Successful Co-IP's

Quite awhile ago, I think I posted something about Co-Immunoprecipitation experiements. I dropped them for about a year or so as other, more pressing, experiments appeared on the horizon. Several months ago, I finally got back to them. I decided to go at them a little differently though. I am convinced that p75 signaling is very context driven. We know that it can cause cell death as well as cell proliferation. At first we were just interested in what the intracellular portion of the receptor was doing but I felt it was important to use a full length receptor and not just the intracellular domain (ICD). I was able to procure a plasmid that expressed full length p75 and have been using that to do double transfections of HEK 293 cells. What is of most interest to us is to see if p75 is interacting with Merlin. Merlin is a tumor suppressor gene that is mutated in the tumors we study. The way I do this experiment is to grow the 293 cells, introduce both Merlin and p75 to the cells, and let them grow for about 24 hours. The 293 cells transcribe and translate the plasmids and presumably process them appropriately. Then, I lyse the cells and isolate the protein. The last step is the actually Co-IP. Each protein has been engineered to be produced with a small peptide tag that readily binds to antibodies. I expose the cell lysate to very small agarose beads to which one of the antibodies is bound. The antibodies bind and immobilize the protein they recognize and theoretically, any protein that is associated with the bound protein should also be secondarily immobilized. After gentle washing to eliminate any random binding, the antibody-antigen association is broken and the eluted off the beads. Finally, I need to visualize the association. This is done with a Western Blot. The eluted proteins are run on a polyacrylamide gel, transferred to nitrocellulose and then probed with an antibody that recognizes a protein that was presumably associated with the original bead-bound protein.


 There is a lot of optimization that has to occur with these experiments. You need to ensure that you have the proper ratio of cell lysate to agarose beads. Lysate-bead incubation time and temperature have to be optimized, The types and amounts of detergents you are using in the lysis buffer and the wash buffer also need to tested. pH is important. Sometimes increasing the pH will increase antibody-protein binding but you also have to keep in mind that temperature can have a dramatic effect on pH depending on the type of buffer you are using. These are a few of the variables that need to be addressed. After countless experiments to control for all of these variables and others, I finally got some results that I trust. The good thing is that I have been able to replicate my results several times with different cells.

Western on the right was the result of a 4h lysate-bead incubation and the one on the left was done for 2h.  Each blot was loaded identically.  Lane 1 (from the right, next to the red dots) is the cell lysate, lane 2 is the unbound protein, lane 3 is a sample of the one of the washes, and lane 4 is the eluted p75 protein.

Now the challenge is to do the experiment in reverse. I just tried that today with my best guess for the conditions but it didn't work. It looks like I will need to take a few steps back and reoptimize for this interaction. If I can get it to work in reverse, I will feel pretty good about saying that Merlin and p75 associate. The question then arises as to whether the interaction is direct p75-Merlin binding or if the association is mediated by other, small adapter proteins such as paxillin.  I will begin to attack this question by performing the Co-IP as above and then labeling the eluted proteins with biotin.  I will then run the eluate on a gel and instead of probing with a specific antibody, I will probe with Streptavidin which binds tightly to biotin.  Streptavidin can be visualized as above.  These data will show me if there are other protein that are binding with p75 and Merlin.  Finally, the next big step will be to determine what those proteins are by mass spectrometry.  These experiements have the potential to shed some real light on how p75 helps to regulate cell growth.

How long have I been doing this?

Every summer we get a new crop of students/employees. I spend the majority of each summer training, handholding, generally getting nothing done. Our lab is a popular lab with research residents and med students looking to get experience. Mostly because my boss is too nice to say no to anyone.

Training research residents is far different than training grad students/post docs. They are generally very driven and not used to failure. All the ones we have had in our lab have been very good and it has been a pleasure to get to know them. One of the things I have to remind them is that it is okay to make mistakes as long as they don't keep making the same mistake over and over. I was teaching a new research resident how to do a Western Blot and stressed a couple of points. I told him that I learned the importance of paying attention to each detail because I had made many mistakes when I let my mind wander or just didn't take the time to follow the protocol. I gave him a couple examples of some of my dumber mistakes and then we moved on. I, of course, was teaching him from the stand point of experience and the belief that I was beyond such mistakes. Well, I made another one today.

