Exosome-Associated Lysophosphatidic Acid Signaling Contributes to Cancer Pain (Workshop Paper 1_jop-164-2684 PDF)

Research Paper PAIN 164 (2023) 2684–2695 Exosome-associated lysophosphatidic acid signaling contributes to cancer pain Iryna A. Khasabovaa, Sergey G. Khasabova, Malcolm Johnsa, Joe Juliettea...

Research Paper PAIN 164 (2023) 2684–2695 Exosome-associated lysophosphatidic acid signaling contributes to cancer pain Iryna A. Khasabovaa, Sergey G. Khasabova, Malcolm Johnsa, Joe Juliettea, Aunika Zhenga, Hannah Morgana, Alyssa Flippena, Kaje Allena, Mikhail Y. Golovkob, Svetlana A. Golovkob, Wei Zhanga,c, James Martid, David Caina, Virginia S. Seyboldd, Donald A. Simonea,* Abstract Pain associated with bone cancer remains poorly managed, and chemotherapeutic drugs used to treat cancer usually increase pain. The discovery of dual-acting drugs that reduce cancer and produce analgesia is an optimal approach. The mechanisms underlying bone cancer pain involve interactions between cancer cells and nociceptive neurons. We demonstrated that fibrosarcoma cells express high levels of autotaxin (ATX), the enzyme synthetizing lysophosphatidic acid (LPA). Lysophosphatidic acid increased proliferation of fibrosarcoma cells in vitro. Lysophosphatidic acid is also a pain-signaling molecule, which activates LPA receptors (LPARs) located on nociceptive neurons and satellite cells in dorsal root ganglia. We therefore investigated the contribution of the ATX–LPA–LPAR signaling to pain in a mouse model of bone cancer pain in which fibrosarcoma cells are implanted into and around the calcaneus bone, resulting in tumor growth and hypersensitivity. LPA was elevated in serum of tumor- bearing mice, and blockade of ATX or LPAR reduced tumor-evoked hypersensitivity. Because cancer cell–secreted exosomes contribute to hypersensitivity and ATX is bound to exosomes, we determined the role of exosome-associated ATX–LPA–LPAR signaling in hypersensitivity produced by cancer exosomes. Intraplantar injection of cancer exosomes into naive mice produced hypersensitivity by sensitizing C-fiber nociceptors. Inhibition of ATX or blockade of LPAR attenuated cancer exosome-evoked hypersensitivity in an ATX–LPA–LPAR-dependent manner. Parallel in vitro studies revealed the involvement of ATX–LPA–LPAR signaling in direct sensitization of dorsal root ganglion neurons by cancer exosomes. Thus, our study identified a cancer exosome- mediated pathway, which may represent a therapeutic target for treating tumor growth and pain in patients with bone cancer. Keywords: Cancer, Exosomes, Pain, Autotaxin, LPA, Nociceptor sensitization 1. Introduction ganglion (DRG) neurons independent of the immune system and Severe pain occurs in more than 60% of patients with primary or physical contact with neurons. metastatic bone cancer.39 The mechanisms underlying bone Cancer cells express high levels of autotaxin (ATX),43 a cancer pain include unique tumorigenic components.48,50 member of the nucleotide pyrophosphatase/phosphodiesterase Experimental models of bone cancer pain have shown that family of ectoenzymes. Expression of the ATX gene, ENPP2, is osteolytic tumor growth results in the sensitization of C-fiber regulated by cytokines, growth factors, and hormones and is nociceptors4,5 and dorsal horn neurons.24,57 Moreover, our in among the top 40 genes upregulated in metastatic cancer.12,45 vitro model of cancer cell–neuron interactions24 showed that Autotaxin has lysophospholipase D activity that cleaves choline factors released by cancer cells directly sensitize dorsal root from lysophosphatidylcholine (LPC) to form lysophosphatidic acid (LPA).42 The high concentration of LPA in plasma and ascites fluid of patients with cancer19 is evidence that cancer cells produce and release LPA. Importantly, LPA is involved in the Sponsorships or competing interests that may be relevant to content are disclosed progression of cancer by stimulating cancer cell proliferation, at the end of this article. angiogenesis, and metastasis,1,47 and an ATX inhibitor is being a Department of Diagnostic and Biological Sciences, University of Minnesota, considered as a possible therapeutic in treating cancer (Clin- Minneapolis, MN, United States, b Department of Biomedical Sciences, University icalTrials.gov Identifier: NCT05586516). of North Dakota, Grand Forks, ND, United States, c MNC, College of Science and Lysophosphatidic acid exerts its effects by binding to LPA1-6 Engineering, University of Minnesota, Minneapolis, MN, United States, d Depart- ment of Neuroscience, University of Minnesota, Minneapolis, MN, United States receptors (LPARs), among which LPARs1,3,5 are widely *Corresponding author. Address. University of Minnesota, Department of Di- expressed by neurons and satellite cells of DRG.20,32,34,46 agnostic and Biological Sciences, 515 Delaware St SE, Minneapolis, MN 55455, LPARs1,3,5 activate downstream pathways including calcium United States. Tel.: 612-625-6464; fax: 612-626-2651. E-mail address: signaling,59 thereby contributing to pain. Intrathecal20,35 or [email protected] (D. A. Simone). intraplantar41 injection of LPA in rodents produces nocifensive Copyright © 2023 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf behaviors and hypersensitivity, and these behaviors are blocked of the International Association for the Study of Pain. This is an open access article with co-administration of an LPAR1 antagonist.52 In humans, distributed under the terms of the Creative Commons Attribution-Non Commercial- No Derivatives License 4.0 (CCBY-NC-ND), where it is permissible to download and intracutaneous injection of LPA produces burning pain and share the work provided it is properly cited. The work cannot be changed in any way hypersensitivity, and sensitizes C-fiber nociceptors.11 Further- or used commercially without permission from the journal. more, LPA is associated with the intensity of neuropathic pain in http://dx.doi.org/10.1097/j.pain.0000000000002967 humans.30 2684 · I.A. Khasabova et al. 164 (2023) 2684–2695 PAIN® December 2023 · Volume 164 · Number 12 www.painjournalonline.com 2685 Exosomes play a central role in ATX–LPA–LPAR signaling. A were injected in 10 mL of phosphate-buffered saline, pH 7.3, significant proportion of secreted ATX in animals and humans is unilaterally into and around the calcaneus bone under isoflurane bound to exosomes, which are crucial for the delivery of LPA to (2%) anesthesia as described previously.23–27 distally located LPA receptors.21,42 Exosomes are spherical, To study the effect of LPA on cancer cells in vitro, cells were extracellular nanovesicles ranging in size from 40 to 150 nm with a seeded into 6-well plates with a density of 3 3 106 cells/well. After high content of CD63, CD9, and CD81 tetraspanins.28 Exosomes 24 hours in culture, the cells were treated with 18:1-LPA or mediate intercellular communication either through fusion with vehicle (0.04% ethanol), cultured for another 24 hours then the target membrane and release of their contents into the trypsinized and counted. cytoplasm of recipient cells or through ligand–receptor interaction on the plasma membrane of the target cell. Exosomes secreted by cancer cells carry cell-specific proteins, mRNAs, microRNAs, 2.4. Blood collection and analysis and DNA fragments that are known to promote angiogenesis, Blood (0.5 mL) was collected and kept for 30 minutes for clotting. tumor cell proliferation, invasion, and metastasis47 as well as Serum was isolated by centrifugation for 10 minutes at 32000g at axonogenesis.36 The contribution of cancer cell exosome genes 4 ˚C. The supernatant was used for measurement of LPA. to pain in oral cancer was recently demonstrated.2,10 Samples were stored at 280˚C until processing. Although LPA is an established mediator of pain, the contributions of LPA and exosome-mediated ATX–LPA–LPAR signaling in cancer pain are unknown. Our results indicate that 2.5. Measurement of endogenous lysophosphatidic acid ATX–LPA–LPAR signaling associated with exosomes contributes The total serum level of LPA was determined using the general to bone cancer pain. Thus, interfering with ATX–LPA–LPAR LPA ELISA kit according to the manufacturer’s protocol signaling may be an effective approach to manage cancer pain (MyBioSource, San Diego, CA). The contribution of various LPA without the limitations of