Membranes of Carrot Protoplasts - NCBI

0 downloads 0 Views 1MB Size Report
Jul 9, 1979 - at the following concentrations: 2 ml of 37% (p = 1.210 g/cm3), 9 mlof27%(p= ... rotor and collected from the bottom in fractions of equal volumes except where noted ... ml substrate solution containing 3 mM disodium ATP, 3 mm. MgSO4, 0.1 ... 30 mM Mes-KOH buffer at pH 6.5 at a concentration of approxi-.

Plant Physiol. (1979) 64, 1005-1011

0032-0889/79/64/1005/07/$00.50/0

Isolation and Characterization of Concanavalin A-labeled Plasma Membranes of Carrot Protoplasts' Received for publication June 8, 1978 and in revised form July 9, 1979

WENDY F. BOSS2 AND ALBERT W. RUESINK3 Department of Biology, Indiana University, Bloomington, Indiana 47401 ABSTRACr s of protoplasts rekased from carrot s The pasma culture cel were labeled with I14Clacetyl-concanavalin A. After homogee faction was isolated in a contiuous nizatio a single labeled isopycic Renografin gradient. The labeled membranes peaked at an apparet density of 1.14 grams per cubic centimeter between the Golgi fraction at a denity of 1.11 grams per cubic centmeter as determined by latent IDPase activity and the mitochondria at a density of 1.16 grams per cubic centimeter as determined by the cytochrome c oxidase activity. Thi method provied a very discrete peak of putative asma On discontinuous Reogrfin gradients a relatively pure fimction of beled plasm m bras could be readly isolated at the 1.122 to 1.146 grams per cubic centmeter interface. Tbe labeled fraction was enriced in both a ATPase (pH 6.5) and a glum synthetase with a pH optimum of 6.5 iu and ceflobiose. Enzyme activwhose actvity was promoted by e labeL ities were not altered by the m

Although the details of the role of the plasma membrane in plant growth and development remain a mystery, the plasma membrane has been implicated in such functional roles as cellulose biosynthesis (21), the mediation of auxin responses (13, 25), and polar auxin transport (11). Based on toxin binding to an isolated membrane fraction, Albersheim (2) hypothesized that the plasma membrane may be involved in a host-pathogen response. To elucidate the roles of the plasma membrane in plant growth and development, a method of isolating relatively large quantities of pure plasma membrane is needed. Such an isolation is dependent upon having a specific marker for the plasma membrane. The most commonly used markers for the plasma membrane have been the phosphotungstic acid-chromic acid stain used in electron microscopy (27) and K+-stimulated ATPase measured at pH 6.5 (17); however, there is much controversy as to the reliability of these markers. In addition to being quite tedious and slow, the phosphotungstic acid-chromic acid stain is sometimes nonspecific (12,30). Although K+-stimulated ATPase activity measured at pH 6.5 is associated with the plasma membrane, it is not restricted to the plasma membrane (5; Boss and Ruesink, unpublished results). Hendriks (15) reported an apparent l25I-iodination of plasma membranes of maize coleoptiles with lactoperoxidase; however, Quail and Browning (23) demonstrated the problems of nonspe-

cific iodination of extracellular protein present in Curcurbita tissue segments with this technique. Extensive attempts with this and other covalent labeling methods have proven unsuccessful in our hands (Ruesink, unpublished results), although radioactive diazosulfanilic acid has labeled the plasma membrane of soybean culture protoplasts (9). Anderson and Ray (3) obtained labeling of the putative plasma membrane of pea stem tissue with radioactive UDPG. The selectivity resulted because the UDPG would not penetrate the plasma membrane dunng labeling and only the plasma membrane incorporated label into membrane-bound polysaccharides. By using protoplasts, we procured the advantage of having the plasma membrane readily accessible to markers. The absence of the cell wall permitted the use of a sugar-specific lectin as a membrane label. Con A,4 which binds glycopyranosides such as a-D-glucopyranoside and a-D-mannopyranoside (1), has been shown to agglutinate carrot protoplasts (10); thus, radioactive Con A bound to glycoproteins and glycolipids on the plasma membrane surface should function as a nondisruptive membrane marker. This paper builds upon a preliminary report of labeling with ["4Clacetyl-Con A (4) by detailing the improved techniques for specifically labeling and isolating plasma membranes from protoplasts by relatively gentle means, by describing the unique properties of Renografin as a gradient medium for membrane isolation, and by characterizing the ATPase and glucan synthetase found with the plasma membrane.

