Pharmacokinetics of Pullulan-Dexamethasone Conjugates in Retinal Drug Delivery

Abstract

The treatment of retinal diseases by intravitreal injections requires frequent administration unless drug delivery systems with long retention and controlled release are used. In this work, we focused on pullulan (≈67 kDa) conjugates of dexamethasone as therapeutic systems for intravitreal administration. The pullulan-dexamethasone conjugates self-assemble into negatively charged nanoparticles (average size 326 ± 29 nm). Intravitreal injections of pullulan and pullulan-dexamethasone were safe in mouse, rat and rabbit eyes. Fluorescently labeled pullulan particles showed prolonged retention in the vitreous and they were almost completely eliminated via aqueous humor outflow. Pullulan conjugates also distributed to the retina via Müller glial cells when tested in ex vivo retina explants and in vivo. Pharmacokinetic simulations showed that pullulan-dexamethasone conjugates may release free and active dexamethasone in the vitreous humor for over 16 days, even though a large fraction of dexamethasone may be eliminated from the eye as bound pullulan-dexamethasone. We conclude that pullulan based drug conjugates are promising intravitreal drug delivery systems as they may reduce injection frequency and deliver drugs into the retinal cells.

Keywords: conjugate; dexamethasone; ocular fluorophotometry; optical coherence tomography; pharmacokinetics; pullulan; retinal drug delivery.

Figure 1

Chemical structures of fluorescently labelled pullulan conjugates. Synthetic details can be found in a recent publication [26] and in Supporting Information SI-1.

Figure 2

Scheme of the kinetic simulation model. The following parameters were used: K_r (first-order release rate constant); K_pD (elimination rate constant of free dexamethasone posteriorly from the vitreous); K_aD (distribution rate constant of free dexamethasone from the vitreous to the anterior chamber); K_vP (distribution rate constant of pullulan–dexamethasone from the vitreous to the anterior chamber); K_aqhP (elimination rate constant of pullulan–dexamethasone from the anterior chamber) and K_aqhD (elimination constant of dexamethasone from the anterior chamber). For detailed parameter values, see Supporting Information SI-4.

Figure 3

Concentrations of fluorescently labelled (A) BDP-pullulan (n = 6 eyes) and (B) BDP-pullulan-DEX (n = 5 eyes) in the vitreous of rats. One compartment model with first-order elimination rate constant was used for curve fitting (lines). The derived kinetic parameters are shown in Table 2. Each line represents the measurement from individual rat eye.

Figure 4

Full color and green fluorescent fundus images of rat eyes before and after intravitreal injection (IVT) of (A) BDP-pullulan and (B) BDP-pullulan-DEX. After one day, the labeled compounds distribute homogeneously in the vitreous. The length of scale bar is 200 µm.

Figure 5

Concentrations of fluorescently labelled BDP-pullulan-DEX in the vitreous and aqueous humor of six rabbit eyes. Each line was fitted for the experimental data of one eye at different time points. One compartment model with first-order elimination rate constant was used for curve fitting.

Figure 6

Simulated and experimental concentrations of BDP-pullulan-DEX in (A) rat, (B) rabbit and (C) human. The blue and red line shows the simulated concentration in vitreous and aqueous humor, respectively. The experimental concentrations in the vitreous and aqueous humor are indicated in the graphs.

Figure 7

Maurice plot of intravitreally administered compounds in the rabbit eyes. The plot shows anteriorly eliminating compounds based on literature data: sucrose (0.342 kDa) [51], FITC-dextran (FD-10.5, FD-67 and FD-157 kDa) [52,53]. The green, red and blue line are derived from Maurice equation by assuming 932, 1150 and 1700 µL as vitreal volume of distribution. Location of BDP-pullulan-DEX NP (~75 kDa) data at close vicinity of the straight lines indicates anterior route of elimination in the rabbit eyes as the main elimination pathway.

