Targeting Pim Kinases and DAPK3 to Control Hypertension

Summary

Sustained vascular smooth muscle hypercontractility promotes hypertension and cardiovascular disease. The etiology of hypercontractility is not completely understood. New therapeutic targets remain vitally important for drug discovery. Here we report that Pim kinases, in combination with DAPK3, regulate contractility and control hypertension. Using a co-crystal structure of lead molecule (HS38) in complex with DAPK3, a dual Pim/DAPK3 inhibitor (HS56) and selective DAPK3 inhibitors (HS94 and HS148) were developed to provide mechanistic insight into the polypharmacology of hypertension. In vitro and ex vivo studies indicated that Pim kinases directly phosphorylate smooth muscle targets and that Pim/DAPK3 inhibition, unlike selective DAPK3 inhibition, significantly reduces contractility. In vivo, HS56 decreased blood pressure in spontaneously hypertensive mice in a dose-dependent manner without affecting heart rate. These findings suggest including Pim kinase inhibition within a multi-target engagement strategy for hypertension management. HS56 represents a significant step in the development of molecularly targeted antihypertensive medications.

Graphical Abstract

Introduction

Hypertension, or elevated arterial blood pressure (BP), affects one-third of American adults, whose risk of acute and chronic morbidity correlates directly with increasing BP (Lifton et al., 2001). Chronic hypertension-related pathologies—cardiac and pulmonary arterial hypertrophy (PAH) (Voelkel and Tuder, 1997, Zanchetti, 2010), kidney disease (Crowley and Coffman, 2014), heart disease, and stroke (Lawes et al., 2008)—are invariably associated with hypertrophic remodeling (Briet and Schiffrin, 2013, Sonoyama et al., 2007) resulting from sustained vascular smooth muscle (VSM) hypercontractility (Bai et al., 2013, Uehata et al., 1997). VSM contractility can be modulated by antihypertensive medications that attempt to restore normal BP and reverse remodeling (Briet and Schiffrin, 2013); for example, NO donors, vasodilating β-blockers, Ca²⁺ channel blockers, and renin-angiotensin system (RAS) blockers (Briet and Schiffrin, 2013, Garcia-Donaire et al., 2011). However, despite widespread usage, antihypertensive medications restore normal BP in fewer than 50% of patients (Crowley and Coffman, 2014, Gu et al., 2012). Side effects (Wiysonge and Opie, 2013) and drug resistance (Calhoun et al., 2008) further reduce the benefits of current drugs.

Approximately 1% of chronic hypertension patients experience acute, rapid elevation of BP (>160/100 mmHg) with potentially fatal end-organ damage, termed hypertensive emergency (HE) (Pollack and Rees, 2008, Tulman et al., 2012). Clinical management of HE is difficult because immediate therapeutic lowering of BP can cause end-organ ischemia. For efficient titration, the ideal HE therapeutic would be dose dependent, with rapid onset and rapid clearance. However, most HE therapies, such as nitroprusside and dihydropyridine analogs, are limited by poor pharmacokinetics or narrow therapeutic windows and are not applicable across the broad range of HE comorbidities. Clevidipine, a Food and Drug Administration (FDA)-approved Ca²⁺ channel antagonist for the treatment of HE, has high therapeutic value; however, side effects (e.g., atrial fibrillation and sinus tachycardia [Keating, 2014], resistance, and drug interactions) have recently been observed (Jacklen et al., 2014). New therapeutic targets and medications for modulating VSM contractility to control both acute and chronic hypertension remain an urgent necessity (Pijacka et al., 2016, Retailleau et al., 2015).

Most antihypertensive medications act at the receptor or second messenger level (RAS, adrenoceptors, Ca²⁺, Na⁺/K⁺) and therefore have broad physiological effects. A more focused and largely unexplored strategy involves inhibition of downstream regulators of VSM contraction. In VSM, contractility is governed by homeostatic control of myosin light chain (LC20) phosphorylation (MacDonald et al., 2016, Somlyo and Somlyo, 1994). LC20 phosphorylation by myosin light-chain kinase (MLCK) induces contraction and is reversed by myosin light-chain phosphatase (MLCP) (Butler et al., 2013, Hartshorne et al., 1998, Sherry and Hartshorne, 1980). Whereas MLCK is dependent upon intracellular free Ca2+ ([Ca2+]i), MLCP is [Ca2+]i independent and controlled by its myosin targeting subunit (MYPT1). MYPT1 phosphorylation inactivates MLCP, suppresses myosin dephosphorylation, and triggers hypercontractility in the absence of elevated [Ca²⁺]i (Ca²⁺ sensitization) (Goulopoulou and Webb, 2014, Kitazawa et al., 1989). Sustained Ca²⁺ sensitization induces remodeling and disease progression (Sonoyama et al., 2007); therefore, MYPT1 and its immediate regulators represent critical points of intervention for antihypertensive medications.

Rho-associated kinase (ROCK)-induced MYPT1 phosphorylation and Ca2+ sensitization has been well characterized (Bai et al., 2013, Uehata et al., 1997, Walsh and MacDonald, 2016). Although ROCK inhibition normalizes BP in hypertensive rodent models, the suboptimal selectivity of many ROCK inhibitors diminishes their therapeutic value (Bain et al., 2007). We have instead focused on death-associated protein kinase 3 (DAPK3), also named zipper-interacting protein kinase (ZIPK) (Carlson et al., 2013, Haystead, 2005, MacDonald et al., 2001a, MacDonald et al., 2016, Turner et al., 2015), a Ser/Thr kinase and downstream target of ROCK that promotes Ca²⁺ sensitization in vivo (Borman et al., 2002) by phosphorylating both MYPT1 and the MYPT1-inhibitor CPI-17 (MacDonald et al., 2001b). The DAPK family (DAPKs 1, 2, and 3) possess identical nucleoside binding residues and are not readily discriminated by ATP competitive inhibitors. However, compared with other kinase families, the DAPK3 catalytic domain contains numerous structural features that render it amenable to selective inhibition (Temmerman et al., 2013). Within the DAPK family only DAPK3 is expressed in VSM, where its effects are regulated by phosphorylation at Thr180, Thr225, and Thr265 (Graves et al., 2005).

Within the kinome, the catalytic domain of DAPK3 is highly similar in sequence and structure to Pim kinases (Manning et al., 2002). These Ser/Thr kinases have been previously associated with cell survival and proliferation by regulation of apoptosis and division (Bachmann and Moroy, 2005, Mukaida et al., 2011, Nawijn et al., 2011). As such, they have been the focus of discovery efforts for cancer therapeutics (Braso-Maristany et al., 2016, Pogacic et al., 2007). In addition to cancer tissues, the Pims (Pim-1, −2, and −3) are constitutively active and transcriptionally regulated (Willert et al., 2010) within human cardiac, skeletal, and VSM tissues (Muraski et al., 2007, Renard et al., 2013) (Figure S1). Pim-1 plays a significant role in VSM remodeling (Liang and Li, 2014) and in the pathogenesis of PAH (Paulin et al., 2011a, Paulin et al., 2011b); however, these effects were not previously linked to contractility, but rather to pro-proliferative effects of the kinase (Katakami et al., 2004, Zippo et al., 2004).

We reasoned that if Pims modulate VSM contractility, then multi-target Pim/DAPK3 inhibition may substantially reduce BP in vivo. To test this hypothesis, we developed key inhibitors for probing Pim/DAPK3 inhibition from our lead molecule, HS38 (1) (Figure 1A). As a specific inhibitor of DAPKs and Pim-3, HS38 displayed higher specificity and greater potency, and lacks chemical liabilities inherent in other DAPK inhibitors (Carlson et al., 2013). In VSM tissues, HS38 delayed agonist-induced contraction and suppressed Ca²⁺ sensitization (MacDonald et al., 2016).

Figure 1. Structure and In Vitro Characterization of Lead Compound and Second-Generation Inhibitors.

(A) Lead compound HS38 with zones of variability (1); dual Pim/DAPK3 inhibitor HS56 (2); DAPK3 inhibitors HS94 (3) and HS148 (4); and inactive analog HS182 (5).

(B) Heatmap representing residual kinase activity following treatment with all analogs (10 μM, n = 2) in a radioactive [³²P]ATP filter-binding kinase assay. (−), no inhibitor. (+), no kinase.

(C) Kinase inhibition isotherms generated from titrating compounds 1–4 against DAPK3, Pim-1, Pim-2, and Pim-3 using radioactive [³²P]ATP filter-binding assay. Data points represent mean ± SEM (n = 2).

Herein we report an investigation into Pim kinases as modulators of VSM contractility. Using a co-crystal structure of HS38 in complex with DAPK3, a second-generation dual Pim/DAPK3 inhibitor HS56 (2) and selective DAPK3 inhibitors HS94 (3) and HS148 (4) were developed and used to gain mechanistic insight into the contribution of Pims to VSM contraction and BP modulation. Characterization in vitro, ex vivo, and in vivo suggests that Pims directly modulate VSM contractility and, together with DAPK3, represent polypharmacological targets for the treatment of chronic hypertension.

Results

Discovery of HS56, HS94, and HS148

Small-molecule inhibitors are essential for understanding the underlying mechanisms of Pim and DAPK activity within VSM and for determining their therapeutic value in hypertensive models. HS56, HS94, and HS148 were developed by the introduction of diverse functionality at three variable zones around the pyrazolo[3,4-d]pyrimidinone scaffold of HS38 (Figure S2). Resulting analogs (Table S1) were evaluated using a radioactive [³²P]ATP filter-binding kinase inhibition assay (Hastie et al., 2006) to determine inhibitory activity ([Analog] = 10 μM) versus DAPK3, Pim-1, Pim-2, and Pim-3 (Figure 1B, a subset of Figure S3). A subset of analogs displaying >80% inhibitory activity toward DAPK3 were titrated in the same assay (Figure 1C, a subset of Figures S4A–S4D) to determine inhibition constants (Ki) (Figure 2A, a subset of Figure S4F).

Figure 2. Potency and Selectivity.

(A) Inhibition constants (Ki) for analogs 1–5. EC₅₀ values derived from kinase inhibition isotherms (Figure 1C) were converted to K_i values using the Cheng-Prusoff equation (Cheng and Prusoff, 1973) (Figure S4G).

(B–D) Primary KINOMEscan profiling of HS56 using a competition binding assay. (B) Full kinome profile of HS56. %Control = 100 × (HS56 signal − positive control)/(negative control − positive control). (C) Subset of data from (B) showing VSM active kinase families (green) and kinases for which %Control is <10 (red). (D) Dendrogram of human kinases showing a subset of data from (B).

(E) Inhibition of ionotropic and G-protein-coupled receptors by HS56. Data points represent mean SEM (n = 4).

Key improvements in DAPK3 potency resulted from HS94 (3) and HS148 (4): K_i = 126 nM and 119 nM, respectively. Dual Pim/DAPK3 inhibitor HS56 (2) maintained potency toward DAPK3 (Ki = 315 nM) and was most potent versus Pim-3 (K_i = 72 nM) with micromolar potency toward Pim-1 (Ki = 1.5 μM) and Pim-2 (Ki = 17 μM). These second-generation inhibitors have high therapeutic potential and served as key molecular probes to investigate the effects of Pim/DAPK3 inhibition and selective DAPK3 inhibition on VSM contractility and hypertension.

HS56 displayed a high degree of selectivity for DAPKs and Pims. HS56 was evaluated in an active site-directed competition binding assay (KINOMEscan; DiscoverX, Fremont, CA). Of the 468 kinases assayed, HS56 competitively inhibited only seven with %Control < 10 (Figure 2B). This subset contained five desired targets (Pim-1, Pim-3, DAPK-1, −2, and −3) and two off-target interactions; non-receptor tyrosine-protein kinase 2 (TYK2) and cyclin-G-associated kinase (GAK) (Figure 2C). TYK2 is a member of the JAK family and is not likely relevant to smooth muscle contraction (Hubbard, 2018). Moreover, HS56 displayed affinity toward inactive TYK2 (JH2domain-pseudokinase) and not catalytically active TYK2 (JH1domain-catalytic) (%Control = 92) (Table S3). GAK regulates endocytosis and uncoating of clathrin-coated vesicles (Neveu et al., 2015) and is also not likely to regulate VSM contraction. Additionally, TKY2 and GAK are located on remote branches of the human kinome dendrogram and are dissimilar in sequence to members of the CAMK subgroup (Figure 2D). The quantitative selectivity score, S(10) (number of non-mutant kinases with %Control <10 )/(number of non-mutant kinases tested) for HS56 is similar to lapatinib (Karaman et al., 2008), the most selective, FDA approved kinase inhibitor (S(10)_HS56 = 0.017 and S(10)_lapatinib = 0.010).

To rule out activity on cell surface receptors, we assayed HS56 by the PDSP database (Department of Pharmacology, UNC, Chapel Hill, NC) against ionotropic and G-protein-coupled receptors (Figure 2E). HS56 displayed no significant (>50%) inhibition or activation of nicotinic, adrenergic, or muscarinic receptors at 10 μM.

Structural Basis of DAPK3 and Pim Inhibition

The crystal structure of human DAPK3 catalytic domain in complex with HS38 (PDB: 5VJA) (Figure 3) (CRELUX, Martinsried, Germany) elucidated intermolecular forces driving nanomolar affinity between HS38 and DAPK3 and delineated uniquely targetable features. Our structure is consistent with the published DAPK3 catalytic domain (PDB: 1P4F) (Velentza et al., 2003) with a root-mean-square deviation (RMSD) of 1.15 Å considering 252 Cα atoms. DAPK3 adopts a bilobal structure typical of kinases with a smaller β-stranded N-terminal lobe and a larger α-helical C-terminal lobe. HS38 sits within the ATP binding pocket at the junction of these two lobes.

Figure 3. Crystal Structure of DAPK3 in Complex with HS38 Defines Domains of Interaction.

(A) Overlaid ribbon and surface diagrams of the DAPK3 catalytic domain (red). HS38 (orange) is bound to the ATP binding cleft (cyan).

(B) Detailed view of HS38 (orange) with interacting resides (cyan sticks), H bonds (black lines), and water molecules (red spheres).

(C) 2D depiction with color-coded domains of DAPK3-HS38 interaction: direct H bonding (dashed black line), hydrophobic interactions (magenta). For data collection and refinement, see Table S2.

DAPK3 contains uncommon features that render it distinct from other members of the human kinome (Figure S5A). Regulatory phosphorylation sites within its ATP-binding pocket are absent. Its gatekeeper residue (Leu93) is conserved in only 18% of kinases (Moffat et al., 2011, Yokoyama et al., 2015). The adenine binding pocket hinge region contains a non-conserved residue (Glu94), which H-bonds to adenine-NH₂ of ATP, forming a canonical donor-acceptor pair. Additionally, the phosphate binding pocket contains a key residue (Lys42), which forms only one H bond to ATP instead of two (Tereshko et al., 2001). Thus, ATP competitive inhibitors may target DAPK3 separately from other closely related kinases (Huang et al., 2010), including the Pims, despite structural similarities and proximity on adjacent sub-branches of the CAMK region of the kinome dendrogram (Manning et al., 2002).

Analysis of non-covalent interactions between HS38 and DAPK3 (Figure 3B) suggested water-mediated H bonds to (1) hinge region backbone between aryl chlorine and Val96/Gly99; (2) adenine binding pocket backbone between pyrazole N2 and Glu94/Val96; (3) phosphate binding pocket between pyrimidinone oxygen and Glu64(N^ɛ)/Phe162; and (4) between thioether carboxamide and Asn144/Asp161. A single direct H bond was present between the thioether and Lys42 (Figure 3C). Notably absent are canonical donor-acceptor H bonds to the hinge region backbone (Figure S5B). The HS38 chlorophenyl ring is sandwiched between Leu19 and Ile160, providing hydrophobic stabilization.

