Differences in Aquaporin Levels Among Cell Types of Radish and Measurement of Osmotic Water Permeability of Individual Protoplasts

Home , Radish

Shinobu Suga 1, 3, Mari Murai 2, 3, Tsuneo Kuwagata 2, 4 and Masayoshi Maeshima 1, 5 1 Laboratory of Cell Dynamics, Graduate School of Bioagricultural Sciences, Nagoya University, Nagoya, 464-8601 Japan 2 Tohoku National Agricultural Research Center, Morioka, 020-0198 Japan

; We investigated tissue- and cell-specific accumulation molecules to cells. Also, CO2 and carbonic ions dissolved in of radish aquaporin isoforms by immunocytochemical anal- water need to be removed from the non-photosynthetic tissues. ysis. In taproots, the plasma membrane aquaporins The water stream depends on the influx and efflux of water to (RsPIP1 and RsPIP2) were accumulated at high levels in the cells through plasma membranes (PM). the cambium, while the tonoplast aquaporin (RsTIP) was Aquaporins facilitate the osmotic water transport across distributed in all tissues. The three isoforms were highly PM and tonoplast (Maurel 1997, Chrispeels et al. 1999, Johans- accumulated in the central cylinder of seedling roots and son et al. 2000, Tyerman et al. 2002). Although water mole- hypocotyls, and rich in the vascular tissue of the petiole of cules can be transported across membranes by simple diffusion mature plants. The results suggest that RsPIP1 and RsPIP2 mechanisms and through ion channels, its rate is limited are abundant in the cells surrounding the sieve tube of the (Haines 1994). This means that water transport, both intercellu- radish plant. The swelling rate of protoplasts in a hypot- lar and intracellular, depends on the activity and quantity of onic solution was determined individually by a newly estab- aquaporins. The distribution of aquaporis in organs and the lished method to compare the osmotic water permeability water channel activity may be one of the factors determining of different cell types. All cells of the cortex and endodermis the pathway and capacity of water flow in plant bodies. in seedlings showed a high water permeability of more than The structural diversity, intracellular localization and –1 300 m s . There was no marked difference in the values physiological responses of plant aquaporins have been investi- between the root endodermis and cortex protoplasts, gated in detail (Jauh et al. 1999, Jiang et al. 2001, Katsuhara et although the RsPIP level was lower in the cortex than in the al. 2002, Morillon et al. 2001, Santoni et al. 2000). Remark- endodermis. This inconsistency suggests two possibilities: able progress has been made into the research on higher-order (1) a low level of aquaporin is enough for high water per- structures of aquaporins (Daniels et al. 1999, de Groot and meability and (2) the water channel activity of aquaporin in Grubmuller 2001, Sui et al. 2001). Furthermore, tissue-specific the tissues is regulated individually. The uneven distribu- gene expression has been demonstrated for the tonoplast aqua- tion of aquaporins in tissues is discussed along with their porins (Barrieu et al. 1998a, Barrieu et al. 1998b, Chaumont et physiological roles. al. 1998, Karlsson et al. 2000, Sarda et al. 1997) and PM aquaporins (Dixit et al. 2001, Otto and Kaldenhoff 2000, Keywords: Aquaporin — Cell specificity — Osmotic water Yamada et al. 1995, Yamada et al. 1997) by the promoter assay, permeability — Plasma membrane — Protoplast assay — in situ hybridization and Northern hybridization. Recently, Radish. quantification of aquaporin proteins has been tried in different cell types of Mimosa pudica (Fleurat-Lessard et al. 1997), soy- Abbreviations: DTT, dithiothreitol; PBS, phosphate-buffered bean nodule (Serraj et al. 1998), Brassica oleracea (Marin- saline; PM, plasma membrane; Pos, osmotic water permeability; PIP, Oliver et al. 2000), Mesembryanthemum crystallinum (Kirch et plasma membrane intrinsic protein; RsPIP, radish plasma membrane aquaporin; RsTIP, radish vacuolar membrane aquaporin; TIP, tono- al. 2000), Nicotiana tabacum (Otto and Kaldenhoff 2000) and plast intrinsic protein. Brassica napus (Frangne et al. 2001). The present study was conducted to investigate the distribution of PM aquaporins, RsPIP1 and RsPIP2 (previous names PAQ1 and PAQ2, respectively) and tonoplast aquaporins, Introduction RsTIP (previous name VM23) among radish tissues and cells. Their mRNA and protein levels have already been investigated Plants retain a large quantity of water in their cells for in several organs at different growth stages by Northern hybrid- metabolic activities and turgor pressure, whilst retaining water ization and immunoblot analysis (Suga et al. 2001, Suga et al. flow in their bodies. Water stream in the bodies is essential for 2002). The protein levels of RsPIP1 and RsTIP on a mem- the supply of nutrients, oxygen, hormones and other signal brane protein basis were relatively constant, whereas the

3 These authors contributed equally to the paper. 4 Present address: National Institute for Agro-environmental Sciences, Tsukuba, 305-8604 Japan. 5 Corresponding author: E-mail, [email protected]; Fax, +81-52-789-4096.

