Day/night regulation of aquaporins during the CAM cycle in Mesembryanthemum crystallinum


R. Vera-Estrella. Fax: +52 777 311 4691; e-mail:


Mesembryanthemum crystallinum exhibits induction of Crassulacean acid metabolism (CAM) after a threshold stage of development, by exposure to long days with high light intensities or by water and salt stress. During the CAM cycle, fluctuations in carbon partitioning within the cell lead to transient drops in osmotic potential, which are likely stabilized/balanced by passive movement of water via aquaporins (AQPs). Protoplast swelling assays were used to detect changes in water permeability during the day/night cycle of CAM. To assess the role of AQPs during the same period, we followed transcript accumulation and protein abundance of four plasma membrane intrinsic proteins (PIPs) and one tonoplast intrinsic protein (TIP). CAM plants showed a persistent rhythm of specific AQP protein abundance changes throughout the day/night cycle, including changes in amount of McPIP2;1, McTIP1;2, McPIP1;4 and McPIP1;5, while the abundance of McPIP1;2 was unchanged. These protein changes did not appear to be coordinated with transcript levels for any of the AQPs analysed; however, they did occur in parrallel to alterations in water permeability, as well as variations in cell osmolarity, pinitol, glucose, fructose and phosphoenolpyruvate carboxylase (PEPc) levels measured throughout the day/night CAM cycle. Results suggest a role for AQPs in maintaining water balance during CAM and highlight the complexity of protein expression during the CAM cycle.


Crassulacean acid metabolism (CAM) evolved as a strategy to concentrate CO2, and as a consequence, water use efficiency in these plants was also significantly increased (Cushman & Bohnert 1999). It is associated with plant adaptation to low-water environments and is thought to be present in approximately 6% of vascular plants (Winter & Smith 1996). CAM is characterized by day/night fluctuations in CO2 uptake; at night, CO2 enters the open stomata and is fixed into oxaloacetate by the enzyme phosphoenolpyruvate carboxylase (PEPc). Oxaloacetate is then reduced, and the resulting malic acid is transported across the tonoplast and stored in the vacuole. During the following light period, when stomata are closed, malate is remobilized from the vacuole, and decarboxylation of the organic acid releases CO2 that is then reassimilated by Rubisco via the normal C3 cycle (Cushman 2001). The halophyte Mesembryanthemum crystallinum shows induction of CAM, primarily in response to environmental stress, that can be accelerated by the developmental state of the plant (Winter & Holtum 2007). The plasticity to interchange between C3 and CAM has made this plant an attractive model system, advancing our knowledge in CAM evolution, physiology, biochemistry and molecular biology (Lüttge 1993; Schmitt et al. 1996; Winter & Smith 1996; Cushman & Bohnert 1999; Cushman 2001; Cushman & Borland 2002; Dodd et al. 2002). In CAM plants, the day/night cyclic transfer of CO2 between organic acids and carbohydrates across several membrane compartments requires an extensive coordination of the enzymes and transport proteins involved. This complex regulation is achieved by an internal circadian rhythm (Boxall et al. 2005) that influences the expression and activity of a night-specific PEPc kinase (PPcK). This PPcK largely controls the phosphorylation state and activity of the CAM-specific form of PEPc (Taybi et al. 2000), which results in measurable changes in PEPc activity in vitro (Scheible, Krapp & Stitt 2000), although the PEPc transcript itself undergoes diurnal changes in expression (Häusler et al. 2000; Scheible et al. 2000; Borland et al. 2006; Cushman et al. 2008). Fluctuations in osmotic potential caused by day/night changes in solute flux must, however, be accompanied by the passive movement of water across membranes.

Aquaporins (AQPs) are membrane proteins that mediate the transport of water, small solutes and gases (Maurel 2007). In higher plants, they belong to four main subgroups, PIP, TIP, SIP and NIP, based on sequence homology and, to a limited extent, membrane location (Barkla et al. 1999b; Ma et al. 2004; Vera-Estrella et al. 2004; Ishikawa et al. 2005; Uehlein et al. 2008). In the moss Physcomitrella patens, genome analysis has revealed additional major intrinsic proteins (MIPs), including a hybrid intrinsic protein (HIP) and two X intrinsic proteins (XIPs) that do not belong to any of the main plant MIP subgroups. Closer inspection revealed that these novel MIPs are also present in other plant species such as Selaginella moellendorffii, Gossypium raimandii, Nicotiana benthamiana, Citrus clementina, Solanum lycopersicum, Vitis vinifera and Ricinus communis (Danielson & Johanson 2008). The Populus trichocarpa genome is particularly high in these non-traditional MIPs as it includes five of these genes (Danielson & Johanson 2008).

The abundance of AQP isoforms in plants (55 in P. trichocarpa, Gupta & Sankararamakrishnan 2009) is attributed to the need for organelle, tissue and developmental specificity, as well as the requirement for the precise control of water transport under specific environmental conditions. An understanding of the significant role that AQPs play in plant water relations is slowly growing, facilitated by the analyses of plants with mutant AQPs or those in which AQP expression has been modified either through gene overexpression, or through reduced expression via antisense or RNAi-mediated changes (Ma et al. 2004; Siefritz et al. 2004; Bots et al. 2005). Light-dependent regulation of AQPs, and more specifically, circadian fluctuations in transcript levels, has been reported in several plant species, with most work focusing on PIPs. Diurnal variation in root hydraulic conductivity in Lotus japonicus was shown to be concurrent with alterations in transcripts for putative AQPs (Henzler et al. 1999). Similarly, transcripts for the ZmPIP1 and ZmPIP2 AQPs from maize roots showed circadian fluctuations associated with alterations in water transport (Lopez et al. 2003). Involvement of AQPs in the circadian rhythm of nyctinastic leaf movement has also been reported. In the leguminous Mimosaceae tree Samanea saman, accumulation of transcript for the PIP SsAQP2 showed highest expression at the beginning of the light period in the pulvinus, which coincided with a requirement for increased water permeability in the motor cells at the beginning of the light phase (Moshelion et al. 2002). In rice, it has been shown that the root-specific AQPs OsPIP2;2, OsPIP2;4 and OsPIP2;5, as well as OsPIP2;1 and OsPIP2;2 expressed both in roots and leaves, showed mRNA and protein diurnal changes that were enhanced by transpiration demand (Sakurai-Ishikawa et al. 2011). More direct evidence has been shown for the role of the tobacco AQP NtAQP1 in diurnal changes in leaf folding in this plant. High expression of NtAQP1 correlated with the beginning of the light phase when leaves unfolded, and NtAQP1 antisense plants showed highly impaired leaf movements (Siefritz et al. 2004). The AQP AtTIP2;1 was tentatively identified in one microarray study, suggesting regulation linked to the circadian rhythm; however, this result requires confirmation (Harmen et al. 2000).

In this study, we demonstrate that in M. crystallinum, changes in water permeability of protoplasts during the day/night CAM cycle corresponds to fluctuations in protein abundance of three PIPs: McPIP1;4, McPIP1;5 and McPIP2;1, and one TIP: McTIP1;2. Upon salt treatment of CAM plants, the day/night fluctuations of McTIP1;2 abundance were arrested while those for McPIP1;4, McPIP1;5 and McPIP2;1 maintained fluctuations similar to non-salt stress conditions.


