Evaluation of ocular tolerability and bioavailability of tonabersat transfersomes ex vivo

Abstract

While transfersomes (TFS) have extensively been investigated as carriers for topical drug delivery to the skin, their application in ocular drug delivery remains largely unexplored. This study aimed to evaluate the tolerability, contact angle, and ocular penetration of tonabersat-loaded TFS using ex vivo models, with a focus on comparing drug distribution in different ocular tissues. A solution of tonabersat in medium chain triglycerides (MCT) was used as the control. Conjunctival tolerability was evaluated using the Hen’s Egg Test on Chorioallantoic Membrane (HET-CAM), while the Bovine Corneal Opacity and Permeability (BCOP) assay was used to establish corneal tolerability. Drop contact angle on freshly excised bovine corneas was measured using a goniometer. Drug penetration into the cornea, conjunctiva, eyelid and sclera-choroid was evaluated using an ex vivo porcine whole eye model under simulated tear flow, 0.25, 0.5, 1, and 2 h after eyedrop application. Both the TFS and MCT formulations exhibited good conjunctival and corneal tolerability with the TFS contact angle on the corneal surface being lower than that of MCT. Significantly greater drug concentrations were achieved in all ocular tissues with the TFS eyedrop, with the C_max from TFS being at least 16-fold higher than that achieved with the MCT solution in the conjunctiva, eyelid and sclera-choroid, with the difference being greatest in the latter. Meanwhile, the corneal C_max was 6-fold greater with TFS. Interestingly, despite simulated tear flow, the T_max was observed at a later timepoint with TFS in all ocular tissues. Overall, this study demonstrates that TFS are well tolerated on the ocular surface and have the potential for sustained and targeted drug delivery to ocular tissues. Thus, they present a promising alternative for safe and effective ocular drug delivery.

Introduction

Topical eyedrops typically result in low drug bioavailability, primarily due to the anatomical tissue barriers and ocular lacrimation leading to rapid drainage and subsequent systemic absorption. Typically, less than 5% of a topically applied drug reaches the site of action [1, 2]. Therefore, in the clinical setting, frequent instillation of eyedrops may be required to achieve the desired therapeutic efficacy, which may result in poor patient compliance and increased toxicity due to overexposure to eyedrop excipients, such as preservatives and surfactants [3]. Several formulation approaches, such as drug-loaded contact lenses, in-situ gels, polymeric micelles, nanoparticles, microemulsions, and liposomes, have been trialled to overcome these challenges [3–7]. Among these, liposomes, due to their ability to encapsulate both hydrophilic and lipophilic drugs, have received significant research interest since they can protect the drug from enzymatic degradation in the tear film, prolong release and enhance drug penetration [8, 9]. However, due to their rigid structure, liposomes typically remain confined in the superficial cell layers as they are unable to penetrate through small intracellular pores without rupture [10]. This has led to the development of more flexible vesicles called transfersomes (TFS) [11, 12].

TFS are ultra-elastic vesicles containing edge activators (EA), such as Tween 80 and Span 80, which significantly reduce the elastic modulus of the phospholipid backbone, rendering them deformable [13]. The stress-responsive deformable nature of TFS allows them to penetrate the stratum corneum as they can squeeze through small intercellular pores without rupture [14]. Moreover, the exceptional surface hydrophilicity of TFS plays a crucial role in their ability to navigate through the skin layers, allowing them to move along osmotic gradients from the lipid-rich epidermis to the well-vascularized hydrophilic dermis via “transpore hydrotaxis” [15–17]. This unique characteristic of TFS has made them the subject of extensive research in transdermal drug delivery.

Although the mechanism of drug penetration into the skin and eyes is often assumed to be similar, ocular tissues differ considerably from the skin in structure and function. As such, despite significant research into TFS as carriers for transdermal drug delivery, research on their ocular applications remains limited. The ocular surface is more vulnerable to excipient mediated toxicity than the skin, therefore, safety and tolerability of TFS on the ocular surface need to be elucidated. Moreover, most studies performed so far, have only focussed on the corneal route of penetration [18–20], neglecting the conjunctival-scleral pathway, which serves as a significant alternative route for ocular drug delivery. As each ocular tissue exhibits unique structural properties and osmotic gradients [21–24], it is hypothesized that TFS penetration and distribution will vary substantially between the cornea, conjunctiva, and sclera. Investigating these differences is crucial to understanding the ocular penetration mechanism of TFS and optimizing them for more efficient and targeted ocular drug delivery.

