QbD-based co-loading of paclitaxel and imatinib mesylate by protamine-coated PLGA nanoparticles effective on breast cancer cells

Abstract

Aim: Paclitaxel and imatinib mesylate are drugs used in the treatment of breast cancer. Conventional drug-delivery systems have limitations in the effective treatment of breast cancer using the drugs.

Materials & methods: Combination index studies were used to identify the optimum ratio of both drugs showing maximum synergistic effect. Using a systematic quality-by-design approach, protamine-coated PLGA nanoparticles co-loaded with paclitaxel and imatinib mesylate were formulated. Further characterization and cell line evaluations were performed.

Results: Encapsulation efficiency obtained was 92.54% for paclitaxel and 75.12% for imatinib mesylate. A sustained (24 h) and controlled zero-order drug release was obtained.

Conclusion: Formulated nanoparticles had a low IC₅₀ value and enhanced cellular uptake.

Graphical Abstract

Plain language summary

Article highlights.

• Paclitaxel and imatinib (PXL and IMA) show synergistic efficacy against the MCF-7 cell line of breast cancer.

• PXL and IMA in the ratio of 4:1 show maximum effect and lowest combination index (CI).

• Quality-by-design methodology led to optimization and systematic formulation of coloaded protamine-coated PLGA nanoparticles.

• Protamine decreased the initial burst release of the formulations.

• A sustained zero-order drug release was obtained for both the formulations.

• Protamine coating was clearly visible on images using transmission electron microscopy analysis.

• Significant reduction in IC₅₀ values showing increased cytotoxicity was observed.

• Roughly 30-fold increase in cellular uptake was obtained as compared with free drugs.

1. Background

Cancer is a life-threatening disease described as the rapid growth of abnormal cells in the human body and is among the major causes of death globally [1]. There has been a rapid increase in number of cancer patients year on year [2]. Out of various cancer types such as lung, ovarian, breast, liver, brain, and colon cancer; the most widespread type of cancer worldwide is breast cancer [3].

Breast cancer is the most common cause of cancer death in females globally [4]. The first line treatment for cancer is chemotherapy, however, there are a few disadvantages associated with this treatment such as high systemic toxicity and low therapeutic effectiveness. Moreover, poor targeting and development of drug resistance is also a major concern allied with chemotherapy [5,6]. Therefore, there is a vital need to develop a novel delivery system to overcome such challenges.

Combination therapy is recognized as one of the most efficient treatments for cancer, especially breast cancer [7]. Combination therapy can prove to be more efficient in treatment if drugs used in combination therapy have synergism [8]. The ratio at which drugs show the maximum synergism should be identified [9]. Combination index studies help identify this ratio [9,10]. Combination therapy using the right choice of drugs and at the right concentrations can lead to efficient breast cancer treatment.

Paclitaxel (PXL) is the fundamental and first line of treatment of breast cancer [11]. However, treatment by PXL has limited efficacy and is marked by several drawbacks [12]. Combination therapy has been explored for increasing of efficacy of PXL against breast cancer [13,14]. Several studies have also focused on co-delivery of PXL along with imatinib mesylate (IMA) [15–17]. However, co-delivery of a hydrophilic and a hydrophobic drug could be challenging for achieving the right encapsulation efficiency of each drug [18]. To improve the delivery of IMA in breast cancer, novel drug-delivery approaches are often explored to increase the efficacy of the treatment.

Nanotechnology can lay the groundwork for future breast cancer treatment by specifically targeting breast cancer cells thereby improving drug delivery [6,19]. PLGA nanoparticles are polymeric nanoparticles that have been recognized as an efficient drug-delivery system in the delivery of anticancer drugs. Several investigations have also explored the role of PLGA in drug delivery for the treatment of breast cancer. Protamine sulfate coating is known to increase cellular uptake and improve the delivery of drugs [20,21]. It has been explored typically in drug delivery of anticancer drugs. It also decreases the initial burst release of the drug and is also known to control the drug release from nanoformulation [22].

This investigation aims to systematically formulate PXL and IMA co-loaded PLGA nanoparticles, coated with protamine sulfate. The first part of the study aims at optimizing the ratio of the two drugs showing maximum synergism, using the principles of combination index analysis. Further, systematic quality-by-design (QbD)-based design of experiments (DoE) methodology has been used for formulation and optimization of protamine-coated co-loaded PLGA nanoformulation. In vitro characterization of the formulation has been done to evaluate drug release. To prove the efficacy of the formulated nanoparticles MTT cytotoxicity and cellular uptake analysis have been performed. QbD-enabled optimization and promising results on MCF-7 cell lines show the translatory potential of formulation, subject to further investigations and trials.

2. Materials & methods

2.1. Materials

PXL was received as a gift sample from Cipla Laboratories Ltd, Mumbai, India, and IMA was kindly gifted by Natco Pharma, India. PLGA (75:25) was received as a gift sample from Evonik, India. Protamine was purchased from TCI Chemicals. Trehalose was purchased from Sigma-Aldrich (MA, USA). Solvents were of analytical grade and were purchased from local vendors.

