Skip to main content
NCBI home page
UKPMC Funders Author Manuscripts logoLink to UKPMC Funders Author Manuscripts
. Author manuscript; available in PMC: 2012 Aug 28.
Published in final edited form as: Glycobiology. 2009 Sep 3;19(12):1537–1546. doi: 10.1093/glycob/cwp132

Development of a microtiter plate-based glycosaminoglycan array for the investigation of glycosaminoglycan-protein interactions

Andrew Marson 1,#, David E Robinson 2,3,#, Patrick N Brookes 4, Barbara Mulloy 5, Michelle Wiles 4, Simon J Clark 1, Helen L Fielder 6, Lisa J Collison 1, Stuart A Cain 1, Catherine M Kielty 1, Sally McArthur 3, David J Buttle 2, Robert D Short 3,§, Jason D Whittle 4, Anthony J Day 1,*
PMCID: PMC3428902  EMSID: UKMS49427  PMID: 19729381

Abstract

The interactions of glycosaminoglycans (GAGs) with proteins underlie a wide range of important biological processes. However, the study of such binding reactions has been hampered by the lack of a simple frontline analysis technique. Previously, we have reported that cold plasma polymerization can be used to coat microtiter plate surfaces with allyl amine to which GAGs (e.g. heparin) can be non-covalently immobilized retaining their ability to interact with proteins. Here we have assessed the capabilities of surface coats derived from different ratios of allyl amine and octadiene (100:0 to 0:100) to support the binding of diverse GAGs (e.g. chondroitin-4-sulfate, dermatan sulfate, heparin preparations and hyaluronan) in a functionally active state. The Link module from TSG-6 was used as a probe to determine the level of functional binding because of its broad (and unique) specificity for both sulfated and non-sulfated GAGs. All of the GAGs tested could bind this domain following their immobilization, although there were clear differences in their protein-binding activities depending on the surface chemistry to which they were adsorbed. On the basis of these experiments 100% allyl amine was chosen for the generation of a microtiter plate-based “sugar array”; X-ray photoelectron spectroscopy revealed that similar relative amounts of chondroitin-4-sulfate, dermatan sulfate and heparin (including two selectively de-sulfated derivatives) were immobilized onto this surface. Analysis of four unrelated proteins (i.e. TSG-6, complement factor H, fibrillin-1 and versican) illustrated the utility of this array to determine the GAG-binding profile and specificity for a particular target protein.

Keywords: glycosaminoglycans, sugar array, glycosaminoglycan-protein interactions, microtiter plate-based assay

Introduction

Glycosaminoglycans (GAGs), unbranched polysaccharides found ubiquitously on the cell surface and in the extracellular matrix, are crucial for numerous biological processes including blood clotting (de Agostini 2006; Gray et al. 2008;), tissue organization (Evanko et al. 2007; Cain et al. 2008; Galtry et al. 2008), the immune response (Kuschert et al. 1999; Mahoney et al. 2005; Clark et al. 2006; Taylor & Gallo 2006; Banerji et al. 2007; Prosser et al. 2007), reproduction (de Agostini 2006; Scarchilli et al. 2007) and development (Camenisch et al. 2000; Lin 2004; Bao et al. 2005; Johnson et al. 2007) . Such processes occur as a direct consequence of the interactions of GAGs with a wide variety of proteins, e.g. proteolytic enzymes, growth factors and chemokines. The ability to identify and characterize these binding reactions is therefore of key importance in biological research. To date, such studies have been hampered by the lack of a simple method for the analysis and preliminary characterization of protein-GAG interactions (Mahoney et al. 2004; Wisniewski et al. 2005).

The lack of adherence of GAGs, such as heparin, to plastic surfaces (e.g. microtiter plates), restricts the use of ELISA-like assays in the investigation of protein-GAG interactions (Mahoney et al. 2004; Robinson et al. 2008). One previous method adopted to address this issue has been the covalent coupling of GAGs (e.g. heparin, heparan sulfate and chondroitin sulfate) to bovine serum albumin (BSA) in order to facilitate their surface immobilization (Najjam et al. 1997; Hasan et al. 1999; Nishioka et al. 2007). This requires the bespoke production of GAG-BSA complexes or the use of commercial preparations (e.g. heparin-BSA) where there is no control over the composition of the GAG attached. Another approach has been to determine the binding of biotinylated GAGs to immobilized proteins (Parkar & Day 1997; Nadesalingham et al. 2003; Mahoney et al. 2005; Harris et al. 2008). However, the methods of biotinylation that are commonly used, i.e. the attachment of biotin molecules either to the GAG carboxylate (Yang et al. 1995; Parkar & Day 1997; Boucas et al. 2008) or amino (Nadesaligham et al. 2003; Mahoney et al. 2005; Harris et al. 2008) groups, have different levels of efficiency and can alter the functional properties of the GAG (Yang et al. 1995; Saito et al. 2005; Boucas et al. 2008). For example, the protein-binding capacity of dermatan sulfate has been observed to be higher when biotin was attached to the free amino groups (which are rare in sulfated GAGs) as compared to the uronic acid (Saito et al. 2005). However, other studies have shown that labeling of carboxylate moieties works efficiently for hyaluronan while giving poor results for heparin (Yang et al. 1995). These studies highlight the fact that the optimal method of biotinylation may be GAG specific such that careful optimization of labeling conditions is necessary for each individual GAG-protein interaction. Additionally, immobilization of proteins on microtiter plates can lead to loss of function through either steric hindrance or v ia conformational perturbation (Marson et al. 2005).

We have shown previously that cold plasma polymerization can be used to coat microtiter plates with allyl amine, generating a surface that allows the non-covalent immobilization of the GAGs heparin, heparan sulfate and dermatan sulfate in a functionally active (i.e. protein binding) state (Mahoney et al. 2004; Mahoney et al. 2005; Clark et al. 2006); this method negates the need for covalent labeling/coupling of the GAG. More recently we have demonstrated that related surfaces (made from different ratios of allyl amine and octadiene monomers) can be used to fabricate a concentration gradient of functional heparin (Robinson et al. 2008). Importantly, a surface chemistry was identified upon which an optimal amount of functional low molecular weight (LMW) heparin could be immobilized so as to give rise to maximum protein binding; this “sweet spot” did not correlate with the conditions where the maximum amount of heparin was bound to the surface.

Here we have used this observation as the basis for a comprehensive study of the protein-binding properties of a diverse group of GAGs (i.e. LMW and unfractionated heparin (UFH), selectively desulphated heparins, chondroitin-4-sulfate, dermatan sulfate and hyaluronan) following their non-covalent immobilization upon a range of surface chemistries (i.e. 100:0 to 0:100% allyl amine:octadiene). This has allowed the development of a simple microtiter plate-based GAG array that has significant utility as a frontline technique in the determination of the GAG-binding profile and specificity of a target protein; e.g. establishing the ability of the protein to interact with sulfated GAGs and the non-sulfated polysaccharide hyaluronan. In this regard, we have used the array to determine the GAG-binding properties of four unrelated proteins to determine and illustrate the utility of this methodology.

