Deciphering the biophysical basis by which glycosaminoglycans control growth factor signaling during development: a biomimetic approach – GlyCON

To characterize the biomimetic platform functionalisation we adopt surface sensitive techniques such as quartz crystal microbalance with dissipation monitoring (QCM-D), spectroscopic ellipsometry (SE), surface plasmon resonance (SPR) and image correlation spectroscopy
Platforms are functionalised or step-by-step by the user or automatically by a liquid handling robot.
To analyse cellular responses we adopted immunofluorescence, western blot, qPCR, confocal microscopy, histological staining and in situ hybridization
To create a library of HS-derived oligosaccharides with a controlled sulfation pattern we use size-exclusion chromatography and strong anion exchange HPLC. The resulting oligosaccharides are biotinylated and their binding to BMP2 tested with surface sensitive techniques.

Biomimetic platforms presenting oriented HS (iHS) with adsorbed BMP2 (aBMP2) or aIhh were developed. These platforms were characterised with QCM-D and with spectroscopic ellipsometry to quantify the amount of each component.
To design platforms for cellular studies we co-functionalised the streptavidin layer with cRGD adhesion peptide and iHS with aBMP2 and Ihh. To facilitate the test of multiple conditions in parallel we created an automated pipetting protocol which is executed by a liquid handling robot inside multi-well plates. A software with a graphical user interface was further developed to pilot the liquid handling robot and to enter specific experimental parameters. For that we re-programmed in VisualBasic language an existing macro developed in the group (Machillot et al 2018) to address each single well of the 96 wellplate with different conditions. The homogeneous distribution of molecules has been assessed by confocal microscopy and the molecular surface density analysed with image correlation spectroscopy. A paper discussing these technical aspects is in preparation.
HS has been digested and HS fragments were first resolved by size-exclusion chromatography, leading to size-uniform oligosaccharide fractions. The dp4 and dp6 fractions were then further fractionated according to charge using strong anion exchange HPLC obtaining different fractions with distinct sulfation pattern.
In tight collaboration with UDE we have looked at the effect of the cell-surface VS extracellular HS by plating cells deficient for Ext1 (KO), the gycosyltransferase required for the elongation of the HS chain, on the biomimetic platforms. We observed that the cellular shape was significantly different on the cRGD platforms between WT and KO CHO cells. In particular, while WT cells rapidly elongated and polarized, KO cells remained round and developed filopodia-like extrusion, indicating that cell-surface HS is required for proper cell adhesion and subsequent polarisation
.

Several papers have been published in the frame of GlyCON project. We will soon publish a paper on the automation of the biomimetic platforms on glass bottom 96 wellplates.
During the next 12 months, we plan to verify the effect of cell-surface versus extracellular HS on BMP2 signaling and cellular adhesion in different cell types such as Mouse Embryonic Fibroblasts and primary chondrocytes. The expression of Ihh and BMP2 carrying mutations in the HS binding site will be performed at UDE and the binding of these mutant proteins s to HS will be performed at CEA. To understand the role of HS sulfation on BMP2 signaling we will finalise the HS tetrasaccharide library with defined sulfations and perform molecular and cellular studies. Platforms presenting CS will be tested on BMP-SMAD and Ihh signalling at different time points. More towards the end of the project we will develop platforms presenting mixed CS and HS to verify the hypothesis of the compensation of CS to the lack of HS. For that we will perform QCM-D and SE measurement to quantify the amount of BMP2 and Ihh on the different mixtures of GAGs. Finally, we will isolate GAG from E15.5 mouse skeletons of wild-type and HS defective mouse mutants. GAG-specific (Heparinases/Chondroitinase ABC) enzymatic digestion will allow to determine the relative amount of either GAG. The sulfation structure of HS and CS will be investigated at IBS. To receive first insight into the impact of the physiologically relevant GAG mixture on Ihh and BMP2 affinity, purified complete GAGs, and isolated HS and CS of wild-type and HS mutant embryos will be biotinylated and immobilized to SAv platforms. The interaction between GF isolated from eukaryotic cell cultures and the immobilized GAGs will be analyzed by ELISA assays using BMP2 and Ihh specific antibodies.

