Inter-domain interactions influence the stability and catalytic activity of the bi-domain protein tyrosine phosphatase PTP99A

Biochimica et Biophysica Acta 1824 (2012) 983–990

Contents lists available at SciVerse ScienceDirect

Biochimica et Biophysica Acta journal homepage: www.elsevier.com/locate/bbapap

Inter-domain interactions influence the stability and catalytic activity of the bi-domain protein tyrosine phosphatase PTP99A Lalima L. Madan, Kapil Goutam, B. Gopal ⁎ Molecular Biophysics Unit, Indian Institute of Science, Bangalore 560 012, India

a r t i c l e

i n f o

Article history: Received 25 October 2011 Received in revised form 30 April 2012 Accepted 3 May 2012 Available online 15 May 2012 Keywords: Silent domain Catalytic activity Activation Phosphatase Allosteric site Domain–domain interaction

a b s t r a c t The two protein tyrosine phosphatase (PTP) domains in bi-domain PTPs share high sequence and structural similarity. However, only one of the two PTP domains is catalytically active. Here we describe biochemical studies on the two tandem PTP domains of the bi-domain PTP, PTP99A. Phosphatase activity, monitored using small molecule as well as peptide substrates, revealed that the inactive (D2) domain activates the catalytic (D1) domain. Thermodynamic measurements suggest that the inactive D2 domain stabilizes the bidomain (D1–D2) protein. The mechanism by which the D2 domain activates and stabilizes the bi-domain protein is governed by few interactions at the inter-domain interface. In particular, mutating Lys990 at the interface attenuates inter-domain communication. This residue is located at a structurally equivalent location to the so-called allosteric site of the canonical single domain PTP, PTP1B. These observations suggest functional optimization in bi-domain PTPs whereby the inactive PTP domain modulates the catalytic activity of the bi-domain enzyme. © 2012 Elsevier B.V. All rights reserved.

1. Introduction Protein tyrosine phosphatases (PTPs) control signaling cascades by selectively dephosphorylating signaling proteins. PTPs are either single domain enzymes or occur as a fusion protein with other domains. Receptor protein tyrosine phosphatases (RPTPs) are membrane associated proteins where the catalytic activity of their cytosolic PTP domain is modulated by stimuli from the extracellular receptor segments. The extracellular receptor segment in RPTPs is generally composed of multiple cell adhesion molecules (CAM) [1]. The phosphatase activity of these proteins is primarily governed by the nucleophilic action of a conserved cysteine residue at the active site of the cytosolic PTP domain. This cysteine residue is a part of the ([I/V]HCGXXR[S/T]G) sequence motif [1]. PTP99A is expressed on almost all axons in the central nervous system of the Drosophila embryo. PTP99A and another RPTP, PTP69D, are responsible for the branching of Segmental Nerves (SN) SNa and SNb in the ventral muscle field of the growing Drosophila embryo [2]. PTP99A is classified as a type III RPTP. In this RPTP, the fibronectin type III repeats in the receptor are connected by a single transmembrane helix to two cytosolic PTP domains. In most dual domain RPTPs (prominent examples include Leukocyte antigen related (LAR) and PTP69D), the membrane proximal domain (D1) is active whereas the membrane distal domain (D2) lacks phosphatase activity [1]. The bi-domain PTPs HPTPα and CD45 are the only exceptions to this feature ⁎ Corresponding author at: Lab 301, Molecular Biophysics Unit, Indian Institute of Science, Bangalore 560012, India. Tel.: + 91 80 2293 3219; fax: + 91 80 2360 0535. E-mail address: [email protected] (B. Gopal). 1570-9639/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.bbapap.2012.05.001

where both D1 and D2 domains in these proteins demonstrate phosphatase activity [3]. While the role of the D1 domain in the case of the Type IIa RPTPs has been characterized [4], the functional significance of the D2 domain remains to be determined. Indeed, while differences in the catalytic parameters of the D1 domain in the presence and absence of the D2 domain have been reported earlier, no mechanism has been proposed for the modulation of catalytic activity due to an inactive PTP domain [5]. Here we describe biochemical features of a bi-domain PTP, PTP99A. Despite apparent sequence and structural similarity (47% similarity in sequence; ca 1.58 Å RMSD in the structural model), the D2 domain is inactive as it lacks the conserved nucleophilic cysteine residue at the active site. In PTP99A, the D1 domain in isolation has several fold less phosphatase activity when compared to the D1–D2 bi-domain construct. As described previously [6], this modulation of activity in the bi-domain constructs could be rationalized by a change in the interaction networks of the Functionally Important Residues (FIRs) of the two domains. While the previous study provided a structural rationale for the modulation of catalytic activity of a PTP domain due to its tandem inactive domain, here we propose a biochemical mechanism for the activation of the D1 PTP domain by the inactive D2 domain. This mechanism is based on a non-essential partially mixed model of enzyme activation and describes in detail the role of inter-domain interface residues in mediating the cross-talk between the two PTP domains. This cross-talk eventually translates into changes in FIR networks which could be visualized by molecular dynamics simulations [6]. Thermal unfolding studies highlight the role of the membrane distal D2 domain in stabilizing the bi-domain (D1–D2) protein construct. We

