Tiny, year-old human brain organoids that mimic the brain of a newborn are providing new clues to potential treatments for a leading genetic cause of childhood epilepsy that typically manifests as the brain matures shortly after birth.
A new study by UC Berkeley researchers points to hyperreactive astrocytes as one of the primary drivers of the inflammation that can cause seizures and highlights possible therapies that could tamp down the inflammation. Astrocytes are one of three main types of glial cells in the brain and help to support and facilitate the transmission of signals through neurons.
Human brain organoids — balls of brain tissue several millimeters across floating in a nutrient liquid — contain only a few types of brain cells and are far from a real brain. But, unlike the flat layer of cells in a traditional cell culture, organoids develop and behave much like cells inside a living brain.
The inherited disorder, called tuberous sclerosis complex, is characterized by potato-shaped lesions in the brain that cause severe seizures often resistant to common antiseizure medications. The disease is also accompanied by debilitating symptoms in other parts of the body, including the heart, skin and kidneys.
The specific genetic cause of TSC's tubers has been suspected for a long time: a child inherits a mutation in one copy or allele of a gene — TSC1 or TSC2 — and then develops a random mutation in the other allele. With both copies of the gene defective, the child develops lesions that interrupt the brain's electrical network. Often, the second mutation occurs in separate regions of the brain, leading to multiple tuberous lesions.
What wasn't known is how mutations in these two genes, which regulate a major developmental and metabolic pathway throughout the body called mTOR, cause tuberous lesions in the brain. The assumption was that they were making neurons overreact, leading to seizures.
In the new study, which appears this week in the journal Nature, Helen Bateup , a Berkeley professor of neuroscience and of molecular and cell biology, and her colleagues show that a key consequence of these mutations is the generation of astrocytes in the brain that are hyperreactive from the moment they form. This suggests that abnormal astrocytes and other glial cells are not just a consequence of seizures — they actually cause the lesions themselves.
"As soon as these astrocytes are born, they are reactive and look like they've been triggered into a disease state. This is arising as a primary result of the mutation," said Bateup, a Weill
Neurohub Investigator. "So now we can rethink the disease pathophysiology. It's not necessarily the case that the neurons are the only cause of seizure activity and the glia become involved later. It could be the other way around, or it could be that both cell types contribute to seizures and epilepsy."
If the hyperactive astrocyte scenario is correct, existing immunosuppressant drugs that target reactive cells could be tried to tamp down the inflammation and perhaps alleviate these intractable seizures.
"If it's really glia-driven and there's all these angry cells causing problems, how much can we fix by just suppressing that?" she said. "Can you calm down the glia and ideally bring them back to a homeostatic state, or if that's not possible, just shut off their ability to cause damage to the surrounding cells? I think that's feasible."
Millimeter-sized human brain organoids
Bateup has been studying TSC and the mTOR pathway for nearly 20 years, initially in mouse models and 2D cell cultures. But mouse models of these disorders don't fully replicate the range of symptoms seen in the human disease — its so-called phenotype. With the advent of human stem-cell-derived brain organoids a decade ago, she shifted her work from 2D cell cultures to 3D organoids.
"Organoids really have a lot of advantages over traditional 2D cultures — we can observe phenotypes that have a lot more similarity to phenotypes patients have," she said.
One difficulty with using human cells, however, is that they develop at a human pace. To see the effects of a developmental disease with onset around the time of birth, the organoids need to be cultivated for at least nine months. As a result, Bateup's graduate students and postdoctoral fellows baby their organoids far longer than most labs.
"The astrocytes don't really acquire their mature properties in humans until around perinatal or early postnatal life, so we grow these organoids for a long time," she said. "Which people think is crazy."
But that coddling allowed her team to see what happens in the organoids as "radial" progenitor cells mature and then switch from their initial job — forming hundreds of billions of neurons — to the production of glial cells, including astrocytes. In the TSC organoid model, the radial progenitor cells start producing reactive astrocytes at a time when they should be making neurons.
"In the organoid model, the most impacted cells are astrocytes. So that's why I'm favoring the idea that they're really driving a lot of the pathology," she said. "Developing the organoid system was a big advance in our ability to more closely mimic patient brain phenotypes. We now have a much better, or really the only robust model to study tuber cell development."
mTORopathies
The mTOR pathway is a fundamental signaling hub in every cell of the body, Bateup said, and a critical regulator of cell growth and metabolism. As such, it has mostly been studied in the context of cancer. But studies have shown that it is also important in brain development and can participate in the pathological changes that occur in neurodegenerative disorders.
In fact, there's a whole range of genetic brain disorders — 14 or more — called mTORopathies, which are caused by mutations in regulators of the mTOR pathway. All lead to hyperactivation of mTOR, resulting in focal epilepsy and often cognitive impairment. Altogether, these mTORopathies are among the most common causes of childhood intractable epilepsy, with TSC being the most typical form — about one in every 6,000 to 10,000 births.
TSC is one of the few mTORopathies with a hereditary component, however; the others appear to be random occurrences — like most cancers. Yet TSC is not a cancer.
"The cells are not dividing more, we know that, but they are impaired in their ability to differentiate into neurons and glia and so they produce these really abnormal cells. These cells form a focal lesion in the brain called a tuber," Bateup said.
The researchers created their brain organoids from human stem cells and introduced mutations in the TSC2 gene of some of the progenitor cells, which propagated into the resultant neurons and glia. They then used single-cell transcriptomics to compare the expressed genes in reactive astrocytes with those in their normal, non-reactive neighbors.
The team found that many of the overexpressed genes in reactive astrocytes were the same as those overexpressed in neurodegenerative diseases such as Alzheimer's. When they looked at tuberous tissue removed from the brains of 10 patients with TSC, they found similar protein expression profiles.
"I think there are some interesting parallels or overlaps with TSC, neurodegenerative disorders, mTOR dysfunction, and astrocyte pathology that are ripe for exploring," she said.
While some mTORopathies can be treated with drugs that inhibit the mTOR pathway, such as rapalogs, which are derivatives of rapamycin, these drugs can have significant side effects, disrupting key metabolic, immune and cellular functions. Ideally, drugs that silence the hyperreactive astrocytes could provide a more targeted therapy without as many side effects.
The work was funded by the National Institute of Neurological Disorders and Stroke (R01NS097823), a Siebel Stem Cell Center Seed grant and a Chan Zuckerberg Biohub investigator award. Thomas Li and John Blair are co-first authors of the paper. Other co-authors are project scientist Taesun Yoo and molecular and cell biology Professor Dirk Hockemeyer of Berkeley and neurosurgeon Gerald Grant and pediatric neurologist Brenda Porter of Stanford University.
