CO 2 -induced chloroplast movement in one cell-layer moss leaves

CO 2-induced chloroplast movement was reported in the monograph by Gustav Senn in 1908: unilateral CO 2 supply to the one cell-layered moss leaves induced the positively CO 2-tactic periclinal arrangement of chloroplasts. However, from the modern criteria, several experimental settings are unacceptable. Here, using a model moss plant Physcomitrium patens, we examined basic features of chloroplast CO 2-tactic relocation with a modernized experimental system. The CO 2 relocation was lightdependent and especially the CO 2 relocation in red light was substantially dependent on photosynthetic activity. Between the cytoskeletons responsible for chloroplast movement of P. patens, the microfilament mainly worked for CO 2 relocation, but the microtubule-based movement was insensitive to CO 2. The CO 2 relocation was induced not only by air with and without CO 2 but also by the more realistic difference in CO 2 concentration between the two sides. In the leaves placed on the surface of a gel sheet, chloroplasts avoided the gel side and positioned in the air facing surface. This was also shown to be photosynthesis dependent. Based on these observations, we propose a working hypothesis that the threshold light intensity between the light-accumulation and -avoidance responses of the photorelocation would be increased by CO 2, resulting in the CO 2-tactic relocation of chloroplast.


Introduction
Chloroplasts in photosynthetic cells move and change positions in response to environmental and/or endogenous cues, resulting in a specific arrangement within a cell. Such phenomenon is called 'chloroplast relocation.' Chloroplast relocation is supposed to optimize the chloroplast position for its physiological activities, represented by photosynthesis. In photosynthesis, pivotal environmental factors are light, temperature, and CO 2 . Among chloroplast relocations in response to these factors, the accumulation and the avoidance responses in photorelocation and the cold avoidance responses (Böhm, 1856;Haberlandt, 1876;Haupt, 1965) have been extensively studied. Studies on chloroplast relocation employing molecular biology have identified the responsible photoreceptors: phototropin in various land plants, phytochrome in mosses, and neochrome in ferns and algae (Sakai et al., 2001;Kawai et al., 2003;Kasahara et al., 2004;Mittmann et al., 2004;Suetsugu et al., 2017), and furthermore revealed the mechanism of temperature perception by phototropin (Kodama et al., 2008;Fujii et al., 2017).
The motility systems for chloroplast relocation have also been attracted attention. It is widely thought that the photorelocation and the cold avoidance are solely driven by the actin microfilament (MF) system and not by the microtubule (MT) system among land plants Kimura & Kodama 2016;Wada & Kong 2018). Several components like CHUP1, KAC and PMI1 have also been identified to be involved in MF-based photorelocation (Suetsugu et al., 2016;Wada & Kong 2018). In protonema cells of Physcomitrium patens , however, not only MF-system but also MT-based motile system actuates chloroplast movement induced by blue light. In the blue light response, the chloroplast movement based on MT system was faster than that by MF-based system. It was further revealed that the red-light response relied only on MT (Sato et al., 2001).
It is widely believed that the photorelocation is dependent on the specific photoreceptors. However, it has been reported that an inhibitor of the photosynthetic electron transport, DCMU, diminishes the phytochromedependent chloroplast motility, anchorage, and thereby the consequent positional changes in Vallisneria gigantea , an aquatic angiosperm (Seitz, 1967;Dong et al., 1995;Dong et al.,1998;Sakai & Takagi, 2005). DCMU also inhibits the neochrome-independent periclinal positioning in ferns (Sugiyama & Kadota, 2011). Hydrogen peroxide, which is partly generated from O 2 molecule reduced by photosystem I especially when chloroplasts are exposed to excess light under stress conditions (Asada, 1999), is also known to enhance the photorelocation (Wen et al., 2008). These observations indicate that the photosynthetic activity is also an important factor for the light response of chloroplast relocation.
Compared with the studies on the light and temperature effects, there have been few recent studies directly examining effects of CO 2 concentration on chloroplast relocation. In porous leaf tissues of land plants, chloroplasts tend to be located along the cell surfaces facing the intercellular space while the positions adjacent to neighboring cells are avoided. Such arrangement of chloroplast is called 'epistrophe' (Haberlandt, 1886). Epistrophe of chloroplasts has been reported broadly from the mesophyll tissues in angiosperm leaves 2 to the assimilation filaments in liverwort thalli (Haberlandt, 1886;Senn, 1908;Evans et al., 1994). This chloroplast position has been argued to facilitate CO 2 supply for photosynthetic carboxylation and to be related to positive CO 2 -tactic movement of chloroplast, as it were the 'CO 2 taxis' (Haberlandt, 1886;Evans et al.,1994). Another interesting example of chloroplast relocation occurs in C4 plant leaves. Chloroplasts in mesophyll cells aggregate towards the bundle sheath cells in response to stresses like high light or drought (Yamada et al., 2009;Kato et al., 2022). CO 2 leak from the bundle sheath cells would increase under such stress conditions (Ubierna et al. 2013), which would lead a hypothesis that this chloroplast aggregation is the result of positive CO 2 -tactic movement. However, neither the epistrophe arrangement of chloroplasts nor the chloroplast aggregation in C4 plants has been proved to be CO 2 relocation. This is due to a difficulty in distinguishing the CO 2 -induced relocation from other chloroplast relocations like photorelocation. Also difficult is to control the CO 2 concentration gradient in one cell in the complicated and compound tissues to test whether the gradient of CO 2 affects the chloroplast position.
There was an observation of CO 2 effect on chloroplast relocation in a filamentous alga Mougeotia sp. A single ribbon-shaped chloroplast in a cell takes a horizontal orientation perpendicular to the light direction in low light and the profile (parallel) orientation in high light. CO 2 elimination caused an increase in the threshold light intensity that caused the orientational change from the horizontal to profile orientations (Mosebach, 1958). However, this plant, in which the chloroplast just rotates and does not migrate, is not suitable for the study of CO 2 -tactic movement.
The phyllid, simple one cell-layered leaf of mosses would enable us to manipulate CO 2 distribution in each cell and provide an ideal system to study the CO 2 -tactic movement of chloroplasts separately from other chloroplast relocation mechanisms. Indeed, taking this advantage of the moss leaf, Senn (1908) assessed the CO 2 -tactic movement of chloroplasts more than a hundred years ago. He first found an interesting chloroplast position in a leaf of the Funaria hygrometrica moss placed on the surface of gelatin gel: chloroplasts were positioned along the surface facing the air, while there were few chloroplasts along the opposite surface attached to the gel. This observation reminded him of the chloroplast epistrophe positioning and the idea of 'CO 2 taxis' and prompted him to test the CO 2 -tactic movement. He devised a double-chamber system, which was made of two glass half-chambers, each of which had a funnel-shape throat on one side. A thin mica sheet having a slit was sandwiched with these half-chambers at their throat parts. The slit was covered with a F. hygrometrica leaf, the margin of which was fixed with gelatin. Into each of the half chambers, the CO 2 -containing air or the CO 2 -free air was introduced, and the leaf was illuminated for enough time. Following microscopic observation of the leaf revealed the chloroplasts tended to accumulate to the periclinal surface facing the CO 2 -containing air. This description by Senn (1908) was the first and the only report which showed explicitly the occurrence of the CO 2 -tactic movement of chloroplast. In his report, however, he did not detail the effects of the leaf adaxial-abaxial polarity, leaf age, direction of the gravity, or the incubation time. He did not finely control light levels, spectra of light, or the CO 2 concentration, either.
Here, we report characteristics of moss leaf chloroplast CO 2 relocation in leaves of Physcomitriumpatens (syn. Physcomitrella patens ), a model moss plant in detail according to observations using a modernized doublechamber system. The system was equipped with monochromatic LED light sources and mass flow controllers to finely control the spectral properties and PFD level of light and the CO 2 and O 2 concentrations. Using this system in a gas | gel and a gas | gas mode, we also examined the photosynthesis dependency and the responsible cytoskeletal system of CO 2 -tactic movement in a pharmacological manner. Based on the results of these observations, we discuss ecological and physiological aspects of the chloroplast CO 2 relocation.

