Neuropathic pain changes the output of rat lamina I spino-parabrachial neurons

Highlights

• •Intact spinal cord preparation allows to study mechanisms of chronic pain.
• •Spared nerve injury specifically amplifies the output of projection neurons.
• •The amplified output may underlie allodynia, hyperalgesia, and spontaneous pain.

Abstract

• Spared nerve injury (SNI) altered the action potential (AP) output of lamina I spino-parabrachial neurons (SPNs) without affecting their resting potential or membrane resistance.

• In one-third of SPNs, high-threshold dorsal root stimulation elicited persistent AP firing which was never observed in cells from naïve animals.

• 38% of SPNs from SNI rats showed spontaneous persistent AP firing.

• After SNI low- and high-output SPNs were no longer nociceptive-specific as part of them responded with APs to low-threshold stimulation.

• These SNI-induced changes of SPN output might represent cellular mechanisms for neuropathy-associated allodynia, hyperalgesia, and spontaneous pain.

Neuropathic pain affects up to 17% of the population and negatively impacts the quality of life [1]. This debilitating condition occurs after a primary lesion or disease of the somatosensory nervous system and is characterized by hyperalgesia (increased response to noxious stimuli), allodynia (abnormal response to innocuous stimuli), and spontaneous pain [1]. Available treatments are limited, and one of the main reasons for that is our incomplete understanding of cellular and molecular mechanisms underlying neuropathic pain.

The sensation of pain is mediated by the whole neuraxis, including the peripheral nervous system, the spinal cord, and the brain. Among various classes of cells involved, spinal projection neurons are particularly important since they are the only ones to transmit sensory information from the spinal cord to supraspinal centers [2]. Projection neurons responsible for nociception are predominantly located in lamina I and mostly target parabrachial area of the brainstem in rodents [2], being spino-parabrachial neurons (SPNs). In vivo extracellular recordings have demonstrated that most lamina I SPNs are nociceptive-specific generating action potentials (APs) in response to noxious heat, cold and mechanical stimuli [3]. Patch clamp experiments have revealed that lamina I projection neurons are functionally heterogeneous [4,5]. In terms of their response to afferent stimulation, lamina I SPNs output could be classified into three groups – low- (LO-SPNs), medium- (MO-SPNs), and high-output (HO-SPNs) neurons [4]. Among those, LO- and HO-SPNs were nociceptive-specific, but their distinct output patterns implied different functional roles. LO-SPNs typically responded to high-threshold dorsal stimulation with a single AP, whereas HO-SPNs encoded stimulation intensity with an increasing number of spikes [4]. Moreover, HO-SPNs showed the ability to integrate the input from both nociceptive afferents and spinal interneurons [4]. Despite recent advances in understanding SPN physiology, little information on neuropathy-induced changes of SPN functioning is available. Yet, such research could elucidate the basic mechanisms of neuropathic pain. Therefore, we sought to find out whether and how peripheral neuropathy alters the output characteristics of lamina I SPNs.

We chose spared nerve injury (SNI) as a model of peripheral neuropathy. The procedure followed the standard one [6] except for one major modification: to produce the most vigorous allodynia and spontaneous pain [7], tibial and sural branches of the sciatic nerve were axotomized and ligated while common peroneal branch was left intact (Fig. 1Aa). SNI was performed on P19–20 rats. Seven days after the SNI onset, Hargreaves plantar test showed no changes in thermal sensitivity: paw withdrawal latency was 15.4 ± 0.5 s for contralateral and 15.9 ± 0.5 s for ipsilateral paw (p = 0.27, n = 9, paired Student's t-test, data not shown). At the same time, von Frey filaments test demonstrated profound mechanical hypersensitivity: paw withdrawal threshold was 5.9 ± 1.0 g for contra- and 3.3 ± 0.6 g for ipsilateral paw (p < 0.001, n = 9, paired Student's t-test, Fig. 1 Ab). All tested SNI animals also exhibited spontaneous pain presented as twitching of the ipsilateral hindlimb that lasted for several seconds and was often followed by paw licking.

Fig. 1.