Results

Immobilization of functional GAGs on a range of allyl amine:octadiene surfaces

Initially assays were performed to study the protein-binding properties of UFH immobilized (in PBS) on the different surface chemistries (100:0, 90:10, 80:20, 70:30, 60:40, 40:60, 20:80, 0:100% allyl amine:octadiene) at a range of concentrations (see Materials and methods). The biological activity of the surface-bound GAG was determined using the Link module domain from human TSG-6 (termed Link_TSG6), a known heparin-binding protein (Mahoney et al. 2005). This domain, which also interacts with the GAGs heparan sulfate, chondroitin-4-sulfate, dermatan sulfate and hyaluronan (Parkar & Day 1997; Mahoney et al. 2004; Mahoney et al. 2005; Milner et al. 2006), was mono-biotinylated (denoted as bA-Link_TSG6) allowing its detection as described previously (Parkar & Day 1997; Mahoney et al. 2004; Mahoney et al. 2005; Robinson et al. 2008). As shown in Figure 1, the amount of functional heparin present on the microtiter plate surface (i.e. as determined from the level of bA-Link_TSG6 binding) was highly dependent on the nature of the surface chemistry. In all cases there was an increase in protein binding with increasing concentration of immobilized GAG (i.e. the protein binding was dose-dependent), but the absolute amount of functional heparin detected differed greatly. For example, there was only a low level of protein binding when the heparin was coated onto 0% and 20% allyl amine (i.e. with 100 and 80% octadiene, respectively), whereas, close to maximal functional heparin was seen on the surfaces coated with 60-90% allyl amine. In general, there was a higher level of protein binding to UFH adsorbed onto the higher ratios of allyl amine:octadiene, with the exception of the 100% allyl amine condition that has an intermediate protein-binding capacity. This relationship can be seen clearly in Figure 2a, which compares the binding of the bA Link_TSG6 protein to UFH and LMW heparin immobilized (at 5-μg/well) on the different surface chemistries. From this it is apparent that these two preparations of heparin behave very similarly.

Figure 1.

Figure 1

Open in a new tab

The protein-binding activities of UFH immobilized on different surface chemistries. Microtiter plates were coated with plasma polymers containing different percentages of allyl amine: 100% (Inline graphic), 90% (▲), 80% (●), 70% (Inline graphic),60% (□), 40% (■), 20% (Inline graphic), corresponding to nitrogen/carbon ratios of 0.22, 0.15, 0.13, 0.10, 0.07, 0.04, 0.02, respectively; the 0% allyl amine (○), i.e. 100% octadiene, does not contain any nitrogen. UFH was incubated with these surfaces at a range of concentrations and the binding capabilities of the immobilized heparin determined using bA-Link_TSG6, an established heparin-binding protein. All values are plotted as mean absorbance (A405nm) ± SEM (n=8), corrected against blank wells (i.e. with no heparin).

Figure 2.

Figure 2

Open in a new tab

The protein-binding activities of various GAGs immobilized on different surface chemistries. Microtiter plates were coated with plasma polymers of allyl amine as described in Fig. 1. GAGs ( a: UFH (■) and LMW (○) heparin; b : chondroitin-4-sulfate (Δ) dermatan sulfate (Inline graphic) and hyaluronan (Inline graphic)) were incubated (5 μg/well) with the coated wells and the binding capabilities of immobilized GAG determined using bA-Link_TSG6; results for LMW heparin have been reported previously (Robinson et al. 2008). For easier comparison the data were normalized against the value for maximum protein binding seen for a particular GAG (i.e. on the optimal surface chemistry). These data are plotted as % maximum protein binding ± SEM (n=8), corrected against blank wells (i.e. with no GAG). The absolute maximum A405nm values obtained for UFH (on 70% allyl amine), LMW heparin (90%), chondroitin-4-sulfate (100%), dermatan sulfate (90%) and hyaluronan (100%) were 1.19, 1.05, 0.82, 1.55, 0.54, respectively.

Given that the Link_TSG6 domain recognises a wide range of GAGs (see above) we used this protein to assess the functional activities of chondroitin-4-sulfate, dermatan sulfate and hyaluronan immobilized on the various plasma-polymerized surfaces (Fig. 2). Whereas UFH and LMW heparin displayed little or no protein binding when immobilized on 0% and 20% allyl amine (previously shown for LMW heparin to be due to the low level of GAG that becomes bound to these surfaces (Robinson et al. 2008)), chondroitin-4-sulfate, dermatan sulfate and hyaluronan all interacted with the bA-Link_TSG6 protein under these conditions. In the case of the 0% allyl amine (100% octadiene) surface, this was shown to be largely due to the interaction of the GAGs with the octadiene polymer, since a lower level of protein binding was seen when they were incubated with identical plates that had not been plasma polymerized (Figure 3). Dermatan sulfate displayed a medium to high level of protein binding on all of the surface chemistries tested (Fig. 2b); the maximum binding seen for this GAG (i.e. on 90% allyl amine) was ~1.3-times higher than that for UFH. In contrast, chondroitin-4-sulfate and hyaluronan displayed reasonable levels of functional activity when immobilized on either low or high percentages of allyl amine with optimal protein binding associated with the 100% surface; the maximal binding of bA-Link_TSG6 to these GAGs was 0.69-and 0.46-fold lower, respectively, than that for UFH. As can be seen from Figure 2b, chondroitin-4-sulfate and hyaluronan had relatively poor protein-binding capabilities when they were immobilized on 70 or 80% allyl amine, indicating that the mechanisms underlying the adsorption of these GAGs may be different on low and high percentages of allyl amine. Overall, these studies demonstrate that all of the GAGs analyzed can be immobilized on plasma-polymerized surfaces and retain their ability to interact with protein. In this regard, the ability of the various surfaces to support functional GAG binding is clearly GAG specific.

Figure 3.

Figure 3

Open in a new tab

Functional GAG binding to untreated surfaces compared to those coated with 100% octadiene. Microtitre plates were plasma polymerized with 100% octadiene or left untreated. These surfaces were incubated (at 5 μg/well) with various GAGs (unfractionated heparin (UFH), dermatan sulfate (DS), chondroitin-4-sulfate (C4S) and hyaluronan (HA)) and their protein-binding capabilities determined using bA-Link_TSG6. Data for the coated (black bars) and uncoated (white bars) surfaces are plotted as mean absorbance at A405nm ± SEM (n=8), corrected against blank wells (i.e. with no GAG).

