Processing

Please wait...

Settings

Settings

Goto Application

1. WO2020109631 - LAMININ IMMOBILIZATION, METHODS AND USES THEREOF

Note: Text based on automatic Optical Character Recognition processes. Please use the PDF version for legal matters

[ EN ]

D E S C R I P T I O N

LAMININ IMMOBILIZATION, METHODS AND USES THEREOF

Technical field

[0001] The present disclosure relates to an approach for site-specific laminin immobilization, methods and uses thereof. Namely the use of a composition for biofunctionalization of 2D and 3D substrates.

Background

[0002] In the framework of regenerative medicine and tissue engineering, much attention has been devoted towards the development of engineered matrices incorporating bioadhesive cues present in stem cell niches, to recapitulate the dynamic nature and biological complexity of these microenvironments, as well as to gain more insight into the function of specific extracellular matrix (ECM) components on stem cell behaviour. The ECM is an essential component of the stem cell niche, as it modulates important biological functions including proliferation, self-renewal and differentiation of stem cells. Among major ECM constituents, laminins play crucial and essential roles in many aspects of tissue physiology and function. These heterotrimeric glycoproteins comprise several bioactive domains involved in the modulation of different biological functions. The latter include the interaction with other ECM proteins (e.g. nidogen, netrin 4 and collagen IV) mediated by the laminin short arms (N-terminus), which contributes to the assembly and stability of basement membranes. These domains are also responsible for the laminin ability to polymerize, even in the absence of other basement membrane components, forming the molecular network that it is in contact with the cellular surface. In addition to its structural role, laminin comprises multiple bioactive domains that interact with growth factors and/or with cell surface receptors (e.g. integrins, dystroglycans, syndecans, and Lutheran), modulating different cell functions including cell adhesion, proliferation, migration and differentiation, as well as ECM deposition.

[0003] Laminin has been incorporated into both two-dimensional (2D) and three-dimensional (3D) cell-instructive microenvironments for applications in regenerative therapies or to get insights into the role of laminin on the modulation of cell behaviour. Strategies explored for laminin immobilization have relied either on its non-specific adsorption or entrapment or, alternatively, on its non-specific covalent immobilization to different substrates through the use of functional groups present in multiple sites of the laminin structure such as amines and thiols. One of the main caveats presented by these strategies is the inability to control orientation and conformation of laminin upon immobilization, which were proven to be crucial for the modulation of cellular behaviour. As such, the exposure of key laminin bioactive epitopes can be compromised. In an attempt to assure the control over the tethering of laminin, in recent years protein immobilization strategies have shifted toward site-specific conjugation, with special focus on biorthogonal chemical reactions (click chemistry), enzymatic ligation and affinity binding, using either unnatural amino acids or engineered site-selective amino acid sequences. These strategies are expected to provide a higher retention of bioactivity, by favouring the access to the active sites of immobilized proteins. To the best of our knowledge, to date, only one study reported the site-selective immobilization of laminin. To control the presentation of full-length ECM proteins without altering their bioactivity, Lee and co-workers explored click chemistry to anchor collagen, fibronectin and laminin onto polyacrylamide gels by their N-terminus. Nevertheless, despite guaranteeing the site-selective immobilization of the proteins, as result of the use of the N-terminus of laminin chains, this approach compromises one of laminin's hallmark features, which is its ability to polymerize. Affinity-binding has been increasingly explored for the site-specific and reversible conjugation of proteins and peptides, because of its versatility and ability to generate dynamic biomimetic systems. However, the successful implementation of this strategy is strongly dependent on the appropriate selection of the binding pairs. Binding systems using high affinity interactions, such as streptavidin and biotin, although often used, require the protein of interest to be either recombinantly or chemically modified. In contrast, the use of natural binding partners constitutes an attractive alternative, as strong non-covalent interactions can be achieved without the need for protein modification.

[0004] These facts are disclosed in order to illustrate the technical problem addressed by the present disclosure.

General Description

[0005] Laminin immobilization into diverse biological and synthetic matrices has been explored to replicate the microenvironment of stem cell niches and gain insight into the role of extracellular matrix (ECM) on stem cell behaviour. However, the site-specific immobilization of this heterotrimeric glycoprotein and, consequently, control over its orientation and bioactivity has been a challenge that has limited many of the explored strategies to date. To address this issue, an affinity-based approach that takes advantage of the native high affinity interaction between laminin and the human N-terminal agrin (hNtA) domain was established. This interaction was expected to

promote the site-selective immobilization of laminin to a substrate, while preserving the exposure of its key bioactive epitopes. Recombinant hNtA (rhNtA) domain was produced with high purity (>90%) and successfully conjugated at its N-terminal with a thiol-terminated poly(ethylene glycol) (PEG) without affecting its affinity to laminin. Self-assembled monolayers (SAMs) of mono-PEGylated rhNtA on gold (mPEG rhNtA-SAMs) were then prepared to evaluate the effectiveness of this strategy. The site-specific immobilization of laminin onto mPEG rhNtA-SAMs was shown to better preserve protein bioactivity in comparison to laminin immobilized on SAMs of thiol-PEG-succinimidyl glutaramide, used for the non-selective covalent immobilization of laminin, as evidenced by its enhanced ability to efficiently self-polymerize and mediate cell adhesion and spreading of human neural stem cells. These results highlight the potential of this novel strategy to be used as an alternative to the conventional immobilization approaches in a wide range of applications, including engineered coatings for neuroelectrodes and cell culture, as well as biofunctionalization of 3D matrices for tissue engineering and regenerative medicine.

[0006] In an embodiment, a strategy that explores the native high affinity interaction (KD = 5 nM) between laminin and the N-terminal agrin (NtA) domain is described. The agrin-binding site in laminin is localized in the central region of its coiled-coil domain and maps to a sequence of 20 conserved residues within the gΐ chain. Interestingly, this interaction requires a coiled-coil conformation of the agrin-binding site. In an embodiment, to assess the ability of this affinity-based approach to immobilize laminin with retention of bioactivity, recombinant human NtA (rhNtA) domain was produced and further conjugated at its N-terminus with a poly(ethylene glycol) (PEG), preferably with a thiol-terminated PEG, to enable the preparation of self-assembled monolayers (SAMs) of rhNtA on gold.

[0007] The specific interactions between mono-PEGylated rhNTA-immobilized laminin and cells are also disclosed. For this purpose, SAMs of mono-PEGylated rhNTA on gold were used as proof-of-concept platforms, since they can be easily produced and specifically tailored to provide a chemically well-defined molecular monolayer. The binding ligand was characterized and its ability to mediate laminin immobilization through a high affinity interaction was shown by solid-binding assay, surface plasmon resonance (SPR) and quartz crystal microbalance with dissipation monitoring (QCM-D). The bioactivity of mono-PEGylated rhNtA-immobilized laminin was subsequently investigated by evaluating its ability to self-polymerize and modulate the behaviour of human neural stem cells (hNSCs). hNSCs were used due to the well described role of laminin on the modulation of neural cell behavior.

[0008] The present disclosure is a versatile approach for the site-selective immobilization of laminin, while preserving its bioactivity, and it is useful for the development of cell instructive microenvironments for tissue engineering and regenerative medicine.

[0009] An aspect of the present disclosure relates to a composition for substrate functionalization comprising: a poly(ethylene glycol) - PEG - for binding to the substrate; an agrin domain comprising a sequence at least 95% identical to SEQ. ID NO: 1 bound to said PEG; a laminin able for binding to a bioactive agent; wherein the laminin is bound to the agrin domain. The composition for substrate functionalization comprises: an agrin domain comprising a sequence at least 95% identical to SEQ. ID. NO: 1 conjugated at its N-terminus to an end-group functionalized PEG to allow the binding of the agrin domain to a substrate while assuring the exposure of the high-affinity laminin binding site of agrin, located on the C-terminal; laminin capable of interacting with a bioactive agent, wherein laminin is bound to the agrin domain by high affinity interaction.

[0010] In an embodiment, the amino acid sequence of recombinant human N-terminal agrin (rhNtA) domain after purification and N-terminal tag cleavage - is the SEQ. ID. 1: GPTCPERALERREEEANVVLTGTVEEILNVDPVQHTYSCKVRVWRYLKGKDLVARESLLDGGNKVVISGFGDPLI CDNQVSTGDTRIFFVNPAPPYLWPAHKNELMLNSSLMRITLRNLEEVEFCVEDKP.

[0011] Methods for the alignment of sequences for comparison are well known in the art, such methods include GAP, BESTFIT, BLAST, FASTA and TFASTA. GAP uses the algorithm of Needleman and Wunsch ((1970) J Mol Biol 48: 443-453) to find the global (over the whole the sequence) alignment of two sequences that maximizes the number of matches and minimizes the number of gaps. The BLAST algorithm (Altschul et al. (1990) J Mol Biol 215: 403-10) calculates percent sequence identity and performs a statistical analysis of the similarity between the two sequences. The software for performing BLAST analysis is publicly available through the National Centre for Biotechnology Information (NCBI). Global percentages of similarity and identity may also be determined using one of the methods available in the MatGAT software package (Campanella et al., BMC Bioinformatics. 2003 Jul 10; 4:29. MatGAT: an application that generates similarity/identity matrices using protein or DNA sequences). Minor manual editing may be performed to optimise alignment between conserved motifs, as would be apparent to a person skilled in the art. The sequence identity values, which are indicated in the present subject matter as a percentage were determined over the entire amino acid sequence, using BLAST with the default parameters.

[0012] In an embodiment for better results, the agrin domain comprises a sequence at least 96% identical to SEQ. ID NO: 1, preferably at least 97% identical to SEQ. ID NO: 1, at least 98% identical to SEQ. ID NO: 1, at least 99% identical to SEQ. ID NO: 1.

[0013] In an embodiment for better results, the agrin domain comprises a sequence identical to SEQ. ID NO: 1.

[0014] In an embodiment for better results, the agrin domain is conjugated at its N-terminal with a thiol-terminated poly(ethylene glycol).

[0015] In an embodiment for better results, the bioactive agent is selected from a list consisting of: a biomolecule, an active ingredient, a growth factor, extracellular vesicle, a hormone, a cell, or mixtures thereof.

[0016] In an embodiment for better results, the biomolecule is another laminin, a different protein, a peptide, a drug, a polysaccharide, or mixtures thereof.

[0017] In an embodiment for better results; the cell is selected from: osteoblast, osteoclast, osteocyte, pericyte, endothelial cells, endothelial progenitor cells, hematopoietic progenitor cell, hematopoietic stem cell, neural progenitor cell, neural stem cell, neural cell, oligodendrocyte progenitor cell, oligodendrocytes, ependymal cells, mesenchymal stromal/stem cell, induced pluripotent stem cell, embryonic stem cell, perivascular stem cells, amniotic fluid stem cell, amniotic membrane stem cell, umbilical cord stem cell, genetically engineered cell, or mixtures thereof.

[0018] In an embodiment for better results, the polyethylene glycol) size varies between 2 and 40 kDa; preferably 3-20 kDa, more preferably 3.5-10 kDa.

[0019] Another aspect of the present disclosure is the use of the composition of the present disclosure in veterinary or human medicine. Namely, for use in the treatment of diseases that involve tissue regeneration or repair, or for use in the prevention of tissue degeneration.

[0020] In an embodiment for better results, the substrate is a naturally-derived or purely synthetic polymer, a synthetically-derived biopolymer, a metal, or mixtures thereof.

[0021] Another aspect of the present disclosure is a hydrogel comprising the composition described in the present subject matter.

[0022] In an embodiment for better results, the hydrogel is selected from the following list: poly(ethylene glycol), poly(vinyl alcohol), chitosan, alginate, collagen, dextran, fibrin, hyaluronic acid, silk fibroin, cellulose, synthetically-derived biopolymers, or mixtures thereof.

[0023] Another aspect of the present disclosure is a substrate coating comprising the composition described in the present subject matter.

[0024] Another aspect of the present disclosure is a coated article comprising the coating composition of the in the present subject matter. In particular wherein the article is an electrode, or a medical device, among others.

[0025] Another aspect of the present disclosure is a kit for screening therapeutic drugs, comprising the composition described in in the present subject matter.

Brief Description of the Drawings

[0026] The following figures provide preferred embodiments for illustrating the description and should not be seen as limiting the scope of invention.

[0027] Figure 1 - Solid-phase assay of Trx-His6-hNtA binding to immobilized msLn-111 (10 pg/mL). The results are shown as fitted curve of the mean ± standard deviation (SD) of three replicate samples from one experiment representative of three independent assays. Non-linear regression analysis revealed a KD value of Trx-HiS6-hNtA to msLn-111 of 6.49 ± 0.58 nM (r2 = 0.98).

