Abstract: “Fungal infections occur when fungus invades the tissue, which can grow and affect the whole body if left untreated. The current antifungal drugs on the market often come with unwanted side effects, and drug resistance will always be a problem. This leads to the necessity for new pathways for inhibiting fungal infections. To this end, we are developing a library of new antifungal agents. An enzyme critical for life, methionine synthase, has a key difference between fungi and humans that can be exploited. An inhibitory molecule can be made to selectively target fungal methionine synthase based on this difference. Utilizing the modelling software Autodock, molecular modelling was done to develop theoretical molecules that target the fungal enzyme. Based on the theoretical modelling, a library of potential inhibitors was synthesized. These compounds were tested in an assay measuring the activity of the fungal enzyme in the presence of our compounds. To further evaluate the activity of each inhibitor, they are tested in a fungal growth assay which show zones of inhibition that prove our molecules are biologically active against fungi.”
East Central IL ACS Undergraduate Research Conference
Thank you for joining us for the 5th Annual ECI-ACS conference!
Are there known inhibitors of MetSyn? If so, how does the docking score of the known inhibitor compare to the proposed compounds?
Great question! There are currently no known inhibitors of MetSyn in the way that we are working with it. There are some studies where groups have identified MetSyn as a potential target for inhibition, but there are no proposed inhibitors in the literature. In terms of determining validity of the modelling, we have matched crystal structure data of the folate natural substrate in MetSyn with a modelled folate through the docking software. We docked an identical molecule of the folate natural substrate and overlaid it with the real X-ray crystal structure data and they matched right on top of each other. This is all a way of saying that the docking software should be accurate in finding the most stable docking conformations. Many of our proposed inhibitors actually end up overlaying in similar ways to the core structures of the folate and homocysteine natural substrates when the models are overlaid with X-ray crystallography data. Something looking ahead in our project could be getting an x-ray crystal structure of MetSyn with our inhibitors bound so we could see exactly how they are binding, and use those models with reference to the crystal structure data. Hopefully that answers your question and sheds some light on our modelling process!
Nice poster, talk, and amount of work! What were your overall yields for these compounds? Also were you able to come up with any type of structural features to explain the observed differences in activity (so why HPN were so active while KLM were relatively inactive)? Finally, since you did the docking studies, did the results from these parallel the observed activities? Thank you.
Those are all great questions! The yields were pretty varying between the compounds with a range between 50-85% for most of them. We’ve had to work with some of our compounds to increase yields in our reactions, but most of them would fall in that range. The structural differences between compounds is a tricky one, because we’ve gotten many structures to show activity. H is essentially the same compound as K and L, but without the carboxylic acid on the end. H is one of our best compounds, but K and L don’t work. However, both N and P do have the carboxylic acids and they do work. N and P are also made with pterins,
a different folate mimic, which could allow the carboxylic acid to fit in the homocysteine binding pocket easier than K or L could. There are subtle differences in the compounds that may change how they’re fitting and binding in the enzyme which could be why there is such a difference in activity, but there isn’t a certain answer right now. If we were able to get X-ray crystallography data for our inhibitors bound in the protein this would shed light on that problem. To answer the last question, in most of the cases that we modelled the inhibitors they ended up working very similar to the Kirby Bauer and fluorescence assay tests. Usually when the modelling shows that the inhibitor can fit and bind well, it ends up being the case with out other tests. For example, H almost perfectly overlaid the natural substrates in the open and closed protein conformation during the modelling, and it has proven to be our best inhibitor. Other inhibitors with good modelling have shown similar results as H. Thank you very much for all the questions and for taking the time to view our poster and presentation!