Share this article on:

Wind-up in lamina I spinoparabrachial neurons: a role for reverberatory circuits

Hachisuka, Junichia,b; Omori, Yua,b; Chiang, Michael C.a,b; Gold, Michael S.a,b; Koerber, H. Richarda,b; Ross, Sarah E.a,b,*

doi: 10.1097/j.pain.0000000000001229
Research Paper
Editor's Choice

Wind-up is a frequency-dependent increase in the response of spinal cord neurons, which is believed to underlie temporal summation of nociceptive input. However, whether spinoparabrachial neurons, which likely contribute to the affective component of pain, undergo wind-up was unknown. Here, we addressed this question and investigated the underlying neural circuit. We show that one-fifth of lamina I spinoparabrachial neurons undergo wind-up, and provide evidence that wind-up in these cells is mediated in part by a network of spinal excitatory interneurons that show reverberating activity. These findings provide insight into a polysynaptic circuit of sensory augmentation that may contribute to the wind-up of pain's unpleasantness.

An excitatory network that shows reverberatory activity contributes to wind-up in lamina I spinoparabrachial output neurons.

aDepartment of Neurobiology, University of Pittsburgh, Pittsburgh, PA, United States

bPittsburgh Center for Pain Research, University of Pittsburgh, Pittsburgh, PA, United States. Dr. Omori is now with the Toray Industries, Inc, Pharmaceutical Research Laboratories, Kamakura, Kanagawa, Japan

Corresponding author. Address: 200 Lothrop St., W1456 BST, Department of Neurobiology, Pittsburgh Center for Pain Research, University of Pittsburgh, Pittsburgh, PA 15213. Tel.: +1 412-624-9178; fax: 412-648-1441. E-mail address: (S.E. Ross).

Sponsorships or competing interests that may be relevant to content are disclosed at the end of this article.

H.R. Koerber and S.E. Ross contributed equally to this work.

Received November 06, 2017

Received in revised form January 25, 2018

Accepted February 20, 2018

Back to Top | Article Outline

1. Introduction

Wind-up is a type of facilitation observed in spinal cord neurons in which the response to repetitive stimulation of peripheral C-fibers increases with each stimulus.18 This mechanism of amplification was first described over 50 years ago in lamina IV spinocervical neurons, where it was shown to be a frequency-dependent phenomenon.22,23 Since then, this form of amplification has been observed in a variety of spinal neurons, including wide-dynamic-range (WDR) neurons in the deep dorsal horn, unidentified neurons in the superficial dorsal horn, and motor neurons in the ventral horn.36,49,50

It has long been speculated that wind-up could contribute to the sensory experience of temporal summation, the increase in the perception of pain in response to repeated stimulation.2,8 Consistent with this possibility, temporal summation in response to electrical, thermal, or mechanical stimulation, is frequency dependent, and requires stimulus intensities capable of activating C-fibers.1,3,20,21,45 Both wind-up and temporal summation are physiological mechanisms of amplification that has the potential to contribute to hypersensitivity and allodynia after injury, as suggested by the observations that wind-up is magnified in mice with peripheral inflammation,17,18,40,43 and temporal summation is heightened in patients with pain.12,26,31,44 Thus, understanding the neural circuits of wind-up is important because these circuits may contribute to pathological pain.11,47

Although wind-up has been studied extensively, there remain 2 major unanswered questions. The first pertains to the affective component of pain. Many earlier studies have focused on amplification of reflexes, using wind-up in the ventral horn and/or motor roots as a primary endpoint with which to tease apart mechanisms underlying this phenomenon.30,36,49 Wind-up also occurs in spinal output neurons that project, directly or indirectly to the thalamus, such as WDR spinothalamic neurons50 and spinocervical neurons.22,23 Because somatosensory input to the thalamus contributes to sensory discrimination, wind-up in these cells could account for the psychophysical phenomenon of temporal summation as manifest by reports of increasing pain intensity with repeated application of a noxious stimulus.20 What remains unclear is whether there is wind-up in sensory affective pathways that mediate the unpleasantness of pain. Spinal output neurons that target the lateral parabrachial nucleus, which convey nociceptive and thermoregulatory information, are believed to contribute to the affective component of pain.5,14 However, whether wind-up occurs in spinoparabrachial (SPB) neurons is unknown. Given that pain unpleasantness shows temporal summation,35 we hypothesized that wind-up would occur in SPB neurons.

The second unresolved question is whether synaptic mechanisms alone are sufficient to account for wind-up. The prevailing theory is that wind-up is mediated by N-methyl-D-aspartate (NMDA) receptors, which are recruited through a relief of magnesium block that occurs on cumulative depolarization in response to repetitive stimulation.8 A key role of NMDA receptors in wind-up is supported by the findings that NMDA receptor antagonists inhibit wind-up of WDR neurons,7,9 ventral horn neurons,41 and motor neurons.48 Accordingly, NMDA receptor antagonists also reduce temporal summation in healthy humans2,32 and abnormal temporal summation in the context of injury-induced pain.13,46 However, the block of wind-up by NMDA receptor antagonists is not always complete, raising the possibility that other mechanisms are likely to contribute.7,9,41

One possibility is a circuit-based mechanism in which an excitatory interneuron network in the dorsal horn acts as an amplifier. Indeed, such a circuit-based mechanism was originally put forth by Mendell in his seminal study, in which he postulated that wind-up may be due to the reverberatory activity in spinal interneurons that is evoked by afferent C-fiber input, reasoning that “if in this period of time another stimulation arrives to the cord, it sums with the ongoing activity to produce a more intense discharge in the interneurons than the one before it 22 .” Consistent with this idea, wind-up is observed in a variety of spinal neurons that do not receive appreciable C-fiber input, indicating that this wind-up must be mediated through a polysynaptic circuit. However, the nature of this circuit, and whether it contributes to (rather than simply propagating) wind-up is not clear.

