IMAP APPLICATION NOTE #6
By Liz Gaudet, Ph.D., Joyce Itatani, Francisco Ramirez and Annegret Boge, Ph.D. Molecular Devices Corporation, 1311 Orleans Dr., Sunnyvale, CA
This application note describes how to set up and run kinase profiling (Profiler) experiments using IMAP. The IMAP fluorescence polarization (FP) assay platform is applicable to a variety of enzymes, especially protein kinases. IMAP is based on the high affinity binding of phosphate to immobilized trivalent metals, rather than on antibody recognition. Thus virtually any peptide substrate can be easily configured to run in an IMAP assay.
This generic feature allows IMAP to be used to test multiple kinases efficiently using IMAP Substrate Finder kits and the IMAP Progressive Binding System. These kinases then become the panel for a Profiler that runs efficiently and precisely in IMAP. You can rapidly test many compounds in the Profiler to assess selectivity, determine IC50s, and assay with a wide range (up F Ser/Thr 2 kit has a protocol and set of peptide substrates distinct from the SF Ser/Thr 1 and SF Tyr kits. These IRAK 4 results are shown in Figure 8, Panel B. The four peptides on the IRAK SF Ser/Thr 2 plate that provided the greatest DmP were ID C3D4, A9B10 and G9H10. A9 B10 is the SF Ser/Thr 2 ID for the same FAM-PKCe pseudosubstrate-derived peptide discussed above, which underscores the reproducibility among IMAP SF kits. There are nine substrates common to both the SF Ser/Thr 1 and SF Ser/Thr 2 kits because they were described as substrates for kinases from both areas of the kinome. (See Figure 1.) Figure 9 shows the corresponding IRAK 4 dilution curves with the SF Ser/Thr 1 and SF Ser/Thr 2 substrates. The FAM-PKCe-derived peptide, of sequence 5FAM-ERMRPRKRQGSVRRRVNH2, had the lowest EC50 of all peptides tested. This peptide was therefore chosen to be the IRAK 4 profiling substrate.
JNKa2 is another kinase that does not yet have available substrates optimized for IMAP. For initial assay development, usable substrates for JNK2a2 were determined with the SF Ser/Thr 2 kit. (See Figure 10.)
As only 0.17 U/mL JNK2a2 was assayed on the SF Ser/Thr 2 plate, few peptides were hits in this assay. The EGFR-derived peptide (well ID E17-F18), an IMAP-validated substrate already in MDC's inventory, provided the sole signal greater than 100 mP. To confirm that this peptide would i ndeed be the profiling substrate chosen, JNK2a2 was titrated with the EGFR-derived peptide and the second- and third-ranked peptides, well IDs C3-D4 and A5-B6, respectively. A5B6, known as CDK7tide, is already in the IMAP inventory. Figure 11 shows that the EGFR-derived peptide performed better than the other two substrates tested, confirming the SF Ser/Thr 2 assay results in Figure 10. This peptide was chosen for the JNK2a2 profiling assay.
The next enzyme on the profiler list in Table 1 is the kinase NEK2. NEK2 can phosphorylate PLM-derived peptide, of sequence 5FAMIRRLSTRRR- COOH, but it would be costefficient to develop a more sensitive assay if possible. To this end, we assayed NEK2 on the SF Ser/Thr 2 plate. (See Figure 12.)
From the histogram shown in Figure 12, peptide with ID G9-H10 had a DmP of 269 mP, the highest substrate signal in the assay. This peptide, of sequence 5FAM-GTFRSSIRRLSTRRRCOOH, has more phosphorylation sites and a sequence extended from the FAM-PLM-derived peptide. The latter peptides DmP in the SF Ser/ Thr 2 assay was only 51 mP (labeled ID G9H10 on the SF Ser/Thr 2 plate). As is evident from Figure 13, peptide G9H10 provides a > 20- fold more sensitive assay than the PLM-derived peptide, significantly reducing the cost-per-well of the NEK2 profile assay.
Another kinase planned for the profiling assay is PAK4. To find the best IMAP substrate for this enzyme, we assayed PAK4 with the SF Ser/Thr 2 kit. (See Figure 14.)
The SF Ser/Thr 2 assay with PAK4 had three peptides with a DmP result greater than 150 mP: well IDs C9D10, O9P10 and A13B14. Peptide C9D10 and O9P10 had very similar results in the SF Ser/Thr 2 assay, which were echoed in the PAK4 titration assay. The peptide with the slightly better EC50 was C9D10, so that substrate was chosen to be the PAK4 profiling peptide.