[0028] Figure 2 - Recombinant human NtA (rhNtA) domain characterization. A) Coomassie Blue stained 15% SDS-PAGE gel and B) Western-blot analysis of 10 pg of purified rhNtA. MW, Molecular Weight marker. C) MALDI-TOF MS spectra of purified rhNtA.

[0029] Figure 3- rhNtA Bioactivity. Solid-phase assay of rhNtA binding to immobilized msLn-111 (10 pg/mL). The results are shown as fitted curve of the mean ± standard deviation (SD) of three replicate samples from one experiment representative of three independent assays. Non-linear regression analysis revealed a KD value of rhNtA to msLn-111 of 5.85 ± 0.43 nM (r2 = 0.99).

[0030] Figure 4- hNSC adhesion to rhNtA-immobilized msLn-111 (10 pg/mL) as a function of rhNtA concentration (0.1 - 40 pg/mL), evaluated performing a cell adhesion centrifugation assay. The number of adherent cells on wells coated with BSA (1% (w/v), negative control), rhNtA (10 pg/mL) and 20 pg/mL poly(ornithine)/ 10 pg/mL msLn-111 (positive control), is also shown. Data represent mean ± SD of 10 to 18 replicates from two independent experiments; one-way ANOVA followed by Dunnett's test (vs. poly(ornithine)/msLn-lll).

[0031] Figure 5 - Isotherm of rhNtA adsorption determined using a direct ELISA assay. The results are shown as fitted curves of the mean ± SD of twelve replicate samples analysed from two independent experiments (r2 = 0.94).

[0032] Figure 6 - Characterization of N-terminal PEGylated rhNtA. A) Cation-exchange chromatogram of the conjugation mixture. The chromatogram was obtained by monitoring the

eluent at 280 nm and using a HiTrap SP HP column (GE Healthcare); 1, di-PEGylated rhNtA; 2, mono-PEGylated rhNtA; 3, Native rhNtA. B) Coomassie Blue stained 4-12% Bis-Tris gel of native and PEGylated rhNtA after purification. M, MW marker; C, Conjugation mixture before cation-exchange chromatography; 1, di-PEGylated rhNtA band at around 36.5 kDa; 2, mono-PEGylated rhNtA band at around 28.8 kDa; 3, native rhNtA at around 14.9 kDa. C) MALDI-TOF MS spectra of native and PEGylated rhNtA. The presented molecular weight corresponds to the value obtained for the high intensity peak - Native rhNtA 14.8 kDa; mono-PEGylated rhNtA 18.4 kDa; di-PEGylated rhNtA 22.1 kDa.

[0033] Figure 7 - Position of conjugated PEG moiety on the purified mono-PEGylated rhNtA determined by peptide mapping. Positive mode MALDI-TOF/TOF identification of the fragments obtained upon digestion of native and mono-PEGylated rhNtA with trypsin. Two peaks correspondent to the N-terminus were identified on both native and mono-PEGylated rhNtA (arrows) - (A) FI - GPTCPER [1-7] m/z 816.3677 and (B) F2 - GPTCPERALER [1-11] m/z 1285.6278.

[0034] Figure 8 - Immobilization of rhLn-521 (ng/cm2) onto mPEG rhNtA-SAMs, as a function of the mono-PEGylated rhNtA concentration. SAMs were prepared using a fixed concentration of EG4 (100.0 mM) and decreasing concentrations of mono-PEGylated rhNtA (2.5, 5.0 and 10.0 mM). rhLn-521 immobilization was monitored in real time by QCM-D analysis and data modeled with application of the Voigt model. Results represent mean ± SD of three independent experiments; one-way ANOVA followed by Bonferroni's test.

[0035] Figure 9 - IRRAS spectra of mPEG rhNtA- and SGA- SAMs. EG4-SAMs and native rhNtA-adsorbed to Au surfaces were used as controls. A baseline correction and smoothing were applied to all the spectra. Two protein characteristic bands were identified - amide I (1690 ± 45 cm 1) and II (1540 ± 60 cm 1).

[0036] Figure 10 - Assessment of the ability of site-selective immobilized laminin to self-polymerize. A) Schematic representation of laminin self-polymerization (not to scale). Adapted from [57] N-terminal domains of a, b and g chains interact to form a polygonal network. The long arm is not involved in the network formation. B) mPEG rhNtA-SAM (left panel) or SGA-SAM (right panel) were incubated with msLn-111 (50 pg/mL) for 24 h, under neutral conditions (pH 7.0), and samples processed for immunofluorescence staining of laminin. Scale Bar = 50 pm. C) Number of aggregates; D) Average size (pm2); E) Perimeter (pm); and F) Maximum Feret's diameter (pm) determined by image processing and quantitative analysis using ImageJ/Fiji and MATLAB® software, respectively. Data represent mean ± SD of 13-16 replicate samples from three independent assays; non-parametric Mann-Whitney U-test.

[0037] Figure 11 - Immobilization of rhLn-521 onto SAMs and its effect on hNSC adhesion and spreading. A) Schematic representation of the prepared surfaces (not to scale) - Au surface (control surface), EG4-SAM (non-fouling surface), mPEG rhNtA-SAM (for site-selective immobilization of laminin) and SGA-SAM (for non-specific covalent immobilization of laminin). B) QCM-D analysis of rhLn-521 immobilization (ng/cm2) on SAMs. Data represent mean ± SD of three independent experiments; one-way ANOVA followed by Bonferroni's test. C) Representative 2D projections of CLSM 3D stack images of adhered hNSCs, after 24h of culture. Scale Bar = 50 pm; D) Total number of adherent cells; and E) Average cell spreading area (pm2) determined by quantitative analysis of images acquired with an IN Cell Analyzer 2000 imaging system. Data represent mean ± standard error of the mean (SEM) of four replicate samples from an experiment representative of three independent assays; one-way ANOVA followed by Bonferroni's test.

[0038] Figure 12 - hNSC adhesion to SAMs determined measuring the average fluorescence intensity of Hoechst 33342 in samples processed for F-actin/DNA staining. Data represent mean ± SEM of four replicate samples from one experiment representative of three independent assays; one-way ANOVA followed by Bonferroni's test.

[0039] Figure 13 - hNSCs average occupied area (pm2) determined by quantitative analysis of images acquired in IN Cell Analyzer 2000 imaging system. Data represent mean ± SEM of four replicate samples of an experiment representative of three independent assays; one-way ANOVA followed by Bonferroni's test.

[0040] Figure 14- Representative 2D projections of CLSM 3D stack images of hNSCs adhered to TCPS and EG4- and SGA-SAMs. Images were acquired after 24h of culture. Scale Bar = 50 pm.

[0041] Figure 15. Schematic representation of the preparation of cell-laden affinity-bound laminin PEG-4MAL hydrogels (not to scale). PEG-4MAL macromer is firstly functionalized with a thiol-containing mono-PEGylated rhNtA (NtA) domain, which will then mediate the binding of laminin with high affinity. Functionalized PEG-4MAL is then reacted with a mixture of protease degradable (MMP2-sensitive peptide) and non-degradable (PEG-dithiol) cross-linkers to allow the formation, under physiological conditions, of a hydrogel network in the presence of cells.

[0042] Figure 16 - Optimization of MMP2-sensitive peptide (SGESPAY^YTA): PEG-dithiol molar ratio (%) in affinity-bound laminin PEG-4MAL hydrogels (50 pM NtA; 20 pg/mL rhLn-521 (Biolamina)). A) Representative 2D projections of CLSM three-dimensional (3D) stack images of cell- laden affinity-bound laminin PEG-4MAL hydrogels at day 14 of cell culture covering a thickness of 200 miti, showing the distribution of live (green) and/or dead (red) cells. Scale Bar = 200 miti. B) Calcein+ cells average occupied area (miti2) after 14 days of culture, as determined by quantitative image analysis of 2D projections of CLSM 3D stack images of cell-laden hydrogels incubated with calcein AM (viable cells) and PI (dead cells). Data represent mean ± standard error of the mean (SEM) of 6-9 images from two independent experiments; one-way ANOVA followed by Dunnett’ s test (vs. 100:0).

[0043] Figure 17 - Effect of NtA concentration on laminin incorporation. A) Density of NtA tethered to PEG-4MAL hydrogel as a function of input NtA concentration. Linear regression: y = 0.9954 x + 0.1115; R2 = 0.9999. Data represent mean ± standard deviation (SD) of three hydrogels prepared independently. B) Laminin incorporation by physical entrapment (Entrap) or by immobilization using an affinity binding ligand (Immob) into PEG hydrogels, determined by ELISA, after 1 and 7 days of incubation in PBS. Data represent mean ± SD of three hydrogels prepared independently. No significant differences were detected (two-way ANOVA).

[0044] Figure 18 - Mechanical and structural properties of PEG-4MAL hydrogels modified with affinity-bound laminin. A) Complex modulus (G*) of hydrogels determined by rheological analysis; and B) Estimation of hydrogels mesh size, based on the measured storage modulus (G'). Data represent mean ± SEM; n = 6; one-way ANOVA followed by Bonferroni's test.

[0045] Figure 19 - Ability of affinity-bound laminin to support hNSC viability and proliferation. A) Representative 2D projections of CLSM 3D stack images of cell-laden hydrogels covering a thickness of 300 pm, showing the distribution of live (in green) and dead (in red) cells at day 7 of cell culture. Scale bar = 300 pm. B) Quantitative analysis of live cells at day 7, as determined by flow cytometry. Data represent mean ± SD of three independent experiments. No significant differences were detected (one-way ANOVA followed by Bonferroni's test). C) Quantitative analysis of total cell number at day 7, as determined by CyQuant® Cell Proliferation kit. The dotted line represents the initial cell density/hydrogel (4 x 104 cells/hydrogel). Data represent mean ± SEM of three independent experiments performed in triplicate; one-way ANOVA followed by Bonferroni's test.

[0046] Figure 20- Affinity-bound laminin PEG-4MAL hydrogels and its effect on hNSC neurite outgrowth. A) hNSCs cultured within unmodified (Unm), laminin physically entrapped (Entrap) and NtA affinity-bound laminin (Immob) PEG-4MAL hydrogels after 14 days of culture, under differentiation conditions. Images show 2D projections of CLSM 3D stack images of cells processed for immunofluorescence labeling of bIII-tubulin/DNA covering a thickness of 200 pm (top images)

and 100 pm (bottom images). Lower panels show magnified views of selected regions highlighted by dashed squares in the upper panel. Arrows depict neurites expressing bIII-tubulin protruding from neurospheres. Scale Bar = 200 pm (top images); 100 pm (bottom images). B) Representative 2D projections of CLSM 3D stack images depicting hNSC outgrowth at day 14 of cell culture, covering a thickness of 80 pm. Scale Bar = 80 pm. C) Total Neurite Length; and D) Total Number of Neurites determined by quantitative analysis of projected CLSM images . Data represent mean ± SD of 12-15 images analyzed per condition; one-way ANOVA followed by Bonferroni's test.

[0047] Figure 21. Phenotypic characterization of hNSC within PEG-4MAL hydrogels. hNSC cultured for 14 days within unmodified (Unm), physically entrapped laminin (Entrap) or NtA affinity-bound laminin (Immob) PEG-4MAL hydrogels. Images show 2D projections of CLSM 3D stack images of cells processed for immunofluorescence labelling of A) Nestin; B) Tau; and C) MAP2. Scale Bar = 100 pm.

Detailed Description

[0048] The present disclosure is also further described, in particular, using embodiments of the disclosure. Therefore, the disclosure is not limited to the descriptions and illustrations provided. These are used so that the disclosure is sufficiently detailed and comprehensive. Moreover, the intention of the drawings is for illustrative purposes and not for the purpose of limitation.

[0049] In this embodiment, the assays and resulting data of the present disclosure are described. In an embodiment, recombinant human N-terminal agrin (rhNtA) domain expression and purification are described. The NtA domain, comprising residues Thr 30 - Pro 157 of the human agrin protein (Uniprot reference 000468), was expressed in Escherichia coli (E. coli) strain BL21(DE3) using the pCoofy2 expression vector (gift from Sabine Suppmann - Addgene plasmid # 43981). pCoofy2 is a derivative of pETM22 vector that contains a thioredoxin-poly-His6 (6x Histidine residues) (Trx-His6) N-terminal tag followed by an HRV 3C recognition site located downstream of the N-terminal tag.