277 278 Aquaporin levels and Pos of radish protoplasts

the taproots and petioles of 1-month-old plants using tissue print immunoblotting. The anti-RsPIP1 and anti-RsPIP2 antibodies recognize the subfamily of RsPIP1 (RsPIP1-1, 1-2, 1-3) and preferentially the RsPIP2-1 isoform, respectively (Suga et al. 2001). It was confirmed that the anti-RsTIP (previous name, anti-VM23) reacts mainly with the RsTIP1 (-VM23) protein using the recombinant RsTIP1 and RsTIP2 proteins (data not shown). As shown in Fig. 1, the distribution of RsTIP was different from that of RsPIPs. The vascular cambium of taproots was stained densely with anti-RsPIP1 and anti-RsPIP2, but not with anti-RsTIP. The rim of the cross-section of the young taproots, which in part are composed of the periderm and sub- periderm (inner cell layer), was clearly stained with the anti- RsTIP (Fig. 1). The region including the xylem was clearly stained with all three antibodies. The tissue print immunoblot for the petiole indicated that the levels of RsPIP1, RsPIP2 and RsTIP were the highest, at least in the vascular bundles, compared with other tissues. These observations were confirmed by the immunocytochemical analysis (Fig. 2). For the analysis we used a thin, lower part of the taproot from a 1-month-old radish to obtain small Fig. 1 Tissue print immunoblot analysis with antibodies to RsPIPs tissues, although large upper parts of the taproot were used for and RsTIP. Discs of radish taproots (0.5–1.5 cm in diameter; a, b, d, e, the tissue print immunoblotting as shown in Fig. 1. The phloem g, h, j) and petioles (c, f, i, k) were prepared from 1-month-old plants and the vascular cambium of the taproot were clearly stained and used for immunoblotting on a polyvinylidene difluoride mem- brown with anti-RsPIP1 (Fig. 2a) and anti-RsPIP2 (Fig. 2d). brane with anti-RsPIP1 (a–c), anti-RsPIP2 (d–f), anti-RsTIP (g–i) and The xylem parenchyma and the central part were also stained the control antibody (j, k). densely. RsTIP was detected in all tissues (Fig. 2g). In the cross-sections of petioles, three major and several minor vascu- RsPIP2 level varied with tissue and stage. However, immunob- lar bundles were observed. RsPIPs and RsTIP were markedly lots of the membrane preparations provide information on the accumulated in these vascular tissues (Fig. 2b, e, h). Further- average content of aquaporins in tissues composed of several more, the periderm and sub-periderm of taproots and petioles cell types. To analyze the distribution of aquaporins, we were stained densely with anti-RsPIP1 and anti-RsTIP. applied tissue print immunoblotting and immunocytochemical methods. We found characteristic distributions of aquaporin Accumulation of aquaporins in the central cylinder of seedlings proteins in different tissues. Fig. 3 shows diagrams of the cross-sections from seedling We developed a new technique to determine the osmotic roots and young taproots. Active cell proliferation occurs in the water permeability (Pos) of protoplasts. There are several meth- cambium of the central cylinder of seedling roots and hypocot- ods to measure Pos of living cells, such as the plasmometric yls, with the central cylinder growing into a taproot. On the method and the pressure probe technique (Maurel 1997). These other hand, the cortex and epidermis of roots and hypocotyls methods are technically difficult to apply to inner or small cells peel away during the growth of the taproots. of tissues. Protoplasts were assayed for the water permeability Fig. 4 shows the distribution patterns of RsPIP1, RsPIP2 in some plants (Maurel 1997, Moshelion et al. 2002). Rama- and RsTIP in seedling roots and hypocotyls. A number of root haleo et al. (1999) reported a new technique using protoplasts hairs distinguish roots from hypocotyls. The central cylinders in a transference chamber. We further improved accuracy and in both organs were stained dark brown. The levels of RsPIP1, simplified the method to measure Pos of protoplasts that were RsPIP2 and RsTIP were high in the cambium of the central cyl- isolated directly from target tissues. We determined Pos using inder and the endodermis of seedling roots (Fig. 4a–f). It is our refined method to compare cells either rich or poor in clear in an enlarged photograph of the central cylinder (Fig. 4i). aquaporins. Here we discuss the relationship between the Pos On the other hand, aquaporin levels were found to be lower in values and the level of aquaporin proteins. the pericycle or cells around the metaxylem. The hypocotyls and seedling roots showed a similar pattern. It should be noted Results that the levels of three aquaporins were higher in the root endodermis than in the hypocotyl (Fig. 4a, c, e). The other feature of Levels of aquaporin proteins in specific cells of young taproots seedling roots is that the epidermis contained relatively high To analyze the distribution of aquaporins, we examined levels of RsTIP (Fig. 4e). Aquaporin levels and Pos of radish protoplasts 279

Fig. 2 High level of RsPIPs and RsTIP proteins in specific cells of radish taproot and petiole. Thin cross-sections of 1-month-old taproots (3– 4 mm in diameter) and petioles were immunostained with anti-RsPIP1 (a, b, c), anti-RsPIP2 (d, e, f), anti-RsTIP (g, h, i) and control antibody (j, k, l). High magnifications of vascular bundles of petioles (b, e, h, k) are shown in the right panels (c, f, i, l). Roman numerals in panel (a) indicate phloem (1), vascular cambium (2), secondary xylem (3) and the central part (4).

