Arabinogalactan proteins are functionally diverse cell wall structural glycoproteins that have been implicated in cell wall remodeling, although the mechanistic actions remain elusive. Here, we identify and characterize two AGP glycoproteins, SLEEPING BEAUTY (SB) and SB-like (SBL), that negatively regulate the gametophore bud initiation in Physcomitrium patens by dampening cell wall loosening/softening. Disruption of SB and SBL led to accelerated gametophore formation and altered cell wall compositions. The function of SB is glycosylation dependent and genetically connected with the class C auxin response factor (ARF) transcription factors PpARFC1B and PpARFC2. Transcriptomics profiling showed that SB upregulates PpARFC2, which in turn suppresses a range of cell wall-modifying genes that are required for cell wall loosening/softening. We further show that PpARFC2 binds directly to multiple AuxRE motifs on the cis-regulatory sequences of PECTIN METHYLESTERASE to suppress its expression. Hence, our results demonstrate a mechanism by which the SB modulates the strength of intracellular auxin signaling output, which is necessary to fine-tune the timing of gametophore initials formation.

Unlike animal cells, plant cells are encased in rigid cell walls that provide mechanical support, but also restrict cell growth and expansion, which are essential for morphogenesis. To negate this problem, cell walls are constantly being remodeled through loosening and reconstruction during plant developmental to confer the much-needed extensibility and plasticity (Anderson and Kieber, 2020). Primary cell walls in plants are composed of three polysaccharides: cellulose, hemicellulose and pectin. Although how these polysaccharides are assembled is still being discussed, it is commonly accepted that cellulose and hemicellulose crosslink to form a network that serves as a scaffold whereas pectin functions as a matrix plasticizer (Wolf et al., 2012). During cell wall loosening, the non-enzymatic α-expansin and hydrolytic endoglucanase Cel12A disintegrate the crosslinks in cellulose-hemicellulose complexes to reduce the elastic and plastic moduli, thereby relaxing structural stress and resulting in a mechanically weaker cell wall (Cosgrove, 2000; McQueen-Mason et al., 1992; Yuan et al., 2001). Conversely, de-esterification of homogalacturonan (HG) pectin by pectin methylesterase (PME) softens the cell wall and increases plasticity (Wang et al., 2020).

Another important component in the cell wall matrix is arabinogalactan proteins (AGPs), a class of glycoprotein that has been implicated in numerous functions throughout plant growth and development (Ellis et al., 2010; Ma et al., 2018; Pereira et al., 2016). Structurally, AGPs belong to the Hyp-rich glycoprotein superfamily and possess distinctive glycosylation motifs with characteristic carbohydrate moieties (Seifert and Roberts, 2007). The extracellular-localized AGPs are characterized by a secretion signal peptide at the N terminus and multiple arabinogalactan glycosylation motifs; often AGPs are tethered to the outer leaflet of plasma membrane by a glycosylphosphatidylinositol (GPI) lipid anchor. Despite the diverse roles AGP has been implicated in, there is a consensus idea that AGPs may play important roles in signaling, an idea that is consistent with the great heterogeneity seen in their polysaccharide substituents. Specifically, AGP has been found to have a binding capacity for calcium in a pH-dependent manner (Lamport and Várnai, 2013). In the Arabidopsis glcat14 mutant, which is compromised in AG-glucoronyltransferase activities and hence displays a strong reduction in glucuronidation, AG calcium-binding capacity is also lowered (Lopez-Hernandez et al., 2020). This indicates that the β-linked glucuronic acid (GlcA) residues in the AG are required for calcium binding and that AGP can act as a Ca2+ capacitor to bind and release apoplastic calcium during the regulation of various developmental processes (Lamport et al., 2014). For example, during extension growth, such as pollen tube elongation and phototropism, which requires dynamic cell wall remodeling, the AGP-Ca2+ can create a periplasmic Ca2+ reservoir that feeds the calcium channel and serves as a major supplier for cytosolic Ca2+ oscillation, which is crucial to incite calcium signaling (Lamport et al., 2014). It has been hypothesized that the AGP-Ca2+ is an integral component of the acid growth model. In this model, the phytohormone auxin rapidly increases the expression of SMALL AUXIN UP-RNA (SAUR), which inhibits PP2C-D phosphatase activities. This decrease in the PP2C-D activity shifts the plasma membrane H+-ATPases toward the C-terminally phosphorylated active state to pump out protons that acidify the apoplastic region and promote cell expansion (Spartz et al., 2014; Takahashi et al., 2012). The acidic environment will aid the dissociation of the carboxylate-bound Ca2+ from AGP-Ca2+, allowing Ca2+ to enter the cytosol via plasma membrane Ca2+ channels to increase Ca2+ abundance.

The renewed acid growth model implies that AGPs play regulatory roles in wall loosening/cell expansion and is consistent with the direct and circumstantial evidence that has been extensively documented (Ding and Zhu, 1997; Lee et al., 2005; Šamaj et al., 1999). First, the backbone proteins of AGPs are post-translationally modified by prolyl-4-hydroxylases (P4Hs) at proline residues before arabinose and/or galactose can be added by O-linked glycosylation (Schultz et al., 2004). Suppressing P4H activities by either pharmacological treatment or genetic perturbation has been shown to lead to retarded root hair growth in Arabidopsis but larger/longer organs in tomato, indicating that sugar moieties on the AGPs are necessary to control plant growth (Velasquez et al., 2011). Second, an Arabidopsis AGP can covalently crosslink to hemicellulose and pectin to form a structure called Arabinoxylan Pectin Arabinogalactan Protein 1 (APAP1), therefore serving as a pectic plasticizer to loosen the pectic network during cell wall remodeling (Leszczuk et al., 2020; Tan et al., 2013). Finally, AGPs have the tendency to self-associate and form agglomerates with adhesive properties (Zhou et al., 2014). Although the renewed acid growth model offers a mechanistic view on the regulation of cell wall loosening, evidence to link auxin and AGP is still lacking and it remains poorly defined how the loosening process is dampened upon reaching an optimal degree of loosening. Here, we show that AGP-like glycoproteins from the early land plant Physcomitrium (formerly Physcomitrella) patens can modulate the strength of auxin signaling output through the class C auxin response factors (ARFs) PpARFC2 and PpARFC1B, which are involved in cell wall loosening/softening during gametophore initials formation. Hence, our finding provides a link between the plasma membrane-localized glycoproteins and intracellular auxin signaling that fine-tune the cell wall remodeling processes during gametophore initiation.

SLEEPING BEAUTY (SB) is a plasma membrane-localized AGP-like glycoprotein

In an attempt to identify novel regulators for cell polarity, we employed a high-quality, full-length cDNA library from which ∼3000 cDNA were transiently overexpressed one by one as untagged proteins in protoplasts to identify gain-of-function mutants that show abnormality in cell polarity during protoplast regeneration. Three cDNA clones, pphn37a15, pphb11i16 and pphn1p04, inhibited protoplast regeneration and arrested the protoplast in a dormant brood cell-like state. Although pphn37a15 and pphb11i16 encode for an ABI3-like transcription factor and a glycosyl transferase, respectively, no obvious homologs from other model plants were found for pphn1p04. We decided to focus on characterizing pphn1p04 and renamed the gene as Sleeping Beauty (PpSB, Pp3c11_14536) on the basis of its overexpressor dormant phenotype (Fig. 1A-C). In order to infer the function of PpSB, we mined various genome databases but found no obvious homologs in other plant lineages, except two homologs in the ceratodon moss (Ceratodon purpureus) and a putative glycoprotein from the pufferfish (Takifugu bimaculatus) that showed 61% identity with PpSB (Fig. 1E). The P. patens genome, however, encodes a putative PpSB homolog, which we named Sleeping Beauty-Like (PpSBL, Pp3c7_8890). Both PpSB and PpSBL (hereafter SB and SBL, respectively) were predicted to contain a signal peptide at the N terminus, numerous putative O-glycosylation sites (indicated by numbers in Fig. 1D) and a transmembrane domain near the C terminal (Fig. 1D). As the protein domain arrangement in SB and SBL greatly resembles that of a classical AGP backbone that is characterized by an over-representation of proline, alanine, serine and threonine, we suspect that SB and SBL could be AGP-like glycoproteins. However, mining the P. patens proteome with SB peptide sequences did not return any classical AGPs (AGP1, Pp3c4_6620; AGP2, Pp3c14_2420; AGP3, Pp3c12_5250; AGP4, Pp3c3_24480; AGP5, Pp3c18_17790) or Fasciclin-like AGPs (FLA1, Pp3c21_10620; FLA2, Pp3c13_8280) that have been described previously (Lee et al., 2005). This is likely a consequence of the intrinsic feature in AGPs, which are characterized by short peptide sequences and low sequence complexity (i.e. lack of clear protein domains). We therefore constructed a phylogenetic tree using known P. patens AGPs and showed that SB and SBL form a clade with FLA2 that represents a major AGP subclass (Fig. 1F), hence inferring SB and SBL as AGP-like glycoproteins.

Fig. 1.

SB is a plasma membrane-localized, AGP-like glycoprotein. (A,B) Protoplasts that inducibly overexpressed an SB coding sequence were incubated on PRM/B supplemented with either DMSO (A) or 1 µM β-estradiol (B). Scale bars: 50 µm. (C) Bar chart summarizing normalized transcript abundance [2^−ΔCT] of SB in iOX:SB grown on BCDAT medium supplemented with DMSO- or 1 µM β-estradiol (β-est). Mean values from three biological replicates are shown. Error bars represent s.d. (D) Schematics of SB and SBL. Both proteins are characterized by an N-terminus signal peptide (SP) and a transmembrane domain (TMD). Numbers indicate putative O-glycosylation sites. (E) Multiple sequence alignment of the most conserved domains between SB, SBL and their homologs in Takifugu bimaculatus (Fugu) and Ceratodon purpureus (Cepur). Amino acid colors are based on ClustalX default coloring. (F) A maximum likelihood phylogenetic tree that depicts the grouping of SB and its homologs from P. patens (Pp), Takifugu bimaculatus (Fugu) and Ceratodon purpureus (Cepur). Numbers indicate bootstrap values and only values above 60 are shown. Branches are colored based on the bootstrap values with color scheme shown on the right. The scale bar represents amino acid substitution rate. (G) Fractionation analysis of SB-Dendra2. Top panel is a Dendra immunoblot (IB) and bottom panel is Coomassie Brilliant Blue (CBB) staining of the corresponding blot. S100 and P100 indicate soluble and microsomal fractions, respectively. Numbers on the left indicate molecular weight in kDa. (H) Dendra, JIM13 and LM6 immunoblots. Microsomal protein fractions of native promoter-driven SB-Dendra2 were analyzed on Dendra (left), JIM13 (middle) and LM6 (right) immunoblots. Numbers on the left indicate molecular weight in kDa. Bottom panel shows CBB staining of respective blots. (I) GFP immunoblots of Yariv-precipitated SBWT-Citrine and SBquintuple-Citrine (top). Protein input is shown in the lower panel. Numbers on the left indicate molecular weight in kDa. (J) Confocal laser-scanning microscopic analyses of protonema cells in a native promoter-driven SB-Dendra2 transgenic moss. Green and red represent Dendra2 and FM4-64 stains, respectively. Asterisks indicate autofluorescence from chloroplasts. Scale bars: 10 μm.

Fig. 1.

