Initiation of asexual reproduction by the AP2/ERF gene GEMMIFER in Marchantia polymorpha

32 Marchantia, asexual reproduction, gemma, AP2/ERF, GEMMIFER, CLE peptide, stem cell 33 34 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 16, 2026. ; https://doi.org/10.64898/2026.01.16.699827doi: bioRxiv preprint

35 Plants have the remarkable ability to clone new individuals from their own bodies. This process, 36 called asexual reproduction, occurs at various structures such as adventitious shoots, bulb ils, 37 tubers and rhizome buds 1. Bryophytes often reproduce asexually through the dispersal of 38 propagules called gemmae with species-specific morphology 2,3. In the liverwort Marchantia 39 polymorpha, discoid-shaped gemmae are produced in specialized cup-shaped structures called 40 gemma cups which are periodically generated along the dorsal midrib of its thalloid body 4. 41 Recent studies on M. polymorpha are beginning to elucidate genetic and hormonal regulation 42 of the gemma and gemma cup development 5. An R2R3-MYB transcription factor, GEMMA 43 CUP-ASSOCIATED MYB1 (MpGCAM1), has been identified as a gene highly expressed in 44 gemma cups 6. Molecular genetic analysis has shown that Mp GCAM1 is required for the 45 formation of gemma cups and gemmae, acting through the control of cell differentiation. 46 Cytokinin signaling promotes the formation of gemma cups through upregulation of 47 MpGCAM1 expression7,8. KARRIKIN INSENSITIVE2 (KAI2)-dependent signaling promotes 48 gemma cup formation by upregulating the expression of a cytokinin biosynthesis enzyme, 49 MpLONELY GUY ( MpLOG), leading to the upregulation of Mp GCAM1 expression9,10. 50 Another R2R3 -MYB transcription factor, SHOT GLASS ( MpSTG), regulates the shape of 51 gemma cup and the gemma development11. In addition, several genes are reported to be required 52 for morphogenesis during early gemma development. MpROOT HAIR DEFECTIVE SIX -53 LIKE1 (MpRSL1) transcription factor gene, targeted by the microRNA FEW RHIZOIDS1 54 (MpFRH1), is required for cellular outgrowth at the floor of gemma cup, which will later 55 develop into gemmae 12,13. The single copy RHO of Plant (MpROP) gene and its regulatory 56 factors are essential for the morphogenesis of various tissues and organs including gemmae and 57 gemma cups 14–17. Signaling of t he plant hormones auxin, ethylene and jasmonate affect the 58 morphology of gemma e18–20. Although these studies have identified a number of genes 59 regulating gemma development, the key genes controlling the initiation of the gemma cell 60 lineage remain unknown. 61 We have previously shown that MpCLA V ATA3/EMBRYO SURROUNDING REGION-62 related 2 (MpCLE2) peptide signaling negatively regulates the formation of gemma cups , in 63 addition to its function in stem cell identity in the meristem located in the apical notch21–23. In 64 a transcriptome analysis, we have identified several differentially expressed transcription factor 65 (DETF) genes in MpCLE2 gain-of-function transgenic lines. Among them, JINGASA (MpJIN), 66 affects stem cell fate in the meristem by promoting periclinal cell division 24. In this study, we 67 report that another DETF, GEMMIFER (MpGMFR)/MpERF14, plays a key role in the 68 initiation of asexual reproduction. This finding would provide a molecular clue to 69 understanding how plant cell fate is regulated in asexual reproduction. 70 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 16, 2026. ; https://doi.org/10.64898/2026.01.16.699827doi: bioRxiv preprint

71 MpGMFR is essential for the formation of gemma cups and gemmae 72 To understand the function of Mp GMFR, we generated loss -of-function alleles using the 73 CRISPR-Cas9 genome editing. Two independent frame-shift alleles (Mpgmfr-1ge and Mpgmfr-74 2ge) were obtained (Fig. S1A and S1B), and both resulted in complete loss of gemma cup and 75 gemma formation. For the quantification, we observed 2 -week-old thallus developed from an 76 explant containing an apical notch. In wild type, all examined plants formed gemma cups on 77 the dorsal surface (Fig. 1A). In contrast, none of the Mpgmfr-2ge plants formed gemma cups 78 and gemmae (Fig. 1B and 1C). This Mpgmfrge phenotype was partially complemented by the 79

of a gRNA-resistant MpGMFR expressed under its own promoter ( Fig. S1C and 80 S1D). To further elucidate the function of MpGMFR, we generated estrogen-inducible artificial 81 microRNA lines targeting Mp GMFR mRNA using the XVE transactivation system 82 (proMpE2F:XVE>>amiR-MpGMFR). In RT –qPCR assays, Mp GMFR mRNA level s were 83 decreased in an estrogen -dependent manner in the proMpE2F:XVE>>amiR-MpGMFR plants 84 compared to wild-type plants (Fig. S1E). The proMpE2F:XVE>>amiR-MpGMFR plants grown 85 for three weeks on estrogen-free medium formed gemma cups, but those on estrogen-containing 86 medium did not (Fig. 1D and 1E). We transferred the 2-week-old plants from estrogen-free to 87 estrogen-containing medium and grown further for 1 week, or vice versa. The transferred plants 88 lost the formation of gemma cups and gemmae in an estrogen-dependent, reversible manner 89 (Fig. 1F and 1G). Collectively, these data show that MpGMFR is required for the formation of 90 gemma cups and gemmae in M. polymorpha. 91 Molecular phylogenetic analysis has shown that M. polymorpha possesses three genes, 92 MpERF1 (Mp1g20040), MpERF14/GMFR (Mp4g00380) and MpERF20/LAXR (Mp5g06970) 93 in class VIII of the AP2/ERF (ERF-VIII) family25,26. A phylogenetic tree with ERF-VIII genes 94 from various land plant species ( Arabidopsis thaliana, Amborella trichopoda, Ginkgo biloba, 95 Salvinia cucullata , Azolla filiculoides , Selaginella moellendorffii , Sphagunum fallax , 96 Physcomitrium patens, Marchantia polymorpha and Anthoceros agrestis), suggests that ERF-97 VIII can be divided into three subgroups, each of which contains single M. polymorpha gene 98 (Fig. S2A). To verify the functional redundancy among ERF-VIII genes in M. polymopha, we 99 conducted complementation test of the Mp gmfrge phenotype by introducing the coding 100 sequences of Mp ERF1 and Mp LAXR, respectively, expressed under Mp GMFR promoter. 