Pre-formulation development of highly potent nontuberculous mycobacterial agents

Abstract

Nontuberculous mycobacteria (NTM) are pathogens that particularly infect patients with structural lung diseases, such as chronic obstructive pulmonary disease or cystic fibrosis. Several novel indole-2-carboxamides (ICs) that were synthesized in our laboratory have demonstrated potent antimicrobial activity against a wide panel of NTM pathogens with in vivo efficacy against Mycobacterium tuberculosis and Mycobacterium abscessus-infected mouse models. Lead ICs suffer poor aqueous solubility and permeability leading to high dosing regimens to achieve efficacy. The objective of this work is to conduct preformulation studies on two of the lead ICs (N2 and N21, US Patent 20180036283A1). Scanning electron microscopy (SEM), thermal analyses [differential scanning calorimetry (DSC), thermogravimetric analysis (TGA)], hot-stage microscopy (HSM)], and X-ray powder diffraction (XRD) were used for their solid-state characterization. Octanol/water partition coefficient (LogP), pKa, solubility, pH stability, intrinsic dissolution and Caco-2 cell monolayer permeability were determined. Thermal analyses and XRD indicated that both ICs are crystalline solids. N21 exists at least in two polymorphic forms and the polymorphic transition was monotropic in nature. However, N2 did not show any polymorphism. These molecules with poor aqueous solubility (<0.2 μg/mL), and high Caco-2 cell monolayer permeability (>18 × 10⁻⁶ cm/s), suggested that both ICs belong to Biopharmaceutics Classification System (BCS) class II and may pose absorption and bioavailability challenges in vivo, primarily due to the poor aqueous solubility. The BCS class II designation is defined by marketed doses, therefore, since N2 and N21 are currently under development, our class II designation is defined by the poor aqueous solubility and high permeability. These molecules are weakly acidic and most stable at pH 6.8 with pKa values between 6.7 and 7.9. Results from the preformulation studies set the stage for the formulation of these drug candidates for clinical investigation and new drug application in the future for the treatment of NTM infections.

1. Introduction

Nontuberculous mycobacteria (NTM) are ubiquitous environmental pathogenic microorganisms that can cause numerous infections, such as progressive pulmonary disease, lymphadenitis, skin/soft tissue infections, and disseminated disease [1,2]. The Mycobacterium avium complex and the Mycobacterium abscessus complex are the two most prevalent NTM infections. Several studies suggest that the number of NTM infections and mortality rates are rising [3,4,5]. According to the American Thoracic Society, there is no proven therapeutic regiment for Mycobacterium abscessus infections. Given that NTM infections prey upon patients with structural lung disorders, including cystic fibrosis and chronic obstructive pulmonary disease (COPD), and significantly reduce quality of life along with hospital admittance with high cost burden, the need for improved therapeutic options for NTM infections is needed.

Both N2 and N21 show potent activity against M. abscessus (ATCC 19977), M. massiliense (CIP108297), M. bolletii (ATCC 14472), and M. tuberculosis (H37Rv) with MIC values ≤0.5 μg/mL [6]. The cytotoxicity study against THP-1 cells, a normal human cell line revealed that these molecules are safe with their IC₅₀ (concentration required to kill 50 % cells) values at least above 50 μM. In the previous study, these lead molecules (N2 and N21 from this study were named as N5 and N25, respectively) achieved efficacy in an in vivo mouse model of M. abscessus infection, supporting the strong translational potential of this series [4,7]. Other ICs were also demonstrated to have efficacy against M. tuberculosis-infected mouse models, further supporting the translational potential of this series. However, the efficacy was achieved with a high oral dose of 300 mg/kg in mouse models suggesting that these molecules were poorly orally absorbed resulting in their poor bioavailability [8,7,9]. The current work was aimed to determine some of the fundamental physicochemical properties of N2 and N21, which is one of the preliminary steps during the preformulation studies. The findings of the present along with the previous preliminary safety and efficacy studies will provide the base for designing rational dosage forms for N2 and N21 to achieve dose reduction and targeted activity.

Preformulation studies play a significant role in drug discovery and development programs for determining their physicochemical and biopharmaceutical properties. ICs are poorly water-soluble [10,11], which is a common barrier in mycobacterial drug discovery and development. The primary objective of this investigation is to establish the physicochemical properties of two new ICs (N2 and N21) to ensure future formulations are stable, effective, and safe. This investigation discloses a preformulation study, including drug solubility, permeability, LogP, dissolution rate, polymorphic forms, and stability for N2 and N21 to determine critical physicochemical properties that ensure the future drug dosage forms to improve PK/PD of these novel compounds to treat mycobacterial infections [12,13,14].

The novelty of this work lies in the fact that developing target specific NTM therapeutics that are highly specific and sensitive to mycobacterial pathogens are unique and cutting edge. Prior to manufacturing new drug products, robust preformulation studies are required for the fundamental understanding of physicochemical properties of API to identify optimal dosing regimens ensuring effective pharmacokinetic/pharmacodynamic properties. There is a clinical need for new therapeutic interventions to treat NTM infections which require lengthy, costly, and generally ineffective treatment modality, and this study has the potential to address these unmet needs. The physicochemical properties of these new compounds were reported for the first time and are essential for future formulation development for clinical investigation and new drug application.

