Transcriptome profiling suggests molecular sexual dimorphism in lumbosacral dorsal root ganglia and sex-specific mechanisms underlying visceral pain.

First, due to the anatomical differences in the urinary tract, urethral catheterization in male mice was challenging and could potentially lead to local (urethral) inflammation. In order to nullify this effect, saline-instilled animals were used as control groups. And naïve groups (no instillations) were not included. Therefore, the “baseline” transcriptome analysis in our data is from mice received three urethral catheterizations. In females, saline instillations slightly but significant increased visceral mechanical sensitivity, suggesting potential saline-induced immune responses in female animals. Second, bulk RNA-seq does not distinguish the cellular sources of DEGs associated with visceral pain symptoms. The specific cell type in lumbosacral DRG responsible for visceral hypersensitivity requires future investigation. Third, only one timepoint were assessed in this study. This timepoint was chosen as the earliest peak of visceral mechanical sensitivity based on previous studies. However, it is possible that other forms of nociceptive responses show their peaks at a different time point. More importantly, although our study demonstrated strong sex differences in lumbosacral DRG during visceral hypersensitivity, we cannot rule out the possibility that similar signaling changes occur in both sexes, but at different time points. Future studies are needed to show if male animals exhibit similar immune and ECM activation in their lumbosacral DRG following VEGF instillation as those in female DRG, only at a later time point.

All mice were subjected to mechanical sensitivity testing before (day 0) and after bladder instillations (day 15) ( Fig. 1A ). Male and female animals displayed similar visceral mechanical sensitivity at the baseline (p=0.3672, males: N=12, females: N=10) before receiving intravesicular instillations ( Fig. 1B ). Force-induced increases in visceral mechanical sensitivity was observed (2-way ANOVA, p<0.0001). Subsequently, male and female mice were randomly assigned into groups to receive either saline or VEGF instillations (male saline: N=5, male VEGF: N=7, female saline: N=5, and female VEGF group: N=5). There was no significant difference in baseline visceral mechanical sensitivity between female or male mice assigned to different experimental groups ( Fig. 1C ; 3-way ANOVA, p=0.8037 between treatment groups). Following bladder instillations, the mice received saline instillations showed a moderate increase in withdrawal frequencies ( Fig. 1D , 3-way ANOVA, p=0.0017). The sex-specific effect of saline instillations also resulted in the sex differences detected in Fig. 1D (3-way ANOVA, p=0.0011). Consistent with our previous reports ( 6 , 7 ), VEGF instillations led to significant increases in visceral mechanical sensitivity in both male and female mice ( Fig. 1E ; 3-way ANOVA, p<0.0001). No sex differences were detected in the VEGF-induced visceral hypersensitivity (2-way ANOVA, p=0.2861). Despite the saline-induced changes in withdrawal frequencies, female animals that received VEGF 165 instillations showed significant increases in withdrawal frequencies compared to those received saline instillations ( Fig. 1F , 2-way ANOVA, p<0.0001 between VEGF and saline-instilled females), indicating VEGF-induced visceral mechanical hypersensitivity. Male animals showed similar VEGF-induced increases in visceral mechanical responses ( Fig. 1G , 2-way ANOVA, p<0.0001 between VEGF and saline instilled males). In summary, bladder VEGF instillations led to visceral mechanical hypersensitivity in both sexes of animals. Following VFT, lumbosacral DRG (L1-L2, L6-S2) were harvested from four animals per experimental group. RNA was extracted from the DRG of each animal and quality check (QC) was performed ( Fig. 2A and Supplementary data 1). Approximately ~16000 genes were annotated in each experimental group (Supplementary data 2). Next, DESEQ2 was used to identify DEGs in the lumbosacral DRG between VEGF-instilled and saline-instilled animals (fold change > 1.5 or < −1.5, p<0.05). A total of 402 DEGs and 555 DEGs were identified in the lumbosacral DRG isolated from male and female VEGF-instilled mice, respectively ( Fig. 2B ). The top changed DEGs in male DRG include genes involved in gene transcription, cell growth and proliferation, apoptosis, cytoskeletal and extracellular matrix organization, energy metabolism, PAR and Ras signaling, ion channels, and neurotransmitter release/uptake ( Fig. 2C and Supplementary data 2). The top changed DEGs in female DRG include genes involved in gene transcription, protein repair and trafficking, cell differentiation, cytoskeleton remodeling, potassium channel, inflammatory responses, and neuropeptide signaling ( Fig. 2D and Supplementary data 2). Interestingly, VEGF-induced DEGs in male and female lumbosacral DRG showed little overlap ( Fig. 2B ). Only 22 VEGF-induced DEGs are found in both sexes, including pseudogenes ( Table 1 ). Four ( 4 ) genes, including Slx4ip , Ryr3 , and Gpr165 were consistently upregulated in VEGF-instilled animal DRG in both sexes (marked in red in Table 1 ). These genes are thought to be involved in DNA maintenance and repair, calcium releases from internal stores, and neuropeptide signaling, respectively. Eight ( 8 ) genes including Emilin3 , Angptl4 , Fcor , Sfrpr4 are consistently downregulated both male and female DRG from VEGF-instilled animals (marked in blue in Table 1 ). These genes are known to be involved in extracellular matrix organization, lipid metabolism, neurodevelopment, and cell growth and differentiation, respectively. The rest of shared DEGs are differentially regulated in male and female DRG, such as Pyroxd2 (regulating mitochondrial function, downregulated in females but upregulated in males) and Myrip (regulating exocytosis, upregulated in females but downregulated in males). These data indicated that bladder VEGF signaling induced sex-specific gene changes in lumbosacral DRG, suggesting distinct biological processes might be responsible for the pathogenesis of visceral hypersensitivity in female and male animals. We next performed hierarchical gene clustering analysis on data from all four experimental groups ( Fig. 3A – B ). This