Characterization of a Red Beet Protein Homologous to the Essential 36-Kilodalton Subunit of the Yeast V-Type ATPase

Abstract

V-type proton-translocating ATPases (V-ATPases) (EC 3.6.1.3) are electrogenic proton pumps involved in acidification of endomembrane compartments in all eukaryotic cells. V-ATPases from various species consist of 8 to 12 polypeptide subunits arranged into an integral membrane proton pore sector (V₀) and a peripherally associated catalytic sector (V₁). Several V-ATPase subunits are functionally and structurally conserved among all species examined. In yeast, a 36-kD peripheral subunit encoded by the yeast (Saccharomyces cerevisiae) VMA6 gene (Vma6p) is required for stable assembly of the V₀ sector as well as for V₁ attachment. Vma6p has been characterized as a nonintegrally associated V₀ subunit. A high degree of sequence similarity among Vma6p homologs from animal and fungal species suggests that this subunit has a conserved role in V-ATPase function. We have characterized a novel Vma6p homolog from red beet (Beta vulgaris) tonoplast membranes. A 44-kD polypeptide cofractionated with V-ATPase upon gel-filtration chromatography of detergent-solubilized tonoplast membranes and was specifically cross-reactive with anti-Vma6p polyclonal antibodies. The 44-kD polypeptide was dissociated from isolated tonoplast preparations by mild chaotropic agents and thus appeared to be nonintegrally associated with the membrane. The putative 44-kD homolog appears to be structurally similar to yeast Vma6p and occupies a similar position within the holoenzyme complex.

V-ATPases are electrogenic proton pumps involved in acidification of endomembrane compartments in all eukaryotic cells (for review, see Finbow and Harrison, 1997). V-ATPases appear to be responsible for acidification of vacuoles, lysosomes, Golgi cisternae, secretory vesicles, and clathrin-coated vesicles. In addition, V-ATPases are associated with the plasma membrane of specialized mammalian cell types. The yeast (Saccharomyces cerevisiae) V-ATPase maintains an acidic environment in the vacuolar lumen and generates a proton gradient that drives the transport of ions such as Ca²⁺ and basic amino acids across the vacuolar membrane (for review, see Nelson and Klionsky, 1996). In plant cells the V-ATPase operates in conjunction with a proton-translocating pyrophosphatase to maintain a proton gradient across the tonoplast that supports vacuolar uptake of K⁺, Ca²⁺, sugar, and other small metabolites by secondary transport systems (Taiz, 1992).

V-ATPases from various species consist of 8 to 12 polypeptide subunits arranged in an integral membrane proton pore sector (V₀) and a peripherally associated catalytic sector (V₁) (Finbow and Harrison, 1997). Thus, V-ATPases display a bipartite structure similar to mitochondrial F₁F₀-ATPases. In yeast, combined biochemical and genetic approaches have identified 10 subunits that range from 14 to 100 kD and are required for holoenzyme assembly and ATPase activity (Table I). The V₁ sector is comprised of 69-, 60-, 54-, 42-, 32-, 27-, and 14-kD polypeptides. The membrane V₀ sector consists of 100- and 17-kD integral membrane proteins and a tightly associated 36-kD peripheral polypeptide (Bauerle et al., 1993). Highly purified preparations of V-ATPase from red beet (Beta vulgaris) tonoplast contain at least 9 polypeptides, of which the 67-, 55-, 52-, 44-, and 32-kD subunits are reported to be peripherally associated (Parry et al., 1989).

Table I.

Subunit composition of V-ATPase from yeast and red beet

Organism	V1 Subunits								V0 Subunits			Ref.
	kD
Yeast	69	60	54		42	32	27	14	100	(36)	17	Graham et al. (1995)
Red beet	67	55	52	(44)	42	32	29		100		16	Parry et al. (1989)

V₁ and V₀ subunits of V-ATPase from yeast and red beet. Subunits are identified based on observed molecular mass. Numbers in parentheses indicate the 36-kD yeast and 44-kD beet subunits examined in this study.

Many V-ATPase subunits are functionally and structurally conserved among all species examined (Finbow and Harrison, 1997). The V₁ 70-kD catalytic and 60-kD regulatory subunits share as much as 60 to 70% identity among diverse species. The 17-kD V₀ proteolipid is among the most highly conserved proteins known, displaying greater than 65% sequence identity among all species examined. This high degree of sequence similarity for multiple subunits of the enzyme reflects the functional importance of this ubiquitous eukaryotic proton pump.