New assay method for Pos of protoplasts with paraffin oil. A target protoplast in the droplet was selected The high levels of aquaporins in specific cells raise a under a microscope, placed into a 0.3-M sorbitol droplet using question: Does the osmotic water permeability of the cells a micromanipulator, and monitored under an inverted micro- reflect the aquaporin level? To answer this question, we deter- scope (Fig. 6B, C). The protoplasts were stable for 2 h in mined Pos of protoplasts individually. Fig. 5 shows typical pro- 0.45 M sorbitol without any additional reagents such as Ca2+. toplasts from the cortex and endodermis of seedling roots. The This is a simple method to determine the swelling rate of the radius of protoplasts isolated from the cortex and endodermis protoplast of interest. were 23–45 m and 18–25 m. Time-course changes of the volume of protoplasts pre- We developed a procedure for the selective isolation of a pared from the cortex and endodermis were compared (Fig. 7). target protoplast and a new assay method for Pos of protoplasts. Discs of roots and hypocotyls were pretreated with a Cells from the same tissue were reproducibly swollen in the maceration medium, and were rinsed in 0.45 M sorbitol (Fig. hypotonic medium at the same rate. The volume gradually 6A). We then selected a particular part in the disc and isolated increased to 150% in the hypotonic 0.30 M sorbitol solution it using a capillary. The small cell cluster was subsequently dis- within 30 s. The steady-state volume in the hypotonic medium persed and washed in 0.45 M sorbitol. A few protoplasts were is comparable to the theoretical value 150% calculated from transferred to a 0.45-M sorbitol droplet in a culture dish filled the osmolarities of 0.45 and 0.30 M sorbitol solutions. 280 Aquaporin levels and Pos of radish protoplasts

containing thicker cell walls proved difficult for protoplast preparation. The method for protoplast preparation and isolation of these special tissues remains to be improved.

Discussion

Aquaporins are highly accumulated in the central cylinder and vascular tissue In the present study, we applied methods of tissue print immunoblotting and immunocytochemistry to examine the distribution of RsPIPs and RsTIP proteins. Immunological quantification provided essential information on the uneven distribution of aquaporins in tissues. The first point is that RsPIPs are highly accumulated in the cambium of taproots, the vascular tissue of petioles, and the central cylinder. The RsPIP level in the cambium was estimated to be several times higher than that of the pericycle and the inner parenchyma cells. High signal intensity in the vascular cambium and the central cylinder was not found to be due to differences in cell size. The levels of RsPIP1 and RsPIP2, but not RsTIP, were high in the vascular cambium of the young taproot (Fig. 2). The cell-specific immu- nostaining was also clearly observed in the central cylinder and endodermis (Fig. 4i). We thought that a high amount of RsPIPs in the cambium supports the rapid flow of water from the phloems, which are located in the central cylinder, toward the pericycle. The high levels of RsPIPs may facilitate water circulation between the Fig. 3 Schematic diagrams of tissues of seedling root and growing xylem and phloem in both the taproot and central cylinder. PM young taproot of radish. The inside part within the endodermis (cen- aquaporins are essential to facilitate the intercellular water flow tral cylinder) grows into a taproot. from the cells surrounding the fine-structured sieve tubes to the relatively large parenchyma cells. Rapid cell growth in this region requires a prompt water supply. The high levels of High Pos values of radish protoplasts RsPIPs in the cambium seemed to be regulated at the transcrip- The Pos values were calculated from the rates of volume tional step since in situ hybridization revealed a high steady- increase under consideration of changes in the cell volume and state mRNA level of RsPIP1 and RsPIP2 in the root cambium the internal solute concentration as described in Materials and (data not shown). The active transcription and translation of Methods. Fig. 8 shows the three theoretical lines of cell vol- RsPIPs may supply a large amount of aquaporin proteins to ume increase, which were calculated for a protoplast with a growing cambium cells. radius of 26.3 m under the assumption that Pos is 180, 280 and The second point is that RsPIP1, RsPIP2 and RsTIP 380 m s–1. The experimental results, shown as open circles, were highly accumulated in the central cylinder, especially in fitted well with the theoretical line of 280 m s–1. This value the cambium and the phloem (Fig. 4). There was a marked was determined as Pos of the protoplast. The Pos values of 20 difference in stained intensity between the cambium and the protoplasts from the endodermis and cortex of the roots metaxylem-surrounding cells, although they were similar in were calculated to be 323M96 and 365M90 m s–1, respectively cell size. The cambium in the central cylinder is active in cell (Fig. 9). The Pos values of the endodermis and cortex cells gave proliferation and growth. The xylem and phloem in the central a single peak on this histogram. A mathematical t-test showed cylinder are the main waterways. A dense distribution of RsPIPs no difference between the Pos values of the endodermis and in the central cylinder may provide rapid dispersion of water cortex cells (p = 0.1721, df = 38). from the phloem to the cortex. It should be noted that the endo- This novel method for determining Pos can be applied to dermis of seedling roots, but not of hypocotyls, were strongly small protoplasts. We tried to determine Pos of cambium cells stained with antibodies to RsTIP, RsPIP1 and RsPIP2. This in the central cylinder but it was hard to select individual proto- may be related to the presence (roots) or absence (hypocotyls) plasts among several cell types, because the central cylinder of the Casparian strip that intercepts the apoplastic water stream. consists of the pericycle, cambium, metaxylem, protoxylem The endodermis of seedling roots contains the Casparian strip, and xylem parenchyma cells (Fig. 3). Also, epidermis cells in which the membrane of the large central vacuoles is a bar- Aquaporin levels and Pos of radish protoplasts 281