984

L.L. Madan et al. / Biochimica et Biophysica Acta 1824 (2012) 983–990

note that a point mutation at the inter-domain interface attenuates interaction between the D1 and D2 domains of PTP99A. These observations suggest multiple roles for the inactive D2 PTP domain involving substrate recognition, activation and thermodynamic stabilization of the catalytic D1 domain. These observations underscore the optimization of conformational features in bi-domain PTPs to modulate catalytic activity.

2. Experimental methods 2.1. Cloning of the PTP99A phosphatase domain constructs The cosmid containing Drosophila ptp99A was obtained from Prof. Kai Zinn (Caltech). The two cytosolic PTP domains of PTP99A were amplified using oligonucleotide primers and procedures reported earlier [6]. The amplified product (corresponding to the amino-acid stretch 450–1050 of PTP99A) had two PTP domains (D1–D2). The PTP domain constructs of PTP99A were amplified such that the recombinant protein corresponding to the D1 domain (residues 450–756) and the D2 domain (incorporating the sequence stretch 740–1050) could be overexpressed separately. The PCR amplified products were ligated between the NheI and XhoI restriction sites of the pET15b vector. The primer 5′ AATGTGTACCAGACGCGGCGTGTCCCACAAGCGA 3′ was used to obtain the mutant. All expression constructs (D1, D2, D1–D2, D1–D2 K990A and D2 K990A) were confirmed by single primer based sequencing (Macrogen, Inc.).

2.2. Expression and purification of PTP99A protein constructs All PTP99A recombinant protein constructs were expressed in E. coli BL21(DE3) cells and purified as reported earlier [7]. Briefly, E. coli BL21(DE3) cells were transformed with the plasmid encoding a PTP99A variant (either D1, D2, D1–D2 or the interface mutants D1–D2 K990A and D2 K990A). The E. coli cells were grown to an optical density (at 600 nm) of 0.5 at 37 °C and induced with 0.2 mM IPTG. Post induction, the temperature was lowered to 15 °C. The cells were lysed in a buffer containing 50 mM Tris–HCl pH 8.0, 250 mM NaCl, 2 mM β-ME and 5% glycerol by sonication. The cell debris was separated from the lysate by spinning at 15,000 rpm at 4 °C. 50 ml of lysate was incubated with 2 ml of Nickel-NTA His-select resin (Sigma-Aldrich, Inc.) for 1 h. Non-specific binding to the resin was minimized by extensive passage of a buffer containing 7 mM Imidazole. The protein was eluted by a gradient of Imidazole concentration ranging from 7 mM to 200 mM (in 50 mM Tris pH 8.0, 250 mM NaCl, 2 mM β-ME and 5% glycerol). The purified protein fractions were concentrated and loaded onto a Sephacryl S-200 column (GE Healthcare) for size exclusion chromatography in a buffer containing 25 mM Tris–HCl pH 8.0, 250 mM NaCl and 5% glycerol. The purified protein was concentrated using an amicon centrifugal concentrator (Millipore) and flash frozen in 7% glycerol to be stored at −80 °C. The molar extinction coefficients (ε at 280 nm) of 41,370 M− 1 cm− 1, 55,350 M− 1 cm− 1 and 98,210 M− 1 cm− 1 were used to estimate the concentrations of the D1, D2 and the D1–D2 protein constructs respectively.

2.3. Substrate peptides The peptide substrates GP1 (KRGERpYRMALL), GP2 (LMRVIpYWLRKR), GP3 (PKVHRpYAPINQ) and GP4 (KNSFVpYQKLSE) were identified from the GP150 protein, a known PTP99A substrate [8]. The Insulin Receptor Peptide (TRDIpYETDYYRK), Cuticle peptide (TAEPDpYGALYE), Nervous fingers (VIGDpYVCRLCK) and Abelson peptide (RDDTpYTAHAG) were designed with inputs from the Peptide Mine server [9]. All peptides were synthesized by GL Biosciences, China.

