- Email: [email protected]
PSD-95 and SAP97
Protein Expression and Purification 68 (2009) 201–207
Contents lists available at ScienceDirect
Protein Expression and Purification journal homepage: www.elsevier.com/locate/yprep
Structural differences between the SH3-HOOK-GuK domains of SAP90/PSD-95 and SAP97 Rama Ramesh Vandanapu a,1, Aditya Kumar Singh a,1, Marina Mikhaylova b, Pasham Parameshwar Reddy b, Michael R. Kreutz b,*, Yogendra Sharma a,* a b
Centre for Cellular and Molecular Biology, Council for Scientific and Industrial Research (CSIR), Hyderabad-500 007, India PG Neuroplasticity, Leibniz Institute for Neurobiology, Brenneckestr. 6, Magdeburg 39118, Germany
a r t i c l e
i n f o
Article history: Received 20 May 2009 and in revised form 14 July 2009 Available online 24 July 2009 Keywords: MAGUK Post synaptic density Thermal unfolding Equilibrium unfolding Hydrophobic interaction chromatography GdmCl, guanidine hydrochloride DSC, differential scanning calorimetry
a b s t r a c t The SH3-HOOK-GUK domains of the postsynaptic scaffolding proteins SAP90/PSD-95 and SAP97 are established targets of synaptic plasticity processes in the brain. A crucial molecular mechanism involved is the transition of this domain to different conformational states. We purified the SH3-HOOK-GUK domain of both proteins to examine variations in protein conformation and stability. As monitored by circular dichroism and differential scanning calorimetry, SAP97 (Tm = 64 °C) is significantly more thermal stable than SAP90/PSD-95 (Tm = 52 °C) and follows a bimodal phase transition. GdmCl-induced equilibrium unfolding of both proteins follows the two-state transitions and thus does not involve the accumulation of stable intermediate state(s). Equilibrium unfolding of SAP97 is highly cooperative from a native state to an unfolded state. In contrast, SAP90/PSD-95 follows a non-cooperative transition from native to unfolded states. A highly cooperative unfolding reaction in case of SAP97 indicates that the protein existed initially as a compact, well-folded structure, while the gradual, non-cooperative melting reaction in case of SAP90/PSD-95 indicates that the protein is in comparison more flexible. Ó 2009 Elsevier Inc. All rights reserved.
Introduction The membrane-associated guanylate kinases (MAGUKs)2, a protein family that has been identified at cell–cell contact sites in different organisms including humans [1], are thought to serve important functions in organizing the postsynaptic molecular meshwork. MAGUKs are modular scaffolds that are involved in the clustering of synaptic membrane receptors and cell adhesion molecules, as well as to downstream signaling components of the synapse [1]. The MAGUKs of brain synapses (i.e. SAP90/PSD-95, SAP97, SAP102, PSD-93/Chapsyn110) are multidomain-proteins composed of three PDZ domains, an SH3 domain, a HOOK region and a guanylate kinase-like domain (GUK domain) (Fig. 1). The crystal structure of the SH3-HOOK-GUK region of SAP90/PSD-95 has been solved previously [2,3]. The GUK domain in MAGUKs lacks key amino acid residues required for binding of nucleoside phosphate and it is assumed that the guanylate kinase-related regions of these multidomain-proteins have adopted a role in protein–protein interactions rather than having an enzymatic role [1]. Accord-
* Corresponding authors. E-mail addresses: [email protected] (M.R. Kreutz), yogendra@ccmb. res.in (Y. Sharma). 1 These authors contributed equally. 2 Abbreviations used: PSD, post synaptic density; MAGuK, membrane associated guanylate kinase; CD, circular dichroism. 1046-5928/$ - see front matter Ó 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.pep.2009.07.007
ingly, a number of protein–protein interactions have been mapped to this region including those to GKAPs [4,5] and SPAR [6]. Interestingly, the SH3 domain in MAGUK proteins interacts intra- and possibly inter-molecularly with the GUK domain [5,7] resulting in oligomerization. It has therefore been proposed that this interaction participates in the dynamic formation of the membrane cytoskeleton [8]. Thus, the transition from closed (intramolecular binding) to open conformations (no intramolecular binding) may influence the accessibility of binding interfaces for protein interactions that are crucial for the scaffolding function of MAGUKs. Previous work has also shown that among others, calmodulin binds to the HOOK region localized between the SH3 and the GUK domains of SAP97 and SAP102 in a Ca2+-dependent manner. It was argued that this binding might be involved in the transition from a closed to an open conformation abrogating the intramolecular interaction [2] and could therefore be an important mechanism for the dynamic organization of the PSD by synaptic Ca2+ transients [8–10]. Analysis of the amino acid sequence of SH3-HOOK-GUK region of SAP90/PSD-95 and SAP97 reveals that both are quite similar, except that there is an insertion of 37 (I2 motif) or 33 amino acid residues (I3 motif) in SAP97 (Fig. 1a and b). SH3 domains in general are intriguing targets for studying signal transduction process, drug design and protein folding dynamics [11,12]. In this paper, we describe an efficient method based on hydrophobic interaction chromatography for the preparation
202
R.R. Vandanapu et al. / Protein Expression and Purification 68 (2009) 201–207
Fig. 1. (a) Domain organization of SAP90/PSD-95 and SAP97. The underlined region was cloned to produce untagged recombinant proteins. (b) Sequence alignment of SAP90/ PSD-95 and SAP97 (isoform CRAa) along with the consensus sequence.
of highly pure untagged SH3-HOOK-GUK modules of SAP90/PSD95 and SAP97 and compared their conformational and unfolding transitions using a number of spectroscopic and calorimetric methods. We report that despite high sequence similarity conformation and stability of these two crucial domains are quite dissimilar.
Materials and methods Sub-cloning of SAP90/PSD-95 and SAP97 in a pET21a vector The DNA encoding SH3-HOOK-GUK region of rat SAP90/PSD-95 (1291–2228 of Accession No. X66474) and SAP97 (2141–3187 of Accession No. U14950, the splice isoform containing the I2 insertion) and cloned in pET21a vector (Novagen) to over-express both protein fragments without any fusion tag. The PCR product was cloned first blunt-end into a TOPO vector (Invitrogen) and then sub-cloned in the pET21a expression vector using the NdeI and EcoRI restriction sites and checked subsequently by DNAsequencing.
Over-expression of untagged proteins The recombinant protein SAP90/PSD-95 was expressed in the bacterial strain Escherichia coli BL21(DE3) (Invitrogen) using LB medium after induction with 0.5 mM isopropyl thio-b-D-galactopyranoside (IPTG) for 10 h at 37 °C containing 100 lg ampicillin. Over-expression of SAP97 in E. coli was induced with 0.2 mM IPTG for 15 h at 25 or 18 °C, respectively.
