Two populations of neurokinin 1 receptor-expressing projection neurons in lamina I of the rat spinal cord that differ in AMPA receptor subunit composition and density of excitatory synaptic input

Abstract

Lamina I of the spinal cord contains many projection neurons that express the neurokinin 1 receptor (NK1r). It has been reported that these cells can undergo long-term potentiation (LTP), which may result from insertion of AMPA-type glutamate receptors (AMPArs) containing GluA1 or GluA4 subunits. We therefore investigated synaptic AMPAr expression on these cells with immunocytochemistry following antigen-retrieval. We also examined their density of glutamatergic input (by analysing AMPAr synaptic puncta and contacts from glutamatergic boutons), and phosphorylation of extracellular signal-regulated kinases (pERKs) following noxious stimulation. Our results indicate that there are two populations of NK1r-expressing projection neurons: large GluA4(+)/GluA1(-) cells with a high density of glutamatergic input and small GluA1(+)/GluA4(-) cells with a much lower input density. Results from pERK experiments suggested that the two groups may not differ in the types of noxious stimulus that activate them. Glutamatergic synapses on distal dendrites of the large cells were significantly longer than those on proximal dendrites, which presumably compensates for the greater attenuation of distally-generated excitatory postsynaptic currents (EPSCs). Both types of cell received contacts from peptidergic primary afferents, however, on the large cells these appeared to constitute over half of the glutamatergic synapses, and were often associated with elongated AMPAr puncta. This suggests that these afferents, which probably contain substance P, provide a powerful, secure synaptic input to large NK1r-expressing projection neurons. These results demonstrate the importance of GluA4-containing AMPArs in nociceptive transmission and raise the possibility that different forms of LTP in lamina I projection neurons may be related to differential expression of GluA1/GluA4.

Fig. 1

Confocal images showing AMPAr subunit immunoreactivity on the dendrites of large and small NK1r-expressing projection neurons in horizontal sections through lamina I. All images are from horizontal sections that had undergone a TSA reaction to reveal NK1r (red) followed by antigen retrieval with pepsin and immunostaining for GluA2 (blue) and either GluA4 or GluA1 (green). (a–d) Part of a dendrite belonging to a large retrogradely labelled cell (soma area 488 μm²) is outlined by NK1r immunostaining and is associated with numerous GluA2-immunoreactive puncta (some indicated with arrows) that are often elongated and that are also stained with the GluA4 antibody. Many GluA2⁺/GluA4⁻ puncta are seen in the surrounding neuropil and these are generally smaller than the puncta on the dendrite. (e–h) Part of the dendrite of another large NK1r⁺ projection cell (soma area 532 μm²) is associated with GluA2⁺ puncta that are not GluA1-immunoreactive (some of these are indicated with arrows). (i–l) The dendrite of a small NK1r⁺ projection neuron (soma area 211 μm²) has GluA2⁺ puncta (arrowheads) that are not GluA4-immunoreactive. (m–p) A dendrite belonging to another small NK1r⁺ projection cell (soma area 228 μm²) has GluA2⁺ puncta that are also GluA1-immunoreactive (arrowheads). Note that the density of GluA2 puncta is much lower on the dendrites of the small cells than on the dendrites of the large cells. Although there are several GluA2⁺ puncta close to the NK1r-immunoreactive dendrite in (m–p), these are not actually in the membrane. The images are projections of 11 (a–d), 10 (e–h), 3 (i–l) or 6 (m–p) optical sections at 0.3 μm z-separation. Scale bar=10 μm.

Fig. 2

Scatter plot showing the relation of soma area to the density of GluA2 puncta on the dendrites for the 32 NK1r-expressing projection neurons that were analysed. Cells were also tested for the presence of either GluA4 or GluA1 and symbols indicate the presence or absence of immunostaining for these subunits at their GluA2 puncta. The dashed line corresponds to a soma area of 300 μm².

