Background
The two ubiquitous Ca2+-dependent, cysteine proteases known as μ-calpain (calpain-1) and m-calpain (capain-2), are the founding members of a gene family comprising 13 genes in mammals [1-3]. Both are heterodimeric enzymes consisting of distinct 80 kDa catalytic subunits, encoded by the Capn1 (μ-80 k) and Capn2 (m-80 k) genes, respectively, that associate with a common 28 kDa regulatory subunit encoded by the Capn4 gene. The μ-80 k and m-80 k subunits share 62% amino acid sequence identity, and are very similar in terms of structure, protein chemistry, and in vitro substrate specificity. Despite these similarities, the differential expression patterns of μ- and m-calpain in mammalian tissues suggest they have some isoform specific and distinct functions. The μ and m designations derive from the levels of Ca2+ required in vitro for optimal activation; 10–50 μM Ca2+ for μ-calpain and 0.3–0.35 mM Ca2+ for m-calpain. It is generally assumed that μ- and m-calpain maintain their differential sensitivities to calcium in vivo, although this has not yet been strictly demonstrated. Furthermore, since the cytoplasmic free Ca2+ concentration is typically less than 1 μM, it is also assumed that other in vivo factors must contribute to regulation of these enzymes [3].
Without apriori knowledge of the factors regulating calpain activity or their relevant substrates, elucidation of biological functions for calpains presents a challenge. Research on calpains has linked them with a wide variety of functions including muscle growth, development, degeneration (3), neuronal growth and neurodegeneration [4], cell cycle progression [5,6], signal cascades triggered by integrins and growth factors [7], membrane protrusion [8], remodeling of the cytoskeleton and cell migration [9-15], and regulation of cell death via both necrosis and apoptosis [16-22]. To date, the literature suggests a complex interplay between caspases and calpains [23,24] and impact of calpain on cell death pathway components [25]. The lack of highly specific cell-permeable inhibitors of calpains contributes to the challenge of investigating and defining calpain functions in these processes. Although over-expression of calpastatin, the endogenous protein inhibitor of μ- and m-calpain provides an important approach for these efforts, it will not distinguish isoform specific functions [24,26,27]. Some work has suggested isoform specific roles, such as a role for m-calpain in epidermal growth factor (EGF)-induced cell motility [28,29] and a role for μ-calpain in interferon-inducible protein 9-induced migration of keratinocytes [28]. A cell permeable calpain inhibitor (which likely inhibits other thiol-proteases as well) has been used to select cells lacking μ-calpain which display reduced proliferation rates [30]. Interestingly, m-calpain expression persisted in these cells, suggesting a possible requirement of m-calpain for cell survival [30].
Targeted gene deletion in mice provides a powerful approach to determining the physiological roles of μ- and m-calpain and the opportunity to approach their isoform specific functions. Initial studies targeted Capn4 based on the prediction that loss of this calpain subunit would abolish activity of both μ- and m-calpain. Capn4-/- murine embryos died between days 10 and 11 of gestation, and there was no detectable μ- or m-calpain activity in these or younger embryos [31]. Capn4-/- murine embryonic fibroblasts (MEFs) could be cultured from these embryos, although they also lacked calpain activity as assessed by casein zymography or by the formation of characteristic spectrin breakdown products, and they displayed migration defects consistent with a role for calpain in release of focal adhesions [9]. An independently derived Capn4 knockout, involving a more extensive deletion of the gene, resulted in an earlier embryonic lethality, apparently at a pre-implantation stage [32]. The different times of embryonic lethality suggested that the first reported Capn4-/- mice [31] were targeted with a hypomorphic mutation, which retained some small level of calpain activity, allowing for their survival to mid-gestation, while the second reported Capn4-/- mice [32] represented a true null mutation. Disruption of Capn1, encoding the μ-calpain catalytic subunit, was subsequently reported to result in fertile, viable mice with some mild defects in the μ-calpain rich platelets relating to their aggregation and clot retraction [33]. The fact that Capn4 null mice die during embryogenesis indicates that at least one of the ubiquitous calpains is essential for development to term. The viability of Capn1-deficient mice does not however distinguish between two possibilities: either that m-calpain is specifically required during embryogenesis, or that either form of calpain alone is sufficient and can compensate for the absence of the other. To resolve this question, we have now knocked out the Capn2 gene encoding the m-80 k subunit in mice. We report here that Capn2 null embryos died prior to the implantation stage, indicating that m-calpain is indispensable for early embryogenesis. This role cannot be fulfilled by μ-calpain, which is expressed in embryonic stem (ES) cells [31] and is assumed to be present at this stage of gestation. This demonstrates unequivocally that m-calpain and μ-calpain have distinct physiological roles during early embryogenesis.