Amorphous Drug Nanoparticles for Inhalation Therapy of Multidrug-Resistant Tuberculosis

Abstract

Tuberculosis (TB) is one of the most prevalent infectious diseases. The global TB situation is further complicated by increasing patient numbers infected with Mycobacterium tuberculosis (M.tb.) strains resistant to either one or two of the first-line therapeutics, promoted by insufficient treatment length and/or drug levels due to adverse reactions and reduced patient compliance. An intriguing approach to improve anti-TB therapy relates to nanocarrier-based drug-delivery systems, which enhance local drug concentrations at infection sites without systemic toxicity. Recently developed anti-TB antibiotics, however, are lipophilic and difficult to transport in aqueous systems. Here, the very lipophilic TB-antibiotics bedaquiline (BDQ) and BTZ (1,3-benzothiazin-4-one 043) are prepared as high-dose, amorphous nanoparticles via a solvent-antisolvent technique. The nanoparticles exhibit mean diameters of 60 ± 13 nm (BDQ) and 62 ± 44 nm (BTZ) and have an extraordinarily high drug load with 69% BDQ and >99% BTZ of total nanoparticle mass plus a certain amount of surfactant (31% for BDQ, <1% for BTZ) to make the lipophilic drugs water-dispersible. Suspensions with high drug load (4.1 mg/mL BDQ, 4.2 mg/mL BTZ) are stable for several weeks. In vitro and in vivo studies employing M.tb.-infected macrophages and susceptible C3HeB/FeJ mice show promising activity, which outperforms conventional BDQ/BTZ solutions (in DMF or DMSO) with an up to 50% higher efficacy upon pulmonary delivery. In vitro, the BDQ/BTZ nanoparticles demonstrate their ability to cross the different biological barriers and to reach the site of the intracellular mycobacteria. In vivo, high amounts of the BDQ/BTZ nanoparticles are found in the lung and specifically inside granulomas, whereas only low BDQ/BTZ-nanoparticle levels are observed in spleen or liver. Thus, pulmonary delivered BDQ/BTZ nanoparticles are promising formulations to improve antituberculosis treatment.

Increasing incidence rates of tuberculosis (TB) patients infected with multidrug-resistant Mycobacterium tuberculosis (M.tb.) strains (MDR-TB) represent an alarming threat to public health on a global level.¹ Development of drug resistance in M.tb. is an evolutionary process promoted by either treatment discontinuation and/or insufficient drug concentrations at the site of infection. First-line anti-TB drugs are often associated with liver or neurotoxicity, lowering patient compliance. Therefore, strategies to increase local drug concentrations at the site of infection in the lungs without systemic adverse effects are required.

Such strategies can be based on recent drugs with higher efficacy and improved pharmacokinetics or advanced drug-delivery strategies to achieve elevated drug concentration levels at the site of the infection.^2,3 Bedaquiline (BDQ) and 1,3-benzothiazin-4-ones (BTZs) are two current drugs effective against MDR-TB.⁴ BDQ blocks the proton pump of the mycobacterial ATP synthase and was approved in 2012 for oral administration.⁵ Although BDQ is practically insoluble in aqueous media and only very slightly soluble under acidic conditions, it shows a high oral bioavailability, especially when taken with meals. However, due to its high lipophilicity (log P = 7.13)⁶ and cationic, amphiphilic nature, BDQ shows extensive accumulation in peripheral tissues after oral administration (terminal elimination half-life of 4–5 months).⁵ Clinical studies with oral bedaquiline have shown mild but significant increases in quantitative tightening (QT) prolongation, a systemic side effect common to many cationic, amphiphilic drugs (CADs). Co-administration with other drugs causing QT prolongation, such as fluoroquinolones and clofazimine, revealed an additive effect, resulting in the restriction of BDQ prescribing to TB control programs that provide QT interval monitoring.⁷ Here, inhalation therapy could reduce systemic adverse effects and expand BDQ treatment options.

BTZs kill M.tb. by inhibition of the enzyme decaprenylphosphoryl-β-d-ribose 2′-epimerase, thereby preventing decaprenylphosphoryl arabinose formation.⁸ BTZ 043 is the most advanced compound in its class and currently undergoing clinical investigation. Like BDQ, BTZ 043 is also poorly soluble (0.014 g/L)⁹ with a moderate lipophilicity (log P = 2.89)⁶ and extensive gastrointestinal/hepatic metabolism. The combination of low solubility and high metabolism is thought to contribute to the low oral bioavailability of the compound seen in animals and humans. The poor oral bioavailability highlights the need for alternative drug-delivery strategies, such as inhalation, to enhance local BTZ lung concentrations and to improve efficacy.

