УДК 581.1


© 2015 A. V. Shuvalov*, J. V. Orlova*, L. A. Khalilova*, N. A. Myasoedov*, I. M. Andreev*, D. V. Belyaev*,**, Y. V. Balnokin*,***

* K.A. Timiryazev Institute of Plant Physiology, Russian Academy of Sciences, Moscow

** Moscow Institute of Physics and Technology, Dolgoprudny, Moscow region

*** Faculty of Biology, M.V. Lomonosov Moscow State University, Moscow

Received April 22, 2014

Cl–/H+ exchange activity in the membranes isolated from the root cells of the halophyte Suaeda altissima (L.) Pall. was originally revealed and characterized. The membrane vesicles were isolated by centrifugation of microsomes in a continuous iodixanol density gradient. The highest activity of latent inosine diphosphatase, a marker of Golgi membranes, was localized in the upper part of the gradient, indicating its enrichment with Golgi membranes. The same part of the gradient was characterized by the highest Cl–/H+ exchange rate. The Cl–/H+ exchange activity was detected as electrogenic ΔрCl-dependent H+ transport monitored as changes in differential absorbance of a ΔpH-probe acridine orange, or as changes in fluorescence excitation spectrum of a pH-probe pyranine loaded into the vesicles. Generation of transmembrane electric potential (Δψ) during the Cl–/H+ exchange was assayed as changes in differential absorbance of a Δψ-probe safranin O. Establishing the transmembrane ΔрCl inward vesicles resulted in H+ efflux sensitive to DIDS (4,4’-diisothiocyano-2,2’-stylbene-disulfonic acid), an inhibitor of chloride transporters and channels, and generation of Δψ negative inside. To maintain the ΔрCl-dependent H+ efflux from the vesicles, either the presence of a penetrating cation tetraphenylphosphonium neutralizing negative charges inside the vesicles or null K+ diffusion potential across the membranes was required. The results demonstrate the activity of an electrogenic Cl–/H+ antiporter in the fraction enriched with Golgi membranes. We hypothesize that the Cl–/H+ antiporter is involved into the regulation of cytoplasmic Cl– concentrations by vesicular trafficking of Cl– from the cytoplasm to the vacuole by endosomes, derivatives of Golgi membranes.


1 This text was submitted by the authors in English.


Abbreviations: AO – acridine orange (N,N,N',N'-tetramethylacridine-3,6-diamine); β-ME – β-mercaptoethanol; DIDS – 4,4’-diisothiocyano-2,2’-stylbene-disulfonic acid; ER – endoplasmic reticulum; mon – monensin; GEF – Golgi-enriched fraction; IDP – inosine diphosphatase; NPPB – 5-nitro-2-(3-phenylpropylamine)benzoic acid; PM – plasma membrane; PMSF – phenylmethylsulfonyl fluoride; PVP-40 – polyvinylpyrrolidone with the average mol wt of 40 000; TPP+ – tetraphenylphosphonium.

Corresponding author: Yurii V. Balnokin. Botanicheskaya ul. 35, Moscow, 127276 Russia; K.A. Timiryazev Institute of Plant Physiology, Russian Academy of Sciences, fax: +7 (499) 977-8018; e-mail: balnokin@mail.ru

Keywords: Suaeda altissima – halophyte – Golgi membranes – anion–/H+ exchange – Cl–/H+ antiporter – vesicular anion trafficking



The transmembrane Cl–/H+ exchange performed by Cl–/H+ antiporters plays an important role in the regulation of Cl– and H+ concentrations in cellular compartments and is implicated in many physiological functions in animals and plants [1, 2]. The Cl–/H+ antiporters belong to the ubiquitous CLC family [3]. Originally, these proteins had been described as anion channels, but later it was shown that many of them function as electrogenic anion/proton antiporters exchanging anions for H+ with the 2 : 1 stoichiometry [4, 5]. 

The CLC genes were found in many plants [3]. Seven CLC genes were discovered in the Arabidopsis thaliana genome [3, 6, 7] with four of them, AtCLCa, AtCLCb, AtCLCc, and AtCLCg encoding the tonoplast proteins. Two Arabidopsis CLC members, AtCLCd and AtCLCf, are targeted to Golgi and one member, AtCLCe, to thylakoid membranes [2, 3, 8]. The products of the CLC genes have not been found in the plant plasma membranes (PM), where various anion channels have been identified by electrophysiological methods [9, 10]. Functions and the physiological role of plant CLCs are being extensively studied [2]. Using patch clamp technique on vacuoles isolated from wild-type and clca knock-out mutant plants, de Angeli et al. [5] showed that AtCLCa functions as a NO3/H+ antiporter and is implicated into the compartmentation of NO3– to the vacuole [5]. AtCLCb, a close relative of AtCLCa, was shown to be capable of conducting outwardly rectifying anionic current when expressed in oocytes [11]. The current was the largest in the presence of nitrate and was accompanied by the H+ counter transport, arguing in favor of AtCLCb being NO3–/H+ antiporter as well. 

Unlike AtCLCa and AtCLCb, another tonoplast protein AtCLCc appears to transport Cl–. AtCLCc was shown to be strongly expressed in stomatal guard cells. The disruption of the AtCLCc gene affected the physiological responses directly related to the transport of chloride ions across the vacuolar membrane. The Cl–/H+ content in guard cells of the AtCLCc mutants was several times lower than in wild type, and the stomata opening in response to light was reduced in the mutants [12]. Compared to wild-type plants, the mutants were more sensitive to NaCl.

Functions and physiological role of AtCLCg, a tonoplast protein, AtCLCd and AtCLCf both targeted to Golgi, and AtCLCe localized in thylakoid membranes are poorly studied in comparison with the tonoplast proteins AtCLCa, AtCLCb, and AtCLCc. 

In addition to NO3– and Cl– compartmentation, the physiological role of A–/H+ antiporters is to compensate the positive charges resulting from electrogenic H+-ATPase operation in organelle membranes. The H+ transport into acidic compartments by the H+-ATPases leads to membrane hyperpolarization that inhibits H+-ATPases and thus retards the compartment acidification. The electrogenic A–/H+ exchange via the antiporter abolishes this inhibition and thereby maintains the luminal acidification. This mechanism is involved in the maintenance and tight control of acidic pH in various intracellular compartments in animal cells [1] and was proposed to work in plants [2]. 

