Plant Physiol. (1987) 85, 360-364 0032-0889/87/85/0360/05/$01.00/0

Localization of Carbohydrate Metabolizing in Guard Cells of Commelina communis1 Received for publication April 8, 1987 and in revised form June 13, 1987 NINA L. ROBINSON2 AND JACK PREISS*3 Department ofBiochemistry and Biophysics, University ofCalifornia, Davis, California 95616

ABSTRACI leaves. The sucrose is either degraded in the apoplast or in the cytoplasm of the storage cell. Sucrose, or its degradation prod- The lliztion ofenzymes involved in the flow of carbon into and out ucts, can be further metabolized to the triose-P or 3-PGA level. of starch was determined in guard cells of Commelina communis. The These compounds may then move into the amyloplast via the guard cell chloroplasts were separated from the rest of the cellular triose-P/Pi translocator and are converted into starch. However, components by a modification of published microfuge methods. The at present, the presence of the triose-P/Pi translocator in amy- enzymes of interest were then assayed in the supernatant and chloroplast loplasts has not been demonstrated. Assuming that the triose-P/ fractions. The chloroplast yield averaged 75% with 10% cytoplasmic Pi translocator is present, the movement of carbon into starch contamination. The enzymes involved in starch biosynthesis, ADPglucose would be a reversal of the enzymic steps occurring in the cyto- pyrophosphorylase, starch synthase, and branching , are located plasm with the last several steps resulting in the direct incorpo- exclusively in the chloroplast fraction. The enzymes involved in starch ration of carbon into starch. The above process will be reversed degradation show a more complex distribution. Phosphorylase is located with starch degradation occurring and the products moving into in both the supernatant and chloroplast fraction, 50% in each fraction. the cytoplasm presumably via the triose-P/Pi translocator. Most of the and debranching enzyme activity is present in the Triose-P and 3-PGA are then further metabolized, as needed, in supernatant (70%) fraction. The majority of the rest of the enzymes the cytoplasm. involved in the degradation of starch to malate and synthesis of starch Using the amyloplast system as a model for guard cells, it was from a hexose precursor were also investigated. All of the enzymes were ofinterest to determine the enzyme activity ofthe starch biosyn- present in the chloroplast except for hexokinase and phosphofructoki- thetic and degradative pathways and their localization. A chlo- nase. The inability to assay these enzymes could possibly have been due roplast and cytoplasmic (supernatant) fraction were obtained to the lack of or low activity of the enzymes or to nonoptimal assay from guard cell using a microfuge technique. En- conditions. zymes involved in the flow ofcarbon into and out of starch were then assayed in the two fractions. MATERIALS AND METHODS Isolation of Guard Cell Protoplasts. The plant material was Commelina communis, 4- to 6-week-old plants. The abaxial One ofthe consequences ofstomatal opening is the breakdown epidermis ofthe leaves was peeled and placed in 0.25 M mannitol of starch during the day into malate and citrate, for use as an with 0.5 mMCaCl2 for at least 30 min. The peels were collected osmoticum. In the late afternoon or evening starch is resynthe- and digested for approximately 1 h in 2% , 10 mm Mes sized. This is the reverse of the situation in mesophyll cells. At (pH 5.3), 0.25 M mannitol, and 0.5 mm CaCl2 at 25°C on an the present time there is no evidence as to the source of carbon orbital shaker at 100 rpm. During this digestion any contami- for the resynthesis of starch due to the lack of Rubisco4 activity nating epidermal and mesophyll cells form protoplasts and burst (22) and the malate formed during opening is not converted due to the low osmoticum. The peels were collected on Miracloth back into starch (19). Based on this, what is the pathway for and transferred to a solution containing 2% cellulase, 0.5% starch synthesis in guard cells and in which ofthe compartments macerase, 10 