From Engine Room to Executioner: Mitochondrial Functions in Cell Death

Webinar Topics:

• Mitochondria in the execution of apoptosis
• How mitochondria drive the apoptosis process
• Methods used to investigate
• Mitochondria and non-apoptotic cell death

About the Presenter:

Dr Stephen Tait carried out his graduate studies at the Institute for Animal Health, in Pirbright, UK where he investigated viral immune evasion strategies. His postdoctoral training was carried out at the Netherlands Cancer Institute, Amsterdam and St. Jude Children’s Research Hospital, Memphis.

Since 2012, he has had his own research group based at the Cancer Research UK Beatson Institute and University of Glasgow. His group is supported, in part, by a fellowship from the Royal Society.

The main research interest of the Tait Lab is investigating deregulation of mitochondrial functions in cancer, while focusing upon cell death and autophagy.

Webinar Transcript:

It is my pleasure to introduce your host for today, Lucy Purser, Senior Events and Marketing Coordinator. Lucy, you now have the floor.

LP:    Thank you, Cara. Hi, welcome to Abcam's webinar today. Today's principal speaker is Dr. Stephen Tait from Cancer Research UK, Beatson Institute and University of Glasgow. Stephen carried out his graduate studies at the Institute for Animal Health in Pirbright, UK. This was where he investigated viral immune evasion strategies. His postdoc training was carried out at the Netherlands Cancer Institute in Amsterdam and at St. Jude Children's Research Hospital, Memphis. Since 2012, Stephen has had his own research group based at the Cancer Research UK, Beatson Institute and the University of Glasgow. His group is supported, in part, by a fellowship from the Royal Society. His lab's main research interest is investigating deregulation of mitochondrial functions in cancer, focusing upon cell death and autophagy.

Joining Stephen today will be David Bruce, Research Area Content Associate from Abcam. David has a BSc in Biomedical Sciences (Pharmacology) from the University of Aberdeen. Before joining Abcam, David studied for his PhD at the University of Dundee, where he examined the regulation of the TGF-beta signaling pathway by novel protein phosphatases and ubiquitin E3 ligases. 

Before we start, just a quick reminder that questions for the Q&A session at the end of the webinar can be submitted at any time via the Q&A panel on the dropdown section at the top of your screen. I will now handover to Stephen who will start this webinar.

ST:    Thanks very much, Lucy, and thanks for the introduction. Today, I'm really going to talk about the mitochondrial pathways of cell death and how to investigate these processes. I'm really going to divide my talk into three separate sections: one, is going to be discussing the mitochondrial or intrinsic pathway of apoptosis, its mechanism, and its dysfunction in disease and really outline a couple of outstanding questions that remain in the field. From that, I'm going to go onto discuss at least some methods that you can use for detecting mitochondrial-dependent apoptosis in the system that you're investigating. Finally, I'm going to showcase a new approach that we've developed to determine mitochondrial function in non-apoptotic cell death.

So, first, I'm going to talk about apoptosis. Apoptotic cell death requires activation of proteases called caspases, which bring about the rapid demise of a cell. Effectively, there are two different pathways, or two main pathways whereby caspases can be activated. One is called the extrinsic pathway of apoptosis, and this typically is activated by death receptor ligands. These bind to a receptor which, in turn, clusters and activates caspase-8, active caspase-8 goes ahead and then cleaves caspase-3 and 7 bringing about rapid cell death. I'm not really going to talk much about the extrinsic pathway of apoptosis. 

Today, we're going to focus on the intrinsic pathway of apoptosis, also known as the mitochondrial pathway of apoptosis, which is activated by one of numerous stresses such as DNA damage or ER stress. This leads to activation of BH3-only proteins that, in turn, activate two proteins in the process Bax and Bak. We'll discuss this in much more detail, but these proteins when they're activated lead to mitochondrial outer membrane permeabilization. This then, in turn, activates caspases through the release of proteins such as cytochrome-C or SMAC and OMI from the mitochondria, these then drive caspase activation in the case of cytochrome-C. SMAC and OMI, on the other hand, inhibits a caspase inhibitor called XIAP, and the net effect is once mitochondrial permeabilization has occurred, also known as MOMP - and I'll refer to it as MOMP through the talk as well - the cell dies within a matter of minutes to a couple of hours.

So just emphasizing the mitochondrial pathway of apoptosis, in healthy cells mitochondria attain various proteins in their mitochondrial intermembrane space, such as cytochrome-c, SMAC and OMI, and when these proteins are released they actively kill the cell. So, in one sense, mitochondria can be considered similar to suicide capsules. When a cell receives a trigger to die, that activates BH3-only protein and one of these is tBid. These, in turn, activate Bax and Bak, and somehow when Bax and Bak are activated they selectively permeabilize the mitochondrial outer membrane. Mitochondrial outer membrane permeabilization, and this has catastrophic consequences for the cell that's mentioned previously. Cytochrome-c binds this adaptor molecule Apaf that, in turn, activates caspases, they cleave hundreds of different proteins bringing about rapid apoptotic cell death. As mentioned, SMAC and OMI inhibit this caspase inhibitor XIAP, and so the net effect is you get rapid, robust caspase activation and rapid cell death.

