3.2.5. Stability Studies of G5 Complex Physical Stability

The stability of the G5 25:1 was studied at daylight and in the dark in different temperature conditions ranging from (4–50 ◦C). For a period of 6 months, the nanoformulations were stored in colorless Eppendorfs and observed for any sign of color change, precipitation, or turbidity. Results are presented in Table S3. This study was performed to investigate which storing condition is suitable for G5 complex 25:1. Based on the result of this investigation, it is clearly indicated that G5 complex 25:1 must be stored in amber containers at 4 ◦C. In addition, these nanoparticles were visualized using SEM to assess their shape and size. As seen in Figure 2B, the complex possessed an almost spherical shape and particle size equivalent to those measured by Zetasizer Nano ZS (Malvern, UK).

#### Drug Content

The results for the stability of RBX in G5 25:1 complex were similar to the results obtained from the G4.5 complex 25:1 stability study. It was noticed that there were no significant differences in RBX peak area between G4.5 and G5 samples. Percent content of RBX in G5 complex 25:1 was calculated and demonstrated in Figure S4B.

#### TEM

G5 complex 25:1 and the empty PAMAM dendrimers G5 were visualized by TEM to determine their morphology and to confirm the loading of RBX. Figure 4 displays the morphology of empty PAMAM dendrimers G5 as well as G5 complex 25:1 complex. *Pharmaceutics* **2022**, *14*, x FOR PEER REVIEW 12 of 23 *Pharmaceutics* **2022**, *14*, x FOR PEER REVIEW 12 of 23

**Figure 4.** TEM overview images of (**A**) empty PAMAM dendrimers G5 and (**B**) G5 complex 25:1. **Figure 4.** TEM overview images of (**A**) empty PAMAM dendrimers G5 and (**B**) G5 complex 25:1. **Figure 4.** TEM overview images of (**A**) empty PAMAM dendrimers G5 and (**B**) G5 complex 25:1.

#### *3.3. Cell Viability Studies (MTS Assay) 3.3. Cell Viability Studies (MTS Assay) 3.3. Cell Viability Studies (MTS Assay)*

3.3.1. Dose–Response Studies in Controlled Mediums 3.3.1. Dose–Response Studies in Controlled Mediums 3.3.1. Dose–Response Studies in Controlled Mediums

Dose–Response of RBX Dose–Response of RBX Dose–Response of RBX

In this work, the safety of RBX was assessed by cell viability studies using MIO-M1 cell line. These investigations were carried out using MTS assay. MIO-M1 cells were cultured in a controlled glucose medium (5.55 mM glucose media) and treated with a range of doses of RBX **(**100 nM–1 μM). After 24 h, viability (%) of the treated cells was assessed and compared to the viability of untreated cells that were cultured in the same controlled conditions. As seen in Figure 5, viability of the treated cells with RBX did not show any significant reduction when compared to viability of untreated cells (*p* > 0.05). This finding indicates that RBX is safe and the investigated doses are tolerable by MIO-M1 cell line in normal conditions. In this work, the safety of RBX was assessed by cell viability studies using MIO-M1 cell line. These investigations were carried out using MTS assay. MIO-M1 cells were cultured in a controlled glucose medium (5.55 mM glucose media) and treated with a range of doses of RBX (100 nM–1 µM). After 24 h, viability (%) of the treated cells was assessed and compared to the viability of untreated cells that were cultured in the same controlled conditions. As seen in Figure 5, viability of the treated cells with RBX did not show any significant reduction when compared to viability of untreated cells (*p* > 0.05). This finding indicates that RBX is safe and the investigated doses are tolerable by MIO-M1 cell line in normal conditions. In this work, the safety of RBX was assessed by cell viability studies using MIO-M1 cell line. These investigations were carried out using MTS assay. MIO-M1 cells were cultured in a controlled glucose medium (5.55 mM glucose media) and treated with a range of doses of RBX **(**100 nM–1 μM). After 24 h, viability (%) of the treated cells was assessed and compared to the viability of untreated cells that were cultured in the same controlled conditions. As seen in Figure 5, viability of the treated cells with RBX did not show any significant reduction when compared to viability of untreated cells (*p* > 0.05). This finding indicates that RBX is safe and the investigated doses are tolerable by MIO-M1 cell line in normal conditions.

Dose–Response of PAMAM Dendrimers G4.5

Dose–Response of PAMAM Dendrimers G4.5

**Figure 5.** Effect of RBX after 24 h exposure on the cell viability of MIO-M1 cells under controlled conditions (mean ± SD, *n* = 3). **Figure 5.** Effect of RBX after 24 h exposure on the cell viability of MIO-M1 cells under controlled conditions (mean ± SD, *n* = 3). **Figure 5.** Effect of RBX after 24 h exposure on the cell viability of MIO-M1 cells under controlled conditions (mean ± SD, *n* = 3).