2.2. Optimizing the ratio of PXL & IMA for maximum efficacy

For combinatorial therapy, the optimum ratio of drugs is an essential parameter for the maximum therapeutic effect [23]. Cytotoxicity studies were performed on MCF-7 (breast cancer cell line) cell lines to optimize the PXL:IMA ratio that should be loaded in the final formulation. The methodology given by Shi et al. [24] was followed with some modifications. Briefly, 96-well plates were used to seed the cells at 5000 cells per well in 200 ml of RMPI-1640 medium (Roswell Park Memorial Institute-1640 medium) and were incubated at 37°C in a 5% CO₂ for 24 h. Further, the culture medium was removed and replaced with the sample solution having mass ratios of PXL: IMA as (a) 1:0, (b) 0.75:0.25, (c) 0.5:0.5, (d) 0.25:0.75 and (e) 0:1, respectively. After incubating it for 48 h, an MTT assay on the cells was performed and the optimum mass ratio was calculated.

2.3. Combination index

The combination index was calculated using the iso-bologram analysis method based on the Chou-Talay method [9,10]. The combination index was calculated using Equation (1).

$$C{I}_x=\left(\frac{C_{c-PXL}}{C_{PXL}}\right)+\left(\frac{C_{C-IMA}}{C_{IMA}}\right)$$ (1)

In Equation (1), CI is the combination index. C_c-PXL and C_c-IMA are the inhibitory concentrations (IC_x) values of drugs PXL and IMA, respectively, in the combination system. C_PXL and C_IMA are the inhibitory concentration values of PXL and IMA drugs alone, respectively. If CI_X value is obtained less than 1 (CI <1) it shows synergism of two drugs in the combination, whereas more than 1 (CI >1) shows antagonism.

2.4. Formulation of co-loaded protamine-coated PLGA nanoparticles

PLGA nanoparticles were formulated by the double emulsion solvent evaporation method. Aqueous phase 1 was prepared by dissolving IMA in distilled water. PLGA and PXL were dissolved in dichloromethane to form the organic phase. Aqueous Phase I was added to the organic phase while stirring using a magnetic stirrer (Remi-RML) to form a W/O emulsion. Further, this emulsion was added to aqueous Phase II (2%-Poloxamer^® 407 solutions) to form a W/O/W emulsion followed by probe sonication for size reduction. Last, the obtained emulsion was subjected to overnight stirring to evaporate the organic phase.

Formed PLGA nanoparticles were coated with protamine by adsorption methodology as reported by Dhami et al. [22]. Briefly, the obtained PLGA nanoparticle emulsion was added dropwise to the protamine solution while being subjected to magnetic stirring for 8 h.

The entrapment efficiency of prepared PLGA nanoparticles was analyzed using the methodology used by Yadav et al. with some modifications [25]. Briefly, the dispersion was subjected to ultracentrifugation at 20,000 r.p.m. for 30 min at 4°C in a refrigerated ultracentrifuge (Beckmann Coulter Optima Max-XP Ultracentrifuge). The supernatant was carefully separated and analyzed for the amount of drug unentrapped using the reversed-phased high-performance liquid chromatography (RP-HPLC) method as described by Peres-Filho et al. [15]. Briefly, PXL and IMA were quantified simultaneously using an HPLC system (Waters) coupled with a PDA detector using Kromasil C-18 column 4.6 × 100 mm. A mobile phase composed of acetate buffer pH 8.5 and acetonitrile in a ratio of 1:1 was used at a flow rate of 0.75 ml/min, and encapsulation efficiency was calculated using the equation given in Equation (2). The pellet was resuspended in water and freeze-dried to obtain a dried powder using trehalose as a cryoprotectant. Trehalose in a concentration of 10% (w/w to PLGA) was used as a cryoprotectant as it is known to maintain particle integrity after lyophilization [26].

$$\begin{array}{c}\mathrm{Encapsulation}\mathrm{Efficiency}=\\ \frac{\mathrm{Total}\mathrm{amount}\mathrm{of}\mathrm{drug}-\mathrm{Amount}\mathrm{unetrapped}}{\mathrm{Total}\mathrm{amount}of\mathrm{Drug}}\times 100\end{array}$$ (2)

2.5. Optimizing the amounts of drugs to be added to obtain drugs in the desired ratio

The ratio of the drug to be entrapped in the final formulation was calculated using cytotoxicity studies [27]. Drugs have different entrapment efficiency, hence amount to be added has to be different to achieve the desired ratio. Using individual entrapment data reverse calculations were performed to find approximate amounts of drugs to be added. Further batches were prepared using the calculated amount and checked for the actual entrapment.