Plant materials and growth conditions

M. crystallinum L. plants were grown from seed as previously described (Barkla et al. 1999a). Three weeks following germination, individual seedlings were transferred to pots containing soil (Metro Mix 500, Sun Grow Horticulture, Bellevue, WA, USA), with two plants per 15-cm-diameter pot, which were watered daily with tap water, and weekly with ½ strength Hoagland's solution. For experiments using root tissue, single plants were transferred to 1 L opaque tubs containing 800 mL of ½ strength Hoagland's medium (Hoagland & Arnon 1938), which was changed weekly. Salt treatment of CAM plants was initiated 6 weeks after germination (approximately 3 weeks following transfer to soil or hydroponic conditions) by inclusion of 200 mm NaCl in the Hoagland's medium or in the tap water. Plants were grown in a glasshouse under natural bright light and photoperiod. The greenhouse photosynthetic photon flux density reached a peak value of 1300 µmol m−2 s−1 during the middle of the day. This level of irradiation ensured that all adult plants were undergoing CAM. Experiments were conducted in the months of February to July at an altitude of 1800 m above the sea level for a period of three consecutive years. Minimum greenhouse temperatures ranged from 20 to 24 °C and maximum temperature was maintained at 25 °C. For experiments in which plants were exposed to 48 h of darkness, greenhouse-grown plants were transferred into a growth chamber maintained at 25 °C in the absence of light.

Membrane isolation and purification

Membranes were isolated from M. crystallinum plants as previously described (Barkla et al. 2002; Vera-Estrella et al. 2004). Leaves and roots of M. crystallinum plants were harvested and sliced into small pieces. Leaf and root material [15 g fresh weight (FW)] was immersed in 150 mL of ice-cold homogenization medium [400 mm mannitol, 10% (w/v) glycerol, 5% (w/v) polyvinylpyrrolidone (PVP)-10, 0.5% (w/v) bovine serum albumin (BSA), 1 mm phenylmethyl sulfonyl fluoride (PMSF), 30 mm Tris, 2 mm dl-dithiothreitol (DTT), 5 mm ethylene glycol tetraacetic acid (EGTA), 5 mm MgSO4, 0.5 mm butylated hydroxytoluene, 0.25 mm dibucaine, 1 mm benzamidine and 26 mm K+-metabisulfite, adjusted to pH 8.0 with H2SO4), and all subsequent operations were carried out at 4 °C. Tissue was homogenized in a commercial blender, filtered through four layers of cheesecloth and centrifuged at 10 000 g (20 min at 4 °C) using a JA20 rotor (Beckman, Mexico City, México) in a high-speed J2-HS centrifuge (Beckman). Pellets were discarded and the supernatants were centrifuged at 80 000 g (50 min at 4 °C) using a fixed angle 45Ti rotor (Beckman) in an L8-M ultracentrifuge (Beckman). The supernatant was aspirated, and the microsomal pellet was resuspended in suspension medium [consisting of 400 mm mannitol, 10% (w/v) glycerol, 6 mm Tris/Mes pH 8.0 and 2 mm DTT]. The microsomal suspension was then layered onto a discontinuous Suc gradient [consisting of a top layer of 9 mL of 22% (w/v) Suc, over 9 mL of 32% (w/v) Suc, on a cushion of 9 mL of 38% (w/v) Suc]. Gradients were centrifuged at 100 000 g (2 h at 4 °C) using a SW28 swinging bucket rotor in an L8-M ultracentrifuge. On a discontinuous Suc gradient tonoplast (TP) from M. crystallinum separates at the 0/22% Suc interface while plasma membrane (PM) is collected from the 32/38% Suc interface. Bands from the gradient were collected, diluted in suspension medium and centrifuged at 80 000 g using a 60Ti rotor (Beckman) in an L8-M ultracentrifuge. The resuspended pellets were collected, frozen in liquid N2 and stored at −80 °C.

Protoplast isolation

Protoplasts of leaves and roots were isolated according to the method of Miedema, Balderas & Pantoja (2000) with some modifications. Epidermal peeled leaf and root tissue pieces were enzymatically digested to remove cell walls with a cocktail (5 mm KCl, 1 mm MgCl2, 0.5 mm CaCl2, 5 mm ascorbic acid, 5 mm 2-(N-morpholino)ethanesulfonic acid (MES), 0.025% BSA (w/v), 0.5% cellulase (w/v), 0.01% (w/v) pectolyase, pH 6.8, 200 or 500 mosmol kg−1 adjusted with sorbitol for leaves and roots, respectively) and shaken for 20 min at 28 °C in a small Petri dish. Following digestion, protoplasts were released from the tissue by tapping the Petri plates five times; the protoplast solution was filtered through one layer of Miracloth (Calbiochem, La Jolla, CA, USA), placed into an Eppendorf tube and centrifuged at 1500 g for 3 min. The protoplasts were washed with a wash solution (5 mm KCl, 1 mm MgCl2, 0.5 mm CaCl2, 5 mm ascorbic acid, 5 mm MES, pH 6.8, with a final osmolarity of 200 or 500 mosmol kg−1 adjusted with sorbitol for leaves and roots, respectively), centrifuged for 3 min at 1500 g and resuspended in 0.5 mL of wash solution.

Preparation of DNA, in vitro transcription mRNA capping and oocyte expression

Plasmid DNA was isolated by alkaline lysis, purified by polyethylene glycol precipitation and digested with Pst1, to produce linear templates. Linearized DNA was extracted with phenol-chloroform, precipitated with ethanol and resuspended in RNase-free water. The complementary RNA (cRNA) was synthesized in vitro using the mMessage mMachine Capping kit (Ambion, Austin, TX, USA) according to manufacturer's instructions. Surgically isolated and defolliculated Xenopus laevis oocytes were injected with 46.0 nL of diethyl pyrocarbonate (DEPC)-H2O or the corresponding cRNA (50 ng in 46.0 nL) using a NANOJECT II automatic injector (Drummond, Broomall, PA, USA). Oocytes were incubated in iso-osmotic (200 mosmol kg−1) Barth's solution [10 mm N-2-hydroxyethyl piperazine-N′-2 ethanesulfonic acid (HEPES)-NaOH, pH 7.4, 88 mm NaCl, 1 mm KCl, 2.4 mm NaHCO3, 0.33 mm Ca(NO3)2, 0.41 mm CaCl2, 0.82 mm MgSO4] for 2 to 5 d post-injection before swelling assays were carried out.

Measurement of protoplast and oocyte water permeability

The osmotic water permeability (Pf) of oocytes and protoplasts was measured by assaying the rates of swelling upon transfer via pipetting into hypo-osmotic shock, from iso-osmotic Barth's solution [10 mm HEPES-NaOH, pH 7.4, 88 mm NaCl, 1 mm KCl, 2.4 mm NaHCO3, 0.33 mm Ca(NO3)2, 0.41 mm CaCl2, 0.82 mm MgSO4] for oocytes or wash buffer (200 or 500 mosmoles kg−1) for protoplasts, to hypo-osmotic dilute Barth's or wash solution (40 mosmol kg−1). Swelling was measured by video imaging on a Nikon Eclipse TE 300 microscope (Nikon, Mexico City, México), equipped with a Hitachi KP-D50 colour video camera (Hitachi Denshi, Ltd, Lake Mary, FL, USA). Images were captured and digitized by the Image-Pro Plus software (Version 4.5, Media Cybernetics, Silver Spring, MD, USA).

