Search

Macrocyclic retinoic acid receptor-related orphan receptor C2 (RORC2) inverse agonists have been designed with favorable properties for topical administration. Inspired by the unanticipated bound conformation of an acyclic sulfonamide-based RORC2 ligand from cocrystal structure analysis, macrocyclic linker connections between the halves of the molecule were explored. Further optimization of analogues was accomplished to maximize potency and refine physiochemical properties (MW, lipophilicity) best suited for topical application. Compound 14 demonstrated potent inhibition of interleukin-17A (IL-17A) production by human Th17 cells and in vitro permeation through healthy human skin achieving high total compound concentration in both skin epidermis and dermis layers., Competing Interests: The authors declare no competing financial interest., (© 2023 American Chemical Society.)

Fast and accurate assessment of small-molecule dihedral energetics is crucial for molecular design and optimization in medicinal chemistry. Yet, accurate prediction of torsion energy profiles remains challenging as the current molecular mechanics (MM) methods are limited by insufficient coverage of drug-like chemical space and accurate quantum mechanical (QM) methods are too expensive. To address this limitation, we introduce TorsionNet, a deep neural network (DNN) model specifically developed to predict small-molecule torsion energy profiles with QM-level accuracy. We applied active learning to identify nearly 50k fragments (with elements H, C, N, O, F, S, and Cl) that maximized the coverage of our corporate compound library and leveraged massively parallel cloud computing resources for density functional theory (DFT) torsion scans of these fragments, generating a training data set of 1.2 million DFT energies. After training TorsionNet on this data set, we obtain a model that can rapidly predict the torsion energy profile of typical drug-like fragments with DFT-level accuracy. Importantly, our method also provides an uncertainty estimate for the predicted profiles without any additional calculations. In this report, we show that TorsionNet can accurately identify the preferred dihedral geometries observed in crystal structures. Our TorsionNet-based analysis of a diverse set of protein-ligand complexes with measured binding affinity shows a strong association between high ligand strain and low potency. We also present practical applications of TorsionNet that demonstrate how consideration of DNN-based strain energy leads to substantial improvement in existing lead discovery and design workflows. TorsionNet500, a benchmark data set comprising 500 chemically diverse fragments with DFT torsion profiles (12k MM- and DFT-optimized geometries and energies), has been created and is made publicly available.

The energetics of rotation around single bonds (torsions) is a key determinant of the three-dimensional shape that druglike molecules adopt in solution, the solid state, and in different biological environments, which in turn defines their unique physical and pharmacological properties. Therefore, accurate characterization of torsion angle preference and energetics is essential for the success of computational drug discovery and design. Here, we analyze torsional strain in crystal structures of druglike molecules in Cambridge structure database (CSD) and bioactive ligand conformations in protein data bank (PDB), expressing the total strain energy as a sum of strain energy from constituent rotatable bonds. We utilized cloud computing to generate torsion scan profiles of a very large collection of chemically diverse neutral fragments at DFT(B3LYP)/6-31G*//6-31G** or DFT(B3LYP)/6-31+G*//6-31+G** (for sulfur-containing molecule). With the data generated from these ab initio calculations, we performed rigorous analysis of strain due to deviation of observed torsion angles relative to their ideal gas-phase geometries. Contrary to the previous studies based on molecular mechanics, we find that in the crystalline state, molecules generally adopt low-strain conformations, with median per-torsion strain energy in CSD and PDB under one-tenth and one-third of a kcal/mol, respectively. However, for a small fraction (<5%) of motifs, external effects such as steric hindrance and hydrogen bonds result in strain penalty exceeding 2.5 kcal/mol. We find that due to poor quality of PDB structures in general, bioactive structures tend to have higher torsional strain compared to small-molecule crystal conformations. However, in the absence of structural fitting artifacts in PDB structures, protein-induced strain in bioactive conformations is quantitatively similar to those due to the packing forces in small-molecule crystal structures. This analysis allows us to establish strain energy thresholds to help identify biologically relevant conformers in a given ensemble. The work presented here is the most comprehensive study to date that demonstrates the utility and feasibility of gas-phase quantum mechanics (QM) calculations to study conformational preference and energetics of drug-size molecules. Potential applications of this study in computational lead discovery and structure-based design are discussed.