There is an inventory substrate available for PKCx, which is a PKC isoform that is further activated by lipids. This FAM-PKCe-derived peptide (5FAM-ERMRPRKRQGSVRRRV-NH2) is also the profiling substrate for IRAK4. Figure 16 shows that DAG and PS are not necessary for a robust PKCx assay. For the kinase profiling assay, the lipid activation step will be omitted to streamline the protocol.
For Ros, the next kinase on our list, the substrate Rostide has already been validated in IMAP, so an SF Tyr assay was not necessary. The TIE2 assay was the last on the Table 1 profiler list not yet optimized for IMAP. TIE2 was assayed on the SF Tyr plate at 0.7 U/mL, and the results are presented in Figure 17, Panel A. There are more than a dozen substrates which can be phosphorylated by TIE2 at that concentration, as indicated by a DmP signal greater than 150 mP, so from these results it is difficult to choose the substrate that provides the most sensitive TIE2 IMAP assay. Testing with a lower concentration of TIE2 increased the stringency of the SF Tyr assay. The SF Tyr results with a reduced concentration (0.1 U/mL) of TIE2 are shown in Figure 17B.Table 3. Top-Ranking TIE2 Substrates SF Tyr Well ID Peptide Sequence MDC Part Number (Name) SF Tyr 0.1 U/mL TIE2 (Figure 18B) Using SF Binding Solution TIE2 Titration (Figure 19) Using Optimized Binding Solution Bkgnd FP DmP Binding Solution Bkgnd FP DmP Binding Solution C13 D14 5FAMEFPIYDFLPAKKK-NH 2 R7188 (FAM-Blk/Lyntide) 92 204 40/60/1:1500 103 286 60/40/1:1200 C7 D8 5FAMKKKKEEIYFFFG- NH2 R7269 (FAM-CSKtide) 164 195 60/40/1:1200 119 324 50/50/1:1200 K17 L18 5FAMQEEEYVFIE- NH2 RP7084 (FAMPDGFRtide) 58 144 10/90/1:1500 115 284 30/70/1:1200
Note to Table 3: The SF Tyr kit assay well ID, peptide sequence and part number/name of peptide are shown in Table 3. The Binding Solutions for the titration assay and the SF Tyr assays are also shown. Binding Solutions are listed as % Buffer A/% Buffer B/dilution of BR. For the TIE2 titration assay, the Binding Solutions were optimized slightly from the SF Tyr protocol to reduce the FP backg round and maximize the FP signal.
The three substrates that provided the greatest DmP in the assay shown in Figure 17B were then chosen for a subsequent TIE2 titration assay, presented in Figure 18. All of these substrates are part of the MDC peptide inventory. In the SF Tyr plate, their well IDs are as follows: C13D14 (Blk/Lyntide), C7D8 (CSKtide) and K17L18 (PDGFRtide); their sequences and Binding Solutions are also given in Table 3. The TIE2 substrate that ranked the best in terms of DmP and EC50 was the FAM-Blk/Lyntide, so this peptide was chosen for the profiling assay.
For kinase selectivity profiling, the screen can be set up in a wide variety of formats. It can be designed to test a single candidate inhibitor or activator compound with different concentrations of ATP, titrate the drug candidate with each enzyme, titrate each enzyme with a single concentration of the compound, test multiple compounds with a single ATP concentration and a single kinase concentration, or other protocols may be devised.
Profiling experiments may vary depending on the type of compounds being tested. For example, if the drug candidates to be profiled are known to be ATP-competitive inhibitors, it may be useful to test with two or more concentrations of ATP. To facilitate profiling experiments, kinases with similar KmATPs can be tested together. For determination of the apparent KmATP for each kinase, please refer to IMAP Application Note #3, Determination of Ki and Km values in IMAP Assays, at www.moleculardevices.com.
