The main function of Bax and Bak is to disrupt mitochondrial membrane integrity in order to release cytochrome c and other pro-apoptotic factors. It was proposed that they mediate this release either by binding to and modifying mitochondrial channel proteins such as VDAC or ANT (Marzo, et al., 1998 ) or by direct pore formation (Antonsson, et al., 1997 ; Eskes, et al., 1998 ). However, the activation of these two pro-apoptotic proteins is still controversial. Two different models have been discussed, the direct activation and the indirect activation. The direct activation model proposes that a subgroup of the BH3-only proteins called “activators” can activate Bak and Bax, whereas the remaining BH3-only proteins, called “sensitizers”, neutralize the anti-apoptotic proteins (Willis, et al., 2007 ). The indirect activation model suggests that Bak and Bax are held in check by being bound to anti-apoptotic proteins. BH3-only proteins displace Bak and Bax from the anti-apoptotic proteins to achieve their activation (Willis and Adams, 2005 ). But no interaction of Bak with BH3-only proteins could be detected. Also, Bax did not bind to BimEL and BimL in co-immunoprecipitation studies (Marani, et al., 2002 ; O'Connor, et al., 1998 ; Yamaguchi, et al., 2003 ). Even though BimS was found to weakly bind to Bax it might not be the physiological truth, since this BimS-Bax complex existed in detergents that alter Bax conformation. Additionally, a BimS mutant that was not able to bind to Bax anymore but could still interact with anti-apoptotic proteins and induce as much apoptosis as the wild type (Willis and Adams, 2005 ; Willis, et al., 2007 ). And finally, the ability of Bcl-2 or Bcl-xL to inhibit apoptosis induced by Bid, Bim or Bad correlates with their ability to bind BH3-only proteins but not Bax and Bak (Cheng, et al., 2001 ). For these reasons it is more likely that the indirect activation of Bax and Bak by the BH3-only proteins is the main process.
Therefore, the investigations described in this study focused in the first part on clarification if Bax and Bak were redundant when Bim is overexpressed. With the help of an antibody specific for an epitope at the N-terminus of Bax or Bak the conformational change was examined. Conformational change of Bax and Bak became evident when elevated fluorescence was measured. Conformational change of Bax could be detected in DU145-Bax cells upon BimL and BimS expression, respectively (figure 11). According to the specific Bak antibody BimL as well as BimS induced Bak activation in DU145-Bak cells as confirmed by the measurement of the conformational change of Bak. Whereas monomeric Bax is translocated from the cytosol to the mitochondria, Bak already resides at the mitochondria. Nevertheless, both proteins undergo conformational change and oligomerize at the mitochondria, which seems to be based on the activation by Bim.
DU145 cells stably expressing EGFP-Bax or EGFP-Bak were used to detect clustering of these two pro-apoptotic proteins. Non-treated DU145 cells, which stably expressed either of both fusion proteins, indicated the localization of Bax and Bak. In DU145 EGFP-Bax cells the fluorescence of the fusion protein appeared diffuse being concordant with the cytosolic localization. EGFP-Bak on the other hand presented a string-like pattern, pointing to its association at the mitochondria (figure 9). When DU145 EGFP-Bax cells were transduced with adenoviral BimL or BimS clustering of Bax could be observed by the punctuated pattern. Expression of either of the two Bim isoforms in DU145 EGFP-Bak cells also led to clustering of the fusion protein, noticeable by the EGFP spots. These results show that Bim induces activation of pro-apoptotic the multi-domain proteins leading to conformational change and clustering of these proteins, which might represent oligomerization.
Although it is not possible to conclude direct or indirect activation of Bax and Bak from this study, it can be said that Bim induces Bax and Bak activation. Bax and/or Bak are needed for execution of apoptosis. Several investigations with bax/bak double knockout cells clarified that without these two proteins apoptosis will not occur (Armstrong, 2006 ; Lindsten, et al., 2000 ; Wei, et al., 2001 ). Expression of BimL resulted in apoptosis in DU145-Bax and DU145-Bak cells, but not in the DU145 mock cells, confirming that induction of apoptosis needs Bax or Bak. BimS joined in on this effect, by inducing cell death in cells expressing Bax or Bak, but not the control cells. Indeed, a higher rate of apoptosis was measured upon BimS expression. Activation of Bax and Bak upon Bim expression ultimately led to apoptosis determined by measurement of DNA-fragmentation. A high percentage of DU145-Bax cells were found to be apoptotic after transfection with AdBimL and AdBimS, respectively. DU145-Bak cells showed similar rates of apoptosis upon BimL or BimS expression. These results demonstrate that BimL as well as BimS are able to kill in a Bax- and Bak-dependent manner.
After establishing the role of Bax and Bak in the Bim induced apoptotic pathway, the focus of this study turned to the relation of Bim to the mitochondria and the ER. The primary objective to use the DU145-Bax targeted Bcl-2 cell system was to assess whether Bim would activate the mitochondria or the ER to induce apoptosis. In transfectant cells Bcl-2 was targeted to the mitochondria by exchanging the C-terminal insertion sequence for an equivalent mitochondrial signal of (DU145-Bax Bcl-2actA) or to the ER by a transmembrane sequence (DU145-Bax Bcl-2cb5). DU145-Bax cells transfected with an empty vector served as a control (DU145-Bax neo). In the absence of Bcl-2 overexpression both BimL and BimS induced apoptosis in DU145-Bax neo cells, indicating their ability to activate Bax. However, in the presence of Bcl-2, Bim was at least partially blocked in its actions and therefore revealed its preferred pathway. Bcl-2 localized at the mitochondria reduced the apoptotic rate upon BimL expression. A similar inhibitory effect of Bcl-2 was found for the apoptotic rate induced by BimS. Even though the mitochondria were protected by Bcl-2 from being permeabilized, Bim was able to induce cell death by other mechanisms. Overexpression of either Bim isoforms in DU145-Bax cells with ER localized Bcl-2 did not end in apoptosis. The anti-apoptotic function of Bcl-2 prevented Bim-induced activation of ER mediated apoptosis. These results implicate that Bim might aim to the ER on its pathway once it is activated by an apoptotic stimuli. This consideration is supported by the report that Bim translocated to the ER in murine myoblast cells (C2C12) upon tunicamycin treatment. There, Bim stimulates the activation of ER stress induced apoptosis, which could be prevented by overexpression of Bcl-xL (Morishima, et al., 2004 ). Furthermore, Bim has been reported to respond to cell death downstream of ER stress and to be induced in response to ER stress. Additionally, it was shown that Bim targeted to the ER initiated apoptosis and can therefore act as a specific death signal in response to ER stress (Morishima, et al., 2004 ). Other studies reported that ER targeted Bcl-2 may sequester Bim, preventing it from interacting with other members of the Bcl-2 family (Egle, et al., 2004 ). This possibility was also stated by other reports, saying that Bim binds and inhibits ER localized Bcl-2 to regulate cell death and is therefore capable of regulating this anti-apoptotic protein at the ER (Kim, et al., 2004 ). It should be mentioned that Bcl-2 targeted exclusively to the ER is more restricted in its anti-apoptotic actions compared to Bcl-2 predominantly expressed at the mitochondria, suppressing cell death induced by ER stress agents and by c-Myc.
