PRIORITY PAPER The Thermoplasma acidophilum Lon protease has a Ser-Lys dyad active site Henrike Besche and Peter Zwickl Department of Structural Biology, Max-Planck-Institute of Biochemistry, Martinsried, Germany A g ene with significant s imilarity to bacterial Lon proteases was identified during the s equencing o f the genome of the thermoacidophilic archaeon Thermoplasma acidophilum. Protein sequence comparison revealed t hat Thermoplasma Lon protease ( TaLon) is more similar t o t he LonB proteases restricted to Gram-positive bacteria than to the widely dis- tributed bacterial L onA. However, th e active site residues of the protease and ATPase domain a re highly conserved i n all Lon proteases. U sing site-directed mutagenesis we s how here that TaLon and EcLon, and probably all other L on pro- teases, contain a Ser-Lys dyad activ e site. The Ta Lon active site mutants were fully assembled and, similar to TaLon wild-type, displayed an apparent molar mass 1 of 430 kDa upon gelfiltration. This would be consistent with a hexameric complex and indeed electron micrographs of TaLon revealed ring-shaped p articles, although of unknown sym- metry. Comparison of the ATPase activity o f Lon wild-type from Therm oplasma or Escherichia coli with respective pro- tease a ctive s ite mutants revealed differences in K m and V values. T his suggests th at in the course of protein d egra- dation by wild-type Lon the protease domain might influ- ence the activity of the ATPase domain. Keywords:AAA + protease; archaea; Lon(La) endopepti- dase; Lon (La) protease; Ser-Lys dyad. Endopeptidase La (EC 3.4.21.53) was the first ATP- dependent proteo lytic enzyme to be identified [1,2]. Later protease La was found to be the product of the Escherichia coli lon gene 2 [3,4] and is now mainly called Lon p eptidase or protease. T his c an be seen by the respective entry and listed references in the MEROPS peptidase database: peptidase family S16 (lon protease family) (http://merops.sanger.ac.uk/famcards/summary/ s16.htm) [5]. The L on protease domain is ubiquitously distr ibuted and m ostly fused to different ATPase domains [6]. However, in the genomes of some Bacteria and A rchaea, standalone Lon protease domains are present, but nothing is known about their biological function. In contrast, plenty of knowledge has accrued about bacterial and mitochondrial Lon pr oteases [7–10], where t he Lon domain is linked t o an N -terminal AAA + domain which, i n turn, is extended N-terminally by a Lon N-terminal (LAN) domain [6]. Certain bacteria contain a second Lon protease, called LonB, which lacks the LAN domain [6] and is assumed to be soluble [11]. Most Archaea contain a LonB homologue, i.e. lacking the LAN domain, which contains two transmembrane-spanning regions and w as shown to b e membrane-bound [12,13]. It has been known for a l ong time that Lon proteases have an active site serine residue [14], but despite extensive mutagenesis the residual catalytic residues r emained e lusive [15]. Ultimately, mutagenesis studies of a viral noncanon- ical Lon protease l acking the ATPase domain revealed the catalytic Ser-Lys dyad ( S-K dyad) for Lon p roteases [16]. We set out to mutate the conserved serine and lysine residues in Thermoplasma acidophilum Lon protease (TaLon) to generalize the S-K dyad for ATP-dependent membrane-bound Lon proteases. In the course o f this work an independent report confirmed the S-K dyad for the E. coli Lon protease (EcLon), but without studying mutual regulation of the ATPase a nd protease domain [ 17]. Based on mutagenesis studies it was p roposed that the ATPase domain of EcLon regulates the protease in a unidirectional manner, i.e. the mutational inactivated protease domain did not influence