Multiple Forms of Phospholipase D Inhibitor from Rat Brain Cytosol

Rat brain cytosol contains proteins that markedly in- hibit the activity of partially purified brain membrane phospholipase D (PLD) stimulated by ADP-ribosylation factor (Arf) and phosphatidylinositol 4,5-bisphosphate (PIP 2 ). Sequential chromatography of the brain cytosol yielded four inhibitor fractions, which exhibited different kinetics to heat treatment at 70 °C. Purification of the most heat-labile inhibitor to homogeneity yielded two preparations, which displayed apparent molecular masses of 150 kDa and 135 kDa, respectively, on SDS-polyacrylamide gels. Tryptic digests of the 150- and 135- kDa proteins yielded similar elution profiles on a C 18 reverse-phase column, suggesting that the 135-kDa form is a truncated form of the 150-kDa form. Sequences of two tryptic peptides were determined. A data base search revealed no proteins with these sequences. The purified 150-kDa inhibitor negated the PLD activ- ity stimulated by Arf, RhoA, or Cdc42. The concentration required for half-maximal inhibition was 0.4 n M . Con- centration dependence on the 150-kDa inhibitor was not affected by changes in the concentrations of Arf, PIP 2 , or phosphatidylcholine used in the assays, suggesting that the inhibition is not due to competition with the activators or substrate for PLD. The purified inhibitor did not affect the PIP 2 -hydrolyzing activity of a phos- pholipase C isozyme that was measured with substrate vesicles of lipid composition identical with that used for the PLD assay. Thus, the mechanism of inhibition appears reductively by treatment with 50 m M Tris (pH 8.0), 6 M guanidine hydrochloride, and 2 m M DTT. The sulfhydryl groups were labeled with 2-nitro-5-thiobenzoate by adding Ellman’s reagent, 5,5 (cid:57) -dithiobis-(2-nitrobenzoic acid), to a final concentration of 10 m M . The 2-nitro-5-thiobenzoate-conjugated proteins were precipitated with 10% trichloroacetic acid, and the pellet was washed with acetone, suspended in 50 m M Tris (pH 8.0), and digested with trypsin overnight at 37 °C. The trypsin-digested peptides were applied to a C 18 column (4.6 (cid:51) 250 mm) that had been equilibrated with 0.1% trifluoroacetic acid. The elution was performed at a flow rate of 1 ml/min with a 60-ml linear gradient of 0 to 60% acetonitrile in 0.1% trifluoroacetic acid. Peptides were detected by measuring absorbance at 215 nm. The sequences of the two indicated peptides were demonstrated to be EANAPAFD ( ) and GSVPLFWE ).

Phospholipase D (PLD) 1 catalyzes the hydrolysis of phospho-lipids at their terminal phosphodiester bond, thus producing phosphatidic acid (PA) and releasing the free polar head group. Stimulation of PLD occurs in a wide variety of cells treated with hormones and growth factors (1)(2)(3)(4)(5)(6). PA has been implicated as a biologically active molecule and can be further metabolized by a PA phosphohydrolase to form diacylglycerol, a protein kinase C activator (3,4,(7)(8)(9)(10). Most mammalian PLD activity is found in association with membranes and appears to be specific for phosphatidylcholine (PC) (11)(12)(13)(14)(15). There is compelling evidence for the existence of multiple isoforms of PLD in mammalian cells, and PLD is known to be activated via multiple pathways involving G proteins, Ca 2ϩ , unsaturated fatty acids, protein kinase C, or protein-tyrosine kinases (1-4, 9, 16 -22). However, there is little information on the biochemical and molecular properties of mammalian PLDs.