Tonabersat (molecular weight 391.8 g/mol and log P 2.9) [25] is a small molecule synthetic hydrophobic drug (solubility in water and medium chain triglycerides around 0.04 mg/mL and 7 mg/mL, respectively) that shows significant promise as a novel therapeutic agent for ocular inflammatory conditions, such as dry eye disease and uveitis, owing to its unique mechanism as a connexin43 hemichannel blocker [26–28]. By inhibiting these channels, tonabersat disrupts the release of pro-inflammatory mediators and the activation of the NLRP3 inflammasome, both of which are key drivers of ocular inflammation [26–28]. The efficacy of tonabersat in retinal inflammatory disorders has already been established after oral administration [26, 29]. It is hypothesized that tonabersat may also have application in the management of chronic inflammation in ocular surface diseases, such as dry eye disease. However, its poor water solubility renders it difficult to formulate as an eyedrop suitable for treating ocular surface inflammation. At present, no commercial formulation of tonabersat is available for ocular application. We have previously developed and optimized tonabersat-loaded TFS eyedrops and demonstrated their stability and ability to enhance ocular penetration in a static ex vivo model [30]. However, since tissue properties, hydration and lacrimal flow can significantly influence osmotic gradients and thus TFS penetration, this study aimed to evaluate and compare the ocular pharmacokinetics of tonabersat-loaded TFS into the cornea, conjunctiva, eyelid and sclera-choroid over 2 h using an ex vivo whole-eye model with simulated tear flow. In addition, ocular tolerability and eyedrop contact angle on the cornea were also evaluated ex vivo to further elucidate the tissue penetration mechanism.

Materials and methods

Materials

Tonabersat was purchased from MedChemExpress (USA). Medium-chain triglycerides (MCT Labrafac Lipophile WL 1349) and Phospholipon^® 90G were gifted by Gattefosse and Lipoid GmbH, Germany, respectively. Tween^® 80, chloroform, methanol and gradient HPLC-grade acetonitrile were purchased from Merck, Germany. Milli-Q Water was obtained by reverse osmosis using a Millipak system (0.22 μm; Millipore, USA). All other chemicals, including phosphate-buffered saline (PBS) tablets (Oxoid Ltd., UK), benzalkonium chloride (Sigma-Aldrich, USA), sodium chloride (Fischer Scientific, UK), calcium chloride dihydrate (VWR chemicals, USA), urea (Carl Roth, Germany), D-glucose monohydrate (Merck, Germany), potassium chloride (Geyer, Germany), sodium carbonate anhydrous (Fischer Scientific, UK), trisodium citrate dihydrate (Fischer Scientific, UK), sodium phosphate tribasic (Thermo Fischer, Germany), potassium hydrogen carbonate (Honeywell, Germany) and hydrochloric acid (Merck, Germany) were of reagent grade and were used as received.

Transfersome preparation

TFS were prepared using the thin film hydration technique described previously [30]. Phospholipon 90 G (180 mg), Tween 80 (20 mg), and tonabersat (10 mg) were dissolved at a ratio of 9:1:0.5 w/w in 10 mL of the organic phase (chloroform: methanol, 2:1 v/v) and the mixture was transferred to a round-bottom flask. The organic phase was evaporated at 60 °C under a vacuum (> 320 mbar) and a rotation speed of 70 rpm using a rotary evaporator (Rotavapor R-215, Buchi, Switzerland). The dried lipid film was then rehydrated with 10 mL of PBS (pH 7.4) containing 0.3 mg of benzalkonium chloride and homogenized for 5 min in an ice bath using a handheld homogenizer (ULTRA-TURRAX^®, IKA Werke GmbH & Co., Germany). The final dispersion containing 1 mg/mL of tonabersat was then extruded ten times using polycarbonate membranes (200 nm, Sigma-Aldrich, New Zealand).

