PEDS Advance Access originally published online on February 3, 2006
Protein Engineering Design and Selection 2006 19(4):155-161; doi:10.1093/protein/gzj014
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Stabilization of the autoproteolysis of TNF-alpha converting enzyme (TACE) results in a novel crystal form suitable for structure-based drug design studies
Department of Structural Chemistry, Schering-Plough Research Institute, 2015 Galloping Hill Road, Kenilworth, NJ 07033, USA
1 To whom correspondence should be addressed. E-mail: brian.beyer{at}spcorp.com
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionReferences |
|---|
The crystallization of TNF-alpha converting enzyme (TACE) has been useful in understanding the structure–activity relationships of new chemical entities. However, the propensity of TACE to undergo autoproteolysis has made enzyme handling difficult and impeded the identification of inhibitor soakable crystal forms. The autoproteolysis of TACE was found to be specific (Y352–V353) and occurred within a flexible loop that is in close proximity to the P-side of the active site. The rate of autoproteolysis was found to be proportional to the concentration of TACE, suggesting a bimolecular reaction mechanism. A limited specificity study of the S1' subsite was conducted using surrogate peptides and suggested substitutions that would stabilize the proteolysis of the loop at positions Y352–V353. Two mutant proteases (V353G and V353S) were generated and proved to be highly resistant to autoproteolysis. The kinetics of the more resistant mutant (V353G) and wild-type TACE were compared and demonstrated virtually identical IC50 values for a panel of competitive inhibitors. However, the kcat/Km of the mutant for a larger substrate (P6 – P6') was
5-fold lower than that for the wild-type enzyme. Comparison of the complexed wild-type and mutant structures indicated a subtle shift in a peripheral P-side loop (comprising the mutation site) that may be involved in substrate binding/turnover and might explain the mild kinetic difference. The characterization of this stabilized form of TACE has yielded an enzyme with similar native kinetic properties and identified a novel crystal form that is suitable for inhibitor soaking and structure determination.
Keywords: autoproteolysis/protein engineering/soakable crystal form/structure-based drug design/TACE
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Tumor necrosis factor-
(TNF) is a membrane bound pro-inflammatory cytokine produced by activated monocytes and macrophages in a highly regulated manner (Bemelmans et al., 1996
; Shanahan and St Clair, 2002). Upon release, TNF participates in the recruitment of inflammatory cells and the stimulation and production of other mediators of pain and cartilage breakdown. An elevated level of circulating TNF has been implicated in several acute and chronic disease states and the reduction of these levels is currently a target for drug development (Shanahan and St Clair 2002; Mikuls and Weaver, 2003).
TNF is synthesized as a membrane anchored 26 kDa precursor protein that is processed by cleavage of the Ala76–Val77 peptide bond to its mature 17 kDa active form (Bemelmans et al., 1996). This processing is performed by a 85 kDa zinc metalloproteinase TNF converting enzyme (TACE) (Black et al., 1997). TACE or ADAM17 is a member of the ADAM family of proteases that contain both disintegrin and metalloprotease (catalytic) domains (Moss et al., 1997). If the cleavage site of TNF is mutated, the protein is not released from cells verifying that processing is required for TNF secretion (Perez et al., 1990; Utsumi et al., 1993).
Because neutralization of TNF has been shown to be therapeutically relevant (Tracey and Cerami, 1994), the inhibition of the catalytic domain of TACE may be useful in the treatment of diseases where release of TNF has been identified as the causative factor. Involvement of this cytokine has been validated in disease states such as arthritis and Crohn's disease and implicated in diverse neuroimmunological pathologies such as multiple sclerosis, Alzheimers and stroke. The association of TACE with disease has spurred the discovery of inhibitors (hydroxamates) that strongly inhibit TACE and have been shown to inhibit TNF secretion from a variety of cell types (Chen et al., 2002; Levin et al., 2002).
Since TACE is a target for drug development, expression and purification of the TACE catalytic domain has been utilized to determine the three-dimensional structure of the enzyme with a bound inhibitor (Maskos et al., 1998). However, the reported crystal form was found to be unsuitable for iterative structure-based drug design studies and therefore warranted a search for a novel crystal form. High-throughput crystallization screening procedures were complicated by the tendency of TACE to undergo autoproteolysis at high concentrations. Autoproteolysis was found to be specific and to occur at a site that produces two 15 kDa fragments.
