How to differentiate apoptotic from necroptotic cell death? Part III.

How to determine cell death?

Various assays exist to determine cell death and all, unfortunately, have their limitations. In this section, I’ll discuss all of the assays that I’ve got experience with. I’ll also indicate whether I deem an assay suitable for high throughput screening (HTS) or not. In general, I’d advise against using an assay that you don’t understand the principles of. Companies such as Promega are never very eager to reveal the underlying principles of their assays, as they’re afraid other companies will copy them, but if you don’t know what, for example, the ‘live-cell protease’ activity is that an assay such as Promega’s  MultiToxFluor assay measures, you can’t possibly determine whether your treatment is indeed affecting the cells’ health or just the activity of this mystery protease. If a company won’t tell me what it is exactly that their assay does and what the buffer components are, I won’t trust that assay.

MTT assay

Figure 1. MTT assay to determine TNFa toxicity on L929 cells.

A rather old fashioned method for determining cell death is the good old MTT assay. This assay depends on the reduction of tetrazole to formazan by oxireductase enzymes in living cells. Formazan forms a purple precipitate that can easily be detected with an absorption spectrometer. However, just like ATP assays, this assay also detects loss of cells, lack of proliferation or reduced metabolic activity, rather than cell death itself. MTT assays are certainly useful because they’re extremely cheap and easy to perform, but shouldn't be used to accurately determine cell death, let alone differentiate between modes of cell death. In addition, the assay only really works well for adherent cells and the cells are lysed in the process and the assay can therefore not be used in a multiplex set up.

Detection method: Absorption

Pros: Cheap and relatively easy

Cons: Sensitivity, not a cell death assay, doesn't discriminate, samples destroyed in the process

Crystal Violet

What I like about using crystal violet is that the assay doesn't depend on metabolic activity of the cells, but exclusively on the number of adherent cells remaining in your well. In addition, the morphology of the stained cells can easily be judged by simple light microscopy. Since the dye can be re-dissolved with methanol after staining and washing, the assay can also be used to accurately quantify the number of cells remaining in your well (Figure 2). Drawbacks are, of course, that the assay can only be used for adherent cells, that the cells are fixed during staining and that it’s hard to integrate the assay in a multiplex procedure. Since the assay only determines loss of cells (which can also be lack of proliferation) no conclusions about the mode of cell death can be drawn from crystal violet staining alone. I find this assay particularly useful for illustrating colony outgrowth after treatment, though it can also be used for routine screening. Because the cell mono-layer easily gets damaged in the execution of the assay during washing, fixing or staining, the assay is not very suitable for small well formats. It works well enough in larger wells, down to a 24-well format, since the relative contribution of some minor damage to the mono-layer won’t influence the outcome of the assay as much in larger wells as in smaller wells.

Figure 2. Crystal violet assay. A: Titration of L929 cells stained with crystal violet to generate a standard curve on a double log scale. Insert shows a macroscopic view of the wells with increasing amounts of cells. B: The reactive oxygen species (ROS) scavenger butylated hydroxyanisole (BHA) protects L929 cells from TNF/zVAD-induced necroptosis. However, higher concentrations of BHA are toxic. C: Macroscopic (top) and microscopic images (bottom) of L929 cells untreated (left), treated with TNF/zVAD (middle) or TNF/zVAD+necrostatin-1 (right). Cells were treated for 4h after which the medium was refreshed and cells were allowed to grow for a week.

Detection methods: Light microscopy, absorption (after re-dissolving dye in methanol)

Pros: Very cheap and relatively easy, objectivity

Cons: Sensitivity, not a cell death assay, doesn't discriminate, only works for adherent cells, cells fixed in the process, prone to errors when the cell mono-layer is scratched, toxicity of the reagents

ATP assays

Luciferase-based assay that determine cellular ATP levels, such as Promega’s Cell Titer Glo, have become very popular, especially in high throughput screens. Ease-of-use is the main selling point of these assays, as they only require a single reagent addition and a short incubation time. In addition, the assay is very sensitive. However, this assay does not determine cell death. Rather, a reduction in ATP levels can reflect several scenarios. First, a reduction in ATP levels in a given well might indicate a reduction in cell numbers as a consequence of death, but could also reflect a reduction in cell growth. If the treated cells expanded more slowly than the control cells, this would be reflected in a relative reduction in ATP levels. Second, overall ATP levels may be transiently reduced in cells under stress without this resulting in cell death. Third, during apoptosis ATP levels actually increase before cell death occurs. Thus, cell death assays based on determining relative ATP levels can very easily lead to invalid results and wrong conclusions. For an initial screen, such an assay might be useful but one should be extremely cautious to draw conclusions based on results obtained from ATP assays alone.

Detection method: Luminescence

Pros: Easy, single addition, suitable for HTS, very sensitive

Cons: Not a cell death assay, doesn't discriminate, samples destroyed in the process

Dye exclusion

In many older papers ‘apoptosis’ is equated with cell permeability for DNA-binding dyes such as propidium iodide (PI) or ethidium bromide (EthBr). However, apoptotic cells only become permeable to such dyes at a very late stage, while necroptotic cells lose plasma membrane integrity much earlier. Thus, dye uptake is either a sign of necroptosis, primary necrosis or late-stage apoptosis (Figure 3). Bear this in mind when reading older papers and don’t make this mistake yourself. Of course, this is not only true for old-fashioned dyes such as PI, but also for newer dyes such as the Sytoxdyes from Life Technologies (formerly Molecular Probes/Invitrogen). Although dye exclusion by itself has limited usability, it’s a method that can easily be combined with other methods in a multiplex assay, as I’ll discuss below.

Detection method: Fluorescence

Pros: Cheap and very easy, single addition and no washing required, suitable for both adherent and suspension cells, many dyes available in different colours, suitable for HTS

Cons: Only indicates necrosis or necroptosis, not early-stage apoptosis

LDH Release

Figure 3. Sytox green staining vs. LDH release from U937 cells
treated with TNFa in the presence of increasing amounts
of zVADfmk for 24 hours.

