Knockdown of m-calpain increases survival of primary hippocampal neurons following NMDA excitotoxicity

Abstract

The calpain family of cysteine proteases has a well-established causal role in neuronal cell death following acute brain injury. However, the relative contribution of calpain isoforms to the various forms of injury has not been determined as available calpain inhibitors are not isoform-specific. In this study, we evaluated the relative role of m-calpain and mu-calpain in a primary hippocampal neuron model of NMDA-mediated excitotoxicity. Baseline mRNA expression for the catalytic subunit of m-calpain (capn2 ) was found to be 50-fold higher than for the mu-calpain catalytic subunit (capn1) based on quantitative real-time PCR. Adeno-associated viral vectors designed to deliver short hairpin RNAs targeting capn1 or capn2 resulted in 60% and 90% knockdown of message respectively. Knockdown of capn2 but not capn1 increased neuronal survival after NMDA exposure at 21 days in vitro. Nuclear translocation of calpain substrates apoptosis inducing factor, p35/p25 and collapsin response mediator protein (CRMP) 2-4 was not detected after NMDA exposure in this model. However, nuclear translocation of CRMP-1 was observed and was prevented by capn2 knockdown. These findings provide insight into potential mechanisms of calpain-mediated neurodegeneration and have important implications for the development of isoform-specific calpain inhibitor therapy.

Figure 1

Calpain knockdown in Rat2 fibroblasts. (A) Rat2 fibroblassts were transfected with three different short interfering RNA (siRNA) sequences targeting the catalytic subunit of μ-calpain (capn1). All three capn1 sequences (lanes A, B and C; see Table 1) reduced expression of μ-calpain compared to controls exposed to transfection reagent alone (Veh), and none of the sequences altered expression of m-calpain. (B) Similarly, three sequences targeting the catalytic subunit of m-calpain (capn2; lanes D, E, and F; see Table 1) decreased expression of that isoform without altering expression of μ-calpain. (C) Sequences B (capn1) and F (capn2) were converted into short hairpin RNA (shRNA) and cloned into a plasmid-based expression system. Transfection of rat2 fibroblasts demonstrated that shRNA reduced expression of the targeted isoform as compared to naïve cells or those transfected with scrambled shRNA. As in the siRNA experiment, knockdown was isoform-specific.

Figure 2

Transduction of primary hippocampal neuron cultures with adeno-associated viral vectors. (A) AAV2/1 vectors expressing shRNA targeting the catalytic subunit of μ-calpain (capn1) and m-calpain (capn2) were able to transduce both neurons and glia. Transduction was identified by expression of ZsGreen and cell type determined by immunolabeling for MAP2 (neurons) or GFAP (glia). (B) Transduction efficiency was determined by dividing the number of ZsGreen expressing cells over the total number of DAPI-labeled nuclei. Vector expressing luc shRNA transduced 96.3±0.4% of cells, capn1 vector transduced 99.1±0.7% of cells and capn2 vector transduced 99.1±1.0% of cells. (C) The majority of cells in the cultures were neurons, as indicated by the fact that 77.8±7.5% of luc-transduced cells, 70.2 ± 8.1% of capn1-transduced cells and 77.6 ± 4.0% of capn2 transduced cells were neurons (mean ± standard deviation). There was no significant difference in the percentage of transduced cells that labeled for the neuronal marker MAP2 between the two conditions (One-way ANOVA, p > 0.05; error bars indicate standard error of the mean).

Figure 3

Calpain isoform knockdown in primary hippocampal neurons. (A) Primary hippocampal neurons were transduced with 1.5×10¹¹GC of vector expressing luc, capn1 or capn2 shRNA. Western blot performed 2 weeks after transduction demonstrates that capn2 shRNA reduces expression of m-calpain, while transduction with luc or capn1 shRNA does not. Expression of μ-calpain could not be detected by Western blot in 20μg of cell lysate from these cultures, even though it can be detected in an equivalent amount of Rat2 fibroblast lysates. (B) Densitometric analysis indicates that capn2 shRNA significantly lowers expression of m-calpain, while transduction with luc and capn1 shRNA has no effect (One-way ANOVA with Scheffe post-hoc analysis, *p < 0.05). (C) Quantitative real-time PCR for calpain 1 and 2 message shows that in control luc shRNA-treated cultures, calpain 2 message is approximately 50-fold more abundant than calpain 1 message. Transduction with capn2 shRNA significantly reduces expression of calpain 2 message compared to luc shRNA control (*p < 0.05, One-way ANOVA with Scheffe post-hoc analysis). (D) Expanded view of calpain 1 message levels. Both capn1 and capn2 shRNA significantly reduced expression of calpain 1 message (*p < 0.05, One-way ANOVA with Scheffe post-hoc analysis), however the physiologic relevance of this knockdown is unknown given the extremely low baseline levels of calpain 1 mRNA.

