The Networking Brain: How Extracellular Matrix, Cellular Networks, and Vasculature Shape the In Vivo Mechanical Properties of the Brain.

Mechanically, the brain is characterized by both solid and fluid properties. The resulting unique material behavior fosters proliferation, differentiation, and repair of cellular and vascular networks, and optimally protects them from damaging shear forces. Magnetic resonance elastography (MRE) is a noninvasive imaging technique that maps the mechanical properties of the brain in vivo. MRE studies have shown that abnormal processes such as neuronal degeneration, demyelination, inflammation, and vascular leakage lead to tissue softening. In contrast, neuronal proliferation, cellular network formation, and higher vascular pressure result in brain stiffening. In addition, brain viscosity has been reported to change with normal blood perfusion variability and brain maturation as well as disease conditions such as tumor invasion. In this article, the contributions of the neuronal, glial, extracellular, and vascular networks are discussed to the coarse-grained parameters determined by MRE. This reductionist multi-network model of brain mechanics helps to explain many MRE observations in terms of microanatomical changes and suggests that cerebral viscoelasticity is a suitable imaging marker for brain disease.

Cortical matrix remodeling as a hallmark of relapsing-remitting neuroinflammation in MR elastography and quantitative MRI.

Multiple sclerosis (MS) is a chronic neuroinflammatory disease that involves both white and gray matter. Although gray matter damage is a major contributor to disability in MS patients, conventional clinical magnetic resonance imaging (MRI) fails to accurately detect gray matter pathology and establish a clear correlation with clinical symptoms. Using magnetic resonance elastography (MRE), we previously reported global brain softening in MS and experimental autoimmune encephalomyelitis (EAE). However, it needs to be established if changes of the spatiotemporal patterns of brain tissue mechanics constitute a marker of neuroinflammation. Here, we use advanced multifrequency MRE with tomoelastography postprocessing to investigate longitudinal and regional inflammation-induced tissue changes in EAE and in a small group of MS patients. Surprisingly, we found reversible softening in synchrony with the EAE disease course predominantly in the cortex of the mouse brain. This cortical softening was associated neither with a shift of tissue water compartments as quantified by T2-mapping and diffusion-weighted MRI, nor with leukocyte infiltration as seen by histopathology. Instead, cortical softening correlated with transient structural remodeling of perineuronal nets (PNNs), which involved abnormal chondroitin sulfate expression and microgliosis. These mechanisms also appear to be critical in humans with MS, where tomoelastography for the first time demonstrated marked cortical softening. Taken together, our study shows that neuroinflammation (i) critically affects the integrity of PNNs in cortical brain tissue, in a reversible process that correlates with disease disability in EAE, (ii) reduces the mechanical integrity of brain tissue rather than leading to water accumulation, and (iii) shows similar spatial patterns in humans and mice. These results raise the prospect of leveraging MRE and quantitative MRI for MS staging and monitoring treatment in affected patients.

Brain inflammation induces alterations in glycosaminoglycan metabolism and subsequent changes in CS-4S and hyaluronic acid.

It remains uncertain how brain glycosaminoglycans (GAGs) contribute to the progression of inflammatory disorders like multiple sclerosis (MS). We investigated here neuroinflammation-mediated changes in GAG composition and metabolism using the mouse model of experimental autoimmune encephalomyelitis (EAE) and sham-immunized mice as controls. Cerebellum, mid- and forebrain at different EAE phases were investigated using gene expression analysis (microarray and RT-qPCR) as well as HPLC quantification of CS and hyaluronic acid (HA). The cerebellum was the most affected brain region showing a downregulation of Bcan, Cspg5, and an upregulation of Dse, Gusb, Hexb, Dcn and Has2 at peak EAE. Upregulation of genes involved in GAG degradation as well as synthesis of HA and decorin persisted from onset to peak, and diminished at remission, suggesting a severity-related decrease in CS and increments in HA. Relative disaccharide quantification confirmed a 3.6 % reduction of CS-4S at peak and a normalization during remission, while HA increased in both phases by 26.1 % and 17.6 %, respectively. Early inflammatory processes led to altered GAG metabolism in early EAE stages and subsequent partially reversible changes in CS-4S and in HA. Targeting early modifications in CS could potentially mitigate progression of EAE/MS.

