Development of β-Amino Alcohol Derivatives that Inhibit Toll-Like Receptor 4 Mediated Inflammatory Response as Potential Antiseptics

Abstract

Toll-like Receptor 4 (TLR4) induced pro-inflammatory signaling has been directly implicated in severe sepsis and represents an attractive therapeutic target. Herein, we report our investigations into the structure-activity relationship and preliminary drug metabolism/pharmacokinetics study of β-amino alcohol derivatives that inhibit the TLR4 signaling pathway. Lead compounds were identified from in vitro cellular examination with µM potency for their inhibitory effects on TLR4 signaling and subsequently assessed for their ability to suppress the TLR4-induced inflammatory response in an ex vivo whole blood model. In addition the toxicology, specificity, solubility, brain-blood barrier permeability, and drug metabolism of several compounds were evaluated. Although further optimizations are needed, our findings lay the groundwork for the future drug development of this class of small molecule agents for the treatment of severe sepsis.

Introduction

The protective functions of innate-immunity rely principally on the ability of an organism to discriminate between endogenous signals and exogenous danger signals of microbial origin.^1,2 Toll-like receptors (TLRs) are a conserved family of integral membrane proteins, which maintain this essential role of innate immunity. At least 10 TLRs have been identified in human and thirteen in mice, each of which recognizes its own group of conserved pathogen associated molecular patterns (PAMPs).

Toll-like Receptor 4 (TLR4), the first of the TLR family to be identified, along with its accessory protein myeloid differentiation factor 2 (MD2), forms a heterodimeric complex. This complex specifically recognizes lipopolysaccharide (LPS), a structurally variant component of gram-negative bacterial cell walls.^3,4 Once LPS is recognized, a homo-heterodimer consisting of two TLR4-MD2-LPS complexes is formed. This homo-heterodimer confers an intracellular signaling cascade that results in downstream activation and nuclear translocation of the transcription factor, Nuclear Factor κB (NF-κB).⁵ Nuclear translocation of NF-κB results in subsequent transcriptional up-regulation of pro-inflammatory cytokines such as interleukin 1 (IL-1), interleukin 6 (IL-6), and tumor necrosis factor-α (TNF-α).^2,5 Given the key role TLR4 plays in regulating cytokines, inhibition of TLR4-mediated signaling is an appealing target for therapeutic development.

Although TLR4-induced cytokine signaling is generally advantageous in fighting infection, TLR4 signaling dysregulation has been directly implicated in a myriad of diseases including sepsis and neuropathic pain.^6–9 Severe sepsis poses a particular threat as a serious medical condition characterized by a systemic inflammation state that results in over 200,000 attributed deaths annually in the United States.¹⁰ Several groups have attempted to design potent anti-sepsis agents by employing the strategy of developing LPS-mimicking antagonists of TLR4.^11,12 One advantage of deriving inhibitors by this method is that antagonists can be developed by rationally modifying the active glycolipid component of LPS to optimize drug efficacy. The disadvantage posed is that these LPS mimetics are typically too large to render the desired pharmacological properties.¹³ A recent example of such a design is Eritoran, a lipid-A mimetic with high affinity for TLR4. Although it reached Phase III clinical trials as an antisepsis agent, Eritoran failed to demonstrate sufficient efficacy in late stage human trials. An alternative strategy, which has the potential for high specificity and bioavailability, is to design low-molecular- weight inhibitors.^14,15 Using this strategy, several small molecules have been developed to target and antagonize TLR4. To the best of our knowledge, none have successfully advanced through late stage clinical trials.^16,17,18 TAK-242 (resatorvid), the most successful TLR4-binding small molecule antagonist, reached Phase III clinical trials as an antisepsis agent but studies were recently discontinued due to a failure to suppress cytokine levels in patients despite showing promising preclinical efficacy in animal models.^19–21 These results might suggest that the current approaches to targeting the TLR4 receptor are flawed. There is an urgent need to develop and validate novel strategies to target the TLR4 pathway for antisepsis therapeutic development.

Using an in silico screen, we have identified β-amino alcohol derivatives (1) as generic inhibitors that disrupt the TLR4/MD-2 complex formation.^22,23 These inhibitors were shown to be effective at suppressing NF-κB activation in TLR4-overexpressing Human Embryonic Kidney (HEK) 293.²² Herein, we further these studies on the preparation, optimization, and biological evaluation of β-amino alcohol derivatives that inhibit LPS-induced inflammation in whole blood. Our findings may lay the groundwork for a new drug development strategy for the treatment of severe sepsis by targeting the TLR4/MD-2 interface and suggest a class of prototype small molecule agents.

Inhibitor Preparation

Synthesis of β-Amino Alcohol Derivatives

A β-amino alcohol of type 1, can be derived from a substituted aryl alkoxy epoxide of type 2 and a functionalized pyrazole compound such as 3 (Figure 1). The epoxide fragments 2 were derived from epichlorohydrin and a phenol derivative, 4. To generate an amine fragment of type 3, a Mannich-type reaction was utilized to further functionalize a 3,5-dimethyl-1H-pyrazole intermediate where paraformaldehyde provided the methylene source and methyl amine hydrochloride.²²

Figure 1.

Retrosynthesis of pyrazole containing β-amino alcohols.

The desired aryl alkoxy oxiranes (2) were synthesized via the reaction of the sodium salt of substituted phenols (4) with epichlorohydrin (Scheme 1).²⁴ Selective incorporation of an allyl ether group was accomplished starting from hydroquinone (R=OH, K₂CO₃, allyl bromide, acetone) to yield the epoxide 2c after addition to epichlorohydrin.²⁵

Scheme 1. Synthesis of β-Amino Alcohol Derivatives.

Methods: A = EtOH, reflux; B = MeCN, K₂CO₃, reflux; C = toluene, reflux; D = EtOH, MW 200 W, 140 °C; E = MeCN, K₂CO₃, MW 200 W, 140 °C. See experimental section for specific methods and yields.

The functionalized heterocycles of type (3) were synthesized using previously established methodology.²¹ Preparation of analogues containing a styrene functionality was achieved using (2-bromobenzyl) pyrazole (6f). A palladium-catalyzed Stille coupling, using Pd(PPh₃)₄ and tributylvinyltin in toluene, generated the styrene pyrazole intermediate (6g).²⁶ A Mannich coupling with methyl amine and paraformaldehyde formed the amine derivative, 3f.

Preparation of the β-amino hydroxyl compounds was carried out by condensing amino substituted pyrazoles with the aryloxy oxiranes. We found that subjecting one equivalent of an aryl alkoxy oxirane (2) and a slight excess (1.1 equilvalent) of the functionalized secondary amine (3) to different reaction conditions, we could improve the isolated yields of the β-amino hydroxyl compounds. Due to limited solubility of many of the starting materials, ethanol was commonly used as the solvent. Unfortunately, we frequently observed the addition of ethanol to the oxirane under refluxing conditions. We successfully supressed this competing reaction in several reactions with the use of either acetonitrile²⁷ or toluene.²⁸

Flitsch and coworkers reported good yields of amino alcohols via the ring opening of epoxides with a variety of amines under microwave irradiation conditions.²⁹ Conducting the coupling reaction under similar conditions (140 °C, 200 W) using a variety of solvent systems, resulted in a significant decrease in reaction times (24–48 hours to less than one hour) and generated excellent yields. We found that either acetonitrile or EtOH worked equally well with under the microwave irradiation conditions with only minor traces of the EtOH addition products.

In order to explore the role the alcohol group of 1 plays in binding to TLR4, compound 1j was methylated utilizing the standard conditions (MeI, NaH) to generate the methyl ether, 7.