101 Neither construct complemented the Mpgmfrge phenotype, suggesting that MpGMFR plays an 102 essential function in the gemma cup and gemma formation that cannot be replaced by either 103 MpERF1 or MpERF20/LAXR (Fig. S2B). 104 105 106 107 108 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 16, 2026. ; https://doi.org/10.64898/2026.01.16.699827doi: bioRxiv preprint 109 Fig.1 | MpGMFR is essential for the formation of gemma cups and gemmae 110 (A and B) Two-week-old thalli grown from the tips of thalli of wild type (A) and Mpgmfr-2ge (B). 111 (C) The number of gemma cups (n=16). (D–G) Effects of MpGMFR knockdown on gemma cup 112 formation in proMpE2F:XVE>>amiR-MpGMFR. Expression of amiR-MpGMFR was induced by 113 β-estradiol (Est) during different periods within the 21 days as indicated below the panels. In 114 A, B, D –G, arrowheads indicate gemma cups. In C, the boxes show the median and 115 interquartile range, and the whiskers show the 1.5x interquartile range. Individual data points 116 are plotted as dots. Two -way ANOVA with Tukey’s post hoc test in C. Means sharing the 117 superscripts are not significantly different from each other, p < 0.05. Scale bars represent 1 cm 118 (A,B,D–G). 119 120 Temporal activation of MpGMFR induces ectopic gemmae formation 121 In M. polyrmopha, gemmae are generated from the floor cells at the bottom of gemma cup. At 122 the initiation of gemma development, a gemma cup floor cell protrudes from the epidermal 123 surface and then divides transversely twice into three cells: a gemma cell, a stalk cell and a 124 basal cell27,28 (Fig. 2A). While the stalk cell stops further divisions, the gemma cell continues 125 to divide and grows into the discoid -shaped gemmae with two apical meristem atic notches at 126 bilateral sites. Thus, the gemma cell acts as the initial cell for the development of gemma. 127 To analyze the ex pression patterns of Mp GMFR during asexual reproduction , we 128 observed transcriptional-fusion reporter lines of Mp GMFR (proMpGMFR:H2B-3xCitrine) 129 under confocal laser scanning microscopy in the cross-section of gemma cup cleared with the 130 iTOMEI method29. The fluorescence signal s of proMpGMFR:H2B-3xCitrine were detected in 131 the gemma cup floor cells. (Fig. 2B–G). In the initial stage of gemma development, the signals 132 were also detected in both the gemma cells and the stalk cell (Fig. 2B), but were lost after the 133 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 16, 2026. ; https://doi.org/10.64898/2026.01.16.699827doi: bioRxiv preprint subsequent several divisions of the gemma cell (Fig. 2C). In a later stage, the signals were 134 scattered within the developing gemmae (Fig. 2D and 2E). After the formation of meristematic 135 notches at bilateral sites of the gemmae, the signals were restricted to cells near the notches 136 (Fig. 2F and 2G ). This pattern is consistent with the promoter activity observed in gemmae 137 (https://mpexpatdb.org/)30. In the vertical longitudinal section of the apical notch of mature 138 plants, the signals were broadly detected at the meristem, except for the central part near the 139 apical cells (Fig. 2H). The signals were detected the floor cells of immature gemma cup but was 140 not in gemmae developing in the cup (Fig. 2I). These data suggest that MpGMFR expression is 141 associated with meristematic region, gemma cup floor cells and the very early stages of gemma 142 development. 143 144 145 146 Fig.2 | Expression patterns of MpGMFR in developing gemma and gemma cup 147 (A) Schematic illustration of early development of gemma (left) and developing gemma cup 148 (right, modified from Barnes and Lang, 1908 28). (B–I) Confocal imaging of proMpGMFR:H2B-149 3xCitrine in developing gemmae ( B–G) and cross -sections of thallus ( H,I). Cell walls were 150 stained with SCRI Renaissance 2200 (SR2200). A sterisks show developing gemmae. Blue 151 and white arrowheads indicate a (sub)apical cell and apical notches, respectively. An arrow 152 indicates the layer of gemma cup floor cells. Scale bars represent 100 μm (B–I). 153 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 16, 2026. ; https://doi.org/10.64898/2026.01.16.699827doi: bioRxiv preprint To understand the function of Mp GMFR, we first generated overexpression lines using 154 constitutive Mp EF1α promoter. The proMpEF1α:MpGMFR plants result in ball-shaped 155 structure composed of small immature thalli (Fig. 3A). To analyze the short-term effects of 156 overexpression, we generated inducible overexpression lines (proMpEF1α:MpGMFR-GR) in 157 which MpGMFR proteins fused with glucocorticoid receptor (GR) are expected to be 158 translocated into the nucleus by dexamethasone (DEX) treatment. In DEX-containing medium, 159 4-day-old proMpEF1α:MpGMFR-GR plants formed protruding cells around the apical notch 160 (Fig. 3B). These cells can be stained by proMpYUC2:GUS and have morphology similar to the 161 gemma cell in the initial stage of gemma development. However , 9-day-old 162 proMpEF1α:MpGMFR-GR plants grown on DEX medium did not form gemmae and only 163 formed ball-shaped structures composed of im mature small thalli , phenocopying the 164 constitutive overexpression lines (Fig. S3). Since MpGMFR expression is suggested to be 165 transiently active in early development of gemmae (Fig. 2A and 2B), we hypothesized that the 166 continuous DEX induction of Mp GMFR compromised normal development of gemma e. To 167 avoid this, 4 -day-old proMpEF1α:MpGMFR-GR plants grow n on DEX -containing medium 168 were transferred to DEX-free medium for further growth. Five days after the transfer, a number 169 of gemmae were formed on both dorsal and ventral side s of thallus (Fig. 3C). The induced 170 gemmae can be detached with water and contain apical notches for further growth (Fig. 3D). 