Global Analysis of Protein Tyrosine Phosphatase Activity with Ultra

Jul 14, 2006 - (C35H40BrN3O9PS2+1) 822.11 (100%), 820.11 (94%), 823.11 (41%); found 822.17 ... phosphatase activity with pNPP, whereas the rest of the...
0 downloads 0 Views 266KB Size
Global Analysis of Protein Tyrosine Phosphatase Activity with Ultra-Sensitive Fluorescent Probes Sanjai Kumar,†,§ Bo Zhou,‡,§ Fubo Liang,‡ Heyi Yang,† Wei-Qing Wang,† and Zhong-Yin Zhang*,‡ Department of Biochemistry and Molecular Biology, Indiana University School of Medicine, 635 Barnhill Drive, Indianapolis, Indiana 46202, and Department of Molecular Pharmacology, Albert Einstein College of Medicine, Bronx, New York 10461 Received December 9, 2005

Protein tyrosine phosphatases (PTPs) consist of a large family of enzymes known to play important roles in controlling virtually all aspects of cellular processes. However, assigning functional significance of PTPs in normal physiology and in diseases remains a major challenge in cell signaling. Since the function of a PTP is directly associated with its intrinsic activity, which is subject to post-translational regulation, new tools are needed to monitor the dynamic activities of PTPs, rather than mere abundance, on a global scale within the physiologically relevant environment of cells. To meet this objective, we report the synthesis and characterization of two rhodamine-conjugated probes that covalently label the active site of the PTPs in an activity-dependent manner, thus providing a direct readout of PTP activity and superior sensitivity, robustness, and quantifiability to previously reported biotinylated probes. We present evidence that the fluorescent probes can be used to identify new PTP markers and targets for potential diagnosis and treatment of human diseases. We also show that the fluorescent probes are capable of monitoring H2O2-mediated PTP inactivation, which should facilitate the study of regulated H2O2 production as a new tier of control over tyrosine phosphorylation-dependent signal transduction. The ability to profile the entire PTP family on the basis of changes in their activity is expected to yield new functional insights into pathways regulated by PTPs and contribute to the discovery of PTPs as novel therapeutic targets. Keywords: PTP • activity-based probes • activity-based proteomics • fluorescent probes • proteomics • PTP activity profiling

Introduction Protein tyrosine phosphatases (PTPs) are important signaling enzymes that serve as key regulatory components in various signal transduction pathways.1,2 Defective or inappropriate regulation of PTP activity leads to aberrant tyrosine phosphorylation, which contributes to the development of many human diseases.3 More than 100 PTP have been identified in the human genome.4 The hallmark that defines the PTP superfamily is the active site amino acid sequence (H/V)C(X)5R(S/T), known as the PTP signature motif, in the catalytic domain. Not surprisingly, the PTPs share a conserved mechanism for phosphotyrosine hydrolysis involving a reactive Cys residue in the PTP signature motif.5 Despite a detailed understanding of PTP catalysis, the functional significance of many members within the PTP family is not completely understood. One of the major challenges of the PTP field is to establish the functional roles for every PTP, both in normal cellular physiology and in pathogenic conditions. * To whom correspondence should be addressed: Tel: (317) 274-8025. Fax: (317) 274-4686; [email protected]. † Albert Einstein College of Medicine. ‡ Indiana University School of Medicine. § These two authors contributed equally to this work.

1898

Journal of Proteome Research 2006, 5, 1898-1905

Published on Web 07/14/2006

Gene knockout approaches have been traditionally applied for functional analysis of PTPs. Although in certain biological systems it is possible to assign functions to specific PTPs by disrupting a desired gene and assessing the resulting phenotype, this process is often tedious, and in cases where multiple PTPs have similar functions, compensation adjustments make the results difficult to interpret. In recent years, RNA interference (RNAi) has been actively used to elucidate the roles of PTPs in cellular pathways, although its effectiveness is limited due to problems in delivery, threshold, and off-target effects. Thus, there is a pressing need for developing and applying novel technologies to probe PTP function. In addition, the one gene at a time approach is clearly inadequate to deal with the dynamics and complexity of PTPs in the complement of proteins within a proteome. Genome-wide measurement of changes in transcription in response to different stimuli allows clustering of genes of similar function based on transcriptional co-regulation. However, transcriptional profiling does not always accurately predict protein expression and activity. Proteomic approaches address some of the gaps in genomic technologies by profiling and identifying bulk changes in protein levels. However, standard proteomics methods are less effective in recording variations 10.1021/pr050449x CCC: $33.50

 2006 American Chemical Society

research articles

Activity-Based Fluorescent PTP Probe

evaluation of two rhodamine-conjugated activity-based probes (compounds II and III, Figure 1) for PTPs. These probes permit direct in-gel fluorescence detection of PTP activities and offer superior sensitivity, robustness, and quantifiability to previous biotinylated probes. The availability of the fluorescent probes should enable global measurements on every PTP encoded by the genome in the whole proteome.

