Different subcellular localizations and functions of Arabidopsis myosin VIII
© Golomb et al. 2008
Received: 16 September 2007
Accepted: 08 January 2008
Published: 08 January 2008
Skip to main content
© Golomb et al. 2008
Received: 16 September 2007
Accepted: 08 January 2008
Published: 08 January 2008
Myosins are actin-activated ATPases that use energy to generate force and move along actin filaments, dragging with their tails different cargos. Plant myosins belong to the group of unconventional myosins and Arabidopsis myosin VIII gene family contains four members: ATM1, ATM2, myosin VIIIA and myosin VIIIB.
In transgenic plants expressing GFP fusions with ATM1 (IQ-tail truncation, lacking the head domain), fluorescence was differentially distributed: while in epidermis cells at the root cap GFP-ATM1 equally distributed all over the cell, in epidermal cells right above this region it accumulated in dots. Further up, in cells of the elongation zone, GFP-ATM1 was preferentially positioned at the sides of transversal cell walls. Interestingly, the punctate pattern was insensitive to brefeldin A (BFA) while in some cells closer to the root cap, ATM1 was found in BFA bodies. With the use of different markers and transient expression in Nicotiana benthamiana leaves, it was found that myosin VIII co-localized to the plasmodesmata and ER, colocalized with internalized FM4-64, and partially overlapped with the endosomal markers ARA6, and rarely with ARA7 and FYVE. Motility of ARA6 labeled organelles was inhibited whenever associated with truncated ATM1 but motility of FYVE labeled organelles was inhibited only when associated with large excess of ATM1. Furthermore, GFP-ATM1 and RFP-ATM2 (IQ-tail domain) co-localized to the same spots on the plasma membrane, indicating a specific composition at these sites for myosin binding.
Taken together, our data suggest that myosin VIII functions differently in different root cells and can be involved in different steps of endocytosis, BFA-sensitive and insensitive pathways, ER tethering and plasmodesmatal activity.
The Arabidopsis myosin gene family contains 17 members. The myosin XI group, which is related to unconventional myosin V [1, 2], includes 13 members while the myosin VIII group, which is related to unconventional myosin V  but also to myosin VI , includes four members. Both myosin groups VIII and XI are specific to plants . Typically, the Arabidopsis myosins contain a conserved motor domain, a number of IQ domains for light-chain binding, a coiled-coil domain that is predicted to facilitate their dimerization and a specific tail to bind the cargo . Plant myosins are generally implicated in cytoplasmic streaming , organelle movement [5–7], cytokinesis [8–10], plasmodesmatal functioning [10, 11], and endocytosis [10, 12, 13].
Myosin VIII (ATM1) was the first plant myosin to be identified and sequenced . A specific antibody raised against a peptide corresponding to its unique tail was used to show, in both immunofluorescence and electron microscopy studies, that ATM1 was localized to plant-specific structures such as the plasmodesmata and plasma membrane of newly formed cell walls in root cells of maize and Arabidopsis [11, 15]. ATM1 has also been found in pit fields in the inner cortex cells of maize root apices, where it was predicted to be involved in fluid-phase endocytosis . In a more recent work, the same antibody was used to show that in cells of maize root caps, ATM1 localized around the nuclei but relocated to amyloplasts and to plasmodesmata following 5 min and 90 min of osmotic stimulus, respectively . The latter suggests acto-myosin involvement in root osmo-sensing . When ATM1 was fused to GFP, its fluorescence was concentrated mostly at the developing cell plate in BY2 cells . The four members of myosin VIII are ATM1, ATM2, myosin VIIIA and myosin VIIIB. While ATM1 is more similar to VIIIA, ATM2 is related to VIIIB both in sequence and expression pattern.
In this work, we followed myosin VIII using GFP fusions. We show that in transgenic plants expressing GFP-ATM1(IQ-tail) fluorescence is differentially localized in different root cells. Sensitivity to BFA also differed between root cells. When transiently expressed in N. benthamiana, GFP-ATM1(IQ-tail) was found in pit fields accumulating callose, and co-localized with the ER. ATM1 also co-localized with internalized FM4-64 and partially co-localized with the endosomal markers ARA6, and rarely with ARA7 and FYVE. Furthermore, truncated ATM1 inhibited the motility of associated ARA6 labeled organelles but could be found on motile FYVE labeled organelles. Only large excess of ATM1 associated with FYVE labeled organelles arrested their motility. Taken together our data suggest that myosin VIII differentially functions in different cells and can be involved in different steps of endocytosis, BFA sensitive and insensitive pathways, ER tethering and in plasmodesmatal activity.
