Extracting Better Ubiquitin Data from Your Samples: Beyond the Cellular Skip

The ubiquitin-proteasome system was discovered at the start of the 1980s, and people have been studying it ever since. Initially, researchers thought that tagging a protein with ubiquitin was the cell’s signal for the protein to be scrapped via the proteasome. But more research has shown that, as with all biology, once you’re up close and personal this post-translational modification is much more complex than we first thought.

Ubiquitin can bind to a lysine on the target protein. Some ubiquitin linkages are signals for protein recycling, some are involved in cell-cycle regulation, some play a role in cell signalling, some are still not well understood. With mono-ubiquitination, multi-monoubiquitination, and poly-ubiquitin chains of different lengths and linkages it can be tricky to know where to start unpicking this system.

If you know your protein is ubiquitinated and you want to investigate that, or a great big smear down your western hints that your protein might be ubiquitinated and you want to find out for sure, then these 10 tips might help you to produce quality data and get to grips with this post-translational modification that quicker.

Stop Before You Even Start

Before you even extract your samples there are things to consider to make sure you don’t miss your ubiquitin signal or that you’re not reading the wrong ubiquitin signals. When it comes to sample preparation, this means inhibitors.

1. Protein ubiquitination is reversible and chains can be lost by the deubiquitinase enzymes lurking inside your cells. Adding deubiquitinase inhibitors to your cell lysis buffer will stop them chewing up the tags on your protein of interest before you can analyze them. You should include ethylenediaminetetraacetic acid (EDTA) or ethyleneglycoltetraacetic acid (EGTA) and N-ethylmaleimide (NEM). Although lots of standard lysis buffer recipes include NEM at concentrations of 5-10 mM, this isn’t enough for some ubiquitin chains. K63 linkages are particularly sensitive and need up to 10 times higher concentrations of NEM to properly preserve them in your sample.
2. Proteasome inhibitors are essential. Aside from K63 and M1 all ubiquitin chains can be targeted to the proteasome. You need to include proteasome inhibitors in your lysis buffer to stop the proteasome shredding your ubiquitinated protein. The most common proteasome inhibitor that works well is MG132, but be careful as use over long periods (12-24 hours) can mean that ubiquitin chains start to arise as part of the cell’s stress response and then you’re measuring something completely different.

On the Western (Blot) Front

With your freshly prepared samples in your hands, one of the most common techniques to try first is western blotting. Each ubiquitin molecule adds 8 kDa to the molecular weight—this can be up to as much as 400 kDa. However, deciphering the number of individual molecules attached and how these are linked can be tricky. A few considerations here before you plunge right in can mean clearer images that will help you work out precisely what is happening to your protein of interest.

1. If you are using single gradient gels for your SDS-PAGE, then 8 % gels teamed with a tris-glycine buffer are a good all-rounder. They give a good separation of all ubiquitin chains right up to the largest of over 20 ubiquitin units. If smaller chains and mono-ubiquitination are more your bag, then using 12 % gels gives better separation for the smaller chains—though, of course, at the expense of your resolution in the upper range.
2. Switching out your western blotting buffer system can also help to discriminate between different size ubiquitin chains. 3-(N-morpholino)propanesulfonic acid (MOPS) is ideal if you want to look at chains of more than 8 ubiquitin units. If your protein of interest is decorated with small ubiquitin chains, between 2 and 5 units, then you need 2-(N-morpholino)ethanesulfonic acid (MES).
3. Always try and use PVDF membranes for ubiquitin blotting as the signal strength from PVDF membranes tends to be higher than for nitrocellulose membranes. Also, a 0.2 µm pore size can help if you’re looking at smaller ubiquitin chains.
4. If you are looking at long ubiquitin chains, then optimizing your transfer step could make all the difference. A transfer at 30 V for 2.5 hours is ideal. Transferring faster than this can cause the ubiquitin chains to unfold, which means that any antibodies you use to look for specific ubiquitin chain linkages might not bind and your data is lost.

Artful Ubiquitin Antibody Analysis

We all know that antibodies are associated with lots of problems and picking the right one to use in your study is no mean feat. This is doubly tricky in the ubiquitin field, because antibodies to lots of the targets just don’t exist, so choice isn’t often a factor and you have to work with what is out there. But never fear, solutions are available.

1. A native or denatured ubiquitin? If your antibody was raised to a native ubiquitin protein, then you need to make sure you don’t denature your protein sample during the assay or you’ll just destroy your signal. On the other hand, ubiquitin is a small globular protein that is difficult to denature. If it isn’t denatured enough during your SDS-PAGE step, then antibodies to a denatured ubiquitin might not recognize it.
   If you are looking at a denatured protein, you can increase the signal strength of your western by giving your PVDF membrane a denaturing treatment before letting it near a primary antibody. This seems scary and it took a lot of convincing for me to try it, but it does work.
   1. Incubate the blot for 15-30 minutes in boiling water.
   2. Incubate for 30 minutes at 4°C in a solution of 20 mM Tris-HCl, pH 7.5, 5 mM ?-mercaptoethanol and 6 M guanidine-HCl.
   3. Autoclave.
2. Mono- or poly-ubiquitin? Most of the ubiquitin antibodies out there bind both mono- and poly-ubiquitin chains. Antibodies specific for poly-ubiquitin have been developed, but don’t seem to recognize the different linkages equally. One study found that the anti-Ub antibody from Dako doesn’t recognize M1-linkages as well as K48 and K63 linkages in poly-ubiquitin chains, while the anti-Ub antibody from Cell Signalling Technology hardly recognizes M1-linkages at all. Obviously, if the ubiquitin chain linkages on your protein are less recognized, then you have less signal. This means that you certainly can’t compare across proteins within a sample.
3. If you want to examine the specific poly-ubiquitin linkages, then antibodies have been developed to most of these. Anti-K6, K11, K33, K48, and K63 are all available to buy off the shelf and seem to get quite good reviews from those using them. Antibodies to M1, K27, and K29 aren’t out there at the moment, but maybe this is your opportunity to develop the next research champion of an antibody?
4. If you can’t find an antibody or the murky world of antibodies has betrayed you before, then you can identify your ubiquitinated protein of interest using specific ubiquitin binding domains. These can be immobilized for pull-down or used as probes in far

If you still can’t work out your protein’s ubiquitin signals and what they mean, then try not to be tempted to go the way of the UPS system and consign the lot to the bin.