Osmoregulatory Isoform of Dihydroxyacetone Phosphate Reductase from Dunaliella tertiolecta: Purification and Characterization

Protein Expression and Purification 24, 404–411 (2002) doi:10.1006/prep.2001.1588, available online at http://www.idealibrary.com on

Osmoregulatory Isoform of Dihydroxyacetone Phosphate Reductase from Dunaliella tertiolecta: Purification and Characterization1 Durba Ghoshal, David Mach, Mamta Agarwal, Archana Goyal, and Arun Goyal Department of Biology, College of Science and Engineering; Department of Biochemistry and Molecular Biology, School of Medicine; and UMD Center for Cell and Molecular Biology, University of Minnesota Duluth, Duluth, Minnesota 55812

Received August 21, 2001, and in revised form November 19, 2001

The osmoregulatory isoform of dihydroxyacetone phosphate (DHAP) reductase (Osm-DHAPR) is an enzyme unique to Dunaliella, photosynthetic unicellular green algae adapted to extreme environments. This is the first report of purification of an isoform of DHAP reductase from Dunaliella, specifically the osmoregulatory isoform that is involved in the synthesis of free glycerol for osmoregulation in extreme environments, such as high salinity. The Osm-DHAPR is cold labile, inactivated by ammonium sulfate, forms a strong complex with Rubisco, and is unstable in the absence of glycerol. These difficulties have been addressed, and a four-step procedure has been developed to purify the Osm-DHAPR from Dunaliella tertiolecta: precipitation of Rubisco by polyethylene glycol, followed by successive chromatography on DEAE cellulose, Sephacryl S200, and Red Agarose. Yield of the purified enzyme was 3.6%, with a specific activity of 938 ␮mol ⴢ minⴚ1 ⴢ mgⴚ1 of protein and a subunit molecular mass of approximately 38 kDa. A maximum specific activity of 2580 ␮mol ⴢ minⴚ1 ⴢ mgⴚ1 of protein could be achieved by assay with 150 mM NaCl. The Osm-DHAPR had little preference for NADH or NADPH, but it is highly specific for DHAP. Other metabolites of glycolysis, the tricarboxylic acid cycle, and the C3 reductive photosynthetic carbon cycle were not reduced by the enzyme. The purified enzyme was stimulated three-fold by 150 to 250 mM NaCl/KCl and by 25 mM MgCl2. Detergents, lipids, or long-chain acyl CoA derivatives, all of 1 This work was supported in part by Pioneer Hi Bred. M.A. was supported by Graduate School Grants in Aid of the University of Minnesota to A.G. D.M. was supported by UROP, University of Minnesota.

404

which inhibited the chloroplastic glyceride form of DHAP reductase, did not affect the activity of OsmDHAPR. The Osm-DHAPR has different properties than the other chloroplastic isoform of DHAP reductase from plants and algae for glycerol phosphate formation and triglyceride synthesis. 䉷 2002 Elsevier Science (USA)

Dihydroxyacetone phosphate (DHAP)2 is the first triose phosphate synthesized during photosynthetic carbon fixation in the chloroplast. The reduction of DHAP to glycerol phosphate (glycerol-P) is catalyzed by dihydroxyacetone phosphate reductase (DHAP reductase), with nomenclature in the reverse reaction to that of glycerol-3P:NAD+ oxidoreductase (EC 1.1.1.8), which initiates the pathway for glycerol, triglyceride, and lipid synthesis. Isoforms of DHAP reductase have been characterized from spinach leaves (4, 5, 18), seeds of Brassica campestris (24), Chlamydomonas reinhardtii (19), and Dunaliella tertiolecta (7–9). Additional isoforms of DHAP reductase of uncharacterized function are present in leaf peroxisomes (18) and in mitochondria (Goyal, unpublished). Dunaliella, a photosynthetic organism from extreme environments, has the remarkable characteristic of surviving large osmotic stresses, largely by virtue of the intracellular accumulation of free glycerol. Glycerol is 2 Abbreviations used: DHAP, dihydroxyacetone phosphate; glycerol-P, glycerol phosphate; Osm-DHAPR, the chloroplast osmoregulatory form of DHAP reductase; Pl, preinoculum; I, inoculum; PEG, polyethylene glycol.

1046-5928/02 $35.00 䉷 2002 Elsevier Science (USA) All rights reserved.

