Crystal structure of the human dual specificity phosphatase 1 catalytic domain

Abstract

The dual specificity phosphatase DUSP1 was the first mitogen activated protein kinase phosphatase (MKP) to be identified. It dephosphorylates conserved tyrosine and threonine residues in the activation loops of mitogen activated protein kinases ERK2, JNK1 and p38‐alpha. Here, we report the crystal structure of the human DUSP1 catalytic domain at 2.49 Å resolution. Uniquely, the protein was crystallized as an MBP fusion protein in complex with a monobody that binds to MBP. Sulfate ions occupy the phosphotyrosine and putative phosphothreonine binding sites in the DUSP1 catalytic domain.

Short abstract

PDB Code(s): 6APX

Introduction

Mitogen‐activated protein kinases (MAPKs), a group of highly conserved tyrosine and serine/threonine protein kinases in eukaryotes,1 play vital roles in a myriad of cellular processes.2, 3, 4 MAPKs are inactive until they become phosphorylated by map kinase kinases (MAPKKs), usually at two sites within their activation loops, which contain a characteristic TxY motif that needs to be phosphorylated on both the threonine and tyrosine residues in order to lock the enzyme into its active conformation. Dephosphorylation, mediated by protein tyrosine phosphatases (PTPs) or dual‐specificity phosphatases (DUSPs), returns the MAPKs to their inactive state. Tight control and balance between phosphorylation and deophosphorylation of MAPKs is pivotal for the proper execution of downstream signaling pathways. DUSPs can dephosphorylate both phosphotyrosine and phosphothreonine in the TxY motif of MAPKs.4 The 26 unique DUSPs encoded by the human genome are classified either as MAPK phosphatases (MKPs) or atypical DUSPs.5 The MKPs are composed of an N‐terminal non‐catalytic MAPK‐binding domain (MKBD) and a highly conserved C‐terminal catalytic domain (CD). Atypical DUSPs have only the CD and show high sequence homology to the Vaccinia virus H1 gene product.6, 7, 8 As key regulators in many MAPK signaling pathways, DUSPs are emerging as potential therapeutic targets in human diseases such as cancer.9

DUSP1, the first MKP demonstrated to be specific for MAPKs, targets ERK2, JNK1 and p38‐alpha.10 It is an inducible nuclear protein11, 12 that is 367 amino acids in length and broadly expressed in a variety of human tissues.13 DUSP1 is an important regulator of both adaptive14 and innate immune systems.15 Mice lacking DUSP1 are resistant to diet‐induced obesity, implicating DUSP1 in metabolic control as well.16 Elevated or diminished expression of DUSP1 has also been observed in various cancers. Elevated DUSP1 levels are associated with melanoma,17 pancreatic,18 prostate,19 glial20 and gastric21 cancer cell lines, whereas decreased DUSP1 expressions are associated with ovarian22 and testicular23 cancers. Moreover, while DUSP1 is overexpressed in early stages of colon, prostate and bladder cancer, its expression is diminished as the tumor becomes more aggressive.24, 25 Overexpression of DUSP1 has also been linked to central nervous system diseases such as multiple sclerosis26 and major depressive disorder.27 In short, the perturbation of DUSP1 expression has been linked to a wide variety of maladies, making it a potential molecular target for therapeutic intervention.

As the archetypal member of the MKP family, DUSP1 has been extensively studied. Even so, the structures of neither the full‐length enzyme nor its CD are known. The lack of structural information may explain why very few inhibitors have been reported to target this enzyme.28 Here, we report the crystal structure of a catalytically inactive DUSP1‐CD (Cys258Ser mutant) at 2.49 Å resolution. We expect that the availability of this structure will facilitate efforts to develop DUSP1 inhibitors and elucidate structure/function relationships for this important enzyme.

Results and Discussion

Construct design and overall structure of the DUSP1‐CD

A truncated DUSP1 protein lacking 53 C‐terminal residues had significantly higher basal phosphatase activity than the full‐length protein.29 Therefore, the CD employed in the present study encompasses residues 172–314 (of 367). Attempts to obtain diffraction quality crystals of the wild‐type DUSP1‐CD or the DUSP1‐CD(Cys258Ser) mutant, either alone or fused to the C‐terminus of E. coli maltose binding protein (MBP), failed. The MBP used in these experiments contained several surface entropy reduction mutations that were designed to facilitate the crystallization of MBP fusion proteins.30 Finally, the MBP‐DUSP1‐CD(Cys258Ser) fusion protein was purified in complex with a “monobody” (YSX1) that specifically binds to the maltose‐free form of MBP.31 This time diffraction quality crystals were obtained and the structure was solved at 2.49 Å resolution with one heterodimer per asymmetric unit [Fig. 1(A)]. Data collection and refinement statistics are summarized in Table 1.