Fig. 4 Immunolocalization of RsPIPs and RsTIP in cross-sections of seedling. Cross-sections through the root (a, c, e, g, i) and hypocotyl (b, d, f, h) of 5-day-old radish seedlings were immunostained with anti-RsPIP1 (a, b), anti-RsPIP2 (c, d), anti-RsTIP (e, f) and control antibodies (g, h). Hairy roots can be observed in root sections (a, c, e, g). The central cylinder of root immunostained with anti-RsPIP2 (panel c) is enlarged in panel (i). rier to water transport. This is because water molecules cross the sure. These growing sub-peridermal cells may be required for cells through plasmodesmata, as discussed for the basal endo- quick water absorption for their growth. Rapid cell growth sperm transfer cells (Barrieu et al. 1998b). Thus, the high level occurs by rapid expansion of the vacuole. A large quantity of of RsTIP provides endodermis cells for rapid transport of water. RsTIP may support a quick water transport across the tono- The abundance in the periderm and sub-periderm is a plast in these cells. characteristic of RsTIP. It is clear that the sub-peridermal cells The results are consistent in part with the previous results may directly sustain the epidermis through high turgor pres- of Mesembryanthemum crystallinum PM aquaporins (Yamada 282 Aquaporin levels and Pos of radish protoplasts

plasts could be selected from the protoplast suspension using microscopy.

The present assay method for Pos has some advantages as compared with other methods. Recently, Ramahaleo et al.

(1999) developed a new method for the Pos measurement of an isolated protoplast that was held by a micropipette in a transfer-

ence chamber. The Pos values were calculated from the initial swelling rate during the first 20% increase in the hypotonic medium. They discussed the systematic errors of underestima-

Fig. 5 A cross-section (A) of young root which was treated with the maceration solution and protoplasts of the cortex (B) and the endodermis (C) of the root of 4-day-old radish seedling. et al. 1995, Yamada et al. 1997). MipA was highly expressed in the developing root xylem cells and the leaf vascular tissue, whereas MipB was expressed in the central cylinder. The MipB promoter has been reported to be active in cells surrounding vascular tissues as well as xylem parenchyma cells and the sub- epidermal cells. Barrieu et al. (1998b) reported the high expression of ZmTIP1 in the subepidermis and endodermis of young roots of maize.

Improvement of selection of protoplast and Pos assay method We developed a method for selection of specific protoplasts and measurement of Pos. A target portion was selected and isolated using a capillary from a tissue segment that was digested in the maceration medium. Thereafter proper proto-

Fig. 6 Protoplast assay system for osmotic water permeability. (A) A schematic diagram of the procedure. Tissue discs were prepared form young roots and hypocotyls of 4-day-old seedlings. The disc was treated with a maceration medium, and then rinsed to remove enzymes. A cluster of protoplasts was picked out from the disc and suspended in 0.45 M sorbitol and 1 mM DTT. The protoplasts obtained were placed in a droplet of 0.45 M sorbitol in a plastic culture dish filled with paraffin oil. A target protoplast in the droplet was picked using a capillary. The capillary was immersed into a droplet of 0.30 M sorbitol and the protoplast was released under an inverted microscope. The swelling of the protoplast in the hypotonic solution was monitored by the microscope and recorded through a CCD video camera. (B) Photograph of a culture dish containing several droplets in paraffin oil. The dish was set on the stage of the inverted microscope. (C) Overall view of apparatus used for the protoplast assay. A micromanipulator, cell injector and CCD video camera were attached to a microscope. Aquaporin levels and Pos of radish protoplasts 283

Fig. 8 Matching of experimental data with a theoretical curve for the time course of changes in protoplast volume. The three theoretical lines were calculated for a protoplast with a radius of 26.3 m under the assumption that Pos is 180 (lower line), 280 (middle line) and 380 (upper line) m s–1 when the protoplast is transferred from 0.45 to 0.30 M sorbitol. The mathematical equation is described in Materials and Methods. Open circles show the experimental results for the protoplast with a radius of 26.3 m after transfer from 0.45 to 0.30 M sorbitol. In this case, the experimental result was matched with the line of –1 280 m s .

Fig. 7 Swelling rates of protoplasts in the hypotonic solution. Proto- plasts were prepared from the hypocotyl cortex (A), the root cortex (B) and the root endodermis (C) of 4-day-old seedlings and assayed for Pos in the hypotonic 0.30-M sorbitol solution. The protoplast volumes (V) calculated from the diameter are shown as the ratio to the original volume (V0) in the 0.45-M sorbitol solution. The traces of typical individual cells are shown separately (open circles, filled circles and squares).

