1 Introduction

Age related macular degeneration and retinitis pigmentosa are one of the major causes for blindness all around the world. It is a slow and progressive condition assisted by activated microglia by aggravating internal inflammation. These activated microglia release chemically active molecules that cause cell damage within the outer retina (Lezzi et al. 2012). A viable treatment for this condition is nearly non-existent. Steroids, like prednisolone, have been used as anti-inflammatory agents. Unfortunately these drugs in present dosage lead to severe side effects. Use of biodegradable polymers for encapsulating the drug will not only help in a sustained release but also help in reducing the severe side effects.

PLGA was selected as the carrier since it is one of the most widely used polymer for drug delivery applications (Danhier et al. 2012). The degradation time can be well controlled by varying the copolymer ratio and molecular weight (Vert et al. 1994; Prokop and Davidson 2008).

The drug prednisolone used in the study, is classified as a synthetic glucocorticoid, used in the treatment of several inflammatory and auto-immune conditions. Prolong use is known to cause depression, insomnia, mood swings and memory loss. Thus optimized nanoparticles of PLGA encapsulating the drug prednisolone (Pred-PLGA-NPs) could significantly help in reducing the side effects. Quality by Design is recognized and implemented by regulatory agencies worldwide. The objective of the study was to develop Pred-PLGA-NPs and analyze the effect of the important factors on the responses like size and drug loading. A 23 factorial design was adopted to formulate the Pred-PLGA-NPs.

2 Materials and Methods

2.1 Materials

PLGA 50:50 (100,000–120,000) was purchased from Durect Corporation AL, USA. Prednisolone, Polyvinyl alcohol and chloroform were purchased from Sigma-Aldrich.

2.2 Experiment Design

A 23 factorial design method was employed for the formulations. The variables considered for the study, along with the levels are shown in Table 1. Design expert software (version 9) was employed to design the formulations.

Table 1 The independent and dependent variables with levels
Full size table

2.3 Preparation of Pred-PLGA-NPs Using Solvent Evaporation Technique

Predetermined quantities of prednisolone and PLGA were dissolved in 1.5 ml of chloroform. An oil-in-water emulsion was formed by emulsifying the polymer solution in 15 ml of 1–3 % w/v aqueous PVA solution using a probe sonicator for 2–4 min over an ice bath. The resultant emulsion was then stirred for 18 h at ambient conditions to remove chloroform. Nanoparticles were recovered by centrifugation (15,000 rpm) and washed three times with deionized water to remove excess PVA, followed by lyophilization prior to storage. The formulation composition is summarized in Table 2.

Table 2 Formulations and their quantities
Full size table

2.4 Measurement of Size and Zeta Potential

Average particle size (z-average), zeta potential of the developed NPs was determined by laser dynamic light scattering using Malvern Zetasizer (Malvern, Worcestershire, UK).

2.5 HPLC Conditions

HPLC was performed using (SHIMADZU Model SPD 20A) equipped with a binary solvent delivery pump. The chromatographic separation was performed on a reverse phase Eclipse plus C18 column, (150 × 4.6 mm, 3.5 μ) and the absorbance was monitored at 254 nm. The system was run with a mobile phase consisting of water +0.1 % TFA: acetonitrile +0.1 % TFA in the ratio of 60:40 (v/v) and was delivered at a flow rate of 0.5 mL/min.

3 Results and Discussion

A set of 8 runs were conducted with 3 independent and 3 dependent variables. All formulations underwent characterization for average particle size, drug loading and zeta potential. The effect of factors on the responses was studied (Table 3) and their respective contour plots were developed.

Table 3 Response data for the complete set of runs
Full size table

3.1 Effect on Zeta Potential

The zeta potential for nanoparticles helps in assessing the stability of the colloidal systems. The analysis found no significant factors but some unaccounted lurking variable may have influenced the response during the experiment. The values ranged from −2.2 to −14 mV. Figure 1 shows the zeta potential and size distribution graphs for the drug loaded PLGA NPs (P5).