Delivery of antibiotics directly to the lung has been shown to achieve high local lung concentrations while reducing the occurrence and severity of systemic side effects. One particular challenge with inhaled antibiotics is the typically higher therapeutic concentration required (mg doses) compared to, e.g., inhaled beta-agonists or corticosteroids (μg doses). High-dose inhalation products are also challenging due to the restricted amount of excipients (i.e., nonactive ingredients), which can be administered to the lung.¹⁰ In addition, studies have shown that inhalation of poorly soluble drugs at high doses can lead to accumulation of nondissolved drug particles, which induce nonspecific side effects, primarily lung inflammation.¹¹

Amorphous drug nanoparticles have highly desirable attributes for the pulmonary delivery of high-dose, poorly soluble antibiotics to the lung. The combined properties of enhanced dissolution rate with increased supersaturation levels (due to high surface area plus amorphous solid state) can achieve ideal lung dissolution kinetics,¹² i.e., slow enough to achieve maintenance of high lung concentrations while rapid enough to avoid side effects due to particle accumulation. Amorphous drug nanoparticles have the further benefit of requiring low excipient amounts and can be administered at high concentrations via nebulization.¹³ In this study, we prepared high-dose, amorphous drug nanoparticle formulations of BDQ and BTZ. Fluorescent labeling showed nanoparticle colocalization with intracellular M.tb. and drug accumulation at infection foci in the lung. In vivo efficacy studies using the pulmonary delivery route in M.tb.-infected mice demonstrated superiority of the nanoformulations compared to nonformulated drugs.

Results and Discussion

Synthesis of BDQ/BTZ Nanoparticles

To obtain stable aqueous nanoparticle suspensions of very lipophilic, water-insoluble drugs is both simple and challenging. Due to the low solubility of lipophilic drugs in water, a high supersaturation can be easily achieved, which supports the formation of nanoparticles.¹⁴ Since the lipophilic surface of the as-prepared nanoparticles dislikes interacting with water, however, the colloidal stability of the lipophilic nanoparticles is low, requiring surface stabilization. With these margin conditions, we have prepared BDQ and BDZ nanoparticles via a solvent-antisolvent approach.^15,16 Although widely applied for micron-sized drugs, examples of amorphous drug nanoparticles are low and typically utilize expensive supercritical CO₂-based technologies and/or polymeric stabilizers,¹⁵⁻²⁰ which are not approved excipients for use in inhalation products. Here, a concentrated “solvent” solution of BDQ/BTZ in DMSO was injected into water as the “antisolvent”. Whereas DMSO is soluble in water, BDQ and BTZ are not and precipitate immediately after injection (Figure 1). The nucleation of BDQ/BTZ as a nanosized solid is supported by low temperature (0 °C) and intense mixing (magnetic stirrer and sonicator). Moreover, ammonium acetate was used to further increase the polarity of the antisolvent and to increase the supersaturation of BDQ/BTZ in the aqueous phase due to the salting-out effect.

Figure 1.

Schematic synthesis of BDQ and BTZ nanoparticles: (a) “Solvent” solution with BDQ/BTZ in DMSO and “antisolvent” solution with TocP/MdP (for BDQ) or SDS (for BTZ) surfactants. (b) Nucleation of the BDQ/BTZ nanoparticles. (c) Structure of the resulting colloidally stable BDQ/BTZ nanoparticles in water.