2.6. Quality-by-design approach

2.6.1. Quality target product profile & critical quality attributes

Desired outcomes intended from formulation are termed as quality target product profile (QTPP) [28]. The selection of the correct profile is essential as it sets the base for the optimization of the formulation [29]. QTPPs were identified based on the published literature, the basics of the formulation, and general knowledge about the dosage form [30].

To meet the desired quality target product profile, critical quality attributes (CQAs) were defined [31]. These are the attributes that need to be achieved to attain the desired profile [32]. These attributes were defined taking into consideration QTPPs and the set target for the formulation [33].

2.6.2. Construction of Ishikawa fishbone diagram

The Ishikawa fishbone diagram is used for mapping CQAs with critical material attributes (CMAs) and critical process parameters (CPPs) [34]. This mapping is based on the principles of the ‘Six-Sigma’ methodology [28]. Ishikawa fishbone diagram was prepared using the Minitab^® 18 software (M/s Minitab, Inc, PA, USA).

2.6.3. Failure mode & effect analysis including risk priority number base ranking

Failure mode and effect analysis (FMEA) is an analysis to identify the factors having the maximum effect on CMAs. FMEA is based on the Risk Priority Number (RPN) methodology [28,30]. RPN was calculated using Equation (3), where S, O, and D are severity, occurrence, and detectability, respectively [30].

$$\text{RPN}=\text{S}\times \text{O}\times \text{D}$$ (3)

2.6.4. Design of experiment & optimization of protamine-coated PLGA nanoparticles

DoE is a powerful methodology for optimization, based on simultaneous variation of variables combined with statistical simulations and predictions to configure parameters for obtaining the optimized formulation having desired attributes (CQAs) [35]. Factors for optimization were screened using the FMEA analysis. Three factors (independent variables) were identified to be optimized to achieve the desired results (dependent variables) [36]. Identified independent variables were: amplitude of probe sonication; amount of PLGA and concentration of protamine solution. The dependent variables identified were mean particle size and zeta potential. For optimization, a 3³ Box-Behnken Design (BBD) was utilized as it can be used to study three factors at three levels with less number of runs as compared with full factorial design [30,37]. DoE simulation was performed using Design Expert (Design Expert 14, Stat-Ease, MN, USA) software. A 3³ BBD-based DoE model works based on the polynomial equation given in Equation (4). ANOVA-based analysis showing the effects of pure, interaction, and quadratic variables on dependent variables was obtained [38]. Further, for optimization of the results, an overlay analysis paired with graphical optimization based on set goals and constraints was done to get an optimized batch having the desired parameters [28]. This batch was repeated in triplicate and the obtained results were compared with the predicted results [39].

$$\begin{array}{c}{\text{Y}}_{\left(\text{n}\right)}={\beta }_0+{\beta }_1{\text{X}}_1+{\beta }_2{\text{X}}_2+{\beta }_3{\text{X}}_3+{\beta }_{12}{\text{X}}_1{\text{X}}_2\\ +{\beta }_{13}{\text{X}}_1{\text{X}}_3+{\beta }_{23}{\text{X}}_2{\text{X}}_3+{\beta }_{11}{\text{X}}_1^2+{\beta }_{22}{\text{X}}_2^2+{\beta }_{33}{\text{X}}_3^2\end{array}$$ (4)

2.7. Characterization of formulated protamine-coated PLGA nanoparticles

2.7.1. Mean particle size, polydispersity index & zeta potential

Mean particle size, polydispersity index, and zeta potential were analyzed using Malvern^® Zeta Sizer [40]. Malvern Zeta Sizer works on the principle of dynamic light scattering [41]. Formulations were appropriately diluted before analysis. Standard disposable cuvettes were used for the analysis.

2.7.2. Transmission electron microscopy

Both protamine-coated and uncoated PLGA nanoparticles were visualized to confirm the surface morphology of the nanoformulations using transmission electron microscopy (TEM) (TEC-12, TECNAI G2 SPIRIT BIOTWIN) operating at 100 kV. Diluted nanoparticles were affixed on carbon-coated copper grids and negatively stained using phosphotungstic acid.

2.7.3. FTIR

Fourier transform infra-red scanning was performed for PXL, IMA, PLGA, protamine, uncoated PLGA nanoparticles, and protamine-coated PLGA nanoparticles using Perkin Elmer FTIR.

2.8. In vitro drug release

In vitro drug release was determined against the dialysis membrane using the methodology given by Bhargawe et al. [40]. For evaluation of drug release, 2 ml of samples (drug solution, PLGA nanoparticles, and protamine-coated PLGA nanoparticles) were poured into different drug-release assemblies. Semi-permeable dialysis membrane was utilized and the dissolution medium was pH 7.4 phosphate-buffered saline (PBS). Samples were withdrawn at pre-decided time points of 0.25, 0.50, 1, 2, 4, 6, 8, 12, 18 and 24 h. Samples were analyzed using the RP-HPLC method. Drug-release kinetic modeling was performed using mathematical modeling (zero-order drug release, first-order drug release, Higuchi kinetics, and Korsmeyer Peppas kinetics) [28,42,43].