Mechanical behavior of the hippocampus and corpus callosum: An attempt to reconcile ex vivo with in vivo and micro with macro properties.

Mechanical properties of brain tissue are very complex and vary with the species, region, method, and dynamic range, and between in vivo and ex vivo measurements. To reconcile this variability, we investigated in vivo and ex vivo stiffness properties of two distinct regions in the human and mouse brain - the hippocampus (HP) and the corpus callosum (CC) - using different methods. Under quasi-static conditions, we examined ex vivo murine HP and CC by atomic force microscopy (AFM). Between 16 and 40Hz, we investigated the in vivo brains of healthy volunteers by magnetic resonance elastography (MRE) in a 3-T clinical scanner. At high-frequency stimulation between 1000 and 1400Hz, we investigated the murine HP and CC ex vivo and in vivo with MRE in a 7-T preclinical system. HP and CC showed pronounced stiffness dispersion, as reflected by a factor of 32-36 increase in shear modulus from AFM to low-frequency human MRE and a 25-fold higher shear wave velocity in murine MRE than in human MRE. At low frequencies, HP was softer than CC, in both ex vivo mouse specimens (p < 0.05) and in vivo human brains (p < 0.01) while, at high frequencies, CC was softer than HP under in vivo (p < 0.01) and ex vivo (p < 0.05) conditions. The standard linear solid model comprising three elements reproduced the observed HP and CC stiffness dispersions, while other two- and three-element models failed. Our results indicate a remarkable consistency of brain stiffness across species, ex vivo and in vivo states, and different measurement techniques when marked viscoelastic dispersion properties combining equilibrium and non-equilibrium mechanical elements are considered.

Mechanical properties of murine hippocampal subregions investigated by atomic force microscopy and in vivo magnetic resonance elastography.

The hippocampus is a very heterogeneous brain structure with different mechanical properties reflecting its functional variety. In particular, adult neurogenesis in rodent hippocampus has been associated with specific viscoelastic properties in vivo and ex vivo. Here, we study the microscopic mechanical properties of hippocampal subregions using ex vivo atomic force microscopy (AFM) in correlation with the expression of GFP in presence of the nestin promoter, providing a marker of neurogenic activity. We further use magnetic resonance elastography (MRE) to investigate whether in vivo mechanical properties reveal similar spatial patterns, however, on a much coarser scale. AFM showed that tissue stiffness increases with increasing distance from the subgranular zone (p = 0.0069), and that stiffness is 39% lower in GFP than non-GFP regions (p = 0.0004). Consistently, MRE showed that dentate gyrus is, on average, softer than Ammon´s horn (shear wave speed = 3.2 ± 0.2 m/s versus 4.4 ± 0.3 m/s, p = 0.01) with another 3.4% decrease towards the subgranular zone (p = 0.0001). The marked reduction in stiffness measured by AFM in areas of high neurogenic activity is consistent with softer MRE values, indicating the sensitivity of macroscopic mechanical properties in vivo to micromechanical structures as formed by the neurogenic niche of the hippocampus.

Different Impact of Gadopentetate and Gadobutrol on Inflammation-Promoted Retention and Toxicity of Gadolinium Within the Mouse Brain.

