Early attrition of drug candidates, including kinase inhibitors, often occurs due to issues that arise during preclinical safety and efficacy evaluation. This problem may be exacerbated by the fact that these studies might fail to consider the basic physiological differences which could exist between human patients and animal models. Little is currently known about how kinase expression can vary between species, and tools to quantitatively measure such differences are lacking. We report the development of a targeted mass spectrometry-based assay capable of monitoring >50 different kinases using peptides conserved in humans and the key preclinical species used in drug development (mouse, rat, dog and cynomolgus monkey). These methods were then used to profile inter-species kinome variability in spleen with three of the current techniques used in targeted proteomics (MRM, PRM and IS-PRM). IS-PRM provided the highest number of kinase identifications, and the results indicate that while this initial set of kinases exhibits high correlation between species for this tissue type, discreet species-specific differences do exist, especially within the cyclin-dependent kinase (CDK) family. Understanding these differences could help rationalize the findings of preclinical studies and have major implications for the selection of these animals as models in kinase drug development.

An UltiMate 3000 UPLC system (Dionex) with the same parameters and configuration across methods and instruments was used to eliminate possible sources of variance. After injection of 2 µL per sample, the peptides were separated on an EASYSpray C18 reverse phase column (75 µm x 25 cm) heated to 50° C with a flow rate of 300 nL/min. The separation was done using a gradient of solvents A (0.1% formic acid, 2% acetonitrile in water) and B (0.1% formic acid, 10% water, 90% acetonitrile) as follows: 1% B for 2 min, then a gradient of 1-45% B over 30 min, then an increase to 90% B for 6 min. After elution from the column, the peptides were introduced to the mass spectrometers with an EASYSpray ion source with an ionization voltage of 1.7 kV. MRM analyses were performed on a TSQ Altis triple quadrupole mass spectrometer (Thermo Fisher Scientific). Collision energy optimization and fragment ion selection (2-4 transitions per peptide) were done by analyzing recombinant kinase digests with Skyline. The MRM method utilized a 1 min scheduling window for each peptide monitored and a cycle time of 1 sec. Alternatively, PRM analyses were performed on an Exploris 480 quadrupole-orbitrap mass spectrometer (Thermo Fisher Scientific). A scheduling window of 1 min was also used for each peptide monitored. The MS2 scans were acquired at a resolution of 30,000 with an automatic gain control (AGC) setting of 1 x 10^5, a maximum injection time (MIT) of 80 msec, and a normalized HCD collision energy (NCE) of 30. Finally, IS-PRM was also performed on an Exploris 480 using the Thermo Scientific SureQuant methodology. A mixture of only SIL peptides in similar biological matrix was used to determine the concentrations and intensity threshold settings needed for optimal triggering. The MS2 scans were acquired at a resolution of 30,000 with an AGC setting of 1 x 10^5, a MIT of 80 msec, and a NCE of 30. The MS1 intensity thresholds for acquisition of the SIL peptides were initially set at 0.005% of the max signal intensity for the SILs obtained from a survey run. These levels were then adjusted empirically to minimize triggering on spurious peaks. An MS1 scan from 300 m/z to 1000 m/z was performed at 120,000 resolution with an AGC target of 1 x 10^6 and a MIT of 50 msec. Upon detection of the SIL peptide precursors in the MS1 scan, a data-dependent MS2 scan of the heavy peptides was acquired at a resolution of 7,500 with an AGC setting of 1 x 10^6, a MIT of 10 msec, and a NCE of 30. Upon detection of 4 out of 6 product ions from the SIL, the light peptides were acquired via a data-dependent MS2 scan with an isolation offset dependent on the terminal amino acid and charge state at a resolution of 60,000 with an AGC setting of 1 x 106, a MIT of 116 msec, and a NCE of 30.