Pim kinases share a high degree of sequence similarity within the Pim family (Bachmann and Moroy, 2005), but only ∼30% overall sequence similarity with other kinases (Jacobs et al., 2005). The activation loop of Pim-1 contains phosphorylation sites; however, Pim kinases are constitutively active, regardless of activation loop phosphorylation, due to stabilizing interactions with the catalytic loop (Merkel et al., 2012, Qian et al., 2005).

Our DAPK3 structure was compared with Pim-1 to give structural insight into the potency of key analogs. The Pim-1 ATP-binding cleft shares ∼75% sequence similarity with DAPK3 (Figures 4A and 4B). Although DAPK3 overlays closely with Pim-1 (RMSD between 171 atom pairs is 1.14 Å), detailed structural analysis revealed several targetable features (Figures 4C and 4D). Like DAPK3, Pim-1 adopts a DFG-in conformation. The glycine-rich hydrophobic pocket at the entrance region of Pim-1 (residues 44–52) is significantly wider than in DAPK3 (residues 19–27) (Jacobs et al., 2005, Yokoyama et al., 2015) giving Pim-1 greater accessibility. The narrower loop of DAPK3 may provide hydrophobic stabilization with the propanamide methyl of HS38, a potential structural basis for selectivity. The wider loop on Pim-1 likely accommodates a wider range of thioether side chains. An unusual hinge region bulge, conserved within Pim kinases, forms only one backbone H bond with ATP due to unique proline resides (Pro123 and Pro125). This unique hinge likely accommodates structural diversity different from that of DAPK3. Constrained rotation also creates an unusual hinge bulge potentially weakening hydrophobic stabilization with the chlorophenyl moiety of HS38, which packs tightly into the hydrophobic cleft of DAPK3.

Figure 4. Structural Analysis of DAPK3 and Pim-1 Reveals Uniquely Targetable Features.

(A) CLUSTAL 2.1 multiple sequence alignment of DAPK3, Pim-1, Pim-2, and Pim-3.

(B) Table of amino acid substitutions within the ATP binding pockets of DAPK3, Pim-1, Pim-2, and Pim-3.

(C) Overlaid catalytic domains of DAPK3-HS38 (red) and Pim-1 bound to AMPPNP (gray, ligand not shown, PDB: 1XR1) (Qian et al., 2005). HS38 (orange) sits within the ATP binding cleft of DAPK3 (cyan). Unique residues within the Pim-1 ATP binding pocket (yellow).

(D) Overlaid ATP binding pockets of DAPK3 (cyan) and Pim-1 with conserved residues (gray) and non-conserved residues (yellow, labels). HS38 (orange) is bound to DAPK3 and water molecules (red spheres) by H bonds (black lines).

Structural Insight into Analog Potency

All analogs share an identical pyrazolo[3,4-d]pyrimidin-4-one core scaffold with three zones of variability (Figure 1A). HS56 displays enhanced potency toward Pims and is identical to HS38 in all respects except its cyanomethyl thioether substituent (zone 2), which likely orients readily within the wider glycine-rich loop of Pims.

HS94 and HS148 are more potent toward DAPK3 and less potent toward all Pims. Both have additional hydrophobic surface area (α-dimethyl and α-ethyl) on zone 2 that likely stabilizes the more confined hydrophobic loop of DAPK3. Both analogs lack an aryl chloride substituent (zone 1), suggesting that smaller meta-aryl substitution may increase potency toward DAPK3 while larger substituents may orient favorably within the hinge region bulge of the Pims. The loss of water mediated H-bonding between HS94/HS148 and Gly99/Val96 of DAPK3 is likely offset by additional hydrophobic interactions.

Zone 3 contains an oxygen substituent in the 4-position that participates in two water-mediated H bonds to DAPK3, contributing substantially to the energetics of binding. This pyrimidone oxygen is otherwise surrounded by hydrophobic residues. Replacement with -NH₂ or alkylated amines proved deleterious to inhibitor potency and led to inactive analog HS182, used here as a negative control.

In general, numerous hydrophobic residues on DAPK3 are replaced with hydrophilic, basic, acidic, and proline residues on Pim-1. Altogether, these combined structural differences likely account for the higher potency of HS94 and HS148 for DAPK3 and the loss of potency toward Pims. HS148 and HS94 were essential for determining the effect of DAPK3 inhibition on VSM contractility in the absence of Pim inhibition. HS56, our most potent Pim inhibitor, remains potent toward DAPK3 and was a key molecular probe for elucidating the effect of dual Pim/DAPK3 inhibition on VSM contractility and hypertension.

Modulation of VSM Contractility

Precise balancing of LC20 phosphorylation modulates the kinetics and intensity of smooth muscle contraction (Walsh, 2011). LC20 phosphorylation and contraction dynamics were characterized in excised arterial VSM. Rat caudal arterial smooth muscle strips were cleaned of adventitia and endothelium denuded. Contraction was induced with the phosphatase inhibitor, calyculin A (Cla) (Figure 5A). Pretreatment with selective DAPK3 inhibitors (HS94, HS148), Pim-3/DAPK3 inhibitor (HS38), and inactive control (HS182) resulted in similar contractile morphologies; Cla-induced force increased to a stable plateau greater than K⁺-induced reference force. Dramatic differences were observed after pretreatment with HS56; force increased to an unstable maximum never greater than K+ reference. LC20 phosphorylation within HS56-treated vessels was reduced to near baseline levels, with unphosphorylated and monophosphorylated LC20 predominating (Figure 5B), whereas no reduction was observed within other treatment groups.

Figure 5. HS56 Delayed Force Onset, Decreased Contractile Force, and Reduced LC20 Phosphorylation in Excised Rat Caudal Arterial VSM Tissues.

(A) Representative calyculin A (Cla)-induced contractile traces following treatment with vehicle (DMSO) or inhibitor.

(B) Phos-tag SDS-PAGE western blot analysis of LC20 phosphorylation from non-phosphorylated (0P) to triphosphorylated (3P) following treatment with vehicle (NH), Cla, or Cla + inhibitor.

(C) Effect of inhibitors on t_{1/2 max}, maximum force, and total phospho-LC20/total LC20 (determined by scanning densitometry quantitation) from Cla-stimulated contractions.

(D) Titration curves derived from multiple contractile traces in the presence of Cla + inhibitor.

Mean ± SEM, n = 3. Significantly different from Cla (ANOVA, Dunnett’s post hoc test, n > 3), *p < 0.002.

Few differences in contraction rates were observed between inhibitors, all of which increased the time required to reach 50% of maximal Cla-induced contraction (t_{1/2 max}) (Figure 5C). Concentration-dependent studies confirmed that HS56, HS94, and HS148 acted similarly to delay force development (Figure 5D). Similar trends were observed for t_{1/2 max} after contraction and latency (Figures S6A and S6B). All three analogs were active at concentrations as low as 1 μM, ex vivo potency typical of bioavailable, ATP-competitive kinase inhibitors with nanomolar activity in vitro, given intracellular ATP ranges from 1 to 10 mM (Beis and Newsholme, 1975).

Remarkable differences between Pim/DAPK3 inhibition and selective DAPK3 inhibition were instead manifested in contractile intensity and LC20 phosphorylation. Maximum Cla-induced force, expressed as percentage of K⁺-induced reference force, was significantly reduced (>75%) by HS56, whereas no reduction was observed with other inhibitors. HS56 caused significant concentration-dependent force reduction at [HS56] >10 μM with ∼3-fold reduction observed at [HS56] >75 μM. HS56 caused 4-fold reduction in LC20 phosphorylation in the same tissues. Similar trends were not evident with HS94 and HS148 at concentrations up to 100 μM. These results provided initial indication that HS56 modulates LC20 phosphorylation and VSM contractility and that Pim/DAPK3 inhibition affects VSM in a fundamentally different manner from selective DAPK3 inhibition.

Mechanistic Insight into Pim Activity

We hypothesized that Pim kinases play a role in VSM contraction by directly phosphorylating regulatory residues on LC20 (Thr18, Ser19) and/or MYPT1 (Thr697, Thr855), thereby suppressing MLCP activity and shifting the balance of myosin to the phosphorylated state. Because all Pims lack a calmodulin regulatory domain, this mechanism of action must be Ca²⁺ independent. Indeed, ex vivo experiments showed that DAPK3 and Pim/DAPK3 inhibitors affect neither kinetics nor intensity of supramaximal Ca²⁺ (pCa4.5)-induced contractions of rat caudal arterial strips. However, whether coincident Pim/DAPK3 inhibition acts to decrease Ca2+ sensitivity with a rightward shift in the pCa-tension response curve remains to be determined (Figures S6C and S6D).

The coulombic surfaces of DAPK3 and Pim-1 (complexed with PimTide) (Pogacic et al., 2007) showed acidic, negatively charged peptide binding grooves traversing their respective C-terminal catalytic lobes (Figures 6A and 6B). Basic substrates align within these grooves with phosphorylation sites oriented toward the γ-phosphate of ATP. DAPK3 displays high affinity for basic substrates within muscle, e.g., MYPT1 and LC20, and Pim kinases display similarly high affinity for basic substrates of the general form RXRHP(S*/T*)G (Pogacic et al., 2007), e.g., Dundee substrate (Bain et al., 2003) and Pimtide. Sequence alignment of these peptide substrates shows numerous basic residues N-terminal to their phosphorylation sites (Figure 6C), which implies that Pim kinases likely display affinity for VSM substrates.

Figure 6. Pim Kinases Directly Phosphorylate VSM Substrates.

(A and B) Structural similarities between peptide binding grooves (yellow box) of DAPK3 and Pim-1. (A) DAPK3 coulombic surface with HS38 (gray sticks). (B) Coulombic surface of Pim-1 with PimTide substrate (gray sticks) (PDB: 2C3I) (Pogacic et al., 2007); (−), acidic (red); (+), basic (blue), neutral (white); black arrows indicate approximate location of peptide phosphorylation site..

(C) Pim substrates (PimTide, Dundee Substrate) and DAPK3 substrates (MYPT1, LC20) contain basic residues N-terminal to their phosphorylation sites (Ser*/Thr*)..

(D) Phos-tag SDS-PAGE western blot depicting phosphorylation of LC20 by MLCK and Pim kinases..

(E) Phos-tag SDS-PAGE western blot depicting phosphorylation of MYPT1 by Pim kinases and regulators of VSM contractility (ROCK, DAPK3). Inhibition by HS56, but not inactive analog HS182; [Analog] = 50 μM.

The ability of Pims to directly phosphorylate LC20 and MYPT1 was investigated in vitro (Figures 6D and 6E, a subset of Figure S7). Phosphorylation of LC20 by Pim-1 was minor, compared with MLCK, after 60 min. No evidence of LC20 phosphorylation by Pim-2 was observed. LC20 phosphorylation by Pim-3 over 60 min was incomplete; however, mono- and diphosphorylated products were clearly visible. Although relatively slow, Pim-3 activity likely contributes to contraction and maintenance of vascular tone.

All three Pims rapidly and robustly phosphorylate MYPT1 at multiple sites on time scales similar to, or faster than, DAPK3 and ROCK. Evidence of phosphorylated products was clear after 30 s and nearly complete conversion to monophosphorylated MYPT1 was apparent after 0.5–5 min. Pim-2 produced primarily mono- and diphosphorylated MYPT1 products while Pim-1 and Pim-3 produced tri- and tetraphosphorylated products. All Pim kinases were inhibited by HS56, but not by inactive analog HS182. DAPK3 activity was also inhibited by HS56, while ROCK activity was unaffected. These results strongly suggest that Pim kinases are regulators of VSM contractility via direct phosphorylation of MYPT1.

HS56 Normalizes BP In Vivo

Having established that HS56 modulates VSM contractility in tissues, we tested whether Pim/DAPK3 inhibition lowered BP in spontaneously hypertensive renin transgene (RenTg) (Caron et al., 2002) and wild-type (WT) mice. Acute systemic vasodilatory effects were observed with HS56, but not HS148. Both HS56 and HS148 were formulated as DMSO solutions. Intravenous infusion of HS56 into RenTG mice induced dose-dependent decreases in systolic blood pressure (ΔSPB = −20 ± 7 and −28 ± 3 mmHg, at 10 and 20 mg/kg, respectively, measured 2.2 min after infusion) without decreasing the heart rate (HR) (Figures 7A–7D). HS56 (20 mg/kg) lowered SBP of RenTg mice to near WT levels (from ∼119 mmHg to ∼91 mmHg) while HR increased slightly (from ∼412 bpm to ∼436 bpm), measured 2.2 min after infusion. HS56-associated HR increases were likely compensatory effects caused by vasodilation. In WT mice, HS56 affected neither HR nor SBP at 10 mg/kg and, at 20 mg/kg, decreased SBP from ∼99 mmHg to ∼76 mmHg (ΔSBP = −23 ± 5 mmHg) without changing HR. Neither HR nor SBP of RenTg and WT mice were significantly affected by HS148 (10 mg/kg).

Figure 7. HS56 Lowers BP in Spontaneously Hypertensive Mice.

HR and BP data recorded during intravenous bolus delivery of vehicle (DMSO; n ≥ 4), HS148 (10 mg/kg; n = 3), and HS56 (10, 20 mg/kg; n ≥ 3) into anesthetized, spontaneously hypertensive RenTg and WT mice.

(A) Representative SBP and HR tracings from single RenTg (top) and WT (bottom) mice. Black arrows indicate time of infusion. DMSO effect on HR (gray arrows) caused reversible lowering of SBP.

(B) Changes in HR and SBP during infusions into RenTg (left) and WT (right) mice. Data points are binned averages of 1 s from every 5 s of data. Black arrows indicate time of infusion and gray arrows the DMSO effect.

(C) Average change in HR at t = 2.5 min.

(D) Average change in SBP at t = 2.5 min.

Data points in (B) to (D) are mean ± SEM. Significantly different from DMSO (two-way ANOVA), *p < 0.05, **p < 0.01, ***p < 0.001.

Interestingly, vehicle (DMSO) alone caused an immediate drop in HR and SBP that was profound, short-lived, and immediately reversible (DMSO effect). This is likely due to jugular vein delivery of undiluted DMSO (30–60 μL) directly into the heart, causing transient slowing of HR, which lowered BP. Upon recovery, HR and BP returned to stable baseline levels (∼0.5 min and ∼1.5 min after infusion, respectively). As expected, the DMSO effect was also observed after infusion of inhibitors; however, HS56 effects persisted for the duration of monitoring and were readily distinguishable from DMSO effects. These results suggest that the combined activity of DAPK3 and Pim kinases contributes to BP regulation and hypertension in vivo and that the pathophysiology of hypertension can be acutely abrogated by HS56.

Discussion

Chronic VSM hypercontractility leads to hypertension, vascular hypertrophy, and significant cardiovascular disease. The contribution of Pim kinases to VSM contractility and pathogenesis of hypertension has been largely uncharacterized. Here we developed and characterized a set of molecular probes, which were utilized to demonstrate that (1) Pims directly target VSM substrates in vitro; (2) Pim/DAPK3 inhibition by HS56 critically modulates LC20 phosphorylation and contractility in tissues; and (3) HS56 regulates BP in vivo without affecting heart rate.