In yeast, the 36-kD V₀ subunit is encoded by the VMA6 gene (Bauerle et al., 1993). The VMA6 gene product, a 36-kD subunit, Vma6p, is required for V-ATPase activity as well as stable holoenzyme assembly. In vma6 mutants lacking the 36-kD subunit, the V₀ sector does not stably assemble, and V₁ particles are unable to associate with the membrane. Biochemical studies revealed that Vma6p is peripherally rather than integrally attached to the vacuolar membrane and thus represents a novel class of peripheral V₀ subunits. A high degree of sequence similarity among Vma6p homologs from various species suggests that this subunit has a conserved role in V-ATPase function (Fig. 1). Subunit homologs from fungal, insect, and mammalian species share 41 to 57% primary sequence identity with Vma6p. In particular, four highly conserved domains display more than 60% identity, corresponding to residue nos. 96 to 133, 172 to 193, 212 to 227, and 314 to 322 in the yeast sequence.

Figure 1.

Amino acid sequence alignment of Vma6p homologs. Dashes indicate spaces added to optimize alignment, periods indicate amino acid identity among all seven sequences examined, and asterisks indicate stop codons. Overall homologies with Vma6p are: S. cerevisiae (S cer), 100% (Bauerle et al., 1993); Neurospora crassa (N cras), 57% (Melnik and Bowman, 1996); Dictyostelium discoideum (D disc), 49% (Temesvari et al., 1994); Manduca sexta (M sex), 47% (Merzendorfer et al., 1997a); Bos taurus (B taur), 46% (Wang et al., 1988); Homo sapiens (H sap), 41% (van Hille et al., 1993); Mus musculus (M musc), 47% (unpublished, GenBank accession no. U21549).  

Although gene homologs of yeast VMA6 have been isolated from various fungal and animal sources, no full-length gene homologs have been reported from any plant source. A search of the Arabidopsis gene-fragment database (using BLAST) identified sequences bearing strong homology to yeast VMA6. By aligning overlapping sequence fragments, we were able to assemble a partial sequence corresponding to a putative Arabidopsis VMA6 homolog (Fig. 2). The deduced primary sequence aligned with Vma6p residue nos. 76 to 345 and displayed all four highly conserved domains (67, 86, 69, and 57% identity with the yeast Vma6p sequence, respectively).

Figure 2.

Amino acid sequence alignment of Vma6p and deduced partial composite sequence from Arabidopsis (A thal). Four highly conserved regions shared between yeast and Arabidopsis sequences are underlined. The sequence homology is 36% in the overlapping region (corresponding to Vma6p residue nos. 76–345). S cer, S. cerevisiae.

The existence of a putative VMA6 homolog in Arabidopsis hints that homologs of this essential yeast subunit may exist in other plant species. Thus, we sought to identify potential Vma6p subunit homologs in V-ATPase-enriched membrane preparations. Here we report the initial characterization of a novel Vma6p homolog from red beet tonoplast membranes. This 44-kD polypeptide was directly identified by immuno-cross-reactivity with antibodies raised against yeast Vma6p. Preliminary characterization suggests that the putative homolog is a peripheral subunit of the V-ATPase that is tightly associated with the membrane.

MATERIALS AND METHODS

Alkaline phosphatase-conjugated antibodies and Kaleidoscope protein molecular mass standards were purchased from Promega. Ready Gels for the MiniProtean II gel system were from Bio-Rad. Nitrocellulose membrane was from Schleicher & Schuell. The BCA reagent kit was from Pierce. All other reagents were from Sigma.

Preparation of rabbit polyclonal antiserum against the yeast (Saccharomyces cerevisiae) 36-kD subunit has been described (Bauerle et al., 1993). Corresponding preimmune serum was collected from the same animal immediately prior to antigen exposure. Rabbit polyclonal antisera prepared against red beet (Beta vulgaris) tonoplast V-ATPase 67- and 57-kD subunits were a generous gift from Dr. Ron Poole (McGill University, Montreal, Quebec, Canada).

