ABSTRACT
Cytokinesis marks the culmination of cell division when the cytoplasm undergoes division to yield two daughter cells. This intricate process encompasses various biological phenomena, including the organization of the cytoskeleton and the dynamics of vesicles. The development of the cell plate aligns with alterations in the cytoskeleton structure and vesicles derived from the trans-Golgi, ultimately resulting in the formation of the planar cell plate. Nevertheless, the coordination of these processes in plants remains to be determined. Here, we introduce HYCCIN-CONTAINING2 (HYC2) as a pivotal cytoskeleton cross-linking protein that accumulates at phragmoplasts and plays a crucial role in cell plate formation. A genetic study involving the depletion of HYC2 post-anaphase in Arabidopsis revealed HYC2’s function in cell plate formation. HYC2 interacted with dynamin-related protein 1A (DRP1A) and SH3 domain-containing protein 2 (SH3P2), essential for membrane tubulation during cell plate formation. The recruitment of SH3P2 and DRP1A to the cell plate and phragmoplast organization was compromised in homozygous hyc2-2 mutant globular embryos. Our results shed light on the cytoskeletal function of HYC2 in assembling the cell plate, potentially by guiding vesicles containing SH3P2–DRP1A to the planar cell plate.
Significance statement During plant cell division, a new membrane is constructed at the division plane, and vesicles fuse to form a transient compartment known as the cell plate. This structure undergoes stabilization via membrane remodeling events, including tubulation, to generate robust membrane intermediates. Apart from DRP1A and SH3P2, our understanding of other protein components involved in membrane remodeling still needs to be completed. HYC2 is a plant counterpart to the human Hyccin protein involved in neuronal myelination and displaying bundling capabilities for microtubules and microfilaments. Our study demonstrates that HYC2 is crucial in recruiting proteins during membrane tubulation in the cell plate formation in developing Arabidopsis embryos. These findings underscore the molecular function and vital role of the Hyccin protein in cell plate formation.
One Sentence Summary Plant HYC2 displays bundling capabilities for microtubules and microfilaments to recruit DRP1A and SH3P2 during membrane tubulation in the cell plate formation in developing Arabidopsis embryos.
Introduction
Within the eukaryotic cell division cycle, cytokinesis is the conclusive step in physically segregating the emerging daughter cells. In metazoans and fungi, this process initiates from the periphery. It progresses toward the center via the formation of a contractile actomyosin ring, leading to the constriction of the cell along the division plane. However, flowering plants, lacking myosin II genes and the formation of a contractile ring, have evolved an alternative cytokinesis strategy. The strategy involves the homotypic fusion of secretory vesicles to the partitioning membrane along dynamic microtubule tracks in the phragmoplast to construct the cell plate (Rasmussen et al., 2011; Jurgens et al., 2015). Vesicles derived from the trans-Golgi network undergo fusion, resulting in hourglass-shaped vesicle intermediates. These intermediates are then integrated into the cell equator, forming a tubulo-vesicular network, a tubular network, and a planar fenestrated sheet (PFS). Concurrently, cell wall polymers such as callose gradually accumulate within the tubulo-vesicular network, thus helping to stabilize its structure up to the PFS. Ultimately, the expanding PFS fuses with the mother cell wall (Smertenko et al., 2018).
Following anaphase, the phragmoplast emerges as a scaffold crucial for orchestrating the cell plate assembly and forming a new cell wall during cytokinesis. Comprising a complex arrangement of microtubules and microfilaments, the phragmoplast plays a pivotal role in facilitating the movement of secreted vesicles and positioning cell wall components for incorporation into the cell plates (Jurgens et al., 2015; Smertenko et al., 2018). In telophase, the phragmoplast manifests as two sets of antiparallel microtubules that converge and overlap at the midline. Materials containing secreted vesicles, essential for constructing the cell plate, traverse along these microtubules and merge to give rise to the cell plate. Consequently, proteins that modulate cytoskeletal dynamics, such as microtubule-associated proteins (MAPs) or actin-binding proteins (ABPs), play a fundamental role in the division and development of plant cells.
In embryo development, only a few cytoskeletal proteins have been identified, including the microtubule-severing protein KATANIN 1 (Luptovciak et al., 2017), microtubule-associated protein 65-3 (MAP65-3; Caillaud et al., 2008; Ho et al., 2011), HINKEL (Strompen et al., 2002), kinesin-related protein 125c (AtKRP125c; Bannigan et al., 2007), and RUNKEL (Krupnova et al., 2009). These cytoskeletal proteins regulate the organization of the phragmoplast microtubule array but not the trafficking of regulatory proteins to the cell plate. The coordination between cytoskeleton proteins and vesicle targeting during cell plate formation remains elusive, as indicated by the use of cytoskeletal inhibitors suggesting an interaction between the two cytoskeletal tracks whereby disrupting one type of filament affects the stability of the other (Maeda et al., 2020).
