Configurationally Locked Ridaifen-B Analogs as Potent Ebola Virus Entry Inhibitors

Abstract

In the face of emerging outbreaks, high mortality rates, and limited treatment options, Ebola virus (EBOV) disease poses a global public health threat, underscoring the need for effective therapies. The unique and critical role that EBOV glycoprotein (GP) plays in the viral life cycle makes it an attractive target for antiviral drug discovery. In this study, we report on the design, synthesis, and evaluation of a series of cyclized ridaifen-B-based inhibitors with enhanced antiviral activity and selectivity against EBOV. Several compounds exhibited nanomolar potency inhibiting EBOV-GP-mediated entry and submicromolar potency against MARV entry. Site-directed mutagenesis indicated inhibitor binding at the EBOV GP1/GP2 fusion loop. The compounds displayed favorable in vitro pharmacokinetic parameters, which encouraged testing the in vivo rapid PK for hit compound 41. This study highlights the potential of 1,2-dihydronaphthalene and tetrahydronaphthalene derivatives as valuable leads for the development of potent antifiloviral agents targeting EBOV among other filoviruses.

Graphical Abstract

Introduction

The COVID-19 pandemic has revealed critical vulnerabilities in our global public health systems, highlighting the pressing and persistent need to develop antiviral therapeutics that can be rapidly deployed against high-consequence pathogens.¹ Among these are filoviruses, which cause severe hemorrhagic fever in humans. Filoviruses, members of the Filoviridae family, are zoonotic, filamentous, non-segmented, negative-sense single-stranded RNA viruses.² The Filoviridae family comprises five genera: Orthoebolavirus, Marburgvirus, Cuevavirus, Striavirus, and Thamnovirus.³ Among these, Orthoebolavirus includes Zaire Ebola virus (EBOV), Sudan virus (SUDV), Bundibugyo virus (BDBV), and Taï Forest virus (TAFV), while Marburgvirus includes Marburg virus (MARV). All these species are of significant concern due to their ability to infect humans and cause fatal disease. Human infections typically occur via contact with animal reservoir hosts, with subsequent human-to-human transmission through infected body fluids.²

EBOV and its relatives have caused frequent outbreaks of hemorrhagic fever disease in humans, with case fatality rates (CFRs) ranging from 30% to 100%.⁴ Since its first identification in 1976, EBOV has triggered recurrent epidemics in Central and West Africa. The 2014–2016 West African Ebola epidemic resulted in over 28,000 reported cases and 11,000 deaths.^5,6 In April 2025, Uganda declared the end of an Ebola disease outbreak, caused by the SUDV, resulting in twelve confirmed cases and four fatalities.⁷ In September 2025, the Democratic Republic of the Congo (DRC) declared its sixteenth outbreak of Ebola virus disease, which had resulted in 64 cases and 45 deaths, (70% CFR) despite the rapid deployment of response teams and emergency vaccination efforts.^8,9 Likewise, MARV remains highly lethal despite its lower incidence, and the 2025–2026 outbreak in Ethiopia, which resulted in 14 confirmed cases and 9 deaths, before being declared contained in January 2026,¹⁰ illustrates its expanding geographic reach and continuing public-health threat.¹¹

As this manuscript is being prepared, the 2026 outbreak of the rare BDBV in the DRC and Uganda has highlighted major gaps in filovirus preparedness, including the lack of approved diagnostics, vaccines, and therapeutics for non-Zaire filoviruses, prompting urgent international efforts to deploy experimental countermeasures and expand genomic surveillance.¹² The World Health Organization (WHO) lists EBOV and related filoviruses among its priority pathogens with epidemic potential and insufficient countermeasures, as outlined in the 2024 R&D Blueprint for epidemic and pandemic pathogens preparedness.^1,13 The United States Centers for Disease Control and Prevention (CDC) also classifies EBOV as a category A (highest priority) biological weapon and bioterrorism threat.^14,15 These concerns have driven extensive efforts toward the development of EBOV-targeted therapeutics, particularly against Zaire ebolavirus, the predominant cause of filovirus outbreaks and epidemics. Central to these antiviral strategies is a detailed understanding of EBOV structure, genome organization, and replication cycle, which provide multiple opportunities for small-molecule intervention.

EBOV has a genome of ~19 kb in length and encodes at least seven structural proteins, including the glycoprotein (GP), nucleoprotein (NP), polymerase complex (L, VP35, VP30), and matrix proteins (VP40, VP24).² EBOV-GP participates in multiple stages of the viral life cycle, functioning as a pivotal determinant of host cell entry and virion assembly (Figure 1A). The viral life cycle begins with GP-mediated host cell attachment and macropinocytosis, followed by endosomal trafficking and proteolytic priming of the GP. Primed GP binds to the endosomal Niemann–Pick C1 (NPC1) receptor, triggering membrane fusion and releasing the nucleocapsid into the cytoplasm (Figure 1A). Replication and transcription proceed via the viral RNA-dependent RNA polymerase (RdRp), with assembly and budding at the plasma membrane driven by VP40 (Figure 1A).^16,17 In later life cycle stages, the GP is synthesized on rough endoplasmic reticulum–bound ribosomes, glycosylated and processed through the ER–Golgi secretory pathway, and incorporated into nascent virions during VP40-mediated budding at the plasma membrane. EBOV-GP is the virus’s sole surface protein exposed to the extracellular environment and is highly conserved across isolates, with >90% amino acid conservation.¹⁸ The essential, unique, and multifunctional nature of EBOV-GP underscores its significance as a molecular target for antiviral drug discovery (Figure 1A).^19,20

Figure 1.

(A) Life cycle of Ebola virus (EBOV), highlighting the essential roles of the viral glycoprotein (GP) and its potential as a target for antiviral drug development. Created in BioRender. Aboukhatwa, S. (2026) https://BioRender.com/38z9j82 (B) Selective Estrogen Receptor Modulators (SERMs) with reported filoviral GP-mediated entry inhibitory activity.

Abbreviations: GP, viral Glycoprotein; NPC1, Niemann-Pick C1 Receptor; RNP, Ribonucleoprotein; vRNA, viral RNA; cRNA, complementary RNA; mRNAs, messenger RNA; RdRp, RNA-dependent RNA polymerase.

The current FDA-approved therapies for EBOV specifically target the EBOV-GP-mediated entry, by neutralizing the virus during the initial stages of host-cell entry. These include the Zaire ebolavirus strain vaccine, Ervebo^®, and the monoclonal antibodies; Inmazeb^® and Ebanga^®;^21,22 however, these interventions have notable limitations, including restricted cross-species activity, high cost, cold-chain supply logistics requirements because of the temperature-sensitive materials, and patient administration challenges. In contrast, small-molecule antivirals offer advantages such as oral delivery, stability, and scalability, which are desirable features for an effective pandemic response in resource-deficient settings.²³ In addition, small-molecule direct-acting antivirals provide broader mechanistic flexibility, with the ability to interfere with several stages of the filovirus life cycle, including viral attachment and early entry,²⁴ endosomal trafficking and late entry,^25,26 genome transcription and replication,^27,28 and viral budding.²⁹

The replication-targeting nucleoside analogs, favipiravir,³⁰ brincidofovir,³¹ and remdesivir²¹ are the most clinically advanced small-molecule candidates evaluated for EBOV infection; however, each ultimately failed to progress further because of inadequate efficacy and/or safety limitations compared to antibody-based therapies (Table S1, Supplementary Information). More recently, obeldesivir demonstrated complete protection in non-human primate models challenged with lethal SUDV infection.³² Previous medicinal chemistry efforts identified and performed structure-activity relationship studies (SAR) for several compound classes targeting EBOV entry, including triphenylethylenes,³³ adamantane dipeptide piperazines,³⁴ adamantane carboxamides,³⁵ benzamide derivatives,^36,37 N-substituted pyrroles,³⁸ thiophenes,³⁹ and diarylsulfides.⁴⁰ Despite these efforts, the discovery of small molecules that specifically target filoviral entry remains underdeveloped.⁴¹ Based on the currently available literature, none of the reported EBOV-GP-targeting small molecules have advanced into clinical trials (Table S1, Supplementary Information).

The clinical studies of EBOV therapeutics are inherently challenging because of the virus’s high lethality, the requirement for BSL-4 containment, sporadic and unpredictable outbreaks, and the logistical constraints of conducting trials in remote, resource-limited settings. Ethical concerns during large outbreaks have also limited the use of randomized controlled trials, resulting primarily in single-arm studies that complicate rigorous assessment of efficacy.^42,43 In addition, many investigational small molecules have shown limited-spectrum, suboptimal pharmacokinetic properties, toxicity liabilities, or reduced efficacy when treatment is initiated after advanced disease onset.

The limited scope of the currently approved anti-EBOV therapeutics and the recent outbreaks caused by non-Zaire filoviruses point to a critical gap in filovirus epidemic preparedness, underscoring the need for potent antivirals with favorable pharmacokinetic and drug-like properties that are also practical to administer and distribute during outbreaks. This study reports the design and structure–activity relationship–driven optimization of a new small-molecule chemotype targeting EBOV glycoprotein–mediated viral entry, yielding inhibitors with enhanced antiviral potency and improved selectivity relative to their cellular cytotoxicity, supporting their suitability for in vivo evaluation and potential development as deployable therapeutic or prophylactic agents during the Ebola virus outbreaks.

Design Rationale

In the pursuit of effective EBOV therapeutics, systematic drug repurposing screens of FDA-approved small molecules have identified multiple agents with potential inhibitory activity against EBOV.⁴⁴ Among these, selective estrogen receptor modulators (SERMs), such as clomiphene (1) and toremifene (2) have demonstrated the ability to inhibit EBOV entry by interfering with the EBOV-GP function, independently of estrogen receptor binding activity, resulting in improved survival in a rodent model of EBOV infection (Figure 1B).²⁵ However, there is no reported evidence that any of the SERMs, including toremifene, have progressed to clinical evaluation, likely due to inherent challenges associated with conducting EVD clinical trials (Table S1, Supporting Information). In a focused screen of SERMs, another triphenylethylene compound, ridaifen-B (compound 3) demonstrated potent in vitro dual inhibitory activity against both EBOV (EC₅₀ = 0.08 μM) and MARV (EC₅₀ = 1.2 μM).⁴⁵ Pilot structural optimization to ablate ER-mediated side effects led to several compounds such as XL-47 (compound 5) that retained potent inhibition against infectious EBOV (EC₅₀ = 0.1 μM) and MARV (EC₅₀ = 0.6 μM) while lacking estrogenic activity.⁴⁵ Accordingly, the present study was designed as a continuation of this validated SERM-derived scaffold optimization strategy to further improve broad-spectrum anti-filovirus activity while minimizing ER-related liabilities.

SERMs such as toremifene (2) bind to EBOV-GP near the fusion loop, specifically at the GP1/GP2 interface (Figure S1, Supporting Information). Inhibitor binding destabilizes the GP and triggers premature release of GP2, preventing fusion between the viral and endosomal membranes (stage 4, Figure 1A).⁴⁶ The co-crystal structure of 2 with EBOV-GP revealed a highly hydrophobic, well-defined binding pocket that accommodates the aromatic rings of 2, allowing its direct interactions with key fusion loop residues (Figure 2A and Figure S2, Supporting Information). Among these, π-π stacking interactions occur between the three aromatic rings of 2 and Y517. Mutagenesis studies have shown that Y517 mutants have a reduced susceptibility to the inhibitory activity of 2.⁴⁷ Additional hydrophobic residues, such as V66, L68, L184, L515, and L558, line the binding pocket and contribute to the binding of the ligand via interactions with its aromatic rings and chloroethyl group.⁴⁶ Thus, the proper orientation of the aromatic rings in 2 is crucial for preserving the interactions with Y517 and attaining optimal engagement with the hydrophobic pocket (Figure 2A and Figure S2, Supporting Information).

Figure 2.

Design of 1,2-dihydronaphthalene analogs as EBOV-GP destabilizers. A) Schematic representation of the binding mode of toremifene (2) in the EBOV GP co-crystal structure (PDB: 5JQ7), highlighting critical interactions and the core cyclization strategy that guided the design of the current series. B) Optimization of triphenylethylene derivatives to improve antifiloviral potency and selectivity.

The dimethylamine group of 2 extends toward the hydrophilic tunnel, where it interacts with polar and charged residues, including R64, E100, T519, T520, and D522 (Figure 2A and Figure S2, Supporting Information).⁴⁶ These interactions contribute to stabilization of the ligand within the fusion loop region and enhance its inhibitory activity; however, the precise role of the basic amine groups is debated.⁴⁷ Some studies suggest that it may not be essential for direct binding to EBOV-GP but instead plays a role in enhancing lysosomal trapping, thereby contributing to the compound’s overall potency.⁴⁷ The functional significance of this basic group in toremifene-like molecules requires further investigation. Unfortunately, the chemical synthesis of triphenylethylene derivatives, such as 1–3 and 5, often produces mixtures of E/Z isomers, with varying ratios for different derivatives and even with batch-to-batch variation for the same compound.^48,49 Triphenylethylenes can readily undergo E/Z isomerization in solution, with rates influenced by their chemical features, solvent nature, and temperature. This intrinsic property complicates the preparation and storage of pure isomers, since such interconversion can occur during purification, in stock solutions, and under biological assay conditions, compromising the accuracy and reproducibility of pharmacological evaluations.⁵⁰ It becomes challenging to compare potencies or correlate activity with receptor binding, since the isomeric forms of triphenylethylenes differ in receptor binding affinity, metabolic liability, and PK properties.^51–54

Therefore, we decided to explore chemical modifications that improve the potency and selectivity while enhancing the stability and PK properties of these triphenylethylene lead compounds. The triphenylethylene SERM derivatives were selected as a starting scaffold for the current study because they possess one of the most extensively validated mechanisms of interaction with EBOV-GP, supported by the available co-crystal structure of toremifene bound to EBOV-GP. In addition, these compounds exhibit the most potent anti-EBOV entry inhibitory activity reported in the literature. The current design aims to rigidify the triphenylethylene core in compound 3 via cyclization to a 1,2-dihydronaphthalene scaffold (Figure 2A). Notably, similar fused triaryl scaffolds, such as those in nafoxidine (compound 4, Figure 1B), showed modest EBOV-inhibitory activity in a previous HTS campaign (EC₅₀ = 0.32±0.03 μM).⁴⁵ By constraining the molecules into a defined geometry, we aimed to (1) eliminate E/Z isomerization of the triphenylethylene scaffold to improve chemical stability, inter-batch reproducibility, and consistency in biological evaluation (2) orient the pharmacophoric aromatic rings A, B, and C to optimize interactions with the EBOV-GP fusion loop and enhance potency; (3) reduce conformational flexibility, thereby lowering the entropic penalties upon binding and improving affinities (4) enhance metabolic stability, potentially improving systemic exposure and PK properties; and (5) reduce off-target cytotoxicity, thereby improving overall selectivity.⁵⁵

As outlined in Figure 2B, the current study builds upon previous optimization of the triphenylethylene-based filovirus inhibitor scaffold. Initial studies identified toremifene (2) as a potent and selective EBOV entry inhibitor; however, limited anti-MARV activity and retained SERM activity restricted its utility as a broad-spectrum lead. Subsequent focused screening of SERMs led to the identification of Ridaifen-B (3) as a dual EBOV/MARV inhibitor, and further modification of the polar ring B substituent yielded XL-47 (5), which retained antiviral potency but lacked detectable estrogen receptor activity.⁴⁵ Based on these findings, the present study explored conformational restriction of the flexible triphenylethylene scaffold through cyclization of both compounds 3 and 5 frameworks. Simplification of the solvent-exposed surface-binding region were incorporated to reduce flexibility and further define structure–activity relationships associated with filovirus activity and selectivity.

Conformational restriction of 3 generated derivative 6, resulting in an approximately three-fold enhancement in EBOV potency while maintaining anti-MARV activity, suggesting that reduced conformational flexibility improved productive target engagement. Integration of both favorable design features, namely scaffold cyclization and optimized ring B substitution, with simplification of C-6 substituent produced derivative 7, which exhibited potent activity against EBOV and high selectivity.

The following sections describe the synthetic strategies, structure–activity relationships (SAR), and antiviral efficacy of the new compounds. The compounds were screened using a biosafety level 2 (BSL-2) pseudovirus assay for both EBOV and MARV. Promising candidates were then subjected to phenotypic screening with live viruses in BSL-4 facilities, leading to the identification of potent GP inhibitors and providing critical SAR insights for effective EBOV entry blockade.

Results & Discussion

Several structural modifications around the 1,2-dihydronaphthalene core were undertaken to enhance potency and selectivity against EBOV or MARV, as illustrated in Figure 2. A diverse array of substituents was introduced via various linkers at the C-1 position of the 1,2-dihydronaphthalene core, at the aryl ring B (Figure 2), which is predicted to occupy the main tunnel region. The core ring was also reduced to a tetrahydronaphthalene scaffold, aiming to improve chemical stability and PK properties.

Chemical Synthesis

Scheme 1 outlines the synthesis of the 1,2-dihydronaphthalene derivatives 12–36, with varying substitutions and linkers at the C-1 position. To install an N-methylpiperidine carboxamide moiety analogous to that in XL-47 (5), the corresponding boronic acid pinacol ester 9 was synthesized as shown in Scheme 1A. This process began with amide coupling between 4-bromoaniline and N-methylpiperidine-4-carboxylic acid using benzotriazol-1-yloxytripyrrolidinophosphonium hexafluorophosphate (PyBOP) as the coupling reagent, which provided the best yield, among others tested. The resulting bromo-amide 8 was then converted, through Miyaura borylation reaction, to the boronic acid pinacol ester 9 via reflux with bis(pinacolato)diboron.

Scheme 1.

Synthesis of C-1 modified 1,2-dihydronaphthalene derivatives 12–36^a

^a Reagents and conditions: (a) PyBOP, i-Pr₂NEt, CH₂Cl₂, rt, 18 h; (b) [(CH₃)₄C₂O₂B]₂, Pd(dppf)Cl₂, KOAc, 1,4-dioxane, 100 °C, 20 h; (c) (1) LiN(Si(CH₃)₃)₂, THF, −78 °C, 45 min; (2) PhN(SO₂CF₃)₂, −78 °C→rt, 4 h; (d) variable Ar-BPin, Pd(dppf)Cl₂, Cs₂CO₃, 1,4-dioxane/water (6:1), 90 °C, 18 h; (e) BBr₃, CH₂Cl₂, 0 °C, 2 h; (f) 2M HCl, 1,4-dioxane, 0 °C→rt, 2 h; (g) 10% Pd/C (50% wet), MeOH, rt, air, 72 h.