MATERIALS AND METHODS Protoplast Isoation. A cell suspension culture of carrot, Daucus carota L., was obtained from Howard Bonnett and grown in a defined medium (7) containing 1.0 mg/l naphthalene acetic acid and 0.02 mg/l kinetin, with subculturing every 4 days. To release protoplasts, approximately 1.0 ml packed volume of carrot suspension culture cells was incubated in 20 ml of 2% (w/v) Driselase in 0.4 molal sorbitol for 2 h at 28 C on a rotary shaker (150 rpm). The released protoplasts were filtered through glass wool to remove large fragments of cytoplasmic debris and centrifuged out of the enzyme at 40g for 5 to 10 min. The osmoticum to be used during labeling (0.43 M mannitol and 0.01 M CaCl2 except where noted) was added slowly and the protoplasts were washed three

times. Protoplast Labling and Homogenization Washed protoplasts were incubated in the presence of [14CJacetyl-Con A (see synthesis below) for 2.5 min at 22 C (100 ,ug [I4CJacetyl-Con A/0.5 ml packed volume of protoplasts). The labeled protoplasts were 1 Supported by National Science Foundation Grants GB 39950 and washed three times with cold (10 C) osmoticum (0.45 M mannitol). PCM 76-15222. Represents part of a thesis presented by W. J. Boss to the Further operations were carried out at 4 C. The washed protoplasts Graduate School of Indiana University in partial fulfillment of the requirements for the Ph.D. degree. 2Present address: Department of Botany, North Carolina State Univer4Abbreviations: Con A: concanavalin A, PVPP: polyvinylpolypyrroliNorth Carolina 27607. done; THM: Tris-homogenizing medium (0.05 M Tris, 10 mm KCL 1.0 mm sitl,ToRaleigh, whom all correspondence should be addressed. EDTA, 0.1 mm MgC12 [pH 7.5J). 1005

1006

BOSS AND RUESINK

(about 0.5 ml) were homogenized with 10 passes of a TenBroeck homogenizer in 8 volumes of 8% (v/v) Renografin and 0.05 g insoluble PVPP in THM. The homogenate was centrifuged at 1,OOOg for 2 min and the supernatant used for further membrane purification. Discontinuous Renografin Gradients. Discontinuous gradients were made immediately before loading using Renografin in THM at the following concentrations: 2 ml of 37% (p = 1.210 g/cm3), 9 ml of 27% (p= 1.146 g/cm3), 8 ml of23% (p= 1.122 g/cm3), 8 ml of 18% (p = 1.093 g/cm3), and 7 ml of 10%o (p = 1.049 g/cm3). The density was calculated from the refractive index according to our empirically determined equation P21° = 3.48IN4 - 3.643. Tubes were centrifuged at 84,000g for 2 h and the bands at all interfaces were collected as noted in Figure 1. An aliquot of each band was measured for radioactivity. The remainder was diluted with 0.05 M Tris at pH 7.5 and either layered onto a continuous gradient for further purification or centrifuged at 30,000g for 1 h, washed twice, and used for enzyme assays. Continuous Renografin Gradients. The supernatant of l,OOOg centrifugation or diluted fractions of the discontinuous gradient were layered onto linear gradients ranging from 10 to 38% (p = 1.049 to 1.216 g/cm3) Renografin in THM as noted above. The gradients were centrifuged at 80,000g for 15 to 18 h in an SW 27 rotor and collected from the bottom in fractions of equal volumes except where noted otherwise. Synthesis of [I4CIAcetyl-Con A. The acetylated derivative of Con A was synthesized using [14Cjacetic anhydride (120 mCi/ mmol) according to the procedure of Noonan and Burger (22). The lyophilized product (376 mCi/mmol, based on the mol wt of the tetramer) was stored at -20 C. Working stock solutions of 1.0 mg/ml in 0.1 mM MnCl2 and 5 mm Tris at pH 7.5 were stored at 4 C for several weeks with no loss of activity. Enzyme Assays. Renografin interferes with both the Pi determinations and the Lowry protein assay; thus, ATPase activity, IDPase activity, and protein content were determined on membranous material which pelleted when gradient fractions were diluted with 0.05 M Tris at pH 7.5 and centrifuged at 30,000g for 1 h. Pellets were washed twice, resuspended in 30 mM Tris-Mes buffer of the appropriate pH containing 0.1 mim DTT, and kept on ice until assayed. ATPase was measured at pH 6.5 using 0.1 ml enzyme and 0.9 ml substrate solution containing 3 mM disodium ATP, 3 mm MgSO4, 0.1 mM DTT, and 55 mM KC1. Latent IDPase was assayed at pH 7.5 by adding 0.2 ml of enzyme solution which had been kept at 4 C for 2 to 4 days to 0.7 ml of substrate solution containing 1.2 mM IDP, I mM MgSO4, 0.1 mm DTT, and 55 mm KCI. In both cases the reaction mixtures were incubated for 15 to 30 min at 38 C, and the reaction terminated with ice-cold trichloroacetic acid (5% final concentration). Inorganic phosphate was determined according to Taussky and Shore (29). UDP-Glucose:glucan glucosyltransferase activity was measured on washed, pelleted membranes which had been resuspended in 30 mM Mes-KOH buffer at pH 6.5 at a concentration of approximately 1OO Lg protein/0.l ml for microsomal fractions and 50 jtg