mm Mes (pH 5.3), 0.4 M mannitol, and 0.5mm are the enzymes localized? These same questions can also be CaCl2. This digestion lasted for approximately 4 h at 25°C on an asked in reference to the starch degradation pathway. orbital shaker at 60 rpm. In storage tissue, which contains starch in amyloplasts, the The guard cell protoplasts were harvested by passing them source of carbon for starch synthesis is sucrose imported from through 20 ,um nylon mesh followed by centrifugation at 200g for 100 min. The supernatant was discarded and the pellet 'Supported in part by a National Science Foundation grant PCM82- resuspended in 2 ml of 0.4 M mannitol with 0.5 mm CaCl2. This 0570. was layered on a 6 ml discontinuous 22/67/90% Percoll gradient 2McKnight Foundation trainee. Present address: Mann Laboratory, (5) and centrifuged for 5 min at 400g. The guard cell protoplasts Department of Vegetable Crops, University ofCalifornia, Davis, CA. were recovered from the 22/67% interface and rinsed with 0.4 3Present address: Department of Biochemistry, Michigan State Uni- mannitol, 10 mm Mes (pH 6.5), 1 mm EDTA, and 0.5 mM CaCl2 versity, East Lansing, MI 48824. (rinsing solution) and collected by a 10 min centrifugation at 4Abbreviations: Rubisco, ribulose 1,5-bisphosphate carboxylase/oxy- 200g. genase; ADPGlc, ADPglucose; RuBP, ribulose bisphosphate; PFK, phos- Isolation of Guard Cell Chloroplasts. The pellet was resus- phofructokinase; PFP, pyrophosphate:fructose-6-P phosphotransferase; pended in 0.3 ml of fresh rinsing solution. This solution was 3-PGA, glyceric acid 3-phosphate; PEP, phosphoenolpyruvate; Glc I-P, placed in an Eppendorf tube separated from 1 ml of 0.7 M 1-phosphate; Fru 6-P, fructose 6-phosphate; Fru 1,6-bisP, fruc- sucrose, 50 mm Tricine (pH 7.9), and 0.5 mm CaCl2 (sucrose tose 1,6-bisphosphate; Fru 2,6-bisP, fructose 2,6-bisphosphate. pelleting material) by a piece of 5 ,m mesh (modification of 360 CARBOHYDRATE METABOLISM IN GUARD CELLS 361 both Refs. 15 and 26). This was centrifuged for 15 s in a overnight in ethanol: 1 M ammonium acetate (pH 3.8, 5:2) after microfuge. The supernatant was removed, containing most of adding sufficient EDTA to make the concentration 20 Mmol. the cell contents except the chloroplasts, and the pellet, mostly This was done to improve the resolution. Standards of frutose, chloroplasts, resuspended in 1 ml of the rinsing solution. The Fru 6-P, and Fru 1,6-bisP were run at 0.1 smol and the location enzyme assays were done on these two fractions. determined with silver nitrate dip (1). The second reaction did Enzyme Assays. Assay of Starch Biosynthetic Enzymes. not go to completion so the label was present in both Fru 1,6- ADPGlc pyrophosphorylase was assayed only in the pyrophos- bisP and Fru 6-P. Both radioactive spots were eluted after soaking phorylase direction (24, 25). The reaction mixture contained 20 the chromatograms in 100% ethanol for 3 h. ,gmol ofglycylglycine (pH 8.0), 1.5 JUmol ofMgCl2, 0.25 umol of Protein was determined using the BCA protein reaction from ADPGlc, 0.5 gmol 32PPi (1000-6000 cpm/nmol), 0.1 mg BSA, Pierce with crystalline BSA as the standard. The cellulase used 0.2 ,mol 3-PGA, and extract in a final volume of 0.25 ml. The in the digestion was obtained from Cooper Biomedical and the reaction mixture was incubated for 30 min at 37°C. Starch macerase was obtained from Calbiochem. The 5 and 20 Mm synthase was assayed as described by Hawker et al. (6). The meshes were obtained from Spectrum Medical Industries, Inc. specific activity of ADP-['4C]Glc was increased to 2000 cpm/ The ['4CJfrutose and ["4C]Glc 1-P were both obtained from nmol and the incubation time was increased to 30 min. Branch- Amersham. The ADP['4C]glc was produced as described by ing enzyme activity was determined as described previously (6). Hawker et al. (6). The coupling enzymes that were used in the In some experiments it was necessary to add a known amount spectrophotometric and fluorometric assays and the Percoll were of rabbit liver glycogen as primer. The reaction was terminated obtained from Sigma. Those enzymes suspended in ammonium