Aging potentiates glioma-driven remodeling of cortico-hippocampal neuron-astrocyte dynamics

Aging potentiates glioma-driven remodeling of cortico-hippocampal neuron-astrocyte dynamics | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Biological Sciences - Article Aging potentiates glioma-driven remodeling of cortico-hippocampal neuron-astrocyte dynamics Steve Ramirez, Heloise Leblanc, Michelle Buzharsky, Michelle He This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8725420/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Age is the strongest determinant of glioblastoma incidence and outcome, yet how aging shapes the functional consequences of tumor burden across the brain remains unclear. Glioblastoma is accompanied by cognitive and affective dysfunction that implicates disruption of neural networks beyond the tumor core, particularly in the aged state. Here, we show that aging modifies how glioma burden progressively reshapes extratumoral brain circuit dynamics. Neurons largely retained behaviorally-aligned activity early in disease, with neuronal alterations such as diminished neuronal event magnitudes emerging with tumor progression in a region- and age-dependent manner. In contrast, astrocytic activity was disproportionately affected by aging and tumor burden, with age amplifying astrocytic engagement under control conditions and glioma burden selectively weakening astrocytic responsiveness, coordination, and long-range synchrony, most prominently in aged hosts and at earlier stages of tumor progression. These functional disruptions were accompanied by transcriptional remodeling in extratumoral regions, with astrocytes exhibiting inflammatory activation and suppression of circuit-supportive programs, and neurons showing complementary, region- and age-dependent alterations in genes related to synaptic function and activity. Together, these findings position astrocytic dysfunction as a dominant feature of age-associated extratumoral circuit vulnerability in glioblastoma and highlight glial mechanisms as key contributors to progressive functional decline in the aging brain. Biological sciences/Cancer/Cancer imaging Biological sciences/Neuroscience/Learning and memory/Hippocampus Biological sciences/Neuroscience/Glial biology/Astrocyte Biological sciences/Neuroscience/Neural ageing Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction Glioblastoma (GBM), is the most common malignant primary tumor of the central nervous system and remains among the most lethal human cancers 1 . Incidence increases sharply with age, and clinical outcome worsens steadily in older patients despite aggressive multimodal therapy 1,2 . Although survival has improved modestly in recent decades, these gains are limited, particularly in the elderly, underscoring that age is not merely a demographic correlate of GBM but a central biological determinant of disease course 3,4 . This age dependence is increasingly recognized as a property of the host brain rather than the tumor alone. When identical tumors are introduced into young and aged mice, the survival disadvantage of aged hosts is preserved, closely mirroring clinical observations, indicating that tumor progression and treatment response are shaped by age-dependent remodeling of the neural and immune environment 2 . Yet most preclinical GBM studies have historically relied on young adult mice, leaving a major gap in understanding how the aged brain’s baseline state alters the functional impact of tumor burden. A critical unanswered question is what these age-dependent effects mean at the multi-circuit level across cell types. Beyond survival, glioblastoma patients frequently experience cognitive, affective, and functional decline, and these impairments are particularly consequential in older individuals, shaping treatment tolerance, quality of life, and clinical decision-making 5–7 . Normal aging is accompanied by functional changes in both neurons and glial cells, including altered excitability, synaptic regulation, neuroinflammatory tone, and glial reactivity, and these changes are not uniform across the brain 8,9 . The prefrontal cortex and hippocampus are particularly vulnerable to aging-related remodeling, and they support stress regulation, affective state, and cognition 10 , domains frequently disrupted in GBM patients 11,12 . Thus, aging may alter tumor outcomes not only by shaping immune responses but also by changing how neural circuits and glial cells respond to tumor-associated perturbation. Astrocytes are well positioned to mediate such age-dependent vulnerability. They couple neuronal activity to vascular and metabolic support, integrate inflammatory and neuromodulatory signals, and dynamically regulate circuit excitability 13–15 . In aging and disease, astrocytes undergo reactive remodeling that can alter their interactions with neurons and reshape state-dependent circuit function 9,16 . Consistent with this framework, we previously found that tumor burden induces astrogliosis that is amplified in aged mice 17 , suggesting that aging heightens astrocyte reactivity to glioma-associated cues. Whether this heightened reactivity translates into altered astrocyte activity dynamics during behavior, and whether aging fundamentally changes neuron–astrocyte coordination under tumor burden, remains unknown. Here, we tested the hypothesis that aging interacts with tumor burden to reconfigure neuron–astrocyte dynamics in vivo within circuits vulnerable to aging and relevant to behavioral state regulation. Using dual-color fiber photometry, we simultaneously recorded neuronal and astrocytic calcium activity in the medial prefrontal cortex (mPFC) and ventral hippocampus (vHPC) during baseline conditions and across paradigms that reliably engage stress- and arousal-related states, including tail suspension 18,19 , elevated zero maze 20,21 , and fear conditioning 19,22 . Importantly, these paradigms are not used to model psychiatric disease per se but to evoke reproducible behavioral states that enable quantification of dynamic neuron–astrocyte interactions during heightened circuit engagement. Given the high prevalence of affective symptoms in GBM patients 11 , probing these dynamics in stress-related contexts may also provide insight into circuit disruptions relevant to clinical neuropsychiatric dysfunction. Finally, to connect these functional phenotypes to underlying molecular mechanisms, we used single-nucleus RNA sequencing to determine how tumor-associated transcriptional programs in these regions differ between young and aged hosts. Together, our work provides, to our knowledge, one of the first in vivo demonstrations that normal aging reshapes neuron–astrocyte coordination and reveals how aging then interacts with tumor burden to disrupt extratumoral circuit dynamics, identifying circuit-level vulnerabilities that may inform function-preserving strategies for older patients. Results To define how aging alters the functional impact of glioma burden on extratumoral brain circuits, we examined neuronal and astrocytic activity dynamics in the medial prefrontal cortex (mPFC) and ventral hippocampus (vHPC), regions central to stress and anxiety regulation and vulnerable to aging. We recorded neuronal activity using jRGECO1a and astrocytic activity using membrane-bound GCaMP8f, enabling cell-type–specific resolution within the same circuits (Fig. 1a). By recording from the hemisphere contralateral to tumor implantation, we isolated how tumor burden reshapes circuit dynamics beyond the tumor core, with a focus on neuron–astrocyte coordination. Histological analysis confirmed robust and comparable and separable expression of jRGECO1a in neurons and GCaMP8f in astrocytes within the mPFC and vHPC across age and tumor burden (Extended Data Fig. 1a-c, f). Quantification of jRGECO1a-positive neurons per region of interest revealed no significant effects of age or tumor burden in either region (Extended Data Fig. 1d,g). Similarly, astrocytic GCaMP8f coverage, quantified as labeled area per region of interest, did not differ significantly by age or tumor status in either region (Extended Data Fig. 1e,h). These data indicate that the functional differences observed in subsequent analyses are unlikely to reflect changes in reporter expression or gross cellular representation. Baseline neuron–astrocyte dynamics during resting state shows region-specific vulnerability To determine whether tumor burden alters circuit function in the absence of overt behavioral challenge, we examined spontaneous neuronal and astrocytic calcium dynamics during home cage behavior at early and later time points following tumor implantation (Extended Data Fig. 2). In the mPFC, tumor-bearing mice showed a progressive reduction in neuronal event rates over time, accompanied by age- and tumor-dependent reductions in astrocytic activity and coordination that were most apparent at the later time point. In contrast, baseline activity and local coordination in the ventral hippocampus were comparatively stable, with age-related but not tumor-associated differences. Specifically, aged mice across both days had higher neuronal rates than young mice, suggesting baseline neuronal hyperactivity in the aged hippocampus. Together, these baseline recordings suggest that tumor burden produces subtle, progressive disruptions in mPFC circuit dynamics that may become more pronounced when circuits are engaged by the behavioral demands of homecage exploration. Aging and tumor burden alter stress-evoked neuronal and astrocytic activity We next asked whether these baseline vulnerabilities are amplified during a potent, externally imposed stressor that recruits prefrontal–hippocampal circuits. During the tail suspension test, we observed clear, behaviorally structured calcium dynamics in both neurons and astrocytes across regions (Fig. 1a–b). Representative dual-color photometry traces illustrate robust neuronal activity fluctuations during struggling epochs, accompanied by slower, larger-amplitude astrocytic calcium events that were temporally coupled to behavioral transitions (Fig. 1b). We then quantified neuronal activity across the full session. In the mPFC, neuronal event rates were consistently higher during struggling than immobility across age and tumor groups (Behavior Effect: F(1,46)=120.6; ****P<0.0001), indicating preserved state-dependent encoding of coping behavior (Fig. 1c). Tumor-bearing mice exhibited a modest increase in overall neuronal event rate independent of behavioral state, while neuronal response magnitude was not significantly altered (Fig. 1c). We next examined astrocytic response magnitude across behavioral states. In the mPFC, astrocytic area under the curve (AUC) increased during struggling in control animals (Behavior Effect: F(1,87)=47.11; ****P<0.0001), with aged controls exhibiting larger responses than young controls (Fig. 1d). This increase was attenuated in young tumor-bearing mice and abolished in aged tumor-bearing mice (Fig. 1d). In the vHPC, neuronal event rates similarly differed between struggling and immobility in young control mice (Behavior Effect: F(1,56)=20.28; ****P<0.0001); however, this behavioral discrimination was reduced with aging (Fig. 1e). Tumor-bearing mice largely resembled age-matched controls, indicating that changes in vHPC neuronal state encoding were driven primarily by aging rather than tumor burden (Fig. 1e). In astrocytes, responses, like the mPFC, increased during struggling (Behavior Effect: F(1,52)=109.1; ****P<0.0001), with aged controls again showing larger event magnitudes overall (Fig. 1f). Likewise, this age-associated amplification was reduced in aged tumor-bearing mice (Fig. 1f). To determine whether these changes in event magnitude were accompanied by altered circuit coordination, we next examined correlations within and across regions. Neuron–neuron correlations between the mPFC and vHPC showed modest age-related reductions but were not strongly influenced by tumor burden (Fig. 1g). In contrast, astrocyte–astrocyte correlations across regions were significantly reduced in tumor-bearing mice, with the largest reductions observed in aged animals (Fig. 1h). Within regions, local neuron–astrocyte coupling in the mPFC was reduced in tumor-bearing mice, driven primarily by a near-significant decrease in aged tumor-bearing animals (Fig. 1i). In the vHPC, neuron–astrocyte correlations also showed a significant effect of tumor burden, although this effect was not attributable to a single age group (Fig. 1j). Together, these findings indicate that tumor burden preferentially disrupts astrocytic coordination locally and across regions, with aging amplifying these effects. Aging and tumor burden alter the temporal coordination of neuron–astrocyte responses during stress Given the strong dependence of cellular responses on behavioral state, we next assessed the temporal organization of neuronal and astrocytic activity aligned to the transition from immobility to struggling. Representative traces (Fig. 2a) and event heatmaps (Fig. 2b-c, Extended Fig. 3) revealed a conserved response sequence to struggling onset across regions, characterized by rapid neuronal activation followed by delayed astrocytic responses (Fig. 2a–c). This overall structure was preserved across age and tumor burden at the population level. Event-triggered average analyses confirmed that neuronal response magnitude and peak amplitude were not significantly altered by age or tumor burden in either region (Fig. 2e-f, k-l). In contrast, astrocytic ETAs revealed pronounced group differences. In the mPFC, astrocytic ETA AUC and peak amplitude were selectively reduced in aged tumor-bearing mice (Fig. 2g-h). In the vHPC, astrocytic ETA magnitude showed an age-associated increase under control conditions, with similar blunting of this response in aged tumor animals (Fig. 2m-n). In young animals, no effect was observed with tumor burden. Together, these findings indicate that tumor burden selectively erodes the astrocytic response to struggling onset in aged mice despite preserved neuronal activation. Given the close relationship between the heightened neuronal activity and subsequent astrocyte response, and the known importance of this astrocytic response for maintenance of neuronal circuit function, we quantified the response timing latency between neuron and astrocyte responses. In the mPFC, neuron–astrocyte peak latency differed across groups, with tumor-bearing mice showing altered delays relative to controls and aged tumor-bearing mice exhibiting the longest delays overall (Fig. 2i). In the vHPC, neuron–astrocyte peak latency was influenced primarily by age rather than tumor status (Fig. 2o). Together, these analyses demonstrate age- and tumor-dependent alterations in the timing and magnitude of astrocytic responses to elevated neuronal activity during stress. Notably, these circuit-level alterations were accompanied by age-specific shifts in stress coping behavior (Extended Fig. 4a-b). Young tumor-bearing mice showed no differences in immobility or latency to immobility, whereas aged tumor-bearing mice exhibited increased immobility and reduced latency to passive coping. The selective convergence of astrocytic dysfunction and behavioral impairment in aged mice supports a link between impaired astrocyte engagement and maladaptive stress responses under tumor burden. Aging and tumor burden alter neuron–astrocyte engagement during anxiogenic exploration Recent studies have demonstrated a central role for astrocytes in shaping anxiety-related behavior 23–25 . In contrast to externally imposed stressors such as tail suspension, exploration of anxiogenic environments reflects a self-directed, internally modulated behavioral state in which astrocytic signaling plays a prominent role in regulating circuit activity 23 . Given the pronounced astrocytic dysregulation we observed during tail suspension, we next asked whether similar circuit vulnerabilities emerge during anxiety-related behavior by recording activity during exploration of the elevated zero maze (EZM) (Fig. 3a). During EZM exploration, we again observed structured calcium dynamics in both neurons and astrocytes across regions. Representative traces illustrate relatively rapid, high-frequency neuronal activity during exploration alongside slower astrocytic calcium transients, with clear differences across age and tumor burden (Fig. 3b-c). We first quantified neuronal activity across EZM areas. In the mPFC, neuronal event rates were overall higher during open-area exploration than during closed-area exploration across all groups (Behavior Effect: F(1,33)=8.965; **P=0.0045), indicating preserved, albeit modest, behavioral state modulation during anxiogenic exploration (Fig. 3d). Overall neuronal rates were generally higher in young mice, but maintained across tumor conditions. However, when neuronal response magnitude was quantified using area under the curve (AUC), both young and aged tumor-bearing mice exhibited reduced overall neuronal AUC relative to age-matched controls, independent of maze area (Fig. 3e). Thus, while neuronal rates were preserved, tumor burden was associated with a global reduction in mPFC neuronal response magnitude rather than a behavior-specific effect. Astrocytic activity in the mPFC showed a distinct pattern. Astrocytic response magnitude generally increased, but exhibited a strong age dependence (Behavior x Age: F(1,39)=7.272; *P=0.0103): aged control mice displayed markedly larger astrocytic AUCs in the open area than young controls (Fig. 3f). This age-associated amplification was absent in tumor-bearing mice. Instead, both young and aged tumor-bearing animals showed reduced astrocytic AUCs relative to controls, independent of maze area (Fig. 3f), indicating a generalized suppression of astrocytic response magnitude. In the ventral hippocampus, across groups, neuronal event rates also distinguish open from closed areas (Behavior Effect: F(1,53)=30.87; ****P<0.0001) (Fig. 3g). Here, tumor-associated effects emerged selectively with aging: aged tumor-bearing mice showed reduced neuronal activity during open-area exploration, whereas closed-area activity remained comparable to controls (Fig. 3g). No tumor-associated differences were observed in young mice, indicating selective attenuation of anxiogenic neuronal engagement in the aged hippocampus. Astrocytic responses in the vHPC trended towards being increased during open-area exploration (Behavior Effect: F(1,52)=2.949; P=0.0919), although mostly driven by a significant increase in aged control animals (Fig. 3h). As in the mPFC, aged control mice exhibited larger astrocytic AUCs than young controls, reflecting age-associated amplification of astrocytic recruitment during anxiogenic exploration (Fig. 3h). Tumor burden selectively blunted this enhancement, with aged tumor-bearing mice showing reduced astrocytic AUCs relative to aged controls, while astrocytic responses in young mice were less affected (Fig. 3h). Thus, tumor effects on astrocytic recruitment extended beyond the tumor-adjacent cortex but were most pronounced with aging. Open-area entry–aligned responses reveal selective loss of astrocytic recruitment Exploratory behavior in the elevated zero maze introduces a more complex set of behavioral variables than the tail suspension, such as rearing and risk assessment without actual engagement of open areas 26 . Therefore, the peaks based method of assessing open-closed area dissociation did not yield as clear results as tail suspension. However, as shown in the representative trace (Fig. 4a), astrocytic events are especially tied to the entry into the open area, just not specific for this transition. Event-aligned heatmaps further illustrated this, as well as group differences in the structure and consistency of neuronal and astrocytic responses surrounding open-area entry in both the mPFC and vHPC (Fig. 4b-c), motivating quantitative analyses of event-triggered response magnitude. Neuronal event-triggered responses were detectable but small in both regions (Fig. 4b-c, Extended Fig. 5a,d). In the mPFC, neuronal ETA AUC and amplitude were low overall and were generally larger in aged mice, but did not differ significantly by tumor burden (Extended Fig. 5a-c). However, qualitatively, the small, organized increase in control mice lost consistency in both tumor groups (Fig. 4b, Extended Fig. 5a,g). Similar to the mPFC, neuronal responses in the vHPC were larger in aged animals (Extended Fig. 5e,f) . In addition, aged tumor mice had much greater variability between events and across mice, and a reduced AUC during open area entry (Fig. 4b, Extended Fig. 5d,e,i), suggesting a loss of stereotyped responses to anxiogenic stimuli selectively in aged tumor mice. Most notably in the EZM, however, was that astrocytic ETAs were substantially larger than neuronal responses in both regions (Fig. 4b-g, Extended Fig. 5h,j). In the mPFC, astrocytic ETA responses at open-area entry were amplified with age in control animals (Fig. 4d-e, Extended Fig. 5h). Tumor burden