Characterization of the PKCε inhibitor CP612. We developed an efficient enantioselective synthesis and carried out structure-activity-relationship studies that led to the development of CIDD-0150612 (CP612; Figure 1A), to improve selectivity and potency for PKCε. CP612 possessed physicochemical properties known to be favorable for CNS drugs (17). These include kinetic aqueous solubility of 7.41 ± 0.12 μM and logD of 1.50 ± 0.75. CP612 showed high plasma protein binding (89.78% ± 1.65%) but good plasma stability and a low microsomal clearance, suggesting low first-pass hepatic metabolism (Supplemental Table 1; supplemental material available online with this article; https://doi.org/10.1172/jci.insight.186805DS1).

Structure and characterization of CP612. (A) Structure and characteristics of CP612. tPSA, topological polar surface area. (B) CP612 inhibited novel PKCs by > 50% at 30 nM but was more potent against PKCε (IC50 = 2.03 nM) than PKCδ (IC50 = 18.8 nM) or PKCθ (30.7 nM). (C) Dose-response curves for inhibition of PKCε by CP612 showing a rightward shift at higher [ATP]. Data are shown as mean ± SEM (n = 3 experiments each done in duplicate or triplicate).

CP612 inhibited PKCε with IC50 = 1.9 ± 0.4 nM (n = 42) and ROCK1 with IC50 = 65.9 ± 10.8 nM (n = 42) making it approximately 55-fold more selective against PKCε over ROCK1. CP612 was screened for kinome specificity at 200 nM against 468 kinases using the scanMax assay panel from Eurofins-DiscoverX. CP612 inhibited 10 other WT kinases besides PKCε, all to < 10% of control: CLK1 (CDC Like Kinase 1), CLK4, PKN1 (Protein kinase N1), PKN2, PKCδ, PKCη, PKCθ, PRKG2 (Protein Kinase CGMP-Dependent 2), ROCK1, and ROCK2. When assayed against several PKC isozymes, CP612 did not inhibit atypical PKCζ, weakly inhibited conventional PKCβII and PKCγ and, within the nPKC subfamily, was most potent against PKCε (Figure 1B).

Since CP612 was derived from Compound 1 and Y-27632, we predicted it would be a competitive inhibitor of ATP binding to PKCε. Therefore, we assayed PKCε activity with increasing concentrations of ATP and in the presence of increasing concentrations of CP612. The EC50 for ATP was 4.2 μM (2.23–7.61 μM, 95% CI) and it increased 40-fold to 168 μM (121–250 μM, 95% CI) in the presence of 300 nM CP612 (Figure 1C). This rightward shift indicates that CP612 inhibits PKCε by competing with ATP.

We next performed a pharmacokinetics study in rats, which showed that CP612 (40 mg/kg) has a plasma half-life of 4.0 ± 0.6 hours after i.v. administration and 6.8 ± 0.7 hours after i.p. administration (Figure 2). The Cmax was 104.5 ± 3.1 μM at the earliest time point (5 minutes) after i.v. administration and was 29.3 ± 1.0 μM between 30 and 60 minutes after i.p. administration. Because binding to plasma protein was approximately 90%, we predicted that the free concentration at each time point was 10 times lower than the measured plasma concentration. After i.v. administration, levels of CP612 decreased between 2 hours and 24 hours by about 70-fold in plasma but only by less than 2-fold in the brain (Figure 2A), resulting in an increase in the brain/plasma concentration ratio from 0.16 ± 0.03 at 2 hours to 9.17 ± 3.19 at 24 hours. After i.p. administration, levels of CP612 decreased between 2 hours and 24 hours by about 30-fold in plasma but remained stable in the brain (Figure 2B), resulting in an increase in the brain/plasma concentration ratio from 0.03 ± 0.009 at 2 hours to 1.18 ± 0.42 at 24 hours. These results indicate that CP612 enters the brain and is cleared from the brain at a much slower rate than from plasma.

