Journal of Biotechnology 119 (2005) 163171

Lon and ClpP proteases participate in the physiological disintegration of bacterial inclusion bodies
Andrea Vera, Anna Ars, Mar Carri , Nuria Gonz lez-Montalb n, o a a Antonio Villaverde
Institut de Biotecnologia i de Biomedicina, Departament de Gen` tica i de Microbiologia, Universitat Aut` noma de Barcelona, e o Bellaterra, 08193 Barcelona, Spain Received 9 September 2004; received in revised form 14 April 2005; accepted 19 April 2005

Abstract Aggregated protein is solubilized by the combined activity of chaperones ClpB, DnaK and small heat-shock proteins, and this could account, at least partially, for the physiological disintegration of bacterial inclusion bodies. In vivo, the involvement of proteases in this process had been suspected but not investigated. By using an aggregation prone -galactosidase fusion protein produced in Escherichia coli, we show in this study that the main ATP-dependent proteases Lon and ClpP participate in the physiological disintegration of cytoplasmic inclusion bodies, their absence minimizing the protein removal up to 40%. However, the role of these proteases is clearly distinguishable especially regarding the fate of solubilized protein. While Lon appears as a minor contributor in the disintegration process, ClpP directs an important attack on the released or releasable protein even not being irreversibly misfolded. ClpP is then observed as a wide-spectrum, main processor of aggregation-prone proteins and also of polypeptides physiologically released from inclusion bodies, even when occurring as soluble versions with a conformation compatible with their enzymatic activity. 2005 Elsevier B.V. All rights reserved.
Keywords: Aggregation; E. coli; Inclusion bodies; Protein folding; Proteolysis

1. Introduction Bacterial inclusion bodies are amorphous aggregates resulting from the deposition of insoluble polypeptide chains (Villaverde and Carri , 2003). o
Corresponding author. Tel.: +34 935812148; fax: +34 935812011. E-mail address: antoni.villaverde@uab.es (A. Villaverde).

They are commonly occurring during the overexpression of foreign genes, whose products are kinetically trapped as stable intermediates of decient folding processes. In bacteria, thermal stress but also the overproduction or misfolding-prone proteins trigger the expression of heat-shock genes, whose products, chaperones and proteases, act coordinately to minimize the occurrence of conformationally aberrant protein forms (Yura et al., 1993; Arsene et al., 2000; Feldman

0168-1656/$ see front matter 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.jbiotec.2005.04.006

164

A. Vera et al. / Journal of Biotechnology 119 (2005) 163171

and Frydman, 2000). Bacterial chaperones DnaK and GroEL (and their cochaperones DnaJ-GrpE and GroES, respectively) address folding of folding reluctant protein forms in a sequential way (Buchberger et al., 1996; Goloubinoff et al., 1999), while small heat-shock proteins such as IbpA/B protect misfolded proteins from irreversible aggregation (Thomas and Baneyx, 1998; Shearstone and Baneyx, 1999; Schlieker et al., 2002). Also, chaperones belonging to the AAA+ family such as ClpB, in cooperation with DnaK and small heat-shock proteins, remove polypeptides from aggregates for their eventual refolding (Thomas and Baneyx, 2000; Mogk et al., 2003a,b; Mogk and Bukau, 2004; Weibezahn et al., 2004). The activity of these and other heat-shock elements might account for an almost complete in vivo disintegration of inclusion bodies in absence of the novo protein synthesis (Carri and Villaverde, 2001). Therefore, o the formation and dissolution of protein aggregates in bacteria results from unbalanced kinetics of protein deposition and removal, both processes being simultaneous and thus producing a continuous physiological reorganization of inclusion body composition even during their volumetric growth (Carri and Villaverde, o 2002). On the other hand, ATP-dependent proteases degrade misfolded polypeptides that are reluctant to chaperone-mediated folding. How substrates are selected for either further folding attempts or digestion is still a matter of discussion (Tomoyasu et al., 2001; Dougan et al., 2002). It is currently accepted that proteolysis is mostly restricted to soluble targets, aggregation resulting then in proteolytic stabilization by limiting the accessibility of protease target sites. However, inclusion body protein is efciently digested in vitro by trypsin (Carri et al., 2000) in a cascade process that o renders progressively shorter products (Cubarsi et al., 2001). In vivo, aggregated polypeptides are cleaved in a site-limited process that is linked to their transfer from the insoluble to the soluble cell fraction (Corchero et al., 1997; Carri et al., 1999). This o observation indicates that proteolysis, apart from the activity of disaggregating chaperones, might be also involved in the physiological dissolution of bacterial protein aggregates, although this possibility has not been further analysed. In this work we have explored the involvement of Lon and ClpP in the formation and disintegration of inclusion bodies. These proteases