shifted astrocytic ETA magnitude downward, with this reduction most evident, and statistically significant, in aged tumor-bearing mice relative to aged controls, while young tumor-bearing mice showed a similar directional decrease that did not reach significance (Fig. 4d-f). A similar pattern was observed in the vHPC, where astrocytic ETA AUCs increased with age under control conditions and were selectively attenuated by tumor burden exclusively in aged animals (Fig. 4g-i, Extended Fig. 5j). This reduction was not driven by decreased peak amplitude (Fig. 4i), as observed in the mPFC, but instead reflected shorter-duration astrocytic responses, as evident from the average event-triggered traces (Fig. 4g). We next assessed coordination across cells and regions during EZM exploration. Local neuron–astrocyte correlations in the mPFC were reduced in tumor-bearing mice, with the strongest reductions observed in aged animals (Fig. 4j). In the vHPC, neuron–astrocyte correlations also showed a significant effect of tumor burden, although this effect was not driven by a specific age group (Fig. 4k). Cross-regional coordination revealed a more selective disruption: neuron–neuron correlations between the mPFC and vHPC showed only modest reductions (Fig. 4l), whereas astrocyte–astrocyte correlations across regions were robustly reduced in tumor-bearing mice of both ages (Fig. 4m), identifying long-range astrocytic synchrony as a major locus of disruption during anxiogenic exploration. Together, these results indicate that during anxiogenic exploration, astrocytic recruitment is amplified with aging and selectively compromised by tumor burden, whereas neuronal responses are comparatively weak and show more limited, region-specific effects. Consistent with these alterations in cellular dynamics, tumor-bearing mice, particularly aged animals, spent more time in the open areas (Extended Fig. 4c) and exhibited reduced risk-assessment behavior (Extended Fig. 4d), linking age-dependent astrocytic dysfunction to altered anxiety-related behavior under tumor burden. Alteration to neuronal and astrocytic responses to foot shock with tumor burden Elevated zero maze exploration revealed marked disruption of astrocytic recruitment in the mPFC across tumor-bearing mice and a selective reduction of astrocytic response magnitude in the vHPC of aged tumor-bearing animals. To determine whether these effects reflected an inability of these circuits to generate large responses or a more specific failure of state-dependent recruitment, we next examined neuronal and astrocytic activity during foot shock, a stimulus known to elicit strong, synchronized activation across limbic circuits 19 . On the same day as EZM testing, mice were subjected to brief foot shocks (Extended Fig. 6). Shock-elicited event-triggered averages revealed tightly time-locked neuronal activation at shock onset, followed by delayed astrocytic responses with prolonged decay in both regions (Extended Fig. 6a–h), consistent with canonical neuron–astrocyte response sequences. Quantification of neuronal response magnitude revealed a significant effect of tumor burden in the medial prefrontal cortex. In both young and aged mice, tumor-bearing animals exhibited reduced neuronal ETA area under the curve (AUC) relative to age-matched controls, indicating attenuated prefrontal neuronal recruitment in response to shock (Extended Fig. 6i). Astrocytic ETA AUCs in the mPFC were likewise significantly reduced in tumor-bearing mice across both age groups (Extended Fig. 8j). In contrast, neuronal and astrocytic ETA AUCs in the ventral hippocampus were not significantly altered by age or tumor status (Extended Fig. 6k–l), indicating preserved hippocampal responsiveness to this highly salient aversive stimulus. These data raise the possibility that reduced cellular responses in the mPFC during elevated zero maze exploration reflect diminished response capacity with advancing tumor burden, whereas reduced responses in the ventral hippocampus of aged tumor-bearing mice may reflect altered recruitment during anxiogenic exploration. Aging reveals transcriptional reprogramming of extratumoral circuits in glioma The circuit-level alterations observed during stress and anxiogenic behavior prompted us to ask whether aging similarly reshapes the transcriptional landscape of extratumoral brain regions under tumor burden. To address this, we performed single-nucleus RNA sequencing of the medial prefrontal cortex (mPFC) across all experimental groups and of the ventral hippocampus (vHPC) from aged control and tumor-bearing mice (Fig. 5a), reflecting the regions and age groups in which functional alterations were most pronounced. This approach captured all major neural, glial, and vascular-associated cell types (Fig. 5b, Extended Fig. 7), enabling systematic assessment of how glioma and aging interact to remodel cellular composition and gene expression beyond the tumor core. Analysis of cell-type proportions in the mPFC revealed tumor-associated changes that differed by age (Fig. 5c). Oligodendrocytes were reduced in both young and aged tumor-bearing mice. Microglia/perivascular macrophages increased with tumor burden, showing a trend toward expansion in young mice and a robust increase in aged animals, consistent with age-amplified inflammatory activation in cortex proximal to the glioma. In contrast, vascular leptomeningeal cells and pericytes, cell populations critical for blood–brain barrier integrity, were selectively reduced in aged tumor-bearing mice, indicating age-dependent vulnerability of the neurovascular compartment. To quantify the extent of transcriptional disruption across cell types, we next applied Augur prioritization analysis. In aged tumor-bearing mice, all captured cell types exhibited significant deviation from baseline classification performance, indicating widespread transcriptional remodeling (Extended Fig. 7b). In contrast, perturbation in young tumor-bearing mice was still broad but slightly more restricted: all cell types except oligodendrocyte precursor cells and vascular leptomeningeal cells showed significant prioritization (Extended Fig. 7b). Direct comparison of perturbation magnitude further revealed cell-type–specific age biases. Oligodendrocytes, astrocytes, excitatory neurons, endothelial cells, and vascular leptomeningeal cells were more strongly perturbed in aged tumor-bearing mice, whereas inhibitory neurons showed greater perturbation in young tumor-bearing mice (Fig. 5d). Microglia and pericytes exhibited comparable perturbation across ages, suggesting shared tumor-associated transcriptional responses independent of aging (Fig. 5d). Given the selective loss of vascular-associated cell types in aged tumor-bearing mice, the central role of astrocytes in neurovascular and metabolic coupling, and the pronounced neuron–astrocyte dysfunction observed in our in vivo recordings, we focused subsequent analyses on neurons and astrocytes as potential integrators of age- and tumor-related circuit vulnerability. Differential expression analyses revealed a pronounced age dependence in both the magnitude and direction of transcriptional change. Across cell types, aged tumor-bearing mice exhibited substantially greater transcriptional downregulation than young tumor-bearing mice (Fig. 5e,h,k). Neurons, particularly excitatory neurons, had downregulated pathways in both age groups predominantly related to synaptic structure and activity (Fig. 5e-g, Extended Fig. 8a). This downregulation was especially pronounced in excitatory neurons of aged tumor mice, with genes such as Slitrk1 and Homer1 , genes essential for synapse structure, significantly downregulated (Extended Fig. 8a,d-e). In contrast, genes upregulated across cell types in both young and aged tumor-bearing mice were consistently enriched for inflammatory and stress-related pathways, indicating a shared tumor-associated inflammatory program that was preserved across age (Fig. 5e-m). In neurons, these pathways included responses to interferon beta, and upregulation of genes such as Stat1 in both age groups, and Stat2 in aged tumor mice (Fig. 5f-g, i-j, Extended Fig. 8b,f). However, this was also associated with downregulation of genes associated with lymphocyte activity (Extended Fig. 8c), especially in aged tumor mice, suggesting a regulatory response by neurons to cytotoxicity. Astrocytes were particularly notable for the extent and specificity of transcriptional suppression in aged tumor-bearing mice. In young tumor-bearing animals, relatively few astrocytic pathways were downregulated, and these were primarily metabolic in nature (Fig. 5i-j). In aged tumor-bearing mice, however, astrocytes exhibited broad suppression of pathways involved in central nervous system differentiation, synaptic support, and neuronal signaling (Fig. 5m, Extended Fig. 9a-b). Genes implicated in astrocyte–neuron communication and synapse stabilization, including Sparcl1 and Megf10 , were among the most strongly suppressed (Extended Fig. 9b,f-g). Upregulated pathways were more universal across age groups, and included canonical markers of astrocyte reactivity, such as Vim and Gfap (Extended Fig. 9c,e) . The selective loss of vascular-associated cell populations in aged tumor-bearing mice raised the possibility that astrocytes may be responding transcriptionally to compromised blood–brain barrier function. To test this directly, we examined curated gene sets derived from prior studies characterizing astrocytic responses to blood–brain barrier disruption. Astrocytes from both tumor-bearing groups, and most prominently from aged tumor-bearing mice, showed substantial overlap with established BBB disruption signatures (Extended Fig. 9d). Of particular interest was the downregulation of Glud1 in aged tumor astrocytes (Extended Fig. 9h), a gene central to astrocytic glutamate metabolism and recycling at the synapse. Reduced expression of this enzyme would be expected to impair astrocytic regulation of extracellular glutamate and neuronal excitability. Aging reveals transcriptional perturbation in circuits distal to tumor burden To determine whether tumor-associated transcriptional changes extend to regions anatomically distant from the tumor site, we performed single-nucleus RNA sequencing of the ventral hippocampus from aged tumor-bearing and age-matched control mice (Fig. 6, Extended Fig. 10). We captured all cell types from the vHPC, as well as cells from adjacent cortex, which were excluded from further analyses (Fig. 6a, Extended Fig. 10a-b). In contrast to the mPFC, cell-type proportions in the vHPC were preserved, with no detectable differences between tumor and control conditions (Fig. 6b, Extended Fig. 10c), indicating that gross cellular composition in this region remains stable under tumor burden. Despite this compositional stability, transcriptional perturbation was widespread. Augur analysis revealed significant prioritization across all major cell populations except pericytes and endothelial cells, demonstrating that tumor burden elicits broad molecular responses in the aged hippocampus even in the absence of overt changes in cell abundance (Fig. 6c). Among neuronal populations, excitatory neurons exhibited the largest number of differentially expressed genes, with the majority showing downregulation (Fig. 6d). Inhibitory neurons displayed a distinct pattern, with a comparable number of upregulated genes but fewer downregulated genes than excitatory neurons (Fig. 6f), suggesting cell-type–specific transcriptional alterations in response to tumor burden. Pathway analysis revealed a conserved induction of mitochondrial translation in both excitatory and inhibitory neurons, accompanied by broad suppression of gene programs related to neuronal structure and excitability (Fig. 6e,g,k). Despite their anatomical distance from the tumor, astrocytes in the ventral hippocampus exhibited robust transcriptional responses (Fig. 6h). Similar to the medial prefrontal cortex, astrocytes showed prominent upregulation of inflammatory gene programs (Fig. 6j), including Ier3 and Gfap , consistent with a distributed glial response to tumor-associated cues. Astrocytes also exhibited increased expression of metabolic genes (Fig. 6j) such as Aldoc , suggesting an altered metabolic state or increased energetic demand. Notably, astrocytes shared with neurons a suppression of pathways associated with synaptic organization, membrane potential regulation, and cognition, implicating astrocytic dysfunction in broader disruption of hippocampal circuit support (Fig. 6e,g,i,k). Of particular interest, both excitatory and inhibitory neurons showed downregulation of Npas4 (Fig. 6k), an activity-dependent immediate early gene, consistent with the reduced neuronal engagement observed during elevated zero maze exploration at later stages of disease. In addition, all cell types exhibited decreased expression of Htr7 (Fig. 6k), a serotonergic receptor enriched in the adult hippocampus and implicated in cognition and mood regulation. Together, these findings demonstrate that glioma burden induces widespread, age-dependent transcriptional remodeling across extratumoral brain regions. Rather than acting solely through local effects near the tumor core, glioma burden engages a distributed molecular response characterized by inflammatory activation, metabolic stress, and suppression of synaptic support programs. These transcriptional alterations parallel the functional disruptions in neuron–astrocyte coordination observed in vivo and reinforce astrocytes as central mediators of glioma-associated vulnerability in the aging brain. Discussion Aging is a dominant determinant of glioblastoma incidence and outcome 27 , yet how the aged brain reshapes circuit-level vulnerability to tumor burden has remained poorly understood. Here, we show that aging fundamentally alters the way glioma burden disrupts extratumoral brain circuits by selectively targeting astrocytic integration rather than broadly suppressing neuronal activity. Across behavioral contexts, astrocytes emerged as the earliest and most sensitive cellular substrate of tumor-associated dysfunction, with aging amplifying both the magnitude and functional consequences of this disruption. A key insight from this work is that tumor burden does not simply abrogate neural activity, but instead degrades the glial-mediated coordination required to transform neuronal signals into coherent, behaviorally appropriate circuit output (Fig. 1 - 4). Neuronal activity largely preserved behavioral state encoding across stress- and anxiety-related paradigms, even in the presence of tumor burden. In contrast, astrocytic responses showed reduced magnitude, altered timing, and weakened synchrony, particularly in aged animals. These findings suggest that glioma-associated circuit dysfunction arises not from loss of neuronal responsiveness per se, but from failure of astrocytes to appropriately integrate, sustain, and coordinate neuronal activity across space and time. This astrocyte-centered vulnerability was shaped by both aging and anatomical proximity to the tumor. Circuits adjacent to the tumor site exhibited the most severe impairments in astrocytic engagement and coordination, while more distal regions showed partial preservation of response structure. Aging markedly intensified these effects, revealing a spatial gradient in which tumor-associated signals interact with age-dependent astrocytic fragility to erode circuit function. Importantly, this pattern was consistent across distinct behavioral states, indicating a generalized impairment in astrocytic circuit support rather than task-specific failure. During anxiety-related behavior, astrocytic dysfunction again dominated the circuit phenotype. Aging was associated with enhanced astrocytic recruitment during anxiogenic states, suggesting that astrocytes normally scale their engagement with increased behavioral demand. Tumor burden selectively blunted this amplification in aged animals, suppressing astrocytic responses during both sustained exploration and discrete behavioral transitions. Neuronal alterations emerged more selectively and later, particularly in the ventral hippocampus, consistent with a model in which neuronal dysfunction arises downstream of impaired astrocytic support rather than as an initiating event. Disruption of astrocytic coordination, rather than local response magnitude alone, emerged as a defining feature of tumor-associated circuit failure. Tumor burden weakened local neuron–astrocyte coupling and, most strikingly, disrupted long-range astrocyte–astrocyte synchrony across regions. This loss of glial coordination occurred even when neuronal activity retained behavioral modulation, suggesting that astrocytes play a critical role in aligning distributed circuit states. Aging amplified this desynchronization, potentially rendering circuits less resilient to tumor-associated perturbation. Moreover, our single-nucleus RNA sequencing data provides a molecular framework for these physiological findings. In the tumor-adjacent cortex, aging dramatically increased the depth and breadth of transcriptional suppression, with astrocytes showing pronounced downregulation of genes involved in synaptic support, neuronal signaling, and metabolic coupling. These changes were accompanied by selective loss of vascular-associated cell types and transcriptional signatures consistent with blood–brain barrier disruption, implicating breakdown of the neurovascular unit as a contributor to astrocytic dysfunction. In distal regions, transcriptional perturbations were evident despite preserved cellular composition, indicating that tumor-associated molecular stress propagates beyond the tumor core without overt structural remodeling. Across regions, astrocytes exhibited the strongest convergence of inflammatory activation and loss of circuit-supportive programs, positioning them as integrators of tumor- and age-related stress. Together, these findings support a model in which aging converts astrocytes from adaptive circuit regulators into points of failure under tumor burden. Early astrocytic dysfunction compromises coordination and sustained engagement, progressively constraining neuronal function and degrading behavioral output. This framework helps reconcile why neuronal activity can appear relatively preserved in early disease stages while behavioral impairment and network instability nonetheless emerge, particularly in older hosts. From a translational perspective, these findings highlight a fundamental tension in glioblastoma therapy. Many pathways that support synaptic maintenance, metabolic coupling, and neuromodulatory signaling are also co-opted by glioma cells to promote tumor growth. In the aged brain, where these circuit-supportive mechanisms are already compromised, further disruption by tumor-directed interventions may exacerbate functional decline even as tumor burden is reduced. Similarly, therapies designed to ameliorate cognitive function in patients could have consequential effects to tumor growth and progression. Our results therefore suggest that preserving astrocytic and neurovascular function may be essential for maintaining circuit integrity and quality of life in older patients alongside anti-tumor strategies. In sum, this work identifies astrocytes as central mediators of age-dependent vulnerability to glioma burden and demonstrates that aging fundamentally reshapes extratumoral circuit dynamics before overt neuronal failure emerges. By revealing astrocytic dysfunction as a primary driver of disrupted network coordination beyond the tumor core, these findings position glial and neurovascular mechanisms as critical determinants of functional outcome in the aging glioblastoma brain. Methods Animals Young and aged male C57BL/6J mice were obtained from The Jackson Laboratory. For fiber photometry experiments, mice were acquired three weeks earlier to accommodate the viral expression period; tumor induction was performed at the same ages across all experiments, with young mice 6–7 weeks old and aged mice 85–86 weeks old at the time of glioma implantation. Mice were housed under standard conditions (12 h light/dark cycle, ad libitum access to food and water). All procedures were approved by the Institutional Animal Care and Use Committee at Boston University and performed in accordance with NIH guidelines. Cell line and mPFC glioma induction SB28-Ohlfest murine glioma cells were maintained under standard culture conditions (37°C, 5% CO₂) and routinely screened for mycoplasma contamination. For stereotaxic implantation, mice were anesthetized with isoflurane and placed in a stereotaxic frame. A 2 µL suspension containing 50,000 SB28 cells in DMEM was injected into the mPFC at 500 nL/min using a 10 µL, 33-gauge Hamilton syringe at DV −2.5 mm. Tumor coordinates (relative to bregma) were AP +1.8 mm and ML −0.5 mm. The syringe was left in place for 1 minute following injection to minimize reflux. Viral delivery and optical fiber implantation To record neuronal and astrocytic