摘要：

Protein tyrosine phosphatases (PTPs) play a critical role in regulation and maintenance of balanced cellular functions. As such, many PTPs are important therapeutic targets. For example, the inhibition of some PTPs,i.e.PTP1B and SHP2, has been shown to be effective against human diseases such as diabetes, obesity and cancer, while the inhibition of their highly homologous counterparts (TCPTP and SHP1, respectively) is detrimental to normal cell growth.1The development of potent and selective drug candidates against members of PTPs has been hampered largely due to the highly conserved active site of these enzymes.2Herein, we report a novel phosphopeptide microarray that enables high-throughput (HT) screening of many putative substrates of PTPs. By making use of PTP “substrate-trapping” mutants (that is, active-site mutants of PTPs capable of noncovalently binding to, but not dephosphorylating, their substrates),3we identified phosphopeptides that confer both strong and specific binding to selected PTPs. This information may aid in the future design of potential PTP inhibitors.Numerous methods exist for comprehensive studies of substrate specificity of PTPs.4,5Combinatorial peptide libraries,4aphage display4band SPOT™ synthesis4cwere some of the early examples. More recently, we and others introduced peptide-based, enzyme substrate microarrays for large-scale studies of phosphatase substrate specificities;5with these new platforms, the enzymatic activity of different classes of protein phosphatases (including PTPs and Ser/Thr phosphatases) could be interrogated, both qualitatively and quantitatively, against a variety of immobilized phosphopeptide substrates. Not withstanding the clear advantages in terms of its simplicity, miniaturization, throughput and sensitivity,6the enzyme substrate microarray suffers from the following drawbacks: (1) inactivation of PTPs often occurs due to the highly oxidizable/nucleophilic cysteine residue in the enzyme active site;7(2) the reproducibility of microarray data is still questionable, despite improvements in the array fabrication and introduction of subarrays in the screening process.5aConsequently, the method cannot be used to discriminate the substrate specificity of highly homologous enzyme pairs, such as PTP1B–TCPTP and SHP1–SHP2 (Fig. S2, ESI).Using our current approach, we developed a “substrate-trapping” microarray to overcome these limitations. As shown inFig. 1, the highly conserved cysteine residue in a PTP active site was mutated to a catalytically inactive serine residue (thus eliminating the inactivation issue in wildtype PTP), generating the so-called “substrate-trapping” mutant, which retained the noncovalent substrate binding ability but prevented subsequent substrate dephosphorylation from occurring.8Previously, these substrate-trapping mutants had been used for proteomic pull-down/identification of potential PTP substrates.3,8bFurthermore, unlike the previous enzyme substrate microarrays where only one enzyme could be applied to the same peptide substrates at any given time,5the substrate-trapping array reported herein is compatible with a newly developed dual-color screening approach (Fig. 1, bottom);9by introducing a screening solution containing two different PTP trapping mutants (each labeled with a spectrally distinct fluorophore) to the same microarray, we were able to identify potential PTP1B- and SHP1-selective peptides.(a) Normal dephosphorylation event catalyzed by a PTP. The active-site cysteine is shown. (b) Substrate-trapping mutants of PTPs bind to the phosphopeptide substrates without catalyzing the dephosphorylation. In a dual-color screening, a pair of PTP mutants was first labeled with Cy5 and Cy3 dye, respectively, and mixed before being applied to the peptide microarray. PTP1B-selective peptides were then visually identified.We synthesized 144 putative peptide substrates of PTPs, each containing eleven amino acid residues with a centrally locatedpTyr moiety flanked by five amino acid residues o

展开