Delivery of Alginate Scaffold Releasing Two Trophic Factors for Spinal Cord Injury Repair

Spinal cord injury (SCI) has been implicated in neural cell loss and consequently functional motor and sensory impairment. In this study, we propose an alginate -based neurobridge enriched with/without trophic growth factors (GFs) that can be utilized as a therapeutic approach for spinal cord repair. The bioavailability of key GFs, such as Epidermal Growth factor (EGF) and basic Fibroblast Growth Factor (bFGF) released from injected alginate biomaterial to the central lesion site significantly enhanced the sparing of spinal cord tissue and increased the number of surviving neurons (choline acetyltransferase positive motoneurons) and sensory fibres. In addition, we document enhanced outgrowth of corticospinal tract axons and presence of blood vessels at the central lesion. Tissue proteomics was performed at 3, 7 and 10 days after SCI in rats indicated the presence of anti-inflammatory factors in segments above the central lesion site, whereas in segments below, neurite outgrowth factors, inflammatory cytokines and chondroitin sulfate proteoglycan of the lectican protein family were overexpressed. Collectively, based on our data, we confirm that functional recovery was significantly improved in SCI groups receiving alginate scaffold with affinity-bound growth factors (ALG +GFs), compared to SCI animals without biomaterial treatment.

Spinal cord injury. The SCI was induced using the modified balloon-compression technique according to our previous study 31 . Briefly, 2-French Fogarty catheter was inserted epidurally at Th8-9 level and the balloon was inflated with 12.5 μ l of saline for 5 min. After compression of spinal cord tissue, catheter was deflated and removed from epidural space. In the sham group (n = 4), the catheter was inserted at the same level of spinal cord, but balloon was not inflated and no lesion was performed. Manual bladder expression was required for 7-14 days after the injury until the bladder reflex was established. No antibiotic treatment was used.
Tissue protein extraction. Fresh frozen spinal cord collected after 3, 7 and 10 days after lesion and controls were embedded (n = 4, each group) on optimal cutting temperature polymer before sectioned using a cryomicrotome (Leica Microsystems, Nantere, France) and subjecting to trypsin digestion. Spinal cord tissue sections (20 μ m thick) were mounted on a parafilm covered glass slide and the tissue was microdissected manually using a binocular. The pieces were extracted by incubating in 20 μ L of 50 mM bicarbonate buffer containing 50 mM dithiothreitol and 1% SDS at 55 °C for 15 min. The extracts were then loaded on 12% polyacrylamide gel and separated at 70 V for 15 min and then 120 V until the dye front reaches the other end of the gel. After migration, the gel was incubated in the gel fixative solution for 30 min and stained with colloidal Coomassie brilliant blue overnight. The stain was removed by washing the gel four times with distilled deionized water 30 . In gel digestion. The gel was cut into ten pieces. Pieces were washed with 300 μ L of distilled deionized water for 15 min, 300 μ L of ACN for 15 min and 300 μ L of NH 4 HCO 3 (100 mM; pH8) for 15 min Scientific RepoRts | 5:13702 | DOi: 10.1038/srep13702 followed by incubation of 300 μ L of NH 4 HCO 3 /acetonitrile (ACN) (1:1, v/v) for 15 min and 300 μ L of ACN for 5 min. Band pieces were dried in a Speedvac for 5 min. The reduction of cystine residues was made with 50 μ L of 10 mM of DTT in NH 4 HCO 3 100 mM (pH8). Pieces were incubated at 56 °C for 1 hour. Alkylation of cystine was made with 50 μ L of of iodo acetamide IAA (50 mM ) in NH 4 HCO 3 (100 mM; pH8). Pieces were incubated at room temperature in the dark for 30 min. Band pieces were washed a second time with 300 μ L of NH 4 HCO 3 100 mM (pH8) for 15 min, then with a mix of 300 μ L of NH 4 HCO 3 /ACN (1:1, v/v) for 15 min and 300 μ L of ACN for 5 min. Band pieces were dried in a Speedvac for 5 min. A digestion of band pieces was made with trypsin (12.5 μ g/mL) in NH 4 HCO 3 20 mM (pH8), enough to cover pieces. Pieces were incubated at 37 °C overnight. Peptides were extracted on shaking platform with 50 μ L of formic acid (FA) 1% two times for 20 min, then 150 μ L of ACN for 10 min. The supernatant was transferred in new tube and dried with Speedvac. Samples were resuspended in 20 μ L of 0.1% trifluoro acetic acid (TFA) and then desalted with Ziptip C18 and eluted with 10 μ L of ACN/0.1% TFA (8:2, v/v). Samples were dried in a Speedvac and resuspended in 15 μ L of ACN/0.1% FA (2:98, v/v).

