Cdk

Gene spots

Simulation

Cell survival

Network

Cell survival

j Drug ]

Fig. 9.1. Example from biological sampling to simulation. Messenger RNAs are extracted from normal or tumor cells. The ratio of their concentrations is measured on a DNA microarray. The mRNA transcribed from the "REG" or "CDK" genes are more abundant in the tumor cells than in normal cells, switching on the corresponding spots (black spot). The opposite holds true for the "APO" gene, giving a spot in a different color. The other spots are unaffected, indicating that the concentrations are similar in both samples for the corresponding gene. These partial results are compatible with the idea that "REGulator" encodes a protein that activates the expression of the "Cell Division Kinase" gene and inhibits that of the "APOptosis" gene. With the help of interaction database, literature surveys and automated inference algorithms49 a portion of the genetic network can be deduced, whereby REG' and REG" encode intermediate (arrow: activation, bar: inhibition). In this network, activation of REG by agents stimulating mitosis yields an hyperactivation of CDK, directly and via REG". This activation results into cell division and tumor proliferation (center bottom). In the absence of a mitogenic agent, division and apoptosis remain balanced and the cell survives without dividing (right top). This equilibrium can be studied according to different kinetics coefficients for each interaction through simulation of a mathematical model constructed to describe the dynamics of the gene expression sustained by the network. The simulation may predict in which direction the network will re-equilibrate when conditions change (for instance, drugs intake). In this simple example, it can easily be seen that REG" inactivation disfavors division and lifts apoptosis inhibition (right bottom). The prediction is that the tumor cell will thus be killed by this drug. Example adapted with authorization from.38

or technological networks such as electronic circuits. Genomics networks are artificial constructions representing some knowledge of system components and their putative interactions. These networks are only effective representations that are supposed to contain the essential properties and logic of the real biological regulatory process in an organism. As an abstrac-

Metabolites

Fig. 9.2. Molecular spaces of macro-molecular networks. The genome is made by the set of the whole DNA sequence and data on the genes. The set of transcribed mRNA and transcription factors constitutes the transcriptome. The Proteome collects data on proteins and their interactions and the metabolome focuses on an organism specific set of metabolic reactions. The different molecular realms overlap and are inter-regulated.

Metabolites

Fig. 9.2. Molecular spaces of macro-molecular networks. The genome is made by the set of the whole DNA sequence and data on the genes. The set of transcribed mRNA and transcription factors constitutes the transcriptome. The Proteome collects data on proteins and their interactions and the metabolome focuses on an organism specific set of metabolic reactions. The different molecular realms overlap and are inter-regulated.

tion they form ideal working objects for the theoretical analysis. The study of interaction interactive networks is in fact currently the main method of investigation in Systems Biology, to the point one may well consider this field as the molecular biology at the network level.

Three types of networks are generally considered that correspond to different molecular spaces. Metabolic networks represent metabolites and the chemical reactions they undergo due to enzymes, a domain which is often called the metabolome. Protein networks collect the interactions among proteins, in particular protein complex formation and dissociation and proteins altering each other, one speaks of the proteome. Gene networks (or genetic regulatory networks), where a gene is linked to another when the protein product of the first, dubbed transcription factor, regulates the ac tivity of the second are in a broad sense the field of the genome (figure 9.2). Several articles and databases have been respectively published and built with this view in mind, for gene networks,26 protein networks31'52'62 and metabolic networks.36'66

These different networks are very much interwoven since genes affect other genes by way of proteins that may well be activated by metabolic reactions. Despite the interconnections between the different levels, the distinction of several molecular spaces is a very common view that finds its origin in the different experimental and analytical tools required for experimenting with each type of molecule. Furthermore it has been quite common to organize research in a layered manner within these different levels, as homogeneous systems with the same kind of elements and interactions are expected to be more tractable. A last reason of this arbitrary segregation is that these molecular spaces are not equally accessible to experimentation. Genetic networks are in the limelight of most current experimental and theoretical efforts in System Biology since molecular biology provides very efficient tools to operate and perform measurements at their level. Despite much development, proteome and metabolome are to this day still less accessible to high throughput experimentation.

In the following discussion we will essentially focus on the genetic network regulating gene transcription and protein expression. This is the type of bio-molecular interaction network that, at this time, is experimentally the most accessible through biomolecular technologies. The work flow of Systems Biology proceeds first by the reconstruction of a network, which is the identification of its components and interactions from genomics raw data. Once obtained, this picture of functional relationships between biological components can itself be considered an object of biology and its topology become the subject of investigations for underlying biological principles. Network structures must finally be complemented with dynamical models in order to gain an understanding of the system's behavior under physiological or pathological conditions (see figure 9.1).

Was this article helpful?

0 0

Post a comment