Crystal structure of human protein tyrosine phosphatase 14 (PTPN14) at 1.65‐Å resolution

Protein-tyrosine phosphatases (PTP) function as critical regulators of many cellular processes, such as metabolism, cell growth, and differentiation, by opposing tyrosine kinase-mediated phosphorylation.1 The 107 human PTPs identified to date can be subdivided into the phosphotyrosine-specific, “classical” PTPs, and those that dephosphorylate phospho tyrosine, threonine, and serine, the dual-specificity PTPs. The 38 classical PTP family members includes receptor-like transmembrane forms and non-transmembrane cytosolic forms.2, 3 Protein-tyrosine phosphatase 14 (PTPN14/PTPD2/PEZ) is a 130-kDa multidomain cytosolic protein, and belongs to the NT6 subtype of PTPs. It is characterized by an N-terminal FERM domain (Band 4.1, ezrin, radixin, moesin homology) and a C-terminal PTP domain with an intervening sequence containing an acidic region and a putative SH3 domain-binding sequence.4, 5 Multiple isoforms of the murine homologue PTP36 have been identified in which parts of these domains are truncated.6 Other mammalian PTPs with a FERM domain include PTPN3, PTPN4, PTPN13, and PTPN21. A feature of FERM domain-containing proteins is that they interact with the plasma membrane and membrane/trans-membrane signaling proteins to regulate organizing the cytoskeleton.7 PTPN14 was first cloned in a screen for PTPs expressed in normal breast tissue, and it is also expressed in varying amounts in kidney, skeletal muscle, lung, and placenta.8 In addition, PTPN14 is highly expressed in human umbilical vein endothelial cells (HUVEC), suggesting it may be a critical enzyme in regulating endothelial cell function.9 On overexpression in HeLa cells, murine PTPN14 is enriched in the membrane-associated cytoskeletal fraction while in HUVEC cells its subcellular localization is dependent on cell density and regulated by serum and TGFβ. In cells grown to confluence, PTPN14 is localized in the cytosol, where it is concentrated at intercellular junctions, but it is predominantly nuclear in sparsely plated cells that have not yet formed extensive cell–cell contacts. TGFβ, which inhibits cell proliferation but not migration, also inhibits translocation of PTPN14 from the cytosol to the nucleus. Furthermore, PTPN14 has been detected in the nucleus in migrating and proliferative cells at the wound edge, suggesting that this phosphatase plays a role in the nucleus during cell proliferation. β-Catenin, a central component of adherens junctions, has been identified as a PTPN14 substrate, and it has been demonstrated that expression of a dominant negative PTPN14 enhances tyrosine phosphorylation of adherens junctions leading to a decrease in cell–cell adhesion and increased cell motility.5, 9-11 Mutational analysis of the tyrosine phosphatase gene family in colorectal cancer identified somatic mutations in several PTPs including missense mutations in PTPN14, such as the Thr1068/Met mutation in the catalytic domain. The functional significance of the PTPN14 mutations has yet to be determined.12 PTPs have been implicated in a variety of human diseases including cancer and cardiovascular and immunological disease, and many of these enzymes are recognized as potential therapeutic targets.13 Here we report the crystal structure of the catalytic domain of human PTPN14 at high resolution. Structural comparison with other PTP family members reveals unique structural features and provides insight into its distinct substrate specificity. Using a PTPN14 cDNA (Purely Proteins Ltd.), multiple expression constructs encompassing the catalytic domain of PTPN14 were generated in the vector pNIC28Bsa4. The vector, derived from pET21a, includes a Tobacco Etch Virus (TEV)-cleavable (*) N-terminal 6×His tag (MHHHHHHSSGVDLGTENLYFQ*SM). A construct expressing residues 886–1187 of (SwissProt accession number Q15678) in the bacteriophage-resistant Escherichia coli BL21(DE3)R3 gave well diffracting crystals. Cells were grown at 18°C after induction with 1 mM isopropyl-thio-β-D-galactopyranoside (IPTG) for 4 h. Cells were harvested and resuspended in a binding buffer (50 mM HEPES pH 7.5; 500 mM NaCl; 5 mM imidazole, 5% glycerol; 0.5 mM