NanoLiquid Chromatography -High Resolution-MS/MS (NanoLC-HR-MS/MS). Samples were
separated by online reversed-phase chromatography using a Thermo Scientific Proxeon Easy-nLC system equipped with a Proxeon trap column (100 μ m ID × 2 cm, Thermo Scientific) and a C18 packed-tip column (100 μ m ID × 10 cm, Nikkyo Technos Co. Ltd). Peptides were separated using an increasing amount of acetonitrile (5%-30% over 120 minutes) at a flow rate of 300 nL/min. The LC eluent was electrosprayed directly from the analytical column and a voltage of 1.6 kV was applied via the liquid junction of the nanospray source. The chromatography system was coupled to a Thermo Scientific LTQ-Orbitrap XL mass spectrometer. The LTQ-Orbitrap XL instrument was set to acquire top 20 MS/MS in data-dependent mode. The survey scans were taken at 70,000 full width at half maximum (FWHM) (at m/z 400) resolving power in positive mode and using a target of 3E6 and default charge state of 2. Unassigned and + 1 charge states were rejected, and dynamic exclusion was enabled for 20 s. The scan range was set to 300-1600 m/z. For the MS/MS, 1 microscan was obtained at 17,500 FWHM and isolation window of 4.0 m/z, using a scan range between 200-2000 m/z 30,32 . Mass Spectra Data Analysis. Tandem mass spectra were processed with Thermo Scientific Proteome Discoverer software version 1.3. Resultant spectra were searched against the Swiss-Prot ® Rattus norvergicus database (version January 2012) using the SEQUEST ® algorithm. The search was performed choosing trypsin as the enzyme with two missed cleavages allowed. Precursor mass tolerance was 10 ppm, and fragment mass tolerance was 0.5 Da. N-terminal acetylation, methionine oxidation and arginine deamination were set as variable modifications. Peptide validation was performed with the Percolator algorithm. Peptides were filtered based on a q-Value below 0.01, which corresponds to a false discovery rate (FDR) of 1%. All the MS data were processed with MaxQuant 33 (version 1.5.1.2) using Andromeda 34 search engine. Proteins were identified by searching MS and MS/MS data against Decoy version of the complete proteome for Rattus norvegicus of the UniProt database [UniProt Consortium. Reorganizing the protein space at the Universal Protein Resource (UniProt). Nucleic Acids Res. 2012, 40 (Database issue), D71− 5.] (Release June 2014, 33675 entries) combined with 262 commonly detected contaminants. Trypsin specificity was used for digestion mode, with N-terminal acetylation and methionine oxidation selected as variable, carbarmidomethylation of cysteines was set as a fixed modification and we allow up to two missed cleavages. For MS spectra an initial mass accuracy of 6 ppm was selected and the MS/MS tolerance was set to 0.5 Th for CID data. For identification, the FDR at the peptide spectrum matches (PSM) and protein level was set to 0.01. Relative, label-free quantification of proteins was done using the MaxLFQ algorithm 35 integrated into MaxQuant with the default parameters.
The data sets used for analysis are deposited at the ProteomeXchange Consortium 36 (http://proteomecentral.proteomexchange.org) via the PRIDE partner repository 37 with the dataset identifier.
Analysis of the proteins identified was done using Perseus software (http://www.perseus-framework. org/) (version 1.5.0.31). The file containing the information from identification was used and hits to the reverse database, proteins only identified with modified peptides and potential contaminants were removed. Then the LFQ intensity were logarithmized (log2(x)). A normalization was achieved using a Z-score with a matrix access by rows. Data coming from control samples were average as the one coming from the lesion part. Six conditions were then analyzed: control (ctrl), lesion part (lesion), segment R1 or C1 seven days (respectively r1_7D and C1_7D) or ten days (r1_10D and C1_10D) after lesion. Only proteins presenting a valid value of LFQ intensity for these six conditions were used for statistical analysis. A Hierarchical clustering was first performed using a Pearson correlation for distance calculation and average option for linkage in row and column trees using a maximum of 300 clusters. For visualization of the variation of proteins expression depending to the segment/time parameter, the profile plot tool was used with a reference profile and an automatic selection of the 10 or 15 correlated profiles.