[0050] In an embodiment, the details on the cloning of human NtA gene and expression vector are disclosed. The human N-terminal agrin (hNtA) domain gene was synthesized by GenScript (Piscataway, NJ, USA) - Clone ID: B25738, from the protein sequence with the Uniprot reference 000468 (Thr 30 - Pro 157). A gene of 395bp length was produced in pUC57 cloning vector and the conditions for expression in Escherichia coli (E. coli) optimized. A high-throughput expression screening was conducted to select, among the several pCoofy expression vectors available the most suitable for the expression of the soluble protein. pCoofy2 expression vector - 6080 bp (gift from Sabine Suppmann - Addgene plasmid # 43981) was the one selected. Both hNtA gene and pCoofy2 expression vector were polymerase chain reaction (PCR) linearized using specific-designed primers and LP1 forward and LP2 reverse primers, respectively (Table 1). Following PCR amplification, the hNtA gene was sub-cloned into the pCoofy2 expression vector using a sequence and ligation independent cloning (SLIC) reaction. The efficiency of DNA insert incorporation into the expression vector was confirmed by colony PCR and DNA sequencing (Macrogen).

[0051] Table 1. Oligonucleotide sequences used for PCR linearization of hNtA gene and pCoofy2 expression vector.

Primer Sequence 5' - 3'

C - LP1 forward vector primer a 5' AAGTTCTGTTCCAGGGGCCC 3'

ccdB - LP2 reverse vector

5' CCCCAG AACAT CAG GTT AATGG CG 3'

primer a

hNtA forward primer (3C) b 5' AAGTTCTGTTCCAGGGGCCCGGCACCTGTCCGGAACGCGC 3

5' CCCCAG AACAT CAGGTT AAT G G CGTT ACGGTTT AT CTT CAACGCAAAATT C 3' hNtA reverse primer (ccdB) b

a Primers for vector amplification were obtained from: J. Scholz, H. Besir, C. Strasser, S. Suppmann, A new method to customize protein expression vectors for fast, efficient and background free parallel cloning, BMC Biotechnol 13 (2013) 12.

bPrimers used for hNtA gene amplification were specifically designed to have gene specific sequences plus 15 to 25 bp extensions complementary to LP1 and LP2 vector primers.

[0052] In an embodiment, the fusion protein (Trx-His6-hNtA) was expressed by induction of log-phase cultures with 0.2 mM of isopropyl b-D-l-thiogalactopyranoside (IPTG; Sigma-Aldrich) at 20°C for 5h and purified by immobilized metal affinity chromatography using a HisTrap column (GE Healthcare). A concentration gradient of 0 to 500 mM of imidazole was applied to elute the fusion protein. The N-terminal Trx-His6 tag was then cleaved during dialysis using HRV 3C protease (kindly provided by Biochemical and Biophysical Technologies (B2Tech) Platform of i3S) in a protease:target protein ratio (unit/pg) of 1:10. Untagged recombinant human NtA (rhNtA) contains additional Gly and Pro residues at its N-terminus (Seq. ID No 1). rhNtA was dialyzed against 10 mM phosphate buffered saline (PBS) pH 7.4, concentrated in a 10 kDa molecular weight cut-off (MWCO) ultrafiltration membrane (GE Healthcare) and stored at -80°C until further use. Final domain concentration was estimated by measuring the absorbance at 280 nm. The purity of the rhNtA domain was assessed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) analysis and Western blotting using the anti-agrin antibody - rabbit polyclonal agrin antibody

(1:5000; Abeam, ab85174). The rhNtA molecular weight was determined by mass spectrometry, as described in "Protein and PEGylated conjugate characterization by mass spectrometry".

[0053] In an embodiment, a solid-phase binding assay was described. Binding of rhNtA to immobilized laminin was assessed performing an enzyme-linked binding assay (ELISA). 96-well microtiter ELISA plates (BD Falcon) were coated with poly(D-lysine) (PDL; 20 pg/mL; Sigma-Aldrich) and then incubated with laminin-111 from mouse Engelbreth-Holm-Swarm sarcoma (msLn-lll;10 pg/mL; Sigma-Aldrich) in 50 mM sodium bicarbonate buffer pH 9.6 (overnight, 4°C). After blocking with 2.5% (w/v) bovine serum albumin (BSA; Biowest; 1.5h, room temperature (RT)), wells were incubated with 2-fold serial dilutions of rhNtA (0.1 - 100 nM; 2h, 37°C). Plates were then washed with PBS/0.05% (v/v) Tween® 20 and incubated with rabbit polyclonal agrin antibody (1:5000; 2h, RT). A goat anti-rabbit IgG (H+L) coupled to horseradish peroxidase (HRP) (1:2000; Life Technologies, A16023) was used as secondary antibody, and colour was developed using 3, 3', 5, 5' tetramethyl benzidine (TMB) as substrate (Biolegend; 30 min, RT). The reaction was stopped with 2 M H SO and the absorbance measured at 450 nm (BioTek® Synergy™ Mx). The dissociation constant (KD) value of rhNtA to msLn-111 was estimated by non-linear regression analysis using GraphPad Prism 6 software (San Diego).

[0054] In an embodiment, real-time interactions between rhNtA and immobilized laminin were detected by surface plasmon resonance (SPR) at 25°C using a BIACORE X100 (GE Healthcare). Laminin was immobilized on a CM5 sensor chip (GE Healthcare) using amine-coupling chemistry. Briefly, the flow cells surface was activated with a 1:1 mixture of 0.1 M N-hydroxysuccinimide (NHS; Sigma-Aldrich) and 0.4 M 3-(N,N-dimethylamino) propyl-N-ethylcarbodiimide (EDC; Sigma-Aldrich) at a flow rate of 10 pL/min. msLn-111 or recombinant human laminin-521 (rhLn-521; Biolamina) (50 pg/mL in 10 mM sodium acetate pH 3.5) were immobilized at a density of 4000 resonance units. The reactive groups in excess were then deactivated with 1 M ethanolamine-HCI pH 8.5 (Sigma-Aldrich). For binding measurements, 5-fold serial dilutions of rhNtA (500 - 0.8 nM) were prepared in 10 mM 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid (HEPES; pH 7.4, containing 150 mM sodium chloride (NaCI; VWR), 3 mM Ethylenediamine tetra-acetic acid (Sigma-Aldrich) and 0.005% (v/v) surfactant P20 (GE Healthcare)) and flowed at 30 pL/min onto the flow cells. Proteins were allowed to associate and dissociate for 120 and 1600 s, respectively. Experimental results were fitted with Langmuir 1:1 binding kinetics within Biacore X100 Evaluation software version 2.0.1 (GE Healthcare).

[0055] In an embodiment, the culture of human neural stem cells (hNSC) is described. hNSC derived from the National Institutes of Health (NIH) approved H9 (WA09) human embryonic stem

cell line were purchased from Life Technologies (N7800-200). Cells were expanded according to the manufacturer's protocol in poly(ornithine)/msLn-lll-coated tissue culture-treated plates in serum-free StemPro® NSC SFM growth medium (Life Technologies) supplemented with basic fibroblast growth factor (bFGF, 20 ng/mL; Life Technologies) and epidermal growth factor (EGF, 20 ng/mL; Life Technologies).

[0053] In an embodiment, to evaluate the ability of rhNtA-immobilized msLn-111 to promote the adhesion of hNSCs, a centrifugation cell adhesion assay was performed. Here, controllable and reproducible detachment forces were applied to adherent cells, providing relative measurements of adhesion strength. 96-well tissue culture plates (Corning) were coated with increased concentrations of rhNtA (0 - 40 pg/mL) overnight at 4°C and blocked with 1% (w/v) BSA for lh at RT to prevent non-specific adhesion to the substrate. msLn-111 was then added at a fixed concentration (10 pg/mL), and the plate incubated for 2h at 37°C. hNSCs were fluorescently-labelled with 2 pM of Calcein AM (Molecular Probes, 10 min, 37QC), suspended in StemPro® NSC SFM medium, and transferred to rhNtA-immobilized msLn-lll-adsorbed surfaces at 3x10s viable cells/cm2. Cells were allowed to adhere for 5 h and, before subjecting the cells to centrifugation, initial fluorescence intensity was measured using a microwell plate reader. After overfilling the wells with 10 mM PBS pH 7.4, until a positive meniscus was observed, the plates were sealed with transparent tape, inverted, and spun at 50 RCF for 5 min. Following cell detachment, the medium was changed, and the wells filled with fresh medium for measurement of final fluorescence. For each well, the adherent cell fraction was determined as the ratio of post-spin (final) to pre-spin (initial) fluorescence readings (lQC = 485 nm; em = 535 nm). Wells incubated with 1% (w/v) BSA or 20 pg/mL poly(ornithine) / 10 pg/mL msLn-111 were used as negative and positive control, respectively. To assess the effect of the domain per se on cell adhesion, wells incubated with 10 pg/mL rhNtA were used.

[0054] I n an embodiment, the rhNtA adsorption isotherm was determined coating 96-well microtiter ELISA plates (BD Falcon) with 2-fold serial dilution of rhNtA (0 - 40 pg/mL; overnight, 4°C) and blocked with 1% (w/v) BSA to prevent non-specific adhesion to substrate. The wells were then incubated with the primary antibody - rabbit polyclonal agrin antibody (1:5000; 2h, RT). Goat anti-rabbit IgG (H+L) coupled to HRP secondary antibody (1:2000) was incubated for 1,5 h at RT. The color was developed using 3, 3', 5, 5' tetramethyl benzidine (TMB) substrate (Biolegend), incubated for 30 min at RT. The reaction was stopped using 2 M H2SO4 and the absorbance measured at 450 nm using a microwell plate reader.

[0055] In an embodiment, N-terminal PEGylation of rhNtA was performed. A 5 mg/mL rhNtA solution in 0.1 M sodium phosphate (Nah PC ) pH 6.5 was added to thiol-PEG-succinimidyl glutaramide (HS-PEG-SGA; 3.5 kDa; purity > 90%, Jenkem USA) at a protei PEG molar ratio of 1:2. The reaction was allowed to proceed for 4h at 4°C and then quenched with 2 M hydroxylamine pH 7.4. The solution was diluted to a final protein concentration of 0.5 mg/mL with 1 mM HCI and the pH adjusted to 3.5 with 1 M HCI. The reaction mixture was then diluted (IOOc) with 20 mM sodium acetate pH 4.0 and loaded onto a HiTrap SP HP cation exchange column (GE Healthcare). A gradient of 0 to 1 M NaCI in 20 mM sodium acetate pH 4.0 was used to elute the different PEGylated fractions. The peak corresponding to the mono-N-terminal conjugate was dialyzed against 10 mM PBS pH 7.4, concentrated in a 10 kDa MWCO ultrafiltration membrane (GE Healthcare) and stored at -80°C until further use. Protein concentration was estimated by measuring the absorbance at 280 nm and the conjugate was characterized by SDS-PAGE and MALDI-TOF mass spectrometry (MALDI-TOF MS).

[0056] In an embodiment, rhNtA and the PEGylated conjugate were characterized by mass spectrometry. The native or mono-PEGylated rhNtA, as well as unmodified HS-PEG-SGA, were analysed by MALDI-TOF MS. Briefly, samples were diluted 1:4 (v/v) in MALDI matrix (sinapinic acid 10 mg/mL, 50% acetonitrile, 0.1% trifluoroacetic acid, all purchased from Sigma-Aldrich), spotted onto a target plate and analysed in linear positive mode. Mass spectra were internally calibrated with thioredoxin (m/z 11674) and apomyoglobin (m/z 16952).