Output characteristics of rat lamina I spino-parabrachial (SPNs) neurons after spared nerve injury (SNI). Aa. Scheme of the SNI. Common peroneal branch of the sciatic nerve was spared, tibial and sural branches were axotomized and ligated. Ab. Paw withdrawal thresholds (assessed by von Frey filaments test) of contra- and ipsilateral hindlimbs 7 days after SNI. Data presented as mean ± SEM. *** - p < 0.001 (paired Student's t-test, n = 9). Ba. Experimental design. Patch-clamp recordings (whole-cell configuration) were acquired from the retrogradely-labelled lamina I SPNs in ex-vivo intact spinal cord preparation (L4-L5 segments). The corresponding dorsal root was stimulated via the suction electrode. Stim – stimulating electrode, rec – recording pipette. Bb. Top: fluorescent image of retrogradely labelled lamina I SPNs. Bottom: same neuron in infrared LED oblique illumination. Ca-f. Characteristic outputs of lamina I SPNs from SNI animals in response to dorsal root stimulation. The strength of the current pulse is indicated near the corresponding trace; open and filled arrows indicate 50 µs and 1 ms pulses, respectively. Note that SNI resulted in the emergence of neuropathy-specific persistent firing output (Ce-f). Cg. Incidence of each type of output in SNI and naïve rats. MO-SPNs – medium-output SPNs (observed only in control conditions [4]). Note that the patterns of SPN output differed significantly between the control and SNI. *** - p < 0.001 (Fisher's exact test). Da. Current clamp recording showing spontaneous switches to persistent firing state. Db. Incidence of spontaneous persistent firing in SPNs from naïve and SNI rats. *** - p < 0.001 (Fisher's exact test).

SPN retrograde labeling was achieved by injecting 200 nl of fluorescent dye (2% aminostilbamidine, Thermo Fisher Scientific, USA) into the parabrachial area contralateral to the injury (stereotaxic coordinates with respect to bregma: 13.2 mm rostrocaudal, 2 mm mediolateral and 5.6 mm dorsoventral, angle 30°) [4] performed immediately after the SNI.

Electrophysiological recordings were conducted 7–10 days after the SNI (P27–30 rats, age-matched naïve animals served as a control) as described previously [4]. Experiments were carried out on ex-vivo intact spinal cord preparation with attached dorsal root; lamina I cells were visualized with oblique infrared LED illumination [8] (Fig. 1Ba-b). SPNs were identified based on fluorescent signal [4]. Patch clamp recordings were obtained from L4-L5 SPNs with 3–5 MΩ borosilicate glass pipettes pulled by P-97 puller (Sutter Instruments, USA) and filled with intracellular solution containing (in mM): 145 K-gluconate, 2.5 MgCl₂, 10 HEPES, 2 Na₂-ATP, 0.5 Na-GTP and 0.5 EGTA (pH 7.3). The extracellular solution had the following composition (in mM): NaCl 125, KCl 2.5, CaCl₂ 2, MgCl₂ 1, NaH₂PO₄ 1.25, NaHCO₃ 26 and glucose 10 (pH 7.4 when bubbled with 95% O₂ and 5% CO₂). Signals were acquired and Bessel filtered at 2.6 kHz with MultiClamp 700B amplifier and digitized at 10 kHz with Digidata 1320 A under the control of pClamp 9 software (all Molecular Devices, USA). Offset potential was compensated before seal formation. Liquid junction potential was not compensated. L4-L5 dorsal roots were stimulated via a suction electrode using ISO-Flex stimulator (AMPI, Israel). Square 50 µs pulses of current were delivered to recruit Aβ/Aδ-fibers; 10–50 µA intensities were considered low-threshold, 60–100 µA stimuli represented Aδ nociceptive range. A 1 ms pulse (30–140 µA) activated all types of fibers, including high-threshold A- and C-afferents.

Whole-cell current clamp recordings coupled with dorsal root stimulation revealed that output characteristics of SPNs after SNI were not identical to the ones observed in naïve rats. As in the neurons of naïve rats [4], we observed LO- (discharging with 1–2 APs in response to high-threshold dorsal root stimulation, Fig. 1 Ca-b) and HO-SPNs (discharging with bursts of 10–20 APs, Fig. 1 Cc-d). However, these two neuronal populations were no longer nociceptive-specific: 3 out of 8 LO-SPNs and 1 out of 4 HO-SPNs also responded to low-threshold stimuli (50 µs / 30–50 µA, Fig. 1 Cb, d). Thus, SNI produced an allodynic shift of the output characteristics of LO- and HO-SPNs.