Plant Physiol. Vol. 64, 1979

protein/0. 1 ml for discontinuous gradient fractions. To 0.1 ml of membrane solution was added 0.02 ml of substrate solution to give final amounts of 2.6 ,umol of MgC12, 2 utmol of cellobiose, 0.2 ,umol of UDP-glucose, and 38.5 pmol of UDP-['4CJglucose (312 mCi/mmol). After 10 min at 28 C (water bath) with intermittent shaking, the reaction was terminated by adding 1.5 ml of CHC13methanol-0.04% (w/v) CaCl2 in H20 (86:14:1) (8). The organic phase was removed and carrier wall (ground tissue culture wall extracted by urea and ethanol) was added to the aqueous phase. The wall material was washed once with methanol, twice with 1 N NaOH (boiling water bath for 5 min), and twice with water. The final pellet was neutralized with 1 drop of 15% (w/v) ascorbic acid and counted in a scintillation counter as described below. Cyt c oxidase and NADH-Cyt c reductase were assayed spectrophotometrically (24). Cyt c oxidase activity was more stable in Renografin than in sucrose and could be maintained on ice for at least 36 h with no decrease in activity. Proteins were estimated by the Lowry procedure (20) with BSA as a standard. Determination of Radioactivity. Radioactivity was measured with a Beckman 233 scintillation counter using a scintillation cocktail of 8% Scintisol-GP in a Beckman fluoralloy-toluene mixture. Renografin concentration was diluted with water to 20% or less prior to counting. Chemicals. Driselase (lot K26018) was purchased from Kyowa Hakko Kogyo Co., Ltd.; Renografm-60 (a mixture of the Nmethyl-D-glucamine and sodium salts of 3,5-diacetamido,2,4,6triiodobenzoic acid) from E. R. Squibb and Co.; ["4Clacetic anhydride (120 mCi/mmol in 5% benzene) and (312 mCi/mmol) from Amersham/Searle; PVPP (water-insoluble) from GAF Corporation; Scintisol-GP from Isolab; Con A (grade IV), disodium ATP, IDP, BSA, DTT, Cyt c, and NADH from Sigma Chemical Co.; and UDP-glucose from Calbiochem.

UDP-[14CJglucose

RESULTS

Protoplast Agglutination and Labeling with Con A. The conditions for the Con A treatment were adjusted to enhance both protoplast viability and plasma membrane labeling. If the protoplasts were labeled for 2, 5, or 10 min with ['4CJacetyl-Con A the specific activity of the isolated plasma membrane was the same. Thus, a 2.5-min incubation time was chosen to minimize the possible uptake of Con A by pinocytosis. Con A agglutinated protoplasts more readily in an osmoticum of 0.05 M Tris, 0.01 M CaC12, and 0.25 M KCI at pH 7.2 (10) than in an osmoticum of 0.43 M mannitol or sorbitol with 0.01 M CaC12 in 0.05 M Tris at pH 7.2 or simply 0.45 M mannitol or sorbitol. In contrast, we found no significant difference in the amount of I14C]acetyl-Con A bound to the protoplasts in any of these osmotica. Since the protoplasts were in the best condition as determined by phase contrast microscopy in an osmoticum of 0.43 M mannitol or sorbitol and 0.01 M CaC12, a mannitol-CaCl2 osmoticum was used for the labeling studies. The final three washes after labeling the protoplasts with Con A were Ca-free to minimize aggregation problems during the membrane isolation procedure. AlIsolation of I ClAceyl-Con A-labeled Plasma Membrane. culture from carrot suspension though protoplast preparations FRACTION p (g/cm3) cells contained less debris than protoplasts of most species, any 1.039 A wall fragments or free plastids in the preparations readily bound lo% L049 Con A. During the membrane isolation procedure, most of this B debris was removed by centrifuging the homogenate at l,000g 18% 1.093 prior to layering it onto a gradient. Differential centrifugation C 1.122 23% proved incapable of giving further purification. Most (75%) of the -* D counts sedimented with the mitochondria at 8,000g. The entire 27% 1.146 1,000g supernatant was therefore used for further purifications on x E 1.211 7:;.: density gradients. Using a sugar specific lectin such as Con A as a membrane FIG. 1. Discontinuous Renografm gradients. Gradients were centrifuged at84,000g for 2 h. Fractions were collected with a disposable pipette label, we were limited in our choice of density gradient media. at the interfaces noted. Renografin proved to be the best.