We can image this whole process by confocal microscopy, and this is shown here. What you're going to see in this movie are the mitochondria being labelled red with the fluorescent fusion protein SMAC-mCherry. They're also expressing YFP-Bax, which is predominantly cytosolic in healthy cells. Now, as the cell dies the cell looks perfectly healthy up until the point the mitochondrial permeabilization occurs. At that point, SMAC-mCherry is released throughout the cell, you'll see it go from this punctate pattern to diffuse pattern and, at the same time, Bax translocates onto the mitochondria. So it moves onto the mitochondria and it goes from this diffuse pattern to punctate pattern. Once mitochondrial permeabilization has occurred, this engages rapid and robust caspase activity that leads to the hallmarks of apoptosis, and you can see that quite nicely in these cells. The cells shrink, the nucleus condenses and, finally, the cell is exposed to phosphatidylserine. This can be detected, because we've added annexin V which is conjugated to APC, and it binds phosphatidylserine as it's flipped from the inner to the outer leaflet of the plasma membrane. So I'll play this movie through a couple of times, just really to emphasize this point that mitochondrial permeabilization is the key event in the mitochondrial pathway of apoptosis. Eventually, these cells, as I say, are exposed to phosphatidylserine and in our bodies these cells are rapidly removed by neighboring phagocytes. Again, at that point, we get massive caspase activation and the cells shrink, membranes bleb, the nucleus condenses and we get phosphatidylserine exposure.

So that would give you the idea that mitochondrial permeabilization leads to cell death through caspase activity, well actually even in the absence of caspase cleavage of substrates, mitochondrial permeabilization often represents a point of no return. This is shown in this movie here in which cells have been transfected with the protein GFP/tBid, which causes mitochondrial permeabilization. These cells have already released SMAC-mCherry throughout the cell. However, these cells have been incubated with a caspase inhibitor and whilst the cells don't die in the short-term, you'll see quite nicely over time the cells undergo massive vascularization, and undergo what looks like necrotic cell death. So even if we prevent caspase activation, providing mitochondrial permeabilization has occurred, cells die. Usually this event causes caspase-dependent apoptosis, but somehow cells in the absence of caspase activity still die. Some studies would argue that this may be an active role for mitochondria in triggering caspase-independent cell death. Alternatively, or even additionally, death may simply be due to a gradual loss of mitochondrial function.

Mitochondrial apoptosis has many roles in health and disease, it's required for proper embryonic development and that's best evidenced in mice that are deficient in mitochondrial permeabilization, so, for example, ones that lack Bax and Bak; it's often embryonic lethal. It's also required amongst many other functions for things like regulating the immune system. So deficiency of mitochondrial permeabilization leads to defective lymphocyte maturation, homeostasis, ultimately triggering autoimmunity. So these are just two examples of the role the mitochondrial apoptosis in maintaining health.

It also has many roles in disease, and perhaps the best evidence role for deregulation of mitochondrial apoptosis is seen in cancer. So one of the hallmarks of cancer is that during the tumorigenic process, cells must have apoptosis. Why is this the case? Well, many tumor suppressor pathways engage apoptosis and kill the cells, it's at risk of becoming malignant, so inhibiting apoptosis promotes transformation. Moreover, it promotes cell survival under hypoxic or nutrient-poor conditions, common conditions that a tumor cell faces. Another aspect whereby apoptosis promotes cancer is many cells when they detach from their substratum undergo a form of apoptosis called anoikis. This prevents cells from moving around the body, and prevents metastasis. So, obviously, inhibiting the apoptosis again would facilitate this process and the movement of cancer cells around their body. Finally, it has this kind of secondary effect, many anti-cancer therapies can induce apoptosis and so by inhibiting apoptotic cell death, a cell can not only become cancerous it can also evade chemo or radiotherapy.

As mentioned a couple of slides back, Bax and Bak are really two key proteins that are essential for MOMP, mitochondrial permeabilization and apoptosis. On a pro-apoptotic trigger, that activates a BH3-only protein leading to Bax and Bak activity, ultimately triggering MOMP. That's best evidenced in mice that lack Bax and Bak, and one example, or physiological example of cell death is the interdigital space in our limbs; this webbing dies through the process of apoptosis, if we knockout Bax and Bak we effectively prevent cell death occurring in these interdigital webs, and you can see the webbing remains. Moreover, if we take cells from animals that are deficient in Bax and Bak, we can see quite nicely the DKO cells, all the intrinsic or mitochondrial triggers of apoptosis, such as staurosporine, etoposide or UV, in the absence of Bax and Bak cells are completely resistant to these stimuli.

So how do these proteins get activated? Well, Bax as shown in that movie a couple of slides ago, predominantly in healthy cells is a cytosolic protein, so this is green GFP Bax shown here. As a cell undergoes mitochondrial permeabilization, undergoes apoptosis, Bax translocates from the cytosolic location onto the mitochondria, as you can see quite nicely here, co-localizing with mitochondria. Upon this translocation it engages mitochondrial permeabilization, and you'll see that nicely in this movie here, Bax translocates, so it goes from this diffuse localization to a punctate mitochondrial localization. Really, simultaneously, mitochondrial permeabilization has engaged only mCherry, which resides in the mitochondrial intermembrane space, is then released from the mitochondria throughout the cell and, ultimately, the cell will die. As you can see quite nicely here, Bax will translocate and, at the same time, mitochondrial permeabilization occurs.

So how do active Bax and Bak cause mitochondrial permeabilization? This is really a matter of much debate. One thing that is clear, is that during the process of apoptosis Bax and Bak form homooligomers, and that's quite evident in the experiments that have been done using chemical cross-linkers. Cells or mitochondria that have been treated with the active form of tBid to engage mitochondrial permeabilization have been treated in the presence of cross-linker. When we look at Bak it goes from this monomeric state in the presence, or when it's activated we can cross-link higher molecular weight of species, and exactly the same thing goes for Bax. So two schools of thought: one, we'd argue that upon activation Bax and Bak form homodimers, dependent kind of, if you like, head-to-head homodimers. These then form higher weight oligomers ultimately causing permeabilization of the mitochondria. Alternatively, active Bax and Bak are being proposed to cause these daisy chain-like homooligomers that ultimately cause mitochondrial permeabilization.

