Identification of downstream genes of MpRRB-mediated cytokinin signaling
A ribonucleic acid-sequence (RNA-seq) analysis was performed to search for genes controlled downstream of MpRRB-mediated cytokinin signaling. Data from three biological replicates of the MpCKX2-overexpressing line and the Mprrb knockout line [15] were used to identify genes with more than a 1.5-fold reduction in expression levels compared to the wild-type Tak-1 and Tak-2, respectively. MpRRA mRNA levels in the two lines were 0.76- and 0.67-fold of those in the wild-type, respectively, indicating the reliability of the obtained data. It is notable that such a mild reduction in MpRRA transcript levels was also observed in quantitative RT-PCR (qRT-PCR) [15]. The genes, whose expression was reduced in both MpCKX2-overexpressing and Mprrb knockout lines, were extracted as those upregulated by MpRRB-mediated cytokinin signaling (Figure 1a). Conversely, genes that represented more than a 1.5-fold increase in expression levels in both lines were listed as those downregulated by the MpRRB-mediated pathway (Figure 1b). As a result, 650 upregulated and 829 downregulated genes were identified (Tables S1 and S2, respectively).
The GO enrichment analysis showed that genes associated with nitrate/nitrite transport were highly enriched among the upregulated genes (Figure 1c). Eighteen upregulated genes were related to nitrate assimilation and transport as follows: one gene for nitrite reductase (orthologous to Arabidopsis NIR1), 12 genes for nitrate transporter (orthologous to NRT2), two genes for NRT2-associating protein (orthologous to NRT3.1), and three genes for ammonium transporter (orthologous to AMT1) (Table S1). Some of the genes were tested by qRT-PCR, which supported the transcriptional change observed in the RNA-seq data (Figure S1). MpRRB-mediated upregulation of the NIR1 ortholog may indicate a promotive effect of cytokinins on nitrate assimilation [23]. In Arabidopsis, the high-affinity nitrate transporter NRT2 is known to form a complex with NRT3.1 (also referred to as NAR2.1), enhancing nitrate absorption through the roots of plants [24] and AMT1 functions in ammonium uptake [25]. Therefore, in M. polymorpha, cytokinins are necessary for the expression of NRT2, NRT3.1, and AMT1 orthologs, and may promote the uptake of nitrate and ammonium from soil. The GO terms related to oxidative stress were significantly enriched in downregulated genes (Figure 1d). Fifty-three downregulated genes encode peroxidase superfamily proteins (Table S2). Although cytokinins also downregulate several peroxidase genes in Arabidopsis [26], the physiological role has not been identified thus far.
MpCKX1 and MpIPT2 were found to be listed in the upregulated and downregulated genes, respectively (Tables S1 and S2). MpCKX1 is one of the two CKX genes in M. polymorpha, and MpIPT2 is orthologous to Arabidopsis IPT9, encoding a tRNA isopentenyltransferase (IPT) that catalyzes the addition of prenyl-moiety to a tRNA-bound adenine nucleotide and contributes to cytokinin biosynthesis [27]. Arabidopsis CKX4 is also known to be induced by cytokinins [4,26,28]. Therefore, to adjust the hormonal level in plant bodies, it is likely conserved in land plants that cytokinin signaling promotes degradation and inhibits biosynthesis of cytokinins. It was observed that a few auxin-related genes displayed cytokinin responses. Namely, MpPIN-FORMED3 (MpPIN3) encoding an auxin efflux carrier was induced by MpRRB-mediated cytokinin signaling, whereas MpGRETCHEN HAGEN 3 (GH3) B (MpGH3B) was downregulated (Tables S1 and S2). In Arabidopsis, GH3 encodes an acyl acid amido synthetase, which catalyzes the conjugation reaction of indole-3-acetic acid (IAA) with amino acids, leading to inactivation of IAA [29]. Thus, to achieve continuous organ growth, M. polymorpha may regulate auxin homeostasis in response to cytokinins.
Cytokinin signaling upregulates the expression of the R2R3-MYB transcription factor GCAM1
Among the upregulated genes downstream of MpRRB-mediated cytokinin signaling, we found Mp6g04830, a gene encoding the R2R3-MYB transcription factor MpGCAM1 (Table S1). Mpgcam1 mutants had defects in gemma cup formation, and these defects were also observed in the Mprrb knockout line [15,18]. This prompted further investigation of MpGCAM1 in terms of cytokinin signaling. The FPKM values of this gene in MpCKX2-overexpressing and Mprrb knockout lines were 0.64 and 0.76, respectively. This represents a 13.8- and 9.4-fold reduction compared to the expression levels in the wild-type Tak-1 and Tak-2, respectively. In support of the RNA-seq data, the qRT-PCR analysis showed that the MpGCAM1 transcript level in the Mprrb line and the MpCKX2-overexpressing line was 25% and 7% of that in the wild-type (Figures 2a, b).
