And my first experiment was a failure …

At Keuka we perform immunohistochemistry in the Cell and Developmental Biology course that I teach each fall semester (taken by sophomore biology majors). We look for one protein at a time and we look for proteins that we know will be there. In our lab the technique is very successful in allowing students to practice the technique and get satisfying results. Some students perform a second experiment with thin slices (we call them sections) of mouse brains of three different types. One type has a mutation in the gene for a protein called dysbindin, which has been implicated in a susceptibility gene for schizophrenia in humans. The other two types of mice have either one normal + one mutant (called heterozygotes) or two normal dysbindin genes. Keuka students are asking the question does the dysbindin genotype have an effect on the presence or amount of another protein of interest. I originally learned this technique from Dr. Cynthia Shannon Weickert (who graduated from Keuka as a biology major in 1987) in whose lab I am currently working.

Here’s what happened when I tried to solve an experimental problem and to practice a more complicated version of an immunocytochemistry technique.

Thursday and Friday of last week I performed an experiment designed to demonstrate that there are estrogen receptors within the neurons found in the brain of monkeys. I only made one procedural mistake and that mistake shouldn’t have eliminated the results. But we will soon try, try again.

I am trying to use a technique called double immunofluorescence microscopy to visualize two proteins NeuN and ER. [Biologist’s use lots of abbreviations. I’ll try to make them clear when necessary]. Neu N is found only in the nucleus on neurons and is therefore a marker protein that identifies specific spots on a slide as the nucleus of a neuron. ERis estrogen receptor alpha; the protein to which the hormone estrogen binds in order to have its effect. We want to know if (well, confirm that) brain neurons have these proteins so we’d know that these cells have the ability to respond to the chemical signal estrogen. When biologists look for specific proteins in cells they often use immmuno techniques to find them. Immuno means that antibodies are used. You probably know that animals produce antibodies to fight off disease-causing organisms like bacteria or viruses. When we get a vaccination we are training our bodies to produce antibodies specific to a particular bacterium or virus. What you may not know is that scientists use animals as factories to produce antibodies to specific proteins so that they can be used in laboratories in many ways including the technique I was trying to practice last week. So, if we use an antibody to NeuN and another antibody to ER those antibodies will attach to the proteins in a thin slice of tissue. Great! Unfortunately we can’t see the antibodies. This is where the fluorescence part of the technique comes in. A fluorescent molecule emits colored light when it is exposed to light. So, if we buy an antibody to one protein with a red fluor covalently bonded to it and an antibody to another protein with a green fluor we can do a double antibody technique on the same tissue and see if the proteins of interest are colocalized. Got it? Good. But it’s more complicated than that.

Antibodies to specific proteins are very expensive. I estimate the amount I used for my one (failed) experiment cost at least $200. The technique that as usually used attaches another antibody to the primary antibody and that secondary antibody is the one with the fluor attached. The primary antibodies to NeuN and ER were made in mice and the secondary antibody was made in donkey and can recognize any mouse protein including the antibodies to NeuN and ERcx.

Y (secondary antibody to mice proteins – made in donkey
Y (primary antibody to NeuN – made in mouse)
NeuN protein

Ok. Do you see a possible problem here? Both my antibodies are made in mouse so if I tried to label them at the same time with the same green secondary, I couldn’t tell the difference! So ultimately we thought we might be able to do the detection sequentially rather than simultaneously if we fixed the tissue in between.

I’m going to stop the detailed description here but I think you can get the flavor of what I’m doing well enough to see that for a working scientist many experiments are intended to see if you can do what you want with the materials you have. Many preliminary experiments are needed to get to the point of doing the “money” experiment, the one that answers your question or confirms your hypothesis and goes into your published paper.

Wish me luck. I’m going to try a variation of last week’s experiment later today.