Pharmacokinetics of CP612. (A and B) Concentrations of CP612 (40 mg/kg) in male mice after i.v. or i.p. administration declined over time in plasma but remained more stable in the brain up to 24 hours afterward. Data are shown as mean ± SEM (n = 2–6 per group).

CP612 reduces PKCε-dependent hyperalgesia. To directly test if CP612 inhibits PKCε-dependent hyperalgesia, we administered the highly selective PKCε activator, ψεRACK, into the dorsum of the hind paw (1 μg/5 μL, intradermally) of rats. Thirty minutes later, rats were administered vehicle (5 μL, intradermally) or CP612 (1 μg/5 μL, intradermally) (Figure 3). We measured nociceptive thresholds at the same site, at baseline, 30 minutes after the administration of ψεRACK (time 0) and 15, 30, and 60 minutes after the administration of vehicle or CP612. Pain thresholds at baseline and at time 0 did not differ across groups (P = 0.999 and P = 1.000, respectively). ψεRACK induced hyperalgesia across groups (P < 0.001), and this was reduced by CP612 at all time points tested (Figure 3) [F PKCinhibition (1,10) = 78.74, P < 0.001; F Time (6,60) = 19.87, P < 0.0001; F TimexPKCinhibition (6,60) = 7.29, P < 0.001]. Given the near-complete reversal, these experiments indicate that CP612 inhibits PKCε-dependent mechanical hyperalgesia.

Hyperalgesia induced by the PKCε activator ψεRACK. Administration of the PKCε activator ψεRACK (1 μg/5 μL, i.d.) induced hyperalgesia in male rats. This was attenuated by CP612 (1 μg/5 μL, i.d.), up to 60 minutes after its administration. Data are shown as mean ± SEM (n = 6 paws per group). **P < 0.01, ***P < 0.001 compared with the same time points in the Vehicle group using Tukey’s post hoc test.

CP612 reduces paclitaxel-induced hyperalgesia. Repeated administration of the cancer chemotherapeutic drug paclitaxel produces long-lasting mechanical hyperalgesia (18), due to CIPN. We tested if CP612 reverses this hyperalgesia. Male rats received repeated injections of paclitaxel (1 mg/kg, i.p. on days 1, 3, 5, and 7). Twenty-four hours after the last dose of paclitaxel, rats were administered vehicle (2 mL/kg, i.p.) or CP612 (20 mg/kg, i.p.). We measured nociceptive thresholds at baseline, 24 hours after the last injection of paclitaxel (time 0), and 15, 30, 60, and 120 minutes as well as 24 hours after administration of CP612 or vehicle (Figure 4). Pain thresholds at baseline and at time 0 did not differ significantly across groups (P = 0.947 and P = 1.000, respectively). Paclitaxel induced hyperalgesia across groups (P < 0.001), and this was attenuated 15, 30, 60, and 120 minutes after the administration of CP612 (Figure 4) [F PKCinhibition (1,10) = 33.74, P < 0.0002; F Time (6, 60) = 59.35, P < 0.0001; F TimexPKCinhibition (6,60) = 10.19, P < 0.0001]. This decrease in hyperalgesia dissipated 1 day after the administration of CP612. These results indicate that CP612 blocks paclitaxel-induced hyperalgesia, an effect that is no longer present after 24 hours.

Hyperalgesia induced by the chemotherapeutic drug paclitaxel. Repeated administration of paclitaxel (1 mg/kg, i.p. every other day for a total of 4 injections) induced hyperalgesia in male rats. This was attenuated by CP612 (20 mg/kg, i.p.), up to 120 minutes after its administration. Data are shown as mean ± SEM (n = 6 paws per group). **P < 0.01, ***P < 0.001 compared with the same time points in the vehicle group using Tukey’s post hoc test.