are responsible for more of 70% of the cellular ATP-dependent proteolysis (Maurizi, 1992) and are in charge of the degradation of aggregation-prone proteins (Tomoyasu et al., 2001; Rozkov and Enfors, 2004), a condition shared by most of bacterially produced foreign polypeptides. The obtained results prove a dramatic role of these proteases in the dissolution process as well as a dissimilar substrate selection in their activity. These data allow the conceptual integration of the activities of chaperones and proteases for the surveillance of protein aggregation and the cellular managing of protein aggregates.

2. Materials and methods 2.1. Bacterial strains and plasmids The Escherichia coli strains used in this work were MC4100 (araD139 (argF-lac) U169 rpsL150 relA1 bB5301 deoC1 ptsF25 rbsR) (Sambrook et al., 1989) and their derivatives JGT19 ClpP (clpP::cat) (Thomas and Baneyx, 1998) and BB2395 Lon ( lon146::miniTn10) (Tomoyasu et al., 2001). Plasmid pJCO46 encodes a soluble, pseudo-wild type E. coli -galactosidase (Corchero et al., 1996), and pJVP1LAC a -galactosidase derivative with the VP1 capsid protein of foot-and-mouth disease virus fused to its amino terminus (Corchero et al., 1996). Both plasmids direct gene expression from lambda pR and pL lytic promoters, controlled by a constitutively expressed temperature-sensitive CI repressor. The presence of the viral stretch largely impairs the folding of -galactosidase and the production of VP1LAC in E. coli results in its aggregation as large cytoplasmic inclusion bodies, only a fraction of the polypeptides reaches the soluble and enzymatically active form (Corchero et al., 1996). 2.2. Media and growth conditions Rich LB medium (Sambrook et al., 1989) was used for cell growth, supplemented with 100 g ml1 ampicillin for plasmid maintenance. Streptomycin at 30 g/ml was also added when necessary. Cultures were performed in shaker asks at 28 C at 250 rpm until the OD550 reached 0.3, and then transferred to a pre-warmed bath at 42 C to trigger recombinant gene

A. Vera et al. / Journal of Biotechnology 119 (2005) 163171

165

expression. Chloramphenicol at 200 g ml1 was added 3 h after induction of gene expression to arrest protein synthesis, and cultures were further incubated at 37 C at 250 rpm. At different times, samples were processed as described below. 2.3. Analysis of protein solubility and determination of -galactosidase activity Solubility of VP1LAC was determined by Western blot analysis of soluble and insoluble cell fractions. Briey, 10 ml-culture samples were centrifuged for 15 min at 12,000 g and resuspended in 500 l of Z buffer without -mercaptoethanol (Miller, 1972), with one tablet of protease inhibitor cocktail (Roche, Ref. 1 836 170) per 10 ml buffer. Samples jacketed in ice were sonicated for 7 min at 50 W under 0.5 s cycles, and centrifuged for 15 min at 12,000 g. Soluble and insoluble fractions were resuspended in denaturing buffer (Laemmli, 1970) up to the same volume. After boiling for 15 min, dened sample volumes were loaded onto SDS-polyacrilamide gels. Immunoreactive bands were visualised by using an anti- -galactosidase polyclonal antibody and quantied, within a linear range, on digitalized blots with the Quantity One software from Bio-Rad. -Galactosidase activity was determined in 0.1 ml-culture samples as described (Ferrer-Miralles et al., 2001). Samples from three independent experiments were analysed independently and values are usually given as the mean plus standard deviation. In some cases, only data series from representative experiments are shown.

