Sex-specific role of RNA-binding protein, pAUF1, on prolonged hypersensitivity after repetitive ischemia with reperfusion injury.

Our previous findings have shown that male and female mice are hypersensitive following an I/R injury to the forepaw [ 57 , 58 ]. These results were replicated here in that when a single I/R injury was performed, both males (p=0.033) and females (p<0.0001) demonstrate acute fore paw guarding behaviors, one day following injury ( Fig. 1A - B ). When a second I/R injury was performed 7 days later in mice that previously experienced a single I/R, a prolonged guarding response is observed in both males (p<0.04) and females (p<0.0001). Males returned to baseline levels by Day 8; however, females still displayed significant paw guarding at the eight- day time point. This suggests that repeated I/R injury causes prolonged pain-like behaviors in both males and females, and this repeated injury model can be used to study the transition from acute to prolonged pain-like responses. Our previous findings have shown specific changes in muscle and DRG gene expression after I/R. For example, there is upregulation of muscle IL1β and its receptor, interleukin type 1 receptor 1 (IL1r1) within DRGs following a single I/R injury [ 56 , 57 ]. We have also reported changes in growth factors, specifically GDNF, to be upregulated in the muscle 1 day after an I/R injury as well as it’s receptor, GFRα1, increased in the affected DRGs of males which contributed to male hypersensitivity [ 53 , 58 ]. We therefore wanted to determine if a similar pattern is observed after repeated I/R. The amount of GDNF and IL1β was therefore quantified in the muscle one day after the second I/R. We found that males have an upregulation of GDNF (p=0.0114) but not IL1β, while females have an upregulation of IL1β (p=0.0122) but not GDNF after a repeated I/R injury ( Fig. 1C ). Immunohistochemical analysis of muscle tissue from males and females with repeated I/R confirm these findings ( Fig. 1D ). These results corresponded to the upregulation of the GDNF co-receptor, GDNF family receptor alpha 1 (GFRα1), at 1 day post single I/R as well as 1 day post double I/R injury in male mice (p<0.05). Upregulation of ASIC3 was also observed in males at both time points (p<0.05), similar to previous literature identifying a role of ASIC3 in altering muscle afferents that lead to pain-related responses [ 53 , 56 ]. No changes in TRPV1 were observed after either injury in males. Additionally, upregulation of IL1r1 is lost following a repeated I/R injury in males compared to the transient I/R injury as shown previously [ 56 ]. This could be due to distinctions in how males and females respond to single vs double I/R. Nevertheless, in female mice, there was observed upregulation of GFRα1 only one day after the second I/R injury. However, there was an upregulation in IL1r1 both 1 day after the first and second I/R injuries (p<0.05) in females. There is also an upregulation in TRPV1 in female DRGs (p<0.05) with no changes in ASIC3 ( Fig. 1E ). Together, this suggests that distinct dynamic signaling and receptor mechanisms in the periphery may underlie observed hypersensitivity in males and females after repeated I/R. To determine possible mechanisms by which distinct receptors could be differentially altered in male and female DRGs after I/R, we used both candidate screening and unbiased proteomics approaches to identify potential regulatory factors in the DRGs. Since initial insults often contribute to prolonged responses after second injuries [ 7 , 15 , 19 ], we compared DRGs from males and females under normal and single I/R conditions. We found several factors that were distinctly expressed in naïve and I/R injured male and female DRGs ( Fig. 2A - D ). Of the 1154 proteins identified to be expressed in our naïve samples, 53 were differentially expressed between uninjured male and female DRGs as shown in both heat maps and volcano plots. Interestingly, of the 507 proteins detected and confirmed in our I/R injured samples, only 16 were differentially altered after I/R between sexes. To complement the unbiased proteomics, we then assessed a series of RNA binding proteins in the DRGs that could also play a role in differential gene expression. We therefore assessed AUF1 and HUR, RNA-binding proteins that have many AU-rich target mRNAs, and found no changes in total HuR, phosphorylated HuR or HNRNPD (aka AU-rich element RNA binding protein 1: AUF1) between male and female DRGs. However, we did find a specific increase in the phosphorylated form of AUF1 (pAUF1) in naïve female DRGs compared to males ( Fig. 2E ). Based on this combination of screening assays and the specific proteins found to be different between males and females, we chose to focus the remaining experiments on pAUF1 as it had a strong potential to play a role in the distinct patterns of gene expression detected between males and females. We first performed IHC to determine if there were any differences in pAUF1 in relation to specific cell types between males and females. There was a significant increase in pAUF1 colocalized with NeuN in female DRGs compared to male DRGs. Often pAUF1+ cells also colocalized with TRPV1 or ASIC3 ( Fig 3A - C ). Since AUF1 has been linked to both stabilization and degradation of RNAs [ 2 , 11 , 23 ] and can be regulated via hormones that are known to be sex specifically expressed [ 5 , 26 , 46 ], we wanted to determine if these inherent differences in AUF1 activation (or expression after I/R) could underlie the differential gene expression patterns observed in male and female DRGs after injury. After a repeated I/R injury, qPCR confirmed an upregulation of GFRα1, ASIC3 and AUF1 in male DRGs. However, females show upregulated GFRα1, IL1r1, TRPV1 and even greater AUF1 ( Figs. 4A - B ). We therefore used our nerve-specific siRNA-mediated knock-down strategy [ 53 , 56 ] to inhibit the expression of AUF1 in sensory neurons after repeated I/R. We saw sex-specific effects