Article Text

Mechanosensitive TRPV4 is required for crystal-induced inflammation
  1. Zhou Lan1,2,3,
  2. Lvyi Chen1,2,4,
  3. Jing Feng1,2,
  4. Zili Xie1,2,
  5. Zhiyong Liu5,
  6. Fang Wang2,6,
  7. Peng Liu4,
  8. Xueping Yue1,2,
  9. Lixia Du1,2,
  10. Yonghui Zhao1,2,
  11. Pu Yang1,2,
  12. Jialie Luo1,2,
  13. Zhe Zhu7,
  14. Xueming Hu1,2,
  15. Liang Cao1,2,
  16. Ping Lu1,2,
  17. Rajan Sah8,
  18. Kory Lavine8,
  19. Brian Kim1,2,6,9,
  20. Hongzhen Hu1,2,6
  1. 1 Department of Anesthesiology, Washington University School of Medicine in St Louis, St. Louis, Missouri, USA
  2. 2 Center for the Study of Itch and Sensory Disorders, Washington University School of Medicine in St Louis, St Louis, Missouri, USA
  3. 3 School of Pharmacy, Hubei University of Chinese Medicine, Wuhan, Hubei, People's Republic of China
  4. 4 School of Pharmaceutical Sciences, South-Central University for Nationalities, Wuhan, Hubei, People's Republic of China
  5. 5 Renmin Hospital of Wuhan University, Wuhan, Hubei, China
  6. 6 Division of Dermatology, Department of Medicine, Washington University School of Medicine in St Louis, St Louis, Missouri, USA
  7. 7 Herbert Irving Comprehensive Cancer Center, Columbia University, New York, NY, USA
  8. 8 Department of Internal Medicine, Cardiovascular Division, Washington University School of Medicine in St Louis, St Louis, Missouri, USA
  9. 9 Department of Pathology and Immunology, Washington University School of Medicine in St Louis, St Louis, Missouri, USA
  1. Correspondence to Dr Lvyi Chen, Washington University in St Louis School of Medicine, St Louis, MO 63130, USA; clyhappy05{at}163.com; Dr Hongzhen Hu; hongzhen.hu{at}wustl.edu

Abstract

Crystal structures activate innate immune cells, especially macrophages and initiate inflammatory responses. We aimed to understand the role of the mechanosensitive TRPV4 channel in crystal-induced inflammation. Real-time RT-PCR, RNAscope in situ hybridisation, and Trpv4eGFP mice were used to examine TRPV4 expression and whole-cell patch-clamp recording and live-cell Ca2+ imaging were used to study TRPV4 function in mouse synovial macrophages and human peripheral blood mononuclear cells (PBMCs). Both genetic deletion and pharmacological inhibition approaches were used to investigate the role of TRPV4 in NLRP3 inflammasome activation induced by diverse crystals in vitro and in mouse models of crystal-induced pain and inflammation in vivo. TRPV4 was functionally expressed by synovial macrophages and human PBMCs and TRPV4 expression was upregulated by stimulation with monosodium urate (MSU) crystals and in human PBMCs from patients with acute gout flares. MSU crystal-induced gouty arthritis were significantly reduced by either genetic ablation or pharmacological inhibition of TRPV4 function. Mechanistically, TRPV4 mediated the activation of NLRP3 inflammasome by diverse crystalline materials but not non-crystalline NLRP3 inflammasome activators, driving the production of inflammatory cytokine interleukin-1β which elicited TRPV4-dependent inflammatory responses in vivo. Moreover, chemical ablation of the TRPV1-expressing nociceptors significantly attenuated the MSU crystal-induced gouty arthritis. In conclusion, TRPV4 is a common mediator of inflammatory responses induced by diverse crystals through NLRP3 inflammasome activation in macrophages. TRPV4-expressing resident macrophages are critically involved in MSU crystal-induced gouty arthritis. A neuroimmune interaction between the TRPV1-expressing nociceptors and the TRPV4-expressing synovial macrophages contributes to the generation of acute gout flares.

  • arthritis
  • experimental
  • cytokines
  • gout
  • immune system diseases

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Key messages

What is already known about this subject?

  • Many exogenous or endogenously formed crystal structures with different compositions can act as damage-associated molecular patterns and in turn trigger NLRP3 inflammasome activation.

  • Ion channels are critical signalling mediators in both innate and adaptive immune cells.

What does this study add?