t he ATPase activity [18]. We investigated this aspect in more detail for TaLon and Ec Lon b y detailed a nalysis o f p rotease active s ite mutants. Materials and methods Sequence alignments The a ctive-site regions of Lon protease protein sequences were aligned with CLUSTAL X [19] on Macintosh PPC. Correspondence to P. Zwickl, Max-Planck-Institute of Biochemistry, Department of Structural Biology, Am Klopferspitz 18, 82152 Martinsried, Germany. Fax: +49 89 85782641, Tel.: +49 89 85782647, E-mail: zwickl@biochem.mpg.de Abbreviations: AAA + , ATPase associated with various cellular activities; DDM, dodecyl-b- D -maltopyranoside; EcFtsH, Escherichia coli FtsH; EcLon, Escherichia c o li Lon; LAN, LonN-terminal; Lonwt, Lon wild-type; S-K, Ser-Lys; TaLon, Thermoplasma acidophilum Lon; TaLonwt, Thermoplasma acidophilum Lon wild-type. Enzyme: e ndopeptidase L a (EC 3.4.21.53). (Received 19 August 2 0 04, revised 1 O cto ber 2004, accepted 6 October 2004) Eur. J. Biochem. 271, 4361–4365 (2004) Ó FEBS 2004 doi:10.1111/j.1432-1033.2004.04421.x Generation of active site mutants Site-directed mutagenesis was performed using the Q uick- ChangeÒ Kit (Stratagene, La Jolla, CA, USA) 3 . pET22b(+)-TaLon-His 6 (S525A and K568A) and Lon wild-type (pLonwt) (K722A) served a s PCR t emplates. The respective primers are listed with the mutated codons underlined: S525A, 5¢-CGAGGGAGTTGAAGGAGAC GCGGCCAGCGTATCAATAGCC-3¢ (sense), 5¢-GGCT ATTGATACGCTGGC CGCGTCTCCTTCAACTCCCT CG-3¢ (antisense); K568A, 5 ¢-CCGGTTGGCGGCGTAAC CGCA GCGGTTGAGGCAGCTATAGAAGC-3¢ (sense), 5¢-GCTTCTATAGCTGCCTCAAC CGCTGCGGTTAC GCCGCCAACCGG-3¢ (antisense); K7 22A, 5¢-GCCGA TGGTGGTTTGAAAGAA GCCCTCCTGGCAGCGCA TCGCG -3¢ (sense), 5¢-CGCGATGCGCTGCCAGGAG GGCTTCTTTCAAACCACCGATCGGC-3¢ (a ntisense). DNA sequencing ( MWG, E bersberg, Germany) 4 of the full-length gene che cked all generated plasmids. Expression, purification and enzymic characterization Wild-type a nd muta nt Ta Lon p roteases were produced, and the hydrolysis of fluorigenic peptides, FITC-labelled casein and ATP were assayed as described [13]. R. G lockshuber (ETH Zu ¨ rich, Switzerland) kindly provided the plasmids pLonwt and pLonS679A [20]. Cells of E. coli BL21(DE3) (Novagen, M adison, WI, USA) 5 harbouring p Lon (-wt, -S679A or -K722A) were grown at 37 °Cin6LLuria– Bertani m edium containing 100 lgÆmL )1 ampicillin. A t an attenuance 6 of 0.6 (600 nm) isopropyl thio-b- D -galactoside was a dded to a final concentration o f 1 m M .Afterfurther incubation for 4 h the cells were harvested b y centrifugation (4000 g,10min4°C),washedwith50m M Tris (pH 7.5) and s tored until purification at )80 °C. Purification of the EcLon variants was performed as described [20]. Results Domain organization of Lon proteases During the sequencing of the T. ac idophilum genome an open reading frame (ORF 1081) was identified, which showed significant sequence similarity to Lon proteases [21] (Fig. 1 A,B). TaLon encompasses an N-terminal ATPase associated with various cellular a ctivities ( AAA + domain) and a C-terminal protease domain, but lacks the N-terminal a-helical domain inherent in most bacterial and eukaryotic Lon homologues (Fig. 1C). However, the AAA + domain of TaLon contains an insert of approximately 9 0 a mino acid residues between the Walker A and Walker B ATPase signatures [22], which is not present in any bacterial or eukaryotic Lon sequence. The 90-residue insert is found in all archaeal L on homologues and is predicted t o contain two consecutive transmembrane h elices, suggesting t hat archaeal Lon proteases are membrane a ssociated [ 21]. Gram-positive bacteria, such as Bacilli and Clostridia, contain a second Lon protease, called LonB which, like t he archaeal Lon homologues, does not contain