All neuronal circuits of the human brain make up the so-called connectome. Within, three spatial dimensions have been identified, ranging from the microscopic to the mesoscopic to the macroscopic scale. Most studies on connectivity focus on the macroscopic scale by performing magnetic resonance imaging (MRI) in and ex vivo. The bottleneck of this method is the limited resolution accompanied by limited accuracy and validity of 3D MRI data. Thereby, the challenge arises to provide anatomical ground-truth information and create microstructure-informed MRI measurements. One way to overcome these resolution differences between macroscopic and microscopic imaging methods is to generate (sub)cellularly resolved (< 100 nm), mesoscopic tissue blocks (mm-sized). This can be reached, by pairing novel tissue clearing techniques with high-end, large-scale microscopic imaging and is the first goal of the presented thesis. In combining different processing and imaging techniques, that are sensitive to different spatial scales and microstructural properties, a comprehensive understanding of the wiring of the human brain can be generated.
Transparency of biological tissue is reached by matching heterogeneous refractive indices (RI) of intra- and extracellular compartments, divided by lipid membranes, to each other and to the RI of their surrounding medium. In doing so, scattering and absorption of light traveling through the sample is reduced. Tissue clearing methods use different approaches to homogenize RI. While hydrophilic and hydrogel-based methods wash out lipids by using strong detergents (e.g., sodium dodecyl sulfate or urea), hydrophobic methods use organic solvents (dibenzylether or ethylcinnamate) to dissolve tissue lipids.
The clearing of human brain tissue is particularly challenging as it is densely packed, highly myelinated, and usually aged (i.e., by donors of high age). Initially seven tissue clearing techniques were tested to determine the most efficient one. Efficiency was defined in a high degree of tissue transparency, preserved ability for immunohistochemical staining before or after clearing to obtain highly resolved microscopic imaging results, and short experimental processes.
All methods were tested on mm-sized, fixed post mortem brain tissue samples and included the CLARITY, CUBIC, iDISCO, MASH, ECi, Visikol and Ce3D protocol. The CLARITY, CUBIC, iDISCO, and MASH techniques were able to clear aged human brain tissue, whereas the ECi, Visikol and Ce3D techniques were not. Generally, water-based CLARITY and CUBIC methods are gentler than solvent-based iDISCO and MASH techniques. The expansion of CLARITY-treated samples appears to be advantageous as well. However, all technical aspects (i.e., hydrogel pre-treatment, electrophoretic chamber acquisition) of the CLARITY method are time and cost consuming. Here, the iDISCO, MASH, and ECi protocols are more efficient. Their material costs are lower and, if successful, the processing duration is short compared to the CLARITY technique. Although, the Visikol (commercial clearing kit) and Ce3D technique claim to be quickly applied and to clear tissue fast, they are less affordable and more complex to perform as well.
After attaining transparency, optical properties of the tissue samples are altered. Hence, the application of immunohistochemistry remains crucial. Microscopic imaging of cleared, immunohistochemically labelled human brain tissue samples revealed that the CLARITY and iDISCO techniques are most suitable. Here, immunohistochemical reagents as well as light penetrated sufficiently deep into the tissue. In addition to the evaluation of different clearing techniques, three microscopic setups, specifically built or equipped for imaging large-scale specimen, were tested to identify suitability, benefits, and downfalls. The results showed that the most suitable and efficient approaches are the combination of i) the CLARITY method coupled with imaging at the Zeiss LSM 880 Airyscan and ii) the iDISCO method coupled with the Miltenyi Biotec Ultramicroscope ll. Both combinations enable three-dimensional histology of aged human brain tissue.
There is growing demand for visualizing and examining human cyto- and myeloarchitecture in 3D. Successfully clearing aged human brain tissue whilst preserving its microstructure and ability for immunohistochemical staining will bridge standard histological thin-sectioning (2D) and non-invasive imaging techniques, such as MRI. Once mastered in post mortem human brain blocks in 3D, tools such as fiber tracking and co-localisation are going to provide a powerful tool for multi-modal validation of in vivo microstructure-informed MRI, unraveling the human connectome.