Figure 1.

(A) Stereoview illustrating the crystal structure of the YSX1 monobody (red) bound to the MBP (green)‐DUSP1‐CD (cyan) fusion protein. (B) Cartoon diagram representing the overall structure of human DUSP1‐CD in stereoview. Secondary structural elements are labeled. (C) Stereoview showing the Cα trace of DUSP1‐CD (green) superimposed with that of DUSP4‐CD (pink). Secondary structure elements are labeled. The point of view is the same as that in (B).

Table 1.

Crystallographic statistics

Data Collection
X‐ray source	MicroMax‐007 HF
Wavelength (Å)	1.5418
Resolution (Å)a	50.00–2.49 (2.53–2.49)
Space group	P43212
Unit‐cell parameters
a, b (Å) =	105.8
c (Å) =	181.4
α, β, γ (degree) =	90
Total reflections	646916
Unique reflectionsa	36681 (1728)
Completeness (%)a	99.7 (95.6)
Multiplicitya	17.6 (9.0)
Mean I/σ (I)a	40.1 (2.0)
Rmerge a Rpim CC1/2 (highest resolution shell)	0.099 (0.830) 0.023 (0.251) 0.881
Refinement
Rwork	0.20
Rfree	0.23
No. of atoms/B‐factors (Å2)
Protein
MBP	2727/66.2
DUSP1‐CD	1119/53.7
Monobody	710/50.7
Ligands
Sulfates	15/57.1
Glycerol	6/76.4
Water	81/49.0
RMSD from ideal geometry
Bond lengths (Å)	0.008
Bond angles (degree)	0.867
MolProbity validationb
MolProbity clash score	5.25 (99th percentile)
MolProbity protein geometry score	1.60 (99th percentile)
Ramachandran plot
Favored regions (%)	89.4
Allowed regions (%)	10.1
Generously allowed region (%)	0.5
Outliers (%)	0.0
PDB code	6APX

^aValues in parentheses are for the highest‐resolution shell.

^bMolProbity was used to analyze the structure.

The fold of the DUSP1‐CD closely resembles those of other MKPs with a central twisted five‐stranded β‐sheet surrounded by six α‐helices [Fig. 1(B)]. One side of the β‐sheet is covered by two α‐helices while four α‐helices straddle the other. In the crystal lattice, the DUSP1‐CD makes crystal contacts with MBP and the monobody but not with any of the neighboring symmetry‐related DUSP1‐CD molecules.

The coordinates of the DUSP1‐CD were submitted to the PDBeFold server32 to identify the three‐dimensional protein structures in the PDB that are most similar to DUSP1. DUSP4,33 DUSP5,34 and DUSP1535 were the most closely related structures with root mean square deviation (RMSD) values of 0.48, 0.52, and 0.57 Å, respectively. When the structure of the DUSP1‐CD is aligned with that of the DUSP4‐CD (with the highest Z‐score of 14.7), 133 Cα atoms can be superimposed with a RMSD of 0.48 Å out of a total of 143 and 144 residues in the DUSP1 and DUSP4 CDs, respectively [Fig. 1(C)]. Thus, the overall fold of DUSP1‐CD is highly similar to that of the DUSP4‐CD.

The active site

The active site of the DUSP1‐CD is composed of His257‐Cys258Ser‐Gln259‐Ala260‐Gly261‐Ile262‐Ser263‐Arg264 in the phosphotyrosine binding loop (PTP‐loop) and Asp227, which acts as a general acid/base during catalysis. In the crystal structure of the DUSP1‐CD, a sulfate ion from the crystallization solution is bound in the PTP‐loop (sulfate1), mimicking the phosphate moiety of phosphotyrosine (Fig. 2). It is held in the active site by an extensive network of hydrogen bonds between O atoms of the sulfate and the backbone amide NH atoms of Ala260 (3.1 Å), Gly261 (3.0 Å), Ile262 (3.2 Å), Ser263 (3.1 Å), and Arg264 (2.9 Å). The guanidinium nitrogens of the highly conserved Arg264 form two hydrogen bonds (3.0 Å, 2.8 Å) with two oxygen atoms of the bound sulfate. The side chain oxygen atom of Ser258 is located 2.6 Å away from the oxygen atom of the sulfate; a cysteine in this position serves as the catalytic nucleophile in the wild‐type enzyme.36 The side chain carbonyl oxygen of Gln259 forms a hydrogen bond (3.0 Å) with the side chain amide nitrogen of Gln72 in MBP.

