BNS808

• Molecular Formula: C25H20Cl3N3O3S
• Molecular Weight: 548.9 g/mol
• InChI: InChI=1S/C25H20Cl3N3O3S/c26-17-9-7-16(8-10-17)24-25-23(29-31(24)22-6-2-1-4-20(22)28)21(5-3-15-34-25)30-35(32,33)19-13-11-18(27)12-14-19/h1-2,4,6-14,21,30H,3,5,15H2
• InChIKey: VSSQVUBXWXIWJG-UHFFFAOYSA-N

BNS808 Peripherally Selective CB1R Antagonist Profile

BNS808 (CAS: 2836313‑12‑1) is an orally active, peripherally selective cannabinoid‑1 receptor (CB1R) antagonist built on a tricyclic scaffold, with molecular weight 548.9 g/mol and formula C₂₅H₂₀Cl₃N₃O₃S [1]. It exhibits sub‑nanomolar binding affinity (Ki = 0.7 nM for both mouse and human CB1R, hIC₅₀ = 0.8 nM) and >2000‑fold selectivity over the CB2 receptor (hCB2R Ki = 1930 nM) [1]. Unlike brain‑penetrant CB1R antagonists such as rimonabant, BNS808 was specifically designed to minimize blood–brain barrier penetration (brain‑to‑plasma ratio of 0.18 at 1 h post‑dose) while retaining potent peripheral anti‑obesity and anti‑steatotic efficacy in diet‑induced obese (DIO) mice [1]. Patent protection is in place under PCT/IL2020/050062 and PCT/IL2022/050276 [1].

1

Peripheral CB1R mechanistic studies — enables dissociation of peripheral metabolic endpoints from central receptor engagement in diet-induced obesity models.

2

High CB1R/CB2R selectivity — supports CB1R-specific assay interpretation with low CB2R-mediated off-target confound.

3

Low brain penetration profile — brain-to-plasma ratio substantially below 1 under reported conditions, minimizing central CB1R confound in metabolic readouts.

• [1] Gammal A, Nassar T, Soae Y, Freeman N, Badihi A, Baraghithy S, Nemirovski A, Tam J, Benita S. Synthesis and Pharmacological Characterization of Novel Peripheral Cannabinoid‑1 Receptor Blockers Based on a Tricyclic Scaffold. J Med Chem. 2025 May 8;68(9):9431‑9445. doi: 10.1021/acs.jmedchem.4c03132. View Source

Why Generic CB1R Antagonists Cannot Substitute for BNS808

Within the CB1R antagonist class, compounds span a wide continuum from fully brain‑penetrant agents (e.g., rimonabant) to peripherally restricted candidates (e.g., AM6545, JD‑5037), yet they cannot be interchanged without compromising experimental validity or safety outcomes [1]. BNS808 occupies a distinct position on this continuum: it delivers sub‑nanomolar CB1R affinity combined with CB2R selectivity exceeding 2000‑fold—a margin that is 7‑fold greater than rimonabant (~285‑fold) and >2.5‑fold greater than JD‑5037 (>700‑fold) when compared across available cross‑study binding data [1]. Critically, its brain‑to‑plasma ratio of 0.18 and absence of CNS‑mediated behavioural effects were demonstrated in direct head‑to‑head experiments against rimonabant within the same study, providing a level of quantitative peripheral‑restriction evidence that is unavailable for many other peripherally restricted CB1R antagonists when referenced against a common brain‑penetrant comparator [1]. Substituting BNS808 with a less‑selective or insufficiently characterised peripherally restricted analog introduces uncontrolled variables in CB2R‑mediated off‑target activity and CNS exposure that can confound data interpretation and regulatory dossier building.

!

Brain penetration ratio differs substantially — most CB1R antagonists lack quantitative peripheral-restriction evidence against a common brain-penetrant comparator; substituting may reintroduce CNS-mediated confounds.

!

CB2R selectivity windows are not interchangeable — inferior CB2R selectivity in alternatives can introduce uncontrolled off-target signals in CB1R-focused functional assays.

!