Hydrophobic interaction chromatography (HIC) and gel filtration A protocol for the purification of untagged proteins was developed. After induction, the harvested bacterial cells were lysed by lysozyme and sonication at 4 °C in lysis buffer (50 mM Tris–HCl, pH 7.0 containing 1 mM EDTA, 1 mM DTT, 100 mM NaCl and 0.1 mM phenyl methyl sulphonyl fluoride (PMSF) [13–15]. The supernatant was precipitated with 20% ammonium sulfate prepared in lysis buffer and centrifuged. After centrifugation, the supernatant was applied to a phenylSepharose column (GE Life sciences), equilibrated with same buffer (lysis buffer) containing 20% ammonium sulfate and then washed with the same buffer but containing 10% ammonium sulfate. Protein was eluted using ammonium sulfate free buffer (50 mM Tris–HCl, pH 7.0, 1 mM EDTA, 1 mM DTT and 100 mM NaCl). The concentrated protein was further purified on a Superdex 75 sizeexclusion column (Pharmacia) equilibrated in 50 mM Tris–HCl (pH 7.5) buffer containing 1 mM EDTA, 1 mM DTT and 100 mM NaCl. Circular dichroism (CD) measurements Near- and far-UV CD spectra were recorded on a Jasco J-815 spectropolarimeter at 25 °C with 5 accumulations in 25 mM Tris– HCl, pH 7.5 buffer containing 100 mM KCl and 1 mM DTT. For far-UV CD, the path length of cuvette used was 0.05 cm with a protein concentration of 0.5 mg/ml. Near-UV CD was performed in the same buffer in a 1 cm path length cell at a protein concentration of 1 mg/ml. All spectra were baseline corrected. Ellipticity data were expressed in millidegree.
R.R. Vandanapu et al. / Protein Expression and Purification 68 (2009) 201–207
203
Differential scanning calorimetry measurements Differential scanning calorimetry measurements were made on a Microcal VP-DSC microcalorimeter (MicroCal, LLC Europe) at a heating rate of 1 °C/min from 10 to 65 °C in case of SAP 90/PSD95 and from 10 to 75 °C in case of SAP97. For all the measurements, protein was prepared in 50 mM Tris pH 7.5, 100 mM KCl. The concentration of protein in measuring cell was 0.5 mg/ml. The temperature dependence of specific heat capacity (Cp) was analysed according to the simple two-state model, assuming the difference between heat capacities of the denatured and native protein states (DCp) is independent of temperature (All values were normalized by protein molecular weight). The experimental data were deconvoluted using ORIGIN software provided by the manufacturer.
Fluorescence spectroscopy Fluorescence emission spectra were recorded in the correct spectrum mode on a F-4500 Hitachi Fluorescence spectrophotometer using an excitation wavelength of 295 nm in 25 mM Tris–HCl, pH 7, containing 100 mM KCl and 1 mM DTT. The excitation and emission band passes were set at 5 nm each. For 8-anilino-1-naphthalene sulfonic acid (ANS) binding experiments, ANS (final concentration 100 lM) was mixed with the protein solution in 25 mM Tris buffer, pH 7 containing 50 mM KCl, incubated for 15 min and spectra were recorded from 400 to 600 nm at the excitation of 365 nm in the correct spectrum mode. ANS buffer blank spectra were also recorded under identical conditions without any protein.
Equilibrium unfolding studies Equilibrium unfolding of both SAP90/PSD-95 and SAP97 proteins (0.1 mg/ml) was carried out using guanidinium chloride (GdmCl) concentrations in the range of 0–6 M with increments of 0.1 M. Trp fluorescence was monitored by exciting the sample at 295 nm and fluorescence signal monitored at 340 nm. Corrected spectra at emission/excitation slits of 5/5 nm were recorded with a response time of 2 s. A two-state model was used for fitting SAP90/PSD-95 and SAP97 proteins. Both fits were evaluated by statistical tests available in the software Graph pad prism. The data were fitted to two or higher state unfolding models described [16] below and best fit parameters were determined.
Y¼
fsn þ sd expððg1 m1 DÞ=RTÞg f1 þ expððg1 m1 DÞ=RTÞg
ð1Þ
where Y is the observed spectroscopic signal, sn and sd represents spectroscopic signal of native and denatured protein, g1 and m1 represents the free energy change and slope of the transition, D is the denaturant concentration, T is the temperature in Kelvin and R is the universal gas constant (value = 1.987 cal K1 mol1).
Thermal unfolding by circular dichroism measurements Circular dichroism measurements were made on a JASCO J-815 spectropolarimeter using 1 cm path length cuvette at 220 nm. Temperature was maintained using peltier heating system at a rate of 1 °C/min. Spectra were recorded in buffer containing 50 mM Tris–Cl (pH 7.5), 100 mM KCl, 1 mM DTT and protein concentration of 0.25 mg/ml was used for thermal unfolding studies and the data were fit using ORIGIN software.
Fig. 2. Schematic representation of the protocol for the purification of SH3-HOOKGUK modules of PSD-95/SAP90 and SAP97 proteins.
Results SH3-HOOK-GUK domains of SAP90/PSD-95 and SAP97 To overcome the problem of fusion tag removal, we cloned both SH3-HOOK-GUK domains (SAP90/PSD-95 and SAP97) in a pET21a expression vector without tag and over-expressed the constructs in E. coli BL21 (DE3) strain. SAP90/PSD-95 was expressed as a soluble protein, when induced at 37 °C with 0.5 mM IPTG. However, in case of SAP97, only a fraction of about 30% expressed as soluble protein while the rest of the protein incorporated into inclusion bodies. To enhance the protein expression in the soluble fraction, we varied the expression conditions, such as induction at low temperature (25 °C) and the use of lower concentrations of IPTG (0.2 mM). At these conditions, the soluble fraction contains more than 60% of the corresponding protein. For subsequent biophysical and biochemical studies we designed a strategy to standardize the method of purification for the untagged SH3-HOOK-GUK region of MAGUKs based on the application of hydrophobic interaction chromatography (HIC). In order to purify SAP90/PSD-95 or SAP97, we loaded the bacterial lysate to a phenyl-Sepharose column in the presence of 20% ammonium sulfate. Both proteins were bound to the resin efficiently in the presence of 20% ammonium sulfate and 10% ammonium sulfate was used for washing the column. Minor impurities were removed by adopting gel filtration as a final step of purification (Fig. 2). We used a Superdex 75 preparative column to obtain a better resolution (Scheme 1). As shown by SDS–PAGE, the proteins thus purified were highly homogenous and stable after gel-filtration chromatography. Fluorescence spectroscopy and surface hydrophobicity SH3-HOOK-GUK domains of SAP90/PSD-95 and SAP 97 have 5 and 3 Trp residues respectively. We used Trp fluorescence to assess the protein tertiary structure and proper folding. Both protein domains exhibit wavelength maxima at about 340 and 343 nm, suggesting that the Trp is not buried inside the protein core (Fig. 3a). Moreover, both proteins exhibited a similar fluorescence spectroscopic pattern suggesting that their Trp microenvironments are comparable as one can expect from their primary structures. We then used extrinsic fluorescence to compare the hydrophobicity of the proteins using ANS as a probe. Both proteins exhibited only
204
R.R. Vandanapu et al. / Protein Expression and Purification 68 (2009) 201–207
Scheme 1. Purification of SAP90 and SAP97 by hydrophobic chromatography: Fractions obtained after HIC were analysed by SDS–PAGE of SAP90/PSD and SAP97. (a) SAP90/ PSD-95, and (b) SAP97. LMW: low molecular weight protein marker, UI: uninduced, TLI: total lysate induced, lanes H1–H6: HIC (Phenyl-Sepharose) purified fractions, GF1 and GF2: gel filtered fractions (Superdex 75). In case of SAP97, 25 °C and 37 °C represent two different induction temperatures.