In an effort to create a more stable, but structurally relevant form of TACE, the internal cleavage site was identified and the specificity of the enzyme was probed for suitable substitutions that would block autoproteolysis. As a result, two mutants were made with single amino acid substitutions adjacent to the internal cleavage site and their effect on the activity and the stability of the enzyme was studied. Screening of this mutant protein complexed with inhibitor yielded a crystal form of TACE that is suitable for inhibitor soaking and consequently structure-based drug design work.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Reagents
Chemicals were obtained from Sigma (St Louis, MO) and Hampton Research (Laguna Niguel, CA); reagents for peptide synthesis were from Aldrich (Milwaukee, WI), Novabiochem (San Diego, CA), Applied Biosystems (Foster City, CA) and Perseptive Biosystems (Framingham, MA). Restriction enzymes for cloning were obtained from New England Biolabs (Beverly, MA), and Bac-to-Bac baculovirus expression kit was purchased from Invitrogen (Carlsbad, CA). Viral titers were determined using the BacPAC Baculovirus Rapid Titer Kit from Clonetech (Palo Alto, CA). Primers were synthesized by Genosys (Woodlands, TX), mutagenesis was performed using QuickchangeTM kit from Stratagene (La Jolla, CA) and sequencing was provided by GeneWiz (New York, NY). P6 buffer exchange columns were from Bio-Rad (Hercules, CA). Peptide chromatography was carried out on a Perseptive Biosystems Integral (Wellesley, MA). Kinetic analysis was conducted using GraphPad Prism 3.0 purchased from GraphPad Software (San Diego, CA). All crystallographic supplies were purchased from either Hampton research or Emerald Biostructures (Bainbridge Island, WA). JMV 390 was purchased from Neosystem (France) and TAPI-2 from EMD Biosciences (San Diego, CA).
Cloning of wt-TACE and TACE mutants
A DNA construct encoding the pre-pro-cat domains of TACE (1-477) was prepared with the C-terminal linker Gly-Ser-(His)6. Site-directed mutagenesis was utilized to prevent N-linked glycosylation by changing Ser266 to an Ala and Asn452 to Gln. The construct was amplified using two PCR primers that added BamH1 (5'-CGCGGATCCATGAGGCAGTCTCTCCTATTCCTG-3') and Kpn1 (5'-CCGGCCTACCTTAGTGATGGTGATGATGGTGGGATC-3') sites to the 5' and 3' ends, respectively. The purified PCR fragment was digested with BamH1 and Kpn1 and subcoloned into the pFastBac1 vector provided in the Bac-to-Bac baculovirus expression system.
The TACE mutants (V353G, V353S) were generated using the QuickChangeTM kit with the native TACE pFastBac1 vector as a template and the following complementary mutagenic primers: (i) V353G, 5'-GGAACTCTTGGATTAGCTTATGGAGGCTCTCCCAGAGCAAAC-3'; (ii) V353G, 5'-GTTTGCTCTGGGAGAGCCTCCATAAGCTAATCCAAGAGTTCC-3'; (iii) V353S, 5'-GGAACTCTTGGATTAGCTTATAGCGGCTCTCCCAGAGCAAAC-3'; (iv) V353S, 5'-GTTTGCTCTGGGAGAGCCGCTATAAGCTAATCCAAGAGTTCC-3'.
The mutagenesis was performed in two steps as previously described (Wang and Malcolm, 2002). In the first stage, two extension reactions were performed in separate tubes with one containing the forward primer and the other containing the reverse primer. After two cycles, the two reactions were mixed and the standard QuickChange mutagenesis procedure was carried out for additional 18 cycles. Following amplification, the parental strand was digested with 1 U of Dpn1 for 1 h and an aliquot was transformed into DH5-alpha cells. All the vectors were sequence confirmed.
Production of recombinant baculovirus
Recombinant baculovirus was produced by utilizing the Bac-to-Bac expression system following the protocols for transposition, isolation and transfection of recombinant bacmid DNA into Sf9 cells for production of viral particles. The virus was amplified to the P3 generation and was titered using the BacPAC Baculovirus Rapid Titer Kit.