[0055] In an embodiment, peptide mapping of native and mono-PEGylated rhNtA was performed to characterize the site of PEG conjugation to rhNtA. Protein bands correspondent to native or mono-PEGylated rhNtA were cut from the gel, washed with ultrapure water and de-stained with 50% (v/v) acetonitrile (ACN) in 50 mM ammonium bicarbonate. The samples were then reduced with 50 mM dithiothreitol (20 min, 56 °C) and alkylated with 55 mM iodoacetamide, (20 min, RT in the dark). Next, protein in gel enzymatic digestion was performed by the addition of 20 ng of trypsin for 3 h at 37 °C in the presence of 0.01% surfactant (ProteaseMAX™, Promega). Resulting peptides were extracted from gel plugs with 2.5% trifluoroacetic acid (TFA; 15 min at 1400 rpm), dried under vacuum and resuspended in 0.1% TFA. Protein identification was performed by a MALDI-TOF/TOF mass spectrometer (4800 Plus, SCI EX). Protein digests were purified by reversed-phase C18 chromatography (ZipTips®, Millipore) following the manufacturer's instructions and eluted in the MALDI sample plate using the MALDI matrix alpha-Cyano-4-hydroxycinnamic acid (CHCA) elution solution at 8 mg/mL in 50% ACN, 0.1% TFA, 6 mM ammonium phosphate. Peptide mass spectra were acquired in positive MS reflector mode in the mass range of m/z 700-5000 and internally calibrated with trypsin autolysis peaks. Individual peptides from both native and mono-PEGylated rhNtA were identified by peptide mass fingerprint (PMF) approach using MS-Bridge software version 5.16.1 (ProteinProspector, University of California, San Francisco, CA) and the protein sequence data (SEQ. ID No. 1). The protein search settings were cysteine carbamidomethylation (constant modification), methionine oxidation (variable modification), up to two missed trypsin cleavages, minimum and maximum digest fragment mass of 500 and 4000 Da, respectively, and a minimum digest fragment length of five amino acids. To estimate the location of the PEG moiety in the conjugate, experimental ion peaks obtained in the peptide map of purified mono-PEGylated rhNtA conjugate were compared to the peptide content of native rhNtA. In addition, an indirect estimation of PEG moiety position in the rhNtA sequence was performed by comparing the mass difference between the conjugated and unmodified PEG molecule.

[0057] In an embodiment, self-assembled monolayers (SAMs) were prepared. Mixed SAMs of mono-PEGylated rhNtA (mPEG rhNtA-SAM) or FIS-PEG-SGA (SGA-SAM) were prepared on gold-coated substrates (0.5 c 0.5 cm2), obtained from Instituto de Engenharia de Sistemas e Computadores - Microsistemas e Nanotecnologias, Portugal (INESC-MN). Prior to use, gold surfaces were cleaned. Gold substrates were then incubated in ethanolic solutions (99.9%; Merck-Millipore) of either mono-PEGylated rhNtA (2.5, 5.0 or 10.0 mM), for laminin affinity binding; or FIS-PEG-SGA (5.0 mM), for non-selective covalent immobilization of laminin. After 4h at RT and under inert conditions (glove chamber saturated with dry nitrogen), SAMs were rinsed with absolute ethanol to remove any physically adsorbed molecules. For proper monolayer packing mPEG rhNtA- and SGA- SAMs were subsequently incubated with a 100.0 mM absolute ethanolic solution of (11-mercaptoundecyl) tetraethylene glycol (EG4, SensoPath Technologies) for 16h at RT and under the inert conditions mentioned above. Afterwards, SAMs were thoroughly rinsed with absolute ethanol (99.9%; Merck-Millipore). EG4-SAMs, used as control surfaces, were prepared by immersing cleaned gold substrates in a 100.0 mM ethanolic solution of EG4 for 16h at RT and under inert conditions.

[0058] In an embodiment, laminin binding to SAMs was followed by Quartz crystal microbala nce with dissipation monitoring (QCM-D). Cleaned Gold-coated QCM-D sensors (Biolin Scientific) were further modified following the procedure reported for the preparation of SAMs. A QCM-D system (Q-Sense E4 instrument, Biolin Scientific) was used to monitor in real time the frequency (Af) and dissipation (AD) shifts related to laminin adsorption. Sensors were pre-incubated with 10 mM H EPES pH 7.4 for 15 min at a constant flow rate of 0.1 pL/min followed by an additional 15 min incubation, under static conditions, to establish the baseline. Then, 300 pL/sensor of rhLn-521 (20 pg/mL) in 10 mM H EPES pH 7.4 was injected at a constant flow rate of 0.1 pL/min, followed by 1 h incubation in static conditions. After, the system was flushed with 10 mM HEPES pH 7.4 during 30 min at a constant flow rate of 0.1 pL/min. All the experiments were conducted at 25 °C. The total amount of immobilized laminin was calculated using the Voigt model in the QTools® V3 software, which takes into account the viscoelastic contributions of the hydrated layer. The fluid density and viscosity of rhLn-521 solution were established at 1000 kg/m3 and 0.001 kg/m-s, respectively and data from the 5th harmonic was used in the analysis. Results are presented as mass per surface area (ng/cm2).

[0059] In an embodiment, infrared reflection absorption spectroscopy (IRRAS) spectra were obtained using a Perkin Elmer FTIR spectrophotometer, model 2000, coupled with a VeeMax II Accessory (PIKE) and a liquid-nitrogen-cooled MTC detector. To avoid water vapor adsorption, instrument was purged with dry nitrogen for 5 min before and during each sample analysis. Au substrates were here used as a background. Incident light was p-polarized and spectra were collected using the 80° grazing angle reflection mode. The analysis of each sample was performed through the acquisition of 100 scans with 4 cm 1 resolution.

[0060] In an embodiment, laminin self-polymerization was assessed. mPEG rhNtA- SAMs were equilibrated in 10 mM HEPES pH 7.4, to favour the affinity interaction with laminin, while SGA-SAMs were equilibrated in 100 mM sodium bicarbonate pH 8.5, to favour the non-selective covalent interaction with laminin amine groups, and then incubated with msLn-111 (20 pg/mL; 2h, RT). The substrates were then thoroughly washed, to remove any unbound laminin. To assess the ability of laminin to self-polymerize, an msLn-111 solution (50 pg/mL) in 20 mM Tris-HCI pH 7.0, containing 1 mM CaCh, was used. The laminin concentration used was below the critical protein concentration necessary to trigger laminin polymerization in neutral pH solution. Mixed SAMs with affinity- (mPEG rhNtA-SAMs) or covalent-bound (SGA-SAMs) msLn-111 were incubated with the laminin solution for 24h at 37°C. Finally, samples were processed for laminin immunostaining. Briefly, samples were fixed with 3.7% (w/v) paraformaldehyde (PFA) for 20 min at RT and washed 3x with 10 mM PBS pH 7.4. Samples were then incubated with blocking buffer (5% (w/v) BSA in PBS for 30 min at RT), followed by incubation with the primary antibody - rabbit polyclonal laminin antibody (1:30; Sigma-Aldrich, L9393), overnight at 4°C. Samples were washed 3x with 10 mM PBS pH 7.4 and incubated with Alexa Fluor® 488 conjugated anti-rabbit secondary antibody (1:300; Life Technologies, A11034), for lh at RT. Samples were observed under confocal laser scanning microscopy (CLSM, Leica TCS SP5II) using a Plan-Apochromat 63x/1.4NA oil objective and acquired z-sections of 15 pm were processed using the ImageJ/Fiji software. Briefly, the rolling ball algorithm was first applied to maximum intensity projection 3D images, to fix uneven background and followed by the application

of a Gaussian blur with s = 1 to reduce noise. Aggregates were segmented through an Otsu thresholding method and a morphological filter was then used to unify structures distanced by less than 1 pm. Lastly, all the structures with an area below 10 pm2 were removed/rejected. For each resulting image, measurements were obtained for number of aggregates, aggregates average size (pm2), perimeter (pm) and maximum Feret's diameter (pm). To evaluate differences on the data resultant from both conditions, statistical analysis of the different variables was conducted on MATLAB® software (R2018a) using the non-parametric Mann-Whitney U-test. Moreover, differences resultant from the combination of the different variables measured were determined by application of the Fisher's combined probability test and an overall p-value determined. This test is applied under the assumption that variables should be independent from each other. As the count is not independent from the other variables, for the Fisher's combined probability test, only average size, perimeter and maximum Feret's diameter of aggregates were considered.

[0061] In an embodiment, cell adhesion and morphology were analysed to determine the impact of the immobilization approach on the cell adhesion-promoting activity of laminin. mPEG rhNtA- or SGA-SAMs were equilibrated in 10 mM HEPES pH 7.4 and then incubated for 2h at RT with rhLn-521 (20 pg/mL). After being washed with 10 mM HEPES pH 7.4 to remove unbound laminin, hNSCs (50,000 viable cells/cm2) were added to the substrates and incubated at 37°C with 5% CO2 for 24h. hNSC adhesion and morphology were assessed after 24 h of culture in samples processed for F-actin/DNA staining. Cells were fixed with 2% (w/v) PFA for 20 min at RT, permeabilized with 0.5% (v/v) Triton X-100 (Sigma-Aldrich) for 10 min at RT and stained with Alexa Fluor® 647 phalloidin (1:50; Biolegend, 424205) for 20 min at RT. The nuclei were counterstained with 0.1 pg/mL of Hoechst 33342 (Life Technologies, H1399) for 20 min at RT. Samples were then imaged with the IN Cell Analyzer 2000 imaging system (GE Healthcare) using a Nikon 20x/0.45 NA Plan Fluor objective. A z-section of 45 pm was acquired per field of view (FOV) and then projected into a 2D image. Images spanning a total area of 4.6 mm2 were analyzed using CellProfiler image analysis software. Briefly, nuclei were segmented through an Otsu thresholding method from the 4',6-diamidino-2-phenylindole (DAPI) channel while cells cytoplasm was segmented from the actin channel through a propagation from the previously identified nucleus, using a minimum cross entropy threshold. Measurements were obtained for total number of adhered cells, average cell spreading area (pm2) and average occupied area (pm2) per FOV. Representative images for each condition were also obtained by CLSM using a Plan-Apochromat 63x/1.4NA oil objective. A z-section of 12 pm was acquired and subsequently projected into a 2D image.

[0062] In an embodiment, cell adhesion was also estimated from Hoechst 33342 average fluorescence intensity of samples processed for F-actin/DNA staining. Briefly, samples were transferred to the wells of black 24-well plates (ibidi) and the fluorescence of each surface measured (lQC = 350 nm; em = 461 nm) using a microwell plate reader. The measurement was performed using the area scan mode with the optics position set to the bottom. The fluorescence intensity (FI) of unseeded SAMs, processed in parallel, was subtracted and the average FI determined.

[0063] Individual experiments with biological replicates were performed at least in duplicate and statistical analysis was performed using GraphPad Prism 6 software (San Diego). Statistically significant differences between two conditions were detected using a non-parametric Mann Whitney U-test. Comparisons between three or more groups were performed with one-way ANOVA analysis, followed by the Bonferroni correction for pairwise comparisons or the Dunnett's two-tailed test for comparisons with the control condition. For all analyses, differences were considered significant at p < 0.05.

[0064] In an embodiment, recombinant human N-terminal agrin (rhNtA) domain was produced. In contrast with most of reports published to date in which the NtA used was derived from chicken agrin, we produced the human variant of this domain because of its higher clinical relevance. This is of particular interest when envisaging the use of this domain in platforms for human-disease modelling, cell culture/delivery or engineered coatings for neuroelectrodes.

[0065] The produced fusion protein Trx-Flis6-hNtA, revealed high affinity binding to msLn-111 with an equilibrium dissociation constant (KD) of 6.49 ± 0.58 nM, as determined by solid-phase binding assay (Fig. 1), similar to that reported for chicken NtA - msLn-111 interaction (KD = 5 nM). Cleavage of the N-terminal tag was subsequently conducted to enable conjugation of a thiol-terminated PEG moiety to the domain N-terminal, for further immobilization of rhNtA to the selected substrate. This immobilization approach was chosen to assure the exposure of the high-affinity laminin binding site of rhNtA, located on the C-terminal. Untagged rhNtA domain was produced with a high purity (> 90%) and with the expected MW 14.5 kDa, as evidenced by Coomassie blue staining (Fig. 2A) and Western blot analysis (Fig. 2B). Molecular weight was further confirmed by MALDI-TOF MS (MW 14.8 kDa) (Fig. 2C).

[0066] rhNtA was shown to mediate a high affinity interaction with msLn-111 (KD = 5.85 ± 0.43 nM), as evidenced by solid-phase binding assay (Fig. 3). To monitor in real time the interaction between rhNtA and immobilized msLn-111 or rhLn-521, SPR was conducted). Native rhNtA was found to interact with high affinity with both msLn-111 and rhLn-521 (KD = 0.98 ± 0.02 and 0.54 ± 0.14 nM, respectively; Table 2), showing that binding affinity towards laminin was retained independently of the laminin isoform. Results from both techniques indicate that the laminin binding site of the produced rhNtA domain preserved its bioactivity, namely the ability to interact with high affinity with laminin.

