Ciliary neurotrophic factor mediated growth of retinal ganglion cell axons on PGS/PCL scaffolds

Ciliary neurotrophic factor (CNTF) promotes survival and/or differentiation of a variety of neuronal cells including retinal ganglion cells (RGCs). Delivery of CNTF requires a suitable medium capable of mediating diffusion and premature release of CNTF within the target tissue. Polymeric tissue-engineered scaffolds have been readily used as substrates for cell transplantation, expansion, and differentiation and, as carriers of cell growth factors. Their functions to CNTF release for RGC proliferation have remained so far unexplored, especially to CNTF affinity to the scaffold and subsequent RGC fate. Electrospun poly(glycerol sebacate)/poly(ϵ-caprolactone) (PGS/PCL) biopolymer scaffolds have recently shown promising results in terms of supporting regeneration of RGC neurites. This work explores covalent immobilization of CNTF on PGS/PCL scaffold and the way immobilised CNTF mediates growth of RGC axons on the scaffold. An ex-vivo three-dimensional model of rodent optic nerve on PGS/PCL revealed that RGC explants cultured in CNTF mediated environment increased their neurite extensions after 20 d of cell culture employing neurite outgrowth measurements. The CNTF secretion on PGS/PCL scaffold was found bio-mimicking natural extracellular matrix of the cell target tissue and, consequently, has shown a potential to improve the overall efficacy of the RGC regeneration process.


Introduction
Optic nerve neuropathies, including optic degenerative disorders and physical injuries to the optic nerve such as traumatic optic nerve crush result in progressive loss of retinal ganglion cells (RGCs) and defects in visual function.RGCs display low regenerative capacity, similar to neural synapses in the central nervous system pathways, and therefore damage induced RGCs apoptosis (e.g. in glaucoma) correlates strongly with irreversible vision loss.Recent studies have shown that RGC regenerative capacity can be enhanced by a temporal micro-environment [1] supported by neurotrophic factors (aka.growth factors)-the endogenous soluble proteins that regulate survival, growth, morphological plasticity and synthesis of proteins for differentiated functions of neurons, including RGCs [2].The growth-factor mediated micro-environment ultimately is capable of preventing degeneration of neurons, and for RGCs, has a potential to slow down progression of vision loss and, eventually stimulating the survival, proliferation and differentiation of RGCs' axons in the target tissue [3,4].Among the available growth factors, the ciliary neurotrophic factor (CNTF) has been shown to provide a lasting long-distance axonal regeneration in adult RGCs owing to a specific RGC receptor to CNTF ligand binding [3,5,6].However, delivery of growth factors, including CNTF, to the RGCs target area is challenging owing to CNTFs' instability, propensity for diffusion away and, rapid in vivo enzymatic degradation in the target area [4,7].Furthermore, often a non-linear transport of neurotrophic factors to the damaged tissue is required and, as such, precise apportioning, maintenance and preservation of a neurotrophic factor gradient is needed to contain its function within a given cellular microenvironment [8].Recent studies on morphogen signalling and various concentration gradients functions have conclusively shown that cells, including the RGCs, grow in a direction where cell positional information is provided [9,10].Consequently, a substrate for the delivery of growth factors that is not only capable of immobilizing cells but also capable of protecting the biological functions of cells in vivo, while maintaining a suitable concentration of neurotrophic factor gradients for cell expansion and differentiation is required [7].
Recent studies have shown that electrospun fibrous biopolymer scaffolds could be promising carriers (aka.cargo vehicles) for the growth factor delivery [11,12] as they display advantageous properties, including benign degradation characteristics and an interconnected porous network structure with a high surface-to-volume ratio that is similar to a natural extracellular matrix (ECM) [13,14].The scaffolds, in general, can promote cell infiltration, adhesion and proliferation and, support transport of nutrients and removal of metabolic wastes in target site, analogous to ECM.Biopolymer scaffolds that degrade without producing toxic or inflammatory by-products and induce minimal inflammatory response are of particular interest [4,7,15,16].
Our most recent work [17] on electrospun poly(glycerol sebacate)/poly(ε-caprolactone) (PGS/PCL) biopolymer scaffolds have shown promising results in terms of supporting regeneration of RGC neurites and, have shown PGS/PCL scaffolds to be far more suitable for the RGCs transplantation compared to several other known biomaterial counterparts based on RGC positive adhesion, cell proliferation and survival [18].Knowing that the PGS/PCL is an appropriate support carrier for RGCs we firmly believe that intracellular signalling pathways activated by CNTF can further enhance RGCs regeneration, however, the specific functions of CNTF grafting and release on PGS/PCL have not yet been explored.
In this study we investigate a gradual release of the growth factor through the CNTF covalent immobilization on PGS/PCL aiming to increase the axonal growth of RGCs neurites.Our aim is to take an advantage of the aligned fibrous PGS/PCL scaffold providing a suitable structural framework for RGC neuronal adhesion and directional growth, assisted by CNTF providing the biochemical features required to maintain the guidance and nourishment to the regenerating RGC axons within the PCL/PGS network.Employing an ex-vivo three-dimensional (3D) model of rodent optic nerve on PGS/PCL scaffold we investigate CNTF secretion on PGS/PCL using the neurite outgrowth measurements and evaluate the PCL/PGS as a potential carrier for the growth factor delivery, specific to the RGCs.Consequently, our work details the design of an experimental 3D model of optic nerve regeneration based on a hypothesis that a controlled CNTF release within the PGS/PCL will enhance RGCs axonal growth.