Design and set up of a microtiter plate-based sugar array

The above studies identified two surface chemistries, namely 100% and 90% allyl amine that are able to support high levels of bA-Link_TSG6 binding to all of the GAGs tested (i.e. heparin, chondroitin-4-sulfate, dermatan sulfate and hyaluronan). This indicates that there is an efficient immobilization of these GAGs in a functional form and, thus, provides the basis for the design of a microtiter plate-based sugar array that can be used to profile the specificity of GAG-protein interactions. Microtiter plates plasma polymerized with 100% allyl amine were chosen for the development of the sugar array in combination with the GAGs investigated above. Our previous binding studies with complement factor H, demonstrated that selectively desulfated heparin preparations, which have lower levels of sulfation and can be used as models of the GAG heparan sulfate, could also be immobilized on the 100% allyl amine surface in a functionally active form (Clark et al. 2006); the factor H Y384 variant (in the context of the CCP6-8 heparin-binding domain) bound reasonably well to all the desulfated heparins tested (i.e. 2-O-desulfated (2-O-deS), 6-O-desulfated (6-O-deS), 2,6-O-desulfated (2,6-O-deS), N-desulfated (N-deS) and N-desulfated/re-N-acetylated (N-deS/R)) revealing that these GAG preparations were being immobilized on the plate surface.

Additional experiments with these same heparin preparations included here indicated that Link_TSG6 binding was more sensitive to changes in sulfation (Figure 4). Furthermore, selective removal of different types of sulfate group caused differential effects on the interaction between Link_TSG6 and heparin. For example, N-deS/R (i.e. the removal of the N-sulfate and its replacement with an acetate moiety) retained the highest level of Link_TSG6 binding, whereas, there was a much lower binding to 2-O-deS (compared to the “parental” UFH from which these were made). Therefore, these two representative desulfated heparin preparations (N-deS/R and 2-O-deS) were also included in the sugar array, i.e. as defined models of heparan sulfate. From Figure 4 it can also be seen that the combined removal of all 2-O-and 6-O-sulfates effectively abolished Link_TSG6 binding. Overall, these data suggest that the 2-O and 6-O-sulfates have a more important role in mediating the binding of heparin to the TSG-6 Link module than the N-sulfate groups. As such it is apparent that this type of assay can provide useful information on the molecular basis of protein-heparin/heparan sulfate interactions.

Figure 4.

Figure 4

Open in a new tab

The effect of desulfation on the binding of heparin to TSG-6. Selectively desulfated heparin preparations (2-O-desulfated: 2-O-deS (●); 6-O-desulfated: 6-O-deS (Inline graphic): 2,6-O-desulfated: 2,6-O deS (□); N-desulfated: N-deS (Δ); N-desulfated/re-N-acetylated: N-deS/R (■)) were incubated with 100% allyl amine-coated microtiter plates at a range of concentrations (0–5 μg/well) followed by the analysis of bA-Link_TSG6 binding. All values are plotted as mean absorbance (A405nm) ± SEM (n=8), corrected against blank wells (i.e. with no heparin).

Thus, in total, seven different GAG preparations were used to generate a prototype sugar array (i.e. UFH, LMW heparin, chondroitin-4-sulfate, dermatan sulfate, hyaluronan, 2-O-deS heparin and N-deS/R heparin). In order to provide an indication on whether the binding of a particular protein to an immobilized GAG is dose-dependent, two concentrations of each GAG (0.2 and 2μg/well) were included in the array. These concentrations were chosen since 2 μg/well gave rise to maximum bA-Link_TSG6 binding for UFH (and the other sulfated GAGs tested; data not shown), whilst binding to the lower concentration of GAG was in all cases suboptimal (see Fig. 1).

X-ray photoelectron spectroscopy (XPS) revealed (from the S2p sulfur signal; Fig. 5a,b) that at both these concentrations, once the relative sulfation levels of each of the GAGs were accounted for, similar absolute amounts of the sulfated GAGs were immobilized on the allyl amine microtiter plate surface (Fig. 5b,d). Thus, there does not appear to be any major effect of the level of GAG sulfation on the amount of GAG immobilized on the allyl amine-coated surface. For example, from Fig. 5 (b and d) it can be seen that there is a similar amount of UFH, N-deS/R and DS bound to the plate surface, where these GAGs have widely different levels of sulfation (i.e. approximately 2.5, 1.5 and 1 sulfates, respectively, per disaccharide). In the case of 2-O-deS (1.5 sulfates per disaccharide) there does appear to be somewhat less of this heparin immobilized (at 0.2 μg/well; Fig. 5b) compared to the other GAGs. The most likely explanation for this is that 2-O-deS heparin is of a lower molecular weight (3.7 kDa) than the other GAG preparations used in the XPS study (all ≥ 10.4 kDa; as determined by size exclusion chromatography (Mulloy 2002)); a consequence of its harsh method of preparation ( Jaseja et al. 1989 ). In this regard, we showed previously that immobilization of low molecular weight preparations of heparin (e.g. dp10 oligomers; ~3 kDa) onto allyl amine at low concentrations (such as 0.2 μg/well) supported less protein binding than higher molecular weight heparins (Mahoney et al. 2004). However, as can be seen from Fig. 5d, the amount of 2-O-deS heparin that becomes bound to the allyl amine-coated surface is equivalent to that of the other sulfated GAGs, when these are immobilized from the higher concentration (i.e. at 2 μg/well). The level of immobilized hyaluronan could not be quantified by XPS from the S2p signal because this GAG does not contain sulfate groups. However, C1s spectra (data not shown) indicated the introduction of peak shifts at higher binding energies (i.e., COO−) consistent with the successful immobilization of the GAG. Therefore, for the sulfated GAGs, at least, any variability in the level of binding by a given protein to the panel of GAGs should be considered to largely reflect its differential binding specificities/affinities.

Figure 5.

Figure 5

Open in a new tab

X-ray photoelectron spectroscopy (XPS) analysis of sulfated GAGs immobilized on microtiter plates. a,c) Unfractionated heparin (UFH), C4S, DS, 2-O-deS and N-deS/R were incubated at either 0.2 (a) or 2.0 (c) μg/well with 100% allyl amine-coated microtiter plates and the amount of GAG bound to the surface was determined by XPS on the basis of the S2p sulfur (S) signal arising from sulfate groups; values are shown as mean S (Atomic %) ± SEM (n=3). (b,d ) In order to determine the relative amount of bound GAGs associated with the plate surface the values in (a,c) were normalized relative to the approximate level of sulfation for each GAG, i.e. 2.5 sulfates per disaccharide in UFH, 1.5 in 2-O-deS and N-deS/R selectively desulfated heparins and 1 in chondroitin-4-sulfate and dermatan sulfate. These normalized values (b, d), i.e. 0.2 and 2.0 μg/well, respectively, shown in arbitrary units, indicate that these sulfated GAGs are all immobilized on the 100% allyl amine surface in approximately equal amounts.