How does the mitochondrial outer membrane permeabilize? Two main models argue that upon activation of Bax and Bak, they either form proteinaceous channels, so they form holes in the mitochondria themselves allowing the release of intermembrane space proteins. The alternative model, and there's certainly a lot of controversy about this, is that rather than directly causing pores in the mitochondrial outer membrane activated Bax and Bak, serve to bend the mitochondrial outer membrane lipids, forming lipidic pores, which ultimately allow the release of cytochrome-c. Exactly how this process occurs, as I mentioned, is still highly debated. 

​

How is it controlled? Simply, what I've discussed so far would give you the idea that it's just unidirectional, provided there's an activating signal Bax and Bak will be activated. However, that's definitely not the case. There are many proteins that restrain Bax and Bak activation, and so, essentially, BCL-2 protein family members regulate mitochondrial permeabilization. This family can be divided into two main subsets: one is the anti-apoptotic BCL-2 protein family, comprised of BCL-2, BCL-W, BCL-XL, A1 and MCL-1 that have four BCL-2 homology domains. Secondly, there are pro-apoptotic BCL-2 proteins and, as I mentioned, Bax and Bak are two of them, and these are effector molecules. Your similarity to these proteins, whether in itself it's an effector protein is unclear. Finally, there's the class of, if you like, apoptosis signaling molecules, the BH3-only proteins, of which there are numerous BID, BIM, BAD, BIK, BMF and so on, that activate apoptosis. So anti-apoptotic BCL-2 proteins restrain the process of mitochondrial permeabilization, and they do this, the whole system works through protein-protein interactions. 

So, for example, this is a structure shown here of MCL-1 binding the BH3 domain of the BH3-only protein BIM. What happens is that the BH3-only domain of BIM binds into the hydrophobic roof of MCL-1, and in doing so MCL-1 can neutralize BIM function. In this part of the slide here, you can see quite nicely these are the anti-apoptic BCL-2 proteins, they bind to varying degrees different BH3-only proteins. So, for example, BIM, PUMA and tBID appear to be able to bind all of them, BAD and NOXA, which are other BH3-only proteins are more restricted in their binding specificity. Secondly, BCL-2 proteins can bind Bax and Bak, and, again, there appears to be some specificity as to which anti-apoptotic BCL-2 proteins can bind to either Bak or Bax.

So exactly how BCL-2 proteins regulate MOMP is still a matter of controversy. Until relatively recently, there were two models suggested to propose this effect. One is that Bax or Bak, for that matter, are directly activated by BH3-only proteins, so, for example, BIM activates Bax and then goes on to drive mitochondrial permeabilization. BCL-2 proteins in this model, the direct model of activation simply served to bind and sequestered these activating BH3-only proteins and, in doing so, they can prevent Bax and Bak activity. The other model, the indirect model of activation is that rather than these proteins such as Bax and Bak being directly activated, they exist in a constitutively active form, and such that BCL-2 family members, anti-apoptotic proteins, serve to block activated Bax and Bak. The BH3-only proteins in this model neutralize anti-apoptotic BCL-2 function, allowing activated Bax and Bak to drive mitochondrial permeabilization. 

More recently, there's been a model I've called the Unified, which is an elaboration of the embedded together model, which really argues that both the direct and indirect models are likely true. In this model what happens is, BIM or another BH3-only protein can activate Bax, leading to Bax activation and mitochondrial permeabilization. BCL-2, anti-apoptotic BCL-2 proteins restrain the process both by binding these activator BH3-only proteins, and by binding activated Bax and Bak. So most likely aspects of both of these models hold true.

Recently, new drugs have been developed to therapeutically target mitochondrial apoptosis in cancer. These drugs have been termed BH3-mimetics, and they work by binding the hydrophobic groove in anti-apoptotic BCL-2 family proteins and neutralize their protective function. So by binding in this hydrophobic groove in, say, for example, BCL-xL this drug Navitoclax ABT-263 prevents BH3-only proteins from binding into the BCL-2 family member, in turn, liberating these BH3-only proteins to go on and activate Bax and Bak.

So there's a variety of BH3-mimetics that have been developed that target a panel, for example, BCL-xL, BCL-W and BCL-2 or BCL-2 family members, so ABT-737, 263 and target all three of them. More recently, new ones have been developed WEHI-539 that target BCL-xL, ABT-199 targets BCL-2. It's important to note, and this is something we'll touch on later as well, that there are various other anti-apoptotic and BCL-2 family members for which we can't specifically target at present.

So how do these drugs work in practice? Well, as I've stated, anti-apoptotic BCL-2 proteins prevent mitochondrial permeabilization. When we add in a BH3-mimetic compound, this targets and neutralizes anti-apoptotic BCL-2 function, and so it removes this break, if you like, upon the apoptotic process. Now, cells are much more sensitive, you can trigger cell death with a lower drug treatment, for example. This leads to release or upregulation of a lower amount of BH3-only proteins, triggering mitochondrial permeabilization. So in the presence of these drugs, many cells and specifically tumor cells are actually more sensitive to cell death. 

Here's some evidence suggesting that they work, and these drugs beyond being very useful as research tools, they appear to be very effective either on their own or in combination with chemotherapy inducing tumor cell death. So, as an example, these are xenograft tumors grown in a nude mouse. They are either grown with vehicle, and you can see quite nicely that the tumor keeps growing, or in the presence of ABT-737. Secondly, in the clinic they've also been used, and this is a patient that's been suffering from a lymphoma pre-treatment, with one of BH3-mimetics, as you can see the lymphoma mass here. Then post-treatment, much of that lymphoma has regressed. So, as I said, they look to be very promising drugs either on their own and in certain lymphoma or leukaemias, or together with additional chemotherapies to specifically kill cancer cells.