When cells lose membrane integrity, their content is spilled in the environment. This includes the enzyme lactate dehydrogenase (LDH). Activity of this enzyme can easily be detected with a commercially available colorimetric assay. The great advantage of this assay is that only the culture medium of the cells is required and that the cells themselves can therefore be used for other assays, for example Western blot or FACS. I found this assay to be very similar in sensitivity to dye exclusion (Figure 3). Of course, just like dye exclusion, the assay only determines the rate of necrosis which can be either a consequence of necroptosis, primary necrosis or secondary necrosis in late-stage apoptosis. An advantage over dye uptake is that you only need an absorption reader for detection and these are usually cheaper than fluorescence readers, although the assay itself is somewhat more expensive. 

Detection method: Absorption

Pros: Cheap and very easy, single addition to culture supernatant, cells can still be used for continued culture or other assays, suitable for both adherent and suspension cells

Cons: Only indicates necrosis or necroptosis, not early-stage apoptosis

Resorufin/Resazurin

The principle of this assay, sold as AlamarBlue or Cell Titer Blue, is the conversion of blue, non-fluorescent, resazurin to red, fluorescent, resorufin by living cells in an oxidation reaction. The assay has several advantages: the cells are not destroyed in the process, the results can be measured with either a fluorometer (red) or an absorbance spectrometer (red/blue) and can easily be combined with other assays that utilize different fluorescence wave lengths or luminescence. In addition, the results are aesthetically pleasing (Figure 3).

Figure 4. Cell Titer Blue on U937 cells treated with various stimuli/inhibitor titrations in a 96-well plate (A). Wells with living cells turn pink, others remain blue. In B I plotted the rate of resorufin turnover by cells treated overnight with TNFa in the presence of increasing amounts of zVAD versus Sytox Green staining.

However, just like ATP assays and such, the assay doesn't indicate cell death but rather cell metabolism. Since different cells display different rates of oxidative metabolism under different circumstances, it will take some experimenting before reliable results can be obtained from this assay. I found that it takes jurkat or U937 cells about 2-4 hours to metabolize resazurin, but I found that L929 cells act rather unpredictably in this assay. The assay can be used as either an end-point read out or kinetic read out. In the form of Amplex Red, resorufin can also be used to determine a rapid oxidative burst. Back when I worked on neutrophils we used Amplex Red routinely to determine neutrophil responses to certain stimuli.

Detection method: Absorption or fluorescence

Pros: Fairly cheap (I only once got a 10 mL sample of Cell Titer Blue and used it for years) and easy, single addition followed by incubation and measurement, suitable for multiplex assays, suitable for HTS, cells remain alive and intact

Cons: Indicates oxidative metabolism, not a cell death assay, doesn’t discriminate

Annexin V Binding

Probably my favourite apoptosis assay is binding of fluorescently labelled Annexin V binding followed by flow cytometry (FACS). Annexin V binds to phosphatidylserine (PS) which is exposed on the outer membrane of apoptotic cells. PS exposure is a passive process and happens when a cell’s active mechanism for retaining PS on the inside of the plasma membrane (the enzyme family collectively known as ‘flippases’) is compromised. Flippases are ATP-dependent enzymes that are sensitive to Ca²⁺. Thus, a drop in cellular ATP levels or a rise in intracellular Ca²⁺ levels will inhibit flippases and lead to PS exposure.

Figure 5. Membrane asymmetry in healthy cells versus PS exposure in apoptotic cells. An image I drew years ago.

Not too much is actually known about the regulation of these flippases during apoptosis, but PS exposure certainly is a very accurate hallmark of early apoptosis. However, PS exposure isreversible, some cells (such as macrophages) constitutively bind low levels of Annexin V and PS exposure can occur under certain rare  conditions (such as Barth syndrome) in the absence of apoptosis. Nevertheless, I’ve generally found Annexin V staining to correlate nicely with apoptosis. Of course, when the cell membrane integrity is compromised, Annexin V will also enter the cell and stain both sides of the membrane. Thus, high Annexin V staining alone can be an indication of either apoptosis or necrosis. Whatever the case, when a cell displays high Annexin V positivity something’s wrong. Annexin V binding can (and should) easily be combined with dye exclusion for accurate differentiation of (early) apoptotic cells and necrotic cells. Bear in mind that Annexin V binding is Ca²⁺-dependent and your binding buffer should therefore always contain ~2.5 mM CaCl₂. If you wash away the Ca²⁺, the Annexin V will also fall off. Finally, living cells will constantly expose low amounts of PS that are actively transported back in side and therefore constant exposure of living cells to Annexin V will very slowly lead to the uptake of the Annexin V and the staining of the cells. However, if you keep the cells on ice, you effectively fix the plasma membranes and the PS levels in the outer membrane won’t change anymore, even if you leave the cells unfixed.

Detection method: Fluorescence

Pros: Accurate assay for apoptosis, sensitive, easily combined with other assays

Cons: Indicates both early apoptosis and necrosis, not suitable for HTS

Figure 5. Annexin V staining of jurkat cells either deficient in FADD (DEF) or reconstituted with FADD (REC) treated with TNFa or TRAIL in the presence of the indicated inhibitors. Samples were taken every 2 hours, stained with Annexin-V-FITC and analyzed by flow cytometry.

Caspase activity

Caspase activity can be determined in a variety of ways and is a fairly reliably indicator of apoptosis. Of course, as I mentioned in the first post of this series, caspases are not exclusively activated during apoptosis and it’s not trivial to tell the activity of one caspase accurately apart from the activity of another caspase. However, caspase-3 activity in particular is certainly a hallmark of apoptosis. Thus, you’ll always find caspase-3 to be very active in apoptotic cells, although limited caspase-3 activation can occur in non-apoptotic cells. To determine caspase-3 activity any assay containing the tetra-peptide substrate ‘DEVD’ will do. I prefer Ac-DEVD-AFC over AMC labelled substrates, since they seem to be more sensitive and AFC will also turn yellow when released from the DEVD moiety, which can even be detected by absorption. Those are available as fluorescent or luminescent assays but you can also buy the substrate and make your own lysis buffer. The substrate is also somewhat cell permeable and can therefore be added to cells before inducing apoptosis and then be used to kinetically determine the increase in apoptosis in a well. However, such an approach is more likely to detect late-stage apoptosis, when the plasma membrane of the cells becomes compromised. In addition, the DEVD is also consumed by the proteasome, so healthy cells will hydrolyse it very slowly.  Therefore, rather than stating that you’re measuring caspase-3 activity when using DEVD as a substrate, state that you’re measuring DEVDase activity as you can’t be absolutely certain that the activity you measure is derived from caspase-3. Still, a fluorescently labelled tetra peptide substrate can easily be combined with dye exclusion and viability (resazurin/resorufin) to determine whether your cells have become apoptotic or necrotic after treatment.