For our IMAP Profiler, we used the EC50 concentration of each kinase with both 10 M and 100 M ATP. Three known kinase inhibitors (H89, SB203580 and staurosporine) were tested with this IMAP Profiler. The plate set-up is shown in Table 5. IMAP Reaction Buffer containing 0.01% Tween 20 rather than 0.1% BSA as a carrier was utilized in this Profiler to avoid inhibitor binding to BSA. This type of binding could reduce the effective inhibitor concentration and result in a right-shifted IC50 curve. Although assays listed in Table 4 that utilized the 100% Buffer A/1:400 BR Binding Solution equilibrated within one hour, those assays with Binding Solutions containing greater than 25% Buffer B required more time to achieve a stable signal. To simplify the protocol, the entire Profiler assay was incubated overnight with Binding Solution. While there is an optimal IMAP Binding Solution for every substrate sequence, it is possible to optimize Binding Solution compositions to accommodate a variety of substrate sequences with good results. This feature can aid in automation of Profiler Systems with IMAP and simplify assay set-up.
Table 4. Setup of Profiling Screen Enzyme EC50 (U/mL) Special Reaction Requirements Substrate Sequence Substrate Part Number (8000 Test Points) Binding Solution (%A/%B/BR Dilution) 1 AMPK 0.01 50 M AMP 5FAM-LKKLRRRLSDANF-NH2 RP7005 100/0/1:400 2 Aurora B 0.04 - 5FAM-LKKLRRRLSDANF-NH2 RP7005 100/0/1:400 3 CDK5/p25 0.05* - 5FAM-GGGPATPKKAKKL-COOH R7252 100/0/1:400 4 CK1 0.008* - 5FAM-HAAIGDDDDAYSITA-NH2 R7311 40/60/1:2500 5 FAK 0.003 2 mM MnCl2 5FAM-GEEIYGEFD-NH2 RP7060 10/90/1:1500 6 IRAK4 0.05 - 5FAM- ERMRPRKRQGSVRRRV-NH2 RP7048 100/0/1:400 7 JNK2a2 0.04 - LVEPLTPSGEAPNQK-5FAM-COOH R7129 60/40/1:1200 8 NEK2 0.01 - 5FAM-GTFRSSIRRLSTRRR-COOH RP7140 100/0/1:400 9 PAK4 0.09 - 5FAM-KKRPQRRYSNVF-COOH RP7008 100/0/1:400 10 PKCx 0.003 - 5FAM-ERMRPRKRQGSVRRRV-NH2 RP7048 100/0/1:400 11 Ros 0.05 2 mM MnCl2 5FAM-KKKSPGEYVNIEFG-NH2 RP7069 75/25/1:600 12 TIE2 0.05 2 mM MnCl2 5FAM-EFPIYDFLPAKKK-NH2 R7188 60/40/1:1200
Note to Table 4: Table 4 shows the details of the Profiling Screen. There are 12 kinases assayed utilizing 10 different substrates. The AMPK requires AMP in the reaction for full activation, and the three tyrosine kinase reactions contain MnCl2. Binding Solutions are listed as % Buffer A/ % Buffer B/ dilution of BR. The EC50s determined for each kinase are listed. For those with an asterisk, the SF assays and/or titration curves are not shown in this note.
Several Ser/Thr kinases in the profiling assay, including AMPK, Aurora B, IRAK4 and PAK4, were inhibited by staurosporine when tested with at least the low dose of ATP. A representative Aurora B assay is shown in Figure 20, Panel A. Notably, the Ser/Thr kinases CK1 (Figure 20, Panel C) and PKCx (Figure 20, Panel D) were relatively resistant to staurosporine inhibition, which correlates well with published reports.(1,2) Staurosporine clearly reduced the activity of CDK5/p25, as shown in Figure 20, Panel B, consistent with the published in vitro staurosporine IC50 for CDK5/p25 of 4 nM.3
The IMAP profiling assay outlined in Appendix A was next performed with the compound SB203580, which is known to be a relatively specific inhibitor of p38. In the profiling assay shown in Figure 21, SB203580 had little or no effect on several of the kinases, including AMPK, which correlates well with Davies results.4 Conversely, casein kinases are strongly inhibited by this compound,6 and the profili to 1 mM) of ATP concentrations.
Note to Table 1: Many Molecular Devices inventory substrates are also available labeled with 5TAMRA. Those listed as untested were indeed previously so, but are now validated IMAP targets (during the course of this work; see text for details). The kinases range across the kinome to showcase the generic nature of the IMAP platform.
c. Add 83 L of the intermediate 1:10 dilution of BR to the Binding Buffer to make a 1:1200 final dilution of the BR.
