Characterization of the Major Integral Protein of Vacuolar Membrane

Plant Physiol. (1992) 98, 1248-1254 0032-0889/92/98/1 248/07/$01 .00/0

Received for publication August 26, 1991 Accepted October 15, 1991

Characterization of the Major Integral Protein of Vacuolar Membrane1 Masayoshi Maeshima Institute of Low Temperature Science, Hokkaido University, Sapporo 060, Japan ABSTRACT

TIP2 has been proposed to be involved in the transport of small molecules. In the present work, a major protein in the vacuolar membranes of radish taproot was isolated and its antibody was prepared to make it available for future cloning and for biochemical studies of cellular function. The protein, tentatively called VM 23, has high hydrophobicity and interacts with DCCD. Reported here are its properties and its presence in vacuolar membranes from other plant species. Also, the properties of VM 23 of central vacuoles and TIP of proteinstorage vacuoles are compared.

The vacuolar membrane of radish (Raphanus sativus) taproot contained a large quantity of a protein of 23 kilodaltons that accounted for more than 25% of the total membrane proteins. The protein, tentatively named VM 23, was purified and characterized. VM 23 tends to aggregate at high temperature even in the presence of 1% sodium dodecyl sulfate. The apparent molecular size of VM 23 was estimated to be about 400 kilodaltons by polyacrylamide gel electrophoresis in the presence of 0.1% Triton X-100. VM 23 was partially extracted from the vacuolar membranes with chloroform:methanol, indicating its high hydrophobicity. The hydrophobic carboxyl modifier N,N'-dicyclohexylcarbodiimide bound covalently to VM 23. The results suggest that VM 23 may act as a secondary transport system coupled with the proton transport. The antibody against radish VM 23 reacted with the major proteins in the vacuolar membranes of mung bean (Vigna radlata) and castor bean (Ricinus communis) hypocotyls and pumpkin (Cucurblta moschats) epicotyl, but not with that of sugar beet (Beta vulgarls) taproot. VM 23 comigrated with vacuolar H+pyrophosphatase on sucrose density gradient centrifugation after sonication of membranes, indicating that it is associated with the vacuolar membrane.

MATERIALS AND METHODS Plant Materials

Taproots of radish (Raphanus sativus L. cv Miyashigedaikon) and sugar beet (Beta vulgaris L. var rapa Dumort) harvested in Sapporo were used to prepare vacuolar membranes. Mung bean ( Vigna radiata) seeds harvested in China were imbibed with water and germinated on nets floated on water in the dark at 26°C for 3.5 d (1 1). Castor bean (Ricinus communis) seeds were soaked for 18 h and germinated in moist vermiculite in the dark at 30°C for 4 d (10). Pumpkin (Cucurbita moschata L. cv Kurokawa Amakuri) seeds were germinated in moist vermiculite in the dark at 26°C for 6 d. Hypocotyls from seedlings of mung bean and castor bean, and epicotyls of pumpkin seedlings were removed and used for membrane preparation.

Plant cells have central vacuoles mostly containing salts, metabolites, and water. The vacuoles play a role in controlling turgor pressure and the cytosolic levels of inorganic ions and metabolites (1). Several kinds of ions and metabolites are accumulated in the vacuole and then released in response to the physiological conditions of cells. The vacuolar membranes have many transport systems specific to each solute (1, 19). Transport mechanisms remain to be elucidated at the molecular level. Among the systems on the vacuolar membrane, two proton pumps, H+-ATPase (15, 16, 18, 25) and H+pyrophosphatase (2, 11, 14, 21), have been purified from plant vacuoles and characterized, but the molecular details of other secondary transport systems are not clear. Biochemical and molecular analyses ofthe protein components of vacuolar membranes may provide important information for understanding the transport machinery and the mechanisms of regulation of the transport activities. Recently, an abundant membrane intrinsic protein in the protein-storage vacuoles of seeds has been characterized and sequenced (5-7, 17). This