To summarize this part, what I've told you so far is that mitochondrial outer membrane permeabilization is the key event in the mitochondrial pathway of apoptosis. It leads to the release of proteins, such as cytochrome-c from the mitochondrial intermembrane space that activate caspases. However, importantly, even in the absence of caspase activity, MOMP typically represents a point of no return, so cells die once it's occurred. BCL-2 family proteins are the main regulators of mitochondrial permeabilization. Pro-apoptotic BH3-only proteins signal the pro-apoptotic signal, or relay the pro-apoptotic signal to activate Bax and Bak, and block protective anti-apoptotic BCL-2 function. Bax and Bak upon activation permeabilize the mitochondrial outer membrane. Anti-apoptotic BCL-2 proteins bind both classes of pro-apoptotic proteins, thereby preventing apoptosis. These, or this network of interactions have been exploited by new drugs that have been designed to specifically bind and neutralize anti-apoptotic BCL-2 proteins.

So I've just outlined three outstanding questions I think that remain regarding mitochondrial pathway of apoptosis, and there are many others but these are at least three that spring to mind; spring to my mind, at least. One is how do activated Bax and Bak permeabilize the mitochondrial outer membrane? As stated, there are two models for this but it's really unclear how this process occurs, and it's critical that we do understand it because it's the really defining event on whether a cell lives or dies. 

Secondly, when we think about targeting this networking disease, can we effectively target BCL-2 family members in cancer? What I mean by this is we can clearly develop drugs to inhibit anti-apoptotic BCL-2 proteins, but are there going to be problems with resistance? For example, upregulation of MCL-1 we can't target that just now. Secondly, if we target many anti-apoptotic BCL-2 proteins simultaneously, is that going to lead to toxicity problems? Moreover, getting away from this idea of sensitizing cells to apoptosis, there are some diseases in which too much apoptosis is a bad thing. For example, various neurodegenerative diseases may be contributed by too much apoptosis. 

So on the flipside, can we target BCL-2 family members to prevent mitochondrial outer membrane permeabilization in cell death? Really, I think that's a relatively unexplored area that deserves more attention.

So next I'm going to discuss with you some methods - certainly not an exhaustive list - but some methods for detecting mitochondrial-dependent apoptosis. Firstly, you have a stimulus of interest that's killing the cells. Does MOMP actually occur after my stimulus of interest? Most of the assays that revolve around detecting mitochondrial permeabilization really rely upon this robust release of intermembrane space proteins, such as cytochrome-c, SMAC and OMI, from the mitochondria into the cytosome. As I showed you earlier in that movie, it's really a black and white event most, if not all, of the mitochondria permeabilize in the cell leading to a massive release of these proteins into the cytosome. So we can use this differential localization to assay for the presence of MOMP.

We can do this in several ways: one is to use immunofluorescent staining for cytochrome-c, SMAC or OMI. It's punctate under normal, healthy conditions and it goes to diffuse cytosolic state when MOMP occurs. This is shown quite nicely in these figures here in which cytochrome-c has been stained either in healthy cells, and you can see it has this punctate localization. Or cells that have been treated to undergo apoptosis with actinomysin D, and you can see here the cells round up that cytochrome-c is released throughout the cell. In the presence of zVAD we see this cytochrome-c release from the mitochondria throughout the cell. zVAD is a caspase inhibitor and this prevents the cells rounding up, and actually makes them more easy to visualize. So I think performing these analyses, a key tip is really to add caspase inhibitor into the experiment to prevent cell rounding and detachment. The important thing to note is that's only going to work provided caspases aren't upstream of mitochondrial permeabilization. So, initially, maybe both ways of doing it, either in the presence or absence of caspase inhibitor would be beneficial.

The pros of it is that it's straightforward, there are very good antibodies to mitochondrial and to membrane space proteins, and it's very obvious when a cell has undergone MOMP, it's not very subjective analysis. You can see quite clearly there's a difference between a healthy and a cell that's undergone MOMP, and it can be very quantitative. I think one of the major cons is that it's quite laborious when quantifying many different samples.

Alternatively, we can detect release of proteins from the intermembrane space throughout the cell using western blotting of cytosolic, and/or mitochondrial membrane fractions. So this is a case of stimulating cells, fractionating by different methods. We often use digitonin-based lysis, but equally as good, hypertonic lysis also works. By using these selective lysis conditions, you can selectively permeabilize the plasma membrane that leave the mitochondrial outer membrane intact. Again, I think it's beneficial to add qVD caspase inhibitor or zVAD into the experiment to prevent cell lysis and cytochrome-c loss, at least from the cytosolic fraction. So, as an example, here these cells have been treated either with control or with etoposide or gamma radiation for different lengths of time. Hypertonically lysed and then a cytosol has been probed for cytochrome-c as a measure of mitochondrial permeabilisation, and you can see quite nicely these are apoptotic triggers and we can detect cytochrome-c quite effectively in the cytosolic extract. 

One of the major pros is that it’s quite straightforward, however, our con to this approach is that you need a larger sample than doing immunostaining, for example, and it's not very quantitative. So from this, obviously, we can't gain a percentage of cells that have undergone mitochondrial permeabilization, only if it has occurred in at least a subset of [recording break]. 