We recently developed an ex vivo somatosensory preparation that enables the recording from lamina I SPB neurons together with cell-type–specific manipulation.16 We therefore set out to test the hypotheses that wind-up occurs in lamina I SPB neurons, and that a network of excitatory interneurons is involved in mediating this amplification. Here, we report that approximately one-fifth of SPB neurons show wind-up on repetitive stimulation of the dorsal root. In addition, we provide evidence that optogenetic activation of excitatory interneurons is sufficient for wind-up. This effect is selective for some excitatory networks because it is observed on activation neurons of the neurotensin-lineage (NtsCre) neurons but not the calretinin lineage (CrCre) neurons. Through optogenetic inhibition, we show that activity in NtsCre neurons is required for dorsal root–evoked wind-up. Finally, our data suggest that this facilitation is mediated, at least in part, by reverberatory activity within an interconnected excitatory network. Together, these studies show the existence of wind-up in lamina I SPB neurons and provide new insight into the underlying neural circuit basis.

Back to Top | Article Outline

2. Methods

2.1. Mouse lines

The mouse lines used for this study were all obtained from The Jackson Laboratories and maintained on C57BL/6J background: NtsCre, a nondisruptive IRES-Cre recombinase knock-in at the endogenous Neurotensin locus (Stock number: 017525); CrCre, a nondisruptive IRES-Cre recombinase knock-in at the endogenous Calretinin locus (Stock number: 010774); Ai9, enabling Cre-dependent expression of tdTomato (Gt(ROSA)26Sor tm9(CAG-tdTomato), Stock number: 007909); Ai32, enabling Cre-dependent expression of an enhanced channelrhodopsin fusion protein, ChR2(H134)/EYFP (Gt (ROSA)26Sortm32(CAG-COP4*H134R/EYFP), Stock number: 012569); and Ai35, enabling Cre-dependent expression of an Archaerhodopsin fusion protein (Gt(ROSA)26Sortm35.1(CAG-aop3/GFP); Stock number: 012735). Genotyping for these alleles was performed with the following primers: for Ai9, TdTR (GGC ATT AAA GCA GCG TAT CC) and TdTR (CTG TTC CTG TAC GGC ATG G) were used to detect TdTomato (196 bp product); for Ai32, ChR2F (ACA TGG TCC TGC TGG AGT TC) and ChR2R (GGC ATT AAA GCA GCG TAT CC) were used to detect ChR2 (212 bp product); for Ai35, XFPF (GCG AGG GCG AGG GCG ATG) and XFPR (CGA TGT TGT GGC GGA TCT TG) were used to detect EYFP (423 bp product), and for the wild-type Rosa allele, RosaWTF (GGA GCG GGA GAA ATG GAT ATG) and RosaWTR (AAA GTC GCT CTG AGT TGT TAT) were used (∼550 bp product).

Four- to seven-week-old mice of both sexes were used in this study. Mice were given free access to food and water and housed under standard laboratory conditions. The use of animals was approved by the Institutional Animal Care and Use Committee of the University of Pittsburgh.

Back to Top | Article Outline

2.2. Immunohistochemistry

Four Nts Cre ; Gt(ROSA)26Sor tm9(CAG-tdTomato) adult mice were perfused with 4% paraformaldehyde and the lumbar spinal cords from L2 to L3 were dissected and subsequently postfixed for 4 hours. Transverse 65-µm thick sections were cut on a vibrating microtome and processed free-floating for immunohistochemistry. Sections were blocked in blocking solution (10% donkey serum, 0.1% Triton in phosphate-buffered saline) for 2 hours and incubated in the following primary antibodies for 14 hours overnight at 4° C: NeuN (1:1000 Millipore MAB377; Millipore, Burlington, MA) and Pax2 (1:1000, Life Technologies 716000; Life Technologies, Carlsbad, CA) as well as Biotin-conjugated Isolectin B4 (1:200, Sigma Aldrich L2140; Sigma Aldrich, St. Louis, MO). After three 20-minute washes with wash buffer (0.1% Triton, 1% donkey serum, 0.3 M NaCl), sections were incubated with Alexa Fluor-conjugated secondary antibodies (1:500; Life Technologies) and Streptavidin-488 (1:500; Thermo Fisher, Waltham, MA) and for 2 hours at room temperature. Next, sections were incubated with Hoechst (1:10,000; Thermo Fisher) incubation for 1 minute to label nuclei. Seven 15-minute washes were performed, and then, sections were mounted on slides and coverslipped. The dorsal horns were imaged through a single optical plane using a Nikon A1R confocal microscope with a 20× objective. Only cells with clearly visible nuclei were counted.

Back to Top | Article Outline

2.3. Labeling spinoparabrachial neurons

Four- to six-week-old mice were anesthetized with isoflurane and placed in a stereotaxic frame. A small hole was made on the skull with dental drill. A glass micropipette was used to inject 100 nL of FAST DiI oil (2.5 mg/mL; Invitrogen, Carlsbad, CA) into left lateral parabrachial nucleus (relative to lambda: anteroposterior −0.5 mm; lateral 1.3 mm; and dorsoventral −2.4 mm). The head wound was closed with stitches. After recovery from the anesthesia, the animals fed and drank normally. The animals were used for electrophysiology 4 to 7 days later.

