CARDIAC ELECTRICAL DYSFUNCTION AND THE LITTLE BRAIN ON THE HEART

J.A.Armour
Université de Montréal, Montréal, Canada

1. Statement of the issue. Central neuronal inputs that regulate cardiac output in response to arterial baroreceptor sensory information, for instance, are usually modeled as external inputs to the intrathoracic nervous system.  Major advances in our knowledge about control of cardiac function paints a richer picture of a hierarchy of computational networks existing within the thorax, operating under the influence of central neuronal (primarily spinal cord/ brainstem), for cardiac control.  Recently, attention has been paid to intrathoracic neuronal mechanisms involved in such control that includes a counterintuitive little brain on the heart - the intrinsic cardiac nervous system.  The latter, residing as it does on the target organ, can operate independently of more centrally located neurons if necessary. 

2.  Cardiac afferent neurons.  In order to understand the complex interactions that occur within each ‘level’ of the cardiac neuroaxis, one must understand the individual cardiovascular sensory inputs that each level receives.  Most cardiac sensory neurites transduce their local chemical milieu as noisy, asynchronous inputs to individual afferent neurons involved in cardiovascular regulation.  The somata of cardiac afferent neurons that generate noisy, asynchronous activity are typically located more centrally, that is adjacent to central neuronal computational centers.  On the other hand, cardiac afferent neurons that generate phasic activity synchronous with cardiodynamics are most abundant in intrathoracic ganglia on or adjacent to the heart.  Individual sensory neurites associated with each cardiac sensory neuron perform the fundamental job of transducing the continuously varying local mechanical and/or chemical milieu to result in the generation of activity patterns that relate to time varied local milieu changes and the spatial distribution of such changes depending on the location of their sensory neurites. 

3. Cardiac neuronal feedback.  The structure of the nervous system that maintains cardiac output can be divided into a hierarchy comprising three main ‘levels’ within the cardiac neuroaxis: 1) the intrinsic cardiac nervous system; 2) the intra-thoracic extracardiac nervous system; and 3) the central nervous systems involving primarily brain stem and spinal cord neurons.  While each level exchanges information continuously to influence one another, they are not redundant in that each takes on the lion’s share of a specific responsibility linked to the type of afferent feedback they receive and the time over which their control is exerted. 

Populations of excitatory and inhibitory neural networks exist in each level of this neuronal hierarchy, those in intrathoracic ganglia being called local circuit neurons.  Local circuit neurons display memory of past events (sensory inputs), extending information processing by an order of magnitude that can last for minutes.  This population evinces activity patterns that shift in a ‘hysteretic’ fashion to new states dependent upon their cardiovascular sensory inputs.  Such memory is useful to reduce multiple, noisy inputs from cardiac sensory neurons to feed-forward information at activity levels appropriate for cardiac efferent neuronal control. 

The transduction properties of cardiac sensory neurites fall roughly into two broad categories: i) fast and ii) slow responding neurons that generate phasic or tonic activity patterns, respectively.  This categorization leads directly to an ansatz on which the functional organization of the cardiac nervous system is quite dependent: i) the physical distance between the somata of cardiac afferent neurons and their sensory neurites and ii) the type of information they transduce.  Intrathoracic neurons are involved primarily in short latency reflexes (as short as 40 msec) that initiate fast-responding (feed-forward) inputs to cardiac efferent neurons.  On the other hand, slow responding afferent neurons generate noisy activity patterns that bear little or no observable relationship to regional cardiac dynamics to provide control on timescales they are suited to the regulation of cardiac output over timescales of several minutes.  Many affect higher computational centers that, in turn, regulate cardiac efferent neurons.  Why are inputs to intrathoracic neurons from higher computational centers typically rerouted through intrathoracic local circuit neurons before reaching the heart?  The little brain on the heart is comprised of redundantly connected populations of excitatory and inhibitory local circuit neurons.  A key feature of any hysteretic population of neurons is their ability to display memory.  Memory stems from the hysteretic nature of sensory inputs transducing cardiac events.  Hysteretic systems, displaying activity patterns dependent upon excitatory and inhibitory neurons, are stable over a wide input range.  Any shift to a new state reflects altered sensory inputs exceeding or going below critical values, providing: i) memory and ii) noise filtering.

4. Cause and effect.  A hysteretic system is thresholded and, as such, when it goes to a new state it stays there until the inputs exceed a threshold.  Dropping back to the old state occurs only when its inputs are reduced to a level much below that which previously caused a shift to the new state.  Superimposed noisy variations can evoke a response to subthreshold inputs that would otherwise cause no response.  The reason is that noisy milieu variations modify the average sensory inputs over a timescale within which a state shift can occur such that there are no longer so-called ‘critical’ levels within which such state shifts occur.  There is a twofold price to pay for such reliance.  It is possible to cycle aimlessly between states when noise transduction become too high.  During excessive memory derived from high inputs, it is easier to enter an excited state than to leave it.  Such excessive memory can be pathological since its forcing of cardiac efferent neurons would not only become excessive, but also difficult to halt.  When the intrinsic cardiac nervous system receives abnormally high inputs, cardiac arrhythmias or even ventricular fibrillation ensue, even in normally perfused hearts.  On the other hand, suppression of intrinsic cardiac neuronal function may act to stabilize cardiac function.  How this complex system does so remains elusive.

Supported by the CIHR and NSHSF.