Scientific paper Thermodynamic Stability of the Dimeric Toxin CcdB Mario [imi}, Gorazd Vesnaver and Jurij Lah* University of Ljubljana, Faculty of Chemistry and Chemical Technology, A{ker~eva 5, 1000 Ljubljana, Slovenia. * Corresponding author: E-mail: jurij.lah@fkkt.uni-lj.si Phone: +386 1 2419 414; fax: +386 1 2419 425 Received: 29-10-2008 Dedicated to Professor Josef Barthel on the occasion of his 80'' birthday Abstract The toxin-antitoxin module ccd located on Escherichia coli plasmid F encodes the antitoxin CcdA and the toxin CcdB. When not complexed with CcdA, CcdB attacks its cellular target gyrase and kills the cell by causing inhibition of both transcription and replication. At physiological conditions CcdB exists as a homodimer. Here we present a study of CcdB unfolding that is focused on the characterization of the structure-thermodynamics relationship needed for understanding the stability and function of CcdB at the molecular level. In this light, thermodynamic parameters of unfolding obtained by global analysis of urea-induced unfolding curves measured at various temperatures by circular dichroism spectros-copy were parsed into the contributions arising from the differences in intra- and inter-molecular interactions of CcdB in the folded dimeric and unfolded monomeric state. According to this parsing the unfolded monomers retain about 30% of the residual structure indicating that the urea-denatured state of CcdB is not a completely unfolded state. Keywords: CcdB, toxin-antitoxin module, stability, structure, circular dichroism, programmed cell death 1. Introduction The ccd operon located on Escherichia coli plasmid F is the toxin-antitoxin module1 encoding the antitoxin CcdA and the toxin CcdB. CcdA inhibits the toxic activity of CcdB by forming a non-covalent CcdA:CcdB complex. CcdA and CcdB are co-expressed in plasmid F-bea-ring cells and their expression is autoregulated at the level of transcription by binding of CcdA:CcdB complex to the promoter DNA.23 Upon plasmid loss, CcdA is quickly degraded by Lon protease, releasing CcdB that by attacking the gyrase kills the cell.1-4 Crystal structure of CcdB5 suggests that at physiological conditions it exists as a homodimer. Recent stability studies on CcdB have indicated that in buffer solution at physiological pH its thermal denaturation is irreversible if no additives that prevent its aggregation are present.6 7 On the other hand, it has been found recently that the chemical unfolding at relatively low temperatures is a reversible process accompanied by dissociation of the dimer.6 7 Although the reported analysis of the measured chemical un- folding curves resulted in estimates of standard thermody-namic parameters of CcdB unfolding, these studies provided almost no information on the structure-thermodynamics relationship needed for understanding the stability and function of CcdB at the molecular level. In this work we performed a global thermodynamic analysis of CcdB urea-induced unfolding curves measured by circular dichroism spectroscopy at various temperatures. An attempt was made to dissect the obtained ther-modynamic stability into enthalpic and entropic contributions that can be further discussed in terms of structural features of CcdB native and denatured states. 2. Experimental Materials. The preparation and purification of CcdB is described elsewhere.5 Its structure and the corresponding amino-acid sequence is given in Figure 1. Since CcdB is a homodimer in solution at non-denaturing conditions its molar concentration is expressed in moldjmer L-1. Solu- tions of purified CcdB were dialyzed extensively against phosphate buffer (0.02 M Na-phosphate, 0.15 M NaCl, 0.001 M EDTA, pH = 7.5). All samples for the urea dena-turation experiments were prepared by mixing 10 M urea and protein stock solution to a final urea concentration between 0 and 8 M. The pH of all solutions was checked and adjusted to 7.5 by addition of NaOH. Concentration of CcdB was determined spectrophotometrically from the