an N-terminal a-helical domain ( Fig. 1C) . In contrast with archaeal Lon proteases the b acterial LonB also lacks the 90-residue insert harbouring the predicted t ransmembrane region and is therefore probably a soluble protease, although this h as not been addressed experimentally [11]. Notably, the archaeon Methanosarcina mazei has t wo lon genes, an ar chaeal-type lon gene containing two transmem- brane helices and a bacterial-type lon gene including an N-terminal domain (Fig. 1). The M. maze i archaeal-type lon gene was not included in a phylogenetic analysis [23] although both genes are a lso p resent in the closely related species Methanosarcina acetivorans and M. barkeri. Simi- larly, the M. mazei genome contains both the comp lete group I and group II chaperonin systems, i.e. the bacterial GroEL/GroES and the archaeal thermosome/prefoldin [24]. In general, approximately 31% of the ORFs in the genomes of M. mazei and its close relatives share t he highest similarity with bacterial genes, which is most probably a result of horizontal gene transfer [23]. In addition to the bacterial and archaeal lon genes the genome of M. mazei contains a noncanonical lon gene (ORF Mm1931), enco- ding for a Lon protease lacking the ATPase domain. Noncanonical Lon proteases are also found in bacteria and viruses (Fig. 1B,C). A separate study failed to detect the ATP-binding region in the archaeal Lon proteases in sequence c omparisons with bacterial homologues and classified the Pyrococcus Lon homologues as ATP-inde- pendent proteases [25]. This is easily explained by t he fact Fig. 1. Sequence a lignment o f s elected r egions of distinct Lon proteases. (A) Alignment of the Walker A an d W alker B motifs of the AAA + ATPase-domain. Identical residues are marked with #, conserved residues with +. (B) A lignment of the S-K dyad of th e p rotease do- main. Labelling as described in ( A). Essential residues o f TaLon pro- tease activity are indicated below. (C) Schematic representation of the domain organization of the prote ins aligned in (A) and ( B) . Bs, Bacillus subtilis;Ec,Escherichia coli;IBDVP2,infectiousbursal disease virus strain P2; M m, Methanosarcina mazei;Ta,Thermoplasma ac idophilum; Tk, Th ermo coccus kodakarensis. The UniProt a ccession numbers of the aligned proteins are: Bs-LonA, P37945; Bs-LonB, P42425; Bs-YlbL, O34470; Ec-Lon, P08177; I BDVP2-VP4, Q82628; Mm-1913, Q8PVP9; Mm-bLon, Q 8PSG1; Mm-Lon, Q8Q0K8; Ta-Lon, Q9HJ89; Tk-Lon, Q8NKS6. 4362 H. Besche and P. Zwickl (Eur. J. Biochem. 271) Ó FEBS 2004 that archaeal Lon p roteases contain a 90 -re sidue insert between the W alker A and Walker B motifs as mentioned above, which impairs alignment of the bacterial with t he archaeal ATPase domain. After removing the 90-residue insert the bacterial and archaeal Walker A and W alker B motifs can b e aligned ( Fig. 1A). Sequence alignment o f the proteolytic core domain s hows that the active site S-K dyad is conserved in archaeal L on, in bacterial LonA and LonB, and in noncanonical Lon proteases ( Fig. 1B). In summary, the Lon protease domain i s an e volutionarily ancient domain, wh ich i s ubiquitously distributed and c an be fused to various other domains [6]. Ta Lon forms ring-shaped complexes Overexpression in E. coli leads to m embrane insertion of the TaLon protease. The isolated membranes were solu- bilized with detergent, and r ecombinant TaLon was purified by Ni 2+ -nitrilotriacetic a cid a ffinity chromatogra- phy [13]. Subsequent size exclusion chromatography revealed an apparent molar mass of 510 kDa. The transmembrane domain of TaLon is surrounded by a dodecyl-b- D -maltopyranoside (DDM) micelle with a mass of  70 kDa. Subtraction of the DDM micelle suggests that six TaLon mo