Experimental Section

Figure 1. Chemical structures of biotin- and lissamine rhodamine B containing PTP probes.

in protein activity. Since protein function is directly associated with its intrinsic activity which is subject to post-translational regulation, novel methods are needed to profile proteins based on their functional properties, rather than mere abundance. Recently, activity-based proteomics has emerged as the method of choice to record dynamics in enzyme activity in complex proteome.6-8 This approach employs chemical probes to covalently label the active sites of enzyme superfamilies in an activity-dependent manner, thus providing a direct readout of catalytic activity. The covalently tagged proteins could then be captured, quantified and identified by chromatographic, electrophoretic and mass spectrometric techniques. Such an approach has been successfully applied to the serine and cysteine hydrolases.6,9 Recent efforts have also succeeded in creating initial probes for the deubiquitinating enzymes10 and phosphatidylinositol 3-kinases.11,12 Because the function of a PTP depends on its phosphatase activity, and the activities of many PTPs are subject to dynamic post-translational regulation, including reversible oxidation of the catalytic cysteine, the development of activity-based probes to selectively monitor the activity of PTPs on a global scale is of great value. To that end, we have developed activity-based PTP probes (e.g., compound I in Figure 1), which consist of a PTP specific R-bromobenzyl phosphonate for covalent attachment to the PTP active site and a linker which connects the reactive group with a biotin affinity tag for visualization and purification of the modified enzyme.13 We showed that the biotin-tagged probes inactivate a broad range of PTPs in a timeand concentration-dependent fashion. We established that this inactivation is active site-directed and irreversible. We provided evidence that these probes form covalent adduct with PTPs, mostly likely involving the active site Cys residue. More importantly, we demonstrated that the probes exhibit extremely high specificity toward PTPs while remaining inert to other proteins, including the whole proteome from E. coli. These properties indicate that the activity-based PTP probes can be used to interrogate the state of PTP activity in the whole proteome, thereby facilitating the simultaneous activity-based profiling of all PTPs in samples of high complexity. Although the biotinylated probes allow direct avidin-based purification of probe-reactive PTPs, they suffer several inherent limitations. In particular, biotin must be visualized indirectly, typically with anti-biotin antibodies or avidin-horseradish peroxidase complexes. These assays are limited in sensitivity, throughput, and dynamic range, thus hindering efforts to rapidly and quantitatively compare large numbers of samples.14 To overcome these limitations, we report the synthesis and

Organic Synthesis. In general, chemicals for organic synthesis were purchased from Sigma-Aldrich Corporation. Lissamine Rhodamine B sulfonyl chloride was purchased from Molecular Probes Inc. All moisture sensitive reactions were performed in a flame-dried glassware under a positive pressure of argon. The reaction progress was followed by thin-layer chromatography using Merck silica gel 60 F254. Flash chromatography was performed using Merck silica gel (230-400 mesh, 60 A). Final products were purified on a semipreparative reverse phase HPLC using a gradient of 0.1% trifluoroacetic acid containing acetonitrile and water. Synthesis of [(4-Aminomethyl-Phenyl)-Bromo-Methyl]Phosphonic Acid Diethyl Ester. This compound was prepared in the following way, with a slight modification of the protocol published elsewhere.13 Thus, in a round-bottom flask, {Bromo[4-(tert-butoxycarbonylamino-methyl)-phenyl]-methyl-phosphonic acid diethyl ester13 (0.24 g, 0.56 mmol) was dissolved in methelene chloride (5 mL) and chilled on ice. To this, was added trifluoroacetic acid (TFA) (99%, 3 mL) and the reaction mixture was stirred vigorously on ice. The duration of the reaction (typically about 20 min) was monitored by following the disappearance of starting material on thin-layer chromatography plate. After the reaction was over, TFA was rapidly removed on a rotory evaporator and subsequently on high vacuum. The resulting brown crude was suspended in methylene chloride (20 mL) and washed with cold saturated sodium bicarbonate (20 mL) solution. The aqueous layer was extracted in 3 × 20 mL methylene chloride. The combined organic layers were dried on anhydrous sodium sulfate and evaporated to dryness on high vacuum. The resulting product (0.16 g), which appeared pure, based on NMR, was directly used in the next step of synthesis. 1H NMR (300 MHz, CDCl3) δ: 1.25 (t, J ) 7 Hz, 3H), 1.35 (t, J ) 7 Hz, 3H), 3.9-4.2 (m, 6H), 4.8 (d, J ) 13 Hz, 1H), 7.4 (d, J ) 8 Hz, 2H), 7.48 (d, J ) 8 Hz, 2H), 8.1 (br s, 2H). Synthesis of Lissamine Rhodamine B Containing Fluorescent PTP Probes (II & III). Probes II and III were prepared by condensing the Lissamine Rhodamine B sulfonyl chloride (commercially available as a mixture of 2 and 4 positional isomers; Molecular Probes Inc.) with [(4-aminomethyl-phenyl)bromo-methyl]-phosphonic acid diethyl ester.13 Thus, to a solution of [(4-aminomethyl-phenyl)-bromo-methyl]-phosphonic acid diethyl ester (100 mg, 0.30 mmol) in DMF (1.0 mL) was added sequentially Lissamine rhodamine B sulfonyl chloride (206 mg, 0.36 mmol) and triethylamine (60 mg, 0.6 mmol), and the reaction mixture was stirred at room temperature for 24 h in an inert atmosphere of argon. After the solvent was completely removed on high vacuum, the crude reaction mixture, that contained the condensed intermediate, as detected by ESI-MS, was treated as following. To an ice-cooled reaction mixture of the crude in DMF (0.5 mL) was added trimethylsilyl bromide (460 mg, 3 mmol) in drops and allowed to react at room temperature for 24 h. The volatiles were removed on rotory evaporator and the reaction mixture was Journal of Proteome Research • Vol. 5, No. 8, 2006 1899

research articles

Kumar et al.