opioids. species was identified by UPLC-MS/MS. Lipids were extracted from serum samples (250 mL) using a 2-phase method of liquid 2. Methods extraction with acetone3,52 and d5-18:2-LPA as an internal standard. The chloroform fraction containing LPA was dried 2.1. Animals under nitrogen and dissolved in 30 mL of 66% methanol and 10 Adult, male and female C3H/He mice (National Cancer Institute, mL of which was analyzed by UPLC-MS/MS. Lysophosphatidic Bethesda, MD; 25-30g) were used. Mice were maintained on a acid was resolved on the ACQUITY UPLC HSS T3 column 12-hour light–dark schedule and had food and water ad libitum. (1.8 mm, pore diameter 100 Å, 2.1 3 150 mm) using the Acquity All procedures were approved by the University of Minnesota UPLC Waters system (Waters, Milford, Massachusetts) at a Institutional Animal Care and Use Committee. At the end of column temperature of 55˚C. Mobile phase A was 2 mm ammonia experiments, mice were euthanized by inhalation of CO2. formate in water and phase B was 2 mm ammonia formate in Experiments were conducted in male and female mice unless methanol. The flow rate was 0.3 mL/minute at an initial indicated otherwise. concentration of 39% B. At 0.5 minute, %B was increased to 80% within 5.5 minutes and to 95% after 9 minutes within 1 minute. At 11 minutes, the column was washed with 99% of 2.2. Drugs solvent B and equilibrated with 39% B for 3 minutes between For mass spectrometry and treatment of fibrosarcoma cells, injections. The Waters Xevo TQ-S triple quadrupole mass linoleoyl lysophosphatidic acid (18:2-LPA) and its deuterated spectrometer controlled by MassLynx V4.1 software was used standard (18:2-LPA-d5), oleoyl-LPA (18:1-LPA), arachidonoyl- to quantify LPA. The following mass transitions were used at a LPA (20:4-LPA), palmitoyl-LPA (16:0-LPA), and collision energy of 21 V: 16:0-LPA 409.2/152.9; 18:0-LPA 437.2/ docosahexaenoyl-LPA (22:6-LPA) were purchased from Echelon 152.9; 18:1-LPA 435.2/152.9; 18:2-LPA 433.2/152.9; 18:2- Biosciences (Salt Lake City, UT). H2L5765834, an antagonist of LPA-d5 438.3/157.9; 20:4-LPA 457.2/152.9; and 22:6-LPA the LPARs1,3,5 was purchased from Tocris (Minneapolis). A stock 481.2/152.9. LPA was quantified against d5-18:2-LPA internal solution (10 mg/100 mL) was prepared in DMSO and diluted in standard using standard curves and expressed in micrograms saline or HEPES buffer to achieve the final dose or concentration. per milliliter of serum. BI-2545, an ATX inhibitor, and BI-3017, an inactive analog (https://opnme.com/molecules/atx-bi-2545; https://opnme. 2.6. Behavioral measures of hypersensitivity com/sites/default/files/opnMe_M2O_profile_ATX_inhibitor_0. pdf) were obtained as a gift from Boehringer Ingelheim In- Mechanical hypersensitivity was defined as an increase in the ternational GmbH and used according to their recommended frequency of paw withdrawal in response to a von Frey mono- guidelines. filament with a bending force of 3.9 mN as we described previously.22,23,25–27 The monofilament was applied 10 times, each for 2 seconds, to random locations on the plantar surface of 2.3. Maintenance and implantation of fibrosarcoma and each hind paw. The number of withdrawal responses evoked by fibroblast cells the monofilament was expressed as a percentage of total stimuli. NCTC clone 2472 fibrosarcoma cells (hereinafter referred to as Before experiments, mice were prescreened for mechanical cancer cells, ATCC, Manassas, VA) and fibroblast clone NCTC hypersensitivity, and animals with withdrawal responses $50% 929 (ATCC, Manassas, VA) were maintained in NCTC 135 were removed from further experimentation.24 Among the mice (Sigma-Aldrich) medium supplemented with 10% horse serum. with implanted cancer cells, only mice with mechanical hyper- Both cell lines are syngeneic with the C3H mice, so administration sensitivity ($50% frequency of withdrawal of the ipsilateral hind of these cells does not cause an immune response and allows paw) were used, and this group was defined as “tumor-bearing” them to grow without rejection. Cancer cells and fibroblasts (2 3 105) (TB) mice. 2686 · I.A. Khasabova et al. 164 (2023) 2684–2695 PAIN® Heat hypersensitivity was defined as a decrease in paw 2.9. Nanoparticle tracking analysis withdrawal latency in response to a radiant heat stimulus applied Quantification of the hydrodynamic diameter distribution and to the plantar surface of each hind paw.16 The intensity of the heat concentration of PLSCExo was performed using the NanoSight source was adjusted so that mice withdrew their hind paws at ;9 LM-10 (Malvern Instruments, Malvern United Kingdom). Purified seconds during baseline testing. Mean withdrawal latencies were exosome samples were diluted 10 times in PBS. Captured obtained from 3 trials separated by at least 2 minutes. A cutoff images were analyzed through the NanoSight software to track time of 16 seconds was chosen to avoid tissue damage. Baseline the motion of each particle over 45 seconds to calculate a measures of paw withdrawal frequency and paw withdrawal diffusion coefficient D of each particle. Using the Stokes–Einstein latency were performed during 3 days before and every second equation, these diffusion coefficients were used to calculate day during 10 days after implantation of fibroblasts and cancer particle size distribution for each sample. Average values of cells. The experimenter conducting the behavioral studies was particle size measures (mean, mode, and median) were blinded to the treatment. calculated from 5 measurements of each sample. 2.7. Real-time quantitative PCR 2.10. Western blot analysis Autotaxin, LPAR1, LPAR3, and LPAR5 mRNAs were detected in Samples of L3-L6 DRG, plantar paw soft tissues or isolated L3-L6 DRG and paw skin from naive and TB mice using qPCR. exosomes were sonicated in single-detergent lysis buffer, and 30 Samples from tumor-bearing mice were ipsilateral to tumors. to 50 mg of protein was subjected to electrophoresis on a Mini- After RNA isolation using Trizol Reagent (Invitrogen), samples PROTEAN TGX 4-20% Stain-Free Gels (Bio-Rad). Samples from were incubated with TURBO DNase (Ambion) and RNA was TB mice were ipsilateral to tumors. The protein loading was purified with the Zymo Research RNA Clean and Concentrator kit confirmed by Revert 700 Total Protein Stain (LI-COR). Immuno- (Thomas Scientific). Then RNA in the samples was quantified by blots were probed with rabbit anti-CD63, anti-CD9, anti-CD81, spectrophotometric analysis with a NanoDrop 2000 (Thermo anti-ENPP2 (ATX, Abcam or Invitrogen), anti-LPAR1 (Bioss), anti- Scientific). cDNA was synthesized using 100 ng of mRNA sample LPAR3 (Invitrogen), and anti-LPAR5 (Millipore) antibodies and with the ProtoScript II First Strand cDNA Synthesis Kit (BioLabs). visualized with IRDye 800CW goat anti-rabbit secondary The real-time quantitative PCR (RT-qPCR) assay was performed antibody (LI-COR). Membranes were imaged using an LI-COR using the Maxima SYBR Green/ROX RT-qPCR Master Mix Odyssey. Total protein and immunoreactivity of bands of interest (Thermo Scientific). A QuantStudio 3 (Applied Biosystems) was were analyzed using ImageJ software. used for quantification of the PCR reaction. All kits were used according to the manufacturers’ protocol. Detected transcript levels were normalized to the b-Actin housekeeping gene and 2.11. Transmission electron microscopy ddCt was calculated.33 Primer pair sequences were as follows: Transmission electron microscopy (TEM) is the most common ATX (IDT), forward primer, 59-GGC TGC ACC TGT GAT AAG A- type of electron microscopy for imaging exosomes.7 A 3 mL 39, and reverse primer, 59-GCA CTG CAG GTC CAT AC-39; aliquot of the purified and concentrated exosome sample was LPAR1 (IDT), forward primer, 59-CTA TGT TCG CCA GAG GAC applied to a glow-discharged 400 mesh ultrathin carbon grid TAT G-39, and reverse primer, 59-GCA ATA ACA AGA CCA ATC (Electron Microscopy Sciences) and stained using 2% uranyl CCG-39; LPAR3 (IDT), forward primer, 59-TGG AGG TGA AAG formate. The grid was then blotted dry with filter paper. The TEM TCA TGC TC-39, and reverse primer, 59-CAG TTC AGG CCG grid was imaged in an FEI Tecnai 300 kV field emission gun TEM TCC AG-39; LPAR5 (IDT), forward primer, 59-ACC AAC GTC TTA (Thermo Fisher Scientific) on a Gatan Summit K2 direct electron TCT CCA CAC-39, and reverse primer, 59-GTA TCT CGA TAG detector