We are in the process of getting a grant ready for submission and are in need of some data. I have been tasked with getting some DNA constructs amplified, purified, and isolated in preparation for a FRET experiement on Friday. Everything was on track until this morning. I prep'd three DNA constructs that we needed to use by transforming E. coli with the DNA constructs (essentially co-opting the bacteria to mass produce my DNA). Once I got bacterial colonies on an agar plate, I picked individual colonies to grow in broth overnight. The growth was successful and I isolated the DNA this morning. I got all the way done with the experiment and then threw out one of the DNA samples by accident. Pure stupidity. It goes to prove that no matter how long you have been doing something, you can still make mistakes if you don't pay attention.

More science humor

ß-galatosidase is an enzyme that has been co-opted by researchers as a reporter gene for successful molecular ligations and senescence assays. We often shorten the name of the assay to ß-gals. A friend of mine who did salmonella research, used to do so many of them that even his wife knew what they were. As a matter of fact, he talked about them so much that his wife called herself his "α-gal". Rather clever, I thought, also funny.

Scientist do have a sense of humor

More often than not, scientist are accused of being a humorless bunch. Most of us, with the possible exception of fruit fly genetists, are indeed somewhat too serious for our own good. For example, we tend to name proteins after how much mass they have. P53, a pivotal tumor suppessor protein is named so because it is a protein (the p) with a mass of 53 kilodaltons (the 53). Fruit fly guys on the the other hand, think they are clever. They name genes things like frizzled, son of sevenless, wingless, hedgehog, and so forth. Just to prove the rest of us do have a sense of humor here is a table for your viewing pleasure. It may not mean too much to anyone who hasn't spent much time in the lab but I think it is spot on and drop dead funny.

ChIP

Finally. I finally have some preliminary results. After 3 months of chasing this down, I have 190 µl of DNA at a concentration of about 15 ng/µl. Whew! Now the hard work of going from an easy to grow cell line to finicky primary cells that we only get about once every two weeks or so. There are so many ways this can go wrong. The experiment takes about three days to go from living cells to purified DNA in a tube. One poorly aimed sneeze and I have to start over. Sometimes I am amazed that science ever moves forward at all.

What worries me about this experiment is that I will be successful in pulling down DNA that is bound to p75 but we won't be able to align it to the human genome. In other words, I will have done a lot of work for no results. A part of this worry stems from the fact that this is essentially an in Vivo test tube experiment. The cells I am currently using bear little resemblence to the real thing. Once a cell has been immortalized, it's genome is frequently messed up. This could mean that any results we obtain do not accurately reflect the interactions in a normal cell. We will be getting around this by using cells that we harvest directly from human patients. However, that only solves one issue. The other issue is the p75 DNA construct I am using. I have modified the original gene by cleaving off the part that normally is in the extracellular space (the ligand binding domain) and tagged it with two different DNA sequences. One tag is for identification purposes and the other is the NLS tag which forces p75 to translocate into the nucleus. This is a rather contrived system. We know that in a normal tumor cell (and I use that term loosely), p75 is a single pass, transmembrane receptor. Once it binds it's ligand, the intracellular part is cleaved and is translocated into the nucleus. My construct skips the ligand binding and normal intracellular trafficking. However, what if the normal trafficking mechanism is necessary for proper DNA binding? What if p75 is associating with another protein as it is being shuttled into the nucleus and it is that protein that is binding to the DNA and p75 is simply acting as a co-factor? Maybe p75 is not associating with another protein but what if it is being modified in some way while in the translocation process thereby allowing it to bind DNA with greater fidelity or greater promiscuity? If the translocation process is important in p75-DNA binding dynamics, I won't know that with my construct.

