SMOC-1 and SMOC-2 proteins are the most recently characterized members of the BM-40 family. Both of them are present in many cell types and are highly expressed during embryogenesis, wound healing, and other physiological processes involving extensive tissue remodeling 
. To identify the molecular mechanisms behind these processes we focused on SMOC-1 protein and especially on its EC domain. Based on the high sequence homology between the heparin-binding regions of SMOC-1 and SMOC-2 (See ), all results presented here for SMOC-1 can likely be extended to SMOC-2. As already shown for EC domains of other BM-40 family proteins, BM-40 
, hevin 
and testican 
the EC module of SMOC-1 is an autonomously folding domain with a distinct affinity for calcium ions. A decrease in the negative molar ellipticity was observed when calcium ions were removed by adding EDTA. This demonstrates that the EF hands of S1EC are active in calcium binding, which is in agreement with previously published data on SMOCs 
The EC domain of SMOCs contains a unique cluster of basic amino acids, representing a potential GAG-binding site that does not appear in other members of the BM-40 family. Our experiments have shown that S1FL as well as S1EC bound HP and HS but not CS or DS. The selectivity for heparin can be explained by the fact that heparin, on behalf of its dense sulfation, carries the amount of negative charges sufficient for the formation of ionic bonds to positively charged amino groups of the protein. It can be therefore seen that it is the lack of negative charge in the case of CS and DS that is causing repulsive forces between the sugar moiety and the protein and thus disabling energetically favorable binding. The likely biological binding partners of the SMOCs are HSs, which exhibit diverse modification patterns 
and multiple “domains” with distinctive sulfation degrees are often present within a single chain. Our results indicate that SMOC-1 binds only to highly sulfated regions of HS which explains the lower stoichiometric ratio of S1FL binding to HS, resulting in higher elution volumes in comparison to the S1FL-HP complex. The data obtained using SEC where S1EC alone eluted at 15 kDa and S1EC in complex with heparin eluted as a broad peak with a maximum at approximately 75 kDa suggest that on average four EC domains can bind to a single heparin molecule. Taking into account the average disaccharide molecular mass of 570 Da, it follows that each EC molecule occupies six disaccharide units. This complies well with the results from molecular modeling where four disaccharide units were necessary to cover the complete binding site. The difference of two disaccharide unites can likely be attributed to hindrance between adjacent EC domains. As in the case of many other ECM proteins 
, S1EC bound heparin in a calcium-independent manner, since chelation of the Ca2+
ions in EF hands did not result in a pronounced decrease in heparin-binding affinity. This indicates that correct conformation of helices involved in Ca2+
-binding is not required for successful heparin binding.
At the sequence level, the proposed heparin-binding site on S1EC is a linear motif of about 20 residues in length which forms two alpha-helices running antiparallel to each other. Based on this, the motif can be described as a blend of a strictly linear motif and a higher order spatial motif. The large basic interacting surface is therefore comprised of two linearly contiguous clusters brought together through correct folding. In comparison to the recently identified heparin-binding site in transglutaminase-2 
, where two clusters of positively charged residues are 300 residues apart in the primary sequence, the basic clusters in S1EC are not distant at the sequence level but correct folding is still necessary for the formation of a functional heparin-binding site. The proposed heparin-binding site in S1EC contains two “CPC clip motifs” comprising one polar and two cationic residues, which has recently been identified as a common heparin-binding motif 
. In addition, the S1EC-binding site also contains non-basic amino acids, known to be involved in heparin binding in other proteins: helix E contains a Glu residue, an amino acid that is supposed to be important for both acidic and bFGF to interact with heparin 
as well as Tyr which may support heparin binding via formation of a hydrogen bond to hydroxyl groups on the heparin 
Great importance is lately attributed to HP and HS-binding proteins. It has been shown that HS can interact with a wide range of proteins due to its high content of charged groups and is essential in various biological processes due to its structural diversity 
