Isolation and Characterization of the Phage T4 PinA Protein, an Inhibitor of the ATP-dependent Lon Protease ofEscherichia coli *

The bacteriophage T4 PinA protein, expression of which leads to inhibition of protein degradation in Escherichia coli cells, has been purified from cells carrying multiple copies of the pinA gene. PinA is a heat-stable protein with a subunit M r of 18,800 and an isoelectric point of 4.6. Under nondenaturing conditions on a gel filtration column, PinA migrated in two peaks corresponding to a dimer and a tetramer. Purified PinA inhibited ATP-dependent protein degradation by Lon protease in vitro; it did not inhibit the activity of otherE. coli ATP-dependent proteases, ClpAP or ClpYQ. Furthermore, PinA did not inhibit ATP-independent proteolysis inE. coli cell extracts. PinA binds with high affinity to Lon protease (K d ∼ 10 nm for dimer binding), and a complex with ∼1 dimer of PinA per tetramer of Lon protease could be isolated by gel filtration. Lon activity was partially restored upon dilution of the PinA-Lon complex to subnanomolar concentrations, indicating that inhibition was reversible and that PinA did not covalently modify Lon protease. PinA was not cleaved by Lon protease, and heating the Lon-PinA complex at 65 °C denatured Lon protease and released active PinA. The properties of PinAin vitro suggest that PinA inhibits protein degradationin vivo by forming a tight, reversible complex with Lon protease.

Lon protease, the product of the lon gene (1), is one of the major ATP-dependent proteases of Escherichia coli. In vivo, Lon is responsible for the degradation of such specific proteins as SulA (2) RcsA (3), and the N protein (4), as well as unfolded and abnormal proteins (5). Purified Lon, subunit M r 87,000, is an oligomeric protein that has been reported to exist in either tetrameric or octameric form (6,7). Sequence analysis suggests that each subunit of Lon contains a proteolytic active site and an ATPase site distinct from the proteolytic site (8). Cleavage of small peptides and some small proteins requires nucleotide binding to Lon but does not require ATP hydrolysis (9 -11), but degradation of high molecular weight proteins requires ATP hydrolysis (12)(13)(14). Under optimal conditions with a variety of protein substrates, two ATP molecules are hydrolyzed per peptide bond cleaved (15).
Hydrolysis of ATP may provide energy to help unfold protein substrates, giving them greater access to the proteolytic active site and making them more susceptible to cleavage (13,16,17). Recent studies of CcdA degradation by Lon in vitro indicated that the absence of stable secondary structure in protein substrates decreased the requirement for ATP hydrolysis (11). Protein substrates bind to two sites on Lon, the proteolytic active site and an allosteric site, which may serve as the site for the protein remodeling function. Occupancy of the allosteric site by substrates such as unfolded polypeptides activates the peptidase activity against small peptides and enhances proteolysis (12). E. coli cells infected with bacteriophage T4 show reduced proteolysis of abnormal proteins and protein fragments (18). Inhibition of protein degradation requires synthesis of T4 proteins made during the first 10 min after infection at 37°C (18 -20). Clones of T4 genes that lead to inhibition of proteolysis in E. coli cells were obtained by Simon and co-workers (19,20), and one specific gene, pinA (proteolysis inhibition A), which resulted in inhibition of abnormal protein degradation, was cloned by Skorupski et al. (21). The target of PinA in vivo appears to be the ATP-dependent Lon protease, because E. coli lon ϩ cells expressing a single copy of the T4 pinA gene are phenotypically Lon Ϫ (21).
We have purified the PinA protein and shown that purified PinA binds to Lon protease and inhibits its ATP-dependent protein degrading activity in vitro. In the accompanying paper (22), we demonstrate that PinA exerts its affect by blocking ATP hydrolysis by Lon.

EXPERIMENTAL PROCEDURES
Materials-All chemicals were obtained from commercial sources unless otherwise specified. [ 3 H]Formaldehyde was obtained from NEN Life Science Products. Clp protease was purified as described previously (23)(24)(25).
