MRK-623

• Molecular Formula: C20H17FN4O
• Molecular Weight: 348.4 g/mol
• InChI: InChI=1S/C20H17FN4O/c1-20(2,26)18-7-9-25-17(12-23-19(25)24-18)13-5-6-16(21)15(10-13)14-4-3-8-22-11-14/h3-12,26H,1-2H3
• InChIKey: GFQMHVYMEKXHDT-UHFFFAOYSA-N

MRK-623: α2/α3-Selective GABAA Agonist

MRK-623 (Compound 14k) is a chemically defined, orally active agonist of the gamma-aminobutyric acid type A (GABAA) receptor, belonging to the imidazo[1,2-a]pyrimidine class of compounds [1]. It functions as a positive allosteric modulator (PAM) at the benzodiazepine binding site, distinguished by its functional selectivity for GABAA receptors containing the α2 and α3 subunits [2]. This profile, often termed 'α1-sparing' due to its near-zero intrinsic efficacy at the α1 subunit, categorizes it as a subtype-selective tool compound for investigating the neuropharmacology of anxiety without the confounding sedative, ataxic, or reinforcing liabilities associated with non-selective benzodiazepine site agonists [3].

α2/α3-subunit functional selectivity

α1-sparing profile for behavioral studies without sedation confound

Orally bioavailable for in vivo anxiety and reinforcement models

Research tool compound for GABAA subtype-specific pharmacology

• [1] Simon C Goodacre, Leslie J Street, David J Hallett, James M Crawforth, Susan Kelly, Andrew P Owens, Wesley P Blackaby, Richard T Lewis, Joanna Stanley, Alison J Smith, Pushpinder Ferris, Bindi Sohal, Susan M Cook, Andrew Pike, Nicola Brown, Keith A Wafford, George Marshall, José L Castro, John R Atack. Imidazo[1,2-a]pyrimidines as functionally selective and orally bioavailable GABA(A)alpha2/alpha3 binding site agonists for the treatment of anxiety disorders. J Med Chem. 2006 Jan 12;49(1):35-8. View Source
• [2] Nina M Shinday, Andrew J R Mawhinney, Michael Z Leonard, Paul W Czoty, Mark A Nader, Michael A Nader, James K Rowlett. Reinforcing effects of compounds lacking intrinsic efficacy at α1 subunit-containing GABAA receptor subtypes in midazolam- but not cocaine-experienced rhesus monkeys. Neuropsychopharmacology. 2013 May;38(6):1006-14. View Source
• [3] John R Atack. GABAA receptor subtype-selective modulators. I. α2/α3-selective agonists as non-sedating anxiolytics. Curr Top Med Chem. 2011;11(9):1176-202. View Source

MRK-623: Why Substitution Fails in Research

Substituting MRK-623 with a seemingly analogous GABAA positive allosteric modulator (PAM) or even another 'α1-sparing' compound is scientifically inadvisable due to profound differences in subunit efficacy profiles that directly translate into distinct behavioral and abuse liability outcomes [1]. The GABAA receptor is a pentameric complex with multiple subunit isoforms (α1-6, β1-3, γ1-3, etc.), and benzodiazepine-site ligands exhibit a spectrum of intrinsic efficacy (degree of receptor activation) across these different subtypes [2]. MRK-623 possesses a unique combination of high binding affinity and a specific functional efficacy fingerprint: near-zero agonism at the α1 subunit, with its highest efficacy concentrated at α2 and α3 subunit-containing receptors [3]. This precise profile contrasts with non-selective agonists (e.g., midazolam, lorazepam) which robustly activate α1 and carry high sedation/abuse risks, and even differs from other α1-sparing compounds like TPA023B (which has efficacy at α5) or TP003 (which has highest efficacy at α3) [3]. These nuanced differences in subtype activation are not academic; they directly correlate with experimental variables such as the degree of self-administration in reinforcement models and the presence of specific motor side effects [1].