If you want to obtain more accurate information about the particular caspase involved, you could consider using an antibody that only detects the cleaved for of the caspase, tag it with a fluorescent label and perform flow cytometry. However, only the executioner caspases (3, 6 and 7) require cleavage for activation and the available antibodies detect other proteins with the same cleavage site as well. These are often cleaved as a consequence of caspase activity (caspases like to cleave their own linkers and will cleave every other protein with the same epitope as well) which is why Western blots with active caspase antibodies will often show a large amount of bands. In flow cytometry or microscopy assays you get no information about the size of the proteins labelled and therefore no accurate information about whether you’re really looking at the caspase or a product of caspase activity. A better method to determine which caspase has been activated is to label all active caspases with a biotinylated substrate, such as bVAD, bEVD or bVEID, perform a pull-down with streptavidin beads and detect your active caspases on Western blot.

Figure 6. Active caspase detection in cell extracts. Extracts were activated by addition of cytochomre c and caspase activity was detected by hydrolysis of the substrate Ac-DEVD-AFC (A) at the indicated time points or active caspases were labeled with bEVD-AOMK, pulled down with streptavidin beads and analyzed on Western blot (B). Active caspase-6 and -3 could be detected. Caspase-8 is cleaved (by caspase-6) but not activated under these conditions. See van Raam et al. for further details.

Detection method: Fluorescence, luminescence or absorbance

Pros: Accurate assay for apoptosis, sensitive, suitable for multiplex, suitable for HTS

Cons: Indicates mostly late apoptosis, risk of false positives

The ultimate multiplex assay?

In a good multiplex assay, you want to combine at least one parameter to detect necroptotic cells and one to detect apoptotic cells. I generally prefer to combine Annexin V binding with dye exclusion on a flow cytometer, as flow cytometry provides a very versatile platform and also provides you with valuable information about cell morphology, besides fluorescence. I've personally come to prefer FITC-labelled Annexin V (I get it from Bender Med, now eBiosciences) with Sytox Red. FITC fluorescence and Sytox Red are excited by different lasers and there’s therefore no need for compensation, while PI and FITC are excited by the same laser and their fluorescent peaks are close together. However, this assay is less suitable for HTS, although most steps can easily be automated. I know the Vandenabeele lab has developed an assay wherein they combine Sytox Green with Ac-DEVD-AMC to detect caspase activity. This seems to work well for them, although I’d combine it with a viability assay in the form of Cell Titer Blue. Assays that can be performed kinetically are always superior to end-point assays, but in the end the use of inhibitors can provide you with the most accurate information.

How to differentiate apoptotic from necroptotic cell death? Part II.

Model systems for necroptosis research

Since necroptosis is what happens when caspase-8 fails to activate, the conditional caspase-8 knockout mouse reveals which tissues are susceptible to this form of cell death during development. Primarily endothelial cells, hematopoietic progenitor cells and leukocytes are susceptible to developmental necroptosis in the absence of caspase-8. CD8⁺ T-cells, monocytes and neutrophils require caspase-8 activity to properly develop and expand and are also quite susceptible to TNFa-induced necroptosis. Ischaemia/reperfusion injury also triggers necroptotic cell death in liver and kidney cells but this may represent a different form of necroptosis, distinct from TNFa-induced necroptosis. Necroptosis can be triggered by several other stimuli, such as immune receptor activation, TLR3 ligation, and RIG-I signalling. There may be other stimuli that induce necroptosis, but suffice to say that various tissues and cells under various conditions are susceptible to this form of cell death. Necroptosis can happen any time, you never know where and you never know when it will strike…

Cellular models of necroptosis

The most commonly used model cell line for necroptosis are the murine fibroblast cells L929. These cells are extremely sensitive to TNFa-induced necroptosis in the presence of the broad-spectrum caspase inhibitor z-VAD-fmk. In fact, they’ll also undergo necroptosis with z-VAD-fmk alone because the L929 cells produce low amounts of TNFa (and other cytokines!) constitutively. L929 cells are useful for the screening of anti- or pro-necroptotic compounds but shouldn’t be considered ‘real’ cells. They really respond very oddly in a number of ways and can’t be trusted entirely, in my experience. They express very high levels of RIPK3, the downstream effector of RIPK1 and this is most likely what makes them so susceptible to necroptotic death.

Mouse embryonic fibroblasts (MEFs) are sometimes susceptible to necroptosis as well, but not always. Bear in mind that authors tend to publish their successful experiments, rather than their failures, so when you see a paper in which necroptosis was induced in MEFs, don’t assume that your MEFs will respond the same way. It will work for some MEFs, but not for others and it remains hard to predict how MEFs will respond.

When dealing with human cells, either primary or cell lines, bear in mind that humans express caspase-10 besides caspase-8. Caspase-10 is activated in the same pathways as 8 and caspase-10 expression appears to be sufficient to prevent necroptosis (since patients deficient in either caspase-8 or -10 are quite viable but often develop Acute Lymhoid Proliferation Syndrome; ALPS) even though caspase-10 can’t substitute entirely for caspase-8. Caspase-10 didn’t evolve in humans but is in fact much older, the rodent lineage simply lost the gene (Figure 1). Presumably, rodent caspase-8 has taken over the functions of both caspase-8 and -10 (see also my review on the subject).

Figure 1: Evolution of caspase-8. Bony fish and their ancestors express two caspase-8 variants: the direct precursor to caspase-8 and -10 ('caspase-810') as well as caspase-18. Caspase-810 splits into two distinct genes (caspase-8 and caspase-10) just before tetrapods, while mammals lost caspase-18. Rodents, finally, lost caspase-10 as well. From my review.