Strains and Culture Conditions

Isogenic yeast strains SEY6211a VMA6 and SEY6211a vma6::LEU2 have been described (Bauerle et al., 1993). Yeast cultures were grown at 30°C with vigorous shaking in 1% yeast extract, 2% Bactopeptone, 2% dextrose buffered at pH 5.0 with 50 mm phosphate/succinate.

Protein Sample Preparation, SDS-PAGE, and Immunoblot Analysis

Whole yeast cell lysates were prepared in sample buffer (8 m urea, 5% SDS, 1 mm EDTA, 50 mm Tris-HCl, pH 6.8, and 5% β-mercaptoethanol) as described previously (Bauerle et al., 1993). Protein concentrations were determined prior to the addition of β-mercaptoethanol by the BCA assay, and 40 μg of protein was loaded per lane.

Proteins were separated on 12, 15, or 10 to 20% gradient gels and electrotransferred to nitrocellulose membranes at a constant 12 V for 30 to 45 min at ambient temperature in a TransBlot semidry transfer cell (Bio-Rad). Western immunoblot analysis was performed as described before (Towbin et al., 1979). Blots were incubated with primary antibodies for 2 to 4 h in TBS containing 0.1% Tween 20 plus 2% nonfat dry milk at dilutions of 1:250, with constant rotation in a hybridization incubator (LabLine, Melrose Park, IL) at 37°C. Alkaline phosphatase-conjugated secondary antibodies were applied at a 1:5000 dilution.

Immunoblot data were quantified by scanning densitometry followed by image analysis using a GS700 imaging densitometer and Molecular Analyst 2.1 software (Bio-Rad). Standard curves for molecular mass estimations were generated by regression analysis of prestained marker proteins using Molecular Analyst 2.1 software.

Preparation of Tonoplast Membrane Vesicles

Vacuolar membranes from red beet storage root were prepared as previously described (Poole et al., 1984), frozen in liquid N₂, and stored at −80°C until use. ATPase and proton-pumping activity were assayed after thawing according to published methods (Poole et al., 1984). Membrane preparations typically displayed specific activities for ATP hydrolysis in the range of 20 μmol mg⁻¹ h⁻¹. Tonoplast membrane vesicles were collected by centrifugation at 100,000g, dissolved in sample buffer, and heated to 95°C for 5 min prior to SDS-PAGE. Protein loads were 20 to 40 μg per lane.

Fractionation of Tonoplast Vesicles

Tonoplast membranes equivalent to 400 μg of protein were thawed, diluted into 1 mL of transport buffer (250 mm sorbitol, 100 mm KCl, and 25 mm BTP-Mes, pH 7.0), and then collected by centrifugation for 30 min at 100,000g in a fixed-angle TLA 120.2 rotor (Beckman). Membrane pellets were suspended to 1 mg/mL protein in transport buffer and then incubated for 30 min on ice in the presence of 5 mm MgSO₄, 0 to 200 mm KNO₃, and ±5 mm Tris-ATP. An aliquot was removed for measuring proton-pumping activity, and the remaining mixture was centrifuged for 30 min to separate membrane and supernatant fractions. The membrane pellet was washed once in transport buffer, and then membranes were collected by centrifugation and resuspended in transport buffer. Wash volumes were pooled and membrane and supernatant protein was precipitated by adding TCA to 10% (v/v) and incubating for 45 min on ice. Protein was pelleted by centrifugation for 15 min at 20,000 rpm in a refrigerated microcentrifuge (Eppendorf). Protein pellets were dissolved in 50 μL of sample buffer and heated to 95°C for 5 min prior to SDS-PAGE.

Partial Purification of Red Beet H⁺-ATPase

Tonoplast membranes equivalent to 200 μg were thawed, diluted in 1 mL of transport buffer, and then collected by centrifugation as described above. Tonoplast vesicles were solubilized with Triton X-100 as previously described (Parry et al., 1989). Briefly, vesicles were resuspended in 0.70 mL of resuspension buffer (1.1 m glycerol, 5 mm Tris-Mes, pH 8.0, 1 mm EDTA, 0.5 mm BHT, and 5 mm DTT), then slowly diluted by dropwise addition of 0.75 mL of solubilization buffer (containing 8% Triton X-100 and 4 mm MgSO₄), and stirred on ice. The resulting mixture was stirred gently for 20 min.