2. Materials and methods

2.1. Materials

N2 (mol wt of 294.4 g/mol) and N21 (mol wt of 298.4 g/mol) (Fig. S1) were synthesized in our laboratory (US Patent: 20180036283A1) using the previously published methods [15]. Unless otherwise specified, N21 referred in this manuscript denotes Form I. Form II was formed after a polymorphic (monotropic) transition. Caco-2 colorectal adenocarcinoma cell line, Dulbecco’s modified eagle medium (DMEM), Penicillin-Streptomycin solution, and Trypsin-EDTA Solution 1× were purchased from ATCC (Manassas, VA). Phosphate-buffer pH 7.4 solution was purchased from Gibco BRL (Gaithersburg, MD), Hank’s Balanced Salt Solution (HBSS) pH 7.4 was purchased from Fisher Scientific (Pittsburgh, PA). Fetal bovine serum (FBS) was purchased from R&D Systems (Minneapolis, MN). Metformin was purchased from EMD Millipore (Burlington MA). Haloperidol was purchased from TCI America (Portland, OR). Deionized water was obtained from Purelab Ultra Water Purification System (ELGA LabWater NA, Woodridge, IL). Solvents of analytical or HPLC grade were used throughout the study. Acetone, acetonitrile, dichloromethane, methanol, and ethanol were purchased from Fisher Scientific, and formic acid from Sigma-Aldrich, Inc. (Burlington MA).

2.2. Methods

2.2.1. Chromatography

Ultra-Performance Liquid Chromatography (UPLC) methods were developed for the quantitation of both N2 and N21 and validated [16]. A Waters Acquity UPLC system (Medford, MA) equipped with the quaternary solvent manager (Model E10QSM461M), sample manager – FTN (Model E10SDI500N), column manager (Model K10CMP305N), PDA detector (Model J08UPL270R) and Empower software. The mobile phase consisted of acetonitrile: 0.1 % (v/v) aqueous formic acid in a ratio of 70: 30 (v/v) at 0.3 mL/min flow rate. Chromatographic separation was achieved on BEH C18 (150 mm × 2.1 mm × 1.7 μm) column (Waters, Medford, MA) at room temperature.

2.2.1.1. Standard solutions and sample preparation.

The standard stock solutions were prepared by dissolving 10 mg of N2 or N21 separately in a 100 mL mobile phase in a volumetric flask. The calibration standard solutions for N2 and N21 were then prepared separately by diluting the standard stock solution (100 μg/mL) with mobile phase to yield nominal concentrations over a concentration range of 0.078–10 μg/mL. The correlation coefficient (r²), intercept, and slope of the regression line were determined. The limit of quantification (LOQ) value was calculated based on a signal-to-noise ratio of 10 [17]. All the standard solutions were stored in a refrigerator.

2.2.1.2. Precision and accuracy.

Within-day precisions of the methods were determined by analysis of replicate (n = 4) standard curves, constructed within a day and was determined from the relative standard deviation (RSD) of the peak areas of each concentration of the calibration range. The day-to-day precision (N = 6) was determined by analyzing calibration standards on six different days over a period of 42 days for N2 and 41 days for N21. For evaluating the accuracy of the proposed methods, different concentrations at three quality control (QC) levels: low (LQC), medium (MQC), and high (HQC) covering the entire calibration range of N2 and N21 were injected on different days. The accuracy of the methods is reported as the percent of measured concentration to the nominal concentration.

2.2.2. Physicochemical characterization

2.2.2.1. Scanning electron microscopy.

For scanning electron microscopy (SEM) analysis, the samples were mounted using conductive carbon tabs (Ted Pella, Redding, CA) and analyzed under a Thermo Scientific Phenom Pharos Desktop SEM (model PW-100–60) equipped with a field emission source at 15 kV using backscattered electron and secondary electron detectors [18].

2.2.2.2. Thermal analyses.

For DSC analysis accurately weighed samples were placed in an aluminum pan and the analysis was performed using a differential scanning calorimeter (DSC-25; TA Instruments, New Castle, DE) equipped with TRIOS software. Each sample was heated from 25 to 300 ^◦C at a heating rate of 10 ^◦C/min, unless otherwise specified, under nitrogen flow (50 mL/min). For N21, the DSC study was further performed by heating samples up to 145 ^◦C, followed by cooling to room temperature and again reheating it from 25 to 300 ^◦C.

For TGA, an accurately weighed samples of N2 and N21, and N21 heated to 145 ^◦C were heated using a thermogravimetric analyzer (TGA-55; TA Instruments, New Castle, DE) at a rate of 10 ^◦C/min from 25 to 300 ^◦C, with a nitrogen flow of 50 mL/min.

