Western blotting is often regarded as some kind of dark magic. You struggle and struggle before eventually making a pact with a dark entity of some kind, and then *poof* your blots work! In reality, a Western blot is as much of a science as anything else in the lab. It just happens to be a particularly finickity one which has so many potential points of failure that getting a good blot (the kind you would submit with a coveted Nature paper) can be a bit more of a war of attrition rather than an instant success. In this article, we will walk you through the somewhat tricky process of Western blotting for EV markers and other EV-associated proteins. To be clear, this is isn’t a protocol – that will differ based upon the equipment you have, your samples, antibodies and a plethora of other factors. However, we will give tips for each step along the way and address points at which things can go wrong.
If you are starting with a liquid sample (e.g., a purified collection volume (PCV) from a qEV column), lysing your EVs is not as simple as adding 1x RIPA buffer (radioimmunoprecipitation assay buffer; a general lysis buffer with an outdated and overly specific name) and hoping for the best. In our opinion, RIPA is still the way to go, but you may need a little trial and error to get the best results. The least harsh lysis method would be to treat your fractions or PCV as the diluent for concentrated RIPA to achieve a final concentration of 1x RIPA in your samples.
If you had a 10x RIPA solution and wanted a final volume of 200 μL per sample with a final concentration of 1x RIPA, this would mean taking 20 μL of 10x RIPA buffer and adding to it 180 μL of your fraction or PCV.
If you have a pellet of EVs (e.g., from the qEV Concentration Kit), simply resuspend in 1x RIPA to reach the desired sample volume. However you achieve lysis, make sure to add protease inhibitors! If you wish to study phosphorylation, phosphatase inhibitors are also needed.
EVs can be tougher to lyse than cells. Using RIPA at a higher concentration than it is intended to be used at (i.e., higher than 1x) can give a harsher lysis. Achieving a final concentration of 5x RIPA can improve lysis but may impact upon your protein of interest. It is always best, therefore, to try a range of RIPA concentrations on non-precious samples to check what works best for you and your proteins.
Some people get good results without the use of RIPA or any other lysis buffer. They just denature whole EVs and seem to get good signal for some proteins. We advise testing in your own hands to see how this might work for your application.
Ideally, you would know the optimal sample amount before lysing. However, this may not be feasible initially, as without knowing just how much protein will be in your lysed samples, it is difficult to make an informed decision on the amount of sample to use. This dilemma only applies if you are opting to load a specific amount of protein per well, however you might choose an alternative approach.
Instead, another option could be to load an equal volume of sample in each well, allowing you to compare fractions or your PCV to the protein-rich later fractions.
Your samples may not be concentrated enough due to low initial EV levels. If this is the case, you can try the qEV Concentration kit and resuspend in just the right amount of RIPA to run your Western blot.
There will be more EVs in total in the PCV than individual fractions, so if you are having trouble seeing proteins in fractions, then trying a concentrated PCV is a good place to start.
If you still need to assess individual fractions and are not getting a high enough concentration of EVs to measure protein with a Western blot then we recommend increasing your original sample volume prior to qEV isolation. This may require switching to a larger column. However, common EV markers in individual fractions are routinely identified with Western blot from 500 µL of plasma, so this is unlikely to be a problem for biofluids.
Now that you have your lysed samples and have decided how much to load per well, it is time to add sample buffer. This will allow your sample to be heavy enough that it won’t just flow upwards out of the well. In addition, it can also be used to reduce the proteins in your samples by adding a reducing agent. This will aid in denaturing your proteins by helping to break disulphide bonds. At this stage, your sample and sample buffer mix should also be heated to denature the proteins and allow them to move more freely through the gel.
You can use β-mercaptoethanol as a reducing agent. EV researchers at Izon have often madesample buffer in 5x strength to reduce the volume we have to add to samples, thereby reducing unnecessary dilution. For heating, 95oC for 5 minutes is usually sufficient.