Preparation of alginate scaffold with affinity-bound factors. Fabrication of the scaffold with the affinity-bound dual growth factors (ALG+ GFs) involved preparing bioconjugates of bFGF and EGF with alginate-sulfate and then mixing both bioconjugate solutions with the solution of a partially calcium-cross-linked alginate. The bioconjugates were prepared by mixing bFGF or EGF with alginate-sulfate solution (1%, w/v) and incubating for 1.5 h at 37 °C, to allow equilibrium binding. The partially calcium-cross-linked alginate solution was prepared as previously described 38 . Briefly, stock solutions of sodium alginate (VLVG, 30-50 kDa, > 65% guluronic acid content, NovaMatrix FMC Biopolymers, Drammen, Norway) and D-gluconic acid/hemi calcium salt were prepared by dissolving the materials in DDW and stirring at room temperature. Each solution was filtered separately through a sterile 0.2 μ m filter membrane into a sterile container in a laminar flow cabinet. Equal volumes from each stock solution (2.08% and 0.68% (w/v) for VLVG alginate and D-gluconic acid, respectively) were combined by extensive homogenization for several minutes to facilitate homogenous distribution of the calcium ions and cross linking of alginate chains. Finally, the bFGF and EGF alginate-sulfate bioconjugates were mixed with the partially cross-linked alginate to yield an bFGF/EGF-containing, affinity-binding alginate scaffold (0.1% alginate-sulfate, 0.9% alginate, 0.3% D-gluconic acid, w/v) (ALG+ GFs). For the control system-lacking GFs, the scaffold was prepared with no affinity-bound factors (ALG).
Intraspinal delivery of alginate scaffold. Seven days after SCI, animals were anesthetized with 1.5-2% halothane and partial laminectomy at Th6-12 level was performed. Using a 50-μ l Hamilton syringe (30G needle, Cole Parmer, Anjou, Quebec) connected to UltraMicroPump III with Micro4 Controller, 4-Channel (World Precious Instruments, Inc., Sarasota FL) and stereotactic device, 4 intraspinal injections/per animal were applied at the lesion site that showed discreet signs of haemorrhage and slight atrophy. In most cases the lesion cavity was apparent through the dorsal site of spinal cord. Bilateral delivery of i) saline, ii) ALG, or iii) ALG+ GFs (2 injections of 2 μ l/per injection/on left and 2 injections of 2 μ l/per injection/on right side with delivery rate of 0.5 μ l/min, loaded with 200 ng/ml of each GF) was performed. Based on our in vitro study results, an affinity-binding alginate scaffold loaded with 200 ng of bFGF/EGF confirmed long term release of GFs 25 . Each delivery was positioned 1 mm from the spinal cord midline and injected at the depth of 1.8-2 mm from the pial surface of the spinal cord. The distance between injections was 1 mm, avoiding vessels. After injecting the dose of alginate scaffold, the needle was maintained in the tissue for an additional 30 seconds. No antibiotic treatment was performed during animal's survival.

Anterograde biotinilated dextran amine (BDA) motor corticospinal tract (CST) axon tracing.
Sham rats (n = 4) and SCI rats (ALG (n = 3), and ALG+ GFs (n = 3)) at 3 weeks post-injury were anesthetized with 2% halothane and placed in a stereotaxic device. The halothane level was maintained at 2-3% throughout the surgery. An incision was made to expose the skull and to identify the bregma and lambda landmarks. Rats received injections of 10% solution of BDA (biotinilated dextran amine 10,000 MW; Molecular Probes, Eugene, OR) in sterile 10 mM sodium phosphate buffer, pH 7.4, injected via glass micropipettes (inner tip diameter of 60-80 μ m) using a controlled pressure device (PicoPump; World Precision Instruments). The injection site was positioned into right and left motor cortex performed at anatomical coordinates: 1.0 mm lateral to bregma, 1.5 mm anterior/posterior to the bregma and 1.5 mm deep to the cortical surface from the pial surface of the brain based on the Stereotaxic Coordinates (Paxinos and Franklin, 2001) (Supplementary Figure 1). A total 8 injections with approximately 0.5 μ l of BDA was injected at each of the four sites at a rate of 80 nl/min during 6-7 min/per injection. The micropipette remained in place for 3 minutes following each injection. After the delivery was completed, the skin overlying the skull was sutured and rats returned to their cages. Behavioral Testing. BBB scoring. Animals were behaviourally tested for 5 min using BBB open-field locomotor test 39 after SCI at day 1, 3, 5, 7, and then in weekly intervals. Each rat was tested for 5 min by two blinded examiners. BBB test measuring locomotor outcome (hindlimb activity, body position, trunk stability, tail position and walking paw placement) of rats by BBB rating scale ranges from 0 -no observable hindlimbs movements to a maximum 21 -plantar stepping, coordination and trunk stability like healthy rats.