The secondary outcome of the overexpression of Bim in the DU145-Bax Bcl-2 system was the ability of membrane associated Bcl-2 to antagonize the killing effect of Bim. Whereas Bcl-2 targeted to the mitochondria only partially reduced Bim-induced apoptosis, localization of this anti-apoptotic protein at the ER almost completely cell death upon Bim expression. Similar results were found in other cell lines with different apoptotic stimuli. It was reported that Bcl-2 targeted to the mitochondria or the ER could protect cells from apoptosis depending on the cell death stimuli (Zhu, et al., 1996 ). Furthermore, it was also shown that Bcl-2 located at the ER was able to block cytochrome c release upon ER stress inducing agents (Hacki, et al., 2000 ).
Mitochondria play a key role in executing the intrinsic pathway, but are also involved in the extrinsic pathway that leads to apoptosis in response to intracellular stress signals. There two types of cells; in cell type I the cell death signal is propagated by caspase cascade initiated by the activation of large amounts of capsase-8 at the DISC following followed by the rapid cleavage of caspase-3 which in turn cleaves vital substrates in the cell. In type II cells no DISC is formed; instead caspase-8 cleaves the pro-apoptotic protein Bid to tBid leading to activation of the mitochondrial cell death pathway (Krammer, 2000 ). The permeabilization of the outer mitochondrial membrane marks the initiation of the intrinsic pathway followed by formation of the apoptosome and the activation of the initiator caspase-9. In the extrinsic pathway the mitochondria are stimulated by the activation of caspase-8 and subsequent cleavage of Bid. Moreover, rupture of the mitochondria can be amplified through caspase-3, caspase-8 and the cleavage of Bid in a feedback loop. The main event of mitochondrial activation is the loss of the membrane potential and consequently the release of cytochrome c and different pro-apoptotic factors. Mitochondria appear to undergo several changes in membrane structure and morphology before releasing cytochrome c. All these changes appear to influence the release of cytochrome c. Several studies indicate that transient openings in the mitochondrial permeability transition pore are important for this event. Bcl-2 family proteins may activate several processes in the mitochondria and ER during apoptosis, for example the reorganization of proteins within the intermembrane space before their passage across the outer mitochondrial membrane (Breckenridge and Xue, 2004 ). There are several models for how cytochrome c is released from the mitochondria. Most of them concentrate on the mechanism by which the outer mitochondrial membrane is permeabilized. The first model claims that non-specific pores at the inner mitochondrial membrane open to cause osmotic swelling of the mitochondrial matrix. This is followed by disruption of the outer membrane and the release of pro-apoptotic factors and cytochrome c (Petit, et al., 1998 ). Model number two says that the members of the Bcl-2 family regulate cytochrome c release. The anti-apoptotic proteins such as Bcl-2 and Bcl-xL were found to prevent efflux of cytochrome c from the mitochondria by inhibition of Bax or Bak function or by binding of BH3-only proteins (Jurgensmeier, et al., 1998 ; Kluck, et al., 1997 ; Yang, et al., 1997 ). In contrast, Bax and Bak stimulated the opening of VDAC and allowed cytochrome c to pass (Shimizu, et al., 1999 ). Thus activation of Bax or Bak appears to be a predominant gateway to mitochondrial activation. The third model proposes a mitochondrial permeability transition-independent mechanism of cytochrome c release. The release may occur by modulation of the mitochondrial volume leading to swelling of this organelle. In this scenario the mitochondria remain intact and maintain their membrane potential (Gogvadze and Orrenius, 2006 ). In the fourth model caspase-2 is the main player mediating cytochrome c release. Caspase-2 was shown to be associated with the mitochondria (Susin, et al., 1999 ) and to disrupt the inner mitochondrial membrane (Von Ahsen, et al., 2000 ). Interestingly, caspase-2 does not need its catalytic activity for this function (Robertson, et al., 2004 ). Although it was not attempted in this study to verify one or possible combinations of any of the models mentioned, cytochrome c efflux was investigated in respect to BimL and BimS. Immunostaining of cytochrome c in DU145-Bax cells revealed a leak of cytochrome c upon Bim expression (figure 14). Infection with AdBimL, staining of cytochrome c and the mitochondria showed in the overlay that cytochrome c had been released from the mitochondria. In mock cells, which are Bax negative there was no sign for cytochrome c depletion upon BimL expression. In general, the same patterns were observed upon BimS expression. But there were more cells detected with cytochrome c depleted mitochondria. In the overlay, the strength of BimS became visible as it also induced cytochrome c release in some of the DU145 mock cells. This was confirmed by Western blot analysis, where cytochrome c could be detected in the cytosolic fraction upon expression of both Bim variants in DU145-Bax cells (figure 16). Release of cytochrome c occurred early, already 14h post infection with AdBimL cytochrome c could be detected in the cytosol. The cytosolic fraction of DU145-Bax cells was found to contain cytochrome c 10h post infection with AdBimS. This rapid activation of the mitochondria shows an effective mode of action of Bim and might with regard to the time course of protein expression indicate that lower amounts of BimS as compared to BimL suffice to trigger the mitochondrial pathway.