To generate an exploratory library for SAR studies at the C-1 position of the 1,2-dihydronaphthalene nucleus, the synthesis began with the enolization of 6-methoxy-2-phenyl tetralone (10) using lithium bis(trimethylsilyl)amide (LiHDMS) as a base, followed by trapping of the corresponding enolate with phenyl triflimide to give the triflate intermediate 11 (Scheme 1B). Although triflic anhydride also functioned as a triflating agent for tetralone 10, it gave lower yields than the milder phenyl triflimide. Enol triflate 11 was subjected to Suzuki–Miyaura coupling with a panel of commercially available or in-house-prepared boronate pinacol esters, affording a focused set of 1,2-dihydronaphthalene derivatives 12 and 14–33. The dimethoxy derivative 12 was subjected to boron tribromide-catalyzed demethylation to give the diphenol derivative 13. Acid-catalyzed Boc deprotection of derivatives 18 and 32 yielded the corresponding free amines 34 and 35, respectively. Compound 36 was produced through palladium-catalyzed N-demethylation of 33 under aerobic conditions (Scheme 1B).^56,57 Selected derivatives (25, 27, 28, and 33) were subjected to catalytic hydrogenation to give the corresponding tetrahydronaphthalene derivatives (37, 38, 39, and 40), respectively, as illustrated in Scheme 2. Acid-promoted Boc deprotection of 40 gave the free amine 41. Thiolation of 40 using Lawesson’s reagent yielded the Boc-deprotected thioamide congener 42 (Scheme 2).

Scheme 2.

Synthesis of the tetrahydronaphthalene derivatives 37–42.^a

^a Reagents and conditions: (a) Pd/C, EtOAc/MeOH, rt, H₂, 50 psi, 24–72 h; (b) 2M HCl, 1,4-dioxane, 0 °C→rt, 2 h; (c) Lawesson’s reagent, THF, 50 °C, 48 h.

The stereochemistry of hydrogenation products (37–42) was assigned based on the well-established stereochemical outcome of Pd/C-catalyzed hydrogenation of sterically hindered cyclic alkenes, which proceeds predominantly through syn addition to afford the cis isomers, a mechanistic principle that is widely recognized in the organic chemistry literature. Accordingly, compounds (37–42) were assigned as the cis racemates (1R,2S/1S,2R). This assignment is also supported by the ¹H NMR spectrum of derivatives 37–42 where the proton at C-1, showing as a doublet at ca. 4.3 ppm, exhibited a coupling constant (J) ranging from 4.4 to 5.3 Hz, consistent with the expected cis relationship between the C-1 and C-2 protons. Attempts to separate the enantiomers of 41 by chiral HPLC using different chiral stationary phases and elution conditions were unsuccessful, consistently resulting in a single broad unresolved peak. Diastereomeric salt resolution using d-tartaric acid provided two fractions with different solubility profiles (Table S2, Supporting Information). We were unable to obtain crystals suitable for X-ray diffraction analysis; therefore, the absolute configuration of the stereocenters could not be definitively assigned. The isolated fractions were evaluated against pEBOV and pMARV (Table S4, Supporting Information).

Structure–Activity Relationship

Establishing SAR from assays conducted in complex cellular systems is inherently challenging. Although a target-specific mechanism of action is presumed, the potential contribution of off-target effects, contributing in less apparent ways, cannot be ruled out. Accordingly, attention was paid to exclude known sources of non-specific antiviral activity, such as lysosomal trapping and intracellular pH disruption. To further support the hypothesis of EBOV-GP–specific binding, a subset of derivatives predicted to have poor binding affinity to EBOV-GP was intentionally included in the series; these compounds exhibited minimal or no antiviral activity. Namely, these compounds are non-basic amine–containing derivatives, including the sterically demanding Boc-protected compounds 18 and 32, methoxy derivative 12, hydroxy derivative 13, tetrahydropyranyl derivative 29, and the small, sterically less demanding primary amine 34.

Compounds 12–40 were evaluated in A549 cells transduced with either HIV/EBOV-GP (pEBOV) or HIV/MARV-GP (pMARV) pseudotyped viruses, using a luciferase reporter assay to assess filoviral entry inhibitory activity. The inhibitory activities of the tested compounds are summarized in Tables 1 and 2, with toremifene (2) included as a positive control in each assay run. Compounds 3–5 were also tested in parallel to investigate the impact of core structure cyclization on both antiviral potency and cytotoxicity.

Table 1.

Filoviral entry inhibitory activity of C-1 modified 1,2-dihydronaphthalene derivatives 12–36 against pEBOV and pMARV.

#	Structure	R	A549 CC50 (μM)	pEBOV EC50 (μM)a	SI pEBOVb	pMARV EC50 (μM)a	SI pMARVb
12	I		>100	0.67±0.26	>149	>100	NA
13	II		58±9	0.48±0.13	120	18±4	3
14	III		>100	>10	ndc	>10	ndc
15	III		26±1	3.44±0.10	7	3.3±0.05	8
16	IV		13±0.4	0.14±0.05	96	0.67±0.06	20
17	IV		48±4	0.64±0.18	76	3.9±0.7	12
18	I		36±8	>100	<0.36	>100	ndc
19	I		>100	1.1±1.0	>90	2.3±0.6	44
20	I		44±4	0.62±0.21	70	>100	<0.4
21	I		>100	6.5±0.9	>15	37±ndc	>3
22	I		28.0±0.8	0.5±0.24	56	2.0±0.6	14
23	I		8.9±0.4	0.15±0.02	58	0.55±0.05	16
24	I		>100	0.90±0.05	>111	9.8±0.9	>10
25	I		8.0±0.1	0.03±0.01	253	1.09±0.11	7
26	I		7.7±0.7	0.11±0.02	72	1.02±0.14	8
27	I		14.0±0.4	0.04±0.03	392	1.6±0.1	9
28	I		29.0±0.6	0.06±0.01	471	4.1±0.4	7
29	I		38±10	24±7	2	32±ndc	1
30	I		>100	>100	ndc	>100	ndc
31	I		13.0±0.2	0.12±0.01	105	5.0±0.3	3
32	I		>100	>100	ndc	4.9±1.6	>20
33	I		3.9±0.1	0.017±0.004	220	1.07±0.08	4
34	I		62±9	2.0±0.7	13	28±0.5	3
35	I		7.5±0.1	0.026±0.005	293	0.44±0.10	17
36	I		7.3±0.1	0.020±0.007	293	1.11±0.05	7
2	Toremifene	21.0±0.3	0.027±0.004	754	2.53±0.03	8
3	Ridaifen-B	3.8±0.03	0.13±0.04	30	0.27±0.02	14
4	Nafoxidine	14.0±0.2	0.22±0.08	64	1.05±0.03	13
5	XL-47	2.1±0.06	0.24±0.12	9	1.52±0.13	1

^aDose–response curves were used to calculate EC₅₀ values using the four-parameter variable slope nonlinear regression analysis in GraphPad Prism 10.2.3. Data are presented as the mean EC₅₀ from three replicates ± standard deviation.

^bSelectivity indices (SI) are calculated as a ratio of CC₅₀/EC₅₀ where CC₅₀ is cytotoxicity measured in a CellTiter-Glo Luminescent Cell Viability Assay.

^cNot determined.

Table 2.

Filoviral entry inhibitory activity of tetrahydronaphthalene derivatives 37–42 compared to their precursor 1,2-dihydronaphthalene (25, 27, 28, 32 and 35) against pEBOV and pMARV.

#	Structure	R	A549 CC50 (μM)	pEBOV EC50 (μM)a	SI pEBOVb	pMARV EC50 (μM)a	SI pMARVb
25	I		8.0 ± 0.1	0.03 ± 0.01	253	1.1 ± 0.1	7
37	II		15.6 ± 0.3	0.26 ± 0.09	60	1.2 ± 0.1	13
27	I		14.1 ± 0.4	0.04 ± 0.03	392	1.6 ± 0.1	9
38	II		57 ± 2	0.74 ± 0.11	77	3.3 ± 0.2	17
28	I		29 ± 0.6	0.06 ± 0.01	471	4.1 ± 0.4	7
39	II		36 ± 1	0.86 ± 0.10	42	5.8 ± 2.2	6
32	I		>100	>100	ndc	4.9 ± 1.6	>20
40	II		>100	>100	ndc	1.5 ± 0.1	>66
35	I		7.5 ± 0.1	0.026 ± 0.005	293	0.44 ± 0.10	17
41	II		6.8 ± 0.2	0.040 ± 0.002	172	0.42 ± 0.03	16
42	II		8.7 ± 0.4	2.9 ± 0.1	3	0.60 ± 0.09	15

^aDose–response curves were used to calculate EC₅₀ values using the four-parameter variable slope nonlinear regression analysis in GraphPad Prism 10.2.3. Data are presented as the mean EC₅₀ from three replicates ± standard deviation.

^bSelectivity indices are calculated as the ratio of CC₅₀/EC₅₀ where CC₅₀ is cytotoxicity measured in a CellTiter-Glo Luminescent Cell Viability Assay.

^cNot determined.

In the pseudoviral assays, derivatives 12–36, modified at the C-1 position, exhibited a wide range of potencies (Table 1). The 1,2-dihydronaphthalene derivative 12, bearing non-basic methoxy substituents on both rings A and B, demonstrated submicromolar activity against pEBOV (EC₅₀ = 0.67 μM). This result suggests that the spatial orientation of the aryl rings is compatible with the EBOV-GP binding site, resulting in significant inhibition. This result is consistent with the previously reported anti-EBOV activity of compounds lacking an ionizable moiety.⁴⁷ Moreover, the antiviral activity of the non-cationic derivatives supports the notion that their mechanism of action is unlikely to be solely mediated through non-specific accumulation within the acidic environment of endosomal compartments.⁵⁸ The biphenolic congener 13 exhibited comparable potency against pEBOV entry (EC₅₀ = 0.48 μM) and a moderate inhibition of pMARV entry (EC₅₀ = 18 μM); however, it displayed increased cytotoxicity, potentially attributable to off-target effects (Table 1).

The chloroimidazole 14, which features a smaller substituent at the C-1 position, was inactive against either pEBOV or pMARV within the tested concentration range, highlighting the importance of a phenyl group at this position for antiviral activity. On the contrary, derivative 15, incorporating an increased steric bulk and an aromatic amine substituent, regained moderate activity against both pseudoviruses (EC₅₀ ≈ 3 μM) compared to the inactive 14; however, it remained approximately 10-fold less potent compared to 12. From another perspective, 15 exhibited greater potency than the unsubstituted aniline 34 against pEBOV and >30-fold more active against pMARV. Compound 34, which bears an ionizable amino group on aryl ring B, was ca. 10-fold less potent against pEBOV (EC₅₀ = 5 μM) compared to 12, further supporting that lysosomal trapping is not the primary mechanism of action for this class of derivatives. Compound 18, the Boc-protected precursor of 34, lacked pseudoviral entry inhibitory activity, likely due to the steric bulk at the aryl ring B and complete masking of the basic amine group. The dimethylamino-substituted derivative 19 exhibited a 5-fold increase in potency against pEBOV entry (EC₅₀ = 1 μM) and >100-fold greater potency against pMARV entry (EC₅₀ = 2 μM), compared to 34 (Table 1).

The 4-pyrrodinyl derivative (20), introducing additional steric bulk at ring B, was ca. 2-fold more potent than 19 against pEBOV. In contrast, the bulkier 4-piperidinyl derivative 21, exhibited ca. 10-fold lower potency than 19. Replacing the 4-piperidinyl of 21 with an N-methylpiperazinyl moiety, introduced an additional ionizable tertiary amine in 22, which exhibited ca. 10-fold and 20-fold increased potency against pEBOV and pMARV, respectively. Introducing a methylene linker to the N-methylpiperazinyl substituent in 23 led to an additional 3-fold increase in potency compared to 22. This finding further supports the critical role of the ionizable amine group in mediating interactions with EBOV-GP, beyond its contribution to endosomal trapping. Compound 16, the meta-substituted analogue of 23 exhibited comparable potency. Compound 24, featuring a less basic morpholine substituent in place of the N-methylpiperazine of 23, exhibited a 10-fold reduction in potency against both pEBOV and pMARV (Table 1).

Linking the N-methylpiperazine moiety via a carbonyl group, as in compound 28, enhanced the activity to the double-digit nanomolar range (pEBOV EC₅₀ = 50 nM). This improvement suggests the potential of an additional hydrogen-bonding interaction that contributes to the potency, despite the reduced basicity resulting from the conversion of one tertiary amine in 23 into an amide. Compound 27, the N-demethylated analogue of 28, achieved comparable potency against pEBOV with an EC₅₀ = 40 nM. To increase polarity, improve aqueous solubility, and enhance hydrolytic stability, a sulfonamide linker was introduced in compound 30, in place of the carboxamide linker present in compound 28. Unexpectedly, compound 30 displayed a complete loss of activity. This may be attributed to the differing hydrogen-bonding capabilities and geometric preferences between sulfonamides and carboxamides. Specifically, the tetrahedral geometry of the sulfonamide nitrogen, in contrast to the sp²-like geometry of the carboxamide nitrogen, could alter the spatial orientation of the linked pharmacophore. Moreover, the loss of a hydrogen-bond donor and the increased acidic character of the sulfonamide may weaken target engagement and reduce endosomal uptake.

Compound 25, a non-cyclic congener of 28, showed an even lower pEBOV EC₅₀ (30 nM), potentially due to increased conformational flexibility, allowing for more favorable interactions with the acidic residues in the target binding site. Compound 26, incorporating a bulkier N-hexyl substituent instead of the N,N-dimethyl amino moiety of 25, exhibited ca.4-fold reduction in pEBOV activity; however, it retained comparable potency against pMARV. Introducing a urea linker, as in 33, in place of the amide linker of 25, resulted in the most potent derivative identified to this point in this SAR study, with a pEBOV EC₅₀ of 17 nM. The N-demethylated congener (36) exhibited comparable potency (pEBOV EC₅₀ = 20 nM) and similar selectivity (Table 1).

The N-methylpiperidinyl counterpart of 33, compound 31, showed a 5-fold reduction in potency against both pEBOV (EC₅₀ = 0.12 μM) and pMARV (EC₅₀ = 5 μM); however, the non-methylated piperidinyl counterpart (35) regained the nanomolar potency comparable to that of compound 33. Analog 32, the N-Boc precursor of 35, was inactive against pEBOV but showed micromolar potency against pMARV. This result suggests that steric bulk is favored in the MARV inhibition. Converting the piperidine ring in compound 35 to a pyran (29), which lacks any basic amine, reduced the potency by ca. 100–1000-fold (pEBOV EC₅₀ = 24 μM, pMARV EC₅₀ = 32 μM) (Table 1). A summary of the results of the initial round of structural optimization is presented in Figure 3(A). Despite the nanomolar potencies of derivatives compounds 25, 28, 33, and 35, which were comparable to toremifene (2) and superior to ridaifen-B (3), their selectivity indices, ranging from 200 to 500-fold, were less favorable than that of toremifene (2).

Figure 3.

(A) Structural variations at C-1 led to significant improvement in pEBOV potency. (B) Tetrahydronaphthalene derivatives are less potent inhibitors of pEBOV than 1,2-dihydronaphthalene derivatives.

One strategy to reduce the cytotoxicity and improve the antiviral selectivity involved hydrogenating the double bond in the 1,2-dihydronaphtahlene core, as depicted in Scheme 2. Conversion from a 1,2-dihydronaphthalene to a tetrahydronaphthalene scaffold for a subset of derivatives generally led to an average 10-fold reduction in potency against pEBOV, while retaining comparable activity against pMARV. Additionally, this structural modification resulted in a modest improvement in cytotoxicity, with an average 2-fold increase in CC₅₀ values (Table 2). The N-methylpiperazine 39 exhibited ca. 15-fold decrease in pEBOV potency compared to the analogous dihydronaphthalene (28). The unalkylated piperazine 38 showed ca. 20-fold reduction in pEBOV potency relative to its analog (27). In contrast, piperidine 41 demonstrated only a modest 1.5-fold decrease in pEBOV inhibitory activity compared to 35.

Although the Boc-protected compound 40 was inactive against pEBOV, it exhibited enhanced potency and selectivity against pMARV, highlighting the favorable interaction of bulkier compounds with the MARV-GP. This behavior was akin to that observed with its 1,2-dihydronaphthalene precursor 32. The conversion of the urea linker in 41 to a thiourea (42) led to ca. 75-fold reduction in potency against pEBOV, while retaining nearly identical pMARV activity and cytotoxicity (Table 2). This observation further supports the specific EBOV-GP binding mode of this series of compounds, highlighting the likely role of hydrogen bonding interactions between the target site residues and the linker region of the molecules. The isolated d-tartrate diastereomeric salts of 41 displayed slight differences in biological activities, which were within the sensitivity limits of the low-micromolar assay (Tables S2 and S3, Supporting Information). A summary of the findings of the second round of structural optimization is presented in Figure 3B.

Overall, the derivatives exhibited higher potency against pEBOV than pMARV. Due to the lack of structural information describing MARV GP in complex with small-molecule entry inhibitors, detailed structural interpretation of the observed SAR differences remains limited. Sequence comparison of EBOV and MARV glycoproteins suggests a conserved GP architecture, with ~35% sequence identity and highly similar structural organization, consistent with a shared mechanism of viral entry among filoviruses.⁵⁹ However, localized divergence within the putative binding region, including substitution of EBOV Y517 with Ile in MARV, may alter steric and hydrophobic properties of the pocket, potentially contributing to the observed differences in antiviral activity and the trend toward increased MARV selectivity with bulkier substituents. Additionally, the presence of an ionizable basic group on the aryl group B seems important for activity. The optimal orientation of the amine within the polar cavity enables electrostatic and hydrogen-bonding interactions with acidic residues, which are essential for maintaining or enhancing activity. Increased bulkiness and total polar surface area of the molecule are favored for pMARV activity.

Entry Inhibitory Activity Against Other Ebolaviruses

Different species of ebolaviruses have been implicated in multiple epidemics over the years. For example, SUDV, the second most prevalent ebolavirus species, was responsible for the two most recent outbreaks in Uganda in 2022 and 2025.^6,7,60 Current outbreaks caused by BDBV in the DRC and Uganda have further emphasized the absence of approved vaccines and therapeutics for non-Zaire filoviruses.¹² Therefore, we decided to evaluate whether the derivatives exhibit broad-spectrum activity against other ebolavirus species. The filoviral entry inhibitory activity of two of the most potent compounds (the dihydronaphthalene 35, and the tetrahydronaphthalene 41) was evaluated against pseudotyped SUDV, BDBV, and TAFV, using toremifene (2) as a reference (Table 3).

Table 3.

Broad-spectrum anti-filoviral activity of selected derivatives against pseudoviral Ebola (pEBOV), Sudan (pSUDV), Bundibugyo (pBDBV), and Tai Forest (pTAFV) viruses.

#	pEBOV EC50 (μM)a	pMARV EC50 (μM)a	pSUDV EC50 (μM)a	pBDBV EC50 (μM)a	pTAFV EC50 (μM)a
35	0.026 ± 0.005	0.44 ± 0.1	1.17 ± 0.13	> 10	1.10 ± 0.11
41	0.04 ± 0.002	0.42 ± 0.03	1.14 ± 0.13	> 10	0.67 ± 0.30
2	0.03 ± 0.004	2.53± 0.03	1.13 ± 0.34	> 10	3.88 ± 1.24

^aDose–response curves were used to calculate EC₅₀ values using the four-parameter variable slope nonlinear regression analysis in GraphPad Prism 10.2.3. Data are presented as the mean EC₅₀ from three replicates ± standard deviation.

