There is no single affinity chromatography method that is universally effective for the purification of calpains. Dye affinity chromatography and calcium-dependent binding to immobilized substrates or inhibitors are the most commonly attempted affinity methods. Our laboratory has used Reactive Red-agarose in the purification of m-and μ-calpain from bovine heart and other tissues for many years (1,2). However, this method has not achieved widespread use due to reported variability in outcomes. The procedure provided below has proven to be highly reproducible. It is particularly effective for the purification of m-calpain, although it is also useful for μ-calpain. Others have described the use of Blue Sepharose-CL-6B for the purification of both m-and μ-calpain, requiring the use of 1 M urea in the elution buffers (3). Our unpublished attempts to use this type of resin were not reproducibly successful, and this method will not be discussed further here. Calcium-dependent binding of calpain either to immobilized substrates, like casein-Sepharose (4,5), or to the inhibitor calpastatin (6), is highly selective but has the inherent risk that autoproteolysis may alter the calpain while it is being purified. Inactivation of calpain through mutation (7) or with covalent inhibitors (8) can be used to avoid autoproteolysis, although the uses for the purified inactive calpain are more limited.KeywordsUnbind ProteinBovine HeartDEAE ColumnCoupling BufferCovalent InhibitorThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

The absolute requirement of Ca(2+) for proteolytic activity is a feature unique to the calpains, a family of heterodimeric cysteine proteases. Conditions are described which give rise to diffraction-quality crystals of m-calpain in two crystal forms, P1 and P2(1). Data have been collected from native crystals of m-calpain in both P1 and P2(1) forms, to 2.6 and 2.15 A, respectively. Selenomethionine-containing crystals have been grown in both forms, and anomalous data from the P2(1) selenomethionine enzyme provided the location of 17 of the 19 Se atoms in the protein.

The hypothesis that calpain subunits dissociate in the presence of Ca2+ has been tested by methods which avoid interference by Ca2+-induced aggregation and large subunit autolysis. Inactive Cys105Ser-m-calpain, bound either to Ni-NTA-agarose or to immobilized casein, after incubation with Ca2+, could be recovered in high yield as a heterodimer. Natural bovine m-calpain, after irreversible inhibition with Z-LLY-CHN2, also bound to immobilized casein and was eluted as a heterodimer. The Ca2+ requirements of calpain containing a small subunit with EF-hand mutations were higher, both before and after autolysis, than those of wild-type calpain. In mixtures of wild-type and mutant enzymes, subunit exchange did not occur in the presence of Ca2+. The results demonstrate that the subunits in both natural and recombinant m-calpain, in the given experimental conditions, remain associated in the presence of Ca2+ both before and after autolysis.

Previous studies suggest that the Ca2+-dependent proteases, calpains, participate in remodeling of the actin cytoskeleton during wound healing and are active during cell migration. To directly test the role that calpains play in cell spreading, several NIH-3T3- derived clonal cell lines were isolated that overexpress the biological inhibitor of calpains, calpastatin. These cells stably overexpress calpastatin two- to eightfold relative to controls and differ from both parental and control cell lines in morphology, spreading, cytoskeletal structure, and biochemical characteristics. Morphologic characteristics of the mutant cells include failure to extend lamellipodia, as well as abnormal filopodia, extensions, and retractions. Whereas wild-type cells extend lamellae within 30 min after plating, all of the calpastatin-overexpressing cell lines fail to spread and assemble actin-rich processes. The cells genetically altered to overexpress calpastatin display decreased calpain activity as measured in situ or in vitro. The ERM protein ezrin, but not radixin or moesin, is markedly increased due to calpain inhibition. To confirm that inhibition of calpain activity is related to the defect in spreading, pharmacological inhibitors of calpain were also analyzed. The cell permeant inhibitors calpeptin and MDL 28, 170 cause immediate inhibition of spreading. Failure of the intimately related processes of filopodia formation and lamellar extension indicate that calpain is intimately involved in actin remodeling and cell spreading.