171 These notches contained apical and subapical cells like the stem cell zone in wild -type plants 172 (Fig. 3E). In contrast to wild-type gemmae that have two apical notches at bilateral sites, the 173 number of apical notches varied between one to four among the individual induced gemmae 174 (Fig. 3F). In addition to the apical notches, the induced gemmae had a trace of stalk , with the 175 cell arrangement similar to those in wild type (Fig. 3E). Although these gemma were smaller 176 than those of wild type in the overall size (ground cover area) , they can grow normally on the 177 medium (Fig. 3D and 3G ). Consistently, 5-day-old pro35S:MpGMFR-mVenus gemmalings 178 formed ectopic gemma on the surface (Fig. 3H). Activity of other ERF-VIII genes, MpERF1 179 and Mp LAXR, were examined by generating proMpEF1α:MpERF1-GR and 180 proMpEF1α:MpLAXR-GR lines. In both cases, gemmae were not induced unlike 181 proMpEF1α:MpGMFR-GR lines, supporting the notion that ERF -VIII genes do not exert 182 redundant function in M. polymorpha (Fig. S4). Collectively, these data show that temporal 183 induction of MpGMFR leads to the ectopic formation of gemmae capable of reproducing as 184 new individuals. 185 186 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 16, 2026. ; https://doi.org/10.64898/2026.01.16.699827doi: bioRxiv preprint 187 Fig.3 | Temporal activation of MpGMFR induces ectopic gemmae formation 188 (A) Morphology of proMpEF1α:MpGMFR plant. The right panel shows the magnified image of 189 a dashed line square in the left panel. (B) Morphology of apical notch with proMpYUC2:GUS 190 marker in 4-day-old plants grown from gemmae. The right panel shows the magnified image 191 of a dashed line square in the center panel . (C) Morphology of 9-day-old plants grown from 192 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 16, 2026. ; https://doi.org/10.64898/2026.01.16.699827doi: bioRxiv preprint gemmae. Wild-type plants were grown for 9 days on normal medium. proMpEF1α:MpGMFR-193 GR plants were first grown on 1 µM DEX -containing medium for 4 days, and transferred to 194 DEX-free medium and grown for 5 days. Overall morphology (left), SEM images of the surface 195 of thallus (center) and 3D-reconstructed images of the cross-section of the thallus in confocal 196 imaging (right) were shown. (D) Morphology of gemmae of wild type (left) and 197 proMpEF1α:MpGMFR-GR (right, two panels). Arrowheads show notches. (E) Confocal imaging 198 of the stem cell zone and a trace of stalk in gemmae of wild type (top) and 199 proMpEF1α:MpGMFR-GR (bottom). The traces of stalk were shown by 3D -reconstructed 200 images. (F) The frequency of gemmae with different numbers of apical notches (n=20–23). (G) 201 Ground cover area of 7 and 14 -day-old plants grown from gemmae (n=18 –20). (H) 3D-202 reconstructed images of 5 -day-old pro35S:MpGMFR-mVenus gemmaling. In F, data are 203 represented by mean (bars) and individual data points (dots). Two -way ANOVA with Tukey’s 204 post hoc test in F. Means sharing the superscripts are not significantly different from each other, 205 p < 0.05. In C and E, cell walls were stained with SR2200. Scale bars represent 1 mm (A and 206 C, left), 200 μm (C, center), 100 μm (B, C, right and D) and 50 µm (G). 207 208 209 To investigate the timing of gemma initiation after MpGMFR induction, we observed 210 apical notches of proMpEF1α:MpGMFR-GR plants grown for 1 to 4 days on 1 µM DEX -211 containing medium. At day 2, protruding cells that have undergone a transverse division was 212 observed at the apical notch (Fig. 4A). The protruding cells divided transversely again to 213 produce a gemma cell and a stalk cell at day 3. At day 4, the gemma cells divided further while 214 the stalk cell remained inactive in cell division, recapitulating the division patterns in early 215 development of wild-type gemmae. In some cases, the cell division patterns of the gemma cell 216 were unusual compared to wild type, which might be the cause of the malformed gemmae (Fig. 217 3D and 3F ). In the 3D -reconstructed images of apical notch area of the parent plants, the 218 induced gemmae can be observed on both dorsal and ventral surfaces in proMpEF1α:MpGMFR-219 GR plants while no gemmae were observed in wild type (Fig. 4B and S5). These data indicate 220 that MpGMFR controls the initiation of gemma cell lineage in the meristem. 221 222 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 16, 2026. ; https://doi.org/10.64898/2026.01.16.699827doi: bioRxiv preprint 223 Fig.4 | MpGMFR initiates the gemma cell lineage in the meristem 224 (A) Confocal images of the apical notch in wild type and proMpEF1α:MpGMFR-GR. Panels in 225 the bottom row show magnified images of proMpEF1α:MpGMFR-GR in the middle row. (B) 3D-226 reconstructed apical notches in 4 -day-old plants of wild type and proMpEF1α:MpGMFR-GR. 227 Cell walls were stained with SR2200. Scale bars represent 50 μm (A). 228 229 Weak induction of MpGMFR promotes gemma cup formation 230 To examine the concentration -dependent effects of MpGMFR, proMpEF1α:MpGMFR-GR 231 plants were grown on growth media containing different concentrations of DEX (Fig. S6). In 232 contrast to the 1 µM DEX treatment that only induced gemma -like tissues , 10 nM DEX 233 treatment induced the formation of semicircular gemma cups near the apical notches , which 234 contains gemma -like tissues inside , in 7 -day-old plants (Fig. 5A). The rim structure of the 235 induced gemma cups in the MpGMFR-GR lines were similar to that of wild-type gemma cups 236 (Fig. 5B). These data suggest that MpGMFR can also induce the formation of gemma cups. 237 To examine the functional relationship between Mp CLE2 and Mp GMFR genes, we 238 generated proMpEF1α:MpGMFR-GR lines in the background of a gain -of-function allele of 239 MpCLE2, proMpYUC2:MpCLE2-221. In this background, overexpression of MpCLE2 results in 240 reduced gemma cup formation and decreased MpGMFR expression level s22,24. In the 241 experiment, gemmae were first grown on DEX-free medium for 12 days , followed by an 242 additional 8 days on 10 nM DEX-containing medium. As a result, two independent MpGMFR-243 GR lines showed a significant increase in the number of gemma cup s compared to the 244