Electrospun fibrous scaffold fabrication and sterilization
The PGS/PCL scaffolds were fabricated according to the previously established protocol [18].Briefly, PGS (M n = 12 000, Sigma Aldrich, Australia) and PCL (M n = 80 000, Sigma Aldrich, Australia) polymers were blended at a weight ratio of 2:1, respectively.The resultant blend was dissolved in a mixture of chloroform (90%) and ethanol (10%) (both Merck KGaA, Darmstadt, Germany) at 15 wt.% concentration and stirred at room temperature (RT) for 4 h.After 2 h of resting time at RT, the PGS/PCL solution was loaded into a 5 ml syringe with a 23-gauge needle (0.6 × 25 mm).The solution feed rate was controlled at 1 (ml hr −1 ) using a syringe pump (220 V, 50 Hz).An electric potential of 15 kV was applied across 0.18 m between a tip of the syringe needle and a rotating drum collector, the latter was covered in removable aluminium foil on which the injected PGS/PCL fibres were spined and later collected.
For purposes of sterilization, the collected PGS/PCL fibres, now in a form of a condensed fibrous mesh (i.e.scaffold), were first washed in a distilled water and treated with a 95% ethanol for 1 h.Secondly, the PGS/PCL fibres were irradiated for 40 min by an ultraviolet light on both sides of each scaffold.Finally, the PGS/PCL scaffolds were immersed in phosphate-buffered saline (PBS) (2x) (Sigma Chemical Co, St. Louis, MO, USA) for 24 h prior to all other steps undertaken.

Immobilization of CNTF on PGS/ PCL scaffolds
To immobilize CNTF on PGS/PCL scaffolds, CNTF was covalently attached to alkaline hydrolysed PGS/PCL as outlined in the previously published protocol [19].Briefly, the PGS/PCL scaffolds (1 × 1 cm 2 ) were hydrolysed by NaOH (Sigma Chemical Co, St. Louis, MO, USA) to cleave the ester linkages present in PGS and PCL polymer backbones, resulting in formation of free carboxyl groups on the scaffolds.Afterwards, the hydrolysed scaffolds were biofunctionalized by CNTF, where CNTF was covalently linked to carboxylic moieties on PGS/PCL.Hydrolysed PGS/PCL scaffolds were obtained through immersion in NaOH (1 N) for 80 min at RT.After hydrolysis, the scaffolds were washed with de-ionized water and dried at 37 • C overnight.Subsequently, the hydrolysed scaffolds were immersed in 2-(N-morpholino) ethanesulfonic acid (MES) (Sigma-Aldrich, Missouri, USA) buffer solution (0.1 M, pH 5.0) for 30 min at RT.Then, the scaffolds were activated with the MES buffer solution, which included 5 mg ml −1 of 1-ethyl-3-(3dimethylaminopropyl) carbodiimide hydrochloride and 2.5 mg ml −1 N-hydroxysulfosuccinimide sodium salt (Sigma-Aldrich, Missouri, USA) for 90 min at RT. Finally, the obtained scaffolds were incubated with 1 µg ml −1 of CNTF for 24 h and, subsequently, repeatedly rinsed with PBS to remove unreacted products and then dried at 37 • C overnight.