Osm-DHAP REDUCTASE FROM Dunaliella

metabolized by two distinct reaction sequences to and from DHAP (1, 12). For glycerol synthesis, the reaction sequence involves DHAP reductase (7–9, 15, 16) and glycerol-P phosphatase (11, 27), and the dissimilation sequence is catalyzed by glycerol dehydrogenase (2) and dihydroxyacetone kinase (20). Previous investigations of the glycerol metabolism in Dunaliella for osmoregulation were based on DHAP reductase activity in the chloroplast (3, 10, 11, 15, 21). We have identified and partially characterized three isoforms of DHAP reductases from D. tertiolecta (8). Two major forms were localized in the chloroplast and a minor form was in the cytosol (7, 8). The major chloroplast isoform is stimulated by NaCl and appears to be specifically involved in the synthesis of glycerol (8). During active growth, the second DHAP reductase in the chloroplast is also active and has properties similar to those of the chloroplastic isoform from higher plants. The cytosolic isoform has characteristics similar to those of spinach leaf cytosolic reductase (7). Based on the elution profile, localization, and characteristics of different isoforms from the leaves of higher plants and from Dunaliella, isoforms of DHAP reductase are designated as follows: (1) the chloroplast osmoregulatory form (Osm-DHAPR) is the isoform from Dunaliella chloroplast that elutes first from the DEAE cellulose column, is stimulated by salts, and is not regulated like the higher plant chloroplast DHAP reductase; (2) the chloroplast glyceride form is the isoform from Dunaliella during active growth that is eluted second from the DEAE cellulose column, and is inhibited by detergents, lipids, long-chain acyl CoA derivatives, and NaCl regardless of whether it was isolated from chloroplasts of Dunaliella or spinach. Consequently higher plant chloroplastic DHAP-reductase is also a “chloroplastic glyceride form”; (3) the cytosolic glyceride form is the third minor isoform from the cytosol of Dunaliella or leaves of higher plants. The cytosolic isoform from higher plants has been purified to homogeneity (18, 24); however, other isoforms have not. In this paper, we report the purification and characterization of the unique osmoregulatory isoform of DHAP reductase (Osm-DHAPR) from D. tertiolecta. MATERIALS AND METHODS Chemicals. Most of the chemicals were purchased from Sigma-Aldrich. DHAP was prepared by hydrolyzing the dimethylketal dimonocyclohexylamine salt with Dowex 50H+ as directed by Sigma and frozen in aliquots until use. DE 52 (DEAE cellulose) was from Whatman, England; Sephacryl S-200 was from Pharmacia; and the dye-ligand agarose (Matrex gel Red Agarose) was from Amicon Corporation. The broad range protein molecular weight marker and the protein assay reagent were purchased from Bio-Rad, U.S.A.

405

Organism and growth conditions. D. tertiolecta (Commonwealth Scientific and Industrial Research Organization Marine Laboratory, Hobart, Australia) was maintained on agar slants containing rich, synthetic “Dunaliella medium” with 0.17 M NaCl as described by Johnson et al. (17) and modified by Goyal et al. (11). Dunaliella cells were adapted to grow at higher salt concentrations (1 M) by gradually increasing the NaCl concentration in the Dunaliella growth media as follows: (1) preinoculum (PI) was seeded from slants and grown for 1 week in continuous light (80 ␮mol ⭈ m⫺2 ⭈ s⫺1) by inoculating 50 mL of growth media containing 0.17 M NaCl; (2) inoculum (I) (100 mL) was seeded with PI (10 mL) in growth medium containing 0.5 M NaCl and grown for 1 week under the same conditions as PI; (3) the experimental cultures were inoculated with I and grown with 1.0 M NaCl for 5 days with continuous shaking in light (150 ␮mol ⭈ m⫺2 ⭈ s⫺1) on a 16:8 h light:dark cycle while bubbling with 5% CO2 in air (v/v). Before harvesting, cultures were examined microscopically for bacterial or fungal contamination. Enzyme purification. The pH of buffers was adjusted at room temperature. Unless otherwise indicated, all steps of the purification were performed at 4⬚C, and all gravity values refer to the maximum radius of the rotors. After 5 days, when cells were in the late log phase or early stationary phase of growth, algae (6-L culture) were harvested by centrifugation at 2530g for 2 min in a Sorvall RT-7 centrifuge using RTH-750 rotor. The algal pellet was washed with fresh iso-osmotic medium, and centrifugation was repeated. The washed cell pellet was resuspended in 100 mL of TDG buffer (100 mM Tris, pH 6.9, 1 mM DTT, 2.5% (v/v) glycerol). D. tertiolecta cells were broken under high pressure in a Yeda press (1500 psi of nitrogen); after 3 min the cell suspension was collected, and the pressate was subjected to the Yeda Press step once again. The resulting homogenate was diluted five-fold by directly collecting in TDG buffer. The homogenate was centrifuged at 40,000g for 30 min and the supernatant was collected (crude extract). Polyethylene glycol (PEG) was added (using a 60% (w/v) stock solution prepared in TDG buffer) to the crude extract slowly, with continuous stirring to a final concentration of 13% (w/v). After the PEG treatment, the resultant suspension was centrifuged at 40,000g for 30 min and the supernatant was collected. It is important that the stock solution of PEG is kept at 50⬚C before use; otherwise it precipitates in cold. The supernatant from the PEG step was mixed with a DE 52 ion exchanger (DEAE cellulose) preequilibrated with TDG buffer for 2 h. The DE 52 saturated with proteins was packed in a column (2.5 ⫻ 25 cm) and washed with TDG buffer (10 to 15 column volumes until A280 was negligible). The bound protein was eluted with a linear

