The brain is often viewed as developing from a single starting pool of cells, but recent research suggests a different story. New evidence indicates that the brain actually arises from two distinct, parallel sets of early cells, one that builds the forebrain and midbrain and another that builds the hindbrain. These cells are restricted to forming specific regions from very early in development. These findings, detailed in a study published in Nature Neuroscience, point to an evolutionary strategy that has shaped nervous system development for hundreds of millions of years.
To understand how the brain forms, scientists study pluripotent stem cells. These are unique cells capable of developing into nearly any tissue in the body. During an early phase of embryonic development known as gastrulation, the basic body plan begins to take shape. At this stage, some stem cells become a tissue called the neural ectoderm, which eventually forms the brain, while other structures like the spinal cord arise from different early tissues.
Generating specific types of brain cells in the lab is a major goal in regenerative medicine. Scientists are particularly interested in the hindbrain, a region that manages vital body processes like breathing, swallowing, and sleep. Motor neuron diseases such as spinal muscular atrophy and amyotrophic lateral sclerosis (ALS), which damage nerve cells in both the brainstem and spinal cord, are often fatal. A 2026 study found that stem cells guided to the same type of neuron through hindbrain-like versus spinal cord-like routes ended up with distinct patterns of gene activity, suggesting a cell’s developmental history leaves a lasting mark.
This highlights why researchers need an accurate roadmap of brain development. Historically, one popular model proposed that all brain regions emerge from a common pool of neural progenitor cells. A progenitor is an early, not-yet-specialized cell in the embryo that divides and gives rise to more specialized cells. Neural progenitors are the early cells whose descendants become the brain and spinal cord. While other early studies mapping cell fates hinted that cells might be dedicated to specific brain regions early on, they could not confirm if the cells were permanently committed to those paths or just following local chemical signals.
“We certainly didn’t expect to find that the brain comes from two different sources! In 1952, the trailblazing Dutch scientist Pieter Nieuwkoop proposed that the entire brain arises from a common source,” study author Kyle M. Loh of Stanford University told PsyPost. “This led to the prevailing model in the field. Therefore, finding two different brain origins surprised us.”
“I also want to emphasize that we built on important work by other scientists in the 1990s who also suggested that there are perhaps multiple origins of the brain,” Loh added. “The origins of the brain remain a longstanding question, with many twists and turns along the way, and we are excited to work with others to continue to explore this area.”
The research, led by Rayyan T. Jokhai, Carolyn E. Dundes, and Loh, aimed to resolve this debate. They wanted to see if the earliest brain cells are already restricted in what they can become. The team looked at both mouse embryos and human stem cells to track how these initial developmental decisions are made.
“The brain is one organ, but we found that it is built from two different sources while a baby is growing in a mother’s womb,” Loh explained. “These two pieces—the front part of the brain and the back part of the brain—are pushed together to create one integrated organ. For instance, a smartphone is built from many different pieces that come from different countries, but they are later connected to form one integrated unit.”
First, the researchers used a technique called genetic lineage tracing in mouse embryos. They applied a fluorescent tag to a specific group of early neural cells right as gastrulation was ending. This tag allowed them to track the exact locations of these cells and their descendants as the embryos continued to grow until just before birth.
The tagged cells, which expressed a specific gene called Gbx2, were almost exclusively found in the developing hindbrain. They were essentially absent from the forebrain and midbrain. This provided evidence that in a living embryo, this specific population of cells is already on a dedicated path to form the hindbrain.
To test if these cells could be forced to change their destiny, the team turned to human pluripotent stem cells grown in lab dishes. They exposed the stem cells to specific mixtures of chemical signals to mimic early embryonic development.
Using different chemical recipes, the researchers turned the stem cells into two separate types of early brain tissue within just two days. One recipe yielded anterior neural ectoderm, which naturally forms the forebrain and midbrain. Another recipe, which included the signaling molecules FGF and retinoic acid, produced posterior neural ectoderm, which naturally forms the hindbrain. The researchers then challenged these cells by exposing them to the opposite developmental signals.
The cells largely resisted the new instructions. When the anterior cells were given chemical signals meant to create a hindbrain, they mostly failed to make the switch. In the same way, the posterior cells could rarely be coaxed into becoming forebrain or midbrain cells. Both cell types generally held their identities even when the two were mixed together in the same dish.
To understand this stubbornness, the researchers looked at the cells’ chromatin, which is the way DNA is packaged and organized. They found that the anterior and posterior cells had markedly different chromatin landscapes. The genes required for forebrain development were accessible and ready for action in the anterior cells, while the hindbrain genes were locked away.
The most important takeaway is “that the neuroectoderm divides into two separate progenitor populations that diverge in their potential to generate different brain regions,” Eduardo Sequerra, a researcher at the Brain Institute at the Universidade Federal do Rio Grande do Norte who was not involved in the study, told PsyPost. He emphasized that this split happens “much earlier than the neural tube formation,” referring to the formation of the early embryonic structure that eventually becomes the brain and spinal cord.
“I am quite confident in their very interesting finding,” Sequerra said. “They joined in vivo developmental studies across different species with in vitro differentiation systems to show their hypothesis is solid in different ways.” In vivo means studies done in living organisms, such as embryos, while in vitro means studies done on cells grown outside the body, such as in a lab dish. Differentiation is the process by which an unspecialized cell, like a stem cell, turns into a specific cell type, such as a neuron.
This early division aligns with older clues in the field. “During the ’90s Nicole Le Douarin’s group performed a series of transplantation studies showing that the folding neural tube is divided into two parts,” Sequerra explained. He noted that one part sits above the notochord, a flexible rod of tissue that helps organize the developing embryo, and “cannot differentiate into anterior structures.” The other part, which lacks a notochord below it, “differentiates into the midbrain and forebrain,” he said.
