Evidence for Two Subpopulations of Cerebrospinal Fluid-Contacting Neurons with Opposite GABAergic Signaling in Adult Mouse Spinal Cord.

Spinal cerebrospinal fluid-contacting neurons (CSF-cNs) form an evolutionary conserved bipolar cell population localized around the central canal of all vertebrates. CSF-cNs were shown to express molecular markers of neuronal immaturity into adulthood; however, the impact of their incomplete maturation on the chloride (Cl-) homeostasis as well as GABAergic signaling remains unknown. Using adult mice from both sexes, in situ hybridization revealed that a proportion of spinal CSF-cNs (18.3%) express the Na+-K+-Cl- cotransporter 1 (NKCC1) allowing intracellular Cl- accumulation. However, we did not find expression of the K+-Cl- cotransporter 2 (KCC2) responsible for Cl- efflux in any CSF-cNs. The lack of KCC2 expression results in low Cl- extrusion capacity in CSF-cNs under high Cl- load in whole-cell patch clamp. Using cell-attached patch clamp allowing recordings with intact intracellular Cl- concentration, we found that the activation of ionotropic GABAA receptors (GABAA-Rs) induced both depolarizing and hyperpolarizing responses in CSF-cNs. Moreover, depolarizing GABA responses can drive action potentials as well as intracellular calcium elevations by activating voltage-gated calcium channels. Blocking NKCC1 with bumetanide inhibited the GABA-induced calcium transients in CSF-cNs. Finally, we show that metabotropic GABAB receptors have no hyperpolarizing action on spinal CSF-cNs as their activation with baclofen did not mediate outward K+ currents, presumably due to the lack of expression of G-protein-coupled inwardly rectifying potassium (GIRK) channels. Together, these findings outline subpopulations of spinal CSF-cNs expressing inhibitory or excitatory GABAA-R signaling. Excitatory GABA may promote the maturation and integration of young CSF-cNs into the existing spinal circuit.

Evidence for PKD2L1-positive neurons distant from the central canal in the ventromedial spinal cord and medulla of the adult mouse.

Neurons in contact with the cerebrospinal fluid (CSF) are found around the medullo-spinal central canal (CC) in adult mice. These neurons (CSF-cNs), located within or below the ependymal cell layer, known as the stem cell niche, present a characteristic morphology with a dendrite projecting to the CC and ending with a protrusion. They are GABAergic, present an intermediate neuronal maturity and selectively express PKD2L1, a member of the transient receptor potential channel superfamily with sensory properties. Using immunohistological and electrophysiological recording techniques in mice, we characterize the properties of a new population of PKD2L1 positive cells that is distant from the CC in a zone enriched with astrocytes and ependymal fibers of the ventro-medial spinal cord and medulla. They appear around embryonic day 16 and their number increases up to early postnatal days. With development and the reorganization of the CC region, they progressively become more distant from the CC, suggesting some migratory capabilities. These neurons share functional and phenotypical properties with CSF-cNs but appear subdivided in two groups. One group, present along the midline, has a bipolar morphology and extends a long dendrite along ependymal fibers and towards the CC. The second group, localized in more ventro-lateral regions, has a multipolar morphology and no apparent projection to the CC. Altogether, we describe a novel population of PKD2L1+ neurons distant from the CC but with properties similar to CSF-cNs that might serve to sense modification in the composition of either CSF or interstitial liquid, a function that will need to be confirmed.

GABAB receptors modulate Ca2+ but not G protein-gated inwardly rectifying K+ channels in cerebrospinal-fluid contacting neurones of mouse brainstem.