2.4. Protein and DNA analysis Total cell protein was determined by analysing cell pellets after microcentrifugation of 3 ml samples and media washing, through standard Bradfords procedure (Sambrook et al., 1989). Correlations between cell protein and optical density were signicant for all the strains, with only slight differences in the slope values. For convenience, optical density was regularly used as a measure of biomass. Intracellular plasmid DNA content was estimated by alkaline lysis extraction, agarose gel electrophoresis and further image analysis as described (Medina et al., 2002). 3. Results 3.1. Enzymatic activity of VP1LAC fusion protein produced in absence of either Lon or ClpP To evaluate the status of the misfolding-prone protein produced in either ClpP and Lon mutants, and to check if VP1LAC could result toxic in absence of these proteases, we monitored biomass evolution during the induction of VP1LAC gene expression. In fact, it was previously observed that the production of this protein is, at some extent, deleterious for E. coli (Corchero and Villaverde, 1998). In absence of ClpP, VP1LAC-producing cells showed a growth curve indistinguishable from that of wild type cells, while in the Lon mutant the optical density increased more slowly (Fig. 1A). On the other hand, in all the tested strains at least a fraction of VP1LAC was enzymatically active,

Fig. 1. Growth (A) and -galactosidase activity (B) of wild type MC4100 (white circles), ClpP JGT19 (black circles) and Lon BB2395 (triangles) cultures during VP1LAC production. Time 0 indicates the induction of VP1LAC gene expression.

166

A. Vera et al. / Journal of Biotechnology 119 (2005) 163171

Fig. 2. Growth of MC4100 (A) and Lon BB2395 (B) without any plasmid (black circles), or containing either pJCO46 (white circles) or pVP1LAC (black triangles). Time 0 indicates the shift from 28 to 42 C to trigger recombinant gene expression in plasmid-bearing cells. Differences at 3 h were observed as not signicant in an ANOVA test (p > 0.1).

and the accumulation of active protein (relative to cell biomass) was again similar in the wild type and the ClpP mutant, but slower in the Lon background (Fig. 1B). The growth of Lon cells might eventually be impaired by the production of the misfolding-prone VP1LAC. To investigate this possibility, we compared the growth of BB2395 Lon producing either VP1LAC or its parental CO46, a totally soluble and functional protein. As observed, the slow growth rate of BB2395 was not associated to the production of any CO46 or VP1LAC plasmid-encoded proteins as it happened in the wild type cells (Fig. 2). This fact was then probably caused by the lon mutation itself, since it is known that SulA, a regulator of cell division, is specically digested by Lon (Mizusawa and Gottesman, 1983) and that the absence of Lon results in signicant perturbations in the cell division (Schoemaker et al., 1984). Therefore, we concluded that the eventual toxicity of VP1LAC was not modied by the absence of either ClpP or Lon. 3.2. Production and solubility of VP1LAC in absence of either Lon or ClpP The total amount of full-length VP1LAC in cell extracts of MC4100 and its derivatives ClpP and Lon was determined during the induction of the encoding gene expression. Interestingly, the inactivation of lon and clpP genes had signicant although opposite effects on the total amount of the produced

fusion protein (Fig. 3A and B). While as expected, the absence of ClpP enhanced VP1LAC yield probably by reducing its proteolytic digestion, the absence of the protease Lon resulted in signicantly lower amounts of recombinant protein per cell biomass. This fact indicated that the proteolytic degradation of VP1LAC was not minimized in absence of Lon but on the contrary it seems to be slightly stimulated. Although the differential growth of Lon cell cultures could eventually have resulted in an unperceived over accumulation of VP1LAC, the precise consideration of biomass for sample processing was expected to minimize any important bias in the analysis. Also, the particular evolution of VP1LAC amounts after induction of gene expression (Fig. 3A) is supportive of enhanced digestion rather than less protein production in Lon cells. On the other hand, and to discard possible inuences of gene dosage in the total protein amount (as eventually modulated by protease gene mutations), plasmid DNA content was estimated in the three strains during induction of VP1LAC gene expression. At 3 h after thermal shift, plasmid amounts were 1.86, 2.04 and 2.48 ng DNA/mg protein for wild type, ClpP and Lon strains, respectively. To further explore the status of VP1LAC under these conditions, its partitioning into soluble and insoluble cell fractions was explored. The evolution of the soluble fraction followed a similar pattern in wild type and Lon strains during protein production, solubility being around 80% but slightly decreasing with the production time. Interestingly, the solubility of VP1LAC