OBJECTIVES: Using a murine model of multiple sclerosis, we previously showed that repeated administration of gadopentetate dimeglumine led to retention of gadolinium (Gd) within cerebellar structures and that this process was enhanced with inflammation. This study aimed to compare the kinetics and retention profiles of Gd in inflamed and healthy brains after application of the macrocyclic Gd-based contrast agent (GBCA) gadobutrol or the linear GBCA gadopentetate. Moreover, potential Gd-induced neurotoxicity was investigated in living hippocampal slices ex vivo. MATERIALS AND METHODS: Mice at peak of experimental autoimmune encephalomyelitis (EAE; n = 29) and healthy control mice (HC; n = 24) were exposed to a cumulative dose of 20 mmol/kg bodyweight of either gadopentetate dimeglumine or gadobutrol (8 injections of 2.5 mmol/kg over 10 days). Magnetic resonance imaging (7 T) was performed at baseline as well as at day 1, 10, and 40 post final injection (pfi) of GBCAs. Mice were sacrificed after magnetic resonance imaging and brain and blood Gd content was assessed by laser ablation-inductively coupled plasma (ICP)-mass spectrometry (MS) and ICP-MS, respectively. In addition, using chronic organotypic hippocampal slice cultures, Gd-induced neurotoxicity was addressed in living brain tissue ex vivo, both under control or inflammatory (tumor necrosis factor α [TNF-α] at 50 ng/µL) conditions. RESULTS: Neuroinflammation promoted a significant decrease in T1 relaxation times after multiple injections of both GBCAs as shown by quantitative T1 mapping of EAE brains compared with HC. This corresponded to higher Gd retention within the EAE brains at 1, 10, and 40 days pfi as determined by laser ablation-ICP-MS. In inflamed cerebellum, in particular in the deep cerebellar nuclei (CN), elevated Gd retention was observed until day 40 after last gadopentetate application (CN: EAE vs HC, 55.06 ± 0.16 µM vs 30.44 ± 4.43 µM). In contrast, gadobutrol application led to a rather diffuse Gd content in the inflamed brains, which strongly diminished until day 40 (CN: EAE vs HC, 0.38 ± 0.08 µM vs 0.17 ± 0.03 µM). The analysis of cytotoxic effects of both GBCAs using living brain tissue revealed an elevated cell death rate after incubation with gadopentetate but not gadobutrol at 50 mM. The cytotoxic effect due to gadopentetate increased in the presence of the inflammatory mediator TNF-α (with vs without TNF-α, 3.15% ± 1.18% vs 2.17% ± 1.14%; P = 0.0345). CONCLUSIONS: In the EAE model, neuroinflammation promoted increased Gd retention in the brain for both GBCAs. Whereas in the inflamed brains, efficient clearance of macrocyclic gadobutrol during the investigated time period was observed, the Gd retention after application of linear gadopentetate persisted over the entire observational period. Gadopentetate but not gadubutrol appeared to be neurotoxic in an ex vivo paradigm of neuronal inflammation.

Liquid-Liver Phantom: Mimicking the Viscoelastic Dispersion of Human Liver for Ultrasound- and MRI-Based Elastography.

OBJECTIVES: Tissue stiffness can guide medical diagnoses and is exploited as an imaging contrast in elastography. However, different elastography devices show different liver stiffness values in the same subject, hindering comparison of values and establishment of system-independent thresholds for disease detection. There is a need for standardized phantoms that specifically address the viscosity-related dispersion of stiffness over frequency. To improve standardization of clinical elastography across devices and platforms including ultrasound and magnetic resonance imaging (MRI), a comprehensively characterized phantom is introduced that mimics the dispersion of stiffness of the human liver and can be generated reproducibly. MATERIALS AND METHODS: The phantom was made of linear polymerized polyacrylamide (PAAm) calibrated to the viscoelastic properties of healthy human liver in vivo as reported in the literature. Stiffness dispersion was analyzed using the 2-parameter springpot model fitted to the dispersion of shear wave speed of PAAm, which was measured by shear rheometry, ultrasound-based time-harmonic elastography, clinical magnetic resonance elastography (MRE), and tabletop MRE in the frequency range of 5 to 3000 Hz. Imaging parameters for ultrasound and MRI, reproducibility, aging behavior, and temperature dependency were assessed. In addition, the frequency bandwidth of shear wave speed of clinical elastography methods (Aplio i900, Canon; Acuson Sequoia, Siemens; FibroScan, EchoSense) was characterized. RESULTS: Within the entire frequency range analyzed in this study, the PAAm phantom reproduced well the stiffness dispersion of human liver in vivo despite its fluid properties under static loading (springpot stiffness parameter, 2.14 [95% confidence interval, 2.08-2.19] kPa; springpot powerlaw exponent, 0.367 [95% confidence interval, 0.362-0.373]). Imaging parameters were close to those of liver in vivo with only slight variability in stiffness values of 0.5% (0.4%, 0.6%), 4.1% (3.9%, 4.5%), and -0.63% (-0.67%, -0.58%), respectively, between batches, over a 6-month period, and per °C increase in temperature. CONCLUSIONS: The liquid-liver phantom has useful properties for standardization and development of liver elastography. First, it can be used across clinical and experimental elastography devices in ultrasound and MRI. Second, being a liquid, it can easily be adapted in size and shape to specific technical requirements, and by adding inclusions and scatterers. Finally, because the phantom is based on noncrosslinked linear PAAm constituents, it is easy to produce, indicating potential widespread use among researchers and vendors to standardize liver stiffness measurements.