The osmotic water permeability (Pf, cm s−1) was calculated by the relation,


where V0 is the initial volume at time zero measured for each individual protoplast or oocyte. In the case of protoplasts, this value ranged between 1.23 10−9 cm3 and 5.97 10−9 cm3, with an average of 2.23 10−9 cm3; V/Vo is the relative volume; Vw is the molar volume of water (18 cm3 mol−1), S is the initial surface area calculated for each individual protoplast or oocyte, osmi is the osmolarity inside the cell and osmo is the osmolarity in the external medium. The time interval applied for the calculations of Pf was within the range of 0–20 s for protoplasts and 0–60 s for oocytes, and in both cases, the initial linear component of the graph within this range was used.

Gas exchange, photosynthesis and light intensity measurements

Single-leaf net photosynthetic rates, stomatal conductance and light intensity were measured with a LI-6400 portable photosynthesis system (LI-COR, Lincoln, NE, USA), operated in the open mode at 25 °C. External air was depleted of CO2 and then mixed with a supply of pure CO2 to produce a reference concentration of 400 µL L−1. Gas flow rate was 500 µmol s−1, and external humidity was 50–60%. A light-emitting diode source was placed on the upper half of the leaf chamber (2 × 3 cm). The middle portion of the youngest, fully expanded leaf was chosen for measurement under natural sunlight, and leaf temperature was maintained at 25 °C. All measurements were taken between 0700 and 2300 h.

Total protein extraction

Plant tissue (roots and whole leaves) was frozen and ground in liquid N2. Powdered samples (2.5 mL) were then homogenized in an equal volume of protein extraction buffer [250 mm mannitol, 10% (w/v) glycerol, 10 mm Tris/MES pH 8, 1 mm ethylenediaminetetraacetic acid (EDTA), 5 mm DTT, 1 mm benzamidine, 1 mm PMSF and 5% (w/v) insoluble PVP] and vortexed for 1 min. The samples were filtered through one layer of Miracloth and the crude protein extracts were centrifuged at 10 000 g for 15 min using an SS34 rotor in a Sorvall RC5C high-speed centrifuge (Dupont, Tlalnepantla, Edo de México, México) to remove cellular debris. Samples were frozen in liquid N2 for further use.

Protein determination

Protein content in purified tonoplast, plasma membrane and total protein extracts was measured by a modification of the dye-binding method (Bradford 1976), in which membrane protein was partially solubilized with 0.5% (v/v) Triton X-100 for 5 min before the addition of the dye reagent concentrate (Bio-Rad, Mexico City, México). BSA was employed as the protein standard.

Quantification of tissue osmolarity, pH, glucose, fructose, malate and pinitol content

Leaves and roots were collected, cut into small pieces and packed in 5 mL syringes containing a Whatman no. 1 filter disk (GE Healthcare, Piscataway, NJ, USA). The material was immediately frozen at −30 °C and the cell sap was obtained in thawed samples by centrifugation at 1200 g for 15 min using a S4180 rotor in a GS-15R centrifuge (Beckman). The osmolarity of the cell sap was measured in 50 µL samples using a cryoscopic osmometre (Osmomat 030: Gonotec, Berlin, Germany). Cell sap pH was measured with a pH microelectrode connected to a pH metre (Accumet, Fisher Scientific, Pittsburgh, PA, USA). Pinitol, fructose and glucose levels were measured by HPLC as previously described by Vernon & Bohnert (1992). These levels were quantified by comparing high-performance liquid chromatography (HPLC) peak areas from the samples with peak areas of pinitol and sugar standards of known concentrations. Malate was quantified enzymatically by a coupled enzyme assay performed according to Hohorst (1965). The reaction medium contained 50 mm glycylglycine (pH 10), 30 mm L-glutamate, 3 mmβ-nicotinamide adenine dinucleotide hydrate (NAD)+, 1 U of glutamate oxaloacetate transaminase (GOT, Sigma-Aldrich, Toluca, Mexico State, México), and 10 U of L-malate dehydrogenase (MDH, Sigma-Aldrich). Malate concentrations were obtained by calculating the difference in the absorbance at 340 nm before and after 20 min incubation at room temperature (RT). Measurements were made on four independent samples, and the results were expressed as µmol malate g−1 fresh weight.

One- and two-dimensional sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS–PAGE) and protein immunoblotting

For one-dimensional (1D) SDS-PAGE, protein was precipitated by dilution of the samples 50-fold in 1:1 (v/v) ethanol : acetone and incubated overnight at −30 °C according to the method of Parry, Turner & Rea (1989). Samples were then centrifuged at 13 000 g for 20 min at 4 °C using an F2402 rotor in a GS-15R table-top centrifuge (Beckman). Pellets were air-dried, resuspended with Laemmli (1970) sample buffer [2.5% (w/v) SDS final concentration], and heated at 60 °C for 2 min before loading and electrophoresis in 12.5% (w/v) linear acrylamide mini-gels.

For two-dimensional (2D) electrophoresis, 100 µg of microsomal protein was desalted and cleaned with the ReadyPrep 2D Cleanup kit (Bio-Rad) according to the manufacturer's instructions. The final protein pellet was resuspended in rehydration buffer [7 m urea, 2 m thiourea, 2% 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate hydrate (CHAPS), 2% amidosulfobetaine-14 (ASB-14), 0.2% Bio-Lyte ampholytes 3/10]. Ready Strip IPG strips (7 cm, linear pH 3–10, Bio-Rad) were layered gel side down onto samples containing microsomal protein isolated from roots or leaves of M. crystallinum and placed in the Protean IEF tray (Bio-Rad). Strips were covered with 1 mL of mineral oil, and active rehydration was carried out for 16 h in a Protean IEF Cell (Bio-Rad) at 50 V and 20 °C. Following strip rehydration, isoelectric focusing (IEF) was carried out with a three-step ramping protocol for a total of 30 000 V-h at 20 °C and a maximum current setting of 50 µA per strip. After IEF the IPG strips were first equilibrated for 15 min at RT in DTT equilibration buffer (6 m urea, 0.375 m Tris-HCl pH 8.8, 2% SDS, 20% glycerol and 2% DTT) on a shaker, followed by an additional 15 min at RT in iodoacetamide equilibration buffer (6 m urea, 0.375 m Tris-HCl pH 8.8, 2% SDS, 20% glycerol and 2.5% iodoacetamide). For the second dimension, IEF gel strips were loaded onto 10% acrylamide gels, and SDS-PAGE was carried out using the Protean Tetra Cell electrophoresis system (Bio-Rad) at 50 V for 4 h at 4 °C.