A. Vera et al. / Journal of Biotechnology 119 (2005) 163171

167

Fig. 3. Total amounts of VP1LAC (A) and relative values of soluble VP1LAC (C) in cell extracts as measured by densitometric units (DU) from a representative experiment. Wild type cells are shown in black, the lon mutant in light grey and the clpP mutant in dark grey bars. Average and standard error values of total VP1LAC from three independent experiments, measured at 3 h after induction of gene expression. (B) A dened value of cell biomass, as adjusted by OD values, was loaded in all the gel lanes. Relative to the total cell protein, VP1LAC content was 266.6 0.6, 102.4 3.3 and 365.9 5.6 DU/mg for wild type, Lon and ClpP strains, respectively. Note that the absolute densitometric unit values from different analysis of Western blot data are not necessarily comparable.

in ClpP cells was much lower than in the other strains and it decreased from around 50% at 1 h to around 30% at 3 h (Fig. 3C). In general, it is accepted that high productivity of a recombinant protein results into lower solubility (and vice versa), because of the higher substrate load for folding assistant proteins. Therefore, the results shown in Fig. 3C are in agreement with the observed differences in the protein yield (Fig. 3A and B), again supporting that the protease ClpP is a main processor of this protein. 3.3. Stability of VP1LAC in absence of either Lon or ClpP The results presented above indicated that the protease ClpP regulates the stability of the aggregation-prone VP1LAC fusion protein. In

recombinant bacteria, inclusion bodies are continuously solubilized in vivo by chaperones, even during their volumetric growth by deposition of misfolded protein chains (Carri and Villaverde, 2002). Protein o hydrolysis has been observed as associated to these antagonistic processes (Corchero et al., 1997; Carri o et al., 1999), but the precise mechanistic involvement of proteases has not been yet explored. Therefore, we evaluated the inuence of ClpP and Lon in the physiological disintegration of VP1LAC inclusion bodies, once protein synthesis is arrested, by monitoring the remaining amount of insoluble VP1LAC. As observed in Fig. 4A, while in wild type cells only 10% of the aggregated protein was reluctant to removal, the fraction of insoluble protein remained stable at around 35% for Lon cells and 50% for ClpP cells. This indicates that the absence of either Lon or ClpP

Fig. 4. Relative amounts of VP1LAC in the insoluble cell fraction (A) and absolute amounts of soluble VP1LAC (B), during in vivo dissolution of protein aggregates. For both A and B panels, wild type cells are shown in black, the lon mutant in light grey and the clpP mutant in dark grey bars. The enzymatic activity of the soluble cell fraction is also presented in wild type (black triangles), lon mutant (white circles) and the clpP mutant (black circles) (C). In all the panels, time 0 indicates the arrest of protein synthesis.

168

A. Vera et al. / Journal of Biotechnology 119 (2005) 163171

proteases impairs the physiological release of protein from inclusion bodies. On the other hand, we had previously shown that most of the polypeptides released from inclusion bodies are highly unstable, and they are nally degraded in the cytoplasm by a cascade process that renders progressively shorter fragments (Carri and o Villaverde, 2001). Therefore, we also explored the fate of the protein removed from aggregates once transferred to the soluble cell fraction. In both wild type and Lon cells, the soluble protein released from aggregates was very unstable (Fig. 4B). However, in absence of ClpP, the stability of the soluble protein was dramatically enhanced, indicating that this protease is processing an important amount of proteins removed from inclusion bodies. In this context, we monitored the evolution of the enzymatic activity in the soluble fraction while aggregates disintegrate. The enzymatic activity in the ClpP strain increased more than two times and it remained highly stable, while contrarily, in the wild type and Lon cells only an extremely slight and transient increase in the activity levels was observed (Fig. 4C). These results indicate that the solubilized VP1LAC, even reaching a conformation compatible with the enzymatic activity, is a substrate for ClpP (although not for Lon), and also that ClpP is digesting an important amount of polypeptides removed from aggregates, as associated to the in vivo disintegration of inclusion bodies.