406

GHOSHAL ET AL.

gradient of 0 to 0.4 M NaCl in TDG buffer (200 mL each). All eluted fractions were assayed for DHAP reductase activity in the absence and the presence of 150 mM NaCl. Salt-stimulated fractions were pooled, desalted, and concentrated by Filtron-30 according to the protocols supplied by the manufacturer (Pal Filtron, U.S.A.). The concentrated protein was applied to a TDG-equilibrated Sephacryl S-200 column (2.5 ⫻ 90 cm), and active fractions were eluted with TDG buffer. Salt concentrations in the eluted fractions were adjusted to 0.5 M NaCl before applying to a Red Agarose column (3 mL) preequilibrated with TDG buffer containing 0.5 M NaCl. The Red Agarose column was washed with the same buffer (10 column volumes), active fractions were eluted with 5 vol of TDG buffer containing 0.5 M NaCl and 5 mM NADH, and purification was followed by SDS– PAGE. Enzyme assay. DHAP reductase reduces DHAP to glycerol-P using NAD(P)H as cofactor. One unit enzyme of activity is defined as the amount of enzyme required to convert 1 ␮mol of NAD(P)H to 1 ␮mol of NAD(P) in 1 min at 25⬚C at pH 6.9. The DHAP reductase activity was assayed spectrophotometrically at room temperature by monitoring the decrease in absorbance of NAD(P)H at 340 nm with time. The assay mixture (1 mL) contained 0.2 mM NAD(P)H and 1 mM DHAP in 33 mM 2-(N-morpholino)ethanesulfonic acid (Mes), 33 mM N-2-hydroxyethylpiperazine-N ⬘-2-ethanesulfonic acid (Hepes), 34 mM tricine buffer at pH 6.9, and the reaction was started by adding the enzyme. When assay was performed in the presence of salt, 150 mM NaCl was included in the assay. Quantitative determinations. Soluble protein was estimated by Bio–Rad protein assay reagent using BSA as standard, and the total chlorophyll was estimated at A652 after extraction with 90% (v/v) ethanol (8). RESULTS

relatively poor growth at 2.0 M NaCl, enzyme was not isolated from the cultures grown at that salt concentration. A more accurate value for the total DHAP reductase activities is the sum of all active fractions after DEAE cellulose chromatography, even though some inactivation from the cold temperature and passage of time may have occurred. When algae were grown with 0.17 M NaCl, addition of NaCl in the assay did not stimulate enzyme activity. However, salt stimulation of activity was apparent when the NaCl concentration in the growth medium was gradually increased to 1 or 1.5 M (Table 1). These results were consistent with our previous observation that the osmoregulatory isoform was activated and the glyceride forms were inhibited by 150 mM NaCl in the assay mixture (8, 11). The possible reason for not observing the salt effect at 0.17 M NaCl was that the salt inhibitory effect of the glyceride form nullified the salt stimulatory effect of the osmoregulatory isoform when the two forms are present in about equal amounts. The salt stimulation of total activity in the crude extract was further resolved into three isoforms by DE 52 ionexchange chromatography. The results indicate that at higher salt concentrations in the growth media (0.5 to 1.5 M NaCl), the osmoregulatory isoform contributes to almost 80% of the total activity. The salt stimulation of total activity in the case of cells grown with higher salt was due to a significant increase of the osmoregulatory form compared to the glyceride form (Table 1). Growth of D. tertiolecta was significantly impaired at 1.5 M (Table 1) and 2.0 M NaCl (data not shown) in the medium, as judged by total chlorophyll. Although with growth at 1.5 M NaCl, the enzyme activity showed salt stimulation and an elution profile similar to growth at 1 M NaCl, the total enzyme activity was significantly less in the crude extract. Therefore, for purifying OsmDHAPR, cells were grown in a growth medium containing 1 M NaCl.

Effect of NaCl in the growth media on the activity of Osm-DHAPR. Previous investigations on DHAP reductase from Dunaliella indicated the presence of two major isoforms in the chloroplasts and a minor cytosolic form (8). In the past, minimal medium with salt has been used to grow experimental cultures; however, the total amount of osmoregulatory isoform of the enzyme was found to be significantly higher if cultures were grown in a rich synthetic Dunaliella growth medium (Table 1). Therefore, before proceeding to purify the Osm-DHAPR, the optimum salt concentration to yield the highest recovery of the enzyme was determined. The amount of NaCl in the growth medium was varied between 0.17 and 2.0 M, and the total activity in the crude extract and three isoforms resolved on an ionexchange column were assayed (Table 1). Because of

Purification of Osm-DHAPR. Purification of the Osm-DHAPR from D. tertiolecta required resolution of several unusual difficulties and challenges. After homogenization of cells, ammonium sulfate fractionation of the crude extract resulted in up to a 90% loss of enzyme activity. Rubisco was strongly associated with DHAP reductase activity throughout the steps of conventional protein purification, and the DHAP reductase activity was extremely sensitive to cold conditions that resulted in continued loss of activity. Inclusion of DTT and glycerol stabilized enzyme activity. The recovery of the total enzyme activity presented in the progress of purification of Osm-DHAPR could have been higher if the enzyme was not sensitive to cold and was not continuously being inactivated.

