The CD44/integrins interplay and the significance of receptor binding and re-presentation in the uptake of RGD-functionalized hyaluronic acid
Introduction
The efficacy of a carrier structure is often determined by the possibility and efficiency of its targeting. This study aims to assess whether the presence of different ligands may produce a synergic effect in increasing the efficacy of their uptake.
Hyaluronic acid (HA) is a natural non-sulfated glycosoaminoglycan (GAG), which is particularly abundant in the extracellular matrices of skin, vitreous humor of the eye, synovial fluid and brain. HA has found extensive application as a biomaterial in form of a soluble polymer (e.g. for viscosupplementation of synovial fluid [1], [2] or as a bioconjugate [3], [4]), as the basis of 3D scaffolds for tissue engineering and cell-based therapies [5], [6], and also in nanomaterials [7], [8], due to its very low toxicity, absence of immune activation and degradability. All these features make it an ideal building block for designing biocompatible artifacts, and indeed HA has often been seen as a degradable alternative to poly(ethylene glycol) (PEG) for conferring a “stealth” (i.e. “invisible” to the immune system) character to a biomaterial [9], [10].
Differently from PEG, however, HA has a complex and specific biological activity based on the presence of HA receptors, also referred to as hyaladerins, which are responsible for its sequestration, internalization, signaling and metabolism. The most studied HA receptors are listed hereafter; they generally bind to HA through Link domains, which vary in composition and size but in most cases show a structural homology with two α-helices and two triple-stranded anti-parallel β-sheets [11]. HARE (HA Receptor for Endocytosis [12], [13]) is a receptor responsible for the uptake and turnover of about 30% of the total body HA and it is mostly expressed in endothelial cells in the liver but also in the lymphatic system [14]. LYVE1 (Lymphatic Vessel Endothelial HA receptor) is another endothelial cell receptor typical of lymphatic circulation [15] but found also in liver and blood vessels [16], [17]; differently from what was initially thought, LYVE1 appears to be prevalently involved in an inflammation-dependent sequestration and presentation of HA, but not in its endocytosis and degradation [18]. TSG-6 (Tumor necrosis factors – Stimulated Gene-6) is also associated with inflammatory phenomena [19]: although not constitutively expressed in healthy tissues, it is produced in a variety of cell types when stimulated by inflammatory cytokines to cause a sort of HA-dependent, anti-inflammatory negative feedback loop. It is worth mentioning that TSG-6 is a soluble receptor, therefore it is not directly involved in endocytic phenomena and therefore cannot be used alone for targeting purposes; however, TSG-6 cross-links HA [20] and this has a significant effect in its presentation to endocytic receptors such as CD44 [21]. RHAMM (Receptor for Hyaluronan-Mediated Motility) is the first isolated HA receptor [22] and it is prevalently expressed in tumoral cells [23] or in cells activated by injurious events, such as a wound [24]. Despite some similarities in the HA-binding motif [25], the RHAMM binding site does not appear to present β-sheets and is therefore not included in the Link family. Besides not completely clarified intracellular functions, RHAMM activates cell motility in an HA-dependent fashion through a mechanism that likely involves integrins [26], but has also an oncogenic role [27] that is also based on the sustained activation of the most ubiquitous HA receptor, CD44.
CD44 is the most widely expressed HA receptor. It is found on the surface of keratinocytes, fibroblasts, chondrocytes and most lympho-hematopoietic cells. CD44 modulates virtually all aspects of HA-based signaling (from embryogenesis [28] to wound healing [29] and other inflammatory responses [30]), including HA-mediated cell adhesion (e.g. neutrophils in liver sinusoids [31]) and HA endocytosis. In particular, CD44 is the main responsible of HA turnover; for example, in CD44 −/− transgenic mice HA unusually accumulates in the epidermal basal layer and in the corneal epithelium contain, while keratinocytes exhibit impaired motility and elongated phenotype [32]. Not surprisingly, keratinocytes harvested from tissues with pathological accumulation of HA of patients suffering from common dermatosis do not express CD44 [33].