Growth of Bacterial Strains-Bacterial strains and plasmids used are described below. Bacteria used to make cell extracts were grown in glucose L broth (31), which contained (per liter) 10 g of tryptone, 5 g of yeast extract, 5 g of NaCl, and 2 g of glucose. Glucose L agar contained glucose L broth with 1.5% (w/v) agar. Ampicillin was used at 50 g/ml as needed.
Purification of Lon Protease-Lon protease was purified from E. coli SG22030 (lon ϩ ) carrying the multicopy plasmid plon ϩ 500, which contains lon under its own promoter (5). The purification method has been described (6). Lon was estimated by SDS-PAGE 1 to be 90 -95% pure and was free of other proteases.
Overexpression and Preparation of PinA-The pinA gene product was expressed from a plasmid constructed by J. Tomaschewski 2 using the expression vectors of Tabor and Richardson (26). The host strain was E. coli LS101, a derivative of K38 (27) carrying a galE mutation to prevent production of excess capsular polysaccharide in the absence of lon function in vivo. Cells were transformed with plasmid pGP1-2, which has T7 RNA polymerase under a heat-inducible promoter (26), and plasmid pT7AT/pin, which has the pinA gene under control of a T7 promoter. To induce expression of the pinA gene, cells grown at 28°C were shifted to 42°C for 30 min and left at 40°C for 2 h. Harvested cells were stored at Ϫ80°C.
For purification of the PinA protein, 24 g of frozen cells were defrosted in an ice bath and suspended in 90 ml of 50 mM Tris-HCl, pH 7.5 (measured at 25°C), containing 2 mM EDTA and either 1 mM ␤-mercaptoethanol or 2 mM DTT (buffer T). Purification steps were carried out at 0 -4°C. Cells were lysed by passage through a French pressure cell at 20,000 p.s.i., and extracts were centrifuged at 27,000 ϫ g for 1 h. Solid ammonium sulfate (30% of saturation) was added to the supernatant, and after 30 min the precipitate was collected by centrifugation at 20,000 ϫ g for 20 min. The resulting pellet was dissolved in 80 ml of buffer T, and the ammonium sulfate precipitation was repeated. The pellet from the second precipitation was stored at Ϫ80°C overnight.
The ammonium sulfate pellet was dissolved in 80 ml of buffer T, and the solution was centrifuged at 20,000 ϫ g for 20 min to remove insoluble material. The supernatant was filtered through a 0.22-m filter (Millipore) in aliquots of 20 ml, and aliquots were loaded onto separate MonoQ HR 10/10 columns. Proteins were eluted with a linear gradient of 0.2-0.6 M KCl in buffer T at a flow rate of 1 ml/min. Fractions containing PinA identified by SDS-PAGE were pooled, and ammonium sulfate was added to 40% saturation. After centrifugation for 15 min at 30,000 ϫ g, the pellet was suspended in 10 -15 ml buffer T with 0.1 M KCl. Insoluble matter was removed by centrifugation, and the resulting supernatant was loaded in aliquots of 2.5 ml onto separate 2.3 ϫ 60-cm TSK250 gel filtration columns equilibrated in buffer T with 0.1 M KCl. Proteins were eluted in the same buffer at a flow rate of 2 ml/min. Purified PinA was stored at Ϫ80°C.
