|Home | About | Journals | Submit | Contact Us | Français|
Organelle movement in plants is dependent on actin filaments with most of the organelles being transported along the actin cables by class XI myosins. Although chloroplast movement is also actin filament-dependent, a potential role of myosin motors in this process is poorly understood. Interestingly, chloroplasts can move in any direction and change the direction within short time periods, suggesting that chloroplasts use the newly formed actin filaments rather than preexisting actin cables. Furthermore, the data on myosin gene knockouts and knockdowns in Arabidopsis and tobacco do not support myosins' XI role in chloroplast movement. Our recent studies revealed that chloroplast movement and positioning are mediated by the short actin filaments localized at chloroplast periphery (cp-actin filaments) rather than cytoplasmic actin cables. The accumulation of cp-actin filaments depends on kinesin-like proteins, KAC1 and KAC2, as well as on a chloroplast outer membrane protein CHUP1. We propose that plants evolved a myosin XI-independent mechanism of the actin-based chloroplast movement that is distinct from the mechanism used by other organelles.
Organelle movement and positioning are pivotal aspects of the intracellular dynamics in most eukaryotes. Although plants are sessile organisms, their organelles are quickly repositioned in response to fluctuating environmental conditions and certain endogenous signals. By and large, plant organelle movements and positioning are dependent on actin filaments, although microtubules play certain accessory roles in organelle dynamics.1,2 Actin inhibitors effectively retard the movements of mitochondria,3–6 peroxisomes,5,7–11 Golgi stacks,12,13 endoplasmic reticulum (ER),14,15 and nuclei.16–18 These organelles are co-aligned and associated with actin filaments.5,7,8,10–12,15,18 Recent progress in this field started to reveal the molecular motility system responsible for the organelle transport in plants.19
Chloroplast movement is among the most fascinating models of organelle movement in plants because it is precisely controlled by ambient light conditions.20,21 Weak light induces chloroplast accumulation response so that chloroplasts can capture photosynthetic light efficiently (Fig. 1A). Strong light induces chloroplast avoidance response to escape from photodamage (Fig. 1B).22 The blue light-induced chloroplast movement is mediated by the blue light receptor phototropin (phot). In some cryptogam plants, the red light-induced chloroplast movement is regulated by a chimeric phytochrome/phototropin photoreceptor neochrome.23–25 In a model plant Arabidopsis, phot1 and phot2 function redundantly to regulate the accumulation response,26 whereas phot2 alone is essential for the avoidance response.27,28 Several additional factors regulating chloroplast movement were identified by analyses of Arabidopsis mutants deficient in chloroplast photorelocation.29–32 In particular, identification of CHUP1 (chloroplast unusual positioning 1) revealed the connection between chloroplasts and actin filaments at the molecular level.29 CHUP1 is a chloroplast outer membrane protein capable of interacting with F-actin, G-actin and profilin in vitro.29,33,34 The chup1 mutant plants are defective in both the chloroplast movement and chloroplast anchorage to the plasma membrane,22,29,33 suggesting that CHUP1 plays an important role in linking chloroplasts to the plasma membrane through the actin filaments. However, how chloroplasts move using the actin filaments and whether chloroplast movement utilizes the actin-based motility system similar to other organelle movements remained to be determined.
Here, we review the recent findings pointing to existence of a novel actin-based mechanisms for chloroplast movement and discuss the differences between the mechanism responsible for movement of chloroplasts and other organelles.
