Optimization of apigenin nanoparticles prepared by planetary ball milling: In vitro and in vivo studies

: This study intended to optimize apigenin ( APG ) nanoparticle formulation prepared by planetary ball milling to enhance its dissolution rate and bioavailability using a design of experiment ( DoE ) . In this study, polyvinyl pyrroli - done ( PVP K30 ) was used as a nanoparticle stabilizer. The independent parameters of milling speed, milling ball size, and drug to solvent ratio were evaluated for their impacts on APG nanoparticles concerning the nanoparticle size ( Y 1 ) , zeta potential ( Y 2 ) , and drug dissolution e ﬃ ciency after 60 min, notated as % DE 60 ( Y 3 ) . The millingball size showed a signi ﬁ cant antagonistic e ﬀ ect ( P = 0.0210 ) on the size of APG nanoparticles, while milling speed had an agonistic e ﬀ ect on the zeta potential values of drug nanoparticles, ranging from low to medium speed levels. In addition, ANOVA analysis indicated that the e ﬀ ect of the drug - to - sol - vent ratio on the % DE 60 of APG from the nanoparticle for - mulations was antagonistically signi ﬁ cant ( P = 0.015 ) , and the quadratic e ﬀ ect of milling speed ( AA ) also had a signi ﬁ - cant antagonistic e ﬀ ect ( P = 0.025 ) on the % DE 60 . Risk assessment analytical tools revealed that milling ball size and milling speed signi ﬁ cantly a ﬀ ect the nanoparticle size. The drug/solvent ratio exerted a strong impact on % DE 60 . Furthermore, the maximum plasma concentration ( C max ) of the optimized APG nanoparticle formula increased by four folds.Inaddition,AUC 0 – t ( ng · mL − 1 · h − 1 ) forAPGnanoparticle ( 353.7 ± 185.3 ng · mL − 1 · h − 1 ) was higher than that noticed in the case of the untreated drug ( 149 ± 137.5 ng · mL − 1 · h − 1 ) by more than two folds.