Estimation of CNTF concentration
Quantitative evaluation of the amount of CNTF in the conditioned medium of CNTF immobilized PGS/PCL scaffolds was conducted employing a plate-based enzyme-linked immunosorbent assay (ELISA).Briefly, the supernatants of CNTFimmobilized PGS/PCL scaffolds were collected at 6 h and, on days 10 and 20 after the immobilization, and were prepared for ELISA assessment to detect the protein levels of CNTF.ELISA measurements were performed according to the manufacturer's instruction (CNTF DuoSet ELISA; R&D Systems Inc., USA; DY557 15 plates).

Preparation of cell culture plate for an ex-vivo model
Circular 3D printed enclosures were employed in the design of the cell culture plate for an ex vivo model of rodent optic nerve damage.The enclosures were comprised of 3D printed poly(vinyl alcohol) (PVA) cylindrical structures 1.1 mm in diameter and 10 mm tall, produced using fused deposition modelling additive manufacturing process (Original Prusa i3 MK3S, Prusa Research, USA).The PVA enclosures were sterilized employing a process described in section 2.1, earlier.Each sterilized 3D PVA enclosure was attached to a PGS/PCL scaffold in each well of the 24-well cell culture plate using Tisseel VH Fibrin sealant (Baxter AG, Vienna, Austria).This was done to pre-position CNTF immobilized PGS/PCL scaffolds right at the centre of each 3D PVA enclosure in each well.Subsequently, circular pieces 1 cm in diameter of CNTF immobilized PGS/PCL scaffolds were placed inside the 3D PVA enclosures ensuring that PGS/PCL scaffolds remained at the bottom of all 24-well cell culture plates, a one 3D PVA enclosure containing one treated scaffold per well.
The REs (5 REs per single well) were seeded on surfaces of PGS/PCL scaffolds at a distance of 1.2 mm (the value was calculated according to rodent's optic nerves) around the 3D PVA enclosures and incubated at 37 • C for 2 h to allow the RE to attach.In this arrangement REs were circumscribed around the 3D PVA enclosures on CNTF immobilised PGS/PCL.During this time, the cell culture plates were manually agitated in a gentle circular motion.The cell culture medium was added to the as required as the permeable structure of the 3D PVA enclosure allowed unrestricted transfer of the culture medium from PGS/PCL scaffold to the REs.The cultured medium was changed by one third of the volume every other day.

Experimental groups and design for comparative analysis
Four groups (Groups I-IV) of different substrates were investigated to cross-compare and ascertain the CNTF-mediated effect on regeneration and directional elongation of RGC neurites.All REs cultivation processes and conditions were kept identical for all groups.The REs cultured on PLO-LA coated TCP, the PGS/PCL scaffold and, CNTF immobilized PGS/PCL scaffold were assigned to Groups I, II and III, respectively.In Group IV, the REs were seeded on PGS/PCL scaffold enclosed in the 3D PVA enclosure as explained in section 2.6.Graphically, the experimental Groups 1-IV are shown in figure 1.
A comparison between Group I and II intended to reveal a role of PGS/PCL scaffold play in the RGCs fate, between Group II and III would show the effects of CNTF grafting on PGS/PCL and, between Group III and IV, the influence of isolated CNTF concentration gradient on PGS/PCL could be further ascertained.On days 1, 10 and 20, following the RE culture, the RGC neurite measurement analysis was performed in all Groups.

Immunofluorescence assay
The immunostaining was carried out for all samples in all Groups after 1, 10 and 20 d of the retinal explanting.The samples were rinsed twice with PBS - and fixed in 4% paraformaldehyde for 20 mins at RT.The REs were permeabilised with 0.2% Triton X-100TM surfactant for microtubule-associated protein 2 (MAP2).The samples were then incubated overnight at 4 • C with the primary antibody: MAP2 (Sigma Chemical Co, St. Louis, MO, USA; MAB378, 1:200).Then, the samples were incubated with fluorescent secondary antibodies: Goat Anti-Mouse IgG antibody; fluorescein isothiocyanate (Sigma Chemical Co, St. Louis, MO, USA; AP124F, 1:100) for 1 h at 37 • C to make cells suited for screening using 532 nm visible light.DAPI (4 ′ ,6diamidino-2-phenylindole) (Sigma Chemical Co, St. Louis, MO, USA; D9564, 3 ng ml −1 ) was used for nuclear counterstaining of fixed cells for 10 min.Finally, the samples were mounted onto glass slides and analysed using a fluorescent microscope (Olympus BX-51, Olympus Corp., USA) equipped with Olympus Optical DP70 12.5MP CCD microscope camera.

