Martha Gschwandtner

Glycosaminoglycans (GAGs) are key ligands for proteins involved in physiological and pathological processes. Specific GAG-binding patterns are rarely identified, with the heparin pentasaccharide as an Antithrombin-III ligand being the best characterized. Generating glycan-specific antibodies is difficult due to their size, pattern dispersion, and flexibility. Single-domain variable new antigen receptors (VNAR nanobodies) from nurse sharks are highly soluble, stable, and versatile. Their unique properties suggest advantages over conventional antibodies, particularly for challenging biotherapeutic targets. Here we have used VNAR semi-synthetic phage libraries to select high-affinity fondaparinux-binding VNARs that did not show cross-reactivity with other GAG species. Competition ELISA and surface plasmon resonance identified a single fondaparinux-selective VNAR clone. This VNAR exhibited an extraordinarily stable protein fold: the beta-strands are stabilized by a robust hydrophobic network, as revealed by heteronuclear NMR. Docking fondaparinux to the VNAR structure revealed a large contact surface area between the CDR3 loop of the antibody and the glycan. Fusing the VNAR with a human Fc domain resulted in a stable product with a high affinity for fondaparinux (Kd = 9.3 × 10⁻⁸ M) that could efficiently discriminate between fondaparinux and other glycosaminoglycans. This novel glycan-targeting screening technology represents a promising therapeutic strategy for addressing GAG-related diseases.

Tenascin-C (TNC) is a complex glycoprotein of the extracellular matrix (ECM) involved in a plethora of (patho-)physiological processes, such as oncogenesis and inflammation. Since chemokines play an essential role in both disease processes, we have investigated here the binding of TNC to some of the key chemokines, namely CCL2, CCL26, CXCL8, CXCL10, and CXCL12. Thereby, a differential chemokine-TNC binding pattern was observed, with CCL26 exhibiting the highest and CCL2 the lowest affinity for TNC. Heparan sulfate (HS), another member of the ECM, proved to be a similarly high-affinity ligand of TNC, with a Kd value of 730 nM. Chemokines use glycosa-minoglycans such as HS as co-receptors to induce immune cell migration. Therefore, we assumed an influence of TNC on immune cell chemotaxis due to co-localization within the ECM. CCL26- and CCL2-induced mobilization experiments of eosinophils and monocytes, respectively, were thus performed in the presence and the absence of TNC. Pre-incubation of the immune cells with TNC resulted in a 3.5-fold increase of CCL26-induced eosinophil chemotaxis, whereas a 1.3-fold de-crease in chemotaxis was observed when monocytes were pre-incubated with CCL2. As both chemokines have similar HS binding but different TNC binding affinities, we speculate that TNC acts as an attenuator in monocyte and as an amplifier in eosinophil mobilization by impeding CCL2 from binding to HS on the one hand, and by reinforcing CCL26 to bind to HS on the other hand.

Bidirectional dialogue between cellular and non-cellular components of the tumor microenvironment (TME) drives cancer survival. In the extracellular space, combinations of matrix molecules and soluble mediators provide external cues that dictate the behavior of TME resident cells. Often studied in isolation, integrated cues from complex tissue microenvironments likely function more cohesively. Here, we study the interplay between the matrix molecule tenascin-C (TNC) and chemokine CCL2, both elevated in and associated with the progression of breast cancer and playing key roles in myeloid immune responses. We uncover a correlation between TNC/CCL2 tissue levels in HER2+ breast cancer and examine the physical and functional interactions of these molecules in a murine disease model with tunable TNC levels and in in vitro cellular and cell-free models. TNC supported sustained CCL2 synthesis, with chemokine binding to TNC via two distinct domains. TNC dominated the behavior of tumor-resident myeloid cells; CCL2 did not impact macrophage survival/activation whilst TNC facilitated an immune suppressive macrophage phenotype that was not dependent on or altered by CCL2 co-expression. Together, these data map new binding partners within the TME and demonstrate that whilst the matrix exerts transcriptional control over the chemokine, each plays a distinct role in subverting anti-tumoral immunity.

We have engineered GPCR (G protein‐coupled receptor) knock‐out and high GAG‐binding affinity into CXCL12α to inhibit CXCL12α‐induced cell migration. Compared to wtCXCL12, the mutant CXCL12α (Δ8 L29K V39K) exhibited a 5.6‐fold and a 2.2‐fold affinity increase for heparin and heparan sulfate, respectively. From NaCl‐based heparin displacement chromatography we concluded that more amino acid replacements would lead to altered GAG (glycosaminoglycan) ligand specificity. GAG silencing by this mutant was shown in a murine seeding model of human cancer cells, whereby a greatly reduced number of liver metastases was detected when the animals were treated intravenously with 1 mg/kg CXCL12α (Δ8 L29K V39K) before cancer cell application.

Chemokine binding to glycosaminoglycans (GAGs) is recognised to be an important step in inflammation and other pathological disorders like tumor growth and metastasis. Although different ways and strategies to interfere with these interactions are being pursued, no major breakthrough in the development of glycan-targeting drugs has been reported so far. We have engineered CXCL8 towards a dominant-negative form of this chemokine (dnCXCL8) which was shown to be highly active in various inflammatory animal models due to its inability to bind/activate the cognate CXCL8 GPC receptors on neutrophils in combination with its significantly increased GAG-binding affinity [1]. For the development of GAG-targeting chemokine-based biopharmaceuticals, we have established a repertoire of methods which allow the quantification of protein-GAG interactions. Isothermal fluorescence titration (IFT), surface plasmon resonance (SPR), isothermal titration calorimetry (ITC), and a novel ELISA-like competition assay (ELICO) have been used to determine K_d and IC₅₀ values for CXCL8 and dnCXCL8 interacting with heparin and heparan sulfate (HS), the proto-typical members of the GAG family. Although the different methods gave different absolute affinities for the four protein-ligand pairs, the relative increase in GAG-binding affinity of dnCXCL8 compared to the wild type chemokine was found by all methods. In combination, these biophysical methods allow to discriminate between unspecific and specific protein-GAG interactions.