Cold allodynia. Cold sensitivity of the hindpaws to the acetone was quantified by foot withdrawal frequency. All animals were tested at day 3 post-injury and then in weekly intervals at day 16, 25, 32 and 49 after SCI. Before testing, the rats were left to acclimatize inside acrylic-plastic cages during the 10-15 min. A drop of acetone (50-100 μ l) was applied to the left and right hindpaws using a plastic syringe, 5 times, with at least 5 min recovery between administrations. The number of brisk foot withdrawals or flinching were considered to be positive behaviour. Data are presented as mean response duration (in seconds). Statistical differences between groups were determined with an unpaired Student's t test.
The sequence of surgical and behavioural procedures performed in time are described in Supplementary  Figure 2.
Serial sagittal sections from each animal were stained for BDA to determine labelled CST axons through the lesion site. Sections were washed in TBS (50 mM Tris/HCl, 150 mM NaCl, pH = 7.6) and subsequently incubated overnight at 4 °C with ABC kit (ABC Vectastain Ellite kit; Vector Laboratories) diluted in TBS. After rinsing 3 times for 10 minutes in TBS, sections were reacted with DAB (consisting of DAB substrate, DAB buffer and 0.6% NISO 4 ) for 20 minutes. Sections were finally washed in TBS and coverslipped in Enthelan. Similarly serial sagittal sections (n = 5) were stained with Luxol Fast Blue to determine the length and cavity size area 40 . Schematic concept of dissected spinal cord segments consisting of rostral (Th5-7), central (Th8-9), caudal (Th10-12) segments processed for immunohistochemical analyses is included in (Supplementary Figure 3).

Quantification analysis.
Immunochemicaly stained sections were analyzed using Olympus BX-50 fluorescent microscope at 4x, 10x 20x and 40x magnifications, captured with digital camera HP Olympus and analyzed by Image J software according to the previous protocol [41][42][43] Figures 4 A,B). Quantification of immunofluorescence intensity (CGRP, Iba1, SYN) rostrally/caudally from lesion site and GFAP, Iba1 also at the site of central lesion) was performed by using ImageJ software. Captured digital images were transformed into monochrome 8-bit images and determined the mean grey level number of black and white pixels within the tissue (value 0-255, when 0 = black pixels, 255 = white pixels). The final result yields the mean ratio of black and white pixels expressed by the histogram. Length of BDA were equally evaluated by Image J and expressed in mm. Morphometric analyses of cavity size were performed on five 1.6 cm sagittal sections from the lesion site of each experimental and sham tissue. Modified Luxol Fast Blue labelling was performed to evaluate cavity area in spinal cord sections 40 . Mean number of cavitation of experimental groups was expressed by mm relative to Sham spinal cord, which was without cavitation's and represent zero (no cavity).
Data and statistical analysis. Obtained data from tissue analyses and behavioural testing were reported as mean ± SEM. Mean values among different experimental groups were statistically compared by one-way ANOVA and Tukey's post hock tests using Graph pad PRISM software. Values of P < 0.05 were considered statistically significant (*P value of < 0.05, **P value of < 0.01, ***P < 0.001).

Results
Spatiotemporal proteomic study of spinal cord tissue after injury. Spinal cord tissues, the rostral, lesion and caudal segments, were collected at 7 and 10 days after lesion and were subjected to tissue microproteomic, MaxQuant proteins analyses followed by Perseus allowed to statistically validate the identification and performed clustering. Figure 1 clearly shows that the control tissue is on a separate branch from lesion, rostral and caudal segments. The time course study reflected that 7 and 10 days are separated from each other. Comparison of the data obtained at 7 days between rostral and caudal segments clearly shows a common cluster of over-expressed groups of proteins (Fig. 1a) and four different clusters between the rostral and caudal of over-or sub-expressed protein groups (Fig. 1b-e). Proteins overexpressed at 7 days (Fig. 1a) are sub-expressed at 10 days (Fig. 1f). Only a cluster of proteins is differentially overexpressed at the caudal level (Fig. 1g). Specific proteins overexpressed in lesion, rostral and caudal segments are presented in Table 1. Radixin, Neural cell adhesion molecule, COP9 signalosome complex (CSN); Cofilin 2; AP-2, dynamin-like, Rab-7a, GST_P are specific proteins that are overexpressed only in caudal regions at 7 and 10 days. In rostral segment 7 and 10 days after SCI, amphiphysin, hydroxyacyl glutathione hydrolase, inositol monophosphatase, Neurofilament heavy polypeptide, Glycogen phosphorylase, Superoxide dismutase [Cu-Zn], phosphoglycerate mutase 1, septin 8, microtubule-associated protein 1A, CD59 glycoprotein, LETM1 and EF-hand domain-containing protein 1, Elongation factor Tu, Succinate dehydrogenase [ubiquinone] flavoprotein subunit, Vesicle-associated membrane protein-associated protein B are the ones specifically overexpressed.
