Direct CNS delivery of proteins using thermosensitive liposome-in-gel carrier by heterotopic mucosal engrafting

Abstract

Delivering therapeutics across the blood-brain barrier (BBB) for treating central nervous system (CNS) diseases is one of the biggest challenges today as the BBB limits the uptake of molecules greater than 500 Da into the CNS. Here we describe a novel trans-nasal mucosal drug delivery as an alternative to the intranasal drug delivery to overcome its limitations and deliver high molecular weight (HMW) therapeutics efficiently to the brain. This approach is based on human endoscopic skull base surgical techniques in which a surgical defect is repaired by engrafting semipermeable nasal mucosa over a skull base defect. Based on endoscopic skull based surgeries, our groups has developed a trans-nasal mucosal rodent model where we have evaluated the permeability of ovalbumin (45 kDa) as a model protein through the implanted mucosal graft for delivering HMW therapeutics to the brain. A thermo sensitive liposome-in-gel (LiG) system was developed for creating a drug depot allowing for a sustained release from the site of delivery to the brain through the implanted nasal graft. We would like to report this as an exploratory pilot study where we are using this novel surgical model to show that the implanted nasal mucosal graft and the LiG delivery system result in an efficient and a sustained brain delivery of HMW proteins. Hence, this study demonstrates that the trans-nasal mucosal engrafting technique could overcome the limitations for intranasal drug delivery and enable the uptake of HMW protein therapeutics into the CNS for the treatment of a wide range of neurodegenerative diseases.

Fig 1. Rat model of heterotrophic mucosal engrafting technique.

(A) A graphical representation of trans-nasal mucosal rodent model with implantation of nasal graft over the craniotomy for delivering high molecule weight proteins to the brain. (B) 3mm craniotomy was outlined using a surgical drill at 1.5 mm anterior posterior and -2mm medial-lateral to bregma. (C) The healed implanted mucosal graft 3 days after the engrafting procedure. (D) A propylene reservoir placed and secured over the graft.

Fig 2

Mason’s Trichome staining of the intact mucosal graft over the craniotomy with no sign of infections at (A) Day 3 and (B) Day 7.

Fig 3. Qualitative and Quantitative uptake of cy5-labeled ovalbumin and ovalbumin in saline in rat brain using mucosal engrafting technique.

(A) Imaging cytometry quantified data from the four selected regions for untreated and Cy5-labeled ovalbumin in saline treated rats (n = 1) using equation total Cy5 intensity = area of Cy5 in selected region in each section * total Cy5 intensity in each selected region in each section. (B) Rat brain cut into 4 equal parts for ELISA with part 3 being the craniotomy site. (C) Ovalbumin detected by ELISA for treatment groups: Untreated (n = 2), ovalbumin saline 48 hours (n = 2), ovalbumin saline 72 hours (n = 2) in the 4 isolated parts of the brain. A 2-way 3 x 4 between and within subjects Anova indicated a significant overall interaction (**** p<0.0001, F (6,9) = 35.75) between different parts (4 brain parts) of the brain and the different treatment groups (untreated, ovalbumin in saline 48 and 72 hours). Post hoc tukey’s tests were further applied to determine differences in ovalbumin uptake in different brain parts. Significant group effects were only found in part 3 where Ovalbumin in saline 48 hours showed significantly greater uptake as compared to untreated rats (** p = 0.0011, DF = 3) and ovalbumin in saline 72 hours(** p = 0.0012, DF = 3). (D) Total ovalbumin (four parts combined) detected by ELISA for treatment groups: Untreated ovalbumin saline 48 hours, ovalbumin saline 72 hours. Ordinary between subjects one way Anova with Tukey’s tests was applied. Ovalbumin in saline (48 hours) showed significantly greater uptake as compared to untreated (*** p = 0.0009, DF = 3) and ovalbumin in saline (72 hours) (** p = 0.0011, DF = 3). A sample size of n = 2 was used for each treatment group.

Fig 4. Transmission electron microscopy images of formulated liposomes.

(A) Anionic Cy5-labeled ovalbumin liposomes (B) cationic Cy5-labeled ovalbumin liposomes (C) and cationic unlabeled ovalbumin liposomes (D) A 30% (w/v) thermo-sensitive Pluronic F-127 solution at 4°C and gel at room temperature.

Fig 5. Quantitative uptake of ovalbumin cationic LiG in rat brain using mucosal engrafting technique.

(A) ICYTE quantified data from the four selected regions for untreated, Cy5- labeled ovalbumin in Pluronic F-127 gel, Cy5-labeled ovalbumin anionic LiG and Cy5-labeled ovalbumin cationic LiG (n = 1) using equation total Cy5 intensity = area of Cy5 in selected region in each section * total Cy5 intensity in each selected region in each section. (B) Ovalbumin detected by ELISA for treatment groups: Untreated (n = 2), ovalbumin LiG 48 hours (n = 2), ovalbumin LiG 72 hours (n = 2) in the 4 isolated parts of the brain. A 3 x 4 between and within subjects 2 way Anova was applied and a significant overall interaction (**** p<0.0001, F (6,9) = 20.15) was found between different parts (4 brain parts) of the brain and the different treatment groups (untreated, ovalbumin in LiG 48 and 72 hours). A post hoc tukey’s test was further applied to determine difference in ovalbumin uptake in different brain parts. Ovalbumin LiG 48 hours shows significant uptake in part 2 as compared to untreated (* p = 0.0154, DF = 3) and ovalbumin LiG 72 hours (* p = 0.0217, DF = 3). Ovalbumin LiG 48 hours also shows significant uptake in part 4 as compared to untreated (** p = 0.0028, DF = 3) and ovalbumin LiG 72 hours (**p = 0.0059, DF = 3). (C) Total Ovalbumin detected (four parts combined) by ELISA for treatment groups: Untreated (n = 2), ovalbumin LiG 48 hours (n = 2), ovalbumin LiG 72 hours (n = 2). Ovalbumin LiG (48 hours) shows significant uptake as compared to untreated (** p = 0.0010, DF = 3) and ovalbumin LiG (72 hours) (** p = 0.0022, DF = 3). (D) Ovalbumin LiG (72 hours) shows significant uptake as compared to untreated (** p = 0.0065, DF = 3) and ovalbumin in saline (72 hours) (* p = 0.0147, DF = 3). A sample size of n = 2 was used for each treatment group.

Fig 6. Diffusion of Cy5-labeled ovalbumin in saline and LiG through human nasal mucosal tissue for up to 3 hours.

A 2x6 multifactor repeated measures anova design was applied and a significant overall interaction between the timepoints and vehicles (cy5-labeled ovalbumin in saline and LiG) was obtained (* p<0.05, F (5,5) = 8.349). Post hoc paired T tests were applied to determine the difference between vehicles at each time point. No significant difference was found.A sample size of n = 2 was used for each treatment group.