CP612 prevents and reverses hyperalgesia induced by morphine withdrawal. Withdrawal from repeated administration of morphine produces hyperalgesia in rodents that is long lasting (19). To examine if CP612 alters this hyperalgesia, we first identified the time points at which hyperalgesia develops in male C57BL/6J mice. Mice received repeated injections of saline (10 mL/kg, i.p.) or morphine (20–100 mg/kg, i.p.) twice daily for 5 days, to induce dependence. Withdrawal from repeated administration of morphine induced hyperalgesia in a time-dependent manner (Figure 5) [F Opioidtreatment (1,10) = 7.85, P < 0.02; F Time (3,30) = 6.03, P < 0.003; F TimexOpioidtreatment (3,30) = 10.67, P < 0.0001]. Specifically, hyperalgesia was not present at 6 hours but was present at 24 hours and 1 week after the last injection of morphine. Repeated administration of saline did not alter pain thresholds at any time tested.

Time course of hyperalgesia induced by morphine withdrawal. After repeated administration of morphine (20–100 mg/kg, i.p.) twice daily for 5 days in male mice, hyperalgesia was not present at 6 hours but was present at 24 hours after the last injection of morphine and persisted for 1 week. No hyperalgesia developed after repeated administration of saline. Data are shown as mean ± SEM (n = 6 per group). **P < 0.01, ***P < 0.001 compared with the same time points in the group receiving repeated injections of saline using Tukey’s post hoc test.

We next investigated if CP612 prevents hyperalgesia induced by withdrawal from repeated administration of morphine, in male and female C57BL/6J mice. Mice received repeated injections of saline (10 mL/kg, i.p.) or morphine (20–100 mg/kg, i.p.) twice daily for 5 days. Six hours after the last injection (i.e., prior to the development of hyperalgesia), mice were administered vehicle (10 mL/kg, i.p.) or CP612 (20 mg/kg, i.p.). Baseline pain thresholds did not differ significantly before treatment with saline, morphine, vehicle, or CP612 and across sexes (all P > 0.99). Withdrawal from repeated administration of morphine produced hyperalgesia that was prevented by administration of CP612, whereas withdrawal from repeated administration of saline did not produce hyperalgesia and was not modified by administration of CP612; these effects did not differ across sexes (Figure 6, A–C) [F Opioidtreatment (1,64) = 50.85, P < 0.001; F PKCinhibition (1,64) = 52.08, P < 0.001; F OpioidtreatmentxPKCinhibition (1,64) = 51.50, P < 0.001; F Time=20.43, P < 0.001; F SexxOpioidtreatmentxPKCinhibition (1,64) = 2.34, P > 0.131; F TimexSexxOpioidtreatmentxPKCinhibition (4,256) = 1.23, P > 0.297]. These findings indicate that a single dose of the PKCε inhibitor can prevent hyperalgesia induced by opioid withdrawal.

Prevention of hyperalgesia due to morphine withdrawal. (A) Experimental timeline. (B and C) Withdrawal from repeated administration of morphine (20–100 mg/kg, i.p.) induced hyperalgesia in male (B) and female (C) mice. This was prevented by CP612, up to 4 weeks after its administration. No hyperalgesia developed after repeated administration of saline. Data are shown as mean ± SEM (n = 11 for each male group, n = 7 for each female group). ***P < 0.001 and **P < 0.01 compared with the same time points in all other groups using Tukey’s post hoc test. For clarity, data from males and females are plotted separately.

We next tested if CP612 reverses hyperalgesia induced by withdrawal from repeated administration of morphine in male and female C57BL/6J mice. Since the pain thresholds of mice receiving repeated administration of saline were not affected by treatment with vehicle or CP612 injection (Figure 6), mice in this experiment were only treated with morphine. Male and female C57BL/6J mice received repeated injections of morphine (20–100 mg/kg, i.p.) twice daily for 5 days. Two weeks after the last injection of morphine (i.e., after the development of hyperalgesia), mice were administered vehicle (10 mL/kg, i.p.) or CP612 (20 mg/kg, i.p.). Baseline pain thresholds did not differ significantly before repeated administration of morphine and across sexes (all P > 0.373). Hyperalgesia was apparent at 24 hours and 1 week after the last injection of morphine (Figure 7, A–C) [F time (5,95) = 0.120.51 P < 0.001]; CP612 reversed this hyperalgesia (Figure 7, A–C) [F PKCinhibition (1,19) = 58.52, P < 0.001], and this occurred similarly across sexes [F sexxPKCinhibition (1,19) = 2.66, P > 0.119] and time [F timexsexxPKCinhibition (5,95) = 0.20, P > 0.960]. Both male and female mice that received vehicle continued to show hyperalgesia up to 4 weeks after the last injection of morphine. In contrast, mice that received CP612 showed a persistent reversal of hyperalgesia. These results indicate that a single dose of the PKCε inhibitor can reverse hyperalgesia induced by opioid withdrawal.

