Abstract
This study explores the influence of pH on the structure and bioavailability of MIDD0301, an orally administered lead compound designed for asthma treatment. MIDD0301, a novel benzodiazepine derivative, targets peripheral GABAA receptors to alleviate lung inflammation and airway smooth muscle constriction. The unique molecular architecture of MIDD0301 incorporates both a basic imidazole and a carboxylic acid within a diazepine framework, resulting in enhanced solubility at physiological pH. Our findings reveal that MIDD0301 exhibits pH-dependent structural interconversion, transitioning between a seven-membered ring form at neutral pH and an acyclic configuration at or below pH 3. Both structural forms exist as two stable conformers in solution, detectable via 1H-NMR spectroscopy at room temperature. Kinetic analysis demonstrated the reversible ring-opening and closing of MIDD0301 at gastric and intestinal pH levels, with varying rate constants. However, in vivo studies indicated that these interconversion kinetics are sufficiently rapid to ensure comparable MIDD0301 concentrations in blood and lung, irrespective of neutral or acidic formulations. Crucially, both acidic and neutral formulations of MIDD0301 exhibited significant lung distribution while maintaining low brain concentrations. These results underscore that MIDD0301’s structural flexibility across physiological pH ranges does not compromise its bioavailability, reinforcing its potential as an effective oral medication for asthma. This research highlights the importance of understanding pH-dependent drug behavior for optimizing benzo-based therapeutics, particularly those developed in research centers like those in Milwaukee.
Keywords: GABAA receptor, benzodiazepine, gastric pH, formulation, asthma, Benzo Milwaukee
Graphical Abstract
[Insert Graphical Abstract Image Here – Refer to original article for image URL if needed]
Introduction
MIDD0301 is an investigational oral drug being developed to manage asthma symptoms by relaxing airway smooth muscle (ASM) and reducing lung inflammation. Its mechanism involves targeting peripheral gamma-aminobutyric acid A receptors (GABAARs). This mechanism is supported by prior observations that propofol, a GABAAR modulator, induces ASM relaxation [1]. Subsequent research has confirmed the presence of GABAARs on ASM cells [2], airway epithelia [3, 4], and airway leukocytes, including CD4+ T cells [5] and alveolar macrophages [6], further justifying the development of GABAAR ligands for asthma treatment. GABAARs are heteropentameric chloride ion channels composed of diverse subunits (α1–6, β1–3, γ1–3, δ, ε, π, θ, and ρ1–3), typically assembled from two α, two β, and one tertiary subunit [7]. MIDD0301 demonstrates binding affinity to these receptors with an EC50 of 72 nM [8] and potentiates GABA-induced chloride flux across various GABAAR subtypes at 100 nM [9]. In vivo studies in asthmatic mice (ovalbumin-sensitized and challenged model) showed that oral administration of MIDD0301 at 100 mg/kg reduced eosinophil, CD4+ T cell, and macrophage counts in the lungs, along with a decrease in airway hyperresponsiveness to methacholine challenge [9]. Further evidence of its anti-inflammatory action was observed in the reduced levels of IL-17A, IL-4, and TNFα in the lungs of these asthmatic mice.
MIDD0301 stands out as the first allosteric GABAAR modulator with limited brain distribution. Sensorimotor assessments at high oral doses (1000 mg/kg) showed no central nervous system (CNS) effects. Recent 28-day immunotoxicity studies at 100 mg/kg also revealed no toxicity and no alterations in humoral immune response to the T-dependent antigen, dinitrophenyl-keyhole limpet hemocyanin [8]. MIDD0301 exhibits a long half-life, exceeding 25 hours in human and 9 hours in mouse microsomal preparations. Pharmacokinetic analysis in mice revealed a lung half-life of nearly 4 hours. A nebulized dose of 0.8 mg/kg MIDD0301 effectively reversed airway resistance induced by 25 mg/ml methacholine [10]. However, oral dosages of 50 mg/kg were needed to achieve pharmacological effects, prompting a deeper investigation into the compound’s gastric stability and bioavailability. This type of detailed pharmacokinetic study is crucial for drug development, especially for compounds like this novel benzo milwaukee derivative.
