SDS-PAGE lets us separate proteins by size, but doesn’t tell us what the proteins we separated are. If you add on a second technique, called a western blot, you can use labeled antibody probes to test for the presence and quantity of proteins of interest. It won’t tell you what all the proteins in your sample are, just the ones you test for (kinda like a molecular game of Go Fish!). And you’ll have to have a lot of patience, but it’s a fundamental biochemistry technique, so let’s look at how it works. 

video added 2/23/22

The basic premise is you take a sample, often a cell lysate (spilled-out cell guts)), and you separate the proteins it contains by size in the SDS-PAGE gel step. Then you shift them out of the gel and onto a membrane (the blot step). And finally, (the worst part), you use labeled antibody probes to see if a protein you’re interested in is on that membrane. It’s the worst part because you have to wash and wash. And, another wash, oh my gosh!  

In addition to being impatient,  I have some bad memories associated with western blotting. So I’m really glad I don’t have to do them a lot. I don’t have to do them a lot because I’m typically studying super purified proteins that I express and specifically purify and *know* is my protein. But it’s a really important technique when you don’t know if your protein of interest is somewhere (like in a cell extract). Or if you want to see if that protein is expressed more or less under certain conditions, at different times, etc. So, like “in the cell” type things. Here’s how it works

I will give a brief overview and then provide some further details to you – but first lets clear up some possible confusion about what “w” you should be usin’! LOWERCASE! There’s a lot you should do for a western blot, but capitalizing the “w” you should not! I sometimes slip up and use a capital W, and you’ll see it written that way a lot, and almost surely won’t get scolded if you write it that way. But, technically it’s a lowercase “w” because it’s not someone’s name. Southern blots (where you’re looking for pieces of DNA) can get the big S cuz they’re named for a person (thanks Edwin Southern!) – the rest of the blotting “compass” was named just for puniness!) But enough of this funniness – nah, who am I kidding, let’s keep things fun… (and more on the Southern blot here: 

The western blot is an experiment used to answer questions like – How much protein X does a cell make under different conditions? The basic premise is – take some mix of proteins (such as the “lysate “you get when you break open (lyse) a cell) -> send them traveling vertically though a gel mesh to separate them by size -> trap them in place -> send them traveling horizontally out of the gel onto a membrane -> use antibodies (proteins that recognize specific other proteins) to probe the membrane to see if and how much of a specific protein is there.

Here’s a general layout of the experiment:

SDS-PAGE – separate proteins by size & trap them in a gel

TRANSFER/BLOT – move the trapped proteins to a membrane that likes to bind proteins – kinda like a “protein duct tape.” The membrane doesn’t feel sticky to your fingers but it does to proteins (but not after you’ve touched it with your fingers – so use gloves and tweezers!)

BLOCK – prevent nonspecific antibody binding by getting “generic” proteins to bind the parts of the membrane where your protein isn’t before the antibody has a chance to. Your protein’s only at small portions of the protein duct tape, so you need to coat the rest of that sticky membrane with something “generic” like bovine serum albumin (BSA) or milk (really! it’s chock full of proteins)

WASH, WASH, WASH – wash off non-bound protein

BIND PRIMARY ANTIBODY – Antibodies are little proteins that recognize & bind specifically to specific parts of other things, with the “other thing” being called an ANTIGEN and the “specific parts” being EPITOPES. In a western blot, the antigen is a protein you’re looking for and the epitope is a specific site on that protein. The primary antibody recognizes your specific protein and binds it, so you get your specificity for the thing you’re looking for – yay!  but it doesn’t have anything “seeable” about it – boo 😢

WASH, WASH, WASH – wash off non-bound primary antibody

BIND SECONDARY ANTIBODY – this recognizes your primary antibody and has some “detectable” quality like a fluorophore so it will emit light or an enzyme like horse radish peroxidase (HRP) that will convert an uncolored compound you add to a colored compound (chromogenic method) or light (chemiluminescence). It works because antibodies have a unique part that recognizes some antigen and a “generic adapter part” – but that generic adapter part’s only generic for the particular animal that made it (i.e. the adapter part’s slightly different in mice & rats). So you can use things like “goat anti-rat” which is a secondary antibody that uses its unique part to recognize the generic adapter part of a primary antibody made by a rat (and if that rat antibody is using its unique part to bind your protein…)