2.9. In vitro cell line studies

2.9.1. MTT cytotoxicity

MTT cytotoxicity studies were performed on MCF-7 cell lines as per the methodology given by Shi et al. [24]. Briefly, 96-well plates were used to seed the cells at 5000 cells per well in 200 ml of RMPI-1640 medium (Roswell Park Memorial Institute-1640 medium) and were incubated at 37°C in a 5% CO₂ for 24 h. Further, the culture medium was removed and replaced with the sample solutions. Sample solutions were (A) PXL-loaded PLGA nanoparticles, (B) IMA-loaded PLGA nanoparticles, (C) PXL and IMA co-loaded PLGA nanoparticles, (D) PXL-loaded PLGA nanoparticles coated with protamine, (E) IMA-loaded PLGA nanoparticles coated with protamine and (F) PXL and IMA co-loaded PLGA nanoparticles coated with protamine. After incubating it for 48 h, an MTT assay on the cells was performed and the optimal mass ratio was calculated.

2.9.2. Cellular uptake

Cellular uptake of the formulation was visualized in MCF-7 cell lines. Cells were subjected to 100 μg/ml test solution for 8 h to assess uptake. At set intervals, the medium was withdrawn and wells were rinsed three times with 200 μl of cold PBS to remove any remaining drugs [44]. Centrifugation (500 × g, 10 min) consolidated cells after trypsinization. After dissolving the pellet in 200 μl of DMSO, the cells were lysed using a bath sonicator for 15 min. Microplate reader Fluo-star Omega, BMG Labtec) was used to measure the fluorescence intensity (Excitation -485 nm and emission -590 nm).

2.10. Stability studies

The stability of the optimized lyophilized formulation was done in type I sealed vials at refrigerated conditions of 2–8°C. Particle size, PDI, zeta potential, and drug release of drugs at 24 h were evaluated as stability-indicating parameters at pre-decided time intervals of 0, 15, 30, and 60 days.

2.11. Statistical analysis

All statistical analysis was performed by ANOVA analysis (two-way ANOVA) using Microsoft Excel^® [45].

3. Results

3.1. Optimization of the ratio of PXL: IMA

Cytotoxicity analysis of different drug ratios on MCF-7 cells is given in Figure 1A. For PXL to IMA ratios of (a) 1:0, (b) 0.75:0.25, (c) 0.5:0.5, (d) 0.25:0.75 and (e) 0:1; IC-50 values were found to be 2.6930, 0.7735, 3.2550, 3.4420 and 3.9380 μg/ml, respectively. The lowest IC₅₀ value was observed in the PXL: IMA ratio of 0.75:0.25.

Figure 1.

Cytotoxicity analysis of different drug ratios of paclitaxel and imatinib (PXL & IMA) on MCF-7 cells. (A) MTT analysis of different drug dose combinations. (B) Combination index analysis.

IMA: Imatinib; PXL: Paclitaxel; P:I: Ratio of paclitaxel and imatinib.

3.2. Combination index

The combination index of formulation at different drug ratios is given in Figure 1B. The lowest combination index was observed for PXL to IMA ratio of 0.75:0.25 (Table 1). On increasing the ratios of IMA, an increase in combination index was observed. These results conform with the optimization of ratio results.

Table 1.

Combination index analysis for different drug dose combinations.

	Free drugs		PXL : IMA :: 0.75:0.25			PXL : IMA :: 0.5:0.5			PXL : IMA :: 0.25:0.75
IC	PXL	IMA	PXL (mcg/ml)	IMA (mcg/ml)	IC	PXL (mcg/ml)	IMA (mcg/ml)	IC	PXL (mcg/ml)	IMA (mcg/ml)	IC
25	1.415	2.64	0.086	0.029	0.053	0.357	0.357	0.388	0.406	1.218	1.015
50	2.693	3.938	0.580	0.193	0.219	1.628	1.628	1.018	0.861	2.582	1.177
75	3.612	5.421	2.111	0.704	0.584	7.073	7.073	3.263	5.539	16.616	5.622

IC: Inhibitory concentration; IMA: Imatinib; PXL: Paclitaxel.

3.3. Optimization of drugs to be added to obtain drugs in the desired ratio

The entrapment efficiency of PXL was around 92.54%, whereas IMA was 75.12%. Using these values, the amount of the drug to be added for loading the drugs in the desired ratio (0.75:0.25 or 3.75 mg PXL and 1.25 mg IMA) was calculated. Amounts to be added were found to be 4.05 and 1.66 mg for PXL and IMA, respectively. This batch was performed in triplicate to calculate the actual amounts entrapped and was found to be 93.14% ± 2.14% (3.77 mg ± 0.09 mg) and 74.15% ± 1.74% (1.23 mg ± 0.03 mg), respectively, for PXL and IMA.