Fluorescent labelling
The PGS/PCL scaffolds were labelled with the rhodamine (RhB) (Merck KGaA, Darmstadt, Germany).First, 8 mg of Bovine serum albumin (BSA) (Merck KGaA, Darmstadt, Germany) was dissolved in 2.0 ml of PBS and 8.4 mg of RhB was dissolved in 0.5 ml dimethyl sulphoxide (Merck KGaA, Darmstadt, Germany).The labelling procedures was performed by the addition of these two solutions together.The obtained mixture was shaken overnight at RT in darkness.Using a dialysis bag with cut-off of 2 kDa, the unreacted enzymes were separated from the mixture, then transferred to the buffer solution (PBS pH 7.4) in darkness and dialyzed with fresh buffer for 24 h.

Statistical analysis
Each specific experiment was conducted independently and duplicated three (3) times.Statistical comparisons between the Groups I-IV were carried out using a one-way analysis of variance followed by posthoc Tukey honest significant difference (HSD) test.The HSD differences were considered statistically significant for a p-value of less than 0.05.All error bars were presented as significant difference (SD).

Bioactive electrospun scaffold fabrication for CNTF delivery
In our study, the bioactive scaffolds were evaluated based on their CNTF release profile and their capacity to elongate RGC axons into the optic nerve head.Our earlier work [17] has shown the aligned PGS/PCL scaffolds provide an appropriate physical micro-environment and structural support for RGCs regeneration and support RGC axonal extensions and neurite outgrowth along with the neurite alignment.In the current study, the PGS/PCL scaffolds were investigated to their capacity to provide a suitable environment in the optic nerve damage area.Physical characterisation of PGS/PCL scaffolds including SEM imaging and PGS/PCL fibre diameter distributions and relative fibre alignment are available in [17].Notably, the average PGS/PCL fibre diameter was found to be 2.3 ± 0.3 µm.All samples studied displayed interconnected pore structures with spatial organization features with an average pore area of 73 ± 5 µm 2 and dimensions that mimic a natural ECM of the retina.Topologically such samples are capable of offering an improved cell attachment during transplantation and accelerating cell migration into the host retinal layers [17].

CNTF incorporation within the scaffolds
Physical adsorption, blend electrospinning, coaxial electrospinning and covalent immobilization are common strategies for incorporating neurotrophic factors within electrospun scaffolds [11,24].Among these methods, covalent immobilization has been investigated as one of the most favourable strategies for a sustained and controlled release of the growth factor [4,11,25] and it has been utilized to graft CNTF on PGS/PCL as described in section 2.2, earlier.
Figure 2 shows an ATR-FTIR spectrum of pure, unmodified PCL/PGS (blue), PLO-LA coated PCL/PGS (red), and PLO-LA coated PCL/PGS functionalised with CNTF (green).Vibrational mode assignments were made using the works of Barth [26] for proteins and, Holland-Moritz and Socrates for the core IR characteristic vibrational mode frequencies [27,28].The spectrum of unmodified PCL/PGS displayed noticeable photoluminescence, which was initially quenched with an application of PLO-LA and reduced even further with an application of CNTF.The spectra are dominated by strong primary polyethylene skeletal stretching bands at ∼1150 cm −1 and C=O ligands at ∼1710-1740 cm −1 that also overlap with PLO-LA and CNTF lipid contributions indicating that PGS/PCL blend displayed a high degree of ordering.The spectra of pure PGS/PCL, PGS/PCL PLO-LA coated and CNTF functionalised are essentially identical except for minor differences observed in vC-COO stretch regions, C-O regions and methylene group frequencies indicating minimal PGS cross-linking with PCL.Contributions from amide I and II bands are not visible in PLO-LA coated PGS/PCL and, PLO-LA and CNTF functionalised samples owing to exceptionally low concentration of PLO-LA and CNTF in solutions.Normally, IR measurements are only able to detect moderate amounts of protein in solution, typically in a range of 10-100 µg (or 0.1-1 mM) [29], whereas the amount of PLO-LA and CNTF used in this study is essentially an order of magnitude lower the IR detectable range and owing to a small volume of a sample, overall making amide I and II bands contributions undetectable.Additionally, active frequency modes from non-conjugated alkenes (C=C stretch) in the 1620-1680 cm −1 range and v 2 bending mode of water in the  1640-1645 cm −1 range overlap amide I and II vibrational contributions [26].
The level of CNTF after 6 h (or 0 d), 10 d and 20 d in the conditioned medium of CNTF immobilized PGS/PCL scaffolds was quantified by ELISA and shown in figure 3, with a number of days (D0, D10, D20) given on the X-axis.The findings confirm the presence of CNTF during the experimental period.
To assess the protein gradient provided by PGS/PCL scaffold, rhodamine (479 Da) was grafted on BSA and observed under fluorescence.Figure 4(a) presents rhodamine on the scaffold, with no protein grafted on the scaffold.As expected, no gradient was detected.Figures 4(b