of the AUF1 knockdown at the DRG level in that upregulation of GFRα1 and ASIC3 in the DRGs after repeated I/R in control injected male mice (PenCON+2xI/R) were not affected by the nerve targeted AUF1 siRNA injections (PenAUF1+2xI/R). However, in the female DRGs, sensory neuron AUF1 knock-down prevented the repeated I/R induced upregulation of both IL1r1 and TRPV1 (p<0.05 vs naïve) after injury. Few effects were observed on GFRα1 expression in females with repeated I/R ( Figs. 4A - B ). Next, we wanted to assess if this effect extended to our behavioral measures to determine if AUF1 had an impact on hypersensitivity post repeated I/R. In both spontaneous paw guarding and mechanical hypersensitivity to fore paw muscle squeezing, control (PenCON+2xI/R) and AUF1 targeted (PenAUF1+2xI/R) males with repeated I/R were unaffected by the AUF1 knockdown ( Fig. 4C - D ). When we analyzed female behavior, we saw significant differences between the two siRNA injected groups in that prolonged paw guarding was partially inhibited by AUF1 knockdown while mechanical hypersensitivity to muscle squeezing was completely inhibited by AUF1 targeting siRNAs in mice with repeated I/R compared to controls ( Fig. 4E - F ). In addition, grip strength was tested in separate cohorts ( Supplementary Fig. 3 ). Normalized to weight, there was no difference in grip strength in males or females between the two groups. However, the other two behavioral measures demonstrate that AUF1 plays a significant role in female specific hypersensitivity following repeated I/R injury. We then wanted to determine if AUF1 knockdown had any effect on TRPV1 or ASIC3 protein in the affected DRGs. Using IHC, in male mice with repeated I/R, both TRPV1 and ASIC3 were unaffected by the AUF1 knockdown ( Fig. 5A - B ). However, in female mice TRPV1 was significantly downregulated in the PenAUF1 group when compared to naïve and the PenCon group (p<0.0036 and p<0.04, respectively) while ASIC3 was unaffected ( Fig. 5C - D ). This shows that AUF1 can regulate gene expression at the DRG level in a sex dependent manner. Based on these data, we wanted to determine if upregulated AUF1 in male DRGs, would induce behavioral and gene expression changes like female mice with repeated I/R. Using an AAV9 containing an AUF1 overexpression construct (AUF1 OE), we injected the median and ulnar nerves in male mice and compared to control AAV9 injected males. We observed that with male paw guarding ( Fig. 6A ), the AAV OE group had prolonged hypersensitivity compared to the AAV injected controls. Although both control and AUF1 OE injected group demonstrated prolonged hypersensitivity to mechanical stimuli after repeated I/R when compared to baseline, AUF1 OE mice did display lower baseline thresholds to muscle squeezing ( Fig. 6B ). At the DRG level ( Fig. 6C ), nerve specific AAV9 mediated AUF1 overexpression increased IL1r1 (195 ± 29%), TRPV1 (404 ± 53%), and GFRα1 (378 ± 14%), but seemed to have prevented ASIC3 upregulation. This demonstrates that AUF1 is playing a role in IL1r1 and TRPV1 upregulation in the DRG following repeated I/R injury and that underlying sex differences in behavioral responses to dual I/R are due in part to altered levels of phosphorylated AUF1 in sensory neurons. Data thus far showed that pAUF1 partially regulated prolonged hypersensitivity following a repeated I/R injury in female mice. However, it was unknown what may be regulating pAUF1 in this model. Hormones are typically an area of interest, especially when sex differences are observed. We therefore tested whether estrogen may be a factor in the sex specific effects observed after repeated I/R injury using female mice with an ovariectomy (OVX). Surprisingly, no difference was observed between the OVX group and the non-OVX group for paw guarding behaviors ( Fig. 7A ). However, it does appear that the OVX group does not return to its original baseline level after the initial I/R like the non-OVX group. Regarding mechanical hypersensitivity, relative to baseline, there is an inhibition of mechanical hypersensitivity in OVX mice with dual I/R compared to controls; however, this may be due to a difference at baseline prior to the second I/R between the OVX and non-OVX group (p=0.0261) ( Supplementary Fig. 4 ). Since estrogen did not appear to play a role in the sex specific effects observed after repeated I/R, we therefore used a prolactin receptor antagonist, ΔPRL, to identify if prolactin may have a role in the observed I/R-induced behaviors and gene expression in females. Using ΔPRL during the second I/R reperfusion phase and prior to D1 behavior, females had partially inhibited guarding behaviors and mechanical hypersensitivity compared to the vehicle group ( Fig. 7B ). Although both groups still have prolonged hypersensitivity, there are significant differences between the two groups at varying timepoints. IHC showed decreased expression of pAUF1 (p<0.0001) in the DRGs of the antagonist group compared to vehicle ( Fig 7C ). Additionally, there was a significant decrease in TRPV1+ cells ( Fig. 7D ). These findings demonstrate a role for prolactin on female related hypersensitivity and DRG expression. Due to observations of upregulated TRPV1 expression in DRGs following injury in the female mice and previous literature demonstrating the regulation of TRP channels via prolactin and IL1r1 [ 47 , 48 ], an siRNA targeting strategy to down regulate expression of TRPV1 ( Fig. 7E ) in our DRGs of interest was used. At day 1 after injury, the PenV1 group had significantly lower guarding scores than the PenCon group (p<0.05) ( Fig. 7F ). Although the PenV1 and the PenCon groups both showed mechanical hypersensitivity over time, the PenV1 group showed less hypersensitivity at 6 days and were statistically different than controls at the 8-day time point. This not only demonstrated a role for TRPV1 in prolonged hypersensitivity in this model, but data also provided evidence of TRPV1 playing a partial role in muscle mechanosensation.