no mers ( 72 kDa) ass emb le in to a hexameric c omplex (Fig. 2A). C onsistently, electron micro- scopic analysis revealed r ing-shaped particles of yet unknown symmetry ( Fig. 2B). The isolated p roteolytic domain of EcLon w as crystallised as a hexameric complex [26]. In contrast, a structural analysis of the yeast mitochondrial Lon protease by analytical ultracentrifuga- tion and electron m icroscopy revealed heptameric ring- shaped complexes [27]. The stoichiometry of the TaLon complex remains to be determined. S-K dyad active site Lon from E. coli has been known for a l ong time to be a serine protease [14,20], but only lately was an active site lysine residue 7 identified [16,17]. These two catalytic residues are conserved among all a rchaeal and bacterial Lon homologues and form an S-K d yad. Very recently, the crystal structure of the hexameric E. coli Lon protease domain has been solved and revealed a unique fold not observed in other S-K dyad peptidases [26]. Only the catalytic core containing the active s ite residues is structur- ally conserved between distinct S-K dyad peptidases [26]. To establish that the Therm oplasma me mbrane-bound Lon protease has the same active site, t he two conserved residues were individually muta ted to alanine (S52 5A and K568A). As with TaLonwt, the S -K mutants were purified as hexamers (data not shown) sustaining substantial ATPase activity (Table 1), but no peptidase activity could be detected (Fig. 3 A). Consequently, TaLonS525A and TaLonK568A showed neither ATP-dependent nor ATP- independent proteolytic activity (Fig. 3B). ATPase activity of protease active site mutants Kinetic analysis of TaLon and its protease active site mutants r evealed t hat t he specific ATPase a ctivity o f Fig. 2. Molar mass and electron microscopy of TaLon. (A) TaLo n (30 mg) was s eparated on Supe rdex 200 HiLoad 26/60 column (25 m M Mes pH 6.2, 300 m M NaCl, 5 m M MgCl 2 ,0.5m M DDM) and detected by UV 280 . M olar mass of marker pro teins an d their respective elution volume s are in dicated. G el filtration r evea led an a pparen t molar mass o f 510 kDa corresponding to a hexameric complex a fter subtraction of 70 kDa for the DDM micelle. (B) Electron micrograph of negatively stained (2% uranylacetate) TaLo n from the Superdex 200 peak fraction, recorded with a Philips C M 200 FEG transmission electron microscope at 16 0 kV. Table 1. ATPase ac tivity of TaLon and EcLon. Protease K m (m M ) V (P i Æmin )1 Ælg )1 ) TaLonwt 0.196 ± 0.004 0.631 ± 0.003 TaLonS525A 0.176 ± 0.014 0.561 ± 0.012 TaLonK568A 0.111 ± 0.003 0.477 ± 0.003 EcLonwt 0.201 ± 0.004 0.554 ± 0.012 EcLonS679A 0.140 ± 0.014 0.273 ± 0.009 EcLonK722A 0.189 ± 0.011 0.782 ± 0.015 Ó FEBS 2004 T. acidophilum Lon is a Ser-Lys dyad protease (Eur. J. Biochem. 271) 4363 TaLonK568A was 25% lower and the affinity for ATP was increased by 43% when compared with wild-type TaLon (Table 1). The K m of TaLonS525A remained unaffected and the specific ATPase activity was only slightly reduced (10%; T able 1). Thus the mutation o f the catalytic protease residues, especially of Lys568 seems to enhance ATP binding and concomitantly slow down ATP hydrolysis. In order to generalize this observation the corresponding E. coli lysine mutant was generated (EcLon- K722A) and purified along with the active site serine mutant (EcLonS679A). B oth m utants were proteolytically inactive (data not shown). I n accordance with our model that the protease domain might have a r egulative influence on the protease domain, both mutants were affected in their ATPase activity i n comparison with the wild-type enzyme. The EcLon Ser679 mutant showed only h alf the wild-type ATPase activity but a 30% higher affinity for ATP, while the lysine mutant