[0067] Table 2. Kinetic parameters (ka, association rate constant; kd, dissociation rate constant) and affinity constant (KD, dissociation constant) of rhNtA to immobilized msLn-111 or rhLn-521, estimated by SPR. Data represent mean ± SD of three independent experiments.


msLn-111 5.25 ± 1.83 5.12 ± 1.71 0.98 ± 0.02 rhLn-521 4.86 ± 1.55 2.52 ± 0.43 0.54 ± 0.14

[0068] In an embodiment, rhNtA-immobilized laminin was shown to mediate cell adhesion of human neural stem cells. Laminin mediates integrin- and non-integrin-dependent adhesion of neural stem cell (NSCs), being commonly used as a surface coating for NSC culture. As the site-specific immobilization of laminin should preserve exposure of its multiple bioactive domains interacting with cell adhesion receptors and, therefore, its cell adhesion-promoting activity, we assessed the ability of rhNtA-immobilized msLn-111 to mediate cell adhesion of hNSCs. These cells were seeded on cell culture wells coated with rhNtA-immobilized msLn-111 and cell adhesion quantified, as a function of rhNtA concentration, using a centrifugation adhesion assay (Fig. 4). The application of a constant centrifugal force revealed an increase in cell adhesion with the increase of the rhNtA concentration from 0.1 to 10 pg/mL (Fig. 4), at which the levels of cell adhesion observed on rhNtA-immobilized msLn-111 reached values similar to those on the msLn-111 positive control (Fig. 4). This suggests that under these conditions, the amount of affinity-immobilized msLn-111 attained a plateau, which is in agreement with the isotherm of rhNtA adsorption (Fig. 5) that shows a saturation plateau at domain concentrations > 2.5 pg /mL. The cell adhesion levels observed in the absence of rhNtA (0 pg/mL rhNtA) are most likely due to the competitive displacement of adsorbed BSA by laminin, an effect also known as the Vroman effect. Cell adhesion to wells incubated only with rhNtA (10 pg/mL) was comparable to that observed on the negative control (BSA; Fig. 4), due to the absence in the rhNtA of bioactive cell adhesion domains. These results demonstrate that, when immobilized through rhNtA, laminin retains its ability to mediate cell adhesion, suggesting the correct exposure of laminin's main bioactive sites interacting with cell surface receptors.

[0069] In an embodiment, rhNtA was shown to be selectively PEGylated at the N-terminal. The rhNtA sequence comprises seven primary amines, including the a-amine of Glyl and the E-amine of six Lys residues (Seq. ID No 1), which can theoretically react with HS-PEG-SGA, more specifically with the SGA group. The site-specific addition of the PEG moiety at the N-terminal domain (ex amine) of rhNtA can be controlled by adjusting the pH at which the PEGylation reaction occurs and by adjusting the protei PEG molar ratio. The conjugation reaction was performed at pH 6.5 and using a proteimPEG molar ratio of 1:2. The slightly acidic pH directs the PEGylation reaction to the cx-amino group at the N-terminal of rhNtA, by taking advantage of the differences in pKa values of the a-amino group (pKa 7.8) versus that of the E-amino groups (pKa 10.1) on the side chains of Lys residues. These conditions favour the protonation of the E-amino groups of Lys, while assuring the availability of the a-amino group at the N-terminal as nucleophile. The cation-exchange chromatogram of the conjugation mixture (Fig. 6A) showed three different fractions, which correspond, as evidenced by molecular weight estimation from SDS-PAGE analysis (Fig. 6B), to rhNtA PEG conjugates containing either a double (MW = 36.5 kDa; di-PEGylated rhNtA, peak 1) or single (MW = 28.8 kDa; mono-PEGylated rhNtA, peak 2) PEG molecule and native rhNtA (MW = 14.5 kDa, peak 3). The efficiency of the PEGylation reaction was determined based on the SDS-PAGE analysis (Fig. 6B). A PEGylation yield of 47% was obtained, from which 94% corresponds to mono-PEGylated rhNtA and the remaining 6% to di-PEGylated rhNtA. Since the large hydrodynamic volume of PEGylated proteins retards their mobility on SDS-PAGE gels, resulting in higher molecular weight estimates, PEGylation reaction products were further analysed by MALDI-TOF MS for an absolute estimation of their molecular weight. The MALDI-TOF MS spectra of native and PEGylated rhNtA (Fig. 6C) indicate a molecular weight centred at 18.4 and 22.1 kDa for mono- and di-PEGylated rhNtA, respectively (the MW of native rhNtA was found to be 14.8 kDa, confirming the previously determined value (Figure 2C)).

[0070] The position of the conjugated PEG moiety on the purified mono-PEGylated rhNtA conjugate was further determined by peptide mapping, comparing the fingerprint of native- and PEGylated rhNtA. The majority of the theoretical fragments, as well as the rhNtA N-terminus were experimentally detected in both samples (Table 3). Also, their molecular masses, which were experimentally determined by MALDI-TOF MS, are very similar to the predicted values obtained in web.expasy.org/peptide_mass (Table 3). Two peaks corresponding to the N-terminus of rhNtA domain were identified in both native and mono-PEGylated rhNtA - GPTCPER (FI) [1-7] m/z

816.3677 (Table 3, Fig. 7A) and GPTCPERALER (F2) [1-11] m/z 1285.6278 uj nm rfc (Table 3, Fig. 7B). In the conjugated sample, a significant decrease in the signal-to-noise ratio was observed for FI, while the peak correspondent to F2 was not detected.

[0071] Table 3. Fragments obtained upon digestion of native- and mono-PEGylated rhNtA with trypsin. The N-terminus sequence of rhNtA fragment is depicted in bold.

Molecular Mass (Da)

Sample Fragments Sequence Calculated0 Experimental6 rhNtA 816.3668 816.3677

1-7 GPTCPER

mono-PEG rhNtA 816.3736 rhNtA 1285.6317 1285.6278

1-11 GPTCPERALER

mono-PEG rhNtA c

rhNtA ALERREEEANVVLTGTVEEILN 3741.8752 3798.9145

8-40

mono-PEG rhNtA VDPVQHTYSCK 3798.9299 rhNtA REEEANVVLTGTVEEILNVDPV 3329.6317 3329.6360

12-40

mono-PEG rhNtA QHTYSCK 3329.6489 rhNtA EEEANVVLTGTVEEILNVDPVQ 3173.5306 3173.5747

13-40

mono-PEG rhNtA HTYSCK 3173.5818 rhNtA 715.4362 715.4423

41-45 VRVWR

mono-PEG rhNtA 715.4426 rhNtA 1162.6930 c

46-55 YLKGKDLVAR

mono-PEG rhNtA 1162.6930 rhNtA 758.4519 758.4529

49-55 GKDLVAR

mono-PEG rhNtA 758.4524 rhNtA 932.4745 c

56-64 ESLLDGGNK

mono-PEG rhNtA 932.4745 rhNtA ESLLDGGNKVVISGFGDPLICD 3263.5848 3263.5940

56-86

mono-PEG rhNtA NQVSTGDTR 3263.6089 rhNtA 2350.1343 2350.1294

65-86 VVISGFGDPLICDNQVSTGDTR

mono-PEG rhNtA 2350.1355 rhNtA 1897.0159 1897.0105

87-102 IFFVNPAPPYLWPAHK

mono-PEG rhNtA 1897.0167 rhNtA 3217.6325 3217.6418

IFFVNPAPPILWPAHKNEL

mono-PEG 87-113

MLNSSLMR 3217.6716 rhNtA

rhNtA 1339.6344 1339.6328 mono-PEG 103-113 NELMLNSSLMR

1339.6374 rhNtA

rhNtA 2091.0426 2091.0317 mono-PEG 114-130 ITLRNLEEVEFCVEDKP

2091.0422 rhNtA

rhNtA 1607.7257 1607.7244 mono-PEG 118-130 NLEEVEFCVEDKP

1607.7286 rhNtA

a Theoretical molecular mass of the fragments calculated in web.expasy.org/peptide_mass. b Experimental molecular masses determined by MALDI-TOF MS. c Fragment not found.

[0072] In an embodiment, N-terminal PEGylated rhNtA was shown to bind with high affinity to laminin. SPR analysis was conducted to evaluate the impact of N-terminal PEGylation on rhNtA laminin binding ability. N-terminal PEGylation led to a decrease of the binding affinity for both msLn-111 (KD = 3.36 ± 0.11 nM) and rhLn-521 (KD = 1.58 ± 0.41 nM) (Table 4), estimated as a 3.4-and 2.9- fold decrease, respectively, compared to native rhNtA (Table 2). This shows that PEG conjugation to rhNtA slowed down the rate of domain association (ka) towards immobilized laminin, while the dissociation rate (kd) was not affected (Table 4). Interestingly, the affinity of both native (Table 2) and mono-PEGylated rhNtA (Table 4) for rhLn-521 was found to be slightly higher than that determined for msLn-111. Moreover, the complex formed with this laminin isoform is more stable, as evidenced by higher kd values (Table 2 and 4). The affinity between NtA and laminin is regulated by laminin chain composition, and that although the agrin binding site is located within the gΐ chain, a and b laminin chains may also contribute to the binding, even though to a lesser extent. Therefore, the differences observed in the binding affinity of native and mono-PEGylated rhNtA to msLn-111 and rhLn-521 can be related to the chain composition of the laminin isoforms tested.

[0073] Table 4. Kinetic parameters (ka and kd) and affinity constant (KD) of mono-PEGylated rhNtA to immobilized msLn-111 or rhLn-521, estimated by SPR. Data represent mean ± SD of three independent experiments.


msLn-111 1.38 ± 0.05 4.64 ± 0.31 3.36 ± 0.11 rhLn-521 1.66 ± 0.63 2.47 ± 0.40 1.58 ± 0.41

[0074] In an embodiment, N-terminal PEGylated rhNtA is used for the site-selective immobilization of laminin with retention of bioactivity. The potential of mono-PEGylated rhNtA to promote the site-selective immobilization of laminin onto substrates was assessed by using mixed SAMs of mono-PEGylated rhNtA and (11-mercaptoundecyl) tetraethylene glycol (EG4) (mPEG rhNtA-SAMs) as a model surface. EG4 was used to prevent nonspecific protein immobilization and thus guarantee signal specificity and maximum bioactivity of the immobilized biomolecules, while also enha ncing the correct monolayer packing. Mixed SAMs of HS-PEG-SGA and EG4 (SGA-SAMs) were also prepared to promote the non-selective covalent immobilization of laminin through the reaction with protein amine groups present in multiple locations within the laminin structure.

[0075] In an embodiment, to avoid steric effects and have a finer control over the binding ligand surface density, which ultimately would determine the amount of affinity-bound laminin, an initial screening was conducted to determine the most suitable mono-PEGylated rhNtA domain concentration. QCM-D analysis was conducted with different mono-PEGylated rhNtA concentrations (2.5, 5.0 and 10.0 mM) while keeping that of EG4 constant (100.0 mM), to evaluate laminin immobilization (Fig. 8). Results show a more effective laminin immobilization when an intermediate concentration of mono-PEGylated rhNtA (5.0 mM) was used (638 ± 29 ng/cm2, compared to the 146 ± 115 and 445 ± 57 ng/cm2, obtained for 2.5 and 10.0 mM mono-PEGylated rhNtA, respectively) (Fig. 8). Based on these results, mixed SAMs were prepared using an intermediate concentration (5.0 mM) of either mono-PEGylated rhNtA or FIS-PEG-SGA.

[0076] In an embodiment, infrared reflection absorption spectroscopy (IRRAS) analysis was performed to characterize the prepared mixed mPEG rhNtA- and SGA-SAMs, using Au surfaces as background. EG4-SAMs and adsorbed native rhNtA to Au surfaces were used as controls. The presence of rhNtA domain was confirmed by the detection of the two protein characteristic absorption bands correspondent to amide I (1690 ± 45 cm 1), which is mainly associated with C=0 stretching vibration and is directly related to the backbone conformation, and amide II (1540 ± 60 cm 1), which results from the N-H and C-N vibration that were observed on both native rhNtA and mPEG rhNtA spectra (Fig. 9).