Ongoing interest in the discovery of selective JAK3 inhibitors led us to design novel covalent inhibitors that engage the JAK3 residue Cys909 by cyanamide, a structurally and mechanistically differentiated electrophile from other cysteine reacting groups previously incorporated in JAK3 covalent inhibitors. Through crystallography, kinetic, and computational studies, interaction of cyanamide 12 with Cys909 was optimized leading to potent and selective JAK3 inhibitors as exemplified by 32. In relevant cell-based assays and in agreement with previous results from this group, 32 demonstrated that selective inhibition of JAK3 is sufficient to drive JAK1/JAK3-mediated cellular responses. The contribution from extrahepatic processes to the clearance of cyanamide-based covalent inhibitors was also characterized using metabolic and pharmacokinetic data for 12. This work also gave key insights into a productive approach to decrease glutathione/glutathione S-transferase-mediated clearance, a challenge typically encountered during the discovery of covalent kinase inhibitors.

The nuclear hormone receptor retinoic acid receptor-related orphan C2 (RORC2, also known as RORγt) is a promising target for the treatment of autoimmune diseases. A small molecule, inverse agonist of the receptor is anticipated to reduce production of IL-17, a key proinflammatory cytokine. Through a high-throughput screening approach, we identified a molecule displaying promising binding affinity for RORC2, inhibition of IL-17 production in Th17 cells, and selectivity against the related RORA and RORB receptor isoforms. Lead optimization to improve the potency and metabolic stability of this hit focused on two key design strategies, namely, iterative optimization driven by increasing lipophilic efficiency and structure-guided conformational restriction to achieve optimal ground state energetics and maximize receptor residence time. This approach successfully identified 3-cyano- N-(3-(1-isobutyrylpiperidin-4-yl)-1-methyl-4-(trifluoromethyl)-1 H-pyrrolo[2,3- b]pyridin-5-yl)benzamide as a potent and selective RORC2 inverse agonist, demonstrating good metabolic stability, oral bioavailability, and the ability to reduce IL-17 levels and skin inflammation in a preclinical in vivo animal model upon oral administration.

Janus kinases (JAKs) are intracellular tyrosine kinases that mediate the signaling of numerous cytokines and growth factors involved in the regulation of immunity, inflammation, and hematopoiesis. As JAK1 pairs with JAK2, JAK3, and TYK2, a JAK1-selective inhibitor would be expected to inhibit many cytokines involved in inflammation and immune function while avoiding inhibition of the JAK2 homodimer regulating erythropoietin and thrombopoietin signaling. Our efforts began with tofacitinib, an oral JAK inhibitor approved for the treatment of rheumatoid arthritis. Through modification of the 3-aminopiperidine linker in tofacitinib, we discovered highly selective JAK1 inhibitors with nanomolar potency in a human whole blood assay. Improvements in JAK1 potency and selectivity were achieved via structural modifications suggested by X-ray crystallographic analysis. After demonstrating efficacy in a rat adjuvant-induced arthritis (rAIA) model, PF-04965842 (25) was nominated as a clinical candidate for the treatment of JAK1-mediated autoimmune diseases.

Bruton tyrosine kinase (BTK) is a key enzyme in B-cell development whose improper regulation causes severe immunodeficiency diseases. Design of selective BTK therapeutics would benefit from improved, in-silico structural modeling of the kinase's solution ensemble. However, this remains challenging due to the immense computational cost of sampling events on biological timescales. In this work, we combine multi-millisecond molecular dynamics (MD) simulations with Markov state models (MSMs) to report on the thermodynamics, kinetics, and accessible states of BTK's kinase domain. Our conformational landscape links the active state to several inactive states, connected via a structurally diverse intermediate. Our calculations predict a kinome-wide conformational plasticity, and indicate the presence of several new potentially druggable BTK states. We further find that the population of these states and the kinetics of their inter-conversion are modulated by protonation of an aspartate residue, establishing the power of MD & MSMs in predicting effects of chemical perturbations.

In an effort to find new and safer treatments for osteoporosis and frailty, we describe a novel series of selective androgen receptor modulators (SARMs). Using a structure-based approach, we identified compound 7, a potent AR (ARE EC 50 = 0.34 nM) and selective (N/C interaction EC 50 = 1206 nM) modulator. In vivo data, an AR LBD X-ray structure of 7, and further insights from modeling studies of ligand receptor interactions are also presented.

