Respiration and the Viscera: Brainstem Rhythm Generators and Visceral Control

01The rhythm core

Breathing is generated in the medulla by a cell population small enough to count

Respiration begins in the medulla oblongata, and the rhythm itself is made by one small region inside the ventral respiratory group. Microsection of the neonatal rat brainstem in vitro identified that region, the pre-Bötzinger complex, and removing it alone eliminated rhythm generation Smith 1991. Slices that keep the region keep oscillating with no lung, no chest wall and no cortex attached.

Output fades gradually and the rhythm stops abruptly

Cumulative ablation of Dbx1-derived pre-Bötzinger neurons in spontaneously rhythmic mouse brainstem slices cut hypoglossal motor output by about 50 percent after roughly 15 cell deletions. The rhythm then decelerated and terminated as the tally reached about 85 neurons Song 2016. Deleting the same class of neuron one stage downstream, in the hypoglossal premotor population of those slices, lowered output steadily and never altered frequency or stopped the output at any tally. Size is held one stage below timing.

A triple oscillator hypothesis assigns inspiration, post-inspiration and active expiration to anatomically distinct networks

Inspiration is generated at the pre-Bötzinger complex, and active expiration at the retrotrapezoid nucleus and parafacial respiratory group Feldman 2013. A triple oscillator account extends that division. It assigns inspiration, post-inspiration and active expiration to three anatomically distinct networks, naming the pre-Bötzinger complex, the post-inspiratory complex and the lateral parafacial nucleus in that order Anderson 2017. The authors put it forward as a hypothesis.

The same lesion reads differently in sleep and in waking

Unilateral ablation of pre-Bötzinger NK1R neurons in adult rats left resting waking breath stable and broke breathing in sleep. Central apneas and hypopneas ran at 36.8 episodes per hour of REM sleep, against 6 per hour before injection McKay 2008. Bilateral ablation of those same cells runs a fixed order, disordering sleep first and ending in an ataxic waking pattern. The unilateral rats stopped at the first step and stayed there through 51 days of recording. Waking supplies drive that a half-strength oscillator can borrow, and REM sleep takes the loan back.

02Findings

What the research shows

03Pontine phrasing

The pons ends inhalation and sets upper airway resistance in one act

The pons modulates the raw medullary rhythm into smooth, adaptable breathing through two paired centers. The upper pons carries the pneumotaxic center, which coordinates the speed of breathing and sends inhibitory impulses that close inspiration and quicken the cadence. The lower pons carries the apneustic center, which coordinates depth and sends stimulatory impulses that prolong inhalation. The pneumotaxic influence overrides it to close the cycle. Together the two centers grade the intensity and frequency of the medullary signal, so the transition between inspiration and expiration arrives smoothly instead of as an on and off switch.

The Kölliker-Fuse nucleus of the dorsolateral pons gates the post-inspiratory phase and by that route ends inspiration. Lesion or pharmacological manipulation of the dorsolateral pons in the rat transforms the pattern into apneusis, a pathological prolonged inspiration produced by a delayed transition from inspiration to expiration Dutschmann 2006. Glutamate microinjected into the Kölliker-Fuse of a perfused rat brainstem produced sustained laryngeal constriction and a longer expiration. A GABA agonist at the same loci abolished post-inspiratory motor activity, triggered apneusis and left the rate variable and slow. The newborn airway shows the same pairing of pattern and caliber.

Phrasing and airway caliber travel together

The Kölliker-Fuse nucleus grades laryngeal resistance in the same act that ends inhalation Dutschmann 2006. A breath that will not end is also a breath meeting a different airway. Post-inspiration, the second of the three respiratory phases Anderson 2017, is where expiratory flow is braked and where speech, swallow and airway protection are built. Apneusis is the reading when that phase fails to arrive, and it shows in the shape of the breath before it shows in the volume. The Kölliker-Fuse also holds large numbers of laryngeal post-inspiratory premotor neurons, so the nucleus that closes an inhalation is also recruited for vocalization.

04Oxygen and carbon dioxide sensors

Two sensor populations set the drive to breathe, one in the neck and one in the medulla

The carotid and aortic bodies read arterial oxygen and answer hypercapnia and acidosis alongside it. A cluster in the rostral medulla reads carbon dioxide through the chemistry of the cerebrospinal fluid. Surgery has separated the two channels in people. Bilateral carotid body tumor resection abolished the normocapnic ventilatory response to hypoxia in all 8 patients studied, with a small residual response under hypercapnic conditions in 2 of them Timmers 2003. Carbon dioxide driven breathing persisted.