SB is a plasma membrane-localized, AGP-like glycoprotein. (A,B) Protoplasts that inducibly overexpressed an SB coding sequence were incubated on PRM/B supplemented with either DMSO (A) or 1 µM β-estradiol (B). Scale bars: 50 µm. (C) Bar chart summarizing normalized transcript abundance [2^−ΔCT] of SB in iOX:SB grown on BCDAT medium supplemented with DMSO- or 1 µM β-estradiol (β-est). Mean values from three biological replicates are shown. Error bars represent s.d. (D) Schematics of SB and SBL. Both proteins are characterized by an N-terminus signal peptide (SP) and a transmembrane domain (TMD). Numbers indicate putative O-glycosylation sites. (E) Multiple sequence alignment of the most conserved domains between SB, SBL and their homologs in Takifugu bimaculatus (Fugu) and Ceratodon purpureus (Cepur). Amino acid colors are based on ClustalX default coloring. (F) A maximum likelihood phylogenetic tree that depicts the grouping of SB and its homologs from P. patens (Pp), Takifugu bimaculatus (Fugu) and Ceratodon purpureus (Cepur). Numbers indicate bootstrap values and only values above 60 are shown. Branches are colored based on the bootstrap values with color scheme shown on the right. The scale bar represents amino acid substitution rate. (G) Fractionation analysis of SB-Dendra2. Top panel is a Dendra immunoblot (IB) and bottom panel is Coomassie Brilliant Blue (CBB) staining of the corresponding blot. S100 and P100 indicate soluble and microsomal fractions, respectively. Numbers on the left indicate molecular weight in kDa. (H) Dendra, JIM13 and LM6 immunoblots. Microsomal protein fractions of native promoter-driven SB-Dendra2 were analyzed on Dendra (left), JIM13 (middle) and LM6 (right) immunoblots. Numbers on the left indicate molecular weight in kDa. Bottom panel shows CBB staining of respective blots. (I) GFP immunoblots of Yariv-precipitated SBWT-Citrine and SBquintuple-Citrine (top). Protein input is shown in the lower panel. Numbers on the left indicate molecular weight in kDa. (J) Confocal laser-scanning microscopic analyses of protonema cells in a native promoter-driven SB-Dendra2 transgenic moss. Green and red represent Dendra2 and FM4-64 stains, respectively. Asterisks indicate autofluorescence from chloroplasts. Scale bars: 10 μm.

We next generated an in-frame Dendra2 knock-in fusion protein, SB-Dendra2, to investigate the localization and properties of SB. Fractionation analysis showed that the SB-Dendra2 was exclusively found in the crude membrane fraction (P100), consistent with the predicted transmembrane domain (Fig. 1G). Furthermore, SB-Dendra2 migrated at a molecular weight larger than the predicted 44.5 kDa (Fig. 1G), suggesting that SB is post-translationally modified. To verify this possibility, we probed the SB-Dendra2 with an established AGP-specific monoclonal antibody, JIM13, which recognizes the glycosyl chains of AGP (Motose et al., 2004; Yates et al., 1996). The P100 protein fraction of SB-Dendra2 was recognized by JIM13 at a molecular weight similar to that of SB-Dendra2 (Fig. 1H), providing evidence that SB is an O-linked glycosylated protein. This reactivity was not observed when a pectin antibody LM6-M was used (Fig. 1H). Furthermore, a Citrine-tagged SB (SB-Citrine) was efficiently precipitated by β-glucosyl Yariv, which is known to bind strongly to AGP (Kitazawa et al., 2013; Lee et al., 2005; Yu and Zhao, 2012) (Fig. 1I); by contrast, an SB variant with five mutated putative glycosylation sites (SBquintuple-Citrine; also see below) was poorly precipitated (Fig. 1I), providing further evidence that SB is an AGP-like protein. We next examined the subcellular localization of SB-Dendra2 and detected SB-Dendra2 on the plasma membrane of protoplasts (Fig. S1A) as well as protonema (Fig. 1J). The SB-Dendra2 also colocalized with FM4-64, a lipophilic probe that stains plasma membrane (Fig. 1J). Collectively, our results demonstrate that SB is a plasma membrane-localized AGP-like glycoprotein.

SB and SBL negatively regulate gametophore initiation

To investigate the function of SB and SBL, we generated single and double disruptants for SB and SBL using homologous recombinant gene replacement approach (Fig. S2A-E). The sb and sbl single disruptants as well as sb sbl double knockout mutant were phenotypically indistinguishable from the wild type (WT) at the protonema stage (Fig. S3A-H). Notably, formation of leafy gametophores was accelerated in the sb sbl double mutant but not in the sb or sbl mutants, although gametophore numbers were reduced in sbl (Fig. 2A-D, Fig. S3I). Quantification of leafy gametophore numbers in BCD medium-inoculated colonies showed that leafy gametophores were visible in sb sbl at least 3 days before the WT, indicating accelerated gametophore formation when SB and SBL were compromised (Fig. 2E). Interestingly, the accelerated gametophore formation in sb sbl did not result in an overall greater number of leafy gametophores (Fig. 2E), suggesting that the accelerated gametophore formation phenotype cannot be ascribed to increased gametophore initial stem cells. Because the chloronema-to-caulonema transition precedes gametophore initiation, we also examined this transitional stage in WT and sb sbl and ruled out the possibility that the accelerated gametophore formation is a consequence of quicker caulonema differentiation (Fig. S3J). Consistent with the knockout phenotype in sb sbl, gametophore formation was severely impaired when SB or SBL was overexpressed in a β-estradiol-inducible manner at different levels to enable phenotypic characterization at a specific developmental stage (Fig 2F-K and 1C, Fig. S3K). Gametophore formation is a multistep process that is preceded by formative divisions, which give rise to a juvenile bud with a three-faced apical cell, followed by its subsequent bushy growth to form the adult leaf-like gametophore (Harrison et al., 2009). We followed the development of individual gametophores in WT and sb sbl (Fig. 2L-S) to pinpoint which developmental step was affected. We found no pronounced difference in the formative divisions of bud initials to attain the tetrahedral shape (Fig. 2L,P). Similarly, bushy growth and timing of leaf initiation were also indistinguishable (Fig. 2M,N,Q,R). Although leaves in the sb sbl gametophores were noticeably smaller compared with WT (Fig. 2O,S), this by no means explains the accelerated gametophore formation seen in sb sbl (Fig. 2B). We suspected that the timing at which the gametophore initials emerge could be the underlying factor. As such, we observed protonema tissues of WT and sb sbl grown in BCD medium and scored the time point at which their first juvenile bud with a three-faced apical cell emerged (Harrison et al., 2009). In WT, the juvenile bud was formed between the 6th and 7th day post-inoculation (dpi) whereas the majority of sb sbl produced their first juvenile bud initial at 5 dpi, with some as early as 4 dpi (Fig. 2T). This suggests that bud initial emergence from the caulonemata was significantly accelerated in sb sbl. Hence, SB and SBL redundantly and specifically act to delay the timing of juvenile bud initiation.

Fig. 2.

SB and SBL inhibit gametophore initiation and regulate cell wall composition. (A-D) Brightfield images of 10-day-old WT (A,C) and sb sbl (B,D) colonies grown on BCD medium. Higher magnifications of A and B are shown in C and D, respectively. Scale bars: 1 mm (A,B); 500 µm (C,D). Arrowheads indicate gametophores. (E) Line chart indicating the number of gametophores per colony in WT (squares) and sb sbl (circles) after inoculation in BCD medium. n indicates the number of colonies scored. **P≤0.001, ***P<0.001 (two-tailed, unpaired Student's t-test). (F-K) Brightfield images of 12-day-old WT (F,G), iOX:SB (H,I) and iOX:SBL (J,K) colonies inoculated in BCD medium supplemented with either DMSO (F,H,J) or 1 µM β-estradiol (G,I,K). Scale bars: 1 mm. Arrowheads indicate gametophores. (L-S) Time-lapse images of gametophore development in WT (L-O) and sb sbl (P-S) from tetrahedral shape (L,P), bushy growth (M,Q) to leaf initiation (N,R) and growth (O,S). Scale bars: 100 µm. (T) Bar chart summarizing the timing at which the first juvenile bud emerges from WT (white bars) and sb sbl (gray bars). Solid and dashed lines are two-point moving average trend lines for sb sbl and WT, respectively. (U) Violin plot indicating the mole percentage (Mol%) of galactose in WT (white polygon) and sb sbl (yellow polygon) protonemata cell walls. White circles indicate the medians; box limits indicate the 25th and 75th percentiles; polygons represent density estimates. *P≤0.05 (two-tailed, unequal variance Student's t-test). n indicates the number of samples analyzed.

Fig. 2.

SB and SBL inhibit gametophore initiation and regulate cell wall composition. (A-D) Brightfield images of 10-day-old WT (A,C) and sb sbl (B,D) colonies grown on BCD medium. Higher magnifications of A and B are shown in C and D, respectively. Scale bars: 1 mm (A,B); 500 µm (C,D). Arrowheads indicate gametophores. (E) Line chart indicating the number of gametophores per colony in WT (squares) and sb sbl (circles) after inoculation in BCD medium. n indicates the number of colonies scored. **P≤0.001, ***P<0.001 (two-tailed, unpaired Student's t-test). (F-K) Brightfield images of 12-day-old WT (F,G), iOX:SB (H,I) and iOX:SBL (J,K) colonies inoculated in BCD medium supplemented with either DMSO (F,H,J) or 1 µM β-estradiol (G,I,K). Scale bars: 1 mm. Arrowheads indicate gametophores. (L-S) Time-lapse images of gametophore development in WT (L-O) and sb sbl (P-S) from tetrahedral shape (L,P), bushy growth (M,Q) to leaf initiation (N,R) and growth (O,S). Scale bars: 100 µm. (T) Bar chart summarizing the timing at which the first juvenile bud emerges from WT (white bars) and sb sbl (gray bars). Solid and dashed lines are two-point moving average trend lines for sb sbl and WT, respectively. (U) Violin plot indicating the mole percentage (Mol%) of galactose in WT (white polygon) and sb sbl (yellow polygon) protonemata cell walls. White circles indicate the medians; box limits indicate the 25th and 75th percentiles; polygons represent density estimates. *P≤0.05 (two-tailed, unequal variance Student's t-test). n indicates the number of samples analyzed.

As AGPs are important structural proteins in the cell wall, we further investigated how disruption of two AGP-like glycoproteins may affect the cell wall compositions by analyzing sb sbl cell wall monosaccharides. Quantification of sb sbl cell wall monosaccharides showed that only galactose is significantly lowered (mole percentage=15.63±0.929) compared with WT (mole percentage=18.20±1.947; Fig. 2U), whereas other monosaccharides remained unaffected (Fig. S4). As galactose is a predominant monosaccharide in the side chains of Rhamnogalacturonan I, a domain of pectin, SB and SBL may redundantly regulate pectin biosynthesis/modeling during cell wall remodeling to repress juvenile bud initiation.

O-linked glycosylations and plasma membrane localization are necessary for SB functions

A key feature of the AGP backbone is the presence of repetitive Ala-Pro, Ser-Pro, Thr-Pro and Val-Pro dipeptide motifs. The Pro residues in these motifs are hydroxylated by P4H before O-linked glycosylation by glycosyltransferases and subsequent additions of type II arabinogalactan to the Hyp residues can take place. The SB protein backbone contains six Ala-Pro (the 85th, 93rd, 95th, 97th, 100th and 108th amino acids), two Thr-Pro (57th and 111th amino acids) and two Val-Pro (54th and 64th amino acids) dipeptide motifs (Fig. 3A). In order to investigate the functional role of these putative O-linked glycosylation sites, we replaced the proline residues with the nonpolar amino acid alanine to abolish the glycosylation sites (crosses in Fig. 3A) and generated a transgenic line, iOX:SBquintuple, that inducibly expressed the SBPro58,86,96,109,112Ala variant. Using the phenotype in iOX:SB overexpressor as a readout of SB function, we examined gametophore initiation in iOX:SBquintuple. Overexpression of WT SB upon 1 µM β-estradiol treatment completely inhibited gametophore initiation (Fig. 3B-D). However, this inhibitory effect was lost in the iOX:SBquintuple variant overexpressor (Fig. 3E-G), indicating that glycosylation sites and the O-linked glycosylations thereof are required for SB to exert this inhibitory effect. Interestingly, SBquintuple remained correctly targeted, as revealed by the plasma membrane localization of SBquintuple-Citrine (Fig. 3H,I), suggesting that carbohydrate moieties in the SB are not required for its secretion but instead may play some functional roles on the plasma membrane. Next, we attempted to validate the functionality of SB on the plasma membrane by fusing an endoplasmic reticulum (ER)-retrieval signal HDEL to the C terminus of SB. Addition of the HDEL signal is known to retain secreted proteins effectively in the ER (Batoko et al., 2000; Gomord et al., 1997). In a transgenic line that inducibly overexpressed SBHDEL, iOX:SBHDEL, the inhibitory effect of gametophore initiation was diminished (Fig. 3J-L), confirming the necessity of plasma membrane targeting for SB to exert any functional roles. Taken together, our results imply that glycosylations on SB and its secretion to the plasma membrane are required for SB functionality.

Fig. 3.