2.4. Assays for phosphatase activity The generic substrate para-NitroPhenyl Phosphate (pNPP) was used to monitor the phosphatase activity in different constructs of PTP99A. All reactions were carried out in CGH (Citrate, Glycine and HEPES) buffer as reported earlier [7]. For phosphatase assays with phosphopeptide substrates, increasing concentrations of the peptide (0.1–2.0 mM) were incubated with 0.2 μM of protein in CGH buffer pH 6.5 containing 5 mM DTT, at 25 °C for 30 min. For titration studies, 1 μM of D1 was incubated with increasing concentrations (0–20 μM) of D2 in CGH buffer pH 6.5 with 5 mM DTT on ice for 1 h. Enzyme assays were performed using 1–25 mM pNPP or 0.1–2.0 mM phosphopeptide substrates. The nonlinear regression module of Sigma-Plot software (Omega Scientific Inc.) was used to fit the data with different kinetic models. 2.5. Thermal denaturation assays Thermal unfolding was monitored using circular dichroism (CD) spectroscopy. These experiments were performed at 2.0 μM protein concentration using a 1 mm cuvette. The change in the mean residue ellipticity at 222 nm was monitored as a function of temperature ranging from 10 to 95 °C at a rate of 1 °C/min in a Jasco-J715A spectropolarimeter. All experiments were performed in a buffer containing 25 mM HEPES pH 7.0, 150 mM NaCl and 2 mM DTT. The averaging time was 3 s and each point was an average of 3 independent readings. The reversibility of unfolding was accessed by decreasing the temperature from 95 to 10 °C at 1 °C/min. As the unfolding was found to be irreversible, aggregation studies were performed to monitor differences in the temperature of melting (Tm) and aggregation (Tagg). These measurements provided a basis to determine the limiting case for the Lumry–Eyring model of protein thermal denaturation. For aggregation studies, the sample contained 2 μM of protein in 50 mM HEPES pH 7.0, 150 mM NaCl and 2 mM DTT. The protein solution was monitored for scattering between 340 and 700 nm in a UV–visible spectrophotometer. Scattering was measured at 403 nm with an increase in temperature from 10 to 95 °C at the rate of 1 °C/min. The turbidity of the solution was calculated as τ = 2.303 (absorbance at 403 nm)/l, where l is the path length of the cuvette [7]. To measure the rate of aggregation, the scattering from 2 μM of protein at 70 °C was monitored at 403 nm as a function of time. 2.6. The Lumry–Eyring model of thermal denaturation Thermal unfolding monitored by CD spectroscopy showed that the thermal denaturation of all PTP99A constructs is irreversible. Aggregation, measured by the turbidity of the samples at 403 nm, showed that all PTP99A constructs denature with an increase in temperature following the limiting case of the Lumry–Eyring model of thermal denaturation [10,11]. Briefly, the thermal denaturation of PTP99A constructs could be defined by N↔U→UN →A

where N, U, UN and A represent the native, unfolded, unfolded nuclei and aggregated forms of the protein. The protein (N) first unfolds to a reactive unfolded state U, which then interacts with other unfolded monomers to form a nucleus UN that irreversibly aggregates to form A. For a monomer, the reversible unfolding transition is defined by the equilibrium constant KU ¼ ½U=½N

ð1Þ

L.L. Madan et al. / Biochimica et Biophysica Acta 1824 (2012) 983–990

whereas the irreversible kinetic constant is defined by dA=dt ¼ k1 ½U:

ð2Þ

The irreversible step follows an Arrhenius dependence at a characteristic temperature T* such that   −1 k1 min ¼ exp½−E=Rð1=T–1=TÞ:

ð3Þ

The data on the PTP99A protein constructs followed the limiting case of this model described by Sanchez-Ruiz [12] where T* ≥ Tm such that unfolding occurs before the protein is deactivated. The T* value was obtained from aggregation studies (also referred to as Tagg), while the Tm was noted from the CD measurements. Also, as described by Zale and Klibanov [13], the observed deactivation constant as measured from aggregation studies is related to the kinetic constant of the k1 by

985

log and saturation phase of the time course to be taken together and treated as a pseudo-first order reaction f ¼ Ao f1− exp½−kobs ðt−to Þg

ð11Þ

where Ao is the maximum value of the turbidity, to is the time at the end of the log phase and k1 is a pseudo first order rate constant. kobs includes the rates kN and kA described in Eq. (8) and is related to the equilibrium constant of thermal unfolding by Eq. (4). The observed rate constant (kobs) obtained from Eq. (11) and the equilibrium constant obtained from Eq. (6) were used to calculate the rate of aggregation k1 from Eq. (4). 2.7. Sequence and conformational analysis

In this simplistic model, the ellipticity reading of the aggregated form is treated to be equivalent to that of the unfolded protein. The terms fN, fU and fA represent the fractional population of the native, unfolded and aggregated protein. For the aggregation studies, the assumption is that fN + fU + fA = 1. Furthermore, as fU = fA, fN + fA = 1 for the CD measurements. The free energy of a two state thermal unfolding of a monomer at a given temperature is described by

Sequences corresponding to bi-domain RPTPs from humans, fly and worm were collated from the PTP database [1]. Sequence alignment and conservation were visualized using the Jalview software [15]. Sequences of individual PTP99A domains were analyzed for their disorder-propensity using IUPred and Disopred algorithms [16,17]. The PTP99A sequence was modeled on the closest homologue PTPγ (PDB: 2NLK [18]; 57% sequence identity) using Modeller version 7.0.7. A change in the accessible surface area upon unfolding was calculated from the model using the Auton and Bolen algorithm [19]. A comparison of the allosteric pocket and the interdomain-interface was examined by superposing the PTP1B structures (PDB: 1SUG [20] and 1T48 [21]) on the PTP99A homology model. All representations of macromolecular structures were made using PyMOL (Delano Scientific Inc.).

