4 Results

↓42

4.1 Cloning of the adenoviral vector

↓43

To clarify the pathway of Bim, full length cDNA of BimL and BimS was amplified by PCR cloned into an adenoviral vector. The use of an adenoviral expression system gives the advantage to over express the protein of interest at similar levels for each experiment in a panel of cell lines. To regulate the expression of these BH3-only proteins, an inducible adenoviral vector based on the Tet-Off system was established in the group (Gossen and Bujard, 1992 ). This system comprises all the functional components needed on a single adenoviral vector the gene of interest, its inducible promoter, the region of the transactivator and its constitutive promoter. The basic adenoviral vector is deleted for the E1 and E3 region. The E3 region was replaced by the expression cassette of the tTA-Tet-Off-transactivator and its constitutive CMV-promoter. To create the pAd1-ΔE1/E3-tTA (Gillissen, et al., 2003 ). The shuttle plasmid pAd2-TRE is used to transfer the gene of interest into the E1 region of the adenovirus type 5. Through modifications, it contains one TRE-element and one BGH-poly(A) signal (Gillissen, et al., 2003 ). Myc-tagged human cDNA of BimL or BimS was cloned by PCR into the shuttle vector using specific primers with BamHI and XbaI for restriction site. Bim was inserted into Ad1-tTA by cotransfection of BJ5183 bacteria for homologous recombination of the shuttle vector containing Bim and the adenoviral vector. Therefore, both plasmids the shuttle vector and Ad5-tTA were linearised with PacI. Homologous recombination in these recombinase proficient bacteria resulted in the final adenoviral vectors Ad5-mycBimL-tTA (Ad-BimL) or Ad5-mycBimS-tTA (Ad-BimS). Production of the virus was achieved by transfecting HEK293 cells with the final adenoviral vector.

Figure 7: Inducible Bim expression mediated by Ad5-mycBim-tTA

A: Genomic structure of recombinant adenovirus Ad5-mycBim-tTA. Ad5 sequences are indicated by black dashes. E1 and E3 regions of Ad5 are replaced by mycBimL and mycBimS expression cassette (white boxes), respectively and tTA expression cassette (grey boxes), respectively (PCMV: immediate early promoter of cytomegalovirus tTA tetracycline-controlled (Tet-Off) transactivator; PTRE: tetracycline-responsive element located 5‘ of the minimal immediate-early CMV promoter).
B: Western blot analysis for BimL and BimS expression. DU145 Bax cells were transduced with Ad5-mycBimL-tTA or Ad5-mycBimS-tTA and cultured in the presence of different doxycyclin concentrations for 24h.

4.2 Titration of the adenoviruses

In order to determine the dose and time dependent effect of both adenoviruses, AdBimL and AdBimS, on cell death, DU145-Bax cells were measured for DNA fragmentation after indicated time points following infection with AdBimL or AdBimS at MOIs mentioned. Cells that were not infected with the adenovirus served as a negative control and did not show any sign of apoptosis at any of the time points. Also, when doxycyclin was added to the media no apoptotic cells could be detected showing that this agent is not toxic for the cells (figure 8). After 24h, DU145-Bax cells, which were infected with either AdBimL or AdBimS with the indicated MOIs and grown under off conditions showed the same level of apoptosis as the control cells. An apoptotic rate of 4% in the presence of doxycyclin demonstrated that doxycyclin suppressed the expression of Bim in case of both AdBimL and AdBimS But once doxycyclin was not present in the media, cell death increased with rising MOIs of each adenovirus. 25 MOI of AdBimL induced apoptosis in 9% of the cells and 25 MOI of AdBimS induced 21% (figure 7A). 50 MOI of adenovirus generated 18% of dead cells after AdBimL overexpression and 34% after BimS overexpression. At 100 MOI the number of dead cells was 33% for AdBimL and 51% for AdBimS infected cells. AdBimL mediated cell death led to 45% apoptosis with 150 MOI and to 57% in cells transduced with AdBimS.

↓44

The 48h time point illustrates the strength of AdBimS, since only 25 MOI of this adenovirus provoked 54% of cell death. Increased MOI of 50 ended up with 81% of apoptotic cells (figure 8A). 100 MOI and 150 MOI even induced apoptosis in 88% of the cells. To a lesser extent BimL overexpression displayed moderately increasing steps in the apoptotic rate with 12% upon 25 MOI and with 24% upon 50 MOI. 39% and 53% of the cells were killed upon 100 MOI and 150 MOI, respectively. In contrast to AdBimS, the cells infected with AdBimL and cultured under off conditions at any MOI rate did not present any noteworthy apoptotic rate. Overexpression of BimS under off conditions revealed leakiness of the system starting with 50 MOI at 17%, mounting to 22%with 100 MOI and 32% of cell death with 150 MOI.

72h post infection cells visibly showed that the off-system was not able anymore to regulate the suppression of Bim properly (figure 8A). AdBimL, under off conditions, induced only some cell death of 10-15%. DU145-Bax cells infected with AdBimS under off conditions caused apoptosis in the range of 14% up to 60%. Under on conditions, with 25 MOI of AdBimL 22% of the cells were found to be apoptotic, with 50 MOI 28%, with 100 MOI 83% and with 150 MOI 89%. AdBimS demonstrated its full power after 72h by killing nearly all the cells (90%) regardless of the MOI used with the exception of 25 MOI, where it induced 47% of cell death. Of note, cells cultured under off conditions and infected with 50 MOI, 100 MOI or 150 MOI of AdBimS showed high numbers of dead cells. In consideration of these results, all the experiments were performed with 25 MOI for both, AdBimL and AdBimS unless mentioned otherwise. To confirm the expression of AdBimL or AdBimS Western blot analyses were performed at four different time points (figure 8B). DU145-Bax cells were transduced with either AdBimL or AdBimS and harvested at the indicated time points. As a control, non-infected cells were used that did not show over expression of Bim. DU145-Bax cells, which were transduced with AdBimL and AdBimS respectively but grown under off conditions, did not show any Bim expression after 16h. Both isoforms were detected for the first time 6h after infection of the cells with the adenovirus. Later time points displayed the increasing expression for both splicing variants. This Western blot analysis displayed the quick induction of Bim.

Figure 8: Dose-response for apoptosis induction by Bim

A, B, C: Left panel: flow cytometric measurement of apoptotic cells with hypodiploid DNA content. DU145-Bax cells were infected with AdBimL at different MOI and cultured in the presence (off) or absence (on) of doxycyclin for 24h, 48h and 72h (top down).
Right panel: DU145-Bax cells were transduced with were infected with AdBimS and treated as described for were infected with AdBimL. Means +/- SD from three independent experiments.
D: Western blot analysis of DU145-Bax neo cells. Cells were transduced with AdBimL or AdBimS and cultured for indicated hours. Cells were harvested after indicated hours and blotted for BimL and BimS, respectively. Actin was used as a loading control.

4.3 Bim activates the intrinsic mitochondrial pathway

4.3.1 Clustering of Bax and Bak

↓45

The multi-domain pro-apoptotic proteins Bax and Bak are known to be involved in the signalling cascade of apoptosis. Once activated these proteins oligomerize at the mitochondria and ensure that the mitochondria are permeabilized by forming pores at the outer membrane.

One way to visualize Bax and Bak clusters at the mitochondria is to make use of green fusion fluorescence proteins. To this end, DU145 cells were stably transfected to express the fusion proteins EGFPBax or EGFPBak, respectively. These cells display green fluorescence even under non-treated (control) conditions (figure 9A and B). Bak is localized to reticular structures since it is associated to the mitochondria (figure 9A). Bax on the other hand is a predominantly cytosolic protein, therefore the fluorescence looks diffuse (figure 9B). The same pictures present themselves under off conditions, when both cell types were infected with the indicated adenovirus, but were cultured in doxycyclin containing media. For both, DU145 EGFPBax and DU145 EGFPBak, regardless whether they are infected with AdBimL or AdBims, did not show any Bax or Bak clustering. 24h post infection with AdBimL or AdBimS under on condition, the fluorescence microscopy revealed a punctuated pattern of EGFPBak and EGFPBax. Both isoforms of Bim caused clustering of Bak and Bax in consequence of oligomerization of these pro-apoptotic proteins. This demonstrates that Bim activates both pro-apoptotic multi-domain proteins. The pictures in the last row represent cells treated with 1µg/ml of epirubicin as a positive control. Epirubicin is an anthracyclin antibiotic, which acts as a DNA damaging agent by inhibiting topoisomerase II. Therefore it triggers the activation of the mitochondrial pathway, which is marked by the assembly of Bak and Bax oligomers at the outer membrane. DU145 EGFPBax and DU145 EGFPBak cells clearly illustrated clustering of Bax and Bak after treatment with epirubicin.