Expression and purification of the catalytic domains of TACE and TACE mutants
Logarithmically growing (2 x 106 cells/ml) Trichoplusia Ni cells (High-5TM cells) were infected with amplified baculovirus at a MOI = 1.0 and grown at 27°C for 48–60 h. Secreted TACE was isolated from the cell culture media after clarification by centrifugation. The pooled supernatants were concentrated 10-fold and the buffer exchanged into 25 mM HEPES and 0.15 M NaCl, pH 7.5, by diafiltration. To the desalted supernatant, 4-aminophenylmerucuric acetate (APMA) was added to 20 µM, lauryl maltoside to 0.05% and imidazole to 25 mM and applied to a Ni-NTA column. The column was washed with 25 mM imidazole in buffer A (50 mM HEPES, 10% glycerol, 0.3 M NaCl and 0.1% NOG, pH 7.5) until a stable baseline was achieved and the protein was eluted with 250 mM imidazole in buffer A. The eluted protein was diluted to 0.1 mg/ml and dialyzed overnight against 25 mM Tris, pH 7.5, 20 µM APMA to digest excess pro-domain. The protein was collected and adjusted to 0.15 M NaCl, concentrated to 10 mg/ml, and applied to a Superdex-75 gel filtration column equilibrated with 25 mM Tris–HCl and 0.2 M NaCl, pH 7.5. Fractions corresponding to the monomer of TACE were pooled, and stored at 4°C. For the screening of crystallization conditions, TACE was concentrated to 15 mg/ml and the salt was removed by passage through a P6 column equilibrated with 25 mM Tris–HCl, pH 7.5. Desalted protein was then immediately complexed with the TACE inhibitor TAPI-2 at a 1 : 1.5 molar ratio. Final purities and molecular weight characterization were done by SDS–PAGE and mass spectrometry.
Autoproteolysis experiments
To identify the autoproteolytic site of TACE, an aliquot of TACE at 10 mg/ml in TBS buffer was exchanged over P6 desalting columns equilibrated in either: (i) 25 mM Tris–HCl, pH 7.5 and 0.15 M NaCl, (ii) 25 mM Tris–HCl, pH 7.5 or (iii) 25 mM Tris–HCl, pH 7.5 + 1.0 mM TAPI-2. The protein was incubated at room temperature for 6 h and analyzed by SDS–PAGE. The protein was transferred to an Immobilon membrane and briefly stained in coomassie to visualize the bands. The intact protein and the two 15 kDa degradation products were identified by N-terminal sequencing for 10 cycles using an Applied Biosystems Procise 494 N-Terminal protein sequencer.
To compare the stabilities of the wild-type and mutant TACE protease stock solutions (10 mg/ml in TBS at 4°C), aliquots of each stock solution were buffer exchanged as listed above. One microliter aliquots were then removed at 3 h and 14 days following buffer exchange. The 15 kDa degradation products were then resolved by SDS–PAGE and the bands visualized by Coomassie staining.
Intermolecular proteolysis of TACE was analyzed as follows: 100 µM (3.0 mg/ml) wt-TACE in TBS was desalted using Bio-Rad P-6 spin columns equilibrated with 25 mM Tris–HCl, pH 7.5 The protein was immediately diluted to 10 and 1 µM in 25 mM Tris–HCl, pH 7.5 and incubated at room temperature. At periodic intervals, samples were removed and NaCl was added from a 3 M stock solution to 0.15 M final concentration to prevent further autoproteolysis. The quenched samples were stored at 4°C and enzyme activity on all samples was determined simultaneously by HPLC as previously described. Decay curves were generated by comparing the remaining activity at a given time (At) with the initial activity (A0). Curves were fit to a simple second-order rate equation At = A0/(1 + A0·k·t), where t is time and k is the rate of inactivation. At each time point of the 10 and 100 µM concentrations, 5 µg aliquots were removed for analysis by SDS–PAGE. The Coomassie blue stained gels were scanned and the remaining intact TACE was quantified by densitometry using a Bio-Rad ChemiDoc. Decay curves were generated as above.