CD44 is a multifunctional and complex receptor. It exhibits affinity to numerous factors, e.g. collagen [34], [35], [36], fibronectin [37], [38], chondroitin sulfate [39], [40], osteopontin [41], etc. In terms of its interactions with HA, CD44 exist in forms that are constitutively binding, constitutively non-binding or switchable; the latter have an inflammation-dependent behavior, which is based on variation in the CD44 glycosylation (sialic residues) that inhibit HA binding [42]; this allows activated T-cells to extravasate through CD44-HA binding [43], or macrophage precursors (CD14-bearing peripheral blood monocytes) to internalize HA only upon activation by inflammatory cytokines, such as TNF-α, IL-1, LPS, or IFN-γ [44], [45], [46], [47].
CD44 is polymorphic. It is encoded by a single gene, which translates to 10 constant exons (first and last 5 ones) that encode for extra-, trans- and intracellular domains, and 10 variable exons whose alternative splicing along with differential N-, and O-glycosylation produce at least 20 isoforms with molar masses from 85 to 230 kDa. The smallest CD44 isoform is also the most common and it is mainly expressed on cells of lympho-hematopoietic lineage; it is devoid of the variable exons and is known as standard CD44s or CD44h. Most mesenchymal cells express CD44E, encoded by three exons of the variable region (CD44V8-10), while the largest CD44 isoform is detected on keratinocytes expressing 8 exons of the variable region (CD44V3-10). The HA affinity varies between isoforms and it is particularly dependent on the CD44 cytoplasmic region [48], [49]: for example, the truncation of the cytoplasmic domain in the CD44st isoform does not allow HA binding and endocytosis [50]. CD44 polymorphism, as well as a variable association to membrane and/or cytoplasmic components may be the reason of a rather elusive mechanism of endocytosis: it has been reported that CD44-HA complexes are not internalized through caveolae or clathrin-coated pits [51], [52], but their endocytosis appears associated to lipid rafts [53], [54] and several protein complexes, e.g. ankyrin [55] but also integrins [54], [56], [57], [58].
Is CD44 a targetable receptor? It is recognized that targeting CD44 would be particularly interesting for cancer- or inflammation-selective delivery of drugs [59]; however, there are major drawbacks: its ubiquitous presence (although in different amounts), its binding to a plethora of possible substrates and its intrinsic variability decrease the selectivity of a targeted action, while its slow turnover (days) [60] can easily produce saturation, therefore allowing only for a relatively low capacity of receptor-mediated uptake of a drug.
Integrins are heterodimeric receptors that mediate cell–cell interactions and cell attachment to the ECM components (fibronectin, laminin, collagen). To date 19 α- and 8 β-subunits are known, and their possible combinations lead to at least 25 different integrins, each of them with different cell expression patterns and ligand specificities [61]. A number of them, in particular those with αv, α8, α5, and αIIb units (αvβ1, αvβ3, αvβ5, αvβ6, αvβ8, α5β1 and αIIbβ3; α5β1 and αvβ3 are the most ubiquitous) show significant affinity to the RGD aminoacidic sequence [62], which has prompted the widespread use of these residues to confer cell adhesion to biomaterials [63], [64]. Due to their common presence, αvβ integrins have a limited selectivity for a pathological condition, but nevertheless they have shown good results for targeting angiogenetic sites, e.g. tumoral vascular endothelia. Most importantly, integrins have a very quick recycling to the cell surface (minutes: see Results and Discussion), therefore they are more difficult to saturate than CD44. Interestingly, both receptors are independently involved in binding and endocytic processes and can be expressed in the same cells, for example in leukocytes (see Fig. 1); since they are often over expressed in the same kind of pathologies (inflammation or cancer), the selectivity of a carrier may be increased by targeting their combination. Here we have evaluated the possibility of a double CD44/αv integrins targeting by soluble carrier structures, i.e. fluorescent HA derivatives of different molecular weights bearing linear RGD peptides as side chains (Scheme 1).