Similar to the previous work, the safety of PAMAM dendrimers G4.5 was assessed

Similar to the previous work, the safety of PAMAM dendrimers G4.5 was assessed

to the viability of untreated cells that were cultured in the same controlled conditions. As seen in Figure 6A, viability of the treated cells with PAMAM dendrimers G4.5 did not

to the viability of untreated cells that were cultured in the same controlled conditions. As seen in Figure 6A, viability of the treated cells with PAMAM dendrimers G4.5 did not

Dose–Response of PAMAM Dendrimers G4.5

Similar to the previous work, the safety of PAMAM dendrimers G4.5 was assessed using different concentrations ranging from (1–50 nM). These investigations were carried out using MTS assay. After 24 h, viability of the treated cells was assessed and compared to the viability of untreated cells that were cultured in the same controlled conditions. As seen in Figure 6A, viability of the treated cells with PAMAM dendrimers G4.5 did not show any significant reduction when compared to viability of untreated cells (*p* > 0.05). These finding indicates that PAMAM dendrimers G4.5 is safe and the investigated doses are tolerable by MIO-M1 cell line in normal conditions. *Pharmaceutics* **2022**, *14*, x FOR PEER REVIEW 13 of 23 show any significant reduction when compared to viability of untreated cells (*p* > 0.05). These finding indicates that PAMAM dendrimers G4.5 is safe and the investigated doses are tolerable by MIO-M1 cell line in normal conditions.

**Figure 6.** Effect of (**A**) PAMAM dendrimers G4.5 and (**B**) PAMAM dendrimers G5 after 24 h. Exposure on the cell viability of MIO-M1 cells under controlled conditions (mean ± SD, *n* = 3). **Figure 6.** Effect of (**A**) PAMAM dendrimers G4.5 and (**B**) PAMAM dendrimers G5 after 24 h. Exposure on the cell viability of MIO-M1 cells under controlled conditions (mean ± SD, *n* = 3).

#### Dose–Response of PAMAM dendrimers G5 Dose–Response of PAMAM dendrimers G5

Similar to the previous work, the safety of PAMAM dendrimers G5 was assessed using different concentrations ranging from (1–50 nM). These investigations were carried out using MTS assay. After 24 h, viability of the treated cells was assessed and compared to the viability% of untreated cells that were cultured in the same controlled conditions. As seen in Figure 6B, viability of the treated cells with PAMAM dendrimers G5 did not show any significant reduction when compared to viability of untreated cells (*p* > 0.05). These findings indicate that the PAMAM dendrimer G5 is safe and the investigated doses are tolerable by MIO-M1 cell line in normal conditions. Similar to the previous work, the safety of PAMAM dendrimers G5 was assessed using different concentrations ranging from (1–50 nM). These investigations were carried out using MTS assay. After 24 h, viability of the treated cells was assessed and compared to the viability% of untreated cells that were cultured in the same controlled conditions. As seen in Figure 6B, viability of the treated cells with PAMAM dendrimers G5 did not show any significant reduction when compared to viability of untreated cells (*p* > 0.05). These findings indicate that the PAMAM dendrimer G5 is safe and the investigated doses are tolerable by MIO-M1 cell line in normal conditions.

#### Dose–Response of G4.5 Complex 25:1 Dose–Response of G4.5 Complex 25:1

It was clearly indicated that RBX and PAMAM dendrimers are well tolerated by MIO-M1 cells. More investigations were carried out to assess the safety of the formulated G4.5 complex 25:1. This was carried out using MTS assay. After 24 h, viability of the treated cells was assessed using different concentrations ranging from (100 nM–1 µM) and compared to the viability of untreated cells that were cultured in the same controlled conditions. As seen in Figure 7A, viability of the treated cells with PAMAM dendrimers G4.5 did not show any significant reduction when compared to the viability of untreated cells (*p* > 0.05). These finding indicates that G4.5 complex 25:1 is safe and the investigated doses are tolerable by MIO-M1 cell line in normal conditions. It was clearly indicated that RBX and PAMAM dendrimers are well tolerated by MIO-M1 cells. More investigations were carried out to assess the safety of the formulated G4.5 complex 25:1. This was carried out using MTS assay. After 24 h, viability of the treated cells was assessed using different concentrations ranging from (100 nM–1 µM) and compared to the viability of untreated cells that were cultured in the same controlled conditions. As seen in Figure 7A, viability of the treated cells with PAMAM dendrimers G4.5 did not show any significant reduction when compared to the viability of untreated cells (*p* > 0.05). These finding indicates that G4.5 complex 25:1 is safe and the investigated doses are tolerable by MIO-M1 cell line in normal conditions.

**Figure 7.** Effect of G4.5 complex 25:1 (**A**) and G5 complex 25:1 (**B**) after 24 h exposure on the cell viability of MIO-M1 cells under controlled conditions (mean ± SD; *n* = 3). **Figure 7.** Effect of G4.5 complex 25:1 (**A**) and G5 complex 25:1 (**B**) after 24 h exposure on the cell viability of MIO-M1 cells under controlled conditions (mean ± SD; *n* = 3).