Significant progress has been made recently on the G protein regulation of PLD. Members of the Arf family of the small G protein (Smg) superfamily were recognized as activators of PLD, and the presence of PIP 2 in substrate vesicles was shown to be essential for the detection of Arf-regulated PLD activity (12,16,(23)(24)(25). Subsequently, a rat brain PLD that is activated by Arf and PIP 2 but inhibited by oleate was separated from another form of rat brain that is completely dependent on oleate for activity but insensitive to Arf and PIP 2 (17). Members (RhoA, Cdc42, and Rac) of another Smg family, Rho, were also found to activate PLD from several tissues. Membrane PLDs in HL60 cell and porcine brain were shown to be synergistically activated by Arf and RhoA (26,27), whereas rat liver PLD was activated by RhoA but not by Arf (28). Thus, whether the target for Arf and Rho proteins is identical or not is not clear. Recently, a group led by Drs. Morris and Frohman (29) cloned a human cDNA corresponding to the PLD (PLD-1) that is activated by Arf and PIP 2 but inhibited by oleate and identified at least two more PLD genes. The same authors reported that the brain PLD-1 could also be activated by members of the Rho family.
Rho-activated PLD is also negatively modulated by Rho GDP dissociation inhibitor (Rho-GDI) which blocks GDP/GTP exchange on certain Smg proteins (26,27,30,31). The present studies were undertaken to investigate the possibility that proteins other than Rho-GDI negatively modulate Smg-dependent PLD activity. We find that rat brain cytosol contains multiple proteins that can inhibit PLD activity stimulated by Arf proteins. We purified one of these inhibitors to homogeneity and studied its effects at various concentrations of Arf, PIP 2 , and PC. * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18

Preparation of Brain Membranes and Ammonium Sulfate Fractionation of Cytosol
All procedures were carried out at 4°C unless otherwise indicated. Rat brains (400 g) were thawed in 4 liters of Buffer A (20 mM Hepes (pH 7.0), 1 mM EGTA, 0.1 mM dithiothreitol (DTT), 1.5 mM phenylmethylsulfonyl fluoride, and leupeptin and aprotinin (each at 1.5 g/ml)) and disrupted three times for 20 s at a time with a Polytron homogenizer at a setting of 8. The homogenate was centrifuged at 1,000 ϫ g for 20 min. The pellet was further homogenized in 2 liters of Buffer A with Polytron homogenizer. Supernatant from the two homogenization steps were pooled and then centrifuged at 100,000 ϫ g for 1 h. The resulting membrane pellet was saved for the purification of PLD. The supernatant was further fractionated by adding solid ammonium sulfate to yield a final saturation of 35%. The solution was stirred gently for 1 h and centrifuged at 5,000 ϫ g for 20 min. The new pellet (0 -35% ammonium sulfate precipitate of cytosol) was saved for the purification PLD inhibitor. The resulting supernatant was brought to 70% saturation with ammonium sulfate, stirred for 1 h, and centrifuged at 5,000 ϫ g for 20 min. The 35-70% ammonium sulfate precipitate of cytosol was stored at Ϫ80°C for the purification of Arf.

Assays of PLD and PLD Inhibitor
Arf-dependent PLD activity was measured using an assay described previously (17) with a minor modification. Briefly, 25 l of mixed lipid vesicles containing PE, PIP 2 , and PC in molar ratio of 16:1.4:1 with [choline-methyl-3 H](pam) 2 PC to yield ϳ 200,000 cpm per assay were added to 2 to 5 l of PLD source, 100 nM Arf, and 5 M GTP␥S in a total volume of 125 l containing 50 mM Hepes (pH 7.5), 3 mM EGTA, 80 mM KCl, 2.5 mM MgCl 2 , and 2 mM CaCl 2 . The final concentration of PC was 3.4 M, and the free Ca 2ϩ concentration is estimated to be 300 nM. Assays were incubated at 37°C for 30 min before addition of 1 ml of stop solution containing CHCl 3 , CH 3 OH, and concentrated HCl (50:50:0.3, v/v) and 0.35 ml of 1 M HCl in 5 mM EGTA. After separation of the organic and aqueous phases by centrifugation, the released [ 3 H]choline in 0.5 ml of the aqueous phase was quantitated by liquid scintillation spectrometry. Unless otherwise indicated, PLD inhibitor was assayed by measuring its capacity to inhibit the Arf-dependent activity measured under the conditions described above. Partially purified PLD (ϳ100 ng) from the TSK-gel Heparin HPLC step (see below) was used as the PLD source, resulting in the production of 20,000 to 30,000 cpm of choline in the absence of inhibitor. Carryover of buffer used in preparation of this PLD source introduced 0.16 mM n-octyl-␤-D-glucopyranoside into the assay.