tion of Pos. The underestimation comes from the assumption that the cell surface and the solute concentration gradient were Fig. 9 Osmotic water permeabilities of radish protoplasts. The Pos values of 20 protoplasts prepared from the root endodermis and cortex kept at their initial values. Actually, the protoplast surface of 4-day-old seedlings were calculated from the swelling rates of the increases and the solute concentration in protoplast decreases protoplasts. The number of protoplasts with the range of the indicated during swelling. To avoid underestimation, we considered the Pos is shown for the endodermis cells (black bars) and the cortex cells changes as shown in equations (4) and (11). When the P val- (stippled bars). The mean values of the endodermis and cortex cells os M M M –1 ues were determined from the data in Fig. 8, using the method were 323 96 (average SD, n = 20) and 365 90 m s (n = 20), respectively. Protoplasts from the endodermis and cortex used for the of Ramahaleo et al. (1999), the value of a protoplast with a P assay were 20–25 m and 30–45 m in radius, respectively. The t- –1 os radius of 26 m was calculated to be 202 m s (25% increase test revealed no significant difference in the Pos values between the in the volume in the first 4 s), which was smaller than the root endodermis and cortex cells (p = 0.1721, df = 38). –1 present result (280 m s ). The experimental results of Pos fitted well with the theoretical curve for the whole assay period. diameter of 25 m was used to trap a protoplast (Ramahaleo et Thus, we concluded that this curve-fitting method is appropri- al. 1999). In the case of small protoplasts but not of large pro- ate for the total process of protoplast swelling. toplasts, the surface area that is attached with a pipette is not The improved method has other advantages. Calculation negligible. Furthermore, the surface area may change during of volume changes of protoplasts is easy using the present swelling and shrinking of the protoplast. Therefore the investi- method, since a free protoplast is used. A glass pipette with a gator should consider this factor. The present method does not 284 Aquaporin levels and Pos of radish protoplasts require a complex transference chamber for the protoplast ulation of each aquaporin isoform should provide novel infor- assay and is applicable to most laboratories. mation on the water stream in plants.

Cortex and endodermis protoplasts have high osmotic water Materials and Methods permeability The protoplasts from the cortex and endodermis gave Plant material –1 significantly high Pos of more than 300 m s . Such high Pos Radish plants (Raphanus sativus L. cv. Tokinashi-daikon) were strongly supports the aquaporin-mediated water transport across grown on the Nagoya University farm for 1 month. Radish seedlings were soaked in water for 1 day and germinated on a net floating on PM. The values are comparable to those of whole red blood 0.3 mM CaSO for 3 or 4 d at 26LC under long-day conditions (light/ –1 4 cells, which exhibit high Pos among living cells (530 m s ) dark regime of 16 h/8 h, cool-white lamp). (Tsai et al. 1991). Typical values for membranes with aquaporins are 100–200 m s–1 (Zeidel et al. 1994). Ramahaleo et Antibody preparation, tissue print immunoblot and immunocyto- al. (1999) reported permeabilities of approximately 10, 400 and chemistry 330 m s–1 for protoplasts of onion leaf, rape hypocotyl and An antibody to RsTIP (anti-RsTIP) (Maeshima 1992) and anti- peptide antibodies to RsPIP1 (anti-RsPIP1) and RsPIP2 (anti-RsPIP2) rape root. The Pos of protoplasts from Samanea motor cells has (Suga et al. 2001) were prepared in advance. The reactivity of antibod- been reported to be regulated diurnally with a range from 3.1 to ies to the aquaporin isoforms was described previously (Suga et al. –1 5.0 m s (Moshelion et al. 2002). Simple diffusion of water 2002). The anti-RsPIP2 antibody is specific to the RsPIP2-1 isoform. molecules across lipid bilayers gave lower values less than The anti-RsPIP1 antibody equally reacts with RsPIP1-1, RsPIP1-2 and 50 ms–1 (Haines 1994), and was dependent on the phase tran- RsPIP1-3. The anti-RsTIP antibody specifically reacts with RsTIP1 but not with RsTIP2. Tissue print immunoblot analysis was carried out sition temperature of the membrane. Plants need to maintain on a polyvinylidene difluoride membrane (Millipore