. These values were

#### Dose–Response of RBX G5 Complex 25:1

Dose–Response of RBX G5 Complex 25:1 Similarly, the safety of G5 complex 25:1 was assessed using MTS assay. After 24 h, viability of the treated cells was assessed using different concentrations ranging from (100 nM–1 µM) and compared to the viability of untreated cells that were cultured in the same controlled conditions. As seen in Figure 7B, viability of the treated cells with G5 complex 25:1 did not show any significant reduction when compared to viability of untreated cells (*p* > 0.05). This finding indicates that G5 complex 25:1 is safe and the investigated doses are tolerable by MIO-M1 cell line in normal conditions. Similarly, the safety of G5 complex 25:1 was assessed using MTS assay. After 24 h, viability of the treated cells was assessed using different concentrations ranging from (100 nM–1 µM) and compared to the viability of untreated cells that were cultured in the same controlled conditions. As seen in Figure 7B, viability of the treated cells with G5 complex 25:1 did not show any significant reduction when compared to viability of untreated cells (*p* > 0.05). This finding indicates that G5 complex 25:1 is safe and the investigated doses are tolerable by MIO-M1 cell line in normal conditions.

#### 3.3.2. Effect of High Glucose Treatment on the Viability of MIO-M1 Cells 3.3.2. Effect of High Glucose Treatment on the Viability of MIO-M1 Cells

A series of experiments were conducted to test the effect of RBX, PAMAM dendrimers, and the proposed complexes on the viability of MIO-M1 cells in two different culture mediums; the controlled medium (5.55 mM) and high glucose medium were approximately five times normal levels of glucose to simulate hyperglycemia for MIO-M1 cells (25 mM). To test the effect of the tested compounds in high glucose treatment, MIO-M1 cells were incubated with the tested compounds for 24, 48, and 72 h. Then, viability was calculated and compared to the relevant controlled medium of that particular time point. As seen in Figure 8, in the presence of the tested compounds, there is no significant loss in cell viability of MIO-M1 cells cultured in high glucose media when compared to relevant controlled media of each studied time point (*p* > 0.05). A series of experiments were conducted to test the effect of RBX, PAMAM dendrimers, and the proposed complexes on the viability of MIO-M1 cells in two different culture mediums; the controlled medium (5.55 mM) and high glucose medium were approximately five times normal levels of glucose to simulate hyperglycemia for MIO-M1 cells (25 mM). To test the effect of the tested compounds in high glucose treatment, MIO-M1 cells were incubated with the tested compounds for 24, 48, and 72 h. Then, viability was calculated and compared to the relevant controlled medium of that particular time point. As seen in Figure 8, in the presence of the tested compounds, there is no significant loss in cell viability of MIO-M1 cells cultured in high glucose media when compared to relevant controlled media of each studied time point (*p* > 0.05). *Pharmaceutics* **2022**, *14*, x FOR PEER REVIEW 15 of 23

**Figure 8.** Effect of the tested compounds after (**A**) 24 h and (**B**) 48 h. Exposure on the cell viability of MIO-M1 cells under controlled and high glucose mediums (mean ± SD, *n* = 3). **Figure 8.** Effect of the tested compounds after (**A**) 24 h and (**B**) 48 h. Exposure on the cell viability of MIO-M1 cells under controlled and high glucose mediums (mean ± SD, *n* = 3).

The effect of RBX and dendrimers (with or without RBX) on the monolayer integrity

inserts before TEER values were measured to evaluate the monolayer

ginning of this experiment, the MIO-M1 monolayer was developed on a porous mem-

integrity. TEER values across the cell monolayer were measured every other day until the

used as controls for each particular insert. A total of 500 nM of RBX and dendrimers (with or without RBX) was incubated for 24 h, and then TEER values were measured again to investigate the influence of complexes on monolayer integrity of MIO-M1 cells as compared to RBX and dendrimers alone. A paired *t*-test was used to determine whether there was a statistically significant difference between TEER values when the MIO-M1 monolayer was treated compared to non-treated cells. As highlighted in Figure 9, TEER values were shown to be approximately 13.0 %, which is the highest reduction (9.7 ± 1.7→8.4 ± 0.7) across the cell monolayer membrane following 24 h of treatment with G4.5 (0.05 nM). Even though the TEER values of G4.5 (0.05 nM) showed the highest reduction across the cell monolayer membrane, it was not statistically significant when compared to TEER values before treatment at a *p*-value less than 0.01. Indeed, the reduction in TEER value of MIO-M1 monolayer, following the treatment of G5 complex 25:1, G4.5 complex 25:1, and RBX, was not statistically significant according to paired student *t*-test (*p* value < 0.01). As detailed in Figure 10, the data exhibited a high Papp (4.0 ± 0.1 × 10−6 cm/s) of the groups without the MIO-M1 monolayer. However, the Papp (1.9 ± 0.0 × 10−6 cm/s) of the developed MIO-M1 without treatment was hardly distinguishable from that observed by Papp of G5 complex and G4.5 complex, at the same condition, were 1.8 ± 0.0 × 10−6 cm/s and 1.7 ± 0.0 × 10−6 cm/s, respectively. In the in vitro permeability assessment, the collected data of cell treated with G5 complex and G4.5 complex showed that the permeability of MIO-M1 did not significantly change compared to that of non-treated MIO-M1 cells. One-way ANOVA analysis revealed that no statistically significant differences were detected among the permeability of the MIO-M1 monolayer with or without treatment of G5 complex, G4.5 complex, and RBX alone. Further statistical analysis by a *post-hoc* test showed that the difference between dendrimers with or without RBX on MIO-M1 permeability