Without certain stabilization, the BDQ/BTZ nanoparticles were—as expected—colloidally instable and agglomerated within 10 min. Therefore, a suitable surfactant was required to guarantee long-term dispersibility of the BDQ/BTZ nanoparticles in the aqueous phase. Pilot investigations revealed that a mixture of monododecylphosphate (MdP) and α-tocopherolphosphate (TocP) provided the best colloidal stabilization for the more aromatic BDQ, whereas sodium dodecyl sulfate (SDS) was specifically effective in stabilizing BTZ nanoparticles (Figure 1). Excess surfactants were removed by centrifugation and redispersion of the BDQ and BTZ nanoparticles prior to use. The presence of the polar phosphate or sulfate head groups of the surfactants on the particle surface is clearly indicated by zeta potential analyses showing negative surface charging of −40 to −60 mV in the biologically relevant pH range of 5 to 9 (Figure 2a,b). The negative charge maintained the colloidal stability of the BDQ/BTZ nanoparticles in the aqueous medium over 4–6 weeks. Moreover, the negative surface charge is advantageous for effective cell uptake, since M.tb.-infected macrophages as the primary reservoir of mycobacteria preferentially take up nanocarriers with negative zeta potential.²¹

Figure 2.

Colloid and particle properties of the BDQ/BTZ nanoparticles: (a,b) Zeta potential analysis with photos of aqueous suspensions. (c,d) STEM images. (e,f) XRD with neat BDQ and BTZ as references. (g) EDXS area scan of BDQ nanoparticles on STEM image with P/Br/C/N element mapping (in (h)) (based on P-K_α, Br-K_α, C-K_α, N-K_α lines). (h) Droplet size of nebulized BDQ/BTZ nanoparticle suspensions in comparison to the aerosol of an isotonic saline.

Material Characterization of the BDQ/BTZ Nanoparticles

Size and morphology of the as-prepared BDQ and BTZ nanoparticles were examined by electron microscopy and dynamic light scattering (DLS). Scanning transmission electron microscopy (STEM) images demonstrate the presence of spherical particles (Figure 2c,d) with mean diameters of 60 ± 13 nm (BDQ) and 62 ± 44 nm (BTZ) (based on statistical evaluation of ∼250 particles, SI: Figure S1). According to DLS, the drug nanoparticles have mean hydrodynamic diameters of 65 ± 20 nm (BDQ) and 69 ± 52 nm (BTZ) in water (SI: Figure S1). The larger hydrodynamic diameter and specifically the broader size distribution in water can be ascribed to the rigid adhesion layer of water molecules on the particle surface because of the high polarity and strong hydrogen bonding of H₂O.

The chemical composition of the BDQ and BTZ nanoparticles was assessed by different analytical methods. According to X-ray powder diffraction, the BDQ/BTZ nanoparticles are predominately amorphous, whereas the nonformulated (neat), micron-sized BDQ/BTZ reference samples are highly crystalline (particle size 20–30 μm, Figure 2e,f). Amorphous drug nanoparticles, in this regard, are advantageous to enhance the dissolution rate.^15,16 Fourier transform infrared (FT-IR) spectra of the as-prepared BDQ/BTZ nanoparticles are in good agreement with pure BDQ and BTZ (SI: Figure S2). Energy-dispersive X-ray spectroscopy (EDXS) proves the colocalization of P, Br, N, and C in the case of the BDQ nanoparticles (Figure 2g). Elemental analysis (EA, C/H/N/S analysis), thermogravimetry (TG, for total organics analysis), and photometry were used to quantify the drug-loading efficiency of the nanoparticles as well as the drug load per volume of suspension (SI: Table S1, Figures S3 and S4). As a result, the drug loading of the BDQ nanoparticles was 69% (with 31% TocP/MdP), and the drug loading of the BTZ nanoparticles was >99% (with <1% of SDS). BDQ and BTZ nanoparticle suspensions used in vitro and in vivo contained 4.10 mg/mL BDQ and 4.20 mg/mL BTZ.

Nebulization of amorphous drug nanoparticle dispersions is one method to administer sufficiently high drug doses to the lung without the need for added pharmaceutical excipients, such as lactose or mannitol, which are commonly used for dry powder-inhaler formulations.²² It is important that drug nanodispersions do not aggregate during nebulization, as this can influence the drug dissolution rate.²³ Equally, the highly concentrated nanosuspensions should not alter aerosol droplet size, as this can influence the deposition pattern in the lung.²³ BDQ/BTZ nanoparticle dispersions were used as prepared with 4.1/4.2 mg/mL BDQ/BTZ and nebulized with a commercially available vibrating-mesh nebulizer (SI: Figure S5). The nebulized BDQ/BTZ nanoparticle aerosol exhibited a volume/mass median geometric droplet size of 4.0 ± 0.1 μm and a size distribution similar to isotonic saline used as a reference solution (Figure 2h). Thus, nebulization is possible with droplet sizes suitable to reach the pulmonary region in the human lung.

