Sunday, September 9, 2012

A little discussion about the ribosomal RNA genes
I mentioned in the last post about the rRNA genes (also called rDNA loci).  The reason why these are important starts with what is called the central dogma of molecular biology, which is that DNA -> RNA -> Protein.  That means that the information for everything in your cell is contained within your DNA.  That information is transcribed into RNA, which is chemically very similar to DNA.  It is also mobile.  It moves from the nucleus of your cells out into the cytoplasm where its information is translated into protein.  The picture below, found at http://bytesizebio.net, illustrates this graphically.  The way that the RNA is translated into protein is by something called the ribosome, http://en.wikipedia.org/wiki/Ribosome .  This ribosome is actually an RNA based machine, that is 3 RNA molecules do the actually synthesizing of the protein. These are called ribosomal RNAs, or rRNAs.  Because every protein in the cell needs the ribosome and these rRNAs to be made, they are obviously found in very high abundance. The way this is possible is because they are in a very high copy number.  Typically you have 2 copies of each gene, one from mom and one from dad.  But you have 200 copies of these genes (100 from mom and 100 from dad).  They are found in what we call head to tail repeats.  This means they come right in a row with the end of one going right up against the start of the next, sorta like floats in a parade.  The interesting thing about these genes is that in actively dividing cells, all 200 are turned "on".  But as cells develop into more mature types, some are turned "off".  Most commonly, half are turned off but I've also seen many cases where most are turned off.
Essentially, one can think of these as the engine that allows a cell to grow and divide.  This has made the regulation of these genes very important to the study of severe metastatic tumors and cancer treatment.  Because it seems that when a cell becomes cancerous, its less likely to become a life threatening and severely aggressive tumor unless its able to turn all these genes "on".
Its an interesting field of study to those of us, like myself, that are interested in basic molecular biological questions because of the problem that differentially regulated repeats pose to the study of the cell.  Sequence is very important to biology. The sequence of DNA is what differentiates one gene from another.  And from these sequences, we can understand how a gene is turned on and off.  Most commonly, whenever you have identical DNA sequences, they behave identically.  However, this is occasionally not true. In this instance, it is the crux of biologists like myself to determine how the cell can tell the difference between two identical sequences. This is a common theme in a study that has recently gained notoriety, called epigenetics. 


A chance to play catch-up

A little summary of the Giles Lab, and how it came to be. 
(also, a summary of some science sprinkled throughout and again at the bottom)
Its been a long time since the last entry into this blog. So I’ll start with happened since the last time. I was a postdoc at the NIH/NIDDK in Gary Felsenfeld’s lab for six years. It seemed like forever. I started there to work on RNAi and its effect on chromatin structure, and that was what I worked on. For anyone whoever reads this I'd like to not sacrifice the science, but I'd love for non-scientists to also have a chance to know what I'm talking about. So, I'll give a sentence or two whenever possible to help clarify things. I can also throw in a link to sites like wikipedia that can explain things too. So that end, RNAi is a process that earned its discover's the nobel prize in 2005(?) or near there. They showed that very small RNAs, things that were previously believed to be nothing more than degraded and useless, had a huge role in things as varied as prevent viral infection to regulating cancer. They had been implicated in effecting genome structure and function in a yeast, but nothing yet in humans. This is what I wanted to work on.
http://en.wikipedia.org/wiki/RNA_interference  