Introduction
Formulation of poorly water-soluble drugs in nanoparticle forms has gained great attention from researchers to enhance their dissolution properties, and, in turn, their bioavailability [1].As per the Noyes-Whitney dissolution theory, there is an indirect relation between particles size and dissolution rate of poorly soluble drugs.Reducing the particle size of the active pharmaceutical ingredient from micronized to a nanosized particles could significantly enhance its dissolution rate [2].This reduction in particle size leads to increase in the surface area, which results in increase in the drug dissolution rate [2].
Nanoparticles could be prepared by two different techniques, by reducing the size of large crystals (topdown techniques) or by increasing the particle size from small to large size by precipitation of dissolved molecules (bottom-up techniques).Ball milling is one of the topdown nanonization techniques, that is based on mixing the drug powder with dispersion media (mostly water) and a suitable stabilizer that helps avoid particles aggregation after the preparation of nanoparticles [3].
Nanoparticles' properties prepared by size reduction using ball milling technologies are affected by several parameters such as the amount of drug, type, and concentration of stabilizers, the amount and size of the ball, speed, time, and temperature [4,5].Many drugs have been nanonized using this method for the production of drug nanoparticles like fenofibrate, naproxen, and ibuprofen [6].Some of these drugs are available in the pharmaceutical markets in tablet forms, such as fenofibrate, ibuprofen, and paliperidone palmitate [5].The addition of a stabilizer to the formulation can play a crucial role in nanoparticle physical stability by reducing the free energy of the system during the nanonization process.The stabilizers used may be either (i) surfactants such as Tween 80 and poloxamer 188, or (ii) polymers such as hydroxypropyl methylcellulose (HPMC) and PVP [7].
Particle size reduction using a planetary ball mill mainly depends on the centrifugal acceleration force rather than the gravitational acceleration force.Thus, the grinding bowls and material are subjected to centrifugal forces in which the change in direction and intensity occur constantly resulting in efficient and fast grinding processes.Several studies that have revealed the efficiency of planetary ball mill in the enhancement of drug dissolution rate and bioavailability were investigated [8].
Pharmaceutical quality by design (QbD) is a systematic approach applied to ensure the quality of the pharmaceutical product according to the compendia guidelines.This could be achieved by studying the effect of the input parameters (critical material attributes; CMA and critical process parameters; CPP) on the critical quality attributes (CQA) of the final products by applying the quality risk management that can affect the quality of the final product.QbD can present several manufacturing advantages such as minimized batch failure, and more efficient and effective control of change, in addition to providing opportunities for more compliant regulatory approaches [9].
Design of experiments (DoE) is a cost-effective and systematic tool to justify the effect of CMA and CPP on the product quality attributes by reducing the number of experiments [4].Moreover, DoE can detect possible interactions between these parameters over a wide range of values without studying all other possible values directly [6].
CQA is defined according to International Conference on Harmonization (ICH) Q8 as a "physical, chemical, biological, or microbiological property or characteristic that should be controlled and be within a suitable limit, range, or distributed to ensure the anticipated final product quality."Identification of CQAs is done through risk assessment as per the ICH guidance Q9 [4].The CQA of the nanoparticles for oral delivery is a small particle, having reasonable zeta potential value as well as increased physical stability.The risk parameters affecting the CQAs of nanoparticles should be evaluated as well.
Apigenin (APG) is a natural bioflavonoid compound present in several dietary plant foods such as vegetables and fruits.It has potential anti-inflammatory, antioxidant, and anticancer activities.Like several phytomedicine compounds, APG is a poorly water-soluble with high permeability compound, classified by the biopharmaceutical classification system as a class II drug [10], which results in its low oral bioavailability (7.06%) in the studied animal models [11].Many authors [12][13][14][15] adopted the theory of particle size reduction to enhance the solubility and the bioavailability of APG.APG nanoparticles have been prepared to utilize the liquid antisolvent precipitation method as described by Zhang et al. [12].This technique produces particles with near-spherical shapes in amorphous form.The solubility of APG nanoparticles was improved in comparison to the raw drug in artificial gastric juice and artificial intestinal juice.However, the preparation of APG nanoparticles by the milling technique and the statistical description of the effects of formulating and process parameters on nanoparticles' attributes had not been described previously.
The aim of the present work was to formulate and optimize APG nanoparticles using planetary ball milling by applying 3 3 full factorial designs.The impact of independent formulation and process parameters (ball size, milling speed, and drug to solvent ratio) on APG nanoparticles based on the CQA was investigated.In addition, the study was carried out based on the optimization procedures based on risk assessment tools by utilizing the Ishikawa fishbone diagram and preliminary hazard analysis (PHA) tool.

Materials
APG was obtained from Beijing Mesochem Technology Co. Pvt.Ltd, Beijing, China.PVP K30 M. Wt. 40,000 was obtained from Loba Chemie, India.Potassium dihydrogen orthophosphate was purchased from Winlab, Leicestershire, United Kingdom.Sodium phosphate dibasic was supplied from Sigma-Aldrich, Missouri, USA.Acetonitrile HPLC grade was obtained from Sigma Aldrich (St. Louis, MO, USA).Other chemicals used were of analytical grade and were used as received.

Design space
The screening was carried out on both CPPs and CMAs that have an influence on the CQAs of APG nanoparticles prepared by planetary ball milling by selecting the proper nanoparticle stabilizer.The effects of three stabilizers (Captisol, Pluronic F-127, and PVP K30) at 5% w/v concentration on APG nanoparticles were evaluated.Other milling conditions were kept constant by using the milling speed of 500 rpm, milling ball size of 0.5 mm, solid-to-solvent ratio of 0.1, and 3 milling cycles (each of 10 min with 5 min pause).

DoE
After defining the design space, DoE was applied to study the impact of selected independent factors (CMAs and CPPs) of high-risk potentials on CQAs of APG nanoparticles manufactured by the milling process.When these parameters were chosen, a design space was subjected to further optimization of the formulation.Designing series of experiments was carried out by applying a 3 3 full factorial design response surface methodology for the measurement and analysis of the nanoparticles CQAs.
Three factors, three levels (3³) full factorial design was used using a software program (Statgraphics Centurion Program Version 17.2.02.).The tested formulation and process independent parameters were the milling speed (A), milling ball size (B), and drug to the solvent ratio (C) (Table 1).The effect of these independent factors on nanoparticle size (Y1), zeta potential (Y2), initial dissolution rate within the first 5 min (IDR; Y3), and dissolution efficiency after 60 min (% DE 60 ; Y4) are shown in Table 1.In addition, the matrix of 3³ full factorial design for APG nanoparticle formulations is tabulated in Table 2.