ΔG ¼ −RT lnðKF Þ

3. Results and discussion

kobs ¼ k1 =ð1 þ ð1=KU ÞÞ:

ð4Þ

ð5Þ

where R is the gas constant (1.98 cal mol − 1), T is the absolute temperature and KF is the equilibrium constant of folding described as KF ¼ 1=KU :

ð6Þ

The free energy of folding at a particular temperature is described by the Gibbs–Helmholtz equation as ΔGðTÞ ¼ ΔHm ð1−T=Tm Þ−ΔCP ððTm –TÞ þ T lnðT=Tm ÞÞ

ð7Þ

where Tm is the temperature when the fraction of unfolded protein (α) is 0.5 (fN = fU). The enthalpy and specific heat capacity at Tm are denoted by ΔHm and ΔCP respectively. The rate of aggregate formation is described by the equation d½A=dt ¼ kU ½N þ kN1 ½U½UN−1  þ kN2 ½UN  þ kA ½U½Un

ð8Þ

where kU, kN1, kN2 and kA are the rate constants for the unfolding, nucleation and aggregation steps respectively. The overall time-course for aggregation can be represented by a stretched exponential function [14]:  n
f ¼ A 1− exp −Bt

ð9Þ

where A is the limiting value of the signal (in this case maximum turbidity), B is the overall rate constant (includes kU, kN and kA) and n is a measure of the sigmoidal degree of the curve. The half time of the reaction is defined by 1=n

t1=2 ¼ ð ln2=BÞ

:

ð10Þ

At the end of the log phase when the nuclei for the aggregates are formed, the rate of the reaction depends on the concentration of the unfolded monomer which attaches to the nuclei. This allows for the

3.1. Purification and oligomeric status of PTP99A constructs All recombinant protein constructs (the D1 and D2 domains, the D1– D2 bi-domain enzyme and the mutant enzymes D1–D2 K990A and D2 K990A) were purified by immobilized metal affinity chromatography (Fig. 1; Supplementary Figure I). We note that while the individual domains (D1 and D2 constructs) form homo-dimers, the D2 K990A is a monomer under oxidizing conditions. The D1–D2 protein was found to be prone to aggregation in the absence of a reducing agent. The D1– D2 K990A mutant, on the other hand, adopts apparent trimeric arrangements in solution in the absence of a reducing agent. Size exclusion chromatography experiments in the presence of a reducing agent (DTT) suggested that all the recombinant protein constructs of PTP99A were monomers (Supplementary Figures I and II). All biochemical experiments were performed in the presence of DTT as a reducing agent. 3.2. Phosphatase assays and the non-essential partially mixed model of enzyme activation Phosphatase assays using pNPP as well as phospho-peptide substrates suggest that the entire phosphatase activity of the PTP99A protein resides in the D1 PTP domain [6]. Interestingly, the catalytic activity of the D1 domain alone was much less when compared to the D1–D2 protein construct. This observation suggested a role for the inactive D2 domain in modulating catalytic activity. We note that addition of the inactive D2 domain to the purified D1 domain sample leads to an increase in the Vmax and Kcat and a decrease in the Km values when compared to the D1 domain in isolation (Fig. 2a, c and d and Supplementary Table I). Addition of a 20 fold molar excess of the D2 domain to the D1 domain increased the Vmax of the D1 catalyzed reaction from 22.02 ± 0.89 × 10 − 2 to 61.89 ± 0.42 × 10 − 2 μmol/min/mg. Correspondingly, the Km decreased from 5.69 ± 0.23 to 0.57 ± 0.03 mM. This activation by the D2 domain was found to have an effect on the D1 domain alone. Neither D1–D2

986

L.L. Madan et al. / Biochimica et Biophysica Acta 1824 (2012) 983–990

Fig. 1. The modular organization of PTP99A. (a) Domain organization of PTP99A. (b) Schematic of the expression constructs used in this study. (c) Sequence conservation at the D1–D2 interface in bi-domain PTPs (collated from the fly, human and worm genome sequences). (d) Homology model of PTP99A (based on PTPγ) and its superposition with the crystal structure of PTP1B. The so-called allosteric pocket in PTP1B is highlighted (yellow). (e) Spatial proximity of the allosteric pocket to the WPD loop dictates the movement of this active-site loop.

