The first brain: Species comparisons and evolutionary implications for the enteric and central nervous systems.

BACKGROUND: The enteric nervous system (ENS) and the central nervous system (CNS) of mammals both contain integrative neural circuitry and similarities between them have led to the ENS being described as the brain in the gut. PURPOSE: To explore relationships between the ENS and CNS across the animal kingdom. We found that an ENS occurs in all animals investigated, including hydra, echinoderms and hemichordates that do not have a CNS. The general form of the ENS, which consists of plexuses of neurons intrinsic to the gut wall and an innervation that controls muscle movements, is similar in species as varied and as far apart as hydra, sea cucumbers, annelid worms, octopus and humans. Moreover, neurochemical similarities across phyla imply a common origin of the ENS. Investigation of extant species suggests that the ENS developed in animals that preceded the division that led to cnidaria (exemplified by hydra) and bilateria, which includes the vertebrates. The CNS is deduced to be a bilaterian development, later than the divergence from cnidaria. Consistent with the ENS having developed independent of the CNS, reciprocal connections between ENS and CNS occur in mammals, and separate neurons of ENS and CNS origin converge on visceral organs and prevertebral ganglia. We conclude that an ENS arose before and independently of the CNS. Thus the ENS can be regarded as the first brain.

Effects of central administration of resistin on renal sympathetic nerve activity in rats fed a high-fat diet: a comparison with leptin.

Similar to leptin, resistin acts centrally to increase renal sympathetic nerve activity (RSNA). In high-fat fed animals, the sympatho-excitatory effects of leptin are retained, in contrast to the reduced actions of leptin on dietary intake. In the present study, we investigated whether the sympatho-excitatory actions of resistin were influenced by a high-fat diet. Further, because resistin and leptin combined can induce a greater sympatho-excitatory response than each alone in rats fed a normal chow diet, we investigated whether a high-fat diet (22%) could influence this centrally-mediated interaction. Mean arterial pressure (MAP), heart rate (HR) and RSNA were recorded before and for 3 hours after i.c.v. saline (control; n=5), leptin (7 µg; n=4), resistin (7 µg; n=5) and leptin and resistin combined (n=6). Leptin alone and resistin alone significantly increased RSNA (71±16%, 62±4%, respectively). When leptin and resistin were combined, there was a significantly greater increase in RSNA (195±41%) compared to either hormone alone. MAP and HR responses were not significantly different between hormones. When the responses in high-fat fed rats were compared to normal chow fed rats, there were no significant differences in the maximum RSNA responses. The findings indicate that sympatho-excitatory effects of resistin on RSNA are not altered by high-fat feeding, including the greater increase in RSNA observed when resistin and leptin are combined. Our results suggest that diets rich in fat do not induce resistance to the increase in RSNA induced by resistin alone or in combination with leptin.

Calbindin-D-28K like immunoreactivity in superficial dorsal horn neurons and effects of sciatic chronic constriction injury.

The neuropathic pain that results from peripheral nerve injury is associated with alterations in the properties of neurons in the superficial spinal laminae. Chronic constriction injury (CCI) of the rat sciatic nerve increases excitatory synaptic drive to excitatory neurons in the substantia gelatinosa while limiting that to inhibitory neurons. Since the calcium-binding protein calbindin D-28K has been associated with excitatory neurons, we examined whether CCI altered the properties of neurons expressing calbindin-like immunoreactivity (Cal+). These account for 30% of the neurons in lamina I and II. Calbindin did not co-localize with any particular electrophysiological phenotype of neuron; in substantia gelatinosa, it was found in some tonic, delay, irregular, phasic and transient firing neurons and in some cells that displayed central, radial or vertical morphology. When neuronal phenotype was defined more precisely in terms of both morphology and electrophysiological properties, no strong correlation with calbindin expression was found. The frequency and amplitude of spontaneous excitatory postsynaptic currents (sEPSC) in calbindin negative (Cal-) neurons was greater than that in Cal+ neurons. CCI did not alter the proportion of Cal+ neurons in the dorsal horn. Although CCI promoted a fourfold increase in sEPSC frequency in Cal+ neurons, sEPSC amplitude was reduced by 22% and charge transfer per second was unchanged. Since synaptic drive to Cal+ neurons is weak and there is no firm correlation between neuronal phenotype and calbindin expression, it is doubtful whether these neurons play a major role in the generation of central sensitization.

Microglia are selectively activated in endocrine and cardiovascular control centres in streptozotocin-induced diabetic rats.

