Direct and Indirect Nociceptive Input from the Trigeminal Dorsal Horn to Pain-Modulating Neurons in the Rostral Ventromedial Medulla

Abstract

The brain is able to amplify or suppress nociceptive signals by means of descending projections to the spinal and trigeminal dorsal horns from the rostral ventromedial medulla (RVM). Two physiologically defined cell classes within RVM, "ON-cells" and "OFF-cells," respectively facilitate and inhibit nociceptive transmission. However, sensory pathways through which nociceptive input drives changes in RVM cell activity are only now being defined. We recently showed that indirect inputs from the dorsal horn via the parabrachial complex (PB) convey nociceptive information to RVM. The purpose of the present study was to determine whether there are also direct dorsal horn inputs to RVM pain-modulating neurons. We focused on the trigeminal dorsal horn, which conveys sensory input from the face and head, and used a combination of single-cell recording with optogenetic activation and inhibition of projections to RVM and PB from the trigeminal interpolaris-caudalis transition zone (Vi/Vc) in male and female rats. We determined that a direct projection from ventral Vi/Vc to RVM carries nociceptive information to RVM pain-modulating neurons. This projection included a GABAergic component, which could contribute to nociceptive inhibition of OFF-cells. This approach also revealed a parallel, indirect, relay of trigeminal information to RVM via PB. Activation of the indirect pathway through PB produced a more sustained response in RVM compared with activation of the direct projection from Vi/Vc. These data demonstrate that a direct trigeminal output conveys nociceptive information to RVM pain-modulating neurons with a parallel indirect pathway through the parabrachial complex.SIGNIFICANCE STATEMENT Rostral ventromedial medulla (RVM) pain-modulating neurons respond to noxious stimulation, which implies that they receive input from pain-transmission circuits. However, the traditional view has been that there is no direct input to RVM pain-modulating neurons from the dorsal horn, and that nociceptive information is carried by indirect pathways. Indeed, we recently showed that noxious information can reach RVM pain-modulating neurons via the parabrachial complex (PB). Using in vivo electrophysiology and optogenetics, the present study identified a direct relay of nociceptive information from the trigeminal dorsal horn to physiologically identified pain-modulating neurons in RVM. Combined tracing and electrophysiology data revealed that the direct projection includes GABAergic neurons. Direct and indirect pathways may play distinct functional roles in recruiting pain-modulating neurons.

Keywords: brainstem; descending control; in vivo electrophysiology; pain-modulation; trigeminal.

Figure 1.

Recording sites within RVM for (A) ArchT and (B) ChR2 experiments. Sites were distributed through the rostro-caudal extent of the RVM between −1.32 and −2.90 mm relative to interaural line. SO: superior olive, VII: nucleus of the facial nerve.

Figure 2.

ChR2 and ArchT expression in Vi/Vc and terminal expression in RVM and PB. A, Expression of ChR2 and ArchT in Vi/Vc cell bodies. B, Expression of ChR2 and ArchT in terminals in RVM. C, Expression of ChR2 in PB. Terminals were densely distributed throughout lateral PB complex (DCN = dorsal cochlear nucleus, lPB = lateral parabrachial nucleus, LRt = lateral reticular nucleus, Pr5VL = principal sensory ventrolateral trigeminal nucleus, py = pyramids, scp = superior cerebellar peduncle, sp5 = spinal trigeminal tract, Sp5C = spinal trigeminal nucleus, caudalis, Sp5I = spinal trigeminal nucleus, interpolaris, Sp5O = spinal trigeminal nucleus, oralis, VII = nucleus of the facial nerve). Scale bars are 200 µm unless otherwise noted. D, Trigeminal nucleus neuronal firing induced by light activation of ChR2 (10- to 50-ms pulses at 7, 10, or 28 Hz). Firing of the neuron reliably followed the light trains. E, Suppression of activity of a trigeminal neuron during light-induced activation of ArchT.

Figure 3.