The mitochondrial (intrinsic) pathway consists of release of cytochrome c and the breakdown of the mitochondrial membrane potential. Release of cytochrome c triggers the binding of dATP to Apaf-1 and the formation of the apoptosome, followed by the activation of caspase-9. Activation of caspase-9 was shown by Western blot analysis and flow cytometric measurements (figure 23, 25). In this study it could be observed that transfection of Bcl-2 and targeting of Bcl-2 to the mitochondria blocked the loss of the mitochondrial membrane potential (figure 22), suggesting that Bim induces the mitochondrial pathway. DNA fragmentation was also partially inhibited by Bcl-2 (figure 20). It can be concluded that the signals leading to mitochondrial activation are at least partially mediated upstream of the mitochondria and can be inhibited by anti-apoptotic proteins such as Bcl-2, e.g. targeted to the ER. This was found to be the case for release of cytochrome c but not for loss of the mitochondrial membrane potential.
It was shown by Western blot analysis that Bcl-2 blocked the Bim induced leak of cytochrome c from the mitochondria. When Bcl-2 was targeted to the mitochondria, only a small amount of cytochrome c was released to the cytosol upon Bim expression (figure 25). A similar effect was achieved when Bcl-2 was present at the ER indicating that ER-mediated apoptosis may lead to secondary activation of the mitochondrial pathway. Both isoforms were blocked in their actions by Bcl-2 and this prevented the release of cytochrome c. Whether Bcl-2cb5 inhibited the effect of Bim by directly binding to this BH3-only protein or whether it blocked the multi-domain proteins Bak and Bax to become activated and therefore inhibited the pore formation at the mitochondrial membrane could not be clarified at this point. Of note, no protein interaction were studied here that might have elucidated the mode of BimL or BimS induced Bak/Bax activation.
There are data describing that cytochrome c release occurs before loss of the mitochondrial membrane integrity (Goldstein, et al., 2000 ). Others claim that cytochrome c can be released independently from breakdown of the mitochondrial membrane potential and activation of the caspases (Bossy-Wetzel, et al., 1998 ). Further it was reported that this process is conducted by the permeability transition pore and can be inhibited by overexpression of Bcl-2 (Susin, et al., 1996 ). It was also proposed that, without cytochrome c, apoptosis is attenuated and, consequently, the apoptosome is not obligatory for stress-induced apoptosis, but only acts as a caspase amplification system that is more important in certain cell types than others (Li, et al., 2000 ; Von Ahsen, et al., 2000 ).
In a next step, investigations concentrated on the question whether Bim-induced mitochondrial activation was dependent of Bax, Bak and inhibited by Bcl-2. In Bax negative mock transfectant cells, no cells were detected with mitochondrial permeability shift. Then again, in DU145-Bax cells, BimL induced breakdown of the mitochondrial potential, and BimS even twice as much. Approximately the same numbers were found in Bax-deficient DU145-Bak cells and their mock transfectants. Apart from the intensity, both isoforms were able to activate the mitochondrial pathway. In DU145-Bax re-expressing and DU145-Bak cells Bim induced breakdown of the mitochondrial membrane potential. In cells, which were either Bax negative or with low Bak expression, no cells could be determined with a mitochondrial shift, suggesting that Bim induces disruption of the mitochondria by activation of Bax or Bak. However, some reports say that Bax and Bak do not play a redundant role in activation of the mitochondria and cannot substitute for each other. BH3-only proteins were shown to induce apoptosis either via Bax or Bak. Nbk was found to kill cells in an only Bax-dependent manner (Gillissen, et al., 2003 ). Nbk is not the only BH3-only protein that is restricted to either Bax or Bak, also other members of this group depend on one of these two pro-apoptotic proteins to induce apoptosis. Therefore it was interesting to discover that Bim does not favour one of these two multi-domain proteins but is equally activating Bax and Bak to cause mitochondrial perturbation.
Studies with Bcl-2 targeted to the ER verified that it protects the mitochondria from a distance. There are established data suggesting that it can inhibit disruption of mitochondrial membrane potential (Annis, et al., 2001 ), release of cytochrome c from mitochondria (Hacki, et al., 2000 ), and oligomerization of Bax (Thomenius, et al., 2003 ). It has been demonstrated that Bax and ER targeted Bcl-2 do not interact during apoptosis, although ER-Bcl-2 inhibited apoptosis (Annis, et al., 2001 ). In another study, Bcl-2 at the ER was able to inhibit the oligomerization of a Bax mutant, which was constitutively localized on the mitochondria (Thomenius and Distelhorst, 2003 ). Since Bcl-2 localized at the ER and mitochondrial Bax are spatially separated, it is unlikely that they interact. However, the data obtained in this study together with the data generated by other groups showed that the mitochondrial cell death pathway induced by Bim could be prevented by Bcl-2. There is also evidence that BH3-only proteins may directly affect the mitochondria (Sugiyama, et al., 2002 ). These findings propose an alternative model in which Bcl-2 lies upstream of BH3-only proteins and impede them in activating Bax and inducing apoptosis. This model is also supported by recent reports that ER-targeted Bcl-2 prevents the mitochondrial localization of the BH3-only protein Bad (Thomenius and Distelhorst, 2003 ).
Expression of Bim in cells with mitochondria-localized Bcl-2 showed a decreased level of DNA-fragmentation (figure 20). The following investigation was supposed to answer the question whether the protection of the mitochondrial membrane potential by Bcl-2 was the reason for the reduced rate of apoptosis. 48h post infection with AdBimL high loss of the mitochondrial membrane potential was observed in the DU145 Bax neo cells. Bcl-2 targeted to the mitochondria reduced this effect by more than half, showing that Bcl-2 was able to protect the cells from BimL induced activation of the mitochondrial pathway. On the other hand, Bcl-2 localized to the ER could not save the mitochondria form losing their membrane potential. In these cells BimL induced a mitochondrial permeability shift almost as much as in cells without Bcl-2. Expression of BimS had a similar but stronger effect. Nearly all DU145-Bax neo cells infected with AdBimS were detected with mitochondrial permeability shift. Targeting Bcl-2 to the ER resulted in loss of the mitochondrial membrane potential for the majority of the cells. At first glance, this result raised doubts since there was no enhanced DNA fragmentation measured in these cells. But a possible solution came from other reports that some cells remain viable after permeabilization of the outer mitochondrial membrane (Holinger, et al., 1999 ). Therefore breakdown of mitochondria might not be necessary or sufficient for apoptosis (von Ahsen, et al., 2000 ). Only Bcl-2 at the mitochondria could protect the cells from losing their mitochondrial integrity upon BimS expression. It seems that the localization of Bcl-2 is important for its inhibitory effect, but the single fact of overexpressing Bcl-2 was not enough to fully neutralize Bim-mediated apoptosis induction at the mitochondria.