Back to Top | Article Outline

2.4. Whole spinal cord preparation

For electrophysiological recordings, we used a modified semi-intact preparation.16 We recorded from neurons in the L2 spinal segment, which are easiest to visualize and record from in this preparation. In brief, 5- to 7-week-old mice were deeply anesthetized with urethane (1.2 g/kg, intraperitoneally). The animals were perfused transcardially through the left ventricle with ice-cold oxygenated (95% O2 and 5% CO2) sucrose-based artificial cerebrospinal fluid (in mM; 234 sucrose, 2.5 KCl, 0.5 CaCl2, 10 MgSO4, 1.25 NaH2PO4, 26 NaHCO3, and 11 glucose). Immediately after perfusion, the skin was incised along the dorsal midline and the spinal cord was quickly excised and placed it into an ice-cold, sucrose-based Krebs solution. Dura and pia-arachnoid membrane were removed after cutting all the ventral and dorsal roots except the L2 root on the right. The spinal cord was placed in the recording chamber and pinned into a chamber wall made from Sylguard. The spinal cord was perfused with Krebs solution saturated with 95% O2 and 5% CO2 at 30 to 31°C. The Krebs solution contained (mM): 117 NaCl, 3.6 KCl, 2.5 CaCl2, 1.2 MgCl2, 1.2 NaH2PO4, 25 NaHCO3, and 11 glucose.

Back to Top | Article Outline

2.5. Patch-clamp recording from dorsal horn neurons

Neurons were visualized using a fixed stage upright microscope (BX51WI Olympus microscope; Olympus, Tokyo, Japan) equipped with a 40× water immersion objective lens, a CCD camera (ORCA-ER; Hamamatsu Photonics, Hamamatsu City, Japan), and monitor screen. A narrow beam infrared LED (L850D-06; Marubeni, Tokyo, Japan, emission peak, 850 nm)was positioned outside the solution meniscus, as previously described.16,28,29,34 To record from NtsCre neurons in optogenetic experiments, mice were generated that harbored both Ai14 and Ai32 alleles, and cells were identified by expression of tdTomato. Spinoparabrachial neurons located within 20 μm from the surface of the dorsal horn were identified by DiI labeling.

Whole-cell patch-clamp recordings were made with an Axopatch 200B amplifier with a Digidata 1322A A/D converter controlled using Clampex software (version 10), all from Molecular Devices. Patch-pipette electrodes had a resistance of 6 to 12 MΩ when filled with a pipette solution of the following composition (mM); 135 potassium gluconate, 5 KCl, 0.5 CaCl2, 5 EGTA, 5 Hepes, 5 ATP-Mg, pH 7.2. Alexa Fluor 488 was added to aid in visualization. The data were low-pass filtered at 2 kHz and digitized at 10 kHz. The liquid junction potential was not corrected.

Cell recordings were made in voltage-clamp mode at holding potentials of −70 mV to record excitatory postsynaptic currents (EPSCs) and current-clamp mode to record action potentials (APs). Frequency of EPSCs and APs was analyzed using MiniAnalysis (Synaptosoft, Inc, Decatur, GA). The L2 dorsal root was stimulated by suction electrode with 100 ms duration. Ad-fiber evoked responses were considered monosynaptic if the latency remained constant when the root was stimulated at 20 Hz and there was no failure, and C-fiber evoked responses were considered monosynaptic if there was no failure at 2Hz.25 To investigate wind-up, dorsal root stimulation (DRS) was applied at 2 Hz and the number of APs after each stimulation (0.5-second window) was counted. The criteria used to define the presence of wind-up were that the maximum number of APs was at least 5 and more than twice as many APs were evoked in response to subsequent stimuli as were evoked in response to the first stimulus.

Back to Top | Article Outline

2.6. Optogenetic activation

During patch-clamp recording, photostimulation was applied to the spinal cord through the objective lens (40×) of the microscope with a Xenon lamp (Lambda DG-4; Sutter Instrument, Novato, CA). A Lamdba DG4 (Sutter Instrument) was used for optogenetic stimulation, where switching between filter positions (0.5 ms) was controlled by a TTL pulse from the output of the A/D converter. We used a GFP filter (centered around 485 nm) for activation of ChR2 and a Cy3 filter (centered around 555 nm) for activation of Arch. Light power on the sample was 1.3 mW/mm2. To examine whether the recorded neuron received monosynaptic or polysynaptic input from NtsCre neurons, we applied 0.1-Hz photostimulation (5 ms). Input was considered monosynaptic if there was no failure and the latency jitter was smaller than 1 ms.16 For blue light–induced wind-up, we applied 2-Hz photostimulation (5 ms), using the same wind-up criteria as root-evoked wind-up. To test whether activation of NtsCre primary afferents caused wind-up, blue light pulse of the same light source was applied on the dorsal root, which was approximately 7 mm away from the spinal cord and far enough to prevent blue light–induced ChR2 activation in the spinal cord.