  • Macrophage-expressed TRPV4 is essential to the generation of crystal-induced joint pain and inflammation and inflammatory responses evoked by diverse crystalline structures.

  • TRPV4 is critically involved in the activation of macrophage NLRP3 inflammasome and production of the inflammatory cytokine interleukin-1β induced by diverse crystalline materials but not non-crystalline NLRP3 inflammasome activators.

  • A neuroimmune interaction between the TRPV1-expressing primary nociceptors and the TRPV4-expressing synovial macrophages is an important cellular mechanism behind acute gout flares.

How might this impact on clinical practice or future developments?

  • Our findings show that macrophage-expressed TRPV4 is a common regulator of crystal-induced inflammatory responses and shed new insights into the development of new therapeutic interventions to treat inflammatory conditions associated with crystal disposition.

Introduction

Innate immune cells recognise a variety of pathogen-associated molecular patterns (PAMPs) and damage-associated molecular patterns (DAMPs) to mount a rapid inflammatory response.1 2 One way in which cells mediate this process is through the NLRP3 inflammasome, an intracellular sensor of both exogenous and endogenous stimuli such as microbes, misfolded proteins and adenosine triphosphate (ATP). Depending on the signal, this pathway activates a variety of cellular processes including inflammatory cytokine production and programmed cell death by pyroptosis to drive an innate antimicrobial response.1 However, as with many host-protective responses, this process can also become highly pathologic.

Many exogenous or endogenously formed crystal structures with different compositions can act as DAMPs and in turn trigger NLRP3 inflammasome activation.3 4 For instance, exogenous silica crystals and asbestosis fibres are well known to cause fibrosis and chronic airway disease. It is also recognised that innate immune cells such as macrophages (MΦs) internalise both endogenous and exogenous crystals and thereby activate the NLRP3 inflammasome, driving many inflammatory diseases such as gout, pseudogout, atherosclerosis and silicosis.5 6 Currently, there are at least three hypotheses for MΦ recognition and NLRP3 inflammasome activation by crystals and nanoparticles: (1) The NLRP3 inflammasome senses the disturbance of cellular homeostasis rather than directly recognising various types of crystals7; (2) Distinct membrane-bound receptors serve as molecular sensors for distinct crystals to initiate NLRP3 inflammasome activation upon interaction5 and (3) Phagocyte membrane engagement by crystals may directly activate the NLRP3 inflammasome in a receptor-independent manner.8 However, the molecular events underlying the activation of the NLRP3 inflammasome induced by cell surface contact with crystals remain unclear.

Growing evidence shows that ion channels are critical signalling mediators in both innate and adaptive immune cells9 and their expression and function on MΦs are tightly associated with NLRP3 inflammasome activation.10 Transient receptor potential ion channels are important signalling components in both neurosensory and inflammatory pathways, giving way to their functions as polymodal sensors of environmental cues and as key mediators in sensory transduction.11–15 Among them, the non-selective multifunctional cation channel TRPV4 is highly expressed in cartilage chondrocytes and activated by hypo-osmolality, heat and arachidonic acid metabolites.16 TRPV4 has also been functionally linked to the skeletal dysplasia.17 18 Although studies reported that TRPV4 is expressed by primary nociceptors and contributes to acute nociception caused by hypotonic stress and neuroinflammation,19–21 emerging evidence also showed that TRPV4 expressed by non-neuronal cells including MΦs is involved in the development of acute/chronic itch and regulation of gastrointestinal motility.22 23 However, it remains unknown whether and how TRPV4 channel is involved in crystal-induced inflammation in MΦs.

In this study, we show that MΦ-expressed TRPV4 is selectively involved in the inflammation and reflexive pain-related responses induced by both endogenous and exogenous crystals through activation of the NLRP3 inflammasome. Our results suggest that TRPV4 is a potential drug target for treating crystal-induced inflammatory disorders.

Materials and methods

Materials and methods are described in online supplemental file 1.