calcium activity, we used AAV-hSyn-jRGECO1a (neurons) and AAV-GfaABC1D-jGCaMP8f-SV40 (astrocytes). Viruses were mixed at a 1:2 dilution (jRGECO1a:jGCaMP8f) and co-injected into the medial prefrontal cortex (mPFC) and ventral hippocampus (vHPC). Injection volumes were 500 nL for mPFC and 750 nL for vHPC. Viruses were injected three weeks prior to tumor implantation and optical fiber placement to allow stable expression before induction of tumor burden and longitudinal recordings. Virus injections and fiber placements were performed in hemispheres contralateral to the tumor injection site. For mPFC, virus injection and recording coordinates were AP +1.8 mm, ML +0.5 mm, DV −2.5 mm (contralateral to tumor). For vHPC, injection coordinates were AP −3.16 mm, ML +3.16 mm, DV −4.5 mm, and the optical fiber was implanted at DV −4.6 mm (contralateral hemisphere to tumor). Optical fibers were 200 µm core diameter, NA 0.37, and were secured using dental cement. Timeline of fiber photometry and behavioral experiments Virus injection occurred three weeks prior to tumor induction and optical fiber implantation. Following tumor induction and fiber implantation, mice underwent longitudinal fiber photometry across baseline and behavioral assays. The timeline was: Home cage 1 (day 8) → tail suspension test (day 10) → home cage 2 (day 12) → elevated zero maze (day 13) → foot shocks, as shown in Fig. 1A. Behavioral assays and quantification Home cage recordings Photometry was recorded in the home cage on day 8 and day 12 of tumor burden to establish baseline neuronal and astrocytic activity during spontaneous behavior. Tail suspension test (TST) Mice were suspended by the tail for 6 min. Behavior was manually scored as struggling or immobility. For peri-event analyses, struggling epochs were required to be ≥2 s in duration to be included. Struggle onset was defined as the first frame of sustained movement following immobility, and struggle offset as the first frame of sustained immobility following a struggling epoch. These labels were used to segment photometry signals for state-dependent quantification. Due to technical limitations of fiber photometry recording, additional mice (non-fiber photometry) were needed to assess latency to immobility, which is reflected in Extended Fig. 4a-b. Latency to immobility was defined as the time elapsed from initial suspension to the first immobility of greater than 3 seconds. Elevated zero maze (EZM) Mice were placed on an elevated circular maze containing two open and two closed quadrants and recorded for 8 min. Behavior was scored using ANY-maze software. Behavioral state was defined as open area or closed area, and transitions were defined by entry into each area. The animal was only counted as having entered the open area if all four legs had crossed the transition point. Risk assessment behavior was characterized by the time in which an animal has its head out into the open area, scanning the environment, without fully engaging into the open area 28,29 . A subset of mice from the pure-behavior experiments for tail suspension also performed the EZM for additional N. For peri-event analyses, open-area entry was used as the alignment event. Foot shock To evoke well-established population level events in both brain regions 19,30 , mice underwent a brief foot shock session in a conditioning chamber. The session duration was 60 s. Foot shocks were delivered at 10 s, 30 s, and 50 s, each lasting 2 s at 1.5 mA. Photometry was recorded throughout the session. Fiber photometry Acquisition Dual-color fiber photometry recordings were performed using the FP3002 system (Neurophotometrics). Excitation light was delivered through implanted optical fibers, and emitted fluorescence was captured by a photodetector/camera system. Astrocytic GCaMP signals were excited at 470 nm and neuronal jRGECO signals were excited at 560 nm, with an output power of 50 µW per channel at the patch cord tip. Calcium-independent isosbestic signals were simultaneously captured by alternating excitation with a 415 nm LED to dissociate motion artifacts, tissue autofluorescence, and photobleaching from calcium-dependent fluorescence changes. All wavelengths were interleaved and collected using Bonsai software (Lopes et al., 2015). Signals were sampled at 30 Hz (30 frames per second). Photometry preprocessing and signal normalization Photometry traces were processed using custom Python scripts ( https://github.com/rsenne/RamiPho ). Raw fluorescence signals were baseline-corrected using an adaptive iteratively reweighted penalized least-squares method (airPLS; Zhang et al., 2010). Signals were then smoothed using kernel-based smoothing to improve signal-to-noise ratio. dF/F was calculated for each channel as: yielding percent deviation from baseline. For peri-event analyses and event-triggered averages (ETAs), photometry signals were z-scored within session prior to alignment and averaging to enable comparison of dynamics across animals and groups. For ETA area-under-the-curve (AUC) quantification, each mouse’s mean ETA trace was integrated over a 10 s post-event window for neuronal signals and a 15 s post-event window for astrocytic signals. Because astrocytic ETAs frequently exhibited large negative deflections, AUC for astrocytes was calculated on the absolute value of the ETA trace to capture full response magnitude. Neuron-astrocyte latency (Δt) was quantified per mouse as the difference between astrocytic and neuronal peak times in the mean ETA trace (Δt = t_astro − t_neuron) for each mouse. Event detection and peak-based quantification Calcium events were detected using a threshold-based peak identification approach with region- and cell-type-specific parameters. Peaks were identified on baseline-corrected, smoothed, and z scored traces using the SciPy function, find_peaks(), with the following parameters: Astrocyte : height ≥ 0.25 SD, prominence ≥ 1 SD, width = 90–500 frames, rel_height = 0.7, wlen = 600 Neuron : height ≥ 1 SD, prominence ≥ 2 SD, width = 15–500 frames, rel_height = 0.5 To quantify behavior-specific event properties, behavioral epochs were first annotated (e.g., struggling versus immobility in the tail suspension test, open versus closed area in the elevated zero maze), and events were assigned to behavioral states based on whether their peak time occurred within a given epoch. Event frequency was computed as events/min for each behavioral state. AUC were averaged across events within each state for each mouse, yielding one per-mouse value per state for group analyses. Peri-event analyses Peri-event analyses were performed by aligning z-scored signals to behavioral transitions, including struggle onset (TST) and open-area entry (EZM). Only trials meeting inclusion criteria (e.g., struggling epochs ≥2 s) were included. ETAs were computed per animal by averaging across trials, then averaged across animals within each group. Confidence bounds for ETAs were computed using a t-confidence interval (tCI) method 31 . For each peri-event time point, a 95% confidence interval was calculated across trials. Lagged correlation analysis Lagged correlations between z-scored photometry traces were computed per mouse using cross-correlation (SciPy correlate) to identify the lag of maximal correlation within predefined bounds. Lag windows were set a priori by pair (neuron–astro: 30–240 frames; neuron–neuron and astro–astro: 0–60 frames; default: 0–150 frames), and neuron–astro pairs were constrained to positive reported lags (astrocytes lagging neurons). Traces were shifted by trimming the leading portion of the appropriate signal at the optimal lag, and coupling strength was quantified as the Pearson correlation coefficient on aligned traces. Single-nucleus RNA sequencing (snRNA-seq) Tissue dissection and experimental design mPFC tissue was dissected on day 13 of tumor burden from young and aged control and tumor-bearing mice (n = 3 per group). vHPC tissue was collected on day 13 from aged control and aged tumor-bearing mice (n = 3 per group). Tissue was rapidly dissected on ice and immediately snap-frozen for nuclei isolation and single-nucleus RNA sequencing. Nuclei isolation, library preparation, and sequencing Nuclei were isolated and processed for 10x Genomics single-nucleus RNA sequencing. FASTQ files were processed using Cell Ranger v7.0.0 (10x Genomics), to perform STAR-based alignment and generate filtered feature-barcode matrices. Preprocessing, quality control, integration, and annotation Downstream analysis was performed in Seurat v4.1.1. Cells were filtered to remove low-quality barcodes and potential multiplets using the following thresholds: 8000 features, and mitochondrial transcript fraction >10%. Doublets were identified and removed using DoubletFinder v2.0.3. Each mouse dataset was normalized by total UMI counts and log-transformed, and variable features were identified using the vst method. Datasets were then integrated across mice using canonical correlation analysis (CCA) in Seurat (FindIntegrationAnchors/IntegrateData). The integrated data were scaled and used for PCA, followed by UMAP and/or t-SNE for visualization. Cell types were annotated based on canonical marker genes. For the ventral hippocampus, excitatory neurons included as “ventral hippocampus” included any cell cluster that included hippocampal subregions and prosubiculum, due to the prosubiculum having similar dorsal-ventral (DV) axis differentiation as the hippocampus and its close proximity to vHPC-proper subregions 32 . Cell Proportions Analysis Differential abundance was first assessed using a propeller-based analysis of sample-level cluster proportions from the speckle package 33 . Post-hoc, logit-transformed cluster proportions were modeled using limma with a group-wise design and empirical Bayes moderation to test within-age Tumor vs Control contrasts. Moderated statistics were extracted for all clusters in Young and Aged cohorts. Differential expression and gene set enrichment analysis Differential expression (DE) was performed within each annotated cell population to identify tumor-associated transcriptional changes relative to age-matched controls. For mPFC, tumor-bearing mice were compared to age-matched controls within each age group, and for vHPC, aged tumor was compared to aged control. DE was conducted using nebula, a negative binomial mixed model framework with mouse identity modeled as a random effect and log library size included as an offset term 34 . Models were fit using the hierarchical likelihood (HL) approach. Genes with θ greater than the 99.5th percentile of the θ distribution within each analyzed cell population were designated as “high-overdispersion” and were excluded from downstream DEG and enrichment analyses. P values were corrected for multiple comparisons using the Benjamini–Hochberg procedure, and genes were considered differentially expressed at FDR < 0.05. Gene Ontology (GO) enrichment was assessed using gene set enrichment analysis (GSEA) implemented in fgsea on ranked gene lists, where genes were ranked by sign(effect) × −log10(adjusted p-value), with enrichment significance determined at FDR < 0.05. To reduce redundancy among significantly enriched Gene Ontology (GO) terms, GO Biological Process terms passing FDR < 0.05 were summarized using REVIGO 35 . GO terms were provided to REVIGO with adjusted P values as the associated scores, and redundancy reduction was performed using the SimRel semantic similarity measure with a medium allowed similarity threshold (0.7). Obsolete GO terms were removed, and the whole UniProt database was used as the reference background. Cell-type prioritization analysis using Augur Cell-type prioritization was performed using Augur, which quantifies the extent to which transcriptional profiles within a given cell type discriminate between experimental conditions via repeated subsampling and supervised classification. For each annotated cell type, Augur repeatedly subsampled an equal number of nuclei per condition and trained a classifier to distinguish Control versus Tumor labels. Classification performance was evaluated using cross-validation and summarized using the area under the receiver operating characteristic curve (ROC AUC). This procedure was repeated across 50 independent subsamples to ensure robustness to cell sampling variability. Augur analyses were performed separately for Young and Aged cohorts. To assess whether individual cell types exhibited classification performance significantly above chance, we tested whether subsample-level ROC AUC distributions exceeded 0.5 (chance performance). For each cell type, ROC AUC values were first averaged across cross-validation folds within each subsample, yielding one independent performance estimate per subsample. A one-sided Wilcoxon signed-rank test was then applied to test whether the median subsample-level ROC AUC was greater than 0.5. P values were corrected for multiple testing across cell types using the Benjamini–Hochberg false discovery rate (FDR) procedure. To identify cell types exhibiting age-dependent differences in tumor-associated transcriptional disturbances, differential prioritization was assessed by comparing observed cell type-specific differences in Augur AUC scores between Young and Aged cohorts to a null distribution of ΔAUC values generated from label-permuted Augur runs. Empirical two-sided permutation p-values were computed for each cell type. Brain tissue processing and immunohistochemistry For validation of viral expression and fiber placement, and histological markers, mice were transcardially perfused with PBS followed by 4% paraformaldehyde. Heads were post-fixed for 48 hours in 4% paraformaldehyde, brains were extracted, and were sectioned at 40 µm. The intensity of the endogenous fluorescent proteins was sufficient so that no amplification step with immunohistochemistry was required. Images were acquired using a Zeiss LSM800 confocal microscope, with 886.5 (3 tiles) x 2068.5 (7 tiles) µm ROI for the mPFC and 1182.0 (4 tiles) x 1469.4 (5 tiles) µm ROI for the vHPC. jRGECO neuron counts were segmented and counted using Cellpose3 36 . Due to the membrane bound nature of the GCaMP protein, segmentation of individual cells was not feasible for astrocytes. Instead, the images were thresholded and the area of positive signal per ROI was obtained using FIJI/ImageJ. Due to the nature of tumor growth and proximity of the optical fiber to the tumor, a subset of mice lost signal over the progression of the disease. To determine whether this reflected a true loss of cellular activity rather than displacement of the optical fiber, we examined confocal images from each mouse. We excluded the first and following recordings in which signal loss occurred if the fiber tip was more than 200 µm from the nearest jRGECO⁺ neuron (excluding neuronal recordings), GCaMP⁺ astrocyte (excluding astrocyte recordings), or both (excluding recordings from both cell types), which corresponds to the loss of 50% of maximum photometry efficiency 37 . Statistical analysis Statistical analyses were performed using GraphPad Prism. When multiple comparisons were performed, Sidak’s correction was applied. Statistical significance was defined as p < 0.05. Data are reported as mean ± SEM unless otherwise indicated. Declarations Data and code availability Custom photometry preprocessing and analysis scripts are freely available at https://github.com/rsenne/RamiPho . Sequencing data will be deposited in Gene Expression Omnibus upon publication. Other data are available from the corresponding authors upon request. Author Contributions Conceptualization: H.L. and S.R.; methodology: H.L. and S.R.; formal analysis: H.L.; investigation: H.L., M.B., M.H.; visualization: H.L.; writing—original draft: H.L. and S.R.; writing—review and editing: H.L., M.B., M.H. and S.R.; funding acquisition: S.R.; resources: S.R.; and supervision: S.R. Competing interests The authors declare no competing interests. Funding and Acknowledgments This work was supported by NIH Transformative Award, the Ludwig Family Foundation, the Air Force Office of Scientific Research award (FA9550-21-1-0310), the Pew Scholars Program in the Biomedical Sciences, the Chan-Zuckerberg Initiative, and Neurophotonics Center at Boston University and the Center for Systems Neuroscience. Behavioral schematics and timelines were created using BioRender. Corresponding author Correspondence to Steve Ramirez ( [email protected] ). References Miller, K. D. et al. Brain and other central nervous system tumor statistics, 2021. CA Cancer J. Clin. 71 , 381–406 (2021). Ladomersky, E. et al. Advanced age negatively impacts survival in an experimental brain tumor model. Neurosci. Lett. 630 , 203–208 (2016). Chen, X., Cui, Y. & Zou, L. Treatment advances in high-grade gliomas. Front. Oncol. 14 , 1287725 (2024). Korja, M. et al. Glioblastoma survival is improving despite increasing incidence rates: a nationwide study between 2000 and 2013 in Finland. Neuro. Oncol. 21 , 370–379 (2019). Karsy, M. et al. Surgical treatment of glioblastoma in the elderly: the impact of complications. J. Neurooncol. 138 , 123–132 (2018). Gately, L., McLachlan, S. A., Dowling, A. & Philip, J. Life beyond a diagnosis of glioblastoma: a systematic review of the literature. J. Cancer Surviv. 11 , 447–452 (2017). Zhang, C. et al. Aging and senescence: Key players in brain tumor progression and drug resistance. Drug Resist. Updat. 81 , 101228 (2025). Mattson, M. P. & Arumugam, T. V. Hallmarks of Brain Aging: Adaptive and Pathological Modification by Metabolic States. Cell Metab. 27 , 1176–1199 (2018). Man, J. H. K. et al. Region-specific and age-related differences in astrocytes in the human brain. Neurobiol. Aging 140 , 102–115 (2024). Raz, N. & Rodrigue, K. M. Differential aging of the brain: patterns, cognitive correlates and modifiers. Neurosci. Biobehav. Rev. 30 , 730–748 (2006). van der Meer, P. B. et al. Depression and anxiety in glioma patients. Neurooncol Pract 10 , 335–343 (2023). Talacchi, A., Santini, B., Savazzi, S. & Gerosa, M. Cognitive effects of tumour and surgical treatment in glioma patients. J. Neurooncol. 103 , 541–549 (2011). Lyon, K. A. & Allen, N. J. From synapses to circuits, astrocytes regulate behavior. Front. Neural Circuits 15 , 786293 (2021). Pearson-Leary, J., Osborne, D. M. & McNay, E. C. Role of Glia in Stress-Induced Enhancement and Impairment of Memory. Front. Integr. Neurosci. 9 , 1–14 (2016). Perez-Catalan, N. A., Doe, C. Q. & Ackerman, S. D. The role of astrocyte-mediated plasticity in neural circuit development and function. Neural Dev. 16 , 1 (2021). Labarta-Bajo, L. & Allen, N. J. Astrocytes in aging. Neuron 113 , 109–126 (2025). Leblanc, H., Buzharsky, M., Varelas, X., Binello, E. & Ramirez, S. Age-dependent glioma progression and functional decline in a syngeneic Murine model: Host vulnerabilities and opportunities for targeted intervention. Neurooncol. Adv. (2025) doi:10.1093/noajnl/vdaf222. Hiraoka, K. et al. Pattern of c-Fos expression induced by tail suspension test in the mouse brain. Heliyon 3 , e00316 (2017). Shin, L. M. & Liberzon, I. The neurocircuitry of fear, stress, and anxiety disorders. Neuropsychopharmacology 35 , 169–191 (2010). Padilla-Coreano, N. et al. Direct ventral hippocampal-prefrontal input is required for anxiety-related neural activity and behavior. Neuron 89 , 857–866 (2016). Lim, S. C., Fusi, S. & Hen, R. Ventral CA1 population codes for anxiety. bioRxivorg (2023) doi:10.1101/2023.09.25.559358. Euston, D. R., Gruber, A. J. & McNaughton, B. L. The role of medial prefrontal cortex in memory and decision making. Neuron 76 , 1057–1070 (2012). Kim, E. et al. Prefrontal cortex astrocytes modulate distinct neuronal populations to control anxiety-like behavior. Nat. Commun. 16 , 7819 (2025). Cho, W.-H. et al. Hippocampal astrocytes modulate anxiety-like behavior. Nat. Commun. 13 , 6536 (2022). Li, J.-T. et al. Astrocytes in the ventral hippocampus bidirectionally regulate innate and stress-induced anxiety-like behaviors in male mice. Adv. Sci. (Weinh.) 11 , e2400354 (2024). Shepherd, J. K., Grewal, S. S., Fletcher, A., Bill, D. J. & Dourish, C. T. Behavioural and pharmacological characterisation of the elevated ‘zero-maze’ as an animal model of anxiety. Psychopharmacology 116 , 56–64 (1994). Pretanvil, J.-A., Salinas, I. Q. & Piccioni, D. E. Glioblastoma in the elderly: treatment patterns and survival. CNS Oncol. 6 , 19–28 (2017). Blanchard, D. C., Blanchard, R. J., Tom, P. & Rodgers, R. J. Diazepam changes risk assessment in an anxiety/defense test battery. Psychopharmacology (Berl.) 101 , 511–518 (1990). Blanchard, R. J., Blanchard, D. C. & Hori, K. An ethoexperimental approach to the study of defense. in Ethoexperimental Approaches to the Study of Behavior vol. 739 114–136 (Springer Netherlands, Dordrecht, 1989). Jimenez, J. C. et al. Contextual fear memory retrieval by correlated ensembles of ventral CA1 neurons. Nat. Commun. 11 , 3492 (2020). Jean-Richard-Dit-Bressel, P., Clifford, C. W. G. & McNally, G. P. Analyzing event-related transients: Confidence intervals, permutation tests, and consecutive thresholds. Front. Mol. Neurosci. 