The effect of CNTF guidance cues on RGCs growth
There is a growing tendency among researchers to utilize explanted organs to reduce the number of animals needed for experimentation and to achieve more accurate control over experimental parameters for regeneration studies [30,31].Also, it is ethically advantageous to use ex vivo models since no postsurgical animal care is needed.In the present study, an ex vivo model was developed to elucidate the effect of CNTF guidance cues on RGCs growth.To conduct a comparative investigation, the REs of rodents were cultured on TCP (Group I), on PGS/PCL (Group II), on the CNTF immobilized PGS/PCL (Group III), and the REs were circumscribed around 3D PVA enclosure on CNTF immobilised PGS/PCL ensuring a sustained release of CNTF from the 3D enclosures towards the REs (Group IV).
To assess the RGCs neurite outgrowth of the explants on all four Groups, the retinas were immunostained for MAP2 as shown in figure 5.
The detailed explant observation study on day 10 revealed that the cells showed proper adhesion and migration on all substrates.The average neurite lengths of REs grown on Group I, II, III and IV substrates were found to be 135 ± 21 µm, 180 ± 17 µm, 418 ± 51 µm and, 374 ± 43 µm, respectively, as displayed in figure 6  Comparison between the control (i.e.Group I) and Group II on day 10 and day 20 revealed that PGS/PCL scaffolds provided an appropriate affinity to RGC cells.Most of the RGCs in the REs exposed to CNTF (i.e.Group III and IV) displayed significantly longer neurite lengths (at p-value < 0.05) compared to Group I and Group II cells on day 10 and day 20 as shown in figure 6(a) for day 10, and figure 6(b) for day 20.Admittedly, no SDs (at p-value > 0.05) were observed between Group III and Group IV from the test (figure 6(a)) on day 10.The direct CNTF exposure on PGS/PCL scaffold facilitated Group III and Group IV to display pronounced neurite outgrowth (at p-value < 0.05) compared to RGCs exposed to CNTF only in the culture medium (i.e.Group I and Group II).On day 20, as shown in figure 6(b), the REs neurites were longer in Group III and IV (at p-value < 0.05) compared to Group I and II cells.The REs on the CNTF-immobilised PGS/PCL scaffolds placed in 3D enclosures (Group IV) displayed a gradual increase on day 20 of cell culture compared to RGCs cultured on the surface of CNTF immobilized scaffolds alone (i.e.Group III), as shown in figure 6(b), confirming that the Group IV model facilitated a stable cytokine secretion that potentially mimics an in vivo gradient micro-environment.
Consequently, to further isolate the effects of CNTF secretion in this model, an image processing algorithm was developed to conduct an area-based measurement to quantify the neurite growth of a dataset of ∼60 images of fluorescently labelled cells over a period of 20 d.In three (3) independent experiments, all REs examined showed an increase in their sizes and extended neurites.Figures 7(a)-(c) present one of Group IV's REs growth over this period.The average percentage of the area occupied by each RE was calculated from a binary image, figure (a')-(c'), of a stained image and presented in figure 7(d).
The centre of gravity data of each RE was calculated from the binary images and presented in figure 8. Figure 8(a) shows an example of a source MAP2 stained image of REs and figure 8(b) a processed binary image of the MAP2 stained image.Figures 8(c) and (d) illustrate the RGC neurite density distribution in the left and the right sites relative to the centre of gravity (equal to the centre of the RE), respectively.Data from figure 8 illustrates the REs display a higher density of neurites on the left side (denoted 'Left') near the 3D enclosure and, a higher concentration of CNTF compared on the right side (denoted 'Right').The average value of definite integrals from centre-to-end of the REs was found to be ca.2.6 × 10 4 pixel 2 (left site) and ca.1.5 × 10 4 pixel 2 (right site).Overall, according to the location examination, the explanted neurons mostly migrated towards the 3D printed enclosure, with higher CNTF concentration.
These findings further support a notion that the neurotrophic factor signalling gradients promote survival of neuronal cell types, including sensory and motor neurons.This study has shown that covalently immobilised CNTF on PGS/PCL offers a possibility of providing an instructional direction to neuronal cells through a permeable 3D enclosure.