Synthesis of Chiral Derivatives

In order to determine if the stereochemistry of the hydroxyl group was essential to the ligand/receptor interaction, the enantiomers of 1s (R=p-Cl, R₁=o-Cl) were synthesized (Scheme 2). Previously synthesized enantiomers of 1j (R=p-OEt, R₁=o-Cl) were also prepared for direct comparison in the current biological assays. Published reports utilized Jacobsen’s hydrokinetic resolution methodology to generate enantioenriched epoxides that could then be opened with the secondary amine nucleophile.³⁰ The active (R,R)(salen)Co(III)OAc catalyst was generated in situ by the aerobic oxidation of [(R,R)-N,N’-bis(3,5-di-tert-butylsalicylidene)-1,2-cyclohexanediamino)]cobalt(II) in the presence of acetic acid. Treatment of the desired racemic epoxides (2b and 2d) with (R,R)(salen)Co(III)OAc catalyst and water yielded the (S)-enantioenriched epoxides (2h, R=p-OEt and 2j, R=p-Cl) and the ring opened (R)-diol products. The corresponding (S,S)(salen)Co(III)OAc catalyst was utilized to generate the desired (R)-epoxides (2i, R=p-OEt and 2k, R=p-Cl).^31,32 We later found that quenching the reaction with pyridinium p-toluene sulfonate (PPTS) (4 eq PPTS/mmol of catalyst) in a 1:1 mixture of MeCN and CH₂Cl₂ removed the catalyst prior to purification, increased the chiral epoxide yields.^33,34

Scheme 2.

Synthesis of Chiral Epoxides.

Generation of the chiral β-amino alcohols (1l, 1k, 1u, and 1t) was accomplished by using similar reaction conditions to their racemic counterparts. We found that both toluene and acetonitrile were effective solvents for these coupling. The addition of a heterogeneous base, K₂CO₃, increased the overall yields and the use of microwave conditions did not affect the enantioselectivity or the yields (Scheme 3). The enantiomeric excess was determined to be greater than 96% by chiral HPLC analysis for each of the derivatives regardless of the reaction conditions.

Scheme 3.

Synthesis of Enantioenriched β-amino Alcohol.

Synthesis of Cyclic Derivatives

A commonly used strategy to improve the potency of flexible small molecule inhibitors is to prepare the cyclic derivatives with increased rigidity, which reduces the entropic penalty upon binding to their protein receptor.³⁵ The macrocycle scaffold (8) was designed based on one of our more active linear compounds, 1q, and is properly substituted to undergo ring closing metathesis through a convenient annulation method (Scheme 4). The desired macrocyclic compound 8 was synthesized by cyclizing the linear precursor 1q using Grubbs II catalyst with a 52% yield and greater than 1:10 E/Z selectivity after purification.³⁶

Scheme 4.

Synthesis of a cyclic derivative.

Biological Evaluation

Inhibition of the LPS-Induced TLR4 Activation in Macrophage Cells

The potency of the β-amino alcohol derivatives as inhibitors of TLR4 signaling was determined by monitoring the reduction of LPS-induced production of nitric oxide (NO) in RAW 264.7 macrophage cells. The IC₅₀ values and the margin of error were determined by a modified Kou-type method from experiments conducted in quadruplicate.³⁷

The compound that lacked substitution on either of the aryl rings (e.g. 1a, R=R₁=H) was ineffective at suppressing NO signaling (Table 1). For this reason, compound 1a was utilized as a negative control for this assay. Mild suppression of NO production was observed when substitutions were made to either of the benzyl rings (1e IC₅₀=51.8 ± 5.8 µM and 1b IC₅₀=64.3 ± 0.1 µM) suggesting substitution of both aryl rings is needed for activity.

Table 1.

Suppression of the LPS-induced NO production in RAW 264 macrophage cells.

	*	R	R1	R2	IC50 (µM)
1a	+/−	H	-H	H	>100
1b	+/−	H	o-Cl	H	64.3±0.1
1c	+/−	p-CH3OCH2CH2-	o-Cl	H	57.3±2.8
1d	+/−	o-CH3O-	o-Cl	H	92.5±0.8
1e	+/−	p-CH3CH2O-	-H	H	51.8±5.8
1f	+/−	p-CH3CH2O-	o-CH3	H	47.7±3.7
1g	+/−	p-CH3CH2O-	o-OCH3	H	>100
1h	+/−	p-CH3CH2O-	p-OCH3	H	67.6±0.3
1i	+/−	p-CH3CH2O-	o-F	H	47.7±2.7
1j	+/−	p-CH3CH2O-	o-Cl	H	27.8±0.3
1k	(R)	p-CH3CH2O-	o-Cl	H	24.8±3.0
1l	(S)	p-CH3CH2O-	o-Cl	H	24.3±3.7
7	+/−	p-CH3CH2O-	o-Cl	CH3	19.8±0.1
1m	+/−	p-CH3CH2O-	m-Cl	H	16.5±0.8
1n	+/−	p-CH3CH2O-	3,4-dichloro	H	21.8±3.1
1o	+/−	p-CH3CH2O-	o-CH=CH2	H	24.7±9.0
1p	+/−	p-CH3CH2O-	o-CF3	H	32.1±2.2
1q	+/−	p-CH2=CHCH2O-	o-CH=CH2	H	23.1±1.1
1r	+/−	p-CH2=CHCH2O-	o-Cl	H	44.6±7.0
1s	+/−	p-Cl-	o-Cl	H	16.1±1.1
1t	(R)	p-Cl-	o-Cl	H	15.5±0.9
1u	(S)	p-Cl-	o-Cl	H	15.6±0.6
1v	+/−	3,4-dichloro	o-Cl	H	20.1±0.1
1w	+/−	p-Cl-	o-CH=CH2	H	23.2±0.7
1x	+/−	p-CF3-	o-CF3	H	27.3±0.0
1y	+/−	p-CF3-	o-Cl	H	23.7±2.6
1z	+/−	p-CN-	o-Cl	H	52.7±2.6

The small structural variations between compounds 1b, 1c, 1e, and 1j and the corresponding biological activity gave preliminary indications that a combination of electronic properties and lipophilicity may play a role in improving the biological activity. Several additional compounds were synthesized to further explore the effect of substitution located on the aromatic ring derived from the epoxide fragment.

Using 1j as the prototype compound, (R=p-OEt, R₁=o-Cl: IC₅₀=27.8 ± 0.3 µM) several analogues were prepared in which the R group was modified to explore the influence of electronics and lipophilicity. Synthesis of derivative 1v, which incorporated an additional chloride substitution (R=3,4-dichloro-, R₁=o-Cl), resulted in a slight decrease of activity from compound 1s (R=p-Cl, R₁=o-Cl: IC₅₀=16.1 ± 1.1 µM). Increasing the lipophilicity (R=p-CF₃, R₁=o-Cl, 1y) generated a slight loss of NO suppression compared with that of 1s. Interestingly, significant loss of activity was observed with compound 1z where R=p-CN (IC₅₀=52.7 ± 2.6 µM).

Based on the current SAR set, it would appear that an electron withdrawing group verses an electron donating group on the phenoxy ring at the para position favors activity (e.g. 1s [R=p-Cl] > 1j [R=p-OEt] > 1a [R=H]). Although there seems to be a slight dependence on the electronic nature of the aromatic ring, activity appears to be influenced more by increased lipophilicity.

The effect of substitution on the aromatic ring attached to the pyrazole ring was also studied. Starting again from 1j as a baseline, the R group was maintained as the p-OEt functionality and the R₁ substituent was varied. The use of a methyl group at the ortho position (1f) resulted in a compound with similar activity to that of the monosubstituted compound 1e (R=p-OEt, R₁=H). Addition of an o-OMe group (1g) gave a compound with no detectable NO suppression. Although relocating the methoxy group to the para position (1h) regenerated some activity (IC₅₀=67.6 ± 0.3 µM) the increase in potency was not significant enough to justify further exploration of electron donating groups on the amine fragment.

Modification of o-R₁ to a fluorine substituent (1i) also did not improve activity. However, incorporation of a more lipophilic, electron withdrawing group such as chlorine (1j) reduced the IC₅₀ to nearly half of that seen with fluorine (47.7 ± 2.7 µM compared to 27.8 ± 0.3 µM). Relocating the chlorine substituent to the meta position (1m) further improved the IC₅₀ to 16.5 ± 0.8 µM, which is comparable to the IC₅₀ of compound 1s. Introducing disubstitution, e.g. R₁=3,4-dichloro (1v), however, did not significantly effect the NO suppression compared to the mono-chloro derivative.