line (Fig. 5C and 5D ), suggesting that Mp CLE2 overexpression phenotypes on 245 gemma cup formation can be suppressed by restoring the reduced MpGMFR expression. 246 247 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 16, 2026. ; https://doi.org/10.64898/2026.01.16.699827doi: bioRxiv preprint 248 Fig.5 | Weak induction of MpGMFR promotes gemma cup formation 249 (A) Effects of different concentrations of DEX on gemma cup formation in 250 proMpEF1α:MpGMFR-GR plants grown for 7 days from gemmae. Panels in the bottom row 251 show the magnified images of dashed line squares in the panels above. (B) SEM images of 252 the gemma cups observed in 20 -day-old wild-type plant and 7-day-old proMpEF1α:MpGMFR-253 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 16, 2026. ; https://doi.org/10.64898/2026.01.16.699827doi: bioRxiv preprint GR plant. (C) Effects of MpGMFR induction in MpCLE2 overexpression plant. Twelve-day-old 254 plants grown on normal medium were transferred to 10 nM DEX -containing medium and 255 further grown for 8 days. Panels in the bottom row show the magnified images of dashed line 256 squares in the panels above. (D) The number of gemma cups (n=4). In D, the boxes show the 257 median and interquartile range, and the whiskers show the 1.5x interquartile range. Individual 258 data points are plotted as dots. Two-way ANOVA with Tukey’s post hoc test in D. Means sharing 259 the superscripts are not significantly different from each other, p < 0.05. Scale bars represent 260 5 mm (A, C) or 500 μm (B). 261 262 Genetic interaction of MpGMFR and MpGCAM1 263 Loss-of-function alleles of MpGCAM1 are deficient in the gemma cup and the gemma6, similar 264 to those of MpGMFR. To analyze the functional interaction between these genes, we introduced 265 frame-shift mutations in Mp GCAM1 by CRISPR -Cas9 genome editing in the 266 proMpEF1α:MpGMFR-GR background (Fig. S7). As expected, gemma and gemma cup 267 formation was completely lost in these lines, and we used one of them for further analysis . To 268 analyze the effects on gemma cup formation , explants containing an apical notch were grown 269 for 7 days on 1 µM DEX -containing medium, followed by an additional 7 days on DEX-free 270 medium. Under these conditions , gemma-like tissues were formed in proMpEF1α:MpGMFR-271 GR plants. In contrast, such gemma-like tissues were not observed in the Mp gcam1ge 272 proMpEF1α:MpGMFR-GR line (Fig. 6A). We also could not observe protruding cells or early-273 stage gemmae in this line, suggesting that both MpGMFR and MpGCAM1 are essential for the 274 initiation of gemma development. 275 To further analyze the relationship, we compared expression patterns of these genes 276 during gemma development. While the MpGMFR expression was detected in the gemma cup 277 floor cell s and early development of gemmae (Fig. 2), the fluorescence signal of the 278 MpGCAM1-Citrine knock-in allele was strongly detected throughout the gemma development 279 but was barely detectable in the gemma cup floor cells6,11 (Fig. S8), suggesting that MpGMFR 280 expression precedes MpGCAM1 expression in gemma development. We further examined the 281 expression levels of each gene in the mutants of the other gene by RT-qPCR assays. In the 6-282 day-old plants grown from explants containing an apical notch, the MpGCAM1 mRNA level 283 was decreased to about 4 % in Mpgmfrge lines compared to wild type (Fig. 6B). In the 4-day-284 old gemmalings of an MpGMFR-GR overexpression line grown with DEX , the MpGCAM1 285 mRNA level was increased to about 750 % compared to mock treatment (Fig. 6C). On the other 286 hand, in the 6 -day-old plants grown from explants, the MpGMFR mRNA level was not 287 significantly affected in a Mpgcam1ko line6 compared to wild type (Fig. 6D). In the 4-day-old 288 gemmalings of an MpGCAM1-GR overexpression line grown with DEX, the MpGMFR mRNA 289 level was decreased to about 74 % compared to mock treatment (Fig. 6E). These data suggest 290 that MpGMFR promotes MpGCAM1 expression directly or indirectly while MpGCAM1 has 291 little impact on MpGMFR expression. Collectively, these data suggest that gemma development 292 is initiated through the stepwise function of MpGMFR and MpGCAM1. 293 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 16, 2026. ; https://doi.org/10.64898/2026.01.16.699827doi: bioRxiv preprint 294 Fig.6 | Genetic interaction of MpGMFR and MpGCAM1 295 (A) Effect of MpGMFR induction in Mp gcam1ge plant. Seven-day-old plants grown on 1 μM 296 DEX-containing medium were transferred to DEX -free medium and further grown for 7 days. 297 Morphology of apical notch (top). SEM images of the surface of plants (bottom). (B–C) Relative 298 expression levels of MpGMFR. Six-day-old thalli of the WT and Mp gcam1ko plants (B). Four-299 day-old gemmalings of proMpEF1α:MpGCAM1-GR on mock or 1 µM DEX-containing medium 300 (C). (D–E) Relative expression levels of Mp GCAM1. Six-day-old thalli (D) and four-day-old 301 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 16, 2026. ; https://doi.org/10.64898/2026.01.16.699827doi: bioRxiv preprint gemmalings (E) are examined. In B–E, the expression levels were normalized by Mp APT 302 expression. In B–E, data are represented by mean (bars) and individual data points (dots). 303 Two-way ANOVA with Tukey’s post hoc test in B and Student’s t test in C–E. Means sharing 304 the superscripts are not significantly different from each other, p < 0.05. In C,D and E, P-value 305 is indicated above each pair of bars. Scale bars represent 1 cm (A) or 500 μm (B). 306 307