absorbance measured at 280 nm in 6 M GdmHCl at 25 °C using extinction coefficients obtained using the method introduced by Gill and von Hippel8 (http://www.expasy. ch). Circular dichroism spectroscopy (CD). CD measurements were performed with an AVIV Model 62A DS spectropolarimeter (Aviv Associates, USA) at various temperatures lower than the one at which CcdB denatures irreversibly. Changes in secondary structure at increasing urea concentrations (0-8 M) were followed by measuring the ellipticity at 225 nm in a 1 cm cuvette at the protein concentration of about 1 pM. 3. Global Thermodynamic Analysis of Urea Unfolding Curves The urea induced unfolding of CcdB from the dime-ric native state (N2) to the denatured monomeric state (D) may be described as a reversible two-state transition (1) The apparent equilibrium constant Kj,u is a function of temperature (T) and urea concentration (u) and can be defined as:9 (2) where is the fraction of CcdB in the native state at given T and u, a^^ = [N2]/c) where c is the total molar Ccd-B dimer concentration and [N2] and [D] are molarities of N2 and D, respectively. According to the model c = [N2] + [D]/2 and the measured ellipticity ) at a given wavelength A, T and u can be expressed in terms of the corresponding contributions 6N and 9DAT.u that characterize pure states N2 and D as: (3) Since 0N A and 6DAcan be estimated at any measured T as linear functions of u (pre- and post-transitional baselines; Figure 2b) the measured a,.^ can be expressed as (Figure 3) (4) On the other hand, aT,u can be connected to the thermodynamics of unfolding through the two-state transition model (equation 1) according to which the linear dependence of the standard Gibbs free energy of unfolding (AG^u) on u can be at any T expressed as: (5) where m is an empirical parameter correlated strongly to the amount of protein surface area exposed to the solvent upon denaturation10 and assumed to be temperature independent. AG^ is the standard Gibbs free energy of unfolding in the absence of urea (u = 0) that may be expressed in terms of the corresponding standard Gibbs free energy (AG^ ) and standard enthalpy of unfolding (Afl^ ) at a reference" temperature To = 25 °C and standard heat capacity of unfolding (AC°) (assumed to be temperature independent) through the Gibbs-Helmholtz relation (integrated form): —(6) It follows from equations 5 and 6 that the model (adjustable) parameters AG^, AH^, AC° and m define, AG^uand also the corresporiding Ktj.u = exp (-AG^u/ RT)). Consequently, the model function for aj,^ derived from equation (7) can be compared to aT,u values determined experimentally from equation 4. The values of adjustable parameters (Table 1) were obtained using the non-linear Levenberg -Marquardt regression procedure.11 Then, the best global fit values of AG^ , AH^ and ACp were used to estimate AGTo (from equation 6), from the Kirchhoff's law (8) and the corresponding entropy contribution, TASTo, from the general relation (9) The temperature dependences of AGTo, AHTo and TASTo are presented in Figure 4. 4. Correlation Between Thermodynamic and Structural Parameters Numerous recent studies on protein stability have shown that for protein unfolding both enthalpy of unfol- ding (AHT) and heat capacity of unfolding (ACp) can be parameterized in terms of changes in solvent accessible polar (AAP) and non-polar (AAN) surface areas associated with the unfolding of the protein.1012-15 Such a parameterization is based on the estimation of the non-polar (AN) and polar (AP) solvent accessible areas of proteins in the folded and unfolded states that were in the case of CcdB calculated with the program NACCESS version 2.1 using the probe size of 1.4 À.16 AN and AP of native (folded) CcdB dimer were obtained from the known crystal struc-ture3, while the AN and AP values of the completely unfolded CcdB monomers were estimated as the sum of the accessibilities of the protein residues located in the Ala-X-Ala tripeptides.1213 The heat capacity (ACp) and