13 , 14 (2020). Ding, S.-L. et al. Distinct transcriptomic cell types and neural circuits of the subiculum and prosubiculum along the dorsal-ventral axis. Cell Rep. 31 , 107648 (2020). Phipson, B. et al. Propeller: Testing for differences in cell type proportions in single cell data. Bioinformatics 38 , 4720–4726 (2022). He, L. et al. NEBULA is a fast negative binomial mixed model for differential or co-expression analysis of large-scale multi-subject single-cell data. Commun. Biol. 4 , 629 (2021). Supek, F., Bošnjak, M., Škunca, N. & Šmuc, T. REVIGO summarizes and visualizes long lists of gene ontology terms. PLoS One 6 , e21800 (2011). Stringer, C. & Pachitariu, M. Cellpose3: one-click image restoration for improved cellular segmentation. Nat. Methods 22 , 592–599 (2025). Pisanello, M. et al. The three-dimensional signal collection field for fiber photometry in brain tissue. Front. Neurosci. 13 , 82 (2019). Additional Declarations There is NO Competing Interest. Supplementary Files FCFigStats.xlsx FC Stats CountsFigStats.xlsx Counts Stats EZMFigStats.xlsx EZM Stats HCFigStats.xlsx HC Stats DEGsFig.56.xlsx DEGs BehaviorFigStats.xlsx Behavior Stats TSTFigStats.xlsx TST Stats GSEAGOBPResults.xlsx GO results ExtendedDataFigs.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8725420","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Biological Sciences - Article","associatedPublications":[],"authors":[{"id":588790975,"identity":"25fb4363-957a-4fbb-a0e0-0d581243819b","order_by":0,"name":"Steve Ramirez","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAArklEQVRIiWNgGAWjYDCCAxBKzgBE8hCphbEBSBmTriVxA9Fa+G6kP3/wsc0mfbtEAuODt21EaJG8kWPYOLMtLXfnjARmw7nEaDE4c4axmefM4dwNNxLYpHmJ03L8YfOfM//TDW4ksP8mTsvxBsNmhooDCUAtbMxEaZE83mM4s6ci2XBnz8NmyTnniNDCd5j9wYcfBnby5uzJBz+8KSNCCxIAx88oGAWjYBSMAqoAAOsqO0iPNwbxAAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0002-9966-598X","institution":"Department of Psychological and Brain Sciences, The Center for Systems Neuroscience, Boston University","correspondingAuthor":true,"prefix":"","firstName":"Steve","middleName":"","lastName":"Ramirez","suffix":""},{"id":588790976,"identity":"3a584a1f-dd74-44e0-86b9-bcdc369a1205","order_by":1,"name":"Heloise Leblanc","email":"","orcid":"https://orcid.org/0000-0001-6355-3575","institution":"Boston University, Harvard University","correspondingAuthor":false,"prefix":"","firstName":"Heloise","middleName":"","lastName":"Leblanc","suffix":""},{"id":588790977,"identity":"1e896aa0-d877-4b0e-82ce-5785ec49763b","order_by":2,"name":"Michelle Buzharsky","email":"","orcid":"https://orcid.org/0000-0001-5781-4054","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Michelle","middleName":"","lastName":"Buzharsky","suffix":""},{"id":588790978,"identity":"94371d0a-8200-4365-9350-95ab47b475d4","order_by":3,"name":"Michelle He","email":"","orcid":"","institution":"Boston University","correspondingAuthor":false,"prefix":"","firstName":"Michelle","middleName":"","lastName":"He","suffix":""}],"badges":[],"createdAt":"2026-01-28 22:45:07","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8725420/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8725420/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":102380060,"identity":"91f4b9f3-d33a-4be1-bacf-4a7db9b5e152","added_by":"auto","created_at":"2026-02-11 06:35:22","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":209852,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eNeuronal and astrocytic activity during tail suspension. (a) \u003c/strong\u003eExperimental schematic illustrating tumor implantation, dual-color fiber photometry recordings from the medial prefrontal cortex (mPFC) and ventral hippocampus (vHPC), and behavioral timeline including tail suspension testing at Day 10. \u003cstrong\u003e(b)\u003c/strong\u003eRepresentative dual-color photometry traces from neurons (left) and astrocytes (right) in the mPFC during tail suspension across experimental groups: young control (YC), young tumor-bearing (YT), aged control (AC), and aged tumor-bearing (AT). \u003cstrong\u003e(c)\u003c/strong\u003e Neuronal event rates in the mPFC during immobility (I) and struggling (S) (left panel; All groups - I v S: ****P\u0026lt;0.0001) and session-averaged rates (right panel inlay; Tumor v Control: P=0.0658); YC: n= 14 mice, YT: n=11 mice, AC: n=13 mice, AT: n=13 mice. \u003cstrong\u003e(d)\u003c/strong\u003eAstrocytic area under the curve (AUC) in the mPFC during immobility (I) and struggling (S) (left panel; YC - I v S: ****P\u0026lt;0.0001, YT - I v S: **P=0.0253, AC - I v S: ****P\u0026lt;0.0001), and comparison between groups (right panel; AT v AC - S: **P=0.0011); YC: n= 14 mice, YT: n=10 mice, AC: n=12 mice, AT: n=13 mice. \u003cstrong\u003e(e) \u003c/strong\u003eNeuronal event rates in the vHPC during immobility (I) and struggling (S) (left panel; YC - I v S: ***P=0.0007, YT - I v S: *P=0.0347) and comparison between groups (right panel; YC v AC - I: *P=0.0178); YC: n= 13 mice, YT: n=15 mice, AC: n=16 mice, AT: n=16 mice.\u003cbr\u003e\n(f) Astrocytic AUC in the vHPC during immobility (I) and struggling (S) (left panel; YC - I v S: ***P=0.0007, YT - I v S: ***P=0.0006, AC and AT - I v S: ****P\u0026lt;0.0001) and comparison between groups (right panel; YC v AC - S: **P=0.0026, AT v AC - S: *P=0.0154); YC: n= 13 mice, YT: n=15 mice, AC: n=16 mice, AT: n=16 mice. (g) Maximum cross-regional neuron–neuron correlations between the mPFC and vHPC across groups. YC: n=14, YT: n=11 mice, AC: n=13 mice, AT: n=12 mice. (h) Maximum cross-regional astrocyte–astrocyte correlations between the mPFC and vHPC across groups (AT v AC: **P=0.0016); YC: n=14, YT: n=10 mice, AC: n=12 mice, AT: n=12 mice. (i) Maximum local neuron–astrocyte correlations within the mPFC across groups. (j) Maximum local neuron–astrocyte correlations within the vHPC across groups (AT v AC: P=0.0529); YC: n=14, YT: n=10 mice, AC: n=12 mice, AT: n=13 mice. (i) Maximum local neuron–astrocyte correlations within the mPFC across groups. YC: n= 13 mice, YT: n=15 mice, AC: n=16 mice, AT: n=16 mice. Data are shown as mean ± SEM with individual animals overlaid. Statistics were assessed using two-way ANOVA (age x tumor) or 3-way ANOVA/mixed-effects models (behavior x age x tumor), as appropriate, with Sidak’s multiple-comparisons tests. *\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.05, **\u003cem\u003eP\u003c/em\u003e\u0026lt; 0.01, ***\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.001,****\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.0001.\u003c/p\u003e","description":"","filename":"image1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8725420/v1/95671f52414e25b6dc5ed6ea.jpg"},{"id":102398208,"identity":"78f39b2e-99d6-43c4-8164-9510e1f92b52","added_by":"auto","created_at":"2026-02-11 10:21:44","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":369561,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eNeuronal and astrocytic responses during the onset of struggling during tail suspension. (a)\u003c/strong\u003e Example dual-region photometry traces showing coordination of neuronal and astrocyte activity in the mPFC and vHPC across the tail suspension session. Shaded regions indicate struggling epochs. Event heatmaps of \u003cstrong\u003e(b) \u003c/strong\u003eneuronal\u003cstrong\u003e \u003c/strong\u003eand \u003cstrong\u003e(c)\u003c/strong\u003e astrocytic activity aligned to struggling onset in the mPFC and vHPC across groups. \u003cstrong\u003e(d, g)\u003c/strong\u003e Mean neuronal (warm colors) and astrocytic (cool colors) ETA traces aligned to struggling onset in the \u003cstrong\u003e(d)\u003c/strong\u003emPFC and \u003cstrong\u003e(j) \u003c/strong\u003evHPC. \u003cstrong\u003e(e-f) \u003c/strong\u003eQuantification of neuronal ETA \u003cstrong\u003e(e)\u003c/strong\u003earea under the curve (AUC) and \u003cstrong\u003e(f) \u003c/strong\u003epeak amplitude in the mPFC; YC: n= 14 mice, YT: n=11 mice, AC: n=13 mice, AT: n=13 mice. \u003cstrong\u003e(g-h) \u003c/strong\u003eQuantification of astrocytic ETA \u003cstrong\u003e(g) \u003c/strong\u003earea under the curve (AUC) (AT v AC: **P=0.0059) and \u003cstrong\u003e(h) \u003c/strong\u003epeak amplitude (AT v AC: *P=0.0341) in the mPFC; YC: n= 14 mice, YT: n=10 mice, AC: n=12 mice, AT: n=12 mice. \u003cstrong\u003e(i) \u003c/strong\u003eQuantification of the latency between the neuronal ETA peak and astrocyte ETA peak in the mPFC (YT v YC: P=0.0511, AT v AC: *P=0.0253); YC: n= 14 mice, YT: n=10 mice, AC: n=12 mice, AT: n=12 . \u003cstrong\u003e(k-l) \u003c/strong\u003eQuantification of neuronal ETA \u003cstrong\u003e(k)\u003c/strong\u003e area under the curve (AUC) and \u003cstrong\u003e(l) \u003c/strong\u003epeak amplitude in the vHPC; YC: n= 14 mice, YT: n=11 mice, AC: n=13 mice, AT: n=13 mice. \u003cstrong\u003e(m-n) \u003c/strong\u003eQuantification of astrocytic ETA \u003cstrong\u003e(m) \u003c/strong\u003earea under the curve (AUC) (Aged v Young: *P=0.0345, Tumor v Control: *P=0.0498) and \u003cstrong\u003e(n) \u003c/strong\u003epeak amplitude (AC v YC: *P=0.0232, AT v AC: *P=0.0481) in the vHPC; YC: n= 13 mice, YT: n=15 mice, AC: n=16 mice, AT: n=16 mice. \u003cstrong\u003e(o) \u003c/strong\u003eQuantification of the latency between the neuronal ETA peak and astrocyte ETA peak in the vHPC (Aged v Young: *P=0.0355). Data are shown as mean ± SEM with individual animals overlaid. Statistics were assessed using two-way ANOVA or mixed-effects models, as appropriate, with Sidak’s multiple-comparisons tests. *\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.05, **\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.01, ***\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.001,****\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.0001.\u003c/p\u003e","description":"","filename":"image2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8725420/v1/c1def52338f8ad80ca1968c9.jpg"},{"id":102398486,"identity":"edb57ea4-d616-41c4-a907-bbf237626572","added_by":"auto","created_at":"2026-02-11 10:23:01","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":335276,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eNeuronal and astrocytic activity during elevated zero maze exploration. (a)\u003c/strong\u003eExperimental schematic illustrating dual-color fiber photometry recordings from the medial prefrontal cortex (mPFC) and ventral hippocampus (vHPC) during elevated zero maze (EZM) exploration in young and aged control (YC, AC) and tumor-bearing (YT, AT) mice. \u003cstrong\u003e(b)\u003c/strong\u003e Representative calcium traces from neurons (jRGECO1a) and \u003cstrong\u003e(c) \u003c/strong\u003eastrocytes (GCaMP8f) in the mPFC and vHPC across experimental groups during EZM exploration. \u003cstrong\u003e(d)\u003c/strong\u003e Neuronal event rate in the mPFC during closed- vs open-area exploration (left panel), and session-averaged comparison across groups (right panel; Aged v Young: F(1, 45)= 8.5,**P=0.0053); YC: n=14, YT: n=10 mice, AC: n=13 mice, AT: n=12 mice. \u003cstrong\u003e(e) \u003c/strong\u003eNeuronal area under the curve (AUC) in the mPFC comparison across groups (YT v YC: **P=0.0057, AT v AC: **P=0.0008); YC: n=14, YT: n=10 mice, AC: n=13 mice, AT: n=12 mice. \u003cstrong\u003e(f) \u003c/strong\u003eAstrocytic AUC in the mPFC during closed- and open-area exploration (left panel; AC v YC - Open Area: **P=0.0031, AC - Open v Closed Area: **P=0.0031), and session-averaged comparison across groups (right panel; YT v YC: **P=0.0049, AT v AC: **P=0.0023); YC: n=13, YT: n=11 mice, AC: n=13 mice, AT: n=11 mice. \u003cstrong\u003e(g)\u003c/strong\u003eNeuronal event rate in the vHPC during closed- and open-area exploration (left panel; YT - Open v Closed Area: #P=0.0606, AC - Open v Closed Area: ***P=0.0001, AT - Open v Closed Area: #P=0.0658), and comparison across groups (AT v AC - Open Area: *P=0.0150); YC: n= 13 mice, YT: n=15 mice, AC: n=16 mice, AT: n=14 mice. \u003cstrong\u003e(h)\u003c/strong\u003e Astrocytic AUC in the vHPC during closed- and open-area exploration (left panel; AC - Open v Closed Area: *P=0.0371), and comparison across groups (AC v YC - Open Area: **P=0.0023, AC v YC - Closed Area: P=0.0600, AT v AC - Open Area: **P=0.0076); YC: n= 13 mice, YT: n=15 mice, AC: n=16 mice, AT: n=14 mice. Data are shown as mean ± SEM with individual animals overlaid. Data are shown as mean ± SEM with individual animals overlaid. Statistics were assessed using two-way ANOVA (age x tumor) or 3-way ANOVA/mixed-effects models (behavior x age x tumor), as appropriate, with Sidak’s multiple-comparisons tests. *\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.05, **\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.01, ***\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.001,****\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.0001.\u003c/p\u003e","description":"","filename":"image3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8725420/v1/7b188a8cb15da8ee21509a5c.jpg"},{"id":102380068,"identity":"b7301760-8cb6-43fd-ae73-3d71e6f41f34","added_by":"auto","created_at":"2026-02-11 06:35:22","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":371466,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eNeuronal and astrocytic responses during open area entry in the elevated zero maze. (a)\u003c/strong\u003eExample dual-region photometry traces showing coordination of neuronal and astrocyte activity in the mPFC and vHPC across EZM exploration. Shaded regions indicate open-area epochs. Event heatmaps of \u003cstrong\u003e(b) \u003c/strong\u003eneuronal\u003cstrong\u003e \u003c/strong\u003eand \u003cstrong\u003e(c)\u003c/strong\u003e astrocytic activity aligned to open-arm entry in the mPFC and vHPC across groups. \u003cstrong\u003e(d, g)\u003c/strong\u003e Mean astrocytic ETA traces aligned to open-arm entry in the \u003cstrong\u003e(d)\u003c/strong\u003e mPFC and \u003cstrong\u003e(g) \u003c/strong\u003evHPC. \u003cstrong\u003e(e-f)\u003c/strong\u003e Quantification of astrocytic\u003cstrong\u003e (e)\u003c/strong\u003e ETA area under the curve (AUC) (AC v YC: *P=0.0241, AT v AC: **P=0.0050) and \u003cstrong\u003e(f) \u003c/strong\u003epeak amplitude (AC v YC: *P=0.0413, AT v AC: *P=0.0394) in the mPFC; YC: n= 14 mice, YT: n=11 mice, AC: n=12 mice, AT: n=9 mice. \u003cstrong\u003e(h-i)\u003c/strong\u003e Quantification of astrocytic ETA \u003cstrong\u003e(h)\u003c/strong\u003e area under the curve (AUC) (AC v AY: ***P=0.0002, AT v AC: *P=0.0189 and peak amplitude (AC v AY: **P=0.0015) in the vHPC. \u003cstrong\u003e(j-k)\u003c/strong\u003e Maximum neuron–astrocyte correlations within the \u003cstrong\u003e(j) \u003c/strong\u003emPFC (AC v AT: ****P\u0026lt;0.0001) and \u003cstrong\u003e(k) \u003c/strong\u003evHPC across groups (Tumor v Control: P=0.068). \u003cstrong\u003e(j–k)\u003c/strong\u003e Maximum cross-regional correlations between the mPFC and vHPC for \u003cstrong\u003e(j) \u003c/strong\u003eneurons (Tumor v Control: *P=0.0112) and \u003cstrong\u003e(k)\u003c/strong\u003e astrocytes (YT v YC: ***P=0.0006, AT v AC: ****P\u0026lt;0.0001). Data are shown as mean ± SEM with individual animals overlaid. Statistics were assessed using two-way ANOVA or mixed-effects models, as appropriate, with Sidak’s multiple-comparisons tests. *\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.05, **\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.01, ***\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.001,****\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.0001.\u003c/p\u003e","description":"","filename":"image4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8725420/v1/08f04103f939530e29a9464b.jpg"},{"id":102380071,"identity":"cbd57430-def0-49ef-b195-7bcf1cc3cfdc","added_by":"auto","created_at":"2026-02-11 06:35:22","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":440584,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAging reshapes tumor-associated transcriptional remodeling in the medial prefrontal cortex. (a)\u003c/strong\u003e Experimental workflow for single-nucleus RNA sequencing (snRNA-seq) of the medial prefrontal cortex (mPFC) from young control, young tumor, aged control, and aged tumor mice, and of the ventral hippocampus (vHPC) from aged control and tumor mice. \u003cstrong\u003e(b) \u003c/strong\u003eUMAP visualization of all mPFC nuclei colored by annotated cell type, including excitatory neurons, inhibitory neurons, oligodendrocytes, oligodendrocyte precursor cells (OPCs), astrocytes, microglia/perivascular macrophages (Microglia/PVM), endothelial cells, vascular leptomeningeal cells (VLMCs), and pericytes. \u003cstrong\u003e(c)\u003c/strong\u003e Proportion of major cell types in the mPFC across experimental groups, shown as stacked bar plots (Oligo- YT v YC: **P=0.0069, AT v AC: ***P=0.00031, VLMC- AT v AC: *P=0.01089, Pericyte- AT v AC: *P=0.01089, Microglia/PVM- AT v AC: *P=0.02797). \u003cstrong\u003e(d)\u003c/strong\u003eCell-type prioritization analysis using Augur, shown as area under the curve (AUC) values comparing tumor versus control conditions in young and aged mice and direct comparison of Augur AUC values between young and aged tumor mice for each cell type (Permutation Test - All P-Values: **P=0.00257). \u003cstrong\u003e(e)\u003c/strong\u003e Upset plots of differentially expressed genes in excitatory neurons from young and aged tumor-bearing mice relative to age-matched controls, shown as counts of upregulated and downregulated genes. \u003cstrong\u003e(f–g)\u003c/strong\u003e Gene set enrichment analysis (GSEA) of Gene Ontology (GO) Biological Process terms in excitatory neurons from \u003cstrong\u003e(f) \u003c/strong\u003eyoung tumor and \u003cstrong\u003e(g) \u003c/strong\u003eaged tumor mice, showing activated and suppressed pathways. \u003cstrong\u003e(h)\u003c/strong\u003e Upset plots of differentially expressed genes in inhibitory neurons from young and aged tumor mice relative to age-matched controls. \u003cstrong\u003e(m)\u003c/strong\u003e GSEA GO Biological Process analysis of inhibitory neurons from \u003cstrong\u003e(i) \u003c/strong\u003eyoung tumor and \u003cstrong\u003e(j) \u003c/strong\u003eaged tumor mice. \u003cstrong\u003e(k)\u003c/strong\u003eUpset plots of differentially expressed genes in astrocytes from young and aged tumor mice relative to age-matched controls. \u003cstrong\u003e(l–m) \u003c/strong\u003eGSEA GO Biological Process analysis of astrocytes from \u003cstrong\u003e(l) \u003c/strong\u003eyoung tumor and \u003cstrong\u003e(m)\u003c/strong\u003e aged tumor mice. Dot size in GSEA plots reflects gene count per pathway; color indicates adjusted \u003cem\u003eP\u003c/em\u003e value for activated or suppressed pathways, as indicated.\u003c/p\u003e","description":"","filename":"image5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8725420/v1/55035cd75d65adfd54b1362a.jpg"},{"id":102398064,"identity":"e51a4f3a-ed5c-4852-babf-c9d21f78c5e6","added_by":"auto","created_at":"2026-02-11 10:20:50","extension":"jpg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":424002,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTumor-associated transcriptional remodeling in the aged ventral hippocampus. (a) \u003c/strong\u003eUMAP visualization of single-nucleus RNA sequencing data from the ventral hippocampus (vHPC) of aged control and aged tumor-bearing mice, colored by annotated cell type. Cells originating from adjacent cortex were identified and excluded from downstream analyses. \u003cstrong\u003e(b)\u003c/strong\u003e Proportion of major cell types in the vHPC from aged control and aged tumor-bearing mice, shown as stacked bar plots\u003cstrong\u003e. (c)\u003c/strong\u003e Augur prioritization analysis showing distributions of ROC AUC values for each cell type comparing aged tumor-bearing versus aged control mice. Dashed line indicates chance-level classification performance (AUC = 0.5) (OPC: **P=0.00154, vHPC Excitatory: **P=0.00154, non-vHPC Excitatory: **P=0.00426, Astrocytes: **P=0.00545, CR Neuron: **P=0.00545, VLMC: *P=0.03667). \u003cstrong\u003e(d)\u003c/strong\u003eNumber of differentially expressed genes in excitatory neurons from aged tumor-bearing mice relative to aged controls, separated by upregulated and downregulated genes. \u003cstrong\u003e(e)\u003c/strong\u003e Gene set enrichment analysis (GSEA) of Gene Ontology (GO) Biological Process terms in excitatory neurons from aged tumor-bearing mice, showing activated and suppressed pathways. \u003cstrong\u003e(f)\u003c/strong\u003eNumber of differentially expressed genes in inhibitory neurons from aged tumor-bearing mice relative to aged controls. \u003cstrong\u003e(g)\u003c/strong\u003e GSEA GO Biological Process analysis for inhibitory neurons from aged tumor-bearing mice. \u003cstrong\u003e(h)\u003c/strong\u003eNumber of differentially expressed genes in astrocytes from aged tumor-bearing mice relative to aged controls. \u003cstrong\u003e(i)\u003c/strong\u003e GSEA GO Biological Process analysis for astrocytes from aged tumor-bearing mice. \u003cstrong\u003e(j)\u003c/strong\u003e Heatmap showing differential expression of genes associated with metabolic processes in astrocytes from aged tumor-bearing mice, displayed as log fold change (logFC). \u003cstrong\u003e(k)\u003c/strong\u003eHeatmap of differentially expressed genes associated with cognition-related pathways across excitatory neurons, inhibitory neurons, and astrocytes in the vHPC, displayed as log fold change (logFC). Dot size in GSEA plots reflects gene count per pathway; color indicates adjusted \u003cem\u003eP\u003c/em\u003e value for activated or suppressed pathways, as indicated. For heatmaps in \u003cstrong\u003e(j)\u003c/strong\u003e and \u003cstrong\u003e(k)\u003c/strong\u003e, asterisks denote statistically significant changes at false discovery rate (FDR) \u0026lt; 0.05.