PinA was also purified on a smaller scale in 20 mM PIPES, pH 6.0 at 4°C, 1 mM ␤-mercaptoethanol, 0.2 M NaCl, 1 mM EDTA, 200 M phenylmethylsulfonyl fluoride, 1 M leupeptin, and 1 M pepstatin. Following cell lysis and centrifugation, the PinA protein was precipitated with ammonium sulfate as described above, and the pellet was suspended in PIPES buffer and dialyzed overnight. The dialysate was centrifuged for 10 min at 14,000 ϫ g in an Eppendorf microcentrifuge. The resulting supernatant was then loaded onto a MonoQ HR 16/10 anion exchange column. The column was developed stepwise with a gradient from 0.2 M to 1 M NaCl in 20 mM PIPES, pH 6.0, at a flow rate of 3 ml/min. When the elution gradient reached 0.3 M NaCl, 0.46 M NaCl, and 1.0 M NaCl, the NaCl concentration was kept constant for 10 min. PinA-containing fractions were pooled, concentrated in a Centriprep-10 concentrator (Amicon), and loaded onto a Superose-12 HR 10/30 gel filtration column (Pharmacia Biotech Inc.). Proteins were eluted from the Superose-12 column in 20 mM PIPES, pH 6.0, containing 0.5 M NaCl, 1 mM ␤-mercaptoethanol, and 1 mM EDTA buffer, at a flow rate of 0.2 ml/min. Column fractions with PinA were pooled, concentrated, and run over the Superose-12 column again in the same buffer. The fractions containing PinA were pooled, dialyzed overnight against 50 mM Tris-HCl, pH 7.5, 2 mM EDTA, 1 mM DTT, and 10% glycerol (buffer B), and stored at Ϫ80°C.
Biochemical Characterization and Physical Properties-One-dimensional SDS-PAGE gels were run essentially as described by Laemmli (28), using Mini-PROTEAN II Ready Gels (Bio-Rad). Except as noted, 12% gels were used. The molecular weight of PinA under nondenaturing conditions was determined by gel filtration on a Superose-12 column by comparing the elution time of the PinA protein to the elution times of proteins of known molecular weights. The subunit molecular weight was estimated by SDS-PAGE. The isoelectric point of the PinA protein was determined using Pharmacia isoelectric focusing 3-9 Phastgels. Protein concentrations were determined by the dye-binding method of Bradford (29) using the reagent supplied by Bio-Rad, with bovine serum albumin as the standard. PinA concentrations were measured from the absorbance using the extinction coefficient.
Amino Acid Analysis-Purified PinA was hydrolyzed in 6 N HCl at 155°C for 30 and 60 min. The hydrolysates were washed, dried, and derivatized with phenylisothiocyanate (PTC). PTC-amino acids were separated on a C18 reverse phase column (4.6 mm ϫ 15 cm) using the solvent system described by Bidlingmeyer et al. (30). The cysteine content of the PinA protein was determined by performic acid oxidation (31), followed by acid hydrolysis and amino acid analysis. Aromatic amino acids were determined spectrophotometrically by second derivative UV spectroscopy in 6 M guanidine hydrochloride, as described by Levine and Federici (32). Using the amino acid content calculated from the DNA-derived sequence of PinA, the extinction coefficient of the protein was determined from the calculated aromatic amino acid con-tent of the protein and the measured absorbance of a standard solution of the protein.
NH 2  Assays for ATP-dependent Proteolytic Activity-Assays for proteolytic activity were performed as described previously (9,23) or as follows. A solution with 9 g of [ 3 H]methyl ␣-casein in 250 l of buffer containing 50 mM Tris-HCl, pH 8.0, 25 mM MgCl 2 , 1 mM DTT, and 4 mM ATP was incubated for 5 min at 37°C, and the reaction was initiated by the addition of 0.5-2.0 g of Lon. Incubation at 37°C was continued for 15-30 min. The reaction was terminated by the addition of 310 l of ice-cold 10% trichloroacetic acid and 40 l of 10 mg/ml bovine serum albumin. Precipitated proteins were separated from trichloroacetic acid-soluble proteins by centrifugation at 4°C in an Eppendorf centrifuge at 14,000 ϫ g for 6 min. Radioactivity was determined by liquid scintillation counting using 0.5 ml of the supernatant in 10 ml of Scintiverse BD (Fisher) or Aquasol (NEN Life Science Products). Assays were performed in duplicate, and the measured activity had a variance of Յ4%.