1007

ISOLATION OF PLASMA MEMBRANES Plant Physiol. Vol. 64, 1979 Membrane on Condnuos Renografin DIstbeton of Gr t Figure 2 shows the isopycnic gradient profile after centrifuging the 1,000g supernatant into a continuous gradient of 11 to 38% Renografin. After ovenight centrifugation, Renografin

gradients are nonlinear at the ends. The major band of radioactivity appears at a density of 1.14 g/cm3, sedimenting between the mitochondria at 1.16 g/cm3 and the Golgi apparatus at 1.11 g/ cm3. The ER, although a broad band in this particular gradient, charcteristically gives a much sharper peak at 1.08 to 1.09 g/cm3

E

06

3.5

0E

Pt ~~040...'

3.0 Z

(e.g. Fig. 4).

p ~~~~~-0260

2.5

d

Although the protoplasts were well washed after being labeled and before being homogenized, some radioactivity was subsequently detected on the upper regions of the gradient. After 2 h of centnrfugation, this band was found at a density of 1.04 to 1.05 g/ cm3 (Fig. 7) while after 18 h the radioactivity had penetrated to 1.05 to 1.06 g/cm3. This shift from 2 to 18 h was typical of soluble protein and not membrane particulates, which reach equilibrium density within 2 h in our system (f. Figs. 2 and 7). It was also determined, as will be disussed in a later section, that there was no corresponding peak of glucan synthetase activity assocated with the upper band of ['4Clacetyl-Con A. These data suggested that the lighter band of Con A was not bound to particulate membrane components. To determine if any Con A released would bind to other membrane fractions during homogenization, the protoplasts were homogenized in the presence of 20,000 cpm of 1I HCacetyl-Con A (Fig. 3). The major peak of radioactivity in Figure 3 corresponds to that observed when ["4CJacetyl-Con A alone was put on a similar continuous gradient (1.057 g/cm3, Boss and Ruesink, unpublished) and to the upper peak found on continuous gradients oflabeled protoplasts. Con A is apparently binding predominantly to soluble materials from the cytoplasm, with no tendency to move to any membrane fraction. Thus, one would not expect any Con A released from the protoplast surface during homogenization to accumulate on other membranes because of differences in binding affinities. The amount of Con A released during homogenization and remaining in an upper band varied with the condition of the protoplasts. When a Tris-KCI-CaCl2 osmoticum (see pargraph 2 under "Results") was used, the amount of low density Con A released during homogenization was greater than with a mannitol


90%o as determined by Cyt c activity) as well as wall debris and plastids which did not sediment at l,000g prior to layering the homogenate on the gradient. Con A binds to isolated cell wall; thus any wail debris present in the protoplast preparation will be labeled and small fragments that do not xediment readily at 1,000g will be carried over into the gradient. The continuous gradient profiles (Figs. 2 and 4) show that labeled debris is present which sediments below the mitochondria and contributes to the radioactivity found in fraction E (Table I). The

ml

10mnm A 21.3 ± 6.3 1.05 + 0.98 Not measured B 13.9± 1.6 1.8± 1.1 23+4 C 16.2+4.4 4.2±+1.1 102± 13 D 21.0±4.8 0.8+0.1 142+ 19 E 6.4±3.5 0.4+0.1 0.7+0.7 a Representative values. b Mean of three or four experiments using different membrane preparations. c Mean of four determinations using the same membrane preparation. D

flu.~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ c

4.0 3.5Z

3.0

4

2.5-

I 6 2.0 x

2

0

240o.

.15 -1.1 4

-I

,~~~~ \

'. * * *i

-1, 2 -1 20 -I .1 9 -1. 8 -.1 7 .1 6

200 E 80

1.13

4 (DE0 .1 2 260 200 ~~~~~~~~-: -240 -1.12

'\

-I 10

1.