The XPS experiments described above also indicated that the relative amounts of the various GAGs that become immobilized on the allyl amine surface are similar at both the 0.2 μg/well and 2.0 μg/well coating concentrations (Fig. 5a,c). To further investigate this, 35S-labelled heparin was used to determine the absolute amount of heparin that becomes bound per well and whether this is dependent on the concentration of the coating solution. As shown in Fig. 6, a similar level of heparin (~100 ng/well) becomes immobilized on the 100% allyl amine plates when incubated with the wells at concentrations between 0.1–2 μg/well (i.e. including the two concentrations of GAG chosen for the array); a higher level of heparin binding (~200 ng/well) was only observed at the highest coating concentration tested (9.5 μg/well). From the data presented in Fig. 1, it can be seen that heparin coated at 2 μg/well has a somewhat higher protein-binding capacity than that coated at 0.2 μg/well. This is consistent with our previous studies revealing that there is not a direct correlation between the amount of immobilized heparin and its protein-binding capability (Robinson et al. 2008). One possible explanation is that some minor components of the heterogeneous heparin preparation (e.g. highly charged species), which have a high-protein binding activity, become preferentially immobilized. In this regard, we have found that there is a small (<20%) enrichment of sulfated species on allyl amine-coated electrospun meshes (as determined from disaccharide analysis of the bound and prebound GAG preparation) when using a highly heterogeneous HS preparation from porcine mucosa (Iduron) at 10 μg/ml coating concentration (i.e. equivalent to 2 μg/well in our assay), but that this increase is only statistically significant for 6-O-sulfation (K. Meade, C.E. Johnson, Holley, R.J., S. Downes, J.D. Whittle, A.J. Day and C.L.R. Merry, in preparation). Therefore, there may be a small degree of preferential immobilization of certain highly sulfated species (e.g. particular 6-O-sulfated motifs) to allyl amine-coated microtiter plates when using heterogeneous heparin preparations, which could explain the higher binding of Link_TSG6 seen at 2 μg compared to 0.2 μg heparin/well (Figure 1). However, given that these are relatively subtle effects and that the XPS experiments with different GAGs (Fig. 5) indicate that binding is largely independent of the level of sulfation, in practical terms this does not appear to limit the use of this method for analysis of the GAG-binding specificity of a target protein.

Figure 6.

Figure 6

Open in a new tab

Determination of the amount of heparin immobilised on 100% allyl amine-coated microtiter plates. The amount of heparin bound on micotiter plate surfaces was determined using 35S-labeled heparin and correlated with the concentration of heparin [heparin] used to coat each well. Data were derived from 3 independent experiments performed in quadruplicate and are shown as average values ± S.E.M.

Analysis of GAG-binding proteins using the sugar array

In order to test the utility of the sugar array four non-related proteins were analyzed. Firstly, the bA-Link_TSG6 protein was tested (Fig. 7a) and shown to give data consistent with the various “pilot” studies described above. Furthermore, these results are in agreement with the known GAG-binding properties of the TSG-6 protein (Parkar & Day 1997; Mahoney et al. 2004; Mahoney et al. 2005; Milner et al. 2006), i.e. it has specificity for the sulfated GAGs heparin, chondroitin-4-sulfate and dermatan sulfate and the non-sulfated GAG hyaluronan. In this regard, it has been found that the TSG-6 Link module domain contains distinct hyaluronan and heparin-binding sites (Mahoney et al. 2005). The GAG-binding profile of the CCP6-8 “heparin”-binding domain from complement factor H obtained with the sugar array (Fig. 7b) was in agreement with previous studies showing that factor H interacts with heparin and dermatan sulfate (Pangburn et al. 1991; Meri & Pangburn 1994; Saito et al. 2005; Clark et al. 2006; Prosser et al. 2007). To our knowledge the direct binding of factor H to chondroitin-4-sulfate and hyaluronan has not been examined previously, however, these GAGs have been concluded to have no affect on factor H function (Meri & Pangburn 1994). Consistent with this we saw no binding of the factor H CCP6-8 domain to hyaluronan at either coating concentration in the sugar array and the low level of protein binding to chondroitin-4-sulfate was not dose-dependent indicating that this is likely to be non-specific; similarly a lack of binding was seen for these GAGs to immobilized CCP6-8 (S. J. Clark & A. J. Day, unpublished results). As described above the desulfation of heparin had a more dramatic effect on the interaction with Link_TSG6 compared to that with factor H. This is also clearly apparent from the sugar array profiles for these two proteins, indicating that TSG-6 and factor H display different structural requirements for heparin binding.

Figure 7.

Figure 7

Open in a new tab

GAG-binding profiles of four human proteins determined with the sugar array. Unfractionated heparin (UFH), LMW heparin (LMWH), chondroitin-4-sulfate (C4S), dermatan sulfate (DS), hyaluronan (HA), 2-O-desulfated heparin (2-O-deS) and N-desulfated/re-N-acetylated heparin (N-deS/R) were incubated with microtiter plates (coated with 100% allyl amine) at 0.2 μg/well (grey) and 2.0 μg/well (black). These immobilized GAGs were then used to determine the GAG-binding profiles of the proteins: a) the Link module from human TSG-6 (TSG-6); b) the Y384 variant of CCP6-8 from human factor H (Factor H); c) the PF1 region from human fibrillin-1 (Fibrillin-1) and d) the G1-domain from human versican (Versican). All values are plotted as mean absorbance (A405nm) ± SEM (n=6), corrected against blank wells (i.e. with no GAG); the data represent two independent experiments performed in triplicate each on a single microtiter plate.

The GAG-binding profiles of two other (less well characterized) proteins were also investigated (Fig. 7c, d), i.e. the N-terminal PF1 region of the elastic fibre protein fibrillin-1 (Marson et al. 2005) and the G1 domain of the proteoglycan versican (VG1) (Matsumoto et al. 2003; Seyfried et al. 2005). Using the prototype sugar array we identified the dose-dependent binding of PF1 to UFH and LMW heparin (Fig. 7c), consistent with earlier studies demonstrating that PF1 contains a heparin-binding site (Cain et al. 2005). Interestingly, there was much weaker binding to 2-O-desulfated heparin (2-O-deS), whereas the interaction with N-deS/R was very similar to that with UFH. These data indicate that 2-O-sulfate groups are likely to have an important role in mediating the interaction of heparin with PF1, whilst N-sulfates are not significantly involved. In addition, we have found for the first time that fibrillin-1 interacts with dermatan sulfate, a characteristic that might correlate with the role of fibrillin-1 in maintaining the elastic integrity of the skin (Marson et al. 2005; Cain et al. 2008). However, the PF1 region of fibrillin-1 showed no binding to the related GAG chondroitin-4-sulfate or to hyaluronan, which is consistent with previous observations (Cain et al. 2005). Conversely, analysis of the GAG-binding properties of VG1 (Fig. 7d) clearly showed that this domain does interact with the non-sulfated GAG hyaluronan. This is as expected given that versican is a well-established hyaluronan-binding protein, where the interaction site has been mapped to its N-terminal G1 domain (Matsumoto et al. 2003; Seyfried et al. 2005). Whilst the protein ligands of versican (a chondroitin sulfate proteoglycan) have been described in detail(Kawashima et al. 2000; Wu et al.2005; Kuznetsova et al. 2006), its binding profile to sulfated GAGs has remained undefined. As can be seen from Figure 7d, in addition to hyaluronan, the VG1 domain interacts in a dose-dependent manner with dermatan sulfate and also exhibits, apparently weaker, binding to heparin; however, no binding to chondroitin-4-sulfate was observed. Further research is now required to determine where the binding sites for these sulfated GAGs are located within the G1 domain and the functional relevance of these interactions.