Inhibition of Lon by PinA-PinA was added to standard assays solutions 1-5 min prior to addition of Lon to initiate the assays. Mixing PinA and Lon before adding them to the assay mixtures did not affect the results. The effect of pH on the inhibitory activity of PinA was determined by substituting the following buffers in the assay mixture: 50 mM MES, sodium salt, 10 mM MgCl 2 , pH 6.0 and 7.0; 50 mM Tris-HCl, 10 mM MgCl 2 , pH 7.0 -9.0; and 50 mM 2-amino-2-methyl-1-propanol HCl, 10 mM MgCl 2 , pH 9.0 -10.5.
Stoichiometry of the Lon-PinA Complex-A 2.5-ml Sephacryl S-200 (Pharmacia) column was equilibrated at room temperature in buffer B. Lon (20 g) was loaded onto the column, and 250-l aliquots of buffer were added at 2 min intervals. Aliquots of 250 l were collected from the column at each step. The same procedure was used for PinA (20 g) and for a mixture of 20 g of Lon and 20 g of PinA. Nucleotide requirement was determined by equilibrating the column in buffer B containing 50 M AMPPNP, and then chromatographing Lon and PinA. Alternatively, Lon, PinA, or the PinA-Lon complex was centrifuged through a Bio-Spin column packed with Sephacryl Superfine S-200 according the manufacturer's instructions.
Lon (250 g), PinA (250 g), and mixtures of the two proteins were also analyzed by gel filtration on a Superose12 column equilibrated in buffer B. Proteins were eluted at a flow rate of 0.4 ml/min in the same buffer. Protein in the fractions was detected by SDS-PAGE, and fractions were assayed for [ 3 H]methyl ␣-casein degrading ability. After SDS-PAGE, the protein was stained with Coomassie Blue and quantitated by densitometry with a Hewlett-Packard ScanJet IIc/ADF using Deskscan and Collage software or by capturing a digital image of the gel with an Eagle Eye Frame Integrator and using NIH Image software. Known amounts of Lon or PinA were used as standards. To determine if formation of the PinA-Lon complex was reversible, the complex isolated by gel filtration was diluted into assay solutions of increasing volume (100 l to 1.6 ml) to reduce the concentrations of PinA and Lon below the apparent K d , and the increase in Lon activity was measured.

RESULTS
Purification of the PinA Protein-PinA was purified to near homogeneity from extracts of cells in which the protein was overproduced. Previous work had shown that PinA was a component of the cytosol and was not associated with the cell membrane. 3 Purified PinA in fractions from the final gel filtration step is shown in Fig. 1A; PinA was estimated to be Ն95% pure in the best fractions, which were stored separately and used for the experiments described below. PinA had a maximum absorbance at 281 nm and an extinction coefficient of 1.95 (mg/ml) Ϫ1 determined by analysis of the second derivative of the UV absorbance spectrum (32).
The pinA gene in the plasmid used for these studies was sequenced and found to be identical to the published DNA sequence (data not shown). Purified PinA had an apparent M r of 20,800 by SDS-PAGE (data not shown) in reasonable agreement with the predicted M r of 18,000. The isoelectric point was 4.6 (data not shown), in good agreement with the predicted pI of 4.3. The amino-terminal amino acid sequence of the protein determined by Edman degradation was found to be MITVDK-WFRINRADTGLCNY, which is identical to that translated from the pinA DNA sequence (35). The amino acid composition of the protein also agreed well with that predicted from the published sequence (data not shown).
In the absence of a reducing agent during purification, or after repeated freeze-thaw cycles, a second band of PinA, apparent M r ϳ19,000, was detected on SDS-PAGE (data not shown). It is not known whether PinA was partially degraded by a contaminating protease or was subject to chemical cleavage or modification. No loss of inhibitory activity of PinA was observed upon prolonged storage of the protein at Ϫ90°C.