Back to Top | Article Outline

3. Results

3.1. Dorsal root stimulation induces wind-up in one-fifth of lamina I spinoparabrachial neurons

To determine whether wind-up develops in SPB neurons, we used an ex vivo spinal cord preparation that preserves intact spinal circuitry and enables whole-cell recordings from SPB neurons in lamina I (Figs. 1A and B), as described previously.16 After the injection of DiI into the lateral parabrachial nucleus, retrogradely labeled lamina I SPB neurons were identified for recording with epifluorescence and then visualized in bright field by oblique infrared LED illumination to establish whole-cell patch recording (Fig. 1C).16,39

Figure 1

Figure 1

Electrical stimulation of L2 dorsal root at 0.5 Hz, elicited a stable response to each stimulus in most cases (Fig. 1D, bottom and Fig. 1E, black trace). However, at 2 Hz stimulation, there was a progressive increase in AP number across the period of stimulation in 12 of 67 lamina I SBP neurons studied (Fig. 1D, top and Fig. 1E, red trace). In these neurons, there was a 299 ± 70% increase in the maximum number of evoked APs by the end of the 10-pulse train relative to the number of APs evoked with the first pulse. This effect was in marked contrast to the absence of a change in the number of evoked APs (6 ± 11%) in the remaining neurons (Fig. 1F). Importantly, and consistent with previous characterizations of wind-up, the increase in the evoked response to DRS significantly reduced by the NMDA receptor antagonist APV (50 μM; Fig. 1G). Thus, wind-up is observed in ∼20% of lamina I SPB neurons and, like in other spinal neurons, it is NMDA receptor dependent.

Back to Top | Article Outline

3.2. Neurotensin-lineage neurons are sufficient for wind-up in lamina I spinoparabrachial neurons

To identify spinal neurons potentially underlying the reverberating circuit hypothesized by Mendell,22 we screened for Cre alleles that would enable genetic access to subsets of excitatory neurons. The NtsCre allele was selected because it is relatively specific for excitatory interneurons and it targets a broad population in the dorsal horn. To determine the distribution of these cells, we analyzed L2 to L3 spinal dorsal horns from mice harboring both the NtsCre and the Ai14 tdTomato reporter alleles, which were costained for NeuN, a marker of neurons, Pax2, a marker of inhibitory neurons, and IB4, a marker of nonpeptidergic afferents that were used to help delineate boundaries within the dorsal horn. Using the ventral aspect of IB4 as a boundary that corresponds approximately to the border between high-threshold C-fiber and low-threshold A-fiber inputs, we found that NtsCre neurons represent 13 ± 1% of neurons within the superficial dorsal horn and 32 ± 2% of neurons within the intermediate dorsal horn (n = 4 mice). Almost all (98 ± 1%) the NtsCre neurons were deemed to be excitatory as evidenced by the absence of Pax2 staining (Fig. 2A).6,30 Because the genetic population defined by NtsCre-mediated recombination is somewhat broader than that defined by neurotensin protein expression in adult mice,15 we refer to the NtsCre population as neurotensin-lineage neurons.

Figure 2

Figure 2

Because we wanted to know whether stimulation of NtsCre neurons alone was sufficient to drive wind-up in lamina I SPB neurons, it was first necessary to confirm that we could consistently activate NtsCre neurons expressing ChR2 with blue light (Fig. 2B). We found that brief (5 ms) blue light exposure typically induced 1 AP in ChR2-expressing NtsCre neurons, which could follow at 2 Hz (Fig. 2C) with a failure rate of less than 5% (4 ± 3%, n = 16 cells).

Having confirmed that it was possible to selectively activate NtsCre neurons, we next sought to determine whether selective activation of these neurons was sufficient to drive wind-up in lamina I (Fig. 2D). For the sake of ease in this preliminary screen, we analyzed unidentified lamina I neurons (rather than retrogradely labeled SPB neurons). Optogenetic stimulation at 2 Hz caused a robust wind-up in 6 of 9 neurons (Figs. 2E and F), indicating that NtsCre interneuron activity is sufficient for wind-up. To determine whether optogenetically induced wind-up by excitatory neurons in the dorsal horn occurs in response to the stimulation of any excitatory neuron subpopulation, we assessed the impact of optogenetic stimulation of CrCre neurons (Fig. 2G). Calretinin is expressed in ∼30% of neurons in the superficial dorsal horn, of which 85% are excitatory.37,38 Although optogenetically evoked APs were observed in lamina I neurons on application of blue light, wind-up was not detected on activation of CrCre neurons (Figs. 2H and I). Thus, the selective activation of some, but not all, excitatory neurons is sufficient to induce wind-up in lamina I neurons.

Next, we turned to the analysis of lamina I SPB neurons because these output neurons may be involved in the affective component of pain. Notably, although some of the input from primary afferents onto lamina I SPB neurons is direct, the majority occurs through polysynaptic connections (Figs. 3A and B), raising the possibility that an excitatory network could contribute to wind-up in lamina I SPB neurons. To test this possibility, we examined the effect of optogenetic simulation of NtsCre neurons (Fig. 3C). Blue light stimulation of NtsCre neurons at 2 Hz caused wind-up in 11 of 18 (61%) lamina I SPB neurons (Figs. 3D and E). Thus, optogenetically induced wind-up was observed in a significantly larger population of lamina I SPB neurons than dorsal root–evoked wind-up (P < 0.01, Fisher exact test). In some neurons, the peak response to the wind-up protocol was achieved by the sixth or seventh stimulus in the train, with a decrease from this peak response observed to the remaining stimuli (Fig. 3D). We suggest that this decrease may be due to a depolarization-induced inactivation of voltage-gated sodium channels. Like dorsal root–evoked wind-up, the NtsCre-mediated wind-up of lamina I SPB neurons was blocked by APV (Fig. 3F) and showed frequency dependence (Figs. 3G and H).