Risk assessment
The PHA tool and Ishikawa fishbone diagram are the risk assessment tools that were applied to illustrate the impact of the studied CMA and CPP on the CQAs of nanoparticle formulations prepared by wet milling procedures.

Preparation of APG nanoparticles (milling procedures)
Zirconium balls with a size of 0.1, 0.5, and 1 mm, and drug-to-solvent ratios of 1 g:25 mL, 3 g:25 mL, and 5 g:25 mL (0.04, 0.12, and 0.2) were used.Also, different milling speeds of 200, 600, and 1,000 rpm were used.The number of cycles was 3, the milling time was 10 min, with a pause time of 5 min.The prepared APG nanosuspensions were then freeze-dried under vacuum pressure of less than 1 mbar to get solid nanoparticles (Alpha 1-4 LD Plus, Martin Christ Gefriertrocknugsanlagen GmbH, Osterode am Harz, Germany).The obtained nanoparticles were frozen and stored in tightly closed containers shielded from light pending further research.

Particle size analysis and zeta potential
The particle size and zeta potential for both nanosuspensions and freeze-dried nanoparticles were calculated with Malvern Zetasizer version 6.02.The prepared nanosuspension samples were diluted with deionized water to a suitable dilution before measuring.Determination of nanoparticle size, zeta potential, and polydispersity index (PDI) were carried out in triplicates.

APG content
The APG content determination in the prepared nanoparticles was carried out in triplicate.About 10 mg freeze-dried APG nanoparticles were dissolved in 10 mL of methanol, from which 1 mL was properly diluted with phosphate buffer, pH 6.8.The absorbance was then measured by using UV spectrophotometry at λ max 336 nm.The APG content in nanoparticles was calculated using Eq. 1 as follows:

%APG in nanoparticles
Amount of APG mg Nanoparticle weight APG PVP mg 100

In vitro dissolution
To assess the in vitro dissolution profile of APG from its nanoparticles, the USP-II dissolution apparatus (Pharma Test, DT 70, Germany) was used.The in vitro dissolution experiment was performed in 900 mL of 6.8 phosphate buffer as a dissolution medium at 37°C and 100 rpm.At pre-determined time intervals (5, 10, 15, 30, 45, and 60 min), 5 mL of the samples was withdrawn using a poroplast-kerze filter, diluted suitably, and the absorbance was measured spectrophotometrically at λ max 336 nm.Dissolution efficiency (DE%) of APG from nanoparticle formulations was calculated by using the trapezoidal rule by calculating the area under the dissolution curve at time (t).The data were expressed as a percentage of the area of the rectangle described by 100% dissolution at the same time [13].

Physicochemical characterization of optimizing
APG nanoparticles

X-ray diffraction analysis
Using X-Ray diffractometry, the crystallinity of the pure APG and lyophilized APG nanoparticles was calculated.The X-ray diffraction spectra were acquired for the powder sample using a RIGAKU diffractometer (Japan) fitted with curved monochromator graphite crystal, automatic divergence slit, and PW/1710 automatic controller.The target used was CuK-based radiation operating at 40 kV and 40 mA (almost = 1.5418Å).The patterns of diffraction were achieved using a continuous mode of scanning with 2°varying from 4°to 60°.

Transmission electron microscopy (TEM)
The optimized nanoparticle formulation was examined for external and internal structure using a TEM (FEI Tecnai G 2 20 TWIN, USA).