Conjunctival tolerability

Conjunctival tolerability was evaluated using the modified Hen’s Egg Test on Chorioallantoic Membrane (HET-CAM) as previously described by Alany et al. [31]. Simulated tear fluid (STF), prepared as described previously [32], was used as the negative control, while 0.1 M sodium hydroxide and 70% ethanol were used as the positive controls. Fertilized Shaver Brown hen’s eggs (Bromley Park Hatcheries Ltd., Auckland, New Zealand) were incubated at 37 ± 0.5 °C and 55 ± 5% RH. On day 3, the eggs were cracked open into custom-made growing chambers and returned to the incubator for an additional 7 days (10 days in total). On day 10, non-viable embryos were discarded, and only healthy, well-developed embryos with highly vascularized chorioallantoic membranes (CAM) were used, applying 0.2 mL of each test formulation and the controls. Any signs of irritation, such as hyperaemia, haemorrhage, and coagulation, were recorded at 0.5, 2, and 5 min and the cumulative irritation potential score was calculated according to Luepke [33, 34]. All experiments were performed in triplicate.

Corneal tolerability

The Bovine Corneal Opacity and Permeability (BCOP) assay was carried out according to the Organisation for Economic Co-operation and Development (OECD) guidelines to identify the irritation potential of cosmetics and pharmaceuticals [35–37]. Freshly harvested bovine eyes were collected from a local slaughterhouse (Auckland Meat Processors Ltd., New Zealand) and examined for epithelial detachment, corneal opacity, and vascularization. Only undamaged eyes were selected for the study. The eyes were placed in square plastic cups and incubated at 32 ± 0.5 °C for 10 min. Custom-made polycarbonate sleeves were placed on the centre of the corneas, and a drop of STF was added. The eyes were incubated for 5 min, and 150 µL of the test formulations were applied, including the negative (STF) and positive (0.1 M sodium hydroxide and 70% ethanol) controls. After 5 min of exposure to the test samples, the corneal sleeves were removed, and the eyes were rinsed with 10 mL of warm PBS (32 ± 0.5 °C). The eyes were incubated for 1 min before evaluating corneal damage, such as corneal opacity and epithelial detachment. Subsequently, 150 µL of sodium fluorescein solution (4 mg/mL; pH 7.4) was applied to the corneas to assess the integrity of the corneal epithelium using a Burton lamp (G. Nissel & Co. Ltd., UK). Observations were documented using individual numerical scores for corneal opacity, corneal fluorescein staining (reflecting epithelial integrity), and epithelial detachment. The mean cumulative score (n = 3) was calculated to determine the irritation potential, following a previously established methodology [36, 38].

Contact angle

The eyedrop contact angle on the ocular surface was evaluated by adapting the method previously described by Agarwal et al. [39]. The contact angle of TFS and MCT was measured on freshly excised bovine corneas using a goniometer (Ossila, UK). Eyes were collected from Auckland Meat Processors Ltd. (New Zealand) and transported in an iced bin. Corneas were scrutinized for any damage and only intact corneas were excised and laid flat on a glass slide after making four radial slits. A drop of STF was first applied to the corneas, followed by a drop of the test formulations, dispensed from a height of 150 mm using a 25-gauge needle. The contact angle was measured automatically using the Ossila contact angle software v3.09.1 (Ossila, UK). All experiments were performed in triplicate. The contact angles of TFS and MCT were statistically compared by ordinary one-way ANOVA followed by Tukey’s multiple comparisons test using GraphPad Prism 9.3.1 with a p-value of 0.05 or less being considered significant. Each test group included three samples (n = 3), and each sample was analyzed in duplicate to ensure reliability and reproducibility. The selected sample size was based on previous literature and standard practices in similar experimental models, which have demonstrated adequate sensitivity for detecting statistically significant differences [39].