407

Osm-DHAP REDUCTASE FROM Dunaliella

TABLE 1 Effect of NaCl Concentration in the Growth Media on DHAP Reductase Activity in Dunaliella tertiolecta Parameter NaCl in the media (M) Total chlorophyll in preparation (mg) Total DHAP reductase activity [␮mol ⭈ min⫺1] Crude extract (⫺NaCl) Crude extract (⫹NaCl) DE 52 (DEAE cellulose) column Chloroplastic osmoregulatory form Chloroplastic glyceride form Cytosolic glyceride form

0.17 20.5 225 220 270 105 155 10

0.5 19.8 210 295 325 235 85 5

(100) (38) (58) (4)

(100) (72) (26) (2)

1.0 24.5 250 510 450 360 80 10

1.5 10

(100) (80) (18) (2)

78 165 155 120 30 5

(100) (77) (19) (3)

Note. After 5 days of algal culture (2 L) growth, activity was measured in the crude extract and in three peaks eluted from the DEAE cellulose column. The activity was assayed with ⫾150 mM NaCl for identifying fractions containing the osmoregulatory isoform (the osmoregulatory isoform was stimulated by 150 mM NaCl, whereas the glyceride forms was inhibited). The data in parentheses are the percentages of activity of three isoforms (100% is the total activity of all isoforms eluted from the DEAE cellulose column). The results represent the total activity (␮mol ⭈ min⫺1) without added NaCl in the assay. The data under DE 52 cellulose represent a sum of activities of all isoforms eluted from the DEAE cellulose column.

The purification of Osm-DHAPR to apparent electrophoretic homogeneity is summarized in Table 2. Rubisco, the major stromal protein, contributes to approximately 60% of the total protein of D. tertiolecta. In a cell-free extract, Rubisco forms a strong complex with DHAP reductases (unpublished observation). However, if the protein concentration in the homogenate was maintained around 1 mg ⭈ mL⫺1, aggregation was minimal. It was crucial to remove a significant amount of Rubisco from the crude extract; otherwise, it was impossible to dissociate Rubisco from DHAP reductase during the course of purification. We took advantage of a Rubisco purification procedure to address this problem. Rubisco can be precipitated with PEG (14). Nearly all Rubisco could be precipitated with 18% PEG (w/v), but a significant amount of DHAP reductase was also coprecipitated. A range of PEG concentration (10 to 16%, w/ v) that would selectively precipitate Rubisco leaving much of DHAP reductase activity in the supernatant was evaluated. A 13% (w/v) PEG concentration was found to be a compromise, because a majority of the DHAP reductase activity remained in the supernatant,

although 20% Rubisco could not be eliminated. Supernatant from this step was used for further purification. The enzyme activity did not bind to a cation exchanger, CM cellulose, but most of the activity was bound to a DEAE cellulose, an anion exchanger. The enzyme activity was eluted with about 200 mM NaCl. The first peak of activity was stimulated by salt, whereas the later peaks were inhibited (Fig. 1). Thus, at this step, the osmoregulatory isoform could be separated from the chloroplastic and cytosolic glyceride isoforms. After the DEAE cellulose step, the total activity in the purification table (Table 2) represents primarily the Osm-DHAPR that constitutes 80% of total enzyme activity. Based on the total enzyme activity of three isoforms at the PEG step, 268 ␮mol ⭈ min⫺1 or less was expected but slightly higher activity (295 ␮mol ⭈ min⫺1) was recovered from the DEAE cellulose column, even though some inactivation of the enzyme may have occurred. It is likely that the DEAE cellulose step has eliminated an unknown inhibitor (15). From this step of purification, the specific activity corresponds to the osmoregulatory isoform.

TABLE 2 Purification of an Osmoregulatory Isoform of DHAP Reductase from D. tertiolecta

Purification step

Total activity (␮mol ⭈ min⫺1)

Total protein (mg)

Specific activity (␮mol ⭈ min⫺1 ⭈ mg⫺1 of protein)

Soluble homogenate PEG supernatant DEAE-cellulose (DE 52) Sephacryl S-200 Red Agarose I Red Agarose II

415 335 296 129 30 15

558 282 34 17 0.07 0.016

0.75 1.2 8.7 7.6 428 938

Recovery of total activity a (%) 100 81 71 31 7.2 3.6

(100) (44) (10) (5)

Purification (fold) 1 1.6 11.6 10 570 1250

a The values in parentheses may be relatively more accurate representations of the recovery of the Osm-DHAPR because after the DEAEcellulose chromatography, three isoforms were separated, and the last three steps of chromatography involved only the Osm-DHAPR.

