It seemed like such an elegant solution at the time; activate the transcription of a glucose transporter in an insulin independent fashion. The goal of current diabetic treatments is to lower blood glucose levels to a point where the glucose no-longer damages other tissues. Insulin moves glucose out of the blood by increasing the presence of a glucose transporter (GLUT4) on the surface of most cells. Drugs like Avandia work by modulating the GLUT4 gene directly, increasing its expression and thus getting more GLUT4 protein to the surface of the cell in a manner that is completely insulin independent. What a boon, we thought, for people with type 2 diabetes who still have some pancreatic function and are not yet taking insulin. Little did we know that those people would suffer from a host of other problems including an increased risk of heart attacks. Indeed, for those of you who follow the news, you may recall the dust up over Avandia a few months ago. The story is far from over, however. An important research paper has just been published in the Journal; Nature that might point the way to a whole new class of diabetes therapeutics which target the same molecule as Avandia.
To understand what the paper shows, we need first to understand how Avandia works. Avandia is a small molecule that binds to a much larger protein called PPARgamma. PPARgamma, in turn, is a member of a class of proteins which bind to specific DNA sequence and regulate whether the gene is transcribed or not. We call proteins like PPARgamma; transcription factors. Keep in mind that genes are the blueprints for proteins. They get copied into messenger RNA (transcribed) and then the messenger RNA is used as a template for structures called ribosomes to construct proteins (translated). The set of all transcription factors within a cell to a great extent define the identity of that cell as neuron, hepatocyte, pancreatic beta cell, or whatever. The mechanism by which transcription factors activate genes is complicated (of course) but one way to model the process is to imagine the enzyme; RNA polymerase. This is the protein that actually reads one strand of the DNA and creates messenger RNA which base-pairs perfectly with the template. Something has to create the initial binding site that will attract RNA polymerase to the beginning of the gene. Transcription factors read signals generated by the cell and when the right signal is given, they change their shape such that they can now bind to DNA in exactly the right place to create a new binding site near the beginning of the gene that attracts RNA polymerase. Genes do not simply have one binding site for one factor. That would be too easy. Instead, they have complex regions where many transcription factors will fight for binding sites. Some act to promote the formation of the huge protein complex that will eventually copy the gene while others act to inhibit the process. It’s sort of like Congress except that things actually get done. So, PPARgamma is one of many transcription factors within the cell. One way to integrate gene expression with whatever is happening at the moment is to require that the transcription factor bind to some small signaling molecule before it can become active and bind DNA. Avandia and all other drugs of this class mimic that small signaling molecule that normally binds to PPARgamma. The complex of Avandia plus PPARgamma can then bind to a DNA sequence very close to the GLUT4 gene and provide a strong motivation for that gene to be copied into RNA.
The problem with Avandia is that the PPARgamma binding sequence can be found on many genes. As a matter of fact, the main job of PPARgamma is not really involved with GLUT4; this is just a side job. Instead, its main job is to regulate growth and development of fat cells. Thus Avandia promotes the expression of many genes which have nothing to do with diabetes. Presumably, some of these newly made proteins contribute to the cardiac side effects and weight gain that plague this therapeutic.
The next concept that we need to introduce is that of the shape of the protein. Each gene codes for a unique string of amino acids which will collapse upon itself in a highly ordered process resulting in it folding into a distinctive shape. Each protein has a very specific shape and it is that shape that will determine what it will do. The reason that proteins fold has to do with the charges on atoms within each amino acid. Positive charges on some atoms will attract the negative charges on other atoms. We call these sort of attractive forces; hydrogen bonds. These hydrogen bonds are often formed between amino acids which are far apart when the string is unfolded. When the protein is properly folded there may be hundreds of different hydrogen bonds stabilizing its shape. Some proteins interact with each other to build large machines and the decision as to whether to interact or not is dependent on shape. Some proteins function as enzymes to catalyze a chemical reaction. Again, the activity of an enzyme is completely dependent on its shape. Some proteins bind to DNA (such as PPARgamma) and, once again, the ability of the protein to bind to DNA is dependent on its shape. One can think of this like pieces of a puzzle fitting together. Proteins, like puzzle pieces, must have complementary shapes to interact.
There are many ways in which cells can regulate the shape of different proteins. Remember that PPARgamma binds to a signaling molecule. When it binds, the shape of PPARgamma changes subtly. Another very important way in which protein shape is regulated is through the chemical attachment of a phosphate group to a specific amino acid. The phosphate group carries a negative charge and by covalently attaching this big negative charge to the protein, other amino acids nearby have to shift or rotate to accommodate that charge. Voila. Shape change. Enzymes that put phosphates on proteins are called kinases. It is a testament to the power of the phosphate group that literally hundreds of different kinases have evolved to regulate virtually every process in our body. Whether you are considering cell division, muscle contraction, the pumping of ions in the kidney, or the zipping about of thoughts in the cerebral cortex; kinases are there – shaping the process by changing protein shapes. Kinases are major control points for the signaling networks within cells.
Now, at last we can consider the paper. PPARgamma is phosphorylated by a kinase at position 273 (i.e. the 273rd amino acid to be translated). The authors of this study were investigating small molecules which bound to PPARgamma and blocked the ability of the kinase (called Cdk5) to phosphorylate it. Apparently the addition of a phosphate group at this position is quite important for fat cells. Remember PPARgamma is primarily interested in helping progenitors to grow up to be fat cells. At any rate, the study concerned looking at the properties of small molecules (possible new drugs) that inhibited the ability of Cdk5 to phosphorylate PPARgamma. Importantly, they found one molecule had potent anti-diabetic activity. This molecule, SR1664, did not affect the heart or promote weight gain (at least in the animal model they were using). Importantly, SSR1664 does not promote the activation of GLUT4 transcription. To make things even more interesting, Avandia not only activated GLUT4 transcription but also inhibited Cdk5 phosphorylation of PPARgamma. So where does this leave us? In terms of mechanism we are pretty much confused. Only one of quite a few experimental compounds that inhibited Cdk5 actually had anti-diabetic activity so just inhibiting Cdk5 is not the answer. What we DO have is a prototype for a new class of drugs. I hope to be writing on clinical trial results in the next year or so for this or some similar compound. Avandia may be in trouble but its target is still ripe with possibilities.