09

+

1.0

206N

O. \0

3

5

20

,,

7

11 9 13 15 FRACTION NUMBER

17

6

08 07 06 05

E

v

19

FIG. 5. Distribution of various markers following a continuous gradient separation of the membranes in fraction D. Fraction D was diluted, layered onto a 15 to 35%, 17-ml continuous Renografm gradient, and centrifuged for 12 h at 84,000g (SW 27 rotor). Equal 0.7-ml fractions were collected from the bottom except for the overlay of 3 ml (fraction 1). Cpm were measured on an aliquot of the fractions. The remaining fractions were diluted, centrifuged at 30,000g for 1 h, washed twice, and assayed for ATPase activity (0- - -0).

Plant Physiol. Vol. 64, 1979

1009

ISOLATION OF PLASMA MEMBRANES

When layered onto a continuous gradient in the same manner as fraction D, fraction C yielded one peak of radioactivity (Fig. 6). Latent IDPase activity was maximum at a density of 1.11 g/cm3 and the peak of radioactivity sedimented slightly below that at 1.125 g/cm3, indicating that the radioactivity in this fraction was not a result of Golgi labeling. When fractions C and D were combined and layered onto a continuous gradient as above, a peak of radioactivity was found at 1.14 g/cm with a definite shoulder at 1.12 g/cm3. The ATPase activity exhibited a similar peak and shoulder. Attempts to show that the shoulder and peak represented two discrete classes of plasma membrane vesicles were inconclusive. Glucan Synthetase Assay. The optimum conditions for the UDP-glucose-glucan glucosyltransferase activity were determined using both a microsomal pellet (8,000-30,000g) and the discontinuous gradient fraction D. The reaction was linear up to 10 min with membranes isolated from protoplasts of cells in the late log phase of growth. The pH optimum was determined to be 6.5 with significant activity found over a broad range from 6.0 to 7.0. Mg and cellobiose were required for optimum activity. A continuous gradient profile of glucan synthetase activity is shown in Figure 7. Although the gradient was centrifuged for only 2.5 h at 84,000g, it is apparent by comparison with Figure 4 that the membrane vesicles had reached their equilibrium density. After measurinq the Cyt c oxidase, NADH-Cyt c reductase, and the amount of '4C]acetyl-Con A in each fraction, the remaining material was diluted with THM, centrifuged at 30,000g for 45 min, washed once in Mes-KOH buffer at pH 6.5, and assayed for glucan synthetase activity. The glucan synthetase activity closely conformed to the peak of ["4Clacetyl-Con A except at the top of the gradient, further evidence that the upper peak represents nonmembrane-bound Con A. Since the glucan had been washed twice in boiling 1 N NaOH, [14Clacetyl-Con A did not interfere with the analysis of the radioactive glucan. ['4C]Acetyl-Con A-labeled boiled enzyme blanks were equivalent to nonlabeled blanks. The distributions on discontinuous gradients (Table I) reveal that both the glucan synthetase activity and the Con A label peak in fractions C and D. Fraction E, the mitochondria-rich fraction, has little glucan synthetase activity, supporting our premise that the majority of the Con A in fraction E is bound to wall debris or plastids and not to plasma membrane. Although Con A cannot be used as a plasma membrane marker for whole cells because it binds to cell walls, it is possible to measure the glucan synthetase distribution from a whole-cell homogenate. Whole cells were ground with sand in a mortar and pestle containing the same homogenizing medium used for protoplasts, and treated in the same manner as the protoplast mem-

FRACTION NUMBER

FIG. 6. Distribution of various markers following a continuous gradient separation of the membranes in fraction C. Procedure exactly as for Figure 5 except that the assay was for latent IDPase instead of ATPase activity.

FRACTION

NUMBER

FIG. 7. Glcnsynthetase activity in a continous 10 to 38% Renografin gradient centrifuged for 2.5 h at 84,OOOqg. Fractions of 1.6 ml each were collected and assayed as in Figure 2. Each remaining fraction was diluted with THM, centrifuged at 30,000g for 1 h, washed once in Mes buffer at pH 6.5, and assayed for glucan synthetase activity measured in nmol of ['4CJglucose incorporated into a NaOH-insoluble pellet in 10 min