Figure 3

Figure 3

An important consideration is that, in addition to excitatory spinal interneurons, the NtsCre allele causes recombination in approximately 10% of primary afferents, which are mainly small diameter cells, and include both peptidergic and nonpeptidergic subtypes (data not shown). To address the potential role of NtsCre primary afferents in optogenetically induced wind-up, we compared the responses of a given lamina I SPB neuron to optogenetic stimulation over the cord (Fig. 3C) with that observed over the root (Fig. 3I). As before, optogenetic stimulation over the cord induced wind-up (Fig. 3K). By contrast, optogenetic stimulation of the NtsCre afferent input was not sufficient for wind-up in lamina I SPB neurons, although in all cases it was sufficient to drive APs in these cells (Figs. 3J and K). Thus, although we cannot rule out a possible contribution from primary afferents, wind-up in lamina I SPB neurons by optogenetic activation of the NtsCre population likely requires activity in neurotensin-lineage neurons in the dorsal horn.

Back to Top | Article Outline

3.3. Neurotensin-lineage neurons are required for wind-up in lamina I spinoparabrachial neurons

Although activation of the local NtsCre network is sufficient for wind-up in lamina I SPB neurons, whether the NtsCre network normally mediates wind-up that is observed on electrical stimulation of C-fibers remained unclear. To address this question, we examined whether dorsal root–evoked wind-up is abolished on optogenetic inhibition of neurotensin-lineage neurons with Archaerhodopsin (Arch), a light-driven proton pump. To determine whether green light–induced activation of Arch was sufficient to inhibit NtsCre neurons, we recorded from these neurons in voltage and current clamp (Fig. 4A). Green light hyperpolarized Arch-expressing NtsCre neurons (10.3 ± 1.4 mV, n = 13 cells) and blocked dorsal root–evoked APs in these cells (Figs. 4B and C), thereby confirming the efficacy of this optogenetic strategy. Next, we addressed whether optogenetic inhibition of NtsCre neurons blocked root-evoked wind-up in lamina I SPB neurons. Towards this end, we identified lamina I SPB neurons that showed wind-up on electrical stimulation of the dorsal root and then, we repeated the stimulation in the presence of green light to inhibit NtsCre neurons (Fig. 4D). We found that optogenetic inhibition of NtsCre neurons significantly reduced root-evoked wind-up (Figs. 4E and F). To gain insight into the underlying mechanism, we performed recordings in voltage clamp to analyze the effect of green light on the input received by lamina I SPB neurons. This analysis revealed that optogenetic inhibition of NtsCre neurons significantly reduced the excitatory input that is observed on dorsal root stimulation, as measured by the net influx of charge (Figs. 4G and H). Taken together, these findings suggest that activity in NtsCre neurons is required for dorsal root–evoked wind-up.

Figure 4

Figure 4

Back to Top | Article Outline

3.4. NtsCre neurons form an extensive excitatory network

Our results suggested that NtsCre neurons are necessary and sufficient for wind-up, but the underlying circuitry remained unclear. To address this, we analyzed ChR2-evoked currents in voltage-clamp mode. We found that, although a small proportion (2 of 30) of lamina I SPB neurons did not receive any input from NtsCre neurons (Fig. 5A), most showed either monosynaptic and polysynaptic EPSCs (16 of 30; Fig. 5B) or polysynaptic EPSCs (12 of 30; Fig. 5C). Thus, most lamina I SPB neurons receive direct or indirect input from NtsCre neurons (Fig. 5D). Next, we examined the connectivity among NtsCre neurons, recording from NtsCre neurons while activating this population optogenetically. In 3 of 25 NtsCre neurons, optogenetic stimulation evoked only a single, large amplitude ChR2-evoked current with very short latency (∼1 ms) because of the opening of ChR2 in the recorded cell (Fig. 5E). However, in most NtsCre neurons, this large amplitude ChR2-evoked current was accompanied by monosynaptic and polysynaptic EPSCs (Fig. 5F), or polysynaptic EPSCs (Fig. 5G). Together, these findings suggest that NtsCre neurons form an excitatory network.

Figure 5

Figure 5

Back to Top | Article Outline

3.5. Activation of NtsCre neurons causes reverberatory activity that correlates with wind-up

The extensive interconnectivity that we observed within the NtsCre network raised the possibility that a polysynaptic circuit also contributes to the wind-up in lamina I SPB neurons. If so, we reasoned that we ought to see not only wind-up of the light-evoked EPSCs in lamina I SPB neurons but also an increase in the subsequent polysynaptic input onto these cells. To investigate this possibility, we recorded from lamina I SPB neurons in voltage-clamp mode in response to optogenetic activation of NtsCre neurons (Fig. 6A). Blue light stimulation at 2 Hz resulted in an increase in both the blue light–evoked EPSC and smaller EPSCs that followed, which seemed to be due to ongoing synaptic input onto the SPB neuron (Fig. 6B). To quantify this effect, we measured the net charge influx between stimuli. This analysis revealed that the net charge increase observed on repeated stimulation was significantly greater in cells that showed wind-up compared with cells that did not (Fig. 6C).