Discussion

As described above we have developed a sugar array that can be used to determine the GAG binding profile of a target protein. Given that microtiter plates coated with 100% allyl amine are commercially available (BD Biosciences Heparin Binding Plate, formerly termed EpranEx plates (Clark et al. 2006)) and that all the GAGs we have used could be obtained from commercial suppliers, we believe that this simple methodology has wide utility. There is currently much interest in defining the carbohydrate-binding properties of proteins and there have been recent technological advances in this area (Fukui et al. 2002; Dendane et al 2008; Tateno et al. 2008; Zhi et al. 2008). However, these more sophisticated technologies (e.g. using evanescent-field fluorescence-assisted detection of protein binding to covalently immobilized glycans (Tateno et al. 2008), or surface patterning of GAGs onto the inner wall of fused-silica capillary tubes (Dendane et al. 2008)), or GAG attachment onto a gold-surface-based platform (Zhi et al. 2008), while of considerable potential value, are not, as yet, widely available, unlike the ELISA-like format that we have described. Importantly, our methodology, which involves the non-covalent association of GAGs with a plasma-polymerized surface, negates the need for labelling or chemical modification of the GAG (which in the other methods would have to be done for all GAG preparations (Fukui et al. 2002)), underlining the simplicity and accessi bility of this approach. In addition to the polysaccharides we chose to evaluate (which are representative of the different major types of GAG), we might have included other GAGs on our prototype sugar array, e.g. chondroitin-6-sulfate or keratan sulfate. It seems likely that these would have bound to the 100% allyl amine surface in a functional state given our results with the other GAGs analyzed, which have a wide variety of sulfation densities (i.e. from 0 to ~2.5 sulfates per disaccharide) and saccharide chemistry/linkage. However, the TSG-6 Link module does not interact with chondroitin-6-sulfate (Parkar & Day 1997; Heng et al. 2008) and its binding to keratan sulfate has not be tested, making these GAGs unsuitable for use during the development stage of the technology. Previously, three different preparations of heparan sulfate (HSI, HSII and heparan sulfate from bovine kidney), which have distinct sulfation patterns, have been used in binding assays on EpranEx plates (Mahoney et al. 2005), indicating that this family of GAGs could be included within this array format. Furthermore, we have shown that a wide-range of selectively desulfated heparin preparations can be successfully analyzed on the allyl amine surface (Clark et al. 2006); see also Fig. 4. It is clear, therefore, that the sugar array methodology described here could be customized by the user to allow a wide variety of analyses to be undertaken. Thus, in summary, this unique tool could be of significant use as a simple, frontline technique for the determination of protein-GAG interactions.

Materials and methods

GAGs and GAG-binding proteins

LMW heparin and hyaluronan were obtained from Sigma Aldrich (catalogue numbers H3400 and H1504, respectively) and chondroitin-4-sulfate was purchased from Calbiochem (catalogue number 230687). UFH corresponded to the 4th International Standard (IS) (Mulloy et al. 2000) and dermatan sulfate, was purified from porcine mucosa and characterized by 1H NMR spectroscopy as described previously (Pavão et al. 1995; Clark et al. 2006). Selectively desulfated heparins (2-O-desulfated (2-O-deS), 6-O-desulfated (6-O-deS), 2,6-O-desulfated (2,6-O-deS), N-desulfated (N-deS), N-desulfated/re-N-acetylated (N-deS/R)) were prepared from 2IS heparin (Mulloy et al. 2000) as before (Jaseja et al. 1989; Mulloy et al. 1994; Baumann et al. 1998; Ostrovosky et al. 2002). Link_TSG6, the Y384 variant of complement factor H CCPs6-8 and the PF1 region of fibrillin-1 were made and biotinylated as we have described previously (Parkar & Day 1997; Cain et al. 2005; Clark et al. 2006). S35-labeled heparin (N sulphonate) was supplied by GE Heathcare. The G1 domain of human versican (VG1) was expressed in E. coli , purified to homogeneity (H. L. Fielder, S. J. Clark, L. J. Collinson, M. E. M. Noble & A. J. Day, Manuscript submitted) and then biotinylated as described before (Kuznetsova et al. 2006).

Plasma polymerization

Different ratios of allyl amine (Sigma Aldrich) and octadiene (Sigma Aldrich) were plasma-deposited onto 96-well microtiter plates using the method described previously (Parry et al. 2006). This was done in a stainless steel reactor 60 cm long and 50 cm in diameter, with an internal electrode; it was pumped to a base pressure of 10−3 mbar by a two-stage rotary pump and liquid nitrogen cold trap and the plasma was excited by a 13.56 MHz radiofrequency generator. The reactor was maintained at an operating pressure of 2 × 10−2 mbar, while varying the partial pressure of the two monomers. Monomer ratios used were 100% allyl amine: 0% octadiene, 90% allyl amine: 10% octadiene, 80% allyl amine: 20% octadiene, 70% allyl amine: 30% octadiene, 60% allyl amine: 40% octadiene, 40% allyl amine: 60% octadiene, 20% allyl amine: 80% octadiene, and 0% allyl amine: 100% octadiene.

Microtiter plate binding assays

Plate assays were performed as we have reported previously (Mahoney et al. 2004; Mahoney et al. 2005; Clark et al. 2006; Robinson et al. 2008). Briefly, GAGs were incubated with untreated microtiter plates, or identical plates coated with diff erent monomer ratios of allyl amine:octadiene (see above), or BD Heparin Binding plates at 0-25 μg/ml (200 μl/well) in PBS for 18 h at room temperature in the dark. Plates were blocked with 1% (w/v) bovine serum albumin (Sigma Aldrich catalogue number A4503) in Standard Assay Buffer 6 (SAB6: 50 mM Sodium Acetate, 100 mM NaCl, 0.2% (v/v) Tween 20, pH 6) as described before (Mahoney et al. 2005). Wells were then incubated with biotinylated protein (e.g. 2 pmol/well bA-Link_TSG6 in SAB6) for 4 h at room temperature and the amount of bound protein, i.e. based on absorbance at 405 nm (A405nm ) after 10 min development, was detected as described in Robinson et al. 2008.

In the case of the sugar array analysis, GAGs (UFH, LMW heparin, chondroitin-4-sulfate, dermatan sulfate, hyaluronan, 2-O-deS and N-deS/R heparin) were incubated in PBS with 100% allyl amine plates at 0.2 or 2 μg/well and blocked as above. Incubation and detection of biotinylated protein was performed as above except factor H and VG1 were incubated (at 10 and 3 pmol/well, respectively), blocked and washed in 20 mM HEPES, 130 mM NaCl, 0.05% Tween 20, pH 7.3, whilst fibrillin-1 PF1 (4 pmol/well) was incubated in 20 mM Tris-HCl, 100 mM NaCl, 1 mM CaCl2 , pH 7.8. The sugar arrays were developed for 10 min.