We use this 2-tiered strategy so that we don’t need “fancy” (detectable) antibodies for every single protein we want to look for. Plus, it increases sensitivity because multiple secondary antibodies can bind each bound primary one so it amplifies the signal. More on antibodies here: 

WASH, WASH, WASH – wash off non-bound secondary antibody

VISUALIZE – detect the detectable using the detectable’s detection method

Some more details: 

It starts off with a typical SDS-PAGE gel (Sodium DodecylSulfate PolyAcrylamide Electrophoresis) to separate proteins in a sample by size. In the case of SDS-PAGE, when I say “separate by size” I really mean “separate by length” – and by “length” I mean the number of amino acid letters. Proteins have multiple levels of structure (more here: ) starting with a long chain of amino acids that folds up based on the unique properties of the letters. You might remember me spending a really fun (at least for me) month talking about them… 

To separate proteins by length you first need to unfold them, which you do with the help of heat, a detergent (artificial soap) called SDS (Sodium Dodecyl Sulfate) and reducing agents like DTT (which break up disulfide bonds – strong bonds between the S’s of Cyses (one of the amino acids)). The heat gives the molecules enough energy to overcome the weak interactions that give the protein chain its overall fold, but not enough energy to break the chain itself. As it unfolds, the SDS slithers in and coats the protein. And the SDS is negatively-charged, so it makes the protein uniformly negatively-charged and prevents refolding because (thanks to that like charges repel thing) it basically makes the whole protein hate itself.

But the negative charge makes it love the positive charge you stick at the bottom of the gel! So you can use the negative charge to direct the protein to travel through a polyacrylamide gel mesh towards a positively-charged electrode. Longer proteins get tangled up more so travel more slowly. 

Unlike in the protein purification technique size exclusion chromatography (SEC)(aka gel filtration) where you use pumps and/or gravity to get the proteins to move, & let the proteins run out (and collect them), with SDS-PAGE you turn the power off before the proteins come out so you trap them in place. (you know when to stop it by watching the dye front and/or a pre-stained ladder).

Speaking of the ladder, it’s just a mix of proteins of known sizes to compare to. For “normal” gels we use an unstained ladder (you can’t see the bands until you stain the gel to see all the bands) cuz they’re cheaper but the western blot ones get the fancy one. The proteins in it are pre-stained so you can see them without having to wait to stain them. This will be important for making sure the blot really happened

When you turn off the power, you remove the electric drive helping your protein swim through the mesh, so you trap the proteins in place. The bigger (longer) ones won’t have gotten as far as the smaller (shorter) ones, so they’ll be trapped higher up.

Now you have a couple of options. If you also want to see all the proteins you can run 2 identical copies of the gel or if you only have a few samples you run the duplicate on the same gel then cut it in half and treat it at 2 little gels afterwards. You use 1 copy for the western blot, where you look for SPECIFIC proteins by using antibodies that BIND SPECIFICALLY to that protein and one copy for traditional all-protein staining with dyes like Coomassie Brilliant Blue (CBB) which stain ALL proteins.

We’ll return to that point later, but for now let’s focus on the one for the blot (transfer of proteins out of the gel and onto a membrane). Why do we need this step anyways? For one thing, as anyone who’s ever run an SDS-PAGE gel can tell you, they like to rip. Not good for lots of washes. Plus – When you turn off the power you trap the proteins in place. But they’re not completely “trapped” – the same gel-ness (liquid-filled mesh) that allows the protein to move through the “solid” gel means the protein can diffuse out (diffusion is a process where molecules, as a result of random movement, end up with a net movement from areas of high concentration to areas of low concentration until they’re equal). more here: 