Little is known about the role of Cl–/H+ antiporters in plant salinity tolerance. The data in this respect have been reported in the above-mentioned paper on AtCLCc [12]. GmCLC1, a tonoplast-localized Cl– transporter from soybean, was shown to be involved into Cl– accumulation in vacuoles. This transporter was shown to alleviate NaCl stress when expressed in BY-2 tobacco cells [13].

The regulation of Cl– concentrations in the cytoplasm is likely to be a task of particular importance for plants growing on salt-affected soils. In the present study, we isolated a membrane fraction enriched with Golgi membranes from the roots of the halophyte Suaeda altissima and demonstrated the electrogenic Cl–/H+ exchange in isolated membranes. These data argue in favor of Cl–/H+ antiporter(s) functioning in Golgi of S. altissima root cells.



Plant material. Seeds of Suaeda altissima (L.) Pall. were collected from the plants in their natural habitat, the shore of the salt lake Elton in the Volgograd region, Russia, and germinated in moist sand. Three-week-old seedlings were potted in 3-L vessels (four plants per vessel) in aerated nutrient solution of Robinson and Downton [14] supplemented with 100 mM NaCl. The further growth of S. altissima occurred in a climatic chamber at 24°C and light intensity of 150 W/m2 from high pressure sodium vapor lamps (Reflux DNAZ-400, Russia) with a 16-h photoperiod. 

Isolation of the partially purified microsomes. All steps of the isolation procedure were carried out at 4°C. Roots of 60- to 65-day-old plants (80 g) were homogenized in 240 mL of medium A containing 0.5 M sucrose, 5 mM DTT, 2.5 mM K2S2O5, 5 mM β-ME, 3 mM EGTА, 50 mM Tris/Mes, pH 7.5, 0.6% PVP-40, 2 mM MgSO4, and 0.5 µL/mL of saturated PMSF solution in ethanol as well as 200 µM pyranine and, when needed, K2SO4 (see below). After filtration, the homogenate was centrifuged at 15 000 g for 20 min and the resulting supernatant was centrifuged at 200 000 g for 1 h (Beckman 70 Ti rotor). The 200 000 g pellet (microsomes) was resuspended in medium B containing 0.5 M sucrose, 5 mM DTT, 5 mM β-ME, 3 mM EGTA, 2 mM MgSO4, and 20 mM BTP/Mes, pH 7.5. The resulting microsomal suspension was layered over 30% (w/v) sucrose cushions prepared in the medium B. After centrifugation at 141 000 g for 1.5 h (Beckman SW 28 rotor); the membrane vesicles were collected at the 0/30% sucrose interface, diluted with the medium B, and centrifuged at 200 000 g for 50 min (Beckman 70 Ti rotor). The resulting pellet was resuspended in the medium B (final volume of 1 mL). This membrane preparation is referred to as the partially purified microsomes.

Golgi-enriched fraction isolation was conducted by membrane flotation in a continuous iodixanol (Opti-PrepTM, “Sigma”, United States) density gradient. The 0–30% iodixanol gradient in the medium B in volume of 10 mL was prepared in 13-mL centrifuge tubes. The partially purified microsomes (5–6 mg of membrane protein in 800 µL) was mixed with 1200 µL of the solution containing 53.5% iodixanol and all components of the medium B, so that the concentration of iodixanol in the microsome suspension was brought to 32%. The microsome suspension was placed at the bottom of the tubes under the iodixanol gradient. The gradients were centrifuged at 240 000 g for 1.5 h (Beckman SW 40 rotor). The 0.5-mL fractions were collected from the gradient upper part, which contained membranes (fig. 1a), and analyzed for their membrane marker activities, Cl–/H+ exchange activity, and protein contents. 

Cl–/H+ exchange detection. Transmembrane Cl–/H+ exchange was observed as alkalization of the vesicular lumen in response to the establishment of the inward Cl– concentration gradient across the membrane. The lumen alkalization was detected as: (1) change in the differential absorbance of a membrane-penetrating ΔpH-probe acridine orange (N,N,N',N'-tetramethylacridine-3,6-diamine, AO) (“Sigma-Aldrich”) [15], or (2) change in the fluorescence excitation spectrum of a non-penetrating across biological membranes ratiometric fluorescent pH-probe pyranine (8-hydroxy-1,3,6-pyrene trisulfonate) (“Invitrogen”) loaded into the vesicles [16]. 

The assays with AO require a preliminary establishment of the acidic-inside ΔpH across the membrane. The acidic-inside ΔpH was established by the addition of a Na+/H+ exchanger monensin (“Sigma-Aldrich”) in the final concentration of 300 nM to the Na+-containing vesicles suspended in the Na+-free reaction mixture. Cl–/H+ exchange and dissipation of ΔpH across the vesicular membrane were induced by the following imposition of the transmembrane concentration Cl– gradient (ΔрCl) by addition of Cl– (70 mM, if not indicated otherwise) to the reaction mixture. The assays were conducted in 1.5 mL of the standard reaction mixture containing 0.5 M sucrose, 20 мМ BTP/Mes (pH 7.5), 3 mM EGTA, 2 mM MgSO4, vesicles (40–100 µg of membrane protein), and 10 µM AO. The changes in differential absorbance of AO were recorded with a Hitachi-557 dual-wavelength spectrophotometer at 492–540 nm. 

For monitoring luminal pH with pyranine, this fluorescent pH-probe was loaded into the vesicles and the measurements of the intravesicular pH changes were performed in the absence of transmenbrane electric potential. This was attained by the establishment of null K+ diffusion potential. The loading was performed by the addition of pyranine at the final concentration of 200 µM to medium A used for root homogenization (see above) and supplemented additionally with 50 mM K2SO4. Alternatively, pyranine was loaded into the vesicles by hypoosmotic shock. To this end, the membrane fractions collected from the iodixanol gradient were diluted with medium B, pelleted at 200 000 g for 50 min (Beckman 70 Ti rotor), and resuspended in the hypotonic medium containing 200 µM pyranine, 70 mM sucrose, 25 mM K2SO4, 2 mM Mg2SO4, 2.5 mM EGTA, 3 mM BTP/Mes, pH 7.5, and 1mM Cl/BTP, pH 7.5. To eliminate pyranine outside the vesicles, 0.5 mL of the final vesicle suspension was passed through the 1  10 cm column with G-50 (fine) Sephadex (“Pharmacia”, Sweden). 