CYP interaction profile differs qualitatively — BNS808 acts as a CYP inhibitor, while rimonabant is a CYP3A4 substrate; substitution may alter drug-drug interaction liabilities in co-administration studies.

• [1] Gammal A, Nassar T, Soae Y, Freeman N, Badihi A, Baraghithy S, Nemirovski A, Tam J, Benita S. Synthesis and Pharmacological Characterization of Novel Peripheral Cannabinoid‑1 Receptor Blockers Based on a Tricyclic Scaffold. J Med Chem. 2025 May 8;68(9):9431‑9445. doi: 10.1021/acs.jmedchem.4c03132. View Source

Quantitative Differentiation Evidence

CB1R Binding Affinity and Functional Antagonism

BNS808 binds to both mouse and human CB1R with identical Ki values of 0.7 nM and a functional hIC₅₀ of 0.8 nM, placing it among the highest‑affinity peripherally selective CB1R antagonists reported [1]. In cross‑study comparison, this affinity is 2.6‑fold higher than rimonabant (Ki = 1.8 nM ), 2.4‑fold higher than AM6545 (Ki = 1.7 nM ), and approximately 2‑fold lower than JD‑5037 (Ki = 0.35 nM )—however, BNS808 achieves this potency while maintaining substantially greater CB2R selectivity than any of these comparators (see Evidence Item 2), a differential that directly impacts assay specificity and off‑target data interpretation.

CB1R Affinity

Context-dependent

Ki = 0.7 nM (m/h)
IC₅₀ = 0.8 nM (h)

Supports high-sensitivity CB1R engagement in cell-based assays.

Cross-study comparison; values from [³H]‑CP55940 binding and GTPγS functional assay.

Evidence Dimension	CB1R binding affinity (Ki) and functional antagonism (IC₅₀)
Target Compound Data	BNS808: mKi = 0.7 nM; hKi = 0.7 nM; hIC₅₀ = 0.8 nM (GTPγS assay)
Comparator Or Baseline	Rimonabant: hKi = 1.8 nM; AM6545: CB1 Ki = 1.7 nM; JD‑5037: CB1 Ki = 0.35 nM
Quantified Difference	BNS808 affinity is 2.6× higher than rimonabant, 2.4× higher than AM6545; 2× lower than JD‑5037 but with >2.8× greater CB2R selectivity than JD‑5037
Conditions	[³H]‑CP55940 competitive binding; hCB1R‑transfected HEK293 membranes; GTPγS functional assay

Why This Matters

Higher affinity at lower concentrations can reduce total compound consumption in screening campaigns and may improve signal‑to‑noise ratios in CB1R‑dependent cellular assays.

• [1] Gammal A, et al. J Med Chem. 2025;68(9):9431‑9445. Table 1: BNS808 mKi 0.7 nM, hKi 0.7 nM, hIC₅₀ 0.8 nM; Table 2: BNS808 hKi 0.7 nM, hIC₅₀ 0.8 nM. View Source

CB2R Selectivity and Assay Specificity

BNS808 exhibits an hCB2R Ki of 1930 nM, yielding a CB1R/CB2R selectivity ratio of approximately 2757‑fold (1930 nM / 0.7 nM), formally reported as >2000‑fold [1]. This selectivity margin is 7‑fold greater than rimonabant (~285‑fold; CB2 Ki = 514 nM ), 9‑fold greater than AM6545 (~308‑fold; CB2 Ki = 523 nM ), and >2.8‑fold greater than JD‑5037 (>700‑fold ). Among the in‑class BNS series compounds directly compared in Gammal et al., BNS808 displayed the highest selectivity: BNS825 yielded only 41‑fold selectivity (CB2 hKi = 127 nM), BNS815 yielded 694‑fold, and BNS807 yielded 1179‑fold [1].

CB2R Selectivity

Context-dependent

CB2 Ki = 1930 nM
Selectivity >2000-fold

Provides low CB2R off-target confound for CB1R-specific pathway interpretation.

Selectivity margin exceeds rimonabant (~7×) and AM6545 (~9×); in-class comparison from same study.