Pharmacokinetic (PK) studies
PK studies for the optimized APG nanoparticle formulation were carried out using male Wistar albino rats (weighing 200-220 g) taken from the college of pharmacy, Experimental animal care center (King Saud University, Riyadh, Saudi Arabia) as described previously [14].The experiments were carried out following an approved protocol (number: KSU-SE-19-66) from the Ethical Committee of the College of Pharmacy, King Saud University, Riyadh, Saudi Arabia.The animals were distributed into two groups (5 rats in each): Group I: rats administered with untreated APG, and Group II: rats administered with the optimized APG nanoparticle formula.The animals had fasted for 24 h before drug administration.The medication was administered orally in a suspension form by using oral gavage.The drug suspensions for each group were prepared with a concentration of 10 mg•mL −1 that was distributed homogenously in an aqueous solution of 0.5% sodium carboxymethyl cellulose (CMC-Na) before administration to the rats.Animals were given free access to food and water 3 h after each sample was orally administered.Blood samples of 0.5 mL were drawn from the retroorbital plexus vein into heparinized test tubes at time intervals of 1, 2, 4, 6, 12, and 24 h.Finally, plasma was isolated at 15,000 rpm by centrifuging the samples for 10 min, then held at −80°C pending analysis.APG was analyzed in rat plasma by using UPLC-MS/MS procedures [15].Plasma separation from blood and extraction of APG from plasma were performed using protein precipitation process.An aliquot of 100 μL of rat plasma was combined with 50 μL of prednisolone (internal standard IS) (200 μg•mL −1 ) and 750 μL of methanol.The collected mixture was vortexed for approximately 1.0 min.The samples underwent centrifugation for about 10 min at 15,000 rpm.After centrifugation, about 800 μL of the supernatant was taken and transferred to a sample vial.For the quantification of APG in plasma samples, about 5 μL of the sample was injected into the UPLC-MS/MS system.To analyze the plasma samples, a validated UPLC-MS/MS assay (UPLC, Waters Acquity, Milford, MA, USA) was used.The chromatographic separation of APG was performed using a BEH C18 column (50, 2.1, and 1.7 mm) with an acetonitrile mobile phase and 0.1% formic acid (35:65% v/v) run at a flow rate of 0.25 mL•min −1 .Tandem mass spectrometry was used to detect the eluted compounds using a TQ detector (Waters Corp., Milford, MA, USA) equipped with an electrospray ionization source that operates in positive ionization mode.The ionization pairs (m/z) were selected as follows: APG: 270.99-152.9(cone voltage 57 V, collision power 34 V) and prednisolone: 403.172-385.224(cone voltage 42 V, collision power 13 V).

Stability studies
Stability studies were carried out for APG nanoparticles according to ICH guidelines that were exposed to accelerated stability study at 40°C ± 2°C/75% residual humidity (RH) ± 5% RH for 6 months.

Statistical analysis
The PK data of six independent studies are presented as mean value ± SD.The data from in vitro and PK experiments were analyzed using one way ANOVA test.For this study, the program used was "Graphpad Instat Program (San Diego, CA, USA)" and was taken as statistically significant.

Experimental design (design space)
To screen the effect of CPPs and CMAs on the CQAs, three stabilizers in 5% concentration were used in the milling procedures (Captisol, Pluronic F-127, and PVP K30).Other milling procedures were kept constant (500 rpm as milling speed, milling ball size of 0.5 mm, solid-to-solvent ratio of 0.1, and 3 milling cycles each of 10 min with 5 min pause).The influences of the tested stabilizers on nanoparticles properties are shown in Table 3.It is clearly evident that milling of APG with 5% PVP resulted in the lowest nanoparticle sizes (277.70 ± 20.18 nm), with an acceptable PDI of 0.23, and a reasonable zeta potential value (−13.70 mV).In addition, 52.3 ± 0.45% of the drug was dissolved after 120 min, which is considered higher than the results obtained with other stabilizers.
Therefore, 5% PVP was selected as a stabilizing agent for APG nanoparticles.In addition, all three milling ball sizes (0.1, 0.5, and 1.0 mm) and three solid-to-solvent ratios were chosen to carry out the 3 3 full factorial designs.It is worth mentioning that milling time parameters (3 milling cycles, each of 10 min with 5 min pause) were kept constant during the whole experiment.