cell monolayers exhibited constant TEER values of more than 8 Ω\*cm<sup>2</sup>

was not statistically significant at a *p* value < 0.01.

3.3.3. In Vitro Permeability Study

brane of Transwell®

#### 3.3.3. In Vitro Permeability Study

The effect of RBX and dendrimers (with or without RBX) on the monolayer integrity of MIO-M1 was examined by measuring TEER values across the cell monolayer membrane in the upper Transwell® chamber before and after treatments (Figure 9). At the beginning of this experiment, the MIO-M1 monolayer was developed on a porous membrane of Transwell® inserts before TEER values were measured to evaluate the monolayer integrity. TEER values across the cell monolayer were measured every other day until the cell monolayers exhibited constant TEER values of more than 8 Ω\*cm<sup>2</sup> . These values were used as controls for each particular insert. A total of 500 nM of RBX and dendrimers (with or without RBX) was incubated for 24 h, and then TEER values were measured again to investigate the influence of complexes on monolayer integrity of MIO-M1 cells as compared to RBX and dendrimers alone. A paired *t*-test was used to determine whether there was a statistically significant difference between TEER values when the MIO-M1 monolayer was treated compared to non-treated cells. As highlighted in Figure 9, TEER values were shown to be approximately 13.0 %, which is the highest reduction (9.7 ± 1.7→8.4 ± 0.7) across the cell monolayer membrane following 24 h of treatment with G4.5 (0.05 nM). Even though the TEER values of G4.5 (0.05 nM) showed the highest reduction across the cell monolayer membrane, it was not statistically significant when compared to TEER values before treatment at a *p*-value less than 0.01. Indeed, the reduction in TEER value of MIO-M1 monolayer, following the treatment of G5 complex 25:1, G4.5 complex 25:1, and RBX, was not statistically significant according to paired student *t*-test (*p* value < 0.01). As detailed in Figure 10, the data exhibited a high Papp (4.0 <sup>±</sup> 0.1 <sup>×</sup> <sup>10</sup>−<sup>6</sup> cm/s) of the groups without the MIO-M1 monolayer. However, the Papp (1.9 <sup>±</sup> 0.0 <sup>×</sup> <sup>10</sup>−<sup>6</sup> cm/s) of the developed MIO-M1 without treatment was hardly distinguishable from that observed by Papp of G5 complex and G4.5 complex, at the same condition, were 1.8 <sup>±</sup> 0.0 <sup>×</sup> <sup>10</sup>−<sup>6</sup> cm/s and 1.7 <sup>±</sup> 0.0 <sup>×</sup> <sup>10</sup>−<sup>6</sup> cm/s, respectively. In the in vitro permeability assessment, the collected data of cell treated with G5 complex and G4.5 complex showed that the permeability of MIO-M1 did not significantly change compared to that of non-treated MIO-M1 cells. One-way ANOVA analysis revealed that no statistically significant differences were detected among the permeability of the MIO-M1 monolayer with or without treatment of G5 complex, G4.5 complex, and RBX alone. Further statistical analysis by a *post-hoc* test showed that the difference between dendrimers with or without RBX on MIO-M1 permeability was not statistically significant at a *p* value < 0.01. *Pharmaceutics* **2022**, *14*, x FOR PEER REVIEW 16 of 23

**Figure 9.** TEER measurements of monolayer cell membrane 24 h after exposure to treatments. Data represent % of TEER mean of the control ± SEM (*n* = 3–4). **Figure 9.** TEER measurements of monolayer cell membrane 24 h after exposure to treatments. Data represent % of TEER mean of the control ± SEM (*n* = 3–4).

**Figure 10.** Apparent permeability coefficient (Papp) of the MIO-M1 monolayer cell after 24 treat-

PAMAM dendrimers have recently been studied to determine their ability to provide an effective and noninvasive drug delivery to the retina. Recent findings of Chang Liu et al. (2016) found that PAMAM dendrimers could rapidly penetrate from the surface of the eye into the vitreous and resided in the retina. This new approach is considered to be important in the development of a safe and noninvasive drug delivery for posterior segment diseases [41]. Herein, we aimed to develop a retinal delivery system by which RBX would be passively delivered from the surface of the eye to the retina. This was achieved through the development of RBX nanoparticles using PAMAM dendrimers. To achieve

ments \* *p* value < 0.01.