Cavity size. During the first week post-injury, a severe inflammatory response occurs at the central lesion. The secondary damage processes lead to cell death and development of cavitations at the epicenter and along the rostrocaudal axis of the spinal cord 40 . In order to fill the cavity and create a permissive environment for regeneration, we administered the liquid form of alginate scaffold directly to the lesion cavity at 7D post injury. Histological assessment of spinal cord sections stained with Luxol Fast Blue revealed cavity area reduction in SCI+ ALG+ GFs and SCI+ ALG groups compared to SCI+ SAL at 42 day post-implantation (Fig. 3). Quantitative stereological analyses of tissue fenestration in 1.6 cm segment revealed significant (**P < 0.01, ***P < 0.001) reduction of cavitation (analysing length and area of cavity) after application of ALG+ GFs (cavity length/3.3 ± 1.5 mm/area/0.56 ± 0.2 mm 2 ) and ALG (cavity length/5.4 ± 1.2 mm/area/1.13 ± 0.2 mm 2 ) compared to the saline treatment (cavity length/7.7 ± 1.4 mm/ area/1.96 ± 0.3 mm 2 ) (Fig. 3).   Caudal/Laminae I-IV 120.7 ± 12.7; Laminae VIII-IX24 ± 3.8) (Fig. 4). The differences in NeuN positive profiles show a statistical significance between individual experimental groups: Sham, SCI+ SAL, SCI+ ALG, SCI+ ALG+ GFs ***P < 0.001, **P < 0.01, *P < 0.05.

ChAT labeled motoneurons. The average number of ChAT positive cells in the SCI+ SAL, SCI+ ALG
and SCI+ ALG+ GFs groups was compared to confirm the hypothesis whether neuronal sparing has included motor neurons of ventral horns. Rostral to the lesion site, the number of spared ChAT+ neurons within the ventral horns significantly increased (*P < 0.05) following alginate biomaterial treatment (10 ± 2.1/SCI+ ALG; 10.9 ± 1.7/SCI+ ALG+ GFs) when compared to the control saline group (7.4 ± 0.9) (Figs 5 and 6). Significant differences in sparing of motor neurons (*P < 0.05, ***P < 0.001) were also recorded among experimental groups caudal to the injury site, although the average number of positive cells had declined (6.9 ± 1.5/SCI+ ALG+ GFs, 5.4 ± 0.9/SCI+ ALG, 2.8 ± 0.8/SCI+ SAL, 11.9 ± 1.9/ Sham) compared to spinal rostral part (Figs 5 and 6). Our results demonstrate that alginate biomaterial implantation resulted not only in common NeuN positive neurons sparing, but also in the specific sparing of endogenous ChAT+ motor neurons.

Synaptic vesicles alterations.
In the spinal cord of Sham and both SCI-SAL and SCI-ALG groups of treated rats, synaptophysin immunoreactivity (SYN+ IR) appeared as numerous diffusely distributed fine dots along the surface of motor neurons and their proximal dendrites, and delineated their polygonal contours (Fig. 7A,B). However, after ALG+ GFs treatment, the density of SYN+ vesicles around remaining CHAT+ motor neurons of the anterior horns strikingly increased when compared to all experimental groups (Fig. 7C,D). The immunoreactive profiles appeared as coarse granules of different size that were also distributed on motor neuron surface. Quantitative analysis of SYN+ vesicle expressed as % of SYN+ positive vesicles within identical fields of anterior horns in all experimental groups confirmed significant increase in ALG+ GFs treated group, particularly caudally to the epicentre of injury (Fig. 7D). Interestingly, we did not see any differences in the density of SYN+ positive vesicles within segments above the lesion site (data not shown).

