Light-Induced Movements of Chloroplasts and Nuclei Are Regulated in Both Cp-Actin-Filament-Dependent and -Independent Manners in Arabidopsis thaliana

Light-induced chloroplast movement and attachment to the plasma membrane are dependent on actin filaments. In Arabidopsis thaliana, the short actin filaments on the chloroplast envelope, cp-actin filaments, are essential for chloroplast movement and positioning. Furthermore, cp-actin-filament-mediated chloroplast movement is necessary for the strong-light-induced nuclear avoidance response. The proteins CHLOROPLAST UNUSUAL POSITIONING 1 (CHUP1), KINESIN-LIKE PROTEIN FOR ACTIN-BASED CHLOROPLAST MOVEMENT 1 (KAC1) and KAC2 are required for the generation and/or maintenance of cp-actin filaments in Arabidopsis. In land plants, CHUP1 and KAC family proteins play pivotal roles in the proper movement of chloroplasts and their attachment to the plasma membrane. Here, we report similar but distinct phenotypes in chloroplast and nuclear photorelocation movements between chup1 and kac1kac2 mutants. Measurement of chloroplast photorelocation movement indicated that kac1kac2, but not chup1, exhibited a clear strong-light-induced increase in leaf transmittance changes. The chloroplast movement in kac1kac2 depended on phototropin 2, CHUP1 and two other regulators for cp-actin filaments, PLASTID MOVEMENT IMPAIRED 1 and THRUMIN 1. Furthermore, kac1kac2 retained a weak but significant nuclear avoidance response although chup1 displayed a severe defect in the nuclear avoidance response. The kac1kac2chup1 triple mutant was completely defective in both chloroplast and nuclear avoidance responses. These results indicate that CHUP1 and the KACs function somewhat independently, but interdependently mediate both chloroplast and nuclear photorelocation movements.


Introduction
Organelle movement is essential for many cellular activities and thus needs to be tightly regulated [1,2]. Because land plants are sessile organisms, the organelle movements should be appropriately regulated by environmental signals, such as light. Among plant organelles, chloroplasts change their position in response to light (chloroplast photorelocation movement). Chloroplasts move towards weak light to capture light efficiently (the accumulation response). Conversely, chloroplasts escape from strong light and move to a position where light absorption is minimized (the avoidance response) [3,4]. Phototropin (phot) is the blue light receptor for chloroplast photorelocation movement. In Arabidopsis thaliana, two phototropins, phot1 and phot2, redundantly mediate the accumulation response, and phot2 primarily regulates the avoidance response [5][6][7].
In general, organelle movement is dependent on actin filaments in land plants. The movements of most organelles, such as mitochondria, Golgi bodies, and endoplasmic reticulum, rely on the cytoplasmic actin cables and motor protein myosins. However, chloroplast movement is mediated by specialized chloroplast-actin (cp-actin) filaments and is likely independent of myosins [8]. Cp-actin filaments are localized mostly on the chloroplast periphery, between chloroplasts and the plasma membrane [9,10]. Blue light regulates the amount, and the positions, of cp-actin filaments. Strong blue light induces the disappearance of cp-actin filaments by severing the filaments in a phot2-dependent manner [10], and then cp-actin filaments are polymerized at the future front regions of the chloroplasts [9][10][11]. Weak blue light induces the polymerization of cp-actin filaments without the detectable severing [9,10]. The blue-lightinduced reorganization of cp-actin filaments is mediated by phototropins [9][10][11], and phot2 is a master regulator of the strong-light-induced response [10]. Other components are involved in the light regulation of cp-actin filaments. A J-domain protein, J-DOMAIN PROTEIN REQUIRED FOR CHLOROPLAST ACCUMULATION RESPONSE 1 (JAC1), plays a particularly important role in the accumulation response [12] and is essential for the reorganization of cp-actin filaments during the weak-light-induced accumulation response [11]. Mutant plants of two interacting coiled-coil proteins, WEAK CHLOROPLAST MOVEMENT UNDER BLUE LIGHT 1 (WEB1) and PLASTID MOVEMENT IMPAIRED 2 (PMI2), are defective in the blue-light-induced reorganization of cp-actin filaments during the avoidance response, and thus, web1 and pmi2 mutants exhibit the attenuated chloroplast avoidance response [13,14]. A C2 domain protein, PLASTID MOVEMENT IMPAIRED 1 (PMI1), is essential for chloroplast movement, and the pmi1 mutant is severely defective in chloroplast photorelocation movement [15]. The cp-actin filaments are labile in pmi1, indicating that PMI1 is necessary for the stability of cp-actin filaments [16]. THRUMIN1 (THRUM1) is an actin-binding and -bundling protein [17]. The thrum1 mutants are partially defective in chloroplast movement and are severely impaired in the accumulation of cp-actin filaments [10,17].