The path to having my own lab
We managed to get one manuscript out on the subject, in Nature Cell Biology. It wasn’t accepted until the fourth try. Its acceptance was greatly helped by a talk that I gave at a Keystone Symposia on epigenetics and chromatin. An editor from the journal was there and approached Gary to say that they would be interested in the work. Its no guarantee, of course, when they approach you like this. But its just a slight wink/nod in the right direction towards publication. We submitted it twice and finally got it in. I needed to follow that work up and I tried to do so by pursuing the genome-wide localization pattern of human Argonaute 2. This technique required the new process known as ChIP-seq, which required the ability to work with linux –based systems. Many of the bench molecular biologists at that time were “doing” chip-seq, but this occurred by virtue of a tight working relationship with a full-time, dedicated biostats colleague. An example of this was the lab next to ours, Dan Camerini’s group. His lab had 3 bioinformatics experts as postdocs. The rest of the lab merely had to do the ChIP and then hand it off to the bioinformatics folks and then the got their beautiful genome-wide binding data back. I had no such luxury and if I was going to stay in this game I knew I had to figure out a way to do these anlyses myself. With the great help of some of Dan Camerini’s postdocs, mainly Kevin Brick and Ivan Gregoritti, I was able to do many of the computational analysis needed from genome-wide, high-throughput sequencing analysis. I was able to put together a manuscript that consisted of the Argonaute 2 genome wide binding and a series of hypothesis driven analysis and comparisons with other publicly available databases. I was able to conclude that the Ago2 binding sites were enriched for repeat regions, and regions enriched for short RNA production and H3K9me3 levels. We sent this manuscript to Molecular Cell and it was reviewed but rejected. They wanted more experimental evidence and clearly thought a bioinformatics journal was more appropriate. This was annoying because less than a year previous, many labs were publishing entire manuscripts that were nothing more than an analysis of the genome wide distributions of a single histone modification. The main lab to do this was Keji Zhao, also at the NIH. But in science, the burden of proof and the bar for what is a good manuscript changes constantly. And although these types of papers were great in 2008, by mid 2009, they were no longer acceptable. I kept up working on this manuscript by re-doing the Ago2 ChIP-seq, including some immunoflourescne done in a collaboration by Gaelle Lefevre and myself, and also doing some small RNA-IP seq. We resubmitted but the reviewers always wanted more controls, and more information. However, this was data was good enough to impress two universities into giving me a job interview, University of Alabama in Huntsville, and University of Alabama at Birmingham. I was given an offer by both places and had no choice but to accept the job at UAB. I was also given a offer to visit the University of North Dakota, but I didn’t actually go visit b/c I accepted the offer at UAB first.  