nor the D1–D2 K990A mutant proteins demonstrated activation upon the addition of the D2 domain (Supplementary Figure III). Another interesting observation is that the PTP99A D2 alone functions as an activator for the D1 domain of PTP99A. This was seen by the effect on PTP99A D1 by the addition of PTP69D D2. PTP69D has similar sequence features as PTP99A; K990 is an arginine in PTP69D (Fig. 1c, Supplementary Figure IV). PTP69D D2, however, does not activate PTP99A D1 (Supplementary Figure V). The activation of the D1 domain by D2 could be explained by the non-essential partially mixed model of enzyme activation [22]. This is represented by a schematic in Fig. 3 (for a more detailed version, please refer to Supplementary Figure VI). The D1–D2 construct represents the case where a non-essential activator is tethered to an enzyme. The activator (A, in this case the D2 domain) enhances the equilibrium constant of the association of the enzyme (E, in this case D1) and its substrate (S) by a factor α. It also affects the turnover rate of the enzyme substrate (ES) complex to enzyme (E) and product (P) by a factor β. For all the substrates examined in this study, β > 1

and α b 1 (Table 1). The rate of the phosphatase reaction catalyzed by the D1–D2 bi-domain construct (Vmax app) is thus β times more than that catalyzed by the D1 domain alone (Vmax). This feature, along with the characteristic inverse hyperbolic plot for Km/Vmax versus activator concentration (Fig. 2b), is consistent with the nonessential partially mixed model of enzyme activation. The small molecule substrate pNPP mimics the phosphotyrosine (pY) residue and owing to its smaller size is more easily accessible to the PTP active site. The phosphopeptides, however, are limited by the physiochemical nature of the amino acids flanking the pY residue in binding to the PTP active site [23]. pNPP is thus a better reporter of changes in the catalytic pocket of the PTP domain independent of the changes on the surface of the active site cleft. This aspect is reflected in the difference of values obtained for the constant β for phosphopeptides as opposed to pNPP (Table 1). The value of β as measured for the peptide substrates was found to be much smaller (2.83 for GP2 and 28.27 for GP1) than that for the small molecule pNPP (β = 113.36).

L.L. Madan et al. / Biochimica et Biophysica Acta 1824 (2012) 983–990

987

Fig. 2. Phosphatase activity measurements for different constructs of PTP99A. (a) Phosphatase activity of the D1 domain with increasing concentrations of the D2 domain (●) and the D2 K990A mutant ( ) using pNPP as substrate. The catalytic activity of D1 domain alone (○) is plotted as a control. (b) Reverse hyperbolic plot for the non-essential partially mixed activation of D1 domain by its cognate D2 domain. (c and d) Michaelis–Menten and Lineweaver–Burk plots for the activation of D1 domain (1 μM concentration) by its D2 domain. 0.0, 1.0, 5.0, 7.5, 10.0 and 20.0 molar excess of D2 domain over D1 domain were used to calculate the Vmax and Km of the D1 catalyzed reaction using pNPP as substrate. The values obtained are tabulated in Supplementary Table I. (e and f) Effect of the K990A mutation on the catalytic parameters (Vmax and Km) for peptide substrates.

3.3. Thermal unfolding and aggregation Both Tm and Tagg of the D1 domain were found to be ca 20 °C less than either the D2 domain or the D1–D2 bi-domain protein construct suggesting that D2 substantially contributes to its thermodynamic stabilization (Fig. 4; Tables 2 and 3). A higher change in the specific heat capacity on unfolding (ΔCp) along with a ca 2000 cal/mol less

free energy of unfolding (ΔG) supports the observation that the D2 domain is more stable compared to the catalytically active D1 domain (Tables 2 and 3). This is also consistent with thermal aggregation measurements that show a difference of 38 s in the t1/2 of aggregation between the D1 and D2 domains. The two state unfolding of the D1–D2 protein and the agreement between the estimates of the free energy of unfolding and enthalpy change on unfolding are consistent with

988

L.L. Madan et al. / Biochimica et Biophysica Acta 1824 (2012) 983–990 Table 2 Thermodynamic parameters for the PTP99A protein constructs. D1

D2

D1–D2

D1–D2 K990Aa

Tm (°C)

Fig. 3. The non-essential partially mixed model of enzyme activation. Please refer to Supplementary Figure VI for a more detailed representation. This schematic is adapted from [22].