Discussion
Axon terminals after injury are unable to uptake and transport neurotrophic factors via the axonal growth cone to the cell body (soma), leading to apoptosis [32], and as such, neurotrophic factor supplementation, including the growth factors, to the RGCs has been proven as a promising strategy for promoting RGC survival and regeneration [1,2,33].Growth factors and endogenous proteins attach to cell-surface receptors and direct cellular functionality, including tissue regeneration.Previous studies have successfully used CNTF to induce axon regeneration following an injury [3,5,24,34,35].Although a bolus injection of the neurotrophic factor is able to protect RGCs from cell death and restore their functions, the inherent protein instability in solution, quick diffusion, and an insufficient half-life of growth factors are the main challenges in this treatment approach [24].The sustained neurotrophic factor expression and maintaining neurotrophic factor bioactivity are critical to controlling the efficient therapeutic neurotrophic factor levels in the target area.Tissueengineered scaffolds were shown to be a potential solution to the aforementioned challenges.Promoted neuron growth has been reported by previous studies which used scaffolds for neurotrophic factor delivery [36,37].Our results also indicate that CNTF immobilised on PGS/PCL is capable of producing concentration of the growth factor within the structural scaffold network leading to the significant improvement in RGCs growth in a presence of grafted CNTF on PGS/PCL substrate.
In addition to providing a nourishing environment, creating a concentration gradient to simulate an ECM is also very effective for the targeted growth and direction of an axon.Biological studies have shown that providing a concentration gradient of growth factors or other biologically effective factors enhances cell survival in the desired path.Cells located in the vicinity of a concentration gradient are able to sense a gradient and respond to the gradient location by their directed outgrowth [38].For example, applying a concentration gradient of a specific neurotrophic factor causes cells that move towards or away from the source of a gradient release according to their specific surface receptors due to chemotoxic induction effects.The induced chemotoxisity can have a key role where a long-distance neurite growth is required.The earlier works by Song et al [39] and Rosoff et al [40] on the effects of concentration gradient on asymmetric growth of axons have shown that a uniform concentration gradient of the desired factor could be attained by utilising 3D gels that provide a gradient over a period of several hours.Also, a study by Matsuzaki et al showed that encapsulation of a nerve growth factor enclosed in a nanoparticle delivery system induces alignment and differentiation of PC12 cell axons [8].
As such, in this study, an ex-vivo experimental model comprised of an electrospun PGS/PCL biopolymer scaffold with covalently attached CNTF pluripotent neurotrophic factor was proposed, and such arrangement was used to generate CNTF concentration gradient that has proven to extended axonal growth of RGCs.The effect of CNTF gradual release on the PGS/PCL scaffold was ascertained based on the results obtained from the image processing algorithm, which confirmed that CNTF concentration was the highest at the source of secretion and reduced gradually as the gradient concentration gradually waned.
The results of measuring the length of RGC neurites in four experimental groups (Groups I-IV) confirmed that CNTF release over a period of 20 d had a measurable effect on RGCs growth.The growth rate of RGC neurites showed that the accumulation of axonal growths in the cultured tissue sections close to CNTF source was significantly higher.We trust the results of this study provide a useful and simple method to aid extracellular substrate simulations that will enable the development of a new generation of tissue-engineered scaffolds suitable not only for regenerating of a lost or a damaged nerve tissue but also for providing an appropriate concentration gradient for other biological applications, which can also focus on the process of improved regeneration of damaged body tissues.

