Arlotta, Paola

professor of stem cell and regenerative biology at Harvard University

Paola Arlotta: Brain Development from Stem Cell to Organoid

Lex Fridman Podcast · Aug 12, 2019

• Paola Arlotta
• brain organoids
• neural tube development
• cerebral cortex
• myelin evolution
• stem cell reprogramming
• single-cell profiling
• neurodevelopmental disorders
• brain plasticity
• ethics of brain organoid research

Core ideas

Arlotta frames human brain development as an astonishing, precisely choreographed process that begins with the neural tube — a simple cube-like structure of stem-like cells spanning the embryo from head to tail — from which all of the brain's thousands of distinct cell types and structures are built over roughly nine months of gestation, followed by roughly twenty more years of postnatal maturation before an adult brain is complete. She stresses that human brain development cannot be fully understood by studying mouse brains, which look nothing like human brains despite decades of valuable mouse research, and notes a striking cross-species pattern: developmental timing itself appears tied to a species' overall pace of life — a mouse brain organoid grown in a dish develops faster than a human one, suggesting the multi-year human developmental timeline (birth through puberty into the twenties) serves a functional purpose tied to how much post-birth learning humans require.

She emphasizes that developmental order is not arbitrary — neurons are made before supporting glial cells, for instance, because cells developing alongside each other continuously influence one another's fate, making isolated post-hoc assembly fundamentally different from true developmental co-formation. This process is governed not just by finely-tuned evolved gene-expression programs but also by mechanical forces (pressure, bending, stretching as the neural tube expands into brain structures), which physically signal cells to activate new genetic programs based on their position and the forces they experience — making brain development a joint chemical and mechanical phenomenon. Arlotta contrasts biological development's apparent "messiness" with engineered systems: while there are many alternative possible routes cells could take, evolution has converged on an extremely tight, reliable developmental pathway such that healthy brains reliably form the same way with very few exceptions — a reliability that is explicitly lost when cells develop outside their normal embryonic environment (e.g., in a lab dish), since the same building blocks placed in a different developmental context predictably produce different, more variable results.

She offers a striking finding on myelin (the axon insulation traditionally understood to speed neural signal transmission): the most evolutionarily recent neurons — those in the upper cerebral cortex responsible for humans' most complex cognition — actually have unusually sparse and irregularly patterned myelin, some having essentially none. Rather than interpreting this as an evolutionary "unfinished" state en route to more myelin, Arlotta argues this may represent a deliberate evolutionary trajectory toward less myelin, since irregular, partial myelination allows finer control over signal timing, adding a layer of variability and flexibility that could support more complex, less predictable neural functions.

On organoids, Arlotta is emphatic that a brain organoid is fundamentally not a brain — it is a much simpler cellular system, growing to a maximum of only a few millimeters in diameter, that mimics only some early aspects of brain development. Their scientific value lies specifically in being the only available system for directly observing human brain development as it happens, since researchers can never directly access or manipulate an actual developing human brain in utero. Organoids are grown by coaxing stem cells (derived, via Nobel Prize-winning reprogramming technology, from a patient's skin or blood sample) through a process that partially self-organizes and partially depends on scientists supplying known developmental cues; they show meaningfully higher variability between individual organoids than real embryos show between individual brains, precisely because the full "code" of development remains only partially understood and organoids lack the full embryonic context. Despite their simplicity, organoids modeling the cerebral cortex specifically (which Arlotta considers the most evolutionarily recent and cognitively significant brain region) can, given months of extended culture, generate multiple distinct cortical neuron types as well as astrocytes — non-neuronal support cells essential for forming synapses (via neurotransmitter release, growth-cone-guided axon extension, and other supporting chemistry) — meaning genuine neuron-to-neuron connectivity and communication do form within organoids, much as isolated cultured neurons will spontaneously form synapses with any nearby neuron given the chance.

The central clinical promise Arlotta describes is using organoids derived from a specific patient's own genetic material — particularly for neurodevelopmental disorders like autism spectrum conditions — to directly observe, for the first time, what may have gone molecularly or cellularly wrong during that individual's own early brain development, rather than inferring from population-level or animal-model data. New single-cell molecular profiling technology, only available in roughly the last five years, allows researchers to compare gene expression across millions of individual cells between affected and control organoids, potentially revealing which specific cell types, developmental stages, or molecular pathways are disrupted — information that could eventually inform genuinely mechanism-based treatment rather than diagnosis based only on symptoms. She frames the field's rapid recent progress (enabled by stem cell reprogramming technology and single-cell profiling, both essentially unavailable even a decade or two ago) as following an exponential trajectory, while firmly rejecting the idea that the field's actual goal is building a complete human brain — the scientific value of organoids doesn't require replicating the whole organ, and higher-order properties like mind, consciousness, or general intelligence are not something current organoid models can capture or study, since these properties emerge only from an actual, complete brain.

On ethics, Arlotta argues that debates about brain organoid research must be continuously grounded in actual current data about what these systems can and cannot do, rather than speculative extrapolation, and should involve an ongoing, evolving conversation among scientists, bioethicists, lawyers, philosophers, and psychiatrists — not a single fixed policy decided once. She argues scientists have a positive moral responsibility to pursue transformative applications (like using a specific patient's own organoid as a drug-screening platform to identify treatments for their particular developmental condition) precisely because this capability didn't exist even five years ago, while acknowledging a legitimate "devil's advocate" concern that language and framing shape public perception disproportionately to the actual state of the science — contrasting how differently the public reacts to the term "brain organoid" versus "human mini-brain" for the identical underlying system, and calling for more careful communication by both scientists and journalists.

Reflecting on nature versus nurture as a parent of two genetically distinct children (despite sharing the same two parents, due to normal genetic recombination), Arlotta describes brain development as analogous to the same musical piece performed by different orchestras and conductors — the same genetic "score" producing different outcomes depending on how it's "played" — with postnatal plasticity (the brain's capacity to change in response to sensory and environmental input, illustrated by how blind individuals' visual cortex develops differently and gets repurposed) providing the mechanism by which lived experience further shapes an already genetically-influenced brain. She closes by speculating that the cerebral cortex — evolution's most recently and heavily invested-in brain region, and the one most defined by plasticity — will likely continue adapting to whatever technologies humans routinely use (drawing an analogy to how animals develop cortical regions finely tuned to species-specific needs, like precise digit control or long-range vision), predicting that AI and other technologies will likely become integrated with human cognition not through some dramatic future merger but through the same ordinary plastic adaptation already visible in phenomena like the brain's near-instantaneous adaptation to virtual reality.

Excerpts