Experimental Eye Research 118 (2014) 42e45
Contents lists available at ScienceDirect
Experimental Eye Research journal homepage: www.elsevier.com/locate/yexer
Short communication
Trans-meningeal drug delivery to optic nerve ganglion cell axons using a nanoparticle drug delivery system Karen Grove a, Julia Dobish b, Eva Harth b, Martha-Conley Ingram c, Robert L. Galloway c, Louise A. Mawn a, * a b c
Vanderbilt Eye Institute, 2311 Pierce Avenue, Nashville, TN 37232-8808, USA Chemistry Department, 7619 Stevenson Center, Vanderbilt University, Nashville, TN 37232, USA Biomedical Engineering, 5824 Stevenson Center, Vanderbilt University, Nashville, TN 37232, USA
a r t i c l e i n f o
a b s t r a c t
Article history: Received 2 June 2013 Accepted in revised form 23 October 2013 Available online 1 November 2013
The purpose of this study was to investigate if neuroprotective drugs can cross the optic nerve sheath in vitro. Four optic nerves were used for this study. Two porcine nerves were harvested at the time of euthanasia and two human nerves were obtained at the time of therapeutic globe enucleation. The optic nerve sheaths were dissected and placed as a membrane in a two chamber diffusion cell to test meningeal penetration by both brimonidine alone and brimonidine encapsulated in nanoparticle (NPbrimonidine). Brimonidine concentration was assayed by UVevis spectrometer measurement of absorbance at 389 nm. Increasing concentration of brimonidine on the receiver side of the chamber was measured in both the brimonidine alone and the brimonidine encapsulated experiments. The human data were fitted with a two parameter exponential regression analysis (brimonidine alone donor r2 ¼ 0.87 and receiver r2 ¼ 0.80, NP-brimonidine donor r2 ¼ 0.79 and receiver r2 ¼ 0.84). Time constant (s) was 10.2 h (donor) and 13.1 h (receiver) in the brimonidine study, and 24.0 h (donor) and 15.9 h (receiver) in the NP-brimonidine study. Encapsulated brimonidine had a longer time to reach equilibrium. Passage of brimonidine through the optic nerve sheath was demonstrated in the experiments. Increase in time constants when comparing the NP-brimonidine with the brimonidine curves in the human studiesindicates that diffusion is delayed by the initial parameter of drug being loaded in NP. Direct treatment of injured optic nerve axons may be possible by trans-meningeal drug diffusion. Ó 2013 Elsevier Ltd. All rights reserved.
Keywords: optic nerve neuroprotection brimonidine drug delivery diffusion nanoparticle human porcine
Nearly all optic neuropathies of varying etiologies including ischemic, traumatic, and Leber’s hereditary optic neuropathy have no therapeutic options (Levin, 2007). Glaucoma is a partial exception, though visual field loss and retinal ganglion cell (RGC) death continue despite well-controlled intraocular pressure (Baltmr et al., 2010). Many neuroprotective strategies are under investigation with goals to protect undamaged RGCs and to rescue their injured counterpart. Many candidate therapies have been successful in animal models of acute and chronic neurodegenerative processes of the optic nerve and tremendous effort continues in pursuit of translating this success (Bessero and Clarke, 2010; Chidlow et al., 2007; Levin, 2007). Brimonidine is of particular interest. Efficacy
* Corresponding author. Tel.: þ1 615 936 1960; fax: þ1 615 936 1540. E-mail addresses: [email protected] (K. Grove), julnmeyer@gmail. com (J. Dobish), [email protected] (E. Harth), martha.e.ingram@gmail. com (M.-C. Ingram), [email protected] (R.L. Galloway), louise.mawn@ vanderbilt.edu (L.A. Mawn). 0014-4835/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.exer.2013.10.016
of neuroprotection has been demonstrated in animal models (Lambert et al., 2011; Saylor et al., 2009). Until recently, clinical trials did not show statistically significant efficacy of brimonidine in the treatment of optic neuropathies (Fazzone et al., 2003; Newman et al., 2005; Wilhelm et al., 2006). However, the Low-Pressure Glaucoma Study Group recently demonstrated less visual field loss in brimonidine-treated patients (9, 9.1%) versus timololtreated patients (31, 39.2% log-rank 12.4, P ¼ .001) in a doublemasked, randomized study. This occurred despite similar intraocular pressure for brimonidine- and timolol-treated patients at all time points. This is consistent with an IOP-independent effect and suggestive of a neuroprotective mechanism (Krupin et al., 2011). Each promising therapy requires a viable route of delivery, a particular challenge regarding central nervous system (CNS) and ocular tissues. Systemic administration of some potential neuroprotective therapeutics is prevented by the side effect profile when high, frequent dosing is required to overcome CNS and ocular barriers (Diebold and Calonge, 2010). The most common options for local treatment include topical drops as well as sub-conjunctival,
K. Grove et al. / Experimental Eye Research 118 (2014) 42e45
sub-tenon’s, and intravitreal injection. However, another route that has not been explored is trans-meningeal drug delivery that especially warrants investigation given that most optic neuropathies are axonal (Levin, 2007). Obstacles to drug delivery to the optic nerve axons include the need for the agent to cross the optic nerve sheath as well as to provide sustained delivery of the neuroprotective drug. The orbital portion of the optic nerve is surrounded by the same meninges as the brain (Miller, 1996). This barrier effectively contributes to the sequestration of its contents from systemic circulation. More specifically, in addition to the bloodebrain barrier provided by the endothelium of brain capillaries, there is an epithelial blood-cerebrospinal fluid (CSF) barrier maintained by the choroid plexuses and the outer arachnoid membrane (Segal, 2000). It is composed of an arachnoid barrier cell layer with numerous tight junctions and a distinct, continuous basal lamina, and is the definitive barrier between the dural circulation and the subarachnoid CSF (Nabeshima et al., 1975; Saunders et al., 2008; Vandenabeele et al., 1996). The parallel development of two unique innovations makes exploration of the trans-meningeal route timely. First, we have developed an image-guided orbital endoscope that facilitates minimally invasive and precise access to the orbital portion of the optic nerve (Atuegwu et al., 2007). Second, our team has created a novel polyester nanoparticle (NP) drug delivery system that allows controlled