Figure 9: Bim induces oligomerization of Bax and Bak in DU145 cells

A: DU145 EGFP-Bak cells were transduced with AdBimL (left panel) and AdBimS (right panel) and cultured for 24h in the presence (off) and the absence (on) of doxycyclin or were treated with epirubicin (epi). Pictures were taken with a fluorescence microscope.
B: DU145 EGFP-Bak cells treated as in A.

4.3.2 Re-expression of Bax and overexpression of Bak in DU145 cells

↓46

Wild type DU145 are deficient for Bax and express only a small amount of Bak. The two cell types used in the following experiments originate from these cells. DU145-Bax transfectants were generated from mock cells to re-express Bax (figure 10A). Mock transfectants were generated in parallel by use of the empty HyTK vector. DU145-Bak cells overexpress exogenous Bak but do not express Bax, like the DU145 wild type cells. Matching mock transfected cells show a low level of Bak expression (figure 10B). These two cell types, one re-expressing Bax and the other overexpressing Bak, gave the possibility to investigate the interplay of Bim with these pro-apoptotic proteins separately.

Figure 10: Expression of Bax and Bak in DU145 cells

Western blot analysis of Bax and Bak expression in DU145-Bax and DU145-Bak cells. Non-treated cells were harvested and blotted for Bax and Bak respectively. Actin was used as a loading control.

4.3.3 Conformational switch of Bax and Bak

For further confirmation that Bim activating Bax and Bak, DU145-Bax, DU145-Bak cells and their corresponding mock transfected control cells were transduced with AdBimL or AdBimS. 16h later, the cells were harvested and stained with conformation specific antibody against the N-terminus of Bax (BaxNT) or the N-terminus of Bak (BakNT). Shift to the higher fluorescence indicating the conformational change of the protein, was determined by flow cytometry (figure 11). First, the cells were examined for a conformational change of Bax. DU145 mock cells, which were not infected with the adenovirus or were grown under off conditions, did not display Bak activation. Also in the corresponding Bax transfectants, no activation could be detected in the control cells or those under of conditions. Overexpression of BimL or BimS in mock cells, which are Bax negative, showed a similarly low number of cells with BaxNT as the controls, namely 8% for BimL and 9% for BimS. After AdBimL expression and under on conditions, in 14% of DU145-Bax cells active Bax was measured. In case of AdBimS, 40% of DU145-Bax cells were stained for Bax conformational change (figure 11A, upper panel). As a control, DU145-Bak and their mock transfectants were investigated for Bax activation, although these cells are Bax negative (figure 11A, lower panel). 4-6% of stained cells positive for BaxNT were detected when the cells were not infected and 8-16% when they were cultured under off conditions. Expression of BimL or BimS in mock cells yielded for both BH3-only proteins in 11% of cells positively stained for conformational changed Bax. Staining of DU145-Bak cells revealed few positive cells, 13% were found after AdBimL and 18% after AdBimS transduction.

↓47

Next, the cells were assayed for a conformational change of Bak. Non-treated DU145 mock as well as DU145-Bax cells did not demonstrate a conformational change of Bak. This was also true for cells, which were infected with either of the adenovirus in the presence of doxycyclin i.e. when the expression was turned off. Only 4-10% of the cells presented staining. Upon BimL expression, 13% of DU145 mock cells were detected BakNT positive and 20% upon BimS expression (figure 11B, upper panel). 30% of DU145-Bax transfectants expressing BimL showed exposure of the BakNT epitope, whereas expression of BimS presented 51% cells with positive staining. The same procedure was performed for DU145-Bak cells (figure 11B, lower panel). 4-11% of DU145 mock and 4-16% of DU145 Bak cells, which were control treated or expression of Bim was suppressed, showed BakNT staining. Under on conditions, determination of activated Bak resulted in high rates for both Bim isoforms in these cells. Expression of BimL revealed 16% of mock cells with Bak conformational change and expression of BimS showed 34%. Staining of DU145-Bak cells displayed 47% positive cells upon BimL and 72% upon BimS expression. These results confirm that Bim is not only able to activate the multi-domain proteins Bax and Bak but can also trigger apoptosis by both proteins, independently from each other.

Figure 11: Bim induces conformational change of Bax and Bak in DU145 cells

A: DU145-Bax and DU145-Bak cells were infected with AdBimL (left panel) or AdBimS (right panel) and grown under off or on conditions. After 16h the cells were measured for conformational change of Bax. Means +/- SD from three independent experiments.
B: DU145-Bax and DU145-Bak cells were infected with AdBimL (left panel) or AdBimS (right panel) and grown under off or on conditions. After 16h the cells were measured for conformational change of Bak. Means +/- SD from three independent experiments.

4.3.4 Induction of apoptosis by Bim is mediated by Bax and Bak

As shown previously Bim activated Bax and Bak by inducing their conformational change and oligomerization at the mitochondria. The next question addressed was to analyze whether this activation of the pro-apoptotic multidomain proteins is functionally linked to activation of apoptosis and consequently DNA-fragmentation. For this purpose both DU145-Bax and DU145-Bak transfectants were used (figure 12). After adenoviral infection with either AdBimL or AdBimS the cells were examined for apoptosis after 48h. Both, BimL and BimS induced apoptosis in Bax dependent fashion. Also, Bak overexpressing, but Bax negative cells underwent apoptosis (figure 12). In the presence of doxycyclin the expression of Bim was suppressed, thus there was no induction of cell death in the Bax or Bak transfectants or their mock controls. Whereas all control treated cells showed an apoptotic rate of 4-6%, cells under off conditions revealed 14-20% of apoptotic cells. Bim is expressed under on conditions, i.e. when no doxycyclin is present in the medium. Bax overexpressing cells underwent apoptosis at 58% after AdBimL expression and at 82% after AdBimS expression. In the mock cells 18% and 26% of apoptosis was found upon overexpression of BimL and BimS, respectively. This might be due to the fact, that these mock cells express a small amount of Bak, but are Bax negative. The same observation was made for the DU145-Bak mock cells. These cells too express a small amount of Bak. In this case, BimL induced 17% and BimS 22% of apoptosis. Interestingly, the apoptotic rate of DU145-Bak cells was similar to the ones obtained in the Bax expressing cells. BimL caused 58% of apoptosis and 72% of the same cells died after infection with AdBimS. These results suggest that overexpression of Bim can trigger apoptosis via Bax or Bak and that both events are independent from each other. Furthermore, both cell types DU145-Bax or DU145-Bak and even their equivalent mock cells show in all cases a higher apoptotic rate after AdBimS infection as compared to AdBimL.

↓48

Figure 12: Bim induces apoptosis in a Bax- and Bak-dependent manner

A, B: Flow cytometric detection of apoptotic cells based on measurement of the cellular DNA content. DU145 cells were transduced with AdBimL or AdBimS and cultured in the presence (off) and the absence (on) of doxycyclin. Control cells were mock treated and grown in the absence of doxycyclin for 72h. Means +/- SD from three independent experiments.

