Growth factor-induced calpain activity under conditions that do not permit increased intracellular calcium. WT NR6 murine fibroblasts were incubated with growth factors (EGF and PDGF) in the presence and absence of BAPTA-AM or calcium free buffer. These treatments were identical to those under which intracellular calcium levels were measured. To measure calpain activity, the cells were subsequently loaded with the substrate Boc-LM-CMAC. Calpain activity is indicated by an increase in fluorescence (images are representative; n = 3).

Calpain activity in S50E m-calpain-expressing cells. (A) As in Fig. 5, control cells possessed the S50E plasmid but were not induced to express it (no dexamethasone) and were not treated with murine m-calpain antisense oligonucleotides. Cells induced to express S50E m-calpain in the presence of endogenous m-calpain down-regulation show increased calpain activity even in the absence of EGF (indicated by increased fluorescence). CI-1 blocks the activity of S50E m-calpain. However, the MEK inhibitor PD98059, which would normally block EGF-induced calpain activity, has no effect (images are representative; n = 3). (B) Cells expressing the S50E-ER hybrid construct were examined for calpain activity. In the absence of induction, tamoxifen (1 μM) did not trigger cleavage of Boc-LM-CMAC. After induction of the S50E-ER construct, tamoxifen was required to note fluorescence. Shown are images taken 60 min after addition of tamoxifen to the cells (images are representative; n = 3).

Effect of intracellular calcium chelation by BAPTA-AM and removal of medium calcium on EGF- and PDGF-induced increases in intracellular calcium. (A) Percentage of cells that demonstrated an increase in intracellular calcium in response to stimulus (i.e., EGF, PDGF, or bradykinin). This response does not occur in cells pretreated with BAPTA-AM, and the response is attenuated in cells perfused in calcium-free buffer. (B) Average increases in intracellular calcium level in response to stimulus (arbitrary fluorescence ratio units) for cells that demonstrated a rise in intracellular calcium levels ± standard errors of the means. The maximum response (bradykinin) was estimated at 500 nM by comparison to ionophore treatment (∼1.3 ratio units; data not shown). The intracellular calcium concentration of BAPTA-AM-treated cells is shown for comparison, although none of these cells were classified as responding. Numbers of cells and numbers of experiments for the different treatments are provided in Materials and Methods (“Experimental analysis”).

S50A m-calpain phosphorylation in vitro and in vivo. Expression of histidine-tagged m-calpains (wild type and S50A) was induced with dexamethasone. (A) Wild-type (WT) and S50A mutant calpains were purified by nickel affinity chromatography. Equal amounts of the purified proteins (as shown by antihistidine blotting) were incubated with activated ERK (300 U) in the presence of [γ-³²P]ATP. S50A m-calpain was radiolabeled to a lesser extent than wild-type m-calpain (n = 2). (B) Quiescent cells expressing wild-type and S50A histidine-tagged calpains were labeled with [³²P]orthophosphate for 2 h prior to EGF treatment. The exogenously encoded calpains were purified by nickel affinity chromatography, and phosphorylated m-calpains were detected by autoradiography. EGF induced phosphorylation of wild-type m-calpain but not S50A m-calpain in vivo (n = 2). Shown are representative antihistidine blots to demonstrate equal loading. The numbers below the bands are relative densities (B) or percentages of wild-type levels (A) that were determined with National Institutes of Health Image.

ERK association with m-calpain in vivo. Coimmunoprecipitation of ERK and m-calpain was performed in the presence of the cross-linking reagent DTBP. WT NR6 murine fibroblasts were treated with cross-linking reagent for 30 min, and some cells were treated with 10 mM CI-1 for 30 min, followed by EGF treatment (controls were not treated with EGF). The cross-linker was then activated with light (365 nm) for 20 min. The cells were lysed and incubated with anti-m-calpain (A and G), anti-pan-ERK (B, C, and F), anti-μ-calpain (D), anti-integrin-linked kinase 1 (ILK1; E), and nonspecific isotype immunoglobulin G (IgG; B, E, F, and G) antibodies, and then the antibody-bound proteins were removed with protein G-agarose beads. The immunoprecipitated (IP) proteins were then isolated and separated on an SDS-10% polyacrylamide gel. WT NR6 whole-cell lysates (A, C, and D) or IgG-immunoprecipitated proteins (B, E, F, and G) were run alongside as a control After transfer to a PVDF membrane, the anti-m-calpain immunoprecipitates were probed for pan-ERK (A), followed by MEK1 (A) (same blot), talin (G), and α-actinin (G) and the anti-ERK immunoprecipitates were probed for m-calpain (B), μ-calpain (C) (separate blots), and also talin (F). Anti-μ-calpain immunoprecipitates were probed unsuccessfully for ERK as an additional control (D). Shown are representative blots from at least two replicates of each.

S50A mutant m-calpain resistance to ERK-induced activation and activity in low and high calcium concentrations. (A) The activity of in vitro-transcribed and -translated wild-type and mutant (S50A) m-calpains was measured against that of recombinant m-calpain (Calbiochem) to determine their ability to cleave the calpain substrate Tau (0.5 μg). Addition of ERK to the in vitro-translated wild type, the T381A mutant m-calpain, and recombinant m-calpain showed equivalent abilities to cleave Tau within 10 min, as shown by loss of the Tau band at 65 kDa. On the other hand, the S50A mutant m-calpain exhibits a significant reduction in ERK activation, as shown by reduced loss of Tau (n = 3). Of the m-calpains tested none were shown to be able to cleave Tau alone (NT) or in the presence of 2 μM calcium (Ca⁺⁺) during a 10-min incubation. (B) Recombinant m-calpain is activated by 2 μM calcium over time. Early time points do not demonstrate noticeable levels of Tau cleavage; however, by 60 min all Tau is cleaved. Wild-type in vitro-translated m-calpain demonstrated a similar time course of calcium-induced activity (not shown). In vitro-translated mutant S50A m-calpain was also activated by calcium, but only at a 100-fold-higher level (200 μM); at 2 μM calcium no loss of Tau was noted over the 60-min time period (data not shown) (n = 3).

