Matricellular proteins are secreted into the extracellular environment, or matrix, but do not play a primary structural role in this location. Rather, these proteins modulate cell function by interacting with cell-surface receptors, proteases, hormones, and other bioeffector molecules, as well as with structural matrix proteins such as collagens. The term ‘matricellular’ was introduced to explain the unusual diversity of functions that were beginning to be recognized in proteins such as thrombospondin-1(TSP-1), SPARC, and tenascin-C. (Sage and Bornstein 1991; Bornstein 1995). The intent of segregating this subclass of secreted proteins was to emphasize that the extracellular environment was a major factor in regulating their synthesis, a process that had been termed ‘dynamic reciprocity’ at an earlier date (Bornstein et al. 1982). A review that summarized many of the considerations that led to the concept of matricellular proteins was published in Methods in Cell Biology (Bornstein 2002).

In additional reviews (Bornstein 2001; Bornstein and Sage 2002) some of the distinguishing characteristics of matricellular proteins were summarized in greater detail. These included: 1) high levels of expression during development and in response to injury; 2) binding to many cell-surface receptors, components of the extracellular matrix, growth factors, cytokines, and proteases; 3) induction of de-adhesion or counter-adhesion in contrast to the adhesivity of most matrix proteins (Murphy-Ullrich 2001; Liu et al. 2009), and 4) a grossly normal or subtle phenotype that is observed in mice with a targeted disruption (knockout) of some matricellular protein genes. The complexity of the functions of most matricellular proteins results from the fact that these functions are, in large part, contextual, i.e. they derive from the different structural proteins, cell-surface receptors, proteases, and cytokines with which these proteins come in contact in the local environment of different tissues.

The unexpected phenotypes of some matricellular protein-null mice have provided clues to the functions of matricellular proteins. Thus, for example, the abnormally shaped and sized collagen fibrils observed in TSP-2-null mice in skin and other connective tissues led to the demonstration that TSP-2 functions as a clearance factor for MMP2 in the pericellular environment (Yang et al. 2001). In turn, the elevated MMP2 levels in the pericellular environment of TSP2-null mice reduced tissue transglutaminase activity and the collagen crosslinks generated by this enzyme (Agah et al. 2005). As another example, the cataracts and increased adipogenesis encountered in SPARC-null mice led to a better understanding of extracellular matrix assembly (Brekken and Sage 2000) and SPARC signaling pathways (Nie and Sage 2009).

Although a useful review of the functions of a limited number of matricellular proteins in bone biology has been published (Alford and Hankenson 2006), and Sangaletti and Colombo (2008) have surveyed the functions of matricellular proteins in inflammation and cancer, this issue of the JCCS provides a more comprehensive coverage of the field and takes advantage of recent mechanistic information that improves our understanding of how matricellular proteins function.

Thus, for TSPs, Bornstein has described the mechanisms that enable TSP-1 and TSP-2 to function as both angiogenic and anti-angiogenic proteins, and the potential clinical applications of these properties; Hankenson and Delany consider the roles of TSP-2 and SPARC in bone formation; Lawler describes the effects of TSPs on the structure of the extracellular matrix; MacLauchlan and Kyriakides review the role of TSPs in wound healing, ischemia, and the foreign body response; and Schellings et al. review the roles of TSPs in cardiac remodeling.

For SPARC and members of the SPARC family, Bradshaw discusses the role of SPARC in extracellular matrix assembly; Arnold and Brekken consider the role of SPARC as a regulator of tumorigenesis, and Nie and Sage summarize the evidence for SPARC as an inhibitor of adipogenesis.

Some matricellar proteins are represented by single presentations. Lund et al. present the role of osteopontin in inflammatory proceses; Midwood and Orend consider the role of tenascin-C in tissue injury and tumorigenesis; Norris et al. describe the many facets of the effects of periostin on cardiac development, remodeling, and pathophysiology ; Merline et al. present evidence for the matricellular functions of small leucine-rich proteoglycans (SLRPS), and Yanagisawa et al. present data supporting the inclusion of fibulin-5 in the matricellular protein family. Finally, Eroglu summarizes the evidence for the roles of matricellular proteins in the development and function of the nervous system.

Clearly, the articles mentioned above do not cover all the proteins that are now considered to be matricellular. Prominent among those not included in this review are several members of the CCN (Cyr-61, Connective tissue growth factor, and Nov family of proteins that were recently reviewed (Chen and Lau 2009; Holburn et al. 2008; and Yeger and Perbal 2007), tenascin- X (Zweers et al. 2004), the galectins (Elola et al. 2007), plasminogen activator inhibitor type 1 (PAI-1) (Maquerlot et al. 2006), and autotaxin (Dennis et al. 2008).

An issue that has not been addressed in the categorization of matricellular proteins is whether to include matrix proteins, the matricellular function of which, as defined above, is revealed only when fragments of the proteins are investigated. It is now commonly recognized that new functions, not present in the intact protein, can be generated by limited physiological proteolysis (Sage 1997; Davis et al. 2000). However, it would seem preferable at this time, in the interest of brevity, to limit our discussion to matrix proteins with matricellular properties that are expressed in the intact proteins.

A PubMed search of the literature, performed at the end of the 2008 calendar year, revealed that there were 348 publications that used the term ‘matricellular’ in the title and/or abstract since its first use in 1995. The increase in recent years has been almost exponential, with 65 citations listed during 2008. A number of these citations concern proteins with credentials as matricellular proteins that are incomplete or questionable, but it is reasonable to expect that the list of matricellular proteins will grow as more information is gathered about other extracellular proteins with non-structural functions.