Skip to main content
European Commission logo
polski polski
CORDIS - Wyniki badań wspieranych przez UE
CORDIS
CORDIS Web 30th anniversary CORDIS Web 30th anniversary
Zawartość zarchiwizowana w dniu 2024-06-18

Find and Bind: Mastering sweet cell-instructive biosystems by copycat nano-interaction of cells with natural surfaces for biotechnological applications

Final Report Summary - FIND AND BIND (Find and Bind: Mastering sweet cell-instructive biosystems by copycat nano-interaction of cells with natural surfaces for biotechnological applications)


Executive Summary:

Living cells are complex entities with remarkable capacity to sense, integrate and respond to environmental cues. Undoubtedly, carbohydrates are a class of molecules which together with the proteins are involved in these processes. Find and Bind intends to explore the potential of this class of molecules to mediate specific recognition events and therefore to provide modulation of certain biological processes. Thus, the project aims to create biological design criteria for the development of new materials and devices constructed from these materials by understanding, mimicking and mastering the nanoscale mechanisms of the interactions of cells and their in vivo environment. The long-term innovation potential of the developed constructs as (i) 3D cell instructive materials able to restore and enhance the functions of healthy tissues; (ii) Biosensors and (iii) Surfaces for selective differentiation of stem cells also are in the targets of this project.

To achieve the project’s scientific and technological objectives, the consortium of Find and Bind is using an interdisciplinary approach, where the convergence of several sciences is exploited:

• The tools of chemistry and biochemistry are employed to obtain specific targeted oligosaccharides under WP1;
• The physical interactions of the obtained macromolecules with specifically designed surface chemistries are used for stable immobilisation of these oligosaccharides onto different platforms (WP2);
• 2D platforms and 3D bioactive oligosaccharide-based structures are built up using the advances of nanotechnology and materials sciences (WP2 and WP4);
• The obtained carbohydrates are evaluated for their effect onto different bioentities crosstalk (among them proteins and cells) using the tools of molecular and basic biology (WP3).

Project Context and Objectives:

Contemporary technology has brought clear improvements in the field of substitution medicine where the developments of artificial joints or first scaffolds for tissue engineering are only examples of its enormous benefits. Although some materials that adequately stimulate the regeneration of functional tissues such as bone are already available, the precise cues to regulate the cellular behaviour and tissue formation are not enough lightened or not yet applied. Therefore, despite the enormous benefits the current technology has brought, the outer limits have been reached and new breakthroughs can only be expected from novel converging technologies that will overcome the shortcomings of the current materials technology. An excellent example of such combined, biology driven approach is the field of nanobiotechnology (NBT), in which existing elements of nature are studied at the nanoscale in order to fabricate new devices for biological, biotechnological or biomedical applications. Using NBT as basis, Find and Bind is aimed at creating biological design criteria for the development of new materials and devices constructed from these materials by understanding, mimicking and mastering the nanoscale mechanisms of the interactions of cells and their in vivo environment. Although these mechanisms involve different bioactive molecules, until now NBT has mainly relied on strategies exploiting small peptide epitopes, copycatting specific sequences of more complex proteins (with tertiary and quaternary structures). The breakthroughs that such approaches brought to biomaterials and tissue engineering fields are undeniable. Nevertheless, the important role that carbohydrates play in tissue structuring and function is still poorly explored and exploited and we believe this is a missing important piece in the NBT puzzle. Hence, the project aims to explore carbohydrates in an integrated approach with other bioactive macromolecules such as proteins. Together, they shall be orchestrated to re-create the in vivo scenario in terms of both signals and timing and to develop third-generation carbohydrate based constructs.

The long-term innovation potential of the developed constructs as (i) 3D cell instructive materials able to restore and enhance the functions of healthy bone and/or cartilage tissue; (ii) Kits to support culturing and differentiation of adult stem cells and (iii) Bio-sensors with emerging applications across biomaterial sciences and biomedical technology fields is also among the targets of the project. By introducing new concepts of high added value, Find and Bind addresses societal objectives such as improving patients’ health care and quality of life.

The scientific and technological objectives of Find and Bind are:

a) Understanding:

• How glycans structure modifications influence the interaction (binding, stability, and release) with other biomolecules (e.g. growth factors, other proteins);
• How nanoscale features influence cellular mechanics and behaviour and how this effect varies between cell types;
• The role of cell-matrix interactions in regulating cell behaviour and differentiation in an integrated approach where materials and surface properties play the key role.

b) Exploring the knowledge

• Elaboration of new synthetic routes for carbohydrates with tailored patterns of introduced functional groups controlling the material properties and biological interactions;
• Application of non-conventional nano-modification methodologies to control cell behaviour through nanoscale engineered surfaces;
• Innovative approaches for development of carbohydrate based systems with tailored stability/degradation via extracellular matrix (ECM) components.

c) Mastering:

• Relevant 2D bio-instructive, nano-tailored surfaces, selective to certain proteins and cell phenotypes applicable as support for bone and adult stem cells culturing and differentiation;
• 3D Cell-instructive systems that re-create both in vivo signals and timing and use those systems in bone tissue regeneration;
• Biosensors as tools for biomolecules and cell studies applicable in the biomedical field with focus on bone tissue disorders.

Project Results:

The project is divided into 4 scientific work packages (WPs) and two additional WPs on Management and on Product exploitation and Intellectual property (IP).

WP1: Sweets mastering

This WP deals with synthesis of bioactive GAG derivatives that are further used in the following WPs. Enzymatic and chemical synthesis were used to obtain sufficient quantities of pure, structurally and chemically well-defined carbohydrates. Libraries of carbohydrates with subtle differences in chemistry (e.g. sulfation), molecular weigth, and sequence were produced for systematic studies.

Glycosaminoglycans (GAGs) are not only structural components of the ECM. They represent also biospecific cues and regulate activity of proteins and cellular functions. Thus, several biomedical uses of GAG or GAG-like molecules can be foreseen. Among them sensing tools for binding proteins including growth factors or bioacitve components of scaffolds for tissue engineering are in the focus of Find & Bind project. Immobilisation of glycans on surfaces to modify sensors or their incorporation as components of scaffolds (e.g. hydrogels) requires modification with reactive binding functionalities or cross-linking moieties to obtain glyco-polymer or glyco-SAM conjugates. We have developed several activation/conjugation strategies to obtain those as can be seen below. Some of the obtained libraries are also listed.

Task 1.1. Synthesis and sequencing of oligosaccharides and short polysaccharides

Products with different molecular weights were obtained by using either enzymatic digestion (hyaluronidase and chondoitinase) or chemical degradation methods such as ozonolysis. Fractionation by ultrafiltration using membranes with different cut off was applied when separation of oligomers with different molecular weight was targeted. Further functionalisation of the obtained oligosaccharides by one of the below described methodologies was used to create libraries of glycans with subtle differences in chemistry, sequence, and branching applicable as 2D supports or 3D matrices.

Functionalisation of GAGs to obtain stable glyco-based bulk materials: Methacrylated GAGs; Acrylated GAGs;

Functionalisation of GAGs for surface immobilisation: Thiolation of polysaccharides; Oxidation of polysaccharides; Hydrazido functionalisation of glycosaminoglycans; Biotinylation of glycosaminoglycans (end-on and side-on); Succinated glycans

Task 1.2. Structural characterisation of the obtained saccharides.

A full structural characterisation of the newly produced carbohydrates was carried out using different and complementary characterisation techniques: nuclear magnetic resonance (NMR), infra-red spectroscopy (FTIR), mass spectroscopy (MS), high-performance liquid chromatography (HPLC), size exclusion chromatography (SEC), dynamic light scattering (DLS) were used among the large pool of available tools.

Task 1.3. Conjugation of the synthesised oligosaccharides with self-assembling block-copolymers or aliphatic chains.

Block-copolymers with excellent control of their molecular weight, polydispersity, functionality, architecture and composition were obtained via a novel methodology: oxime click chemistry. Once the synthetic and purification procedures were established, we have studied the self-assembly of these hybrid natural-synthetic diblock copolymers with poly-L-lysine as a model polyelectrolyte. Previous studies have described the the formation of spherical, nanosized interpolyelectrolyte complexes by the interaction of hyaluronic acid-block-poly(ethylene glycol) (HA-b-PEG) with poly-L-lysine (PLL) at stoichiometric charge-to-charge ratio. These nanoassemblies have hydrodynamic radii ranging from 45 to 150 nm with very low dispersity indices (D.I. = 0.02-0.05) but increasing ionic strength to the physiological lvels leaded to disruption of the complexes and crosslinking is needed to obtain stability at physiological ionic strength. Thus, we focused on the influence of sulfated groups in the polysaccharide on the fomration of the IPECs and the stability of IPEC against salt addition. The formation of IPECs from different polyelectrolytes ratios were studied in detail by dynamic light scattering (DLS). The stability of the formed IPECs was studied at different ionic strengths and STEM.