The same operation cost those patients a second function. Baroreflex sensitivity averaged 6.4 ms/mmHg against 14.7 in 12 controls, and the authors describe that failure as heterogeneous across the group. One organ in the neck reports for the breath and for the pressure, and the pressure loop is worked through on autonomic regulation.

The carbon dioxide sensor has a name and a gene

The retrotrapezoid nucleus is a rostral medullary cluster of glutamatergic neurons identified by coexpression of Phox2b and Nmb transcripts. Those cells read brain PCO2 as hydrogen ion concentration through the proton sensors TASK-2 and GPR4, with an astrocytic and vascular contribution alongside Guyenet 2022. PHOX2B mutations impair the development of the nucleus and cause congenital central hypoventilation syndrome. Silencing it contributes to periodic breathing and to central sleep apnea, a failure of the rhythm rather than the airway collapse described on sleep apnea and the nervous system.

Losing the sensor does not lose the defended value

Conditional Phox2b mutant mice lack a retrotrapezoid nucleus, and their breathing is not stimulated by elevated carbon dioxide at least to postnatal day 9. They barely respond as juveniles. They survive, breathe normally beyond the first days after birth, and hold blood PCO2 within the normal range, with chemosensitivity recovering in part by adulthood Ramanantsoa 2011. The carotid bodies carry the defended PCO2 while the central sensor is absent. In this model the set point sits in the arrangement, and the retrotrapezoid nucleus is its fastest route to the value.

05Three vagal receptor families

Lung afferents rewrite the timing of the breath before the next one begins

Vagal lung afferents divide into three families, and their central terminations occupy largely nonoverlapping regions of the caudal half of the nucleus tractus solitarius Kubin 2006. They are the slowly adapting stretch receptors, the rapidly adapting stretch receptors and the bronchopulmonary C fibers, and the mapping of their terminal fields comes from tracing work in animals.

Pump cells of the nucleus tractus solitarius mediate the Breuer-Hering reflex and project onward to ventrolateral medullary premotor and rhythm-generating neurons and to the pontine parabrachial region Kubin 2006. Airway mechanoreceptors through the trachea, lungs and pulmonary vessels feed inflation into that pathway, and metaboreceptors in skeletal muscle add a report on the cost of the work being done.

Only the timing half of the reflex hides from awareness

Pseudorandom unloading in 17 awake normal subjects showed that vagal feedback modulates respiratory timing on a breath-by-breath basis, as long as there is no awareness of the stimulus BuSha 2001. Stimuli above and below the perceptual threshold both raised tidal volume and inspiratory flow rate. Only the subthreshold ones shortened inspiratory time. Volume answered either way, and timing was the part awareness took. Breathing is the one visceral output where volition and reflex compete for a single motor pool.

Chest compression lengthens inspiratory time in term infants and shortens it in preterm infants

Chest compression below functional residual capacity triggers the deflation reflex, and the answer depends on maturity. Term infants prolonged inspiratory time by 15 percent while preterm infants shortened it by 24 percent. Expiratory time shortened by 35 percent in the term infants against 18 percent in the preterm infants Hannam 1998. Fifteen of 20 preterm infants tended to short apneas of 2 to 5 seconds under the same hands, 88 percent of them central. The infants who paused were the ones whose inspiratory time shortened most. The newborn version of this circuitry is taught in the first breath.

Defensive afferents become the symptoms of airway disease

Bronchopulmonary nociceptors trigger defensive reflexes against inhaled irritants and are implicated in asthma. Inflammatory mediators activate these vagal C fibers and produce coughing, chest tightness, changed breathing pattern and reflex bronchoconstriction Undem 2020. Those responses defend a healthy lung. In inflamed tissue they contribute to the signs and symptoms of airway disease instead. The afferent has not changed. The tissue state reading it has.