4. Discussion The controlled degradation of intracellular proteins is a critical activity within the complex cell biology network. Apart from removing damaged and/or deleterious proteins for quality control purposes, proteases are involved in important regulatory mechanisms (Gottesman and Maurizi, 1992), among others the adaptation to stationary-phase (Weichart et al., 2003) and starvation (Schweder et al., 1996), the heat-shock (Bukau, 1993) and general stress (Zhou et al., 2001) responses, cell division (Schoemaker et al., 1984), mutagenesis (Frank et al., 1996), phage (Gayda and Markovitz, 1978) and plasmid replication (Maas, 2001) and capsule synthesis (Torres-Cabassa and Gottesman, 1987). The ATP-dependent serine proteases ClpP and Lon are responsible for more than

70% of the proteolysis in E. coli (Maurizi, 1992) covering most of the regulatory functions listed above, and they are also critical for the removal of misfolded or otherwise damaged proteins (Tomoyasu et al., 2001; Rosen et al., 2002). Although in general it has been believed that aggregated polypeptides are, at least to a great extent, protected from proteolysis, data obtained in our laboratory indicated that they are targets for in vivo proteolytic attack (Corchero et al., 1997; Carri et o al., 1999; Carbonell and Villaverde, 2002), that could be mechanistically associated to protein removal from inclusion bodies during their physiological disintegration (Carri and Villaverde, 2001). Here we have shown o the involvement of both ClpP and Lon proteases in this process, since the absence of any of them impairs, although at different extent, the release of misfoldingprone VP1LAC protein (Fig. 4A). The comparative analysis of the protein amount that remains associated to inclusion bodies in absence of de novo protein synthesis, indicates that the absence of ClpP impairs protein removal of more than 50% of inclusion body protein, while in absence of Lon this gure is around 30%. Indirect effects of protease mutations on inclusion body disintegration cannot be discarded, since the higher load of misfolded cell and recombinant proteins in these cells can reduce the amounts of free disaggregating chaperones. However, in a double DnaK ClpB mutant protein release is only reduced 44.5% when compared with the wild type (Carri and o Villaverde, 2003), strongly suggesting that the effects of protease mutations on protein release have to be mostly direct. On the other hand, in the ClpP mutant, the amount of VP1LAC that appears in the soluble cell fraction during the inclusion body disintegration is dramatically higher than in the wild type (Fig. 4B). This fact clearly indicates that ClpP is responsible for the attack of very recently refolded (or refolding) polypeptides. In addition, ClpP might act through in situ proteolysis of VP1LAC chains exposed on the solvent-exposed interface of bodies. Irrespectively of the precise location of the digestion event, the concomitant raise of -galactosidase enzymatic activity in absence of ClpP (Fig. 4C) indicates that the substrates for this protease are not path-off folding intermediates, but polypeptide chains that can still reach their native conformation or already properly folded and assembled in tetramers.