rate of drug release (Van der Ende et al., 2008). Many NP drug delivery systems for use in ocular disease are under investigation (Diebold and Calonge, 2010), though a trans-meningeal approach to the optic nerve axons has not been studied in any regard. We propose the possibility of a novel route of delivery through the optic nerve sheath. Drug encapsulated in NP can be customized for controlled released over an extended period of time. Therefore, a retrobulbar depot could deliver continuous drug to the RGC axons through the optic nerve sheath. The NP has the potential to facilitate high drug concentration directly to optic nerve axons, extended duration of therapeutic effect, and minimal systemic effects. In this study, we have developed an in vitro system to detect and measure trans-meningeal passage of neuroprotective drugs. We have chosen brimonidine as the drug cargo given the evidence presented above and as a surrogate for other neuroprotective therapeutics in development. To optimize the human in vitro assay, the experimental protocol was first conducted in an animal model using fresh porcine optic nerve meninges. Eyes were obtained from pigs euthanized by other investigators at our institution directly at the time of death. For the human tissue experiments, patients undergoing therapeutic enucleation donated the orbital portion of the optic nerve. The protocol was developed in accordance with the Declaration of Helsinki. Vanderbilt University Institutional Review Board and Institutional Animal Care and Use Committee approval were obtained prior to initiation of the study. Signed consent was obtained from patients prior to enucleation of the eye. Porcine and human eyes with optic nerve stumps of over 15 mm were enucleated. The eyes with attached optic nerves were stored in Hank’s basic salt solution (HBSS) with glucose (Invitrogen, Carlsbad, CA) at 5 C for no longer than 12 h. Dissection of the meninges was performed under the operating microscope in the following manner. The optic nerve was cleared of perineural soft tissue with fine-toothed forceps and spring-loaded ophthalmic operating scissors. A longitudinal cut was made along the length of the nerve through the optic nerve sheath. When dissecting porcine nerves, a false plane was created deep to several layers of axons and followed circumferentially around the optic nerve segment. This was necessary given the lack of significant subarachnoid space in which to dissect. That is, in order to insure the integrity of the
43
arachnoid layer, which in large part defines the bloodeCSF barrier, axons deep to this layer were included in the dissection. The subarachnoid space was easily identifiable on the human optic nerves and, therefore, axons were not included in the dissection from the optic nerve. At completion, a sheet of contiguous meningeal tissue was removed. The dural-arachnoid sheath orientation was maintained. A diffusion cell with a 5 mm in diameter orifice and chamber volumes of 1 ml (PermeGear Hellertown, PA) was selected to test meningeal penetration by both brimonidine alone and brimonidine encapsulated nanoparticle. Meningeal samples were mounted between the donor and receiver chambers with available diffusion area of 19.6 mm2. The dural side was placed facing the donor and the subarachnoid side toward the receiver. Each meningeal sample was used for only one experiment. The donor chamber was filled with a 1 ml solution of either brimonidine or NP-brimonidine in HBSS with glucose. The receiver chamber had only HBSS with glucose. Both were constantly stirred with a magnetic stir bar. 25 ml samples were taken from both the donor and receiver chambers at time intervals. To prepare the brimonidine solution, brimonidine (Sigma Aldrich St. Louis, MO) was measured into an Eppendorff tube and HBSS with glucose was used to solubilize the drug. After ample sonication, the drug solution was filtered and 1 mL was measured into the donor side of the cell whereas 1 mL HBSS with glucose was measured into the receiver side. The drug-loaded nanoparticle was prepared as follows. The 50-nm polyester nanoparticles were prepared as previously described (Van der Ende et al., 2008). In brief, the nanoparticle is synthesized from dvalerolactone. Upon oxidation of the allyl groups to epoxides, a crosslinking reaction with short peg diamines facilitates the nanoparticle structure (Van der Ende et al., 2008). The nanoparticle is loaded after formation with brimonidine, a major advantage over other drug delivery systems, and leads to a formulated drug that can be administered (Van der Ende et al., 2009). Brimonidine was encapsulated into the nanoparticle via a developed and published nanoprecipitation method (Van der Ende et al., 2010). Briefly, because the nanosponges are soluble in organic solvents as well as drugs such as brimonidine, a homogenous solution is formed and precipitated into water that contains Vitamin E ePEG (1%). The product is centrifuged and washed with water to remove traces of the solvent. The product can be readily resuspended in buffer. The formation of the nanoparticle was controlled to allow for a 50 nm particle with 7% crosslinking and 7%% drug load. Nanoparticle with encapsulated brimonidine (50 nm) was measured into an Eppendorff tube and HBSS with glucose was added. After sonication, 1 ml of the drugloaded nanoparticle suspension was measured into the donor side of the cell whereas 1 mL HBSS with glucose was measured into the receiver side. 2.4 Lyophilized samples from each chamber at paired time points were dissolved in 25 ml DMSO. 2 ml of sample solution were pipetted onto the pedestal of a UVevis spectrophotometer (NanoDropÔ) and the absorbance measured at 389 nm. Measurements were made in triplicate and the average absorbance of each sample was used to plot diffusion of drug through the meninges over time. Absorbance was plotted against time points for both donor and receiver samples for each experiment demonstrating change in concentration over time. In a two chamber diffusion experiment the expected concentrationetime curve in the receiver cell would be governed by the equation:
PS CReceiver ¼ CDonor 1 e V t
44
K. Grove et al. / Experimental Eye Research 118 (2014) 42e45
Fig. 1. Diffusion of brimonidine through porcine meninges with brimonidine alone (left) and NP-Brimonidine (right). Absorbance measured by UVevis with a NanoDropÔ Spectrophotometer at 389 nm as a function of time indicates change in concentration of each chamber.
