Does the core circadian clock in the moss Physcomitrella patens (Bryophyta) comprise a single loop?

Abstract

Background: The endogenous circadian clock allows the organism to synchronize processes both to daily and seasonal changes. In plants, many metabolic processes such as photosynthesis, as well as photoperiodic responses, are under the control of a circadian clock. Comparative studies with the moss Physcomitrella patens provide the opportunity to study many aspects of land plant evolution. Here we present a comparative overview of clock-associated components and the circadian network in the moss P. patens.

Results: The moss P. patens has a set of conserved circadian core components that share genetic relationship and gene expression patterns with clock genes of vascular plants. These genes include Myb-like transcription factors PpCCA1a and PpCCA1b, pseudo-response regulators PpPRR1-4, and regulatory elements PpELF3, PpLUX and possibly PpELF4. However, the moss lacks homologs of AtTOC1, AtGI and the AtZTL-family of genes, which can be found in all vascular plants studied here. These three genes constitute essential components of two of the three integrated feed-back loops in the current model of the Arabidopsis circadian clock mechanism. Consequently, our results suggest instead a single loop circadian clock in the moss. Possibly as a result of this, temperature compensation of core clock gene expression appears to be decreased in P. patens.

Conclusions: This study is the first comparative overview of the circadian clock mechanism in a basal land plant, the moss P. patens. Our results indicate that the moss clock mechanism may represent an ancestral state in contrast to the more complex and partly duplicated structure of subsequent land plants. These findings may provide insights into the understanding of the evolution of circadian network topology.

Figure 1

Characterization of putative CCA1/LHY orthologs in P. patens. Characterization of putative CCA1/LHY orthologs in P. patens. (A) Domain structure of CCA1 and LHY in Arabidopsis with the single Myb-domain. (B) An unrooted maximum-likelihood phylogenetic tree constructed from an alignment of the Myb-domain of the SHAQKYF subtype in protein sequences found in A. thaliana, O. sativa, S. moellendorffii, C. reinhardtii and P. patens. (C) Comparative analysis of the oscillating profiles of AtCCA1, PpCCA1a and PpCCA1b in A. thaliana and P. patens. P. patens cultures were sampled every fourth hour for one day in LD (16 h light/8 h dark) and for two days in DD (constant dark). Comparative data for AtCCA1 was downloaded from the DIURNAL database (see Methods). Quantitative RT-PCR expression data was normalized (CT_target - CT_reference) and the minimum transcription level in each time series was arbitrarily set to 0. Grey and white bars surrounded by dashed lines indicate subjective night and day in DD. Plots including error bars indicating standard deviation from duplicate runs of all genes can be found in Additional file 6.

Figure 2

Characterization of putative PRR orthologs in P. patens. Characterization of putative PRR orthologs in P. patens. (A) Domain structure of TOC1 in Arabidopsis with the pseudo-receiver domain at the N-terminal and CCT motif at the C-terminal. (B) An unrooted maximum-likelihood phylogenetic tree was constructed from the amino acid sequences of the pseudo-receiver and the CCT motif in A. thaliana, O. sativa, S. moellendorffii, C. reinhardtii, O. tauri and P. patens. (C) Comparative analysis of the expression profiles of TOC1 and the four PRRs in Arabidopsis (left panel), and the expression profiles of PpPRR1-4 in P. patens (right panel). See legend to figure 1 for further details.

Figure 3

Identification and characterization of putative LUX orthologs in P. patens. Identification and characterization of putative LUX orthologs in P. patens. (A) Domain structure of LUX in Arabidopsis with the Myb DNA binding motif toward the C-terminal. (B) An unrooted maximum-likelihood phylogenetic tree was constructed from an alignment of the Myb domain in A. thaliana, O. sativa, S. moellendorffii, C. reinhardtii and P. patens. (C) Comparative analysis of the expression profile of LUX in Arabidopsis and the putative ortholog Phypa-47310 in P. patens. See legend to figure 1 for further details.

Figure 4

Identification and characterization of a putative ELF4 ortholog in P. patens. Identification and characterization of a putative ELF4 ortholog in P. patens. (A) Domain structure of ELF4 in Arabidopsis with the DUF1313 domain covering most of the amino acid sequence. (B) An unrooted maximum-likelihood phylogenetic tree was constructed from an alignment of the DUF1313 region in A. thaliana, O. sativa, S. moellendorffii, C. reinhardtii and P. patens. AtEFL1 is At2g29950, AtEFL2 is At1g72630, AtEFL3 is At2g06255 and AtEFL4 is At1g17455. (C) Comparative analysis of the expression profile of ELF4 in Arabidopsis and the putative ortholog Phypa-49622 in P. patens. See legend to figure 1 for further details.

Figure 5

Identification and characterization of putative ELF3 orthologs in P. patens. Identification and characterization of putative ELF3 orthologs in P. patens. (A) An unrooted maximum-likelihood phylogenetic tree was constructed from an alignment of conserved regions of ELF3-like amino acid sequences in A. thaliana, O. sativa, S. moellendorffii and P. patens. (B) Comparative analysis of the expression profile of ELF3 in A. thaliana and the two putative orthologs, Phypa-66647 and Phypa-165364, in P. patens. Phypa-233510 did not display a rhythmic expression pattern (data not shown). See legend to figure 1 for further details.

Figure 6

Protein architecture of the ZTL-family of genes. Protein architecture of ZTL family-like genes in A. thaliana, O. sativa, S. moellendorffii and P. patens.

Figure 7

Time series expression pattern of putative clock genes in different temperatures. Normalized expression levels (CT_target - CT_reference) for putative circadian clock genes in P. patens measured under three different temperatures in DD (constant dark). The light regime was identical to previous experiments (see legend to figure 1 for further details). Temperature was changed one day (ZT-24) before the onset of DD.

Figure 8

Core clock network comparison. Comparison of the core clock network in Arabidopsis and the version of the core clock in P. patens proposed in this study. (A) Locke's three-loop model where the inferred but as yet unidentified component "X" linking the feedback from TOC1 to LHY/CCA1 has been omitted for clarity. LUX has been placed in the vicinity of the morning-phased loop since interactions with LHY/CCA1 has been experimentally verified (see text for further details). Similarly, the ZTL family of components has been placed by the evening-phased loop. ELF4 has been positioned below the central loop since recent results suggest that it interacts with components of both the morning- and evening-phased loop (see text for further details). The positioning of ELF3 is relative since the exact means by which it interacts with the input light pathway, and feed back rhythmic information from the core clock, is not yet fully understood. (B) A suggested version of the P. patens circadian clock network. The single loop corresponds to the morning-phased loop of the Arabidopsis clock mechanism in seedlings and the "simplified slave version" of the Arabidopsis root clock. PpELF3L1-2 refer to Phypa_66647 and Phypa_165364. PpELF4L and PpLUXL refer to Phypa_49622 and Phypa_47310.