Figure 2.

(A) Stereoview of the fit of the residues of the PTP‐loop of DUSP1‐CD to the final 2mF_o – DF_c electron density map (blue, 2.49 Å resolution, contoured at the 1σ level) with the bound sulfate ion. C atoms are colored green, O atoms red, N atoms blue, and the S atom orange. (B) Stereoview illustrating the interactions of the bound sulfate ion in the PTP‐loop of DUSP1‐CD in ball and stick representation. Hydrogen bonds are indicated by red dashes. The point of view is the same as that in (A).

Bound sulfate ions

The DUSP1‐CD was crystallized in 75 mM MES pH 5.9 and 2.4 M ammonium sulfate. In the current structure, three sulfate ions are bound to the DUSP1‐CD. As detailed above, sulfate1 is bound in the PTP‐loop. To identify the putative phosphothreonine‐binding site in DUSP1, the structure of the DUSP1‐CD was aligned with that of the phosphorylated p38 peptide bound to VHR [PDB code: 1J4X].37 Sulfate2 of the DUSP1‐CD is situated close to the region where the phosphate of the peptide phosphothreonine binds to VHR. Therefore, it seems feasible that sulfate2 mimics the binding of the phosphothreonine phosphate in the TxY motif of MAPKs to the DUSP1‐CD. However, whereas the p38 phosphothreonine binds weakly in the VHR/peptide complex, sulfate2 in the DUSP1‐CD structure exhibits strong H‐bonding interactions with the side chain of Arg292 (2.9 Å, 2.8 Å), the main chain amide of Gly172 (3.0 Å) and a water molecule (2.9 Å). A third sulfate (Sulfate3) was found in a pocket adjacent to the PTP‐loop. This sulfate makes H‐bonds with the main chain of Asn209 (3.0 Å) and the side chains of Arg264 (2.7 Å, 3.2 Å) and Ser207 (2.4 Å) of DUSP1, as well as with Gln72 (2.7 Å) of MBP. Arg264 is a conserved residue of the PTP‐loop and its guanidinium side chain is pivotal for the binding of the phosphate in the active site.

Conclusions

In summary, we have determined the crystal structure of the DUSP1‐CD at 2.49 Å resolution. To obtain this structure, it was necessary to employ a monobody that binds to MBP as a co‐crystallization chaperone. In chaperone‐assisted crystallography, an oft‐raised question is whether the chaperone has induced any changes in the structure of the target protein. The fact that the structure of DUSP1‐CD aligned quite well with that of the DUSP4‐CD, which was crystallized in the absence of chaperones, implies that no structural changes in the DUSP1‐CD were caused by the monobody. Although inhibitors of several DUSPs have been described, the overall progress in this field has been limited by the conserved PTP‐loop architecture. In this regard, the structure of the DUSP1‐CD may be useful for the development of specific inhibitors against the DUSP family.

Materials and Methods

Cloning, expression, and purification

The DUSP1‐CD, harboring a Cys258Ser mutation and including a C‐terminal His₆‐tag was cloned into the pMALX(E) vector (a gift from Dr. Lars Pedersen, NIEHS) at the NotI and BamHI restriction sites.30 The open reading frame of monobody YSX1 was synthesized by Biomatik and supplied in a pUC57 vector. It was digested with NdeI and BamHI and ligated between the same two restriction sites in pET‐11a. The MBP‐DUSP1‐CD(Cys258Ser)‐His₆ and YSX1 proteins were expressed separately in E. coli strain BL21‐CodonPlus(DE3)‐RIL (Agilent Technologies). Cultures were grown to mid‐log phase (OD₆₀₀ = 0.4–0.6) at 37°C in Luria‐Bertani medium supplemented with 100 μg mL⁻¹ ampicillin, 30 μg mL⁻¹ chloramphenicol and 0.2% glucose. Proteins were induced by isopropyl β‐D‐1‐thiogalactopyranoside (IPTG) at a concentration of 1 mM. The MBP‐DUSP1‐CD(Cys258Ser)‐His₆ fusion protein was induced at 37°C for 4 h. To produce the YSX1 monobody, first the culture was cooled to 18°C for 30 min and then induced overnight. The cells were harvested by centrifugation and stored at −80°C until further use.