w as stimulated with resp ect to ATP hydrolysis (40%) and hardly affected in its affinity for ATP (Table 1). Discussion The Lon protease is ubiquitously distributed and was found to exist as a standalone protein as well as fused to a AAA + ATPase-domain (Fig. 1C ) [6]. Whereas the b acterial Lon protease and its homologues i n eukaryal organelles are soluble, the archaeal c ounterpart is membrane-attached [13]. Bacterial LonB proteases, like archaeal Lon, lack the LAN domain p resent in bacterial L onA pro teases, but like bacterial LonA, also miss the membrane anchor foun d in archaeal Lon proteases. Taken together the Lon protease domain is h ighly versatile and can function in different contexts, i.e. standalone or fused t o an ATPase, and s oluble or membrane-bound. Solubilized TaLon has a m olar mass of 430 kDa corresponding to a hexameric complex. T he s ame o ligome- rization state w as revealed r ecently in t he crys tal structure of the hexameric EcLon protease domain [26], whereas electron microscopy of yeast m itochondrial Lon proteaese showed a heptameric complex [27]. Using site-directed mutagenesis we established t he S-K dyad for the membrane-bound Ta Lon and confirmed it f or the soluble EcLon. Subsequently, we characterized the TaLon and EcLon protease-deficient mutants lacking the catalytic serine o r lysine residue. K inetic analysis and comparison of their r espective ATPase activity revealed a potential regulation of ATPase hydrolysis depending on the p roteolytic cycle. Though the unidirectional regulation of the peptidase activity by the ATPase has been described for the EcLon protease several years ago [18], the reciprocal regulatio n of the ATPase b y the protease remained unrecognized, altho ugh two reports described impaired ATP hydrolysis of FtsH m utants lacking p rotease activity. Karata et al . [28] reported that mutation of the zinc-binding residue His421 in Escherichia coli FtsH (EcFtsH) completely abolished protease activity an d reduced the ATPase activity to 23% of wild-type FtsH. In this case, it c an be claimed that the prevention of zinc ion b inding might l ead to structural perturbations that reduce the ATPase activity. However, i n two independent studies mutations of the catalytic aspartate in the conserved HEAGH motif of EcFtsH [29] and Bacillus subtilis FtsH [30] were generated. The conserved aspartate residue activates a w ater molecule for the nucleo philic attack but does not affect zinc ion binding. Again these mutant Fts H proteins were reported to lack proteolytic activity and showed a 20% reduced ATPase activity. Taken together with our observation that mutation of the Ta Lon and EcLon protease active site residues dec reases the ATPase activity, we propose that bidirectional crosstalk between the ATPase a nd peptidase domains is necessary for controlled protein degradation. Fig. 3. Pr oteoly tic activity of TaLonwtandtheactivesitemutants TaLonS525A and TaLonK568A. (A) TaLon (46 n M )andThermo- plasma acidophilum p roline iminopeptidase (1.83 l M )wasincubated with 100 l M succinyl-LLVY-7-amido-4-methylcoumarin in assay buffer ( 50 m M Mes pH 6.2, 20 m M MgCl 2 ,0.5m M DDM) at 60 °C. TaLonwt and the ATPase d eficient mu tant Ta LonK63A (compare [13]) served as control; the wild -type activity was set equal to o ne . (B) TaLon (116 n M )wasincubatedwith5l M fluorescein isothiocyanate- casein i n the assay buffer with or without 2 m M ATP a t 60 °C. For the S-K dyad m utants n o diffe renc e w as obse rved in the presence or absence o f ATP. 4364 H. Besche and P. Zwickl (Eur. J. Biochem. 