I have now spent three weeks trying and failing to do a technique called a Western Blot. It’s a pretty neat procedure where you can take all the protein in a sample, separate out all the proteins by their size, then mark certain proteins. It’s a way of visualizing protein X on a piece of paper (well, membrane) that approximates how much of protein X is in your organism (or system).
First, here’s a picture of what results could look like (where protein X in this case is a protein called Bcl-xL*):
Orientation/Explanation of that image: the numbers on the left tell you the size of the protein (see, biggest on the top, smaller further down). kDa means kilodaltons, and it’s a unit of size (or mass). The lefthand column of the blot is a protein ladder. This is a known mixture of proteins of known size that orient you to the approximate size of your protein.
The top, larger box is the experimental blot: this is the one with your actual results. The bottom, smaller box is the control. If you look at the row of blobs on the top, you see that there’s a dark blob (=lots that protein Bcl-xL) in the second column, and very faint blobs (=very little Bcl-xL) in the third and fourth columns. If you look at the very bottom of the picture, there’s a line with – and +. This tells you whether or not this thing called siRNA was added (+ means it was added, – means it was not). In the simplest terms, siRNA blocks production of a specific protein (in this case Bcl-xL). So, as you would expect, when you add siRNA, you get very little of the protein! The bottom box is just the control to show that the lighter blobs weren’t just because the researchers accidentally added very little sample, based on the fact that a different protein (which should not be influenced by the siRNA) was present in the same amounts every time.
I’m trying to do basically the same thing: perform a western blot for my protein to confirm whether or not my morpholino (which, for our intents and purposes, does about the same thing siRNA does: prevent protein production) worked! This is along the vein I was talking about last week: it’s a control to convince you that any results I get are actually from my manipulations, and not a fluke or due to some other factor. Because if my western blot shows that my morpholino didn’t decrease protein levels, then I would have to conclude that 1) something is wrong with my morpholino or with how I’m using it and/or 2) any effects on the neurons are not due to the reduction of my protien, but are instead due to something else I’m doing.
Okay, so I’ve been trying (and failing) to get a western blot to work for weeks now. And when I say my western blots are failing, I don’t mean that I’m not seeing reduction of my protein or my blot isn’t showing the results I expect (those would mean either my morpholino likely isn’t working or my hypothesis is wrong). I mean that my blots aren’t showing anything at all. No bands at all except for the protein ladder.
This is a simplification of the procedure I’m doing:
1) Dissect out dorsal sections (the backs) of frog embryos.
2) Break open the cells and separate out the majority of the protein from other stuff in the cells.
3) Denature (unwind) the proteins. This gives them a negative charge. So I can now pull them through a gel by having a positive charge on one end and negative on the other, so they get attracted down the length of the gel. Because the gel resists the movement of the proteins, they end up getting separated by size (the biggest proteins move the slowest, and therefore stay at the top of the gel).
4) Transfer the proteins from the gel to a membrane (again using charge to move them over).
5) Staining certain proteins with antibodies (conceptually very similar to immunocytochemistry). I put on antibodies that bind specifically to the protein I want to visualize, and then add chemicals that allow me to visually see where on the membrane that protein is.
Now, at every single one of those steps, one (or ten) things could be going wrong. And western blots are notoriously obnoxious for this reason. There are a ton of steps (the real protocol [read: recipe] I’m following has 35 steps). And each of those steps can be tweaked slightly (maybe one person uses ethanol instead of methanol; there are several ways to break up your proteins; there are lots of options for which antibodies to use to probe for your protein). So it’s very difficult to pinpoint where my problem is. Which is extremely frustrating when the whole process takes about 10-15 hours each time, and I don’t know for sure if it worked until the very end. There are a couple of ways I can find out if everything has at least worked up to a certain point (there are chemicals that will stain all the protein, so I will be able to tell whether any protein is actually present on the membrane at all). So slowly, but surely, I can narrow down my problems. But it could take awhile.
And this is one reason research is expensive and takes a long time: oftentimes things are just very tricky or very complicated. Advancing technology makes things easier (western blots are much easier and faster than they were a few decades ago). But of course then you have the time and capability to do more, so we get more ambitious. At one time, a good western blot plus a handful of other experiments were worthy of a publication. Now they often show up in papers as footnote controls to bolster the other, much more major, experiments.
So I’ll probably be spending several more weeks getting this to work. I’m meeting with a grad student next week who does western blots all the time, and she should be able to help me further narrow down my problems, so I’m looking forward to that.
Also one somewhat unrelated note: as much as I’d love to explain exactly what I’m doing, I have to withhold some information. You’ve probably heard the term “publish or perish”. And a lot of publications will not allow data to be published if it has been published anywhere else (including online in a blog like this). In addition, there’s a risk that another researcher somewhere would stumble upon my blog and see what I’m doing and think, “Hey, that’s a good idea, I’ll do that really quickly and publish on it!” And another frequent requirement from publishers is that the research is novel, so if somebody already has a paper on that (even if they stole the idea from me), I wouldn’t be able to publish. And as much as I hate this culture of secrecy and publishing games, revealing too many specifics here could seriously hurt my research and my career. So unfortunately I don’t think I’ll be able to tell you about the specific protein I’m working on right now (I’ll just call it “my protein”). But the protocols I’m using are pretty universal, and that really gives you a good idea of what I’m actually physically doing in lab, even if some of the intellectual puzzle pieces are missing.
*for anyone curious, Bcl-xL is part of the Bcl-2 protein family (a protein family is just a categorization method by which we group proteins with similar structures and functions into families). This protein family is important in regulating a process called “programmed cell death” or “apoptosis”. This type of cell death is actually really important because it’s a method by which cells can determine when something is very wrong (major DNA damage, viral infection, etc) and destroy themselves in a way that prevents other cells from getting damaged, too. But the process is also a part of normal growth and development: it’s how you have unwebbed fingers and toes! When your hands are developing, they’re actually webbed, and eventually all the cells in between what will become your fingers go through programmed cell death!