SB functions on the plasma membrane in a glycosylation-dependent manner. (A) Schematic of proline dipeptide motifs that could be hydroxylated by P4H. Crosses in parentheses indicate mutated proline residues. (B-D) Fifteen-day-old colonies of iOX:SB grown on BCD medium supplemented with DMSO (B) or 1 μM β-estradiol (C). Arrowheads indicate gametophores. Scale bars: 1 mm. The number of gametophores in colonies grown in the presence of DMSO (circles) or 1 μM β-estradiol (diamonds) are summarized in D. n indicates the number of colonies scored. ***P<0.001 (two-tailed, unpaired Student's t-test). (E-G) Fifteen-day-old colonies of iOX:SBquintuple grown on BCD medium supplemented with DMSO (E) or 1 μM β-estradiol (F). Arrowheads indicate gametophores. Scale bars: 1 mm. The number of gametophores in colonies grown in the presence of DMSO (circles) or 1 μM β-estradiol (diamonds) summarized in G. n indicates the number of colonies scored. No significant difference was detected by two-tailed, unpaired Student's t-test (P>0.001). (H,I) Confocal laser-scanning microscopic analyses of iOX:SB-Citrine (H) and iOX:SBquintuple-Citrine (I) after 16 h of 1 μM β-estradiol induction. Images are z-stack projection (maximum intensity). Green and magenta represent Citrine and chloroplast autofluorescence, respectively. Scale bars: 10 μm. (J-L) Fifteen-day-old colonies of iOX:SBHDEL grown on BCD medium supplemented with DMSO (J) or 1 μM β-estradiol (K). Scale bars: 10 μm. The number of gametophores in colonies grown in the presence of DMSO (circles) or 1 μM β-estradiol diamonds are summarized in L. n indicates the number of colonies scored. No significant difference was detected by two-tailed, unpaired Student's t-test (P>0.001).

Fig. 3.

SB functions on the plasma membrane in a glycosylation-dependent manner. (A) Schematic of proline dipeptide motifs that could be hydroxylated by P4H. Crosses in parentheses indicate mutated proline residues. (B-D) Fifteen-day-old colonies of iOX:SB grown on BCD medium supplemented with DMSO (B) or 1 μM β-estradiol (C). Arrowheads indicate gametophores. Scale bars: 1 mm. The number of gametophores in colonies grown in the presence of DMSO (circles) or 1 μM β-estradiol (diamonds) are summarized in D. n indicates the number of colonies scored. ***P<0.001 (two-tailed, unpaired Student's t-test). (E-G) Fifteen-day-old colonies of iOX:SBquintuple grown on BCD medium supplemented with DMSO (E) or 1 μM β-estradiol (F). Arrowheads indicate gametophores. Scale bars: 1 mm. The number of gametophores in colonies grown in the presence of DMSO (circles) or 1 μM β-estradiol (diamonds) summarized in G. n indicates the number of colonies scored. No significant difference was detected by two-tailed, unpaired Student's t-test (P>0.001). (H,I) Confocal laser-scanning microscopic analyses of iOX:SB-Citrine (H) and iOX:SBquintuple-Citrine (I) after 16 h of 1 μM β-estradiol induction. Images are z-stack projection (maximum intensity). Green and magenta represent Citrine and chloroplast autofluorescence, respectively. Scale bars: 10 μm. (J-L) Fifteen-day-old colonies of iOX:SBHDEL grown on BCD medium supplemented with DMSO (J) or 1 μM β-estradiol (K). Scale bars: 10 μm. The number of gametophores in colonies grown in the presence of DMSO (circles) or 1 μM β-estradiol diamonds are summarized in L. n indicates the number of colonies scored. No significant difference was detected by two-tailed, unpaired Student's t-test (P>0.001).

SB function is associated with PpARFC2

The accelerated gametophore formation and reduced galactose level in the sb sbl cell wall composition provides a link between cell wall remodeling and gametophore initiation. Indeed, various gametophore developmental defects have been observed in cell wall biosynthesis mutants that are defective in Cellulose Synthase (CESA) and Hydroxyproline O-arabinosyltransferases (HPAT) (Goss et al., 2012; MacAlister et al., 2016). Given that side chains in Rhamnogalacturonan I of pectin are heavily decorated by galactose, we posit that SB and SBL may regulate pectin compositional changes during cell wall remodeling to negatively control gametophore initiation. Because the phytohormone auxin plays a pivotal role in cell wall loosening, an important aspect of cell wall remodeling, we attempted to establish a connection between SB and auxin signaling in the control of cell wall remodeling and gametophore initiation. This postulation is supported by the fact that gametophore formation is inhibited in tissues treated with the exogenous auxin analog 1-naphthaleneacetic acid (NAA) and in a mutant that completely lacks auxin/indole-3-acetic acid proteins (AUX/IAAs), repressor proteins that inhibit auxin-specific transcription factors [AUXIN RESPONSE FACTOR (ARF) proteins] (Lavy et al., 2016).

Some ARFs are constitutively repressed by the AUX/IAA co-repressors when active auxin is scarce. A higher active auxin level will trigger the proteasome-mediated degradation of AUX/IAA and de-represses the ARFs to initiate transcription of target genes (Lavy and Estelle, 2016). To understand how SB may affect the auxin response, we next quantified the transcript abundance of ARFs in sb sbl and iOX:SB. The P. patens genome encodes 17 ARFs that can be grouped into three main classes (A-C, which are shared between all land plants) and an additional class D that lacks the DNA-binding domain (Lavy et al., 2016). We chose to quantify the transcript abundance of one or two selected ARFs from each main class in WT, sb sbl and iOX:SB after 16 h of 10 µM NAA treatment. Whereas PpARFA3, PpARFA7, PpARFA8 and PpARFB1 did not exhibit any discernible changes in their expression patterns (Fig. S5A-D), PpARFC2 was upregulated in iOX:SB in a dosage-dependent manner (Fig. 4A). By manipulating the SB expression level using different concentrations of β-estradiol, we found that transcript abundance of PpARFC2 exhibited positive correlations with the SB protein abundance (Fig. 4A, Fig. S5E). This raises the possibility that abundance of the plasma membrane-localized SB could activate an auxin signaling pathway that is mediated by PpARFC2. It is unclear why the PpARFC2 transcript in sb sbl was not further downregulated compared with WT (Fig. 4A), but we suspect this could be an issue of detection limit as endogenous PpARFC2 expression was low. We decided to characterize further the thus-far-unknown roles of PpARFC2 in cell wall remodeling.

Fig. 4.

SB function is associated with PpARFC2. (A) Bar chart that summarizing transcript abundance [log2(fold change)] of PpARFC2 in WT, sb sbl and iOX:SB that was induced by 0.25 pM or 1 μM β-estradiol after 16 h of 10 µM NAA treatment. Log2(fold change) shown is relative to mock-treated (no NAA) samples of respective genotypes. Mean values from three biological replicates are shown. Error bars represent s.d. *P<0.05, **P<0.01 (two-tailed, unequal variance Student's t-test). n.s., not significant. (B) Nine-day-old colonies of WT (left) and pparfc2 (right) inoculated in BCD medium. Arrowheads indicate gametophores. Scale bars: 1 mm. (C) Line chart summarizing the number of gametophores per colony in WT (squares) and pparfc2 (triangles). Error bars represent s.d. n indicates the number of colonies scored. ***P<0.001 (two-tailed, unpaired Student's t-test). n.s., not significant. (D,E) Eleven-day-old colonies (D) and 3-day-old protoplasts (E) of iOX:PpARFC2 grown on either BCD (D) or PRM/B (E) medium supplemented with DMSO (left) or 1 μM β-estradiol (right). Arrowheads indicate gametophores. Scale bars: 1 mm (D); 50 µm (E). (F) Violin plot summarizing the D-Galactose composition (mole percentage) in protonemata cell walls of WT (white), pparfc2 (blue), and iOX:PpARFC2 treated with DMSO (pink) or 1 μM β-estradiol (red). White circles show the medians; box limits indicate the 25th and 75th percentiles; polygons represent density estimates. Three independent samples for each genotype were analyzed. *P<0.05 (one-tailed, equal variance Student's t-test).

Fig. 4.

SB function is associated with PpARFC2. (A) Bar chart that summarizing transcript abundance [log2(fold change)] of PpARFC2 in WT, sb sbl and iOX:SB that was induced by 0.25 pM or 1 μM β-estradiol after 16 h of 10 µM NAA treatment. Log2(fold change) shown is relative to mock-treated (no NAA) samples of respective genotypes. Mean values from three biological replicates are shown. Error bars represent s.d. *P<0.05, **P<0.01 (two-tailed, unequal variance Student's t-test). n.s., not significant. (B) Nine-day-old colonies of WT (left) and pparfc2 (right) inoculated in BCD medium. Arrowheads indicate gametophores. Scale bars: 1 mm. (C) Line chart summarizing the number of gametophores per colony in WT (squares) and pparfc2 (triangles). Error bars represent s.d. n indicates the number of colonies scored. ***P<0.001 (two-tailed, unpaired Student's t-test). n.s., not significant. (D,E) Eleven-day-old colonies (D) and 3-day-old protoplasts (E) of iOX:PpARFC2 grown on either BCD (D) or PRM/B (E) medium supplemented with DMSO (left) or 1 μM β-estradiol (right). Arrowheads indicate gametophores. Scale bars: 1 mm (D); 50 µm (E). (F) Violin plot summarizing the D-Galactose composition (mole percentage) in protonemata cell walls of WT (white), pparfc2 (blue), and iOX:PpARFC2 treated with DMSO (pink) or 1 μM β-estradiol (red). White circles show the medians; box limits indicate the 25th and 75th percentiles; polygons represent density estimates. Three independent samples for each genotype were analyzed. *P<0.05 (one-tailed, equal variance Student's t-test).

Using the homologous recombination approach, we replaced the PpARFC2 locus with an NPTII gene to generate a pparfc2 knockout mutant (Fig. S6A,B). The pparfc2 mutant phenocopied sb sbl and exhibited accelerated gametophore initiation by at least 3 days compared with WT (Fig. 4B,C). The quicker gametophore initiation in pparfc2 is unlikely to have resulted from an accelerated caulonema transition as chloronema-to-caulonema differentiation in pparfc2 was similar to that in WT (Fig. S6D). Notably, overexpression of PpARFC2 also phenocopied the iOX:SB overexpressor (Figs 1A,B and 2H-I) and showed inhibition of gametophore initiation as well as repressed apical growth during protoplast regeneration (Fig. 4D,E). The failure of gametophore initiation in iOX:PpARFC2 may reflect an increase in cell wall stiffness that is pectin related. Indeed, the galactose level in the cell wall was also marginally reduced in the pparfc2 (mole percentage of 16.33±0.776 versus 18.2±1.95 in WT; Fig. 4F). Remarkably, unlike sb sbl, which only showed altered monosaccharide composition in galactose (Fig. 2U), cell wall compositions in pparfc2 exhibited a multitude of affected monosaccharide levels in arabinose, mannose and xylose (Fig. S7). This points to an important role of PpARFC2 in mediating cell wall remodeling to dampen cell loosening/softening. Unexpectedly, iOX:PpARFC2 also exhibited a low galactose level (11.63±0.586 mole percentage) that was further exacerbated (β-estradiol 9.36±1.443 mole percentage) after PpARFC2 overexpression (Fig. 4F), suggesting that reduction in the galactose level alone could not account for the phenotype observed. We suspect that the reduced galactose level in iOX:PpARFC2 could be a consequence of prolonged leakiness in PpARFC2 overexpression that triggers a compensatory mechanism in the cell wall to reduce it. Collectively, phenotypic similarities between knockout and overexpressor mutants of SB and PpARFC2 provide a genetic basis for their connection in playing common roles in a genetic pathway that regulates cell wall remodeling and subsequently gametophore initiation.

