X26 - The Lung-Brain Inflammatory Nexus: Development of a microphysiological system for modeling respiratory disease effects on the brain

Saturday, October 11, 2025

10:00 AM - 11:00 AM PDT

Location: Exhibit Hall F, G & H

Presenting Author(s)

Zachary Greenwell

Undergraduate Research Assistant
University of Rochester, United States

Co-Author(s)

Gram S. Hepner

Graduate Student
University of Rochester, United States
BH

Brian Hassard

Graduate Student
University of Rochester, United States

Primary Investigator(s)

BM

Benjamin Miller

Dean's Professor of Dermatology
University of Rochester, United States

Co-Author(s)

SM

Sebastião Martin

Graduate Student
University of Rochester

Introduction: : Amidst recent efforts and legislation designed to reduce the need for animal testing in disease modeling and drug discovery, microphysiological systems (MPS) have emerged as promising tools capable of mimicking complex, biologically relevant human systems on a microscale. Previous work has yielded individual microphysiological systems modeling the alveolar-capillary (lung) barrier (µSiM-ACB) and the cerebro-vascular (brain) barrier (µSiM-CVB). We are designing a system that combines these two MPS platforms to monitor cytokine levels in real-time via integrated photonic biosensors, providing valuable information about the effect of lung tissue infection on the brain. Evidence suggests that some specific biomarkers indicative of viral lung infection include the pro-inflammatory mediators IL-6, IL-1β, TNFα, IFNγ, and CXCL10. Our goal is to produce a device that leverages specific binding via antigen-antibody interactions to quantify brain barrier responses to respiratory infections, providing mechanistic insight into the lung-brain nexus in the context of viral infection.

Materials and

Methods: : The functional sensing component of this microphysiological system is a photonic-integrated circuit (PIC) equipped with 18 ring resonators, with three ring resonators corresponding to each of six photonic waveguide channels. These sensors are functionalized by printing biomolecules onto the ring resonators using a piezoelectric microarrayer. For the purposes of initial sensor calibration experiments, sensors were functionalized with capture probes consisting of antibodies corresponding to the antigens of interest (anti-IL-6, anti-IL-1β, anti-TNFα, anti-IFNγ, and anti-CXCL10) as well as control probes to account for non-specific binding between proteins (anti-FITC, anti-Mouse IgG, and bovine serum albumin). Once functionalized, the sensors are incorporated into microfluidic devices, which allow for fluid flow over the sensor. The sensor is then aligned to a tunable laser, which performs 20-nm wavelength sweeps about every 6 seconds as fluid flows through the device, providing data about the resonant wavelengths of light present in each photonic channel. Each experiment begins by flowing endothelial cell media with 2% fetal bovine serum through the device followed by flow of an antigen cocktail consisting of the previously mentioned biomarkers in the same endothelial cell media. Photonic analysis software is used to quantify the specific antigen-antibody interactions occurring throughout each experiment.

Results, Conclusions, and Discussions:: Preliminary results show that, upon introduction of an antigen cocktail, both specific and non-specific binding occur in significant amounts. TNFα and IFNγ have consistently proven to be an effective capture probe, with resulting redshifts (increases in wavelength) of up to 80 pm. IL-6 and IL-1β have produced more muted responses, with redshifts often falling in the 30 to 40 pm range. CXCL10 has been the most problematic capture probe with little evidence that specific binding occurs; as such, it has been abandoned in our most recent experiments. All the tested control probes (anti-FITC, anti-Mouse IgG, and BSA) redshift by around 20 pm when exposed to the antigen cocktail, indicative of non-specific interactions between these controls and the antigens within the cocktail, though anti-Mouse IgG and BSA exhibit more consistency between experiments. The sensor calibration experiments performed thus far have provided strong evidence that our sensor is indeed working, allowing us to detect multiple antigens in solution at once. As we move forward, we aim to generate even more robust calibration curves of the antigens of interest, paving the way for sensor integration into cell-seeded brain chips. Upon validation of sensor functionality in the µSiM-CVB devices, our focus will be shifted toward implementation of the brain chip into the complete dual lung/brain system. Based on sensor data collected to this point, we anticipate strong antigen detection capabilities both in the individual µSiM-CVB device as well as in the integrated lung-brain inflammatory nexus, at which point our sensors will be used to reliably quantify the inflammatory response of the cerebro-vascular barrier in response to viral respiratory infection of the alveolar-capillary barrier.

Acknowledgements and/or References (Optional): :