Where C is a concentration, P is the material permeability, S is the surface used for diffusion and V is the receiver volume. This equation presumes a simple diffusion model which the biological tissue will not exactly emulate. Two parameter exponential regression analysis was performed and goodness-of-fit evaluated by correlation coefficient (r2) for the human studies. Nuclear magnetic resonance (NMR) spectroscopy was performed on donor and receiver samples. Tissue analysis was performed with light microscopy. Porcine meningeal tissue harvested directly at the time of death without refrigeration and immediately fixed with 4% formalin was compared with both tissue harvested from eyes stored in Hank’s BSS with glucose at 5 C for 12 h and tissue from completed previous experiments to ensure histologic integrity after storage and experiments. The tissue samples were fixed in 4% neutral-buffered formalin, embedded in paraffin using standard procedures, cut into 4-mm-thick sections on a microtome, and conventionally stained with hematoxylineeosin. Samples were viewed at 10 and 20. Freezing of the tissue was avoided as to not disrupt its structure or function as this has been shown to increase the permeability of spinal meninges (Thompson and Bernards, 1998).
The preliminary porcine experiments demonstrated decreasing concentration of brimonidine in the donor chambers with corresponding rise in concentration of the receiver chambers indicating diffusion toward equilibrium. A delay occurred in time to equilibrium of brimonidine that was introduced into the donor chamber encapsulated in NP (Fig. 1). With human meninges, brimonidine was detected with increasing absorbance in the receiver chamber over time demonstrating diffusion through the meninges whether testing brimonidine or brimonidine encapsulated in nanoparticle (Fig. 2). Exponential regression curves are demonstrated in Fig. 2 with the following correlation coefficients (r2): Brimonidine alone (left) donor r2 ¼ 0.87 and receiver r2 ¼ 0.80, NP-brimonidine (right) donor r2 ¼ 0.79 and receiver r2 ¼ 0.84. With analysis of the time constant (s), we find that s is 10.2 h (donor) and 13.1 h (receiver) in the brimonidine study, whereas the NP-brimonidine study reveals s is 24.0 h (donor) and 15.9 h (receiver). NMR showed evidence of brimonidine in both chambers; NP was noted only in the donor chamber in the NP experiments. Comparison of porcine meningeal tissue harvested, fixed, and processed just after death without refrigeration, tissue harvested
Fig. 2. Diffusion of brimonidine through human meninges with brimonidine alone (left) and NP-brimonidine (right). Absorbance measured by UVevis as a function of time indicates change in concentration of each chamber. Exponential regression curves are plotted and evaluated by r2. Brimonidine alone (left) donor r2 ¼ 0.87 and receiver r2 ¼ 0.80. NPbrimonidine (right) donor r2 ¼ 0.79 and receiver fit r2 ¼ 0.84.
K. Grove et al. / Experimental Eye Research 118 (2014) 42e45
from eyes stored in Hank’s BSS with glucose at 5 C for 12 h, and tissue from completed experiments revealed that all specimens were grossly and histologically intact. Our hypothesis is that direct neuroprotection of the optic nerve axons is possible if drugs can be transported across meninges. We were able to show that brimonidine crosses the selectively permeable barrier of the optic nerve sheath in pigs and humans. Both the brimonidine alone and the NP-brimonidine experiments demonstrated penetration of the meningeal barrier of both species by brimonidine. An important consideration is that our assay measures brimonidine whether it is encapsulated or is present as free drug. Our NMR data show that NP is not present in the receiver chamber suggesting the drug had to be released from the NP on the donor side prior to permeation. That is, drug must be released before diffusion leading to a flattening of the slope due to this additional step. Also, brimonidine encapsulated in NP would decrease the effective concentration in the donor chamber further increasing time to equilibrium given the direct relationship of concentration and diffusive flux. Finally, comparison of the time constants of the human NP-brimonidine with the human brimonidine alone study indicates that diffusion is delayed by the initial parameter of drug being loaded in NP. Ultimately, the data show that brimonidine crosses porcine and human optic nerve meninges lending viability to a new route of optic neuropathy treatment: retrobulbar, controlled-release depot for diffusion of drug once unbound from NP encapsulation. We chose to test brimonidine with the NP. Extensive research is underway for many potential neuroprotective therapeutics, but no drug has yet to definitively translate the neuroprotective paradigm to optic neuropathy treatment in clinical trials. We chose brimonidine encouraged by the recent findings of the LowPressure Glaucoma Study Group (Krupin et al., 2011). Introducing a drug to the retrobulbar space implicates the intraorbital segments of RGC axons as the site of action or uptake. Though the pathways of the neuroprotective effects of brimonidine are not completely known, they are likely facilitated by binding of alpha2adrenergic receptors. Each of three subtypes is found in ocular tissue in varying distribution. In humans, subtype 2B is found on RGC axons in addition to the somata of RGCs and glia, as the axons course along the contour of the eye (Woldemussie et al., 2007). Though the presence of alpha2-adrenergic receptors on axonal segments comprising the orbital portion of the optic nerve has not been specifically investigated, confirmed presence along the intraocular portions of RCG axons is promising. Brimonidine