The Primary Cilium and its Hedgehog Signaling in Nociceptors Contribute to Inflammatory and Neuropathic Pain

Animals

All experiments using rats were performed on 300 to 400 g male Sprague-Dawley rats (Charles River Laboratories, Hollister, CA). Rats were housed 3 per cage under a 12-h light/dark cycle in a temperature- and humidity-controlled room at the University of California, San Francisco animal care facility. Food and water were available in home cages, ad libitum. Protocols for rat experiments were approved by the University of California, San Francisco, Institutional Animal Care and Use Committee, and adhered to the National Institutes of Health (NIH) Guidelines for the care and use of laboratory animals.

All studies on mice were performed with C57BL/6J male and female mice (The Jackson Laboratory, Bar Harbor, ME). Experiments in mice were conducted under a protocol approved by the Institutional Animal Care and Use Committee at the University of New England, in accordance with the Guide for the Care and Use of Laboratory Animals as adopted and promulgated by the NIH. Mice were housed 3–5 per cage under a 12-h light/dark cycle in a temperature- and humidity-controlled room in the animal care facility at the University of New England, Biddeford campus. Food and water were available in home cages, ad libitum.

Immunohistofluorescence analysis of primary cilia in lumbar dorsal root ganglia (DRGs)

Rats were anesthetized with 5% isoflurane followed by perfusion through the left ventricle with 100 ml of cold phosphate buffered saline (PBS) containing 10U heparin/mL PBS, followed by 300 ml of methanol-buffered 10% formalin (Thermo Fisher, Waltham, MA). Bilateral L4 and L5 DRGs were dissected out, transferred into PBS containing 30% sucrose and 0.02% NaN₃. These DRGs were then mounted in red tissue freezing medium (Electron Microscopy Sciences, Hatfield, PA) and 15 micrometers (μm) sections cut with a cryostat (Leica CM1900). Sections were then mounted on glass slides and used for immunohistofluorescence analyses.

Presence and length of primary cilia on DRG neurons were investigated using antibody staining of histological sections of adult rat DRGs and acutely-dissociated cultures prepared from surgically-isolated fresh mouse DRGs. Slides containing either rat DRG sections or coverslips with mouse DRG neuronal cultures were rehydrated and washed three times with 1X PBS, followed by a 30-min incubation in blocking buffer (1X PBS, 0.3% TritonX-100, 5% normal goat serum). Blocking buffer was then aspirated and primary antibodies, in antibody incubation buffer (1x PBS, 0.3% TritonX-100, 1% bovine serum albumin), were applied and allowed to incubate in a humid, dark chamber overnight at 4 °C. Slides were allowed to equilibrate to room temperature, and primary antibody solution aspirated. Slides were then washed three times with 1X PBS, secondary antibodies in antibody incubation buffer were applied and allowed to incubate at room temperature in a humid, dark chamber for 45 min. Slides were then washed five times with 1X PBS, followed by one rinse in a Coplin jar containing DI H₂0, and finally cover slipped with DAPI mounting medium (Abcam, Waltham, MA) before imaging with a confocal microscope. The following primary antibodies were employed: rabbit anti-ARL13B (ADP-ribosylation factor-like 13B; ProteinTech, Rosemont, IL) 1:1000, mouse anti-ɣ-tubulin (clone GTU-88, Sigma-Aldrich, St. Louis, MO) 1:1000, goat anti-TrkA (tropomyosin receptor kinase A; R&D Systems, Minneapolis, MN) 1:300, rabbit anti-ACIII (adenylyl cyclase III; Encor, Gainesville, FL) 1:4000, mouse NeuN (RNA binding fox-1 homolog 3; ProteinTech) 1:1000, and Alexa 647-coupled IB4 (isolectin GS-IB₄; Invitrogen) 1:500. The following secondary antibodies were employed: goat anti-rabbit IgG (H+L) (Alexa 488-coupled, Invitrogen, Waltham, MA) 1:1000, and goat anti-mouse IgG (H+L) (Alexa 555-coupled, Invitrogen) 1:1000.

