Biochimica et Biophysica Acta 1794 (2009) 905–912
Contents lists available at ScienceDirect
Biochimica et Biophysica Acta j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / b b a p a p
Attenuation of ionic interactions profoundly lowers the kinetic thermal stability of Pyrococcus furiosus triosephosphate isomerase Sanjeev Kumar Chandrayan, Purnananda Guptasarma ⁎ Division of Protein Science and Engineering, Institute of Microbial Technology, Chandigarh 160 036, Council of Scientific and Industrial Research, New Delhi, India
a r t i c l e
i n f o
Article history: Received 12 September 2008 Received in revised form 8 February 2009 Accepted 5 March 2009 Available online 21 March 2009 Keywords: Protein thermal stability Protein unfolding Kinetic stability Cold denaturation Salt bridge and ionic interaction
a b s t r a c t We investigate here the high structural stability of Pyrococcus furiosus triosephosphate isomerase (PfuTIM) by exploring the effects – upon the protein's structure and kinetic thermal stability – of modulation of its ionic interactions through pH variations, and mutations. PfuTIM shows comparable structural contents at pH 3.0, 7.0 and 10.0. However, at pH 3.0, subtle changes are seen in the protein's surface hydrophobicity and association status, and its kinetic thermal stability is profoundly reduced (as evidenced by its facile heat- and cold-mediated denaturation, characterized by a high degree of hysteresis and irreversibility). Increase in ionic strength through addition of salt counters the reduction of stability, and reversal of pH facilitates partial refolding. Further, a mutated form of PfuTIM (mPfuTIM) lacking 4 key charged residues involved in ionic interactions displays a structural content identical to PfuTIM but profound reduction in kinetic stability to thermal and chemical denaturation, as well as evidence of partial unfolding at temperatures between 90 °C and 100 °C, unlike PfuTIM. We conclude, therefore, that ionic interactions (which are known to determine protein thermodynamic stability) can also contribute significantly to protein kinetic thermal stability. © 2009 Elsevier B.V. All rights reserved.
1. Introduction The inordinately high thermal stability of proteins sourced from organisms living in extreme environments remains intriguing, as well as incompletely understood [1–3]. Proteins from hyperthermophiles are commonly observed to be highly kinetically stable, i.e., kinetic barriers separate the folded and unfolded states of the protein, causing rates of unfolding to be extremely slow [4–6], and unfolding to be irreversible [7]. Our group has been examining recombinant and engineered forms of mesophile, and hyperthermophile, triosephosphate isomerase (TIM), a tetrameric protein with an eight-fold beta/alpha barrel subunit structure. We have proposed that the high kinetic thermal (conformational) stability of Pyrococcus furiosus TIM (PfuTIM), discernible in its inordinately slow unfolding, owes to the high structural autonomy of its beta/alpha unit ‘substructures’, or supersecondary structural elements [8,9]. Briefly, our view is that the cooperativity characterizing the global unfolding transition of any protein reduces in direct proportion with the increase in the structural autonomy of its substructures which, in turn, rise in direct proportion with the ‘thermophilicity’ of the organism from which the protein is sourced [10,11]. Any reduction in cooperativity that owes to increased substructural autonomy thus translates directly into longer times taken for both (a) unfolding (explaining why hyperthermophile
proteins show higher kinetic stability), and (b) refolding (explaining the apparent irreversibility of unfolding of these proteins). One mechanism by which substructures – such as individual helices, helix-strand assemblies, or larger supersecondary structural elements like beta-alpha units – could achieve increased structural autonomy is through additional surface ionic interactions occurring within one, or more, such substructures within a protein. We have earlier demonstrated that the highly stable structural ‘fortress’ of PfuTIM can be breached by denaturant-based destruction of ionic interactions, to such an extent that initial denaturation effected by heating leads to subsequent cold