Based on molecular properties, kinome profiling, kinase assays, and ex vivo studies, HS56 satisfies the criteria for high-quality chemical probes (Arrowsmith et al., 2015). In brief, these include: (1) defined mode of action; (2) selective: S(10)_HS56 = 0.017 from 468 assays; (3) synthetically available; (4) defined structurally inactive analog (HS182); and (5) demonstrated effects in tissues at 1 μM. Based on our findings, HS56 is suitable for characterization of dual DAPK/Pim inhibition in VSM systems and hypertension.

HS94 and HS148 show a high degree of selectivity for DAPK3 (2.5- to 3-fold more potent than HS56), low potency for Pim-3 (15- to 35-fold less potent than HS56), and low micromolar to no activity versus Pim-1 and Pim-2. Thus, HS94 and HS148 are sufficiently selective for probing DAPK3 inhibition in the absence of Pim inhibition in tissues and animals.

All Pim kinases phosphorylated MYPT1 in vitro (Figure 6E). HS56 substantially inhibited MYPT1 phosphorylation by the Pims and DAPK3, but not ROCK. Although phosphorylation of LC20 by Pim-3 occurred slowly compared with MLCK, we reasoned that this activity was physiologically relevant based on tissue studies. In VSM tissues, HS56, HS94, and HS148 delayed force-onset kinetics; however, insignificant differences among their effects suggest that the role of Pim kinases is not yet kinetically discernible from that of DAPK3. Remarkably, selective inhibition of DAPK3 alone (HS94, HS148) was insufficient to reduce contractile force, while concurrent Pim/DAPK3 inhibition by HS56 reduced maximum force and drove LC20 phosphorylation to near baseline levels (Figure 5). In these experiments, Cla inhibits MLCP and, in the absence of phosphatase activity, LC20 phosphorylation induced contraction. Pretreatment with potent inhibitors of MYPT1 phosphorylation (e.g., HS94, HS148) was sufficient to lengthen t_{1/2 max}, but force eventually reached a stable maximum due to persistent activities of MLCK and other LC20 kinases. The ability of HS56 to reduce maximum force, by deduction, necessarily involves immediate or upstream inhibition of LC20 phosphorylation.

Most notably, Pim/DAPK3 inhibition by HS56 restored near normal SBP in RenTg mice (ΔSBP = −28 ± 3 mmHg; ∼23% reduction) without diminishing HR, while DAPK3 inhibition by HS148 affected neither SBP nor HR (Figure 7). HS56 displayed dose-dependent, rapid lowering of SBP that persisted for several minutes; characteristics that are consistent with ideal HE therapeutics (Tulman et al., 2012) and compare favorably with, or surpass those of ROCK and P2X3 antagonists reported in notable hypertension studies. ROCK inhibition by Y-27632 (IC₅₀ = 800 nM) produced a 50- to 80-mmHg (20%–33%) drop in BP in hypertensive mice (Uehata et al., 1997); however, slow onset (∼1 hr), long duration of action (>7 hr), and off-target effects (e.g., PRK2; IC50 = 600 nm) (Bain et al., 2007) diminish its therapeutic value for both chronic hypertension and HE. The P2X3 receptor antagonist, AF-219, produced a 28-mmHg (∼19%) drop in SBP in hypertensive rats, but this was accompanied by a significant drop in HR (∼10%) (Pijacka et al., 2016), in dramatic contrast to HS56, which had little impact on the heart.

These studies provide mechanistic biological insight into the polypharmacology of Pims and DAPK3 in the context of VSM contraction; establishing Pim kinases as potentially critical elements of a multi-target engagement strategy for controlling hypertension. HS56 is key molecular probe for elucidating the effect of Pim/DAPK3 inhibition on VSM contractility and represents a significant first step in the development of next-generation antihypertensive medications. If HS56 or future generations of Pim/DAPK3 inhibitors produce clinically predictable dose-dependent reduction in systemic BP without altering HR, this certainly would constitute a significant advance in the molecular management of acute hypertension, thereby reducing morbidity and mortality rates associated with the condition. We anticipate that chronic administration of HS56 may also mitigate and reverse hypertrophic remodeling and chronic hypertension-related pathologies.

Significance

Investigations into underlying mechanisms governing VSM contractility remain vitally important for developing molecularly targeted antihypertensive medications. Here we demonstrated that Pim kinases directly target VSM substrates in vitro. Additionally, we developed a set of inhibitors that were utilized as molecular probes to demonstrate that dual Pim/DAPK3 inhibition by HS56 (1) critically modulates LC20 phosphorylation and contractility in tissues and (2) regulates BP in vivo without affecting heart rate. These studies provide mechanistic insight into the polypharmacology of VSM contraction, establishing Pim kinases as potentially critical elements of a multi-target engagement strategy for controlling hypertension.

STAR★Methods

Contact for Reagent and Resource Sharing

Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Timothy Haystead (https://www.cell.com/action/showMethods?pii=S2451–9456%2817%2930269–6timothy.haystead@duke.edu).

Experimental Model and Subject Details

Bacteria Culture

Escherichia coli BL21 (DE3)-R3 cells (Strain-Damerell et al., 2014) and Escherichia coli BL21(DE3)pLysS (Sutherland et al., 2016) were cultured in either terrific broth (TB) or Luria-Bertami (LB) medium (supplemented with 50 μg/mL of appropriate antibiotic) at 37°C.

Tissue Preparation

Tissues were harvested from male Sprague-Dawley rats (∼300 g, age 10–12 weeks) (Charles River Laboratories, Crl:SD Strain code:400) that had been housed, anesthetized, and euthanized as previously described (MacDonald et al., 2016, Moffat et al., 2011, Sutherland et al., 2016) according to protocols approved by the University of Calgary Animal Care and Use Committee consistent with the standards of the Canadian Council on Animal Care.

Animal Models and Welfare

Hypertensive adult male mice (age 4–6 months) bearing a transgene expressing renin (RenTg) and normotensive (WT) littermates (Duke Animal Models Core, also available fromThe Jackson Laboratory, Cat# 007853) were bred for use in experiments as previously described (Caron et al., 2002). All animal experiments were conducted in compliance with institutional policies and appropriate regulations and were approved by the institutional animal care and use committees for Duke University School of Medicine (Duke IACUC A129–16-06) and Durham VAMC (1506–002).

Method Details

Human Protein Atlas and Expression Atlas

A bioinformatic analysis of the Human Protein Atlas database (HPA) (Uhlen et al., 2015) using search terms (Pim1, Pim2, and Pim3) provided graphical representations of HPA data in the form of a Tissue Atlas (Figure S1A). Each bar on the resulting Tissue Atlas represents the highest expression score found in a particular group of tissues. Expression scores were automatically calculated based on archived datasets within the HPA to represent (i) Pim kinase RNA expression from RNA-seq results, reported as number of transcripts per million (TPM) and (ii) Pim kinase protein expression score based on a best estimate of the “true” protein expression from knowledge-based annotation of the HPA database. Detailed institutional methods for HPA data collection, analysis, and scoring can be found online at https://www.proteinatlas.org/about/assays+annotation#ihk. Representative results and Tissue Atlas for each Pim kinase can be found online at:

http://www.proteinatlas.org/ENSG00000137193-PIM1/tissue

http://www.proteinatlas.org/ENSG00000102096-PIM2/tissue

http://www.proteinatlas.org/ENSG00000198355-PIM3/tissue

A second bioinformatic analysis of baseline RNA expression for Pim kinases in human tissues was conducted using the Expression Atlas (Petryszak et al., 2016) provided by the European Bioinformatics Institute, found online at https://www.ebi.ac.uk/gxa/home, using a Boolean combination of search terms (Pim# AND Homo sapiens; where # = 1, 2, or 3). Data from individual searches were visualized, downloaded, and depicted as heatmaps representing Pim gene expression values, which were converted into a color-scale image, providing a visual representation of gene expression levels across different biological locations and different independent experiments (Figure S1B). Data representing RNA expression for Pim-1, Pim-2, and Pim-3 within smooth muscle tissues was found in datasets entitled 68 RANTOM5 and 32 Uhlen’s Lab.

Expression and Purification of Kinases

DAPK3 (ZIPK)

Human DAPK3 sequence (NCBI accession: NP_001339) was obtained as cDNA clone (Genome Systems Inc., St Louis, MO, Cat# AI660136) (Borman et al., 2002) and then subcloned into bacterial expression vector pGEX-6P1 (GE Healthcare, Cat# 28954648). The catalytic domain of constitutively-active DAPK3 consisting of N-terminal residues 1–320 was expressed in E. coli (BL21(DE3)pLysS (Promega, Cat# L1195) as a recombinant GST-fusion (GST-DAPK3). Bacteria were grown in standard Luria-Broth and protein production was induced by addition of 0.1 mM IPTG for 4 hours at 37°C. Bacteria were collect by centrifugation and resuspended in Buffer A, frozen at - 80°C and then lysed at room temperature with lysozyme (5 mg/25 ml culture). After centrifugation at 10,000 × g for 30 min, the supernatant was removed and GST-DAPK3 (1–320) was captured with glutathione-Sepharose affinity chromatography. After extensive washing (Buffer A + 0.5M NaCl), the bead slurry was incubated with PreScission Protease for GST-tag removal (GE Healthcare, Cat# 27084301) for 16 hours at 4°C. Eluted fractions were recovered and analyzed by SDS-PAGE to identify pure DAPK3(1–320) material.

Pim Kinases

cDNAs encoding human full length Pim kinases (Pim-1,−2,−3) were obtained from synthetic sources and used as templates to amplify kinase domain-containing sequences and further sub-cloned into different expression vectors, using ligation independent cloning (Strain-Damerell et al., 2014). Pim-1 (aa 1–312) was cloned into PLIC-SGC1 [pET expression vector with His₆ tag in a 23 aa-N-terminal fusion peptide, with TEV protease cleavage site, (Amp⁺)] and co-expressed with λ-phosphatase in Escherichia coli BL21 (DE3)-R3 cells. Pim-2 (aa 1–311) and Pim-3 (aa 1–326) were cloned into pNIC28-Bsa4 [pET expression vector with His₆ tag in a 22 aa-N-terminal fusion peptide, with TEV protease cleavage site, (Kan⁺)] and co-expressed with λ-phosphatase in Escherichia coli BL21 (DE3)-R3 cells. Transformed cells were initially cultured (from an overnight pre-culture) in either terrific broth (TB) medium (supplemented with 50 μg/mL of appropriate antibiotic) to OD₆₀₀ of ∼1.6 at 37°C, followed by additional growth while cooling to 18°C to an OD₆₀₀ of ∼3 before induction with 0.5 mM IPTG overnight, or cultured in Luria-Bertami (LB) medium to OD₆₀₀ of ∼0.4 at 37°C, followed by additional growth while cooling to 18°C to an OD₆₀₀ of ∼0.7 before induction with 0.5 mM IPTG overnight. Cells were harvested by centrifugation (JLA 8,100 rotor Beckman Coulter, Avanti J-20 XP centrifuge) and were frozen at −20°C. Cells expressing His₆-tagged proteins were re-suspended in lysis buffer (50 mM HEPES pH 7.5, 500 mM NaCl, 10 mM Imidazole, 5% glycerol and 0.5 mM TCEP (Tris(2-carboxyethyl)phosphine hydrochloride) in the presence of protease inhibitors cocktail (1 μL/mL) and lysed by sonication using a 750 W Sonics Vibra-Cell sonicator, with amplitude set to 35 %, with bursts of 5 sec on-10 sec off, for 5 minutes, on ice. PEI (polyethyleneimine) was added to a final concentration of 0.15 % and lysates were transferred to centrifuge tubes and centrifuged at 53,000×g using a JA-25.50 rotor, for at least 45 minutes, at 4°C. After centrifugation, the clarified supernatant was passed through a gravity column of 5 mL Ni-Sepharose resin, IMAC (GE Healthcare, Cat# 17–5268-01), previously equilibrated in lysis buffer. The resin was first washed with 50 mL of lysis buffer containing 1 M NaCl and 30 mM imidazole, then with 25 mL of Lysis Buffer containing 100 mM imidazole and finally the protein was eluted with 25 mL of Lysis Buffer containing 300 mM imidazole. The eluted proteins were collected and treated overnight with TEV (Tobacco Etch Virus) protease (Sigma-Aldrich, Cat# T4455) at 40°C to remove the N-terminal tag. Digested proteins were loaded onto a nickel column again to remove the cleaved hexa-histidine expression tag protease used. The flow-through containing the cleaved proteins was collected and concentrated to 5 mL (Amicon) and injected onto a Superdex 75 or 200 (16/60) gel filtration column on an AKTA system (GE Healthcare) pre-equilibrated into GF Buffer (50 mM HEPES pH7.5, 300 mM NaCl, 5% glycerol, and 0.5 mM TCEP). The resulting pure protein was stored at −80°C in 50 mM HEPES, pH 7.5, 300 mM NaCl, 0.5 mM TCEP and 5% glycerol. The correct mass and purity for all protein constructs was confirmed by an Agilent 1100 Series LC/MSD TOF (Agilent Technologies Inc., Palo Alto, CA).

MLCK

MLCK was purified from chicken gizzard as previously described (Ngai et al., 1984).

ROCKα

Constitutively-active N-terminal His6-tagged and N-terminal HA-tagged, recombinant rat ROCKα (ROCK2; residues 2–543; 67.2 kDa), expressed by baculovirus in Sf21 insect cells, was purchased from EMD Millipore (catalog number 14–338).

PKA

PKA was purified from bovine heart as previously described (Grassie et al., 2012).

Preparation of Kinase Assay Substrates

Peptide Substrate for DAPK3

MYPT1 substrate peptide (KKKRQSRRSTQGVTL), corresponding to Arg690 to Lys701 of MYPT1, was synthesized by University of Calgary Peptide Services (Calgary, Alberta), confirmed by amino acid analysis, and shown to be 95% pure by analytical high-performance liquid chromatography. (MacDonald et al., 2001a)

Peptide Substrate for Pim Kinases

RSRHSSYPAGT was custom synthesized by Biomatik (Wilmington, Delaware, USA).

LC20

Smooth muscle myosin regulatory light chains (LC20, also termed MLC2; NCBI accession: NP_990609) were purified from chicken gizzard as described (Hathaway and Haeberle, 1983). Briefly, tissue was homogenized in a blender in buffer containing MOPS, pH 6.8, 20 mM KCl, 1 mM dithiothreitol, 1 mM EGTA, 1 mM MgCl2, and 0.2% Triton X- 100 (Buffer A plus Triton). Subsequent separation of whole myosin was achieved by a series of centrifugations (12,000 × g) and resuspensions in Buffer A plus Triton. Pellets from the final centrifugation were solubilized by sonication in buffer containing 20 mM MOPS and 5 mM DTT, pH 7.0. This solution was adjusted to contain final concentrations of urea (8 M), guanidine HCl (1 M), and SDS (0.05%) at pH 8.0. Essential and regulatory myosin light chains (LC17 and LC20, respectively) were enriched in solution by the addition of absolute ethanol at room temperature, followed by centrifugation of unwanted precipitate at 12,000 × g. The supernatant was dialyzed against buffer containing 5 mM ammonium bicarbonate and 0.4 mM DTT. Following dialysis, the solution was adjusted to contain 20 mM MOPS, 0.6 M NaCl, 1 mM EDTA, 1 mM DTT, pH 7.5 (equilibration buffer) and loaded onto a column of phenyl-Sepharose. After washing with equilibration buffer, LC20 was isolated from LC17 and other protein contaminants by elution with deionized water containing 1 mM DTT. Eluted fractions were recovered and analyzed by SDS-PAGE.