Results

TRPV4 is required for monosodium urate (MSU) crystal-induced reflexive pain-related responses and inflammation in a mouse model of gout

To determine the role of TRPV4 in inflammatory joint disease, we first employed the well-established mouse model of gout.24 As expected, wild-type (wt) mice treated with intra-articular (IA) injections of MSU crystals (0.8 mg per site) developed severe joint swelling and reflexive mechanical and thermal pain-related responses (figure 1A–C). Strikingly, intraperitoneal (i.p.) administration of GSK219, a potent and selective TRPV4 inhibitor,23 25 either 3 days before injections of MSU crystals (figure 1A–E) or 5 hours after MSU crystal injections (online supplemental figure 1a–e) significantly reduced MSU crystal-induced reflexive pain-related responses as well as inflammation-related parameters including ankle oedema, leucocyte infiltration and myeloperoxidase (MPO) activity. The inhibitory effect of GSK219 was comparable to that of colchicine, a standard of care treatment for acute gout flares (figure 1A–E). Moreover, genetic ablation of TRPV4 function significantly alleviated MSU crystal-induced reflexive pain-related responses, and joint inflammation (ankle oedema, synovium lesion/inflammation, neutrophil infiltration and MPO activity) when compared with Trpv4+/+ control mice (figure 1F–N), suggest that TRPV4 function is critically required for the generation of MSU crystal-induced gouty arthritis.

Figure 1

Pharmacological inhibition or genetic ablation of TRPV4 function reduces inflammation and reflexive pain-related responses in a mouse model of gout. (A–C) IA injections of MSU crystals (0.8 mg/site) in wt C57BL/6J mice produced reflexive mechanical (A) and heat (B) pain-related responses as well as joint swelling (C) in a time-dependent manner. Mice were treated with either GSK219 (5 mg/kg, i.p.) or vehicle 3 days before injections of MSU crystals. Colchicine (1 mg/kg, i.p.) was used as a positive control. n=6–8 for each group. (D, E) Increased leucocyte infiltration (D) and myeloperoxidase activity (E) were detected 6 hour after IA injections of MSU crystals in the presence of GSK219 (5 mg/kg, i.p.) or vehicle. Colchicine (1 mg/kg, i.p.) was used as a positive control. n=6 for the vehicle control and 7–8 for all other groups. (F–H) Time courses for reflexive mechanical (F) and heat (G) pain-related responses as well as joint swelling (H) induced by IA injections of MSU crystals in the Trpv4+/+ and Trpv4-/- mice, n=7–8 for each group. (I, J) Both total leucocyte infiltration (I) and myeloperoxidase activity (J) were measured 6 hours after IA injections of MSU crystals, n=7–8 for each group. (K) Representative HE-staining showing oedema and inflammation of ankles in both Trpv4+/+ and Trpv4-/- mice 6 hours after IA injections of MSU crystals. Magnifications, ×5 (middle) and ×20 (right). Asterisks indicate the infiltrating leukocytes. n=3 independent repeats with similar results for both groups. (L) Representative FACS plots of neutrophils in the articular cavity from the Trpv4+/+ and Trpv4−/− mice subjected to IA injections of MSU crystals. (M, N) Summarised data on the right show the percentile of neutrophils within the CD45+ population (M) and the comparison of the neutrophil counts (N) between the Trpv4+/+ and Trpv4−/− mice subjected to IA injections of MSU crystals. n=6 for all groups. Statistical significance was determined using Tukey post hoc tests (multiple comparisons, (A–C), one-way ANOVA with Tukey’s post-test (D–E), two-way ANOVA followed by Bonferroni’s post hoc test (F–H), and Student’s t test (I, J, M, N). When compare to Control + Vehicle group, ##P<0.001, ###P<0.001. When compare to other groups, *P<0.05, **p<0.01, ***p<0.001. ANOVA, analysis of variance; FACS, fluorescence activated cell sorting; IA, intra-articular; i.p, intraperitoneal; MSU, monosodium urate; wt, wild-type.

