Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
Review
, 14 (1), 53-67

TNFalpha: A Trigger of Autonomic Dysfunction

Affiliations
Review

TNFalpha: A Trigger of Autonomic Dysfunction

Gerlinda E Hermann et al. Neuroscientist.

Abstract

During disease, infection, or trauma, the cytokine tumor necrosis factor alpha (TNF alpha) causes fever, fatigue, malaise, allodynia, anorexia, gastric stasis associated with nausea, and emesis via interactions with the central nervous system. Our studies have focused on how TNF alpha produces a profound gastric stasis by acting on vago-vagal reflex circuits in the brainstem. Sensory elements of this circuit (i.e., nucleus of the solitary tract [NST] and area postrema) are activated by TNF alpha. In response, the efferent elements (i.e., dorsal motor neurons of the vagus) cause gastroinhibition via their action on the gastric enteric plexus. We find that TNF alpha presynaptically modulates the release of glutamate from primary vagal afferents to the NST and can amplify vagal afferent responsiveness by sensitizing presynaptic intracellular calcium-release mechanisms. The constitutive presence of TNF alpha receptors on these afferents and their ability to amplify afferent signals may explain how TNF alpha can completely disrupt autonomic control of the gut.

Figures

Fig. 1
Fig. 1
Components of gastric vago-vagal reflexes. Top: Coronal histological section through caudal hindbrain. Components of the dorsal vagal complex have been high-lighted by the uptake of horseradish peroxidase via the vagus nerve. Bottom: Schematic describing essential details of vago-vagal gastric motility control reflexes. 1) Vagal mechanosensors from the esophagus, antrum, and duodenum are activated by contact with food. These afferents enter the brainstem and collect as the solitary tract. Vagal afferents terminate on neurons in the nucleus of the solitary tract (NST). 2) Neurons in the NST are activated by vagal afferents that release glutamate onto postsynaptic NST neurons. Different phenotypes of NST neurons direct different aspects of the efferent vagal response. 3) For example, inhibitory NST neurons (RED) use GABA or NE to inhibit tonically active and excitatory DMN neurons that normally act to increase gastric tone and motility. 4) Other excitatory NST neurons (GREEN: probably glutamatergic) activate an inhibitory (NANC) vagal efferent pathway to the stomach. 5) The net result is a vagally mediated inhibition of gastric motility. The esophageal-gastric relaxation reflex (Rogers and others 2005), for example, uses this essential circuitry. 6) This reflex circuitry is contained in a circumventricular region of the brainstem out-side the blood-brain barrier. Considerable evidence suggests that neural elements in this vago-vagal reflex network can be dramatically modulated by hormones, cytokines, and other chemical agents in the circulation that arrives via fenestrated capillaries (BLUE ARROWS). Not shown are the multitude of neural inputs from the brainstem and hypothalamus that adjust vagal reflex functions appropriate to changes in behavior and internal state, as well as parallel ascending projections from the NST associated with the control of feeding behavior (Berthoud 2002). Also not shown is the parallel reflex path connecting gastric distension with a vagally mediated increase in gastric acid output (Rogers and others 2005). AP = area postrema; DMN or DMV = dorsal motor nucleus of the vagus; NA = nucleus ambiguus; NST = nucleus of the solitary tract; ST = solitary tract of vagal fibers.
Fig. 2
Fig. 2