Hypomyelination and congenital cataract (HCC) was initially identified as a human white-matter disorder characterized by the insufficient myelination of both the central and peripheral nervous systems, thus resulting in progressive neurological decline and the presence of congenital cataracts (Zara et al., 2006). HCC is attributed to a deficiency in the Hyccin protein, a novel neuronal protein primarily expressed in the central nervous system and localized explicitly in neurons rather than myelinating cells (Zara et al., 2006; Ugur and Tolun, 2008; Gazzerro et al., 2012; Traverso et al., 2013). The plasma membrane-anchored phosphatidylinositol 4-kinase scaffolding complex comprises Hyccin, PI4KIIIα, and its adaptors tetratricopeptide repeat domain 7 (TTC7) and pho eighty-five requiring 3 (EFR3) (Baskin et al., 2016). Dysfunction of the Hyccin protein in fibroblasts from HCC patients and the oligodendroglial lineage of knockout mice feature reduced levels of PI4KIIIα, TTC7A/B, and EFR3A, which indicates the role of Hyccin in regulating PtdIns(4)P production. In the Arabidopsis genome, two HYCCIN-CONTAINING genes, HYC1 and HYC2, encode proteins as components of the PI4Kα1 complex. HYC1 is essential for the male gametophyte, whereas HYC2 is crucial for embryogenesis (Noack et al., 2022). However, the underlying molecular mechanism of HYC2 in embryogenesis remains to be elucidated.
In a search to identify genes crucial for embryogenesis in Arabidopsis, we obtained the hyc2 mutant, which features embryonic arrest at the late globular stage, forming aborted seeds. Biochemical analysis revealed the molecular role of HYC2 as a protein that bundles microfilaments and microtubules. HYC2 protein accumulates in the phragmoplast and is involved in cell plate formation, suggesting its involvement with microtubules and microfilaments. Additionally, we observed an interaction between HYC2 and the SH3P2– DRP1A complex, recognized for its role in membrane tubulation in cell plate formation. In hyc2-2/- globular embryos, signals from SH3P2 and DRP1A were absent at the developing cell plate. Hence, the essential function of the HYC2 protein in the cytoskeleton is crucial for recruiting the SH3P2–DRP1A complex to the cell plate in Arabidopsis.
Results
Disruption of HYC2 results in the arrest of globular embryos and the formation of cell wall stubs in Arabidopsis embryos
As the results found by Noack et al., 2022, we observed seed abortion phenotype in the heterozygous mutants of hyc2-2 and hyc2-3 (Figure S1A), and a quarter of the embryo was arrested at the globular stage when most had reached the heart stage in individual hyc2/+ siliques (Figure S1B). Genotyping of grown seedlings from heterozygous hyc2/+ progenies revealed a 1:2 ratio of wild type to heterozygotes, indicating defective embryogenesis (Table S1). The primary amino acid sequence of HYC2 shares 35.8% identity with the Hyccin domain (pfam0970), which is present in eukaryotic Bilateria and plants but absent in unicellular yeast. Protein alignment demonstrated high conservation of the Hyccin domain, with divergence in the C-terminal region (Figure S2). HYC2 was ubiquitously expressed in all tissues, with the highest expression in siliques (Figure S3A). The seed abortion phenotype of hyc2-2 was successfully rescued by its transgene HYC2p::HYC2-2XGFP (Figure S1A). Genetic and complementation analyses suggested that the arrested globular seeds in hyc2-2/+ siliques are hyc2-2 homozygotes (hyc2-2). The seed abortion phenotype of hyc2-2 could be rescued by full-length HYC2, HYC21-417, and HYC21-405, but not HYC21-398 (Figure S3B). JPred secondary structure prediction (Drozdetskiy et al., 2015) indicated that the penultimate helix of HYC2 (375-395 aa) is crucial for complementation (Figure S2A). These findings underscore the essential role of HYC2 in proper embryo development in Arabidopsis.
Because of arrested globular embryos in homozygous hyc2-2/-, we used SCRI Renaissance 2200 (SCRI), a dye for cell wall staining, to delineate the developing embryo outline (Smith and Long, 2010). SCRI staining revealed that cell organization was well coordinated in the wild type, and cell wall development proceeded normally (Figure 1A). However, in the hyc2-2/- embryo, cell wall formation exhibited irregularities and occasional discontinuities. Some embryonic cells featured a noticeable cell wall discontinuation phenotype (indicated by arrowheads in Figure 1A), reminiscent of observations in cytokinesis-deficient mutants (Sollner et al., 2002). Moreover, the cell number in the hyc2-2/- embryo exceeded that of the wild type at the globular stage and approached that of the wild type at the transition stage (Figure 1A, Table S2). We also investigated potential cytokinesis defects in hyc2-2/- embryos using ultrastructural observation (transmission electron microscopy [TEM] images in Figure 1B). In the wild-type embryo, the cell wall appeared smooth and compact, whereas, in the hyc2-2/-embryo, the cell wall featured looseness, protrusions, and occasional failure to fuse with the parental cell wall. The SCRI and TEM results, portraying discontinuous cell plates with undulating membrane sacs linked to incomplete cell plate formation, underscore the crucial role of HYC2 during cell plate formation in Arabidopsis.
