Addressing an ongoing controversy: Do C1 neurons contribute to resting blood pressure?

Control of sympathetic vasomotor tone by catecholaminergic C1 neurones of the rostral ventrolateral medulla oblongata.

Marina N, Abdala AP, Korsak A, Simms AE, Allen AM, Paton JF, Gourine AV.

This study was designed to investigate the role of rostral ventral medullary catecholaminergic neurons in the control of resting blood pressure. We all know that most of supraspinal resting vasomotor tone comes from some neurons that "live" in the rostral portion of the ventrolateral medulla (RVLM). We also know that many neurons in RVLM are "C1" neurons, that is, they can synthesize catecholamines. Finally, we know that both C1 and non-C1 neurons send axons to sympathetic preganglionic neurons in the spinal cord, with most being C1. At the time of these discoveries it was assumed that C1 neurons were a major source of resting and relfex RVLM-mediated sympathetic activity. However, studies done in the past ten years have been able to investigate the function of C1 neurons by specifically targeting them with antibodies and viruses. Schreihofer and Guyenet showed that when about 75% of C1 neurons were ablated, sympathoexcitatory responses to direct or reflex-mediated stimulation were markedly reduced, but resting tone of SNA and MAP were unaffected. Further work from this lab showed that specific direct activation of C1 neurons produced increases in SNA and MAP. Also, work from another lab showed that specific "re-expression" of angiotensin II receptors in C1 neurons of an angiotensin receptor knockout mouse model restored SNA responses to angiotensin II in the RVLM. Together, these data suggest a prominent role for C1 neurons in SNS responses but doesn't fully address the role of C1 neurons in the maintanance of resting tone.

The very innovative and unique aspect of this study was the system the authors designed to test direct acute inhibition of C1 neurons in vivo. This was accomplished using lentiviruses specific for C1 neurons, but instead of the virus encoding for something that kills the cell, it encodes an inhibitory G protein-coupled receptor that has a specific ligand (allatostatin) which is not present in mammals. After RVLM injection of the virus and a 5 to 6 week recovery period, rats were anesthetized and instrumented to record MAP and renal SNA. They showed convincingly that acute specific inhibition of C1 neurons in vivo causes a reversible fall in MAP and RSNA in animals that recieved the virus but not in sham rats. They then tested the role of C1 neurons in the SNS response to hypercapnia, or high levels of arterial CO2. Interestingly, if C1 neurons were inhibited before experimentally varying CO2 levels, the response was the same as control. If C1 neurons were inhibited during high CO2, the peak response was attenuated. Also, they suggest that blockade of ionotropic glutamate receptors did not affect this phenomenon. I engourage anyone reading this post to look at the results of this second-to-last experiment and decide for yourself what the results suggest. Finally, they used an isolated heart-brainstem preparation to show that C1 neurons generate much of the respiratory-related SNA bursts.

These data contrast with previous work described. However, the authors note an important distinction in the discussion. By using a toxin that killed C1 neurons over the course of days, other vasomotor centers had the opportunity to increase their activity in compensation. Importantly, they showed in this study that the inhibition was immediate and reversible. Additionally, it appears (to me) that glutamatergic inputs on C1 neurons do have some role in the sympathetic response to hypercapnia, although I admit that the results from this study don't clearly suggest whether they do or do not. Finally, C1 neurons clearly are responsible for respiratory-related bursts of SNA in the authors' in situ  heart-brainstem prep. They could confirm these results in vivo by performing experiments similar to those from Ann Schreihofer's lab of the past 5 years or so.

C1 neurons in the rat rostral ventrolateral medulla differentially express vesicular monoamine transporter 2 in soma and axonal compartments

C. P. Sevigny, J. Bassi, A. G. Teschemacher, K. S. Kim, D. A. Williams, C. R. Anderson and A. M. Allen

European Journal of Neuroscience, Vol. 28 pp. 1536-1544, 2008

C1 neurons of the RVLM are defined as producing adrenaline, but the effect that adrenaline is used to is still not quite worked out; they are generally accepted as using glutamate as their major neurotransmitter.  To this end, they express a vesicular glutamate transmitter.  They also express a vesicular adrenaline transporter: VMAT2.  The authors of this paper looked at the differential expression of VMAT2 in the cell soma and axon terminals to determine if rostrally-located, bulbospinal C1 neurons release adrenaline somatodendritically (which requires cell body expression of VMAT2), synaptically, or both.

A lentivirus was used to identify the C1 neurons in one group of rats, rhodamine fluorescent microspheres were injected into the T3/T4 level of the spinal cord to identify spinally projecting neurons in a second group of rats, and anti-PMNT and anti-VMAT2 antibodies were used to identify PMNT (again, to identify C1 neurons) and VMAT2 in tissue sections.  Both brainstem sections and spinal cord sections were examined, to compare cell soma expression of VMAT2 in spinally projecting C1 nuerons to axonal expression in the same neurons.  Additionally, spinally projecting C1 neurons were compared to rostrally projecting C1 neurons.

Rostrally projecting C1 neurons did have VMAT2 present in their cell bodies, but bulbospinal C1 neurons, which were identified by the presence of the rhodamine fluorescent microspheres, rarely did.  VMAT2 was present in the axon varicosities in the spinal cord, though, suggesting that the bulbospinal C1 neurons do release adrenaline as a neurotransmitter.  While bulbospinal C1 neurons may not release adrenaline somatodendritically, though, some C1 neurons are thought to, and bulbospinal neurons seem to be excited to alpha-2-adrenoceptor blockade, suggesting that adrenaline may have a role in regulating the tonic activity of the RVLM.  In short, some C1 neurons, but not the bulbospinal C1 neurons, likely release adrenaline within the RVLM (which has an effect on tonic activity in the RVLM) while C1 bulbospinal neurons release adrenaline along with glutamate at their axon terminals.