Imaging and data processing

Microscopy was conducted with a Leica Stellaris laser scanning confocal microscope utilizing Leica LAS X software version 4.6 and Plan Apo 40x/1.3NA and 63x/1.4NA objectives. 405, 488, 561, and 638 nm lasers with appropriate beam splitters (substrate or TD 488/561/633) were activated and emission ranges were selected for each fluorophore. Sequential scanning between frames was then activated at 1024×1024 resolution using frame average = 2. The channel containing the ciliary marker was then used to establish the appropriate z-volume. Optimal step size was established using the Nyquist Sampling Theorem calculated by LAS X. A z-stack of interest was then collected, titled, and saved for processing. Data was transferred as .lif files from the confocal software to FIJI. Channels were split individually, and stacks were projected to maximum intensity. LUT was set to grayscale for each of the channels to adjust brightness and contrast. LUTs were then set for cell nuclei in blue, cilia in green, basal body in red, and grayscale or yellow for neuronal markers.

DRG neuron cultures

Acute cultures of dissociated DRG were prepared from C57BL/6J mice aged 4 – 6 weeks (The Jackson Laboratory), as described by Molliver and colleagues²⁴. In anesthetized mice, after transcardial perfusion of ice-cold phosphate-buffered saline solution, and removal of overlying dorsal skin, spinal ligaments and paraspinal muscles, laminectomy was performed with spring-loaded micro-iridectomy scissors and lumbar DRG were harvested. After desheathing, DRGs were pooled and treated, first with papain (Worthington Biochemical, Lakewood, NJ) for 10 minutes (min) at 37 °C, followed by a digestion with collagenase, type 2, (Worthington Biochemical) and dispase II (Roche, Mannheim, Germany) for 10 min at 37° C. After trituration with a fire-polished Pasteur pipette, neurons were plated onto individual 12-mm diameter round glass cover slips (Warner Instruments, Holliston, MA) that had been coated with poly-D-lysine and laminin and cultured in F12 medium with 5% fetal calf serum and penicillin-streptomycin (Thermo Fisher). Neurons were cultured for 2–5 days before fixation with 4% paraformaldehyde in a cytoskeletal-stabilizing buffer²⁵ at room temperature for 20 min, followed by immunohistofluorescence analysis.

RNA extraction and reverse transcription quantitative real-time polymerase chain reaction (RT-qPCR)

Total RNA from rat L4 and L5 DRG was extracted using Trizol (Invitrogen) and the PureLink^™RNA Mini Kit (Invitrogen) according to the manufacturer’s instruction. The RNA concentration in each sample was determined with a spectrophotometer (Shimadzu, Santa Clara, CA). RNA was transcribed into cDNA using the iScript^™ Advanced cDNA Synthesis Kit for RT-qPCR (Bio-Rad, Hercules, CA). Real time PCR (Polymerase Chain Reaction) was performed with the SsoAdvanced Universal SYBR Green Supermix (Bio-Rad) and specific primers for IFT88 (intra-flagellar transport protein 88) (Bio-Rad, assay ID: qRnoCID0004310) and IFT52 (Bio-Rad assay ID: qRnoCID0002245) on the CFX 96^™ real-time PCR detection system (Bio-Rad). The PCR program was run with the following conditions: 2 min at 95 °C, followed by 40 cycles of 15 s at 95 °C, and 30 s at 60 °C. Each sample was run in triplicate. Change in gene expression was determined by the 2^−ΔΔCT method and expressed as relative change with respect to control level. Glycerinaldehyde-3-phosphate dehydrogenase (GAPDH) was used as a reference gene (Bio-Rad, assay ID: qRnoCID0057018).