The assays with pyranine were performed in 1.5 mL of the reaction mixture containing the vesicles (30–50 µg of membrane protein). For monitoring luminal pH changes with pyranine loaded in the vesicle lumen during the vesicle isolation, the reaction mixture was identical to the one used in the experiments with AO but free of AO and supplemented with 200 nM valinomycin and K+ (K2SO4) equivalent to the K+ concentration in the vesicular lumen. The total luminal K+ comprised of endogenous and added K+ was estimated to be 77 ± 9 mM (n = 8) (see below). The reaction was initiated by the addition of 70 mM Cl/BTP (pH 7.5) to the vesicular suspension. When pyranine was loaded into the vesicles by hypotonic shock, the reaction mixture was identical to the hypotonic loading medium described above but free of pyranine and supplemented with 200 nM valinomycin, and the reaction was initiated by the addition of 25 mM Cl/BTP (pH 7.5). The excitation fluorescence spectra of pyranine were recorded with a Hitachi 850 fluorescence spectrophotometer at λem = 510 nm. The ratio of emission intensities at two excitation wavelengths (F458/F405) was calculated. To convert fluorescence intensity ratios into intravesicular pH values, the calibration was performed as follows. Hypotonic media buffered with 25 mM BTP/Mes to various pHs were loaded into the vesicles. The 4-parameter Hill function was fitted to the calibration data over pH range from 6.7 to 8.2 (fig. 2) using Sigma Plot 11.0 software. Luminal pH changes mediated by Cl–/H+ exchange were calculated as the differences between luminal pHs before and after 25 mM Cl/BTP addition to the vesicle suspensions. 

Detection of transmembrane electric potential generation. Generation of interior-negative or interior-positive electric potentials (Δψ) was detected by monitoring differential absorbance of Δψ-probes safranin O (2,8-dimethyl-3,7-diamino-phenazineat) (“Sigma-Aldrich”) at 524–554 nm [17] or oxonol VI (bis-(3-propyl-5-oxoisoxazol-4-yl)pentamethine oxonol) (“Invitrogen”) at 590–610 nm [18], respectively, using a Hitachi 557 spectrophotometer. The measurements were carried out in 1.5 mL of the standard reaction medium (see above) containing the vesicles (70–100 µg of membrane protein) and 8 µM safranine O or 2 µM oxonol VI. The interior-negative Δψ generation was initiated by the addition of 70 mM Cl– in the form of various salts (pH 7.5) and the generation of interior-positive Δψ was initiated by the addition of a penetrating cation tetraphenylphosphonium (TPP+). 

Estimations of intravesicular K+ and Na+ concentrations. K+ and Na+ concentrations in vesicular lumen were estimated as the concentrations of the ions in the supernatant obtained after centrifugation of the vesicle suspension at 200 000 g during the partially purified microsome isolation procedure. K+ and Na+ concentrations in the supernatant were determined using a flame photometer Leki FP 640 (“LEKI Instruments”, Finland).

Analytical methods. Cyt c oxidase (EC and antimycin A-insensitive NADP(H)-dependent cyt c reductase (EC were assayed spectrophotometrically according to [19] at room temperature in the reaction mixture containing 20 mM BTP/Mes (pH 7.2), 1 µM antimycin A, 10 mM NaN3, 200 µM NADPH, and 20 µM cyt c. For determination of cyt c oxidase and antimycin A-insensitive NADP(H)-dependent cyt c reductase activities, DDT and β-ME were omitted from the media A and B used in the procedure of the membrane vesicle isolation. Protein content was measured by the method of [20] with BSA as a standard. The activities of vanadate- and nitrate-sensitive ATPases as well as latent IDPase were assayed at 28°C for 60 min in 0.2 mL of the corresponding reaction medium and calculated from the amount of inorganic phosphate released from the added substrate. Vanadate- and nitrate-sensitive ATPase activities were assayed in the reaction media containing 40 мМ BTP/Mes (pH 6.5 for vanadate-sensitive or pH 7.8 for nitrate-sensitive ATPase), 1 мМ EGTA, 1 мМ MgSO4, 1 mM ATP, the vesicles (10 µg membrane protein) and 500 µM vanadate or 100 mM NO3/BTP (pH 7.8). The reaction mixture for determination of latent IDPase activity contained 0.5 M sucrose, 40 мМ BTP/Mes (pH 7.0), 1 мМ EGTA, 1 мМ MgSO4, 50 mM KCl, 1 mM IDP, 0.5% Triton X-100, and the vesicles (10 µg membrane protein). The released inorganic phosphate content was determined by the method described in [21]. 

Statistics. All the experiments were repeated 5–8 times with the membrane vesicles from independent preparations. The results of typical experiments are presented in the figures. Where presented, the data are expressed as means ± SD.



Centrifugation of the microsomes through 30% sucrose cushion resulted in a membrane fraction referred to as the partially purified microsomes (table). About 35 mg of microsomal protein were obtained from 80 g of the roots and approximately one third of this amount was collected from the 30% sucrose cushion. Most of microsomal vanadate- or nitrate-sensitive ATPase activities, markers of the membranes containing P-type ATPases or vacuolar/endosomal membranes, respectively, were retained by the partially purified microsomes. Specific activities of these enzymes in the fraction were approximately two times higher than in the initial microsomes. All latent IDPase activity, a known marker of Golgi, being in the microsomes was found in the fraction of partially purified microsomes, and the specific activity of the enzyme was 2.5 times higher in this fraction than in the microsomes. On the other hand, only one third of the microsomal antimycin A-insensitive NAD(P)H-cyt c reductase activity, an ER marker, was collected from the 0/30% interface of the sucrose gradient with specific activity of this enzyme being equal to its initial level. Only 8% of total microsomal cyt c oxidase activity was detected in the partially purified microsomes, and the specific activity of this enzyme lowered by 78%, indicating removing the bulk of mitochondrial membranes. Thus, the considerable portion of the membranes (63% on the protein basis) was removed by centrifugation through the 30% sucrose cushion. However, Golgi and tonoplast, the main candidates to contain Cl–/H+ antiporters, were mostly retained in the partially purified microsomes (table).

