (C) Western blot analysis showing the levels of WC-1 in the wild-type, ies-1^KO, ino80^KO and ies-1^KO strains. The strains were grown in 2% glucose liquid media. (D) Western blot analysis showing the phosphorylation of WC-2 in the wild-type, ies-1^KO, ino80^KO and ies-1^KO strains. The strains were grown in 2% glucose liquid media. (E) Western blot analysis showing the levels of WC-2 in the wild-type, ies-1^KO, ino80^KO and ies-1^KO strains. The strains were grown in 2% glucose liquid media. (F) Western blot analysis of FRQ or WC-1 in the wild-type, ies-1^KO, wc-1^RIP (bd), and ies-1^KO wc-1^RIP strains. The strains were grown in 2% glucose liquid media. (G) Northern blot analysis showing the levels of frq mRNA in the wild-type, ies-1^KO, wc-1^RIP (bd), and ies-1^KO wc-1^RIP strains. The strains were grown in 2% glucose liquid media.

Fig 5. The establishment of nucleosomal barriers at the frq promoter by the INO80 complex prevents RNA pol II initiation.

(A) ChIP analysis showing H3 density at the C-box, TSS or ORF middle regions of the frq locus in the wild-type, ies-1^KO, and ino80^KO strains. The strains were grown in 2% glucose liquid media. Significance was assessed by two-tailed t-test. *P<0.05, **P<0.01. Error bars show the mean ±S.D. (n = 3). (B) ChIP analysis showing the recruitment of SET-2 and enrichment of H3K36me3 at the ORF 3’ of frq in the wild-type and ino80^KO strains. The strains were grown in 2% glucose liquid media. Significance was assessed by two-tailed t-test. *P<0.05, **P<0.01. Error bars show the mean ±S.D. (n = 3).