Most primary human leukocytes appear to be susceptible to necroptosis, although no reports of a clear comparison has been published. In the older literature, TNFa or Fas-induced cell death in the presence of z-VAD-fmk is mentioned several times (here, here, here and here, for example) and it seems safe to assume that in most of these cases the cells succumbed to necroptotic cell death.

Certain clones of Jurkat cells are susceptible to necroptosis, but not all of them. In my experience, the FADD-deficient clone 5C3 is highly susceptible to necroptotic death induced by TNFa. The ‘wild type’ jurkat cell line A3 is not and the caspase-8 deficient line I9.2 is only mildly susceptible by itself, but, surprisingly, becomes more susceptible upon addition of z-VAD-fmk, suggesting that caspase-8 is not essential to prevent necroptosis in these cells. The RIPK1-deficient jurkat cell line is not susceptible to necroptosis unless reconstituted with RIPK1 harbouring a cleavage site mutation, The parental clone doesn’t undergo necroptosis upon TNFa stimulation in the presence of z-VAD-fmk. However, the RIPK1 deficient cells are extremely susceptible to all forms of apoptosis, suggesting either an important role of RIPK1 in preventing apoptosis or that these cells lack another anti-apoptotic factor in addition to RIPK1. These cells were initially generated by selecting randomly mutated jurkat clones against the ability to activate NF-kB, although later research has shown that RIPK1 is dispensable for NF-kB activation downstream of TNFa signalling.

The monocytic cell line U937 is also extremely susceptible to TNFa/z-VAD-fmk-induced necroptosis. These cells, as mentioned before ,will also produce high amount of TNFa in a RIPK1-dependent manner when stressed with a variety of stimuli in the presence of z-VAD-fmk. Other monocytic cell lines I tried, THP1 and NB4, are not susceptible to necroptotic cell death. Interestingly, in contrast to jurkat cells, U937 cells also undergo necroptosis when stimulated with TRAIL in the presence of z-VAD-fmk (Figure 2), even though jurkat cells are susceptible to TRAIL-induced apoptosis when FADD is expressed. This suggests that TRAIL may signal differently in U937 cells than it does in jurkat cells. I don’t know why this is, but I’ll share my observation and if anyone has an explanation, I’d be happy to collaborate.

Figure 2: TRAIL induces necroptosis in U937 cells.Treating U937 cells with TRAIL induced necrosis (Sytox Red uptake) which could not be prevented by zVAD, unlike TRAIL-induced apoptosis in Jurkat (A3) cells (A). Necrostatin only prevented necrosis in TRAIL+zVAD treated U937s (B). On Western blot, RIP1 cleavage was barely affected by zVAD in U937 cells, whereas it could be prevented in Jurkat cells. PARP cleavage was affected equally in both cell types (C and D).

These are all the necroptosis models that I’m experienced with. I’m sure there are others but you should run some tests to determine whether or not your favourite cell line is susceptible to necroptotic cell death. In the next chapter, I’ll outline several methods for determining cell death that are suitable for necroptosis research.

What determines whether a given cell is susceptible to necroptosis?

The most important factor that determines a cell’s susceptibility to necroptosis is the expression level of RIPK3 as well as expression of the downstream effector MLKL (mixed lineage kinase domain-like). Those cells that are most susceptible to necroptosis, appear to be those that express the highest levels of RIPK3 (such as L929 cells) while RIPK1 levels are normally quite stable among different cells. In fact, ordinary cells such as HeLa cells can be made susceptible to necroptosis by over-expression of RIPK3.

A recent paper in Science indicates that, indeed, RIPK3 kinase activity is required for necroptotic signalling, as mice expressing a kinase death mutant of RIPK3 did not succumb to necroptosis in the absence of caspase-8.  However, in the presence of caspase-8, these mice succumbed to massive caspase-8-dependent apoptosis. Thus, the kinase activity of RIPK3 both induces necroptosis while RIPK3 can act as a scaffold to promote apoptosis in the absence of kinase activity.

How to differentiate apoptotic from necroptotic cell death? Part I.

Figure 1. Example of a necroptotic cell (left) versus an
apoptotic cell (right). Image courtesy of the Vadenabeele lab.

Since the discovery that at least two forms of programmed cell death exist, apoptosis and necroptosis, a need has arisen to accurately discriminate between these two forms of cell death. Determining whether a cell is death or alive seems straightforward enough but all commonly used cell death assays have certain caveats and exceptions that one should be aware of before accurate conclusions can be drawn. In the following essays, I’ll go over the different cell death assays that I’m familiar with and discuss their applicability in discriminating between apoptotic and necroptotic cell death as well as common pitfalls and limitations using examples from my own experiments.

What discriminates apoptotic from necroptotic cell death?

Simply put: apoptosis is caspase-dependent cell death, whereas necroptosis is RIP-kinase-dependent cell death (for a brief review see Walsh, 2014) . Apoptosis occurs in an orderly manner: the cell’s DNA is digested into chunks of roughly 300 base pairs and the cell contents are packaged in small vesicles, the apoptotic bodies, that are phagocytosed by neighbouring cells and tissue macrophages for recycling. During necroptosis, on the other hand, the plasma membrane ruptures and the cell contents are spilled into the environment, stimulating a local inflammatory response. Apoptosis is the dominant form of cell death, since caspases have the ability to inactivate the RIPK signalling pathway. Thus, if your cells are dying while caspases are active, they’re most likely dying by apoptosis. However, caspase activity also occurs outside apoptosis and although vendors might claim that their product accurately detects the activity of one caspase or another, no simple method exists to make this distinction.

In general, caspase activity can be measured by the rate of cleavage of a tetra peptide linked to a fluorophore. For example, products to determine caspase-3/7 activity are usually based on the tetra peptide ‘DEVD’ linked to, for example, an AMC or AFC fluorescent moiety. Active caspase-3 will cleave the peptide after the last ‘D’, releasing the fluorescent moiety. An increase in fluorescence can then be said to correlate with an increase of caspase-3 activity. However, no single tetra peptide is exclusively cleaved by one caspase or another. DEVD is indeed a preferred substrate of both caspase-3 and -7 but can also be processed readily by caspase-8 and even the proteasome. Thus, an increase in DEVDase activity in your sample doesn’t necessarily indicate an increase in caspase-3 activity. The same holds true for every other tetra-peptide-based assay.