1.Sefkow-Werner J, Machillot P., Sales A., Castro-Ramirez E., Degardin M., Boturyn D., Cavalcanti-Adam A., Albiges-Rizo C., Picart C. and Migliorini E. Heparan sulfate co-immobilized with cRGD ligands and BMP2 on biomimetic platforms promotes BMP2-mediate d osteogenic differentiation. Acta Biomaterialia, 114:90-103, 2020.
2. Migliorini,E., Guevara-Garcia, A., Albiges-Rizo, C., and C. Picart. Learning from BMPs and their biophysical extracellular matrix microenvironment for biomaterial design. Bone 141 2020. 115540.
3.Khodr, V. Machillot, P., Reiser, JB, Migliorini, E. and C. Picart. High throughput measurements of bone morphogenetic protein (BMP)/BMP receptors interactions using bio-layer interferometry, Biointerphases 16, 031001 (2021).
4.Sales Ramos, A., Khodr, V., Machillot, P., Fourel, L., Migliorini, E., Albiges-Rizo, C, and Picart, C. Presentation of bone morphogenetic proteins to cells at their basal side reveals their specificity in the initiation of cell adhesion and bone differentiation. Submitted 1er mars 2021.
5.Sefkow-Werner J., Wang I, Picart C, Migliorini E, Delon A. Photobleaching and correlation spectroscopy for in situ quantification of multi-labelled molecules on surfaces. May 2021, European Conferences on Biomedical Optics (conference proceeding)
6. El Masri R, Seffouh A, Roelants C, Seffouh I, Gout E...Vivès R. Extracellular endosulfatase Sulf-2 harbours a chondroitin/dermatan sulfate chain that modulates its enzyme activity. Cell reports, in revision.

During recent years, increasing evidence has been accumulated that the bioavailability of growth factors (GFs) in the tissue is determined by the extracellular matrix, in particular by the heparan sulfate (HS) chains of proteoglycans, which interact with distinct basic domains of the GFs. Nevertheless, it is largely unknown how the HS structure (the sequence of N- and O- sulfation) and the binding sites of the GFs (the sequence of basic amino acids) affect the interaction at the molecular level. Moreover, chondroitin sulfate (CS), a glycosaminoglycan (GAG) with structural similarities to that of HS also binds to many GFs, although with reduced affinity. Preliminary evidence gained by the consortium indicates that CS levels are increased in chondrocytes of mouse mutants with a defective HS structure pointing to a compensatory function of CS at least in chondrocytes. How the composition of GAGs in a tissue affects GF activity has however not been systematically investigated.
With GlyCON, we have established a multidisciplinary project bridging material science and developmental biology. Three international teams interact in a tight collaboration to gain basic insight into the mechanism by which HS and CS control the activity, presentation and distribution GFs. Based on the scientific expertise of the partners, we will use the process of embryonic endochondral ossification as a model to investigate the molecular mechanisms that determine the interaction of HS and CS with bone morphogenetic proteins (BMP) and Indian hedgehog (Ihh). In our multi-disciplinary approach, we will decipher basic principles of how the GAG structure controls the bioactivity of these GFs and how alterations in the GAG composition affect developmental processes, tissue homeostasis and HS related diseases, like multiple osteochondromas (MO) in vivo.
Specifically, we will create a library of HS oligosaccharides exhibiting distinct sulfation patterns and immobilize these with the same orientation as in the in vivo proteoglycans on biomimetic platforms. These platforms will be developed for biophysical studies of the binding affinity and stoichiometry of the HS-BMP and HS-Ihh interactions and for subsequent biological investigation of how the HS structure determines the bioactivity of the GF towards co-cultured cells. The impact of changes in the HS structure on tissue homeostasis will be investigated by the in vivo analyses of mouse mutants carrying specific defects in HS synthesizing enzymes. In the second part, we will analyze the biophysical interaction of GF with CS and determine how the relative composition of both GAGs (HS and CS) affects the bioactivity of GFs in vitro using cells cultured on biomimetic platforms and in vivo in primary cells of mouse mutants.
Our final aim is to generate detailed knowledge on the molecular basis of the GAG–GF interaction, thereby paving the way for future studies of other GF, tissues and rare diseases associated with the biosynthesis of GAGs, and ultimately for new therapeutic approaches.