Partial Purification of PLD
Rat brain membrane pellet was stirred for 1 h with 3 volumes (1200 ml, w/v) of Buffer A containing 1% Triton X-100 and 0.3 M NaCl. Insoluble material was discarded after centrifugation at 100,000 ϫ g for 1 h. Solubilized proteins (8 g of protein in 1200 ml) were applied to a heparin-Sepharose CL-6B column (5 ϫ 30 cm) that had been equilibrated with Buffer B (20 mM Hepes (pH 7.0), 1 mM EGTA, 0.1 mM DTT, and 0.1% Triton X-100) containing 0.3 M NaCl. Unbound proteins were washed with 600 ml of Buffer B containing 0.3 M NaCl. Bound proteins were eluted from the column with a 1200-ml linear gradient of 0.3 to 2.0 M NaCl in Buffer B. Fractions of 20 ml were collected and assayed for PLD activity in the presence of GTP␥S and Arf purified from rat brain (see Purification of Arf below). Peak fractions (600 mg of protein in 300 ml) were pooled and diluted with 4 volumes of Buffer B and loaded onto a TSK-gel CM-5PW cation-exchange column (21.5 ϫ 150 mm, Toso-Haas) equilibrated in Buffer B. Proteins were eluted, at a flow rate of 5 ml/min, by application of 300 ml of Buffer C (20 mM Hepes (pH 7.0), 1 mM EGTA, 0.1 mM DTT, and 0.7% n-octyl-␤-D-glucopyranoside) and then a 200-ml linear gradient of 0 to 1.0 M NaCl in Buffer C. The column was washed then with 100 ml of Buffer C containing 1.0 M NaCl. Fractions of 5 ml were collected and analyzed for PLD activity. Peak fractions (fractions 33-44) were pooled and dialyzed against Buffer D (20 mM Hepes (pH 7.4), 1 mM EGTA, 0.1 mM DTT, and 0.7% n-octyl-␤-D-glucopyranoside) containing 2.5 g/ml aprotinin. After clarification by centrifugation, the supernatant containing 25.8 mg of protein was applied to a Mono Q HR 5/5 anion-exchange column (Pharmacia Biotech Inc.) that had been equilibrated with Buffer D. Proteins were eluted at 1.0 ml/min with a 40-ml linear gradient of 0 to 1.0 M NaCl in Buffer D. Fractions of 1 ml were collected. Peak fractions (fractions 12-31) were diluted 4-fold with Buffer C and applied to a Mono S HR 5/5 cation exchange column (Pharmacia) equilibrated with Buffer C. Proteins were eluted with a 40-ml linear gradient of 0 to 1.0 M NaCl in Buffer C. The pool of peak fractions (fractions 25-36) from Mono S column was diluted 4-fold with Buffer C and loaded onto a TSK-gel Heparin-5PW column (7.5 ϫ 75 mm) equilibrated with Buffer C. Proteins were eluted by applying a linear gradient from 0 to 0.5 M NaCl for 15 min and a second linear gradient from 0.5 to 1.0 M for 45 min. Fractions of 1 ml were collected at a flow rate of 1 ml/min and assayed for PLD activity. Peak fractions (fractions 40 -51) were pooled to yield 700 g of protein in 12 ml, which was then stored at Ϫ70°C as a 50-l aliquot to be used as the source of PLD. Once thawed, a sample was used in the same day.