analysis first compares each experimental group to all other experimental groups, in order to identify enriched gene in this sex/experimental condition, then groups genes with similar expression patterns into a hierarchy, in order to identify relationships and patterns in gene expression data. Interestingly, sex-biased gene enrichment can be detected not only in VEGF-induced gene expression changes, but also between saline-instilled animals ( Fig. 3B ). The heatmap in Fig. 3B represents enrichment scores of specific gene expressions in each individual animal. Genes known for their immune functions are highly enriched in saline-instilled female animals (yellow box), such as genes encoding interferon gamma ( Ifnγ ), Z-DNA binding protein 1 ( Zbp1 ), suppressor of cytokine signaling 1 (Socs1), macrophage inflammatory protein 1 gamma ( Mip1γ/Ccl9 ), interferon regulatory factor 1 ( Irf1 ), thrombospondin-1 ( Thbs1/Tsp1 ), insulin-like growth factor 1 (Igf1), prostaglandin D2 synthase ( Ptgds ), B-cell CLL/lymphoma 10 ( Bcl10 ), and tyrosine kinase 2 ( Tyk2 ). The observed female-specific enrichment of immune response genes is consistent with a previous report on sex-biased immune gene expression across tissues ( 17 ). Our data also showed strong extracellular matrix (ECM) gene enrichment in saline-instilled female DRG, such as genes encoding collagen I and III ( Col1a1 , Col1a2 , Col6a1 , Col6a2 , Col6a3 ) as well as Tsp1 . ECM changes has been implicated in tissue homeostasis ( 19 , 20 ) and cell-cell communication in the nervous system ( 21 ). It is unsure how female enrichment of ECM genes at the baseline means to lumbosacral DRG changes during visceral pain. Not surprisingly, VEGF instillations induced increased nociceptive gene expression in both male and female lumbosacral DRG ( Fig. 3B , green dots). However, the enrichment has a strong sex-bias towards male DRG (green boxes). These genes include well-known nociceptor genes such as Trpv1 , Trpa1, and Tmem100 ( 22 ), mechanosensitive ion channel gene Peizo2 and Kcnk1 , cation channel genes such as Trpm7 , Trpm8 , Cacna1b and cacna1e , genes encoding neurotransmitter and peptide receptors such as purinergic receptor P2X2, adenosine A2B receptor, and neuropeptide Y. These gene upregulation in VEGF-instilled animal DRG remains consistent with previous report on nociceptor upregulation in other chronic pain models. However, the sex-biased enrichment of these genes in male DRG suggest the potential sex differences in the contribution of nociceptive signaling pathways to visceral pain. Next, we used gene set enrichment analysis (GSEA) ( 23 ) to identify significantly changed cellular functions and biological processes associated with visceral hypersensitivity in the lumbosacral DRG ( Fig. 3A ). We first performed additional GSEA analysis to compare the DEGs between female and male DRG from saline-instilled experimental groups. Our analysis suggested a strong female-biased ECM-receptor interactions at the baseline (Supplementary Fig. 1). For example, female DRG showed more potent collagen and laminin interaction with VLA proteins, suggesting more active leukocyte function. Our data also indicated more fibronectin and Thrombospondin-1 (THBS1) expression in saline-instilled female DRG at the baseline, suggesting possible stronger cell adhesion, migration, and differentiation in female DRG at the baseline compared to male DRG. Next, we analyzed VEGF-induced changes in male and female DRG, resulting in 39 significantly altered cellular and biological processes ( Fig. 3C – D , enrichment score >=4.0). Striking sex differences were detected in these biological processes, with all the glial- and lipid-related functions changed only in males, and nearly all the immune- and ECM function-related functions changes only in females ( Fig. 3C ). Actin cytoskeleton pathways were identified in both sexes. Figure 3D illustrates all the 39 identified biological functions with enrichment score coded by color. The most robust changes of biological processes are the innate immune response, which only changed in female DRG. Taken together, these data suggested strong immune activation and elevated Treg cell responses in female DRG following bladder VEGF instillations, which is likely response for the development of visceral hypersensitivity in female animals following VEGF-induced immune responses in the urinary bladder and sensory afferent. On the other hand, male DRG showed strong activation of glial activation and lipid metabolism, suggesting alternative mechanisms in increasing afferent activity. After gene and gene set analysis, we were set out to identify molecular signaling pathways that drive the biological process changes identified. Ingenuity Pathway Analysis (IPA) was used to evaluate the canonical signaling pathway changes in lumbosacral DRG associated with visceral hypersensitivity ( Fig. 4A ). Due to the high numbers of genes needed for performing IPA, DEGs were re-identified in Partek using p<0.1 as a criteria instead of p<0.05 (Supplementary data 3) resulting in 656 male DEGs and 957 female DEGs. We first performed IPA analysis based on the DEGs with pseudogenes included and the DEGs without pseudogenes. Our analysis showed an almost identical outcome between these two conditions (Supplementary figure 2); therefore, we will be discussing IPA results without pseudogenes included in the DEGs (p1.5-fold changes). Given the significant baseline difference (females had higher baseline levels of ECM and immune-responsive genes), it was challenging to identify pathways altered exclusively due to VEGF- instillation. Both of the ECM and immune pathways are recognized as major drivers in the development of chronic pain ( 24 , 25 ). To offset the effect of baseline sex differences and identify pathways specifically altered in the lumbosacral DRG due to VEGF treatment, we conducted a comparative analysis combining male (n=8) and female (n=8) samples, with treatment (saline vs. VEGF) as the variable parameter ( Fig. 4B ). The activation z-score are assigned considering changes in the hundreds of upstream and downstream genes and master regulators at the network level (≥2 considered significant). This analysis revealed seven pathways altered in both sexes due to VEGF treatment. Compared to saline-treated DRG, VEGF-treated DRG showed activation of