The hot stage microscopic analysis of N21 was performed on a hot-stage microscope (Analysis Microscope Leica DM2500 M, Leica Microsystems Inc. Deerfield, IL). About 0.1 mg of the sample was placed on a glass slide and heated at a controlled rate of 2 ^◦C from 25 to 180 ^◦C. Any physical changes during heating were visually observed and digital images were captured at those and some predetermined temperatures.

2.2.2.3. X-ray powder diffraction.

N2 and N21 were characterized by using an X-ray powder diffractometer (D8 ADVANCE, Bruker AXS, WI) equipped with a variable-temperature stage (TTK 450, Anton Paar, GrazStraβang, Austria) and a Si strip one-dimensional (1D) detector (LynxEye, Bruker AXS, WI). The measurements were performed at room temperature. For N21, the sample was further analyzed using variable-temperature X-ray diffraction (VTXRD). The heating rate was 10 ^◦C/min, and the sample was maintained under isothermal conditions during each scan. The measurements were performed in 10 ^◦C increments from room temperature to 170 ^◦C using Co Kα radiation (35 kV × 45 mA), the step size was 0.02^◦. The data were collected over the range of 2 to 35^◦2θ and 1 s dwell time.

2.2.2.4. Equilibrium solubility.

The thermodynamic equilibrium solubility of N2, and N21 (form I and II) was determined in deionized water at 37 ^◦C. An excess amount of the sample (N = 3) was added to 5 mL of water in borosilicate screw-capped glass bottles and shaken continuously in a MaxQ™ 4450 Benchtop Orbital Shaker (Fisher Scientific, Pittsburgh PA). After 24 h, the samples were collected and centrifuged at 4000 rpm for 5 min, and the supernatants were collected, and filtered through 0.22 μm syringe filter before injecting to UPLC for analysis. This filtration step was adopted to prevent entry of any particulate matter during UPLC analysis.

The equilibrium solubility of N2 and N21 was also determined at 25 ^◦C in organic solvents (acetone, acetonitrile, dichloromethane, ethanol, and methanol) using a modified isothermal saturation shake flask technique. The equilibrium solubility of N21 form II was determined in acetonitrile and methanol. The excess amount of the samples (N = 3) was added to respective solvents and shaken for 48 h at 25 ^◦C. Samples were collected and centrifuged, supernatants were diluted suitably in the mobile phase. The concentrations of N2 and N21 in saturated samples of both aqueous and organic solvents were determined using the developed UPLC methods.

2.2.2.5. Intrinsic dissolution rate (IDR).

Intrinsic dissolution rates of N2 and N21 were determined using a modified Wood’s intrinsic dissolution apparatus as described by Saboo et al., 2020 [19], with minor modification of the instrument. Briefly, 100 mg of the powders were compacted in an 8 mm circular die at a compression pressure of 2100 psi in a hydraulic press (Carver Inc., Menomonee Falls, WIS) held for 60 s for N2. The optimized held time for N21 was 90 s for compact formation. The die containing drug compacts were attached to a shaft and holder and submerged in 100 mL phosphate buffered saline (PBS; pH 7.4) at 37 ^◦C with the rotation speed of 100 rpm. At pre-determined time points, 0.2 mL of the dissolution media were withdrawn from each vessel (n = 3), and volume sampled was replaced with equal volume of fresh PBS. The samples were analyzed for N2 and N21 content using previously reported LC-MS/MS methods with warfarin as an internal standard [7]. The IDR (μg/min/cm²) was calculated using the equation [19]:

$$\mathrm{IDR}=\frac{\mathrm{k}\times \mathrm{V}}{\mathrm{S}\times \mathrm{N}}$$

Where k is the slope of the regression line (ng/mL.min), V is the volume of dissolution medium (100 mL), S is the tablet surface area (0.5 cm²), and N is the weight fraction of N2 or N21 in the compact.

2.2.2.6. Partition coefficient determination.

The LogP values of N2 and N21 was determined by the miniaturized shake flask method with some modifications [20]. This method was used in because of its simplicity allowing a rapid way of determining lipophilicity using a very low drug amount. N2 was added at a theoretical concentration of 5 mg/mL and N21 at 50 mg/mL in the pre-saturated N-octanol: DI water, 1:1 ratio. The resultant mixtures were shaken in the incubator at 25 ^◦C to reach equilibrium. After 4 h, both aqueous and octanol phases were carefully separately collected followed by centrifugation at 4000 rpm for 5 min. Supernatants were collected, suitably diluted with the mobile phase, and analyzed on UPLC. The LogP value was determined using the following equation:

$$\mathrm{Log\; P}=\mathrm{Log}\left[\frac{\left(\text{Concentration of drug in octanol}\right)}{\left(\text{Concentration of drug in water}\right)}\right]$$

The average partition coefficient with standard deviation were reported from triplicate samples.

2.2.2.7. pKa determination.