The calmodulin binding proteins, caldesmon and calponin, are cleaved by both major isoforms of calpain in vitro. The patterns of fragments generated by each enzyme are essentially identical for a given substrate. Qualitatively, the cleavage pattern of each substrate is unchanged by the presence or absence of calmodulin suggesting that the interaction between calmodulin and these calmodulin-binding proteins does not alter substrate recognition by calpain. However, calmodulin (at microM concentrations) does have a small, but significant, inhibitory effect directly on calpain as evidenced by slower rates of cleavage of alpha-casein, a protein that does not bind calmodulin. Inhibition is more pronounced with mu-calpain (15-25%) than with m-calpain (6-10%). In order to demonstrate, unequivocally, that substrate recognition does not require an interaction between calpain and a substrate's calmodulin-binding domain, recombinant, full-length caldesmon and a mutant lacking the calmodulin binding domain were tested as substrates for calpain in the presence and absence of calmodulin. Calpain produced similar cleavage patterns of the baculovirus expressed caldesmon and the truncated mutant. Competition experiments demonstrated that calpain does not discriminate between the truncated mutant and full length caldesmon. This suggests that substrate recognition by calpain was not altered significantly by the absence of the calmodulin-binding domain. Cleavage of a second calmodulin-binding protein, calponin was also examined. The rate of calponin cleavage was increased in the presence of calmodulin, an observation that is also inconsistent with any requirement for calpain to bind to its calmodulin-binding site. These results demonstrate that calmodulin-binding domains do not provide substrate recognition sites for calpains. It seems likely that the calmodulin-like regions of calpain function to bind calcium and to regulate enzyme conformation as required for activity and that they do not interact directly with most substrates.

Death of dystrophin-deficient muscle purportedly results from increases in [Ca]in that cause the activation of calpains. We have tested whether calpains play a role in this process by assaying for changes in calpain concentration and activation in peak necrotic mdx mice (4 weeks of age) and in completely regenerated mdx mice (14 weeks of age). Biochemical fractionation and immunoblotting with epitope-specific antisera allowed measurement of the concentrations of m- and mu-calpains and the extent of autoproteolytic modification. Our findings show that total calpain concentration is elevated in both 4-week and 14-week mdx mice. This increase in concentration was shown to result primarily from a significant increase in m-calpain concentration at 4 weeks. Northern analysis demonstrated that neither m- nor mu-calpain mRNA concentrations differed between mdx and controls suggesting that the increased calpain concentration results from post-translational regulation. Immunoblotting with antibodies directed against amino-terminal peptides revealed an increase in autoproteolysis of mu-calpain, indicative of increased activation. The extent of autoproteolysis of mu-calpain returns to control levels during regeneration. This is not a consequence of increased calpastatin mRNA or protein. The findings reported here support a role for calpains in both the degenerative and regenerative aspects of mdx dystrophy.

Our studies suggest that protein kinase C is involved in low calcium (Ca2+)-stimulated secretion of parathyroid hormone (PTH) but not directly in high Ca(2+)-stimulated intracellular degradation of PTH to secreted carboxyl-terminal fragments (C-PTH), an important component of Ca(2+)-regulated PTH secretion. The present study was undertaken to determine the presence of calcium-activated proteases, 84 kDa (micro)-calpain and 80 kDa (milli)-calpain, in the bovine parathyroid, and whether they could degrade PTH to C-terminal fragments. Immunocytochemistry of bovine parathyroid tissue using antibodies raised against bovine heart micro- and milli-calpain detected both isoforms of calpain. Western blotting of total bovine parathyroid cell protein prepared from primary cell cultures confirmed the presence of both isoforms of calpain, demonstrated by specific milli- and micro-calpain bands. Purified bovine PTH (bPTH) was incubated in vitro with human erythrocyte micro-calpain and the cleavage products were separated by reverse-phase HPLC. Eluant fractions were assayed with an RIA with equimolar sensitivity to C-PTH and bPTH, and peak areas integrated. Micro-calpain produced a C-PTH peak from bPTH which co-eluted with the major C-PTH secreted by parathyroid cells in culture. C-PTH production by micro-calpain, expressed as per cent area under the curve, increased from 0% in the absence of either micro-calpain or Ca2+, to 71.5% when a 5:1 molar ratio of bPTH to calpain was used. Amino acid sequencing and analysis of the immunoreactive PTH cleavage products indicated the presence of two fragments of bPTH in the C-PTH peak, bPTH47-48 and bPTH69-84. In summary, both isoforms of calpain are present in the bovine parathyroid and calpains may play a role in the Ca(2+)-dependent degradation of PTH to secreted C-terminal fragments.