Ocular drug penetration

Ocular drug penetration from TFS containing 1 mg/mL tonabersat was evaluated using an ex vivo porcine whole-eye model with simulated tear flow [40]. A 1 mg/mL solution of tonabersat in MCT was used as the control. Fresh porcine eyes were collected from a local abattoir (Dahua Supermarket, New Zealand) and transported to the laboratory in PBS in an iced bin. All eyes were observed for corneal or scleral damage, and only intact eyes were selected for the study. Excessive fatty tissues were removed using a scalpel and scissors, and eyes were prepared for the study.

The porcine eye model was set up by placing the eyes into a twelve-well microplate. Eyes were oriented to expose the cornea and bulbar conjunctiva to the test formulation. Excised eyelid tissue was placed on top of each eye, with the tarsal conjunctiva facing the cornea. Careful measures were taken to prevent leakage between the eye globe and the donor compartment using customized 3D-printed plates and screws. STF (50 µL/min) was added between the tarsal conjunctiva and cornea using a peristaltic pump and the entire set-up (except the pump) was placed at 35 ± 3°C in a shaking water batch (Jeio Tech, South Korea). The rationale for selecting 35 ± 3°C is to mimic the ocular surface temperature [41]. An STF flow rate of 50 µL/min was chosen corresponding to approximately 16.7% of the tear turnover rate for a 300 µL dose [40]. These parameters help ensure that the experimental setup yields more relevant and translatable data on drug penetration across ocular tissues. After equilibration for 10 min, 300 µL of the test formulations (n = 6) was applied between the tarsal conjunctiva and the cornea. At pre-determined timepoints (0.25, 0.5, 1 and 2 h after formulation application; six eyes per formulation and per timepoint), the eyes were removed and thoroughly rinsed with PBS to remove any excess formulation. The eyes were carefully dissected and separated into the cornea, conjunctiva and sclera-choroid. All tissues, including the eyelid, were weighed and stored at -20 °C until further analysis by HPLC.

For HPLC analysis, the drug was extracted from the tissues as described previously [42]. The tissues were thawed and homogenized in 500 µL of Milli-Q water using a Precellys Evolution Homogenizer (Bertin Technologies, France). Homogenization consisted of 6 × 20 s at 6800 rpm followed by 6 × 30 s at 5000 rpm, each with a 60-s pause. After homogenization, 500 µL of chilled acetonitrile was added to the samples, which were then sonicated for 20 min at 5 ± 3 °C, followed by centrifugation (Eppendorf, Germany) at 13,000 rpm for 0.5 h at 4 °C. The clear supernatant was then assayed for tonabersat using the stability-indicating HPLC-UV method detailed previously [43]. Subsequently, the tonabersat concentration at each timepoint was calculated as the amount of tonabersat recovered (µg) per tissue weight (g). Tonabersat penetration in each tissue was compared by plotting the tonabersat concentration against time and determining the mean AUC over 2 h (AUC_(0–2 h)). The tissue concentrations obtained with each formulation at each timepoint were statistically compared using two-way ANOVA followed by Šídák’s multiple comparisons test. C_max and AUC (0–2 h) were statistically compared using paired and unpaired t-tests, respectively. All statistical analyses were performed using GraphPad Prism 9.3.1 software, and a p-value of 0.05 or less was considered significant. Each test group included six samples for each tissue (n = 6), and each tissue sample was analyzed in duplicate to ensure reliability and reproducibility. While a formal power analysis was not conducted, the chosen sample size was based on previous literature and standard practices in similar experimental models, which have demonstrated adequate sensitivity for detecting statistically significant differences [13].

Results and discussion

Transfersome preparation

The detailed physicochemical characterization of TFS and MCT formulations, including particle size, polydispersity index, zeta potential, and morphology, has been reported in our previous publication [30]. Briefly, TFS exhibited a mean particle size of 126 ± 1 nm with a PDI of 0.16 ± 0.01, representing a narrow size distribution and neutral zeta potential of 0 mV. Morphological analysis using transmission electron microscopy confirmed the presence of spherical, well-defined vesicles. In contrast, the MCT formulation was a clear solution.