The concentration of protein is high in the gel but low in the surroundings, so the protein can start to diffuse. This diffusion’s slower cuz you don’t have the electric pull helping it overcome friction & gel-tangled-upness but it’s also multi-directional – since their’s no “opposite end” to travel to, it’ll just wander randomly

So you put the gel next to a membrane then change where you put the charge – instead of putting it at the bottom you put it “on the side” to direct the proteins to move horizontally instead of vertically. This is a much shorter trip and you’re not trying to separate any proteins, just move your protein to a safer place to work with it. We’ll talk more on that “Work with it” part later – but first…

It’s not quite as simple as just moving the electrodes (external charge sources). There are different methods for blotting. Older methods include things like capillary transfer, heat-accelerated convectional transfer, vacuum blotting, & diffusion transfer (I thought western blots were slow – but these guys take 2 days just for the blotting part!) Today the dominant method is electroblotting (electrotransfer) which is faster and more efficient.

There are a few types of electroblotting (wet, semi-dry, & dry). I’m going to tell you about WET TRANSFER, which is the kind I’ve always used. Semi-dry & dry are faster, but the transfer’s often not as good

We use a module that fits into the same tank we use for the SDS-PAGE. It’s kinda confusing that in both cases the parts of that get plugged in are both at the top – but that’s just how they get access to the power – the charge they generate is either from the bottom (for SDS-PAGE) or the back (for blotting)

This module fits the sandwich we make, which consists of sponge/filter paper/gel/membrane/filter paper/sponge all wrapped up in a plastic cassette (with holes for buffer to be able to get through) that gets stuck into an “electric oven” – an electrode assembly (blotting module). I called it an oven because it’s somewhere you’re putting the sandwich, but it’s also fitting because it can get hot when it generates the electrical field needed to get the proteins to move. We do the blotting in a cold room. Another option is to use an ice pack

And by wet, I mean it – it’s important to keep everything submerged in transfer buffer when preparing the sandwich. It’s easy for this to get messy but you want to avoid mess-making because the transfer buffer contains methanol, which is hazardous & has to go into the hazardous waste disposal bottle. We usually do the prep in a big glass baking pan

When I first learned how to do a western blot, my teacher was very emphatic about not leaving air bubbles which interfere with the protein’s transfer – proteins can “swim” but they can’t fly! She had us use a glass rod (or a piece of a pipette) to roll out the bubbles. That message stuck with me and let’s just say that, in my first grad school lab experience, I got a little over-zealous with the rolling one time – which happened to be a figure I had to show at a lab meeting during a lab rotation…

You want to make sure that the membrane is on the side with the positive electrode (labeled red) or else you’ll move the proteins out of the gel and onto the filter paper – which is just there to help wick the liquid through, not to catch the protein! The protein-catching’s done by the membrane – these have really little holes (pores) that let through salts, etc, but not your proteins. So they act as a physical barrier preventing your protein from going too far (through the membrane and into the filter paper). But you don’t want to just block the proteins’ path – you need to get the proteins to actually stick to the membrane barrier and stay stuck through all the washes you’re going to subject it to. So you want a membrane that proteins find “sticky.” A couple common options are: nitrocellulose & polyvinylidene difluoride (PVDF). More here: 

Once you get the sandwich set up you turn the power on to generate an electric field to pull the protein out of the gel and onto the membrane (commonly 200mA for 1-2 hrs)

How do you know if the transfer worked? That fancy-dancey pre-stained ladder that we only bring out for special occasions. When you turn off the power and go to remove the membrane, you should see all of the ladder on the membrane and none still in the gel. We can’t see our proteins, but we assume that if the ladder transferred, they did too. Not all proteins transfer as efficiently, however, so you may want to reversibly stain the membrane to check you got good transfer first.

Transfer efficiency depends on a number of things- gel composition, field strength, etc and also protein size -> although they only have to travel through the gel the short direction this time (which is usually just 0.75 or 1.5mm) the same friction gel-tangling-upness applies so it will take bigger (longer) proteins longer to transfer

If all looks good you go from the blot to the block… unlike the antibodies which we choose for their extreme pickiness, when it comes to proteins, the membrane holding the proteins we send the antibodies out probing for is protein-sticky, but it isn’t picky, so you have to take great care to avoid false alarms! 