4.4 Activation of pro-apoptotic proteins by Bim leads to the breakdown of the mitochondrial membrane potential

Additionally to the activation of Bax and Bak the integrity of the mitochondrial membrane potential can give evidence to the activation of the intrinsic apoptotic pathway. The cationic carbocyanine dye JC-1 is a fluorescent marker for mitochondrial depolarization during apoptosis. Accumulated JC-1 in the matrix changes its fluorescence emission light from green to red. Cells were infected with AdBimL or AdBimS, harvested after 24h, loaded with JC-1 and the samples were than analyzed for loss of red fluorescence by flow cytometry (figure 13). Non-treated control cells for both DU145-Bax and DU145-Bak transfectants and their control transfectants did not show any disruption of the mitochondria. The same result was obtained for cells under off conditions. Expression of BimL as well as BimS led to the breakdown of the mitochondrial membrane potential. Mock showed low mitochondrial membrane potential in 16% of the cells after BimL expression and in 34% after BimS expression. BimL induced reduction of the mitochondrial potential in 32% of DU145-Bax cells and in 63% of the cells after BimS expression. DU145-Bak and the respective mock transfectants displayed a similar rate of apoptosis induction. AdBimL induced breakdown of the mitochondrial potential in 12% of the mock transfectants and in 30% of Bak overexpressing cells. In 32% of the mock transfectants mitochondrial breakdown of the membrane potential was detected and in 61% of the DU145-Bak cells following BimS expression. Thus, Bim triggers the mitochondrial pathway as shown by the decreased numbers of cells with intact mitochondria. Furthermore, BimS induced the mitochondrial death pathway more effectively than BimL. Moreover, this occurs via either Bax or Bak and this may explain the relatively high rates of mitochondrial membrane potential dissipation and cell death in both mock transfectant types. The very low level at which these cells express Bak is apparently for BimS to cause disruption of the mitochondrial membrane potential and mitochondrial permeabilization.

Figure 13: Bim-induced loss of ∆Ψm in a Bax- and Bak-dependent manner

A, B: Flow cytometric detection of cells with lowered mitochondrial membrane potential. Indicated DU145 cells were transduced with AdBimL or AdBimS and cultured in the presence (off) and the absence (on) of doxycyclin. Control cells were mock treated and grown in the absence of doxycyclin. Measurements were performed after 24h. Means +/- SD from three independent experiments.

↓49

Measurement of the mitochondrial membrane potential can give information about the disruption of the outer membrane of the mitochondria. This leads to the release of proteins that are normally harboured in the intermembrane space of the mitochondria, into the cytosol. Additional confirmation that cytochrome c release did actually occur was obtained by immunofluorescence staining of cytochrome c (Figure 14). Mitochondria were stained with Mito Tracker Red, cytochrome c was stained in green and the nuclei were made visible with DAPI staining. DU145-Bax and their mock controls were infected with either AdBimL or AdBimS and cytochrome c release was determined after 16h. Under off conditions no cytochrome c release was visible in neither of the cell types regardless of the virus used. The overlay revealed by the yellow colour that cytochrome c is still present in the mitochondria. But expression of Bim in DU145-Bax cells caused liberation of cytochrome c as can be seen in the overlay. The higher potency of BimS to induce apoptosis is obvious under on conditions in the mock cells. Even with a low level of Bak expression in the mock transfectants, AdBimS was able to trigger cytochrome c release to a noticeable degree.

Figure 14: Bim induces cytochrome c release dependent of Bax

A, B: Immunofluorescence: indicated DU145 cells were transduced with AdBimL or AdBimS. Cells were stained for mitochondria with MitoTracker Red CMXRos, for cytochrome c with anti-cytochrome c followed by Alexa Fluor 488-conjugated chicken anti-goat antibodies (green fluorescence) and for the nuclei with DAPI staining. Pictures were taken with fluorescence microscope 16h after infection with the corresponding adenovirus.

4.5 Initiation of the caspase cascade by Bim 

The activation of Bax and Bak and the disruption of the mitochondrial membrane are events that take place prior to caspase activation. The following studies aimed to identify the caspases involved in the intrinsic death pathway that is induced by Bim. After adenoviral expression of AdBimL or AdBimS and treatment with specific or broad spectrum caspase inhibitors, the apoptotic rate of the cells was determinates by flow cytometry at 48h following transduction. In all experiments the cells were incubated with 20µM of caspase inhibitor (figure 15). Both mock transfectants and likewise DU145-Bax and DU145-Bak, the latter not transduced with the adenoviral vector did not show apoptotic signs, the apoptotic rate amounted to a range of 3-6%. When expression of Bim was suppressed in these cells, 4-8% underwent cell death. In DU145-Bax cells 2-11% of apoptosis was measured, under control or off conditions whereas in DU145-Bak cells 2-7% of the cells died with evidence of genomic DNA fragmentation. In 12% and 6% of the DU145 mock transfectants DNA–fragmentation was detected upon AdBimL or AdBimS expression, respectively. Their DU145-Bax counterparts revealed cell death in 33% of the cells after transduction with AdBimL in the absence of doxycyclin. AdBimS expression induced apoptosis in 41% of these cells. 35% of the DU145-Bak cells cultured under on conditions were found to be apoptotic following AdBimL expression, while 41% of the cells died when AdBimS was expressed. However, in this setting, both isoforms could only trigger 8% of apoptosis in mock cells.

↓50

In a first step, using the broad caspase inhibitor zVAD-fmk all caspases were blocked to see if the caspases play functional role in the Bim pathway. Additional treatment with zVAD-fmk reduced apoptosis to 3% in DU145-Bax cells transduced with AdBimL as compared to Ad BimL expression alone in the absence of doxycyclin. This is representative for all non-treated (control) cells and the one under off conditions. In the same setting, also the short version of Bim failed to induce cell death, 6% apoptotic cells were measured. Overexpression of AdBimL or AdBimS and addition of zVAD-fmk in DU145-Bak cells led to a similar outcome with 4% of dead cells. To identify the caspases that play a role for Bim triggered cell death caspase inhibitors were employed. The caspase-3 inhibitor zDEVD-fmk caused reduction of apoptosis under on conditions in DU145-Bax cells after AdBimL expression. Only 17% of the cells died under these circumstances. The same cells overexpressing AdBimS and grown in the presence of zDEVD-fmk showed 14% of apoptosis. Again, a similar result was found in the DU145-Bak cells: with caspase-3 being blocked, AdBimL could only induce 20% and AdBimS 13% of apoptosis. Next, the impact of caspase-9 was examined with the zLEHD-fmk inhibitor. Only 10% of the DU145-Bax cells, which were grown in the presence of the inhibitor went to apoptosis upon expression of AdBimL and 12% died upon expression of AdBimS. DU145-Bak overexpressing cells behaved the similarly, presenting also 10% of apoptosis after infection with AdBimL and 16% with AdBimS. And finally caspase-8 was blocked with zIETD-fmk inhibitor, where a similar effect was achieved. 17% and 12% of dead Bax expressing cells were detected upon BimL and BimS expression. Their DU145-Bak counterparts reacted the same way after BimL or BimS expression, namely with 13% of apoptosis.

In general, it can be said, that the specific caspase inhibitors saved around 50% of the cells in both cell types. Thus, the initiator caspases-8 and 9 and the effector caspase-3 are important for Bax and Bak mediated apoptosis upon Bim expression, especially since the pan-caspase inhibitor diminished cell death to the level of not infected cells.

Figure 15: Bim-induced apoptosis is caspase dependent

A: DU145-Bax and DU145 mock cells were infected with AdBimL or AdBimS in the presence (off) and the absence (on) of doxycyclin and treated with 20μM of the indicated caspase inhibitors. Cells were cultured for 48h and measured for hypodiploid DNA content with flow cytometry. Means +/- SD from three independent experiments.
B: DU145-Bak and DU145 mock cells treated as in A. Means +/- SD from three independent experiments.