308 Due to its nature of autonomous asexual reproduction, the liverwort Marchantia polymorpha 309 has become an important model organism for studying the molecular mechanisms of asexual 310 reproduction. In this study, we identified MpERF14/GMFR as a key regulator of the initiation 311 of asexual reproduction. Molecular genetic analysis shows that Mp GMFR is required for the 312 development of gemma cups and gemmae while overexpression of MpGMFR induces gemma 313 cups or gemmae in a dose -dependent manner. When induced weakly, M pGMFR acts as a 314 positive regulator of gemma cup development, which is consistent with the inhibitory function 315 of MpCLE2 in gemma cup development because the expression of MpGMFR is suppressed by 316 MpCLE2 peptide signaling . Strong short-term induction of Mp GMFR activity results in the 317 formation of ectopic gemmae with a junction to the stalk and meristem atic notches, which 318 enable gemmae to be detached from the parental plant and grow independently in a remote 319 location. In contrast, c ontinued overexpression of Mp GMFR interferes with normal 320 development of gemmae, suggesting the importance of spatio-temporal regulation of MpGMFR 321 expression during gemma development. Consistently, the expression of MpGMFR is observed 322 in the gemma cup floor cells and the early stages of gemma development , but it is gradually 323 restricted near the meristems during gemma development. 324 Previous studies show that Mp GCAM1 is required for gemma and gemma cup 325 formation6,8–10. In this study, we show that Mp GCAM1 expression is substantially reduced in 326 the MpGMFR loss-of-function alleles while it is increased after the induction of Mp GMFR. 327 The formation of ectopic gemmae in the MpGMFR-GR lines requires MpGCAM1, suggesting 328 that MpGCAM1 may act in the gemma cell lineage to reinforce and/or maintain the cell identity. 329 In contrast, Mp GMFR expression was not significantly affected in the MpGCAM1 loss-of-330 function allele . In the expression analysis with fluorescent protein reporters, Mp GMFR 331 expression slightly precedes the Mp GCAM1 expression during asexual reproduction. 332 Collectively, these data suggest that early stage of gemma development can be further divided 333 into two steps, controlled by MpGMFR and MpGCAM1, respectively. 334 Our phylogenetic analysis suggests that class VIII of AP2/ERF family proteins can be 335 classified into three subgroups, each comprising genes from major lineages of land plants 336 including single M. polymorpha gene (Mp ERF1, Mp ERF20/LAXR and Mp ERF14/GMFR). 337 Thus, three distinct ERF-VIII genes were already present in the last common ancestor of land 338 plants while their closest homolog is unclear in the ERF family of streptophyte algae 25. In 339 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 16, 2026. ; https://doi.org/10.64898/2026.01.16.699827doi: bioRxiv preprint Arabidopsis thaliana, AtERF84 which belongs to the subgroup containing MpGMFR has been 340 reported as a positive regulator in drought re sistance while its function in development is 341 unclear31. In M. polymorpha , MpERF20/LAXR regulates cellular reprogramming during the 342 tissue regeneration in response to the reduction of the auxin level after the decapitation of 343 meristematic notch32. MpERF20/LAXR is a homolog of ENHANCER OF SHOOT 344 REGENERATION 1/DORNRÖSCHEN (ESR1/DRN) in A. thaliana, which promotes stem cell 345 formation in the shoot regeneration33. In the moss P. patens, PpESR genes promote gametophore 346 apical cell identity whose expression is regulated by cytokinin signaling 34, implying deep 347 evolutionary conservation of these genes in controlling the stem cell identity in land plants. The 348 function of MpERF14/GMFR in asexual reproduction has a similarity to the ESR/LAXR genes 349 in terms of regulation of cell identity leading to the formation of new meristems even though 350 MpERF20/LAXR cannot replace the activity of MpERF14/GMFR. 351 In summary, we have presented the identification of MpERF14/GMFR as a molecular 352 trigger of the asexual reproduction in M. polymorpha . This work provides a foundation for 353 future studies to elucidate how plants evolved the formation of extra meristems from their body 354 that may have made a significant contribution to their prosperity on land. 355 356

357 We thank Ikuko Nakanomyo, Jutarou Fukazawa, Sayaka Matsui, Yuuki Sakai for technical 358 assistance. This work was supported by JSPS KAKENHI (Grant Number JP22H02676) to Y .H., 359 Takeda Science Foundation to Y .H., GteX Program Japan (JPMJGX23B0) to K.I., the Program 360 for Forming Japan's Peak Research Universities (J -PEAKS) from JSPS to KI, BBSRC 361 BB/T007117/1 to J.H and BBSRC BB/F011458/1 for confocal microscopy. F.R. is a 362 Leverhulme Early Career Fellow (ECF -2023-534) funded by the Leverhulme Trust and the 363 Isaac Newton Trust (23.08(f)), I.B. is funded by the Herschel Smith Fund studentship. 364 365 Author contributions 366 G.T. and Y .H. conceived and designed the research. G.T., S.Y ., F.R., I.B. and M.S. performed 367 the experiments. G.T., S.Y ., F.R., M.S. and Y .H. analyzed the data. K.I. contributed materials 368 and analysis tools. F.R., T.K., J.H. and Y .H. supervised the project. G.T. and Y .H. wrote the 369 manuscript. All authors reviewed and edited the manuscript. 370 371 Declaration of Interests 372 The authors declare no competing interests. 373 374 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 16, 2026. ; https://doi.org/10.64898/2026.01.16.699827doi: bioRxiv preprint