**4. Discussion**

represent % of TEER mean of the control ± SEM (*n* = 3–4).

**Figure 10.** Apparent permeability coefficient (Papp) of the MIO-M1 monolayer cell after 24 treatments \* *p* value < 0.01. **Figure 10.** Apparent permeability coefficient (Papp) of the MIO-M1 monolayer cell after 24 treatments \* *p* value < 0.01.

**Figure 9.** TEER measurements of monolayer cell membrane 24 h after exposure to treatments. Data

#### **4. Discussion 4. Discussion**

PAMAM dendrimers have recently been studied to determine their ability to provide an effective and noninvasive drug delivery to the retina. Recent findings of Chang Liu et al. (2016) found that PAMAM dendrimers could rapidly penetrate from the surface of the eye into the vitreous and resided in the retina. This new approach is considered to be important in the development of a safe and noninvasive drug delivery for posterior segment diseases [41]. Herein, we aimed to develop a retinal delivery system by which RBX would be passively delivered from the surface of the eye to the retina. This was achieved through the development of RBX nanoparticles using PAMAM dendrimers. To achieve PAMAM dendrimers have recently been studied to determine their ability to provide an effective and noninvasive drug delivery to the retina. Recent findings of Chang Liu et al. (2016) found that PAMAM dendrimers could rapidly penetrate from the surface of the eye into the vitreous and resided in the retina. This new approach is considered to be important in the development of a safe and noninvasive drug delivery for posterior segment diseases [41]. Herein, we aimed to develop a retinal delivery system by which RBX would be passively delivered from the surface of the eye to the retina. This was achieved through the development of RBX nanoparticles using PAMAM dendrimers. To achieve our aim, several complexes were formulated using 2 different generations and surface functional groups (i.e., anionic PAMAM dendrimers G4.5 and neutral PAMAM dendrimers G5). Then, these complexes were characterized and assessed for their cytotoxicity effects using the MIO-M1 cell line. The ideal nanosystem for ocular therapy formulations must have high loading efficiency, small particle size, and fast drug release [42]. In the current study, the diameter of PAMAM G4.5 and G5 was found to be 186.6 ± 2.3 and 214.9 ± 8.5 nm, respectively, supported by the findings of Yavuz (2015) [1]. Furthermore, nanoparticles with a size range typically < 400.0 nm are suitable for ophthalmic use according to Gorantla et al. (2020) [43]. In addition, the results of this study revealed that the mean particle size values of G4.5 and G5 loaded with RBX varied from 289.4 ± 39.9 nm to 482.4 ± 28.7 nm, indicating the loaded polymer is suitable as an ophthalmic formulation. Results indicated that the proposed complexes have significantly increased particle size when compared to the empty PAMAM dendrimers (*p* < 0.05), which is attributed to entrapment of RBX molecules within the internal cavities of PAMAM dendrimers. Sizes of both complex formulations of each generation were enlarged in a range approximately between 100.0–280.0 nm with the exception of PAMAM G5 5:1 complex which was measured as 669.8 ± 35.3 nm. High particle size of this formulation might be a result of aggregation. Such results were in line with the findings of Yavuz et al. [1].

The PDI is usually measured to determine the heterogeneity of a sample based on the size. PDI is the potential result of size distribution or aggregation of the particles in a sample during isolation or analysis [44,45]. It was demonstrated that PDI values of empty PAMAM dendrimers G4.5 and G5 were 0.297 ± 0.055 and 0.350 ± 0.008, respectively. Surfing the literature revealed that PDI < 0.35 is suitable for ocular drug delivery [46]. G4.5 and G5 complexes possessed a PDI value ranging from (0.335–0.394). Such findings revealed that there are no

significant changes in PDI of the dendrimers after loading it with RBX (*p* > 0.05). Findings of this work were indicative of monodispersity of PAMAM dendrimers. Specifically, G4.5 and G5 complex 25:1 possessed the least PDI among all studied nanoformulations (PDI ≤ 0.35) indicating high particle homogeneity. In fact, the lower PDI value is much closer to achieving a monodisperse system; so, the latter complexes were considered the most monodisperse systems that is suitable for ocular drug delivery of RBX. However, an exception to the aforesaid findings was observed with G5 complex 5:1. This complex possessed a significantly high PDI value (0.587 ± 0.106), which was previously discussed in possessing a high particle size as well. This observation could be related to the aggregation of the nanoparticles. According to Tawfik et al. (2018), PDI measurements of all studied vardenafil/PAMAM complexes were characterized by small PDI values ≤ 0.25 [21]. Another study by Peng et al. demonstrated that the PDI of the PAMAM formulation was found to be in a range of (0.230–0.339), which supports our findings [47].