camera (Gatan, Inc). TCA GGG CAC-39; and b-Actin (IDT), forward primer, 59-GAT CAG CAA GCA GGA GTA CGA-39, and reverse primer, 59-AAA 2.12. Electrophysiology ACG CAG CTC AGT AAC AGT C-39. Electrophysiological recordings were made from single nocicep- tive C fibers from the tibial nerve in control mice anesthetized with 2.8. Exosome isolation and characterization isoflurane as described previously.4,5,53,54,56 Mice were anes- Murine fibrosarcoma (cancer) cell clone NCTC 2472 and thetized with 1% to 3% isoflurane through a nosecone. Once fibroblast clone NCTC 929 (ATCC, Manassas, VA) were plated adequately anesthetized, the hair around the left hind leg was in P75 flasks at the same density (11 3 103 cells/cm2) and removed and an incision was made in the skin over the maintained in NCTC 135 (Sigma) and EMEM (Thermo Fisher) gastrocnemius muscle, which was dissected and removed to medium, respectively. Exosomes were isolated from culture access the tibial nerve. The skin was then sutured to a stainless media conditioned by either mouse fibroblast (control) or steel ring (1.3-cm inner diameter) to form a pool that was filled fibrosarcoma (cancer) cell lines (from 5 to 15 passages in with mineral oil. Dental impression material (COE-FLEX, GC, culture) as described.38 In brief, cells that achieved 80% of America) was applied to the skin and around the ring to prevent oil confluence were cultured for another 48 hours in appropriate from leaking out of the pool during the experiment and to stabilize fresh medium. The medium was collected, and exosomes were the hind paw. Action potentials from individual fibers were isolated by sequential centrifugation (10 minutes 3200g; 30 amplified, audiomonitored, and visualized on a oscilloscope and minutes 310,000g and twice 60 minutes 3100,000g). Each a PC. Extracellular recordings were obtained only from single pellet was resuspended in phosphate-buffered saline (PBS), fibers easily discriminated according to amplitude and shape of and cells were trypsinized. The individual masses of exosomes their action potential using Spike 2.0 software (CED, Cambridge, and cells were determined by total protein (BCA Protein Assay United Kingdom). The receptive field (RF) of each fiber was kit, Sigma). Final exosome mass was expressed as a ratio of identified using a calibrated suprathreshold von Frey (VF) mono- total protein of exosomes to total protein of cells recovered from filament (149 mN force). Nociceptors were identified by their the same flask. responses to different types of mechanical stimuli including light December 2023 · Volume 164 · Number 12 www.painjournalonline.com 2687 touch (brush with a cotton-tipped applicator) and mild pinch. frequency of response was defined as a percentage ([number of Fibers responding to pinch but not light touch were studied. neurons that responded/number of neurons tested] 3 100%). Conduction velocity (CV) was determined by electrically stimu- lating the RF and measuring the conduction latency and the 2.15. Statistical analyses distance between the RF and the recording electrode. Fibers with a CV ,1.3 m/second were considered C fibers. Data are reported as the mean 6 SEM when normally distributed. Spontaneous activity (1 minutes) and responses evoked by For these data, multiple groups were analyzed by one-way and controlled noxious mechanical stimuli applied to the RF were two-way analysis of variances (ANOVAs) (with repeated mea- obtained before and at 30 and 60 minutes after injection of sures [RM] when applicable), followed by Bonferroni t tests for exosomes from either cancer cells or fibroblasts (300 mg/10 mL), multiple comparisons. Two groups were tested with the two- or vehicle (10 mL of PBS) into the RF. Calibrated VF monofila- tailed unpaired Student t test. Data that were not normally ments were used to deliver mechanical stimuli. Mechanical distributed are reported as the median (25th to 75th percentile response threshold (mN) was defined as the minimum force that range), and effects of treatments were analyzed using the evoked a response in at least 50% of the trials. Responses Mann–Whitney rank sum test or the Kruskal–Wallis one-way evoked by the 149 mN VF monofilament applied to the RF for 5 ANOVA to determine differences between groups. Data for the seconds were expressed as the mean of 3 trials, each separated frequency of response were analyzed using Chi-square tests. by 60 seconds. To further characterize nociceptors functionally, a Statistical analyses were performed using SigmaPlot 14.5 Peltier device (25 mm2 probe) was used to deliver computer- software (Inpixon, CA). Values of P , 0.05 were considered controlled heat (50˚C for 5 seconds) and cold (0˚C for 10 seconds) statistically significant. to the RF from a base temperature of 32˚C and a ramp rate of 18˚C/second with ;2 minutes between stimuli. 3. Results 3.1. LPA signaling contributed to hypersensitivity produced 2.13. Isolation of adult murine dorsal root ganglion neurons by bone cancer Primary cultures of dissociated DRG were prepared from DRG Consistent with our previous reports,22,25 cutaneous mechanical dissected from all levels of the spinal cord of adult C3H mice as and heat hypersensitivity developed within 10 days after we described previously.24 After enzymatic and mechanical implantation of mouse NCTC clone 2472 fibrosarcoma (cancer) dissociation, the final cell suspension was plated on laminin- cells unilaterally into and around the calcaneus bone of male and coated glass coverslips (Fisher Scientific) and maintained in 2.5 female mice (F1,14 5 326.35, 2-way RM ANOVA, P , 0.001, n 5 mL of Ham F12/DMEM medium supplemented with l-glutamine 8 mice/group, followed by Bonferroni t test) and did not differ (2 mm), glucose (40 mm), penicillin (100 U/mL), streptomycin (100 between sexes (P 5 0.15). Implantation of mouse clone NCTC mg/mL), and DNase I (0.15 mg/mL, Sigma). Cells were 929 fibroblasts did not produce the appearance of a visible tumor maintained in a humidified atmosphere of 5% CO2 at 37˚C for and hypersensitivity in naive mice of either sex (F1,305 1.74, 20 to 24 hours before use. 2-way RM ANOVA, P 5 0.20, n 5 6-8 mice/group, followed by Bonferroni t test). Tumor-bearing (TB) mice exhibited an elevated level of LPA in 2.14. Measure of free intracellular calcium ([Ca21]i) serum (Fig. 1A). This change did not occur in mice injected with Measures of [Ca21]i were made in the soma of single neurons nonmalignant fibroblasts. Of the 5 species of LPA, only 3, 18:1-, using a dual-emission microfluorimeter (Photoscan, Photon 18:2-, and 20:4-LPA, were elevated in TB mice (Fig. 1B). No Technology International) to monitor the Ca21-sensitive fluores- changes occurred in the levels of 16:0- and 22:6-LPA. Because cent indicator indo-1. Only neurons with a soma area ,500 mm2 few mice (,2%) did not develop hypersensitivity after implantation were selected and are referred to as small in this report. Small of cancer cells, LPA levels were not determined in this group. The neurons are most likely to be nociceptive.18,44 Neurons were promitogenic effect of 18:1-LPA (10 mM, 24 hours) on fibrosar- loaded with indo-1 by incubating the cultures with 3 mM indo-1 coma (cancer) cells in vitro was confirmed by counting cells before acetoxymethyl ester (Molecular Probes, Eugene, OR) for 60 plating and after treatment (Fig. 1C). The concentration of LPA and minutes at 37˚C in HEPES-buffered Hank balanced salt solution duration of treatment were based on a previous study.14 containing 2% bovine serum albumin. The HEPES-Hank’s Because LPA was elevated in TB mice and targets LPARs1,3,5 solution contained (in mM) HEPES (20), NaCl (137), KCl (5.4), in DRG, we tested whether an LPAR antagonist, H2L5765834, NaHCO3 (3.0), CaCl2 (1.3), MgSO4 (0.4), MgCl2 (0.5), KH2PO4 would reduce hypersensitivity in TB mice. Intraplantar (i.pl.) (0.4), and glucose (5.6), pH 7.35, and was adjusted to 335 to 340 injection of H2L5765834 (5 mg/kg, i.pl.) into the tumor-bearing mOsm with sucrose. After removal of the loading solution, cells paw attenuated mechanical hypersensitivity and blocked heat were maintained in HEPES buffer containing (in mM) HEPES (25), hypersensitivity as compared to injection of vehicle (Figs. 1D and NaCl (135), CaCl2 (2.5), KCl (3.5), MgCl2 (1), and glucose (3.3), pH E). A lower dose (1 mg/kg, i.pl.) had no effect (P 5 