In addition, as I mentioned earlier, there are many scenarios wherein moderate caspase activity is not followed by cell death. Caspase-3 activity, for example, normally associated with end-stage apoptosis, also plays a role in memory formation in the brain in the absence of cell death. Caspase-7 may be involved in inflammation and initial activation of caspase-8 signals cell survival, rather than death. Thus, although apoptosis is invariably associated with caspase activity, caspase activity does not necessarily lead to apoptosis.

Unfortunately, no simple methods exists to determine the activity of the RIP kinases or their downstream effectors. However, there are several potent inhibitors of RIPK1 on the market: the necrostatins. The first of these, necrostatin-1, has now been shown to inhibit at least one additional enzyme, indoleamine-2,3-dioxygenase (IDO; Vandenabeele et al), and should therefore be used with caution. An alternative is now available in the form of necrostatin-1s (Nec-1s) which still prevents necroptotic cell death, in the absence of IDO inhibition. However, we can’t know what we haven’t looked for and even this inhibitor may have off-target effects.

Notwithstanding the fact that all chemical inhibitors may, to a greater or lesser extent, have off-target effects, utilization of these inhibitors still provides us with relatively simple means of discriminating the two forms of cell death. Thus, if you want to investigate whether a death-inducing compound kills your cells by apoptosis or necroptosis you could do a control experiment in the presence of a broad spectrum caspase inhibitor, such as z-VAD-fmk or boc-D-fmk, and/or necrostatin. If the caspase inhibitor rescues the death phenotype, the cells were most likely killed by apoptosis, if the necrostatin rescues the phenotype, the cells were most likely dying by necroptosis. However, there is a caveat. Certain substances have the ability to induce necroptotic cell death, but only in the absence of caspase activity as a consequence of autocrine TNFa signalling. To determine whether your compound induces such ‘secondary necroptosis’ I advice using a combination of z-VAD-fmk and necrostatin as a control besides z-VAD-fmk and necrostatin alone. If your compound appears to induce secondary necroptosis, perform a control in the presence of an anti-TNFa antibody to make sure that the observed cell death isn't a consequence of autocrine TNFa signalling. 

Figure 2. Secondary necroptosis in U937 cells. For this experiment, U937 cells where stimulated with the DNA-damaging agent etoposide in the presence of various inhibitors, as indicated. As you can see in panel A, the cells produce large amounts of TNFa upon etoposide stimulation in the presence of zVAD. Addition of necrostatin-1 (Nec1) inhibits this TNFa production. In panel B, you can see that the TNFa production is so high that the cells undergo secondary necroptosis. Vability is restored by addition of necrostatin-1 or an anti-TNFa antibody (aTNF). See van Raam et al., 2012 for a detailed description.

Your Western blots deceive you!

Ever since the technique to blot protein samples on nitrocellulose membrane was developed, Western blot has been a go-to technique for every molecular biologist. The method is simple enough and the results seem straightforward, but correct interpretation of your results is a crucial and often overlooked step.

In Figure 1, I present an example of a straightforward Western blot result. For this experiment, I differentiated the myeloid cell line PLB into functioning granulocytes by stimulating them with dimethylformamide (DMF) for 6 days . Afterwards, I took a sample of my stimulated and control cells as well as a sample of primary human granulocytes (PMN), dissolved these in sample buffer, ran the samples next to each other on an SDS-PAGE gels, blotted the gel to a membrane and probed that membrane with antibodies against the NADPH-oxidase components p67 and p47. Finally, I probed the blot with fluorescently labelled antibodies and scanned it on a Li-Cor Odyssey infrared scanner to detect the signal. As you can see, the neutrophils (PMN) express high levels of both p67 (in red) and p47 (in green) whereas the control PLB cells express nothing and the treated PLB cells express levels comparable to the neutrophils, demonstrating that the experiment was successful. I also probed the blot for β-actin to demonstrate equal loading of the samples. This is a perfectly straightforward result: the protein is either there or not there at all. No discussion about it! However, most blots won’t be so straightforward.

Figure 1. Western blot probed for p67phox (red) and p47phox (green) or actin (lower panel). Loading control on the far left lane with molecular weight bands as indicated in kDa. A sample of human neutrophils (PMN) is shown as a control as compared to non-differentiated PLB cells and PLB cells differentiated with 0.5% DMF.

In most cases, your protein of interest won’t be either there or not there. The level of expression will be reduced or increased by a certain margin. Western blots can be very deceptive in demonstrating such subtle effects, as I will outline below.

Controls

First and foremost in scientific practice come the controls. Your control and experiment samples should have received the exact same treatment to be valid and run side by side on a gel for proper comparison. I know it happens far too often that you go through the whole procedure only to discover a nasty spot in either your sample or control lane. Then you go and repeat the whole procedure, only to have the other lane smudged this time. If only you cropped both pictures and spliced them next to each other, you’d have a perfect result… Unfortunately, this is not allowed. Such practice reeks of scientific fraud and makes it impossible to judge whether or not your result was real, even if you mean well. Nope, my friend, either it’s back to the lab, or show the smudges! There’s nothing wrong with showing smudges, one should not be afraid of that. Smudges are much preferred over fraud!

Detection limits

The second issue to consider before stating whether or not a certain protein has disappeared or not (for example after siRNA treatment) is the detection limit of your protein of interest. The problem with all commercial antibodies is that nobody really knows just how strongly they interact with their target protein. Anybody who’s tried several different antibodies against the same target knows that some will give you a very strong signal and low background, while others will give you a weak signal and high background. Anything in between, in any possible combination, can also be encountered. It’s completely random! This indicates that the strength of the signal on your blot is not so much an indication of relative protein abundance as of antibody quality. Though, of course, protein abundance also helps.