The integrin β4 subunit often undergoes proteolytic cleavage within its long cytoplasmic tail to yield a characteristic protein pattern of 205, 165, and 125 kDa. The results in this study suggest that β4 cleavage often occurs during or after cell lysis, where it was readily inhibitable by calcium chelators (EDTA, EGTA) and inhibitors of cysteine proteases (E64e, leupeptin). The cleavage of β4 is catalyzed by a calpain-like enzyme because (i) it requires calcium, (ii) it is mimicked by purified milli-calpain, and (iii) it is inhibited by several calpain inhibitors including the calpain-specific inhibitor calpastatin. Within intact cells, cleavage of β4 was cell type-specific and observed only when the cells were made permeable to calcium. Substantial cell viability was retained during β4 cleavage induced by ionomycin plus calcium, indicating that cleavage within intact cells was not necessarily a consequence of cell death. However, manipulations of cells including suspension, synchronization, and stimulation with serum, phorbol esters, or other agents all failed to induce cleavage, suggesting that if cleavage is physiologically relevant, it is not easily duplicated in vitro. Analysis of multiple cell types showed a wide variation in β4 sensitivity to proteolytic cleavage, suggesting that this process might be differentially regulated depending on the cellular environment.

The peptide EKLGERDDTIPPEYRELLEKKTGV was synthesized to mimic the central consensus sequence of calpastatin, the specific, endogenous inhibitor of the calpains (EC 3.4.22.17). The peptide competitively inhibits hydrolysis of casein by either micro- or milli-calpain but does not affect the activity of other proteases. This inhibitory peptide was preferentially cross-linked to milli-calpain in the presence of calcium using the heterobifunctional cross-linking reagent m-maleimidobenzoyl-N-hydroxysuccinimide ester. Cross-linking of the peptide was blocked by calpastatin. The site of crosslinking for the peptide within milli-calpain was localized using random chemical cleavage of the enzyme-peptide complex at cysteine residues. Calpain fragments were identified as amino-terminal fragments through reactivity with a peptide-specific antiserum or as non-amino-terminal fragments through incorporation of C-14 from (CN)-C-14. Analysis of the control and cross-linked fragments, from experiments using both milli-calpain and micro-calpain, maps the chemical cross-linking site to cysteine-497 and localizes the binding site for the calpastatin-like peptide to this highly conserved region of domain III of calpains catalytic subunit.

The insulin receptor substrate 1 (IRS-1) contains at least 11 sequence motifs that are rich in proline (P), glutamic acid (E), serine (S), and threonine (T), i.e., PEST regions. Proteins with PEST regions turn over rapidly. IRS-1 is degraded rapidly in vivo upon exposure of 3T3-L1 adipocytes to insulin. The intracellular, calcium-dependent, neutral proteases known as cal-pains are one possible mechanism by which IRS-1 may be degraded. To begin to investigate this possibility, purified exogenous calpain was shown to degrade IRS-1 in cell-free extracts from basal and insulin-treated cells and rat recombinant IRS-1 in vitro. Only two proteolytic fragments could be detected. One had a mol wt of ∼79 kDa, arising from the C-terminus end, and the second had a mol wt of ∼90 kDa arising from near the N-terminus, possibly a product of the same cleavage event, since the mol wt of IRS-1 from insulin-treated cells was ∼170 kDa. These results suggest that IRS-1 may serve as a substrate for calpain in vivo, accounting for its rapid degradation.

