USF Health Byrd Alzheimer's Institute

Antifolate resistance in Plasmodium falciparum dihydrofolate reductase (pfDHFR) remains a major challenge for malaria control. To understand how this enzyme maintains function under antifolate selection, we developed PfPATH, a computational framework that integrates mutational fitness measurements, evolutionary interactions, and structural information to map adaptive trajectories. PfPATH adaptive walks reveal that evolution is constrained to a narrow ridge in sequence space defined by a small subset of residues. Most resistance mutations are individually deleterious and become viable only when supported by stabilizing changes, restricting adaptation to a few high-resistance pathways. Structure network analyses show that these mutations do not disrupt the global enzyme architecture but instead reorganize internal communication to preserve catalytic function. Consistent with this, molecular dynamics and free-energy analyses indicate that resistant variants remain stable while sampling a broader ensemble of low-energy conformational substates. Together, these results reveal a narrow and predictable route to antifolate escape.

Ionizable lipid nanoparticles (iLNPs) have revolutionized Ribonucleic acid (RNA) therapeutics by enabling precise and efficient delivery of nucleic acids. However, their clinical translation remains challenged by batch-to-batch variability, complex lipid - RNA interactions, and stringent regulatory requirements. This review highlights how advanced microfluidic technologies address these issues by providing precise control over iLNP fabrication through engineered mixer geometries, optimized flow dynamics, and pH-dependent self-assembly. Comparative analyses of hydrodynamic flow focusing (HFF), and staggered herringbone mixers (SHM) demonstrate their distinct influence on particle size, polydispersity index (PDI), and encapsulation efficiency. Furthermore, the integration of design-of-experiments (DoE) methodologies, computational fluid dynamics (CFD) modeling, and machine learning (ML)-assisted optimization enables predictive formulation design and adaptive process control, enhancing reproducibility and scalability. Collectively, this review underscores microfluidics and ML as synergistic technologies that bridge laboratory innovation with Good Manufacturing Practice (GMP)-compliant, large-scale production paving the way for the next generation of intelligent, personalized RNA nanomedicines.

The ErbB4 gene is a schizophrenia (SCZ) risk gene that interacts with PSD-95 via its C-terminus, a connection disrupted in SCZ patients. To investigate the functional significance of this interaction, we generated a zygotic mutant mouse lacking the terminal valine “V” residue from the ErbB4 TVV motif. The homozygous (homo) mice exhibited disrupted ErbB4‒PSD-95 interactions and SCZ-relevant behavioral deficits, including impairments in motor function, sensory processing, and memory performance. Structural computational analysis further revealed that the mutation altered the structural conformation of the ErbB4 C-terminus, which affected its binding affinity for PSD-95. Mechanistically, the mutation led to up-regulated but less activation of ErbB4 and down-regulated but overactivation of PSD-95, possibly representing a failed compensatory response aiming to maintain the ErbB4-PSD-95 interaction. Additionally, homo mice presented NMDAR2A subunit specific hypofunction and reduced GAD67 expression. These findings highlight that the ErbB4–PSD–95 interaction is a critical molecular link in the synaptic dysfunction and behavioral abnormalities associated with SCZ.Supplementary InformationThe online version contains supplementary material available at 10.1186/s13041-025-01238-2.

The gene BIN1 is the second-largest genetic risk factor for late-onset Alzheimer’s disease (LOAD). It is expressed in neurons and glia in the brain as cell-type-specific and ubiquitous isoforms. BIN1 is an adaptor protein that regulates membrane dynamics in many cell types. Previously, we reported that BIN1 predominantly localizes to presynaptic terminals in neurons and regulates presynaptic vesicular release. However, the function of neuronal BIN1 in relation to LOAD is not yet fully understood. A significant gap in the field is the unbiased characterization of neuronal BIN1-interacting proteins and proximal neighbors. To address this gap and help define the functions of neuronal BIN1 in the brain, we employed TurboID-based proximity labeling to identify proteins biotinylated by the neuronal BIN1 isoform 1-TurboID fusion protein (BIN1iso1-TID) in cultured mouse neuroblastoma (N2a) cells in vitro and in adult mouse brain neurons in vivo. Label-free quantification-based proteomic analysis of the BIN1iso1-TID biotinylated proteins led to the discovery of 360 proteins in N2a cells and 897 proteins in mouse brain neurons, identified as BIN1iso1-associated (proximal) or interacting proteins. A total of 92 proteins were common in both datasets, indicating that these are high-confidence BIN1-interacting or proximity proteins. SynapticGO analysis of the mouse brain dataset revealed that BIN1iso1-TurboID labeled 159 synaptic proteins, with 60 corresponding to the synaptic vesicle cycle. Based on phosphorylation site analysis of the neuronal BIN1iso1-TID interactome and related kinase prediction, we selected and validated AAK1, CDK16, SYNJ1, PP2BA, and RANG through immunostaining and proximity ligation assays as members of the BIN1 interactome in the mouse brain. This study establishes a foundation for further investigations into the function of neuronal BIN1 by identifying several previously unknown proximal and potential interacting proteins of BIN1.

We present an idea of protein molecules that challenges the traditional view of proteins as simple molecular machines and suggests instead that they exhibit a basic form of “intelligence”. The idea stems from suggestions coming from Integrated Information Theory (IIT), network theory, and allostery to explore how proteins process information, adapt to their environment, and even show memory-like behaviors. We define protein intelligence using IIT and focus on how proteins integrate information (in terms of the parameter Φ coming from IIT) and balance their core (stable, ordered regions) and periphery (flexible, disordered regions). This balance allows proteins to remain stable while adapting to changes and operating in a critical state where order and disorder coexist. We summarize recent findings on conformational memory, allosteric regulation, protein intrinsic disorder, liquid-liquid phase separation, and critical transitions, and compare protein behavior to other complex systems like ecosystems and neural networks. While our perspective offers a unified framework to understand proteins, it also raises questions about applying intelligence concepts to molecular systems. We discuss how this understanding could advance protein engineering, drug design, and synthetic biology, while at the same time acknowledging the challenges of creating adaptive, “intelligent” proteins. This concept bridges the gap between mechanistic and systems-level views of proteins and offers a comprehensive understanding of their dynamic and adaptive nature. We have tried to redefine the traditionally metaphorical concept of “intelligence” in biochemistry as a measurable property while simultaneously establishing the material foundation of protein intelligence through the identification of fundamental elements such as memory and learning in molecular systems.