Fig. 3. Comparison of intrinsic and extrinsic fluorescence and CD of SAP90/PSD-95 and SAP97. (a) represents intrinsic Trp fluorescence and (b) represents the ANS protein complex. The ANS concentration used was 100 lM with a protein concentration of 100 lg in 25 mM Tris buffer, pH 7.5, containing 50 mM KCl. (c + d) Far-UV CD spectra of SAP90/PSD-95 and SAP97. Protein concentration was 150 lg (SAP90/PSD-95) and 300 lg (SAP97) with a pathlength of 0.1 cm. (e) and (f) represent near-UV CD of SAP90/PSD95 and SAP97, respectively. Protein concentration was 1 mg/ml and pathlength was 1 cm. Data are presented in millidegrees.
a marginal shift during binding of ANS as compared to buffer suggesting a very weak binding to the dye (Fig. 3b). This indicates that only few exposed hydrophobic residues are available for binding on the surface of the protein, which is rather uncommon for a recombinant protein of this size and might be due to an intramolecular or intermolecular interaction of the SH3 domain with the GUK domain, which in turn could reduce the surface hydrophobicity of the proteins in solution.
Secondary and tertiary structure To examine if the purified proteins fold properly, the native state of the proteins were monitored by far- and near-UV CD. The secondary structure shows minima around 220 and 208 nm, which indicates that both proteins are largely in a a-helical conformation and properly folded (Fig. 3c and d). Though SH3 and GUK domains consist of prominent b-strands, their contribution in
205
R.R. Vandanapu et al. / Protein Expression and Purification 68 (2009) 201–207
far-UV CD is not visible and is overlapped by helical content. While comparing the far-UV CD spectra of both proteins, we found that the intensity of the 208 nm band varies in both proteins, which is probably due to the differences in their aromatic amino acid contents. The near-UV CD spectra of SAP90/PSD-95 and SAP97 show broad bands in the range of 290 and 275–290 nm, indicating peaks for aromatic amino acids. Both proteins are rich in aromatic amino acids; SAP90/PSD-95 harbors 5 W, 15 F, 10 Y residues where as SAP97 contains 3 W, 12 F and 14 Y residues. Few peaks in the range of 265–275 nm for Phe and Tyr were also seen in near-UV CD of SAP90/PSD-95 (Fig. 3e and f). Thus, both proteins can be distinguished based on their near-UV CD spectra and appear to be properly folded.
Table 1 Thermal unfolding temperature for D50% and chi-square value for best fit curves of SAP90 and SAP97 monitored by circular dichroism.
Parameters Preferred model
SAP90/PSD-95 Two-state (M2 state)
SAP97 Two-state bimodal (MN2 state)
Higher thermal stability of SAP97 than SAP90/PSD-95 as monitored by CD
Chi-square Tm1(°C) DH1 (cal/mol) DH2 (cal/mol) DHv1 (cal/mol) DHv2 J/mol) Tm2 (°C)
1.4E6 52.2 ± 0.08 1.2E5 ± 1.6E3 — — — —
1.9E5 64.4 ± 0.02 6.6E4 ± 2.06E3 1.7E5 ± 3.5E3 2.2E5 ± 5.8E3 4.4E4 ± 8.2E3 57.4 ± 0.19
The thermal stability of both MAGUKs was compared by monitoring the loss in secondary structure by far-UV CD measurements at 220 nm. The obtained data fit to a two-state model. As depicted in Fig. 4, there is almost no change in the ellipticity till 45 °C. Upon further increasing the temperature, the ellipticity values at 220 nm started decreasing suggesting that the protein is undergoing unfolding. Both proteins were completely unfolded beyond 70 °C. The D50% values calculated for SAP97 and SAP90/PSD-95 were 64 and 52 °C respectively. These data suggest that SAP97 is thermally more stable than SAP90/PSD-95. Higher thermostability of SAP97 can also be seen from the higher negative value of free energy (GU increases with respect to GN).
Parameters
SAP90/PSD-95
SAP97
Chi-square Temperature at D50% Standard error
0.0011 51.85 0.17
5.1E4 62.34 0.13
Table 2 Melting temperature and enthalpy of transitions of two proteins SAP90/PSD-95 and SAP97 monitored by differential scanning calorimetry.
Table 3 Thermodynamic parameters for equilibrium unfolding of SAP90/PSD-95 and SAP97 in presence of various concentrations of GdmCl monitored by Trp fluorescence emission. Proteins
Model used for fitting
DG° (kcal/ mol K)
m (kcal/ mol2 K)
D50% (M)
SAP90/PSD95 SAP97
Two-state
2.4
8.5
1.14
Two-state
2.1
8.1
1.05
Bimodal transition of thermal unfolding of SAP97 by DSC To follow up this potentially important difference between both proteins, we performed a calorimetric study of thermal denaturation of SAP90/PSD-95 and SAP97 in aqueous solutions. Thermal unfolding of SAP90/PSD-95 and SAP97 carried out using DSC corroborated the highly stable nature of the proteins seen by CD (Fig. 4a). Importantly, DSC also shows essential differences in the thermostability of both proteins (Fig. 4b, c). As seen in Table 1, the thermodynamic parameters for thermal unfolding of both proteins are significantly different from each other. SAP90/PSD-95 shows a single endotherm peak in the DSC curve near 52 °C with an enthalpy change of about 1.2E5 ± 1.6E3 kJ mol1. However, SAP97 melts in two different temperature ranges: near 56 and 64 °C with enthalpy changes of about 1.7E5 ± 3.5E3 and
6.6E4 ± 2.0E3 kJ mol1, respectively, demonstrating a bimodal transition (Table 1). Deconvolution analysis shows that the single endotherm is well approximated as the sum of two-state transitions. Thus, the two transitions of the bimodal DSC curve for SAP97 are not of a two-state type. SAP97 melts probably as two structurally independent parts, the peak with smaller enthalpy is at low temperature (56 °C) whereas the peak with larger enthalpy is at higher temperature (64 °C) (Fig. 4). Although thermal unfolding of both MAGUKs follows a two-state unfolding model it is in case of SAP97 a two-state bimodal transition. The Tm of both transitions of thermal unfolding of SAP97 were 56 °C and 64 °C, which are significantly higher than the single Tm of SAP90, which is 53 °C (Table 1).
Fig. 4. Thermal unfolding of SAP90/PSD-95 observed by CD and DSC. (a) Thermal unfolding of SAP90/PSD-95 and SAP97 monitored by circular dichroism spectroscopy. Figure represents the fraction of SAP90/PSD-95 and SAP97 proteins unfolded with increase in temperature. The best fit values for the curves were represented in Table 1. (b) and (c) represent the best fit curves for the thermal unfolding of the proteins SAP90/PSD-95 and SAP97 monitored by differential scanning calorimetry, respectively. Tm and best fit values for the data are represented in Table 2.
206
R.R. Vandanapu et al. / Protein Expression and Purification 68 (2009) 201–207
Cooperative unfolding of SAP97 while SAP90/PSD-95 undergoes noncooperative unfolding These data suggest that SAP97 might have a more compact structure with higher conformational stability. The equilibrium unfolding of both MAGUKs by GdmCl was monitored by Trp fluorescence. The data were best fitted into equations representing two-state transition (Fig. 5). The transition phase begins from 0.8 M GdmCl and ends at 1.3 M GdmCl. The protein remains largely in a native state by 0.8 M GdmCl and is completely unfolded beyond 1.3 M. This pattern is not seen in case of SAP90/PSD-95 indicating that as opposed to SAP97 non-cooperative folding transitions occur. The D50% for SAP90/PSD-95 and SAP97 were found at 1.14 and 1.05 M, respectively, further suggesting a relatively higher conformational stability of SAP97. Moreover, the data do not indicate the presence of partially folded intermediate states in SAP97, which is surprising given the large size of the protein (35 kDa) with its multiple domains. The cooperativity of an unfolding reaction is generally measured qualitatively by the width and shape of the unfolding transition. Interestingly, the global unfolding of both proteins was also found to be significantly different. As seen in Fig. 5, a sudden change in the fluorescence signal in case of SAP97 is visible, suggesting very high cooperativity between domains during unfolding transitions. A highly cooperative unfolding reaction in case of SAP97 indicates that the protein existed initially as a compact, well-folded structure, while a gradual, non-cooperative melting reaction in case of SAP90/PSD-95 indicates that the protein existed initially as a flexible, partially unfolded protein or as a heterogeneous population of folded structures.