MIDD0301 incorporates a carboxylic acid moiety linked to a basic imidazobenzodiazepine scaffold. This structural feature leads to pH-dependent solubility and a reversible conversion from a seven-membered ring structure to an open-ring benzophenone form under acidic conditions. This equilibrium has been previously studied for 1,4-benzodiazepin-2-ones using NMR (diazepam [11], fludiazepam [11], and flurazepam [12]) and UV absorption (diazepam [13], prazepam [14], temazepam [15], and nitrazepam [16]). Additionally, triazolobenzodiazepine estazolam [17] and imidazobenzodiazepine midazolam [18] have been evaluated under acidic conditions using NMR and UV absorption, respectively. Sternbach initially reported the formation of 1,4-benzodiazepin-2-ones under basic conditions and suggested that the less pronounced pharmacological effects of acyclic aminoacetoamido compounds might be due to in vivo cyclization [19]. Later research from Hoffmann-LaRoche confirmed the ring-opening reaction for imidazobenzodiazepines in acidic environments and the isolation of a dihydrochloride salt [20].
This paper presents a comprehensive evaluation of MIDD0301’s pH-dependent solubility, lipophilicity (logD), and structural changes across different pH levels. We determined the kinetics of the ring-opening and closing reactions at gastric and intestinal pH using NMR and spectrophotometry. Finally, we investigated the impact of different formulations on the in vivo absorption of MIDD0301, examining whether gastrointestinal pH variations affect its bioavailability. This research is particularly relevant to the development of benzodiazepine-based therapeutics originating from or being studied in research hubs like Milwaukee.
Experimental Section
pH Dependent Absorbance Measurements
For pH-dependent absorbance measurements, 2 μL of a 5 mM solution of either closed or open MIDD0301 (synthesis details in Supporting Information) were added to 198 μL of buffered water at varying pH levels in a 96-well polypropylene plate (Nunc, 249944). Buffers used, adjusted to 0.1M ionic strength with KCl, were: pH 9 (12.5 mM borate), pH 8, 7, and 6 (50 mM phosphate), pH 5, 4, and 3 (50 mM acetate), and pH 2 and 1 (50 mM phosphoric acid). Solutions were mixed and aliquots (2 x 40 μL) were transferred to a 384-well plate (Coring UV star, 781801). After a 24-hour incubation at room temperature, absorbance was measured from 230 to 400 nm in 2 nm increments using a Tecan M1000 plate reader. Background subtraction was performed using 198 μL of buffered water with 2 μL DMSO. Absorbance averages from duplicate measurements were plotted against wavelength.
Determination of Solubility at Different pH
To assess solubility, 2 mg of closed MIDD0301 was added to 200 μL of buffered water at different pH and agitated in a closed vial on a horizontal shaker for 24 hours. Mixtures were then centrifuged at 16000 × g for 5 min, and the supernatant pH was measured. Three 10 μL aliquots of supernatant were diluted in 190 μL of 50:50 methanol/buffer water (25 mM phosphate buffer, pH 10), mixed, and 40 μL were transferred to a 384-well plate (Coring UV star, 781801) for UV detection at 270 nm (Tecan M1000). Each pH level was tested in triplicate. Concentrations were determined using a calibration curve in 50:50 methanol/buffer water (25 mM phosphate, pH 10).
Mass Spectrometry Analysis of pH Dependent Ratio of Closed and Open MIDD0301
For MS analysis, 10 mL of 50 μM aqueous closed MIDD0301 solution was adjusted to pH 1 using 0.1 M HCl. Before adjusting to pH 2 with 10% ammonium hydroxide, a 400 μL aliquot was taken (pH 1 sample). This process was repeated up to pH 9. Aliquots were kept at room temperature for 24 hours, followed by final pH measurement. Solutions were directly injected into a Shimadzu 2020 mass detector via syringe, needle adapter, and connector, without additional solvent. Peak heights at 414 m/z and 432 m/z were measured to determine ratios.