The partially purified microsomes were further fractionated in a continuous iodixanol density gradient. After centrifugation, the membranes resided in the upper 2/3 of the gradient (fig. 1a). The part of the gradient that contained membranes was divided into 0.5-mL fractions. Typical distributions of the marker enzyme activities, the Cl–/H+ exchange activity, and protein content among the fractions are shown in figs. 1b and 1c. The protein content exhibited two small maxima in the fractions 1–7 and 8–11 (numbered from the top down). Two groups of enzyme activities corresponded to the protein maxima. The activity of nitrate-sensitive ATPase, a marker of the endomembranes, was the highest at the lower protein maximum (fractions 8–11), whereas latent IDPase, a Golgi marker, displayed the highest activity in the upper fractions (1–7). Vanadate-sensitive ATPase activity, like nitrate-sensitive ATPase activity, displayed the maximum at the fractions 8–11 with a shoulder in the region of fractions 4–5.

All membrane fractions collected from the iodixanol gradient were tested for their Cl–/H+ exchange activities using a pH-probe pyranine loaded into the vesicles at the first stage of the vesicle isolation. The Cl–/H+ exchange was detected at null K+ diffusion potential (see MATERIALS AND METHODS section) in the presence of valinomycin. The highest activity of the Cl–/H+ exchange was found in the upper part of the gradient. The profile of this activity was essentially coincident to the profile of latent IDPase activity, indicating that the Cl–/H+ exchange occurs largely in the Golgi membranes.

Next, we pooled the iodixanol gradient fractions 1–6 characterized by the highest latent IDPase activity into one fraction named Golgi-enriched fraction (GEF). The contamination of GEF with other membranes can be judged from the data presented in the table. Both total and specific vanadate-sensitive ATPase (P-type ATPase), nitrate-sensitive ATPase (V-ATPases of vacuolar, Golgi, and endosomal membranes), and antimycin A-insensitive NAD(P)H-cyt c reductase activities (endoplasmic reticulum) were substantially lower in GEF than in the partially purified microsomes. Only 1.5 and 8.8% of total and specific cyt c oxidase activities (mitochondria), respectively, measured in partially purified microsomes, were found in GEF. Although total IDPase activity was considerably lost during the iodixanol gradient fractionation, its specific activity remained at the level found after centrifugation through 30% sucrose cushion and was 2.5 times higher than in the initial microsomes. 

Next, the Cl–/H+ exchange activity in GEF was characterized. A pH probe AO and a pH probe pyranine were used for testing the alkalization of the vesicle lumen caused by Cl–/H+ exchange. The analysis of the lumen alkalization with AO required the preliminary acidification of the lumen because AO permits ΔрH measurement only if pH in the lumen is lower than in the external medium [15]. To meet this condition, we took advantage of the presence of endogenous sodium ions inside the isolated vesicles. The Na+ concentration in the vesicle lumen was estimated to be 19 ± 4 mM (n = 8). The addition of a Na+/H+ exchanger monensin to the vesicular suspension resulted in the efflux of Na+ from the vesicle lumen in exchange for H+ and thereby to generation of ΔрН across the vesicular membrane, which was monitored as a decrease in the AO differential absorbance (fig. 3a). 

With Cl– in the reaction medium, the operation of Cl–/H+ antiporter should result in the H+ efflux from the vesicular lumen and dissipation of ΔрН that was preliminary established across the membrane. However, the increase in AO differential absorbance, which would indicate ΔрН dissipation, was hardly detectable when Cl– ions were introduced into the vesicle suspension together with BTP, a cation known to weakly penetrate membranes (fig. 3a). The increase in differential absorption of AO was observed only after the addition of a lipophilic cation tetraphenylphosphonium (TPP+). The effect of TPP+ increased with its concentration. We suggested that in the presence of Cl/BTP the generation of a transmembrane electric potential, negative inside vesicles, as a result of the electrogenic Cl–/H+ exchange performed by a putative Cl–/H+ antiporter inhibited its further operation. It is evident that TPP+, as a cation easily penetrating through the membranes, neutralized negative charges inside the vesicles and thereby kept the antiporter functioning. 

The above suggestion on the inhibition of Cl–/H+ antiporter by the accumulated negative charges was validated by the experiments, in which the generation of negative-inside transmembrane electric potential was observed when Cl– ions in the form of Cl/BTP were added to the vesicle suspension. To directly detect the negative-inside Δψ [17] generation, a Δψ-sensitive probe safranine O was used at the same conditions as in fig. 3a. After generation of the negative-inside Δψ in response to Cl/BTP addition, TPP+ introduction to the medium led to rapid shifts of Δψ to more positive values with their subsequent slower changes in the opposite direction until steady-state Δψ levels were achieved (fig. 3b). The TPP+-induced shifts of Δψ and the final steady state of Δψ levels were higher with more TPP+ added. However, Δψ did not achieve the values observed before the Cl/BTP addition even at the highest TPP+ concentration of 2 mM.

To verify whether these abrupt shifts of safranine O differential absorbance in response to TPP+ really reflect the changes in Δψ, we compared the changes in differential absorbance of safranine O with Cl/BTP and (Cl/BTP + TPP+) added to the vesicles. The decrease in ΔA524–554 was smaller in response to addition of (Cl/BTP + TPP+) than of Cl/BTP alone indicating the Δψ shift to more negative value in the last case (fig. 3d). The intravesicular pH changes under Cl/BTP or (Cl/BTP + TPP+) were in accordance with the respective Δψ shifts (figs. 3c, 3d). The alkalization of the vesicle lumen was observed when Cl– was added with TPP+, but in fact there was no alkalization when Cl– ions were added without the permeable cation.