Polygraph records of strain gauge activity monitoring gastric motility: Effects of TNFα. Basal gastric motility and tone of a food-deprived and anesthetized animal are minimal. Therefore, gastric motility was maximally stimulated by application of the peptide thyrotropin-releasing hormone (TRH; i.e., vagal, cholinergic stimulation) directly onto the exposed brainstem (TRH-icv). (A) Top trace: When stimulated gastric motility has plateaued (~ 5 to 10 minutes after TRH application), unilateral microinjection of 20nl of phosphate buffered saline (PBS; vehicle) was made into the left dorsal vagal complex (DVC). No change in motility was observed. Bottom trace: When the same volume of TNFα (0.02 femtomoles) was microinjected into the DVC, gastric motility was rapidly suppressed for prolonged periods of time (in excess of two hours in this example). (B) Central TNFα effects on gastric motility depend on intact vagal pathways. Polygraph tracing of gastric motility in a unilaterally vagotomized animal: Minimal basal gastric motility is maximally stimulated by application of TRH onto the brainstem surface (TRH-icv). Unilateral microinjection of the highest dose of TNFα (20 femtomoles) into the DVC of the vagotomized side is no longer effective in suppressing gastric motility. Adapted with permission from Hermann and Rogers (1995).
Fig. 3
Fig. 3
Endotoxin challenge provokes production and release of proinflammatory cytokines including TNFα into the systemic circulation. Pretreatment with the TNF-adsorbant construct (TNFR:Fc) directly within the dorsal vagal complex blocked cFOS-activation (right panel) of NST cells in response to systemic lipopolysaccharide (LPS) exposure. In contrast, pretreatment with only the Fc fragment did not prevent cFOS-induction (left panel). The middle panel outlines structures of interest in adjacent panels. AP = area postrema; DMN = dorsal motor nucleus of the vagus; NST = nucleus of the solitary tract; ST = solitary tract. Scale bar = 400 micron. Adapted with permission from Hermann and others (2003).
Fig. 4
Fig. 4
Raw polygraph records of strain gauge activity monitoring gastric motility. Both records were obtained at ~90 minutes after systemic (IV) administration of lipopolysaccharide (LPS) i.e., circulating proinflammatory cytokine levels are elevated. (A) This animal had received LPS (IV) and continuous ventricular perfusion (intracerebroventricular [ICV]) of the Fc fragment vehicle control. ICV application of thyrotropin-releasing hormone (TRH) is unable to override the suppressive effects of the endogenously produced TNFα. (B) This animal received LPS (IV) and continuous ventricular perfusion of the TNFR:Fc adsorbant construct (ICV). Presumably, the endogenously produced TNFα is adsorbed and effectively neutralized as it passes through the ventricular flow. In this case, application of TRH to the floor of the fourth ventricle is able to maximally stimulate gastric motility. Adapted with permission from Hermann and others (2002).
Fig. 5
Fig. 5
Photomicrographs of coronal histological sections through the NST immunohistochemically processed for cFOS-activation. Immunoreactivity for cFOS-activation is characterized by darkly stained nuclei. (A) control injection of PBS (20nL) directly into the DVC results in few cFOS-(+) cells in the NST and DMN; (B) injection of TNFα (0.2 femtomoles) induces a significant increase in cFOS expression; (C) co-injection of the same amount of TNFα with the AMPA glutamate receptor antagonist, NBQX, suppresses cFOS production to control levels; (D) co-injection of TNFα with the NMDA antagonist, MK-801, also reduces cFOS activation. These data support the hypothesis that TNFα may excite NST neurons by increasing presynaptic (i.e., vagal afferents) glutamate neurotransmission. AP = area postrema; DMN = dorsal motor nucleus of the vagus; NST = nucleus of the solitary tract; ST = solitary tract. Adapted with permission from Emch and others (2001).
Fig. 6
Fig. 6
TNFα activation of adrenergic cells in the nucleus of the solitary tract (NST). Preliminary data suggest that TNFα may specifically cause the activation of adrenergic neurons in the NST. This figure shows that microinjection of TNFα into the medial NST provokes cFOS activation (seen as brown nuclear staining) of virtually every adrenergic neuron (tyrosine hydroxylase [Th]–positive immunore-activity is seen as black cytoplasmic staining) in the area. These neurons do not express cFOS at rest nor fallowing vehicle (PBS) microinjection (not shown here, but see Fig. 5).