Detection limit becomes especially relevant when your signal is very weak. This means you’re drawing conclusions close to the detection limit of your assay. Thus, when comparing two lanes on a blot, it may seem that in one lane the protein is there and in the other it’s absent, while in reality the difference is no more than 10% or so. What makes the blot deceptive is the detection limit of your protein. It could just happen that you loaded 10% less of your experimental sample and that this is enough to make your protein of interest undetectable, even though it’s still there. If you now compare the control and your experimental lane, it will look like your protein of interest has disappeared completely after treatment and since we usually use highly abundant proteins, such as actin, as our loading controls, this 10% difference in loading will go unnoticed. I see these kind of results quite often with siRNA controls; very weak signals that appear to show a black and white difference. Results collected on old fashioned X-ray film are particularly sensitive to this form of deception, since a pixel on the film is either black or not. The grey values on a film are entirely derived from pixel density. In contrast, results collected with a fluorescence imager are much less sensitive to deception, since every pixel can have a wide range of values. On the Li-Cor Odyssey, for example, a pixel can have a value between 1 and 40,000, providing a huge dynamic range.

Post-processing errors

Every digital image can be manipulated to show only what one desires others to see. Of course, manipulating only part of the image (say the control lane) and not the rest is fraud, but there’s a huge grey area of manipulation that is allowed, but not quite correct. I provide an example in Figure 2. For this experiment, I tried to knock down a gene with siRNA. The blot shows my control sample (Ctrl) and two different siRNA’s (1 and 2). Panel A is what the blot actually looks like after I acquired the data with the Odyssey infrared imager, panel B is manipulated to show what I want to see (protein levels decreased after siRNA treatment) and panel C is my loading control (HSP90). As you can see, the effect of siRNA treatment appears to be much greater in panel B than in panel A, even though both show the exact same blot. What I did to generate panel B was to adjust the image display curves, rather than merely the contrast. You should never, ever do that! Ever. Because by adjusting the curves you’re discarding data you don’t like. You’re telling the program that you don’t care about values above or below a certain threshold, thus you get rid of pixels with very low or very high values. Of course, this eliminates your background, but you also lose information that might be valuable, such as the weak bands visible in panel A below and above the main bands. In addition, you enhance and multiply a small difference to make it seem much greater. This might indeed help you get your work published in Nature, wherein most blots are suspiciously squeaky clean, but it really isn't the way to go.

Figure 2. Control sample (Ctrl) and samples treated with siRNA 1 or 2. A) unprocessed image, B) processed image, C) loading control (HSP90). 

Loading controls

Another obvious problem is with the loading control. As I mentioned before, we tend to choose a highly abundant protein as our loading control, such as Actin. However, if your control is much more abundant than your protein of interest, the result may be highly deceptive. I provide an example in Figure 3. For this experiment, I simply took a cell lysate and diluted this with sample buffer in 10%-steps (thus, lane 1 is 100%, lane 2 90%, 80% etc.). Then I probed the blot with an anti-actin antibody, followed by a fluorescently-labelled antibody and I scanned the blot with the Odyssey Infrared scanner. The nice thing about this technology is that I can actually quantify my signal and draw a graph, as shown in the figure. Now, you’ll notice that the difference between the first 5 lanes is very hard to spot with the naked eye, even though there’s an almost 2-fold difference in loading (100% vs 60%)! In addition, the computer can barely tell the difference between the first 3 lanes, even though I loaded only 80% of the sample in lane 3. Thus, also the intensity of your loading control can be very deceptive and you should be aware that the same phenomenon occurs for every protein you blot for. If you’re trying to draw conclusions outside the linear range of detection, whether by eye or computer, you’re going to be deceived.   

Figure 3. Decreasing amounts of sample were loaded on gel with steps of 10%. The blot was probed for human Actin and data acquired on the Li Cor Odyssey.

Finally, the molecular weight of your controls and experimental samples matter. Not every protein in your sample is going to transfer equally well and molecular weight is an important determinant for transfer efficiency. Large proteins tend to transfer slower whereas small proteins can actually be transferred straight through your blot if your transfer time is too long. Thus, ideally, your loading control and protein of interest should be of both similar size and similar abundance.

Solutions

It’s up to you to provide the right controls for your experiments and to make sure your samples and controls are on the same blot. You can not, and should not, compare samples from different blots as they may have been differently exposed and processed.

You should know the detection limit of your protein of interest and the affinity of your antibody. When you first start using a new antibody, run a control blot with a dilution range of your sample to make sure you’re measuring in the linear range of your protein.

Image manipulation is allowed, of course, you’re already doing it when you’re exposing your film for different lengths of time. However, only manipulate the whole blot and never discard data. Better to show some noise and background. Find some useful hints and tips on this website. Nowadays, I much prefer using fluorescence scanners, such as the Odyssey, the acquisition Western blot data because the scanned imaged can be quantified and I can easily make different exposures of my blot, simply by setting the exposure level. When using the Odyssey, always set the acquisition for the maximal possible exposure, without over-exposing the image (shown as white pixels in the colour view or blue pixels in the grey scale view). That way, you collect as much data as possible and you can always post-process your image to get a prettier picture. More important than a pretty picture, however, is the ability to graph and calculate your data.  

I hope these examples and guidelines give you some idea of how to interpret Western blot results, be it your own data or published work. 

How to make the mutation you want and succeed every time

Good research is highly reliant on good tools. Among the most important tools in molecular biology and biochemistry are mutant versions of the genes we study. These mutants help us understand how the proteins encoded by the genes function, both in vivo and in vitro. Today, I shall discuss the site-directed mutagenesis protocol I've been using to get my mutants. My protocol works. Every. Single. Time.

When it comes to site-directed mutagenesis many people prefer the easy way out and use a kit. The QuickChange kit from Stratagene, for example, is a popular choice. However, this kit is rather expensive, unnecessarily complex and really not as good as advertised. The whole procedure can be performed with materials you probably already have in your lab at a fraction of the cost, if you follow the exact steps outlined below.

1. Primer Design

The first step towards generating a successful mutation is designing your primers. The same rules of thumb apply when designing a mutagenesis primer as for designing a cloning primer: you're looking for a primer that extents at least 9 bases on both sites of the mutation, with a GC content of 40-60% and a melting temperature of about 52-55C. I always use complementary primers, a pair of primers that are each others exact opposites. For example, in figure 1, I show you the primer I designed to mutate an arginine residue, R54, to a cysteine in human Ubiquitin B. I did this by changing a single 'C' in the primer to a 'T', as indicated by the yellow bar. The codon 'CGC' encodes an arginine, whereas ' TGC' encodes a cysteine. Easy as that!