It was reported that Bcl-2 or Bcl-xL can block mitochondrial membrane permeabilization induced by ectopically expressed BH3-only proteins or by Bax and/or Bak (Gross, et al., 1999 ). Therefore, it is possible that the apoptotic pathway is divided in two processes. Activation of pro-apoptotic proteins results in an initial release of approximately 10% of the cytochrome c from the intermembrane space followed by caspase activation. The second mechanism depends on the permeability transition leading to matrix swelling and consequent remodelling of the mitochondria, whereby all the cytochrome c is dumped (Bernardi and Azzone, 1981 ). These findings may underline the importance of the ER in the Bim pathway in the present study. Even though the mitochondria lost their membrane potential upon Bim expression (figure 22) in cells where Bcl-2 was targeted to the ER, these cells did not undergo apoptosis upon Bim expression (figure 20). It seems that the ER has to be accessible for Bim in order to induce apoptosis. If the ER is protected by Bcl-2, Bim can activate alternative pathways such as the mitochondrial cell death pathway as is indicated by the loss of the mitochondrial membrane potential, but this does not seem sufficient to lead to cell death. Further it might also support the model of the mitochondrial remodelling. This model implies that the permeabilization of the outer mitochondrial membrane does not lead to complete depletion of cytochrome c but requires the remodelling of the inner membrane. The structural reorganization of the mitochondria seems to be dependent of the permeability transition following activation of pro-apoptotic proteins to guarantee that cytochrome c release during apoptosis is quick and absolute. Studies have demonstrated that pro-apoptotic Bcl-2 family members remodel the mitochondrial structure (Scorrano, et al., 2002 ). The fact that Bcl-2 localized at the mitochondria did not completely inhibit apoptosis might support the theory that there are some alternative pathways and more importantly that the ER is available for Bim. Whereas Bcl-2 solidly prevented loss of the mitochondrial membrane potential, it did not inhibit apoptosis in the same strong way upon Bim expression. This could be another hint that the ER is an important station for Bim to fulfil its apoptotic function and that it does not rely exclusively on mitochondria as it is proposed for other BH3-only proteins.
Interestingly, transfected and targeted Bcl-2 could, at least partially, inhibit mitochondrial cell death but the processing of the caspases was merely blocked. Although this result was obtained by two different methods, the question needed to be answered how these caspases could be activated if cytochrome c release is prevented. It should be expected that caspase activation is completely blocked if there is no cytochrome c release. But it can be speculated that overexpressed Bcl-2 is able to inhibit cytochrome c release but nevertheless small amounts of cytochrome c could reach the cytosol, which is enough to activate the caspases to a certain extent. Small amounts of cytochrome c are not detectable in Western blots. And, as already mentioned the assumption that only a small amount of the cytochrome c resides in the intermembrane space and most of it is found in the matrix (Bernardi and Azzone, 1981 ) might lead to the idea that permeabilization of the outer membrane and the liberation of few of these molecules is sufficient to activate the caspase cascade. For these reasons it is plausible that overexpression of Bcl-2 cannot fully inhibit cytochrome c release.
It should also be taken to account that in the case of BimS, inhibition of caspase-8 led to a partial protection of the mitochondria when Bcl-2 was overexpressed in the cell, regardless of its localization. Addition of the caspase-8 inhibitor reduced the number of cells with a mitochondrial permeability shift by half (figure 26). It was reported that caspase-8 can be recruited to the ER (Ng, et al., 1997 ) and can also be found at the outer mitochondrial membrane (Chandra, et al., 2004 ). Bcl-2 localized at the mitochondria diminished the amount of cells with lowered mitochondrial potential although BimS was present. Additional treatment with the caspase-8 inhibitor boosted this effect, only small amounts of cells were detected, which had lost the mitochondrial membrane potential. Based on these findings it could be speculated that BimS directly or indirectly activates caspase-8 to induce the breakdown of the mitochondrial membrane potential. Caspase-8 does not seem to be a target for BimL as inhibition of this caspase did not have an effect on mitochondrial activation (figure 26), but it can not be ruled out that caspase-8 is involved in its pathway through an amplification loop, e.g. through cleavage of Bid to tBid. Further aspects of caspase-8 and Bim will be discussed below.
Various caspases are activated in different manner at different points of the apoptotic pathway, but nevertheless, activation of the caspases marks a crucial step in apoptosis. Mitochondrial membrane permeabilization and release of pro-apoptotic factors including cytochrome c, is involved in the activation of the caspases. However, there are debates about whether permeabilization of the outer mitochondrial membrane itself relies on or can occur independently of caspase activity (Breckenridge and Xue, 2004 ). Intermembrane space proteins are released at different time points following the apoptotic stimulus and some depend on caspases. This finding suggests that decisions are made upstream of mitochondria on whether and which pro-apoptotic factor is released form the intermembrane space (Breckenridge and Xue, 2004 ).
Emphasis of investigations in this thesis was put on caspase-3, caspase-9 and caspase-8. Caspase-9 is the key initiator caspase of the mitochondrial pathway and is recruited by APAF-1 into the apoptosome. Active caspase-9 cleaves thereby and activates caspase-3. Caspase-8 is essential for the death receptor pathway and mediates the mitochondrial amplification loop. Caspase-3 represents the main downstream effector caspase and can be activated by either caspase-9 or caspase-8.
The activation of the three mentioned caspases was measured indirectly by detection of their cleavage products by use of Western blot analyses. The initiation of the caspase cascade was observed in DU145-Bax and DU145-Bak cells, presenting the capability of Bim to mediate cell death through both Bax and Bak (figure 16). DU145-Bax cells transfected with ER-targeted Bcl-2 showed resistance to caspase processing upon Bim expression (figure 23). These results were consistent with the measurements of DNA-fragmentation, where it was shown that these cells do not undergo apoptosis upon Bim stimulation in consequence of inhibition of the ER-mediated death pathway by Bcl-2cb5. Based on this finding it was concluded that the first station of Bim on this pathway is the ER, from where the death signal is transported to other compartments (figure 20). Further confirmation was achieved by flow cytometric measurements. Both, activation and inhibition of caspase-9, -3 and-8 were measured. To measure caspase activation, a FITC labelled caspase specific peptide substrate was used that binds to the activated enzyme. In all DU145 transfectants, an activation of the caspases could be measured (figure 23). The only exception was DU145-Bax Bcl-2cb5 cells, where none of the caspases were activated due to inhibition by Bcl-2cb5. The inverse picture was seen using caspase inhibitors individually for each of the three caspases. The apoptotic rate was dramatically reduced once the caspase inhibitor zVAD-fmk was added to the cells infected with AdBimL or AdBimS (figure 24). Lesser levels of inhibition were exerted by zDEVD-fmk and zIETD-fmk inhibition with zLEHD-fmk inhibition showing the least effect.