↓51

Apparently, loss of the mitochondrial membrane potential happens within 24h after Bim expression (figure 13). Once the outer membrane is broken apoptosis associated proteins are released, such as AIF, Smac and cytochrome c. These proteins aim to different targets and carry on the apoptotic process. Cytochrome c for example is part of the apoptosome where caspase-9 is activated which in turn leads to the activation of the caspase cascade. Thereby, the pro-form of the caspases is cleaved to become an active caspase. Since Bim is obviously activating the caspase cascade, cleavage products of the caspases should be detectable on the protein level. Hence, total lysates of cells were collected at different time points to perform western blot analysis after infection of DU145-Bax cells with adenoviral Bim. With this time course the generation of the active caspases could be followed (figure 16). As a control, DU145-Bax cells were transduced with AdBimL or AdBimS and cultured in the presence of doxycyclin for 24h. Western blot analysis showed that in case Bim is not expressed, there is no caspase activation, cytochrome c release or PARP cleavage in these cells (figure 15A, B, first lane). One of the initiator caspases that is activated at early time points of the apoptotic process is caspase-9. Upon BimL expression caspase-9 is cleaved after 14h, this time point marks the formation of the apoptosome (figure 16A). 10h after transduction of AdBimS activation of caspase-9 could be observed (figure 16B). For both isoforms, AdBimL and AdBimS the cleavage products could not be detected after 24h, probably due to their degradation. The impact of Bim on the main effector caspase-3 demonstrated the same effect. 14h after expression of BimL, pro-caspase-3 was processed to its cleavage products of 20kDa and 18kDa. But at the 24h time point there are no more active caspases detectable by Western blot analyses. BimS was more effective by inducing caspase-3 activation already 10h post adenoviral infection. Disruption of the outer mitochondrial membrane potential causes the release of cytochrome c and is upstream of caspase activation events. According to the results of caspase activation, cytochrome c is detectable in the cytosol (cytosolic fraction) after 14h for AdBimL and after 10h for AdBimS expression. PARP (poly-ADP-ribose-polymerase) is a DNA repair enzyme. It is cleaved upon apoptosis for a substrate of caspases-3/7 and is therefore a good marker for this type of caspases mediated cell death. After expression of BimL in Bax positive DU145 cells, PARP is cleaved dominantly after 14h and upon BimS expression after 10h. Over time the pro-form of PARP disappears and after 18h and 14h respectively, only its cleaved form can be found. The results, gained by this time course, indicate that BimS activates the apoptotic pathway faster than BimL.

Figure 16: Bim activates proteins of the intrinsic pathway

A: Western blot analysis for BimL. DU145-Bax cells were transduced with recombinant adenovirus AdBimL and cultured for indicated hours in the presence (off) and the absence (on) of doxycyclin.
B: Western blot analysis for BimS, as in A.

4.6 Role of Bcl-2 in Bim induced cell death

Bcl-2 is an anti-apoptotic protein and can antagonize the actions of pro-apoptotic proteins, although the exact mechanism is still strongly debated (Borner, 2002). Therefore, the next step in these investigations was to establish the role of Bim and the potentially involved organelles. The following three kinds of Bax re-expressing DU145 cells were used for the studies. Cells were transfected stably to express Bcl-2 mutants either at the mitochondria (Bcl-2actA) or at the endoplasmic reticulum (Bcl-2cb5). The control cells (neo) were transfected with the empty vector and do not overexpress Bcl-2 (figure 17). To target Bcl-2 to the mitochondria, the membrane anchor was replaced by the mitochondrial insertion signal protein actA of Listeria monocytogenes. To achieve localization of Bcl-2 to the ER, the rat hepatic ER cytochrome b5 isoform targeting sequence was fused to Bcl-2 instead of its C-terminal transmembrane region. The resulting products were transduced into DU145-Bax cells. The purpose of these transfectants was to investigate whether Bcl-2, being an anti-apoptotic protein would be able to protect the cells from Bim-induces apoptosis through a mitochondrial and/or ER pathway.

↓52

Figure 17: Expression of Bcl-2, Bax and Bak expression in DU145-Bax cells

Western blot analysis of Bcl-2, Bax and Bak expression in DU145-Bax cells.

4.6.1 Bcl-2 targeted to the mitochondria or the ER prevents cell death upon certain apoptotic stimuli 

As shown by Western blot analysis, DU145-Bax Bcl-2-transfected cells overexpress Bcl-2 (figure 17). To confirm that these cells express Bcl-2 at the targeted organelle, mitochondria and ER, respectively, immunostaining for subcellular localization was performed. The cells were stained for Bcl-2, mitochondria or endoplasmic reticulum and examined with a fluorescence microscope (figure 18). After fixation and permeabilization the cells were incubated with the primary antibody specific for mitochondria, for the ER or for Bcl-2. Addition of the second fluorescent antibody stained the mitochondria and the ER red or Bcl-2 green. Nuclei were DAPI stained, and thereby visualized in blue. As for the neo control transfected cells there was no staining for Bcl-2 at all, showing that these cells do at best express low levels of Bcl-2 and can be used as a control. DU145-Bax Bcl-2actA cells reveal in the overlay that Bcl-2actA localized at the mitochondria (figure 18, left panel), but was not found at the ER (figure 18, right panel). Vice versa in DU145-Bax Bcl-2cb5 cells, here Bcl-2cb5 inserted into the endoplasmic reticulum (figure 18, right panel), but did not co-localize with the mitochondria (figure 18, left panel).

Figure 18: Localization of targeted Bcl-2 in DU145-Bax transfectants

Cells were stained for mitochondria with MitoTracker Red CMXRos or ER with mouse anti-BAP31 followed by Alexa Fluor 594-conjugated chicken anti-rat (red fluorescence). Bcl-2 was visualized by anti-Bcl-2 followed by Alexa Fluor 488-conjugated chicken anti-mouse (green fluorescence). The nuclei were stained in blue with DAPI.

↓53

The subsequent drugs, known for their involvement in cell death pathways and stimulation of different types of cell death were applied to confirm the functionality of cell system used in the following experiments. Epirubicin is a DNA damaging agent that inhibits topoisomerase II and triggers the mitochondrial apoptotic pathway. Tunicamycin, a protein N-glycosylation inhibitor, and thapsigargin, which blocks SERCA, are both causing ER stress induced apoptosis. After 72h, the cells were subjected to flow cytometric measurement of the cellular DNA content to determine the rate of apoptotic cells (figure 19). DU145-Bax neo cells showed 77% apoptotic cells after treatment with Epirubicin. Induction of ER stress also led to apoptosis in these cells. 74% of dead cells were detected after tunicamycin and 37% after thapsigargin treatment and showed partial inhibition of apoptosis as compared to the control transfectants. Epirubicin induced in 45% of DU145-Bax Bcl-2actA (mito) cells apoptosis by DNA damage. Treatment of DU145-Bax Bcl-2actA cells with tunicamycin and thapsigargin respectively induced in 52% and 47% of the cells apoptosis by ER stress. The highest inhibition of apoptosis by Bcl-2 was achieved in DU145-Bax Bcl-2cb5 (ER) cells. Whereas 50% of the cells died upon Epirubicin treatment, tunicamycin and thapsigargin had a minor effect on the cells by revealing only 25% and 22% of apoptotic cells. Thus, ER stress inducing agents induced much less apoptosis in cells that express Bcl-2 targeted to the endoplasmic reticulum. Additionally, these measurements suggested that Epirubicin does not engage to the ER to induce apoptosis.

Figure 19: Functionality of the DU145-Bax Bcl-2 transfectants

DU145-Bax mock and Bcl-2 transfectants were treated with epirubicin (1μg/ml), tunicamycin (1μM), or thapsigargin (10μM). Apoptotic cells were determined after 72h by measurement of the genomic DNA content.

4.6.2 Inhibition of Bim induced apoptosis by Bcl-2

All three DU145-Bax cell lines were transduced with AdBimL and AdBimS respectively (figure 20). 48h post transduction, the apoptotic cells were measured by flow cytometry. Non-treated cells and cells under off conditions did not show any pertinent apoptosis neither for BimL nor for BimS. In DU145-Bax neo cells AdBimL induced 50% of cell death, whereas AdBimS managed to kill 63% of the cells. But when Bcl-2 was present, this anti-apoptotic protein was able to overcome the induction of apoptosis by Bim. Localized at the mitochondria Bcl-2 partially protected the cells against Bim. DU145-Bax Bcl-2actA cells died by 31% over BimL and by 45% over BimS expression. In consideration of the DU145-Bax neo cells, this signifies an inhibition of cell death by approximately 1.5 fold for both BimL and BimS. The most significant effect of Bcl-2 was detected in the DU145-Bax Bcl-2cb5 cells. The apoptotic rate was only 17% for BimL and 15% for BimS. Compared to the control cells this means an inhibition factor of 3 times for AdBimL and even 4 times for AdBimS. Although Bim induced apoptosis in cells without Bcl-2 expression, as could be observed in the DU145-Bax neo cells, Bcl-2 protected the cells from death, especially when located at the endoplasmic reticulum. In this experiment as well as in the following it should be noted that BimS induced a higher level of cell death as has already be seen in the other DU145 cells. Once again the results confirm the current model that BimS induces apoptosis more effectively than BimL.