375 Plant materials and growth conditions 376 Marchantia polymorpha male Takaragaike-1 (Tak-1) accession was used as wild type in this 377 study. M. polymorpha plants were grown on half -strength Gamborg B5 medium (pH 5.5) 378 solidified with 1.4% agar. M. polymorpha plants were grown at 22 ℃ under continuous white 379 light. Transgenic plants are listed in Table S1. 380 Plasmid Construction 381 Primers used in this study are listed in Table S2. All plant transformation vectors were generated 382 using the Gateway cloning system (Thermo Fisher Scientific, Waltham, MA, United States). 383 Gateway destination vectors are described in Ishida et al. (2022)32, Ishizaki et al. (2015)35 and 384 Sugano et al. (2018)36. 385 For genome editing of Mp GMFR/MpERF14 and Mp GCAM1, guide RNA s were 386 designed using CRISPRdirect37. The plasmids for genome editing were constructed according 387 to Sugano et al. (2018)36. 388 For the complementation study of genome editing alleles for Mp GMFR, a 5026 bp 389 MpGMFR promoter sequence was PCR amplified with a primer pair of 390 MpGMFR_prom_F_InFusion_XbaI and MpGMFR_prom_R_InFusion_XbaI, and cloned into 391 the Xba I digestion site of pMpGWB30 1 and pMpGWB31 2 vectors using In -Fusion HD 392 Cloning Kit (Takara Bio , Shiga, Japan ) to produce pMpGWB3 01-proMpGMFR and 393 pMpGWB312-proMpGMFR, respectively . The coding sequence of MpGMFR was PCR 394 amplified from M. polymorpha cDNA with a primer pair of Mp GMFR_CDS_F and 395 MpGMFR_CDS_R_+stop or MpGMFR_CDS_R_-stop, and cloned into pENTR 4 Dual 396 Selection Vector (Thermo Fisher Scientific). To introduce gRNA -resistant mutation to the 397 resulting plasmids, pENTR_MpGMFR_CDS_+stop and pENTR_Mp GMFR_CDS_-stop, site 398 directed mutagenesis was performed. The plasmids were PCR amplified with mutagenesis 399 primers, MpGMFR_CDS_gRNAres_F and Mp GMFR_CDS_gRNAres_R, and subjected to 400 digestion with Dpn I (Takara Bio), followed by transformation of Escherichia coli. 401 Mutagenized plasmids were selected by DNA sequencing. pENTR-402 MpGMFR_CDS_gRNAres_+stop was transferred to pMpGWB301-proMpGMFR, and pENTR-403 MpGMFR_CDS_gRNAres_-stop was transferred to pMpGWB3 12-proMpGMFR using 404 Gateway LR Clonase II Enzyme mix (Thermo Fisher Scientific). 405 For production of the estrogen -inducible artificial microRNA (amiRNA) lines, an 406 amiRNA target sequence was designed at coding sequence of MpGMFR using amiRNA Design 407 Helper38 (https://marchantia.info/tools/amir_helper/) to be inserted into MpMIR160 backbone. 408 Cloning into t he pMpGWB36832 plasmid to generate estrogen-inducible amiRNA construct 409 was performed according to Sakai et al. (2024)17. 410 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 16, 2026. ; https://doi.org/10.64898/2026.01.16.699827doi: bioRxiv preprint For promoter reporter analysis, a 5026 bp DNA fragment of Mp GMFR promoter 411 sequence flanking the translational initiation site was PCR amplified with a primer pair of 412 MpGMFR_prom_F and MpGMFR_prom_R, and cloned into pENTR/D-TOPO vector (Thermo 413 Fisher Scientific). The resulting plasmid, pENTR -proMpGMFR, was transferred to the 414 pMpGWB323-H2B vector24 using Gateway LR Clonase II Enzyme mix. 415 For production of MpGMFR overexpression alleles, pENTR-MpGMFR_CDS_+stop was 416 transferred to pMpGWB 303 using Gateway LR Clonase II Enzyme mix . For production of 417 inducible MpGMFR overexpression alleles, pENTR-MpGMFR_CDS_-stop was transferred to 418 pMpGWB113. For the pro35S:MpGMFR-mVenus, the CDS of Mp GMFR was synthesi zed 419 (Genewiz) as a L0_CDS12 part and directly cloned into the pBy12 vector as described in 420 Romani et al. (2024)39. 421 For the functional analysis for Mp ERF1 and MpERF20/LAXR, the coding sequences of 422 MpERF1 and MpERF20/LAXR were PCR amplified from M. polymorpha cDNA with a primer 423 pair of Mp ERF1_CDS_F and Mp ERF1_CDS_R_-stop or MpLAXR_CDS_F and 424 MpLAXR_CDS_R_-stop, and cloned into pENTR4 Dual Selection Vector. The resulting 425 plasmid, pENTR -MpERF1_CDS_-stop or pENTR-MpLAXR_CDS_-stop were transferred 426 into pMpGWB313 or pMpGWB312-proMpGMFR using Gateway LR Clonase II Enzyme mix. 427 Generation of transgenic plants 428 Agrobacterium-mediated transformation of M. polymorpha was performed using regenerating 429 thalli according to Kubota et al. (2013)40. CRISPR/Cas9-based genome editing was performed 430 according to Sugano et al. (2018)36. Mutations in the guide RNA target loci were examined by 431 direct sequencing of PCR product amplified from genome DNA samples with primers listed in 432 Table S2. 433 Imaging and phenotypic measurement 434 For the analysis of overall plant morphology and the measurement for the number of gemma 435 cup, plants were imaged under a digital microscope (DMS1000, Leica Microsystems, Wetzlar, 436 Germany) or a digital camera (TG-6, Olympus). 437 For fluorescence observation in confocal imaging, plants were fixed and cleared with 438 iTOMEI protocol29 as described in Takahashi et al. (2023)24. The cleared samples were mounted 439 in the mounting solution and observed under a confocal laser scanning microscopy (Fluoview 440 FV3000, Olympus). For the observation of developing gemmae and gemma cups, hand-sections 441 were prepared with a scalpel. For 3D -reconstruction, Z-series images were processed using 442 ‘3D-viewer’ or ‘3D-projection’ function of Fiji software41. 443 For the scanning electron microscopy imaging, plants were pre -fixed with 4 % 444 glutaraldehyde in 0.05 M phosphate buffer (pH 7.2) for 2 hours at room temperature, followed 445 by washing in the phosphate buffer. Pre-fixed plants were post-fixed with 1 % osmium tetroxide 446 in 0.05 M phosphate buffer (pH 7.2) for 2 hours at 4 ℃. Fixed plants were dehydrated using 447 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 16, 2026. ; https://doi.org/10.64898/2026.01.16.699827doi: bioRxiv preprint ethanol series at room temperature. The plants were immersed in t-butyl alcohol and freeze -448 dried in an evacuator (VFD -21S, Vacuum Device Inc.) until completely dry. Finally, samples 449 were coated with gold by an ion sputtering device (JFC1500, JEOL) and observed under a 450 scanning electron microscopy (JSM-T220A, JEOL). 