Reversal of hyperalgesia due to morphine withdrawal. (A) Experimental timeline. (B and C) Withdrawal from repeated administration of morphine (20–100 mg/kg, i.p.) induced hyperalgesia in male (B) and female (C) mice. This was reversed by CP612, administered 2 weeks after hyperalgesia was established and lasted for an additional 2 weeks. Data are shown as mean ± SEM (n = 5–7 per group). ***P < 0.001 compared with the same time points in the vehicle group using Tukey’s post hoc test. For clarity, data from males and females are plotted separately.

CP612 has a low addictive potential. We tested the addictive potential of CP612 by examining the extent to which rats are willing to work to obtain CP612. Rats were initially trained to self-administer either saline, morphine (500 μg/kg/i.v. infusion), or 1 of 3 different doses of CP612 (75, 150, or 300 μg/kg/i.v. infusion) for 4 days at fixed ratio 1 (FR1, where 1 nose poke delivers 1 infusion). During this phase, rats learned self-administration behavior, as attested by greater responding in the active hole than the inactive hole [F holetype (1,37) = 129.11, P < 0.001, data not shown]. Intake differed across different treatment groups [F Group (4,37) = 8.50, P < 0.01] in that the group self-administering morphine took fewer infusions (average intake, 9.1 infusions) compared with all other groups, which had similar levels of intake (average intake, 28.0, 29.3, 30.8, and 29.9 for rats self-administering saline or CP612 at 75, 150, or 300 μg/kg/i.v. infusion, respectively).

During the second phase, the ratio to obtain an infusion increased geometrically every other day (FR1, FR3, FR6, FR12, FR24, FR48, and FR96). This increase in ratio produced an increase in responding in the active hole (Figure 8A) [F Ratio (6,222) = 7.75, P < 0.001], and this occurred differentially across groups [F RatioxGroup (24,222) = 15.71, P < 0.001]. Thus, rats self-administering saline or any dose of CP612 did not significantly increase responding in the active hole when the ratio to obtain an infusion increased, whereas rats self-administering morphine showed increased responding with the increase in ratio (Figure 8A). There were no group differences in responding in the inactive hole [F Group (4,37) = 1.40, P > 0.25].

Self-administration of CP612. (A) Increasing the ratio to obtain an infusion across self-administration sessions produced a concomitant increase in responding in the active hole in rats self-administering morphine (500 μg/kg/i.v. infusion) but not in rats self-administering vehicle or different doses of CP612 (75, 150, and 300 μg/kg/i.v. infusion). (B) Rats self-administering morphine had a lower elasticity value (α = 0.4 × 103) compared with all other groups (α = 1.1, 0.9, 0.9, 1.2 × 103 for vehicle and CP612 at 75, 150, and 300 μg/kg/i.v. infusion). They also had a higher price value to switch from inelastic to elastic behavior (Pmax = 50.1) compared with all other groups (Pmax = 4.1, 6.5, 6.1, 2.5 for vehicle and 75, 150, and 300 μg/kg/i.v. infusion). Data are shown as mean ± SEM (n = 10 for vehicle, n = 9 for morphine, n = 7–8 for each CP612 group). ###P < 0.001 compared with their infusions at FR1 and the same ratios between rats self-administering morphine and all other groups using Tukey’s post hoc test. ***P < 0.001 for Pmax and ††P < 0.01 for α, between rats self-administering morphine compared with all other groups using Tukey’s post hoc test. AH, active hole; IH, inactive hole.