Fig. 7
Fig. 7
Effects of gastric distension, PBS, or TNFα microinjection on a single neuron of the nucleus of the solitary tract (NST). A–C are records from the same neuron. (A) Upper trace on left: Raw oscillograph record of the identification of a gastric distension related neuron in the NST. Distension of the antrum (lower trace) produces a brisk increase in the firing rate of the NST neuron; phase-locked to the stimulus. Upper trace on right: Integrated rate-meter record of the same event. Inset: 20 superimposed NST spikes; profile of identified NST neuron responding to gastric distension alone. (B) integrated record of neuronal activity. (Upper) Control injection of PBS (3nL delivered at arrow) has no effect on the neuron’s firing rate; same neuron as recorded in A. Inset shows the neuron’s signature spike profile. (Lower) TNFα (0.03 femtomoles) microinjected onto the same NST neuron causes activation of identified neuron. Scale bar = 1 minute. (C) Response to gastric distension of the same NST neuron depicted in A, now 30 minutes after exposure to and recovery from the TNFα injection seen in B. Note the greatly potentiated response to gastric distension that significantly outlasts the stimulus. Inset shows superimposed, individual NST spikes identifying this neuron as the same as that seen in A. Scale bars for insets: 200uV/2.5msec. Adapted with permission from Emch and others (2000).
Fig. 8
Fig. 8
TNFα inhibits tonically active gastric dorsal vagal motor nucleus (DMN) neurons. (A) A raw oscilloscope spike record shows a reduction in neuronal activity (upper trace) during gastric balloon distension (lower trace), identifying the cell as a DMN neuron. (B) Response of the same DMN neuron to application of TNFα (0.03 femtomoles) at the arrow (note the difference in lime scales); spontaneous activity is inhibited. (C) One hour after TNFα application, DMN firing rate of the same neuron in response to gastric distension returns to normal. (D) Intracellular recordings of in vitro brain slices also demonstrate that spontaneous activity of DMN neurons is inhibited by TNFα (10pM). Adapted with permission from Emch and others (2001).
Fig. 9
Fig. 9
Constitutive expression of TNF receptor type 1 immunoreactivity (TNFR1-ir) on afferents in the medullary brainstem of the rat. (A) Ponto-medullary junction area illustrating the entry of vagal sensory nerve fibers staining positively for TNFR1-ir. Trigeminal afferents in the spinal nucleus of the trigeminal nerve are also TNFR1-ir positive. (B) Vagal afferent fibers in the solitary tract (ST) staining for TNFR1-ir. (C) Vagal afferent fibers diverging from the ST and entering the medial solitary nucleus (mNST). (D) TNFR1-ir positive vagal afferent fibers continue through the mNST and ramify widely through the area postrema (higher magnification in inset on left). (E) TNFR1-ir on afferents in the dorsal root entry zone (the section shown is the cervical spinal cord). (F) TNFR1-ir labeled trigeminal afferent fibers penetrate the surface of the inferior cerebellar peduncle (ICP) and diverge within the spinal trigeminal nucleus (SN V). (G) Higher power detail of TNFR1-ir–positive fibers on the brainstem surface penetrating the ICR (H) Higher power detail of TNFR1-ir afferent fibers in the SN V. Abbreviations: AP = area postrema; ICP = inferior cerebellar peduncle; mNST = medial solitary nucleus; SN V = spinal nucleus of trigeminal nerve; ST = solitary tract; X aff = vagal sensory nerve fibers. Scale bars: F = 1000 micron; A = 500 micron; B, C, G, H = 100 micron; D = 200 micron; E = 20 microns. Adapted with permission from Hermann and others (2004).
Fig. 10
Fig. 10