31.32 ± 0.53 55.42 ± 0.05 53.91 ± 0.11 39.66 ± 0.23 71.41 ± 1.39 40.66 ± 0.17 57.21 ± 0.49 56.28 ± 0.98 49.29 ± 0.71 Tagg (°C) Change in enthalpy 28.85 ± 1.02 80.88 ± 1.42 67.57 ± 1.78 54.18 ± 3.03 ΔHm (kcal/mol) 22.25 ± 2.96 Change in heat capacity 1.31 ± 0.68 3.49 ± 0.16 2.44 ± 0.15 1.78 ± 0.62 ΔCp (kcal/mol/K) 0.62 ± 0.02 a The two values of Tm, ΔCp and ΔHm in the case of the D1–D2 K990A mutant correspond to the two transitions observed in the thermal unfolding profiles for this protein. These parameters correspond to the reversible unfolding step of the Lumry–Eyring nucleated polymerization model.

the suggestion that the stability of the D1–D2 protein is enhanced by the presence of the D2 domain.

3.4. Analysis of the D1–D2 domain interface and the role of Lys 990 In an effort to understand the interaction between the D1 and D2 PTP domains, a sequence alignment was performed for all dual domain PTPs to identify conserved residues at the interface of the two PTP domains (Fig. 1c). While the linker between the two PTP domains is not conserved across homologues, the C-terminal regions of both the D1 and D2 domains (structurally located at the inter-domain interface) revealed conserved glutamate and lysine residues. Based on the observations from the single domain constructs and previous reports where the removal of residues from the carboxyl terminus of D2 was shown to alter the activity of its D1 domain [5], Lys 990 of PTP99A was mutated to an alanine residue to evaluate the role of this residue in inter-domain interactions. We note that this point mutation at the D1–D2 interface affected the phosphatase activity of the D1–D2 construct by lowering the reaction rates for all substrates (Fig. 2e and f and Supplementary Table II). These kinetic parameters are closer to those observed for the D1 domain in isolation. Indeed, addition of the D2 K990A mutant protein to the D1 domain could not reproduce the activation profile (Fig. 2a). Thermal denaturation studies showed that this point mutation disrupts the cooperativity in the unfolding of the two domains in the D1–D2 protein construct (Fig. 4, Tables 2 and 3). In fact, the denaturation of the D1–D2 K990A mutant protein suggests a three state unfolding pattern with the two transitions closer to those obtained for the individual domains of PTP99A.

A comparison between the biochemical features of the bi-domain (D1–D2) PTP99A protein construct, the catalytically active D1 domain and the inactive D2 PTP domain reveals that the D2 domain both activates and stabilizes the D1 PTP domain. It thus appears likely that the D2 domain could modulate signaling events mediated by PTP99A. A sequence analysis of the D1 and D2 domains of PTP99A for disorder propensities suggests that the D2 domain has a more stable core despite being catalytically inactive (Supplementary Figure VII). The mean residue disorder score for D2 was seen to be ca 35% less than that of D1. We also note that the net change in accessible surface area upon unfolding was found to be 30% more for the D2 domain, which is consistent with a larger change in specific heat capacity observed in the thermal denaturation experiments. A single point mutation (K990A) at the D1–D2 domain interface uncouples the interactions between the two PTP domains of PTP99A. The D1–D2 K990A mutant protein is also less active than the native enzyme. A superimposition of the PTP99A model with the structure of the extensively characterized PTP, PTP1B, revealed that the D1–D2 interface (and the conserved Lys990) superposed on to the so-called allosteric pocket of single domain PTPs [21] (Fig. 1d and e). While the structural location of Lys990 in the allosteric site of the PTP domain is unambiguous, the chemical environment is different between the PTP99A interface and PTP1B. In particular, this site, located between helices α3 and α6 is primarily hydrophobic (Phe196 and Phe280; PDB: 1T4J) in PTP1B. In either case, this allosteric pocket is present just behind the active site and controls the

Table 1 Parameters for the non-essential partially mixed model of enzyme activation of the catalytic D1 domain by the inactive D2 PTP domain. Substrate

Vmaxa (μmol/min/mg)

app b Vmax (μmol/min/mg)

βc

Km (μM)

app Km (μM)

αd

pNPP GP1 KRGERpYRMALL GP2 LMRVIpYWLRKR GP3 PKVHRpYAPINQ GP4 KNSFVpYQKLSE Insulin receptor TRDIpYETDYYRK Cuticle TAEPDpYGALYE Nervous fingers VIGDpYVCRLCK AbelsonRDDTpYTAHAG

0.22 ± 0.01 1.07 ± 0.09 0.95 ± 0.23 1.61 ± 0.13 1.95 ± 0.23 2.37 ± 0.16 1.26 ± 0.04 4.42 ± 0.14 0.83 ± 0.05

24.89 ± 0.52 30.25 ± 2.93 2.69 ± 0.28 12.42 ± 1.49 12.72 ± 1.25 45.98 ± 0.83 17.10 ± 0.66 8.89 ± 0.38 2.59 ± 0.07

