Predicted to double in prevalence within the next two decades alone, Alzheimer’s disease (AD) represents one of the most significant ongoing challenges to global health. Despite decades of extensive research efforts, there remains no cure, with the majority of promising preclinical treatment strategies ultimately failing in human patients. Progress is to some extent limited by the existing systems used to investigate AD in the lab, which may not comprehensively reflect the human condition. The following article explores a burgeoning approach to modelling this disease: induced pluripotent stem cells (iPSCs). Produced by reprogramming the adult skin cells of human Alzheimer’s patients back into their primitive, versatile developmental states, these stem cells can be used to generate samples of human brain tissue—seemingly ideal for investigating disease mechanisms and testing candidate drugs. Thus heralded by some as a more biologically faithful model of human AD, how do iPSCs compare against the trusty lab rat?When you close your eyes, you are 13 years old again … The room smells faintly of a cloying perfume you begged your mum to buy—notes of vanilla, jasmine, and unwavering teen spirit. You leave eager, uneven footprints down the stairs into the bathroom and inspect yourself in the mirror, admiring with pride what, in hindsight, is quite possibly the ugliest outfit of all time. Minutes later, when your eyes at last reopen, you are thankfully 21 once more, acid-wash jeggings and that terrible metallic crop top safely exiled to the past. Well, exiled from all but enduring memory…From regenerating doctors to endlessly repeating days, so many of the stories we read, write, and dream seem to represent an abiding desire to revisit the past. Although a long way from flux capacitors and time-turners, every day we enjoy the remarkable ability to revisit people, places, and moments long departed, stitching the past to the present in the act of remembering. However, it is in disorders of memory such as AD, as this portal to the past begins to close, that we truly appreciate just how important these tiny acts of time travel are. Without them, we lose not only names and stories but the very sense of who we are.With Alzheimer’s on the rise, scientists are turning to their own cellular version of time travel: iPSCs. Given the right molecular ingredients, a patient’s adult skin cells can be reverted to their earliest, most malleable form before being reprogrammed into neurons reflecting the donor’s specific genetic makeup. This technique may yield new and improved disease models for AD: systems that mimic aspects of the human illness, allowing drug testing and investigation of disease mechanisms. But does this “brain-in-a-dish” approach have what it takes to one day dethrone the trusty lab rat as the face of AD modelling?The leading cause of dementia, AD affects over 50 million people worldwide, with cases estimated to double within the next 20 years alone.1 Primarily affecting the elderly, it is a late-onset disorder, characterized by the gradual death of brain cells (neurodegeneration).Although the precise origins of AD remain in debate, two key proteins are consistently found at the scene of the crime: amyloid-β (Aβ) and tau. A fragment of the much larger amyloid precursor protein (APP), Aβ is usually mopped up and broken down by the brain’s waste disposal machinery. In Alzheimer’s, however, as these clearance systems deteriorate with age, accumulating Aβ fragments join to form plaques around neurons. Tau, normally a scaffolding protein maintaining neuronal structure, similarly aggregates, twisting into intracellular clumps called neurofibrillary tangles (NFTs). Researchers propose that Aβ plaques and NFTs trigger the immune system, leading to inflammation that damages and ultimately kills neurons.2 However, as the severity of plaques, tangles, inflammation, and behavioral symptoms is not always correlated, it is difficult to conclusively categorize these hallmarks as either causes or effects of the disease.3, 4The first casualties of AD are the brain’s memory and learning hubs: the hippocampus and entorhinal cortex.5, 6 Cell death here is largely specific to a population of neurons producing a signaling molecule (neurotransmitter) called acetylcholine.7 As neurodegeneration spreads to the parietal and frontal lobes, cognitive impairments progress to mood, awareness, and communication deficits.With no cure available, current AD treatments target Aβ plaques, acetylcholine levels, and excessive neuronal firing (hyperexcitation), achieving variable success in slowing symptom development. 2Alzheimer’s research has historically employed genetically engineered (transgenic) mouse models. While approximately 80% of human AD cases are sporadic (sAD), arising unpredictably in later life, there is also an earlier-onset, genetic form of Alzheimer’s: familial AD (fAD).8 Over 200 fAD-associated mutations have been identified, with most of the affected genes encoding proteins required to convert APP into Aβ, transport fatty molecules between neurons, or support the brain’s immune cells (glia).9 Created using genetic editing techniques such as CRISPR, mice carrying these fAD mutations develop hallmark features of human AD, producing tangle-prone forms of tau and Aβ plaques and exhibiting progressive cognitive decline.10While mouse models have offered vital insights into the disease-causing (pathogenic) mechanisms of Aβ and tau, shaping the field’s prevailing hypotheses, they are notably limited by interspecies differences in lifespan, neuron structure, and immune function. Similarly, fAD