PpARFC2 represses cell wall remodeling-related genes

Thus far, our genetic evidence indicates an essential role for SB and PpARFC2 in affecting cell wall remodeling and composition that is crucial for gametophore initiation, possibly through negative regulation of cell wall loosening/softening. To understand better the transcriptional reprogramming upon PpARFC2 overexpression, we performed RNA-sequencing (RNA-seq) analyses using iOX:PpARFC2 to unravel the transcriptional targets of PpARFC2. Upon a 16 h β-estradiol induction, a time point before the overexpressor phenotype became apparent, a total of 2284 differentially expressed genes (DEGs) were identified between the DMSO- and β-estradiol-treated samples. Gene Ontology (GO) analyses of these DEGs showed that cell wall modification-related GO terms are highly enriched (FDR adjusted P≤0.005) in all three biological replicates (Fig. 5A,B). For example, the GO Biological Processes ‘cell wall organization/biogenesis’ (GO:0071554, GO:0071555, GO:0042545, GO:0071669), ‘pectin metabolic/catabolic processes’ (GO:0045490, GO:0045488) and ‘xyloglucan metabolic process’ (GO:0010411) were amongst the most significantly enriched (Fig. 5A,B, Fig. S8). In line with the transcription factor role of PpARFC2, the Molecular Functions ‘DNA binding transcription factor activity’ (GO:0003700) and ‘transcription regulator activity’ (GO:0140110) were also enriched, in addition to ‘pectin esterase activity’ (GO:0030599) and ‘xyloglucan:xyloglucasyl transferase activity’ (GO:0016762) (Fig. 5A,B). In summary, a large-scale downregulation of cell wall loosening and modification-associated genes, such as EXPANSIN, PECTIN METHYLESTERASES (PME), COBRA and xyloglucan-modifying enzymes, was observed (Fig. 5C), suggesting that PpARFC2 functions as a repressor-ARF in cell wall remodeling. Of particular interest is the pectin methylesterases (PMEs), which are encoded by at least 42 loci in the P. patens genome. A short induction (16 h) of PpARFC2 overexpression led to at least nine PME genes being repressed in all three biological replicates (Fig. 5C), with an additional three PME genes downregulated in one of the biological replicates. Misregulation of multiple PME genes may account for the iOX:PpARFC2 phenotype in which the gametophore initiation and apical growth was affected (Fig. 4D,E), likely arising from cell wall stiffening. Depending on the mode of enzymatic actions, PME de-methylesterifies the methylated carboxyl group of pectin to allow pectin degradation by endo-polygalacturonase or pectate lyase, hence rendering a more fluidic homogalacturonan for tissue softening (Micheli, 2001). By contrast, the de-methylesterified pectin can also crosslink with Ca2+, leading to rigidification of cell walls to suppress growth (Willats et al., 2001). These putative transcriptional targets argue for a role of PpARFC2 in suppressing cell wall loosening/softening and is in line with the phenotype observed in iOX:PpARFC2 in which gametophore initiation was severely inhibited (Fig. 4D).

Fig. 5.

PpARFC2 transcriptionally represses cell wall remodeling genes. (A) Bubble plot depicting enriched GO terms (FDR-adjusted P≤0.005) that are related to biological processes (blue), cellular compartments (beige) and molecular functions (green) after 16 h of PpARFC2 overexpression. Terms are shown as −log (adj P-value) (y-axis) and z-scores (x-axis) that predict the bias in gene regulation. The size of the bubbles is proportional to the number of DEGs. Numbers in bubbles are GO identifiers that correspond to the table in B. Results shown are from biological replicate #2. (B) Table summarizing significantly enriched GO terms (FDR-adjusted P≤0.005) plotted in the bubble plot in A. Colors correspond to those in A. (C) Heatmap showing cluster analysis of DEGs (−1≤log2 Fold change≥1) in the GO term ‘cell wall organization or biogenesis’ (GO: 0071554). Gene identifiers are listed on the right and colored as blue to indicate pectin esterases, red to indicate expansins, violet to indicate COBRAs and green to indicate xyloglucan:xyloglucasyl transferases.

Fig. 5.

PpARFC2 transcriptionally represses cell wall remodeling genes. (A) Bubble plot depicting enriched GO terms (FDR-adjusted P≤0.005) that are related to biological processes (blue), cellular compartments (beige) and molecular functions (green) after 16 h of PpARFC2 overexpression. Terms are shown as −log (adj P-value) (y-axis) and z-scores (x-axis) that predict the bias in gene regulation. The size of the bubbles is proportional to the number of DEGs. Numbers in bubbles are GO identifiers that correspond to the table in B. Results shown are from biological replicate #2. (B) Table summarizing significantly enriched GO terms (FDR-adjusted P≤0.005) plotted in the bubble plot in A. Colors correspond to those in A. (C) Heatmap showing cluster analysis of DEGs (−1≤log2 Fold change≥1) in the GO term ‘cell wall organization or biogenesis’ (GO: 0071554). Gene identifiers are listed on the right and colored as blue to indicate pectin esterases, red to indicate expansins, violet to indicate COBRAs and green to indicate xyloglucan:xyloglucasyl transferases.

PpARFC2 binds to the promoter of PMEs to repress their expressions

Molecular, genetic and biochemical studies have revealed binding motifs on the promoter region to which ARFs bind. These DNA target sequences are termed auxin response cis-elements (AuxREs) and are typified by the hexanucleotide motif 5′-TGTCNN-3′, with different ARFs showing variable affinities (Boer et al., 2014; Franco-Zorrilla et al., 2014; Freire-Rios et al., 2020). Our RNA-seq experiment indicates that PpARFC2 may suppress the transcription of cell wall remodeling-related genes by binding to their promoter regions. We explored this possibility by examining the presence of hexanucleotide AuxREs 5′-TGTCGG-3′ and 5′-TGTCTC-3′ in the 2-kb promoter regions of all putative transcriptional targets (Boer et al., 2014; Franco-Zorrilla et al., 2014). Of the 48 putative transcriptional targets (Fig. 5C) that were misregulated in iOX:PpARFC2, one PME (Pp3c5_12660) promoter region contains three 5′-TGTCGG-3′ and two 5′-TGTCTC-3′ motifs, fitting the required presence of multiple AuxREs for ARF effective binding (Table S1) (Freire-Rios et al., 2020). In order to demonstrate the binding between PpARFC2 and the promoter region of this targeted PME (PMEpro), we generated C- and N-terminal yellow fluorescent protein (YFP)-tagged fusion proteins for PpARFC2, driven by an Arabidopsis UBQ10 promoter (Grefen et al., 2010). Whereas PpARFC2-YFP yielded no fluorescence, YFP-PpARFC2 was localized to the nucleus when transiently expressed in P. patens protoplasts (Fig. 6A,B), supporting the notion that PpARFC2 is a transcription factor. By co-expressing YFP-PpARFC2 and a PMEpro:firefly luciferase (FLuc) fusion construct, we performed a promoter transactivation assay in tobacco leaves. FLuc activity was normalized using a co-infiltrated Renilla luciferase (RLuc) construct. We detected FLuc activity when PMEpro:FLuc was expressed alone, likely reflecting basal expression that was driven by endogenous ARFs in the tobacco leaves (Fig. 6C). When co-expressed with YFP-PpARFC2, we observed a significant reduction of FLuc activity (Fig. 6C). This is consistent with the RNA-seq data (Fig. 5C) and implies that PpARFC2 represses the PME expression and may directly bind to the PMEpro region. To validate the PpARFC2-PMEpro binding, we expressed a maltose-binding protein (MBP)-tagged recombinant protein of the PpARFC2 B3 DNA-binding domain (MBP-B3) and tested for its effect on the mobility retardation of biotinylated DNA probes that are specific to the ten AuxREs-like motifs (5′-TGTCNN-3′) found in PMEpro (Fig. 6D). We found that MBP-B3 caused mobility retardation of DNA probes for motifs 3 (5′-TGTCGT-3′), 4 (5′-TGTCAA-3′) and 10 (5′-TGTCTC-3′) (Fig. 6E, Fig. S9A), confirming that PpARFC2 directly binds to these motifs on the PME regulatory sequences to suppress its expression during cell wall remodeling to affect gametophore morphogenesis.

Fig. 6.

PpARFC2 binds to the promoter regions of PME to repress its expression. (A,B) Confocal laser-scanning microscopic analyses of protoplasts transiently expressing YFP-PpARFC2. Fluorescent (A) and respective brightfield (B) images are z-stack projections (sum). Green and magenta represent YFP and chloroplast autofluorescence, respectively. Scale bars: 10 µm. (C) Bar chart summarizing the promoter transactivation activities of PMEpro:FLuc by YFP-PpARFC2. The F-Luc activities were normalized to RLuc expression and expressed as FLuc/R-Luc ratios on the x-axis. Results are from four biological replicates in two independent experiments. Error bars represent s.d. (D) Schematic of the PpPME genomic locus. Green and blue boxes represent 5′ UTRs and exons, respectively. AuxREs on the promoter regions are marked as Motif 1-10 with their respective nucleotide positions in the parentheses. Motifs highlighted in yellow are AuxREs located on the antisense strand. (E) Streptavidin immunoblots from EMSA in which interactions between the PpARFC2 DNA-binding domain (MBP-B3) and biotinylated DNA probes of three AuxRE motifs were shown. Unlabeled competitor probes were supplied in 800-fold molar excess. Asterisk indicates unspecific binding signals. (F) Schematic of the PpARFC2 genomic locus and cis-regulatory sequences used in transcriptional fusions. Green and blue boxes represent 5′ UTRs and exons, respectively. Numbers indicate the nucleotide positions with +1 detonating the transcription start site. (G-N) Confocal laser-scanning microscopic analyses of caulonemata (G) and various developmental stages – three-cell stage (H), tetrahedral shape stage (I,J) as well as leafy buds (K-N) – in PpARFC2pro:NLS-GFP-GUS. The GFP intensity is represented by a 16-color LUT, with calibration bars shown at top right of each image. Magenta represents propidium iodide stain. Images are z-stack projections (sum). Scale bars: 10 µm.

Fig. 6.

PpARFC2 binds to the promoter regions of PME to repress its expression. (A,B) Confocal laser-scanning microscopic analyses of protoplasts transiently expressing YFP-PpARFC2. Fluorescent (A) and respective brightfield (B) images are z-stack projections (sum). Green and magenta represent YFP and chloroplast autofluorescence, respectively. Scale bars: 10 µm. (C) Bar chart summarizing the promoter transactivation activities of PMEpro:FLuc by YFP-PpARFC2. The F-Luc activities were normalized to RLuc expression and expressed as FLuc/R-Luc ratios on the x-axis. Results are from four biological replicates in two independent experiments. Error bars represent s.d. (D) Schematic of the PpPME genomic locus. Green and blue boxes represent 5′ UTRs and exons, respectively. AuxREs on the promoter regions are marked as Motif 1-10 with their respective nucleotide positions in the parentheses. Motifs highlighted in yellow are AuxREs located on the antisense strand. (E) Streptavidin immunoblots from EMSA in which interactions between the PpARFC2 DNA-binding domain (MBP-B3) and biotinylated DNA probes of three AuxRE motifs were shown. Unlabeled competitor probes were supplied in 800-fold molar excess. Asterisk indicates unspecific binding signals. (F) Schematic of the PpARFC2 genomic locus and cis-regulatory sequences used in transcriptional fusions. Green and blue boxes represent 5′ UTRs and exons, respectively. Numbers indicate the nucleotide positions with +1 detonating the transcription start site. (G-N) Confocal laser-scanning microscopic analyses of caulonemata (G) and various developmental stages – three-cell stage (H), tetrahedral shape stage (I,J) as well as leafy buds (K-N) – in PpARFC2pro:NLS-GFP-GUS. The GFP intensity is represented by a 16-color LUT, with calibration bars shown at top right of each image. Magenta represents propidium iodide stain. Images are z-stack projections (sum). Scale bars: 10 µm.

Next, we generated two PpARFC2 promoter fusion reporters that encompass the −360∼+3620 nt (PpARFC2pro:nls-GFP) and −3000∼−1 nt (PpARFC2proII:nls-GFP) regulatory sequences (Fig. 6F) to gain insight into the PpARFC2 expression pattern and how it contributes to gametophore morphogenesis. Whereas no detectable signal was observed in the PpARFC2proII:nls-GFP reporter line (Fig. S9B-E), PpARFC2pro:nls-GFP displayed unique spatial expression patterns in developing gametophores (Fig. 6G-N). Overall, PpARFC2 was expressed ubiquitously at protonema stage with no pronounced differences between caulonemata and chloronemata (Fig. 6G). At the three-cell stage of gametophore initials, in which the gametophore cell fate can be ascertained, PpARFC2 expression was not detected (Fig. 6H), implying that the absence of PpARFC2 may be a prerequisite for gametophore bud initiation. Thereafter, PpARFC2 exhibited differential spatial expressions in the gametophore bud tissues throughout its development, often showing the highest expression levels at the apical stem cells (Fig. 6I-N), reflecting its role in regulating the gametophore growth.