The myosin motor proteins move various cargos along the actin filaments. Plants have evolved large multigene families encoding myosins of which class XI myosins are the most numerous and best studied.35–37 It was found that the transport of most plant organelles is abolished by potent myosin inhibitors such as N-ethylmaleimide (NEM) and 2,3-butanedione monoxime (BDM),4,6,8,10,11,13,15,16 although these data should be treated with caution due to potential off-target effects of these inhibitors. Using primarily ectopic expression of the myosin domain fusions to fluorescent reporters, it was suggested that some myosins are associated with organelles.15,38–43 Furthermore, recent analyses using myosin gene knockouts and the dominant-negative inhibition in Arabidopsis and tobacco revealed important roles for myosins XI in transport of Golgi stacks, peroxisomes, mitochondria and ER.14,15,37,42–46 It was also demonstrated that for at least Golgi stacks and peroxisomes, processive transport relies entirely on the four highly expressed class XI myosin motors.47 Among these, XI-K plays the predominant roles in the organelle transport, whereas myosins XI-1 (MYA1) and XI-2 (MYA2) and XI-I provide more limited contributions.14,15,37,44,45,47 Interestingly, two studies suggested that the myosin XI-dependent ER flow causes drag of cytoplasm historically defined as cytoplasmic streaming.14,15 Elimination of the three or all four of these myosins in Arabidopsis results in virtual arrest of the organelle trafficking and cytoplasmic streaming.15,47
Because of the association of chloroplasts with the long actin cables,48–50 myosin localization on plastids,38,42,51–53 and the inhibition of chloroplast movement by myosin inhibitors,51,54–56 it has been thought that chloroplast movement is also myosin-dependent. However, detailed microscopic analyses of chloroplast photorelocation movement induced by microbeam irradiation did not support this assumption.57,58 First, chloroplasts can move in any direction without rotation within a short lag time,57,58 a fact that argued against utilization of preexisting actin filaments. Second, light-induced reorganization of cytoplasmic actin cables (e.g., actin cables elongating in the direction of movement) has never been observed. Third, the velocity of moving chloroplasts is much lower than those of other organelles. Indeed, the myosin XI-dependent transport of the mitochondria, peroxisomes, Golgi stacks and ER occurs with the mean velocity on the order of several µm/sec3-15,37,44 matching a velocity of recombinant myosin XI-1 head translocation along F-actin in vitro.59 In contrast, the velocities of chloroplast movement are approximately several µm/min.57,58,60,61 Furthermore, none of the studied Arabidopsis myosin knockouts or tobacco myosins knockdowns exhibited any detectable defects in chloroplast photorelocation movement,37,44 although the abnormal chloroplast distribution in a myosin gene-silenced tobacco mesophyll cells was reported recently.42 Our preliminary analysis of the chloroplast photorelocation movement in Arabidopsis myosin XI double, triple and quadruple knockout lines known to be defective in transport of other organelles and F-actin organization15,45,47 showed normal accumulation and avoidance responses (N. Suetsugu, V. V. Dolja and M. Wada, unpublished data). Although the possibility that the myosins XI and VIII that remained present in these lines can collectively mediate chloroplast movement cannot be ruled out, these results indicate that the functional requirements of chloroplast movement are distinct from those of the other organelles.
Although the actin filament arrangements likely involved in chloroplast anchorage to the plasma membrane were detected in several plant species,62–68 no light-induced changes of actin filament dynamics correlated with the directions of chloroplast movement have been previously reported.29,31,32,50
Very recently, we have identified such a long-sought-after, novel arrangement of F-actin associated with chloroplast photorelocation movement. These short actin filaments were detected by detailed analysis of the transgenic Arabidopsis expressing the GFP-talin fusion protein.69 Notably, these filaments located at the chloroplast periphery, but not the cytoplasmic F-actin, showed dynamic, light-induced changes during chloroplast photorelocation. Similar behavior of the chloroplast-associated short actin filaments was observed using either tdTomato-fimbrin69 or life-ACT-Venus fusion reporters for F-actin visualization (SG Kong, N. Suetsugu and M. Wada, unpublished data). Because these latter reporters are known to have little if any effect on F-actin structure and dynamics, our data indicate that the light-dependent dynamics of short actin filaments was not an artifact caused by GFP-talin reporter. Confocal microscopic analysis revealed that these short actin filaments are localized exclusively at the interface between the chloroplast and the plasma membrane. We named these actin filaments as chloroplast-actin filaments (cp-actin filaments).69 When irradiated with high-intensity blue light to induce the avoidance response, the cp-actin filaments on the irradiated chloroplasts transiently disappeared. Shortly after that, the newly formed cp-actin filaments began to appear at the leading edge of the chloroplasts showing “biased” localization (Fig. 2). When irradiated with low-intensity blue light to induce the accumulation response, biased localization of the cp-actin filaments occurred without the transient disappearance. Such localized accumulation of cp-actin filaments in the front region of chloroplasts occurs prior to chloroplast movement. The extension of chloroplast envelope at the leading edge of chloroplast suggests that cp-actin filament accumulation generates the pulling force. When chloroplasts stop moving, the cp-actin filaments redistribute around the entire chloroplast periphery. Importantly, the differences in the amount of cp-actin filaments between the front and the rear halves of chloroplasts were closely correlated with the velocities of the accumulation or avoidance movements; the greater the difference in the cp-actin filament amounts, the faster the chloroplast movement.69 Moreover, the increase in the blue light intensity was found to result in the greater difference in cp-actin filament amounts and the concomitant growth in the velocity of the avoidance movement.69 Since this velocity also depends on the phot2 abundance,61,70 the cp-actin filament regulation is likely to involve the activity of phot2.