HYC2 is a component of the PI4Kα1 complex and is critical for the completion of cell plate formation
HYC2 interacts with No Pollen Germination (NPG) proteins and serves as a component of the PI4Kα1 complex (Noack et al., 2022). To elucidate HYC2’s role within the PI4Kα1 complex, we examined PI4P signals using the fluorescent signal of theYFP-PHFAPP1 (Vermeer et al., 2009). In planta, PI4P signals primarily accumulated at the plasma membrane in wild-type embryos from the globular to the linear stage (Figure S4A). In contrast, in hyc2-2 globular embryos, the PI4P signal decreased at the plasma membrane and began accumulating in the cytosol (Figure S4A), which suggests HYC2’s involvement in recruiting PI4P to the plasma membrane. In dividing tobacco BY-2 cells, YFP-PHFAPP1 labeled the growing cell plates after detecting FM4-64 signals (Vermeer et al., 2009). Similarly, in Arabidopsis root cells, PI4P signals were gathered at the expanding cell plate, already marked by HYC2 signals (Figure S4B). HYC2 signals appeared in the cell plate before the PI4P signals, implying HYC2’s role in cell plate formation.
To gain insights into the cellular role of HYC2 in cell plate formation, we examined its localization in dividing cells. In Arabidopsis embryogenesis and root cells, HYC2 was found in the cytosol and plasma membrane (Figure 2). The fluorescent signals of HYC2-2XGFP, driven by the native promoter, exhibited robust HYC2 expression after the 8-cell embryo stage (Figure 2A). Additionally, time-lapse fluorescence microscopy revealed HYC2 accumulation at the cell plate during the division of root cells (Figure 2B). HYC2-GFP appeared in the midsection of the cell, where cell plate formation initiates, before labeling the emerging cell plate by FM4-64 signals (Dhonukshe et al., 2006). It gradually surrounded the FM4-64 signals until the cell plate formation was completed. The organization of the cytoskeleton, including microtubules and microfilaments, plays a regulatory role in cell division. To explore the connection between HYC2 and the cytoskeleton in cell plate formation, we used immunofluorescence analysis at different stages of mitosis. HYC2 was co-localized with the microtubule array after metaphase (Figure S5A) and with F-actin signals after telophase (Figure S5B). Hence, HYC2’s function is crucial in cell plate formation.
To further investigate the role of HYC2 in cell plate formation, we rescued hyc2-2/+ embryos by using GFP-tagged HYC2 fused with the N-terminal region of CYCB1;2, containing the destruction box (D box) responsible for protein degradation in anaphase (Criqui et al., 2000). HYC2-CYCB1;2 was progressively degraded concomitant with the emergence of the cell plate marked by FM4-64 (Figure 3A). As compared with our previous findings indicating the rescue of the seed-aborted phenotype of hyc2-2/+ by HYC2pro::HYC2-2XGFP (Figure 1A), 26.8% of seeds were arrested at the globular stage in HYC2pro::HYC2-CYCB1;2-2XGFP/hyc2-2/+ and in hyc2-2/+ (Figure 3B). This result underscores the need for the presence and function of HYC2 after the onset of anaphase for successful cell plate formation and embryogenesis in Arabidopsis.
HYC2 exhibits binding and bundling properties for both microtubules and microfilaments
To elucidate the biological function of HYC2, we tried to identify the potential HYC2-interacting proteins by co-IP and liquid chromatography-tandem mass spectrometry (LC-MSMS) analyses. Numerous tubulin and actin proteins were co-immunoprecipitated with HYC2 (Table S3), and their interactions were confirmed by immunoblotting assays (Figure 4A). To determine the roles of HYC2 in tubulin and microtubule binding ability, microscale thermophoresis (MST) and co-sedimentation analysis (Figure 4B) were further used. The results illustrate the direct interaction of recombinant HYC2 with tubulin, thereby suggesting its role as a tubulin-binding protein. Low-speed co-sedimentation assay between microtubule and HYC2 revealed the microtubule bundling activity of HYC2 (Figure 4D upper panel). In vitro bundling experiments clearly showed clustered microtubule accumulation in the presence of HYC2, and these HYC2-bundled microtubules were resistant to cold treatment-induced disassembly (Figure 4D lower panel).
The MST and co-sedimentation assays were also applied to demonstrate the direct interaction of recombinant HYC2 with actin and F-actin (Figure 4C). The F-actin co-sedimentation assay showed the F-actin bundling activity of HYC2 (Figure 4E upper panel). After co-incubation of F-actin and HYC2 in vitro, bundled F-actin filaments were observed in the presence of GST- and 6XHis-tagged recombinant HYC2 but not denatured HYC2 (Figure 4E lower panel). The above results suggest that HYC2 is an F-actin binding and can bundle F-actin into higher-order structures in vitro.
Given that HYC2 can interact with microfilaments and microtubules, we speculated that it might interact with both simultaneously. Indeed, incubating microfilaments and microtubules with HYC2 resulted in bundle formation (Figure 4F upper panel). In vitro immunofluorescence analyses confirmed the co-localization of HYC2 with cross-linked microfilaments and microtubules (Figure 4F) in contrast to no primary antibody control in the lower panel of Figure 4F. All these results collectively indicate that HYC2 functions as a cytoskeletal cross-linking protein, simultaneously binding and bundling microfilaments and microtubules.