Type 1 and 2 diabetes are associated with dysfunction in multiple hormone systems, as well as increased sympathetic nerve activity, which may contribute to the development of diabetic complications. In other pathologies, such as myocardial infarction, increased sympathetic drive is associated with neuroinflammation and microglial activation in the hypothalamic paraventricular nucleus (PVN), a brain region that regulates sympathetic drive and multiple endocrine responses. In the present study, we used immunohistochemistry to study microglial and neuronal activation in the PVN and related brain regions in streptozotocin (STZ)-induced diabetic rats. As expected, STZ treatment was associated with elevated blood glucose within 1 week. STZ injections also caused neuronal activation in the PVN and superoptic nucleus (SON) but not in the nucleus tractus solitarius (NTS), which was evident by 6 weeks. STZ-treated rats showed increased plasma osmolarity, which would be expected to activate PVN and SON neurones. There was no apparent increase in histochemical markers of microglial activation, including phospho-p38, phospho-extracellular signal regulated kinase, P2X4 receptor or interleukin 1-ß even at 10 weeks after STZ-treatment. However, we did see a significant increase in the percentage of microglia with an activated morphology in the PVN, SON and NTS, although not in surrounding hypothalamic, brainstem or cortical regions. These morphological changes included a significant reduction in microglial process length and were evident by 8 weeks but not 6 weeks. The delayed onset of microglial changes compared to neuronal activation in the PVN and SON suggests the over-excitation of neurones as a mechanism of microglial activation. This delayed microglial activation may, in turn, contribute to the endocrine dysregulation and the elevated sympathetic nerve activity reported in STZ-treated rats.

Analysis of purinergic and cholinergic fast synaptic transmission to identified myenteric neurons.

Types and projections of neurons that received cholinergic, purinergic and other fast excitatory synaptic inputs in myenteric ganglia of the guinea-pig distal colon were identified using combined electrophysiological recording, application of selective antagonists, marker dye filling via the recording microelectrode, and immunohistochemical characterisation. Fast synaptic inputs were recorded from all major subtypes of uniaxonal neurons including Dogiel type I neurons, filamentous interneurons, circular muscle motor neurons and longitudinal muscle motor neurons. Fast excitatory postsynaptic potentials were completely blocked by the nicotinic receptor antagonists hexamethonium or mecamylamine in 62% of neurons tested and were partially inhibited in the remaining neurons. The P2 purine receptor antagonist, pyridoxal-phosphate-6-azophenyl-2',4'-disulfonic acid, reduced the amplitudes of fast excitatory postsynaptic potentials in 20% of myenteric neurons. The 5-hydroxytryptamine(3) receptor antagonist granisetron reduced the amplitude of fast excitatory postsynaptic potentials in only one of 15 neurons tested. In five of five neurons tested, the combination of a nicotinic antagonist, pyridoxal-phosphate-6-azophenyl-2',4'-disulfonic acid, granisetron and 6-cyano-7-nitroquinoxaline-2,3-dione did not completely block the fast excitatory postsynaptic potentials. Immunohistochemical studies of the neurons that had been identified electrophysiologically and morphologically imply that P2X(2) receptors may mediate fast transmission in some neurons, and that other P2X receptor subtypes may also be involved in fast synaptic transmission to myenteric neurons of the guinea-pig distal colon. Neurons with nicotinic and pyridoxal-phosphate-6-azophenyl-2',4'-disulfonic acid-sensitive fast excitatory postsynaptic potentials were present in both ascending and descending pathways in the distal colon. Thus, neither cholinergic nor mixed cholinergic/purinergic synaptic responses are confined to a particular class of neuron. The results indicate that acetylcholine and ATP are the major fast excitatory neurotransmitters in guinea-pig distal colon myenteric ganglia.

Effects of substance P on excitability and ionic currents of normal and axotomized rat dorsal root ganglion neurons.

Substance P (SP) may act within dorsal root ganglia (DRG) to modulate the transmission of nociceptive information. Because peripheral nerve injury (axotomy) alters the peptide content of sensory neurons, we used whole-cell recording to examine the effects of sciatic nerve section on the sensitivity of rat lumbar DRG neurons to SP (0.3--1 microM). At 1 microM, SP increased the excitability of 'small', putative nociceptive neurons but had little effect on the excitability of 'large' neurons. Two-four weeks after sciatic nerve section, however, the effect of SP on 'large' axotomized neurons was increased and its effect on 'small' neurons was decreased. SP did not affect Ca(2+) channel currents in control or axotomized neurons. The effects of SP on the current-voltage (I--V) relationship of 77% of neurons involved increased inward current at potentials below -30 mV and suppressed outward current at potentials above -20 mV. The effects of SP on the I--V relationship were similar in control and in axotomized neurons and the altered sensitivity of 'small' and 'large' cells could not be attributed to axotomy-induced changes in input resistance or membrane potential. The possible relevance of alterations in sensitivity, of 'large' DRG neurons to SP, to the generation of neuropathic pain is discussed.