Effect of independent variable parameters on particle size
In the development of APG nanoparticles, the particles size is one of the most important parameters that should be optimized.The influence of independent parameters (milling speed; A, milling ball size; B, and drug-to-solvent ratio; C) on nanoparticles size are displayed in Table 4.
Ball size (B) showed a significant antagonist effect (P = 0.0210) on the particles size of the APG nanoparticles.Noticeable, but insignificant, effects on APG nanoparticles size were observed in the case of milling speed (A) and drug-to-solvent ratio (C), the calculated P values were 0.4538 and 0.8760, respectively, which are >0.05.Moreover, the milling speed quadratic effect (AA) revealed a highly significant agonist effect of APG nanoparticles size (P = 0.007).
Increasing the milling ball size resulted in reducing the APG nanoparticles size significantly (P = 0.021).In contrast, increasing the milling speed resulted in significantly reducing the nanoparticle size only at low values.In addition, milling speed and solid solvent ratio (AC) showed a slight interactive effect on APG nanoparticle sizes, but all these interactive effects are statistically insignificant.
Figure 1a shows the response surface plot illustrating the impact of speed and ball size on the particle size of APG at fixed concentrations of the drug/solvent ratio (0.12, i.e., 3 g APG:25 mL milling solvent).It is clear that the milling speed and ball size have the most effect on APG nanoparticle size.Increasing the milling speed resulted in reducing nanoparticles size at speed value up to  600 rpm, thereafter the effect is In addition, increasing the milling ball size resulted in reducing nanoparticle size at all milling speeds.As shown in Figure 2, the largest particle size was observed in the case of nanoparticle formula #8 and #13, in which the smallest ball size (0.1 mm) was used along with constant solid/solvent ratio (0.04) at milling speeds of 1,000 and 200 rpm, respectively.In contrast, the smallest nanoparticle sizes were recorded for formulae #15 and #25, in which milling ball sizes (0.5 and 0.1 mm) were used, respectively, with constant milling speed (600 rpm) and different solid/solvent ratios (0.2 and 0.04, respectively) in the milling procedures.
The relation between the size of the nanoparticle and the velocity of the milling could be explained based on the energy produced during rotation.Alshora et al. [8] revealed that low milling speed might reduce nanoparticle size due to the higher energy that causes particle impaction and breakage.Further increase in the milling speed may lead to particle agglomeration, mainly due to the breakage of the interarticular repulsive force caused by the stabilizer, resulting in an expansion in the nanoparticles' sizes [8].Increasing the solid content did not influence the size of nanoparticles significantly.In contrast, Ghosh and colleagues found that rising the solid content from 2% to 5% substantially reduced the particle size [16].

Effect of independent variable parameters on zeta potentials
The results of the impact of different independent factors (milling speed, milling ball size, and drug to solvent ratio) on the zeta potential values of APG nanoparticle formulations are displayed in Table 4.The quadratic effect of milling speed (AA) on zeta potential is significant (P = 0.0157).The other parameters milling ball size (B) and drug solvent ratio (C) had insignificant effects on the zeta potential of the APG.Also, the interactive effect (AC) showed a pronounced antagonistic effect on the nanoparticle's zeta potential, but this effect is slightly insignificant (P = 0.07).Also, neither of the studied individual factors (A, B, C) exhibited solely a significant effect on APG nanoparticle zeta potential values.
The response surface plot analyzing the impact of milling speed and solid/solvent ratio on APG nanoparticles zeta potential values at constant milling ball size (0.5 mm) is displayed in Figure 1b.Milling speed showed an agonistic effect on the drug nanoparticle zeta potential values from lower to medium speed levels.This effect was reversed to be antagonistic at higher milling speed, but these effects are insignificant (P = 0.44).Milling ball size and solid/solvent ratio did not exhibit any noticeable effect on nanoparticles' zeta potential values.
Figure 3 illustrates the zeta potential values of different APG nanoparticle formulations.The data showed that the highest zeta potential value (−21.9 mV) was recorded in the case of formula #8, in which a milling speed of 200 rpm, a milling ball size of 0.1 mm, and a solid/solvent ratio of 0.12 was applied as a milling condition.In contrast, the lowest zeta potential value (−10.6 mV) was observed in the case of nanoparticle formula #12, in which the milling conditions were: 200 rpm milling speed, 1 mm milling ball size, and 0.12 solid/solvent ratio, as shown in Tables 2 and 4. In a previous study [8], it was revealed that using a small ball size (0.1-0.5 mm) resulted in an increased zeta potential of the prepared nanoparticles significantly.They demonstrated that for rosuvastatin calcium nanoparticle formulations that were milled at 800 rpm using solid-to-solvent ratio of 0.625 with different ball sizes, the negative value of zeta potential decreased (from −28.5 to −22.9 mV) with increase in the ball size from 0.1 to 1 mm.