Evidence Dimension	CB1R/CB2R selectivity ratio
Target Compound Data	BNS808: hCB2R Ki = 1930 nM; selectivity ratio = 2757‑fold (>2000‑fold reported)
Comparator Or Baseline	Rimonabant: CB2 Ki = 514 nM, ~285‑fold; AM6545: CB2 Ki = 523 nM, ~308‑fold; JD‑5037: >700‑fold; BNS825: CB2 Ki = 127 nM, 41‑fold
Quantified Difference	BNS808 selectivity exceeds rimonabant by ~7×, AM6545 by ~9×, JD‑5037 by >2.8×, and BNS825 by ~67×
Conditions	[³H]‑CP55940 competitive binding; hCB2R‑transfected HEK293 membranes

Why This Matters

A >2000‑fold selectivity window minimises CB2R‑mediated confounding signals in CB1R‑focused functional assays and reduces the probability that observed phenotypic effects arise from unintended CB2R engagement, thereby lowering the risk of false‑positive hits in obesity and metabolic disease screening cascades.

• [1] Gammal A, et al. J Med Chem. 2025;68(9):9431‑9445. Table 1: BNS808 hCB2R Ki = 1930 nM; BNS825 hCB2R Ki = 127 nM. Discussion: BNS808 >2000‑fold selectivity. View Source

Brain Penetration vs. Rimonabant

In a direct head‑to‑head in vivo biodistribution study, lean C57Bl/6 mice received a single oral dose of 1 mg/kg of either BNS808 or rimonabant. At 1 h post‑administration, BNS808 achieved a brain concentration of approximately 13.7 ng/g and a brain‑to‑plasma (B/P) ratio of 0.18, while rimonabant reached 128 ng/g in brain with a B/P ratio of 4.28 [1]. This represents a 9.3‑fold lower absolute brain concentration and a 23.8‑fold lower B/P ratio for BNS808 relative to rimonabant under identical dosing and sampling conditions. After chronic 24‑day oral dosing (1 mg/kg/day), BNS808 brain levels remained at approximately 20 ng/g [1]. BNS808 preferentially distributed to the liver (liver/plasma ratio of 3.62 at 1 h) [1].

Brain Penetration

Head-to-head

BNS808: Brain 13.7 ng/g, B/P 0.18
Rimonabant: Brain 128 ng/g, B/P 4.28

Low brain exposure under identical dosing; supports peripheral-restriction interpretation.

1 mg/kg PO, 1 h post-dose in C57Bl/6 mice; n≥3.

Evidence Dimension	Brain concentration and brain‑to‑plasma (B/P) concentration ratio
Target Compound Data	BNS808: brain = 13.7 ng/g; B/P ratio = 0.18 at 1 h post 1 mg/kg PO; chronic brain ~20 ng/g; liver/plasma ratio = 3.62
Comparator Or Baseline	Rimonabant: brain = 128 ng/g; B/P ratio = 4.28 at 1 h post 1 mg/kg PO; liver/plasma ratio = 7.4
Quantified Difference	9.3‑fold lower brain concentration; 23.8‑fold lower B/P ratio for BNS808 vs. rimonabant
Conditions	Lean male C57Bl/6 mice; acute oral gavage 1 mg/kg; drug levels measured at 1, 2, 4, 8 h post‑dose; n ≥ 3 per group

Why This Matters

A B/P ratio substantially below 1 and absolute brain levels 9‑10× lower than rimonabant under identical experimental conditions provide quantitative confidence that BNS808 will not confound in vivo metabolic efficacy readouts through unintended central CB1R engagement—a critical criterion for peripherally restricted mechanism‑of‑action studies.