Transgenic Arabidopsis plants expressing GFP-ATM1(IQ-tail) were generated. Among tens of seedlings resistant to kanamycin, only one seedling expressed detectable levels of GFP-ATM1(IQ-tail), suggesting some cytotoxicity of the mutant molecule. GFP-ATM1(IQ-tail) was expressed and passed on to subsequent generations but no significant phenotypes or changes in cell morphology relative to wild-type (wt) plants were observed. Interestingly, fluorescence of GFP-ATM1 (IQ-tail) was visible in the roots but almost undetectable in the shoots and leaves.
Here, we studied transgenic plants expressing a chimera of GFP fused to a truncated myosin VIII – the IQ-tail domain of ATM1. This mutant molecule, lacking the head, motor domain that bind actin but containing the neck and tail domains, is expected to bind to its cellular targets and function as dominant negative by blocking wt myosin binding. Indeed, among many seedlings expressing the gene for antibiotic resistance, only one was found expressing the fluorescent chimera of myosin, at very low levels. This might be the result of a cytotoxic effect of the mutant. The fact that no detectable phenotypes were observed in the plants expressing the mutant myosin was a surprise, but might be explained by the low level of expression. In this regard, it should be mentioned that when other truncated myosins were expressed in plant cells, no significant inhibition of organellar movement was observed  suggesting redundancy in myosin's function. In the transgenic plants, we show differential localization and function of ATM1 in root cells. Similar differential localization was shown for myosin VIII in maize roots stained with anti myosin VIII antibody where it was found to be distributed diffusely in root cap cells but appeared as fine spots in the distal part of the apical meristem, in cells of the inner cortex, and in the distal part of the elongation region .
In view of its differential sensitivity to BFA in different root cells and its co-localization with different endosomal markers, we propose that ATM1 participates in various stages of endosomes biogenesis and function. We confirm previous data that myosin VIII resides at the plasma membrane and is enriched in plasmodesmata [11, 12, 15, 33]. In addition, we confirm a previous conclusion that myosin VIII is involved in endocytosis  and participates in tethering cortical elements of the ER to the plasma membrane .
Based on their findings, Dieter Volkmann and co-workers concluded that myosin VIII may be less important for intracellular motility and more involved in the anchoring of actin filaments at cell peripheries . They predicted another possible role for myosin VIII in forming a structural support for the cortical ER elements tightly underlying the plasma membrane, both outside and inside the plasmodesmata . They also suggested that myosin VIII might drive invagination of the plasma membrane during fluid-phase endocytosis .
While ATM1 is more or less equally expressed in all plant organs, ATM2 show specificity to male reproductive organs (Additional Figure 2). Other gene families of cytoskeletal proteins show differential expression in vegetative and reproductive tissues, such as actin , profilin [35, 36] and myosin XI (genevestigator expression profiles ). We provide evidence here that both ATM1 and ATM2 co-localize at specific sites on the plasma membrane when transiently expressed in N. benthamiana leaf. This is the first time that two plant myosins have been shown to co-localize at the same spots on the plasma membrane, suggesting a unique composition of the plasma membrane at these sites. The myosin VIII-labeled spots on the plasma membrane might be sites where the cortical actin fibers are linked to the plasma membrane such as "focal contacts" [10, 37, 38], sites of endocytosis , sterol-enriched complexes  or something else that we do not yet know about. Since our myosin constructs do not contain the head domain which is the actin-binding domain, we could not use them to address the question of actin binding to the plasma membrane.
In addition, we show co-localization of ATM1 (Figure 5G–I) with the ER in N. benthamiana leaves. When BDM (2,3-Butanedione 2-monoxime) which is a general inhibitor of myosin ATPases of eukaryotic cells, was applied to growing maize roots, alterations of the typical distribution patterns of myosin VIII, actin filaments and cortical ER elements associated with plasmodesmata and pit fields were observed . Nevertheless, as shown here, the expressed GFP-ATM1 which is a mutant molecule lacking the head, actin binding domain, did not disrupt the ER network. This suggests that myosin VIII is necessary but not essential for anchoring the cortical ER to the plasma membrane  and that other proteins with overlapping functions are there. Since ER spans the plasmodesmata  it further suggests close relationships between ATM1 and the ER. Plasmodesmatal myosin VIII was postulated to be involved in regulating conductivity by the force it can generate to control the spacing between desmotubules and plasma membrane . Our data suggest that the cross talk between ER and myosin VIII is not limited to plasmodesmata because colocalization was observed at the outer membrane of epidermal cells. Two models have been proposed for the role of myosin V in ER localization and movement in yeast and animal cells; in model-A Myosin V actively transport ER tubules, while in model-B myosin V plays a role in tethering ER to the cell surface . The prediction is that in plants, myosin XI plays the role of model-A while myosin VIII the role of model-B.
