Jose Diaz, Universidad Autónoma del Estado de Morelos
29 August 2013
Every cell is continuously monitoring their environment and receiving information from it. This information has the form of chemical, electrical or mechanical signals that should induce a specific adaptive cell response. The first problem for the cell, concerning the interpretation of this information, is how the message must be coded in order to be readable by the genetic machine.
In the case of hormonal and growth factor inputs, information is somehow coded in the fraction of specific activated receptors. In the case of electric signals, information is generally coded in the frequency of membrane depolarization. Electric and chemical signals are not mutually exclusive, but they can interact among them giving rise to an additional level of complexity for signaling coding and interpretation.
Focusing on hormonal and growth factor signals, the activation of specific receptors at the cell surface is, in most of the cases, a switch process with a time course that depends on the kinetic properties of the ligand-receptor complex formation and, in some cases, of the (ligand-receptor)-(ligand-receptor) dimmer formation. Biochemical switches are practically immune to noise, so the coding of information at the cell surface seems to be a high-fidelity process that is virtually unaffected by external stochastic factors.
The transmission of the message content to the nucleus must be done with high fidelity. Cascades of phosphorylation reactions are good examples of this task. The ultra-sensitivity and high specificity of these cascades, which implies switches with high Hill exponents, protect the message content from intracellular noise. Stable limit cycles are also good candidates for protecting encoded information from noise. In excitable systems, the information is transmitted in a robust form by bursts of depolarization, while the duration of the inter bursts interval is more susceptible to noise. Crosstalk between signaling pathways also conveys an exchange of information that adjust the cellular performance in response to diverse simultaneous environmental inputs.
The final goal of this flow of information is the activation or inhibition of a specific set of genes, with correct timing and duration. At this point, the message content must be decoded by the translational machinery. It is possible to measure the uncertainty in the response of a set of genes to an input by using the Shannon entropy H. This value can be utilized to calculate the value of I, i.e., the content of information transmitted through the communication channel.
A plausible method for measuring H consists in the determination of the probability of expression of each gene in response to a given input. This method can be applied to any set of genes, independently of the structure of the gene regulatory network. However, each topology corresponds to a different probability distribution and as consequence, has a specific value of H. Probabilities can change in time in response to a time-varying signal, thus H also becomes a time-dependent variable.
In this form, cells can be pictured as stochastic machines in which their gene regulatory networks exhibit a particular probability distribution of gene activation for each specific input. Thus, for each input, the uncertainty in the cell response is measured by H, and this uncertainty is due to noise in the communication channel.
It is also possible to estimate the bitrate and capacity of a cellular communication channel under different inputs by using the Systems Analysis tools together with Information Theory. The bitrate is related to the rate of information transfer to the nucleus due to the flow of protein molecules, which act as regulators of gene expression. It can be used to estimate the number of molecules that enter the nucleus per unit of time. If we know the rate of flow of molecules into the nucleus, the number of bits of information that these molecules carry per unit of time can be estimated.
In this form, Information Theory is beginning to be useful for cell biologists interested in the complex dynamical processes that sustain the flow and interpretation of information from the environment. It will enter its maturity as soon as theoretical and experimental biologists work together in the elucidation of: 1) the molecular mechanisms that sustain the flow of information over the cell signaling network, under noisy conditions, and 2) the rules that lead to specific cell responses to each type of input to this network.
This will help us to understand how cells take decisions about the proliferation/differentiation balance, and about the adjustments that they must do in their current dynamical state to counteract harmful environmental cues, which is of major importance in the case of plants.
Is Information Theory useful for cell biologists?
29 August 2013
Every cell is continuously monitoring their environment and receiving information from it. This information has the form of chemical, electrical or mechanical signals that should induce a specific adaptive cell response. The first problem for the cell, concerning the interpretation of this information, is how the message must be coded in order to be readable by the genetic machine.
In the case of hormonal and growth factor inputs, information is somehow coded in the fraction of specific activated receptors. In the case of electric signals, information is generally coded in the frequency of membrane depolarization. Electric and chemical signals are not mutually exclusive, but they can interact among them giving rise to an additional level of complexity for signaling coding and interpretation.
Focusing on hormonal and growth factor signals, the activation of specific receptors at the cell surface is, in most of the cases, a switch process with a time course that depends on the kinetic properties of the ligand-receptor complex formation and, in some cases, of the (ligand-receptor)-(ligand-receptor) dimmer formation. Biochemical switches are practically immune to noise, so the coding of information at the cell surface seems to be a high-fidelity process that is virtually unaffected by external stochastic factors.
The transmission of the message content to the nucleus must be done with high fidelity. Cascades of phosphorylation reactions are good examples of this task. The ultra-sensitivity and high specificity of these cascades, which implies switches with high Hill exponents, protect the message content from intracellular noise. Stable limit cycles are also good candidates for protecting encoded information from noise. In excitable systems, the information is transmitted in a robust form by bursts of depolarization, while the duration of the inter bursts interval is more susceptible to noise. Crosstalk between signaling pathways also conveys an exchange of information that adjust the cellular performance in response to diverse simultaneous environmental inputs.
The final goal of this flow of information is the activation or inhibition of a specific set of genes, with correct timing and duration. At this point, the message content must be decoded by the translational machinery. It is possible to measure the uncertainty in the response of a set of genes to an input by using the Shannon entropy H. This value can be utilized to calculate the value of I, i.e., the content of information transmitted through the communication channel.
A plausible method for measuring H consists in the determination of the probability of expression of each gene in response to a given input. This method can be applied to any set of genes, independently of the structure of the gene regulatory network. However, each topology corresponds to a different probability distribution and as consequence, has a specific value of H. Probabilities can change in time in response to a time-varying signal, thus H also becomes a time-dependent variable.
In this form, cells can be pictured as stochastic machines in which their gene regulatory networks exhibit a particular probability distribution of gene activation for each specific input. Thus, for each input, the uncertainty in the cell response is measured by H, and this uncertainty is due to noise in the communication channel.
It is also possible to estimate the bitrate and capacity of a cellular communication channel under different inputs by using the Systems Analysis tools together with Information Theory. The bitrate is related to the rate of information transfer to the nucleus due to the flow of protein molecules, which act as regulators of gene expression. It can be used to estimate the number of molecules that enter the nucleus per unit of time. If we know the rate of flow of molecules into the nucleus, the number of bits of information that these molecules carry per unit of time can be estimated.
In this form, Information Theory is beginning to be useful for cell biologists interested in the complex dynamical processes that sustain the flow and interpretation of information from the environment. It will enter its maturity as soon as theoretical and experimental biologists work together in the elucidation of: 1) the molecular mechanisms that sustain the flow of information over the cell signaling network, under noisy conditions, and 2) the rules that lead to specific cell responses to each type of input to this network.
This will help us to understand how cells take decisions about the proliferation/differentiation balance, and about the adjustments that they must do in their current dynamical state to counteract harmful environmental cues, which is of major importance in the case of plants.
Competing interests
None declared