Both derivatives 35, and 41 exhibited comparable or greater potency than toremifene (2) in inhibiting pEBOV entry and showed micromolar inhibitory activity against pSUDV, with EC₅₀ values ≤1 μM, comparable to that of toremifene (2) (Table 3). Both displayed similar potency against pTAFV, with EC₅₀ values comparable to those observed for pSUDV and higher potency than toremifene (2). Submicromolar potency of both 35, and 41 was also observed against pMARV, highlighting their potential for further development as broad-spectrum antifiloviral drugs. Interestingly, these compounds inhibited pBDBV entry with lower potency relative to other tested filoviruses, with EC₅₀ values >10 μM, which may be attributed to the genetic variability compared to GPs of other ebolavirus species.^61–63 This result was unexpected, given the high overall sequence similarity among filoviral glycoproteins; however, in the absence of high-resolution inhibitor-bound 3D structures for MARV, BDBV, and SUDV GPs, a detailed structural explanation for these differences cannot be established. The observed variability, therefore, likely reflects subtle, unresolved differences in the glycoprotein-binding environment that are not captured by available sequence comparisons alone. These findings suggest that even closely related filoviruses may display distinct sensitivity profiles to this chemotype.

Structural Basis for Viral Entry Inhibition Activity

As described above, the derivatives were designed based on the structures of known EBOV GP binders, such as toremifene (2). To verify that these compounds act via binding to the GP in a mechanism similar to that of toremifene, we conducted site-directed mutagenesis studies and docking analyses. Although direct biophysical characterization could not be performed due to limitations in obtaining sufficient quantities of properly folded glycoprotein, the current pseudovirus entry inhibition, site-directed mutagenesis, and molecular docking data collectively support interaction within the GP1/GP2 fusion loop region.

Mutagenesis studies of EBOV GP validate GP binding as the mechanism of viral entry inhibition

Site-specific single amino acid mutagenesis of pEBOV GP was used to elucidate the molecular basis of entry inhibition by the new derivatives. Previous studies have shown that Y517 mutants of EBOV GP exhibit significantly reduced susceptibility to toremifene (2) compared to wild-type (WT) GP. Y517 is highly conserved across EBOV species and is located in the internal fusion loop of the EBOV GP. Its substitution with either serine or isoleucine is known to destabilize toremifene (2) binding, greatly reducing EBOV susceptibility to inhibition by 2.⁴⁵

Derivatives 35 and 41 were tested for pseudoviral entry inhibitory activity against Y517S and Y517I mutants (Table 4). Both 35 and 41 showed a rightward shift in the dose–response curves—hence decreased potency, relative to WT GP, indicating that Y517 is critical to the binding mechanism of these inhibitors (Table 4). Consistent with previous reports, toremifene (2) showed substantially decreased activity against Y517 mutants, with EC₅₀ values 576-fold and 314-fold higher for Y517S and Y517I, respectively, relative to WT EBOV GP (Table 4).

Table 4.

Inhibitory activity of derivatives 35, 41, and 2 against pEBOV expressing wild-type (WT) and mutated GPs.^a

pEBOV GP Mutants	35	41	2
	EC50 (μM)	EC50 (μM)	EC50 (μM)
	[Fold change]	[Fold change]	[Fold change]
WT	0.026±0.005	0.040±0.002	0.027±0.004
Y517S	0.68±0.25	1.12±0.04	15.56±4.64
	[26]	[28]	[576]
Y517I	0.27±0.08	0.61±0.03	8.48±2.94
	[10]	[15]	[314]

^aDose–response curves were used to calculate EC₅₀ values using the four-parameter variable slope nonlinear regression analysis in GraphPad Prism 10.2.3. Data are presented as the mean EC₅₀ from three replicates ± standard deviation.

Binding Mode Insights: EBOV-GP Inhibitors Compared to Toremifene

To get insights into the binding mode of our inhibitors, in silico docking studies were performed using the X-ray crystal structure of EBOV GP in complex with 2 (PDB ID 5JQ7, 2.69 Å), as the template.⁴⁶ The biological assembly is a trimer of GP1–GP2 heterodimers, with three molecules of 2 occupying three equivalent pockets formed at the internal fusion-loop at the GP1/GP2 interfaces (Fig. 3A). The pocket is amphipathic, with a largely hydrophobic entrance that funnels into a more polar channel toward the trimer core (Figure 4B).

Figure 4.

Binding Mode of EBOV-GP Inhibitors (A) EBOV GP trimer composed of three dimers: chains A and B (blue and purple), chains C and D (red and orange), and chains E and F (green and cyan), with toremifene (space-filling, grey) bound at the GP1–GP2 interface of each subunit. (B) Docking site at the GP1/GP2 interface. (C) Binding interactions of toremifene (2) at the GP1–GP2 interface. (D) Predicted binding interactions of 33 at the GP1–GP2 interface. (PDB ID: 5JQ7)

In the unrefined crystallographic model, 2 is anchored by a network of polar and aromatic contacts (Figure 4C). The tertiary amine forms a salt bridge with E100, ring B engages R64 via a cation–π contact, and the ether linkage on ring B accepts a hydrogen bond from T519. Aromatic ring stacking is supplied by edge-to-face π–π contacts with Y517 by rings A and B (centroid distances = 5.32 Å and 5.24 Å, respectively), placing ring B between Y517 (π–π) and R64 (π–cation). Upon energy minimization of the complex, the polar interaction network is reinforced by the formation of an additional salt bridge with D522 while π–π stacking is modestly attenuated. After minimization, π–π contacts with Tyr517 persist for ring B (centroid = 5.39 Å, angle ≈ 51°) and ring A (centroid = 5.18 Å, angle ≈ 50°), with a weaker contribution from ring C (centroid = 5.65 Å, angle ≈ 61°), which is more solvent-exposed (Figure 4C).

Docking simulations of the urea derivative 33 returned poses that preserve the key interactions observed for 2, but with a pronounced shift toward polar stabilization (Figure 4D). Urea 33’s terminal amine, ionized at physiological pH, is predicted to form dual salt bridges with E100 and D522, and the urea nitrogen (N1) to donate a hydrogen bond to E100. These polar anchors drive compound 33 deeper into the polar portion of the fusion-loop pocket. Aromatic interactions are largely retained: edge-to-face π–π interactions with Y517 involve ring B (centroid distance = 5.39 Å, angle ≈ 54°) and ring C (centroid distance = 4.85 Å, angle ≈ 51°). In the top five ranked poses, ring B is consistently sandwiched between Y517 (π–π) and R64 (π–cation), and ring C further reinforces binding via additional π–π contacts. Ring A contributes to binding affinity through multiple hydrophobic contacts with residues lining the entrance to the pocket; its methoxy substituent remains solvent-exposed. Thus, urea 33 adopts a binding mode that combines deeper polar anchoring with preserved aromatic contacts relative to 2, increasing the likelihood of the ligand’s interference with prefusion-trimer stability (Figure 4D).

The predicted binding mode of 33 is strongly supported by the SAR described above (Figure 3A). In particular, the amine substituent on phenyl ring B is a primary determinant of potency, since modifications at this position produce large potency shifts (Table 1 and Figure 3A). Interestingly, MARV potency trends diverge from EBOV in ways that map onto structural differences between the two glycoproteins: MARV potency is relatively improved by bulkier ring C substituents, which suggests the presence of an accommodating hydrophobic subsite in the MARV GP that can accept larger aromatic or lipophilic groups. In addition, molecular docking studies showed comparable binding scores for both enantiomers of 41. The predicted binding mode suggests that the relatively spacious and flexible binding pocket can accommodate either enantiomer while maintaining the key interactions with Y517 and neighboring polar residues, despite small differences in binding geometry (Table S4, Supporting Information).

Site-directed mutagenesis studies corroborates the docking models. Substitution of Tyr517 with serine (Y517S), and to a lesser extent with isoleucine (Y517I; retains hydrophobicity but not aromaticity), reduced susceptibility to all inhibitors tested (Table 4), highlighting the central role of Y517 π–π interactions. The docking poses explain why newer derivatives show somewhat reduced dependence on Y517 relative to toremifene: by trading some aromatic proximity for additional polar contacts within the cavity, these derivatives retain high potency even when stacking is weakened.

Overall, the combined SAR, mutagenesis studies, and docking simulation affirm that the new derivatives engage conserved elements of the GP fusion-loop pocket, notably π–π stacking with Y517, cation–π interaction with R64, and salt-bridge formation with E100 and D522. While 2 relies predominantly on aromatic stacking, urea 33 penetrates deeper into the polar pocket towards the trimer center. This positioning perturbs GP trimer stability more efficiently and ultimately enhances modulation of viral entry with a higher potency.

Derivatives Display Specific Antifiloviral Activity

To assess the specificity of the antiviral mechanism, two representative derivatives were counter-screened for viral entry inhibition activity against pseudoviral H5N1 influenza (pH5N1), a member of the Orthomyoxoviridae family and vesicular stomatitis virus (pVSV); a member of the Rhabdoviridae family, for (Table 5). Both pseudoviruses utilize the same HIV core, as for the pEBOV, while expressing their corresponding viral surface glycoproteins.

Table 5.

Antiviral counter-screen against pVSV and pH5N1 by compounds 35 and 41.

Compound	pEBOV EC50 (μM)	pMARV EC50 (μM)	pVSV-G EC50 (μM)	pH5N1 EC50 (μM)
35	0.026 ± 0.005	0.44± 0.1	11.7 ± 0.2	24.8 ± 7.0
41	0.040 ± 0.002	0.42±0.03	10.2 ± 1.0	11.3 ± 0.6

The dihydronaphthalene derivative 35 showed markedly reduced inhibition against pVSV and pH5N1, with 450- and 950-fold decreases in activity, respectively, relative to pEBOV. Similarly, the tetrahydronaphthalene derivative 41 exhibited 250- and 280-fold lower activity against pVSV and pH5N1, respectively, compared to pEBOV. The selectivity against pMARV was less pronounced; however, the compounds still exhibited approximately 10- to 50-fold lower activity against pVSV and pH5N1. These findings confirm that the antiviral activity of these compounds is not related to the HIV core or to a common host factor shared among enveloped viruses, supporting a preferential antiviral effect against filoviruses rather than nonspecific viral inhibition.

Inhibition of Replication of Authentic EBOV and MARV

The antiviral activity of the most promising derivatives, selected based on high potency and favorable selectivity indices, was evaluated against infectious EBOV and MARV in BSL-4 settings (Table 6). Consistent with the pseudoviral assay results, the derivatives exhibited higher potency against EBOV than MARV. For EBOV, derivatives 31, 35, 36, and 41 achieved lower EC₅₀ values relative to the reference compound ridaifen-B (3) (Table 6). Testing against infectious MARV confirmed several promising candidates with submicromolar EC₅₀ values. The most active was 41 with an EC₅₀ of 0.38 μM (Table 6). These results are encouraging given the limited number of reported small-molecule MARV inhibitors, drawing attention to the potential of these compounds for further development.

Table 6.

Inhibitory activity of selected derivatives in the infectious EBOV and MARV assay

#	Hela CC50 (μM)	EBOV EC50a	SI EBOVb	MARV EC50a	SI MARVb
31	8.1±0.2	0.05±0.01	162	ndc	ndc
33	7.2±0.2	0.27±0.04	27	ndc	ndc
35	8.2±0.3	0.07±0.01	117	1.3 ± 1.6	6
36	8.3±0.4	0.03±0.00	330	ndc	ndc
41	4.6±0.1	0.04±0.01	120	0.38 ± 0.44	12
3	3.9±0.1	0.08±0.02	48	1.2 ± 0.2	3
Remdesivir	> 12	0.12±0.04	> 104	0.18 ± 0.07	70

^aDose–response curves were used to calculate EC₅₀ values using the four-parameter variable slope nonlinear regression analysis in GraphPad Prism 10.2.3. Data are presented as the mean EC₅₀ from three replicates ± standard deviation.

^bSelectivity indices (SI) are calculated as a ratio of CC₅₀/EC₅₀ where CC₅₀ is cytotoxicity measured in a Cell Titer-Glo Luminescent Cell Viability Assay.

^cnd: not determined.

Discrepancies were observed between the pseudoviral and infectious virus assay results, with all derivatives exhibiting higher EC₅₀ values in the infectious assay settings (Table 6). This difference likely reflects inherent variations in viral morphology and infectivity between the two assay systems, which was observed previously by our group.³⁸. Moreover, the observed discrepancy may be partly attributable to differences in endocytic uptake efficiency between A549 cells employed in pseudovirus assays and HeLa cells used in infectious assays. In addition, the intrinsic fluorescence of the 1,2-dihydronaphthalene derivatives may have interfered with signal readouts in the infectious assay; however, inspection of the raw imaging data and the low compound concentrations used indicate no significant contribution from compound autofluorescence. In contrast, the subset of compounds with a hydrogenated core 36–41, displayed consistent antiviral activities across both assays (Table 6). These tetrahydronaphthalene derivatives are non-fluorescent due to the hydrogenation of the central double bond which disrupts the conjugated π-system. Such discrepancies are not surprising as differences in cell lines and virion morphology between pseudotyped and infectious EBOV systems have been reported previously for other EBOV entry inhibitors.^35,38 Furthermore, the pseudotyped virus assay primarily measures inhibition of viral entry, whereas the infectious virus assay encompasses additional stages where GP can be targeted, including GP posttranslational processing and viral morphogenesis. Nevertheless, the observed potencies of these derivatives remain superior or comparable to previously reported EBOV entry inhibitors targeting GP-mediated fusion, both from our in-house library and from the literature.⁴¹

Despite this discrepancy, the data indicates a qualitative correlation between pEBOV and infectious virus readouts, although the relationship is not strictly linear. The pEBOV and infectious virus assays show a moderate positive correlation (R² ≈ 0.6), while a stronger correlation is observed for selectivity indices (R² ≈ 0.9), supporting consistent SAR trends across both systems (Figure S3, Supporting Information). EC₅₀-based stratification, separating compounds into active, medium, and inactive clusters, showed a good correlation between the mean EC₅₀ values between both infectious EBOV and pEBOV assays, supporting the use of pEBOV assay results for hit prioritization (Table S5 and Figure S4, Supporting Information). Overall, pEBOV alone cannot fully define potency but is highly informative for ranking compounds.

Metabolic Stability and CYP Inhibition Potential

To identify potential lead candidates suitable for in vivo efficacy studies, the metabolic stability of selected compounds was assessed in vitro. Structurally diverse analogues were evaluated for their stability in human and mouse plasma and liver microsomes at a drug concentration of 2 μM. The tested compounds showed good plasma stability in both human and mouse (Table 7). Most derivatives with promising antiviral potency e.g. 31, 33, 35, and 41, demonstrated high resistance to hepatic microsomal degradation in human liver microsomes (>85% drug remaining after 60 min co-incubation), indicating favorable metabolic profiles and potential for desirable pharmacokinetics. In mouse liver microsomes, 31, 33, 35, and 41, demonstrated moderate resistance to degradation (>50% drug remaining after 60 min co-incubation).

Table 7.

Metabolic Stability and CYP450 Inhibition of Selected Derivatives.

#	Metabolic Stability (% Remaining)a				CYP450 Isoforms Inhibition %b
	Plasma		Liver Microsome		1A2	2C9	2D6	3A4
	Human	Mouse	Human	Mouse
20	100	85	79	64	7	45	0	0
22	100	100	100	37	1	39	11	0
25	100	100	100	70	0	16	14	2
27	100	100	100	76	6	88	7	24
29	100	100	10	42	11	100	6	0
31	100	100	100	51	40	55	23	41
33	100	100	87	75	16	38	8	54
34	100	100	59	9	73	84	3	55
35	100	100	87	76	17	27	0	42
41	100	100	100	61	24	56	15	25
CYP# Positive Control c–f					86c	90d	90e	98f

^a% Compound remaining after 60 min as compared to t₀. Final compound concentration is 2 μM.

^b% Inhibition at 10 μM.

^cFurafylline,

^dSulfaphenazole,

^eQuinidine, and

^fKetoconazole were used as positive controls in the CYP450 inhibition assays. The results are from three replicates.

In parallel, CYP inhibition assays at 10 μM drug concentrations, revealed low to moderate inhibitory activity against four major CYP isoforms, CYP1A2, CYP2C9, CYP2D6, and CYP3A4 for most derivatives compared to reference standard CYP inhibitors (Table 7). Together, these findings suggest a low potential for drug–drug interactions and support continued evaluation of this compound series in preclinical in vivo studies.

Assessment of Potential Human Ether-à-go-go-Related Gene (hERG) Inhibition in vitro

The compounds with a pharmacophore of a basic amine and multiple aromatic rings are potential red flags as drug candidates, because of potential hERG potassium channel inhibition.⁶⁴ Potent hERG inhibition is a major safety concern because it can prolong the QT interval and cause life-threatening arrhythmias and cardiac arrest. The US FDA guidelines recommend assessing hERG inhibition potential in early drug discovery to minimize failure risk at advanced drug development stages.⁶⁵ The biochemical hERG inhibition assay gives a quick, cost-effective rough idea of the potential in vivo risk. The IC₅₀ of 41 in fluorescence polarization assay against hERG was estimated to be 2.93 μM (Figure 5). This can be regarded as a moderate risk, that may require structural optimization or monitoring.^66,67

Figure 5.

Assessment of hERG inhibition potential of compound 41 in vitro compared to reference standard E-4031.

The potency of 41 in EBOV inhibition (0.04 μM) is 73-fold higher than its hERG inhibition IC₅₀, but only ca. 14-fold when compared to its EC₉₀ (0.2 μM). To achieve viral suppression, a common starting point in in vivo studies is to target ca. 5× the EC₉₀ (0.6 μM), which in this case corresponds only to ca. 3-fold relative to the hERG IC₅₀ (Figure 5). Therefore, this compound should be advanced with caution for further optimization to reduce hERG safety margin.

In vivo Pharmacokinetic Parameters of Compound 41

Compound 41 displayed potent dual antifiloviral activity against both EBOV and MARV (0.04 and 0.4 μM), respectively, in the infectious assay (Table 6). It also demonstrated favorable in vitro metabolic stability in both human and mouse models and acceptable CYP inhibition profile (Table 7). Despite that the in vitro hERG inhibition of 41 is concerning, its potent antifiloviral potency prevails and makes it a promising starting point for further development. Prior to advancing 41 for in vivo antiviral efficacy studies, it was deemed necessary to get an insight into its in vivo pharmacokinetic behavior.