mutations expressed in mouse models drive disease by increasing production of toxic, plaque-forming Aβ variants, whereas human sAD arises mainly from age-related failures in Aβ disposal.10 Together, these biological and mechanistic differences may explain why transgenic mice do not develop NFTs or the extensive neurodegeneration seen in human AD. 10 What’s more, unable to ask about memories of childhood and the hallowed halls of the cages in which they were raised, we resort to potentially reductive spatial tests of memory when assessing treatment outcomes in mice.11 Consequently, of the hundreds of candidate drugs that prove effective in mouse models, very few successfully translate to human Alzheimer’s patients.2Interspecies differences in size and sensibility…Put simply, stem cells are a biological “blank slate.” Unlike most of your adult cells, which have already committed to a fixed identity and function (cell fate), stem cells retain the ability to grow into numerous types of cells. Initially capable of becoming almost any cell type (pluripotent), the cells of the human embryo lose access to particular fates during development, eventually becoming unipotent. This process, termed differentiation, is directed by networks of transcription factors: molecular switches that turn genes on and off to shape a cell’s structure and function. Notably, all the cells in your body contain the same genes, but they only switch on those that are relevant to their function; this is why your neurons and red blood cells, for example, look and behave so differently.12A theoretically infinite supply of cells of any type, the therapeutic potential of embryonic stem cells in replacing damaged or diseased tissue was unfortunately confounded by ethical concerns associated with harvesting them.13Thus, in 2006, when Japanese researcher Shinya Yamanaka successfully reprogrammed adult human skin cells (fibroblasts) back into pluripotent stem cells, and the field got its second wind. These “iPSCs” were created by exposing the fibroblasts to transcription factors typically present in the human embryo, switching on the genes that enable pluripotency. A blank slate once again, the iPSCs can then be differentiated into any cell type, simply by choosing the correct combination of transcription factors.14Back to the glory days (pluripotency)…Using skin samples from fAD patients, researchers can produce dishes of iPSC-derived neurons called cultures, where each cell carries the donor’s set of genes (genome), including their predisposition to AD. Simply by adjusting the recipe of transcription factors used, separate cultures are created containing specific neuron or glial cell subtypes. To more faithfully represent the diverse cellular landscape of the human brain, these are often combined in specific ratios to create a co-culture.15IPSC-derived models offer a rare window into the earliest, most mystifying stages of AD, allowing investigation of the causal and chronological relationships between disease events such as amyloid and tau aggregation, hyperexcitation, and inflammation. As a renewable source of human neurons, perhaps most important is the potential of iPSC-derived AD models in high-throughput screening: using automated systems to test the effects of thousands of potential AD treatments in parallel. 16 Given their identical genetic content, iPSC-derived models can be used to predict the treatment response of the human donor more reliably than transgenic mice and may also help to explain why existing AD treatments work better for some patients than others.However, grown outside of a living organism (in vitro), iPSC-derived models are not particularly useful for studying the behavioral aspects of Alzheimer’s. In this respect, mouse models remain superior… Since most iPSC cultures are derived from fAD, rather than sAD patients,15 they also share the corresponding limitations seen in mouse models.One unique pitfall of iPSC-derived models is cellular immaturity. As a disease of aging, it is important that the neurons in AD models are as old and wise as those of human patients. Noticing that iPSC-derived neurons appear smaller than those of adult humans, researchers discovered that the reprogramming process used to induce pluripotency has a de-aging effect, erasing the molecular marks acquired by the fibroblast over time.17, 18 Even once differentiated, iPSC-derived cells exhibit fetal behavior, for example, producing immature tau rather than the aggregation-prone isoform—as such, iPSC models do not develop NFTs.19 Emerging solutions aim to bypass pluripotency altogether, converting adult skin cells directly into neurons to preserve their maturity. 20In any case, these in vitro models constitute an exciting new arrival in the arena of AD modelling, more likely a valuable companion to existing mouse models than a hostile adversary!I have long been committed to translating whatever knowledge I accrue into accessible, evidence-based science communication, aiming to bridge the ever-widening gap between academia and the public. I was drawn to the Biochemical Society’s Science Communication Prize as an invaluable opportunity to contribute to this larger goal. I chose to write about what I know—and what fascinates me—basing my entry around my university dissertation project investigating stem cell models of Alzheimer’s disease. My ultimate hope is that this article lifts to some degree the curtain on one of many ways in which scientists actually conduct the research that makes headlines. - Priya Prakash, University College London.
Priya Prakash (Thu,) studied this question.