enthalpy of unfolding (AH^ = AH^^ + AC°p(T - T^ )) can be expressed as the sum of non-polar (subscript N) and polar (subscript P) contributions12-15 ic;, = + = aMn + bAAp and (10) (11) Parameters a = 1.88 J mol1 K1 À2, b = - 1.09 J mol-1 K-1 À-2, c = -35.3 J mol1 À-2and d = 131.4 J mol1 À-2 are obtained from Murphy and Freire12 and Xie & Freire,17 while AH^^ is parameterized as AH^^ = cAAN + dAAP17 and represents the enthalpy of unfolding observed with most global proteins at their median transition temperature of TH = 60 °C. The corresponding entropy of unfolding (ASo) can be expressed as12-15,17,18 (12) The solvation contribution (ASTo,solv) that describes the exposure of polar and non-polar groups to the solvent upon unfolding and dissociation of the CcdB dimer may be estimated as ASo,sojv = AC^ ln(T/TS),12,19 where TS » 112 °C is the estimated reference temperature at which AST soiv is assumed to be equal to zero. The second term (AS.oojher) can be considered as the sum of changes in configuratio-nal entropy (AS.oconf) and translational and rotational entropy (ASo,r+j) that accompany CcdB dimer unfolding and dissociation. ASTo,other was estimated by subtracting the calculated ASo,solv = ACp ln(T/TS) from the corresponding measured AST (equation 12; Figure 5). Moreover, AS.or+j can be estimated as an entropy change accompanying (rigid-body) dissociation of the dimer (AS.or+t = 209 J K1 mol-1)14 while ASo,conf can be estimated as AS:o,conf = {N) 18 J K-1 (mol residue)-1 where {N) is the average number of amino-acid residues participating in the unfolding process and 18 J K-1 (mol residue)-1 is the average overall configu- ration entropy change obtained from the thermodynamic database for unfolding of monomeric proteins.12 As shown recently,20-24 the thermodynamics of a protein unfolding and some association processes can be correlated with its structural features through AAN and AAP values calculated from equations 10 and 11 using the experimentally obtained values for ACP and AH^. This enables the dissection of ACP and AH^ into contributions due to interactions of non-polar (A C^^ N, AH^ N) and polar (AC^ P, AH^ P) surfaces (see equations 10 and 11). Furthermore, the average number of the unfolded residues, {N), can be estimated as {N) = {AAn + AAp) ■o E m CM (M ® -16 ii ® i / AVx • V. ÌÌ6 • 20 40 60 80 T/X b) TJ E E C m CM (N ® Figure 2. Thermal and urea induced denaturation of CcdB monitored by spectropolarimetry. (a) Thermal denaturation profile and the corresponding spectra (inset) measured at 25 °C before (full line) and after heating to 90 °C (dotted line); (b) Urea denaturation profile measured at 5 °C. Pre- and post-transitional baselines (dotted lines) define ellipticities of the native and denatured state over the whole range of urea concentrations, u (see equation 4). Inset represents spectra measured at u = 3.5 M (full line), 7.0 M (disconnected line) and at 3.5 M (dotted line; after dilution from 7.0 M and corresponding multiplication of spectra by the dilution factor) indicating high degree of reversibility of urea denaturation. Global fitting of the reversible two-state model function (equation 7) shows good agreement with experimental aj.u versus u curves measured at various temperatures (Figure 3). The resulting best-fit thermodynamic parameters AG^ , AHT , ACpo are in relatively good agreement with the correospondoing literature values (Table 1). 0.8 0.4 0.0 □ • É s " 1 'S T/X 5 15 25 35 40 ^^n 1 ' - * 1 ■ ■ • ■ 0 2 4 .6 8 u I mol L"^ Figure 3. Global model analysis of urea induced denaturation profiles. The measured fraction of native CcdB dimer (points; see equation 4) and the corresponding best global fit of the two-state model function (lines; see equation 7). The best fit parameters are presented in Table 1. The thermodynamic profiles of CcdB unfolding estimated from the best-fit parameters (Table 1) are presented in Figure 4. The observation that CcdB folding in the standard state at physiological temperatures is an enthalpy driven