\u003c/p\u003e","description":"","filename":"image6.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8725420/v1/cc6e547a19cacd2518a6d350.jpg"},{"id":106401607,"identity":"889638f6-6b4c-4470-8cb0-347a9faaab6a","added_by":"auto","created_at":"2026-04-08 09:08:02","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3775502,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8725420/v1/42ecd198-cb26-45d9-a863-14a1fd7f64dc.pdf"},{"id":102398506,"identity":"7656393d-3781-444e-b983-0b9e0c693085","added_by":"auto","created_at":"2026-02-11 10:23:06","extension":"xlsx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":15827,"visible":true,"origin":"","legend":"FC Stats","description":"","filename":"FCFigStats.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8725420/v1/4231639095592becff9fcbc1.xlsx"},{"id":102380063,"identity":"b1475ba9-2841-4514-a240-29de25485bcb","added_by":"auto","created_at":"2026-02-11 06:35:22","extension":"xlsx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":15683,"visible":true,"origin":"","legend":"Counts Stats","description":"","filename":"CountsFigStats.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8725420/v1/f1abdfb6b767f5c0f069f0ba.xlsx"},{"id":102398135,"identity":"a245cc6e-dc36-43a8-b685-0183763d7512","added_by":"auto","created_at":"2026-02-11 10:21:10","extension":"xlsx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":26663,"visible":true,"origin":"","legend":"EZM Stats","description":"","filename":"EZMFigStats.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8725420/v1/9c275fd74bdd495ef50c11b7.xlsx"},{"id":102380065,"identity":"1667aa40-67fa-4573-819f-1b4c788e6187","added_by":"auto","created_at":"2026-02-11 06:35:22","extension":"xlsx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":26261,"visible":true,"origin":"","legend":"HC Stats","description":"","filename":"HCFigStats.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8725420/v1/178f341f88db5ba21cbc3d10.xlsx"},{"id":102398494,"identity":"8939900a-2da4-4c6e-aa60-400d718d25d5","added_by":"auto","created_at":"2026-02-11 10:23:03","extension":"xlsx","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":403773,"visible":true,"origin":"","legend":"DEGs","description":"","filename":"DEGsFig.56.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8725420/v1/f47929559fd850fa99d15de0.xlsx"},{"id":102398472,"identity":"fb069ce0-2745-464b-b91a-9b3ac8d9898d","added_by":"auto","created_at":"2026-02-11 10:22:57","extension":"xlsx","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":15740,"visible":true,"origin":"","legend":"Behavior Stats","description":"","filename":"BehaviorFigStats.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8725420/v1/77aba8f2dffe08b9e2561781.xlsx"},{"id":102380066,"identity":"22f1cc6d-a94e-49b9-8a54-2ff5ac03b896","added_by":"auto","created_at":"2026-02-11 06:35:22","extension":"xlsx","order_by":7,"title":"","display":"","copyAsset":false,"role":"supplement","size":36219,"visible":true,"origin":"","legend":"TST Stats","description":"","filename":"TSTFigStats.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8725420/v1/60f6e9535d65341725c057df.xlsx"},{"id":102380069,"identity":"c7a12409-138e-4ea7-89a3-93722c102ffb","added_by":"auto","created_at":"2026-02-11 06:35:22","extension":"xlsx","order_by":8,"title":"","display":"","copyAsset":false,"role":"supplement","size":119574,"visible":true,"origin":"","legend":"GO results","description":"","filename":"GSEAGOBPResults.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8725420/v1/f3dbe94d8fb2569275f70c4d.xlsx"},{"id":102380073,"identity":"7ab73151-9c0d-4637-9e93-736c538e2b5d","added_by":"auto","created_at":"2026-02-11 06:35:22","extension":"docx","order_by":9,"title":"","display":"","copyAsset":false,"role":"supplement","size":3926948,"visible":true,"origin":"","legend":"","description":"","filename":"ExtendedDataFigs.docx","url":"https://assets-eu.researchsquare.com/files/rs-8725420/v1/f1ec1a6b90a9d86458e1f6c8.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Aging potentiates glioma-driven remodeling of cortico-hippocampal neuron-astrocyte dynamics","fulltext":[{"header":"Introduction","content":"\u003cp\u003eGlioblastoma (GBM), is the most common malignant primary tumor of the central nervous system and remains among the most lethal human cancers\u003csup\u003e1\u003c/sup\u003e. Incidence increases sharply with age, and clinical outcome worsens steadily in older patients despite aggressive multimodal therapy\u003csup\u003e1,2\u003c/sup\u003e. Although survival has improved modestly in recent decades, these gains are limited, particularly in the elderly, underscoring that age is not merely a demographic correlate of GBM but a central biological determinant of disease course\u003csup\u003e3,4\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eThis age dependence is increasingly recognized as a property of the host brain rather than the tumor alone. When identical tumors are introduced into young and aged mice, the survival disadvantage of aged hosts is preserved, closely mirroring clinical observations, indicating that tumor progression and treatment response are shaped by age-dependent remodeling of the neural and immune environment\u003csup\u003e2\u003c/sup\u003e. Yet most preclinical GBM studies have historically relied on young adult mice, leaving a major gap in understanding how the aged brain\u0026rsquo;s baseline state alters the functional impact of tumor burden.\u003c/p\u003e\n\u003cp\u003eA critical unanswered question is what these age-dependent effects mean at the multi-circuit level across cell types. Beyond survival, glioblastoma patients frequently experience cognitive, affective, and functional decline, and these impairments are particularly consequential in older individuals, shaping treatment tolerance, quality of life, and clinical decision-making\u003csup\u003e5\u0026ndash;7\u003c/sup\u003e. Normal aging is accompanied by functional changes in both neurons and glial cells, including altered excitability, synaptic regulation, neuroinflammatory tone, and glial reactivity, and these changes are not uniform across the brain\u003csup\u003e8,9\u003c/sup\u003e. The prefrontal cortex and hippocampus are particularly vulnerable to aging-related remodeling, and they support stress regulation, affective state, and cognition\u003csup\u003e10\u003c/sup\u003e, domains frequently disrupted in GBM patients\u003csup\u003e11,12\u003c/sup\u003e. Thus, aging may alter tumor outcomes not only by shaping immune responses but also by changing how neural circuits and glial cells respond to tumor-associated perturbation.\u003c/p\u003e\n\u003cp\u003eAstrocytes are well positioned to mediate such age-dependent vulnerability. They couple neuronal activity to vascular and metabolic support, integrate inflammatory and neuromodulatory signals, and dynamically regulate circuit excitability\u003csup\u003e13\u0026ndash;15\u003c/sup\u003e. In aging and disease, astrocytes undergo reactive remodeling that can alter their interactions with neurons and reshape state-dependent circuit function\u003csup\u003e9,16\u003c/sup\u003e. Consistent with this framework, we previously found that tumor burden induces astrogliosis that is amplified in aged mice\u003csup\u003e17\u003c/sup\u003e, suggesting that aging heightens astrocyte reactivity to glioma-associated cues. Whether this heightened reactivity translates into altered astrocyte activity dynamics during behavior, and whether aging fundamentally changes neuron\u0026ndash;astrocyte coordination under tumor burden, remains unknown.\u003c/p\u003e\n\u003cp\u003eHere, we tested the hypothesis that aging interacts with tumor burden to reconfigure neuron\u0026ndash;astrocyte dynamics in vivo within circuits vulnerable to aging and relevant to behavioral state regulation. Using dual-color fiber photometry, we simultaneously recorded neuronal and astrocytic calcium activity in the medial prefrontal cortex (mPFC) and ventral hippocampus \u0026nbsp;(vHPC) during baseline conditions and across paradigms that reliably engage stress- and arousal-related states, including tail suspension\u003csup\u003e18,19\u003c/sup\u003e, elevated zero maze\u003csup\u003e20,21\u003c/sup\u003e, and fear conditioning\u003csup\u003e19,22\u003c/sup\u003e. Importantly, these paradigms are not used to model psychiatric disease per se but to evoke reproducible behavioral states that enable quantification of dynamic neuron\u0026ndash;astrocyte interactions during heightened circuit engagement. Given the high prevalence of affective symptoms in GBM patients\u003csup\u003e11\u003c/sup\u003e, probing these dynamics in stress-related contexts may also provide insight into circuit disruptions relevant to clinical neuropsychiatric dysfunction. Finally, to connect these functional phenotypes to underlying molecular mechanisms, we used single-nucleus RNA sequencing to determine how tumor-associated transcriptional programs in these regions differ between young and aged hosts.\u003c/p\u003e\n\u003cp\u003eTogether, our work provides, to our knowledge, one of the first \u003cem\u003ein vivo\u003c/em\u003e demonstrations that normal aging reshapes neuron\u0026ndash;astrocyte coordination and reveals how aging then interacts with tumor burden to disrupt extratumoral circuit dynamics, identifying circuit-level vulnerabilities that may inform function-preserving strategies for older patients.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eTo define how aging alters the functional impact of glioma burden on extratumoral brain circuits, we examined neuronal and astrocytic activity dynamics in the medial prefrontal cortex (mPFC) and ventral hippocampus (vHPC), regions central to stress and anxiety regulation and vulnerable to aging. We recorded neuronal activity using jRGECO1a and astrocytic activity using membrane-bound GCaMP8f, enabling cell-type\u0026ndash;specific resolution within the same circuits (Fig. 1a). By recording from the hemisphere contralateral to tumor implantation, we isolated how tumor burden reshapes circuit dynamics beyond the tumor core, with a focus on neuron\u0026ndash;astrocyte coordination.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eHistological analysis confirmed robust and comparable and separable expression of jRGECO1a in neurons and GCaMP8f in astrocytes within the mPFC and vHPC across age and tumor burden (Extended Data Fig. 1a-c, f). Quantification of jRGECO1a-positive neurons per region of interest revealed no significant effects of age or tumor burden in either region (Extended Data Fig. 1d,g). Similarly, astrocytic GCaMP8f coverage, quantified as labeled area per region of interest, did not differ significantly by age or tumor status in either region (Extended Data Fig. 1e,h). These data indicate that the functional differences observed in subsequent analyses are unlikely to reflect changes in reporter expression or gross cellular representation.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eBaseline neuron\u0026ndash;astrocyte dynamics during resting state shows region-specific vulnerability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo determine whether tumor burden alters circuit function in the absence of overt behavioral challenge, we examined spontaneous neuronal and astrocytic calcium dynamics during home cage behavior at early and later time points following tumor implantation (Extended Data Fig. 2). In the mPFC, tumor-bearing mice showed a progressive reduction in neuronal event rates over time, accompanied by age- and tumor-dependent reductions in astrocytic activity and coordination that were most apparent at the later time point. In contrast, baseline activity and local coordination in the ventral hippocampus were comparatively stable, with age-related but not tumor-associated differences. Specifically, aged mice across both days had higher neuronal rates than young mice, suggesting baseline neuronal hyperactivity in the aged hippocampus. Together, these baseline recordings suggest that tumor burden produces subtle, progressive disruptions in mPFC circuit dynamics that may become more pronounced when circuits are engaged by the behavioral demands of homecage exploration.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAging and tumor burden alter stress-evoked neuronal and astrocytic activity\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe next asked whether these baseline vulnerabilities are amplified during a potent, externally imposed stressor that recruits prefrontal\u0026ndash;hippocampal circuits. During the tail suspension test, we observed clear, behaviorally structured calcium dynamics in both neurons and astrocytes across regions (Fig. 1a\u0026ndash;b). Representative dual-color photometry traces illustrate robust neuronal activity fluctuations during struggling epochs, accompanied by slower, larger-amplitude astrocytic calcium events that were temporally coupled to behavioral transitions (Fig. 1b).\u003c/p\u003e\n\u003cp\u003eWe then quantified neuronal activity across the full session. In the mPFC, neuronal event rates were consistently higher during struggling than immobility across age and tumor groups (Behavior Effect: F(1,46)=120.6; ****P\u0026lt;0.0001), indicating preserved state-dependent encoding of coping behavior (Fig. 1c). Tumor-bearing mice exhibited a modest increase in overall neuronal event rate independent of behavioral state, while neuronal response magnitude was not significantly altered (Fig. 1c). We next examined astrocytic response magnitude across behavioral states. In the mPFC, astrocytic area under the curve (AUC) increased during struggling in control animals (Behavior Effect: F(1,87)=47.11; ****P\u0026lt;0.0001), with aged controls exhibiting larger responses than young controls (Fig. 1d). This increase was attenuated in young tumor-bearing mice and abolished in aged tumor-bearing mice (Fig. 1d).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn the vHPC, neuronal event rates similarly differed between struggling and immobility in young control mice (Behavior Effect: F(1,56)=20.28; ****P\u0026lt;0.0001); however, this behavioral discrimination was reduced with aging (Fig. 1e). Tumor-bearing mice largely resembled age-matched controls, indicating that changes in vHPC neuronal state encoding were driven primarily by aging rather than tumor burden (Fig. 1e). In astrocytes, responses, like the mPFC, increased during struggling (Behavior Effect: F(1,52)=109.1; ****P\u0026lt;0.0001), with aged controls again showing larger event magnitudes overall (Fig. 1f). Likewise, this age-associated amplification was reduced in aged tumor-bearing mice (Fig. 1f).\u003c/p\u003e\n\u003cp\u003eTo determine whether these changes in event magnitude were accompanied by altered circuit coordination, we next examined correlations within and across regions. Neuron\u0026ndash;neuron correlations between the mPFC and vHPC showed modest age-related reductions but were not strongly influenced by tumor burden (Fig. 1g). In contrast, astrocyte\u0026ndash;astrocyte correlations across regions were significantly reduced in tumor-bearing mice, with the largest reductions observed in aged animals (Fig. 1h). Within regions, local neuron\u0026ndash;astrocyte coupling in the mPFC was reduced in tumor-bearing mice, driven primarily by a near-significant decrease in aged tumor-bearing animals (Fig. 1i). In the vHPC, neuron\u0026ndash;astrocyte correlations also showed a significant effect of tumor burden, although this effect was not attributable to a single age group (Fig. 1j). Together, these findings indicate that tumor burden preferentially disrupts astrocytic coordination locally and across regions, with aging amplifying these effects.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAging and tumor burden alter the temporal coordination of neuron\u0026ndash;astrocyte responses during stress\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eGiven the strong dependence of cellular responses on behavioral state, we next assessed the temporal organization of neuronal and astrocytic activity aligned to the transition from immobility to struggling. Representative traces (Fig. 2a) and event heatmaps (Fig. 2b-c, Extended Fig. 3) revealed a conserved response sequence to struggling onset across regions, characterized by rapid neuronal activation followed by delayed astrocytic responses (Fig. 2a\u0026ndash;c). This overall structure was preserved across age and tumor burden at the population level.\u003c/p\u003e\n\u003cp\u003eEvent-triggered average analyses confirmed that neuronal response magnitude and peak amplitude were not significantly altered by age or tumor burden in either region (Fig. 2e-f, k-l). In contrast, astrocytic ETAs revealed pronounced group differences. In the mPFC, astrocytic ETA AUC and peak amplitude were selectively reduced in aged tumor-bearing mice (Fig. 2g-h). In the vHPC, astrocytic ETA magnitude showed an age-associated increase under control conditions, with similar blunting of this response in aged tumor animals (Fig. 2m-n). In young animals, no effect was observed with tumor burden. Together, these findings indicate that tumor burden selectively erodes the astrocytic response to struggling onset in aged mice despite preserved neuronal activation.\u003c/p\u003e\n\u003cp\u003eGiven the close relationship between the heightened neuronal activity and subsequent astrocyte response, and the known importance of this astrocytic response for maintenance of neuronal circuit function, we quantified the response timing latency between neuron and astrocyte responses. In the mPFC, neuron\u0026ndash;astrocyte peak latency differed across groups, with tumor-bearing mice showing altered delays relative to controls and aged tumor-bearing mice exhibiting the longest delays overall (Fig. 2i). In the vHPC, neuron\u0026ndash;astrocyte peak latency was influenced primarily by age rather than tumor status (Fig. 2o). Together, these analyses demonstrate age- and tumor-dependent alterations in the timing and magnitude of astrocytic responses to elevated neuronal activity during stress.\u003c/p\u003e\n\u003cp\u003eNotably, these circuit-level alterations were accompanied by age-specific shifts in stress coping behavior (Extended Fig. 4a-b). Young tumor-bearing mice showed no differences in immobility or latency to immobility, whereas aged tumor-bearing mice exhibited increased immobility and reduced latency to passive coping. The selective convergence of astrocytic dysfunction and behavioral impairment in aged mice supports a link between impaired astrocyte engagement and maladaptive stress responses under tumor burden.\u003c/p\u003e\n\u003ch3\u003e\u003cstrong\u003eAging and tumor burden alter neuron\u0026ndash;astrocyte engagement during anxiogenic exploration\u003c/strong\u003e\u003c/h3\u003e\n\u003cp\u003eRecent studies have demonstrated a central role for astrocytes in shaping anxiety-related behavior\u003csup\u003e23\u0026ndash;25\u003c/sup\u003e. In contrast to externally imposed stressors such as tail suspension, exploration of anxiogenic environments reflects a self-directed, internally modulated behavioral state in which astrocytic signaling plays a prominent role in regulating circuit activity\u003csup\u003e23\u003c/sup\u003e. Given the pronounced astrocytic dysregulation we observed during tail suspension, we next asked whether similar circuit vulnerabilities emerge during anxiety-related behavior by recording activity during exploration of the elevated zero maze (EZM) (Fig. 3a). During EZM exploration, we again observed structured calcium dynamics in both neurons and astrocytes across regions. Representative traces illustrate relatively rapid, high-frequency neuronal activity during exploration alongside slower astrocytic calcium transients, with clear differences across age and tumor burden (Fig. 3b-c).\u003c/p\u003e\n\u003cp\u003eWe first quantified neuronal activity across EZM areas. In the mPFC, neuronal event rates were overall higher during open-area exploration than during closed-area exploration across all groups (Behavior Effect: F(1,33)=8.965; **P=0.0045), indicating preserved, albeit modest, behavioral state modulation during anxiogenic exploration (Fig. 3d). Overall neuronal rates were generally higher in young mice, but maintained across tumor conditions. However, when neuronal response magnitude was quantified using area under the curve (AUC), both young and aged tumor-bearing mice exhibited reduced overall neuronal AUC relative to age-matched controls, independent of maze area (Fig. 3e). Thus, while neuronal rates were preserved, tumor burden was associated with a global reduction in mPFC neuronal response magnitude rather than a behavior-specific effect.