Thus, exploitation of the roles of key anti-inflammatory cytokines may enable M. bovis to evade elimination by the host adaptive immune response, enabling disease progression. FNI modules adopt a characteristic fold with a minor b-sheet, a major b-sheet, two conserved disulfide bonds, and a hydrophobic core containing a conserved tyrosine and tryptophan. In contrast, the Ile-to-Ala mutations alter binding of a monoclonal antibody directed towards an epitope in 9FNI. These results raise the possibility that mutations of the IGD motifs cause structural changes in FNI modules that impair the picomolar interactions of 70K with the cell components that mediate MSF activity. The 70K fragment of FN binds to cell-surface FN Armepavine assembly sites with the same nanomolar affinity as full-length FN, and at a thousand-fold lower picomolar concentration, 70K stimulates migration of fibroblasts into type I collagen. Full-length FN and the N-3FNIII construct lack MSF activity, probably because the sites required for the activity are obscured in the quaternary structure. Signaling mediating MSF activity seems to involve inhibition of AKT. The cell-surface molecules that initiate such signaling remain to be identified. Experiments blocking avb3 integrins with antibodies indicate a role for this integrin in migration, but whether avb3 interacts with MSF to initiate signaling or engages binding sites in the supporting collagen gel to mediate migration is not known. We are not aware of avb3 interacting with any ligands with picomolar affinity. In addition, there is no evidence that the nanomolar binding of 70K to the cell surface assembly sites is dependent on avb3. A cyclic RGD peptide that inhibits avb3 cell adhesive activity does not block binding of 70K at nanomolar concentrations to the cell-surface. Further, although NGR motifs in 3FNI and 5FNI spontaneously convert to integrin-binding isoDGR sequences, and isoDGR can interact with avb3 integrins, the conversion of NGR to isoDGR is incomplete, and mutagenesis experiments indicate that the sequences are not responsible for binding of 70K to assembly sites. Previous NMR studies showed little structural alteration in 9FNI with mutations in the IGD motif; these changes involved Ser575 and the disulfide connecting the A- and D-strands. Our results showed that the I572A mutation results in decreased binding of mAb 5C3 to its epitope, which contains Gly567 five amino acids away from Ile572. Further, mutations in the IGD motif, but not the 30 differences between rat and human 70K or mutations in NGR motifs, affected the ability of 70K to interact with the Batimastat bacterial peptide FUD. As with other FN-binding sequences from bacteria, FUD appears to bind to the E-strand of FNI modules. Because the IGD sequence is in the loop connecting B and C strands, it is unlikely that mutations in IGD motifs interfere directly with FUD binding. Instead, because the conserved isoleucine is part of the hydrophobic core of FNI modules, we hypothesize that the Ile-to-Ala mutations disrupt the hydrophobic core of the FNI module which in turn deforms the modules and alters the ability of FUD to bind. The IGD motif is part of the sequence YX G found in nine of 12 FNI modules. It is noteworthy that mutation of the tyrosine, which is found in all 12 FNI modules and contributes to the hydrophobic core, to serine in any of the five N-terminal FNI modules is deleterious to secretion of 70K and binding of 70K to matrix assembly sites or to Staphylococcus aureus, an interaction that requires b-zipper formation.