Glycolipid analogues were also developed under this WP. Palmitoyl xanthan with various conjugation ratios was successfully synthesised to obtain an amphiphilic polysaccharide. The amphiphilic character imparted to native xanthan only by a certain palmitoyl ratio allows the self-assembly into stable hollow capsule structures in cell-friendly conditions, i.e. in the presence of physiological ion concentration and pH. Microcapsules with long-term stability and ability to support cell viability and proliferation in vitro over prolonged time were obtained using this self-assembling system. In summary, we have demonstrated the use of self-assembly to provide triggered activation for capsule formation. This approach provides an easy and versatile strategy for the construction of synthetic matrices capable of encapsulating living cells that could be applied in cell delivery therapies.

WP2: Two-dimensional models

The complexity of carbohydrates and our lack of understanding of their regulation lead to questions of how such a diverse and seemingly unsystematic set of biomolecules can encode information. To answer this query we must study not only the structures of glycans but also their molecular interactions. WP2 is focussed on creating suitable models for these studies and validating these in studies with proteins (e.g. growth factors) and cells.

The models were created from both natural and synthetic/modified carbohydrates obtained in WP1. Overall, the approaches presented below guarantee a precise modification of surfaces involving structuring of polysaccharides based on self-organisation processes. All techniques are based on adsorption from solutions and therefore, in contrast to other physical techniques like electron or laser beam irradiation, applicable to devices of virtually any shape and structure. While the first task of this WP is focused on homogeneous coated surfaces, which are very useful for model studies, the second one presents more complex nano-structured surfaces, which are mimicking closer the natural occurring distinct bioactive domains.

Task 2.1: Design of homogenous coated surfaces by electrostatic interactions.

When studying interactions at the molecular, cell and tissue level, the native properties of the immobilised molecule, i.e. its biofunctionality, need to be retained upon immobilisation. The introduction of functional groups allowing immobilisation could have unwanted effects on the studied molecule, but that is not always straightforward to evaluate. Thus, we have tested several approaches for carbohydrates immobilisation onto solid supports. These approaches can be generally classified as: (i) non-covalent immobilisation of unconjugated carbohydrates on unmodified surfaces; (ii) covalent immobilisation of chemically unmodified carbohydrates on modified surfaces and (iii) covalent immobilisation of conjugated carbohydrates on derivatised surfaces. Some of these examples are: Coupling of carbohydrates via hydrazido groups;Coupling of carbohydrates via biotin-streptavidin specific interactions; Amino coupling via naturally occurring carboxyl groups or introduced aldehyde groups: physical adsorption (A); immobilisation via EDC/NHS activation (B); deposition of oxidised HA via reductive amination (C) and end-side immobilisation of HA via reductive amination (D); End-on immobilisation via the reducing end. The efficiency of the immobilisation in all cases was studied by QCM-D. We have demonstrated that biotin-streptavidin specific interactions can be used in creating feasible models. As a part of this study, we have contributed to the development of a new QCM sensor.

Hyaluronan (HA) and chondroitin sulfate (CS), and synthetically sulfated derivatives of the two, were immobilised by biotin-streptavidin binding on this new sensor. The degree of biotinylation and the placement of biotin groups (end-on/side-on) were varied. Our results showed that end-on biotinylated HA was efficiently degraded by hyaluronidase, whereas already a low degree of side-on biotinylation destroyed the degrading ability of the enzyme. Synthetically introduced sulfate groups also had this effect. Moreover, for sulfated derivatives (b-sHA end-on and b-sCS 1%), mass adsorption rather than mass loss was seen upon addition of hyaluronidase in QCM-D. This adsorption was reversible upon addition of a 10 mM solution of NaOH. Thus, the enzyme still binds to the GAG, but it can no longer degrade the chain. For the over-sulfated CS, this was confirmed by HPLC, where sCS and b-sCS were not degraded into smaller fragments, indicating that they were not recognized as CS. Taken together, the results from surface based (QCM-D) and bulk (HPLC) degradation showed great resemblance. This indicates that the functionalisation of the immobilised GAGs (biotinylation or sulfation), rather than the surface immobilisation as such, is most likely the main factor when a lowered degree of degradation is observed and that the action of hyaluronidase efficiently probes the biofunctionality of immobilised GAGs. On the other hand, binding experiments with the proteoglycan aggrecan emphasized the influence of protein size and surface orientation of the GAGs for in-depth studies of GAG behavior. While in studies considering cell surface GAGs the end-on configuration is considered to be optimal (as shown above) since it resembles the structure of a proteoglycans, in ECM this is not always the case. HA is free in the ECM and the aggrecan binding is not seen as end-on immobilised in HA–aggrecan interaction. In this case, side-on immobilisation of HA on the surface gives the highest interaction with aggrecan. Too high functionalisation and sterical hindrance is a likely explanation of the decreased binding seen in the other two cases.

Task 2.2: Design of nano-structured surfaces.

Electrostatic interactions are also in the base of layer-by-layer (LbL) assembly. In this approach a charged surface and polycounterions in solution interact. Because natural GAGs are generally negatively charged they can be subject to this assembly without any further modifications. After adsorption of the first polyion, the surface charge usually converts and the adsorption of a polyion with opposite charge is possible. This procedure can be repeated several times until the desired layer number is achieved. In this way thick but soft layers can be generated on a large variety of material surfaces. The conformation and binding of weak “polyelectrolytes” namely charged polysaccharides is highly dependent on the environmental conditions like pH value, ionic strength and temperature during the assembly. Therefore, the properties of these stable multilayer surfaces can be tailored by the number of layers, type of polysaccharide(s) and complexing conditions (e.g. pH value or ionic strength), which will have a major impact on how proteins adsorb. We have used three polycations under this project: (i) poly-L-lysine since is the most used polycation (ii) chitosan because is the only polusaccharide with a positive charge and (iii) collagene as it is part of ECM together with GAGs. In all cases the build up was followed by QCM-D.

In the case of collagen (Col) deposition, AFM images of Col1/CS and Col1/HN showed a febrile structure of the films. The fibrils in the Col1/CS system increased in diameter from 10.1±1.4 nm for (Col1/CS)2 up to ~19.2±3.25 nm for (Col1/CS)10. Individual fibrils of about 1-2 μm in length were observed after 2 bilayers. The length of the fibrils increased with number of layers deposited and after 10 bilayers fibrils of 4-5 μm in length could be observed. In the Col1/HN system, films had sparse fibrils of 9.2±1.2nm diameter for (Col1/HN)2. A meshwork of fibrils was observed for (Col1/HN)5 and (Col1/HN)10 with no significant change in fibrils` diameter. The addition of a collagen layer clearly increased the thickness of the fibrils in both systems.

In order to investigate cell adhesion to our synthesised substrates and not extraneous proteins, cells were incubated with 1) cycloheximide, to avoid influences of adhesive proteins synthesized by the cells themselves and 2) in serum free media to prevent influences of serum proteins on cell adhesion. Under these conditions cell spreading was superior on ECM films compared to PDMS and tissue culture plastic (TCP), suggesting that normal spreading on TCP is dependent on proteins synthesized by the cells or serum proteins. The average cell area on the (Col1/CS)2 film was 2025±1085 μm2 which was significantly greater than the average cell area on TCP 661±611 μm2 using the Student’s t test (p>0.05; α=0.05). Cells under these conditions do not adhere well to PDMS, however, the ECM film modified PDMS allowed strong cell adhesion, where even application of 10% tensile strain to a PDMS membrane for 24 hrs was possible without noticeable cell detachment. In order to prove that cell adhesion was integrin-mediated, cells were incubated with antibodies to alpha1, alpha3, alpha5, beta1 integrins and IgG as control and then cultured on films of (Col1/CS)2. Beta1 integrin forms dimers with several alpha subunits and controls interaction with several ECM molecules including collagen, fibronectin and vitronectin. Alpha1 binds the monomeric collagen fibril and laminin, alpha3 binds laminin and thrombospondin, and alpha5 binds fibronectin and fibrinogen. Cell adhesion and spreading were hindered only with the beta1 integrin antibodies, while the other integrins had no effect on adhesion or cell spreading. Less cell adhesion was observed when beta1 integrin antibodies were used compared to the control and other antibodies. Furthermore, cell area was significantly lower for cells incubated with beta1 integrin antibodies (364±142 μm2 compared to IgG 1927±733μm2) whereas incubation with other antibodies caused no significant difference in average cell area.