Sexual Dimorphism in Extracellular Matrix Composition and Viscoelasticity of the Healthy and Inflamed Mouse Brain.

Magnetic resonance elastography (MRE) has revealed sexual dimorphism in brain stiffness in healthy individuals and multiple sclerosis (MS) patients. In an animal model of MS, named experimental autoimmune encephalomyelitis (EAE), we have previously shown that inflammation-induced brain softening was associated with alterations of the extracellular matrix (ECM). However, it remained unclear whether the brain ECM presents sex-specific properties that can be visualized by MRE. Therefore, here we aimed at quantifying sexual dimorphism in brain viscoelasticity in association with ECM changes in healthy and inflamed brains. Multifrequency MRE was applied to the midbrain of healthy and EAE mice of both sexes to quantitatively map regional stiffness. To define differences in brain ECM composition, the gene expression of the key basement membrane components laminin (Lama4, Lama5), collagen (Col4a1, Col1a1), and fibronectin (Fn1) were investigated by RT-qPCR. We showed that the healthy male cortex expressed less Lama4, Lama5, and Col4a1, but more Fn1 (all p < 0.05) than the healthy female cortex, which was associated with 9% softer properties (p = 0.044) in that region. At peak EAE cortical softening was similar in both sexes compared to healthy tissue, with an 8% difference remaining between males and females (p = 0.006). Cortical Lama4, Lama5 and Col4a1 expression increased 2 to 3-fold in EAE in both sexes while Fn1 decreased only in males (all p < 0.05). No significant sex differences in stiffness were detected in other brain regions. In conclusion, sexual dimorphism in the ECM composition of cortical tissue in the mouse brain is reflected by in vivo stiffness measured with MRE and should be considered in future studies by sex-specific reference values.

Contribution of Tissue Inflammation and Blood-Brain Barrier Disruption to Brain Softening in a Mouse Model of Multiple Sclerosis.

Neuroinflammatory processes occurring during multiple sclerosis cause disseminated softening of brain tissue, as quantified by in vivo magnetic resonance elastography (MRE). However, inflammation-mediated tissue alterations underlying the mechanical integrity of the brain remain unclear. We previously showed that blood-brain barrier (BBB) disruption visualized by MRI using gadolinium-based contrast agent (GBCA) does not correlate with tissue softening in active experimental autoimmune encephalomyelitis (EAE). However, it is unknown how confined BBB changes and other inflammatory processes may determine local elasticity changes. Therefore, we aim to elucidate which inflammatory hallmarks are determinant for local viscoelastic changes observed in EAE brains. Hence, novel multifrequency MRE was applied in combination with GBCA-based MRI or very small superparamagnetic iron oxide particles (VSOPs) in female SJL mice with induced adoptive transfer EAE (n = 21). VSOPs were doped with europium (Eu-VSOPs) to facilitate the post-mortem analysis. Accumulation of Eu-VSOPs, which was previously demonstrated to be sensitive to immune cell infiltration and ECM remodeling, was also found to be independent of GBCA enhancement. Following registration to a reference brain atlas, viscoelastic properties of the whole brain and areas visualized by either Gd or VSOP were quantified. MRE revealed marked disseminated softening across the whole brain in mice with established EAE (baseline: 3.1 ± 0.1 m/s vs. EAE: 2.9 ± 0.2 m/s, p < 0.0001). A similar degree of softening was observed in sites of GBCA enhancement i.e., mainly within cerebral cortex and brain stem (baseline: 3.3 ± 0.4 m/s vs. EAE: 3.0 ± 0.5 m/s, p = 0.018). However, locations in which only Eu-VSOP accumulated, mainly in fiber tracts (baseline: 3.0 ± 0.4 m/s vs. EAE: 2.6 ± 0.5 m/s, p = 0.023), softening was more pronounced when compared to non-hypointense areas (percent change of stiffness for Eu-VSOP accumulation: -16.81 ± 16.49% vs. for non-hypointense regions: -5.85 ± 3.81%, p = 0.048). Our findings suggest that multifrequency MRE is sensitive to differentiate between local inflammatory processes with a strong immune cell infiltrate that lead to VSOP accumulation, from disseminated inflammation and BBB leakage visualized by GBCA. These pathological events visualized by Eu-VSOP MRI and MRE may include gliosis, macrophage infiltration, alterations of endothelial matrix components, and/or extracellular matrix remodeling. MRE may therefore represent a promising imaging tool for non-invasive clinical assessment of different pathological aspects of neuroinflammation.