[0077] In an embodiment, site-selective immobilized laminin was shown to retain its ability to self-polymerize. Laminin polymerization is a key process to direct basement membrane assembly and organization and occurs by a thermally reversible process dependent on the presence of calcium ions. Previous studies exploring this hallmark feature of laminin demonstrated that this ECM protein is able to spontaneously self-polymerize in solution at neutral pH, requiring a minimal protein concentration of approximately 60 nM. The three-arm interaction model has been proposed by Cheng and co-workers (Fig. 10A) to explain the mechanism of laminin polymerization. This model proposes that the three laminin N-terminal short arms are able to interact with the globular N-terminal domains of other laminins to form a polygonal network, contributing to the basement membrane architecture. The long arm of the laminin heterotrimer, in turn, is predicted not to be involved in the network formation, being free to interact with cells out of the plan of the polymer. Based on this evidence, we evaluated the ability of laminin to self-polymerize when immobilized onto mPEG rhNtA-SAMs (Fig. 5B, left panel) or SGA-SAMs (Fig. 10B, right panel). Differences between the two conditions were assessed by image processing and quantitative analysis, using ImageJ/Fiji and MATLAB® software, respectively. Four different variables were

measured, including number of aggregates (Fig. IOC), aggregates average size (Fig. 10D), perimeter (Fig. 10E) and maximum Feret's diameter (Fig. 10F). More information about the specific meaning of each measured variable can be found in https://imagej.nih.gov/ij/docs/menus/analyze.html. The relation between the number of aggregates, the aggregates average size and the perimeter give us an indication about laminin ability to form larger organized polygonal-shape structures versus several small unstructured aggregates, while maximum Feret's diameter are good indicators of network organization. An organized polygonal network consisting mainly of lamellar aggregates was consistently observed when laminin was selectively immobilized onto mPEG rhNtA-SAMs (Fig. 10B, left panel). Data from image quantitative analysis (Fig. 10C-F) support this observation, as a higher tendency for the formation of larger polygonal-based (Fig. 10C-E) and organized (Fig. 10F) structures is evident as result of laminin polymerization. Moreover, the formed homogeneous matrix presents a fractal-like organization (Fig. 10B, left panel) and resembles the pattern of the networks reported in literature. On the other hand, the non-selective covalent immobilization of laminin onto SGA-SAMs, led to the formation of a matrix consisting mainly of several smaller aggregates (Fig. 10C-E), which are extended over several focal plans with no observable formation of polygon-based structures (Fig. 10B, right panel). The inability to form a more organized structure is evidenced by the lower values of maximum Feret's diameter (Fig. 10F). The combination of the three independent variables (average size, perimeter and maximum Feret's diameter) and its analysis using the Fisher's combined probability test, also revealed statistically significant differences (p < 0 0001 on laminin self-polymerization between mPEG rhNtA- and SGA-SAMs. Overall, the results obtained showed that depending on the binding ligand (mono-PEGylated rhNtA vs. FIS-PEG-SGA), structural distinct matrices can be formed. Moreover, we clearly demonstrate that only when we promote the site-selective immobilization of laminin (mPEG rhNtA-SAMs), one retains its natural/intrinsic ability to self-polymerize.

[0078] In an embodiment, site-selective immobilized rhLn-521 was found to support hNSC adhesion and spreading. rhLn-521 has a key role in the modulation of neural cell behaviour, including neuronal adhesion, viability and network formation. Indeed, documents show the potential of this laminin isoform to be used as a robust substratum for the culture and renewal of human embryonic and pluripotent stem cells. To assess laminin ability to mediate hNSC adhesion and spreading upon its binding to mPEG rhNtA-SAMs, rhLn-521was used. QCM-D analysis was first conducted to follow the kinetics of rhLn-521 immobilization onto SAMs (Fig. 11 A-B). Then, to characterize whether the proposed strategy favoured the exposure of rhLn-521 cell-adhesive domains, hNSC adhesion and spreading was assessed (Fig. 11 C-E).

[0079] In an embodiment, QCM-D monitoring of rhLn-521 immobilization onto mPEG rhNtA-SAMs indicated a protein immobilization (638 ± 29 ng/cm2, Fig. 11B), with mass per surface area values close to the theoretical value calculated for laminin immobilization under these conditions (952 ng/cm2). This suggests that the immobilized mono-PEGylated rhNtA retained a 3D conformation that allowed the selective interaction with laminin. Moreover, mPEG rhNtA SAMs led to an immobilization degree statistically similar to that observed for uncoated Au surfaces (1130 ± 348 ng/cm2), in which laminin may randomly adsorb over the entire surface. SGA-SAMs, in contrast, presented lower amounts of immobilized rhLn-521 (342 ± 44 ng/cm2) (Figure 11B), when compared to mPEG rhNtA-SAMs. As HS-PEG-SGA moieties can randomly interact with different amine groups within a single laminin molecule, these results suggest that the conformation of covalently-bound laminin and its spatial arrangement at the surface were distinct from that of affinity-bound laminin, leaving less area for the immobilization of other laminin molecules. As expected, the amount of adsorbed rhLn-521 on EG4-SAMs, was very low (20 ± 1 ng/cm2) (Fig. 11B).

[0080] In an embodiment, site-selective immobilization of laminin onto mPEG rhNtA-SAMs led to hNSC adhesion comparable to that observed on rhLn-521-coated Au surfaces, used here as a control surface for hNSC adhesion (Fig. 11D and Fig. 12). After 24 h of culture, cells were well-spread (Fig. HE and Fig. 13), exhibiting a morphology (Fig. 11C) comparable to that observed for cells seeded on rhLn-521-coated Au surfaces surfaces (Fig. 11C). Moreover, cells adhered to mPEG rhNtA-SAMs exhibit an average spreading area of 2812 ± 89 pm2 (Fig. 11E), comparable to that observed for cells cultured on rhLn-521-coated Au (3089 ± 101 pm2; Figure HE). TCPS surfaces, which are conventional used in adhesion studies, were used herein as additional control surfaces. The morphology of cells seeded on rhLn-521- site-selective immobilized on SAMs or -coated on Au surfaces is comparable to that observed on rhLn-521-modified TCPS surfaces (Fig. 14). In addition, the average cell spreading area in these conditions is similar/equivalent to that observed in TCPS surfaces (3496 ± 483 pm2). The non-selective immobilization of laminin onto SGA-SAMs, in turn, resulted in a significantly lower cell adhesion (p = 0.0191 ; Fig. HD and Fig. 12) and average spreading area (1430 ± 161 pm2; p < 0.0001 ; Fig. HE and Fig. 13) when compared to mPEG rhNtA-SAMs. Moreover, cells with both spread (Fig. HC) and round (Fig. 14) morphology were observed. These results suggest that the non-selective nature of the immobilization process partially compromises the exposure of laminin cell binding domains, despite the considerable amount of rhLn-521 bound to SGA-SAMs (342.3 ± 43.6 ng/cm2) (Fig. HB). As expected, on EG4-SAMs, the few adherent cells (Fig. HD and Fig. 12) presented mainly a round morphology (Fig. HC and Fig. 14), as evidenced by the low average cell spreading area (881 ± 70 pm2, Fig. HE). Overall, these results clearly demonstrate that site-specific immobilization of laminin through the use of mono-PEGylated rhNtA better preserved laminin bioactivity in terms of ability to self-polymerize and to mediate hNSC adhesion and spreading, when compared to non-selective covalent immobilization approaches.

[0081] The present disclosure demonstrates the potential of mono-PEGylated rhNtA as an effective natural affinity binding ligand for site-selective immobilization of laminin, allowing the preservation of laminin ability to self-polymerize and mediate cell adhesion and spreading. This affinity binding strategy overcomes several drawbacks associated with the currently available strategies for laminin immobilization. Moreover, this approach is highly versatile, as result of the ability of NtA to bind the different laminin isoforms that comprise the gΐ chain, which represent more than 50% of the isoforms identified to date, with variations in affinity imposed by a and b chains. Therefore, this strategy enables the immobilization of different laminin isoforms, which can be of interest for particular cell types and for application in specific disease contexts. Overall, the proposed strategy is important for a broad range of applications, including 2D coatings for cell culture, functionalization of 3D matrices for cell and/or drug delivery, engineered coatings for neuroelectrodes, among others.

[0085] In an embodiment, rhNtA was explored for attaining the site-selective immobilization of laminin within a degradable synthetic hydrogel. A four-arm maleimide terminated poly(ethylene glycol) (PEG-4MAL) macromer was used as the hydrogel base material (Fig. 15), due to excellent cytocompatibility and in vivo tolerance, and well characterized biochemical and biophysical properties. PEG-4MAL macromer was functionalized with a thiol-containing mono-PEGylated rhNtA (NtA) domain, for the site-selective immobilization of laminin to the hydrogels (Fig. 15). Laminin-111 (at a concentration of 100 pg/mL) was the isoform selected for the development of PEG-4MAL hydrogels with affinity-bound laminin, since it was already shown to successfully promote NSC neuronal outgrowth and differentiation. To allow the hNSC 3D culture that would span the time frame of hNSC proliferation and neuronal differentiation, a mix of protease degradable (cysteine-flanked matrix metalloproteinase 2 (MMP2)-sensitive peptide) and nondegradable (PEG-dithiol) cross-linkers was used to mediate, under physiological conditions, the formation of a degradable hydrogel network (Fig. 15).

[0082] In an embodiment, the hydrogel components are described. A four-arm maleimide terminated poly(ethylene glycol) (PEG-4MAL; 40 kDa, > 95% purity, > 90% substitution, Jenkem USA) was chosen as the hydrogel base material. A macromer solution was prepared in 10 mM 4-(2-hydroxyethyl)piperazine-l-ethanesulfonic acid (HEPES) in phosphate buffered saline (PBS, pH 7.4) at a final polymer concentration of 10.0% (w/v). Mono-PEGylated recombinant human N-terminal

agrin (NtA) domain was produced as previously described (D. Barros, P. Parreira, J. Furtado, F. Ferreira-da-Silva, E. Conde-Sousa, A.J. Garcia, M.C.L. Martins, I.F. Amaral, A.P. Pego, An affinity-based approach to engineer laminin-presenting cell instructive microenvironments, Biomaterials 192 (2019) 601-611). Laminin-111 from mouse Engelbreth-Holm-Swarm sarcoma (msLn-111) was obtained from Sigma-Aldrich. A cysteine-flanked matrix metalloproteinase-2 (MMP2)-sensitive peptide (Acetyl (AcJ-GDCDDSGESPAY^YTADDCDG-Amide (NF ), in which the symbol -J- depicts the MMP2 cleavage site; 2.1 kDa, > 85% purity, GenScript) and a PEG-dithiol (3.5 kDa, > 95% purity, Jenkem USA) were used as the degradable and non-degradable cross-linkers, respectively.

[0083] In an embodiment, the tethering efficiency of NtA to PEG-4MAL was determined by quantifying free/unreacted thiols in the reaction mixture using the Measure-iT thiol assay kit (Invitrogen) according to the manufacturer's instructions. Briefly, PEG-4MAL was functionalized with the NtA domain (input NtA concentration: 10, 50 or 100 mM) for 15 min in 10 mM HEPES in PBS (pH 7.4). The samples were then mixed with thiol-quantitation reagent in a black 96-well plate and the fluorescence measured (lQC = 494 nm; em = 517 nm) using a microwell plate reader. A calibration curve of glutathione (0 - 55 mM) was used to calculate the concentration of free thiols.

[0084] In an embodiment, the preparation of affinity-bound laminin PEG-4MAL hydrogels is disclosed (Fig. 15). Briefly, PEG-4MAL was first functionalized with the NtA domain (input NtA concentration in the gel: 10, 50, or 100 mM) for 15 min in 10 mM HEPES in PBS (pH 7.4). msLn-111 (input laminin concentration in the gel: 100 pg/mL) was then added to the mixture and allowed to bind to NtA for 30 min. Functionalized PEG-4MAL precursors were then cross-linked into a hydrogel by addition of a mix of a MMP2-sensitive peptide and a PEG-dithiol prepared in 10 mM HEPES in PBS (pH 6.5), at different % molar ratios (100:0; 80:20; 50:50; 20:80; 0:100). The cross-linkers were added at a 1:1 molar ratio of cysteine residues on the cross-linkers to remaining maleimide groups on NtA-functionalized PEG-4MAL. Hydrogels were polymerized at 37 °C, 5% CO2 for 15 min. Unmodified PEG-4MAL gels and PEG-4MAL gels containing entrapped msLn-111 (100 pg/mL) were also prepared and used as controls.

[0085] In an embodiment, laminin incorporation within PEG-4MAL hydrogels was quantified by an enzyme-linked immunosorbent assay (ELISA). The hydrogels were prepared as previously described and incubated at 37 5C in 125 pL of PBS (pH 7.4). At day 1 and 7, the supernatant was collected, stored at -80 5C and then analyzed using a mouse laminin ELISA kit (Abeam, abll9572) according to the manufacturer's instructions. Absorbance values were measured at 450 nm using a microwell plate reader.

