PinA Inhibits ATP Hydrolysis and Energy-dependent Protein Degradation by Lon Protease*

The bacteriophage T4 PinA protein inhibited degradation of [3H]α-methyl casein by purified Lon protease from Escherichia coli, but inhibition was noncompetitive with respect to casein. PinA did not inhibit cleavage of the fluorogenic peptide,N-glutaryl-alanylalanylphenylalanyl-3-methoxynaphthylamide and, moreover, did not block the ability of protein substrates, such as casein, to activate cleavage of fluorogenic peptides by Lon. Thus, PinA does not block the proteolytic active site or the allosteric protein-binding site on Lon. Inhibition of basal ATPase activity was variable (50–90%), whereas inhibition of protein-activated ATPase activity was usually 80–95%. Inhibition was noncompetitive with respect to ATP. PinA did not block activation of peptide cleavage by nonhydrolyzable analogs of ATP. These data suggest that PinA does not bind at the ATPase active site of Lon and does not interfere with nucleotide binding to the enzyme. PinA inhibited cleavage of the 72-amino acid protein, CcdA, degradation of which requires ATP hydrolysis, but did not inhibit cleavage of the carboxyl-terminal 41-amino acid fragment of CcdA, degradation of which does not require ATP hydrolysis. PinA thus appears to interact at a novel regulatory or enzymatic site involved in the coupling between ATP hydrolysis and proteolysis, possibly blocking the protein unfolding or remodeling step essential for degradation of high molecular weight protein substrates by Lon.

strates, such as unfolded polypeptides, activates ATPase activity and enhances cleavage of low molecular weight peptides (5,8). Yeast Lon mutants lacking proteolytic activity but having an intact ATPase can promote assembly of protein complexes (9), indicating that Lon can catalyze conformational changes in proteins. The allosteric protein-binding site is likely to be where specific degradation motifs in proteins are recognized and to be part of the catalytic domain for unfolding proteins. The ability to alter protein conformation may allow Lon to partially unfold proteins to give them greater access to the proteolytic active sites (10,11).
Cells expressing the T4 pinA gene are phenotypically Lon Ϫ , suggesting that PinA functions to inhibit the Lon protease in vivo (12). In the preceding paper, we showed that purified PinA forms a complex with Lon protease and inhibits its casein degradation activity in vitro (13). Here, we investigate the mechanism of Lon inhibition by PinA and suggest that PinA interferes with the coupling between ATP hydrolysis and protein degradation, possibly by preventing protein unfolding by Lon.

EXPERIMENTAL PROCEDURES
Purification Procedures and Standard Methods-The purification of PinA and Lon, assay for casein degradation, buffer composition, and methods for SDS-PAGE 1 and gel filtration were described in the preceding paper (13). An ATP-regenerating system consisting of 50 mM creatine phosphate and 20 g/ml phosphocreatine kinase was used in some assays.
Degradation of CcdA and N Protein-phage N protein was purified and degraded as described previously (14). For degradation, 2 g of N protein was added to 250 l of buffer containing 50 mM Tris-HCl, pH 8.0 at 25°C, 25 mM MgCl 2 , 1 mM DTT, and 4 mM ATP; in other assays, 0.5 mM AMPPNP or deionized H 2 O was substituted for ATP. The solutions were incubated for 5 min at 37°C, and degradation was initiated by the addition of 2 g of Lon. After 45 min at 37°C, the reaction was terminated by adding 310 l of ice-cold 10% trichloroacetic acid. Insoluble protein was collected by centrifugation for 15 min at 14,000 ϫ g in an Eppendorf centrifuge. The pellet was washed twice in 0.5 ml of acetone, air-dried, and dissolved in 10 l of SDS-PAGE sample buffer. Proteins were separated by SDS-PAGE, stained with Coomassie Blue, and quantitated by scanning on a Pharmacia-LKB UltroScan XL densitometer using GelScan XL software.
Purified CcdA and synthetic CcdA41, consisting of the carboxylterminal 41 amino acids of CcdA, were degraded by Lon protease as described by Van Melderen et al. (7). For quantitation of CcdA and CcdA41 degradation, the reactions were quenched with 4 M guanidine-HCl and degradation products were analyzed by reverse phase chromatography as described previously (7).