adrenergic and serotonin receptor signaling, neutrophil extracellular trap signaling, eicosanoid signaling, DHA signaling, CCXCR4 signaling, and synaptic long-term depression signaling pathways, suggesting their role in visceral hypersensitivity development. Based on the baseline effect observed in figure 3B , we anticipated differences in VEGF response between male and female mice. To investigate this, we compared male and female dataset separately ( Fig. 4C ). In this analysis, pathways upregulated due to VEGF instillation will have positive Z-score and pathways downregulated due to VEGF instillations will have negative Z-score. Pathways related to cholesterol metabolism like LXR/RXR activation, PL metabolism were affected in male mice. On the other hand, pathways related to ECM and immune system like Antigen presenting pathway, Interferon alpha/beta signaling, Assembly of collagen fibrils and other multimeric structures were affected in female mice. Interestingly, pathways like Acute phase signaling and Ion channel transport are affected in both sexes. However, the opposite directionality indicates the sexual dimorphic response to VEGF-instillation. Furthermore, our analysis further illustrates the sex differences at the baseline (between saline-instilled groups) and during visceral pain symptoms (between VEGF-instilled groups) ( Fig. 4C ). Compared to male DRG, female DRG showed significant less active adrenergic and serotonin receptor signaling, neutrophil extracellular trap signaling, eicosanoid, DHA, and CCXCR4 signaling, as well as synaptic long term depression signaling pathway. However, in VEGF-instilled mice, all these signaling pathways are significantly more active in female DRG compared to male DRG, suggesting their contribution in the development of visceral hypersensitivity in female animals. We next focused on VEGF-induced changes in nociceptive genes in the lumbosacral DRG using IPA ( Fig. 5 ). Twenty-one ( 21 ) significantly changed DEGs were identified in male DRG and sixteen ( 16 ) were identified in female DRG ( Table 2 ). There were two overlapping nociceptive DEGs between male and female groups ( Table 2 , highlighted in yellow), p2ry6 (encoding pyrimidinergic receptor P2Y6) and ptpn5 (encoding protein tyrosine phosphatase non-receptor type 5). However, only p2ry6 were downregulated in both male and female DRG, whereas ptpn5 is upregulated in female DRG but downregulated in male DRG. The VEGF-induced nociceptive DEGs in male DRG include genes encoding neuropeptide and growth factors ( Nts , Penk , Cck , Hgf , Wnt5A , and Agt ), ion channels ( Kcnq1 ), transporters ( Slc6a4 , Slca7l1 , and Slc6a91 ), and GPCRs ( P2yr6 , Drd2 , Rqamp1 , Grm3 , Ntsr2 , and Cx3cr1 ) ( Fig. 5A ). Transcription regulator gene Btg2 was also found to be differentially expression in male DRG ( Fig. 5A ). In female DRG, VEGF induced nociceptive DEGs encoding cytokines ( Edn1 , Ngf , and Clec11a ), ion channel protein Scn4a , GPCRs ( Oxtr , Calcrl , Hcrtr2 , Mmgprf , and P2ry6 ) ( Fig. 5B ). Several peptidase genes including Mme , Psmb8 , and Ctss were also differentially expressed in female DRG ( Fig. 5B ). VEGF-induced changes in nociceptive signaling pathways in lumbosacral DRG are illustrated in Figure 5C – D . Male DRG DEGs strongly suggested the activation of nociception and mechanical nociception signaling pathways as well as decreased analgesia signaling ( Fig. 5C ) following VEGF instillations. These data suggest nociceptive response in male DRG underlies the visceral pain symptoms in mice. On the other hand, female DEGs in their lumbosacral DRG suggested strong activation of analgesia pathways and inhibition in the hyperalgesia signaling pathways, with only mild activation of nociception pathways ( Fig. 5D ). These data suggest that the active effort of resolving visceral hyperalgesia in female DRG following VEGF instillations ( Fig. 5D ). Ngf expression is significantly downregulated in female DRG following VEGF instillations ( Fig. 5 ). Next, we performed analysis on NGF signaling pathway. A total of 77 DEGs were identified in NGF signaling pathways (Supplementary File 4). The VEGF-induced DEG list showed strictly sex-specific regulation, with females displaying more active transcriptional engagement (24 upregulated and 22 downregulated) and males showing weaker activation (12 upregulated and 19 downregulated). Notably, all genes in this dataset were sex-unique, with no overlap between female and male DEGs. Both sexes exhibited downregulation of the immediate-early transcription factor Egr1 , indicating a shared attenuation linked to neurite sprouting. However, IPA analysis also showed that female DRG displayed broader activation of NGF signaling, including upregulation of Acta1 , Atp2a1 , Atp2a2 , Mef2c , Map2 , Raf1 , Edn1 , and Zeb1 , reflecting enhanced cytoskeletal remodeling, calcium handling, and transcriptional regulation ( 26 , 27 ) (Supplementary Figure 4). In contrast, males showed selective upregulation of Atm , Fanca , Fgf2 , Met , Sox11 , Wnt5a , and Yes1 , indicating greater activation of DNA repair, growth factor, and stress-adaptive pathways ( 28 ) but a weaker NGF pathway engagement (Supplementary Figure 4). IPA also revealed striking sex-specific differences in the neurite growth/sprouting signaling pathway, which was strongly activated in male DRG but inhibited in female DRG following VEGF instillations ( Fig. 6 ). We next investigated to what extent the androgen and estrogen signaling pathways changed in lumbosacral DRG following intravesical instillations of VEGF 165 . Androgen signaling pathway showed pronounced sex-biased transcriptional programs, with of genes (82 out of 108) uniquely regulated in one sex (Supplementary data 5). The relatively small, shared subset and the presence of genes with opposite regulatory patterns ( Atp2a1, Atp2a2, Ccn2, Fcmr, Ly6a, Nupr1, Ptgds, Synm and Twist1 ), suggesting sexually dimorphic AR-dependent transcriptional responses (Supplementary Figure 5). In female DRG, Sting1, Ngf, Edn1, Ptgds, Mme and Scn4a were differentially expressed, suggesting enhanced modulation of nociception ( 34 , 36 , 38 ) . In males, Penk, Nts, Kcnq1, and Slc7a11, Btg2 , and Hap1 were differentially expressed, indicating androgen-driven divergence in nociceptive processing. Together, these findings suggested that sex hormones intersect