Co., Bedford, MA, the water flow in their organs, especially in the cambium and U.S.A.) to investigate the distribution of aquaporins following a stand- cells surrounding the sieve tube, even at low temperatures. ard procedure (Cassab and Varner 1987). Thus, the PM and tonoplast must contain a large quantity of Immunocytochemical analysis was carried out after fixation of aquaporins in plants. tissues with glutalaldehyde. Small pieces cut from the roots, hypocotyls, taproots and petioles were fixed with 2% glutalaldehyde in phos- In the radish, the tonoplast may not limit Pos of proto- phate-buffered saline (PBS) for 4 h at 4LC. The tissue pieces were plasts, since Pos of tonoplast, determined by the stopped-flow embedded in 5% agar, and thereafter thin sections of them were pre- spectrophotometrical method, was significantly higher than pared. The thickness was 90 m for seedling roots, 110 m for seed- that of PM (Ohshima et al. 2001). Thus, the Pos values of ling hypocotyls, 75 m for taproots and 200 m for petioles. The sec- protoplasts may be closely related to the water channel tions were dehydrated in a series of ethanol (30, 50, 70, 90 and 100% in sequence) and rehydrated by exchanging the ethanol solution (100, activity through RsPIPs. From this point of view, there are two 90, 70, 50 and 30%). The sections were treated with 1% defatted milk, explanations why Pos do not reflect the aquaporin level. First, 0.05% Tween-20 and 0.1% NaN3 in PBS and then treated with the pri- quick water transport in protoplasts occurs even at low RsPIP mary antibody. After rinsing with 0.5% defatted milk, 0.05% Tween- levels in the cortex. Indeed, red blood cells show high Pos 20 in PBS, the blotted membranes were treated with horseradish per- (530 m s–1) (Tsai et al. 1991) even though the red cell mem- oxidase-linked protein A (Amersham Biosciences Co., Piscataway, NJ, U.S.A.). The blots were visualized in 0.06% (w/v) 3,3>-diaminobenzi- branes of humans and rats contain only 2.4% and 3.7%, respec- dine tetrahydrochloride, 0.03% (w/v) CoCl2 and 0.03% (v/v) H2O2 and tively (Nielsen et al. 1993). In the radish, RsPIPs and RsTIP then photographed. occupy 10% and 40% of the total protein in PM and tonoplast, respectively (Higuchi et al. 1998, Maeshima 2001, Ohshima et Protoplast preparation and water permeability measurement of proto- al. 2001). Therefore, the RsPIP levels in radish are estimated to plasts be at least a few percentages of the membrane proteins. One Tissue segments (200–300 m in thickness) cut from roots (upper portion) and hypocotyls (central portion) from 4-day-old seedlings can estimate that the low content is enough for expression of were incubated in 2 ml of a maceration solution. The maceration solu- high Pos. tion contained 2% (w/v) Cellulase Onozuka RS (Yakult Pharmaceuti- Secondly, the PM aquaporins are functionally regulated. cal Ind. Co., Tokyo, Japan), 0.1% (w/v) Macerozyme R-10 (Yakult There is a possibility that the water channel activity of RsPIPs Pharmaceutical Ind. Co., Tokyo, Japan), 0.03% (w/v) Pectolyase Y23 in the endodermis is suppressed under the assay conditions. (Kikkoman Co., Noda, Japan), 20 mM MES-KOH, pH 5.5, 2 mM dithiothreitol (DTT) and 0.39 M sorbitol at 25LC for 20 to 60 min. The Several possible factors may regulate the function of aquapor- cell wall was partially hydrolyzed by this treatment. 2+ ins; namely, the cytoplasmic Ca concentration, phytohor- These discs were then carefully rinsed in 2 ml of 0.45 M sorbitol mones, protein kinase, protein phosphatase, the external and 1 mM DTT to remove the maceration enzymes. DTT was added to osmotic pressure and light (Chrispeels et al. 1999, Johansson et prevent oxidation of cell components and the pH of the sorbitol was about 5.8. The target parts such as the cortex were sucked up with a al. 2000, Maurel 1997). The Pos of protoplasts isolated from small capillary (100–150 m in internal diameter). For the endoder- different tissues under physiological conditions should be mis, the central cylinder containing endodermis cells was taken from determined in consideration of these factors. Further quantita- the disc using a relatively large capillary (400 m in internal diame- tive studies of the water channel activity and its functional reg- ter), thereafter the endodermis cells lined around the central cylinder Aquaporin levels and Pos of radish protoplasts 285