3.4. Quality-by-design approach

3.4.1. Quality target product profile & critical quality attributes

QTPPs identified are listed in Supplementary Table S1. QTPPs were identified based on knowledge space using preliminary knowledge, literature research, and preliminary lab trials. CQAs identified were optimum drug ratio (having maximum synergy), particle size (less than 220 nm so that sterilization can be done by membrane filtration), zeta potential (in the range of 20–30 mV, so as to obtain maximum stability), entrapment efficiency (maximum), drug release (sustained and controlled), stability (long term), IC₅₀ value (low) and cellular uptake (maximum).

3.4.2. Ishikawa fishbone diagram

The Ishikawa fishbone diagram prepared is given in Supplementary Figure S1. Ishikawa fishbone also known as the cause-effect analysis diagram having six arms gives all the parameters that can affect CQAs [46].

3.4.3. Failure mode & effects analysis based on risk priority number ranking

FMEA analysis based on the RPN scoring method is given in Table 2. Values having an RPN score of more than 20 were identified as critical factors [28].

Table 2.

Failure mode & effects analysis using risk priority number scoring.

Factors		Severity	Occurrence	Detectability	RPN
Drug
PXL	Dose	4	2	1	8
	Log P	4	1	1	4
	pKa	3	1	1	3
	Molecular weight	3	1	1	3
	Hydrophilicity/hydrophobicity	3	1	1	3
IMA	Dose	4	2	1	8
	Log P	4	1	1	4
	pKa	3	1	1	3
	Molecular weight	3	1	1	3
	Hydrophilicity/hydrophobicity	3	1	1	3
Materials
PLGA	Amount	4	4	2	32
	Molecular weight	4	1	3	12
	Grade	4	1	3	12
	Lactic acid: glycolic acid ratio	4	1	3	12
	Viscosity grade	3	1	3	9
	Purity	3	1	1	3
Protamine sulphate	Molecular weight	4	2	2	16
	Salt form	4	1	2	8
	Grade	3	2	2	12
	Amount (Conc of protamine solution)	5	3	2	30
Instrumental parameters
Magnetic stirrer	Make	1	2	1	2
	RPM	3	2	2	12
	Temperature	2	1	1	2
Ultracentrifugation	Make	1	2	1	2
	RPM	3	2	1	6
	Temperature	2	2	1	4
	Time	4	2	1	8
Process parameters
Preparation technique	Solvent evaporation	4	1	1	4
	Precipitation	4	1	1	4
Organic phase	Solvent	3	2	1	6
	Amount	3	2	2	12
Aq. Phase	Poloxamer type	3	2	1	6
	Concentration of poloxamer used	3	2	1	6
	Amount	4	2	1	8
Size reduction method
Probe sonication	Number of cycles	5	2	2	20
	Amplitude	5	3	2	30
	Pulse	4	2	2	16
	Duration of each cycle	4	2	2	16
Miscellaneous
Environmental Factors	Temperature	2	4	1	8
	Humidity	3	4	1	12

Severity: The nature of product severity causing harm. Severity scale (1–5) with 1 being not noticed by a customer and 5 being hazardous or life-threatening and could place the product survival at risk.

Occurrence: The quality of the product based on analysis or tests.

Occurrence Scale (1–5) with 1 being highly unlikely and 5 being almost certain.

Detectability: Standard operating procedures, or those procedures that have been proposed.

Detection Scale (1–5) with 1 being almost certain to detect and 5 is almost impossible.

Severity in FMEA analyzes a failure mode's impact on a system or process, crucial in pharmaceuticals for patient safety and treatment efficacy. Severity ratings on a scale of 1–5:

1: Minimal Impact – Small dose differences, no safety or treatment impact.

2: Minor Effects – Changes affect medication efficacy, minor adjustments may be needed.

3: Moderate Effect – Adjustments impact therapy, and may require treatment changes.

4: Significant Effect – Dose variations compromise safety and effectiveness.

5: Critical Impact – Extreme deviations risk the patient's life, requiring immediate action.

Severity ratings guide for mitigation efforts. Protamine sulphate's molecular weight significantly affects its RPN score, indicating its impact on failure modes. Changes may affect efficacy, safety or stability, leading to severe consequences. Regulating molecular weight is crucial for product quality and patient safety.

Aq.: Aqueous; FMEA: Failure mode & effects analysis; IMA: Imatinib; PLGA: Poly(lactic-co-glycolic acid); PXL: Paclitaxel; RPM: Revolutions per minute.

3.4.4. Design of experiments

Responses for batches planned as per design of experiments methodology are summarized in Table 3. Mean particle size varied between 145.76 and 242.17 nm whereas zeta potential varied between 6.5 and 44.2 mV. For the Design of Experiments (DoE), ANOVA perturbation, 3D response surfaces, and overlay plots along with desirability analysis are given in Figure 2. The results of both responses were significant as per ANOVA analysis and hence were further used to investigate the design space. Using overlay analysis along with desirability analysis, optimized batch and required values of independent variables were selected.

