Biology Reference
In-Depth Information
adopted over recent years to study how patterns are
generated within roots and shoots
[1]
.
Here, we review how both components of systems
biology have been adapted to the study of plant development
and sketch how this is increasing our understanding of the
factors and the logic that drive plant development.
longitudinal axis with 50% just above the stem cell niche,
then 40 units of expression would be assigned to the
epidermal cells in this location. A more sophisticated
approach made use of an expectation maximization algo-
rithm, which treated the longitudinal and radial datasets as
a matrix in which the sums for each row and column must
be constant
[4]
. Validation efforts have generally been
consistent with predictions of the expectation maximiza-
tion approach. However, for the long term it will be
important to determine directly the expression levels in
cells at different developmental stages along each of the
cell files in the root.
Several alternative approaches have been used to assay
genome-wide cell type-specific expression in plants. These
include the use of laser capture microdissection and tagged
ribosomal proteins coupled with immunoprecipitation.
Despite being tested on Arabidopsis
[5]
, laser capture
microdissection combined with microarray analysis has
been used most extensively for rice
[6]
. Immunoprecipita-
tion of polysomes involves fusing an epitope tag such as the
antigenic portion of the Myc protein to a ubiquitously
expressed ribosomal protein expressed behind a tissue-
specific promoter. The construct is introduced into plants,
allowing polysomes to be immunoprecipitated with anti-
bodies specific to the epitope. This was performed on
Arabidopsis root tissues with several tissue-specific
promoters driving the epitope-tagged ribosomal protein
[7]
.
The ability to assay polysomes makes this approach well
suited to identifying the mRNAs that are actively under-
going translation. A more recently developed technique
involves isolating nuclei that have been tagged using tissue-
specific promoters
[8]
. To date, this approach has been
primarily used for chromatin immunoprecipitation assays
(see Chapter 4), although it would be interesting to perform
a side-by-side comparison of RNAs found in the nucleus
and those located in the cytoplasm.
Analysis of Gene Activity in Space and Time
An early effort to assay gene expression in individual cell
types within a plant organ involved the use of five trans-
genic lines in which individual cell populations were
marked with green fluorescent protein (GFP)
[2]
. Enzymes
were used to dissociate the cells of the Arabidopsis root,
which were subsequently sorted using a fluorescence-acti-
vated cell sorter (FACS). From the cells enriched for GFP
expression, RNA was extracted and used for microarray
analysis. To profile expression along the developmental
timeline, roots were dissected into sections corresponding
to developmental zones and RNA was extracted from the
isolated sections and used for microarray analysis. A more
recent analysis extended this Root Map to include nearly all
of the known cell types in the root
[3]
. Because root growth
is not precisely synchronous, to gain a more precise
knowledge of developmental stages, 13 sections along the
longitudinal axis (
Figure 20.2
) were cut from individual
roots and mRNA was extracted and used for microarray
analysis.
The microarray data sets were first analyzed sepa-
rately for the cell types and developmental stages. In
each case clusters of co-regulated genes were identified
that were enriched for biological functions. In the case of
the developmental stage-specific data a surprising finding
was that a large number of the co-regulated clusters
exhibited fluctuating behavior such that their expression
peaked at two or more different stages of development
[3]
. This was unexpected because development is
normally thought of as a unidirectional, progressive
process. It is an example of the type of finding that could
not have been foreseen prior to performing this type of
discovery-driven experiment. More recently, the same
approach has been used to profile small RNAs in specific
cell types and developmental stages. At least 70 novel
microRNAs (see Chapter 2) were identified using this
approach, and many previously characterized microRNAs
were shown to be expressed in a cell type- or develop-
mental stage-specific manner.
Two different approaches have been used to combine
the radial and longitudinal datasets to infer expression for
any cell within the root. The first approach used the
developmental stage data to divide expression for each cell
type proportionally along the longitudinal axis
[2]
. For
example, if expression for gene X was 80 units in the
epidermis and its expression was distributed along the
The Use of Spatiotemporal Specific
Expression Data to Analyze Development
The primary utility of genome-wide expression datasets is
to generate hypotheses that will shed light on biological
processes. A straightforward hypothesis is that a gene
expressed specifically in a particular cell type and/or
developmental stage is involved in a process performed
specifically in that cell or stage. An example of testing this
type of hypothesis is the identification of a family of genes
required for Casparian strip formation, a terminal differ-
entiation property of endodermal cells, based on their
specific expression in maturing endodermal cells
[9]
.
Another example is an effort to identify transcriptional
regulators of the transition from cellular proliferation to
differentiation. The immediate progeny of stem cells