Conjunctival tolerability

The HET-CAM test is a valuable alternative for assessing the conjunctival irritation potential of the developed TFS because it is versatile, sensitive and allows for the testing of both hydrophilic and hydrophobic substances with good in vivo correlation [44]. As expected, STF (negative control), MCT, and TFS eyedrops did not result in any vascular response, such as hyperaemia, haemorrhage or coagulation (Fig. 1A). Meanwhile, both 0.1 M NaOH (cumulative score: 19 ± 2, severely irritant) and 70% ethanol (cumulative score: 14 ± 1, severely irritant) elicited significant vascular responses (hyperaemia, haemorrhage and coagulation). Our results are consistent with the findings of Agarwal et al. [36], who used the HET-CAM method to test the conjunctival tolerability of semifluorinated alkanes. It has to be noted that, although the HET-CAM test is an ethical and efficient alternative to traditional animal-based methods for assessing the ocular irritation potential of eyedrops, it has some limitations, such as it cannot show important in vivo responses such as pain, blinking, tissue healing, or the effects of long-term use.

Fig. 1.

Ex vivo tolerability of MCT and TFS. (A) Assessment of conjunctival tolerability using the HET-CAM method: No vascular responses were observed upon application of STF, MCT and TFS but significant hyperaemia, haemorrhage and coagulation were observed with the positive controls, 0.1 M NaOH and 70% ethanol. (B) Assessment of corneal tolerability using the BCOP assay: No opacity, corneal staining or epithelial detachment was observed upon application of STF, MCT and TFS, but considerable opacity, corneal staining and epithelial detachment were observed in eyes treated with the positive controls, 0.1 M NaOH and 70% ethanol (STF: simulated tear fluid; MCT: Medium chain triglycerides, TFS: Transfersomes)

Corneal tolerability

The OECD recommends the BCOP assay as an alternative to the Draize test for the evaluation of corneal tolerability of test formulations [45, 46]. Compared to other alternative ex vivo tests, the BCOP assay directly measures corneal damage by assessing opacity (tissue integrity), permeability (barrier function), and epithelial detachment, mimicking key in vivo responses [47]. Corneal edema, vacuolation and protein denaturation are typically observed as increased corneal opacity while epithelial compromise can be observed by fluorescein staining [13]. In the BCOP assay, the application of STF (negative control), MCT and TFS did not result in any opacity, sodium fluorescein staining, epithelial detachment or stromal edema (cumulative score: 0.0, practically non-irritant), suggesting that both formulations are biocompatible and practically non-irritant on the corneal surface. In contrast, treatment with the positive controls resulted in significant fluorescein staining, corneal opacity, and epithelial detachment, which corresponded to cumulative scores of 4.5 for 0.1 M NaOH and 8.5 for 70% EtOH, indicating severe irritation (Fig. 1B). Microscopy of the corneal sections showed complete detachment of the corneal epithelium with significant stromal vacuolation and edema in corneas exposed to 0.1 M NaOH. Edema and epithelial detachment were also observed in corneas exposed to 70% ethanol. Similar results were previously observed with curcumin loaded Phospholipon 90G and Tween 80 TFS, which caused no changes in the corneal surface morphology [13]. While the BCOP assay is effective for identifying substances that cause severe corneal damage, the test often produces false negatives for less irritating compounds, therefore while useful for initial screening further additional in vivo safety studies are needed to fully evaluate the long-term safety and tolerability of TFS formulations.

Contact angle

The angle of contact between test formulations and bovine cornea is illustrated in Fig. 2, with the contact angle of MCT being relatively greater than that of TFS. While the corneal epithelial surface is hydrophobic, it is always covered by a liquid film (tear film), which provides a new hydrophilic surface for the interaction of eyedrops [48, 49]. The tear film comprises two layers, a hydrophilic mucoaqueous (more than 90% of total tear film volume) and a hydrophobic lipid layer [49–52]. The hydrophilic mucoaqueous layer is more compatible with water-based (hydrophilic) formulations, allowing them to spread more easily, resulting in a lower contact angle [48]. Moreover, compared to the lipophilic MCT formulation, the TFS water-based formulation has a lower surface tension on the hydrophilic corneal surface, enabling better wetting and spreading [48]. In addition, the excipient or vesicle surface charges present in the formulation also play a vital role. At physiological pH, the corneal epithelium behaves as a negatively charged membrane [53], which repels negatively charged formulations and attracts neutral or positively charged formulations [54]. Previously, Lallemand et al. [55] performed static and dynamic contact angle measurements and surface tension studies on harvested rabbit eyes and showed that the Novasorb cationic emulsion had a better spreading coefficient on the cornea and conjunctiva than anionic eyedrops. As previously described, the developed TFS suspended in PBS carry a neutral to positive surface charge [30], thus spreading efficiently on the corneal surface. On the other hand, MCT shows a relatively higher contact angle likely due to its higher surface tension and viscosity. The smaller contact angle of the TFS formulation is beneficial for drug delivery as it improves ocular surface wettability, reducing tear washout and ultimately enhancing drug penetration [55].