Therefore, rapid PK study was carried out in mouse animal model to estimate the in vivo pharmacokinetic parameters of 41. A single administration (bolus) of compound 41 was performed in male BALB/c mice at 5 mg/kg intravenously (i.v.) and 50 mg/kg orally (p.o.). From the i.v. treated mice group, the apparent volume of distribution (V_ss) was estimated to be 14 L/kg and the half-life (T_1/2) to be 6.8 h (Figure 6). The initial concentration following i.v. bolus dosing (C₀) was 1224 ng/mL, equivalent to 2.8 μM, which is 10-fold the EBOV EC₉₀ (0.2 μM), indicating that the therapeutic dose could potentially be reduced without compromising the therapeutic outcome. Peak plasma concentration following p.o. administration (T_max) was observed at 4 h (Figure 6). Unfortunately, the oral bioavaialability of 41 was only 6.5%, precluding its use via oral administration. But the favorable in vivo half-life and clearance would allow administration of the 41 i.v. for efficacy studies. Importantly, the projected oral C_max (0.48 μM) is below the hERG IC₅₀ (2.9 μM), giving an approximate 6-fold safety margin, which warrants further optimization to mitigate hERG inhibition potential and improve cardiac safety profile. These in vivo results validate the in vitro estimates of plasma and liver microsomal stability. No adverse effects or toxicity were observed on administration through either route.

Figure 6.

Pharmacokinetic Parameters for 41 in mice following i.v. bolus and p.o. bolus.

Conclusions and Future Directions

Current therapeutic options for Ebola virus disease (EVD) remain limited, emphasizing the unmet need for new antiviral strategies against a virus that is gradually expanding its geographic footprint and posing a sustained global health threat. In this study, we designed new EBOV entry inhibitors by cyclizing the triphenylethylene core of the SERM, ridaifen-B (3). Further structural modifications around the cyclic core were performed to improve the antiviral potency while minimizing cellular toxicity, initially assessed via pseudovirus assay screening. Structure–activity relationships (SAR) analyses provided insight into molecular features critical for antiviral activity. Mechanistic studies including mutagenesis and molecular modeling confirmed that these compounds act by targeting the EBOV glycoprotein (EBOV-GP). Among the derivatives, 1,4-dihydronaphthalenes 31, 35, and 36, as well as the tetrahydronaphthalene derivative 41 exhibited nanomolar EBOV entry inhibition in infectious EBOV assays, with compound 41 also demonstrating submicromolar potency against MARV. In vitro pharmacokinetic evaluations revealed good metabolic stability and a favorable CYP450 inhibition profile for most derivatives. Rapid in vivo pharmacokinetic study confirmed that compound 41 possesses a favorable PK profile following i.v. administration. Collectively, these results position compound 41 as a promising candidate for future in vivo efficacy trials, and a strong starting point for further optimization toward the development of orally bioavailable anti-Ebola therapies.

Despite its potent antifiloviral profile, compound 41 exhibits a narrow safety margin driven by hERG liability. Ongoing structural optimization efforts are focused on reducing hERG liability while preserving antiviral activity and improving the overall therapeutic index. Further optimization will also aim to refine the PK profile to enable oral administration, as well as assess blood-brain barrier permeability to ensure adequate central nervous system exposure for potential neurological manifestations associated with EBOV infection. Although ER activity was previously evaluated for lead compounds and the antiviral mechanism appears to be independent of classical ER modulation, potential ER-related off-target effects will be considered during future lead optimization and preclinical development. Although molecular modeling suggests that both enantiomers of 41 can adopt similar binding modes within the EBOV glycoprotein binding pocket, potential enantiomer-specific differences in pharmacological, pharmacokinetic, or toxicological properties cannot be excluded. Future studies will therefore focus on further stereochemical characterization and exploration of analogs with reduced stereochemical complexity while maintaining antiviral potency. Although this work focuses primarily on EBOV, the main scaffold may also provide a foundation for the development of broad-spectrum antifiloviral agents, highlighting its potential impact in combating emerging filovirus threats.

Experimental Section

General Chemistry Methods

All reagents and solvents were purchased from commercial suppliers and used without further purification. Air- and moisture-sensitive reactions were conducted in flame-dried glassware under an inert atmosphere of argon. All synthesized final products were subjected to purification through flash chromatography and, when required, HPLC. Compound identities were confirmed through ¹H, ¹³C, and ¹⁹F NMRs as well as HRMS. Purity level, indicated by analytical HPLC, was confirmed to be at least 95% prior to their submission for biological evaluations. Hydrogenation reactions were performed in a Parr shaker apparatus. Reaction progress was monitored by thin-layer chromatography (TLC) using Macherey-Nagel silica gel 60 F254 precoated glass plates, visualized under UV light (254 and 360 nm) or after special staining. Products were purified on a Teledyne ISCO CombiFlash R_f⁺ system using silica prefilled Teledyne RediSep cartridges, with UV detection at 254 nm and 280 nm.

¹H, ¹³C, and ¹⁹F NMR spectra were recorded on a Bruker AVANCE AVIII-400 HD NMR spectrometer, equipped with a 5 mm room-temperature SmartProbe at 400, 100, and 376 MHz, respectively. Spectra for compounds 12, 14, 18, 20, and 36–38 were recorded on a Bruker AVANCE AVIII-600. Chemical shifts (δ) are reported in parts per million (ppm), relative to the residual solvent peak: CDCl₃ (¹H = 7.26 ppm, ¹³C = 77.2 ppm), CD₃OD (¹H = 3.31 ppm, ¹³C = 49.0 ppm), CD₃COCD₃ (¹H = 2.05 ppm, ¹³C = 29.8 ppm), (CD₃)₂SO (¹H = 2.50 ppm, ¹³C = 39.5 ppm), and CD₃CN (¹H = 1.94 ppm, ¹³C = 1.3 ppm). Coupling constants are reported in Hertz (Hz). Signal multiplicities are denoted as: s, singlet; br s, broad singlet; d, doublet; t, triplet; q, quartet; and m, multiplet.

Either semi-preparative or preparative reversed-phase (RP) HPLC was used to purify final compounds. Semi-preparative HPLC was performed using a Phenomenex Luna 5 μm C₁₈-RP column (250 × 10 mm) with a flow rate of 4 mL/min and a gradient of 20–95% acetonitrile in water containing 0.1% formic acid. Detection was performed at 254 and 280 nm using a Shimadzu SPD-20A detector. Preparative HPLC was conducted using an ACE 5 AQ C₁₈-RP column (150 × 30 mm) with a flow rate of 17 mL/min, employing a 20–100% acetonitrile in water gradient with 0.1% formic acid. The separation was monitored at 254 and 280 nm using a Shimadzu SPD-10A VP detector. All tested compounds were confirmed to be ≥ 95% pure by analytical HPLC on a Shimadzu LC-20AB system equipped with Ace 5 C₈-RP column (150 × 4.6 mm) and a Shimadzu SPD-20A detector at 254 and 280 nm, eluting with 20–95% acetonitrile in water gradient with 0.1% formic over 15 min.

LC–MS analyses were performed on an Agilent InfinityLab LC/MSD instrument equipped with a ZORBAX SB-C₁₈-RP column (1.8 μm, 4.6 × 50 mm) and a photodiode array UV detector, using a 10–95% acetonitrile in water gradient with 0.1% formic acid. High-resolution mass spectrometry (HRMS) of the final compounds was conducted at the Mass Spectrometry Laboratory, University of Illinois at Urbana-Champaign, using a Waters^™ Synapt G2-Si mass spectrometer. Observed molecular weights of the compounds were within 5 ppm of the calculated values.

N-(4-Bromophenyl)-1-methylpiperidine-4-carboxamide (8).

In a round-bottom flask, 1-methylpiperidine-4-carboxylic acid (0.65 g, 4.6 mmol) was suspended in dichloromethane (15 mL). N,N-diisopropylethylamine (2.4 mL, 14 mmol) and benzotriazol-1-yloxytripyrrolidinophosphonium hexafluorophosphate (PyBOP) (2.6 g, 5.0 mmol) were added. The mixture was stirred at room temperature for 30 min until a clear solution was formed. 4-bromoaniline (0.79 g, 4.6 mmol) was then added, and the reaction was stirred for 18 h at room temperature. The reaction mixture was adsorbed to celite and purified by flash chromatography on a 40 g silica gel column, eluting with a gradient of 0−20% MeOH/DCM. The product (8) was obtained as white solid (950 mg, 70%). ¹H NMR (400 MHz, DMSO-d₆) δ 10.16 (s, 1H), 7.57 (d, J = 8.9 Hz, 2H), 7.49 (d, J = 8.9 Hz, 2H), 3.49 (d, J = 11.4 Hz, 2H), 2.97 (t, J = 12.0 Hz, 2H), 2.80 (s, 3H), 2.56 (t, J = 12.0 Hz, 1H), 2.03 (d, J = 13.2 Hz, 2H), 1.82 (dd, J = 13.7, 11.5 Hz, 2H). ¹³C NMR (101 MHz, DMSO-d₆) δ 172.4, 138.8, 132.0, 121.6, 115.4, 55.4, 53.2, 43.1, 26.4. HRMS(ESI-TOF) for C₁₃H₁₈N₂OBr m/z [M+H]⁺ calcd: 297.0603, found: 297.0602. The bromine isotope pattern is observed with an m/z of 279.0602 [M(⁷⁹Br)+H]⁺ and 299.0583 [M(⁸¹Br)+H]⁺ in a 1:1 ratio.

1-Methyl-N-(4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)phenyl)piperidine-4-carboxamide (9).

Miyaura Borylation Reaction: In a flame-dried, argon-flushed round bottom flask, a mixture of 9 (75 mg, 0.25 mmol), 4,4,4’,4’,5,5,5’,5’-octamethyl-2,2’-bi(1,3,2-dioxaborolane) (90 mg, 0.35 mmol), and [1,1´-bis(diphenylphosphino)ferrocene]dichloropalladium(II) (Pd(dppf)Cl₂) (10 mg, 2 mol%) was suspended in anhydrous degassed dioxane (2 mL). An aqueous solution of 1.5 M cesium carbonate (0.35 mL, 0.5 mmol) was added dropwise, and the reaction mixture was refluxed for 2 h at 110 °C under argon. After cooling to room temperature, the mixture was filtered and concentrated under vacuum to yield the crude pinacol ester, which was directly used in the next step without further purification.

6-Methoxy-2-phenyl-3,4-dihydronaphthalen-1-yl trifluoromethanesulfonate (11).

In a flame-dried, argon-flushed round-bottom flask, 6-methoxy-2-phenyl-3,4-dihydronaphthalen-1(2H)-one (2.00 g, 7.93 mmol) was suspended in anhydrous THF (30 mL) and stirred at −78 °C. A 1.0 M solution of lithium bis(trimethylsilyl)amide in THF (9.12 mL) was added dropwise, and the mixture was stirred at −78 °C for 45 min. A solution of 1,1,1-trifluoro-N-phenyl-N-((trifluoromethyl)sulfonyl)methanesulfonamide (2.86 g, 8.01 mmol) in anhydrous THF (4 mL) was added dropwise to the reaction mixture at −78 °C. The reaction mixture was allowed to warm to room temperature and stirred for 3 h before quenching with 10 mL saturated ammonium chloride solution. The reaction mixture was extracted with ethyl acetate (3 × 30 mL), the organic layers were combined, washed with brine (10 mL), dried over anhydrous sodium sulfate, and concentrated in vacuo to give a crude brown oil that was purified via flash chromatography on a 40 g silica column eluting with 0−30% EtOAc/hexanes. The desired product was obtained as a white waxy solid (2.30 g, 75%), TLC (25% EtOAc/hexanes) R_f = 0.58. ¹H NMR (400 MHz, CDCl₃) δ 7.52 – 7.42 (m, 4H), 7.42 – 7.34 (m, 2H), 6.85 (dd, J = 8.6, 2.4 Hz, 1H), 6.81 (d, J = 2.6 Hz, 1H), 3.87 (s, 3H), 3.01 (t, J = 7.9 Hz, 2H), 2.80 (t, J = 7.9 Hz, 2H). ¹³C NMR (101 MHz, CDCl₃) δ 160.0, 141.2, 137.8, 136.3, 129.1, 129.0, 128.4, 128.3, 123.5, 123.0, 113.9, 111.5, 55.4, 30.1, 28.2. ¹⁹F NMR (376 MHz, CDCl₃) δ −74.56. LCMS (ESI) for C₁₈H₁₅F₃O₄S m/z [M+H]⁺ = 385.37, [M+Na]⁺ = 407.36.

7-Methoxy-4-(4-methoxyphenyl)-3-phenyl-1,2-dihydronaphthalene (12).

In a flame-dried, argon-flushed round bottom flask, a mixture of 6-methoxy-2-phenyl-3,4-dihydronaphthalen-1-yl trifluoromethanesulfonate (11) (50 mg, 0.13 mmol), 2-(4-methoxyphenyl)-4,4,5,5-tetramethyl-1,3,2-dioxaborolane (34 mg, 0.13 mmol), and [1,1′-bis(diphenylphosphino)ferrocene]dichloropalladium(II) (3.8 mg, 5 mol%) was suspended in anhydrous dioxane (2 mL). An aqueous solution of 1.5 M dicesium carbonate (0.26 mL, 0.4 mmol) was added dropwise, and the reaction mixture was stirred for 18 h at 90 °C under an argon atmosphere. Water (3 mL) and ethyl acetate (6 mL) were added, and the organic phase was separated, dried over anhydrous sulfate, filtered, and concentrated under reduced pressure. The crude mixture was adsorbed to celite and purified by flash chromatography on a 4 g silica column eluting with dichloromethane to give 12 as a white solid (30 mg, 67%). TLC (30% EtOAc/hexanes) R_f = 0.5. ¹H NMR (600 MHz, CDCl₃) δ 7.16 – 7.09 (m, 2H), 7.07 – 7.01 (m, 4H), 6.99 – 6.93 (m, 2H), 6.80 – 6.76 (m, 3H), 6.74 (d, J = 8.5 Hz, 1H), 6.60 (dd, J = 8.6, 2.8 Hz, 1H), 3.81 (s, 2H), 3.79 (s, 2H), 2.98 – 2.92 (m, 2H), 2.79 (dd, J = 8.9, 6.5 Hz, 2H). ¹³C NMR (151 MHz, CDCl₃) δ 158.5, 158.2, 143.4, 137.8, 134.9, 134.4, 132.3, 132.2, 130.6, 128.4, 127.7, 127.6, 125.7, 113.5, 113.3, 110.9, 55.4, 55.2, 30.9, 29.1. HRMS (ESI-TOF) for C₂₄H₂₂O₂ m/z [M+H]⁺ calcd: 342.1620, found: 342.1619.

5-(4-Hydroxyphenyl)-6-phenyl-7,8-dihydronaphthalen-2-ol (13).

In a flame-dried, argon-flushed round bottom flask, 12 (0.40 g, 1.17 mmol) was dissolved in anhydrous dichloromethane (20 mL) and cooled in an ice-bath. Next, 11 mL of a 1 M boron tribromide (BBr₃) solution in dichloromethane was added. After completion, as indicated by TLC, reaction was quenched with an aqueous sodium bicarbonate solution and pH was adjusted to ca. 8. The aqueous layer was extracted with ethyl acetate (3 × 20 mL). The combined organic extracts were concentrated, washed with brine, and dried over anhydrous sodium sulfate. The crude mixture was purified on a 24-g silica column eluting 0−50% EtOAc/hexanes. The title compound 13 was isolated as an off-white solid (0.23 g, 63%), TLC (30% EtOAc/hexanes) R_f = 0.24. ¹H NMR (400 MHz, CDCl₃) δ 7.18 – 7.10 (m, 2H), 7.09 – 7.00 (m, 3H), 6.96 – 6.91 (m, 2H), 6.76 – 6.68 (m, 4H), 6.54 (dd, J = 8.5, 2.7 Hz, 1H), 4.68 (s, 1H), 4.62 (s, 1H), 2.99 – 2.90 (dd, J = 9.05, 604, 2H), 2.79 (dd, J = 9.3, 6.2 Hz, 2H) HRMS(ESI-TOF) for C₂₂H₁₉O₂ m/z [M+H]⁺ calcd: 315.1385, found: 315.1377.

2-Chloro-5-(6-methoxy-2-phenyl-3,4-dihydronaphthalen-1-yl)-1-methyl-1H-imidazole (14).

In a flame-dried, argon-flushed round bottom flask, a mixture of 11 (134 mg, 0.35 mmol), 2-chloro-1-methyl-5-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)-1H-imidazole (70 mg, 0.29 mmol), and [1,1′-bis(diphenylphosphino)ferrocene]dichloropalladium(II) (10 mg, 5 mol%) was suspended in anhydrous dioxane (3.5 mL). An aqueous solution of 1.5 M cesium carbonate (0.58 mL, 0.98 mmol) was added dropwise, and the reaction mixture was stirred for 18 h at 90 °C under an argon atmosphere. Water (3 mL) and ethyl acetate (6 mL) were added, and the organic phase was separated, dried over anhydrous sulfate, filtered, and concentrated under reduced pressure. The crude mixture was adsorbed to celite and purified by flash chromatography on a 4-g silica column eluting with 0−50% EtOAc/hexanes to give 14 as a white solid (70 mg, 69%). TLC (25% EtOAc/hexanes) R_f = 0.5. ¹H NMR (600 MHz, CDCl₃) δ 7.24 – 7.20 (m, 2H), 7.18 – 7.13 (m, 1H), 7.06 (d, J = 7.2 Hz, 2H), 6.80 – 6.76 (m, 2H), 6.73 (s, 1H), 6.66 (dd, J = 8.6, 2.6 Hz, 1H), 3.83 (s, 3H), 3.07 (s, 3H), 3.02 – 2.91 (m, 2H), 2.91 – 2.78 (m, 2H). ¹³C NMR (151 MHz, CDCl₃) δ 159.2, 141.9, 140.6, 137.1, 131.7, 131.5, 128.5, 128.5, 128.3, 127.2, 127.1, 126.7, 123.2, 113.8, 111.4, 55.4, 31.3, 30.5, 28.6. HRMS(ESI-TOF) for C₂₁H₂₀N₂OCl m/z [M+H]⁺ calcd: 351.1264, found: 351.1264.

2-(6-Methoxy-2-phenyl-3,4-dihydronaphthalen-1-yl)-9H-carbazole (15).

In a flame-dried, argon-flushed round bottom flask, a mixture of 11 (100 mg, 0.26 mmol), 2-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)-9H-carbazole (76 mg, 0.26 mmol), and [1,1′-bis(diphenylphosphino)ferrocene]dichloropalladium(II) (10 mg, 5 mol%) was suspended in anhydrous dioxane (2 mL). An aqueous solution of 1.5 M cesium carbonate (0.3 mL, 0.78 mmol) was added dropwise, and the reaction mixture was stirred for 18 h at 90 °C under argon. Water (2 mL) and ethyl acetate (4 mL) were added, and the organic phase was separated, dried over anhydrous sulfate, filtered, and concentrated under reduced pressure. The crude mixture was adsorbed to celite and purified by flash chromatography on a 4 g silica column eluting with 0−50% EtOAc/hexanes to give 15 as a white solid (60 mg, 57%). TLC (25% EtOAc/hexanes) R_f = 0.57. ¹H NMR (400 MHz, CDCl₃) δ 7.99 – 7.91 (m, 2H), 7.85 – 7.80 (m, 1H), 7.45 – 7.35 (m, 2H), 7.24 (d, J = 0.6 Hz, 2H), 7.19 (ddd, J = 8.0, 6.4, 1.8 Hz, 1H), 7.10 – 6.93 (m, 6H), 6.82 (d, J = 2.7 Hz, 1H), 6.76 (d, J = 8.6 Hz, 1H), 6.57 (dd, J = 8.6, 2.7 Hz, 1H), 3.81 (s, 3H), 3.02 (t, J = 7.7 Hz, 2H), 2.85 (t, J = 7.6 Hz, 2H). ¹³C NMR (101 MHz, CDCl₃) δ 158.6, 143.5, 139.9, 138.4, 137.9, 135.8, 134.5, 131.3, 129.6, 128.5, 127.9, 127.6, 125.9, 125.7, 123.6, 123.4, 122.9, 120.5, 119.5, 113.3, 110.9, 110.7, 110.3, 55.4, 31.0, 29.3. HRMS(ESI-TOF) for C₂₉H₂₄NO m/z [M+H]⁺ calcd: 402.1858, found: 402.1853.