These inhibition studies of caspase-9 revealed that caspase-9 participates in Bim induced apoptosis in DU145-Bax and DU145-Bak cells (figure 15). Addition of zLEHD-fmk reduced apoptosis in both cell lines upon BimL expression by approximately one third. Upon BimS expression and treatment with caspase-9 inhibitor almost four times less apoptosis was measured in DU145-Bax as well as DU145-Bak cells. Yet, presence of Bcl-2 interfered with caspase-9 activation in principle, so that use of zLEHD-fmk did not show any major effect, no matter at which organelle Bcl-2 was localized. Measurement of caspase-9 activity upon Bim expression confirmed these data. Whereas cells without Bcl-2 showed elevated activation of caspase-9, Bcl-2 at the mitochondria diminished active caspase-9 by half. Bcl-2 targeted to the ER restrained the activation of caspase-9 even more. Compared to DU145-Bax neo, only one third of the cells were detected with activated caspase-9. These data suggest that both Bim variants seem to induce caspase-9 activation to the same extent. Caspase-3 is cleaved and activated by caspase-9 in the mitochondrial pathway. Therefore, this caspase should be a part of the Bim induced apoptotic pathway. Bax-dependent cell death was reduced in half of the cells when they were, additionally to the adenovirus, treated with the caspase-3 inhibitor. Similar effects were found in DU145-Bak cells upon Bim expression and addition of zDEVD-fmk. Caspase-3 activity was also determined in context to Bcl-2 overexpression. Roughly, for both Bim isoforms cells overexpressing Bcl-2 resulted in a 50% reduction of the caspase-3 activity as compared to cells with no Bcl-2 overexpression. These results do not necessarily mean that caspase-3 has a bigger role than caspase-9 in respect to Bim. Higher caspase activation rates and stronger inhibition of apoptosis with zDEVD-fmk could be the result of an amplification loop. Caspase-8 connects the extrinsic and intrinsic pathway via Bid and can amplify the mitochondrial activation. Therefore, involvement of caspase-8 in the Bim pathway was also investigated. DU145-Bax cells treated with zIETD-fmk to inhibit caspase-8 presented upon expression of either of the Bim isoforms the same apoptotic rate as the ones treated with zDEVD-fmk. The same experiments in DU145-Bak revealed comparable apoptotic rates as detected in the DU145-Bax cells. Hence, it can be said that caspase-8 has a role in Bim induces apoptosis. So far in the work presented, both Bim splicing variants had a similar impact on cells except that BimS was more important in induction of cell death related events. Measurement of casapse-8 activation though, suddenly presented a difference between the two isoforms, which might have a differential impact on their pathway. Expression of BimL in DU145-Bax neo cells showed only a small quantity of cells with caspase-8 activity. Even less cells were detected with caspase-8 activity when Bcl-2 was present regardless of the localization. Expression of BimS in DU145-Bax neo cells demonstrated more than double as much cells with active caspase-8 as induced by BimL. But much less Bcl-2 transfectants with caspase-8 activation were observed when expressing BimS. These results gave the first clue that caspase-8 has a role in BimS induced apoptosis, which does not seem to be case for BimL. In consideration of reports that caspase-8 is located at the mitochondria and might have a great influence on mitochondrial activation, it was investigated whether inhibition of caspase-8 would have an effect on the loss of the mitochondrial integrity induced by Bim. Transduction of AdBimS in DU145-Bax neo cells led to 61% of cells with lowered mitochondrial membrane potential which did not change when zIETD-fmk was added to the cells. But overexpression of Bcl-2 and inhibition of caspase-8 prevented, however, the cells from loss of the mitochondrial membrane potential. In this constellation the caspase-8 inhibitor reduced the cells detected for mitochondrial permeability shift by half, no matter whether Bcl-2 was targeted to the mitochondria or the ER. Therefore, it may be assumed that Bcl-2 participates in the mechanism of caspase-8 activation initiated by BimS. Additional clues were given by the report that activation of caspase-8 might be achieved by a cytochrome c dependent way. It was proposed that caspase-6, an effector caspase like caspase-3, is the only caspase with the ability to activate caspase-8 after cytochrome c release. In this case, caspase-8 would not need interaction with FADD and no formation of the DISC to be processed (Cowling and Downward, 2002 ). But as already discussed above no cytochrome c release could be detected by Western blot analysis in cells overexpressing Bcl-2, which makes this theory less likely, unless the minor part of cytochrome c kept in the intermembrane space would be enough to trigger this pathway. To make this point clear, further investigations especially about caspase-6 as the activator of caspase-8 would be necessary, but were not done in this work.
Detection of processed caspase-9, -3 and -8 by Western blot analysis revealed that these caspases are activated upon both BimL and BimS expression independently of the Bcl-2 expression level (figure 25). One possibility for the activation of these caspases despite of the presence of Bcl-2 could be a Bcl-2 insensitive feedback loop, where caspase-9 is activated, which in turn leads to activation of caspase-3. This effector caspase on the other hand is able to activate caspase-8. The other explanation could be taken from the proposition that only about 15% of the cytochrome c is located in the intermembrane space and the rest resides in the matrix (Bernardi and Azzone, 1981 ). Given that the inner mitochondrial membrane might not be permeabilized upon Bim expression, this small amount of cytochrome c might be able to induce caspase-9 activation and thereby trigger the activation of caspase-3 and -8. It should be noticed, that for the same time point (30h) there was more cleavage product detected for caspase-8 in cells transduced with AdBimL than for cells expressing BimS. It could be speculated that due to the strength of BimS and the fact that it is not post-translationally modified, active caspase-8 is already degraded at this time point.