A. Vera et al. / Journal of Biotechnology 119 (2005) 163171

169

While ClpP shows a major role in the proteolysis coupled to inclusion body disintegration, the absence of the protease Lon does not result in any detectable increase of either VP1LAC amount or activity in the soluble cell fraction (Fig. 4), indicating different roles or substrate specicities of both enzymes. Also, the absence of Lon does not enhance the total yield of VP1LAC, contrarily to that observed in the ClpP mutant (Fig. 3A and B). These observations could indicate that in absence of Lon, ClpP (or eventually other cell proteases) could degrade VP1LAC, and that this complementing activity might be enhanced due to higher substrate load for folding assistant proteins in a Lon background. ClpP would be then, a secondary protease for the substrates commonly attacked by Lon during inclusion body disintegration. In this context, ClpP interacts with chaperones ClpA and ClpX to form either ClpAP, ClpXP or ClpXAP complexes (Ortega et al., 2004). In these complexes, the chaperone component confers distinct substrate specicity by regulating the accessibility to the protease active site (Hoskins et al., 1998). Contrarily, the specicity of Lon is self-determined by an amino terminal conserved domain within the same polypeptide chain (Ebel et al., 1999). However, it has also been proved that Lon and ClpP have similar and at least partially overlapping targets (Wickner and Maurizi, 1999). In fact, the suspected redundancy in targeted proteolysis has been already proven experimentally for Lon and ClpYQ (Wu et al., 1999). Interestingly, the enhanced VP1LAC protein yield observed in ClpP cells (Fig. 3) is irrespective of ClpA (results not shown, and (Carri o and Villaverde, 2002)), as also observed for specic targets of ClpP (Weichart et al., 2003). The connection between protein disaggregation and degradation reported here and identied in previous studies (Corchero et al., 1997; Carri et al., o 1999) could lie on two different but not exclusive mechanisms. First, it is known that the proteolytic attack on inclusion bodies is not surface-restricted, but it generates an intense fragmentation of the bodies due to heterogeneous stabilities of the embedded polypeptides (Carri et al., 2000). This digestion pato tern dramatically enlarges the solvent-surface of the aggregates during their digestion, what offers extended interfaces for the activity of refolding chaperones. Therefore, a mild proteolytic attach could largely favour non-proteolytic protein release from inclusion

bodies. Second, the disaggregating activity of refolding chaperones could progressively offer new target sites to cell proteases thus initiating the proteolytic cascade attack on polypeptides that are being solubilized. Finally, the roles of Lon and ClpP proteases in processing inclusion bodies are distinguishable from those in the managing thermal aggregates, in which Lon but not ClpP might have a main role in a low-specicity digestion process (Rosen et al., 2002). The physiological features of these two kinds of aggregates appear indeed dissimilar and are currently under study in our laboratory.

Acknowledgements We are grateful to A. Mogk and F. Baneyx for generously providing strains BB2395 and JGT20, respectively. This work has been supported by grants BIO2004-00700 (MEC, Spain), 2002SGR-00099 (AGAUR, Spain), and by Maria Francesca de Roviralta Foundation.

References
Arsene, F., Tomoyasu, T., Bukau, B., 2000. The heat shock response of Escherichia coli. Int. J. Food Microbiol. 55 (13), 39. Buchberger, A., Schroder, H., Hesterkamp, T., Schonfeld, H.J., Bukau, B., 1996. Substrate shuttling between the DnaK and GroEL systems indicates a chaperone network promoting protein folding. J. Mol. Biol. 261 (3), 328333. Bukau, B., 1993. Regulation of the Escherichia coli heat-shock response. Mol. Microbiol. 9 (4), 671680. Carbonell, X., Villaverde, A., 2002. Protein aggregated into bacterial inclusion bodies does not result in protection from proteolytic digestion. Biotechnol. Lett. 24, 19391944. Carri , M.M., Corchero, J.L., Villaverde, A., 1999. Proteolytic digeso tion of bacterial inclusion body proteins during dynamic transition between soluble and insoluble forms. Biochim. Biophys. Acta 1434 (1), 170176. Carri , M.M., Cubarsi, R., Villaverde, A., 2000. Fine architecture of o bacterial inclusion bodies. FEBS Lett. 471 (1), 711. Carri , M.M., Villaverde, A., 2001. Protein aggregation as bacterial o inclusion bodies is reversible. FEBS Lett. 489 (1), 2933. Carri , M.M., Villaverde, A., 2002. Construction and deconstruction o of bacterial inclusion bodies. J. Biotechnol. 96 (1), 312. Carri , M.M., Villaverde, A., 2003. Role of molecular chaperones in o inclusion body formation. FEBS Lett. 537 (13), 215221. Corchero, J.L., Cubarsi, R., Enfors, S., Villaverde, A., 1997. Limited in vivo proteolysis of aggregated proteins. Biochem. Biophys. Res. Commun. 237 (2), 325330.