451 GUS staining 452 GUS staining was performed according to Hirakawa et al. (2020)21. Briefly, individual plants 453 were stained separately in 30 –50 µL GUS staining solution (50 mM sodium phosphate buffer 454 pH7.2, 1 mM potassium-ferrocyanide, 1 mM potassium-ferricyanide, 10 mM EDTA, 0.01 % 455 Triton X-100 and 1mM 5-bromo-4-chloro-3-indolyl-b-D-glucuronic acid) at 37 ℃ in dark. 456 GUS-stained samples were washed with water, cleared with ethanol, and mounted with clearing 457 solution (chloral hydrate-glycerol-water, 8:1:2) for imaging under a light microscope (BX51, 458 Olympus, Tokyo, Japan). 459 RT-qPCR 460 To quantify MpGMFR and MpGCAM1 mRNA levels, total RNA was extracted from thallus 461 using NucleoSpin RNA Plant (Macherey-Nagel, Duren, Germany) according to manufacturer’s 462 instruction. First-strand cDNAs were prepared using ReverTra Ace qPCR RT Master Mix with 463 gDNA Remover (TOYOBO, Osaka, Japan). RT -qPCR was performed with the reagents, TB 464 Green Premix Ex Taq II (Takara Bio) and THUNDERBIRD Next SYBR qPCR Mix 465 (TOYOBO), using the devices, Step One Plus Real -Time PCR System (Thermo Fisher 466 Scientific) and CFX Connect (Bio-Rad Laboratories, California, USA). MpAPT (Mp3g25140) 467 was used as a reference gene. 468 Phylogenetic analysis 469 Protein sequences were retrieved from the following databases: MarpolBase 42 470 (https://marchantia.info), Phytozome 43 (https://phytozome-next.jgi.doe.gov/), TAIR 471 (http://www.arabidopsis.org/) and GinkgoDB 44 (https://ginkgo.zju.edu.cn/genome/). 472 Alignment was performed on the amino acid sequences of the AP2 domain using CLUSTALW 473 (https://www.genome.jp/tools-bin/clustalw). After manually removing the alignment gaps 474 using SeaView45, phylogenetic analysis was performed on the alignment using MrBayes3.2.746. 475 Two runs with four chains of Markov chain Monte Carlo (MCMC) iterations were performed 476 for 6,000,000 generations, keeping one tree every 100 generations. The first 25% of the 477 generations were discarded as burn -in and the remaining trees were used to calculate a 50% 478 majority-rule tree. The standard deviation for the two MCMC iteration runs was below 0.01, 479 suggesting that it was sufficient for the convergence of the two runs. Convergence was assessed 480 by visual inspection of the plot of the log likelihood scores of the two runs calculated by 481 MrBayes47. Character matrix used to run the Bayesian phylogenetic analysis is provided in Data 482 S1. 483 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 16, 2026. ; https://doi.org/10.64898/2026.01.16.699827doi: bioRxiv preprint Data visualization 484 The statistical software R version 4.5.0 was used for data visualization. 485 Quantification and statistical analysis 486 For phenotypic quantification, JMP pro 18 ( JMP Statistical Discovery LLC , North Carolina, 487 USA) was used for statistical tests. Statistical details including the type of test, sample size and 488 statistical significance can be found in figure legends. 489 490 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 16, 2026. ; https://doi.org/10.64898/2026.01.16.699827doi: bioRxiv preprint

491 1. Niklas, K. J. & Cobb, E. D. The evolutionary ecology (evo -eco) of plant asexual reproduction. Evol 492 Ecol 31, 317–332 (2017). 493 2. Imura, S. Vegetative Diaspores in Japanese Mosses. J Hattori Bot Lab 77, 177–232 (1994). 494 3. Laaka-Lindberg, S., Korpelainen, H. & Pohjamo, M. Dispersal of Asexual Propagules in Bryophytes. J 495 Hattori Bot Lab 93, 319–330 (2003). 496 4. Shimamura, M. Marchantia polymorpha : Taxonomy, Phylogeny and Morphology of a Model System. 497 Plant Cell Physiol 57, 230–256 (2016). 498 5. Kato, H., Yasui, Y . & Ishizaki, K. Gemma cup and gemma development in Marchantia polymorpha. 499 New Phytol 228, 459–465 (2020). 500 6. Yasui, Y . et al. GEMMA CUP-ASSOCIA TED MYB1, an Ortholog of Axillary Meristem Regulators, Is 501 Essential in Vegetative Reproduction in Marchantia polymorpha. Curr Biol 29, 3987–3995.e5 (2019). 502 7. Aki, S. S. et al. Cytokinin Signaling Is Essential for Organ Formation in Marchantia polymorpha. Plant 503 Cell Physiol 60, 1842–1854 (2019). 504 8. Aki, S. S. et al. R2R3-MYB transcription factor GEMMA CUP -ASSOCIATED MYB1 mediates the 505 cytokinin signal to achieve proper organ development in Marchantia polymorpha. Sci Rep 12, 21123 506 (2022). 507 9. Komatsu, A. et al. Control of vegetative reproduction in Marchantia polymorpha by the KAI2 -ligand 508 signaling pathway. Curr Biol 33, 1196–1210.e4 (2023). 509 10. Komatsu, A. et al. KAI2-dependent signaling controls vegetative reproduction in Marchantia 510 polymorpha through activation of LOG-mediated cytokinin synthesis. Nat Commun 16, 1263 (2025). 511 11. Sakai, Y . et al. SHOT GLASS, an R2R3 -MYB transcription factor, promotes gemma cup and 512 gametangiophore development in Marchantia polymorpha. New Phytol 247, 2678–2696 (2025). 513 12. Proust, H. et al. RSL Class I Genes Controlled the Development of Epidermal Structures in the Common 514 Ancestor of Land Plants. Curr Biol 26, 93–99 (2016). 515 13. Honkanen, S., Thamm, A., Arteaga-Vazquez, M. A. & Dolan, L. Negative regulation of conserved RSL 516 class I bHLH transcription factors evolved independently among land plants. eLife 7, e38529 (2018). 517 14. Hiwatashi, T. et al. The RopGEF KARAPPO Is Essential for the Initiation of Vegetative Reproduction 518 in Marchantia polymorpha. Curr Biol 29, 3525–3531.e7 (2019). 519 15. Rong, D. et al. ROP signaling regulates spatial pattern of cell division and specification of meristem 520 notch. Proc Natl Acad Sci USA 119, e2117803119 (2022). 