Grantee Research Project Results

Final Report: A Neurovascular Unit on Chip for Reducing Animals in Organophosphate Neurotoxicology

EPA Grant Number: R839504
Title: A Neurovascular Unit on Chip for Reducing Animals in Organophosphate Neurotoxicology
Investigators: Cliffel, David , May, Jody , Neely, M. Diana
Institution: Vanderbilt University , Vanderbilt University Medical Center
EPA Project Officer: Callan, Richard
Project Period: August 1, 2019 through July 31, 2022 (Extended to July 31, 2024)
Project Amount: $850,000
RFA: Advancing Actionable Alternatives to Vertebrate Animal Testing for Chemical Safety Assessment (2018) RFA Text |  Recipients Lists
Research Category: Chemical Safety for Sustainability

Objective:

In this project, we developed an organ on a chip platform entitled the NeuroVascular Unit on chip (NVU) for neuronal development and toxicity studies that faithfully replicates the blood-brain barrier. This NVU is multicell type-based with dynamic metabolic and signaling methods of assessing chemical and drug toxicity without the limitations of animal studies or standard cell death/morphology toxicology assays.  Our objective was to demonstrate that the NVU provides equal if not superior information than conventional animal toxicity assays for organophosphate compounds of interest like chlorpyrifos. We developed the microclinical analyzer and cellular signaling assays to measure in situ both the acute and chronic responses of the multiple types of cell lines that recreate neuronal tissues to exposures to potential drugs and other chemicals.  Our approach has had the advantage of providing quantitative information regarding a variety of cellular activities, including metabolism, membrane transport, protein translation, and hence, provides a comprehensive approach to absorption, distribution, metabolism, excretion (ADME) and toxicological (TOX) profiles.   The major project activities were the creation of a NVU platform that used only cells derived from human sources, improved bioanalytical sensors for targeting biomarkers in microfluidic outflow, chlorpyrifos exposure studies using astrocytes and neurons, and testing potential drug rescue studies to chlorpyrifos exposure. 

Summary/Accomplishments (Outputs/Outcomes):

Over the course of this project, we have achieved a number of important outcomes, addressed significant challenges, and published a number of resulting outputs.  We continue to work on additional manuscripts to detail the outcomes for the project.   The most important outcome was the goal of creating a human-based NVU for toxicological studies.  Our resulting development was almost entirely successful as we were able to create a wide variety of neuronal subtypes and astrocytes.  Unfortunately, the most important neuronal subtype, acetyl cholinergic neurons, were not able to be created from human stem cells reliably.  Electrical activity measurements confirmed the functionality of the neurons but they did not show any effect from chlorpyrifos exposure.  We successfully developed a multianalyte microclinical analyzer for measuring 8 analytes simultaneously including 4 neurotransmitters, and are working to complete on two journal manuscripts on this. We have expanded the analytes to include protein targets and published a paper addressing challenges overcome to do this.  Finally, we tried continuous monitoring of BBB membrane integrity and permeability experiments within the NVU using in situ transendothelial electrical resistance (TEER) measurements.  Unfortunately, these measurements were not helpful as the electrical resistance was too low to represent the real biological system.  Finally, we explored the role of chlorpyrifos in glutamate uptake experiments in astrocytes to address two different literature pathway hypotheses, and were able to use our methodology to confirm one pathway over the other.  We are working on final manuscript to describe these results.  A description of these findings for each project activity are detailed in the following paragraphs. 

The use of rat neurons was critiqued in the proposal review and kickoff meeting as a potential weakness and not in line with the overall goals of the EPA STAR program.  We addressed that weakness by adding Dr. Diana Neely and her ability to create human-based neurons from stem cells.  The Neely laboratory has already developed protocols and routinely differentiates glutamatergic cortical neurons, mesencephalic dopamine neurons and astrocytes from human stem cells. During the project, Dr. Neely was able to create and supply three neural sub-types from inducible pluripotent stem cells that differentiated into neurons for dopamine, GABA, and glutamate neurotransmitters.  Also, stem cells were differentiated into two different types of human astrocytes for chlorpyrifos exposure studies.  Unfortunately, the ability to differentiate stem cells into acetyl cholinergic neurons was not achieved despite multiple approaches over three years.  Literature methodology to achieve this differentiation into cholinergic neurons was not able to be reproduced in our labs.  This limited our chlorpyrifos exposure studies that were initially focused on acute exposure effects on cholinergic neurons.  As an alternative approach, we shifted our exposure studies to the effects of chlorpyrifos and its metabolite chlorpyrifos oxon on glutamate pathways in neurons and astrocytes.   