All the results of the caspase investigations point to the conclusion that Bim requires mostly caspase-9 and -3 to execute apoptosis. BimS additionally seems to involve caspase-8 into these steps of cell death induction. Interestingly, blocking caspase-8 did not affect the activation of the mitochondrial and therefore the breakdown of the mitochondrial membrane potential by BimS. BimS induced permeabilization of the mitochondria could, however, be partially inhibited by usage of the caspase-8 inhibitor. DU145-Bax cells containing Bcl-2actA and also cells with Bcl-2cb5 showed a reduced number of cells with mitochondrial membrane potential loss. It seems that at this very early part of apoptosis induction BimS is able to activate an alternative, parallel apoptotic pathway where caspase-8 is activated upstream via the ER to trigger secondary activation of the mitochondrial pathway.
The unfolded protein response is mediated through three ER transmembrane receptors, pancreatic ER kinase (PKR)-like ER kinase (PERK), activation transcription factor 6 (ATF6), and inositol-requiring enzyme 1 (IRE1). In healthy cells, all three ER stress receptors are maintained in an inactive state through binding to the ER chaperone BiP/GRP78. On accumulation of unfolded proteins, BiP dissociates from the three receptors, which leads to their activation and triggers the unfolded protein response. The unfolded protein response provides protection form cell death by reducing the accumulation of unfolded proteins and restores normal ER functioning (Schroder and Kaufman, 2005 ). However, when ER stress is excessive and prolonged cell death mechanisms are activated. Several mechanisms have been proposed for ER-induced cell death, including direct activation of proteases, kinases, transcriptions factors and Bcl-2 family proteins. Several investigations of ER stress induced apoptosis were performed in murine systems focusing on caspase-12, which is specifically cleaved upon induction of ER stress (Nakagawa, et al., 2000 ). There is no evidence of functional caspase-12 in human, although its mRNA was detected, but due to a frame shift mutation and disruption by a stop-codon and it is enzymatically inactive (Saleh, et al., 2004 ). Human caspase-4 was proposed to take the place of murine caspase-12 in humans in ER stress mediated cell death (Hitomi, et al., 2004 ). Caspase-4 was predominantly found at the ER membrane, but also at the mitochondria, which might imply additional ER-independent functions. Processing of caspase-4 was observed after exposure to ER stress inducing agents such as thapsigargin and tunicamycin in human neuroblastoma cells but not upon inducers of mitochondria-dependent dell death such as UV-irradiation and DNA damaging agents. However, caspase-4 knock down in HeLa cells had little effect on apoptosis induced by ER stress, implying that the importance of this caspase in ER stress is cell type dependent and so not sufficiently understood. In the investigations presented here, expression of Bim and inhibition of caspase-4 did not fundamentally block DNA-fragmentation (figure 24), questioning its suggested role, at least in Bim induced apoptosis. When Bcl-2 was not present, expression of BimL and treatment with caspase-4 inhibitor a minor reduction in cell death was observed. In the same settings with BimS a similarly decreased apoptotic rate was measured. In cells, where Bcl-2 was targeted to the mitochondria, the caspase-4 inhibitor to an only minor extent diminished apoptosis, when these cells were transduced with AdBimL or AdBimS. These results reveal that caspase-4 is not the key caspase in ER stress-induced apoptosis, its inhibition helped to decrease cell death, but not in a major way. In the presence of Bcl-2 the effect of the caspase-4 inhibitor was even smaller, probably because overexpressed Bcl-2 anyhow partially inhibited apoptosis. When Bcl-2 was targeted to the ER, no effect of the caspase-4 inhibitor could be detected, neither upon BimL nor BimS expression. This was not surprising since these cells did not show relevant apoptosis induction or show by analyses of DNA-fragmentation upon Bim expression. Further studies are needed to clarify a potential role of caspase-4 in Bim induced ER stress. The influence of capase-4 inhibition on mitochondrial membrane potential loss and detection of caspase-4 cleavage products could give some insight on participation of caspase-4. A potential candidate for the role as the mediator of ER stress induced apoptosis is caspase-8. It is thought to act as initiator caspase at the ER and to contribute to mitochondria-ER crosstalk. It was found to localize to the outer mitochondrial membrane in an active state. Caspase-8 localized at mitochondria may activate caspase-3, which then processes caspase-9 and thus might contribute to the amplification loop. As another substrate of casapse-8, Bap31 (Bcl-2 associated protein 31) was identified. It was observed in a regulatory complex at the ER with Bcl-2 and caspase-8 (Ng, et al., 1997 ). Cleavage of this integral ER protein generates a p20 fragment that has pro-apoptotic features. The p20 fragment, which remains in the ER membrane, may mediate mitochondrion-ER crosstalk through a Ca2+-dependent mechanism (Chandra, et al., 2004 ). It causes the release of Ca2+ from the ER, which is in turn taken up by mitochondria and triggers the release of cytochrome c and caspase activation (Rao, et al., 2004 ). Overexpression of Bax or Bak led to Ca2+ efflux from the ER, Ca2+ influx into the mitochondria resulting in cell death, that can be prevented by Bcl-2 (Nutt, et al., 2002 ). The expression of Bcl-2 decreased the amount of Ca2+ that could be released from intracellular stores, regardless of the mode of store depletion, the cell type or the species from which Bcl-2 was derived. The response to ER stress agent thapsigargin revealed that Bcl-2 increased the permeability of the ER membrane. Bcl-2 is said to inhibit apoptotic mechanisms downstream of cytochrome c, probably at the level of the ER (Foyouzi-Youssefi, et al., 2000 ). The ER-mitochondria crosstalk might be used for mitochondrial amplification of ER initiated apoptosis. Considering the findings described, the following pathway might apply for BimS: Active caspase-8 in the outer mitochondrial membrane cleaves and activates Bap31 to p20, upon which Ca2+ is released from the ER. Liberated Ca2+ is absorbed by the mitochondria causing their permeabilization, which might lead to apoptosis. Several studies have demonstrated that members of the Bcl-2 family also integrate into the ER membranes, where they regulate the transfer of ER Ca2+ to mitochondria and ER stress signals (Oakes, et al., 2006 ). Whereas anti-apoptotic members of the Bcl-2 family can prevent alteration of Ca2+ homeostasis, pro-apoptotic members promote Ca2+ mobilization from the ER to mitochondria during apoptosis, perhaps by regulating of the activity of the ER inositol trisphosphate receptor. This might be important for the mitochondrial permeability transition pore opening and intermembrane space protein release (Breckenridge and Xue, 2004 ; Mathai, et al., 2005 ; White, et al., 2005 ).