Non-selective benzodiazepines (e.g., diazepam, midazolam)

Engage α1 subunit, which may introduce sedation, motor impairment, and abuse-liability confounds in behavioral readouts.

Other α2/α3-preferring agonists (e.g., MRK-898)

Subunit selectivity or pharmacokinetic differences may shift anxiolytic endpoint interpretation; direct interchange not supported.

• [1] Nina M Shinday, Andrew J R Mawhinney, Michael Z Leonard, Paul W Czoty, Mark A Nader, Michael A Nader, James K Rowlett. Reinforcing effects of compounds lacking intrinsic efficacy at α1 subunit-containing GABAA receptor subtypes in midazolam- but not cocaine-experienced rhesus monkeys. Neuropsychopharmacology. 2013 May;38(6):1006-14. View Source
• [2] John R Atack. GABAA receptor subtype-selective modulators. I. α2/α3-selective agonists as non-sedating anxiolytics. Curr Top Med Chem. 2011;11(9):1176-202. View Source
• [3] John R Atack. GABAA receptor alpha2/alpha3 subtype-selective modulators as potential nonsedating anxiolytics. Curr Top Behav Neurosci. 2010;2:331-60. View Source

MRK-623 Differentiation: Quantitative Evidence

Functional Subunit Selectivity Profile

In radioligand binding assays using [3H]flumazenil at human recombinant GABAA receptors, MRK-623 demonstrates high affinity across α1, α2, α3, and α5 subunits [1]. However, its differentiation from a non-selective comparator like MRK-696 lies in the *pattern* of affinity across subtypes. MRK-623 shows a marked selectivity for α5 (Ki = 0.53 nM) and α1 (Ki = 0.85 nM) over α2/α3 (Ki = 3.7-4.0 nM), whereas MRK-696 exhibits a flat, non-selective high-affinity profile across all four subtypes (Ki range: 0.11-0.22 nM) [2].

Submit Selectivity

Head-to-head

α2/α3-selective vs. TPA023B (α2/α3/α5), TP003 (α3)

Supports attribution of α2/α3-mediated behavioral endpoints

Isolates contributions from α5 or α3 alone in experimental design

Evidence Dimension	Binding affinity (Ki)
Target Compound Data	α1: 0.85 nM; α2: 3.7 nM; α3: 4.0 nM; α5: 0.53 nM
Comparator Or Baseline	MRK-696: α1: 0.14 nM; α2: 0.22 nM; α3: 0.11 nM; α5: 0.14 nM
Quantified Difference	MRK-623 affinity for α2 is ~17-fold lower than MRK-696 (3.7 vs 0.22 nM)
Conditions	Human recombinant GABAA receptors (αxβ3γ2) expressed in L(tk-) cells; radioligand binding with [3H]flumazenil [2]

Why This Matters

This difference in binding profile is the molecular basis for the distinct functional and behavioral pharmacology, allowing researchers to probe the role of α2/α3 subunits without the confounding high-affinity α1 binding of non-selective tools.

• [1] Simon C Goodacre, Leslie J Street, David J Hallett, James M Crawforth, Susan Kelly, Andrew P Owens, Wesley P Blackaby, Richard T Lewis, Joanna Stanley, Alison J Smith, Pushpinder Ferris, Bindi Sohal, Susan M Cook, Andrew Pike, Nicola Brown, Keith A Wafford, George Marshall, José L Castro, John R Atack. Imidazo[1,2-a]pyrimidines as functionally selective and orally bioavailable GABA(A)alpha2/alpha3 binding site agonists for the treatment of anxiety disorders. J Med Chem. 2006 Jan 12;49(1):35-8. View Source
• [2] Nina M Shinday, Andrew J R Mawhinney, Michael Z Leonard, Paul W Czoty, Mark A Nader, Michael A Nader, James K Rowlett. Reinforcing effects of compounds lacking intrinsic efficacy at α1 subunit-containing GABAA receptor subtypes in midazolam- but not cocaine-experienced rhesus monkeys. Neuropsychopharmacology. 2013 May;38(6):1006-14. View Source

Subtype Binding Affinity (Ki) Comparison

In a direct head-to-head comparison using a progressive-ratio schedule of intravenous self-administration in rhesus monkeys with a midazolam-training history, the reinforcing potency (ED50) of MRK-623 was quantified and compared to the non-selective partial agonist MRK-696 [1].