with sex-specific nociceptive networks ( 40 , 43 , 46 , 49 , 52 ). With respect to the estrogen signaling pathways, we considered 262 genes related to Esr1 (Estrogen Receptor Alpha) and Esr2 (Estrogen Receptor Beta), both nuclear hormone receptors mediating estrogen effects. In the female DRG, 103 genes were upregulated and 76 downregulated, whereas 51 DEGs were upregulated and 40 downregulated in male DRG (Supplementary Figure 6). In female DRG, Ptgds and Ptgis (prostaglandin-mediated inflammatory pain), Ctss and Edn1 (neuroinflammation and immune activation), and Mme and Oxtr (neuropeptide and opioid modulation) were differentially expressed, suggesting potential modulation of nociception ( 41 , 44 ) In male DRG, Slc6a4 and Slc7a11 were upregulated, indicating enhanced central sensitization and oxidative stress–linked pain pathways ( 47 , 50 ). Together, these results show that estrogen signaling intersects with sex-specific nociceptive networks, where females exhibit a predominantly analgesic transcriptional profile, while males show a pro-nociceptive bias. The pathway analysis for androgen signaling pathways can be viewed in Supplemental Figure 5. IPA showed strong androgen receptor signaling activation without nociception activation, suggesting that androgen signaling is not a strong driver for visceral hypersensitivity in our model. Similarly, estrogen receptor signaling are preferentially activated in female DRG without strong connection to nociception pathway in both male and female DRG (Supplemental Figure 6-7). Overall, our data showed androgen and estrogen receptor signaling activation in their corresponding sexes, including DEGs known to be involved in nociception and pain processing, but without strong connection to nociception according to IPA. We next used STRING analysis to predict potential protein-protein interactions in lumbosacral DRG using Partek identified DEG data ( Fig. 6A ). Proteins responsible for chemotaxis, differentiation, and activation of peripheral immune cells were predicted to be associated with VEGF-induced visceral hypersensitivity ( Fig. 6B – D ). Our data showed that Secreted Frizzled-Related Protein 4 ( Sfrp4 ) and angiopoietin-like protein 4 ( Angtl4 ) were dysregulated in the lumbosacral DRG in both sexes following VEGF instillations. These proteins were found to interact with a key Wnt signaling protein beta-catenin (Ctnnb1) and peroxisome proliferator-activated receptor gamma (PPARγ), suggesting potential common mechanisms in both sexes involved in visceral hypersensitivity ( Fig. 6B ). In male DRGs following VEGF instillations, our data showed decreased expression of PPARγ, cholecystokinin (Cck), proenkephalin (Penk), angiotensinogen (Agt), metabotropic glutamate receptor 3 (Grm3), dopamine receptor D2 (Drd2), neurotensin receptor 2 (Ntsr2), and serpin family E member 1 (Serpine1), as well as increased expression of neurotensin (Nts). These proteins are mostly neurotransmitter receptors and hormones, suggesting dysregulation of neurotransmission in male DRGs during visceral hypersensitivity ( Fig. 6C ). In female DRGs, interferon-gamma-inducible GTPase protein 47 (Ifi47), allograft inflammatory factor 1 (Aif1), CXC motif chemokine receptor 5 (Cxcr5), suppressor of cytokine signaling 1 (Socs1), signal transducer and activator of transcription 1 (Stat1), and interferon regulatory factor (Irf) are predicted to be downregulated following VEGF instillations, suggesting immune resolution when mechanical hypersensitivity is still present ( Fig. 6D ).

Twenty-two young adult C57BL6/J mice (RID:IMSR_JAX:000664; Jackson Laboratory, Bar Harbor, ME) were used in this study. Male (n=12) and female (n=10) mice were both included. All animals were between 8-10 weeks at the beginning of the experiments and generally weighed between 24-30 grams. Intravesical instillations of saline or VEGF 165 were performed in all animals as previously described ( 6 ) to induce visceral hypersensitivity (VEGF male and female groups; n=7 and 5, respectively), while saline was instilled as experimental controls (saline male and female groups, N=5 in each group). All mice were housed in a temperature-controlled environment at the University of Colorado Anschutz Medical Campus (CU-AMC) (Aurora, CO) on a 14-hour light/10-hour dark cycle with ad libitum access to food and water. All instillations, mechanical sensitivity testing, and DRG dissections took place during the light cycle. All animal procedures were submitted to and approved by the IACUC of CU-AMC (No. 01147). Intravesical instillations of VEGF 165 (ProSpec, Rehovot, Israel; #cyt-336) were performed to induce visceral hypersensitivity as previously described ( 6 ). In short, mice were anesthetized with isoflurane (1.5%) and kept on a warming pad during instillations and recovery. Female mice were transurethrally catheterized with a sterile 24-gauge BD Insyte Autoguard ™ Vialon ™ catheter (Becton Dickinson, USA) ( 6 ). Male mice were catheterized with a sterile Polyethylene PE10/10 catheter (Warner Instruments, USA #64-0750) following the previously published protocol ( 14 ). After removing urine, VEGF 165 (6.41 nM in 100 μL saline) or saline (100μL) were instilled into the urinary bladder via the urethral catheter with a syringe attached to one end. To ensure consistent contact of the substances with the bladder lumen and to avoid reflux or leakage, catheters were occluded and left in place for 30 minutes. After each instillation, mice were removed from isoflurane inhaler and left on a heating pad to recover before being returned to their respective cages. All mice received three bladder instillations on three separate days (day 1, 5, and 9, Fig. 1A ). Animals experienced bleeding during catheterization would be excluded from the study. No animals in this study were excluded (no attrition). Manual Von Frey testing (VFT) was used to assess visceral mechanical sensitivity in VEGF groups and saline groups as previously described ( 6 ). Briefly, mice were individually placed in a clear glass chamber (7.5 cm x 7.5 cm x 15 cm) on an elevated wire grid floor (0.5 cm 2 grid size) for habituation and testing. Following habituation, a series of force-generating filaments (0.04, 0.16, 0.4, 1, 1.4, and 2 g; Stoelting, USA) were applied to the mouse lower abdominal area in the vicinity of the urinary bladder. A sharp