Discussion Although a rather specialized chromatographic method for protein purification, HIC is now being exploited for various applications [13,14]. We have described now the application and optimization of a simple method of HIC for the purification of the SH3-HOOK-GUK domain of MAGUKs. The current method is quite efficient and its advantages are a high degree of purity, a high protein yield and the possibility of scaling up the procedure. Moreover, the purification can be performed in a couple of hours and does not require extensive equilibration of gel matrix and sample as required in case of ion-exchange chromatography. Thus, the application of HIC for the purification of soluble parts of multiple domain proteins like SAP90/PSD-95 or SAP97 without fusion tag is a major step forward to ease for instance biophysical studies on the interaction of these proteins with their binding partners. Subsequent studies revealed that SAP97 is thermally more stable than SAP90/PSD-95, follows cooperative unfolding transitions, has higher intrinsic stability with a bimodal curve suggesting a more compact structure than SAP90/PSD-95. It appears that hydrophobic domain interactions may be playing important role in imparting the higher stability of SAP97. Given the high sequence conservation and structural similarities, this is very interesting. A potential reason might be the insertion of the I2 motif in the HOOK region. This will extend the helical HOOK region and might bring the SH3 and GUK domains closer together, which could then shield the hydrophobic surfaces. These results suggest that conformational flexibility may play an important role in recognition of binding partners of both MAGUKs. While SAP90/PSD-95 is a highly abundant component of the PSD, located in the centre of the post-
Fig. 5. Equilibrium unfolding of SAP90/PSD-95 and SAP97 by guanidine chloride. (a and d) best fits for the unfolding data for the SAP90/PSD-95 and SAP97 proteins, (b and e) residuals represented to check the validity of the data, (c and f) fractions of proteins present in different states from native to unfolded state with varying concentration of guanidine chloride respectively. Unfolding was monitored by fluorescence emission at 340 nm with an excitation wavelength of 295 nm, and varying concentrations of GdmCl (0–6 M) were used. A two-state model was used for fitting the data for both proteins. For values of free energies, refer to Table 3. fN-fraction in native state, fU-fraction in unfolded state. For SAP90/PSD-95 and SAP97 two-state model was the best fit with D50% concentrations of 1.14 M and 1.05 M, respectively.
R.R. Vandanapu et al. / Protein Expression and Purification 68 (2009) 201–207
synaptic scaffold and thought be involved in its organization [17], SAP97 containing the I2 insertion is to a large degree located extrasynaptic and also seems to be involved in other functions like trafficking of AMPA-receptors from Golgi to the synapse [17]. These different locations and functions might impose the necessity of a more rigid and compact structure of the SH3-HOOK-GUK domain in case of SAP97 and a more flexible conformation in case of SAP90/PSD-95. In summary our data suggest that an insertion in the SH3-HOOK-GUK region can influence the structural organization and folding dynamics of MAGUK proteins and the presented production and purification scheme for recombinant SH3-HOOKGUK protein modules will help to elucidate further the molecular dynamics of protein–protein interactions in this region. Acknowledgments A.K.S. was supported by the DBT-PDF. The experimental work was supported by a joint DBT-BMBF and DAAD-DST PPP Grant award to Y.S. and M.R.K. References [1] J.M. Montgomery, P.L. Zamorano, C.C. Garner, MAGUKs in synapse assembly and function: an emerging view, Cell Mol. Life Sci. 61 (2004) 911–929. [2] A.W. McGee, S.R. Dakoji, O. Olsen, D.S. Bredt, W.A. Lim, K.E. Prehoda, Structure of the SH3-guanylate kinase module from PSD-95 suggests a mechanism for regulated assembly of MAGUK scaffolding proteins, Mol. Cell 8 (2001) 1291– 1301. [3] G.A. Tavares, E.H. Panepucci, A.T. Brunger, Structural characterization of the intramolecular interaction between the SH3 and guanylate kinase domains of PSD-95, Mol. Cell 8 (2001) 1313–1325. [4] E. Kim, S. Naisbitt, Y.P. Hsueh, A. Rao, A. Rothschild, A.M. Craig, M. Sheng, GKAP, a novel synaptic protein that interacts with the guanylate kinase-like domain of the PSD-95/SAP90 family of channel clustering molecules, J. Cell Biol. 136 (1997) 669–678.
207
[5] H. Wu, C. Reissner, S. Kuhlendahl, B. Coblentz, S. Reuver, S. Kindler, E.D. Gundelfinger, C.C. Garner, Intramolecular interactions regulate SAP97 binding to GKAP, EMBO J. 19 (2000) 5740–5751. [6] D.T. Pak, S. Yang, S. Rudolph-Correia, E. Kim, M. Sheng, Regulation of dendritic spine morphology by SPAR, a PSD-95-associated RapGAP, Neuron 31 (2001) 289–303. [7] A.W. McGee, D.S. Bredt, Identification of an intramolecular interaction between the SH3 and guanylate kinase domains of PSD-95, J. Biol. Chem. 274 (1999) 17431–17436. [8] L. Funke, S. Dakoji, D.S. Bredt, Membrane-associated guanylate kinases regulate adhesion and plasticity at cell junctions, Annu. Rev. Biochem. 74 (2005) 219–245. [9] I. Paarmann, O. Spangenberg, A. Lavie, M. Konrad, Formation of complexes between Ca2+-calmodulin and the synapse-associated protein SAP97 requires the SH3 domain-guanylate kinase domain-connecting HOOK region, J. Biol. Chem. 277 (2002) 40832–40838. [10] N. Masuko, K. Makino, H. Kuwahara, K. Fukunaga, T. Sudo, N. Araki, H. Yamamoto, Y. Yamada, E. Miyamoto, H. Saya, Interaction of NE-dlg/SAP102, a neuronal and endocrine tissue-specific membrane-associated guanylate kinase protein, with calmodulin and PSD-95/SAP90. A possible regulatory role in molecular clustering at synaptic sites, J. Biol. Chem. 274 (1999) 5782– 5790. [11] A. Levitzki, Targeting signal transduction for disease therapy, Curr. Opin. Cell Biol. 8 (1996) 239–244. [12] D.C. Dalgarno, M.C. Botfield, R.J. Rickless, SH3 domains and drug design: Ligands, structure and biological function, Biopolymers 43 (1998) 383– 400. [13] B.F. Roettger, M.R. Ladisch, Hydrophobic interaction chromatography, Biotechnol. Adv. 7 (1989) 15–29. [14] J.R. Fisher, Y. Sharma, S. Iuliano, R.A. Piccioti, D. Krylov, J. Hurley, J. Roder, A. Jeromin, Purification of myristoylated and nonmyristoylated neuronal calcium sensor-1 using single-step hydrophobic interaction chromatography, Protein Express. Purif. 20 (2000) 66–72. [15] M.K. Jobby, Y. Sharma, Rapid purification of recombinant bB2-crystallin using hydrophobic interaction chromatography, Protein Express. Purif. 28 (2003) 158–164. [16] H.C. Hung, Y.H. Chen, G.Y. Liu, H.J. Lee, G.G. Chang, Equilibrium protein folding–unfolding process involving multiple intermediates, Bull. Math. Biol. 65 (2003) 553–570. [17] T.M. Boeckers, The postsynaptic density, Cell Tissue Res. 326 (2006) 409– 422.
