So, thank you for the opportunity to present our work today on FGF21 therapy in a neonatal hypoxic ischemic brain injury model. For my disclosures, I'm also a T32 postdoctoral fellow at the University of Pittsburgh. I have had some internal funding through the Safra Center for Resuscitation Research as well as external funding through the Laird, Olin, Zoll Foundations. And then my two main mentors for this project, Pat Kohanek and Travis Jackson, have a pending patent related to some of the work that I'm going to present. Therapeutic hypothermia is highly effective for neonates after hypoxic ischemic brain injury with clear benefits and subsequent neurodevelopmental outcomes. In children and adults after cardiac arrest, however, the benefits of therapeutic hypothermia versus targeted temperature management with strict fever prevention are less clear. Recently, studies have identified cold stress hormones, which are hormones that are upregulated by cold exposure, involved in thermogenesis, and may be neuroprotective. Cold stress growth factor 21, or FGF21, is one such cold stress hormone that's produced by healthy human adults when exposed to prolonged cold, as shown by this figure, showing serum levels of FGF21 in adults exposed to 24 and 19 degrees centigrade for over 12 hours. FGF21 acts via binding to the beta-clotho co-receptor on the cell surface to trigger downstream cellular pathways. Our lab recently explored the expression of the beta-clotho receptor in the human hippocampus and prefrontal cortex throughout the lifespan, and we found the beta-clotho is highly expressed in the hippocampus and prefrontal cortex of infants, with decreased expression throughout the lifespan and virtually no expression in the adult hippocampus and prefrontal cortex. Earlier work from our lab also showed that exogenous FGF21, when given in vitro to developmentally immature rat neurons, upregulates the neuroprotective cold shock protein RNA binding motif 3, or RBM3. Another group recently reported that exogenous FGF21 also improves outcomes after neonatal hypoxic ischemic injury. The aim of this study was to identify pathways involved in FGF21-mediated neuroprotection in the Rice-Vinucci model of neonatal hypoxic ischemic encephalopathy. We chose the Rice-Vinucci model based off our earlier receptor data, showing that beta-clotho is most highly expressed in the brains of infants and neonates, with decreased expression throughout the lifespan. For those of you who are not familiar with the Rice-Vinucci model, here is a schematic. Postnatal day 10 mice undergo anesthesia, and we dissect down to the right common carotid artery and ligate the artery. The pups then recover from surgery, are replaced back with their dams, and then undergo global hypoxia, placed in an 8% hypoxic chamber for 25 minutes. And then finally, they're monitored for two hours post-injury. This injury model causes injury through a combination of relative ischemia during the global hypoxic episode. Because we're interested in the influence of temperature on brain injury pathways, we monitor the temperatures of the pups throughout the surgery and during the injury and post-injury phase in order to maintain strict normothermia. Before we could give FGF21 as a therapy, we first sought to characterize our model. And what's unique about the Rice-Vinucci model is that in order to see brain injury, you need a combination of both relative ischemia and hypoxia. In the figures to the left, you can see that we looked at the hemispheres and hippocampal volume of brains in sham animals versus those who underwent the hypoxic ischemic injury. You can see in sham animals that the two hemispheres and hippocampuses are the same size. But in injured animals, it's only the ipsilateral side, the side that has had the right common carotid artery ligated, that shows significant hemispheric and hippocampal volume loss. To the right, you can see representative images at 14 days post-injury of sham and injured animals. We then sought to identify cell death pathways that may play a role in the Rice-Vinucci model. Spectrum breakdown products are markers of cellular necrosis. You can see that at one day post-injury, there are significant increases in spectrum breakdown products in the injured animals compared to sham animals, which is reflected graphically to the right. And total PARP can be broken down as part of cellular apoptosis, and at 24 hours and four days post-injury, there were significant decreases in total PARP, suggestive of ongoing cellular apoptosis. We then sought to characterize the FGF21 axis in the model prior to giving exogenous FGF21 therapy. And because the levels of hormones and receptors may be developmentally mediated, we examined FGF21 serum levels as well as expression of the beta-clotho and RBM3 proteins in both sham and injured animals at 24 hours, four days, and eight days post-injury. When we looked at serum FGF21 levels, we saw a significant increase in FGF21 at 24 hours post-injury compared to shams. We were particularly excited by this finding as we recently reported that in pediatric post-arrest patients, there were significant increases in FGF21 at zero to 24 hours post-arrest compared to healthy controls. We then examined levels of the beta-clotho co-receptor in the hippocampus and found that there were significant decreases in beta-clotho expression at four days post-injury compared to shams, with subsequent significant increases at eight days post-injury compared to shams. And finally, we examined the expression of the neuroprotective cold shock protein RBM3. We found a trend towards a decrease in RBM3 levels at 24 hours post-injury in injured animals compared to shams. And subsequently, at eight days post-injury, we saw a significant increase in RBM3 levels. Having fully characterized the model, we then sought to examine the effect of exogenous FGF21 supplementation. And for these studies, animals were randomized to one and a half mg per kg of either subcutaneous FGF21 or vehicle at five minutes post-injury. For studies that extended out past one day, they received daily injections of either vehicle or FGF21. We found that looking at the same cell death pathways that we had examined previously, there were significant increases in spectrum breakdown products in almost all of the vehicle treated animals. But only in one of the FGF21 treated animals had significant increases in spectrum breakdown products. So there was a significant difference in spectrum breakdown products between sham and vehicle animals, but no significant difference between FGF21 treated and sham animals. We then examined levels of total PARP, marker of cellular apoptosis, and found significant increases in total PARP in FGF21 treated animals compared to vehicle treated animals. We then examined the influence of FGF21 supplementation on the neuroprotective cold shock protein RBM3. At 24 hours and four days post-injury, we saw significant decreases in RBM3 levels in vehicle treated animals, but we saw no significant differences between sham animals and FGF21 treated animals, suggesting that FGF21 supplementation attenuates the loss of RBM3. And finally, we examined the effect of FGF21 supplementation on another cold stress hormone, cold inducible RNA binding protein, or CIRBP. CIRBP is not downstream of the beta-clotho receptor. And when we examined CIRBP levels, there was a significant difference on non-parametric ANOVA at 24 hours post-injury, but no significant individual post-hoc differences. But at four days post-injury, there were significant increases in FGF21 treated animals compared to the sham treated animals. This suggests that there may be an interrelationship between the two cold shock proteins, or the FGF21 may have non-beta-clotho-mediated impacts on other cold stress hormones. So in summary, in this study, we found that FGF21, exogenous FGF21 therapy in the Rice-Vinucci model decreases markers of cellular apoptosis and necrosis, attenuates the loss of the neuroprotective cold shock protein RBM3, and at late time points, increases the neuroprotective cold shock protein cold-inducible RNA binding protein. I'd like to thank my mentors, Pat Kohanek and Travis Jackson, as well as all the members of the Kohanek and Jackson lab who helped with this project.

FGF21 therapy

neonatal hypoxic ischemic brain injury

therapeutic hypothermia

cardiac arrest

cold stress hormone