Fig. 2.

Ex vivo corneal spreading dynamics of TFS and MCT using bovine corneas. (A) Schematic illustration of the contact angle measurement. (B) Graphical representation of contact angle measurements (mean ± SD; n = 3); ***p˂0.001. (TFS: Transfersomes, MCT: Medium chain triglycerides). TFS showed a lower contact angle than the control (MCT), indicating better wettability

Ocular drug penetration

Ex vivo tissue penetration studies showed that TFS significantly improved tonabersat penetration in all ocular tissues investigated (Fig. 3), with the C_max and AUC_(0–2 h) in all tissues being several-fold higher than those of the MCT solution (Table 1). As the primary aim of this study was to evaluate the influence of TFS on drug penetration in different ocular tissues, no gelling agents were included in the vehicle to maintain non-occlusive conditions on the ocular surface and facilitate osmotic gradient driven transport of TFS. Despite the simulated tear flow and absence of gelling agents or viscosity builders, the TFS formulation had a higher T_max than the MCT control, demonstrating its potential for targeted and sustained drug delivery.

Fig. 3.

Tonabersat concentrations after application of a single dose of the TFS and MCT formulations in the (A) cornea, (B) conjunctiva, (C) eyelid, and (D) sclera-choroid for up to 2 h (mean ± SEM; n = 6); *p˂0.05, **p˂0.01, ***p˂0.001, ****p˂0.0001. Ex vivo tissue penetration studies showed that TFS significantly improved tonabersat penetration in all ocular tissues compared to MCT

Table 1.

Ex vivo ocular drug pharmacokinetics (mean ± SEM; n = 6)

Parameter	Tissue	TFS	MCT
Cmax (µg/g)	Cornea	14.6 ± 1.6	2.2 ± 0.3
	Conjunctiva	80.9 ± 8.9	5.0 ± 0.6
	Eyelid	55.6 ± 6.8	3.3 ± 1.0
	Sclera-choroid	5.6 ± 0.7	0.3 ± 0.2
Tmax (h)	Cornea	1.0	0.25
	Conjunctiva	2.0	0.25
	Eyelid	2.0	0.25
	Sclera-choroid	0.5	0.25
AUC(0–2 h) (µg.h/g)	Cornea	21.3 ± 3.1	0.6 ± 0.1
	Conjunctiva	101.9 ± 12.5	2.7 ± 1.3
	Eyelid	86.6 ± 11.1	4.9 ± 1.6
	Sclera-choroid	7.4 ± 1.2	0.1 ± 0.1