1-(3-(6-Methoxy-2-phenyl-3,4-dihydronaphthalen-1-yl)benzyl)-4-methylpiperazine (16).

In a flame-dried, argon-flushed round bottom flask, a mixture of 11 (100 mg, 0.26 mmol), 1-methyl-4-(3-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)benzyl)piperazine (82 mg, 0.26 mmol), and [1,1′-bis(diphenylphosphino)ferrocene]dichloropalladium(II) (10 mg, 5 mol%) was suspended in anhydrous dioxane (2 mL). An aqueous solution of 1.5 M cesium carbonate (0.3 mL, 0.78 mmol) was added dropwise, and the reaction mixture was stirred for 18 h at 90 °C under argon. Water (2 mL) and ethyl acetate (4 mL) were added, and the organic phase was separated, dried over anhydrous sulfate, filtered, and concentrated under reduced pressure. The crude mixture was adsorbed to celite and purified by flash chromatography on a 4 g silica column eluting with 0–20% MeOH/DCM to give 16 as a white solid (50 mg, 45%). TLC (10% MeOH/DCM) R_f = 0.3. ¹H NMR (400 MHz, CDCl₃) δ 7.21 (t, J = 7.5 Hz, 1H), 7.14 (dt, J = 7.7, 1.4 Hz, 1H), 7.10 – 6.96 (m, 6H), 6.94 (t, J = 1.8 Hz, 1H), 6.78 (d, J = 2.6 Hz, 1H), 6.70 (d, J = 8.6 Hz, 1H), 6.58 (dd, J = 8.6, 2.7 Hz, 1H), 3.80 (s, 3H), 3.37 (s, 2H), 2.96 (t, J = 7.7 Hz, 2H), 2.79 (t, J = 7.3 Hz, 2H), 2.47 – 2.19 (m, 11H). ¹³C NMR (101 MHz, CDCl₃) δ 158.5, 143.2, 139.8, 137.7, 137.6, 135.3, 134.6, 132.2, 130.2, 129.7, 128.4, 127.8, 127.5, 127.4, 125.6, 113.2, 110.9, 62.8, 55.3, 55.1, 52.7, 46.0, 30.7, 29.0. HRMS(ESI-TOF) for C₂₉H₃₃N₂O m/z [M+H]⁺ calcd: 425.2593, found: 425.2587.

(3-(6-Methoxy-2-phenyl-3,4-dihydronaphthalen-1-yl)phenyl)(4-methylpiperazin-1-yl)methanone (17)

In a flame-dried, argon-flushed round bottom flask, a mixture of 11 (100 mg, 0.26 mmol), (4-methylpiperazin-1-yl)(3-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)phenyl)methanone (86 mg, 0.26 mmol), and [1,1′-bis(diphenylphosphino)ferrocene]dichloropalladium(II) (10 mg, 5 mol%) was suspended in anhydrous dioxane (2 mL). An aqueous solution of 1.5 M cesium carbonate (0.3 mL, 0.78 mmol) was added dropwise, and the reaction mixture was stirred for 18 h at 90 °C under argon. Water (2 mL) and ethyl acetate (4 mL) were added, and the organic phase was separated, dried over anhydrous sulfate, filtered, and concentrated under reduced pressure. The crude mixture was adsorbed to celite and purified by flash chromatography on a 4 g silica column eluting with 0–20% MeOH/DCM to give 17 as a light brown solid (60 mg, 53%). TLC (10% MeOH/DCM) R_f = 0.3. ¹H NMR (400 MHz, CDCl₃) δ 7.39 – 7.23 (m, 3H), 7.13 – 6.96 (m, 5H), 6.90 (t, J = 1.4 Hz, 1H), 6.77 (d, J = 2.6 Hz, 1H), 6.66 (d, J = 8.6 Hz, 1H), 6.56 (dd, J = 8.6, 2.7 Hz, 1H), 3.79 (s, 3H), 3.76 – 3.55 (m, 2H), 2.95 (br s, 2H), 2.78 (br s, 4H), 2.37 (br s, 2H), 2.26 (s, 3H), 2.02 (br s, 2H), 1.22 (s, 2H). ¹³C NMR (101 MHz, CDCl₃) δ 170.1, 158.7, 143.1, 139.9, 137.5, 135.7, 135.2, 134.8, 132.3, 129.63, 129.56, 128.5, 127.8, 127.5, 126.0, 125.8, 113.5, 110.8, 55.3, 54.7, 47.1, 46.0, 41.9, 30.7, 28.9, 24.9. HRMS(ESI-TOF) for C₂₉H₃₁N₂O₂ m/z [M+H]⁺ calcd: 439.2386, found: 439.2378.

tert-Butyl (4-(6-methoxy-2-phenyl-3,4-dihydronaphthalen-1-yl)phenyl)carbamate (18).

In a flame-dried, argon-flushed round bottom flask, a mixture of 11 (150 mg, 0.39 mmol), tert-butyl (4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)phenyl)carbamate (149 mg, 0.47 mmol), and [1,1′-bis(diphenylphosphino)ferrocene]dichloropalladium(II) (15 mg, 5 mol%) was suspended in anhydrous dioxane (3 mL). An aqueous solution of 1.5 M cesium carbonate (0.5 mL, 0.78 mmol) was added dropwise, and the reaction mixture was stirred for 18 h at 90 °C under an argon atmosphere. Water (3 mL) and ethyl acetate (6 mL) were added, and the organic phase was separated, dried over anhydrous sulfate, filtered, and concentrated under reduced pressure. The crude mixture was adsorbed to celite and purified by flash chromatography on a 12 g silica column eluting with dichloromethane to give 18 as a white solid (135 mg, 81%). TLC (DCM) R_f = 0.6. ¹H NMR (600 MHz, CDCl₃) δ 7.23 (d, J = 8.1 Hz, 2H), 7.14 – 7.09 (m, 2H), 7.07 – 7.01 (m, 3H), 7.01 – 6.96 (m, 2H), 6.78 (d, J = 2.6 Hz, 1H), 6.73 (d, J = 8.6 Hz, 1H), 6.59 (dd, J = 8.6, 2.8 Hz, 1H), 6.45 (s, 1H), 3.81 (s, 3H), 2.98 – 2.92 (m, 2H), 2.79 (dd, J = 9.0, 6.5 Hz, 2H), 1.53 (s, 9H). ¹³C NMR (151 MHz, CDCl₃) δ 158.6, 152.8, 143.3, 137.8, 136.8, 134.8, 134.7, 134.6, 131.9, 130.4, 128.4, 127.7, 127.6, 125.9, 118.1, 113.4, 110.9, 80.9, 55.4, 30.9, 29.1, 28.5. HRMS (ESI-TOF) for C₂₈H₂₉NO₃ m/z [M+H]⁺ calcd: 427.2147, found: 427.2139.

4-(6-Methoxy-2-phenyl-3,4-dihydronaphthalen-1-yl)-N,N-dimethylaniline (19).

In a flame-dried, argon-flushed round bottom flask, a mixture of 11 (75 mg, 0.19 mmol), N,N-dimethyl-4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)aniline (48 mg, 0.19 mmol), and [1,1′-bis(diphenylphosphino)ferrocene]dichloropalladium(II) (10 mg, 5 mol%) was suspended in anhydrous dioxane (1.5 mL). An aqueous solution of 1.5 M cesium carbonate (0.26 mL, 0.39 mmol) was added dropwise, and the reaction mixture was stirred for 18 h at 90 °C under an argon atmosphere. Water (2 mL) and ethyl acetate (4 mL) were added, and the organic phase was separated, dried over anhydrous sulfate, filtered, and concentrated under reduced pressure. The crude mixture was adsorbed to celite and purified by flash chromatography on a 4-g silica column eluting with 0−30% EtOAc/hexanes to give 19 as a colorless oil (40 mg, 58%). TLC (10% EtOAc/hexanes) R_f = 0.5. ¹H NMR (400 MHz, CDCl₃) δ 7.16 – 7.10 (m, 1H), 7.10 – 7.01 (m, 1H), 6.92 (dd, J = 19.5, 7.6 Hz, 1H), 6.81 (dd, J = 9.4, 5.8 Hz, 1H), 6.70 – 6.64 (m, 0H), 6.62 (dd, J = 8.6, 2.7 Hz, 1H), 3.83 (s, 1H), 3.00 – 2.92 (m, 3H), 2.85 – 2.75 (m, 1H). ¹³C NMR (101 MHz, CDCl₃) δ 158.5, 143.8, 138.0, 135.3, 133.8, 132.3, 132.0, 131.0, 130.3, 128.5, 127.8, 127.7, 125.6, 113.2, 112.3, 110.9, 77.5, 77.2, 76.8, 55.4, 40.8, 31.0, 29.2. HRMS(ESI-TOF) for C₂₅H₂₆NO m/z [M+H]⁺ calcd: 356.2014, found: 356.2014.

1-(4-(6-Methoxy-2-phenyl-3,4-dihydronaphthalen-1-yl)phenyl)pyrrolidine (20).

In a flame-dried, argon-flushed round bottom flask, a mixture of 11 (125 mg, 0.33 mmol), 1-(4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)phenyl)pyrrolidine (89 mg, 0.33 mmol), and [1,1′-bis(diphenylphosphino)ferrocene]dichloropalladium(II) (12 mg, 5 mol%) was suspended in anhydrous dioxane (3 mL). An aqueous solution of 1.5 M cesium carbonate (0.65 mL, 0.98 mmol) was added dropwise, and the reaction mixture was stirred for 18 h at 90 °C under an argon atmosphere. Water (3 mL) and ethyl acetate (6 mL) were added, and the organic phase was separated, dried over anhydrous sulfate, filtered, and concentrated under reduced pressure. The crude mixture was adsorbed to celite and purified by flash chromatography on a 4-g silica column eluting with 0−20% MeOH/DCM to give 20 as a brown solid (80 mg, 65%). TLC (10% MeOH/DCM) R_f = 0.5. Further purification by precipitation from a concentrated methanol solution yielded 20 as a beige solid (15 mg). ¹H NMR (600 MHz, CDCl₃) δ 7.16 – 7.10 (m, 2H), 7.09 – 7.01 (m, 3H), 6.90 (d, J = 7.7 Hz, 2H), 6.83 (d, J = 8.4 Hz, 1H), 6.77 (d, J = 2.6 Hz, 1H), 6.59 (dd, J = 8.6, 2.8 Hz, 1H), 6.44 (d, J = 7.0 Hz, 2H), 3.81 (s, 3H), 3.27 (s, 4H), 2.97 – 2.91 (m, 2H), 2.78 (dd, J = 9.0, 6.5 Hz, 2H), 2.00 (s, 4H). ¹³C NMR (151 MHz, CDCl₃) δ 158.4, 146.5, 143.9, 138.0, 135.4, 133.4, 132.1, 131.1, 130.3, 128.5, 127.8, 127.6, 126.6, 125.4, 113.1, 111.3, 110.8, 55.4, 47.7, 31.0, 29.2, 25.6. HRMS(ESI-TOF) for C₂₇H₂₈NO m/z [M+H]⁺ calcd: 382.2171, found: 382.2165.

1-(4-(6-Methoxy-2-phenyl-3,4-dihydronaphthalen-1-yl)phenyl)piperidine (21).

In a flame-dried, argon-flushed round bottom flask, a mixture of 11 (100 mg, 0.26 mmol), 1-(4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)phenyl)piperidine (75 mg, 0.26 mmol), and [1,1′-bis(diphenylphosphino)ferrocene]dichloropalladium(II) (10 mg, 5 mol%) was suspended in anhydrous dioxane (2 mL). An aqueous solution of 1.5 M cesium carbonate (0.35 mL, 0.52 mmol) was added dropwise, and the reaction mixture was stirred for 18 h at 90 °C under an argon atmosphere. Water (2 mL) and ethyl acetate (4 mL) were added, and the organic phase was separated, dried over anhydrous sulfate, filtered, and concentrated under reduced pressure. The crude mixture was adsorbed to celite and purified by flash chromatography on a 4-g silica column eluting with 0−30% EtOAc/hexanes to give 21 as a colorless oil (70 mg, 68%). TLC (10% MeOH/DCM) R_f = 0.4. ¹H NMR (400 MHz, CDCl₃) δ 7.17 – 7.08 (m, 2H), 7.06 (d, J = 6.7 Hz, 3H), 7.01 – 6.84 (m, 2H), 6.84 – 6.72 (m, 4H), 6.61 (dd, J = 8.6, 2.7 Hz, 1H), 3.83 (s, 3H), 3.16 (s, 4H), 3.00 – 2.91 (m, 2H), 2.80 (dd, J = 9.2, 6.2 Hz, 2H), 1.73 (s, 4H), 1.59 (s, 2H). ¹³C NMR (101 MHz, CDCl₃) δ 158.5, 143.7, 138.0, 135.2, 134.0, 131.9, 130.9, 128.5, 127.8, 127.7, 127.1, 125.6, 116.8, 116.0, 113.2, 110.9, 55.4, 50.8, 50.7, 31.0, 29.2, 26.1, 26.0, 24.5. HRMS(ESI-TOF) for C₂₈H₃₀NO m/z [M+H]⁺ calcd: 396.2327, found: 396.2323.

1-(4-(6-Methoxy-2-phenyl-3,4-dihydronaphthalen-1-yl)phenyl)-4-methylpiperazine (22).

In a flame-dried, argon-flushed round bottom flask, a mixture of 11 (100 mg, 0.26 mmol), 1-methyl-4-(4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)phenyl)piperazine (79 mg, 0.26 mmol), and [1,1′-bis(diphenylphosphino)ferrocene]dichloropalladium(II) (10 mg, 5 mol%) was suspended in anhydrous dioxane (2 mL). An aqueous solution of 1.5 M cesium carbonate (0.35 mL, 0.52 mmol) was added dropwise, and the reaction mixture was stirred for 18 h at 90 °C under an argon atmosphere. Water (2 mL) and ethyl acetate (4 mL) were added, and the organic phase was separated, dried over anhydrous sulfate, filtered, and concentrated under reduced pressure. The crude mixture was adsorbed to celite and purified by flash chromatography on a 4-g silica column eluting with 0−20% MeOH/DCM to give 22 as a white solid (70 mg, 65%). TLC (5% MeOH/DCM) R_f = 0.22. ¹H NMR (400 MHz, CDCl₃) δ 7.13 (t, J = 7.3 Hz, 2H), 7.10 – 6.88 (m, 5H), 6.79 (d, J = 8.3 Hz, 3H), 6.74 (d, J = 8.6 Hz, 1H), 6.62 (dd, J = 8.6, 2.2 Hz, 1H), 3.82 (s, 3H), 3.41 (d, J = 4.8 Hz, 4H), 3.14 (d, J = 5.2 Hz, 4H), 2.96 (t, J = 7.6 Hz, 2H), 2.84 – 2.72 (m, 5H). ¹³C NMR (101 MHz, CDCl₃) δ 158.5, 148.1, 143.2, 137.8, 134.6, 134.5, 132.8, 132.1, 130.4, 128.3, 127.6, 127.5, 125.7, 116.3, 113.2, 110.8,, 55.3, 54.6, 47.6, 44.8, 30.8, 29.0. HRMS(ESI-TOF) for C₂₈H₃₁N₂O m/z [M+H]⁺ calcd: 411.2436, found: 411.2434.

1-(4-(6-Methoxy-2-phenyl-3,4-dihydronaphthalen-1-yl)benzyl)-4-methylpiperazine (23).

In a flame-dried, argon-flushed round bottom flask, a mixture of 11 (100 mg, 0.26 mmol), 1-methyl-4-(4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)benzyl)piperazine (82 mg, 0.26 mmol), and [1,1′-bis(diphenylphosphino)ferrocene]dichloropalladium(II) (10 mg, 5 mol%) was suspended in anhydrous dioxane (2 mL). An aqueous solution of 1.5 M cesium carbonate (0.3 mL, 0.78 mmol) was added dropwise, and the reaction mixture was stirred for 18 h at 90 °C under argon. Water (2 mL) and ethyl acetate (4 mL) were added, and the organic phase was separated, dried over anhydrous sulfate, filtered, and concentrated under reduced pressure. The crude mixture was adsorbed to celite and purified by flash chromatography on a 4 g silica column eluting with 0−20% MeOH/DCM to give 23 as a white solid (50 mg, 45%). TLC (10% MeOH/DCM) R_f = 0.3. ¹H NMR (400 MHz, CDCl₃) δ 7.16 (d, J = 7.7 Hz, 2H), 7.10 – 6.96 (m, 7H), 6.78 (d, J = 2.4 Hz, 1H), 6.72 (d, J = 8.5 Hz, 1H), 6.59 (dd, J = 8.6, 2.6 Hz, 1H), 3.80 (s, 3H), 3.48 (s, 2H), 2.96 (dd, J = 9.1, 5.9 Hz, 2H), 2.84 – 2.75 (m, 2H), 2.48 (s, 8H), 2.31 (s, 3H). ¹³C NMR (101 MHz, CDCl₃) δ 158.6, 143.3, 138.8, 137.7, 136.2, 135.3, 134.6, 131.1, 130.4, 129.0, 128.4, 127.6, 127.6, 125.8, 113.3, 111.0.96, 62.96, 55.40, 55.24, 53.11, 46.14, 30.81, 29.12. HRMS(ESI-TOF) for C₂₉H₃₃N₂O m/z [M+H]⁺ calcd: 425.2593, found: 425.2592.

4-(4-(6-Methoxy-2-phenyl-3,4-dihydronaphthalen-1-yl)benzyl)morpholine (24).