Beside LbL described, we have also applied micro-contact printing to obtain patterns derived from different GAGs. The main challenge in printing GAGs is associated with their wet-ability (swelling and causing lose of shape) or/and solubility in aqueous media (e.g. cell culture media). Thus, we needed to develop several methods to minimise or ideally to eliminate these undeserved effects. Another complication arises from the need to use different supports (e.g. glass, gold, vinyl) for GAGs immobilisation because of the compatibility with different characterisation techniques. Polydimethylsiloxane (PDMS) was used for the fabrication of variously structured stamps. PDMS stamps must be treated with oxygen plasma to make them hydrophilic otherwise the GAG solution would not spread and attach to the PDMS. In the first trial, we have printed fluorescein-labelled, thiolated GAGs (FL-t-GAGs) on hydrophobic vinyl-terminated glass surfaces. The second approach enables to print the vinylsilane (7-Octenyldimethylchlorosilane, ABCR, Germany) onto cleaned glass cover slips, backfilling the remaining glass surface with a PEG-silane (2-[Methoxy(polyethyleneoxy)propyl]trimethoxysilane, ABCR) and to immerse into a GAG solution (2mg/mL in TRIS-HCl, pH 7.4) for further UV-irradiation (365nm, 4h). The printing process can be monitored with an inverted light microscope and the printing result was observed with fluorescence or confocal laser scanning microscopy. The results show that both methods are suitable for obtaining distinctive patterns of GAG. However, their effectiveness is different: with similar microscopy conditions the printed FL-tCS seems to give more intense print but with some aggregation on the outer part of the pillars. The method that enables to print the silane first with a subsequent immobilisation of the FL-tCS shows a better distribution of the labelled GAG.

In another methodology, we have used oxidised GAGs (WP1) that were immobilised on amino-functionalised surfaces (glass or gold with amino-terminated self-assembing monolayers, SAMs). Patterns on these surfaces were obtained by μCP of adhesive (fibronectin, FN) or non-adhesive (albumin, BSA) proteins available in the human plasma. Oxidised hyaluronan (ox-HA), hyaluronan sulfate (ox-HAS) and heparin (ox-Hep) were used in these trials. After the μCP of fluorescent BSA and FN over the immobilised ox-GAGs substrates, the transferred patterns were characterised by fluorescence microscopy and AFM. In general, the binding of proteins was reduced on HA surfaces when compared to glass, but from the fluorescence microscopy images it can be seen that the obtained patterns of both proteins were homogeneous. The weaker fluorescence of the FN pattern could be caused by the lower concentration of fluorescent label. The AFM image shows the topography of the protein-printed and non-printed areas. A 4-5 nm step was measured by AFM in semi-contact mode, which is consistent with the height of a protein layer. The stability of the protein pattern was confirmed by the immersion of printed surfaces in PBS for 24 h and analysed again by fluorescence microscopy. We found that the patterns were perfectly preserved and also proving that the protein adsorption to ox-HA and ox-HAS was strong enough to allow future cell culture studies. We have further confirmed the functionality of these surfaces by culturing adipose stem cells on them. When FN was used on the ox-GAG, the cells adhere and elongate only on the protein pattern. In this case the adhesion between the protein and the underlying GAG is quite strong (SEM). When BSA was used instead, the cells were more elongated compared to the controls but not as much alined along the pattern as in the case of FN pattern. It must be noticed that in the case of BSA, the cells form focal adhesions only with the ox-GAG and they do not adhere on the surface area patterned with proteins. On some of the samples is visible that cells extend their filopodia from one unpatterned area to another, crossing the zones with BSA without adhering on them.

Task 2.3: Carbohydrates cross-talks in bioenvironment.

The surfaces designed in task 2.1 were tested for their ability to bind specifically to growth factors such as FGF-2 and BMP-2 by SPR, ellipsometry or QCM. The short half-life of growth factors, and their tendency to aggregate in solution aggravates the usage, but combining positively charged BMP-2 with negatively charged GAGs, such as CS, has proven to stabilise them. In order to visualise GAG-protein interaction, a study with QCM-D was carried out using the biotin/SAM-platform developed during the first reporting period. BMP-2 was added to the CS immobilised in two different ways. Both conditions induced binding of BMP-2 to CS immobilised via the two strategies. Whether the interaction was specific or not was dependent on the underlying SLB and the ionic strength. Unspecific binding to the lipid bilayer could be seen at low ionic strength when SLB-COOH was used. In high ionic strength buffer, the unspecific interaction to SLB-COOH was diminished. BMP-2 bound to a higher extent to h-CS immobilised using strategy (a) when added in low ionic strength buffer compared to when added in high ionic strength buffer. When SLB-NH2 was used, negligible unspecific interaction could be seen for both the low and the high ionic strength buffers. Here, BMP-2 bound to a higher extent to CS when added in high ionic strength buffer. The CS-BMP-2 layer was more rigid compared to the CS-collagen layer. Since CS is negatively charged, BMP-2 binding was observed to the CS-modified SLBs using either strategy (a) or (b) at low ionic strength, but also to the SLB-COOH. When increasing the ionic strength the non-specific binding to SLB-COOH was suppressed and a decreased but presumably specific binding was observed to CS immobilised to SLB-COOH. The amount of BMP-2 bound to SLB-COOH only, SLB-COOH with h-CS and SLB-NH2 with CS is roughly the same at low ionic strength, even though the coupled amounts of underlying CS differ, indicating a primarily electrostatic interaction. When strategy (b) was used to immobilize CS, higher amounts were coupled and CS is also expected to be in a more native form using this strategy (see above). The large amounts of BMP-2 that bound to CS immobilised using this strategy at high ionic strength confirmed this assumption. The reaction did not reach equilibrium under a reasonable time, likely due to structural rearrangements occurring in the CS layer on a slower time scale and perhaps also BMP-2 aggregation. The layer after addition of BMP-2 was quite compact, indicated by the relatively low dissipation, consistent with the expected crosslinking effect of the homodimer BMP-2 on the underlying CS layer.

Fibronectin (FN) is another protein of interest since it is adhesive protein and it also has a specific binding domain for Heparin (HEP) – the most sulfated natural GAG. Therefore, we have investigated the interactions of FN with GAGs with different degree of sulfation (HA, HAS, CS, heparin) by SPR. These experiments were carried out using gold sensors modified with thiolated GAGs (see previous report for the procedure), which were exposed to increasing concentrations of FN (1-200μg/mL) to obtain an adsorption isotherm. It can be seen that with increasing concentration of FN the angle shift of the SPR is also increasing, i.e. an increased amount of adsorbed protein onto GAG is detected. The highest adsorption of FN was measured on tHep and tCS and the adsorption isotherms indicate that there could be some more FN bound to the corresponding GAGs. On the other hand tHA and tHAS already seem to be saturated with FN after a concentration of 100 μg/mL. For the generation of quantitative data about the binding of FN we used Langmuir linear regression to plot vs. c according to the formula: to obtain the binding constant K and the maximum surface concentration Γmax. The results of the fitting (Table 1.3.3) confirm that tHep binds the highest amount of FN with a Γmax of 0.95 ng/mm2 (95 ng/cm2) followed by tCS with 0.78 ng/mm2. Surprisingly tHAS has the lowest capacity (0.55 ng/mm2) to bind FN. Other specific proteins such as aggrecan were also tested (see above).

WP3: Cells as tools

This WP aims to select specific bioactive carbohydrates that direct cellular mechanisms to the desired pathways and to employ this ability in a development of support(s) for controlled stem cell differentiation.

Osteoblasts, chondrocytes and human mesenchymal stem cells (hMSCs) both from bone marrow and from fat were used under the project. Although bone marrow hMSCs are the most studied adult stem cells, the adipose derived stem cells have been also demonstrating a great potential in bone tissue regeneration. Intially, we have optimised the protocols for the isolation and characterisation of these cells. After each isolation, the cells were characterised by flow cytometry for expression of the markers CD105, CD73, CD90, CD45 and CD34 as well for their osteogenic (alizarin red staining, gene expression of osteocalcin, osteopontin, alkaline phosphatase and osteonectin) and chondrogenic potential (alcian blue staining, sulfated GAGs production).