As seen with the modifications made to phenol group, the addition of an o-CF₃ group resulted in a modest decrease in activity (1p, IC₅₀=31.1 ± 2.2 µM) compared to that of 1j. Interestingly, when a vinyl group (1o) was incorporated into the structure, the inhibitory potency was reduced to an IC₅₀ of 24.7 ± 9.0 µM. These results imply that the ortho position may be more sensitive to sterics than to electronics.

The effect of the hydroxyl group’s stereochemistry on NO production was also investigated. Surprisingly, no significant difference in activity between the racemic derivatives and their stereoisomers (1j, 1k and 1l) and (1s, 1t and 1u) was observed. The results suggest that the role of the alcohol in the small molecule-protein interaction remains unclear and may not be critical for potency. This hypothesis was supported by the methyl ether analogue, 7. The ether derivative demonstrated reasonable activity with an IC₅₀ of 19.8 ± 0.1 µM, implying that an alcohol substitution may not be essential but the C–O– functionality may be important.

An interesting result was obtained with the macrocyclic compound 8. Compared to the acyclic system (1q, IC₅₀=23.1 ± 1.1 µM) the macrocycle retained its potency, (IC₅₀=16.0 ± 2.0 µM) suggesting the possibility of an alternative scaffold.

Inhibition of Inflammatory Response in Human Whole Blood

The biological properties of compounds 1j and 1a were further investigated in an effort to determine if this class of compounds would be suitable for further studies as antisepsis agents.³⁸ In human whole blood, 1j and the control compound, 1a, showed differential efficiency in inhibiting the LPS-induced cytokine release. Compound 1j was more effective at reducing IL-8 secretion than 1a, and this difference was significant at concentrations of 50 and 250 µM (p<0.05, T-test) (Figure 2A and 2B). Suppression of TNFα release was also observed for 1j at a dose of 250 µM. However, at concentrations of 500 µM, both compounds reduced the levels of all of the studied cytokines equally well, suggesting that non-specific inhibition becomes prevalent at this concentration. These inhibitory effects by 1j have also been shown to be cytokine-specific. For example, 1j and 1a had no observable differences in their inhibitory effects on IL-6 cytokine release.

Figure 2.

(A) Dose-dependent inhibitory effects of 1a and 1j on LPS-mediated cytokine release in human whole blood. (B) Inhibitory effects of compounds 1a and 1j at a fixed concentration of 250 µM on LPS-induced cytokine release in human whole blood. Significant difference between the two compounds was observed for inhibiting IL-8 and TNFα, but not for IL-6 (** p<0.01, * p<0.05, t test).

Compound 1j reduced the plasma levels of IL-8 to 35% and those of TNFα and IL-6 to 57% and 50%, respectively, at a concentration of 250 µM. This data is consistent with the estimated IC₅₀ values for compound 1j of 110.5 µM for IL-8, 315.6 µM for TNF-α, and 318.4 µM for IL-6 using the human whole blood model and the secretion of the respective cytokines. The elevated IC₅₀ values seen in the whole blood assay compared to the results derived from the cell-culture based assays are perhaps due to the complexity of the ex vivo whole blood system. Consistent with the in vitro assay results, the control compound 1a showed a significantly lower potency to inhibit the inflammatory response in the whole blood model. Importantly, no agonistic effects with respect to cytokine release or hemolysis in the presence of up to 1 mM of either compound were observed. Taken together, our observations imply that 1j is effective to suppress LPS-induced inflammatory response in human whole blood.

Drug Metabolism/Pharmacokinetics (DMPK)

To preliminarily evaluate the drug potential of 1j, the metabolic stability, blood brain barrier (BBB) permeability, and lipophilicity were investigated. To determine the lipophilicity of 1j, the octanol/water distribution coefficient was measured at the physiologically relevant pH of blood serum (pH=7.4). The logD value of a drug at pH 7.4 is known as a predictor of ADME/TOX properties. The logD of 1j was determined to be 3.36, a value marginally above the optimum for drug development, but still within a functional range, showing potential for a balance of permeability and solubility.^39,40

The BBB permeability of 1j was evaluated using the MDR1-MDCK permeability assay, a cell culture model system shown to be an accurate predictor of BBB penetration potential. To determine penetration efficiency, the ability of 1j to pass through an MDR1-MDCK monolayer in both the apical to basolateral and basolateral to apical directions was measured. At 5 µM, 1j was determined to have a brain-blood barrier permeability classification of “high”.

The metabolic stability of 1j was also evaluated. The intrinsic clearance (CL_int) values of 1j (1 µM) were determined in the presence of human liver microsomes and human S9 fractions as well as rat S9 fractions. The observed values of CL_int in each system were determined to be >0.70, 0.3, and >0.35 mL/min/mg protein, respectively. This data suggests that 1j is likely to be metabolized in the liver at a relatively rapid rate. Generally, rapid metabolism for a drug candidate is undesirable and this issue should be addressed as SAR studies continue.

Further, a preliminary cytochrome P450 (CYP) screening was also conducted. Compound 1j was tested for its ability to inhibit four key cytochrome enzymes involved in drug metabolism. The IC₅₀ values for CYP1A2, CYP2C9, CYP2C19, and CYP2D6 were found to be greater than 100, 43, 20, and 32 µM, respectively. These data, in conjunction with the metabolism results indicate that further optimization of the specificity of 1j is needed before advancement to more clinically relevant in vivo models can be considered.

Specificity and Toxicity Tests

The β-amino alcohol derivatives were further examined for their specificity to TLR4 and screened against 12 representative kinases (AKT1, CAMK1, DDR2, GSK-3α, MAPK1, MET, PAK1, PDGFRB, PIM1, PKC-γ, PLK4, and SRC). The two active lead compounds, 1j and 1s, were screened (Figure 3) using a KinaseSeeker™ assay. (A luciferase fragment complementation assay that provides a direct measurement of their interactions with each respective kinase). Neither 1j or 1s inhibited a kinase by more than 27% at their respective IC₅₀ concentrations determined in the RAW264.7 macrophage cells, while the majority of the kinases examined retained >95% of their activity. The results suggest that the active dose of 1j or 1s required for TLR4 inhibition does not substantially perturb the activity of these biologically essential enzymes.

Figure 3.

Kinase Selectivity Screen for 1j (28 µM) and 1s (16 µM). Both compounds were profiled against a panel of 12 kinases using the KinaseSeeker™ assay. (A) Treatment with 1j resulted in retention of >95% activity for a majority of the kinases, and the most inhibited kinase, PKC-γ, was inhibited by 22.4%. (B) Treatment with 1s also resulted in the retention of 95% activity for a majority of the kinases and the most inhibited kinase, MARK1, was inhibited by 26.7%.

Toxicity assessments for the β-amino alcohol derivatives were carried out using a variety of methods. We have observed that the cell toxicity was variable and cell line dependent (Supplementary Figure 1 and 2). Most importantly, negligible hemolysis was observed at concentrations as high as 1 mM in human whole blood.

Conclusions

In summary, we have reported initial results regarding the development of a series of low molecular weight β-amino alcohol derivatives that appear to suppress TLR4-mediated inflammation response. Preliminary structure activity relationship studies revealed variation of functionality on both aryl rings could increase the potency. As part of our evaluation process, we have also demonstrated these compounds are capable of suppressing LPS-induced inflammatory response in both macrophage cells and a human whole blood ex vivo model.

Although the potency values of the compounds described herein are modest, it should be noted that they presumably function by disrupting the association of TLR4 and MD-2.²³ Energetically, it is a challenging goal to compete with two protein binding partners and block their association with small molecule agents.^15,41,42 The micromolar inhibitory activities observed with the β-amino alcohol derivatives render a promising starting point for further drug development optimization.

It is anticipated that as improvements in potency, metabolic stability, and safety occur, these β-amino alcohol derivatives will serve as useful probes in studying TLR4-mediated inflammation and could pave the way for future antisepsis therapeutics.