The amount of cp-actin filaments on stationary chloroplasts determines the extent of chloroplast anchorage to the plasma membrane.69 High-intensity blue light-induced disappearance of cp-actin filaments correlates with the increase of chloroplast movement in random direction. Conversely, low-intensity blue light induces the increase of the amount of cp-actin filaments and consequently the motility of the irradiated chloroplasts decreases. The analysis of a chup1 mutant line, previously isolated as a mutant deficient in both chloroplast photorelocation and positioning,22,29,33 revealed highly variable and rapid motility of chloroplasts. Importantly, chup1 plants lacked cp-actin filaments, supporting that the role of cp-actin filaments in chloroplast anchorage to the plasma membrane. CHUP1 is localized on the chloroplast outer membrane and can interact with F-actin, G-actin and profilin in vitro.29,33,34 Thus, CHUP1 is a key factor of cp-actin filament regulation that likely links chloroplasts to cp-actin filaments.
The phototropins are also important players in regulation of the blue light-mediated cp-actin filaments. High-intensity blue light-induced disappearance of cp-actin filaments did not occur in phot2 mutant plants69 deficient in chloroplast avoidance response,27 suggesting that phot2-dependent transient disappearance of cp-actin filaments is essential for this response. Although phot1 mutant plants showed normal cp-actin filament dynamics during both accumulation and avoidance responses,69 phot1phot2 double mutant completely lacked the blue light-regulation of cp-actin filaments69 (including high-intensity light-induced disappearance, low-intensity light-induced increase and the increased accumulation at the leading edge of the chloroplasts). As a result, the double mutant plants were impaired in chloroplast photorelocation movement26 and the change of motility.69 These data indicate that phototropins mediate chloroplast photorelocation and the anchorage to the plasma membrane via the regulation of cp-actin filaments.
Detailed observation of the disappearance and reappearance of cp-actin filaments by confocal microscopy revealed that cp-actin filaments shortened towards the chloroplast periphery and disappeared at the edge and that the reappearance of cp-actin filaments occurred at the chloroplast edge and possibly at the same position where cp-actin filaments disappeared.69 These observations suggest that the assembly sites of cp-actin filaments are located on the chloroplast edge and that the cp-actin filaments elongate at the center of the interface between chloroplasts and the plasma membrane. The microscopy using higher optical and temporal resolution is required to further characterize the dynamics of cp-actin filaments generation, elongation and disassembly.