TRPV4 expression and function are increased in the setting of acute gout flares

Although TRPV4 is reported to be expressed by the primary nociceptors,26 27 surprisingly, we did not detect TRPV4 expression and function in the dorsal root ganglion (DRG) cell bodies and nerve terminals in the skin using the Trpv4eGFP mice, RNAscope in situ hybridisation, and live-cell Ca2+ imaging on ex vivo DRG explants (online supplemental figure 2a–i). On the other hand, TRPV4-eGFP+ cells were found in the synovial membrane and overlapped with well-known synovial MΦ markers such as F4/80, Cx3cr1 and CD68 (online supplemental figure 3a,b). To determine the functionality of TRPV4 in the synovial MΦs, we used live-cell Ca2+ imaging on sort-purified CD45+ synovial MΦs using CD11b/MHCII (figure 2A–C). Application of a potent and selective TRPV4 activator GSK101 elicited a robust [Ca2+]i response in the synovial MΦs isolated from the Trpv4+/+ mice,28 29 which was nearly abolished by GSK219 (figure 2A). GSK101-induced [Ca2+]i response was absent in synovial MΦs from either the Trpv4-/- mice or the Cre+ MΦ-specific Trpv4 cKOs (Cx3cr1CreERT;Trpv4f/f ) in which TRPV4 expression is selectively ablated in MΦs after tamoxifen induction22 (figure 2B,C). Moreover, GSK101 activated TRPV4-like whole-cell currents in the eGFP+ synovial MΦs isolated from the Trpv4eGFP mice, which was inhibited by GSK219 (figure 2D). Flow cytometry further revealed that most of the TRPV4-eGFP+/CD45+ cells were CD11b+/F4/80+ (figure 2E). Strikingly, both TRPV4 mRNA transcripts (figure 2F) and the number of TRPV4-eGFP+ cells were markedly increased in the joints 6 hours after IA injections of MSU crystals (0.8 mg/site) (figure 2G and online supplemental figure 3c). Consistent with the data from mouse studies, TRPV4 expression was significantly increased in PBMCs, the most relevant cell type of resident MΦs, from patients with acute gout flares when compared with that from healthy subjects (online supplemental figure 3d). In marked contrast, the expression of TRPV1, TRPA1, or TRPM8 was not significantly different between these two groups (online supplemental figure 3d). Moreover, PBMCs from gout patients displayed a stronger [Ca2+]i response upon application of TRPV4 agonist GSK101 (100 nM) when compared with that from healthy control subjects (online supplemental figure 3e,f).

Figure 2

TRPV4 is functionally expressed by the synovial MΦs and TRPV4 expression and the number of TRPV4+ MΦs are increased by MSU crystals. (A) Representative traces showing GSK101 (0.3 µM)-elicited [Ca2+]i response in freshly dissociated synovial MΦ single-cell suspensions from the Trpv4+/+ mice. Preapplied and coapplied GSK219 (1 µM) abolished the GSK101 action. Summarised data on the right show the reduction of the percentage of GSK101-responsive synovial MΦs by GSK219. (B) Representative traces showing that GSK101 induced a [Ca2+]i response in the synovial MΦ single-cell suspensions from the Trpv4+/+ but not Trpv4-/- mice. Summarised data on the right show that genetic ablation of TRPV4 function reduces the percentage of GSK101-responsive synovial MΦs. (C) Representative traces showing the GSK101-induced [Ca2+]i response in the synovial MΦ single-cell suspensions from the Cre- but not Cre+ Cx3cr1CreERT; Trpv4f/f mice. Summarised data on the right show reduced the percentage of GSK101-responsive synovial MΦs isolated from the MΦ-specific TRPV4 cKOs. (D) Left: Representative time course of membrane currents evoked by GSK101 (0.3 µM) at +100 mV and −100 mV membrane potentials with and without coapplied GSK219 (1 µM). Horizontal bars denote the time courses for applications of GSK101 and GSK219. Middle: Representative current-voltage curves taken at time points i, ii, iii and iv (colour coded) from the time course on the left. A ramp protocol elicited by a voltage ramp from −100 mV to +100 mV was used. Quantification of the effect of GSK219 on GSK101-activated whole-cell membrane current recorded at +100 mV is shown on the right. (E) Flow cytometry shows that the number of TRPV4-eGFP+ cells increased significantly 6 hours after IA injections of MSU crystals (0.8 mg/site). (F) TRPV4 mRNA expression in the synovial lining of the Trpv4eGFP mice treated with vehicle or MSU crystals. (G) Quantification of the number of the F4/80+/eGFP+ synovial MΦs in response to treatment with MSU crystals or vehicle. Statistical significance was determined using Student’s t-test (A–H), *p<0.05, **p<0.01, ***p<0.001. n=5–6 per group. IA, intra-articular; MSU, monosodium urate.

MSU crystal-induced inflammation and reflexive pain-related responses are attenuated in MΦ-specific Trpv4 cKOs

To test whether MΦ-expressed TRPV4 is required for MSU crystal-induced gouty arthritis, we used the MΦ-specific Trpv4 cKOs (Cx3cr1CreERT; Trpv4f/f ).22 Indeed, the Cre+ mice displayed significantly reduced reflexive pain-related responses, joint inflammation and synovium lesion when compared with the Cre- littermates in response to IA injections of MSU crystals (figure 3).