ChR2-induced activation of trigeminal cell bodies mimics noxious stimulation, activating ON-cells and suppressing the firing of OFF-cells. A, Light was delivered to Vi/Vc cell bodies while recording from RVM neurons. B, Vector expression was found throughout the trigeminal complex and grouped by placement (ventral = red, dorsal = blue). Optical fiber placements are also mapped (gray circles). ON-cell (C) OFF-cell (D) and NEUTRAL-cell (E) responses to noxious mechanical stimulation of face and light delivered to Vi/Vc cell bodies. Ratemeter records (1-s bins) with mechanical stimulation of the whisker pad (left, black triangle) and light stimulation (right, blue bar and shading, 30 s) show mechanical-related and optogenetically-evoked responses recorded from RVM. F, Group data show trigeminal cell body activation significantly increased the firing rate of ON-cells (p < 0.0001, t₍₁₃₎ = 9.64, Cohen's d = 1.76), and significantly decreased the firing rate of (G) OFF-cells as compared with prelight firing rates (p < 0.0001, t₍₁₃₎ = 8.55, Cohen's d = 2.35). H, NEUTRAL-cells showed no significant change in firing rate on trigeminal cell body activation (p = 0.59, t₍₇₎ = 0.57). ****p < 0.0001 compared with prelight firing rate, ns: not significant, paired t test. Blue lines = cells from male animals, black lines = cells from female animals. sp/s = spikes per second, dotted lines = cells from animals with virus restricted to dorsal Vi/Vc.

Figure 4.

ChR2-induced activation of RVM terminals arising from ventral Vi/Vc and Vc Lamina V neurons, but not dorsal Vi/Vc, mimics noxious stimulation, activating ON-cells and suppressing the firing of OFF-cells. A, Trigeminal terminals in RVM were activated while recording from RVM neurons. B, Vector expression was found throughout the trigeminal complex and grouped by placement (ventral = red, dorsal = blue). ON- cells (C) and OFF-cells (D) were more likely to respond to light in RVM in animals with vector found in ventral Vi/Vc and Vc V than dorsal Vi/Vc and Vc I–IV. E, G, Ratemeter records (1-s bins) with mechanical stimulation of the whisker pad (left, black triangle) and light stimulation (right, blue bar and shading, 30 s) show mechanical-related and optogenetically-evoked responses recorded from an RVM (E) ON-cell and (G) OFF-cell in animals with vector expression in ventral Vi/Vc. Activation of RVM terminals in animals injected in ventral Vi/Vc and Vc V significantly increased (F) ON-cell firing (p < 0.0001, t₍₂₉₎ = 6.39, Cohen's d = 0.48), and (H) significantly decreased OFF-cell firing (p < 0.0001, t₍₂₅₎ = 4.84, d = 0.69) compared with prelight firing rate. I, K, Ratemeter records (1-s bins) with mechanical stimulation of the whisker pad (left, black triangle) and light stimulation (right, blue bar and shading, 30 s) show mechanical-related and optogenetically-evoked responses recorded from RVM (I) ON-cell and (K) OFF-cell in animals with vector expression in dorsal Vi/Vc. Light activation in animals with viral expression restricted to dorsal Vi/Vc resulted in no significant change on (J) ON-cell (p = 0.20, t₍₁₄₎ = 1.35) or (L) OFF-cell firing (p = 0.89, t₍₁₃₎ = 0.14). ****p < 0.0001 compared with prelight firing rate, paired t tests. Blue lines = cells from male animals, black lines = cells from female animals. sp/s = spikes per second, ns = not significant.

Figure 5.

ArchT-induced inhibition of RVM terminals arising from ventral Vi/Vc attenuates ON-cell and OFF-cell noxious evoked activity. A, Ventral Vi/Vc terminals in RVM were inhibited during noxious mechanical stimulation of the face. Representative examples show (B) ON-cell and (C) OFF-cell activity during noxious mechanical stimulation during uninhibited trials (“Light OFF”), compared with during Arch-T induced inhibition of Vi/Vc terminals in RVM (“Light ON”). D, ON-cell evoked spikes in response to noxious mechanical stimulation were significantly attenuated during terminal inhibition (p = 0.0068, t₍₁₈₎ = 3.06, Cohen's d = 0.21), while (E) ongoing activity was not affected (p = 0.77, t₍₁₈₎ = 0.30). F, The OFF-cell pause in response to noxious mechanical stimulation was significantly attenuated during terminal inhibition (p = 0.0001, t₍₁₇₎ = 4.97, Cohen's d = 0.49), while (G) ongoing activity was not affected (p = 0.13, t₍₁₇₎ = 1.58). **p < 0.01 and ***p < 0.001 compared with Light OFF trials, paired t test. Blue lines = cells from male animals, black lines = cells from female animals. sp/s = spikes per second. ns = not significant.