↓54

Figure 20: Bcl-2 at the endoplasmic reticulum protects cells from Bim-induced apoptosis

B, D: flow cytometric detection of apoptotic cells based on measurement of the cellular DNA content. DU145-Bax cells were transduced with AdBimL or AdBimS and cultured in the presence (off) and the absence (on) of doxycyclin. Control cells were mock treated and grown in the absence of doxycyclin for 48h. Means +/- SD from three independent experiments.
A, C: representative histograms of experiments in B and D.

4.6.3 Necrotic cell death

Since there were almost no apoptotic cells found among DU145-Bax Bcl-2cb5 cells upon Bim expression, the possibility of a necrotic cell death had to be ruled out. By double staining of the cells with Annexin-V-FITC and propidium iodide (PI), viability of the cells can be monitored and necrotic versus apoptotic cells identified. PI is used as a marker for the integrity of the cell membrane in PI negative cells. Annexin-V-FITC points at early apoptotic cells by binding to phosphatidylserine on the outer side of the membrane. It should be noted that in late stages of apoptosis cells may also lose their cell membrane integrity and can therefore be double stained. In that case, late apoptotic cells cannot be distinguished from necrotic cells. DU14- Bax neo, DU145-Bax Bcl-2actA and DU145-Bax Bcl-2cb5 cells were harvested 16h after transduction and were evaluated for double positive cells (figure 21). Non-treated cells that were used as a control, showed 5% necrosis in DU145-Bax neo, 8% in DU145-Bax Bcl-2actA and 3% in DU145-Bax Bcl-2cb5 cells. Culture of adenovirally transduced cells in the presence of doxycyclin to inhibit induction of BimL or resulted in a similarly low amount of necrotic cells. 5-6% was determined in DU145-Bax neo, 8-9% in DU145-Bax Bcl-2actA and 4-5% in DU145-Bax Bcl-2cb5 cells. Expression of BimL led to elevated, but not high levels of double positive cells. In DU145-Bax neo 21% of double positive cells were measured, in DU145-Bax Bcl-2actA 14% and in 16% DU145-Bax Bcl-2cb5 cells. The short variant of Bim did not particularly change the number of cells stained for Annexin-V-FITC and PI. Quantification of DU145-Bax neo, DU145-Bax Bcl-2actA and DU145-Bax Bcl-2cb5 cells revealed 24%, 20% and 21% of necrotic cells upon Bim expression.

These data clarified that Bcl-2actA or cb5 expression did not convert the cell death from apoptosis to necrosis but maintained viability upon BimL or BimS expression. The most relevant result though, was that in DU145-Bax Bcl-2cb5 cells Bim did not induce necrosis, nor did it cause apoptosis as could be observed by measurement of DNA-fragmentation (figure 20). Bcl-2 localized at the endoplasmic reticulum seems to prevent apoptotic death induced by Bim.

↓55

Figure 21: Bim does not induce necrosis in DU145 cells

A, C: representative dot plots of experiments presented in B, D.
B,   D: Cells were stained with Annexin-V-FITC, counterstained with PI and measured by flow cytometry. PI positivity is a sign of necrosis, whereas cells positive for Annexin-V, but negative for PI are defined as apoptotic. DU145 cells were transduced with AdBimL or AdBimS and cultured in the presence (off) and the absence (on) of doxycyclin. Control cells were mock treated and grown in the absence of doxycyclin for 48h. Means +/- SD from three independent experiments.

4.7 Disruption of the mitochondrial membrane potential

The breakdown of mitochondrial membrane potential occurs prior to DNA-fragmentation and cell death and marks the activation of the intrinsic mitochondrial apoptotic pathway. Permeabilization of the mitochondrial outer membrane results in breakdown of the mitochondrial membrane potential. The next question was whether Bcl-2 was able to avoid cell death by preventing the breakdown of the mitochondrial membrane potential. DU145-Bax neo, DU145-Bax Bcl-2actA and DU145-Bax Bcl-2cb5 cells were incubated with AdBimL or AdBimS for 48h and the loss of the mitochondrial membrane potential was measured by flow cytometry.

Control treated cells of all three cell transfectants (neo, Bcl-2actA, Bcl-2cb5) did not show loss of the mitochondrial potential in the cells measured. This was still true when the expression of Bim was prevented under off conditions in AdBim transduced cells (figure 22). Under on conditions on the other hand, 62% of DU145-Bax neo cells were detected with mitochondrial permeability shift upon BimL and 79% upon BimS expression. Overexpression of Bcl-2 at the mitochondria reduced the cells with low mitochondrial membrane potential after Bim expression. Only 28% of D145-Bax Bcl-2actA cells were found to be affected by AdBimL, whereas 35% of the cells showed loss of mitochondrial membrane potential in the case of AdBimS. This indicated that these cells are protected by Bcl-2 at the mitochondrial level from Bim induced breakdown of the mitochondrial membrane potential. Conversely, in DU145-Bax Bcl-2cb5 cells, where Bcl-2 is associated to the endoplasmic reticulum, 54% of the cells had loss of their mitochondrial potential upon AdBimL infection. AdBimS caused in 67% of the same cells reduction of the mitochondrial potential. Thus, Bcl-2 could only to a minor extent prevent disruption of the mitochondria upon Bim expression, when it was targeted to the ER, although it dramatically inhibited DNA-fragmentation at the same position.

↓56

Figure 22: Bcl-2 targeted to the mitochondria prevents loss of ∆Ψm induced by Bim

A, B: Flow cytometric detection of cells with lowered mitochondrial membrane potential. Indicated DU145 cells were transduced with AdBimL or AdBimS and cultured in the presence (off) and the absence (on) of doxycyclin. Control cells were mock treated and grown in the absence of doxycyclin for 48h. Means +/- SD from three independent experiments.

4.8 Caspases are crucial in the Bim pathway

A connection between Bim and the caspases was already established in Bax or Bak expressing cells. Overexpressed Bcl-2 had been shown to save cells from apoptosis. Therefore, the question was addressed whether Bcl-2 would have an influence on the initiator caspases-8 and -9, and the effector caspase-3, which can be cleaved and activated by both initiator caspases. DU145-Bax neo, DU145-Bax Bcl-2actA and DU145-Bax Bcl-2cb5 cells were transduced with AdBimL or AdBimS and grown under off or on conditions. Cells were measured after 36h by flow cytometry for activated caspases using a cell permeable FITC labelled tetrapeptide of the corresponding caspase. The caspase activity was quantified according to elevated fluorescence. First, caspase-9 was analyzed in the Bim-induced apoptosis pathway. All three DU145-Bax transfectants, showed no caspase-9 activation upon mock treatment as determined by binding of the LEHD-FITC substrate (figure 23). Also, none of the cells showed caspase-9 activation when the expression of either Bim isoforms was suppressed by the presence of doxycyclin. Transduction of BimL in DU145-Bax neo resulted in 23% of cells with caspase-9 activity. In DU145-Bax Bcl-2actA transfectants 14% of cells were found with caspase-9 activation and in 8% of the cells with Bcl-2 localized at the ER. The same procedure upon BimS expression revealed caspase-9 activation in 44% of the DU145-Bax neo cells, 25% of the DU145-Bax Bcl-2actA transfectants and 16% of the DU145-Bax Bcl-2cb5 transfectants.

As the effector caspase-3 is cleaved and activated by caspase-9, caspase-3 activation should occur upon Bim expression. Cells, which were mock treated or transduced with AdBimL and then grown under off conditions, did not display caspase-3 activation as determined by binding of fluorescin conjugated DEVD peptide. 39% of DU145-Bax neo cells presented caspase-3 activation upon BimL expression. AdBimS induced caspase-3 activation in 62% of the DU145-Bax neo cells. In cells, where Bcl-2 was targeted to the mitochondria, BimL and BimS expression stimulated in 17% and 37% of the cells caspase-3 activation, respectively. Similar numbers were obtained for DU145-Bax Bcl-2cb5 cells; upon BimL expression 22% and upon BimS expression 34% of the cells were determined.

