Peripheral gating of mechanosensation by glial diazepam binding inhibitor

We report that diazepam binding inhibitor (DBI) is a glial messenger mediating crosstalk between satellite glial cells (SGCs) and sensory neurons in the dorsal root ganglion (DRG). DBI is highly expressed in SGCs of mice, rats, and humans, but not in sensory neurons or most other DRG-resident cells. Knockdown of DBI results in a robust mechanical hypersensitivity without major effects on other sensory modalities. In vivo overexpression of DBI in SGCs reduces sensitivity to mechanical stimulation and alleviates mechanical allodynia in neuropathic and inflammatory pain models. We further show that DBI acts as an unconventional agonist and positive allosteric modulator at the neuronal GABAA receptors, particularly strongly affecting those with a high-affinity benzodiazepine binding site. Such receptors are selectively expressed by a subpopulation of mechanosensitive DRG neurons, and these are also more enwrapped with DBI-expressing glia, as compared with other DRG neurons, suggesting a mechanism for a specific effect of DBI on mechanosensation. These findings identified a communication mechanism between peripheral neurons and SGCs. This communication modulates pain signaling and can be targeted therapeutically.

time spent responding to the tape was recorded.Biting, scratching, or visibly attempting to remove the tape from the paw was scored as a response.
Response to pinprick.Mice were acclimated in von Frey testing chambers for 1 hour.A 27-gauge needle was applied to the glabrous skin of the hindpaw, taking care not to pierce through the skin.
Ten trials per mouse were performed with 1-min intervals between each trial.Paw withdrawal, shaking, or licking was scored as a response and reported as percentage for the total number of trials.
A new needle was used for each mouse.
Alligator clip.Response to blunt pressure application was assessed with the alligator clip assay, as previously described ( 76), modified to hind paw.Briefly, mice were first acclimated for 5 min in round plexiglass containers.An alligator clip (with constant pressure) was applied to the middle of the paw, and the mouse was placed back into the plexiglass container.A response was scored when the mouse showed awareness of the clip by biting, vocalization, grasping of the paw, or a jumping response.A cut-off of 90 s was applied to prevent tissue damage.The time to respond was recorded and reported as latency for each mouse.

Cold allodynia (dry ice test).
The test was performed as previously described (77).Each animal was placed in a clear acrylic container separated by black opaque dividers, which were positioned on top of 3/16" borosilicate glass (Stemmerich Inc, USA) and allowed to acclimate for 20 minutes before testing.A dry ice pellet was applied to the hind paw through the glass and the time until hind paw withdrawal was recorded; three trials were performed for each mouse at 5-min intervals and mean withdrawal latency was calculated.
Hargreaves assay.Heat sensitivity was assessed using the Hargreaves method (78).The radiant heat source was applied to the plantar surface of the mouse hind paw and the latency to withdrawal was used to determine the heat sensitivity threshold, with a cut-off of 20 s to prevent tissue damage; three trials were performed for each mouse at 5-min intervals and mean withdrawal latency was recorded.
Rotarod assay.Motor function was assessed using a rotarod apparatus (IITC Life Science) and measured as the latency of the mouse to fall off the rotating platform, with an acceleration of 0.2 rpm/sec over a two-minute period.
Quantitative real-time RT-PCR (qPCR).DRGs were extracted and total RNA was extracted using Trizol reagent (Invitrogen).Isolated RNA was dissolved in 10 l of DEPC-treated water and reversetranscribed using a reverse transcription reagent kit (PrimeScript RT Reagent Kit with gDNA Eraser, Takara) and a thermal cycler (Mastercycler, Eppendorf).Quantitative PCR (qPCR) analysis was performed in the Real-time Detection System (FQD-48A(A4), BIOER) by SYBR Premix Ex TaqII (Takara).The PCR products were also run on a 2% agarose gel and were visualized using a gel imager (TFP-M/WL, Vilber Lourmat).Primers for qPCR analysis are listed in Suppl.Table 1.