stirred in the presence of 9:1 methanol/water (10 mL) for 2 h. The final two products were separated and purified from the crude by reverse phase semipreparative HPLC (a linear gradient of 0.1% TFA in acetonitrile and water) in equal ratio as two separate 2 and 4 positional isomers II and III (26% yield) (Figure 1). The characterization of both II & III positional isomers were achieved using ESI-MS (+): m/z calculated for (C35H40BrN3O9PS2+1) 822.11 (100%), 820.11 (94%), 823.11 (41%); found 822.17 (100%), 820.14 (94%), 823.23 (41%). Kinetic Characterization of the Fluorescent PTP Probes. The fluorescent PTP probes were dissolved in dimethyl sulfoxide (DMSO) to a 100 mM stock solution and stored at -20 °C. For assessing kinetic mechanism and the relevant kinetic constants, time dependent inactivation kinetics was performed using the Yersinia PTP YopH. All the kinetic experiments were performed at 25 °C in a pH 6.0 buffer containing 50 mM sodium succinate, 1 mM EDTA, 1 mM DTT, and ionic strength of 150 mM adjusted with NaCl. The inactivation reaction was initiated by the addition of a 5 µL aliquot of PTP stock to a 45 µL solution containing appropriately diluted probe (final DMSO concentration 5%). At appropriate time intervals, an aliquot of 2 µL was withdrawn from the reaction and added into a 200 µL of 20 mM pNPP in pH 6.0 buffer at 30 °C. The pseudo-first-order inactivation constants (kobs) at various concentrations of the probe were obtained by fitting the data to eq 1:

(

)

A0 - A∞ -kobs‚t At A∞ ) e A0 A0 A0

(1)

where A0, At, and A∞ are activities of the enzyme at incubation time zero, t and infinity, respectively. A nonlinear regression fit of kobs vs probe concentration to eq 2 yielded the maximal inactivation constant ki and dissociation constant Ki: kobs )

ki × [I] KI + [I]

(2)

Covalent Labeling with the Fluorescent Probes. All of the labeling reactions with PTPs and non-PTP enzymes were initiated by adding fluorescent PTP probe(s) (final concentration 0.2 mM) to the pre-equilibrated enzymes (final concentration 1 µM) at 25 °C for 15 min in appropriate buffer. All labeling reaction involving PTPs were carried out in 50 mM sodium succinate buffer (pH 6.0) that contained 1 mM EDTA, 1 mM dithiothreitol (DTT), and adjusted to an ionic strength of 150 mM with NaCl. Labeling reaction involving non-PTP enzymes were performed in their respective optimum conditions and thus, the following buffers were used, adjusted to ionic strength of 150 mM - potato acid phosphatase, 100 mM sodium acetate, pH 5.0, 1 mM EDTA; PP1, 50 mM 3,3-dimethylglutarate, pH 7.0, 2 mM MnCl2; alkaline phosphatase, 50 mM Tris, pH 9.0, 1 mM MgCl2; glyceraldehyde-3-phosphate dehydrogenase, 15 mM sodium pyrophosphate, pH 8.5, 7.5 mM NAD, 1 mM DTT; calpain, 100 mM imidazole, pH 7.3, 10 mM CaCl2, 1 mM DTT; trypsin, 100 mM Tris, pH 8.5, 1 mM EDTA; chymotrypsin, 50 mM Tris, pH 7.8, 50 mM CaCl2; thermolysin, 50 mM HEPES, pH 7.0, 5 mM CaCl2; elastase, 100 mM Tris, pH 7.5; glutathioneS-transferase, 50 mM 3,3-dimethylglutarate, pH 7.0, 1 mM EDTA, 1 mM DTT. The Yersinia PTP was chosen to assess the detection limit that is achievable with the fluorescent PTP probes, II and III, in comparison with what could be obtained using previously reported PTP probe I which required antibody-based detection.13 Parallel labeling reactions were set up that involved 1900