After electrophoresis, SDS-PAGE-separated proteins were transferred onto nitrocellulose membranes [enhanced chemiluminescence (ECL), GE Healthcare, Cd. de Mexico, México] as previously described (Vera-Estrella et al. 1999). Following transfer, membranes were blocked with TBS (100 mm Tris, 150 mm NaCl) containing 0.02% (w/v) Na-azide and 5% (w/v) fat-free milk powder (Sveltie, Nestle, Saltillo, Coahuila, México) for 2 h at room temperature. Blocked membranes were incubated for a minimum of 3 h at room temperature with the appropriate primary antibodies, followed by the addition of a 1:5000 dilution of secondary antibodies (goat anti-rabbit) conjugated to horseradish peroxidase. Immunodetection was carried out using the chemiluminescent ECL Western blotting analysis system (GE Healthcare). Mean intensity of the immunodetected protein bands were calculated as a relative calibrated measurement of the total band size and intensity using ECL molecular weight markers as loading control standards (Fermentas, Glen Burnie, MD, USA). Images were analysed using ImageJ 1.37v image analysis software [National Institute of Health (NIH), Bethesda, MD, USA] (Abramoff, Magelhaes & Ram 2004).

Total RNA isolation and semi-quantitative RT-PCR

Plant tissue (roots or whole leaves) was ground to a powder under liquid nitrogen followed by RNA extraction using Trizol according to the manufacturer's procedure (Invitrogen, GE healthcare, Piscataway, NJ, USA). Total RNA was treated with RNase-free DNase (Fermentas) for 30 min at 37 °C, then 5 µg was used to synthesize the first strand cDNA with the addition of 0.5 µg oligo(T) and 200 u RevertAid H Minus M-MuLC reverse transcriptase (Fermentas) and incubated at 42 °C for 1 h. PCR was performed in a volume of 50 µL containing 1.25 units Taq DNA polymerase (Fermentas), 0.2 mm dNTP mix, 2 mm MgCl2, and 0.4 µm gene-specific primers. Primer sequences used for the different AQPs, the McPPC1 (CAM-specific M. crystallinum PEPc), and the McUBQ (M. crystallinum ubiquitin) are shown in Supporting Information Table S1. PCR mixtures were initially denatured at 95 °C for 2 min. Cycling parameters were as follows: denaturing at 94 °C for 1 min, primer annealing at 58 °C (McUBQ) or 60 °C (AQPs and McPPC1) for 1 min, and extension at 72 °C for 1.5 min. The number of cycles was 23 for McUBQ; 28 for McPIP1;4, McPIP1;5, McTIP1;2 and McPIP2;1; and 32 for McPIP1;2 and McPPC1. Finally, PCR mixtures were extended at 72 °C for 10 min. Primer specificity was determined by sequencing the PCR amplification products for each gene. PCR products were size separated by electrophoresis on 1% agarose gels and visualized by staining with ethidium bromide. Gene expression was normalized to the expression of McUBQ for each time and treatment.

PEPc activity

PEPc activity in total protein extracts was measured spectrophotometrically at 340 nm and 30 °C in a 1 mL assay medium containing 50 mm Tris/MES pH 8.0, 1 mm EDTA, 10 mm MgSO4, 10 mm NaHCO3, 50 µmβ-nicotinamide adenine dinucleotide, reduced (NADH), 5 units malate dehydrogenase and 2 mm PEP. The rate of NADH oxidation in the presence of PEP, malate dehydrogenase and 25 µg total protein was calculated using the NADH molar extinction coefficient of 6220 cm s−1. Values were expressed as µmol mg−1 protein h−1.

Primary and secondary antibodies

Peptides representing the second extracellular loop (McPIP1;4, McPIP2;1, McPIP1;2) or the carboxyl terminus (McTIP1;2, McPIP1;5) of the deduced amino acid sequence of the M. crystallinum AQPs were synthesized, coupled to keyhole limpet haemocyanin and antibodies were generated as previously described (Kirch et al. 2000). Anti-McPIP1;4 and McPIP2;1 antibodies were used at a dilution of 1:250 whereas antisera for the detection of McTIP1;2, McPIP1;2 and McPIP1;5 were used at a dilution of 1:500.


Characterization of C3 and CAM plants

M. crystallinum clearly shows distinguishable juvenile and adult growth phases with a dramatic change in development as it switches in the adult phase to CAM metabolism upon exposure to stress (high light, temperature, salt or water deficit) (Adams et al. 1998; Winter & Holtum 2007). In this study, CAM was induced in well-watered adult plants by natural high light (>800 µmol m−2 s−1) and data is presented for CAM plants grown for 2 weeks in the absence or presence of 200 mm NaCl. To confirm the metabolic state of the plants, net photosynthesis rate and stomatal conductivity were measured simultaneously over a day/night cycle in young (4-week-old, C3) and adult (8-week-old, CAM) untreated and salt-treated (200 mm NaCl for 2 weeks, CAM-S) M. crystallinum plants (Fig. 1). In young plants (C3), rates of photosynthesis and stomatal conductance increased during the onset of the light period reaching a peak at around midday and steadily declined throughout the afternoon; by 1700 h rates were generally 50% of those recorded at midday. Only negligible photosynthesis rates and stomatal conductance were recorded during the dark period (Fig. 1a,b). Untreated (CAM) or salt-treated (CAM-S) adult CAM plants showed two distinct, temporal peaks of photosynthesis and stomatal conductance, one at the beginning of the light period and a second at the start of the dark period (Fig. 1a,b). Both, photosynthesis and stomatal conductance were very low during the hours of highest temperature and light intensity in the CAM plants (Fig. 1a,b,c, CAM and CAM-S), as expected; during this time, the CAM plants underwent noticeable wilting (Winter & Holtum 2007). The greenhouse photosynthetic photon flux density was highest during the middle of the day, reaching a peak value of 1300 µmol m−2 s−1.

Figure 1.

Day/night patterns of net photosynthesis (a), stomatal conductance (b) and photosynthetic photon flux density (PPFD) (c) in Mesembryanthemum crystallinum juvenile C3 plants, adult untreated Crassulacean acid metabolism (CAM) plants, and adult salt-treated CAM plants (CAM-S; 200 mm NaCl for 2 weeks). Measurements were taken at 1 h intervals throughout the day/night period from 7 until 23 h on the middle portion of the youngest fully expanded leaf from each plant. Yellow horizontal bars indicate the light period; black horizontal bars indicate the dark period. Each curve is representative of three replicate experiments that were carried out over consecutive 24 h periods on different plants.

Day/night changes in the swelling rate and water permeability of leaf and root protoplasts from C3 and CAM plants

In CAM plants, fluctuations in osmotic potential driven by day/night changes in solute flux and the subcellular compartmentation of these compounds, requires the compensatory control of water movement into cells and across endomembranes. This may result in measurable changes in the water permeability of membranes during the CAM cycle. To measure membrane water permeability, we isolated protoplasts from leaves and roots of juvenile C3, adult CAM untreated and adult CAM salt-treated plants at different times throughout the day, and measured the protoplast swelling response to a hypotonic challenge. The rate of swelling of protoplasts isolated from C3 plants was constant throughout the day/night cycle in both leaves (Fig. 2a) and roots (Fig. 2b), with similar rates being obtained at the different time points measured. In contrast, the rate of swelling of protoplasts from untreated adult CAM plants varied greatly depending on the time of day the protoplasts were isolated. Compared with protoplasts from leaves of C3 plants, leaf protoplasts from untreated adult CAM plants showed a decrease in swelling rate at 0700, 1100 and 1500 h, while at 1700 h, the swelling rate was similar in protoplasts from both C3 and CAM leaves (Fig. 2c). At the end of the light period (1900 h) the swelling rate was almost twofold higher in protoplasts from leaves of adult CAM untreated plants compared with protoplasts from leaves of juvenile C3 plants (Fig. 2c). In roots, the swelling rate of protoplasts from CAM untreated or C3 plants was similar at 0700 and 1100 h (Fig. 2b,d). However, the protoplasts from roots of CAM untreated plants showed a 2.2-, 2.6- and 4.0-fold increase in swelling rate compared with protoplasts isolated from roots of C3 plants at 15, 17 and 19 h, respectively (Fig. 2b,d). The rates of swelling of protoplasts isolated from leaves and roots of salt-treated CAM plants were similar to those observed for the untreated CAM plants (Fig. 2e,f).