brane preparations. If we had altered the density of the plasma membrane during protoplast release and labeling, the specific activity of the glucan synthetase in discontinuous fractions should be changed. This was not the case. For instance, the specific activities in fractions C, D, and £ of whole-cell homogenates were 193, 268, and 10 nmol of glucose incorporated/mg protein in 10 min compared to values of 152, 188, and 16 nmol/mg protein found for isolated membranes of protoplasts released from cells harvested 4 h earlier on the same day. Since fraction E of the protoplast homogenate incorporated more glucose than a similar fraction from whole cells, it is apparent that removing sugars enzymically from the plasma membrane has not lightened the membranI e marked th above results also show that Renografin can be used to separate homogenates of whole cells as well as of protoplasts. Effects of Con A on Plasma Membrane. Testing if Con A altered the plasma membrane of carrot suspension cell protoplasts, we found no difference in the distribution or amount of ATPase, latent IDPase, Cyt c oxidase, or NADH-Cyt c reductase activity on discontinuous gradients from protoplasts treated with and without Con A, and there was no change in the appearance of membranes on the gradients. Since glucan synthetase seemed to be the most specific enzyme marker for the plasma membrane and sincmthe substrate for the enzyme (ruDP-glucose) might be particularly sensitive to sugar-specific Con A, the enzyme was studied in more detail. Protoplasts were labeled with Con A and the glucan synthetase activity was measured on a microsomal membrane fraction and fraction D of the discontinuous gradients. Values are expressed as umol of glucose incorporated/mg protein in 15 mmn. Using the 8,000 to 3C,oOOg pellet, we measured values of 100 with Con A and 101 in its absence. With fraction D we found 175 with Con A and 166 without it. Due to the inherently high variability (an average of 10%), individual data points showed considerable scatter, however, at no time was there a significant difference in the synthetase activity with or without Con A either in the results summarized above or in the activity of membranes isolated on continuous gradients. Con A does not alter protoplasts structurally as determined by phase and Nomarski optics in this laboratory and by electron microscopy in other laboratories (6, 33). In short, no effect of the Con A treatment could be detected on the enzyme activities or gradient distributions of any of the membranes studied.

1010

BOSS AND RUESINK

DISCUSSION Radioactive Con A can be used as an effective, specific, nondestructive readily detectable plasma membrane label for carrot protoplasts. In the absence of cell walls, it exhibits excellent qualities as a plasma membrane label, altering neither the position of membranes on density gradients nor the activity of the enzymes tested. Because protoplasts are washed well before and after labeling, the problem of labeling extracellular protein (15, 23) is In our hands, Con A-labeled plasma membranes did not sediment in large sheets at low g-forces as did those of Neurospora (28), for reasons that are not clear. Whether the Con A is binding to the plasma membrane proper or to nascent wall materials associated with it has not been determined and is not important for identification purposes so long as the label does not move around. Other work has shown that there is no simple relationship between the amount of new wall present and the amount of Con A bound (32). Renografin was chosen as a gradient medium because of its unique properties and because other common gradient media could not be used. Sucrose displaces Con A from the membrane and glycerol proved unsatisfactory due to its high viscosity and to the high concentrations needed for the required densities. Solutions of Renografm are denser but less viscous than solutions of glycerol, sucrose, or Ficoll of equal concentrations (w/v). In solutions of equal densities, Renografm has a lower osmolality than sucrose and glycerol but higher than Ficoll. In addition to permitting Con A labeling, the combination of Renografin separation with carrot protoplasts appears to provide a better separation of plasma membrane vesicles from Golgi and mitochondria than that reported by many others (3, 9, 17, 24). The apparent densities of membranes isolated on isopycnic gradients will depend on the mass of the membrane vesicle as well as the hydrated volume of the vesicle, and thus will vary with the permeability and osmotic activity of the gradient medium (31). Vesicles in permeant gradient media (glycerol or CsCb sediment at a density close to that of the membrane itself, in contrast to the situation in impermeant media where vesicles equilibrate at lower densities. In the latter, osmotic effects are important. Sucrose has much more osmotic activity than FicolL shrinks the vesicles, and causes them to sediment to higher densities than in Ficoll. In solutions of equal densities, Renografin has an osmotic effect intermediate between sucrose and Ficoll (less than 0.5 that of sucrose) and vesicles should equilibrate accordingly. Renografin does not appear to alter intact or isolated plasma membranes and has been used by others to isolate a number of kinds of cells and organelles (see 26). Due to its low osmotic activity and low permeability, Renografin enhances the separation of vesicles of different volumes (31). Although Renogafin is in many ways an excellent gradient medium for separating membranes, there are several problems. The low viscosity of Renografin affords rapid equilibrium centrifugation; however, it permits some undesirable mixing of the gradient medium at the discontinuous gradient interfaces. Renografin absorbs light at 280 nm, and it interferes with the Lowry assay. It precipitates in trichloroacetic acid and with some ions such as molybdate and high concentrations of Mg. At concentrations greater than 20% (v/v) it temporarily quenches scintillation fluor. Renografin is an excellent gradient medium but must be removed before many standard biochemical assays can be carried out on the isolated membranes. Because ATPase has been proposed as a potential plasma membrane marker, its distribution in our system was studied with great interest Our maximum rate compares favorably with reports of other plasma membrane systems (e.g. 18) on the basis of phosphate released per mg protein. In this carrot tissue culture system, the ATPase seems somewhat more restricted to the plasma