The detergent-solubilized mixture was partially purified by gel filtration on Sephacryl S-400 as previously described (Parry et al., 1989). A 60- × 0.75-cm-diameter column packed with Sephacryl S-400 was preequilibrated with running buffer containing 10% glycerol, 0.3% Triton X-100, 0.05 mg/mL phospholipid (Type IV-S, Sigma), 5 mm DTT, 1 mm Tris-EDTA, 4 mm MgCl₂, and 5 mm Tris-Mes, pH 8.0. The entire sample volume was loaded and the column was run at a rate of 40 mL/h at 4°C. The detergent mixture was not centrifuged to remove nonsolubilized membrane particles prior to loading. Typically, 40 1-mL fractions were collected and 50-μL aliquots were assayed for ATPase activity and protein concentration.

Protein Determination

The protein concentration of yeast extract was determined by BCA assay according to the supplier's instructions (Pierce). The protein concentration of tonoplast vesicle preparations and gel-filtration fractions was determined by the method of Bradford (1987).

Determination of ATPase and Proton-Pumping Activity

ATPase activity was determined by measuring the amount of Pi liberated from ATP at 37°C in a 20-min reaction using the Ames method as previously described (Parry et al., 1989). Gel-filtration column fractions were supplemented with sonicated 1.33-mg/mL type IV-S phospholipid to preserve V-ATPase activity.

Proton-pumping activity in tonoplast vesicles was measured by the method of Giannini et al. (1995). Typically, vesicles equivalent to 10 to 20 μg of protein were suspended in transport buffer containing 250 mm sorbitol, 50 mm KCl, 5 mm MgSO₄, and 5 μm acridine orange. The reaction was initiated by the addition of Tris-ATP to 5 mm final concentration, and proton pumping was monitored by observing the decrease in acridine orange A₄₉₀ with a UV/visible light spectrophotometer (model DU 640, Beckman) in “kinetics/time mode.” Rates were calculated from data collected at 10-s intervals during a 3-min reaction period.

Sequence Analysis

VMA6 homologous protein sequences were aligned using Align Plus 2.0 from Scientific and Educational Software (State Line, PA). Alignment parameters were determined according to the work of Myers and Miller (1988). Arabidopsis sequence fragments homologous to yeast VMA6 were identified by a BLAST search of GenBank (Gish and States, 1993). The following Arabidopsis fragments were used to generate a derived partial amino acid sequence: T13399, Z26026, Z24482, Z30468, T20646, H36140, H36163, H37538, T13974, R30209, T44170, N97286, AA042689. The derived partial sequence reported in Figure 2 was confirmed by identifying at least two overlapping fragments along the entire length of the sequence.

RESULTS

A 44-kD Polypeptide from Red Beet Exhibits Immuno-Cross-Reactivity with Yeast Vma6p and Cofractionates with Tonoplast V-ATPase

Immunoblot analysis with antiserum generated against the yeast 36-kD V₀ subunit (α-Vma6p) consistently revealed a single band in lanes containing red beet tonoplast protein (Fig. 3, top). Immunoblot analysis using the corresponding preimmune serum did not recognize either the yeast 36-kD subunit or the cross-reactive protein in tonoplast vesicle lanes (Fig. 3, bottom). Neither α-Vma6p nor preimmune sera recognized a cross-reactive band in protein extracts from yeast vma6 mutant cells lacking the 36-kD subunit. Thus, the reactivity observed in both yeast and red beet samples appears to be specifically due to α-Vma6p polyclonal antibodies present in the immune serum. The molecular mass of the cross-reacting beet protein was estimated to be 44 kD by comparison with prestained molecular mass protein standards in an adjacent lane.

Figure 3.

Immuno-cross-reactivity between yeast Vma6p and beet 44-kD polypeptide. Proteins were separated on a 12% gel. Top, Immunoblot with polyclonal antiserum prepared against yeast Vma6p (αVma6p). Lane 1, Forty micrograms of whole cell protein from wild-type yeast VMA6; lane 2, 40 μg of protein from yeast vma6 mutant; and lane 3, 20 μg of red beet tonoplast protein. Relative positions of molecular mass standards are indicated. Bottom, Corresponding immunoblot with related preimmune serum.

