Ex vivo imaging of active caspase 3 by a FRET-based molecular probe demonstrates the cellular dynamics and localization of the protease in cerebellar granule cells and its regulation by the apoptosis-inhibiting protein survivin

Abstract

Background: Apoptosis takes place in naturally occurring neuronal death, but also in aging, neurodegenerative disorders, and traumatic brain injuries. Caspase 3 (Casp3) is the most important effector protease in apoptosis: being inactive inside the cell, it undergoes enzymatic cleavage and - hence - activation once the apoptotic cascade is triggered. Immunological techniques with antibodies against cleaved Casp3 (cCasp3) or assays with colorimetric/fluorogenic substrates are commonly in use, but they do not allow to directly follow the dynamics of activation in alive neurons that may be committed to die.

Results: By combined biolistic transfection, confocal microscopy, and fluorescence resonance energy transfer (FRET), we have implemented a methodology to dynamically monitor Casp3 activation in organotypic cerebellar slices from postnatal mice. After transfection with pSCAT3 FRET probes, we measured the ratio of the emissions of the donor/acceptor pair (ECFPem/Venusem) in fixed or alive cultures. In so doing, we i. discriminated the cellular compartment(s) of enzyme activation (nucleus, perikaryon, neurites); ii. demonstrated that Casp3 was constitutively active in the granule cells; iii. followed the fluctuations of ECFPem/Venusem, and its response to 25 mM KCl depolarization, or to increased intracellular Ca(++) after NMDA (1 mM), kainic acid (1 mM), or A23187 (100-200 μM). The specificity of the active pSCAT3-DEVD probe was confirmed with RNA interference and after inhibition of Casp3 with Ac-DEVD-CMK (100 μM), as both sets of experiments brought ECFPem/Venusem to the values recorded with the control probe pSCAT3-DEVG. After double-transfection with pSCAT3-DEVD + pHcRed1-C1-survivin, we also showed a 44-56% reduction of basal Casp3 activity in cells overexpressing survivin, a protein-member of the family of apoptosis inhibitors, with augmented survival (2.82 folds). Survivin-rescued cells were sensitive to 5 mM H2O2 oxidative stress but died without intervention of Casp3.

Conclusions: This ex vivo FRET-based methodology provides quantitative information on the functional and histological dynamics of Casp3 activation in individual neurons at a cell level resolution. Not only it can be combined with experimental manipulation of the apoptotic machinery inside the cell, but offers several advantages over existing protocols for monitoring apoptosis in live mammalian neurons, and has potential to be transferred in vivo. Due to the pivotal role of Casp3 in apoptosis, our approach is relevant for a better comprehension of molecular neurodegeneration in the normal and pathological brain.

Keywords: Apoptosis; Biolistic transfection; Caspase 3; Cerebellum; Confocal microscopy; FRET; Live imaging; Neurons; Organotypic cultures; Survivin.

Fig. 1

Visualization of Casp3 activation in fixed OCCs after biolistic transfection. a Low magnification image of a double-transfected OCC (pSCAT3-DEVD + pHcRed1-C1) after excitation with the 588 nm argon laser line. HcRed1 expression permits an easy visualization, localization, and identification of successfully transfected cells. The red-dotted line indicates the border of the culture. b-g Exemplificative images of two CGCs in the IGL (b-d) and two CGCs (e-g) in the EGL after pSCAT3-DEVD transfection showing the emissions of the FRET pair at 475 nm (ECFP) and 530 nm (Venus). The cell at right in b-d is a CGC in the vertical bipolar stage of migration and displays a well visible axon (asterisk) that bifurcates to give origin to a parallel fiber. The two cells in e-g are CGCs at the horizontal bipolar stage of migration. The cell at right displays some enlargements of its processes with high Casp3 activity (arrowheads). Note that to better show the distribution of ECFP and Venus images are taken at different laser excitation powers. As an example, the true fluorochrome emissions during FRET recording are shown in black and white in the inserts of panels b and c. In d and g cells are imaged in pseudocolor using a logarithmic scale to display the ECFP_em/Venus_em ratio. Note the cellular resolution of the FRET probe. h Combined ICC for the marker NeuN (green channel) and biolistic transfection with pHcRed1-Surv (red channel) shows two transfected CGCs in the IGL. Both cells are in the vertical bipolar stage and their axons have been labeled by the asterisks. Image has been modified and reproduced with permission from [29]. i-k Combined ICC for cCasp3 (red channel) and biolistic transfection with pSCAT3-DEVD (green channel) after induction of apoptosis with 1 mM NMDA for 48 h shows several cCasp3 immunoreactive cells. The pattern of cellular localization of the 17/19 kDa fragment of the protease is different among cells, one of which (arrow) displays a highly condensed cCasp3 positive nucleus. The larger cell transfected with pSCAT3-DEVD displays cytoplasmic cCasp3 immunoreactivity, but the nucleus (arrowhead) is negative. Abbreviations: cCasp3 = cleaved caspase 3; CGC = cerebellar granule cell; ECFP = enhanced cyan fluorescent protein; EGL = external granular layer of forming cerebellar cortex; IGL = internal granular layer of forming cerebellar cortex; NeuN = nuclear nuclei antigen; NMDA = N-methyl-D-aspartate; Surv = survivin; Venus = mutated yellow fluorescent protein. Scale bars: A = 500 μm; b-k: 10 μm