The effect of homogeneously immobilised glycans on the proliferation and differentiation of mesenchymal stem cells were studied. Immobilised oxidised glycans (WP2) provided different coded information and can affect cell growth and differentiation. MSC seeded on GAG-modified surfaces were confluent after 3 days and differentiation media for osteogenic and chondrogenic induction were loaded twice a week. The osteogenic differentiation medium was prepared by adding recombinant human bone morphogenetic protein (rhBMP-2) in DMEM with 10% FBS. Chondrogenic differentiation medium was prepared by the addition of transforming growth factor-β1 (TGF-β1) in DMEM without FBS. Cells adhere on all surfaces modified with GAGs and display a significant degree of cell spreading. The organisation of actin fibers is an indicator of the strength of cell-substratum attachment and the extent of cell spreading. When cells were grown on CS and ox-CS, the presence of vinculin at the end of actin fibers can be seen clearly without adding of any protein in the medium. The presence of vinculin at the ends of actin fibers suggests integrin involvement in cell adhesion to the GAGs immobilised surfaces. The effect of GAGs surfaces on MSC differentiation was also studied. Osteogenic differentiation can be observed on ox-HAS1 and ox-HEP modified surfaces, while on ox-HA-25 modified surface, the staining is faint. Chondrogenic differentiation of MSC was induced by the addition of transforming growth factor- β1 (TGF-β1) to the proliferation medium (without FBS). While MSC formed aggregates typical for chondrocytes on HA, ox-HAS1, ox-HEP modified surfaces, no aggregates can be seen on CS and ox-CS modified surfaces. In a summary, sulfated ox-GAGs in cooperation with growth factors, like BMP-2 and TGF-ß, induce strong osteogenesis but weak chondrogenesis on MSC on have shown. Moreover, these studies show that immobilised polysaccharides still provide a bioactivity comparable to the native molecules.

Next, we have tested the synergistic effect of patterns and glycans on the cellular attachment and differentiation. In the first experiments we have tested different patterns (distance and depth) and different glycans separately. The second set of experiments was devoted to the synergistic effect of both patterns and glycans on the same surface. In these experiments we have selected patterns and glycans based on the results from the initial tests.Cell cytotoxicity assays (ISO10993) demonstrated that all the tested samples are not toxic. The results from the Alizarin red staining (2 weeks in culture) demonstrated lack of osteoinduction for all of the tested samples. On the other hand, some of the samples induced differentiation into chondrocytes phenotype. The results demonstrated that periodicity has more significant influence on the cell differentiation when compared to the depth of the pattern. From the tested sizes, the intermediate periodicity of 6.65μm showed the highest ability to induce chondrogenic differentiation. Among the tested glycans, we could not identify any that induce osteogenic differentiation. However, some of tested materials supported chondrogenesis. Based on these data, we have selected several combinations of glycan/pattern and further tested the synergetic effect of chemistry/topography on the stem differentiation. Our results have demonstrated that some GAGs, namely tCS, tHA and oxHA 10% induce differentiation of ASCs into the chondrogenic lineage in absence of differentiation medium. This result is supported by Alcian blue staining and PCR analysis. The experiments carried out over fused silica demonstrated that the substrate itself induce ASCs differentiation into chondrogenic lineage. However, some topographies may accelerate the differentiation as was observed for channels with 1000 nm depth for all analysed periodicities. None of the tested substrates demonstrated potential to induce osteogenic differentiation in absence of differentiation medium.

Finally, under this WP we have been also looking for alternative sources and conditions for culturing stem cells. Stromal vascular fraction (SVF) of adipose tissue is a rich source of preadipocytes, mesenchymal stem cells (MSC), endothelial progenitor cell, T cells, B cells, mast cells as well as adipose tissue macrophages. The objective of this study was to optimise the cryopreservation procedure for hSVF in a way to obtain high post-thaw viability of the ADSC (> 70%). Eleven liposuction samples, donated from local clinics with informed consent for R&D were processed using GMP grade reagents. It must be noticed that we have consolidated SVF isolation procedure to obtain yield >0.2 million viable nucleated cells per gram of adipose tissue (AT). Isolated hSVF was cryopreserved within xeno-free and cGMP conditions. Cell viability was assessed by two methods: Flow cytometry analysis and hemocytometer cell counting using Trypan Blue exclusion dye. Preliminary data demonstrated post-thaw viable mononuclear cell recovery of 72.6±1.8 %. Adipose stem cells (ASC), the adherent fraction of these cells, were purified and expanded for characterisation and 3D studies. Both hSVF cells, and correspondent ASCs were expanded up to passage 3 (P3) in distinct media compositions – 10% Fetal Bovine Serum, Low serum (2% FBS) or Xeno-free media, in order to assess proliferation rates, cell morphology and cell tri-lineage differentiation potential. We observed cell morphology in monolayer similar for all FBA-cultures groups, yet smaller and thinner when cultured in xeno-free media. An increased ASC yield with xeno-free expansion media, with population doublings occurring at < 40h was detected. Additionally to proliferation rate of ASC along passaging, differentiation potential into osteogenic, adipogenic and chondrogenic lineages – included as minimum criteria for hMSC, was tested and validated. Alizarin red staining demonstrated mineralisation following osteogenic differentiation, oil red O was used to stain lipid droplets after adipogenic differentiation, while alcian blue staining of glycosaminoglycans was performed to evidence chondrogenic differentiation. The differentiation was followed up to 21 days at passage 1-2 and 2-3. When comparing all groups, an increased ASC differentiation potential was observed in xeno-free expansion media for all tested groups. Reproducibility was further assessed with 2 more donors. Consistently, increased cell proliferation and fibroblast- like morphology was obtained with xeno-free expansion media. Evaluation of reproducibility of ASC immunophenotype was also determined for the different expansion media. Concomitant expression of CD90+/CD105+/CD73+ mesenchymal stem cell markers were obtained for all expansion media tested, yet different staining intensities were verified, evidencing distinct expression levels of these markers.

In a summary we can stat that SVF was reproducibly isolated from adipose tissue lipoaspirates in GMP facilities following xeno-free and cGMP procedures. Adipose stem cells (ASC) were further purified from the SVF and characterised regarding their proliferation capacity, cell morphology in monolayer, immunophenotype - >90% expression of positive markers, < 10% negative markers; and differentiation capacity. A xeno-free media for cell expansion was also identified with optimal outcomes.

WP4: The third dimension

WP4 aims at merging the information obtained under WP2 and WP3 and at translation of the developed techniques under these WPs into mastering of 3D systems able to guide and stimulate cells to form tissues.

Although 2D models are indispensable and give unique information about the effect of polysaccharides nano-shaped surfaces on the structuring of ECM components, they are still a long way from recreating the complexity and dynamism of the natural three-dimensional (3D) environment of cells, their ECM. Nowadays, the major challenge lies in fabricating structures that replicate the native 3D architecture with the appropriate biochemical composition to successfully modulate cellular adhesion, proliferation, and differentiation, as well as to regenerate native tissue functions. The ability to produce controlled, reproducible and indistinguishable microenvironments for a large number of cells is ever more sought after for both fundamental biological studies (e.g. understanding cues for focal adhesion and cellular polarisation) as well as for the future of cell-based assays for drug development and screening. To enable 3D-applications of different polysaccharide derivatives (WP4) we established several methods to introduce cross-linkable groups into the carbohydrate backbone (see the protocols under WP1). As an example, methacrylated carbohydrates (photocrosslinkable) were obtained by a reaction between a polysaccharide and glycidyl methacrylate at 50 °C for 24 hrs. Beside the synthesis of the methacrylated samples, we established a protocol for their cross-linking under mild conditions compatibles with cell culturing. We have assessed the cytotoxicity of different photocrosslinking agents on cells encapsulated in methacrylated alginate. Methacrylated alginate (Alg-Me) was chosen for these studies because it can be crosslinked without initiator (control) by using CaCl2. Passage 3 bovine chondrocytes were encapsulated at 6 million/mL density in Alg-Me gels synthesised as described above (WP1). The alginate/cell constructs were mixed thoroughly with:

(i) 0.05% (w/v) Irgacure 2959; A 30uL volume of the mixture was then cast into cylinders and allowed to crosslink for 5min under a UV;
(ii) 0.05% (w/v) of lithium phenyl-2,4,6-trimethylbenzoylphosphinate (LAP). A 30uL volume of the mixture was then cast into cylinders and allowed to crosslink for 5min under a UV;
(iii) 2% (W/V) VA-086; A 30uL volume of the mixture was then cast into cylinders and allowed to crosslink for 5min under a UV.
(iv) TEA/Eosin as photocrosslinking agents. Hydrogels were prepared with either a high concentration of TEA/Eosin (15 μL of 5M TEA in water, 3 μL of 0.3% (W/V) eosin Y in N-vinyl-pyrrolidone in 250 μL of 2% Alg-Me solution), or a low concentration of TEA/Eosin (15 μL of 5M TEA in water, 3 μL of 0.3% (W/V) eosin Y in N-vinyl-pyrrolidone in 250 μL of 2% Alg-Me solution) and photocrosslinked with an 18W neon lamp for 10 min.
(v) A solution of Alg-Me crosslinked using 102mM CaCl2 was used as control.