Experimental Section

Secreted Nitric Oxide Assay

RAW264.7 cells were grown in RPMI supplemented with 10% FBS, Pen-Strep, and L-Glutamine. RAW264.7 cells were then planted in 96 well plates at 100,000 cells per well and grown for 24 hrs in the media described previously.^43,44 After 24 h, the media was removed and replaced with unsupplemented RPMI (Invitrogen, Carlsbad, CA). Lanes were doped with the TLR agonists and simultaneously doped with small molecule drugs as specified for each experiment. Plates were then incubated at 37 °C for 18–24 h. Following 18–24 h, incubation 100 µL of media was removed and added to flat black 96 well microfluor plates (Thermo Scientific, Hudson, NH). Ten µL of 2,3-diaminonaphthalene (0.05 mg/mL in 0.62 M HCl) was added to each well and incubated for 15 min. Five µL 3 M NaOH was added to stop the reaction and the plate was read on Beckman Coulter DTX880 reader with excitation set at 365 nm and emission analyzed at 450 nm. Nitrite concentration was determined from nitrite standard curve.

Trypan Blue Toxicity Assay

Human Embryonic Kidney 293 (HEK293) cells were stably transfected with TLR4 and necessary assembly and signaling proteins (MD2, CD-14, LPSBP, etc.). Cells were cultured in DMEM supplemented with 10% FBS, penicillin (100 U/mL), streptomycin (100 mg/mL), l-glutamine (2 mM), 1 mg/mL normocin (InvivoGen, San Diego, CA) and 2 unit of HEK-Blue Selection solution (InvivoGen, San Diego, CA). Cells were implanted in 6 cm plates and grown to 65–85% confluency by incubating at 37 °C prior to drug treatment. On the day of treatment, the media was removed from the 6 cm plate and replaced with CSF buffer supplemented with drug treatment. After 24 h incubation at 37 °C, CSF was removed, and cells were agitated with 1× Trypsin and re-suspended in fresh DMEM supplemented media. After re-suspension, a 100 µL sample was taken from each 6 cm plate, mixed gently with 100 µL Trypan Blue and allowed to sit for 5 min. The ratio of blue stained cells to total cells was then quantified using a Bright Line 0.1 mm depth hemocytometer under a Nikon TMS light microscope.

Whole Blood Cytokine Release Assay

The following studies were approved by the regional ethics committee in Norway. Human whole blood was drawn from healthy individuals after giving their written consent. The whole blood model of sepsis has been described in detail previously.³⁸ Briefly, blood was drawn from healthy human donors into 4.5 mL cryo tube vials (Nunc, Roskilde, Denmark) containing lepirudin (Refludan^®, Pharmion, Copenhagen, Denmark) in a final concentration of 50 µg/mL. Lepuridin is a highly specific thrombin inhibitor preventing coagulation only, not interfering with other biological activities. Small molecule drugs or PBS with Ca²⁺ and Mg²⁺ (Sigma, D8662) were administered to 1.8 mL cryo tube vials prior to blood sampling. Immediately after the sample was obtained, blood was added to each of the respective tubes and preincubated with inhibitors at 37°C for five to ten minutes. Afterwards they were supplied with 100 ng/mL phenol-extracted ultra-pure LPS (upLPS) from E. coli 0111:B4 (InvivoGen, San Diego, CA, USA), or endotoxin-free PBS with Ca²⁺ and Mg². Samples were incubated rotating at 37°C for two hours. Then ethylenediaminetetraacetic acid (EDTA) (Sigma) was added to a final concentration of 10 mM, and the samples were centrifuged 15 min at 3220×g, 4°C. Plasma was obtained and stored at −80 °C until further analysis.

Plasma levels of TNF-α, IL-8, and IL-6 were determined using singleplex Bio-plex cytokine kits from BioRad (Hercules, Ca, USA). The analyses were performed according to the manufacturer’s recommendations. The plasma cytokine concentration measured upon LPS stimulation in absence of inhibitor was set to 100%. For statistical analyses, based on the dataset of n = 3 (Figure 2A) or n=9 (Figure 2B), a two-sided, paired Student’s T-test was performed based on single data points of % plasma cytokine levels.

Metabolic Stability

Metabolic Stability in Human Liver Microsomes

A concentration of 1 µM of compound 1j with <0.25% DMSO was incubated at 37 °C in buffer with 0.5 mg/mL of microsomal protein. The reaction was initiated by the addition of cofactors at 5 different time points (0, 10, 20, 30, and 60 min). A positive control (5 µM testosterone) was run in parallel with sampling at 1, 10, and 30 min. After the final time point, fluorimetry was used to confirm the addition of NADPH to the reaction mixture. LC-MS/MS was used to determine the peak area response ratio (peak area corresponding to 1j divided by that of an internal standard) without running a standard curve. The in vitro intrinsic clearance CL_{int in-vitro}=k /P (ml/min•mg protein) was determined, where k is the elimination rate constant (min⁻¹) and P is the protein concentration (mg/mL).

Metabolic Stability in Liver S9 Fractions

A concentration of 1 µM of compound 1j in <0.25% DMSO was incubated at 37 °C in buffer with 1 mg/mL of pooled S9 protein from both rat and human and 1 mM of NADPH, UDPGA, PAPS, and GSH. The reaction was initiated by the addition of cofactors at 5 different time points (0, 10, 20, 30, and 60 min). A positive control (20 µM testosterone and 20 µM 7-hydroxycoumarin) was run in parallel with sampling at 1, 10, 30, and 60 min. After the final time point, fluorimetry was used to confirm the addition of NADPH to the reaction mixture. LC-MS/MS was used to determine the peak area response ratio (peak area corresponding to 1j divided by that of an internal standard) without running a standard curve. The in vitro intrinsic clearance CL_{int in-vitro}=k /P (ml/min•mg protein) was determined, where k is the elimination rate constant (min⁻¹) and P is the protein concentration (mg/mL).

CYP IC₅₀ in Human Liver Microsomes

The IC₅₀ values of 1j for five CYP enzymes (1A2, 2C9, 2C19, 2D6, and 3A4) were determined using a pool of mixed-gender human liver microsomes (minimum of 10 donors). Compound 1j was incubated with pooled human liver microsomes (0.25 mg/mL) at 37 °C in the presence of phosphate buffer (100 mM, pH 7.4), MgCl₂ (5 mM), CYP-specific probe (at approximately Km), and NADPH (1 mM) at eight concentrations (0–100 µM). After a period of incubation, the samples were treated by the addition of protein precipitation solvent and centrifuged. The CYP enzyme activities were measured by determining the formation of the CYP probe metabolites by LC-MS/MS. The individual CYP isoform (specific probe/metabolite formed/known inhibitor used as the positive control) tested: CYP 1A2 (Phenacetin (150 µM)/acetaminophen/α-naphthoflavone); CYP 2C9 (Diclofenac (6 µM)/4’-OH diclofenac/sulfaphenazole); CYP 2C19 (S-mephenytoin (40 µM)/4’-OH mephenytoin/(+)-N-3-benzylnirvanol); CYP 2D6 (bufuralol (7 µM)/1’-OH bufuralol/Quinidine); CYP 3A4 (Testosterone (75 µM)/6 β-OH testosterone/Ketoconazole); CYP 3A4 (Midazolam (1.4 µM)/1’-OH midazolam/Ketoconazole). The IC₅₀ was estimated by fitting the experimental data to a sigmoidal model and non-linear regression analysis using GraphPad Prism^® software.