Adult male and female Swiss Webster mice (Charles River) 3–6 weeks of age were used throughout all experiments. Mice were housed in a barrier facility maintained on a 14:10-h light-dark cycle with a temperature-controlled environment and given food/water ad libitum. All procedures were approved by the Cincinnati Children’s Hospital Institutional Animal Care and Use Committee and adhered to NIH Standards of Animal Care and Use under the Association for Assessment and Accreditation of Laboratory Animal Care International (AAALAC)-approved practices. Mice were anesthetized with 2–3% isoflurane for surgical procedures. For I/R injuries, the brachial artery in the right forelimb was exposed and the artery was ligated using a 7.0 silk suture. Incisions were closed with a 6.0 silk suture after the occlusion. After 6 hours, animals were re-anesthetized for removal of the arterial occlusion suture. A second I/R injury was then performed 7 days after the first. Sham mice were also utilized by performing an incision and loosening of the vessel on each surgery day without performing the occlusion. To determine the most effective siRNA used for in vivo experiments, knockdown efficiency was tested via neuro-2A cells in vitro . In a 12 well plate, approximately 10 4 cells/well were seeded and incubated overnight (18–24hours). TransIT-TKO (Mirus) transfection reagent was prepared adding 2.5μL to 100μl of media without serum. Next, 10 μM siRNA stock solution from one of four different targeting duplexes or a non-targeting control was added to the transfection reagent mixture to create a 10nM final concentration per well. This was incubated at room temperature for approximately 20 minutes. Then each of the four siRNA duplexes (or control duplex) were distributed dropwise into separate wells containing cells in complete growth medium (Eagle’s Minimum Essential Medium, 10% Fetal Bovine Serum, and 1% Penicillin-streptomycin). Cells were incubated for approximately 24 hours and harvested for RT-PCR analysis to determine knockdown efficiency. The AUF1 siRNA found to have the greatest targeting efficiency in vitro ( Supplementary Fig. 1A - B ) was modified for binding to the lipophilic peptide, Penetratin-1 similar to our previous reports [ 52 , 53 , 56 ]. To knockdown AUF1 mRNA in vivo in mice with dual I/R injury, animals were injected with 0.1–0.2 μL of Penetratin-linked AUF1 (PenAUF1) targeting siRNAs or non-targeting controls (PenCon) into the median and ulnar nerves [ 53 , 56 ] on days 4 and 5 after the initial I/R using a quartz microelectrode connected to a picospritzer. Wounds were then closed with 6.0 silk sutures. To knockdown TRPV1 in peripheral nerves in vivo , these procedures were followed with a Penetratin-conjugated TRPV1 targeting siRNA purchased from Dharmacon (PenV1, Supplementary Fig. 2 ) or PenCon into the median and ulnar nerves. To overexpress AUF1 in select peripheral nerves in vivo , a similar injection procedure was followed except adeno-associated viruses, serotype 9 (AAV9), containing an AUF1 overexpression construct were injected into the median and ulnar nerves 3–4 weeks prior to the first I/R injury. AAVs with AUF1 the overexpression construct (AAV9-CAG-m-HNRNPD-IRES-eGFP, SKU: AAV-261522) or controls (AAV9-CAG-EGFP, Cat No: 7076) were purchased from Vector Biolabs and injected into the median and ulnar nerves in 1uL volumes as described above. After injections, incision sites were closed using 6.0 silk sutures. Prolactin receptor (Prlr) antagonist Δ1–9-G129R-hPRL (ΔPRL, MyBioSource, Cat#: MBS400130) is a modified PRL blocking the function of Prlr. 100 μl of ΔPRL (5μg per 100μl) or vehicle (RNase free water) were given via an intraperitoneal injection during the reperfusion phase of the second I/R injury and approximately 30 minutes before D1 behavior. Overall, mice received two doses of the ΔPRL or vehicle. Mice were anesthetized under 2–3% isoflurane and received dorsolateral incisions between the ribs and hips. The ovarian pad was pulled through the incision sites on each side and the ovaries were removed by applying pressure with a hemostat. The severed oviduct, uterine horn, and fat were returned to the abdominal cavity. The incision site was closed using 6.0 silk sutures, which were removed 7d post-surgery. Non-OVX mice received sham surgeries during this process. Following behavioral protocols, the uterus was removed and weighed to determine successful ovariectomies compared to the control group (see Supplementary Fig. 3 ). IR/shams were performed as described above, followed by behavior or tissue collection. Mice were tested for baseline behavior one day before injury as described previously [ 50 , 54 ]. Mice were analyzed at baseline prior to the initial I/R and then again at day 1 and 6 post I/R. After the second I/R, animals were again tested at days 1, 3, 6 and 8 post second injury. Each day, two behavioral tasks were performed: assessment of spontaneous paw guarding (non-evoked) and analysis of mechanical withdrawal thresholds to fore paw muscle squeezing (evoked). Mice were first placed individually in raised chambers with a mesh bottom and allowed to habituate for at least 15 minutes. Fore paw guarding was conducted by observing the mice for one hour, at five-minute intervals, assigned a score from 0–2 (0- mouse using all paws without reservation, 1- mouse isn’t bearing all weight on paws equally, 2- mouse holds paw completely off the mesh) after observation for 1 min. The average of the 12 trials was determined each day and then averaged across all mice within an experimental group. Mechanical hypersensitivity was performed on both forepaws using a set of digital calibrated forceps with a blunt probe. To assess paw withdrawal thresholds, mice were gently held and then had their plantar surface of the forepaw placed on the blunt probe side of the forceps. The paw was squeezed until the mouse elicited a withdrawal response. The force at which the animal withdrew was recorded. The cut-off intensity was set at 300g. Withdrawal thresholds were determined as a set of three trials, performed three times with a five-minute interval. The average of the three trials were used for analysis. In addition to this, fore paw grip strength was also analyzed in separate cohorts using a grip strength meter (BioSeb, France). Animals were held by the tail above a metal grid and allowed to grip the grid with only their forepaws. They were then pulled back horizontally until they could no longer retain their grip. The grip strength was measured in grams for three rounds of three trails with 5 minutes in between. The average of all nine trials was taken per testing day and corrected for weight differences for analysis. The experimenter was blinded to the treatment groups during all behavioral assays. Right DRGs (C7, C8, & T1) were collected from both male and female Swiss Webster mice following a single I/R injury and samples were run through comparative profiling to identify protein changes via SWATH nanoLC-MS/MS using runs from each sample to generate quantitative comparisons. The frozen tissue was placed in an Eppendorf homogenizer (Kontes749520–0000) and then 25 μl of homogenization buffer (10mM Tris with Roche complete protease inhibitor) was added and the sample homogenized for several minutes before an additional 10μl of homogenization buffer was added again and re-homogenized. Next, 25 μl of 2X Invitrogen sample buffer without dye was added and the sample was sonicated in a water bath for 1 minute followed by cooling on ice for 15 min for two cycles. The samples were then centrifuged at 1000xg for 10 minutes and the supernatant was placed in a new tube. Next the Pierce 660nm Protein assay was performed to determine the protein concentration using a BSA as a standard. 20 μg of each sample in 40 μl of Laemmli buffer were run 1.5 cm into a 1D, 1.5 mm 4–12% BT gel using MOPS running buffer. Pre-stained protein markers were used in surrounding lanes and the regions between the markers and the dye front were excised for trypsin digestion following a standard gel protocol. The resulting peptides were extracted, dried and prepped for mass spectrometry. 2.5 μg of each sample was run on the nanoLX-MS/MS in DDA mode and the combined DDA runs were searched using Protein Pilot to create a protein spectral library. From this, 1702 proteins were identified with a false discovery rate (FDR) of less than 1% at peptide and protein levels. A matched SWATH-MS method in DIA mode of the samples was used to collect quantitative data for each of the samples for comparative profiling (8 runs) and the total ion chromatograms were generated. SVVATH-D data analysis workflow was used to validate the data set and detect any significant quantitative changes. From this, 1175 proteins were identified and quantified with 99% confidence with FDR of less than 1%. A heat map of the 30 proteins with significant changes between the groups was generated (p<0.05) as well as a volcano plot of significant proteins (p value1.25 fold change) to show upregulated and down regulated proteins. Forepaw muscle was dissected from anesthetized mice and flash frozen in liquid nitrogen followed by storage at −80°C. After homogenization in protein lysis buffer containing 1% SDS, 10mM Tris-HCL (pH 7.4) and protease inhibitors (1μg/mL pepstatin, leupeptin, aprotinin, 1 mM sodium orthovanadate, and 100 μg/ml phenylmethylsulfonyl fluoride; Sigma-Aldrich). 