In the first three protocols we have used the same UV light source (B-100AP 100 Watt High Intensity Lamp, 365nm wavelength). Following polymerisation, the gels were washed 3 times (5 min each) in DMEM containing 10% FBS, 1% Antibiotic/Antimycotic and 50 μg/mL L-ascorbic acid. The media was changed after 24 h and then once every 3 days. Cells were also encapsulated using the same procedure but omitting any washing steps in order to investigate the effect of washing on the cytotoxic response. Finally, cells were encapsulated in methacrylated hyaluronic acid (HA-Me) hydrogels using the same methods to test the applicability of the above concentrations for HA-Me gelation. The hydrogels were cultured for 1, 4 and 7 days and cell viability was determined with a live/dead assay (Invitrogen) according to the manufacturer’s protocol. Images of live/dead stained cells showed a viability of above 80% of encapsulated cells using all the different methods. The viability was similar in all conditions and comparabble to CaCl2 gelled alginate. However, cells encapsulated using VA-086 exhibited the best results. Gels were stable during the 7 days period of study but the stiffness of the gels appeared to vary according to the used method. The estimated stiffness was from strong to weak as follows LAP > TEA/Eosin > CaCl2 > VA-086 > Irgacure. Hydrogels prepared using Irgacure gelled only in the middle but the sides did not gel. Hydrogels prepared using VA-086 had many bubbles due to the N2 released from the crosslinking agent after activation. Bubbles can be reduced by decreasing the UV exposure to 2 min or decreasing the UV source intensity, bubbles disappeared after 1 or 2 days of culture. To investigate the effect of the photocrosslinking agents on the cell viability, as well as applicability of the above methods for gelling HA-Me, cells were cultured for 24 hours in the hydrogels prepared using the methods above without any washing steps. The viability in gels crosslinked with a high concentration of TEA/Eosin was less than 20%, when a low concentration of TEA/Eosin was used the viability increased to over 70%. Hydrogels crosslinked using LAP had also a similar viability to TEA/Eosin low. Cells encapsulated in gels crosslinked with VA-086 showed the best viability with over 90% viable cells. It is worth to note that the LAP crosslinked gels were the stiffest among all conditions and can maintain a relatively high viability without any bubbles. The viability within LAP hydrogels might be lower due to the fact that it is stiffer than the other gels. An UV exposure for less time might improve the viability of cells encapsulated using LAP photoinitiators.

Under this WP, we have also studied chondrocyte activity in sulfated hydrogels because we hav demonstrated that in 2D models, sulfation degree influcnes significantly cell spreading. Passage 3 chondrocytes were cultured for 7, 14 and 21 days. The morphology of cells encapsulated in the different hydrogels was similar except for cells cultured in 100% sulfated alginate (Alg-S) which had a more spread morphology typical of fibroblastic chondrocytes even after just 7 days of culture. Cells cultured in the 100% Alg-S hydrogels showed a drastic increase in proliferation. Cells encapsulated in 75% Alg-S exhibited minor spreading and protruding filopodia. Cells in 50% and 75% Alg-S also showed an increase in cell proliferation compared to lower Alg-S concentrations tested and DAPI staining revealed multiple nuclei within cell nodules. Proliferation of encapsulated cells was quantified with a BrdU cell proliferation assay: cell encapsulated in alginate sulfate 2% have 436% higher proliferation when compared to 2% alginate and was significantly higher than the other studied conditions. The reference gene (RPL13) used for these experiments had a significantly lower expression in the 100% Alg-S sample. This is an uncommon observation in alginate grown chondrocytes and data from our lab suggest the reference gene RPL13 to be the most stable for chondrogenesis studies among many commonly used reference genes.Literature search suggested that sulfate groups remaining in the cDNA could interfere with the RT-qPCR amplification process. These reports also suggest that addition of BSA would eliminate the effect of sulfate groups in the amplification process. We added 1μL of 10% BSA in a total of 20 μL RT-qPCR reaction volume which brought the expression of RPL13 in the 100% Alg-S sample up to the level of other samples. The Agg and Col2 expression showed a down-regulatoin when Alg-S was present in the encapsulating matrix. IGF was up-regulated with increase in degree of sulfation attaining a maximum of 2-fold with 10% Alg-S. IL1b was down-regulated 2-fold in all sulfated samples however the Ct value was very low which makes the interpretation of these results very difficult.

In order to decouple the effect of gel stiffness from the spreading behavior cells were cultured in 0.4% alginate and in 2% alginate sulfate. Alginate 2% (Alginate sulfate 0%) presented the highest stiffness while the pure alginate sulfate was similar to the soft 0.4% alginate gels. The 0.4% alginate gels did not show any spreading compared to alginate sulfate which indicates that spreading in the alginate sulfate gels is not due to the lower stiffness of the gel. Next, we decided to investigate the molecular mechanism possibly responsible of the spreading morphology and high proliferation in alginate sulfate hydrogels. In particular, we focused our attention on β1 integrins which have a central role in mediating adhesion to the ECM proteins, motility and proliferation of chondrocytes. Furthermore we looked into the activity of RhoA GTPase, an important regulator of chondrocyte proliferation and differentiation. Incubation with the beta1 integrin blocking antibody AIIB2 showed a remarkable inhibition of cell spreading and proliferation in alginate sulfate hydrogels, while no inhibition was observed in the IgG treated controls. This evidence indicates that integrin signaling plays a role in the physical interaction of the cells with the hydrogel and the increase in cell proliferation.Regarding the RhoA activation state, we found that RhoA activity was significantly higher in 2% alginate sulfate compared to pure alginate. Moreover, β1 integrin blocking resulted in a significant inhibition of RhoA activity compared to the non-treated control, while IgG did not have any significant effect. Those findings let us infer that β1 integrin plays an important role in cell protrusion and proliferation within alginate sulfate and could act through the RhoA GTPase pathway.

Finally, acrylated hyaluronan (Ac-HA) and acrylated sulfated hyaluronan (Ac-S-HA) gels were investigated as 3D environment for chondrocyte re-differentiation. 2% (w/v) Me-S-Alg gels presented homogeneous structure and cell distribution. However, we were not able to follow the cellular spreading and proliferation due to the instability of the gels. 2% (w/v) Me-Alg gels were used as control. They were stable in culture up to three weeks. However, the cell presented round morphology during the three weeks of culturing. Sulfated gels were always more swalled (bigger in size) compared to non-sulfated gels. Thus, we tried to stabilise thm by additional cross-linking: incubation of methacrylated alginate sulfate gels in 100 mM CaCl2 for 10 min after UV crosslinking or addition of 3 mM CaCl2 to the culture media did not improve the stability problem of the gels. Gel obtained by mixting 95% (v/v) of 2% (w/v) Me-S-Alg and 5% (v/v) of 2% (w/v) Me-Alg was stable for three weeks in culture. However, the cell morphology was very similar to pure Me-Alg with no cell spreading. Higher concentration resulted in increased stability: 2.5% (w/v) Me-S-Alg gels were stable for up to one week of culture. 4% (w/v) Ac-S-HA (DSAC= 0.7 DSS= 1.5) and 4% (w/v) low molecular weight Ac-HA (DSAC= 0.9) gels were also tested. Ac-S-HA gels got very soft and dissolved in less than a week whereas non-sulfated gels were intact and cultured up to three weeks. Cell morphology in non-sulfated gels was round and no spreading was observed. 5% (w/v) Ac-S-HA, 5% (w/v) Me-S-Alg, 5% (w/v) Ac-S-HA + 2% (w/v) methacrylated dextran, 5% (w/v) Me-S-Alg + 2% (w/v) methacrylated dextran gels were tested for stability in culture and cell morphology. All of the gels were intact and stable up to four weeks in culture. Ac-S-HA gels were the most swollen compared to other gels. Cell morphology was found to be round in the first week for all gels. Cells started to spread in sulfated hyaluronan gels within the second week and by the third week cell morphology was found to be highly spreading and elongated. In the alginate gels and dextran-mixed gels, cells preserved their round morphology during the cultured period. In general, cells looked better in the acrylated hyaluronan samples than the methacrylated polymers.Ac-HA and Ac-S-HA hydrogels were also switable for MSCs encapsulation. Both hydrogels were superior to centrifuged pellets for expression of chondrogenic genes collagen 2 and Sox9. On the other hand, Ac-HA promoted upregulation of hypertrophic marker Collagen 10. Similar result was observed for Ac-S-HA although less pronounced. Moreover, expression of MMP-13 and Runx2 in Ac-S-HA gels was lower than in centrifuged pellets. These results are showing that Ac-S-HA is a promising material for achieving chondrogenesis with prevention of hypertrophic differentiation.