Significant work has been dedicated to the discovery of JAK kinase inhibitors resulting in several compounds entering clinical development and two FDA approved NMEs. However, despite significant effort during the past 2 decades, identification of highly selective JAK3 inhibitors has eluded the scientific community. A significant effort within our research organization has resulted in the identification of the first orally active JAK3 specific inhibitor, which achieves JAK isoform specificity through covalent interaction with a unique JAK3 residue Cys-909. The relatively rapid resynthesis rate of the JAK3 enzyme presented a unique challenge in the design of covalent inhibitors with appropriate pharmacodynamics properties coupled with limited unwanted off-target reactivity. This effort resulted in the identification of 11 (PF-06651600), a potent and low clearance compound with demonstrated in vivo efficacy. The favorable efficacy and safety profile of this JAK3-specific inhibitor 11 led to its evaluation in several human clinical studies.

PF-06651600, a newly discovered potent JAK3-selective inhibitor, is highly efficacious at inhibiting γc cytokine signaling, which is dependent on both JAK1 and JAK3. PF-06651600 allowed the comparison of JAK3-selective inhibition to pan-JAK or JAK1-selective inhibition, in relevant immune cells to a level that could not be achieved previously without such potency and selectivity. In vitro, PF-06651600 inhibits Th1 and Th17 cell differentiation and function, and in vivo it reduces disease pathology in rat adjuvant-induced arthritis as well as in mouse experimental autoimmune encephalomyelitis models. Importantly, by sparing JAK1 function, PF-06651600 selectively targets γc cytokine pathways while preserving JAK1-dependent anti-inflammatory signaling such as the IL-10 suppressive functions following LPS treatment in macrophages and the suppression of TNFα and IL-1β production in IL-27-primed macrophages. Thus, JAK3-selective inhibition differentiates from pan-JAK or JAK1 inhibition in various immune cellular responses, which could potentially translate to advantageous clinical outcomes in inflammatory and autoimmune diseases.

Anisole and fluoroanisoles display distinct conformational preferences, as evident from a survey of their crystal structures. In addition to altering the free ligand conformation, various degrees of fluorination have a strong impact on physicochemical and pharmacokinetic properties. Analysis of anisole and fluoroanisole matched molecular pairs in the Pfizer corporate database reveals interesting trends: 1) PhOCF3 increases log D by ~1 log unit over PhOCH3 compounds; 2) PhOCF3 shows lower passive permeability despite its higher lipophilicity; and 3) PhOCF3 does not appreciably improve metabolic stability over PhOCH3 . Emerging from the investigation, difluoroanisole (PhOCF2 H) strikes a better balance of properties with noticeable advantages of log D and transcellular permeability over PhOCF3 . Synthetic assessment illustrates that the routes to access difluoroanisoles are often more straightforward than those for trifluoroanisoles. Whereas replacing PhOCH3 with PhOCF3 is a common tactic to optimize ADME properties, our analysis suggests PhOCF2 H may be a more attractive alternative, and greater exploitation of this motif is recommended., (© 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim.)

ITK (interleukin-2-inducible T-cell kinase) is a critical component of signal transduction in T-cells and has a well-validated role in their proliferation, cytokine release and chemotaxis. ITK is an attractive target for the treatment of T-cell-mediated inflammatory diseases. In the present study we describe the discovery of kinase inhibitors that preferentially bind to an allosteric pocket of ITK. The novel ITK allosteric site was characterized by NMR, surface plasmon resonance, isothermal titration calorimetry, enzymology and X-ray crystallography. Initial screening hits bound to both the allosteric pocket and the ATP site. Successful lead optimization was achieved by improving the contribution of the allosteric component to the overall inhibition. NMR competition experiments demonstrated that the dual-site binders showed higher affinity for the allosteric site compared with the ATP site. Moreover, an optimized inhibitor displayed non-competitive inhibition with respect to ATP as shown by steady-state enzyme kinetics. The activity of the isolated kinase domain and auto-activation of the full-length enzyme were inhibited with similar potency. However, inhibition of the activated full-length enzyme was weaker, presumably because the allosteric site is altered when ITK becomes activated. An optimized lead showed exquisite kinome selectivity and is efficacious in human whole blood and proximal cell-based assays.

