The role of the Arabidopsis morning loop components CCA1, LHY, PRR7, and PRR9 in temperature compensation

Abstract

A defining, yet poorly understood characteristic of the circadian clock is that it is buffered against changes in temperature such that the period length is relatively constant across a range of physiologically relevant temperatures. We describe here the role of PSEUDO RESPONSE REGULATOR7 (PRR7) and PRR9 in temperature compensation. The Arabidopsis thaliana circadian oscillator comprises a series of interlocking feedback loops, and PRR7 and PRR9 function in the morning loop. The prr7 prr9 double mutant displays a unique phenotype that has not been observed before in other Arabidopsis clock mutants. In the prr7 prr9 mutant, the effects of temperature are overcompensated, apparently due to hyperactivation of the transcription factors CIRCADIAN CLOCK ASSOCIATED1 (CCA1) and LATE ELONGATED HYPOCOTYL (LHY). Inactivation of CCA1 and LHY fully suppresses the overcompensation defects of prr7 prr9 mutants and rescues their long period phenotype. Overcompensation in prr7 prr9 mutants does not rely on FLOWERING LOCUS C, a previously identified gene required for temperature compensation. Together, our results reveal a role of PRR7 and PRR9 in regulating CCA1 and LHY activities in response to ambient temperature.

Figure 1.

Conditional Loss of Entrainment of prr7 prr9 by Thermocycles. (A) Mean circadian traces for CCA1pro:LUC and TOC1pro:LUC activity in Col-2 and prr7 prr9 seedlings in constant light at 22°C following entrainment in constant light and thermocycles consisting of 12 h at 12°C, followed by 12 h at 22°C. Phase values are shown in the right-side panel. RAE, Relative amplitude error; RAE values close to zero are indicative of strong rhythms, and RAE = 1 defines the limit of statistically significant rhythmicity. (B) Mean circadian traces for CCA1pro:LUC and TOC1pro:LUC activity in Col-2 and prr7 prr9 seedlings in constant light at 22°C following entrainment in constant light and thermocycles consisting of 12 h at 22°C, followed by 12 h at 28°C. Phase values are shown in the right-side panel. (C) Mean circadian traces for CCA1pro:LUC activity in Col-2 and prr7 prr9 seedlings in constant darkness at 22°C following entrainment in constant darkness and thermocycles consisting of 12 h at 12°C, followed by 12 h at 22°C. Phase values are shown in the right-side panel. All data (luciferase activity and circadian periods) are shown as mean ± se (s.e.m.; n = 12 to 24).

Figure 2.

The prr7 prr9 Double Mutant Overcompensates. Seedlings carrying a CCA1pro:LUC or TOC1pro:LUC circadian reporter (Salomé and McClung, 2005) were entrained to light-dark cycles for 7 to 8 d at 23°C. Luciferase activity was then recorded from day 10 at 12, 16, 22, or 30°C as described in Methods. All data (luciferase activity and circadian periods) are shown as mean ± se (s.e.m.; n = 12 to 24). (A) Mean circadian traces for CCA1pro:LUC activity in Col-2 and prr7 prr9 at 12, 22, and 30°C. (B) Mean circadian period of the CCA1pro:LUC reporter in Col-2 and prr7 prr9 as a function of temperature. (C) Temperature compensation behavior of prr5 prr7 and prr7 prr9 double mutants. (D) Temperature compensation behavior of the gi-1 (weak) and gi-201 (strong) mutant alleles. (E) Temperature compensation behavior of the long period mutants arr3 arr4 and ztl-4. (F) Temperature compensation behavior of the core clock component mutants lhy-20 and toc1-101. (G) Temperature compensation behavior of the flc-3 mutant in the Col-0 background.

Figure 3.

Targeted Knockdowns of Arabidopsis Clock Genes by amiRNAs. amiRNAs (Schwab et al., 2006) were designed to target each of the clock genes CCA1, LHY, TOC1, and ZTL. Randomly chosen T2 transgenic lines were characterized for the period of the circadian reporter CCA1pro:LUC, which is present in the same T-DNA as the 35S:amiRNA cassette (see Methods for details). All data (luciferase activity and circadian periods) are shown as mean ± se (s.e.m.; n = 12). (A) Mean period length for ZTL amiRNA lines. The period length of TOC1pro:LUC in the T-DNA insertion allele ztl-4 (Michael et al., 2003) is shown as reference. (B) Mean period length for TOC1 amiRNA lines and the corresponding loss-of-function phenotype in the toc1-101 mutant (Kikis et al., 2005). (C) Mean period length for CCA1 amiRNA lines. Note that no true loss-of-function allele exists for CCA1 in the Col-0 background. Ws, Wassilewskija. (D) Mean period length for LHY amiRNA lines and the corresponding loss of function phenotype in the lhy-20 T-DNA insertion allele (Michael et al., 2003). (E) Mean period length for CCA1-LHY tandem amiRNA lines. Ler, Landsberg erecta.

Figure 4.

Rescue of Overcompensation by Targeted Knockdowns of the Transcription Factors CCA1 and LHY. Constructs expressing amiR-CCA1, amiR-LHY, or amiR-CCA1-LHY (shown in Figure 3) were introduced in prr7 prr9 by Agrobacterium tumefaciens–mediated transformation. The circadian phenotype of multiple T2 lines was assessed with the CCA1pro:LUC reporter included on the T-DNA. All data (luciferase activity and circadian periods) are shown as mean ± se (s.e.m.; n = 12). (A) Progressive rescue of the long period phenotype of prr7 prr9 by amiR-CCA1-LHY. Seedlings were grown at 22°C. (B) Mean circadian traces of strong amiR-CCA1/LHY prr7 prr9 lines, when assayed at 30°C. (C) Mean period lengths of the genotypes shown in (B).

Figure 5.

Conservation of the Role of PRR7 and PRR9 in Temperature Compensation across Arabidopsis Accessions. A construct driving the expression of a tandem amiRNA targeting PRR7 and PRR9 was introduced in Arabidopsis accessions. At least three independent T2 transgenic lines were assayed for each condition and genetic background. All data (luciferase activity and circadian periods) are shown as mean ± se (s.e.m.; n = 12). (A) Knockdown of PRR7 and PRR9 leads to a long circadian period in several accessions. (B) Knockdown of PRR7 and PRR9 results in overcompensation in all accessions tested. Circadian parameters for the lines shown in (A) were scored at 22 and 30°C.

Figure 6.

Loss of PRR7 and PRR9 Results in the Accumulation of Morning Loop Genes but Not Evening Loop Components. (A) Normal expression of the clock genes CCA1, LHY, and TOC1 is restored in the prr7 prr9 mutant at 12°C. (B) prr7 prr9 accumulates more CCA1 and LHY mRNA at higher temperatures than Col-0. Expression values were integrated over the whole time course. (C) HSP70 expression is under the control of the clock. LHY expression is shown as control for a strongly rhythmic gene. Data extracted from Gould et al. (2006) and shown as mean ± sd (n = 3). (D) Induction of HSP70 expression by ambient temperature is maintained in prr7 prr9. Left panel: expression values from publicly available microarray data (Gould et al., 2006). Middle and right panels: HSP70 expression determined by quantitative PCR in Col-0 and prr7 prr9 at Zeitgeber Time 72 (ZT72). Expression values are shown for one biological replicate as mean ± sd (n = 2).