denaturation upon cooling [10,11], with no discernible kinetic stability left in the parts of PfuTIM that have undergone such cold denaturation. We have also shown that such parts show heat-renaturation; however, with such heatrenaturation failing to restore PfuTIM completely to native structure, indicating the presence of kinetic barriers protecting the native state. In the present communication, we establish that a profound reduction in PfuTIM's stability can be also achieved by other means of destroying surface electrostatic interactions, namely: (a) changes in pH, and (b) mutation-based destruction of 4 key ionic interactions within a single ‘substructure’. 2. Materials and methods 2.1. PfuTIM samples, mutagenesis of PfuTIM and mPfuTIM samples
⁎ Corresponding author. Tel.: +91 172 2636680x3301; fax: +91 172 2690585. E-mail address:
[email protected] (P. Guptasarma). URL: http://www.geocities.com/guptasarma/ (P. Guptasarma). 1570-9639/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.bbapap.2009.03.005
Recombinant PfuTIM containing an N-terminal 6xHis affinity tag was expressed in Escherichia coli M15 and purified, as described
906
S.K. Chandrayan, P. Guptasarma / Biochimica et Biophysica Acta 1794 (2009) 905–912
earlier [10,11]. The purified protein containing the 6xHis affinity tag was used for all experiments. To perform site-directed mutations destroying 4 ionic interactions in helix 4 of the 4th beta/alpha unit of PfuTIM, a previously described vector containing the gene encoding PfuTIM [10], was used as a template to introduce the following mutations: Arg111Thr (R111T), Arg112Ala (R112A), Glu114Lys (E114K) and Gly115Gly (E115G), using the Quick Change site-directed mutagenesis kit (Stratagene). The details of the mutagenesis are in the supplementary file. All experiments with PfuTIM used identical protein concentrations of 0.28 mg/ml, except for experiments comparing unfolding of PfuTIM and mPfuTIM, for which a PfuTIM concentration of 0.1 mg/ml was used. All experiments with mPfuTIM used protein concentrations of 0.1 mg/ml. PfuTIM incorporating the 6xHis affinity tag has an isoelectric point (pI) of 6.9. For the pH experiments involving three different values of pH, the middle pH value (pH 7.0) was chosen to be close to the protein's pI (pH 6.9), since it is well known that proteins are most thermodynamically stable at their pI. It may be noted that the nature of the 4 mutations in the mutated PfuTIM (called mPfuTIM) was such that two positively charged and two negatively charged residues were deleted, resulting in no change in the pI. Given the possibility of some unfolding of mPfuTIM due to its destabilization through mutations, and the greater likelihood of intermolecular associations between any partiallyunfolded molecules created upon heating, we decided to perform comparisons of PfuTIM and mPfuTIM at a pH slightly away from their pI of 6.9, to reduce chances of intermolecular associations involving mPfuTIM (in the event of its proving to be less conformationally stable than PfuTIM). Thus, we performed the comparisons of the two
proteins at pH 8.0. All results shown are of representative samples. Further, we also created several single-site, or single point, mutants individually incorporating the mutations referred to above. These are described in the supplementary file. 2.2. Chemicals and buffers Guanidium hydrochloride (Gdm.HCl) and other chemicals were from GE Healthcare (USB Chemicals), USA; ANS (8-anilino, 1naphthalenesulfonic acid) from Sigma Chemical Co., USA. Buffers in the pH range 3.0–6.0 used citrate; in the range 9.0–10.0 used carbonate; and in the pH range 7.0–8.0 used Tris or phosphate. 2.3. Circular dichroism Far- and near-UV CD spectra were acquired with solutions of identical protein concentration on a JASCO J-810 spectropolarimeter, using 1 mm and 4 mm cuvettes, and a spectral bandpass of 4 nm. 2.4. Temperature/time scans and thermal melting Raw ellipticity (at 222 nm) or whole CD spectral data (between 250 and 190/195 nm) were acquired as a function of temperature, or time, using fixed temperatures (25 °C, 95 °C or 98 °C) for time scans, and constant heating rates of 3 °C/min and 1 °C/min for temperature scans, using Peltier heating and 1.0 mm and 2.0 mm cuvettes heated by a 9 mm metal spacer blocks.
