Biology Reference
In-Depth Information
Box 9.1
Feedforward loops are network motifs
In a FFL, the first factor of the loop regulates the last factor directly (short arm)
and indirectly (long arm) through an intermediate factor. Four variables
determine the dynamic behavior of a FFL. These variables are (i) the overall
regulatory effect of the arms. The long arm can be positive or negative. (ii)
Whether arms are coherent or incoherent (the same or opposite sign). (iii) The
logic gate that integrates the signals from the two upstream nodes. This
determines whether the downstream node can be regulated by both arms
(AND gate) or by either arm (OR gate). (iv) The type of change in signal that
passes through the loop. Does the change in the signal directed from an active
to a repressed state (ON-to-OFF), or its inverse? Depending on the combina-
tion of these four variables, FFLs can act as pulsers, persistence detectors of
change, or accelerators of response time. For example, consider the Type 3
coherent FFL where the upstream node directly and indirectly represses the
downstream node through a second inhibitor activated by the first. When this
specific architecture operates with an OR Gate, the resulting FFL produces a
delayed response to a decrease (ON-to-OFF step) in the upstream node. As
result, this type of FFL can filter ON-to-OFF pulses and only respond to
persistent changes in the upstream node. This type of architecture exists in the
Yan network linking
Pnt-P1
,
miR-7
, and
YAN
and is predicted to buffer YAN
expression from stochastic decreases of
Pnt-P1
.
mutually exclusive expression of the miRNAs and their targets. miRNAs can
also be found in single-negative FBLs, where themiRNA represses an activator
of the miRNA. A limitation of these genome-wide studies has been their
ability to experimentally verify the existence of computationally predicted
links. For example,
Martinez
et al.
(2008)
derived their network from a
combinationof computational predictions ofmiRNA targets andY1Hexperi-
ments that established which TFs bound a library of DNA elements containing
predicted promoters for miRNA genes.
Several experimentally verified examples of miRNA-containing FBLs
have been described. Some of these are listed in
Table 9.1
and have been
reviewed extensively. Instead, we focus on two examples where a role for
miRNA-mediated robustness has been shown. In
Drosophila melanogaster
,
the transcription repressor YAN binds and represses the transcriptional
enhancer of the
miR-7
gene. In turn, miR-7 binds and represses the
protein expression of
YAN
. YAN and miR-7 are part of a network that
regulates the transition from multipotent retinal progenitor cells to differ-
entiated photoreceptors (
Li and Carthew, 2005
;
Li
et al
., 2009
). The YAN
network is a bistable system that transitions from a high YAN/low miR-7