Inhibitory Activity of PinA-Inhibition of Lon protease activity was examined using [ 3 H]methyl ␣-casein as the substrate and by following the ATP-dependent release of trichloroacetic acid-soluble peptide fragments. Extracts of cells in which PinA was overproduced contained an inhibitor of Lon protease, and no such inhibitory activity was seen with extracts of control cells with the plasmid vector, which did not contain PinA (data not shown). The inhibitory activity against Lon protease coeluted with the PinA protein during purification (Fig. 1B). Fig. 2 shows the concentration dependence of inhibition of Lon-dependent casein degradation by purified PinA. Inhibition was Ն90% in the presence of a sufficient excess of PinA. In this experiment, PinA was added to the assay solutions in increasing amounts prior to starting the reaction by addition of Lon protease. The amount of PinA required to inhibit Lon was less than 1% of the casein present in the assay solution, and, there-fore, inhibition was not due to interaction of PinA with the substrate. PinA apparently binds very rapidly to Lon, since prior incubation of the two proteins together did not increase the degree of inhibition (data not shown). PinA has a very high affinity for Lon under the conditions used for these assays; half-maximal inhibition occurred at about 20 nM PinA subunit or only 10 nM PinA dimer. PinA purified by both procedures described under "Experimental Procedures" had comparable inhibitory properties.
In vivo studies suggested that other ATP-dependent E. coli proteases were not affected by PinA (22). As shown in Table I, PinA did not inhibit casein degradation by the ATP-dependent proteases, ClpAP or ClpYQ (HslUV). Thus, PinA appears to specifically recognize Lon protease.
Stability of PinA-PinA was stable when heated. Table II shows that PinA inhibited Lon to the same extent before and after heating for 10 min at ϳ100°C. Inhibition of Lon by PinA required the intact PinA protein, because digestion of PinA by trypsin or chymotrypsin inactivated PinA (data not shown). This experiment also demonstrated that Lon inhibition was not due to nonproteinaceous inhibitors contaminating the PinA.
To determine if PinA was cleaved by Lon, PinA was incubated with Lon for 5 min at 37°C, and the reaction was terminated by boiling the proteins in SDS. The proteins were separated by SDS-PAGE and quantitated by densitometry after staining with Coomassie Blue. As shown in Fig. 3, no loss of PinA was detectable after incubation with Lon. Thus, PinA does not appear to be a substrate for Lon.
Oligomeric Structure of PinA-Freshly purified PinA migrated with an apparent molecular weight of 40,000 on a Superose 12 gel filtration column in the presence of 0.1-0.2 M KCl and thus appears to be a dimer. After storage at Ϫ20°C for more than 1 year, PinA species that appeared by gel filtration to be tetramers (see Fig. 4) and octamers (data not shown) predominated. No noticeable effect on the ability of PinA to inhibit Lon protease accompanied these changes in oligomeric state, and it is possible that the aggregated PinA dissociated to dimers and tetramers at the dilutions used for assays. High salt concentrations (Ն0.3 M KCl) cause PinA to dissociate into monomers and, when present in assay solutions, decrease the inhibitory effects of PinA (data not shown). These last data suggest either that the monomeric form of PinA does not inhibit Lon or that the interaction between PinA and Lon is disrupted by high ionic strength.
PinA inhibition of Lon was optimal between pH 8 and 9 (data not shown). Above pH 9.5, inhibitory activity decreased considerably, but the decreased inhibition could reflect conforma- tional changes in either protein, because both Lon and PinA appear to lose activity at higher pH. PinA was not irreversibly inactivated at pH 10.5, because PinA preincubated at pH 10.5 was able to inhibit Lon when added to the standard pH 8.0 assay buffer (data not shown).
Demonstration of a Lon-PinA Complex-PinA binding to Lon was shown by isolation of a complex of the two proteins on a Sephacryl S-200 gel filtration column. Fig. 4 shows the fractions containing PinA and Lon protease when the two proteins were run over the column either separately or after mixing together. When the proteins were chromatographed separately, the elution position of PinA, which was a mixture of dimers and tetramers, was much later than that of Lon. When the proteins were run together, PinA and Lon were found together in fractions eluting slightly ahead of the elution position of Lon alone, and there was a decrease in the protein peak at the elution position of PinA alone (Fig. 4A). There was considerable overlap in the positions of the Lon-PinA complex and Lon alone, partly due to the tendency of Lon to trail severely in these columns. The complex could be isolated in the presence or absence of nucleotide (data not shown).