In the cornea, the TFS resulted in a statistically significant improvement in C_max (approximately 6.6-fold) and AUC_(0–2 h) (35.5-fold) over the control (Table 1). Previously, Janga et al. [56] too observed significantly higher corneal penetration of natamycin in vivo from TFS than from free drug in suspension. The corneal epithelium, comprising 5–7 layers of epithelial cells connected by tight epithelial junctions, is the major barrier to penetration of topically applied drugs [57]. The increased penetration observed with TFS may likely be attributed to their stress-responsive deformable nature which enables them to squeeze through the tight epithelial junctions of the corneal epithelium. Our results are in alignment with previous studies performed by Barbalho et al. [13] who compared the penetration of curcumin from TFS and MCT in different corneal layers. They reported that with MCT, the drug was primarily localised in the superficial layers of the epithelium due to MCT’s tendency to form a thin film on the ocular surface, for which lipophilic drugs typically have greater affinity than for the corneal epithelium [13, 58]. Meanwhile, TFS not only facilitated greater drug penetration but also ensured that the drug was uniformly distributed through the epithelial layers. No significant drug penetration was observed beyond the epithelial layers at the 0.25 h timepoint tested by Barbalho et al. [13], likely due to the tendency of lipophilic drugs to form a depot in the corneal epithelium, from which they are distributed into the hydrophilic stroma in a more sustained fashion [59–61]. It should be noted, however, that curcumin (BCS class IV drug) is known for its poor bioavailability due to its inherently poor penetration capacity [62], which may not necessarily be the case for tonabersat, which is a BCS class II drug [28, 63]. In the present study, the T_max was observed at 1 h in the cornea, after which the corneal concentration appeared to reduce slightly, likely due to drainage of the TFS eyedrop from the ocular surface by the simulated tear flow mechanism. Considering that drug penetration in the retina (0.37 ± 0.25 µg/g; data not shown) was only observed at the 2 h timepoint, we believe that the slight reduction in corneal concentration at 2 h is due to drug diffusion and distribution from the cornea into deeper ocular tissues.

In the conjunctiva and the eyelid (lined by the tarsal conjunctiva), tonabersat penetration from the TFS appeared to increase steadily over the 2-h test period. TFS resulted in an approximately 16.2- and 16.8-fold higher C_max and a 37.7- and 17.7-fold higher AUC_(0–2 h) in the conjunctiva and eyelid, respectively, than the control with this difference being statistically significant. Conjunctival penetration is relatively unexplored, which is surprising as the conjunctiva offers a considerably larger surface area for drug absorption than the cornea [64, 65]. The conjunctiva is composed of two layers: an outer epithelium and an inner stroma. The conjunctival epithelium is more hydrophilic than the corneal epithelium [66] and the intercellular spaces are approximately 230-fold larger than those in the cornea [66]. As such, it is anticipated to be more amenable to intercellular transport of TFS. This is reflected in the observations made in the present study, where both formulations showed significantly greater drug penetration in the conjunctiva and eyelid than in the cornea. The magnitude of difference between corneal and conjunctival penetration was most pronounced for the TFS, which showed an approximately 5.5-fold higher C_max and a 4.8-fold higher AUC_(0–2 h) in the conjunctiva than in the cornea. In the eyelid, the TFS showed an approximately 3.8-fold higher C_max and a 4.0-fold higher AUC_(0–2 h) than in the cornea. Our results align with previously published data, where significantly higher penetration of hydrophobic curcumin was observed from TFS in the conjunctiva and tarsal conjunctiva (eyelid) than in the cornea [13]. The authors of that study noted that the hydrophilic nature of the conjunctiva allows for a more thermodynamically favourable interaction with the amphiphilic TFS compared to the cornea, thus enhancing penetration. It should be noted, however, that a large fraction of drug absorbed into the conjunctiva may be lost into the systemic circulation via conjunctival blood capillaries and lymphatic vessels, thus reducing ocular bioavailability [67]. While our ex vivo model does not account for this systemic loss and may thus slightly overestimate results, it remains a valuable tool for comparing drug penetration from various formulations and for evaluating how different tissue morphologies influence penetration.

As observed in the cornea, in the sclera-choroid, the drug concentration from TFS increased for the first 0.5 h and then decreased, possibly due to diffusion into deeper tissues. This too may have contributed to drug penetration into the retina by the 2-h timepoint (0.37 ± 0.25 µg/g; data not shown). The penetration enhancing effect of TFS was significant with the C_max and AUC_(0–2 h) being approximately 18.6- and 74.4-fold greater than that observed with the MCT control. Penetration into the sclera occurs through aqueous pores; therefore, the sclera is generally more permeable to hydrophilic molecules than lipophilic ones [60, 68]. This may explain the greater permeation of the amphiphilic TFS than the lipophilic MCT solution. Our findings are consistent with previous studies showing that the natamycin concentrations from TFS and a TFS gel were significantly higher in scleral tissue than those from a simple drug suspension [56]. The permeability of the sclera is similar to that of the corneal stroma, showing no apparent dependence on the distribution coefficient but a strong dependence on molecular radius [69]. Compared to the conjunctiva, TFS exhibited an approximately 14.5-fold lower C_max and 13.7-fold lower AUC_(0–2 h) in the sclera-choroid. Considering that the scleral composition is similar to that of the corneal stroma (more hydrophilic) [69], it may be hypothesized that together the conjunctiva and sclera exhibit similar osmotic gradients as those observed in the cornea and in the skin. Our study suggests that the TFS form a depot in the conjunctiva, followed by sustained drug diffusion driven by transpore hydrotaxis into the sclera, as TFS inherently tend to migrate to more hydrophilic regions to stay hydrated.