Figure 1. A primer designed  to mutate arginine 54 to a cysteine (R54C) in human Ubiquitin B.

Notice how the primer extents for 10 bases to both side of the mutation? Shorter primers also work, but try to extent it at least 9 bases on both sides of the mutation while keeping the GC content within 40-60% (in this case, it's 52.4%) and the melting temperature between 52-55C (53.5C, in this case). Bear in mind that the mutation you make will not count towards the melting temperature of the primer, as the mutation represents a mismatch! Of course, you can introduce more than one mutation, but the primer will have to be proportionally longer. I've successfully introduced up to seven point mutations with a single primer pair in one reaction using this method.

2. PCR

Once your primer order has come in and you have your template ready, it's time to run a PCR reaction. For site-directed mutagenesis reactions, it's wise to use a reliable polymerase with proof-reading. Several options are available, but my favorite enzyme is Phusion from New England Biolabs. This enzyme is twice as fast as a regular DNA polymerase and a whopping four times as fast as other proof-reading enzymes, such as Pfu. It has, in my experience, also proven to be highly reliable. Using the Phusion enzyme and buffer, set up the following reaction:

200 ng DNA template

125 ng forward primer

125 ng reverse primer 

2 µL dNTP (2.5 mM each)

10 µL 5X Phusion buffer

1 µL Phusion

 x µL water (to 50 µL total)

The ideal amount of template to be used depends on a lot of factors, but I find 50-200 ng to be sufficient. If you're worried about background you can use a little less, but more generally works better.

Set up one reaction with and one without polymerase as a control! This is important, so make sure not to forget this control.

The following PCR program can be used for Phusion (Figure 2), adjust to primer melting temperature (2 degrees below the melting temperature of your primers is ideal) and plasmid size (30 seconds/kb, remember you're amplifying the entire plasmid plus your insert!). 18 cycles is generally sufficient.

Figure 2. General PCR program used for site-directed mutagenesis with Phusion. 

It's best to run the PCR in the morning. The program only takes about 2 hours in total, when using Phusion; enough time for a hearty lunch and a cup of coffee!

3. Analyze Product on Gel 

After the PCR is done, take 10 µL of your product and analyze it on a 1% agarose gel. This is a very important step in the mutagenesis process! If the mutagenesis reaction has worked you'll see a signifcant increase in the amount of DNA in your reaction with polymerase, as compared to your control reaction without polymerase. If you don't see this increase, the reaction hasn't worked and there's no point in proceeding! Have a good look at your primer design or try again with more template DNA. If the reaction has worked, you should see something similar to Figure 3.

Figure 3. PCR product of the mutagenesis reaction analyzed on gel. Samples 1-3 were run without polymerase, samples 4-6 were run with Phusion polymerase added to the reaction.

4. Digest Product with DpnI

If, and only if, you got significant amplification of your template after PCR, you can digest your product with the restriction enzyme DpnI. This enzyme only digests methylated DNA, so your template (which was amplified in and purified from E. coli) gets digested, whereas your mutation-bearing product, formed in the PCR reaction, remains intact. Simply add 1µL DpnI (I get mine from Promega these days, but any DpnI is fine) directly to your PCR tubes and put them at 37C for 2-3 hours. Don't forget them and leave them overnight, as there will be nothing left (I know from experience...).

5. Transform  into DH5a and Mini-prep

After digestion, simply take 5-10 µL of your product and transform into competent DH5a E. coli to amplify the DNA. Any strain of competent cells will work, really. Plate the cells in LB agar plates after the transformation and let them grow overnight at 37C. The plate of the no polymerase control should have no colonies, whereas the other plate should have plenty.

I know from experience that if you digest a product that did not look like it was amplified on gel you will still get no colonies on the no polymerase control plate and some on the other plate. However, none of these colonies will harbor the mutation. I don't know why this is, yet it happens.

The following day, pick some 12 colonies from your plate and grow mini-prep cultures overnight. You'll need enough DNA for sequencing (50 ng/µL) and I usually grow and prep 2 mL culture for high copy plasmids and 4 mL culture for low copy plasmids. I like to use 2xTY medium for my cultures, rather than LB, as the yield is higher.

For your mini-preps you can use any kit you like, but be aware that the popular Qiagen kits are expensive but absolutely no better than any other kit out there. I generally like the kits from Machery-Nagel or, if you really want to safe some money, you can buy the Econo-Prep columns from Epoch Life Sciences and make your own buffers. Mini-prep columns can be recycled and re-used several times, so there's really no use in spending a lot of money here.

6. Sequencing

After you've done your mini-preps, it's time to send out your DNA for sequencing. Most labs these days use commercial sequencing services, and prices have dropped significantly in recent years, so I don't find it necessary to safe money in this step at the risk of losing time. In general, the mutagenesis reaction is rather efficient, but I like to sequence about twelve clones of each reaction to be on the safe side. I find that at least one in four clones generally contains the desired mutation. 

If you don't have access to fast and cheap sequencing facilities, you can also consider designing your mutation so that you introduce a restriction site. That way, you can analyze your clones by restriction digest. However, nowadays, this is hardly cheaper, faster or more reliable than sequencing and you'll have to sequence anyway in the end.

Conclusion

Well, that's it! With this protocol, my mutagenesis reactions always work, provided I designed my primers right and used enough template. If you're introducing multiple mutations with very long primers, you might consider adding 1% DMSO to your PCR reaction. This prevents the formation of DNA super structures during amplification, but also decreases the efficiency of the polymerase. Good luck making some mutants of your own!

Apoptosis vs. Necroptosis

Cell death is a highly regulated process. Ask any cancer or stroke patient. In the former case, too little cell death is causing problems, in the latter it's too much cell death that's doing the damage. Every day, approximately 100 billion (!) cells die in your body and every day, all those cells are replaced. Every day you die a little and you never even noticed.