271) Ó FEBS 2004 Acknowledgements We tha nk Ulf Klein (MPI Martinsried) f or assistance, Oana Mihalache (MPI Martinsried) for e lectron microscopy, Erik Roth (MPI Martins- ried) for calibration of the Sup erdex 200 column and Rudi Glockshuber (ETH Zu ¨ rich) for providing EcLon expression plasmids. We are indebted to Wolfgang Baumeister (MPI Martinsried) for g enerous a nd continuous support. Fin ally, we want to thank the referees for their valuable suggestions. This work was supported by a grant from the DFG t o Peter Z wickl (Z w58/3-2). References 1. Swamy, K.H. & Goldberg, A.L. (1981) E. coli conta ins e i ght soluble proteolytic activities, o ne being ATP dependent. Nature 292, 652–654. 2. Chung, C.H. & Goldberg, A.L. (2004) Endopeptidase La. In Handbook of Proteolytic E nyzmes (Barret,A.J.,Rawlings,N.D.& Woessner, J.F., eds), pp. 1998–2002. Elsevier Academic Press, London, UK. 3. Chung, C.H. & Goldberg, A.L. (1981) The product of the lon (capR)geneinEsch erichia co li is the A TP-dep endent pro tease, protease La. Pro c. Natl Acad. Sci. USA 78, 493 1–4935. 4. Charette, M.F., Henderson, G.W. & Markovitz, A. (1981) ATP hydrolysis-dependent protease activity of the lon (capR) protein of Escherichia coli K-12. Proc. Natl Acad. Sci. USA 78, 4728–4732. 5. Barrett, A.J., R awlings, N.D. & O’Brien, E.A. ( 2001) The MEROPS database as a protease information system. J. Struct. Biol. 134, 95–102. 6. Iyer, L.M., Leipe, D.D., Koonin, E.V. & Aravind, L. (2004) Evolutionary history a nd higher ord er classification of AAA + ATPases. J. Struct. Biol. 146, 1 1–31. 7. Goldberg, A.L. (1992) The mechanism and functions of ATP- dependent proteases i n bacterial and animal cells. Eur. J. Biochem. 203, 9–23. 8. Gottesman, S. (1996) P roteases and t heir targets i n Esche richia coli. Annu. Rev . Genet. 30 , 465–506. 9. Ka ¨ ser, M. & Langer, T. (2000) Protein degradation in mitochondria. Se min. Cell Dev . Biol. 11 , 181–190. 10. Liu, T., L u, B., Lee, I ., Ondrovicova, G., Kutejova, E. & S uzuki, C.K. (2004) DNA and RNA binding b y the m itochondrial lon protease is regulated by nucleotide and protein substrate. J. Biol. Chem. 279, 13902–13910. 11. Serrano,M.,Hovel,S.,Moran,C.P.,Henriques,A.O.&Vo ¨ lker, U. (2001) F orespore-specific transcription o f t he lonB gene during sporulation in Bacillus subtilis. J. Bacteriol. 183, 2995–3003. 12. Fukui, T., Eguchi, T ., Atomi, H. & Imanaka, T. (2002) A mem - brane-bound archaeal Lon p rotease displays AT P-indepen dent proteolytic activity towards unfo lded p roteins an d ATP-depen - dent activity fo r folded p ro teins. J. Bacteriol. 184, 3689–3698. 13. Besche, H., Tam ura, N ., Tam ura, T . & Zwickl, P. (2004) Muta- tional analysis of conserved AAA + residues in the archaeal Lon protease from Thermoplasma acidophilum. FEBS Lett. 574, 161–166. 14. Amerik, A.Y., Antonov, V.K., Gorbalenya, A.E., Kotova, S.A., Rotanova, T.V. & Shimbarevich, E.V. (1991) Site- directed muta- genesis of La protease. A catalytically active serine residue. FEBS Lett. 287, 211–214. 15. Rotanova, T.V. (1999) Structural and functional peculiarities of the ATP-dependent L on protease f rom Escherichia coli. Russ. J. Bioorganic C hem. 25, 883–891. 16. Birghan, C., Mundt, E. & Gorbalenya, A.E. (2000) A non-cano - nical L on proteinase lac king the ATPase domain employs the Ser-Lys c atalytic dyad to exercise broad control ove r the l ife c ycle of a d ouble -stranded RNA virus. EMBO J. 19, 114–123. 17. Rotanova, T.V., Mel’nikov, E.E. & Tsirulnikov, K.B. (20 03) A catalytic Ser-Lys d yad in the active site of the ATP -dependent Lon protease from Escherichia coli. Russ. J. Bioorganic Chem. 29, 85–87. 18. Fischer, H. & Glockshuber, R . (1994) A point mutation w ithin the ATP-binding site inactivates both catal ytic functions of the ATP- dependent protease La (Lon) f rom Escherichia coli. FEBS Lett. 356, 1 01–103. 