X-ray photoelectron spectroscopy (XPS)

For XPS analysis, GAGs (10 μg/ml in PBS, 200 μl/well) were incubated with microtiter plates coated with 100% allyl amine. Plates were then placed in the dark for 18 h at room temperature, after which surfaces were washed twice with PBS and left overnight to dry. Well bottoms were punched out from the plate for analysis. XPS data were acquired on a Kratos Axis ULTRA ‘DLD’ X-ray photoelectron spectrometer equipped with a monochromatic Al-Kα X-ray source and operating with a base pressure in the range 10−8 mbar to 10−10 mbar. Survey spectra (1200 eV – 0 eV) were acquired at a pass energy of 160 eV and an X-ray power of 150 W. Sulfur S2p spectra were collected separately at a pass energy of 160eV. High-resolution C1s spectra were acquired at a pass energy of 40 eV and an X-ray power of 150 W. The analysis area was approximately 300 μm × 700 μm. All samples were run as insulators. Subsequent processing was carried out using CasaXPS (www.casaxps.com). Spectral data were corrected for instrumental transmission and charge corrected to the C1s peak at 285.0 eV. Quantitative data were then obtained using theoretical Scofield photo-ionisation cross sections. The level of adsorption of sulfated GAGs to the microtiter plate surface was determined by measuring the sulfur (S2p) signal.

Binding of35S-labeled heparin to allyl amine-coated plates

35S-labeled heparin (143 μCi/mg) and unlabeled LMW heparin (1:9 ratio) were added to 100% allyl amine plates at between 0.1 μg/well and 9.5 μg/well)in PBS (200 μ1/well) and incubated for 18 h at room temperature in the dark. The plates were then washed twice with PBS. Bound heparin was released by incubation of wells with 3M NaCl (200 μl) for 1 h and the amount of radoactivity determined by scintillation counting where background from plates incubated with PBS alone was subtracted.

Acknowledgements

We would like to thank Professor John T. Gallagher for helpful advice throughout this project. This work was funded by a BBSRC Small Business Research Initiative grant (BB/C51922X/1). AJD and LJC gratefully acknowledge support from the Arthritis Research Campaign (16539). XPS was carried out at the University of Sheffield Surface Analysis Centre.

Abbreviations

2 IS

second international standard of heparin

2-O-deS

2-O-desulfated heparin

2,6-O-deS

2,6-O-desulfated heparin

6-O-deS

6-O-desulfated heparin

BSA

bovine serum albumin

C4S

chondroitin-4-sulfate

DS

dermatan sulfate

GAG

glycosaminoglycan

HA

hyaluronan

HS

heparan sulfate

Link_TSG6

the Link module from human TSG-6

LMW

low molecular weight

N-deS

N-desulfated heparin

N-deS/R

N-desulfated and re-N-acetylated heparin

TSG-6

tumor necrosis factor-stimulated gene-6

UFH

unfractionated heparin

VG1

G1-domain from human versican

XPS

X-ray photoelectron spectroscopy.

Footnotes

Conflict of interest statement

JDW, MW and PNB are employees of BD Biosciences, which is continuing the development of the technology described in this paper

