GSDMD gene knockout alleviates hyperoxia-induced hippocampal brain injury in neonatal mice

Background: Neonatal hyperoxia exposure is associated with brain injury and poor neurodevelopment outcomes in preterm infants. Our previous studies in neonatal rodent models have shown that hyperoxia stimulates the brain’s inflammasome pathway, leading to the activation of gasdermin D (GSDMD), a key executor of pyroptotic inflammatory cell death. Moreover, we found inhibition of GSDMD activation attenuates hyperoxia-induced brain injury in neonatal mice. We hypothesized that GSDMD plays a pathogenic role in hyperoxia-induced neonatal brain injury and that GSDMD gene knockout (KO) will alleviate hyperoxia-induced brain injury. Methods: Newborn GSDMD knockout mice and their wildtype (WT) littermates were randomized within 24 h after birth to be exposed to room air or hyperoxia (85% O2) from postnatal day 1 to 14. Hippocampal brain inflammatory injury was assessed in brain sections by immunohistology for allograft inflammatory factor 1 (AIF1), a marker of microglial activation. Cell proliferation was evaluated by Ki-67 staining, and cell death was determined by TUNEL assay. RNA sequencing of the hippocampus was performed to identify the transcriptional effects of hyperoxia and GSDMD-KO, and qRT-PCR was performed to confirm some of the significantly regulated genes. Results: Hyperoxia-exposed WT mice had increased microglia consistent with activation, which was associated with decreased cell proliferation and increased cell death in the hippocampal area. Conversely, hyperoxia-exposed GSDMD-KO mice exhibited considerable resistance to hyperoxia as O2 exposure failed to increase either AIF1+ or TUNEL+ cell numbers, nor decrease cell proliferation. Hyperoxia exposure differentially regulated 258 genes in WT and only 16 in GSDMD-KO mice compared to room air- exposed WT and GSDMD-KO, respectively. Gene set enrichment analysis showed that in the WT brain, hyperoxia differentially regulated genes associated with neuronal and vascular development and differentiation, axonogenesis, glial cell differentiation, and core development pathways hypoxia-induced factor 1, and neuronal growth factor pathways. These changes were prevented by GSDMD-KO. Conclusion: GSDMD-KO alleviates hyperoxia-induced inflammatory injury, cell survival and death, and alterations of transcriptional gene expression of pathways involved in neuronal growth, development, and differentiation in the hippocampus of neonatal mice. This suggests that GSDMD plays a pathogenic role in preterm brain injury, and targeting GSDMD may be beneficial in preventing and treating brain injury and poor neurodevelopmental outcomes in preterm infants.


Introduction 65
Each year more than 15 million infants are born preterm worldwide [1]. Extremely premature 66 infants born at less than 28 weeks of gestational age are at great risk of having a multi-organ 67 injury that predominantly involves the lung and brain [2][3][4][5]. Born with immature lungs, these 68 premature infants suffer respiratory failure soon after birth and often require oxygen (O 2) therapy 69 and mechanical ventilation to survive. However, this life-sustaining high-concentration O2 therapy 70 (hyperoxia) can cause lung inflammation that ultimately leads to bronchopulmonary dysplasia 71 (BPD), characterized by disrupted alveolar and vascular development and reduced lung function 72 [2,3]. The immature brains in these premature infants are also affected by the hyperoxia that 73 results in inflammation leading to short-term and long-term neurodevelopmental sequelae such 74 as intraventricular hemorrhage, encephalopathy of prematurity, cerebral palsy, intellectual 75 disability, and cognitive deficits [4,5]. Therefore, survivors of BPD are known to suffer not only 76 from long-term lung disease but also suffer from long-term sequelae involving the brain leading 77 to long-term neurodevelopmental impairment (NDI). Moreover, there is mounting clinical evidence 78 bp deletion in exon 5 of the GSDMD gene. Experiments were done with homozygous KO mice 131 and their WT littermates. 132 133