The pH-metric technique (based on potentiometric titration) was used for the determination acid-base dissociation constant (pKa) of N2 and N21 using OrionStar T910 pH titrator (Fisher Scientific, Pittsburgh, PA) [21]. Briefly, 25 mg of N2 and N21 were dissolved in 20 mL of methanol and 5 mL of 1 N HCl was added. The apparent pH of the resultant solution was adjusted to ~2.0. The samples were titrated with 1 N NaOH solution at room temperature using a pH electrode. The course of titrations was monitored up to about pH 12. A minimum of three independent measurements were conducted. The pKa values were calculated from the corresponding inflection point of titration curves [22]. The pKa of haloperidol was determined as a reference standard under the similar experimental condition.

2.2.2.8. pH-dependent solubility and stability.

The pH-dependent saturation solubility of N2 and N21 was determined in Britton-Robinson buffers pH ranging from 1.2 to 11.0 by saturation shake-flask method [23]. The buffer solution (a mixture of acetic, phosphoric, and boric acids, each at 0.04 M) was prepared, and 0.2 M NaOH was added to raise the pH required for each shake-flask experiment except for the pH 1.2 which was adjusted with 1 N HCl. Excess amounts of N2 and N21 were added in 1 mL of each buffer and water (N = 3) and the samples were kept for shaking at 25 ^◦C in a shaking incubator. After 48 h, the samples were collected, and undissolved particles were settled by centrifugation. The supernatants were filtered through 0.22 μm PVDF membrane filters to ensure any undissolved particles did not interfere. The concentration of N2 and N21 was determined by LC-MS/MS. Remaining filtrates of preselected pH from the pH-dependent solubility study were utilized for stability study. Samples were incubated at 25 ^◦C in sealed vials over a period of 4 weeks. At definite time intervals, 0.1 mL sample was collected and analyzed by LC-MS/MS.

2.2.2.9. Caco-2 cell culture.

Caco-2 cells were incubated in a humidified incubator at 37 ^◦C with 5 % CO₂. The cell culture media contained DMEM supplemented with 100 U/mL penicillin, 100 μg/mL streptomycin, and 20 % fetal bovine serum. The media on the Caco-2 cells was replaced three times a week and the cells were passaged at or before 80 % confluency with 0.25 % trypsin/0.53 mM thylenediaminetetraacetic acid (EDTA) solution. The cells were initially seeded at a density of 50,000 cells/cm² in the 24-well Transwells^® (Thermo Scientific).

2.2.2.9.1. Caco-2 cell permeability study.

This study was performed as described previously with some modifications [24]. The transepithelial electrical resistance (TEER) was measured in each monolayer using a Millicell^®-ERS apparatus (Millipore Corp., Bedford, MA). TEER values (Ω _× cm²) were calculated after subtracting the resistance of blank filters. The monolayers having TEER values greater than 250 Ω _× cm² were included in this study and were formed 26 to 28 days after plating.

For absorptive (AP → BL) permeability experiments, the transport media (HBSS pH 7.4 supplemented with 0.1 % BSA without phenol red) was aspirated from the AP (donor) chamber of each well and replaced with transport media with drug molecules. Metformin was used as a control. The plates were placed in a 37 ^◦C incubator with 5 % CO₂ and agitated on a plate shaker and the samples were collected at 15, 30, 45, 60, 90, and 120 min. For secretive transport (BL → AP) permeability experiments, the transport media was aspirated from the BL (donor) chamber of each well and replaced drug solutions in transport media. The plates were incubated for the absorptive experiments with the same time points. Samples were removed from the AP chamber at the designated time points and replaced with the same volume of pre-warmed media. All samples were stored at − 20 ^◦C until analysis by UPLC. The cumulative transported drug concentration was plotted versus sampling time to obtain a value for the initial slope of the resulting curve. Apparent permeability (Papp ×10⁻⁶ cm/s) was calculated from the following equation [24]:

$$\mathrm{Papp}=\left[\frac{{\mathrm{V}}_{\mathrm{R}}}{\mathrm{A}\times {\mathrm{C}}_0}\right]\times \left[\frac{\mathrm{dC}}{\mathrm{dt}}\right]$$

Where V_R is the volume in mL in the receiver chamber, A is the filter membrane surface area in cm², C₀ is the initial drug concentration in μM in the donor chamber, and dC/dt is the initial slope of the concentration versus time (sec) curve.

3. Results and discussion

3.1. UPLC method validation

A prerequisite for quantification of novel compounds from various in vitro biological assays is a robust, reliable, and reproducible quantification method. For these purposes, we have developed a validated UPLC method for quantification of N2 and N21 in various biological and physicochemical assays.

3.1.1. Specificity

A method to separate N2 and N21 where the retention time was 3.4 ± 0.05 min and 3.7 ± 0.05 min for N2 and N21, respectively was developed. The run time for both methods was 6.0 min. The chromatograms of blank mobile phase did not show any interference with that of N2 or N21 (Fig. S2). The λ_max was 295 nm and 301 nm for N2 and N21, respectively. The peak purity index of N2 and N21 were determined using a PDA detector to determine the specificity of the methods. It was found to be >0.998 for both N2 and N21 indicating the specificity of methods and noninterference of any other peak of degradation product or impurity [25].