Figure 6

Figure 6

One of the factors that could lead to increased net charge influx is reverberatory activity within the network. We found that brief (5 ms) blue light stimulation induced EPSCs (Fig. 6D) that continued for the prolonged period (eg, >1 second). Given the extended time course of this reverberatory activity, we reasoned that a second stimulus during this time could sum with previous activity, thereby contributing to wind-up as previously suggested.22 To test this idea, we compared the reverberating activity evoked by a single stimulus between lamina I SPB neurons that subsequently showed wind-up to those that did not. In this retrospective analysis, we found that in lamina I SPB neurons where no wind-up was observed, there was no difference in the EPSC frequency during the 1000 to 500 ms before optogenetic stimulation compared with the EPSC frequency during the 500 to 1000 ms after optogenetic stimulation. By contrast, in lamina I SPB neurons where wind-up was observed, there was a significant increase in the EPSC frequency observed after optogenetic stimulation compared with before (Figs. 6E and F). Together, these findings suggest that wind-up in lamina I SPB neurons is mediated, at least in part, by a polysynaptic circuit that shows reverberatory activity (Fig. 6G).

Back to Top | Article Outline

4. Discussion

Wind-up is a mechanism of facilitation that is believed to account for the intensification of pain observed with repeated or prolonged application of noxious stimuli, a psychophysical phenomenon referred to as temporal summation. Importantly, both the sensory/discriminative (intensity) and the emotional/affective (unpleasantness) components of pain are subject to temporal summation. Spinothalamic neurons, which are believed to contribute to intensity coding, undergo wind-up, but whether SPB neurons, which are believed to be contribute to unpleasantness coding, are subject to wind-up, was less clear. Our study reveals that lamina I SPB neurons show wind-up in response to repetitive stimulation of the dorsal root. In addition, we provide evidence that this wind-up is mediated, at least in part, by NtsCre neurons because selective activation of these cells was sufficient to evoke wind-up in lamina I SPB neurons, whereas optogenetic inhibition of these cells inhibited dorsal root–evoked wind-up. Finally, our data suggest that the duration of reverberatory input onto lamina I SPB neurons contributes to the manifestation of wind-up. These findings provide new insight into the circuit mechanisms that could underlie the temporal summation of pain's unpleasantness, and uncover a long-speculated contribution of an excitatory network in this sensory augmentation (Fig. 6E).

Our study reveals that wind-up in lamina I SPB neurons on optogenetic stimulation of NtsCre neurons was more frequent in occurrence than that observed on electrical stimulation of the dorsal root. Although the reason for this difference is not known, we suggest that the key role of inhibitory neurons within nociceptive circuits of the dorsal horn is likely to account for the difference. That is, primary afferent input normally recruits activity in inhibitory interneurons to effect feed-forward, feedback, and lateral inhibition. By contrast, our optogenetic stimulation of NtsCre excitatory interneurons may have bypassed this inhibition, thereby shifting the balance of activity in favor of excitation, leading to an increase in the prevalence of wind-up.

Our data suggest that reverberatory activity within a neural circuit contributes to wind-up lamina I SPB neurons. It has long been speculated that interconnectivity among excitatory interneurons may provide a substrate for reverberatory activity, allowing for summation.22 Under physiological conditions, this summation is limited by activity in inhibitory interneurons, as evidenced by the observation that disinhibition by γ-aminobutyric acid and glycine antagonists causes long-lasting EPSCs by DRS.4,10,27,42 Thus, inhibition may normally limit wind-up. Because a decrease in inhibitory tone is frequently seen on injury,19,24 this loss of inhibition may contribute to abnormal reverberatory activity, giving rise to both heightened wind-up and abnormally elevated wind-up that are observed in pathological conditions.18

Back to Top | Article Outline

Conflict of interest statement

The authors have no conflict of interest to declare.

Research reported in this publication was supported by the National Institute of Arthritis and Musculoskeletal and Skin Diseases of the National Institutes of Health under Award Number R01 AR063772 and R21 AR064445 to S.E. Ross; the National Institute of Neurological Disorder and Stroke under Award Number NS023725 and NS096705 to H.R. Koerber, NS083347 to M.S. Gold, and NS096860 to M.M. Chiang; and the National Institutes of Diabetes and Digestive and Kidney Diseases under Award Number DK107966 to M.S. Gold. Part of this work was supported by a grant from the Rita Allen Foundation to S.E. Ross who is a Rita Allen Foundation Pain Scholar.