process, accompanied by a negative entropy contribution and negative heat capacity change is a general feature of globular proteins.13 More interesting is the dissection of the measured thermodynamic parameters of unfolding at 25 °C to various contributions based on equations 10 - 12 (Figure 5). Parsing of enthalpic contribution indicates that the contribution due to the changed interactions of polar surfaces that favor folding overcompensates the corresponding contributions of non-polar surfaces that on average favor unfolding. Furthermore, parsing of entropic contributions indicates that the contribution due to the changes of configurational, translational and rotational freedom (AS.oojher) that favors unfolding slightly prevails over the solvation contribution (ASTsojv) which favors CcdB folding. , As shown recently,20,21 the correlation of the experimentally determined thermodynamics of protein unfolding with its structural features can be carried out using an approach in which AAN AAP and {N} that accompany unfolding and dissociation of CcdB dimer are calculated from combination of experimental thermodynamic data and parametrized equations 10 and 11. These calculations show that upon urea denaturation of CcdB both AAn/AAn,st and AAP/AAP,ST ratios are about 70%. Similarly, the {N)/{N)st in which {N} is determined either from a) b) Figure 4. Thermodynamic profile of CcdB unfolding. (a) Standard Gibbs free energy, AG^^; (b) enthalpy, AH^ (full line), and entropy contribution, taso (dotted line) in the absence of urea (u = 0) as functions of temperature, T, were estimated from the best-fit parameters (Table 1) by using equations 6, 8 and 9. dered only as reasonably good approximations since they comprise errors of the empirical parameterization and those of the measured thermodynamic quantities. Nevertheless, we believe that using this approach one can explain, at least in a semi-quantitative way, the correlation between the thermodynamics of urea-induced unfolding of its and the structural features of CcdB folded and unfolded state. Table 1. Comparison of the thermodynamic parameters of CcdB unfolding at To = 25 °C obtained from global fitting of the model function (equation 7)a to the chemical unfolding data (Figure 3) with the corresponding parameters obtained from other thermody-namic studies6,7 Figure 5. Thermodynamic profile of CcdB unfolding at T = 25 °C. The standard thermodynamic quantities AGTo (black), AHTo (grey), T asso (black) are presented in kJ moldjmer-1, while for clarity reasons ACO (grey) is presented in 10 kJ moldjm,r-1 K-1. Contributions to AHO due to the changed interactions of polar (hatched horizontally) and non-polar surfaces (hatched vertically). T ASTo contributions due to the differences in solvation of the folded and unfolded state (hatched vertically) and other contributions that contain changes of conformational, translational and rotational freedom upon unfolding (hatched horizontally). Contributions to ACPo due to the exposure of polar (hatched horizontally) and non-polar (hatched vertically) surfaces. For clarity, the estimated AHTo and TASTo contributions (see equations 10-12) are divided by a factor of four. equation 13 ({N) = 140) or 14 «N) = 150) is about 70%. Evidently, these AAN/AAN,ST, AAp/AAp,ST and {N)/{N)ST values show that the degree of urea induced unfolding is significantly lower then unity. The resulting conclusion is that the urea-denatured state is not a completely unfolded state which is in accordance with our far-UV CD results (Figure 2). Moreover, reasonable agreement between the {N) values determined from two different relations (equations 13 and 14) suggests that the presented enthalpy and entropy contributions have a real physical meaning. We are well aware that these contributions determined from the described combination of experimental thermodynamics and structure-based parameterization can be consi- denaturant AG0/ kJ mtil1 AH^/ kJ m(ol-1 ACP/ kJ mol1 K1 urea (this work) 89 105 11 GdmHCl (reference 6) 87 166 13 GdmHCl (reference 7) 85 31 12 aThe parameter errors are estimated to be about ± 5% for AG^ , ± 15% for AHOo and ± 10% for AC^. O Our model analysis resulted in the best-fit value of parameter m (equation 5) of 12.6 (±1.3) kJ mol-1 M-1 6. Acknowledgment We thank prof. R. Loris for providing the purified toxin CcdB used in this work that was supported by the Ministry of Higher Education, Science and Technology and by the Agency for Research of Republic of Slovenia through the Grants No. P1-0201 and J1-6653. 