Considering the data obtained in this thesis, the just mentioned process does not seem to be main pathway, but lead to an amplification and acceleration of the death signal. For further investigations about the role of Bim in ER stress, Bim should be expressed in the presence of the caspase-8 inhibitor and subsequent detection for Bap31 should not show the p20 fragment. This would be a further clue that at least BimS mediates apoptosis by inducing ER stress. This hypothesis is supported by the finding that upon BimS expression DU145-Bax neo and DU145-Bax Bcl-2actA cells showed upregulation for Bap31, but not in the cells, that express Bcl-2 at the ER (figure 28). This confirms that Bcl-2 is able to inhibit induction of ER stress when it is localized at the ER. Further, this is in line with the finding that DU145-Bax neo and DU145-Bax Bcl-2actA cells displayed Ca2+ release upon BimS expression (figure 27). Surprisingly, Western blot analysis of these cells transduced with AdBimL presented under on conditions the same picture as described for BimS. It might therefore be speculated that BimL does also target the ER, leading to upregulation of Bap31, but fails to induce ER stress and thus also calcium release. The slight induction of Bap31, which was detected under off conditions might come from the transduction of the cells with the adenovirus. Supplementary clues that this hypothesis of BimS but not BimL acting through the ER pathway were given by data showing that capase-8 is activated upon BimS expression but not in cells infected with AdBimL (figure 23). Moreover, perturbation of the mitochondria could be diminished by capase-8 inhibition when BimS was expressed in the cells, which was not the case upon BimL expression (figure 26). Overexpression of the p20 Bap31 fragment triggered Ca2+ release and sensitized mitochondria for caspase-8 induced apoptosis (Breckenridge, et al., 2003 ). Activation of the ER and the mitochondria might nevertheless be partially caspase-8 independent, although the mechanisms described for Bim so far, relies on caspase activation. Caspase-8, Bap31 and Bcl-2 or Bcl-xL that were found to form a complex (Ng, et al., 1997 ). Formation of this complex can be blocked by the BH3-only protein Spike (Mund, et al., 2003 ).
Another ER stress-induced cell death modulator is CHOP, a transcription factor induced during ER stress (Kaufman, 1999 ; (Wang, et al., 1996 ). Lack of CHOP provides partial resistance to ER stress-induced apoptosis (Zinszner, et al., 1998 ). CHOP promotes ER stress mediated apoptosis by repressing Bcl-2 gene expression, which increases the proportion of pro-apoptotic proteins in the cell and facilitates their activation. How CHOP induces apoptosis though, is unclear. While capable of inducing apoptosis and contributing to cell death in several scenarios involving ER stress, CHOP is not essential for cell death induced by ER stress (Harding, et al., 2003 ). Overexpression of CHOP has been shown to induce apoptosis, which was linked to the activation and mitochondrial translocation of Bax. Vice versa, overexpression of Bcl-2 could block CHOP induced apoptosis (McCullough, et al., 2001 ). This could not be confirmed by overexpression of Bim in DU145 cells. Under on conditions, both BimL and BimS induced upregulation of CHOP independently of the Bcl-2 status. A further ER protein involved in ER stress is the chaperone protein GRP78/BiP (immunoglobin heavy chain binding protein). It maintains ER function, facilitates protein folding and protects cells from toxic insults, hence has anti-apoptotic abilities. BiP is localized inside the ER lumen, but it is also found on the cell surface, especially in prostate cancer cells (Mintz, et al., 2003 ). BiP functions as a main regulator of the unfolded protein response by binding to and preventing the activation of all three stress sensors IRE1, PERK, ATF6 (Kaufman, 1999 ). Upon ER stress these sensors are released form BiP and become activated. Additionally, BiP inhibits apoptotic signals at least in part by blocking caspase activation (Rao, et al., 2002 ). All in all, BiP has protective function against cytotoxic insults, it can protect cells against apoptosis caused by disturbance of ER homeostasis. Detection of BiP by Western blot analysis demonstrated increased levels of this protein upon expression of both Bim isoforms. No differences were found in-between cells with or without Bcl-2 overexpression. The strong expression of BiP might indicate that this protein is induced and regulates ER stress induced by Bim. BiP is also believed to regulate ER Ca2+ storage (Lievremont, et al., 1997 ). Measurement of elevated cytosolic calcium revealed that the short form of Bim induced calcium release form the ER in a time dependent manner. DU145-Bax neo transduced with AdBimS showed increased calcium levels making regulation of calcium fluxes by BAP31 in this setting unlikely. Within 8h the cytosolic Ca2+ levels increased considerably upon BimS expression in these cells. Bcl-2 localized at the mitochondria partially interfered with BimS induced calcium leak. Noticeable, but still less calcium fluxes were measured in DU145-Bax Bcl-2actA cells increasing slightly in the same time frame. It was therefore not surprising that DU145-Bax Bcl-2cb5 cells, which express Bcl-2 at the ER were resistant to BimS induced calcium release Consequently, when the ER was blocked by Bcl-2, BimS was not able to promote leakage of calcium from this organelle. Expression of BimL did not induce calcium release in none of the three cell transfectants at any time point. Taking in account that BiP can block ER signalling proteins, Ca2+ release and caspase activation, the long variant of Bim might not be able to counteract. In various tumour cells, among them prostate cancer cell line DU145, which were used in the present studies, highly augmented BiP expression was determinated (Arap, et al., 2004 ; Misra, et al., 2005 ). The ability of BimS to induce stronger caspase-8 activation as compared to BimL might lead to calcium release from the ER. The second reason might be the strength of BimS to induce apoptosis in comparison to BimL, which was constantly evident throughout the investigations. Therefore, BimS might be able to overcome a putative protective effect of BiP.
Bcl-2 family members including Bcl-2 Bcl-xL, Bax, Bak and Nbk have been shown to associate with the ER, suggesting that Bcl-2 family proteins operate at the ER to regulate calcium homeostasis and apoptotic cell death (Breckenridge, et al., 2002 ; Ng, et al., 1997 ).