Back to Top | Article Outline


[1]. Andrew D, Greenspan JD. Peripheral coding of tonic mechanical cutaneous pain: comparison of nociceptor activity in rat and human psychophysics. J Neurophysiol 1999;82:2641–8.
[2]. Arendt-Nielsen L, Petersen-Felix S. Wind-up and neuroplasticity: is there a correlation to clinical pain? Eur J Anaesthesiol Suppl 1995;10:1–7.
[3]. Arendt-Nielsen L, Petersen-Felix S, Fischer M, Bak P, Bjerring P, Zbinden AM. The effect of N-methyl-D-aspartate antagonist (ketamine) on single and repeated nociceptive stimuli: a placebo-controlled experimental human study. Anesth Analg 1995;81:63–8.
[4]. Baba H, Ji RR, Kohno T, Moore KA, Ataka T, Wakai A, Okamoto M, Woolf CJ. Removal of GABAergic inhibition facilitates polysynaptic A fiber-mediated excitatory transmission to the superficial spinal dorsal horn. Mol Cell Neurosci 2003;24:818–30.
[5]. Bourgeais L, Gauriau C, Monconduit L, Villanueva L, Bernard JF. Dendritic domains of nociceptive-responsive parabrachial neurons match terminal fields of lamina I neurons in the rat. J Comp Neurol 2003;464:238–56.
[6]. Cheng L, Arata A, Mizuguchi R, Qian Y, Karunaratne A, Gray PA, Arata S, Shirasawa S, Bouchard M, Luo P, Chen CL, Busslinger M, Goulding M, Onimaru H, Ma Q. Tlx3 and Tlx1 are post-mitotic selector genes determining glutamatergic over GABAergic cell fates. Nat Neurosci 2004;7:510–17.
[7]. Davies SN, Lodge D. Evidence for involvement of N-methylaspartate receptors in “wind-up” of class 2 neurones in the dorsal horn of the rat. Brain Res 1987;424:402–6.
[8]. Dickenson AH. A cure for wind up: NMDA receptor antagonists as potential analgesics. Trends Pharmacol Sci 1990;11:307–9.
[9]. Dickenson AH, Sullivan AF. Evidence for a role of the NMDA receptor in the frequency dependent potentiation of deep rat dorsal horn nociceptive neurones following C fibre stimulation. Neuropharmacology 1987;26:1235–8.
[10]. Duan B, Cheng L, Bourane S, Britz O, Padilla C, Garcia-Campmany L, Krashes M, Knowlton W, Velasquez T, Ren X, Ross S, Lowell BB, Wang Y, Goulding M, Ma Q. Identification of spinal circuits transmitting and gating mechanical pain. Cell 2014;159:1417–32.
[11]. Eide PK. Wind-up and the NMDA receptor complex from a clinical perspective. Eur J Pain 2000;4:5–15.
[12]. Eide PK, Jorum E, Stenehjem AE. Somatosensory findings in patients with spinal cord injury and central dysaesthesia pain. J Neurol Neurosurg Psychiatry 1996;60:411–15.
[13]. Eide PK, Jorum E, Stubhaug A, Bremnes J, Breivik H. Relief of post-herpetic neuralgia with the N-methyl-D-aspartic acid receptor antagonist ketamine: a double-blind, cross-over comparison with morphine and placebo. PAIN 1994;58:347–54.
[14]. Gauriau C, Bernard JF. Pain pathways and parabrachial circuits in the rat. Exp Physiol 2002;87:251–8.
[15]. Gutierrez-Mecinas M, Furuta T, Watanabe M, Todd AJ. A quantitative study of neurochemically defined excitatory interneuron populations in laminae I-III of the mouse spinal cord. Mol Pain 2016;12:1744806916629065.
[16]. Hachisuka J, Baumbauer KM, Omori Y, Snyder LM, Koerber HR, Ross SE. Semi-intact ex vivo approach to investigate spinal somatosensory circuits. Elife 2016;5:e22866.
[17]. Hedo G, Laird JM, Lopez-Garcia JA. Time-course of spinal sensitization following carrageenan-induced inflammation in the young rat: a comparative electrophysiological and behavioural study in vitro and in vivo. Neuroscience 1999;92:309–18.
[18]. Herrero JF, Laird JM, Lopez-Garcia JA. Wind-up of spinal cord neurones and pain sensation: much ado about something? Prog Neurobiol 2000;61:169–203.
[19]. Koerber HR, Brown PB. Quantitative analysis of dorsal horn cell receptive fields following limited deafferentation. J Neurophysiol 1995;74:2065–76.
[20]. Koltzenburg M, Handwerker HO. Differential ability of human cutaneous nociceptors to signal mechanical pain and to produce vasodilatation. J Neurosci 1994;14:1756–65.
[21]. Lundberg LE, Jorum E, Holm E, Torebjork HE. Intra-neural electrical stimulation of cutaneous nociceptive fibres in humans: effects of different pulse patterns on magnitude of pain. Acta Physiol Scand 1992;146:41–8.
[22]. Mendell LM. Physiological properties of unmyelinated fiber projection to the spinal cord. Exp Neurol 1966;16:316–32.
[23]. Mendell LM, Wall PD. Responses of single dorsal cord cells to peripheral cutaneous unmyelinated fibres. Nature 1965;206:97–9.
[24]. Moore KA, Kohno T, Karchewski LA, Scholz J, Baba H, Woolf CJ. Partial peripheral nerve injury promotes a selective loss of GABAergic inhibition in the superficial dorsal horn of the spinal cord. J Neurosci 2002;22:6724–31.
[25]. Nakatsuka T, Park JS, Kumamoto E, Tamaki T, Yoshimura M. (1999). Plastic changes in sensory inputs to rat substantia gelatinosa neurons following peripheral inflammation. Pain 1999;82:39–47.
[26]. Nikolajsen L, Hansen CL, Nielsen J, Keller J, Arendt-Nielsen L, Jensen TS. The effect of ketamine on phantom pain: a central neuropathic disorder maintained by peripheral input. PAIN 1996;67:69–77.
