Limits...
The role of noise and positive feedback in the onset of autosomal dominant diseases.

Bosl WJ, Li R - BMC Syst Biol (2010)

Bottom Line: Autosomal dominant (AD) diseases result when a single mutant or non-functioning gene is present on an autosomal chromosome.These diseases often do not emerge at birth.Model pathways for two autosomal dominant diseases, AD polycystic kidney disease and mature onset diabetes of youth (MODY) were simulated and the results are compared to known disease characteristics.

View Article: PubMed Central - HTML - PubMed

Affiliation: Harvard Medical School, Boston, MA 02115, USA. william.bosl@childrens.harvard.edu

ABSTRACT

Background: Autosomal dominant (AD) diseases result when a single mutant or non-functioning gene is present on an autosomal chromosome. These diseases often do not emerge at birth. There are presently two prevailing theories explaining the expression of AD diseases. One explanation originates from the Knudson two-hit theory of hereditary cancers, where loss of heterozygosity or occurrence of somatic mutations impairs the function of the wild-type copy. While these somatic second hits may be sufficient for stable disease states, it is often difficult to determine if their occurrence necessarily marks the initiation of disease progression. A more direct consequence of a heterozygous genetic background is haploinsufficiency, referring to a lack of sufficient gene function due to reduced wild-type gene copy number; however, haploinsufficiency can involve a variety of additional mechanisms, such as noise in gene expression or protein levels, injury and second hit mutations in other genes. In this study, we explore the possible contribution to the onset of autosomal dominant diseases from intrinsic factors, such as those determined by the structure of the molecular networks governing normal cellular physiology.

Results: First, simple models of single gene insufficiency using the positive feedback loops that may be derived from a three-component network were studied by computer simulation using Bionet software. The network structure is shown to affect the dynamics considerably; some networks are relatively stable even when large stochastic variations in are present, while others exhibit switch-like dynamics. In the latter cases, once the network switches over to the disease state it remains in that state permanently. Model pathways for two autosomal dominant diseases, AD polycystic kidney disease and mature onset diabetes of youth (MODY) were simulated and the results are compared to known disease characteristics.

Conclusions: By identifying the intrinsic mechanisms involved in the onset of AD diseases, it may be possible to better assess risk factors as well as lead to potential new drug targets. To illustrate the applicability of this study of pathway dynamics, we simulated the primary pathways involved in two autosomal dominant diseases, Polycystic Kidney Disease (PKD) and mature onset diabetes of youth (MODY). Simulations demonstrate that some of the primary disease characteristics are consistent with the positive feedback-stochastic variation theory presented here. This has implications for new drug targets to control these diseases by blocking the positive feedback loop in the relevant pathways.

Show MeSH

Related in: MedlinePlus

Y haploinsufficiency. Simulation results for each of the networks in Figure 1 when Y is haploinsufficient and Gaussian noise is added to X production.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC2902440&req=5

Figure 3: Y haploinsufficiency. Simulation results for each of the networks in Figure 1 when Y is haploinsufficient and Gaussian noise is added to X production.

Mentions: All of the possible positive feedback loops that may be derived from a three-component network are shown in Figure 1. A positive feedback loop is defined here as a loop that encompasses at least two nodes and has an even number of negative regulators, either zero or two, in the loop. In each case, X is the signalling protein at the top of the loop, Y an intermediate protein, and D is a marker for the disease state. Simulations were run with either X or Y production reduced by one half, as a consequence of heterozygosity, and Gaussian noised added to the expression of the affected gene product. Time steps of one half week, and thus random variation on the same time scale, were similar to the lower end of random variations reported earlier [21]. Figure 2 shows time courses over an 80-year period for all of the loops in the X-heterozygosity case and Figure 3 shows the Y-heterozygosity cases. This time span was chosen to approximately represent a normal human lifespan. Network structures that cause D to increase to high levels early and irreversibly may be characterized as allowing or causing a high level of disease symptoms. These include 2.a, 2.c and 3.b for X-heterozygosity and all of the type 3 loops for Y-heterozygosity. Milder or variable disease symptoms may be expected when the D time course is somewhat repressed or arises intermittently. These include 1.c and 2.b for X- heterozygosity and 1.a, 1.c, all type 2 loops for Y heterozygosity. No disease occurs in networks 1.a, 1.b, 3.a and 3.c for X- heterozygosity; no such cases occur for Y-heterozygosity, although 1.b has a very low level of variable disease presence. These results are summarized in Table 1. The characterization of network dynamics in terms of disease onset as mild is somewhat general, with a wide range of dynamic responses possible. The severe labels (++) indicate cases where disease onset would be expected to be early, symptoms relatively stable and the likelihood of occurrence almost certain.