Task 4.1. Coating of 3D polymer based systems with bioactive glycans

The simplest way to translate the 2D surfaces designed and developed under WP2 into 3D structures containing bioactive glycans is by coating of various 3D matrices. Recently, emphasis has been given to non-harmful and versatile techniques to modify polymeric substrates: the self-assembly sequential adsorption of proteins and polysaccharides, known as layer-by-layer (LbL) adsorption, is one of the most promising today. It is a simple and versatile technique which allows producing robust coatings, even in substrates with complex geometries. Because it discards the need of organic and harmful solvents, it is an attractive technique for tissue engineering applications, including for biomimetic composite-like coatings, to control the release of drugs, or to manipulate the adhesion, differentiation, proliferation and even function of attached cells. Under this WP, we developed a new method allowing the application of the LbL (WP2) for building up 3D interconnected, porous structures. The concept is based on the use of packed particles acting as templates, which are leached out after being coated with a multilayer biocompatible film. Perfusion technique was used instead of the common dipping or spraying methodologies for building up the multilayers. The process was performed by drop by drop addition of the polyelectrolyte medium over the top of free standing packing of the charged wax particles.Mechanically stable and interconnected scaffolds were obtained after 10 bilayers. Fine 3D multilayer network was obtained after leaching of the spheres and the freeze drying leads to the formation of desired scaffold. The scaffolds were able to support cell attachment and proliferation even in the inner regions of the structure. Cells attach to the surface of the pores and remind viable after 3 and 7 days of culture.

Task 4.2. Non conventional processing techniques

Under this task we have developed different 3D structures by combining different approaches such as self-assembly and supercritical fluid extraction of nanoparticles encapusaltion in LbL assembly. An example is presented below. We developed of liquified multilayer sacs containing microparticles as cell supports. Due to the high versatility of the LbL technique, different properties of the sacs can be easily tailored, e.g. the permeability of the shell can be controlled by varying the number of layers deposited and the surface of the sacs can be customised by endowing nanoparticles, lipids, viruses, among others. Cells were mixed with an alginate solution (solution A) or with an alginate solution containing poly(L-lactic acid) microparticles (PLLA; solution B). Both solutions were extruded separately, originating alginate particles by ionotropic gelation. Alginate particles obtained with solution A were used as control. Afterwards, an external shell was developed at the surface of the obtained particles by layer-by-layer technique. The cross-linked alginate core was further liquified by ethylenediaminetetraacetic acid (EDTA) treatment, originating sacs encapsulating cells (ALG capsules), or encapsulating cells and PLLA microparticles (PLLA capsules). The capability of sacs to adapt to different structures, such as syringes, was also assessed. Sacs were collected, demonstrating a certain jamming resistance when passing through thinner regions of a syringe. The adaptable shape of the sacs is of huge interest for encapsulation systems since it offers devices with possibility of being implanted by minimal invasive methods in transplantation procedures. The average diameter of the sacs was 1.8 ± 0.08 mm and the diameter of the PLLA microparticles covered the range of 20–100 m. PLLA lacks natural cell recognition sites, and its hydrophobicity and low surface energy are known to affect cell attachment and proliferation. Therefore, prior to biological assays, PLLA microparticles were surface modified to enhance cell adhesion affinity. Cell−material interactions of the encapsulated cells with microparticles were visualised by SEM. At day 1, cells were adhered and spread across the surface of microparticles. At day 7, large aggregates of microparticles and cells could be detected. The encapsulated cells maintained their elongated fibroblast-like morphology.In addition, a fluorescence viability staining was performed to assess living and dead cells dispersed in the sacs environment, using a live-dead assay. Overall, MTS and live-dead data suggest a good viability for all formulations, particularly for sacs with PLLA particles. Control and empty sacs exhibited a higher number of dead cells compared to sacs with PLLA particles. This difference becomes more evident as the time of culture increases, particularly from day 14. This may be due to the fact that L929 cells are anchorage dependent and, thus, the cell adhesion sites provided by PLLA microparticles increased the surface area for cell growth. Consequently, sacs with PLLA particles presented an enhanced metabolic activity and cell viability. Moreover, the good viability of the encapsulated cells suggested by MTS and live-dead data are a indicator that the membrane allows the diffusion of nutrients and exchange of waste products for a long time period. The distribution and structural organization of encapsulated cells within the cross-linked core of control particles or the liquefied environment of capsules was assessed by DAPI-phalloidin fluorescence assay. DAPI-phalloidin showed that, interestingly, cells started to form large aggregates within the sacs with PLLA microparticles constructing their own 3D cell culture assembly system. Cell proliferation was assessed as a function of culture time up to 28 days by DNA quantification assay. Sacs with PLLA particles had the highest cell proliferation rate during the culture time with a continuous increase up to 28 days of culture. On the other hand, at 28 days, cell proliferation in the empty sacs did not increase and a significant decrease in control particles was observed. Besides the lack of adhesion sites, these results may also corroborate the jeopardized diffusion rate offered by the crosslinked core of control particles. In a summary, DAPI-phalloidin and DNA data suggest an improved value provided by the sacs with PLLA particles regarding cell−material interactions and cell proliferation. In a summary, we can state that we were successful in developing sacs encapsulating cells and PLLA microparticles that allow the diffusion of essential molecules for cell survival. The proposed encapsulation system not only alleviates the current bioencapsulation issues but also allows tailoring membrane permselective ability, construct mechanical integrity, as well as optimising and customising a cell friendly environment inside the capsule. The developed approach can potentially be used to encapsulate anchorage dependent cells and control the phenotype of primary cells or stem cells fate. Moreover, different microparticles can be encapsulated which may be loaded with growth factors and other biomolecules of interest, customised to the type of cell encapsulated.

Task 4.3. Assembling 3D glycan structures by electrostatic interactions

Glycosaminoglycans (GAGs) are negatively charged molecules. In order to design freestanding structures using supramolecular chemistry (e.g. electrostatic interactions) at least one positively charged polymer should be considered. Below we describe some examples of the application of this approach to create functional nanoparticles or hydrogels.

Chitosan/chondroitin sulfate

Chitosan/Chondroitin Sulfate nanoparticles (CH/CS NPs) were prepared by polyelectrolyte complexation. The NPs suspension spontaneously occurred when the CH solution was added to the CS solution, under strong magnetic stirring at room temperature. By Atomic Force Microscopy (AFM) analysis, it was possible to observe that we can obtain nanoparticles from both formulations (CH/CS 1/1 and CH/CS 2/1) with a spherical shape and with dimensions inferior to 150 nm. The assays with L929 cells showed that the NPs do not provoke a cytotoxic response. To prepare the protein-loaded CH/CS NPs, BSA (pI 5.3) was dissolved in water and then mixed in the CS solution. Following a 10 min centrifugation at 11,500 g, the supernatant was discarded and the NPs were isolated. The protein concentrations used for this study were such that allowed the preparation of NPs with 15% (w/w) of protein respective to the NP content. The obtained systems were characterized by dynamic light scattering. The NPs’ dimensions were inferior to 200 nm for both CH/CS 1/1 and 2/1 formulations and the encapsulation of BSA only affected the latter, reducing the size of the NPs. The encapsulation efficiency was higher when BSA loading of 20% was used (weight of protein/weight of chitosan used to prepare the NPs). An increase of protein concentration causes a decrease in the NPs yield as it disturbs the formation of the complex between chitosan and chondroitin sulfate, which ultimately contribute to the decrease in the sustainment of protein in the nanoparticulate complex. The BSA release from the NPs was sustained for a period of weeks. Despite the typical burst release during the first hours of the test, the release rate stabilizes during the following days of the release study. This is a promising result if we take in consideration that the ultimate application of this nanoparticulate system is the entrapment of biological macromolecules such as growth factors. These bioactive agents act towards tissue healing in prolonged biological processes and lose their stability quite easily. Therefore, the application of protein controlled release systems, capable of delivering the growth factor in a slow and controlled fashion, is quite attractive for tissue engineering applications.