Assessment of Hippocampal Cell Proliferation and Death 152
Cell proliferation was assessed by immunofluorescent staining for Ki67, a nuclear proliferation 153 marker, and the proliferating index was calculated as the average percentage of Ki67-positive 154 nuclei in total nuclei in 5 random HPV on hippocampal sections from each animal. Cell death was 155 studied using a TUNEL assay, and the cell death index was calculated as the average percentage 156 of TUNEL-positive nuclei in total nuclei in 5 random HPV on hippocampal sections from each 157 We first showed that GSDMD is expressed in the room air-exposed WT (WT-RA) hippocampal 188 sections, and it was increased in the hyperoxia-exposed WT (WT-O2) hippocampal sections. But 189 GSDMD was undetectable in RA-exposed GSDMD-KO (KO+RA) brains and significantly 190 decreased in hyperoxia-exposed GSDMD-KO (KO+O2) brains (Fig. 1A). GSDMD gene 191 expression measured by qRT-PCR showed that hyperoxia-exposed WT hippocampus had similar 192 expression compared to RA-exposed WT hippocampus, and its expression was extremely low in 193 the RA-exposed and hyperoxia-exposed GSDMD-KO hippocampus (Fig. 1B). These results 194 confirmed GSDMD deficiency in the KO brains, and hyperoxia increased GSDMD protein 195 expression in the hippocampus. 196

GSDMD-KO Reduces Hippocampal Inflammation in Hyperoxia-exposed Neonatal Mice 198
We next examined hippocampal sections for microglial infiltration by immunostaining to assess 199 whether GSDMD-KO affects hyperoxia-induced hippocampal inflammation. Histologically, there 200 were many microglial cells in the WT+O2 brains compared to the other three groups ( Fig. 2A). 201 These cells had activated microglia features such as enlargement of the cell body, reduction in 202 the territory, irregular cell shape, and self-association for each other ( Fig. 2A). Quantitative 203 analysis showed that the microglial count was 4-fold higher in the WT-O2 group than in the 204 GSDMD-KO groups (P < 0.001, Fig. 2B). Thus, GSDMD-KO ameliorated hyperoxia-induced 205 microglial cell activation in the hippocampus. 206

GSDMD Deficiency Improves Cell Survival and Decreases Cell Death in Hyperoxia-exposed 207
Brains 208 GSDMD is a key executor of inflammasome-induced pyroptosis, and hyperoxia is known to 209 reduce cell survival and cause cell death in hyperoxia-induced brain injury models. We found that 210 the WT-O2 group showed a 67% decrease in cell proliferation compared to WT-RA (P<0.01). 211 However, the KO-O2 group had approximately 58% increased cell proliferation compared to WT-212 O2 group (P < 0.05, Fig. 3A and B). When we assessed cell death, our data showed that the WT-213 O2 group had a nearly 1.5-fold increase in cell death compared to the other three groups (P < 214 0.01, Fig. 4A and B). with 146 genes upregulated and 112 genes downregulated (Fig. 6A), whereas in GSDMD-KO 229 animals hyperoxia differentially regulated only 16 genes ( Fig. 7A-B). Histogram of P-values in 230 KO-O2 vs. KO-RA showed uniform distribution, suggesting the few differentially expressed genes 231 identified are likely false discoveries. We performed an overrepresentation analysis on Topcluster 232 to identify biological processes and pathways for the genes induced and suppressed by hyperoxia 233 in WT animals. The bar graph in Fig. 6B shows the top Gene Ontology Biological Processes, and 234 KEGG and Reactome pathways associated with genes induced and suppressed by hyperoxia in 235 WT animals. Genes induced by hyperoxia were associated with neuroprojection morphogenesis, 236 neuron differentiation, neuron development, axonogenesis, blood circulation, hypoxia-induced 237 factor 1 (HIF-1) pathway, cell growth, chemotaxis, angiogenesis, and vascular development. 238 Suppressed genes were associated with neural growth factor (NGF) stimulated transcription, 239 memory, nuclear events kinase and transcription factor activation, short-term memory, response 240 to corticosterone, cell surface receptor signaling pathway involved in cell-cell signaling, ligand-241 activated transcription factor activity, and response to hypoxia. Network plots for the top 242 differentially induced gene pathways in WT brains were axonogenesis, neuron projection 243 guidance, developmental growth involved in morphogenesis, axon guidance, and developmental 244 cell growth (Fig. 6C). The top differentially suppressed gene pathways included cognition, 245 learning or memory, and muscle dell development (Fig. 6D). blood vessel diameter maintenance, vascular process in circulatiory system, reponse to BMP, 256 and oligodendrocyte differentiation (Fig. 7B). Network plots for the top differentially induced gene 257 pathways in the KO brains were regulation of neurogenesis, axonogenesis, synapse organization, 258 dentate gyrus development, and limbi system development (Fig. 7C). Top suppressed gene 259 pathways included regulation of blood circulation, glial cell differention, axon guidance, and 260 myelination (Fig. 7D). These findings suggest that in the setting of hyperoxia, GSDMD-KO 261 modulated important developmental pathways in the hippocampus. 262