Plant material
Physcomitrium patens (syn. Physcomitrella patens ) Cove-NIBB line (Nishiyama et al., 2000) was cultured on the BCD medium supplemented with 1 mM CaCl 2 and 1.0% [w/v] agar, which contained 1 mM KNO 3 , 1.84 mM KH 2 PO 4 , 1.32 mM CaCl 2 , 1 mM MgSO 4 , 0.23 μM CoCl 2 , 0.22 μM CuSO 4 , 45 μM FeSO 2 , 0.1 μM Na 2 MoO 4 , 2 μM MnCl 2 , 0.19 μM ZnSO 4 , 1 μM H 3 BO 3 , 0.17 μM KI and 1.0% [w/v] agar (Hereafter we call this medium simply the BCD agar medium and the medium without agar as the BCD medium) and was adjusted at pH6.5 with KOH, in continuous white fluorescent light (FL40SEX-N-HG, NEC Lighting Ltd., Tokyo, Japan) at PPFD 39˜45 μmol m -2 s -1 with no day/night cycle at 18°C and 80% relative humidity in a growth chamber (LPH-350-SP, Nippon Medical & Chemical Instruments, Osaka, Japan). A bunch of protonemata was transplanted on the BCD agar medium and cultured for 9 to 16 days as a colony. Then several shoots emerged in each colony. From each shoot, bearing about 10 leaves with midribs reaching to the leaf tips, three leaves including the largest leaf, and its neighboring younger and older leaves were collected. Since the leaf was folded at its midrib, one side of the lamina from each of the leaves was cut off at the midrib and the half lamina retaining the midrib was used for experiments. The midribs were utilized for handling. Hereafter, we simply call a half lamina retaining the midrib 'a leaf.' Sufficiently flat leaves were selected and used in the experiments. Leaves were immersed in the half strength BCD liquid medium and kept in the dark for one night at 18°C. After the one-night dark treatment, healthy leaves showed the dark arrangement of anticlinal chloroplast positioning (Senn, 1908;Suetsugu et al., 2017; see Before experiment data in Figure  4), and thus, the leaves with anticlinal chloroplast arrangement were used for the experiments.