Step 4. Prepare 10 mL 40/60/1:2000 Binding Solution.
a. Mix 4 mL of 1X Buffer A with 6 mL 1X Buffer B to make Binding Buffer.
b. Add 50 L of the intermediate 1:10 dilution of BR prepared in Step B to the Binding Buffer to make a 1/2000 final dilution of the BR.
Step 5. Prepare 10 mL 10/90/1:1500 Binding Solution.
a. Mix 1 mL of 1X Buffer A with 9 mL 1X Buffer B to make Binding Buffer.
b. Add 67 L of the intermediate 1:10 dilution of BR prepared in Step B to the Binding Buffer to make a 1:1500 final dilution of the BR.
Step 6. Mix each Binding Solution thoroughly and add 60 L of the appropriate Binding Solution to the appropriate wells, as indicated in Table B.
Note: Those wells to which Binding Solutions containing =<25% Buffer B were added may be read after a 1- hour incubation. Those assays with Binding Solution containing >25% Buffer B will require a longer incubation time to achieve a stable signal. To simplify the protocol, you can incubate the entire Profiler assay overnight and then read the FP just once.
Excitation filter: 485nm-20fwhm
Emission filter: 530nm-25fwhm
Dichroic mirror: 505nm
Excitation filter: 530nm-25fwhm
Emission filter: 590nm-20fwhm
Dichroic mirror: 561nm
Settings for the MDC SpectraMax M5:
Excitation: 485 nm
Emission: 530 nm
Auto cut-off: 515 nm
Excitation: 530 nm
Emission: 590 nm
Auto cut-off: 561
Step 2. Subtract the background value from both S and P raw data.
Step 3. Calculate FP and plot FP against compound concentration.
The histogram in Figure 2 shows the initial SF Ser/Thr 1 assay for AMPK, the first kinase on the Table 1 panel for which a substrate for IMAP has not yet been determined. Please refer to www. moleculardevices.com for a more detailed Substrate Finder assay protocol. The concentration of kinase is expressed in U/mL for all data in this note, where 1 U = 1 nmol phosphate incorporated into a substrate per minute.
Several considerations may influence the kinase concentration chosen for Substrate Finder experiments. 1.0 U/mL of AMPK was selected because it would provide sufficient enzyme to pick up substrate hits but still enable a significant dilution of the stock in the assay. In some of the subsequent Substrate Finder assays presented in this note, we chose to use a lower concentration of kinase, either because our stock of enzyme was limiting, or we had prior knowledge of the sensitivity of at least one of the substrates in the assay plate. In a few cases, it may be necessary to repeat a Substrate Finder assay using a different kinase concentration in order to obtain useful information. In most of the work described in this note, lower enzyme concentrations were employed to increase the stringency of the assay and reduce the number of substrate hits.
The results in Figure 2 show that several of the tested substrates will provide an AMPK activity assay with a DmP of greater than 300 mP at 1 U/mL AMPK. These peptides were then compared with an enzyme titration to identify the most sensitive assay as measured by the EC50. (See Figure 3.) The lowest EC50 in the assay was 0.01 U/mL, obtained with the peptide 5FAM-LKKLRRRLSDANF-NH2 (ID G3H4). This peptide was chosen as the AMPK profiling substrate because its assay both conserved enzyme and maintained a robust signal. As AMPK is maximally active in the presence of AMP, AMP was added to the enzyme titration assay and will be used in the profiling assay at a concentration of 50 M.
The Aurora B assay on the Table 1 panel had also not yet been developed with IMAP. This kinase was tested with the SF Ser/Thr 1 kit to determine efficient substrates.
The Aurora B assays with peptides ID K7L8 (5FAM-GRTGRRNSI-NH2, PKAtide, IMAP validated) and ID G3H4 (5FAMLKKLRRRLSDANF- NH2) resulted in the greatest DmP values on the S F Ser/Thr 1 plate, as shown in Figure 4. Because only one kinase concentration is tested in the Substrate Finder, the SF Ser/Thr 1 results do not completely reflect the efficiency of each substrate; the EC50s must be determined. Therefore, we titrated Aurora B with PKAtide and peptide G3-H4. These results are shown in Figure 5. The performance of the two substrates is very comparable, so we chose the G3-H4 peptide as our Aurora B profiling substrate. Aurora B is the substrate of choice for AMPK, another of our profiling enzymes, thereby reducing the number of different substrates required for the profiling screen.