after 2 h at 30C. sulfate that were used in the same reaction mixture as the extracts Assay ofStarch Degradative Enzymes. Both amylase and R- were dialyzed overnight in 50 mM Hepes (pH 7.5), 1 mM EDTA, enzyme were assayed by determining the concentration ofreduc- and 1 mM GSH or DTE. Rabbit muscle phosphorylase a, used ing sugars present at the completion of the reaction (18). Both in the branching enzyme assay, was obtained from Boehringer- reactions were terminated after 2 h at 37°C and the concentration Mannheim. ofreducing sugars determined using Nelson's (16) assay. Hexokinase was assayed using two different methods (27, RESULTS AND DISCUSSION The assay was converted to a 29). phosphorylase (18) two-step Several methods were tried before the microfuge method was fluorometric assay (14) because ofthe low activity ofthe enzyme. found to provide the highest chloroplast yields and lowest cyto- In the first step of the reaction, Glc 1-P was formed while in the plasmic contamination in the chloroplast fraction ofa guard cell second step the concentration of Glc 1-P was determined. The communis first was incubated at for 15 C. preparation. The chloroplast yield and step 37C min. The second step cytoplasmic contamination were determined by assaying the went to completion at room temperature. Phosphoglucomutase activity of a cytoplasmic, PEP carboxylase, and chloroplastic, was assayed spectrophotometrically (3). ADPGlc pyrophosphorylase, marker in both fractions. ADPglc Assays ofGlycolytic Enzymes. Aldolase was assayed spectro- pyrophosphorylase was used as the chloroplast marker due to photometrically over a 10 min period (10). Fructose bisphospha- the lack of Rubisco activity in guard cells (22). The chloroplast tase was assayed in two ways, fluorometrically and radioactively, yield is the percentage ofthe ADPGlc pyrophosphorylase activity using a modification of Kelly et al. (12). The fluorometric assay present in the chloroplast fraction. The cytoplasmic contamina- was converted to a two-step reaction. In the first step Fru 6-P tion is the percent of PEP carboxylase activity present in the was formed while in the second step the Fru 6-P concentration chloroplast fraction. The microfuge method described above was fluorometrically determined. The radioactive assay used resulted in 74% chloroplast yield (n = 48; SD = 6.2) and 10% ['4C]Fru-1,6-bisP. After 10 min at room temperature the prod- cytoplasmic contamination (n = 36; SD = 3.6) ofthe chloroplast ucts of the reaction were determined using descending paper fraction. chromatography in ethanol: 1 M ammonium acetate (5:2, pH 3.8) Starch Biosynthetic Pathway. Results. The enzymes involved solvent system and rechromatographed in butanol: in the conversion of Glc 6-P to triose-P, and 3-PGA in the propanol:acetone:80% formic acid:30% TCA (40:20:25:25:1.5) cytoplasm were assayed (Table I) and found to be present in the with 0.5 g of EDTA (100 ml solvent)-'. The location of the cytoplasm. PFK and PFP catalyze the conversion of Fru 6-P to standards were determined (1). The location of the radioactivity Fru 1,6-bisP. The highest activities for PFP were obtained only was determined by counting 1 cm sections in a toluene based in the presence of the activator, Fru 2,6-bisP. In the absence of cocktail. PEP carboxylase was determined using a modification Fru 2,6-bisP there was a 75% decrease in activity. of Bahr and Jensen's (2) Rubisco assay. The reaction mixture The triose-P/Pi translocator found in the chloroplast mem- contained 1 gmol of PEP in place of RuBP and the reaction was brane of mesophyll cells exchanges triose-P or 3-PGA for phos- terminated after 10 min at room temperature. PFK was assayed phate into or out ofthe chloroplasts (7). Presumably this translo- spectrophotometrically (4, 11) with 25 gmol of Pi (pH 7.5) cator also functions in guard cells. Once the three-carbon units included when the cytoplasmic isozyme was assayed. Phospho- are inside the chloroplast they could be converted into starch by glucose isomerase was assayed spectrophotometrically (17). PFP a partial reversal of the steps outlined