Fig. 1. (Panels A–C) Lanes 1–8, respectively, in the 12% SDS-PAGE electrophoretograms in panels A–C, correspond to identical PfuTIM (25 kDa) samples incubated for 1 h each in buffers of pH 10, 9, 8, 7, 6, 5, 4, and 3, prior to electrophoresis. Samples were mixed with SDS-PAGE loading buffer and electrophoresed without boiling, as described previously [10], to examine differences in SDS-mediated unfolding at room temperature (resulting in transformation from ∼80 kDa to ∼20 kDa). Lane 9 shows pre-stained molecular weight markers of 20, 26, 34, 47, and 86 kDa (from bottom to top). Panel A shows effects of incubation of PfuTIM in absence of salt. In Panel B, solutions contained 300 mM KCl. In Panel C, solutions contained 1 M Gdm.HCl. (Panels D–F) Each panel describes a different regime of treatment of PfuTIM, in respect of alterations of pH or temperature (results in subsequent figures). The starting point of each regime of treatment is marked by the number ‘1’, with successive numbers outlining the further stages of treatment.
S.K. Chandrayan, P. Guptasarma / Biochimica et Biophysica Acta 1794 (2009) 905–912
907
2.5. Fluorescence spectroscopy Emission spectra of protein (tryptophan), or 8-anilino, 1-naphthalene sulfonic acid (ANS), were collected using an FMO-427 monochromator attached to a JASCO J-810 spectropolarimeter, with excitation at 280 nm (for tryptophan), or 355 nm (for ANS), using bandwidths of 4 or 8 nm. 2.6. Dynamic light scattering DLS data collection was done on a Wyatt, Protein Solutions DynaPro 800 instrument after filtration of the protein sample through a 0.1 μm filter, using a protein concentration of approximately 0.27 mg/ml, and using a single-angle scatter of 824 nm laser radiation. Approximately twenty 10 second averages of scattering data were used for the analyses. 2.7. Chromatographic and electrophoretic studies Gel-filtration chromatography was performed on a pre-equilibrated analytical SMART Superdex-200 column (Pharmacia), to compare hydrodynamic volumes of PfuTIM at different pH values. SDS-PAGE electrophoretic assays exploring PfuTIM's pH-dependence of stability to SDS-induced unfolding at room temperature were carried out as described previously [10]. It may be noted that in these assays, the low mobility form at ∼80 kDa shows more than one band (usually two bands), which have been proposed to result from conformational heterogeneity. 3. Results 3.1. PfuTIM is susceptible to destabilization by SDS at pH 3.0 We have previously described an SDS-PAGE based assay for the structural destabilization of PfuTIM [10], examining the protein's transformation from a low-mobility form (∼80 kDa), to a normalmobility form (∼ 20 kDa), upon mixing with SDS-PAGE loading buffer at room temperature, without boiling of samples. This assay was used to examine the response of PfuTIM to pH. Fig. 1A shows that PfuTIM's mobility is not altered to ∼ 20 kDa upon lowering of pH from 10.0 to 8.0, but that at pH 7.0 a minor ∼ 20 kDa population is observed which increases in abundance as pH is lowered to 6.0, and 5.0, with no further change occurring upon lowering of pH to 4.0, and 3. Fig. 1B and C show that SDS-unfolding at low pH is partially opposed by the increase of ionic strength through addition of 300 mM KCl, or 1 M Gdm.HCl (which acts as an electrolyte, and not as a denaturant of PfuTIM, at this concentration [10]). PfuTIM's stability is thus clearly established to be affected by modulation of ionic interactions. This was explored further by comparing PfuTIM's conformational characteristics as a function of the schemes of heating, cooling and alteration of pH, as described in Fig. 1D–F. 3.2. PfuTIM has similar secondary structure but different surface features at pH 3.0 Supplementary Fig. 1A shows that far-UV CD spectra of PfuTIM at pH 3.0, 7.0 and 10.0, are identical, indicating identical secondary structural contents. Near-UV CD spectra (Supplementary Fig. 1B) suggest, however, that there are subtle changes in the disposition of PfuTIM's single surface tryptophan. Supplementary Fig. 1C shows that fluorescence emissions from PfuTIM's tryptophan at pH 10.0 and 7.0 are nearly identical, but that at pH 3.0 a marked reduction in emission intensity occurs together with blue shifting of the wavelength of maximal emission (emλmax). Therefore, while secondary structure does not change, there are subtle changes in the environment of PfuTIM's lone surface tryptophan, together with changes in surface
Fig. 2. Changes in PfuTIM's quaternary structure with pH. Panels A–C show dynamic light scattering (DLS) data indicating PfuTIM's hydrodynamic characteristics at different pH (marked alongside). (Panel D) Gel-filtration chromatography of PfuTIM at different pH (marked alongside).