Each subunit of calpain (EC 3.4.22.17) is proteolytically modified when the enzymes are exposed to calcium. These cleavages appear to be important for regulating the proteolytic activity and calcium-sensitivity of the proteinases. We have synthesized peptides that correspond to the sites of autoproteolytic modification within the catalytic subunit of each calpain. Polyclonal antisera raised against these peptides are highly specific for the unmodified catalytic subunit of each calpain. The antiserum specific for the N-terminal epitope of milli-calpain was used to demonstrate an inverse relationship between the presence of this N-terminal peptide and casein hydrolysis. The antiserum specific for the N-terminal epitope of micro-calpain was used to demonstrate proteolytic modification of the catalytic subunit of mu-calpain in rat erythrocytes treated with ionomycin and calcium.

Eukaryotic initiation factor 4F (eIF-4F) is a multisubunit protein that functions in the first step of the binding of capped mRNAs to the small ribosomal subunit. Its largest polypeptide component, p220, is cleaved following poliovirus infection. This is thought to inactivate eIF-4F function, thereby preventing cap-dependent initiation of translation of cellular mRNAs. In this report, we show that p220 in extracts of uninfected HeLa cells is specifically lost in the presence of calcium. The responsible activities have been partially purified and identified as the calcium-dependent, neutral, cysteine proteases calpains I and II. In addition, a third calcium-dependent activity was resolved from the calpains and also results in the loss of p220. This activity has properties similar to a transglutaminase and copurifies with tissue transglutaminase through several chromatographic steps. None of these calcium-dependent activities appears to mediate p220 cleavage in poliovirus-infected cells.

The intracellular calcium-dependent proteolysis of fodrin has been postulated to be central to the regulation of plasticity of the cortical cytoskeleton of many eukaryotic cells. The close proximity of the sites of calmodulin (CaM) binding and calcium-dependent protease I (CDP-I) cleavage in mammalian alpha-fodrin suggested that their action may be linked. In hypotonic and isotonic buffers, CDP-I proteolysis of the beta subunit of fodrin was absolutely dependent upon the presence of active CaM. The stimulation by CaM was inhibited by CaM antagonists. The rate of CDP-I proteolysis of both subunits was enhanced by CaM, while the rate of fodrin proteolysis with other proteases was not influenced by CaM. The increase in the susceptibility of fodrin to CDP-I proteolysis was half-maximal at 80 nM CaM, and maximal at 200 nM CaM. The unusual and differential susceptibility of alpha- and beta-fodrin to proteolysis by CDP-I in the absence of CaM was exploited to investigate the quaternary structure of fodrin in which only the alpha subunit was cleaved. Cleavage of the alpha subunit alone did not destroy the tetrameric form of the molecule, whereas CDP-I cleavage of both subunits rendered the molecule incapable of reforming tetramers. These results provide structural and functional evidence that CaM and CDP-I act synergistically in the regulated proteolysis of fodrin.

Fodrin (brain spectrin) binds calmodulin and is susceptible to proteolysis by calcium-dependent protease I (CDP-I, calcium-activated neutral protease I, or calpain I). Both events involve the central region of the alpha-fodrin subunit, and calmodulin binding enhances the sensitivity of fodrin to CDP-I mediated proteolysis. Fragments of fodrin, generated chemically or proteolytically, which retain calmodulin binding activity have been identified and analyzed by two-dimensional peptide mapping and by direct protein sequencing. Both CDP-I and calmodulin interact with the terminal portion of the eleventh repetitive unit in fodrin, which is at the center of the molecule. CDP-I cleavage occurs between Tyr104 and Gly105 and preserves the calmodulin binding activity of the carboxyl-terminal fragment. In contrast, chymotryptic cleavage at Trp120 reduces the ability of this fragment to bind calmodulin, and tryptic cleavage beyond Trp120 completely eliminates calmodulin binding activity. It is concluded that Ser-Lys-Thr-Ala-Ser-Pro-Trp-Lys-Ser-Ala-Arg-Leu-Met-Val-His-Thr-Val-Ala- Thr- Phe-Asn-Ser-Ile-Lys, a 24-residue peptide which bridges repeats 11 and 12 of brain alpha spectrin contains the high affinity calmodulin binding domain.