may ultimately be an appropriate drug for sustained trans-meningeal delivery by retrobulbar depot, but it also acts as a surrogate for agents that prove to be significantly neuroprotective to the optic nerve axons. Encouraging preliminary data is presented here. Porcine studies gave qualitative evidence of meningeal permeability to brimonidine and allowed optimization of the protocol, but given the necessity to include a layer of axons of variable thickness, quantitative analysis of replicate diffusion studies will vary significantly. The next phase of this work continues with additional in vitro studies with human tissue and use of an animal model to demonstrate effective targeting and tissue concentration in vivo. This study puts forth the potential of a novel route of drug delivery to optic nerve axons. It is understood that the meninges of the CNS are permeable to some drugs, i.e. opioids used in epidural anesthesia (Moore et al., 1982). We propose that attempts to exploit the trans-meningeal route in the development of therapies for optic neuropathies is now timely given the confluence of three conditions: significant progress in the pursuit of neuroprotection for
45
ophthalmic disease, a novel nanoparticle that offers solubilization and controlled-release of drug, and an image-guided orbital endoscope capable of low-risk access to the orbital portion of the optic nerve. Financial support Supported by a Fight For Sight/North American Neuroophthalmology Award. Supported in part by Research to Prevent Blindness Unrestricted Grant and Physician Scientist Award, NIH R21EB009223-02 and NIH 5R21RR025806-02. Conflict of interest No conflicting relationship exists for any author. References Atuegwu, N.C., Mawn, L., Galloway, R., 2007. Transorbital Endoscopic image guidance, Engineering in Medicine and Biology Society, 2007. In: EMBS 2007. 29th Annual International Conference of the IEEE, pp. 4663e4666. Baltmr, A., Duggan, J., Nizari, S., Salt, T.E., Cordeiro, M.F., 2010. Neuroprotection in glaucoma e is there a future role? Exp. Eye Res. 91, 554e566. Bessero, A.C., Clarke, P.G., 2010. Neuroprotection for optic nerve disorders. Curr. Opin. Neurol. 23, 10e15. Chidlow, G., Wood, J.P.M., Casson, R.J., 2007. Pharmacological neuroprotection for glaucoma. Drugs 67, 725e759. Diebold, Y., Calonge, M., 2010. Applications of nanoparticles in ophthalmology. Prog. Retin. Eye Res. 29, 596e609. Fazzone, H.E., Kupersmith, M.J., Leibmann, J., 2003. Does topical brimonidine tartrate help NAION? Br. J. Ophthalmol. 87, 1193e1194. Krupin, T., Liebmann, J.M., Greenfield, D.S., Ritch, R., Gardiner, S., 2011. A randomized trial of brimonidine versus timolol in preserving visual function: results from the low-pressure glaucoma treatment study. Am. J. Ophthalmol. 151, 671e681. Lambert, W., Ruiz, L., Crish, S., Wheeler, L., Calkins, D., 2011. Brimonidine prevents axonal and somatic degeneration of retinal ganglion cell neurons. Mol. Neurodegen. 6, 4. Levin, L.A., 2007. Axonal loss and neuroprotection in optic neuropathies. Can J. Ophthalmol. 42, 403e408. Miller, N.R., 1996. The optic nerve. Curr. Opin. Neurol. 9, 5e15. Moore, R.A., Bullingham, R.E.S., McQuay, H.J., Hand, C.W., Aspel, J.B., Allen, M.C., Thomas, D., 1982. Dural permeability to narcotics: in vitro determination and application to extradural administration. Br. J. Anaesth. 54, 1117e1128. Nabeshima, S., Reese, T.S., Landis, D.M., Brightman, M.W., 1975. Junctions in the meninges and marginal glia. J. Comp. Neurol. 164, 127e169. Newman, N.J., Biousse, V., David, R., Bhatti, M.T., Hamilton, S.R., Farris, B.K., Lesser, R.L., Newman, S.A., Turbin, R.E., Chen, K., Keaney, R.P., 2005. Prophylaxis for second eye involvement in leber hereditary optic neuropathy: an openlabeled, nonrandomized multicenter trial of topical brimonidine purite. Am. J. Ophthalmol. 140, 407e415. Saunders, N.R., Ek, C.J., Habgood, M.D., Dziegielewska, K.M., 2008. Barriers in the brain: a renaissance? Trends Neurosci. 31, 279e286. Saylor, M., McLoon, L.K., Harrison, A.R., Lee, M.S., 2009. Experimental and clinical evidence for brimonidine as an optic nerve and retinal neuroprotective agent: an evidence-based review. Arch. Ophthalmol. 127, 402e406. Segal, M.B., 2000. The choroid plexuses and the barriers between the blood and the cerebrospinal fluid. Cell Mol. Neurobiol. 20, 183e196. Thompson, S.J., Bernards, C.M., 1998. Barrier properties of the spinal meninges are markedly decreased by freezing meningeal tissues. Anesthesiology 89, 1276e1278. Van der Ende, A.E., Kravitz, E.J., Harth, E., 2008. Approach to formation of multifunctional polyester particles in controlled nanoscopic dimensions. J. Am. Chem. Soc. 130, 8706e8713. Van der Ende, A., Croce, T., Hamilton, K., Sathiyakumar, V., Harth, E., 2009. Tailored polyester nanoparticles: post-modification with dendritic transporter and targeting units via reductive amination and thiol-ene chemistry. Soft Matter 5, 1417e1425. Van der Ende, A.E., Sathiyakumar, V., Diaz, R., Hallahan, D.E., Harth, E., 2010. Linear release nanoparticle devices for advanced targeted cancer therapies with increased efficacy. Polym. Chem. 1, 260e271. Vandenabeele, F., Creemers, J., Lambrichts, I., 1996. Ultrastructure of the human spinal arachnoid mater and dura mater. J. Anat. 189 (Pt 2), 417e430. Wilhelm, B., Ludtke, H., Wilhelm, H., 2006. Efficacy and tolerability of 0.2% brimonidine tartrate for the treatment of acute non-arteritic anterior ischemic optic neuropathy (NAION): a 3-month, double-masked, randomised, placebocontrolled trial. Graefes Arch. Clin. Exp. Ophthalmol. 244, 551e558. Woldemussie, E., Wijono, M., Pow, D., 2007. Localization of alpha 2 receptors in ocular tissues. Vis. Neurosci. 24, 745e756.