We present a novel series of selective androgen receptor modulators (SARMs) which shows excellent biological activity and physical properties. 1-(2-Hydroxy-2-methyl-3-phenoxypropanoyl)-indoline-4-carbonitriles showed potent binding to the androgen receptor (AR) and activated AR-mediated transcription in vitro. Representative compounds demonstrated diminished activity in promoting the intramolecular interaction between the AR carboxyl (C) and amino (N) termini. This N/C-termini interaction is a biomarker assay for the undesired androgenic responses in vivo. In orchidectomized rats, daily administration of a lead compound from this series showed anabolic activity by increasing levator ani muscle weight. Importantly, minimal androgenic effects (increased tissue weights) were observed in the prostate and seminal vesicles, along with minimal repression of circulating luteinizing hormone (LH) levels and no change in the lipid and triglyceride levels. This lead compound completed a two week rat toxicology study, and was well tolerated at doses up to 100 mg/kg/day, the highest dose tested, for 14 consecutive days.

Due to the high attrition rate of central nervous system drug candidates during clinical trials, the assessment of blood-brain barrier (BBB) penetration in early research is particularly important. A genetic approximation (GA)-based regression model was developed for predicting in vivo blood-brain partitioning data, expressed as logBB (log[brain]/[blood]). The model was built using an in-house data set of 193 compounds assembled from 22 different therapeutic projects. The final model (cross-validated r(2) = 0.72) with five molecular descriptors was selected based on validation using several large internal and external test sets. We demonstrate the potential utility of the model by applying it to a set of literature reported secretase inhibitors. In addition, we describe a rule-based approach for rapid assessment of brain penetration with several simple molecular descriptors.

A series of 4-(3-biaryl)quinolines with sulfone substituents on the terminal aryl ring (8) was prepared as potential LXR agonists. High affinity LXRbeta ligands with generally modest binding selectivity over LXRalpha and excellent agonist potency in LXR functional assays were identified. Many compounds had LXRbeta binding IC(50) values <10 nM while the most potent had EC(50) values <1.0 nM in an ABCA1 mRNA induction assay in J774 mouse cells with efficacy comparable to T0901317. Sulfone 8a was further evaluated in LDL (-/-) mice and shown to reduce atherosclerotic lesion progression., (2010 Elsevier Ltd. All rights reserved.)

A series of quinoline-3-carboxamide containing sulfones was prepared and found to have good binding affinity for LXRbeta and moderate binding selectivity over LXRalpha. The 8-Cl quinoline analog 33 with a high TPSA score, displayed 34-fold binding selectivity for LXRbeta over LXRalpha (LXRbeta IC(50)=16nM), good activity for inducing ABCA1 gene expression in a THP macrophage cell line, desired weak potency in the LXRalpha Gal4 functional assay, and low blood-brain barrier penetration in rat., (Copyright 2009 Elsevier Ltd. All rights reserved.)

A series of (hetero)arylpyrimidines agonists of the Wnt-beta-catenin cellular messaging system have been prepared. These compounds show activity in U2OS cells transfected with Wnt-3a, TCF-luciferase, Dkk-1 and tk-Renilla. Selected compounds show minimal GSK-3beta inhibition indicating that the Wnt-beta-catenin agonism activity most likely comes from interaction at Wnt-3a/Dkk-1. Two examples 1 and 25 show in vivo osteogenic activity in a mouse calvaria model. One example 1 is shown to activate non-phosphorylated beta-catenin formation in bone., (Copyright 2009 Elsevier Ltd. All rights reserved.)

A series of 4-(3-aryloxyaryl)quinolines with sulfone substituents on the terminal aryl ring (7) was prepared as LXR agonists. High affinity LXR ligands with excellent agonist potency and efficacy in functional assays of LXR activity were identified. In general, these sulfone agonists were equal to or superior to previously described alcohol and amide analogs in terms of affinity, functional potency, and microsomal stability. Many of the sulfones had LXRbeta binding IC(50) values <10nM while the most potent compounds in an ABCA1 mRNA induction assay in J774 mouse cells had EC(50) values <10nM and were as efficacious as T0901317., (Copyright 2009 Elsevier Ltd. All rights reserved.)