Next to unraveling the human connectome based on myelinated axonal fiber pathways, the second goal of this thesis is to investigate the distribution patterns of astrocytes. As they support oligodendrocytes during myelination processes, astrocytes are proposed to be co-localised with myelinated fibers. Revealing their spatial organization offers another great approach to study the wiring of the human brain. Astrocytes are the most common subtype of glia cells in the central nervous system and can serve many functions depending on their interaction partner and surroundings. Those include but are not limited to the regulation of synaptic plasticity, ion and pH homeostasis of neurons and the vasculature. Moreover, astrocytes can adapt morphologically, physiologically, and molecularly during their response to neurodegeneration or demyelination. This remodelling is facilitated, amongst others, by the glial fibrillary acidic protein (GFAP), a cytoskeletal protein and the primary intermediate filament of mature astrocytes. Hence, antibodies against GFAP are commonly and validly used. Furthermore, great heterogeneity occurs among astrocytes with distinguishable morphologies in white matter (fibrous astrocytes) and grey matter areas (protoplasmic astrocytes). Accordingly, GFAP is expressed differently depending on the brain region.
Focusing on the co-localisation of myelinated fibers and astrocytes, immunohistochemical analyses were performed to identify myelin- and GFAP-positive structures in aged human cerebral and brainstem areas. Up to three different regions of interest of four different human brains were examined to analyse myelinated fiber and astrocyte content in white and grey matter areas. Subsequently, two image processing approaches were deployed to extract fluorescence intensity values. The first, semi-manual approach, performed with the Zeiss ZEN blue software, generated distribution patterns within a wider range. Hereby, each channel underwent linear unmixing to co-localize the detected fluorescent signals to each other. The second, semi-automated approach, performed with the Segmensation application software, generated distribution patterns within a narrower range. As GFAP dominates in fibrillary astrocytes of white matter regions consistent results were obtained with both image analysis tools. As GFAP expression in grey matter astrocytes is limited, less consistent results were obtained. Therefore, immunohistochemical analyses of different white and grey matter regions of the human brain align with regional distribution patterns of GFAP-positive astrocytes. Remarkably, as proposed in white matter regions, astrocytes and myelinated axons share a similar spatial organization across all investigated regions of interest.
In 1992, Suzuki and Raisman first observed this intertwined organization of axons and glial processes in the rat’s brain, coining the term glial framework. The presented thesis confirms this spatial organization of myelinated fibers and astrocytic branches in the human brain. Although this phenomenon was known as well, it was unclear whether this is a global feature. Moreover, the scarcity of studies on the human glial framework, especially in brainstem areas, needs to be acknowledged. This thesis starts to fill the gap, providing evidence for the presence of the glial framework, created by GFAP-positive astrocytes, within white matter regions of cerebral and brainstem areas of the aged human brain. Additionally, the application of immunohistochemistry in cleared, mesoscopic-sized brain tissue samples adds another promising perspective to successfully unraveling the human glial framework globally. As there is growing demand for sensitizing non-invasive imaging techniques to the heterogeneity of astrocytes, studying the spatial organization of myelinated fibers in correlation with astrocytes will provide deeper insights into the impact astrocytes have on neurological integrity and degeneration.
Both points of focus of the presented thesis aim to untangle and give a clearer understanding of the human connectome. Myelin, as the primary component of structural and functional integrity of all neuronal fiber pathways, needs to remain the centre of investigations. However, expanding these investigations by studying the glial framework is going to propel our understanding of the human myeloarchitecture forward. Using the coherent approach as is tissue clearing, is a great way to examine structural and functional correlations in 3D and needs to remain a central element of future studies.
| Identifer | oai:union.ndltd.org:DRESDEN/oai:qucosa:de:qucosa:92182 |
| Date | 21 June 2024 |
| Creators | Rusch, Henriette |
| Contributors | Universität Leipzig |
| Source Sets | Hochschulschriftenserver (HSSS) der SLUB Dresden |
| Language | English |
| Detected Language | English |
| Type | info:eu-repo/semantics/publishedVersion, doc-type:doctoralThesis, info:eu-repo/semantics/doctoralThesis, doc-type:Text |
| Rights | info:eu-repo/semantics/openAccess |
| Relation | https://doi.org/10.1016/j.neuroimage.2021.118832 |
Page generated in 0.003 seconds