[0086] In an embodiment, the rheological properties of PEG-4MAL hydrogels were characterized using a strain-controlled Kinexus Pro Rheometer (Malvern Instruments Ltd, Malvern, UK) and an 8 mm diameter parallel-plate geometry. Hydrogels for each condition (0 8 mm) were prepared as described above, and allowed to hydrate in PBS (pH 7.4) at 37 °C for 24 h. The samples were then tested in a humidified environment at physiological temperature (37 °C) and with application of 30% compression (oscillatory measurement gap). Amplitude strain sweeps (0.1 - 100% at 0.1 Hz) were initially performed for each condition to define the linear viscoelastic region (LVR). Frequency sweeps (0.01 - 10 Hz at 1% strain) were then performed and the storage (G'), loss (G") and complex (G*) modulus, as well as the phase angle (d, °) were determined within the LVR, by averaging all data points acquired from a 0.01 - 1 Hz interval. The relative mesh size (nm) value was estimated using the following equation:


where G' = storage modulus in Pa, A = Avogadro's constant (6.022140857 x 1023 mol-1), R = gas constant (8.314 m3.Pa.mol _1.K _1), T = temperature (37 °C = 310.15 K).

[0087] In an embodiment, the culture of human neural stem cells (hNSC) is described. hNSC (N7800-200, Life Technologies) were expanded according to the manufacturer's protocol in expansion medium - StemPro® NSC serum-free medium (SFM; Life Technologies) supplemented with bFGF (20 ng/mL) and EGF (20 ng/mL).

[0088] In an embodiment, the culture of hNSC within affinity-bound laminin PEG-4MAL hydrogels is described. hNSC were dissociated into single cells using StemPro accutase cell dissociation reagent (Life Technologies) and further suspended in the solution containing laminin-functionalized PEG-4MAL precursors (4 x 10s viable cells/mL). Cell-laden hydrogels were formed by mixing the PEG-4MAL precursor solution containing cells with the MMP2-sensitive peptide and PEG-dithiol cross-linkers, dissolved in 10 mM HEPES in PBS (pH 6.5), at different molar ratios (%), and subsequently incubating the polymerizing gels at 37 °C, 5% CO2 for 15 minutes. hNSCs were initially cultured in expansion medium and then induced to differentiate along the neuronal lineage by growth factor withdrawal. Briefly, at day 2 of culture, the medium was switched to the StemPro NSC SFM media-Neurobasal/B27 (Life Technologies) (1:1) mix, without growth factors. Briefly, at At day 8, half of the medium was replaced by the StemPro NSC SFM-Neurobasal/B27 (1:3) mix supplemented with 10 ng/mL of brain-derived neurotrophic factor (BDNF; Peprotech) and 500 mM of N6, 2'-0-Dibutyryladenosine 3', 5' -cyclic monophosphate sodium salt (dibutyril cAMP; Sigma).

Half of the medium was changed every other day, up to 14 days. hNSCs cultured within unmodified and PEG-4MAL gels containing entrapped msLn-111 (100 pg/mL) were herein used as controls.

[0089] In an embodiment, the evaluation of cell viability is described. The distribution of viable and dead cells within PEG-4MAL hydrogels was assessed at day 7 of cell culture using a Live/Dead assay. Cell-hydrogel matrices were rinsed with warm PBS pH 7.4 and incubated with 4 mM calcein AM (Life Technologies) and 6 mM propidium iodide (PI; Life Technologies) at 37 °C for 30 min, for detection of viable and dead cells, respectively. After incubation, the samples were rinsed twice with PBS pH 7.4, transferred to the culture medium, and immediately observed under confocal laser scanning microscopy. Quantitative analysis of live and dead cells was also assessed using immunocytometry performed in six-pooled hydrogels, after hydrogel dissociation. Briefly, the hydrogels were sequentially incubated with 1.25 mg/mL of collagenase type II (Gibco; lh at 37 °C) and StemPro accutase cell dissociation reagent (20 min at 37 °C) under stirring (70 rpm). Cells were mechanically dissociated by pipetting, diluted in Glasgow Minimal Essential Medium (GMEM; Life Technologies) supplemented with 10% (v/v) inactivated fetal bovine serum (FBS) and centrifuged. For live/dead staining, cells were then transferred to a round-bottomed 96-well plate and stained with calcein AM (67 nM, 20 min at 37 °C) and propidium iodide (PI; 6 mM, 10 min at 37 °C). Cells were finally washed trice and suspended in FACS buffer (PBS pH 7.4 supplemented with 2% (v/v) FBS) for flow cytometry analysis on BD FACSCantoTM II (BD Biosciences). Unlabeled cells were used to set the fluorescence gates and cells stained with calcein AM and PI only were used to establish the compensation settings. For each flow cytometry analysis, 10,000 events were acquired inside the respective gate.

[0090] In an embodiment, the screening of degradable: non-degradable cross-linker molar ratio (%) is disclosed. After 14 days in culture, the viability of hNSC within hydrogels cross-linked with different molar ratios (%) of cysteine-flanked MMP2-sensitive peptide (SGESPAY^YTA) and PEG-dithiol was assessed using a Live/Dead assay, performed as described above. Average occupied area (pm2) by live (Calcein+) cells was assessed in the projected two-dimensional (2D) images through the application of an Otsu thresholding method and determination of the total area occupied by cells in each image.

[0091] In an embodiment, the growth of hNSC cultured within PEG-4MAL hydrogels was estimated from total cell number after 7 days of cell culture, determined using the CyQUANT® cell proliferation assay kit (Life Technologies) according to the manufacturer's instructions. Briefly, cells were extracted from the hydrogels, through sequential incubation with 1.25 mg/mL of collagenase type II (Gibco; 1 h at 37 °C) and StemPro accutase cell dissociation reagent (Life Technologies; 20 min at 37 5C) under stirring (70 rpm). Cells were mechanically dissociated by pipetting, diluted in Glasgow Minimum Essential Medium (GMEM; Life Technologies) supplemented with 10% (v/v) inactivated FBS, centrifuged and the cell pellet stored at -80 5C. The cell pellet was then thawed at room temperature (RT) and incubated with CyQUANT® GR dye/cell lysis buffer. Fluorescence (lQC = 480 nm; em = 520 nm) was measured, after mixing with CyQUANT GR dye, in a microwell plate reader. The total number of cells for each condition was estimated from a standard curve generated with a known amount of hNSCs over a range of 50 to 250,000 cells. Cell samples used to generate the standard curve were measured in triplicate. For each condition, hydrogels without cells, cultured in parallel and processed similarly to those with cells, were used as blanks and their background fluorescence values subtracted.

[0092] In an embodiment, hNSC neurite outgrowth and phenotypic analysis is described. The effect of laminin site-selective immobilization on neurite outgrowth and cell phenotype was assessed after 14 days of culture, in samples processed as whole-mounts for immunofluorescence staining of Nestin (neural stem/progenitor cells), bIII-tubulin (developing and mature neurons) microtubule-associated protein 2 (MAP2; neuronal dendrites and cell bodies), and Tau (mature neurons, axonal marker). Cell-laden hydrogels were fixed in 3.7% (w/v) paraformaldehyde (PFA) solution diluted 1:1 in culture media (30 min; 37 °C) and permeabilized with 0.2% (v/v) Triton X-100 in PBS (45 min; RT). Samples were then incubated with blocking buffer (5% (v/v) BSA in PBS) for 1 h at RT, followed by incubation with the primary antibody - mouse anti-pill-tubulin (1:500; Biolegend, 801201); mouse anti-Nestin (1:200; Abeam, AB22035); mouse microtubule-associated protein 2 (MAP2; 1:200; Invitrogen, 13-1500); rabbit anti-Tau (1:100; Sigma-Aldrich, T6402), overnight (ON) at 4 °C. To detect primary antibodies, samples were incubated with Alexa Fluor® 488 conjugated anti-rabbit secondary antibody (1:1000; Life Technologies, A21206) or Alexa Fluor® 594 conjugated anti-mouse secondary antibody (1:1000; Life Technologies, A11020) ON at 4 °C. The nuclei were counterstained with 0.1 pg/mL Hoechst 33342 for 20 min at RT. Samples were observed under confocal laser scanning microscopy, and z-sections covering a thickness of 200 pm or 80 pm were acquired using an HC Plan APO CS lOx / 0.40 NA or HC Plan APO CS 20x / 0.70 NA objective, respectively. Neurite outgrowth was assessed in the projected two-dimensional (2D) images stained for bIII-tubulin/DNA using a developed ImageJ script. Briefly, nuclei were segmented through an Otsu thresholding method, while neurites were segmented through subtraction of the previously identified nucleus, using a Huang thresholding method. Measurements were obtained for total neurite length (pm) and total number of neurites per image.

[0093] Statistical analysis was performed using GraphPad Prism 6 software (San Diego). Sample distribution was initially tested for normality using the Kolmogorov-Smirnov test. Comparisons between three or more groups were performed with one-way ANOVA analysis, followed by the Bonferroni correction for pairwise comparisons or the Dunnett's two-tailed test for comparisons with control conditions. For all analysis, differences were considered significant at p < 0.05.

[0094] In an embodiment, PEG-4MAL hydrogels were shown to provide independent control of different biochemical and biophysical properties. A hydrogel platform based on a PEG-4MAL macromer was selected as the base material for this work, as we have seen it exhibits excellent in vitro and in vivo biocompatibility with different cell types, including muscle stem cells and pancreatic islets. Moreover, the modular nature of PEG-4MAL hydrogels allows independent control over different biochemical and biophysical properties such as type and density of cell-adhesive ligands, mechanical and structural properties, and protease-dependent degradation. Therefore, the use of these matrices is especially advantageous to assess the effect of a particular biochemical or biophysical cue on the modulation of cell behavior. The described properties allowed us to initially define the molecular weight (40 kDa) and the final polymer density (10% (w/v) of the PEG-4MAL macromer), that would result in hydrogels with mechanical properties within the range of values preferred for NSC proliferation (100 - 1000 Pa) and neurite branching and extension (200 - 400 Pa).

[0095] In an embodiment, control over affinity-bound laminin PEG-4MAL hydrogel degradability was shown to be required for hNSC long-term culture. Hydrogel degradability, a critical factor for matrix remodelling, is essential for in vitro cell survival and proliferation within synthetic niches, and, therefore, is a key feature to take into consideration when designing cell instructive microenvironments. In this context, the incorporation of protease-sensitive peptide cross-linkers has been widely explored to better control material degradation. Among the different protease-degradable peptides characterized to date, a fast-degrading sequence shown to be relatively specific for MMP2 (J. Patterson, J.A. Hubbell, Enhanced proteolytic degradation of molecularly engineered PEG hydrogels in response to MMP-1 and MMP-2, Biomaterials 31(30) (2010) 7836-45) was used, as this metalloproteinase is expressed at high levels by NSCs. The fast gelation kinetics of the thiol-maleimide reaction (Michael-type addition) may lead to the formation of heterogeneous hydrogels, which in turn can translate into variable cell biological outcomes. Consequently, to increase the pKa of the thiol group and thereby reduce the availability of the reactive thiolate, the di-thiol-containing MMP2-sensitive cross-linker peptide used in this work was designed to include negatively charged amino acids flanking the thiol-bearing cysteine residue (DCDD). This strategy allowed the thorough mixing of the hydrogel precursor solutions prior to use.

[0096] The fast-degrading nature of the selected MMP2 sequence, combined with the high levels of MMP2 production by NSCs, led us to hypothesize that fully degradable hydrogels might not be able to support the 3D culture of hNSCs, up to 14 days. To test this hypothesis, affinity-bound laminin hydrogels cross-linked by addition of a MMP2-sensitive peptide (degradable) and a PEG-dithiol (non-degradable) at different % molar ratios (100:0; 80:20; 50:50; 20:80; 0:100) were evaluated and the combination allowing the hNSC 3D culture that would span the time frame of hNSC proliferation and neuronal differentiation established. hNSCs showed a significantly higher number of viable cells, with cellular extensions at day 14 of cell culture, in gels formulated with a combination of degradable and non-degradable cross-linkers, specifically for the 80% MMP2-sensitive peptide to 20% PEG-dithiol cross-linker molar ratio, when compared to fully degradable hydrogels (p = 0.018 vs. 100:0 cross-linker molar ratio (%); Fig. 16). As such, all subsequent studies were performed using this degradable and non-degradable cross-linkers molar ratio (%).

