Using Yeast To Discover New Drugs
Joshua Kritzer, PhD, joined the Department of Chemistry at Tufts University in 2009 following a postdoctoral fellowship at the Whitehead Institute for Biomedical Research. He approaches biomedical research from his broad training in chemical engineering (BE, The Cooper Union), biophysical chemistry (PhD, Yale University), and yeast genetics. The goal of the Kritzer lab is to discover novel bioactive molecules and use them to understand the molecular biology of human disease.
“I’m a chemist, and I was trained as a chemist. We synthesize molecules and test them, more or less in a drug development mentality, except that we use very nontraditional types of molecules to block nontraditional types of drug targets,” says Kritzer. The nontraditional molecules are peptides that can bind to nontraditional but very important protein targets, such as transcription factors, which are involved in many human diseases. “When a protein binds to DNA and turns a gene on and off, or when a protein binds another protein, they bind over a large surface. This is less like the traditional lock-and-key picture of molecule binding, and more like two pieces of Velcro coming together,” says Kritzer. Most traditional drugs are small molecules that bind to a target as a key enters a lock. These small-molecule drugs don’t work on large protein surfaces, which is why many desirable drug targets have been dismissed as “undruggable” by the pharmaceutical industry. As a PhD candidate at Yale University, Kritzer worked on synthesizing molecules that could bind to and inhibit these larger, Velcro-type interactions. Chemically synthesizing such molecules, however, was an expensive and time-consuming process, as was screening them for possible drug candidates.
“Then I did my postdoc in a cell biology lab that used yeast cells as a model for human disease,” says Kritzer. “Yeast is a great organism. It is very genetically tractable, meaning you can take genes out of it or put genes into it very easily. Its entire genome has been sequenced. It only has about 6,000 genes [compared to our 30,000], so most of the pathways in this organism are pretty well understood.” Kritzer combined his training in chemical synthesis with that in the genetic manipulation of yeast to develop an inexpensive and rapid process for synthesizing and screening new bioactive molecules. The hope is that these new molecules will unlock doors to previously hidden biology, and to new treatments for human diseases such as cancer and neurodegenerative disorders.
Kritzer’s strategy is to direct the model organism (yeast) to synthesize and screen the molecules, thereby identifying molecules of interest from among millions of possibilities. His research group does this by making large collections of DNA that encode instructions for making peptides of interest. “So we have a little plastic tube with DNA in it, with trillions of pieces of DNA, each of which has the instructions to make a different peptide, and we can put that in yeast very easily,” says Kritzer. Any yeast that takes up this DNA will start making a peptide that can be screened for function by plating the yeast on a selective medium. The selection could be life or death, in that only yeast colonies with certain DNA will grow, or it could be a color change or some other identifying reporter. “We can plate 5 million yeast and maybe we get 100 or 200 hits,” says Kritzer. Yeast colonies of interest are picked off the plate, the DNA is purified and sequenced, and the DNA that codes for interesting peptides is put into E. coli host bacteria for production and storage. The DNA can be used at any time to produce quantities of specific peptides that can then be characterized for effectiveness in altering a particular disease-associated protein interaction.
“Yeast is so easily manipulated that this is pretty much just ‘plug and play’ with any yeast model that’s out there,” says Kritzer. For example, he worked on a yeast model of Parkinson’s disease in his postdoc lab. “We were expressing in the yeast a human protein that is known to cause Parkinson’s disease, and that would kill the yeast.” Using the process described above, Kritzer incorporated billions of pieces of DNA into the yeast, plated them out and waited for colonies to grow. Only yeast cells with DNA that made a peptide that prevented Parkinson’s-like cell death were able to grow in the selective medium. “You pick those up, and you can sequence the DNA within the cells and tell what your molecule is right away,” says Kritzer. “So not only do the yeast make the molecules for us, they also tell us which ones are interesting, and all of that takes about a week.”
Another advantage of this method is that once researchers have identified a peptide of interest, they can genetically change the peptide easily by changing the nucleotide bases in the DNA, then redo the yeast screen to see the effects of the change. “Normally if you want to optimize a compound, you have to synthesize 60, 70, 80 versions of it, and test them one by one,” says Kritzer. “But genetic modification is very easy. It allows us to get a lot of information about how these peptides work very quickly.”
Kritzer and his research group are currently focusing on cyclic peptides, which are like natural peptides in that they are made up of amino acids, but are less prone to degradation and are typically more potent due to the cyclization. They are synthesizing and screening for cyclic peptides that modify many molecular interactions involved in cancer. Some of his current targets include the MYC, STAT3 and HSF1 transcription factors, which are being investigated by yeast two-hybrid screening, heat-shock protein 90, a chaperone protein that is required for proper folding of many of the major drivers of oncogenesis, and histone deacetylase proteins, which can activate or deactivate proteins.
“My most common collaborators are people who are studying a specific protein that is critical for some sort of disease or other state that they are investigating,” says Kritzer. “The best way to test the role of that protein is to have an inhibitor—but most inhibitors are hard to come by. This is a very rapid way of getting multiple milligrams of a weak inhibitor in a form that is chemically tractable, easy to use, very soluble, and won’t kill cells (because it's made of amino acids). This method is very attractive for target validation.” Kritzer is working to turn his lab into an efficient pipeline of peptide discovery, optimization, and maturation into more drug-like compounds. Future plans include introducing his modified peptides into mammalian and human cells.
For more information, please go to