Binding Affinity (Ki)

Head-to-head

α1 0.85 nM, α2 3.70 nM, α3 4.00 nM, α5 0.53 nM

Distinct affinity fingerprint vs. non-selective MRK-696 (α2 0.22, α3 0.11 nM)

[³H]flumazenil binding, human recombinant receptors

Evidence Dimension	Reinforcing potency (ED50 for self-administration)
Target Compound Data	ED50 = 0.09 ± 0.01 mg/kg/injection
Comparator Or Baseline	MRK-696: ED50 = 0.0011 ± 0.0009 mg/kg/injection
Quantified Difference	MRK-623's ED50 is approximately 82-fold higher than MRK-696, indicating a substantially lower reinforcing potency and a reduced abuse liability profile.
Conditions	Rhesus monkeys (N=4) trained to self-administer midazolam (0.03 mg/kg/inj) under a progressive-ratio schedule of i.v. drug injection [1]

Why This Matters

For researchers modeling anxiety, this directly quantifies the reduced risk of confounding reinforcement effects with MRK-623 compared to a non-selective GABAA PAM, enabling cleaner interpretation of anxiolytic efficacy in behavioral assays.

• [1] Nina M Shinday, Andrew J R Mawhinney, Michael Z Leonard, Paul W Czoty, Mark A Nader, Michael A Nader, James K Rowlett. Reinforcing effects of compounds lacking intrinsic efficacy at α1 subunit-containing GABAA receptor subtypes in midazolam- but not cocaine-experienced rhesus monkeys. Neuropsychopharmacology. 2013 May;38(6):1006-14. View Source

In Vivo Anxiolytic Potency & BZ Site Occupancy

Using whole-cell patch-clamp electrophysiology in L(tk-) cells expressing human recombinant GABAA receptors, the functional efficacy of MRK-623 was determined across α1, α2, α3, and α5 subunit-containing receptor subtypes [1]. MRK-623 exhibits a unique functional fingerprint compared to other α1-sparing compounds: it has near-zero efficacy at α1, its highest efficacy at α2/α3, and lower efficacy at α5 [2]. This contrasts with TPA023B, which retains significant efficacy at α5, and TP003, which has its highest efficacy at α3 with near-zero efficacy at α2 and α5 [3].

BZ Site Occupancy

Reported

78% occupancy at 3 mg/kg p.o.; ID50 0.8 mg/kg

Dose-occupancy benchmark for in vivo anxiolytic endpoint studies

Rat elevated plus maze, oral gavage

Evidence Dimension	Functional efficacy profile across GABAA subtypes
Target Compound Data	Near-zero efficacy at α1; highest efficacy at α2/α3; lower efficacy at α5
Comparator Or Baseline	TPA023B: functional selectivity for α2/α3/α5; TP003: functional selectivity for α3 (near-zero at α2, α5); MRK-696: no differences in efficacy across subtypes
Quantified Difference	Qualitative difference in subtype efficacy profile; specifically, MRK-623's combination of α2/α3 efficacy with low α5 efficacy distinguishes it from TPA023B.
Conditions	Human recombinant GABAA receptors (αxβ3γ2) expressed in L(tk-) cells; whole-cell patch-clamp electrophysiology [1]

Why This Matters

The specific α2/α3 functional selectivity, combined with the lack of α5 efficacy, makes MRK-623 a more precise tool than TPA023B for isolating the roles of α2 and α3 subunits in anxiolysis and other behaviors, minimizing potential confounding effects from α5 modulation (which is implicated in learning, memory, and sedation).