Acid-Dependent Conversion of Closed MIDD0301 Determined by 1H-NMR
For NMR studies, 1 mg of closed MIDD0301 was dissolved in 500 μL d6-DMSO, followed by 500 μL D2O. A 1H-NMR spectrum was immediately recorded. Subsequently, DCl (5.125 M in D2O) was added in small increments (0.1, 1, 2, 3, 4, 5, 8, 11, 21 μL). The NMR tube was vigorously shaken after each addition, and a 1H-NMR spectrum was recorded.
Stability Measurement of Closed and Open MIDD0301 at Different pH
Stability was assessed by adding 5 μL of 1 mM closed or open MIDD0301 solutions to 75 μL of buffered water at pH 2 or 8 in a 384-well plate (Coring UV Star). Solutions were mixed, and absorbance at 246 nm was recorded every 30 seconds for 260 minutes using a Tecan M1000. Background subtraction used buffered water with 5 μL DMSO. Absorbance vs. time plots were analyzed using first-order kinetics to determine rate constants (k) and half-lives. NMR-based stability measurements are detailed in Supporting Information.
Determination of logD
LogD was determined by adding 200 μL of 2.5 mM closed MIDD0301 in 1-octanol to 300 μL of water-saturated 1-octanol and 500 μL of 1-octanol-saturated buffered water at various pH levels in screw-top vials. Vials were vigorously shaken overnight and allowed to separate for 1 hour. Octanol layer aliquots (3 x 10 μL) were diluted with 70 μL octanol, and water layer aliquots (3 x 70 μL) were diluted with 10 μL of corresponding buffer. Absorbance was measured at 270 nm using a 384-well plate (Coring UV star) and a Tecan M1000. Calibration curves in octanol and water were used to determine MIDD0301 concentrations. Assays were performed in triplicate.
Molecular Dynamics Using MOE (Chemical Computing Group)
Molecular dynamics simulations were conducted using MOE software. Closed and open MIDD0301 structures were built and protonated using the protonate 3D function. The imidazole nitrogen protonation state was manually adjusted to +1 for both structures. Conformational searches were performed using LowModeMD with varying energy windows.
Pharmacokinetic Studies in Mice
Six-week-old female Swiss Webster mice (Charles River Laboratory, WIL, MA) were housed under specific pathogen-free conditions with controlled humidity, temperature, and a 12-hour light/dark cycle, with free access to food and water. All animal experiments were approved by the University of Wisconsin – Milwaukee Institutional Animal Care and Use Committee (IACUC). Mice received intragastric gavage of closed or open MIDD0301 at 25 mg/kg in different formulations. After 60 minutes, blood was collected via cardiac puncture into heparinized tubes. Lungs and brains were harvested and stored in liquid nitrogen (n=4).
Blood samples were thawed, vortexed, and 100 μL aliquots were mixed with 400 μL cold methanol containing 300 nM XHE-III-74A internal standard (I.S.). Samples were vortexed, centrifuged at 16000 × g for 10 min at 4 °C, and supernatants were transferred, evaporated, reconstituted with 400 μL methanol, and spin-filtered. 4,5-diphenylimidazole was added as an instrument standard. Injection volume was 5 μL (LC-MS/MS, Shimadzu 8040). Brain and lung tissues were thawed, weighed, homogenized in methanol with I.S., centrifuged, spin-filtered, evaporated, and reconstituted in methanol. 4,5-diphenylimidazole was added post-filtration. Injection volume was 5 μL (LC-MS/MS, Shimadzu 8040).
HPLC was performed using Shimadzu Nexera X2 LC30AD series pumps with an Agilent RRHD Extend-C18 column (2.1 mm × 50 mm, 1.8 μm particle size) and gradient elution (methanol and water with 0.1% formic acid) at 0.5 mL/min. Analytes were monitored in positive mode using a Shimadzu 8040 triple quadrupole mass analyzer with electrospray and atmospheric pressure ionization (DUIS). MRM transitions were optimized for closed MIDD0301, open MIDD0301, XHE-III-74A, and 4,5-diphenylimidazole. Data acquisition was performed using LabSolutions software. Standard curves were fitted by linear regression, and validation samples were calculated back. Accuracy was assessed by comparing calculated concentrations to nominal values.