The increase in ratio also changed the number of self-infusions [F Ratio (6,222) = 171.68, P < 0.001], and this occurred differentially across groups [F RatioxGroup (24,222) = 6.49, P < 0.001]. Thus, rats self-administering saline or any dose of CP612 showed a similar decrease in the number of self-infusions when the ratio to obtain an infusion increased. In contrast, rats self-administering morphine did not significantly decrease the number of self-infusions until FR98.

Data (number of self-infusions) were also fitted in an economic-demand curve, which provides an estimate of the motivation to obtain drugs and their addictive potential (20). We measured elasticity (α), which is the relative change in consumption as a function of price (i.e., ratio of nose pokes required to obtain an infusion). Behavior is considered “inelastic” when consumption is insensitive to price and “elastic” when consumption is sensitive to price. We also measured Pmax, which is the price at which intake switches from inelastic to elastic. Rats self-administering morphine had a lower elasticity (α) compared with all other groups [F Group (4,37) = 5.12, P < 0.01] and had a greater Pmax [F Group (4,37) = 18.67, P < 0.001] (Figure 8B). Taken together, these findings indicate that CP612 is highly unlikely to be addictive.

CP612 does not alter morphine self-administration. Because many patients with chronic pain take opioids for pain management (8, 9), we wanted to know if CP612 would change opioid intake. Therefore, we investigated if adding CP612 to the morphine solution during self-administration (i.e., a cocktail of morphine and CP612) modifies self-administration of morphine. We tested rats from the prior experiment for self-administration of morphine (500 μg/kg/i.v. infusion). Rats were first tested under a FR6 to reestablish responding. Then, we added vehicle or 3 different doses of CP612 (75, 150, or 300 μg/kg/i.v. infusion) to the morphine solution. We examined its effects during a 4-hour session using a within-session progressive ratio schedule of reinforcement, in which the ratio to obtain an infusion is increased semilogarithmically within a self-administration session as 1, 2, 4, 6, 9, 12, 15, 20, 25, etc., thereby allowing testing over a single day (21). At all tested doses, CP612 did not modify responding, measured as the number of infusions and breaking point (i.e., the highest ratio rats complete to earn an infusion of drug) (Figure 9) [F PKCinhibition (3,29) = 1.67, P > 0.195 and = 1.21, P > 0.324 for infusions and breaking point, respectively].

Self-administration of morphine when adding CP612 to the morphine solution. Adding CP612 (75, 150, and 300 μg/kg/i.v. infusion) to the morphine solution (500 μg/kg/i.v. infusion) during self-administration did not modify responding in a progressive ratio test, where the ratio to obtain an infusion was increased progressively within a self-administration session. This was measured as breaking point (highest ratio reached to earn an infusion of drug) and number of self-infusions. Data are shown as mean ± SEM; dots are scores from individual rats (n = 7–10 per group).

We next investigated if CP612, administered prior to a self-administration session, modifies self-administration of morphine. A separate group of rats was first trained to self-administer morphine (500 μg/kg/i.v. infusion) at FR1 and then at FR3. Rats learned self-administration behavior, as attested by greater responding in the active hole versus the inactive hole [F Holetype (1,17) = 173.96, P < 0.001]; this occurred to a similar extent in rats that would later receive vehicle or CP612 (Figure 10A) [F HoletypexPKCinhibition (1,17) = 0.08, P > 0.78]. The average intake of morphine was 9.26 infusions, and this intake was similar in rats that would later receive vehicle or CP612 [F PKCinhibition (1,17) = 0.02, P > 0.89; F DaysxPKCinhibition (8,126) = 0.65, P > 0.73].

Self-administration of morphine after administering CP612 six hours beforehand. (A) Rats acquired morphine self-administration behavior at FR1, and this responding was increased when the ratio to obtain morphine was increased to FR3. This occurred to the same extent in rats that would later receive vehicle or CP612. AH, active hole; IH, inactive hole. During the progressive ratio tests, the ratio to obtain an infusion was increased progressively within a self-administration session. Administration of CP612 (30 mg/kg, i.p.) 6 hours beforehand did not modify morphine intake or responding. (B–D) This was measured as breaking point (highest ratio reached to earn an infusion of drug) and number of self-infusions the day after the last self-administration session (B), after 1 day of withdrawal (C), or after the administration of naloxone (D) (0.03 mg/kg, s.c.). Data are shown as mean ± SEM; dots are scores from individual rats (n = 6–13 per group). ***P < 0.001 compared with the inactive hole using Tukey’s post hoc test.

