- © 2019 American Society of Plant Biologists. All Rights Reserved.
Abstract
Secretory trafficking is highly conserved in all eukaryotic cells and is required for secretion of proteins as well as extracellular matrix components. In plants, the export of cuticular waxes and various cell wall components relies on secretory trafficking, but the molecular mechanisms underlying their secretion are not well understood. In this study, we characterize the Arabidopsis (Arabidopsis thaliana) dwarf eceriferum11 (cer11) mutant and we show that it exhibits reduced stem cuticular wax deposition, aberrant seed coat mucilage extrusion, and delayed secondary cell wall columella formation, as well as a block in secretory GFP trafficking. Cloning of the CER11 gene revealed that it encodes a C-TERMINAL DOMAIN PHOSPHATASE-LIKE2 (CPL2) protein. Thus, secretory trafficking in plant cells in general, and secretion of extracellular matrix constituents in developing epidermal cells in particular, involves a dephosphorylation step catalyzed by CER11/CPL2.
The primary aerial surfaces of land plants are covered by a lipidic cuticle, which provides protection against nonstomatal water loss, serves as a barrier to pathogen invasion, and prevents organ fusion. The cuticle is synthesized by shoot epidermal cells and mainly consists of cutin polyester matrix and cuticular waxes. The biosynthesis of cuticular waxes has been studied for many years and the major biosynthetic enzymes have been identified and localized to the endoplasmic reticulum (ER; reviewed by Samuels et al., 2008; Bernard and Joubès, 2013; Haslam and Kunst, 2013). By comparison, relatively little is known about the export of wax molecules from their site of synthesis in the ER to the plasma membrane (PM), and from the PM through the cell wall to the plant surface. Currently, the only specific components of the wax export machinery known are two ATP-binding casette (ABC) transporters ECERIFERUM5 (CER5)/ABCG12 (Pighin et al., 2004) and ABCG11(Bird et al., 2007), which function as dimers in the PM (McFarlane et al., 2010), and two glycosylphosphatidylinositol (GPI)-anchored lipid transport proteins that also reside in the PM (DeBono et al., 2009; Kim et al., 2012). However, the delivery mechanism of wax molecules to the PM has not been resolved.
One attractive hypothesis is that wax components are transported from the ER to the PM by Golgi-mediated vesicular trafficking through the secretory pathway (reviewed by Kunst and Samuels, 2003). The secretory pathway and the endomembrane system are conserved in eukaryotes, but the plant endomembrane system differs from the mammalian system in several important ways. For example, the organization of endosomes in plant cells is different from that in mammalian cells. The trans-Golgi network (TGN), an independent and highly mobile organelle in plant cells, serves as the main secretory and endocytic hub equivalent to early endosomes (Dettmer et al., 2006; Viotti et al., 2010; Gendre et al., 2015). Although secretion of proteins and cell wall polysaccharides via secretory trafficking in plants has been studied to some extent (Viotti et al., 2010; Driouich et al., 2012; reviewed by Drakakaki and Dandekar, 2013), the question of whether cuticular waxes are transported to the PM along the secretory pathway is still a matter of debate. Recently, wax secretion has been shown to be dependent on the function of GNOM-LIKE1 and ECHIDNA, proteins required for Golgi- and TGN-mediated vesicle trafficking, providing evidence that the secretory pathway is involved (McFarlane et al., 2014).
The isolation of wax-deficient cer mutants led to the identification of a number of genetic loci required for the biosynthesis and deposition of wax (Koornneef et al., 1989). One of these mutants, cer11-1 (hereafter cer11), stood out from the rest because its wax deficiency cosegregated with a dwarf bushy stature resembling that of secretory trafficking mutants (e.g. echidna; Fig. 1A; Gendre et al., 2011).
Phenotypes of the cer11 mutant. A, Images of 5-week–old wild-type (Ler) and cer11 plants show that the cer11 plant is dwarf and bushy. Scale bar = 5 cm. B, Stem cuticular wax load determined by GC-FID. Wax analysis data are expressed as the mean percentage ± sd (n = 4; data shown is one of three independent experiments). Asterisk indicates that, statistically, this value was significantly different from wild type at P < 0.05 (Student’s t test). C, SEM images showing wax crystals deposited on the surface of wild-type and cer11 stems. Scale bar = 5 μm. D, Wild-type (Ler) and cer11 seeds after hydration with shaking in water or EDTA, and subsequent staining with ruthenium red. Lower magnification images show phenotypes of multiple seeds and higher magnification images show single representative seeds. Scale bar = 0.6 mm. E, Monosaccharide analysis of mucilage by HPAEC. The monosaccharide composition of mucilage extracted from seeds using or Na2CO3 or from AIR prepared from whole seeds was determined. Differences in rhamnose (Rha) and GalUA (GalA) content between wild type and cer11 are indicated by black arrows. Values are the mean ± sd (n = 4; independent seed lots for each genotype extracted and processed at the same time; data shown is one of three independent experiments) and are expressed as nmoles sugar normalized to milligrams of seed used for mucilage extraction or milligrams of AIR hydrolyzed (whole seed). Asterisks indicates that, statistically, this value was significantly different from wild type at P < 0.05 (Student’s t test). F, SEM images of wild-type and cer11 whole seeds (upper) and individual seed coat epidermal cells (lower). Columellae are indicated by white arrows. Scale bar = 20 µm. G, Light micrographs of cross sections through wild-type (upper) and cer11 (lower) seed coats stained with toluidine blue showing epidermal cells at 4, 7, 10, and 15 DPA. Development of the cer11 seed coat epidermal cells at 4, 7, and 10 DPA is indistinguishable from the wild type. However, at 15 DPA, cer11 columellae formation is not complete and the seed coat does not rupture upon hydration. Mucilage pockets are indicated by black arrowheads; a columella is indicated by a black arrow. Scale bar = 20 μm. WT, wild type.
In this study we show that cer11 wax deficiency cosegregates with a failure of seed coat epidermal cells to properly extrude pectinaceous mucilage. Like wax, mucilage deposition in the seed epidermis has been shown to be dependent on ECHIDNA function (McFarlane et al., 2013). A possible explanation for the complex cer11 phenotype is that the mutant is defective in secretion. In this study we investigate this hypothesis by determining the effect of the cer11 mutation on the trafficking of secretory GFP (secGFP), a secreted form of Aequorea victoria GFP that serves as a tool for genetic analysis of trafficking in plant cells (Batoko et al., 2000; Zheng et al., 2004). Further, we identify the protein encoded by CER11 using map-based cloning techniques. Our results reveal that CER11 encodes a phosphatase with a previously unknown role in secretion and show that C-TERMINAL DOMAIN PHOSPHATASE-LIKE (CPL)-mediated dephosphorylation is an essential step in cuticular wax deposition and seed coat epidermal cell wall formation.
RESULTS
The cer11 Mutant Is Defective in Wax Deposition
The cer11 wax-deficient mutant was identified by a visual screen for stem surface glossiness (Koornneef et al., 1989). Detailed analysis of stem and leaf wax by gas chromatography with flame ionization detection (GC-FID) demonstrated that the mutant accumulated only 53% of wild-type stem wax, and 75% of wild-type leaf wax (Rashotte et al., 2001). The cer11 mutant also exhibited severe wax deficiency under our growth conditions, as evidenced by its glossy, bright green stems (Supplemental Fig. S1A) and reduced stem wax accumulation reaching ∼50% of wild-type levels (Fig. 1B). In addition, scanning electron microscopy (SEM) showed that cer11 has fewer epicuticular wax crystals on the stem surface than wild type (Fig. 1C). Compositional analyses of the stem wax did not reveal any changes in the cer11 stem wax profile (Supplemental Fig. S1B).