3.1.2. Linearity and range

The concentration range where compound concentration and absorbance were linear were selected and plotted. The calibration curves were constructed by the least square regression analysis by plotting peak area versus concentration. It showed a linear relationship with a correlation coefficient of >0.999 for both N2 and N21 over the concentration range of 0.078–10 μg/mL.

3.1.3. Precision and accuracy

Precision and accuracy of an analytical method ensure the quantification of N2 and N21 are reproducible and true. The precision at each concentration is presented in Table S1. The within-day variability was ≤3.98 % and ≤ 2.04 % for N2 and N21, respectively. The day-to-day variance for N2 was ≤6.36 % and for N21 ≤ 4.31 %. The low RSD values indicated that both the methods are precise and reproducible. The accuracy was determined by injecting three QC standards at three QC levels each (N = 9) for N2 and N21 and it was found to be 99.72–101.74 % for N2 and 98.31–103.28 % for N21 (Table S2). These results indicated the accuracy of the methods as the measured concentrations were in close agreement with their actual values.

3.2. Physicochemical characterization

3.2.1. SEM imaging

The SEM is utilized for topographical analysis, as it can provide high-quality, nanometer-scale images, and size distribution of powders and particles [26]. From SEM analysis we observed that N2 is cylindrical in shape with particle size between ~2–200 μm (Fig. 1A). N21 was irregular in shape with particle sizes ranging from ~5–100 μm (Fig. 1B).

Fig. 1.

Scanning electron microscopic image of N2 and N21.

3.2.2. Thermal and XRD analysis

The DSC measures the amount of energy absorbed (endothermal event) or released (exothermal event) by a sample when a physical event takes place during controlled heating. Melting point, heat capacity, heat of fusion, and polymorphic transitions can be revealed from DSC analysis. The DSC thermogram (Fig. 2A) of N2 showed a sharp endothermic event at 233 ^◦C indicating its melting. When the solid exists in more than one polymorphic forms, a DSC thermogram may show different thermal events such as solid-solid transition, recrystallization, and possible decomposition [27]. The DSC thermogram of N21 showed an endothermic peak at 139 ^◦C, followed by an exothermic peak at 143 ^◦C, and a second endothermic peak at 168 ^◦C. The multiple events in the thermogram could be explained by the fact that N21 (Form I) melts at 139 ^◦C, followed by recrystallization at 143 ^◦C to a different polymorphic form (Form II). Form II finally melts at 168 ^◦C. However, when N21 was reheated to 300 ^◦C after initial heating to 145 ^◦C the recrystallization event was absent. It showed a single endothermic peak with a peak temperature of 168 ^◦C indicating that once Form II was formed the transformation of Form II to Form I during cooling did not occur. The FDA has an increasing need for characterization of drugs that exist in various polymorphic forms as they may have unique stability and PK/PD profiles. For example, drugs that have a polymorphic form that melts at a higher temperature tend to have lower aqueous solubility (potentially leading to lower absorption, thus poor PK) and higher stability (potentially leading to longer shelf-life).

Fig. 2.

(A) Differential scanning calorimetry thermograms of N2, N21, and reheated N21. (B) Thermogravimetric curves of N2 and N21.

TGA is an analytical technique used to determine a material’s thermal stability and its fraction of volatile components by measuring the weight change that occurs as a sample is subjected to controlled heating. Therefore, TGA is an effective analytical method to analyze the decomposition, dehydration, desolvation, or sublimation processes when the sample is subjected to heat [27]. The TGA thermograms as shown in Fig. 2B for both N2 and N21, revealed non-significant weight loss during all the thermal events, confirming that the endothermic events are neither desolvation nor dehydration, suggesting good thermal stability. However, weight loss at higher temperature in both TGA thermograms may represent sample decomposition after melting.

The XRD technique is very useful in establishing whether a compound exists in an amorphous or crystalline state. The XRD analysis confirmed the crystalline nature of both N2 (Fig. 3A) and N21 (Fig. 3B). To confirm the polymorphic transition events of N21 as described earlier, further analytical methods (HSM and PXRD) were carried out [28,29]. HSM is a thermal analysis technique that combines both properties of thermal analysis and microscopy and serves as a complimentary analytical method that is used to support DSC and TGA data. It further helps to detect small changes in the sample that may be missed by DSC and TGA during thermal analyses [29]. The HSM of N21 was performed from 25 to 190 ^◦C at a controlled heating rate to visually observe the effect of heating on N21. Both the melting events of 139 ^◦C and 168 ^◦C from DSC studies were visually observed by HSM confirming a possible polymorphic transition and complete melting around 174 ^◦C (Fig. 4). However, the exothermic peak at 143 ^◦C was not detected by HSM. To confirm this exothermic event, variable temperature XRD (VTXRD) was used (Fig. 3B). There was no change in the crystal structure of N21 upto 140 ^◦C. However, at 140 ^◦C, some new peaks started to emerge, and some of the original peaks of N21 disappeared indicating the formation of another crystal form (polymorph) of N21. VTXRD enabled the direct identification of the temperature-dependent crystallizing polymorphic transition of N21.

