Mechanism of dorsal root ganglion stimulation for pain relief in painful diabetic polyneuropathy is not dependent on GABA release in the dorsal horn of the spinal cord

Abstract Aims It is hypothesized that dorsal root ganglion stimulation (DRGS), sharing some of the mechanisms of traditional spinal cord stimulation (SCS) of the dorsal columns, induces γ‐aminobutyric acid (GABA) release from interneurons in the spinal dorsal horn. Methods We used quantitative immunohistochemical analysis in order to investigate the effect of DRGS on intensity of intracellular GABA‐staining levels in the L4‐L6 spinal dorsal horn of painful diabetic polyneuropathy (PDPN) animals. To establish the maximal pain relieving effect, we tested for mechanical hypersensitivity to von Frey filaments and animals received 30 minutes of DRGS at day 3 after implantation of the electrode. One day later, 4 Sham‐DRGS animals and four responders‐to‐DRGS received again 30 minutes of DRGS and were perfused at the peak of DRGS‐induced pain relief. Results No significant difference in GABA‐immunoreactivity was observed between DRGS and Sham‐DRGS in lamina 1‐3 of the spinal levels L4‐6 neither ipsilaterally nor contralaterally. Conclusions Dorsal root ganglion stimulation does not induce GABA release from the spinal dorsal horn cells, suggesting that the mechanisms underlying DRGS in pain relief are different from those of conventional SCS. The modulation of a GABA‐mediated “Gate Control” in the DRG itself, functioning as a prime Gate of nociception, is suggested and discussed.

(hereafter labeled SCS) has resulted in significant pain reduction in many intractable PDPN patients. [9][10][11] However, in about 40% of these patients SCS treatment is not effective. [9][10][11] Conventional dorsal root ganglion stimulation (hereafter named DRGS) acts at the level of the dorsal root ganglion (DRG), a promising new location for neuromodulation in managing selected pain conditions-among those also PDPN. 12 DRGS enables the physician to target a more peripheral, selective communication station for all nociceptive signaling from the peripheral nervous system to the dorsal horn in the spinal cord, from where the information is further processed via the spinothalamic tracts to the brain. 13 The DRG plays an important role in both neuropathic and nociceptive pain conditions. 14 Therefore, while SCS can theoretically only modulate Aβ fiber signaling, DRGS might also be able to modulate Aδ-and C-type fiber signaling. 15 A key molecule in the processing and modulation of the nociceptive signal in the spinal dorsal horn is γ-aminobutyric acid (GABA). 16,17 Experimental data demonstrated decreased extracellular GABA levels and increased intracellular GABA levels in the dorsal horn after peripheral nerve injury. 18,19 GABA has been proposed to have a major role in nociceptive and non-nociceptive processing according to the Gate Control Theory. 20 Based on this, it was suggested to stimulate non-nociceptive Aβ fibers in the dorsal columns (SCS) aiming to turn on the Gate mechanisms and modulate the nociceptive input. In an experimental study of allodynic rats, increased levels of extracellular GABA were indeed found in the dorsal horns in response to SCS. 18,21,22 The pivotal role of GABA in the analgesic effect of SCS has furthermore been confirmed by the fact that the application of the GABA B agonist baclofen changed "non-responders-to-SCS" into "responders-to-SCS." [23][24][25][26] The results of an experimental study of Janssen et al 27 demonstrated additionally that responders-to-SCS showed decreased levels of intracellular GABA-immunoreactivity (GABA-IR) in the spinal dorsal horn in comparison with non-responders-to-SCS and Sham-SCS animals. A relation between the release of intracellular accumulated GABA in the spinal cord dorsal horn, and the analgesic effect of SCS was therefore hypothesized. [27][28][29][30] The mechanisms underlying DRGS and its ensuing pain relief are as yet unknown, and it is likely that it shares some spinal and supraspinal mechanisms with SCS, dependent on Aβ fibers activated by both types of stimulation. The present study is therefore aimed to investigate the hypothesis that DRGS induces GABA release from spinal dorsal horn cells. We used quantitative immunohistochemical analysis in an animal model of PDPN to study the effect of one single DRGS paradigm on the levels of spinal dorsal horn intracellular GABA.

| Ethical statement
The protocols for this study were approved by the Animal Care

| Animals
The study was performed on 48 adult female Sprague Dawley rats, weighing 170-230 g and 8 weeks old at the start of the experiments (Charles River, Maastricht, The Netherlands).They were preoperatively housed in pairs and postoperatively individually, in transparent plastic cages situated in a climate controlled room under a 12-hour light/dark cycle with food and water ad libitum. All efforts were made to minimize the number of animals used and their suffering, and alternatives to in vivo techniques were considered.

| Induction of diabetes mellitus
All animals fasted overnight before the induction of diabetes.
Streptozotocin (STZ, Sigma-Aldrich, Schnelldorf, Germany) was freshly dissolved in sterile NaCl 0.9% to a solution of 65 mg/mL. Animals were intraperitoneally injected with the STZ solution (65 mg/kg) to induce diabetes mellitus (DM). On day 4 post-STZ injection, blood glucose levels were measured from the saphenous vein, using an Accu-Chek Aviva® glucometer (Roche Diagnostics GmbH, Mannheim, Germany). Rats developing diabetes, defined as a blood glucose level of ≥15 mmol/L, 31 were included in the study. withdrawal threshold (WT) was determined by use of the up-down method, 32 as previously described. 33 The cutoff value was defined as the absence of the paw withdrawal response to a 28.84 g force to prevent tissue damage. The registered 50% WT's (measured in grams) were then multiplied by 10 000 and logarithmically transformed to conform with Weber's law 34 and in order to obtain a linear scale.