908
S.K. Chandrayan, P. Guptasarma / Biochimica et Biophysica Acta 1794 (2009) 905–912
Table 1 Dynamic light scattering (DLS) based estimates of the hydrodynamic radius, percentage polydispersity of size, and molecular weight, of PfuTIM in buffers of different pH, showing transformation of the ∼25 kDa PfuTIM polypeptide from a tetramer at pH 7.0 or 10.0, into a 20-mer at pH 3.0. pH
Peak
Radius (nm)
% Polydispersity
Molecular weight (kDa)
% Mass
3.0 7.0
1 1 2 1 2
8.5 4.1 16.8 4.2 21.7
18.5 17.7 20.1 10.5 0
497 91 2468 96 4508
100.0 99.4 0.6 99.8 0.2
10.0
shown) which also indicates the occurrence of some partial refolding. However, changes in fluorescence emission spectra at pH 3.0, or 10.0, are not significantly reversed (Fig. 4C) by restoration of pH to 7.0. The single tryptophan in helix 8 (within the 8th and last beta/alpha unit) of PfuTUM may be expected to dominate both its near-UV CD spectrum and its fluorescence emission spectrum. The partial reversal of the near-UV CD characteristics without corresponding reversal of fluorescence emission characteristics is thus intriguing; however, at this point, we would not wish to speculate upon the reasons for this. All experiments described in the above sections were also performed in 300 mM KCl (Supplementary Figs. 3–7), which noticeably mitigated the observed conformational effects seen in the absence of salt.
hydrophobicity, since the dye, ANS, also binds to PfuTIM's surface at pH 3.0 (Supplementary Fig. 1D), but not at other pH values. 3.3. Tetrameric PfuTIM turns into a 20-mer at pH 3.0 PfuTIM is known to be a tetramer at neutral pH [12]. Fig. 2A shows that PfuTIM's hydrodynamic volume is increased at pH 3.0, relative to that at pH 7.0 (Fig. 2B) or pH 10.0 (Fig. 2C). This corresponds to an association of five tetramers into a 20-mer particle of ∼500 kDa and ∼ 8.5 nm radius (Table 1), owing presumably to hydrophobic associations. The increased hydrophobicity at pH 3 causes the 20mer to associate with the matrix of the Superdex-200 column during gel filtration, causing it to elute with a much-delayed elution volume (Fig. 2D), relative to the wild-type tetrameric PfuTIM. 3.4. PfuTIM shows increased susceptibility to thermal unfolding at pH 3.0 At pH 7.0, PfuTIM does not unfold thermally between 25 °C and 95 °C, as temperature is increased at a rate of 3 °C/min (Fig. 3A). At pH 10.0, limited unfolding is seen. At pH 3.0, profound unfolding (nearly 50% loss of far-UV CD signal) is seen. When PfuTIM is placed directly at 95 °C and monitored (Fig. 3B), again there is a profound loss of signal at pH 3.0 over a period of ∼150 s, while no corresponding loss is seen either at pH 7.0, or at pH 10.0. This suggests that the kinetic stability of PfuTIM is dramatically lowered upon lowering of pH to 3.0. Notably, lowering of kinetic stability (manifesting as increase in rates of unfolding) as a consequence of pH changes, has previously been proposed and demonstrated in at least one other instance, in a thermostable ferredoxin [6]. Changes wrought by heating are also clearly evident in PfuTIM's far-UV CD spectra at 95 °C (Fig. 3C), which reveal greater contribution from random coil structures at pH 3.0. 3.5. Cooling from 95 °C induces cold denaturation in partially denatured PfuTIM Supplementary Figs. 2A and B show that heating and cooling at pH 3.0 cause PfuTIM to initially display heat-denaturation followed by cold denaturation, even as earlier demonstrated with PfuTIM in 2 M or 4 M Gdm.HCl [11]. Heating and cooling also have effects on PfuTIM's fluorescence at pH 10.0 and 3.0 (Supplementary Fig. 2C), and cause the pH 10.0 sample to also now reveal surface hydrophobicity (in addition to the pH 3.0 sample) in the form of ANS binding (Supplementary Fig. 2D). 