\u003c/p\u003e\n\u003cp\u003eAstrocytic activity in the mPFC showed a distinct pattern. Astrocytic response magnitude generally increased, but exhibited a strong age dependence (Behavior x Age: F(1,39)=7.272; *P=0.0103): aged control mice displayed markedly larger astrocytic AUCs in the open area than young controls (Fig. 3f). This age-associated amplification was absent in tumor-bearing mice. Instead, both young and aged tumor-bearing animals showed reduced astrocytic AUCs relative to controls, independent of maze area (Fig. 3f), indicating a generalized suppression of astrocytic response magnitude.\u003c/p\u003e\n\u003cp\u003eIn the ventral hippocampus, across groups, neuronal event rates also distinguish open from closed areas (Behavior Effect: F(1,53)=30.87; ****P\u0026lt;0.0001) (Fig. 3g). Here, tumor-associated effects emerged selectively with aging: aged tumor-bearing mice showed reduced neuronal activity during open-area exploration, whereas closed-area activity remained comparable to controls (Fig. 3g). No tumor-associated differences were observed in young mice, indicating selective attenuation of anxiogenic neuronal engagement in the aged hippocampus.\u003c/p\u003e\n\u003cp\u003eAstrocytic responses in the vHPC trended towards being increased during open-area exploration (Behavior Effect: F(1,52)=2.949; P=0.0919), although mostly driven by a significant increase in aged control animals (Fig. 3h). As in the mPFC, aged control mice exhibited larger astrocytic AUCs than young controls, reflecting age-associated amplification of astrocytic recruitment during anxiogenic exploration (Fig. 3h). Tumor burden selectively blunted this enhancement, with aged tumor-bearing mice showing reduced astrocytic AUCs relative to aged controls, while astrocytic responses in young mice were less affected (Fig. 3h). Thus, tumor effects on astrocytic recruitment extended beyond the tumor-adjacent cortex but were most pronounced with aging.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eOpen-area entry\u0026ndash;aligned responses reveal selective loss of astrocytic recruitment\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eExploratory behavior in the elevated zero maze introduces a more complex set of behavioral variables than the tail suspension, such as rearing and risk assessment without actual engagement of open areas\u003csup\u003e26\u003c/sup\u003e. Therefore, the peaks based method of assessing open-closed area dissociation did not yield as clear results as tail suspension. However, as shown in the representative trace (Fig. 4a), astrocytic events are especially tied to the entry into the open area, just not specific for this transition. Event-aligned heatmaps further illustrated this, as well as group differences in the structure and consistency of neuronal and astrocytic responses surrounding open-area entry in both the mPFC and vHPC (Fig. 4b-c), motivating quantitative analyses of event-triggered response magnitude.\u003c/p\u003e\n\u003cp\u003eNeuronal event-triggered responses were detectable but small in both regions (Fig. 4b-c, Extended Fig. 5a,d). In the mPFC, neuronal ETA AUC and amplitude were low overall and were generally larger in aged mice, but did not differ significantly by tumor burden (Extended Fig. 5a-c). However, qualitatively, the small, organized increase in control mice lost consistency in both tumor groups (Fig. 4b, Extended Fig. 5a,g). Similar to the mPFC, neuronal responses in the vHPC were larger in aged animals (Extended Fig. 5e,f) . In addition, aged tumor mice had much greater variability between events and across mice, and a reduced AUC during open area entry (Fig. 4b, Extended Fig. 5d,e,i), suggesting a loss of stereotyped responses to anxiogenic stimuli selectively in aged tumor mice.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eMost notably in the EZM, however, was that astrocytic ETAs were substantially larger than neuronal responses in both regions (Fig. 4b-g, Extended Fig. 5h,j). In the mPFC, astrocytic ETA responses at open-area entry were amplified with age in control animals (Fig. 4d-e, Extended Fig. 5h). Tumor burden shifted astrocytic ETA magnitude downward, with this reduction most evident, and statistically significant, in aged tumor-bearing mice relative to aged controls, while young tumor-bearing mice showed a similar directional decrease that did not reach significance (Fig. 4d-f). A similar pattern was observed in the vHPC, where astrocytic ETA AUCs increased with age under control conditions and were selectively attenuated by tumor burden exclusively in aged animals (Fig. 4g-i, Extended Fig. 5j). This reduction was not driven by decreased peak amplitude (Fig. 4i), as observed in the mPFC, but instead reflected shorter-duration astrocytic responses, as evident from the average event-triggered traces (Fig. 4g).\u003c/p\u003e\n\u003cp\u003eWe next assessed coordination across cells and regions during EZM exploration. Local neuron\u0026ndash;astrocyte correlations in the mPFC were reduced in tumor-bearing mice, with the strongest reductions observed in aged animals (Fig. 4j). In the vHPC, neuron\u0026ndash;astrocyte correlations also showed a significant effect of tumor burden, although this effect was not driven by a specific age group (Fig. 4k). Cross-regional coordination revealed a more selective disruption: neuron\u0026ndash;neuron correlations between the mPFC and vHPC showed only modest reductions (Fig. 4l), whereas astrocyte\u0026ndash;astrocyte correlations across regions were robustly reduced in tumor-bearing mice of both ages (Fig. 4m), identifying long-range astrocytic synchrony as a major locus of disruption during anxiogenic exploration.\u003c/p\u003e\n\u003cp\u003eTogether, these results indicate that during anxiogenic exploration, astrocytic recruitment is amplified with aging and selectively compromised by tumor burden, whereas neuronal responses are comparatively weak and show more limited, region-specific effects. Consistent with these alterations in cellular dynamics, tumor-bearing mice, particularly aged animals, spent more time in the open areas (Extended Fig. 4c) and exhibited reduced risk-assessment behavior (Extended Fig. 4d), linking age-dependent astrocytic dysfunction to altered anxiety-related behavior under tumor burden.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAlteration to neuronal and astrocytic responses to foot shock with tumor burden\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eElevated zero maze exploration revealed marked disruption of astrocytic recruitment in the mPFC across tumor-bearing mice and a selective reduction of astrocytic response magnitude in the vHPC of aged tumor-bearing animals. To determine whether these effects reflected an inability of these circuits to generate large responses or a more specific failure of state-dependent recruitment, we next examined neuronal and astrocytic activity during foot shock, a stimulus known to elicit strong, synchronized activation across limbic circuits\u003csup\u003e19\u003c/sup\u003e. On the same day as EZM testing, mice were subjected to brief foot shocks (Extended Fig. 6). Shock-elicited event-triggered averages revealed tightly time-locked neuronal activation at shock onset, followed by delayed astrocytic responses with prolonged decay in both regions (Extended Fig. 6a\u0026ndash;h), consistent with canonical neuron\u0026ndash;astrocyte response sequences.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eQuantification of neuronal response magnitude revealed a significant effect of tumor burden in the medial prefrontal cortex. In both young and aged mice, tumor-bearing animals exhibited reduced neuronal ETA area under the curve (AUC) relative to age-matched controls, indicating attenuated prefrontal neuronal recruitment in response to shock (Extended Fig. 6i). Astrocytic ETA AUCs in the mPFC were likewise significantly reduced in tumor-bearing mice across both age groups (Extended Fig. 8j). In contrast, neuronal and astrocytic ETA AUCs in the ventral hippocampus were not significantly altered by age or tumor status (Extended Fig. 6k\u0026ndash;l), indicating preserved hippocampal responsiveness to this highly salient aversive stimulus.\u003c/p\u003e\n\u003cp\u003eThese data raise the possibility that reduced cellular responses in the mPFC during elevated zero maze exploration reflect diminished response capacity with advancing tumor burden, whereas reduced responses in the ventral hippocampus of aged tumor-bearing mice may reflect altered recruitment during anxiogenic exploration.\u003c/p\u003e\n\u003ch3\u003e\u003cstrong\u003eAging reveals transcriptional reprogramming of extratumoral circuits in glioma\u003c/strong\u003e\u003c/h3\u003e\n\u003cp\u003eThe circuit-level alterations observed during stress and anxiogenic behavior prompted us to ask whether aging similarly reshapes the transcriptional landscape of extratumoral brain regions under tumor burden. To address this, we performed single-nucleus RNA sequencing of the medial prefrontal cortex (mPFC) across all experimental groups and of the ventral hippocampus (vHPC) from aged control and tumor-bearing mice (Fig. 5a), reflecting the regions and age groups in which functional alterations were most pronounced. This approach captured all major neural, glial, and vascular-associated cell types (Fig. 5b, Extended Fig. 7), enabling systematic assessment of how glioma and aging interact to remodel cellular composition and gene expression beyond the tumor core.\u003c/p\u003e\n\u003cp\u003eAnalysis of cell-type proportions in the mPFC revealed tumor-associated changes that differed by age (Fig. 5c). Oligodendrocytes were reduced in both young and aged tumor-bearing mice. Microglia/perivascular macrophages increased with tumor burden, showing a trend toward expansion in young mice and a robust increase in aged animals, consistent with age-amplified inflammatory activation in cortex proximal to the glioma. In contrast, vascular leptomeningeal cells and pericytes, cell populations critical for blood\u0026ndash;brain barrier integrity, were selectively reduced in aged tumor-bearing mice, indicating age-dependent vulnerability of the neurovascular compartment.\u003c/p\u003e\n\u003cp\u003eTo quantify the extent of transcriptional disruption across cell types, we next applied Augur prioritization analysis. In aged tumor-bearing mice, all captured cell types exhibited significant deviation from baseline classification performance, indicating widespread transcriptional remodeling (Extended Fig. 7b). In contrast, perturbation in young tumor-bearing mice was still broad but slightly more restricted: all cell types except oligodendrocyte precursor cells and vascular leptomeningeal cells showed significant prioritization (Extended Fig. 7b). Direct comparison of perturbation magnitude further revealed cell-type\u0026ndash;specific age biases. Oligodendrocytes, astrocytes, excitatory neurons, endothelial cells, and vascular leptomeningeal cells were more strongly perturbed in aged tumor-bearing mice, whereas inhibitory neurons showed greater perturbation in young tumor-bearing mice (Fig. 5d). Microglia and pericytes exhibited comparable perturbation across ages, suggesting shared tumor-associated transcriptional responses independent of aging (Fig. 5d).\u003c/p\u003e\n\u003cp\u003eGiven the selective loss of vascular-associated cell types in aged tumor-bearing mice, the central role of astrocytes in neurovascular and metabolic coupling, and the pronounced neuron\u0026ndash;astrocyte dysfunction observed in our \u003cem\u003ein vivo\u003c/em\u003e recordings, we focused subsequent analyses on neurons and astrocytes as potential integrators of age- and tumor-related circuit vulnerability.\u003c/p\u003e\n\u003cp\u003eDifferential expression analyses revealed a pronounced age dependence in both the magnitude and direction of transcriptional change. Across cell types, aged tumor-bearing mice exhibited substantially greater transcriptional downregulation than young tumor-bearing mice (Fig. 5e,h,k). Neurons, particularly excitatory neurons, had downregulated pathways in both age groups predominantly related to synaptic structure and activity (Fig. 5e-g, Extended Fig. 8a). This downregulation was especially pronounced in excitatory neurons of aged tumor mice, with genes such as \u003cem\u003eSlitrk1\u003c/em\u003e and \u003cem\u003eHomer1\u003c/em\u003e, genes essential for synapse structure, significantly downregulated (Extended Fig. 8a,d-e). In contrast, genes upregulated across cell types in both young and aged tumor-bearing mice were consistently enriched for inflammatory and stress-related pathways, indicating a shared tumor-associated inflammatory program that was preserved across age (Fig. 5e-m). In neurons, these pathways included responses to interferon beta, and upregulation of genes such as \u003cem\u003eStat1\u003c/em\u003e in both age groups, and \u003cem\u003eStat2\u0026nbsp;\u003c/em\u003ein aged tumor mice\u003cem\u003e\u0026nbsp;\u003c/em\u003e(Fig. 5f-g, i-j, Extended Fig. 8b,f). However, this was also associated with downregulation of genes associated with lymphocyte activity (Extended Fig. 8c), especially in aged tumor mice, suggesting a regulatory response by neurons to cytotoxicity.\u003c/p\u003e\n\u003cp\u003eAstrocytes were particularly notable for the extent and specificity of transcriptional suppression in aged tumor-bearing mice. In young tumor-bearing animals, relatively few astrocytic pathways were downregulated, and these were primarily metabolic in nature (Fig. 5i-j). In aged tumor-bearing mice, however, astrocytes exhibited broad suppression of pathways involved in central nervous system differentiation, synaptic support, and neuronal signaling (Fig. 5m, Extended Fig. 9a-b). Genes implicated in astrocyte\u0026ndash;neuron communication and synapse stabilization, including \u003cem\u003eSparcl1\u003c/em\u003e and \u003cem\u003eMegf10\u003c/em\u003e, were among the most strongly suppressed (Extended Fig. 9b,f-g). \u0026nbsp;Upregulated pathways were more universal across age groups, and included canonical markers of astrocyte reactivity, such as \u003cem\u003eVim\u003c/em\u003e and \u003cem\u003eGfap\u003c/em\u003e (Extended Fig. 9c,e) .\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe selective loss of vascular-associated cell populations in aged tumor-bearing mice raised the possibility that astrocytes may be responding transcriptionally to compromised blood\u0026ndash;brain barrier function. To test this directly, we examined curated gene sets derived from prior studies characterizing astrocytic responses to blood\u0026ndash;brain barrier disruption. Astrocytes from both tumor-bearing groups, and most prominently from aged tumor-bearing mice, showed substantial overlap with established BBB disruption signatures (Extended Fig. 9d). Of particular interest was the downregulation of \u003cem\u003eGlud1\u003c/em\u003e in aged tumor astrocytes (Extended Fig. 9h), a gene central to astrocytic glutamate metabolism and recycling at the synapse. Reduced expression of this enzyme would be expected to impair astrocytic regulation of extracellular glutamate and neuronal excitability.\u003c/p\u003e\n\u003ch3\u003e\u003cstrong\u003eAging reveals transcriptional perturbation in circuits distal to tumor burden\u003c/strong\u003e\u003c/h3\u003e\n\u003cp\u003eTo determine whether tumor-associated transcriptional changes extend to regions anatomically distant from the tumor site, we performed single-nucleus RNA sequencing of the ventral hippocampus from aged tumor-bearing and age-matched control mice (Fig. 6, Extended Fig. 10). We captured all cell types from the vHPC, as well as cells from adjacent cortex, which were excluded from further analyses (Fig. 6a, Extended Fig. 10a-b). In contrast to the mPFC, cell-type proportions in the vHPC were preserved, with no detectable differences between tumor and control conditions (Fig. 6b, Extended Fig. 10c), indicating that gross cellular composition in this region remains stable under tumor burden.\u003c/p\u003e\n\u003cp\u003eDespite this compositional stability, transcriptional perturbation was widespread. Augur analysis revealed significant prioritization across all major cell populations except pericytes and endothelial cells, demonstrating that tumor burden elicits broad molecular responses in the aged hippocampus even in the absence of overt changes in cell abundance (Fig. 6c). Among neuronal populations, excitatory neurons exhibited the largest number of differentially expressed genes, with the majority showing downregulation (Fig. 6d). Inhibitory neurons displayed a distinct pattern, with a comparable number of upregulated genes but fewer downregulated genes than excitatory neurons (Fig. 6f), suggesting cell-type\u0026ndash;specific transcriptional alterations in response to tumor burden.\u003c/p\u003e\n\u003cp\u003ePathway analysis revealed a conserved induction of mitochondrial translation in both excitatory and inhibitory neurons, accompanied by broad suppression of gene programs related to neuronal structure and excitability (Fig. 6e,g,k). Despite their anatomical distance from the tumor, astrocytes in the ventral hippocampus exhibited robust transcriptional responses (Fig. 6h). Similar to the medial prefrontal cortex, astrocytes showed prominent upregulation of inflammatory gene programs (Fig. 6j), including \u003cem\u003eIer3\u003c/em\u003e and \u003cem\u003eGfap\u003c/em\u003e, consistent with a distributed glial response to tumor-associated cues. Astrocytes also exhibited increased expression of metabolic genes (Fig. 6j) such as \u003cem\u003eAldoc\u003c/em\u003e, suggesting an altered metabolic state or increased energetic demand. Notably, astrocytes shared with neurons a suppression of pathways associated with synaptic organization, membrane potential regulation, and cognition, implicating astrocytic dysfunction in broader disruption of hippocampal circuit support (Fig. 6e,g,i,k). Of particular interest, both excitatory and inhibitory neurons showed downregulation of \u003cem\u003eNpas4\u0026nbsp;\u003c/em\u003e(Fig. 6k), an activity-dependent immediate early gene, consistent with the reduced neuronal engagement observed during elevated zero maze exploration at later stages of disease. In addition, all cell types exhibited decreased expression of \u003cem\u003eHtr7\u0026nbsp;\u003c/em\u003e(Fig. 6k), a serotonergic receptor enriched in the adult hippocampus and implicated in cognition and mood regulation.\u003c/p\u003e\n\u003cp\u003eTogether, these findings demonstrate that glioma burden induces widespread, age-dependent transcriptional remodeling across extratumoral brain regions. Rather than acting solely through local effects near the tumor core, glioma burden engages a distributed molecular response characterized by inflammatory activation, metabolic stress, and suppression of synaptic support programs. These transcriptional alterations parallel the functional disruptions in neuron\u0026ndash;astrocyte coordination observed \u003cem\u003ein vivo\u003c/em\u003e and reinforce astrocytes as central mediators of glioma-associated vulnerability in the aging brain.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eAging is a dominant determinant of glioblastoma incidence and outcome\u003csup\u003e27\u003c/sup\u003e, yet how the aged brain reshapes circuit-level vulnerability to tumor burden has remained poorly understood. Here, we show that aging fundamentally alters the way glioma burden disrupts extratumoral brain circuits by selectively targeting astrocytic integration rather than broadly suppressing neuronal activity. Across behavioral contexts, astrocytes emerged as the earliest and most sensitive cellular substrate of tumor-associated dysfunction, with aging amplifying both the magnitude and functional consequences of this disruption.\u003c/p\u003e\n\u003cp\u003eA key insight from this work is that tumor burden does not simply abrogate neural activity, but instead degrades the glial-mediated coordination required to transform neuronal signals into coherent, behaviorally appropriate circuit output (Fig. 1 - 4). Neuronal activity largely preserved behavioral state encoding across stress- and anxiety-related paradigms, even in the presence of tumor burden. In contrast, astrocytic responses showed reduced magnitude, altered timing, and weakened synchrony, particularly in aged animals. These findings suggest that glioma-associated circuit dysfunction arises not from loss of neuronal responsiveness per se, but from failure of astrocytes to appropriately integrate, sustain, and coordinate neuronal activity across space and time.