Assays for Peptidase Activity-Cleavage of the fluorogenic tetrapeptide, glutaryl-Ala-Ala-Phe-MNA, was assayed in a solution containing 50 mM Tris-HCl, pH 8.0 at 25°C, 10 mM MgCl 2 , 1 mM DTT, 4 mM ATP, 50 M peptide, 1 g of Lon, and 5 g of ␣-casein (6). The 50-l reaction * This work was supported by National Science Foundation Grant 8818950 (to L. D. S.) and by a Robert Wood Johnson predoctoral fellowship (to J. J. H.). 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  mixture was incubated in the dark at 37°C for 20 min, after which the reaction was terminated by the addition of 0.2 M sodium borate, pH 9.1, and 2 mM EDTA. Fluorescence was read on a LS-3B Perkin-Elmer fluorometer at an excitation wavelength of 335 nm and an emission wavelength of 415 nm. To examine the effect of nucleotide on peptidase activity, nucleotide was omitted, or 50 M AMPPNP was substituted for ATP.
Assays for ATPase Activity-ATPase activity was measured by the release of [ 32 P]orthophosphate from [␥-32 P]ATP. Reaction mixtures (50 l) contained 50 mM Tris-HCl, pH 8.0 at 25°C, 25 mM MgCl 2 , 1 mM DTT, and 1 mM [␥-32 P]ATP (specific activity 2 Ci mol Ϫ1 ) (15). The mixture was incubated at 37°C for 5 min, and the reaction was initiated by the addition of 1 g of Lon. After 20 min, the reaction was terminated by the addition of 200 l of an ice-cold stop solution consisting of a 1:1 mixture of 5 mM silicotungstic acid in 1 M H 2 SO 4 and 5% ammonium molybdate in 2 M H 2 SO 4 . The phosphomolybdate was extracted with 0.5 ml of solvent composed of a 1:1 mixture of isobutanol in toluene (16), and was centrifuged for 3 min at 14,000 ϫ g at 4°C. The organic phase containing [ 32 P]orthophosphate (250 l) was counted in 10 ml of Scintiverse BD scintillation fluid (Fisher).

RESULTS
Inhibition of ATP-dependent Proteolysis-Casein degradation by Lon was measured in the presence of varying amounts of PinA and casein ( Fig. 1). PinA inhibited the V max for casein degradation by Lon, but the apparent K m for casein (2.6 M) was unaffected by the presence of excess PinA. Thus, PinA is not competitive with respect to casein and apparently does not bind in the protein substrate-binding site. The inset to Fig. 1 shows the Dixon plot for PinA inhibition of casein degradation; the calculated K i for PinA was 4 Ϯ 1 nM.
PinA Inhibits Substrate-stimulated ATPase Activity of Lon-The effect of PinA on the ATPase activity of Lon was examined in the presence or absence of protein substrate. The PinA concentration required for inhibition of Lon ATPase activity was similar to that required for inhibition of casein degradation (Fig. 2). In the experiment shown, PinA inhibited about 60% of the basal ATPase activity of Lon, whereas inhibition of substrate-stimulated ATPase activity was much greater (Fig. 2). In different experiments, inhibition of basal activity varied between 50 and 80%, and inhibition of substrate-stimulated ATPase activity was usually greater, about 80 -95%. The residual substrate-stimulated activity was similar to uninhibited basal ATPase activity.
To determine if PinA inhibition of ATPase activity was competitive with ATP, assays were performed with a fixed amount of PinA and varying concentrations of ATP. As shown in Table  I, increasing the concentration of ATP to 1 mM (20 times the K m ) did not overcome PinA inhibition. No changes in the ATP concentration dependence for either basal or substrate-stimulated ATPase activity were seen in the presence of PinA (data not shown). Thus, PinA lowered the V max of ATP hydrolysis by Lon but did not appear to affect the affinity of ATP binding to Lon.