The last way, at least, using an antibody immunostaining approach, I'm going to discuss today is the detect release of proteins from the intermembrane space throughout the cell by FACS analysis, or cytochrome-c, or this can be done also for SMAC or OMI released from the mitochondria. So this method was developed by Nigel Waterhouse several years ago now, and the rationale behind it is here we can take a healthy cell, when you selectively permeabilize the plasma membrane the mitochondria are still intact, so we have no loss of mitochondrial intermembrane space protein signal, and we can immunostain these cells and stain them by FACS, and so these stain high. However, when a cell has undergone apoptosis, the intermembrane space proteins leak throughout the cell, and we can then selectively permeabilize the plasma membrane and this protein, if you like, could be cytochrome-c, can be released from the cell into the cytosol. We stain these cells up, and these stain low and so, in essence, we can get two peaks: one healthy, no MOMP has occurred; the second unhealthy, MOMP has occurred. We can distinguish this in a very quantitative manner, and that's showing quite nicely here from Nigel's paper. These are healthy cells here, and not undergone MOMP; these are the ones that have undergone MOMP and we can very easily quantitate that, as shown here from some work I did a few years ago.

The pros of this approach is it's straightforward, quantitative and, actually, you don't need very many cells for it as well. You can do many different samples at the same time. The cons is, at least, my experience of this approach is sometimes the positive and negative peaks overlap somewhat, and that complicates quantification. Secondly, digitonin concentration must be optimized on a power cell type basis. 

The final approach for detecting MOMP I'm going to discuss for you today, is really to use live cell imaging. So we can fluorescently label mitochondria with proteins such as cytochrome-c-GFP, SMAC-mCherry, SMAC-GFP or OMI-mCherry. These proteins obviously reside in the mitochondrial intermembrane space in healthy cells, but we can image these cells repeatedly over time and the proteins are then released as a cell undergoes mitochondrial permeabilization. We can detect the release of these proteins throughout the cell.

I think this is shown quite nicely here in which these are cells, MCF-7 cells that have been treated to undergo apoptosis, they express matrix, mitochondrial matrix-targeted dsRED which stays in the mitochondria even when they undergo mitochondrial permeabilization. However, these cells are also expressed in SMAC-GFP and you'll see as the cells undergo mitochondrial permeabilization, SMAC-GFP is really released in an explosive manner throughout the cell very rapidly. You can see that quite nicely there. So it's a relatively straightforward methodology to look at mitochondrial permeabilization in real-time. 

One of the huge advantages of it, is we can look on the power cell basis, we can image these cells, and we can quite easily quantify by gating around the individual cells when a cell has undergone mitochondrial permeabilization. So we apply a so-called punctate diffuse index, and this is averaged over many different cells here. This is the start of mitochondrial permeabilization, and really within 10 min most cells have undergone complete release, complete mitochondrial permeabilization, and then ultimately die.

So the pros of this approach is it's quantitative, small cell numbers acquired can easily multiplex with other probes. So, for example, we can look at caspase activation and BACS translocation at the same time. The cons are that it's relatively low throughput, the analysis time can be considerable, and at least for some of these reporter proteins, they may have to be introduced stably into cells. So transient transfections, especially with things like cytochrome-c it tends to mislocalize, and so it doesn't localize properly to the mitochondria. So, in that sense, it may take a while to generate a cell line that's of much use. 

So moving on from that, I want to really ask the question or give some hints how can I tell if my treatment is inducing mitochondrial-dependent apoptosis? Well first you want to address, whether your cell is undergoing apoptosis and commonly we do this in the lab by flow cytometry. Here we make use of cell impermeant dye such as propidium iodide, which is taken up into dying cells and all dying cells take up propodium iodide whether they undergo apoptosis or not. However, we co-labelled cells with annexin V, together with PI and this allows us to specifically stain the apoptotic population, which is annexin V positive and usually PI negative, but, ultimately, when cells are left long enough they'll take up PI. So an example of this is shown here, these cells have been engaged to undergo apoptosis, these are the healthy cells. We have PI staining along the bottom here and annexin V staining along the side. Cells, as they undergo apoptosis they become annexin V positive, and this is because of this phosphatidylserine exposure that I mentioned at the beginning of my talk that occurs during apoptosis. It's flipped from the inner to the outer leaflet of the plasma membrane, and we can detect that with annexin V. Ultimately, these cells that are annexin V positive will take up PI as well. But we can specifically measure apoptosis by looking at the annexin V positive, PI negative population.

Secondly, you can address whether apoptosis is occurring by adding in chemical caspase inhibitors of which there are many too that we use often in the lab, qVD or zVAD. These should, if caspase-dependent process is occurring, caspase-dependent apoptosis, it should at least slow cell death. With the one caveat that MOMP can also lead to slower caspase-independent cell death, which makes it important to do extensive kinetics. Moreover, you can ask whether mitochondrial permeabilization is required for cell death, by generating cell lines that overexpress anti-apoptotic BCL-2 proteins. So an example here at my lab where we've expressed our BCL-xL in U2OS cells, BCL-xL is an anti-apoptotic BCL-2 family member, then we've measured cell death fall in proteasome inhibitor. You can see quite nicely you get cell death, and this is by looking at annexin V positivity, BCL-xL inhibits this process. Which we'd argue that mitochondrial - that the proteasome inhibitor MG132 is triggering cell death through a mitochondrial-dependent effect. Alternatively or additionally, in addition to overexpressing anti-apoptotic BCL-2 proteins, we can also knockdown Bax or Bak. With the caveat that in most situations either Bax or Bak is sufficient to drive apoptosis, and so in most cases both will need to be knocked down. 

I stated just a minute ago, I think it's important to do cell death assays, in many cases, with extensive kinetics. This is just really highlighting this example, and we can overexpress BCL-xL, treat cells with staurosporine, which is an apoptotic trigger, looking at the percentage of mitochondrial permeabilization, and all these different concentrations of staurosporine BCL-xL effectively block cell death, at least in the short-term. If these cells are left longer, you can see quite nicely only at the lower concentrations of staurosporine does BCL-xL allow clonagenic survival. Clearly, this compound staurosporine and many other triggers are like this as well. In a longer-term scenario, these apoptotic stimuli or pro-death stimuli can trigger other forms of cell death, and so it's important to do wherever possible these analyses with extensive kinetics.