Changes in the action potential in sensory neurones after peripheral axotomy in vivo.

Following nerve injury, modified somatic ion channels may underlie ectopic activity in axotomized A-type neurones in dorsal root ganglia (DRGs) leading to abnormal pain signalling. Using intracellular microelectrodes both in vivo and in vitro, action potentials (APs) were recorded in rat DRG neurones classified by axonal conduction velocity. After lesions to L5 spinal or sciatic nerves, APs in both A alpha/beta and A delta cells were wider, and those in A alpha/beta neurones more frequently showed inflections during repolarization, than APs in cells in undamaged ganglia. AP amplitudes and dV/dt(max) were not significantly altered by axotomy. These results confirm previous observations in intact ganglia in vitro but differ from those reported for dissociated neurones using patch recording techniques.

Are there functional P2X receptors on cell bodies in intact dorsal root ganglia of rats?

P2X purinoceptors have been suggested to participate in transduction of painful stimuli in nociceptive neurons. In the current experiments, ATP (1-10 mM), alpha,beta-methylene-ATP (10-30 microM) and capsaicin (10 nM-1 microM) were applied to neurons impaled with high resistance microelectrodes in rat dorsal root ganglia (L4 and L5) isolated in vitro together with the sciatic nerve and dorsal roots. The agonists were either bath applied or focally applied using a picospritzer. GABA (100 microM) and 40-80 mM K+ solutions gave brisk responses when applied by either technique. Only three of 22 neurons with slowly conducting axons (C cells) showed evidence of P2X-purinoceptor-mediated responses. Only two of 13 cells which responded to capsaicin (putative nociceptors), and none of 29 cells with rapidly conducting axons (A cells), responded to the purinergic agonists. When acutely dissociated dorsal root ganglion cells were studied using patch-clamp techniques, all but four of 30 cells of all sizes responded with an inward current to either ATP or alpha,beta-methylene-ATP (both 100 microM). Our data suggest that few sensory cell bodies in intact dorsal root ganglia express functional purinoceptors.

Inhibitory transmission to the longitudinal muscle of the mouse caecum is mediated largely by nitric oxide acting via soluble guanylyl cyclase.

In this work we describe a region of mouse intestine, the caecum, in which inhibitory transmission to the longitudinal muscle is predominantly due to nitric oxide. In the presence of muscarinic receptor blockade, electrical stimulation of intramural nerves in the longitudinal muscle of the mouse caecum evoked a relaxation. The relaxation was reduced to about 25% of the control amplitude by the nitric oxide synthase inhibitor L-NMMA (NG-methyl-L-arginine), but was unaffected by D-NMMA. In the presence of the nitric oxide scavenger oxyhaemoglobin, the relaxation was reduced to less than 10% of the control amplitude. In the circular muscle of the caecum and the longitudinal muscle of the ileum, colon and rectum, electrical field stimulation either evoked only small relaxations, or relaxations that were unaffected by L-NMMA. Nitric oxide synthase-containing neurons in the caecum were localized immunohistochemically using an antibody to neuronal nitric oxide synthase or with NADPH diaphorase histochemistry. Reactive nerve cell bodies were observed in the myenteric plexus, and varicose nerve fibres were present in the longitudinal and circular muscle layers of the caecum. The transduction mechanism of the nitric oxide-mediated relaxation in the longitudinal muscle of the caecum was examined using ODQ (1 H-[1,2,4]oxadiazolo[4,3,-alpha-]quinoxalin-1-one), a selective inhibitor of soluble guanylyl cyclase. ODQ abolished the relaxations induced by applied sodium nitroprusside (0.01-1 mM) and reduced the relaxation induced by electrical stimulation to about 40% of control values. However, ODQ reduced the relaxations induced by electrical stimulation to a lesser extent than L-NMMA. Hence, although the relaxation in this tissue mediated by NO (or an NO-related substance) is largely via soluble guanylyl cyclase, an action of NO on other targets cannot be ruled out.

Electrophysiological mapping of fast excitatory synaptic inputs to morphologically and chemically characterized myenteric neurons of guinea-pig small intestine.