. Since HS, and not HP, is present in ECM, we investigated the physiological significance of S1EC binding by determining the role of S1EC in supporting the adhesion to epithelial cells. This property was recently described for SMOC-2 EC (S2EC) domain and integrins αvβ1 and αvβ6 were identified as the receptors involved in S2EC-mediated cell adhesion 
. Since HaCaT cells exhibit the same calcium-dependent behavior in adhesion to S1EC, it seems reasonable to assume that these receptors are also involved in SMOC-1-mediated cell adhesion. Indeed, colocalization experiments using S1EC have identified integrin β6 at the sites of focal adhesions (see ). Interestingly, formation of focal adhesions was not observed in other members of BM-40 family. Conversely, abrogation of focal adhesions occurred in the presence of BM-40 as well as hevin/SC1 
Since it is known that HSPGs can function as co-receptors in integrin-mediated cell adhesion, we have investigated whether cell adhesion on S1EC depends on its heparin-binding activity. Data acquired using several approaches indicated that the heparin-binding site must be available and functional to interact with cell-surface HS. Firstly, addition of soluble heparin reduced adhesion of HaCaT cells to S1EC in a dose dependent manner, ruling out the possibility that it is the conformational change caused by heparin binding that accounts for successful binding to integrin as recently shown for focal adhesion kinase 
. The value of IC50
for heparin, estimated from cell adhesion assays is similar to the Kd
value, calculated from intrinsic fluorescence measurements. Secondly, damaging the integrity of cell surface HSs by addition of sodium chlorate, which inhibits sulfation of GAGs, resulted in abrogated adhesion, and thirdly, the DBMUT showed reduced HaCaT adhesion activity. These data collectively suggest that it is the interaction with cell surface HSPGs that is a major factor for S1EC-mediated cell adhesion. The most likely targets are proteoglycans of the syndecan or glypican families. Especially the former are well known to play important roles as co-receptors in integrin-mediated cell adhesion 
. However, we can conclude that HSPGs on the cell surface do not suffice for successful binding, since under chelating conditions HaCaT cells failed to adhere to S1EC even though recombinant S1EC retains its heparin-binding activity under these conditions. HSPGs can thus be considered as co-receptors, which stabilize the formed adhesion complex. While the interaction between S1EC and HSPGs alone may not be tight enough to promote cell adhesion, it is likely sufficient to serve a regulatory purpose. This hypothesis is further supported by the fact that S1EC retained a certain degree of cell-binding activity even in the presence of saturating concentrations of heparin () and when the heparin-binding site was completely removed ().
Comparison with other known heparin-binding proteins shows that the interaction is of intermediate affinity, which supports the regulatory or accessory role of these interactions in cell adhesion. The affinity of SMOC-1 for heparin was measured using soluble heparin. The heparan sulfate found in the pericellular space is bound to the protein backbone which is in turn bound to the cell membrane. For this reason the heparan sulfate is not freely diffusible and its activity is not directly comparable to the activity of soluble heparin (the activity of pure solids is 1). Moreover, multiple spatially adjacent HSPG molecules can create the effect of a very high local “concentration” of heparan sulfate. Therefore, measurements with soluble heparin performed in dilute solutions are only a measure to compare the affinity of SMOC-1 for heparin to other known heparin-binding proteins and are not truly representative of in vivo situations. These are more closely simulated by the cell adhesion experiments which we have performed and which have clearly shown that the amounts of HSPGs on the cell surface are certainly sufficient to influence cell adhesion to SMOC-1.
Several HSPGs, including collagen XVIII, perlecan and agrin 
, are also abundant in the BM, a complex form of ECM, where SMOC-1 was shown to localize. Based on currently available data it is yet impossible to predict which of these HSPGs SMOC-1 can interact with. On the other hand not enough data is yet available to exclude the possibility of SMOC-1 binding to other proteins where it could fine-tune the physiological processes as an adaptor protein. Further information at the molecular level is therefore necessary to assess the role of SMOC-1 in ECM.