Pharmacological reagents

Prostaglandin E₂ (PGE₂), cyclopamine hydrate, and paclitaxel (purified from Taxus yannanensis) were purchased from Sigma-Aldrich. Recombinant rat Sonic hedgehog (Shh) was purchased from Abcam. Stock solutions of PGE₂ (1 μg/μl) in absolute ethanol was diluted to 20 ng/μl) with 0.9% NaCl immediately before injection. The ethanol concentration of the final PGE₂ solution was ~2% and the injection volume 5 μl. Paclitaxel was prepared in absolute ethanol and polyethoxylated castor oil (Cremophor EL; 1:1; Sigma-Aldrich) and further diluted in saline, to a final concentration of 1 mg/kg). Cyclopamine hydrate was diluted in 0.9% NaCl containing 2% DMSO and was administered by intradermal (i.d.) or intra-ganglionic (i.gl.) routes of administration, both 10 μg), respectively. Recombinant Shh was reconstituted in 0.9% NaCl at 0.1 mg/ml and further diluted in 0.9% NaCl to be administered i.gl. or i.d. at 200 ng/animal. All dose selections were based on previous studies that established effectiveness at their targets^17,26–28.

Intrathecal administration of siRNA

Small interfering (si) RNA was used to silence the mRNA expression of Ift88 (mIFT88, Thermo Fisher, Ambion in vivo pre-designed siRNA, siRNA ID number S157132) and Ift52 (mIFT52, Thermo Fisher Ambion in vivo predesigned siRNA pn4404010, siRNA ID number. S187205) in vivo, by intrathecal (i.t.) administration. A scrambled sequence was used as the negative control (nc-siRNA, Thermo Fisher, Ambion in vivo pre-designed negative control siRNA). siRNAs were mixed with PEI (In Vivo Jet-PEI; Polyplus Transfection) according to the manufacturer’s instructions. The N:P ratio (number of nitrogen residues of In Vivo Jet-PEI per DNA phosphate) used was 8 (1 μg of siRNA was mixed with 0.16 μl of In Vivo Jet-PEI). siRNA targeting IFT88, IFT52 and nc-siRNA were administered i.t. for 3 consecutive days. To administer siRNAs, rats were briefly anaesthetized with 2.5% isoflurane and a 30-gauge hypodermic needle inserted into the subarachnoid space, on the midline, between the L4 and L5 vertebrae. The i.t. site of injection was confirmed by the elicitation of a tail flick, a reflex that is evoked by accessing the subarachnoid space and bolus i.t. injection²⁹. This procedure produces not only reversible inhibition of the expression of the relevant proteins in DRG neurons but also modulation of nociceptive behavior²⁷.

Intra-ganglion administration of pharmacological agents

Injection of compounds targeting the Hh signaling pathway into the L5 DRG was performed with a gingival needle (30-gauge) prepared as previously described^30,31, attached to a 50 μl Hamilton syringe by a short length of PE-10 polyethylene tubing; modification of the injecting needle decreases the risk of tissue damage when its tip penetrates the DRG. Rats were lightly anesthetized by inhalation of isoflurane, and the fur over the lower back shaved. The site of skin penetration was 1.5 cm lateral to the vertebral column and ~0.5 cm caudal from a virtual line passing between the rostral edges of the iliac crests. To facilitate the insertion of the injecting needle through the skin, an initial cutaneous puncture was made with a larger (16-gauge) hypodermic needle. In sequence, the injecting needle was inserted through the puncture wound in the skin and oriented toward the region of the intervertebral space between the fifth and sixth lumbar vertebrae, until it touched the lateral region of the vertebrae. To reach the space between the transverse processes of the fifth and sixth lumbar vertebrae, small incremental movements of the needle tip were performed, until the resistance provided by the bone was diminished. When it reached the correct position, the tip of the needle felt “locked in place” (through the intervertebral space) and a flinch of the ipsilateral hindpaw was observed³⁰, indicating that it had penetrated the DRG of the fifth lumbar spinal nerve³¹. The procedure, starting from the beginning of anesthetic inhalation, until withdrawal of the injection needle, takes ~3 min. Animals regained consciousness ~2 min after anesthesia was discontinued. Importantly, no change in the nociceptive mechanical paw-withdrawal threshold of the ipsilateral hindpaw, after a single injection or daily repeated injections in the L5-DRG, was observed, and immunofluorescence analysis of the L5-DRG after injections has shown no signs of cell damage³⁰. Likewise, no significant change in the mechanical nociceptive threshold in the hindpaw, after i.gl. injection of vehicle (single or repeated injections), was observed, indicating a lack of injection-induced damage (*PMID: 23401543*). This technique has been used by us and others (PMIDs: 25878283, 23401543, 25489099, 29626652, 32523543, 23776243, 30537520).