Fig. 1
figure 1

Zeta potential distribution and size distribution graph of drug loaded PLGA NPs (P5)

Full size image

3.2 Effect on z-Average

The factor Drug/polymer ratio (30.35 %) and Sonication time (33.98 %) were found to have significant effect on the size of the NPs. There were no interaction effects. z-Average of developed NPs were in the range of 347 d nm (P5)–2354 d nm (P3) for different variable combinations (Fig. 2).

Fig. 2
figure 2

Contour plot showing effect of D/P ratio and PVA on size of the particle

Full size image

The model with an F-value of 97.34 was found to be significant (p < 0.0001) along with the R-square values, which were in reasonable agreement (Predicted: 0.9345, Adjusted: 0.9857). The observed adequate precision ratio was 27.311, which is well above the minimum prescribed ratio of 4, resulting in a satisfactory signal-to-noise ratio. Thus, this model could be used to navigate the design space.

$$\begin{aligned} {\text{z-average}} & = 4467 - 7710*{\text{A}} + 429.58333*{\text{B}} - 1180.25*{\text{C}} \\ & \quad - 764.16667*{\text{A}}*{\text{B}} + 2267.5*{\text{A}}*{\text{C}} \\ \end{aligned}$$
(1)

The analysis revealed that the Drug/polymer ratio and Sonication time had a significant but negative effect on the z-average of the polymeric NPs. During emulsification, an increase in drug/polymer ratio may lead to an optimum amount of PLGA being utilized for the encapsulation of the drug (Feczkó et al. 2011; Seju et al. 2011), thus helping in reducing the size of the NPs. The particle size also reduced with the increase in sonication time as finer droplets were being formed by the prolonged and concentrated supply of energy. The factor PVA, showed a slight positive effect as compared to the other factors.

3.3 Effect on Drug Loading

The concentration of the drug loaded (µg drug/mg of NP) was found in the range of 5.8 (P5)–26.892 (P8). The model with an F-value of 26.38 was found to be significant (p < 0.0001) along with the R-square values, which were in reasonable agreement (Predicted: 0.7610, Adjusted: 0.9477). The observed adequate precision ratio was 12.761, which is well above the minimum prescribed ratio of 4, resulting in a satisfactory signal-to-noise ratio. Thus, this model could be used to navigate the design space.

$$\begin{aligned} {\text{Drug}}\,{\text{loading}} & = 20.06838 + 27.26750*{\text{A}} - 9.41179*{\text{B}} \\ & \quad - 5.14638*{\text{C}} + 8.64083*{\text{A}}*{\text{B}} + 2.34063*{\text{B}}*{\text{C}} \\ \end{aligned}$$
(2)

The factor Drug/polymer ratio was found to have the most significant effect (83.89 %) on drug loading. The other factors showed no significant effect on the response (Fig. 3).

Fig. 3
figure 3

Contour plot showing effect of D/P ratio and PVA on drug loading

Full size image

The drug loading increased with the increase in the drug to polymer ratio. The drug-polymer interaction and drug miscibility in the organic solution affects the percentage drug entrapment in the NPs (Panyam et al. 2004). An increase in concentration of the drug and polymer in the organic phase causes the viscosity of the organic phase to increase, leading to lesser drug movement into aqueous phase. This may have brought about an increase in the amount of drug inside the NPs (Song et al. 2008).

By applying the constraints (minimize size and maximize the drug loading) on the dependent factors, an optimum Pred-PLGA-NPs formulation was selected. Keeping desirability factor as a basis, Design Expert software was employed to predict the process parameters for the optimized NPs. These Pred-PLGA-NPs obtained by employing the optimized process parameters (0.5:1, 3 % and 4 min) were characterized for z-average (549.8 nm), zeta potential (−15.6 mV) and drug loading (21.169 µg/mg). The response values were in good agreement with the software-predicted values (568.5 nm, −9.950 mV and 25.930 µg/mg), thereby reaffirming the validity of the model.

4 Conclusion

To determine the effects of process variables, a 3-factorial 2-level design and analysis was been carried out. Drug/polymer ratio and sonication time had the most significant effect on the size of the NPs. Maximum levels of Drug/polymer ratio and sonication time showed to produce the smallest particle size. Drug/polymer ratio alone was found to have the most significant and also the largest positive effect on drug loading. The study also found that none of the selected process variables had any significant effect on the zeta potential of the particles.