As a control, we examined the effect of TPP+ added to the vesicle suspension in the absence of Cl–. TPP+ without Cl– resulted in lumen alkalization as well (fig. 3e). We suggested that a motive force for the H+ efflux in this case was a positive-inside diffusion potential generated by TPP+. To check the validity of this suggestion, the shift of Δψ upon the TPP+ addition in the absence of Cl– was studied with another Δψ probe, oxonol VI that allows monitoring positive-inside Δψ (fig. 3f). Indeed, in this case, TPP+ added to the vesicle suspension led to a decrease in the differential absorbance of oxonol VI, indicating the generation of positive-inside TPP+ diffusion potential. The positive sign of the potential is also confirmed by the fact that K+ ionophore valinomycin added after TPP+ to K+ containing vesicles increased the differential absorbance of oxonol VI, indicating a shift in Δψ to a more negative value (fig. 3f). Despite the vesicles being suspended in K+-free media, namely, in the medium B and in the standard reaction medium, some K+ remained inside the vesicles throughout the vesicle isolation procedure and the ion transport experiments. The presence of K+ inside the vesicles was evidenced by the acidification of the vesicle lumen in response to K+/H+ exchanger nigericin (data not shown). The TPP+-induced lumen alkalization in the presence of Cl– as opposed to that in its absence was accompanied by the generation of negative-inside Δψ (figs. 3b, 3d) that is consistent with the functioning of Cl–/H+ antiporter in the vesicle membrane. 

In the following experiments the Cl–/H+ exchange activity was studied with a ratiometric pH-sensitive probe pyranine, using the isolated GEF vesicles with specified composition of the intravesicular medium. Pyranine and other components of the intravesicular medium were loaded in the isolated GEF vesicles by hypotonic shock. K+ concentrations inside the vesicles and in the reaction medium were equalized and a K+ ionophore valinomycin was added to the reaction medium, allowing the maintenance of null potassium diffusion potential through the vesicular membrane. An advantage of such an approach is the avoidance of the negative charge accumulation inside the vesicles in the course of the Cl–/H+ exchange, which would have inhibited the further antiporter operation. After Cl/BTP addition, the alkalization of the vesicle lumen was observed as changes in the fluorescence excitation spectrum of pyranine (fig. 4a). Figure 4b demonstrates the kinetics of the fluorescence intensity at λex = 458 nm and λem = 510 nm in response to Cl/BTP addition. Intravesicular pH changes (fig. 4c) were determined from the ratio of emission intensities at two excitation wavelengths (F458/F405) using calibration curve (fig. 2). Mes/BTP added as a negative control at the same conditions led only to slight alkalization of the lumen (fig. 4c). 

The operation of a Cl– transporter in the vesicle membranes was confirmed by the experiments with inhibitors of anion transport, such as DIDS – a known inhibitor of chloride transporters and channels in diverse organisms [2, 22, 23] and NPPB – an inhibitor of anion channels in mammals and plants [9, 10, 24]. In the present work, DIDS in the concentration of 50 µM suppressed the ΔpCl-dependent lumen alkalization to the level observed with Mes/BTP (fig. 4c). The reaction was progressively inhibited as DIDS concentration increased with IC50 about 7 µM (fig. 5). The ΔpCl-dependent lumen alkalization was insensitive to NPPB within the concentration range applied from 0 up to 50 µM. 

We examined other anions for their ability to drive H+ efflux from the vesicles. The inward concentration gradients of Cl–, NO3–, and Br– led to almost identical alkalizations of the vesicle lumen, while the gradients of malate or acetate elicited the acidification of the intravesicular medium (fig. 6). The alkalization of the vesicular lumen was higher with the higher Cl– concentration in the reaction medium, and it demonstrated saturated kinetics with KM for Cl– equal to 18.5 ± 1.7 mM (n = 4) (fig. 7). 



The results of this work demonstrate Cl–/H+ exchange in Golgi-enriched fraction (GEF) isolated from the roots of the halophyte S. altissima. The fraction was obtained by the centrifugation of the microsomes through 30% sucrose cushion and then in continuous iodixanol gradient. The specific activity of the latent IDPase in GEF was 2.5 times higher than in the initial microsomes (table), and a distribution of the Cl–/H+ exchange activity in the iodixanol gradient coincided with the activity of the latent IDPase (fig. 1b), indicating the localization of the Cl–/H+ antiporter in Golgi membranes.

The vesicle lumen of GEF was alkalized when the concentration gradient of Cl– was established across the membrane (figs. 3, 4, 6, 7), indicating ΔpCl–-induced H+ efflux from the vesicles. Taking into account that in A. thaliana CLC proteins have been shown to be localized in the tonoplast, Golgi, and thylakoids, but not in any other cell membranes [2, 3, 8] and assuming the same localization of these proteins in S. altissima cells, we suggested that the Cl–/H+ exchange observed in GEF was mediated by an electrogenic Cl–/H+ antiporter functioning predominantly in the Golgi membranes. The experiments with the application of a Δψ-probe safranine O directly demonstrated that the ΔpCl-dependent alkalization of the vesicle lumen in GEF was accompanied by the generation of the negative-inside Δψ across the vesicle membrane (fig. 3b), indicating that a positive diffusion potential of BTP or another cation, for instance K+, could not be responsible for the H+ efflux from the vesicle lumen. If the ΔpCl-dependent generation of the negative-inside Δψ was the result of the establishment of the negative Cl– diffusion potential, then the H+ transport would occur from the medium into the lumen and acidification but not alkalization of the lumen should be observed. The results obtained are in agreement with the functioning of electrogenic Cl–/H+ antiporter(s) in GEF membranes that generate electric potential negative inside the vesicles. 