0.058, one-way 7.4 and 335 to 340 mOsm. For microfluorimetry, the coverglass RM ANOVA, N 5 6/group, data not shown). Furthermore, was mounted in a superfusion chamber attached to an inverted administration of the antagonist into the contralateral hind paw fluorescence microscope. Cells were superfused with HEPES had no effect on mechanical hypersensitivity (P 5 0.85, one-way buffer at a rate of 1.7 mL/minute and a temperature of 25 6 1˚C. A RM ANOVA, N 5 6/group), indicating the absence of a systemic stable baseline for each neuron was obtained during superfusion effect and the occurrence of receptors in the area of the tumor. with HEPES buffer before testing responses to chemical stimuli: Given evidence that LPARs1,3,5 are expressed by neurons and KCl (25 mM, 10 seconds) or capsaicin (200 nM, 30 seconds). The satellite cells of DRG,20,32,34,46 we determined the effect of tumor threshold for defining the occurrence of a Ca21 transient was an growth on the expression of LPAR1, LPAR3, and LPAR5 in L3-L6 increase in [Ca21]i.50% of the basal [Ca21]i. After testing, DRG using RT-qPCR and western blots. There were no changes neuronal viability was confirmed by a response to 50 mM KCl. in mRNA for any receptor subtype (P 5 0.66, P 5 0.15 and P 5 Only neurons responding to 50 mM KCl were included. The 0.50, Mann–Whitney rank sum test, N 5 7-12 mice/group) or 2688 · I.A. Khasabova et al. 164 (2023) 2684–2695 PAIN® Figure 1. Lysophosphatidic acid signaling contributed to hypersensitivity produced by bone cancer. (A) Level of LPA was higher in serum of tumor-bearing (TB) mice compared with naive (N) mice or mice injected with fibroblasts (Fb). Results are expressed as a percentage of the average for naive mice. *Different from naive and fibroblast-injected mice at P , 0.001, one-way ANOVA followed by the Kruskal–Wallis test on ranks. The numbers in the bars indicate the number of mice. (B) Elevated level of LPA in TB mice was due primarily to changes in 18:1-, 18:2-, and 20:4 species of LPA. The results are expressed as a percentage of the average for the same species of LPA in naive mice. *Different from values for the naive group within each species of LPA at P 5 0.03, P 5 0.004, and P 5 0.04, respectively (Student t-test). The error bar in the naive group represents the maximum variance exhibited across the individual species of LPA. (C) 18:1-LPA (10 mM, 24 hours) stimulated proliferation of fibrosarcoma (cancer) cells compared with cells treated with vehicle (0.04% ethanol; 99.96% saline). Results are expressed as a percentage of the mean for vehicle and reported as the median and the 25th to 75th percentile range; the numbers in the bars indicate the number of culture preparations that was treated. *Different from vehicle at P # 0.001, Mann–Whitney rank sum test. (D) H2L5765834, the antagonist of LPARs1,3,5 (5 mg/kg, i.pl.) reduced the withdrawal response to the mechanical stimulus in TB mice but had no effect in naive mice (F[2,133] 5 104, 2-way RM ANOVA, P , 0.001, n55 to 9 mice/group, followed by Bonferroni t test). The vehicle (15% DMSO, 5% Tween-80, and 80% saline) had no effect. *Different from baseline (BL) in the same treatment group at P 5 0.001; #different from vehicle-treated mice at P 5 0.032; tdifferent from naive mice treated with H2L5765834 at P 5 0.032. (E) H2L5765834 blocked heat hypersensitivity in TB mice (F[2,65] 5 10.33, 2-way RM ANOVA, P , 0.001, n 5 5 to 6 mice/group, followed by Bonferroni t test). *Different from BL at P , 0.001; #different from vehicle-treated mice at P 5 0.005; tdifferent from naive mice treated with H2L5765834 at P 5 0.001. (F) Systemic administration of BI-2545, an ATX inhibitor (10 mg/kg, i.p.), blocked mechanical hypersensitivity in TB mice (F[4,217] 5 49.48, 2-way RM ANOVA, P , 0.001, n 5 7 to 9 mice/group, followed by Bonferroni t test). Neither vehicle nor BI-3017 (negative control) had an effect. BI-2545 produced no effect in naive mice. *Different from BL at P , 0.001; #different from vehicle (10% DMSO, 40% PEG400, 5% Tween-80, and 45% saline) and BI-3017-treated mice at P 5 0.002; tdifferent from naive mice treated with BI-2545 at P , 0.001. (G) BI-2545 also blocked heat hypersensitivity in TB mice (F[4,116] 5 16.66, 2-way RM ANOVA, P , 0.001, n 5 6 to 9 mice/group, followed by Bonferroni t test). BI-2545 had no effect in naive mice. BI-3017 and vehicle had no effects in TB mice. *Different from BL at P , 0.001; #different from vehicle and BI-3017-treated TB mice at P 5 0.05; tdifferent from naive mice treated with BI-2545 at P 5 0.007. (H) Representative examples of ATX protein detected by western blot in soft tissues of plantar paw from naive (N) and TB mice. (I) In densitometric analyses of images, the density of the ATX band was normalized to the total loaded protein. The results are expressed as a percentage of values from naive mice. *Different from naive mice at P 5 0.007 (Student t test). The numbers in the bars indicate the number of mice. their corresponding proteins in TB mice (P 5 0.54, Mann– administration of BI-2545 (10 mg/kg, i.p.) reduced hypersensi- Whitney rank sum test, N55-6 mice/group for LPA1 and P 5 tivity, normalizing mechanical and heat sensitivity to the level in 0.22, Student t-test for LPA5). The LPAR3 protein was not naive mice (Figs. 1F and G). It has been shown previously that detected in either TB or naive mice but was detected in the spinal this dose of BI-2545 reduces the plasma level of LPA in vivo by cord as a positive control for the antibody. Thus, although there 90%.29 BI-3017, the inactive analog for BI-2545, served as the were no differences in the expression of proteins for any receptor control and had no effect on hypersensitivity. We also examined subtype in DRG between naive and TB mice, H2L5765834 the expression of ATX mRNA and ATX protein in relevant tissues reduced hypersensitivity in TB mice without altering withdrawal to identify sources of increased LPA synthesis. No difference was responses in naive mice (Figs. 1D and E). detected in ATX mRNA between naive and TB mice in L3-L6 DRG Because LPA is produced primarily by the enzyme ATX, we (P 5 0.16, Student t test, N 5 9-10 mice/group), and the ATX determined whether inhibition of ATX would attenuate tumor- protein was not detected in DRG in both groups of mice. evoked hypersensitivity. Because of the high viscosity of the However, the ATX protein was higher in soft tissues from the solvent (10% DMSO, 40% PEG400, 5% Tween-80, and 45% plantar paw of TB mice compared with naive mice (Figs. 1H and saline), the ATX inhibitor, BI-2545, could not be injected into the I), and malignant fibrosarcoma (cancer) cells expressed more paw and therefore was given systemically. Systemic ATX mRNA than nonmalignant fibroblasts (ddCt for cancer cells: December 2023 · Volume 164 · Number 12 www.painjournalonline.com 2689 1.15 6 0.06; fibroblasts: 0.92 6 0.06, P 5 0.02, Student t test, diameters of 74 nm and 96 nm, respectively. The standard N 5 5 cultures/group). Given the occurrence of ATX in the TB paw exosomal markers CD63, CD81, and CD9 were present in both and the known association of extracellular ATX with exosomes,21 types of exosomes (Fig. 2D). It is noteworthy that the we investigated the contribution of exosome-related ATX to production of exosomes by cancer cells was twice that of tumor-evoked hypersensitivity. fibroblasts during the same 48-hour period (Fig. 2E), and they contained more ATX protein than the fibroblast exosomes (Figs. 2F and G). 