A cut above: the critical roles of DICER-LIKE genes in Marchantia development

Abstract

Small RNAs (sRNAs), such as microRNAs (miRNAs) and small-interfering RNAs (siRNAs), play important roles in plant development, abiotic stress responses and immunity. These RNAs are generated by DICER-LIKE (DCL) ribonucleases, which cleave long RNA precursor molecules into 21–24 nucleotide double-stranded RNA fragments. In the case of miRNAs and siRNAs, these fragments are subsequently incorporated into RNA-induced silencing complexes (RISCs) and convey target specificity, resulting in transcriptional gene silencing or post-transcriptional repression through mRNA degradation or translational inhibition.

A single DCL gene is present in green algae, whereas the gene family is diversified in streptophytes (Bélanger et al., 2023). DCL1, present in streptophyte algae and all land plant lineages, produces mature miRNAs from pri-miRNA precursors. DCL2, DCL3 and DCL4, responsible for generating various types of siRNAs, emerged in bryophytes (DCL3/4) and vascular plants (DCL2 in addition to DCL3/DCL4), respectively. Additionally, DCL5 represents a monocot-specific DCL that produces reproductive-phased siRNAs in anthers (Bélanger et al., 2023; Liu et al., 2009). While DCLs have been extensively studied in the model dicot Arabidopsis thaliana (Arabidopsis), their roles in early diverging plant lineages remain largely unexplored.

Wolfgang Frank has a particular interest in bryophyte sRNAs; as some of the first plants to colonise terrestrial habitats, bryophytes had to develop specific molecular mechanisms to thrive on land. The diversification of sRNAs likely contributed to these adaptive processes. PhD student Erika Csicsely and post-doctoral researcher Oguz Top joined Frank's research group to investigate the molecular adaptations underlying the process of terrestrialisation. Top conducts comparative studies on the moss Physcomitrium patens and the liverwort Marchantia polymorpha (Marchantia) to uncover regulatory mechanisms that might have facilitated land plant adaptation. For the highlighted publication, he collaborated with Csicsely to investigate the role of DCL genes in Marchantia.

Like other bryophytes, Marchantia possesses four DCL genes (Bélanger et al., 2023), and Csicsely et al. confirmed that these belong to the DCL1, DCL3 and DCL4 subclades, with two DCL1 sequences (MpDCL1a, MpDCL1b). The active site of DCL proteins is formed by a PAZ (Piwi/Argonaute/Zwille) domain, two ribonuclease III (RNase III) domains and a double-stranded RNA-binding site, and many DCLs harbour additional domains such as Helicase C, Dicer-dimer and restriction enzyme subunit III domains (Liu et al., 2009). Most of these domains are present in MpDCL1a, MpDCL3 and MpDCL4, but MpDCL1b only contains a PAZ and two RNase III domains. It thus appears that MpDCL1a is the canonical homologue of seed plant DCL1, while MpDCL1b constitutes a novel subclade of DCL1 proteins. Reciprocal BLAST searches identified MpDCL1b-like sequences in the genomes of other bryophytes and ferns, while they are conspicuously absent in the genomes of streptophyte algae and seed plants.

To understand the function of Marchantia's DCL genes, Csicsely et al. generated loss-of-function mutants using CRISPR/Cas9 gene editing. They were able to obtain large deletions in MpDCL1b, MpDCL3 and MpDLC4. However, only small insertions or deletions were obtained for MpDCL1a, none of which resulted in a frameshift or early stop codon. This suggests that a full knock-out of MpDCL1a may be lethal, as observed for dcl1 mutants in Arabidopsis (Henderson et al., 2006).

Despite not representing a full knock-out, the Mpdcl1a mutant displayed the most striking phenotype among all gene-edited lines, showing severely reduced growth and forming a callus-like structure that lacked the characteristic dichotomous thallus branching of the wild type (Figure 1a). Additionally, the mutant produced exposed gemmae (buds for asexual propagation) without surrounding gemma cup walls. Mpdcl1b and Mpdcl4 mutant thalli appeared similar to the wild type, while Mpdcl3 displayed increased thallus branching (Figure 1a). Furthermore, Mpdcl1a, Mpdcl3 and Mpdcl4 mutants failed to produce gametangiophores, which are essential for sexual reproduction; even in Mpdcl1b, these organs were much smaller (Figure 1b). These findings suggest that MpDCL1a is most critical for vegetative growth, while all the MpDCL genes are essential for reproductive development.

DCL genes and sRNAs have also been linked to phytohormone signalling and abiotic stress responses (Li et al., 2020; Sunkar et al., 2007). At high salt concentrations, mutations in MpDCL1a and MpDCL1b increased survival compared to the wild type, suggesting that these two genes regulate Marchantia's sensitivity to salinity. Several Mpdcl mutants also displayed altered hormone sensitivity: Mpdcl3, in particular, was less sensitive to the synthetic auxin NAA, lacking the enhanced rhizoid formation observed in wild-type plants. Mpdcl1a, Mpdcl3 and Mpdcl4 were hypersensitive to abscisic acid (ABA), showing a further reduction in thallus size.

To understand the molecular basis for the observed changes in development and stress responses, Csicsely et al. investigated sRNA and mRNA production in all the Mpdcl mutants. Mpdcl1a had the highest number of differentially expressed sRNAs, followed by Mpdcl3, and these changes inversely correlated with changes in their predicted mRNA targets. In agreement with their wild-type-like appearance, Mpdcl1b and Mpdcl4 showed fewer differentially expressed sRNA:mRNA modules. Mpdcl1b, however, failed to show salt-induced changes in miRNA and mRNA accumulation seen in wild type. This observation agrees with the mutant's altered salt stress response and suggests a specific role for MpDCL1b under abiotic stress.

Taken together, Csicsely et al. provide a comprehensive molecular and phenotypic analysis of dcl mutants in Marchantia, highlighting the critical roles of MpDCL genes in development and the salt stress response. Notably, they identified a previously unrecognised DCL subclade (DCL1b) specific to ferns and bryophytes. The authors propose that DCL1b may have helped early land plants to cope with terrestrial stresses such as salinity or desiccation. As seed plants evolved, this function may have been replaced by the emergence of DCL2 and further diversification of DCL3 and DCL4, ultimately leading to the loss of DCL1b. It will be interesting to see whether studies of DCL1b homologues in other fern and bryophyte species can corroborate this hypothesis.