Figure 2.

Relative volume changes (V1/Vo) and water permeability (Pf) values calculated for protoplasts isolated from leaves and roots of Mesembryanthemum crystallinum plants during the day/night cycle. Time course of protoplast swelling measured as the change in the relative volume of protoplasts isolated from leaves and roots of juvenile C3 (a, b), adult untreated Crassulacean acid metabolism (CAM) (c, d) and adult salt-treated CAM (CAM-S; e, f) plants after hypo-osmotic shock. The relative volume changes (V1/Vo) were measured with reference to initial volume (normalized to one) and plotted as a function of time after the introduction of protoplasts into a 40 mosmol hypotonic solution at time zero. Osmotic water permeability (Pf) values calculated from the initial linear component of the rate of protoplast swelling from leaves (g) and roots (h) as described in the Materials and Methods section. Values are the mean ± standard deviation (SD; n = 10).

The osmotic water permeability coefficient (Pf) was calculated from the rates of swelling of protoplasts isolated from leaves and roots of C3, untreated (CAM) and salt-treated (CAM-S) CAM plants for the different time points throughout the light period (Fig. 2g,h). In both roots and leaves of C3 plants, the water permeability remained constant throughout the day. In contrast, the water permeability in roots and leaves of CAM untreated and salt-treated plants showed a peak towards the beginning of the dark period, with the increase in roots preceding by several hours the increase observed in leaves (1500 h compared with 1900 h) (Fig. 2g,h).

AQP water transport activity and analysis of antibody specificity

During CAM, diurnal changes in AQP expression are thought to control passive water movement to help maintain cellular osmotic homeostasis (Lopez et al. 2004). To assess the role of AQPs in the observed day/night changes in membrane permeability, we chose five M. crystallinum AQPs, representative of both the TIP and PIP families that may have a role in plant water relations. We previously reported that while classified as a PIP, McPIP1;4 co-localizes with low-density tonoplast fractions, distinct from plasma membrane, on discontinuous sucrose density gradients of microsomal membranes from control and salt-treated M. crystallinum leaves and roots (Barkla et al. 1999a). More specifically, we have recently found that McPIP1;4 appears to co-localize with the prevacuolar compartment marker PEP12 (Oliviusson et al. 2006), in a discrete membrane fraction, when microsomal membranes are separated by surface charge using the technique of free flow zonal electrophoresis (Vera-Estrella, unpublished data). The remaining four AQPs localized to the membrane compartment suggested by their nomenclature; McPIP1;2, McPIP1;5 and McPIP2;1 to the plasma membrane and McTIP1;2 to the tonoplast (Barkla et al. 1999b). While McPIP1;2 and McTIP1;2 are present in both leaves and roots, McPIP1;5 is leaf specific and McPIP2;1 is only detected in roots.

Previous studies have demonstrated that McTIP1;2 and McPIP2;1 show high water permeability when heterologously expressed in X. laevis oocytes (Vera-Estrella et al. 2004; Amezcua-Romero, Pantoja & Vera-Estrella 2010). Here, we confirm this and also demonstrate that McPIP1;4, McPIP1;5 and McPIP1;2 mediate the movement of water to varying degrees (Supporting Information Fig. S1), although with water permeability values on the order of half those measured in the highly water permeable TIP and PIP2 isoforms, McTIP1;2 and McPIP2;1 (Table 1). PIP1 class AQPs typically show low water permeability, although antisense studies have shown they play a significant role in cellular and whole-plant water relations (Siefritz et al. 2002).

Table 1.  Osmotic water permeability (Pf) of Xenopus laevis oocytes injected with 50 ng of the indicated cRNA, or RNase free water (H2O) under hypo-osmotic shock
cRNA injectedPf (10−3 cm s−1)
  1. Pf values were calculated from the initial linear component of the swelling rate of oocytes as described in the Materials and Methods section. Values are mean ± SE (n = 15). The Vo values were measured for each independent oocyte.

H2O1.3 ± 0.78
McPIP1;28.25 ± 1.26
McPIP1;46.78 ± 1.76
McPIP1;57.39 ± 0.87
McPIP2;113.9 ± 0.98
McTIP1;216.0 ± 1.20

Western blot analysis of leaf and root microsomal membranes separated by 2D-PAGE (Fig. 3a,b, respectively), using peptide-specific antibodies to the different AQPs (Kirch et al. 2000) was used to evaluate isoform-specific recognition. As observed in Fig. 3c,e–g, the antibodies raised against McPIP1;4, McTIP1;2, McPIP1;5 and McPIP2;1 recognized a single protein spot on immunoblots of 2D-PAGE-separated proteins, corresponding to 40, 34, 30 and 31 kDa, respectively. In contrast, the antibody raised against a peptide of McPIP1;2 recognized four discrete protein spots (45, 34, 22 and 15 kDa; Fig. 2d), suggesting it may either cross-react with several PIP family members with distinct isoelectric points and molecular mass or other unrelated proteins, particularly the 22 and 15 kDa polypeptides that are below the expected molecular mass for AQPs (30–41 kDa). We have detected a polypeptide of 34 kDa corresponding to McPIP1;2 and a non-specific band of 70 kDa when using the McPIP1;2 antibody on Western blots of isolated X. leavis oocytes plasma membrane when expressing this AQP (Supporting Information Fig. S2a), which matches the mass of one of the proteins detected on the 2D gel (Fig. 3d).

Figure 3.

Analysis of antibody specificity to the aquaporins. Mesembryanthemum crystallinum microsomal membranes from leaves (a) and roots (b) were separated by two-dimensional polyacrylamide gel electrophoresis (PAGE), and Western blot analysis was carried out using peptide-specific antibodies against McPIP1;4 (c), McPIP1;2 (d), McTIP1;2 (e), McPIP1;5 (f) and McPIP2;1 (g). Microsomal protein was isolated by centrifugation and separated in the first dimension by isoelectric focusing (IEF) followed by sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS-PAGE) in the second dimension. Protein from leaves (c, d, e and f) and roots (g) was then blotted onto ECL-nitrocellulose membranes. The peptides recognized by each of the antibodies correspond to 40 kDa (McPIP1;4), 30 kDa (McPIP1;5), 31 kDa (McPIP2;1), 45, 34, 25, 15 kDa (McPIP1;2) and 34 kDa (McTIP1;2). Blots are representative of three independent experiments.