PinA Does Not Inhibit Nucleotide-activated Peptidase Activity of Lon-Hydrolysis of both peptide and protein substrates occurs at the same active site in Lon (6). Cleavage of glutaryl-Ala-Ala-Phe-MNA is activated by nucleotide binding but does not require ATP hydrolysis. PinA did not inhibit the peptidase activity of Lon on glutaryl-Ala-Ala-Phe-MNA (Table II). PinA also did not inhibit cleavage of oxidized insulin B chain or a synthetic peptide, FAPHMALVPV, in the presence of AMPPNP (data not shown). Thus, PinA does not prevent binding of nucleotides to Lon, nor does it interfere with the allosteric activation of peptidase activity upon nucleotide binding. To test whether inhibition by PinA is mediated through ADP, a potent  inhibitor of both peptide and protein degradation, casein degradation assays were conducted in the presence of an ATPregenerating system. No difference in the extent of inhibition or in the PinA concentration dependence was observed, indicating that the Lon is not acutely sensitive to inhibition by ADP in the presence of PinA. Binding of protein substrates, such as casein, to an "allosteric site" on Lon stimulates peptidase activity of Lon (8). This substrate-stimulated peptidase activity was also unaffected by excess PinA (Table II). These results indicate that protein substrate binding at the proteolytic active site and at the allosteric site on Lon is not affected by PinA.
PinA Inhibition of N Protein Degradation-The ability of PinA to inhibit degradation of natural substrates of Lon was examined first with N protein, which is degraded by Lon in the presence of ATP (14). N protein degradation also occurs in the presence of nonhydrolyzable ATP analogs but at Յ25% of the rate seen with ATP (10). Fig. 3A shows that PinA inhibited degradation of N protein in the presence of Lon and ATP. In the experiment shown, inhibition was greater than 50%, but a low rate of degradation was seen even in the presence of excess PinA. Incubation of N protein with Lon in the presence of AMPPNP resulted in degradation of ϳ50% of the N protein in 60 min, and PinA inhibited this degradation to a lesser extent (20 -25%) (Fig. 3B). Since N protein degradation is only partially dependent on ATP hydrolysis, these experiments suggested that PinA inhibition is more complete when ATP hydrolysis is required. This conclusion was confirmed by the effects of PinA on degradation of a different physiological Lon substrate, CcdA.
PinA Blocks Degradation Coupled to ATP Hydrolysis-CcdA, a protein made by F factor, is degraded in vitro by Lon only when accompanied by ATP hydrolysis (7). A truncated form of CcdA, CcdA41, lacks stable secondary structure and can be degraded by Lon in the presence of nonhydrolyzable analogs of ATP (7). PinA completely blocked CcdA degradation in the presence of ATP but did not inhibit degradation of CcdA41 in the presence of AMPPNP (Fig. 4). CcdA41 degradation occurs at a similar rate in the presence of ATP or AMPPNP, and PinA was unable to inhibit degradation in either case (Table III). Thus, PinA blocks a step in Lon-dependent degradation that requires ATP hydrolysis, probably energy-dependent conformation rearrangement (or unfolding) of substrate proteins. Since PinA blocks ATP hydrolysis, these data also show that ATP binding produces a similar allosteric activation of Lon activity as is produced by nonhydrolyzable analogs. DISCUSSION Purified PinA inhibited casein degradation by Lon protease, and inhibition appeared to be noncompetitive with casein. The apparent K i for PinA (3-4 nM) is consistent with tight binding between the two proteins, as had been suggested by the isola-

TABLE II PinA effects on peptidase activity of Lon
The assay solutions (50 l) contained 50 M glutaryl-Ala-Ala-Phe-MNA and 1 g of Lon with or without 1 g of PinA. The ATP concentration was 4 mM, and that of AMPPNP was 20 M. To examine substrate-stimulated peptidase activity, 5 g of ␣-casein was added. tion of a stoichiometric complex of PinA and Lon by gel filtration. PinA affinity for Lon is comparable to that observed for other low molecular weight protein protease inhibitors and their targets (Table IV). The very high affinity of PinA for Lon suggests a specific function for PinA in inhibiting Lon activity in vivo. Unlike most protease inhibitors, PinA apparently does not inhibit protein degradation by interfering with substrate binding or peptide bond hydrolysis at the proteolytic active site of its target protease.