References

  1. de Agostini AL. An unexpected role for anticoagulant heparan sulfate proteoglycans in reproduction. Swiss Med. Wkly. 2006;136:583–590. doi: 10.4414/smw.2006.11368. [DOI] [PubMed] [Google Scholar]
  2. Banerji S, Wright AJ, Noble M, Mahoney DJ, Campbell ID, Day AJ, Jackson DG. Structures of the Cd44-hyaluronan complex provide insight into a fundamental carbohydrate-protein interaction. Nat. Struct. Mol. Biol. 2007;14:234–239. doi: 10.1038/nsmb1201. [DOI] [PubMed] [Google Scholar]
  3. Bao X, Mikami T, Yamada S, Faissner A, Muramatsu T, Sugahara K. Heparin-binding growth factor, pleiotrophin, mediates neuritogenic activity of embryonic pig brain-derived chondroitin sulfate/dermatan sulfate hybrid chains. J. Biol. Chem. 2005;280:9180–9191. doi: 10.1074/jbc.M413423200. [DOI] [PubMed] [Google Scholar]
  4. Baumann H, Scheen H, Huppertz B, Keller R. Novel regio- and stereoselective O-6-desulfation of the glucosamine moiety of heparin with N-methylpyrrolidinone-water or N,N-dimethylformamide-water mixtures. Carbohydr. Res. 1998;308:381–388. doi: 10.1016/s0008-6215(98)00097-4. [DOI] [PubMed] [Google Scholar]
  5. Boucas RI, Trindade ES, Tersariol IL, Dietrich CP, Nader HB. Development of an enzyme-linked immunosorbent assay (ELISA)-like fluorescence assay to investigate the interactions of glycosaminoglycans to cells. Anal. Chim. Acta. 2008;618:218–226. doi: 10.1016/j.aca.2008.04.059. [DOI] [PubMed] [Google Scholar]
  6. Cain SA, Baldock C, Gallagher JT, Morgan A, Bax DV, Weiss AS, Shuttleworth CA, Kielty CM. Fibrillin-1 interactions with heparin. Implications for microfibril and elastic fiber assembly. J. Biol. Chem. 2005;280:30526–30537. doi: 10.1074/jbc.M501390200. [DOI] [PubMed] [Google Scholar]
  7. Cain SA, Baldwin AK, Mahalingam Y, Raynal B, Jowitt TA, Shuttleworth CA, Couchman JR, Kielty CM. Heparan Sulfate Regulates Fibrillin-1 N- and C-terminal Interactions. J. Biol. Chem. 2008;283:27017–27027. doi: 10.1074/jbc.M803373200. [DOI] [PubMed] [Google Scholar]
  8. Camenisch TD, Spicer AP, Brehm-Gibson T, Biesterfeldt J, Augustine ML, Calabro A, Kubbalak S, Klwer SE, McDonald JA. Disruption of hyaluronan synthase-2 abrogates normal cardiac morphogenesis and hyaluronan-mediated transformation of epithelium to mesenchyme. J. Clin. Invest. 2000;106:349–360. doi: 10.1172/JCI10272. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Clark SJ, Higman VA, Mulloy B, Perkins SJ, Lea SM, Sim RB, Day AJ. His-384 allotypic variant of factor H associated with age-related macular degeneration has different heparin binding properties from the non-disease-associated form. J. Biol. Chem. 2006;281:24713–24720. doi: 10.1074/jbc.M605083200. [DOI] [PubMed] [Google Scholar]
  10. Dendane N, Hoang A, Renaudet O, Vinet F, Dunmy P, Defrancq E. Surface patterning of (bio)molecules onto the inner wall of fused-silica capillary tubes. Lab Chip. 2008;8:2161–2163. doi: 10.1039/b811786a. [DOI] [PubMed] [Google Scholar]
  11. Evanko SP, Tammi MI, Tammi RH, Wight TN. Hyaluronan-dependent pericellular matrix. Adv. Drug Deliv. Rev. 2007;59:1351–1365. doi: 10.1016/j.addr.2007.08.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Fukui S, Feizi T, Galustian C, Lawson AM, Chai W. Oligosaccharide microarrays for high-throughput detection and specificity assignments of carbohydrate-protein interactions. Nature Biotech. 2002;20:1011–1017. doi: 10.1038/nbt735. [DOI] [PubMed] [Google Scholar]
  13. Galtry CM, Kwok JCF, Carulli D, Rhodes KE, Fawcett JW. Distribution and synthesis of extracellular matrix proteoglycans, hyaluronan, link proteins and tenascin-R in the rat spinal cord. Eur. J. Neurosci. 2008;27:1373–1390. doi: 10.1111/j.1460-9568.2008.06108.x. [DOI] [PubMed] [Google Scholar]
  14. Gray E, Mulloy B, Barrowcliffe TW. Thromb. Heparin and low-molecular-weight heparin Haemost. 2008;99:807–818. doi: 10.1160/TH08-01-0032. [DOI] [PubMed] [Google Scholar]
  15. Harris EN, Weigel JA, Weigel PH. The human hyaluronan receptor for endocytosis (HARE/Stabilin-2) is a systemic clearance receptor for heparin. J. Biol. Chem. 2008;283:17341–17350. doi: 10.1074/jbc.M710360200. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Hasan M, Najjam S, Gordon MY, Gibbs RV, Rider CC. IL-12 is a heparin-binding cytokine. J. Immunol. 1999;162:1064–1070. [PubMed] [Google Scholar]
  17. Heng BC, Gribbon P, Day AJ, Hardingham TE. Hyaluronan binding to link module of TSG-6 and to G1 domain of aggrecan is differently regulated by pH. J. Biol. Chem. 2008;283:32294–32301. doi: 10.1074/jbc.M804155200. [DOI] [PubMed] [Google Scholar]
  18. Jaseja M, Rej RN, Sauriol F, Perlin AS. Novel regio- and stereoselective modifications of heparin in alkaline solution. Nuclear magnetic resonance spectroscopic evidence. Can. J. Chem. 1989;67:1449–1456. [Google Scholar]
  19. Johnson CE, Crawford BE, Stavridis M, Ten Dam G, Wat AL, Rushton G, Ward CM, Wilson V, van Kuppevelt TH, Esko JD, Smith A, Gallagher JT, Merry CL. Essential alterations of heparan sulfate during the differentiation of embryonic stem cells to Sox1-enhanced green fluorescent protein-expressing neural progenitor cells. Stem Cells. 2007;25:1913–1923. doi: 10.1634/stemcells.2006-0445. [DOI] [PubMed] [Google Scholar]
  20. Kawashima H, Hirose M, Hirose J, Nagakubo D, Plaas AHK, Miyasaka M. Binding of a large chondroitin sulfate/dermatan sulfate proteoglycan, versican, to L-selectin, P-selectin, and CD44. J. Biol. Chem. 2000;275:35448–35456. doi: 10.1074/jbc.M003387200. [DOI] [PubMed] [Google Scholar]
  21. Kuschert GS, Coulin F, Power CA, Proudfoot AE, Hubbard RE, Hoogewerf AJ, Wells TN. Glycosaminoglycans interact selectively with chemokines and modulate receptor binding and cellular responses. Biochemistry. 1999;38:12959–12968. doi: 10.1021/bi990711d. [DOI] [PubMed] [Google Scholar]
  22. Kuznetsova SA, Issa P, Perruccio EM, Zeng B, Sipes JM, Ward Y, Seyfried NT, Fielder HL, Day AJ, Wight TN, Roberts DD. Versican-thrombospondin-1 binding in vitro and colocalization in microfibrils induced by inflammation on vascular smooth muscle cells. J. Cell Sci. 2006;119:4499–4509. doi: 10.1242/jcs.03171. [DOI] [PubMed] [Google Scholar]
  23. Lin X. Functions of heparan sulfate proteoglycans in cell signaling during development. Development. 2004;131:6009–6021. doi: 10.1242/dev.01522. [DOI] [PubMed] [Google Scholar]
  24. Mahoney DJ, Mulloy B, Forster MJ, Blundell CD, Fries E, Milner CM, Day AJ. Characterization of the interaction between tumor necrosis factor-stimulated gene-6 and heparin: implications for the inhibition of plasmin in extracellular matrix microenvironments. J. Biol. Chem. 2005;280:27044–27055. doi: 10.1074/jbc.M502068200. [DOI] [PubMed] [Google Scholar]
  25. Mahoney DJ, Whittle JD, Milner CM, Clark SJ, Mulloy B, Buttle DJ, Jones GC, Day AJ, Short RD. A method for the non-covalent immobilization of heparin to surfaces. Anal. Biochem. 2004;330:123–129. doi: 10.1016/j.ab.2004.03.055. [DOI] [PubMed] [Google Scholar]