CGRP positive fibres. CGRP immunoreactivity was observed in all experimental groups
(SCI+ ALG+ GFs, SCI+ ALG, SCI+ SAL) in fibres and punctuate terminals of superficial dorsal horn (Laminae I-III) and LT (LT-Lissauer's tract) area 44 located along the lateral edge of the dorsal horn and medial grey mater (Fig. 8). Moreover, depending on the experimental group, few individual CGRP positive fibres extending from Lamina III toward Laminae V (0.226 ± 0.099 mm/SCI+ SAL) and VII (Figs 8  and 9) were detected. The longest CGRP+ fibres with the average of length 0.301 ± 0.103 mm were observed after administration of alginate biomaterial alone and alginate biomaterial with affinity-bound GFs to the injured spinal cord (0.301 ± 0.103 mm/SCI+ ALG; 0.27 ± 0.053 mm/SCI+ ALG+ GFs). Sham spinal cord didn't contain CGRP positive fibres extended into the intermedia spinal cord layers; however CGRP terminals within superficial dorsal horn were frequently observed. The differences associated with length of fibres show statistical significance (**P < 0.01, *P < 0.05) only between Sham and other experimental groups (SCI+ SAL, SCI+ ALG, SCI+ ALG+ GFs) (Fig. 9). The number of immunolabeled CGRP fibres varied among individual experimental groups and areas of spinal cord. The most numerous CGRP positive fibres, forming bundles -like structures were observed in rostral segments from the lesion site after the delivery of alginate biomaterial with the affinity-bound GFs (6.7 ± 2.3/SCI+ ALG+ GFs, 4.5 ± 4/SCI+ ALG, 4.2 ± 3.3/SCI+ SAL, 1 ± 1/Sham). Average number of positive fibres was slightly decreased caudally to the epicentre of injury (5.9 ± 4/SCI+ ALG+ GFs, 4.8 ± 4.5/SCI+ ALG, 4.1 ± 3/SCI+ SAL, 1 ± 1/Sham) (Figs 8 and 9). Among the individual experimental groups in both studied parameters we did not observe statistical differences (Fig. 9).
Axonal sprouting via BDA tracing. BDA delivery to the sensorimotor cortex served to label descending CST axons of spinal cord. In sham animals, BDA-labelled CST axons were detected along the entire length of sagittal sections (16 mm) of the spinal cord; more specifically in the ventral part of the dorsal column, where stripe of organized BDA positive axons occurred (16 mm ± 0) (Figs 10 and  11). After spinal cord injury, CST axons appeared disorganized, ended above the lesion site and many cut BDA axons formed terminal structures like buttons. Re-growth of CST fibres into denervated areas of spinal cord was monitored following alginate administration. Moreover, the alginate biomaterial alone and with affinity-bound EGF/bFGF promoted increased re-growth of few BDA positive fibres through the central lesion with occasional innervations below the lesion site (2.9 mm ± 0.7 from a total 16 mm length of section) compared to saline treatment (0.6 mm ± 0.1) (Figs 10 and 11) (*P < 0.05).

Angiogenesis. For visualization of the vascular structures, the endothelial cell marker von Willebrand
Factor (vWF) was used (Supplementary Figure 7). Numerous positive blood vessels were observed in the Sham group, mainly in the white matter compared to the injured spinal cord groups (SCI+ SAL or SCI+ ALG), where the density of blood vessels decreased in close vicinity of the lesion site (Supplementary Figure 7). Treatment with alginate biomaterial with the affinity-bound GFs resulted in an increase of vWF positive blood vessels in the white matter and in grey matter as well, at lesion site. Results obtained from immunohistochemical analyses suggest that GFs-enriched alginate biomaterial created a suitable environment for blood vessels survival or reconstruction, but without significant differences between treatment groups (data not shown).

Discussion
Currently, the field of SCI neurotherapeutics is still in its infancy and there are no effective ad approved therapies for SCI in humans. A contributing factor for such failed neuroregenerative processes has been attributed partly to the development of the optimal regeneration-supportive microenvironment that can initiate a neurobridge connecting disconnected spinal cord segments.
The present study clearly demonstrates that the local delivery of injectable alginate biomaterial capable of increasing the bioavailability of key growth factors such as bFGF and EGF and their appropriate presentation improve the repair of SCI through multiple mechanisms, such as: i) reducing the central lesion cavity, ii) increasing the number of surviving neurons including ChAT+ motor neurons and their synaptic connections, iii) enhancing outgrowth of CST axons, iv) preserving or stimulating formation of new blood vessels, and v) attenuating inflammation; which altogether enhance the functional recovery after SCI without sensory impairments.