3 3 3 were picked and chosen using a small capillary (100 m) under micro- [r* /(r* + b ) ] P (dr*/dt) = –CoPosVw/r0,(8) scopy. A few protoplasts in the rinse medium were transferred to a droplet (30 l) of 0.45 M sorbitol in a culture dish (Falcon) filled with where: paraffin oil (see Fig. 6). The droplets of sorbitol solution were stable on the bottom of the dish. The temperature in the paraffin oil was mon- r* V (r/r0), (9) 1/3 itored with a thermometer. A target protoplast in the droplet was b V (– Co0/Co) .(10) picked up using microscopy with a micro-capillary (100 m in internal diameter) pre-filled with paraffin oil. The tip of the capillary was After integrating equation (8) under the initial condition (r = r0 for t = 0) moved into a droplet (100 l) of hypotonic solution of 0.30 M sorbitol we obtain in the same dish and the microscope was focused on the tip of the cap- b $4 3 3 illary. The protoplast was then released into the droplet from the capil- r – 1 – ----%5ln ------r + b – ln ------1 + b 6 3 3 3 lary. The micro-capillary attached with a cell injector was handled &6r+ b 1 + b with a micromanipulator (see Fig. 6C). When a protoplast was released b $4–1.2r – b –1.2 – b .CoPosVw into a droplet of the hypotonic solution, a small volume of 0.45-M –------%5tan ------0– tan ------0= – ------0t (11) sorbitol solution also entered into the droplet. The volume was negligi- 3&63b 3b r0 ble, however, being only 1–10 nl. All sorbitol solutions contained 1 mM DTT to protect membrane components from oxidation during When the values of r0, Co0 and Co have been determined, we can calcu- the experiment. late the theoretical time course of changes in V for any value of Pos Protoplasts were viewed using an inverted microscope (Nikon, from equation (11). The value of Pos can then be determined from the Diaphoto TMD, Tokyo, Japan) equipped with a differential interfer- fitted curve (Fig. 8). ence contrast (DIC) system and recorded by a charge-coupled device (CCD) camera (Sony, DXC-107A, Tokyo, Japan). The assay was car- Acknowledgments ried out at 25LC and the diameter and volume of the protoplasts were calculated from the images on the video monitor using a NIH/Image We are grateful to Dr. Yasuo Shioya of National Institute of Live- program. stock and Grassland Science for his technical assistance in establish- ing our method of water permeability measurement. This work was Calculation of osmotic water permeability supported by Grants from the Japanese Ministry of Education, Sci- –1 Osmotic water permeability Pos (m s ) of protoplasts was deter- ence, Sports and Culture for Scientific Research (nos. 13640645 and mined by the swelling rate of protoplasts after exposure to hypotonic 10219203) and COE Research to M. Maeshima. S.S. is a recipient of conditions. The rate of change in protoplast volume V is given by Japan Society for the Promotion of Science for Young Scientists Research Fellowship 14084401. dV/dt = Pos SVw(C – Co)(1) References where t is time (s), S is the surface area of the protoplast (m2), C and Co are the solute (osmotically active particles) concentration inside and –3 Barrieu, F., Thomas, D., Marty-Mazars, D., Charbonnier, M. and Marty, F. outside the protoplast, respectively (mol m ). P is assumed to be os (1998a) Tonoplast intrinsic proteins from cauliflower (Brassica oleracea L. independent of changes in protoplast surface area. Vw is the partial –6 3 –1 var. botrytis): immunological analysis, cDNA cloning and evidence for molar volume of water (18P10 m mol ). If the protoplast is assumed expression in meristematic tissues. Planta 204: 335–344. to be a perfect sphere, S and V are written as Barrieu, F., Chaumont, F. and Chrispeels, M.J. (1998b) High expression of the tonoplast aquaporin ZmTIP1 in epidermal and conducting tissues of maize. S = 4 r2,(2)Plant Physiol. 117: 1153–1163. V = 4/3 r3,(3)Cassab, G.I. and Varner, J.E. (1987) Immunocytolocalization of extensin in developing soybean seed coats by immunogold-silver staining and by tissue printing on nitrocellulose paper. J. Cell Biol. 105: 2581–2588. where r is the radius of the protoplast, and the circular constant. Chaumont, F., Barrieu, F., Herman, E.M. and Chrispeels, M.J. (1998) Character- Assuming that C decreases inversely proportional with increasing vol- ization of a maize tonoplast aquaporin expressed in zones of cell division and ume of the protoplast, then C is expressed as elongation. Plant Physiol. 117: 1143–1152. Chrispeels, M.J., Crawford, N.M. and Schroeder, J.I. (1999) Proteins for trans- C = (V0/V) Co0,(4)port of water and mineral nutrients across the membranes of plant cells. Plant Cell 11: 661–675. where Daniels, M.J., Chrispeels, M.J. and Yeager, M. (1999) Projection structure of a plant vacuole membrane aquaporin by electron cryo-crystallography. J. Mol. 3 Biol. 294: 1337–1349. V = 4/3 r ,(5) 0 0 de Groot, B.L. and Grubmuller, H. (2001) Water permeation across biological membranes: mechanism and dynamics of aquaporin-1 and GlpF. Science 294: and V0, Co0 and r0 are initial values of V, Co and r, respectively. Com- 2353–2357. bining equation (4) with equation (1) yields Dixit, R., Rizzo, C., Nasrallah, M. and Nasrallah, J. (2001) The Brassica MIP- MOD gene encodes a functional water channel that is expressed in the stigma dV/dt = PosSVw[(V0/V) Co0 – Co]. (6) epidermis. Plant Mol. Biol. 45: 51–62. Fleurat-Lessard, P., Frangne, N., Maeshima, M., Ratajczak, R., Bonnemain, J.- Next, substituting equations (2), (3) and (5) into equation (6) yields L. and Martinoia, E. (1997) Increased expression of vacuolar aquaporin and H+-ATPase related to motor cell function in Mimosa pudica L. Plant Physiol. 3 2 3 114: 827–834. dr /dt = 3r PosVw[(r0/r) Co0 – Co]. (7) Frangne, N., Maeshima, M., Schäffner, A.R., Mandel, T., Martinoia, E. and Bonnemain, J.L. (2001) Expression and distribution of vacuolar aquaporin in 3 2 Since dr /dt = 3r (dr/dt), equation (7) can be rewritten as young and mature leaf tissues of Brassica napus in relation to water fluxes. 286 Aquaporin levels and Pos of radish protoplasts