The light-induced reorganization of cp-actin filaments was found in the fern Adiantum capillus-veneris [18] and the moss Physcomitrella patens [19], indicating that the cp-actin-filament-based chloroplast movement is conserved among land plants. Two protein families, CHLOROPLAST UNUSUAL POSITIONING1 (CHUP1) and KINESIN-LIKE PROTEIN FOR ACTIN-BASED CHLOROPLAST MOVEMENT (KAC), are indispensable for the polymerization and/or maintenance of cp-actin filaments and have conserved functions in land plants [20][21][22][23][24][25]. In Arabidopsis, mutants deficient in CHUP1 or two KAC proteins (KAC1 and KAC2) show severe defects in chloroplast movement and positioning [20,21,22,26]. The chloroplasts of these mutant plants are detached from the plasma membrane, resulting in aberrant positioning and the absence of directional photorelocation movement because they lack detectable cpactin filaments [9,10,11,22]. CHUP1 is a multi-domain protein that consists of an N-terminal hydrophobic region, a coiled-coil domain, an actin-binding motif, a proline-rich region and a conserved C-terminal region [20]. CHUP1 is localized on the chloroplast outer envelope through the N-terminal hydrophobic region [20,21,27]. CHUP1 interacts with F-actin, Gactin, and profilin in vitro probably through the C-terminal region, which includes the actinbinding motif and the proline-rich region [20,27]. The N-terminal coiled-coil domain serves as a dimerization domain [28] and is essential for the binding of CHUP1 to the plasma membrane [21]. KAC is a microtubule motor kinesin-like protein. Although KAC belongs to the kinesin-14 family, including minus end-directed motors with a C-terminal motor domain, no detectable microtubule motor activity was observed [22,29]. Similar phenotypes between chup1 and kac1kac2 in Arabidopsis suggest that CHUP1 and KAC proteins coordinately mediate cp-actin-mediated chloroplast movement and positioning, although the mechanism is unknown.
The movement of nuclei is also regulated by blue light [30] and dependent on phototropins in Arabidopsis [31] and the fern A. capillus-veneris [32]. In Arabidopsis' mesophyll and pavement cells the nuclei are situated on the cell bottom in the darkness (dark position) and move to the side walls (light position) in response to strong blue light as a result of the nuclear avoidance response [31,33,34]. We recently demonstrated that the nuclear avoidance response is mediated by the movement of plastids and chloroplasts attached to the nuclei in Arabidopsis' pavement and mesophyll cells, respectively [34]. Using green fluorescent protein (GFP)-talin lines, the position of the nucleus and the associated actin filaments were visualized simultaneously [34]. When the side of the nucleus along the long axis is partially irradiated with strong blue light and the pattern of nuclear movements was analyzed, two types of light-induced nuclear movements are observed. One is "the avoidance movement" in which the nuclei escape from the irradiated side towards the non-irradiated side. The avoidance movement depends on cp-actin-filament-dependent movement of plastids attached to the nuclei and is induced specifically by blue light in a phot2-dependent manner [34]. The other is "the parallel movement" in which the nuclei move parallel to the actin bundles. The parallel movement is independent of plastids and is induced mostly in a photosynthesis-dependent manner [34]. Mutant analyses have indicated that the nuclear avoidance response is dependent on CHUP1 and PMI1, in addition to phot2 [16,34]. The nuclear avoidance response in mesophyll cells is mediated by PMI1 alone, but the response in the pavement cells depends on both PMI1 and the homologous PMI1-RELATED1 (PMIR1), but not on PMIR2 [16]. However, the role of the KAC proteins in the nuclear avoidance response remained to be determined because the blue-lightinduced movements of plastids and nuclei had not been analyzed in kac1kac2 pavement cells.