Next, we tested whether MpGCAM1 responds to exogenously applied cytokinins. When thallus tips of wild-type plants were treated with 50 µM of trans-zeatin (tZ), MpGCAM1 transcripts increased by more than 1.4-fold (Figure 2a). However, such an increase was not observed in the Mprrb line, and a similar trend was also seen for MpRRA mRNA (Figure 2a, c), indicating that MpGCAM1 as well as MpRRA is induced by cytokinins through the MpRRB-mediated pathway. Furthermore, the MpGCAM1 mRNA level in the MpRRB-overexpressing line and the Mprra knockout line in which the cytokinin signaling is activated was 3.1- and 2.6-fold higher, respectively (Figure 2d, e). These results support the idea that MpGCAM1 is controlled by cytokinin signaling.
To examine whether MpGCAM1 is a direct target of MpRRB, we treated wild-type plants with 10 µM of cycloheximide (CHX), a protein synthesis inhibitor, and conducted a qRT-PCR analysis. As shown in Figure 2c, cytokinin-dependent induction of MpRRA was observed even in the presence of CHX, suggesting that MpRRA is directly upregulated through the MpRRB-mediated pathway. It is notable that the application of CHX alone increased MpGCAM1 and MpRRA transcripts. This is likely due to an indirect effect of the inhibition of protein synthesis (Figure 2a, c). Conversely, the level of MpGCAM1 mRNA slightly decreased as a result of the tZ treatment in the presence of CHX (Figure 2a). This result suggests that in response to cytokinins, MpGCAM1 is not a direct target of MpRRB and requires de novo protein synthesis for its induction.
Yasui et al. [18] previously constructed a b-glucuronidase (GUS)-reporter gene for MpGCAM1, in which the 5215-bp upstream and 378-bp downstream from start codon of MpGCAM1 were fused in-frame to the GUS gene. In the transgenic line, the GUS signal was detected in the apical notch, the floor of gemma cups and developing gemma, a similar expression pattern to that of the MpRRBpro:GUS line [15,18]. Mature thalli of MpGCAM1pro:GUS were treated with 50 µM of tZ and used for histological GUS staining (Figure S2). However, the signals neither increased nor spatially expanded upon tZ treatment. This is likely due to their limited sensitivity to exogenously applied cytokinins in M. polymorpha, which contains a relatively high level of cytokinins [15]. To overcome this problem, the expression of MpGCAM1pro:GUS in the MpCKX2-overexpressing line containing a reduced amount of endogenous cytokinins was observed [15]. Five independent lines showed the same GUS expression pattern. Therefore, the representative lines #2 and #3 with different MpCKX2 expression levels were used for detailed analysis (Figure 3a). GUS staining showed that the signals were missing at the apical notch in both lines (Figures 3b–g). The effect of cytokinins on MpGCAM1 expression in gemma cups could not be estimated since gemma cups were not formed in the two lines. Nevertheless, the result of the apical notch suggests that cytokinins upregulate the MpGCAM1 expression in thallus tissues.
MpGCAM1 is involved in cytokinin-dependent gemma cup formation
To examine genetic interaction between MpGCAM1 and cytokinin signaling, double knockout mutants were generated by crossing the male Mpgcam1 line and the female Mprra line. The double mutant lines #15 and #17 never formed gemma cups like the Mpgcam1 line (Figure 4a, b), implying that enhanced gemma cup formation caused by the defect in MpRRA was suppressed in the Mpgcam1 mutant. This result suggests that MpGCAM1 functions downstream of cytokinin signaling in terms of gemma cup formation. Curled thalli were produced in the Mprra knockout line [15], whereas Mpgcam1 did not exhibit any curling phenotype (Figure 4a). Interestingly, the Mpgcam1 Mprra double mutants also produced curled thalli (Figure 4a), indicating that MpGCAM1 is dispensable for cytokinin-dependent differential growth of thalli. It is notable that rhizoid formation was comparable in the wild-type and single mutants as previously described [15,18], and no difference was found between the wild-type and double mutants (Figure 4a).
Cytokinin signaling promotes thallus development from dedifferentiated cell clumps
To investigate the role of cytokinin signaling in organ formation, we took advantage of the dedifferentiation phenotype caused by MpGCAM1 overexpression. Transgenic plants overexpressing MpGCAM1 with a dexamethasone (DEX)-inducible system in the wild-type and the Mprrb knockout line were generated. MpGCAM1 fused to the hormone-binding domain of the glucocorticoid receptor remains inactive due to its cytoplasmic localization in the absence of DEX, while it moves to the nucleus and exerts the function in the presence of DEX [18]. Seven independent lines for the wild-type and nine independent lines for the Mprrb line were isolated, and two lines were selected for each one with similar MpGCAM1 expression levels for further analysis (Figure 5a). Mprrb knockout lines exhibited no gemma cup-phenotype in the absence of DEX, as observed in the original mutant (Figure 5b). However, thallus growth was severely inhibited, and cell clumps and few rhizoids were generated mainly in the tip region of both wild-type and Mprrb when plants were grown in the presence of 5 µM of DEX (Figure 5b). This suggests that the overexpression of MpGCAM1 is sufficient to reprogram and produce dedifferentiated cells in the absence of cytokinin signaling. A number of small thalli were generated at random positions in the wild-type when the DEX-treated plants were transferred to a DEX-free medium. Similarly, small thalli were produced in the Mprrb line, while the thallus size was much smaller than that of the wild-type (Figure 5b). This result suggests that MpRRB-mediated cytokinin signaling is involved in activating cell proliferation after initiating thallus development in a DEX-free medium.