Although further investigations must be conducted to elucidate the molecular basis and interactions of BimL and BimS, the following scenario is supported by the data provided in this study: A cell death signal activates BimL resulting in its release from the microtubule and translocation to the ER. There it activates an ER-associated apoptosis signal that results in activation of Bak and Bax. It is unclear whether this occurs at the mitochondria or the ER. At the mitochondrion Bak and Bax mediate loss of the mitochondrial membrane potential and the release of cytochrome c. The liberation of cytochrome c activates the apoptosome and triggers the activation of caspase-9, which mediates the cleavage of caspase-3. It could not be clarified whether ER and mitochondrial events happen sequentially or represent two separate pathways. Although is was recently shown by cell survival assays that Bim is the most critical initiator of ER stress induced apoptosis (Puthalakath, et al., 2007 ), the mechanism of ER/mitochondria crosstalk is nevertheless enigmatic.
BimL can be antagonized by Bcl-2, although this anti-apoptotic protein cannot entirely block the induction of apoptosis. The prevention of cell death by Bcl-2 seems to depend on its localisation. Expressed at the mitochondria, Bcl-2 could only partially inhibit apoptosis. While cytochrome c release from mitochondria was inhibited by both Bcl-2actA and Bcl-2cb5, only Bcl-2actA interfered with loss of the mitochondrial membrane potential. The reason for this might be that BimL still manages to induce mitochondrial activation to a certain extent, which is reflected in the reduced but yet detectable DNA-fragmentation. When Bcl-2 resides at the ER, BimL was not able to conduct apoptosis, suggesting that ER-stress induced apoptosis is parallely activated to the mitochondrial apoptosis pathway.
BimS seems to mainly use the same pathway as BimL by activation of the mitochondrial cell death pathway. Just as for BimL, Bcl-2 expression at the mitochondria can prevent activation of the mitochondrial apoptosis pathway and therefore apoptosis. But there is one dramatic difference between the pathway of BimL and BimS. According to the findings made in this project, caspase-8 is the chief helper of BimS. Upon activation, BimS not only induces the activation of the multi-domain proteins Bak and Bax but additionally also of caspase-8, whether this might be in a direct or indirect way. Only upon BimS expression IETD-fmk could inhibit loss of the mitochondrial membrane potential. Moreover, BimS induced a higher percentage of cells with caspase-8 activation as shown by binding of fluorescing IETD peptide substrate. Localization of Bcl-2 at the ER fully blocked BimS mediated apoptosis. Activation of the ER apoptosis signal seems to be triggered simultaneously to the activation of the mitochondrial apoptotic events. Bcl-2 seems to grant survival at this position even against such a strong killer as BimS. By targeting the ER, BimS causes the depletion of the calcium stores causing ER stress as an additional apoptotic pathway to its activation of the mitochondrial cell death pathway. BimL does activate caspase-8 only in a small number of cells although it induces upregulation of the ER stress proteins such as BiP, CHOP and Bap31. ER stress is said to be followed by calcium release and subsequent uptake by the uniporters of the mitochondria. Uptake of calcium into the mitochondria causes swelling of the matrix and can lead to disruption of the membranes. Active capase-8 was found by others to be integrated into the outer mitochondrial membrane, where it assists on permeabilization of the mitochondria (Chandra, et al., 2004 ). Additionally, it is also possible that caspase-8 activates Bid by cleavage to tBid, which can not only trigger the activation of the multi-domain pro-apoptotic proteins Bak and Bax but can also form pores at the outer mitochondrial membrane enhancing the disruption of these organelles. Therefore, BimS might have three possibilities or any combinations of those to induce and most importantly to ensure cell death once it is activated. Based on the results obtained in the present study, the following model is proposed:
|Figure 29: Model for Bim induced apoptosis|
Although some steps of Bim induced apoptotic pathway were illuminated, there are still a lot of questions to be answered. For example, if other anti-apoptotic proteins such as Bcl-xL are also able to inhibit the pathway of Bim or how Bim causes ER stress. Further it would be interesting to know, by which mechanism BimS induces the release of Ca2+ from the ER and the activation of caspase-8. The ER might mark the point where BimL and BimS follow different pathways. BimL might activate Bax and Bak already at the ER and not later at the mitochondria which then leads to the permeabilization of the mitochondria followed by the release of cytochrome c and finally to apoptosis. BimS does not only activate Bax and Bak just like BimL, but also induces a leak of calcium from the ER, which is taken up by the mitochondria and enhances the death signal. It can be speculated if this additional step of calcium mobilisation might be the reason for the potency of BimS.
One reason for the strength of BimS might lay in the fact that it is not bound and therefore not regulated by the dynein motor complex of the microtubule. An incoming death signal might activate BimS directly, whereas BimL would have to be released from the microtubule. Another explanation could be taken from the report that BimS is subjected to any post-translational modification, but is directly active after expression as opposed to BimL and BimEL (Weber, et al., 2007 ). Finally, the structure of BimS could be responsible for its strong killing ability. It was speculated that the intensity by which a BH3-only protein induced the apoptotic pathway is depending on its affinity to anti-apoptotic proteins. No such data are, however, at presence available comparing BimS and BimL. Binding assays revealed that Bim and Puma, which are both strong killers, had high affinity to all anti-apoptotic family members and could neutralize all the anti-apoptotic proteins, whereas the less potent BH3-only proteins had a more restricted binding spectrum (Chen, et al., 2005 ). On the other hand, Noxa only bound to Mcl-1 and Bfl-1/A1 and turned out to be a weak killer, Nbk neutralized Bcl-w, Bcl-xL, but not Bcl-2 or Mcl-1. Therefore it was suggested that limited targeting of anti-apoptotic proteins correlates with the apoptotic strength of the BH3-only proteins (Chen, et al., 2005 ). Such a mechanism might therefore also explain the higher potency of BimS to induce apoptosis as compared to BimL. The exact interplay with the anti-apoptotic proteins needs further extensive investigations and cannot be explained at this point. Further, it can be speculated that higher induction of apoptosis by BimS might be due to the additional amplification of the death signal by the release of calcium from the ER. As presented in this study, BimS did not only upregulate ER stress proteins but also triggered elevated cytosolic calcium levels. Calcium released by the ER is known to be absorbed by the mitochondria (Szalai, et al., 1999 ), which in turn contributes to mitochondrial permeabilization and the release of cytochrome c and other apoptotic factors.
The data presented in this thesis can promote Bim as a physiologically relevant target in tumour therapy.
|© 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 4.0||Zertifizierter Dokumentenserver|
der Humboldt-Universität zu Berlin
|HTML-Version erstellt am: