Brain HealthResearch PaperOpen Access

Boosting Brain Blood Flow Rescues the Brain's Waste-Clearance System During Sepsis

Cerebral hypoperfusion drives glymphatic dysfunction in systemic inflammation — and restoring blood flow with levosimendan reverses it.

Thursday, May 14, 2026 0 views
Published in Adv Sci (Weinh)
A microscopy image of mouse brain cross-section showing glowing perivascular spaces around blood vessels with fluorescent tracer flowing through brain tissue, on a dark background in a neuroscience lab

Summary

Researchers discovered that systemic inflammation caused by bacterial endotoxin (LPS) disrupts the brain's glymphatic waste-clearance system in a two-phase pattern: first suppressing, then dramatically over-driving CSF inflow while blocking outflow. The culprit appears to be reduced cerebral blood flow (hypoperfusion). Using the cardiac drug levosimendan to boost brain perfusion, the team restored normal glymphatic dynamics, improved clearance of amyloid-beta, reduced neuroinflammation, and reversed behavioral deficits in mice — all without breaching the blood-brain barrier. The key molecular mediator was AQP4, a water channel that becomes mislocalized during inflammation and is rescued by restored blood flow.

Detailed Summary

The glymphatic system — the brain's perivascular waste-clearance network driven by the water channel AQP4 on astrocyte endfeet — is increasingly recognized as central to brain health and neurodegeneration. Disruption of this system has been implicated in Alzheimer's disease, traumatic brain injury, and stroke. Sepsis-associated encephalopathy (SAE), which affects up to 70% of severe sepsis patients and carries high mortality and long-term cognitive consequences, had been suspected to involve glymphatic dysfunction, but the temporal dynamics, mechanisms, and therapeutic implications were previously unknown.

This study used a mouse model of systemic inflammation (intraperitoneal LPS at 5 mg/kg) and developed novel near-infrared II (NIR-II) fluorescent probes — including a 66.9 kDa biomimetic fluorescent protein (BSA@IR-780) — to non-invasively image CSF flow through the glymphatic system in living mice. Imaging was performed at 1, 3, 24, and 72 hours post-LPS. At 1 hour, glymphatic influx around the middle cerebral artery (MCA) was significantly reduced compared to controls. By 3 hours, influx reversed and became progressively elevated, peaking at 24 hours before returning to baseline at 72 hours (p<0.05 to p<0.001 across timepoints). Simultaneously, CSF efflux through the olfactory bulb/nasopharyngeal lymphatic plexus was continuously suppressed from 1 to 72 hours. Brain homogenate quantification at 30 minutes post-tracer confirmed significantly elevated tracer levels in LPS-treated mice at 24 hours. Critically, at the 5 mg/kg dose used, Evans Blue extravasation assays confirmed BBB integrity was preserved, demonstrating that elevated glymphatic influx was not an artifact of barrier breakdown.

Cerebral blood flow (CBF) measured by laser Doppler showed an inverse relationship with glymphatic influx trends after LPS: CBF dropped significantly at 3 hours (matching the influx surge) and remained depressed. The investigators then tested levosimendan — a calcium-sensitizing cardiac inotrope and vasodilator approved in many countries for acute heart failure — as a means to restore CBF. Levosimendan treatment significantly increased CBF, normalized glymphatic influx and efflux dynamics, and restored amyloid-beta clearance (assessed with a small-molecule NIR-II probe mimicking Aβ). Crucially, when levosimendan was combined with bilateral carotid artery stenosis surgery to re-induce hypoperfusion, its glymphatic benefits were completely abolished — definitively establishing cerebral hypoperfusion as the causal mediator of LPS-induced glymphatic dysfunction.

Mechanistically, AQP4 normally concentrates at astrocytic endfeet surrounding blood vessels (perivascular polarization). LPS caused significant AQP4 depolarization — redistribution away from endfeet — which would impair the directional CSF-ISF exchange the glymphatic system depends on. Levosimendan's restoration of CBF prevented this depolarization, maintaining AQP4 at endfeet. When AQP4 was pharmacologically inhibited with TGN-020, levosimendan's benefits on Aβ clearance and neuroinflammation suppression were completely blocked, confirming AQP4 as the essential downstream effector. Neuroinflammation markers (microglial activation, pro-inflammatory cytokines) were elevated by LPS and significantly attenuated by levosimendan in an AQP4-dependent manner. Behavioral testing showed levosimendan also significantly ameliorated LPS-induced neurological deficits.

The findings establish a mechanistic chain: systemic inflammation → cerebral hypoperfusion → AQP4 depolarization → glymphatic dysfunction → impaired waste clearance and neuroinflammation → SAE. This is the first study to temporally characterize glymphatic dynamics across 72 hours of systemic inflammation using in vivo NIR-II imaging, and the first to causally link hypoperfusion to glymphatic failure and test a clinically available drug to reverse it. The identification of levosimendan as a potential SAE therapeutic is particularly notable given its existing clinical safety profile.

Key Findings

  • LPS (5 mg/kg) caused a biphasic glymphatic influx pattern: suppressed at 1h, then significantly elevated at 3h and 24h (p<0.001), returning to baseline at 72h, while CSF efflux via olfactory bulb was continuously suppressed across all 72 hours
  • BBB integrity was confirmed intact at 5 mg/kg LPS via Evans Blue assay, ruling out barrier breakdown as the cause of elevated glymphatic influx
  • Cerebral blood flow showed an inverse temporal relationship with glymphatic influx — CBF dropped significantly at ~3h post-LPS, coinciding with the influx surge, consistent with a hypoperfusion-driven mechanism
  • Levosimendan restored CBF and normalized both glymphatic influx and efflux; when CBF restoration was negated by bilateral carotid stenosis, all glymphatic benefits were abolished, causally implicating hypoperfusion
  • LPS induced AQP4 depolarization (redistribution away from astrocytic endfeet) that was prevented by levosimendan-mediated CBF restoration, implicating AQP4 mislocalization as the molecular mechanism
  • AQP4 inhibitor TGN-020 completely blocked levosimendan's improvements in amyloid-beta clearance and neuroinflammation suppression, confirming AQP4 as the essential effector in this pathway
  • Levosimendan significantly attenuated LPS-induced neuroinflammatory markers and neurological behavioral deficits in mice, demonstrating therapeutic relevance beyond glymphatic mechanics alone

Methodology

Wild-type C57BL/6J mice received intraperitoneal LPS (1 or 5 mg/kg) to model systemic inflammation; glymphatic function was assessed non-invasively at 1, 3, 24, and 72 hours post-LPS using novel NIR-II fluorescent probes (including 66.9 kDa BSA@IR-780 for large-molecule CSF tracing and small-molecule probes for Aβ mimicry) injected into the cisterna magna (7 µL at 1 µL/min, n=6–8/group). Cerebral blood flow was measured by laser Doppler; BBB integrity by Evans Blue; AQP4 localization by immunofluorescence. Interventions included levosimendan administration and bilateral carotid artery stenosis surgery as a causal test of hypoperfusion. Statistical analyses included two-way repeated-measures ANOVA, one-way ANOVA, and Student's t-tests with data presented as mean ± SEM.

Study Limitations

This study was conducted entirely in mice, and translation of levosimendan's glymphatic effects to human sepsis patients requires clinical validation, particularly given species differences in glymphatic anatomy and AQP4 distribution. The NIR-II probe imaging approach, while innovative, is currently a research tool not available for human use. The authors do not report any conflicts of interest, and the study was funded by the National Natural Science Foundation of China and regional institutions.

Enjoyed this summary?

Get the latest longevity research delivered to your inbox every week.