↓57

When caspase-3 is active, it can process caspase-8, which in turn cleaves Bid to tBid. Truncated Bid (tBid) can provoke activation of the mitochondria via Bak and Bax, thereby creating an amplification loop. To see whether Bim could induce such a feedback loop, the activation of caspase-8 was measured by staining with a fluorescin labelled IETD peptide. Control treated cells of all three DU145-Bax transfectants were monitored for caspase-8 activation and showed no induction of such activation. Likewise, adenovirally transduced cells cultured in the presence of doxycyclin did not show caspase-8 activation. Transduction of DU145-Bax neo cells with AdBimL resulted in 30% of cells showing caspase-8 activation as compared to 14% of DU145-Bax Bcl-2actA cells and 15% of DU145-Bax Bcl-2cb5 cells were found. BimS had a much stronger effect on caspase-8 activation as compared to BimL. 64% of the DU145-Bax neo cells showed caspase-8 activation. But when Bcl-2 was present at the mitochondria caspase activation found in only 36% of the cells and when Bcl-2 was expressed at the ER, 25% of the cells showed caspase-8 activation.

Figure 23: Bcl-2 targeted to the ER can prevent activation of the caspases upon Bim expression

A, B, C: Flow cytometric detection of cells with increased caspase activation. Indicated DU145 cells were transduced with AdBimL or AdBimS and cultured in the presence (off) and the absence (on) of doxycyclin. Control cells were mock treated and grown in the absence of doxycyclin. Measurements were performed after 36h. Means +/- SD from three independent experiments.

As observed in previous experiments BimS exerted caspase activation in a higher percentage of cells (figure 23). Caspase activation for all caspases analyzed could be inhibited by both Bcl-actA and Bcl-2cb5 to a similar extent. To summarize these findings, BimL and BimS induced activation of caspase-3, -8, and -9, cells were infected with adenoviral Bim and treated with several irreversible caspase inhibitors. 48h later the cells were harvested and validated for DNA-fragmentation. Control treated DU145-Bax neo cells displayed apoptosis in 3 to 4% of the cells (figure 24). Mock treated DU145-Bax Bcl-2actA cells showed 4-6% of apoptotic cells and DU145-Bax Bcl-2cb5 cells 2-6%. When these cells were infected with AdBimL or AdBimS, but cultured in doxycyclin containing media, 5-7% DU145-Bax neo cells were found to be apoptotic, in DU145-Bax Bcl-2actA 4-8%, and in DU145-Bax Bcl-2cb5 5-11% of the cells. When the expression of BimL was turned on DU145-Bax neo cells genomic DNA fragmentation was measured in 52% of the cells. Additional treatment with zDEVD-fmk, a caspase-3 inhibitor, diminished the percentage of apoptotic cells to 25% establishing the relevance of the effector caspase-3 in this setting. When zLEHD-fmk was used to inhibit caspase-9, DNA fragmentation was measured in 44% of the cells. Caspase-8 inhibition by the zIETD-fmk inhibitor revealed an apoptotic rate of 31%. Caspase-4 has been shown to be located at the ER and was suggested to be only involved in apoptosis if the ER pathway is activated. Treatment with the caspase-4 inhibitor, zLEVD-fmk, marginally inhibited apoptosis with only 40% of cells showing DNA fragmentation. Finally, treatment with a pan-caspase inhibitor, zVAD-fmk, strongly reduced cell death to 10%, indicating that BimL cannot mediate apoptosis without caspases.

↓58

Even though apoptosis was decreased in cells expressing Bcl-2actA compared to cells with neo control transfectants showing only low Bcl-2 expression, additional caspase inhibition could further decrease apoptosis. DU145-Bax Bcl-2actA cells presented an apoptotic rate of 31% upon BimL expression, which fell to 22%, when the cells were additionally treated with zDEVD-fmk for caspase-3 inhibition. BimL triggered apoptosis in 27% of the cells in the presence of the caspase-9 inhibitor zLEHD-fmk. When caspase-8 was inhibited 14% of the DU145-Bax Bcl-2actA cells revealed DNA-fragmentation and inhibition of caspase-4 resulted in 23% of dead cells. Addition of the pan-caspase inhibitor zVAD-fmk decreased the rate of the apoptotic DU145-Bax Bcl-2actA cells to a level compared to the control treated cells, namely 7%. None of the caspase inhibitors had a major effect on cells overexpressing ER-targeted Bcl-2, since BimL alone could not induce cell death in more than just 16% of the cells. Addition of the caspase-3 inhibitor zDEVD-fmk led to 13% of cell death in these cells transduced with AdBimL and upon blocking of caspase-9 by zLEHD-fmk to 15%. Inhibition of caspase-8 by zIETD-fmk led to DNA-fragmentation in 9% of the cells and addition of the caspase-4 inhibitor zLEVD-fmk resulted in apoptotic DNA fragmentation in 13% of the cells. At last, when the cells were treated with the broad spectrum caspase inhibitor zVAD-fmk, 7% of apoptotic cells were measured.

The same settings were used for investigations of caspases involved in BimS mediated apoptosis. BimS caused apoptosis in 69% of DU145-Bax neo cells. Once caspase-3 was block by the inhibitor, the apoptotic rate was lowered to 44%. Caspase-9 inhibition still resulted in 62% of apoptotic cells. When the casapse-8 inhibitor was added, BimS induced DNA fragmentation in 46% of the cells. Treatment with the casapse-4 inhibitor zLEVD-fmk decreased apoptosis to 48%. The pan-caspase inhibitor zVAD-fmk strongly reduced cell death to 15% of the cells. DU145-Bax Bcl-2actA cells displayed in 50% of the cells apoptotic DNA fragmentation upon BimS expression, showing that Bcl-2 was partially able to guard the cells. When casapse-3 was inhibited with zDEVD-fmk, 27% of apoptotic cells were observed. In case of additional caspase-9 inhibition, 45% of these cells underwent apoptosis. 26% of apoptosis happened in cells exposed to zIETD-fmk to block caspase-8 and 34% of cells died upon additional caspase-4 inhibition. In the presence of the pan-caspase inhibitor zVAD-fmk 20% of these cells showed DNA-fragmentation. Protection by Bcl-2cb5 was so strong that only 15% of apoptotic cells were detected in AdBimS infected cells. There was no relevant reduction in the amounts of apoptotic cells after additional treatment with the various caspase inhibitors. 11% of the cells were quantified for DNA-fragmentation upon BimS expression after addition of the caspase-3 inhibitor.14% of the cells underwent apoptosis when they were treated with casapse-9 inhibitor. Inhibition of caspase-8 and caspase-4 resulted in 12% and 13% apoptotic cells, respectively. The 12% of apoptotic cells detected in cells treated with the broad spectrum caspase inhibitor zVAD-fmk showed that the residual apoptosis induced by BimS in Bcl-2cb5 expressing cells is caspase independent.

Figure 24: Bim acts in a caspase-dependent manner

A: DU145-Bax cells were infected with AdBimL in the presence (off) and the absence (on) of doxycyclin and treated with 20µM of indicated caspase inhibitors. Cells were cultured for 48h and measured for hypodiploid DNA content by flow cytometry. Means +/- SD from three independent experiments.
B: DU145-Bax cells infected with AdBimS and treated as in A. Means +/- SD from three independent experiments.