In a next step, we encapsulated platelet lysates (PL). PL have been increasingly used for TE strategies as they consist of an enriched concentration suspension of GFs isolated from the blood platelets. When activated, these platelets release different types of GFs, such as platelet derived growth factor (PDGF), transforming growth factor (TGF), insulin-like growth factor (IGF), among others. PL-loaded NPs were obtained using exactly the same procedure as the one described for the BSA. The PL cumulative release during the culture period was quantified by microBCA analysis. The kinetics of PL delivery was similar to the one reported above for BSA. During the 7 days time range, we could observe a higher release in the first day of culture, caused by the typical burst release for both formulations (slightly higher for CH/CS 1/1). After this phase, the rate of release stabilizes towards a more controlled profile in both formulations. One of the aims of this study was to evaluate if the PL released from CH/CS NPs would still be active and still enhance cell viability and proliferation. After the first day of culture, there was no difference in viability between all the tested formulations. However, at day 3 and 7, two formulations show clearly higher viability: the PL-loaded NPs (both CH/CS 1/1 and CH/CS 2/1) cultured in presence of FBS. This result is expected since in these cases, PL acts as a protein supplement for cell culture, offering a GF-richer environment for the hASCs. Moreover, it was also possible to observe that the cells cultured with the PL-loaded NPs in absence of FBS could achieve viability levels close to the ones observed for the control (cells only + FBS). This indicates that the PL released from the NPs might act as a replacement for FBS, thus avoiding the contact with proteins from animal sources.The PL-loaded NPs were cultured in vitro in direct contact with hASCs during 7 days. The evaluation of morphology and proliferation was assessed by fluorescent labelling of the cells and further visualization by microscopy. It was difficult to find differences after the first day of culture, however after day 7, the cells in contact with the PL-loaded CH/CS NPs in presence of bovine serum present enhanced proliferation whereas the hASCs cultured with the PL-loaded NPs without serum show similar proliferation levels to the control, thus reinforcing the result obtained by the viability assay. The hASCs morphology doesn’t change with the presence of PL, however, the high levels of confluence turned the cells into a more fibroblast-like shape.

Succinated chitosan/oxidised heparin

We have developed a new class of injectable polysaccharide-based hydrogels that can be used in the cartilage repair and regeneration. The cross-linked hydrogel is formed from two functional polysaccharides, namely succinyl chitosan (s-CHI) and oxidised chondroitin sulfate (ox-CS) without employing any extraneous cross-linking agent. S-Chi displays good water-soluble property at various pH values. The determined substitution degree of S-Chi was 24.2%. Determination of the aldehyde content of ox-CS revealed an oxidation degree of 26.2%. The formation of hydrogel is due to the Schiff base reaction between amino groups of s-Chi and aldehyde groups of ox-CS. As expected, the gelation time decreased with the increase of ox-CS content. On the other hand, the gel storage modulus increased with the increase of cross-linker content, which indicates the formation of denser cross-linking network, i.e. the more rigid hydrogel.There is no significant difference of the metabolic activity of cells cultured in contact with the hydrogel when compared to the control sample. However, the absorbance of all hydrogel samples increased with the increase of time, which indicated that the hydrogel could act as a barrier that decreased the diffusing rate of the XTT testing molecules. Cells cultured on different hydrogels displayed large and flattened morphologies after 7days incubation. As growth continued, cellular colonies gradually expanded in size and the adjacent ones interconnected with each other. Hydrogel degraded after one week of incubation with cells, of which the degraded products seem to be converted to nutrition and take part in the metabolic activity of cells. Overall, hydrogel and its degraded products exhibit a non-cytotoxic nature, which contribute to the cell survival and growth. Quantitative XTT assay and FDA vial staining using 3T3 cell line as a model were also carrid out. Cells were well distributed inside the hydrogel scaffold and possessed spherical morphology at the beginning stage. As growth continued, cells gradually elongate and expand in morphology, which provide the adjacent ones interconnected with each other

Succinated gelatine/oxidised heparin

In a similar manner succinated gelatine (S-GL) was obtained by the introduction of succinyl groups into primary amino groups of gelatin enabling an increased water-solubility at room temperature. We have used a commercial gelatin type A with isoelectric point (IEP) at about 8 to 9. The succinylation resulted is a shifting of the IEP to a value around 4 as determined by zeta potential measurement. Crosslinked s-GL/ox-Hep nanoparticles were prepared by the one-step procedure using water/acetone as solvent. The formation of nanoparticles was influenced by the s-GL concentration and the presence of surfactant Pluronic F68. Non-ion Pluronic F68 was chosen as surfactant due to general preference in protein stabilization and its biocompatibility. We found that flocculation appeared under the preparation of either high concentration of s-GL solution (5% w/v) or lower one (0.5% w/v) without surfactant at both pH=2.5 and pH=7.4. The stable nanoparticles were formed at low concentration of s-GL (0.5% w/v) with different concentrations of Pluronic F68 of 0.1% and 1%, respectively. Therefore, amphiphilic surfactant, like F68, can contribute to form stable nanoparticles in water/acetone system

The size and polydispersity index (PDI) of nanoparticles were measured by DSL. The hydrodynamic radius (z-average) of nanoparticles with 0.1% w/v surfactant F68 was around 200nm at both pH=2.5 and pH=7.4. However, the PDI was narrower at low pH value compared to the one prepared at neutron solution. The nanoparticles size and its PDI increased up to 372.3nm at high concentration of surfactant F68 (1.0% w/v) due to the increase of surfactant molecules incorporated into one nanoparticle. Highly negative charged oxidised heparin possesses multiple aldehydes on its backbone, which can be used as a cross-linker to finally stabilize nanoparticles via forming imine linkages. pH 2.50 was used to find the critical molar ratio of aldehyde groups of ox-Hep to amino groups of s-GL. When the molar ratio is greater than 1:2.3 there was a large amount of aggregation appearing in the solution, which could be formed by the cross-linking reaction took place among several particles. With this critical molar ratio condition, 4 different pH solutions were compared. It was found that there was a huge aggregation with the addition of acetone at pH=5 because that s-GL easily precipitates at which the pH is close to its IEP (pH=4) and s-GL has minimum solubility. We have investigated the size and PDI of s-GL and s-GL/ox-Hep cross-linked nanoparticles at different pH conditions. s-GL at different pH values showed very broad range distribution before cross-linking, whose PDIs are nearly 1. After adding ox-Hep, cross-linked nanoparticles were formed, whose diameter was in the range of 200 to 300 nm depending on pH. At pH 2.50 the average size of nanoparticles is around 196.5nm of narrow distribution of 0.123 which is good for using in the application of drug delivery. With the increase of pH value, the particles size slightly increase and PDI is getting larger compared to the one formed at pH 2.5. And there were still some uncross-linked s-GL molecules in the solution which is shown by the peak with low intensity except for the sample at pH 2.5. The cross-linking degree of s-GL/ox-Hep nanoparticles at pH 2.50 and 7.40 was measured using TNBS test. Both experimental degrees of cross-linking (13.3% and 19.4%) were lower than expected (43.5%). The degree of cross-linking at pH 2.50 (13.3%) is lower than the one at pH 7.40 (19.4%) because their differences in their cross-linking mechanisms. In general, the mechanism of cross-linking was mediated by the reaction between the aldehyde functional groups and free nonprotonated ε-amino groups (-NH2) of lysine or hydroxylysine through a nucleophilic addition-type reaction. Investigation with TEM found that nanoparticles synthesized at pH 2.50 had been heavily shrinking, from 200nm in solution measured by DLS down to 50nm at dry status. However, those synthesized at pH 7.40 were found to be magnifying by 200~300nm, which were composed of clear small particles of about 50nm in size. Nanoparticles are formed by succinyl-gelatin and oxidized heparin, which possess amino groups, carboxyl and sulphate groups on the surface. We assume that at low pH=2.50 all the groups are protonated with weak affinities among –NH3+, -COOH and –SO3H. The positive –NH3+ groups form a layer of net charge on the particle surfaces, which might be dominant and repel each other thus resulting in well distribution of nanoparticles.This also was confirmed by measurements of zeta potential of nanoparticle at pH 2.5 whose surface charges were positive with potential in the range of 6 to 10mV. However, at neutral pH=7.4 -COOH and –SO3H groups are deprotonated with the presence of negative charges, meanwhile –NH2 is in free status and might show relatively stronger interaction between –COO- or –SO32-, such as the formation of hydrogen bonds, which might cause the formation of aggregation.