Intradermal administration of pharmacological agents

For the i.d. injection of cyclopamine and recombinant Shh, a hypotonic shock (2 μl of distilled water, separated in the syringe by an air bubble) preceded their injection, to facilitate entry of membrane-impermeable compounds into nerve terminals^32–34 on the dorsum of the hind paw. In these experiments, a total volume of 5 μl was injected i.d. on the dorsum of the hind paw.

Measuring mechanical nociceptive threshold

Mechanical nociceptive threshold was quantified using an Ugo Basile Analgesymeter (Stoelting, Wood Dale, IL), to perform the Randall–Selitto paw withdrawal test^35–37. This device applies a mechanical force that increases linearly with time, to the dorsum of the rat’s hind paw. Rats were placed into cylindrical acrylic restrainers with lateral ports that allow access to the hind paw, as previously described³⁸, to acclimatize them to the testing procedure. Mechanical nociceptive threshold is defined as the force in grams at which a rat withdraws its paw. Baseline threshold is defined as the mean of 3 readings taken before injection of test agents; each experiment was performed on different groups of rats. Data are presented as mechanical nociceptive threshold (in grams) and as percentage change from preintervention baseline.

Paclitaxel chemotherapy-induced peripheral neuropathy (CIPN) in the rat

Using an established rat CIPN model, paclitaxel was injected intraperitoneally (i.p.) every other day for a total of 4 doses (1 mg/kg × 4 i.p.); control animals received the same volume of vehicle with proportional amounts of Cremophor EL and ethanol diluted in saline^28,39. Presence of paclitaxel-induced painful peripheral neuropathy was confirmed by demonstrating a decrease in mechanical nociception threshold in the Randall-Selitto paw withdrawal test^28,40–42.

Inflammatory pain model: PGE2-induced mechanical hyperalgesia in the rat

PGE₂ (100 ng) was injected i.d., in a volume of 5 μl, using a 30-gauge hypodermic needle attached to a microsyringe (Hamilton, Reno NV). PGE₂-induced hyperalgesia was quantified as the percentage change in mechanical nociceptive threshold, calculated from the baseline, pre-injection, mechanical nociceptive threshold versus the threshold after injection of the test agent. The dosing protocol and timing of mechanical nociceptive threshold assessment were based on previous studies^26,43.

Statistical analyses

In each behavioral experiment, only one hind paw per rat was used. In these experiments, data are presented as percentage change from baseline mechanical nociceptive threshold. Repeated-measures one-way analysis of variance (ANOVA), followed by the Bonferroni post hoc multiple comparison test or the Student t test was used for statistical analyses. Prism 8.0 (GraphPad Software, San Diego, CA) was used for the graphics and to perform statistical analyses; P < 0.05 was considered statistically significant. Data are presented as mean ± SEM.

Adult nociceptors elaborate soma primary cilia, in vivo and in vitro

Primary cilia in adult DRG neurons were detected using immunohistofluorescence staining with confocal microscopy. Adult rat DRGs (Fig. 1A–D, Suppl. Fig. 1B), and cultures of neurons from acutely-dissociated adult mouse DRGs (Suppl. Fig. 1A), were labeled with an antibody recognizing the ciliary protein Arl13b (ADP-ribosylation factor-like 13B)⁴⁴, to identify the plasma membrane surrounding the central ciliary axoneme, and an antibody recognizing ɣ-tubulin to label the basal body from which the primary cilium grows^45,46. To identify the cells expressing these ciliary markers as neurons, we employed antibodies recognizing Fox-3/NeuN (Fig. 1A), TrkA (Fig. 1B) or calcitonin gene-related peptide (CGRP) (Fig. 1C). We also identified neurons binding isolectin IB4, which labels non-peptidergic nociceptors (Fig. 1D). In DRG cultures neurons were identified by their double birefringent phase-contrast appearance and isolectin B4 (IB4) reactivity (Suppl. Fig. 1A). Clear indication of a primary cilium was found on many neurons in the rat DRG and elaborated by cultured mouse DRG neurons. In rat DRGs, the large number of satellite cells surrounding neurons are also ciliated (Suppl. Fig. 1B), with cilia that appear to be oriented facing neuronal cell bodies.