Contents lists available at ScienceDirect
Experimental Eye Research journal homepage: www.elsevier.com/locate/yexer
Short communication
Trans-meningeal drug delivery to optic nerve ganglion cell axons using a nanoparticle drug delivery system Karen Grove a, Julia Dobish b, Eva Harth b, Martha-Conley Ingram c, Robert L. Galloway c, Louise A. Mawn a, * a b c
Vanderbilt Eye Institute, 2311 Pierce Avenue, Nashville, TN 37232-8808, USA Chemistry Department, 7619 Stevenson Center, Vanderbilt University, Nashville, TN 37232, USA Biomedical Engineering, 5824 Stevenson Center, Vanderbilt University, Nashville, TN 37232, USA
a r t i c l e i n f o
a b s t r a c t
Article history: Received 2 June 2013 Accepted in revised form 23 October 2013 Available online 1 November 2013
The purpose of this study was to investigate if neuroprotective drugs can cross the optic nerve sheath in vitro. Four optic nerves were used for this study. Two porcine nerves were harvested at the time of euthanasia and two human nerves were obtained at the time of therapeutic globe enucleation. The optic nerve sheaths were dissected and placed as a membrane in a two chamber diffusion cell to test meningeal penetration by both brimonidine alone and brimonidine encapsulated in nanoparticle (NPbrimonidine). Brimonidine concentration was assayed by UVevis spectrometer measurement of absorbance at 389 nm. Increasing concentration of brimonidine on the receiver side of the chamber was measured in both the brimonidine alone and the brimonidine encapsulated experiments. The human data were fitted with a two parameter exponential regression analysis (brimonidine alone donor r2 ¼ 0.87 and receiver r2 ¼ 0.80, NP-brimonidine donor r2 ¼ 0.79 and receiver r2 ¼ 0.84). Time constant (s) was 10.2 h (donor) and 13.1 h (receiver) in the brimonidine study, and 24.0 h (donor) and 15.9 h (receiver) in the NP-brimonidine study. Encapsulated brimonidine had a longer time to reach equilibrium. Passage of brimonidine through the optic nerve sheath was demonstrated in the experiments. Increase in time constants when comparing the NP-brimonidine with the brimonidine curves in the human studiesindicates that diffusion is delayed by the initial parameter of drug being loaded in NP. Direct treatment of injured optic nerve axons may be possible by trans-meningeal drug diffusion. Ó 2013 Elsevier Ltd. All rights reserved.
Keywords: optic nerve neuroprotection brimonidine drug delivery diffusion nanoparticle human porcine
Nearly all optic neuropathies of varying etiologies including ischemic, traumatic, and Leber’s hereditary optic neuropathy have no therapeutic options (Levin, 2007). Glaucoma is a partial exception, though visual field loss and retinal ganglion cell (RGC) death continue despite well-controlled intraocular pressure (Baltmr et al., 2010). Many neuroprotective strategies are under investigation with goals to protect undamaged RGCs and to rescue their injured counterpart. Many candidate therapies have been successful in animal models of acute and chronic neurodegenerative processes of the optic nerve and tremendous effort continues in pursuit of translating this success (Bessero and Clarke, 2010; Chidlow et al., 2007; Levin, 2007). Brimonidine is of particular interest. Efficacy
* Corresponding author. Tel.: þ1 615 936 1960; fax: þ1 615 936 1540. E-mail addresses: [email protected] (K. Grove), julnmeyer@gmail. com (J. Dobish), [email protected] (E. Harth), martha.e.ingram@gmail. com (M.-C. Ingram), [email protected] (R.L. Galloway), louise.mawn@ vanderbilt.edu (L.A. Mawn). 0014-4835/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.exer.2013.10.016
of neuroprotection has been demonstrated in animal models (Lambert et al., 2011; Saylor et al., 2009). Until recently, clinical trials did not show statistically significant efficacy of brimonidine in the treatment of optic neuropathies (Fazzone et al., 2003; Newman et al., 2005; Wilhelm et al., 2006). However, the Low-Pressure Glaucoma Study Group recently demonstrated less visual field loss in brimonidine-treated patients (9, 9.1%) versus timololtreated patients (31, 39.2% log-rank 12.4, P ¼ .001) in a doublemasked, randomized study. This occurred despite similar intraocular pressure for brimonidine- and timolol-treated patients at all time points. This is consistent with an IOP-independent effect and suggestive of a neuroprotective mechanism (Krupin et al., 2011). Each promising therapy requires a viable route of delivery, a particular challenge regarding central nervous system (CNS) and ocular tissues. Systemic administration of some potential neuroprotective therapeutics is prevented by the side effect profile when high, frequent dosing is required to overcome CNS and ocular barriers (Diebold and Calonge, 2010). The most common options for local treatment include topical drops as well as sub-conjunctival,
K. Grove et al. / Experimental Eye Research 118 (2014) 42e45
sub-tenon’s, and intravitreal injection. However, another route that has not been explored is trans-meningeal drug delivery that especially warrants investigation given that most optic neuropathies are axonal (Levin, 2007). Obstacles to drug delivery to the optic nerve axons include the need for the agent to cross the optic nerve sheath as well as to provide sustained delivery of the neuroprotective drug. The orbital portion of the optic nerve is surrounded by the same meninges as the brain (Miller, 1996). This barrier effectively contributes to the sequestration of its contents from systemic circulation. More specifically, in addition to the bloodebrain barrier provided by the endothelium of brain capillaries, there is an epithelial blood-cerebrospinal fluid (CSF) barrier maintained by the choroid plexuses and the outer arachnoid membrane (Segal, 2000). It is composed of an arachnoid barrier cell layer with numerous tight junctions and a distinct, continuous basal lamina, and is the definitive barrier between the dural circulation and the subarachnoid CSF (Nabeshima et al., 1975; Saunders et al., 2008; Vandenabeele et al., 1996). The parallel development of two unique innovations makes exploration of the trans-meningeal route timely. First, we have developed an image-guided orbital endoscope that facilitates minimally invasive and precise access to the orbital portion of the optic nerve (Atuegwu et al., 2007). Second, our team has created a novel polyester nanoparticle (NP) drug delivery system that allows controlled rate of drug release (Van der Ende et al., 2008). Many NP drug delivery systems for use in ocular disease are under investigation (Diebold and Calonge, 2010), though a trans-meningeal approach to the optic nerve axons has not been studied in any regard. We propose the possibility of a novel route of delivery through the optic nerve sheath. Drug encapsulated in NP can be customized for controlled released over an extended period of time. Therefore, a retrobulbar depot could deliver continuous drug to the RGC axons through the optic nerve sheath. The NP has the potential to facilitate high drug concentration directly to optic nerve axons, extended duration of therapeutic effect, and minimal systemic effects. In this study, we have developed an in vitro system to detect and measure trans-meningeal passage of neuroprotective drugs. We have chosen brimonidine as the drug cargo given the evidence presented above and as a surrogate for other neuroprotective therapeutics in development. To optimize the human in vitro assay, the experimental protocol was first conducted in an animal model using fresh porcine optic nerve meninges. Eyes were obtained from pigs euthanized by other investigators at our institution directly at the time of death. For the human tissue experiments, patients undergoing therapeutic enucleation donated the orbital portion of the optic nerve. The protocol was developed in accordance with the Declaration of Helsinki. Vanderbilt University Institutional Review Board and Institutional Animal Care and Use Committee approval were obtained prior to initiation of the study. Signed consent was obtained from patients prior to enucleation of the eye. Porcine and human eyes with optic nerve stumps of over 15 mm were enucleated. The eyes with attached optic nerves were stored in Hank’s basic salt solution (HBSS) with glucose (Invitrogen, Carlsbad, CA) at 5 C for no longer than 12 h. Dissection of the meninges was performed under the operating microscope in the following manner. The optic nerve was cleared of perineural soft tissue with fine-toothed forceps and spring-loaded ophthalmic operating scissors. A longitudinal cut was made along the length of the nerve through the optic nerve sheath. When dissecting porcine nerves, a false plane was created deep to several layers of axons and followed circumferentially around the optic nerve segment. This was necessary given the lack of significant subarachnoid space in which to dissect. That is, in order to insure the integrity of the
43
arachnoid layer, which in large part defines the bloodeCSF barrier, axons deep to this layer were included in the dissection. The subarachnoid space was easily identifiable on the human optic nerves and, therefore, axons were not included in the dissection from the optic nerve. At completion, a sheet of contiguous meningeal tissue was removed. The dural-arachnoid sheath orientation was maintained. A diffusion cell with a 5 mm in diameter orifice and chamber volumes of 1 ml (PermeGear Hellertown, PA) was selected to test meningeal penetration by both brimonidine alone and brimonidine encapsulated nanoparticle. Meningeal samples were mounted between the donor and receiver chambers with available diffusion area of 19.6 mm2. The dural side was placed facing the donor and the subarachnoid side toward the receiver. Each meningeal sample was used for only one experiment. The donor chamber was filled with a 1 ml solution of either brimonidine or NP-brimonidine in HBSS with glucose. The receiver chamber had only HBSS with glucose. Both were constantly stirred with a magnetic stir bar. 25 ml samples were taken from both the donor and receiver chambers at time intervals. To prepare the brimonidine solution, brimonidine (Sigma Aldrich St. Louis, MO) was measured into an Eppendorff tube and HBSS with glucose was used to solubilize the drug. After ample sonication, the drug solution was filtered and 1 mL was measured into the donor side of the cell whereas 1 mL HBSS with glucose was measured into the receiver side. The drug-loaded nanoparticle was prepared as follows. The 50-nm polyester nanoparticles were prepared as previously described (Van der Ende et al., 2008). In brief, the nanoparticle is synthesized from dvalerolactone. Upon oxidation of the allyl groups to epoxides, a crosslinking reaction with short peg diamines facilitates the nanoparticle structure (Van der Ende et al., 2008). The nanoparticle is loaded after formation with brimonidine, a major advantage over other drug delivery systems, and leads to a formulated drug that can be administered (Van der Ende et al., 2009). Brimonidine was encapsulated into the nanoparticle via a developed and published nanoprecipitation method (Van der Ende et al., 2010). Briefly, because the nanosponges are soluble in organic solvents as well as drugs such as brimonidine, a homogenous solution is formed and precipitated into water that contains Vitamin E ePEG (1%). The product is centrifuged and washed with water to remove traces of the solvent. The product can be readily resuspended in buffer. The formation of the nanoparticle was controlled to allow for a 50 nm particle with 7% crosslinking and 7%% drug load. Nanoparticle with encapsulated brimonidine (50 nm) was measured into an Eppendorff tube and HBSS with glucose was added. After sonication, 1 ml of the drugloaded nanoparticle suspension was measured into the donor side of the cell whereas 1 mL HBSS with glucose was measured into the receiver side. 2.4 Lyophilized samples from each chamber at paired time points were dissolved in 25 ml DMSO. 2 ml of sample solution were pipetted onto the pedestal of a UVevis spectrophotometer (NanoDropÔ) and the absorbance measured at 389 nm. Measurements were made in triplicate and the average absorbance of each sample was used to plot diffusion of drug through the meninges over time. Absorbance was plotted against time points for both donor and receiver samples for each experiment demonstrating change in concentration over time. In a two chamber diffusion experiment the expected concentrationetime curve in the receiver cell would be governed by the equation:
PS CReceiver ¼ CDonor 1 e V t
44
K. Grove et al. / Experimental Eye Research 118 (2014) 42e45
Fig. 1. Diffusion of brimonidine through porcine meninges with brimonidine alone (left) and NP-Brimonidine (right). Absorbance measured by UVevis with a NanoDropÔ Spectrophotometer at 389 nm as a function of time indicates change in concentration of each chamber.