Kinase Profiling

The kinase profiling was performed by Luceome Biotechnologies, Tucson, AZ. Compounds 1j and 1s were dissolved in DMSO and tested at concentrations of 28 and 16 µM, respectively. Each compound was first evaluated for false positive against split luciferase. If they did not inhibit luciferase control than they were profiled in duplicate against the following kinases: AKT1, CAMK1, DDR2, GSK3α, MARK1, MET, PAK1, PDGFRB, PIM1, PKC-γ, PKL4, and SRC using the protocol described by Ghosh and coworkers.⁴⁵ The percent inhibition and percent activity remaining was calculated using the following equation:

$$\begin{array}{c}\hfill \%\text{Inhibition}=[(\text{ALUcontrol}-\text{ALU sample})/\text{ALU control}]\times 100\%\hfill \\ \hfill \%\text{Activity Remaining}=100-\%\text{Inhibition}\hfill \end{array}$$

General Chemistry Methods

All reactions were run under an inert atmosphere of either N₂ or Ar gas. Reaction solvents were purchased anhydrous and of HPLC quality. All other reagents were purchased from Sigma Aldrich and used without further purification. Yields were calculated for material judged homogenous by thin layer chromatography (TLC) and nuclear magnetic resonance (NMR). TLC was performed on Merck Kieselgel 60 F₂₅₄ plates, eluting with the solvent indicated, visualized by a 254 nm UV lamp, and stained with an ethanolic solution of phosphomolybdic acid-hydrate. Glassware for reactions was oven dried at 125 °C prior to use. Column flash chromatography was performed using silica gel (SiO₂) Premium R_f, 60 Ǻ, 200 × 400 mesh from Sorbent Technologies. Nuclear magnetic resonance spectra were acquired on a Bruker spectrometer (300 MHz for ¹H and 75 MHz for ¹³C) or a Varian Inova-400 (400 MHz for ¹H and 100 MHz for ¹³C). Chemical shifts for ¹H NMR spectra are reported in parts per million (ppm) and referenced to the signal of residual CDCl₃ at 7.26 ppm. Chemical shifts for ¹³C NMR and DEPT spectra are reported in parts per million (ppm) and referenced to the center line of the residual CDCl₃ triplet at 77.23 ppm. Chemical shifts of the unprotonated carbons (‘C’) for DEPT spectra were obtained by comparison with the ¹³C NMR spectrum. Enantiomeric excess was determined by chiral HPLC analysis using a Daicel Chiralcel AD-H silica column (length=25 cm) eluting with a mobile phase of an indicated percentage of i-PrOH / hexanes with 0.1% Et₂NH and at a flow rate of 0.5 mL/min. Retention times for the major and minor enantiomers were detected with Shimdzu SPD-6A UV spectrometric detector at 254 nm. Optical rotations were obtained on a Jasco P1038 polarimeter (Na D line) using a microcell with a 1 dm path length. Specific rotations ([α]_D²⁰, Unit: °cm²/g) are based on the equation α = (100• α) / (l•c) and are reported as unitless numbers where the concentration c is in g/100 mL and the path length l is in decimeters. High Resolution Mass Spectrometry data were obtained at the mass spectrometer facility of the University of Colorado Boulder, Department of Chemistry & Biochemistry on an ESI-qTOF-MS (electrospray-triple quadrupole-time-of-flight mass spectrometer) from Applied Biosystems, PE SCIEX/ABI API QSTAR Pulsar i Hybrid LC/MS/MS. All compounds tested have a purity of ≥95% as determined by TLC, NMR, HRMS, and chiral HPLC for chiral compounds. Compounds were named using ChemBioDraw Ultra 11.0.

General Synthetic Procedures for Racemic Epoxides of Type 2

2-((4-chlorophenoxy)methyl)oxirane (2d)

Into a 50 mL round bottom flask, (4-chloro)phenol (1.00 g, 7.78 mmol, 1 eq) was taken up into acetone (20 mL, 0.4 M) and K₂CO₃ (3.22 g, 23.3 mmol, 3 eq) and epichlorohydrin (2.34 mL, 31.1 mmol, 4 eq) were added consecutively. The reaction mixture was set to stir at reflux for 24 h. At this time an additional 4 eq of epichlorohydrin was added and the solution was allowed to stir at reflux for an additional 24 h.

The reaction was cooled to room temperature and the solids were filtered off. The solvent was removed under reduced pressure and the resulting oil was taken up into toluene (20 mL). The organic layer was washed with H₂O (1 × 20 mL), 1 M aqueous NaOH solution (1 × 20 mL) and H₂O (1 × 30 mL). The organic layer was dried over Na₂SO₄, filtered, and the solvent was removed under reduced pressure. The resulting yellow oil was purified via flash SiO₂ chromatography (4.5 × 5 cm, 10% EtOAc/hexanes) to yield the desired epoxide 2d as a yellow oil (1.19 g, 84%); R_f = 0.22 (10% EtOAc / hexanes); ¹H NMR (400 MHz, CDCl₃) δ 7.25 – 7.19 (m, 2H), 7.25 – 7.19 (m, 2H), 6.87 – 6.81 (m, 2H), 6.86 – 6.81 (m, 2H), 4.20 (dd, J = 2.95, 11.03 Hz, 1H), 3.88 (dd, J = 5.82, 11.03 Hz, 1H), 3.33 (ddt, J = 2.81, 2.81, 4.13, 5.72 Hz, 1H), 2.89 (dd, J = 4.18, 4.86 Hz, 1H), 2.73 (dd, J = 2.66, 4.90 Hz, 1H); ¹³C NMR (101 MHz, CDCl₃) δ 157.20 (C), 129.49 (CH), 126.19 (C), 116.04 (CH), 69.17 (CH₂), 50.17 (CH), 44.70 (CH₂); HRMS (ESI⁺) = calcd C₉H₉ClO₂Li (M+Li⁺) = 191.0445, found = 191.0440.

General Synthetic Procedure for the Pyrazole Fragments of Type 6

1-(3-chlorobenzyl)-3,5-dimethyl-1H-pyrazole (6c)

In a 25 mL round bottom flask with DMSO (14.5 mL), KOH (0.88 g, 15.6 mmol, 1.5 eq) and 3,5-dimethylpyrazole (1.00 g, 10.4 mmol, 1 eq) were combined and set to stir at room temperature. This mixture was then heated at 80 °C for 1 h. The reaction was then cooled to room temperature and 3-chloro-benzylchloride (1.32 mL, 10.4 mmol, 1 eq) was added and the reaction mixture was allowed to stir for an additional 2 h.

Upon completion, the reaction was poured into H₂O (50 mL) and the solution was extracted with CHCl₃ (4 × 50 mL). The resulting organic layers were washed with H₂O (4 × 100 mL) to remove DMSO, dried over Na₂SO₄, filtered, and concentrated under reduced pressure. The resulting material was purified via flash SiO₂ column chromatography (5.5 × 5 cm, 5% EtOAc / hexanes) to yield the desired pyrazole 6c as an yellow oil (2.11 g, 92%); R_f = 0.139 (10% EtOAc / hexanes); ¹H NMR (300 MHz, CDCl₃) δ 7.23 – 7.18 (m, 2H), 7.05 – 7.01 (m, 1H), 6.96 – 6.89 (m, 1H), 5.85 (s, 1H), 5.16 (s, 2H), 2.24 (s, 3H), 2.14 (s, 3H); ¹³C NMR (75 MHz, CDCl₃) δ 148.06 (C), 139.62 (C), 139.34 (C), 134.78 (C), 130.13 (CH), 127.83 (CH), 126.81 (CH), 124.85 (CH), 105.93 (CH), 52.06 (CH₂), 13.69 (CH₃), 11.24 (CH₃); HRMS (ESI⁺) = calcd C₁₂H₁₃ClN₂Na (M+Na⁺) = 243.0659, found = 243.0652.

General Synthetic Procedure for Methyl Amine Fragment of Type 3

1-(1-(3-chlorobenzyl)-3,5-dimethyl-1H-pyrazol-4-yl)-N-methylmethanamine (3c)

In a 25 mL round bottom flask, methylamine-HCl (0.92 g, 13.6 mmol, 3 eq) and paraformaldehyde (0.82 g, 27.2 mmol, 6 eq) were combined in EtOH (9 mL). This solution was stirred for 2 h at 60 °C. At this point, the pyrazole derivative 1-(3-chlorobenzyl)-3,5-dimethyl-1H-pyrazole (6c) (1.00 g, 4.5 mmol, 1 eq) was added and the reaction was heated to 75 °C and stirred for 21 h.