| Development and assessment of mechanical hypersensitivity
Mechanical hypersensitivity was tested at baseline (pre-STZ injection) and weekly during 4 weeks post-STZ injection. Mechanical hypersensitivity, caused by PDPN, was defined to be present in case of a decrease in ≥0.2 unit of the log 10 (10 000 × 50% WT) when compared to baseline. Only animals that developed mechanical hypersensitivity were selected for this study and treated with DRGS.

| Implantation of DRGS electrode
Under general anesthesia, bipolar electrode was implanted unilaterally at the L5 DRG, according to the earlier described procedure adapted from Pan et al. 35,36 In short, via a paravertebral incision the intervertebral foramen was exposed, the foramen was opened, and the anode and cathode were implanted at the DRG. The electrode was then secured into the transverse process followed by closure of the incision in layers.

| Immunohistochemical detection of GABA
Slides were allowed to dry at room temperature for 2 hour before being washed with Tris-buffered saline (TBS, 0.1 M, pH 7.6), including 0.3% Triton X-100 (TBS-T), TBS, and TBS-T. Thereafter, the slides were blocked for anti-GABA immunohistochemistry by the incubation in 2% normal donkey serum (Sigma-Aldrich, Zwijndrecht, The Netherlands, D9663) for 1 hour, diluted in TBS-T, and then incubated with rabbit anti-GABA polyclonal antibody (1:5000 diluted in TBS-T; Sigma-Aldrich, Zwijndrecht, The Netherlands, A2052) for 48 hour.

| Immunoreactivity analysis
After the GABA-staining protocol was completed, sections were observed under an Olympus AX-70 microscope and immunohistochemical analysis was performed, as previously described by Janssen et al. 27 Firstly, photomicrographs were taken from both ipsilateral and contralateral dorsal horns for the spinal levels L4-L6 using the Provis

| Statistical analysis
The assay and the data analysis were performed in a blinded fashion in an identical mode as published before. 27

| Description of cohorts of animals
Forty-three of the 48 animals (90%) that were injected with STZ developed DM (blood glucose ≥15 mmol/L). Twenty-two out of the 43 diabetic animals developed painful neuropathy (51%, ≥0.2 decrease in log 10 (10 000 × 50% WT) and were implanted with a DRGS system. Two animals were excluded from the study because of high MT (MT > 1mA) and two due to a connector breakage. Eleven animals were selected for treatment with DRGS and seven for treatment with Sham-DRGS. 36 Of these animals, 4 responders-to-DRGS animals and 4 Sham-DRGS animals were selected for the GABA-IR analyses.
We described the efficacy of DRGS in an earlier publicized study. 36 In short, DRGS induced a complete reversal of mechanical hyper-

| Comparison of mean gray values L4-L6
The anti-GABA immunohistochemical analysis showed a strong GABA-IR, prevalently in laminae 1-3 of the spinal dorsal horn

| D ISCUSS I ON
This is the first experimental study to assess intracellular GABA with the Gate Control Theory, 18,27 while our results instead point to a different mechanism involved in DRGS and its production of pain relief. The assay and data analysis utilized during our study were performed in an identical way as the previous study regarding GABA release in SCS, 27 with the difference that the previous study did not concern PDPN animals but neuropathic pain animals with a partial sciatic nerve ligation (according to the Seltzer model). From these data, it was concluded that the assay used for this study is sensitive enough to detect changes in GABA-IR in the spinal dorsal horn.
The dorsal horn is an essential and second-order relay station for the integration and modulation of pain. 16  Conventional SCS is not successful in all patients, and pain relieving effects can decline over the years. 9,11 Lack of anatomic specificity of the painful area and positional variations in stimulation is furthermore well-known drawbacks of conventional SCS therapy. The DRG appears to be an appealing site for neurostimulation, 45 and clinical evidence indicates that DRGS provides efficacious pain relief in neuropathic pain sufferers, [46][47][48][49][50] which is confirmed by experimental studies. 35,36 In comparison with SCS, DRGS has been demonstrated to have a better anatomic specificity of the painful area. 50 Furthermore, with DRGS the electrical fields have a direct effect on the neural tissues that are pathophysiologically involved in the chronic pain disease condition.
Since the location of stimulation is completely different with DRGS (local stimulation at the DRG) as compared to SCS (stimulation of the dorsal column), the question remains whether both stimulation paradigms act via the same mechanism. Our quantitative immunohistochemical data indicate that, even though DRGS and SCS resulted in a similar decrease in PDPN, 36  Clearly, DRGS does probably not only directly act via modulation of Aβ fibers, but also of Aδ-and C-type fibers. 15  In conclusion, DRGS does not induce GABA release in spinal dorsal horn of neuropathic (PDPN) rats. With this observation, we suggest that the mechanism underlying DRGS-induced pain relief is different from that of dorsal column SCS. Further research is warranted to elucidate the mechanism underlying DRGS in pain relief. The modulation of a GABA-mediated "Gate Control" in the DRG, functioning as a prime Gate of nociception, is suggested.

ACK N OWLED G M ENT
Prof. Roberto SGM Perez had an important role in the initiation of this study and helped substantially with the conception and design of the work. To our greatest regret, he passed away on 07.09.2017.

D I SCLOS U R E
This work was financially supported by a research grant from Abbott