Effect of independent variables on APG % DE 60
The dissolution efficiency of APG from nanoparticle formulations within 60 min (% DE 60 ) describes both the rate and magnitude of drug dissolution from nanoparticle formulations and its profile within 60 min.The impacts of different independent factors on the % DE 60 of APG from the nanoparticle formulations are displayed in Figure 4.The analysis of variance for the effect of independent variables on % DE 60 of APG from the nanoparticle formulations (Table 4) revealed that the solid/solvent ratio exhibited a highly significant antagonistic effect on % DE 60 of APG from the nanoparticles formulations (P = 0.015) and the quadratic effect of milling speed (AA) also showed a significant antagonistic effect (P = 0.025) on % DE 60 .Moreover, milling speed and milling ball size showed an insignificant agonistic effect (P > 0.05).
The response surface plot estimating the influence of the independent variables, namely, milling speed and solid/solvent ratio on the % DE 60 of APG from the nanoparticle formulations at constant milling ball sizes (0.5 mm) is illustrated in Figure 1c.Increasing the solvent-to-drug ratio led to a pronounced slowing of drug dissolution from the nanoparticle formulations and, in turn, reduced its % DE 60 .In addition, the highest values of % DE 60 of the drug were observed for nanoparticles formulation #25, in which medium milling speed (600 rpm), smallest milling ball size (0.1 mm), and lowest solid-to-solvent ratio (0.04) were used concomitantly.This might be attributed to decreasing nanoparticle size by increasing milling speed along with minimizing solid/solvent ratio as explained previously.Figure 4 illustrates the % DE 60 of APG from different nanoparticle formulations.Amongst all the tested 27 nanoparticle formulations, the dissolution rate was found to be the highest from formula F25 (% DE 60 was 64.86 ± 7.52%).These formulations were milled by using 600 rpm as a milling speed, a milling ball size of 0.1 mm, and a very low solid/solvent ratio of 0.04.In contrast, the slowest drug dissolution was exhibited in the case of nanoparticle formulas F9 and F19 which showed % DE 60 values of 2.47 ± 1.13 and 1.12 ± 0.54, respectively.
In our previous study on rosuvastatin calcium nanoparticles prepared by planetary ball milling [8], an antagonistic interaction effect between solid content and the ball size on the drug dissolution efficiency was reported.However, increasing the ball size or solid content led to a decrease in the drug's initial dissolution rate.
Rezaei et al. [17] found that the dissolution rate of the nonsteroidal anti-inflammatory drug indomethacin nanoparticles was about four times higher than micronized drug mixture with PVP within 30 min in comparison to the corresponding drug-polymer physical mixture.The enhancement of drug dissolution rate from the prepared nanoparticle formulations might be attributed to reducing the particle size, and minimization of drug crystallinity due to the effect of hydrophilic polymers on increasing drug wettability.In addition, Liu and coworkers [18] combined the alteration of the crystal environment with a reduction in particle size to synthesize nanoparticles from celecoxib (CXB).Due to the reduction in particle size and modification of the crystal environment, CXB nanoparticles showed markedly improved dissolution rate and oral bioavailability in their research.Moreover, Kakran et al. [19] enhanced the in vitro dissolution rate of quercetin, a poorly water-soluble antioxidant, by manufacturing drug nanoparticles using PVP and pluronic F127 as nanoparticles' stabilizers.

PHA
The PHA tool (Table 5) and Ishikawa fishbone diagram (Figure 5) showed various crucial factors that can possibly impact the CQAs of nanoparticle formulations prepared by wet milling procedures.However, it is impossible practically to control or screen the impact of all the independent variables on the quality attributes of APG nanoparticles prepared by planetary ball wet milling.
Therefore, it is important to specifically clarify and study a large proportion of experimental variations by those variables, which are considered to have a strong or substantial effect on the product's quality attributes.Thus, the PHA method (Table 5) was used to conduct a risk assessment matrix showing various levels of risk related to these variables.
From the literature of published related research and the preliminary studies, it was obvious that stabilizer type and concentration have a high impact on CQAs of the nanoparticles produced by milling procedures, while particle size, particle shape, and crystallinity exhibited medium impact [20][21][22][23][24][25].
Moreover, milling time factors (milling cycle, number of milling cycles, and pause time) are considered as CMA and CPP that have a low impact on quality attributes and can be easily controlled.
In the present study, upon investigating the influence of the tested independent parameters the CQAs of APG nanoparticles, the risk assessment analytical tools (Table 5 and Figure 5) indicated that both milling speed and milling ball size exhibited high risk on the nanoparticle size, while solid/solvent ratio showed a medium effect on the studied response.Solid/solvent ratio and milling ball size exhibited high risk on nanoparticle zeta potential, while milling speed showed low risk.Moreover, the solid/solvent ratio exerted high risk on both IDR and % DE 60 , while milling speed and milling ball size showed medium risk in these responses.