• [1] Gammal A, et al. J Med Chem. 2025;68(9):9431‑9445. Figure 1: BNS808 brain ~13.7 ng/g, B/P ratio 0.18, liver/plasma ratio 3.62; rimonabant brain 128 ng/g, B/P ratio 4.28. Results text: brain penetrance limited, average 13.7 ng/g. View Source

CNS Side-Effect Liability vs. Rimonabant

BNS808 was assessed for CNS‑mediated behavioural effects in direct comparison with the brain‑penetrant CB1R antagonist rimonabant. At doses of 1 and 10 mg/kg PO in C57Bl/6J mice, BNS808 did not induce any significant changes in locomotor activity, whereas rimonabant (10 mg/kg PO) produced pronounced hyperactivity [1]. BNS808 was unable to antagonise the hypoactivity induced by the CB1R agonist HU‑210 (30 μg/kg IP), and only rimonabant (10 mg/kg PO) blocked the cataleptic behaviour mediated by WIN‑55,212—a brain‑CB1R‑dependent effect [1]. Correspondingly, at 1 h post‑dose, brain rimonabant levels were 600 ng/g versus only 13 ng/g for BNS808 [1].

CNS Behaviour

Head-to-head

BNS808: No locomotor, hypoactivity reversal, or catalepsy changes
Rimonabant: Hyperactivity, reversed hypoactivity, blocked catalepsy

Functional dissociation from central CB1R-mediated behavioural effects in mouse model.

1–10 mg/kg PO; Promethion system; brain rimonabant 600 ng/g vs BNS808 13 ng/g.

Evidence Dimension	CNS‑mediated behavioural effects (locomotor activity, agonist‑induced hypoactivity reversal, catalepsy blockade)
Target Compound Data	BNS808: no locomotor changes at 1 or 10 mg/kg PO; failed to reverse HU‑210 hypoactivity; failed to block WIN‑55,212 catalepsy; brain level 13 ng/g
Comparator Or Baseline	Rimonabant: significant hyperactivity at 10 mg/kg PO; reversed HU‑210 hypoactivity; blocked WIN‑55,212 catalepsy; brain level 600 ng/g
Quantified Difference	Complete absence of three CNS‑mediated behaviours for BNS808 vs. positive detection for rimonabant; 46‑fold lower brain drug level for BNS808
Conditions	Wild‑type male C57Bl/6J mice; Promethion High‑Definition Behavioural Phenotyping System; n = 4–8 per group

Why This Matters

The functional demonstration that BNS808 does not recapitulate rimonabant's CNS side‑effect profile—even at doses 10× the efficacious metabolic dose—provides direct experimental evidence for peripheral restriction. For procurement purposes, this validates BNS808 as a tool compound that can dissociate peripheral metabolic benefits from central neuropsychiatric liabilities, eliminating the safety confound that led to rimonabant's market withdrawal.

• [1] Gammal A, et al. J Med Chem. 2025;68(9):9431‑9445. Figure 2: BNS808 vs. rimonabant locomotor activity (A–C), HU‑210 hypoactivity reversal (D–F), WIN‑55,212 catalepsy blockade (G), brain drug levels (H). View Source

Cardiotoxicity and Hepatocellular Safety Margins

BNS808 demonstrated a hERG IC₅₀ of 5.39 μM against the cardiac potassium channel, yielding a safety margin of approximately 6,737‑fold relative to its CB1R functional IC₅₀ of 0.8 nM—substantially exceeding the standard acceptance criterion of >100‑fold [1]. In HepG2 hepatocyte viability assays, BNS808 showed a relative IC₅₀ of 16.84 μM, corresponding to a therapeutic selectivity index of approximately 24,057‑fold (16.84 μM / 0.7 nM) [1]. The compound was non‑mutagenic in the mini‑Ames assay at doses from 31.25 to 1000 μg/well across TA98 and TA100 strains both with and without S9 metabolic activation [1]. By comparison, the reference toxic compound staurosporine exhibited a HepG2 IC₅₀ of 0.06 μM—280‑fold more cytotoxic than BNS808 [1].

Cardiac/Liver Safety

Class-level

hERG IC₅₀ = 5.39 μM
HepG2 IC₅₀ = 16.84 μM

Reported hERG and hepatocyte viability endpoint context; high margin relative to CB1R IC₅₀.

Mini-Ames non-mutagenic; staurosporine HepG2 IC₅₀ 0.06 μM.