170

A. Vera et al. / Journal of Biotechnology 119 (2005) 163171 Mizusawa, S., Gottesman, S., 1983. Protein degradation in Escherichia coli: the lon gene controls the stability of sulA protein. Proc. Natl. Acad. Sci. USA 80 (2), 358362. Mogk, A., Bukau, B., 2004. Molecular chaperones: structure of a protein disaggregase. Curr. Biol. 14 (2), R78R80. Mogk, A., Deuerling, E., Vorderwulbecke, S., Vierling, E., Bukau, B., 2003a. Small heat shock proteins. ClpB and the DnaK system form a functional triade in reversing protein aggregation. Mol. Microbiol. 50 (2), 585595. Mogk, A., Schlieker, C., Friedrich, K.L., Schonfeld, H.J., Vierling, E., Bukau, B., 2003b. Refolding of substrates bound to small Hsps relies on a disaggregation reaction mediated most efciently by ClpB/DnaK. J. Biol. Chem. 278 (33), 3103331042. Ortega, J., Lee, H.S., Maurizi, M.R., Steven, A.C., 2004. ClpA and ClpX ATPases bind simultaneously to opposite ends of ClpP peptidase to form active hybrid complexes. J. Struct. Biol. 146 (12), 217226. Rosen, R., Biran, D., Gur, E., Becher, D., Hecker, M., Ron, E.Z., 2002. Protein aggregation in Escherichia coli: role of proteases. FEMS Microbiol. Lett. 207 (1), 912. Rozkov, A., Enfors, S.O., 2004. Analysis and control of proteolysis of recombinant proteins in Escherichia coli. Adv. Biochem. Eng. Biotechnol. 89, 163195. Sambrook, J., Fritsch, E., Maniatis, T., 1989. Molecular Cloning. A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Schlieker, C., Bukau, B., Mogk, A., 2002. Prevention and reversion of protein aggregation by molecular chaperones in the E. coli cytosol: implications for their applicability in biotechnology. J. Biotechnol. 96 (1), 1321. Schoemaker, J.M., Gayda, R.C., Markovitz, A., 1984. Regulation of cell division in Escherichia coli: SOS induction and cellular location of the sulA protein, a key to lon-associated lamentation and death. J. Bacteriol. 158 (2), 551561. Schweder, T., Lee, K.H., Lomovskaya, O., Matin, A., 1996. Regulation of Escherichia coli starvation sigma factor (sigma s) by ClpXP protease. J. Bacteriol. 178 (2), 470476. Shearstone, J.R., Baneyx, F., 1999. Biochemical characterization of the small heat shock protein IbpB from Escherichia coli. J. Biol. Chem. 274 (15), 99379945. Thomas, J.G., Baneyx, F., 1998. Roles of the Escherichia coli small heat shock proteins IbpA and IbpB in thermal stress management: comparison with ClpA, ClpB, and HtpG in vivo. J. Bacteriol. 180 (19), 51655172. Thomas, J.G., Baneyx, F., 2000. ClpB and HtpG facilitate de novo protein folding in stressed Escherichia coli cells. Mol. Microbiol. 36 (6), 13601370. Tomoyasu, T., Mogk, A., Langen, H., Goloubinoff, P., Bukau, B., 2001. Genetic dissection of the roles of chaperones and proteases in protein folding and degradation in the Escherichia coli cytosol. Mol. Microbiol. 40 (2), 397413. Torres-Cabassa, A.S., Gottesman, S., 1987. Capsule synthesis in Escherichia coli K-12 is regulated by proteolysis. J. Bacteriol. 169 (3), 981989. Villaverde, A., Carri , M.M., 2003. Protein aggregation in recomo binant bacteria: biological role of inclusion bodies. Biotechnol. Lett. 25 (17), 13851395.