In a flame-dried, argon-flushed round bottom flask, a mixture of 11 (100 mg, 0.26 mmol), 4-(4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)benzyl)morpholine (79 mg, 0.26 mmol), and [1,1′-bis(diphenylphosphino)ferrocene]dichloropalladium(II) (10 mg, 5 mol%) was suspended in anhydrous dioxane (2 mL). An aqueous solution of 1.5 M cesium carbonate (0.3 mL, 0.78 mmol) was added dropwise, and the reaction mixture was stirred for 18 h at 90 °C under argon. Water (2 mL) and ethyl acetate (4 mL) were added, and the organic phase was separated, dried over anhydrous sulfate, filtered, and concentrated under reduced pressure. The crude mixture was adsorbed to celite and purified by flash chromatography on a 4 g silica column eluting with 20−70% EtOAc/hexanes to give 24 as a brown oil (70 mg, 65%). TLC (50% EtOAc/hexanes) R_f = 0.37. ¹H NMR (400 MHz, CDCl₃) δ 7.18 (d, J = 8.1 Hz, 2H), 7.10 – 6.97 (m, 7H), 6.80 (d, J = 2.6 Hz, 1H), 6.72 (d, J = 8.5 Hz, 1H), 6.61 (dd, J = 8.6, 2.7 Hz, 1H), 3.81 (s, 3H), 3.73 (t, J = 4.6, 4H), 3.48 (s, 2H), 2.97 (dd, J = 9.1, 6.4 Hz, 2H), 2.81 (dd, J = 9.2, 6.1 Hz, 2H), 2.44 (t, J = 4.5, 4H). ¹³C NMR (101 MHz, CDCl₃) δ 158.6, 143.2, 139.0, 137.7, 135.7, 135.2, 134.6, 131.1, 130.3, 129.6, 129.0, 128.4, 127.6, 125.8, 113.4, 111.0, 67.1, 63.4, 55.4, 53.7, 30.8, 29.1. HRMS(ESI-TOF) for C₂₈H₃₀NO₂ m/z [M+H]⁺ calcd: 412.2277, found: 412.2274.

4-(6-Methoxy-2-phenyl-3,4-dihydronaphthalen-1-yl)-N-(2-(methylamino)ethyl)benzamide (25).

In a flame-dried, argon-flushed round bottom flask, a mixture of 11 (90 mg, 0.23 mmol), N-(2-(methylamino)ethyl)-4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)benzamide (75 mg, 0.23 mmol), and [1,1′-bis(diphenylphosphino)ferrocene]dichloropalladium(II) (9 mg, 5 mol%) was suspended in anhydrous dioxane (2 mL). An aqueous solution of 1.5 M cesium carbonate (0.3 mL, 0.47 mmol) was added dropwise, and the reaction mixture was stirred for 18 h at 90 °C under argon. Water (2 mL) and ethyl acetate (4 mL) were added, and the organic phase was separated, dried over anhydrous sulfate, filtered, and concentrated under reduced pressure. The crude mixture was adsorbed to celite and purified by flash chromatography on a 4 g silica column eluting with 0−20% MeOH/DCM to give 25 as a light brown solid (90 mg, 90%). TLC (5% MeOH/DCM) R_f = 0.43. ¹H NMR (400 MHz, CDCl₃) δ 7.73 (d, J = 8.2 Hz, 2H), 7.21 – 7.04 (m, 6H), 7.04 – 6.98 (m, 2H), 6.81 (d, J = 2.2 Hz, 1H), 6.63 (d, J = 8.6 Hz, 1H), 6.59 (dd, J = 8.6, 2.5 Hz, 2H), 3.83 (s, 3H), 3.60 (q, J = 5.4 Hz, 2H), 2.99 (dd, J = 8.9, 6.5 Hz, 2H), 2.82 (dd, J = 9.2, 6.2 Hz, 2H), 2.67 (t, J = 5.4 Hz, 2H), 2.40 (s, 7H). ¹³C NMR (101 MHz, CDCl₃) δ 167.4, 158.6, 143.7, 142.7, 137.6, 135.3, 134.5, 132.1, 131.3, 129.8, 128.2, 127.7, 127.3, 126.9, 126.0, 113.4, 110.8, 58.2, 55.3, 44.9, 36.7, 30.8, 28.9. HRMS(ESI-TOF) for C₂₈H₃₁N₂O₂ m/z [M+H]⁺ calcd: 427.2386, found: 427.2383.

4-(6-Methoxy-2-phenyl-3,4-dihydronaphthalen-1-yl)-N-(2-(piperidin-1-yl)ethyl)benzamide (26).

In a flame-dried, argon-flushed round bottom flask, a mixture of 11 (90 mg, 0.23 mmol), N-(2-(piperidin-1-yl)ethyl)-4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)benzamide (84 mg, 0.23 mmol), and [1,1′-bis(diphenylphosphino)ferrocene]dichloropalladium(II) (9 mg, 5 mol%) was suspended in anhydrous dioxane (2 mL). An aqueous solution of 1.5 M cesium carbonate (0.3 mL, 0.47 mmol) was added dropwise, and the reaction mixture was stirred for 18 h at 90 °C under argon. Water (2 mL) and ethyl acetate (4 mL) were added, and the organic phase was separated, dried over anhydrous sulfate, filtered, and concentrated under reduced pressure. The crude mixture was adsorbed to celite and purified by flash chromatography on a 4 g silica column eluting with 0−20% MeOH/DCM to give 26 as a light brown solid (75 mg, 69%). TLC (5% MeOH/DCM) R_f = 0.4. ¹H NMR (400 MHz, CDCl₃) δ 7.72 (d, J = 8.1 Hz, 2H), 7.25 – 7.04 (m, 6H), 7.02 (d, J = 6.8 Hz, 2H), 6.81 (d, J = 2.1 Hz, 1H), 6.64 (d, J = 8.6 Hz, 1H), 6.60 (dd, J = 8.6, 2.4 Hz, 1H), 3.83 (s, 3H), 3.65 – 3.57 (m, 2H), 3.03 – 2.94 (m, 2H), 2.82 (dd, J = 9.0, 6.4 Hz, 2H), 2.71 – 2.64 (m, 2H), 2.57 (s, 4H), 1.71 – 1.65 (m, 4H), 1.52 (s, 2H). ¹³C NMR (101 MHz, CDCl₃) δ 167.3, 158.6, 143.6, 142.7, 137.6, 135.3, 134.5, 132.3, 131.3, 129.8, 128.2, 127.7, 127.3, 126.8, 126.0, 113.4, 110.8, 57.4, 55.3, 54.4, 36.1, 30.8, 28.9, 25.4, 23.9. HRMS(ESI-TOF) for C₃₁H₃₅N₂O₂ m/z [M+H]⁺ calcd: 467.2699, found: 467.2695.

(4-(6-Methoxy-2-phenyl-3,4-dihydronaphthalen-1-yl)phenyl)(piperazin-1-yl)methanone (27).

In a flame-dried, argon-flushed round bottom flask, a mixture of 11 (90 mg, 0.23 mmol), piperazin-1-yl(4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)phenyl)methanone (74 mg, 0.23 mmol), and [1,1′-bis(diphenylphosphino)ferrocene]dichloropalladium(II) (9 mg, 5 mol%) was suspended in anhydrous dioxane (2 mL). An aqueous solution of 1.5 M cesium carbonate (0.3 mL, 0.47 mmol) was added dropwise, and the reaction mixture was stirred for 18 h at 90 °C under argon. Water (2 mL) and ethyl acetate (4 mL) were added, and the organic phase was separated, dried over anhydrous sulfate, filtered, and concentrated under reduced pressure. The crude mixture was adsorbed to celite and purified by flash chromatography on a 4 g silica column eluting with 0−20% MeOH/DCM to give 27 as a light brown solid (31 mg, 31%). TLC (5% MeOH/DCM) R_f = 0.42. ¹H NMR (400 MHz, CDCl₃) δ 7.26 (s, 2H), 7.18 – 6.96 (m, 7H), 6.83 – 6.78 (m, 1H), 6.68 (d, J = 8.4 Hz, 1H), 6.64 – 6.57 (m, 1H), 3.82 (s, 3H), 3.83 – 3.44 (m, 4H), 3.09 – 2.89 (m, 4H), 2.87 – 2.77 (m, 2H). ¹³C NMR (101 MHz, CDCl₃) δ 170.6, 158.7, 142.7, 142.4, 137.6, 135.5, 134.4, 132.6, 131.5, 129.6, 128.3, 127.6, 127.3, 126.9, 126.1, 113.5, 110.7, 55.3, 44.8, 30.7, 28.9. HRMS(ESI-TOF) for C₂₈H₂₉N₂O₂ m/z [M+H]⁺ calcd: 425.2229, found: 425.2233.

(4-Methylpiperazin-1-yl)(4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)phenyl)methanone (28).

In a flame-dried, argon-flushed round bottom flask, a mixture of 11 (90 mg, 0.23 mmol), (4-(6-methoxy-2-phenyl-3,4-dihydronaphthalen-1-yl)phenyl)(4-methylpiperazin-1-yl)methanone (77 mg, 0.23 mmol), and [1,1′-bis(diphenylphosphino)ferrocene]dichloropalladium(II) (9 mg, 5 mol%) was suspended in anhydrous dioxane (2 mL). An aqueous solution of 1.5 M cesium carbonate (0.3 mL, 0.47 mmol) was added dropwise, and the reaction mixture was stirred for 18 h at 90 °C under argon. Water (2 mL) and ethyl acetate (4 mL) were added, and the organic phase was separated, dried over anhydrous sulfate, filtered, and concentrated under reduced pressure. The crude mixture was adsorbed to celite and purified by flash chromatography on a 4 g silica column eluting with 0−20% MeOH/DCM to give 28 as a light brown solid (70 mg, 68%). TLC (5% MeOH/DCM) R_f = 0.4. ¹H NMR (400 MHz, CDCl₃) δ 7.29 (d, J = 5.9 Hz, 2H), 7.17 – 7.04 (m, 5H), 7.03 – 6.97 (m, 2H), 6.81 (d, J = 2.4 Hz, 1H), 6.70 (d, J = 8.6 Hz, 1H), 6.62 (dd, J = 8.6, 2.7 Hz, 1H), 3.83 (s, 3H), 3.81 – 3.33 (m, 4H), 3.03 – 2.95 (m, 2H), 2.82 (dd, J = 9.1, 6.3 Hz, 2H), 2.49 (s, 4H), 2.39 (s, 4H). ¹³C NMR (101 MHz, CDCl₃) δ 170.4, 158.7, 142.8, 142.0, 137.6, 135.3, 134.5, 133.4, 131.3, 129.7, 128.3, 127.6, 127.4, 126.9, 126.0, 113.4, 110.8, 55.3, 54.9, 47.4, 45.8, 41.8, 30.7, 28.9. HRMS(ESI-TOF) for C₂₉H₃₁N₂O₂ m/z [M+H]⁺ calcd: 439.2386, found: 439.2389.

N-(4-(6-Methoxy-2-phenyl-3,4-dihydronaphthalen-1-yl)phenyl)tetrahydro-2H-pyran-4-carboxamide (29).

In a flame-dried, argon-flushed round bottom flask, a mixture of 11 (100 mg, 0.26 mmol), N-(4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)phenyl)tetrahydro-2H-pyran-4-carboxamide (86 mg, 0.26 mmol), and [1,1′-bis(diphenylphosphino)ferrocene]dichloropalladium(II) (10 mg, 5 mol%) was suspended in anhydrous dioxane (2 mL). An aqueous solution of 1.5 M cesium carbonate (0.35 mL, 0.5 mmol) was added dropwise, and the reaction mixture was stirred for 18 h at 90 °C under an argon atmosphere. Water (3 mL) and ethyl acetate (6 mL) were added, and the organic phase was separated, dried over anhydrous sulfate, filtered, and concentrated under reduced pressure. The crude mixture was adsorbed to celite and purified by flash chromatography on a 12-g silica column eluting with 20−70% EtOAc/hexanes to give 29 as a white solid (120 mg, 85%). TLC (50% EtOAc/hexanes) R_f = 0.5. ¹H NMR (400 MHz, CDCl₃) δ 7.41 (d, J = 8.4 Hz, 2H), 7.13 (t, J = 7.2 Hz, 3H), 7.09 – 7.00 (m, 5H), 6.80 (d, J = 2.3 Hz, 1H), 6.72 (d, J = 8.5 Hz, 1H), 6.60 (dd, J = 8.5, 2.6 Hz, 1H), 4.09 (d, J = 11.1 Hz, 2H), 3.83 (s, 3H), 3.48 (t, J = 10.5 Hz, 2H), 3.01 – 2.93 (m, 2H), 2.80 (dd, J = 9.0, 6.3 Hz, 2H), 2.49 (td, J = 11.1, 5.4 Hz, 1H), 2.02 – 1.78 (m, 4H). ¹³C NMR (101 MHz, CDCl₃) δ 172.2, 158.5, 143.1, 137.7, 136.1, 134.8, 134.5, 131.8, 130.2, 128.3, 127.6, 127.4, 125.8, 119.3, 113.3, 110.8, 67.2, 55.3, 43.3, 30.8, 29.3, 29.0. HRMS (ESI-TOF) for C₂₉H₃₀NO₃ m/z [M+H]⁺ clad: 440.2226, found: 440.2221.

1-((4-(6-Methoxy-2-phenyl-3,4-dihydronaphthalen-1-yl)phenyl)sulfonyl)-4-methylpiperazine (30).

In a flame-dried, argon-flushed round bottom flask, a mixture of 11 (90 mg, 0.23 mmol), 1-methyl-4-((4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)phenyl)sulfonyl)piperazine (86 mg, 0.23 mmol), and [1,1′-bis(diphenylphosphino)ferrocene]dichloropalladium(II) (9 mg, 5 mol%) was suspended in anhydrous dioxane (2 mL). An aqueous solution of 1.5 M cesium carbonate (0.3 mL, 0.47 mmol) was added dropwise, and the reaction mixture was stirred for 18 h at 90 °C under argon. Water (2 mL) and ethyl acetate (4 mL) were added, and the organic phase was separated, dried over anhydrous sulfate, filtered, and concentrated under reduced pressure. The crude mixture was adsorbed to celite and purified by flash chromatography on a 4 g silica column eluting with 20−80% EtOAc/Hexanes to give 30 as a light brown solid (52 mg, 47%). TLC (50% EtOAc/Hexanes) R_f = 0.28. ¹H NMR (400 MHz, CDCl₃) δ 7.65 – 7.57 (m, 2H), 7.28 – 7.23 (m, 2H), 7.16 – 7.03 (m, 3H), 6.99 – 6.91 (m, 2H), 6.83 (d, J = 2.6 Hz, 1H), 6.65 (dd, J = 8.6, 2.7 Hz, 1H), 6.59 (d, J = 8.6 Hz, 1H), 3.84 (s, 3H), 3.07 (s, 4H), 2.99 (dd, J = 9.0, 6.5 Hz, 2H), 2.83 (dd, J = 9.2, 6.1 Hz, 2H), 2.56 – 2.49 (m, 4H), 2.33 (s, 3H). ¹³C NMR (101 MHz, CDCl₃) δ 158.8, 145.5, 142.2, 137.6, 136.3, 133.8, 133.4, 131.9, 129.3, 128.2, 127.9, 127.4, 127.1, 126.4, 113.5, 111.1, 55.3, 54.1, 46.0, 45.8, 30.8, 28.8. HRMS(ESI-TOF) for C₂₈H₃₁N₂O₃S m/z [M+H]⁺ calcd: 475.2055, found: 475.2053.

N-(4-(6-Methoxy-2-phenyl-3,4-dihydronaphthalen-1-yl)phenyl)-1-methylpiperidine-4-carboxamide (31).

In a flame-dried, argon-flushed round bottom flask, a mixture of 11 (100 mg, 0.26 mmol), intermediate 9 (90 mg, 0.26 mmol), and [1,1′-bis(diphenylphosphino)ferrocene]dichloropalladium(II) (10 mg, 5 mol%) was suspended in anhydrous dioxane (2 mL). An aqueous solution of 1.5 M cesium carbonate (0.35 mL, 0.52 mmol) was added dropwise, and the reaction mixture was stirred for 18 h at 90 °C under argon. Water (2 mL) and ethyl acetate (4 mL) were added, and the organic phase was separated, dried over anhydrous sulfate, filtered, and concentrated under reduced pressure. The crude mixture was adsorbed to celite and purified by flash chromatography on a 4 g silica column eluting with 0−20% MeOH/DCM to give 31 as a light brown solid (50 mg, 45%). TLC (10% MeOH/DCM) R_f = 0.5. ¹H NMR (400 MHz, CDCl₃) δ 8.33 (s, 1H), 7.45 (d, J = 8.2 Hz, 2H), 7.15 – 7.08 (m, 2H), 7.03 (t, J = 6.7 Hz, 5H), 6.79 (s, 1H), 6.69 (d, J = 8.6 Hz, 1H), 6.58 (dd, J = 8.6, 2.1 Hz, 1H), 3.82 (s, 3H), 3.49 – 3.26 (m, 2H), 3.03 – 2.82 (m, 4H), 2.82 – 2.74 (m, 3H), 2.69 (s, 3H), 2.29 – 2.06 (m, 4H). ¹³C NMR (101 MHz, CDCl₃) δ 172.2, 158.5, 143.1, 137.6, 136.2, 136.1, 134.8, 134.6, 131.7, 130.2, 129.4, 128.3, 127.7, 127.4, 126.7, 125.8, 119.5, 113.3, 110.8, 55.3, 53.4, 44.4, 30.8, 29.0, 26.4. HRMS(ESI-TOF) for C₃₀H₃₃N₂O₂ m/z [M+H]⁺ calcd: 453.2542, found: 453.2546.

tert-Butyl 4-((4-(6-methoxy-2-phenyl-3,4-dihydronaphthalen-1-yl)phenyl)carbamoyl)piperidine-1-carboxylate (32).

In a flame-dried, argon-flushed round bottom flask, a mixture of 11 (100 mg, 0.26 mmol), tert-butyl 4-((4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)phenyl)carbamoyl)piperidine-1-carboxylate (112 mg, 0.26 mmol), and [1,1′-bis(diphenylphosphino)ferrocene]dichloropalladium(II) (10 mg, 5 mol%) was suspended in anhydrous dioxane (2 mL). An aqueous solution of 1.5 M cesium carbonate (0.35 mL, 0.5 mmol) was added dropwise, and the reaction mixture was stirred for 18 h at 90 °C under an argon atmosphere. Water (3 mL) and ethyl acetate (6 mL) were added, and the organic phase was separated, dried over anhydrous sulfate, filtered, and concentrated under reduced pressure. The crude mixture was adsorbed to celite and purified by flash chromatography on a 12-g silica column eluting with 20−70% EtOAc/hexanes to give 32 as a white solid (120 mg, 85%). TLC (50% EtOAc/hexanes) R_f = 0.5. ¹H NMR (400 MHz, CDCl₃) δ 7.40 (d, J = 8.4 Hz, 2H), 7.18 – 7.00 (m, 8H), 6.80 (d, J = 2.6 Hz, 1H), 6.71 (d, J = 8.6 Hz, 1H), 6.60 (dd, J = 8.6, 2.7 Hz, 1H), 4.20 (s, 2H), 3.82 (s, 3H), 3.01 – 2.92 (m, 2H), 2.80 (dd, J = 9.2, 6.2 Hz, 4H), 2.38 (tt, J = 11.3, 3.5 Hz, 1H), 1.92 (d, J = 11.9 Hz, 2H), 1.76 (qd, J = 12.4, 4.3 Hz, 2H), 1.49 (s, 9H). ¹³C NMR (101 MHz, CDCl₃) δ 172.4, 158.5, 154.7, 143.1, 137.7, 136.1, 134.8, 134.5, 131.8, 130.2, 128.3, 127.6, 127.4, 125.8, 119.3, 113.3, 110.8, 79.7, 55.3, 44.5, 43.2, 30.8, 29.0, 28.7, 28.5. HRMS (ESI-TOF) for C₃₄H₃₈N₂O₄ m/z [M+H]⁺ calcd: 538.2832, found: 538.2836.