above in the cytoplasm. was assayed as described previously (13). Glyceraldehyde 3-P The activity of the chloroplastic isozymes of NADH triose-P dehydrogenase was assayed spectrophotometrically (8) as was dehydrogenase, triose-P isomerase, and aldolase are lower than triose-P isomerase (28). for their cytoplasmic isozymes. The conversion of Fm 1,6-bisP Production of/4JFru-1,6-BisP. The ['4C]Fru-1,6-bisP used in to Fru 6-P is catalyzed by fructose bisphosphatase. Fructose the radioactive fructose bisphosphatase assay was made from bisphosphatase was difficult to assay because of its low activity. ['4C]Fru using yeast hexokinase (Sigma) and rabbit muscle Fru Two methods were used as assays. A two-step fluorometric assay 6-P kinase (Sigma) in a two-step procedure. The reaction mixture monitoring the production of NADH and a radioactive assay. for the hexokinase reaction contained 50 smol of Tris-HCl (pH When the radioactive assay was used the product ofthe reaction 8.0), 1 ,umol of frutose, 10 Mmol of ATP, 10 ,umol of MgCl2, 1.7 was shown to be Fru 6-P. The enzymes necessary for the con- units of hexokinase, and 120 uCi of ['4C]frutose in 1 ml. After version of Fru 6-P to Glc I-P, the substrate for ADPGlc pyro- 15 min at room temperature 0.02 unit of Fru 6-P kinase was phosphorylase, were present in the chloroplast. added and the reaction proceeded for an additional 30 min at The last steps in the starch biosynthetic pathway result in the room temperature. The reaction mixture was chromatographed direct incorporation of carbon into starch. The first enzyme in 362 ROBINSON AND PREISS Plant Physiol. Vol. 85, 1987 Table I. Distribution and Activity ofEnzymes Involved in the Flow ofCarbon into and out ofStarch in Guard Cells ofC. communis The protoplasts and chloroplasts were prepared by modification and combination of Miills and Joy (15) and Shimazaki and Zeiger (26). Enzyme Supernatant Chloroplast Percent in nmol (mg protein. min)-' Marker Enzymes PEP carboxylase 96.5 10.0 10.4 (36)a ADPglucose pyrophosphorylase 11.3 42.8 73.7 (48) Starch Degradative Enzymes Amylase 29.5 16.3 (2) 34.5 (6)b Phosphorylase 41.8 34.0 (2) 42.8 (6)b R-enzyme 22.1 6.2 (3) 24.7 (6)b Starch Biosynthetic Enzymes ADPglucose pyrophosphorylase 11.3 42.8 73.7 (48) Starch synthase 0.4 1.8 (2) 76.8 (4)b Branching enzyme 1 mgprimer 78.5 324.4(1) 75.8(4)b 0.01 mg primer 42.3 136.5 (1) 69.0 ( I)b Glycolytic Enzymes Aldolase 27.0 9.1 28.3 (3) Fructose-1,6-bisphosphatase 7.6 10.4 58.0 (4) Phosphoglucose isomerase 283 46.0 17.7 (3) Phosphoglucomutase 192 52.5 26.0 (3) PFK 8.9 <0.97c 0.0 (3) PFP 9.5 <0.97c 0.0 (2) Glyceraldehyde I-P dehydrogenase NADPH 57.9 80.3 64.7 (3) NADH 555.0 69.8 15.0 (3) Triose-P isomerase 7853 2944 25.0 (3) aNumber in parentheses is the number of chloroplast preparations studied for enzyme distribu- tion. b Protein determinations were not made in every case in which the percent distribution was deter- mined. 'There was no activity detected in the chloroplast fraction. this portion ofthe pathway is ADPGlc pyrophosphorylase which cator, assuming that the triose-P/Pi translocator is present in converts Glc 1-P to ADPGlc. This enzyme is present exclusively guard cell chloroplast membranes and it functions as in meso- in the chloroplast. ADPGlc is incorporated into an al,4-glucan phyll chloroplasts. This must be verified for guard cells. Experi- by starch synthase. The formation of the a 1,6-branch points ments were attempted to determi:ne what compounds move into result from branching enzyme action. It was occasionally neces- the chloroplast but they were unsuccessful (N Robinson, unpub- sary to assay branching enzyme in the presence of a primer, lished results). It was assumed initially that ADPGlc pyrophos- rabbit liver glycogen. The activity at two different primer con- phorylase would be the control point in the conversion ofthree- centrations is given in Table I. All three enzymes are located carbon units into starch within the chloroplast compartment. If exclusively in the chloroplast. Several experiments were con- this is true, ADPGlc pyrophosphorylase should have