PpARFC2 and PpARFC1B functions overlap downstream of SB

Phenotypic similarities between sb sbl and pparfc2 suggest that SB and PpARFC2 may share a genetic pathway that regulates cell wall remodeling during gametophore morphogenesis. To determine whether there is epistasis between SB and PpARFC2, and whether PpARFC2 functions downstream of SB, we generated a pparfc2 disruptant in the iOX:SB background and investigated whether the iOX:SB overexpressor phenotype could be rescued. Contrary to our expectation, gametophore formation was not restored in iOX:SB/pparfc2 when SB overexpression was induced (Fig. S10A,B), implying that PpARFC2 function may overlap with an additional class C ARF, namely PpARFC1A and/or PpARFC1B, downstream of SB. We therefore overexpressed PpARFC1A and PpARFC1B to look for a similar gametophore inhibition phenotype. Although overexpression of PpARFC1B inhibited gametophore formation (Fig. 7D,H,L), similar inhibitory effects were not observed in an A-type PpARFA3 (Fig. 7A,E,I), a B-type PpARFB1 (Fig. 7B,F,J) or another C-type PpARFC1A (Fig. 7C,G,K) overexpressor, even though these transgenes were overexpressed to various extents (Fig. S11). This suggests that PpARFC1B and PpARFC2 may be functionally homologous. Consistent with this, when PpARFC1B was disrupted in the iOX:SB/pparfc2 background (Fig. S10C,D), gametophore inhibition phenotype was partially rescued (Fig. 7M-X). We ruled out the possibility that the phenotype reversal observed in iOX:SB/pparfc2/pparfc1b was a consequence of reduced SB overexpression as SB transcript abundance in iOX:SB/pparfc2/pparfc1b was comparable to that in the iOX:SB parental line (Fig. S12A). As most ARFs possess a C-terminal dimerization domain that allows homo- and heterodimerization to be formed (Guilfoyle and Hagen, 2007; Kato et al., 2015; Vernoux et al., 2011), we further explored the possibility of PpARFC2 forming a heterodimer with PpARFC1B in a bimolecular fluorescence complementation (BiFC) assay. No YFP reconstitution was observed when cYFP-PpARFC1B and nYFP-PpARFC2 were co-expressed in tobacco leaves (Fig. S12B), suggesting that PpARFC2 and PpARFC1B are likely to function independently as transcription factors to regulate an overlapping set of targets; this is also in agreement with the genetic interaction observed between these two C-ARFs.

Fig. 7.

PpARFC2 and PpARFC1B function redundantly downstream of SB. (A-H) Ten-day-old colonies of iOX:PpARFA3 (A,E), iOX:PpARFB1 (B,F), iOX:PpARFC1A (C,G) and iOX:PpARFC1B (D,H) inoculated in BCD medium supplemented with DMSO (A-D) or 1 µM β-estradiol (E-H). Arrowheads indicate gametophores. Scale bars: 1 mm. (I-L) Line charts indicating the number of gametophores per colony in iOX:PpARFA3 (I), iOX:PpARFB1 (J), iOX:PpARFC1A (K) and iOX:PpARFC1B (L) over the course of 17 days of inoculation in BCD medium supplemented with DMSO (circles) or 1 µM β-estradiol (diamonds). n indicates the number of colonies scored. ***P<0.001 (two-tailed, unpaired Student's t-test. (M-T) Eleven-day-old colonies of WT (M,Q), iOX:SB (N,R), iOX:SB/pparfc2 (O,S) and iOX:SB/pparfc2/pparfc1b (P,T) inoculated in BCD medium supplemented with DMSO (M-P) or 1 µM β-estradiol (Q-T). Arrowheads indicate gametophores. Scale bars: 1 mm. (U-X) Line charts indicating the number of gametophores per colony in WT (U), iOX:SB (V), iOX:SB/pparfc2 (W) and iOX:SB/pparfc2/pparfc1b (X) over the course of 17 days of inoculation in BCD medium supplemented with DMSO (circles) or 1 µM β-estradiol (diamonds). n indicates the number of colonies scored. ***P<0.001 (two-tailed, unpaired Student's t-test.

Fig. 7.

PpARFC2 and PpARFC1B function redundantly downstream of SB. (A-H) Ten-day-old colonies of iOX:PpARFA3 (A,E), iOX:PpARFB1 (B,F), iOX:PpARFC1A (C,G) and iOX:PpARFC1B (D,H) inoculated in BCD medium supplemented with DMSO (A-D) or 1 µM β-estradiol (E-H). Arrowheads indicate gametophores. Scale bars: 1 mm. (I-L) Line charts indicating the number of gametophores per colony in iOX:PpARFA3 (I), iOX:PpARFB1 (J), iOX:PpARFC1A (K) and iOX:PpARFC1B (L) over the course of 17 days of inoculation in BCD medium supplemented with DMSO (circles) or 1 µM β-estradiol (diamonds). n indicates the number of colonies scored. ***P<0.001 (two-tailed, unpaired Student's t-test. (M-T) Eleven-day-old colonies of WT (M,Q), iOX:SB (N,R), iOX:SB/pparfc2 (O,S) and iOX:SB/pparfc2/pparfc1b (P,T) inoculated in BCD medium supplemented with DMSO (M-P) or 1 µM β-estradiol (Q-T). Arrowheads indicate gametophores. Scale bars: 1 mm. (U-X) Line charts indicating the number of gametophores per colony in WT (U), iOX:SB (V), iOX:SB/pparfc2 (W) and iOX:SB/pparfc2/pparfc1b (X) over the course of 17 days of inoculation in BCD medium supplemented with DMSO (circles) or 1 µM β-estradiol (diamonds). n indicates the number of colonies scored. ***P<0.001 (two-tailed, unpaired Student's t-test.

Glycoproteins are indispensable components of plant cell walls and have been implicated in a myriad of developmental processes, although their downstream intracellular mechanistic actions remain largely elusive (Ellis et al., 2010; Ma et al., 2018). Here, we demonstrate that AGP-like glycoproteins from P. patens are able to positively induce the transcriptional upregulation of a class C ARF, which in turn represses the expression of a wide range of cell wall remodeling-related genes to dampen cell wall softening/loosening. Negative regulations of the SB-PpARFC2 signaling pathway in gametophore initiation are reflected by the accelerated juvenile bud formation when either SB or PpARFC2 was disrupted (Figs 2A-D and 4B). As the upregulation of PpARFC2 by SB is dosage dependent (Fig. 4A), we propose that abundance of the plasma membrane-localized SB is an important indicator to signal for a temporal inhibition on cell wall softening/loosening. This inhibitory function of SB on cell expansion is clearly demonstrated by the iOX:SB overexpressor phenotype (Fig. 2F-I) and seemingly corroborated by its weak expression in all developmental stages (Fig. S13A), as a high expression level would be detrimental for organ morphogenesis. We argue that during auxin-mediated acid growth, increased SB abundance in the expanding cell wall, as a result of de novo biosynthesis and secretion, could potentially act as a proxy/index to impose temporary pauses on cell expansion through PpARFC2 signaling output. In support of this, 16 h of 1 µM NAA treatment upregulated SB and PpARFC2 expression (Fig. S13B), lending support to the idea that SB and PpARFC2 upregulation during the auxin-mediated acid growth is necessary to halt uncontrolled cell expansion. Intriguingly, under the same duration of 1 µM cytokinin treatment, SB and PpARFC2 expressions were repressed (Fig. S13B). Although poorly defined, cytokinin was previously shown to increase cell wall extensibility (an indication of wall loosening) in excised cotyledons of radish and cucumber (Thomas et al., 1981). Furthermore, cytokinin transcriptionally affects a wide range of cell wall remodeling genes, such as expansin, laccase, pectin-modifying and xyloglucan-modifying genes (Brenner et al., 2012; Downes and Crowell, 1998), strongly indicating a role in cell wall remodeling. Why would treatment of two phytohormones that promote cell wall loosening lead to an opposite outcome in the expression levels of SB and PpARFC2? It is tempting to suggest that crosstalk between auxin and cytokinin may create a periodic transcriptional oscillation of SB and PpARFC2 in a feedback loop to fine-tune cell wall loosening/softening during gametophore morphogenesis, thereby contributing to the anisotropic growth of leafy gametophores in P. patens.

The function of class C ARF in auxin signaling is enigmatic compared with other canonical A- and B-ARF. Domain-swapping experiments using a set of minimal ARFs (one ARF from each of the three classes) in the liverwort Marchantia polymorpha shows that auxin-dependent A-ARF activates a set of common target genes that are antagonized by the auxin-independent B-ARF, whereas C-ARF is neither an explicit activator nor a repressor (Kato et al., 2020). Hence, C-ARF is seemingly excluded from the canonical auxin-TIR1-AUX/IAA repression mechanism, despite showing in vivo interactions with the AUX/IAA repressor (Kato et al., 2015). Analysis using a constitutive auxin-responsive mutant that completely lacks AUX/IAA repressors also concluded that this system is broadly conserved across land plants, including P. patens (Lavy et al., 2016). Another interesting feature of C-ARF is that its transcripts are targeted by microRNA MIR160-directed post-transcriptional regulation (Mallory et al., 2005). This MIR160/C-ARF module regulates a range of developmental processes in plants that includes leaf shape/size formation (Ben-Gera et al., 2016; Hendelman et al., 2012), root meristem patterning (Bennett et al., 2014; Ding and Friml, 2010; Wang et al., 2005) and floral determinacy (Liu et al., 2010). Our results expanded the functional roles of C-ARF to include cell wall remodeling, a function that is shared by one of the three Arabidopsis C-ARFs, ARF17 (Yang et al., 2013). ARF17 is able to directly bind to the promoter region of CALLOSE SYNTHASE 5, a major callose biosynthesis gene, to regulate pollen wall patterning (Yang et al., 2013). Data from our promoter transactivation assay and electrophoresis mobility shift assay (EMSA) showed that PpARFC2 also binds directly to the PME promoter region to suppress its expression (Fig. 6C-E), leading to increased cell wall stiffening and cell growth inhibition as shown by the iOX:PpARFC2 phenotype (Fig. 4D,E). This implies that PpARFC2-downregulated PMEs are required for cell wall softening. The role of PMEs in cell wall remodeling has been contentious with reports showing that in vitro PME treatment softens the cell wall to impact loosening indirectly (Wang et al., 2020). Consistent with this, localized de-esterification of pectin HG at the shoot apical meristem renders the tissues at leaf initiation sites softer (Braybrook and Peaucelle, 2013; Peaucelle et al., 2011). Our data suggest that PpARFC2-regulated PMEs also play a similar softening role during cell wall remodeling. This is supported by two lines of evidence: (1) sb sbl exhibited downregulated PpARFC2 expression (and hence higher PME abundance) and showed accelerated gametophore formation (Figs 2A-E and 4A); (2) in pparfc2, PME expression levels are presumably upregulated, causing the cell wall to soften and subsequently leading to easier/accelerated gametophore bud initiation (Fig. 4B,C). It is interesting to note that the optimal pH for PME activity in Arabidopsis is pH 7.5 (Hongo et al., 2012), indicating that the acidic environment created by auxin-induced acid-growth loosening is unfavorable for PME activity. How loosening and softening of cell wall are synchronized to achieve cell growth requires more careful examination.

We propose that SB abundance on the plasma membrane imparts signaling information to activate downstream PpARFC2 and PpARFC1B. How does SB achieve this? It is conceivable that SB may form agglomerates (Zhou et al., 2014) at the incipient sites of bud initials to interact with cell wall RLKs to initiate a signaling cascade that leads to the upregulation of PpARFC2, possibly through phosphorylation of transcription factor(s) that directly promote the transcription of PpARFC2 or downregulation of MIR160, which is predicted to target PpARFC transcripts (https://www.pmiren.com/) (Guo et al., 2020). However, physical interaction between cell wall glycoproteins and surface-localized RLKs has never been demonstrated in plants, although a genetic interaction between an AGP, SALT OVERLY SENSITIVE5 (SOS5), with two LRR-RLKs was reported (Basu et al., 2016). It will be essential to identify the binding receptor(s) for SB to characterize further the SB-PpARFC2 signaling module.

Plant materials and growth conditions

The moss Physcomitrium patens of Bruch & Schimp subsp. patens was used as WT throughout this study. Homogenized protonematal cells were cultured on BCDAT agar medium overlaid with cellophane (Futamura Chemical Company) under continuous white light (∼25 μmol photon m−2 s−1) at 25°C. To induce leaf-like gametophore formation and score gametophore number, lumps of protonematal cells were inoculated onto BCD agar medium, and newly formed gametophores at the peripheral areas of the colonies were manually scored between 12 and 15 dpi. For transgene induction, medium was supplemented with either DMSO or 1 µM β-estradiol (Merck/E1024). Protoplasts were collected from 5-day-old protonema that had been digested in 1% w/v Driselase (in 8% mannitol) for 30 min. To generate transgenic moss, protoplasts of the appropriate genetic background were prepared and transformed using the polyethylene glycol (PEG)-mediated transformation method (Nishiyama et al., 2000). Final stable transformants were selected twice on BCDAT agar medium containing appropriate antibiotics.