Contents lists available at ScienceDirect
Protein Expression and Purification journal homepage: www.elsevier.com/locate/yprep
Structural differences between the SH3-HOOK-GuK domains of SAP90/PSD-95 and SAP97 Rama Ramesh Vandanapu a,1, Aditya Kumar Singh a,1, Marina Mikhaylova b, Pasham Parameshwar Reddy b, Michael R. Kreutz b,*, Yogendra Sharma a,* a b
Centre for Cellular and Molecular Biology, Council for Scientific and Industrial Research (CSIR), Hyderabad-500 007, India PG Neuroplasticity, Leibniz Institute for Neurobiology, Brenneckestr. 6, Magdeburg 39118, Germany
a r t i c l e
i n f o
Article history: Received 20 May 2009 and in revised form 14 July 2009 Available online 24 July 2009 Keywords: MAGUK Post synaptic density Thermal unfolding Equilibrium unfolding Hydrophobic interaction chromatography GdmCl, guanidine hydrochloride DSC, differential scanning calorimetry
a b s t r a c t The SH3-HOOK-GUK domains of the postsynaptic scaffolding proteins SAP90/PSD-95 and SAP97 are established targets of synaptic plasticity processes in the brain. A crucial molecular mechanism involved is the transition of this domain to different conformational states. We purified the SH3-HOOK-GUK domain of both proteins to examine variations in protein conformation and stability. As monitored by circular dichroism and differential scanning calorimetry, SAP97 (Tm = 64 °C) is significantly more thermal stable than SAP90/PSD-95 (Tm = 52 °C) and follows a bimodal phase transition. GdmCl-induced equilibrium unfolding of both proteins follows the two-state transitions and thus does not involve the accumulation of stable intermediate state(s). Equilibrium unfolding of SAP97 is highly cooperative from a native state to an unfolded state. In contrast, SAP90/PSD-95 follows a non-cooperative transition from native to unfolded states. A highly cooperative unfolding reaction in case of SAP97 indicates that the protein existed initially as a compact, well-folded structure, while the gradual, non-cooperative melting reaction in case of SAP90/PSD-95 indicates that the protein is in comparison more flexible. Ó 2009 Elsevier Inc. All rights reserved.
Introduction The membrane-associated guanylate kinases (MAGUKs)2, a protein family that has been identified at cell–cell contact sites in different organisms including humans [1], are thought to serve important functions in organizing the postsynaptic molecular meshwork. MAGUKs are modular scaffolds that are involved in the clustering of synaptic membrane receptors and cell adhesion molecules, as well as to downstream signaling components of the synapse [1]. The MAGUKs of brain synapses (i.e. SAP90/PSD-95, SAP97, SAP102, PSD-93/Chapsyn110) are multidomain-proteins composed of three PDZ domains, an SH3 domain, a HOOK region and a guanylate kinase-like domain (GUK domain) (Fig. 1). The crystal structure of the SH3-HOOK-GUK region of SAP90/PSD-95 has been solved previously [2,3]. The GUK domain in MAGUKs lacks key amino acid residues required for binding of nucleoside phosphate and it is assumed that the guanylate kinase-related regions of these multidomain-proteins have adopted a role in protein–protein interactions rather than having an enzymatic role [1]. Accord-
* Corresponding authors. E-mail addresses: [email protected] (M.R. Kreutz), yogendra@ccmb. res.in (Y. Sharma). 1 These authors contributed equally. 2 Abbreviations used: PSD, post synaptic density; MAGuK, membrane associated guanylate kinase; CD, circular dichroism. 1046-5928/$ - see front matter Ó 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.pep.2009.07.007
ingly, a number of protein–protein interactions have been mapped to this region including those to GKAPs [4,5] and SPAR [6]. Interestingly, the SH3 domain in MAGUK proteins interacts intra- and possibly inter-molecularly with the GUK domain [5,7] resulting in oligomerization. It has therefore been proposed that this interaction participates in the dynamic formation of the membrane cytoskeleton [8]. Thus, the transition from closed (intramolecular binding) to open conformations (no intramolecular binding) may influence the accessibility of binding interfaces for protein interactions that are crucial for the scaffolding function of MAGUKs. Previous work has also shown that among others, calmodulin binds to the HOOK region localized between the SH3 and the GUK domains of SAP97 and SAP102 in a Ca2+-dependent manner. It was argued that this binding might be involved in the transition from a closed to an open conformation abrogating the intramolecular interaction [2] and could therefore be an important mechanism for the dynamic organization of the PSD by synaptic Ca2+ transients [8–10]. Analysis of the amino acid sequence of SH3-HOOK-GUK region of SAP90/PSD-95 and SAP97 reveals that both are quite similar, except that there is an insertion of 37 (I2 motif) or 33 amino acid residues (I3 motif) in SAP97 (Fig. 1a and b). SH3 domains in general are intriguing targets for studying signal transduction process, drug design and protein folding dynamics [11,12]. In this paper, we describe an efficient method based on hydrophobic interaction chromatography for the preparation
202
R.R. Vandanapu et al. / Protein Expression and Purification 68 (2009) 201–207
Fig. 1. (a) Domain organization of SAP90/PSD-95 and SAP97. The underlined region was cloned to produce untagged recombinant proteins. (b) Sequence alignment of SAP90/ PSD-95 and SAP97 (isoform CRAa) along with the consensus sequence.
of highly pure untagged SH3-HOOK-GUK modules of SAP90/PSD95 and SAP97 and compared their conformational and unfolding transitions using a number of spectroscopic and calorimetric methods. We report that despite high sequence similarity conformation and stability of these two crucial domains are quite dissimilar.
Materials and methods Sub-cloning of SAP90/PSD-95 and SAP97 in a pET21a vector The DNA encoding SH3-HOOK-GUK region of rat SAP90/PSD-95 (1291–2228 of Accession No. X66474) and SAP97 (2141–3187 of Accession No. U14950, the splice isoform containing the I2 insertion) and cloned in pET21a vector (Novagen) to over-express both protein fragments without any fusion tag. The PCR product was cloned first blunt-end into a TOPO vector (Invitrogen) and then sub-cloned in the pET21a expression vector using the NdeI and EcoRI restriction sites and checked subsequently by DNAsequencing.
Over-expression of untagged proteins The recombinant protein SAP90/PSD-95 was expressed in the bacterial strain Escherichia coli BL21(DE3) (Invitrogen) using LB medium after induction with 0.5 mM isopropyl thio-b-D-galactopyranoside (IPTG) for 10 h at 37 °C containing 100 lg ampicillin. Over-expression of SAP97 in E. coli was induced with 0.2 mM IPTG for 15 h at 25 or 18 °C, respectively.
Hydrophobic interaction chromatography (HIC) and gel filtration A protocol for the purification of untagged proteins was developed. After induction, the harvested bacterial cells were lysed by lysozyme and sonication at 4 °C in lysis buffer (50 mM Tris–HCl, pH 7.0 containing 1 mM EDTA, 1 mM DTT, 100 mM NaCl and 0.1 mM phenyl methyl sulphonyl fluoride (PMSF) [13–15]. The supernatant was precipitated with 20% ammonium sulfate prepared in lysis buffer and centrifuged. After centrifugation, the supernatant was applied to a phenylSepharose column (GE Life sciences), equilibrated with same buffer (lysis buffer) containing 20% ammonium sulfate and then washed with the same buffer but containing 10% ammonium sulfate. Protein was eluted using ammonium sulfate free buffer (50 mM Tris–HCl, pH 7.0, 1 mM EDTA, 1 mM DTT and 100 mM NaCl). The concentrated protein was further purified on a Superdex 75 sizeexclusion column (Pharmacia) equilibrated in 50 mM Tris–HCl (pH 7.5) buffer containing 1 mM EDTA, 1 mM DTT and 100 mM NaCl. Circular dichroism (CD) measurements Near- and far-UV CD spectra were recorded on a Jasco J-815 spectropolarimeter at 25 °C with 5 accumulations in 25 mM Tris– HCl, pH 7.5 buffer containing 100 mM KCl and 1 mM DTT. For far-UV CD, the path length of cuvette used was 0.05 cm with a protein concentration of 0.5 mg/ml. Near-UV CD was performed in the same buffer in a 1 cm path length cell at a protein concentration of 1 mg/ml. All spectra were baseline corrected. Ellipticity data were expressed in millidegree.