Summary and conclusions
Recent research findings focused on the optic nerve neuropathies have shown that providing an appropriate growth factor supplementation, including CNTF, can be highly promising to RGC regeneration on artificial biopolymer scaffolds.Various substrates have been assessed for a suitable growth factor delivery, each aiming to offer sustained concentrations of growth factor in the target area while mimicking a natural ECM environment.In this study, CNTF covalently immobilized on a surface of PGS/PCL scaffolds and enclosed in a simulated physical 3D model was investigated for gradual CNTF release and was found to encourage the RGC neurite outgrowth and providing a directional instruction to RGC axons.The RGC REs cultured on the model significantly increased their neurite extensions after 20 d of cell culture.The increased RGC neurite dimensions and a positive neurite outgrowth have shown that CNTF mediated PGS/PCL scaffold can provide a continuous concentration of the gradient within the limits of the proposed ex vivo experimental 3D model of rodent optic nerve degeneration.However, additional in vivo studies are required to explore further the potential therapeutic effects of such findings in restoring vision.In this context, identifying strategies to optimize the tissue-engineered bioactive scaffold's regenerative capacity is the next challenge, and it needs to be addressed prior to the implementation of this and/or a similar type of grafted factor arrangements in clinical applications.It also implies that the studied PGS/PCL:CNTF system should undergo an additional examination to determine the degree and extent of functional group interactions and, the estimation of the CNTF concentration across individual areas on the scaffold surface.

Figure 1 .
Figure 1.Experimental groups (Group I-IV) including, Group I with the REs were cultured on TCP, Group II with REs seeded on PGS/PCL, Group III with the REs cultured on CNTF immobilized PGS/PCL and, Group IV with the REs circumscribed around the 3D PVA enclosure on PGS/PCL in the presence of CNTF gradient.
) and (c) show the gradient of BSA in 30 mins and, in 1 h, respectively, after the initial burst release.The fluorescence intensity was calculated for the functionalized PGS/PCL scaffolds with rhodamine along the Y-axis and shown in figure4(d).The downward trend in the fluorescence intensity plot confirmed formation of concentration gradient.

Figure 4 .
Figure 4. Fluorescence imaging of (a) rhodamine, and the gradient of BSA in 30 mins and (b) in 1 h, and (c) after the initial burst release on PGS/PCL; (d) the fluorescence intensity of the functionalized PGS/PCL scaffolds with rhodamine 2 h after the initial burst release.

Figure 5 .
Figure 5.The MAP2 staining and DAPI images of REs on day 10 and day 20 of culture on control (Group I) (top), unmodified PGS/PCL polymeric scaffold (Group II) (second from the top), on CNTF immobilized PGS/PCL (Group III) scaffold (second from the bottom) and, on PGS/PCL with CNTF gradient positioned within the 3D enclosure (Group IV) (bottom images).

Figure 6 .
Figure 6.The average neurite lengths of neurons in µm on (a) day 10 and (b) day 20 on control (Group I), unmodified PGS/PCL polymeric scaffold (Group II), on CNTF immobilized PGS/PCL (Group III) scaffold and, on PGS/PCL with CNTF gradient positioned within the 3D enclosure (Group IV).

Figure 7 .
Figure 7. MAP2 staining of one of Group IV's REs shown on left image column, images (a)-(c) and related binary images shown on right monochrome image column, images (a')-(c') on the designed ex vivo model after (a) 1 d, (b) 10 d, (c) 20 d of culture and (d) the area occupied by the REs over 20 d in three independent tests.

Figure 8 .
Figure 8.The REs on the designed ex vivo model (a) MAP2 staining with the CNTF gradient decreasing from the left to the right of the image; the CNTF immobilized on PGS/PCL are placed inside the 3D printed enclosure, and (b) a binary processed image of the MAP2 stained image; (c)-(d) density distribution profile of the REs showing the left, image (c), and the right, image (d), of the explant.