How To Study the Degenerating Brain
Neurodegenerative diseases are caused by the pathological death of neurons – each condition defined by a characteristic pattern of loss affecting different classes of neurons and neural circuits. The causes of some neurodegenerative conditions are established – in particular, those caused by specific DNA abnormalities. But the etiology of the more common, sporadic forms of neurodegenerative diseases remains mysterious. For nearly all such maladies, the search for ways to prevent or halt degeneration is ongoing.
A better understanding of how and why each disease progresses is necessary both for developing potential treatment strategies and for monitoring the effectiveness of interventions during clinical trials. Below are some of the main techniques deployed to study these debilitating and increasingly common diseases.
Post-mortem brain tissue
Detailed histological examination of post-mortem brains remains the gold standard for assessing the neuropathological extent of a disease. Schemes for analyzing microscopic and macroscopic features to divide disease progression into well-defined stages have been developed for each major condition. For example, Braak staging is used for both Alzheimer’s Disease (AD) and Parkinson’s Disease (PD) and Vonsattel and colleagues established five grades of Huntington’s Disease (HD) neuropathology.
Comparing the extent of pathology to clinical records of people’s symptoms is an important application of these data. Such analyses can, however, reveal variable relationships between the brain structure and clinical signs. For example, Vonsattel et al noted, in their original report, that five people with clinically manifest HD had no discernible structural abnormalities , whereas people with pronounced AD or PD pathology can display a wide range of symptoms.
Staging is an essential component of studies seeking to identify contributors to disease biology in post-mortem human brains. Such studies use immunohistochemistry and in situ hybridization to look at disease markers and molecules of interest to see how their expression and distribution change. Most diseases are marked by disease-specific protein accumulations but there are also pan-disease processes associated with cell stress, cell death and inflammation. Another important tool for such studies is staining techniques for labeling cells that are dying or that indicate the mechanism of cell death.
Post-mortem tissue can also be used for large-scale analyses of gene expression or protein abundance using contemporary genetic or proteomic profiling methods.
In contrast to post-mortem studies that provide a single, final snapshot of a brain in time, clinical and basic researchers need methods for tracking the real-time progression of neurodegenerative diseases in living people.
The best such biomarkers will correlate with disease progression strongly enough to provide diagnostic and prognostic information. However, methods not sensitive enough at an individual level to inform clinical decisions can still be useful for research purposes. And biomarker monitoring in individual patients is an increasingly vital element of clinical trials. Additionally, the possibility of identifying biomarkers indicating that an asymptomatic person is in the prodromal stage of a disease would facilitate the testing of early interventions.
Magnetic resonance imaging (MRI) can reveal both structural changes to the brain as neurodegeneration progresses and changes in function. Individual diseases are associated with specific patterns of atrophy that can be monitored using MRI. For example, shrinkage of striatal regions is a biomarker of HD progression, whereas various nodes in the motor system decrease in size over the natural history of motor neuron disease (MND). Diffusion tensor imaging (DTI) is a variant of MRI that is useful for studying white matter changes.
Functional MRI also plays an essential role in tracking disease-associated changes in neurophysiology and how they develop over time. For example, AD is robustly associated with decreased cerebral blood flow to temporal and frontal cortices and this can be used to distinguish AD from other causes of dementia.
PET imaging offers particular advantages for monitoring neurodegenerative diseases because it allows the labeling and visualization of specific molecules in the living human brain. For example, radioactively labeled molecules that bind to amyloid peptides and tau are used to image the accumulation and distribution of these molecules in AD, while there are a range of ligands that bind to components of the dopaminergic system used to look at dopaminergic neuronal loss and its consequences in PD.
Neuroimaging can be used alongside clinical testing to relate structural and functional brain changes to symptomology. Moreover, PET and MRI approaches can be combined to relate molecular changes to atrophy or blood flow changes.
In addition to directly imaging the brain – which is expensive and not widely available – basic researchers and clinicians agree that more accurate and easy-to-monitor biomarkers are needed for tracking disease progression.
Electrophysiological techniques – such as EEG, MEG and nerve conduction studies – can aid in tracking neurophysiological changes, such as changes in communication between brain regions or outputs to muscles, with high temporal resolution.