Full-length cDNA library construction and overexpression screen

High-quality, full-length cDNA libraries of P. patens were constructed from non-hormone-treated, auxin-treated or cytokinin-treated protonemal cells as described previously (Nishiyama et al., 2003; Seki et al., 1998). To conduct the overexpression screen, we chose about 3000 full-length cDNA candidates for which functions could not be predicted by both ends of the cDNA sequences. These cDNA candidates were subcloned as SfiI fragments into pTFH22.4 and transiently expressed as untagged proteins in protoplasts (Fujita et al., 2004). Transformed cells were selected 10-20 days after transformation based on green fluorescence encoded by the GFP expression cassette in the vector backbone. These transformed cells were screened for the dormant brood cell-like phenotype.

AGP precipitation by β-glucosyl Yariv

Total proteins were extracted from β-estradiol-treated (16 h) mosses that stably expressed SB-Citrine (#77) or SBquintuple-Citrine (#3) by grinding protonemata tissues in a two-volume homogenization buffer (50 mM HEPES-KOH pH 7.5, 250 mM sucrose, 5% v/v glycerol, 10 mM EDTA). Tissue lysate was filtered through a 50 µm cell strainer and cell debris was cleared by 10 min centrifugation at 8000 g, 4°C. The supernatant was further ultracentrifuged at 100,000 g for 30 min at 4°C to obtain microsomal fractions, which were resuspended in 300 µl of 1 mg/ml β-glucosyl Yariv (Cyrusbioscience) and incubated overnight at 4°C while rotating. The overnight suspension was spun down at 21,000 g for 10 min at 4°C and the resulting pellet was washed twice in 500 µl of 1% w/v NaCl. Thereafter, the resulting pellet was further washed twice by adding serially 250 µl DMSO, 750 µl cold acetone, 10 µl 2% w/v NaCl, and then spun down at 21,000 g for 10 min at 4°C. The Yariv-precipitated pellet was resuspended in SDS loading buffer and resolved on SDS-PAGE with 10% bis-acrylamide gels.

Immunoblotting

To prepare the microsomal protein (P100) fraction, protonemal tissues were ground to fine powder in liquid nitrogen followed by addition of one volume of chopping buffer (0.4 M sucrose, 10 mM HEPES). Cell debris was cleared from the lysate by centrifugation at 3500 g for 5 min, followed by ultracentrifugation of the resulted supernatant at 100,000 g for 30 mins at 4°C. The resulting pellet (P100 fraction) was resuspended in one volume of sample buffer (250 mM Tris pH 6.8, 4% SDS, 20% glycerol, 0.75 M β-mercaptoethanol) and denatured at 68°C for 5 min before being resolved on 10-12% SDS-PAGE bis-acrylamide gels. Proteins were immunoblotted with the following antibodies: rabbit polyclonal anti-Dendra2 (antibodies-online, ABIN361314; 1:2500), mouse monoclonal anti-GFP (Living Colors Av JL-8, Clontech; 1:1000), rat monoclonal JIM13 (gift from Prof. Paul Knox, University of Leeds, UK; 1:1000), rat monoclonal LM6-M (gift from Prof. Paul Knox; 1:1000), peroxidase-conjugated anti-rat (Proteintech, SA00001-15; 1:5000) and peroxidase-conjugated anti-rabbit (ECL, NIF824; 1:5000).

Cell wall composition analysis

To extract cell wall sugar content, 5-day-old protonematal tissues were homogenized in the presence of absolute ethanol (10% w/v). The alcohol-insoluble residue (AIR) was prepared from the homogenate as described (Goubet et al., 2009). The AIR was hydrolyzed with 2 M trifluoroacetic acid (TFA) at 120°C for 1 h. After removal of TFA with a SpeedVac (Thermo Fisher Scientific), the sugar composition was determined by high-performance anion-exchange chromatography with pulsed amperometric detection (HPAEC-PAD) using a Dionex ICS-5000+ liquid chromatograph equipped with a CarboPac PA-1 column and a pulsed amperometric detector (Thermo Fisher Scientific) as described previously (Ishikawa et al., 2000). Mole percentage (Mol%) represents relative concentrations of monosaccharides and were calculated as surface area under each monosaccharide peak, with respect to a standard mixture of nine monosaccharides (Fuc, Rha, Ara, Glc, Gal, Man, Xyl, GlcA and GalA).

Phylogenetic tree construction and molecular cloning

Amino acid sequences and respective gene identifiers were retrieved from Phytozome v13 (https://phytozome-next.jgi.doe.gov) and NCBI GenBank (Table S2). These sequences are aligned with MAFFT v7 (https://mafft.cbrc.jp/alignment/software/) (Katoh et al., 2018) before being curated using NOISY algorithms embedded on Phylogeny (https://ngphylogeny.fr) (Lemoine et al., 2019). The best substitution model (with the lowest Bayesian information criterion, BIC score) for these aligned and curated amino acid sequences were then predicted using the Model Selection tab on IQ-TREE web server (http://iqtree.cibiv.univie.ac.at) (Trifinopoulos et al., 2016). Maximum likelihood analysis of the amino acid sequences was then performed using the Tree Inference tab on IQ-TREE web server by choosing Blosum62 substitution matrices and standard bootstrap analysis, with 100 bootstrap alignments. See supplementary Materials and Methods for further details on cloning protocols for knock-out, knock-in and overexpression constructs.

qRT-PCR, semi-quantitative RT-PCR and RNA-seq

Total RNA was extracted from 5-day-old protonemal tissues that were blotted dry and ground in liquid nitrogen according to the manufacturer's protocol for the RNeasy Mini Kit (Qiagen). First-strand cDNA was synthesized with oligo (dT) and random primers using ReverTra Ace qPCR RT kit (Toyobo) following manufacturer's instructions. qRT-PCR was performed on an Applied Biosystems 7300 Real-Time PCR System (Thermo Fisher Scientific) with Power SYBR Green PCR Master Mix (Thermo Fisher Scientific). The specificity of all amplicons was determined using dissociation curve analyses. Relative fold changes were calculated based on at least three technical and three biological replicates using the formula 2−[C(t) Gene of Interest−C(t)Ub Conjugating E2] to normalize gene expression levels to Ubiquitin Conjugating Enzyme E2. Oligonucleotides used in the qRT-PCR are summarized in Table S3. For RNA-seq, total RNA from DMSO- and 1 μM β-estradiol-treated protonemal tissues of iOX:PpARFC2 was extracted as above. RNA-seq was carried out using the DNBseq platform at BGI Genomics with DEGs between DMSO-treated and β-estradiol-induced samples identified using the EBSeq method [−1≤log2fold change ≥1, posterior probability of being equivalent expression (PPEE)≤0.05] (Leng et al., 2013). Fragment counts of each gene was normalized by fragments per kb per million (FPKM). GO functional enrichment of DEGs was performed using the Hypergeometric Distribution test (phyper) function in R to obtain P-values for each GO term. GO terms with false discovery rate (FDR)<0.01 are defined as significantly enriched. See supplementary Materials and Methods for further details on semi-quantitative RT-PCR.

Promoter transactivation assay and EMSA

Agrobacterium tumefaciens (GV3101) strains that harbor pGWB35-PMEpro:FLuc, UBQ10:PpARFC2-YFP and 35S:RLUc were co-infiltrated (at OD600 0.5) into the abaxial side of 3-week-old tobacco (Nicotiana benthamiana) leaves. Leaves were harvested 40 h post-infiltration and ground in liquid nitrogen, followed by the addition of three volumes of extraction buffer (25 mM NaH2PO4, 2 mM EDTA, 10% glycerol, 1% Triton X-100, 2 mM DTT, 25 mM Tris-HCl pH 8.0). Protein lysate was spun down at 20,000 g and supernatant was collected. FLuc and RLuc activities were detected using Dual-Luciferase® Reporter Assay system (Promega) in a luminometer (AB-2250 Luminescencer-MCA).

To perform EMSA, the B3 DNA-binding domain (136th∼229th amino acids) of PpARFC2 was PCR-amplified with primers #TOK_100 and #TOK_101, cloned into pMALC2x (NEB) and expressed as a MBP-tagged recombinant protein (MBP-B3) in Escherichia coli strain BL21. Soluble MBP-B3 was purified by amylose-agarose affinity (NEB) and the eluted protein was dialyzed overnight in 100 mM Tris (pH 7.5), 100 mM KCl, 5 mM MgCl2. Complementary DNA probes were 3′ biotinylated using the Biotin 3′ End DNA Labeling Kit (Thermo Fisher Scientific). Oligonucleotides (#TOK_198∼#TOK_217) of 100 nM were individually biotinylated at the 3′ in reactions that consisted of TdT Reaction Buffer (100 mM cacodylic acid, 2 mM CoCl2, 0.2 mM DTT), 0.5 µM Biotin-11-UTP and 0.15 U/µl terminal deoxynucleotidyl transferase (TdT). The reactions were incubated at 37°C for 30 min followed by the addition of 2.5 µl 2.5 M EDTA to stop the reaction. Biotinylated oligonucleotides were cleaned using chloroform:isoamyl alcohol extraction and complementary pairs of oligonucleotides were annealed by 5 min incubation at 95°C followed by gradual cooling to room temperature at the rate of −0.1°C/s. EMSA was performed according to the instruction manual of the LightShift Chemiluminescent EMSA Kit (Thermo Fisher Scientific). In brief, each binding reaction contained 7 µl dialyzed MBP-B3, 1× binding buffer (10 mM Tris pH 7.5, 50 mM KCl, 1 mM DTT, 2.5% glycerol), 1 µg Poly (dI.dC) and 20 fmol biotinylated DNA probes in a 20 µl reaction volume. In control reactions, specific unlabeled DNA probe competitors were supplied at 800-fold molar excess. After a 20 min incubation at room temperature, the resulting protein–DNA complexes were resolved on an 8% TBE non-denaturing polyacrylamide gel and then transferred to a Hybond N+ nylon membrane (Cytiva) and crosslinked at 120 mJ/cm2 for 1 min. Detection of the biotinylated DNA probes was carried out using the Chemiluminescent Nucleic Acid Detection Module (Thermo Fisher Scientific).

BiFC assays

Agrobacterium tumefaciens (GV3101) strains that harbor the pUBN-nYFP: nYFP-PpARFC2 and pUBN-cYFP: cYFP-PpARFC1B constructs were co-infiltrated (at OD600 0.5) into the abaxial side of 3-week-old tobacco (Nicotiana benthamiana) leaves. The positive control used was a basic helix-loop-helix transcription factor, PpSMF1, which has been shown to form a homodimer (Chater et al., 2016). Forty hours post-infiltration, leaves were imaged using confocal laser-scanning microscope.

Microscopy imaging

Confocal images were acquired using a Zeiss laser-scanning microscope 980. GFP was excited using laser 514 nm (laser power 2-5%) with emission signal collected within a 508-579 nm spectrum at 1 Airy Unit, in conjunction with a main dichroic beam splitter (MBS) 514/639.

We thank Prof. Paul Knox (University of Leeds, U.K.) for his generous donation of JIM13 and LM6-M antibodies. We are indebted to Dr Yukako Chiba (Hokkaido University, Japan) and Prof. Tsuyoshi Nakagawa (Shimane University, Japan) for sharing the 35S:RLuc construct and pGWB35 vector. We acknowledge the excellent services provided by IPMB Cell Biology Core Lab and G-TeC DNA Analysis Division.

Author contributions

Conceptualization: T.F.; Methodology: O.T., P.S., J.R., M.A., T.K., T.F.; Validation: B.P.S., T.F.; Investigation: O.T., P.S., J.R., L.H., M.A., B.P.S., Y.W., T.K.; Resources: O.T.; Writing - original draft: O.T.; Writing - review & editing: O.T., J.R., M.A., T.K., T.F.; Visualization: O.T., J.R.; Supervision: O.T., T.F.; Funding acquisition: O.T., T.F.