effectors ducted without obtaining protein data for specific activity results. and exhibit the lowest activity in the pathway. Robinson et al. The percent distribution results from these experiments were still (24) and Outlaw and Tarcznski (21) have shown that ADPGlc used to obtain the distribution values given in Table I for both pyrophosphorylase from guard cells is controlled by the ratio of starch synthase and branching enzyme. (This is also true for [3-PGA] to [Pi]. ADPGlc pyrophosphorylase, however, did not phosphorylase, amylase, and R-enzyme.) show the lowest activity. Fructose bisphosphatase, aldolase, and Discussion. The carbon source for starch biosynthesis in guard starch synthase all had lower activities. The activity of ADPGlc cells is not known. Guard cells have no Rubisco activity (22) pyrophosphorylase reported in Table I is in the presence of 1mM and malate is not reconverted into starch at physiologically 3-PGA, the activator. In the absence of 3-PGA, in guard cells, significant rates (19). The first enzyme assayed in the starch there is no activity. The activity in vivo will fluctuate from 0 to biosynthetic pathway therefore was phosphoglucose isomerase 42.81 nmol (min mg)-' depending on the ratio of [3-PGA] to with Glc 6-P as the substrate. The enzymes necessary for the [Pi] and can still control flux through this portion ofthe pathway. conversion of Glc 6-P to triose-P are present in the cytoplasm Starch synthase showed the lowest activity, 1.81 nmol (mg and the enzymatic control point is the conversion of Fru 6-P to min)-', of all the enzymes in this portion of the pathway. Based Fru 1,6-bisP. Two enzymes catalyze this conversion and which on assumptions from Outlaw (19) in Vicia faba, 1385 nmol is responsible for the in vivo activity is not known. It is possible (anhydroglucosyl equivalents) (mg protein)-' is needed for each that both enzymes are active in the cell with one more active cycle of stomatal opening and closing. If these figures are in the under specific metabolic conditions than the other. Both ofthese same order of magnitude in C. communis and if starch synthase enzymes are controlled by a number of effectors which are an is the rate limiting step for starch synthesis it would take approx- indication ofthe energy status of the cell. imately 13 h to form sufficient starch for one cycle of opening Once triose-P or 3-PGA are formed in the cytoplasm they and closing. ADPGlc pyrophosphorylase could be inhibited over presumably move into the chloroplast via the triose-P/Pi translo- 90% and still not be the rate limiting step. However, the activity CARBOHYDRATE METABOLISM IN GUARD CELLS 363 ofstarch synthase is probably low due to two factors: (a) and chloroplast fraction with no success. The enzyme may be present in the chloroplast fraction could degrade the product of present in such low amounts that it was not possible to detect its the starch synthase reaction as it forms and (b) there are two activity by the assay used or optimal assay conditions were not forms of starch synthase, soluble and granule bound. Only the used. Forty percent of the phosphorylase activity and approxi- soluble starch synthase was assayed. Both of these factors would mately 25% of the phosphoglucomutase activity are present in lead to erroneously low activities for starch synthase. the chloroplast. Assaying the isozymes of fructose bisphosphatase was difficult The conversion of Glc 6-P to triose-P and 3-PGA for export due to their low activity. This enzyme was of particular interest from the chloroplast is a reversal of the starch biosynthetic because Hedrich et al (9) found no fructose bisphosphatase pathway except for the conversion of Fru 6-P to Fru 1,6-bisP. activity in guard cell chloroplasts of V. faba or the Argenteum This step is catalyzed by the chloroplast isozyme ofPFK. Various mutant of Pisum sativum. They proposed that compounds pro- assay conditions were tried but no PFK activity could be detected duced by the cytosolic fructose bisphosphatase or substrates for with the assay procedure used. PFK or PFP are transported into and out of the