(49a) Organotypic Brain Slices As a Platform for Measuring Response to Biological Stimuli

Introduction: Chronic diseases are the leading cause of death and disability, affecting 6 in 10 adults in the United States. Developing novel therapeutics is necessary to combat chronic disease and reduce this burden, but the process to take a drug from discovery to marketplace is slow and costly. Failure to translate outcomes in animal models to success in human clinical trials is a major contributor to the time and cost associated with bringing drugs to market and is in part due to the inability of animal models to recapitulate the physiology of the disease in question. 2D in vitro cell culture models are simple and inexpensive, but lack contact between neighboring cells and native tissue architecture. While in vivo models are the most physiologically relevant, there are multiple disadvantages to this approach, including hefty animal care/housing expenses and inability to gain mechanistic insights, particularly when using larger, non-mouse models. Therefore, there is motivation to develop platforms that accurately recapitulate in vivo physiology, but are higher throughput and tailorable. Organotypic slice models have advantages over primary cell culture, as multicellular interactions are retained and bias in cell type is eliminated. Additionally, tissue slices retain 3D architecture and the slices mature in vitro in a similar manner to native conditions, allowing slices to be used long term in culture without sacrificing physiological relevance. A single animal can yield numerous healthy organotypic slices, which reduces the required number of animals required of in vivo animal testing. Relative to in vivo models, the extracellular environment can be better controlled, which allows for high throughput screening of various therapeutics or to study the biological response to any stimulus, including toxins and injury/disease models. Here, we demonstrate the use of organotypic brain slice cultures to probe cellular response to oxygen-glucose deprivation (OGD), an injury model for hypoxia and ischemia, and subsequent treatment of that injury with a therapeutic enzyme (superoxide dismutase).

Methods: Postnatal day (P) 7 Sprague-Dawley rats were sacrificed and the brain was promptly extracted. Hemispheres were separated and cut coronally into 300-micron slices. Slices were separated in ice cold dissection media (HBSS, 5 g/L D-glucose) and plated onto 0.4 um semi-porous membranes (3 slices per membrane) in a 6 well plate with fresh media (50% HBSS, 44% MEM, 5% horse serum, 2 mM GlutaMAX, 1% penicillin streptomycin) and cultured for 1 week. Media was replaced daily and supernatant was immediately stored at -80 °C. Media samples were analyzed by lactate dehydrogenase (LDH) assay to assess total cellular death as a function of culture time. Slices were stained with propidium iodide (PI), washed thoroughly, and fixed in 10% phosphate buffered formalin on days 1, 2, 3, 5, and 7. These slices were permeabilized with Triton X-100 (1%), counter stained with DAPI (1:10,000) and imaged by confocal microscopy to assess cellular death in the cortex, thalamus, and hippocampus. To analyze cell-cell interaction, bolistic transfection of neurons was used. Gene gun particles were prepared by adding 10 mg gold (1-micron size), 50 ul of 0.05M spermidine, 20 ug mCherry plasmid, and 50 uL of 1M CaCl₂ , incubated, and the pellet washed with 100% ethanol. Bullets were formed using Tefzel tubing and then dried with nitrogen, cut with the Helios® Tubing Cutter and stored at 4 °C with a desiccation pellet until use. Slices are bombarded using a Helios® Gene Gun. A helium pulse at 200 psig propels mCherry-coated, 1-micron gold particles at the tissue to transfect individual cells on the tissue. Following 2 days of incubation, slices are stained with Cd11b+ (1:500) for microglia. OGD was performed on 3 DIV (days in vitro) slices by: 1) removing brain slices from glucose-treated media; 2) depriving slices of oxygen by purging a sealed chamber containing the well plates with nitrogen, and 3) incubating at 37 °C for 0.5h to 3h. Following injury, non-transfected slices were cultured in normal culture media doped with superoxide dismutase (0.1 mg/slice) for 24 h. Culture media was aspirated and stored at -80 C for use with LDH assay. Slices were stained with PI and imaged to assess cellular death in the cortex, thalamus, and hippocampus. The response protein level was measured with ELISA. Following injury, mCherry-transfected slices were transferred into a perfusion chamber/confocal microscopy system and culture media is perfused at 20 ul/h using a 2-way syringe pump. Images containing neurons and microglia were acquired every 15 min for 6 h and the interactions between neurons and microglia were analyzed.

Results: We found that the optimal time to begin the OGD injury model for P7 rats was at 3 DIV, based on cell death data from LDH assay and PI staining. In the first few days of culture, slices undergo an inflammatory response due to the stress of slicing. As a result, cell death is high initially and the conditions do not make for an adequate experimental window to measure the response of a stimulus until initial death has subsided. By varying the total OGD time, we demonstrated the ability to precisely control a stimulus, measuring a dose-dependent response to the injury. Slices subjected to OGD without any treatment showed significantly more cellular death than to the healthy control group, as expected based on previous OGD studies with 2D cell culture and tissues. Additionally, OGD slices treated with superoxide dismutase (SOD) showed significantly less cellular death than the untreated OGD slices. This was also expected, as SOD is a therapeutic enzyme that works against reactive oxygen species (ROS). Although ROS are present in healthy tissue, OGD causes their presence to increase above the basal level, which leads to oxidative stress and downstream cellular damage. We demonstrated that organotypic brain slices can be cultured with high viability for several hours in a controlled perfusion environment compatible with multi-channel confocal imaging. The combination of time lapse imaging with biolistic transfection on organotypic brain slices to probe multicellular interactions has not previously been reported in the literature.

Conclusions: Organotypic brain slices can be paired with novel time-lapse imaging techniques to probe multicellular interactions with physiologically relevant cytoarchitecture and precise control of the extracellular environment. While we demonstrate the use of OGD with the organotypic brain slice platform, there are a myriad of other possible stimuli, including other injury and disease models, toxins, and therapeutics. Furthermore, the severity of the stimulus is controllable; thus, one can correlate the response of a disease or toxin to multiple levels of severity. With this approach, we can gather essential information about the underlying mechanism, using that data to drive our research direction toward a therapeutic that will be safe and effective. Finally, we show that organotypic brain slices can be used to screen therapeutics, and in this study, evaluated SOD. Because many brain slices are garnished from a single animal, this platform is compatible with the screening of multiple drugs at the same time without the need for excessive sacrifice and animal housing costs. We anticipate that whole hemisphere brain slices will be used to further the development and translation of therapeutics for combatting neurological disease.