408

GHOSHAL ET AL.

FIG. 1. Separation of three isoforms of DHAP reductase from homogenates of Dunaliella tertiolecta by DE 52 (DEAE cellulose) ionexchange chromatography. The supernatant from the PEG step was mixed with DEAE cellulose preequilibrated with TDG buffer for 2 h. The DEAE cellulose saturated with proteins was packed in a column (2.5 ⫻ 25 cm) and washed with TDG buffer. The bound protein was eluted with a linear salt gradient of 0 to 0.4 M NaCl in TDG buffer (200 mL each). All eluted fractions were assayed for DHAP reductase activity in the absence and the presence of 150 mM NaCl. Each fraction contained 3 mL. The first peak is the chloroplastic osmoregulatory form, the second peak is the chloroplastic glyceride form, and the third peak is the cytosolic form. The cells were grown in Dunaliella medium containing 1 M NaCl to stationary phase.

Ammonium sulfate inactivates the Osm-DHAPR; therefore, active fractions from the DEAE cellulose column were desalted and concentrated by the Filtron concentration system. The active concentrated fractions (2.0 mL) were resolved on a Sephacryl S-200 column. The Sephacryl S-200 step achieved some purification (Fig. 2), but a significant loss of activity occurred (56% of activity from the pooled DEAE cellulose); therefore, if desired, this step may be omitted. Fractions containing Osm-DHAPR activity were adjusted for optimal salt concentration for binding prior to applying to a Red Agarose column, which tightly bound all activity. Several combinations of salt and substrates for equilibration buffer (DHAP/glycerol-P/NaCl) and elution buffer (NAD(P)H/ NaCl/ KCl) were evaluated. The optimum salt concentration for binding was 0.5 M NaCl, and elution with 5 mM NADH and 0.5 M NaCl could recover about 25% of the activity (Table 2) with a few minor peptide bands. The Red Agarose step was repeated once again to achieve further purification. After the second Red Agarose column, a further 50% loss of activity occurred but the enzyme was purified to an apparent electrophoretic homogeneity as judged from SDS–PAGE (Fig. 2). In order to save the small amount of protein for the second Red Agarose column, normally fractions were not analyzed by SDS–PAGE. Characterization of the Osm-DHAPR. Properties of the Osm-DHAPR are summarized in Table 3. This isoform, like two other isoforms in Dunaliella, in vitro, can utilize both NADH and NADPH to reduce DHAP

to glycerol-P. The Km of NADH is not significantly different than the Km of NADPH. The lack of pyridine nucleotide specificity for the osmoregulatory isoform is different from the glyceride isoform from spinach (9). However, under physiological conditions, NADPH is the only available substrate in the chloroplasts of Dunaliella that can be utilized by Osm-DHAPR. The enzyme is active at a broad pH range (5.5 to 8), with a pH optimum of 6.9. The Osm-DHAPR has negligible activity of glycerol-P dehydrogenase (oxidation of glycerolP to DHAP) even under nonphysiological conditions, such as pH 9 with 30 mM glycerol-P and up to 5 mM NAD(P). Higher plants, like spinach, have a minimal NAD-linked glycerol-P dehydrogenase that is inhibited by salt (23). Detergents, lipids, or long-chain acyl CoA derivatives (oleoyl CoA, palmitoyl CoA, and stearoyl CoA) inhibit glyceride isoforms of DHAP reductase from plants and algae (6, 8), but the Osm-DHAPR was not inhibited by these molecules (Table 3). Although DTT and thioredoxin stimulate the chloroplastic glyceride isoform from both higher plants and algae, these molecules had no effect on the activity of Osm-DHAPR. Fructose-2,6-bisphosphate (FBP) did not affect the activity of Osm-DHAPR. Our results differ from the findings of Haus and Wegmann (16), who reported that “FBP, an inhibitor of mammalian muscle GPDH, activates the enzyme from Dunaliella in the direction of DHAP reduction.” These authors suggested that the activation was not strong enough to account for regulation. They further suggested that FBP activation was a remarkable effect, especially in relation to glycerol synthesis via degradation of starch, where FBP is an intermediate. Haus and Wegmann did not purify the Osm-DHAPR; their preparation also contained glyceride isoforms. This difference may account for the possible reason for the discrepancy with our results. Substrate specificity. The Osm-DHAPR had little preference for NADH or NADPH (Table 3), but it is highly specific for DHAP. The enzyme was not active

FIG. 2. Evaluation of purification of Osm-DHAPR by SDS–PAGE after each step of purification. A band at 38 kDa was visible for the purified Osm-DHAPR.