20μg samples were boiled in gel loading buffer containing β-mercaptoethanol and sodium lauryl sulfate as a reducing agent and loaded onto AnyKD precast polyacrylamide gels (Bio-Rad 4569033) for protein separation. Proteins were transferred to polyvinylidene difluoride membranes (PVDF; Merck Millipore Ltd.) at 35 V overnight in a cold room at 4°C. The following day, the membrane was washed, blocked with buffer (Odyssey; LiCor 927–40000) diluted in PBS (1:4) and incubated in 2x PBS with 0.2% Tween and blocking buffer (1:1) with primary antibodies rabbit anti-AUF1 (Abcam, 1:5000), rabbit anti-phospho-AUF1 (Abcam, 1:2000), rabbit anti-Human antigen R (HuR) (Abcam,1:5000), rabbit anti-phospho-HuR (Abcam, 1:200), mouse anti-interleukin 1β (IL1β) (R&D Systems, 1:500), rabbit anti-glial cell line-derived neurotrophic factor (GDNF) (Alomone, 1:500), and chicken anti-glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (Abcam, 1:2000). After incubation over-night, the membranes were washed and incubated in 2x PBS with 0.2% Tween and 0.01% SDS and blocking buffer (1:4) with appropriate infrared-conjugated secondary antibodies (LiCor IRDye 680RD Donkey anti-Chicken 926–68075 Lot no. C80717 –13, IRDye 800CW Donkey anti-Rabbit 926–32213 Lot No. c80929–05, Goat anti-mouse IRDye 800CW Lot No C31021 –01). Membranes were visualized on LiCor Odyssey CLx protein imaging system. Exposure times were kept consistent between runs and gain was set to 1.0. Band intensity was quantified using ImageJ software (NIH) via previous procedures [ 58 ]. The immunoreactive bands were then analyzed via densitometry and then quantified through ImageJ software. The optical density was normalized to GAPDH and reported as protein quantification. To extract tissue, mice were first anesthetized via an intramuscular injection of Ketamine/Xylazine Solution (9 mg/mL Ketamine + 0.9 mg/mL Xylazine). For DRG extraction, post perfusion with ice cold 0.9% saline solution, a laminectomy was performed to expose the spinal cord and then the C7, C8, and T1 DRGs were isolated. For muscle extraction, skin and digits were removed from the right forepaw and the muscle was isolated and placed in disposable vinyl specimen molds with optimal cutting temperature (O.C.T.) compound. After tissue dissection, DRGs were immersion fixed in 4% paraformaldehyde for 30 min prior to being embedded in 10% gelatin (in MQ water) and fixed again overnight in 3% PFA. The gelatin blocks are then rinsed and placed in 20% sucrose (in 0.1M PB) overnight or until ready for use while muscle sections were flash frozen in liquid nitrogen. 45μm DRG sections were cut using a HM 430 sliding microtome, while muscle tissue was cut using a cryostat at 12 microns. Slides were mounted with subzero freezing medium (Mercedes Scientific) and allowed to dry on a slide incubator for at least 5 minutes. Cryostat sectioned tissue was fixed on the slides with 4% PFA in 0.01 M PSB for 15 min. Gelatin embedded DRG sections were placed in a 12-well plate and underwent the same treatment listed below. Slides/wells were rinsed in 0.1 M PBS and then underwent antigen retrieval, for 20 minutes in citrate buffer, at 70°C. Slides/wells were then blocked in 0.01 M PBS with 4% donkey serum, 4% horse serum, 1% bovine serum albumin, and 0.01M PBS with 0.1% Tween-20 for 1 hour. Then the slides/wells were incubated in the above blocking buffer overnight at room temperature with gentle rocking with the primary antibodies as indicated guinea pig anti-ASIC3 (1:2000, Millipore), rabbit anti-TRPV1 (1:2000, Alomone) or goat anti-TRPV1 (1:2000, Novus), mouse anti-IL1β (1:400, R&D Systems), rabbit anti-phospho-AUF1 (1:100, Abcam), rabbit anti-NeuN (1:500, Abcam) and rabbit anti-GDNF (1:100, Alomone). Next, slides/wells were washed in 0.01M PBS and incubated in the same blocking buffer but with appropriate secondary antibodies (647 donkey anti-guinea pig, 1:400, 488 donkey anti-mouse and 594 donkey-anti rabbit, Jackson Immunoresearch). The gelatin embedded DRG sections in the wells were then placed carefully on gelatin coated slides. The slides were then rinsed in 0.01M PBS and cover slipped using Fluro media with DAPI to stain nuclei. Labeling of cells was identified and characterized on a Nikon A1 inverted confocal microscope with sequential scanning. Images were then compiled and prepared for publishing via NIS-elements imaging software. For quantification, non-sequential sections of samples were counted to prevent any overlapping cells from being quantified more than once. Cells that were positive for one or more antibodies were counted separately and each marker was quantified for each section. Counts were performed on 3–5 sections per single animal and the average counts per condition were used for analysis (n=3–4). Positive cells were counted via ImageJ using each individual channel, as well as combined channels to assess colocalization. The percentage of an indicated marker colocalized with NeuN positive cells was reported where indicated. Negative controls were performed by staining the slides with only secondary antibodies (i.e., no primary) to confirm some specificity. To note, there is not an available knockout of AUF1 to test primary antibody specificity for this marker. Additionally, positive pAUF1 cells were counted via nuclear or cytoplasmic staining as it is present in both cellular compartments. Using previously published procedures and primers [ 52 , 53 , 56 , 57 ], C7, C8, and T1 DRGs were dissected as described previously and frozen on dry ice until all dissections were completed (maximum of 4 hours). Samples were then used for RNA isolation using Qiagen RNeasy Mini kits according to the manufactures’ directions. 