The spleen tissue samples were thawed on ice, diced into small pieces and then homogenized with a Storm Pro bullet blender (Next Advance). First, the pieces were transferred to 1.5 mL tubes (Eppendorf) and then washed 3X with 500 µL of lysis buffer (50 mM EPPS (Alfa Aesar), 100 mM NaCl, 1X Halt protease and phosphatase inhibitor cocktail). After removing the last wash supernatant, fresh lysis buffer was added to each tube which was approximately twice the volume of the tissue piece. One half-scoop of 0.5 mm zirconium oxide beads (Next Advance) was added to each and the tubes were placed in the bullet blender. The samples were homogenized at 4° C with the speed setting at 10 and the time setting at 3, and then visually inspected to ensure absence of particulates. The samples were then spun down at 10,000 g for 10 min at 4° C. The supernatant was moved to fresh tubes with an equal volume of lysis buffer containing 10% sodium dodecyl sulfate (SDS) (Alfa Aesar) preheated to 70° C. The samples were further heated to 90° C for 10 min. Next, the samples were sonicated in a 505 Sonic Dismembrator (Thermo Fisher Scientific) with 30 pulses of one second duration at 90% amplitude followed by a one second pause to facilitate chromatin disruption. The samples were then spun down again at 14,000 g at room temperature for ten min and the supernatant was transferred to fresh tubes. The protein content was quantified using a BCA assay (Thermo Fisher Scientific) according to the manufacturer’s instructions and then aliquoted into 130 µg amounts in 96 µL total volumes (using additional lysis buffer to dilute). The protein was then reduced with the addition of 1 µL of 500 mM dithiothreitol (DTT) and incubated at 56° C for 30 min in a ThermoMixer C (Eppendorf) set at 1000 rpm. The protein was next alkylated with the addition of 3 µL of 500 mM 2-iodoacetamide and incubated at room temperature in the dark for 30 min. The SDS and other chemicals were removed from the samples through a precipitation procedure by first adding 400 µL of methanol, 200 µL of chloroform and 300 µL of water to crash the protein out of solution. After gentle vortexing, the samples were centrifuged at 2000 g at room temperature for 5 min. The top phase was removed and then 300 µL of methanol cooled to -20° C was added to the samples which were then gently vortexed and centrifuged at 9000 g at 4° C for 2 min. After removal of all liquid, a fresh 500 µL of methanol cooled to -20° C was added to the samples. Again, after vortexing, the samples were centrifuged at 9000 g at 4° C for 2 min. Most of the methanol was removed, except for just enough volume to cover the protein pellet, and then the samples were frozen and stored at -80° C until needed. The protein pellets were thawed, and any remaining methanol was removed from the samples. The pellets were resuspended in digestion buffer (50 mM EPPS, pH 8.5) and then spiked with a stock mixture of the stable isotope-labeled (SIL) heavy peptide standards to give final concentrations of 25, 40 or 62.5 nM. These heavy standards containing C-terminal 15N and 13C-labeled arginine or lysine residues, were purchased as >95% pure, 10 µM aliquots in 30% acetonitrile containing 0.1% formic acid (New England Peptide). The samples were diluted further with digestion buffer to a final volume of 100 µL. Any remaining particulates from the pellets were dissolved with sonication in a 505 Sonic Dismembrator (Thermo Fisher Scientific) using 10-15 pulses of one sec duration at 90% amplitude followed by a one sec pause. Finally, 2 µg of Trypsin/Lys-C mix (Promega) was added and the samples were digested overnight at 37° C in a ThermoMixer C (Eppendorf) set at 1000 rpm. The next day, the samples were acidified to 0.1% trifluoroacetic acid (TFA) and then desalted using C18 spin columns (Nest Group, Inc.) according to the manufacturer’s instructions. The peptides were taken to dryness on a Savant speedvac (Thermo Fisher Scientific), and then reconstituted in water containing 0.1% formic acid (Honeywell) to give a final peptide concentration of 0.5 µg/µL (or in the case of the dilution experiments using human peptide digests: 0.5, 0.25, 0.125, 0.05, or 0.025 µg/µL).