MYPT1

Full-length chicken MYPT1 sequence (PPP1R12A, NCBI accession no. NP_990454) was generously provided by Dr. David Hartshorne (University of Arizona). MYPT1 was sub-cloned into bacterial expression vector pGEX-6P1 (GE Healthcare, Cat# 28954648) and transformed into E. coli strain BL21(DE3)pLysS (Promega, Cat# L1195). The recombinant GST-tagged MYPT1 was purified as previously described (Sutherland et al., 2016). Briefly, GST-MYPT1 protein production was induced in E. coli BL21(DE3)pLysS by addition of IPTG for 4 hours at 37°C. Bacteria were collect by centrifugation and resuspended in 30 mM Tris-HCl, pH 7.5, 1 mM EGTA, 1 mM EDTA, 150 mM NaCl, 0.2 mM DTT with protease inhibitors (Buffer A), frozen at - 80°C and then lysed at room temperature with lysozyme (5 mg/25 ml culture). After centrifugation at 10,000 × g for 30 min, the supernatant was removed, diluted 1:3 with Buffer A, and loaded onto an SP-Sepharose column. After development with a 0.15 to 0.5 M NaCl gradient, fractions containing GST-MYPT1 were pooled and captured with glutathione-Sepharose affinity chromatography. After extensive washing (Buffer A + 0.5 M NaCl), the bead slurry was incubated with PreScission Protease for GST-tag removal (GE Healthcare, Cat# 27084301) at 4°C for 3 days. Eluted fractions were recovered and analyzed by SDS-PAGE to identify full-length MYPT1 material. The MYPT1 protein was confirmed to be free of kinase activity by an in vitro phosphorylation assay performed in the absence of exogenous kinase.

Kinase Assays

General Protocol

Analogs were formulated as 10 mM solutions in DMSO and assayed against protein kinases using radiolabeled ATP in a protocol adapted from the International Center for Kinase Profiling (Dundee, UK) (Hastie et al., 2006). Protein kinase, dissolved in 25 mM HEPES pH 7.4, 10 mM MgCl₂, 1 mM dithiothreitol (HEPES-Mg buffer), was added to a solution of peptide substrate, ATP, analog, and DMSO in HEPES-Mg buffer. The final reaction volume of 40 μL contained 50 – 350 ng kinase, 140 – 300 μM peptide substrate, 5 – 50 μM ATP (Bio Basic, Markham, ON, Canada, Cat# AB0020) with 5 – 20 μCi of [γ−³²P]-ATP (Perkin Elmer, Waltham, MA, USA, Cat# NEG035C005MC), 10 μM analog (initial screen) or 0.001 μM – 100 μM analog (titration), and 2.5 % DMSO. For exact compositions, see below. After 10 min, the reaction was terminated by addition of concentrated H₃PO₄ (10 μL). A 5 μL aliquot of each mixture was spotted onto P81 ion exchange cellulose chromatography paper (Whatman Cat# 3698325), washed 3× with 200 mM H₃PO₄ and then 1× with acetone. After thorough drying, the papers were placed in uncapped scintillation vials and Cherenkov counted for 0.5 min using tritium isotope detection (Beckman LS6500 Scintillation Counter).

DAPK3 Kinase Assay

The final reaction contained 52 ng GST-DAPK3, 140 μM MYPT1 peptide substrate, 50 μM ATP with 5 – 20 μCi of [γ−³²P]-ATP, 10 μM or 0.001 μM – 100 μM analog, and 2.5 % DMSO in HEPES-Mg buffer.

Pim-1 Kinase Assay

The final reaction contained 100 ng Pim-1, 300 μM Pim peptide substrate, 50 μM ATP with 5 – 20 μCi of [γ−³²P]-ATP, 10 μM or 0.001 μM – 100 μM analog, and 2.5 % DMSO in HEPES-Mg buffer.

Pim-2 Kinase Assay

The final reaction contained 350 ng Pim-2, 300 μM Pim peptide substrate, 5 μM ATP with 5 – 20 μCi of [γ−³²P]-ATP, 10 μM or 0.001 μM – 100 μM analog, and 2.5 % DMSO in HEPES-Mg buffer.

Pim-3 Kinase Assay

The final reaction contained 150 ng Pim-3, 300 μM Pim peptide substrate, 50 μM ATP with 5 – 20 μCi of [γ−³²P]-ATP, 10 μM or 0.001 μM – 100 μM analog, and 2.5 % DMSO in HEPES-Mg buffer.

Determination of DAPK3 K_M for ATP

Using a variation of the kinase assay general protocol, GST-DAPK3 with saturating amounts of peptide substrate was assayed against several concentrations of ATP. A stock solution of 4 mM ATP, containing 550 μCi/mL of [γ−³²P]-ATP, was serially diluted with HEPES-Mg buffer and aliquoted into the reactions. Experimental reactions: final volumes of 40 μL contained 52 ng GST-DAPK3, 140 μM MYPT1 peptide substrate, and ATP ranging from 1 μM – 500 μM with 0.006 – 3 μCi of [γ−³²P]-ATP. Negative control: each experiment reaction was repeated without GST-DAPK3, and the counts per minute were subtracted to remove background signal. Data was analyzed by non-linear curve fitting (Graphpad Prism 7) and Lineweaver-Burk double reciprocal plot (Figure S4F).

IC₅₀ and K_i Values

IC₅₀ values were determined from titration curves (Figure S4) using log(inhibitor) vs. response - Variable slope (four parameters) (Graphpad Prism 7). The IC₅₀ for HS56 titrated against Pim-2 was estimated from curve fitting X-min = −3 (R² = 0.901). IC₅₀ values were converted to inhibition constants (K_i) using the Cheng Prusoff equation (Cheng and Prusoff, 1973) (Figure S4G). Previously reported Michaelis constants (K_M) of Pim kinases for ATP were used; 160 μM (Pim-1), 4 μM (Pim-2), and 40 μM (Pim-3) (Burger et al., 2013). For GST-DAPK3, the K_M for ATP was determined experimentally to be 20 μM (Figure S4E).

Tissue Preparation and Force Measurements

Caudal arteries were removed from male Sprague-Dawley rats (∼300 g) that had been anesthetized and euthanized according to protocols approved by the University of Calgary Animal Care and Use Committee consistent with the standards of the Canadian Council on Animal Care. The arteries were cleaned of excess adventitia, denuded of endothelium, and cut into helical strips (1.5 mm × 6 mm). Muscle strips were mounted on a Grass isometric force transducer (FT03C) and force was recorded as previously described (MacDonald et al., 2016, Moffat et al., 2011, Sutherland et al., 2016). Native tissues were maintained in HEPES-buffered Tyrode’s (H-T) solution containing 137 mM NaCl, 2.7 mM KCl, 1 mM MgCl₂, 1.8 mM CaCl₂, 5.6 mM glucose, and 10 mM HEPES, pH 7.4. Prior to induction, baseline contractile force was identified with application of H-T buffer containing 87 mM KCl and then washed and pre-incubated with vehicle (DMSO), or inhibitor (HS38, HS56, HS94, HS148, HS182) at concentrations of 0–100 μM. Contraction was induced with Calyculin A (Cla, 0.5 μM) (EMD Millipore, Cat# 208851) in H-T solution. Skinned (demembranated) tissues were generated by incubation with 1% (v/v) Triton X-100 as described (Sutherland et al., 2016). Skinned tissues were washed (3 × 5 min each) in pCa 9 solution (4 mM K₂EGTA, 5.83 mM MgCl₂, 0.5 mM DTT, 20 mM TES, pH 6.9, and an ATP regenerating system composed of 3.9 mM Na₂ATP, 7.56 mM potassium propionate, 16.2 mM phosphocreatine, and 30 U/mL creatine kinase) and then Ca²⁺-dependent contraction was induced with pCa 4.5 solution (4 mM CaEGTA, 5.66 mM MgCl2, 0.5 mM DTT, 20 mM TES, pH 6.9, and the ATP-regenerating system). Muscle strips were flash frozen in 10% (w/v) trichloroacetic acid, 10 mM dithiothreitol in acetone, followed by 3 × 10 second washes in 10 mM DTT in acetone. Tissues were then lyophilized overnight prior to protein extraction.

LC20 Phosphorylation in Tissues

LC20 phosphorylation in excised smooth muscle strips was quantified as previously described (MacDonald et al., 2016). Briefly, rat caudal artery VSM proteins were extracted from lyophilized tissues in 1 mL of SDS-gel sample buffer with constant shaking for 16 hours at 4°C prior protein resolution by Phos-Tag SDS-PAGE using 12.5% acrylamide and 50 μM Phos-Tag reagent (FujiFilm Wako; Cat# AAL-107). Following transfer to PVDF membrane (Roche Life Science, Quebec, Canada), proteins were fixed by incubation for 20 minutes with 0.5% glutaraldehyde in phosphate-buffered saline and non-specific binding sites were blocked with 5% (w/v) nonfat dry milk. Membranes were incubated overnight with anti-LC20 (Santa Cruz Biotechnology, Cat# sc-48414, Santa Cruz, CA) and then incubated for 1 hour with horseradish peroxidase–conjugated secondary antibody (Chemicon, Temecula, CA). All western blots were visualized with West Femto enhanced chemiluminescence reagent (Thermo Fisher, Cat# 34095) using a LAS4000 Imaging Station (GE Healthcare), ensuring that the representative signal occurred in the linear range. Quantification was performed by densitometry with ImageQuant TL software (GE Healthcare). LC20 phosphorylation stoichiometry was calculated as follows: mol Pi/mol LC20 = [LC20–1P + (2 × LC20–2P)] / [LC20–0P + LC20–1P + LC20–2P].

In Vitro Phosphorylation Studies

Reactions were performed at 30°C in a final volume of 140 μL, initiated by addition of ATP (final [ATP] = 0.5 mM), and terminated at the indicated time by addition of SDS-PAGE sample buffer and boiling. After termination, samples were further diluted 1:10 and run on Phos-Tag SDS-PAGE gels as described above. Experiments were repeated in the presence of inhibitor (50 μM) as indicated.

Phosphorylation of LC20 by MLCK

MLCK reactions contained 10 μg/mL LC20, 2 μg/mL MLCK, 25 mM HEPES pH 7.4, 10 mM MgCl₂, purified bovine calmodulin (10 μg/mL), and 0.1 mM CaCl₂.

Phosphorylation of LC20 by Pim Kinases

Reactions contained 10 μg/mL LC20, 2 μg/mL Pim kinase, 25 mM HEPES pH 7.4, and 10 mM MgCl₂.

Phosphorylation of MYPT1 by ROCK

25 mM TRIS-HCl pH 7.5, 5 mM MgCl₂, 5 mM EGTA, 0.895 ng/mL ROCK, 20 μg/mL MYPT1, 0.5 mM ATP. Reactions were repeated with HS182 and 50 μM HS56.

Phosphorylation of MYPT1 by Pim-1

25 mM HEPES pH 7.4, 10 mM MgCl2, 2 μg/mL PIM1, 20 μg/mL MYPT1, 0.5 mM ATP. Reactions were repeated with HS182 and HS56.

Phosphorylation of MYPT1 by Pim-2

25 mM HEPES pH 7.4, 10 mM MgCl₂, 2.8 μg/mL PIM1, 20 μg/mL MYPT1, 0.5 mM ATP. Reactions were repeated with HS182 and HS56.

Phosphorylation of MYPT1 by Pim-3

25 mM HEPES pH 7.4, 10 mM MgCl2, 3 μg/mL PIM1, 20 μg/mL MYPT1, 0.5 mM ATP. Reactions were repeated with HS182 and HS56.

Phosphorylation of MYPT1 by DAPK3

25 mM HEPES pH 7.4, 10 mM MgCl₂, 2.5 μg/mL GST-DAPK3, 20 μg/mL MYPT1, 0.5 mM ATP. Reactions were repeated with HS182, HS56, HS38, and HS94.

Phosphorylation of MYPT1 by PKA

25 mM TRIS-HCl pH 7.5, 5 mM MgCl₂, 5 mM EGTA, 1 μg/mL PKA, 20 μg/mL MYPT1, 0.5 mM ATP.

Western Blotting

Phosphorylation of LC20 (from purified chicken gizzard) by MLCK or PIM kinases was resolved using 12.5% acrylamide and 50 μM Phos-Tag reagent (FujiFilm Wako; Cat# AAL-107) as described above. MYPT1 phosphorylation was quantified on optimized Phos-Tag SDS-PAGE gels as previously described (Sutherland et al., 2016). Proteins were transferred to nitrocellulose membranes (0.2 μm) overnight at 28 V and 4°C in 25 mM TRIS-HCl, pH 7.5, 192 mM glycine, and 10% (v/v) methanol. After blocking with 5% (w/v) nonfat dry milk, membranes were incubated with primary antibody (MYL2 Antibody, Cat# sc-517414, Santa Cruz Biotechnology, Santa Cruz, CA; anti-panMYPT1, custom made peptide-directed antibody, New England Peptide, Gardner, MA; 1:10,000 dilution) in TBST for 2 h. Chemiluminescence signal detection using the SuperSignal West Femto reagent (Thermo Fisher, Cat# 34095) was used following incubation with HRP-coupled secondary antibody (1:5,000 dilution). The emitted light was detected and quantified with a chemiluminescence imaging analyzer (LAS3000mini, Fujifilm), and images were analyzed with Multi-Gauge version 3.0 software.

X-Ray Crystallography

Expression, Purification, and Crystallization

Expression and purification of the DAPK3 catalytic domain for crystallographic studies was performed using proprietary methods by Crelux GmbH (Martinsried, Germany). Briefly, human DAPK3^Val9-Gly288 was expressed as a fusion protein with TEV (tobacco etch virus protease) cleavable N-terminal His₆ affinity tag in Escherichia coli strain BL21 (DE3). The fusion protein was purified by Ni affinity chromatography and His₆-TEV domain was removed by treatment with TEV protease. DAPK3 was isolated for crystallization studies by size exclusion chromatography. Crystals of DAPK3 in complex with HS38 were obtained using hanging drop vapor diffusion setups: 0.7 μL of DAPK3 (9.0 mg/mL in 50 mM HEPES, 500 mM NaCl, 5 mM DTT, 5 % glycerol, pH 7.5), pre-incubated with 5.4-fold molar excess of HS38 (150 mM in DMSO) for 2 h, mixed with 0.7 μL of reservoir solution (1.9 M sodium malonate, pH 6.8) and equilibrated at 4°C over 400 μL of reservoir solution. Crystals appeared within 1 day and grew over 3–4 days to full size.

Data Collection

A complete 2.5 Å resolution data set of a DAPK3-HS38 crystal was collected at the ESRF synchrotron radiation source, id29, (Grenoble, France) using a Pilatus 6M (Dectris) detector at 0.97625 Å wavelength. Data collection and refinement statistics are listed in Table S2. Data reduction was carried out with MOSFLM, and scaling with SCALA. Phasing was carried out using AMoRE.

Structure Determination and Refinement

Initial structure determination was carried out by molecular replacement using a published DAPK3 structure as search model (PDB: 1P4F) (Huber et al., 2012). Alternating manual re-building and refinement with REFMAC5 resulted in the final model. Chains A, B, C, and D are coincident with each other. Residues 49 to 53 in chain A, residues 47 to 53 in chain B, residues 52 to 54, 105 to 106 and 150 to 151 in chain C as well as residues 51 to 55, 104 to 106 and 150 to 151 in chain D could not be modelled due to flexibility. The final model allowed for unambiguous placement of HS38.