We identified two homologous genes, Kinesin-like protein for actin-based chloroplast movement 1 (KAC1) and KAC2 through the screening of the mutants deficient in chloroplast accumulation response.71 Unexpectedly, KAC genes encode microtubule motor kinesin-like proteins. KAC proteins belong to the kinesin-14 family,72 which includes minus end-directed motors with a C-terminal motor domain, whereas conventional kinesins are plus-end directed motors with a N-terminal motor domain. Previously, KAC1 was identified by yeast two-hybrid screening as a protein interacting with a cyclin-dependent kinase,73 a gemini-virus protein,74 and a katanin.75 Thus, KAC1 is implicated in cell cycle and/or the regulation of microtubules. However, kac1kac2 double mutant plants show normal stature, development and cortical microtubule organization.71 Furthermore, neither microtubule-binding, nor ATPase activities of KAC motor domains are experimentally detectable.71 Bioinformatics analysis indicated that several conserved amino acid residues essential for the motor activity are absent from the core domains of KAC proteins, e.g., a switch I arginine necessary for the salt bridge formation (Fig. 3A) and the two basic residues in L12/α5 region forming a microtubule-interacting surface (Fig. 3B). Interestingly, KAC proteins of a moss Physcomitrella patens retain the conserved switch I motif (Fig. 3A) and at least one basic residue in L12/α5 (Fig. 3B). In line with this and unlike higher plants, P. patens utilizes both actin filaments and microtubules for chloroplast movement.76,77 These observations suggest that, at least in moss, KAC proteins could function as the microtubule motor for chloroplast movement. Amino acids involved in neck/motor core interaction essential for the directionality of minus end-directed motors78 are highly conserved in KAC proteins. When the corresponding KAC1 amino acids were mutated to permit the minus end-directed kinesin to move in both directions78 and the mutant KAC1 was expressed in kac1kac2 plant line, the kac1kac2 mutant phenotypes were partially rescued, but the directionality change of chloroplast photorelocation was not observed.71 In addition, KAC proteins harbor a conserved central coiled-coil region that is likely involved in oligomerization.73
Detailed analyses of chloroplast photorelocation movements revealed that the kac1 mutant plants are severely impaired in the accumulation response and showed slower avoidance response compared to the wild type. Whereas kac2 mutant plants show no discernible defects, kac1kac2 double mutant plants completely lack both the accumulation and the avoidance responses.71 Therefore, KAC1 and KAC2 redundantly mediate chloroplast photorelocation movement with KAC1 playing a principal role. Arabidopsis KAC1 and KAC2 genes reside in the duplicated regions on chromosome 5,79 and the corresponding proteins are very similar. However, KAC1 mRNA and protein accumulate at much higher level than those of KAC2. Interestingly, when KAC2 cDNA was expressed from KAC1 promoter in kac1kac2 mutant background, KAC2 protein accumulated at a level similar to that of KAC1 and rescued the defects in chloroplast photorelocation movement.71 Thus, the difference between kac1 and kac2 phenotypes is due to the difference in mRNA expression levels but not in protein stability or functional activity of KAC1 and KAC2. In addition to the defect in chloroplast photorelocation, kac1kac2 mutants are impaired in chloroplast attachment to the plasma membrane. Whereas wild-type chloroplasts are stationary with slight Brownian-like movements in the absence of blue light irradiation, chloroplasts in kac1 and kac2 plants exhibit certain motility. In kac1kac2 double mutants, chloroplasts partially aggregate near nucleus and also move with the flow of cytoplasm.71 It can be concluded that, similar to CHUP1, KAC proteins are essential for both chloroplast photorelocation and attachment to the plasma membrane.
Similar to chup1 mutants, kac1kac2 mutants completely lack cp-actin filaments further supporting the causal connection between filament formation and chloroplast movement (Fig. 2).71 The kac2 mutants showed normal cp-actin filament dynamics (Fig. 2), whereas fewer cp-actin filaments were detected in kac1 mutants (Fig. 2). When kac1 mutant plants were irradiated with the weak blue light, biased localization of cp-actin filaments did not occur and chloroplast accumulation response was not induced. Although the avoidance response in the kac1 mutants did take place, it was slower with weaker localized accumulation of cp-actin filaments compared to the wild type. Collectively, these results indicated that KAC proteins are necessary for the cp-actin filament formation or maintenance, processes that are critical for chloroplast photorelocation and plasma membrane anchorage. Furthermore, the abundance of KAC proteins defines the amount of cp-actin filaments and the speed of chloroplast movement.
Although the bulk of KAC proteins localizes in cytoplasm,71,73,80 a small fraction of them is associated with the plasma membrane80 and chloroplasts.81 Since cp-actin filaments exist at the interface between the chloroplast and the plasma membrane,69 this small fraction of KAC proteins is likely to function in cp-actin filament regulation.
