Multi-nuclear magnetic resonance imaging and spectroscopy of lithium in the brain
by Tor Rasmus Memhave
Date of Examination:2023-12-08
Date of issue:2024-02-23
Advisor:Prof. Dr. Susann Boretius
Referee:Prof. Dr. Susann Boretius
Referee:Prof. Dr. Dr. Hannelore Ehrenreich
Persistent Address:
http://dx.doi.org/10.53846/goediss-10358
Files in this item
Name:Memhave_PhDThesis_upload.pdf
Size:7.27Mb
Format:PDF
The following license files are associated with this item:
Abstract
English
Equipped with only three protons, lithium stands as one of the most unexpected drug discoveries in history. Despite its initial FDA approval in 1970, lithium's precise mechanism for regulating mood remains largely elusive. Lithium is generally considered a modulator of neurotransmission, an inhibitor of enzymes in the signaling cascade of inositol (1,4,5)-triphosphate and diacylglycerol, and neuroprotective. Yet, the true challenge lies in understanding how lithium effectively treats patients with bipolar disorder. Bipolar disorder is a psychiatric illness that is characterized by recurrent periods of mania and depression, often accompanied by a high suicide rate (12-fold higher than the population average) and various comorbidities. While lithium is an effective drug for some bipolar patients (between 30-50%), it also carries an increased risk of renal disease. It is therefore essential to understand the early stages of lithium treatment and find a way to predict whether a treatment is successful. However, despite extensive research, biomarkers for lithium treatment response have largely eluded scientists. In this thesis, I aim to tackle a part of this problem, focusing on the effect of lithium on the healthy brain. By isolating the drug from its pathological context, I have focused on characterizing lithium treatment in mice during the first four weeks of treatment. My investigation encompasses three key aspects: lithium distribution in the brain, brain metabolism, and changes in water diffusivity in the brain. Employing magnetic resonance imaging and spectroscopy, I have looked for biomarkers with a translational perspective. The primary isotope of lithium, lithium-7, is a spin-3/2 nucleus with a nuclear receptivity of 0.29 compared to 1 of protons. Lithium-7 is therefore magnetically active making it possible to detect lithium directly in the brain. Previous studies in humans and animals have revealed an inhomogeneous distribution of lithium in the brain. However, no in vivo imaging of lithium in mice has previously been performed. Through optimizing the magnetic resonance sequence and hardware, I successfully performed the first in vivo lithium-7 magnetic resonance imaging of the mouse brain (Chapters 2-3). I found lower lithium concentrations in the olfactory bulb and cerebellum compared to the rest of the brain. Furthermore, I successfully estimated the lithium concentration in the brain of mice. To investigate brain metabolism, I employed magnetic resonance spectroscopy – a technique that relies on the principle of chemical shift. The conventional approach to analyze spectroscopy data is to fit it with a linear combination of basis spectra (model functions of different metabolites). I performed this analysis using a software package called LCModel, which is currently considered the gold standard for in vivo spectroscopy. I could show that this software has non-ideal behavior when adding noise to spectra (Chapter 4). To better characterize the impact of adding noise on the quality of the fit, I incorporated data from five different species: mice, rats, marmosets, macaques, and humans. I also developed an open-source toolbox for simulating cortical spectra from these species at various noise levels. To analyze 3D astrocyte cell cultures within an MRI scanner I developed a magnetic resonance-compatible bioreactor linked to a compact incubator, enabling precise control of gas mixing and exchange (Chapter 5). Using the bioreactor, I investigated the metabolic response of astrocytes to lithium-enriched media. I found that astrocytes exposed to therapeutic concentrations of lithium (0.78 mM) exhibited reduced levels of myo-inositol and glutamate plus glutamine, while these levels increased at higher lithium concentrations (4.38 mM). Finally, I established a magnetic resonance-based characterization of the effects of lithium treatment in mice, incorporating structural analysis, metabolic assessments, and diffusion-related changes. I found that mice on a lithium-enriched diet showed elevated levels of myo-inositol, decreased N-acetylaspartate, and reduced diffusivity. At first glance these changes looked bleak – considering that these suggest increased glial content, decreased neuronal health, and inflammation. However, upon further investigation, I found that these were accompanied by a change in the glutamate-to-glutamine ratio and increased neurite density index – changes associated with a potential shift in cell population rather than neurotoxicity. This thesis starts by developing methods enabling the study of lithium treatment in mice in vivo. The second half demonstrates the application of these methodologies and presents evidence for a shift in cell population following lithium treatment. While I obviously cannot claim to have solved the mechanism of lithium action, I have succeeded in delineating a magnetic resonance-based profile of lithium treatment in both astrocytes and mice.
Keywords: magnetic resonance imaging (MRI); magnetic resonance spectroscopy (MRS); Lithium-7 (7Li); mice; astrocytes; X-nuclei