Preparation of Vacuolar Membranes Vacuolar membranes were prepared basically as described previously (1 1). The parenchymatous tissues of taproots from radish and sugar beet were homogenized in a grinding medium using a grater. The hypocotyls of mung bean and castor bean and the epicotyls of pumpkin were chopped with a razor blade and then ground in a mortar. The grinding medium contained 0.25 M sorbitol, 5 mM EGTA, 1 mM PMSF, 1% (w/v) PVP, 1% (w/v) ascorbic acid (neutralized with KOH), and 50 mM Tris-acetate, pH 7.5. The homogenate was filtered and centrifuged at 3600g for 10 min. The supernatant fluid was centrifuged at 1 20,000g for 20 min. The precipitate was resuspended in 15 mL of 20 mM Tris-acetate, pH 7.5, 0.31 M sucrose, 1 mm EGTA, 1 mM MgCl2, and 2 mm DTT, and 2 Abbreviations: TIP, tonoplast integral protein; DCCD, N,N'dicyclohexylcarbodiimide; VM 23, vacuolar membrane integral pro-

' Supported by Grants-in-Aid for Scientific Research (No. 03257201 and No. 03660072) from the Ministry of Education, Science and Culture of Japan.

tein of 23 kilodaltons. 1248

INTEGRAL PROTEIN OF VACUOLAR MEMBRANE placed in 40 mL centrifuge tubes. The suspension was overlaid with 10 mL of 0.25 M sorbitol, 1 mm EGTA, 1 mM MgCl2, 2 mM DTT, and 20 mm Tris-acetate, pH 7.5. After centrifugation at 120,000g for 30 min, vacuolar membrane vesicles forming bands at the interface between the two solutions were collected and suspended in 10 mL of 0.25 M sorbitol, 1 mM EGTA, 1 mM MgCl2, 2 mM DTT, and 20 mM Tris-acetate, pH 7.5. The suspension was then centrifuged at 1 30,000g for 20 min, and the white pellet was suspended in 2 mL of the same buffer. The resulting suspension was the highly purified preparation of vacuolar membranes as judged by the specific activities of marker enzymes for other organelles such as plasma membrane (1 1, 16). Purification of VM 23 To the vacuolar membrane suspension (2 mg protein/mL) prepared from radish were added solid KCI and 5% sodium deoxycholate to final concentrations of 50 mm and 2 mg/mg of protein, respectively, and the suspension was centrifuged at 1 50,000g for 30 min. The pellet was suspended in 20 mM Tris-acetate, pH 7.5, 20% (w/v) glycerol, 1 mM DTT, 1 mM EGTA, 2 mM MgC12, and 0.4% (w/v) lysophosphatidylcholine (egg yolk, type I) to make up half the volume of the original membrane suspension. The suspension was centrifuged at 150,000g for 40 min at 8°C after it had been stirred for 10 min at 25°C. The supernatant fraction was applied to a column (bed volume, 3 mL) of QAE-Toyopearl (Tohso, Japan), which had been preequilibrated with 20 mM Tris-acetate, pH 7.5, 20% glycerol, 1 mM DTT, 1 mM EGTA, and 2 mM MgCl2 (Tris/GDEM). The column was washed with 10 mL of Tris/GDEM buffer containing 0.12 M NaCl, and VM 23 was eluted from the column with 10 mL of Tris/GDEM buffer containing 0.22 M NaCl at a flow rate of 0.4 mL/min. The content of VM 23 in the fractions was estimated by SDSPAGE. Further purification of VM 23 was carried out by SDS-PAGE and electroelution. Gel Electrophoresis and Immunoblotting

SDS-PAGE in a 12% (w/v) polyacrylamide gel or 5 to 15% gradient gel containing 0.1% SDS was carried out by the method of Laemmli (8). Samples were heated in 50 mM TrisHCl, pH 6.8, 2% SDS, and 1% 2-mercaptoethanol at 70°C for 10 min before electrophoresis. The gels were stained with Coomassie blue and scanned at 650 nm with a densito-pattern analyzer to estimate the relative content of VM 23. PAGE was done in the presence of 0.1% (w/v) Triton X- 100 in a 5 to 18% linear gradient gel by a modified version of the method of Tomida et al. (22) as described previously (13). Antibody against the vacuolar H+-pyrophosphatase from mung bean was prepared as described previously (1 1). Antibody against VM 23 was raised in a rabbit by injection of the purified preparation (0.4 mg). The immunoglobulin G fractions were prepared from the sera by ammonium sulfate fractionation and DEAE-cellulose column chromatography (10). Immunoblotting was performed by a modified version of the method of Towbin et al. (23). For immunostaining, the proteins were electroblotted from the SDS-polyacrylamide gel to nitrocellulose using a semidry blotting apparatus (Transblot