Mucilage Release and Cell Wall Formation Are Defective in Seed Coat Epidermal Cells of cer11
During differentiation, seed coat epidermal cells synthesize and secrete a large volume of pectinaceous mucilage. Upon exposure of mature seeds to an aqueous solution the mucilage expands, ruptures the cell wall, and extrudes to form a gel-like capsule surrounding the seed (for review, see Haughn and Western, 2012; North et al., 2014). During our analysis of the cer11 mutant, we noticed that upon water imbibition the seed coat mucilage capsule was much smaller than that of wild-type seeds (Fig. 1D). After treatment with EDTA, a chelator that loosens mucilage pectin, a larger capsule was observed; however, it did not appear to be as big as wild type. Previous analyses of mutants with a similar phenotype have shown that the changes in mucilage capsule size can result from (1) loosening of the mucilage such that the adherent layer separates from the seed, (2) a compositional change in pectin that prevents mucilage extrusion, or (3) a decrease in the amount of pectin made and/or deposited. To distinguish among these possibilities, we quantified the mucilage monosaccharide composition by high-performance anion-exchange chromatography (HPAEC; Fig. 1E). The HPAEC data revealed that there was little qualitative difference in the spectrum of sugars found in cer11 and wild-type mucilage extracted from seeds using either water or Na2CO3, a mild base used to extract cell wall polysaccharides. However, the amount of the rhamnogalacturonan-I backbone sugars rhamnose and galacturonic acid (GalUA), the major monosaccharides of seed coat mucilage, was significantly decreased (Fig. 1E), suggesting that there is less extractable mucilage in cer11 seeds. To investigate this further, we next performed monosaccharide analysis on alcohol-insoluble residue (AIR) samples prepared from whole seeds including mucilage. As shown in Figure 1E, significantly less rhamnose and GalUA were obtained from whole cer11 seeds compared to wild type, indicating that less mucilage is present in cer11 seeds. As mucilage synthesis and secretion are likely to be interdependent, this result is consistent with impaired mucilage synthesis or secretion.
Examination of mucilage-producing seed epidermal cells of cer11 by SEM revealed that the columella, a volcano-shaped secondary cell wall, is not properly formed (Fig. 1F). Sectioning of developing seeds showed that there were no major anatomical differences between cer11 and the wild type during the early (4 days post anthesis [DPA]), middle (7 DPA), and late (10 DPA) stages of epidermal cell development, but that at maturity (15 DPA), the cer11 columella was hollow (Fig. 1G), causing it to collapse at the end of development when the seed coat desiccates. These data suggest that the deposition of columella secondary cell wall is slower in cer11 than in wild type and fails to complete development by seed maturity. In summary, seed coat mucilage deposition and release, as well as secondary cell wall formation, are impaired in cer11 seeds.
The cer11 Mutant Is Defective in Protein Secretion
Stunted growth of the cer11 mutant, reduced wax accumulation on the stem surface, reduced mucilage content in the seed epidermis, and delayed columella secondary cell wall deposition can all be explained by a defect in the secretory pathway. Many studies of secretory trafficking have taken advantage of secGFP, a visual marker transported from the ER to the apoplast where its green fluorescent signal is greatly diminished in the acidic apoplastic environment of plant cells (Batoko et al., 2000; Zheng et al., 2004). If secretion from the ER to the PM is inhibited, increased secGFP accumulation in the symplast will result in a strong fluorescence signal at the site of the block. To determine if the cer11 mutation affects secGFP movement through the secretory pathway to the PM, we introduced secGFP into wild-type and cer11 plants and compared its localization. The secGFP gene was transcribed in both wild-type and cer11 mutant backgrounds (Fig. 2A). However, as shown in Figure 2B, the cer11 mutant exhibited an enhanced GFP signal in the ER network and fusiform bodies compared to the wild type, indicative of compromised ER to PM trafficking.
Secretion of secGFP and MUM2-YFP is disrupted in cer11. A, RNA blot analysis of secGFP transcript levels in wild-type and cer11 seedlings expressing secGFP. The RNA blot was performed using total RNA from seedlings, with ribosomal RNA used as a control. B, Confocal laser scanning microscopy images of wild-type and cer11 hypocotyl cells expressing secGFP show increased levels of GFP fluorescence in the ER network and fusiform bodies in cer11. Scale bar = 10 μm. C to M, MUM2-YFP localization by confocal microscopy in wild type (C–F) and cer11 (G–M) seed coat epidermal cells at the 4-DPA (C, G), 7-DPA (D, H), and 10-DPA (F, J) seed coat developmental stages in transverse section. YFP-MUM2 localization in wild type (E) and the cer11 (I) seed coat epidermal cells at 7 DPA in longitudinal section. Overlay (M) of MUM2-YFP (K) and the PM dye FM4-64 (L), enlarged from the white box in (I). The developmental stages of the embryos are shown in the insets. Abbreviations: mucilage pocket (m), cytoplasmic column (cc), primary cell wall (pcw), secondary cell wall (scw). Scale bars = 20 μm. WT, wild type.
Secretion of MUM2, a Cell Wall-Modifying Enzyme, But Not CESA5 or Wax Export Proteins, Is Delayed in the cer11 Mutant
CER11 may be required for secretory trafficking of cargo such as wax molecules and cell wall components, and/or secretion of components of the molecular machinery involved in transport of wax and cell wall components. To investigate these two possibilities, we examined whether the cer11 mutation changed the localization of fluorophore-tagged proteins involved in wax export (CER5/ABCG12, Pighin et al., 2004; ABCG11, Bird et al., 2007; and LTPG, DeBono et al., 2009), and seed coat polysaccharide deposition (CESA5; Harpaz-Saad et al., 2011; Mendu et al., 2011; Sullivan et al., 2011) or modification (MUM2; Supplemental Methods; Dean et al., 2007; Macquet et al., 2007). No obvious differences in the localization of the wax export proteins or GFP-CESA5 were detected between cer11 and wild type by confocal microscopy (Supplemental Fig. S2), whereas the localization of MUM2-YELLOW FLUORESCENT PROTEIN (YFP) was altered in seed-coat epidermal cells (Fig. 2, C–M). In wild-type seeds at 4 DPA, YFP fluorescence in seed-coat epidermal cells was present in the primary cell wall, while no YFP signal was detected in cer11 cells (Fig. 2, C and G). At 7 DPA, MUM2-YFP fluorescence in the wild type was evenly distributed throughout the mucilage pocket and in the primary cell wall (Fig. 2, D and E), while the YFP signal in cer11 seeds was concentrated at the inner edge of the mucilage pocket close to the cytoplasmic column (Fig. 2, H and I). Longitudinal optical sections through the epidermal cells, combined with staining of the PM using the sterol dye FM4-64, confirmed the observation that the MUM2 protein in cer11 seeds was concentrated on the inner side of the mucilage pocket (Fig. 2, I, and K–M). Finally, at 10 DPA, the MUM2-YFP signal was detected in the columella secondary cell wall of both wild-type and cer11 seed coats, but the signal was much weaker in cer11 seeds (Fig. 2, F and J). Overall, the MUM2-YFP subcellular localization at different stages of seed-coat epidermal cell development is consistent with the hypothesis that secretion of the MUM2 protein in the cer11 mutant is delayed compared to the wild type, resulting in uneven distribution of MUM2 signal in the primary cell wall and mucilage pocket.
CER11 Gene Encodes CPL2
To determine the molecular identity of CER11 and obtain clues about its biochemical function, a positional cloning approach was used to isolate the CER11 gene. The cer11 mutant in the Landsberg erecta (Ler) background was crossed to the Columbia-0 wild type to generate a mapping population. Segregation analysis of the cer11 phenotype relative to known Simple Sequence Length Polymorphism markers in the F2 progeny revealed that the cer11 wax deficiency was linked to a ∼300-kb region between 21,140 bp and 325,700 bp on chromosome 5 that contains 87 genes (Supplemental Fig. S3A).
The cer11 mutant was generated by fast neutron mutagenesis (Koornneef et al., 1982), and therefore could carry a deletion or chromosomal rearrangement. As such, the mutation might be detectable by changes in PCR amplification relative to wild type. To test this possibility, PCR was used to amplify different chromosomal fragments within the 300-kb region of interest. Using this strategy, the gene At5g01800 was identified as having a breakpoint on chromosome 5 at position 308,246 in cer11. Furthermore, two additional breakpoints in the cer11 mutant were identified by inverse PCR, one on chromosome 1 at position 714,924 and the other on chromosome 5 at position 110,325 (Supplemental Fig. S3B). The identified chromosomal rearrangements include an inversion of 197,912 bp on chromosome 5 between positions 110,335 bp and 308,246 bp, and a reciprocal translocation between chromosomes 5 and 1 joining chromosome 5 from 110,325 bp to chromosome 1 at 714,927 bp and chromosome 1 from 714,924 bp to chromosome 5 at 308,246 bp (Supplemental Fig. S3B). Furthermore, two nucleotides (AA at chromosome 1 from 714,925 to 714,926 bp), nine nucleotides (TTACTTGAA at chromosome 5 from 110,326 to 110,334 bp), and two nucleotides (AG at chromosome 5 from 308,247 to 308,249 bp) were missing at the three break points, respectively (Supplemental Fig. S3B). The rearrangements were confirmed by PCR amplification (Supplemental Fig. S3C).
Three genes in cer11 were affected by this chromosomal rearrangement: At1g03060 (SPIRRIG, SPI), At5g01270 (C-TERMINAL DOMAIN [CTD] PHOSPHATASE-LIKE2 [CPL2]), and At5g01800 (SAPOSIN B DOMAIN-CONTAINING PROTEIN, SAPLIP). To determine which of these three genes is CER11, we examined the phenotypes of homozygous transfer DNA (T-DNA) insertion lines in SPI or CPL2 (no T-DNA mutant of SAPLIP was available), and carried out transgene complementation tests for CPL2 and SAPLIP.
The SPI gene was eliminated as a CER11 candidate because two T-DNA mutants in SPI (spi-1 and spi-2), with insertions in the 10th and 17th exon, respectively (Supplemental Fig. S4A), did not show any of the cer11 phenotypes (Supplemental Fig. S4, B–D).
To determine whether CPL2 or SAPLIP is CER11, genomic complementation constructs CPL2g and SAPLIPg were generated and transformed into the cer11 mutant. The CPL2g transgene rescued all the cer11 phenotypes including the dwarfism, the stem wax deficiency, the seed coat mucilage phenotype, and the altered columella morphology (Supplemental Fig. S5, A–E), whereas the SAPLIPg transgene did not. In addition, the high intracellular GFP fluorescence of the cer11 mutant carrying the secGFP transgene was suppressed by CPL2g, indicating that CPL2 is able to complement the cer11 secretory defect (Supplemental Fig. S5, F and G). These data strongly suggested that CER11 is CPL2.
Additional evidence that CPL2 is the CER11 gene was obtained through phenotypic analyses of plants homozygous for two SALK T-DNA alleles, cpl2-1 and cpl2-2 (Ueda et al., 2008), and a GABI-KAT T-DNA allele that we designate cpl2-3 (Fig. 3A). Wild-type CPL2 transcripts could not be detected in any of these cpl2 alleles (Fig. 3B). All three mutants were bushy and dwarf in stature, similar to the cer11 mutant (Fig. 3C); and they have reduced total wax loads on the stem surface (Fig. 3D) and exhibit a cer11-like defective mucilage phenotype (Fig. 3E). Defects in cpl2-1 and cpl2-2 seed mucilage were not as apparent as in cpl2-3, possibly due to the fact that cpl2-1 and cpl2-2 are partial loss-of-function alleles (Ueda et al., 2008). Moreover, the seeds of all three cpl2 alleles have collapsed columellae, like the cer11 mutant (Fig. 3F). In summary, all three cpl2 alleles phenocopy the cer11 mutant. On the basis of all the accumulated evidence, we conclude that CER11 is CPL2. We refer to this gene as CER11/CPL2 and rename the cer11-1 allele cer11-1/cpl2-4.
Phenotypes of the cpl2 mutants. A, The structure of the CER11/CPL2 gene (At5g01270) shows exons as black boxes, introns as solid lines, 5′-UTR as a white box, and 3′UTR as a black triangle. The locations of the T-DNA insertion of cpl mutants SALK_149234 (cpl2-1), SALK_059753 (cpl2-2), and GK-433F07 (cpl2-3) were mapped and are indicated by arrowheads. The location of the breakpoint in the cer11 mutant is indicated by a vertical black arrow. The locations of primers used for RT-PCR in (B) are indicated with short horizontal arrows. B, RT-PCR analysis of steady-state CER11/CPL2 transcript levels in wild type (Col-0) and mutant (cpl2-1, cpl2-2, and cpl2-3) leaves. RT-PCR was performed using total leaf RNA, and GAPC was used as a control. C, Images of 5-week–old wild-type (Col-0) and mutant (cpl2-1, cpl2-2, and cpl2-3) plants showing all the cpl2 mutant plants are dwarf and bushy compared to the wild type. Scale bar = 5 cm. D, Cuticular wax analysis for wild type and the mutant stems by GC-FID. Wax analysis data are expressed as the mean percentage ± sd (n = 4; data shown is one of three independent experiments). The asterisk indicates that, statistically, the mutant was significantly different from wild type (P < 0.05; Student’s t test). E, Mucilage extrusion from wild-type (Col-0) and cpl2-1, cpl2-2, and cpl2-3 seeds after hydration in water and staining with ruthenium red. Scale bar = 0.5 mm. F, SEM images of seed coat epidermis of seeds from wild type (Col-0) and mutants. Scale bar = 20 µm. WT, wild type.
CER11/CPL2 Is Expressed Throughout the Plant and CER11/CPL2 Protein Is Localized in the Nucleus and in the Cytoplasm
A previous study revealed that CER11/CPL2 encodes a CTD phosphatase, which can dephosphorylate the RNA polymerase II complex in vitro (Koiwa et al., 2004; Ueda et al., 2008). However, the role of CER11/CPL2 in secretion, wax export, and cell wall formation has not been previously described. As a first step toward investigating the CER11/CPL2 function in this context, the transcription profile of the CER11/CPL2 gene in Arabidopsis (Arabidopsis thaliana) was examined by reverse transcription quantitative PCR (RT-qPCR). The RT-qPCR results demonstrate that the CER11/CPL2 gene is ubiquitously expressed in all plant organs tested, including leaf, stem, flower, root, seed, and developing seedlings (Fig. 4A). The highest levels of expression were detected in leaves, seeds, and 10-d–old seedlings.
Expression pattern of the CER11/CPL2 gene and localization of the CER11/CER2-GFP protein and its colocalization with VHA-C-YFP. A, Expression levels of the CER11/CPL2 gene in different organs. Total RNAs from leaf, stem, flower, root, 10-d–old seedling, and developing seeds at 4 DPA, 7 DPA, and 10 DPA from wild type (Col-0) were analyzed for CER11/CPL2 and GAPC gene expression by RT-qPCR. CER11/CPL2 gene expression was normalized to GAPC expression values. Error bars represent sd (n = 4). B and C, Confocal microscopy of CER11/CPL2-GFP localization in stem epidermal cells (B) and seed coat epidermal cells at 7 DPA (C). A representative embryo for the developmental stage is shown in the inset. Scale bars = 20 µm. d and E, The dwarf phenotype of the VHA-C mutant det3-1 is complimented by VHA-C-YFP. In 6-week–old plants (D), the short stature of det3 is returned to wild-type height when VHA-C-YFP is introduced (E). Letters show statistical significance for n = 5, one-way ANOVA followed by Tukey's Honestly Significant Difference posthoc analysis performed using SPSS 25 software, P < 0.0001. Scale bar = 7 cm. F to K, Colocalization of CPL2-RFP and VHA-C-YFP in stem epidermal cells (F–H) and seed coat epidermal cells (I–K). Scale bar = 20 μm. WT, wild type.
To determine the subcellular localization of the CER11/CPL2 protein, a translational fusion between the CER11/CPL2 genomic sequence and GFP (CER11/CPL2g-GFP) was transformed into cer11-1/cpl2-4 plants and shown to rescue the mutant phenotype (Supplemental Fig. S5H). CER11/CPL2-GFP localization in the stem (Fig. 4B) and the seed coat epidermal cells (Fig. 4C) of the complemented plants was then examined by confocal microscopy. In both organs, the CER11/CPL2-GFP expression was detected in the nucleus and in the cytoplasm, which is consistent with a previous study using a CER11/CPL2-GFP transgene driven by the CaMV35S promoter (Koiwa et al., 2004).
What Is the Substrate of CER11/CPL2?
To detect the potential substrate(s) of CER11/CPL2, we carried out a yeast-2-hybrid screen of an Arabidopsis complementary DNA (cDNA) library using BD-CER11/CPL2 as bait, and identified 23 potential interactors (Supplemental Table S1). Two of them, CELL WALL-PLASMA MEMBRANE LINKER PROTEIN (CWLP, encoded by At3g22120) and VACUOLAR-TYPE H+-ATPASE SUBUNIT C (VHA-C/DET3, encoded by At1g12840) were further investigated because of their potential link to secretory trafficking.
Because the prey clones of CWLP only contained the 3′ end of the gene, a construct containing a full-length CWLP (FL-CWLP) conjugated with an activation domain (AD) was generated to verify whether CWLP is a true CER11/CPL2 interactor in the yeast-2-hybrid system. As shown in Supplemental Fig. S6A, the C-terminal CWLP (CWLP166–334), but not the FL-CWLP, can bind to CER11/CPL2 in yeast (Saccharomyces cerevisiae). We also tested whether CWLP and CER11/CPL2 are colocalized in vivo. For this experiment, an full-length CWLP with YFP inserted between the signal peptide and the remainder of the coding region and driven by the CWLP native promoter (CWLPp::YFP-CWLP), as well as the CER11/CPL2-RED FLUORESCENT PROTEIN (RFP) driven by the CER11/CPL2 native promoter (CER11/CPL2g-RFP), were coexpressed in wild-type (ecotype Columbia-0 [Col-0]) plants. Visualization of the YFP-CWLP and CER11/CPL2-RFP fluorescence patterns in the stem and the seed coat epidermal cells by confocal microscopy showed that YFP-CWLP is not expressed in the stem or seed coat epidermis (Supplemental Fig. S6, B–D and H–J), but is observed in cells in the leaf epidermis (Supplemental Fig. S6, E–G) and seed palisade cells (Supplemental Fig. S6, K–M).
In addition, two T-DNA mutant alleles of CWLP, cwlp-1 and cwlp-2, were examined for their wax and seed coat mucilage phenotypes. Insertions in both cwlp-1 and cwlp-2 were located in exons, and wild-type CWLP transcripts could not be detected in either cwlp-1 or cwlp-2 homozygous lines (Supplemental Fig. S7, A and B). The GC-FID wax analysis data showed that stem wax loads of the cwlp mutants were similar to the wild type (Supplemental Fig. S7C), while ruthenium red staining of cwlp mutant seeds immersed in water demonstrated that they released normal amounts of mucilage (Supplemental Fig. S7D). Taken together, these data indicate that CWLP is not a substrate of CPL2 during cuticular wax export or seed coat mucilage release.
To determine whether VHA-C and CPL2 colocalize in Arabidopsis epidermal cells, VHA-C genomic sequence was fused to the YFP reporter gene (VHACg-YFP) and coexpressed with CER11/CPL2g-RFP in wild-type (Col-0) plants. Proper function of the YFP-tagged VHA-C was demonstrated by complementation (Fig. 4, D and E) of the dwarf phenotype of the VHA-C mutant de-etiolated3-1 (det3-1; Schumacher et al., 1999). Detection of fluorescence by confocal microscopy established that VHA-C-YFP and CER11/CPL2-RFP colocalize in the nucleus and cytoplasm of the stem and seed coat epidermis (Fig. 4, F–K).
To confirm binding of VHA-C to CER11/CPL2 in yeast, the N-terminal fragment (VHA-C1–175), C-terminal fragment (VHA-C176–375), or full-length VHA-C protein conjugated with ADs were coexpressed with BD-CER11/CPL2 in yeast cells. This revealed that only the C-terminal VHA-C fragment could bind to CER11/CPL2 in the yeast-2-hybrid system (Fig. 5A). To further delineate the binding domain of CER11/CPL2, a series of truncations of CER11/CPL2 were generated and binding activity toward the C-terminal of VHA-C was examined by yeast-2-hybrid assay. As shown in Figure 5B, the N-terminal fragment CER11/CPL21–599 is able to bind to VHA-C, whereas the shorter fragment CER11/CPL21–392 is not. To assess whether these two proteins could interact in plants as well as a heterologous system, a split-luciferase assay was performed in Nicotiana benthamiana. Full-length VHA-C was fused with the C-terminal luciferase (cLuc-VHA-C). Three N-terminal luciferase constructs were generated, corresponding to those used in the yeast-2-hybrid experiments (CER11/CPL2-nLuc, CER11/CPL21–599-nLuc, and CER11/CPL21–392-nLuc). Based on the previous experiment, CER11/CPL2-nLuc and CER11/CPL21–599-nLuc were hypothesized to interact with cLuc-VHA-C, while CER11/CPL21–392-nLuc was not. To test for nonspecific interactions, full-length CER11/CPL2 was also coexpressed with free cLuc or MITOGEN ACTIVATED PROTEIN KINASE6 (MPK6), a similarly localized protein not expected to interact with CER11/CPL2 (Smékalová et al., 2014; Zhang et al., 2016). After infiltration into N. benthamiana leaves, strong luciferase activity was detected in cells coexpressing cLuc-VHA-C with either full-length CER11/CPL2-nLuc or CER11/CPL21–599-nLuc, and there was a substantial decrease in luciferase activity for CER11/CPL21–392-nLuc, the shortest CPL2-nLuc construct assayed (Fig. 5C). Luciferase activity in the negative controls was consistent with noninteraction of the protein pairs, and expression of MAPK6-cLuc was confirmed by western blot (Supplemental Fig. S8). Taken together, these data indicate that there is a robust interaction between VHA-C and CER11/CPL2 in planta.
CPL2 interacts with VHA-C and can dephosphorylate VHA-C in vitro. A, Yeast-2-hybrid assay shows that the C-terminal domain of VHA-C (AD-VHA-C176–375) can bind to CPL2 (BD-CPL2). AD, GAL4 activation domain fusions; BD, GAL4 DNA binding domain fusions. Serial 1:10 dilutions are shown. B, Yeast-2-hybrid assay shows that the N-terminal domain of CER11/CPL2 (CPL21–599) is sufficient for VHA-C binding. The domain structure of each of the CPL2 truncations tested are shown, with two major functional domains indicated. Catalytic FCP1 homology domain (blue); dsRNA binding motif (red). C, Split luciferase assay. Luciferase activity in ∼2-cm regions of N. benthamiana leaves coinfiltrated with Agrobacterium strains containing various nLuc and cLuc constructs. cLuc-VHA-C was tested for interaction with each of CER11/CPL2-nLuc (full length, FL), CER11/CPL21–392-nLuc (1–392), CER11/CPL21–599-nLuc (1–599), and nLuc, as well as between CER11/CPL2-nLuc (FL), and MPK6-cLuc or cLuc. Robust luciferase activity was detected when cLuc-VHA-C was coexpressed with CER11/CPL2-nLuc (FL) or CER11/CPL21–599-nLuc. Results are representative of 10 replicates (Reps), from two experiments. D, In vitro dephosphorylation assay. VHA-C-YFP was extracted from 10-d–old transgenic seedlings expressing VHACp::VHAC-YFP and incubated with buffer-only, 6×His-CER11/CPL2, or LPP. The total VHA-C-YFP and phosphorylated VHA-C-YFP (Phospho-VHA-C-YFP) were analyzed by western blotting with anti-GFP antibody and anti-Phospho-Ser/Thr antibody, respectively. E, Measurement of the band intensity of Phospho-VHA-C-YFP relative to Total VHA-C-YFP in (D) was used for quantitation.
VHA-C Protein Localization Is Not Altered in the cer11/cpl2 Mutant
Even though VHA-C and CER11/CPL2 can interact in vivo, the effect of their interaction on the function of VHA-C is not known. One possibility is that binding to CER11/CPL2 might change VHA-C localization in the cell. To test this hypothesis, the VHACg-YFP construct was transformed into the cer11-1/cpl2-4 mutant, and VHA-C-YFP subcellular localization in both wild-type and cer11/cpl2 genetic backgrounds was evaluated. To ensure differences in the location of the VHAC-YFP insertion did not create differences in expression, protein levels of VHAC-YFP were examined and found to be comparable in the wild-type and cer11/cpl2 backgrounds (Supplemental Fig. S9). Confocal microscopy of VHA-C-YFP in the stem and seed coat epidermis showed similar localization between the cer11/cpl2 mutant and wild-type plants (Supplemental Fig. S9).
CER11/CPL2 Can Dephosphorylate VHA-C In Vitro
To assess if VHA-C is a substrate of CER11/CPL2, an in vitro dephosphorylation assay was carried out. Phosphorylated VHA-C-YFP was prepared by immunoprecipitation with an anti-GFP antibody from 10-d–old transgenic seedlings expressing VHACg-YFP, and incubated in the presence of buffer alone, purified 6×His-CER11/CPL2, or a Lambda protein phosphatase (LPP) positive control. Both CER11/CPL2 and LPP dephosphorylated VHA-C under our experimental conditions (Fig. 5, D and E).
The det3-1 Mutation does Not Affect Stem Wax and Seed Coat Mucilage Deposition
V-ATPase has been shown to be essential for secretory trafficking, and null alleles of Arabidopsis V-ATPase genes cause gametophytic or embryo lethality (Dettmer et al., 2006). The det3-1 mutant allele of V-ATPase subunit C is a weak allele caused by a T/A transversion in the first intron of the VHA-C gene (Fig. 6A). It results in reduced transcript accumulation, together with the presence of an additional, longer transcript, indicating that the det3-1 mutation alters transcript splicing (Fig. 6B; Schumacher et al., 1999). The det3-1 mutant exhibits a 2-fold decrease in V-ATPase activity, and an associated defect in the TGN/early endosome causing severe growth retardation (Fig. 6C; Schumacher et al., 1999; Brux et al., 2008). Because null mutants in V-ATPase subunits are not viable, we attempted to confirm the involvement of V-ATPase in wax and seed coat mucilage secretion using the det3-1 allele. However, our analysis of the stem wax load and seed coat mucilage released upon immersion in water or EDTA did not reveal any differences between the det3-1 and wild type (Fig. 6, D and E). Furthermore, the columella structure in the mutant seed coat epidermal cells was indistinguishable from that in the wild type (Fig. 6F).
Wax and seed coat phenotypes of the det3-1 mutant. A, The structure of the VHA-C gene (At1g12840) shows exons as black boxes, introns as black lines, 5′-UTR as a white box, and 3′UTR as a black triangle. The location of the point mutation in det3-1 is indicated by a vertical arrow. The locations of primers used for RT-PCR in (B) are indicated with short horizontal arrows. B, RT-PCR analysis of steady-state VHA-C transcript levels in wild type (Col-0) and the det3-1 leaves. RT-PCR was performed using total leaf RNA, and the expression level of GAPC was used as a control. Two different sizes of VHA-C transcripts were detected in the det3-1, which are indicated by black arrows. C, A comparison of 5-week–old wild-type (Col-0) and the det3-1 plants shows that the det3-1 mutation causes a severe growth defect. Scale bar = 5 cm. D, Cuticular wax analysis for 5-week–old wild-type and the det3-1 stems by GC-FID. Wax analysis data are expressed as the mean percentage ± sd (n = 4; data shown is one of three independent experiments). E, Mucilage extrusion from wild-type and the det3-1 seeds after hydration in water or EDTA, and staining with ruthenium red. Scale bars = 0.5 mm. F, SEM images of seed coat epidermis of seeds from wild type and the mutant (det3-1). Scale bars = 20 μm. WT, wild type.
DISCUSSION
CER11/CPL2 Plays an Important Role in Secretion
In plants, export of extracellular matrix components such as cuticular waxes and cell wall polysaccharides relies on secretory trafficking, but molecular details of their secretion remain largely obscure. In this study, we characterize the cer11/cpl2 mutant, which exhibits dwarfism, reduced stem cuticular wax deposition, aberrant seed coat mucilage extrusion, and delayed secondary cell wall formation. In addition, cer11/cpl2 mutation disrupts secretory protein trafficking of the cell wall-modifying enzyme MUM2 and secGFP. The cer11/cpl2 phenotype suggests that CER11/CPL2 plays an important role in secretion. The pleiotropic nature of the cer11/cpl2 phenotype may be due to impaired secretion of the extracellular matrix materials (cargo) and/or proteins required for the translocation and processing of matrix materials.