19. Thompson, J.D., G ibson, T.J., Plewniak, F., J eanm ougin, F. & Higgins, D.G. (1997) The C lustal-X windows interfa ce – flex ible strategies for multiple sequence alignment aided by quality ana- lysis tools. Nucleic Acids Re s. 25, 4 876–4882. 20. Fischer, H. & Glockshuber, R. (1993) ATP hydrolysis is not stoichiometrically linked with proteolysis in the ATP-dependent protease La from Escherichia coli. J. Biol. Chem. 268, 22502– 22507. 21. Ruepp, A., Graml, W., S antos-Martinez , M.L., K oretke, K.K., Volker, C ., Mewes, H.W., Frishman, D., Stocker, S., Lupas, A.N. & Baumeister, W. (2000) The genome sequence of the thermo- acidophilic scavenger Thermoplasma acidophilum. Nature 407, 508–513. 22.Walker,J.E.,Saraste,M.,Runswick,M.J.&Gay,N.J.(1982) Distantly Related Sequences i n the a-andb-Subunits of ATP synthase, myosin, kinases and other ATP-requiring enzymes and a common nucleotide binding f old. EMBO J. 1, 945–951. 23. Deppenmeier, U ., Johann, A., Hartsch, T., Merkl, R., Schmitz, R.A., Martinez-Arias, R., Henne, A., Wiezer, A., Baumer, S., Jacobi,C.,Bruggemann,H.,Lienard,T.,Christmann,A.,Bom- eke, M., Steckel, S., B hattach aryya, A., Ly kidis, A., O verbee k, R., Klenk,H.P.,Gunsalus,R.P.,Fritz,H.J.&Gottschalk,G.(2002) The genome of Methanosarcina mazei: e vidence for lateral g ene transfer between bac teria an d archae a. J. Mol. Microbiol. B io- technol. 4, 4 53–461. 24. Klunker,D.,Haas,B.,Hirtreiter,A.,Figueiredo,L.,Naylor,D.J., Pfeifer, G., Mu ¨ ller, V., D eppe nmeier, U ., G ottschalk, G ., H artl, F.U. & Hayer-Hartl, M. (2003) Coexistence of Group I and Group II chaperonins in th e archaeon Methanosarcina mazei. J. Bi ol. Chem. 278, 33256–33267. 25. Ward, D.E., Shockley, K.R., Ch ang, L.S., Levy, R.D., Michel, J.K., Conners, S. B. & Kelly, R.M. ( 2002) Proteolysis in hyperthermophilic microorganisms. Ar chaea 1, 63–74. 26. Botos, I., Melnikov, E.E., Cherry, S., Tropea, J.E., Khalatova, A.G.,Rasulova,F.,Dauter,Z.,Maurizi,M.R.,Rotanova,T.V., Wlodawer, A. & Gustchina, A. (2004) The catalytic domain of Escherichia coli Lon protease h as a unique f old and a Ser-Lys dyad in the a ctive site. J. Bio l. Chem. 279, 8 140–8148. 27. Stahlberg, H., K utejova, E. , S uda, K., Wolpensinger, B., Lustig, A., Schatz, G., Engel, A. & Suzuki, C.K. (1999) Mitochondrial Lon o f Saccharomyces cerevisiae is a ring-shaped protease with seven fexible subunits. Proc. Natl A cad. Sci. USA 96, 6787–6790. 28. Karata, K ., Inagawa, T., Wilkinson, A.J. , Tatsuta, T. & Ogura , T. (1999) Dissecting the role of a con served mo tif (the secon d region of homology) in the AAA family of ATPases – Site-directed mutagenesis of the ATP-depe ndent prot ease FtsH. J. Bi ol. Chem . 274, 2 6225–26232. 29. Jayasekera, M.M.K., Foltin, S.K., Olson, E.R. & Holler, T.P. (2000) Escherichia coli requires the protease activity of FtsH for growth. Arch. Biochem. Biophys. 380, 1 03–107. 30. Kotschwar,M.,Harfst,E.,Ohanjan,T.&Schumann,W.(2004) Construction and analyses of mutant ftsH alleles of Bacillus subtilis involving the ATPase- a nd Zn-binding domains. Curr. Microbiol. 49, 1 80–185. Ó FEBS 2004 T. acidophilum Lon is a Ser-Lys dyad protease (Eur. J. Biochem. 271) 4365 . wild-type Lon the protease domain might influ- ence the activity of the ATPase domain. Keywords:AAA + protease; archaea; Lon( La) endopepti- dase; Lon (La) protease; Ser-Lys dyad. Endopeptidase La (EC. [21] (Fig. 1 A, B). TaLon encompasses an N-terminal ATPase associated with various cellular a ctivities ( AAA + domain) and a C-terminal protease domain, but lacks the N-terminal a- helical domain inherent. acterial Lon protease and its homologues i n eukaryal organelles are soluble, the archaeal c ounterpart is membrane-attached [13]. Bacterial LonB proteases, like archaeal Lon, lack the LAN domain