Most of those cells die by a process called apoptosis; a programmed form of cell death. Every cell is genetically programmed to undergo apoptosis, a sequence of orchestrated events leading to the cells demise, when the cell has either suffered irreparable internal damage or receives an extrinsic stimulus from its environment. The extrinsic signal to die is given when a cell is, for example, infected with a virus, has become old or redundant or has become dislodged from its usual place in the body. The cell is basically told to quietly commit suicide, dismantle itself and allow its remains to be recycled. Cells are equipped with 'death receptors', members of the TNFa super family, that receive the death signal.

As mentioned above, apoptosis can also be triggered intrinsically. When, for example, a cell's genome or essential organelles have suffered irreparable damage a sequence of events leads to the release of toxic proteins from the mitochondria, such as cytochrome c and SMAC, that induce the cell to undergo apoptosis and remove itself from the population.

Extrinsic versus intrinsic apoptosis (image by author)

Apoptosis is a quiet, dignified form of cell death that does not trigger an inflammatory response. Apoptosis depends on the sequential activation of the caspases; a family of cysteine proteases. At the top of the chain are the initiator caspases (caspase-8 and -10 for death-receptor-induced apoptosis, caspase-9 for intrinsically-triggered apoptosis), at the bottom are the executioner caspases (caspase-3, -6 and -7) that dismantle the cell. Diametrically opposed to apoptosis is necrosis; a messy form of cell death -wherein the cell's contents are spilled into the environment- that does elicit an inflammatory response.

Surprisingly, necrosis can also follow a genetically-encoded program, similar to apoptosis. However, programmed necrosis, now widely known as 'necroptosis', does not depend in caspase activity, but on the activity of a kinase: Receptor Interacting Protein Kinase 1 (RIPK1). RIPK1 functions as the initiator of the pathway, while several downstream kinases (most notable RIPK3) serve as the executioners. We don't know much about necroptosis yet, but new findings on this form of cell death are published almost daily. What we do know for sure is that caspase activity is essential to prevent it. Necroptosis only occurs in the absence of caspase activity (for a free review by some friends of mine, see here).

The initiators of death-receptor-dependent apoptosis, caspase-8 and caspase-10, have both been shown to cleave and inactivate RIPK1 (reviewed by me here). Thus, if those caspases are activated, RIPK1 is inactivated. The opposite is also true: When the gene for caspase-8 is knocked out in mice (mice don't have the gene for caspase-10), the embryo deficient in caspase-8 dies on the eleventh day after gestation (Varfolomeev et al. 1998). This is a crucial day in the development of the murine embryo, since at that time the embryo's own blood circulation kicks in. In the absence of caspase-8 activity, the hearth and blood vessels of the embryonic mice fail to develop. Experiments with conditional knockout mice, mice that are only deficient for caspase-8 in certain tissues, have revealed that caspase-8 activity is also essential for the development of the immune system (see Kang et al.).

This failure of caspase-8 deficient embryos to develop, is entirely due to the activity of RIPK1 and its downstream effector RIPK3. Knock either one of these genes out concomitantly with caspase-8 or the adapter protein FADD (essential for the activation of caspase-8) and the mouse develops just fine (here, here, here and here; only the last one, the one I'm on, is free) . Or at least; it develops past this crucial stage, past day 11, for knockout of RIPK1 is lethal in itself. RIPK1 has pro-life as well as pro-death functions, but knockout of RIPK3 is relatively safe. Mice deficient in RIPK3 as well as caspase-8 develop into relatively healthy individuals. They still have some problems, of course, caspase-8 and RIPK3 are not entirely useless genes you can just dispose of. They have problems dealing with viral infections, for example, and their T cells proliferate unchecked. They may have other problems too, that haven't surfaced yet.

Thus, caspase-8 has a crucial pro-survival role in shutting off RIPK1 and preventing it from inducing necroptosis. But how, then, does a cell wherein caspase-8 is activated not die by apoptosis instead? How does it live to develop into a healthy mouse or human? Caspase-8 activates through dimerization; two molecules of caspase-8 are forcefully brought together to form an active complex. The previously mentioned adapter protein FADD is essential for initiating this process of dimerization, but recent evidence has shown that once a few dimers are formed around clusters of FADD, more caspase-8 dimers can form independent of FADD. An important clue comes from the observation that caspase-8 does not only activate when it dimerises with itself to form a homodimer, but can also when it forms a dimer with its cousin, FLIP (FLICE-like Inhibitory Protein), to form a heterodimer. FLIP is similar to caspase-8 but has no protease activity, it is an inactive caspase homologue. The heterodimer is active, but has a restricetd substrate repertoire; it cleaves the pro-apoptotic substrates of caspase-8 with very low efficiency, while it cleaves the non-apoptotic substrates of caspase-8 just as efficiently as the homodimer. I recently published a very readable review on the proliferative versus the apoptotic functions of caspase-8, find it here.

Apoptosis vs. Necroptosis vs. Survival  (image by author)

Now, this is not the end of it. There is much more to RIPK1 signaling than necroptosis; it plays an important role in immune activation and development too. In addition, recent evidence suggests that direct cleavage of RIPK1 cleavage by caspase-8 may not even be the key to prevention of necroptosis. Instead, caspase-8 may cleave CYLD, a de-ubiquitinating enzyme and an important regulator of RIPK1 activity. As long as CYLD is active, RIPK1 can promote necroptosis but if CYLD is inactivated RIPK1 is more likely to promote cell survival. However, caspase-8 is very bad at cleaving either CYLD or RIPK1. The paracaspase MALT1 can also cleave CYLD, an event that is crucial for the activation of T cells, but it is not so very good at it either. Could there be another caspase, downstream of caspase-8, that cleaves and inactivates RIPK1? Does cleavage of RIPK1 really lead to its inactivation or do the two fragments gain a different function? Does the caspase-8/FLIP heterodimer have other substrates, besides RIPK1 and CYLD? These and other important questions are currently under investigation. We're not even sure yet what the relevance of necroptosis is for either normal human physiology or pathology.

Surely these are exciting times for the fields of cell death and inflammation! I will use this blog to review the latest findings in these fields, both by myself and by others. I hope to attract opinionated readers both inside and outside the fields and get some discussions going.