S100A10: a complex inflammatory role

In this issue of Blood, O'Connell and colleagues demonstrate that macrophage recruitment in response to an inflammatory stimulus is markedly decreased in S100A10-deficient mice compared with wild-type mice, establishing the plasminogen receptor, S100A10, as a major mediator of this key plasmin-dependent component of the inflammatory response.¹ 

Activation of plasminogen to the broad-spectrum serine proteinase, plasmin, is markedly enhanced when plasminogen is bound to the cell surface.²  The plasmin formed remains associated with the cell where plasmin is protected from inactivation by its primary inhibitor, α₂-antiplasmin.²  This results in arming cells with the broad-spectrum proteolytic activity of plasmin. Because macrophages have a high capacity for plasminogen, no single molecule can account for the entire plasminogen-binding capacity of these cells. However, a specific subset of plasminogen receptors, exposing C-terminal lysines on the cell surface, is responsible for promoting plasminogen activation on macrophages and other eukaryotic cells.³ 

In addition to its well-known role in thrombolysis, plasmin(ogen) is required for optimal recruitment of macrophages in the response to inflammatory stimuli.^4,5  During macrophage recruitment, cell-surface plasmin activity functions to directly degrade components of the extracellular matrix as well as to activate pro–matrix metalloproteinase-9 for degradation of collagen, thus allowing passage of cells through tissues.⁵  Plasminogen-dependent macrophage recruitment is regulated by plasma carboxypeptidase B⁶  (whose active form removes C-terminal basic arginyl and lysyl residues) consistent with a requirement for interaction of plasminogen with the C-terminal lysine residues of plasminogen receptors on the cell surface.

S100A10 belongs to the S100 family of calcium-binding proteins and is usually found in cells bound to its ligand, annexin A2, in the heterotetrameric complex (S100A10)₂–(annexin A2)₂, AIIt.¹  S100A10 is synthesized with a C-terminal lysine and binds plasminogen.¹ 

O'Connell and colleagues show that the number of macrophages recruited to the peritoneal cavity in response to thioglycollate is 53% less in S100A10 ^−/− mice, compared with wild-type mice. The functions of other macrophage plasminogen receptors that are synthesized with C-terminal lysines have been evaluated, previously, in the thioglycollate-induced peritonitis model after intravenous injection of specific antibodies. These receptors include histone H2B (48% less macrophage recruitment compared with injection of nonimmune control),⁷  α-enolase (24% less recruitment),⁷  and, recently, Plg-R_KT (58% less recruitment).⁸  In another model of inflammation, monocytoid cells overexpressing α-enolase exhibit increased recruitment to the acutely inflamed lung compared with cells overexpressing a mutant α-enolase that does not bind plasminogen (lacking the 30 C-terminal amino acids).⁹  The study by O'Connell is the first examination of macrophage recruitment in a specific plasminogen receptor knockout mouse model.

In concordance with the foregoing studies, the O'Connell study emphasizes that multiple plasminogen receptors with C-terminal basic residues participate in macrophage invasion. S100A10^−/− macrophages exhibited impaired plasmin-dependent invasion through Matrigel (an extract of murine tumor-derived basement membrane, commonly used to assess the role of proteinases in cell migration). When macrophages from either wild-type or S100A10^−/− mice were treated with carboxypeptidase B, a major reduction in Matrigel invasion was observed. O'Connell and colleagues interpreted the effect of carboxypeptidase B on S100A10^−/− macrophages to indicate that other plasminogen receptors with C-terminal basic residues also participate in macrophage invasion.

It is apparent that the sum of the effects of functional blockade of specific plasminogen receptors, analyzed to date, is greater than a 100% reduction in plasminogen-dependent macrophage recruitment. Thus, it is possible that each specific plasminogen receptor may be required at different steps in the inflammatory response; for example, crossing different layers of peritoneal tissue at which different contributions of direct plasmic cleavage of the extracellular matrix or activation of matrix metalloproteinase-9 for collagen degradation⁵  may predominate.

In the O'Connell study, S100A10^−/− mice also showed a marked decrease in neutrophil recruitment in response to thioglycollate. However, neutrophil recruitment in response to thioglycollate is not diminished in plasminogen^−/− mice compared with wild-type mice.^4,5  O'Connell and colleagues are currently investigating the possibility that S100A10 may regulate proteases other than plasmin on the neutrophil surface. Thus, it is likely that the contribution of S100A10 to cellular inflammation is both complex and multifaceted.

S100A10 is present on most cells in a heterotetrameric complex with its ligand, annexin A2: AIIt composed of 2 molecules of S100A10 and 2 molecules of annexin A2.¹  Although annexin A2 is not synthesized with a C-terminal lysine, it has been proposed that annexin A2 may directly bind plasminogen after processing at a specific site to reveal a C-terminal lysine.¹⁰  Distinguishing between the contributions of S100A10 and annexin A2 to plasminogen binding has been very difficult because annexin A2^−/− mice also exhibit markedly reduced S100A10 protein expression.¹¹  Surprisingly, in the O'Connell study, knockout of S100A10 also reduced cell-surface expression of annexin A2.¹  O'Connell and coworkers used 2 approaches to sort out the respective contributions of S100A10 and annexin A2 to functions dependent on plasminogen binding. First, cell-surface annexin A2 on thioglycollate-stimulated peritoneal macrophages was analyzed in Western blotting. Processing of annexin A2 was not detectable, suggesting that annexin A2 was not directly binding plasminogen on the recruited peritoneal macrophages. Second, when S100A10 was added back to S100A10^−/− macrophages (that exhibited reduced invasion across a Matrigel barrier), invasiveness was increased. However, addition of annexin A2 to S100A10^−/− macrophages did not significantly increase invasion. These data suggest that annexin A2 most likely serves as a cell-surface binding partner of the plasminogen receptor, S100A10, but does not directly bind plasminogen under these experimental conditions.

The study by O'Connell provides a strong rationale for future studies to address the interplay between the contributions of multiple plasminogen receptors to macrophage recruitment in the inflammatory response. In addition, the distinct contributions of these plasminogen receptors should also be investigated in other physiologic and pathologic processes requiring cell migration.