To address whether the cross-reactive protein is specifically associated with V-ATPase in tonoplast vesicles, we partially purified the enzyme and monitored cofractionation of the cross-reactive protein with peak V-ATPase fractions. Tonoplast vesicles were detergent solubilized and proteins were separated by gel-filtration chromatography. The mixture was not centrifuged prior to loading on the column, a step normally taken to remove any residual unsolubilized membranes. This allowed us to observe the distribution of the 44-kD polypeptide between fully and partially solubilized fractions. Under these conditions, solubilized V-ATPase eluted as a single broad peak separated from bulk tonoplast protein, as evidenced by protein and ATPase activity profiles (Fig. 4). ATPase specific activity (corresponding to fractions 19–21 in Fig. 4) was typically enriched greater than 10-fold in peak column fractions compared with tonoplast vesicles (Table II). A sharp early peak (fractions 13–16 in Fig. 4) with relatively lower ATPase specific activity corresponded to incompletely solubilized tonoplast vesicles eluting with the void volume.

Figure 4.

Gel-filtration purification of V-ATPase from detergent-solubilized red beet tonoplasts. Tonoplast membranes were detergent solubilized and proteins were separated by gel filtration as described in Methods. Equivalent aliquots of even-numbered fractions were assayed for ATPase activity (micromoles per hour of PO₄³⁻ liberated) and total protein concentration (milligrams) as described in Methods. ▪, Relative protein concentration; □, relative ATPase activity. Peak ATPase specific activity (fraction no. 20) was approximately 140 μmol mg⁻¹ h⁻¹, representing a 7-fold enrichment.

Table II.

Partial purification of V-ATPase from red beet tonoplast membranes

Preparation	Specific Activity	Enrichment
	μmol mg−1 h−1	-fold increase
Tonoplast vesicles	21.7  ± 4.0	1.0
Peak column fraction	271  ± 40	12.5  ± 0.8

V-ATPase was partially purified from detergent-solubilized tonoplast vesicles by gel-filtration chromatography as described in Methods. Equivalent aliquots of column fractions were assayed for total protein and ATPase activity, and the fraction containing the highest ATPase specific activity was identified. The results below represent the average of three separate purifications.

Column fractions were probed with polyclonal antiserum directed against the 67-kD subunit of the red beet V-ATPase (α-67 kD) to determine the distribution of this peripheral V₁ subunit. Immunoblot results confirmed that the distribution of the 67-kD V₁ subunit closely correlated with V-ATPase activity (Fig. 5, top). The 67-kD subunit distributed between both the solubilized and unsolubilized V-ATPase peaks. Very little of this V₁ subunit was observed in fractions containing the bulk of soluble tonoplast protein released by detergent treatment (fraction no. 24), indicating that this subunit remained primarily associated with the enzyme complex during column purification.

Figure 5.

Distribution of 67-kD subunit and 44-kD polypeptide in gel-filtration fractions of detergent-solubilized beet tonoplasts. Column fractions from the experiment described in Figure 4 were analyzed. Fraction numbers are indicated at the top. Arrows mark peak ATPase and protein fractions. The asterisk indicates the fraction containing the peak ATPase specific activity. Top, Equivalent volumes of each sample were separated on a 10 to 20% gradient gel and then immunoblotted with anti-67-kD antiserum. A doublet band pattern was typically observed with α67-kD antiserum when samples were separated on gradient gels. Bottom, Same samples probed with αVma6p antiserum.

Distribution of the 44-kD cross-reactive protein was similar to that of the 67-kD V₁ peripheral subunit and closely correlated with the observed ATPase activity peaks (Fig. 5, bottom). We were unable to detect any of the 44-kD protein in peak fractions of soluble tonoplast protein. The 44-kD protein pattern correlated closely with both the V-ATPase activity profile and the 67-kD peripheral V₁ subunit pattern. This suggests that the 44-kD cross-reactive protein was tightly associated with the partially purified V-ATPase fraction from detergent-solubilized tonoplast membranes.