Corchero, J.L., Viaplana, E., Benito, A., Villaverde, A., 1996. The position of the heterologous domain can inuence the solubility and proteolysis of beta-galactosidase fusion proteins in E. coli. J. Biotechnol. 48 (3), 191200. Corchero, J.L., Villaverde, A., 1998. Plasmid maintenance in Escherichia coli recombinant cultures is dramatically, steadily, and specically inuenced by features of the encoded proteins. Biotechnol. Bioeng. 58 (6), 625632. Cubarsi, R., Carri , M.M., Villaverde, A., 2001. In situ proteo olytic digestion of inclusion body polypeptides occurs as a cascade process. Biochem. Biophys. Res. Commun. 282 (2), 436 441. Dougan, D.A., Mogk, A., Bukau, B., 2002. Protein folding and degradation in bacteria: to degrade or not to degrade? That is the question. Cell. Mol. Life Sci. 59 (10), 16071616. Ebel, W., Skinner, M.M., Dierksen, K.P., Scott, J.M., Trempy, J.E., 1999. A conserved domain in Escherichia coli Lon protease is involved in substrate discriminator activity. J. Bacteriol. 181 (7), 22362243. Feldman, D.E., Frydman, J., 2000. Protein folding in vivo: the importance of molecular chaperones. Curr. Opin. Struct. Biol. 10 (1), 2633. Ferrer-Miralles, N., Feliu, J.X., Vandevuer, S., Muller, A., CabreraCrespo, J., Ortmans, I., Hoffmann, F., Cazorla, D., Rinas, U., Prevost, M., Villaverde, A., 2001. Engineering regulable Escherichia coli beta-galactosidases as biosensors for anti-HIV antibody detection in human sera. J. Biol. Chem. 276 (43), 4008740095. Frank, E.G., Ennis, D.G., Gonzalez, M., Levine, A.S., Woodgate, R., 1996. Regulation of SOS mutagenesis by proteolysis. Proc. Natl. Acad. Sci. USA 93 (19), 1029110296. Gayda, R.C., Markovitz, A., 1978. Altered bacteriophage lambda expression in cell division mutants capR(lon) of Escherichia coli K-12. Mol. Gen. Genet. 159 (1), 111. Goloubinoff, P., Mogk, A., Zvi, A.P., Tomoyasu, T., Bukau, B., 1999. Sequential mechanism of solubilization and refolding of stable protein aggregates by a bichaperone network. Proc. Natl. Acad. Sci. USA 96 (24), 1373213737. Gottesman, S., Maurizi, M.R., 1992. Regulation by proteolysis: energy-dependent proteases and their targets. Microbiol. Rev. 56 (4), 592621. Hoskins, J.R., Pak, M., Maurizi, M.R., Wickner, S., 1998. The role of the ClpA chaperone in proteolysis by ClpAP. Proc. Natl. Acad. Sci. USA 95 (21), 1213512140. Laemmli, U.K., 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227 (259), 680685. Maas, R., 2001. Change of plasmid DNA structure, hypermethylation, and Lon-proteolysis as steps in a replicative cascade. Cell 105 (7), 945955. Maurizi, M.R., 1992. Proteases and protein degradation in Escherichia coli. Experientia 48 (2), 178201. Medina, M.G., Carbonell, X., Villaverde, A., 2002. Connection between gene dosage and protein stability revealed by a highyield production of recombinant proteins in an E. coli LexA1(Ind) background. Biotechnol. Bioeng. 78 (7), 722730. Miller, J.H., 1972. Experiments in Molecular Genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

A. Vera et al. / Journal of Biotechnology 119 (2005) 163171 Weibezahn, J., Bukau, B., Mogk, A., 2004. Unscrambling an egg: protein disaggregation by AAA+ proteins. Microb. Cell Fact. 3 (1), 1. Weichart, D., Querfurth, N., Dreger, M., Hengge-Aronis, R., 2003. Global role for ClpP-containing proteases in stationary-phase adaptation of Escherichia coli. J. Bacteriol. 185 (1), 115125. Wickner, S., Maurizi, M.R., 1999. Heres the hook: similar substrate binding sites in the chaperone domains of Clp and Lon. Proc. Natl. Acad. Sci. USA 96 (15), 83188320.

171

Wu, W.F., Zhou, Y., Gottesman, S., 1999. Redundant in vivo proteolytic activities of Escherichia coli Lon and the ClpYQ (HslUV) protease. J. Bacteriol. 181 (12), 36813687. Yura, T., Nagai, H., Mori, H., 1993. Regulation of the heatshock response in bacteria. Annu. Rev. Microbiol. 47, 321 350. Zhou, Y., Gottesman, S., Hoskins, J.R., Maurizi, M.R., Wickner, S., 2001. The RssB response regulator directly targets sigma(S) for degradation by ClpXP. Genes Dev. 15 (5), 627637.