1 249

SD, Bio-Rad). Immunoblot analysis was performed using horseradish peroxidase-linked protein A and a peroxidase color reaction (3, 1 1). Extraction of Proteins from Vacuolar Membranes with Organic Solvent

Solvent extraction of proteins was performed according to the method of Rea et al. (20). Two milliliters of a cold mixture ofacetone:ethanol (1: 1) were added to 50 ,uL of the membrane Ig) to remove the membrane lipids, and the suspension (100 mixture was incubated at -20C for 3 h. The aggregated protein was sedimented by centrifugation at 15,000g for 10 min. The supernatant fluid was decanted, and the pellet was dried under a stream of N2. The pellet was resuspended in 20 AL of H20, and 0.6 mL of ice-cold chloroform:methanol (2:1) was added. The mixture was stirred gently at 4°C for 4 h and centrifuged at 15,000g for 10 min. The supernatant fraction (chloroform:methanol extract) was rotary-evaporated to near dryness. Labeling with

[14C]DCCD

Labeling of the proteins in the vacuolar membranes with

['4C]DCCD was performed by the method of Uchida et al. (24) with a few modifications as described previously (1 1). The vacuolar membranes (200 Ag) suspended in 0.8 mL of 10 mM Tris-acetate, pH 7.5, 0.5 mM PMSF, and 1 mM EDTA were incubated with 22 AM of ['4C]DCCD (37 kBq, 2.1 GBq/ mmol) at 25C for 75 min. Proteins were precipitated with cold 10% TCA then subjected to SDS-PAGE. The gel was incubated in Amplify (Amersham) for 30 min and dried. Radioactive bands were detected with Hyperfilm-MP (Amersham) after 15 d of storage in the dark at -80°C.

Sucrose Density Gradient Centrifugation The vacuolar membranes were subjected to sonication for 20 s and overlaid on a linear gradient of 12 to 40% (w/w) sucrose. The sucrose solutions contained 1 mM DTT, 1 mM MgCl2, 1 mm EGTA, and 20 mM Tris-acetate, pH 7.5. After centrifugation for 14 h at 26,000 rpm (90,000g) at 4°C in a Hitachi RPS 27 rotor, the gradient was collected in 0.9-mL fractions. Analytical Measurements The purified preparation of 23-kD protein (200 pmol) was thoroughly dialyzed against 0.02% SDS and used for Nterminal amino acid sequence analysis performed with a pulsed liquid phase sequenator (Applied Biosystems, model 477A) and an amino acid analyzer (model 120A). Activity of vacuolar inorganic pyrophosphatase was measured in a reaction medium (0.25 mL) containing 1 mm sodium PPi, 1 mM MgSO4, 50 mM KCI, 1 mm sodium molybdate, 0.02% Triton X-100, and 30 mM Tris-Mes, pH 7.2, as described previously (11). One unit was defined as the amount of enzyme that hydrolyzed 1 ,umol of PPi/min under the assay conditions. Protein concentration was determined by the method of Lowry et al. (9) after precipitation with cold 10% TCA.

Plant Physiol. Vol. 98, 1992

MAESHIMA

1 250

RESULTS

femne rdat r-t

A

5/

~

Isolation of VM 23

J -ob I-

~

As shown in Figure lA, a large amount of protein in the vacuolar membranes from radish is detected at 23 kD after SDS-PAGE. The vacuolar membranes from mung bean hypocotyls also contain a major protein of 22 kD as described below (see Fig. 6). The major protein of 23 kD in the radish vacuolar membranes, which was tentatively named VM 23, was purified by selective solubilization from the membranes with deoxycholate and lysophosphatidylcholine, and QAEToyopearl column chromatography. The amino terminal sequence of the isolated VM 23 was as follows: 10 I NH2-Pro-Ile-( )-Asn-Ile-Ala-Ile-Gly-Gly-Val-Gln-Glu-Glu20 Thr-Thr-His-Pro-Ser-Leu-LeuThe third residue could not be determined. VM 23 did not seem to be a glycoprotein according to the present findings; that is, it was not stained with horseradish peroxidase-linked Con A or affected by endoglycosidase F.