409

Osm-DHAP REDUCTASE FROM Dunaliella

TABLE 3 Properties of Osm-DHAPR and Its Comparison with the Chloroplast Glyceride Form from D. tertiolecta Osmoregulatory Form

Chloroplast Glyceride Form

pH Optimum NADH or NADPH Km DHAP (␮M) Km NADH/ NADPH (␮M) Vmax (␮mol ⭈ min⫺1 ⭈ mg⫺1 of protein) ⫺ NaCl Vmax (␮mol ⭈ min⫺1 ⭈ mg⫺1 of protein) ⫹ NaCl NaCl/KCl

6.9 23.4 23/19.3 938 2580 ⬃Three-fold stimulation by 150 to 250 mM

MgCl2 (25 mM)

Two- to three-fold stimulation; activity is inhibited by higher concentrations No effect No effect No effect

with substrates/metabolites of glycolysis (glycerladehyde-3-phosphate, 3-phosphoglycerate, pyruvate, fructose-6-phosphate, glucose-6-phosphate, fructose-1,6bisphosphate), the tricarboxylic acid cycle (citrate, oxaloacetate, malate), phosphoglycolate, hydoxypyruvate, or dihydroxyacetone. Effect of salts on Osm-DHAPR. The purified enzyme was stimulated with increasing salt concentrations in the assay, with a maximum threefold increase by 150 to 250 mM NaCl or KCl. If the concentration of salt was further increased, stimulation was progressively reduced to the basal level at 500 mM (Fig. 3). A similar salt stimulation could be achieved by 25 mM MgCl2 but higher concentrations inhibited the enzyme activity. The salt stimulation is unique to this isoform, compared to the glyceride isoform that is inhibited by salt (Gee et al., 1993). The glyceride isoform from Dunaliella is relatively more tolerant to inhibition by salt than spinach. The activity of the glyceride isoform in Dunaliella was inhibited 50% by 100 mM NaCl compared to severe inhibition of the spinach isoform by 25 mM NaCl (9). Antibodies raised against the cytosolic glyceride isoform of DHAP reductase from spinach did not crossreact with the Osm-DHAPR. Therefore, based on differences in the properties with glyceride isoforms, we propose that the Osm-DHAPR in Dunaliella is unique and different from glyceride isoforms from higher plants and algae.

may be present. In higher plants and algae, glycerol3-phosphate dehydrogenase is referred to as DHAP reductase, because at physiological pH and substrate concentrations the enzyme is essentially inactive as a dehydrogenase (4, 5). We have discovered and now purified 300 NaCl

Osm-DHAP Reductase (% Activity)

Effect of detergents (triton X-100, Chaps, and ␤ -octylglucoside) (50–100 ␮M) Fatty acyl CoA CoA (palmitoyl CoA, stearoyl CoA, and oleoyl CoA] (100 ␮M) DTT (5 mM) and Escherichia coli reduced thioredoxin (5 ␮g)

6.9 80 57/63 ND ND 10% stimulation by 50 mM 50% inhibition by 100 mM Stimulates 10%, activity is inhibited by higher concentrations 50–60% inhibition, 90% inhibition by Chaps 80% inhibition, 100% Inhibition by oleoyl CoA Stimulates two-fold

KCl

250

MgCl2

200

150

100

50

0 0

100

200

300

400

500

Salt [mM]

DISCUSSION All organisms need glycerol phosphate for the synthesis of glycerolipids; therefore, isoforms of glycerol-3phosphate dehydrogenase with different properties

FIG. 3. Effect of salts (MgCl2, KCl, and NaCl) on the Osm-DHAPR from Dunaliella. The reaction in 1.0 mL contained 33 mM Mes, 33 mM Hepes, 34 mM tricine at pH 6.9, 1 mM DHAP, and 200 ␮M NADH. The pattern of stimulation was similar if the activity was assayed with 200 ␮M NADPH. The enzyme was assayed immediately after purification from the second Red Agarose column. The activity without salt is shown as 100%.

410

GHOSHAL ET AL.