Results and Discussion
MIDD0301 is synthesized by reacting the corresponding ethyl ester under basic conditions followed by acidification with acetic acid [9]. A white precipitate forms at pH 5 and is removed by filtration. To systematically assess MIDD0301’s aqueous solubility across a range of pH, we used a “shake flask” method and quantified concentrations via UV absorption. The results are shown in Figure 1.
Figure 1.
A) Absorbance spectra of MIDD0301 (50 μM in buffered aqueous solution with 1% DMSO); B) Solubility of MIDD0301 in water at different pH after 24 h shake flask method, determined by UV absorption at 270 nm.
UV measurements of 50 μM MIDD0301 in buffered water (pH 1-9) after 24-hour equilibration revealed isosbestic points at 270 nm, indicating similar total absorbance for both MIDD0301 structures at this wavelength (Figure 1A). Similar pH-dependent absorbance patterns with isosbestic points have been reported for midazolam (255 nm) [18]. Subsequent solubility tests using the shake flask method and UV absorbance at 270 nm (Figure 1B) showed excellent solubility of MIDD0301 at neutral pH (>3 g/L), but significantly lower solubility (<100 mg/L) at pH 3.9. This contrasts with midazolam, which is less soluble at neutral pH (70 mg/L) but more soluble at lower pH due to imidazole nitrogen protonation [18]. MIDD0301, in addition to a basic imidazole nitrogen, possesses a carboxylic acid group, leading to the formation of a neutral zwitterionic species around pH 5 (Figure 3). Analogous to amino acids, MIDD0301 exhibits its lowest solubility at its isoelectric point [21].
Figure 3.
[Insert Figure 3 Image Here – Refer to original article for image URL if needed]
MIDD0301 structures and pH-dependent forms.
To characterize MIDD0301’s acid-base behavior, we employed two methods: titration of a basic MIDD0301 solution with 1 N HCl and a spectrophotometric method using data from Figure 1A [22]. Results are presented in Figure 2.
Figure 2.
A) Acid titration of 50 μM MIDD0301 in 0.1 N NaOH/water with 30% DMSO. pH measured after 1 N HCl additions; inset: pH transition determination by non-linear regression. B) Spectral difference plot for MIDD0301 solutions at varying pH; inset: absorbance difference vs. pH plot to determine spectrophotometric transition pH. Absorbance difference is the sum of differences at 252 and 292 nm. pH of transition determined by nonlinear regression.
Acid titration of basic MIDD0301 solution showed a sharp pH decrease after NaOH neutralization (Figure 2A). A minor curve change between pH 7 and 6 was less pronounced than the change between pH 5 and 4. Nonlinear regression estimated the inflection point at pH 4.69. Spectrophotometric analysis of pH-dependent absorbance (Figure 1A), plotting absorbance changes at 252 and 292 nm against pH, yielded a transition pH of 4.34 (Figure 2B) [23], consistent with the titration result. Midazolam, an imidazobenzodiazepine, has reported pKa values for the imidazole nitrogen of 6.15 [20] and 6.04 [18]; however, the carboxylic acid at the 3-position in MIDD0301 reduces the imidazole nitrogen’s basicity.
We used mass spectrometry to investigate whether MIDD0301 undergoes closed-to-open form conversion even under mildly acidic conditions, potentially explaining the observed absorbance and pKa changes, consistent with imine-to-primary amine conversion (Figure 3). Imine protonation is expected to activate the electrophilic carbon, enabling nucleophilic attack by water. The resulting carbinolamine is then expected to form a ketone via proton transfer and carbon-nitrogen bond cleavage.
