Early evolution of the land plant circadian clock

Summary While angiosperm clocks can be described as an intricate network of interlocked transcriptional feedback loops, clocks of green algae have been modelled as a loop of only two genes. To investigate the transition from a simple clock in algae to a complex one in angiosperms, we performed an inventory of circadian clock genes in bryophytes and charophytes. Additionally, we performed functional characterization of putative core clock genes in the liverwort Marchantia polymorpha and the hornwort Anthoceros agrestis. Phylogenetic construction was combined with studies of spatiotemporal expression patterns and analysis of M. polymorpha clock gene mutants. Homologues to core clock genes identified in Arabidopsis were found not only in bryophytes but also in charophytes, albeit in fewer copies. Circadian rhythms were detected for most identified genes in M. polymorpha and A. agrestis, and mutant analysis supports a role for putative clock genes in M. polymorpha. Our data are in line with a recent hypothesis that adaptation to terrestrial life occurred earlier than previously expected in the evolutionary history of charophyte algae. Both gene duplication and acquisition of new genes was important in the evolution of the plant circadian clock, but gene loss has also contributed to shaping the clock of bryophytes.


Fig. S1
Alignments used for phylogenetic construction          Table S1 Gene names, family/sub-clade, gene ID or accession number Table S2 Oligonucleotides used in this study Methods S1 Supplemental materials and methods describing sequence retrieval, sequence analysis and phylogenetic reconstruction

Fig. S1
Alignments used for phylogenetic construction. Amino acid sequences were aligned using the M-Coffee algorithm in T-Coffee (Notredame et al., 2000;Wallace et al., 2006).

Inventory of putative bryophyte and charophyte circadian clock genes
Homologs to Arabidopsis thaliana circadian clock genes were first identified in the liverwort Marchantia polymorpha, in the hornwort Anthoceros agrestis and in the charophyte Klebsormidium flaccidum. Initial gene family classification was supported by BLAST e-values, reciprocal BLAST searches to the Arabidopsis genome, characterization of conserved protein domains and by BLAST searches against the oneKP database (www.onekp.com).

Sequence retrieval
M. polymorpha genes were identified using Arabidopsis amino acid sequences as queries in tBLASTn searches in the publically available M. polymorpha genome v. 3.1 (Phytozome 11; https://phytozome.jgi.doe.gov). A. agrestis genes were likewise obtained searching the oneKP database (www.onekp.com). Algal gene sequences were retrieved from the publically available K. flaccidum genome V1.0 database (http://www.plantmorphogenesis.bio.titech.ac.jp/) and from transcriptome NGS data for the streptophyte algae Cylindrocystis cushleckae (Cc; Acc. No. ERR364373) and Coleochaete irregularis. (Ci; Acc. No. ERR364367), which were obtained from the NCBI short read archive (SRA). Sequences were assembled using Trinity r20131110 (Haas et al., 2013), tBLASTn searches were performed and gene prediction was done using the webbased FGENESH+ program. Homologs were also retrieved from the following species: A. thaliana, Oryza sativa, Selaginella moellendorffii, Picea abies and P. patens. Arabidopsis sequences were downloaded from TAIR (www.arabidopsis.org). Genes of other species were identified by BLASTp searches with Arabidopsis queries against databases at JGI (http://www.jgi.doe.gov/), NCBI (http://ncbi.nlm.nih.gov) and plantGDB (http://www.plantgdb.org). To investigate the distribution of identified clock homologs among charophyte algae, hornworts, liverworts and mosses, BLAST searches were performed against the oneKP database (www.onekp.com). All previously identified clock genes as well as genes identified in this study are listed with accession numbers in Supporting Information Table S1.
Branches resulting in conflicting topologies from the two methods were collapsed. Trees were rooted with charophyte sequences, except for the CCA1/LHY/RVE and PRR families that were midpoint rooted.