Fig. 3.

(A) XRD plot of N2 and (B) XRD plot N21 at different temperatures.

Fig. 4.

Hot-stage microscopic images of N21 after heating to different temperatures.

It was further investigated that whether the polymorphic transition observed for N21 was either monotropic or enantiotropic. As per the heat-of-fusion rule, “if the higher melting form has lower heat of fusion, the two forms are usually enantiotropic, otherwise, they are monotropic.” The results in Table 1 show that irrespective of the sample size, the polymorphic form with the higher melting point (~ 168 ^◦C) had a higher enthalpy of fusion than the polymorphic form with the lower melting point (~ 139 ^◦C). N21 had no transition point below 143 ^◦C. Above this temperature, the only observed event was the melting of Form II at 168 ^◦C. Also, in the case of monotropy, a higher melting form may not revert to a lower melting form on cooling, which was observed in the case of N21 Form II (Fig. 2A, N21 reheated) [30,31]. Therefore, the polymorphic transition observed in N21 is monotropic.

Table 1.

Effect of sample size on the enthalpies of the thermal events in N21.

	Weight (mg)	Enthalpy of fusion (J/g)a	Enthalpy of crystallization (J/g)b	Enthalpy of fusion (J/g)c
Sample 1	5.0	26.7	− 12.6	56.6
Sample 2	7.2	25.6	− 11.5	51.0
Sample 3	6.1	27.9	− 12.8	49.4
Sample 4	4.8	23.0	− 10.2	54.8
Average (± SD)	5.8 ± 1.1	25.8 ± 1.2	− 11.8 ± 0.7	52.9 ± 3.8

^anEndotherm at 139 ^◦C (refer Fig. 2A)

^bExotherm at 143 ^◦C

^cEndotherm at 168 ^◦C

3.2.3. Experimental solubility data of N2 and N21

The aqueous and organic equilibrium solubility characterization of an API is of fundamental importance in determining absorption, its interaction with putative therapeutic targets, and its ultimate metabolism and excretion [32]. The aqueous equilibrium solubility at 37 ^◦C (to mimic human body temperature) for N2 was 0.16 ± 0.02 μg/mL, and 0.17 ± 0.03 μg/mL for N21, suggesting that both molecules are practically insoluble in water [33,34]. The aqueous solubility values in μg/mL after 48 h of stirring at 25 ^◦C were 0.18 ± 0.03 μg/mL for N2, 0.20 ± 0.02 μg/mL for N21 Form I, and 0.31 ± 0.03 μg/mL for N21 Form II.

The solubility of N2 and N21 in various pharmaceutically important organic solvents was also determined at 25 ^◦C, as this solubility profile assists in future formulation development. The results of this study (Table 2) N2 is soluble in dichloromethane, sparingly soluble in methanol and ethanol, and slightly soluble in acetone and acetonitrile. N21 is soluble in dichloromethane, soluble in ethanol and sparingly soluble in acetonitrile, methanol, and acetone. The solubility of N21 Form II was higher than Form I.

Table 2.

Solubility of N2 and N21 in selected organic solvents (mean ± SD, n = 3).

Solvent	Observed solubility (mg/mL)
	N2	N21 form I	N21 form II
Acetonitrile	9.0 ± 1.3	26.3 ± 1.0	40.3 ± 0.8
Methanol	22.1 ± 1.0	14.9 ± 2.0	22.3 ± 2.4
Acetone	6.7 ± 0.4	14.6 ± 1.2	–
Ethanol	20.3 ± 2.3	76.4 ± 17.8	–
Dichloromethane	50.1 ± 4.8	108.4 ± 17.8	–

- Not determined

1 mg/mL of N2 = 3.396 mM

1 mg/mL of N21 = 3.351 mM

In the case of N21 Form II, its observed solubility ratio was more than Form I by 1.55 times in water, 1.53 times in acetonitrile and 1.50 times in methanol indicating that it is a metastable form. The free energy difference (ΔF_T) between the metastable and stable form of N21 was calculated according to the equation [35]:

$$\mathrm{\Delta }\mathrm{FT}=RTln\frac{\mathrm{solubility\; of\; the\; metastable\; form}}{\mathrm{solubility\; of\; the\; stable\; form}}$$

In this equation, R is the gas constant = 1.987 cal/mol⋅K; T is the temperature in which the solubilities were determined (298 K). The ΔF_T values based on the solubilities of N21 Form I and Form II in water, acetonitrile and methanol was 259.5, 251.8 and 239.8 cal/mol, respectively.

3.2.4. Intrinsic dissolution

The aim of this study was to determine the intrinsic dissolution (IDR) of N2 and N21 to profile how quickly these compounds will be released into the dissolution medium from a solid with constant surface area. Fig. 5 shows their dissolution profiles in PBS at pH 7.4 Both molecules dissolve slowly with <15 ng/mL dissolved within two hours, as determined by LC-MS/MS. The IDR determined from dissolution profiles was 16.93 ± 0.95 ng/min/cm² for N2 and 21.00 ± 0.99 ng/min/cm² for N21. The IDR method is closest in predicting the behavior of drugs in in vivo conditions. As per BCS, IDR below 100 μg/min/cm² can be assigned as low soluble [36]. The results from this study suggest that solubility will be a limiting factor for bioavailability of N2 and N21.