TCEP). After cell lysis with an Emulsiflex high-pressure homogenizer and centrifugation of the lysate the supernatant was purified by gravity flow on DEAE cellulose (DE52, Whatman) followed by Ni-NTA affinity chromatography (Qiagen). The eluted protein was further purified on a Superdex 200 16/60 gel-filtration column equilibrated in 50 mM HEPES, pH 7.5, 150 mM NaCl, 10 mM DTT run on an ÄktaXpress system, and the protein was concentrated to 8 mg/mL using an 10-kDa cut-off concentrator (Amicon). LC-ESI-TOF mass spectrometry confirmed the correct mass predicted for this construct (37565 Da). Crystals were grown at 20°C using the sitting drop method, and initial trials were carried out using in-house screens and commercial screening kits (Hampton Research). The best crystals were obtained following optimization using 200 nL sitting drops mixing 150 nL of protein with 50 nL of a solution containing 0.1 M Bis-Tris HCl, pH 5.0; 0.2 M Li2SO4, and 25% PEG 3350. The crystals were cryo-protected using 20% ethylene glycol, which was added to the drop 30 s prior to mounting. X-ray diffraction data were collected at SLS synchrotron single wavelength beamline-X10 at 100 K using a MarResearch CCD detector to a maximum resolution of 1.65 Å. Structure solution and refinement: Diffraction images were indexed, integrated and scaled using HKL200014 in the CCP415 suite of programs. The statistics for data collection and refinement are summarized in Table I. The structure of PTPN14 was solved by molecular replacement with Phaser16 using SHP1 (PTPN6) [Protein Data Bank (PDB) code: 1FPR] as a search model. Iterative rounds of restrained refinement with TLS using refmac5,17 against maximum likelihood targets, were interspersed by manual rebuilding of the model using Coot.18 The atomic coordinates of the final structure have been deposited in the PDB with the code 2BZL. The crystal structure of PTPN14 has been refined to high-resolution, low R-values and satisfactory geometry (Table I). The structure determined here contains the catalytic domain residues Glu895 to Ser1184. The entire structure was very well defined in the electron density except for a small region close to the N-terminus (Glu905–Gly907), three loop residues (Leu1026–S1028), and the last three C-terminal residues. The asymmetric unit contains one protein molecule, two sulfate ions, and two alternative conformations of an ethylene glycol molecule that was used as a cryo-protectant. Val1049 located in the loop between sheet β8/9 and sheet β10 has unusual backbond angles; however, the conformation of this residue is well defined in the electron density. The overall structure of the catalytic domain closely resembles the classical architecture of tyrosine phosphatases,19 consisting of highly twisted mixed β-sheets flanked by α-helices (Fig. 1). The main secondary structure elements of PTPN14 resemble closely the structures determined for related phosphatases PTP1B (sequence identity 32%; RMSD of Cα atoms 3.2 Å) and PTPN6 (sequence identity 32%; RMSD of Cα atoms 3.8 Å); however, differences are present in a number of loop regions connecting conserved secondary structure elements. The main difference in PTPN14 is the presence of an extended α3–β12 loop of 12 residues (Thr1103–Pro1115), starting in a 310-helix, which is much smaller in other phosphatases whose structures have been determined to date (Fig. 1). Sequence alignments indicate, as expected, that the closely related phosphatase PTPN21 (PTPD1) has a similar sequence insertion.3 Interestingly, the phosphatases PTPN9 (MEG2) and the D2 domain of PTPRC (CD45) share this extended loop sequence. In the CD45 structure this 11-amino acid basic loop is flexible in nature, and is thought to play a role in substrate recruitment.20 (A) Three-dimensional structure of PTPN14. Secondary structure elements were determined using the program ICM (Molsoft) and labeled using the nomenclature in ref.19. The WPD loop is shown in yellow, α-helices are in red, β-strands in green, and the 310-helices in magenta. The active site Cys1121 and cocrystallizing sulfate ions are shown as a ball-and-stick representation. (B) Structure-based sequence alignment of PTPN14 and PTP1B (PDB code: 