Targeted knockdown of Ift88 by siRNA attenuates Ift88 in rat DRG, leading to a loss of primary cilia and a decrease in the average length of retained cilia

Two classes of IFT protein complexes, B and A, are responsible for anterograde and retrograde trafficking of proteins into and out of the primary cilium, respectively⁴⁷. Ift88 forms an important component of the B complex⁴⁸. Primary cilia in DRG neurons can be disrupted through the targeted inactivation of genes encoding proteins essential for maintenance of primary cilia structure, such as Ift88, an approach that we have employed successfully^49–52. In the present experiments male rats were treated i.t. with siRNA targeting Ift88, in a dose of 2, 5 or 10 μg/day, for 3 consecutive days (Fig. 3A–D). Because rats treated with the highest dose of siRNA showed the strongest behavioral phenotype, L4 and L5 DRGs were harvested 7 days after the last injection of Ift88 siRNA, from euthanized rats, to prepare mRNA, or rats were transcardially perfused with 4% paraformaldehyde to prepare tissue for immunohistofluorescence analysis. Total RNA, isolated from L4 and L5 lumbar DRGs, from rats treated with 10 μg siRNA, was used to generate cDNA through reverse transcription. Quantitative RT-PCR, performed on the resultant cDNA revealed a significant decrease in Ift88 expression in animals treated with the 10 μg dose of siRNA targeting Ift88 (Fig. 2A). To visualize and quantify the primary cilium at the single-cell level, DRGs were harvested from paraformaldehyde-perfused rats, 7 days after the final injection of the 10 μg dose of siRNA targeting Ift88. Then 15-μm thick sections of cryo-embedded L4 and L5 DRGs were labeled with antibodies recognizing Arl13b and ɣ-tubulin, followed by confocal microscopy (Fig. 2B). Counting of cells with and without cilia, and length measurement of remaining cilia revealed a significant decrease in both the prevalence of cilia in DRG cells (Fig. 2C) as well as the average length of remaining cilia (Fig. 2D), in animals treated with siRNA targeting Ift88, compared to controls treated with a negative siRNA.

Effect of Ift88 siRNA on baseline mechanical nociceptive threshold and pain associated with paclitaxel CIPN

Male rats were treated i.t. with siRNA targeting Ift88, 2, 5 or 10 μg/day for 3 consecutive days. Twenty-four, 48 and 72 h after the last administration of siRNA, nociceptive threshold was evaluated by the Randall-Selitto paw-withdrawal test. At 24, 48 and 72 h, nociceptive threshold was significantly increased compared to rats receiving the same dose of a negative control siRNA (Fig. 3A and B).

We next employed paclitaxel CIPN as a preclinical model of neuropathic pain, to evaluate the role of the nociceptor primary cilium in neuropathic pain. A group of male rats were treated with siRNA targeting Ift88 (2, 5 or 10 μg/day, i.t.) for 3 consecutive days. Seventy-two hours after the last administration of siRNA, paclitaxel (1 mg/kg, i.p.) was administered on days 0, 2, 4 and 6. Mechanical nociceptive threshold was evaluated before siRNA treatment was started (baseline), and again on days −3, −2, −1, 3, 5, 7,14, 21 and 28 dose after the last dose of paclitaxel, from day 3 of behavioral testing, with the highest dose of Ift88 siRNA bringing animals essentially back to baseline values, as compared with rats receiving an i.t. dose of a negative control siRNA (Fig. 3C and D).