Journal of Proteome Research • Vol. 5, No. 8, 2006

incubating YopH with both fluorescent and biotin-tagged PTP probes (2 mM) for 60 min in pH 6.0 buffer at 25 °C. All labeling reactions were quenched by addition of an equal volume of 2X SDS loading buffer (reducing) and heating the resulting mixture at 75 °C for 5 min. The proteins were resolved by SDS-PAGE. The separated proteins on SDS-PAGE were either directly scanned for fluorescence or analyzed using western blot analysis. The amount of labeling with II and III were analyzed using a Typhoon 9400 scanner (Amersham Biosciences Inc.) in fluorescence mode, with green (532 nm) and red (633 nm) lasers with pre-set optimum emission filters. The scanned images were quantified with ImageQuant 3.3 (Molecular Dynamics). Effect of H2O2 on PTP Activity Monitored by the Fluorescent Probes. To determine the effect of H2O2 on PTP activity, 1 µM PTP1B and 1 mM H2O2 were incubated in a 20 µL reaction at 25 °C in the pH 6.0 buffer. At appropriate time intervals, 2 µL aliquots were withdrawn for measurement of residual phosphatase activity with pNPP, whereas the rest of the reaction mixtures were treated with 2 µL of the fluorescent probes (2 mM) at 25 °C for 15 min. The following procedures were used to determine the effect of H2O2 on PTPs in the cell. Human colon carcinoma derived HCT-116 cells were treated with or without H2O2 (1 mM) for 5 min, washed immediately with ice-cold PBS, and lysed and reacted on ice in buffer A (50 mM MES, pH 6.0, 150 mM NaCl, 1.0% Triton X-100, 0.1% SDS, 10% Glycerol, 5 mM EDTA, 10 µg/mL leupeptin, and 5 µg/mL aprotinin, 1 mM PMSF) containing 2 mM fluorescent probes for 30 min. A control experiment without H2O2 treatment was performed simultaneously for direct comparison. The two cell lysates were then cleared by spinning at 13 000 × g for 15 min and the soluble fractions were quantified using Bradford assay kit (Amersham Biosciences). The proteins were separated by SDS-PAGE and the gel was scanned for fluorescence. PTP Profiling in Cancer Cells. The following cancer cells from different origins were chosensMOA231 (lung), MDA-MB435S (breast), MCF-7 (breast), HepG2 (liver), SKOV3 (ovary), HeLa (cervices) and HCT-116 (colon). All cells were grown to ∼80% confluency in a Dulbecco’s modified Eagle’s minimum essential medium (DMEM) supplemented with 10% fetal bovine serum (Life Technologies, Inc.), penicillin (50 units/mL), streptomycin (50 µg/mL), and L-glutamine (2 mM) under a humidified atmosphere containing 5% CO2. Cells were then washed with ice-cold PBS and lyzed in buffer B (50 mM MES, pH 6.0, 150 mM NaCl, 1.0% Triton X-100, 0.1% SDS, 10% Glycerol, 5 mM EDTA, 10 µg/mL leupeptin, 5 µg/mL aprotinin, 1 mM PMSF and 1 mM DTT) on ice for 30 min. The cell lysate was then cleared by centrifugation at 13 000 × g for 15 min. The soluble portion of the proteomes was isolated and the protein concentration determined. The labeling reaction of the cancer cell proteomes was carried out by incubating 50 µg of protein from the different cell lysates with 1 mM of the fluorescent probe for 1 h at 25 °C in 50 mM sodium succinate buffer (pH 6.0) that contained 1 mM EDTA, 1 mM dithiothreitol (DTT), and adjusted to an ionic strength of 150 mM with NaCl.

Results and Discussion Synthesis of Lissamine Rhodamine B Containing PTP Probes. The biotinylated PTP probes reported previously13 were based on R-bromobenzyl phosphonate, an irreversible inhibitor of the Yersinia PTP YopH.15 Subsequent experiments demonstrated that R-bromobenzyl phosphonate and the biotinylated PTP probes could serve as mechanism-based inactivators for

Activity-Based Fluorescent PTP Probe

research articles

Figure 2. Synthetic route for the synthesis of II and III. The fluorescent PTP probes II and III were synthesized from [(4aminomethyl-phenyl)-bromo-methyl]-phosphonic acid diethyl ester and Lissamine Rhodamine B sulfonyl chloride.

all PTPs tested. Although valuable for affinity purification of probe-reactive PTPs, the biotinylated probes lack the required sensitivity for successful identification of low abundant/low activity PTPs, and the robustness for rapid comparison of large numbers of proteomic samples. Fluorescent-based detection is known to offer superior sensitivity and better quantitative accuracy than the biotin/avidin-based measurement.16 To increase the sensitivity, throughput, and dynamic range of the PTP probes, we decided to conjugate the R-bromobenzyl phosphonate group to a fluorescent tag, which is used to visualize the PTPs modified by the probe. Indeed, such fluorescent-labeled probes have been prepared for serine proteases and metalloproteases.17-20 To introduce a fluorescent tag at the 4 position of the phenyl ring, we have developed a seven-step procedure for the synthesis of [(4-aminomethyl-phenyl)-bromo-methyl]-phosphonic acid diethyl ester that contained a naked primary amine.13 We prepared the rhodamine-containing PTP probes by condensing the primary amine of [(4-aminomethyl-phenyl)bromo-methyl]-phosphonic acid diethyl ester with the commercial available Lissamine Rhodamine B sulfonyl chloride (provided as a mixture 2 & 4 positional isomers) in DMF at room temperature. The final step involved deprotection of the phosphonate diester using bromotrimethylsilane to yield the fluorescent PTP probes II and III (Figure 2). The two positional isomers, namely II and III, were isolated and purified using semipreparative reverse phase HPLC and confirmed by mass spectrometry. Kinetic Characterization of Fluorescent Probes II and III. To characterize the two positional isomers II and III as activitybased PTP probes, we examined for their effect on the activity of Yersinia PTP YopH using pNPP as a substrate. Compounds II and III inactivated YopH in a nearly identical time- and concentration-dependent manner (data not shown). We should note that using the same strategy we have also synthesized a Cy5-labled R-bromobenzyl phosphonate probe, which also exhibited very similar reactivity toward PTPs as did the rhodamine-containing probes (data not shown). Apparently, the size and orientation of the fluorescent tags have little effect on the probe’s binding and reactivity toward the PTP active site. Thus, unless indicated otherwise, all subsequent experiments were performed with a mixture of II and III in equimolar ratio, referred as rhodamine-based or fluorescent PTP probes. In addition to YopH, the rhodamine-based probes also inactivated a broad range of PTPs in a time- and concentrationdependent fashion. Analysis of the pseudo-first-order rate constant as a function of probe concentration showed that the fluorescent probes-mediated YopH inactivation displayed satu-