Purification of Arf
The 35-70% ammonium sulfate precipitate of rat brain cytosol was suspended in ϳ400 ml of buffer E (10 mM Tris (pH 8.0), 1 mM EDTA, 1 mM DTT, and leupeptin and aprotinin (each at 1 g/ml)), dialyzed against the same buffer, and centrifuged to remove insoluble particles. The supernatant (10 g of protein in 400 ml) was loaded onto a DEAE-Sephacel column (5 ϫ 20 cm) that had been equilibrated with Buffer E. The column was washed with 800 ml of Buffer E, and the bound proteins were eluted with a 2-liter linear gradient of 0 to 0.5 M NaCl in Buffer E. Fractions (20 ml) were assayed for PLD-activating activity in the presence GTP␥S and for Arf protein by immunoblot analysis with antibodies to Arf. Peak fractions of Arf (800 mg of protein in 100 ml) were pooled and were brought to 1.5 M ammonium sulfate by adding solid salt, and the mixture was centrifuged to remove particles. The supernatant was applied to a TSK-gel Phenyl-5PW (21.5 ϫ 150 mm) that had been equilibrated with Buffer F (20 mM Hepes (pH 7.0), 1 mM EGTA and 0.1 mM DTT) containing 1.5 M ammonium sulfate. Proteins were eluted, at a flow rate of 5 ml/min, by application of a decreasing ammonium sulfate linear gradient from 1.5 to 0.6 M for 20 min and a second decreasing ammonium sulfate linear gradient from 0.6 to 0 M for 10 min. The column was then washed with ammonium sulfate-free Buffer F. Peak fractions (fractions 32-36) were pooled, concentrated to 5 ml in an ultrafiltration apparatus fitted with a YM 10 membrane (Amicon) and applied to AcA 54 gel filtration column (2.5 ϫ 110 cm) that had been equilibrated with Buffer G (20 mM Hepes (pH 8.0), 1 mM EDTA, 100 mM NaCl, 5 mM MgCl 2 , 1 mM sodium azide, 2 mM DTT, and 250 mM sucrose). Proteins were eluted at a flow rate of 0.5 ml/min. Fractions of 6 ml were collected and assayed for Arf protein and PLDactivating activity. Peak fractions (fractions 61-66) were pooled and stored in aliquots at Ϫ70°C. This preparation (12.6 mg) was greater than 90% pure based on SDS-PAGE.

Separation of Heat-labile and Heat-stable PLD Inhibitors and Purification of the Heat-labile Form
Anion-exchange Chromatography on DEAE-Sephacel Column-The 0 -35% ammonium sulfate precipitate of rat brain cytosol was suspended in Buffer H (20 mM Tris (pH 7.5), 1 mM EDTA, 1 mM DTT, and leupeptin and aprotinin (each at 1 g/ml)), extensively dialyzed overnight against the same buffer, and centrifuged to remove insoluble materials. The resulting supernatant (10 g of protein in 400 ml) was applied to a DEAE-Sephacel anion-exchange column (5 ϫ 50 cm) that had been equilibrated with Buffer H. After washing with 400 ml of Buffer H, proteins were eluted with a 2-liter linear gradient of 0 to 0.5 M NaCl in Buffer H. Fractions of 20 ml were collected and assayed for their capacity to inhibit Arf-dependent PLD activity. Three partially resolved activity peaks were observed as shown in Fig. 1A. The three peak fractions (fractions 73-77, 82-88, and 94 -98, which were named inhibitor I, II, and III, respectively) were pooled separately and kept frozen at Ϫ70°C for further purification and characteriztion.
Anion-exchange Chromatography on DEAE-5PW Column-The inhibitor I pool (700 mg of protein in 140 ml) from the DEAE-Sephacel column was dialyzed against Buffer H and centrifuged to remove insoluble materials. The resulting supernatant was divided into three equal portions, which were then chromatographed independently. Each portion was applied to a TSK-gel DEAE-5PW column (21.5 ϫ 150 mm) that had been equilibrated with Buffer H. The column was eluted at 5 ml/min with a 400-ml linear gradient from 0 to 0.3 M NaCl in Buffer H followed by a 200-ml linear gradient from 0.3 to 1.0 M NaCl in the same Buffer. Fractions of 5 ml were collected and assayed for PLD-inhibiting activity. The peak fractions (fractions 39 -47) from all three chromatography runs were pooled (Fig. 1B).