~A

..

ts

0O1% SDS 1% SDS

B,

55

98

55

Properties of VM 23

70 98

van- Nmk

a="

_

C-~.

^R f

__

2

3

4

5

6

Figure 1. SDS-PAGE of vacuolar membranes and VM 23. A, Vacuolar membranes (30 .ig) from radish were suspended in 2% SDS and 1% 2-mercaptoethanol, incubated at the indicated temperature for 10 min, and subjected to SDS-PAGE in a 12% polyacrylamide gel. B, VM 23 was isolated by QAE-Toyopearl column chromatography and then totally purified by SDS-PAGE and electroelution. The purified preparation of VM 23 (80 jig) was incubated at the indicated temperature for 10 min in the presence of 0.1% (lanes 2 and 3) or 1% SDS (lanes 4-6). Each sample contained 1% 2-mercaptoethanol and 50 mM Tris-HCI, pH 6.8. Samples were subjected to SDS-PAGE in a 5 to 15% linear gradient (lanes 1-3) or a 12% polyacrylamide gel (lanes 4-7). Lanes 1 and 7, molecular mass standards (kD).

Figure 1 shows the SDS-PAGE analysis of the vacuolar membranes and the purified VM 23 after dissociation at various temperatures. A temperature of 37TC or below was not enough to dissociate some of the polypeptides with molecular masses of 50 to 70 kD even in the presence of 2% SDS (Fig. lA). By treatment at high temperature (980C), the amount of protein at 23 kD decreased, and the proteins at 40, 60, 80 kD or the top of the gel increased. The aggregation of VM 23 was observed in the purified preparation. As shown in Figure B (lane 2), the purified preparation of VM 23 gave a major band at 23 kD and a diffuse band at 40 kD. The sample used in this experiment had been electroeluted from the slice of SDS-gel at 23 kD. The gel-purified protein at 40 kD, when reelectrophoresed on SDS-PAGE, yielded not only the 40-kD band but also the 23-kD band (data not shown). The results indicate that the diffuse band at 40 kD is not a different polypeptide from VM 23. If the sample was heated at 980C in the presence of 0.1% SDS for 10 min, the VM 23 aggregated as a high molecular mass complex (Fig. 1B, lane 3). Aggregation by heat is not an unusual phenomenon for very hydrophobic proteins. Heating of the sample in 1% SDS prevented the aggregation of VM 23 as a large complex, but the bands at 40 and 60 kD did not disappear (Fig. IB, lanes 4-6). The protein bands at 40 and 60 kD may represent various levels of heat aggregation of VM 23. The molecular size of the native form of VM 23 was estimated by PAGE in the presence of Triton X- 100. Figure 2 shows a two-dimensional polyacrylamide gel of the purified preparation after QAE-Toyopearl column chromatography. After PAGE in the presence of 0.1% Triton X- 100, the gel slice was subjected to SDS-PAGE to detect the VM 23 protein. The VM 23-detergent complex migrated at the position of about 400 kD. The molecular mass of the micelle of Triton X- 100 is about 90 kD (4). The value obtained in this experi-

INTEGRAL PROTEIN OF VACUOLAR MEMBRANE

I St

669I 440

232 140 I

67I

,1

V'

cs 1f _

*

_

97- _ 6 7-

-

4 2-

30-

the radish membranes is so low compared with the mung bean membranes that a clear band could not be observed at the position of the enzyme by fluorography. During the study on vacuolar H+-ATPase of red beet, Rea et al. (20) detected the 64-, 40-, 27-, 23-, and 16-kD polypeptides in the organic solvent extract of the vacuolar membranes. They reported that only the 16-kD protein (DCCD-binding subunit of ATPase) among the five polypeptides was labeled with ['4C]DCCD (20). The 23-kD protein of red beet may differ from radish VM 23. Localization of VM 23 on Vacuolar Membranes