retraction of the abdomen, jumping up, or licking or scratching the lower abdomen were considered as positive responses. Each force-generating filament was used ten times per mouse to evaluate the mechanical sensitivity to the particular force. The percentage of positive responses was recorded as withdrawal frequency for each force (0-100%). Higher withdrawal frequency indicates higher visceral mechanical sensitivity. To evaluate VEGF-induced changes in visceral mechanical sensitivity, withdrawal frequencies to all filaments were averaged within experimental groups and compared using 2-way or 3-way ANOVA with sex, treatment (saline vs VEGF), or filament forces as factors (indicated in each figure). The most recent version of GraphPad Prism (RRID:SCR_002798; GraphPad Software Inc., California, USA) was used for statistical comparison and figure generation. All data were presented as mean ± SEM. Following the final day of Von Frey testing (Day 16), four animals from each group were euthanized and their lumbosacral L1, L2, L6, S1, and S2 DRG were harvested immediately ( 15 ). Total RNA was extracted from lumbosacral DRG using a RNAeasy miniprep kit (Qiagen, CA, USA), following manufacturer’s instructions. Extracted RNA was resuspended in 15 μl of RNase-free water and stored at −80°C. RNA Quality check was performed on the Agilent 2100 bioanalyzer (Thermoscientific, USA). RNA was quantified using the Qubit fluorometer 3.0 (Thermo Fisher Scientific), and 300 ng of RNA per sample was used for library preparation. The universal mRNA Poly-A tail prep kit was used for library preparation following manufacturer instructions. QC passed libraries were sequenced at a depth of 40 million paired-end reads per sample on a Novaseq platform (Illumina, USA). RNA quality check, cDNA library preparation and sequencing were performed by the CU-AMC Genomics Core facility (Aurora, CO). RNA-seq data analysis was conducted in Partek ® Flow ® (Version 10.0, Partek Inc., St. Louis, USA). Raw FASTQ files uploaded to the software underwent QC and read alignment to a mouse reference genome (mm39) using STAR aligner (STAR - 2.7.8a). HTSeq (0.11.0, RRID:SCR_005514) was used for annotation and quantification. Gene expression analysis was performed using DESeqQ2 (RID:SCR_015687). Datasets were analyzed using sex (male vs female) and treatment (saline vs VEGF) as determinants. DEGs were identified as fold changes >1.5 or <-1.5 with p<0.05. Subsequently, hierarchical gene clustering and gene set enrichment analysis (GSEA) were performed using Partek ® Flow ® . The Kyoto Encyclopedia of Genes and Genomes (KEGG, RRID:SCR_012773) database were used for GSEA. The enrichment score was set at 4 (p≤0.02) instead of 3 (p1.5 or <-1.5 with P<0.1. Gene expression data were then downloaded as an Excel file and imported into IPA. In addition, Basepair (Basepair Inc., New York, USA) was used to generate heatmap in Figure 3 . STRING analysis ( 16 ) was used to examine and predict protein-protein interactions in Figure 7 and was performed using a publicly available database (Version 12.0, RRID:SCR_005223). The graphical representation of the RNA-seq work pipeline were created in Microsoft PowerPoint and the graphical abstract were created using BioRender (RRID:SCR_018361; BioRender, Toronto, Canada).

The current study focused on mapping transcriptome changes in lumbosacral DRG associated with visceral pain symptoms. The lumbosacral DRG (L1-L2, L6, and S1-S2) supply major sensory afferents from the bladder ( 35 ). In this study, we conducted bulk RNA-seq in these DRG in a bladder-origin visceral pain mouse model. Single-cell RNA-seq has been used to study molecular changes in specific cell types in the DRG in animal models of somatic pain ( 37 ). However, the cell dissociation required for single-cell RNA-seq affects cell transcriptomes ( 39 ), raising concerns about how closely transcriptomic data obtained from dissociated and sorted cells reflects the physiological state. High-throughput bulk RNA-Seq provides a snapshot of the whole DRG gene expression profile and is beneficial in identifying the molecular mechanisms associated with genes differentially expressed across conditions ( 42 ). Using Bulk RNA method, we detected significant changes in genes and signaling pathways associated with visceral hypersensitivity in lumbosacral DRG. Sex differences in DRG transcriptome in both baseline (saline-instilled animals) and VEGF-instilled mice were identified. Compared to male DRG, female DRG showed more prevalent immune gene expression at the baseline and more immune pathway activation in animals exhibiting visceral pain symptoms. Active extracellular matrix regulation was also observed in female but not male DRG in animals experiencing visceral hypersensitivity. On the other hand, male DRG displayed more active glial activation and nociceptive responses compared to female DRG in animals exhibiting visceral pain symptoms. Our data strongly suggested the heightened immune response may be the driver of visceral hypersensitivity in female mice, whereas the nociceptor upregulation and glia-neuron interactions are likely underpinning male visceral hypersensitivity. ( 45 )( 48 , 51 )( 53 )Our data suggested that female mice displayed a relatively more active immune status at the baseline than saline-instilled males. Following VEGF instillations, our data also suggested that immune and ECM signaling pathways are more active in female DRG compared to male DRG. These results are consistent with a recent unbiased quantitative proteomic analysis of urine from patients diagnosed with UCPPS ( 54 ). Compared to the urine from healthy controls, those from UCPPS patients showed significant reduction of CD99, COL3A1, and COL1A2, which are involved in leukocyte migration and extracellular matrix remodeling ( 54 ). Heparan sulfate N-deacetylase/N-sulfotransferase 1 and Golgiglycoprotein1 (GLG1), which potentially modulates leukocyte migration were increased in female UCPPS urine samples ( 54 ). We observed a similar trend in the IPA analysis. Neither of these genes were significantly altered in males following VEGF 165 instillations, suggesting sex-specific modulation of selectin signaling in females. In addition, the Multidisciplinary Approach to the Study of Chronic Pelvic Pain (MAPP) Research Network found that female patients diagnosed with