Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2017 Jul 27;11:70.
doi: 10.3389/fncom.2017.00070. eCollection 2017.

Striatal Network Models of Huntington's Disease Dysfunction Phenotypes

Affiliations
Free PMC article

Striatal Network Models of Huntington's Disease Dysfunction Phenotypes

Pengsheng Zheng et al. Front Comput Neurosci. .
Free PMC article

Abstract

We present a network model of striatum, which generates "winnerless" dynamics typical for a network of sparse, unidirectionally connected inhibitory units. We observe that these dynamics, while interesting and a good match to normal striatal electrophysiological recordings, are fragile. Specifically, we find that randomly initialized networks often show dynamics more resembling "winner-take-all," and relate this "unhealthy" model activity to dysfunctional physiological and anatomical phenotypes in the striatum of Huntington's disease animal models. We report plasticity as a potent mechanism to refine randomly initialized networks and create a healthy winnerless dynamic in our model, and we explore perturbations to a healthy network, modeled on changes observed in Huntington's disease, such as neuron cell death and increased bidirectional connectivity. We report the effect of these perturbations on the conversion risk of the network to an unhealthy state. Finally we discuss the relationship between structural and functional phenotypes observed at the level of simulated network dynamics as a promising means to model disease progression in different patient populations.

Keywords: Huntington's disease; STDP; dynamics; homeostatic plasticity; medium spiny neuron; network; neurodegenerative; striatum.

Figures

Figure 1
Figure 1
An example of healthy winnerless network dynamic. Shown are 11 out of the total 50 FN neuron network simulated. As in striatum, the network shows episodic “bursts of bursts” which are organized in neuronal activations that alternate, and which include no silent neurons. The network, derived from a preliminary evolution of a randomly initialized, unidirectional set of connections between neurons, meets our criteria for health.
Figure 2
Figure 2
A healthy network shows winnerless activity (left) which abruptly changes when 1 neuron in a 50 neuron network is silenced, simulating neurodegeneration. Neurons which fire prolonged bursts are considered “unhealthy.” The network dynamics and trajectories shift to winner-take-all (right) in which some neurons fire continuously while others fall silent. In a simple three neuron motif, over-firing in unhealthy neurons inhibits downstream MSNs, producing silent “losers”, which fail in turn to provide proper inhibition to downstream neurons, thus producing a second winner. Multiple competitive interactions in this complex network makes the final outcome of the winner-take-all dynamics difficult to predict.
Figure 3
Figure 3
Probability of developing an unhealthy dynamic in at least one neuron in a 500 neuron network, as a function of fraction of neurons silenced (simulating neuron death in a real striatum). Network conversion risk is calculated using 20 simulations with different neurons silenced. For each data point the proportion of simulations in which an unhealthy neuron was observed is plotted. The red line represents a sigmoidal fit to the data, predicting a particular trajectory for disease progression.
Figure 4
Figure 4
A healthy network again shows winnerless activity (left) which abruptly changes when a bidirectional connection is added to 1% of already connected neurons. 10 out of the 500 neurons are plotted. Bidirectional connectivity is observed to increase dramatically among MSNs in Huntington's disease model animals.
Figure 5
Figure 5
Probability of developing unhealthy dynamic in at least one neuron in a 500 neuron network simulation, as a function of the fraction of already connected neurons to which a bidirectional connection is added. For each curve a different weight for the bidirectional connection is used. All other network weights were developed by a preliminary refinement phase of the simulation using iSTDP. Network conversion risk is calculated as in Figure 3.
Figure 6
Figure 6
Network conversion risk by neuron index constructed using simulations from Figure 5. Because all simulations start from the same developed and healthy network, statistics for individually identified neurons can be pooled across simulations. For each simulation in which a network was classified unhealthy by at least one neuron showing unhealthy dynamics, neurons causing this classification due to prolonged firing were tabulated. The table was then used to construct the histogram showing their participation rate in causing the unhealthy network dynamics.
Figure 7
Figure 7
Simulated of HD therapeutics. Network conversion risk from Figure 5, wij = 0.001, plotted again in blue. Repeating the same simulation under different conditions aimed at modeling different secondary perturbation to the network. For each secondary perturbations, amelioration or exacerbation of the network conversion risk is observed. For a detailed description of each secondary perturbation, see Section 2.4.1.
Figure 8
Figure 8
Topological changes in a 500 neuron network across time as a results of iSTDP. (A) Connection fraction decreases. (B) Bidirectional connection fraction decreases. (C) Loops of length 3 decrease in frequency.

Similar articles

See all similar articles

Cited by 2 articles

References

    1. Angulo-Garcia D., Berke J. D., Torcini A. (2016). Cell assembly dynamics of sparsely-connected inhibitory networks: a simple model for the collective activity of striatal projection neurons. PLoS Comput. Biol. 12:e1004778. 10.1371/journal.pcbi.1004778 - DOI - PMC - PubMed
    1. Berke J. D. (2011). Functional properties of striatal fast-spiking interneurons. Front. Syst. Neurosci. 5:45. 10.3389/fnsys.2011.00045 - DOI - PMC - PubMed
    1. Cao Y., Bartolomé-Martiń D., Rotem N., Rozas C., Dellal S. S., Chacon M. A., et al. . (2014). Rescue of homeostatic regulation of striatal excitability and locomotor activity in a mouse model of Huntington's Disease. Proc. Natl. Acad. Sci. U.S.A. 112, 2239–2244. 10.1073/pnas.1405748112 - DOI - PMC - PubMed
    1. Cepeda C., Galvan L., Holley S. M., Rao S. P., Andre V. M., Botelho E. P., et al. . (2013). Multiple sources of striatal inhibition are differentially affected in Huntington's Disease mouse models. J. Neurosci. 33, 7393–7406. 10.1523/JNEUROSCI.2137-12.2013 - DOI - PMC - PubMed
    1. Cepeda C., Wu N., Andre V. M., Cummings D. M., Levine M. S. (2007). The corticostriatal pathway in Huntington's Disease. Prog. Neurobiol. 81, 253–271. 10.1016/j.pneurobio.2006.11.001 - DOI - PMC - PubMed

LinkOut - more resources

Feedback