Crystallographic Data Analysis

Molecular graphics were produced using the UCSF Chimera (Pettersen et al., 2004) package from the Computer Graphics Laboratory, University of California, San Francisco (supported by NIH P41 RR-01081). H-bonding was analyzed using FindHBond. Structural overlays were created using MatchMaker. Analysis and 2D depiction of non-covalent protein-ligand interactions was performed using PoseView (Stierand and Rarey, 2010) (ZBH Center for Bioinformatics, University of Hamburg, Germany).

Acute Infusion Studies in Mice

Acute systemic vasodilatory effects of HS148 and HS56 were examined in hypertensive RenTg mice and normotensive WT littermates under 2% isoflurane anesthesia (Duke Animal Models Core) (Caron et al., 2002). A catheter (PE-50) was inserted into the left jugular vein for the administration of basal fluids and test compounds. A second catheter, pulled down PE-50 attached to a pressure transducer (MLT844, ADInstruments, Colorado Springs, CO), was placed into the left carotid artery. Intra-arterial BP was continuously recorded through the carotid catheter using the PowerLab data acquisition system and LabChart software (ADInstruments, Colorado Springs, CO). HS148, HS56 or vehicle (DMSO) were injected intravenously into the internal jugular vein at a volume of 1 μl/g body weight (25–30 μl total volume depending on weight) followed immediately by an equivalent bolus of vehicle (total infusion 2 μl/g body weight, 50–60 μl). Vehicle-only mice received two identical vehicle infusions in rapid succession so that mice were exposed to the same volumes. HS148, HS56 and vehicle were examined in separate mice to minimize residual effects of the drug. Intra-carotid BP data were analyzed by averaging 1 second of data every 5 seconds.

Chemical Synthesis

Abbreviations

DMF N,N-Dimethylformamide

THF Tetrahydrofuran

DIEA N,N-Diisopropylethylamine

DTT Dithiothreitol

EtOH Ethanol

MeOH Methanol

For synthesis of compounds 1–71, All reagents were purchased from commercial sources and used without further purification. Unless otherwise stated, the commercial source all commonly available reagents was Sigma-Aldrich Corp. (St. Louis, MO, USA), VWR (Radnor, PA, USA), or Matrix Scientific (Elgin, SC, USA).

General Procedure A (Scheme 1), amino-1H-pyrazoles (6a – 6n)

Scheme 1.

Preparation of 6a – 6n.

Ethyl-2-cyano-3-ethoxyacrylate (1.50 g, 8.87 mmol, 1.0 equiv.) (Sigma-Aldrich, Cat# E28005) was combined with N,N-diisopropylethylamine (1.7 mL, 9.8 mmol, 1.1 equiv.) and the appropriate hydrazine (9.8 mmol, 1.1 equiv.) in ethanol (10 mL). The mixture was heated to 80°C for 12 hours or until starting hydrazine was consumed. Silica gel chromatography or crystallization yielded 6a – 6n.

Ethyl 5-amino-1-phenyl-1H-pyrazole-4-carboxylate (6a)

General procedure A was followed using phenylhydrazine (Sigma-Aldrich, Cat# P26252). The mixture was condensed to dryness and dissolved in dichloromethane (5 mL). Silica gel chromatography with gradient elution (20–40% ethyl acetate in hexane) yielded 6a (96% yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₂H₁₃N₃O₂ 232.0; found 232.0.

Ethyl 5-amino-1-(2-chlorophenyl)-1H-pyrazole-4-carboxylate (6b)

General procedure A was followed using 2-chlorophenylhydrazine hydrochloride (Sigma-Aldrich, Cat# 109509). The mixture was condensed to dryness and dissolved in dichloromethane (5 mL). Silica gel chromatography with gradient elution (20–40% ethyl acetate in hexane) yielded 6b (99% yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₂H₁₂ClN₃O₂ 266.1; found 266.0.

Ethyl 5-amino-1-(3-chlorophenyl)-1H-pyrazole-4-carboxylate (6c)

General procedure A was followed using 3-chlorophenylhydrazine hydrochloride (Matric Scientific, Cat# 075249). The mixture was condensed to dryness and dissolved in dichloromethane (5 mL). Silica gel chromatography with gradient elution (0–50% ethyl acetate in hexane) yielded 6c (86% yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₂H₁₂ClN₃O₂ 266.1; found 266.0.

Ethyl 5-amino-1-(4-chlorophenyl)-1H-pyrazole-4-carboxylate (6d)

General procedure A was followed using 4-chlorophenylhydrazine hydrochloride (Sigma-Aldrich, Cat# C65807). The mixture was condensed to dryness and dissolved in dichloromethane (5 mL). Silica gel chromatography with gradient elution (0–20% ethyl acetate in dichloromethane) yielded 6d (95% yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₂H₁₂ClN₃O₂ 266.1; found 266.0.

Ethyl 5-amino-1-isopropyl-1H-pyrazole-4-carboxylate (6e)

General procedure A was followed using isopropylhydrazine hydrochloride (Matrix Scientific, Elgin, SC, USA, Cat# 007266). The mixture was condensed to dryness and dissolved in dichloromethane (5 mL). Silica gel chromatography with gradient elution (20–40% ethyl acetate in hexane) yielded 6e (95% yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₉H₁₅N₃O₂ 198.1; found 198.0.

Ethyl 5-amino-1-(o-tolyl)-1H-pyrazole-4-carboxylate (6f)

General procedure A was followed using o-tolylhydrazine hydrochloride (Alfa Aesar, Cat# A15767). The mixture was condensed to dryness and dissolved in dichloromethane (5 mL). Silica gel chromatography with gradient elution (20–30% ethyl acetate in hexane) yielded 6f (99% yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₃H₁₅N₃O₂ 246.1; found 246.1.

Ethyl 5-amino-1-(m-tolyl)-1H-pyrazole-4-carboxylate (6g)

General procedure A was followed using m-tolylhydrazine hydrochloride (Acros, Cat# 139150010). The mixture was condensed to dryness and dissolved in dichloromethane (5 mL). Silica gel chromatography with gradient elution (20–30% ethyl acetate in hexane) yielded 6g (86% yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₃H₁₅N₃O₂ 246.1; found 246.1.

Ethyl 5-amino-1-(3-methoxyphenyl)-1H-pyrazole-4-carboxylate (6h)

General procedure A was followed using 3-methoxyphenylhydrazine hydrochloride (Matrix Scientific, Elgin, SC, USA, Cat# 021395). The mixture was condensed to dryness and dissolved in dichloromethane (5 mL). Silica gel chromatography with gradient elution (20–30% ethyl acetate in hexane) yielded 6h (53% yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₃H₁₅N₃O₃ 262.1; found 262.1.

Ethyl 5-amino-1-(4-methoxyphenyl)-1H-pyrazole-4-carboxylate (6i)

General procedure A was followed using 4-methoxyphenylhydrazine hydrochloride (Matrix Scientific, Elgin, SC, USA, Cat# 021131). The mixture was condensed to dryness and dissolved in dichloromethane (5 mL). Silica gel chromatography with gradient elution (20–40% ethyl acetate in hexane) yielded 6i (86% yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₃H₁₅N₃O₃ 262.1; found 262.1.

Ethyl 5-amino-1-(3-cyanophenyl)-1H-pyrazole-4-carboxylate (6j)

General procedure A was followed using 3-cyanophenylhydrazine hydrochloride (Ark Pharm, Inc., AK-39869). The mixture was condensed to dryness and dissolved in dichloromethane (5 mL). Silica gel chromatography with gradient elution (0–40% ethyl acetate in hexane) yielded 6j (71% yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₃H₁₂N₄O₂ 257.1; found 257.1.

Ethyl 5-amino-1-(4-cyanophenyl)-1H-pyrazole-4-carboxylate (6k)

General procedure A was followed using 4-cyanophenylhydrazine hydrochloride (Sigma-Aldrich, Cat# 453471). Addition of water (5 mL) to the reaction mixture resulted in precipitation of an off white solid, which was isolated by filtration and washed with ethanol and water to yield 6k (94% yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₃H₁₂N₄O₂ 257.1; found 257.1.

Ethyl 5-amino-1-(3-fluorophenyl)-1H-pyrazole-4-carboxylate (6l)

General procedure A was followed using 3-fluorophenylhydrazine hydrochloride (Matrix Scientific, Elgin, SC, USA, Cat# 003925). The mixture was condensed to dryness and dissolved in dichloromethane (5 mL). Silica gel chromatography with gradient elution (10–20% ethyl acetate in hexane) yielded 6l (88% yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₂H₁₂FN₃O₂ 250.1; found 250.1.

Ethyl 5-amino-1-(3-(trifluoromethyl)phenyl)-1H-pyrazole-4-carboxylate (6m)

General procedure A was followed using 3-(trifluoromethyl)phenylhydrazine hydrochloride (Matrix Scientific, Elgin, SC, USA, Cat# 005466). The mixture was condensed to dryness and dissolved in dichloromethane (5 mL). Silica gel chromatography with gradient elution (10–20% ethyl acetate in hexane) yielded 6m (84% yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₃H₁₂F₃N₃O₂ 300.1; found 300.1.

4-(5-amino-4-(ethoxycarbonyl)-1H-pyrazol-1-yl)benzoic acid (6n)

General procedure A was followed using 4-hydrazinobenzoic acid hydrochloride (Alfa Aesar, Ward Hill, MA, USA, Cat# B20326). The mixture was condensed to dryness and dissolved in dichloromethane with 10% methanol (5 mL). Silica gel chromatography with gradient elution (0–16% methanol in dichloromethane) yielded 6n (99% yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₃H₁₃N₃O₄ 276.1; found 276.1.

General Procedure B (Scheme 2), 6-mercapto-1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-4-ones (7a – 7n)

Scheme 2.

Preparation of 7a – 7n.

Amino-1H-pyrazoles (6a – 6n) (3 – 14 mmol, 1.0 equiv.) were combined with benzoyl isothiocyanate (1.2 equiv.) in tetrahydrofuran (50 mL). The heated to reflux under Ar(g) until starting material was consumed (6 – 18 hours). The resulting thiourea was added dropwise to a refluxing solution of ethanol (50 mL) and sodium ethoxide (21 wt. % in ethanol, 10 mL). The mixture was stirred for 30 min. at 90°C and condensed to dryness. Product was precipitated by the addition of water (50 mL). After stirring overnight, the off white solid was isolated by filtration and washed with water, then 50% ether in heptane to give 7a – 7n.

6-mercapto-1-phenyl-1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-4-one (7a)

General procedure B was followed using 6a (1.0 g, 4.3 mmol) to produce 7a (0.91 g, 86 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₁H₈N₄OS 245.1; found 245.0.

1-(2-chlorophenyl)-6-mercapto-1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-4-one (7b)

General procedure B was followed using 6b (1.2 g, 4.5 mmol) to produce 7b (0.92 g, 73 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₁H₇ClN₄OS 279.0; found 279.0.

1-(3-chlorophenyl)-6-mercapto-1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-4-one (7c)

General procedure B was followed using 6c (3.8 g, 14.3 mmol) to produce 7b (3.7 g, 92 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₁H₇ClN₄OS 279.0; found 279.0.

1-(4-chlorophenyl)-6-mercapto-1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-4-one (7d)

General procedure B was followed using 6d (0.86 g, 3.2 mmol) to produce 7b (0.65 g, 72 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₁H₇ClN₄OS 279.0; found 279.0.

1-isopropyl-6-mercapto-1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-4-one (7e)

General procedure B was followed using 6e (1.6 g, 8.4 mmol) to produce 7e (0.88 g, 50 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₈H₁₀N₄OS 211.1; found 211.0.

6-mercapto-1-(o-tolyl)-1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-4-one (7f)

General procedure B was followed using 6f (2.2 g, 9.0 mmol) to produce 7e (1.95 g, 84 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₂H₁₀N₄OS 259.1; found 259.1.

6-mercapto-1-(m-tolyl)-1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-4-one (7g)

General procedure B was followed using 6g (1.8 g, 7.2 mmol) to produce 7g (1.63 g, 86 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₂H₁₀N₄OS 259.1; found 259.1.

6-mercapto-1-(3-methoxyphenyl)-1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-4-one (7h)

General procedure B was followed using 6h (1.2 g, 4.7 mmol) to produce 7h (0.83 g, 65 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₂H₁₀N₄O₂S 275.1; found 275.1.

6-mercapto-1-(4-methoxyphenyl)-1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-4-one (7i)

General procedure B was followed using 6i (2.0 g, 7.7 mmol) to produce 7i (2.01 g, 96 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₂H₁₀N₄O₂S 275.1; found 275.1.

Synthesis of 3-(6-mercapto-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-1-yl)benzonitrile (7j)

General procedure B was followed using 6j (2.2 g, 8.4 mmol) to produce 7j (1.1 g, 50 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₂H₇N₅OS 270.0; found 270.0.

4-(6-mercapto-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-1-yl)benzonitrile (7k)

General procedure B was followed using 6k (2.1 g, 8.3 mmol) to produce 7k (2.1 g, 96 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₂H₇N₅OS 270.0; found 270.0.

1-(3-fluorophenyl)-6-mercapto-1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-4-one (7l)

General procedure B was followed using 6l (1.0 g, 4.2 mmol) to produce 7l (0.80 g, 73 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₁H₇FN₄OS 263.0; found 263.0.

6-mercapto-1-(3-(trifluoromethyl)phenyl)-1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-4-one (7m)

General procedure B was followed using 6m (0.60 g, 2.0 mmol) to produce 7m (0.53 g, 85 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₂H₇F₃N₄OS 313.0; found 313.1.

4-(6-mercapto-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-1-yl)benzoic acid (7n)

General procedure B was followed using 6n (0.14 g, 0.52 mmol) to produce 7n (0.064 g, 42 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₂H₈N₄O₃S 289.0; found 289.0.

2-((1-(3-chlorophenyl)-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)acetonitrile, HS56 (2)

To a solution of 7c (1.0 g, 3.6 mmol, 1.0 equiv.) in anhydrous tetrahydrofuran (10 mL) with N,N-diisopropylethylamine (1.25 mL, 7.2 mmol, 2.0 equiv.) under Ar(g) was added chloroacetonitrile (0.227 mL, 3.6 mmol, 1.0 equiv.) dropwise over 2 minutes. After stirring for 30 minutes the reaction was condensed to a tan oil and stirred under 50% ethyl acetate in heptane (50 mL) for 5 hours. The resulting off white precipitate was collected by filtration and washed 50% ethyl acetate in heptane followed by water to give HS56 (2) (1.05 g, 83 % yield). ¹H NMR (500 MHz, dmso-d₆) δ (ppm): 13.07 (s, 1H), 8.33 (s, 1H), 8.23 (s, 1H), 8.16 (d, J = 7.1 Hz, 1H), 7.57 (t, J = 7.7 Hz, 1H), 7.46 (d, J = 7.6 Hz, 1H), 4.36 (s, 2H); ¹³C NMR (500 MHz, dmso-d₆) δ (ppm): 158.9, 157.1, 151.2, 139.2, 136.7, 133.7, 130.9, 126.8, 120.6, 119.6, 117.3, 105.2, 16.4.

1-(3-chlorophenyl)-6-((2-hydroxyethyl)thio)-1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-4-one, HS43 (8)

Compound 8 was prepared as previously described (Carlson et al., 2013). To a solution of 7c (0.10 g, 0.36 mmol, 1.0 equiv.) in anhydrous dimethylformamide (1.0 mL) with N,N-diisopropylethylamine (0.188 mL, 1.1 mmol, 3.0 equiv.) was added 2-bromoethanol (0.028 mL, 0.39 mmol, 1.1 equiv.). The reaction was stirred for 12 hours and formic acid (0.30 mL) was added. The acidified mixture was purified by preparative HPLC (40 – 100 % methanol in water with 0.2 % formic acid modification over 18 minutes) to give HS43 (8) (0.067 g, 58 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₃H₁₁ClN₄O₂S 323.0; found 323.1.