Other aspects that can be investigated, one can look at Bax activation, for example, by looking at translocation onto mitochondria, making use of an active conformation-specific antibody 6A7 that was developed in Richard Youle's lab many years ago to look at activated Bax, and that can be used by imaging or flow cytometry. Equally, Bak activity can be looked at by making use of a conformation-specific antibody Ab-2. 

Moreover, extensive studies have been carried out into MOMP and its role in cell death, making use of in vitro mitochondrial permeabilization assays. So these are, in many cases, using mitochondria that are isolated from mouse liver, although they can be isolated also from many cell lines as well. Mouse liver mitochondria can be incubated in an Eppendorf tube, along with different BH3-only proteins, but this is just an example here. You incubate it along with these proteins that engage MOMP, engage mitochondrial permeabilization, then you can fractionate out quite easily simply by centrifugation the mitochondrial membrane pellet, or the supernatant. You can see quite nicely here then we block for cytochrome-c. In the control situation there is no MOMP or cytochrome-c remains in the pellet. In the activated situation where we have mitochondrial permeabilization, cytochrome-c is completely released from the mitochondria. So I would refer you to this paper for more detailed methods on how to analyze mitochondrial permeabilization in an in vitro manner. 

So for the last few minutes I'm going to talk to you, or discuss with you some new approaches that we've been utilizing to look at mitochondrial importance in non-apoptotic forms of cell death. We've been using this approach to look at the role of mitochondria in necroptosis, and this is a former cell death that's been first described several years ago now, and it's strange in that it's a form of cell death that's actually inhibited by caspases. I'm not going to go into extensive detail about how caspase inhibit the process, but this really demonstrates that we can treat cells with tumor necrosis factor, and many cell types are fine in the presence of tumor necrosis factor. cVAD caspase inhibitor is co-added and then we see massive extent of cell death, and this is detected by uptake of cytox green and it's one of these membrane impermeant dyes when cells die, we take it up and they light up green. The textbook view of cell death is that it can really be divided into two forms: one is necrotic, which is a passive, unprogrammed form of cell death. It's not regulated by any set of proteins. Apoptosis is clearly an active form of cell death, it's programmed. There's certainly some discussions as to whether this is true, but traditionally, apoptosis is thought of as a non-inflammatory form of cell death, whereas necrosis is considered inflammatory.

Necroptosis falls somewhere in between in that it's a programmed form of cell death with necrotic-like features. It's thought that this form of cell death may actually promote inflammation. So it's a non-apoptotic programmed form of cell death that's elicited by various stimuli such as viruses, tumor necrosis, factor pathogen recognition receptors and, undoubtedly, many other stimuli can also activate necroptosis. So far they all require activation of this kinase called RIPK3. This, in turn, activates another protein in a cascade called MLKL, which leads to cell death with necrosis-like features. We were interested in how cell death is executed.

Many papers over the years that have implicated a central role for mitochondria necroptosis execution. So, for example, during necroptosis ROS, reactive option spaces have been shown to be increased, and which may come from the mitochondria. There's a loss in ATP and, obviously, mitochondria being a powerhouse of the cell, loss of mitochondrial function would lead to loss of ATP, again implicating mitochondrial dysfunction in necroptosis. More recently, mitochondrial proteins such as PGAM5 or Drp-1 have been implicated in necroptotic cell death, and clearly based on previous form in the apoptotic signaling, there's some precedence that mitochondria may be involved in actively causing cell death. 

So we got interested in addressing whether mitochondria were involved in necroptotic cell death, really from initial movies that we'd carried out looking at mitochondrial function as cells undergo necroptosis. So these cells have been stained up with a potentiometric dye called TMRE, which stains up healthy mitochondria, it stains them up red. We imaged these cells as they undergo necroptotic cell death, and so over time what you'll see is these cells eventually burst, they then take up annexin V 488 which goes into the cell and lights up the plasma membrane from the inside. I just want to emphasize, and I'll play it a couple of times, is that as the cells undergo death, at that point they lose their TMRE, they lose their membrane potential. Really arguing that loss of mitochondrial function is a late event in necroptosis, as you can see there. So that's just a still taken from that movie and really shown at the point of lysis, these mitochondria still retain at least some membrane potential. But this movie at least doesn't implicate or discount the role that mitochondria may have in necroptosis. 

So we thought it would be really cool and really important if we could take cells and ask if we deplete mitochondria from them, does it impart any resistance to necroptosis? So the way we've done that is make use of enforced Parkin-mediated mitophagy. So Parkin is an E3 ubiquitin ligase that was described back in 2008 by Richard Youle's lab in mammalian cells to cause mitophagy, which is selective removal of damaged mitochondria through the process of autophagy. So if you disrupt transmembrane potential of mitochondria by an uncoupler, such as CCCP or FCCP, Parkin is recruited onto these mitochondria and then ubiquinates them, and this drives their removal from the cell. 

So we've taken this approach to stably generate cell lines or cells that stably express YFP-Parkin, and then treated them with uncoupler and then looked at mitochondrial content. You can see quite nicely here, the mitochondria are in red here, the cells are expressed YFP-Parkin, which is in green. Once they're treated with uncoupler, in most cases we failed to detect mitochondria left. By confocal analysis we've extended this also to do EM analysis, mitochondria highlighted in red here, these are YFP-Parkin control cells. You can see they're full of mitochondria, and once these cells are treated with uncoupler for a couple of days, we completely deplete mitochondria from most of the cells. So we can generate cell lines that really homogenously deplete mitochondria in a really effective manner. 