High throughput microsomal stability assays have been widely implemented in drug discovery and many companies have accumulated experimental measurements for thousands of compounds. Such datasets have been used to develop in silico models to predict metabolic stability and guide the selection of promising candidates for synthesis. This approach has proven most effective when selecting compounds from proposed virtual libraries prior to synthesis. However, these models are not easily interpretable at the structural level, and thus provide little insight to guide traditional synthetic efforts. We have developed global classification models of rat, mouse and human liver microsomal stability using in-house data. These models were built with FCFP&#95;6 fingerprints using a Naïve Bayesian classifier within Pipeline Pilot. The test sets were correctly classified as stable or unstable with satisfying accuracies of 78, 77 and 75% for rat, human and mouse models, respectively. The prediction confidence was assigned using the Bayesian score to assess the applicability of the models. Using the resulting models, we developed a novel data mining strategy to identify structural features associated with good and bad microsomal stability. We also used this approach to identify structural features which are good for one species but bad for another. With these findings, the structure-metabolism relationships are likely to be understood faster and earlier in drug discovery.

A series of 4-(3-aryloxyaryl)quinolines with alcohol substituents on the terminal aryl ring was prepared as potential LXR agonists, in which an alcohol group replaced an amide in previously reported amide analogs. High affinity LXR ligands with excellent agonist potency and efficacy in a functional model of LXR activity were identified, demonstrating that alcohols can substitute for amides while retaining LXR activity. The most potent compound was 5b which had an IC(50)=3.3 nM for LXRbeta binding and EC(50)=12 nM (122% efficacy relative to T0901317) in an ABCA1 mRNA induction assay in J774 mouse cells.

A series of cinnolines/quinolines was prepared and it was found that 4-phenyl-cinnoline/quinolines with either a 2',3' or 2',5'-disubstituted benzyloxy moiety or the 1-Me-7-indole methoxy moiety on the meta position of the 4-phenyl ring showed good binding selectivity for LXRbeta over LXRalpha. The LXRbeta binding selective modulators displayed good activity for inducing ABCA1 gene expression in J774 macrophage cell line and poor efficacy in the LXRalpha Gal4 functional assay. 26, 37 and 41 were examined for their ability to induce SREBP-1c gene expression in Huh-7 liver cell line and they were weak partial agonists.

Rotigaptide (3) is an antiarrhythmic peptide that improves cardiac conduction by modifying gap-junction communication. Small molecule gap-junction modifiers with improved physical properties were identified from a Zealand Pharma peptide library using pharmaceutical profiling, established SAR around 3, and a putative pharmacophore model for rotigaptide. Activity of the compounds was confirmed in a mouse cardiac conduction block model of arrhythmia. Dipeptide 9f (GAP-134) was identified as a potent, orally active gap-junction modifier for clinical development.

A series of 4-(amido-biarylether)-quinolines was prepared as potential LXR agonists. Appropriate substitution with amide groups provided high affinity LXR ligands, some with excellent potency and efficacy in functional assays of LXR activity. Novel amide 4g had a binding IC(50)=1.9 nM for LXRbeta and EC(50)=34 nM (96% efficacy relative to T0901317) in an ABCA1 gene expression assay in mouse J774 cells, demonstrating that 4-(biarylether)-quinolines with appropriate amide substitution are potent LXR agonists.

A series of substituted 2-benzyl-3-aryl-7-trifluoromethylindazoles were prepared as LXR modulators. These compounds were partial agonists in transactivation assays when compared to 1 (T0901317) and were slightly weaker with respect to potency and efficacy on LXRalpha than on LXRbeta. Lead compounds in this series 12 (WAY-252623) and 13 (WAY-214950) showed less lipid accumulation in HepG2 cells than potent full agonists 1 and 3 (WAY-254011) but were comparable in efficacy to 1 and 3 with respect to cholesterol efflux in THP-1 foam cells, albeit weaker in potency. Compound 13 reduced aortic lesion area in LDLR knockout mice equivalently to 3 or positive control 2 (GW3965). In a 7-day hamster model, compound 13 showed a lesser propensity for plasma TG elevation than 3, when the compounds were compared at doses in which they elevated ABCA1 and ABCG1 gene expression in duodenum and liver at equal levels. In contrast to results previously published for 2, the lack of TG effect of 13 correlated with its inability to increase liver fatty acid synthase (FAS) gene expression, which was up-regulated 4-fold by 3. These results suggest indazoles such as 13 may have an improved profile for potential use as a therapeutic agent.

A series of novel 7-(5'-cyanopyrrol-2-yl) substituted benzo[1,4]oxazepin-2-ones were prepared and tested for their progesterone receptor (PR) agonist or antagonist activity in the alkaline phosphatase assay using the human T47D breast carcinoma cell line. Both PR agonists and antagonists were achieved with an appropriate choice of 5-substitution. Several analogs were potent PR agonists (e.g., 12 and 13) or PR antagonists (e.g., 18) with good selectivity over other steroid receptors.