06The bladder switch

The pons that phrases the breath also holds the switch between storing and voiding urine

Barrington's nucleus, the pontine micturition center in the dorsomedial pontine tegmentum, directly excites bladder motoneurons and indirectly inhibits urethral sphincter motoneurons through inhibitory interneurons in the medial sacral cord Blok 1998. A ventrolateral pontine L-region holds continence. The periaqueductal gray receives bladder filling information and switches the system between storage and voiding.

Three peripheral nerves carry the result. The sympathetic hypogastric nerve stimulates urethral smooth muscle for storage. The parasympathetic pelvic nerve stimulates the detrusor and relaxes the urethra for voiding, and the somatic pudendal nerve drives the external urethral sphincter Loufopoulos 2026. The supraspinal control areas were mapped first in the cat, and human imaging has since found activity in the same areas Blok 1998.

The sphincter column and the bladder column are continuous tissue

Human morphometry places Onuf's cell group X between the middle of S2 and the rostral one third of S3, a slender longitudinal column in the ventral horn Konno 1986. The sacral parasympathetic intermediolateral nucleus sits chiefly at S3 and S4, with a cellular bridge running between the two. Pudendal outflow to the external sphincter is classically given as S2 to S4. The somatic column that holds the sphincter shut and the autonomic column that empties the bladder are neighbors in one stretch of gray matter.

The voiding command has been triggered directly

Optogenetic stimulation of glutamatergic Barrington's neurons in mice triggers voiding, while stimulating the Crh-expressing subpopulation causes bladder contraction that usually fails to empty the bladder Verstegen 2019. Ablating the glutamatergic population produces profound urinary retention, while deleting Crh from those same neurons changes nothing. Excitatory drive arrives from the midbrain periaqueductal gray and the lateral hypothalamus, the same two structures that human filling scans separate into volume and urge.

Filling and urge are two different reports

Eleven healthy men were scanned by PET at six bladder volumes, twelve scans each. Activity rose with bladder volume in the periaqueductal gray, the midline pons, the mid-cingulate cortex and both frontal lobes Athwal 2001. Falling urge to void tracked a different portion of the cingulate cortex, the premotor cortex and the hypothalamus. Volume and urge are read by separate networks in the same skull. The interior is regulated at the brainstem while awareness receives an edited summary.

The bladder reports its filling the way the lung does

The urothelium transduces stretch through ENaC, Piezo1, Piezo2 and TRPV4 channels, then releases ATP, nitric oxide and acetylcholine onto adjacent afferents Loufopoulos 2026. Two hollow organs with different mechanical jobs report their own filling through their own epithelium into the same converging brainstem tissue.

07The central autonomic network

One integrated network controls the breath, the vessel and the sphincter together

Respiration is the integrative readout of the brainstem's autonomic command. The central autonomic network has seven named components. Its forebrain members are the insular cortex, the amygdala and the hypothalamus. Its brainstem members are the periaqueductal gray matter, the parabrachial complex, the nucleus of the tractus solitarius and the ventrolateral medulla. The network controls preganglionic sympathetic and parasympathetic, neuroendocrine, respiratory and sphincter motoneurons together, and its activity is state dependent Benarroch 1993. A single medullary lesion can produce orthostatic hypotension, paroxysmal hypertension and sleep apnea in one patient. The convergence is visible in the failure.

The overlap reaches the level of cell populations. Unilateral saporin ablation in the rat removed 97 percent of NK1R-immunoreactive cells within the pre-Bötzinger complex and rostral ventral respiratory group, sparing catecholaminergic neurons. That ablation cut both the depressor and the tachypneic responses to chemical stimulation of the ventral respiratory group Wang 2002. The authors state that no evidence yet shows the two functions belonging to the same neurons. Rhythm and pressure share the tissue, and the claim stops there.

Sympathetic outflow is gated inside every breath

Microneurography during graded leg cycling in 13 adults recorded muscle sympathetic burst frequency at 20.0 bursts per minute at rest. It held there at 40 percent of maximal workload, then rose to 31.6 at 60 percent and 44.7 at 80 percent. At baseline and at every exercise intensity, bursts were fewest at end-inspiration and most frequent from mid to end expiration Katayama 2021. The within-breath swing widened from rest to 60 percent of maximum and widened no further above it. Vascular drive is phase locked to the respiratory cycle.