In addition to our efforts to increase cholinergic differentiation efficiency, we have assessed the functional property of our neuronal cultures measuring electrical activity using microelectrode arrays. Our observations demonstrate that the derived neurons are electrically active and therefore functional.   Chlorpyrifos exposure is well known for its effects on acetylcholine neural transmission. In addition to affecting acetylcholine neurotransmission, chlorpyrifos has also been shown to affect glutamatergic and dopaminergic neurons as well as astrocytes. Since our neuronal cultures are composed of different neuron types secreting glutamate and GABA in addition to limited acetylcholine, we assessed the effects of chlorpyrifos on the electrical activity of our neuron cultures.  These experiments found that even at relatively high concentrations chlorpyrifos did not significantly affect the electrical activity of these cultures at any time point of exposure and regardless of differentiation protocol. 

Temporal measurement of cellular response is vital to resolving the specific mechanisms and pathways that are affected by neurotoxins. We designed eight-channel screen-printed electrodes (SPE) which provide a low cost, real-time, and robust monitoring system for a maximum of eight analytes simultaneously. Each working electrode is separated into a single microfluidic channel using a custom built microfluidic chamber made out of polydimethylsiloxane. Polymeric and enzymatic films were deposited and cross-linked onto the working electrodes to make each electrode specific to one analyte. Electrochemical measurements were used to determine changes in analyte concentration reflecting the cellular response. Our previous work was focused on biosensors for cellular bioenergetics: glucose and oxygen consumption, and lactate and acid production.  In this project, we have developed electrochemical biosensors to selectively and quantitatively measure four neurotransmitters, including glutamate, acetylcholine, adenosine, and dopamine. This biosensor array provided a platform to obtain bioinformation of neurological functions particularly following treatment with chlorpyrifos and other neurological toxicants.  We also developed the ability to create biosensors for cytokine targets. 

One critical aspect of the neurovascular unit on a chip is the integrity of its blood-brain barrier.   The ex situ work from our lab suggested that endothelial barrier disruption in a chimeric NVU device comprising human astrocytes, human pericytes, hBMVECS, and rat neurons, occurred after a low dose chlorpyrifos exposure to the vascular side. We continued to work to improve on these ex situ measurements of BBB membrane integrity and permeability experiments by creating an NVU device composed entirely of human cells and by adding in situ transendothelial electrical resistance (TEER) measurements, in addition to the expanded neurotransmitter biosensor array.  TEER measurements would provide a direct measurement of the resistance of the BBB membrane, and thus can immediately identify any changes to the membrane permeability resulting from chlorpyrifos exposure.  As an associated area of exploration, we are also re-examined the foundational aspects of membrane-on-a-chip design. These included cell monolayer edge effects, device orientation, and coculture cell line interaction specificity.  Quantification in all these studies incorporated TEER measurements using a World Precision Instruments Epithelial Volt/Ohm (TEER) Meter 3 (EVOM3) as well as permeability assays titrating FITC-dextrans in multiple molecular weights.  Unfortunately, the TEER measurements in the organ-on-a-chip device resulted in relatively low electrical resistivities for the endothelial barrier, likely from edge effects and cellular cohesion limitations.  While these resistivities were similar to those reported by other researchers using TEER, they were at least an order of magnitude smaller than seen in human blood-brain barriers.  This reflected both a limitation in TEER as a method and in our design to maximize the ability to continuously monitor the integrity of the NVU BBB.   

Astrocytes as an important glial cell on the neuronal side of the barrier control multiple neuronal metabolisms and modulate brain bioenergetics and blood flow. Glutamate uptake and the glutamate-glutamine cycle are particularly contributed to our study of neurotransmitters. Glutamate plays a vital role as the excitatory neurotransmitters in the central nervous system; however, its residual is potentially neurotoxic residual in the extracellular space. In this year, we focused on the role of astrocytes in mediating the response to chlorpyrifos.  Two different theories on the role of astrocytes in chlorpyrifos present in the literature were evaluated with our approach.  Our results showed that the net glutamate uptake efficiency of two subtypes of human astrocytes were significantly reduced by chlorpyrifos at high concentrations compared to controls. Similarly, exposure to chlorpyrifos-oxon, the active metabolite of chlorpyrifos, also resulted in a significant reduction in net-astrocytic glutamate uptake. However, in lower concentrations glutamate uptake was not affected significantly by either chlorpyrifos or chlorpyrifos-oxon. 