Effect of Ift52 siRNA on CIPN pain

To independently validate the role of primary cilia in nociceptor function we evaluated the role of a second intra-flagellar protein, Ift52, in paclitaxel CIPN. Male rats were treated with siRNA targeting Ift52 (10 μg, i.t., for 3 consecutive days). Seventy-two hours after the last administration of Ift52 siRNA rats were euthanized and L4 and L5 DRG harvested, for preparation of RNA. Total RNA isolated from L4 and L5 DRGs was used to generate cDNA through reverse transcription. Quantitative RT-PCR performed upon the resultant cDNA revealed a significant decrease in Ift52 expression in animals treated with siRNA targeting Ift52 (Fig. 4A). We again employed paclitaxel CIPN as our neuropathic pain model, to evaluate the role of the nociceptor primary cilium in neuropathic pain. Male rats were treated with siRNA targeting Ift52 for 3 consecutive days in a dose of 10 μg/day (i.t.). Seventy-two hours after the last siRNA injection, paclitaxel (1 mg/kg, i.p.) was administered on days 0, 2, 4 and 6. Mechanical nociceptive threshold was evaluated before siRNA treatment was started (baseline), and again on days −3, −2, −1, 3, 5, 7,14, 21 and 28 days after the last dose of paclitaxel. Similar to the effect of Ift88 siRNA, the siRNA targeting Ift52 reduced paclitaxel-induced hyperalgesia from day 3 of behavior testing, as compared with rats receiving an i.t. dose of a negative control siRNA (Fig. 4B and C).

Primary cilium-dependence of the pronociceptive effect of Sonic hedgehog

While many effects of Hh are primary cilium-dependent^14,53, the role of the primary cilia in Hh-induced pain has not been studied. To test for the dependence of the pronociceptive effects of Shh on the primary cilium, male rats were treated with siRNA targeting Ift88, in a dose of 10 μg/day for 3 consecutive days. Seventy-two hours after administration of the last dose of Ift88 siRNA, rats received an i.gl. (Fig. 5A) or i.d. (Fig. 5B) injection of recombinant Shh (200 ng) and nociceptive threshold was evaluated. Treatment of rats with siRNA for Ift88, reduced hyperalgesia by both i.gl. and i.d. Shh (Fig. 5A and B).

To investigate the role of Hh signaling in nociceptor sensitization, we again employed the CIPN model of neuropathic pain induced by paclitaxel. To induce CIPN, male rats were treated with paclitaxel (1 mg/kg x 4, i.p.) on days 0, 2, 4 and 6. On day 7, at which time hyperalgesia was fully established, rats were treated with cyclopamine (10 μg) i.gl. (Fig. 5D) or i.d. (Fig. 5C), which inhibits Hh signaling by binding to the Smo co-receptor⁵⁴. Nociceptive threshold was evaluated 30 min and, 1 and 3 h after cyclopamine treatment (N=6). Administration of cyclopamine, by both the i.gl. and i.d. route, attenuated paclitaxel-induced hyperalgesia (Fig. 5A and B).

siRNA targeting Ift88 attenuates inflammatory pain

Finally, we examined the role of nociceptor primary cilia in inflammatory pain, as produced by i.d. injection of PGE₂, a prototypical direct-acting pronociceptive inflammatory mediator^55–57. Rats were treated i.t. with siRNA targeting Ift88 (10 μg/day, i.t.) for 3 consecutive days. 72 h after the final siRNA injection, PGE₂ was administered on the dorsum of the hindpaw (100ng/5μl/paw, i.d.), at the site of nociceptive threshold testing. Nociceptive threshold was evaluated before siRNA treatment was started (baseline), again 72 h after the last siRNA treatment, and 30 minutes after i.d. injection of PGE₂ (Fig. 6). As was observed with baseline thresholds in the control rat (Fig. 3A, B), siRNA targeting of Ift88 led to an increase in the baseline nociceptive threshold (Fig. 6A). Treatment with Ift88 siRNA also resulted in a marked reduction in PGE₂-induced mechanical hyperalgesia, evaluated after the last dose of Ift88 siRNA (Fig. 6B), and 72 h after the last dose of Ift88 siRNA (Fig. 6C).