To assess the impact of HYC2’s F-actin–bundling activity on microfilament organization in planta, we compared F-actin organization in wild-type and hyc2-2/- embryos by staining F-actin with Alexa Fluor 488 phalloidin. Wild-type globular embryos exhibited strong and elongated fluorescent signals, whereas F-actin structures in hyc2-2/- embryos appeared shorter with cytoplasmic decorated punctate structures (Figure S6A). To quantify the extent of F-actin bundling in globular embryos, we measured the skewness, kurtosis, and occupancy as previously described (Figure S5B; Higaki et al., 2010). Quantitative analysis revealed reduced alignment of actin bundles in hyc2-2 mutants, as indicated by decreased skewness and kurtosis. Additionally, the density of F-actin was higher in hyc2-2/- than in wild-type embryos, as evidenced by increased F-actin occupancy (Figure S6B). The distortion of F-actin organization in hyc2-2/- embryos underscores the role of HYC2 in F-actin bundling.
The absence of HYC2 resulted in faulty phragmoplasts in globular embryos of Arabidopsis
In plant cell division, microtubules play a more crucial role than microfilaments, as indicated by drug treatment and molecular studies (Higaki et al., 2008; Li et al., 2015). We examined the organization of microtubules in hyc2-2/- embryos, particularly during cell plate formation. Compared to wild-type cells, hyc2-2/- cells displayed abnormal cell division without overt irregularities in cortical microtubules (Figure 5A-B). To assess the impact of HYC2 in cell division, we observed the plant-specific cytoskeletal structures known as phragmoplasts by using mGFP:AtTUA6 (Liao and Weijers, 2018) driven by an embryo-specific WOX2 promoter in arrested globular embryos of hyc2-2/- compared to wild-type control embryos (Figure 5C-F).
At the later stages of cell plate formation, as the tubular structure of the cell plate develops, microtubule bundles typically extend toward the growing margin of the cell plate until cytokinesis is completed (Figure 5C, E). We identified dividing cells through SCRI-stained floated cell plates (Figure 5C, D) and cell wall stubs in hyc2-2/- cells (Figure 5F). In wild-type dividing cells, the microtubule-based phragmoplast accumulated at the leading edge of the cell plate (Figure 5C). However, in approximately 67% of dividing cells in hyc2-2/-, the microtubule signals of the phragmoplast were absent (Figure 5G). In all hyc2-2/- embryonic cells with cell wall stubs, no phragmoplast signals labeled by mGFP-TUA6 were detected (Figure 5F). Our findings suggest that HYC2 is essential for microtubule organization during cell plate formation. HYC2, with its cytoskeleton-bundling function, regulates microtubule organization in cell plate formation.
HYC2 interacts with SH3P2 and DRP1A and is crucial for recruiting SH3P2 and DRP1A to the cell plate
To understand the role of HYC2 in cell plate formation, we investigated its protein– protein interactions with other known participants in cell plate formation by using yeast two-hybrid (Y2H) methods (Figure 6A, Figure S7). HYC2 exhibited a weak interaction with SH3P2, which is involved in membrane tubulation (Figure 6A). The HYC2–SH3P2 interaction was further confirmed by co-IP (Figure 6B). Because SH3P2, forming a complex with DRP1A, plays a vital role in establishing a tubular-vesicular network (Ahn et al., 2017), we also explored the interaction between HYC2 and DRP1A. DRP1A did not interact with HYC2 in the Y2H system, whereas co-IP assays (Figure 6B) and co-IP/LCMSMS analysis (Table S3) revealed an interaction between HYC2 and DRP1A. Thus, HYC2 likely forms a complex with SH3P2 and DRP1A during cell plate formation.
To investigate the impact of HYC2 on SH3P2 and DRP1A localization, we generated DRP1A-GFP and SH3P2-GFP transgenic plants and observed their signals in the dividing cells of hyc2-2/- globular embryos. SH3P2 signals driven by its native promoter accumulated in the cell plate in wild-type embryos but were reduced in hyc2-2/- globular embryos (Figure 6C-D). Similarly, DRP1A-GFP signals in wild-type dividing cells were substantially accumulated in the cell plate (Figure 6E), whereas approximately 89% of dividing cells in hyc2-2/- globular embryos showed reduced DRP1A signals (Figure 6F). In hyc2-2/- cells, SH3P2 and DRP1A signals were dispersed in the cytosol and occasionally accumulated in dot-like structures (Figure 6C, E). This finding suggests that the cytoskeleton-associated HYC2 may contribute to the localization of SH3P2 and DRP1A to the cell plate, and the leaky cell plate phenotype in hyc2-2/- may result from reduced signals of SH3P2 and DRP1A, thus preventing the formation of a stable tubulo-vesicular network structure.