Figure 3. Kinetics of YopH inactivation by the fluorescent probes. (a) Time and concentration dependence of YopH inactivation by the fluorescent probes. The experimental points are represented by various symbols and the line connecting them is the fitted line to eq 1 to yield pseudo-first-order rate constant kobs. Probe concentrations were as follows: O, 0.2 mM; b, 0.5 mM; 4, 1.0 mM; [, 1.5 mM; 2, 2.0 mM; and 0, 3.0 mM. (b) A plot of kobs vs concentration of the fluorescent probes. The points are experimental and the line connecting them is the fitted line to eq 2. This yielded the kinetic parameters, Ki (0.69 mM) and ki (0.52 min-1).

ration kinetics with respect to the probe concentration (Figure 3), yielding values for the equilibrium binding constant KI and the inactivation rate constant ki of 0.69 mM and 0.52 min-1, respectively. These values compare favorable to the biotinylated probe I, which exhibits a Ki of 0.74 mM and a ki of 0.17 min-1 toward YopH.13 Thus, similar to the biotinylated probes, the fluorescent probes act as active site-directed affinity agents whose mode of action likely involves at least two steps: binding to the PTP active site followed by covalent modification of active site residue(s). Reactivity of the Fluorescent Probes toward a Panel of PTPs. To further investigate the probe’s reactivity toward the PTPs, a panel of PTP enzymes from different PTP-subfamilies was chosen including PTP1B, SHP-2, and HePTP for classical cytosolic PTPs, DEP-1 for receptor-like PTP, VHR and PRL-3 for dual-specific phosphatases, and the low molecular PTP (LMW-PTP). Upon incubating the fluorescent probes with the PTPs (10 ng each), the modified PTPs were separated by SDSPAGE and the extent of labeling was assessed by direct in-gel measurement of the rhodamine tag on the fluorescent probes using a Typhoon 9400 flatbed laser-induced fluorescence scanner. As shown in Figure 4, all of the PTPs were covalently labeled and detected. It is most likely that the phenyl phosphonate group serves as a phosphotyrosine mimic that directs Journal of Proteome Research • Vol. 5, No. 8, 2006 1901

research articles

Kumar et al.

Figure 5. Comparison of the sensitivity between the rhodamineand biotin-containing PTP probes. (a) A fluorescent scanned image of the SDS-PAGE gel containing various amount of labeled YopH bythe fluorescent probes. (b) YopH labeling by the biotincontaining PTP probe I, as detected by Western blot using antibiotin antibody under identical experimental conditions.

Figure 4. Labeling of PTPs with the fluorescent probes. Upper panel, direct in-gel fluorescent-based detection of PTPs covalently labeled with the fluorescent PTP probes. Lower panel, silver staining for assessment of loaded protein amount (10 ng).

the probe to the active site, and the latent leaving group (bromide) at the R position unleashes its reactivity upon binding to the PTP active site. Since the PTPs have a conserved phosphotyrosine binding site and share a common mechanism for catalysis, it is not surprising that the probes react covalently with all PTPs, likely involving the catalytic Cys residue.13 Detection Limit of the Fluorescent PTP Probes. To assess the detection limit of the fluorescent probes, a fixed amount of purified YopH was treated with the rhodamine-tagged probes (2 mM) at pH 6.0 and 25 °C for 60 min. pH 6 was chosen for the experiments in order to conserve materials because the R-bromobenzyl phosphonate based probes undergo significant solvolysis at pH values >7.13 However, it is important to point out that the probes display similar reactivity to the PTPs at both pH 6 and 7. For comparison, a parallel reaction was set up under identical conditions with the biotinylated probe I that required Western-blot analysis for detection. After quenching the reaction, the labeled YopH was serially diluted and separated from excess amount of unreacted probes on SDS-PAGE. The gel containing YopH labeled with the fluorescent probes was scanned for green fluorescence while the other was used for biotin/anti-biotin antibody based detection. Much to our satisfaction, the fluorescent-based detection was found to be at least 1000-fold more sensitive than that of the biotinconjugated probes (Figure 5). A similar detection limit was also achieved with a Cy5-based PTP probe (data not shown). Thus, the fluorescent PTP probes could detect as low as 10 pg of YopH, while the limit of detection obtained by the biotincontaining probe was only 10 ng. This enhanced detection limit obtained with the fluorescent PTP probes is quite impressive and should enable the measurement of low abundant and low activity PTPs in complex proteome mixture. Specificity of the Fluorescent PTP Probes. Given the extremely high level of detection limit of the rhodaminecontaining PTP probes, it was imperative to check if they crossreact with other proteins. Thus, we examined the reactivity of 1902

Journal of Proteome Research • Vol. 5, No. 8, 2006

Figure 6. Specificity of the fluorescent PTP probes. The reactivity of the probes was monitored in-gel using a Typhoon 9400 flatbed fluorescence scanner (top panel). A total of 10 ng protein was used for each sample (bottom panel). The lane assignments are as follows: 1, YopH; 2, potato acid phosphatase; 3, alkaline phosphatase; 4, PP1; 5, glyceraldehyde-3-phosphate dehydrogenase; 6, calpain; 7, trypsin; 8, chymotrypsin; 9, elastase; 10, thermolysin; and 11, GST.

the fluorescent PTP probes toward a panel of mechanistically distinct classes of enzymes including potato acid phosphatase, alkaline phosphatase, Ser/Thr protein phosphatase PP1, glyceraldehyde-3-phosphate dehydrogenase, calpain, trypsin, chymotrypsin, elastase, thermolysin, and glutathione S transferase, under their reported optimum activity conditions. As evident from Figure 6, none of these enzymes were labeled by the fluorescent probes under comparable conditions used for YopH. Thus, the fluorescent probes exhibit extremely high selectivity toward PTPs while remaining inert to other proteins. The lack of reactivity of the probe toward nonphosphatases is easy to understand because these enzymes do not recognize aryl phosphates as substrates. What is remarkable is that although both alkaline and acid phosphatases hydrolyze aryl phosphates and possess an active site nucleophile (Ser and His, respectively), neither could be labeled by the probe. The source of this specificity may relate to the highly reactive active site Cys in PTPs and the special affinity of the nucleophilic sulfur for phosphorus. Interestingly, the probe also failed to label glyceraldehyde-3-phosphate dehydrogenase, calpain, and glu-