Planta 212: 270–278. (2001) Low aquaporin content and low osmotic water permeability of the Haines, T.H. (1994) Water transport across biological membranes. FEBS Lett. plasma and vacuolar membranes of a CAM plant Graptopetalum paraguay- 346: 115–122. ense: comparison with radish. Plant Cell Physiol. 42: 1119–1129. Higuchi, H., Suga, S., Tsuchiya, T., Hisada, H., Morishima, S., Okada, Y. and Otto, B. and Kaldenhoff, R. (2000) Cell-specific expression of the mercury- Maeshima, M. (1998) Molecular cloning, water channel activity and tissue insensitive plasma-membrane aquaporin NtAQP1 from Nicotiana tabacum. specific expression of two isoforms of radish vacuolar aquaporin. Plant Cell Planta 211: 167–172. Physiol. 39: 905–913. Ramahaleo, T., Morillon, R., Alexandre, J. and Lassalles, J.P. (1999) Osmotic Jiang, L., Phillips, T.E., Hamm, C.A., Drozdowicz, Y.M., Rea, P.A., Maeshima, water permeability of isolated protoplasts: modifications during develop- M., Rogers, S.W. and Rogers, J.C. (2001) The protein storage vacuole: a ment. Plant Physiol. 119: 885–896. unique compound organelle. J. Cell Biol. 155: 991–1002. Sarda, X., Tousch, D., Ferrare, K., Legrand, E., Dupuis, J.M., Casse-Delbart, F. Johansson, I., Karlsson, M., Johanson, U., Larsson, C. and Kjellbom, P. (2000) and Lamaze, T. (1997) Two TIP-like genes encoding aquaporins are The role of aquaporins in cellular and whole plant water balance. Biochim. expressed in sunflower guard cells. Plant J. 12: 1103–1111. Biophys. Acta 1465: 324–342. Santoni, V., Gerbeau, P., Javot, H. and Maurel, C. (2000) The high diversity of Jauh, G.Y., Phillips, T.E. and Rogers, J.C. (1999) Tonoplast intrinsic protein iso- aquaporins reveals novel facets of plant membrane functions. Curr. Opin. forms as markers for vacuolar functions. Plant Cell 11: 1867–82. Plant Biol. 3: 476–481. Karlsson, M., Johansson, I., Bush, M., McCann, M.C., Maurel, C., Larsson, C. Serraj, R., Frangne, N., Maeshima, M., Fleurat-Lessard, P. and Drevon, J.-J. and Kjellbom, P. (2000) An abundant TIP expressed in mature highly vacu- (1998) A -TIP cross-reacting protein is abundant in the cortex of soybean olated cells. Plant J. 21: 83–90. N2-fixing nodules. Planta 206: 681–684. Katsuhara, M., Akiyama, Y., Koshio, K., Shibasaka, M. and Kasamo, K. (2002) Suga, S., Imagawa, S. and Maeshima, M. (2001) Specificity of the accumula- Functional analysis of water channels in barley roots. Plant Cell Physiol. 43: tion of mRNAs and proteins of the plasma membrane and tonoplast aquapor- 885–893. ins in radish organs. Planta 212: 294–304. Kirch, H.H., Vera-Estrella, R., Golldack, D., Quigley, F., Michalowski, C.B., Suga, S., Komatsu, S. and Maeshima, M. (2002) Aquaporin isoforms respon- Barkla, B.J. and Bohnert, H.J. (2000) Expression of water channel proteins in sive to salt and water stresses and phytohormones in radish seedlings. Plant Mesembryanthemum crystallinum. Plant Physiol. 123: 111–124. Cell Physiol. 43: 1229–1237. Maeshima, M. (1992) Characterization of the major integral protein of vacuolar membrane. Plant Physiol. 98: 1248–1254. Sui, H., Han, B.G., Lee, J.K., Walian, P. and Jap, B.K. (2001) Structural basis of Maeshima, M. (2001) Tonoplast transporters: organization and function. Annu. water-specific transport through the AQP1 water channel. Nature 414: 872– Rev. Plant Physiol. Plant Mol. Biol. 51: 469–497. 878. Marin-Oliver, M., Chevalier, T., Fobis-Lisy, I., Dumas, C. and Gaude, T. (2000) Tsai, S.T., Zhang, R. and Verkman, A.S. (1991) High channel-mediated water Aquaporin PIP genes are not expressed in the stigma papillae in Brassica permeability in rabbit erythrocytes: characterization in native cells and oleracea. Plant J. 24: 231–240. expression in Xenopus oocytes. Biochemistry 30: 2087–2092. Maurel, C. (1997) Aquaporins and water permeability of plant membranes. Tyerman, S.D., Niemietz, C.M. and Bramley, H. (2002) Plant aquaporins: multi- Annu. Rev. Plant Physiol. Plant Mol. Biol. 48: 399–429. functional water and solute channels with expanding roles. Plant Cell Envi- Moshelion, M., Becker, D., Biela, A., Uehlein, N., Hedrich, R., Otto, B., Levi, ron. 25: 173–194. H., Moran, N. and Kaldenhoff, R. (2002) Plasma membrane aquaporins in the Yamada, S., Katsuhara, M., Kelly, W.B., Michalowski, C.B. and Bohnert, H.J. motor cells of Samanea saman: diurnal and circadian regulation. Plant Cell (1995) A family of transcripts encoding water channel proteins: tissue- 14: 727–739. specific expression in the common ice plant. Plant Cell 7: 1129–1142. Morillon, R., Catterou, M., Sangwan, R.S., Sangwan, B.S. and Lassalles, J.P. Yamada, S., Nelson, D.E., Ley, E., Marquez, S. and Bohnert, H.J. (1997) The (2001) Brassinolide may control aquaporin activities in Arabidopsis thal- expression of an aquaporin promoter from Mesembryanthemum crystallinum iana. Planta 212: 199–204. in tobacco. Plant Cell Physiol. 38: 1326–1332. Nielsen, S., Smith, B.L., Christensen, E.I., Knepper, M.A. and Agre, P. (1993) Zeidel, M.L., Nielsen, S., Smith, B.L., Ambudkar, S.V., Maunsbach, A.B. and CHIP28 water channels are localized in constitutively water-permeable seg- Agre, P. (1994) Ultrastructure, pharmacologic inhibition, and transport selec- ments of the nephron. J. Cell Biol. 120: 371–383. tivity of aquaporin channel-forming integral protein in proteoliposomes. Bio- Ohshima, Y., Iwasaki, I., Suga, S., Murakami, M., Inoue, K. and Maeshima, M. chemistry 33: 1606–1615.

(Received October 15, 2002; Accepted December 24. 2002)