4.8.1 Processing of caspases upon Bim expression

↓59

Caspases were shown to be functionally involved in Bim induced apoptosis. This was shown on one hand by measurement of active caspases and on the other hand by the impact of inhibition of these caspases on apoptosis execution as measured by DNA fragmentation. Thus, detection of caspase subunits by Western blot analysis was used to show proteolytic processing of the caspase proforms (figure 25). All three DU145-Bax transfectants were transduced with the adenoviral vectors and harvested after 30h. The first lane demonstrates that, under on conditions, both Bim isoforms were expressed at equal levels in all cell lines. It should be noted that cells, which were infected with the adenovirus but kept under off conditions did not reveal Bim expression indicating tight conditional expression. Since expression of Bim in DU145-Bax neo cells induced loss of the mitochondrial membrane potential, cytochrome c release was analyzed. Cells, which were either control treated or grown under off conditions following adenoviral transduction, did not show cytochrome c release. In both cases, total cytochrome c was found in the mitochondrial fraction, regardless of the cell types and the Bim isoform used. In the cytosolic fraction of DU145-Bax neo cells, cytochrome c could be detected upon both BimL and BimS indicating cytochrome c release expression. Expression of Bcl-2 at the mitochondria mostly prevented release of cytochrome c into the cytosol upon BimL expression. The same result could be found upon BimS expression where a subtle band for cytochrome c was seen in the cytosolic fraction. Surprisingly, expression of Bcl-2cb5 at the ER inhibited mitochondria from releasing cytochrome c in the presence of BimL and BimS, respectively. This indicates a crosstalk of the ER pathway feeding into mitochondrial apoptosis signalling. Downstream of cytochrome c release formation of the apoptosome and the activation of caspase-9 occur. Uncovering of caspase-9 subunits would indicate processing of caspase-3 by the mitochondrial pathway although caspase-3 was also shown to cleave the caspase-9 zymogen. BimL expression in DU145-Bax neo cells led to processing of caspase-9 as could be observed by the reduced proform. In the same cells, BimS also induced cleavage of caspase-9. Detection of decreased pro-caspase-9 in DU145-Bax Bcl-2actA cells upon expression of BimL showed that caspase-9 was processed even in the presence of Bcl-2. Transduction of these cells with AdBimS under on conditions led to the detection of the processed caspase-9 subunit, indicating activation of caspase-9. In DU145-Bax Bcl-2cb5 cells, neither expression of BimL nor BimS caused cleavage of caspase-9. No reduction of the proform and no cleavage product of caspase-9 could be detected upon Bim expression. Bcl-2 localized at the ER might interfere with Bim mediated caspase-9 activation in a yet unknown way. According to casapse-9 processing, caspase-3 cleavage product was seen upon BimL expression in Bcl-2 deficient cells. Processing of caspase-3 proform was found in DU145-Bax neo cells upon BimS expression. DU145-Bax Bcl-2actA and DU145-Bax Bcl-2cb5 cells revealed a slight occurrence of caspase-3 p20 fragment in case of BimL or BimS expression. To enhance the death signal, caspase-3 can process and activate caspase-8 creating a positive feedback loop. No matter whether Bcl-2 was overexpressed in the cells or not, expression of BimL or BimS resulted in the appearance of a processed caspase-8 intermediate. This might indicate that the amplification loop was initiated independently of the Bcl-2 status or that caspase-8 processing is initiated upstream of a Bcl-2 inhibitable event. Finally, cleavage of PARP was examined as it is cleaved by caspase-3 upon apoptotic stimuli. DU145-Bax neo cells transduced with AdBimL revealed PARP cleavage under on conditions. In the same settings, upon expression of BimS only the cleavage product of PARP could be detected indicating massive caspase-3 activation. Despite the expression of Bcl-2actA, BimL as well as BimS expression induced PARP cleavage to an extent comparable to the DU134-Bax neo controls. Furthermore, BimL and BimS triggered PARP cleavage under on conditions in DU145-Bax Bcl-2cb5 cells, although to a lesser extent. This indicates that caspase activation leading to PARP processing via caspase-3 like caspases occurs within the ER pathway downstream of Bcl-2 inhibitable events.

Figure 25: Bcl-2 partially suppresses processing of caspases

A: Western blot analysis for indicated pro-apoptotic proteins. DU145-Bax cells were transduced with AdBimL and cultured for 30h in the presence (off) and the absence (on) of doxycyclin. Cell lysates were blotted for the indicated proteins with the appropriate antibodies.
B: DU145-Bax cells were transduced with AdBimS, and treated as in A.

4.9 Caspase-8 is involved in BimS induced apoptosis

Although the extrinsic death-receptor initiated pathway was not investigated with regard to Bim mediated apoptosis, it is possible that caspase-8 is important in the Bim pathway since it was reported that caspase-8 is involved in ER stress induced apoptosis (Chandra, et al., 2004 ).

↓60

DU145-Bax neo, DU145-Bax Bcl-2actA and DU145-Bax Bcl-2cb5 cells were incubated with AdBimS for 24h and cells with low mitochondrial membrane potential were measured by flow cytometry (figure 26). DU145-Bax neo cells mock treated or grown under off conditions did not show cells with low mitochondrial membrane potential. Expression of BimS led to 60% of cells, which had loss of their mitochondrial potential and 58% were detected when these cells were treated with casapse-8 inhibitor. DU145-Bax Bcl-2actA cells, which were either not infected with the adenovirus AdBimS or were infected, but cultured in the presence of doxycyclin showed 4-6% of cells with low mitochondrial membrane potential. 46% of these cells were detected with mitochondrial permeability shift upon BimS expression. Additional inhibition of caspase-8 reduced the effect by half. Similar data was obtained in cells, where Bcl-2 is targeted to the ER. Here, 7% of cells showed mitochondrial membrane potential loss when control cultured or transduced with AdBimS in the presence of doxycyclin. BimS induced breakdown of the mitochondrial membrane potential in 58% of the cells. This number was reduced to 32% when caspase-8 was blocked. Thus, BimS mediates cell death through caspase-8 in a Bcl-2 dependent way. According to these data, no difference could be detected with respect to caspase-8 inhibition in BimS induced mitochondrial permeabilization whether Bcl-2 was expressed at the mitochondria or at the ER. The fact that inhibition of caspase-8 did not influence the breakdown of the mitochondrial membrane potential might suggest that Bcl-2 is upstream of caspase-8 activation.

Figure 26: BimS mediates caspase-8 activation

DU145 cells were transduced with AdBimS and cultured in the presence (off) and the absence (on) of doxycyclin or were additionally treated with capase-8 inhibitor for 24h. Control cells were mock treated and grown in the absence of doxycyclin. Means +/- SD from three independent experiments.

4.10 BimS induces calcium release into the cytosol

The main calcium store in the cell is the endoplasmic reticulum. Ca2+ is pumped from the cytosol into the lumen of the endoplasmic reticulum by the SERCA (sarcoplasmic endoplasmic reticulum calcium ATPase). ER stress causes the depletion of calcium from the store. Cytosolic Ca2+ is mostly absorbed by the mitochondria through their Ca2+ uniporters. The positively charged calcium ions disturb the mitochondrial membrane potential. At a certain threshold, accumulated Ca2+ in the matrix triggers the opening of the permeability transition pore, which results in swelling of the matrix, the loss of the membrane potential and finally in the destruction of the outer mitochondrial membrane. Consequently, apoptosis inducing factors are released form the mitochondria, such as cytochrome c, Omi/Htr2 and Smac/Diablo. Since Bcl-2 can antagonize the apoptosis promoting effect of Bim, it was investigated whether Bcl-2 is inhibiting apoptosis (see 3.5.2) by avoiding ER stress and interfering with Ca2+ release. Considering that a leak of calcium from the ER is one of the first events of the apoptotic process, even upstream of mitochondrial activation upon ER stress, Ca2+ fluxes were measured Bim expression. Cytosolic Ca2+ was measured by using Fluo-3AM, which binds to Ca2+ ions resulting in increased fluorescence. Time course experiment of Bim expression showed that Bim can be already found 8h after infection with the adenoviral vector and is increasing with time (figure 27). So, cytosolic Ca2+ elevations were determined at 8h, 16h and 24h by flow cytometry. As a positive control for ER tress, thapsigargin was used. This agent inhibits SERCA, and induces an unfolded protein response and increased Ca2+ levels in the cytosol. DU145-Bax neo, DU145-Bax Bcl-2actA or DU145-Bax Bcl-2cb5 cells were infected with either AdBimL or AdBimS or treated with thapsigargin and subjected to flow cytometric measurement (figure 27). Control cells with low Bcl-2, which were control treated or were transduced with AdBimL but grown under off conditions did not show any sign of calcium leak at any of the indicated time points. Turning on AdBimL expression did not result in major changes in these results. Even after 24h, only 10% of the cells were detected with elevated Ca2+ levels, implying that BimL cannot trigger Ca2+ release. The same cells were infected with AdBimS and displayed another picture. In DU145-Bax neo, mock treated and cells under off conditions did not show increased release of Ca2+ at any time points. But under on conditions, AdBimS provoked after 16h calcium fluxes in 19% of the cells. This number increased to 30% when the cells were exposed to AdBimS for 24h. Treatment of the cells with thapsigargin only, proved that ER stress could be induced in these cells. 8h after addition of thapsigargin, 15% of the cells were found to have released Ca2+. 16h later, 36% and at 24h 44% of the cells were affected.