In summary, the TFS formulation significantly enhanced ocular penetration of tonabersat showing a higher AUC_(0–2 h), C_max as well as T_max. Our findings suggest that the composition and structure of the TFS formulation significantly enhance penetration through the tight epithelial junctions of the corneal and conjunctival epithelium to achieve targeted drug delivery (Fig. 4). This study provides a comprehensive and dynamic view of drug distribution using TFS over time. Penetration studies at different timepoints allow the real-time monitoring of drug penetration, highlighting changes in distribution patterns and concentrations across the tissues. This study enabled the detection of variability in drug penetration using TFS across different tissues of the eye and observed how penetration adapts to changes in the anatomy and physiology of the different tissues. These findings demonstrate the enhanced penetration potential of the TFS compared to the conventional formulation and highlight the translational relevance of the ex vivo studies. The ex vivo penetration study is a critical intermediate step between in vitro testing and in vivo evaluation. Using both formulations, it directly compares tonabersat penetration across intact ocular tissues under controlled conditions. Although our results offer meaningful comparisons between formulations under controlled conditions, they should be interpreted with caution when extrapolating to in vivo conditions due to the lack of systemic clearance, larger dose volume, absence of blinking mechanism and anatomical differences between the porcine and human eye [70]. These limitations reinforces the model’s role as an indicative screening tool rather than an exact predictor of in vivo bioavailability. Therefore, future studies will be necessary to evaluate the in vivo penetration efficacy as well as the long-term tolerability the TFS.

Fig. 4.

Schematic representation of the corneal penetration of MCT and TFS formulations. The figure only represents the penetration in the cornea, but it would be expected to be similar in the other tissues tested here. (Figure created using BioRender (https://biorender.com/)

Conclusion

Although the penetration enhancing effect of TFS is well characterized in the skin, their application in ocular drug delivery remains largely unexplored. In this study, the tolerability of TFS eyedrops was demonstrated ex vivo. The TFS eyedrop, due to its low surface tension and viscosity, appeared to spread easily on the corneal surface and had good wettability as demonstrated by the small contact angle. Compared to the control, the TFS formulation enabled significantly greater drug penetration into all tested tissues. TFS penetration was highest in the conjunctiva, followed by the eyelid (tarsal conjunctiva), the cornea and the sclera-choroid, respectively. Despite a simulated tear clearance mechanism, sustained drug delivery was observed with TFS, with significant penetration being observed even in deeper ocular tissues, likely due to the deformable nature of the TFS, which allows them to squeeze easily through narrow pores and travel along the osmotic gradient to the interior of the eye. Overall, this study shows that TFS provide a promising avenue for targeted and sustained ocular delivery of hydrophobic drugs in a safe and efficient manner.

Author contributions

All authors contributed to the manuscript. SB: Methodology, Validation, Investigation, Writing – original draft, Writing – review & editing, Visualization. IDR: Conceptualization, Methodology, Writing- review & editing, Resources. PA: Conceptualization, Methodology, Writing- review & editing, Supervision, Resources.

Funding

Open Access funding enabled and organized by CAUL and its Member Institutions

The authors thank the Health Research Council of New Zealand for supporting this project, including SB’s PhD scholarship (project grant 20/317) and PA’s salary (project grants 20/317 and 24/741). IDR’s directorship was funded by the Buchanan Charitable Foundation.

Data availability

The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

Not Applicable.

Consent for publication

The authors declare that they have provided their consent for publishing this manuscript.

Competing interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.