3.6. Restoration of pH to 7.0 in cooled samples leads to partial refolding Dialysis of heated and cooled samples against pH 7.0 buffer at 25 °C to restore pH to 7.0, leads to a detectable reduction in the random coil content of the samples earlier exposed to pH 3.0 and 10.0 (Fig. 4A), suggesting that there is a partial reversal of changes in secondary structural content. Similarly, near-UV CD spectra of such pH 7.0restored samples (Fig. 4B) reveal that the changes seen earlier, through exposure to pH 3.0, or pH 10.0, are partially reversed, and that this is accompanied by a decrease in surface hydrophobicity (data not
Fig. 3. Thermal melting of PfuTIM in buffers of different pH (marked as insets) monitored through far-UV CD spectroscopy. (Panel A) Loss of negative ellipticity at 222 nm as a function of heating (3 °C/min) at different pH. (Panel B) Different rates of unfolding of PfuTIM at different pH, assessed through monitoring of changes in negative ellipticity at 222 nm, as a function of time. (Panel C) Far-UV CD spectra of PfuTIM collected at 95 °C in buffers of different pH.
S.K. Chandrayan, P. Guptasarma / Biochimica et Biophysica Acta 1794 (2009) 905–912
909
Fig. 4. Partial reversal of structural changes in PfuTIM (following heating and cooling, at different pH; see insets for details) resulting from restoration of pH to 7.0. (Panel A) Far-UV CD spectra. (Panel B) Near-UV CD spectra. (Panel C) Fluorescence emission spectra. (Panel D) Fluorescence emission of ANS in control solutions (marked pH 3.0 C, pH 7.0 C, and pH 10.0 C) and in the presence of PfuTIM, at pH 3.0, 7.0 and 10.0.
3.7. Removal of key ionic networks to create mPfuTIM We also examined the effects of destroying key ionic interactions in PfuTIM, involving helix 4 in the 4th beta/alpha unit of PfuTIM's eight-fold beta/alpha unit structure, consisting of a short helix, a loop and a longer helix containing 6 charged residues: Asp105, Glu107, Arg111, Arg112, GLu114 and Glu115. Five of these six residues form ion pair networks involving specific ionic interactions between some of these residues, and other residues such as Lys71. In all there are four ion pair interactions within this region, involving Glu107–Arg111, Arg111–Glu114, Arg112–Glu115, and Glu115–Lys71 [13]. We made four site-directed mutations, R111T, R112A, E114K and E115G, to destroy all four ion pair interactions, by replacing four of the charged residues in helix 4 with structurally analogous residues from the psychrophile Methanococcoides burtonii TIM to create a mutant, mPfuTIM (for further details, see Supplementary Figs. 8A, 8B, 9A and 9B). 3.8. mPfuTIM is structurally identical to PfuTIM, but less stable mPfuTIM was discovered to be identical to PfuTIM (wild-type) in respect of secondary structure (Supplementary Fig. 10A), hydrodynamic volume (Supplementary Fig. 10D), and fluorescence characteristics (Supplementary Fig. 10C); however, differences were seen in near-UV CD (Supplementary Fig. 10B), suggesting subtle differences in association of aromatic residues with chiral structures. Relative to PfuTIM, a profound lowering of stability is seen in mPfuTIM, which undergoes irreversible, cooperative thermal melting between 92 °C