Plant Physiol. Vol. 64, 1979

membrane than with some other systems (5) but it still is inadequate to locate the plasma membrane defnitively, especially in the absence of information about the location of tonoplast ATPases (19). K+ ion was included regularly in the incubation medium even though our ATPase exhibits little K+ stimulation. Toward the end of this work, the pH optimum was found to be closer to 6.0 than to the 6.5 used for these experiments. Although both the K+ATPase and glucan synthetase have been suggested as plasma membrane markers (see 16), they could sometimes be separated somewhat in the present work and when maize coleoptile membranes were fractionated on sucrose (16). In both cases a portion of the ATPase activity remained near the top of some gradients for reasons not yet clear. Here, the glucan synthetase seems better correlated with plasma membrane than does the ATPase. Although other workers use pH 8.0, the pH optimum for this glucan synthetase is distinctly 6.5, a pH closer to that expected at the cell surface. The nature of the product has not yet been determined. Because of its association with the plasma membrane, we are investigating the possibility that it is cellulose. Tobacco protoplasts were used for our early work with Con A. Although they agglutinated very well, they were more fragile and difficult to work with than carrot protoplasts. Similar gradient profiles were obtained with the tobacco, but the recovery of radioactive Con A was much less, with more of the label being lost during homogenization. The vacuolar membrane has not been identified on our gradients. There is no reason whatever to think that it is the Con Alabeled moiety, but on the other hand it is impossible to rule out the possibility that the tonoplast is contaminating our plasma niembrane preparation as it has contaminated others (e.g. 14). Attempts to isolate vacuoles from the carrot protoplasts so as to determine the location of the tonoplast on Renografm gradients have been unsuccessful. The following evidence indicates that we have isolated plasma membrane vesicles: (a) Con A bound to the plasma membrane of protoplasts and caused them to agglutinate; (b) labeled protoplasts which were treated with glutaraldehyde gave the same peak of labeled membranes on continuous gradients as those not fixed; (c) marker enzymes purported to be associated with the plasma membrane were found to co-sediment with the Con A-labeled membrane fraction; (d) there was no change in the distribution of activity of the marker enzymes used whether the protoplasts were treated with or without Con A so that the enzyme analysis was not-an artifact of labeling; (e) the labeled membrane fraction did not co-sediment with any known marker for other intracellular membranes such as mitochondria, Golgi, or ER. These studies have determined another way to label and isolate the higher plant cell plasma membrane. They provide the ground work for future research on the roles of the plasma membrane in plant growth and development. LITERATURE CITED 1. AGRAWAL BBL, I} GOLDSTEN 1972 Concanavalin A, the jack bean (Canavalia ensformis) phytohemagglutinin. Methods Enzymol 28: 313-317 2. ALBELusm P, AJ ANDMRSON-PROUTY 1975 Carbohydrates, proteins, cell surfaces and the biochmistry of pathogenesis. Annu Rev Plant Physiol 26: 31-52 3. ANDmtsoN R, PM RAY 1978 Labeling of the plasma membrane of pea cells by a surfacelcalized glucan synthetase. Plant Physiol 61: 723-730 4. Boss WF, AW RuEssnc 1976 Concanavalin A as a plasma membrane labeL Plant Physiol 57: S-17 5. Bowvs DJ, H KAuss 1976 Characterization, enzymatic and lectin properties of isolated membranes from Phawohls awws. Biochim Biophys Acta 443: 360-374 6. BuRGEss J, PJ LINSTEAD 1976 Ultrastructural studies of the binding of concanavalin A to the

plasmalemma of higher plant protoplasts. Planta 130: 73-79 7. EssoN T 1965 Studies on the growth requirements and growth measurements of cell cultures of Haplopappus gracils. Physiol Plant 18: 976-993 8. FOLCH J, M Los, GHS STANLEY 1957 A simple method for the isolation and purification of total lipids from animal tissues. J Biol Chem 226: 497-509 9. GALB.ArrH DW, DH NoRTHcoTE 1977 The isolation of plasma membrane from protoplasts of soybean suspension cultures. I Celi Sci 24: 295-310