interstitial cystitis/bladder pain syndrome (IC/BPS) and chronic overlapping pain conditions (COPCs) display heightened cytokine release in response to toll-like receptor-4 (TLR4) stimulation, suggesting a link between immune priming and central sensitization in IC/BPS in female patients ( 45 ). Similar studies ( 48 , 51 ), along with animal studies demonstrating the role of TLRs in central nervous system in pain sensitization ( 53 ), suggested that TLR responsiveness in circulating immune cells may play a role in pelvic pain centralization in these female patients. ( 55 ) Interestingly, we also observed upregulation of genes associated with autoimmune conditions, increased antibody synthesis, B cell activation, and antigen presentation (Supplementary Figure-3) in VEGF-instilled mice, whereas none of these gene expressions were detected in saline-instilled mice. A recent study using a chronic constriction injury (CCI) pain model quantitatively demonstrated increased IgG accumulation in DRG which co-localized with macrophages and sensory neurons ( 55 ). The authors hypothesized that peripheral nerve injury induces de novo expression of autoantigens, which are detected by antigen-presenting cells, leading to B cell activation. This activation subsequently triggers immune and autoimmune responses, along with Fc receptor (FcR) signaling in the DRG of mice exhibiting increased mechanical hypersensitivity. Following saline instillations, we observed a slight change in visceral sensitivity only in females. Further analysis revealed that the saline instillation-induced visceral hypersensitivity was due to the significant increases in female visceral sensitivity (2-way ANOVA, p<0.0001), whereas male mice did not show any saline-induced changes in withdrawal frequencies (2-way ANOVA, p=0.2861). Transurethral catheterization procedure is technically challenging in males ( 14 ) but not in females. We think that the saline-instillations induced moderate increases in visceral mechanical sensitivity in female animals are not due to the transurethral catheterization procedure, but rather the primed immune functions in female mice compared to male mice ( 17 ) and more robust stress responses to handling ( 18 ). A recent study showed that saline produces a greater corticosterone responses in female mice compared to male mice, suggesting that female mice are more sensitivity to handling ( 18 ). Immune pathways are more active in health female animals than males ( 17 , 55 ), suggesting that stronger immune responses are likely responsible for the female visceral sensitivity change following saline instillations. The role of infiltrating immune cells in the development of pain has already been reported ( 56 , 57 ). Our gene set enrichment analysis indicated that T regulatory response was enhanced in female mice instilled with VEGF 165 . However, no Treg cell response gene enrichment was observed in male mice subjected to VEGF-instillations. Similar results were observed in a recent study of CSF1R −/− mice, showing Treg cells play role in preventing the development of hypersensitivity in female mice but not in male mice ( 58 ). Overall, our results showed that female mice have highly activated immune pathways at the baseline. The differences could be attributed to the cell signaling status and not the number of immune cells present, as no differences were observed in the number of Cd45 + through flow cytometry at the baseline ( 59 ). Our data indicates that several glial signaling pathways were upregulated in VEGF-instilled male DRG, whereas very few glial signaling pathways were changed in VEGF-instilled female DRG. Our findings complement the study by Mogil et al., where inhibiting glial pathways using minocycline, fluorocitrate, and propentofylline reversed hypersensitivity only in male mice but not in female mice ( 60 ). It is worth pointing out that the significantly changed glial signaling pathways are labeled as “microglia” function related. This is because that majority of the glial cells data in the KEGG database (used in our GSEA analysis) originated from the central nervous system (CNS). Therefore, the signaling pathway related to glia functions were labeled with CNS glial type such as microglia. Resident macrophages have been identified as the counterpart of microglia in DRG ( 61 ) and express similar genes and signaling pathway expression with microglia in the CNS. Therefore, the potential sex-biased DRG resident macrophage immune gene expression may underlie the observed sex differences in immune genes and pathways in our study. Male DRG also showed more and stronger nociceptive gene changes following VEGF instillations, suggesting that nociceptive sensitization in the lumbosacral DRG might be the main driver for male visceral hypersensitivity. Using IPA, we observed the prediction of biological process activation such as analgesia, hyperalgesia, and nociception pathways based on changes in gene expression. While female mice showed predicted activation of genes that promote analgesia, male mice displayed a predicted upregulation of hyperalgesia and nociceptive genes ( Fig. 5 ). To name a few, mu-opioid receptor gene ( Oprm1), Bdnf, and Tac1 nociceptive genes were upregulated in the male VEGF 165 -instilled mice; the upregulation of these genes have been reported in hyperalgesia ( 62 ). Specifically, Tac1 is involved in neuroinflammation and hyperalgesia via activating the neurokinin 1 (Nk1) receptor ( 63 – 65 ). In addition, the genes known to encode pain-inhibiting proteins such as Adora1, Adora2a, Mmp24 ( 66 , 67 ), were upregulated in female VEGF 165 -instilled mice and downregulated in male VEGF 165 -instilled mice. Genes involved in analgesic pathways, such as Cnr1 and Gabbr1 ( 68 , 69 ), were upregulated in female VEGF-instilled mice while downregulated in male VEGF 165 -instilled mice. Lastly, Grin1 is upregulated only in male VEGF 165 -instilled mice; Grin1 is commonly found in Functional Abdominal Pain (FAP) and Irritable Bowel Syndrome (IBS) ( 70 ). A downregulation of pain inhibitory genes, like Npyr1 ( 71 ), was observed only in male VEGF 165 -instilled mice. Lastly, STRING analysis showed how changes in the expression of CCK in male mice can affect signal transduction of pain processes through Fos/Jun proteins ( 72 ). Our data showed little overlap between male and female