  26. Marson A, Rock MJ, Cain SA, Freeman LJ, Morgan A, Mellody K, Shuttleworth CA, Baldock C, Kielty CM. Homotypic fibrillin-1 interactions in microfibril assembly. J. Biol. Chem. 2005;280:5013–5021. doi: 10.1074/jbc.M409029200. [DOI] [PubMed] [Google Scholar]
  27. Matsumoto K, Shionyu M, Go M, Shimizu K, Shinomura T, Kimata K, Watanabe H. Distinct interaction of versican/PG-M with hyaluronan and link protein. J. Biol. Chem. 2003;278:41205–41212. doi: 10.1074/jbc.M305060200. [DOI] [PubMed] [Google Scholar]
  28. Meri S, Pangburn MK. Regulation of alternative pathway complement activation by glycosaminoglycans: specificity of the polyanion binding site on factor H. Biochem. Biophys. Res. Commun. 1994;198:52–59. doi: 10.1006/bbrc.1994.1008. [DOI] [PubMed] [Google Scholar]
  29. Milner CM, Higman VA, Day AJ. TSG-6: a pluripotent inflammatory mediator? Biochem. Soc. Trans. 2006;34:446–450. doi: 10.1042/BST0340446. [DOI] [PubMed] [Google Scholar]
  30. Mulloy B. Gel permeation chromatography of heparin. In: Volpi N, editor. Analytical Techniques to Evaluate the Structure and Function of Natural Polysaccharides: Glycosaminoglycans. Research Signpost; Trivandrum: 2002. pp. 115–128. [Google Scholar]
  31. Mulloy B, Gray E, Barrowcliffe TW. Characterization of unfractionated heparin: comparison of materials from the last 50 years. Thromb. Haemostasis. 2000;84:1052–1056. [PubMed] [Google Scholar]
  32. Mulloy B, Forster MJ, Jones C, Drake AF, Johnson EA, Davies DB. The effect of variation of substitution on the solution conformation of heparin: a spectroscopic and molecular modelling study. Carbohydr. Res. 1994;255:1–26. doi: 10.1016/s0008-6215(00)90968-6. [DOI] [PubMed] [Google Scholar]
  33. Nadesalingam J, Lopez Bernal A, Dodds AW, Willis AC, Mahoney DJ, Day AJ, Reid KBM, Palaniyar N. Identification and characterization of a novel interaction between pulmonary surfactant protein D and decorin. J. Biol. Chem. 2003;278:25678–25687. doi: 10.1074/jbc.M210186200. [DOI] [PubMed] [Google Scholar]
  34. Najjam S, Gibbs RV, Gordon MY, Rider CC. Characterization of human recombinant interleukin 2 binding to heparin and heparan sulfate using an ELISA approach. Cytokine. 1997;9:1013–1022. doi: 10.1006/cyto.1997.0246. [DOI] [PubMed] [Google Scholar]
  35. Nishioka S, Aikawa J, Ida M, Matsumoto I, Street M, Tsujimoto M, Kojima-Aikawa K. Ligand-binding activity and expression profile of annexins in Caenorhabditis elegans. J. Biochem. 2007;141:47–55. doi: 10.1093/jb/mvm009. [DOI] [PubMed] [Google Scholar]
  36. Ostrovsky O, Gallagher JT, Mulloy B, Fernig DG, Berman B, Ron D. Differential effects of heparin saccharides on the formation of specific fibroblast growth factor (FGF) and FGF receptor complexes. J. Biol. Chem. 2002;277:2444–2453. doi: 10.1074/jbc.M108540200. [DOI] [PubMed] [Google Scholar]
  37. Pangburn MK, Atkinson MA, Meri S. Localization of the heparin-binding site on complement factor H. J. Biol. Chem. 1991;266:16847–16853. [PubMed] [Google Scholar]
  38. Parkar AA, Day AJ. Overlapping sites on the Link module of human TSG-6 mediate binding to hyaluronan and chrondroitin-4-sulphate. FEBS Lett. 1997;410:413–417. doi: 10.1016/s0014-5793(97)00621-2. [DOI] [PubMed] [Google Scholar]
  39. Parry KL, Shard AG, Short RD, White RW, Whittle JD, Wright A. ARXPS characterisation of plasma polymerised surface chemical gradients. Surf. Interf. Anal. 2006;38:1497–1504. [Google Scholar]
  40. Pavão MSG, Mourão PAS, Mulloy B, Tollefsen DM. A unique dermatan sulfate-like glycosaminoglycan from ascidian. Its structure and the effect of its unusual sulfation pattern on anticoagulant activity. J. Biol. Chem. 1995;270:31027–31036. doi: 10.1074/jbc.270.52.31027. [DOI] [PubMed] [Google Scholar]
  41. Prosser BE, Johnson S, Rovers P, Herbert AP, Blaum BS, Tyrrell J, Jowitt TA, Clark SJ, Tarelli E, Uhrin D, Barlow PN, Sim RB, Day AJ, Lea SM. Structural basis for complement factor H linked age-related macular degeneration. J. Exp. Med. 2007;204:2277–2283. doi: 10.1084/jem.20071069. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Robinson DE, Marson A, Short RD, Buttle DJ, Day AJ, Parry P, Wiles M, Highfield P, Mistry A, Whittle JD. Surface Gradient of functional heparin. Adv. Mater. 2008;20:1166–1169. [Google Scholar]
  43. Saito A, Munakata H. Factor H is a dermatan sulfate-binding protein: identification of a dermatan sulfate-mediated protease that cleaves factor H. J. Biochem. 137:225–233. doi: 10.1093/jb/mvi025. [DOI] [PubMed] [Google Scholar]
  44. Scarchilli L, Camaioni A, Botazzi B, Negri V, Doni A, Deban L, Bastone A, Salvatori G, Mantovani A, Siracusa G, Salustri A. PTX3 interacts with inter-alpha-trypsin inhibit or: implications for hyaluronan organization and cumulus oophorus expansion. J. Biol. Chem. 2007;282:30161–30170. doi: 10.1074/jbc.M703738200. [DOI] [PubMed] [Google Scholar]
  45. Seyfried NT, McVey GF, Almond A, Mahoney DJ, Day AJ. Expression and purification of functionally active hyaluronan-binding domains from human cartilage link protein, aggrecan and versican: formation of ternary complexes with defined hyaluronan oligosaccharides. J. Biol. Chem. 2005;280:5435–5448. doi: 10.1074/jbc.M411297200. [DOI] [PubMed] [Google Scholar]
  46. Tateno H, Mori A, Uchiyama N, Yabe R, Iwaki J, Shikanai T, Angata T, Narimatsu H, Hirabayashi J. Glycoconjugate microarray based on an evanescent-field fluorescence-assisted detection principle for investigation of glycan-binding proteins. Glycobiology. 2008;18:789–798. doi: 10.1093/glycob/cwn068. [DOI] [PubMed] [Google Scholar]
  47. Taylor KR, Gallo RL. Glycosaminoglycans and their proteoglycans: host-associated molecular patterns for initiation and modulation of inflammation. FASEB J. 2006;20:9–22. doi: 10.1096/fj.05-4682rev. [DOI] [PubMed] [Google Scholar]
  48. Wisniewski HG, Snitkin EV, Mindrescuet C, Sweet MH, Vilcek J. TSG-6 protein binding to glycosaminoglycans: formation of stable complexes with hyaluronan and binding to chondroitin sulfates. J. Biol. Chem. 2005;280:14476–14484. doi: 10.1074/jbc.M411734200. [DOI] [PubMed] [Google Scholar]
  49. Wu YJ, La Pierre DP, Wu J, Yee AJ, Yang BB. The interaction of versican with its binding partners. Cell Res. 2005;15:483–494. doi: 10.1038/sj.cr.7290318. [DOI] [PubMed] [Google Scholar]
  50. Yang B, Yang BL, Goetinck PF. Biotinylated hyaluronic acid as a probe for identifying hyaluronic acid-binding proteins. Anal. Biochem. 1995;228:299–306. doi: 10.1006/abio.1995.1354. [DOI] [PubMed] [Google Scholar]
  51. Zhi ZL, Laurent N, Powell AK, Karamanska R, Fais M, Voglmeir J, Wright A, Blackburn JM, Crocker PR, Russell DA, Flitsch S, Filed RA, Turnbull JE. A versatile gold surfcae approach for fabrication and interrogation of glycoarrays. ChemBioChem. 2008;9:1568–1575. doi: 10.1002/cbic.200700788. [DOI] [PubMed] [Google Scholar]

RESOURCES