Table 3.

Design space for the design of experiments approach.

Independent variables			Dependent variables
Factor 1	Factor 2	Factor 3	Response 1	Response 2	PDI
A: Amplitude of probe sonication	B: Amount of PLGA	C: Concentration of protamine solution	Mean particle size (n = 3)	Zeta potential (n = 3)
%	mg	%	nm	mV
40	120	5	205.74 ± 2.54	35.2 ± 1.2	0.174
25	90	5	185.14 ± 4.10	27.5 ± 1.6	0.156
10	90	7.5	242.17 ± 8.14	36.8 ± 2.1	0.142
25	90	5	197.24 ± 6.01	26.8 ± 0.8	0.117
25	120	7.5	225.14 ± 10.02	38.7 ± 0.6	0.145
10	120	5	224.12 ± 1.66	19.4 ± 1.0	0.154
40	90	7.5	197.45 ± 4.12	44.2 ± 1.4	0.114
25	90	5	184.62 ± 12.11	28.4 ± 0.9	0.241
10	90	2.5	172.14 ± 16.42	6.5 ± 1.4	0.141
25	60	7.5	205.11 ± 7.55	40.7 ± 2.1	0.104
40	60	5	191.45 ± 5.14	37.6 ± 0.4	0.124
25	90	5	189.1 ± 0.47	27.1 ± 0.7	0.236
25	90	5	193.7 ± 8.14	30.1 ± 0.9	0.124
25	60	2.5	158.6 ± 6.62	13.4 ± 0.4	0.174
40	90	2.5	145.78 ± 5.04	17.6 ± 1.2	0.220

PDI: Polydispersity index.

Figure 2.

Design of experiment responses (A1–A5). Responses for mean particle size; (B1–B5) responses for zeta potential; (C1–C3) overlay plots for optimization and (D1–D3) desirability plots.

3.5. Transmission electron microscopy

Transmission electron microscopy images of coated and uncoated PLGA nanoparticles are given in Figure 3A & B, respectively.

Figure 3.

Characterization of nanoparticles. (A) Transmission electron microscopy images of drug-loaded PLGA nanoparticles. (B) Transmission electron microscopy images of drug-loaded protamine-coated PLGA nanoparticle. (C) In-vitro drug release of paclitaxel drug solution (PXL DS), paclitaxel PLGA nanoparticle (PXL PLGA), and protamine-coated paclitaxel PLGA nanoparticle (PXL PLGA PT) (n = 3). (D) In-vitro drug release of imatinib mesylate drug solution (IMA DS), imatinib mesylate PLGA nanoparticle (IMA PLGA), and protamine-coated imatinib mesylate PLGA nanoparticle (IMA PLGA PT) (n = 3). (E) FTIR scanning for PXL, IMA, PLGA, protamine, PLGA nanoparticles, and protamine-coated PLGA nanoparticles.

CDR: Cumulative drug release; PXL DS: Paclitaxel drug solution; PXL PLGA: Paclitaxel PLGA nanoparticle; PXL PLGA PT: Protamine-coated paclitaxel; PLGA nanoparticle.

3.6. In-vitro drug release

In-vitro drug release from drug solution, PLGA nanoparticles, and protamine-coated nanoparticles for PXL and IMA separately are given in Figure 3C & D, respectively. Formulation of PLGA nanoparticles led to drug release getting sustained from 2 to 24 h. For both the drugs PXL and IMA drug-release mechanism changed from first-order to zero-order kinetics when nanoparticles were coated with protamine. Korsmeyer-Peppas's modeling showed the drug release of both drugs followed a non-Fickian zero-order drug release. Supplementary Table S2 gives kinetic modeling data for all the drugs.

3.7. FTIR

FTIR scanning analysis of drugs, PLGA, and formulations is given in Figure 3E. Characteristics peaks of drugs and polymer were identified in both PLGA nanoparticles and protamine-coated PLGA nanoparticles, confirming them in the final formulation.

3.8. In-vitro cell line studies

3.8.1. Cytotoxicity

IC₅₀ values for pure drugs PXL and IMA were 2.693 and 3.938 mcg/ml, respectively. These values were reduced to 0.6871 and 1.9710 mcg/ml when drugs were loaded in PLGA nanoparticles. The obtained IC₅₀ value of 0.09891 mcg/ml was because of synergism on co-loading into nanoparticles. With protamine coating, there was a fivefold reduction in the IC₅₀ value of PXL-loaded PLGA nanoparticles. Protamine-coated PLGA nanoparticles loaded with IMA had an IC₅₀ value of 0.3545. Co-loaded protamine-coated PLGA nanoparticles had the lowest IC₅₀ value of 0.03824 mcg/ml showing that it had maximum cytotoxicity. Results are summarized in Figure 4A.

Figure 4.

Cytotoxicity and cellular uptake studies. (A) IC₅₀ values for MTT analysis. (B) Cellular uptake analysis.