Conversely, we explored whether HYC2 localization depended on DRP1A and SH3P2 during cell plate formation. SH3P2 RNAi and drp1a-2 T-DNA mutants have defects in cell plate formation, including irregular cell organization and cell wall stubs found in hyc2-2/- globular embryos (Mravec et al., 2011; Ahn et al., 2017). Loss of SH3P2 and DRP1A in SH3P2 RNAi plants and drp1a-2 T-DNA mutants did not affect HYC2 localization (Figure 6G-J), which suggests that HYC2 localization may be independent of DRP1A and SH3P2. HYC2’s role in regulating phragmoplast formation and its impact on SH3P2 and DRP1A localization to the cell plate underscores its vital role in cell plate formation. The phytohormone auxin plays a critical role in establishing embryo patterns, regulating cell division, and determining cell identity in plants (Blilou et al., 2005; Moller and Weijers, 2009; Schlereth et al., 2010; Lau et al., 2012). PIN-FORMED (PIN) proteins, particularly PIN1, contribute to establishing auxin gradients in globular embryos. PIN1 and DRP1A form transient complexes during cell plate formation, and the polarization of PIN1 depends on DRP1A (Mravec et al., 2011). Observing that hyc2-2/- embryos failed to progress to the heart stage, we examined auxin gradients and PIN1 localization in globular embryos by crossing a DR5 reporter and PIN1-GFP with hyc2-2/+ mutants. In the wild type, the DR5 reporter was expressed in the apical cell and its descendants until the 32-cell stage, after which DR5 activity shifted to the uppermost suspensor cells (Friml et al., 2003). Although DR5 activity was present in the uppermost suspensor cells in hyc2-2 globular embryos, auxin signals abnormally accumulated in the upper tier and part of the lower tier (Figure 7), thus indicating altered auxin distribution in the hyc2-2/- mutant. In the wild-type globular embryo, PIN1 signals accumulated at the plasma membrane of the upper tier and provascular cells. PIN1-GFP still localized to the plasma membrane in hyc2-2/- mutants but exhibited a slight increase in cytosolic accumulation (Figure 7). These findings suggest that the developmental defects in hyc2-2/- embryos are linked to abnormal auxin distribution and PIN1 polarization. Because PIN1 recruitment to the plasma membrane remains intact, HYC2 is critical for recruiting SH3P2 and DRP1A to the cell plate instead of PIN1 trafficking along the secretory pathway.
Discussion
We elucidated the molecular role of HYC2 as a cytoskeleton-bundling protein crucial for cell plate formation during early Arabidopsis embryogenesis. The HYC2 mutation resulted in cytokinesis defects, forming arrested globular embryos and aborted seeds. After metaphase, HYC2 exhibited accumulation around the FM4-64-labeled cell plate and phragmoplasts. Depletion of HYC2 after anaphase caused embryo arrest, which highlights the indispensable role of HYC2 in cell plate formation. Moreover, HYC2 played a regulatory role in phragmoplast formation and influenced the recruitment of SH3P2 and DRP1A to the cell plate. These findings underscore the essential role of HYC2, with its cytoskeleton-bundling activity, in phragmoplast formation and the recruitment of SH3P2 and DRP1A to the developing cell plate.
FAM126A, a human protein analogous to HYC2, interacts with the PI4K complex to regulate PI4P production in humans. However, the precise mechanism governing the association between FAM126A and the PI4K complex remains unclear (Yao et al., 1999; Stevenson-Paulik, Love and Boss, 2003). A recent study indicated that HYC2 is a constituent of the PI4Kα1 complex and localizes to Arabidopsis plasma membrane nanodomains (Noack et al., 2022). Notably, except for hyc2 mutants, mutants of various other components, including PI4Kα1, HYC1, NPG proteins, and EFOP (EFR3-OF-PLANTS) proteins, exhibit gametophyte-defective phenotypes (Noack et al., 2022). Loss of HYC2 leads to embryo lethality at the globular stage in Arabidopsis. These lethal phenotypes pose challenges in exploring the function of the PI4Kα1 complex in plants. Furthermore, the molecular mechanism and biological role of HYC2 remain unexplored, even in animals. Our study initiates the investigation of the Hyccin family, shedding light on its role in regulating cytoskeleton dynamics during cell plate formation and influencing the recruitment of proteins to the cell plate in Arabidopsis.
Cell plate formation relies on lipid determinants that delineate plasma membrane identity. These determinants encompass phosphatidylinositol 4-phosphate (PI4P), synthesized by PI 4-kinases, including PI4Kα1, PI4Kβ1, and PI4Kβ2 (Hammond et al., 2012; Nakatsu et al., 2012; Baskin et al., 2016). PI4Kα1 primarily functions at the plasma membrane, whereas PI4Kβ1 and PI4Kβ2 generate the PI4P pool at the trans-Golgi network (Preuss et al., 2006; Lin et al., 2019). PI4Kβ1 and PI4Kβ2 play roles in cell plate formation, as evidenced by the pi4kβ1pi4kβ2 mutant displaying over-stabilized phragmoplast microtubules, mis-localization of MAP65-3 and defects in the membrane trafficking of KNOLLE and PIN2 (Lin et al., 2019). In contrast, our study revealed that HYC2, a component of the PI4Kα1 complex, protects microtubules against disassembly, and hyc2-2/- embryonic cells lose microtubule-based phragmoplasts. Despite PI4Kβ1/PI4Kβ2 and HYC2 influencing membrane trafficking, they appear to regulate phragmoplasts by distinct mechanisms. In yeast, PI4Kα and PI4Kβ groups are non-redundant, with no substitution possible between them (Audhya et al., 2000). Our findings support that these two groups are not functionally redundant in plants.
The creation of the cell plate encompasses the incorporation of membrane structures into an existing plasma membrane, a process tightly governed by the secretory system’s regulation of materials delivered to the developing cell plate (Backues et al., 2007). This regulation is intricately linked to the organization of microtubules in phragmoplasts and microtubule-associated proteins, including the Kinesin-12 subfamily and MAP65-3, which are associated with the phragmoplast. Kinesin-12 proteins are recognized for their role in transporting cargo along microtubules (Muller and Livanos, 2019). KNOLLE, a syntaxin-like protein, is implicated in cytokinesis (Lauber et al., 1997), and in kin12akin12b double mutants, the failure to organize microtubule bundles into an antiparallel phragmoplast array resulted in the loss of KNOLLE’s localization to the cell plate (Lee, Li and Liu, 2007). Conversely, MAP65-3 influenced microtubule-plus ends in phragmoplasts but did not affect the membrane trafficking of KNOLLE and DRP1A (Ho et al., 2011). These observations suggest that cytoskeleton-associated proteins play crucial roles in phragmoplast organization but may not be directly involved in transporting cargo to the cell plate. Notably, our results demonstrate that cytoskeleton-associated HYC2 regulates both the formation of phragmoplasts and the delivery of cargo to the cell plate during cell plate formation.
To figure out if the recruitment is related to PI4P production, we treated Arabidopsis seedlings of DRP1A-GFP transgenic plants with phenylarsine oxide (PAO), PI4P inhibitor to detach PI4P from plasma membrane to cytosol. Our results showed that both DRP1A and SH3P2 signals were still in the cell plate after PAO treatment in contrast to the PI4P sensor, mRFP-PHFAPP1 signals were released from the plasma membrane to the cytosol (Figure 6K). These suggest that PI4P signals have less impact on targeting DRP1A and SH3P2 to the cell plate. However, we could not rule out the effect of PI4P on DRP1A and SH3P2 because both DRP1A-GFP and SH3P2-GFP fluorescent intensity were slightly reduced after PAO treatment, and previous studies showed that SH3P2 binds to PI4P (Zhuang et al., 2013; Mravec et al., 2011; Ahn et al., 2017). Accordingly, the cytoskeleton binding function of HYC2 plays a critical role in SH3P2 and DRP1A recruitments during cell plate formation.
Using biochemical interaction studies and in planta investigations, we observed impaired localization of SH3P2 and DRP1A in incomplete cell plates in hyc2-2/- embryonic cells. These findings indicate that HYC2, with its cytoskeleton-bundling activity, is essential for adequately localizing SH3P2 and DRP1A to the cell plate. Previous research has demonstrated the interaction between DRP1A and PIN1 in correcting the PIN1 distribution (Mravec et al., 2011). Notably, in hyc2-2/- mutants, we did not observe a depletion of PIN1 at the plasma membrane, although DRP1A localization was altered. HYC2 plays a role in both phragmoplast formation and the trafficking of cargo, including SH3P2 and DRP1A, to the cell plate. Given that HYC2 is a part of the PI4Kα1 complex, further investigation into the relation between the PI4Kα1 complex and the SH3P2–DRP1A complex promises to be intriguing.
Methods
Plant material and growth conditions
All Arabidopsis plants and materials used in this study were in the Col-0 ecotype background. T-DNA insertion mutants of hyc2-2 (SALK_003807) and hyc2-3 (SALK_040977) were obtained from the Arabidopsis Biological Resource Center (Ohio State University, USA). Seeds from Arabidopsis thaliana were surface-sterilized after stratification for 3 days at 4°C and germinated on half-strength Murashige and Skoog (MS) medium (Duchefa Biocheme, The Netherlands) containing 1% sucrose and 0.7% phytoagar (Duchefa Biocheme, The Netherlands) or grown in the soil inside a growth chamber with a 16-h light/8-h dark cycle at 22°C.
Confocal microscopy and imagining
Microscopic analysis of embryonic cells from hyc2-2 or transgenic plants carrying SH3P2-GFP, DRP1A-GFP, DR5::GFP, PIN1-GFP, and mGFP:AtTUA in hyc2-2 were observed under a Zeiss LSM880 confocal microscope. The outline of the embryonic cell wall was visualized by SCRI staining. Primary roots from transgenic plants with HYC2-2XGFP were stained with FM4-64 (Invitrogen) for root cells. GFP signals from HYC2, FM4-64/PI, and SCRI were recorded using GFP, rhodamine, and DAPI filters. Line graphs of the signal intensities were processed by using ImageJ.
Ultrastructural analysis
For transmission electron microscopy (TEM), developing embryos of Arabidopsis were dissected from siliques. The samples were fixed in 2.5% glutaraldehyde and 4% paraformaldehyde in 0.1 M sodium phosphate buffer, pH 7.0, at room temperature for 4 h. After three 20-min buffer rinses; the samples were post-fixed in 1% OsO4 in the buffer for 4 h at room temperature and rinsed for 20 min three times. The samples were then dehydrated in an acetone series, embedded in Spurr’s resin, and sectioned with a Lecia Reichert Ultracut S or Lecia EM UC6 ultramicrotome. The ultra-thin sections (70-90 nm) were stained with uranyl acetate and lead citrate. A Philips CM 100 transmission electron microscope at 80 KV was used for viewing, and images were captured using a Gatan Orius CCD camera.
Globular embryo observation
To examine the early embryogenesis of hyc2 and HYC2pro::HYC2-CYCB1;2-2XGFP/hyc2-2/+ transgenic plants, young siliques were fixed in ethanol: acetic acid (8:1) for 30 min. They were cleared in Hoyer’s solution (chloral hydrate: distilled water: glycerol=8:3:1). Developing seeds were dissected from siliques and photographed by using an Olympus BX51 camera and a Zeiss Axio Imager Z1 Microscope.