Figure 3

MΦ-specific Trpv4 cKOs display reduced joint swelling and reflexive pain-related responses in MSU crystal-induced gouty arthritis. (A–C) Time courses for reflexive mechanical (A) and heat (B) pain-related responses as well as joint swelling (C) induced by IA injections of MSU crystals (0.8 mg/site) in both Cre+ and Cre- Cx3cr1CreERT; Trpv4f/f mice (Experiments were performed 1 week after the last tamoxifen administration). n=6–8 for each group. (D, E) Changes of total leucocyte infiltration (D) and myeloperoxidase activity (E) in both Cre+ and Cre- Cx3cr1CreERT; Trpv4f/f mice 6 hours after IA injections of MSU crystals (0.8 mg/site), n=6–8 for each group. (F) Representative HE-staining showing oedema and inflammation of ankle sections 6 hour after IA injections of MSU crystals (0.8 mg/site) in the Cre+ and Cre- Cx3cr1CreERT; Trpv4f/f mice. Magnifications, ×5 (middle) and ×20 (right). Asterisks indicate the infiltrating leukocytes. n=3 independent experiments with similar results for both groups. (G) Representative FACS plots of synovial neutrophils from both Cre and Cre+ Cx3cr1CreERT; Trpv4f/f mice subjected to IA injections of MSU crystals. (H, I) Summarised data show the comparison of the percentile of neutrophils within the CD45+ population (H) and the neutrophil count (I) between the Cre and Cre+ Cx3cr1CreERT; Trpv4f/f mice subjected to IA injections of MSU crystals. n=5 for each group. Statistical significance was determined using two-way ANOVA, followed by Bonferroni’s post hoc test (A–C), and Student’s t-test (D, E, H, I), *p<0.05, **p<0.01, ***p<0.001. ANOVA, analysis of variance; IA, intra-articular; MSU, monosodium urate.

Since the synovial MΦs can arise from either tissue-resident or circulating monocyte-derived MΦs, we also tested the inducible MΦ-specific Trpv4 cKOs 4 weeks after tamoxifen administration when tissue-resident MΦs retain their identity but the monocyte-derived MΦs are replaced with wt MΦs produced by bone-marrow-derived progenitors as shown by fate mapping studies.30 Both reflexive pain-related responses and ankle swelling were still significantly reduced in the Cre+ Cx3cr1CreERT; Trpv4f/f mice when compared with their Cre- littermates (online supplemental figure 4). These results recapitulate the reduced reflexive pain-related responses and inflammation phenotypes displayed by the Trpv4-/- mice and further demonstrate that TRPV4 expression in the tissue-resident MΦs is the primary contributor to MSU crystal-induced gouty arthritis.

TRPV4 is also functionally expressed by neutrophils31 and articular chondrocytes that are involved in age-related and post-traumatic osteoarthritis (OA).32 Indeed, in addition to synovial MΦs, the number of eGFP+ neutrophils was also markedly increased by MSU crystals as revealed by flow cytometry and immunofluorescence (online supplemental figure 5a–c). Notably, we and others have used neutrophil infiltration as an inflammation marker for MSU crystal-induced gouty inflammation.33 To test if TRPV4 expression in these cell types also contributes to gouty inflammation and reflexive pain-related responses, we generated the neutrophil-specific (S100a8Cre; Trpv4f/f ) and cartilage-specific (Col2a1CreERT+; Trpv4f/f ) Trpv4 cKOs. Neither of these Cre+ cKOs displayed changes in ankle swelling and reflexive pain-related responses after IA injections of MSU crystals compared with their respective Cre- groups (online supplemental figure 5d–i). Consistent with the absence of TRPV4 expression and function in primary nociceptors, genetic ablation of TRPV4 function in the TRPV1 lineage neurons (Trpv1Cre; Trpv4f/f ) had no effect on MSU crystal-induced gouty arthritis (online supplemental figure 5j–l). Together, these results suggest that neutrophil-expressed, articular chondrocyte-expressed or sensory nociceptor-expressed TRPV4 is dispensable for the development of MSU crystal-induced gouty arthritis.