MTT: 3-[4,5-dimethylthiazol-2-yl]-2,5 (diphenyl tetrazolium bromide) assay; RFU: Relative fluorescence unit.

3.8.2. Cellular uptake

Figure 4B gives the cellular uptake of pure drugs, co-loaded PLGA nanoparticles, and coloaded protamine-coated PLGA nanoparticles. There was a significant increase in cellular uptake of drugs when loaded in PLGA nanoparticles, and a further increase in cellular uptake when coated with protamine.

3.9. Stability studies

No significant change in stability indication parameters (mean particle size, PDI, zeta potential, PXL & IMA drug release in 24 h) was observed. Very small variations (nonsignificant) were observed in the parameters with a time of up to 60 days. Supplementary Table S3 gives the results of stability studies at different time intervals. Excellent stability was attributed to a well-executed lyophilization technique, ensuring the preservation of the structural integrity of all freeze-dried nanoparticle compositions without any collapse.

4. Discussion

4.1. Optimization of the ratio of PXL: IMA

Drugs in ratio 0.75:0.25 showed the lowest IC₅₀ value, because of maximum inhibition shown at the ratio. This led us to proceed with a PXL:IMA ratio of 0.75:0.25 to be loaded in the final formulation.

4.2. Combination index

Combination index of less than 1 (CI <1) means synergism between the drugs at that concentration [9,10]. The lowest CI was observed for the PXL: IMA ratio of 0.75:0.25, showing maximum synergism at the concentration. Henceforward this concentration was finalized.

4.3. Optimization of the amounts of drugs to be added to obtain drugs in the desired ratio

For the formulation of co-loaded nanoparticles, if 4.05 mg of PXL and 1.66 mg of IMA are added, the amounts of drugs loaded are 3.77 and 1.23 mg, respectively. The amounts of drugs loaded are in ratio (0.753:0.247) very close to the desired ratio of 0.75:0.25. Hence these amounts were used in the batches ahead.

4.4. Quality-by-design approach

4.4.1. Quality target product profile & critical quality attributes

QTPPs and CQAs were identified and targets were set for the formulation [30,47].

4.4.2. Failure mode & effects analysis based on risk priority number ranking

The amount of PLGA, concentration of protamine solution, and amplitude of probe sonication were identified as the critical factors to be utilized using the design of experiments methodology.

4.4.3. Design of experiments

4.4.3.1. Mean particle size

ANOVA analysis of the dependent variable mean particle size showed an F-value of 36.28 and a p-value of 0.0005, showing that the model is significant [48]. Factors A and C had both linear and quadratic effects significantly contributing to mean particle size, whereas for factor B only the quadratic effect was significant. No significant interaction effect was observed for mean particle size. The highest F-value of 176.71 was obtained for factor C showing that it has the maximum effect. This can also be seen in the perturbation graph given in Figure 2 (A2). A strong R² value of 0.985 was observed showing that the observed and predicted values are close and the model is good for exploration [30,47,49]. The quadratic equation obtained for predicting the responses obtained is:

$$\begin{array}{cc}\text{MPS}=& 289.643-5.410\times \text{A}+30.991\times \text{C}\\ & +0.069\times {\text{A}}^2+0.010\times {\text{B}}^2-2.57\times {\text{C}}^2\end{array}$$ (5)

Effects of independent variables on dependent variables can be visualized on the dependent variable mean particle size in Figure 2 (A3–A5). Linear effects of variables A and C and quadratic effects of A, B, and C can be seen in the 3D response surfaced plots.

4.4.3.2. Zeta potential

Zeta potential response was also significant as an F value of 70.35 was obtained for the model in the ANOVA analysis. Factors A, B, C, and C² were significant and had a significant effect on the zeta potential because their p-value was less than 0.05. Factor C (protamine concentration) had a very high F-value of 409.64, showing a high linear effect on the zeta potential. This can be attributed to a net positive charge on the protamine, contributing to the positive effect on the zeta potential. A comparison of various factors can be seen in the perturbation plot where linear and quadratic effects of C can be seen. 3D response surfaces were also plotted shown in Figure 2 (B3 to B5). R² value of 0.978 was obtained and a difference between the adjusted and predicted R² value of less than 0.2 was obtained indicating that the model can be explored [28,30,50]. The obtained quadratic equation for zeta potential response was:

$$\begin{array}{cc}\text{Zeta Potential}=& 11.982-0.508\times \text{A}-0.438\times \text{B}\\ & +11.054\times \text{C}-0.640{\text{ C}}^2\end{array}$$ (6)

4.4.3.3. Overlay & desirability

Overlay and desirability analysis were performed using the principles of response surface methodology. Both the responses were overlayed and an area of interest was highlighted as shown in Figure 2 (C1 to C3). Further desirability plot was plotted shown in Figure 2 (D1 to D3). The batch having the highest desirability was predicted by software based on statistical analysis. A batch was predicted to have a mean particle size of 168.748 nm and a zeta potential of 25 mV, which can be obtained by using parameters as A (amplitude of probe sonication) -34, B (amount of PLGA) -82.1 mg, and C (concentration of protamine solution) -3.82%. On performing the same, the actual results obtained were 167.3 ± 2.45 nm mean particle size and 24.7 ± 0.8 mV ZP. This batch was identified as an optimized batch which was used for further investigation.