Hydrophobic Chromatography on Phenyl-5PW Column-Saturated ammonium sulfate solution was added to the combined peak fractions from the DEAE-5PW column to give a concentration of 0.8 M ammonium sulfate, and the mixture was then centrifuged to remove insoluble materials. The supernatant containing 200 mg of protein was applied to a TSK-gel Phenyl-5PW column (21.5 ϫ 150 mm) that had been equilibrated with Buffer H containing 0.8 M ammonium sulfate. Proteins were eluted at a flow rate of 5.0 ml/min with a decreasing ammonium sulfate linear gradient from 0.8 to 0 M in Buffer H for 60 min. Two activity peaks, centered at fractions 20 and 49, were observed. Peak fractions 18 -23 and 46 -52, designated inhibitors IA and IB, were pooled separately (Fig. 1C).
Cation-exchange Chromatography on Heparin-5PW Column-Peak fractions of inhibitor IA (7.4 mg of protein) and inhibitor IB (38.4 mg of protein) from the Phenyl-5PW HPLC column were dialyzed against Buffer H and loaded separately onto a TSK-gel Heparin-5PW column (7.5 ϫ 75 mm) that had been equilibrated with the same buffer. Both activities were eluted at 1.0 ml/min with a linear NaCl gradient from 0 to 0.6 M in Buffer H for 60 min. Fractions of 1.0 ml were collected. The inhibitor IA was eluted in a peak centered at fraction 30 (Fig. 1D), whereas the inhibitor IB was eluted in a peak centered at fraction 26 (Fig. 1E).
Gel Filtration Chromatography Inhibitor IA on Superose 12 PC 3.2/30 Column-Each of the inhibitor IA fractions 29 (170 g of protein) and 32 (230 g of protein) from the TSK-gel Heparin-5PW HPLC was concentrated to 30 l on a Microcon-30 (Amicon), and applied to a Superose 12 PC 3.2/30 column (3.2 ϫ 300 mm, Pharmacia). This sizing chromatography was performed on a Pharmacia-LKB SMART System equipped with a separation unit and a precision pump. The column was eluted at the flow rate of 40 l/min with 20 mM Hepes (pH 7.5) containing 150 mM NaCl. Fractions of 40 l were collected and assayed for PLD-inhibiting activity and analyzed on SDS-polyacrylamide gel (Fig. 3).
The above purification procedure starting with the 0 -35% ammonium sulfate precipitate was repeated three times to accumulate enough purified inhibitor IA for characterization.

RESULTS
Multiple Forms of PLD Inhibitor-We found that most PLDinhibiting activity in rat brain cytosol could be precipitated by adding ammonium sulfate to 35% saturation (data not shown). Fractionation of the precipitated proteins on a DEAE-Sephacel column yielded three partially resolved PLD-inhibiting activity peaks I, II, and III, which eluted sequentially as shown in Fig.  1A. Rho-GDI had previously been identified as an inhibitor of Rho-stimulated PLD but appeared not to inhibit Arf-stimulated PLD activity (27). Furthermore, immunoblotting with antibodies to Rho-GDI did not detect any reactive bands in these three peak fractions (data not shown).
In an effort to identify any physical property that distinguishes one inhibitory activity from the others, the kinetics of heat inactivation of the three fractions was evaluated. Upon incubation at 70°C, peak I lost ϳ50% of its inhibiting activity in 3 min but retained ϳ40% of its activity after 30 min. However, the activities of peaks II and III increased slightly and remained at that level during the incubation period ( Fig. 2A). At 85°C, the three inhibitor fractions lost their activities completely within 30 min but with different kinetics. Whereas peaks II and III showed slow, monophasic decreases in activity, peak I displayed a biphasic time course with a rapid inactivation phase followed by a slow one (data not shown). These results suggested that the inhibitor in peak I is most likely distinct from those in peaks II and III.