AQP abundance during day/night cycle in C3 and CAM plants

In the leaves (Fig. 4) and roots (Fig. 5) of young, C3 plants, no evidence of day/night regulation of AQPs was detected with all proteins exhibiting a similar level of abundance throughout the day/night cycle. However, in leaves of adult CAM plants, differential regulation was observed for four of the five AQPs under study (Fig. 4). Leaf abundance of McPIP1;4 was low during the light period, with an increase in abundance observed towards the end of the day, which peaked at the end of the light period, and into the beginning of the dark period; protein levels were again low at 0700 h the following morning (Fig. 4). McPIP1;5 in CAM plant leaves was not detectable during the beginning of the light period up to 1500 h. At 1500 h the protein appeared and increased in abundance with maximal amounts observed between 1700 and 1900 h. Protein levels then decreased following this short peak, and by 2100 h, protein was again barely detectable (Fig. 4). McPIP1;2 (34 kDa protein) appeared to be unregulated by both metabolic status of the plant and day/night cycle as abundance in leaves was unchanged throughout the 24 h period in CAM plants similar to that of C3 plants (Fig. 4). It is important to note that none of the other three proteins recognized by the McPIP1;2 antibody showed changes in abundance over the day/night cycle (data not shown). McTIP1;2 abundance in leaf tissue of CAM plants was high at the onset of the light period but rapidly declined, with low levels maintained throughout the day until 1700 h, where levels were seen to increase again into the dark period (Fig. 4).

Figure 4.

Expression analysis of aquaporins (AQPs) during the day/night cycle in Mesembryanthemum crystallinum leaf tissue. Membrane protein fractions and RNA were isolated at 2 h intervals over a period of 24 h from C3 juvenile and adult Crassulacean acid metabolism (CAM) untreated or salt-treated (200 mm NaCl for 2 weeks) leaf tissue. McPIP1;4 (intracellular compartment), McPIP1;2 (plasma membrane), McTIP1;2 (tonoplast) and McPIP1;5 (plasma membrane) were detected in leaves. The membrane fractions were isolated on discontinuous sucrose density gradients at the 0/22% and 34/38% sucrose interface, for tonoplast/intracellular compartment and plasma membrane, respectively. Western blot analysis and RT-PCR were carried out as described in the Materials and Methods section using peptide-specific antibodies directed against the different AQPs and gene-specific primers. Blots and photographs are representative of three independent experiments. AQP transcript accumulation was normalized by dividing mean relative values by the mean relative McUBQ values (Supporting Information Fig. S2b). Protein abundance mean intensity is the average of three independent blots.

Figure 5.

Expression analysis of aquaporins (AQPs) during the day/night cycle in Mesembryanthemum crystallinum root tissue. Membrane fractions and RNA were isolated at 2 h intervals over a period of 24 h from C3 juvenile and adult Crassulacean acid metabolism (CAM) untreated or salt-treated (200 mm NaCl for 2 weeks) root tissue. McPIP1;4 (intracellular compartment), McPIP1;2 (plasma membrane), McTIP1;2 (tonoplast) and McPIP2;1 (plasma membrane) were detected in roots. The membrane fractions were isolated on a discontinuous sucrose density gradient at the 0/22% and 34/38% sucrose interface, for tonoplast/intracellular compartment and plasma membrane, respectively. Western blot analysis and RT-PCR were carried out as described in the Materials and Methods section using peptide-specific antibodies directed against the different AQPs. Blots and photographs are representative of three independent experiments. AQP transcript accumulation was normalized by dividing mean relative values by the mean relative McUBQ values (Supporting Information Fig. S2b). Protein abundance mean intensity is the average of three independent blots.

The levels of McPIP1;4 in root tissue of CAM plants also exhibited a day/night regulation, but in this tissue, peak protein abundance was observed earlier in the day between 1300 and 1900 h (Fig. 5). At the onset of the dark period and early in the light period, protein was not detectable (Fig. 5). We did not detect McPIP1;5 in root tissue under any of the conditions studied (Supporting Information Fig. S2d). As observed in leaves, the levels of McPIP1;2 were unchanged throughout the 24 h period in CAM plants, similar to that of C3 plants (Fig. 5). Root amounts of McTIP1;2 in CAM plants showed a sharp peak of protein abundance at 1900 h and otherwise low levels throughout the day/night cycle (Fig. 5). The root-specific McPIP2;1 in CAM plants was observed until 1300 h and showed increasing abundance during the day, reaching a peak between 1700 and 1900 h and dropping to undetectable levels during the first hours of the night period (Fig. 5).

In order to confirm that the changes in AQP abundance were because of day/night regulation during the CAM cycle, CAM plants were placed under conditions of continuous darkness for 48 h, and samples were taken after the 48 h period at the same time as for the day/night experiments. Under this condition, the AQPs maintained the same patterns of protein abundance changes as those observed in CAM plants grown under normal photoperiod (Supporting Information Fig. S3).

Effect of salinity on the abundance of AQPs in CAM plants during the day/night cycle

Previously, we have reported that the abundance of several AQPs in adult M. crystallinum plants is differentially regulated by salt. In leaves of salt-treated plants, McTIP1;2 was down-regulated, while that of McPIP1;4 was up-regulated, and McPIP1;2 abundance was not affected (Vera-Estrella et al. 2000). In roots, the abundance of McPIP2;1, McTIP1;2 and McPIP1;4 was up-regulated by salt while the abundance of McPIP1;2 was not affected (Kirch et al. 2000). In these earlier studies, plants were collected at the beginning of the light period. Here, we examined the day/night changes in abundance and expression of these AQPs in CAM plants in the presence or absence of 200 mm NaCl for 2 weeks. In leaves of salt-treated CAM plants, McPIP1;4 protein levels were seen to be high during the 0900–2100 h period with no protein detected at 0700 h, in contrast with the peak of protein abundance that was observed in CAM untreated plants at the beginning of the dark period (Fig. 4). However, in roots, the abundance and regulation of McPIP1;4 was similar between CAM untreated and CAM salt-treated plants (Fig. 5). The abundance of McPIP1;2 in leaves (Fig. 4) and roots (Fig. 5) was not modified by salt. Salt treatment drastically reduced the amounts of McTIP1;2 in leaves throughout the day/night cycle (Fig. 4), whereas roots showed a gradual increase in abundance of McTIP1;2 through the day, reaching a peak at the beginning of the night period, similar to that of untreated CAM plants (Fig. 5). A slight increase in the leaf-specific McPIP1;5 peak protein abundance was observed between control and salt-treated CAM plants, as well as a shift in the peak by 2 h to an earlier time (Fig. 4). Changes in abundance of the root-specific AQP McPIP2;1 in CAM salt-treated plants was different from that of non-treated CAM plants as the increases in protein abundance occurred 2 h earlier, at 1100 h, with maximal protein levels observed at 2100 h (Fig. 5). These results support the view that individual AQPs have specific roles within the plant cell, with each contributing independently to the water status of the plant cell at defined locations and periods and under specific stress conditions.