The lack of inhibition of Lon peptidase activity and the noncompetitive nature of inhibition of casein degradation indicate that PinA binds outside of the proteolytic active site on Lon. Protein substrates bind to Lon and stimulate hydrolysis, suggesting that there is an allosteric binding site for protein substrates on Lon (8). PinA did not interfere with the ability of casein to stimulate peptidase activity (Table II), indicating that PinA does not block casein binding at the allosteric site, and thus PinA does not appear to bind, or at least does not completely occupy, the allosteric protein-binding site on Lon.
PinA does not completely inhibit nucleotide hydrolysis, as evidenced by the 50% basal ATPase activity remaining at concentrations of PinA that block 90 -95% of protein degradation (Fig. 2). However, protein-activated ATPase activity was completely blocked by PinA. Inhibition of both basal and substrate-stimulated ATP hydrolysis is independent of ATP concentration. Peptidase activity and CcdA41 degradation are stimulated by nucleotide binding and were unaffected by PinA. PinA is not competitive with ATP or ATP analogs and does not bind at the nucleotide site. Thus, PinA binds to a unique site distinct from previously described functional sites on Lon.
Nucleotide binding and hydrolysis have distinct roles in Lon activity. Binding of nucleotides, such as ATP, nonhydrolyzable ATP analogs, or CTP, induces a conformation change which enhances peptidase activity of Lon (6). However, nucleotide binding is not sufficient to allow degradation of proteins with stable secondary or tertiary structure, which requires ATP hydrolysis (7). Since Lon peptidase activity is stimulated by nucleotides in the presence of PinA, PinA does not appear to affect the nucleotide-induced conformational change in Lon to the "activated" state.
Binding of protein substrates at the allosteric site induces a conformational change in Lon that releases bound ADP (3). The release of ADP allows Lon to continue the catalytic cycle of peptide bond cleavage, and failure to release ADP would "freeze" Lon in a conformationally inactive state (3). Since PinA did not inhibit peptide or CcdA41 degradation when either ATP or the nonhydrolyzable analog, AMPPNP, was used, PinA does not lock Lon in an inactive ADP-bound state.
Recent data suggest that energy-dependent proteases need ATP hydrolysis to drive enzyme catalyzed conformational changes or unfolding of protein substrates. Yeast Lon protease and the ATPase domains of Clp proteases have chaperone-like activities and can remodel protein structures (9,10,17). Studies with E. coli Lon have shown that the presence of stable secondary structure in a substrate is the determinant of the requirement for ATP hydrolysis (7). With Clp proteases and the 26 S proteasome, disruption of the secondary structure of protein substrates is apparently required to allow substrates to get to the proteolytic active sites, which are located in an interior aqueous cavity accessible through relatively narrow channels. Although the structure of Lon is not known, it is likely that the active sites of Lon are similarly inaccessible to folded proteins.
PinA specifically blocks the steps in protein degradation that require ATP hydrolysis. Degradation of CcdA (which requires unfolding) but not of CcdA41 (which does not require unfolding) was inhibited, suggesting that PinA inhibits the chaperone activity of Lon. We propose that PinA prevents the unfolding of protein substrates by interfering with ATP hydrolysis and the conformational changes in Lon that accompany ATP hydrolysis. These conformational changes in Lon may alter its interactions with bound proteins, which in turn result in structural changes in the protein substrate.
Our previous data showed that a dimer of PinA binds to a tetramer of Lon (13). Lon has four potential ATPase sites per tetramer, and inhibition of all stimulated ATPase activity by a single PinA dimer implies that Lon subunits act in a cooperative manner. Since it does not block access to the active sites for several protein substrates, PinA probably does not bind in the substrate channel to the active sites. We propose that PinA binds to a flexible region on Lon that is important in cooperative ATP-driven conformational changes that are coupled to unfolding of protein substrates. This binding site on Lon may represent a functional site that is occupied by other effectors of Lon in vivo, and its identification should give additional information about physiological regulation of Lon.  (20). d N-Carbobenzoxy-Phe-Arg-7-amido-4-methyl coumarin. e N-Benzoyl-DL-Arg-p-nitroanilide. f Not applicable. g SSI, Streptomyces subtilisin inhibitor.