Figure 8

Simulation of released dexamethasone concentration in the vitreous and in the aqueous humor after intravitreal injection of BDP-pullulan-DEX in rat (A), rabbit (B) and human (C). The simulated doses of BDP-pullulan-DEX were 30, 500 and 500 µg per eye for rats, rabbits and humans, corresponding to DEX doses of 3, 50 and 50 µg per eye for rats, rabbits and humans, respectively. The dotted line (- - -) shows the minimum effective intravitreal concentration of dexamethasone for inhibiting the expression of VEGF, which is 1 nM (or 0.394 ng/mL) [54].

Figure 9

Ex vivo retinal explants of mice were treated with 15 μL of fluorescently labelled Cy3-pullulan-DEX (0.7, 1.4 and 1.9 mg/mL) and Cy3-pullulan (1.7 mg/mL). Untreated retina (NT) was used as control explant. TUNEL-positive nuclei in (A) the inner nuclear cell layer (INL), and (B) the outer nuclear cell layer (ONL) were counted and plotted as percentage of all nuclei in the INL and ONL areas. The number of cell rows in (C) INL and (D) ONL are also presented. Bars indicate standard deviations of means. One-way ANOVA **** p < 0.0001.

Figure 10

Microglial cells in sections of the ex vivo mouse retina labelled with an antibody against Iba-1 (green). Untreated retina, retina treated with Cy3-pullulan-DEX (15 μL, 1.9 mg/mL) and Cy3-pullulan (15 μL, 1.7 mg/mL) are shown. Cy3 fluorescence is shown as red. Nuclei were stained with DAPI (blue). RG: overlay of red and green channels for microglial cell, nanoparticle or conjugate colocalization. RGB: overlay of red, green and blue channels. Bar size: 20 μm.

Figure 11

Fundus and optical coherence tomography (OCT) images of rat vitreous and retina. The figures are before and after intravitreal injections (1, 3 and 5 days) of BDP-pullulan (A) and BDP-pullulan-DEX (B). The length of scale bar for fundus images are 200 µm. In the case of OCT, vertical and horizontal scale bars in OCT are 110 and 130 µm, respectively.

Figure 12

Müller glial cells in sections of the ex vivo mouse retina labelled with an antibody against glutamine synthetase (green). Untreated retina, retina treated with Cy3-pullulan-DEX (15 μL, 1.9 mg/mL) and Cy3-pullulan (15 μL, 1.7 mg/mL) are shown. Cy3 fluorescence is shown as red. Nuclei were stained with DAPI (blue). RG: overlay of red and green channels for Müller glial cell, nanoparticle or conjugate colocalization. RGB: overlay of red, green and blue channels. Bar size: 20 μm.

Figure 13

Retinal distribution of BDP-pullulan-DEX (100 μL, 5 mg/mL) in the vitreo-retinal ex vivo bovine explant 24 h after intravitreal injection. Representative confocal microscopy images of cryosections display the penetration of BDP-pullulan-DEX in the retinal layers (green). The inner limiting membrane (ILM) was labelled with rabbit anti-collagen type IV antibody (Col-IV ab, red). The vitreous can be seen as transparent layer in transmission imaging which is well aligned along the ILM while it appears in bright green color due to the high load of BDP-pullulan-DEX in merged channel mode. Nuclei are stained with Hoechst (blue). The white bar on the right bottom corner of each picture indicates the bar size: 50 μm. Abbreviations indicated in image: ganglion cell layer (GCL), inner plexiform layer (IPL), inner nuclear layer (INL), outer plexiform layer (OPL), outer nuclear layer (ONL), overlay of red/green/blue channels (RGB).

Figure 14

Images of retinal sections from two-months old mice after intravitreal injection (1 μL) of 5 mg/mL Cy3-pullulan-DEX (cyanine3, red). Müller glial cells were labelled with an antibody against glutamine synthetase (green). Nuclei were stained with DAPI (blue). RG: overlay of red and green channels for Müller glial cell and nanoparticle colocalization. RGB: overlay of red, green and blue channels. Bar size: 20 μm.