500ng of total RNA was then reverse transcribed into cDNA by first incubating in 10x DNase buffer, RNasin, and DNase I to eliminate any residual genomic DNA, followed by reaction with superscript II reverse transcriptase (0.25μL) for 50 minutes at 42°C. Samples were stored at −80°C until needed for realtime PCR. To assess quantification and purity, an Epoch microplate reader was used with the program Gen5. To perform real-time RT-PCR 20 ng of cDNA per well was used with SYBR-Green real-time PCR reactions (applied biosystems) and necessary primers on a StepOne real-time PCR System (Applied Biosystems). Ct values were obtained and analyzed by the ΔΔCt method after normalization to GAPDH. Expression differences are determined from the normalized ΔΔCt values and standard error of the difference in means. Fold change between conditions is determined and values were converted to a percent change where 2-fold = 100% change [ 53 ]. MIQE guidelines were followed to ensure integrity [ 9 ]. Data were analyzed using GraphPad Prism 9.5.0 and SigmaPlot. Critical significance was set to α < 0.5. All data were checked for normal distribution and equal variance using the Shapiro-Wilk normality tests and then parametric or nonparametric tests were used accordingly. All behavioral tests are specifically indicated in the figure legends. Behavioral analyses were analyzed via a two-way repeated measures ANOVA with Holm-Sidak and/or Tukey’s post hoc tests. Analysis of protein quantification from western blots were measured via unpaired T-test or corresponding non-parametric test across groups. RT-PCR was analyzed via one-way ANOVA with Holm-Sidak post hoc test. All analyses that passed the omnibus test were further assessed by Tukey’s, Holm-Sidak, or Dunn’s post hoc test analysis as noted in figure legends. One instance of a statistical outlier in the control group from the TRPV1 siRNA experiments defined as being more than two standard deviations away from the mean was removed. In all studies, the researcher was blinded by co-investigators or by the unknown genotype of each animal.

This study investigated the sex-specific role of sensory neuron pAUF1 in a newly developed prolonged ischemic myalgia model. We found that repeated I/R induced prolonged hypersensitivity in both male and female mice [ 57 , 58 ], but males recovered more rapidly than females ( Fig. 1 ). Currently, it is unclear when females may resolve this hypersensitivity, although by day 8 there is a trend towards baseline. Our previous findings showed specific muscle and DRG gene expression changes after I/R. For example, there was upregulation of muscle IL1β and GDNF and their respective receptors, IL1R1 and GFRα1, within DRGs following a single I/R injury [ 53 , 56 , 57 ]. Interestingly, after repeated I/R, we detected sex-specific muscle and DRG expression patterns ( Figs. 1 – 5 ). Specifically, there was an upregulation of GDNF (but not IL1β) in the muscles as well as an upregulation of ASIC3 and GFRα1 in the DRG of males, supporting previous literature identifying a role of ASIC3 and GFRα1 in altering muscle afferents that underlie pain-related behaviors [ 53 , 56 ]. Females however, showed an upregulation of IL1β (but not GDNF) in the muscles and increased expression of IL1r1 and TRPV1 in the DRGs with no changes in ASIC3 ( Fig. 1 ). Studies have noted an association between IL1β and TRPV1, particularly in ischemic conditions [ 74 ]. Furthermore, in chronic inflammatory diseases there is an upregulation of IL1β and high expression of IL1r1 in TRPV1+ DRG neurons [ 36 ]. Together with the fact that IL1β can increase TRPV1 expression in sensory neurons [ 33 , 36 ], this suggests that distinct signaling and receptor mechanisms in the periphery may underlie observed hypersensitivity in males and females after repeated I/R. Studies of acute ischemic injury showed that IL1β/IL1r1 signaling may have a role in regulating hypersensitivity [ 56 ]. Here we showed repeated I/R increased muscle IL1β and DRG IL1r1 expression in females but not in males ( Fig. 1 ). The sex-specific activation of this pathway after repeated I/R injury may be due to several currently undetermined factors, but data herein suggests enhanced p-AUF1 plays a significant role ( Figs. 2 – 6 ). There is a reported connection between AUF1 and IL1β/IL1r1 signaling [ 10 , 34 , 61 ], and AUF1 has 7 binding motifs on IL1r1 mRNA. Therefore, it is possible that female specific IL1β/IL1r1 signaling in the DRGs may be mediated via differential AUF1 activation. siRNA knockdown of AUF1 in female sensory neurons during repeated I/R displayed inhibited behavioral hypersensitivity and downregulation of IL1r1 in the DRGs. There was no effect of AUF1 knockdown on male behavior or DRG gene expression ( Figs. 4 – 5 ). It is important to note that AUF1 knockdown only affected IL1r1 and TRPV1 but did not regulate expression of GFRα1 leading us to postulate that AUF1 may uniquely regulate distinct factors in DRGs. Nevertheless, overexpression of AUF1 in male sensory neurons prolonged I/R-related behaviors and increased