Kinetic characterization of TACE
TACE activity was measured by incubating 25 µM TNF-alpha substrate (Ac-SPLAQA*VRSSSR-COOH) in 25 mM Tris–HCl, pH 7.5, in a volume of 100 µl, with enzyme (final concentration 2 nM) for 30 min at room temperature. The reaction was quenched with an equal volume of 1% TFA and the reaction products were resolved by HPLC on a Zorbax C-8 column (4.6 x 50 mm) using a 15–35% water/acetonitrile gradient in the presence of 0.1% TFA. The area of the P-side product was measured at 215 nm and was compared with a standard curve produced by synthetic P-side product peptide. The concentration of the substrates and product standards were verified by amino acid analysis. For kcat/Km measurements, purified 2 nM TACE or 10 nM mutant TACE was incubated with substrate and 100 µl aliquots were removed at 0, 5, 10, 20, 30, 45 and 60 min in order to obtain initial velocities. kcat/Km calculations were performed by assuming 100% enzyme activity.
Inhibition assays were performed using a newly designed fluorescent peptide K(Mca)SPLAQA-VRSSSRK(Dnp)-NH2 [derived from previous reports (Jin et al., 2002)] in 96-well plates. Dilutions of inhibitor were added to 25 µM substrate and the reaction was initiated with the addition of 2 nM wild-type TACE or 5 nM V353G TACE. The fluorescence was measured every 30 s for 30 min (
em = 340, ex = 380) to obtain initial velocities. IC50's (Table I) were calculated using Prizm Graph Pad software.
|
Crystallization and structure determination of V353G TACE
The TACE mutant V353G TACE could be crystallized under similar conditions as wild-type TACE. V353G TACE was concentrated to 15 mg/ml in 150 mM NaCl, 25 mM Tris–HCl, pH 7.5. After desalting the V353G TACE with Bio-Rad P-6 columns, TAPI-2 was added to a molecular ratio of 1 : 2 (enzyme : inhibitor). The complex was crystallized using the hanging drop vapor diffusion technique. Equal amounts of V353G TACE and inhibitor solution were mixed with the reservoir solution, containing 15% polyethylene glycol 4000, 10% 2-propanol, 100 mM sodium citrate (pH 5.6) and equilibrated at 295 K. Crystals were observed after 7 days and belong to space group P212121.
V353G TACE crystals were washed using the reservoir solution. For cryopreservation, glycerol was added to a final concentration of 15% to the reservoir solution. Crystals were flash frozen in liquid propane and X-ray diffraction was collected at 100 K using internal or synchrotron sources. HKL2000 (Otwinowski, 1993) and CCP4 (CCP4, 1994) were used for data processing and reduction. The crystal structure of V353G TACE was solved by molecular replacement using the program AMORE (Navaza, 1994) and refined using REFMAC (CCP4, 1994). The published TACE structure was used as the starting model (PDB code 1BKC
[PDB]
; Maskos et al., 1998). The diffraction data and model statistics are given in Table II. Soaking with compounds was performed by incubation of the crystals in the reservoir solution in the presence of up to 1 mM of the respective inhibitor. Replacement of the co-crystallized inhibitor was verified by analyzing differences in electron density maps (fo – fc). Figures were generated using Molscript (Kraulis, 1991) and RASTER3D (Merrit and Murphy, 1994).
|
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Mutagenesis of position 353 in TACE dramatically affects overexpression levels
As previously reported (Moss et al., 1997), wild-type TACE overexpression results in the protein being processed and secreted into the expression media at concentrations of 10 mg/l. This expression level resulted in final purified amounts of 2–4 mg/l with the majority of protein loss coming from aggregation during processing. However, mutagenesis of position 353 to either Gly or Ser resulted in an increase in aggregation and resulted in final protein yields of approximately one-fifth that of the wild-type. Purification schemes and subsequent losses were consistent for all three recombinant forms of the protein with the exception of additional aggregate observed during gel filtration purification of the mutant forms. Final purified TACE (wild-type and mutant forms) was determined to be monomeric by dynamic light scattering and analytical gel filtration. This protein was not observed to be further susceptible to aggregation during final concentration for crystallization.
Autoproteolysis of TACE is specific and occurs through a bimolecular mechanism
Purified TACE was found to be stable for at least 2 weeks at 10–15 mg/ml when either co-complexed with inhibitors or stored in the presence of 0.15 M NaCl (Milla et al., 1999). Protein in the absence of salt or inhibitor showed a rapid and specific cleavage between residues Y352-V353 which resides within the surrounding sequence TLGLAY-VGSPRN (Figure 1). Cleavage at this site yields two distinct products having predicted molecular weights of 16 057 and 14 624 Da for the protein fragments 215–352 and 353–477, respectively.