The proteolytic activity of Lon in fractions containing the isolated complex was assayed and compared with that present in similar fractions when Lon protease was run alone. As shown in Fig. 4B, the casein degrading activity of the PinA-Lon complex was only 5-10% of the activity of Lon protease alone.
The amount of PinA bound to Lon was determined in a separate experiment by mixing the two proteins in a ratio of five PinA dimers per subunit of Lon and isolating the complex by gel filtration on a Sephacryl S-200 column. Protein in the fractions corresponding to the complex was quantitated by comparing the intensity of the Coomassie-stained bands to known amounts of Lon or PinA that had been run and stained in parallel. Approximately 1-2 dimers of PinA were bound to 1 tetramer of Lon. Increasing the amount of PinA added to Lon did not increase the PinA found complexed to Lon (data not shown).  Release of Active PinA from the Lon-PinA Complex-The release of active PinA from the PinA-Lon complex was demonstrated by taking advantage of the thermal stability of PinA. Aliquots of Lon and the PinA-Lon complex from the Sephacryl spin-column eluates were incubated at 65°C for 10 min, after which the aliquots were immediately added to reaction mixtures with and without fresh Lon. Table III shows that, after 10 min at 65°C, Lon alone or in the complex with PinA had no proteolytic activity. Addition of the heated PinA-Lon complex to assay solutions with fresh Lon resulted in inhibition. The inhibition was not due to interference from the heated and presumably denatured Lon, since the addition of heated Lon alone to the reaction mixture had no effect on casein degradation by fresh nondenatured Lon. Thus, no irreversible change in PinA accompanies binding to and inhibition of Lon protease.
Reversibility of Lon Inhibition by PinA-The PinA-Lon complex was isolated by gel filtration on Sephacryl S-200. Eluates were collected and were assayed for ATP-dependent casein degradation in reaction solutions of different volumes. Table IV shows that Lon in the complex isolated from the column was inhibited Ͼ90% when assayed at high concentrations of the complex, but dilution of the complex resulted in a progressive increase in the Lon activity. Similar results were obtained without isolation of the complex by gel filtration. PinA and Lon were premixed in a ratio sufficient to cause Ͼ90% inhibition of Lon activity when assayed in 100-l reaction solutions. Dilution of the same mixture into larger volumes for assay resulted in dissociation of the complex and a 40% gain in Lon activity (Table IV). In other experiments, 40 -85% of Lon activity was recovered at 8 -16-fold dilution of the PinA-Lon complex into assay solutions (data not shown). These results suggest that, although the PinA-Lon complex is quite stable, complex formation is reversible. DISCUSSION The bacteriophage T4 PinA protein has been purified from E. coli cells carrying a multicopy plasmid with the pinA gene under control of a T7 promoter. Purified PinA inhibits ATP-dependent casein degradation by Lon protease by more than 90%. PinA shares several properties with other polypeptide protease inhibitors. It is relatively small (M r 20,000), heat-stable, and acidic (pI ϭ 4.6); however, PinA does not inhibit proteases such as trypsin, chymotrypsin, subtilisin, or pepsin but is degraded and inactivated by treatment with these proteases. PinA ap-pears to target Lon protease specifically, and elsewhere (22) we demonstrate that PinA inhibits ATP-dependent protein degradation by Lon protease by blocking the coupling between ATP hydrolysis and peptide bond cleavage. Inhibition of [ 3 H]methyl ␣-casein degradation by Lon protease occurs at very low concentrations of PinA, and an apparent K i of about 5 nM was calculated under standard assay conditions. The low K i implies that PinA binds tightly to Lon, which was confirmed by showing that it was possible to isolate a complex of Lon and PinA after gel filtration chromatography. The complex was formed in the absence of nucleotide as well as in the presence of the nonhydrolyzable ATP analog, AMPPNP, indicating that tight binding of PinA to Lon does not require any of the energy-driven steps involved in protein degradation.