R.R. Vandanapu et al. / Protein Expression and Purification 68 (2009) 201–207
203
Differential scanning calorimetry measurements Differential scanning calorimetry measurements were made on a Microcal VP-DSC microcalorimeter (MicroCal, LLC Europe) at a heating rate of 1 °C/min from 10 to 65 °C in case of SAP 90/PSD95 and from 10 to 75 °C in case of SAP97. For all the measurements, protein was prepared in 50 mM Tris pH 7.5, 100 mM KCl. The concentration of protein in measuring cell was 0.5 mg/ml. The temperature dependence of specific heat capacity (Cp) was analysed according to the simple two-state model, assuming the difference between heat capacities of the denatured and native protein states (DCp) is independent of temperature (All values were normalized by protein molecular weight). The experimental data were deconvoluted using ORIGIN software provided by the manufacturer.
Fluorescence spectroscopy Fluorescence emission spectra were recorded in the correct spectrum mode on a F-4500 Hitachi Fluorescence spectrophotometer using an excitation wavelength of 295 nm in 25 mM Tris–HCl, pH 7, containing 100 mM KCl and 1 mM DTT. The excitation and emission band passes were set at 5 nm each. For 8-anilino-1-naphthalene sulfonic acid (ANS) binding experiments, ANS (final concentration 100 lM) was mixed with the protein solution in 25 mM Tris buffer, pH 7 containing 50 mM KCl, incubated for 15 min and spectra were recorded from 400 to 600 nm at the excitation of 365 nm in the correct spectrum mode. ANS buffer blank spectra were also recorded under identical conditions without any protein.
Equilibrium unfolding studies Equilibrium unfolding of both SAP90/PSD-95 and SAP97 proteins (0.1 mg/ml) was carried out using guanidinium chloride (GdmCl) concentrations in the range of 0–6 M with increments of 0.1 M. Trp fluorescence was monitored by exciting the sample at 295 nm and fluorescence signal monitored at 340 nm. Corrected spectra at emission/excitation slits of 5/5 nm were recorded with a response time of 2 s. A two-state model was used for fitting SAP90/PSD-95 and SAP97 proteins. Both fits were evaluated by statistical tests available in the software Graph pad prism. The data were fitted to two or higher state unfolding models described [16] below and best fit parameters were determined.
Y¼
fsn þ sd expððg1 m1 DÞ=RTÞg f1 þ expððg1 m1 DÞ=RTÞg
ð1Þ
where Y is the observed spectroscopic signal, sn and sd represents spectroscopic signal of native and denatured protein, g1 and m1 represents the free energy change and slope of the transition, D is the denaturant concentration, T is the temperature in Kelvin and R is the universal gas constant (value = 1.987 cal K1 mol1).
Thermal unfolding by circular dichroism measurements Circular dichroism measurements were made on a JASCO J-815 spectropolarimeter using 1 cm path length cuvette at 220 nm. Temperature was maintained using peltier heating system at a rate of 1 °C/min. Spectra were recorded in buffer containing 50 mM Tris–Cl (pH 7.5), 100 mM KCl, 1 mM DTT and protein concentration of 0.25 mg/ml was used for thermal unfolding studies and the data were fit using ORIGIN software.
Fig. 2. Schematic representation of the protocol for the purification of SH3-HOOKGUK modules of PSD-95/SAP90 and SAP97 proteins.
Results SH3-HOOK-GUK domains of SAP90/PSD-95 and SAP97 To overcome the problem of fusion tag removal, we cloned both SH3-HOOK-GUK domains (SAP90/PSD-95 and SAP97) in a pET21a expression vector without tag and over-expressed the constructs in E. coli BL21 (DE3) strain. SAP90/PSD-95 was expressed as a soluble protein, when induced at 37 °C with 0.5 mM IPTG. However, in case of SAP97, only a fraction of about 30% expressed as soluble protein while the rest of the protein incorporated into inclusion bodies. To enhance the protein expression in the soluble fraction, we varied the expression conditions, such as induction at low temperature (25 °C) and the use of lower concentrations of IPTG (0.2 mM). At these conditions, the soluble fraction contains more than 60% of the corresponding protein. For subsequent biophysical and biochemical studies we designed a strategy to standardize the method of purification for the untagged SH3-HOOK-GUK region of MAGUKs based on the application of hydrophobic interaction chromatography (HIC). In order to purify SAP90/PSD-95 or SAP97, we loaded the bacterial lysate to a phenyl-Sepharose column in the presence of 20% ammonium sulfate. Both proteins were bound to the resin efficiently in the presence of 20% ammonium sulfate and 10% ammonium sulfate was used for washing the column. Minor impurities were removed by adopting gel filtration as a final step of purification (Fig. 2). We used a Superdex 75 preparative column to obtain a better resolution (Scheme 1). As shown by SDS–PAGE, the proteins thus purified were highly homogenous and stable after gel-filtration chromatography. Fluorescence spectroscopy and surface hydrophobicity SH3-HOOK-GUK domains of SAP90/PSD-95 and SAP 97 have 5 and 3 Trp residues respectively. We used Trp fluorescence to assess the protein tertiary structure and proper folding. Both protein domains exhibit wavelength maxima at about 340 and 343 nm, suggesting that the Trp is not buried inside the protein core (Fig. 3a). Moreover, both proteins exhibited a similar fluorescence spectroscopic pattern suggesting that their Trp microenvironments are comparable as one can expect from their primary structures. We then used extrinsic fluorescence to compare the hydrophobicity of the proteins using ANS as a probe. Both proteins exhibited only
204
R.R. Vandanapu et al. / Protein Expression and Purification 68 (2009) 201–207
Scheme 1. Purification of SAP90 and SAP97 by hydrophobic chromatography: Fractions obtained after HIC were analysed by SDS–PAGE of SAP90/PSD and SAP97. (a) SAP90/ PSD-95, and (b) SAP97. LMW: low molecular weight protein marker, UI: uninduced, TLI: total lysate induced, lanes H1–H6: HIC (Phenyl-Sepharose) purified fractions, GF1 and GF2: gel filtered fractions (Superdex 75). In case of SAP97, 25 °C and 37 °C represent two different induction temperatures.
Fig. 3. Comparison of intrinsic and extrinsic fluorescence and CD of SAP90/PSD-95 and SAP97. (a) represents intrinsic Trp fluorescence and (b) represents the ANS protein complex. The ANS concentration used was 100 lM with a protein concentration of 100 lg in 25 mM Tris buffer, pH 7.5, containing 50 mM KCl. (c + d) Far-UV CD spectra of SAP90/PSD-95 and SAP97. Protein concentration was 150 lg (SAP90/PSD-95) and 300 lg (SAP97) with a pathlength of 0.1 cm. (e) and (f) represent near-UV CD of SAP90/PSD95 and SAP97, respectively. Protein concentration was 1 mg/ml and pathlength was 1 cm. Data are presented in millidegrees.
a marginal shift during binding of ANS as compared to buffer suggesting a very weak binding to the dye (Fig. 3b). This indicates that only few exposed hydrophobic residues are available for binding on the surface of the protein, which is rather uncommon for a recombinant protein of this size and might be due to an intramolecular or intermolecular interaction of the SH3 domain with the GUK domain, which in turn could reduce the surface hydrophobicity of the proteins in solution.