Unlike buffered solutions used in spectrophotometry, MIDD0301 solutions for MS analysis were pH-adjusted with 0.1 M HCl to minimize ion suppression. After 24 hours, solutions were directly injected into the MS to avoid LC-induced equilibrium shifts. MS spectra for representative pH values are in Figure 4A, and peak height ratios (414 m/z, closed MIDD0301; 432 m/z, open MIDD0301) vs. pH are plotted in Figure 4B.
Figure 4.
A) Representative MS spectra of MIDD0301 at different pH; B) Peak height ratios of 414 m/z (closed MIDD0301) and 432 m/z (open MIDD0301). At pH 4.3, both forms are present at equal peak heights.
For positive ion detection, closed MIDD0301 shows primary mass peaks at 414 and 416 m/z (79Br/81Br isotopes). Open MIDD0301 exhibits equivalent peaks at 432 and 434 m/z. At pH 7 and above, 414/416 m/z is dominant (98% closed form), consistent with in vivo quantification [9]. At pH 6, open MIDD0301 increased to 17% after 24 hours, indicating slow hydrolysis of the imine to amine at this pH. Equilibrium between open and closed MIDD0301 was observed between pH 4-5. The equilibrium pH of 4.3 closely matched the transition pH from titration (4.69) and UV spectrophotometry (4.34) (Figure 2). At pH 3 and lower, open MIDD0301 predominated, but even at pH 1 with 0.1 N HCl, 10% closed MIDD0301 remained after 24 hours. At pH 1, a dimeric species of open MIDD0301, responsible for the 415/417 m/z MS signal, was detected.
Complete closed-to-open MIDD0301 conversion was achieved in 4 N aqueous HCl at 60 °C for 5 hours. Open MIDD0301 was isolated and analyzed by NMR in d6-DMSO, confirming two rotamers, consistent with previous reports for diazepam and fludiazepam [11]. Pronounced 1H-NMR chemical shift differences were observed for the R-methyl group (1.45 ppm major, 1.49 ppm minor) and methine protons (4.11 ppm major, 4.23 ppm minor) in an 81:19 ratio. Distinct 13C-NMR signals were also observed for major and minor rotamer carbons, including R-methyl carbon (14.94 ppm major, 17.75 ppm minor) and methine carbon (41.45 ppm minor, 49.35 ppm major). (NMR signal assignments in Supporting Information).
Open MIDD0301 in D2O/d6-DMSO partially converted to closed MIDD0301 (50:50 ratio by 1H-NMR), consistent with the MS ratio at pH 4 (Figure 4). In 0.1 M DCl/D2O/d6-DMSO, 10% closed MIDD0301 was observed by 1H-NMR, aligning with MS data at pH 1-2 (Figure 4). Diazepam under similar conditions showed a 20:80 open-to-closed ratio, while fludiazepam (2’-F substituent, like MIDD0301) showed a 75:25 ratio [11]. The 2’-F substituent significantly favors ring-opening and open form equilibrium for benzodiazepines at gastric pH.
1H-NMR of open MIDD0301 in 0.1 M DCl/D2O/d6-DMSO showed two baseline-separated doublets for the R-methyl group (1.38 ppm major, 1.47 ppm minor) and a multiplet at 4.20 ppm for the methine proton for both rotamers (60:40 ratio), similar to fludiazepam’s cis/trans rotamer ratio (55:45) [11]. Molecular dynamics revealed two stable conformers with different imidazole-phenyl orientations (Figure 5 A and B), structure A (P stereochemistry) and structure B (M configuration).
Figure 5.
[Insert Figure 5 Image Here – Refer to original article for image URL if needed]
Stable conformers of open MIDD0301 (A, B) and closed MIDD0301 (C, D) determined by molecular dynamics and detected by NMR.
The energy difference between conformers A and B was only 1.43 kcal/mol. Conformational search showed interconversion only above a 12 kcal/mol energy window, confirming NMR findings of similar rotamer ratios and resolvable 1H-NMR and 13C-NMR peaks.