Activity-Based Fluorescent PTP Probe

Figure 7. Comparative analysis of PTP activity in cancer cells with the fluorescent probes. Left panel, in-gel fluorescence analysis of global PTP activity profiles obtained from reactions between cancer cell lysates (50 µg total protein) and the rhodamine-containing probes. Right panel, Coomassie Blue staining of the protein in cell lysates for comparison. Lane assignments are as follows: 1, MOA231; 2, MDA-MB-435S; 3, MCF-7; 4, HepG2; 5, SKOV3; and 6, Hela.

tathione S transferase, enzymes that also have active site Cys, indicating that noncovalent interactions between the probe and the PTP active site is also important in controlling labeling specificity. This level of selectivity indicates that the fluorescent probes are suitable for global analysis of PTP activity in a cell. Global Analysis of PTP Activity in Cancer Cells. Activitybased probes are emerging as important discovery tools for identification of functional proteins at the whole proteome level.6,17,21 They are also being developed for use in early diagnosis of many life-threatening diseases by directly comparing the activity of many target proteins at the proteome level.20,22 Cancer is marked by aberrant tyrosine phosphorylation, and many known oncogenes are protein tyrosine kinases.23 With such clear links between deregulated protein tyrosine kinase activity and cancer, it is no surprise that much of the effort in kinase drug discovery has been focused on these enzymes. Theoretically, abnormal regulation of PTP activity could also lead to aberrant tyrosine phosphorylation. Unfortunately, the biological significance of the PTPs in cancer remains not well understood. Given the extraordinary selectivity and sensitivity of the fluorescent PTP probes, it is expected that effective application of these probes will accelerate the functional characterization of PTPs in cancer. A better understanding of the roles of PTPs in the onset and progression of tumor will provide a unique perspective in targeting various diseases at the molecular level. The identification of PTP activities selectively expressed by tumor cells or tissues may also provide new biomarkers and targets for the diagnosis and treatment of cancer. To determine the utility of the fluorescent probes to monitor PTP activity at the whole proteome level, we treated cellular lysates (50 µg) derived from various cancer cell lines including MOA231 (lung), MDA-MB-435S (breast), MCF-7 (breast), HepG2 (liver), SKOV3 (ovary), and HeLa (cervices) with 1 mM fluorescent probes for 1 h at pH 6.0 and 25 °C. After the reaction was stopped, the proteins were resolved by SDS-PAGE and the gel was scanned for fluorescence to determine the extent of labeling by the rhodamine-containing probes, which is proportional to PTP activity. As shown in Figure 7, it is clear that the fluorescent probes are capable of recording PTP activity in crude cell extracts. As controls, lysates from MCF-7 cells were

research articles either pretreated with iodoacetate (10 mM for 10 min), which alkylates the PTP active site Cys,24 or preheated in 2% SDS buffer for 95 °C for 10 min before the addition of the PTP probe. As expected, iodoacetate or heat treatment, which should inactivate the PTP activity in the cell lysate, completely blocked the ability of the PTP probe to label the cell lysate (data not shown). Interestingly, it is evident that the activities of members of the PTP superfamily span a large dynamic range within a single cell type. Strikingly, although the patterns of total protein expression were not grossly different among the panel of cancer cell lines, the global PTP activity profiles for each cell type were drastically different. The results indicate that different cancer cell lines can be distinguished by the distinctive activity profiles of the PTPs according to their respective tissue of origin, which will significantly contribute to the molecular diagnosis, the assessment of the cancer risk and prognosis, and the understanding of the molecular basis of cancer. Finally, the ability to profile the entire PTP family on the basis of their activity should greatly accelerate both the assignment of PTP function in cancer and the identification of potential therapeutic targets. Application of the Fluorescent Probes to Monitor H2O2Mediated PTP Inactivation. One important feature of activitybased probes is the ability to measure changes in intrinsic enzyme activity in samples of interest that may take place independent of variations in protein abundance.17 Since the intrinsic activity of PTPs can be altered as a result of posttranslational modification, the use of the fluorescent probes may provide a direct readout of the functional state of the PTPs involved in signaling pathways. In this regard, we note that a substantial body of literature now points to the importance of regulated H2O2 production as a new tier of control over tyrosine phosphorylation-dependent signal transduction, and PTPs have been identified as an important target of H2O2 inside the cell.25-27 The PTPs have an invariant Cys residue in the active site, which functions as a nucleophile in catalysis.5 The active site Cys exhibits an unusually low pKR,24 which enhances its nucleophilicity but also renders it susceptible to oxidation. Oxidation of the active site Cys by H2O2 generates cysteine sulfenic acid (Cys-SOH), which abrogates its nucleophilic properties, thereby inhibiting PTP activity.28 Cysteine sulfenic acid can be readily reduced back to the sulfhydryl state by various cellular reductants, ensuring that H2O2 could function as a ubiquitous intracellular messenger at physiological concentrations.29 One potential function of H2O2 produced following agonist stimulation is to inactivate transiently the PTP(s) that provide an inhibitory constraint upon the system, thereby facilitating the induction of an optimal tyrosine phosphorylation response to that stimulus. To define how PTP activity can be regulated by H2O2mediated transient oxidation in vivo and to gain further insight into the biological importance of this novel regulatory mechanism in various signaling contexts, one needs sensitive and robust methods to measure reversible oxidation of the PTPs. One method that has been reported in the literature is based on the fact that the H2O2-modified Cys in the PTP active site is resistant to alkylation by iodoacetate and can be reactivated by treatment with dithiothreitol, whereas PTPs that have not been oxidized by H2O2 in the cell are irreversibly inactivated by iodoacetate which targets the active site Cys.30 Reactivated PTPs, i.e., H2O2-oxidized PTPs, are then detected on the basis of their ability to dephosphorylate an artificial [32P]phosphatelabeled peptide substrate in an in-gel assay. Obviously, this method is indirect and rather laborious and time-consuming. Journal of Proteome Research • Vol. 5, No. 8, 2006 1903