The putative Cl–/H+ antiporter, being an electrogenic system, should be regulated by the transmembrane electric potential. Both the Cl– influx from the exterior into vesicle lumen and the H+ efflux in the opposite direction via the putative Cl–/H+ antiporter should lead to negative charge accumulation inside the vesicles. Without penetrating cations in the reaction medium, the accumulation of negative charges in the vesicle lumen must evidently stop the Cl–/H+ antiporter operation. Indeed, ΔpCl-dependent alkalization of the vesicle lumen went on only when Cl/BTP was added to the reaction medium together with a penetrating cation TPP+ neutralizing the negative charges inside the vesicles (figs. 3a, 3c) or when the null K+ diffusion potential across the membrane was established (figs. 4–7). The operation of the Cl–/H+ antiporter is validated also by progressively stimulating ΔpCl-dependent alkalization of the vesicle lumen and Δψ shifting to less negative values with increasing TPP+ concentration in the reaction medium (figs. 3a, 3b). 

The experiments with the inhibitors of anion transports argue also in the favor of Cl–/H+ antiporter activity in the membranes of GEF. DIDS, a known inhibitor of Cl– transporters and Cl– channels in mammals, bacteria, and plants [10, 22, 24], inhibited the ΔpCl-dependent lumen alkalization in the vesicles of the fraction with IC50 = 7 µM. At the same time, the alkalization of the vesicle lumen proved to be insensitive to another inhibitor of anion transport, NPPB, known as an inhibitor of anion channels [10, 24] (figs. 4c, 5). 

The putative Cl–/H+ antiporter did not appear to be specific for Cl–. The alkalization of the vesicular lumen also took place when NO3– or Br– gradients were imposed upon the membrane (fig. 6). The low selectivity to anions was demonstrated for the CLC transporters of A. thaliana vacuolar membrane by patch-clamp technique [3, 5, 11]. An alternative interpretation of our result is the functioning of several antiporters differing in selectivity to anions in the membranes of GEF. The observed low anion selectivity of the A–/H+ exchange may reflect the ability of S. altissima cells to store NO3–, an essential nutrient, and/or to accumulate various anions along with Cl– for lowering osmotic potential in vacuolar compartment under condition of high salinity.

According to our results, the gradients of malate or acetate caused an opposite effect, namely the acidification of the vesicle lumen, indicating the involvement of other mechanism(s) different from the A–/H+ antiporter, into the H+ transport. The anion selective channel involved into malate accumulation in vacuoles has been found in plant vacuolar membranes. Malate uptake through this malate influx system is energized by the vacuolar proton pump and the driving force for this process is the H+ electrochemical potential difference between the cytosol and the vacuole [25]. It should be noted that the acidification of the vesicle lumen mediated by weak acids may be performed without participation of transporters. Acetic acid and other weak acids are capable of crossing the membranes in their non-dissociated form releasing their proton at the trans side of the membranes and thereby lowering pH in the inner media [26].

The questions concerning a physiological role of the Cl–/H+ antiporter targeted to Golgi in root cells of S. altissima deserve some attention. Apart from neutralizing negative charges accumulated in the cytoplasm during the H+-ATPase operation and pH regulation in organelles [1, 2], Cl–/H+ antiporters probably take part in Cl– homeostasis and contribute to plant salinity tolerance. Under high salinity, the equilibrium Cl– potential across PM can be more negative than transmembrane electric potential, i.e., situations are possible when Cl– ions move from the extracellular medium into the cytoplasm down gradient of electrochemical potential [27]. In these cases, the Cl–/H+ antiporter is capable of transporting Cl– from the cytoplasm into extracellular space or any intracellular compartments, in particular, vacuole. KM of the lumen alkalization for Cl– in GEF is revealed to be of 18.5 mM (fig. 7), the value that corresponds to Cl– concentrations in the cytoplasm of a related species Suaeda maritima [28]. 

Our findings do not exclude the possibility that a Cl–/H+ antiporter also functions in the tonoplast of S. altissima. The relatively high Cl–/H+ exchange activity was observed in the iodixanol gradient fractions 8–11 where nitrate-sensitive ATPase activity had a maximum (fig. 1b). The Cl– transport from the cytoplasm into vacuoles with the participation of the Cl–/H+ antiporter is possible either directly through the operation of Cl–/H+ antiporter of the tonoplast or via vesicular trafficking performed by endosomes, derivatives of trans-Golgi network that previously accumulated Cl– by their own Cl–/H+-antiporter operation. The implication of the vesicle traffic in the compartmentation of Na+ into vacuoles has been shown in Arabidopsis root cells [29]. The specific isoforms of Na+/H+ antiporters, NHX5 and NHX6, were found to be associated with trans-Golgi network and early endosomes and the double mutant nhx5 nhx6 was characterized by reduced growth and increased sensitivity to NaCl [30]. In accordance with the putative vesicular trafficking of Cl–, our previous electron microscopic studies of S. altissima exposed to NaCl showed an increased production of vesicles and multivesicular bodies in leaf and root cells at the boundaries of the cytoplasm and the central vacuole, as well as of the cytoplasm and PM. The local accumulation of Cl– in these zones was demonstrated by the electron cytochemistry method based on the formation of electron-dense AgCl granules after Ag+ treatment of tissues of plants grown on NaCl [31]. The question arises about the relevance of the ion vesicles trafficking. Martinoia et al. [32] hypothesized that vesicular trafficking of ions is beneficial in comparison with direct transport, because small vesicles would be energized more efficiently owing to their larger surface-to-volume ratios.

The authors are grateful to Dr. L. G. Popova for her valuable comments on the manuscript. 

The research was supported by the Russian Foundation for Basic Research, project no. 12-04-00987-а.


1.Jentsch T.J. Chloride and the endosomal–lysosomal pathway: emerging roles of CLC hloride transporters // J. Physiol. 2007. V. 578. P. 633–640. 

2.Barbier-Brygoo Н., de Angeli A., Filleur S., Frachisse J.-M., Gambale F., Thomine S., Wege S. Anion channels/transporters in plants: from molecular bases to regulatory networks // Annu. Rev. Plant Biol. 2011. V. 62. P. 25–51.

3.Marmagne A., Vinauger-Douard M., Monachello D., de Longevialle A.F., Charon C., Allot M., Rappaport F., Wollman F.A., Barbier-Brygoo H., Ephritikhine G. Two members of the Arabidopsis CLC (chloride channel) family, AtCLCe and AtCLCf, are associated with thylakoid and Golgi membranes, respectively // J. Exp. Bot. 2007. V. 58. P. 3385–3393.