In contrast, 1H-NMR of closed MIDD0301 at pH 8 in D2O/d6-DMSO identified two rotamers (80:20 ratio). R-methyl protons appeared at 1.09 ppm (major) and 1.82 ppm (minor), and the methine proton at 4.20 ppm (minor) and 6.47 ppm (major). 13C-NMR signals for the R-methyl (15.8 ppm) and methine carbon (50.8 ppm) were identical for both rotamers, possibly due to longer carbon nuclei relaxation times. 1H-13C HSQC NMR showed coupling for both rotamers but absent carbon signals (Supporting Information). Molecular dynamics for closed MIDD0301 showed a larger energy difference between rotamers (ΔE = 5.32 kcal/mol, 80:20 ratio) but interconversion at a 6 kcal/mol energy window, explaining unified 13C-NMR signals (1.5 s relaxation time). The more stable rotamer (Figure 5C) shows a short methine proton-carboxylate distance, suggesting hydrogen bonding, potentially causing the 2.47 ppm chemical shift difference between rotamers.
Closed-to-open MIDD0301 conversion kinetics were studied under acidic conditions using 1H-NMR. Closed MIDD0301 in D2O:d6-DMSO (pH 8) was titrated with DCl/D2O. Representative 1H-NMR spectra are in Figure 6.
Figure 6.
1H-NMR spectra showing conversion of closed MIDD0301 sodium salt to open MIDD0301 by DCl (5.125 M in D2O) addition. Bottom spectra: closed MIDD0301 sodium salt. Subsequent spectra: same solution after 1, 2, 3, 4, and 21 μL DCl additions. Top spectra: open MIDD0301.
As expected, adding DCl (5.125 M in D2O) progressively converted closed MIDD0301 sodium salt to open MIDD0301. Signals at 1.09 ppm and 6.47 ppm disappeared, and signals at 1.38/1.47 ppm and 4.20 ppm emerged. Imidazole hydrogen deshielded due to imidazole nitrogen protonation (8.22 ppm to 8.41 ppm). Hydrogen ortho to the bromine substituent deshielded upon ring-opening (7.26 ppm to 7.89 ppm). Partially deuterated primary amine protons appeared at 8.67 ppm.
Kinetics of closed-to-open MIDD0301 conversion in 0.1 N HCl (gastric pH) and open-to-closed conversion at pH 8 (intestinal pH) were investigated using 1H-NMR. Closed MIDD0301 in D2O:d6-DMSO was treated with DCl (5.125 M) to achieve 0.1 N HCl. 1H-NMR spectra were recorded every minute. Figure 7A shows the integrated area (1.24–1.06 ppm) vs. time. Similarly, open MIDD0301 in D2O:d6-DMSO was adjusted to pH 8 with NaOD (7.23 M in D2O), and integrated areas (1.24–1.06 ppm) are plotted vs. time in Figure 7B.
Figure 7.
Kinetics of open/closed MIDD0301 interconversion at different pH. A) Time-dependent conversion of 2.4 mM closed MIDD0301 in 50:50 D2O:d6-DMSO with 0.1 N DCl (1H-NMR). Methyl proton peak areas vs. time. B) Time-dependent conversion of 2.4 mM open MIDD0301 in 50:50 D2O:d6-DMSO at pH 8 (1H-NMR). Methyl proton peak areas vs. time. C) Time-dependent conversion of 62.5 μM closed MIDD0301 in 0.1 N HCl (UV absorbance at 246 nm). D) Time-dependent conversion of 62.5 μM open MIDD0301 in 50 mM phosphate buffer pH 8 (UV absorbance at 246 nm). (n=4)
Assuming pseudo-first-order kinetics, closed-to-open conversion in 0.1 N DCl had a kacid of 0.050 min−1, while open-to-closed conversion at pH 8 had kbase of 0.007 min−1. Ring-opening was approximately seven times faster than ring-closing. Half-lives were 13.7 min and 96.2 min, respectively. Spectrophotometry at lower concentrations (62.5 μM) in aqueous solutions (no DMSO) and faster acquisition was used for comparison. At 246 nm, first-order kinetics gave kacid 0.142 min−1 and kbase 0.024 min−1 (Figure 7C, D). Ring-opening was again seven times faster. Both reactions were accelerated at lower concentration without DMSO, suggesting a charged transition state. The half-life for closed MIDD0301 in 0.1 N HCl was 4.8 min, indicating significant open MIDD0301 presence in the stomach. The half-life of open MIDD0301 at pH 8 was 28.8 min. Identical UV and NMR spectra were obtained for closed and open MIDD0301 at equilibrium at the same pH (Supporting Information).