|
To investigate the mechanism of TACE degradation, the autoproteolysis of TACE was measured by both enzyme activity and SDS–PAGE at three protein concentrations 1, 10 and 100 µM (0.03, 0.3 and 3.0 mg/ml). Monitoring the enzyme activity and amount of uncleaved protein over a period of 24 h clearly demonstrated a concentration dependence of inactivation (Figure 2). When the amount of remaining enzyme activity at a given time (At/A0) was fitted to a simple decay equation, the t1/2's were 4.5, 15 and 28 h for 100, 10 and 1 µM enzyme, respectively (Figure 2A). This data correlated with results found when the amount of full-length protein remaining was analyzed by SDS–PAGE with t1/2's of 1.3 and 16 h for 100 and 10 µM enzyme (Figure 2B). The concentration-dependent half-life is consistent with a second-order intermolecular cleavage event (Rose et al., 1993
).
|
Wild-type TACE demonstrates unique specificity at the
subsite
In order to probe the substrate specificity of wild-type TACE, several residues containing different side-chain characteristics (size, hydrophobicity, charge, etc.) were substituted at the 
position of the TNF-alpha substrate (Table III). Although none of the substitutions produced better substrates than the native sequence, several substitutions were accommodated at a lesser efficiency. Those peptides that were cleavable (
, Ala, Phe and Asn) exhibited specificity constants (kcat/Km) that were 10- to 70-fold less than the wild-type residue. The substitutions of Gly or Asp yielded substrates that were not cleavable under the assay conditions tested.
|
Mutagenesis of position 353 stabilizes the autodegradation of wild-type TACE
Utilizing the specificity data for wild-type TACE in combination with energy minimization studies, two mutant proteins were created with the replacement of the native Val at position 353 with either a Ser or a Gly. A Gly was chosen because it completely blocked substrate hydrolysis and a Ser was chosen as an intermediate substitution (resulted in 10-fold reduction in substrate proteolysis) to observe what stabilizing effect a poor yet cleavable substrate would have on protein stability. The resulting proteins V353S TACE and V353G TACE were purified and characterized to assess their stability in the absence of salt or inhibitor. SDS–PAGE characterization of these mutant proteins indicated that both were more stable with the Gly substitution (as expected) having a more dramatic effect on limiting autoproteolysis. In Figure 3, no degradation of either mutant was observed after 3 h at 15 mg/ml in the absence of salt or inhibitor. Contrary to this, 10% of wild-type TACE was degraded in the same time frame. After 2 weeks, 95% of wild-type TACE was degraded while 50% of V353S TACE and 10% of V353G ACE underwent autoproteolysis as determined by densitometry. Thus, V353G TACE was selected for further evaluation. The analysis of V353S TACE was to determine the impact of a substitution that was both conservative in nature however not completely inhibitory.
|
V353G TACE has similar kinetic properties to wild-type TACE
V353G TACE and wild-type TACE were compared kinetically with the TNF-alpha substrate peptide and a series of competitive inhibitors (Table I). kcat/Km was approximately five times lower for the mutant enzyme suggesting that the mutated residue might have an impact on the binding/turnover of the substrate. Comparison of the structures of the two proteins indicated a shift in a structural loop containing this residue that might explain the observed enzymatic differences (Figure 4). Inability to fully saturate TACE with substrate prohibited the actual determination of true binding constants and turnover numbers. However, comparison of a panel of structurally diverse competitive inhibitors indicated similar binding to the two enzymes. IC50 values for the panel were experimentally within 2-fold of each other.
|
Crystallization of V353G TACE results in a soakable crystal form
The structure of V353G TACE is similar to wild-type TACE. Using the listed crystallization conditions, crystals were obtained within 7 days of set-up. Although V353G TACE crystallizes under similar conditions to wild-type TACE, the space group was found to be different. The V353G TACE crystals belong to space group P212121 (a = 73, b = 75, c = 103 Å) and were found to contain two molecules within the asymmetric unit. The crystals diffracted to a resolution of 1.7 Å. Overall structural comparison of wild-type TACE with V353G TACE indicated that there are three regions that exhibited conformational change between the two structures. Figure 4 illustrates these differences (circled in red) and indicates that the most dramatic changes occur within the loop from Val353 to His361 (center circle). Residue Pro356 within this loop shifts the most with a distance of 13 Å from its original position.
