Bursting Pacemaker Neurons Based on Models of Respiratory

Bursting Pacemaker Neurons Based on: Models of Respiratory Rhythm Generation in the Pre-Botzinger Complex. I. Bursting Pacemaker Neurons Robert. J. Butera, John Rinzel, Jeffery C. Smith

Introduction Pre-Bötzinger complex is the hypothesized site for respiratory rhythm generation p It is housed in the rostral ventrolateral medulla p Part B, Photo labeling the ventrolateral medulla (pre-Bötzinger complex area approximated by dashed line).

Pacemaker Neurons p p Pre-Bötzinger complex houses the pacemaker neurons It is hypothesized that contribution of both a pacemaker-based kernel and a pattern-formatting network driven by the kernel is responsible for the respiratory rhythm generation (Hybrid model). Some pacemaker neurons receive tonic excitatory inputs (from the mundane neurons) necessary to bring the membrane potential into the voltage window where bursting occurs. These neurons are classified as conditional bursting pacemakers.

Background of the Research Paper In earlier models respiratory rhythm generation was postulated to arise from network interactions, specifically inhibitory connections. But in such models the rythmicity ceased when synaptic inhibition was blocked. p In the hybrid model, for which this paper is a segway, inhibitory interactions are not essential, mimics the actual in vitro and en bloc experiment results. p The objective of this paper is modeling the rhythm and inspiratory burst generation in the kernel operating in vitro. p

Model Development Two models have been proposed for neurons responsible for rhythm and inspiratory burst generation in vitro. Model 1 Model 2 • Based on one-compartment Hodgkin-Huxley model. • Bursting occurs by virtue of fast activation and slow activation of a persistent Sodium current INa-P • Based on model 1. • Bursting occurs by virtue of fast-activating persistent Sodium current INa-P (inactivation term “h” removed) and slow activation of Potassium current IKs

Model 1 It is composed of five ionic currents across the plasma membrane: a fast sodium current, INa; a delayed rectifier potassium current, IK; a persistent sodium current, INa. P; a passive leakage current, IL; and a tonic current, Itonic_e (although this last current is considered to be inactive in these models)

Model 1 - Formulation Where, x∞Є {m. P, m, h, n} and x Є {h, n}

Model 2 The second model is identical to the model 1 except with the addition of a slow K+ current, IKS. (The removal of the inactivation term "h" from INa. P is not visible in the model diagram. )

Model 2 - Formulation Where, x∞Є {m. P, m, k, n} and x Є {k, n}

How does model 1 work? Dynamic response of model 1 as a function of EL: membrane potential EL=-60 m. V A closer look…

(in) activation Nullclines - m∞ 3 , n∞ 4, h∞ V (m. V)

EL=-57. 5 m. V EL=-54 m. V

Results from research article.

Bursting to tonic spiking EL=-60 m. V EL=-57. 5 m. V EL=-54 m. V

Model 1 - Animation The two kinds of currents in this model are: 1) Spike generating currents - INa & IK 2) Sub threshold currents (INa. P+ IL called Isub). The bursting cycle can be understood like this: • When g. Na. P increases beyond a critical value, Isub is large enough to initiate a burst. • The firing of action potentials gradually inactivates h (slow variable) • The bursting terminates when INa. P is inactivated sufficiently and the cell hyperpolarizes. • Now h gradually de-inactivates increasing Isub, to trigger another burst and so on… Isub vs Time

Bifurcation Mechanism EL=-65 m. V Silent EL=-58 m. V Bursting SN bif. - saddle-node bifurcation HC bif - saddle-homoclinic bifurcation sub. H - subcritical Andronov-Hopf bifurcation. EL=-55 m. V Beating

Model 2 -Results EL=-59. 5 m. V EL=-50 m. V EL=-40 m. V

Results from research article

Model 2 -working p Model 2 operates in a very similar fashion as Model 1, the difference being the slow activation persistent sodium current INa. P is replaces by a slow activation of potassium current IKS

Difference between Models 1 & 2 1 2 p p p Burst initiated by INa. P, terminated by inactivating INa. P Membrane conductance gm increases through the silent phase The membrane potential remains flat during the inter burst interval Burst duration decreases with depolarization Supports bursting over a small range of EL (-60 to -54 m. V) p p p Burst initialed by INa. P, terminated by activating IKS sufficiently Membrane conductance gm decreases through the silent phase The membrane potential interval not as flat during the inter burst interval Burst duration does not decrease with depolarization Supports bursting over double the range of EL as model 1 (-59. 5 to -40 m. V)

Both models go through a regime of silence bursting and beating In both models a minimum value of g. Na. P is required to support bursting. If g. Na. P is too low, only quiescence , or for higher values beating are supported.

Miscellaneous Comments p The same effects of chaging EL can be obtained by fixing EL and varying the parameter gtonic (Itonic) or Iapp ( g. L (v-EL) in model 2). Some of these results shown below.

Summary p p Model 1 is found to be more consistent with experimental data. The relative flat interburst interval is due to the fact that the subthreshold currents are all balanced and add up to zero. These are minimal models that provide a believable explanation for generating multistate, voltage-dependent behavior observed in the Pre-Botzinger pacemaker neurons. Although the actual burst generating currents still need to be unidentified in the Pre-Botzinger neurons

Questions?

Happy Holidays

Neuron Xmas tree!!

References p RT-PCR reveals muscarinic acetylcholine receptor m. RNA in the pre-Bötzinger complex, Jiunu Lai*, Xuesi M. Shao*, Richard W. Pan, Edward Dy, Cindy H. Huang, and Jack L. Feldman p p Models of Respiratory Rhythm Generation in the Pre-Bo¨tzinger Complex. I. Bursting Pacemaker Neurons, ROBERT J. BUTERA, JR. , 1, 2 JOHN RINZEL, 1– 3 AND JEFFREY C. SMITH 1 The Dynamic Range of Bursting in a Model Respiratory Pacemaker Network , Janet Best, Alla Borisyuk, Jonathan Rubin, David Terman, Martin Wechselberger p All simulations performed using Matlab 7. 0 , with a ode 15 s solver and absolute and relative tolerance of 10 -6.