To understand the role of KAC proteins, especially the relationship between KAC and other proteins, we generated multiple mutant plants between kac1kac2 and other mutants and analyzed light-induced movement of chloroplasts and nuclei in these mutants. Here, we found clear differences in chloroplast and nuclear movements between chup1 and kac1kac2 in Arabidopsis, although previous studies suggested that CHUP1 and KAC proteins function as the same pathway [22,24].

Plant materials and plant growth
Arabidopsis seeds (Columbia) were sown on one-third-strength Murashige and Skoog culture medium containing 1% (w/v) sucrose and 0.8% (w/v) agar. After incubation for 2 d at 4°C, the plants were cultured under white light at approximately 100 μmol m -2 s -1 under a 16/8-h light/ dark cycle at 23°C in a growth chamber. Approximately 2-week-old plants were used for the analyses of chloroplast and nuclear photorelocation movements. To observe the chloroplast distribution, plants were cultured on soil (Metro Mix 350; Sun Gro, Vancouver, BC, Canada) under white light at approximately 80 μmol m -2 s -1 under a 16/8-h light/dark cycle in a growth chamber. The N7 nuclear marker line [35] was provided by the Arabidopsis Biological Stock Center. Double-and triple-mutant plants were generated by genetic crossings. Mutant lines containing the N7 nuclear marker and GFP-mouse-talin [9,10] were generated by genetic crossings.

Analyses of chloroplast photorelocation movements
Chloroplast photorelocation movement was examined by measuring changes in leaf transmittance as described previously [36]. The detached third leaves from 16-day-old plants were placed on 1% (w/v) gellan gum in a 96-well plate. Samples were dark-adapted for at least 1 h prior to transmittance measurements. Blue light was supplied from a blue light-emitting diode illuminator (LED-mB; EYELA). The red light transmittance was automatically measured every 2 min using a microplate reader (VersaMax; Molecular Devices). To disrupt actin filaments, the detached third leaves were treated with 10 μM latrunculin B. The inhibitor treatment was performed by floating the leaves on the inhibitor solution for 12 h.

Observation of the chloroplast distribution
Three-week-old soil-grown plants were dark-adapted for 8 h. Then, plants were kept in the darkness for 2 additional h or irradiated with blue light at 30 μmol m -2 s -1 for 2 h. The leaves were collected and fixed as described previously [34].

Analyses of nuclear photorelocation movements
Time-course experiments for the nuclear photorelocation movement were performed as described previously [34]. After 2-week-old plants were dark-adapted for 24 h, they were irradiated with 50-μmol m -2 s -1 blue light for 12 h. The leaves were collected and fixed at 0, 3, 6, 9, and 12 h after light irradiation, as described previously [34]. Data for wild type and chup1 from [34] were used for comparison because those for kac1kac2 and chup1kac1kac2 were acquired during the same experimental period.

Confocal laser scanning microscopy
The cp-actin filaments, nuclear movement, and the plastids attached to the nucleus were observed under a confocal microscope (SP5; Leica Microsystems) as described previously [10,34]. Fluorescence was observed at 500-550 nm for GFP and 650-710 nm for chlorophyll autofluorescence.

Strong-blue-light-induced changes in leaf transmittance in the kac1kac2 double mutant
To further examine the defective chloroplast photorelocation movements in chup1 and kac1kac2, we examined light-induced changes in leaf transmittance in these mutants ( Fig 1A). This method allows the non-invasive measurement of chloroplast photorelocation movements in multiple leaf cells and layers [37], and defective chloroplast movement in various Arabidopsis mutants were reliably detected [9,14,16,38]. As indicated in Fig 1A, in the wild type, the decrease in leaf transmittance was induced by weak blue light (3 μmol m -2 s -1 ; white arrow in Fig 1) as a result of the accumulation response, whereas strong blue light (20 or 50 μmol m -2 s -1 ; sky blue or blue arrows in Fig 1, respectively) induced an increase in leaf transmittance as a result of the avoidance response. After the light was extinguished (black arrow in Fig 1), the transmittance rapidly decreased and reached the basal level (the dark recovery response). As expected, in the chup1 mutant, no detectable light-induced changes in leaf transmittance were detected ( Fig 1B and S1 Fig). However, unexpectedly, we detected clear leaf transmittance changes in kac1kac2. Although weak blue light did not induce any changes in leaf transmittance, kac1kac2 exhibited a clear increase in leaf transmittance in response to strong blue light ( Fig 1B and S1 Fig), indicating that the avoidance response-like chloroplast movements still occurs in the kac1kac2 double mutant. Furthermore, after the light was extinguished, the decrease in the transmittance was very slow compared with the wild type ( Fig 1B and S1 Fig).
To reveal the regulators involved in mediating the chloroplast movement in the kac1kac2 double mutants, we crossed various mutants defective in chloroplast photorelocation movement with kac1kac2 double mutants and analyzed leaf transmittance changes in the resulting triple or quadruple mutant plants. phot1kac1kac2 displayed a similar phenotype as kac1kac2 (Fig 1D and 1E and S1 Fig; Student's t test, P > 0.08 for kac1kac2 vs. phot1kac1kac2 at 20 μmol m -2 s -1 ). Consistent with phot2 mainly functioning under strong blue light conditions [38], the phot2kac1kac2 mutants were severely deficient in the strong-light-induced-increase in leaf transmittance but retained a subtle weak-light-induced decrease and strong-light-induced increase in leaf transmittance compared with phot1phot2 ( Fig 1D and 1E). These responses were phot1-dependent because phot1phot2kac1kac2 exhibited no detectable changes in leaf transmittance irrespective of the light intensity, similar to phot1phot2 (Fig 1D and S1 Fig; Student's t test, P > 0.1 for phot1phot2 vs. phot1phot2kac1kac2 at all fluence rates). These results indicate that phot2 is the primal photoreceptor for the strong-light-induced chloroplast movements in kac1kac2.
The jac1 mutant plants were unresponsive to weak blue light but responsive to strong blue light as a result of the avoidance response [12,14]. Although the pattern of changes in the leaf transmittance was similar between jac1 and kac1kac2, the increase in the leaf transmittance in response to strong light, especially at 20 μmol m -2 s -1 , was much larger in jac1 than in kac1kac2. Furthermore, the jac1kac1kac2 triple mutant plants resembled the kac1kac2 (S1 Fig). These results indicate that the avoidance response induced in jac1 mutant plants is largely KAC-independent and vice versa. Both web1 and pmi2pmi15 mutant plants were defective in the strong-light-induced avoidance response [14], and thus, only slight increases were induced in both mutants (S1 Fig). However, web1kac1kac2 and pmi2pmi15kac1kac2 exhibited larger increases in the transmittance change when compared with web1 and pmi2pmi15, respectively (S1 Fig). The phenotypes of web1kac1kac2 and pmi2pmi15kac1kac2 somewhat resembled those of kac1kac2, indicating that the kac1kac2 mutation suppressed the weak avoidance response in web1 and pmi2pmi15, similar to the jac1 mutation [14]. Because jac1kac1kac2, web1kac1kac2, and pmi2pmi15kac1kac2 showed a strong-light-induced increase in (a) Analysis of chloroplast photorelocation movements by measuring the light-induced changes in leaf transmittance. A black trace represents a typical light-induced change in wild type under our light irradiation condition (indicated by a colored bar above a graph). Transmittance changes are indicated as Δ transmittance (a transmittance value at the indicated time minus the transmittance at 0 min). After the measurement was performed for 10 min in darkness, samples were sequentially irradiated with continuous blue light at 3, 20, and 50 μmol m -2 s -1 for 60, 40, and 40 min, as indicated by white, sky blue, and blue arrows, respectively. The light was extinguished at 150 min (black arrow). Double-headed arrows indicate the chloroplast movements induced under the indicated light conditions. Red lines indicate the initial linear fragments (from 2 to 6 min after changes in light fluence rate) which were used for the calculation of the percentage transmittance change over 1 min (i.e., speed of leaf transmittance changes). transmittance similar to that of kac1kac2, this indicated that JAC1, WEB1, and PMI2 (and PMI15) are dispensable for the strong-light-induced chloroplast movements in kac1kac2.
Similar to chup1, chup1kac1kac2 is defective in any chloroplast photorelocation movement (Fig 1A and 1B; Student's t test, P > 0.1 for chup1 vs chup1kac1kac2 at all fluence rates). Interestingly, pmi1kac1kac2 and thrum1kac1kac2 synergistically impaired chloroplast photorelocation movements compared with kac1kac2, pmi1, and thrum1. They were similar to that of chup1 (Fig 1E and 1F; Student's t test, P > 0.2 for chup1 vs. pmi1kac1kac2 or thrum1kac1kac2 at all fluence rates), indicating that PMI1 and THRUM1 are required for the strong-lightinduced chloroplast movements in kac1kac2. Table 1 summarizes the mutant phenotypes (Table 1).

Strong-blue-light-induced changes in leaf transmittance in the kac1kac2 double mutant depends on actin filaments
Our previous analyses indicated that kac1kac2 lacks detectable cp-actin filaments [8,22]. Thus, it is plausible that KAC-independent chloroplast movement is also independent of cp-actin filaments. To examine whether KAC-independent chloroplast movement is still actin-dependent, we investigated chloroplast photorelocation movement in wild type and kac1kac2 mutant plants treated with the actin polymerization inhibitor latrunculin B, which inhibits blue-lightinduced movements of chloroplasts and nuclei in Arabidopsis wild type plants [21,33]. Treatment with latrunculin B completely abrogated any blue-light-induced changes in leaf transmittance in both wild type and kac1kac2 mutant plants, although the leaf transmittance continued to increase in a light-independent manner (Fig 2 and Table 1). Thus, the strong-light-induced chloroplast movements in kac1kac2 are also dependent on actin filaments.
Subtle strong-blue-light-induced chloroplast movement from the periclinal to anticlinal walls in the kac1kac2 double mutant Very few chloroplasts are localized on the mesophyll cell surface even under weak light conditions, when chloroplasts cover the cell surface in the wild type as a result of the accumulation response [22]. Furthermore, experiments with microbeam irradiation revealed no detectable directional movements in response to the strong blue light in kac1kac2 [22]. However, when we closely observed the chloroplast distribution in dark-adapted or strong-light-irradiated Table 1. Summary of the data on chloroplast photorelocation movements in Fig 1 and Fig 2. Mutant phenotypes or inhibitor effect in wild type or kac1kac2 mutant backgrounds are indicated. Ac: accumulation response, Av: avoidance response, Av (kac-like): the avoidance response found in kac1kac2 mutants, -: no or severely defective responses. Here, strong blue-light-induced response in jac1 is categorized into kac-like avoidance response.  plants, we noticed slight light-induced changes in the chloroplast distribution in kac1kac2 but not in chup1 or chup1kac1kac2 (Fig 3). In the wild type, some chloroplasts were situated on the cell surface and anticlinal cell walls in the darkness, whereas chloroplasts covered the whole anticlinal wall surface under the strong light condition (Fig 3A), resulting in an increase in the ratio of the area occupied by chloroplasts (called the "chloroplast area") to the area of the whole cell surface in the dark (Fig 3B; Student's t test, P < 0.05 for dark vs. strong light). In the darkness, the chloroplast area was slightly smaller in both chup1 and kac1kac2 than in the wild type, and in response to the strong blue light the area did not change in chup1 (Fig 3B; Student's t test, P > 0.3 for dark vs. strong light). Statistically insignificant, but the area slightly increased in kac1kac2 (Fig 3B; Student's t test, 0.05 < P < 0.1 for dark vs. strong light). Most chloroplasts of the chup1kac1kac2 mutant did not localize on the cell surface. The chloroplast area was much smaller than in chup1 or kac1kac2 and did not change in response to light (

KAC proteins and CHUP1 redundantly mediate the nuclear avoidance response in pavement cells
In approximately 70% of wild type pavement cells, nuclei were situated on the cell bottom in darkness. A strong blue light induced nuclear movements from the cell bottom to the anticlinal walls, and nuclei were in the light position in 70% of the pavement cells after 6 h. Thus, a 40% increase in nuclei in the light position occurred in response to the strong blue light in wild type (Fig 4). Consistent with our previous results [34], chup1 mutants were severely impaired in the nuclear avoidance response and exhibited a subtle increase in the nuclei in the light position. However, nuclei in only 35% of the cells were in the light position even after 12 h (Fig 4). The kac1kac2 mutant showed a slight defect in dark positioning and nuclei were positioned on the cell bottom in 60% of the pavement cells, which was 10% less than in wild type (Fig 4). Furthermore, the avoidance response was partially defective in kac1kac2. In response to a strong blue light, only a 20% increase in nuclei in the light position occurred, and the nuclei were in the light position in 60% of the pavement cells even after 12 h (Fig 4). Thus, kac1kac2 is partially defective in the nuclear avoidance response although it exhibited much weaker defects than chup1, which is similar to the chloroplast photorelocation movement, as mentioned above. The chup1kac1kac2 triple mutant was completely impaired in the nuclear avoidance response and did not display any light-induced changes in the nuclear position. Interestingly, 45-50% of nuclei were in the light position irrespective of light conditions (Fig 4).
To further analyze the light-induced nuclear movement, the side of the nucleus along the long axis was irradiated with strong a blue light microbeam in lines expressing GFP-talin (see Introduction) [34]. In wild type, approximately 80% of the nuclei exhibited light-induced movement and about 40% of the nuclei showed the avoidance movement (Fig 5A; data from [34]). The chup1 mutant had severely attenuated the avoidance movements and the parallel movement was prominent in chup1, indicating that cp-actins existed on the plastids attached to the nuclei, which is essential for the avoidance response and to suppress parallel movements (Fig 5A; data from [34]; chi-square test, P > 0.1 for wild type vs. chup1). The kac1kac2 was similar to wild type (chi-square test, P < 0.0005 for wild type vs. kac1kac2), but only approximately 20% of the nuclei showed the avoidance movement (Fig 5A and 5B). This subtle defect in the avoidance movement explains the partial defects in the nuclear avoidance response in kac1kac2. The chup1kac1kac2 triple mutant plants were similar to the chup1 plants in that the parallel movement was dominant and almost no avoidance movement was observed (Fig 5A  and 5B; chi-square test, P > 0.1 for chup1 vs. chup1kac1kac2). However, compared with chup1, the proportion of parallel movement was reduced and the absence of movement was increased.
Approximately half of the plastids were attached to the nucleus in wild type pavement cells and more plastids were associated with the nucleus in chup1 mutant pavement cells [34]. The relative adhesive power between plastids and the plasma membrane may be dependent on cp-actin filaments, which determined the number of plastids attached to the nucleus [34]. Interestingly, all of the plastids were attached to the nucleus in pavement cells of chup1kac1kac2 triple mutant plants ( Fig  5C; Student's t test, P > 0.5 for attached to nucleus vs. whole cell), indicating that the chup1kac1kac2 triple mutant plants are completely defective in the attachment of plastids to the plasma membrane.

Distinct functions of CHUP1 and KAC proteins in chloroplast photorelocation movement
In this report, we found distinct phenotypes for chup1 and kac1kac2 with regard to chloroplast photorelocation movement and the nuclear avoidance response. Detailed microbeam experiments indicated that kac1kac2 mutants were totally defective in the "directional" chloroplast photorelocation movement [22], similar to chup1 mutant plants [9]. The biased localization of cp-actin filaments at the front region of the chloroplast is essential for the directional chloroplast movement, and the greater difference in the amount of cp-actin filaments between the front and rear regions is necessary for efficient chloroplast movement [9,10,11,22]. Together with the reduced attachment of chloroplasts from the plasma membrane, these phenotypes are largely attributable to the lack of cp-actin filaments in both mutants [9,10,11,22]. The measurement of leaf transmittance changes revealed that a clear chloroplast movement in response to strong blue light was observed in kac1kac2, but not in chup1 or chup1kac1kac2. Additionally, we detected subtle but significant light-induced changes in the chloroplast distribution pattern on the uppermost mesophyll surface in kac1kac2, but not in chup1 or chup1-kac1kac2. Although our attempts to detect the light-induced chloroplast movement in kac1kac2 under the microscope failed, we could detect the light-induced movement in kac1kac2 through the measurement of leaf transmittance in which the sum of the subtle changes in the multiple cell layers is detectable.
In addition to Arabidopsis, a functional difference between CHUP1 and KAC was found in the moss P. patens [23][24][25]. The blue-light-induced chloroplast movement is dependent on both actin filaments and microtubules, and thus, is inhibited only when they are both disrupted by inhibitors [39]. Among the three P. patens CHUP1 genes (PpCHUP1A, B, and C), PpCHUP1A is essential for actin-dependent chloroplast movement [23]. Because the structure and dynamics of short actin filaments on chloroplasts similar to Arabidopsis cp-actin filaments were found in P. patens [19], at least PpCHUP1A is likely necessary for the regulation of these filaments in P. patens. The PpCHUP1A single and PpCHUP1A/B double knockout plants had impaired actin-dependent chloroplast movement but retained normal microtubule-dependent chloroplast movement, and they showed the normal attachment of chloroplasts to the plasma membrane [23]. Plants containing knockouts of two P. patens KAC genes (PpKAC1 and PpKAC2) exhibited severe defects, such as no photorelocation movement and the strong aggregation of chloroplasts [24]. Although the complete aggregation of chloroplasts in the PpKAC1/ 2 double knockout precluded the analysis of the involvement of PpKACs in actin-filament-or microtubule-dependent chloroplast movement, the analyses of PpKAC1/2 double RNAi lines revealed that PpKACs primarily mediate actin-filament-chloroplast movement [25]. Therefore, CHUP1 and KACs coordinately regulate actin-filament-dependent chloroplast movement but also show some independent functions in land plants.
In Arabidopsis, CHUP1 and KACs exhibited quite distinct localization patterns although some parts of these proteins likely co-localized. When the functional fusion with florescent proteins were examined, CHUP1 localized on the chloroplast outer envelope [20,21,27,34], whereas KAC localized primarily in the cytosol and partly on the plasma membrane [22,40,41]. A western blot analysis using fractionated protein showed the difference between CHUP1 and KACs; CHUP1 was detected in the microsomal fraction but KACs were detected primarily in the soluble fraction [16,22,40]. Nevertheless, KACs might mediate cp-actin-filamentdependent chloroplast movement only at the CHUP1-localized site, i.e., at the interface between the plasma membrane and the chloroplast where cp-actin filaments are found [9,10], because CHUP1 is essential for KAC-dependent chloroplast movement.
KAC-independent chloroplast photorelocation movement is mediated by phot2, CHUP1, PMI1, and THRUM1 and requires intact actin filaments The strong-light-induced chloroplast movement found in kac1kac2 was severely attenuated in phot2kac1kac2, chup1kac1kac2, pmi1kac1kac2, and thrum1kac1kac2, indicating that phot2, CHUP1, PMI1, and THRUM1 are required for the KAC-independent chloroplast movement (Fig 6). Although phot2 is the main photoreceptor under the strong light condition, a very slight light-induced chloroplast movement was detected in phot2kac1kac2 but not in phot1-phot2kac1kac2, indicating that phot1 can mediate chloroplast movement in the absence of KACs, although inefficiently. Similarly, the phot1-dependent chloroplast avoidance response was observed in the phot2 mutant in response to very strong white light [42]. Compared with phot1, phot2 localized on the chloroplast outer envelope at a higher amount [43]. Thus, irrespective of KAC-dependence, strong-blue-light-induced chloroplast movements might depend on phototropins localized on the chloroplast outer envelope.
Although CHUP1, PMI1, and THRUM1 are required similarly in the KAC-independent chloroplast movement, the mutant phenotypes of chup1, pmi1, and thrum1 are quite different. The chup1 mutant lacked any light-induced chloroplast movement (Fig 1) [10,20]. Cp-actin filaments have not been detected in the chup1 mutant under any light conditions [9][10][11]. Besides regulating cp-actin filaments, CHUP1 has the ability to associate with the plasma membrane through the coiled-coil region [21]. The CHUP1 N-terminal region, including the N-terminal chloroplast targeting signal sequence and the coiled-coil domain, is sufficient to mediate the attachment of chloroplasts to the plasma membrane, although the F-actin binding domain and the C-terminal region are implicated in the cp-actin filament regulation [21]. Thus, the connection between the chloroplast and the plasma membrane via the CHUP1 Nterminal region might play an important role in the KAC-independent chloroplast movement. Cp-actin filaments were not detected in thrum1, although thrum1 exhibited a weak but significant chloroplast movement (both the accumulation and avoidance responses) (Fig 1) [10,17]. THRUM1 has the ability to bundle actin filaments in vitro [17] and to interact with actin filaments in vivo [10,17]. Because THRUM1 interacts with both cp-actin filaments and cytoplasmic actin filaments, THRUM1 might mediate KAC-independent chloroplast movement through the interaction with cytoplasmic actin filaments. Consistent with a weak chloroplast photorelocation movement, cp-actin filaments were unstable and detected only under certain light conditions in pmi1 mutants [16]. However, the biochemical function of PMI1 remained to be determined, and thus, it is difficult to imagine the role of PMI1 in KAC-independent and cp-actin-filament-dependent chloroplast movements.

CHUP1 and KACs are indispensable for the nuclear avoidance response
The nuclear avoidance response is dependent on cp-actin-filament-mediated photorelocation movement of plastids in pavement cells and in mesophyll cells [16,34]. The chup1 mutant lacking cp-actin filaments was severely defective in the nuclear avoidance response but we detected A model for strong-blue-light-induced avoidance response. In wild type, phot2 (and also phot1) perceive strong blue light and mediates cp-actin-filament-mediated avoidance response in CHUP1-and KAC-dependent manners. At least in kac1kac2 mutant plants, light-induced chloroplast movement is induced in actin-filament-and CHUP1-dependent manners (but cp-actin-filament-independent). PMI1 and THRUM1 regulate both cp-actin-dependent or-independent responses, but the relationship between these proteins, CHUP1, and KAC is unclear.
doi:10.1371/journal.pone.0157429.g006 a very subtle nuclear avoidance response in the chup1 mutant. The kac1kac2 mutant exhibited a weak but substantial nuclear avoidance response, and the chup1kac1kac2 triple mutant exhibited no light-induced changes in the nuclear distribution pattern, indicating that CHUP1 and KACs redundantly mediate the nuclear avoidance response in pavement cells. When examined using microbeam irradiation, two types of light-induced nuclear movements were observed, phot2-dependent "avoidance movement" from the irradiated side towards the non-irradiated side, which depends on the cp-actin-filament-dependent movement of plastids attached to the nuclei, and the photosynthesis-dependent "parallel movement" along the actin bundles that are independent of the plastids [34]. The avoidance movement was attenuated in kac1kac2, and was abrogated in chup1 and chup1kac1kac2, but the parallel movement was enhanced in these mutant plants compared with in wild type, indicating that the parallel movement is independent of CHUP1 and KAC. Compared with chup1 single mutant plants [34], the proportion of the parallel movement was slightly decreased in chup1kac1kac2 and, consequently, the proportion of nuclei unresponsive to light was higher in chup1kac1kac2. Because KAC1 shows actin-binding activity [22], the KACs might mediate parallel movement through the actinbinding, at least in the absence of CHUP1. Interestingly, approximately half of the nuclei are in the light position irrespective of the light condition. This phenotype is highly similar to that of the pmi1pmir1 double and pmi1pmir1pmir2 triple mutant plants [16], indicating that CHUP1, KAC, and PMI1/PMIR1 should function at a similar step. The phot2 mutant also exhibited no light-induced changes in the nuclear distribution pattern, but only approximately 30% of the nuclei were in the light position regardless of the light condition, which was similar to darkadapted wild type plants (70% in the dark position) [34], indicating that CHUP1, KAC, and PMI1/PMIR1 mediate nuclear positioning independent of phot2 in the darkness. Although cpactin filaments were not detected on pavement plastids in chup1 and pmi1pmir1pmir2 mutant plants [16,34], phot2 retained cp-actin filaments on the pavement plastids [34]. Therefore, cpactin filaments should be required for the nuclear positioning on the cell bottom in darkness.

Conclusion
We revealed that CHUP1 and KAC proteins cooperatively, but somewhat independently, function during strong-blue-light-induced movements of chloroplasts and nuclei. Although cpactin filaments are required for the fast directional movements of chloroplasts [9][10][11], our present findings indicate that another actin-dependent mechanism control light-induced changes in chloroplast positioning independently of cp-actin filaments. This cp-actin filament-independent mechanism still be dependent on the factors involved in the regulation of cp-actin filaments, i.e., CHUP1, PMI1, and THRUM1. Together with phototropin and PMI1/PMIR, CHUP1 and KAC are highly conserved factors in Streptophytes and thus these proteins are core factors for chloroplast movement in Streptophytes [44]. Unraveling the molecular properties of these factors is required for understanding the molecular mechanism.
Supporting Information