In root hairs (Figure 1D) GFP-ATM1 formed punctuate pattern along the cell. The clear zone of root hairs tip is rich in membrane recycling activity and organelles marked by the endosomal marker FYVE . Myosin XI was also found in the clear zone of pollen tubes . The presence of ATM1 in the clear zone of elongating root hairs should be carefully analyzed.
When ATM1 was co-expressed with different endosomal markers, partial co-localization was found with ARA6, and rarely with ARA7 and FYVE. Endosomes in mammalian cells are categorized into four classes; early endosomes, late endosomes, recycling endosomes and lysosomes . In plants there is no similar classification, and the functional differences between endosomes is not clear. The Rab5 small GTPases typically regulate early endosomes  and ARA6 and ARA7, which are plant Rab5 orthologs, localize to different and partially overlapping sub-populations of endosomes [28, 46]. While ARA6 co-localized with SNARE proteins characteristic of the pre-vacuolar compartment (PVC), ARA7 did not . This suggested that ARA7 labels an earlier endosomal compartment . Indeed it has been shown that in plants, the routes of endocytosis and vacuolar transport merge at the PVC . However, when ARA7 was co-expressed with other PVC markers – PS1-GFP  or AtPEP12p:HA , it was also detected in the PVC. In our working system, ATM1 was mainly detected at the plasma membrane in a punctate pattern; however, it was also observed in the cytoplasm, albeit rarely. The dots of cytoplasmic GFP-ATM1 were co-localized with internalized FM4-64 and rarely with the markers FYVE and ARA7. Co-localization of ATM1 with ARA6 was pronounced and seemed to be at the plasma membrane. Indeed it was previously indicated that GFP-ARA6 resides on the plasma membrane and also on the ER  whereas Ara7 and Rha1 are different . By showing preference of ATM1 to ARA6 labeled endosomes and by demonstrating that each ARA6 labeled organelle loose its motility when associated with truncated ATM1 while FYVE labeled organelles can remain mobile although associated with ATM1 and only large excess of truncated ATM1 could arrest their motility, we provide another evidence for the presence of different subpopulation of endosomes . Our data suggests that ATM1 is more crucial for the motility of ARA6 containing endosomes and plays only a minor role, if at all, in the function of FYVE/ARA7 containing endosomes. Importantly, the observation that ARA6-GFP partially co-labeled BFA-induced structures in roots  is in agreement with our findings of the association of ATM1 with ARA6-GFP on the one hand, and with BFA bodies in specific root cells on the other hand.
The partial localization of ATM1 to different vesicles marked by ARA6, ARA7 or FYVE and to BFA bodies suggests that it has a role in the motility of endosomes at different stages of maturation and in endosomal recycling to the plasma membrane [19, 20, 22–24]. A dual role in endocytosis, both at the plasma membrane and in endosomal motility, has been suggested for myosin VI [50, 51], which is phylogenetically close to myosin VIII . Myosin VI is implicated in both the formation of clathrin-coated vesicles at the plasma membrane and the movement of nascent uncoated vesicles from the actin-rich cell periphery to the early endosome in animal cells [50, 51]. Interestingly, myosin VI was found to move toward actin's minus end  in a processive manner . The latter functions are still not known for myosin VIII.
Plant roots are very sophisticated sensors that are able to perceive various different environmental and soil cues such as gravitation force, touch, water potential and osmolarity. Gravity-sensing is done by specialized cells that are located within the columella root cap. The signal is perceived by specialized cells, statocytes, containing specific amyloplasts, statoliths that sediment in these cells according to the gravitation force . The gravity perceived signal is transmitted from the columela cells by a mobile auxin signal to the cells at the elongation zone . The actin meshwork is believed to be the cellular structure that sedimenting amyloplasts pass through or interact with to trigger the downstream signaling events leading to root orientation . There is also accumulating evidence that the transport of auxin, is regulated by actin [24, 57–59]. Thus the differential localization of ATM1 in root cap and elongating cells might reflect the different roles that it plays as part of the actin cytoskeleton during signaling perception. We did not detected particular phenotypes related to gravitropism in the GFP-ATM1 (IQ-tail) plants. Although ATM1 seems to have polarized localization to transverse walls in cells at the elongation zone its role in the polarized accumulation of auxin transporters [19, 24] is still to be determined. It was also shown that touch stimulation but not gravity-stimulation led to transient increases in Ca2+ ions in root cells  and that the increase induced in the cap cells was larger and longer-lived than in cells in the meristematic or elongation zone . The touch induced calmodulin like protein 2 (TCH2)  was found by us to interact with the IQ domains of ATM1 in a calcium regulated manner . Thus ATM1 can be involved in differential touch responses in the different root cells, being regulated by TCH2 as a light chain. ATM1 was also shown to be responsive to osmo-signals specifically in root-cap cells where it was found to be recruited to plastids surfaces following stimulation . The dotted pattern seen in some of the root cells was identified by an anti myosin VIII antibody in maize roots[11, 15] as plasmodesmata enriched pit fields. Using a specific stain for callose we confirmed the presence of ATM1 in pit fields of N. benthamiana leaves, however, in the transgenic plants, ATM1 fluorescent dots are scattered all over the cell. Thus we don't know precisely, what is the different function that ATM1 plays in these particular cells. Also, not all cells in the "belt" of cells showing dotted pattern of GFP-ATM1 were dotted, the reason might be the absence of synchronization in their developmental stage.