You can confirm that in various other ways: one is to do an immunoblot for different mitochondrial proteins. Only in the presence of Parkin and only after uncoupler is added, we see depletion of different mitochondrial proteins. I think it's important to note that if you're carrying out this sort of approach in the lab, it's really beneficial to blot against different mitochondrial proteins. For example, TOM 20 has been shown to be ubiquitinated by Parkin, and removed from mitochondria in the absence of mitophagy. So it's important to look at matrix proteins and into membrane space proteins. If you have affected mitochondrial depletion, these should also be removed from the cells. 

So, importantly, at least some cell lines can actually survive short-term falling mitochondrial depletion. You see that quite nicely here. We've taken cells that express Parkin uncoupled to mitochondria, and we get effective mitochondrial depletion and we usually analyze these cells at 40 hours post-addition of CCCP. The cells actually stick around, and if we look at cell viability here, for many days after depletion. Ultimately, they all go on and die.

So we know it generated a cell line that can survive one term in the absence of mitochondria. We do see a robust inhibition of clonogenic survival, but I think the important thing to note is we have this window of opportunity where we have cells that have mitochondria or not, and then we can ask different questions. So we've used this approach to ask if we deplete mitochondria, do we have an effect upon necroptosis?

Here, we've used the system that was developed by Andrew Oberst, that's based over at the University of Washington in Seattle, where he's developed really nice systems to cleanly activate caspase-8 and trigger apoptosis, or cleanly activate RIPK3 and trigger programmed necrosis or necroptosis. So here he's fused FKBP domains for either caspase-8 or RIPK3. In the presence of dimerizer drug this brings caspase-8 together causing its activation, which then allows apoptosis. The same goes for RIPK3, it actually causes oligomerization of RIPK3 and can trigger cell death. So using this approach, we've either engaged apoptosis or necrosis and depleted mitochondria to ask if we get rid of mitochondria, does it affect the kinetics or the extent of cell death?  

You can see quite nicely here these are caspase-8 dimerizer cells. You add the chemical dimerizer, you trigger rapid apoptotic cell death. However, if you deplete mitochondria from the cells and then trigger caspase-8 activation, you completely block the ability of caspase-8 to trigger apoptosis in these cells. Which argues at least in these cells, that caspase-8 requires mitochondrial involvement to trigger apoptotic cell death. 

On the other hand, when we do this with RIPK3, we dimerize or oligomerize it, we get rapid levels of necroptosis. You do the same thing when cells are being depleted of mitochondria, and we see exactly the same extent of cell death. We can quantitate that very nicely. These are cells that have been depleted of mitochondria, and triggered to undergo necroptosis. These are the ones that are mitochondrial replete, and undergo necroptosis. Effectively, you can overlay the two graphs. Really arguing that mitochondrial elimination does not affect cell death triggered by direct activation of RIPK3.

So what I want to just highlight in this last part is the mitochondrial elimination does not prevent necroptosis either driven by direct activation of RIPK3, or we've also done it after TNF-induced necroptosis. Beyond this, I think Parkin-driven mitophagy is a powerful tool to define mitochondrial importance in any given biological process, not solely cell death. So RIPK3 triggers necroptosis we believe independently of the mitochondria, and actually subsequent to our work there's evidence that RIPK3 through MLKL can trigger necroptosis by directly causing plasma membrane permeabilization. 

I'd just like to finish up by acknowledging the guys in my lab that are certainly helping with this work, and continuation of it. Moreover, I've mentioned during this presentation Andrew Oberst's developed these clean ways of killing cells either by apoptosis or necroptosis. The initial mitophagy work was started in Doug Green's lab, and these are the guys that fund my lab. With that I'd like to thank you for listening to me and I'll hand you back to David, who's going to talk to you about some seminar-related products. Thanks very much.

DB:    Thank you, Stephen, for a really great talk. I'm sure you'll have plenty of questions from our listeners. Hello, I would like to take this opportunity to tell you a bit more about some of our resources and products that Abcam has available for cancer research.

Our products can be reviewed by customers and we call these reviews AbReviews. AbReviews are unbiased where we publish all positive and negative reviews; these are also quick and easy to complete.

Abcam has a dedicated multilingual scientific support team to help you with any problems or questions about our products. All of our products are covered by our AbPromise, where we guarantee that a product will work in the applications and species listed on the datasheets. We also offer testing discounts where a species or application not listed on our reagents datasheet can be tested. 

Abcam offers the largest range of rabbit monoclonal antibodies. These combine the benefits of monoclonal specificity with the robust rabbit immune system. Since these are raised from rabbits they show no cross-reactivity when used on lysed tissue, so why not try a RabMAb from Abcam?

We have a broad range of secondary antibodies to complement our primary antibodies, including our new range of Alexa Fluor® conjugated secondaries. We produce these in-house and they are available pre-adsorbed to different species, and in a range of Alexa Fluor® fluorochrome conjugates. 

We also offer cell-staining reagents, including our CytoPainter range of reagents which are ideal for multicolor staining. These allow you to visualize subcellular components such as the mitochondria or lysosome, and can be used on live and fixed cells. 

Abcam offers a number of dyes for fluorescence imaging, these include DRAQ5 and DRAQ7. These are small molecules which label double stranded DNA, or permeabilized cells. CyTRAK Orange, which stains the nucleus and cytoplasm. Nuclear yellow and nuclear green, which are both DNA dyes and you can find more information about these on our website.

