U.Va. computer models help in understanding C. difficile, antibiotic resistance - University of Virginia The Cavalier Daily

Each year, Clostridium difficile infects approximately half a million people in the U.S., most of whom are hospitalized patients. C. difficile is a toxin-producing bacterium able to cause symptoms such as severe diarrhea and colitis. Because of growing antibiotic resistance of C. difficile, Biomedical Engineering Prof. Jason Papin is using computer models to study and better understand the infection.

The rise in prevalence of C. difficile stems from the bacteria's ability to rebound after antibiotic treatment, contributing to it becoming the leading cause of hospital acquired infections. Antibiotics are a common treatment for patients with bacterial infections given their ability to slow the growth of the bacteria or kill it. Bacteria, however, can develop strategies to defend against the antibiotics, known as resistance mechanisms. The presence of these resistance mechanisms can result in antibiotic resistance in which antibiotics are ineffective and infections can become not treatable.

Matthew Jenior, a postdoctoral research fellow in Papin's lab, said the drugs used to treat the bacterium create a self-perpetuating cycle.

"Unfortunately, the drugs that we used to treat C. difficile are now growing more cases of resistance to it, so it leads to this cycle of recurrent infection where you get infected, you treat it, you stop antibiotics and it comes back again," Jenior said.

As a result of the C. difficile's growing antibiotic resistance, the Papin lab uses computers to model systems that occur in bacteria to better understand specific cellular metabolic pathways, or series of interactions between genes and their products that form and break down molecules for essential cellular processes. 

"The focus of the lab is [on] building computer models of networks of proteins and metabolites inside of cells," Papin said. 

These computer models are primarily C. difficile specific, but the lab's past and current research of similar bacteria and other pathogens, or disease-causing agents, has assisted in the research about C. difficile. 

For example, the Papin Lab compares the hypervirulent to virulent forms of C. difficile. Hypervirulent strains of bacteria, viruses or fungi are immensely pathogenic and capable of producing severe illness as compared to regular virulent strains that produce the typical illness. Jenior describes comparing both forms as taking what they have learned about the virulent strains and seeing if that applies to the hypervirulent strains and whether they are associated with worse patient outcomes.  

Papin's lab originally investigated C. difficile bacteria in isolation, but began including its interactions and behavior using advanced computer models within a host about four years ago when Jenior joined the team.

These advanced computer models are called GENREs, or genome-scale metabolic network reconstructions. To develop the computer models, Papin said complex mathematical equations are programmed into the computer. 

In sum, the mathematical equations represent different kinds of chemical reactions that could occur, which helps researchers predict which conditions they should test with C. difficile. By testing various substances with C. difficile and how it reacts to them, researchers in Papin's lab can discover which compounds are necessary for the growth and survival of the bacteria. 

A recent article by Papin and his lab noted that the expression of specific genes in the bacteria when responding to different metabolic conditions provided insights into C. difficile. 

"[The] different ways that C. difficile turns on its ability to be a pathogen [and cause infections] were linked in different ways to metabolism," Jenior wrote.   

Due to the correlation between the C. difficile's metabolism and its pathogenic properties, Papin's lab was able to test these conditions in the real world. Papin explains how the lab is able to hypothesize whether or not the bacteria will be able to survive under particular conditions. 

"You make predictions about if this bacteria can grow if I give it this kind of compound or sugar or not," Papin said.  

As a result of the computer's prediction making abilities, Jenior mentions how the lab was able to manipulate the environment of the bacteria to compare if what the computer predicted was accurate.  

The computer models require a great deal of computing power and, therefore, the lab benefits from collaborations with other research labs. The lab team collaborated with several colleagues, including Rita Tamayo at the University of North Carolina, Chapel Hill and Bill Petri Jr.  at the University's Division of Infectious Diseases and International Health, to confirm the accuracy of their computer models after they had been developed. 

Fourth-year Engineering student Mary Dickenson has worked on this C. difficile research, as well as projects on other bacteria and pathogens. 

"The C. diff. project is what I came in and started working on initially," Dickenson said. "So I did a lot of the modeling and troubleshooting … and not just either computational or experimental, there's this nice hybrid approach." 

Dickenson said her research experience was an all hands-on process that involves several key players. 

"[Papin's lab] is a very collaborative environment, so we all kind of work together to solve any problems that come up," Dickenson said.

From her research experience, Dickenson has been able to learn several new skills in the lab, including different computational techniques, which the C. difficile computer models introduce. 

These computer models and other dry lab tools, such as applied or computational mathematical analyses, can increase the efficiency of wet labs which involve the use and analysis of drugs, chemicals or other types of biological matter using various liquids. 

"[Computer models] can really accelerate testing in a wet lab environment, [by] narrowing down a lot of the possibilities that you have, [so you can have] a more targeted approach to doing wet lab testing," Dickenson said. 

Papin said there is potential for using computer models more generally in the field of medicine.

"The computer model can predict, hopefully, which genes are essential for the microbe, and so, if you know which genes are essential, you know the different targets that you would develop a drug to hit," Papin said. 

For example, if a drug can alter or destroy the necessary genes that C. difficile needs to reproduce, then that drug could be an effective treatment option for the infection. 

Papin's modeling is vital for not just understanding C. difficile reproduction, but also the bacterium's survival. The computer models were recently able to predict a specific cellular process that C. difficile heavily relies on for using carbon as a food source. 

"If we find the most important carbon sources among [the bacteria species in the healthy gut microbiome]," Jenior said. "Then we can start to look for the species present in healthy people that are the best at taking [carbon sources] away from C. diff. and then build out proto-targeted probiotics [with these good bacteria to] completely avoid antibiotics"— ultimately starving C. difficile of its needed carbon.

One 2021 study using probiotics to treat C. difficile found that the rate of hospital-acquired Clostridioides difficile infection improved by 39 percent at a Quebec hospital when 70 percent of antibiotic users took a three-strain Lactobacillus probiotic under a pharmacy-driven protocol.

Given the growing understanding of the capabilities of the computer models, Papin and his lab hope that doctors in the future will be able to treat patients more effectively based on new computer model-generated predictions. 

By investigating the biochemical processes of the C. difficile bacteria, the Papin Lab makes it possible to apply the understanding of the nature of the bacteria to new treatment options for C. difficile bacterial infection. 

Adblock test (Why?)