4.5. Transmission electron microscopy

PLGA nanoparticles can be visualized in the transmission electron microscopic images. They were visualized to have spherical morphology. Protamine coating is distinctly visible in the images of coated nanoparticles.

4.6. In-vitro drug release

Formulation of PLGA nanoparticles sustained the drug release of both drugs from 2 to 24 h. Similar results of sustained release by formulation of PLGA nanoparticles have been reported earlier in multiple studies [22,51,52]. Protamine coating reduced the burst release of both drugs. For PXL, 52.48% drug was released in 4 h from uncoated PLGA nanoparticles, whereas from coated nanoparticles only 16.45% drug was released. IMA values for the same were 41.96 and 14.72% from PLGA and coated PLGA nanoparticles respectively. This reduction in the amount of drug released in the first 4 h shows that there is suppression of burst release of both drugs. Similar results have also been reported [53–55]. For PXL there was a change in drug-release kinetics mechanism from first-order kinetics to zero-order kinetics. Similarly, for IMA mechanism changed from first-order to zero-order drug-release kinetics. Korsmeyer-Peppas modeling showed a non-Fickian zero order drug release for both the formulations from protamine-coated formulations. Co-loading of PXL and IMA in protamine-coated PLGA nanoparticles led to a controlled zero-order sustained drug release of both drugs.

4.7. FTIR

Characteristics peaks of PXL and IMA were present in the FTIR scanning of PLGA nanoparticles and protamine-coated PLGA nanoparticles. The peak of PLGA was also clearly visible in the scanning of the formulations. The presence of characteristic peaks confirms the presence of drugs and PLGA in the final formulation.

4.8. Cell line studies

4.8.1. MTT cytotoxicity

MTT cytotoxicity analysis showed a significant increase in cytotoxicity of the formulation as compared with the free drugs. IC₅₀ values of PXL and IMA were 2.693 and 3.938 mcg/ml, respectively. There was a significant reduction in IC₅₀ value for co-loaded PLGA nanoparticles (0.09891 mcg/ml). Similar results have been reported earlier in many findings [20,54]. Further protamine coating led to a three-time reduction in MTT cytotoxicity analysis, similar to that reported in earlier studies [53,55].

4.8.2. Cellular uptake

Loading of drugs in PLGA nanoparticles leads to approximately 24-fold increase in cellular uptake which explains the reason for the decrease in IC-50 value after the formulation of PLGA nanoparticles [53]. Further, there was an increase in cellular uptake because of protamine coating since it is known to increase cellular uptake of the nanocarriers [53,56].

4.9. Stability studies

Stability studies data showed stability of the formulation at temperatures of 2–8°C for a time interval of 60 days. This was attributed to the application of trehalose used as a cryoprotectant as it is known to maintain particle integrity after lyophilization [57]. Freeze-drying is a crucial technique for preserving the stability of PLGA-NPs by preventing their degradation by hydrolysis [58]. The inclusion of trehalose in PLGA nanoparticles enabled the effective regulation of residual moisture [59]. This was attributed to the fluid-like behavior of the frozen mass, rather than it being solid [60], which eventually provided improved mechanical protection for the nanoparticles thereby improving the stability [61] of the formulations.

5. Conclusion

PXL and IMA in the ratio of 4:1 were found to have a maximum synergistic effect. The combination index at this ratio was the lowest indicating the maximum efficacy. Entrapment efficiencies obtained for PXL and IMA were 92.54 and 75.12%, respectively. QbD-based approach coupled with DoE methodology was successfully implemented to obtain an optimized batch having mean particle size and zeta potential of 167.3 ± 2.45 nm and 24.7 ± 0.8 mV, respectively.

TEM imaging showed PLGA nanoparticles of spherical morphology and protamine coating was clearly visible. A sustained and controlled zero-order drug release was obtained for both drugs. IC₅₀ value for final formulation was found to be 0.038 mcg/ml. An increase in cellular uptake of 30-fold was obtained as compared with the individual drug alone. Formulated formulation was stable for a period of 60 days at refrigerated conditions (2–8°C).

The formulation had promising results against the MCF-7 cell lines and the robust systematic optimization methodology developed shows the translatory potential of the formulation subject to further studies and clinical trials.

Supplemental material

Supplemental data for this article can be accessed at https://doi.org/10.1080/17435889.2024.2353557

Financial disclosure

The authors have no financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Competing interests disclosure

The authors have no competing interests or relevant affiliations with any organization or entity with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Writing disclosure

No writing assistance was utilized in the production of this manuscript.