CP612 (30 mg/kg, i.p.), administered 6 hours prior, did not modify morphine intake or responding, measured as the number of morphine infusions and breaking point; this was true for all successive progressive ratio tests: test 1 was done the day after initial self-administration (Figure 10B) [F PKCinhibition (1,17) = 0.01, P > 0.94 and = 0.02, P > 0.88 for infusions and breaking point, respectively], test 2 was done after 1 day of withdrawal from self-administration (Figure 10C) [F PKCinhibition (1,17) = 0.16, P > 0.69 and = 0.34, P > 0.56 for infusions and breaking point, respectively], and test 3 was done after the administration of naloxone (0.03 mg/kg, s.c.) (Figure 10D) [F PKCinhibition (1,17) = 1.66, P > 0.219 and = 1.80, P > 0.19 for infusions and breaking point, respectively]. Similar results were obtained when CP612 was administered 18 hours prior to the self-administration session (Supplemental Figure 1).

CP612 reduces somatic signs of morphine withdrawal but not conditioned place aversion. In addition to hyperalgesia, withdrawal from repeated administration of morphine produces somatic signs of withdrawal and conditioned place aversion. We tested the effects of CP612 on these responses. A pilot experiment using C57BL/6 mice showed poor conditioned place aversion, so we used DBA/2J mice for these experiments, since they show more robust conditioned place aversion (22). Male DBA/2J mice received repeated injections of saline (10 mL/kg, i.p.) or morphine (20–100 mg/kg, i.p.) twice daily for 5 days. On day 6 (the conditioning day), mice were pretreated with vehicle (10 mL/kg, i.p.) or CP612 (different doses, i.p.), followed by a single injection of saline (10 mL/kg, i.p.) or morphine (100 mg/kg, i.p.) 4 hours later. Two hours after this last injection of saline or morphine, mice were administered naloxone (5 mg/kg, i.p.) to precipitate somatic signs of withdrawal and conditioned place aversion. Withdrawal from repeated administration of morphine evoked somatic signs of withdrawal (Figure 11, A–C, and Supplemental Figure 2); only the highest dose of CP612 (40 mg/kg) reduced them (Figure 11, A–C) [H PKCinhibition (3, n = 62) = 33.34, P < 0.001]. Withdrawal from repeated administration of morphine also evoked conditioned place aversion (Figure 11, A and C) [F PKCinhibition (5,78) = 7.16, P < 0.001; mice receiving repeated administration of saline with vehicle versus mice receiving repeated administration of morphine with vehicle, P < 0.001], and this was not modified by CP612 at any dose tested (all P > 0.079).

Morphine withdrawal and conditioned place aversion (CPA) precipitated by naloxone. (A) Experimental timeline. (B) On conditioning day, naloxone (5 mg/kg, i.p.) evoked signs of withdrawal in male mice that received repeated administration of morphine (20–100 mg/kg, i.p.) and were treated with vehicle (n = 25) or a low dose of CP612 (10 mg/kg, i.p., n = 15), but this was decreased in mice treated with a high dose of CP612 (40 mg/kg, i.p., n = 10). (C) On test day, prior administration of naloxone produced CPA in mice that received repeated administration of morphine, and this occurred similarly in mice that were treated with vehicle (n = 30) or different doses of CP612 (5, 10, 20, 40 mg/kg, i.p., n = 6, 18, 7, 11, respectively). For B, box and whiskers are median and 25% interquartile intervals; dots are scores from individual mice. For C, Data are shown as mean ± SEM; dots are scores from individual mice. ##P < 0.01 and ###P < 0.001 compared with Saline + Vehicle: ***P < 0.001 compared with Morphine + Vehicle using Tukey’s post hoc test.