Much effort is also being expended into identifying biofluid molecules that correlate with disease status. Hence, blood, urine and cerebrospinal fluid (CSF) are potential sources of informative metabolites or other biomolecules. As an example, in HD, CSF levels of neurofilament light protein correlate with disease progression and can distinguish premanifest from manifest disease in carriers of pathogenic mutations, and CSF levels of mutant huntingtin protein are also useful. In AD, serum neurofilament light protein might covary with pathology. Ongoing research in AD, PD and MND all aims to establish better biomarkers for these conditions.
Certain neurodegenerative disorders are always the direct result of genetic abnormalities. HD, for example, follows pathogenic expansion of a sequence of CAG repeats in the huntingtin gene, whereas spinal muscular atrophy (SMA) comes in various forms, each associated with different genetic lesions.
For more common, neurodegenerative disorders there are often inherited subtypes of disease – which can be clinically indistinguishable from sporadic forms – that arise from specific genetic mutations. Hence, early onset Alzheimer’s can result from mutations in the amyloid precursor protein or presenilin 1 or 2 genes. And mutations in superoxide dismutase 1 cause a familial form of MND.
Often, however, the relationship between particular gene mutations and disease incidence is less straightforward. Gene variants, instead, can be thought of as risk factors for diseases or modulators of disease progression.
Either way, defining the genetic architecture of risk for neurodegenerative diseases at an ever-finer granularity remains an active pursuit. Studying the biology of the implicated genes, their altered protein products and the systems these affect should lead to better understanding of disease pathophysiology. Albeit, such goals have proved frustratingly difficult, as exemplified by the failure (so far) of the amyloid hypothesis of AD to yield effective interventions.
In order to conduct the kind of invasive mechanistic studies that are impermissible in people, neuroscientists have long sought to develop model systems that recapitulate the key features of human neurodegenerative conditions. Having valid model systems – be they in vivo animal models or in vitro systems – permits investigations of pathogenesis and pathophysiology, electrophysiological investigations of how neuronal loss affects network function, and early stage screening of compounds that might halt degeneration or otherwise protect or restore brain function.
In earlier decades, animal models based on neurotoxins that selectively targeted specific neuronal populations predominated. A prominent example would be the MPTP model of Parkinsonism. More recently, investigators have favored models based on genetics, arguing that such models better mimic human scenarios.
Pathogenic mutations discovered in human populations can be introduced as faithfully as possible into mice or other animals and their effects studies. This hasn’t, though, always been as straightforward as researchers would have liked or predicted, with mutations that are disease-causing in people not always leading to disease states in animals. Consequently, disease-associated genes may need to be overexpressed in animal models or manipulated in ways not seen in human populations to induce similar conditions.
The validity of any model, therefore, needs to be carefully assessed. These models do nevertheless provide invaluable opportunities for testing potential therapies, such as the development of antisense molecules for HD.
Human tissue models
An emerging alternative to animal models of neurodegenerative disease is the use of brain cells derived from human-derived induced pluripotent stem cells (iPSC). Such cells can be grown in monolayers or used to form three-dimensional brain organoids. Briefly, fibroblasts taken from human subjects are reprogrammed into iPSCs and these, in turn, are transformed into neurons or other cell types. Such cells are then capable of self-organizing into three-dimensional structures which develop certain characteristics of brain tissue.
Starter cells can be taken from any human, giving researchers the option to select patients and controls according to how they wish to run their experiments. That is, subjects can be selected according to either their disease phenotype or genotype of interest, and controls can be age-matched or be family members who control for genetic factors.
Such research is at an early stage, but initial results suggest promise for enabling researchers to investigate core aspects of disease biology in not just human tissue, but tissue derived from people affected by the disease in question. For example, the consequences of manipulating the AD risk gene APOE4 in iPSC-derived neurons, glia and organoids all showed specific disease-related phenotypes.
While each neurodegenerative condition has its own etiology and definitional pattern of neuronal loss, recent years have seen a surge in interest in the role played by the immune system, autophagy and inflammation in neurodegeneration. In particular, the contribution microglia make to degeneration has attracted attention. Studying inflammation and microglia can involve staining of post-mortem tissue, neuroimaging approaches and even making microglia-like cells from iPSCs and seeding brain organoids with them.