AQP transcript accumulation during the day/night cycle in C3, CAM and CAM-salt plants

Transcript accumulation profiles for the M. crystallinum AQP genes under study were obtained from leaves and roots of C3, CAM and salt-treated CAM plants by using semiquantitative RT-PCR of RNA isolated at different time points of the day/night cycle. Transcript expression of McUBQ was used as a constitutive control, and all data were normalized to the McUBQ signal (Supporting Information Fig. S2b). As an indicator for CAM induction/metabolism, we monitored the transcript accumulation of McPPC1 (Supporting Information Fig. S2c). Analysis of McPIP1;4, McPIP1;2, McPIP1;5, McPIP2;1 and McTIP1;2 genes in C3, CAM and salt-treated CAM plants showed diverse and independent changes in transcript accumulation. In leaves, transcript levels of McPIP1;4 showed a significant reduction between 1900 and 2100 h in all conditions, the only difference being that in CAM plants, the reduced accumulation was maintained throughout the dark period to the start of the following light period (0700 h) (Fig. 4). In roots, McPIP1;4 maintained stable levels in both C3 and CAM plants throughout the day/night period, with salt-treated CAM plants showing a transient reduction in transcript levels at 2100 h (Fig. 5). Expression levels of McPIP1;2 in leaves showed no change in either C3 or salt-treated CAM plants, while in CAM plants, the expression was minimal at 0700 h, increasing to a plateau at 0900 h where it was maintained throughout the subsequent day/night period to decrease again at 0700 h the following day (Fig. 4). In roots, the levels of expression of McPIP1;2 were maintained throughout the day, with a slight decrease observed at 2100 h in C3 plants (Fig. 5). The levels of expression of McTIP1;2 in leaf and root tissue were maintained throughout the day/night cycle, with only a small dip in expression observed at 2100 h in roots from C3 plants (Figs 4 & 5). A similar small dip in expression was observed in leaves of C3 plants for McPIP1;5 at 2100 h, but otherwise, little change in expression throughout the day/night cycle in CAM or salt-treated CAM plants was observed for this AQP (Fig. 4). McPIP2;1 transcript accumulation was unchanged throughout the day/night cycle in the roots (Fig. 5).

CAM responses to day/night cycle

To enable us to relate the changes observed in AQP protein amount and membrane water permeability to specific events occurring during the CAM cycle, we studied one of the key enzymes in CAM, PEPc. In vivo, circadian regulation of PEPc occurs through post-translational changes in phosphorylation state and allosteric regulation by malate (Nimmo 2000). However, it has also been shown to present day/night variations in activity when measured in vitro (Scheible et al. 2000), which corresponds closely to alterations in transcript levels (Häusler et al. 2000; Scheible et al. 2000; Borland et al. 2006; Cushman et al. 2008). We also followed the abundance of inositol methyl transferase (IMT), an enzyme in the pinitol biosynthesis pathway that, like PEPc, follows a day/night cycle of expression (Ishitami et al. 1996; Nelson, Rammesmayer & Bohnert 1998), and we measured the levels of pinitol at similar time points. In addition, at 2 h intervals over a 24 h period, we measured cell sap osmolarity and the levels of glucose in C3 and CAM plants to determine their precise pattern of regulation during the CAM cycle.

As expected for C3 plants, PEPc activity in total protein extracts remained at similar levels throughout the day/night cycle (Fig. 6a), while in CAM plants, the classic day/night rhythm was observed with low levels of activity measured during the middle of the light period and activities similar to those observed at 0900 and 1100 h during the dark period (Fig. 6b, see Scheible et al. 2000). The PEPc activity pattern in CAM plants exposed to continuous darkness was similar to that observed in CAM plants grown under a normal day/night cycle (Fig. 6c), confirming the anticipation and persistence of the oscillations (Wilkins 1992). The higher absolute values of PEPc activity measured in the dark-grown plants may be attributable to the difference in CO2 and humidity of the growth chamber to which the plants were moved for the dark treatment, as compared with those conditions encountered in the green house.

Figure 6.

Phosphoenolpyruvate carboxylase (PEPc) activity in Mesembryanthemum crystallinum plants. PEPc activity was measured in total protein extracts isolated at 2 h intervals throughout the day/night cycle from leaves of (a) C3 plants, (b) untreated adult Crassulacean acid metabolism (CAM) plants and (c) adult CAM plants exposed to continuous darkness for 48 h (CAM-dark). Grey bars represent the light period and black bars represent the dark period. Values for PEPc activity are means of three independent experiments with standard deviation (SD) not exceeding 10%.

IMT abundance and pinitol levels also showed a circadian rhythm exclusive to CAM plants (Fig. 7a), although opposite of that observed for PEPc activity (Fig. 6b). In CAM plants, highest IMT protein levels and pinitol amounts (10.7 µmol g−1 FW) were detected during the middle of the light period, while the lowest levels of pinitol (7.9 µmoles) were detected at the beginning and end of the dark period (Fig. 7a, CAM). Salt-treated CAM plants showed the same pattern of changes in pinitol levels (Supporting Information Fig. S4a). This polyol functions to restore the osmotic potential of the cytoplasm to drive water uptake for turgor maintenance (Stoop, Williamson & Pharr 1996). The levels of glucose and fructose, additional sugars that may also be involved in balancing changes in cellular osmotic potential during the day/night CAM cycle, were also investigated (Fig. 7b,c). Similar to pinitol, the amounts of glucose (70 nmol g−1 FW) and fructose (40 nmol g−1 FW) remained constant in juvenile C3 plants over a 24 h photoperiod (Fig. 7b,c, C3), whereas, in CAM plants, the amount of glucose and fructose increased from 62.8 and 15.5 nmol g−1 FW at 9 h to a maximum level of 222.4 and 238.0 nmol g−1 FW detected at 1700 h, respectively (Fig. 7b,c, CAM). Similar results were found when CAM plants were salt-treated (Supporting Information Fig. S4c,d).

Figure 7.

Quantitation of inositol methyl transferase (IMT) protein levels and pinitol, glucose and fructose amounts in cell sap from leaves of Mesembryanthemum crystallinum plants during the day/night cycle. Pinitol (a), glucose (b) and fructose (c) were measured in leaves of juvenile C3 and adult Crassulacean acid metabolism (CAM) plants at 2 h intervals throughout the day night cycle. Grey bars represent the light period and black bars represent the dark period. Values are expressed as means of three independent experiments with standard deviation (SD) not exceeding 10%. IMT abundance in total protein was detected using a polyclonal antibody at a 1/2000 dilution against the 40 kDa protein (a, top panels). Blots are representative of three independent experiments.

These results suggested that during CAM changes in sugar biosynthesis may lead to changes in cellular osmotic potential, and this appears to occur independently of whether the plant is salt-treated or not. To determine if the leaf water potential changed during the day/night cycle, the osmolarity of the cell sap was measured in C3 and CAM plants. While the cell sap osmolarity from C3 plants was not altered (Fig. 8a, C3), CAM plants exhibited a rise in osmolarity at the beginning of the light period that reached a maximum value at 1500 h, which was followed by a decrease in osmolarity into the dark period (Fig. 8a, CAM). This pattern matched closely to that observed for levels of pinitol, glucose and fructose (Fig. 7). The cell sap osmolarity of salt-treated CAM plants was similar to those observed in CAM plants (Fig. 8a, CAM-S). In contrast, cell sap osmolarity in roots was not affected by CAM in either untreated (Supporting Information Fig. S5a) or salt-treated plants (Supporting Information Fig. S5b).

Figure 8.

Osmolarity, pH and malate determination in cell sap from leaves of Mesembryanthemum crystallinum plants during the day/night cycle. (a) Osmolarity was measured in leaves of juvenile C3 (C3), adult Crassulacean acid metabolism (CAM) untreated and salt-treated (CAM-S) plants. (b) Malate (bar graphs) and pH (scatter plot) were measured in leaves of juvenile C3 (C3) and adult untreated CAM plants at 2 h intervals throughout the day/night cycle. Grey bars represent the light period and black bars represent the dark period. Values are expressed as means of three independent experiments with standard deviation (SD) not exceeding 10%.