Evidence Dimension	hERG channel inhibition, hepatocyte cytotoxicity, and bacterial mutagenicity
Target Compound Data	BNS808: hERG IC₅₀ = 5.39 μM (safety margin ~6,737×); HepG2 IC₅₀ = 16.84 μM (selectivity index ~24,057×); mini‑Ames: non‑mutagenic
Comparator Or Baseline	Staurosporine: HepG2 IC₅₀ = 0.06 μM. Acceptance criterion: hERG IC₅₀ > 100 × CB1R IC₅₀. BNS807: hERG IC₅₀ = 18.04 μM.
Quantified Difference	BNS808 hERG safety margin 67‑fold above acceptance threshold; 280‑fold less cytotoxic than staurosporine; selectivity index >24,000× for CB1R over hepatocyte toxicity
Conditions	Automated patch clamp (SyncroPatch 384PE) for hERG; HepG2 cell viability (CellTiter‑Glo); mini‑Ames fluctuation assay (TA98, TA100 ± S9)

Why This Matters

A high hERG safety margin and a hepatocyte selectivity index exceeding 24,000‑fold reduce the probability of compound‑attributable cardiac or hepatic toxicity signals during in vivo efficacy studies, enabling cleaner interpretation of metabolic phenotype data and lowering compound‑related attrition risk in preclinical development pipelines.

• [1] Gammal A, et al. J Med Chem. 2025;68(9):9431‑9445. Table 3: BNS808 hERG IC₅₀ 5.39 μM; HepG2 IC₅₀ 16.84 μM; mini‑Ames non‑mutagenic. Text: selectivity index discussion. View Source

CYP Inhibition and Drug-Drug Interaction Risk

In human liver microsome assays, BNS808 exhibited moderate inhibition of CYP3A4 (IC₅₀ = 2.99 μM), weak‑to‑moderate inhibition of CYP2C9 (IC₅₀ = 8.33 μM) and CYP2C19 (IC₅₀ = 18.0 μM), and weak inhibition of CYP1A2 and CYP2D6 (IC₅₀ > 50 μM) [1]. In contrast, rimonabant is itself a substrate of CYP3A4 and amidohydrolase, and its systemic exposure is sensitive to co‑administered CYP3A4 modulators (e.g., ketoconazole increased rimonabant AUC by 104%) [2], creating a bidirectional DDI liability. The authors explicitly note that BNS808 has a reduced likelihood of causing significant drug–drug interactions compared to rimonabant [1].

CYP Inhibition

Class-level

CYP3A4 IC₅₀ = 2.99 μM
CYP2C9 IC₅₀ = 8.33 μM
CYP2C19 IC₅₀ = 18.0 μM

CYP inhibition profile may support DDI risk assessment in co-administration studies.

Rimonabant is a CYP3A4 substrate, creating distinct DDI profile.

Evidence Dimension	CYP450 isoform inhibition profile and DDI liability
Target Compound Data	BNS808: CYP3A4 IC₅₀ = 2.99 μM (moderate); CYP2C9 IC₅₀ = 8.33 μM; CYP2C19 IC₅₀ = 18.0 μM; CYP1A2/CYP2D6 IC₅₀ > 50 μM
Comparator Or Baseline	Rimonabant: CYP3A4 substrate; AUC increased 104% by ketoconazole; metabolised via CYP3A and amidohydrolase
Quantified Difference	BNS808 is a CYP inhibitor (not a sensitive substrate); rimonabant is a CYP3A4 substrate with exposure sensitive to CYP3A4 modulators—qualitatively distinct DDI risk profiles
Conditions	Human liver microsomes; isoform‑selective probe substrates; LC‑MS/MS quantification

Why This Matters

In research settings requiring co‑administration with CYP‑modulated compounds (e.g., in polypharmacology or disease model studies), BNS808's CYP inhibition profile—rather than substrate susceptibility—offers a more predictable pharmacokinetic behaviour, reducing the risk of uncontrolled drug–drug interaction confounds.