\u003c/p\u003e\n\u003cp\u003eThis astrocyte-centered vulnerability was shaped by both aging and anatomical proximity to the tumor. Circuits adjacent to the tumor site exhibited the most severe impairments in astrocytic engagement and coordination, while more distal regions showed partial preservation of response structure. Aging markedly intensified these effects, revealing a spatial gradient in which tumor-associated signals interact with age-dependent astrocytic fragility to erode circuit function. Importantly, this pattern was consistent across distinct behavioral states, indicating a generalized impairment in astrocytic circuit support rather than task-specific failure.\u003c/p\u003e\n\u003cp\u003eDuring anxiety-related behavior, astrocytic dysfunction again dominated the circuit phenotype. Aging was associated with enhanced astrocytic recruitment during anxiogenic states, suggesting that astrocytes normally scale their engagement with increased behavioral demand. Tumor burden selectively blunted this amplification in aged animals, suppressing astrocytic responses during both sustained exploration and discrete behavioral transitions. Neuronal alterations emerged more selectively and later, particularly in the ventral hippocampus, consistent with a model in which neuronal dysfunction arises downstream of impaired astrocytic support rather than as an initiating event.\u003c/p\u003e\n\u003cp\u003eDisruption of astrocytic coordination, rather than local response magnitude alone, emerged as a defining feature of tumor-associated circuit failure. Tumor burden weakened local neuron–astrocyte coupling and, most strikingly, disrupted long-range astrocyte–astrocyte synchrony across regions. This loss of glial coordination occurred even when neuronal activity retained behavioral modulation, suggesting that astrocytes play a critical role in aligning distributed circuit states. Aging amplified this desynchronization, potentially rendering circuits less resilient to tumor-associated perturbation.\u003c/p\u003e\n\u003cp\u003eMoreover, our single-nucleus RNA sequencing data provides a molecular framework for these physiological findings. In the tumor-adjacent cortex, aging dramatically increased the depth and breadth of transcriptional suppression, with astrocytes showing pronounced downregulation of genes involved in synaptic support, neuronal signaling, and metabolic coupling. These changes were accompanied by selective loss of vascular-associated cell types and transcriptional signatures consistent with blood–brain barrier disruption, implicating breakdown of the neurovascular unit as a contributor to astrocytic dysfunction. In distal regions, transcriptional perturbations were evident despite preserved cellular composition, indicating that tumor-associated molecular stress propagates beyond the tumor core without overt structural remodeling. Across regions, astrocytes exhibited the strongest convergence of inflammatory activation and loss of circuit-supportive programs, positioning them as integrators of tumor- and age-related stress.\u003c/p\u003e\n\u003cp\u003eTogether, these findings support a model in which aging converts astrocytes from adaptive circuit regulators into points of failure under tumor burden. Early astrocytic dysfunction compromises coordination and sustained engagement, progressively constraining neuronal function and degrading behavioral output. This framework helps reconcile why neuronal activity can appear relatively preserved in early disease stages while behavioral impairment and network instability nonetheless emerge, particularly in older hosts.\u003c/p\u003e\n\u003cp\u003eFrom a translational perspective, these findings highlight a fundamental tension in glioblastoma therapy. Many pathways that support synaptic maintenance, metabolic coupling, and neuromodulatory signaling are also co-opted by glioma cells to promote tumor growth. In the aged brain, where these circuit-supportive mechanisms are already compromised, further disruption by tumor-directed interventions may exacerbate functional decline even as tumor burden is reduced. Similarly, therapies designed to ameliorate cognitive function in patients could have consequential effects to tumor growth and progression. Our results therefore suggest that preserving astrocytic and neurovascular function may be essential for maintaining circuit integrity and quality of life in older patients alongside anti-tumor strategies.\u003c/p\u003e\n\u003cp\u003eIn sum, this work identifies astrocytes as central mediators of age-dependent vulnerability to glioma burden and demonstrates that aging fundamentally reshapes extratumoral circuit dynamics before overt neuronal failure emerges. By revealing astrocytic dysfunction as a primary driver of disrupted network coordination beyond the tumor core, these findings position glial and neurovascular mechanisms as critical determinants of functional outcome in the aging glioblastoma brain.\u003c/p\u003e"},{"header":"Methods","content":"\u003ch2\u003e\u003cu\u003eAnimals\u003c/u\u003e\u003c/h2\u003e\n\u003cp\u003eYoung and aged male C57BL/6J mice were obtained from The Jackson Laboratory. For fiber photometry experiments, mice were acquired three weeks earlier to accommodate the viral expression period; tumor induction was performed at the same ages across all experiments, with young mice 6\u0026ndash;7 weeks old and aged mice 85\u0026ndash;86 weeks old at the time of glioma implantation. Mice were housed under standard conditions (12 h light/dark cycle, ad libitum access to food and water). All procedures were approved by the Institutional Animal Care and Use Committee at Boston University and performed in accordance with NIH guidelines.\u003c/p\u003e\n\u003ch2\u003e\u003cu\u003eCell line and mPFC glioma induction\u003c/u\u003e\u003c/h2\u003e\n\u003cp\u003eSB28-Ohlfest murine glioma cells were maintained under standard culture conditions (37\u0026deg;C, 5% CO₂) and routinely screened for mycoplasma contamination. For stereotaxic implantation, mice were anesthetized with isoflurane and placed in a stereotaxic frame. A 2 \u0026micro;L suspension containing 50,000 SB28 cells in DMEM was injected into the mPFC at 500 nL/min using a 10 \u0026micro;L, 33-gauge Hamilton syringe at DV \u0026minus;2.5 mm. Tumor coordinates (relative to bregma) were AP +1.8 mm and ML \u0026minus;0.5 mm. The syringe was left in place for 1 minute following injection to minimize reflux.\u003c/p\u003e\n\u003ch2\u003e\u003cu\u003eViral delivery and optical fiber implantation\u003c/u\u003e\u003c/h2\u003e\n\u003cp\u003eTo record neuronal and astrocytic calcium activity, we used AAV-hSyn-jRGECO1a (neurons) and AAV-GfaABC1D-jGCaMP8f-SV40 (astrocytes). Viruses were mixed at a 1:2 dilution (jRGECO1a:jGCaMP8f) and co-injected into the medial prefrontal cortex (mPFC) and ventral hippocampus (vHPC). Injection volumes were 500 nL for mPFC and 750 nL for vHPC. Viruses were injected three weeks prior to tumor implantation and optical fiber placement to allow stable expression before induction of tumor burden and longitudinal recordings.\u003c/p\u003e\n\u003cp\u003eVirus injections and fiber placements were performed in hemispheres contralateral to the tumor injection site. For mPFC, virus injection and recording coordinates were AP +1.8 mm, ML +0.5 mm, DV \u0026minus;2.5 mm (contralateral to tumor). For vHPC, injection coordinates were AP \u0026minus;3.16 mm, ML +3.16 mm, DV \u0026minus;4.5 mm, and the optical fiber was implanted at DV \u0026minus;4.6 mm (contralateral hemisphere to tumor). Optical fibers were 200 \u0026micro;m core diameter, NA 0.37, and were secured using dental cement.\u003c/p\u003e\n\u003ch2\u003e\u003cu\u003eTimeline of fiber photometry and behavioral experiments\u003c/u\u003e\u003c/h2\u003e\n\u003cp\u003eVirus injection occurred three weeks prior to tumor induction and optical fiber implantation. Following tumor induction and fiber implantation, mice underwent longitudinal fiber photometry across baseline and behavioral assays. The timeline was:\u0026nbsp;Home cage 1 (day 8) \u0026rarr; tail suspension test (day 10) \u0026rarr; home cage 2 (day 12) \u0026rarr; elevated zero maze (day 13) \u0026rarr; foot shocks, as shown in Fig. 1A.\u003c/p\u003e\n\u003ch2\u003e\u003cu\u003eBehavioral assays and quantification\u003c/u\u003e\u003c/h2\u003e\n\u003ch3\u003e\u003cem\u003eHome cage recordings\u003c/em\u003e\u003c/h3\u003e\n\u003cp\u003ePhotometry was recorded in the home cage on day 8 and day 12 of tumor burden to establish baseline neuronal and astrocytic activity during spontaneous behavior.\u003c/p\u003e\n\u003ch3\u003e\u003cem\u003eTail suspension test (TST)\u003c/em\u003e\u003c/h3\u003e\n\u003cp\u003eMice were suspended by the tail for 6 min. Behavior was manually scored as struggling or immobility. For peri-event analyses, struggling epochs were required to be \u0026ge;2 s in duration to be included. Struggle onset was defined as the first frame of sustained movement following immobility, and struggle offset as the first frame of sustained immobility following a struggling epoch. These labels were used to segment photometry signals for state-dependent quantification. Due to technical limitations of fiber photometry recording, additional mice (non-fiber photometry) were needed to assess latency to immobility, which is reflected in Extended Fig. 4a-b. \u0026nbsp;Latency to immobility was defined as the time elapsed from initial suspension to the first immobility of greater than 3 seconds.\u003c/p\u003e\n\u003ch3\u003e\u003cem\u003eElevated zero maze (EZM)\u003c/em\u003e\u003c/h3\u003e\n\u003cp\u003eMice were placed on an elevated circular maze containing two open and two closed quadrants and recorded for 8 min. Behavior was scored using ANY-maze software. Behavioral state was defined as open area or closed area, and transitions were defined by entry into each area. The animal was only counted as having entered the open area if all four legs had crossed the transition point. Risk assessment behavior was characterized by the time in which an animal has its head out into the open area, scanning the environment, without fully engaging into the open area\u003csup\u003e28,29\u003c/sup\u003e. A subset of mice from the pure-behavior experiments for tail suspension also performed the EZM for additional N. For peri-event analyses, open-area entry was used as the alignment event.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eFoot shock\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eTo evoke well-established population level events in both brain regions\u003csup\u003e19,30\u003c/sup\u003e, mice underwent a brief foot shock session in a conditioning chamber. The session duration was 60 s. Foot shocks were delivered at 10 s, 30 s, and 50 s, each lasting 2 s at 1.5 mA. Photometry was recorded throughout the session.\u003c/p\u003e\n\u003ch2\u003e\u003cu\u003eFiber photometry\u003c/u\u003e\u003c/h2\u003e\n\u003cp\u003e\u003cem\u003eAcquisition\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eDual-color fiber photometry recordings were performed using the FP3002 system (Neurophotometrics). Excitation light was delivered through implanted optical fibers, and emitted fluorescence was captured by a photodetector/camera system. Astrocytic GCaMP signals were excited at 470 nm and neuronal jRGECO signals were excited at 560 nm, with an output power of 50 \u0026micro;W per channel at the patch cord tip. Calcium-independent isosbestic signals were simultaneously captured by alternating excitation with a 415 nm LED to dissociate motion artifacts, tissue autofluorescence, and photobleaching from calcium-dependent fluorescence changes. All wavelengths were interleaved and collected using Bonsai software (Lopes et al., 2015). Signals were sampled at 30 Hz (30 frames per second).\u003c/p\u003e\n\u003ch2\u003e\u003cem\u003ePhotometry preprocessing and signal normalization\u003c/em\u003e\u003c/h2\u003e\n\u003cp\u003ePhotometry traces were processed using custom Python scripts (\u003cu\u003ehttps://github.com/rsenne/RamiPho\u003c/u\u003e). Raw fluorescence signals were baseline-corrected using an adaptive iteratively reweighted penalized least-squares method (airPLS; Zhang et al., 2010). Signals were then smoothed using kernel-based smoothing to improve signal-to-noise ratio. dF/F was calculated for each channel as:\u003c/p\u003e\n\u003cp\u003e\u003cimg width=\"171\" height=\"39\" src=\"data:image/png;base64,R0lGODlhAAE7AHcAMSH+GlNvZnR3YXJlOiBNaWNyb3NvZnQgT2ZmaWNlACH5BAEAAAAALAAAAgAAATUAhgAAAAAAAAEAAAAAOgAAZgEAZgAA/wA6OgA6ZgE6ZgA6kABmtgD//zoAADoBADoAOjo6ADo6Ojo6kDpmZjpmkDpmtjqQtjqQ2zqQ2mYAAGYBAGYAOmYBOmY6AGY6OmY6ZmaQtmaQ22a222a2/2a2/pA6AJA6AZA6ZpBmAJBmAZBmOpC225Db/5Da/7ZmALdmALZmOraQZra2Zrbb/7b//7b+/9uQOtuROtuQZtuRZtu2Ztu3Ztu2kNu3kdv/29v//9r///+2Zv+3Zv/bkP/bkf/btv/b2///tv//t///2////wECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwf/gACCg4SFhoeIiYqLjI2KRR4WiTgIK46XmJmIQQEBCj+aoQCQko45CZaiqqusgy6dsLAENIU2AyKLPB0VoK2+oZwLv4y2uIMvsbEENYI9u73D0dKCSSgBF4NHHcKEtiyNRxnc0+SGQw3Y2Rmz0t6FSSkBGINI24RIGuPl+6ouA9+u0gniNMJQOH2CzgnkR44goXMIWzk09GJAC0Iv5hEi4kAjw4+OkpRgh0jkJ0MTC7kgCTKaDZblTEIblMTEMkYvbrbciejcyZ7oDBUp8e8Qp4U8Wbn4uU/hIY4KgDASIsBj0qtHBR3ZAHCQO0KvYhVVF7EVjgbXRnUIUJQHhAAI/2gR4uGhE4KuAJLEQDtggrhB5wLA1Is2QAUZEkCdTVtk7d1uYwtR1YiEw8VD+MpeBWkjQEEAL+VSGymabEnSLkH4KKFAB69gOEKATjsIRgALP3z2EnkrbwnPDxvo441LZABhNlSzdv0jiHCaqA3dEEBC0I0CzA7V1Ll5p/FYTAGECz+Q9iF/eH+JPFAhIdp0WQV1TidT0KvPANAHX3jfK/DR7Lk3zngzUWNCMlElUtFl3bUUDjsiIUWgIZ3hZ4h+KCWjoScFAtYASbrJ91+IAEDkHlMrlVahhyhGdg5J8YmXAXkA4HNTTVZRZFGDOwXjioUykicSTCpFNkyMK9o31v98ABixV1wi4lefKywl6RtTSP4XZIdUjfNCdQruyCNn5h0yISFnnmfkL1/lN9aE3wVwwASp+OZiUOr8xFtXJno11pCipTnIdDkqsuCYIGGIiKAlPpeIoupFN2WfgGoXXX5aNjpOpfZpyWmajAJwqCOjIrrPg6UZwukg8eEA5Kq/nNnnbJ+tOkRivsEYgZFWchrErl2dGWOuqdqYHSPbHWsqOToIlh6Fa/qoA65oztghKzFmOUIRH9DQmTE6PNDVK7joEAJ6PPBCK7e0kAuAuei+Zt5RRZwAUJuC7CAAAQwugk+CyzYEy5qFDJuXCmzhliFSvlipaGAUzCDIYgNETIj/cb05Z1gvQ6xlsW9s4aKxulYe8RsCEpe3EFWdiLnIZAEjOqUjKcZs88yh5KRshhyuUrPNGQKpiFNAx5ySJhwVmqE+YWlIpHv8bShYqlfhm8hBRQNttSOZNUI0NdYIpI1m+FZj3timFpNLA+pmbbPamPTA9rVBF4mXCwxfWSCkeC9LyiSVuJ3136agckloF186NJ7QPS3445AnGh6Ji1hNeeSYZ+4LYWz5hVB8Wz2r1V8F0xa65qinHhJRxf0GZJKIG5VplPI5zqrU4NGt+u5J9ackXnF2QqN9RAbfM+/I7075z6NDWELeMkaEqm/Q+4L79dhnr/323Hfv/ffghx8A/63QheejfULTXvo4LqSf/PuC7+khUkwugrOI1cuO/fDw9x/Nqr5zBcHMwTiwDNB/CCyar4BVLdtB5lnTS6AEBecueP0jXaBoFgFElzgaaZCDEwzhmDAmMrSoixMtAyEAjlaeFIrwSMLTnTQI1whK1Cl1zHvhR873Ebh1akMk0UXbNIc1HbbkazJyoCisZjax2cM0qGOhEfdxtFn9QopuuhtSkAg5WE1xH7Hbx/2IxYgcPo6LXyzHUmToCzQ2in+EMFgaIbcYbDSGLd9wC1xSRRe7AG8vnSNdo6Z2MUAaBjGKeY9a4IKXra3QdFxZlCDnCLnkrKY18lpAbGazENvghv9ExAGZhawYSuMgRzmYbI6jyEih/4TxHYqj5OPW055Bwsc89ZtSACGFxAD2qgQB0lSeOmQ8OEJKloJ7kVxIlCQS9Wl5MLHS8u4Eo7NZyyDroEWEFHFMZGZNWwK8V1qc1AAokW804WHeL7GESy2Fioft4+YBvRmzNmEITr/pxJzqJD+oVUtPROHTKkHzp0uFqn6N6CY9Y8apSTnKi6NhSQCFGVHRBPBT1xwdjRSKCI4u1FSyWmWSbIWrBfLKU9H5VWSEVSYvRhBZsfxowLLVzm1161uCCNe4AlCuc12QZJ5hF6Z6Gq/mzOsa9brXmj7oiFDJNG1aehhaPkaxj40mZCtGNCHHPJYyEmZ1Y+cUz8lS9siFoBCPjZDjU9caKTgmVIlsjSswZocJN8r1rkucpyKKiNe+ssKHjuDB3PxK2FXQkBE25EkgAAA7\" alt=\"image\"\u003e\u003c/p\u003e\n\u003cp\u003eyielding percent deviation from baseline.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFor peri-event analyses and event-triggered averages (ETAs), photometry signals were z-scored within session prior to alignment and averaging to enable comparison of dynamics across animals and groups. For ETA area-under-the-curve (AUC) quantification, each mouse\u0026rsquo;s mean ETA trace was integrated over a 10 s post-event window for neuronal signals and a 15 s post-event window for astrocytic signals. Because astrocytic ETAs frequently exhibited large negative deflections, AUC for astrocytes was calculated on the absolute value of the ETA trace to capture full response magnitude. Neuron-astrocyte latency (\u0026Delta;t) was quantified per mouse as the difference between astrocytic and neuronal peak times in the mean ETA trace (\u0026Delta;t = t_astro \u0026minus; t_neuron) for each mouse.\u003c/p\u003e\n\u003ch2\u003e\u003cem\u003eEvent detection and peak-based quantification\u003c/em\u003e\u003c/h2\u003e\n\u003cp\u003eCalcium events were detected using a threshold-based peak identification approach with region- and cell-type-specific parameters. Peaks were identified on baseline-corrected, smoothed, and z scored traces using the SciPy function, find_peaks(), with the following parameters:\u003c/p\u003e\n\u003cul class=\"decimal_type\"\u003e\n \u003cli\u003e\u003cem\u003eAstrocyte\u003c/em\u003e\u003cstrong\u003e:\u003c/strong\u003e height \u0026ge; 0.25 SD, prominence \u0026ge; 1 SD, width = 90\u0026ndash;500 frames, rel_height = 0.7, wlen = 600\u003cbr\u003e\u0026nbsp;\u003cbr\u003e\u0026nbsp;\u003c/li\u003e\n \u003cli\u003e\u003cem\u003eNeuron\u003c/em\u003e\u003cstrong\u003e:\u003c/strong\u003e height \u0026ge; 1 SD, prominence \u0026ge; 2 SD, width = 15\u0026ndash;500 frames, rel_height = 0.5\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003eTo quantify behavior-specific event properties, behavioral epochs were first annotated (e.g., struggling versus immobility in the tail suspension test, open versus closed area in the elevated zero maze), and events were assigned to behavioral states based on whether their peak time occurred within a given epoch. Event frequency was computed as events/min for each behavioral state. AUC were averaged across events within each state for each mouse, yielding one per-mouse value per state for group analyses.\u003c/p\u003e\n\u003ch2\u003e\u003cem\u003ePeri-event analyses\u003c/em\u003e\u003c/h2\u003e\n\u003cp\u003ePeri-event analyses were performed by aligning z-scored signals to behavioral transitions, including struggle onset (TST) and open-area entry (EZM). Only trials meeting inclusion criteria (e.g., struggling epochs \u0026ge;2 s) were included. ETAs were computed per animal by averaging across trials, then averaged across animals within each group. Confidence bounds for ETAs were computed using a t-confidence interval (tCI) method\u003csup\u003e31\u003c/sup\u003e. For each peri-event time point, a 95% confidence interval was calculated across trials.\u003c/p\u003e\n\u003ch2\u003e\u003cem\u003eLagged correlation analysis\u003c/em\u003e\u003c/h2\u003e\n\u003cp\u003eLagged correlations between z-scored photometry traces were computed per mouse using cross-correlation (SciPy correlate) to identify the lag of maximal correlation within predefined bounds. Lag windows were set a priori by pair (neuron\u0026ndash;astro: 30\u0026ndash;240 frames; neuron\u0026ndash;neuron and astro\u0026ndash;astro: 0\u0026ndash;60 frames; default: 0\u0026ndash;150 frames), and neuron\u0026ndash;astro pairs were constrained to positive reported lags (astrocytes lagging neurons). Traces were shifted by trimming the leading portion of the appropriate signal at the optimal lag, and coupling strength was quantified as the Pearson correlation coefficient on aligned traces.\u003c/p\u003e\n\u003ch2\u003e\u003cu\u003eSingle-nucleus RNA sequencing (snRNA-seq)\u003c/u\u003e\u003c/h2\u003e\n\u003ch3\u003e\u003cem\u003eTissue dissection and experimental design\u003c/em\u003e\u003c/h3\u003e\n\u003cp\u003emPFC tissue was dissected on day 13 of tumor burden from young and aged control and tumor-bearing mice (n = 3 per group). vHPC tissue was collected on day 13 from aged control and aged tumor-bearing mice (n = 3 per group). Tissue was rapidly dissected on ice and immediately snap-frozen for nuclei isolation and single-nucleus RNA sequencing.\u003c/p\u003e\n\u003ch3\u003e\u003cem\u003eNuclei isolation, library preparation, and sequencing\u003c/em\u003e\u003c/h3\u003e\n\u003cp\u003eNuclei were isolated and processed for 10x Genomics single-nucleus RNA sequencing. FASTQ files were processed using Cell Ranger v7.0.0 (10x Genomics), to perform STAR-based alignment and generate filtered feature-barcode matrices.\u0026nbsp;\u003c/p\u003e\n\u003ch3\u003e\u003cem\u003ePreprocessing, quality control, integration, and annotation\u003c/em\u003e\u003c/h3\u003e\n\u003cp\u003eDownstream analysis was performed in Seurat v4.1.1. Cells were filtered to remove low-quality barcodes and potential multiplets using the following thresholds: \u0026lt;200 detected genes, genes expressed in \u0026le;3 cells, \u0026gt;8000 features, and mitochondrial transcript fraction \u0026gt;10%. Doublets were identified and removed using DoubletFinder v2.0.3. Each mouse dataset was normalized by total UMI counts and log-transformed, and variable features were identified using the vst method. Datasets were then integrated across mice using canonical correlation analysis (CCA) in Seurat (FindIntegrationAnchors/IntegrateData). The integrated data were scaled and used for PCA, followed by UMAP and/or t-SNE for visualization. Cell types were annotated based on canonical marker genes. For the ventral hippocampus, excitatory neurons included as \u0026ldquo;ventral hippocampus\u0026rdquo; included any cell cluster that included hippocampal subregions and prosubiculum, due to the prosubiculum having similar dorsal-ventral (DV) axis differentiation as the hippocampus and its close proximity to vHPC-proper subregions\u003csup\u003e32\u003c/sup\u003e.