The fact that the cer11/cpl2 mutants are viable indicates that CER11 is not essential for secretion. Instead, it may play a regulatory role at times or in cell types where a high level of secretion is required. Differential regulation of secretion has been previously documented during cell plate formation, cytokinesis, pollen tube growth, and root development (Wang et al., 2016; Li et al., 2017; Ravikumar et al., 2018). In support of this hypothesis, we detected a delay in MUM2 secretion to the apoplast in cer11/cpl2. This finding points to inefficient secretory trafficking as the underlying cause of the complex cer11/cpl2 mutant phenotype. Such inefficiency would be expected to impact wax and cell wall deposition, as well as cell expansion. Indeed, defects in secretion are known to have a variety of effects on cell biology that lead to stunted growth (e.g. det3-1, Schumacher et al., 1999; echidna, Gendre et al., 2011).
CER11/CPL2 Interacts with VHA-C and Can Dephosphorylate It In Vitro
Cloning of the CER11 gene revealed that it encodes CPL2, a predicted CTD phosphatase containing an FCPH catalytic domain homologous to the yeast FCP1 CTD phosphatase (Koiwa et al., 2002, 2004). In Arabidopsis, more than 20 genes are predicted to encode plant CTD phosphatases (Bang et al., 2006). Among them, CPL1/FIERY2 (FRY2) and CER11/CPL2 are unique to plants and belong to the CPL1-like class because, in addition to the FCPH domain, these proteins also contain two or one double-stranded RNA-binding motifs, respectively (Koiwa et al., 2004). CPL1/FRY2 and CER11/CPL2 have in vitro dephosphorylation activity (Fig. 5, D and E; Koiwa et al., 2004), and were shown to be required for plant growth, response to abiotic stresses such as cold, salt, and abscisic acid, and micro RNA biogenesis (Koiwa et al., 2002; Ueda et al., 2008; Manavella et al., 2012). The male gamete lethality of the cpl1cpl2 double mutant indicates that these two proteins function redundantly for at least some essential processes (Koiwa et al., 2004).
However, the phenotypes of cer11/cpl2 mutants are not identical to cpl1/fry2, and include additional leaf morphology phenotypes, early flowering (Ueda et al., 2008), dwarfism, and stem wax-deficiency (this study), which were not observed in the cpl1/fry2. Furthermore, CPL1 protein is present only in the nucleus, while CPL2 is localized in both the nucleus and cytoplasm (Fig. 4, B and C; Koiwa et al., 2004). Taken together, these results suggest that CPL1/FRY2 and CER11/CPL2 are only partially redundant, and that CER11/CPL2 has functions that are distinct from CPL1/FRY2. Our data demonstrate that this is indeed the case and provide evidence that CER11/CPL2 also plays an important role in general secretory trafficking.
To elucidate CER11/CPL2 function in secretion, we carried out a yeast-2-hybrid screen in an attempt to identify the CER11/CPL2 substrate(s). We recovered a number of potential CER11/CPL2 binding partners, including the VHA-C subunit of the vacuolar H+-ATPase (V-ATPase), a multisubunit enzyme with a well-established role in secretion (Gaxiola et al., 2007). V-ATPase was of particular interest because it not only functions in acidification of the endomembrane compartments and maintains the proton gradient across membranes, but is also directly involved in secretory transport from the TGN (Dettmer et al., 2006; Schumacher and Krebs, 2010). Besides the demonstrated interaction of CER11/CPL2 with VHA-C in yeast, several additional lines of evidence suggested that the VHA-C may be the substrate of the CER11/CPL2 phosphatase. First, results from the binding study with truncated fragments indicate that CER11/CPL2 interacts with VHA-C in the same region that contains the FCPH domain (Fig. 5B). Second, VHA-C interacts with CER11/CPL2 in planta (Fig. 5C). Third, VHA-C is coexpressed and colocalizes with CER11/CPL2 in the stem and seed coat epidermis (Fig. 4, F–K). Fourth, and finally, CER11/CPL2 can dephosphorylate VHA-C in vitro as shown by a direct assay (Fig. 5, D and E). We therefore hypothesized that CER11/CPL2 may influence secretion, at least in part, by affecting the phosphorylation state of VHA-C. VHA-C is known to be phosphorylated by a With No Lys Kinase8 (Hong-Hermesdorf et al., 2006), but the effect of phosphorylation on the function of VHA-C has not been determined. In yeast and insect cells, phosphorylation and dephosphorylation control reversible assembly of the V-ATPase, a process that has been suggested to be an important mechanism controlling V-ATPase activity (Toei et al., 2010). Similarly, phosphorylation may also be involved in modulating the activity level of plant V-ATPase in a tissue-specific manner, during development, or in response to changing environmental conditions.
VHA-C Mutant det3-1 Does Not Exhibit Wax- or Seed Coat-Related Phenotypes
To directly assess if the detected interaction between CER11/CPL2 and VHA-C is indicative of V-ATPase involvement in secretion of cuticular waxes and seed coat epidermal cell wall components, we examined the phenotypes of det3-1, the only known viable mutant of VHA-C (Cabrera y Poch et al., 1993). The det3-1 plants exhibit stunted growth (Fig. 6C; Cabrera y Poch et al., 1993), with hypocotyls that are considerably shorter than the wild type. The dwarfism of det3-1 is a consequence of defective cell expansion resulting from reduced V-ATPase activity in the TGN (Schumacher et al., 1999; Brux et al., 2008; Luo et al., 2015). The det3-1 plants also have cellulose synthesis defects (Cano-Delgado et al., 2003), and the mutant is hypersensitive to the cellulose synthesis inhibitor isoxaben (Brux et al., 2008). Isoxaben treatment led to a rapid loss of CESA6 from the PM, suggesting that isoxaben may affect trafficking of CESAs to the PM (Paredez et al., 2006), leading to reduced cellulose synthesis and decreased cell expansion. However, cuticular wax deficiency, defects in seed coat mucilage release, and alterations in seed coat epidermal cell wall formation have not been described for det3-1. Our analysis of det3-1 indicates that this mutant does not have any obvious alterations in wax accumulation or cell wall deposition in seed coat epidermal cells as observed in the cer11/cpl2 mutant. This may be due to the fact that det3-1 still has 40% of wild-type V-ATPase activity (Schumacher et al., 1999), which may be sufficient for normal cuticular wax accumulation and seed coat development. Another possibility is that CER11/CPL2-mediated dephosphorylation of VHA-C is not required for trafficking of cuticular waxes, seed coat epidermal cell wall components, secGFP, or MUM2-GFP. Unique wax and cell wall-related phenotypes observed in cer11/cpl2 mutants, but not det3-1, suggest that there are other secretion-related proteins besides VHA-C whose dephosphorylation by CER11/CPL2 is required for cuticle/cell wall formation in the plant epidermis.
In summary, our work provides in vivo demonstration of the importance of CER11/CPL2-dependent dephosphorylation for efficient secretory trafficking and deposition of extracellular matrix constituents in developing epidermal cells. Identification of CER11/CPL2 substrate(s) is the critical next step that will allow us to establish the role of CER11/CPL2 and investigate how changes in phosphorylation state of its substrate(s) affect the flux of cargo/proteins down the secretory pathway.