Neurons within the myenteric plexus of the guinea-pig ileum were impaled using conventional intracellular electrodes. Points of stimulation within the surrounding ganglia and connectives which gave rise to fast excitatory synaptic potentials were mapped using a movable monopolar stimulating electrode. Cells were then injected with the intracellular marker, biocytin, and processed for multiple label immunohistochemistry to reveal their morphologies, chemical contents and, hence, their functional classes. Of 65 neurons belonging to the S electrophysiological class, 53 received fast excitatory synaptic inputs from stimulation at sites at least 2 mm away in a directly circumferential direction. These inputs almost certainly arise from stimulation of the circumferentially-directed axons of the Dogiel type II/AH-neurons, which are thought to be intrinsic sensory neurons. The majority of cells which projected anally and were immunoreactive for nitric oxide synthase (19/25), all neurons which ramified in the tertiary plexus and were identified as longitudinal muscle motor neurons (6/6) and all neurons identified as excitatory motor neurons innervating the circular muscle (12/12) received inputs from these circumferentially-directed pathways. However only one of six descending filamentous interneurons impaled received such inputs, suggesting they may be differentially innervated. The conduction velocities of circumferentially-directed axons giving rise to fast excitatory post synaptic potentials were estimated to be 0.41 +/- 0.10 m/s (mean +/- standard deviation, n = 21). The conduction velocities estimated for longitudinally-directed pathways were 0.55 +/- 0.25 m/s (n = 29). Thus, the majority of myenteric neurons receive fast excitatory synaptic input from putative intrinsic sensory neurons which project circumferentially around the intestine.

Cholinergic, somatostatin-immunoreactive interneurons in the guinea pig intestine: morphology, ultrastructure, connections and projections.

The shape, projection, ultrastructure and chemistry of interneurons that were initially identified by their immunoreactivity for somatostatin in the small intestine of the guinea pig were examined. Somatostatin immunoreactive nerve cell bodies and nerve fibres were located in the myenteric plexus. Simultaneous labelling for 2 antigens revealed that the somatostatin immunoreactive interneurons were also immunoreactive for choline acetyltransferase, but not for calbindin or neuropeptide Y. Cell shapes were determined by immunohistochemistry, ultrastructural analysis and intracellular dye filling. The neurons had large cell bodies (38 x 14 microns) which gave rise to long branched filamentous dendrites and a single axon. The projections of the somatostatin immunoreactive interneurons were determined by analysis of patterns of fibre loss and survival following degenerative section of myenteric nerve pathways and by analysis of individual neurons that were injected intracellularly with dye. The axons projected anally, sometimes after a short oral excursion, and were confined to the myenteric plexus. They gave rise to multiple varicose branches within myenteric ganglia: in the majority of cases the first branch was within 100 microns of the cell body. Synaptic inputs to the cells were examined by light and electron microscopy. All somatostatin immunoreactive neurons received numerous somatostatin immunoreactive inputs on the cell body and all filamentous processes. Ultrastructural investigation indicated that these constituted the majority of all inputs. It is concluded that cholinergic, somatostatin immunoreactive, interneurons have a unique soma morphology and form synaptically connected chains that run anally in the myenteric plexus.

Electrophysiological analysis of the convergence of peripheral inputs onto neurons of the coeliac ganglion in the guinea pig.

The convergence of intestinofugal axons from different intestinal regions onto individual neurons in the coeliac ganglion of the guinea pig was investigated using intracellular recording methods in vitro. Peripheral nerve trunks from the distal ileum, the most proximal colon and the colon near the colonic flexure were electrically stimulated along with preganglionic fibres running in the splanchnic nerve. Fast cholinergic excitatory synaptic potentials (EPSPs) were seen in ganglion cells in response to stimulation of each nerve trunk. Roughly half of 78 neurons impaled received inputs from stimulation of peripheral nerves, and almost all of these received input from the proximal colon. Most cells responded to stimulation of more than one peripheral nerve indicating that coeliac neurons receive converging inputs from intestinofugal neurons located in more than one intestinal region. In a second series of experiments, segments of intestine were left attached to the ganglion and distended with saline to stimulate peripheral mechanosensory input to the coeliac ganglion. In each experiment, two segments were stimulated. A subgroup of ganglion cells exhibited spontaneous fast EPSPs and the frequency of these potentials was increased by distension of one or other of the attached intestinal segments. However, few neurons responded to distension of both of the attached intestinal segments suggesting that some of the intestinofugal inputs to the coeliac ganglion identified by electrical stimulation may be sensitive to sensory modalities other than distension. Hexamethonium (0.5 mM) applied to the intestine, and not to the coeliac ganglion, reduced the frequency of the spontaneous synaptic potentials seen in coeliac ganglion cells, but did not abolish the response to distension of the colon (n = 8). When the Ca2+ concentration of the solution bathing the proximal colon was reduced to block all synaptic transmission in the enteric plexuses the background synaptic input was further depressed, but again the response to distension was little changed (n = 4). This suggests that at least some of the neurons projecting from the colon to the coeliac ganglion are first order mechanosensory neurons.