In Vitro Dissolution Kinetics

As stated in the introduction, one of the advantages of amorphous drug nanoparticles is to modulate drug dissolution kinetics in the lung. To establish baseline information on BDQ/BTZ-nanoparticle dissolution kinetics, a dialysis system was used to study drug dissolution and diffusion into a reservoir compartment containing bovine serum albumin in 0.1 M PBS (10.8 mg/mL BSA), which was necessary to bind the lipophilic compounds and achieve sink conditions (SI: Figure 6).⁹ Nonformulated (neat), crystalline drug powders (particle size 20–30 μm) were used as controls. The highly lipophilic BDQ (both formulated nanoparticles and nonformulated neat drug) was found to rapidly saturate the receiver fluid and adsorb to all surfaces in the system within the first few hours of the experiment, necessitating further experimental optimization beyond the scope of the current study (SI: Figure 6a). The BTZ dissolution rate could be measured reliably with this method and showed that the BTZ-nanoparticle formulation increased the drug-release rate substantially (release at t_24h ≈ 18%) compared to nonformulated, micron-sized neat BTZ (release at t_24h ≈ 8%; SI: Figure 6b). Interestingly, the BTZ dissolution rate from the amorphous nanoparticles was slower than that of a similar BTZ compound encapsulated in human serum albumin nanoparticles (release at t_24h = 50%),⁹ indicating that formulation attributes can effectively modulate drug-release kinetics.

In Vitro Intracellular Localization and Efficacy

To allow in vitro and in vivo monitoring, the BDQ/BTZ nanoparticles were labeled with the lipophilic fluorescent dye Lumogen Red (LR) (Figure 3a). LR is known as an extremely efficient dye with a quantum yield near 100%,²⁴ so that very low amounts of <0.1 mol % LR per total solids are sufficient. During nanoparticle preparation, LR was added to the “solvent” solution (Figure 1). The presence of LR is indicated by the weak red color and the intense red luminescence that are visible even for the naked eye (Figure 3b). Excitation and emission spectra show the characteristic absorption at 380 to 620 nm and the deep-red emission at 590 to 800 nm peaking at 610 nm (Figure 3c). Similar to BDQ and BTZ, the concentration of LR can be quantified by photometry and resulted in 19 and 14 μg/mL for BDQ-LR and BZT-LR nanoparticles, respectively (SI: Figure S4). The fluorescence characteristics of the BDQ-LR/BTZ-LR nanoparticles are similar to dissolved LR in DMSO solution. In water, however, LR is insoluble, and aqueous suspensions of LR do not show any measurable emission (Figure 3c).

Figure 3.

Fluorescence of exemplary Lumogen Red (LR)-labeled BDQ nanoparticles and localization within M.tb.-infected macrophages: (a) Scheme of BDQ-LR nanoparticles. (b) Suspensions of BDQ-LR in daylight and with excitation (377 nm). (c) Excitation and emission spectra of BDQ-LR nanoparticles (with LR solution in DMSO and LR solution in water as references). (d,e) Murine BMMO-infected with GFP-expressing M.tb. (green) for 2 h was subsequently incubated with BDQ-LR nanoparticles (gray) and analyzed after 24 (c) and 48 h (d) by confocal fluorescence laser scanning microscopy with three-dimensional reconstruction of stacks from consecutive scans (blue: nuclei stained with blue fluorescent DAPI, red: LAMP-1-positive membranes).

To compare the localization of internalized BDQ-LR or BTZ-LR nanoparticles and intracellular M.tb., murine bone-marrow-derived macrophages (BMMO, 37 °C, 7.5% CO₂) were infected with M.tb.-GFP (GFP: green fluorescent protein). After 2 h postinfection (p.i.), cells were exemplarily incubated with BDQ-LR nanoparticles, and confocal laser scanning fluorescence microscopy was performed 24 and 48 h later. At both time points, the macrophages showed uptake of BDQ-LR nanoparticles (Figure 3d,e), whereby the LR signal was found in certain but not all LAMP-1-positive (LAMP-1: lysosomal associated membrane protein 1) endosomal compartments within the cytosol but excluded from nuclei (stained with blue fluorescent 4′,6-diamidino-2-phenylindole, DAPI) and nuclear membranes. In several cells, the BDQ-LR signals colocalized with intracellular M.tb.-GFP. Three-dimensional reconstructions of stacks from scans of single-confocal-microscopy layers show the BDQ-LR signal (gray) and mycobacteria (green) enwrapped by LAMP-1-positive phagosomal membranes (red) in close proximity to the nuclei (blue) (Figure 3d,e). These findings indicate that (i) LR was still incorporated within the BDQ-LR nanoparticles but not in all intracellular membranes, as otherwise, the LR signal would appear in a diffuse pattern across cells, and (ii) the nanoparticles were likely taken up into the same phagosomal compartments as those harboring intracellular mycobacteria. High drug concentrations in localized phagosomal compartments are thought to enhance the killing effect of the antibiotic and to reduce resistance development.