↓61

Overexpressed Bcl-2actA targeted to the mitochondria did not change the situation for the cells. Cells, which served as a negative control or in which the expression of either of the Bim isoforms was kept suppressed in the presence of doxycyclin, did not show Ca2+ fluxes. Expression of BimL did not have an impact on the cells in respect to Ca2+ release. BimS on the other hand, triggered Ca2+ release from the ER in DU145-Bax Bcl-2actA cells. 8h post infection nothing had happened yet, another 8h later though in 17% of these cells higher amounts of Ca2+ were measured. Cells, which were collected 24h after transduction with AdBimS showed 20% of cells with calcium release. Once these cells were treated with thapsigargin as a positive control, they showed high release rates of 42% at 8h, 55% at 16h and finally 58% at 24h. DU145-Bax Bcl-2actA cells seemed to be more sensitive for the ER stress agent. The fact, that these cells are transfectants might be the answer for this phenomenon. Finally, DU145-Bax Bcl-2cb5 cells were investigated. Since these cells express Bcl-2 at the ER, they should be protected against ER stress and therefore should not release Ca2+. Mock treated cells and cells, which were infected with either of the adenovirus but grown in the presence of doxycyclin, could not be monitored for Ca2+ fluxes. Indeed, neither of the Bim variants could induce Ca2+ release at any of the indicated time points under on conditions. Bcl-2 targeted to the ER was able to protect the cells from ER stress generated by BimS. Moreover, even thapsigargin was not strong enough to break through the protective wall of Bcl-2. After the longest incubation time of 24h with thapsigargin, 20% of the cells were detected with elevated Ca2+ levels, revealing the functionality of the cells. These results point to a link between BimS and Bcl-2 at the ER, which has yet to be identified.

Figure 27: Ca2+-fluxes from the ER upon Bim expression

A, C: representative histograms of Ca2+ release measurement after 24h.
B, D: DU145-Bax cells were transduced with AdBimL or AdBimS and cultured with doxycyclin (off) or without doxycyclin (on) for indicated time points. As a positive control the cells were treated with 10µM thapsigargin (thaps). Increased cytosolic Ca2+ levels were measured by flow cytometry. Means +/- SD from three independent experiments.

4.10.1 Bim induces upregulation of ER stress proteins

According to the results obtained so far, it was speculated that Bim might associate with the endoplasmic reticulum where it would induce a stress response and initiate apoptosis. Therefore, detection of ER protein levels associated with ER stress responses might give some insights into the mechanisms by which Bim mediates apoptosis. Total lysates were collected 16h post infection with either AdBimL or AdBimS and subjected to Western blot analysis (figure 28).

↓62

Figure 28: Bim induces upregulation of ER proteins

A: Western blot analysis of the expression of ER associated proteins. DU145-Bax cells were transduced with AdBimL and cultured for 30h in the presence (off) and the absence (on) of doxycyclin. Cell lysates were blotted for the indicated proteins and developed with the appropriate antibodies.
B: DU145-Bax cells were transduced with AdBimS, and treated as described in A.

Detection of BimL (figure 28A) after transduction of AdBimL of all three cells lines confirmed that BimL was expressed under on conditions at comparable levels. None of the cell lines, which were mock treated or grown in the presence of doxycyclin, did show any BimL expression. Additionally, DU145-Bax neo, DU145-Bax Bcl-2actA and DU145-Bax Bcl-2cb5 were examined for Bcl-2 expression. Only cells that stably express Bcl-2, namely DU145-Bax Bcl-2actA and DU145-Bax Bcl-2cb5, revealed Bcl-2 expression. Since Bim activates Bax directly or indirectly expression levels might be changed upon Bim expression, especially in cells, which express Bcl-2 and might neutralize Bim. But Detection of Bax revealed that it was equally expressed, no matter which cell type was examined. Moreover, cells under on conditions showed the same Bax level as cell under control or off conditions. For the same reason Bak expression was examined. AdBimL induced upregulation of Bak in DU145-Bax neo cells under on conditions. While BimL expression also resulted in upregulation of Bak in DU145 BaxBcl-2actA cells, Bak proteins levels remained the same when Bcl-2 was targeted to the ER. CHOP is a transcription factor induced during ER stress and promotes apoptosis. BimL expression triggered upregulation of CHOP in all three cell transfectants, indicating that BimL may have caused ER stress. The slightly elevated CHOP level detected under off conditions was most likely set off by the adenovirus. Another ER stress protein is BAP31which is a resident integral membrane protein of the endoplasmic reticulum. It regulates the export of other integral membrane proteins to the downstream secretory pathway. ER stress and other apoptotic stimuli lead to generation of a p20 fragment, preferably cleaved by caspase-8. Also BAP31 was found to be up regulated, but only in DU145-Bax neo and DU145-Bax Bcl-2actA cells, whereas in DU145-Bax Bcl-2cb5 cells no such upregulation could be determined upon BimL expression. BiP/GRP78, another ER stress protein, is involved in protein folding and assembly, targeting misfolded protein for degradation, ER Ca2+- binding and controlling the activation of trans-membrane ER stress sensors. Further, due to its anti-apoptotic character, it is a component of the unfolded protein response. Upon BimL expression higher amounts of BiP were detected in all three cell transfectants. Cells, which were transduced with AdBimL in the presence of doxycyclin, presented a slightly stronger band than the corresponding control cells. This could be explained by a potential stress of infection with the adenovirus. The anti-apoptotic protein Mcl-1 is known to bind to Bim with high affinity, although its preferred binding partner is the BH3-only protein Noxa. Influence or binding of Bim to Mcl-1 cannot be ruled out, but nevertheless total Mcl-1 was equal in all three cell lines. Even expression of BimL did not influence Mcl-1 amounts. β-Actin detection was used as a loading control.

Detection of the same proteins was performed upon BimS expression. Transduction of AdBimS in all three transfectants under on conditions proved that AdBimS was equally expressed in all of them (figure 28B). Bcl-2 could not be found in DU145-Bax neo cells, but was present in the transfectants. Also the short version of Bim did not induce changes in the total amount of Bax expression. The cells expressed similar levels whether they were treated or not. Its multi-domain relative Bak on the other hand was upregulated in cells without Bcl-2 and also in cells, where Bcl-2 was localized at the mitochondria under on conditions. Control cells and cells, which were grown in the presence of doxycyclin showed unchanged Bak levels. In DU145-Bax Bcl-2cb5 cells expressing BimS no difference could be determined. The protein band for Bak in BimS transduced cells in the absence of doxycyclin was alike to the ones observed under control conditions. BimS expression resulted in up regulation of CHOP in cells where Bcl-2 was absent. But Bcl-2 targeted to the mitochondria or the ER could not prevent induction of this ER protein under on conditions for BimS expression. Detection of BAP31 showed that up regulation occurs in DU145-Bax neo cells upon BimS expression and when Bcl-2 is localized at the mitochondria. Just as for BimL, no upregulation of BAP31 could be monitored in cells that express Bcl-2 at the ER. BiP displayed a higher expression in all three cells types upon BimL or BimS expression under on conditions, regardless of the Bcl-2 status. AdBimS transduced cells showed a minor upregulation of the ER stress proteins CHOP and BiP under off conditions. Expression of BimS did not influence Mcl-1 on its expression level. All samples were quantified for equivalent expression. β-Actin detection served as a loading control for the Western blot analysis.


© Die inhaltliche Zusammenstellung und Aufmachung dieser Publikation sowie die elektronische Verarbeitung sind urheberrechtlich geschützt. Jede Verwertung, die nicht ausdrücklich vom Urheberrechtsgesetz zugelassen ist, bedarf der vorherigen Zustimmung. Das gilt insbesondere für die Vervielfältigung, die Bearbeitung und Einspeicherung und Verarbeitung in elektronische Systeme.
DiML DTD Version 4.0Zertifizierter Dokumentenserver
der Humboldt-Universität zu Berlin
HTML-Version erstellt am:
03.04.2012