Predicting the Structure of the YscF Polymer in the Type III Secretion System of Yersinia pestis Using AlphaFold3

Author

Andrew Utinans, Nova Southeastern UniversityFollow

Defense Date

11-19-2025

Document Type

Thesis

Degree Type

Master of Science

Degree Name

Biological Sciences

First Advisor

Julie Torruellas Garcia, Ph.D.

Second Advisor

Arthur Sikora, Ph.D.

Third Advisor

Navi Gill, Ph.D.

Keywords

Type Three Secretion System, Yersinia pestis, Protein, Protein Modeling, AlphaFold, Artificial Intelligence

Abstract

The Type III Secretion System (T3SS) is a major virulence factor in Yersinia pestis, the causative agent of the bubonic and pneumonic plague. The T3SS enables the bacterium to inject effector proteins into host cells through a needle-like filament composed of repeating subunits of the protein YscF. Despite the lethality of the plague and the critical role YscF plays in pathogenesis, the polymerized structure of YscF has never been elucidated. This study utilized AlphaFold3 and high-performance computing resources to generate in silico models of the wildtype (WT) YscF and 2 mutants, constitutive secreting (CS) D28A and non-secreting (NS) D77A. Stoichiometries ranging from 6 to 24 subunits were modeled. Statistical analyses and qualitative comparisons to homologous structures indicated that the 24mer is the most physiologically relevant form of YscF. While AlphaFold3 produced consistent monomeric folds and confidence estimates, the absolute inter-chain confidence (ipTM) remained modest indicating limitations in resolving long range interactions. Mutant analyses revealed that while D28A closely resembled the WT YscF with a slightly wider pore, D77A did not polymerize into a physiologically relevant structure. Additionally, patterns in Arg73 making up the interior of the pore in the WT and CS suggest that it may play a key role in the overall structural integrity of YscF. These findings provide the first computational model of polymerized YscF, establish a framework for comparing YscF to other T3SS needle variants, and show the potential of AI driven structural biology methods for studying complex protein structures.

NSUWorks Citation

Andrew Utinans. 2025. Predicting the Structure of the YscF Polymer in the Type III Secretion System of Yersinia pestis Using AlphaFold3. Master's thesis. Nova Southeastern University. Retrieved from NSUWorks, . (223)
https://nsuworks.nova.edu/hcas_etd_all/223.