The membrane is kinda like “duct tape” that’s only sticky for proteins – but it’s sticky for ALL proteins. When you transfer the proteins from the gel to the membrane, your protein sticks to the membrane, but only at the spots where it’s present in the gel (the proteins should be moving straight horizontally because the electric field you generate is straight from back to front).

The antibodies you’re going to use are also proteins, so they think the membrane’s sticky too…so if they stuck it’d kinda defeat the whole point of the western – to detect specific proteins. You use a primary antibody that recognizes & binds to your protein, then a secondary antibody that binds to the primary antibody, amplifies the signal, and, because its conjugated (“permanently” attached) to something detectable, allows you to visualize where the primary antibody bound. 

So it’s important that the places where the primary antibody is bound are only where your protein is. If you add the primary antibody without blocking, it’ll still bind your protein, but it’ll also bind all the exposed sticky surface. So you need to first coat the sticky surface with some “boring” protein. By “boring” I mean it won’t interact with any of the antibodies and/or interfere with any detection methods, etc. Basically, you want something that’s there just to de-stickify the membrane so that the whole membrane is coated in proteins – but be so unreactive that it’s like nothing’s there. A couple common options are BSA (Bovine Serum Albumin) and nonfat milk (really!) Milk’s cheap but can have clumping and cross-reactivity problems. more here: 

I don’t run many western blots (thankfully! I hate these things…) because I’m working with recombinantly-expressed, super-purified protein. The “recombinantly-expressed” part refers to me “recombining” DNA – sticking the genetic instructions for making the protein I’m interested in studying into a new home – a vector plasmid – a circular piece of DNA that I can stick into insect cells or bacterial cells, etc. to have them make the protein for me and then I put it through lots of purification steps. more here: 

I care how much protein’s there because I want a good yield so I have lots of protein I can study – but I don’t need the western to tell me how much protein there is because “all” the protein is (hopefully) my protein by the time I’m through with purifying it. However, for cell biologists, they often want to just see whether and how much of a certain protein a cell makes under different conditions. 

The instructions for making proteins are written in DNA (as genes). But in order to make the corresponding protein, the cell first makes messenger RNA (mRNA) copies of those genes and those mRNA copies are used by the protein-making machinery (ribosomes). So, DNA -> mRNA -> protein.

Different techniques try to measure “gene expression” by looking at different steps along that pathway. For example, RNA seq and qPCR look at mRNA expression to see how many copies of the recipe are put into circulation. The western blot, however, looks at how many protein products are actually made from those recipes. There’s often a good correlation, but mRNA is subject to regulation, such as by miRNA, which can cause reduced production of the corresponding protein. more here: 

Also unlike RNA seq, which tells you everything that’s getting made, with a western blot you can only tell if something’s there if you know to look for it. This is why I hate when scientists show western blots without also showing the corresponding Coomassie (gel stained to showi *all* the proteins present) – especially when they’re claiming they’ve purified the protein first! So please, don’t be that scientist that makes me mad!

tech note: If you want to use a western blot to try to compare the amount of a protein present in different samples, you usually use some sort of “loading control” to show that you loaded the same amount of total protein in each lane (i.e. if one lane has a stronger band, that’s because the protein was expressed more in that sample, not just that you loaded more of the sample into the well). You can also normalize to that control – basically adjust your measurements in relationship to how much of the control there is in the sample. This works because the loading control is something like tubulin, which makes up a system of microtubules that help things move in an orderly fashion throughout cells, and is present in consistent amounts. In order to make use of your loading control you’ll have to play another round of Go Fish! – use a different labeled antibody, which is specific to that loading control and do the bind and wash routine again. 

more on other types of blots (e.g. Southern blots look for specific DNA & northern blots look for specific RNA) 

more on some topics mentioned (and others) #365DaysOfScience All (with topics listed) 👉

Leave a Reply

Your email address will not be published.