Double-chamber system
A brass double chamber having glass windows on both sides was custom made (Figure 1a). The volume of each half chamber was 0.7 mL. Mass flow controllers (MFC) (EL-FLOW Select, Bronkhorst, Ruurlo Netherlands, and SEC-400MK2, HORIBA, Kyoto, Japan) were used to obtain gas mixture. The gas was humidified by bubbling in the bottle ( Figure 2). The gas was introduced to each half-chamber at 80 mL min -1 . The concentrations of CO 2 and O 2 of the gas were checked with an infrared gas analyzer (LI-840, LI-COR, Lincoln, NE, U.S.A.) and an oxygen sensor (3080-O 2 , Walz, Effeltrich, Germany). The high vacuum grease (Dow Corning Asia Ltd., Hong-Kong, China) was smeared between the half chambers for airtightness. The temperature of the double chamber was kept at 19-21°C. To illuminate the leaf samples, we used blue (λ = 442 nm) or red (λ = 625 nm) LEDs (10W High Power LED, LED Generic, Yamanashi, Japan) introduced via the four-branched optic fibers (originally provided for a PAM fluorometer 101 system by Walz). A sheet of tracing paper (STP-B5K-105, SAKAE TECHNICAL PAPER, Tokyo, Japan) was placed between the half chamber and the light source for even illumination. Spectra of light sources at the position of the sample in the chamber measured with a light analyzer (LA-105, Nippon Medical & Chemical Instruments) are shown in Figure 1b. The photon flux density (PFD) was measured with a quantum sensor for photosynthetically active radiation (400-700 nm, LI-250A, LI-COR). When continuously used, the PFD tended to increase linearly: the increase after illumination for 3 h was within 5%. The experimental system was covered with black cloth to eliminate light from other light sources. The experiments were conducted in two modes: the gas | gas mode and the gas | gel mode as detailed below. During the experiment, the leaves were placed vertically with the leaf base upwards and illuminated equally from both sides to avoid any unilateral effects of the gravity and light gradient ( Figure 2).

Chloroplast relocation in a gas | gas mode
A piece of aluminum foil (12 μm thickness, UACJ Foil Co., Tokyo, Japan) was fixed over the aperture of the double chamber using the high vacuum grease (Dow Corning Asia) ( Figure 2a). Four or five rectangular windows (approx. 300 μm × 600 μm) were opened in the foil. A leaf sample moistened with the half strength BCD medium was attached over each of the windows with surface tension of the liquid medium (Figure 2c, e). On each of the apical and basal ends of the leaves, a piece of 0.8% agar (FUJIFILM Wako Pure Chemical Co., Osaka, Japan) was placed as water reservoir to prevent desiccation (*in Figure 2c, e). The leaves were kept in the dark in the chamber for 30 minutes and subject to the experimental treatments. Periclinal leaf surfaces were exposed to the gas mixed separately for each side. The PFD level, the gas concentrations and treatment period are shown in each figure.

Chloroplast relocation in a gas | gel mode
A piece of aluminum foil was fixed over the aperture of the double chamber using the high vacuum grease (Dow Corning Asia Ltd.). A square window (6 mm × 6 mm) was opened in the foil. A square sheet in 2 mm thick and 8 mm width of 0.6% [w/v] Gellan gum (FUJIFILM Wako Pure Chemical Co.) gel containing the half strength BCD medium was placed over the window (Figure 1d, and 10c). Four leaves were placed on the gel sheet. Before the experiment, we checked the leaf surface was not covered with the liquid exuded from the gel sheet. The samples were kept in the dark for 30 min in the chamber before the onset of experiments. The same gas mixture was introduced to both half chambers for 2 h. The PFD level and the gas concentration are shown in each figure.

Pharmacological treatments
To inhibit photosynthesis, we used 3-(3,4-dichlorophenyl)-1,1-dimethylurea (DCMU) (Sigma Chemical Co.), a specific inhibitor of electron transport in photosystem II (Trebst 1979), at the final concentration of 100 μM in the half strength BCD medium containing 1% [v/v] DMSO. The leaves after the one-night dark treatment were immersed in the half strength BCD medium containing 1% [v/v] DMSO with or without 100 μM DCMU for 1 h. Chlorophyll fluorescence transients in these leaves were measured with a MINI-PAM-II (Waltz). Complete inhibition of PSII electron transport was confirmed by this DCMU treatment (data not shown). The agar gel used for water reservoir in the gas | gas mode experiment and the gellan gum sheet used in the gas | gel mode experiment were immersed in the half strength BCD medium containing 1% [v/v] DMSO with or without 100 μM DCMU for 1 h.
Oryzalin and latrunculin A (both from FUJIFILM Wako Pure Chemical) were used to disrupt the polymerization of MT and that of actin MF (Yi & Goshima, 2020). Stock solutions of oryzalin and latrunculin A were 100 mM and 10 mM in DMSO. In the experiments, oryzalin and latrunculin A were used at 50 μM and 10 μM in the half strength BCD medium containing 1% [v/v] DMSO. Before the inhibitor treatment, the collected leaves were immersed in the half strength BCD medium for 5 hours in the dark at 18°C. After this dark treatment, the chloroplasts were in the anticlinal arrangement, typical of the arrangement in the dark. These leaves were transferred to the half strength BCD medium containing 1% [v/v] DMSO with or without inhibitor(s) and incubated in the dark overnight at 18°C. The pieces of agar gel used for water reservoirs were immersed in the corresponding inhibitor media for 1 h before the experiments.