521 16. Mulvey, H. & Dolan , L. RHO GTPase of plants regulates polarized cell growth and cell division 522 orientation during morphogenesis. Curr Biol 33, 2897–2911.e6 (2023). 523 17. Sakai, Y . et al. Regulation of ROP GTPase cycling between active and inactive states is essential for 524 vegetative organogenesis in Marchantia polymorpha. Development 151, dev202928. (2024). 525 18. Kato, H. et al. The Roles of the Sole Activator -Type Auxin Response Factor in Pattern Formation of 526 Marchantia polymorpha. Plant Cell Physiol 58, 1642–1651 (2017). 527 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 16, 2026. ; https://doi.org/10.64898/2026.01.16.699827doi: bioRxiv preprint 19. Monte, I. et al. A Single JAZ Repressor Controls the Jasmonate Pathway in Marchantia polymorpha. 528 Mol Plant 12:185–198 (2019). 529 20. Li, D. et al. Ethylene-independent functions of the ethylene precursor ACC in Marchantia polymorpha. 530 Nat Plants 6, 1335–1344 (2020). 531 21. Hirakawa, Y . et al. Induction of Multichotomous Branching by CLA V ATA Peptide in Marchantia 532 polymorpha. Curr Biol 30, 3833–3840.e4 (2020). 533 22. Takahashi, G., Betsuyaku, S., Okuzumi, N., Kiyosue, T. & Hirakawa, Y . An Evolutionarily Conserved 534 Coreceptor Gene Is Essential for CLA V A TA Signaling in Marchantia polymorpha. Front Plant Sci 12, 535 (2021). 536 23. Hirakawa, Y . Evolution of meristem zonation by CLE gene duplication in land plants. Nat Plants 8, 537 735–740 (2022). 538 24. Takahashi, G., Kiyosue, T. & Hirakawa, Y . Control of stem cell behavior by CLE-JINGASA signaling 539 in the shoot apical meristem in Marchantia polymorpha. Curr Biol 33, 5121–5131.e6 (2023). 540 25. Bowman, J. L. et al. Insights into Land Plant Evolution Garnered from the Marchantia polymorpha 541 Genome. Cell 171, 287–304.e15 (2017). 542 26. Montgomery, S. A. et al. Chromatin Organization in Early Land Plants Reveals an Ancestral Association 543 between H3K27me3, Transposons, and Constitutive Heterochromatin. Curr Biol 30, 573–588.e7 (2020). 544 27. Mirbel C.-F. Researches anatomiques et physiologiques sur le Marchantia polymorpha. Mém. Acad. R. 545 Sci. Inst. France 13: 337–436 (1835). 546 28. Barnes, C. R. & Land, W. J. G. Bryological Papers. II. The Origin of the Cupule of Marchantia. Bot Gaz 547 46, 401–409 (1908). 548 29. Sakamoto, Y . et al. Improved clearing method contributes to deep imaging of plant organs. Commun 549 Biol 5, 1–12 (2022). 550 30. Romani F. et al. The landscape of transcription factor promoter activity during vegetative development 551 in Marchantia. Plant Cell 36, 2140–2159 (2024). 552 31. Zhang, P. et al. The long non-coding RNA DANA2 positively regulates drought tolerance by recruiting 553 ERF84 to promote JMJ29-mediated histone demethylation. Mol Plant 16, 1339–1353 (2023). 554 32. Ishida, S. et al. Diminished Auxin Signaling Triggers Cellular Reprogramming by Inducing a 555 Regeneration Factor in the Liverwort Marchantia polymorpha. Plant Cell Physiol 63, 384–400 (2022). 556 33. Banno, H., Ikeda, Y ., Niu, Q.-W. & Chua, N.-H. Overexpression of Arabidopsis ESR1 Induces Initiation 557 of Shoot Regeneration. Plant Cell 13, 2609–2618 (2001). 558 34. Hata, Y . et al. snRNA-seq analysis of the moss Physcomitrium patens identifies a conserved cytokinin-559 ESR module promoting pluripotent stem cell identity. Dev Cell 60, 1884–1899 (2025). 560 35. Ishizaki, K. et al. Development of Gateway Binary Vector Series with Four Different Selection Markers 561 for the Liverwort Marchantia polymorpha. PLoS One 10, e0138876 (2015). 562 36. Sugano, S. S. et al. Efficient CRISPR/Cas9-based genome editing and its application to conditional 563 genetic analysis in Marchantia polymorpha. PLoS One 13, e0205117 (2018). 564 37. Naito, Y ., Hino, K., Bono, H. & Ui-Tei, K. CRISPRdirect: software for designing CRISPR/Cas guide 565 RNA with reduced off-target sites. Bioinformatics 31, 1120–1123 (2015). 566 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 16, 2026. ; https://doi.org/10.64898/2026.01.16.699827doi: bioRxiv preprint 38. Flores-Sandoval, E., Dierschke, T., Fisher, T. J. & Bowman, J. L. Efficient and Inducible Use of 567 Artificial MicroRNAs in Marchantia polymorpha. Plant Cell Physiol 57, 281–290 (2016). 568 39. Romani, F. et al. The minimal cell-cycle control system in Marchantia as a framework for understanding 569 plant cell proliferation. bioRxiv 2025.03. 12.642684 (2025). 570 40. Kubota, A., Ishizaki, K., Hosaka, M. & Kohchi, T. Efficient Agrobacterium-Mediated Transformation 571 of the Liverwort Marchantia polymorpha Using Regenerating Thalli. Biosci Biotechnol Biochem 77, 572 167–172 (2013). 573 41. Schindelin, J. et al. Fiji: an open-source platform for biological-image analysis. Nat Methods 9, 676–574 682 (2012). 575 42. Tanizawa, Y . et al. MarpolBase: Genome database for Marchantia polymorpha featuring high quality 576

genome sequences. bioRxiv 2025.03.30.646155 (2025). 577 43. Goodstein, D.M. et al. Phytozome: a comparative platform for green plant genomics. Nucleic Acids Res 578 40, D1178– D1186 (2012). 579 44. Gu, K. J., Lin, C. F., Wu, J. J. & Zhao, Y. P. GinkgoDB: an ecological genome database for the living 580 fossil, Ginkgo biloba. Database (Oxford) 2022, baac046 (2022). 581 45. Gouy, M., Tannier, E., Comte, N. & Parsons, D. P . Seaview Version 5: A Multiplatform Software for 582 Multiple Sequence Alignment, Molecular Phylogenetic Analyses, and Tree Reconciliation. Methods 583 Mol Biol 2231, 241–260 (2021). 584 46. Ronquist, F. et al. MrBayes 3.2: Efficient Bayesian Phylogenetic Inference and Model Choice Across a 585 Large Model Space. Syst Biol 61, 539–542 (2012). 586 47. Gelman, A. & Rubin, D. B. Inference from Iterative Simulation Using Multiple Sequences. Statist Sci 587 7, 457–472 (1992). 588 589 .CC-BY-NC-ND 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted January 16, 2026. ; https://doi.org/10.64898/2026.01.16.699827doi: bioRxiv preprint