1-(3-chlorophenyl)-6-(isobutylthio)-1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-4-one, HS44 (9)

To a solution of 7c (0.040 g, 0.14 mmol, 1.0 equiv.) in anhydrous dichloromethane (1.0 mL) with N,N-diisopropylethylamine (0.075 mL, 0.43 mmol, 3.0 equiv.) was added 1-bromo-2-methylpropane (0.062 mL, 0.57 mmol, 4.0 equiv.). The reaction was heated to 30°C for 4 hours, condensed to remove solvent, and dissolved in DMSO. Preparative HPLC (5 – 100 % methanol in water with 0.2 % formic acid modification over 20 minutes) provided HS44 (9) (0.010 g, 21 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₅H₁₅ClN₄OS 335.1; found 335.1.

Ethyl 4-((1-(3-chlorophenyl)-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)butanoate, HS49 (10)

To a solution of 7c (0.040 g, 0.14 mmol, 1.0 equiv.) in anhydrous dichloromethane (1.0 mL) with N,N-diisopropylethylamine (0.075 mL, 0.43 mmol, 3.0 equiv.) was added ethyl 4- bromobutyrate (0.083 mL, 0.57 mmol, 4.0 equiv.). The reaction was heated to 30°C for 4 hours, condensed to remove solvent, and dissolved in DMSO. Preparative HPLC (5 – 100 % methanol in water with 0.2 % formic acid modification over 20 minutes) provided HS49 (10) (0.016 g, 28 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₇H₁₇ClN₄O₃S 393.1; found 393.2.

2-((1-(3-chlorophenyl)-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)-2-methylpropanoic acid, HS50 (11)

To a solution of 7c (0.040 g, 0.14 mmol, 1.0 equiv.) in anhydrous dichloromethane (1.0 mL) with N,N-diisopropylethylamine (0.075 mL, 0.43 mmol, 3.0 equiv.) was added 2-bromo-2-methylpropionic acid (0.096 g, 0.57 mmol, 4.0 equiv.). The reaction was heated to 30°C for 4 hours, condensed to remove solvent, and dissolved in DMSO. Preparative HPLC (5 – 100 % methanol in water with 0.2 % formic acid modification over 20 minutes) provided HS50 (11) (0.032 g, 61 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₅H₁₃ClN₄O₃S 365.0; found 365.2.

Ethyl 4-((1-(3-chlorophenyl)-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)-3-oxobutanoate, HS51 (12)

To a solution of 7c (0.040 g, 0.14 mmol, 1.0 equiv.) in anhydrous dichloromethane (1.0 mL) with N,N-diisopropylethylamine (0.075 mL, 0.43 mmol, 3.0 equiv.) was added ethyl 4-chloroacetoacetate (0.079 mL, 0.57 mmol, 4.0 equiv.). The reaction was heated to 30°C for 4 hours, condensed to remove solvent, and dissolved in DMSO. Preparative HPLC (5 – 100 % methanol in water with 0.2 % formic acid modification over 20 minutes) provided HS51 (12) (0.031 g, 53 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₇H₁₅ClN₄O₄S 407.1; found 407.2.

1-(3-chlorophenyl)-6-((3-hydroxypropyl)thio)-1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-4-one, HS53 (13)

To a solution of 7c (0.050 g, 0.18 mmol, 1.0 equiv.) in anhydrous dichloromethane (1.0 mL) with N,N-diisopropylethylamine (0.094 mL, 0.54 mmol, 3.0 equiv.) was added 3-bromo-1-propanol (0.017 mL, 0.20 mmol, 1.1 equiv.). The reaction was heated to 30°C for 12 hours and then partitioned between water (2 mL) and ethyl acetate (2 mL). The resulting precipitate was isolated by filtration to give HS53 (13) (0.044 g, 73 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₄H₁₃ClN₄O₂S 337.1; found 337.1.

1-(3-chlorophenyl)-6-((2-oxo-2-phenylethyl)thio)-1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-4-one, HS54 (14)

To a solution of 7c (0.050 g, 0.18 mmol, 1.0 equiv.) in anhydrous dichloromethane (1.0 mL) with N,N-diisopropylethylamine (0.094 mL, 0.54 mmol, 3.0 equiv.) was added 2-bromoacetophenone (0.039 g, 0.20 mmol, 1.1 equiv.). The reaction was heated to 30°C for 12 hours and then partitioned between water (2 mL) and ethyl acetate (2 mL). The resulting precipitate was isolated by filtration to give HS54 (14) (0.046 g, 65 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₉H₁₃ClN₄O₂S 397.1; found 397.2.

1-(3-chlorophenyl)-6-(((6-chloropyridin-3-yl)methyl)thio)-1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-4-one, HS58 (15)

To a solution of 7c (0.050 g, 0.18 mmol, 1.0 equiv.) in anhydrous dichloromethane (1.0 mL) with N,N-diisopropylethylamine (0.094 mL, 0.54 mmol, 3.0 equiv.) was added 2-chloro-5-(chloromethyl)pyridine (0.032 g, 0.20 mmol, 1.1 equiv.). The reaction was heated to 30°C for 12 hours and then partitioned between water (2 mL) and ethyl acetate (2 mL). The resulting precipitate was isolated by filtration to give HS58 (15) (0.026 g, 36 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₇H₁₁Cl₂N₅OS 404.0; found 404.1.

Methyl 2-((1-(3-chlorophenyl)-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)acetate, HS61 (16)

To a solution of 7c (0.050 g, 0.18 mmol, 1.0 equiv.) in anhydrous dichloromethane (1.0 mL) with N,N-diisopropylethylamine (0.094 mL, 0.54 mmol, 3.0 equiv.) was added methyl bromoacetate (0.018 mL, 0.20 mmol, 1.1 equiv.). The reaction was heated to 30°C for 12 hours and then partitioned between water (2 mL) and ethyl acetate (2 mL). The resulting precipitate was isolated by filtration to give HS61 (16) (0.035 g, 56 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₄H₁₁ClN₄O₃S 351.0; found 351.1.

1-(3-chlorophenyl)-6-((2-oxo-2-(pyridin-2-yl)ethyl)thio)-1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-4-one, HS62 (17)

To a solution of 7c (0.050 g, 0.18 mmol, 1.0 equiv.) in anhydrous dichloromethane (1.0 mL) with N,N-diisopropylethylamine (0.094 mL, 0.54 mmol, 3.0 equiv.) was added 2-bromo-1-(2-pyridinyl)-1-ethanone hydrobromide salt (0.055 g, 0.20 mmol, 1.1 equiv.). The reaction was heated to 30°C for 12 hours and then partitioned between water (2 mL) and ethyl acetate (2 mL). The resulting precipitate was isolated by filtration to give HS62 (17) (0.024 g, 34 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₈H₁₂ClN₅O₂S 398.0; found 398.2.

1-(3-chlorophenyl)-6-((2-(2-hydroxyethoxy)ethyl)thio)-1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-4-one, HS63 (18)

To a solution of 7c (0.050 g, 0.18 mmol, 1.0 equiv.) in anhydrous DMF (1.0 mL) with N,N-diisopropylethylamine (0.094 mL, 0.54 mmol, 3.0 equiv.) was added 2-(2-chloroethoxy)-ethanol (0.021 mL, 0.20 mmol, 1.1 equiv.). The reaction was heated to 50°C for 12 hours and then partitioned between water (5 mL) and ethyl acetate (5 mL). The organic layer was isolated and washed with water (2 × 5 mL) and condensed to a solid. Trituration with ether (5 mL) gave HS63 (18) (0.026 g, 39 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C¹⁵H¹⁵ClN⁴O³S 367.1; found 367.2.

1-(3-chlorophenyl)-6-(methylthio)-1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-4-one, HS73 (19)

To a solution of 7c (0.050 g, 0.18 mmol, 1.0 equiv.) in anhydrous DMF (1.0 mL) with N,N-diisopropylethylamine (0.094 mL, 0.54 mmol, 3.0 equiv.) was added iodomethane (0.012 mL, 0.20 mmol, 1.1 equiv.). The reaction was heated to 60°C for 12 hours and then diethyl ether (5 mL) was added to precipitate the product. The resulting precipitate was isolated by filtration and washed with water and diethyl ether to give HS73 (19) (0.035 g, 67% yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₂H₉ClN₄OS 293.0; found 293.1.

Methyl 2-((1-(3-chlorophenyl)-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)-2-methylpropanoate, HS74 (20)

To a solution of 7c (0.050 g, 0.18 mmol, 1.0 equiv.) in anhydrous DMF (1.0 mL) with N,N-diisopropylethylamine (0.094 mL, 0.54 mmol, 3.0 equiv.) was added methyl 2-bromoisobutyrate (Aldrich, Cat# 17457) (0.024 mL, 0.20 mmol, 1.1 equiv.). The reaction was stirred for 12 hours and then partitioned between ethyl acetate (5 mL) and water (5 mL). The organic layer was washed with water, condensed, and the resulting solid was triturated with a 1:1 solution of diethyl ether in hexane to give HS74 (20) (0.049 g, 72% yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₆H₁₅ClN₄O₃S 379.1; found 379.1.

Methyl 2-((1-(3-chlorophenyl)-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)butanoate, HS75 (21)

To a solution of 7c (0.050 g, 0.18 mmol, 1.0 equiv.) in anhydrous DMF (1.0 mL) with N,N-diisopropylethylamine (0.094 mL, 0.54 mmol, 3.0 equiv.) was added methyl 2-bromobutyrate (Aldrich, Cat# 237310) (0.022 mL, 0.20 mmol, 1.1 equiv.). The reaction was heated to 60°C for 12 hours and then partitioned between ethyl acetate (5 mL) and water (5 mL). The organic layer was washed with water, condensed, and the resulting solid was dissolved in dichloromethane (1 mL) and precipitated by addition of diethyl ether (20 mL). The resulting solid was isolated by filtration and washed with diethyl ether to give HS75 (21) (0.023 g, 34% yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₆H₁₅ClN₄O₃S 379.1; found 379.1.

6-((2-hydroxyethyl)thio)-1-phenyl-1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-4-one, HS83 (22)

To a solution of 7a (0.050 g, 0.20 mmol, 1.0 equiv.) in anhydrous DMF (1.0 mL) with N,N-diisopropylethylamine (0.107 mL, 0.61 mmol, 3.0 equiv.) was added 2-bromoethanol (0.016 mL, 0.23 mmol, 1.1 equiv.). The reaction was stirred for 12 hours and then water (0.2 mL) and AcOH (0.050 mL) were added. Preparative HPLC (5 – 100 % methanol in water with 0.2 % formic acid modification over 20 minutes) provided HS83 (22) (0.021 g, 36 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₃H₁₂N₄O₂S 289.1; found 289.1.

2-((4-oxo-1-phenyl-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)acetonitrile, HS85 (23)

To a solution of 7a (0.100 g, 0.40 mmol, 1.0 equiv.) in anhydrous DMF (2.0 mL) with N,N-diisopropylethylamine (0.214 mL, 1.2 mmol, 3.0 equiv.) was added chloroacetonitrile (0.026 mL, 0.40 mmol, 1.0 equiv.). The reaction was stirred for 12 hours and then water (0.2 mL) and AcOH (0.050 mL) were added. Preparative HPLC (5 – 100 % methanol in water with 0.2 % formic acid modification over 20 minutes) provided HS85 (23) (0.060 g, 52 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₃H₉N₅OS 284.1; found 284.1

Methyl 2-((1-(3-chlorophenyl)-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanoate, (24)

To a solution of 7c (1.80 g, 6.5 mmol, 1.0 equiv.) in anhydrous THF (40 mL) with N,N-diisopropylethylamine (2.3 mL, 13 mmol, 2.0 equiv.) was added methyl 2-bromopropionate (Aldrich, Cat# 167185) (0.71 mL, 6.39 mmol, 0.98 equiv.). The reaction was stirred for 12 hours and condensed to a solid, which was triturated with water and diethyl ether to give 24 (2.04 g, 86 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₅H₁₃ClN₄O₃S 365.1; found 365.1.

Methyl 2-((4-oxo-1-phenyl-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanoate, HS87 (25)

To a solution of 7a (0.050 g, 0.20 mmol, 1.0 equiv.) in anhydrous DMF (1.0 mL) with N,N-diisopropylethylamine (0.107 mL, 0.61 mmol, 3.0 equiv.) was added methyl 2-bromopropionate (0.025 mL, 0.23 mmol, 1.1 equiv.). The reaction was stirred for 12 hours and then water (0.2 mL) and AcOH (0.050 mL) were added. Preparative HPLC (5 – 100 % methanol in water with 0.2 % formic acid modification over 20 minutes) provided HS87 (25) (0.038 g, 56 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₅H₁₄N₄O₃S 331.1; found 331.1.

Methyl 2-((4-oxo-1-phenyl-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)acetate, HS88 (26)

To a solution of 7a (0.100 g, 0.40 mmol, 1.0 equiv.) in anhydrous DMF (2.0 mL) with N,N-diisopropylethylamine (0.214 mL, 1.2 mmol, 3.0 equiv.) was added methyl bromoacetate (0.038 mL, 0.40 mmol, 1.0 equiv.). The reaction was stirred for 12 hours and then water (0.2 mL) and AcOH (0.050 mL) were added. Preparative HPLC (5 – 100 % methanol in water with 0.2 % formic acid modification over 20 minutes) provided HS88 (26) (0.078 g, 60 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₄H₁₂N₄O₃S 317.1; found 317.1.

Methyl 2-methyl-2-((4-oxo-1-phenyl-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanoate, HS89 (27)

To a solution of 7a (0.050 g, 0.20 mmol, 1.0 equiv.) in anhydrous DMF (1.0 mL) with N,N-diisopropylethylamine (0.107 mL, 0.61 mmol, 3.0 equiv.) was added methyl 2-bromoisobutyrate (0.029 mL, 0.23 mmol, 1.1 equiv.). The reaction was stirred for 12 hours and then formic acid (0.25 mL) was added. Preparative HPLC (5 – 100 % methanol in water with 0.2 % formic acid modification over 20 minutes) provided HS89 (27) (0.048 g, 68 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₆H₁₆N₄O₃S 345.1; found 345.1.

Methyl 2-((4-oxo-1-phenyl-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)butanoate, HS90 (28)

To a solution of 7a (0.050 g, 0.20 mmol, 1.0 equiv.) in anhydrous DMF (1.0 mL) with N,N-diisopropylethylamine (0.107 mL, 0.61 mmol, 3.0 equiv.) was added methyl 2-bromobutyrate (0.026 mL, 0.23 mmol, 1.1 equiv.). The reaction was stirred for 12 hours and then formic acid (0.25 mL) was added. Preparative HPLC (5 – 100 % methanol in water with 0.2 % formic acid modification over 20 minutes) provided HS90 (28) (0.036 g, 51 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₆H₁₆N₄O₃S 345.1; found 345.1.

Ethyl 3-oxo-4-((4-oxo-1-phenyl-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)butanoate, HS91 (29)

To a solution of 7a (0.050 g, 0.20 mmol, 1.0 equiv.) in anhydrous DMF (1.0 mL) with N,N-diisopropylethylamine (0.107 mL, 0.61 mmol, 3.0 equiv.) was added ethyl 4-chloroacetoacetate (0.037 mL, 0.23 mmol, 1.1 equiv.). The reaction was stirred for 12 hours and then formic acid (0.25 mL) was added. Preparative HPLC (5 – 100 % methanol in water with 0.2 % formic acid modification over 20 minutes) provided HS91 (29) (0.075 g, 98 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₇H₁₆N₄O₄S 373.1; found 373.1.