Once boiled and reduced/not reduced, your samples are ready to be loaded onto the gel.
Whether to reduce or not reduce your proteins is a big decision. Some antibodies will only detect non-reduced forms of the protein, whereas some others will only detect the reduced form. You may not know this until you try it out. Generally, cytosolic proteins may require reducing conditions whereas membrane spanning proteins like tetraspanins may not.
However, this is definitely not a hard and fast rule. If you are unsure whether reducing or non-reducing conditions are required for your protein/antibody pairing then it is best to do a test blot with a non-precious sample tested in both reducing and non-reducing conditions.
This bit will very much depend on the size of the protein you want to look for and your Western blot setup, so there is very little we can advise on here.
If you are running individual fractions, you need to decide whether you are running an equal volume or an equal mass of protein for each sample. Running an equal mass of protein will be very difficult for some fractions where the protein concentration is very low.
There are a few controls that you can add to your gel. The first is a raw, pre-column sample. This should give you an idea of the presence or absence of any protein of interest in your starting sample.
It is also important to know where your protein of interest is present. If not looking at individual fractions, this can be done by running the buffer volume, PCV and post-PCV samples, to check that your protein of interest is not, for example, associated with the free proteins.
One thing you can do before you look for an individual protein is to use a reversible membrane stain to confirm that there are proteins present in your blot.
You can use ponceau S solution as a reversible membrane stain. This is simply poured onto your membrane, left to agitate over the surface and then washed off, allowing you to see the bands of proteins for each well in your blot membrane. Eventually it will have all washed off, leaving your blot as pristine as it was before staining.
If you are running individual fractions with equal volumes then it is normal to see fainter/more intense bands and different bands in different fractions.
If individual protein bands (i.e., those detected by antibodies) or whole protein staining (i.e., with a reversible membrane stain) show proteins to be at the very top of the gel only, it is possible that your lysis was inefficient and whole EVs prevented protein migration into the gel. In this case, you may want to increase your RIPA concentration as suggested in Step 1.
(Also if using a precast gel, make sure you remove any packaging strips at the bottom of the gel which prevent the current running properly!)
After blocking you will probe the membrane with antibodies specific to your protein of interest, followed by a suitable secondary antibody. This should be done following the protocol of the antibody manufacturer, but may require some tweaking in your hands to achieve the best signal.
As a positive control, you can run a recombinant version of your protein of interest on your gel. This control may be smaller as it will likely lack modifications such as glycosylation.
If you do not detect the protein in your samples, try concentrating samples and trying again. If it is still absent, your EVs probably do not contain that marker – or at least not at a detectable concentration.
If you are seeing a smear of CD63 rather than a clear band, do not worry as this is normal. CD63 is highly glycosylated, meaning that there will be a wide range of sizes of this protein plus various amounts of sugar moieties. If you are having trouble detecting CD63, we recommend trying another tetraspanin such as CD9 or CD81 which tend to give more distinct bands.
Interpreting your results may be straightforward, but if they are different than you were expecting then this can be more troublesome. Before any scientific interpretations are made, it’s best to run the blot again with new samples to make sure that you see the same thing again.
If you do not see the same proteins when isolating EVs using a Gen 2 column as you did with a Legacy column, you can refer to this guide. Likely either you using an old protocol (we’ve changed some volumes to suit the new resin) or your protein of interest is non-EV associated and the newer, purer columns remove this protein from EV isolates.
Whilst the Legacy columns removed >97% of soluble protein, the Gen 2 columns remove ~99% of soluble protein. This seems a tiny difference, but this might make the difference between seeing a soluble protein in your EV fractions or PCV and not seeing that protein. In this case, the results from the Gen 2 columns will be more accurate due to increased purity. Check the protein-rich fractions to see if your protein is instead located there.
If you are still having trouble after trying all the tips and tricks in this article, get in touch with our support team.