7. References 1. L. Buts, J. Lah, M.-H. Dao-Thi, L.Wyns, R. Loris, Trends Biochem. Sci. 2005, 30, 672-679. 2. P. Bernard, M. Couturier, Mol. Biol. 1992, 226, 735-745. 3. T. Miki, J. A. Park, K. Nagao, N. Murayama, T. Horiuchi, J. Mol. Biol. 1992, 225, 39-52. 4. M.-H. Dao-Thi, L.Van Melderen, E.De Genst, H. Afif, L. Wyns, R. Loris, J. Mol. Biol. 2005, 348, 1091-1102. 5. R. Loris, M.-H. Dao-Thi, E. M. Bahassi, L. Van Melderen, F. Poortmans, R. Liddington, M. Couturier, L. Wyns, J. Mol. Biol. 1999, 285, 1667-1677. 6. M.-H. Dao-Thi, J. Messens, L. Wyns, J. Backmann, J. Mol. Biol. 2000, 299, 1373-1386. 7. K. Bajaj, G. Chakshusmathi, K. Bachhawat-Sikder, A, Surolia, R. Varadarajan, Biochem J. 2004, 380, 409-417. 8. S. C. Gill, P. H. von Hippel, Anal. Biochem. 1989, 182, 319-326. 9. I. Drobnak, M.Serucnik, J. Lah and G. Vesnaver, Acta Chim. Slov. 2007, 54,445-451. 10. J. K. Myers, N. C. Pace, M. J. Scholtz, Protein Sci. 1995, 4, 2138-2148. 11. W. H. Press, B. P. Flannery, S. A. Teukolsky, W. T. Vetterling, Numerical Recipes 1992 Cambridge University Press, Oxford. 12. K. P. Murphy and E. Freire, Adv. Protein Chem. 1992, 43, 313-361. 13. G. I. Makhatadze, P. L. Privalov, Adv. Protein Chem. 1995, 47, 307-425. 14. R. S. Spolar, M. T. Jr. Record, Science 1994, 263, 777-784. 15. A. D. Robertson, K. P. Murphy, Chemical Reviews 1997, 97, 1251-1267. 16. S. J. Hubbard, and J. M. Thornton, NACCESS computer program 1993, University College London. 17. D. Xie, E. Freire, Proteins Struct. Funct. Genet. 1994, 19, 291-301. 18. K. H. Lee, D. Xie, E. Freire, L. M. Amzel, Proteins Struct. Funct. Genet. 1994, 20, 68-84. 19. R. L. Baldwin, Proc. Natl. Acad. Sci. USA 1986, 83, 8069-8072. 20. J. Lah, M. Simic, G. Vesnaver, I. Marianovsky, G. Glaser, H. Engelberg-Kulka and R. Loris, J. Biol. Chem. 2005, 280, 17397-17407. 21. J. Lah, I. Prislan, B. Kržan, M. Salobir, A. Francky and G. Vesnaver, Biochemistry 2005, 44, 13883-13892. 22. J. Lah, I. Marianovsky, G. Glaser, H. Engelberg-Kulka, J. Kinne, L. Wyns and R. Loris, J. Biol. Chem. 2003, 278, 14101-14111. 23. J. Lah, M. Bešter-Rogač, T.-M. Perger, G. Vesnaver, G. J. Phys. Chem. B 2006, 110, 23279-26291. 24. J. Lah, I. Drobnak, M. Dolinar, G. Vesnaver, Nucleic Acids Res. 2008, 36, 897-904. 25. E. F. Pettersen, T. D. Goddard, C. C. Huang, G. S. Couch, D. M. Greenblatt, E. C. Meng, and T. E. Ferrin, J. Comput. Chem. 2004, 25, 1605-1612. Povzetek Operon ccd iz E. coli kodira antitoksin CcdA in toksin CcdB ter predstavlja t.i. modul toksin-antitoksin. Če CcdA in CcdB nista medsebojno povezana v kompleks, prosti CcdB napade celično tarčo, DNA girazo, in z inhibicijo prepisovanja in podvajanja DNA povzroči smrt bakterijske celice. Pri fizioloških pogojih CcdB obstaja kot homodimer. Za razliko od nedavno raziskanega razvitja CcdB, se v naši študiji osredotočamo na povezavo med strukturo in termodinamiko denaturacije, ki je potrebna za razumevanje stabilnosti in delovanja CcdB na molekularnem nivoju. V tej luči so bili ter-modinamski parametri razvitja dobljeni s pomočjo globalne analize denaturacijskih krivulj merjenih pri različnih temperaturah s spektropolarimetrijo. Razčlenjeni so bili na prispevke, ki izhajajo iz razlik v intra- in inter-molekularnih interakcijah, ki jih CcdB lahko tvori v zvitem dimernem in razvitem monomernem stanju. Ta razčlemba pokaže, da razviti monomeri ohranijo približno 30 % strukture, kar pomeni, da opaženo denaturirano stanje doseženo z visoko koncentracijo uree ne ustreza popolnoma razvitemu stanju proteina.