• [1] Simon C Goodacre, Leslie J Street, David J Hallett, James M Crawforth, Susan Kelly, Andrew P Owens, Wesley P Blackaby, Richard T Lewis, Joanna Stanley, Alison J Smith, Pushpinder Ferris, Bindi Sohal, Susan M Cook, Andrew Pike, Nicola Brown, Keith A Wafford, George Marshall, José L Castro, John R Atack. Imidazo[1,2-a]pyrimidines as functionally selective and orally bioavailable GABA(A)alpha2/alpha3 binding site agonists for the treatment of anxiety disorders. J Med Chem. 2006 Jan 12;49(1):35-8. View Source
• [2] Nina M Shinday, Andrew J R Mawhinney, Michael Z Leonard, Paul W Czoty, Mark A Nader, Michael A Nader, James K Rowlett. Reinforcing effects of compounds lacking intrinsic efficacy at α1 subunit-containing GABAA receptor subtypes in midazolam- but not cocaine-experienced rhesus monkeys. Neuropsychopharmacology. 2013 May;38(6):1006-14. View Source
• [3] John R Atack. GABAA receptor subtype-selective modulators. I. α2/α3-selective agonists as non-sedating anxiolytics. Curr Top Med Chem. 2011;11(9):1176-202. View Source

Self-Administration Reinforcing Effects

Following oral administration, MRK-623 achieves dose-dependent occupancy of the benzodiazepine binding site on GABAA receptors in the rat brain, providing a direct link between administered dose and target engagement in the central nervous system [1].

Self-Administration ED50

Head-to-head

0.09 mg/kg/inj (vs MRK-696: 0.0011 mg/kg/inj)

Indicates lower reinforcing potency; relevant to abuse-liability profiling

Rhesus monkey IV self-administration, midazolam baseline

Evidence Dimension	Benzodiazepine site occupancy in rat brain
Target Compound Data	78% occupancy at 3 mg/kg (oral); ID50 = 0.8 mg/kg
Comparator Or Baseline	Baseline: 0% occupancy at vehicle control
Quantified Difference	Achieves 78% receptor occupancy at a 3 mg/kg oral dose, with a calculated half-maximal occupancy dose (ID50) of 0.8 mg/kg.
Conditions	Rat; oral gavage administration; measurement via in vivo [3H]flumazenil binding assay [1]

Why This Matters

This data allows researchers to confidently select oral doses that achieve substantial (>50%) target engagement for behavioral studies, ensuring that observed anxiolytic effects are linked to GABAA receptor modulation and not due to off-target activities or insufficient exposure.

• [1] Simon C Goodacre, Leslie J Street, David J Hallett, James M Crawforth, Susan Kelly, Andrew P Owens, Wesley P Blackaby, Richard T Lewis, Joanna Stanley, Alison J Smith, Pushpinder Ferris, Bindi Sohal, Susan M Cook, Andrew Pike, Nicola Brown, Keith A Wafford, George Marshall, José L Castro, John R Atack. Imidazo[1,2-a]pyrimidines as functionally selective and orally bioavailable GABA(A)alpha2/alpha3 binding site agonists for the treatment of anxiety disorders. J Med Chem. 2006 Jan 12;49(1):35-8. View Source

MRK-623 Research Application Scenarios

Isolating α2 vs. α3 Subunit Functions

Researchers seeking to isolate the specific contribution of α2- and α3-containing GABAA receptors to anxiety-related behaviors should select MRK-623 as a primary tool compound. Its near-zero efficacy at the α1 subunit (the primary mediator of benzodiazepine-induced sedation and ataxia) and its distinct functional profile at α5 (unlike TPA023B) allow for the interrogation of anxiolytic pathways with minimal confounding by motor impairment or memory effects [1]. The established oral bioavailability and target engagement data (78% occupancy at 3 mg/kg, p.o.) [2] provide a validated dosing paradigm for rodent anxiety models such as the elevated plus maze and fear conditioning.