[27]. Peirs C, Williams SP, Zhao X, Walsh CE, Gedeon JY, Cagle NE, Goldring AC, Hioki H, Liu Z, Marell PS, Seal RP. Dorsal horn circuits for persistent mechanical pain. Neuron 2015;87:797–812.
[28]. Pinto V, Szucs P, Derkach VA, Safronov BV. Monosynaptic convergence of C- and Adelta-afferent fibres from different segmental dorsal roots on to single substantia gelatinosa neurones in the rat spinal cord. J Physiol 2008;586: 4165–77.
[29]. Pinto V, Szucs P, D. Lima D, Safronov BV. Multisegmental A{delta}- and C-fiber input to neurons in lamina I and the lateral spinal nucleus. J Neurosci 2010;30: 2384–95.
[30]. Price DD. Characteristics of second pain and flexion reflexes indicative of prolonged central summation. Exp Neurol 1972;37:371–87.
[31]. Price DD, Long S, Huitt C. Sensory testing of pathophysiological mechanisms of pain in patients with reflex sympathetic dystrophy. PAIN 1992;49:163–73.
[32]. Price DD, Mao J, Frenk H, Mayer DJ. The N-methyl-D-aspartate receptor antagonist dextromethorphan selectively reduces temporal summation of second pain in man. PAIN 1994;59:165–74.
[33]. Punnakkal P, von Schoultz C, Haenraets K, Wildner H, Zeilhofer HU. Morphological, biophysical and synaptic properties of glutamatergic neurons of the mouse spinal dorsal horn. J Physiol 2014;592:759–76.
    [34]. Safronov BV, Pinto V, Derkach VA. High-resolution single-cell imaging for functional studies in the whole brain and spinal cord and thick tissue blocks using light-emitting diode illumination. J Neurosci Methods 2007;164: 292–98.
    [35]. Sarlani E, Grace EG, Reynolds MA, Greenspan JD. Sex differences in temporal summation of pain and after sensations following repetitive noxious mechanical stimulation. PAIN 2004;109:115–23.
    [36]. Schouenborg J, Sjolund BH. Activity evoked by A- and C-afferent fibers in rat dorsal horn neurons and its relation to a flexion reflex. J Neurophysiol 1983;50:1108–21.
    [37]. Smith KM, Boyle KA, Madden JF, Dickinson SA, Jobling P, Callister RJ, Hughes DI,Graham BA. Functional heterogeneity of calretinin-expressing neurons in the mouse superficial dorsal horn: implications for spinal pain processing. J Physiol 2015;593:4319–39.
    [38]. Smith KM, Boyle KA, Mustapa M, Jobling P, Callister RJ, Hughes DI, Graham BA. Distinct forms of synaptic inhibition and neuromodulation regulate calretinin-positive neuron excitability in the spinal cord dorsal horn. Neuroscience 2016;326:10–21.
    [39]. Szucs P, Pinto V, Safronov BV. (2009). Advanced technique of infrared LED imaging of unstained cells and intracellular structures in isolated spinal cord, brainstem, ganglia and cerebellum. J Neurosci Methods 2009;177: 369–80.
    [40]. Thompson SW, Dray A, McCarson KE, Krause JE, Urban L. Nerve growth factor induces mechanical allodynia associated with novel A fibre-evoked spinal reflex activity and enhanced neurokinin-1 receptor activation in the rat. PAIN 1995;62:219–31.
    [41]. Thompson SW, King AE, Woolf CJ. Activity-dependent changes in rat ventral horn neurons in vitro; summation of prolonged afferent evoked postsynaptic depolarizations produce a d-2-amino-5-phosphonovaleric acid sensitive windup. Eur J Neurosci 1990;2:638–49.
    [42]. Torsney C, MacDermott AB. Disinhibition opens the gate to pathological pain signaling in superficial neurokinin 1 receptor-expressing neurons in rat spinal cord. J Neurosci 2006;26:1833–43.
    [43]. Traub RJ. Spinal modulation of the induction of central sensitization. Brain Res 1997;778:34–42.
    [44]. Vase L, Svensson P, Nikolajsen L, Arendt-Nielsen L, Jensen TS. The effects of menthol on cold allodynia and wind-up-like pain in upper limb amputees with different levels of phantom limb pain. Neurosci Lett 2013;534:52–7.
    [45]. Vierck CJ Jr, Cannon RL, Fry G, Maixner W, Whitsel BL. Characteristics of temporal summation of second pain sensations elicited by brief contact of glabrous skin by a preheated thermode. J Neurophysiol 1997;78:992–1002.
    [46]. Warncke T, Stubhaug A, Jorum E. Ketamine, an NMDA receptor antagonist, suppresses spatial and temporal properties of burn-induced secondary hyperalgesia in man: a double-blind, cross-over comparison with morphine and placebo. PAIN 1997;72:99–106.
    [47]. Woolf CJ. Windup and central sensitization are not equivalent. PAIN 1996;66:105–8.
    [48]. Woolf CJ, Thompson SW. The induction and maintenance of central sensitization is dependent on N-methyl-D-aspartic acid receptor activation; implications for the treatment of post-injury pain hypersensitivity states. PAIN 1991;44:293–9.
    [49]. Woolf CJ, Wall PD. Relative effectiveness of C primary afferent fibers of different origins in evoking a prolonged facilitation of the flexor reflex in the rat. J Neurosci 1986;6:1433–42.
    [50]. Zhang DX, Owens CM, Willis WD. Intracellular study of electrophysiological features of primate spinothalamic tract neurons and their responses to afferent inputs. J Neurophysiol 1991;65:1554–66.

    Wind-up; Nociception; Neurotensin; Optogenetics; Spinal cord; Dorsal horn

    © 2018 International Association for the Study of Pain