Partition coefficient logDoct/wat of MIDD0301 at different pH was determined to understand oral absorption. Closed MIDD0301 in 1-octanol was partitioned with buffered aqueous solutions (pH 1-9) for 24 hours. Concentrations in each layer were determined by spectrophotometry. Results are in Figure 8.
Figure 8.
Lipophilicity (logDoct/wat) of MIDD0301 at different pH. 2.5 mM closed MIDD0301 in 1-octanol partitioned with buffered aqueous solutions (pH 1-9) for 24 h. UV absorbance at 270 nm used for concentration determination in both layers. (n=3)
Lowest lipophilicity was observed at pH 1 (open MIDD0301, multiple charges) and pH 8+ (closed MIDD0301 carboxylate, Figure 3), consistent with high aqueous solubility (Figure 1B). Highest 1-octanol concentration was between pH 2-5, reflecting neutral charge and potential hydrogen bonding, and correlating with lowest aqueous solubility (Figure 1B).
Bioavailability of MIDD0301 formulated at different pH was assessed in mice. Closed or open MIDD0301 (25 mg/kg oral doses) were administered to mice (n=4). Blood, lung, and brain concentrations of both forms were quantified by LC-MS/MS after 60 minutes. Results are in Table 1.
Table 1.
Concentrations of closed and open MIDD0301 60 minutes post-oral administration with different formulations.
| Nr | Compound | Vehicle | Concentration of closed MIDD0301 (nM)[d] | Concentration of open MIDD0301 (nM)[d], % of closed MIDD0301 |
|—|—|—|—|—|—|—|—|—|
| | | | Blood | Lung | Brain | Blood | Lung | Brain |
| 1 | Closed MIDD0301 | 2.5 % PEG400[a]98 % HPMC[b](2 % in water) | 322±65 | 287±46 | 87±27 | 9±33 % | 5±22 % | 3±12 % |
| 2 | Closed MIDD0301 | 40 % propylene glycol60 % HPMC[b](2 % in water) | 239±37 | 455±47 | 104±42 | 6±23 % | 9±22 % | 3±13 % |
| 3 | Closed MIDD0301 Na+ salt | 10 % PEG40090 % HPMC(2 % in water) | 154±58 | 306±106 | 118±58 | 5±23 % | 5±12 % | 2±12 % |
| 4 | Closed MIDD0301 | 2.5 % PEG400[a]98 % HPMC[b](2 % in PBS[c]) | 413±116 | 1745±1121 | 130±24 | 18±144 % | 28±192 % | 4±33 % |
| 5 | Pretreatment with Omeprazole | 100 % HPMC[b](2 % in carbonate buffered water, pH 9) | 516±180 | 2046±1155 | 159±33 | 20±84 % | 32±212 % | 4±23% |
| 6 | Open MIDD0301 | 2.5 % PEG400[a]98 % HPMC[b](2 % in 0.1 N HCl) | 371±104 | 1792±296 | 130±42 | 14±74 % | 35±82 % | 4±23 % |
[a] polyethylene glycol 400, [b] hydroxypropyl methylcellulose, [c] phosphate buffered saline, pH 7.4, [d] data (n = 4) as average±StD.