Sequential fractionation of peak I proteins on a preparative DEAE-5PW HPLC column (Fig. 1B) and a Phenyl-5PW column ( Fig. 1C) yielded two inhibitory peaks, IA and IB. Both peak IA and IB proteins were chromatographed separately under identical conditions on a Heparin-5PW column. IA activity eluted as a single peak centered at fraction 30 (Fig. 1D), whereas IB activity eluted in a main peak centered at fraction 26 and two minor peaks (Fig. 1E). The kinetics of heat inactivation was followed at 70°C using peak fraction 30 (IA) and peak fraction 26 (IB) from the Heparin-5PW column steps (Fig. 2B). More than 95% of IA activity was abolished within 1 min. Under the same conditions, IB lost rapidly only ϳ30% of its activity and retained ϳ65% of its activity after 30 min. The peak fraction from the DEAE-5PW step (mixture of IA and IB) lost ϳ50% of its activity in the rapid phase and retained ϳ35% after 30 min in a fashion similar to the result shown in Fig. 2A with the post-DEAE-Sephacel fraction I. These results suggest that the rapid phase in the inactivation of the post-DEAE Sephacel fraction I was due largely to the inactivation of inhibitor IA. Whether the biphasic time course of IB is an intrinsic property of a single inhibitor or reflects yet another mixture of inhibitors with different heat stabilities awaits further purification of IB.
Further purification of the heat-stable peak II and III inhibitors from the DEAE-Sephacel column is underway but has not progressed enough to be reported here. Nevertheless, the two peak activities remained heat-stable through several chromatography steps (data not shown).
Purification of Heat-labile Inhibitor IA-Peak fraction 26 (IB) from the Heparin-5PW column showed a number of protein bands on SDS-polyacrylamide gels (figure not shown) and was not further purified. In contrast, SDS-PAGE analysis of 8 fractions across the peak from the Heparin-5PW column containing IA showed three broad bands with apparent molecular masses of 150, 135, and 120 kDa. The 120-kDa band and probably the 135-kDa band also appeared to comprise multiple proteins with different mobility (Fig. 1F). No other protein bands were apparent in the silver-stained gel. Furthermore, while repeating the purification procedure, we noticed that the relative abundance of the three proteins bands changed and that the inhibitory activity correlated with none of the three bands but rather with the sum of the three. This suggests that inhibitor IA activity is attributable to multiple proteins and that smaller proteins might have been derived from the 150-kDa protein by proteolysis. Indeed, when protease inhibitors were not added to buffers during the first trial of purification or when the purification time was longer, the intensity of the lower molecular mass bands became stronger (figure not shown).
Two post-Heparin-5PW column fractions, fractions 29 and 32 in Fig. 1D, were subjected to gel filtration chromatography (Fig. 3A). Fraction 29 yielded two protein peaks, one small and one large, centered at fractions with an apparent molecular mass of 300 kDa and 140 kDa, respectively. In contrast, fraction 32 yielded a large peak at the fraction corresponding to 300 kDa followed by a shoulder. For both chromatographies, the PLD-inhibiting activity profiles were parallel to protein elution profiles. SDS-PAGE analysis of the fractions across the gel filtration peaks revealed that the early peak is due to a 150-kDa protein and the late peak (or shoulder) to a mixture of 135and 120-kDa proteins (Fig. 3B). These results clearly indicated that the inhibitory activity is associated with both the 150-kDa protein and the mixture of 135-and 120-kDa proteins. It was also apparent that the 150-kDa protein exists in a dimeric form, whereas the 135-and 120-kDa proteins exist in monomeric forms. Thus, the segment cleaved by proteolysis might be involved in dimerization.
