user login 

BITS2007 Meeting
BITS2007 Meeting

26-28 April 2007 Napoli, Italy

email support
Home > Contribution details
get PDF of this contribution get XML of this contribution get ICal of this contribution
An agent-based model of microRNA-mediated gene regulation: simulation and interpretation of observational data
Gene regulatory networks, especially those involving non-coding RNA, represent the
paradigm of complex systems: many simple processes, under selective pressures,
synergistically interact to produce a desired global behaviour, often exhibiting
hierarchical self-organization. The mathematical modelling of this kind of systems by
differential equations is an hard task, mainly because the great number of required
reaction parameters, for which reliable values are rarely available. Agent-based
models,  allowing us to abstract from biochemical detail, are an useful tool to
investigate gene regulatory mechanisms. Moreover, the agent approach provides models
that are inherently multi-scale and stochastic, addressing two main problems of
Systems Biology. In this work  we report about an agent based-model developed to
interpret a set of observational data, simulating some different scenarios on the
basis of available biological knowledge.

MicroRNAs are short single-stranded RNA molecules that negatively regulate gene
expression at post-transcriptional level. In the Cell Biology Laboratory, we depleted
v by decoy transfection -  two microRNAs (miRNA 221 and 222) in Human Umbilical Vein
Endothelial Cells. Correspondently, we found an important change in the overall 
microRNA signature vnot in the depleted microRNAs only- and in the expression of a
housekeeping gene, c-kit, that is under the post-transcriptional control of miRNA 221
and 222.   The c-kit expression showed a marked oscillatory pattern that vanished with
time, notably reaching  lower values after microRNAs depletion. 
This kind of behaviour is the hallmark of a regulation feedback, whose aim is to
maintain stable the number of c-kit molecules inside the perturbed cell. The
regulation is consistent with several molecular pathways depicted by  three distinct
scenarios: 1. the auto-regulation of c-kit by a  negative feedback loop, 2. the
auto-regulation of the microRNA network by a negative feedback loop, and 3. a
combination of both the feedback loops. 
The three scenarios were simulated by an agent-based model of the system. The model
was developed using the NetLogo 3.1.4  platform,  with a decentralized swarm approach
and multiple agents locally interacting in agreement with simple rules, dictated by
biological knowledge.  To account for the cooperative and aspecific action of
microRNAs, another coding gene under microRNAs control, besides c-kit, was
The cell, in a simplified 2-dimensional representation, was divided in nucleus and
cytoplasm. Agents were created to represent a microRNA network constituted by miRNA
221, miRNA 222 and other 50 microRNAs, that regulated c-kit and the other coding gene
vin the model, gene alfa,   but were not depleted by decoy transfection. Other agents
were created to represent  c-kit mRNA, c-kit protein, gene alfa mRNA, gene alfa
protein, ribosomes, decoy molecules, transcription factors of microRNAs and coding
genes, promoter regions of microRNAs and coding genes. To model the negative feedback
loops, we allowed proteins and microRNAs to occupy, for a limited amount of time,
their  promoters, preventing the attachment and firing action of the transcription
factors. Agents moved randomly inside the cell and received a stochastic life-time.

The simulations clearly showed the plausibility of a microRNA post-trascriptional
regulation network, auto-regulated by a negative feedback and synergistically
interacting with the coding gene transcriptional regulation.  The observed
time-varying expression of c-kit was interpreted as a macro-behaviour emergent from
In our experience, agent-based modelling makes the inherent complexity of  the
multiple, interacting processes of gene regulation manageable, by choosing an
appropriate level of abstraction. Moreover, the behaviour rules, that constitute the
backbone of the agent-based model, are  made of the same natural language narratives
commonly used by biologists to explain phenomena, greatly strengthening the
interdisciplinary nature of the research. In our opinion, abstraction and
interdisciplinariety makes the agent-based modelling approach suitable to address the
question about the interplay between molecular pathways in gene regulatory networks.
Id: 199
Place: Napoli, Italy
Centro Congressi "Federico II"
Via Partenope 36
Starting date:
-- not yet scheduled --   
Duration: 01h00'
Contribution type: Poster
Primary Authors: RIPOLI, Andrea (Institute of Clinical Physiology CNR, Pisa)
Co-Authors: MERCATANTI, Alberto (Institute of Clinical Physiology CNR, Pisa)
RAINALDI, Giuseppe (Institute of Clinical Physiology CNR, Pisa)
Presenters: RIPOLI, Andrea
Included in session: Poster Session
Included in track: Gene expression and system biology | Last modified 08 July 2009 10:35 |

Powered by Indico 0.90.3