and 100 °C (Fig. 5A), during heating, whereas PfuTIM shows no such behavior. We heated mPfuTIM at different heating rates of 3 °C/min and 1 °C/min, but found no difference in the profiles (Supplementary Fig. 11). Fig. 5B shows far-UV CD spectra of cooled mPfuTIM and PfuTIM, establishing the extent, and irreversibility (also see Fig. 5C, presenting fluorescence emission data) of mPfuTIM's unfolding. The shoulders in the emission spectra come from the solvent Raman scatter peaks, which show up when the signal from the protein is lower as e.g., with the use of lower protein concentrations. It may be noted that concentrations of 0.1 mg/ml were used for comparisons of PfuTIM with mPfuTIM, whereas concentrations of 0.28 mg/ml were used for the other experiments involving PfuTIM. Lack of ANS binding by mutant PfuTIM (Supplementary Fig. 12C) also suggests irreversibility of unfolding, since completely unfolded chain sections would display no exposed hydrophobic clusters. Fig. 5D shows that Gdm.HCl unfolds mPfuTIM with an apparent Cm of 4.8 M, and PfuTIM with an apparent Cm of 5.4 M. Detailed data may be seen in Supplementary Fig. 12A and B. 3.9. mPfuTIM is less kinetically stable than wild-type PfuTIM Fig. 6A plots changes in the CD signal of mPfuTIM and PfuTIM at 222 nm as a function of time, during incubation at 98 °C. It may be noted that the figure shows data from the time point at which the solution reached thermal equilibrium at 98 °C. So, some structural change already occurs in both the PfuTIM and mPfuTIM samples, by the time the solutions reach 98 °C, during heating from room
910
S.K. Chandrayan, P. Guptasarma / Biochimica et Biophysica Acta 1794 (2009) 905–912
Fig. 5. (Panel A) Unfolding-induced changes in negative ellipticity at 222 nm of mPfuTIM and PfuTIM (wild-type) accompanying heating (to 100 °C) and cooling (to 25 °C). (Panel B) Far-UV CD spectra of mPfuTIM and PfuTIM after heating and cooling. (Panel C) Fluorescence emission spectra of mPfuTIM and PfuTIM after heating and cooling. (Panel D) Equilibrium Gdm.HCl-induced unfolding of mPfuTIM and PfuTIM (48 h incubations).
temperature (as seen also in Fig. 5A). Notably, once 98 °C is reached and the monitoring is started, PfuTIM shows no further changes. In contrast, mPfuTIM unfolds with a fast phase between 0 and ∼ 300 s, and a subsequent slower phase, indicating that it has been kinetically destabilized, as is evident also in its unfolding by 5 M, and 6 M, Gdm. HCl, at 25 °C (Fig. 6C and D). Notably, however, both mPfuTIM and PfuTIM remain resistant to unfolding by SDS according to the SDSPAGE-based assay (Fig. 6B), indicating that the mutations do not destabilize mPfuTIM as much as transfer to pH 3.0, in this regard. To dissect out this observation further, we created two single point mutants, E114K (to abolish the R111–E114 ionic interaction) and E115G (to abolish the R112–E115 and K71–E115 interactions), to examine the individual roles of these residues. Supplementary Fig. 13 shows that neither the E114K mutant, nor the E115G mutant, displays any reduction in kinetic stability, nor any melting at a temperature below 100 °C, as was seen earlier with mPfuTIM. Interestingly, although both mutants have the same far-UV CD and fluorescence emission characteristics as PfuTIM (just as was observed with mPfuTIM), the near-UV CD spectrum of each of the two mutants is also indistinguishable from that of PfuTIM. In contrast, significant changes were seen in the near-UV CD spectrum of mPfuTIM, relative to PfuTIM. Thus, clearly neither E114K nor E115G elicits any sign of the changes wrought by the collective making of four mutations in mPfuTIM, although these mutations individually abolish 1, and 2, ion pairs, respectively, from amongst the four ion pairs abolished by the mutations introduced in mPfuTIM. It is also interesting that E114K fails to have any effect despite the introduction of a charge of opposite character. Our data thus suggests that an additive effect is at play, i.e., that the abolishing of 1, or 2, ion pairs, fails to cause the effect obtained