Funding

This research was funded by the NOVARTIS Foundation (Japan) for the Promotion of Science (3133 to O.T.), the Ministry of Education, Culture, Sports, Science and Technology/Japan Society for the Promotion of Science (JP19K06701 to O.T.; JP16K14747, JP18H04829 and JP20H04878 to T.F.), an IPMB intramural grant (Academia sinica to O.T.), and a National Science and Technology Council (Taiwan) research grant (111-2311-B-001-007-MY3 to O.T.). J.R. and P.S. were supported by scholarships from the China Scholarship Council and Japanese Government Monbukagakusho Ministry of Education, Culture, Sports, Science and Technology.

Data availability

RNA-seq datasets in this study have been deposited in the NCBI GEO database with the following accession numbers: GSE213920, GSM6596485, GSM6596486, GSM6596487, GSM6596488, GSM6596489.

Anderson
,
C. T.
and
Kieber
,
J. J.
(
2020
).
Dynamic construction, perception, and remodeling of plant cell walls
.
Annu. Rev. Plant Biol.
71
,
39
-
69
.
Basu
,
D.
,
Tian
,
L.
,
DeBrosse
,
T.
,
Poirier
,
E.
,
Emch
,
K.
,
Herock
,
H.
,
Travers
,
A.
and
Showalter
,
A. M.
(
2016
).
Glycosylation of a Fasciclin-Like Arabinogalactan-Protein (SOS5) mediates root growth and seed mucilage adherence via a cell wall receptor-like kinase (FEI1/FEI2) Pathway in Arabidopsis
.
PLoS ONE
11
,
e0145092
.
Batoko
,
H.
,
Zheng
,
H.-Q.
,
Hawes
,
C.
and
Moore
,
I.
(
2000
).
A Rab1 GTPase is required for transport between the endoplasmic reticulum and golgi apparatus and for normal Golgi movement in plants
.
Plant Cell
12
,
2201
-
2217
.
Ben-Gera
,
H.
,
Dafna
,
A.
,
Alvarez
,
J. P.
,
Bar
,
M.
,
Mauerer
,
M.
and
Ori
,
N.
(
2016
).
Auxin-mediated lamina growth in tomato leaves is restricted by two parallel mechanisms
.
Plant J.
86
,
443
-
457
.
Bennett
,
T.
,
van den Toorn
,
A.
,
Willemsen
,
V.
and
Scheres
,
B.
(
2014
).
Precise control of plant stem cell activity through parallel regulatory inputs
.
Development
141
,
4055
-
4064
.
Boer
,
D. R.
,
Freire-Rios
,
A.
,
Van Den Berg
,
W. A. M.
,
Saaki
,
T.
,
Manfield
,
I. W.
,
Kepinski
,
S.
,
López-Vidrieo
,
I.
,
Franco-Zorrilla
,
J. M.
,
De Vries
,
S. C.
,
Solano
,
R.
et al.
(
2014
).
Structural basis for DNA binding specificity by the auxin-dependent ARF transcription factors
.
Cell
156
,
577
-
589
.
Braybrook
,
S. A.
and
Peaucelle
,
A.
(
2013
).
Mechano-chemical aspects of organ formation in arabidopsis thaliana: the relationship between Auxin and Pectin
.
PLoS ONE
8
,
e57813
.
Brenner
,
W. G.
,
Ramireddy
,
E.
,
Heyl
,
A.
and
Schmülling
,
T.
(
2012
).
Gene regulation by cytokinin in Arabidopsis
.
Front. Plant Sci.
3
,
1
-
22
.
Chater
,
C. C.
,
Caine
,
R. S.
,
Tomek
,
M.
,
Wallace
,
S.
,
Kamisugi
,
Y.
,
Cuming
,
A. C.
,
Lang
,
D.
,
MacAlister
,
C. A.
,
Casson
,
S.
,
Bergmann
,
D. C.
et al.
(
2016
).
Origin and function of stomata in the moss Physcomitrella patens
.
Nat. Plants
2
,
16179
.
Cosgrove
,
D. J.
(
2000
).
Loosening of plant cell walls by expansins
.
Nature
407
,
321
-
326
.
Ding
,
Z.
and
Friml
,
J.
(
2010
).
Auxin regulates distal stem cell differentiation in Arabidopsis roots
.
Proc. Natl. Acad. Sci. USA
107
,
12046
-
12051
.
Ding
,
L.
and
Zhu
,
J.-K.
(
1997
).
A role for arabinogalactan-proteins in root epidermal cell expansion
.
Planta
203
,
289
-
294
.
Downes
,
B. P.
and
Crowell
,
D. N.
(
1998
).
Cytokinin regulates the expression of a soybean β-expansin gene by a post-transcriptional mechanism
.
Plant Mol. Biol.
37
,
437
-
444
.
Ellis
,
M.
,
Egelund
,
J.
,
Schultz
,
C. J.
and
Bacic
,
A.
(
2010
).
Arabinogalactan-proteins: key regulators at the cell surface?
Plant Physiol.
153
,
403
-
419
.
Franco-Zorrilla
,
J. M.
,
López-Vidriero
,
I.
,
Carrasco
,
J. L.
,
Godoy
,
M.
,
Vera
,
P.
and
Solano
,
R.
(
2014
).
DNA-binding specificities of plant transcription factors and their potential to define target genes
.
Proc. Natl. Acad. Sci. USA
111
,
2367
-
2372
.
Freire-Rios
,
A.
,
Tanaka
,
K.
,
Crespo
,
I.
,
Van der Wijk
,
E.
,
Sizentsova
,
Y.
,
Levitsky
,
V.
,
Lindhoud
,
S.
,
Fontana
,
M.
,
Hohlbein
,
J.
,
Roeland Boer
,
D.
et al.
(
2020
).
Architecture of DNA elements mediating ARF transcription factor binding and auxin-responsive gene expression in Arabidopsis
.
Proc. Natl. Acad. Sci. USA
117
,
24557
-
24566
.
Fujita
,
T.
,
Nishiyama
,
T.
,
Hiwatashi
,
Y.
and
Hasebe
,
M.
(
2004
).
Gene tagging, gene- and enhancer-trapping, and full-length cDNA overexpression in Physcomitrella Patens
. In
New Frontiers in Bryology
(ed.
A. J.
Wood
,
M. J.
Oliver
and
D. J.
Cove
), pp.
111
-
132
.
Kluwer Academic
.
Gomord
,
V.
,
Denmat
,
L.-A.
,
Fitchette-Laine
,
A.-C.
,
Satiat-Jeunemaitre
,
B.
,
Hawes
,
C.
and
Faye
,
L.
(
1997
).
The C-terminal HDEL sequence is sufficient for retention of secretory proteins in the endoplasmic reticulum (ER) but promotes vacuolar targeting of proteins that escape the ER
.
Plant J.
11
,
313
-
325
.
Goss
,
C. A.
,
Brockmann
,
D. J.
,
Bushoven
,
J. T.
and
Roberts
,
A. W.
(
2012
).
A CELLULOSE SYNTHASE (CESA) gene essential for gametophore morphogenesis in the moss Physcomitrella patens
.
Planta
235
,
1355
-
1367
.
Goubet
,
F.
,
Barton
,
C. J.
,
Mortimer
,
J. C.
,
Yu
,
X.
,
Zhang
,
Z.
,
Miles
,
G. P.
,
Richens
,
J.
,
Liepman
,
A. H.
,
Seffen
,
K.
and
Dupree
,
P.
(
2009
).
Cell wall glucomannan in Arabidopsis is synthesised by CSLA glycosyltransferases, and influences the progression of embryogenesis
.
Plant J.
60
,
527
-
538
.
Grefen
,
C.
,
Donald
,
N.
,
Hashimoto
,
K.
,
Kudla
,
J.
,
Schumacher
,
K.
and
Blatt
,
M. R.
(
2010
).
A ubiquitin-10 promoter-based vector set for fluorescent protein tagging facilitates temporal stability and native protein distribution in transient and stable expression studies
.
Plant J.
64
,
355
-
365
.
Guilfoyle
,
T. J.
and
Hagen
,
G.
(
2007
).
Auxin response factors
.
Curr. Opin. Plant Biol.
10
,
453
-
460
.
Guo
,
Z.
,
Kuang
,
Z.
,
Wang
,
Y.
,
Zhao
,
Y.
,
Tao
,
Y.
,
Cheng
,
C.
,
Yang
,
J.
,
Lu
,
X.
,
Hao
,
C.
,
Wang
,
T.
et al.
(
2020
).
PmiREN: A comprehensive encyclopedia of plant miRNAs
.
Nucleic Acids Res.
48
,
D1114
-
D1121
.
Harrison
,
C. J.
,
Roeder
,
A. H. K.
,
Meyerowitz
,
E. M.
and
Langdale
,
J. A.
(
2009
).
Local cues and asymmetric cell divisions underpin body plan transitions in the moss Physcomitrella patens
.
Curr. Biol.
19
,
461
-
471
.
Hendelman
,
A.
,
Buxdorf
,
K.
,
Stav
,
R.
,
Kravchik
,
M.
and
Arazi
,
T.
(
2012
).
Inhibition of lamina outgrowth following Solanum lycopersicum AUXIN RESPONSE FACTOR 10 (SlARF10) derepression
.
Plant Mol. Biol.
78
,
561
-
576
.
Hongo
,
S.
,
Sato
,
K.
,
Yokoyama
,
R.
and
Nishitani
,
K.
(
2012
).
Demethylesterification of the primary wall by PECTIN METHYLESTERASE35 provides mechanical support to the Arabidopsis stem
.
Plant Cell
24
,
2624
-
2634
.
Ishikawa
,
M.
,
Kuroyama
,
H.
,
Takeuchi
,
Y.
and
Tsumuraya
,
Y.
(
2000
).
Characterization of pectin methyltransferase from soybean hypocotyls
.
Planta
210
,
782
-
791
.
Kato
,
H.
,
Ishizaki
,
K.
,
Kouno
,
M.
,
Shirakawa
,
M.
,
Bowman
,
J. L.
,
Nishihama
,
R.
and
Kohchi
,
T.
(
2015
).
Auxin-mediated transcriptional system with a minimal set of components is critical for morphogenesis through the life cycle in Marchantia polymorpha
.
PLoS Genet.
11
,
e1005365
.
Kato
,
H.
,
Mutte
,
S. K.
,
Suzuki
,
H.
,
Crespo
,
I.
,
Das
,
S.
,
Radoeva
,
T.
,
Fontana
,
M.
,
Yoshitake
,
Y.
,
Hainiwa
,
E.
,
van den Berg
,
W.
et al.
(
2020
).
Design principles of a minimal auxin response system
.
Nature Plants
6
,
473
-
482
.
Katoh
,
K.
,
Rozewicki
,
J.
and
Yamada
,
K. D.
(
2018
).
MAFFT online service: Multiple sequence alignment, interactive sequence choice and visualization
.
Brief. Bioinform.
20
,
1160
-
1166
.
Kitazawa
,
K.
,
Tryfona
,
T.
,
Yoshimi
,
Y.
,
Hayashi
,
Y.
,
Kawauchi
,
S.
,
Antonov
,
L.
,
Tanaka
,
H.
,
Takahashi
,
T.
,
Kaneko
,
S.
,
Dupree
,
P.
et al.
(
2013
).
β-galactosyl yariv reagent binds to the β-1,3-galactan of arabinogalactan proteins
.
Plant Physiol.
161
,
1117
-
1126
.
Lamport
,
D. T. A.
and
Várnai
,
P.
(
2013
).
Periplasmic arabinogalactan glycoproteins act as a calcium capacitor that regulates plant growth and development
.
New Phytol.
197
,
58
-
64
.
Lamport
,
D. T. A.
,
Varnai
,
P.
and
Seal
,
C. E.
(
2014
).
Back to the future with the AGP-Ca2+ flux capacitor
.
Ann. Bot.
114
,
1069
-
1085
.
Lavy
,
M.
and
Estelle
,
M.
(
2016
).
Mechanisms of auxin signaling
.
Development
143
,
3226
-
3229
.
Lavy
,
M.
,
Prigge
,
M. J.
,
Tao
,
S.
,
Shain
,
S.
,
Kuo
,
A.
,
Kirchsteiger
,
K.
and
Estelle
,
M.
(
2016
).
Constitutive auxin response in Physcomitrella reveals complex interactions between Aux/IAA and ARF proteins
