Biology Reference
In-Depth Information
controlled by a small set of flowering time genes, which in turn activate the floral meristem identity
genes
APETALA1
(
AP1
) and
LEAFY
(
LFY
). These meristem identity genes stimulate the expression of
floral organ identity genes. Organ identity genes act in a combinatorial fashion to specify the different
types of floral organs: sepals, petals, stamens and carpels. According to the 'floral quartet model', the
proteins encoded by floral homeotic genes assemble into distinct multimeric complexes in an organ-
type specific manner [Theissen and Saedler, 2001]. All of the floral homeotic proteins present in those
complexes belong to the MADS-box family of transcription factors. Evidence from yeast n-hybrid studies
suggests that higher-order complex formation is mediated mostly by members of the SEPALLATA (SEP)
subfamily of MADS-domain proteins [Honma and Goto, 2001; Immink
et al.
, 2009], which are required
for specification of the identities of all 4 types of floral organs. The
SEPALLATA
subfamily consists of
4 largely redundant genes (
SEP1
-
SEP4
). Combined loss-of-function mutations in each of the 4 genes
lead to homeotic conversion of all types of floral organ to leaf-like organs [Ditta
et al.
, 2004].
The formation of multimeric protein complexes seems to be not only required for the regulation of
downstream targets, but may also play a role in positive autoregulation of floral homeotic genes. In
addition to initial upregulation by upstream factors, autoregulation has been observed for key floral
homeotic genes, like
APETALA3
(
AP3
) and
PISTILLATA
(
PI
) [Jack
et al.
, 1994; Krizek and Meyerowitz,
1996; Hill
et al.
, 1998; Honma and Goto, 2000],
AGAMOUS
(
AG
) [Gomez-Mena
et al.
, 2005] and
SEPALLATA3
(
SEP3
) [Kaufmann
et al.
, 2009].
The requirement for autoregulation involving heterodimer formation has so far been characterized
primarily for AP3 and PI, which are the two floral homeotic B function proteins specifying petal and
stamen identity. However, genome-wide binding data for SEP3 indicate that it binds to the promoters
of almost all of the floral homeotic genes, and induction experiments also show that it can upregulate
the expression of these genes [Kaufmann
et al.
, 2009]. Since SEP3 is a key mediator of heteromeric
higher-order complex formation between floral homeotic proteins, and autoregulation is observed for
nearly all floral homeotic proteins, this likely indicates that autoregulation is mediated by higher-order
complexes, although the initial expression of floral homeotic genes is unaffected in
sepallata
triple
mutants [Pelaz
et al.
, 2000]. Thus, the combination of results from protein-protein and protein-DNA
interaction studies as well as genetic evidence suggest a complex scenario for the establishment of the
different floral organ identities by multiple direct protein-protein and regulatory interactions. Despite
the fact that not all interactions have been confirmed
in planta
yet, current evidence allows us to generate
a model for interactions during early flower development. Furthermore, other recent evidence suggests
post-transcriptional control mechanisms in the network, such as the role of the microRNA miR172 in
the translational repression of the spatial regulator APETALA2 (AP2) [Chen, 2004].
The complexity of direct molecular interactions necessitates the use of novel computational tools
to understand the flowering process, optimally those which would allow for the explicit modelling of
transcription, translation and protein binding reactions. A limitation at this point is a general lack of
quantitative data for these different processes, restricting the modelling to generic estimates.
Ordinary differential equations are widely accepted as a modelling method for biological pathways
[Sun and Zhao, 2004]. However, this method carries the disadvantage of being difficult to represent
schematic information such as pathway models illustrated using biological elements such as mRNAs
and proteins. The estimation of the required parameters for a simulation, especially in the case of a gene
regulatory network, is also an open issue. In this regard, Petri nets offer an attractive alternative for
simple construction, visualization and simulation of gene regulatory networks.
A Petri net is a mathematical model used for the representation and analysis of concurrent processes.
Petri nets are described in part by the visual elements “place”, “transition”, “arc”, and “token” (Fig. 1).