Where C is a concentration, P is the material permeability, S is the surface used for diffusion and V is the receiver volume. This equation presumes a simple diffusion model which the biological tissue will not exactly emulate. Two parameter exponential regression analysis was performed and goodness-of-fit evaluated by correlation coefficient (r2) for the human studies. Nuclear magnetic resonance (NMR) spectroscopy was performed on donor and receiver samples. Tissue analysis was performed with light microscopy. Porcine meningeal tissue harvested directly at the time of death without refrigeration and immediately fixed with 4% formalin was compared with both tissue harvested from eyes stored in Hank’s BSS with glucose at 5 C for 12 h and tissue from completed previous experiments to ensure histologic integrity after storage and experiments. The tissue samples were fixed in 4% neutral-buffered formalin, embedded in paraffin using standard procedures, cut into 4-mm-thick sections on a microtome, and conventionally stained with hematoxylineeosin. Samples were viewed at 10 and 20. Freezing of the tissue was avoided as to not disrupt its structure or function as this has been shown to increase the permeability of spinal meninges (Thompson and Bernards, 1998).
The preliminary porcine experiments demonstrated decreasing concentration of brimonidine in the donor chambers with corresponding rise in concentration of the receiver chambers indicating diffusion toward equilibrium. A delay occurred in time to equilibrium of brimonidine that was introduced into the donor chamber encapsulated in NP (Fig. 1). With human meninges, brimonidine was detected with increasing absorbance in the receiver chamber over time demonstrating diffusion through the meninges whether testing brimonidine or brimonidine encapsulated in nanoparticle (Fig. 2). Exponential regression curves are demonstrated in Fig. 2 with the following correlation coefficients (r2): Brimonidine alone (left) donor r2 ¼ 0.87 and receiver r2 ¼ 0.80, NP-brimonidine (right) donor r2 ¼ 0.79 and receiver r2 ¼ 0.84. With analysis of the time constant (s), we find that s is 10.2 h (donor) and 13.1 h (receiver) in the brimonidine study, whereas the NP-brimonidine study reveals s is 24.0 h (donor) and 15.9 h (receiver). NMR showed evidence of brimonidine in both chambers; NP was noted only in the donor chamber in the NP experiments. Comparison of porcine meningeal tissue harvested, fixed, and processed just after death without refrigeration, tissue harvested
Fig. 2. Diffusion of brimonidine through human meninges with brimonidine alone (left) and NP-brimonidine (right). Absorbance measured by UVevis as a function of time indicates change in concentration of each chamber. Exponential regression curves are plotted and evaluated by r2. Brimonidine alone (left) donor r2 ¼ 0.87 and receiver r2 ¼ 0.80. NPbrimonidine (right) donor r2 ¼ 0.79 and receiver fit r2 ¼ 0.84.
K. Grove et al. / Experimental Eye Research 118 (2014) 42e45
from eyes stored in Hank’s BSS with glucose at 5 C for 12 h, and tissue from completed experiments revealed that all specimens were grossly and histologically intact. Our hypothesis is that direct neuroprotection of the optic nerve axons is possible if drugs can be transported across meninges. We were able to show that brimonidine crosses the selectively permeable barrier of the optic nerve sheath in pigs and humans. Both the brimonidine alone and the NP-brimonidine experiments demonstrated penetration of the meningeal barrier of both species by brimonidine. An important consideration is that our assay measures brimonidine whether it is encapsulated or is present as free drug. Our NMR data show that NP is not present in the receiver chamber suggesting the drug had to be released from the NP on the donor side prior to permeation. That is, drug must be released before diffusion leading to a flattening of the slope due to this additional step. Also, brimonidine encapsulated in NP would decrease the effective concentration in the donor chamber further increasing time to equilibrium given the direct relationship of concentration and diffusive flux. Finally, comparison of the time constants of the human NP-brimonidine with the human brimonidine alone study indicates that diffusion is delayed by the initial parameter of drug being loaded in NP. Ultimately, the data show that brimonidine crosses porcine and human optic nerve meninges lending viability to a new route of optic neuropathy treatment: retrobulbar, controlled-release depot for diffusion of drug once unbound from NP encapsulation. We chose to test brimonidine with the NP. Extensive research is underway for many potential neuroprotective therapeutics, but no drug has yet to definitively translate the neuroprotective paradigm to optic neuropathy treatment in clinical trials. We chose brimonidine encouraged by the recent findings of the LowPressure Glaucoma Study Group (Krupin et al., 2011). Introducing a drug to the retrobulbar space implicates the intraorbital segments of RGC axons as the site of action or uptake. Though the pathways of the neuroprotective effects of brimonidine are not completely known, they are likely facilitated by binding of alpha2adrenergic receptors. Each of three subtypes is found in ocular tissue in varying distribution. In humans, subtype 2B is found on RGC axons in addition to the somata of RGCs and glia, as the axons course along the contour of the eye (Woldemussie et al., 2007). Though the presence of alpha2-adrenergic receptors on axonal segments comprising the orbital portion of the optic nerve has not been specifically investigated, confirmed presence along the intraocular portions of RCG axons is promising. Brimonidine