(T1030-07-42) Follicular Delivery of 4-phenylbutyric Acid PLGA Nanoparticles to Mitigate Chemical Warfare Agent-Induced Skin Injury: A Quality by Design Approach

Purpose: Lewisite, an arsenic-based chemical warfare agent and vesicant, causes early skin irritation, and prolonged pain, inflammation, and blistering. To mitigate these effects, we repurposed 4-phenylbutyric acid (4-PBA)—an FDA-approved drug for urea cycle disorders has shown promise by inhibiting the underlying pathogenic mechanism, particularly unfolded protein response. The present study aims to develop a topical formulation for enhanced follicular delivery of 4-PBA. Encapsulation of 4-PBA in poly(lactic-co-glycolic acid) (PLGA) nanoparticles (NPs) is hypothesized to improve follicular delivery to the site of injury, while a topical gel formulation would prolong drug contact with the affected skin. To establish a systematic and robust approach for formulation development, a Quality by Design (QbD) methodology and Design of Experiments (DoE) were implemented. Critical material attributes (CMAs) and critical process parameters (CPPs) of the PLGA nanoparticle formulations were identified and correlated with the critical quality attributes (CQAs) of the final product to ensure the desired quality target product profile (QTPP).

Methods: 4-PBA-loaded PLGA NPs were prepared using a nanoprecipitation technique. The organic phase consisted of varying ratios of 4-PBA and PLGA dissolved in dichloromethane (DCM), while the aqueous phase comprised 1% w/w polyvinyl alcohol (PVA) in deionized water. The organic phase was injected into the aqueous phase using a syringe pump at different rates, which were systematically varied and optimized. The resulting mixture was stirred for three hours to ensure complete evaporation of DCM. The precipitated nanoparticles were then collected by ultracentrifugation at 13,400 rpm for 10 minutes. A custom response surface design (RSD) was generated using JMP® Pro software to evaluate the effects of two independent variables: the drug-to-polymer ratio and the injection rate. The impact of these variables was assessed on four critical quality attributes: drug loading (%), polydispersity index (PDI), particle size (nm), and zeta potential (mV) of the 4-PBA-loaded PLGA nanoparticles. Prediction profiler analyses were performed to understand the relationships between the independent variables and the measured responses. The optimized formulation predicted by the model was validated experimentally, with confirmation runs (n=3) demonstrating robust prediction. The optimized nanoparticle batch (NP1) was subsequently incorporated into gel formulations using carboxymethylcellulose (CMC) and Carbopol 980NF as gelling agents. The NP1-loaded gels were evaluated for in vitro permeation (IVPT) using Franz diffusion cells, with full-thickness porcine skin, respectively (n = 4). Unencapsulated 4-PBA in PBS served as the control. Receptor samples were taken at predetermined time points. The amounts of 4-PBA in the hair follicles and skin were determined using a differential tape-stripping technique and analyzed using a validated HPLC method.

