Researchers have differentiated human embryonic stem cells into thyroid cells for the first time, according to a report.
By overexpressing two transcription factors — PAX8 and NKX2-1 — Terry Davies, MD, of Mount Sinai Medical Center in New York City, and colleagues said they were able to induce stem cells into thyroid cells. They reported their results at the American Thyroid Association meeting.
Exposing the differentiated cells to activin A and then thyroid-stimulating hormone (TSH) further matured them into functional thyroid follicle cells, Davies told.
“We’ve shown very clearly that by pushing human embryonic stem cells into thyroid cells, these cells form follicles, and they function,” Davies said. “They take up radioiodine and look like normal thyroid cells.”
The necessary steps to transform human stem cells into thyroid cells are still unclear, but mouse models have shown that overexpression of the regulatory transcription factors pax8 and nkx2-1 allows murine embryonic stem cells to differentiate into thyroid follicular cells.
Davies and his group decided to take those results into human cells, using lentiviral vectors to insert PAX8 and NKX2-1 into human embryonic stem cells.
Although overexpressing these transcription factors did enable the stem cells to differentiate into thyroid cells, these cells were not yet functional, Davies said.
“What’s interesting is that while putting transcription factors into human embryonic stem cells does start the program, those cells don’t make thyroglobulin,” Davies said. “They show a very little bit of thyroglobulin mRNA, but we’ve not been able to detect protein in those cells. We have to mature the cell further.”
To do that, Davies and his team exposed the cells to activin A, moving them into a different differentiation phase in which they transformed into endodermal cells. The final step is to “hit them with thyroid stimulating hormone,” Davies said.
“That activates the TSH receptor and takes them through to make thyroid follicles,” he said. “Those thyroid follicles express lots of thyroglobulin and lots of thyroid peroxidase, and they take up radioiodine. So they seem to be good-looking thyroid follicles.”
In their study, the researchers also found that HEX, which is regulated by PAX8 and is important for provoking thyroid cell differentiation, was markedly underexpressed in the stem cells compared with normal human thyroid cells. But the calcium-dependent transcription repressor Downstream Regulatory Element Antagonist Modulator (DREAM) was suppressed as the cells differentiated.
This suggests that the procedure worked via the induction of HEX and suppression of DREAM in undifferentiated stem cells, the researchers said.
“This is a model system that has a lot of potential in a number of different ways,” Davies said. “First, it can teach us about the differentiation of the thyroid gland and how the thyroid gland develops. Second, it can be used for screening of drugs to look for effects on human thyroid cells.”
Davies said his group currently has a program for developing molecules for diagnostic and therapeutic purposes, with two lead candidates published this month in Thyroid. Some of these uses include helping radioiodine to get into metastases, as well as replacing recombinant TSH when assessing patients with thyroid cancer.
“These molecules can also be used to aid the differentiation of thyroid cells, so these two programs come together,” Davies said.
Ultimately, the goal is to build a working thyroid that can replace a diseased one. Although this is still many years away, he acknowledged, it does promise the opportunity to personalize treatment.
“You can take stem cells from each individual, differentiate them into thyroid cells, and replace a damaged or removed thyroid gland,” he said.
Bryan Haugen, MD, of the University of Colorado in Denver, who was not involved in the study, expressed some cautions about forcing cells to overexpress genes.
“When you manipulate a gene and overexpress it, now it’s no longer recapitulating normal biology,” Haugen said. “You’ve forced something that could make it aggressive or possibly turn it into a cancer, versus doing something try to follow the natural progression and activate it more naturally.”
“This is a bit of an artificial system,” he added, “but the good thing is that the way it’s set up, it looks very promising for making normal thyroid follicular cells.”
Haugen also added that the work can help researchers better understand thyroid biology and pathophysiology. “It’s a new, very useful tool for our field. We can have a model of thyroid development to see where things can go right and where they can go wrong.”
Davies said his next steps include turning the working thyroid follicles into an actual gland. “We’re using structural supports to try and develop a larger in vitro thyroid gland, then get them transplanted into mice.”
He and his team are also further developing their small molecule program, to find candidates that can activate thyroid cell differentiation and be used for other applications.