One blockade moves the breath and the pressure together

Lumbar intrathecal fentanyl blocked group III and IV muscle afferents in five men during rhythmic exercise. Ventilation fell, with VE/VCO2 down 8 to 17 percent through a reduced breathing frequency, and end-tidal PCO2 rose 4 to 7 Torr. Mean arterial pressure fell 8 to 13 percent and heart rate 2 to 8 percent Amann 2010. One afferent stream was silenced and both outputs moved, in awake people.

08Segmental input and the oscillator

Afferent traffic entering a thoracic segment reaches the respiratory oscillator and can reset it

Stimulating the thoracic sympathetic chain in a perfused juvenile rat preparation lengthened expiration, prolonged postinspiratory vagal discharge and modulated late-expiratory abdominal activity Dhingra 2019. Rhythmic 400 ms trains delivered near the intrinsic respiratory rate entrained the central respiratory rhythm outright. The traffic entered at a thoracic segment, nowhere near the medulla.

High thoracic spinal transection abolished the evoked respiratory modulation in the same preparation, which places the relay for these afferents in the thoracic spinal cord Dhingra 2019. The modulation habituates during sustained stimulation, so the oscillator answers a changing input and ignores a constant one.

Slow breathing moves one autonomic arm and not the other

Fifteen minutes of device-guided breathing at 5 breaths per minute lowered muscle sympathetic burst frequency by 7.8 bursts per minute in prehypertensive veterans with post-traumatic stress disorder. Thirteen breathed slowly against 10 paced by a sham device Fonkoue 2018. Sympathetic baroreflex sensitivity roughly doubled, from -1.08 to -2.29 bursts per 100 heart beats per mmHg. Heart rate and cardiovagal baroreflex sensitivity did not move. Driving the breath deliberately reached the sympathetic arm and left the cardiac vagal arm where it was. What paced breathing does to arterial pressure, and for how long, is set out on blood pressure and the nervous system.

A matched segmental input changes the timing of the breath before it changes its size

Chiropractic care works at thoracic segments, the entry point the perfused rat preparation used to reach the oscillator. The Unified Model of Tone takes the entrainment result as its prediction. What decided the outcome in that preparation was when the train landed in the respiratory cycle and how near its interval sat to the intrinsic rate. The trains that entrained the rhythm ended up aligned with late expiration. So a segmental input in a person should register first as a longer expiration and a later inspiratory off-switch, with tidal volume moving last. A heavier thrust delivered at mid-inspiration should do less to the cycle than a lighter one delivered into late expiration.

Breathing that can lengthen, brake at the top, hand itself to volition and take itself back is holding its range, and that freedom to move and return is the health. Breathing pinned at one setting has left the range. It presents as apneusis when the off-switch fails to arrive, as periodic breathing when the carbon dioxide sensor loses its grip, and as central apnea when a half-strength oscillator meets REM sleep.

09Tone

How this system expresses tone

Tone in the respiratory system is read as phase timing. Every input the breath answers lands somewhere in the cycle, and where it lands decides whether the rhythm shifts, holds or entrains. Oscillation, coupling and input quality carry the signature here.

Each of the other foundations has a value somewhere in the respiratory cycle. Gain: removing both carotid bodies erased the normocapnic hypoxic ventilatory response in all 8 patients, and a small response to hypoxia survived under hypercapnia in 2 of them. One sensor channel sets the size of one answer. Prediction: blocking group III and IV muscle afferents dropped the ventilatory equivalent for carbon dioxide, VE/VCO2, by 8 to 17 percent during exercise. End-tidal PCO2 rose 4 to 7 Torr, because the breath had been set from the work rather than from the blood. Constraint: quiet exhalation has no motor act available and runs on elastic recoil, so forced expiration is driven by a separate oscillator at the lateral parafacial nucleus. Load: heavy work more than doubles sympathetic burst frequency, from 20.0 to 44.7 bursts per minute, and the within-breath swing stops widening above 60 percent of maximum work rate. Time course: vagal timing corrections land inside one breath, slow breathing moves sympathetic traffic in 15 minutes, and carbon dioxide sensitivity in mutant mice recovers only in part, and only by adulthood. Set point: conditional Phox2b mutant mice barely answer carbon dioxide as juveniles and still hold arterial PCO2 in the normal range, so the value is defended by the arrangement.

10Across the library

How this page relates to the rest of the library

11Frequently asked

Questions about this topic

12The sources

References

Sources: primary literature, linked inline.