In another preparation that proceeded swiftly, we resolved an inhibitor IA preparation into a fraction containing nearly homogeneous 150-kDa protein (Ͼ95%) and a fraction contain- FIG. 1. Purification of PLD-inhibiting proteins from rat brain cytosol. The 0 -35% ammonium sulfate precipitate from the cytosol of 400 g of rat brains was subjected to sequential chromatography on a DEAE-Sephacel column (A), a preparative HPLC DEAE-5PW column (B), a preparative HPLC Phenyl-5PW column (C), and HPLC Heparin-5PW columns (D and E). Detailed procedures are described under "Experimental Procedures." Bars above the elution profiles indicate those inhibitor fractions that were pooled for subsequent purification or characterization. Fractions thus pooled are hereafter collectively referred to as a "peak fraction." Peak fractions I, II, and III refer respectively to fractions 73-77, 82-88, and 94 -98 of A. Peak fraction IA (fractions 27-34) from the HPLC Heparin-5PW column was subjected to SDS-PAGE on an 8% gel, and proteins were visualized by silver staining (F).
ing largely 135-kDa protein (ϳ90%) with a small amount of 120-kDa protein (ϳ10%) (Fig. 4A). These two preparations were digested separately with trypsin and subjected to a C 18 column. The elution profiles were similar as shown in Fig. 4B, in agreement with the notion that the 135-kDa protein is a truncated fragment of the 150-kDa protein. At the present time, we do not know whether the 135-kDa form exists in cells or is an artifact of the purification procedure. However, the 120 kDa proteins(s) appeared to be generated during purification.
The sequences of two peptides through the first eight residues were determined to be EANAPAFD and GSVPLFWE. A search of the data base did not reveal any proteins with these sequences. Upon incubation at 70°C, both the 150-and 135-kDa proteins rapidly lost their inhibitory activity as did the mixture of both proteins as shown in Fig. 2B.
Characterization of Inhibitor IA-In addition to Arf proteins, other small G proteins, including RhoA and Cdc42, are known to activate PLD (12,16,17,(25)(26)(27)30). Our brain PLD preparation could be activated by Cdc42 and RhoA, and, in both cases, the resulting activation was diminished by inhibitor IA, as in the case of Arf-stimulated activity (Fig. 5). The Arf-and Cdc42-stimulated PLD activities were inhibited to a similar extent (ϳ75%) by 0.5 nM inhibitor IA, whereas the same concentration of inhibitor reduced the RhoA-stimulated PLD activity by 50%.
The concentration-dependent inhibition of Arf-stimulated PLD activity by the 150-kDa inhibitor IA was studied at 3 different concentrations of Arf (Fig. 6A). Half-maximal inhibition was observed at 0.4 nM inhibitor IA (calculated based on monomeric molecular mass) in the presence of 100 nM Arf, and the inhibition profile was nearly unchanged when the Arf concentration was reduced by 20-fold. The inhibitor concentration dependence was also measured with respect to PIP 2 and PC (Fig. 6, B and C). Neither PIP 2 nor PC concentration variance affected the profiles of inhibitor IA concentration dependence. DISCUSSION Frequently, inhibitors identified for the enzymes acting on lipid substrates are hydrophobic molecules that interact nonspecifically with lipid vesicles. However, the inhibitory action of inhibitor IA is unlikely to be due to a hydrophobic interaction

FIG. 2. Effect of heat treatment on PLD-inhibiting activity.
Peak fractions from various chromatography steps were each diluted 5-fold with Buffer H and incubated at 70°C. Aliquots of the mixture were assayed at the indicated times for PLD-inhibiting activity as described under "Experimental Procedures." A, peak fractions I (qOOq), II (åOOå), and III (fOOf) from the DEAE-Sephacel column (Fig. 1A). B, peak fraction I from the DEAE-5PW column (Fig. 1B) (qOOq), peak fraction IA from the Heparin-5PW column (Fig. 1D) (åOOå), and peak fraction IB from the Heparin-5PW column (Fig. 1E) (fOOf). because the inhibitor is a cytosolic protein that elutes in very early fractions from a hydrophobic column (Fig. 1C). Another possible interaction target for the inhibitor is PIP 2 , which is a cofactor for the PLD reaction. A number of cytosolic proteins are known to bind PIP 2 tightly and inhibit the hydrolysis of PIP 2 by phospholipase C (PLC) (33). However, the concentrations (up to 5 nM) of inhibitor IA used in this study are too low to cause inhibition directly by binding PIP 2 and do not interfere with the presumed binding of PIP 2 to PLD (Fig. 6B). Furthermore, when the effect of inhibitor IA was studied with a phospholipase C (PLC) isoform, PLC-␥1, using substrate vesicles of lipid composition identical to that used for the PLD assay, PLC activity was not affected by inhibitor IA (data not shown).