N-(4-(6-Methoxy-2-phenyl-3,4-dihydronaphthalen-1-yl)phenyl)-4-methylpiperazine-1-carboxamide (33).

In a flame-dried, argon-flushed round bottom flask, a mixture of 11 (100 mg, 0.26 mmol), 4-methyl-N-(4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)phenyl)piperazine-1-carboxamide (90 mg, 0.26 mmol), and [1,1′-bis(diphenylphosphino)ferrocene]dichloropalladium(II) (10 mg, 5 mol%) was suspended in anhydrous dioxane (2 mL). An aqueous solution of 1.5 M cesium carbonate (0.35 mL, 0.5 mmol) was added dropwise, and the reaction mixture was stirred for 18 h at 90 °C under an argon atmosphere. Water (3 mL) and ethyl acetate (6 mL) were added, and the organic phase was separated, dried over anhydrous sulfate, filtered, and concentrated under reduced pressure. The crude mixture was adsorbed to celite and purified by flash chromatography on a 4-g silica column eluting with 0−20% MeOH/DCM to give 33 as a brown solid (70 mg, 60%). TLC (10% MeOH/DCM) R_f = 0.5. Further purification by precipitation from a concentrated methanol solution yielded 33 as a beige solid (15 mg). ¹H NMR (400 MHz, CDCl₃) δ 7.24 (d, J = 8.5 Hz, 2H), 7.16 – 7.09 (m, 2H), 7.09 – 6.97 (m, 5H), 6.82 – 6.73 (m, 2H), 6.60 (dd, J = 8.6, 2.6 Hz, 1H), 6.32 (s, 1H), 3.83 (s, 3H), 3.56 – 3.49 (m, 4H), 3.01 – 2.92 (m, 2H), 2.80 (dd, J = 9.2, 6.3 Hz, 2H), 2.50 – 2.43 (m, 4H), 2.35 (s, 3H). ¹³C NMR (101 MHz, CDCl₃) δ 158.5, 154.9, 143.2, 137.7, 137.3, 134.7, 134.6, 134.5, 131.7, 130.3, 128.3, 127.6, 125.7, 119.3, 113.3, 110.7, 77.3, 77.0, 76.7, 55.3, 54.7, 46.1, 44.1, 30.8, 29.0. HRMS (ESI-TOF) for C₂₉H₃₂N₃O₂ m/z [M+H]⁺ calcd: 454.2495, found: 454.2491.

4-(6-Methoxy-2-phenyl-3,4-dihydronaphthalen-1-yl)aniline (34).

To a solution of 18 (167 mg, 0.04 mmol) in 1,4-dioxane (2 mL) was added a solution of 4 M HCl in 1,4-dioxane (2 mL). The reaction mixture was stirred for 6 h until completion, as monitored by TLC. The solvent was dried, and the HCl salt of 34 (128 mg, quant.) was desalted on a sodium bicarbonate SPE column. ¹H NMR (400 MHz, CD₃OD) δ 7.30 (d, J = 7.7 Hz, 2H), 7.22 (d, J = 7.8 Hz, 2H), 7.14 – 6.99 (m, 6H), 6.85 (s, 1H), 6.65 – 6.52 (m, 2H), 3.80 (s, 3H), 3.33 (s, 1H), 2.98 (t, J = 7.6 Hz, 2H), 2.78 (t, J = 7.6 Hz, 2H). ¹³C NMR (101 MHz, CD₃OD) δ 159.0, 142.7, 141.4, 137.5, 135.9, 133.8, 132.7, 129.2, 128.9, 128.0, 127.4, 126.6, 125.8, 122.4, 113.0, 110.6, 54.3, 30.6, 28.4. HRMS(ESI-TOF) for C₂₃H₂₂NO m/z [M+H]⁺ calcd: 328.1701, found: 328.1696.

4-((4-(6-Methoxy-2-phenyl-3,4-dihydronaphthalen-1-yl)phenyl)carbamoyl)piperidin-1-ium chloride (35)

To a solution of 32 (140 mg, 0.26 mmol) in 1,4-dioxane (1 mL) was added a solution of 4 M HCl in 1,4-dioxane (1 mL). The reaction mixture was stirred for 6 h until completion as monitored by TLC. The solvent was dried, and the HCl salt, was washed with ether to yield 35 as a white solid (120 mg, quant.). ¹H NMR (400 MHz, DMSO-d₆) δ 10.11 (s, 1H), 7.50 (d, J = 8.4 Hz, 2H), 7.13 (t, J = 7.3 Hz, 2H), 7.08 – 6.99 (m, 3H), 6.93 (d, J = 8.4 Hz, 2H), 6.85 (d, J = 2.3 Hz, 1H), 6.64 (dd, J = 8.6, 2.5 Hz, 1H), 6.53 (d, J = 8.6 Hz, 1H), 3.75 (s, 3H), 3.70 (dd, J = 11.4, 3.7 Hz, 1H), 3.57 (s, 2H), 3.49 (dd, J = 11.0, 4.1 Hz, 1H), 3.04 – 2.80 (m, 4H), 2.67 (M, 3H), 2.02 – 1.92 (m, 2H), 1.90 – 1.75 (m, 2H). ¹H NMR (400 MHz, CD₃OD) δ 7.44 (d, J = 8.4 Hz, 2H), 7.11 – 7.04 (m, 2H), 7.05 – 6.98 (m, 3H), 6.95 (d, J = 8.4 Hz, 2H), 6.79 (d, J = 2.6 Hz, 1H), 6.62 (d, J = 8.6 Hz, 1H), 6.58 (dd, J = 8.6, 2.5 Hz, 1H), 3.77 (s, 3H), 3.65 (s, 1H), 3.48 (dt, J = 12.9, 3.9 Hz, 2H), 3.09 (td, J = 12.6, 3.4 Hz, 2H), 2.92 (dd, J = 9.2, 6.3 Hz, 2H), 2.80 – 2.69 (m, 3H), 2.10 (dd, J = 14.7, 3.8 Hz, 2H), 1.98 (dtd, J = 14.8, 11.4, 4.1 Hz, 2H). ¹³C NMR (101 MHz, CD₃OD) δ 174.1, 160.1, 144.5, 138.9, 138.1, 137.4, 136.1, 132.6, 131.3, 129.4, 128.6, 128.3, 126.8, 120.7, 114.2, 111.9, 55.7, 44.3, 41.6, 32.0, 29.9, 26.6. HRMS(ESI-TOF) for C₂₉H₃₀N₂O₂ m/z [M+H]⁺ calcd: 439.2386, found: 439.2388.

N-(4-(6-Methoxy-2-phenyl-3,4-dihydronaphthalen-1-yl)phenyl)piperazine-1-carboxamide (36).

A hydrogenation vessel was evacuated and backfilled with argon three times. 33 (17 mg, 0.04 mmol) and 10% Palladium on carbon (50% wet with water, 10 mg, 10 mol%) were placed in the vessel under argon, followed by methanol (3 mL). The vessel was initially placed on a Parr hydrogenator under a 50-psi hydrogen atmosphere and shaken for 18 h. The hydrogen was vented accidentally, and the reaction mixture was left under air for 70 h. Under these aerobic conditions, instead of reduction as the main product, a product of N-demethylation of the piperazine ring was obtained, consistent with reported aerobic N-demethylation reactions mediated by Pd(0) catalyst.⁵⁶ The reaction mixture was filtered through a pad of Celite and concentrated to give 36 as a white solid (17 mg, quant.), which was further purified by preparative HPLC. ¹H NMR (600 MHz, Methanol-d₆) δ 7.29 – 7.24 (m, 2H), 7.11 – 7.07 (m, 2H), 7.06 – 7.00 (m, 3H), 6.99 – 6.93 (m, 2H), 6.81 (d, J = 2.6 Hz, 1H), 6.67 (d, J = 8.5 Hz, 1H), 6.60 (dd, J = 8.6, 2.7 Hz, 1H), 3.82 – 3.76 (m, 7H), 3.29 (dd, J = 5.9, 4.6 Hz, 4H), 2.98 – 2.92 (m, 2H), 2.75 (dd, J = 9.0, 6.6 Hz, 2H). ¹³C NMR (151 MHz, Methanol-d₆) δ 158.7, 156.0, 143.3, 137.5, 135.1, 134.8, 134.5, 131.2, 130.0, 128.0, 127.2, 127.0, 125.4, 120.0, 112.7, 110.4, 54.3, 43.1, 41.1, 30.6, 28.5. HRMS(ESI-TOF) for C₂₈H₃₀N₃O₂ m/z [M+H]⁺ calcd: 440.2338, found: 440.2332.

(±)-(1R,2S/1S,2R)-N-(2-(Dimethylamino)ethyl)-4-(6-methoxy-2-phenyl-1,2,3,4-tetrahydronaphthalen-1-yl)benzamide (37).

A hydrogenation vessel was evacuated and backfilled with argon three times. 25 (43 mg, 0.1 mmol) in ethyl acetate (2 mL) and 10% Palladium on carbon (50% wet with water, 40 mg, 20 mol%) were placed in the vessel under argon. Methanol (5 mL) was added, and the vessel was placed on a Parr hydrogenator under 50 psi hydrogen atmosphere and shaken for 24 h. The reaction was monitored for completion by LCMS as both the starting material and the product have identical R_f value on TLC. Upon complete reduction, the reaction mixture was filtered through a pad of Celite and concentrated to give 37 as a white solid (40 mg, quant.), which was further purified by preparative HPLC. ¹H NMR (600 MHz, CDCl₃) δ 7.69 (s, 1H), 7.52 (d, J = 7.8 Hz, 2H), 7.17 (s, 3H), 6.89 – 6.77 (m, 4H), 6.69 (d, J = 7.7 Hz, 1H), 6.52 (d, J = 7.8 Hz, 2H), 4.36 (d, J = 4.4 Hz, 1H), 3.84 (s, 4H), 3.67 (s, 2H), 3.45 (d, J = 9.5 Hz, 1H), 3.31 – 3.02 (m, 2H), 2.88 (s, 2H), 2.55 (s, 6H), 2.20 (dd, J = 12.3, 5.7 Hz, 1H), 1.88 (s, 1H). ¹³C NMR (151 MHz, CDCl₃) δ 167.7, 158.1, 146.5, 143.7, 137.9, 131.4, 131.3, 131.2, 130.6, 128.0, 127.9, 126.3, 125.9, 113.1, 112.7, 58.0, 55.3, 51.0, 45.3, 44.4, 36.1, 30.1, 21.9. HRMS(ESI-TOF) for C₂₈H₃₃N₂O₂ m/z [M+H]⁺ calcd: 429.2542, found: 429.2538.

(±)-(1R,2S/1S,2R)-(4-(6-Methoxy-2-phenyl-1,2,3,4-tetrahydronaphthalen-1-yl)phenyl)(piperazin-1-yl)methanone (38).

A hydrogenation vessel was evacuated and backfilled with argon three times. 27 (17 mg, 0.04 mmol) in ethyl acetate (1 mL) and 10% Palladium on carbon (50% wet with water, 20 mg, 20 mol%) were placed in the vessel under argon. Methanol (3 mL) was added, and the vessel was placed on a Parr hydrogenator under 50 psi hydrogen atmosphere and shaken for 24 h. The reaction was monitored for completion by LCMS as both the starting material and the product have identical R_f value on TLC. Upon complete reduction, the reaction mixture was filtered through a pad of Celite and concentrated to give 38 as a white solid (15 mg, quant.), which was further purified by preparative HPLC. ¹H NMR (600 MHz, CDCl₃) δ 7.18 – 7.14 (m, 3H), 7.06 (d, J = 8.2 Hz, 2H), 6.87 – 6.79 (m, 4H), 6.71 (dd, J = 8.4, 2.7 Hz, 1H), 6.53 (d, J = 8.1 Hz, 2H), 4.36 (d, J = 5.3 Hz, 1H), 3.84 (s, 3H), 3.59 (br s, 4H), 3.47 (ddd, J = 13.3, 5.4, 2.3 Hz, 1H), 3.12 (dtd, J = 23.5, 17.3, 5.8 Hz, 2H), 2.86 (s, 5H), 2.16 (qd, J = 13.0, 6.2 Hz, 1H), 1.90 (dd, J = 12.9, 5.7 Hz, 1H). ¹³C NMR (151 MHz, CDCl₃) δ 170.9, 158.2, 146.1, 143.6, 138.0, 131.5, 130.8, 130.6, 128.0, 127.9, 126.4, 126.0, 113.2, 112.8, 55.3, 51.0, 45.2, 43.6, 30.1, 21.8. HRMS(ESI-TOF) for C₂₈H₃₀N₂O₂ m/z [M+H]⁺ calcd: 441.2542, found: 441.2537.

(±)-(1R,2S/1S,2R)-(4-(6-Methoxy-2-phenyl-1,2,3,4-tetrahydronaphthalen-1-yl)phenyl)(4-methylpiperazin-1-yl)methanone (39).

A hydrogenation vessel was evacuated and backfilled with argon three times. 28 (44 mg, 0.1 mmol) in ethyl acetate (2 mL) and 10% Palladium on carbon (50% wet with water, 40 mg, 20 mol%) were placed in the vessel under argon. Methanol (5 mL) was added, and the vessel was placed on a Parr hydrogenator under 50 psi hydrogen atmosphere and shaken for 24 h. The reaction was monitored for completion by LCMS as both the starting material and the product have identical R_f value on TLC. Upon complete reduction, the reaction mixture was filtered through a pad of Celite and concentrated to give 39 as a white solid (44 mg, quant.), which was further purified by preparative HPLC. ¹H NMR (400 MHz, CDCl₃) δ 7.19 – 7.13 (m, 3H), 7.05 (d, J = 8.1 Hz, 2H), 6.86 (d, J = 8.5 Hz, 1H), 6.84 – 6.78 (m, 4H), 6.71 (dd, J = 8.5, 2.6 Hz, 1H), 6.49 (d, J = 8.1 Hz, 2H), 4.34 (d, J = 5.3 Hz, 1H), 3.84 (s, 3H), 3.45 (ddd, J = 13.1, 5.4, 2.2 Hz, 3H), 3.11 (qd, J = 17.4, 8.8 Hz, 4H), 2.79 – 2.39 (m, 4H), 2.37 (s, 3H), 2.20 (tq, J = 13.1, 6.6 Hz, 1H), 1.88 (dd, J = 11.7, 4.2 Hz, 1H). ¹³C NMR (101 MHz, CDCl₃) δ 170.5, 158.1, 144.6, 143.8, 137.9, 132.8, 131.5, 131.2, 130.4, 128.1, 127.8, 126.2, 125.8, 113.1, 112.7, 55.2, 54.8, 50.9, 47.3, 45.7, 45.3, 41.7, 30.1, 21.9. HRMS(ESI-TOF) for C₂₉H₃₃N₂O₂ m/z [M+H]⁺ calcd: 441.2542, found: 441.2538.

(±)-(1R,2S/1S,2R)-tert-Butyl 4-((4-(6-methoxy-2-phenyl-1,2,3,4-tetrahydronaphthalen-1-yl)phenyl)carbamoyl)piperidine-1-carboxylate (40).

A hydrogenation vessel was evacuated and backfilled with argon three times. 32 (150 mg, 0.28 mmol) in ethyl acetate (2 mL) and 10% Palladium on carbon (50% wet with water, 60 mg, 10 mol%) were placed in the vessel under argon. Methanol (10 mL) was added, and the vessel was placed on a Parr hydrogenator under 50 psi hydrogen atmosphere and shaken for 24 h. The reaction was monitored for completion by LCMS as both the starting material and the product have identical R_f value on TLC. Upon complete reduction, the reaction mixture was filtered through a pad of Celite and concentrated to give 40 as a white solid (150 mg, quant.), which was further purified by preparative HPLC. ¹H NMR (400 MHz, CDCl₃) δ 7.22 – 7.12 (m, 5H), 7.07 (s, 1H), 6.89 – 6.81 (m, 3H), 6.79 (d, J = 2.6 Hz, 1H), 6.69 (dd, J = 8.4, 2.7 Hz, 1H), 6.39 (d, J = 8.6 Hz, 2H), 4.29 (d, J = 5.0 Hz, 1H), 4.17 (dd, J = 15.5, 8.5 Hz, 2H), 3.84 (s, 3H), 3.41 (ddd, J = 13.4, 5.4, 2.3 Hz, 1H), 3.09 (dtt, J = 17.4, 11.9, 6.2 Hz, 2H), 2.77 (t, J = 12.0 Hz, 2H), 2.32 (ddt, J = 11.5, 7.3, 3.7 Hz, 1H), 2.20 (tq, J = 13.1, 6.7 Hz, 1H), 1.87 (d, J = 13.0 Hz, 3H), 1.72 (tt, J = 12.1, 6.5 Hz, 2H), 1.48 (s, 9H). ¹³C NMR (101 MHz, CDCl₃) δ 172.3, 158.0, 154.7, 144.1, 138.7, 137.8, 135.5, 131.8, 131.5, 131.0, 128.1, 127.8, 126.1, 118.3, 113.0, 112.6, 79.7, 55.3, 50.5, 45.3, 44.4, 43.1, 30.1, 28.6, 28.5, 21.9. HRMS(ESI-TOF) for C₃₄H₄₀N₂O₄ m/z [M+Na]⁺ calcd: 563.2886, found: 563.2888.

(±)-(1R,2S/1S,2R)-N-(4-(6-Methoxy-2-phenyl-1,2,3,4-tetrahydronaphthalen-1-yl)phenyl)piperidine-4-carboxamide (41).