MATERIALS AND METHODS
Plant Material and Growth Conditions
Arabidopsis (Arabidopsis thaliana) ecotypes Col-0, Landsberg erecta (Ler), and the cer11 mutant (Koornneef et al., 1989) were used in this study. SALK T-DNA insertion lines (Alonso et al., 2003), SALK_149234 (cpl2-1, Ueda et al., 2008), SALK_059753 (cpl2-2, Ueda et al., 2008), SALK_131421 (spi-1), and SALK_116367 (spi-2) were obtained from the Arabidopsis Biological Resource Center (www.arabidopsis.org). The T-DNA lines, GK-433F07 (cpl2-3), GK-382D01 (cwlp-1), and det3-1 (Cabrera y Poch et al., 1993) were obtained from The Nottingham Arabidopsis Stock Centre (www.arabidopsis.info). The SK T-DNA line SK_31605 (cwlp-2) was obtained from Agriculture and Agri-Food Canada (http://aafc-aac.usask.ca/FST/, Robinson and Parkin, 2009). Homozygous lines for each gene were identified by genotyping using primers listed in Supplemental Table S2. The MUM2-YFP line was generously provided by Dr. Erin Gilchrist (Haughn Lab, University of British Columbia; Supplemental Methods). The secGFP, CER5-GFP, and GFP-CESA5 lines were previously described (Batoko et al., 2000; Pighin and Zheng et al., 2004; Bischoff et al., 2011). The YFP-ABCG11 and YFP-LTPG lines were kindly provided by Dr. Lacey Samuels (University of British Columbia, BC, Canada; Bird et al., 2007; DeBono et al., 2009). All of these marker lines, and MUM2-YFP, were crossed into cer11. Seeds were germinated on Arabidopsis thaliana medium (Haughn and Somerville 1986) supplemented with agar (7 g/L) and appropriate antibiotics. The 7- to 10-d–old seedlings were transferred to soil and grown at 20°C under continuous light with a light intensity of 100 μmol photons m−2 s−1.
Plasmid Construction and Plant Transformation
DNA fragments containing CER11/CPL2g, CER11/CPL2g-RFP, SAPLIPg, CWLPp::YFP-CWLP, VHACg-YFP, CER11/CPL2-nLuc, CER11/CPL21–599-nLuc, CER11/CPL21–392-nLuc, and cLuc-VHA-C were generated and introduced into binary vectors as described in Supplemental Methods. Plasmids were electro-transformed into Agrobacterium tumefaciens GV3101 cells and Arabidopsis wild-type (Col-0) plants or cer11 mutant plants transformed by the floral spray method (Chung et al., 2000). The T1 transgenic seeds were screened on Arabidopsis thaliana medium (Haughn and Somerville, 1986) supplemented with agar (7 g/L) and appropriate antibiotics.
DNA fragments for the construction of CPL2, CWLP, and VHA-C yeast-2-hybrid plasmids were amplified from cDNA as described in Supplemental Methods. To generate the construct for 6×His-CER11/CPL2 protein expression and purification, the fragment of CER11/CPL2 cDNA was released from the BD-CER11/CPL2 construct and ligated into the EcoRI and SalI sites of the vector pET28a.
Wax Analysis by GC-FID
Wax analysis was done on the top 10 centimeters of 4- to 5-week–old primary inflorescence stems as outlined in Supplemental Methods.
Monosaccharide Analysis by HPAEC
Monosaccharide composition of cell wall material from mucilage extracted with water or Na2CO3 and whole seeds was determined by HPAEC using a protocol similar to Dean et al. (2007) and described in detail in Supplemental Methods.
Map-Based Cloning
To clone the CER11 gene, the cer11-1 mutant in the Ler background was crossed to the Columbia-0 wild type to generate a mapping population. Twenty-one F2 generation plants with cer11 wax phenotype were used for rough mapping, and 1,100 F2 generation plants with cer11 wax phenotype were used for fine mapping. The four insertion-deletion markers used for fine mapping on chromosome 5 are shown in Supplemental Figure S3. They are M21140 at position 21,140 bp, M82750 at position 82,750 bp, F7A7-1 at position 211,000 bp, and T20L15 at position 325,700 bp. The primer sequences are shown in Supplemental Table S2.
Inverse PCR
One microgram of genomic DNA from the wild type (Ler) or cer11 was digested by TaqI restriction enzyme at 65°C for 2 h, followed by the inactivation of the enzyme at 80°C for 20 min. After purification, the digested DNAs were re-ligated with T4 DNA ligase at 18°C overnight. The self-ligated DNAs were used as templates for inverse PCRs with primers annealing to the known sequences close to the chromosome break point.
RT-qPCR
Rosette leaves, stems, flowers, roots, seedlings, and developing seeds of the wild type (Col-0) were collected and immediately frozen in liquid nitrogen. Total RNA was isolated by using TRIZOL Reagent (Invitrogen Life Technologies), and reverse transcription was performed by using iScript cDNA Synthesis Kit (Bio-Rad). Gene-specific primers used in RT-qPCR are listed in Supplemental Table S2. iQ SYBR Green Supermix (Bio-Rad) was used to perform RT-qPCR in an iQ5 Multicolor Real-Time PCR Detection System as specified by the manufacturer (Bio-Rad, www.bio-rad.com).
Microscopy
Confocal Microscopy
Subcellular localization was carried out on a model no. UltraviewVoX spinning disk confocal microscope (Perkin Elmer) using glycerol immersion lens for live imaging. Dissected segments from the apical ∼3 cm of stem or developing seeds were mounted in distilled water and immediately imaged. The cer11 developing seeds expressing MUM2-YFP were stained with 5 μg/mL of FM4-64 (Molecular Probes) for 10 min. GFP was detected using a 488-nm laser with a 507- to 543-nm filter, YFP was detected using a 514-nm laser with a 525- to 555-nm filter, and RFP and FM4-64 were detected using a 561-nm laser with a 570- to 620-nm filter. The confocal images were taken using the software Volocity (Perkin Elmer) and edited using the software Fiji (https://imagej.net/Fiji).
SEM
Dry seeds or segments from the apical 1 cm of dry stem were mounted onto stubs and sputter-coated with gold particles for 10 min in a model no. SEM Prep 2 sputter coater (Nanotech). The coated samples were viewed under a model no. S4700 field emission SEM (Hitachi) at an accelerating voltage of 5 kV and a working distance of 12 mm.
Mucilage Staining
Dry seeds were hydrated in water or 50 mm of EDTA with shaking at room temperature for 1 h, and washed with water twice before staining with 0.01% (w/v) aqueous solution of ruthenium red (Sigma-Aldrich) for 1 h. The low magnification images were taken with a model no. SZX16stereomicroscope (Olympus) and the high magnification images were taken with an AxioScop 2 light microscope (Carl Zeiss).
Resin Embedding and Sectioning
Developing seeds from wild-type (Ler) and cer11 plants were dissected and fixed with 3% (v/v) glutaraldehyde (Canemco) in 0.1 m of P buffer overnight at 4°C. The samples were then postfixed in 1% (v/v) osmium tetroxide in 0.05 m of P buffer for 2 h After dehydration in a series of increasing concentrations of ethanol, the seeds were embedded in Spurr’s resin. The embedded samples were sectioned (50 μm) using an Ultracut E microtome (Reichert) and stained with 1% (w/v) toluidine blue. Images were taken under a model no. AxioSkop 2 light microscope (Carl Zeiss).
Yeast-2-hybrid Screen
Yeast-2-hybrid screening was carried out using the GAL4-based yeast-2-hybrid system. The CER11/CPL2 was fused to the C-terminal end of the GAL4 promoter-binding domain (BD-CER11/CPL2). The BD-CER11/CPL2 as bait was introduced into the yeast strain YPH1347 (kindly provided by Dr. Xin Li, University of British Columbia, BC, Canada), in which GAL4 promoter drives the His synthetic gene, and used to screen an Arabidopsis cDNA expression library (kindly provided by Dr. Yuelin Zhang, University of British Columbia, BC, Canada) encoding Arabidopsis proteins as C-terminal fusions to GAL4 transcription AD. Putative CER11/CPL2 interacting proteins were selected based on His prototrophy.