Optimized formula of APG nanoparticles
The following desirability parameters were selected for the tested independent factors: minimum particle size, maximum zeta potential, and maximum DE 60 (Table 6).
Based on the modeling made by Statgraphics Centurion, version 17, and a desirability factor equal to 95%, the suggested conditions for optimum nanoparticles formulation are as follows: milling speed (A) = 666 rpm, milling ball size (B) = 0.1 mm, and drug/solvent ratio (C) = 0.04, which means 1 g of APG: 25 mL of water containing 5% PVP as the milling solvent.The observed CQA values were found close to the predicted optimized values for the nanoparticles' formula.The observed particle size was 252.7 ± 13.34 nm (the predicted particle size is 216.60 nm).Zeta potential observed value was −13.50 ± 1.61 compared to the predicted value (−16.44).In addition, the % DE observed value was 59.9 ± 0.91 (predicted is 56.08%).

TEM
The structure and morphology of the APG nanoparticle formula was studied using TEM analysis.The images obtained from TEM indicated and spherical nanoparticles (Figure 6).Apigenin nanoparticles prepared by planetary ball milling  11

X-ray diffraction analysis
The crystalline nature of the raw material, physical mixture, and APG nanoparticles were studied by X-ray diffractometer.The X-ray powder diffraction pattern untreated APG showed two distinctive peaks at 2θ degrees of 7.1, 11.2, 14.2, 15.0, and 15.9.These sharp diffraction peaks indicate the crystalline state of APG [15] (Figure 7).In contrast, PVP did not show distinct diffraction peaks in its spectrum.In the case of APG-PVP physical mixture, the drug diffraction peaks did not exhibit a noticeable change, but the diluting effect of the polymer did.A partial reduction in the intensities of the APG crystalline peaks was detected in the case of the optimized nanoparticle formula.This might be contributed to the effect of PVP in inhibiting the drug crystallization [26] in addition to the drug dispersion in the polymeric matrix of the stabilizer.

PK data analysis
The comparison of PK parameters of APG from the optimized nanoparticle formula with the untreated drug, calculated by using WinNonlin Software (Pharsight Co., Mountain View, CA, USA) program is listed in Table 7 and presented in Figure 8.The non-compartmental PK model was used for calculating different PK parameters including maximum plasma concentration (C max ) and maximum concentration time (T max ).APG nanoparticle formula showed plasma concentration significantly higher than untreated drug.The calculated C max of orally administered APG nanoparticle formulation was found to be about four folds higher than that observed in untreated APG.Also, T max was 1.1 h for the nanoparticles' formula compared to that of the untreated drug (6 h), indicating rapid absorption.Also, AUC 0−t (ng•mL −1 •h −1 ) for APG nanoparticle (353.7 ± 185.3 ng•mL −1 •h −1 ) was higher than that noticed in the case of the untreated drug (149 ± 137.5 ng•mL −1 •h −1 ) by more than two folds.

