Hi Sneha,

While using serum as a cheap positive control might sound like a good idea, I don’t think this would work all that well. Keep in mind that EVs make up only a minute portion of all serum proteins, which renders detecting them in crude serum very hard. If you were to run a protein extract of total serum, 90+ % of this protein sample would be albumin… I’ve never tried this, but I’d be surprised to see robust detection of even the most highly abundant EV markers in a total serum extract. There are, however, commercially available serum EV standards that can be used as a convenient positive control for WB.

In terms of lyzing vs. whole EVs, I think there is a lengthy discussion about this question somewhere on ResearchGate. If I remember correctly, some people just add loading buffer before running SDS-PAGE, but most use RIPA or similar buffers to disrupt EVs prior to protein quantification and electrophoresis.

Best,

Dominik

Hello Dominik Buschmann

Thank you very much for your suggestions. I was just curious and also gave a try last week and as you said there was no band detected for the whole serum when tested against TSG101. I am surprisingly also not seeing a positive banding for my exosome samples, despite having a good prep associated with this. I am trying to validate its size using DLS. Although the size on DLS seems convincing I am not seeing any positive data on WB. Could you suggest any possible explanation for this?

In my previous lab, I got positive results for TSG101 (Santa Cruz antibody), but now I am unable to get any band no matter how many modifications I perform on my samples (FYI TSG101 antibody now is from Thermofisher). I don't know if this could be an antibody issue since there is a positive banding on my whole cell lysates. Any guidance or suggestions would be hugely helpful!

Well, there are still many potential explanations for you not getting any bands for TSG101 – the most obvious one being that being that the protein is just not present in your lysates at any discernable concentration.

DLS just substantiates particles; this does not necessarily mean that they are EVs, let alone exosomes. Did you check any other EV markers on the same gel? It’s a good sign that the antibody produced a nice band for the positive control; this should rule out at least some AB issues.

A few things that come to mind:

- How much protein was loaded? Can you increase the amount?

- Was protein correctly transferred to membrane? Check Ponceau stain before blocking.

- Are any other markers detected on the membrane?

- Did you already increase AB concentration and/or incubation time?

Here’s a thread that goes into more detail on using WB to detect EV proteins:

https://www.researchgate.net/post/What_is_the_best_Western_blotting_protocol_for_characterization_of_urinary_exosomes

Anecdotally, we’ve observed the same issue with our serum EV preps for years: markers such as alix, tetraspanins and syntenin are nicely detected, but lysates are always negative for TSG101 (Abcam AB)…

Thank you very much for providing me with your detailed suggestions Dominik Buschmann . I am going to try blotting for CD81 this week to see what they look like. And to answer your questions:

a) How much protein was loaded? Can you increase the amount?

60 ug of protein was loaded this time and ponceau was also checked. It appeared like the proteins transferred but nothing showed up on the blot.

b) Are any other markers detected on the membrane?

I am yet to validate the same in the coming weeks.

c) Did you already increase AB concentration and/or incubation time?

I am already using 1:250 of the antibody and do an overnight incubation at 4 degrees like I usually do.

Thanks so much for providing me with the paper. I will go through it right away.

Oh wow, 60 ug of protein is A LOT for serum EVs. This actually makes me assume that your EVs were rather crude (e.g. isolated by precipitation, differential UC or a similar method) instead of a pure preparation?

The problem with “dirty” EV preps (as you’ll see in the paper) is that even though you load a decent amount of protein on the gel (e.g. 60 ug), the majority of this protein is not actually EV protein, but some kind of contaminant (serum albumin and lipoproteins are likely candidates in blood-derived EV preparations).

This non-EV protein thus “dilutes” the concentration of genuine EV markers, and makes it hard to detect them; particularly if said marker had a low expression to begin with.

I’d definitely recommend checking your membranes for albumin (easily visible in Ponceau staining) and/or lipoproteins. Should these be present in relevant amounts, try purifying your EVs (by e.g. SEC and/or density gradient) and re-check for marker proteins.

Good luck,

Dominik

Thank you very much, Dominik, for your suggestions. I am trying CD81 this week, I will let you know what it looks like. Also, I had exosome preparations that were performed with 0.2-micron filtration and then overnight centrifugation which still ended up showing no bands on western for TSG101. I had used a similar prep back in my Master's and they have worked very well with TSG101 (as I mentioned, it was a different antibody). Having said that, I am finding it really hard to troubleshoot as I am unable to find out where exactly things are messing up. I also tried performing an SEC, but the fractions were so dilute that it was very hard for me to get a decent protein concentration to load them on western blot.

Also, I do not see a visible pellet after overnight centrifugation or short-term centrifugation for exosomes isolated from serum/plasma. My fear is resuspending them in PBS and performing a second short spin would result in a huge loss of exosomes? If you have any ideas, please suggest.