Figure 6.

Both RVM and PB receive GABAergic projections from ventral Vi/Vc. Representative animal (A–D) and group data (E–H). A, Vi/Vc was injected with AAV9-mGAD65(delE1)-GFP and (i) GFP-labeled GABAergic cells were found in ventral Vi/Vc. Injection of Fluorogold (FG) in RVM (B) and CTb (C) in PB revealed cell bodies in ventral Vi/Vc (ii and iii, respectively). D, GAD65+ Vi/Vc RVM projecting neurons (solid arrow) and PB projecting neurons (empty arrow) were identified (iv). Vi/Vc cells projecting to both RVM and PB could be identified (square). E, Injection sites in RVM and in PB in four animals: two male and two female. F, Anatomical maps showing co-localization of GAD65+ Vi/Vc neurons projecting to RVM, (G) GAD65+ Vi/Vc neurons projecting to PB, (H) and Vi/Vc neurons projecting to both PB and RVM in four animals. LRt = lateral reticular nucleus. Scale bars are 200 µm unless otherwise noted.

Figure 7.

PB neurons are activated or inhibited by Vi/Vc cell body stimulation, while unaffected by RVM terminal activation. ChR2-induced activation of trigeminal terminals in PB mimics noxious stimulation, activating ON-cells and suppressing OFF-cells. A, Vi/Vc cell bodies and terminals in RVM were activated while recording from PB neurons. B, Recording locations in PB and (C) optical fiber placements and vector expression in Vi/Vc were mapped. D, Ratemeter records with mechanical stimulation of the whisker pad (left, black triangle) and light stimulation (right, blue bars and shading, 30 s) in RVM and Vi/Vc show mechanical-related and optogenetically-evoked responses recorded from PB neurons in cells that were (D) activated or (F) inhibited by noxious stimulation and Vi/Vc cell body activation. E, G, PB neurons did not change in activity in response to light in RVM, while activation of Vi/Vc cell bodies either (E) increased or (G) decreased PB neuron activity. H, Vi/Vc terminals in PB were activated while recording from RVM neurons. I, Optic fiber placements and (J) viral expression locations were mapped. K, L, Ratemeter records with mechanical stimulation of the whisker pad (left, black triangle) and light stimulation (right, blue bar and shading, 30 s) in PB show mechanical-related and optogenetically-evoked responses recorded from RVM (K) ON and (L) OFF-cells. M, Activation of terminals in PB significantly increased ON-cell firing (p < 0.0001, t₍₁₄₎ = 6.91, Cohen's d = 1.09), and (N) significantly decreased OFF-cell firing (p = 0.0002, t₍₁₁₎ = 5.45, Cohen's d = 0.82). ***p < 0.001 and ****p < 0.0001 compared with prelight firing rate, paired t test. Blue lines = cells from male animals, black lines = cells from female animals. sp/s = spikes per second.

Figure 8.

RVM cell response timing differs with activation of Vi/Vc terminals in RVM versus PB. A, Representative ON-cell and (B) OFF-cell responses to RVM terminal stimulation versus (C) ON-cell and (D) OFF-cell responses to PB terminal stimulation. Ratemeter records (1-s bins) show optogenetically-evoked responses (blue bar and shading, 30 s) recorded from RVM. Line segment after light termination indicates the length of the after-response. E, ON-cell after-responses were significantly longer with activation of Vi/Vc terminals in PB compared with activation of Vi/Vc terminals in RVM (p = 0.0315, t₍₃₀₎ = 2.26, Cohen's d = 0.53). F, The OFF-cell pause after-response was also longer with activation of terminals in PB (p = 0.0085, t₍₁₈₎ = 2.96, Cohen's d = 1.08). *p < 0.05 and **p < 0.005 compared between RVM and PB terminal activation, unpaired t test. sp/s = spikes per second.