Effect of storage on the stability of APG nanoparticles
Table 8 illustrates the effect of storage in accelerated stability conditions (40°C ± 2°C/75% RH ± 5% RH) on particle size, % DE 60 , and zeta potential of APG optimized nanoparticle formulation during 6 months.The particle size for freshly prepared nanoparticles was 252.7 ± 13.34 nm with a zeta potential of −13.5 ± 1.61.After 3 months of accelerated stability, the particles size increased to 381.6 ± 42.82, but no further increase in particle size was noticed after 6 months.Zeta potential values during storage were in the range of −10.8 ± 4.04 and −13.6 ± 0.643 after 3 and 6 months, respectively.In addition, the drug exhibited an IDR of 89.9 ± 0.91 from the freshly prepared nanoparticles with a % DE 60 of 82.66 ± 1.88.However, the IDR of the drug from nanoparticle was reduced to 28.7 ± 0.57 and 28.0 ± 2.6 after storage for 3 and 6 months, respectively, with % DE 60 values of 56.96 ± 3.88 and 51.69 ± 2.2, respectively.Figure 9 shows the dissolution profiles of APG dissolution from the optimized nanoparticles formulation at 3 and 6 months compared to the dissolution of fresh APG and untreated powder.The results showed a decrease in the drug dissolution rate from nanoparticle formulation after 3 and 6 months, but still significantly higher than that of the untreated drug.The slight slowing of the drug dissolution rate from nanoparticle formulation after 3 and 6 months might be due to enlargement in particle size.
The nanonization of the drug particles to the nanosize range might result in increasing particles' surface area in comparison to microparticles or coarse particles, which affects nanoparticles' stability during storage [27].Physical instability problems like sedimentation, crystal growth, agglomeration, or change in crystallinity state are the most common problems with nanonization that should be limited or avoided.A dry state like solid dosage forms usually has better stability than suspensions.Therefore, to limit or avoid the stability problem of nanosuspension, they should be changed to powder form [27].
Chemical stability problems like hydrolysis and oxidation can also affect the stability of nanosuspension [28].One of the popular methods to improve chemical stability is either by changing the nanosuspension from a liquid state to solids on the shelf where the stability was achieved better than nanosuspension.The second method to enhance the chemical stability of the drug during nanonization is by increasing the concentration of the nanosuspension [3].
One of the important approaches to stabilize the nanoparticles is the addition of a suitable stabilizer to the formulation during the nanonization process [29].The addition of a stabilizer to the nanosuspension formulation can play an important role to limit agglomeration by   reducing the free energy of the system during the nanonization process.
To determine the optimum storage conditions for the nanoparticles, a stability study has performed to spotlight any physical or chemical instability issue that can occur during the shelf life of the prepared nanoparticle formulations.Accelerated stability study for 6 months at 40°C and 75% RH is recommended in the guidelines of the ICH, representing the long-term stability of nanoparticles [30].

Conclusion
Enhancing the dissolution rate of APG by particle size reduction using milling or grinding method has not been approached.QbD persuades the pharmaceutical industry to use risk management and science-based manufacturing principles to earn process and product understanding and thus assures the quality of the product.
The study concluded that the application of the QbD approach in the nanonization of APG by planetary ball milling could help in defining the CMAs and CPPs affecting the final nanoparticle product attributes (CQAs).In addition, the nanonization of APG resulted in improving the in vitro drug dissolution rate, and, in turn, enhanced its oral bioavailability using the rats' model.

Figure 1 :
Figure 1: Response surface plot for the effect of milling speed and solid/solvent ratio on APG nanoparticle size (a), zeta potential (b), and % DE 60 (c) (ball size was kept constant at medium level).

Figure 2 :
Figure 2: Particles size of different APG nanoparticle formulations.

Figure 3 :
Figure 3: Zeta potential values of different APG nanoparticle formulations.

Figure 4 :
Figure 4: % DE 60 values of APG from the different nanoparticles' formulations.
high risk, yellow colormedium risk, and green colorlow risk.

Figure 5 :
Figure 5: Ishikawa Fishbone diagram for the effect of CMAs and CPPs on the CQAs of APG nanoparticle formulation prepared by planetary ball wet milling.

Figure 8 :
Figure 8: Drug concentration-time profile curve of APG after oral administration of its optimized nanoparticle formulation in comparison to the pure drug.

Figure 9 :
Figure 9: Dissolution profile of APG nanoparticles at initial nanoparticles, 3 months stability, and 6 months stability compared to untreated powder.

Table 1 :
Independent factors and dependent parameters (responses) for APG nanoparticle formulation prepared by wet milling

Table 2 :
Matrix of 3³ full factorial design for APG nanoparticles formulations prepared by planetary ball milling

Table 3 :
Screening data for the effect of different stabilizers on APG nanoparticles prepared by planetary ball milling by using 500 rpm as milling speed, milling ball size of 0.5 mm, solid/solvent ratio of 0.1, and 3 milling cycles (each of 10 min with 5 min pause)

Table 4 :
Analysis of variance for particle size, zeta potential, and % DE 60 of APG nanoparticle formulations prepared by planetary ball milling * Significant differences; p value less than 0.05

Table 5 :
Risk assessment of potential factors affecting CQAs of APG nanoparticles prepared by planetary ball milling by PHA tool

Table 8 :
Effect of storage on particles size, zeta potential, and % DE 60 values of optimized APG nanoparticle formula