To a solution of 40 (60 mg, 0.11 mmol) in 1,4-dioxane (1 mL) was added a solution of 4 M HCl in 1,4-dioxane (1 mL). The reaction mixture was stirred for 6 h until completion as monitored by TLC. The solvent was dried, and the deprotected Boc-deprotected HCl salt intermediate (49 mg, quant.) was desalted on a sodium bicarbonate SPE column. ¹H NMR (400 MHz, CDCl₃) δ 8.94 (s, 1H), 8.46 (s, 1H), 7.19 (d, J = 8.1 Hz, 2H), 7.14 – 7.02 (m, 3H), 6.79 (dd, J = 7.4, 2.3 Hz, 2H), 6.75 (d, J = 3.3 Hz, 1H), 6.63 (dd, J = 8.4, 2.6 Hz, 1H), 6.35 (d, J = 8.2 Hz, 2H), 4.23 (d, J = 4.9 Hz, 1H), 3.77 (s, 3H), 3.41 – 3.34 (m, 1H), 3.28 (s, 2H), 3.06 (s, 2H), 2.77 (s, 2H), 2.46 (s, 1H), 2.25 – 2.09 (m, 1H), 1.87 (s, 6H). ¹³C NMR (101 MHz, CDCl₃) δ 172.20, 169.16, 157.93, 144.00, 138.70, 137.78, 135.89, 131.86, 131.40, 130.90, 128.08, 127.82, 126.09, 118.56, 113.04, 112.66, 55.23, 50.47, 45.26, 43.00, 41.29, 30.08, 25.79, 21.88. ¹H NMR (400 MHz, CD₃OD) δ 7.18 – 7.06 (m, 5H), 6.86 – 6.74 (m, 4H), 6.65 (dd, J = 8.5, 2.6 Hz, 1H), 6.36 (d, J = 8.5 Hz, 2H), 4.27 (d, J = 5.1 Hz, 1H), 3.78 (s, 3H), 3.46 (dt, J = 12.4, 2.6 Hz, 2H), 3.37 (ddd, J = 13.2, 5.0, 1.9 Hz, 2H), 3.17 – 2.97 (m, 4H), 2.67 (tt, J = 10.7, 3.8 Hz, 1H), 2.22 (qd, J = 13.0, 6.4 Hz, 1H), 2.05 (dd, J = 14.5, 3.1 Hz, 2H), 1.93 (qd, J = 13.2, 3.9 Hz, 2H), 1.85 – 1.76 (m, 1H). ¹³C NMR (101 MHz, CD₃OD) δ 173.9, 159.6, 145.6, 140.3, 138.9, 137.3, 133.0, 132.4, 131.9, 129.2, 128.8, 127.0, 119.8, 119.7, 113.9, 113.7, 55.6, 51.9, 46.7, 44.3, 41.5, 31.1, 26.6, 23.3. HRMS(ESI-TOF) for C₂₉H₃₃N₂O₂ m/z [M+H]⁺ calcd: 441.2542, found: 441.2534.

(±)-(1R,2S/1S,2R)-N-(4-(6-Methoxy-2-phenyl-3,4-dihydronaphthalen-1-yl)phenyl)piperidine-4-carbothioamide (42).

To a flame-dried 25 mL round-bottom flask equipped with a magnetic stir bar and a reflux condenser under an atmosphere of argon was added 32 (20 mg, 37 μmol) and anhydrous THF (2 mL). 2,4-bis(4-methoxyphenyl)-1,3,2,4-dithiadiphosphetane 2,4-disulfide, Lawesson’s reagent (30 mg, 74 μmol) was added in three portions. The reaction mixture was heated at 120 °C for 48 h, where both conversion into the thioamide and deprotection of the Boc group were complete. The reaction mixture was allowed to cool to rt and was concentrated under reduced pressure. The residue was dissolved in ethyl acetate (20 mL) and washed with saturated aqueous sodium bicarbonate (2 × 5 mL) to remove phosphorus-containing byproducts, then with brine (5 mL). The organic layer was dried over anhydrous sodium sulfate, filtered, and concentrated in vacuo. The crude material was purified by flash column chromatography on silica gel (gradient elution: hexanes → hexanes/EtOAc 4:1 → 1:1) to afford 42 as a yellow solid (10 mg, 59%). ¹H NMR (400 MHz, CDCl₃) δ 9.78 (br s, 1H), 8.51 (br s, 1H), 7.88 – 7.67 (m, 1H), 7.47 (d, J = 7.4 Hz, 2H), 7.15 (d, J = 6.6 Hz, 3H), 6.92 – 6.78 (m, 2H), 6.76 (d, J = 2.3 Hz, 1H), 6.66 (dd, J = 9.6, 1.5 Hz, 1H), 6.42 (d, J = 7.9 Hz, 2H), 4.28 (d, J = 4.8 Hz, 1H), 3.80 (d, J = 3.5 Hz, 3H), 3.73 (d, J = 2.5 Hz, 1H), 3.50 – 3.32 (m, 2H), 3.24 – 2.86 (m, 3H), 2.82 (s, 1H), 2.36 (s, 2H), 2.25 – 2.03 (m, 2H), 2.00 (s, 1H), 1.84 (d, J = 8.8 Hz, 1H). ¹³C NMR (101 MHz, CDCl₃) δ 203.4 (C=S), 158.2, 144.0, 141.2, 138.0, 136.9, 131.7, 130.8, 128.3, 128.0, 126.3, 121.8, 113.3, 112.8, 55.5, 55.4, 53.0, 50.8, 45.5, 44.0, 30.2, 28.1, 22.1. HRMS(ESI-TOF) for C₂₉H₃₃N₂OS m/z [M+H]⁺ calcd: 457.2314, found: 457.2310.

Cell Lines and Reagents

Human A549 lung epithelial cells (ATCC CCL-185) and human 293T embryonic kidney cells (ATCC CRL-1573) were cultured in high-glucose Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 100 μg/mL streptomycin, 100 U penicillin (P/S) and 10% fetal bovine serum (FBS) at 37°C and 5% CO₂. HeLa cells (ATCC, Cat# CCL-2) were cultured in DMEM (ThermoFisher, Cat# 11995073) supplemented with 10% FBS (GeminiBio, Cat# 100-106-500), and kept at 37 °C, 5% CO₂ in a humidified incubator.

Production of Ebola and Marburg Pseudoviruses

Ebola Zaire (pEBOV) and Lake Victoria Marburg pseudoviruses (pMARV) containing a luciferase reporter gene were produced by transiently co-transfecting 293T cells with EBOV or MARV GP glycoprotein plasmids and HIV-1 pro-viral vector pNL4-3.Luc.R⁻E⁻ (BEI Resources) using a polyethyleneimine (PEI)-based transfection protocol. After a 5 h incubation period, the transfection media was changed to phenol red-free DMEM with 1% P/S and 2% FBS. Twenty-four hours post-transfection the supernatant containing the pseudoviruses was collected and filtered through a 0.45 μM pore size surfactant-free cellulose acetate membrane for immediate use.

Dose-response Analysis for Pseudoviral Entry Inhibition and Cytotoxicity

For pseudoviral entry inhibition assay, A549 cells were seeded at 5,000 cells per well in a white-bottom opaque 96-well plates and incubated overnight at 37 °C and 5% CO₂. The cells were transduced with the pseudovirus and treated with 3-fold serial dilutions of the test compounds previously dissolved in DMSO and further diluted in a way that the DMSO concentration never exceeds 1%. Pseudovirions with 1% DMSO were used as negative controls. After 48h of incubation, the cells were lysed using the Neolite Reporter Gene Assay System (Revvity) for luciferase activity measurement according to manufactures’ instructions. Luminescent signal was read on an EnVision plate reader (PerkinElmer). For the evaluation of compound’s cytotoxicity, A549 cells were seeded in 96-well plates the day before the treatment with 2-fold dilutions of the compounds. Cell culture media with 1% DMSO was used as negative control.

For cell viability assessment, A549 or HeLa cells were seeded at 5,000 cells per well in white-bottom 96-well plates and incubated overnight. Cells were treated with serial dilutions of the test compound or vehicle control. 48 h post-treatment, luminescence activity was measured using a CellTiter-Glo Luminescent Cell Viability Assay (Promega). Luminescence data values were normalized to the negative control to determine the percent of viral entry inhibition or cytotoxicity values. Normalized values were plotted and analyzed using nonlinear four-parameter variable slope regression to estimate the half-maximum inhibitory activity (EC₅₀) and the 50% cytotoxic concentration (CC₅₀) of each compound. All samples were assayed in triplicate to estimate the standard deviation. The selectivity indices were calculated as the ratio between EC₅₀ and CC₅₀ values.

EBOV and MARV Infectious Virus Assays

HeLa cells sourced from ATCC were seeded into 384-well plates at 4,000 cells per well, 24 hours prior to infection. The control compound, remdesivir, was sourced commercially (Tocris). remdesivir was resuspended from powder to 5 mM in DMSO. Experimental compounds were resuspended to 5 mM in DMSO. All compounds, both control and experimental, underwent a second dilution to 200 μM with DMSO. Media was then aspirated from the 384-well plates and replaced with Dubecco’s Modified Eagle Medium (DMEM, Gibco), with 2% Fetal Bovine Serum (FBS, GeminiBio) and 100 units/mL Penicillin-Streptomycin (P/S, Gibco) at a volume of 40 μL.

Compounds were dispensed into the plate using the I.DOT Liquid Handler (Dispendix). Both compound dilutions were loaded into the I.DOT source plate along with DMSO. The I.DOT was programmed to complete a 10-point dose curve, in triplicate, ranging from 20 μM to 40 nM. Compounds found to have low EC₅₀ values and incomplete dose-response curves were retested using a different I.DOT program to complete a 10-point dose curve starting at 0.625 μM. All wells were backfilled with DMSO to a final volume of 200 nL. To avoid edge effects, the outer wells did not receive any compounds or DMSO. Within the BSL-4 laboratory, either wild-type Ebola Mayinga (EBOV) or Marburg Angola (MARV) was mixed with DMEM, 2% FBS, 1x P/S at an MOI of 0.2, and 10 μL of the inoculum was loaded onto the HeLa plates, one plate for each virus type. All plates were incubated for 40 h at 37 °C. Plates were submerged in 10% neutral-buffered formalin to inactivate the viruses and removed from the BSL4 laboratory for further processing.

After exiting the BSL-4 laboratory, plates were washed three times with 1x phosphate buffered saline (PBS, Fisher Scientific), then membranes were permeabilized in 0.1%v/v Triton X (Acros Organics). Cells were blocked using 3.5%w/v bovine serum albumin (BSA, Sigma Aldrich), and the primary antibody, suspended in 3.5% BSA was applied. The EBOV antibody was Mouse anti-EBOV GP 4F3 (IBT Bioservices), and the MARV antibody was Rabbit anti-MARV VLP (IBT Bioservices). Following incubation with the primary antibody, plates were washed in 1x PBS, and the secondary antibody was applied. For EBOV, the secondary antibody was AlexaFluor488 goat-anti-mouse (Invitrogen), and for MARV, the secondary antibody was AlexaFluor488 goat-anti-rabbit (Invitrogen).

Following the incubation of the secondary antibody, nuclei were stained with Hoescht 33342 (Thermo Fisher) and plates imaged on the Cytation 1 Imaging Reader (BioTek) using the DAPI and GFP channels. The number of antigen-positive infected cells and total cell nuclei were counted using CellProfiler image processing software (Broad Institute, MA). Infection efficiency (rate of infection) was determined using Microsoft Excel and normalized to the DMSO vehicle-treated controls for each compound at each dose. Nuclei count for each compound at each dose was also calculated. Dose–response curves and nuclei counts were fitted using a four-parameter variable slope inhibitor response curve via nonlinear regression using GraphPad Prism 10.2.3.

CYP Inhibition Assay

The assay was conducted as previously described.⁶⁸ A master solution containing phosphate buffer, ultra-pure water, human liver microsomes (Corning. Microsomes Cat. 452117; Lot. 38298), and test compound was prepared for a final compound concentration of 10 μM. For CYP2C9 inhibition, Tolbutamide was added at a final concentration of 200 μM to the above solution. For CYP2D6 inhibition, Dextromethorphan was added at the final concentration of 10 μM to the above solution. For CYP3A4 inhibition, Midazolam was added at the final concentration of 5 μM to the above solution. For CYP1A2 inhibition, Phenacetin was added at the final concentration of 40 μM to the above solution. The mixture was pre-warmed at 37 °C for 5 min and the reaction was initiated by the addition of 10 mM NADPH solution at the final concentration of 1 mM. The reaction was stopped by the addition of 200 μL of cold quench solution (acetonitrile containing internal standard 10 μM Reserpine) at the designated time point (20 min). Experiments were performed in duplicate. Samples were vortexed for 30 seconds and centrifuged at 12,000 RPM for 15 min at 4 °C. Following centrifugation, 100 μl of the supernatant was transferred to an autosampler tray for subsequent LC-MS analysis. Average ratios of test compounds compared to vehicle control were used to calculate percent inhibition values.

Liver Microsome Stability Assay

As previously described,³⁷ the samples at a final concentration of 2 μM were incubated with human (Gibco HMMCPL, Lot no. PL0501) or mouse (Xenotech BALB/c, M3000, Lot no. 2210125) liver microsomes (0.5 mg protein/ml) in the presence of 20 mM NADPH) in 100 mM phosphate buffer at a 37 °C. Reactions were started with the addition of samples and terminated by removing 50 μl aliquots at selected time points (0, 30 and 60 min) and quenching with 200 μl of acetonitrile containing internal standard (0.6 μM Reserpine). Samples were vortexed for 30 seconds and centrifuged at 12,000 RPM for 15 min at 4 °C. Supernatants (100 μl) were transferred to autosampler trays for LC-MS/MS analysis. Experimental controls included Verapamil (lot #LRAD0939) tested under the same conditions as the test articles, incubation without NADPH, and matrix only incubation. All assays were performed in duplicate.

Plasma Stability Assay

As previously described,³⁷ the test articles (2 μM) were incubated with human (BioIVT HUMANPLK2–0000291, Lot no. HMN1218024) or mouse (BioIVT MSE02PLK2–0101314, Lot no. MSE51890) plasmas at 37 °C. Reactions were started with the addition of the test articles and terminated by removing 40 μl aliquots at selected time points (0 and 60 min) and quenching with 160 μl of acetonitrile containing internal standard (0.6 μM Reserpine). Samples were vortexed for 30 seconds and centrifuged at 12,000 RPM for 15 min at 4 °C. Supernatants (100 μl) were transferred to autosampler trays for LC-MS/MS analysis. Experimental controls included Verapamil (lot #LRAD0939) tested under the same conditions as the test articles. All assays were performed in duplicate.

LC-MS analysis was conducted with an Agilent Single Quad GG125B (turbo spray, ESI) and a Waters XBridge BEH C18 (2.1 × 100 mm, 2.5 μm) column. The samples were run on a gradient consisting of MilliQ Water with 0.1% (v) formic acid (mobile phase A) and 30% Acetonitrile (mobile phase B), going from 35% mobile phase B to 95% mobile phase B with a flow rate of 0.35 mL/min. LC-MS/MS analysis was conducted with a Shimadzu triple quadrupole LCMS-8050 (turbo spray, ESI) on a Waters XBridge BEH C18 (2.1 × 50 mm, 2.5 μm) column. The samples were run on a gradient consisting of MilliQ Water with 0.1% (v) formic acid (mobile phase A) and 30% Acetonitrile with 0.1% (v) formic acid (mobile phase B), going from 30% mobile phase B to 90% mobile phase B with a flow rate of 0.25 mL/min. MRM monitoring was performed for all samples.

Solubility Testing

Stock solutions of each compound (10 mM in DMSO) were diluted to 100 μM in Tris-buffered saline (TBS), phosphate-buffered saline (PBS), deionized water, cell culture medium, and simulated gastric fluid. Samples were vortexed and centrifuged (12,400 RPM, 2 min), and the supernatants were analyzed by HPLC (Shimadzu CBM20A, Waters Poroshell 120 EC-C18, 3 × 50 mm, 2.7 μm) using a 20–100% acetonitrile/water gradient containing 0.1% formic acid. Solubility was determined by interpolating peak areas against a 12.5–500 μM calibration curve prepared from serial dilutions of stock solutions in acetonitrile.

Computational experimental details

Molecular docking studies were performed using the X-ray crystal structure of the Ebola virus glycoprotein in complex with toremifene (2) at 2.7 Å resolution (PDB ID: 5JQ7).⁴⁶ A single GP1–GP2 protomer was extracted from the trimeric assembly and used as the receptor model, as the targeted binding site is contained within each protomer. The protein was prepared and protonation was adjusted using the Schrödinger Protein Preparation Wizard. Ligand structures were minimized using LigPrep. Compounds were docked into the toremifene binding site using Glide within Maestro 13.5 (Schrödinger, LLC).

In vivo pharmacokinetic properties

Rapid PK studies were conducted at Pharmaron Beijing Co. Ltd. after ethical consideration and review by Pharmaron’s Institutional Animal Care and Use Committee (IACUC # PK-M-07182024). Pharmacokinetic profiles of compound 41 were determined via administration to male BALB/c mice as 5 mg/kg i.v. bolus (n = 3) and 50 mg/kg p.o. bolus (n = 3). The animals were 6–8 weeks old and weighed 22–28 g at the time of drug administration. Compounds were formulated for dosing as 1 and 5 mg/mL solutions in 20% aqueous 2-Hydroxypropyl-β-cyclodextrin (HP-β-CD), for i.v. and p.o. dosing, respectively. Blood samples following drug administration were collected at 0.083 (i.v. only), 0.25, 0.5, 1, 2, 4, 8, and 24 h. The plasma samples from each animal group were pooled together and analyzed by LC-MS/MS. Calibration curves were constructed from analyzing standard compounds solutions (0.5–1000 ng/mL) in 50% acetonitrile/water diluted 1:3 in blank BALB/c mouse plasma. 15 μL of each of standards, and animal plasma samples were added to 200 μL of acetonitrile containing dexamethasone as an internal standard, for precipitating protein. The samples were vortexed for 30 s, centrifugated at 4 °C, 4000 rpm for 15 min, and the supernatant was diluted 1:3 with water. 6 μL of diluted supernatant was injected into a Shimadzu AB API 5500 LC-MS/MS system for quantitative analysis, employing a Raptor Biphenyl (3 ×30 mm, 2.7 μm) LC column. PK parameters were estimated by non-compartmental model using Phoenix WinNonlin 8.3 software.

All animal experiments were carried out at Pharmaron Beijing Co., Ltd. (IACUC # PK-HAM-06012024) in compliance with the Animal Care Policies and Procedures of UIC. Animal care and experimental procedures were approved by the Institutional Animal Ethics Committee (ACC No 22–205).

Supplementary Material

• EBOV clinical trials drug list, EBOV-GP 3D structure, Compound 41 enantiomers, pseudoviral and infectious assay readouts correlation, Representative dose–response curves, and HPLC trace analysis, NMR spectra and HRMS spectra of key compounds (PDF)

• Molecular string file for all final compounds (CSV)

• 5jq7 Prepared protein for docking simulation (PDB)

Abbreviations

BSL

biosafety level

CDC

Centers for Disease Control and Prevention

EBOV

ebolavirus Zaire species

GP

glycoprotein

MARV

Marburg virus

NPC1

Niemann-Pick Type C1

Pd(dtbpf)Cl₂

[1,1′-bis(di-tert-butylphosphino)ferrocene]dichloropalladium(II)

Pd(dppf)Cl₂

[1,1′-bis(diphenylphosphino)ferrocene]dichloropalladium(II)

pEBOV

pseudotyped ebolavirus Zaire

PK

pharmacokinetic

pMARV

pseudotyped Marburg virus

PyBOP

benzotriazol-1-yloxytripyrrolidinophosphonium hexafluorophosphate

RdRp

RNA-dependent RNA polymerase

RNP

ribonucleoprotein

SAR

structure-activity relationship

SERMs

selective estrogen receptor modulators

SUDV

ebolavirus Sudan species

TLC

thin-layer chromatography

WHO

The World Health Organization