Displacement studies of TAPI-2 indicated that V353G TACE was amenable to soaking for structure-based drug design efforts. Figure 5 illustrates the electron density for the soaked in inhibitor JMV 390. The mutation of position 353 from a valine to a glycine caused a conformational change (Figure 6) that facilitates crystallization in a new space group giving a crystal form that is soakable.
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
The regulation of protease activity is often achieved through autoproteolytic processing events that can result in either the activation of an enzyme or its eventual inactivation. For example, the aspartic proteases can undergo autoprocessing to facilitate the removal of the propeptide and result in fully active biologically relevant catalytic domains (Deinum et al., 1998
; Wittlin et al., 1999). Contrary to this, autolytic cleavage also can result in inactivation. Activated -trypsin is autoproteolyzed at Lys176-Asn177 resulting in a lack of affinity for native substrates (Smith and Shaw, 1969). This may serve as a means of regulating the activity of the enzyme following activation.
In the case of TACE, the instability of the uninhibited enzyme at higher concentrations made protein handling challenging. The crystal form of the published X-ray structure (Maskos et al., 1998) was found unsuitable for iterative structure-based drug design studies. Therefore, a new crystal form was sought. In order to investigate the broadest conditions possible, we wanted to be able to screen the protein in the absence of inhibitor or salt. To facilitate this objective, the autoproteolysis of the enzyme was investigated.
Since the autoproteolysis site appears to be only loosely related to the native substrate cleavage site of TNF (SPLAQA-VRSSSR), the mechanism of action of cleavage was investigated. Proteolysis of TACE by a contaminant enzyme was ruled out through the observation that the inhibition of TACE activity (by inhibitor or salt) preserved the integrity of the enzyme at all concentrations. Therefore, the mechanism of inactivation can be determined by identifying the order of the reaction. Autoproteolysis could either occur through an intermolecular cleavage event (cleavage of one TACE molecule by another) or by one that is intramolecular (one molecule of TACE cleaving itself). In the first case, the half-life of the enzyme would be dependent on enzyme concentration (second-order) while intramolecular autoproteolysis would be constant over a broad range of concentrations (first-order). In the case of TACE autoproteolysis, the cleavage of the enzyme was concentration-dependent and therefore a second-order process.
Since it appeared that the sequence was acting as a pseudo-substrate for the active site, and that only the and 
residues match the consensus substrate sequence, a limited study examining the specificity of the residue was performed to suggest substitutions that would be resistant to cleavage. Previous investigations identified a general preference of the subsite for hydrophobic residues (Jin et al., 2002) but did not evaluate a panel of substrates with diverse substitutions. Therefore, using a panel of surrogate peptides based on the native sequence of TNF-, the specificity of the subsite was explored. Substitution of the valine with hydrophobic residues (Phe, Ser, Ala) yielded kcat/Km values that were 10% of that for the Val containing peptide. Asn at the position was highly resistant to cleavage with a kcat/Km of 1% of Val, while Gly and Asp showed no detectable cleavage under assay conditions (Table III). Based on these data we surmised that it would be possible to design mutants of TACE that would be highly resistant to autoproteolysis. We therefore examined the impact of V353G and V353S on the stability and catalytic activity of TACE. These substitutions were chosen based on the decrease in hydrolysis of substrates containing them at and modeling experiments that suggested compatibility.
The initial observation that V353G TACE was more stable (Figure 3) than wild-type TACE or V353S TACE led us to focus our characterization and screening efforts solely on this construct. Greater than 90% of the starting material for this construct remained intact following a 2 week incubation period at 4°C under conditions where the enzyme was active. Determination of the three-dimensional structure of this construct complexed with TAPI-2 indicated that there were three regions in the structure that exhibited differences from those of complexed wild-type TACE (Figure 4).
To confirm that these observed structural differences do not dramatically affect the enzymatic properties of TACE, we evaluated the kinetics of the two enzymes with several ligands. The overall profile for the mutant protein indicated that the mutation only slightly affected the activity of the enzyme. Metalloprotease inhibitors (Table 1) binding to the prime-side of the active site exhibited virtually identical binding constants when characterized with wild-type TACE and V353G TACE. In contrast, the kcat/Km values for wild-type TACE and V353G TACE differed by 5-fold (V353G TACE is less) when assayed with a peptide containing the TNF- sequence. This observed reduction can be explained by the interaction of the substrate with a different conformation in the non-prime region caused by the V353G mutation. Comparison of the two structures indicated (Figure 4) that in V353G TACE, the loop from residue 353 to 361 shifts toward the catalytic zinc narrowing the non-prime-side of the active site cleft. However, since the binding of prime-side inhibitors is not affected by the V353G mutation, the mutant protein demonstrated identical structure–activity relationships in this region and is therefore a useful tool for structure-based drug design work.