Specifically, we have focused on whether the uptake of the carrier or any of its phases (surface binding, endocytosis, intracellular localization) would be operated in synergy or dominated by one of the two receptors. It is worth mentioning that a synergic behavior is probably desirable in terms of surface binding, because of the better selectivity in targeting two over expressed receptors; however an integrin-dominated endocytosis may be advantageous in that it could speed up the internalization of a carrier and reduce/avoid saturation phenomena.
Section snippets
Chemicals
HA sodium salt with 64, 234, or 1100 * 103 g/mol (from static light scattering measurements) was purchased from Medipol SA (Lausanne, CH). N-(3-dimethylpropyl)-N′-ethylcarbodiimide hydrochloride (EDC) was purchased from Fluka (St-Louis, MO), fluoresceinamine, acetaldehyde, cyclohexyl isocyanide from Sigma (St-Louis, MO); N-(3-Aminopropyl)methacrylamide hydrochloride = from Polysciences (Warrington, PA). The peptide MeOGCGRGDSNH2 was bought from Genscript (Piscataway, NJ).
Cell culture
J744.2 murine
Preparation of HA derivatives
All the reactions described hereafter were performed on HA with three molecular weights (64, 234, 1100 kDa). The RGD sequence was introduced as a side chain of hyaluronic acid by performing a Michael-type addition of the cysteine-containing MeOGCGRGDSNH2 peptide onto HA that was previously functionalized with methacrylamide groups (∼10% of carboxy groups, calculated on the basis of 1H-NMR). An outline of the reaction sequence and the corresponding IR and 1H-NMR spectra for 64 kDa are presented
Conclusions
The introduction of RGD peptides on the HA backbone allows the modulation of the uptake of the polysaccharide through the binding to both CD44 and integrins. The endocytosis of HA-RGD appears to be mechanistically very similar to that of native HA: they seem to share endocytic machinery, intracellular destination (lysosomes), and endosomal acidification-dependent CD44-mediated trafficking: the latter is jammed by the action of Bafilomycin (inhibition of acidification) and shows a saturation
Acknowledgments
The authors would like to thank Prof. Tony Day (University of Manchester) for the advice and the discussions about HA receptors and Ms. Ghislaine Robert-Nicoud and Mr. Christopher Cadman (University of Manchester) for scientific discussions regarding the synthetic aspects of this study. SO would like to thank gratefully BBSRC for a studentship.
References (83)
- et al.
Semisynthetic resorbable materials from hyaluronan esterification
Biomaterials
(1998) - et al.
Photopolymerized hyaluronic acid-based hydrogels and interpenetrating networks
Biomaterials
(2003) - et al.
Target-specific intracellular delivery of siRNA using degradable hyaluronic acid nanogels
J Control Release
(2007) - et al.
Hyaluronan: a multifunctional, megadalton, stealth molecule
Curr Opin Cell Biol
(2000) - et al.
Hyaluronan-binding proteins: tying up the giant
J Biol Chem
(2002) - et al.
Expression, processing, and glycosaminoglycan binding activity of the recombinant human 315 kDa hyaluronic acid receptor for endocytosis (HARE)
J Biol Chem
(2007) - et al.
Endocytic function, glycosaminoglycan specificity, and antibody sensitivity of the recombinant human 190 kDa hyaluronan receptor for endocytosis (HARE)
J Biol Chem
(2004) - et al.
Mouse LYVE-1 is an endocytic receptor for hyaluronan in lymphatic endothelium
J Biol Chem
(2001) - et al.
The inflammation-associated protein TSG-6 cross-links hyaluronan via hyaluronan-induced TSG-6 oligomers
J Biol Chem
(2011) - et al.
TSG-6 modulates the interaction between hyaluronan and cell surface cd44
J Biol Chem
(2004)