While a truncated GFP fusion of ATM1 lacking the head domain can be expressed by plants that remain normal, it was found to be a useful probe for ATM1's behavior in root cells. Using these plants it is shown here, in live cells, that ATM1 changes its localization in root epidermal cells as they develop. This change in localization is accompanied by a change in sensitivity to BFA, indicating on a functional modification. Further this work provide evidence using microscopy of live cells, that Myosin VIII is preferentially involved in the function of ARA6 associated endosomes and is localized to ER and pit fields rich in plasmodesmata.
Arabidopsis thaliana ecotype Col-0 were seeded on MS (Murashige & Skoog) media, incubated 4 days at 4°C in the dark and then transferred to a growth room at 24°C under 16 h light/8 h dark. After 7–10 days, seedlings were transferred to pots with peat and grown in a temperature-controlled (23°C) greenhouse under continuous light.
Nicotiana benthamiana plants were grown in peat in a controlled growth room at 25°C with optimum light for 16 h daily.
In order to fuse the IQ-tail domain of ATM1 to GFP, we used a cDNA clone kindly provided by Dieter Volkmann (University of Bonn, Germany) and the following primers: Fwd-GGG GTA CCC GTA CTC TCC ACG GCA TT and Rev-CGG GAT CCG TGC TTG GGA ATG CTG CC. The resulting fragment was ligated downstream of GFP using KpnI and BamHI into the plasmid ART7 containing GFP with a linker of 10XAla at its C terminus. Similarly, ATM1 was ligated to ART7 containing RFP cherry  with a linker of 10XAla at its C terminus. The ATM2 IQ-tail domain was isolated from Arabidopsis RNA using RT-PCR and the following primers: Fwd-GGG GTA CCA GGA AAA AGG TTC TTC AAG GC and Rev-CGG GAT CCC TAG CCT CTT TTT CCC CA. Similar clones of myosin VIIIA were isolated using these primers: Fwd-GGGGTACCCAGA TTGGGGTTCTTGAAGAT, and Rev-CGGGATCCTTAATACCTAGTA CTCCTCAA. And for myosinVIIIB Fwd-GGGGTACCGTAATTAGCG TCCTTGAGGAA, and Rev-CGGGATCCTCAATAACTTTTCTTGCACCA. These were fused to either GFP or RFP as described for ATM1.
The entire expression cassettes containing the fluorescent chimeras under the regulation of the 35S promoter were then transferred to the binary vector pART27 using a NotI cleavage. The plasmid encoding DsRED-FYVE was provided by Josef Samaj from the University of Bonn . Plasmids encoding GFP fusions of ARA7 and ARA6 were provided by Takashi Ueda from RIKEN, Japan . The plasmid encoding the GFP fusion of the H/KDEL receptor, ERD2-GFP, was provided by Chris Hawes from Oxford Brookes University, UK .