Co-immunoprecipitation, LC-MS/MS analysis, and immunoblot analysis
Cell lysates were obtained from 5 g of 7-day-old seedlings harboring HYC2p:HYC2-2XGFP and GFP in extraction buffer (100 mM Tris, pH 8.0, 150 mM NaCl, 5 mM EDTA, 5 mM EGTA, 10 mM DTT, 0.1% NP-40, and protease inhibitor cocktail [Roche]). Then, 50 μl GFP-Trap A (Chromo-Tek) was washed with extraction buffer and added to cell lysates for incubation at 4°C for 2 h with gentle rotation. After washing three times with extraction buffer, proteins were eluted for LC-MS/MS analysis to identify the potential interacting proteins. The acquired MS/MS data were analyzed using the SEQUEST search program (BioWorks Browser 3.3, Thermo Fisher Scientific) against the Arabidopsis database downloaded from the National Center for Biotechnology Information (NCBI). Immunoblots were detected by chemiluminescence (SuperSignal West Pico, Thermo) for immunoblotting.
Purification of HYC2 recombinant protein from E. coli
To construct the HYC2 recombinant protein, PCR products using primer sets (HYC2 ORF w/o SP BamHI F and HYC2 ORF SalI R; HYC2 ORF-F BamH1 and HYC2 ORF-R XhoI) were ligated into the vector pGEX 4T-1 and pET30a to generate a fusion protein carrying the GST and 6XHis-tag, respectively. The recombinant proteins were expressed in E. coli strain BL21 and incubated with resins (GE Healthcare) according to the manufacturer’s instructions. The resins were intensively washed with incubation buffer before eluting the recombinant proteins. The eluted protein was further buffer-exchanged using a PD-10 column (GE Healthcare).
Microscale thermophoresis (MST) measurements
MST analysis was performed using a Microscale Thermophoresis Monolith NT.115Pico instrument. Actin, Tubulin, or HYC2 was labeled with Alexa-647 using a Monolith NT™ Protein Labelling Kit RED-NHS. The protein interaction was determined after 10 min incubation at room temperature in Tris buffer (50 mM Tris, pH 8.0, 100 mM NaCl). Finally, 3 to 5 μl was loaded into standard capillaries, and thermophoresis was determined.
MT binding and bundling assay
According to the manufacturer’s manual, microtubules were obtained from tubulin (T240, Cytoskeleton). All proteins were centrifuged at 100,000 × g for 30 min at 4°C before use. The reaction was buffered with 80mM PIPES, pH6.9, 2mM MgCl2, 0.5mM EGTA, 10% glycerol and 1mM GTP. In the high-speed experiments, samples were centrifuged at 100,000 × g for 30 min to pellet the MTs. The presence of HYC2 in the supernatants and pellets was analyzed via SDS-PAGE and Coomassie Brilliant Blue R staining. In the low-speed experiments, the samples were centrifuged at 12,500 × g for 30 min to pellet high-order MTs. MT in the supernatants and pellets was revealed via SDS-PAGE and Coomassie Brilliant Blue R staining.
Actin binding and bundling assay
High- and low-speed co-sedimentation experiments were performed to assess the F-actin binding and bundling activities of HYC2. All proteins were centrifuged at 100,000 × g for 30 min at 4°C before use. The reaction medium was buffered with 5 mM MES (pH 6.2) and supplemented with 10μM CaCl2. In the high-speed experiments, samples were centrifuged at 100,000 × g for 30 min to pellet the actin filaments. The presence of HYC2 in the supernatants (F-actin unbound fraction) and pellets (F-actin bound fraction) was analyzed via SDS-PAGE and Coomassie Brilliant Blue R staining. In the low-speed experiments, the samples were centrifuged at 12,500 × g for 30 min to pellet high-order F-actin structures. The presence of F-actin in the supernatants and pellets was analyzed via SDS-PAGE and Coomassie Brilliant Blue R staining.
In vitro visualization of microtubules and microfilaments
The bundles of microtubules and actin filaments in the samples were also visualized directly by using a Zeiss LSM510 confocal microscope. Rhodamine-red–labeled microtubules were incubated with 1 μM HYC2-6XHis at 22°C for 30 min or on ice for 30 min. For F-actin bundles, preformed F-actin was incubated with 1 μM GST-HYC2, HYC2-6XHis, and denatured GST-HYC2 at 22°C for 30 min. The HYC2 recombinant protein was boiled for 10 min to denature the proteins. Actin filaments were stained with 350 nM Alexa 488-phalloidin (Invitrogen) and observed under the same excitation and emission conditions as GFP fluorescence under the confocal microscope.
In vitro immunofluorescence
In vitro immunofluorescence was performed to investigate the location of HYC2 with cytoskeleton bundles. Preformed microfilaments and rhodamine-red–labeled microtubules were incubated with HYC2-6XHis at room temperature for 5 min. Anti-6XHis antibody (Invitrogen 46-0693, 1:50) and anti-mouse IgG-TRITC antibody (Sigma T5393, 1:50) were added by incubation at room temperature for 15 min. Actin filaments were stained with 350 nM Alexa 633-phalloidin (Invitrogen). HYC2 incubated with a secondary antibody alone was a negative control. Aliquots were transferred to slides pretreated with poly-L-Lys and visualized by confocal microscopy.