Fig. 5.

Dissolution profiles of N2 and N21 from the discs in PBS pH 7.4 at 37 ^◦C.

3.2.5. Partition coefficient

The LogP for N2 and N21 as drugs must possess a balance of lipophilic and hydrophilic solubilities for absorptions leading to optimized efficacy and bioavailability [37,38,32,39]. The miniaturized shake-flask approach was used in this study to determine the partition coefficient with a small sample sizes. The observed LogP value for N2 was 4.88 ± 0.02 meaning it is partitioned more than 10,000 times in the n-octanol than in water. The LogP for N21 was 5.76 ± 0.07 meaning it is partitioned more than 100,000 times in the n-octanol than water. This highly lipophilic properties of N2 and N21 indicates the likelihood of crossing lipophilic biological membranes is high in a molecularly dispersed state.

3.2.6. pKa determination

This well-established association between pKa and pharmacokinetic parameters has resulted in the requirement for pKa value determination for a new drug candidate by the FDA. Additionally, pKa determination has traditionally had profound effects on PK/PD profiles. In this pH-metric experiment, Haloperidol was used as a reference standard (pKa = 8.22 ± 0.09) [40]. The pKa value of N2 was determined to 7.86 ± 0.08 for N2 and 7.62 ± 0.08 for N21, suggesting both compounds are weakly acidic. The amide and indole nitrogen atoms are weakly acidic, because the lone pair of electrons on the nitrogen atoms are in resonance with the amide carbonyl and the pi-bond network in the indole ring, respectively. In addition, the lone pair of electrons on the indole nitrogen are necessary for the indole ring’s aromaticity. Therefore, in both functional groups, the nitrogen atom is not basic because the lone pair of electrons are delocalized through resonance/aromaticity. This electron delocalization/resonance capability of both functional groups also contributes to the weak acidity of both groups in that after the proton (the nucleus of the hydrogen atom) has been donated, the resulting anionic conjugate is stabilized by resonance as well. These pKa values suggest that both molecules are more likely to be at an equilibrium between unionized and ionized forms at pH 7.4, pH of blood and most tissues.

3.2.7. pH-dependent equilibrium solubility and stability

In vivo, there are many environments with different pH ranges, therefore, we evaluated the solubility of N2 and N21 in various biologically relevant pH environments (pH of 1.2 to 11.0). In general, the solubility showed a pH-dependent solubility profile of a weak acid with higher solubility in basic pH (Fig. 6A). The solubility was observed to be highest at pH 11.0. However, both molecules were still practically insoluble at all the tested pH values (solubility values were <550 ng/mL for N2 and < 300 ng/mL for N21).

Fig. 6.

(A) The pH solubility of N2 and N21 in Britton-Robinson buffers. (B) The pH rate profiles of N2 and N21 in solution at 25 ^◦C in Britton-Robinson buffers of different pH over the period of 28 days.

To determine storage capabilities of N2 and N21 in solution, the stability of N2 and N21 at 25 ^◦C in water and the selected Britton-Robinson buffers was evaluated over a period of 4 weeks. The stability of both compounds was determined at pH 1.2, 4.5, and 6.8 according to the pH conditions of gastrointestinal track [41]. The pH values up to 11 were included to determine their stability under extreme conditions and determining appropriate storage conditions to prevent degradation. The results of this study (Fig. 6B) suggested that both molecules are less stable in gastric pH (1.2) than at intestinal (4.5 and 6.8) pH. Overall, both molecules were stable between the pH 4.5–10.0 and their degradation rate followed first-order kinetics. The highest degradation was observed at pH 11.0. The first-order degradation rate constant for N2 and N21 was calculated from the slope of the lines. It was the lowest at pH 6.8 among the studied pH buffers with 0.00415 ± 0.00122 day⁻¹ for N2 and with 0.00468 ± 0.00083 day⁻¹ for N21 at 25 ^◦C, indicating they are the most stable at pH 6.8. These data suggest that N2 and N21 will remain stable after oral administration and will not degrade significantly in the stomach and intestine prior to absorption.