It is known that some kinesins are able to interact directly with the actin filaments.82–86 and that the conserved, KAC-specific, C-terminal domain can interact with the F-actin in vitro.71 Intriguingly, KAC proteins are functionally similar to the yeast Smy1p kinesin; both of them bind F-actin and do not require microtubules and motor activity.86,87 Although it is known that Smy1p cooperates with myosin class V Myo2p motor to transport cargo along actin bundles,86 it is not clear whether a similar mechanism operates in chloroplast movement.
The actomyosin systems in animals and yeast are known to generate either myosin-dependent or actin polymerization-dependent forces that are responsible for the transport of organelles, as well as some intracellular pathogenic bacteria and viruses.88 We cannot exclude the possibility that the plant myosins or other, yet unknown, actin-dependent motors generate the motive force for chloroplast movement. However, it seems more likely that the actin polymerization itself is responsible for this process. The Arp2/3 complex-dependent actin polymerization at the cargo edge (i.e., actin comet tail formation) is the best characterized mechanism for the movements of organelles, bacteria and viruses.89 However, Arp2/3 complex-dependent mechanism appears to be intrinsically and mechanistically different from the cp-actin filament-dependent mechanism. Cp-actin filaments are formed at the leading edge of chloroplasts69 possibly generating the chloroplast pulling force, whereas actin comet tails are polymerized at the lagging edge of cargo to generate the pushing force. More importantly, chloroplast movement and cp-actin filament regulation are not affected in Arabidopsis arp2/3 mutant lines.69
Since the cp-actin filaments elongate from the edge of chloroplasts,69 the cp-actin nucleators may also be localized at the edge of chloroplasts and polymerize the cp-actin filaments towards the inward of chloroplast. This polymerization may push the nucleators towards the moving direction of the chloroplast and generate the motive force if the cp-actin filaments are firmly but temporally anchored to the plasma membrane. Extension of chloroplast envelope observed at the front side of chloroplast before its movement supports this hypothesis.69 The disassembly of cp-actin filaments may occur at their farthest ends.69 KAC proteins are one of the likely candidates for anchoring cp-actin filaments to the plasma membrane, because they are localized at the plasma membrane and can interact with F-actins.71,80 This model is highly speculative and thus further investigation of the cp-actin filament dynamics and the regulating factors (e.g., CHUP1 and KAC) is required for better understanding of the involved molecular mechanism.
Each chloroplast can move autonomously and independently from other chloroplasts suggesting that the movement regulation mechanism is chloroplast-centered rather than belongs to the nucleus or cytosol per se. This mechanism is able to read and act upon the environmental light conditions that fluctuate over time. Using the cp-actin filaments localized at their periphery, chloroplasts move towards the weak light to maximize photosynthesis, but away from the excessively strong light to avoid photodamage. On the other hand, plants, fungi and animals have evolved an alternative, myosin-dependent, long-distance transport mechanism for such ubiquitous organelles as mitochondria, peroxisomes, Golgi stacks and ER. We propose that this mechanistic difference is due to a need for each individual chloroplast to respond independently to the very localized change of light intensity. Such precise and prompt chloroplast repositioning in response to the environmental changes necessitates an autonomous rather than centrally controlled motility system. Conspicuously, the genes encoding the three key components of the cp-actin filament-mediated chloroplast movement, PHOT, CHUP1 and KAC, already evolved in the basal lineage of land plants represented by liverwort Marchantia polymorpha.90,91 It seems possible that this unique mechanism of chloroplast movement have facilitated the explosive evolutionary diversification and proliferation of the land plants.
We thank Dr. Sam-Geun Kong for unpublished results and confocal microscopic analysis of cp-actin filaments. This work was supported in part by the Japanese Ministry of Education, Sports, Science and Technology (MEXT 13139203 and 17084006 to M.W.) and the Japan Society of Promotion of Science (JSPS 13304061, 16107002 and 20227001 to M.W.; 20870030 to N.S.). Work in V.V.D. lab is supported in part by NIH ARRA award No. GM087658.
Previously published online: www.landesbioscience.com/journals/psb/article/12802