Plant Physiol. Vol. 64, 1979

ISOLATION OF PLASMA MEMBRANES

10. GLDMLIUS K, A WAU,L1, T EIxsON 1974 Ag inAtintg effects of oncanavalin A on isolated protoplas of Daucus carola. Physiol Plant 31: 225-230 11. GoLDwsam MHM 1977 The polar tran of auxiL Annu Rev Plant Physiol 28: 439-478 12. Hu JIL, Tl FLowaas 1976 Properties of membranes from the halophyte Suaeda marbna. I. Cytocheical aining of membrnes in relation to vaidity of membrae marker. I Exp Bot 27: 658672 13. H2UDiN JW, JH CmHY, DJ Moazt, CA L1ates 1972 Enhancement of RNA polymerae activity by a factor released by auxin from plam membrae. Proc Nat Acad Sci USA 69: 3146-3150 Of 14. HEasoN SL, WA CAmsE, DJ Moai 1976 Evidence for an inCre in plasma membranes from soybean hypocotyla induced by the plant hormone, indoI-3-acetic acid. Plant Physiol 58: 548-551 15. HsEmRs T 1976 lodination of maize coleotil: a possible method for identifying plant plasma membranes. Plant Sci Lett 7: 347-357 16. HENDR T 1978 The distribution of glucan synthetas in maiZe colepl: a comparison with K-ATPase. Plat Sci Left 11: 261-274 17. HODGES TK, RT LIONAD 1974 Purification of plsma membrane-bound adenosin triphosphatase from plant root. Methods Enzymol 32: 392-406 18. LEONAD RT, WI VANDRWouDE 1976 solation of plasm membrnes from corn roots by sucrose denty gradient cetristi An anomalos effect of FicolL Plant Physiol 57: 105-114 19. LIN W, GJ WAGNE, HW SIBGEAtaN, G HtND 1977 Membrane-bound ATPmw of intact vacuoks and tonoplas isolased frm mature plant tse. Biocim Biophys Acta 465: 110117 with the 20. LowRy OH, NJ RosmstowH, AL FA, RJ RANDALL 1951 Protein m Folin phenol reaent. I Biol Chem 193: 265-275 21. MoNTEzNos D, RM BROWN JR 1976 Surface architecture of the plant celL biogenesis of the

22. 23. 24.

25. 26. 27.

28. 29. 30. 3 1. 32. 33.

1011

cell wall, with special emphais on the role ofthe plasma membrane in celluloe biwynthes. J Supmol Strt 5: 277-290 NOONAN KD, MM BURGER 1974 An asay for labeled lctin binding to cell srfac Methods Enzymol 32: 621-625 QUAL PH, A BROWNING 1977 Failure of oxdse to odinae speficay the plasm membrne of Cwwbita time segments, Plant Physo 59: 759-766 RAY PM 1977 Auxin-binding sites of mae coleptile are lcalized on membranes of the endoplasmc ricL Plant Physiol 59 594-599 nt by hydrogen ions Curr Top RAYLE DL, R CLELND 1977 Control of plant oel enla Dev Bjol 11: 187-214 fugtlon in non-ionic media, In GD Birnle, D Rickwood, RIcKWooD, D 1978 Ioycnic eds, Centrifugal Sqeao in Molcular and Cell Biology. Butteorth, London, pp 219250 RoAND IC, CA I e, DJ Moaa 1972 Phosphsotungc acid-chromic acid as a selecive eectro-den stain for plama membranes of plant ceil Stain Technol 47: 195-200 zation of Nervspo cram plasma memSCARBOROUGH GA 1975 Isolation and branes J Biot Chem 250: 1106- 111 TAusssY HH, E SHosx 1953 A microcolormic method for derminaton of inognic pkosphor. I Blol Chem 202. 675-685 THoM M, WM IAErscHs, A MAREm 1975 Isolaton of membrnes from sugarcane cel suapenalos evidene for a plasma membrane enrichd fracto Plant Sci Lut 5: 245-253 WAU,ACH DFM 1967 Iolti of plm membranes of animal ceall BD Davir, L Warrn, eds, The Specificty of Cdl Surface Prentice-Hall Englewood Cliffs, NJ, pp 145-150 WnALON FA 1979 Co valin A binding sites an the plam membrne of leck stem pr nplast. Planta 144: 209-215 WAsusON FA, LC Fows:A FC CoNsrABEm, OL GAsaoG 1976 Labeing of co valin 89:305-316 A sites on the plasma membrne of soybean protopasts. Pt