113.36 28.27 2.83 7.72 6.52 19.40 13.57 4.27 3.12

5.60 ± 0.23 × 103 367.82 ± 34.32 873.67 ± 210.68 428.24 ± 36.24 74.47 ± 8.89 1017.34 ± 161.84 309.04 ± 33.32 1361.68 ± 94.68 121.69 ± 26.61

2.27 ± 0.17 × 103 115.15 ± 11.18 367.08 ± 39.11 87.42 ± 13.2 54.54 ± 5.2 35.75 ± 2.08 112.42 ± 10.14 179.86 ± 15.02 16.29 ± 1.55

0.421 0.314 0.420 0.204 0.732 0.035 0.364 0.132 0.133

Please refer to Supplementary Figure III for a schematic description of the non-essential partially mixed model of enzyme activation. a Vmax, Km: Enzymatic constants obtained for the D1 domain (in the absence of the activator D2 domain). b app app Vmax , Km : Enzymatic activity constants obtained for the D1 domain in the presence of the activator. c ß: fold increase in the specific activity of D1 when tethered with the D2 domain. d α: fold decrease in Km of D1 in the D1–D2 bi-domain protein.

L.L. Madan et al. / Biochimica et Biophysica Acta 1824 (2012) 983–990

989

Table 3 Kinetic and thermodynamic parameters obtained for the aggregation of the PTP99A protein constructs. D1 Maximum turbidity τ403 nm (cm− 1) Degree of sigmoid behavior n Lag time to (s) Half time t½ (s) Pseudo first order rate constant kobs (× 10− 2) s− 1 Free energy of unfolding ΔG (kcal/mol) Equilibrium constant at 70 °C KU (×102) Rate constant k1 (×10− 3) s− 1

D2

D1–D2

D1–D2 K990A

0.46 ± 0.02

1.73 ± 0.06

2.12 ± 0.04

1.76 ± 0.05

1.46 ± 0.07

5.00 ± 0.38

8.58 ± 0.54

5.30 ± 0.32

22.0 ± 2.0

50.0 ± 2.0

46.0 ± 2.0

40.0 ± 2.0

49.97 ± 0.66

60.35 ± 0.25

60.67 ± 0.16

54.92 ± 0.18

3.86 ± 0.15

9.22 ± 0.22

5.71 ± 0.24

12.74 ± 0.45

− 6.76 ± 0.69

− 4.71 ± 0.36

− 4.28 ± 0.39

− 7.67 ± 0.35

70.70 ± 0.34

4.76 ± 0.29

2.74 ± 0.33

234.80 ± 0.52

38.61 ± 1.26

92.39 ± 4.53

57.31 ± 1.89

127.40 ± 4.33

These parameters are based on the Lumry–Eyring nucleated polymerization model.

Fig. 4. Thermal unfolding of PTP99A constructs. (a) Thermal unfolding of PTP99A constructs as measured by far UV CD spectroscopy. (b) Aggregation of the PTP99A constructs as monitored by light scattering at 403 nm. (c) Real time kinetics of aggregation at 70 °C for various PTP99A protein constructs. 2.0 μM of protein sample (D1 domain (○), D2 domain (□), D1–D2 protein (△), D1–D2 K990A mutant (▽) and externally reconstituted D1 and D2 domains (⋄)) was used in these experiments.

990

L.L. Madan et al. / Biochimica et Biophysica Acta 1824 (2012) 983–990

opening and closing of its WPD loop [1], directly affecting phosphatase activity.

[8]

4. Conclusions

[9]

To summarize, the biochemical characterization of PTP99A suggests an evolutionary optimization of catalytic activity in bi-domain PTPs. It thus appears likely that functional optimization in PTPs involves multiple, perhaps overlapping, roles for an associated inactive PTP domain encompassing substrate recognition, stabilization and modulation of catalytic activity.

[10]

[11] [12] [13]

Appendix A. Supplementary data [14]

Supplementary data to this article can be found online at http:// dx.doi.org/10.1016/j.bbapap.2012.05.001.

[15]

References

[16]