Methyl 2-((1-(2-chlorophenyl)-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanoate, HS107 (30)

To a solution of 7b (0.100 g, 0.36 mmol, 1.0 equiv.) in anhydrous DMF (2.0 mL) with N,N-diisopropylethylamine (0.188 mL, 1.1 mmol, 3.0 equiv.) was added methyl 2-bromopropionate (0.040 mL, 0.36 mmol, 1.0 equiv.). The reaction was stirred for 12 hours and then formic acid (0.30 mL) was added. Preparative HPLC (40 – 100 % methanol in water with 0.2 % formic acid modification over 18 minutes) provided HS107 (30) (0.100 g, 76 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₅H₁₃ClN₄O₃S 365.0; found 365.1.

1-(2-chlorophenyl)-6-((2-hydroxyethyl)thio)-1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-4-one, HS108 (31)

To a solution of 7b (0.050 g, 0.18 mmol, 1.0 equiv.) in anhydrous DMF (2.0 mL) with N,N-diisopropylethylamine (0.070 mL, 0.54 mmol, 3.0 equiv.) was added methyl 2-bromoethanol (0.013 mL, 0.18 mmol, 1.0 equiv.). The reaction was stirred for 12 hours and then formic acid (0.30 mL) was added. Preparative HPLC (40 – 100 % methanol in water with 0.2 % formic acid modification over 18 minutes) provided HS108 (31) (0.041 g, 71 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₃H₁₁ClN₄O₂S 323.0; found 323.1.

1-(2-chlorophenyl)-6-((3-hydroxypropyl)thio)-1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-4-one, HS109 (32)

To a solution of 7b (0.050 g, 0.18 mmol, 1.0 equiv.) in anhydrous DMF (2.0 mL) with N,N-diisopropylethylamine (0.070 mL, 0.54 mmol, 3.0 equiv.) was added methyl 3-bromo-1-propanol (0.016 mL, 0.18 mmol, 1.0 equiv.). The reaction was stirred for 12 hours and then formic acid (0.30 mL) was added. Preparative HPLC (40 – 100 % methanol in water with 0.2 % formic acid modification over 18 minutes) provided HS109 (32) (0.031 g, 52 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₄H₁₃ClN₄O₂S 337.0; found 337.1.

Methyl 2-((1-(4-chlorophenyl)-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanoate, HS121 (33)

To a solution of 7d (0.100 g, 0.36 mmol, 1.0 equiv.) in anhydrous DMF (2.0 mL) with N,N-diisopropylethylamine (0.125 mL, 0.72 mmol, 2.0 equiv.) was added methyl 2-bromopropionate (0.040 mL, 0.36 mmol, 1.0 equiv.). The reaction was stirred for 12 hours, condensed to dryness, and water (10 mL) was added. The resulting precipitate was isolated by filtration, washed with water and ether to give HS121 (33). (0.129 g, 98 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₅H₁₃ClN₄O₃S 365.0; found 365.1.

Methyl 2-((4-oxo-1-(o-tolyl)-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanoate, HS122 (34)

To a solution of 7f (0.100 g, 0.39 mmol, 1.0 equiv.) in anhydrous DMF (2.0 mL) with N,N-diisopropylethylamine (0.125 mL, 0.72 mmol, 2.0 equiv.) was added methyl 2-bromopropionate (0.040 mL, 0.36 mmol, 1.0 equiv.). The reaction was stirred for 12 hours and then water (10 mL) was added. After stirring overnight, the resulting precipitate was isolated by filtration, washed with water and ether to give HS122 (34). (0.132 g, 98 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₆H₁₆N₄O₃S 345.1; found 345.1.

Methyl 2-((4-oxo-1-(m-tolyl)-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanoate, HS123 (35)

To a solution of 7g (0.100 g, 0.39 mmol, 1.0 equiv.) in anhydrous DMF (2.0 mL) with N,N-diisopropylethylamine (0.125 mL, 0.72 mmol, 2.0 equiv.) was added methyl 2-bromopropionate (0.040 mL, 0.36 mmol, 1.0 equiv.). The reaction was stirred for 12 hours and then water (10 mL) was added. After stirring overnight, the resulting precipitate was isolated by filtration, washed with water and ether to give HS123 (35). (0.118 g, 87 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₆H₁₆N₄O₃S 345.1; found 345.1.

Methyl 2-((1-isopropyl-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanoate, HS125 (36)

To a solution of 7e (0.100 g, 0.48 mmol, 1.0 equiv.) in anhydrous THF (2.0 mL) with N,N-diisopropylethylamine (0.166 mL, 0.95 mmol, 2 equiv.) was added methyl 2-bromopropionate (0.053 mL, 0.48 mmol, 1.0 equiv.). The reaction was stirred for 2 hours and condensed to dryness. Silica gel chromatography with gradient elution (0 – 5% methanol in dichloromethane) yielded HS125 (36). (0.119 g, 84 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₂H₁₆N₄O₃S 297.1; found 297.1.

Methyl 2-((1-(3-methoxyphenyl)-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanoate, HS126 (37)

To a solution of 7h (0.100 g, 0.36 mmol, 1.0 equiv.) in anhydrous THF (2.0 mL) with N,N-diisopropylethylamine (0.166 mL, 0.95 mmol, 2.6 equiv.) was added methyl 2-bromopropionate (0.040 mL, 0.36 mmol, 1.0 equiv.). The reaction was stirred for 2 hours and condensed to dryness. Silica gel chromatography with gradient elution (0 – 5% methanol in dichloromethane) yielded HS126 (37). (0.120 g, 92 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₆H₁₆N₄O₄S 361.1; found 361.1.

Methyl 2-((1-(4-methoxyphenyl)-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanoate, HS127 (38)

To a solution of 7i (0.100 g, 0.36 mmol, 1.0 equiv.) in anhydrous THF (2.0 mL) with N,N-diisopropylethylamine (0.166 mL, 0.95 mmol, 2.6 equiv.) was added methyl 2-bromopropionate (0.040 mL, 0.36 mmol, 1.0 equiv.). The reaction was stirred for 2 hours and condensed to dryness. Silica gel chromatography with gradient elution (0 – 5% methanol in dichloromethane) yielded HS127 (38). (0.112 g, 86 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₆H₁₆N₄O₄S 361.1; found 361.1.

Methyl 2-((1-(4-cyanophenyl)-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanoate, HS128 (39)

To a solution of 7k (0.100 g, 0.37 mmol, 1.0 equiv.) in anhydrous THF (2.0 mL) with N,N-diisopropylethylamine (0.166 mL, 0.95 mmol, 2.6 equiv.) was added methyl 2-bromopropionate (0.040 mL, 0.36 mmol, 1.0 equiv.). The reaction was stirred for 2 hours and condensed to dryness. Silica gel chromatography with gradient elution (5 – 10% methanol in dichloromethane) yielded HS128 (39). (0.130 g, 99 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₆H₁₃N₅O₃S 356.1; found 356.1.

Methyl 2-((1-(3-fluorophenyl)-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)butanoate, HS144 (40)

To a solution of 7l (0.100 g, 0.38 mmol, 1.0 equiv.) in anhydrous THF (4.0 mL) with N,N-diisopropylethylamine (0.133 mL, 0.76 mmol, 2.0 equiv.) was added methyl 2-bromobutyrate (0.044 mL, 0.38 mmol, 1.0 equiv.). The reaction was stirred for 2 hours and condensed to dryness. Silica gel chromatography with gradient elution (1 – 10% methanol in dichloromethane) yielded HS144 (40). (0.138 g, 99 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₆H₁₅FN₄O₃S 363.1; found 363.1.

Methyl 2-((1-(3-fluorophenyl)-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanoate, HS145 (41)

To a solution of 7l (0.100 g, 0.38 mmol, 1.0 equiv.) in anhydrous THF (4.0 mL) with N,N-diisopropylethylamine (0.133 mL, 0.76 mmol, 2.0 equiv.) was added methyl 2-bromopropionate (0.042 mL, 0.38 mmol, 1.0 equiv.). The reaction was stirred for 2 hours and condensed to dryness. Silica gel chromatography with gradient elution (1 – 10% methanol in dichloromethane) yielded HS145 (41). (0.111 g, 83 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₅H₁₃FN₄O₃S 349.1; found 349.1.

Methyl 2-((4-oxo-1-(3-(trifluoromethyl)phenyl)-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanoate, HS146 (42)

To a solution of 7m (0.100 g, 0.32 mmol, 1.0 equiv.) in anhydrous THF (4.0 mL) with N,N-diisopropylethylamine (0.112 mL, 0.64 mmol, 2.0 equiv.) was added methyl 2-bromopropionate (0.036 mL, 0.32 mmol, 1.0 equiv.). The reaction was stirred for 2 hours and condensed to dryness. Silica gel chromatography with gradient elution (1 – 10% methanol in dichloromethane) yielded HS146 (42). (0.129 g, 100 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₆H₁₃F₃N₄O₃S 399.1; found 399.1.

Methyl 2-((1-(3-cyanophenyl)-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanoate, HS177 (43)

To a solution of 7j (0.060 g, 0.22 mmol, 1.0 equiv.) in anhydrous DMF (2.0 mL) with N,N-diisopropylethylamine (0.078 mL, 0.44 mmol, 2.0 equiv.) was added methyl 2-bromopropionate (0.037 mL, 0.22 mmol, 1.0 equiv.). The reaction was stirred for 12 hours and condensed to a solid which was triturated with water and diethyl ether to yield HS177 (43). (0.064 g, 81 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₆H₁₃N₅O₃S 356.1; found 356.1.

4-(6-((1-methoxy-1-oxopropan-2-yl)thio)-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-1-yl)benzoic acid, HS179 (44)

To a solution of 7n (0.064 g, 0.22 mmol, 1.0 equiv.) in anhydrous DMF (2.0 mL) with N,N-diisopropylethylamine (0.078 mL, 0.44 mmol, 2.0 equiv.) was added methyl 2-bromopropionate (0.037 mL, 0.22 mmol, 1.0 equiv.). The reaction was stirred for 12 hours and condensed to a solid which was triturated with water and diethyl ether to yield HS179 (44). (0.084 g, 100 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₆H₁₄N₄O₅S 375.1; found 375.1.

General Procedure C (Scheme 4), 2-((4-oxo-1-phenyl-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)acetamides (1), (3 – 4), and (45 – 61)

Scheme 4.

Preparation of 1, 3 – 4, 45 – 61.

Methyl esters 16, 21, 24–28, 30, 33–44 were dissolved in methanolic ammonia (7 N, 5 mL) and heated to 60 – 90°C in a sealed vessel until starting material was consumed. Reactions were condensed to dryness and purified by crystallization from water and trituration with ethyl ether to give carboxamides 1, 3 – 4, 45 – 61.

2-((1-(3-chlorophenyl)-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanamide, HS38 (1)

General procedure C was followed using 24 to produce HS38 (1) as previously described (Carlson et al., 2013).

2-methyl-2-((4-oxo-1-phenyl-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanamide, HS94 (3)

General procedure C was followed using 27 (15 mg, 0.038 mmol) to produce HS94 (3) (16 mg, 100 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₅H₁₅N₅O₂S 330.1; found 330.1.

2-((1-(3-fluorophenyl)-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)butanamide, HS148 (4)

General procedure C was followed using 40 (86 mg, 0.24 mmol) to produce HS148 (4) (49 mg, 57 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₅H₁₄FN₅O₂S 348.1; found 348.1.

2-((1-(3-chlorophenyl)-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)acetamide, HS64 (45)

General procedure C was followed using 16 (8 mg, 0.023 mmol) to produce HS64 (45) (8 mg, 100 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₃H₁₀ClN₅O₂S 336.0; found 336.1.

2-((1-(3-chlorophenyl)-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)butanamide, HS76 (46)

General procedure C was followed using 21 (10 mg, 0.026 mmol) to produce HS76 (46) (10 mg, 100 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₅H₁₄ClN₅O₂S 364.1; found 364.1.

2-((4-oxo-1-phenyl-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanamide, HS92 (47)

General procedure C was followed using 25 (15 mg, 0.045 mmol) to produce HS92 (47) (16 mg, 100 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₄H₁₃N₅O₂S 316.1; found 316.1.

2-((4-oxo-1-phenyl-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)acetamide, HS93 (48)

General procedure C was followed using 26 (15 mg, 0.047 mmol) to produce HS93 (48) (16 mg, 100 % yield). LCMS (ESI Ion Trap) m/z: [M+H^]+ calcd for C₁₃H₁₁N₅O₂S 302.1; found 302.1.

2-((4-oxo-1-phenyl-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)butanamide, HS95 (49)

General procedure C was followed using 28 (15 mg, 0.044 mmol) to produce HS95 (49) (16 mg, 100 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₅H₁₅N₅O₂S 330.1; found 330.1.

2-((1-(2-chlorophenyl)-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanamide, HS112 (50)

General procedure C was followed using 30 (50 mg, 0.14 mmol) to produce HS112 (50) (48 mg, 100 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₄H₁₂ClN₅O₂S 350.0; found 350.1.

2-((1-(4-chlorophenyl)-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanamide, HS134 (51)

General procedure C was followed using 33 (60 mg, 0.16 mmol) to produce HS134 (51) (48 mg, 83 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₄H₁₂ClN₅O₂S 350.0; found 350.1.

2-((4-oxo-1-(o-tolyl)-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanamide, HS135 (52)

General procedure C was followed using 34 (60 mg, 0.17 mmol) to produce HS135 (52) (45 mg, 79 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₅H₁₅N₅O₂S 330.1; found 330.1.

2-((4-oxo-1-(m-tolyl)-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanamide, HS136 (53)

General procedure C was followed using 35 (60 mg, 0.17 mmol) to produce HS136 (53) (51 mg, 89 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₅H₁₅N₅O₂S 330.1; found 330.1.

2-((1-isopropyl-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanamide, HS137 (54)

General procedure C was followed using 36 (60 mg, 0.20 mmol) to produce HS137 (54) (53 mg, 93 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₁H₁₅N₅O₂S 282.1; found 282.1.

2-((1-(3-methoxyphenyl)-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanamide, HS138 (55)

General procedure C was followed using 37 (60 mg, 0.17 mmol) to produce HS138 (55) (49 mg, 84 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₅H₁₅N₅O₃S 346.1; found 346.1.

2-((1-(4-methoxyphenyl)-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanamide, HS139 (56)

General procedure C was followed using 38 (60 mg, 0.17 mmol) to produce HS139 (56) (55 mg, 95 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₅H₁₅N₅O₃S 346.1; found 346.1.

2-((1-(4-cyanophenyl)-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanamide, HS140, (57)

General procedure C was followed using 39 (60 mg, 0.17 mmol) to produce HS140 (57) (44 mg, 77 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₅H₁₂N₆O₂S 341.1; found 341.1.

2-((1-(3-fluorophenyl)-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanamide, HS149 (58)

General procedure C was followed using 41 (55 mg, 0.16 mmol) to produce HS149 (58) (46 mg, 87 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₄H₁₂FN₅O₂S 334.1; found 334.1.

2-((4-oxo-1-(3-(trifluoromethyl)phenyl)-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanamide, HS150 (59)

General procedure C was followed using 42 (65 mg, 0.16 mmol) to produce HS150 (59) (43 mg, 68 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₅H₁₂F₃N₅O₂S 384.1; found 384.1.

