Magnetic Resonance Elastography Detects Brain Stiffness Reductions during Acute Lipopolysaccharide-Induced Neuroinflammation

Introduction: : Neuroinflammation, a defining characteristic of neurodegenerative diseases (Bergs, 2024), is a natural immune response caused by pathogens and/or tissue damage. Bacterial and viral infections that trigger neuroinflammation—although typically self-resolving—have been associated with an increased risk of developing neurodegenerative diseases, suggesting that inflammation may persist unresolved in some individuals (Bruno, 2023; Onisiforou, 2024; Cossu, 2018). Given that neuroimmune cell infiltration alters tissue mechanics, magnetic resonance elastography (MRE)—a noninvasive technique that measures viscoelastic properties—shows promise for monitoring these changes in vivo. While studies have shown brain stiffness reductions in diseases with chronic neuroinflammation where tissue ultimately degenerates (Murphy, 2011; Streitberger, 2012; Wuerfel, 2010), MRE’s sensitivity to acute neuroinflammation remains unclear. Advancing MRE for therapeutic monitoring requires mechanistic studies validating MRE-derived mechanical properties with specific changes in brain mechanobiology, best achieved through animal models.
Few MRE studies have examined acute neuroinflammatory events. Riek et al. (2012) reported a transient decrease in whole brain stiffness during peak inflammation in a rodent experimental autoimmune encephalomyelitis (EAE) model over the course of two weeks. In our study, we used a systemic bacterial infection model using lipopolysaccharides (LPS) and employed high-resolution MRE and evaluation of cellularity to determine sensitivity to inflammation during a shorter time frame. We hypothesized that brain stiffness will decrease at the peak of neuroinflammation due to infiltration of neuroimmune cells, ECM remodeling, and edema, with a return to baseline upon subsided immune activity.

Materials and

Methods: : Eleven adult, female Long-Evans rats were used for this study. Rats were MRE scanned at baseline (TP1), then received either systemic injection with saline or E. Coli-derived LPS (saline: n=5; LPS: n=6; Figure 1). Rats were rescanned using the same protocol at two-days post-injection during peak neuroinflammation (TP2) and at five-days post-injection during the established recovery period (TP3). MRE and anatomical data were acquired using a custom EPI-MRE sequence with a 9.4T Bruker Biospec MRI scanner. A piezoelectric actuator was attached to a custom bite bar to deliver 700 Hz vibrations to anesthetized rats maintained on 1-3% isoflurane in oxygen and body temperature at 32-36°C for a total 1-hour scan protocol (Milbocker et al. 2024). MRE imaging parameters included: TR/TE = 4000/60 ms; 20x20 mm FOV; 80x80 matrix; 40 slices, 0.5 mm thick; 0.25x0.25x0.50 mm3 resolution. Stiffness maps were estimated from displacement images with nonlinear inversion formulated without boundary conditions (McGarry, 2012; Kurtz, 2024).
Histological analysis was performed on a separate cohort of rats euthanized 1 day post-injection (saline: n=8; LPS: n=9). Brains were extracted, fixed, cryoprotected and sectioned at 40 μm in the coronal plane. Sections containing the dorsal hippocampus were labelled with primary antibodies for versican—a large proteoglycan—and CS56—an antibody that labels specific sulfation patterns of glycoprotein chains.

Results, Conclusions, and Discussions:: This study demonstrates that MRE is highly sensitive to rapid, inflammation-induced changes in brain tissue mechanics. We show that LPS-induced acute neuroinflammation causes a reduction in brain stiffness, returning to baseline within 5 days, as neuroinflammation resolves. Repeated measures ANOVA tests were performed to compare whole brain stiffness pre-injection and 2 days post-injection within rat, revealing brain stiffness decreased by 8.9% in LPS-treated rats (ηp2=0.67; µTP1=6.58 kPa ±0.50; µTP2=5.99±0.56 kPa; p< 0.01). A second set of repeated measures ANOVA tests were performed to compare pre-injection and 5 days post-injection within rat. After recovery, brain stiffness returned to pre-treatment levels in LPS-treated rats (µTP3=6.52±0.61 kPa; p=0.194, ηp2=0.179; Figure 2). No significant differences were observed in the saline-treated group. These results highlight MRE’s ability to detect dynamic, very short-term changes in tissue stiffness associated with acute neuroinflammation.
Softening is likely the result of the influx of non-integrated cellular components in the tissue. One-way ANOVA comparing the optical density of ECM components in hippocampal CA1 between treatment groups revealed no significant difference in LPS compared to saline-treated rats during peak inflammation for neither versican (saline:0.676±0.330; LPS:0.849±0.107; p=0.187) or CS56 (saline: 0.753±0.186; LPS: 0.841± 0.144, p=0.307) (Figure 3). However, comparison of the coefficient of variance for each ECM component showed that LPS treatment reduced the variance between subjects, ultimately leading to a larger number of rats with a high density of Versican and CS56 in dorsal hippocampal CA1. The variability in reported Versican density between rats decreased from 49% to 13%, while the variability in CS56 was reduced from 25% to 17%. Analysis of neuroimmune cells is ongoing.
This study established that acute inflammation softens brain tissue, which may play a role in tissue damage during chronic inflammation, and that MRE can monitor these events in vivo. We also demonstrated that an increase of proteoglycan density may contribute to tissue softening. Further research is needed to determine how short-term softening from acute neuroinflammation might contribute to long-term changes, especially in older or vulnerable populations.

Acknowledgements and/or References (Optional)::