Non-peptidergic primary afferents are presynaptic to neurokinin-1 receptor immunoreactive lamina I projection neurons in rat spinal cord

Abstract

Background: Pain-related (nociceptive) information is carried from the periphery to the dorsal horn of the spinal cord mostly by two populations of small diameter primary afferents, the peptidergic and the non-peptidergic. The peptidergic population expresses neuropeptides, such as substance P and calcitonin gene-related peptide, while the non-peptidergic fibers are devoid of neuropeptides, express the purinergic receptor P2X3, and bind the isolectin B4 (IB4). Although it has been known for some time that in rat the peptidergic afferents terminate mostly in lamina I and outer lamina II and non-peptidergic afferents in inner lamina II, the extent of the termination of the latter population in lamina I was never investigated as it was considered as very minor. Because our preliminary evidence suggested otherwise, we decided to re-examine the termination of non-peptidergic afferents in lamina I, in particular with regards to their innervation of projection neurons expressing substance P receptors (NK-1r). We used retrograde labeling of neurons from the parabrachial nucleus combined with lectin IB4 binding and immunocytochemistry. Samples were examined by confocal and electron microscopy.

Results: By confocal microscopy, we studied the termination of non-peptidergic afferents in lamina I using IB4 binding and P2X3 immunoreactivity as markers, in relation to CGRP immunoreactivy, a marker of peptidergic afferents. The number of IB4 or P2X3-labeled fibers in lamina I was higher than previously thought, although they were less abundant than CGRP-labeled afferents. There were very few fibers double-labeled for CGRP and either P2X3 or IB4. We found a considerable number of IB4-positive fiber varicosities in close apposition to NK-1r-positive lamina I projection neurons, which were distinct from peptidergic varicosities. Furthermore, we confirmed at the ultrastructural level that there were bona fide synapses between P2X3-immunoreactive non-peptidergic boutons and neurokinin-1 receptor-positive lamina I dendrites.

Conclusions: These results indicate the presence of direct innervation by non-peptidergic nociceptive afferents of lamina I projection neurons expressing NK-1r. Further investigations are needed to better understand the role of these connections in physiological conditions and chronic pain states.

Figure 1

CGRP, IB4 and P2X3 staining in transverse spinal cord sections.A and B show low magnification confocal images of CGRP-IR and IB4 positive (A) or P2X3-IR (B) fibers. C and D represent high magnification confocal images from the middle third of the latero-medial extent of the superficial dorsal horn. In C, note that there is limited co-localization of IB4 and CGRP (in yellow). Arrowheads show axonal varicosities (boutons) from non-peptidergic fibers in lamina I, which do not co-localize CGRP immunoreactivity. The framed regions in A and B indicate the approximate regions from where C and D, respectively, were obtained (the latter originate from other sections). CGRP (in green); IB4 (in red); P2X3 (in red). Scale bar (A, B) = 200 μm; scale bar (C, D) = 20 μm.

Figure 2

Confocal images at high magnification obtained from parasagittal spinal cord sections showing CGRP-IR (green) and P2X3-IR (red) varicosities in the superficial dorsal horn. P2X3-IR varicosities were present in considerable number in lamina I (LI) but were more highly concentrated in inner lamina II (LIIi). Scale bar = 20 μm.

Figure 3

Confocal images at high power obtained from horizontal spinal cord sections. In a confocal optical section from lamina I adjacent to the white matter (A), note the relatively abundant P2X3-IR fibers with varicosities (boutons). CGRP-IR fibers and boutons were considerably more abundant in this lamina. In a confocal optical section from inner lamina II (B), note the very high density of P2X3-IR fibers and varicosities, higher than that of CGRP-IR fibers in lamina I. Note that most varicosities display either P2X3 or CGRP immunoreactivity, although some co-localization is observed (yellow). Scale bar (A, B) = 20 μm.

Figure 4

Confocal triple-labeling image showing the innervation of lamina I spinoparabrachial (A) multipolar, (B) fusiform and (C) pyramidal neurons by IB4+ varicosities (indicated by arrowhead). IB4 (red); NK-1r (green); spinoparabrachial neurons labeled with CTb (white). Scale bar (A-C) = 20 μm.

Figure 5

Example of a quadruple labeling observed at the confocal level using a multi-track approach. In this image the following signals were simultaneously detected: CGRP (green); IB4 binding (red); CTb transported retrogradely from the parabrachial nucleus (blue); NK-1r (white). A fusiform neuron, double labeled with CTb and NK-1r, is innervated by CGRP-IR boutons (arrowhead) and IB4+ (arrow) boutons, which represent distinct populations. However, a small population of varicosities co-labeled for CGRP and IB4 (curved arrow) was detected. Scale bar (A-E) = 20 μm.

Figure 6

Quantification of IB4-positive boutons in direct apposition to the cell body and primary dendrites of NK-1r-IR lamina I spinoparabrachial neurons. Results are expressed in number of appositions from boutons per 100 μ of membrane length of NK-1r-IR cell. N = 6 animals.

Figure 7

Electron micrograph showing: in A, a P2X3-IR bouton (arrow) forming an axo-somatic contact onto a lamina I cell body (CB), which is also postsynaptic to unlabeled axonal boutons (V); in B, a P2X3-IR central bouton of a type Ia glomerulus (CI P2X3+) presynaptic to 3 lamina I dendrites (D) and apposed to an axonal bouton (V2). Scale bar = 0.5 μm.

Figure 8

Electron micrographs showing axo-dendritic contacts between P2X3-positive boutons (arrows) and NK-1r-positive dendrites (D) (A and B). Note the silver-intensified gold particles along the plasma membrane of the dendrites, representing NK-1r-IR sites. B, unlabeled axonal bouton. Scale bar = 0.5 μm.