The antimycobacterial activity was subsequently assessed in BMMO infected with M.tb. 2 h prior to being treated with BDQ-LR/BTZ-LR nanoparticle suspensions (0.1 and 1.0 μg/mL drug; incubation up to 72 h). Several controls were included: (i) the surfactant solutions without drug (TocP/MdP or SDS, negative controls), (ii) BDQ/BTZ dissolved in DMF with subsequent dilution into cell-culture medium (0.1 and 1.0 μg/mL, positive controls), (iii) pure DMF (negative control), and (iv) nontreated infected cells (negative control). As expected, the negative controls of the surfactant solutions as well as DMF did not affect the mycobacterial growth, which was similar to nontreated cells as indicated by similar numbers of colony forming units per milliliter (CFU/mL) (Figure 4a–c). This also indicates that the surfactants TocP/MdP and SDS as such did not cause any observable cytotoxic, apoptotic, or necrotic effect to the infected cells. In contrast, both BDQ-LR and BTZ-LR nanoparticle systems showed similar or even a slightly higher antimycobacterial killing efficacy as solutions with BDQ/BTZ in DMF (Figure 4b,c). These results demonstrate the nanoparticle-formulated drugs to reach the intracellular mycobacteria at similar drug concentrations as the positive controls of the freely dissolved drugs.

Figure 4.

In vitro studies with BDQ-LR/BTZ-LR nanoparticles: (a) Schematic drawing of antibiotic activity tests in M.tb. H37Rv-infected macrophages by determining the CFU/mL. (b,c) Antimycobacterial activity of BDQ-LR (b) or BTZ-LR (c) nanoparticles after 48 and 72 h at 1.0 and 0.1 μg/mL drug concentrations (red bars, aqueous suspension) compared to the following controls: surfactant solutions (light green bars, TocP/MdP or SDS used with an identical concentration as for the BDQ-LR/BTZ-LR nanoparticle suspensions, negative controls), BDQ/BTZ dissolved in DMF (light orange bars, positive controls), pure DMF (blue bars, negative control), and infected, nontreated BMMO cells (gray bars, negative control). Data represent mean ± standard deviations from independent n = 3 experiments.

In Vivo Localization and Pharmacodynamic Activity

Based on the promising in vitro results, the efficacy of the BDQ/BTZ nanoparticles was assessed versus drug solutions prepared in DMF and diluted in isotonic saline (positive control) in a murine TB model using the susceptible C3HeB/FeJ mouse strain. The advantage of this strain is that it develops necrotic granulomas similar to the lesions observed in active TB in humans.²⁵ Animals were infected with H37Rv, a virulent strain of M.tb., via an aerosol chamber (see SI). After 30 days post infection, 40 μL of undiluted BDQ-LR or BTZ-LR nanoparticle suspensions equal to 4.1/4.2 mg/mL BDQ/BTZ (6.56/6.72 mg/kg) drug concentration were administered intranasally (i.n.), followed by six further doses every second day for 2 weeks. Intranasal administration of volumes >35 μL (up to 70 μL) were shown to result in 40–50% lung deposition of the administered dose.²⁶ Since other forms of pulmonary administration either (i) result in extremely low lung deposition (e.g., ∼8% aerosol delivery with nose-only inhalation chambers)²⁷ or (ii) required injectable anesthesia and are therefore not suitable for pilot studies with frequent dosing schedules, i.n. administration was considered the most appropriate noninvasive administration method for this pilot study. Control groups were either treated with equal amounts of BDQ/BTZ dissolved in DMF (5%; 4.1/4.2 mg/mL or 6.56/6.72 mg/kg BDQ/BTZ) (positive controls) or not treated at all (negative controls). Two weeks (14 days) after the first treatment, the mice were sacrificed, and organs removed for fluorescence imaging and CFU determination (Figure 5a).