Secondary and tertiary structure To examine if the purified proteins fold properly, the native state of the proteins were monitored by far- and near-UV CD. The secondary structure shows minima around 220 and 208 nm, which indicates that both proteins are largely in a a-helical conformation and properly folded (Fig. 3c and d). Though SH3 and GUK domains consist of prominent b-strands, their contribution in
205
R.R. Vandanapu et al. / Protein Expression and Purification 68 (2009) 201–207
far-UV CD is not visible and is overlapped by helical content. While comparing the far-UV CD spectra of both proteins, we found that the intensity of the 208 nm band varies in both proteins, which is probably due to the differences in their aromatic amino acid contents. The near-UV CD spectra of SAP90/PSD-95 and SAP97 show broad bands in the range of 290 and 275–290 nm, indicating peaks for aromatic amino acids. Both proteins are rich in aromatic amino acids; SAP90/PSD-95 harbors 5 W, 15 F, 10 Y residues where as SAP97 contains 3 W, 12 F and 14 Y residues. Few peaks in the range of 265–275 nm for Phe and Tyr were also seen in near-UV CD of SAP90/PSD-95 (Fig. 3e and f). Thus, both proteins can be distinguished based on their near-UV CD spectra and appear to be properly folded.
Table 1 Thermal unfolding temperature for D50% and chi-square value for best fit curves of SAP90 and SAP97 monitored by circular dichroism.
Parameters Preferred model
SAP90/PSD-95 Two-state (M2 state)
SAP97 Two-state bimodal (MN2 state)
Higher thermal stability of SAP97 than SAP90/PSD-95 as monitored by CD
Chi-square Tm1(°C) DH1 (cal/mol) DH2 (cal/mol) DHv1 (cal/mol) DHv2 J/mol) Tm2 (°C)
1.4E6 52.2 ± 0.08 1.2E5 ± 1.6E3 — — — —
1.9E5 64.4 ± 0.02 6.6E4 ± 2.06E3 1.7E5 ± 3.5E3 2.2E5 ± 5.8E3 4.4E4 ± 8.2E3 57.4 ± 0.19
The thermal stability of both MAGUKs was compared by monitoring the loss in secondary structure by far-UV CD measurements at 220 nm. The obtained data fit to a two-state model. As depicted in Fig. 4, there is almost no change in the ellipticity till 45 °C. Upon further increasing the temperature, the ellipticity values at 220 nm started decreasing suggesting that the protein is undergoing unfolding. Both proteins were completely unfolded beyond 70 °C. The D50% values calculated for SAP97 and SAP90/PSD-95 were 64 and 52 °C respectively. These data suggest that SAP97 is thermally more stable than SAP90/PSD-95. Higher thermostability of SAP97 can also be seen from the higher negative value of free energy (GU increases with respect to GN).
Parameters
SAP90/PSD-95
SAP97
Chi-square Temperature at D50% Standard error
0.0011 51.85 0.17
5.1E4 62.34 0.13
Table 2 Melting temperature and enthalpy of transitions of two proteins SAP90/PSD-95 and SAP97 monitored by differential scanning calorimetry.
Table 3 Thermodynamic parameters for equilibrium unfolding of SAP90/PSD-95 and SAP97 in presence of various concentrations of GdmCl monitored by Trp fluorescence emission. Proteins
Model used for fitting
DG° (kcal/ mol K)
m (kcal/ mol2 K)
D50% (M)
SAP90/PSD95 SAP97
Two-state
2.4
8.5
1.14
Two-state
2.1
8.1
1.05
Bimodal transition of thermal unfolding of SAP97 by DSC To follow up this potentially important difference between both proteins, we performed a calorimetric study of thermal denaturation of SAP90/PSD-95 and SAP97 in aqueous solutions. Thermal unfolding of SAP90/PSD-95 and SAP97 carried out using DSC corroborated the highly stable nature of the proteins seen by CD (Fig. 4a). Importantly, DSC also shows essential differences in the thermostability of both proteins (Fig. 4b, c). As seen in Table 1, the thermodynamic parameters for thermal unfolding of both proteins are significantly different from each other. SAP90/PSD-95 shows a single endotherm peak in the DSC curve near 52 °C with an enthalpy change of about 1.2E5 ± 1.6E3 kJ mol1. However, SAP97 melts in two different temperature ranges: near 56 and 64 °C with enthalpy changes of about 1.7E5 ± 3.5E3 and
6.6E4 ± 2.0E3 kJ mol1, respectively, demonstrating a bimodal transition (Table 1). Deconvolution analysis shows that the single endotherm is well approximated as the sum of two-state transitions. Thus, the two transitions of the bimodal DSC curve for SAP97 are not of a two-state type. SAP97 melts probably as two structurally independent parts, the peak with smaller enthalpy is at low temperature (56 °C) whereas the peak with larger enthalpy is at higher temperature (64 °C) (Fig. 4). Although thermal unfolding of both MAGUKs follows a two-state unfolding model it is in case of SAP97 a two-state bimodal transition. The Tm of both transitions of thermal unfolding of SAP97 were 56 °C and 64 °C, which are significantly higher than the single Tm of SAP90, which is 53 °C (Table 1).
Fig. 4. Thermal unfolding of SAP90/PSD-95 observed by CD and DSC. (a) Thermal unfolding of SAP90/PSD-95 and SAP97 monitored by circular dichroism spectroscopy. Figure represents the fraction of SAP90/PSD-95 and SAP97 proteins unfolded with increase in temperature. The best fit values for the curves were represented in Table 1. (b) and (c) represent the best fit curves for the thermal unfolding of the proteins SAP90/PSD-95 and SAP97 monitored by differential scanning calorimetry, respectively. Tm and best fit values for the data are represented in Table 2.
206
R.R. Vandanapu et al. / Protein Expression and Purification 68 (2009) 201–207
Cooperative unfolding of SAP97 while SAP90/PSD-95 undergoes noncooperative unfolding These data suggest that SAP97 might have a more compact structure with higher conformational stability. The equilibrium unfolding of both MAGUKs by GdmCl was monitored by Trp fluorescence. The data were best fitted into equations representing two-state transition (Fig. 5). The transition phase begins from 0.8 M GdmCl and ends at 1.3 M GdmCl. The protein remains largely in a native state by 0.8 M GdmCl and is completely unfolded beyond 1.3 M. This pattern is not seen in case of SAP90/PSD-95 indicating that as opposed to SAP97 non-cooperative folding transitions occur. The D50% for SAP90/PSD-95 and SAP97 were found at 1.14 and 1.05 M, respectively, further suggesting a relatively higher conformational stability of SAP97. Moreover, the data do not indicate the presence of partially folded intermediate states in SAP97, which is surprising given the large size of the protein (35 kDa) with its multiple domains. The cooperativity of an unfolding reaction is generally measured qualitatively by the width and shape of the unfolding transition. Interestingly, the global unfolding of both proteins was also found to be significantly different. As seen in Fig. 5, a sudden change in the fluorescence signal in case of SAP97 is visible, suggesting very high cooperativity between domains during unfolding transitions. A highly cooperative unfolding reaction in case of SAP97 indicates that the protein existed initially as a compact, well-folded structure, while a gradual, non-cooperative melting reaction in case of SAP90/PSD-95 indicates that the protein existed initially as a flexible, partially unfolded protein or as a heterogeneous population of folded structures.