Modulation of vagal afferent terminals responses to ATP by exposure to TNFα. A low-power field brainstem slice contains a section of the nucleus of the solitary tract (NST); numerous calcium green-labeled central vagal afferent terminals are visible. A–D depict the exact same field of this brainstem slice under different perfusion conditions. A and B were taken before exposure of the slice to 1nM TNFα; C and D were taken just after 10-minute exposure to TNFα. (A) Varicosities at rest just before perfusion of the slice with ATP (100uM). (B) Varicosities in the same field as A showing the peak effect of ATP. (C) Varicosities at rest. (D) Varicosities in the same field as C showing the peak effect of ATP after the slice was exposed to TNFα. (E) Relative fluorescence plot of the initial response to ATP to elevate the calcium signal in the varicosity that is indicated by the arrow (see B) before TNFα exposure. (F) Relative fluorescence plot of the response to ATP to elevate the calcium signal in the same varicosity indicated by the arrow (see D) after TNFα exposure. This TNF amplification of the calcium signal is sensitive to ryanodine (see Fig. 11). Adapted with permission from Rogers, Van Meter, and Hermann (2006).
Fig. 11
Fig. 11
TNFα can amplify terminal calcium signaling by activating ryanodine channels. ATP acts at terminal ligand-gated (P2X3 receptor) cation channels to produce an increase in intraterminal calcium. This increase, in turn, activates calcium-induced calcium release (CICR) mechanisms. TNFα apparently amplifies this effect via the generation of cAOP ribose (cADPR), a known ryanodine channel agonist molecule. CD38 generates cADPR; TNFα drives CD38 transcription (Iqbal and others 2006) and may enhance CD38 trafficking. Adapted with permission from Rogers, Van Meter, and Hermann (2006).
Fig. 12
Fig. 12
Neurodegeneration after vagotomy. (A) Montage of micrographs illustrating the overall effects on TNF receptor type 1 immunoreactivity (TNFRI-ir) staining within the medulla two weeks after unilateral vagotomy (left side) at the supra-nodose level. Supra-nodose vagotomy effectively disconnects the primary afferent cell bodies from central afferent fibers. These unilateral vagotomies allowed each animal to serve as its own control; comparisons between left (vagotomized) and right (control) sides could be made directly. (Right) Control side: brainstem side contralateral to the vagus nerve section. Note the constitutive presence of TNFR1-ir vagal afferents in the solitary tract (ST) and tenth nerve (X) afferent pathway. Also note the lack of TNFR1-ir in the dorsal motor nucleus of the vagus (DMN) on this side of the brainstem. (Left) Side ipsilateral to the supra-nodose vagotomy. Note the complete absence of TNFR1-ir in the afferent fibers of the ST on this side. In contrast, the neurons in the DMN on this side show enhanced TNFR1-ir. Following vagotomy there is also an increase in staining of the efferent fibers leaving the DMN neurons (at *) relative to the control side. Scale bar = 0.5 mm. (B) Photomicrographs of medullary brainstem sections demonstrating the time course of the up regulation of TNFR1-ir in DMN neurons on the side ipsilateral to the vagotomy (left side) relative to the control (right) side. Within three days following unilateral vagotomy, neurons of the DMN (and nucleus ambiguus [NA], data not shown), ipsilateral to vagal section, demonstrated a dramatic increase in TNFR1-ir throughout the entire rostro-caudal extent of these nuclei. This increase in the number of neurons expressing TNFR1-ir was quantified across time postvagotomy; the number of neurons expressing TNFR1-ir in each nucleus peaked at one week and began to decline over time. The absolute number of DMN neurons on the vagotomized side also declined over time. Panels i–v illustrate changes in the appearance of TNFR1-ir in DMN neurons on the side of the vagotomy (left) versus control (right) over the course of time. Scale bar = 500 microns. (C) High-power views of the DMN overtime. Note Wallerian degeneration of the DMN as characterized by the large, eccentric nuclei. Scale = 100 micron. Adapted with permission from Hermann and others (2004).

Similar articles

See all similar articles

Cited by 25 PubMed Central articles

See all "Cited by" articles

Publication types

MeSH terms

Substances

Feedback