3.2. Verification of exosomes Like most mammalian cells, NCTC clone 2472 fibrosarcoma 3.3. Cancer exosomes produced hypersensitivity cells and NCTC 929 clone fibroblasts secrete exosomes, nanoparticles that are spherical in shape (Figs. 2A–C). Exo- The ability of cancer exosomes to produce hypersensitivity was somes secreted by fibrosarcoma cells (cancer exosomes) and determined in naive male and female mice. Because the fibroblasts (fibroblast exosomes) were similar in size with mean hypersensitivity and sensitization of nociceptors in TB mice diameters of 129 6 4.28 and 125 6 4.39 nm and modal occurred at the tumor site,4,5,15,55 exosomes were injected into Figure 2. Verification of exosomes. (A) Representative transmission electron microscopy (TEM) images of a fibrosarcoma (cancer) cell surrounded by extracellular vesicles identified as exosomes by size and shape. (B) Collage of 2 TEM images from the cell shown in (A). Arrows point to exosomes. (C) Exosomes exhibit a spherical morphology. (D) Nanoparticle tracking analysis (NTA) indicated a similar size of exosomes of both cell types (P 5 0.46, Student t test), although the concentration (particles/mL) of cancer exosomes was twice that of fibroblast exosomes. Red and black traces represent the average of data from 5 independent cultures in each group. Both populations of exosomes expressed exosomal markers. Tetraspanins CD63, CD81, and CD9 were determined by western blotting on 50 mg of protein in cancer (C) and fibroblast (Fb) exosomes. (E) Cancer cells secreted more exosomes. *Different from fibroblast exosomes at P 5 0.009, Student t test, n 5 5 cultures/group. (F) Representative images of ATX-immunoreactive bands detected in exosomes (50 mg of total protein) isolated from medium conditioned by cancer (C) and fibroblast (Fb) cells. (G) Quantification of ATX protein in cancer (C) and fibroblast (Fb) exosomes was based on the ratio between the fluorescence of the ATX band and the total loaded protein. The results are expressed as a percentage of fibroblast exosomes and are presented as the median and the 25th to 75th percentile range. The numbers in the bars indicate the number of cultures/group. *Different from fibroblast exosomes at P 5 0.005, Mann–Whitney rank sum test. 2690 · I.A. Khasabova et al. 164 (2023) 2684–2695 PAIN® the plantar hind paw. A single injection (i.pl.) of cancer exosomes The hypersensitivity produced by cancer exosomes in naive produced dose-dependent, acute mechanical and heat hyper- mice was mediated by LPARs1,3,5 because co-administration of sensitivity in the injected paw of naive male mice (Figs. 3A and B). cancer exosomes (300 mg) and H2L5765834, the LPAR1,3,5 Mechanical and heat hypersensitivity occurred after doses of 100 antagonist (5 mg/kg), into the paw blocked the development of and 300 mg exosomes/10 mL. Mechanical hypersensitivity mechanical and heat hypersensitivity produced by the exosomes persisted for more than 3 hours and returned to baseline values alone (Figs. 3E and F). The association of cancer exosomes with at 24 hours, whereas heat hypersensitivity lasted for 2 hours after ATX was verified by treating isolated exosomes with the ATX injection. The highest dose (300 mg) of fibroblast exosomes did inhibitor, BI-2545 (1 mM),29 for 1 hour at 37˚C in vitro, before not produce hypersensitivity. A single injection of cancer injecting them into the paw at a dose of 300 mg. Cancer exosomes (300 mg/10 mL, i.pl.) produced similar mechanical exosomes pretreated with BI-2545 did not produce mechanical and heat hypersensitivity in female mice (Figs. 3C and D), or heat hypersensitivity (Figs. 3G and H). indicating comparable effects in both sexes. Both types of cells were sensitive to serum deprivation and were cultured with 10% 3.4. Cancer exosomes sensitized nociceptors horse serum; however, only cancer exosomes caused hyper- sensitivity, negating the possible effect of contamination with To determine whether exosomes from cancer cells produced serum exosomes. hypersensitivity by sensitizing nociceptors, recordings were Figure 3. Cancer exosomes produced hypersensitivity. (A) Single intraplantar (i.pl.) injection of cancer exosomes (30, 100, or 300 mg/10 mL) produced a dose- dependent increase in the withdrawal frequency of the injected paw of naive male mice. The increased response persisted for at least 3 hours (F[4,287] 5 21.68, 2- way RM ANOVA, P , 0.001, n57 to 15 mice/group, followed by Bonferroni t test). Fibroblast exosomes injected at the highest dose (300 mg/10 mL) did not produce hypersensitivity (P 5 0.46, n 5 7 mice). *Different from baseline (BL) at P , 0.001; #different from saline at P 5 0.05. (B) Heat hypersensitivity was evident 1 hour after i.pl. administration of cancer exosomes in naive male mice (F[4,186] 5 4.15, 2-way RM ANOVA, P 5 0.005, n 5 12 to 16 mice/group, followed by Bonferroni t test). The highest dose of fibroblast exosomes (300 mg/10 mL) had no effect (P 5 0.46, n 5 15 mice). *Different from BL at P 5 0.005; #different from saline at P 5 0.007. (C and D) As in male mice, cancer exosomes (300 mg/10 mL, i.pl.) induced acute mechanical (F[2,112] 5 68.78, 2-way RM ANOVA, P , 0.001, n 5 6 to 7 mice/group, followed by Bonferroni t test) and heat hypersensitivity (F[2, 45] 5 3.92, 2-way RM ANOVA, P , 0.05, n 5 6 mice/group, followed by Bonferroni t test) in female mice. The same dose of fibroblast exosomes (300 mg/10 mL) had no effect (P 5 0.55 and P 5 0.20 for mechanical and heat hypersensitivity, respectively, n 5 6 mice/group). *Different from BL at P , 0.001; #different from saline at P , 0.01. (E and F) Hypersensitivity produced by cancer exosomes was mediated by LPARs1,3,5. H2L5765834, the LPAR1,3,5 antagonist (5 mg/kg, i.pl.), co-injected with cancer exosomes eliminated the mechanical (F1,60 5 8.59, 2-way RM ANOVA, P , 0.02, followed by Bonferroni t test, n 5 6 mice/mixed-sex group) and heat hypersensitivity (F [1,30] 5 7.54, 2-way RM ANOVA, P , 0.03, n 5 6 mice/group, followed by Bonferroni t test) evoked in naive mice by cancer exosomes treated with vehicle (15% DMSO, 5% Tween-80, and 80% saline). *Different from BL at P , 0.01; #different from mice treated with intact cancer exosomes at P , 0.02. (G and H) Pretreatment of cancer exosomes in vitro with the ATX inhibitor BI-2545 (1 mM, 1 hour, 37˚C) also eliminated the mechanical (F [1,48] 5 20.63, 2-way RM ANOVA, P 5 0.002, n 5 6 mice/group, followed by Bonferroni t test) and heat hypersensitivity (F [1,24] 5 11.03, 2-way RM ANOVA, P , 0.011, n 5 6 mice/mixed-sex group, followed by Bonferroni t test) produced by cancer exosomes pretreated with vehicle (10% DMSO, 40% PEG400, 5% Tween-80, and 45% saline). *Different from BL at P , 0.05; #different from mice treated with intact cancer exosomes at P , 0.02. December 2023 · Volume 164 · Number 12 www.painjournalonline.com 2691 made from 39 C-fiber nociceptors from the tibial nerve in naive of cancer exosomes on evoked responses of nociceptors were male and female mice. All nociceptors had RF areas on the independent of sex (F2,26 5 0.244, 2-way RM ANOVA, P 5 plantar hind paw and were responsive to mechanical stimulation. 0.786). Of the 39 nociceptors, 6 were excited by heat, 5 were excited by cold, and 8 were excited by both heat and cold stimuli. We did not 3.5. ATX–LPA–LPAR pathway contributes to the examine effects of exosomes on responses to heat and cold sensitization of DRG neurons evoked by cancer exosomes because of the relatively low number of temperature-sensitive fibers in each experimental group. None of the nociceptors To determine whether exosomes had a direct effect on the exhibited spontaneous activity before injection. Injection of excitability of small DRG neurons (area , 500 mm2), we measured vehicle into the RF did not evoke spontaneous activity in any of chemically evoked Ca21 transients in a bioassay of neuronal the 10 tested nociceptors, and injection of fibroblast exosomes sensitization. We previously showed that this assay detected (300 mg/10 mL) caused a low level of spontaneous activity (0.001 sensitization of small DRG neurons measured as increased 6 0.01 impulses/sec) only in 2 of the 14 nociceptors that responses evoked by 25 mM KCl and capsaicin in noncontact persisted for 60 minutes. By contrast, all 15 nociceptors exposed co-cultures of DRG neurons and fibrosarcoma (cancer) cells for to cancer exosomes (300 mg/10 mL) developed spontaneous 24 hours.24,26 Isolated DRG neurons (24 hours in vitro) were activity (0.2 6 0.09 impulses/sec at 30 minutes, and 0.1 6 0.05 incubated (60 minutes, 37 ˚C) with different concentrations of impulses/sec at 60 minutes, Figs. 4A and B). cancer exosomes (10, 30, and 100 mg/mL in HEPES) in the Exosomes from cancer cells decreased the mechanical presence of 3 mM indo-1. Cancer exosomes (30 and 100 mg/mL) response threshold of nociceptors (Fig. 4C). Thresholds of C increased basal [Ca21]i (P 5 0.002, one-way ANOVA, followed by fibers did not differ between the groups before any injection and Kruskal–Wallis test on ranks for the 2 highest concentrations), were not altered at any time after injection of vehicle or exosomes whereas fibroblast (control) exosomes had no effect (P 5 0.09). from fibroblasts. By contrast, mechanical thresholds decreased