DRG gene expression changes following VEGF-induced visceral hypersensitivity. Specifically, two genes, the secreted frizzled-related protein 4 ( Sfrp4 ) and angiopoietin-like protein 4 ( Angtl4 ) were downregulated in the lumbosacral DRG in both sexes during visceral hypersensitivity. STRING analysis suggested that these proteins are both involved in PPARγ signaling pathway, which has been implied in DRG during nerve regeneration and functional recovery after peripheral nerve injury ( 73 ). PPARs are nuclear receptors that bind to lipid molecules, transduce signals from the metabolic environment, and control gene expression ( 54 ). GSEA analysis showed that genes involved in fatty acid metabolic processes such as Adipoq, Aqp7, Cd36, Lpl, Pck1, Plin1, Plin4, and Pparγ were upregulated in VEGF 165 -instilled female mice, as were PPAR target genes such as Hmgcs2 and Scd1 . Importantly, no differences in PPAR signaling pathways were identified between male and female mice at the baseline. Therefore, PPAR signaling pathway might contribute to the visceral pain symptoms in both male and female mice. ( 74 ) Interestingly, Hmgcs2 and Scd1 are typically expressed in satellite glial cells, connective tissue, and Schwann cells, but not sensory neurons. Based on previously published single-cell RNA-seq data ( 75 ), these genes are not expressed in immune cells but are strongly upregulated in satellite glial cells following peripheral nerve injury ( 73 ). Our data suggest that the VEGF-induced changes in Hmgcs2 and Scd1 are possibly due to changes in lumbosacral satellite glial cells. It is worth mentioning that two genes, sfrp4 and angtl4 were significantly downregulated in lumbosacral DRG of both VEGF-instilled males and females. SFRP4 and ANGTL4 can interact with key proteins involved in the Wnt signaling pathway, such as β-Catenin, and PPARγ ( 76 ). SFRP4, the secreted frizzled related protein 4, is a modulator of Wnt signaling pathway ( 77 ) , which has been implicated in chronic pain condition ( 78 , 79 ). Therefore, SFRP4’s ability to inhibit Wnt can contribute to chronic pain ( 80 ). Recent single-cell RNA-seq data showed that in DRG, mesenchymal stem cells (MSCs) are the primary cell population expressing sfrp4 ( 37 ). Therefore, the downregulation of Sfrp4 in our data is likely due to changes in MSCs. MSCs are present in the DRG of humans, mice, and rats ( 81 ), and are involved in maintaining tissue homeostasis, influencing the surrounding cellular milieu, and potentially contributing to pain-related processes ( 82 ). MSCs in DRG can also interact with immune cells and glial cells, participate in immune modulation and inflammation regulation ( 72 ). In addition, MSCs have been shown to modulate sensory neuron function through paracrine signaling, releasing factors that activate neuron excitability and pain perception and overall sensory processing ( 83 , 84 ). Wnt signaling pathway is highly conserved in MSCs ( 85 , 86 ). The significant downregulation of sfrp4 suggests Wnt signaling activation in MSCs, which may be a common mechanism for immune activation and nociception in lumbosacral DRG. Other translatome studies on DRG also reported SFRP4 expression to be sexually dimorphic in response to a nociceptive signaling molecule, prostaglandin (PGE2) ( 87 ). More recently, a comprehensive single-cell RNA-seq study categorized female patients into two distinct groups - those experiencing adenomyosis pain and those who are pain-free revealed intriguing insights ( 76 ). Notably, this study identified a novel cell population cluster, the SFRP4 + natural killer T cells exclusively presented in patients with adenomyosis pain. An increased expression of SFPR4 is an indication of decreased Wnt signaling leading to suppression of cell differentiation. Additionally, the study identified that another distinct cluster enriched with multipotent stem cell markers presented exclusively within the pain group. Pseudo-timeline analysis implied that this cluster of multipotent progenitor cells has the potential to differentiate into neural progenitor cells, indicating the significant role of SFRP4 in modulating the stemness in the DRG. Besides Sfrp4 , Angiopoietin-like 4 (Angptl4 ) was also downregulated in both male and female VEGF 165 -instilled mice compared to saline-instilled mice. ANGPTL4 interacts with syndecans and forms a complex with the Wnt co-receptor LRP6. The ANGPTL4/LRP6 complex is internalized via clathrin-mediated endocytosis and degraded in lysosomes to attenuate Wnt/β-catenin signaling ( 88 ). Therefore, ANGPTL4 can also function as a Wnt signaling antagonist. Angptl4 is highly expressed in adipose tissues and the liver of humans and mice, and to a lesser extent, in the heart, muscle, kidney, skin, and other tissues ( 89 , 90 ) In mouse DRGs, satellite glial cells, MSCs, and endothelial cells express ANGPTL4 ( 37 ). In addition to its role in lipid metabolism, ANGPTL4 regulates inflammatory signaling ( 91 , 92 ) in a bidirectional fashion ( 93 ); the exact mechanism is not fully understood. A recent study on Angptl4 −/− mice suggested the role of ANGPTL4 in promoting macrophage cell differentiation and neutrophil clearance during wound healing ( 94 ). ANGPTL4 deficiency affects cell-matrix interactions that potentially regulate monocyte-to-macrophage differentiation and prolonged neutrophil persistence after injury ( 94 ). We observed a significant downregulation in the Angptl4 gene in the VEGF 165 -instilled female mice and male mice, suggesting a possible perturbance in neutrophil clearance. Various mechanisms of neutrophil involvement in sensory neuronal sensitization have been described, such as sensitization via neutrophil elastase and chemokines ( 95 , 96 ). Interestingly, VEGF 165 -instilled female mice, but not male mice, showed differences in expression of interferon-gamma-inducible GTPase protein 47 (Ifi47), platelet factor 4 (Pf4), and allograft inflammatory factor 1 (Aif1); all of which are responsible for chemotaxis, differentiation, and activation of peripheral immune cells. Interestingly, these genes are all downregulated in VEGF 165 -instilled female mice, which explains the comparatively low expression of immune-responsive genes shown in Figure 3C in the VEGF 165 -instilled female mice compared to the saline-instilled female mice. Neutrophils are primary mediators