a unique chloroplastic Osm-DHAPR from Dunaliella that appears to be present as a complex with glycerolP phosphatase to synthesize free glycerol from DHAP (unpublished results). Metabolic intermediates of lipid biosynthesis had no effect on the activity of OsmDHAPR, and the two chloroplastic isoforms were regulated differently in Dunaliella, suggesting that this isoform may not involved in glyceride metabolism. The purification of the Osm-DHAPR required resolution of several unusual combinations of challenges, including loss of activity with ammonium sulfate precipitation; strong association of Rubisco; and the sensitivity of enzyme to the cold. The first two problems were addressed by using PEG to eliminate Rubisco while leaving the enzyme activity in the solution. Additionally, once the algal cells were homogenized all steps of purification were performed at a stretch, including 2.5% (v/v) glycerol in buffers stabilizing the enzyme activity. Dunaliella uses glycerol for osmoregulation and for protecting proteins from dehydration, a protective effect of glycerol on the enzyme activity was expected. The purified enzyme was stable for about one month at ⫺20⬚C, if stored after desalting in 2 M glycerol. We have used a molecular sieving (Sephacryl S-200) column for which sample preparation and chromatography required more than 16 h, during which time a significant loss of activity occurred. Therefore, enzyme may be directly applied to an affinity chromatography medium. The photosynthetic carbon reduction cycle is the primary biochemical pathway that contributes carbon to glycerol synthesis in Dunaliella; however, in an elevated stress condition, carbon may also be derived from the products of starch breakdown (13). The triose-P DHAP is the central molecule that is utilized for both lipid metabolism and osmoregulation in Dunaliella. Previous investigations of the glycerol metabolism for osmoregulation in Dunaliella were based on DHAP reductase activity in the chloroplast (3, 10, 21). The presence of two isoforms in the chloroplast was not recognized until we separated three isoforms by isolating enzyme from the cells that were adapted to a high salt concentration and grown to stationary phase, where only the osmoregulatory isoform predominates (8). Thus, previous investigators were not aware of the presence of multiple isoforms of DHAP reductases. The presence of two distinctly different isoforms of DHAP reductase in the chloroplast further suggests suborganelle enzymatic localization of glycerol-P synthesis, one for the synthesis of lipids and one for the synthesis of glycerol (8). The salt stress inhibits glyceride forms and stimulates the Osm-DHAPR for glycerol production from DHAP (8). With only one form of DHAP reductase, it would be difficult to regulate two different functions. We have observed that during initial steps of purification, some glycerol-P phosphatase activity is often coeluted with the Osm-DHAPR. It is likely that

the two chloroplastic isoforms of two diverse functions may be separated by forming a complex with other enzymes of the glycerol or glyceride synthesis pathway via protein–protein interactions. In fact, the protein– protein interactions to form a complex for metabolic/ substrate channeling makes sense; otherwise, how would the two isoforms of the DHAP reductase be separated in the chloroplasts and compete for the glycerolP from a single metabolite pool? Protein–protein interactions are intrinsic to virtually every cellular process (25). A well-known example is the pyruvate dehydrogenase complex (22). A similar multienzyme complex formation via protein–protein interactions has been reported for other chloroplastic enzymes (26). Because of lack of substantial biochemical data on physical interaction between two proteins, we recognize that our observation of coelution is not strong evidence for complex formation. Therefore, future work could be directed toward characterization of glycerol-P phosphatase, protein–protein interactions between glycerol-P phosphatase and the Osm-DHAPR, its regulation, and genetic engineering of plants with Osm-DHAPR and glycerolP phosphatase to induce better osmoregulation with glycerol. In conclusion, this is the first report of purification of an isoform of DHAP reductase from Dunaliella, specifically the osmoregulatory isoform that is involved in the synthesis of free glycerol for osmoregulation. OsmDHAPR is a unique enzyme from an algal species that is adapted to extreme environments. Therefore, understanding the biochemistry of this enzyme is not only important from the point of view of biology, but it may have significant practical applications in agriculture. ACKNOWLEDGEMENTS We dedicate this paper to the late Professor N. E. Tolbert, who showed a minor shoulder in the chromatogram of the chloroplastic peak and had challenged AG to discover the new isoform. The authors thank Dr. Neil Nelson for critically reading the manuscript.

REFERENCES 1. Avron, M. (1986) The osmotic components of halotolerant algae. Trends Biochem. Sci. 11, 5–6. 2. Ben-Amotz, A., and Avron, M. (1974) Isolation, characterization, and partial purification of a reduced nicotinamide adenine dinucleotide phosphate-dependent dihydroxyacetone reductase from the halophilic algae Dunaliella parva. Plant Physiol. 53, 628– 631. 3. Brown, A. D., Lilley, R. McC, and Marengo, T. (1982) Osmoregulation in Dunaliella. Intracellular distribution of enzymes of glycerol metabolism. Z. Naturforsch. 37c, 1115–1123. 4. Gee, R. W., Byerrum, R. U., Gerber, D. W., and Tolbert, N. E. (1988) Dihydroxyacetone phosphate reductase in plants. Plant Physiol. 86, 98–103. 5. Gee, R. W., Byerrum, R. U., Gerber, D. W., and Tolbert, N. E.