\u003c/p\u003e\n\u003ch3\u003e\u003cem\u003eCell Proportions Analysis\u003c/em\u003e\u003c/h3\u003e\n\u003cp\u003eDifferential abundance was first assessed using a propeller-based analysis of sample-level cluster proportions from the speckle package\u003csup\u003e33\u003c/sup\u003e. Post-hoc, logit-transformed cluster proportions were modeled using limma with a group-wise design and empirical Bayes moderation to test within-age Tumor vs Control contrasts. Moderated statistics were extracted for all clusters in Young and Aged cohorts.\u003c/p\u003e\n\u003ch3\u003e\u003cem\u003eDifferential expression and gene set enrichment analysis\u003c/em\u003e\u003c/h3\u003e\n\u003cp\u003eDifferential expression (DE) was performed within each annotated cell population to identify tumor-associated transcriptional changes relative to age-matched controls. For mPFC, tumor-bearing mice were compared to age-matched controls within each age group, and for vHPC, aged tumor was compared to aged control. DE was conducted using nebula, a negative binomial mixed model framework with mouse identity modeled as a random effect and log library size included as an offset term\u003csup\u003e34\u003c/sup\u003e. Models were fit using the hierarchical likelihood (HL) approach. Genes with \u0026theta; greater than the 99.5th percentile of the \u0026theta; distribution within each analyzed cell population were designated as \u0026ldquo;high-overdispersion\u0026rdquo; and were excluded from downstream DEG and enrichment analyses. P values were corrected for multiple comparisons using the Benjamini\u0026ndash;Hochberg procedure, and genes were considered differentially expressed at FDR \u0026lt; 0.05.\u003c/p\u003e\n\u003cp\u003eGene Ontology (GO) enrichment was assessed using gene set enrichment analysis (GSEA) implemented in fgsea on ranked gene lists, where genes were ranked by sign(effect) \u0026times; \u0026minus;log10(adjusted p-value), with enrichment significance determined at FDR \u0026lt; 0.05. To reduce redundancy among significantly enriched Gene Ontology (GO) terms, GO Biological Process terms passing FDR \u0026lt; 0.05 were summarized using REVIGO\u003csup\u003e35\u003c/sup\u003e. GO terms were provided to REVIGO with adjusted P values as the associated scores, and redundancy reduction was performed using the SimRel semantic similarity measure with a medium allowed similarity threshold (0.7). Obsolete GO terms were removed, and the whole UniProt database was used as the reference background.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eCell-type prioritization analysis using Augur\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eCell-type prioritization was performed using Augur, which quantifies the extent to which transcriptional profiles within a given cell type discriminate between experimental conditions via repeated subsampling and supervised classification. For each annotated cell type, Augur repeatedly subsampled an equal number of nuclei per condition and trained a classifier to distinguish Control versus Tumor labels. Classification performance was evaluated using cross-validation and summarized using the area under the receiver operating characteristic curve (ROC AUC). This procedure was repeated across 50 independent subsamples to ensure robustness to cell sampling variability. Augur analyses were performed separately for Young and Aged cohorts.\u003c/p\u003e\n\u003cp\u003eTo assess whether individual cell types exhibited classification performance significantly above chance, we tested whether subsample-level ROC AUC distributions exceeded 0.5 (chance performance). For each cell type, ROC AUC values were first averaged across cross-validation folds within each subsample, yielding one independent performance estimate per subsample. A one-sided Wilcoxon signed-rank test was then applied to test whether the median subsample-level ROC AUC was greater than 0.5. P values were corrected for multiple testing across cell types using the Benjamini\u0026ndash;Hochberg false discovery rate (FDR) procedure.\u003c/p\u003e\n\u003cp\u003eTo identify cell types exhibiting age-dependent differences in tumor-associated transcriptional disturbances, differential prioritization was assessed by comparing observed cell type-specific differences in Augur AUC scores between Young and Aged cohorts to a null distribution of \u0026Delta;AUC values generated from label-permuted Augur runs. Empirical two-sided permutation p-values were computed for each cell type.\u003c/p\u003e\n\u003ch2\u003e\u003cu\u003eBrain tissue processing and immunohistochemistry\u003c/u\u003e\u003c/h2\u003e\n\u003cp\u003eFor validation of viral expression and fiber placement, and histological markers, mice were transcardially perfused with PBS followed by 4% paraformaldehyde. Heads were post-fixed for 48 hours in 4% paraformaldehyde, brains were extracted,\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003eand were sectioned at 40 \u0026micro;m. \u0026nbsp;The intensity of the endogenous fluorescent proteins was sufficient so that no amplification step with immunohistochemistry was required. Images were acquired using a Zeiss LSM800 confocal microscope, with 886.5 (3 tiles) x 2068.5 (7 tiles)\u0026nbsp;\u0026micro;m ROI for the mPFC and 1182.0 (4 tiles) x 1469.4 (5 tiles)\u0026nbsp;\u0026micro;m ROI for the vHPC. jRGECO neuron counts were segmented and counted using Cellpose3\u003csup\u003e36\u003c/sup\u003e. Due to the membrane bound nature of the GCaMP protein, segmentation of individual cells was not feasible for astrocytes. Instead, the images were thresholded and the area of positive signal per ROI was obtained using FIJI/ImageJ.\u003c/p\u003e\n\u003cp\u003eDue to the nature of tumor growth and proximity of the optical fiber to the tumor, a subset of mice lost signal over the progression of the disease. To determine whether this reflected a true loss of cellular activity rather than displacement of the optical fiber, we examined confocal images from each mouse. We excluded the first and following recordings in which signal loss occurred if the fiber tip was more than 200 \u0026micro;m from the nearest jRGECO⁺ neuron (excluding neuronal recordings), GCaMP⁺ astrocyte (excluding astrocyte recordings), or both (excluding recordings from both cell types), which corresponds to the loss of 50% of maximum photometry efficiency\u003csup\u003e37\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cu\u003eStatistical analysis\u003c/u\u003e\u003c/p\u003e\n\u003cp\u003eStatistical analyses were performed using GraphPad Prism. When multiple comparisons were performed, Sidak\u0026rsquo;s correction was applied. Statistical significance was defined as p \u0026lt; 0.05. Data are reported as mean \u0026plusmn; SEM unless otherwise indicated.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003e\u003cu\u003eData and code availability\u003c/u\u003e\u003c/h2\u003e\n\u003cp\u003eCustom photometry preprocessing and analysis scripts are freely available at \u003cu\u003ehttps://github.com/rsenne/RamiPho\u003c/u\u003e. Sequencing data will be deposited in Gene Expression Omnibus upon publication. Other data are available from the corresponding authors upon request.\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003eAuthor Contributions\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eConceptualization: H.L. and S.R.; methodology: H.L. and S.R.; formal analysis: H.L.; investigation: H.L., M.B., M.H.; visualization: H.L.; writing—original draft: H.L. and S.R.; writing—review and editing: H.L., M.B., M.H. and S.R.; funding acquisition: S.R.; resources: S.R.; and supervision: S.R.\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003eFunding and Acknowledgments\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eThis work was supported by NIH Transformative Award, the Ludwig Family Foundation, the Air Force Office of Scientific Research award (FA9550-21-1-0310), the Pew Scholars Program in the Biomedical Sciences, the Chan-Zuckerberg Initiative, and Neurophotonics Center at Boston University and the Center for Systems Neuroscience.\u003c/p\u003e\n\u003cp\u003eBehavioral schematics and timelines were created using BioRender.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCorresponding author\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCorrespondence to Steve Ramirez ([email protected]).\u0026nbsp;\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eMiller, K. D. \u003cem\u003eet al.\u003c/em\u003e Brain and other central nervous system tumor statistics, 2021. \u003cem\u003eCA Cancer J. Clin.\u003c/em\u003e \u003cstrong\u003e71\u003c/strong\u003e, 381\u0026ndash;406 (2021).\u003c/li\u003e\n\u003cli\u003eLadomersky, E. \u003cem\u003eet al.\u003c/em\u003e Advanced age negatively impacts survival in an experimental brain tumor model. \u003cem\u003eNeurosci. Lett.\u003c/em\u003e \u003cstrong\u003e630\u003c/strong\u003e, 203\u0026ndash;208 (2016).\u003c/li\u003e\n\u003cli\u003eChen, X., Cui, Y. \u0026amp; Zou, L. Treatment advances in high-grade gliomas. \u003cem\u003eFront. Oncol.\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, 1287725 (2024).\u003c/li\u003e\n\u003cli\u003eKorja, M. \u003cem\u003eet al.\u003c/em\u003e Glioblastoma survival is improving despite increasing incidence rates: a nationwide study between 2000 and 2013 in Finland. \u003cem\u003eNeuro. Oncol.\u003c/em\u003e \u003cstrong\u003e21\u003c/strong\u003e, 370\u0026ndash;379 (2019).\u003c/li\u003e\n\u003cli\u003eKarsy, M. \u003cem\u003eet al.\u003c/em\u003e Surgical treatment of glioblastoma in the elderly: the impact of complications. \u003cem\u003eJ. Neurooncol.\u003c/em\u003e \u003cstrong\u003e138\u003c/strong\u003e, 123\u0026ndash;132 (2018).\u003c/li\u003e\n\u003cli\u003eGately, L., McLachlan, S. A., Dowling, A. \u0026amp; Philip, J. Life beyond a diagnosis of glioblastoma: a systematic review of the literature. \u003cem\u003eJ. Cancer Surviv.\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 447\u0026ndash;452 (2017).\u003c/li\u003e\n\u003cli\u003eZhang, C. \u003cem\u003eet al.\u003c/em\u003e Aging and senescence: Key players in brain tumor progression and drug resistance. \u003cem\u003eDrug Resist. Updat.\u003c/em\u003e \u003cstrong\u003e81\u003c/strong\u003e, 101228 (2025).\u003c/li\u003e\n\u003cli\u003eMattson, M. P. \u0026amp; Arumugam, T. V. Hallmarks of Brain Aging: Adaptive and Pathological Modification by Metabolic States. \u003cem\u003eCell Metab.\u003c/em\u003e \u003cstrong\u003e27\u003c/strong\u003e, 1176\u0026ndash;1199 (2018).\u003c/li\u003e\n\u003cli\u003eMan, J. H. K. \u003cem\u003eet al.\u003c/em\u003e Region-specific and age-related differences in astrocytes in the human brain. \u003cem\u003eNeurobiol. Aging\u003c/em\u003e \u003cstrong\u003e140\u003c/strong\u003e, 102\u0026ndash;115 (2024).\u003c/li\u003e\n\u003cli\u003eRaz, N. \u0026amp; Rodrigue, K. M. Differential aging of the brain: patterns, cognitive correlates and modifiers. \u003cem\u003eNeurosci. Biobehav. Rev.\u003c/em\u003e \u003cstrong\u003e30\u003c/strong\u003e, 730\u0026ndash;748 (2006).\u003c/li\u003e\n\u003cli\u003evan der Meer, P. B. \u003cem\u003eet al.\u003c/em\u003e Depression and anxiety in glioma patients. \u003cem\u003eNeurooncol Pract\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, 335\u0026ndash;343 (2023).\u003c/li\u003e\n\u003cli\u003eTalacchi, A., Santini, B., Savazzi, S. \u0026amp; Gerosa, M. Cognitive effects of tumour and surgical treatment in glioma patients. \u003cem\u003eJ. Neurooncol.\u003c/em\u003e \u003cstrong\u003e103\u003c/strong\u003e, 541\u0026ndash;549 (2011).\u003c/li\u003e\n\u003cli\u003eLyon, K. A. \u0026amp; Allen, N. J. From synapses to circuits, astrocytes regulate behavior. \u003cem\u003eFront. Neural Circuits\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 786293 (2021).\u003c/li\u003e\n\u003cli\u003ePearson-Leary, J., Osborne, D. M. \u0026amp; McNay, E. C. Role of Glia in Stress-Induced Enhancement and Impairment of Memory. \u003cem\u003eFront. Integr. Neurosci.\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 1\u0026ndash;14 (2016).\u003c/li\u003e\n\u003cli\u003ePerez-Catalan, N. A., Doe, C. Q. \u0026amp; Ackerman, S. D. The role of astrocyte-mediated plasticity in neural circuit development and function. \u003cem\u003eNeural Dev.\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 1 (2021).\u003c/li\u003e\n\u003cli\u003eLabarta-Bajo, L. \u0026amp; Allen, N. J. Astrocytes in aging. \u003cem\u003eNeuron\u003c/em\u003e \u003cstrong\u003e113\u003c/strong\u003e, 109\u0026ndash;126 (2025).\u003c/li\u003e\n\u003cli\u003eLeblanc, H., Buzharsky, M., Varelas, X., Binello, E. \u0026amp; Ramirez, S. Age-dependent glioma progression and functional decline in a syngeneic Murine model: Host vulnerabilities and opportunities for targeted intervention. \u003cem\u003eNeurooncol. Adv.\u003c/em\u003e (2025) doi:10.1093/noajnl/vdaf222.\u003c/li\u003e\n\u003cli\u003eHiraoka, K. \u003cem\u003eet al.\u003c/em\u003e Pattern of c-Fos expression induced by tail suspension test in the mouse brain. \u003cem\u003eHeliyon\u003c/em\u003e \u003cstrong\u003e3\u003c/strong\u003e, e00316 (2017).\u003c/li\u003e\n\u003cli\u003eShin, L. M. \u0026amp; Liberzon, I. The neurocircuitry of fear, stress, and anxiety disorders. \u003cem\u003eNeuropsychopharmacology\u003c/em\u003e \u003cstrong\u003e35\u003c/strong\u003e, 169\u0026ndash;191 (2010).\u003c/li\u003e\n\u003cli\u003ePadilla-Coreano, N. \u003cem\u003eet al.\u003c/em\u003e Direct ventral hippocampal-prefrontal input is required for anxiety-related neural activity and behavior. \u003cem\u003eNeuron\u003c/em\u003e \u003cstrong\u003e89\u003c/strong\u003e, 857\u0026ndash;866 (2016).\u003c/li\u003e\n\u003cli\u003eLim, S. C., Fusi, S. \u0026amp; Hen, R. Ventral CA1 population codes for anxiety. \u003cem\u003ebioRxivorg\u003c/em\u003e (2023) doi:10.1101/2023.09.25.559358.\u003c/li\u003e\n\u003cli\u003eEuston, D. R., Gruber, A. J. \u0026amp; McNaughton, B. L. The role of medial prefrontal cortex in memory and decision making. \u003cem\u003eNeuron\u003c/em\u003e \u003cstrong\u003e76\u003c/strong\u003e, 1057\u0026ndash;1070 (2012).\u003c/li\u003e\n\u003cli\u003eKim, E. \u003cem\u003eet al.\u003c/em\u003e Prefrontal cortex astrocytes modulate distinct neuronal populations to control anxiety-like behavior. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 7819 (2025).\u003c/li\u003e\n\u003cli\u003eCho, W.-H. \u003cem\u003eet al.\u003c/em\u003e Hippocampal astrocytes modulate anxiety-like behavior. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 6536 (2022).\u003c/li\u003e\n\u003cli\u003eLi, J.-T. \u003cem\u003eet al.\u003c/em\u003e Astrocytes in the ventral hippocampus bidirectionally regulate innate and stress-induced anxiety-like behaviors in male mice. \u003cem\u003eAdv. Sci. (Weinh.)\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, e2400354 (2024).\u003c/li\u003e\n\u003cli\u003eShepherd, J. K., Grewal, S. S., Fletcher, A., Bill, D. J. \u0026amp; Dourish, C. T. Behavioural and pharmacological characterisation of the elevated \u0026lsquo;zero-maze\u0026rsquo; as an animal model of anxiety. \u003cem\u003ePsychopharmacology \u003c/em\u003e\u003cstrong\u003e116\u003c/strong\u003e, 56\u0026ndash;64 (1994).\u003c/li\u003e\n\u003cli\u003ePretanvil, J.-A., Salinas, I. Q. \u0026amp; Piccioni, D. E. Glioblastoma in the elderly: treatment patterns and survival. \u003cem\u003eCNS Oncol.\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, 19\u0026ndash;28 (2017).\u003c/li\u003e\n\u003cli\u003eBlanchard, D. C., Blanchard, R. J., Tom, P. \u0026amp; Rodgers, R. J. Diazepam changes risk assessment in an anxiety/defense test battery. \u003cem\u003ePsychopharmacology (Berl.)\u003c/em\u003e \u003cstrong\u003e101\u003c/strong\u003e, 511\u0026ndash;518 (1990).\u003c/li\u003e\n\u003cli\u003eBlanchard, R. J., Blanchard, D. C. \u0026amp; Hori, K. An ethoexperimental approach to the study of defense. in \u003cem\u003eEthoexperimental Approaches to the Study of Behavior\u003c/em\u003e vol. 739 114\u0026ndash;136 (Springer Netherlands, Dordrecht, 1989).\u003c/li\u003e\n\u003cli\u003eJimenez, J. C. \u003cem\u003eet al.\u003c/em\u003e Contextual fear memory retrieval by correlated ensembles of ventral CA1 neurons. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 3492 (2020).\u003c/li\u003e\n\u003cli\u003eJean-Richard-Dit-Bressel, P., Clifford, C. W. G. \u0026amp; McNally, G. P. Analyzing event-related transients: Confidence intervals, permutation tests, and consecutive thresholds. \u003cem\u003eFront. Mol. Neurosci.\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 14 (2020).\u003c/li\u003e\n\u003cli\u003eDing, S.-L. \u003cem\u003eet al.\u003c/em\u003e Distinct transcriptomic cell types and neural circuits of the subiculum and prosubiculum along the dorsal-ventral axis. \u003cem\u003eCell Rep.\u003c/em\u003e \u003cstrong\u003e31\u003c/strong\u003e, 107648 (2020).\u003c/li\u003e\n\u003cli\u003ePhipson, B. \u003cem\u003eet al.\u003c/em\u003e Propeller: Testing for differences in cell type proportions in single cell data. \u003cem\u003eBioinformatics\u003c/em\u003e \u003cstrong\u003e38\u003c/strong\u003e, 4720\u0026ndash;4726 (2022).\u003c/li\u003e\n\u003cli\u003eHe, L. \u003cem\u003eet al.\u003c/em\u003e NEBULA is a fast negative binomial mixed model for differential or co-expression analysis of large-scale multi-subject single-cell data. \u003cem\u003eCommun. Biol.\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e, 629 (2021).\u003c/li\u003e\n\u003cli\u003eSupek, F., Bo\u0026scaron;njak, M., \u0026Scaron;kunca, N. \u0026amp; \u0026Scaron;muc, T. REVIGO summarizes and visualizes long lists of gene ontology terms. \u003cem\u003ePLoS One\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, e21800 (2011).\u003c/li\u003e\n\u003cli\u003eStringer, C. \u0026amp; Pachitariu, M. Cellpose3: one-click image restoration for improved cellular segmentation. \u003cem\u003eNat. Methods\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e, 592\u0026ndash;599 (2025).\u003c/li\u003e\n\u003cli\u003ePisanello, M. \u003cem\u003eet al.\u003c/em\u003e The three-dimensional signal collection field for fiber photometry in brain tissue. \u003cem\u003eFront. Neurosci.\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 82 (2019).\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-8725420/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8725420/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Age is the strongest determinant of glioblastoma incidence and outcome, yet how aging shapes the functional consequences of tumor burden across the brain remains unclear. Glioblastoma is accompanied by cognitive and affective dysfunction that implicates disruption of neural networks beyond the tumor core, particularly in the aged state. Here, we show that aging modifies how glioma burden progressively reshapes extratumoral brain circuit dynamics. Neurons largely retained behaviorally-aligned activity early in disease, with neuronal alterations such as diminished neuronal event magnitudes emerging with tumor progression in a region- and age-dependent manner. In contrast, astrocytic activity was disproportionately affected by aging and tumor burden, with age amplifying astrocytic engagement under control conditions and glioma burden selectively weakening astrocytic responsiveness, coordination, and long-range synchrony, most prominently in aged hosts and at earlier stages of tumor progression. These functional disruptions were accompanied by transcriptional remodeling in extratumoral regions, with astrocytes exhibiting inflammatory activation and suppression of circuit-supportive programs, and neurons showing complementary, region- and age-dependent alterations in genes related to synaptic function and activity. Together, these findings position astrocytic dysfunction as a dominant feature of age-associated extratumoral circuit vulnerability in glioblastoma and highlight glial mechanisms as key contributors to progressive functional decline in the aging brain.","manuscriptTitle":"Aging potentiates glioma-driven remodeling of cortico-hippocampal neuron-astrocyte dynamics","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-02-11 06:35:17","doi":"10.21203/rs.3.rs-8725420/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"ddb51d22-1649-4f4d-9111-6f5196ec4c3b","owner":[],"postedDate":"February 11th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":62639503,"name":"Biological sciences/Cancer/Cancer imaging"},{"id":62639504,"name":"Biological sciences/Neuroscience/Learning and memory/Hippocampus"},{"id":62639505,"name":"Biological sciences/Neuroscience/Glial biology/Astrocyte"},{"id":62639506,"name":"Biological sciences/Neuroscience/Neural ageing"}],"tags":[],"updatedAt":"2026-03-27T11:31:42+00:00","versionOfRecord":[],"versionCreatedAt":"2026-02-11 06:35:17","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8725420","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8725420","identity":"rs-8725420","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}