[1] J.N. Andersen, O.H. Mortensen, G.H. Peters, P.G. Drake, L.F. Iversen, O.H. Olsen, P.G. Jansen, H.S. Andersen, N.K. Tonks, N.P. Moller, Structural and evolutionary relationships among protein tyrosine phosphatase domains, Mol. Cell. Biol. 21 (2001) 7117–7136. [2] C.J. Desai, Q. Sun, K. Zinn, Tyrosine phosphorylation and axon guidance: of mice and flies, Curr. Opin. Neurobiol. 7 (1997) 70–74. [3] P. Johnson, H.L. Ostergaard, C. Wasden, I.S. Trowbridge, Mutational analysis of CD45. A leukocyte-specific protein tyrosine phosphatase, J. Biol. Chem. 267 (1992) 8035–8041. [4] H. Cho, R. Krishnaraj, M. Itoh, E. Kitas, W. Bannwarth, H. Saito, C.T. Walsh, Substrate specificities of catalytic fragments of protein tyrosine phosphatases (HPTP beta, LAR, and CD45) toward phosphotyrosylpeptide substrates and thiophosphotyrosylated peptides as inhibitors, Protein Sci. 2 (1993) 977–984. [5] M. Streuli, N.X. Krueger, T. Thai, M. Tang, H. Saito, Distinct functional roles of the two intracellular phosphatase like domains of the receptor-linked protein tyrosine phosphatases LCA and LAR, EMBO J. 9 (1990) 2399–2407. [6] L.L. Madan, S. Veeranna, K. Shameer, C.C.S. Reddy, R. Sowdhamini, B. Gopal, Modulation of catalytic activity in multi-domain protein tyrosine phosphatases, PLoS One 6 (2011) e24766. [7] L.L. Madan, B. Gopal, Addition of a polypeptide stretch at the N-terminus improves the expression, stability and solubility of recombinant protein tyrosine

[17]

[18]

[19]

[20]

[21]

[22] [23]

phosphatases from Drosophila melanogaster, Protein Expr. Purif. 57 (2008) 234–243. S.J. Fashena, K. Zinn, Transmembrane glycoprotein gp150 is a substrate for receptor tyrosine phosphatase DPTP10D in Drosophila cells, Mol. Cell. Biol. 17 (1997) 6859–6867. K. Shameer, L.L. Madan, S. Veeranna, B. Gopal, R. Sowdhamini, PeptideMine—a webserver for the design of peptides for protein-peptide binding studies derived from protein–protein interactomes, BMC Bioinforma. 11 (2010) 473. J.M. Andrews, C.J. Roberts, A Lumry–Eyring nucleated polymerization model of protein aggregation kinetics: 1. Aggregation with pre-equilibrated unfolding, J. Phys. Chem. B 111 (2007) 7897–7913. R. Lumry, H. Eyring, Conformation changes of proteins, J. Phys. Chem. 58 (1954) 110–120. J.M. Sanchez-Ruiz, Theoretical analysis of Lumry–Eyring models in differential scanning calorimetry, Biophys. J. 61 (1992) 921–935. S.E. Zale, A.M. Klibanov, On the role of reversible denaturation (unfolding) in the irreversible thermal inactivation of enzymes, Biotechnol. Bioeng. 25 (1983) 2221–2230. A.M. Morris, M.A. Watzky, R.G. Finke, Protein aggregation kinetics, mechanism, and curve-fitting: a review of the literature, Biochim. Biophys. Acta 1794 (2009) 375–397. A.M. Waterhouse, J.B. Procter, D.M.A. Martin, M. Clamp, G.J. Barton, Jalview Version 2—a multiple sequence alignment editor and analysis workbench, Bioinformatics 25 (2009) 1189–1191. Z. Dosztanyi, V. Csizmok, P. Tompa, I. Simon, The pairwise energy content estimated from amino acid composition discriminates between folded and intrinsically unstructured proteins, J. Mol. Biol. 347 (2005) 827–839. J.J. Ward, J.S. Sodhi, L.J. McGuffin, B.F. Buxton, D.T. Jones, Prediction and functional analysis of native disorder in proteins from the three kingdoms of life, J. Mol. Biol. 337 (2004) 635–645. A.J. Barr, E. Ugochukwu, W.H. Lee, O.N. King, P. Filippakopoulos, I. Alfano, P. Savitsky, N.A. Burgess-Brown, S. Muller, S. Knapp, Large-scale structural analysis of the classical human protein tyrosine phosphatome, Cell 136 (2009) 352–363. M. Auton, D.W. Bolen, Application of the transfer model to understand how naturally occurring osmolytes affect protein stability, Methods Enzymol. 428 (2007) 397–418. A.K. Pedersen, G.G. Peters, K.B. Moller, L.F. Iversen, J.S. Kastrup, Water-molecule network and active-site flexibility of apo protein tyrosine phosphatase 1B, Acta Crystallogr. D: Biol. Crystallogr. 60 (2004) 1527–1534. C. Wiesmann, K.J. Barr, J. Kung, J. Zhu, D.A. Erlanson, W. Shen, B.J. Fahr, M. Zhong, L. Taylor, M. Randal, R.S. McDowell, S.K. Hansen, Allosteric inhibition of protein tyrosine phosphatase 1B, Nat. Struct. Mol. Biol. 11 (2004) 730–737. I.H. Segel, Enzyme Kinetics: Behavior and Analysis of Rapid Equilibrium and Steady State Enzyme Systems, Wiley, New York, London, 1975. Z. Jia, D. Barford, A. Flint, N. Tonks, Structural basis for phosphotyrosine peptide recognition by protein tyrosine phosphatase 1B, Science 268 (1995) 1754–1758.