Fig. 3

Sizes of AMPAr-immunoreactive puncta on NK1r-expressing lamina I projection neurons compared to those not associated with these cells. The histograms show the frequency distribution of puncta of different sizes for: (a) those that were not associated with NK1r⁺ profiles (n=300 puncta), (b) those present on the dendrites of small projection cells (soma area <300 μm², n=154 puncta on 10 cells) and (c) those on large projection cells (soma area >300 μm², n=671 puncta on nine cells). (d) shows the frequency distribution of puncta associated with the four projection neurons that were analysed in sections reacted for CGRP, GluA2 and GluA4. The grey portions of each bar represent puncta that were apposed to CGRP-immunoreactive boutons, while the white parts correspond to puncta not associated with these boutons. For further details, see text.

Fig. 4

Confocal images showing the relation of CGRP-immunoreactive boutons to GluA2⁺/GluA4⁺ puncta on the dendrites of a retrogradely-labelled lamina I neuron. (a) A low magnification image reveals several Fluorogold (FG) labelled cells in a horizontal section through lamina I. One of these is identified with an asterisk. The box indicates the region shown at higher magnification in (b). (b) Part of the dendrite of the cell marked with the asterisk is seen (between arrowheads) and contains several bright granules of FG. Boxes represent areas shown in the subsequent images. (c, e, g, i) and (d, f, h, j) show two regions of the dendrite of this cell which have been scanned to reveal Fluorogold (white), CGRP (red), GluA2 (blue) and GluA4 (green). Both these regions have several GluA2⁺/GluA4⁺ puncta. Some of these (arrowheads) are apposed to CGRP-immunoreactive boutons, while others (arrows) are not. The images are projections of seven optical sections at 1 μm z-separation (a) and of five (b), three (c, e, g, i) or two (d, f, h, j) optical sections at 0.5 μm separation. Scale bars=25 μm (a), 10 μm (b), 5 μm (c–j).

Fig. 5

Confocal images showing contacts from CGRP- and VGLUT2-containing boutons onto the dendrites of NK1r-expressing lamina I projection neurons. (a) shows the soma (asterisk) and part of a dendrite belonging to a small (soma area=276 μm²) projection neuron, which are outlined with the NK1r (green). (b) The corresponding field scanned to reveal CGRP (blue) and VGLUT2 (red). (c) The merged image shows scattered contacts from CGRP- and VGLUT2-immunoreactive boutons (some of which are indicated with arrowheads and arrows, respectively). (d) Part of the dendrite of a large projection neuron (soma area 498 μm²) is outlined by the NK1r. (e) This field scanned to reveal CGRP and VGLUT2. (f) The merged image shows numerous contacts from CGRP- and VGLUT2-immunoreactive boutons (some of which are indicated with arrowheads and arrows, respectively). The images are projections of two (a–c) and three (d–f) optical sections at 0.5 μm separation. Scale bar=10 μm.

Fig. 6

Scatter plot showing the relation of soma area to the density of contacts from boutons that were CGRP- and/or VGLUT2-immunoreactive on the dendrites for the 24 NK1r-expressing projection neurons that were analysed. The dashed line corresponds to a soma area of 300 μm².

Fig. 7

Confocal images showing pERK-immunoreactivity in NK1r-expressing lamina I neurons following injection of capsaicin into the hindpaw. (a) A projected confocal image stack from a horizontal section through the ipsilateral side of the L5 segment. Several pERK-immunoreactive neurons can be seen and five of these are numbered. (b–e, b′–e′) show more limited projections through cells 1 (b, b′), 2, 3 (c, c′), 4 (d, d′) and 5 (e, e′). Each of these cells is immunostained for both NK1r and pERK, and there is internalization of the receptor. Cells 1, 3 and 5 had somata >300 μm² cross-sectional area (519, 309, 614 μm², respectively), while cells 2 and 4 were in the 200–300 μm² range (203, 246 μm², respectively). A small NK1r-immunoreactive cell (soma area 149 μm²) that was negative for pERK is marked with an arrowhead in (d). Images are projections of nine (a), four (b, e) or three (c, d) optical sections at 2 μm z-spacing. Scale bar=20 μm.