All purification steps were performed at 4°C using a Bio Rad NGC Chromatography System with pre‐packed chromatography columns (GE Healthcare). Co‐lysis was employed to prepare the monobody/MBP‐DUSP1‐CD(Cys258Ser)‐His₆ complex. Cells were suspended in buffer A (50 mM Tris‐HCl pH 7.5, 150 mM NaCl, 10% (v/v) glycerol and 25 mM imidazole) supplemented with complete EDTA‐free protease inhibitor cocktail tablets (Roche Molecular Biochemicals), lysed using an APV‐1000 homogenizer (Invensys APV Products) and centrifuged at 30,000 g for 30 min. The supernatant was filtered through a 0.22 μm polyethersulfone membrane (Thermo Scientific) and applied onto a 5 ml His Trap FF column pre‐equilibrated in buffer A. The column was washed with 10 column volumes of buffer A and eluted with a linear gradient of imidazole in buffer A to 500 mM. Fractions containing the fusion protein/monobody complex were pooled, concentrated by ultrafiltration (30 kDa cutoff, Millipore) and diluted with 25 mM Tris‐HCl pH 7.5 (buffer B) to reduce the NaCl concentration to about 25 mM. The diluted sample was applied onto a 5 ml Hi Trap Q HP anion exchange column equilibrated with buffer B. The column was washed with 10 column volumes of buffer B containing 25 mM NaCl and the bound protein was eluted with a linear gradient of NaCl to 1 M. The eluted fractions of the complex were pooled, concentrated and applied onto a HiPrep 26/60 Sephacryl S‐200 HR column equilibrated in 25 mM Tris pH 7.5, 150 mM NaCl, 2 mM Tris (2‐carboxyethyl) phosphine (TCEP), 10% (v/v) glycerol (buffer C). The peak fractions containing the complex were pooled, concentrated to 33 mg mL⁻¹, centrifuged at 15,700 g for 10 min and filtered through a 0.22 μm Ultra free centrifugal filter unit (Millipore). Aliquots were flash frozen in liquid nitrogen and stored at −80°C.

Crystallization and data collection

The purified complex (33 mg mL⁻¹) was screened for crystals using a Gryphon crystallization robot (Art Robbins Instruments) against commercially available crystal screens at 19°C by the sitting drop vapor diffusion method. Crystals suitable for data collection were optimized in a 2 μL drop (1 μL protein + 1 μL crystallization solution) against 500 μL of reservoir solution by the hanging drop method in 75 mM MES pH 5.9, 2.4 M ammonium sulfate. After a week, a single crystal was retrieved, cryo‐protected with a short sweep in the mother liquor supplemented with 10% (v/v) glycerol and immediately flash cooled in liquid nitrogen. X‐ray diffraction data were collected using a MAR345 detector mounted on Rigaku MicroMax‐007 HF high intensity microfocus generator equipped with VariMAX HF optics (Rigaku Corporation) and operated at 40 kV and 30 mA. Crystals were held at 93K. The data were integrated and scaled with HKL‐3000.38

Structure solution and refinement

The structure of the MBP fusion protein/monobody complex was solved by molecular replacement using PHASER39 from the CCP4 program suite. A search model containing three ensembles comprised of the coordinates of MBP (PDB code 3H4Z),40 the monobody YSX1 (PDB code 3CSB),31 and DUSP4 (PDB code 3EZZ)33 was employed as implemented in PHASER. The model was rebuilt manually in COOT41 and refined with phenix.refine in the PHENIX software suite.42 Model validation was performed with MolProbity.43 The Ramachandran plots were prepared with PROCHECK44 and they revealed that 88.4% of the residues fall within the most favored region, 10.1% are in the allowed region, and 0.5% are in the generously allowed region. There are no residues in the disallowed regions. Figures were prepared with PyMOL (The PyMOL Molecular Graphics System, Version 1.8 Schrödinger, LLC).

Conflict of Interest Statement

The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commerical products or organizations imply endorsement by the US Government.

Data Deposition

Coordinates and structure factors were deposited in the Protein Data Bank under accession code 6APX.