Changes in Liver Mechanical Properties and Water Diffusivity During Normal Pregnancy Are Driven by Cellular Hypertrophy.

During pregnancy, the body's hyperestrogenic state alters hepatic metabolism and synthesis. While biochemical changes related to liver function during normal pregnancy are well understood, pregnancy-associated alterations in biophysical properties of the liver remain elusive. In this study, we investigated 26 ex vivo fresh liver specimens harvested from pregnant and non-pregnant rats by diffusion-weighted imaging (DWI) and magnetic resonance elastography (MRE) in a 0.5-Tesla compact magnetic resonance imaging (MRI) scanner. Water diffusivity and viscoelastic parameters were compared with histological data and blood markers. We found livers from pregnant rats to have (i) significantly enlarged hepatocytes (26 ± 15%, p < 0.001), (ii) increased liver stiffness (12 ± 15%, p = 0.012), (iii) decreased viscosity (-23 ± 14%, p < 0.001), and (iv) increased water diffusivity (12 ± 11%, p < 0.001). In conclusion, increased stiffness and reduced viscosity of the liver during pregnancy are mainly attributable to hepatocyte enlargement. Hypertrophy of liver cells imposes fewer restrictions on intracellular water mobility, resulting in a higher hepatic water diffusion coefficient. Collectively, MRE and DWI have the potential to inform on structural liver changes associated with pregnancy in a clinical context.

The influence of body temperature on tissue stiffness, blood perfusion, and water diffusion in the mouse brain.

While hypothermia of the brain is used to reduce neuronal damage in patients with conditions such as traumatic brain injury or stroke, little is known about how temperature affects the biophysical properties of in vivo brain tissue. Therefore, we measured shear wave speed (SWS), apparent diffusion coefficient (ADC), and cerebral blood flow (CBF) in the mouse brain at different body temperatures to investigate the relationship between temperature and tissue stiffness, water diffusion, and blood perfusion in the living brain. Multifrequency magnetic resonance elastography (MRE), diffusion-weighted imaging (DWI), and arterial spin labeling (ASL) were performed in seven mice while increasing and recording body temperature from hypothermia (28-30 °C) to normothermia (36-38 °C). SWS, ADC, and CBF were analyzed in regions of whole brain, cortex, hippocampus, and diencephalon. Our results show that SWS decreases while ADC and CBF increase from hypothermia to normothermia (whole brain SWS: -6.2%, ADC: +34.0%, CBF: +80.2%; cortex SWS: -10.1%, ADC: +30.9%, CBF: +82.4%; all p > 0.05). We found a significant inverse correlation between SWS and both ADC and CBF in all analyzed regions except diencephalon (whole brain SWS-ADC: r = -0.8, p < 0.005; SWS-CBF: r = -0.84, p < 0.005; cortex SWS-ADC: r = -0.74, p < 0.05; SWS-CBF: r = -0.65, p < 0.05). These results show that in vivo brain stiffness is inversely correlated with temperature, extracellular water mobility, and microvascular blood flow. Regional differences indicate that cortical areas are more markedly affected by hypothermia than central regions such as diencephalon. Temperature should be considered as a confounder in elastographic measurements, especially in preclinical settings. STATEMENT OF SIGNIFICANCE: Hibernating mammals lower their body temperature and metabolic activity. A hypothermic state can also be induced for medical purposes to reduce the risk of neural damage in patients with neurological disease or injury. However, little is known how physical soft-tissue properties of the in-vivo brain such as water diffusion, blood perfusion or mechanical parameters correlate with each other when temperature changes. Our study demonstrates for the first time that those quantitative imaging markers are tightly linked to changes in body temperature. While water diffusion and blood perfusion are reduced during hypothermia, brain stiffness significantly increases, suggesting that multiparametric quantitative MRI should be used for the noninvasive assessment of brain metabolic activity.