Antibodies
The sources for the antibodies are as follows, with the working dilution in parentheses: GFP (1:1000, Santa Cruz, sc-9996), actin (1:500, Pierce, MA1-744), mouse IgG (1:1000, Jackson ImmunoResearch, 115-035-003), 6XHis (1:50, Invitrogen, 46-0693), FLAG (1:1000, Sigma, F7425) and mouse IgG-TRITC (1:50, Sigma, T5393)
Author information
Y.-W. H., H.-J.W., and G.-Y.J. conceptualized and designed the experiments. Y.-W. H., C.-L.G., and C.-T. J. conducted the experiments, and Y.-W. H. and G.-Y.J. wrote the article.
Supplementary information
SI Methods
Complementation of the hyc2-2 mutants
To complement the hyc2-2/- mutants, the genomic sequence corresponding to locus At5g64090, which includes a 2.5-kb putative promoter and the 1.3-kb coding region, was amplified using Phusion polymerase (Finnzymes, http://www.finnzymes.com) and the HYC2 gDNA KpnI F and HYC2 ORF BamHI R primers. Subsequently, the PCR product was inserted into a modified pPZP221 vector containing two C-terminal GFP sequences followed by the Nopaline synthase (NOS) terminator. Cloning primers are detailed in Table S4. The confirmed construct was introduced into hyc2-2 heterozygous plants via a floral dip method 1. For antibiotic selection, T1 generation seeds were cultivated on solid half-strength MS medium (Duchefa) supplemented with 1% sucrose, 0.7% phytoagar (Duchefa), and G418 antibiotic at 100 μg/μl. After genotyping, seeds from plants with a hyc2-2/- homozygous background carrying HYC2p:HYC2-2XGFP were collected for further investigation.
RNA isolation and RT-PCR analysis
Total RNA was extracted using the Qiagen RNeasy Mini Kit (QIAGEN) and treated with TURBO DNA-free DNase (Ambion) to eliminate genomic DNA contamination. Reverse transcription of RNA was carried out using the M-MLV transcriptase system (Invitrogen) and oligo-dT primers. Primer sequences for HYC2 expression (HYC2 RT-F and HYC1 RT-R) are provided in Table S4.
Visualization of F-actin in globular embryos
To visualize F-actin in globular embryos, young siliques were fixed in 4% paraformaldehyde in phosphate-buffered saline (PBS), pH 7.4, for 1 h. Subsequently, the buffer was replaced with PBS buffer, pH 7.4, containing 100 nM Alexa 488-phalloidin. After 1 h of incubation, SCRI Renaissance 2200 (SCRI) was added at a 10,000X dilution for 30 min. Embryos were dissected on slides and observed using a Zeiss LSM510 or LSM880 confocal microscope. The skewness, kurtosis, and occupancy of F-actin in the globular embryos were calculated as previously described2. Statistical analysis involved the examination of at least 20 globular embryos from both wild-type and hyc2-2/-, analyzed with ImageJ software (http://imagej.nih.gov/ij/).
Immunofluorescence
Immunofluorescence was performed on whole-mounted roots following the protocol by Sauer and Friml3. Arabidopsis seedlings were mounted on slides with anti-GFP (Abcam, 1:400), anti-tubulin (Sigma, 1:400), and anti-actin (Thermo, 1:100) antibodies. Secondary antibodies for GFP and tubulin were anti-rabbit IgG-FITC and anti-mouse IgG-TRITC, respectively. Before placing coverslips, 100 μM DAPI was used for nuclei staining. Cells were imaged using the Zeiss LSM510 confocal microscope.
Acknowledgments
We thank Dr. Bo Liu for the helpful discussion about our study. We thank Mrs. Mei-Jane Fang (DNA Analysis Core Laboratory, Academia Sinica) for assistance with the confocal microscopy analyses and Mr. Wann-Neng Jane (Plant Cell Biology Core Laboratory, Academia Sinica) for the ultrastructural analysis. We acknowledge the use of the Microscale Thermophoresis Monolith NT.115Pico in the Biophysics Core Facility, funded by the Academia Sinica Core Facility and Innovative Instrument Project (AS-CFII-111-201). The proDR5:GFP and proPIN1:PIN1-GFP transgenic lines were generous gifts from Dr. Klaus Palme (Benkova et al., 2003). The pWOX2: mGFP:AtTUA transgenic lines were kindly provided by Dr. Dolf Weijers (Liao and Weijers, 2018). The construct of pDRP1A: DRP1A and the seeds of drp1a-2 (SALK_06977) were kindly provided by Dr. Sebastian Y. Bednarek (Mravec et al., 2011). The seeds of SH3P2 RNAi transgenic plants were kindly provided by Dr. Liwen Jiang (Zhuang et al., 2013). This work was supported by research grants from Academia Sinica (AS-GCS-111-L06) and the National Science and Technology Council (MOST 108-2311-B-001-036-MY and MOST 110-2313-B-001 −007 -MY3) to G.-Y. Jauh.
Footnotes
Material distribution footnote: The author(s) responsible for the distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (https://academic.oup.com/plcell/pages/General-Instructions) is Guang-Yuh Jauh.