Fig. 2

FRET measurements of Casp3 activity in fixed OCCs and specificity controls. a calculation of FRET_eff of the SCAT3 probes by acceptor photobleaching (experiment #3). There is a statistically significant difference in the ECFP_em/Venus_em ratio before and after photobleaching of pSCAT3-DEVG. The mean value of FRET_eff was 0.22 ± 0.02 (see text and Additional file 1 for further information). b A comparison of the ECFP_em/Venus_em mean values after transfection with pSCAT3-DEVD (FRET probe) or pSCAT3-DEVG (control probe) shows a reduction of about 56 % in ECFP_em/Venus_em in cells transfected with the control probe. As in the control probe cellular Casp3 cannot cleave the FRET pair, this experiment demonstrates that the enzyme is constitutively active in CGCs in the absence of any pharmacological treatment. After transfection, OCCs were maintained in medium 2, containing Neurobasal and B27 supplement. c Specificity of the pSCAT3-DEVD probe for Casp3 is demonstrated after RNAi and Ac-DEVD-CMK inhibition. After multiple transfection with pSCAT3-DEVD and the mix of the four shRNA plasmids targeting the Casp3 gene or with pSCAT3-DEVD in the presence of the caspase inhibitor Ac-DEVD-CMK (100 μM) there is a significant reduction in the mean value of ECFP_em/Venus_em in comparison to that measured from OCCs transfected with pSCAT3-DEVD alone (pastel blue bars). There are notably no differences in mean values of ECFP_em/Venus_em when the pSCAT3-DEVG control probe was employed (blue bars), with the exception of OCCs transfected with a combination of pSCAT3-DEVG and the RNAi control clone [pSCAT3-DEVG 0.33 ± 0.01 (294 cells); pSCAT3-DEVG + RNAi control 0.27 ± 0.02 (100 cells); P = 0.01]. This result is somewhat puzzling as pSCAT3-DEVG is insensitive to Casp3 cleavage. It is also worth noting that after co-transfection with pSCAT3-DEVD + RNAi control values of ECFP_em/Venus_em displayed higher variance (0.21) and, consequently, SEM was much higher (0.15 – see bars in graph) than in all other tested conditions. To explain these data we can only speculate that transfection with a Casp3-unrelated RNAi control clone of unknown specificity (such as that provided by manufacturer) may somehow interfere with protein synthesis machinery in cells. d TCD after RNAi and Ac-DEVD-CMK inhibition of Casp3. There is a statistically significant increase of TCD after multiple transfection with shRNA plasmids that epigenetically inhibit the casp3 gene together with pSCAT3-DEVD (pSCAT3-DEVD 13.13 ± 1.80, n. cells = 584; pSCAT3-DEVD + RNAi 33.89 ± 4.09, n. cells = 840; P = 0.000926) or pSCAT3-DEVG (pSCAT3-DEVG 18.84 ± 3.72, n. cells = 941; pSCAT3-DEVG + RNAi 27.37 ± 3.91, n. cells = 1391; P = 0.04279), but not in RNAi control experiments and after Ac-DEVD-CMK [n. cells = 213 (pSCAT3-DEVD + RNAiCONTR), 585 (pSCAT3-DEVD+ Ac-DEVD-CMK), 133 (pSCAT3-DEVG + RNAiCONTR), 794 (pSCAT3-DEVG+ Ac-DEVD-CMK)]. In these experiments, we have counted a total of 5,841 transfected cells and measured their areas. Mean area was 165.71 ± 2.16 μ², corresponding to a mean diameter of 15 μm for a circular object. e Transfection with pSCAT3-DEVD or pSCAT3-DEVG demonstrates the existence of a subpopulation of CGCs resistant to induction of apoptosis by ionotropic glutamate receptor agonists and A23187 after measurement of mean values of ECFP_em/Venus_em (pSCAT3-DEVD 0.59 ± 0.04; n. cells = 251; pSCAT3-DEVD + KA 0.30 ± 0.01; n. cells = 103; pSCAT3-DEVD + NMDA 0.34 ± 0.01; n. cells = 105; pSCAT3-DEVD + KA + NMDA 0.28 ± 0.01; n. cells = 62; pSCAT3-DEVD + A23187 0.27 ± 0.01; n. cells = 132; pSCAT3-DEVG 0.33 ± 0.01 n. cells = 294; pSCAT3-DEVG + KA 0.29 ± 0.01; n. cells = 98; pSCAT3-DEVG + NMDA 0.33 ± 0.01; n. cells = 80; pSCAT3-DEVG + KA + NMDA 0.29 ± 0.01; n. cells = 68; pSCAT3-DEVG + A23187 0.27 ± 0.005; n. cells = 7). Error bars = SEM. *P-value 0.05–0.01 ** P-value < 0.01–0.001 % *** P-value < 0.001 %