Observation on CO 2 -relocation arrangement in transverse sections
After the gas | gas mode experiment with one-sided CO 2 exposure, the leaf samples were immediately immersed in a fixation buffer containing 1% [w/v] glutaraldehyde, 4% formalin and 50 mM sodium phosphate buffer at pH 7.0 for one h on ice in the dark and then in a refrigerator for three days at 4°C to harden the leaves. The samples, being immersed in the fixative, were sectioned transversely by hand at ca. 50 μm thick with a razor blade (Item No. 99027, FEATHER Safety Razor Co., Osaka, Japan) under a stereo microscope (SZX16 equipped with SDF PLAPO 1XPF Objective, Olympus, Tokyo, Japan). Sections were observed under a light microscope equipped with a CCD camera (BX50 equipped with DP71, Olympus).

Quantification of distribution of chloroplasts in each cell
To quantitatively examine chloroplast distribution, we recognized five chloroplast positions: adaxial periclinal position (ad), anticlinal position (ant), abaxial periclinal position (ab), intermediate between ad and ant (ad'), and intermediate between ab and ant (ab') ( Figure 3b). The leaves for quantification were observed under the light microscope (BX50 and DP71) immediately (< 15 min) after the experimental treatment. We took seven to ten photos for one leaf, shifting the focal plane sequentially from the adaxial end to abaxial end ( Figure 3c). On the sequential photos, we counted the numbers of chloroplasts in these categories for each cell. The number of chloroplasts per cell was 27.6 ± 7.6 (Mean ± SD) for 2,381 cells counted for the data shown in the present paper. Here, the proportion of chloroplasts in each category to total chloroplasts of the cell was calculated for each cell: where i is either ad, ad' ant, ab' or ab. The proportions thus obtained are shown in figures.

Statistical analyses 5
To analyze the data statistically, the one-sidedness index: e = 2 * r ad + r ad' -r ab' -2 * r ab , and the flatness index: f = 2 * r ad + r ad' + r ab' + 2 * r ab were calculated for each cell. e represents the one-sided periclinal arrangement of chloroplasts to the adaxial side, while f represents the tendency of periclinal positioning. Differences in these indexes were statistically examined. When the results of two conditions were compared, the Mann-Whitney U test was used. For comparison among more than three conditions, the Kruskal-Wallis test and thepost hoc pair-wise comparison by the Dwass-Steel-Critchlow-Fligner (DSCF) multiple comparison method were employed. Additionally, to see the respective effect of the PFD and the CO 2 concentration, the Spearman's rank correlation was also used. The results of analyses are shown in figures. Analyses were conducted on JAMOVI (Version 2.3, the jamovi project 2022), the software for statistics based on R packages (Version 4.1, 2021).

Chloroplasts CO 2 relocation in Physcomitrium patens
We first checked whether the CO 2 -tactic movement occurred in the leaves of P. patens using the doublechamber system in the gas | gas mode (Figure 1c). One-sided CO 2 supply for 2 h in the light induced the periclinal chloroplast positioning on the CO 2 -supplied side (Figure 3a). This periclinal positioning to CO 2 -supplied side was observed irrespective of the CO 2 supply side was adaxial or abaxial. These results suggest that positive CO 2 -tactic movement occurred regardless of the cell polarity. This one-sided periclinal arrangement was observed in both blue and red light ( Figure 4). However, strong blue light prohibited the chloroplasts from taking the periclinal positions suggesting that the strong blue light induced avoidance response, which masked the CO 2 -tactic movement. In the dark, CO 2 -tactic movement was not observed and the chloroplasts remained at the anticlinal positions (Figure 4).

Effect of DCMU on chloroplast CO 2 relocation
Effects of inhibition of photosynthetic electron transport on chloroplast CO 2 relocation were examined with double-chamber system in the gas | gas mode. As shown in Figure 5, the CO 2 -tactic movement in blue light was not inhibited by DCMU, while that in red light was substantially inhibited. These indicate that the CO 2 -tactic movement occurred depending on at least two pathways, blue-light pathway that would be independent of photosynthesis and the other pathway that would at least partly depend on photosynthesis.
On the other hand, to obtain some insight into roles of ROS, we conducted a gas | gas mode experiment using O 2 -free gases containing 700 ppm and 0 ppm CO 2 and observed no inhibitory effect of the O 2 elimination on CO 2 -tactic movement (Figure 6a). We also checked effects of O 2 -containing and O 2 -free gases. There was no indication of O 2 -tactic chloroplast movements (Figure 6b).