chloroplast Discussion. The enzymes necessary for the phosphorolytic rather than three-carbon units, and are then converted into pathway are present in the chloroplast but there was no hexoki- starch. There is no evidence in mesophyll cells for a hexose-P nase activity which is necessary for the hydrolytic pathway. translocator (7). Fructose bisphosphatase activity is present in However, neither pathway can be ruled out at this time because the chloroplast as indicated both by the fluorometric and radio- it is possible that optimal assay conditions for hexokinase were active assays. The observed activity was not due to aldolase not used. It may be possible to detect the enzyme activity with a because no ['4C]triose-P were found when ['4C]Fru 1,6-bisP was fluorometric procedure which would increase the Insitivity 10 used. The activity was also not due to PFP because this enzyme to 100 times over that of the spectrophotometric assay. Using is located exclusively in the guard cell cytoplasm. The fructose the calculation presented earlier it is possible to determine how bisphosphatase and aldolase activity results reported here are long it would take to degrade starch for one cycling of opening. approximately 10 nmol (mg min)-' for the chloroplast isozymes. If the phosphorolytic pathway is used and R-enzyme is not rate If fructose-1,-6-bisphosphatase and aldolase are the rate limiting limiting, it wouid take about 40 min to degrade sufficient starch step in starch synthesis, instead of starch synthase, it would take for one cycle of opening. If only the hydrolytic pathway is 3 h to convert a sufficient amount ofanhydroglucosyl equivalents functioning it would take about 1.5 h to degrade sufficient starch. into starch for one cycle of openin, and closing based on the If both pathways are functioning time is reduced to 30 min. same calculation used above. For comparison, the activity of Before determining starch levels, Outlaw and Manchester (20) ADPGlc pyrophosphorylase could be inhibited 80% and still not allowed 1.5 h to elapse from fully closed to fully open stomata. be rate limiting under these conditions. A time course of starch The above calculations are consistent with this time period. The concentrations changes during stomatal opening and closing has time required for stomatal opening varies depending on environ- not been done at the present time. These experiments would be mental conditions. In general, stomatal opening and therefore difficult to do because of the inherent limitations of working starch degradation occur within a few hours. A time course, with guard cells. The guard cells have the entire dark cycle in under specified environmental conditions, of starch degradation which to convert sufficient carbon into starch for the next cycle has not been done at the present time. Probably both degradation of opening and the fructose bisphosphatase and aldolase results pathways function in vivo, but which dominates is not clear at are in this order of magnitude. the present time, in either mesophyll or guard cells. Which Another unanswered question is the source ofcarbon for starch pathway dominates may well depend on the plant material biosynthesis because of the lack of Rubisco activity (22) and investigated. In pea mesophyll chloroplasts, the dominant path- malate is not reconverted to starch at physiologically significant way is phosphorolytic (27) but in spinach mesophyll chloroplasts rates (19). The most likely source is sucrose. Outlaw (19) has both pathways function (18) as evidenced by the enzymes pres- proposed that sucrose could be a cytoplasmic osmoticum. When ent. the stomata begin to close the sucrose could be converted into The other problem in the degradation pathway is the absence starch. Even if sucrose is not the cytoplasmic osmoticum it is of the chloroplast isozyme of PFK. Kelly and Latzko (11) were produced in the surrounding mesophyll and exported out of the unable to assay the chloroplast form ofPFK in a crude mesophyll cells to the phloem. Guard cells could then take the sucrose up mixture. They were only able to detect