CAM plants also showed the classic day/night changes in leaf cell sap pH, directly related to the malic acid content of the leaf shown in Fig. 8b (CAM), while C3 plants showed no changes in leaf cell sap pH or L-malic acid content (Fig. 8b, C3). Leaves of salt-treated CAM plants showed similar results as those observed in untreated CAM plants for cell sap pH and malic acid content (Supporting Information Fig. S4b).


The daily carbon cycling in CAM plants requires tight regulation of water transport between cellular compartments and may imply that water permeability of cellular membranes will vary with the time of day, depending on the requirement for osmotic adjustment in a manner different from that which is observed in C3 plants.

Evidence for this difference was provided from measurements of swelling rates in protoplasts isolated from CAM plants compared with C3 plants. As expected, in C3 plants, which do not exhibit either large daily fluctuations in cellular osmolarity (Fig. 8a), or alterations in pinitol, glucose and fructose accumulation (Fig. 7a–c), plasma membrane water permeability (Pf) of protoplasts isolated from leaf and root tissue remained constant throughout the day/night cycle (Fig. 2g,h). In contrast, protoplasts from CAM plants showed marked differences depending on the time point during the day in which the protoplasts were isolated, with cellular Pf being highest at the end of the day (Fig. 2g,h). These CAM-specific changes in swelling rate and water permeability can be correlated with changes in abundance of distinct AQPs in defined compartments within the cell, which would function to meet the cell's changing demands for water flow at defined times during day/night cycle (Figs 4 & 5). M. crystallinum plants responded to CAM by altering the protein abundance pattern of several AQPs, including McPIP1;5 in the plasma membrane of leaves, McPIP1;4 in an unidentified compartment of leaves and roots that co-migrates with the tonoplast under specific fractionation conditions; McTIP1;2 in the tonoplast of leaves and roots and McPIP2;1 in the plasma membrane of roots (Figs 4 & 5).

Changes in abundance pattern were unique to CAM plants as juvenile plants undergoing C3 photosynthesis showed no such day/night regulation of any of the AQPs under study. These CAM-induced changes in protein abundance were altered upon imposition of salt-stress to adult CAM plants with McTIP1;2 showing the most dramatic response in the leaves where a clear decrease in abundance was induced by salt (Fig. 4). In some cases, salt-treatment not only altered the abundance of the protein but also drastically affected the day/night rhythm of protein abundance (Fig. 4). In particular, McPIP1;4 under salt-treatment was abundant during the daylight hours between 0700 and 1900 h but was undetectable at 0700 h. In CAM plants, in the absence of salt stress, the protein showed low abundance during the morning and into the early afternoon (0700–1500 h) followed by a transient peak of abundance at the end of the light period (1700–2100 h). Only one of the AQPs under study, McPIP1;2 found in the plasma membrane of both roots and leaves, showed no alteration in the amount of protein throughout the day/night cycle and maintained its constitutive high levels regardless of the plant's age, metabolic state or stress exposure, suggesting that this protein plays an important role in the basal metabolism of the plant (Figs 4 & 5). These results highlight the complexity of AQP responses occurring in CAM plants under salt stress and suggest a tight regulation of water transport. Furthermore, they imply that salt stress induces non-CAM-related changes in the abundance or expression of some AQPs, independent of those observed in CAM plants, indicating a role in leaves for the AQPs McPIP1;4 and McTIP1;2 that responded differentially to CAM and high salt treatment.

Of the two leaf plasma membrane localized PIPs investigated in this study, only one, McPIP1;5, showed strong day/night variations in protein abundance in CAM plants, presenting the possibility that this AQP is directly involved in osmotic adjustments during the light/dark cycle and may be responsible for the changes recorded in Pf in protoplasts from CAM plants. This proposal is strengthened by the observation that levels of McPIP1;5 remained unchanged and Pf values were constant in C3 plants (Fig. 4). However, we cannot rule out the role of other yet unidentified PIPs or the possible regulation of AQPs by post-translational modifications, which would directly affect the activity of existing AQPs (reviewed in Maurel et al. 2008). The observed peak in Pf and McPIP1;5 abundance also matched the peak of glucose, fructose, pinitol and IMT expression measured in leaves of CAM plants (Fig. 7, CAM), suggesting that water flow through McPIP1;5 is driven passively by the increase in osmolytes linked to the cycling of carbon in the cells.

The peak of protein abundance for McTIP1;2 and McPIP1;4, which are expressed in intracellular membranes of the leaf in CAM plants, occurred 2 h later than the peak observed for McPIP1;5, at the beginning of the dark period (Fig. 4). In the case of McPIP1;4, this peak appeared to be transient, in contrast to the peak for McTIP1;2 which was sustained throughout the dark period until the beginning of the light period. It is possible that the observed differences in AQP protein abundance at a particular time in a particular subcellular compartment during the day/night cycle may be one of the main mechanisms that the cell uses for osmotic adjustment in response to the metabolic changes that occur in the plants during CAM, as observed by the dynamic changes in levels of pinitol, glucose and fructose (Fig. 7, CAM). However, without direct measurements of the membrane permeability of the compartments in which these intracellular AQPs reside, it is difficult to assign a role for these proteins in CAM or salt-stress.

Plant root AQPs also appear to have a role in cellular day/night water relations in CAM plants with three of the four root-expressed AQPs showing day/night changes in protein abundance (Fig. 5), albeit all with unique rhythms, that were absent in juvenile C3 plants. With the exception of McPIP1;4, the application of salt stress to CAM plants further affected the day/night changes in the abundance of AQPs in the root (Fig. 5) but did not result in such large changes as was observed for the leaf AQPs (Fig. 4). The peak of protein accumulation in the root of McPIP2;1 (Fig. 5) followed well the increases in plasma membrane water permeability observed in protoplasts isolated from roots of CAM plants (Fig. 2h), suggesting that this AQP may participate in the absorption of water during the CAM cycle. These changes in abundance of AQPs in different membranes and tissues, at distinct times during the CAM cycle, highlight both the requirement to control osmotic potential in specific cellular compartments in phase with metabolic partitioning, and the requirement for regulated control of water flow throughout the plant.

Analysis of transcript accumulation and comparison with protein abundance showed a large discrepancy between the two, suggesting that there is little coordination between the time of AQP gene expression and when its translation product is required by the cell. This suggests that post-transcriptional mechanisms involved in mRNA stability/degradation may play a large role in regulating protein levels. The ability to maintain transcript abundance at sufficient levels would allow the rapid production of proteins when they are required, thus eliminating delays required for transcription. This level of regulation, along with post-translational mechanisms such as phosphorylation and protonation (Chaumont, Moshelion & Daniels 2005; Törnroth-Horsefield et al. 2006; Maurel et al. 2008; Prak et al. 2008; Van Wilder et al. 2008; Amezcua-Romero et al. 2010) would allow for the rapid and precise control of water movement via AQPs, which appears to be required during specific stages of the CAM cycle and in response to salt stress.


We thank Guadalupe Muñoz for her technical support. This research was funded by DGAPA grant IN221308 to R.V.-E., CONACYT grants 57685 to R.V.-E. and 49735 to B.J.B., and FOMIX Morelos-CONACyT grant to O.P.