to cadherin and gained an affinity for extracellular matrix components via integrins. Cadherin was removed from the cell surface as
the skeletogenic cell went through EMT and was immediately
endocytosed [5, 6].
While these efforts were productive, we were bothered by the
approach. It seemed as if we were stuck playing hunches about
molecules and it did not seem like we were gaining a broad understanding of how the EMT process was coordinately regulated. We
did not understand the detailed cell biology behind the changes,
and we did not understand what was directly controlled by the
regulatory apparatus. Productive genetic approaches were being
taken in other embryonic systems, most notably by Maria Leptin
in Drosophila [7], but the sea urchin lacked the use of genetics as a
tool for discovery. So, my choices were either to switch to a more
tractable genetic model, or to find a way to move forward using the
sea urchin system.
At about that time new discoveries in the lab sent us down a
different path. We began identifying genes that regulated developmental specification. Because of that, Eric Davidson at Caltech,
with a longstanding focus on mechanisms of gene regulation,
asked if I might be interested in joining him and his lab in identifying the gene regulatory network (GRN) that established early
specification of the sea urchin embryo. I thought about it overnight
and realized that if we were able to identify the regulatory network
governing specification we would also be generating the regulatory
apparatus controlling the EMT. With that goal in mind I called Eric
back and eagerly entered into an entirely new direction for the lab.
Other members of the sea urchin community quickly joined in and
over the next 10 years our understanding of the complexity of early
specification grew enormously. The skeletogenic cell lineage led the
way with a detailed regulatory sequence [8].
With a fairly detailed GRN model in hand we returned to the
EMT problem. Could we use the GRN to understand how the
EMT is regulated and coordinated? Our strategy was fairly simple.
Since we knew when the EMT began we started by asking what
regulatory events occurred in the 2 h prior to launching the EMT.
We systematically knocked down each of the transcription factors in
the GRN that were activated within that two-hour time period and
asked whether, in the absence of that transcription factor, was the
EMT crippled? To assess the outcome, we designed several simple
assays, and used time-lapse microscopy to watch the cell behavior
during the EMT. To focus on the EMT only, the assays employed
fluorescently tagged skeletogenic cells in embryos that were unlabeled. We could also assess whether each EMT event was cell
autonomous or if there might be a non-autonomous component
to the process. Lindsay Saunders, a graduate student, and I did the
analysis [9]. We learned that at least 10 transcription factors were
Perspective on Epithelial-Mesenchymal Transitions in Embryos
9
the skeletogenic cell went through EMT and was immediately
endocytosed [5, 6].
While these efforts were productive, we were bothered by the
approach. It seemed as if we were stuck playing hunches about
molecules and it did not seem like we were gaining a broad understanding of how the EMT process was coordinately regulated. We
did not understand the detailed cell biology behind the changes,
and we did not understand what was directly controlled by the
regulatory apparatus. Productive genetic approaches were being
taken in other embryonic systems, most notably by Maria Leptin
in Drosophila [7], but the sea urchin lacked the use of genetics as a
tool for discovery. So, my choices were either to switch to a more
tractable genetic model, or to find a way to move forward using the
sea urchin system.
At about that time new discoveries in the lab sent us down a
different path. We began identifying genes that regulated developmental specification. Because of that, Eric Davidson at Caltech,
with a longstanding focus on mechanisms of gene regulation,
asked if I might be interested in joining him and his lab in identifying the gene regulatory network (GRN) that established early
specification of the sea urchin embryo. I thought about it overnight
and realized that if we were able to identify the regulatory network
governing specification we would also be generating the regulatory
apparatus controlling the EMT. With that goal in mind I called Eric
back and eagerly entered into an entirely new direction for the lab.
Other members of the sea urchin community quickly joined in and
over the next 10 years our understanding of the complexity of early
specification grew enormously. The skeletogenic cell lineage led the
way with a detailed regulatory sequence [8].
With a fairly detailed GRN model in hand we returned to the
EMT problem. Could we use the GRN to understand how the
EMT is regulated and coordinated? Our strategy was fairly simple.
Since we knew when the EMT began we started by asking what
regulatory events occurred in the 2 h prior to launching the EMT.
We systematically knocked down each of the transcription factors in
the GRN that were activated within that two-hour time period and
asked whether, in the absence of that transcription factor, was the
EMT crippled? To assess the outcome, we designed several simple
assays, and used time-lapse microscopy to watch the cell behavior
during the EMT. To focus on the EMT only, the assays employed
fluorescently tagged skeletogenic cells in embryos that were unlabeled. We could also assess whether each EMT event was cell
autonomous or if there might be a non-autonomous component
to the process. Lindsay Saunders, a graduate student, and I did the
analysis [9]. We learned that at least 10 transcription factors were
Perspective on Epithelial-Mesenchymal Transitions in Embryos
9