for the innate host defense ( 97 ). Neutrophil infiltration into DRG after peripheral injury has previously been observed in murine models of chronic pain ( 98 ). Contrary to neutrophils from healthy controls, neutrophils from patients with pain conferred mechanical hypersensitivity in recipient naïve mice ( 99 ). Adoptive transfer of immunoglobulin, serum, lymphocytes, or monocytes from the chronic pain mice did not affect the pain behavior of the naïve mice. Furthermore, systemic depletion of neutrophils before and after insult prevents the development of persistent mechanical hypersensitivity in mice ( 99 ). In our data, NGF is significant downregulated in female DRG, but not in male DRG ( Fig. 5 ), suggesting potential pain resolution in female DRG ( 36 ). Our data showed sex-unique NGF signaling pathway-related gene changes in lumbosacral DRG following VEGF instillations, with no overlap between female and male DEGs (Supplementary data 4). IPA analysis also showed that female DRG displayed broader activation of NGF signaling, including transcriptional regulation. On the other hand, males display weaker NGF signaling with greater activation in genes related to DNA repair, growth factor, and stress-adaptive pathways. IPA identified inhibitory effect on the immediate-early transcription factor Egr1 in lumbosacral DRG in both sexes, suggesting attenuated neurite sprouting in both sexes (Supplementary Figure 4). Taken together, these results indicate that while Egr1 -driven structural remodeling is dampened in both sexes, female DRG maintain a more robust NGF-driven excitatory and structural signaling state, whereas male DRG favor stress-responsive and survival-associated signaling. NGF signaling is critical in neurite growth/sprouting, which has been implied in chronic pain conditions ( 100 – 102 ). In our previous study, we have also demonstrated that chemical sympathectomy in the peripheral nervous system abolished Complete Freund’s Adjuvant (CFA)-induced hind paw mechanical hyperalgesia ( 103 ). IPA on neurite growth signaling pathways revealed that that neurite growth/sprouting pathways were inhibited in female DRG but activated in male DRG following VEGF instillations ( Fig. 6 ). Female DRG showed broad upregulation of genes involved in synaptic organization, cytoskeletal remodeling, and MAPK/AKT signaling, indicating a transcription-dependent strategy for structural reorganization ( 29 ). In contrast, males exhibited upregulation of lipid signaling and post-translational regulators, including phosphoinositide kinases, PLC–PKC pathways, and Rho GTPases ( 30 ). These data suggested that in female DRG, neurite growth/sprouting requires transcriptional support ( 31 ), whereas in male DRG, neurite growth/sprouting relies on lipid-mediated and post-translational mechanisms, consistent with previous reports ( 32 , 33 ).

Our research provides the first characterization of signaling pathway alterations in lumbosacral DRG that occur during the development of UCPPS symptoms in a VEGF 165 -induced mouse model. Our data not only revealed significant sex differences in gene expression between male and female mice at baseline, but also identified sex-specific signaling mechanisms contributing to visceral pain symptoms. Our data also highlighted the important roles of ECM remodeling and organization in visceral hypersensitivity, especially in females.

Visceral pain originates from internal organs and is typically dull, diffused and poorly localized ( 1 ). Abdominal and pelvic pain are the most common types of visceral pain and are often associated with diagnoses of lower urinary tract, gastrointestinal (GI) tract, or reproductive organ disorders ( 2 ). Chronic visceral pain can persist without inflammation or apparent tissue pathology ( 3 ), leaving chronic analgesics the main treatment option ( 4 , 5 ). Existing data suggest neuronal plasticity involving the sensitization of sensory afferent in pelvic organs, known as visceral hypersensitivity as a widely accepted mechanism underpinning chronic abdominal and pelvic pain ( 4 ). However, how visceral hypersensitivity is developed or maintained is still poorly understood, hindering the development of alternative therapies ( 4 , 5 ). In this study, we characterized the gene and signaling pathway changes in lumbosacral DRG associated with bladder-originated visceral hypersensitivity using a mouse model of Urological Chronic Pelvic Pain Syndrome (UCPPS) ( 6 , 7 ). UCPPS encompasses interstitial cystitis (IC)/bladder pain syndrome (BPS) in females and chronic prostatitis (CP)/chronic pelvic pain syndrome (CPPS) in males ( 8 ). Elevated vascular endothelial growth factor (VEGF) level has been detected in the urine of UCPPS patients ( 9 ); Increased VEGF levels were also found in bladder biopsy samples from IC patients compared with controls and correlated with pain severity ( 10 , 11 ). Intravesical instillations of VEGF 165 , a mouse recombinant of VEGF A reliably induce nociceptor sensitization and visceral mechanical hypersensitivity in mice ( 6 , 7 ), serving as an animal model of bladder-originated visceral pain without apparent tissue pathology ( 12 ). Bladder afferents transmit bladder mechanical sensation and nociceptive signals to the lumbosacral spinal cord via five sets of lumbar and sacral DRG (L1-L2, and L6-S2) ( 13 ). Therefore, we performed transcriptome profiling to uncover changes in gene networks in lumbosacral DRG in mice instilled with VEGF 165 (VEGF animals) comparing to their saline-instilled littermates (control animals). Our data showed significant sex differences in differentially expression genes (DEGs) in L1-L2 and L6-S2 DRG in VEGF animals when compared to control animals, as well as sex differences in the molecular profile of lumbosacral DRG in saline-instilled control animals. We also identified distinctive signaling pathways associated visceral hypersensitivity in female and male animals. Overall, our data showed sex-specific immune and nociceptive gene expression at the baseline in lumbosacral DRG, which we called “molecular sexual dimorphism,” as well as sex-specific gene and signaling pathway activation associated with visceral hypersensitivity. This foundational knowledge provides a valuable source for targeting specific signaling pathways to prevent or reverse visceral hypersensitivity in both sexes.