Moreover, cancer exosomes increased the number of small DRG after injection of cancer exosomes from 1.5 6 0.25 g before neurons in which KCl (25 mM, 10 seconds) and capsaicin (200 injection to 0.8 6 0.12 g at 30 minutes and to 0.7 6 0.1 g at 60 nM, 30 seconds) evoked a Ca21 transient, indicating sensitization minutes after injection. Exosomes from cancer cells also in- (Figs. 5A and B). Fibroblast exosomes (100 mg/mL) had no effect creased responses evoked by a suprathreshold von Frey on the frequency of evoked Ca21 transients. monofilament that delivered a force of 149 mN (Figs. 4D and The contribution of LPA was determined by incubating DRG E). The number of evoked impulses before any injection did not neurons with cancer exosomes that were pretreated for 1 hour differ between the groups and was not altered after injection of with the ATX inhibitor, BI-2545 (1 mM), which has been shown to vehicle or fibroblast exosomes. However, cancer exosomes decrease levels of LPA in plasma by approximately 90%.29The increased the number of evoked impulses from 47 6 6.4 contribution of LPAR to neuronal sensitization produced by impulses before injection, to 101 6 12.1 impulses at 30 minutes, cancer exosomes was tested by superfusing cancer exosome- and 68 6 10.9 impulses at 60 minutes after injection. The effects treated DRG neurons with the antagonist H2L5765834 (10 mM) Figure 4. Exosomes from cancer cell–sensitized nociceptors. (A) Representative examples of spontaneous activity of single C-fiber nociceptors before and at 30 minutes after i.pl. injection of vehicle or exosomes from fibroblasts and cancer cells (300 mg in 10 mL). (B) Discharge rate of spontaneous activity before (baseline, BL) and at 30 and 60 minutes after injections. Discharge rate increased after injection of cancer exosomes (F[4,72] 5 3.05, 2-way RM ANOVA with Bonferroni t test, P, 0.05, n 5 10 to 15 nociceptors/group). (C) Mechanical response thresholds before and at 30 and 60 minutes after injection of vehicle, or exosomes from fibroblasts or cancer cells. Mechanical thresholds decreased after injection of cancer exosomes (F[4,72] 5 5.26, 2-way RM ANOVA with Bonferroni t test, P, 0.001). Data are reported as the mean 6 SEM. (D) Representative examples of responses of single C-fiber nociceptors evoked by a von Frey monofilament with a bending force of 149 mN applied to the receptive field for 5 seconds (horizontal bar) before and at 30 minutes after vehicle, fibroblast exosomes, or cancer exosomes. (E) Number of impulses evoked by 149 mN before and at 30 and 60 minutes after vehicle or exosomes from fibroblasts or cancer cells (F[4,72] 5 12.99, 2-way RM ANOVA with Bonferroni t test, P, 0.001). Data are reported as the mean 6 SEM. *Different from baseline; #different from group treated with fibroblast exosomes at specific time points; ^different from group treated with vehicle at corresponded time points. **** #### ^^^^ P , 0.001; *** P , 0.005; * # ^ P , 0.05. 2692 · I.A. Khasabova et al. 164 (2023) 2684–2695 PAIN® Figure 5. Cancer exosomes sensitized small DRG neurons in vitro. Sensitization of neurons was defined as an increase in the occurrence of a Ca21 transient evoked by KCl (25 mM) or capsaicin (200 nM). The occurrence of a Ca21 transient in response to superfusion with KCl (A) or capsaicin (B) increased in small DRG neurons treated with cancer (C) exosomes (10, 30, or 100 mg of total protein/mL, 60 minutes). The highest concentration of fibroblast (Fb) exosomes (100 mg/mL) had no effect (P 5 0.27, chi-square test). (A)*Different from vehicle (HEPES) at P , 0.03; #different from the fibroblast exosome-treated group at P , 0.02, chi- square test. (B) *Different from vehicle at P 5 0.01; #different from the fibroblast exosome-treated group at P 5 0.002, chi-square test. The sample size for each treatment group seems within each bar. (C and D) Superfusion of cancer exosome-treated cultures (100 mg/mL, 60 minutes) with either the LPAR1,3,5 antagonist H2L5765834 (10 mM, 10 minutes) or the ATX inhibitor BI-2545 (1 mM, 10 minutes) reduced the frequency of Ca21 transients evoked in small DRG neurons in response to KCl (C) or capsaicin (D). *Different from other groups at P , 0.05, chi-square test. DRG neurons incubated with cancer exosomes for 60 minutes were then superfused with H2L5765834 (10 mM, 10 minutes) or vehicle (0.004% DMSO in HEPES) before testing responses to KCl or capsaicin. H2L5765834 (10 mM) alone had no effect on intact DRG neurons (P 5 1.0, chi-square test). Cancer exosomes were pretreated with BI-2545 (1 mM, 1 hour) or vehicle (10% DMSO, 40% PEG400, 5% Tween-80, and 45% saline) before incubation with DRG neurons. 10 minutes before stimulation with KCl or capsaicin. Both the ATX We showed that intraplantar administration of cancer exo- inhibitor (Fig. 5C) and the LPAR1,3,5 antagonist (Fig. 5D) reduced somes produced acute local mechanical and heat hypersensi- the sensitization of small DRG neurons produced by cancer tivity in naive mice through ATX–LPA–LPAR-mediated exosomes. The concentration of H2L5765834 (10 mM) used was sensitization of nociceptors. Cancer exosomes induced compa- based on its ability to attenuate calcium transients.58 rable nociceptor sensitization and hypersensitivity in male and female mice, indicating a common, sex-independent mechanism related to the malignancy of the cells because nonmalignant 4. Discussion fibroblast exosomes had no effect in either sex. Hypersensitivity Nerves are an essential element of the tumor microenvironment. induced by cancer exosomes was mediated by LPARs1,3,5 and Perineural invasion underlies tumor proliferation, angiogenesis, was related to ATX activity because hypersensitivity did not and inflammation. In turn, the tumor sensitizes nociceptors and develop if exosomes were treated with an ATX inhibitor in vitro produces pain by mechanical compression and the release of before administration. algogenic mediators.49 The contribution to pain of some Despite the well-defined oncogenic roles of LPA, including algogenic mediators of cancer exosomes was recently estab- cancer cell proliferation, angiogenesis, and metastasis,47 the lished.2,10 Our study extended these observations by addressing contribution of LPA signaling to cancer pain has received little a mechanism underlying cancer exosome-induced pain, namely, attention.60 Our data confirm the involvement of LPA signaling in ATX–LPA–LPAR signaling associated with cancer exosomes, as both fibrosarcoma (cancer) cell proliferation and fibrosarcoma- summarized in Figure 6. induced hypersensitivity. Namely, hypersensitivity in TB mice was December 2023 · Volume 164 · Number 12 www.painjournalonline.com 2693 Figure 6. Schematic representation of the pronociceptive ATX–LPA–LPAR pathway associated with cancer exosomes. ATX bound to cancer exosomes cleaves choline from lysophosphatidylcholine (LPC) to form LPA, an agonist of LPARs1,3,5. Activation of LPARs1,3,5 on DRG neurons causes sensitization of nociceptors and pain. Inhibition of ATX by BI-2545 or blocking of LPA1,3,5 receptors with the antagonist H2L5765834 reduces tumor-evoked hypersensitivity. associated with an elevated plasma level of LPA and was reduced an increased basal [Ca21]i in small DRG neurons. A similar by blocking LPARs1,3,5. Unlike an earlier report60 using a change was observed in neurons co-cultured with fibrosarcoma mammary gland carcinoma cell line, we did not observe an cells and neurons isolated from TB mice.24 Elevated basal increase in LPAR1,3,5 expression in DRG at either the mRNA or [Ca21]i can be attributed to LPARs, TRPV1, and Nav1.8 protein levels. It is noteworthy that of the 3 LPA species elevated activation by LPA associated with cancer exosomes and may in TB mice, 18:1- and 20:4-LPA have clinical relevance, underlie neuronal sensitization. Cancer exosomes sensitized correlating with the intensity of neuropathic pain of various DRG neurons in an LPAR-dependent manner because their etiologies in humans.30,35 Indeed, both LPA species showed high effect was blocked by an LPAR antagonist. In addition, neuronal agonist potency at LPAR1 and LPAR3 in an animal model of sensitization did not occur when cancer exosomes were neuropathic pain.35 The residual hypersensitivity in the presence pretreated with an ATX inhibitor. of the LPAR antagonist may be due to direct LPA binding to Thus, our studies demonstrate