Figure 5.

In vivo studies with BDQ-LR/BTZ-LR nanoparticles: (a) Scheme showing the treatment of C3HeB/FeJ mice with M.tb. H37Rv infection and BDQ-LR/BTZ-LR nanoparticle treatment. (b) LR-associated fluorescence in lung, spleen, and liver after 14 days of treatment. (c) CFU/mL in lung and spleen after 14 days of treatment (seven doses every second day) of either BTZ/BDQ nanoparticles (164/168 μg/40 μL i.n. dose; red bars) compared to BTZ/BDQ dissolved in DMF and then diluted in saline (164/168 μg/40 μL i.n. dose; positive control; orange bars). Nontreated, infected mice were used as a negative control (gray bars). Data represent mean ± standard deviations from n = 5 mice (statistics relates to unpaired t test).

Fluorescence LR signals were detectable in the lungs, especially of BDQ-LR-treated animals, whereas no signals were detected in spleens and livers of nanoparticle-treated animals (Figure 5b). Significant pharmacodynamic activity could be observed for both the positive control and the BDQ/BTZ nanoparticles in comparison to the untreated group (Figure 5c). Intranasal administration of BDQ-LR nanoparticles resulted in the reduction of M.tb. counts in lungs and spleens comparable to what was seen upon administrating the free drug. However, equal dosages of BTZ-LR nanoparticles resulted in a significant 40–50% reduction of M.tb. counts in the lungs as well as slightly but not significantly lower CFUs in spleens when compared to mice, which received the free drug (Figure 5c, lower panel). These findings indicate the nanoparticle-mediated drug efficacy following i.n. administration against both pulmonary as well as splenic mycobacteria. Most importantly though, the antimycobacterial activity of nanoparticle-formulated BTZ in the lung was superior over the freely dissolved drugs (positive control) despite much lower fluorescent signals in the lungs when compared to signals from BDQ-LR nanoparticles (Figure 5b,c).

Fluorescence microscopy of lung tissue sections of infected C3HeB/FeJ mice treated with BDQ-LR nanoparticles revealed granulomas as indicated by the density of DAPI stained nuclei of infiltrating inflammatory cells (Figure 6; compare to SI: Figure S8a). Distribution of the deep-red fluorescence demonstrates that the BDQ-LR nanoparticles not only reach the deep lung and accumulate primarily in direct vicinity to the granulomas but also in close proximity to M.tb (SI: Figure S7). More importantly, BDQ-LR nanoparticles were present in macrophages, the prime host cell of M.tb. as revealed by cell morphology (Figure 6c,d). The majority of these cells stained positive for the alveolar macrophage marker F4/80. Interestingly, nanoparticle aggregates were mostly associated with F4/80 negative cells (Figure 6e,f: lower excisions). Finally, it should be mentioned that the pathological alterations observed in the lungs did not crossly differ between BDQ-LR/BTZ-LR nanoparticle-treated, free-drug-treated, and nontreated mice indicating that treatment neither affected pathological sequelae nor induced any obvious adverse tissue reactions (SI: Figure S8a). Quantification of histopathological alterations, however, revealed that lungs of mice treated with BTZ-LR nanoparticles had smaller inflamed areas than other treatment groups, which goes along the lower CFU counts in these mice (Figures S8b and S9). Parallel studies conducted in noninfected Balb/c mice did not show evidence that intranasal administration of BTZ-NP caused local lung irritancy (as assessed by lung proinflammatory cytokine levels, neutrophilic influx into the lung, or markers of tissue damage) up to 48 h following a single i.n. administration (SI: Figure S10).

Figure 6.

BDQ-LR nanoparticles in lung tissue: (a,f) Different fluorescence microscopy images showing lung tissue with inflamed regions (granuloma) on different scales of magnification with BDQ-LR nanoparticles (red), DAPI-stained nuclei (blue), and F4/80-stained M.tb. (green).