Here we applied a well-characterized compression model of spinal cord injury leading to overall impairment of motor and sensory functions associated with loss of corresponding neuronal pools, overreaction of microglia/astrocytes and inability of axonal regrowth through the lesion site 45 . Using a tissue micro-proteomic approach, we established that in the time period after lesion, the nature of proteins varied throughout the spinal rostro-caudal axis. Particularly, the proteins found in caudal and rostral segments at 7 and 10 days after SCI were different compared to 3 days post injury 40 . Previous data clearly show that three days after lesion, the factors secreted in the lesion and rostral segments are anti-inflammatory and neurotrophic, while in the caudal region a cocktail of apoptotic and neurotoxic proteins predominate 30 . The present study shows that on days 7 and 10 after SCI, in the caudal segments neurotrophic factors are overexpressed, as well as adhesion molecules and signalling proteins. In contrast, in rostral segments the proteins overexpressed are involved in metabolism at the level of the mitochondria or the cytoplasm, as well as in intracellular signalling. This clearly indicates that real differences exist between the rostral and caudal segments in terms of physiological and molecular processes, and that these differences are dynamic in time. Importantly, the results indicate that the caudal region possesses all the factors that can stimulate neurite outgrowth, but these seem to be insufficient in amount and are blocked by proteoglycans. Taking into account these ex vivo data, we attempted to connect the rostral and caudal segments through the lesion by constructing an alginate biomaterial bridge loaded with GFs. Thus all immunohistochemical and tracing analyses were performed along the rostro-caudal axis, to better understand and define differences in pathological or regenerative processes above and below the lesion site after biomaterial treatment.
Our strategy is in line with recent pre-clinical studies performed after incomplete/complete injuries, and attempting to reconnect links with the tissue below the injury site, either bypassing the central lesion or rebuilding tissue in a cyst mediated via the application of biomaterials [46][47][48][49][50] . The novelty of our strategy is the combination of biomaterial used as a bridge together with sustained delivery of key growth factors for SCI repair. In vitro, this combination was found to be effective in promoting cell retention and expansion, while also enabling neural progenitor lineage differentiation in situ 25 . In continuity with these findings, our in vivo results document significant spinal cord tissue sparing, resulting in neuronal sparing that may lead to enhanced plasticity and reorganization of preserved neuronal circuits. Furthermore, the sparing of ChAT positive motor neurons may correlate with the trend of motor function improvement observed during the whole survival period in SCI rats treated with the alginate scaffold with or without GFs, in comparison to animals treated with saline.
Physiological locomotion is governed by motoneurons that receive synaptic inputs from local interneurons, descending pathways and proprioceptive sensory neurons. The convergence of proper excitatory and inhibitory inputs on motoneurons mediated by synaptic connections is required for motor control, reflexes and tonic firing of the motoneurons. Disruption of the cellular components and/or synaptic connectivity in this spinal circuitry has been implicated in motoneuron spasticity and various motoneuron disorders 45 . For this reason in this study we followed the response of ChAT motor neuron-related synaptic vesicles in segments above and below the central lesion using synaptophysin immunohistochemistry (SYN+ IR). Synaptophysin is the most abundant integral membrane protein of synaptic vesicles 51 and can be used as a marker protein of synaptic vesicles in the central and peripheral nervous systems 52 . The present data document that SYN+ IR around motor neurons in the anterior horns showed similar patterns in most experimental groups, except for the group receiving ALG+ GFs. In these rats, we observed more intense SYN+ IR in the caudal compared to the rostral segments. These results may be linked with our proteomic data, confirming the higher expression of neurotrophic and synaptogenetic factors in the caudal segment, thus producing a favourable environment for synaptic rebuilding reflected by increased SYN+ IR after ALG+ GF delivery. Although our proteomic findings respond to 10 days survival following SCI, the higher level of synaptogenetic factors may be further potentiated with a GF-enriched environment, as most likely seen in the present study with ALG+ GF delivery. The mechanisms mediating increase in SYN+ IR may reflect several processes such as: i) up-regulation of synaptic functions after SCI, which is more likely related to the release of excitatory amino acids, or ii) may indicate plastic changes associated with formation of new synapses. Thus, to further understand changes in motoneuron synaptic connectivity after SCI treatment, the transporter systems such as vesicular glutamate and glycine transporters (VGluT1/VGluT2, GlyT2) need to be further studied.