Novel 7-aryl benzo[1,4]oxazepin-2-ones were synthesized and evaluated as non-steroidal progesterone receptor (PR) modulators. The structure activity relationship of 7-aryl benzo[1,4]oxazepinones was examined using the T47D cell alkaline phosphatase assay. A number of 7-aryl benzo[1,4]oxazepinones such as 10j and 10v demonstrated good in vitro potency (IC(50) of 10-30 nM) and selectivity (over 100-fold) at PR over other steroidal receptors such as glucocorticoid and androgen receptors (GR and AR). Several 7-aryl benzo[1,4]oxazepinones were active in the rat uterine decidualization model. In this in vivo model, compounds 10j and 10u were active at 3 mg/kg when dosed orally.

A series of potent and binding selective LXRbeta agonists was developed using the previously reported non-selective LXR ligand WAY-254011 as a structural template. With the aid of molecular modeling, it was found that 2,3-diMe-Ph, 2,5-diMe-Ph, and naphthalene substituted quinoline acetic acids (such as quinoline 33, 37, and 38) showed selectivity for LXRbeta over LXRalpha in binding assays.

We have recently discovered 5-(3-cyclopentyl-2-thioxo-2,3-dihydro-1H-benzimidazol-5-yl)-1-methyl-1H-pyrrole-2-carbonitrile (14) as a potent, selective, and orally active non-steroidal progesterone receptor (PR) agonist. Compound 14 and its analog 13 possessed sub-nanomolar in vitro potency (EC(50) 0.1-0.5nM) in the T47D alkaline phosphatase assay, similar to that of the steroidal PR agonist medroxyprogesterone acetate (MPA). In contrast to MPA, 14 was highly selective (>500-fold) for the PR over both glucocorticoid and androgen receptors. In the rat uterine decidualization and complement component C3 models, 14 had oral ED(50) values of 0.02 and 0.003mg/kg, respectively, and was from 6- to 20-fold more potent than MPA. In the monkey ovulation inhibition model, compound 14 was also highly efficacious and potent with an oral ED(100) of 0.03mg/kg.

A series of phenyl acetic acid based quinolines was prepared as LXR modulators. An SAR study in which the C-3 and C-8 positions of the quinoline core were varied led to the identification of two potent LXR agonists 23 and 27. Both compounds displayed good binding affinity for LXRbeta and LXRalpha, and increased expression of ABCA1 in THP-1 cells. These two compounds also had desirable pharmacokinetic profiles in mice and displayed in vivo efficacy in a 12-week Apo E knockout mouse lesion model.

A structure-based approach was used to optimize our new class of quinoline LXR modulators leading to phenyl acetic acid substituted quinolines 15 and 16. Both compounds displayed good binding affinity for LXRbeta and LXRalpha and were potent activators in LBD transactivation assays. The compounds also increased expression of ABCA1 and stimulated cholesterol efflux in THP-1 cells. Quinoline 16 showed good oral bioavailability and in vivo efficacy in a LDLr knockout mouse model for lesions.

A variety of C17 amide-substituted 6-azaandrost-4-en-3-ones were prepared and tested versus human type 1 and 2 steroid 5 alpha-reductase (5AR) and human adrenal 3 beta-hydroxy-delta 5-steroid dehydrogenase/3-keto-delta 5-steroid isomerase (3BHSD) in order to optimize potency versus both isozymes of 5AR and selectivity versus 3BHSD. Two series of potent and selective C17 amides were discovered, 2,5-disubstituted anilides and (arylcycloalkyl)amides. Compounds from each series with picomolar IC50's versus human type 2 5AR and low nanomolar to picomolar IC50's versus human type 1 5AR possessing 100-500-fold selectivity versus 3BHSD were identified. A conformational model to predict 3BHSD potency was developed which could rationalize 3BHSD potency within three different series of compounds. Evaluation of some optimal compounds from this series in a chronic castrated rat model of 5AR inhibitor induced prostate involution, and pharmacokinetic measurements identified compounds (9, 12, 16, and 29) with good in vivo efficacy and half-life in the dog. An intact rat model of in vivo selectivity for 5AR versus 3BHSD inhibition was also developed. Dual inhibitors of both human 5AR's may show advantages over type 2 selective 5AR inhibitors, such as finasteride (1), in the treatment of disease states which depend upon dihydrotestosterone.