• [1] Gammal A, et al. J Med Chem. 2025;68(9):9431‑9445. CYP inhibition section: BNS808 IC₅₀ values; discussion comparing DDI risk vs. rimonabant. View Source
• [2] Rimonabant. RxReasoner drug interaction summary. CYP3A4 substrate; ketoconazole co‑administration increases AUC by 104%. View Source

Optimal Research and Industrial Applications

Dissociating Peripheral vs. Central CB1R in DIO Models

BNS808 is the compound of choice for studies requiring unambiguous separation of peripheral metabolic effects from central CB1R‑mediated confounds. In DIO C57Bl/6J mice (60% high‑fat diet, 16‑week induction), chronic oral BNS808 at 1 mg/kg/day for 24 days reduced body weight, decreased total body fat, increased lean body mass, ameliorated hypertriglyceridemia and hypercholesterolemia, reversed hepatic steatosis, and lowered ALT, AST, and ALP liver enzymes—all without detectable CNS‑mediated behavioural effects [1]. This profile enables researchers to attribute metabolic improvements to peripheral CB1R blockade with a level of confidence not achievable with brain‑penetrant comparators.

MASLD Target Validation and Drug Discovery

BNS808's preferential liver distribution (liver/plasma ratio of 3.62 at 1 h post‑dose [1]) and demonstrated reversal of hepatic steatosis—histologically confirmed reduction in fat vacuole deposition, significant decrease in liver triglyceride content, and reduction in hepatocellular injury markers ALT, AST, and ALP —position it as a high‑value tool compound for target validation studies focused on hepatic CB1R in MASLD/MASH. The compound's favourable hepatocyte safety margin (selectivity index >24,000‑fold [1]) further supports its use in chronic dosing paradigms where liver health is the primary endpoint.

CB1R-Mediated DDI Liability Assessments

BNS808's CYP inhibition profile—moderate CYP3A4 (IC₅₀ = 2.99 μM), weak‑to‑moderate CYP2C9 (IC₅₀ = 8.33 μM) and CYP2C19 (IC₅₀ = 18.0 μM), with weak CYP1A2/CYP2D6 inhibition [1]—contrasts with rimonabant's status as a CYP3A4 substrate susceptible to pharmacokinetic modulation by CYP3A4 inhibitors and inducers [2]. This makes BNS808 the preferred CB1R antagonist for preclinical studies where co‑administration with CYP‑modulated agents is required, as the compound acts as a perpetrator (inhibitor) rather than a victim (substrate) of CYP‑mediated interactions, yielding a more tractable DDI risk assessment.

Reference Standard for High-Throughput Screening Selectivity

With a CB1R/CB2R selectivity ratio exceeding 2000‑fold—the highest reported among major peripherally restricted CB1R antagonists [1]—and confirmed neutral antagonism (no inverse agonist activity in GTPγS assays [1]), BNS808 serves as an ideal reference compound for calibrating CB1R‑selective screening assays and for benchmarking the selectivity of new chemical entities. Its >2000‑fold window ensures that CB2R‑driven false positives are minimised when used as a positive control in CB1R‑targeted high‑throughput screens.

Application

Selection Property

Validation Focus

Peripheral CB1R mechanistic studies (DIO model)

Peripheral restriction profile

Body weight, fat mass, and liver enzyme endpoints in diet-induced obese mice

MASLD target validation research

Preferential liver distribution and hepatocyte safety margin

Liver triglyceride content and hepatocellular injury marker reduction

Drug-drug interaction liability studies

CYP inhibition profile (perpetrator, not substrate)

CYP3A4/2C9/2C19 interaction risk in co-administration paradigms

High-throughput screening selectivity reference

CB1R/CB2R selectivity window

Minimization of CB2R-driven false positives in CB1R-targeted assays

• [1] Gammal A, et al. J Med Chem. 2025;68(9):9431‑9445. Figures 1–6; Tables 1–3; Results and Discussion. View Source
• [2] Rimonabant. RxReasoner drug interaction summary. CYP3A4 substrate; exposure modulated by CYP3A4 inhibitors/inducers. View Source