an essential role for cancer TRPV1,6,40,41 or modulation of Nav1.8 channels in DRG exosomes in the generation of hypersensitivity through the neurons.31 ATX–LPA–LPAR pathway in a murine model of bone cancer Importantly, hypersensitivity in TB mice was reduced, and pain. Although we focused on the key roles of ATX and LPA, other blocked at some time points, by inhibition of ATX, a major LPA- enzymes and bioactive lipids contained in exosomes may also producing enzyme.42 In our study, the increase in LPA in TB mice contribute to hypersensitivity. For example, exosomes contain all was associated with expression of ATX in the tumor mass. 3 A2 phospholipases that hydrolyze glycerophospholipids to Because of the greater capacity of fibrosarcoma (cancer) cells to produce arachidonic acid. Exosomes express constitutive and synthesize ATX, these cells are likely the main source of ATX inducible cyclooxygenases that convert arachidonic acid into production in the tumor mass, but the role of activated immune prostaglandin PGH2, which is subsequently transformed to the and stromal cells cannot be excluded.8 Reduced degradation of algogen prostaglandin PGE2.9 In addition, cancer exosomes may LPA by extracellular phosphate phosphatases may also enhance be involved in pain through the matrix metalloproteinase-1 the pool of LPA.17 expressed on their surface.2,10 Given that a significant proportion of secreted ATX in animals Thus, we conclude that exosomes secreted by cancer cells and humans is present in an exosome-bound form that is affect tumor-evoked pain. Moreover, pharmacological targeting considered physiological,21 it is not surprising that the ATX of ATX–LPA–LPAR signaling can have a double clinical benefit: to protein was identified in exosomes secreted by both cancer cells reduce the pain caused by the tumor and to reduce the and fibroblasts. Isolated exosomes exhibited representative proliferation of cancer cells. characteristics that include spherical shape, size (40-200 nm), and the presence of standard exosomal markers CD63, CD81, Conflict of interest statement and CD9.37 Importantly, exosomal ATX retains the ability to synthesize LPA. Binding of ATX to membrane integrins on target The authors have no conflict of interest to declare. cells increases and localizes the catalytic activity of ATX on target cells providing the generation of LPA in the vicinity of its Acknowledgements receptors.13 On binding of ATX to integrins on the surface of target cells, LPA is released and then generated again.21 The authors thank Jacob Gable for his help in studying the Consistent with earlier studies,2 we showed that fibrosarcoma effect of exosomes on neuronal sensitization in vitro. LC-MS/MS (cancer) cells secreted more exosomes than nonmalignant analysis was performed at the UND SOMHS MS core facility fibroblasts. Moreover, cancer exosomes contained more ATX supported by the SOMHS Dean’s office. Electron microscopy per unit of exosome mass defined by protein. Evidence that was performed in the Characterization Facility, University of exosome-bound ATX is not only a synthetic enzyme of LPA but Minnesota, which receives support from the NSF through the also carries LPA to distally located LPAR21,42 confers an MRSEC (Award Number DMR-2011401) and the NNCI (Award exceptional role for cancer exosomes in LPA-mediated Number ECCS-2025124) programs. Particle analysis was con- hypersensitivity. ducted in the Minnesota Nano Center, which is supported by the Although we cannot rule out the contribution of immune and NSF through the National Nanotechnology Coordinated In- glial cells to nociceptor sensitization produced by cancer frastructure (NNCI) under Award Number ECCS-2025124. exosomes in vivo, their direct effect on nociceptor sensitization This work was supported by NIH grants CA241627 to D.A.S., has been demonstrated in vitro. Specifically, we demonstrated CA263777 to S.G.K., and NS119279 to M.Y.G. 2694 · I.A. Khasabova et al. 164 (2023) 2684–2695 PAIN® I.A.K. designed, performed, analyzed, interpreted data, and Hargreaves K, Dubner R, Brown F, Flores C, Joris J. A new and sensitive wrote the manuscript. S.G.K. and D.C. designed, performed and method for measuring thermal nociception in cutaneous hyperalgesia. PAIN 1988;32:77–88. analyzed electrophysiological data, created the associated figure, Harper K, Brochu-Gaudreau K, Saucier C, Dubois CM. Hypoxia and edited the manuscript. M.J. and J.J. performed the RT-qPCR downregulates LPP3 and promotes the spatial segregation of ATX and and the western blot studies. J.J., K.A., and A.F. performed LPP1 during cancer cell invasion. Cancers (Basel) 2019;11:1403. behavioral experiments. A.F. performed cell cultures and con- Hiura A, Sakamoto Y. Quantitative estimation of the effects of capsaicin on the mouse primary sensory neurons. Neurosci Lett 1987;76:101–6. tributed to LPA treatment of fibrosarcoma cells in vitro. M.Y.G. Hwang H, Kim EK, Park J, Suh PG, Cho YK. RhoA and Rac1 play and S.A.G. performed the mass spectrometric assay and independent roles in lysophosphatidic acid-induced ovarian cancer associated data analysis. M.J. performed LPA ELISA analysis. chemotaxis. Integr Biol (Camb) 2014;6:267–76. W.Z. designed and performed transmission electron microscopy. Inoue M, Rashid MH, Fujita R, Contos JJ, Chun J, Ueda H. Initiation of J.M. designed and performed the nanoparticle analysis. V.S.S. neuropathic pain requires lysophosphatidic acid receptor signaling. Nat Med 2004;10:712–718. contributed to design of experiments, interpretation of data, Jethwa SA, Leah EJ, Zhang Q, Bright NA, Oxley D, Bootman MD, Rudge creation of figures, and writing of the manuscript. D.A.S. initiated SA, Wakelam MJ. Exosomes bind to autotaxin and act as a physiological and coordinated the study, and contributed to the electrophys- delivery mechanism to stimulate LPA receptor signalling in cells. J Cel Sci iology experiments and writing of the manuscript. 2016;129:3948–57. Khasabov SG, Hamamoto DT, Harding-Rose C, Simone DA. Tumor- evoked hyperalgesia and sensitization of nociceptive dorsal horn neurons Article history: in a murine model of cancer pain. Brain Res 2007;1180:7–19. Received 22 February 2023 Khasabova IA, Chandiramani A, Harding-Rose C, Simone DA, Seybold Received in revised form 5 April 2023 VS. Increasing 2-arachidonoyl glycerol signaling in the periphery Accepted 11 April 2023 attenuates mechanical hyperalgesia in a model of bone cancer pain. Pharmacol Res 2011;64:60–7. Available online 6 June 2023 Khasabova IA, Stucky CL, Harding-Rose C, Eikmeier L, Beitz AJ, Coicou LG, Hanson AE, Simone DA, Seybold VS. Chemical interactions between fibrosarcoma cancer cells and sensory neurons contribute to cancer pain. References J Neurosci 2007;27:10289–98. Balijepalli P, Sitton CC, Meier KE. Lysophosphatidic acid signaling in Khasabova IA, Golovko MY, Golovko SA, Simone DA, Khasabov SG. cancer cells: what makes LPA so special? Cells 2021;10:2059. Intrathecal administration of Resolvin D1 and E1 decreases Bhattacharya A, Janal MN, Veeramachaneni R, Dolgalev I, hyperalgesia in mice with bone cancer pain: involvement of Dubeykovskaya Z, Tu NH, Kim H, Zhang S, Wu AK, Hagiwara M, Kerr endocannabinoid signaling. Prostaglandins Other Lipid Mediators AR, DeLacure MD, Schmidt BL, Albertson DG. Oncogenes 2020;151:106479. overexpressed in metastatic oral cancers from patients with pain: Khasabova IA, Khasabov SG, Harding-Rose C, Coicou LG, Seybold BA, potential pain mediators released in exosomes. Sci Rep 2020;10: Lindberg AE, Steevens CD, Simone DA, Seybold VS. A decrease in 14724. anandamide signaling contributes to the maintenance of cutaneous Brose SA, Golovko SA, Golovko MY. Brain 2-arachidonoylglycerol levels mechanical hyperalgesia in a model of bone cancer pain. J Neurosci are dramatically and rapidly increased under acute ischemia-injury which 2008;28:11141–52. is prevented by microwave irradiation. Lipids 2016;51:487–95. Khasabova IA, Xiong Y, Coicou LG, Piomelli D, Seybold V. Peroxisome Cain DM, Wacnik PW, Eikmeier L, Beitz A, Wilcox GL, Simone DA. proliferator-activated receptor a mediates acute effects of Functional interactions between tumor and peripheral nerve in a model of palmitoylethanolamide on sensory neurons. J Neurosci 2012;32: cancer pain in the mouse. Pain Med 2001;2:15–23. 