research articles

Kumar et al.

with ice-cold buffer, lysed and reacted on ice with 2 mM fluorescent probes for 30 min. The results indicate that transient exposure of the cells to H2O2 significantly decreased the fluorescent intensity of many bands as compared to the control, although H2O2 had no appreciable effect on the protein expression patterns (Figure 8c). The decrease in labeling by the fluorescent probes in the presence of H2O2 most likely occurred as a result of PTP oxidation by H2O2. Collectively, the data show that the fluorescent probes can be used to monitor the redox state of the PTP active site cysteine in the cell, thereby providing important tools for studying how PTP activity can be regulated by transient production of H2O2 triggered by the activation of various cell surface receptors.

Conclusions

Figure 8. H2O2-mediated oxidation of PTP visualized by the fluorescent probes. (a) Time-dependent loss of PTP1B activity as a result of oxidation of the catalytic Cys by H2O2. Inactivation was initiated by adding H2O2 (1 mM) to PTP1B (1 µM) at 25 °C. PTP activity was determined using pNPP as substrate. (b) PTP1B labeling by the fluorescent probes. At appropriate time intervals PTP1B treated with H2O2 was allowed to react with 2 mM fluorescent PTP probes for 15 min. The extent of labeling was visualized by in-gel fluorescence scanning (top panel). The bottom panel is silver stain of the amount of PTP1B loaded. (c) A fluorescent scanned image of the gel depicting a reduced level of reactivity of the fluorescent probes toward proteome derived from human colon carcinoma derived cell, HCT-116, without (lane1) and with (lane 2) H2O2 treatment.

Since covalent labeling of the PTPs by the R-bromobenzyl phosphonate-based probes requires the active site thiol group (Kumar et al., 2004), the active site Cys oxidized PTPs should not react with the PTP probes. Consequently, the fluorescent PTP probes should provide a sensitive, direct, and more convenient method to monitor the reversible oxidation of PTPs. To test this hypothesis, PTP1B (1 µM) was incubated with 1 mM H2O2 at 25 °C. At appropriate time intervals, aliquots of PTP1B were withdrawn to measure for residual PTP activity. Simultaneously, the H2O2-treated PTP1B was also allowed to react with the fluorescent probes. The amount of fluorescence labeling on PTP1B and thus the levels of PTP activity remaining was visualized by direct in-gel fluorescence scanning. As shown in Figure 8, the level of PTP1B activity and the fluorescence intensity of the labeled PTP1B diminished close to zero within 6 min of incubation with H2O2. Thus, there is perfect correlation between the amount of catalytically competent PTP1B and the extent of fluorescent probes incorporated into PTP1B, because oxidized PTP1B is catalytically inactive and therefore unreactive to the probes. In another experiment, we sought to determine whether the fluorescent probes are capable of measuring the changes in PTP activities in a transiently oxidized proteome. Human colon carcinoma derived HCT-116 cells were treated with or without 1 mM H2O2 for 5 min, washed immediately 1904

Journal of Proteome Research • Vol. 5, No. 8, 2006

We have synthesized and characterized two fluorescent rhodamine-containing PTP probes that are highly sensitive for direct in-gel visualization of PTP activity. Kinetic analyses suggest that these probes are active site directed and inactivate a broad range of PTPs in a time- and concentration-dependent fashion. Direct in-gel fluorescence scanning indicate that the fluorescent probes form a covalent adduct with the PTPs and the amount of labeling correlate with PTP activity. As expected, the fluorescent probes can detect on the order of 100 attomole of rhodamine-labeled PTP, a detection limit nearly 3 orders of magnitude more sensitive than that of the biotin-conjugated probes. Moreover, the fluorescent probes also exhibit extremely high selectivity toward PTPs while remaining inert to other proteins. Thus, the rhodamine-containing probes provide superior sensitivity, quantifiability, and throughput for activitybased PTP profiling. Initial proof-of-concept experiments show that the fluorescent probes are capable of monitoring simultaneously the activity levels of PTPs at the whole proteome level and that the activity profiles of PTPs are significantly different among a panel of human cancer cell lines. This highlights the potential to use the fluorescent probes to identify new PTP markers and targets for the diagnosis and treatment of human diseases. Finally, it is shown that the fluorescent probes are capable of monitoring H2O2-mediated PTP inactivation, which should facilitate the study of regulated H2O2 production as a new tier of control over tyrosine phosphorylation-dependent signal transduction. Further application of the activity-based fluorescent probes will accelerate global characterization of PTPs, thereby increasing our understanding of PTPs in cell signaling and in diseases. Abbreviations. PTP, Protein tyrosine phosphatase; DMF, N,N-dimethylformamide; LRSC, lissamine rhodamine B sulfonyl chloride; NHS, n-hydroxysuccinimide ester; GST, glutathione S transferase; DTT, dithiothreitol; pNPP, paranitrophenyl phosphate; EDTA, ethylenediaminetetraacetic acid; TFA, trifluoacetic acid; TEA, triethylamine; TMS-Br, trimethylsilyl bromide; DMSO, dimethyl sulfoxide.