2HNP). Secondary structure elements are colored as above. Structural studies, together with enzyme kinetic studies of other PTPs, have provided important insights into the mechanism of substrate recognition and catalysis.21, 22 In PTPN14, the active site cysteine (Cys1121) superimposes extremely closely on that of PTP1B, and is located in the well-conserved phosphate binding loop [HCSXGXGR(T/S)G]. Also, the position of Gln1169, of the Q-loop, which H-bonds with the active site water molecule is highly conserved as in other PTPs. The catalytically important WPD loop (Trp1077–Ser1091) is in an unhindered open conformation. An ethylene glycol molecule is found in close proximity to the active-site cysteine, and might mimic a hydrophilic substrate phosphate moiety that is expected to bind at this site. The phosphotyrosine recognition loop, with the sequence Tyr–Arg–Asp (YRD) in PTP1B, which delineates a boundary of the active site pocket, is replaced by Ile939–Arg940–Glu941 in PTPN14. The change of the aromatic tyrosine residue to aliphatic isoleucine is significant, because in PTP1B, Tyr46 mediates (π–π) packing with the phenyl ring of the substrate phosphotyrosine,23, 24 and this interaction is not possible in a PTPN14 substrate complex, leading to reduced affinity for a phosphotyrosine substrate. However, the hydrogen bonds formed by Asp48 of PTP1B to the substrate are likely maintained by the conserved glutamic acid residue of PTPN14. Residues Met258 and Gly259 of PTP1B form an open cleft, allowing direct access of substrates to the active site; however, bulkier residues in this position, such as in PTPRA(alpha), cause steric hindrance, and are key determinants of substrate selectivity.25, 26 In PTPN14, Met1161 and Phe1162 are present in the corresponding positions, and the bulky phenylalanine is likely to restrict substrate specificity to phosphopeptides with a small amino acid in the pY+1 position. The secondary phosphotyrosine binding pocket found in PTP1B is not evident in PTPN14 due to the different loop conformation.24, 27 Several somatic mutations have been described for PTPN14 that occur in various cancer tissues.12 One of the mutations frequently detected in colorectal cancer (Thr1068Met) is located in the catalytic domain (Fig. 2). This residue is surface exposed, and on the opposite face of the protein from the active site, making it unlikely that the catalytic activity of PTPN14 is directly affected by this mutation. The mutation is in the proximity of the extended loop, and together, these residues may constitute a surface involved in protein–protein interactions. The functional consequences of this mutation has yet to be determined; however, because inactivation of PTPN14 using a dominant negative approach leads to increased phosphorylation of adherens junctions,11 one possibility is that a downregulation of PTPN14 function causes decreased cell–cell adhesion and increased cell migration, thereby contributing to cancer cell invasion and metastasis. Alternatively, the mutations may upregulate PTPN14 function, and consequently, endothelial cell proliferation and angiogenesis in cancer. As such, PTPN14 regulates cellular pathways that may be amenable to therapeutic intervention. These results provide a high-resolution crystal structure of the phosphatase PTPN14, thereby providing insight into its distinct substrate specificity and identifying unique structural features. Superimposition of PTP1B (PDB code: 2HNP) colored red and PTPN14 (PDB code: 2BZL) in yellow showing the extended loop. The phosphate binding loop of the active site is shown in blue for orientation and the position of the Thr1068/Met mutation is shown. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.] We thank the crystallography group for collecting diffraction data and the Biotechnology Group for providing the expression vector. The Structural Genomics Consortium is a registered charity (number 1097737) funded by the Wellcome Trust, GlaxoSmithKline, Genome Canada, the Canadian Institutes of Health Research, the Ontario Innovation Trust, the Ontario Research and Development Challenge Fund, and the Canadian Foundation for Innovation.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.