ζ-potential measurements determine the electrostatic potential at the electrical double layer surrounding a nanoparticle in a solution. In fact, ζ-potential plays an important role in determining the stability of nanoparticles as well as knowing the intensity of electrostatic attraction between biomolecules and the nanoparticles. Nanoparticles with a ζ-potential between −10.0 and +10.0 mV are considered approximately neutral, while nanoparticles with ζ-potentials of greater than +30.0 mV or less than −30.0 mV are considered to be strongly cationic and strongly anionic, respectively. Since most of eye cellular membranes are negatively charged, ζ-potential can affect a nanoparticle's tendency to permeate membranes, with cationic particles generally displaying more toxicity associated with cell wall disruption [48]. A study carried out by Yavuz (2015) found that ζ-potential values of blank dendrimer generations G4.5 was strongly anionic (−45.9 + 8.8) [1]. On the other hand, nanosynthons; manufacturer of PAMAM dendrimers G5, has stated that PAMAM dendrimers G5 possesses a neutral charge as they were synthesized with aminoethanol surface. In the current work, the ζ-potential measurement of G4.5 was −44.0 mV, which is similar to the result demonstrated by Yavuz [1]. Additionally, it was found that the surface charge of G5 was −0.2 mV which is in line with the information provided by nanosynthons. In this work, measurements of ζ-potential for G4.5 and G5 complexes showed that, ζ-potential values, either negative or neutral, were increased in the presence of RBX. Moreover, G4.5 complexes 25:1 and 1:1 remained anionic in contrast to PAMAM dendrimers used in complex 2.5:1 and 5:1, which exhibited a neutral charge after the loading process. Results obtained from this work also showed that zeta potential values for neutral complexes (i.e., G5 complexes) increased yet did not significantly change after the loading process (*p* > 0.05). However, it was clear that the loading of RBX, a weakly basic drug, into PAMAM dendrimers had a positive impact on these ζ-potential findings. Results demonstrated by Tawfik et al. (2018) supported our findings as vardenafil, a weakly basic drug (similar to RBX), was encapsulated within PAMAM dendrimers and ζ-potential measurements were increased after the loading process [21]. Drugs, such as RBX, could be encapsulated or conjugated to dendrimers. The morphology of G4.5 and G5 complexes was observed by TEM. It was shown that complexes as well as blank PAMAM dendrimers showed a discrete spherical morphology. Investigation of drug–polymer binding revealed that RBX was encapsulated within PAMAM dendrimers and no binding to PAMAM surface occurred. Furthermore, the ability of a fixed amount of RBX to be incorporated into various dendrimer generations and concentrations was investigated to estimate DE%. The drug loading efficiency is influenced by certain factors. These include the nature of the polymer, the encapsulated drug molecules, and polymer–drug ratio [25]. DE% of RBX in G4.5 and G5 complexes varied from 88.8% to 98.68%. It was demonstrated that drug loading efficiency of RBX did not show any significant differences among the proposed nanoparticles (*p* > 0.05). However, increasing the amount of RBX did not significantly show any differences compared to lower concentrations in the proposed nanoparticles (*p* > 0.05). Nabavizadeh et al. (2016) prepared different concentrations of 5:1, 5:2, 5:3, 5:4, and 5:5 of capecitabine: PAMAM dendrimers and studied DE% of these formulations. They

found that DE% of capecitabine decreased with higher PAMAM dendrimer concentration. Increasing PAMAM dendrimer concentration has led to stronger electrostatic interactions between capecitabine and PAMAM dendrimers as a result [22,23]. In contrast to the findings reported by the later study, no significant differences occurred between DE% of G4.5 complexes as well as G5 complexes. Our findings were not similar to Nabavizadeh et al. (2016) results and were not affected by the factor of conjugation [22]. We supported our findings by visualizing the loading pattern of the complexes under TEM, which showed that RBX was only encapsulated within PAMAM dendrimers and not conjugated to PAMAM surface. Tawfik et al. (2018) studied vardenafil DE% among the proposed nanoparticles and revealed that direct correlation between the dendrimer concentration and vardenafil DE% was not held true for all concentrations. Statistically, no significant differences in vardenafil DE% were observed between the proposed nanoformulations [21]. These finding are in line with our results, which showed no significant differences in DE% were observed between 1:1, 2.5:1, 5:1, and 25:1 RBX: PAMAM.