The First Year of the Giles Lab
The job offer at UAB came around April of 2011 and I moved there around August. I did a couple extra experiments before anyone was in the lab and we resubmitted the manuscript with Gaelle and Gary but it was reviewed and rejected. I think brought in a graduate student, Blake Atwood, who developed a QPCR assay system to use for RT-QPCR and ChIP-QPCR analysis of the rDNA locus. I have gotten a little bit ahead of myself here. Our analysis had always suggested that Argonaute 2 was specifically localized to the repetitive regions of the human genome. However, these analyses leave out the ribosomal RNA genes, of which there are ~200 per human cell. This highly repetitive nature of these genes makes it difficult to analyze them. However, we were able to use a single consensus repeat unit of the locus and observe that the tags from the ChIP-seq experiment were highly enriched for the coding region of this gene. So, Blake enabled us to do ChIP-QPCR and RT-QPCR throughout the rDNA locus and we analyzed the effect of knocking down Ago2 on RNA levels throughout the locus. I also brought on a postdoc , named Mariana Saint Just Ribeiro. She was previously in Suming Huang’s lab at Florida. Suming was a postdoc along with me in Gary’s lab so I knew I could trust his recommendation. He made it clear that Mariana was perfectly capable of doing the work but that he didn’t have sufficient time to mentor her. So I took the leap and brought her into the lab. She quickly made an impact by demonstrating a strong effect on histone modifications when Ago2 is reduced. She joined the lab around the start of January 2012. Blake joined a couple months before that. With these new additions in personnel and data, we resubmitted the manuscript to Nature Genetics, and it was triaged. We immediately resubmitted it to Nature Cell Biology, which also triaged it. We immediately resubmitted it to Genes and Development and they sent it out. However, the reviews were mostly negative. Although the reviewers like the area of research and would like to see more work, they weren’t convinced. So now its September of 2012 and I have a new manuscript on this project. We have mostly just rearranged how the data is to be presented. We added some ChIP-QPCR of both endogenous and tagged Ago2 to the rDNA locus to confirm that it is actually bound there as suggested by CHIP-seq. We have also added a new analysis of the sRIP-seq, which demonstrates some information of the biogenesis of the small RNAs to which Ago2 are bound. We utilized a metabolic labeling technique to demonstrated the a loss of Ago2 actually does have an effect on the synthesis rate of the rRNA genes. It also illustrates a change in the processing rate when Ago2 is knocked down. We have added some very interesting data demonstrating that Ago2 binds to SETDB1, a histone lysine methyltransferases. We can show that this is found at the rDNA locus and that its localization depends on Ago2. So now the manuscript is in the hands of Gary and we hope to resubmit it soon, for the umpteenth time. Luckily, this isn’t the only thing that lab has going on. I have also been lucky to be in a collaboration with Hengbin Wang, also at UAB in the Dept. of Biochemistry &Molecular Genetics. His lab initiated a very large scale protein purification project to screen HeLa nuclear extract for its histone deubiquitination activity. Histone ubiquitiination is modification that we don’t know much about, so being able to find the enzymes that regulate it is a huge step. He ended the project with ChIP-seq and RNA-seq data that was done by his Chinese collaborators. However, their analysis didn’t make much sense, so he needed me to take a look at it. I did, and it was quickly apparent that it was going to be a ton of work. Hengbin quickly offered me the chance to be a coauthor. I jumped at the chance. My analysis showed that the loss of this new protein Usp49, caused ~10,000 introns to be retained in the mature mRNA. Furthermore, these introns have a uH2B containing nucleosome positioned at the 5’ splice site. These introns are also highly enriched for uH2B. We submitted this manuscript to cell, where it was triaged. Then we sent it to Science, where it was reviewed but rejected. The reviewers loved the biochemistry but thought that we hadn’t proved the splicing connection. We addressed their concerns to prove that a direct effect on splicing was occurring and then resubmitted it to Nature. They rejected it with the reviews being of the exact opposite nature. They seemed to love the splicing connection but hated the biochemistry. This only goes to show how broken our system of peer-review is! I’ll say that again, its broken! How could their be any truth to a process that could be so varied between 2-5 experts ranging between 2 of the premier journals in science? We are currently hoping to resubmit this to science. My lab currently has 3 people in it, after another very bright graduate student named Jessica Makofske joined the group. Her project is to purify specific chromatin fragments from the rDNA gene. This is no easy feat and is something that would revolutionize chromatin biology. To date, if you want to know which proteins are bound at a given site at a given time you have to have a-priori knowledge of the proteins existence and then check for it. This would most likely be done by ChIP-QPCR or a gel shift analysis. However, if one were able to biochemically purify certain chromatin fragments, mass-spec could be done on the entire locus and we would be able to know all the proteins that are present, as well as all the histone modifications. Our approach is the utlize a combination of nuclease accessibility, velocity sedimentation rates and differential restriction digestion to purify these fragments. This project has already yielded some interesting data.
A summary of some science:
ChIP.  Chromatin immunoprecipitation. First of all, what is chromatin. Well, most people know what DNA is.  But DNA doesn't look like that inside your cells.  Its actually wrapped up like a string wrapped around beads.  The beads are proteins called histones.  And when you have a group of histones combined with the DNA thats wrapped around it, its called a nucleosome.  When you string together a bunch of nucleosomes together and add in all the other non-histone proteins that bind to your DNA, you have chromatin. 
immunoprecipitation is a fancy way for doing the following, taking an antibody that recognizes a protein and attaching it to a bead.  The bead is the size of very very small pebble.  The antibodies are the same kind of antibodies that your cell makes to fend off infections. When this antibody is attached to the very small pebble, you can incubate the antibody:bead complex with your cells and then give it a spin.  The proteins recognized by the antibody:bead will pellet (or precipitate) to the bottom of the tube.  This technique is called immunoprecipitation  If the antibody is desinged to pull down chromatin, its called chromatin immunoprecipitation. 
ChIP-QPCR.  Is doing ChIP and then measuring how much of a given DNA sequence is in the tube using quantitative PCR.http://en.wikipedia.org/wiki/Real-time_polymerase_chain_reaction.  ChIP-seq is a ChIP experiment where instead of doing the PCR step to quantify how much of a given DNA sequence is associated with a given protein, you sequence all the DNA that comes down in the tube.  This will typically yield between 10 and 150 million small sequences and requires quite a large computer to analyze.