Abuse Liability and Reinforcement Studies

For behavioral pharmacologists studying the neurobiological underpinnings of substance use disorders, MRK-623 serves as a critical comparator in reinforcement studies. Its direct, head-to-head comparison against non-selective GABAA PAMs like MRK-696 (ED50 0.09 vs. 0.0011 mg/kg/inj) [3] provides a quantitative benchmark for reduced abuse potential. Procurement of MRK-623 is essential for studies designed to test the hypothesis that eliminating α1-subunit efficacy reduces the reinforcing properties of benzodiazepine-site ligands in both drug-naïve and drug-experienced non-human primate models [3].

In Vivo Receptor Occupancy Neuroimaging

Electrophysiology laboratories investigating the pharmacology of recombinant GABAA receptors should utilize MRK-623 as a reference agonist for α2/α3-containing complexes. Its functional efficacy profile, as established via whole-cell patch-clamp recordings in L(tk-) cells [1], makes it an ideal positive control for validating the functional expression of α2β3γ2 and α3β3γ2 receptors. Its contrasting profile to compounds like TP003 (α3-selective) and TPA023B (α2/α3/α5-selective) [4] enables a panel approach to map the subtype composition of native neuronal GABAA receptors in brain slice preparations.

Benchmarking Target Engagement for Novel Anxiolytic Candidates

In industrial drug discovery settings, MRK-623 provides a well-characterized benchmark for assessing target engagement of novel, orally bioavailable anxiolytic candidates. Its in vivo receptor occupancy profile (ID50 of 0.8 mg/kg, p.o.) [2] serves as a performance standard when using [3H]flumazenil or PET imaging ligands to measure CNS penetration and GABAA receptor binding of new chemical entities. Comparing a novel compound's occupancy data to MRK-623's established parameters offers a quantifiable, data-driven method for prioritizing candidates with optimal brain exposure and on-target pharmacology.

Application

Selection Property

Validation Focus

GABAA subunit deconvolution research

α2/α3-selective functional profile

Submit-specific pharmacology in anxiety models

Abuse-liability behavioral pharmacology

Well-characterized self-administration profile

Reinforcement endpoint benchmarking vs. non-selective BZDs

Receptor occupancy imaging (PET/SPECT)

Reported dose-occupancy relationship

Target engagement calibration without sedation confound

• [1] Simon C Goodacre, Leslie J Street, David J Hallett, James M Crawforth, Susan Kelly, Andrew P Owens, Wesley P Blackaby, Richard T Lewis, Joanna Stanley, Alison J Smith, Pushpinder Ferris, Bindi Sohal, Susan M Cook, Andrew Pike, Nicola Brown, Keith A Wafford, George Marshall, José L Castro, John R Atack. Imidazo[1,2-a]pyrimidines as functionally selective and orally bioavailable GABA(A)alpha2/alpha3 binding site agonists for the treatment of anxiety disorders. J Med Chem. 2006 Jan 12;49(1):35-8. View Source
• [2] Simon C Goodacre, Leslie J Street, David J Hallett, James M Crawforth, Susan Kelly, Andrew P Owens, Wesley P Blackaby, Richard T Lewis, Joanna Stanley, Alison J Smith, Pushpinder Ferris, Bindi Sohal, Susan M Cook, Andrew Pike, Nicola Brown, Keith A Wafford, George Marshall, José L Castro, John R Atack. Imidazo[1,2-a]pyrimidines as functionally selective and orally bioavailable GABA(A)alpha2/alpha3 binding site agonists for the treatment of anxiety disorders. J Med Chem. 2006 Jan 12;49(1):35-8. View Source
• [3] Nina M Shinday, Andrew J R Mawhinney, Michael Z Leonard, Paul W Czoty, Mark A Nader, Michael A Nader, James K Rowlett. Reinforcing effects of compounds lacking intrinsic efficacy at α1 subunit-containing GABAA receptor subtypes in midazolam- but not cocaine-experienced rhesus monkeys. Neuropsychopharmacology. 2013 May;38(6):1006-14. View Source
• [4] John R Atack. GABAA receptor subtype-selective modulators. I. α2/α3-selective agonists as non-sedating anxiolytics. Curr Top Med Chem. 2011;11(9):1176-202. View Source