DNA was prepared as follows: two to three fresh mature leaves were ground in liquid nitrogen to a fine powder and dissolved in a mixture containing: extraction buffer (350 mM sorbitol, 100 mM Tris PH 7.5, 5 mM EDTA pH 7.5), Nuclei lysis buffer (200 mM Tris PH 7.5, 50 mM EDTA, 2 M NaCl, 2% CTAB) and 5% N-lauroylsarcosine at 1:1:0.4 ratio. After 15 min incubation at 65°C, chloroform/isoamylalcohol extraction was performed. DNA was precipitated in isopropanol and re-suspended in sterile ddH2O. Total RNA was prepared as follows: 2.5 gr of plant tissue was ground in liquid nitrogen to a fine powder and incubated 10 min in 20 ml hot (65°C) CTAB buffer (2% PVP, 2% CTAB, 2 M NaCl, 25 mM EDTA PH8, 0.1 M Tris PH8). Chloroform/isoamylalcohol extraction was followed and RNA was precipitated in 2.5 M LiCl. The pellet was re-dissolved in SSTE buffer (1 mM EDTA, 10 mM Tris PH8 1 M NaCl, 0.5% SDS) and additional chloroform/isoamylalcohol extraction was performed. RNA was precipitated in ethanol and NH4Ac and the pellet was dissolved in sterile double distilled ddH2O. Poly-A RNA was isolated using Oligotex kit of Qiagen (Cat No. 7002) according to the manufacturer's instructions. RT-PCR was performed using the SuperScript reverse transcriptase of Invitrogen (Cat No. 18064-022).
The amount of 250 mg of seedlings of transgenic plants was grinded to a fine powder in liquid nitrogen. The powder was boild in 200 μl of Laemmli's protein sample buffer  for 10 min and then centrifuged 10 min at 14 krpm in room temp. A sample (30 μl) of the extract was separated on SDS-PAGE and blotted onto a nitrocellulose membrane. GFP fusion proteins were detected with anti GFP antibody (Santa Cruz), and a secondary HRP conjugated antibody (Jacksom ImmunoResearch). For chemiluminescent reaction, the SuperSignal kit (Pierce) was used.
An IX81/FV500 laser-scanning microscope (Olympus) was used to observe fluorescently labeled cells. The following filter sets were used: for observing GFP, 488 nm excitation and BA505-525; RFP, 543 nm excitation and BA610; FM4-64, 488 or 515 nm excitation and BA660. The objective used was PlanApo 60 × 1.00 WLSM 8/0.17. To observe aniline blue we used 405 nm excitation and BA430-460, using objective UPlanSApo 60 × 1.35 oil, 8/0.17 FN26.5. When GFP and RFP were detected in the same sample, we used DM (dichroic mirror) 488/543 and when aniline blue was added, DM 405/488/543/633 was used. In all cases, where more than one color was monitored, sequential acquisition was performed. FM4-64 (Molecular Probes) staining was performed at a final concentration of 8 μM for 5–20 min. Callose was stained by incubating leaf segments for 40 min in a mixture of 0.1% aniline blue in ddH2O and 1 M glycine, pH 9.5, at a volumetric ratio of 2:3, and pre-mixed for at least 1 d before use . BFA (Sigma) was used at a final concentration of 50 μM for 50 min. in the presence of 4 μM of FM4-64.
The fluorescent chimeras were expressed using Agrobacterium infiltration. Briefly: Agrobacterium tumefaciens strain GV3101 was transformed with the plasmid and grown at 28°C for 48 h. Bacterial culture originating from one colony was grown, precipitated and dissolved to an OD600 of 0.5 in the following buffer: 50 mM MES, pH 5.6, 0.5% glucose, 2 mM NaPO4 and 100 μM acetosyringone (cat. no. D13440-6, Sigma Aldrich). Leaves of 3-week-old N. benthamiana plants were infiltrated with the bacterial culture using a 1-ml syringe. Expression of the fluorescent chimera in the leaf cells was detectable after 24 h, peaking at 48 h.
Arabidopsis plants were transformed as previously described . A culture of Agrobacterium GV3101 was grown in 1 liter of LB for 24 h at 28°C. The bacteria were precipitated at 5000 rpm for 15 min and dissolved to an OD600 of 0.8 in the following buffer: 5% sucrose, 10 mM MgCl2, and 0.044 μM BA (benzylamino purine). The detergent Silwet L-77 (LHELE SEEDS Texas USA) was added to a final concentration of 0.03%. Plants at the closed-flower-bud stage were dipped in the Agrobacterium solution for 5 min. The plants were then covered with plastic bags and transferred to controlled growth conditions. After 48 h, the bags were removed and after 1 month, seeds were collected. Seeds were sterilized in 1% NaClO for 5 min and washed three times with sterile H2O. Seeds (~5000) were plated on 10-cm Petri dishes with MS, the relevant antibiotics and 500 μg/ml claforan (Aventis) to prevent Agrobacterium growth. Antibiotic-resistant seedlings were transferred to a new Petri dish and scanned for GFP or RFP expression under a fluorescent binocular.
This research was supported by a Research Grant from the Israeli Science Foundation (ISF) 752/05. The authors thank Dieter Volkmann, Frantisek Baluska, Josef Samaj, Takashi Ueda and Chris Hawes for plasmids and Amnon Schwartz for his sage advice.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.