Figure 4

Characterization of NMDA injury model. (A) Primary hippocampal neurons were exposed to 10μM, 100μM, 200μM, 400μM or 800μM NMDA or HBS vehicle for 30 minutes. Twenty-four hours after injury, live cells were labeled with calcein-AM and dead cells with PI. The percentage of surviving cells was significantly reduced in cultures treated with NMDA concentrations of 200μM and higher (*p < 0.05, One-way ANOVA with Scheffe post-hoc analysis). No additional increase in injury was observed at concentrations greater than 200μM, so this concentration was used for all subsequent studies. (B) Primary hippocampal neurons were treated with 1.0μM MDL-28170 and subsequently injured with 200μM NMDA as described. Twenty-four hours after injury, cells were stained with DAPI to label nuclei and stained with propidium iodide to identify dead cells. Treatment with the calpain inhibitor MDL-28170 (1.0μM MDL) significantly increased cell survival compared to injured cultures treated with vehicle (Student's t-test *p < 0.05).

Figure 5

Calpain activity in neurons transduced with AAV2/1 expressing luc, capn1 or capn2 shRNA. (A) Calpain activity was assessed by immunolabeling cells for calpain-cleaved spectrin (Ab38, red) after 30 minutes of NMDA exposure. Calpain activity was observed in cells transduced with all three shRNA sequences, however the distribution in capn2 shRNA treated cultures was primarily limited to processes, while luc and capn1 shRNA treated cultures displayed labeling both in the soma and processes. (B) Quantification of the percentage of ZsGreen expressing cells that displayed any labeling for calpain-cleaved spectrin. NMDA injury (black bars) increased the percentage of transduced cells with calpain activity in all three transduction conditions as compared to vehicle treated controls (grey bars). No significant difference in the percentage of cells with calpain activity was observed between injured cultures transduced with control (luc), capn1 or capn2 shRNA (One-way ANOVA with Scheffe post-hoc analysis, p > 0.05). (C) Quantification of the percentage of Ab38 positive cells with labeling in the cell body. A significantly smaller percentage cells transduced with capn2 shRNA immunolabeled for calpain activity in the cell body (One-way ANOVA with Scheffe post-hoc analysis, *p < 0.05).

Figure 6

Cell survival after NMDA injury. (A) Twenty-four hours after NMDA exposure, cultures transduced with AAV2/1 vector expressing luc (luc NMDA) or capn1 shRNA (capn1 NMDA) displayed abundant PI labeling and morphologic signs of cell death, including swelling and breakdown of processes. In contrast, cultures transduced with AAV2/1 vector expressing capn2 shRNA (capn2 NMDA) did not differ qualitatively from control cultures treated with HBS (luc HBS). (B) Quantification of cell survival was performed by calculating the ratio of PI-negative, ZsGreen expressing cells to the total number of ZsGreen expressing cells. Survival in uninjured cultures did not differ between transduction conditions. Cultures transduced with luc and capn1 shRNAs displayed a significant decrease in cell survival following NMDA injury (black bars). The cell death associated with injury in these two transduction conditions did not differ from that observed following NMDA injury in non-transduced controls. In cultures transduced with AAV2/1 vector expressing capn2 shRNA, survival did not differ between vehicle (grey bar) and NMDA treatment (black bar). Furthermore, capn2 knockdown significantly increased survival after NMDA exposure compared to naïve, luc and capn1 shRNA transduced cultures (One-way ANOVA with Scheffe post hoc analysis, * p < 0.05 vs. naïve NMDA, luc NDMA and capn1 NMDA).

Figure 7

Translocation of calpain substrates following NMDA exposure. Hippocampal cultures were injured by application of 200μM NMDA for 30 minutes. Immediately following injury or 6 or 24 hours later, cultures were fixed and immunolabeled for the calpain substrates AIF, p35, CRMP-1, CRMP-2, CRMP-3 and CRMP-4. Only CRMP-1 displayed differential cellular localization following injury. Uninjured cultures immunolabeled for CRMP-1 in the cytoplasm, with staining excluded from the nucleus. Following 30 minutes of NMDA exposure, CRMP-1 immunoreactivity was visible throughout both the cell body and nucleus. Nuclear CRMP-1 labeling remained at 6 hours after injury and was visible in dead cells (as identified by PI labeling) 24 hours after NMDA application.

Figure 8

Effect of m-calpain knockdown on CRMP-1 translocation. Primary hippocampal neurons were transduced with AAV2/1 vector expressing luc, capn1 or capn2 shRNA and injured by application of 200μM NMDA two weeks later. Twenty-four hours after injury cells were fixed and immunolabeled for CRMP-1. A significantly smaller percentage of cells transduced with capn2 shRNA displayed nuclear CRMP-1 labeling compared to luc and capn1 shRNA transduced cultures (One-way ANOVA with Scheffe post-hoc analysis, *p < 0.05).