by abolishing 4 ion pairs collectively. To abolish the only ion pair not affected by the mutations E114K or E115G, i.e., R111–E107, we introduced the mutation R111T; however, although DNA sequencing confirmed the introduction of this mutation, the protein failed to be detectably expressed in the same bacterial host to make all the other protein variants. 4. Discussion As charge–charge interactions are extremely distance-sensitive, subtle alterations in a protein's structure – which can remain undetected by spectroscopic methods such as CD or fluorescence spectroscopy – could conceivably be associated with significant alterations of the strengths of ionic interactions on a protein's surface, through reduced coulombic attractions (both as the cause, and the consequence, of the destruction of such interactions). Thus, upon the loss of certain ionic interactions, a protein could potentially retain its native structure at room temperature and still have become substantially structurally destabilized. From a purely thermodynamic perspective, such destabilization is well known, and there are many descriptions of reductions in the thermodynamic stabilities of proteins based on the destruction of ionic interactions [6,14,15]. In this paper, however, our focus has been on showing that destruction of ionic interactions profoundly destabilizes a hyperthermostable protein, PfuTIM, in terms of its kinetic thermal (conformational) stability, without profoundly altering its fold. As discussed in the Introduction, one manner in which destruction of ionic interactions through lowering of pH could make denaturation more facile, in kinetic terms, could be through lowering of the
S.K. Chandrayan, P. Guptasarma / Biochimica et Biophysica Acta 1794 (2009) 905–912
911
Fig. 6. (Panel A) Rates of unfolding of mPfuTIM and PfuTIM at 98 °C, at pH 8.0, assessed through monitoring of changes in negative ellipticity at 222 nm, as a function of time. (Panel B) SDS-PAGE electrophoretograms of PfuTIM (lanes 2 and 3) and mPfuTIM (lanes 4 and 5) mixed with SDS-PAGE loading buffer, and electrophoresed without (lanes 3 and 5), and with (lanes 2 and 4) boiling. Lane 1 shows molecular weight markers of 14.4, 18.0, 25.0, 35.0, 45.0, 66.0 and 116.0 kDa (bottom to top). (Panels C and D) Rates of unfolding of mPfuTIM and PfuTIM at 25 °C, in 5 M and 6 M Gdm.HCl, respectively, at pH 8.0, assessed through monitoring of changes in negative ellipticity at 222 nm, as a function of time.
structural autonomy of individual substructures (e.g., beta/alpha units). In the results presented above, the destruction of ionic interactions in the 4th beta/alpha unit of PfuTIM can be assumed to have drastically reduced the individual thermodynamic stability of this substructure within the protein's overall structure, and hence also the autonomy with which this substructure is formed and retained. We have shown that there is no discernible change in structure at room temperature in mPfuTIM; however, there is clearly a drastic decrease in its overall conformational stability that is also evident in terms of speeding up its unfolding. Therefore, our results may be taken to be circumstantial evidence for a mechanistic role for ionic interactions in mediating kinetic thermal stability in this hyperthermophile protein through decrease in the substructural autonomy of a defined substructure, because all the 4 ionic interactions destroyed in mPfuTIM are within the 4th beta/alpha unit. Likewise, there are other substructures also in PfuTIM within which surface ionic interactions similarly occur, and it may be surmised that the abolishing of such interactions (in addition to those abolished in the work described here) would further lower the kinetic thermal stability of PfuTIM. Acknowledgements SKC thanks CSIR, New Delhi, for a doctoral research fellowship. PG thanks CSIR and DBT, New Delhi, for grants to study protein folding and protein engineering.