DRG expression of AUF1, IL1r1, TRPV1, and GFRα1 ( Fig. 6 ). Interestingly, GFRα1 was upregulated, comparable to what is seen in males following I/R injury, but injury-enhanced ASIC3 was blocked. Currently, there is no literature on interactions of AUF1 and ASIC3. We could speculate that artificially enhanced AUF1 mediated alterations in IL1R1 in male DRGs caused a compensatory inhibition of the ASIC3 upregulation, but how AUF1 overexpression in male sensory neurons modulates ASIC3 will need to be investigated in future studies. Additionally, the mechanism by which AUF1 is phosphorylated, as total AUF1 between males and females is similar, could involve prolactin. Prolactin is shown to be upregulated both at the site of injury and systemically after inflammation [ 47 ]. Using a prolactin antagonist, we show this intervention inhibited pain-related behaviors in our model ( Fig. 7 ). Additionally, there is a link between prolactin and TRP channels [ 48 ]. TRPV1 normally contributes to thermal sensitivity from the skin. However, literature has demonstrated that TRPV1 contributes to mechanical hypersensitivity from the muscles [ 68 , 73 , 74 ]. Further, our reports [ 26 ] have shown that heat sensitive muscle afferents do not often contain TRPV1. In fact, some are mechanically sensitive. When TRPV1 is reduced in the DRG, we see partially inhibited mechanical hypersensitivity ( Fig. 7 ). Similar studies have shown a relationship between IL1β, IL1R1 and TRPV1, in that when IL1R1 is deleted in TRPV1+ nociceptors, mechanical allodynia is prevented [ 36 ]. These data indicate that pAUF1 is a factor partially driving sex-specific differences in I/R-related hypersensitivity which may be regulated via prolactin, IL1β signaling and TRPV1 ( Fig. 8 ). Inflammatory cytokines, like IL1β and its receptor, are involved in the pathogenesis of endometriosis, an estrogen-dependent disease often associated with chronic pain [ 27 ]. The involvement of IL1β/IL1r1 has also been shown in breast cancer studies which show that IL1β/IL1r1 expression was mediated via estrogen [ 4 , 30 ]. Literature has also demonstrated that estrogen and prolactin can regulate IL1β production from macrophages, which we have suggested to be a source of IL1β [ 56 ], while testosterone may inhibit its release [ 55 , 62 ]. However, behavioral data here does not support a strong role for estrogen in sex specific I/R related hypersensitivity as OXV mice did not show significant alterations compared to controls ( Fig. 7 , Suppl. Fig. 4 ). It is possible that repeated injuries mask any effect estrogen may have on acute pain-like behaviors. The roles that estrogen and testosterone may play in repeated I/R injury induced hypersensitivity are crucial in future studies. However, we demonstrate a role of prolactin after repeated I/R. Prolactin has been linked to numerous pain disorders like rheumatoid arthritis and lupus, which disproportionately affects women [ 25 ]. Prolactin has been suggested to have a role in sensitizing channels, like TRPV1 after inflammation [ 48 ], but our data shows there are likely other targets in models of ischemic myalgia ( Fig. 7 ). AUF1 can be regulated via hormones [ 5 , 24 , 46 ], so prolactin may be part of the mechanism disproportionately activating AUF1 (and subsequently TRPV1) in female DRGs compared to the males. This report is the first to describe a role for RNA binding proteins in sensory neurons that modulate a sex specific gene expression program which regulates pain-like hypersensitivity after repeated I/R injury. A new model of prolonged hypersensitivity was also established with these studies. We uncovered a novel DRG protein with enhanced phosphorylation in females that appears to uniquely modulate IL1R1 and TRPV1 expression in DRGs. This was determined using our novel nerve-specific siRNA mediated knockdown strategy and AAV mediated overexpression technologies. Thus, innate differences in male and female DRGs can contribute to I/R related hypersensitivity through distinct gene regulatory mechanisms. It would be important in the future to determine the neurochemical and physiological phenotypes of the sensory neurons driving pAUF1-mediated effects in our model of prolonged I/R-related hypersensitivity as well as the exact mechanism linking pAUF1, IL1r1 and TRPV1. Although AUF1 has been extensively studied, the effects that each isoform can have in any cell type is not fully clear. It has been shown to both stabilize or degrade early response genes, cell cycle regulators, as well as cytokines [ 69 ]. AUF1 is known to have functional flexibility as it can bind with high affinity but with widely varying functional effects that can also be altered with protein modifications, like phosphorylation. This study also did not determine how AUF1 was sex-specifically activated in female DRGs. It is plausible that hormones could regulate this differential activation pattern [ 23 , 46 , 47 , 55 , 57 , 63 ] but this will need to be determined in the future along with how AUF1 or hormones can regulate