.
eLife
5
,
e13325
.
Lee
,
K. J. D.
,
Sakata
,
Y.
,
Mau
,
S.-L.
,
Pettolino
,
F.
,
Bacic
,
A.
,
Quantrano
,
R. S.
,
Knight
,
C. D.
and
Knox
,
J. P.
(
2005
).
Arabinogalactan proteins are required for apical cell extension in the moss Physcomitrella patens
.
Plant Cell
17
,
3051
-
3065
.
Lemoine
,
F.
,
Correia
,
D.
,
Lefort
,
V.
,
Doppelt-Azeroual
,
O.
,
Mareuil
,
F.
,
Cohen-Boulakia
,
S.
and
Gascuel
,
O.
(
2019
).
NGPhylogeny.fr: New generation phylogenetic services for non-specialists
.
Nucleic Acids Res.
47
,
W260
-
W265
.
Leng
,
N.
,
Dawson
,
J. A.
,
Thomson
,
J. A.
,
Ruotti
,
V.
,
Rissman
,
A. I.
,
Smits
,
B. M. G.
,
Haag
,
J. D.
,
Gould
,
M. N.
,
Stewart
,
R. M.
and
Kendziorski
,
C.
(
2013
).
EBSeq: an empirical Bayes hierarchical model for inference in RNA-seq experiments
.
Bioinformatics
29
,
1035
-
1043
.
Leszczuk
,
A.
,
Kalaitzis
,
P.
,
Blazakis
,
K. N.
and
Zdunek
,
A.
(
2020
).
The role of arabinogalactan proteins (AGPs) in fruit ripening—a review
.
Horticulture Res.
7
,
176
.
Liu
,
X.
,
Huang
,
J.
,
Wang
,
Y.
,
Khanna
,
K.
,
Xie
,
Z.
,
Owen
,
H. A.
and
Zhao
,
D.
(
2010
).
The role of floral organs in carpels, an Arabidopsis loss-of-function mutation in MicroRNA160a, in organogenesis and the mechanism regulating its expression
.
Plant J.
62
,
416
-
428
.
Lopez-Hernandez
,
F.
,
Tryfona
,
T.
,
Rizza
,
A.
,
Yu
,
X. L.
,
Harris
,
M. O. B.
,
Webb
,
A. A. R.
,
Kotake
,
T.
and
Dupree
,
P.
(
2020
).
Calcium binding by arabinogalactan polysaccharides is important for normal plant development
.
Plant Cell
32
,
3346
-
3369
.
Ma
,
Y.
,
Zeng
,
W.
,
Bacic
,
A.
and
Johnson
,
K.
(
2018
).
AGP Through time and space
.
Annu. Plant Rev.
1
,
1
-
38
.
MacAlister
,
C. A.
,
Ortiz-Ramírez
,
C.
,
Becker
,
J. D.
,
Feijó
,
J. A.
and
Lippman
,
Z. B.
(
2016
).
Hydroxyproline O-arabinosyltransferase mutants oppositely alter tip growth in Arabidopsis thaliana and Physcomitrella patens
.
Plant J.
85
,
193
-
208
.
Mallory
,
A. C.
,
Bartel
,
D. P.
and
Bartel
,
B.
(
2005
).
MicroRNA-directed regulation of Arabidopsis Auxin Response Factor17 is essential for proper development and modulates expression of early auxin response genes
.
Plant Cell
17
,
1360
-
1375
.
McQueen-Mason
,
S.
,
Durachko
,
D. M.
and
Cosgrove
,
D. J.
(
1992
).
Two endogenous proteins that induce cell wall extension in plants
.
Plant Cell
4
,
1425
-
1433
.
Micheli
,
F.
(
2001
).
Pectin methylesterases: Cell wall enzymes with important roles in plant physiology
.
Trends Plant Sci.
6
,
414
-
419
.
Motose
,
H.
,
Sugiyama
,
M.
and
Fukuda
,
H.
(
2004
).
A proteoglycan mediates inductive interaction during plant vascular development
.
Nature
429
,
873
-
878
.
Nishiyama
,
T.
,
Hiwatashi
,
Y.
,
Sakakibara
,
K.
,
Kato
,
M.
and
Hasebe
,
M.
(
2000
).
Tagged mutagenesis and gene-trap in the moss, Physcomitrella patens by shuttle mutagenesis
.
DNA Res.
7
,
9
-
17
.
Nishiyama
,
T.
,
Fujita
,
T.
,
Shin-I
,
T.
,
Seki
,
M.
,
Nishide
,
H.
,
Uchiyama
,
I.
,
Kamiya
,
A.
,
Carninci
,
P.
,
Hayashizaki
,
Y.
,
Shinozaki
,
K.
et al.
(
2003
).
Comparative genomics of Physcomitrella patens gametophytic transcriptome and Arabidopsis thaliana: Implication for land plant evolution
.
Proc. Natl. Acad. Sci. USA
100
,
8007
-
8012
.
Peaucelle
,
A.
,
Braybrook
,
S. A.
,
Le Guillou
,
L.
,
Bron
,
E.
,
Kuhlemeier
,
C.
and
Höfte
,
H.
(
2011
).
Pectin-induced changes in cell wall mechanics underlie organ initiation in Arabidopsis
.
Curr. Biol.
21
,
1720
-
1726
.
Pereira
,
A. M.
,
Lopes
,
A. L.
and
Coimbra
,
S.
(
2016
).
Arabinogalactan proteins as interactors along the crosstalk between the pollen tube and the female tissues
.
Front. Plant Sci.
7
,
1
-
15
.
Šamaj
,
J.
,
Braun
,
M.
,
Baluška
,
F.
,
Ensikat
,
H.-J.
,
Tsumuraya
,
Y.
and
Volkmann
,
D.
(
1999
).
Specific localization of arabinogalactan-protein epitopes at the surface of maize root hairs
.
Plant Cell Physiol.
40
,
874
-
883
.
Schultz
,
C. J.
,
Ferguson
,
K. L.
,
Lahnstein
,
J.
and
Bacic
,
A.
(
2004
).
Post-translational modifications of arabinogalactan-peptides of Arabidopsis thaliana
.
J. Biol. Chem.
279
,
45503
-
45511
.
Seifert
,
G. J.
and
Roberts
,
K.
(
2007
).
The Biology of Arabinogalactan Proteins
.
Annu. Rev. Plant Biol.
58
,
137
-
161
.
Seki
,
M.
,
Carninci
,
P.
,
Nishiyama
,
Y.
,
Hayashizaki
,
Y.
and
Shinozaki
,
K.
(
1998
).
High-efficiency cloning of Arabidopsis full-length cDNA by biotinylated CAP trapper
.
Plant J.
15
,
707
-
720
.
Spartz
,
A. K.
,
Ren
,
H.
,
Park
,
M. Y.
,
Grandt
,
K. N.
,
Lee
,
S. H.
,
Murphy
,
A. S.
,
Sussman
,
M. R.
,
Overvoorde
,
P. J.
and
Gray
,
W. M.
(
2014
).
SAUR inhibition of PP2C-D phosphatases activates plasma membrane H+-ATPases to promote cell expansion in Arabidopsis
.
Plant Cell
26
,
2129
-
2142
.
Takahashi
,
K.
,
Hayashi
,
K. I.
and
Kinoshita
,
T.
(
2012
).
Auxin activates the plasma membrane H+-ATPase by phosphorylation during hypocotyl elongation in Arabidopsis
.
Plant Physiol.
159
,
632
-
641
.
Tan
,
L.
,
Eberhard
,
S.
,
Pattathil
,
S.
,
Warder
,
C.
,
Glushka
,
J.
,
Yuan
,
C.
,
Hao
,
Z.
,
Zhu
,
X.
,
Avci
,
U.
,
Miller
,
J. S.
et al.
(
2013
).
An Arabidopsis cell wall proteoglycan consists of pectin and arabinoxylan covalently linked to an arabinogalactan protein
.
Plant Cell
25
,
270
-
287
.
Thomas
,
J.
,
Ross
,
C. W.
,
Chastain
,
C. J.
,
Koomanoff
,
N.
,
Hendrix
,
J. E.
and
Van Volkenburgh
,
E.
(
1981
).
Cytokinin-induced wall extensibility in excised cotyledons of radish and cucumber
.
Plant Physiol.
68
,
107
-
110
.
Trifinopoulos
,
J.
,
Nguyen
,
L.-T.
,
von Haeseler
,
A.
and
Minh
,
B. Q.
(
2016
).
W-IQ-TREE: a fast online phylogenetic tool for maximum likelihood analysis
.
Nucleic Acids Res.
44
,
W232
-
W235
.
Velasquez
,
S. M.
,
Ricardi
,
M. M.
,
Dorosz
,
J. G.
,
Fernandez
,
P. V.
,
Nadra
,
A. D.
,
Pol-Fachin
,
L.
,
Egelund
,
J.
,
Gille
,
S.
,
Harholt
,
J.
,
Ciancia
,
M.
et al.
(
2011
).
O-glycosylated cell wall proteins are essential in root hair growth
.
Science
332
,
1401
-
1403
.
Vernoux
,
T.
,
Brunoud
,
G.
,
Farcot
,
E.
,
Morin
,
V.
,
Van Den Daele
,
H.
,
Legrand
,
J.
,
Oliva
,
M.
,
Das
,
P.
,
Larrieu
,
A.
,
Wells
,
D.
et al.
(
2011
).
The auxin signalling network translates dynamic input into robust patterning at the shoot apex
.
Mol. Syst. Biol.
7
,
508
.
Wang
,
J.-W.
,
Wang
,
L.-J.
,
Mao
,
Y.-B.
,
Cai
,
W.-J.
,
Xue
,
H.-W.
and
Chen
,
X.-Y.
(
2005
).
Control of root cap formation by MicroRNA-targeted auxin response factors in Arabidopsis
.
Plant Cell
17
,
2204
-
2216
.
Wang
,
X.
,
Wilson
,
L.
and
Cosgrove
,
D. J.
(
2020
).
Pectin methylesterase selectively softens the onion epidermal wall yet reduces acid-induced creep
.
J. Exp. Bot.
71
,
2629
-
2640
.
Willats
,
W. G. T.
,
Orfila
,
C.
,
Limberg
,
G.
,
Buchholt
,
H. C.
,
Van Alebeek
,
G.-J. W. M.
,
Voragen
,
A. G. J.
,
Marcus
,
S. E.
,
Christensen
,
T. M. I. E.
,
Mikkelsen
,
J. D.
,
Murray
,
B. S.
et al.
(
2001
).
Modulation of the degree and pattern of methyl-esterification of pectic homogalacturonan in plant cell walls: Implications for pectin methyl esterase action, matrix properties, and cell adhesion
.
J. Biol. Chem.
276
,
19404
-
19413
.
Wolf
,
S.
,
Hématy
,
K.
and
Höfte
,
H.
(
2012
).
Growth control and cell wall signaling in plants
.
Annu. Rev. Plant Biol.
63
,
381
-
407
.
Yang
,
J.
,
Tian
,
L.
,
Sun
,
M.-X.
,
Huang
,
X.-Y.
,
Zhu
,
J.
,
Guan
,
Y.-F.
,
Jia
,
Q.-S.
and
Yang
,
Z.-N.
(
2013
).
AUXIN RESPONSE FACTOR17 is essential for pollen wall pattern formation in Arabidopsis
.
Plant Physiol.
162
,
720
-
731
.
Yates
,
E. A.
,
Valdor
,
J.-F.
,
Haslam
,
S. M.
,
Morris
,
H. R.
,
Dell
,
A.
,
Mackie
,
W.
and
Knox
,
J. P.
(
1996
).
Characterization of carbohydrate structural features recognized by anti-arabinogalactan-protein monoclonal antibodies
.
Glycobiology
6
,
131
-
139
.
Yu
,
M.
and
Zhao
,
J.
(
2012
).
The cytological changes of tobacco zygote and proembryo cells induced by beta-glucosyl Yariv reagent suggest the involvement of arabinogalactan proteins in cell division and cell plate formation
.
BMC Plant Biol.
12
,
126
.
Yuan
,
S.
,
Wu
,
Y.
and
Cosgrove
,
D. J.
(
2001
).
A fungal endoglucanase with plant cell wall extension activity
.
Plant Physiol.
127
,
324
-
333
.
Zhou
,
L. H.
,
Weizbauer
,
R. A.
,
Singamaneni
,
S.
,
Xu
,
F.
,
Genin
,
G. M.
and
Pickard
,
B. G.
(
2014
).
Structures formed by a cell membrane-associated arabinogalactan-protein on graphite or mica alone and with Yariv phenylglycosides
.
Ann. Bot.
114
,
1385
-
1397
.

Competing interests

The authors declare no competing or financial interests.

Supplementary information