Split-Luciferase Assay
The split-luciferase assay system and appropriate controls have been described in Chen et al. (2008) and Gehl et al. (2011). Briefly, CER11/CPL2-nLuc, CER11/CPL21–599-nLuc, CER11/CPL21–392-nLuc, cLuc-VHA-C, MPK6-cLUC (Smékalová et al., 2014; Zhang et al., 2016), and empty vectors were transformed into A. tumefaciens GV3101 cells. The Agrobacterium cells were grown to log phase, then collected and resuspended in an infiltration medium (1× Murashige and Skoog, 10 mm of MES at pH 5.6, and 150 µM of acetosyringone) to a final concentration of OD600 = 0.6. The 3-4–week-old Nicotiana benthamiana leaves were infiltrated using 1-mL syringes with different combinations of bacterial suspensions. The plants were incubated at 22°C in the dark for 48 h. 1 mm of d-Luciferin (GoldBio) dissolved fresh in water, 10 mm of MgCl2, and 10 mm of MES at pH 5.6 were infiltrated into the Agrobacterium-infiltrated leaves. Entire leaves were imaged every 30 s for 15 min with a CCD camera that had a low-light imaging system (ChemiDocTM XRS+; Bio-Rad; http://www.bio-rad.com/), using 3×3 binning settings for all images.
Protein Expression and Purification
The 6×His-CER11/CPL2 was expressed in Escherichia coli Rosetta DE3 (Novagen). The 6×His-CER11/CPL2 was purified under denaturing conditions with HisPur Ni-NTA Resin (Thermo Fisher Scientific) and eluted in lysis buffer (20 mm of sodium P, 300 mm of sodium chloride, and 6 M of urea at pH 7.4) supplemented with 250 mm of imidazole. The purified protein was renatured by buffer exchange to P-buffered saline + 10% (v/v) glycerol. LPP was obtained from New England Biolabs.
In Vitro Dephosphorylation Assay
To obtain phosphorylated VHA-C protein, 0.5 g of 10-d–old transgenic seedlings expressing VHACp::VHAC-YFP was lysed in extraction buffer (50 mm of Tris-HCl, 10 mm MgCl2, 500 mm NaCl, 0.1% [w/v] NP40, 1 mm phenylmethylsulfonyl fluoride, 1× protease inhibitor, and 1× PhosSTOP phosphatase inhibitor; Roche). Insoluble matter was removed by centrifugation (15,000g, 15 min at 4°C). Protein concentration was determined by Protein Assay Reagent (Bio-Rad). A quantity of 1.2 mg of total proteins was added to 1.2 μg of Anti-GFP antibody (Roche) coupled to Protein A/G PLUS-Agarose (Santa Cruz Biotechnology) and incubated overnight at 4°C. After washing with extraction buffer without protease inhibitor and phosphatase inhibitor, the agarose beads were aliquoted equally into three tubes and incubated with phosphatase buffer (1× PMP buffer and 1 mm of MnCl2; New England Biolabs) only, 1 μg of 6×His-CER11/CPL2, or 1 μg of LPP as a positive control at 30°C for 2 h. The reaction was stopped by adding SDS sample buffer. The total VHA-C-YFP protein and phosphorylated VHA-C-YFP protein were analyzed by western blotting with Anti-GFP antibody (Roche) and Anti-Phospho-Ser/Thr antibody (BD Biosciences), respectively. The band intensity ratio of the protein bands was quantified in the software Photoshop (Adobe).
Statistical Analyses
Analysis of wax and monosaccharide data, including calculation of sd (sd) and Student’s t test, was performed using the software Excel 2007 (Microsoft). One-way ANOVA and Tukey’s Honestly Significant Difference posthoc analysis to compare plant heights was performed in the software SPSS 25 (IBM).
Accession Numbers
Sequence data from this article can be found in the GenBank/European Molecular Biology Laboratory databases under the following accession numbers: CER11/CPL2 (At5g01270), VHA-C/DET3 (At1g12840), MUM2 (At5g63800), CER5/ABCG12 (At1g51500), ABCG11 (At1g17840), CESA5 (At5g09870), LTPG (At1g27950), SPI (At1g03060), SAPLIP (At5g01800), CWLP (At3g22120).
Supplemental Data
The following supplemental materials are available.
Supplemental Figure S1. Phenotypes of the cer11-1 mutant.
Supplemental Figure S2. Localization by confocal microscopy of CER5-GFP, YFP-ABCG11, and YFP-LTPG in wild-type and mutant (cer11-1) stem epidermal cells, and CESA5-GFP localization in wild-type and mutant (cer11-1) seed coat epidermal cells at 7 DPA.
Supplemental Figure S3. Chromosomal rearrangements identified in the cer11-1 mutant.
Supplemental Figure S4. Structure of the SPI gene and phenotypes of spi-1 and spi-2 mutants.
Supplemental Figure S5. CPL2p::CPL2, but not SAPLIPp::SAPLIP, can complement the phenotypes of the cer11-1 mutant.
Supplemental Figure S6. CER11/CPL2 and CWLP do not interact in the epidermis.
Supplemental Figure S7. Phenotypes of the cwlp mutants.
Supplemental Figure S8. Western blot of cLuc proteins from N. benthamiana leaf tissue used in the split luciferase assay.
Supplemental Figure S9. VHA-C-YFP localization in wild-type and cpl2-3 mutant stem and seed coat epidermal cells, and protein levels in leaves.
Supplemental Table S1. Genes identified from Yeast-2-hybrid screening.
Supplemental Table S2. PCR primers used in this study.
Supplemental Methods. Plasmid construction, wax analysis by GC-FID, and monosaccharide analysis by HPAEC.
ACKNOWLEDGMENTS
We thank the Genomic Analysis Laboratory of the Salk Institute for sequence-indexed Arabidopsis T-DNA insertion mutants, Dr. Erin Gilchrist for providing MUM2-YFP, Dr. Shawn D. Mansfield with access to and assistance with HPAEC, Dr. Yuelin Zhang for the MPK6-cLuc strain, and the University of British Columbia BioImaging Facility for the use of microscopes and assistance with confocal imaging. We are grateful to Dr. Abel Rosado for critical reading of the article and insightful comments.
Footnotes
The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (www.plantphysiol.org) is: Ljerka Kunst (kunst{at}mail.ubc.ca).
L.S., L.K., and G.W.H. designed research; L.S., G.H.D., H.Z., M.J.M., and T.M.H. performed research and analyzed data; L.S., L.K., and G.W.H. wrote the article.
↵1 This work was supported by the Natural Sciences and Engineering Research Council of Canada (Strategic Project Grant to L.K. and G.W.H.), and the Pei-Huang Tung and Tan-Wen Tung Graduate Fellowship (to L.S.) and University of British Columbia Faculty of Science Graduate Award (to L.S.).
↵2 Present address: Department of Molecular Biology, Massachusetts General Hospital, and Department of Genetics, Harvard Medical School, Boston, Massachusetts, 02114.
↵3 Present address: Department of Biology, McGill University, 1205 Dr Penfield Avenue, Montreal, Quebec, H3A 1B1, Canada.
↵5 Senior author.
↵[OPEN] Articles can be viewed without a subscription.
- Received June 25, 2019.
- Accepted August 22, 2019.
- Published September 4, 2019.
REFERENCES