4.Accardi A., Miller C. Secondary active transport mediated by a prokaryotic homologue of ClC Cl− channels // Nature. 2004. V. 427. P. 803–807. 

5.De Angeli A., Monachello D., Ephritikhine G., Frachisse J.M., Thomine S., Gambale F., Barbier-Brygoo H. The nitrate/proton antiporter AtCLCa mediates nitrate accumulation in plant vacuoles // Nature. 2006. V. 442. P. 939–942. 

6.Hehenberger M, Schwappach B., Fischer W.N., Frommer W.B., Jentsch T.J., Steinmeyer K. A family of putative chloride channels from Arabidopsis and functional complementation of a yeast strain with a CLC gene disruption // J. Biol. Chem. 1996. V. 271. P. 33 632–33 638.

7.Lv Q.D., Tang R.J., Liu H., Gao X.S., Li Y.Z., Zheng H.Q., Zhang H.X. Cloning and molecular analyses of the Arabidopsis thaliana chloride channel gene family // Plant Sci. 2009. V. 176. P. 650–661. 

8.Von der Fecht-Bartenbach J.V.D., Bogner M., Krebs M., Stierhof Y.D., Schumacher K., Ludewig U. Function of the anion transporter AtCLC-d in the trans-Golgi network // Plant J. 2007. V. 50. P. 466–474. 

9.Roberts S.K. Plasma membrane anion channels in higher plants and their putative functions in roots // New Phytol. 2006. V. 169. P. 647–666. 

10.Tavares B., Domingos P., Dias P., Feijo J.A., Bicho A. The essential role of anionic transport in plant cells: the pollen tube as a case study // J. Exp. Bot. 2011. P. 62. P. 2273–2298. 

11.Von der Fecht-Bartenbach J.V.D., Bogner M., Dynowsky M., Ludewig U. CLC-b-mediated NO3–/H+ exchange across the tonoplast of Arabidopsis vacuoles // Plant Cell Physiol. 2010. V. 51. P. 960–968.

12.Jossier M., Kroniewitcz L., Dalmas F., le Thiec D., Ephritikhine G., Thomine S., Barbier-Brygoo H., Vavasseur A., Filleur S., Leonhardt N. The Arabidopsis vacuolar anion transporter, AtCLCc, is involved in the regulation of stomatal movements and contributes to salt tolerance // Plant J. 2010. V. 64. P. 563–576. 

13.Wong T.H., Li M.W., Yao X.Q., Lam H.M. The GmCLC1 protein from soybean functions as a chloride ion transporter // J. Plant Physiol. 2013.V. 170. P. 101–104. 

14.Robinson S.P., Downton W.J.S. Potassium, sodium and chloride ion concentrations in leaves and isolated chloroplasts of the halophyte Suaeda australis R. Br. // Aust. J. Plant Physiol. 1985. V. 12. P. 471–479. 

15.Palmgren M.G. Acridine orange as a probe for measuring pH gradients across membranes, mechanism and limitations // Analyt. Biochem. 1991. V. 192. P. 316–321. 

16.Clement N.R., Gould J.M. Pyranine (8-hydroxy-1,3,6-pyrenetrisulfonate) as a probe of internal aqueous hydrogen ion concentration in phospholipid vesicles // Biochemistry. 1981. V. 20. P. 1534–1538. 

17.Akerman K.E., Wikstrom M.K. Safranine as a probe of the mitochondrial membrane potential // FEBS Lett. 1976. V. 68. P. 191–197. 

18.Bashford C.L., Chance B., Prince R.C. Oxonol dyes as monitors of membrane potential. Their behavior in photosynthetic bacteria // Biochim. Biophys. Acta. 1979. V. 545. P. 46–57. 

19.Hodges T.K., Leonard R.T. Purification of a plasma membrane-bound adenosine triphosphatase from plant roots // Methods Enzymol. 1974. V. 32B. P. 392–406. 

20.Simpson I.A., Sonne O. A simple, rapid and sensitive method for measuring protein concentration in subcellular membrane fractions prepared by sucrose density ultracentrifugation // Analyt. Biochem. 1982. V. 119. P. 424–427. 

21.Saheki S., Takeda A., Shimazu T. Assay of inorganic phosphate in the mild pH range, suitable for measurement of glycogen phosphorylase activity // Analyt. Biochem. 1985. V. 148. P. 277–281. 

22.Matulef K., Maduke M. Side-dependent inhibition of a prokaryotic ClC by DIDS // Biophys. J. 2005. V. 89. P. 1721–1730. 

23.Howery A.E., Elvington S., Abraham S.J., Choi K.H., Dworschak-Simpson S., Phillips S., Ryan C.M., Sanford R.L., Almqvist J., Tran K., Chew T.A., Zachariae U., Andersen O.S., Whitelegge J., Matulef K., Bois J.D., Maduke M.C. A designed inhibitor of a CLC antiporter blocks function through a unique binding mode // Chem. Biol. 2012. V. 19. P. 1460–1470. 

24.Jentsch T.J., Stein V., Weinreich F., Zdebik A.A. Molecular structure and physiological function of chloride channels // Physiol. Rev. 2002. V. 82. P. 503–568. 

25.Hafke J.B., Hafke Y., Smith J.A.C., Lüttge U., Thiel1 G. Vacuolar malate uptake is mediated by an anion-selective inward rectifier // Plant J. 2003. V. 35. P. 116–128. 

26.Antonenko Y.N., Bulychev A.A. Measurements of local pH changes near bilayer lipid membrane by means of pH microelectrode and a protonophore dependent membrane potential. Comparison of the methods // Biochim. Bophys. Acta. 1991. V. 1070. P. 279–282.

27.Teakle N., Tyerman S.D. Mechanisms of Cl– transport contributing to salt tolerance // Plant Cell Environ. 2010. V. 33. P. 566–589. 

28.Hajibagheri M.A., Flowers T.J. X-ray microanalyses of ion distribution within root cortical cells of the halophyte Suaeda maritima (L.) Dum. // Planta. 1989. V. 177. P. 131–134. 