These are the things I think about as I am trying to fall asleep at night. At least they are entertaining and intellectually provoking.

ChIP update

Science moves at a snail's pace most times. I finally have the first part of this experiment nailed down (DNA isolation and cleavage). I was using a sonicator to break up the DNA but was getting poor results. I went to an enzymatic method called Micrococcal Nuclease digestion and have had some success. I am going to run through this once more tomorrow and attempt to clean the results up by combining the MNase digestion with sonication. Next week I will move on to infecting the cells with virus so that p75 will be expressed and then work on getting p75 binding to beads for the immunoprecipitation part of the experiment. Once I have that worked out, I will change cell types and make whatever small changes I need to make before really hitting this experiment for results.

New experiment

I am still working on the experiment I wrote about in the last post but I am now adding on another experiment that takes the next step. The first experiment (from the last post) will confirm that my protein of interest, or rather a portion of my protein of interest, is exerting intranuclear control. The protein I am looking at is called p75 which is a single pass transmembrane receptor. This means that a portion of the protein sticks out into the environment, a small portion passes through the cell membrane and the remaining portion (the intracellular domain or ICD) sticks into the cell. Now, unlike fruit fly scientists who, in attempt to be funny, name their genes things like sonic hedgehog or SOS (son of sevenless), most of the rest of us are pretty boring. P75 is protein (the P) that has a mass of 75 kilodaltons. P75 is a receptor protein that functions to bind neurotrophins such as Nerve Growth Factor (NGF). NGF is call a ligand. The whole purpose of a ligand binding to a receptor is to cause a change in gene expression so that a cell can respond to it's surroundings. This process of intracellular communication or cell signaling is what fascinates me. This process is very dynamic and can be relatively simple or incredibly complex. The process often depends on the environment and the context in which the signal is received by the cell. For example, the accepted dogma for p75 signaling is that when p75 binds to a ligand, an enzyme called gamma secretase clips off the ICD which then through some mechanism is transported to the nucleus where it causes the cell to undergo apoptosis (programmed cell death or more literally, cell suicide). However, this is not always the correct. We have shown that our tumor cells actually proliferate under treatment with certain p75 ligands. Also, recent research from another lab in NY has shown that the p75ICD binds to the promoter region of a gene called Cyclin E1. Cyclin E1 is the protein that jump starts the synthesis of DNA replication in a cell in order for that cell to undergo mitosis. In light of these data, we have decided to see to what else the p75ICD is binding once it translocates to the nucleus. We are doing this by a method called ChIP Seq. That stands for Chromatin Immunoprecipitation sequencing.

In short (the protocol I am using is 15 pages long, so yes, this will be short), I will be forcing p75 to go into the cell nucleus where it will presumably bind to whatever sequence of DNA it binds to. I will then crosslink p75ICD to the DNA with formaldehyde, do a nuclear isolation and then clip the DNA into lengths of about 200-500 base pairs with a sonicator. Once I have confirmed the length and amount of DNA with it's associated proteins, I will apply the DNA/protein to some magnetic beads that are coated with an antibody that recognizes only p75. Once p75 and it's associated DNA is bound, I will isolate the beads, wash away any extra, non-bound protein/DNA, finally elute the p75/DNA into a small tube. Once I have done this, I will reverse the crosslinking, degrade any residual protein and RNA and hopefully be left with nothing but strands of ~500 base pairs long segments of DNA that were bound to p75. The trick now is to sequence the DNA that I have obtained. To do this, I need a primers which are small stretches of DNA of a known sequence. These primers will bind to my DNA and enable me to enhance and then sequence the DNA. The problem is that I don't know the sequence to begin with so how can I design primers? I will do that by adding a stretch of DNA to each 500 base pair piece of DNA I isolated. Once I have done that, I can enhance each strand of DNA by PCR and then have the amplified DNA sequenced. The results will hopefully show a number of genes that were bound to the p75ICD in Vivo.

I began the first set of experiments on Monday and am having problems getting enough DNA to move on to the immunoprecipitation step. I'll keep you updated.