Enzyme Linked Immunosorbent Assay (ELISA).
DRGs from all spinal levels of C57BL/6 mice were rapidly extracted into HBSS on ice and washed once.Standard DRG incubation solution (500 µl) containing (in mM):160 NaCl, 2.5KCl, 5CaCl2, 1 MgCl2, 10 HEPES, 8 D-glucose (pH adjusted to 7.4 with NaOH; all from Sigma) was added and the ganglia were incubated for 30 minutes at room temperature.The 'High-K + ' solution was produced by equimolar replacement of 150 mM NaCl with 150 mM KCl.After incubation, the supernatants were collected for DBI detection using DBI ELISA kit from Abbexa Ltd (abx153899), according to the manufacturer's instructions.Fluostar Omega microplate reader (BMG LABTECH, Germany) was used to analyse the samples.Movie S1 Light-sheet microscopy of cleared rat lumbar DRG immunolabeled with NF200 (green), peripherin (red) and DBI (white).Staining, iDISCO clearance and imaging was performed as described in (6).

Supplemental
data identifying DBI expression pattern in DRG.A, Immunofluorescence (IF) staining of a section from an adult human DRG (red -DBI, blue -DAPI, greenautofluorescence).B, IF staining of a section from a foetal human DRG (red -DBI, green -TUBB3, blue -DAPI).C, D, Combined fluorescence in situ hybridisation (FISH) and IF analysis of mouse DRG sections.Co-labelling of DBI (FISH, white) and FABP7 (IF, red) is shown in (C); Colabelling of DBI (FISH, white) and GS (IF, red) is shown in (D).E, IF co-labelling rat DRG sections with antibodies against DBI (red) and S100B (green).F, IF co-labelling of rat DRG neuron section with antibodies against DBI (green), NF200 (blue) and peripherin (red).G. IF co-labelling of mouse sciatic nerve section with antibodies against DBI (red) and S100B (green).Supplemental Figure 3.Additional DBI siRNA knockdown experiments.A, Confirmation of the intrathecal siRNA knockdown efficiency of DBI; example immunofluorescence staining of DRG sections from mice receiving siRNA against Dbi or non-targeting control oligonucleotide (2 g/site).Immunofluorescence was quantified as percentage of immunolabeled area per section.B-E, Experiments on male mice.B, Analysis of the von Frey force vs. withdrawal response rate for mice intrathecally injected with either siRNA against Dbi (2 g/site; blue) or non-targeting control oligonucleotide (black); behavioural tests were performed 48 h after the injection.C, Timeline of the mechanical sensitivity changes following a single intrathecal injection of siRNA against Dbi (blue) or non-targeting control oligonucleotide (black).In (A, B) **, *** indicate significant difference from matched control group (p<0.01,p<0.001; two-way repeated-measures ANOVA with Sidák post-hoc test).D, E, Summary of the results of the pinprick (C) and rotarod test (D) of mice intrathecally Supplemental Figure 4. Analysis of transgene expression in the DRG following viral construct injections.A, Examples of fluorescence micrographs of mouse DRG 8 weeks after DRG injection with AAV9-U6-shDBI-CAG-EGFP virions; GFP fluorescence (green) and FABP7 (upper panels) and TUBB3 (lower panels) immunofluorescence (red) are shown individually (as labelled) or overlaid ('merge').B, Examples of fluorescence micrographs of mouse DRG 8 weeks after DRG injection with AAV5-gfaABC1D-DBI-EGFP virions.Presentation and labelling as in panel A. 11 Supplemental Figure 5. Genetic manipulations of DBI expression by viral DRG injections do not significantly affect spinal levels of DBI.A, B, GFP fluorescence (green) is not detectable in the spinal cord sections corresponding to L4 DRG of mice 8 week after DRG injection with either AAV9-U6-shDBI-CAG-EGFP virions (A) or AAV5-gfaABC1D-DBI-EGFP virions (B).Coimmunolabelling with FABP7 (upper panels) and TUBB3 (lower panels) is shown in red; DAPI nuclear staining