As expected, when used at concentrations comparable to those in standard assays, the PinA-Lon complex isolated by gel filtration showed little proteolytic activity against [ 3 H]methyl ␣-casein. However, dilution of the complex resulted in a small increase in proteolytic activity, suggesting that inhibition of Lon by PinA is at least partially reversible. Heating the PinA-Lon complex released PinA from denatured Lon protease. The free PinA was able to bind to Lon and inhibit proteolytic activity. SDS gel analysis of PinA after forming a complex with Lon or released from the complex by heating showed no evidence of cleavage of the PinA.
The classical protease inhibitors from plants and microbial organisms are characterized by either reversible or irreversible mechanisms (37). Reversible inhibitors have a specific peptide bond, which combines with the active site of the target protease and is then cleaved by the protease. Hydrolysis does not proceed to completion, but an equilibrium between intact and cleaved peptide bonds is established. Irreversible inhibitors combine with the protease and are cleaved at a specific peptide bond. However, the acyl intermediate between the inhibitor and the protease is not hydrolyzed and the inhibitor remains covalently bound to the protease. The latter inhibitor-enzyme complex resists dissociation by urea and SDS. The results of the thermal inactivation and dilution studies of the PinA-Lon complex suggest that PinA may differ from other reversible inhibitors in that it is not cleaved by the protease.
Skorupski et al. (21) have shown that lon ϩ cells lysogenic for in which the pinA gene has been cloned behave phenotypically like lon mutants. These lysogens produce mucoid colonies, filament in response to DNA damage, permit efficient plaque formation by Ots phage at 40°C, and exhibit reduced levels of b PinA is retained in the column and is not present in the eluate.

Reversibility of Lon inhibition by PinA
The PinA-Lon complex was formed, and the Lon was assayed at different dilutions. In part A, the PinA-Lon complex formed with saturating PinA was isolated by gel filtration on Sephacryl S-200 and assayed in increasing reaction volumes to dilute the complex. In part B, PinA was titrated into Lon to produce Ͼ90% inhibition in the smallest reaction volume, and identical aliquots of the mixture were assayed in solutions of increasing volume. Results from single experiments are shown. abnormal protein degradation, all typical of E. coli cells lacking functional Lon protease (17,38). Expression of pinA has no detectable effect on abnormal protein degradation in E. coli null lon strains (21). Our finding that PinA does not inhibit ClpAP or ClpYQ activity in vitro is consistent with the idea that PinA displays specificity for Lon protease alone. ATP-dependent proteases tend to be high molecular weight, multimeric enzymes with potential binding sites for regulatory components. Several endogenous inhibitors that bind reversibly to the eukaryotic 20 S and 26 S proteasomes have been described, each of which appears to have a unique mode of action. For example, an inhibitor described by Chu-Ping et al. (39) apparently acts allosterically since it inhibits three distinct peptidase activities of the 20 S proteasome, whereas a ubiquitinated inhibitor from rabbit reticulocytes blocks ATPdependent degradation of ubiquitinated proteins by the 26 S proteasome but has only partial activity against peptidase activities (40). A multimeric factor, CF2, combines with the 20 S proteasome and appears to inhibit peptidase activity (41). Since CF2 and the 20 S proteasome are both components of the 26 S proteasome, this inhibition probably reflects changes in accessibility of the proteolytic active sites during assembly of the 26 S proteasome, which has ATP-dependent proteolytic activity and has a more stringent specificity, preferentially degrading ubiquitinated proteins. No inhibitors of the E. coli ClpAP or ClpXP proteases have yet been described. The CIII protein of inhibits the ATP-dependent FtsH (HflB) protease; however, in vivo data suggest that CIII simply acts as a competitive substrate for the protease, and CIII does not inhibit Lon protease in vitro. 4 A search of the GenBank and Swiss-Prot data bases (42) revealed no homologies to pinA at either the DNA or amino acid sequence level. PinA, therefore, appears to be a novel protease inhibitor, highly specific for Lon protease. Further characterizations of the effects of PinA on the proteolytic, peptidase and ATPase activities of Lon are described in the accompanying paper (22).