[0097] In an embodiment, Mono-PEGylated rhNtA (NtA) was shown to mediate efficient site-selective laminin incorporation. Maleimide groups in PEG-4MAL macromer efficiently react with thiol-containing peptides through Michael-type addition. This enables a fine-tuning over ligand incorporation and cross-linking efficiency, ultimately leading to the formation of structurally defined matrices. To favor the site-selective immobilization of laminin, PEG-4MAL was first reacted with a thiol-containing NtA domain (Fig. 15). This domain was tethered onto the PEG-4MAL macromer with high yields, for all the concentrations tested (10, 50 and 100 mM), evidencing a precise control over NtA density (Fig. 17A). The effect of different NtA concentrations on laminin incorporation within hydrogels was then assessed. Efficient laminin incorporation was observed for all the NtA concentrations tested, as evidenced by ELISA results (incorporation efficiency > 95%, determined after 1 and 7 days of incubation in PBS; Fig. 17B). We hypothesize that the lack of significant differences between laminin entrapped vs. immobilized can be explained by limitations on protein diffusion imposed by the reduced hydrogel mesh size (20 - 30 nm), which will hinder the diffusion of laminin due to its high hydrodynamic diameter (= 42 nm). As free thiol groups in laminin could also react with maleimide groups from PEG-4MAL and thus contribute to the laminin immobilization into PEG-4MAL hydrogels, the availability of free/unreacted thiols in laminin stock solution (pH 7.4) was assessed. Free thiols were not detected (< 2.75 mM, the lower detection limit of the assay), indicating that the molar ratio of free thiols in laminin was less than 0.22%, and assuring minimal contribution of laminin free thiol groups to laminin incorporation into PEG-4MAL hydrogel.

[0098] In an embodiment, laminin-modified hydrogels revealed mechanical properties within the values preferred for NSC growth and differentiation. Hydrogel mechanical and structural (equilibrium swelling and mesh size) properties have a key role on the modulation of different cellular functions ( e.g . survival, proliferation, and differentiation. The fine control of these properties is, therefore, crucial to direct neural stem cell fate. The effect of the binding ligand (NtA) density and laminin incorporation on hydrogel mechanical properties and structure (mesh size) was assessed through rheological studies. All the prepared hydrogels presented Storage modulus (G') > Loss modulus (G") (Table 5), showing that PEG-4MAL has transitioned from a viscous liquid to a gelled state. Moreover, the hydrogels presented a "solid-like" behavior after 24h at 37 °C, as evidenced by the values of phase angle (d) below 10° (Table 5), meaning that the viscoelastic properties of the hydrogels were stable for the conditions tested.

[0099] Table 5. Storage modulus (G'), Loss modulus (G") and Phase angle (d,°) of hydrogels determined by rheological analysis. Data represent mean ± Standard error of the mean (SEM); n = 6.

Input NtA

Storage Modulus Loss Modulus Phase Angle

Hydrogel concentration

(G', Pa) (G", Pa) (d, ») (mM)

Unm 259.1 ± 13.8 10.6 ± 1.7 2.3 ± 0.2


[00100] The covalent immobilization of increasing concentrations of NtA (from 10 to 100 mM) led to a trend for reduced in the storage modulus (G') (Table 5) and to the formation of a slightly looser polymer network, as evidenced by the tendency for larger mesh size values (Fig. 18B). The tethering of NtA reduces the number of maleimide-terminated PEG chains available for cross-linking, thus reducing the number of cross-linking points. This is in an excellent agreement with the trend for a lower stiffness observed with increasing concentrations of NtA (Fig. 18A). These results suggest that the approach used for laminin incorporation (physical entrapment vs. affinity immobilization) impacts the hydrogel mechanical and structural properties. Results presented are consistent with previous studies, exploring PEG-diacrylate hydrogels, in which it was shown that the covalent

immobilization of laminin through a PEG chain significantly reduces the hydrogel stiffness, as opposed to the physical entrapment of the protein (L. Marquardt, R.K. Willits, Neurite growth in PEG gels: effect of mechanical stiffness and laminin concentration, J Biomed Mater Res A 98(1) (2011) 1-6).

[00101] In an embodiment, affinity-bound laminin PEG-4MAL hydrogels revealed to support hNSC culture and biological function. The potential of PEG-4MAL hydrogels incorporating affinity-bound laminin to support hNSC biological function, namely viability, proliferation, and outgrowth, was subsequently investigated. Laminin incorporation within PEG-4MAL hydrogels using 10 and 50 mM NtA, favored the formation of stiffer hydrogels (Fig. 18A), able to better support hNSC culture for longer periods (up to 14 days). As result, these hydrogels were used to perform the qualitative and quantitative analysis of cell behavior. PEG-4MAL hydrogels unmodified or containing physically entrapped laminin were also prepared and used as controls. To get insight into the effect of controlled immobilization of laminin on hNSC survival and proliferation, cells were cultured for 7 days within degradable PEG-4MAL hydrogels with either entrapped or immobilized laminin. Confocal microscopy showed the presence of live cells widely distributed throughout all hydrogels, often growing as cellular spheroids (Fig. 19A). The quantitative analysis of live/dead cells, assessed by flow cytometry, showed values of average cell viability greater than 75% for all the conditions tested, evidencing the good cytocompatibility of the proposed matrices (Fig. 19B) (I.O. f. Standardization, ISO 10993-5:2009: Biological evaluation of medical devices Tests for in vitro cytotoxicity, 2009). In contrast to the physical entrapment of laminin, site-selective immobilization of laminin using 10 mM and 50 mM NtA supported higher hNSC growth compared to unmodified hydrogels, with significantly higher cell numbers being observed in 10 mM NtA - laminin immobilized hydrogels (p = 0.0180) (Fig.l9C ). An increase in the cell number of 1.5-, 1.9-, 4.2- and 3.1-fold (vs. initial cell density - 4 x 104 cells/hydrogel) was observed for unmodified, laminin entrapped, 10 mM-and 50 mM NtA - laminin immobilized hydrogels, respectively (Fig. 19C). Since laminin retention in PEG-4MAL hydrogels was similar regardless of being physically entrapped or affinity-bounded (Fig. 17B), these results suggest that the site-selective immobilization of laminin through the use of 10 mM of NtA, better preserved the exposure of laminin domains interacting with cell adhesion receptors, as compared to physically entrapped laminin.

[00102] In an embodiment, controlled laminin immobilization promoted neurite outgrowth. Laminin presents several bioactive domains with neurite outgrowth promoting ability, and, as a result, plays a key role in mediating NSC migration, differentiation and neurite extension both in vitro and in vivo. The effect of laminin site-selective immobilization on hNSC neurite outgrowth was assessed after 14 days of culture under differentiation conditions (Fig. 20). A population of cells expressing the neuronal marker bIII-tubulin, with the evident formation of neuronal processes, was observed within 10 mM NtA affinity-immobilized laminin PEG-4MAL hydrogels (Fig. 20A). Differences between the conditions tested were assessed by image processing and quantitative analysis, using ImageJ/Fiji. Total neurite length (pm) and total number of neurites were quantified (Fig. 20B-D). The controlled immobilization of laminin using 10 mM NtA supported a more pronounced neurite extension, compared with any of the other conditions tested, as evidenced by the higher values of total neurite length (p = 0.0035 vs. Unm; p = 0.0152 vs. Entrap; Fig. 20B) and total number of neurites (p = 0.0007 vs. Unm; p = 0.0079 vs. Entrap; Fig. 20C). The use of 50 mM NtA, despite leading to similar amounts of laminin incorporation (Fig. 20B), resulted in significantly lower values of total neurite length (p = 0.0056; Fig. 20C) and total number of neurites (p = 0.0073; Fig. 20D) compared to 10 mM NtA affinity-bound laminin. The reduction in neurite outgrowth (Fig. 20C-D) and the reduction in cell number (Fig. 19C) observed for cells cultured within hydrogels with 50 mM NtA - immobilized laminin may be related with the inability of the protein to bind the NtA domain, as result of steric effects. The results obtained contrast to previous studies in which no significant differences on neurite outgrowth were observed when laminin was covalently immobilized using a non-selective approach, vs. physically entrapped laminin (L. Marquardt, R.K. Willits, Neurite growth in PEG gels: effect of mechanical stiffness and laminin concentration, J Biomed Mater Res A 98(1) (2011) 1-6). Overall, these results clearly demonstrate that the site-selective immobilization of laminin through the use of 10 mM NtA better preserved protein bioactivity in terms of ability to promote hNSC proliferation and outgrowth, when compared to physically entrapped laminin.

[00103] In an embodiment, controlled laminin immobilization was shown to support hNSC sternness and neuronal maturation. A more detailed phenotypic analysis was performed at day 14 of cell culture by immunocytochemistry (Fig. 21). At this time point, Nestin+ cells (Fig. 21A) were observed for all the conditions tested, thus evidencing the potential of the proposed hydrogels to support hNSC sternness over time. This is a key feature when envisaging the use of hydrogels for cell transplantation, as they should provide an adequate microenvironment for the maintenance of a pool of NSC, upon transplantation. The ability of the proposed hydrogels to support hNSC neuronal maturation was further assessed using markers for mature neurons, more specifically MAP2 and Tau. The qualitative analysis of 2D projected CLSM images revealed the presence of differentiated neurons expressing MAP2 and Tau in all the conditions. Yet, axonal extensions staining positively for Tau were more evident in 10 mM NtA affinity-immobilized laminin PEG-4MAL hydrogels, which suggests that these hydrogels are more permissive to hNSC neuronal maturation when compared to unmodified gels and gels containing physically-entrapped laminin, in line with their higher neurite-promoting ability.

[00104] The present disclosure demonstrates the potential of a PEG-4MAL hydrogel modified with affinity-bound laminin as a dynamic 3D platform enabling NSC proliferation, neuronal differentiation, and neurite extension. The proposed strategy allows the oriented and controlled immobilization of laminin, with preservation of protein bioactivity, thus overcoming some of the main drawbacks associated with the currently available strategies (i.e. physically entrapment and non-selective covalent immobilization). Moreover, the immobilization approach used in this work is highly versatile, as result of the ability of NtA to bind with high affinity to different laminin isoforms comprising the gΐ chain, which represents more than 50% of the isoforms identified to date. Therefore, the reported 3D matrices can be used for the site-selective immobilization of different laminin isoforms and used for the analysis of cell-laminin interactions occurring in different stem cell niches and disease contexts. These hydrogels can be also of interest to provide a controlled niche with instructive cues to improve survival, engraftment and long-term function of transplanted stem cells; for the creation of tissue-engineered constructs; and for the development of cell expansion and biomanufacturing systems.

[00105] Where ranges are given, endpoints are included. Furthermore, it is to be understood that unless otherwise indicated or otherwise evident from the context and/or the understanding of one of ordinary skill in the art, values that are expressed as ranges can assume any specific value within the stated ranges in different embodiments of the invention, to the tenth of the unit of the lower limit of the range, unless the context clearly dictates otherwise. It is also to be understood that unless otherwise indicated or otherwise evident from the context and/or the understanding of one of ordinary skill in the art, values expressed as ranges can assume any subrange within the given range, wherein the endpoints of the subrange are expressed to the same degree of accuracy as the tenth of the unit of the lower limit of the range.

[00106] The disclosure should not be seen in any way restricted to the embodiments described and a person with ordinary skill in the art will foresee many possibilities to modifications thereof.

[00107] The above described embodiments are combinable.

[00108] The following claims further set out particular embodiments of the disclosure.

REFERENCES

1. Aumailley, M., The laminin family. Cell Adh Migr, 2013. 7(1): p. 48-55

2. Taylor, A.C., et al., Surface functionalisation of nanodiamonds for human neural stem cell adhesion and proliferation. Sci Rep, 2017. 7(1): p. 7307.

3. Stabenfeldt, S.E., A.J. Garcia, and M.C. LaPlaca, Thermoreversible laminin-functionalized hydrogel for neural tissue engineering. J Biomed Mater Res A, 2006. 77(4): p. 718-25.

4. Lee, J.P., et al., N-terminal specific conjugation of extracellular matrix proteins to 2- pyridinecarboxaldehyde functionalized polyacrylamide hydrogels. Biomaterials, 2016. 102: p. 268-76

5. Mascarenhas, J.B., et al., Mapping of the laminin-binding site of the N-terminal agrin domain (NtA). EMBO J, 2003. 22(3): p. 529-36.

6. Denzer, A.J., et al., Electron microscopic structure of agrin and mapping of its binding site in laminin-1. EMBO J, 1998. 17(2): p. 335-43.