By comparing the R_F against a standard curve generated by regression analysis of prestained marker proteins, we calculated the apparent molecular mass of the cross-reacting protein to be 44.9 kD. This is within close range of 44- and 42-kD accessory subunits of beet V-ATPase previously described (Parry et al., 1989). To determine whether the cross-reactive protein comigrated with one of these previously reported accessory subunits, α-Vma6p immunoblots were compared with the protein band pattern of V-ATPase. Partially purified V-ATPase protein fractions were separated by electrophoresis and visualized directly by Coomassie staining or were transferred to nitrocellulose and then stained with amido black to reveal the subunit band pattern. Identical samples were transferred to nitrocellulose and then immunoblotted with α-Vma6p antibodies. The cross-reactive band migrated closely with the prominent 44-kD subunit observed in partially purified V-ATPase fractions (Fig. 6). Thus, the previously described 44-kD accessory subunit of beet V-ATPase appeared to be selectively immuno-cross-reactive with yeast Vma6p antibodies.

Figure 6.

Comigration of cross-reactive band with 44-kD accessory subunit of V-ATPase. Partially purified V-ATPase proteins were separated on a 15% polyacrylamide gel, and then duplicate lanes either were visualized by Coomassie staining (A) or were transferred to nitrocellulose and immunoblotted with α-Vma6p antiserum (B). Coomassie-stained and immunoblotted band patterns were compared by aligning band patterns of prestained marker proteins loaded in adjacent lanes. Stained gel and developed immunoblots were digitally imaged and analyzed using a scanning imaging densitometer.

The 44-kD Polypeptide Is Dissociated from Tonoplast Membranes by Urea Treatment

Treating yeast vacuolar membranes with urea quantitatively strips Vma6p from the membrane, unlike other V₀ subunits that are integrally associated with the membrane (Bauerle et al., 1993). To address whether the cross-reactive 44-kD polypeptide behaved in a similar fashion, we incubated isolated beet tonoplast vesicles in transport buffer ± 8 m urea and examined the distribution of the 44-kD polypeptide in supernatant and membrane pellet fractions (Fig. 7). Similar to yeast Vma6p, the 44-kD polypeptide was quantitatively removed from membranes treated with 8 m urea. Peripheral V₁ 57- and 67-kD subunits were also completely dissociated from the membrane under these conditions (C. Magembe, unpublished observations). Thus, the 44-kD polypeptide behaved like a peripherally attached subunit in membrane fractionation experiments with a strong chaotrope.

Figure 7.

Dissociation of 44-kD polypeptide from tonoplast membranes by urea treatment. Tonoplast membranes equivalent to 50 μg of protein were incubated in buffer alone (top) or buffer containing 8 m urea (bottom) for 30 min on ice. Membranes were collected by centrifugation and protein samples from both membrane pellet and supernatant fractions were prepared as described in Methods. Proteins were separated on 12% gels and immunoblotted with α-Vma6p antiserum. Developed immunoblots were digitally imaged and analyzed using a scanning imaging densitometer. S, Supernatant; P, membrane pellet.

KNO₃ Treatment Inactivates V-ATPase and Causes Partial Dissociation of 44- and 67-kD Polypeptides

Mild chaotropic agents such as KNO₃ have been shown to disrupt V-ATPase activity by causing specific dissociation of V₁ peripheral subunits from the V₀ sector (Ward et al., 1992). It is interesting that the dissociation effect is largely dependent on the presence of MgATP and thus appears to be specifically correlated with the ATP-dependent active state of the holoenzyme. The use of such chaotropes can provide useful information about the relationship between V-ATPase activity and subunit assembly.

Figure 8 illustrates the effect of KNO₃ on proton pumping in tonoplast vesicles. The ATP-dependent proton gradient was rapidly dissipated by addition of gramicidin, confirming that tonoplast vesicles were tightly sealed under incubation conditions. Proton-pumping activity was vanadate insensitive (less than 15% inhibition in the presence of 200 μm vanadate), indicating that tonoplast vesicles were relatively free of contaminating plasma membrane ATPase activity (not shown). Proton-pumping activity was strongly inhibited by the presence of 200 mm KNO₃ in the transport assay. Proton-pumping activity was similarly prevented by preincubating tonoplast vesicles with 200 mm KNO₃ plus MgATP prior to performing the assay. Following dilution of preincubated vesicles into the assay medium, the resulting KNO₃ concentration during the assay was approximately 4 mm, well below the observed K_i of 8 mm (refer to Fig. 10, top). Thus, inhibition in this case was due to inactivation of the V-ATPase during preincubation rather than inhibition during the assay. The inhibitory effect of KNO₃ preincubation was much less pronounced in the absence of MgATP; preincubation with 200 mm KNO₃ in the presence of either Mg²⁺ or Tris-ATP alone inhibited proton-pumping activity by less than 50% (data not shown).