263
We performed qRT-PCR to verify select genes differentially regulated by hyperoxia and GSDMD-264 KO. Representative genes whose expression was increased by hyperoxia in the WT brains but 265 reduced by GSDMD-KO included basic helix-loop-helix family member e40 (Bhlhe40), endothelin 266 1 (Edn1), immediate early response 3 (Ier3), and serpine1 which are involved in the neurovascular 267 injury, synaptic plasticity, apoptosis, and cellular senescence ( GSDMD-laden extracellular vesicles (EV) derived from hyperoxia-exposed rat models into healthy 277 neonatal rats induced pathological hallmarks of BPD, and these GSDMD-laden EVs can cross 278 the blood-brain barrier causing inflammatory brain injury [13]. In this study, we focused our 279 investigations on the effects of GSDMD-KO in a neonatal mouse model of hyperoxia-induced 280 brain injury. We provided evidence, for the first time to the best of our knowledge, that GSDMD 281 deficiency ameliorates hyperoxia-induced inflammation and cell death in the hippocampus. We 282 also report the effects of GSDMD-KO on hyperoxia-modulated transcriptomes and distinctive 283 enriched biological pathways in the hippocampus. 284 It is well known that the etiology of lung injury and BPD in preterm infants is multifactorial. 285 However, hyperoxia is thought to be a significant contributor to the inflammatory response 286 mediated by macrophages and neutrophils, which invade the endothelium and alveolar spaces of 287 premature lungs, causing lung injury and subsequent development of BPD [3]. A previous study 288 from our laboratory demonstrated that hyperoxia-exposed GSDMD-KO animals had significantly 289 less alveolar macrophage and neutrophil infiltration, an improvement in alveolarization and gas 290 exchange surface area, improved vascularization and less vascular remodeling/muscularization 291 compared to the WT mice indicating improvements in the hyperoxia-induced lung injury, and 292 deranged alveolar and vascular development that are seen in BPD [17]. 293

294
In addition, mounting evidence suggests hyperoxia is an important trigger of brain injury. Studies 295 have shown that the developmental stages of the lung and brain in rodent models are comparable 296 to preterm humans [25][26][27]. In most rodent models, lung injury has been noticed after exposure 297 to hyperoxia for 7 to 14 days, whereas brain injury has been detected after exposure to hyperoxia 298 for just 6 hours to 48 hours. In our study, rodent models were exposed to hyperoxia for 14 days, 299 aiming to investigate the chronic effects of hyperoxia on brain injury, and laying the foundation for 300 further potential investigations of the complex lung-brain axis interactions which result in multiple 301 comorbidities in preterm infants.  immunostaining showed GSDMD expression was increased in the hippocampal area of 596 hyperoxia-exposed wildtype (WT-O2) brain compared to the hippocampal areas of the room air-597 exposed WT (WT-RA), room air-exposed GSDMD-KO (KO-RA), and hyperoxia-exposed Quantification of cell death index (percentage of apoptotic nuclei divided by total nuclei) revealed 624 that WT hippocampus had increased cell death when exposed to hyperoxia. In contrast, 625 hyperoxia-exposed KO hippocampus had significantly less cell death. n=5/group. **P < 0.01, WT-626 RA vs. WT-O2. ## P < 0.001, WT-O2 vs. KO-O2. 20x magnification. Scale bars: 50 µm. animals. A. Volcano plot of genes differentially expressed in the hippocampus of room air-637 exposed WT mice compared to hyperoxia-exposed WT mice (fold change >1.25 and FDR<0.1). 638 B. Gene set enrichment analysis for Gene Ontology term and KEGG pathways of genes 639 differentially expressed by hyperoxia-exposed WT mice. Hyperoxia-induced genes were 640 associated with neuroprojection morphogenesis, neuron differentiation, neuron development, 641