Upon completion, the reaction was cooled to room temperature and the solvent was removed under reduced pressure. The resulting oil was taken up into 50 mL of CHCl₃ and washed with a saturated aqueous solution of NaHCO₃ (1 × 20 mL). The resulting aqueous layer was then extracted with CHCl₃ (3 × 30 mL) and then dried over Na₂SO₄, filtered, and concentrated under reduced pressure. The resulting yellow oil was purified via flash SiO₂ chromatography (5 × 4 cm, 40% EtOAc / hexanes with 2% Et₃N) to yield the desired methyl amine 3c as a yellow oil (0.467 g, 40%); R_f = 0.09 (60% EtOAc / hexanes with 2% Et₃N); ¹H NMR (300 MHz, CDCl₃) δ 7.23 – 7.18 (m, 2H), 7.01 – 6.96 (m, 1H), 6.95 – 6.88 (m, 1H), 5.18 (s, 2H), 3.27 (s, 2H), 2.87 (s, 1H), 2.23 (s, 3H), 2.12 (s, 3H), 2.11 (s, 3H); ¹³C NMR (75 MHz, CDCl₃) δ 147.92 (C), 139.82 (C), 137.98 (C), 134.82 (C), 130.13 (CH), 127.81 (CH), 126.75 (CH), 124.79 (CH), 114.56 (C), 52.21 (CH₂), 49.02 (CH₂), 40.64 (CH₃), 12.28 (CH₃), 9.89 (CH₃); HRMS (ESI⁺) = calcd C₁₄H₁₉ClN₃ (M+H⁺) = 264.1262, found = 264.1263.

Methods for the Synthesis of β hydroxy Amide Compounds of Type 1

Method A: 1-(((1-benzyl-3,5-dimethyl-1H-pyrazol-4-yl)methyl)(methyl)amino)-3- phenoxypropan-2-ol (1a)

To a 10 mL round bottom flask, 3.33 mL of absolute EtOH was added. To this the amine, 1-(1-benzyl-3,5-dimethyl-1H-pyrazol-4-yl)-N-methylmethanamine, (3a) (0.17 g, 0.73 mmol, 1.1 eq), and the commercially available epoxide 2a (0.10 g, 0.66 mmol, 1.0 eq) were combined. The reaction mixture was set to stir for 22 h at reflux.

At which time the reaction was complete, it was cooled to room temperature and the solvent was removed under reduced pressure. The resulting oil was taken up in CHCl₃ (30 mL), washed with a saturated aqueous solution of NaHCO₃ (1 × 20 mL). The resulting aqueous layer was extracted with CHCl₃ (3 × 30 mL). The combined organic layers were then dried over Na₂SO₄, filtered, and the solvent was removed under reduced pressure. The resulting oil was purified via flash SiO₂ column chromatography (3.0 × 4.0 cm, 50% EtOAc / hexanes with 2% Et₃N) to yield the desired alcohol 1a as a clear oil (0.13 g, 50 %); R_f = 0.46 (50% EtOAc / hexanes with 2% Et₃N); ¹H NMR (300 MHz, CDCl₃) δ 7.33 – 7.19 (m, 4H), 7.08 – 7.00 (m, 2H), 6.98 – 6.86 (m, 4H), 5.21 (s, 2H), 4.16 – 4.01 (m, 2H), 3.95 (s, 2H), 3.53 (br s, 1H), 3.37 (ABq, J = 12.90 Hz, Δv = 45.41 Hz, 2H), 2.60 (dd, J = 9.41, 12.24 Hz, 1H), 2.48 (dd, J = 4.77, 11.80 Hz, 1H), 2.25 (s, 1H), 2.23 (s, 1H), 2.11 (s, 1H); ¹³C NMR (75 MHz, CDCl₃) δ 158.96 (C), 147.50 (C), 138.18 (C), 137.54 (C), 129.65 (CH), 128.93 (CH), 127.67 (CH), 126.69 (CH), 121.14 (CH), 114.73 (CH), 113.62 (C), 70.53 (CH₂), 66.29 (CH), 59.58 (CH₂), 53.02 (CH₂), 51.73 (CH₂), 42.05 (CH₃), 12.35 (CH₃), 10.05 (CH₃); HRMS (ESI⁺) = calcd C₂₃H₂₉N₃O₂Na (M+Na⁺) = 402.2152 found = 402.2168.

Method B: 1-(((1-(2-chlorobenzyl)-3,5-dimethyl-1H-pyrazol-4-yl)methyl)(methyl)amino)-3-(4-ethoxyphenoxy)propan-2-ol (1j)

Into a 25 mL round bottom flask, the epoxide, 2-((4-ethyoxyphenoxy)methyl)oxirane), (2b) (0.300 g, 1.54 mmol, 1 eq), amine, 1-(1-(2-chlorobenzyl)-3,5-dimethyl-1H-pyrazol-4-yl)-N-methylmethanamine (3b), (0.45 g, 1.70 mmol, 1.1 eq), and K₂CO₃ (0.23 g, 1.62 mmol, 1 eq) were combined in acetonitrile (5.7 mL, 0.27 M). This solution was heated to reflux for 24 h, at which time, the reaction was complete. The solvent was removed under reduced pressure and the resulting yellow oil was taken up into 10 mL CHCl₃ and partitioned into a saturated aqueous solution of NaHCO₃ (10 mL) and CHCl₃ (10 mL). The layers were separated and the aqueous layer was back extracted with CHCl₃ (4 × 30 mL). The organic layers were combined and dried over Na₂SO₄, filtered and the solvent removed under reduced pressure.

The resulting yellow oil was purified via flash SiO₂ column chromatography (2.5 × 4 cm, 40% EtOAc/hexanes with 2% Et₃N) to yield the desired alcohol (1j) as a white solid (0.612 g, 87%); R_f = 0.300 (60% EtOAc / hexanes with 2% Et₃N); ¹H NMR (300 MHz, CDCl₃) δ 7.38 – 7.33 (m, 1H), 7.22 – 7.09 (m, 2H), 6.87 – 6.78 (m, 4H), 6.51 – 6.47 (m, 1H), 5.30 (s, 2H), 4.12 – 4.03 (m, 1H), 3.98 (q, J = 6.96, 6.98, 6.98 Hz, 2H), 3.90 (d, J = 4.98 Hz, 2H), 3.40 (ABq, J = 13.19, Δv = 47.60 Hz, 2H), 2.61 (dd, J = 9.58, 12.25 Hz, 1H), 2.47 (dd, J = 4.18, 12.24 Hz, 1H), 2.26 (s, 3H), 2.25 (s, 2H), 2.12 (s, 2H), 1.39 (t, J = 6.99 Hz, 3H); ¹³C NMR (75 MHz, CDCl₃) δ 153.52 (C), 153.09 (C), 148.08 (C), 138.70 (C), 135.36 (C), 131.97 (C), 129.47 (CH), 128.82 (CH), 127.63 (CH), 127.53 (CH), 115.68 (CH), 115.58 (CH), 113.66 (C), 71.30 (CH₂), 66.40 (CH), 64.20 (CH₂), 59.57 (CH₂), 51.74 (CH₂), 50.33 (CH₂), 42.12 (CH₃), 15.16 (CH₃), 12.40 (CH₃), 9.86 (CH₃); HRMS (ESI⁺) = calcd C₂₅H₃₃ClN₃O₃ (M+H⁺) = 458.2205, found = 458.2199.

Method C: (S)-1-(((1-(2-chlorobenzyl)-3,5-dimethyl-1H-pyrazol-4-yl)methyl)(methyl)amino)-3-(4-ethoxyphenoxy)propan-2-ol (1l)

Into a 5 mL reaction vial, the epoxide, (S)-2-((4-ethoxyphenoxy)methyl)oxirane (2h), (0.017 g, 0.087 mmol, 1 eq), the amine, 1-(1-(2-chlorobenzyl)-3,5-dimethyl-1H-pyrazol-4-yl)-N-methylmethanamine (3b), (0.023 g, 0.087 mmol, 1.1 eq) were taken up into toluene (0.3 mL, 0.3 M) and set to reflux for 48 h.