In conclusion, we describe a mutation of TACE that both stabilizes the protein from autoproteolysis and causes a conformational change that facilitates crystallization in a new space group that is amenable to soaking. The replacement of Val 353 with either Gly or Ser resulted in a stable form of uninhibited TACE with reduced rates of autoproteolysis. The ability to predict stabilization mutations based on enzyme specificity and cleavage site characterization suggests a general method of stabilizing proteases that otherwise undergo autoproteolysis.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Bemelmans,M.H.A., VanTits,L.J.H. and Buurman,W.A. (1996) Crit. Rev. Immunol., 16, 1–11.[Web of Science][Medline]
Black,R.A. et al. (1997) Nature, 385, 729–733.[CrossRef][Medline]
Chen,J.M., Guixian,J., Sung,A. and Levin,J.I. (2002) Bioorg. Med. Chem. Lett., 12, 1195–1198.[CrossRef][Medline]
Collaborative Computational Project No. 4. (1994) Acta Crystallog. sect. D, 50, 760–763.
Deinum,J., Derkx,F.H. and Schalekamp,M.A. (1998) Biochim. Biophys. Acta., 1388, 386–396.[CrossRef][Medline]
Jin,G., Huang,X., Black,R., Wolfson,M., Rauch,C., McGregor,H., Ellestad,G. and Cowling,R. (2002) Analytical Biochemistry, 302: 269–75.[CrossRef][Web of Science][Medline]
Kraulis,P.J. (1991), MOLSCRIPT: a program to produce both detailed and schematic plots of protein structures. J. Appl. Crystallog., 24, 946–950.[CrossRef]
Levin,J.I. et al. (2002) Bioorg. Med. Chem. Lett., 12, 1199–1202.[Medline]
Maskos,K. et al (1998) Proc. Natl Acad. Sci. USA, 95, 3408–3412.
Merritt,E.A. and Murphy,M.E.P. (1994) Acta Crystallogr. Sect. D Biol. Crystallogr., 50, 869–873.[CrossRef][Medline]
Mikuls,T.R. and Weaver,A.L. (2003) Curr. Rheumatol. Rep., 5, 270–277.[Medline]
Milla,M.E. et al. (1999) J. Biol. Chem., 274, 30563–30570.
Moss,M.L. et al. (1997) Nature, 385, 733–736.[CrossRef][Medline]
Navaza,J. (1994) Acta Crystallog. Sect. A, 50, 157–163.[CrossRef][Web of Science]
Otwinowski,Z., Sawer,L., Isaacs,N. and Bailey,S. (1993) Proceedings of the CCP4 Study Weekend: Data Collection and Processing. Daresbury Laboratory, Warrington, UK.
Perez,C., Albert,I., DeFay,K., Zachariades,N., Gooding,L. and Kriegler,M. (1990) Cell, 63, 251–258.[CrossRef][Web of Science][Medline]
Rose,J.R., Rafael,S. and Craik,C.S. (1993) J. Biol. Chem., 268, 11939–11945.
Shanahan,J.C. and St Clair,E.W. (2002) Clin. Immunol., 103, 231–242.[CrossRef][Web of Science][Medline]
Smith,R.L. and Shaw,E. (1969) J. Biol. Chem., 244, 4704–4712.
Tracey,K.J. and Cerami,A. (1994) Annu. Rev. Med., 45, 491–503.[CrossRef][Web of Science][Medline]
Utsumi,T., Levitan,A., Hung,M.C. and Klostergaard,J. (1993) J. Biol. Chem., 268, 9511–9516.
Wang,W. and Malcolm,B.A. (2002) Methods Mol. Biol., 182, 37–43.[Medline]
Wittlin,S., Rosel,J., Hofmann,F. and Stover,D.R. (1999) Eur. J. Biochem., 265, 384–393.[Web of Science][Medline]
Received September 20, 2005; revised December 16, 2005; accepted December 18, 2005.
Edited by Mirek Cygler
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||