Based on the results of DE%, the appropriate ratio of RBX-PAMAM dendrimer was selected for stability studies and ex vivo cell line studies, which was found to be 25:1 for each studied PAMAM generation. Moreover, Yavuz et al. (2015) studied the in vitro drug release of dexamethasone nanoparticles using different generations of PAMAM dendrimers. Since the nanoparticles are expected to be cleared from the eye in approximately an hour, it was desired that dexamethasone should be immediately released to penetrate to the posterior segment of the eye. Thus, a 3 h long in vitro drug release study was conducted in PBS at 37 ◦C. In a period of 3 h, the studied complexes released dexamethasone in a range of 40.0 to 80.0% [1]. In our work, a period of 8 h in vitro release study was carried out since the complex formulations were designed as fast release nanoparticles to be applied topically and their ocular retention time is suggested to be shorter. After the release study period has ended, it was found that G4.5 complexes released RBX in a range from 47.0–86.0%, while G5 complexes released RBX in a range of 25.0–81.0%. Cumulative release (%) of RBX from the studied nanoparticles showed that the drug was released in a sustained manner. In this study, it was clearly noticed that the highest release of RBX was demonstrated with G4.5 complex 25:1 and G5 complex 25:1 up to 86.0% and 81.0% respectively *(p* < 0.05). The release of RBX from the suggested complexes was compared to other in vitro release studies of PAMAM dendrimers. Gothwal et al. (2017) formulated bendamustine nanoparticles of PAMAM dendrimers G4. In this study, PAMAM dendrimers released the drug in a sustained manner up to 72 h [49,50]. Another study presented by Yesil-Celiktas et al. (2017) reported that PAMAM dendrimers provided a sustained release of BCA for six consecutive days [51]. All of these studies are in line with our findings of this study which demonstrated that PAMAM dendrimers are found to provide a sustained release of RBX in all studied formulations. In another study, Kesharwani et al. (2015) observed that the curcumin release decreased as PAMAM generation increased despite having a higher loading capacity [52]. The result of this study supports our findings in which RBX was released in higher percentages from the lower generation (G4.5) when compared to RBX release from the higher generation (G5).

Furthermore, the stability of G4.5 and G5 25:1 complexes was inspected visually to evaluate the appearance of any color change, precipitation, or turbidity. This was carried out to determine the disintegration of the proposed nanoparticles when exposed to light and/or high temperatures. Results obtained from this work showed that after the study period has ended, it was clearly observed that with increasing the storage temperature, physical changes such as color change and precipitation were recorded. Our findings revealed that the nanoformulations were found to be physically stable at 4 ◦C, whether kept in the dark or the light. Shadrack et al. (2015) has investigated physical stability of tetramethylscutellarein nanoparticles using PAMAM dendrimers at different temperature conditions (i.e., 0, 27 and 40 ◦C) for two months. Investigations showed a sign of color change, turbidity and precipitate formation were observed for formulations stored in light indicating that light had an influence on their stability [25]. These findings come in agreement with results

obtained from our study that suggested that light exposure have can influence the stability of RBX nanoparticles. Findings of the later study found that nanoparticles were relatively stable when kept in the dark at high temperature (i.e., 40 ◦C). In contrast to these findings, we have observed physical changes of the nanoformulation at higher temperatures even when kept in the dark. This phenomenon was described by Prajapati et al. (2009) who explained that the formation of precipitates observed may be attributed to the opening of dendritic structures at higher temperatures [53]. By that means, we suggest that RBX complexes should be kept in the dark at 4 ◦C to provide the maximum stability of the proposed formulations. Also, another measurement was carried out to investigate the stability of RBX in the proposed complexes when exposed to different storage conditions. When each study period has ended, drug content was determined in each studied complex by using UPLC MS/MS. After that, percent content of RBX was calculated to estimate the loss of RBX in each tested complex. Results obtained from UPLC-MS/MS showed that RBX in G4.5 complex 25:1 remained above 90% stability at 4 ◦C for 6 months, the stability was slightly decreased over time from one month to 6 months. Moreover, percent content of RBX was apparently decreased by more than 20% with increasing the temperature. The results of this experiment revealed that the percent content of RBX is higher at 4 > 25 > 37 > 50 ◦C. Similar finding were observed with G5 complex 25:1. All studied nanoformulations were more stable when stored in the dark despite the influence of temperature over time. As a result of these observations, it is suggested that RBX nanoparticles should be stored in amber containers at 4 ◦C for maximum stability. RBX is a very sensitive drug that should be kept away from light and higher temperatures. To avoid RBX degradation, this drug must be stored in an amber container at −20 ◦C. In this work, it was observed that formulating RBX in a nanoparticulate system using PAMAM dendrimers can improve the stability of RBX in addition to its tendency to provide a facile retinal drug delivery.