may ultimately be an appropriate drug for sustained trans-meningeal delivery by retrobulbar depot, but it also acts as a surrogate for agents that prove to be significantly neuroprotective to the optic nerve axons. Encouraging preliminary data is presented here. Porcine studies gave qualitative evidence of meningeal permeability to brimonidine and allowed optimization of the protocol, but given the necessity to include a layer of axons of variable thickness, quantitative analysis of replicate diffusion studies will vary significantly. The next phase of this work continues with additional in vitro studies with human tissue and use of an animal model to demonstrate effective targeting and tissue concentration in vivo. This study puts forth the potential of a novel route of drug delivery to optic nerve axons. It is understood that the meninges of the CNS are permeable to some drugs, i.e. opioids used in epidural anesthesia (Moore et al., 1982). We propose that attempts to exploit the trans-meningeal route in the development of therapies for optic neuropathies is now timely given the confluence of three conditions: significant progress in the pursuit of neuroprotection for
45
ophthalmic disease, a novel nanoparticle that offers solubilization and controlled-release of drug, and an image-guided orbital endoscope capable of low-risk access to the orbital portion of the optic nerve. Financial support Supported by a Fight For Sight/North American Neuroophthalmology Award. Supported in part by Research to Prevent Blindness Unrestricted Grant and Physician Scientist Award, NIH R21EB009223-02 and NIH 5R21RR025806-02. Conflict of interest No conflicting relationship exists for any author. References Atuegwu, N.C., Mawn, L., Galloway, R., 2007. Transorbital Endoscopic image guidance, Engineering in Medicine and Biology Society, 2007. In: EMBS 2007. 29th Annual International Conference of the IEEE, pp. 4663e4666. Baltmr, A., Duggan, J., Nizari, S., Salt, T.E., Cordeiro, M.F., 2010. Neuroprotection in glaucoma e is there a future role? Exp. Eye Res. 91, 554e566. Bessero, A.C., Clarke, P.G., 2010. Neuroprotection for optic nerve disorders. Curr. Opin. Neurol. 23, 10e15. Chidlow, G., Wood, J.P.M., Casson, R.J., 2007. Pharmacological neuroprotection for glaucoma. Drugs 67, 725e759. Diebold, Y., Calonge, M., 2010. Applications of nanoparticles in ophthalmology. Prog. Retin. Eye Res. 29, 596e609. Fazzone, H.E., Kupersmith, M.J., Leibmann, J., 2003. Does topical brimonidine tartrate help NAION? Br. J. Ophthalmol. 87, 1193e1194. Krupin, T., Liebmann, J.M., Greenfield, D.S., Ritch, R., Gardiner, S., 2011. A randomized trial of brimonidine versus timolol in preserving visual function: results from the low-pressure glaucoma treatment study. Am. J. Ophthalmol. 151, 671e681. Lambert, W., Ruiz, L., Crish, S., Wheeler, L., Calkins, D., 2011. Brimonidine prevents axonal and somatic degeneration of retinal ganglion cell neurons. Mol. Neurodegen. 6, 4. Levin, L.A., 2007. Axonal loss and neuroprotection in optic neuropathies. Can J. Ophthalmol. 42, 403e408. Miller, N.R., 1996. The optic nerve. Curr. Opin. Neurol. 9, 5e15. Moore, R.A., Bullingham, R.E.S., McQuay, H.J., Hand, C.W., Aspel, J.B., Allen, M.C., Thomas, D., 1982. Dural permeability to narcotics: in vitro determination and application to extradural administration. Br. J. Anaesth. 54, 1117e1128. Nabeshima, S., Reese, T.S., Landis, D.M., Brightman, M.W., 1975. Junctions in the meninges and marginal glia. J. Comp. Neurol. 164, 127e169. Newman, N.J., Biousse, V., David, R., Bhatti, M.T., Hamilton, S.R., Farris, B.K., Lesser, R.L., Newman, S.A., Turbin, R.E., Chen, K., Keaney, R.P., 2005. Prophylaxis for second eye involvement in leber hereditary optic neuropathy: an openlabeled, nonrandomized multicenter trial of topical brimonidine purite. Am. J. Ophthalmol. 140, 407e415. Saunders, N.R., Ek, C.J., Habgood, M.D., Dziegielewska, K.M., 2008. Barriers in the brain: a renaissance? Trends Neurosci. 31, 279e286. Saylor, M., McLoon, L.K., Harrison, A.R., Lee, M.S., 2009. Experimental and clinical evidence for brimonidine as an optic nerve and retinal neuroprotective agent: an evidence-based review. Arch. Ophthalmol. 127, 402e406. Segal, M.B., 2000. The choroid plexuses and the barriers between the blood and the cerebrospinal fluid. Cell Mol. Neurobiol. 20, 183e196. Thompson, S.J., Bernards, C.M., 1998. Barrier properties of the spinal meninges are markedly decreased by freezing meningeal tissues. Anesthesiology 89, 1276e1278. Van der Ende, A.E., Kravitz, E.J., Harth, E., 2008. Approach to formation of multifunctional polyester particles in controlled nanoscopic dimensions. J. Am. Chem. Soc. 130, 8706e8713. Van der Ende, A., Croce, T., Hamilton, K., Sathiyakumar, V., Harth, E., 2009. Tailored polyester nanoparticles: post-modification with dendritic transporter and targeting units via reductive amination and thiol-ene chemistry. Soft Matter 5, 1417e1425. Van der Ende, A.E., Sathiyakumar, V., Diaz, R., Hallahan, D.E., Harth, E., 2010. Linear release nanoparticle devices for advanced targeted cancer therapies with increased efficacy. Polym. Chem. 1, 260e271. Vandenabeele, F., Creemers, J., Lambrichts, I., 1996. Ultrastructure of the human spinal arachnoid mater and dura mater. J. Anat. 189 (Pt 2), 417e430. Wilhelm, B., Ludtke, H., Wilhelm, H., 2006. Efficacy and tolerability of 0.2% brimonidine tartrate for the treatment of acute non-arteritic anterior ischemic optic neuropathy (NAION): a 3-month, double-masked, randomised, placebocontrolled trial. Graefes Arch. Clin. Exp. Ophthalmol. 244, 551e558. Woldemussie, E., Wijono, M., Pow, D., 2007. Localization of alpha 2 receptors in ocular tissues. Vis. Neurosci. 24, 745e756.