_

2014

1251

The vacuolar membranes of radish were subjected to sucrose density gradient centrifugation and then the distribution of VM 23 was determined by immunostaining as shown in Figure 5. The vacuolar membrane fraction used in this study

_mi

1

Figure 2. PAGE of VM 23 in the presence of Triton X-1 00. The purified VM 23 (15 jig) after QAE-Toyopearl column chromatography was subjected to PAGE in a 5 to 18% linear gradient polyacrylamide gel containing 0.1% Triton X-100 (the first dimension, upper panel). The sample solution contained 20 mm Tris-acetate, pH 7.5, 20% glycerol, 1 mm DTT, 1 mm EGTA, and 2 mm MgCI2. Next, a gel slice was overlaid onto the second gel containing 0.1% SDS and electrophoresed (the second dimension, lower panel). The figures on the first and second dimensions indicate the molecular masses (kD) of the size markers. Arrowhead in the second dimension points to VM 23.

ment was much larger than would have been expected if the VM 23 had existed as a single polypeptide. These results suggest that the purified VM 23 occurs in an oligomeric form in the detergent solution. The vacuolar membranes from radish and mung bean were pretreated with acetone:ethanol (1:1) to remove the membrane lipid and then extracted with chloroform:methanol (2:1). As shown in Figure 3, radish VM 23 and the 22-kD polypeptide of mung bean were partially extracted from the membranes into chloroform:methanol (2:1). This revealed that both proteins are very hydrophobic molecules. Whether VM 23 is a proteolipid is not yet known. Incubation of vacuolar membranes from radish with [14C] DCCD for 75 min yielded three labeled products upon SDSPAGE and fluorography: bands of 16, 23, and 40 kD (Fig. 4). The intense band at 16 kD corresponds to the DCCD-binding subunit of vacuolar H+-ATPase as reported previously (11, 16, 20). The band at 23 kD corresponds to VM 23, and the band at 40 kD was an aggregate of VM 23 as described above. The 22-kD polypeptide of vacuolar membranes from mung bean was also labeled with ['4C]DCCD as reported previously (11). The results indicate that VM 23 of radish and the 22kD protein of mung bean have a DCCD-reactive domain in a hydrophobic pocket. As reported previously (11), DCCD interacted and inhibited the vacuolar H+-pyrophosphatase of mung bean. However, the content of H+-pyrophosphatase in

2

3

6

5

4

7

.......

._.t _

j l.,-..::

:OolAs 14 j'l

". .1

radish

mung bean

Figure 3. SDS-PAGE analysis of fractions prepared by chloroform:methanol extraction of vacuolar membranes. The membranes were prepared from radish and mung bean, and treated with acetone:ethanol (1:1) to remove the membrane lipids. The membranes were then resuspended in H20, and ice-cold chloroform:methanol (2:1) was added. The mixtures were stirred for 4 h at 40C and centrifuged at 18,000g for 10 min. The supernatant fluids (chloroform:methanol extracts) were rotary-evaporated. Both the supernatant and precipitate fractions were subjected to SDS-PAGE. Lane 1, radish vacuolar membranes (75 Mg); lane 2, radish membranes treated with chloroform:methanol; lane 3, chloroform:methanol extract of radish membranes; lane 4, molecular mass standards (97, 67, 42, 30, 20, and 14 kD); lane 5, vacuolar membranes of mung bean (75 Mg); lane 6, mung bean membranes treated with chloroform:methanol; lane 7, chloroform:methanol extract of mung bean membranes.

Plant Physiol. Vol. 98, 1992

MAESHIMA

1252

2 ij . ;

1.

"

L.1

bottony

.

..t

5i

4-t

6 ^/

'0 (-,' .F,

.4--

0.

A

40

0

A.. a

CL

3 ()e

:> racti *fr s mbei b w

VM 23

20-

* 60 WO ". -

_