RESEARCH HIGHLIGHTS

Our research highlights serve as a collection of feature articles detailing recent scientific achievements on GCS HPC resources. 

Computational Scientists Open New Front on Antibiotic-Resistant Bacteria
Research Highlight –

Researchers at Heinrich Heine University and Forschungszentrum Jülich are using high-performance computing in tandem with experimental research to better understand a common, antibiotic-resistant infection prevalent in healthcare settings.

While drug development has helped humans overcome countless illnesses over the decades, certain viruses and bacteria have remained stubborn adversaries. In hospitals and other healthcare facilities, doctors and nurses must vigorously protect immunocompromised patients from several antibiotic-resistant infections. One of them, Pseudomonas aeruginosa, is a bacteria well-suited for spreading so-called nosocomial infections, or infections that most readily spread in healthcare environments among patients with weakened immune systems. 

P. aeruginosa is one of the candidates that the World Health Organization considers among the more threatening bacteria,” said Prof. Holger Gohlke, who heads Heinrich Heine University’s (HHU’s) Computational Pharmaceutical Chemistry and Molecular Bioinformatics group. “Our experimental collaborators research this bacterium in a Level-2 safety lab. As a healthy person, you can enter the lab with little risk, but for immunocompromised people, it can cause severe infections.” 

Gohlke and his experimental collaborator, Prof. Karl-Erich Jäger at HHU’s Institute of Molecular Enzyme Technology at Forschungszentrum Jülich, were funded by a German Research Foundation Collaborative Research Centre grant to focus on studying membrane dynamics in cellular systems. Their goal is to improve the drug development pipeline and better combat challenging infections like those brought about by P. aeruginosa. Gohlke and his research group use high-performance computing (HPC) resources at the Jülich Supercomputing Centre (JSC) to run molecular dynamics simulations aimed at better understanding one of the bacterium’s enzymes responsible for cell membrane damage. This research aims at improving our understanding of P. aeruginosa and, in the medium term, discovering new methods to bolster our pharmaceutical defenses against antibiotic-resistant bacteria. The team’s research made the cover of JACS Au.

Fatty acids provide clues to combating antibiotic resistance

One of the main characteristics of P. aeruginosa that makes it dangerous to human health lies in one class of enzymes. Specifically, so-called type-A phospholipases, or PLA1, are enzymes in P. aeruginosa that damage healthy cell membranes and disrupt the signaling networks cells use to combat infection. Gohlke and his collaborators identified a specific enzyme, PlaF, as one that can make the bacteria more dangerous for immunocompromised patients. 

In prior research, the team had identified how certain types of fatty acids—specifically, medium-chain free fatty acids (FFAs)—could blunt PlaF activity in cells, but did not fully understand the specific reasons for its success in inactivating PlaF. 

To better understand how FFAs can combat PlaF’s role in increasing P. aeruginosa’s virulence, Gohlke, Jäger, and their team used a combination of highly complex molecular dynamics simulations in concert with in vitro and in vivo experiments. With access to the JUWELS supercomputer at JSC, the researchers uncovered that FFAs not only indirectly influenced PlaFs by alternating its shape, but also by directly impacting the enzyme’s center, which might lessen its influence on nearby healthy cells. In both cases, the researchers opened a potential path for molecular biologists, chemists, and other scientists to develop new pharmaceuticals to combat P. aeruginosa infections. 

“The research community has known that fatty acids can have functions as signaling molecules,” Gohlke said. “This is one strong example of the role that they can play in regulating protein activity by influencing its configuration and very directly by serving as a blocking agent for active sites in an enzyme. This is another factor to consider in the future as we look at similar bacterial systems when we are investigating new ways that pharmaceuticals can fight off infection.” 

HPC resources improve drug development pipeline

Gohlke has a joint appointment at JSC’s parent organization, Forschungszentrum Jülich. He has also been a long-time user of JSC’s computational resources. For him, having a long-running, collaborative relationship with JSC staff has provided a positive boost for his research. 

“JSC has always provided our team excellent user support,” Gohlke said. “Oftentimes, PhD students or postdoctoral researchers may be accessing these systems for the first time, and we feel confident in their abilities to help us with troubleshooting our application on their system whenever we need it.” 

Gohlke’s team primarily uses the Amber code for its research, which was one of the first applications of its kind to embrace the use of GPUs in supercomputers. JSC hosted some of Europe’s first GPU-accelerated systems, and as the center moves from its current flagship machine, JUWELS, to its next-generation system, JUPITER, there will be even more GPU-accelerated computing at researchers’ disposal. 

He indicated that the team often uses small numbers of GPUs in its simulations, but that they need to run many iterations of their simulations for meaningful insights. JUPITER will have more GPU nodes available to researchers, meaning that Gohlke’s team and other researchers will have more nodes available for their research. 

With its increased access to GPUs, the team is also exploring ways that machine learning methods might be able to support its drug discovery research more generally. Gohlke indicated that the team was already looking into ways to use machine learning to more effectively predict enzyme function and better judge how certain mutations could either help or hinder particular treatment methods. 

Since the JACS Au paper came out, Gohlke’s team has continued to leverage HPC resources to seek out small molecules that could ultimately impact PlaF specifically and PLA1 enzymes generally. 

“We have identified compounds that show promise for positively impacting these systems,” Gohlke said. “I would not claim that we have new antibiotics out of this research, but we do already have promising compounds that can act as a template for further development. 

-Eric Gedenk

Related Publication: Gentile, R.et al (2024). “Molecular Mechanisms Underlying Medium-Chain Free Fatty Acid-Regulated Activity of the Phospholipase PlaF from Pseudomonas aeruginosa”JACS Au. DOI: https://doi.org/10.1021/jacsau.3c00725 

Funding for JUWELS was provided by the Ministry of Culture and Research of the State of North Rhine-Westphalia and the German Federal Ministry of Education and Research through the Gauss Centre for Supercomputing (GCS).

This article originally appeared in the summer 2025 issue of InSiDE magazine

Tags: Heinrich Heine Universität Düsseldorf JSC bioengineering Health and Medicine