(190bn) Network Motif Properties Influence Transmission of Autosomal Allelic Imbalance to Phenotype Relevant Signals

Background: In a typical diploid cell, maternal and paternal alleles (copies) of a gene are often assumed to be expressed equally. This is true for many genes that are present on the non-sex chromosomes (autosomes). However, recent advances in genome-wide allele specific sequencing and targeted FISH (Fluorescence in-situ hybridization) measurements show that, in many cases, one allele may not be fully active and that this information is encoded epigenetically [1,2,3]. Interestingly, experiments show that this autosomal ‘monoallelic expression (MAE)’ is set early on in cellular differentiation processes and is stable under continuous mitotic divisions [1]. This indicates that maintenance of this imbalance may have important biological consequences. Inquiries into the establishment, causes and consequences of allelic imbalance and its role in disease phenotypes are areas of active research [4].

Objectives: An important question in the field is, When is allelic imbalance consequential? To answer this, one needs to look at the transmission of allelic imbalance to variables that affect the behavior of signaling networks which ultimately influences cellular homeostasis and state changes. Therefore, we did a systematic theoretical analysis of the influence of allelic imbalance on network topology, its signal-response relationship and functional behavior.

Brief Methods: We chose a set of network motifs commonly found in signaling networks. Then we modeled the dynamics of nodes (biochemical participants) in these motifs using differential equations and chose parameter sets that define a particular function and signal-response relationship. We defined biallelic (both alleles equally active) and monoallelic (one allele active) groups using expression ratio constraints. Allelic imbalance was introduced when setting the initial conditions for the individual nodes independently or in pairwise combinations. Simulations were performed by generating Latin Hypercube Samples of the free parameters defining allelic imbalance. Then we derived important signaling features from the simulated dynamics for the biallelic and monoallelic groups.

Key Results: One striking observation was that whether the monoallelic and biallelic groups exhibit different behaviours depends on the network motif and the signal-response behavior. Further, the presence of regulatory loops in a signaling motif may redistribute the effect of allelic imbalance even if it occurs at one node of the motif. In some cases, monoallelic and biallelic groups showed differences in the range of signal sensitivity, even though the range of expression levels were similar. This theoretical analysis shows that to assess its consequence, allelic imbalance needs to be considered in the context of the signaling motif in which it appears.

Future directions: Our analysis provides information to test above observations experimentally and better inform current genome-wide sequencing and targeted FISH experiments. In addition, our analysis will help future efforts to understand the relative contribution of allelic imbalance to cellular variability as opposed to stochastic sources [5].

References:

[1] Gimelbrant, A. et al. (2007) Science, 318(5853), pp 1136-1142

[2] Savova, V. et al. (2013) Curr Opin Genetics & Development, 23(6), pp 642-648

[3] Savol, A. et al. (2017) PLoS One, 12(8), e0182568

[4] Savova, V. et al. (2017) Mol Psychiatry 22, pp 1785-1794

[5] Swain, P. et al. (2002) PNAS 9(20), pp 12795-12800