Figure 8.

Effect of KNO₃ preincubation on proton transport in beet tonoplast vesicles. Proton transport activity was monitored as a percent decrease in acridine orange fluorescence measured at 490 nm (% F). Tonoplast vesicles equivalent to 10 μg of protein were diluted in 1 mL of transport buffer containing acridine orange and equilibrated for 3 min. The reaction was started by adding 5 mm Tris-ATP, and absorbance measurements were collected every 10 s for 3 min. At the end of the reaction Gramicidin D (Gram) was added to a final concentration of 3 μm. a, Nonpreincubated vesicles assayed in the absence of KNO₃ for 3 min and then for 3 min after Gramicidin D addition; b, nonpreincubated vesicles assayed in the presence of 200 mm KNO₃; c, vesicles preincubated with 5 mm MgATP; and d, vesicles preincubated with 5 mm MgATP plus 200 mm KNO₃.

Figure 10.

V-ATPase disruption as a function of KNO₃ concentration. Tonoplast vesicles equivalent to 50 μg of protein were preincubated with 5 mm MgATP and 0 to 200 mm KNO₃ for 30 min on ice. Aliquots equivalent to 10 μg of protein were removed for proton transport assays in the presence of 5 mm MgATP. The remaining sample was separated into membrane pellet and supernatant fractions and processed as described for Figure 8. Bands corresponding to 67- and 44-kD polypeptides were quantified by imaging densitometry as described in Methods. Top, Proton transport activity following preincubation; middle, distribution of 67-kD polypeptide between membrane pellet and supernatant fractions following preincubation; bottom, distribution of 44-kD polypeptide following preincubation.

Dissociation of the 44-kD Polypeptide Is Dependent on ATP and KNO₃ Concentration

A series of membrane fractionation experiments were then conducted using KNO₃ to further examine the association of the 44-kD polypeptide with tonoplast vesicles. Specifically, we sought to correlate association of the 44-kD polypeptide with proton-pumping activity. Tonoplast vesicles were preincubated at 0°C with KNO₃ in the presence of MgATP, and then aliquots equivalent to 20% of the total sample were assayed for proton transport activity. The remaining membranes were pelleted, and both membrane and supernatant fractions were examined for the 67-kD V-ATPase subunit and the 44-kD polypeptide by immunoblot analysis. When compared with the control, preincubation of tonoplast vesicles with 200 mm KNO₃ plus MgATP resulted in substantial dissociation of the 67-kD V₁ subunit (Fig. 9, top). The 44-kD polypeptide was also partially removed from the membrane by preincubation with nitrate, although to a lesser extent than the 67-kD peripheral subunit (Fig. 9, bottom).

Figure 9.

Dissociation of 67-kD subunit and 44-kD polypeptide by MgATP and KNO₃. Tonoplast vesicles equivalent to 50 μg of protein were preincubated with 5 mm MgATP ± 200 mm KNO₃ for 30 min on ice. Membranes were collected by centrifugation and protein samples from both membrane pellet and supernatant fractions were prepared as described in Methods. Proteins were separated on 15% gels and then immunoblotted. P, Membrane pellet; S, supernatant. Top, Membrane pellet and supernatant fractions probed with α67-kD serum. Bottom, Same fractions probed with α-Vma6p serum.

Dissociation of both the 67-kD V₁ subunit and the 44-kD polypeptide appeared to be a function of preincubation nitrate concentration in the presence of MgATP (Fig. 10, middle and bottom). At the highest preincubation concentration of KNO₃ tested (200 mm), 60% of the 67-kD subunit and 50% of the 44-kD polypeptide were dissociated from the membrane. It is interesting that complete inhibition of proton-pumping activity was achieved at much lower concentrations, with an observed preincubation K_i for nitrate inhibition of approximately 8 mm. Thus, although we consistently observed concentration-dependent dissociation of V-ATPase subunits, it appeared to be only generally correlated with loss of enzyme activity. In the absence of ATP during preincubation, the observed K_i for nitrate inhibition was approximately 260 mm. Some corresponding dissociation of V-ATPase subunits was observed; maximally, 25% of the 67-kD polypeptide and 10% of the 44-kD polypeptide were released during preincubation in the absence of MgATP (not shown). However, the amount of subunit dissociation observed in the absence of MgATP was not dependent on KNO₃ concentration and therefore was not clearly correlated with KNO₃ effects on the active V-ATPase holoenzyme.