Finally, under this WP we reveal for first time the potential of xanthan (bacterial extracellular polysaccharide) as cell encapsulation system. Its carboxymethyl and palmitoyl derivates were synthesised and stable 3D structures were formed from them in the presence of physiological ion concentration and pH. The optimised process conditions enabled generating microcapsules with long-term stability and ability of supporting the survival and functionality of encapsulated cells (chondrocytic cell line) over prolonged time. The simplicity of conjugation process at cell friendly conditions makes these functionalised polysaccharides a promising alternative to the conventional cell encapsulation systems applied in cell-delivery therapies.

Product exploitation and intellectual properties

Under Find and Bind project we have developed several procedures, technologies and products that currently are at different grade of Technology Readiness Levels (TRL):

TRL 9: Actual technology system qualified through successful mission operations.
TRL 8: Actual technology system completed and qualified through test and demonstration.
TRL 7: Technology prototype demonstration in an operational environment.
TRL 6: Technology demonstration in a relevant environment.
TRL 5: Technology validation in relevant environment.
TRL 4: Technology validation in laboratory.
TRL 3: Analytical and experimental critical function and/or characteristic proof-of-concept.
TRL 2: Technology concept and/or application formulated.
TRL 1: Basic principles observed.

WP1

EP 12007934.8 Modified alginate hydrogels for tissue engineering and regenerative medicine G Palazzolo, R Mhanna, J Becher, S Möller, M Schnabelrauch, M Zenobi-Wong (TRL 4)

As described above, these materials have been patented for cartilage regeneration. However, they have not been tried in animals/human yet, which we assume is the relevant environment. Based on these assumptions we have attribute TRL4 of these products.

Methacrylated gellan gum (TRL 9)

Under the project upscalling of the methacrylation of the gellan gum (GG-MA) was upscalled and validated. The manufacturing process is implemented and a product based on GG-MA is currently available under the brands mimsys® and irisbiosciences. The product mimsys G is marketed for 3D cell culture applications in the biotech and pharma sectors.

WP2

Biotinylated QCM-D sensor (TRL9)

The biotin sensor has been launched as a product and has gone through several tests to show its performance in a broad variety of applications. We therefore assume that it is at the final TRL9 level.

QCM-D setup for cell experiments (TRL3)

The cell module on the other hand is still in the very early stages of prototype development. Since we have mainly used this prototype in our scientific research and not made the transition fully into R&D, we believe TRL3 represents most appropriate the stage of development of QCM-D cell module.

WP3

Xeno-free conditions for cell culture (TRL7)

Protocols for cell manipulation (isolation and expansion of mesenchymal stem cells) under a xeno-free mindset were developed. However, some of the protocols have not yet been fully implemented under GMP conditions, which impairs its readiness from therapeutic application perspective.

Stromal vascular fraction (TRL9)

A process for isolation and storage of stromal vascular fraction (SVF) was developed. Currently, this process is GMP compliant and is implemented in a manufacturing environment that follows Cell and Tissue EU Guidelines, and EU GMP Guide. An authorization has been granted by Portuguese Health Authorities regarding the compliance of these cells with transplantation requirements. These cells are also qualified as starting material for the manufacturing of ATMPs. Research use only SVF is also made available by Stemmatters for experimental and other pre-clinical research activities. These cells are currently offered under the brand irisbiosciences.

WP4

EP 13002227.6 Microtissues C Millan and M Zenobi-Wong (TRL 5)

The researchers involved in the development of this work intent to begin a startup company based on the patent. This is an in vitro technology, i.e. it is already tried in a relevant environment. Thus, TRL 5 describs best the stage of this innovation.

EP 12006202.1 Neo-cartilage formed via reconstitution of cross-linked, surface-modified tissue fragments: Applications for tissue engineering and regenerative medicine C Millan, D Miranda-Nieves, Y Yang, T Groth, M Zenobi-Wong (TRL 4)

This approach has been tested only in vitro. No further validation in animals/human, i.e. relevant environment, has been performed yet. It therefore falls under TRL4.

PCT/EP2013/001931 Process of Cartilage Repair C Millan, D Miranda-Nieves, Y Yang, T Groth, M Zenobi-Wong (TRL 4)

This is a followup of EP 12006202.1. As mentioned above only in vitro tests have been carried out. No further validation in animals/human has been performed yet.

Potential Impact:

Scientific and technological prospects

One of the main drives of the project is to understand how the natural environment influences cell behaviour and to translate this understanding into industrial applications. For this purpose, Find and Bind is using the already existing knowledge about cell-ECM interaction mechanisms but at the same time is looking at elucidating the role of some sugars in the cell-material and cell-cell communication at the nanoscale level, estimating relevant novel 2D and 3D systems by mimicking physiological chemistry and nano-morphological cues.

Novel cell-instructive materials: The concept of tissue engineering involves implantation of scaffolds/cells constructs into the patients and subsequent resorption of the scaffolds with simultaneous repair and neotissue formation by the host tissue. The leading role of the scaffold in encouraging and directing the cells in their invasion and differentiation is critical for its clinical efficiency. Therefore, Find and Bind is focussing much of the effort into giving clues (including for invasion and differentiation) on what cells should do in these scaffolds to make the successful regeneration and integration.

Materials to support culturing and differentiation of adult stem cells: Stem cells based therapies are probably the most important potential application of human stem cells. Stem cells, instructed to differentiate into specific cell types, offer the possibility of a renewable source of replacement cells to treat different diseases. However, many adult stem-cell populations remain difficult to propagate and expand ex vivo. Nowadays, there are yet no standard culture conditions for supporting these stem cells, i.e. to help their propagation and differentiation. Keeping in mind the further clinical application and therefore the need to eliminate the use of animal components and feeders, Find and Bind developed several tools that were validated to work with adult stem cells. The proposd approaches can be a promising alternative of the inconvenient, expensive and widely used methodology via different growth factors and proteins enriched culture mediums.

Carbohydrates-based biosensors: Access to appropriate diagnostic tools is an essential component in the evaluation and improvement of global health. Diagnostics are crucial for identifying the presence and cause of diseases at both the individual and population levels. However, despite the key role of diagnostics, they tend to receive less attention than research efforts focused on novel therapeutics or preventive strategies. Current diagnostic tools are unable to provide practical low-cost monitoring for treatment efficiency. Carbohydrates have the potential to overcome these problems.Hence the third long-term targeted application of Find and Bind addresses these issues and focuses on better and more accessible diagnostic tools.

Economical impact

The created new knowledge under Find and Bind represent new manufacturing concepts for high added-value third generation polysaccharide based devices with different application across the biomedical field. The close industry/research interactions stimulated the industrial take-up of the developed novel solutions. As a result, new end-use opportunities for the carbohydrates with improved performance for different biomedical applications (scaffolds, sensors, etc.) were created with a value that exceeds the total cost of their preparation, thus generating a profit margin. All these inputs together with the constantly increased internal know-how exchange and personnel improved qualification enabled enhanced SMEs' competitiveness across EU and beyond.

Societal objectives

This project intends to create a basis for tissue engineering strategies and therefore it aims to improve the quality of life of the patients suffering from musculoskeletal disorders, the most common cause of severe long-term pain and physical disability affecting hundreds of millions of people around the world. Trends indicate that these disorders are likely to become even more important over the coming years with a progressively aging population. These disorders significantly affect the psychosocial status of injured people, giving rise to enormous healthcare expenditures. Nowadays, the available therapies for these diseases suffer from serious limitations such as lack of donor tissue, rejection and susceptibility to infection. Find and Bind considers two approaches to overpass these limitations: tissue engineering and stem cells based therapies, and thus proposes new technologies to achieve breakthroughs in these fields.

Environmental assessment

Concerning the environmental impact, Find and Bind makes use of natural origin materials. The raw materials used are in many cases extracted from fishing industry by-products that otherwise would have been wasted. The origin of these natural polymers are potentially renewable sources (although the management of those resources are outside the scope of this proposal), and could also decrease the price of currently used synthetic biodegradable medical polymers.

List of Websites:

www.findandbind.eu

Project coordinator:

Prof. Rui Luis Gonçalves dos Reis, University of Minho, tel: +351 253 510 907, fax: +351 253 510 909; e-mail: rgreis@dep.uminho.pt