“This article shows that the split into two populations occurs much earlier than the established notochord and neural folds stage,” Sequerra added, referring to the ridges of tissue on the embryo’s back that rise up and merge to form the neural tube. “This split possibly occurs when the future notochord-mesoderm cells are still migrating.” The mesoderm is the middle layer of cells in an early embryo, which goes on to form muscles, bones, blood, and other tissues.
Using this new knowledge, the research team successfully directed the posterior neural ectoderm cells to grow into specialized hindbrain motor neurons. These particular neurons are responsible for controlling muscles in the face and neck, including those necessary for swallowing. The lab-grown neurons showed spontaneous calcium activity and fired electrical impulses when stimulated, signs that they were functional.
“We are excited by the ability to create human hindbrain neurons in a Petri dish,” Loh said. “The hindbrain—the back part of the brain, otherwise known as the brainstem—controls life-sustaining functions such as sleep, wakefulness, consciousness, hunger, and so on. The ability to create human hindbrain neurons in a Petri dish may help us better understand these fundamental, life-sustaining processes, and diseases that impact them.”
“Both amyotrophic lateral sclerosis and spinal muscular atrophy are deadly diseases,” Loh noted. “In particular, spinal muscular atrophy is the #1 genetic cause of death in babies under 1 year of age. Both diseases kill hindbrain motor neurons that control muscles involved in eating, swallowing, and speech.”
“We cannot sample living motor neurons from a patient’s hindbrain to watch how these diseases attack these neurons. Growing these hindbrain motor neurons in a dish gives us a way to investigate these diseases,” Loh explained. “We hope to explore how amyotrophic lateral sclerosis, spinal muscular atrophy, and other diseases affect human hindbrain motor neurons, with the hope of using this system to discover and test potential treatments.”
“The development of a method to differentiate motor neurons in vitro has the potential to bring hope for those with neuromuscular disorders,” Sequerra said. However, he cautioned that this approach relies on studying patient-specific cells in a dish, rather than replacing damaged tissue in the body. “Of course it has the potential of becoming a way of studying specific patients’ motor neurons through iPSCs [induced pluripotent stem cells] but it definitely is not close to a stem cell therapy for regeneration.”
The researchers also examined the embryos of several other species, including macaques, chickens, zebrafish, and acorn worms. Acorn worms are small marine worms that burrow in the seafloor and are distant relatives of animals with backbones. The researchers found distinct anterior and posterior neural cell groups in all of them. This suggests that the mechanism of using two parallel starting populations to build a brain has been conserved across different species for roughly 550 million years.
“We hope to learn more about how the brain was built during evolution,” Loh said. “In humans, mice, and other types of animals, the two parts of the brain develop next to one another, even though they originate from different sources. However, other pioneering scientists discovered that jellyfish—which diverged from us 600 million years ago—have two separate nerve rings at different ends of the body.”
“Our speculation is that early during evolution, there was perhaps no ‘grand plan’ to create a single brain, but rather potentially two separate nervous systems were created,” Loh continued. “Later in evolution, these two nervous systems may have been pushed together to give rise to the appearance of one integrated brain. It is important to emphasize that this is only an idea, and it might change as scientists learn more.”
“The expansion of the midbrain-forebrain progenitor field was probably an important phenomenon in the vertebrate nervous system evolution,” Sequerra noted. Moving forward, he said “it will be nice to go deeper into the mechanisms for the segregation of the two populations in both vertebrates and invertebrate chordates, or even other deuterostomes.” Invertebrate chordates are animals without a backbone that still share key body features with vertebrates, such as a notochord at some stage of life. Examples include sea squirts and lancelets, which belong to the broader deuterostome group alongside vertebrates.
As with all research, there are a few things to keep in mind. Growing human cells in a plastic dish does not perfectly recreate the complex, three-dimensional environment of a developing embryo. The cells might behave slightly differently when isolated from the intricate chemical exchanges that happen naturally in the body.
Additionally, the researchers emphasize that their findings should not be misinterpreted as suggesting the brain lacks unity. “Some people reading our paper sometimes think it suggests the brain is two separate organs. More precisely, the brain is one organ, but it is built from two different sources,” Loh explained. “By the analogy, the heart is likewise built from two different sources: one source creates the right ventricle, and another source creates the three other chambers of the heart. But ultimately, all four chambers of the heart work together, similar to how different parts of the brain all work together to give rise to all the amazing things that the brain can do.”
In addition, while the lab-grown cells appeared strongly committed to their specific paths, other unknown chemical combinations might exist that could force them to switch identities. Testing this in living mammals is technically difficult and would require placing labeled progenitor cells into the wrong location in a growing mammalian embryo. Despite these caveats, the ability to predictably grow specialized hindbrain neurons offers a highly useful tool for studying fatal neurodegenerative diseases in the lab. These are diseases in which nerve cells gradually break down and die over time, such as ALS.
The study, “Two parallel neural ectoderm progenitors contribute to the developing brain,” was authored by Rayyan T. Jokhai, Carolyn E. Dundes, Hadia S. Ahsan, Rachel S. Kang, Rachel E.A. Salomon-Shulman, Arjun Rajan, Yoon Seok Kim, Liam J. Stanton, Christine Xu, Stephanie Do, Brennan D. McDonald, José Miguel Andrade López, Hugo A. Urrutia, Hannah Greenfeld, Alicia Wong, Yimiao Qu, Andrew S. Petkovic, Yi Miao, K. Christopher Garcia, Michelle Monje, Daniel E. Wagner, Marianne E. Bronner, Christopher J. Lowe, and Kyle M. Loh.