The reaction mixture was cooled to room temperature, transferred to a 25 mL round bottom flask, and the solvent was removed under reduced pressure. The resulting orange oil was purified via flash SiO₂ column chromatography (1.5 × 4 cm, 60% EtOAc / hexanes with 2% Et₃N) to yield the desired chiral alcohol (1l) as a clear oil (0.025 g, 63%); R_f = 0.210 (50% EtOAc / hexanes with 2% Et₃N); Assay of enantiomeric excess: HPLC (Chiralcel AD 25 cm column, 20% i-PrOH / hexanes with 0.1% Et₂NH; 1.0 mL / min) t_r (major) = 4.56 min, t_r (minor) = 5.7 min; >96% ee; [α]_D²⁵ = −15.06 (c=0.649, CHCl₃); ¹H NMR (400 MHz, CDCl₃) δ 7.35 (dd, J = 1.42, 7.79 Hz, 1H), 7.15 (dtd, J = 1.55, 7.43, 7.45, 20.61 Hz, 1H), 6.85 – 6.79 (m, 2H), 6.51 – 6.43 (m, 1H), 5.30 (s, 1H), 4.08 (dddd, J = 5.33, 5.33, 5.33, 9.28 Hz, 1H), 3.97 (q, J = 6.97, 6.98, 6.98 Hz, 1H), 3.90 (d, J = 4.96 Hz, 1H), 3.42 (s, 1H), 3.40 (ABq, J = 13.19 Hz, Δv = 63.80 Hz, 1H), 2.61 (dd, J = 9.70, 12.18 Hz, 1H), 2.47 (dd, J = 4.13, 12.24 Hz, 1H), 2.26 (s, 1H), 2.25 (s, 1H), 2.12 (s, 1H); ¹³C NMR (101 MHz, CDCl₃) δ 153.45 (C), 153.03 (C), 148.05 (C), 138.68 (C), 135.32 (C), 131.91 (C), 129.43 (CH), 128.79 (CH), 127.55 (CH), 127.51 (CH), 115.61 (CH), 115.50 (CH), 113.70 (C), 71.22 (CH₂), 66.33 (CH), 64.13(CH₂), 59.50 (CH₂), 51.69 (CH₂), 50.30 (CH₂), 42.07 (CH₃), 15.15 (CH₃), 12.39 (CH₃), 9.84 (CH₃); HRMS (ESI⁺) = calcd C₂₅H₃₃ClN₂O₃ (M+H⁺) = 458.2205, found = 458.2191.

Method D: 1-(4-(allyloxy)phenoxy)-3-(((3,5-dimethyl-1-(2-vinylbenzyl)-1H-pyrazol-4-yl)methyl)(methyl)amino)propan-2-ol (1q)

Into a 5 mL Biotage Microwave tube, the epoxide 2-((4-(allyloxy)phenoxy)methyl)oxirane (2c) (0.038 g, 0.182 mmol, 1 eq) and the amine 1-(3,5-dimethyl-1-(2-vinylbenzyl)-1H-pyrazol-4-yl)-N-methylmethanamine (3f) (0.051 g, 0.20 mmol, 1.1 eq) were combined in absolute ethanol (0.455 mL). The tube was then sealed with the appropriate crimp cap and the tube placed into the microwave. After a total of 15 min (200 watts, 140 °C, 4–5 bar) the reaction was compete.

The reaction mixture was transferred to a round bottom with ethyl acetate and the solvent was removed under reduced pressure. The resulting oil was diluted with CHCl₃ (20 mL) and washed with a saturated aqueous solution of NaHCO₃ (1 × 10 mL). The aqueous layer was extracted with CHCl₃ (2 × 20 mL) and the combined organic layers were dried over Na₂SO₄, filtered, and the solvent was removed under reduced pressure. The resulting material was purified via flash SiO₂ column chromatography (1.5 × 6.5 cm, 60% EtOAc / hexanes with 2% Et₃N) to yield the desired alcohol 1q (0.072 g, 86%); R_f = 0.25 (60% EtOAc / hexanes with 2% Et₃N); ¹H NMR (300 MHz, CDCl₃) δ 7.46 (dd, J = 1.46, 7.61 Hz, 1H), 7.25 – 7.10 (m, 2H), 6.96 (dd, J = 10.95, 17.29 Hz, 1H), 6.83 (s, 4H), 6.47 (dd, J = 0.88, 7.59 Hz, 1H), 6.12 – 5.97 (m, 1H), 5.65 (dd, J = 1.37, 17.27 Hz, 1H), 5.39 (d, J = 1.33 Hz, 1H), 5.37 – 5.35 (m, 1H), 5.33 (dq, J = 1.42, 1.53, 1.53, 51.83 Hz, 1H), 5.30 – 5.27 (m, 2H), 4.48 (dt, J = 1.52, 1.52, 5.31 Hz, 2H), 4.14 – 4.02 (m, 1H), 3.90 (d, J = 5.03 Hz, 2H), 3.39 (ABq, J = 13.20 Hz, Δv = 47.40 Hz, 2H), 3.27 (br s, 1H), 2.60 (dd, J = 9.53, 12.19 Hz, 1H), 2.47 (dd, J = 4.21, 12.25 Hz, 1H), 2.25 (s, 2H), 2.24 (s, 2H), 2.07 (s, 2H); ¹³C NMR (75 MHz, CDCl₃) δ 153.29 (C), 153.14 (C), 147.68 (C), 138.56 (C), 135.92 (C), 134.59 (C), 133.76 (CH), 133.74 (CH), 128.44 (CH), 127.73 (CH), 126.33 (CH), 126.30 (CH), 117.71 (CH₂), 117.38 (CH₂), 115.86 (CH), 115.63 (CH), 113.60 (C), 71.28 (CH₂), 69.67 (CH₂), 66.36 (CH), 59.54 (CH₂), 51.72 (CH₂), 50.70 (CH₂), 42.06 (CH₃), 12.38 (CH₃), 9.97 (CH₃); HRMS (ESI⁺) = calcd C₂₈H₃₆N₃O₃ (M+H⁺) = 462.2751 found = 462.2750.

Method E: 1-(((1-(2-chlorobenzyl)-3,5-dimethyl-1H-pyrazol-4-yl)methyl)(methyl)amino-3-(4-Chlorophenoxy)propan-2-ol (1s)

Into a 5 mL Biotage Microwave tube the epoxide, 2-((4-chlorophenoxy)methyl)oxirane (2d, (0.05 g, 0.27 mmol, 1 eq), the amine, 1-(1-(2-chlorobenzyl)-3,5-dimethyl-1H-pyrazol-4-yl)-N-methylmethanamine (3b), (0.078 g, 0.300 mmol, 1.1 eq), and K₂CO₃ (0.036 g, 0.271 mmol, 1 eq) were all taken up in acetonitrile (1 mL, 0.27 M). The tube was then sealed with the appropriate crimp cap and the tube placed into the microwave. After a total of 50 min (200 watts, 180 °C, 9 bar) the reaction was compete. The reaction tube was cooled to room temperature, cap was removed, and the reaction mixture was transferred to a 15 mL round bottom flask with EtOAc. The solvent was removed under reduced pressure and the resulting oil was diluted with CHCl₃ (10 mL). This solution was partitioned into a separatory funnel containing a 1:1 solution of CHCl₃ and a saturated aqueous solution of NaHCO₃ (10 mL each). The layers were separated and the organic layer was extracted with CHCl₃ (2 × 10 mL). The combined organic layers were dried over Na₂SO₄, filtered, and the solvent was removed under reduced pressure to yield an orange oil.

This oil was purified via flash SiO₂ column chromatography (2.5 × 4 cm, 60% EtOAc / hexanes with 2% Et₃N) to yield the desired alcohol (1s) as a yellow oil (0.121 g, quant.); R_f = 0.25 (60% EtOAc / hexanes with 2% Et₃N); ¹H NMR (300 MHz, CDCl₃) δ 7.33 (dd, J = 1.44, 7.71 Hz, 1H), 7.23 – 7.16 (m, 2H), 7.17 – 7.06 (m, 2H), 6.85 – 6.74 (m, 2H), 6.48 (dd, J = 1.60, 7.51 Hz, 1H), 5.28 (s, 2H), 4.18 – 3.96 (m, 1H), 3.97 – 3.79 (m, 2H), 3.38 (ABq, J = 13.18 Hz, Δv = 45.8 Hz, 3H), 2.58 (dd, J = 9.54, 12.17 Hz, 1H), 2.45 (dd, J = 4.29, 12.23 Hz, 1H), 2.24 (s, 3H), 2.24 (s, 3H), 2.10 (s, 3H); ¹³C NMR (101 MHz, CDCl₃) δ 157.55 (C), 148.02 (C), 138.66 (C), 135.29 (C), 132.07 (C), 129.67 (CH), 128.82 (CH), 127.56 (CH), 115.97 (C), 113.65 (C), 70.84 (CH₂), 66.15 (CH), 59.26 (CH₂), 51.70 (CH₂), 50.59 (CH₂), 42.07 (CH₃), 12.39 (CH₃), 9.74 (CH₃); HRMS (ESI+) = calcd C₂₃H₂₈Cl₂N₃O₂Na (M+H⁺) = 448.1538 found = 448.1553.