It is worth noting that, overall, the AMPK SF Ser/Thr 1 assay background and calibrator control FPs (Figure 2) were reproducible in the Aurora B SF Ser/Thr 1 assay (Figure 4). This highlights the ease of assay development with IMAP.
IMAP assays for the kinases CDK5/p25 and Casein Kinase 1 (CK1) have already been developed with MDC inventory substrates, as shown in Table 1. In contrast, a validated IMAP substrate for FAK had not yet been found, so the IMAP SF Tyr plate was used to find one. The SF Tyr assay has a distinct protocol and set of peptide substrates that are different from those of the SF Ser/Thr 1 and SF Ser/Thr 2 kits. As shown in Figure 6, the initial assay with 0.07 U/mL FAK (Panel A) was then repeated with 0.02 U/mL (Panel B). Assaying with the lower concentration of enzyme reduced the number of substrate hits on the plate and facilitated in choosing the profiling substrate.
Table 2. Top-Ranking FAK Substrates SF Tyr Well ID Peptide Sequence MDC Part Number (Name) SF Tyr 0.02 U/mL FAK (Figure 6B) Using SF Binding Solutions FAK Titration (Figure 7) Using Optimized Binding Solutions Bkgnd FP DmP Binding Solution Bkgnd FP DmP Binding Solution G13 H14 5FAM-GEEIYGEFDNH 2 RP7060 (FAM-Src- Familytide) 231 171 40/60/ 1:1500 73 285 10/90/ 1:1500 A11 B12 5FAMGEEPLYWSFPAKKKNH 2 RP7095 (FAMSrctide) 165 146 60/40/ 1:1200 81 257 30/70/ 1:1200 C15 D16 5FAMAEEEIYGEFEAKKKKNH 2 RP7093 230 151 40/60/ 1:1500 73 229 0/100/ 1:1500
Note to Table 2: The SF Tyr kit assay well ID, peptide sequence, and part number /name of peptide are shown. The Binding Solutions for the titration assay and for the SF Tyr assays are also shown. Binding Solutions are listed as % Buffer A/% Buffer B/dilution of Binding Reagent. For the FAK titration assay, the Binding Solutions were optimized slightly from the SF Tyr protocol in order to reduce the FP background and maximize the FP signal.
Serial dilutions of FAK were then tested with the three top substrates ranked by DmP from Figure 6, Panel B. Results of the enzyme titration assay are shown in Figure 7, and Table 2 lists the peptide sequences and their names and well IDs. Two of the top peptide hits with 0.02 U/mL are presently peptides in MDC's inventory; these are the FAM-Srctide and FAM-Src-Family-tide (well IDs A11B12 and G13H14, respectively). As expected the SF Tyr assay with the lesser FAK concentration was a more accurate predictor of EC50 than the 0.07 U/mL assay. The FAK substrate that ranked the best in terms of DmP and EC50 was the FAMSrc- Family-tide, so this peptide was chosen for the profiling assay.
As shown in Table 2, the FAM-Src-Family-tide on the SF Tyr plate had an FP background of > 200 mP with the 40% Buffer A, 60% Buffer B, 1:1500 BR Binding Solution as directed in the SF Tyr protocol. This FP background was optimized by testing the peptide only with different Binding Solutions (data not shown), as outlined in "IMAP Application Note #1: The IMAP Progressive Binding System." To achieve the FP background of 73 mP in the FAK titration assay, we used a Binding Solution of 10% Buffer A, 90% Buffer B and 1:1500 BR. This simple Binding Solution optimization reduced the background FP and maximized the FP signal. The Binding Solutions recommended by the SF Tyr kit protocol are usable but may not be optimal for each substrate; they were devised to allow for acceptable assay windows for all SF substrates while keeping the total number of SF Tyr assay Binding Solutions to a minimum.
The next kinase on the profiling panel without a known IMAP substrate is IRAK 4. As shown in Figure 8, Panel A, three substrates performed significantly better than the others in a SF Ser/ Thr 1 assay with this kinase. These peptides were: ID A11B12, C11D12 and K11L12 on the SF Ser/Thr 1 plate. K11L12 is the FAM-PKCe pseudosubstrate-derived peptide, of sequence 5FAM-ERMRPRKRQGSVRRRV-NH2, and is already available in the IMAP substrate inventory.
We also tested IRAK 4 with the SF Ser/Thr 2 kit for CK1, TKL and CMGC
Ser/Thr Kinases because IRAK4 is a member of the TKL family. The S