KEY POINTS: Medullo-spinal CSF contacting neurones (CSF-cNs) located around the central canal are conserved in all vertebrates and suggested to be a novel sensory system intrinsic to the CNS. CSF-cNs receive GABAergic inhibitory synaptic inputs involving ionotropic GABAA receptors, but the contribution of metabotropic GABAB receptors (GABAB -Rs) has not yet been studied. Here, we indicate that CSF-cNs express functional GABAB -Rs that inhibit postsynaptic calcium channels but fail to activate inhibitory potassium channel of the Kir3-type. We further show that GABAB -Rs localise presynaptically on GABAergic and glutamatergic synaptic inputs contacting CSF-cNs, where they inhibit the release of GABA and glutamate. Our data are the first to address the function of GABAB -Rs in CSF-cNs and show that on the presynaptic side they exert a classical synaptic modulation whereas at the postsynaptic level they have an atypical action by modulating calcium signalling without inducing potassium-dependent inhibition. ABSTRACT: Medullo-spinal neurones that contact the cerebrospinal fluid (CSF-cNs) are a population of evolutionary conserved cells located around the central canal. CSF-cN activity has been shown to be regulated by inhibitory synaptic inputs involving ionotropic GABAA receptors, but the contribution of the G-protein coupled GABAB receptors has not yet been studied. Here, we used a combination of immunofluorescence, electrophysiology and calcium imaging to investigate the expression and function of GABAB -Rs in CSF-cNs of the mouse brainstem. We found that CSF-cNs express GABAB -Rs, but their selective activation failed to induce G protein-coupled inwardly rectifying potassium (GIRK) currents. Instead, CSF-cNs express primarily N-type voltage-gated calcium (CaV 2.2) channels, and GABAB -Rs recruit Gßγ subunits to inhibit CaV channel activity induced by membrane voltage steps or under physiological conditions by action potentials. Moreover, using electrical stimulation, we indicate that GABAergic inhibitory (IPSCs) and excitatory glutamatergic (EPSCs) synaptic currents can be evoked in CSF-cNs showing that mammalian CSF-cNs are also under excitatory control by glutamatergic synaptic inputs. We further demonstrate that baclofen reversibly reduced the amplitudes of both IPSCs and EPSCs evoked in CSF-cNs through a presynaptic mechanism of regulation. In summary, these results are the first to demonstrate the existence of functional postsynaptic GABAB -Rs in medullar CSF-cNs, as well as presynaptic GABAB auto- and heteroreceptors regulating the release of GABA and glutamate. Remarkably, postsynaptic GABAB -Rs associate with CaV but not GIRK channels, indicating that GABAB -Rs function as a calcium signalling modulator without GIRK-dependent inhibition in CSF-cNs.

The role of intraspinal sensory neurons in the control of quadrupedal locomotion.

From swimming to walking and flying, animals have evolved specific locomotor strategies to thrive in different habitats. All types of locomotion depend on the integration of motor commands and sensory information to generate precisely coordinated movements. Cerebrospinal-fluid-contacting neurons (CSF-cN) constitute a vertebrate sensory system that monitors CSF composition and flow. In fish, CSF-cN modulate swimming activity in response to changes in pH and bending of the spinal cord; however, their role in mammals remains unknown. We used mouse genetics to study their function in quadrupedal locomotion. We found that CSF-cN are directly integrated into spinal motor circuits. The perturbation of CSF-cN function does not affect general motor activity nor the generation of locomotor rhythm and pattern but results in specific defects in skilled movements. These results identify a role for mouse CSF-cN in adaptive motor control and indicate that this sensory system evolved a novel function to accommodate the biomechanical requirements of limb-based locomotion.

Loss of inhibitory synapses causes locomotor network dysfunction of the rat spinal cord during prolonged maintenance in vitro.

The isolated spinal cord of the neonatal rat is widely employed to clarify the basic mechanisms of network development or the early phase of degeneration after injury. Nevertheless, this preparation survives in Krebs solution up to 24 h only, making it desirable to explore approaches to extend its survival for longitudinal studies. The present report shows that culturing the spinal cord in oxygenated enriched Basal Medium Eagle (BME) provided excellent preservation of neurons (including motoneurons), glia and primary afferents (including dorsal root ganglia) for up to 72 h. Using DMEM medium was unsuccessful. Novel characteristics of spinal networks emerged with strong spontaneous activity, and deficit in fictive locomotion patterns with stereotypically slow cycles. Staining with markers for synaptic proteins synapsin 1 and synaptophysin showed thoroughly weaker signal after 3 days in vitro. Immunohistochemical staining of markers for glutamatergic and glycinergic neurons indicated significant reduction of the latter. Likewise, there was lower expression of the GABA-synthesizing enzyme GAD65. Thus, malfunction of locomotor networks appeared related to loss of inhibitory synapses. This phenomenon did not occur in analogous opossum preparations of the spinal cord kept in vitro. In conclusion, despite histological data suggesting that cultured spinal cords were undamaged (except for inhibitory biomarkers), electrophysiological data revealed important functional impairment. Thus, the downregulation of inhibitory synapses may account for the progressive hyperexcitability of rat spinal networks despite apparently normal histological appearance. Our observations may help to understand the basis of certain delayed effects of spinal injury like chronic pain and spasticity.