To address limitations in current NVU instrumentation, we focused on addressing the stability and lifetimes of the instrumentation, enzymatic and ionic-selective sensors, TEER measurements and NVU hardware.  Newer gravity fed designs will likely replace PDMS based valves and pumps in future NVUs, as the continual wear on the PDMS leads to hardware lifetime issues.  The building of microfluidic prototypes by 3D printing is likely to replace our current NVU design process.  Biosensor design is another area in which 3D printing is likely to make significant changes to the current sensor arrays. 

Conclusions:

Overall, the technical effectiveness of NVU technologies applied to the study of organophosphate toxicity is limited by the difficulty in obtaining cholinergic neurons from human stem cells.  While there are scattered literature references to creating human cholinergic neurons, those methods were not successful in our experiments.  Additionally, the technical effectiveness of using transendothelial electrical resistance for continuously monitoring the integrity of the blood-brain barrier in a neurovascular unit on a chip was also demonstrated to be lacking.  The technical effectiveness of electrochemical biosensors for both small molecular targets and important cytokines in microfluidic devices was successfully demonstated.  These biosensors are helpful to understanding metabolic pathways involved in specific mechanisms of organophosphate toxicity.  The ability of NVU methodologies to study astrocytic metabolic pathways was demonstrated successfully to compare two existing hypotheses regarding the astrocyte glutamate uptake response to chlorpyrifos.  Given the importance of organophosphate exposure in human neurotoxicity, detailed metabolic pathway analysis is likely to continue to be very important in deciding the careful balance between which if any organophosphate pesticides to use and at what concentration unless suitable alternatives can be developed in the agricultural and home pest control industries.  In any case, the legacy of past organophosphate exposures in humans will continue to have health consequences for many years. 

In terms of economic feasibility of human-based neurovascular unit-on-a-chip methods for organophosphate neurotoxicology studies, the cost and time needed for these studies has demonstrated to be significantly higher than conventional animal models.  The cost for human stem cells and the compounds needed for their differentiation process is significantly higher than conventional animal work.  Additionally, the time required for the differentiation of the human stem cells is over 3 months that extends the project study duration past many comparable animal models.  Improved differentiation methods may be an avenue to explore this in the future, and brain organoids may also offer a faster and less costly differentiation process.  The potential for a much richer scientific dataset and an alternative to using animals in research using the organ-on-a-chip model must be weighed against its increased cost of supplies and time needed for the project. 

Journal Articles on this Report : 3 Displayed | Download in RIS Format

Publications Views

Other project views:	All 4 publications	4 publications in selected types	All 4 journal articles

Publications

Type	Citation	Project	Document Sources
Journal Article	Miller DR, McClain ES, Dodds JN, Balinski A, May JC, McLean JA, Cliffel DE. Chlorpyrifos disrupts acetylcholine metabolism across model blood-brain barrier. Frontiers in Bioengineering and Biotechnology 2021;9: 622175.	R839504 (2021) R839504 (2022) R839504 (Final)	Full-text: PMC Full Text - HTML Abstract: PubMed Abstract- HTML
Journal Article	Buckey G, Owens OE, Gabriel AW, Downing CM, Calhoun MC, Cliffel DE. Adsorption and electropolymerization of p-Aminophenol reduces reproducibility of electrochemical immunoassays. Molecules 2022; 27(18):6046.	R839504 (2022) R839504 (Final)	Full-text: MDPI - Full Text HTML Exit Abstract: PubMed - Abstract HTML
Journal Article	Buckey G, Owens OE, Richards HA, Cliffel DE. Electrochemical immunomagnetic assay for interleukin-6 detection in human plasma. Sensors & Diagnostics 2024;3(6):1039-1043.	R839504 (Final)	Full-text: Full Text HTML Abstract: Abstract HTML

Progress and Final Reports:

Original Abstract

2020 Progress Report

2021 Progress Report

2022 Progress Report

2023 Progress Report