Discussion Although a rather specialized chromatographic method for protein purification, HIC is now being exploited for various applications [13,14]. We have described now the application and optimization of a simple method of HIC for the purification of the SH3-HOOK-GUK domain of MAGUKs. The current method is quite efficient and its advantages are a high degree of purity, a high protein yield and the possibility of scaling up the procedure. Moreover, the purification can be performed in a couple of hours and does not require extensive equilibration of gel matrix and sample as required in case of ion-exchange chromatography. Thus, the application of HIC for the purification of soluble parts of multiple domain proteins like SAP90/PSD-95 or SAP97 without fusion tag is a major step forward to ease for instance biophysical studies on the interaction of these proteins with their binding partners. Subsequent studies revealed that SAP97 is thermally more stable than SAP90/PSD-95, follows cooperative unfolding transitions, has higher intrinsic stability with a bimodal curve suggesting a more compact structure than SAP90/PSD-95. It appears that hydrophobic domain interactions may be playing important role in imparting the higher stability of SAP97. Given the high sequence conservation and structural similarities, this is very interesting. A potential reason might be the insertion of the I2 motif in the HOOK region. This will extend the helical HOOK region and might bring the SH3 and GUK domains closer together, which could then shield the hydrophobic surfaces. These results suggest that conformational flexibility may play an important role in recognition of binding partners of both MAGUKs. While SAP90/PSD-95 is a highly abundant component of the PSD, located in the centre of the post-
Fig. 5. Equilibrium unfolding of SAP90/PSD-95 and SAP97 by guanidine chloride. (a and d) best fits for the unfolding data for the SAP90/PSD-95 and SAP97 proteins, (b and e) residuals represented to check the validity of the data, (c and f) fractions of proteins present in different states from native to unfolded state with varying concentration of guanidine chloride respectively. Unfolding was monitored by fluorescence emission at 340 nm with an excitation wavelength of 295 nm, and varying concentrations of GdmCl (0–6 M) were used. A two-state model was used for fitting the data for both proteins. For values of free energies, refer to Table 3. fN-fraction in native state, fU-fraction in unfolded state. For SAP90/PSD-95 and SAP97 two-state model was the best fit with D50% concentrations of 1.14 M and 1.05 M, respectively.
R.R. Vandanapu et al. / Protein Expression and Purification 68 (2009) 201–207
synaptic scaffold and thought be involved in its organization [17], SAP97 containing the I2 insertion is to a large degree located extrasynaptic and also seems to be involved in other functions like trafficking of AMPA-receptors from Golgi to the synapse [17]. These different locations and functions might impose the necessity of a more rigid and compact structure of the SH3-HOOK-GUK domain in case of SAP97 and a more flexible conformation in case of SAP90/PSD-95. In summary our data suggest that an insertion in the SH3-HOOK-GUK region can influence the structural organization and folding dynamics of MAGUK proteins and the presented production and purification scheme for recombinant SH3-HOOKGUK protein modules will help to elucidate further the molecular dynamics of protein–protein interactions in this region. Acknowledgments A.K.S. was supported by the DBT-PDF. The experimental work was supported by a joint DBT-BMBF and DAAD-DST PPP Grant award to Y.S. and M.R.K. References [1] J.M. Montgomery, P.L. Zamorano, C.C. Garner, MAGUKs in synapse assembly and function: an emerging view, Cell Mol. Life Sci. 61 (2004) 911–929. [2] A.W. McGee, S.R. Dakoji, O. Olsen, D.S. Bredt, W.A. Lim, K.E. Prehoda, Structure of the SH3-guanylate kinase module from PSD-95 suggests a mechanism for regulated assembly of MAGUK scaffolding proteins, Mol. Cell 8 (2001) 1291– 1301. [3] G.A. Tavares, E.H. Panepucci, A.T. Brunger, Structural characterization of the intramolecular interaction between the SH3 and guanylate kinase domains of PSD-95, Mol. Cell 8 (2001) 1313–1325. [4] E. Kim, S. Naisbitt, Y.P. Hsueh, A. Rao, A. Rothschild, A.M. Craig, M. Sheng, GKAP, a novel synaptic protein that interacts with the guanylate kinase-like domain of the PSD-95/SAP90 family of channel clustering molecules, J. Cell Biol. 136 (1997) 669–678.
207
[5] H. Wu, C. Reissner, S. Kuhlendahl, B. Coblentz, S. Reuver, S. Kindler, E.D. Gundelfinger, C.C. Garner, Intramolecular interactions regulate SAP97 binding to GKAP, EMBO J. 19 (2000) 5740–5751. [6] D.T. Pak, S. Yang, S. Rudolph-Correia, E. Kim, M. Sheng, Regulation of dendritic spine morphology by SPAR, a PSD-95-associated RapGAP, Neuron 31 (2001) 289–303. [7] A.W. McGee, D.S. Bredt, Identification of an intramolecular interaction between the SH3 and guanylate kinase domains of PSD-95, J. Biol. Chem. 274 (1999) 17431–17436. [8] L. Funke, S. Dakoji, D.S. Bredt, Membrane-associated guanylate kinases regulate adhesion and plasticity at cell junctions, Annu. Rev. Biochem. 74 (2005) 219–245. [9] I. Paarmann, O. Spangenberg, A. Lavie, M. Konrad, Formation of complexes between Ca2+-calmodulin and the synapse-associated protein SAP97 requires the SH3 domain-guanylate kinase domain-connecting HOOK region, J. Biol. Chem. 277 (2002) 40832–40838. [10] N. Masuko, K. Makino, H. Kuwahara, K. Fukunaga, T. Sudo, N. Araki, H. Yamamoto, Y. Yamada, E. Miyamoto, H. Saya, Interaction of NE-dlg/SAP102, a neuronal and endocrine tissue-specific membrane-associated guanylate kinase protein, with calmodulin and PSD-95/SAP90. A possible regulatory role in molecular clustering at synaptic sites, J. Biol. Chem. 274 (1999) 5782– 5790. [11] A. Levitzki, Targeting signal transduction for disease therapy, Curr. Opin. Cell Biol. 8 (1996) 239–244. [12] D.C. Dalgarno, M.C. Botfield, R.J. Rickless, SH3 domains and drug design: Ligands, structure and biological function, Biopolymers 43 (1998) 383– 400. [13] B.F. Roettger, M.R. Ladisch, Hydrophobic interaction chromatography, Biotechnol. Adv. 7 (1989) 15–29. [14] J.R. Fisher, Y. Sharma, S. Iuliano, R.A. Piccioti, D. Krylov, J. Hurley, J. Roder, A. Jeromin, Purification of myristoylated and nonmyristoylated neuronal calcium sensor-1 using single-step hydrophobic interaction chromatography, Protein Express. Purif. 20 (2000) 66–72. [15] M.K. Jobby, Y. Sharma, Rapid purification of recombinant bB2-crystallin using hydrophobic interaction chromatography, Protein Express. Purif. 28 (2003) 158–164. [16] H.C. Hung, Y.H. Chen, G.Y. Liu, H.J. Lee, G.G. Chang, Equilibrium protein folding–unfolding process involving multiple intermediates, Bull. Math. Biol. 65 (2003) 553–570. [17] T.M. Boeckers, The postsynaptic density, Cell Tissue Res. 326 (2006) 409– 422.