Fig. 3

Exemplificative images of the different morphologies of pSCAT3-DEVD and pHcRed1-C1-survivin double-transfected cells. a-d A CGC with the typical vertical bipolar stage morphology in the IGL. The arrow-heads indicate the axon and its branching into the parallel fiber. e-h A presumptive Golgi cell in the IGL can be identified as being larger and displaying a multipolar morphology. i-l Two CGCs in the IGL display the typical features of early apoptosis: retraction of processes and a tendency to cell shrinkage. Note that to better show the distribution of ECFP and Venus as well as the cell morphology, images are taken at different laser excitation powers. Abbreviations: CGC = cerebellar granule cell; ECFP = enhanced cyan fluorescent protein; IGL = internal granular layer of forming cerebellar cortex; Surv = survivin. Scale bars: 10 μm

Fig. 4

a-b Effect of survivin overexpression on the ECFP_em/Venus_em values in OCCs cultivated in serum-containing medium (a) or Neurobasal + B27 supplement (b). In both culture conditions, survivin overexpression significantly reduces the constitutive activity of Casp3 in CGCs. c Survivin overexpression results in a statistically significant increase in the intensity of 618 nm fluorescence emission (in arbitrary units) of HcRed1 (IF_emHcRed1) in CGCs transfected with pHcRed1-C1 or pHcRed1-C1-survivin. d-e Correlation between the intensity of 618 nm fluorescence emission (in arbitrary units) of HcRed1 (IF_emHcRed1) and ECFP_em/Venus_em in CGCs double transfected with pSCAT3-DEVD and pHcRed1-C1 (d) or pSCAT3-DEVD and pHcRed1-C1-survivin (e). Each dot represents a transfected cell. Note that IF_emHCRed1is not correlated with Casp3 activity in cells transfected with the survivin control vector (d), whereas IF_emHCRed1, and hence survivin level inside the cell (see text), is inversely correlated to Casp3 activity when CGCs are engineered to overexpress survivin (e). f-g Effect of H₂O₂ oxidative stress (5 mM) on the density of CGCs (f) or on the ECFP_em/Venus_em value (g) after double transfection with pSCAT3-DEVD and pHcRed1-C1, or pSCAT3-DEVD and pHcRed1-C1-survivin. Note that oxidative stress has no effect on either the number/area of transfected cells or the ECFP_em/Venus_em values when the cellular levels of survivin are not genetically manipulated. Survivin overexpression results in a statistically significant increase in cell density, and reduced, albeit not significantly, the ECFP_em/Venus_em values after H₂O₂. Error bars = SEM. ** P-value 0.01–0.001 % *** P-value < 0.001 %

Fig. 5

FRET measurements of Casp3 activity in live OCCs. a-b Repetitive imaging with a 40x lens results in an artifact increase of the ECFP_em/Venus_em values recorded from the same cells in a 24-h interval. a progressive increase in the ECFP_em/Venus_em ratio in six CGCs monitored at subsequent intervals up to 24 h. b comparison of the ECFP_em/Venus_em ratios in two groups of cells undergoing single (left - blue dots) or repeated (right - red dots) laser excitation to capture FRET pair images. c Imaging with a 20x lens and repetitive laser excitations did not significantly alter the ECFP_em/Venus_em ratio in seven CGCs that were photographed up to ten times to acquire FRET pair images. d-f Exemplificative traces of the fluctuations in the ECFP_em/Venus_em ratio in healthy (d), suffering (e) or dying/dead (f) CGCs during a four-hour follow up. FRET was measured in these eleven cells during the course of the same experiment. Cell 2 and 3 in f died soon after one hour from start, and could not be subsequently identified when the culture was scanned for the acquisition of images at 120 m. It should be noted that live imaging can be made difficult by X-Y axis and focus (Z axis) drifting of the sample. If in-focus images cannot be captured at all time-points, cell(s) must be discharged from subsequent analysis. g = Effect of 60 mM KCl depolarization on the ECFP_em/Venus_em ratio during a three-hour follow up (see text for further explanation). h Exemplificative traces of the fluctuations in the ECFP_em/Venus_em ratio during a three-hour follow up in eight CGCs challenged with 100 mM H₂O₂. i Lack of a statistically significant response to 100 mM H₂O₂ in a subpopulation of CGCs that appeared to be resistant to oxidative stress (see text for discussion). Error bars = SEM. * P-value 0.05–0.01 %