First, closed MIDD0301 in PEG400/HPMC-water suspension (pH 6.1) (Table 1, Entry 1) resulted in 322 nM blood and 287 nM lung concentrations after 1 hour. Brain concentration was 30% of lung, indicating good tissue selectivity. Open MIDD0301 levels were low (2-3% of closed form). Propylene glycol/HPMC-water solution (pH 5.9, Table 1, Entry 2) yielded lower blood (239 nM) but higher lung (455 nM) closed MIDD0301 concentrations, with similar open MIDD0301 levels. Closed MIDD0301 sodium salt (pH 8.2, Table 1, Entry 3) in PEG400/HPMC-water gave the lowest blood concentration (154 nM) but comparable lung (306 nM) and brain (118 nM) concentrations. Open MIDD0301 levels remained low. A neutral formulation (HPMC-PBS/PEG400, pH 7.4, Table 1, Entry 4) significantly increased lung concentration (1745 nM) compared to non-buffered formulation, with blood at 413 nM. Brain concentration was 7.8% of lung, showing excellent selectivity. Omeprazole pretreatment (pH 6.7 gastric pH, Table 1, Entry 5) slightly increased closed MIDD0301 concentrations in all tissues using the neutral formulation. Acidic formulation of open MIDD0301 (HPMC-0.1 N HCl/PEG400, pH 1.8, Table 1, Entry 6) showed comparable blood (371 nM), lung (1792 nM), and brain (130 nM) concentrations to the neutral closed MIDD0301 formulation (Table 1, Entry 3). Open MIDD0301 levels were again low. Consistent with kinetic studies, >90% open MIDD0301 converts to closed form within 1 hour at intestinal pH 8 (Figure 7D). Thus, MIDD0301 interconversion kinetics are rapid enough for similar tissue concentrations regardless of initial formulation pH.
Conclusions
The carboxylic acid functionality of MIDD0301 significantly enhances its aqueous solubility at neutral pH compared to classic benzodiazepines. This feature also reduces lipophilicity at neutral pH, limiting brain penetration to <10% of lung concentration, crucial for peripheral GABAAR targeting in asthma without CNS side effects [9]. MIDD0301 can be orally administered in either open (acidic formulation, ring-closing at intestinal pH) or closed form (neutral formulation) with comparable bioavailability. In all biological samples, open MIDD0301 was ≤4% of the total compound, suggesting the closed form primarily mediates pharmacodynamic effects. Further research into benzo milwaukee derivatives like MIDD0301 could pave the way for more targeted asthma therapies.
Supplementary Material
Supporting information
NIHMS1585656-supplement-Supporting_information.pdf (1.9MB, pdf)
Acknowledgements
We thank Dr. Beryl R. Forman and Jennifer L. Nemke (Animal Resource Center at UWM) for their guidance and support. This work was supported by the National Institutes of Health (USA) R41HL147658 (L.A.A.), R03DA031090 (L.A.A.), R01NS076517 (J.M.C., L.A.A.), R01HL118561 (J.M.C., L.A.A., D.C.S.), R01MH096463 (J.M.C., L.A.A.) as well as the University of Wisconsin-Milwaukee, University of Wisconsin-Milwaukee Research Foundation (Catalyst grant), the Lynde and Harry Bradley Foundation, and the Richard and Ethel Herzfeld Foundation. In addition, this work was supported by grant CHE-1625735 from the National Science Foundation, Division of Chemistry.
Abbreviations
ASM airway smooth muscle
GABAAR gamma amino butyric acid type A receptor
IL-17A interleukin 17A
IL-4 interleukin 4
TNFα tumor necrosis factor alpha
CNS central nervous system
NMR nuclear magnetic resonance
LC-MS liquid chromatography-mass spectrometry
MS mass spectrometry
DMSO dimethyl sulfoxide
TLC thin layer chromatography
HRMS high resolution mass spectrometry
ESI electron spray ionization
IT-TOF ion trap-time of flight
HSQC heteronuclear single quantum coherence
Footnotes
Supporting Information
Supporting Information is free of charge on the ACS Publication web site at DOI: – Synthesis of open and closed MIDD0301
- NMR analysis of MIDD0301 at different pH
- Table of NMR signals
- Time dependent NMR spectra of closed and open MIDD0301 at pH 2 and 8, respectively
References
[List of references from original article – keep the numbering and links if possible]
Associated Data
Supplementary Materials
Supporting information
NIHMS1585656-supplement-Supporting_information.pdf (1.9MB, pdf)
