change polarity, invade through the basement membrane,
de-adhere from the adherens junction, and the plasma membrane
is remodeled), though differences are observed in different model
systems. What is clear is that in both the carcinoma and embryo
systems, the molecular basis of this complex cellular event called
EMT is incompletely understood. Indeed, of the several thousand
papers a year on EMT, most focus on the epiphenomenon, that is,
does the phenotypic change occur to an epithelial culture, or layer,
under applied experimental conditions? Far fewer papers focus on
the functional mechanics of that EMT in molecular detail.
A major reason for not understanding the EMT process in
greater detail is that most systems are asynchronous, that is, the
cells undergoing an EMT are at different states at any given time
making it difficult to deduce the precise sequence of molecular
events. A few cases of EMT in embryonic systems do offer synchrony, but each of these also has shortcomings. For example,
ventral furrow formation in Drosophila melanogaster (Drosophila)
provides a near synchronous EMT of mesoderm cells, and some
genes necessary for the process have been identified. However, the
difficulty in that system is that the number of mesoderm cells is
small relative to the remaining cells of the embryo, and the EMT
occurs relatively early in development, at a time when many maternally expressed genes are still expressed. This makes it difficult to
exploit the power of Drosophila genetics to discover the genes
mechanistically involved specifically in the EMT process [1]. Anchor
cell invasion in Caenorhabditis elegans is another embryonic EMT
in which one cell invades through the basement membrane as part
of vulval assembly [2]. In this case the system is genetically tractable
and a number of genes involved in the process have been identified.
There is no question of synchrony, since only the one cell participates. However, a shortcoming of this system for EMT analysis is
that the anchor cell does not complete an EMT. It breaches the
basement membrane in a manner similar to that utilized by cells
undergoing EMT in other systems, but it does not de-adhere from
the epithelium. The sea urchin embryo also has a population of cells
that undergo EMT at a precise time in early development and a
gene regulatory network of specification is well established for
those cells, making this a useful model system for understanding
control of the process [3]. Nevertheless, this system also has shortcomings in that the skeletogenic cells that go through the EMT are
only 5% of the population of cells in the embryo, making it a
challenge to determine the sequence of molecular events in that
small population.
Here we describe a method that can be used on any system to at
least partially overcome some of the shortcomings possessed by
many systems. Single cell RNA-sequencing (scRNA-seq) has
advanced to the point where one can obtain a profile of expressed
RNA in each cell. Computational approaches along with a temporal
304
Abdull J. Massri et al.
de-adhere from the adherens junction, and the plasma membrane
is remodeled), though differences are observed in different model
systems. What is clear is that in both the carcinoma and embryo
systems, the molecular basis of this complex cellular event called
EMT is incompletely understood. Indeed, of the several thousand
papers a year on EMT, most focus on the epiphenomenon, that is,
does the phenotypic change occur to an epithelial culture, or layer,
under applied experimental conditions? Far fewer papers focus on
the functional mechanics of that EMT in molecular detail.
A major reason for not understanding the EMT process in
greater detail is that most systems are asynchronous, that is, the
cells undergoing an EMT are at different states at any given time
making it difficult to deduce the precise sequence of molecular
events. A few cases of EMT in embryonic systems do offer synchrony, but each of these also has shortcomings. For example,
ventral furrow formation in Drosophila melanogaster (Drosophila)
provides a near synchronous EMT of mesoderm cells, and some
genes necessary for the process have been identified. However, the
difficulty in that system is that the number of mesoderm cells is
small relative to the remaining cells of the embryo, and the EMT
occurs relatively early in development, at a time when many maternally expressed genes are still expressed. This makes it difficult to
exploit the power of Drosophila genetics to discover the genes
mechanistically involved specifically in the EMT process [1]. Anchor
cell invasion in Caenorhabditis elegans is another embryonic EMT
in which one cell invades through the basement membrane as part
of vulval assembly [2]. In this case the system is genetically tractable
and a number of genes involved in the process have been identified.
There is no question of synchrony, since only the one cell participates. However, a shortcoming of this system for EMT analysis is
that the anchor cell does not complete an EMT. It breaches the
basement membrane in a manner similar to that utilized by cells
undergoing EMT in other systems, but it does not de-adhere from
the epithelium. The sea urchin embryo also has a population of cells
that undergo EMT at a precise time in early development and a
gene regulatory network of specification is well established for
those cells, making this a useful model system for understanding
control of the process [3]. Nevertheless, this system also has shortcomings in that the skeletogenic cells that go through the EMT are
only 5% of the population of cells in the embryo, making it a
challenge to determine the sequence of molecular events in that
small population.
Here we describe a method that can be used on any system to at
least partially overcome some of the shortcomings possessed by
many systems. Single cell RNA-sequencing (scRNA-seq) has
advanced to the point where one can obtain a profile of expressed
RNA in each cell. Computational approaches along with a temporal
304
Abdull J. Massri et al.