29.Hamaji K., Nagira M., Yoshida K., Ohnishi M., Oda Y., Uemura T., Goh T., Sato M.H., Morita M.T., Tasaka M. Dynamic aspects of ion accumulation by vesicle traffic under salt stress in Arabidopsis // Plant Cell Physiol. 2009. V. 50. P. 2023–2033. 

30.Bassil E., Ohto M.A., Esumi T., Tajima H., Zhu Z., Cagnac O., Belmonte M., Peleg Z., Yamaguchi T., Blumwald E. The Arabidopsis intracellular Na+/H+ antiporters NHX5 and NHX6 are endosome associated and necessary for plant growth and development // Plant Cell. 2011. V. 23. P. 224–239. 

31.Балнокин Ю.В., Куркова Е.Б., Халилова Л.А., Мясоедов Н.А., Юсуфов А.Г. Пиноцитоз в клетках корня соленакапливающего галофита Suaeda altissima и его возможное участие в транспорте ионов Cl // Физиология растений. 2007. Т. 54. С. 892–901. 

32.Martinoia E., Meyer S., de Angeli A., Nagy R. Vacuolar transporters in their physiological context // Annu. Rev. Plant Biol. 2012. V. 63. P. 183–213.


Distribution of total and specific marker enzymatic activities in membrane fractions isolated from the roots of Suaeda altissima


The total activities of vanadate-sensitive ATPase, nitrate-sensitive ATPase, and latent IDPase are given in nmol Pi/(fraction s); the specific activities of these enzymes are in nmol Pi/(µg protein h). The total activities of antimicin A-insensitive NAD(P)H-cyt c reductase and cyt c oxidase are given in nmol cyt c/(fraction s); the specific activities of these enzymes are in nm cyt c/(µg protein h).



Fig. 1. Centrifugation of S. altissima root cell membranes collected from 30% sucrose cushion in a continuous iodixanol gradient. 

a – a view of the iodixanol gradient after centrifugation of the vesicle suspension; b – distribution of marker enzyme and Cl–/H+ exchange activities in the iodixanol gradient: (1) latent IDPase, (2) vanadate-sensitive ATPase, (3) nitrate-sensitive ATPase, and (4) Cl–/H+ exchange. For all enzymes, the highest activity among the fractions is taken as 100, corresponding to 18.2 nmol Pi/(fraction min) for latent IDPase, 38.9 nmol Pi/(fraction min) for vanadate-sensitive ATPase, and 63.2 Pi/(fraction min) for nitrate-sensitive ATPase. The Cl–/H+ exchange activities were measured with pyranine loaded into the vesicles in the course of their isolation. The measurements were performed at null K+ diffusion potential. The Cl–/H+ exchange activities are given as the intravesicular pH change after imposing ΔpCl upon the membranes and normalized to the highest fraction activity taken as 100; c – distribution of protein (5) and fraction densities (6).


Fig. 2. Calibration curve: the fluorescence intensity (λem = 510 nm) ratio of pyranine at two excitation wave lengths (F458/F405) as a function of intravesicular pH. 


Fig. 3. Induction of the ΔpCl-dependent vesicle lumen alkalization in GEF by a well-penetrating across membranes cation tetraphenylphosphonium (TPP+). 

The GEF vesicles were suspended in the standard reaction medium (see MATERIALS AND METHODS section). Changes of ΔpH were monitored with a ΔpH-probe AO (a, c, e) and Δψ with a Δψ-probe safranine O (b, d) or with a Δψ-probe oxonol VI (f); a, b – TPP+ (0.5 mM (1), 1 mM (2) or 2 mM (3)) was added after sequential additions of monensin (300 nM) and Cl/BTP (70 mM, pH 7.5); c, d – Cl/BTP (70 mM, pH 7.5) (1) or Cl/BTP (70 mM, pH 7.5) + TPP+ (1 mM) (2) were introduced into the vesicle suspension following the lumen acidification with monensin (300 nM); e, f – TPP+ (1 mM) was added to the vesicle suspension in the absence of Cl–.


Fig. 4. Alkalization of the vesicular lumen in GEF in response to Cl/BTP addition at null K+ diffusion potential and effect on it of an anion transport inhibitor DIDS (50 µM). 

The alkalization was monitored with a fluorescent ratiometric pH-sensitive probe pyranine loaded into the vesicle lumen by hypotonic shock in the medium containing 200 µM pyranine, 70 mM sucrose, 25 mM K2SO4, 2 mM Mg2SO4, 2.5 mM EGTA, 3 mM Mes/BTP, pH 7.5, and 1 mM Cl/BTP, pH 7.5. The reaction medium was the loading medium supplemented with 200 nM valinomycin. The alkalization was initiated by the addition of 25 mM Cl/BTP (pH 7.5). For a negative control, Mes/BTP was used instead of Cl/BTP in the loading medium and for the initiation of the reaction. a – excitation spectra of pyranine before (1) and after (2) the Cl/BTP addition; b – typical kinetics of the fluorescence intensity change at λex = 458 nm and λem = 510 nm in response to Cl/BTP addition; c – mean intravesicular pH changes for 8 min after Cl/BTP (25 mM), Mes/BTP (25 mM), or MES/BTP (25 mM) + DIDS (50 µM) additions.


Fig. 5. Concentration profiles of effects of anion transport inhibitors, DIDS (1) and NPPB (2), on ΔpCl-dependent alkalization of the vesicular lumen in GEF. 

The conditions for the reaction are the same as in the caption to fig. 4. 


Fig. 6. Intravesicular pH changes in the suspension of the GEF vesicles in response to addition of various 25 mM anion solutions: chloride (1), nitrate (2), bromide (3), malate (4), acetate (5) buffered with BTP to pH 7.5. 

The pH changes were monitored at null K+ diffusion potential with pyranine loaded in the vesicular lumen. The conditions of the loading and the reaction are described in fig. 4.


Fig. 7. Vesicular lumen alkalization in GEF as a function of Cl– concentration in the reaction medium. 

The pH changes were monitored at null K+ diffusion potential with pyranine loaded in the vesicular lumen. The conditions of the loading and the reaction are described in fig. 4. Cl– was added to the vesicular suspension in the form of Cl/BTP (pH 7.5).