One-third of tested SNI SPNs exhibited neuropathy-specific outputs (Fig. 1 Ce-g). A single high-threshold stimulus switched these cells to a state of persistent firing (PF-SPNs, Fig. 1 Ce-f) which often lasted for tens of seconds. In 2 out of 6 PF-SPNs such a switch could also be triggered by low-threshold pulses (Fig. 1 Cf). Given the emergence of PF-SPNs, output patterns of SPNs differed significantly between SNI and control (Fig. 1 Cg, p < 0.001, Fisher's exact test). Transitions to a persistent firing state happened even in the absence of any stimulation (Fig. 1 Da). Spontaneous PF was detected in 38% (10 out of 26) of SPNs from SNI rats, whereas it was absent in SPNs of naïve rats (Fig. 1 Db, p < 0.001, Fisher's exact test). Interestingly, neither LO- nor HO-SPNs showed spontaneous persistent firing, suggesting cellular specificity of such type of output.

Neuropathy-induced increase in SPN responses to peripheral stimuli has been observed in vivo previously [9]. We found that SNI altered lamina I SPN output in a cell-specific manner, compromising the functional roles that SPNs play in physiological conditions. First, SNI promoted an allodynic shift of SPN output characteristics: in total, 33% of tested SPNs generated spikes in response to the activation of low-threshold fibers. Moreover, LO- and HO-SPNs were no longer nociceptive-specific as part of them discharged to low-threshold stimulation. Second, SNI dramatically increased SPN output in response to high-threshold stimulation. While HO-SPNs were responsible for about 70% of APs in naïve rats [4], PF-SPNs (undetected in uninjured animals) were the main contributors to the overall AP output after the SNI. Lastly, SPNs showed spontaneous persistent firing states which were not observed in the neurons of naïve rats. Altogether, these changes of SPN output might represent cellular mechanisms for three classic neuropathy-associated pain conditions - allodynia, hyperalgesia, and spontaneous pain.

SNI-induced increase of SPN output could be attributed to synaptic and/or intrinsic neuronal mechanisms. Likewise to spinal interneurons [10], we did not find any neuropathy-associated changes of SPN resting membrane potential or membrane resistance: these parameters (-61 ± 1 mV and 0.86 ± 0.11 GΩ, respectively, n = 26) did not differ significantly (unpaired Student's t-test, data not shown) from the ones for naïve SPNs (reported earlier [4]). Unaltered passive membrane parameters suggested other causes for changes in SPN excitability. It has recently been shown that lamina I SPNs differ in voltage-gated potassium currents [11]. SNI could add up to these differences, leading to the development of neuropathy-specific discharge phenotypes. Synaptic mechanisms underlying altered SPN output could be quite diverse and involve both primary afferents and the spinal cord. For instance, the synergy between nociceptor hyperexcitability resulting from the upregulation of sodium voltage-gated channels [12,13] and microglia-dependent increase in AMPA and NMDR-receptor-dependent responses [14] could be the reason for the allodynic shift of SPN output characteristics. This shift could similarly arise from a postsynaptic boost of glutamatergic synaptic transmission caused by glial mediators released by spinal astrocytes [15]. The latter could also contribute to changed SPN functioning because of the downregulation of glutamate transporters [15] leading to a prolonged presence of glutamate within the synaptic cleft. Lastly, neuropathy-induced impairment of inhibitory tone within the dorsal horn circuitry [12] could play an important role in maintaining evoked and spontaneous persistent firing of the SPNs. Given that changes of SPN output are cell-specific, one may assume that so are the underlying mechanisms. Therefore, further studies of synaptic and intrinsic cellular mechanisms are necessary for a better understanding of altered SPN output in particular and neuropathic pain in general.

Ethics statement

All experimental procedures were approved by the Animal Ethics Committee of the Bogomoletz Institute of Physiology (Kyiv, Ukraine) and performed in accordance with the EU Directive 2010/63/EU for animal experiments, ethical guidelines of the International Association for the Study of Pain, and the Society for Neuroscience Policies on the Use of Animals and Humans in Neuroscience Research.

Author contributions

VK: the concept of the study, research design, behavioral testing, electrophysiological recordings, data analysis and interpretation, manuscript preparation. KA and MK: electrophysiological recordings, data analysis, manuscript preparation. SR, SK, and MK: data analysis, manuscript preparation. PB and NV: conceiving the study, data analysis and interpretation, and manuscript revision.

Declaration of Competing Interest

The authors declare no competing interests.

Data availability

• Data will be made available on request.