cytokine production in other cell types linked to pain [ 20 , 31 , 57 ]. Lastly, our results show slight differences in pain-related behaviors between the groups receiving repeated I/R injuries and the groups that also received siRNA nerve injections ( Figs. 1 , 4 , 6 ). It is possible that four procedures may have influenced hypersensitivity (and gene expression) between these two distinct injury groups. However, proper controls within experiments do indicate a strong role for Prlr, pAUF1, and TRPV1 in female related I/R hypersensitivity. Myalgia affects multiple biological systems [ 45 , 60 , 65 , 66 ]. Repeated I/R injury occurs in many disorders with MSK pain as an underlying feature, including peripheral vascular disease, sickle cell anemia, fibromyalgia, and complex regional pain syndrome [ 1 , 13 , 16 , 18 , 28 , 51 , 67 ]. Many MSK disorders that cause chronic pain affect men and women differently [ 6 , 14 , 39 , 42 , 43 , 50 , 57 , 63 ], and in some conditions (e.g., migraines, fibromyalgia, etc.), women often exhibit greater pain severity [ 6 , 35 , 57 , 64 , 65 , 70 ]. Our results are some of the first to study sex differences in ischemic myalgias, a common cause of pain in diseases that affect women more than men [ 50 , 72 ]. Current treatment strategies for diseases associated with pain and I/R injuries include opioids and exercise. However, these therapies are often not feasible due to exercise intolerance or pharmacological ineffectiveness in patients with I/R-related injuries [ 21 , 32 , 60 ]. This highlights the importance of better understanding how pain develops in these unique perfusion disorders. Regardless, the current report indicates a sex specific role for DRG expressed RNA binding proteins in I/R-related hypersensitivity.

Musculoskeletal pain is a major health concern and is experienced more than any other type of pain worldwide [ 3 , 17 , 22 , 38 , 40 , 50 ]. Chronic musculoskeletal pain (myalgia) specifically, defined as pain persisting beyond normal healing time, is a prevalent burden causing years of disability [ 8 ]. Pain is often widespread throughout the body, making strategies to treat the underlying issue difficult. Myalgia can be due to numerous underlying causes: overuse or strain of muscle, traumatic injury, or disease. A major factor of disease-based myalgia is ischemia with reperfusion (I/R) injury. Repeated I/R injury occurs in many disorders such as peripheral vascular disease, sickle cell anemia, or even fibromyalgia and complex regional pain syndrome [ 1 , 13 , 18 , 28 , 51 , 67 ]. Diseases like these are often reported more in the female population, where pain tolerance is also lower [ 12 , 16 , 20 , 41 , 50 , 71 , 72 ]. The mechanisms of sexually dimorphic pain hypersensitivity have been linked in part to the immune system [ 37 , 55 ] with far less known about sex differences in the peripheral nervous system. At the cellular level, there is emerging evidence that different male and female mechanisms may govern pain responses [ 37 , 44 , 51 ]. Although animals of both sexes show similar pain-like behaviors following I/R injury, analysis of the dorsal root ganglia (DRGs) showed that I/R evoked an increase in acid-sensing ion channels (ASIC) 1 and 3 in males, while females showed an upregulation of transient receptor potential vanilloid type 1 (TRPV1) as well as TRP Melastatin 8 (TRPM8) [ 57 ]. This suggests that distinct molecular pathways may underlie pain-like behaviors after I/R between males and females. One possible factor that may affect the sex-specific mechanisms within DRG neurons leading to prolonged hypersensitivity could involve RNA binding proteins which modulate gene expression [ 2 , 29 , 49 ]. One such protein is AU-rich element RNA binding protein factor 1 (AUF1, aka, HNRNPD) which has been shown to regulate cytokine-related mRNAs and can have either stabilizing or degrading effects, depending on the isoform [ 59 , 73 ]. AUF1 binds transcripts encoding cytokine receptors and plays a role in inflammatory responses [ 69 ], which have been suggested to be important regulators of female specific responses to I/R [ 57 ]. The regulation of AUF1 is of interest as post-translational modifications, like phosphorylation, can alter its activity which can affect its ability to regulate gene expression levels [ 69 ]. Additionally, AUF1 is known to be regulated via hormones [ 23 , 24 ], including prolactin (PRL) which is increased in numerous pain conditions and plays a prominent role in inflammatory pain in females specifically [ 47 , 48 ], possibly through TRPV1 neurons [ 48 ]. PRL acts through the JAK./STAT pathway [ 54 ], which can regulate proteins such as AUF1 [ 23 ]. We therefore wanted to investigate the role of sensory neuronal pAUF1 and PRL signaling in a model of prolonged hypersensitivity after repeated I/R injury. Highlighting the potential sex specific mechanisms leading to pain under these conditions could be beneficial for future preclinical and clinical studies and lead to novel sex specific therapies for ischemic myalgia.