12735–43. Cain DM, Wacnik PW, Turner M, Wendelschafer-Crabb G, Kennedy WR, Kowal J, Arras G, Colombo M, Jouve M, Morath JP, Primdal-Bengtson B, Wilcox GL, Simone DA. Functional interactions between tumor and Dingli F, Loew D, Tkach M, Théry C. Proteomic comparison defines novel peripheral nerve: changes in excitability and morphology of primary markers to characterize heterogeneous populations of extracellular afferent fibers in a murine model of cancer pain. J Neurosci 2001;21: vesicle subtypes. Proc Natl Acad Sci U.S.A 2016;113:E968–77. 9367–76. Kuttruff CA, Ferrara M, Bretschneider T, Hoerer S, Handschuh S, Nosse Canul-Sánchez JA, Hernández-Araiza I, Hernández-Garcı́a E, Llorente I, B, Romig H, Nicklin P, Roth GJ. Discovery of BI-2545: a novel autotaxin Morales-Lázaro SL, Islas LD, Rosenbaum T. Different agonists induce inhibitor that significantly reduces LPA levels in vivo. ACS Med Chem Lett distinct single-channel conductance states in TRPV1 channels. J Gen 2017;8:1252–7. Physiol 2018;150:1735–46. Kuwajima K, Sumitani M, Kurano M, Kano K, Nishikawa M, Uranbileg B, Chuo ST, Chien JC, Lai CP. Imaging extracellular vesicles: current and Tsuchida R, Ogata T, Aoki J, Yatomi Y, Yamada Y. Lysophosphatidic acid emerging methods. J Biomed Sci 2018;25:91. is associated with neuropathic pain intensity in humans: an exploratory Corn KC, Windham MA, Rafat M. Lipids in the tumor microenvironment: study. PLoS One 2018;13:e0207310. from cancer progression to treatment. Prog Lipid Res 2020;80:101055. Seung Lee W, Hong MP, Hoon Kim T, Kyoo Shin Y, Soo Lee C, Park M, Donoso-Quezada J, Ayala-Mar S, González-Valdez J. The role of lipids in Song JH. Effects of lysophosphatidic acid on sodium currents in rat dorsal exosome biology and intercellular communication: function, analytics and root ganglion neurons. Brain Res 2005;1035:100–4. applications. Traffic 2021;22:204–20. Lin ME, Rivera RR, Chun J. Targeted deletion of LPA5 identifies novel Dubeykovskaya ZA, Tu NH, Garcia PDR, Schmidt BL, Albertson DG. Oral roles for lysophosphatidic acid signaling in development of neuropathic cancer cells release vesicles that cause pain. Adv Biol 2022;6:e2200073. pain. J Biol Chem 2012;287:17608–17. Düll MM, Stengel M, Ries V, Strupf M, Reeh PW, Kremer AE, Namer B. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using Lysophosphatidic acid activates nociceptors and causes pain or itch real-time quantitative PCR and the 22DDCt method. Methods 2001;25: depending on the application mode in human skin. PAIN 2022;163: 402–8. 445–60. Ma L, Uchida H, Nagai J, Inoue M, Chun J, Aoki J, Ueda H. Euer N, Schwirzke M, Evtimova V, Burtscher H, Jarsch M, Tarin D, Weidle Lysophosphatidic acid-3 receptor-mediated feed-forward production of UH. Identification of genes associated with metastasis of mammary lysophosphatidic acid: an initiator of nerve injury-induced neuropathic carcinoma in metastatic versus non-metastatic cell lines. Anticancer Res pain. Mol Pain 2009;5:1744-8069-5-64. 2002;22:733–40. Ma L, Nagai J, Chun J, Ueda H. An LPA species (18:1 LPA) plays key roles Fulkerson Z, Wu T, Sunkara M, Kooi CV, Morris AJ, Smyth SS. Binding of in the self-amplification of spinal LPA production in the peripheral autotaxin to integrins localizes lysophosphatidic acid production to neuropathic pain model. Mol Pain 2013;9:1744-8069-9-29. platelets and mammalian cells. J Biol Chem 2011;286:34654–63. Madeo M, Colbert PL, Vermeer DW, Lucido CT, Cain JT, Vichaya EG, Goldsmith ZG, Ha JH, Jayaraman M, Dhanasekaran DN. Grossberg AJ, Muirhead D, Rickel AP, Hong Z, Zhao J, Weimer JM, Lysophosphatidic acid stimulates the proliferation of ovarian cancer Spanos WC, Lee JH, Dantzer R, Vermeer PD. Cancer exosomes induce cells via the gep proto-oncogene G 12. Genes Cancer 2011;2:563–75. tumor innervation. Nat Commun 2018;9:4284. Hamamoto DT, Khasabov SG, Cain DM, Simone DA. Tumor-evoked Mathieu M, Névo N, Jouve M, Valenzuela JI, Maurin M, Verweij FJ, sensitization of C nociceptors: a role for endothelin. J Neurophysiol 2008; Palmulli R, Lankar D, Dingli F, Loew D, Rubinstein E, Boncompain G, 100:2300–11. Perez F, Théry C. Specificities of exosome versus small ectosome December 2023 · Volume 164 · Number 12 www.painjournalonline.com 2695 secretion revealed by live intracellular tracking of CD63 and CD9. Nat squamous cell carcinoma contributes to cancer pain and associated Commun 2021;12:4389. inflammation. PAIN 2017;158:2396–409. McDonald MK, Capasso KE, Ajit SK. Purification and microRNA profiling Schmidt BL. The neurobiology of cancer pain. Neuroscientist 2014;20: of exosomes derived from blood and culture media. J visualized 546–62. experiments: JoVE 2013;76:e50294. Schmidt BL, Hamamoto DT, Simone DA, Wilcox GL. Mechanism of Mercadante S. Malignant bone pain: pathophysiology and treatment. cancer pain. Mol Interventions 2010;10:164–78. PAIN 1997;69:1–18. Seeger DR, Golovko SA, Golovko MY. Blood–brain barrier is the major Morales-Lázaro SL, Serrano-Flores B, Llorente I, Hernández-Garcı́a E, site for a rapid and dramatic prostanoid increase upon brain global González-Ramı́rez R, Banerjee S, Miller D, Gududuru V, Fells J, Norman ischemia. Lipids 2020;55:79–85. D, Tigyi G, Escalante-Alcalde D, Rosenbaum T. Structural determinants Ueda H, Ueda M. Lysophosphatidic acid as an initiator of neuropathic of the transient receptor potential 1 (TRPV1) channel activation by pain: biosynthesis and demyelination. Clin Lipidol 2011;6:147–58. phospholipid analogs. J Biol Chem 2014;289:24079–90. Uhelski ML, Bruce D, Speltz R, Wilcox GL, Simone DA. Topical Nieto-Posadas A, Picazo-Juárez G, Llorente I, Jara-Oseguera A, application of loperamide/oxymorphindole, mu and delta opioid Morales-Lázaro S, Escalante-Alcalde D, Islas LD, Rosenbaum T. receptor agonists, reduces sensitization of C-fiber nociceptors that Lysophosphatidic acid directly activates TRPV1 through a C-terminal possess NaV1.8. Neuroscience 2020;446:102–12. binding site. Nat Chem Biol 2011;8:78–85. Uhelski ML, Bruce DJ, Séguéla P, Wilcox GL, Simone DA. In vivo Nishimasu H, Okudaira S, Hama K, Mihara E, Dohmae N, Inoue A, Ishitani optogenetic activation of Nav1.81 cutaneous nociceptors and their R, Takagi J, Aoki J, Nureki O. Crystal structure of autotaxin and insight responses to natural stimuli. J Neurophysiol 2017;117:2218–23. into GPCR activation by lipid mediators. Nat Struct Mol Biol 2011;18: Uhelski ML, Cain DM, Harding-Rose C, Simone DA. The non-selective 205–12. cannabinoid receptor agonist WIN 55,212-2 attenuates responses of C- Panupinthu N, Lee HY, Mills GB. Lysophosphatidic acid production and fiber nociceptors in a murine model of cancer pain. Neuroscience 2013; action: critical new players in breast cancer initiation and progression. Br J 247:84–94. Cancer 2010;102:941–6. Uhelski ML, Gupta K, Simone DA. Sensitization of C-fiber nociceptors in Pearce RJ, Duchen MR. Differential expression of membrane currents in mice with sickle cell disease is decreased by local inhibition of dissociated mouse primary sensory neurons. Neuroscience 1994;63: anandamide hydrolysis. PAIN 2017;158:1711–1722. 1041–56. Urch EC, Donovan-Rodriguez T, Dickenson HA. Alterations in dorsal horn Peyruchaud O, Saier L, Leblanc R. Autotaxin implication in cancer neurones in a rat model of cancer-induced bone pain. PAIN 2003;106: metastasis and autoimunne disorders: functional implication of binding 347–56. autotaxin to the cell surface. Cancers (Basel) 2019;12:105. Williams JR, Khandoga AL, Goyal P, Fells JI, Perygin DH, Siess W, Parrill Rivera RR, Lin ME, Bornhop EC, Chun J. Conditional Lpar1 gene AL, Tigyi G, Fujiwara Y. Unique ligand selectivity of the GPR92/LPA5 targeting identifies cell types mediating neuropathic pain. FASEB J 2020; lysophosphatidate receptor indicates role in human platelet activation. 34:8833–42. J Biol Chem 2009;284:17304–19. Sanderson RD, Bandari SK, Vlodavsky I. Proteases and glycosidases on Yung YC, Stoddard NC, Chun J. LPA receptor signaling: pharmacology, the surface of exosomes: newly discovered mechanisms for extracellular physiology, and pathophysiology. J Lipid Res 2014 55:1192–214. remodeling. Matrix Biol 2019;75-76:160–9. Zhao J, Pan HL, Li TT, Zhang YQ, Wei JY, Zhao ZQ. The sensitization of Scheff NN, Ye Y, Bhattacharya A, MacRae J, Hickman DN, Sharma AK, peripheral C-fibers to lysophosphatidic acid in bone cancer pain. Life Sci Dolan JC, Schmidt BL. Tumor necrosis factor alpha secreted from oral 2010;87:120–5.