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.bbapap.2009.03.005.
References [1] K.A. Luke, C.L. Higgins, P. Wittung-Stafshede, Thermodynamic stability and folding of proteins from hyperthermophilic organisms, FEBS J. 274 (2007) 4023–4033. [2] A. Razvi, J.M. Scholtz, Lessons in stability from thermophilic proteins, Protein Sci. 15 (2006) 1569–1578. [3] C. Vieille, G.J. Zeikus, Hyperthermophilic enzymes: sources, uses, and molecular mechanisms for thermostability, Microbiol. Mol. Biol. Rev. 65 (2001) 1–43. [4] K. Ogasahara, M. Nakamura, S. Nakura, S. Tsunasawa, I. Kato, T. Yoshimoto, K. Yutani, The unusually slow unfolding rate causes the high stability of pyrrolidone carboxyl peptidase from a hyperthermophile, Pyrococcus furiosus: equilibrium and kinetic studies of guanidine hydrochloride-induced unfolding and refolding, Biochemistry 37 (1998) 17537–17544. [5] J.K. Kaushik, K. Ogasahara, K. Yutani, The unusually slow relaxation kinetics of the folding–unfolding of pyrrolidone carboxyl peptidase from a hyperthermophile, Pyrococcus furiosus, J. Mol. Biol. 316 (2002) 991–1003. [6] P. Wittung-Stafshede, Slow unfolding explains high stability of thermostable ferredoxins: common mechanism governing thermostability? Biochim. Biophys. Acta 1700 (2004) 1–4. [7] S. D'Amico, J.C. Marx, C. Gerday, G. Feller, Activity–stability relationships in extremophilic enzymes, J. Biol. Chem. 278 (2003) 7891–7896. [8] S. Maiti, M. Luthra-Guptasarma, P. Guptasarma, Phenomenological perspectives on the folding of beta/alpha-barrel domains through the modular formation and assembly of smaller structural elements, IUBMB Life 54 (2002) 213–221. [9] A. Shukla, P. Guptasarma, Folding of beta/alpha-unit scrambled forms of S. cerevisiae triosephosphate isomerase: evidence for autonomy of substructure
912
S.K. Chandrayan, P. Guptasarma / Biochimica et Biophysica Acta 1794 (2009) 905–912
formation and plasticity of hydrophobic and hydrogen bonding interactions in core of (beta/alpha)8-barrel, Proteins 55 (2004) 548–557. [10] S. Mukherjee, S. Sharma, S. Kumar, P. Guptasarma, Slow irreversible unfolding of Pyrococcus furiosus triosephosphate isomerase: separation and quantitation of conformers through a novel electrophoretic approach, Anal. Biochem. 347 (2005) 49–59. [11] S.K. Chandrayan, P. Guptasarma, Partial destabilization of native structure by a combination of heat and denaturant facilitates cold denaturation in a hyperthermophile protein, Proteins 72 (2008) 539–546. [12] M. Kohlhoff, A. Dahm, R. Hensel, Tetrameric triosephosphate isomerase from hyperthermophilic archaea, FEBS Lett. 383 (1996) 245–250.
[13] H. Walden, S.G. Bell, R.J.M. Russell, B. Siebers, R. Hensel, G.L. Taylor, Tiny TIM: a small tetrameric hyperthermostable triosephosphate isomerase, J. Mol. Biol. 306 (2001) 745–757. [14] O.J. Halskau, R. Perez-Jimenez, B. Ibarra-Molero, J. Underhaug, V. Munoz, A. Martinez, J.M. Sanchez-Ruiz, Large-scale modulation of thermodynamic protein folding barriers linked to electrostatics, Proc. Natl. Acad. Sci. U. S. A. 105 (2008) 8625–8630. [15] A.L. Pey, D. Rodriguez-Larrea, S. Bomke, S. Dammers, R. Godoy-Ruiz, M.M. GarciaMira, J.M. Sanchez-Ruiz, Engineering proteins with tunable thermodynamic and kinetic stabilities, Proteins 71 (2008) 165–174.