2-((1-(3-cyanophenyl)-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanamide, HS178 (60)

General procedure C was followed using 43 (32 mg, 0.090 mmol) to produce HS178 (60) (21 mg, 68 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₅H₁₂N₆O₂S 341.1; found 341.1.

4-(6-((1-amino-1-oxopropan-2-yl)thio)-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-1-yl)benzoic acid, HS180 (61)

General procedure C was followed using 44 (40 mg, 0.11 mmol) to produce HS180 (61) (35 mg, 92 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₅H₁₃N₅O₄S 360.1; found 360.1.

2-((1-(3-chlorophenyl)-4-oxo-4,5-dihydro-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)-N-(2-morpholinoethyl)propanamide, HS147 (62)

To a stirred solution of 24 (89 mg, 0.22 mmol) in tetrahydrofuran (5 mL) with N,N-diisopropylethylamine (0.078 mL, 0.44 mmol, 2.0 equiv.) was added 4-(2-aminoethyl)morpholine (0.145 mL, 1.11 mmol, 5.0 equiv.). The solution was heated to 65°C for 12 h and then condensed to a solid. Silica gel chromatography with gradient elution (1–10% methanol in dichloromethane), followed by trituration with methanol yielded HS147 (62) (67 mg, 65% yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₂₀H₂₃ClN₆O₃S 463.1; found 463.1.

1-(3-chlorophenyl)-6-((2,4-dinitrophenyl)thio)-1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-4-one (63)

To a stirred solution of 7c (2.2 g, 7.9 mmol) in methanol (20 mL) with N,N-diisopropylethylamine (4.2 mL, 24 mmol, 3.0 equiv.) was added 2,4-dinitrochlorobenzene (Aldrich, Cat# 138630) (1.92 g, 9.5 mmol, 1.2 equiv.). The solution was stirred for 24 h and condensed to a yellow solid. Silica gel chromatography with gradient elution (0–10% methanol in dichloromethane), followed by trituration with ethyl ether yielded 63 (1.7 g, 50% yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₇H₉ClN₆O₅S 445.0; found 445.0.

4-chloro-1-(3-chlorophenyl)-6-((2,4-dinitrophenyl)thio)-1H-pyrazolo[3,4-d]pyrimidine (64)

A Vilsmeier complex was generated by mixing phosphorus(V) oxychloride (Aldrich, Cat# 201170) (6.5 mL, 710 mmol, 18 equiv.) and DMF (21 mL, 272 mmol, 70 equiv.). The complex was added to a stirred suspension of 63 (1.73 g, 3.9 mmol, 1 equiv.) in chloroform (20 mL). The mixture was heated to reflux for 4 h and then cooled in an ice bath. Water (50 mL) was added slowly and the mixture was extracted with dichloromethane (4 × 100 mL). The combined organic layers were dried over Na2SO4, filtered, and condensed to give 64 (1.75 g, 97 % yield) as a yellow solid. LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₇H₈Cl₂N₆O₄S 463.0; found 462.9.

General Procedure D (Scheme 5), methyl 2-((4-amino-1-(3-chlorophenyl)-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanoates (65 – 68)

Scheme 5.

Thiol protection, Vilsmeier chlorination, and preparation of 65 – 68.

To a stirred solution of 64 (1.0 equiv.) in DMF was added amine (R4-NH₂, 2.0 – 6.0 equiv.) followed by N,N-diisopropylethylamine (3.0 equiv.). The solution was heated (40 – 60°C) for 2 – 12 h until starting chloride was consumed. Dithiothreitol (0.5 equiv.) was added and the mixture stirred for 2 – 4 h until removal of dinitrophenyl thioether was complete. Methyl 2-bromopropionate (1.0 eq) was then added. The reaction was stirred for 12 h and condensed to dryness. Silica gel chromatography with gradient elution (0–10% methanol in dichloromethane) yielded 65 – 68.

Methyl 2-((1-(3-chlorophenyl)-4-((2-methoxyethyl)amino)-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanoate, HS181 (65)

General procedure D was followed using 64 (100 mg, 0.22 mmol) and 2-methoxyethylamine (Aldrich, Cat# 241067) (0.44 mmol, 2.0 equiv.) to produce HS181 (65) (88 mg, 97 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₈H₂₀ClN₅O₃S 422.1; found 422.1.

Methyl 2-((4-amino-1-(3-chlorophenyl)-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanoate, HS188 (66)

General procedure D was followed using 64 (100 mg, 0.22 mmol) and methanolic ammonia (7 N, 0.063 mL, 0.44 mmol, 2.0 equiv.) to produce HS188 (66) (72 mg, 91 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₅H₁₄ClN₅O₂S 364.1; found 364.1.

Methyl 2-((1-(3-chlorophenyl)-4-((2-hydroxyethyl)amino)-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanoate, HS189 (67)

General procedure D was followed using 64 (100 mg, 0.22 mmol) and ethanolamine (0.027 mL, 0.44 mmol, 2.0 equiv.) to produce HS189 (67) (80 mg, 91 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₇H₁₈ClN₅O₃S 408.1; found 408.1.

Methyl 2-((1-(3-chlorophenyl)-4-(methylamino)-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanoate, HS194 (68)

General procedure D was followed using 64 (150 mg, 0.32 mmol) and methylamine (33 % wt in ethanol, 0.302 mL, 1.94 mmol, 6.0 equiv.) to produce HS194 (68) (98 mg, 80 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₆H₁₆ClN₅O₂S 378.1; found 378.1.

General Procedure E (Scheme 6), methyl 2-((4-amino-1-(3-chlorophenyl)-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanamides (5, 69 – 71)

Scheme 6:

Preparation of 5, 69 – 71.

Methyl esters 65–68 were dissolved in methanolic ammonia (7 N, 5 mL) and heated to 60 – 90°C in a sealed vessel until starting material was consumed. Reactions were condensed to dryness and purified by silica gel chromatography or trituration with methanol to give carboxamides 5, 69 – 71.

2-((1-(3-chlorophenyl)-4-((2-methoxyethyl)amino)-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanamide, HS182 (5)

General procedure E was followed using 65 (44 mg, 0.10 mmol). Silica gel chromatography with gradient elution (0–6% methanol in dichloromethane) yielded HS182 (5) (30 mg, 71 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₇H₁₉ClN₆O₂S 407.1; found 407.1.

2-((4-amino-1-(3-chlorophenyl)-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanamide, HS191 (69)

General procedure E was followed using 66 (36 mg, 0.10 mmol). Trituration with methanol (3 mL) yielded HS191 (69) (26 mg, 74 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₄H₁₃ClN₆OS 349.1; found 349.1.

2-((1-(3-chlorophenyl)-4-((2-hydroxyethyl)amino)-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanamide, HS192 (70)

General procedure E was followed using 67 (40 mg, 0.10 mmol). Silica gel chromatography with gradient elution (0–10% methanol in dichloromethane) yielded HS192 (70) (29 mg, 74 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₆H₁₇ClN₆O₂S 393.1; found 393.1.

2-((1-(3-chlorophenyl)-4-(methylamino)-1H-pyrazolo[3,4-d]pyrimidin-6-yl)thio)propanamide, HS195 (71)

General procedure E was followed using 68 (50 mg, 0.13 mmol). Trituration with methanol (3 mL) yielded HS195 (71) (46 mg, 96 % yield). LCMS (ESI Ion Trap) m/z: [M+H]⁺ calcd for C₁₅H₁₅ClN₆OS 363.1; found 363.1.

Quantification and Statistical Analysis

Kinase Assays

Graphpad Prism 7 was used for statistical analysis of kinase assay data. For each analysis, total n and SEM are presented in the figure legend. Curves were plotted using variable slope (four parameters) non-linear fit.

Tissue and Animal Studies

All values are presented as the mean ± S.E.M., with n indicating the number of animals (tissue experiments). Data were analyzed by Student’s t test (two-tailed). For comparison of multiple groups, significance was determined by two-way analysis of variance with Dunnett’s post hoc test. Differences were statistically significant when P < 0.05.

Data and Software Availability

All data are available upon request to the Lead contact.

The accession number for the crystal structure of DAPK3 in complex with HS38 reported in this paper is 5VJA. For statistics see Table S2.

Scheme 3.

Preparation of 2, 8–44.

Key Resources Table.

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
anti-MYL2	Santa Cruz Biotechnology	Cat# sc-517414
anti-panMYPT1	New England Peptide	Custom
anti-LC20	Santa Cruz Biotechnology	Cat# sc-48414; RRID:AB_2148042
Bacterial and Virus Strains
Escherichia coli BL21 (DE3)-R3	Structural Genomics Consortium, University of Oxford (Strain-Damerell et al., 2014)	NA
Escherichia coli BL21(DE3)pLysS	Promega	Cat# L1195
Chemicals, Peptides, and Recombinant Proteins
HS38	Own Production (Carlson et al., 2013)	N/A
Analogs of HS38	Own Production (This Paper)	N/A
MLCK	Justin MacDonald, University of Calgary Chicken Gizzard (Ngai et al., 1984)	N/A
ROCKα (ROCK2; residues 2–543)	EMD Millipore	Cat# 14–338
PKA	Justin MacDonald, University of Calgary Bovine Heart (Grassie et al., 2012)	N/A
Peptide substrate for DAPK3, KKKRQSRRSTQGVTL, corresponding to Arg690 to Lys701 of MYPT1	University of Calgary Peptide Services (MacDonald et al., 2001a)	Custom
Peptide substrate for Pim Kinases, RSRHSSYPAGT	Biomatik	Custom
LC20	Justin MacDonald, University of Calgary Chicken Gizzard (Hathaway and Haeberle, 1983)	NCBI accession: NP_990609
PreScission Protease for GST-tag removal	GE Healthcare	Cat# 27084301
[γ-32P]-ATP	Perkin Elmer	Cat# NEG035C005MC
ATP disodium salt	BioBasic	Cat# AB0020
Ni Sepharose Resin	GE Healthcare	Cat# 17–5268–01
Calyculin A	EMD Millipore	Cat# 208851
Phos-Tag SDS-PAGE	FujiFilm Wako	Cat# AAL-107
West Femto enhanced chemiluminescence reagent	Thermo Fisher	Cat# 34095
TEV Protease	Sigma-Aldrich	Cat# T4455
Ethyl-2-cyano-3-ethoxyacrylate	Sigma-Aldrich	Cat# E28005
phenylhydrazine	Sigma-Aldrich	Cat# P26252
2-chlorophenylhydrazine hydrochloride	Sigma-Aldrich	Cat# 109509
3-chlorophenylhydrazine hydrochloride	Matrix Scientific	Cat# 075249
4-chlorophenylhydrazine hydrochloride	Sigma-Aldrich	Cat# C65807
isopropylhydrazine hydrochloride	Matrix Scientific	Cat# 007266
o -tolylhydrazine hydrochloride	Alfa Aesar	Cat# A15767
m -tolylhydrazine hydrochloride	Acros	Cat# 139150010
3-methoxyphenylhydrazine hydrochloride	Matrix Scientific	Cat# 021395
4-methoxyphenylhydrazine hydrochloride	Matrix Scientific	Cat# 021131
3-cyanophenylhydrazine hydrochloride	Ark Pharm, Inc	Cat# AK-39869
4-cyanophenylhydrazine hydrochloride	Sigma-Aldrich	Cat# 453471
3-fluorophenylhydrazine hydrochloride	Matrix Scientific	Cat# 003925
3-(trifluoromethyl)phenylhydrazine hydrochloride	Matrix Scientific	Cat# 005466
4-hydrazinobenzoic acid hydrochloride	Alfa Aesar	Cat# B20326
benzoyl isothiocyanate	Sigma-Aldrich	Cat# 261653
chloroacetonitrile	Sigma-Aldrich	Cat# C19651
methyl 2-bromopropionate	Sigma-Aldrich	Cat# 167185
methyl 2-bromoisobutyrate	Sigma-Aldrich	Cat# 17457
methyl 2-bromobutyrate	Sigma-Aldrich	Cat# 237310
2,4-dinitrochlorobenzene	Sigma-Aldrich	Cat# 138630
phosphorus(V) oxychloride	Sigma-Aldrich	Cat# 201170
2-methoxyethylamine	Sigma-Aldrich	Cat# 241067
Critical Commercial Assays
Active Site-Directed Competition Binding Assay against 468 Kinases	Eurifins DiscoverX Corp	KINOMEscan Scanmax Screen
Ionotropic and G Protein-Coupled Receptor Assay	PDSP, Dept. of Pharmacology, UNC Chapel Hill https://pdspdb.unc.edu/pdspWeb/	Primary Binding (% Inhibition) Screen
Deposited Data
Human Protein Atlas	https://www.proteinatlas.org/ (Uhlen et al., 2015)	N/A
Expression Atlas European Bioinformatics Institute	https://www.ebi.ac.uk/gxa/home (Petryszak et al., 2016)	N/A
Crystal Structure of DAPK3 bound to HS38	(This Paper)	PDB: 5VJA
Crystal Structure of DAPK3 used as search model	(Huber et al., 2012)	PDB: 1P4F
Crystal Structure of Pim-1 bound to AMPPNP	(Qian et al., 2005)	PDB: 1XR1
Crystal Structure of Pim-1 bound to PimTide	(Pogacic et al., 2007)	PDB: 2C3I
Crystal Structure of DAPK1 bound to AMPPNP	(McNamara et al., 2009)	PDB: 3F5U
Experimental Models: Organisms/Strains
Sprague-Dawley rats	Charles River Laboratories,	Crl:SD Strain code:400
Renin Transgene (RenTg) Mice and Normotensive (WT) littermates	Duke Animal Models Core (Caron et al., 2002) Also available from Jackson Laboratory	Jackson Lab Cat# 007853
Oligonucleotides
Bacterial expression vector pGEX-6P1	GE Healthcare	Cat# 28954648
Recombinant DNA
cDNA encoding DAPK3 (a.a. 1–320)	Genome Systems Inc (Borman et al., 2002)	Cat# AI660136 NCBI accession: NP_001339
cDNA encoding Pim-1 (aa 1–312)	Structural Genomics Consortium, University of Oxford	Addgene Cat# 39132 UniProtKB: P11309
cDNA encoding Pim-2 (aa 1–311)	Structural Genomics Consortium, University of Oxford	Addgene Cat# 38867 UniProtKB: Q9P1W9
cDNA encoding Pim-3 (aa 1–326)	Structural Genomics Consortium, University of Oxford	UniProtKB: Q86V86
cDNA encoding full-length chicken MYPT1	Dr. David Hartshorne, University of Arizona	PPP1R12A, NCBI accession: NP_990454
Software and Algorithms
Graphpad Prism 7	Graphpad Software	N/A
ImageQuant TL software	GE Healthcare	N/A
Chemiluminescence Analysis Multi-Gauge version 3.0	FujiFilm	N/A
Crystallographic Data Reduction	MOSFLM	N/A
Crystallographic Data Scaling	SCALA	N/A
Crystallographic Data Phasing	AMoRE	N/A
UCSF Chimera	https://www.cgl.ucsf.edu/chimera/ (Pettersen et al., 2004)	N/A
PoseView, ZBH Center for Bioinformatics	https://poseview.zbh.uni-amburg.de/ (Stierand and Rarey, 2010)	N/A
Powerlab Data Acquisition System with Labchart	ADInstruments	N/A
Other
P81 ion exchange cellulose chromatography paper	Whatman	Cat# 3698325
PVDF membrane	Sigma	Cat# 3010040001