DISCUSSION

In this report we present evidence supporting the identification of a red beet protein homologous to Vma6p. The putative 44-kD homolog was specifically cross-reactive with α-Vma6p polyclonal antibodies, indicating substantial sequence similarity with the yeast V₀ subunit. The 44-kD polypeptide cofractionated with peak V-ATPase activity as well as the 67-kD subunit of the V₁ sector, indicating that it was tightly associated with V-ATPase isolated from red beet. A 44-kD polypeptide was previously reported to be associated with highly purified V-ATPase preparations from red beet tonoplast (Parry et al., 1989). This protein was released, along with other V₁ subunits, by cold inactivation conditions, leading the authors to conclude that it is a peripheral V-ATPase subunit. Our results provide additional support for the identification of this subunit and suggest further that it shares significant similarity with a homologous subunit in yeast. The existence of Vma6p homologs among animal, plant, and fungal species suggests that this conserved subunit is important in the assembly and function of V-ATPase holoenzyme.

The yeast 36-kD subunit is a tightly associated peripheral component of the V₀ sector of V-ATPase. Our results indicate that the 44-kD unit is also peripherally attached and thus susceptible to removal with chaotropic agents such as KNO₃. KNO₃ has been shown to inhibit V-ATPase activity by dissociation of V₁ subunits from the membrane (Rea et al., 1987; Tu et al., 1987). At KNO₃ concentrations sufficient to completely inhibit V-ATPase activity in the presence of MgATP, neither the 67- nor the 44-kD subunit was quantitatively released from the membrane. Thus, the release of peripheral subunits may be a secondary consequence of a nitrate-induced conformational shift that leads to enzyme inactivation. Alternatively, some peripheral subunits may remain nonspecifically associated with the tonoplast membrane. For instance, if tonoplast membrane preparations contain a fraction of inside-out vesicles, then a portion of V-ATPase would likely be protected from dissociation by nitrate (Rea et al., 1987). Further biochemical studies are under way to determine the membrane association of this putative Vma6p homolog.

Given the ubiquitous involvement of V-ATPase in acidifying internal, and in some cases, external compartments, there has been much interest in elucidating the mechanism(s) by which proton-pumping activity is regulated. V-ATPase activity appears to be regulated in part by posttranslational modifications of V₁ subunits (for reviews, see Forgac, 1996; Merzendorfer et al., 1997a, 1997b). In addition, several reports have provided evidence supporting the hypothesis that V-ATPase cellular activity may also be modulated by regulated assembly-disassembly of V₁ and V₀ sectors. For example, selective release of peripheral V-ATPase subunits has been correlated with in vivo enzyme inactivation in response to chilling in mung bean seedlings (Matsuura-Endo et al., 1992). In Manduca sexta, regulation of V-ATPase activity during larval development appears to be correlated with a loss of V₁ subunits (Sumner et al., 1995). Recently, Kane (1995) described in vivo assembly-disassembly of peripheral V-ATPase subunits in yeast cells in response to Glc deprivation. In yeast, active V-ATPase assembles by attachment of preexisting V₁ particles from a cytoplasmic pool onto the V₀ membrane sector. Hence, V-ATPase components may be directly triggered to mediate assembly or disassembly in response to intracellular signals. The positioning of Vma6p as a peripherally associated V₀ subunit required for V₁ attachment suggests a role in such a regulatory mechanism.

We conclude that the 44-kD protein in red beet is a subunit of the tonoplast V-ATPase holoenzyme and is homologous to a tightly associated peripheral V₀ subunit previously described in yeast. Future studies will focus on a more detailed biochemical characterization of this novel V₀ subunit to understand its role in V-ATPase assembly.

Abbreviations:

BCA

bicinchoninic acid

BHT

butylated hydroxytoluene

V-ATPase

V-type proton-translocating ATPase