Conclusions

Amorphous bedaquiline (BDQ) and 1,3-benzothiazin-4-one 043 (BTZ) nanoparticles were prepared by a simple solvent-antisolvent approach. The nanoparticles have extraordinary drug loads with 69% BDQ and >99% BTZ of total nanoparticle mass. They exhibit mean diameters of 60 ± 13 nm (BDQ) and 62 ± 44 nm (BTZ) are available as stable aqueous suspensions with high drug loads of 4.1–4.2 mg/mL. Both nanoparticle systems are labeled with Lumogen Red (LR), whereby an intense deep-red fluorescence could be achieved to enable localization of nanoparticles both in vitro and in vivo.

The amorphous BDQ-LR/BTZ-LR nanoparticles are highly effective in killing M.tb. both in macrophages in vitro (BMMO) as well as in vivo using a M.tb. susceptible murine model (C3HeB/FeJ mice). In vitro studies demonstrate the ability of the BDQ-LR/BTZ-LR nanoparticles to efficiently cross the different biological barriers and to reach the phagosomal site of intracellular mycobacteria with high concentration. In vivo studies reveal that the BDQ-LR/BTZ-LR nanoparticles are differentially distributed across the lung with accumulations around as well as within granulomas. When comparing BDQ-LR/BTZ-LR nanoparticles with the BDQ/BTZ positive controls (freely dissolved drugs in DMF diluted in saline), pulmonary administration of BTZ-LR nanoparticles reduces the number of lung-associated M.tb. CFU/mL by 50%, whereas BDQ-LR nanoparticles lead to similar CFU counts as the free drug. These results demonstrate a superior in vivo efficacy of the BTZ-LR nanoparticles over the nonformulated drug. In summary, amorphous drug nanoparticles for inhalation therapy of poorly soluble antibiotics using nebulizer systems overcome many of the previously reported limitations associated with inhaled antibiotics. They can be administered in therapeutically relevant quantities to infected sites in the lung, and they are stable during storage as well as nebulization and show desirable dissolution kinetics for lung administration. Further development of this therapeutic strategy provides promising solutions to further improve treatment schemes by both, BDQ and BTZ.

Materials and Methods

LR-Labeled BDQ Nanoparticles (BDQ-LR)

12 mg of BDQ (21.6 μmol, >98%, MedKoo Biosciences, USA) and 45 μg of LR (88.0 nmol, Kremer Pigmente, Germany) were dissolved in 1 mL of DMSO (≥99.9%, Sigma-Aldrich, Germany) as the solvent. Moreover, 2.5 mg of disodium α-tocopherolphosphate (4.5 μmol, 97%, Alfa Aesar, Germany), 3.7 mg of sodium monododecylphosphate (12.3 μmol, >75%, TCI Chemicals, Germany), and 30 mg of ammonium acetate (0.39 mmol, VWR, Germany) were dissolved in in 10 mL of demineralized water as the antisolvent. The antisolvent solution was cooled to 3 °C in an ice bath. Thereafter, 0.4 mL of the solvent solution containing 4.8 mg of BDQ and 18 μg of LR were injected into the antisolvent solution via a syringe with vigorous stirring (750 rpm). In addition, the antisolvent solution was mixed by a sonicator with a sonication amplitude of 50% (HD2070, Bandelin, Germany) during and until 10 s after the injection of the solvent solution. Finally, the as-prepared BDQ nanoparticles were separated by centrifugation (25 000 rpm, 20 min). Thereafter, the BDQ nanoparticles were resuspended in 0.8 mL of the antisolvent solution and sonicated for 2 s. The suspensions should be stored in the dark at 3–5 °C.

LR-Labeled BTZ Nanoparticles (BTZ-LR)

BTZ-LR nanoparticles were prepared similarly to the aforementioned BDQ nanoparticles. Instead of 12 mg of BDQ, 10 mg of BTZ (30.8 μmol, 99.66%, MedChemExpress, Germany), and 75 μg of LR (69.5 nmol, Kremer Pigmente, Germany) were dissolved in 1 mL of DMSO (≥99.9%, Sigma-Aldrich, Germany) as the solvent. Moreover 5 mg of sodium dodecylsulfate (SDS, 17.3 μmol, 99%, abcr, Germany) and 30 mg of ammonium acetate (0.39 mmol, VWR, Germany) were dissolved in 10 mL of demineralized water as the antisolvent.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsnano.3c01664.

• Details related to the analytical techniques, material characterization, aerosolization, in vitro studies, and in vivo studies (PDF)

The authors declare no competing financial interest.