Acknowledgment. This work was supported by National Institutes of Health Grant DK68447 and the G. Harold and Leila Y. Mathers Charitable Foundation. References (1) Tonks, N. K.; Neel, B. G. Curr. Opin. Cell. Biol. 2001, 13, 182195. (2) Li, L.; Dixon, J. E. Semin. Immunol. 2000, 12, 75-84. (3) Zhang, Z.-Y. Curr. Opin. Chem. Biol. 2001, 5, 416-423. (4) Alonso, A.; Sasin, J.; Bottini, N.; Friedberg, I.; Friedberg, I.; Osterman, A.; Godzik, A.; Hunter, T.; Dixon, J. E.; Mustelin, T. Cell 2004, 117, 699-711.

research articles

Activity-Based Fluorescent PTP Probe (5) Zhang, Z.-Y. Prog. Nucleic Acid Res., Mol. Biol. 2003, 73, 171220. (6) Cravatt, B. F.; Sorensen, E. J. Curr. Opin. Chem. Biol. 2000, 4, 663-668. (7) Jeffery, D. A.; Bogyo, M. Curr. Opin. Biotechnol. 2003, 14, 87-95. (8) Kozarich, J. W. Curr. Opin. Chem. Biol. 2003, 7, 78-83. (9) Greenbaum, D.; Medzihradszky, K. F.; Burlingame, A.; Bogyo, M. Chem. Biol. 2000, 7, 569-581. (10) Borodovsky, A.; Ovaa, H.; Kolli, N.; Gan-Erdene, T.; Wilkinson, K. D.; Ploegh, H. L.; Kessler, B. M. Chem. Biol. 2002, 9, 11491159. (11) Yee, M. C.; Fas, S. C.; Stohlmeyer, M. M.; Wandless, T. J.; Cimprich, K. A. J. Biol. Chem. 2005, 280, 29053-29059. (12) Liu, Y.; Shreder, K. R.; Gai, W.; Corral, S.; Ferris, D. K.; Rosenblum, J. S. Chem., Biol. 2005, 12, 99-107. (13) Kumar, S.; Zhou, B.; Liang, F.; Wang, W.-Q.; Huang, Z.; Zhang, Z.-Y. Proc. Natl. Acad. Sci., U.S.A. 2004, 101, 7943-7948. (14) Adam, G. C.; Sorensen, E. J.; Cravatt, B. F. Mol. Cell. Proteomics 2002, 1, 781-790. (15) Taylor, W. P.; Zhang, Z.-Y.; Widlanski, T. S. Bioorg. Med. Chem. 1996, 4, 1515-1520. (16) Patricelli, M. P.; Giang, D. K.; Stamp, L. M.; Burbaum, J. J. Proteomics 2001, 1, 1067-1071. (17) Liu, Y.; Patricelli, M. P.; Cravatt, B. F. Proc. Natl. Acad. Sci., U.S.A. 1999, 96, 14694-14699.

(18) Saghatelian, A.; Jessani, N.; Joseph, A.; Humphrey, M.; Cravatt, B. F. Proc. Natl. Acad. Sci., U.S.A. 2004, 101, 10000-10005. (19) Kidd, D.; Liu, Y.; Cravatt, B. F. Biochemistry 2001, 40, 4005-4015. (20) Adam, G. C.; Sorensen, E. J.; Cravatt, B. F. Mol. Cell. Proteomics 2002, 1, 828-835. (21) Kessler, B. M.; Glas, R.; Ploegh, H. L. Mol. Immunol. 2002, 39, 171-179. (22) Jessani, N.; Liu, Y.; Humphrey, M.; Cravatt, B. F. Proc. Natl. Acad. Sci., U.S.A. 2002, 99, 10335-10340. (23) Hunter, T. Philos. Trans. R. Soc. London B Biol. Sci. 1998, 353, 583-605. (24) Zhang, Z.-Y.; Dixon, J. E. Biochemistry 1993, 32, 9340-9345. (25) Finkel, T. Curr. Opin. Cell Biol. 2003, 15, 247-254. (26) Tonks, N. K. Cell 2005, 121, 667-670. (27) Rhee, S. G.; Kang, S. W.; Jeong, W.; Chang, T. S.; Yang, K. S.; Woo, H. A. Curr. Opin. Cell Biol. 2005, 17, 183-189. (28) Denu, J. M.; Tanner, K. G. Biochemistry 1998, 37, 5633-5642. (29) Barrett, W. C.; DeGnore, J. P.; Keng, Y. F.; Zhang, Z.-Y.; Yim, M.B.; Chock, P. B. J. Biol. Chem. 1999, 274, 34543-34546. (30) Meng, T. C.; Fukada, T.; Tonks, N. K. Mol. Cell 2002, 9, 387399.

PR050449X

Journal of Proteome Research • Vol. 5, No. 8, 2006 1905