A growing body of evidence suggests the existence of multiple PLD isoforms, and the PLD preparation used in this study is likely to contain more than one isoform of PLD (25,26,29). The effect of Arf and RhoA on rat brain PLD was shown to be additive. However, it is not clear whether the additivity is due to the existence of separate PLD enzymes or separate interaction sites for Arf and RhoA on the same PLD. In the former case, inhibitor IA inhibits both Arf-and RhoA-dependent enzymes. In the latter case, inhibitor IA reduces the activity of a specific PLD independently of the activator. Neither mechanism eliminates the possibility that inhibitor IA inhibits basal activity in the absence of activators like PIP 2 and Smg proteins. However, we could not evaluate the effect of inhibitor 1A on the basal activity, because the basal activity of our partially purified PLD preparation is too low to be measured quantitatively.
The concentration of inhibitor IA required for half-maximal inhibition is 0.4 nM, while the cellular concentration of inhibitor IA appears to be much higher than 0.4 nM, as homogeneous preparation of inhibitor IA from cytosol required only several thousandfold purification according to rough estimates. Furthermore, the interaction of inhibitor IA with PLD is not affected by the presence of activated (GTP␥S-bound) Arf. Therefore, an increase in the concentration of a GTP-bound Smg protein in response to extracellular signals is not likely to be sufficient to relieve the inhibition. A specific mechanism by which receptor signaling can modulate the interaction between the inhibitor and PLD might be involved. Modulation by phosphorylation is one possibility. Considering that inhibitor 1A is extremely sensitive to proteolysis, degradation of the inhibitor in response to a PLD-activating signal is also a possibility.
While our work was in progress, Geny et al. (34) reported the identification of a bovine brain cytosolic protein that inhibits Arf-dependent PLD activity associated with HL-60 cell membrane. This inhibitory protein was eluted from a Superose 12 column in the void volume which corresponds to an apparent molecular mass larger than 300 kDa. The partially purified bovine brain inhibitor was heat-stable. Thus, inhibitor IA is unlikely to be a rat homolog of the heat-stable bovine brain inhibitor. Preliminary results from gel filtration chromatography of the peak II inhibitor from the DEAE-Sephacel column (Fig. 1A) indicate that the apparent size of inhibitor II is much The sulfhydryl groups were labeled with 2-nitro-5-thiobenzoate by adding Ellman's reagent, 5,5Ј-dithiobis-(2-nitrobenzoic acid), to a final concentration of 10 mM. The 2-nitro-5-thiobenzoate-conjugated proteins were precipitated with 10% trichloroacetic acid, and the pellet was washed with acetone, suspended in 50 mM Tris (pH 8.0), and digested with trypsin overnight at 37°C. The trypsin-digested peptides were applied to a C 18 column (4.6 ϫ 250 mm) that had been equilibrated with 0.1% trifluoroacetic acid. The elution was performed at a flow rate of 1 ml/min with a 60-ml linear gradient of 0 to 60% acetonitrile in 0.1% trifluoroacetic acid. Peptides were detected by measuring absorbance at 215 nm. The sequences of the two indicated peptides were demonstrated to be EANAPAFD (E) and GSVPLFWE (Ⅺ). larger than that (300 kDa) of untruncated IA. The peak II inhibitor, which was also heat-stable, is a better candidate for a rat homolog of the bovine brain inhibitor.