Furthermore, SCI-induced secondary pathological processes also cause interruption of the CST tract 53,54 , leading to partial or complete impairment of descending motor pathways for skilled movements below the injury 2,55 . The compression model used in the present experimental study carried out at thoracic levels caused interruption of axon fibres corresponding to both hindpaws with some degree of spontaneous regeneration and behavioural improvements 56 . The behavioural outcome can be enhanced by promoting the axonal integrity and plasticity of the corticospinal tract and descending serotonergic pathways via GF delivery 57 . In according with these finding, our data confirm significant re-growth of BDA positive fibres observed after intraspinal injection of GF-enriched alginate biomaterial at the central lesion 58 . In contrast, delivery of alginate biomaterial alone did not induce the same effect as the GF-enriched biomaterial. The neuroprotective effect of biomaterial on axon regrowth has been described in many other in vitro and in vivo studies, where in vitro studies demonstrate that biomaterial promotes neural cell attachment and neurite outgrowth [59][60][61] while in vivo studies show only partial regeneration after gel is implanted in the injured spinal cord 23,62,63 . The explanation for differences in axonal outgrowth seen between in vitro and in vivo may be given by multiple factors associated with inhibitory, immune, endocrine processes that are typical for the complex in vivo environment 21,64,65 . In addition, optimal regeneration of axons requires preserved vascular supply. Our data indicate that an alginate scaffold may provide an appropriate substrate also for the survival and re-growth of blood vessels. Furthermore, growth factors affinity-bound to the alginate scaffold promote the survival, proliferation and differentiation of microvascular cells 66,67 , which results in extensive collateral branching of damaged vessels and thickening of vessels within the lesion site.
Another important issue in damaged spinal cord pathology is the development of central sensitization, which often contributes to hyperalgesia and allodynia typically associated with inflammatory pain 68 . In the present study therefore we addressed the response of sensory fibres expressing calcitonin gene-related peptide (CGRP) following treatment with alginate biomaterial. Our data show that significant increase in CGRP+ fibres occurred in the dorsal horns and lateral grey matter after ALG+ GF delivery. These most likely represent unmyelinated pelvic afferent fibres that convey thermal and nociceptive information 69 . Plastic re-organization of spinal neural circuitry and morphological changes in the spinal reflex pathway (primary afferent fibres and spinal interneurons) may be responsible for serious post-injury complications that could lead to uncontrolled excitatory activity of glutamate-driven sympathetic preganglionic neurons, and similarly to loss of inhibitory GABAergic/glycinergic interneurons that could have an impact on increased bilateral hind limb sensitivity to cold. Although the administration of GF-enriched alginate biomaterial promoted extending of CGRP positive fibres, we did not observe adverse sensory response to cold, such as observed after saline or pure alginate delivery. Responses of the hind limbs were relatively stable in the SCI+ ALG+ GF treated group during the whole survival period, with intensity similar to that in the sham controls. From our results we can speculate that alginate biomaterial with affinity-bound growth factors enhanced changes in CGRP fibres, but without behavioural adverse sensory response.
Central sensitization of spinal neurons or neuronal hyper-responsiveness and alterations in behavioural pain thresholds may be also in close correlation with microglial activation, as pointed out in some recent studies [70][71][72] . It is known that release of excitatory amino acids 73 , interleukin-1 74 , and prostaglandin E2 75 by microglia actively participate in the generation of central sensitization after SCI. On the basis of this hypothesis we can conclude that significant microglia response after saline delivery could induce an increase in central sensitization of spinal neurons and promote the kind of adverse sensory response to cold detected in the present study. However, ALG-GF treatment causing attenuation of microglia may lead to normalization of sensory behaviour.
Reactive astrocytes together with microglia and meningeal fibroblasts are known to participate in scar formation, representing a mechanical and chemical barrier for nerve tissue regeneration 76 . Significant differences in GFAP+ IR between the experimental groups were observed spatially, mainly in the central and caudal segments. However, treatment with ALG and ALG+ GFs did not attenuate the activation and proliferation of astrocytes after SCI, which may ultimately contribute to glia scarring at the central lesion site.
In the present experimental study we tried to define the efficacy of usage of alginate itself as well alginate enriched with GFs for spinal cord repair. EGF and bFGF were selected due to their stable and high binding properties, as well as long-term sustained GF release from alginates, confirmed in vitro 25 . Although these factors are important for their mitotic and partially differential properties for endogenous neural progenitors and their ability to accelerate neovascularisation, they may also contribute to astrogliosis and tissue scarring. In future experiments therefore, other neurotrophins such as BDNF, GDNF, NT-3 with neuroprotective action, or VEGF, PDGF with angiogenic properties should be considered for incorporation into one alginate device. In addition, simultaneous digestion of Chondroitin Sulfate Proteoglycans by chondroitinase ABC (ChABC) should be considered as well 8,9,43 . From the surgical point of view, the alginate scaffold described herein was injected into irregular spinal cord cavities, where it adjusted itself into the cavity shape and in the presence of calcium ions could undergo gelation in situ. This type of non-invasive technique for vehicle administration is potentially advantageous in particular when considering the fragility of the spinal cord site.