3.2.8. Caco-2 permeability of N2 and N21

To determine permeability of N2 and N21 across cellular membranes, we utilized the Caco-2 cell line to determine the transport of these compounds across the human intestinal epithelium. We observed a time-dependent increase in the transport of N2 and N21 (Fig. 7) from the apical→basal (AP → BL) side. The apparent permeability value of N2 was 17.76 ± 0.88 × 10⁻⁶ cm/s and N21 was 18.31 ± 1.83 × 10⁻⁶ cm/s. Metformin was used as a control and its apparent permeability value from the AP → BL side was 1.83 ± 0.47 × 10⁻⁶ cm/s and was within the reported range [42,43,44]. In the case of BL → AP, the apparent permeability of both N2 and N21 was low and below the LOQ level. These data suggest that both N2 and N21 are more effectively transported from the intestinal lumen (apical side) into the bloodstream (basolateral side) than being actively transported back from the bloodstream into the intestinal lumen. Volpe and colleagues have reported [24], that drugs have high permeability if the apparent permeability values are >14.0 × 10⁻⁶ cm/s. This study demonstrated high permeable nature of N2 and N21. These results also suggest that passive diffusion could be the mechanism underlying permeability of these molecules across Caco-2 cells.

Fig. 7.

Results of Caco-2 permeability study of N2 and N21, shown as time versus concentration graph (n = 9). AP → BL denotes apical to basolateral and BL → AP as basolateral to apical transport. The initial concentration in the donor compartment 69.9 ± 0.2 μM for N2 and 58.1 ± 0.7 μM for N21.

4. Conclusions

Our objective was to develop and validate an UPLC method and perform preformulation studies including solid-state characterization of two novel lead N2 and N21 ICs that are active against a wide panel of NTM pathogens. The developed UPLC methods were specific for the respective molecules and were used in determining their concentration from the study samples. Both molecules exist in crystalline state, N2 exhibited a sharp melting peak at 233 ^◦C without any polymorphic transitions. The DSC analysis, along with variable temperature XRD and HSM, confirmed that N21 can exists at least in two polymorphic forms and the polymorphic transition was monotropic in nature. According to regulatory guidelines, dissolution, solubility, permeability, and pKa are core components of the BCS class drugs and this study determined these physical properties during the preformulation work for both ICs. Both N2 and N21 possess low aqueous solubility (<0.2 μg/mL) and IDR (<22 ng/min/cm²) but high lipophilicity (LogP of 5–6) and permeability (18 × 10⁻⁶ cm/s). These weakly acidic drugs with a pKa of 7.86 for N2 and 7.62 for N21, respectively were most stable at pH 6.8. These preformulation findings will be crucial for developing rational formulations for both N2 and N21 for future clinical studies. Due to the low acidic nature of N2 and N21, future formulation as salts will be challenging, however these data support nanoparticle formulation or amorphous solid formulation. Since N2 and N21 have hydrogen-bonding capabilities, co-crystal formulation can also be a lead option. These physicochemical properties will be essential for developing dosage forms for these two potential drug candidates for the treatment of NTM infections. In order to help IC accumulation in the lung and overcome the poor oral drug absorption, liposomal formulation optimized for inhalation delivery will be a viable and a much-needed method for treatment of NTM infections. The future project will lead to develop IC loaded liposomes that are optimized for inhalation delivery directly to lungs.

The oral bioavailability of drugs depends on factors such as aqueous solubility, drug permeability, dissolution rate, first-pass metabolism, pre-systemic metabolism, and susceptibility to efflux mechanisms. However, the most frequent causes of low oral bioavailability are due to poor solubility and low permeability. Any drug to be absorbed must be present in the form of an aqueous solution, i.e. in the dissolved state at the site of absorption to achieve the desired concentration of the drug in systemic circulation for achieving the required pharmacological response [45]. Poorly water-soluble drugs often require high doses to reach therapeutic plasma concentrations after oral administration. Low aqueous solubility is the major problem encountered with the formulation development of APIs. The results of the present work showed that both N2 and N21 have high permeability but low aqueous solubility. Thus, their potent antimicrobial activity against mycobacterial species in vitro is a challenge for translating in in vivo. The improvement of drug solubility, thereby their bioavailability will be the major goal in the future study. There are many approaches available and reported in the literature to enhance the solubility of poorly water-soluble drugs. These techniques are chosen based on certain criteria such as the properties of the drug under consideration, the nature of excipients to be used, and the nature of the intended dosage form. Development of formulation with the goal to increase the solubility and the targeted delivery of N2 and N21 are underway. In order to improve water solubility of these drugs, future plan is to employ bioisosteric replacement of the indole ring resulting in an amino acid core, thus introducing an ionizable primary amine. In addition, to overcome aqueous solubility issues, we will formulate poorly soluble ICs and active amino acid-based analogs to improve oral absorption or deliver lead compound directly to the lung through inhalation delivery. Liposomes are assembled phospholipid vesicles that have been used to encapsulate both hydrophilic and hydrophobic drugs. Hydrophobic compounds, which the IC series is classified, can be trapped inside the hydrophobic lipid bilayer of the liposome for sustained drug release. Drug loaded liposomes have been shown to be preferentially taken up by macrophages, where NTM reside, thus delivering the anti-NTM drug in high concentrations to the mycobacteria. In order to help IC accumulation in the lung and overcome the poor oral drug absorption, liposomal formulation optimized for inhalation delivery will be a viable method for the treatment of NTM infections.

Supplementary Material

Supplementary data to this article can be found online at https://doi.org/10.1016/j.rechem.2025.102307.

Data availability

No data was used for the research described in the article.