The role of noise and positive feedback in the onset of autosomal dominant diseases.

Bosl WJ, Li R - BMC Syst Biol (2010)

Y haploinsufficiency. Simulation results for each of the networks in Figure 1 when Y is haploinsufficient and Gaussian noise is added to X production.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC2902440&req=5

Figure 3: Y haploinsufficiency. Simulation results for each of the networks in Figure 1 when Y is haploinsufficient and Gaussian noise is added to X production.
Mentions: All of the possible positive feedback loops that may be derived from a three-component network are shown in Figure 1. A positive feedback loop is defined here as a loop that encompasses at least two nodes and has an even number of negative regulators, either zero or two, in the loop. In each case, X is the signalling protein at the top of the loop, Y an intermediate protein, and D is a marker for the disease state. Simulations were run with either X or Y production reduced by one half, as a consequence of heterozygosity, and Gaussian noised added to the expression of the affected gene product. Time steps of one half week, and thus random variation on the same time scale, were similar to the lower end of random variations reported earlier [21]. Figure 2 shows time courses over an 80-year period for all of the loops in the X-heterozygosity case and Figure 3 shows the Y-heterozygosity cases. This time span was chosen to approximately represent a normal human lifespan. Network structures that cause D to increase to high levels early and irreversibly may be characterized as allowing or causing a high level of disease symptoms. These include 2.a, 2.c and 3.b for X-heterozygosity and all of the type 3 loops for Y-heterozygosity. Milder or variable disease symptoms may be expected when the D time course is somewhat repressed or arises intermittently. These include 1.c and 2.b for X- heterozygosity and 1.a, 1.c, all type 2 loops for Y heterozygosity. No disease occurs in networks 1.a, 1.b, 3.a and 3.c for X- heterozygosity; no such cases occur for Y-heterozygosity, although 1.b has a very low level of variable disease presence. These results are summarized in Table 1. The characterization of network dynamics in terms of disease onset as mild is somewhat general, with a wide range of dynamic responses possible. The severe labels (++) indicate cases where disease onset would be expected to be early, symptoms relatively stable and the likelihood of occurrence almost certain.

Bottom Line: Autosomal dominant (AD) diseases result when a single mutant or non-functioning gene is present on an autosomal chromosome.These diseases often do not emerge at birth.Model pathways for two autosomal dominant diseases, AD polycystic kidney disease and mature onset diabetes of youth (MODY) were simulated and the results are compared to known disease characteristics.

View Article: PubMed Central - HTML - PubMed

Affiliation: Harvard Medical School, Boston, MA 02115, USA. william.bosl@childrens.harvard.edu

ABSTRACT

Background: Autosomal dominant (AD) diseases result when a single mutant or non-functioning gene is present on an autosomal chromosome. These diseases often do not emerge at birth. There are presently two prevailing theories explaining the expression of AD diseases. One explanation originates from the Knudson two-hit theory of hereditary cancers, where loss of heterozygosity or occurrence of somatic mutations impairs the function of the wild-type copy. While these somatic second hits may be sufficient for stable disease states, it is often difficult to determine if their occurrence necessarily marks the initiation of disease progression. A more direct consequence of a heterozygous genetic background is haploinsufficiency, referring to a lack of sufficient gene function due to reduced wild-type gene copy number; however, haploinsufficiency can involve a variety of additional mechanisms, such as noise in gene expression or protein levels, injury and second hit mutations in other genes. In this study, we explore the possible contribution to the onset of autosomal dominant diseases from intrinsic factors, such as those determined by the structure of the molecular networks governing normal cellular physiology.

Results: First, simple models of single gene insufficiency using the positive feedback loops that may be derived from a three-component network were studied by computer simulation using Bionet software. The network structure is shown to affect the dynamics considerably; some networks are relatively stable even when large stochastic variations in are present, while others exhibit switch-like dynamics. In the latter cases, once the network switches over to the disease state it remains in that state permanently. Model pathways for two autosomal dominant diseases, AD polycystic kidney disease and mature onset diabetes of youth (MODY) were simulated and the results are compared to known disease characteristics.

Conclusions: By identifying the intrinsic mechanisms involved in the onset of AD diseases, it may be possible to better assess risk factors as well as lead to potential new drug targets. To illustrate the applicability of this study of pathway dynamics, we simulated the primary pathways involved in two autosomal dominant diseases, Polycystic Kidney Disease (PKD) and mature onset diabetes of youth (MODY). Simulations demonstrate that some of the primary disease characteristics are consistent with the positive feedback-stochastic variation theory presented here. This has implications for new drug targets to control these diseases by blocking the positive feedback loop in the relevant pathways.

Show MeSH
Related in: MedlinePlus