Osm-DHAP REDUCTASE FROM Dunaliella (1988) Differential inhibition and activation of two leaf dihydroxyacetone phosphate reductases. Plant Physiol. 87, 379–383. 6. Gee, R. W., Byerrum, R. U., Gerber, D. W., and Tolbert, N. E. (1989) Changes in the activity of the chloroplastic and cytosolic forms of dihydroxyacetone phosphate reductase during maturation of leaves. Plant Physiol. 89, 305–308. 7. Gee, R. W., Goyal, A., Byerrum, R. U., and Tolbert, N. E. (1989) Two isozymes of dihydroxyacetone phosphate reductase in Dunaliella. Plant Physiol. 91, 345–351. 8. Gee, R. W., Goyal, A., Byerrum, R. U., and Tolbert, N. E. (1993) Two isoformes of dihydroxyacetone phosphate reductase from the chloroplasts of Dunaliella. Plant Physiol. 103, 243–249. 9. Gee, R. W., Goyal, A., Gerber, D. W., Byerrum, R. U., and Tolbert, N. E. (1988) Isolation of dihydroxyacetone phosphate reductase from Dunaliella chloroplasts and comparison with isozymes from spinach leaves. Plant Physiol. 88, 896–903. 10. Gimmler, H., and Lotter, G . (1982) The intracellular distribution of enzymes of the glycerol cycle in the unicellular algae Dunaliella parva. Z. Naturforsch. 37c, 1107–1114. 11. Goyal, A., Brown, A. D., and Gimmler, H. (1987) Regulation of salt induced starch degradation in Dunaliella tertiolecta. J. Plant Physiol. 127, 77–96. 12. Goyal, A., Lilley, R. McC, and Brown, A. D. (1986) The size and the turnover of the glycerol pool in Dunaliella. Plant Cell Environ. 9, 703–706. 13. Goyal, A., Lilley, R.McC, and Brown, A. D. (1986) Osmoregulation in Dunaliella : Relative contribution of photosynthesis and starch breakdown to glycerol biosynthesis during a salt stress in light, in “Contemporary Themes in Biochemistry” (Kon, O. L., Chung, MC-M, Hwang, P. L. H., Leong, S. F., Loke, K. H., Thiyagarajah, P., and Wong, P. TH., Eds.), pp. 336–337, Cambridge Univ. Press, Cambridge. 14. Hall, N. P., and Tolbert, N. E. (1978) A rapid procedure for the isolation of ribulose bisphosphate carboxylase/oxygenase from spinach leaves. FEBS Lett. 96, 167–169. 15. Haus, H., and Wegmann, K. (1984) Glycerol 3 phosphate dehydrogenase (EC 1.1.1.8) from Dunaliella tertiolecta. I. Purification and kinetic properties. Physiol. Plant. 60, 283–288.

411

16. Haus, H., and Wegmann, K. (1984) Glycerol 3 phosphate dehydrogenase (EC 1.1.1.8) from Dunaliella tertiolecta. II. Influence of phosphate esters and different salts on the enzymatic activity with respect to osmoregulation. Physiol. Plant. 60, 289–293. 17. Johnson, M. K., Johnson, E. J., MacElroy, R. D., Spear, H. Z., and Bruff, B. S. (1968) Effect of salts on the halophilic algae Dunaliella viridis. J. Bactriol. 95, 1461–1468. 18. Kirsch, T., Gerber, D. W., Byerrum, R. U., and Tolbert, N. E. (1992) Plant dihydroxyacetone phosphate reductases: Purification, characterization and localization. Plant Physiol. 100, 352–359. 19. Klo¨ck, G., and Kreuzberg, K. (1989) Kinetic properties of a snglycerol-3-phosphate dehydrogenase purified from the unicellular alga Chlamydomonas reinhardtii. Biochim. Biophys. Acta 991, 347–352. 20. Lerner, H. R., and Avron, M. (1977) Dihydroxyacetone kinase activity in Dunaliella parva. Plant Physiol. 59, 15–17. 21. Marengo, T., Lilley, R.McC, and Brown, A. D. (1985) Osmoregulation in Dunaliella. Catalysis of glycerol 3 phosphate dehydrogenase reaction in a chloroplast enriched fraction of Dunaliella tertiolecta. Arch. Microbiol. 142, 262–268. 22. Randall, D. D., Miernyk, J. A., Fang, T. K., Budde, J. A., and Schuller, K. A. (1989) Regulation of the pyruvate dehydrogenase complexes in plants. Ann. N. Y. Acad. Sci. 573, 193–205. 23. Santora, G. T., Gee, R., and Tolbert, N. E. (1979) Isolation of a snglycerol-3-phosphate: NAD oxidoreductase from spinach leaves. Arch. Biochem. Biophys. 196, 403–411. 24. Sharma, N., Phutela, A., Malhotra, S. P., and Singh, R. (2001) Purification and characterization of dihydroxyacetone phosphate reductase from immature seeds of Brassica campestris L. Plant Sci. 160, 603–610. 25. Srere, P. A. (1987) Complexes of sequential metabolic enzymes. Annu. Rev. Biochem. 56, 89–124. 26. Su¨ss, K. H., Arkona, C., Manteuffel, R., and Adler, K. (1993) Calvin cycle multienzyme complexes are bound to chloroplast thylakoid membranes of higher plants in situ. Proc. Natl. Acad. Sci., USA 90, 5514–5518. 27. Sussman, I., and Avron, M. (1981) Characterization and partial purification of DL-glycerol-1-phosphatase from Dunaliella salina. Biochim. Biophys. Acta 661, 199–204.