activity ofthe chloroplast from the apoplast. isozyme after ammonium sulfate precipitation. These experi- The enzymes necessary for the production of triose-P and 3- ments were not conducted with guard cell extracts because of PGA are present in the cytoplasm and the enzymes involved in the limited material. Hedrich et al. (9) did detect activity of the the conversion ofthese three-carbon units into starch are present chloroplastic PFK in both V. faba and Argenteum pea. in the chloroplast. The enzymes directly involved with starch The distribution of the enzymes directly involved in starch biosynthesis, ADPGlc pyrophosphorylase, starch synthase, and degradation are more complex than for the starch biosynthetic branching enzyme, are all localized exclusively within the guard enzymes. The degradative enzymes are present in both cellular cell chloroplast. The distribution results on the starch biosyn- compartments but more of the activity is in the cytoplasm than thetic enzymes are similar to the results obtained by Okita et al. chloroplast. Okita et al. (18) found this same distribution pattern (18) for mesophyll cells. in spinach mesophyll cells. Why these enzymes are present in Starch Degradation Pathway. Results. There are two possible the cytoplasm of mesophyll and guard cells is not understood. pathways for the degradation of starch in vivo (Table I). One One possibility is in the regulation of the starch degradative pathway is hydrolytic involving amylases, R-enzyme, and hexo- enzymes. Thus far (23), no strong evidence has been found for kinase and the other is phosphorolytic involving phosphorylase, the regulation of amylase, by effectors, in various mesophyll R-enzyme, and phosphoglucomutase. It is not known at the tissues, suggesting the possibility ofa 'futile cycle' ofsimultaneous present time which pathway is dominant in most plants in vivo degradation and synthesis of starch. Since starch is localized in but both are probably functioning. R-enzyme was assayed and the chloroplast the degradation rate could be controlled by the found to be present with most of the activity in the cytoplasm. quantity ofphosphorylase and amylase present in the chloroplast. The same situation was seen with amylase. Hexokinase is also The control point would then be entry of the enzymes into the necessary for the hydrolytic pathway. Two different sets ofassay chloroplast. The substrates for both phosphorylase and amylases conditions were used to assay hexokinase in both the supernatant are amylose and amylopectin which are only present in the 364364ROBINSON AND PREISS Plant Physiol. Vol. 85, 1987 chloroplast. Even though the starch degradative enzymes are 7. HEBER U, M HEBER 1981 The chloroplast envelope: Structure, function, and role in leafmetabolism. Annu Rev Plant Physiol 32: 139-168 present in the cytoplasm there is no substrate available. 8. HEBER U, NG PON, M HEBER 1963 Localization of carboxydismutase and triosephosphate dehydrogenase in chloroplasts. Plant Physiol 38: 355-360 CONCLUSION 9. HEDRICH R, K RASCHKE, M STITr 1985 A role for fructose 2,6-bisphosphate in regulating carbohydrate metabolism in guard cells. Plant Physiol 79: 977- The possible pathway for starch biosynthesis and 982 degradation 10. HORECKER BL 1975 Fructose bisphosphatase aldolase from spinach. Methods have been investigated in guard cells of C. communis (Table I). Enzymol 42: 234-239 The enzymes of the starch biosynthetic pathway are present in 11. KELLY GJ, E LATzKO 1977 Chloroplast phosphofructokinase. I. Proof of the appropriate compartments with sufficient activities to ac- phosphofructokinase activity in chloroplasts. Plant Physiol 60: 290-294 count 12. KELLY GJ, G ZIMMERMANN, E LAzxKO 1982 Fructose-bisphosphatase from for starch synthesis. The starch degradative pathway is spinach leafchloroplast and cytoplasm. Methods Enzymol 90: 371-378 more complex because there are two possible pathways, hydro- 13. KRUGER NJ, E KOMBRINK, H BEEVERS 1983 Pyrophosphate:fructose 6-phos- lytic and phosphorolytic. Also, no activity could be detected for phate phosphotransferase in germinating castor bean seedlings. FEBS Lett the chloroplast isozymes ofhexokinase or PFK in C. communis. 