Results: The optimized nanoparticle formulation (NP1), with a drug:polymer ratio of 1:1 and an injection rate of 1 mL/min, was predicted with a desirability factor of 0.89 and confirmed experimentally. NP1 exhibited robust physicochemical properties, including a drug loading of 44.75 ± 0.01%, a polydispersity index (PDI) of 0.3 ± 0.06, a mean particle size of 359 ± 33.54 nm, and a zeta potential of –11.2 mV (table. 1). Actual versus predicted plots for all responses showed p-values lesser than 0.05, except for zeta potential, indicating that zeta potential was not a critical quality attribute and was thus not considered when getting the prediction profiler data (Fig 1.) Scanning electron microscopy confirmed uniform, spherical nanoparticles, while FTIR analysis demonstrated potential molecular interactions between 4-PBA and PLGA, supporting stable encapsulation. IVPT indicated significant enhancement in follicular and total skin delivery with NP1 gels. NP1 in Carbopol gel achieved a 24.79% increase in follicular delivery (58.14 ± 16.54 cm²) and a 70.59% increase in total skin delivery (90.43 ± 5.12 cm²), while NP1 in CMC gel showed a 90.53% increase in follicular delivery (88.77 ± 16.05 cm²) and a 93.12% increase in total skin delivery (102.43 ± 13.64 cm²), compared to unencapsulated control (46.59 ± 16.54 cm²), (Fig 3.). All gel formulations were clear, homogeneous, and skin-compatible in pH (Carbopol: 4.9–5.8; CMC: 6.6–7.0). NP1 slightly reduced the density of Carbopol gels (0.91 vs. 0.96) and increased it in CMC gels (1.02 vs. 1.00), indicating minor structural changes. Texture analysis showed NP1 reduced hardness and increased adhesiveness in Carbopol gels (3.3 ± 1.2 g vs. 5.5 ± 0.1 g; –24.7 ± 0.4 g·s vs. –8.4 ± 2.5 g·s), while in CMC gels, it slightly increased hardness (1.8 ± 0.4 g vs. 1.5 g) and significantly enhanced adhesiveness (–11.1 ± 0.1 g·s vs. 0.05 g·s). Both gels retained good cohesiveness, with CMC showing greater springiness and resilience. Overall, the NP1-loaded with the CMC-based gel showed slightly superior performance in drug delivery and texture attributes.

Conclusion: Using a QbD approach, 4-PBA-loaded PLGA nanoparticles were optimized for enhanced skin and follicular delivery. Incorporation into CMC and Carbopol gels produced stable, clear formulations with favorable mechanical properties. The CMC gel showed superior drug delivery and texture, supporting its potential as a topical strategy for mitigating chemical warfare agent-induced skin injury.

References: 1. Li C, Srivastava RK, Weng Z, Croutch CR, Agarwal A, Elmets CA, et al. Molecular Mechanism Underlying Pathogenesis of Lewisite-Induced Cutaneous Blistering and Inflammation: Chemical Chaperones as Potential Novel Antidotes. American Journal of Pathology. 2016 Oct 1;186(10):2637–49.
2. Kshirsagar S, Dandekar A, Srivastava RK, Khan J, Muzaffar S, Athar M, et al. Microneedle-mediated transdermal delivery of N-acetyl cysteine as a potential antidote for lewisite injury. Int J Pharm. 2023 Nov 25;647.
3. Dandekar A, Vora D, Yeh JS, Srivastava RK, Athar M, Banga AK. Enhanced Transdermal Delivery of N-acetylcysteine and 4-phenylbutyric acid for Potential Use as Antidotes to Lewisite AAPS PharmSciTech. AAPS PharmSci Tech. 2023 Jan 31;
4. Vora D, Dandekar AA, Srivastava RK, Athar M, Banga AK. Development and Evaluation of a Topical Foam Formulation for Decontamination of Warfare Agents. Mol Pharm. 2022 Dec 5;

Acknowledgements: This project was supported by NIH/NIAMS U01AR078544 to M.A. The sponsors had no involvement in the design of the study, data collection, analysis, or interpretation.


Table 1. The custom RSD matrix for development of 4-PBA loaded PLGA NPs, which generates combinations of three levels and two factors, comprising 3 center points, and response values were drug loading, particle size, PDI, and zeta potential after experimentally running the matrix (n=3).


Fig 1. 4-PBA loaded PLGA NPs drug loading% %, particle size, PDI, generated by the JMP software. The graphs provide information about the values of the response variables predicted when using a particular combination of values of the factors. It also gives a desirability parameter that predicts how likely it is for the combination to match the predicted values. The value closer to 1 implies a higher probability of getting values close to the prediction.


Fig 2. Follicular delivery using differential tape stripping and total skin delivery of 4-PBA nanoparticles in NP 1 loaded gels (3 mg/mL in finite dose) in full-thickness porcine skin.