Skip to main content
2024 BMES Annual Meeting Main banner
Icon Legend
This session is not in your schedule.
This session is in your schedule. Click again to remove it.

Bioinformatics, Computational and Systems Biology

Bioinformatics, Computational and Systems Biology - Poster Session E

Poster A4 - Targeting peptoids based antimicrobials for treating ear infections

Saturday, October 26, 2024
10:00 AM - 11:00 AM EST
Location: Exhibit Hall E, F & G

Introduction:  Hearing loss is an exceedingly common concern, affecting around 1.5 billion people, with numbers expected to increase considerably in the coming years. One of its most preventable causes is persistent infection by viruses, bacteria, or fungi, which can have serious adverse effects extending beyond hearing loss, such as hindering educational, communicational, and professional capacity and impeding overall physical and mental health. Due to the ubiquity of the issue, a variety of approaches such as intratympanic injections and systemic drug deliveries, have been developed to combat these infections. However, the location and size of the region of interest tend to make treatment difficult and unreliable, leading to the development of antibiotic resistance and over $750 billion in losses yearly from drug overuse. A relatively contemporary method, nanotherapeutics, has become increasingly prevalent due to their accurate targeting ability, which reduces dosages and side effects. One promising class is the peptoid, a biomimetic of antimicrobial peptides that is less likely to be degraded by the host’s immune system while also being easily synthesized, low cost, and highly configurable. Furthermore, previous literature has suggested that peptoids could be efficacious against ESKAPE pathogens, the six bacterial species most responsible for nosocomial infections. Particularly, bactericidal properties emerge when the peptoid is tetramerized into an ellipsoidal morphology. Consequently, this study proposes the development of peptoids capable of clustering specifically to neutralize these microbes inducing ear infections.

Materials and

Methods:  This study has been done in three steps: (1) building peptoid structures in atomistic representation, (2) evaluating hydropathy of the peptoid structures and generating training data for the machine learning algorithm, and finally, (3) identifying potential antimicrobial candidates. In step 1, we built seven unit long peptoid structures using monomers with three types of sidechains--charged, aromatic, and brominated aromatic. We generated all combinations of a 7-mer peptoid composed of three monomer types, leading to 2187 peptoid sequences. Rather than performing molecular dynamics simulation for all the sequences, which was computationally challenging, we performed hydropathy calculations, which are computationally affordable and can be used to predict the self-assembly properties of the sequences. In step 2, we performed the hydropathy calculations using the PARCH scale approach, which assigns each monomer of a peptoid a value between 0 and 10, where 0 means hydrophobic and 10 is hydrophilic. Molecular dynamics simulations were also performed for 400 peptoid sequences to train an ML data set on the correlation of parch values and the tendency of peptoids to self-assemble in small or big clusters. In step 3, we will employ machine learning algorithms to identify the top 10 peptoid candidates that self-assemble into small clusters of trimers and tetramers. These peptoid sequences will then be synthesized by Barron Lab, an experimental collaborator at Stanford University, and tested for antimicrobial activity in vitro and in vivo.

Results, Conclusions, and Discussions: Results
The hydropathy analysis of the 2187 peptoid sequences revealed a strong dependence on the net charge on the peptoid. For sequences bearing a high positive charge +7, +6, or +5, the peptoids had high parch values indicating hydrophilicity. Such peptoids are expected not to self-assemble. Conversely, in the sequences bearing a no or low charge, i.e., 0, +1, and +2, the peptoids had low parch values, indicating hydrophobicity and a tendency to self-assemble into large clusters, which make peptoids ineffective as antimicrobials. Through the hydropath calculations, we have narrowed our targets to ~210 sequences, which is ~10% of our initial 2187 set of peptoids. We are now investigating the effect of the location of the charges along the 7-mer peptoid sequence and using machine learning to train the locations and charge data to identify the top 10 peptoid candidates for in vitro and in vivo investigation in our collaborator’s lab.

Discussion
This study has focused on exploring the design space of peptoid-based antimicrobials using a systematic machine-learning approach that considers the chemical and structural aspects of monomers in determining the efficacy of the peptoid as an antimicrobial. The workflow developed in this research can affordably be extended to longer peptoids and other biomolecules.

Conclusion
We have narrowed the design space of 7-mer peptoids to 10% of our initial set by performing molecular-level simulations and hydropathy calculations. Our machine-learning approach will further reduce the potential candidates to the top 10 that will be tested for efficacy in the experimental labs to eliminate infection-causing bacteria.

Acknowledgements (Optional):