is shown in blue.C, D, Example RT-PCR analyses (C) and summarised RT-PCR data (D) on the Dbi transcript level in the lumbar spinal cord sections of mice 8 week after DRG injection with either AAV9-U6-shDBI-CAG-EGFP virions, or AAV5-gfaABC1D-DBI-EGFP virions (as indicated).Supplemental Figure 6.Additional data for the manipulations with DBI expression by DRGdelivered viral constructs.A, Schematic of the experimental timeline for the experiments with DRG injection AAV9-U6-shDBI-CAG-EGFP virions (1.1 -1.2x10 12 vg/ml; 2 µl).B, C, Example RT-PCR analyses (B) and summarised RT-PCR data (C) on the Dbi transcript levels in the DRG after control or AAV9-U6-shDBI-CAG-EGFP virion injection; ***indicates significant difference from control group (p<0.001,unpaired t-test).D, E, Mechanical (von Frey) and thermal (Hargreaves) sensitivity tests during 42 days after the DRG injection of AAV9-U6-shDBI-CAG-EGFP virions or GFP control virions performed on the contralateral paws; the results of the corresponding ipsilateral paw tests are shown in Fig. 2H, I. F-I, Contralateral paw mechanical (von Frey) and thermal (Hargreaves) sensitivity tests on mice received DRG injections of AAV5-gfaABC1D-DBI-EGFP virions or GFP control virions (1.1 -1.2x10 12 vg/ml; 2 µl).Six weeks after viral injections mice underwent either SNI surgery (F, G) or hind paw CFA injections (H, I).The mice were tested for mechanical and thermal sensitivity thresholds during subsequent 14 days; the results of the corresponding ipsilateral paw tests are shown in Fig. 3G-J.J, Similar to panels (A-D) but GABAA channels were studied, together with 3(H126R) or 2(F77I) or their combination (as indicated); other conditions and labelling as in panels (A-D).K, L, summarise the data from experiments in (G-J), similar to panels (E, F). *,**,***indicate significant difference from  GABAA group (p<0.05,p<0.01, p<0.001, respectively; one-way ANOVA with Dunnett post-hoc test).(25 l) or the SNI surgery.Behavioural tests were performed on days 3, 5, 7, 10, and 14 thereafter.C, D, Mechanical (C) and heat (D) sensitivity was monitored after the SNI induction to the mice implanted with osmotic mini-pumps delivering either DBI (red; 200 μM, 0.5 μl/h) or mutant DBI(K33A) (green; 200 μM, 0.5 μl/h).*, ***, indicate significant difference from time-matched control group (at p<0.05, or p<0.001, respectively; two-way repeated-measures ANOVA with Tukey's post-hoc test).E, F, Mechanical (E) and heat (F) sensitivity was monitored after the CFA induction to the mice pre-implanted with osmotic mini-pumps delivering either DBI or mutant DBI(K33A).Other conditions as in panels (B, C) *, **, indicate significant difference from timematched control group (at p<0.05, or p<0.01, respectively; two-way repeated-measures ANOVA with Tukey's post-hoc test).G-K, benzodiazepine antagonist, flumazenil, but not TSPO antagonist, PK11195, antagonises anti-nociceptive properties of DBI.G, Schematic of the experimental.The DRG cannula implantation was performed and, at the same time, mice received either the hindpaw injection of CFA (25 l) or the SNI surgery.H, I, Mechanical (B) and heat (C) sensitivity was monitored after the SNI induction to the mice implanted with DRG cannulas.0.5 hrs before last measurements DBI (200 μM, 2 l) was co-injected with either PK11195 (red; 200 μM, 2 l total volume) or flumazenil (blue; 200 μM, 2 l total volume).*, indicates significant difference from time-matched control group (p<0.05;two-way repeated-measures ANOVA with Tukey's post-hoc test).J, K, Experiments similar to these shown in panels I, J but CFA inflammatory pain model was used instead of SNI.* indicates significant difference from time-matched control group (p<0.05;twoway repeated-measures ANOVA with Tukey's post-hoc test).