ERK activation and phosphorylation of m-calpain in vitro and in vivo. (A) Purified activated ERK (300 U) and calpain (10 μg) enzymes were incubated together with or without calcium chloride (2 μM). The presence of ERK increased m-calpain activity over fivefold (*, P < 0.01). This occurred in both the presence and absence of calcium (error bars, standard errors of the means; n = 4). (B) As a control, the upstream enzyme MEK was substituted for ERK and had no effect on m-calpain activity (n = 4). (C) The in vitro activation assay was repeated in the presence or absence of ERK (300 U) and calcium chloride (2 mM). The presence of ERK or calcium increased m-calpain activity over threefold (*, P < 0.01), but there was no difference between these states (n = 4). This series of experiments were performed with different batches of m-calpain, and this may be the reason for higher levels of ERK-induced activation in panel A. (D) Adding [γ-³²P]ATP to the mixture of ERK and calpain enzymes caused radiolabeling of the large subunit of m-calpain, whereas MEK did not cause radiolabeling (n = 3). (E) In vivo, [³²P]orthophosphate labeling of WT NR6 murine fibroblasts demonstrates increased radiolabeling of histidine-tagged m-calpain upon EGF stimulation. This increase was eliminated by pretreatment with the MEK inhibitor PD98059 (2 μM), suggesting that the increase in phosphorylation seen with EGF treatment is mediated by ERK (n = 3). An antihistidine blot is shown to demonstrate equal protein levels. In panels D and E representative blots are shown.

Expression of the S50A m-calpain mutant inhibits EGF-induced motility and deadhesion in fibroblasts. (A) WT NR6 cells expressing the S50A mutant m-calpain were analyzed for their ability to migrate into a denuded space. The cells were pretreated for 24 h with either antisense oligonucleotides against murine (mouse AS) or human (human AS) m-calpain to down-regulate either the endogenous murine or the exogenous human m-calpain. A central swath was denuded in the confluent layer of cells, which were then treated with or without EGF (10 nM) in the presence of antisense oligonucleotides (20 μM) and incubated for 24 h. The area of the wound was measured, and the decreased area from the control (mouse antisense treatment only) was subtracted as background. The decreased area from the cells treated with human antisense oligonucleotides and EGF was set to 100%, and results for the other treatments are percentages of this value. The exogenous wild-type (WT) human m-calpain demonstrated the same EGF-induced motility as the endogenous murine m-calpain (P was not significant). However, the S50A mutant m-calpain showed a significant decrease in motility compared to the endogenous murine m-calpain (P < 0.05; n = 3). (B) WT NR6 cells stably expressing wild-type (black bars) or the S50A (striped bars) human m-calpain were incubated with antisense oligonucleotides against murine m-calpain and then treated with or without CI-1 and/or EGF. The cells were then inverted and centrifuged (1,643 × g) for 5 min at 37°C. The adherent cells were counted before and after centrifugation. The number of remaining cells are expressed as a percentage of the starting number of cells. Cells expressing wild-type human m-calpain and treated with EGF showed an enhanced ability to detach from the substratum compared to the S50A-expressing cells (P < 0.01; n = 6). Cells treated with diluent or not treated were incubated with PBS, the vehicle for EGF.

EGF-induced calpain activity in S50A m-calpain-expressing cells. (A) Activity of S50A m-calpain was measured in the face of down-regulation of endogenous murine m-calpain. Antisense oligonucleotides were designed to be specific for murine m-calpain. These did not block the expression of S50A m-calpain, which is human in origin. Antisense treatment in the presence of EGF for 12 h was adequate to down-regulate endogenous m-calpain (lanes 3 and 4 [from the left]), and exogenous expression of S50A was not blocked (lanes 5 to 7; n = 3). (B) Calpain activity was measured in cells expressing S50A m-calpain with the substrate Boc-LM-CMAC (5 μM). Cells that possessed the S50A plasmid but that were not induced to express it were used as a control (no dexamethasone). Control cells were not treated with murine m-calpain antisense oligonucleotides. All other cells were induced with dexamethasone (2 μM) for 18 to 24 h and treated with murine m-calpain antisense oligonucleotides for the same time period. This produced conditions under which the activity of the S50A m-calpain construct could be measured. Control cells show normal EGF-induced calpain activity (indicated by increased fluorescence, equivalent to lightness of tone in micrographs). S50A-expressing cells do not exhibit EGF-induced calpain activity, and CI-1 and the MEK inhibitor PD98059 have no additional effects (images are representative; n = 3). (C) Calpain activity was determined by a calpain activity assay kit (BioVision). WT NR6 cells expressing wild-type (WT) or S50A human m-calpain-His were pretreated for 24 h with either antisense oligonucleotides against murine (mouse AS) or human (human AS) m-calpain to down-regulate either the endogenous murine or the exogenous human m-calpain and then treated with or without CI-1 and/or EGF. The cells were lysed, and the lysates were analyzed for calpain activity. The results from this assay correlate with those from the Boc assay showing that the mutation of serine 50 to alanine significantly reduces EGF-induced m-calpain activity (P < 0.05 for endogenous mouse antisense down-regulation compared to human antisense down-regulation). Cells treated with diluent were incubated with PBS, the vehicle for EGF.