153:409-412 14. LOwRY OH, JV PASSONNEAU 1972 A Flexible System of Enzymatic Analysis. Because of the difficulty in assaying PFK, in the chloroplast, Academic Press, New York and since Hedrich et al. (9) reported lack of activity of the 15. MILLS WR, KW Joy 1980 A rapid method for isolation ofpurified, physiolog- chloroplast isozyme of fructose bisphosphatase, their suggestion ically active chloroplsts, used to study the intracellulardistribution ofamino made earlier that compounds produced by the cytoplasmic fruc- acids in pea leaves. Planta 148: 75-83 or PFK or PFP are 16. NELSON N 1944 A photometric adaptation of the Somogyi method for the tose bisphosphatase substrates for transported determination of glucose. J Biol Chem 153: 375-380 into and out ofthe chloroplast rather than three-carbon units is 17. NOLTMANN EA 1964 Isolation of crystalline phosphoglucose isomerase from of interest. Presumably, sucrose, if it is the source of carbon for rabbit muscie. J Biol Chem 239: 1545-1550 starch biosynthesis, would be degraded in the cytoplasm to 18. OKITA TW, E GREENBERG, DN KUHN, J PREISS 1979 Subcellular localization of the starch degradative and biosynthetic enzymes of spinach leaves. Plant substrates ofPFK or PFP, enterthe chloroplast, and be converted Physiol 64: 187-192 into starch. When the starch is degraded to the level of Fru 6-P 19. OUTLAW WH JR 1982 Carbon metabolism in guard cells. Recent Adv Phyto- it could be exported out of the chloroplast. This sequence of chem 16: 185-222 events would eliminate the problems with both the lack of the 20. OuTLAw WH JR, J MANCHESTER 1979 Guard cell starch concentration quan- tiatively related to stomatal aperture. Plant Physiol 64: 79-82 chloroplast isozyme ofPFK reported here and the lack offructose 21. OUTLAW WH JR, MC TARCZNSKI 1984 Guard cell starch biosynthesis regulated bisphosphatase reported by Hedrich et al. (9). Unfortunately, by effectors of ADPglucose pyrophosphorylase. Plant Physiol 74: 424-429 there is no evidence for his kind of shuttle in mesophyll chloro- 22. OUTLAW WH JR, MC TARCZNSKI, LC ANDERSON 1982 Taxonomic survey for plasts. the presence ofribulose-1,5-bisphosphate carboxylase activity in guard cells. Plant Physiol 70: 1218-1220 23. PREiss J 1982 Regulation of the biosynthesis and degradation of starch. Annu LITERATURE CITED Rev Plant Physiol 33: 431-454 24. ROBINSON NL, E ZEIGER, J PREIss 1983 Regulation of ADPglucose synthesis 1. ANET EFUJ, TM REYNOLDS 1954 Isolation of mucic acid from fruits. Nature in guard cells of Commelina communis. Plant Physiol 73: 862-864 174: 930 25. SHEN L, J PRmSS 1964 The activation and inhibition of bacterial adenosine- 2. BAHR JT, RG JENSEN 1978 Activation ofribulose bisphosphate carboxylase in diphosphateglucose pyrophosphorylase. Biochem Biophys Res Commun 17: intact chloroplasts by CO2 and light. Arch Biochem Biophys 185: 39-48 424-429 3. BERGMEYER HU 1974 Phosphoglucomutase In Methods of Enzymatic Analy- 27. SHIMAzAKI K, E ZEIGER 1985 Cyclic and noncycic photophosphorylation in sis, Ed 2, Vol 1, Academic Press, New York, pp 499-500 isolated guard cell chloroplasts from Vicia faba L. Plant Physiol 78: 211- 4. CsEKE C, AN NISHAWA, BB BUCHANAN 1982 Modulation of chloroplast 214 phosphofructokinase by NADPH. A mechanism for linking light to the 27. STrrr M, PV BULPIN, T AP REEs 1978 Pathway of starch breakdown in regulation ofglycolysis. Plant Physiol 70: 658-661 photosynthetic tissues of Pisum sativum. Biochim Biophys Acta 544: 200- 5. FrrzsimoNs PJ, JDB WEYEmS 1983 Separation and purification of protoplast 214 types from Commelina communis L. leafepidermis. J Exp Bot 34: 55-66 28. TOMLINSON JD, JF TURNER 1979 Pea seed triose phosphate isomerase. Phy- 6. HAWKER JS, IH OzBuN, H OzAiu, E GREENBERG, J PREISS 1974 Interaction tochemistry 18: 1959-1962 ofspinach leafadenosine diphosphate glucose al ,4-al ,4-glucosyl transferase 29. TURNER JF, L COPELAND 1981 Hexokinase II of pea seeds. Plant Physiol 68: in synthesis ofbranched-glucan. Arch Biochem Biophys 160: 530-551 1123-1127