Abcam also has an extensive offering of products to support your immunohistochemistry protocols. These include various types of detection systems such as EXPOSE-IHC kits for high sensitivity staining, as well as multi-colored detection systems or chromogens for staining multiple markers. You can find a wide range of reagents to support your IHC, including blocking reagents and sera, antigen retrieval buffers, as well as histology stains to assist you in identifying tissue morphology. 

We have a number of western blotting antibody cocktails, including our apoptosis cocktail, and these reagents allow you to detect multiple proteins in one immunoblot. We also have apoptosis detection kits that monitor the flipping of phosphatidylserine or PS from the interface of the plasma membrane to the cell surface after apoptosis has been initiated.

Last year we launched a new range of ELISA kits or SimpleStep ELISA kits. The time taken to develop our SimpleStep ELISA kits is reduced when compared to traditional ELISAs. This is because our new kits only require one washing step. This reduced washing also has the benefit of improving their sensitivity, as well as their specificity. 

Abcam also offers a wide range of pharmacological tools, these include over 4,000 enzyme receptor ion channel inhibitors and activators. Our range of pharmacological tools include some of the compounds Stephen spoke about today, including the caspase substrate inhibitors, and species EP. 

We are currently pulling together our different cancer research resources onto one page at abcam.com/cancer. We have brought together our products, articles, pathways, as well as relevant webinars and events. This is a work-in-progress and we're continually aiming to improve this page.

I'd like to finish by highlighting one of our upcoming meetings: Chromothripsis, Clustered Mutations and Complex Chromosome Rearrangements. This is taking place in Cambridge, Massachusetts in September. If you'd like to find out more information on this meeting, please visit our website at abcam.com/events. I'll now pass you back over to Stephen who is ready to answer your questions we've received during the webinar. Thank you for your attention.

ST:    Thanks very much, David. So thanks guys that sent in the questions, I'll try my best to answer some of them. One was from Victoria, and that was: What concentration of CCCP was I using for these experiments? Typically, we've used in the lab 12.5 µM. Whilst we've been using CCCP, we've moved over a little bit more now to using antimycin A and oligomycin. The reason being is that CCCP also effects lysosomal functions, so moving or confirming our results using other methods has been really beneficial.

This is another one: Does necroptosis induce inflammation? I think there's certainly a few mouse models that lack RIPK3, for example, that we'd argue that necroptosis doesn't induce inflammation, so it reduces inflammatory bowel disease. Well, at least, I guess, that argues that RIPK3 is required for it. I think there's still a lot of debate as to whether these mouse models that are implying that necroptosis has a role in inflammation, whether it's really to do with necroptosis or the ability of RIPK3 to actually induce cytokine production. So I think that's still very much a matter of debate. Moreover, some labs would argue that necroptosis serves to actually inhibit inflammation by killing the cell. So I think it's still, as I said, still an active area of debate whether necroptosis is inflammatory per se.

So another question: During multiplexing, which probe is suitable? I think that's obviously relative to the imaging approaches that I was discussing earlier, whereby we're looking at mitochondrial permeabilization. So the annexin V conjugates that are used for PS or to look at phosphatidylserine exposure, they come in a range of colors. So if you image looking at, say, SMAC-mCherry release from the mitochondria, you can certainly use annexin V coupled to Alexa Fluor® 488 to look at phosphatidylserine exposure. Moreover, there are various caspase activity probes that can be used: one, or many of these are based on FRET or loss of FRET, so fluorescence resonance energy transfer, that upon caspase cleavage you get a loss of signal. These can typically be combined with SMAC-mCherry, and there's a few of those reporters out there. To look at plasma membrane integrity, simply PI is a very good compound to look at that. It's taken up very quickly and very robustly when cells undergo apoptotic cell death, ultimately, the end of the whole process, they take up PI and so that can be, again, multiplexed with many of these SMAC-GFP or SMAC-mCherry type live cell imaging approaches.

I'm seeing if there's other questions here. Can it effect the other function of cells if a cell doesn't have mitochondria? Undoubtedly that's the case, this is something where we're investigating now, the role of mitochondria in, for example, protein secretion, vesicular trafficking we're quite interested in addressing these questions. In the absence of mitochondria are these functions actually affected? Not simply through the role of mitochondria in providing energy, but do mitochondria have a signaling function related to organelle transport, vesicular trafficking in cells? I think, undoubtedly, that's going to be the case, you remove mitochondria from a cell it's going to have many different effects on that cell.

I'll just finish up with a couple of other questions, one is: Do all cancer therapies kill via apoptosis? This is certainly a matter of debate as well. As I mentioned in one of my slides, apoptotic cell death is often the primary way in which cells die, but clearly during cancer therapy, cells often undergo necroptotic or necrotic cell death as well. So if cells are exposed to high enough levels of drugs or radiation, whilst apoptosis may be the primary way in which they undergo cell death, ultimately, the cells will die by other means as well. 

Another question is: How does mitochondrial permeabilization kill the cell in the absence of caspase activity? Really, we don't know how this occurs, I think the best evidence in the literature would simply be that there's a progressive loss in mitochondrial function, which ultimately leads to cell death. So with that, I'd like to thank you for your questions and thanks for tuning into the webinar, and I would like to hand, well, myself back to David. Thanks very much.

DB:    Thanks, Stephen, and thank you for that. I'll now pass you over to Lucy who is going to just tie up the end of the seminar.

LP:    Thanks, David. We would like to thank everyone who attended the webinar today, and a copy of this presentation will be emailed to you very shortly. If you have any questions about anything that has been discussed today or have any technical enquiries, the Abcam support team would be very happy to help you. They can be emailed at technical@abcam.com, and more details about relevant phone numbers can also be found on the Abcam website. We hope you have found this webinar informative and useful to your work, and we hope we can welcome you to another Abcam webinar in the future. Thank you, again, for attending and good luck with your research!