Motility system of chloroplast CO 2 relocation
To assess involvement of the MFs and/or MTs in CO 2 -induced chloroplast relocation, we examined the effects of oryzalin and/or latrunculin A (Figure 7). The leaves showing anticlinal chloroplast arrangement in the dark were treated with the inhibitor(s) and subject to the one-sided CO 2 supply in blue light. The simultaneous treatment with oryzalin and latrunculin A prevented chloroplasts from relocating from their dark positions. Oryzalin-treated leaves showed CO 2 -dependent one-sided periclinal chloroplast arrangement, implying the contribution of MF to CO 2 relocation. The one-sidedness of periclinal arrangement of these leaves was greater than that of control leaves treated with DMSO. On the other hand, interestingly, chloroplasts in the latrunculin A-treated leaves showed the periclinal arrangement like the light-accumulation response but the one-sided movement towards the CO 2 -supplied side was absent, viz . the MT-driven movement was not responsible for CO 2 relocation in this moss.
3.4 Time course of chloroplast CO 2 relocation and effect of CO 2 concentration In the previous observations of CO 2 -tactic chloroplast movements (Senn, 1908), the duration of the one-sided CO 2 gas treatment or how much difference in CO 2 concentration needed for the one-sided CO 2 relocation 6 was not described. Here we addressed these problems. Starting from the anticlinal dark arrangement, chloroplasts moved to periclinal positions one-sidedly up to 2 h after the onset of the one-sided CO 2 supply. The longer treatment for 4 h, instead, increased the chloroplasts in the CO 2 free side (Figure 8a). When the CO 2 concentrations of two half-chambers were 700 ppm | 0 ppm or 400 ppm | 0 ppm, chloroplasts moved towards CO 2 -supplied sides. When the CO 2 concentrations were 200 ppm | 0 ppm, chloroplasts were still attracted to the CO 2 side, but the proportion of periclinal position was less biased. At the CO 2 concentrations of 400 ppm | 200 ppm and 700 ppm | 400 ppm, one-sided chloroplast relocation to high CO 2 side was observed. However, one-sidedness indexes decreased considerably (Figure 8b, see one-sidedness index e in Table 1). These observations indicate that one-sided chloroplast relocation occurred when the CO 2 concentrations between the two sides differed, and the one-sidedness would decrease with the decrease in the CO 2 concentration difference and/or the increase in the absolute CO 2 level.

Effects of the CO 2 concentration and PFD level of blue light supplied to both sides on chloroplast relocation
For the CO 2 -tactic movement to occur, it might be possible that the CO 2 elimination inhibits the chloroplasts from taking the periclinal positions, while the CO 2 supply attracts chloroplasts in the periclinal positions. Thus, we examined the effect of the same CO 2 concentration supply to both sides to check the effect of the absolute CO 2 concentration on the chloroplast positioning (Figure 9). At a given PFD level of blue light, CO 2 enhanced the periclinal positioning of chloroplasts and increased the index f , but the effect of PFD level was also marked (Table 2). At low CO 2 concentrations, there were substantial number of chloroplasts showed periclinal positionings. However, with the increase in the PFD level of the blue light such chloroplasts decreased. There was also a clear tendency that the number of chloroplasts in the central area of periclinal cell surface decreased with the increase in the PFD level and/or the decrease in the CO 2 concentration (Figure 9).

CO 2 -dependent chloroplast movement in the gas | gel mode and inhibitory effect of DCMU
The chloroplast arrangement reported for the moss leaf placed on the gel by Senn (1908) would be caused by the CO 2 gradient caused by photosynthetic CO 2 consumption by the chloroplasts. The lower diffusivity of CO 2 in the liquid phase than in the air would result in a decrease in CO 2 concentrations around the chloroplasts due to their CO 2 fixation in the light. The window of the double chamber was covered with a transparent gellan gum sheet, on which P. patens leaf samples were placed (Figure 10c). The same gas was introduced to both half-chambers. In blue light, chloroplasts showed periclinal arrangement on the air side in a CO 2 -dependent manner (Figure 10a). In the leaf samples treated with DCMU, however, one-sided chloroplast relocation was not observed (Figure 10b).

A possible explanation for light-dependent CO 2 -tactic chloroplast movement
In the present study, we revealed basic characteristics of CO 2 relocation, and the light dependence would be one of the most important ones. The first question would be how the CO 2 relocation is dependent on light conditions. It was reported that CO 2 changed the light intensity threshold between the low-and highlight responses in Mougeotia sp. (Mosebach, 1958). The idea of the threshold shift of photorelocation by CO 2 is an attractive explanation for the light dependent CO 2 -tactic movement. If the increase in the CO 2 concentration increases the threshold PFD level for light-avoidance response, the one-sided CO 2 supply induces the chloroplasts in the lower CO 2 side to show the light-avoidance response while chloroplasts in higher CO 2 side to show the light-accumulation response, which results in a CO 2 -dependent one-sided periclinal arrangement. Indeed, the threshold shift by CO 2 may explain clearly not only the present result in Figure 9 and Table 2 but also all the other data obtained in the present study. However, it is interesting to point out that, in Mougeotia sp., the CO 2 elimination increased the threshold light intensity of chloroplast orientation change (Mosebach, 1958): the threshold shift occurred in the direction contrary to the direction observed in the present study.
The experiments using DCMU inferred the presence of a photosynthesis-dependent CO 2 relocation mechanism, in addition to the photosynthesis-independent mechanism ( Figure 5). The red-light dependence of the chloroplast CO 2 relocation would be at least partly owing to photosynthesis driven by red light. Since it was known that the phytochrome-dependent chloroplast photorelocation was inhibited by far-red light and DCMU inVallisneria gigantea , an aquatic angiosperm (Dong et al., 1995), and that the red-light response in P. patens was dependent on phytochrome (Mittmann et al., 2004), it is necessary to examine involvement of phytochrome in the photosynthesis-dependent CO 2 relocation pathway in P. patens leaves.
Preceding studies on chloroplast relocation of P. patens have been carried out using the protonemata. When P. patensprotonemata were grown in weak red light, red light became effective in inducing chloroplast photorelocation, while, in white-light-grown protonemata, the red-light response was not observed (Kadota et al., 2000). In the present report, using the leaves of P. patens grown in white light, we revealed that not only blue light, but also red light induced chloroplast CO 2 relocation (Figure 4). The PFD threshold between the blue-light-accumulation and -avoidance responses in land plants ranges from 10 to 50 μmol m -2 s -1 (Zurzycki, 1962;Zurzycki, 1967;Yatsuhashi & Wada, 1990;Trojan & Gabrys, 1996), but the threshold in protonemata of P. patens was reported to be above 360 μmol m -2 s -1 (Kadota et al., 2000, Sato et al., 2001. In the present study using leaves of P. patens , we found that the threshold was around 60˜70 μmol m -2 s -1 (Figure 9). The critical difference in the present report and the previous studies remains unanswered. We also note that the light avoidance overrode the one-sided periclinal arrangement induced by CO 2 (Figure 4). We suppose this observation can be compared to the fact that the light avoidance suppressed the epistrophe in Arabidopsis thalianaleaves (Tholen et al., 2008).

Motility system of CO 2 relocation
The blue-light response of chloroplast relocation in P. patensprotonemata is known to rely on the MF and MT motility systems (Sato et al., 2001). We also observed both the oryzalin-insensitive MF-based movement and latrunculin A-insensitive MT-based chloroplast movement. More importantly, the former was responsible for the CO 2 -tactic movement, while the latter was not (Figure 7). Furthermore, the oryzalin-treated leaves showed more active CO 2 -tactic movement than the control leaves, indicating that the oryzalin-sensitive MT-system competitively suppressed the CO 2 -tactic movement (see the difference in the eindex in Figure  7). This may also be related to the smaller effect of CO 2 supplied to both sides at the same concentration than the effect of light irradiation ( Figure 9 and Table 2b). In this experiment as well, the CO 2 effect on chloroplast movement would be suppressed by the CO 2 -insensitive MT system.

Ecological aspect of CO 2 relocation
We observed the CO 2 -tactic chloroplast movement in the gas | gas mode experiments, in which two sides of the leaves were supplied with the gases of different CO 2 concentrations. In nature, the heterogeneity in CO 2 distribution within a plant cell is thought to be mainly created by the photosynthetic consumption of CO 2 by chloroplasts. In particular, it was noteworthy that the biased chloroplast arrangement observed in the gas | gel mode experiments was lost by inhibiting photosynthesis by DCMU even in blue light ( Figure  10). This observation would confirm that the heterogeneity of CO 2 concentration in the cell was caused by photosynthesis. Obeying Fick's law, CO 2 diffuses from the ambient air to the chloroplasts along the concentration gradient. Since the CO 2 diffusion coefficient in the gas phase is greater than that in the liquid phase by four orders of magnitude (Terashima, et al., 2006) and the CO 2 concentration is determined by the CO 2 supply and the CO 2 assimilation rate of each chloroplast at its position (Warren et al., 2007;Tholen & Zhu 2011), the CO 2 concentration in the cell steeply decreases near the gas-liquid boundary while it is rather homogenous in the places distant from such boundary (Ho et al., 2012). Thus, for the CO 2 -tactic movement to occur, there should be the gas-liquid boundary around the photosynthetic cells. In terrestrial bryophytes and other land plants with gaseous intercellular spaces, such interfaces commonly exist in photosynthetic tissues, while there are no such interfaces in aquatic algae. In particular, in plants like mosses, which have photosynthetic tissues exposed to the external environment, water drops can cover the cell surface and change the CO 2 supply (Williams & Flanagan, 1996). This situation would result in the conditional CO 2 heterogeneity, and chloroplasts would show the CO 2 -tactic movements towards the water-free surfaces, 8 which would be effective not only in increasing the CO 2 supply to the chloroplasts but also in avoiding photoinhibition of the chloroplasts. The cost-benefit analysis of CO 2 -relocation will require the discussion in context of the cytoskeletal energetics (Hill & Kirschner, 1982;Okamoto & Lightfoot, 1992;Leighton & Sivak, 2022) and evaluation of the cellular CO 2 conductance (Mizokami, et al., 2022).

Concluding remarks
In the present study, we observed the CO 2 -tactic chloroplast movement in P. patens moss leaves by utilizing the double side chamber with which we controlled CO 2 concentrations in the gases separately supplied to the adaxial and abaxial surfaces. The chloroplast relocation occurred either in the blue-or red-light, and not in the dark. There would be the photosynthesis-dependent and -independent CO 2 relocation mechanisms, and the photosynthesis-dependent relocation occurred only in the red light. We also identified MF system as the motility system responsible for the CO 2 relocation in the blue light, while MT system would be suppressive ( Figure 11). The photorelocation of the land plant is broadly reported to be based on MF system, while, among Streptophyta, MT-based relocation is unique to Bryophyte likeP. patens . Until now, the chloroplast CO 2 -tactic movement has only been reported for two species Bryophyta sensu strict . One is F. hygrometrica (Senn, 1908) and the second is P. patens by the present study. However, our on-going extensive examinations have already revealed CO 2 relocation in some other land plant phyla: CO 2 -tactic chloroplast movement would be a more general phenomenon than it has been thought.  Tables   Table 1. Effects of CO 2 concentration on the e andf indexes of CO 2 -tactic chloroplast movement shown in Figure 8b The Kruskal-Wallis test's p < 0.001 (N > 4 samples x 5 cells, df = 5) for both indexes, e and f . The alphabets are the same of shown in Figure 8b, which indicate the significant differences respectively for e and f .    The double-chamber system and its two experimental modes. A schematic diagram of the double-chamber system showing the gas flows in half chambers (a). Spectra of monochromatic blue and red LEDs (b). Two experimental modes: the gas | gas mode (c) and the gas | gel mode (d). In the gas | gas mode, 4˜5 leaves were attached to the rectangular windows opened in a piece of aluminum foil. Two agar gel blocks (*) placed to cover apical and basal ends of the leaves were used as water reservoirs (bar: 1 mm) (e). The gas flow diagram of the double-chamber system. The leaves placed over the windows open in an aluminum foil piece or those placed on a gel sheet placed on the foil window, was set in the double chamber. The chamber was illuminated from both side by monochromatic LED light via the optic fibers. The chamber and thus the leaves were placed vertically. The mixture of N 2 and O 2 was humidified and then mixed with 1% CO 2 in N 2 . Two gas mixtures were prepared and introduced into the respective half-chambers. The CO 2 and O 2 concentrations were measured before the experiments (a). Optical fibers hold by the brass blocks delivered light from LEDs to illuminate the double chamber from both sides (blue arrows), and gas mixtures were introduced to both half chambers (red block arrows) (bar: 5 cm) (b). In practice, a piece of tracing paper was inserted between the fiber-end and the chamber on each side. Chloroplast relocation to the high CO 2 concentration side. The cross-sectional views of one-sided chloroplast arrangement are shown (a). The gas containing 700 ppm CO 2 and 21% O 2 was supplied to the adaxial/abaxial side and that containing 0 ppm CO 2 and 21% O 2 gas was to the abaxial/adaxial side for 2 h at 50 μmol m -2 s -1 blue light irradiation (25 μmol m -2 s -1 each side) (a-1/a-2). The upper side is the adaxial surface. The bar denotes 50 μm. A cell in a single cell-layer tissue (b inset), chloroplast positions in a cell were categorized into ad, ad ', ant, ab', and ab as shown in (b). Sequential photographs were taken from the adaxial end (c-1) to the abaxial end (c-7) of a leaf treated under the same conditions as (a-1). The boundary of the focusing cell is shown in (c-8). The chloroplasts were categorized into the ad (blue circles), ad' (blue-outlined circles) (c-2), ant (open circles) (c-3, 4, 5), ab' (red-outlined circle), and ab (red circle) (c-7). Bar indicates 20 μm (c). Effects of light on chloroplast CO 2 -relocation. The proportion of the number of chloroplasts in each positional category to total number of chloroplasts was calculated for each cell in the leaves before and after the gas | gas mode experiment for 2 h. The mean proportion is shown with a bar indicating ± S.E. in a cumulative column. The CO 2 concentrations were 0 ppm (shown as -) | 700 ppm (shown as CO 2 ) and the O 2 concentration was 21%. PFD levels were 50 μmol m -2 s -1 (25 μmol m -2 s -1 from each side) for moderate-blue-light, 50 μmol m -2 s -1 (25 μmol m -2 s -1 from each side) for red-light irradiation, 120 μmol m -2 s -1 (60 μmol m -2 s -1 from each side) for high-blue-light, and 0 μmol m -2 s -1 for dark condition.
The Kruskal-Wallis test's p < 0.001 for both indexes,e and f . Different uppercase (for e ) /lowercase (for f ) letters beside cumulative bars indicate significant differences by the post-hoc DSCF pairwise comparisons, p< 0.01 or interpreted as p < 0.02 with the Bonferroni correction for concurrent analyses in e and f . Effects of DCMU treatment on CO 2 relocation. The proportion of the number of chloroplasts in each positional category after the gas | gas mode experiment for 2 h is shown as in Figure 4. Leaves were treated either with 100 μM DCMU or 1% DMSO. CO 2 concentrations were 700 ppm | 0 ppm for the adaxial and abaxial sides. O 2 was 21%. The PFD level was 40 μmol m -2 s -1 (20 μmol m -2 s -1 from each side) for both blue and red light. According to Kruskal-Wallis test, p< 0.001 for the differences in e index. Different lowercase letters beside the cumulative bars indicate the significant differences in e by the post-hoc DSCF comparisons (p < 0.01). Effect of O 2 on chloroplast movement. The proportion of the number of chloroplasts in each positional category after the gas | gas mode experiment for 2 h was shown (a) as in Figure 4. The CO 2 concentrations were 0 ppm (shown as -) | 700 ppm (shown as CO 2 ). Leaves were illuminated with 40 μmol m -2 s -1 (20 μmol m -2 s -1 from each side) blue light. The difference in index e was significant (p < 0.001, Mann-Whitney U test) (a). In (b), the O 2 was supplied one-sidedly for 2 h. The O 2 concentrations were 0% (shown in figure as -) | 21% (shown in figure as O 2 ). The CO 2 concentration was 0 ppm. Samples were illuminated with blue light at 40 μmol m -2 s -1 (20 μmol m -2 s -1 from each side). The difference in index e between the two O 2 treatment was not significant (p = 0.957, Mann-Whitney U test). Effect of inhibition of cytoskeleton systems on CO 2 relocation. The proportion of the number of chloroplasts in each positional category after the gas | gas mode experiment for 2 h was shown as in Figure 4. The concentrations of the cytoskeleton inhibitors were 50 μM for oryzalin and 10 μM for latrunculin A. The CO 2 concentrations were 0 ppm (shown in figure as -) | 700 ppm (shown in figure as CO 2 ). O 2 was 21%. Leaves were illuminated with blue-light at 40 μmol m -2 s -1 (20 μmol m -2 s -1 from each side). Differences in e and f were both statistically significant respectively (p < 0.001, Kruskal-Wallis test). The different alphabets besides the columns indicate significant differences as in Figure 4. The photographs of the adaxial surfaces of leaves treated with 700 ppm CO 2 on adaxial side and 0 ppm CO 2 on abaxial side are shown (b). Bar denotes 20 μm. The time course of CO 2 relocation (a) and the effect of CO 2 concentration (b). The proportion of the number of chloroplasts in each positional category after the gas | gas mode experiment for 1 to 4 h was shown (a) as in Figure 4. The CO 2 concentrations were 700 ppm | 0 ppm for abaxial side. O 2 was 21%. Samples were illuminated with 60 μmol m -2 s -1 (30 μmol m -2 s -1 from each side) blue light (a). In (b), the effect of CO 2 concentration for gas | gas experiment was closely inspected. The respective CO 2 concentration were shown in the panel. O 2 was 21%. Samples were illuminated with 50 μmol m -2 s -1 (25 μmol m -2 s -1 from each side) blue light for 2 h (b).
Respectively, in both (a) and (b), the differences in indexes, eand f were statistically significant (p < 0.001, Kruskal-Wallis test). The different alphabet letters besides the columns indicate significant differences as in Figure 4. For values of eand f indexes in (b), see Table 1. Surface views of P. patens leaves after 1 h gas | gas mode experiments treated with the same concentration of CO 2 on both sides of the leaves. The photographs of the adaxial surfaces are shown with the CO 2 concentrations and the light intensities. O 2 was 21%. A bar denotes 50 μm. Different alphabet letters at the left bottom of photos indicate the significant differences in f shown in Table 2. Effects of the CO 2 concentrations and DCMU on the gas | gel mode experiments. The proportion of the number of chloroplasts in each positional category after the gas | gel mode experiments for 2 h was shown as in Figure 4. O 2 was 21%. Samples were illuminated with blue light at 50 μmol m -2 s -1 (25 μmol m -2 s -1 from each side) for 2 h. In (a) the biased chloroplast arrangements induced by the gas | gel experiment are shown. The same concentration of CO 2 was supplied to both the samples-attached surface and the opposite surface of the gel sheet. The respective CO 2 concentrations are shown in figure. The gel-attached surface of the samples for each experiment is represented as gel in panel. The differences in e values were statistically significant (p< 0.001, Kruskal-Wallis test). Different alphabet letters beside columns indicate significant differences as in Figure 5. (a). The gas | gel experiment was conducted with 1% DMSO or DCMU treated samples (b). The CO 2 was 700 ppm. The abaxial surface was attached to the gel sheet. Asterisks indicate the significant difference in index e of the two experiments atp < 0.001 (Mann-Whitney U test) (b). (c) shows the four leaves placed on a gellan gum gel sheet and sealed on a window of aluminum foil. The bar denotes 5 mm. A model for CO 2 effect on chloroplast relocation in relation to the light dependency and the motility system in P. patens leaf lamina cells revealed in this paper. Effects of CO 2 concentration on the e and f indexes of CO 2 -tactic chloroplast movement shown in Figure 8b The Kruskal-Wallis test's p < 0.001 (N > 4 samples × 5 cells, df = 5) for both indexes, e and f. The alphabets are the same of shown in Figure 8b, which indicate the significant differences respectively for e and f .