N-((1-(2-chlorobenzyl)-3,5-dimethyl-1H-pyrazol-4-yl)methyl)-3-(4-ethoxyphenoxy)-2-methoxy-N-methylpropan-1-amine (7)

Into a 10 mL round bottom flask, the alcohol, 1-(((1-(2-chlorobenzyl)-3,5-dimethyl-1H-pyrazol-4-yl)methyl)(methyl)amino)-3-(4-ethoxyphenoxy)propan-2-ol (1j), (0.020 g, 0.044 mmol, 1 eq) was taken up in THF (2.2 mL). To this, NaH (60%, 0.023 g, 0.96 mmol, 22 eq) was added and the mixture was stirred for 10 min. MeI (0.08 mL, 1.3 mmol, 30 eq) was added dropwise. The solution was allowed to stir at room temperature for 30 min, at which time, it appeared complete by TLC. The reaction mixture was quenched by the slow addition of 2 mL of a saturated aqueous solution of NH₄Cl. Once the effervescence ceased (about 1 hr) the mixture was diluted with ether and the layers were separated. The aqueous layer was back extracted with ether (1 × 20 mL) than CH₂Cl₂ 2 × 20 mL). The combined organic layers were dried over Na₂SO₄, filtered, and the solvent was removed under reduced pressure. The resulting dark brown oil was purified via flash SiO₂ column chromatography (2 × 3 cm, dry loaded on SiO₂, 40% EtOAc / hexanes with 2% Et₃N) to yield the desired methy ether (7) as a clear oil (0.0078 g, 37%); R_f = 0.212 (60% EtOAc / hexanes with 2% Et₃N); ¹H NMR (400 MHz, CDCl₃) δ 7.34 (dd, J = 1.25, 7.87 Hz, 1H), 7.17 (td, J = 1.77, 7.66, 7.87 Hz, 1H), 7.10 (td, J = 1.38, 7.45, 7.53 Hz, 1H), 6.83 – 6.76 (m, 4H), 6.47 (d, J = 6.15 Hz, 1H), 5.25 (s, 2H), 4.03 (dd, J = 3.50, 9.98 Hz, 1H), 3.97 (q, J = 6.97, 6.98, 6.98 Hz, 2H), 3.87 (dd, J = 5.51, 10.00 Hz, 1H), 3.48 (s, 3H), 3.39 – 3.25 (m, 2H), 2.64 – 2.56 (m, 1H), 2.53 – 2.45 (m, 1H), 2.24 (s, 3H), 2.23 (s, 3H), 2.07 (s, 3H), 1.38 (t, J = 6.99 Hz, 3H); ¹³C NMR (101 MHz, CDCl₃) δ 153.34, 153.15, 148.16, 135.47, 131.91, 129.42, 128.75, 127.58, 127.47, 115.56, 115.49, 69.51, 64.15, 57.99, 57.40, 52.32, 50.21, 43.19, 29.92, 15.17, 12.33, 9.76; HRMS (ESI⁺) = calcd C₂₆H₃₅ClN₃O₃ (M+H⁺) = 472.2382, found = 472.2361.

Macrocycle 8

To a 50 mL round bottom flask, the β-amino alcohol 1q (0.05g, 0.11 mmol, 1 eq) was dissolved in toluene (21 mL). The solution was degassed and Grubbs II catalyst (0.0009g, 0.011 mmol, 0.1 eq) was added and reaction mixture was degassed again. This mixture was stirred for 7 h at 110 °C. At this time, the reaction was cooled to room temperature and the solvent was removed under reduced pressure. The dark brown oil was purified via flash SiO₂ column chromatography (1.5 × 7.0 cm, 60% EtOAc / hexanes with 2% Et₃N) to yield 8 as an yellow oil (0.024 g, 52%); R_f = 0.38 (60% EtOAc / hexanes with 2% Et₃N); ¹H NMR (300 MHz, CDCl₃) δ 7.20 – 7.08 (m, 4H), 6.96 – 6.88 (m, 2H), 6.84 (d, J = 6.65 Hz, 1H), 6.74 (d, J = 5.27 Hz, 1H), 6.74 – 6.69 (m, 2H), 5.77 (dt, J = 5.83, 5.83, 16.01 Hz, 1H), 4.86 (d, J = 5.66 Hz, 2H), 4.62 (ABq, J = 16.50 Hz, Δv = 61.90 Hz, 2H), 3.98 (d, J = 7.24 Hz, 1H), 3.71 – 3.58 (m, 2H), 3.38 (d, J = 13.29 Hz, 1H), 3.34 – 3.19 (m, 2H), 2.56 (dd, J = 6.61, 12.46 Hz, 1H), 2.45 (s, 3H), 2.22 (s, 3H), 1.34 (s, 3H); ¹³C NMR (75 MHz, CDCl₃) δ 155.82, 153.37, 151.17, 147.10, 135.73, 134.78, 133.47, 130.00, 128.32, 128.23, 127.45, 127.08, 126.93, 119.18, 115.18, 69.31, 68.63, 66.64, 56.95, 52.03, 50.58, 12.13, 9.49; HRMS (ESI⁺) = calcd C₂₆H₃₂N₃O₃ (M+H⁺) = 434.2438, found = 434.2423.

Formation of β-hydroxy Amine HCl Salts

If required, the free bases of the following compounds were made into the HCl salts via the following procedure. The compound was dissolved into 0.5 M Et₂O. To this solution, an equal volume of 2 M HCl in Et₂O was added dropwise until the HCl salt precipitated out of solution. This mixture was allowed to stir for an additional 10 min then the solvent was evaporated off using a stream of N₂ gas. The HCl salt was then dried on high vacuum overnight and stored at room temperature over Dryrite^®.

Abbreviation List

AKT1

RAC-alpha serine/threonine-protein kinase

BBB

blood brain barrier

CAMK1

calcium/calmodulin-dependent protein kinase type 1

CSF Buffer

cerebrospinal fluid buffer

CD14

cluster of differentiation 14

CYP 450

cytochrome P450

DDR2

discoidin domain-containing receptor 2

DMEM

Dulbecco's modified eagle medium

DMSO

dimethyl sulfoxide

FBS

fetal bovine serum

GSK3α

glycogen synthase kinase-3 alpha

HEK293

human embryonic kidney 293 cells

IL-1

interleukin 1

IL-1β

interleukin 1-beta

IL-6

interleukin 6

LPS

lipopolysaccharide

MARK1

MAP/microtubule affinity-regulating kinase 1

MeCN

acetonitrile

MD2

myeloid differentiation factor 2

MW

microwave

NADPH

nicotinamide adenine dinucleotide phosphate

NF-κB

nuclear factor κB

NMR

nuclear magnetic resonance

NO

nitric oxide

PAPS

3'-phosphoadenosine-5'-phosphosulfate

GSH

glutathione

PAK1

P21/Cdc42/Rac1-activated kinase 1

PDGFβ

platelet-derived growth factor-β

PBS

phosphate buffered saline

PIM1

proto-oncogene serine/threonine-protein kinase

PKC-γ

protein kinase γ

PLK4

polo-like kinase 4

PPTS

pyridinium p-toluenesulfonate

RPMI medium

Roswell Park Memorial Institute medium

SRC

sarcoma kinase

TLC

thin layer chromatography

TLR-4

toll-like receptor 4

TNF-α

tumor necrosis factor-α

UDPGA

uridine 5'-diphospho-glucuronosyltransferase