Furthermore, ex vivo cell line studies were conducted to evaluate the safety of the formulated nanoparticles in MIO-M1 cells. It was found that the drug RBX in a range of concentrations did not affect the cell viability in both controlled and high glucose mediums. Similar results were observed with the work of Aldarwesh (2016) [9]. Also, empty and loaded PAMAM dendrimers G4.5 and G5 were investigated for their cytotoxicity effect in MIO-M1 cell line. The nanoparticles, in different concentrations, did not affect the viability of these cells, despite the type of cell medium (i.e., controlled or high glucose medium). To our knowledge, this is the first cytotoxicity study that was carried out to assess ex vivo behavior of PAMAM dendrimers using the MIO-M1 cell line. To assess the negative effect of complexes on the MIO-M1 monolayer integrity as a simple blood–retinal barrier (BRB) model, an in vitro permeability study on Transwell® and TEER was carried out. The cells were developed on the upper surface of the semi-permeable membrane in the Transwell® culture chambers in order to quantify the disruption level of new complexes on a monolayer of MIO-M1 cells. The difference in the TEER values of dendrimers and complexes was not statistically significant compared to TEER values before treatment at a *p*-value less than 0.01. Indeed, the reduction in TEER value of MIO-M1 monolayer, following the treatment of G5 complex 25:1, G4.5 complex 25:1, RBX, was not statistically significant according to paired Student *t*-test (*p* < 0.01). These findings suggest that both G5 complex 25:1 and G4.5 complex 25:1 did not significantly damage the monolayer integrity of MIO-M1 cells. The value of apparent permeability of developed MIO-M1 monolayers was confirmed by comparing it to the literature value for MIO-M1 cells [54]. Furthermore, the permeability values of about 10–6 cm/s have been previously observed in various in vitro BRB studies [32,38]). After confirmation of the monolayer integrity of MIO-M1, the monolayer integrity of MIO-M1 was inspected by calculating Papp of the barriers using FITC-dextran. During in vitro permeability assay, the monolayer cells were exposed to one of the G5 complex 25:1, G4.5 complex 25:1, and RBX for 24 h. The effect of the complexes on the barrier integrity was examined and compared to RBX, non-treated cells, and no cell groups with regards to the impact of treatment on the apparent permeability coefficient of the MIO-M1 monolayer. As shown in Figure 10, the Papp value of non-seeded insert was significantly higher than the

Papp value of untreated cell, indicating the formation of MIO-M1 monolayer membrane on the Transwell®. Furthermore, the results showed no statistically significant differences among the permeability value of the MIO-M1 monolayer with or without treatment of dendrimers and complexes. The difference between dendrimers with or without RBX on MIO-M1 permeability was not statistically significant (*p* >0.05). These findings suggest that both G5 complex and G4.5 complex did not have a negative effect on the monolayer integrity of MIO-M1 cells. In addition, the above mentioned data support the cell viability studies which indicate the safety of G5 complex and G4.5 complex on MIO-M1 cell line in normal conditions.

#### **5. Conclusions**

RBX is a newly potent anti-VEGF drug that has been recently investigated for the treatment of DR. At present, anti-VEGF therapy is usually delivered to the retina via intravitreal injections carrying many ocular complications that cause patient noncompliance. To provide the maximum therapeutic outcomes of treating diabetic retinopathy, a non-invasive therapy of RBX was designed. This was carried out using nanoparticles of PAMAM dendrimers. Several methodologies have been implemented and the noninvasive nanoparticles that show the best in vitro characterization were formulated and characterized. The proposed nanoparticles will overcome ocular complications associated with the invasive therapy of diabetic retinopathy and the result will help in increasing patient adherence. The future direction of this study is to assess the in vivo behavior of RBX nanoparticles and to evaluate their permeability properties and ability to provide non-invasive retinal delivery of the potent anti-VEGF, RBX.

**Supplementary Materials:** The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/pharmaceutics14071444/s1, Figure S1: Particle size (PS) of (A) G4.5 and (B) G5 complexes along with the empty PAMAM dendrimers G4.5 and G5 (mean ± SD, *n* = 3; ANOVA\* *p* ≤ 0.05); Figure S2: Polydispersity index (PDI) of (A) G4.5 complexes and (B) G5 complexes along with the empty PAMAM dendrimers G4.5 and G5 (mean ± SD, *n* = 3; ANOVA\* *p* ≤ 0.05); Figure S3: Drug loading efficiency (DE%) of (A) G4.5 complexes and (B) G5 complexes (mean ± SD, *n* = 3; ANOVA\* *p* ≤ 0.05); Figure S4: Stability studies of RBX in (A) G4.5 complexes and (B) G5 complexes under different storage conditions (mean ± SD, *n* = 3); Table S1: Different storage conditions used to assess the stability of G4.5 and G5 complexes; Table S2: Effect of heat and light on the stability of G4.5 complex 25:1; Table S3: Effect of heat and light on the stability of G5 complex 25:1.

**Author Contributions:** Conceptualization, R.A.A. and I.A.A.; methodology, F.S.A., A.A., F.Y.A., B.A., R.A.A. and N.H.; software, F.S. and W.A.M.; validation, F.S.A., F.Y.A. and H.G.A.; formal analysis, B.A. and N.H.; investigation, F.S., R.A.A. and Q.H.A.; resources, F.S.A.; data curation, W.A.M.; writing original draft preparation, F.S.; writing—review and editing, B.A., F.S.A., and F.Y.A.; visualization, F.S.A.; supervision, I.A.A. and F.S.A.; project administration, I.A.A.; funding acquisition, F.S.A. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research project was supported by Researchers Supporting Project number RSP-2021/340, King Saud University, Riyadh, Saudi Arabia.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Acknowledgments:** Authors are thankful to the Researchers Supporting Project number RSP-2021/340, King Saud University, Riyadh, Saudi Arabia for supporting this work.

**Conflicts of Interest:** The authors declare no conflict of interest.