TRPV4 function is critically involved in MSU crystal-induced NLRP3 inflammasome activation and interleukin-1β production

Activation of NLRP3 inflammasome is a hallmark of MΦ-mediated inflammatory responses, which requires at least two signalling events.7 34 The first event involves nuclear factor-κB (NF-κB)-dependent upregulation of NLRP3 along with pro-interleukin (IL)-1β, which is triggered by PAMPs including Toll-like receptors, DAMPs or cytokines. Of note, NF-κB signalling also promotes the production and release of tumour necrosis factor-α (TNF-α), which can be independent of NLRP3 inflammasome activation. The second event involves the assembly of a complex of multiple proteins including NLRP3, ASC (the adaptor molecule apoptosis-associated speck-like protein), and pro-caspase-1, resulting in the activation of caspase-1. Subsequently, active caspase-1 processes pro-IL-1β to mature IL-1β, which is then released from dying MΦs.

Strikingly, application of the TRPV4 activator GSK101 was sufficient to induce the production of IL-1β in a concentration-dependent manner in the Trpv4+/+ bone marrow derived macrophages (BMDMs) primed with lipopolysaccharide (LPS) but not in LPS-unprimed group, which could be blunted by either pharmacological inhibition or genetic ablation of TRPV4 function (figure 4A,B and online supplemental figure 6a), although GSK101 did not affect pro-IL-1β and pro-caspase-1 in the lysates (Input) of LPS-primed BMDMs isolated from the Trpv4+/+ and Trpv4−/− mice (online supplemental figure 6a). Similar to GSK101, application of MSU crystals also markedly increased the IL-1β production from LPS-primed wt BMDMs (figure 4C and online supplemental figure 6a) while the levels of both pro-IL-1β and pro-caspase-1 were comparable in lysates (Input) of LPS-primed and MSU crystal-treated BMDMs isolated from either Trpv4+/+ or Trpv4−/ mice (online supplemental figure 6a). GSK219 suppressed MSU crystal-induced IL-1β production from LPS-primed BMDMs in a concentration-dependent manner (figure 4C). MSU crystal-induced IL-1β production was also abolished in the LPS-primed BMDMs isolated from either the Trpv4−/− mice or Cre+ MΦ-specific Trpv4 cKOs when compared with their respective control groups (figure 4D,E and online supplemental figure 6a). Moreover, LPS-primed BMDMs from mice deficient in NLRP3 or caspase-1 were unable to release cleaved IL-1β in response to either MSU crystals (figure 4F) or GSK101 (figure 4G). These results suggest that TRPV4 is involved in the second event of NLRP3 inflammasome activation in LPS-primed BMDMs and NLRP3-caspase-1 signalling is required for TRPV4-mediated IL-1β production induced by both MSU crystals and GSK101. Consistent with results from mouse BMDMs, MSU crystals also induced robust inflammasome activation in cultured primary human MΦs (THP-1 cells) and human PBMCs isolated from healthy subjects, which was significantly reduced by GSK219 (figure 4H,I, and online supplemental figure 6b,c), suggesting that TRPV4 is a key mediator of MSU crystal-induced NLRP3 inflammasome activation in innate immune cells from both rodents and humans.

Figure 4

TRPV4 function is critical to MSU crystal-induced NLRP3 inflammasome activation and IL-1β production. (A) GSK101 promoted IL-1β production in LPS-primed but not in LPS-unprimed BMDMs in a concentration-dependent manner. (B) MSU crystal-induced IL-1β production from LPS-primed BMDMs was severely suppressed by TRPV4 antagonism and genetic ablation of TRPV4 function in both global Trpv4 KO mice and MΦ-specific Trpv4 cKOs compared with their respective controls. (C) MSU crystal-induced IL-1β production from LPS-primed BMDMs was inhibited by GSK219 in a concentration-dependent manner. (D) MSU crystal-induced IL-1β production was markedly reduced in LPS-primed BMDMs isolated from the Trpv4−/− mice when compared with the Trpv4+/+ mice. (E) MSU crystal-induced IL-1β production was markedly reduced in LPS-primed BMDMs isolated from the Cre+ Cx3cr1CreERT; Trpv4f/f mice when compared with their Cre- littermates. Note PBS, LPS, and MSU crystals alone did not increase IL-1β production in (D, E). (F) MSU crystal-induced IL-1β production in LPS-primed BMDMs from NLRP3-deficient mice and Casp-1-deficient mice and their respective control mice. (G) ELISA analysis of IL-1β in supernatants from PMA-differentiated THP-1 cells treated with various concentrations of GSK219 and then stimulated with MSU crystals (200 µg/mL). (H) GSK101-induced IL-1β production in LPS-primed BMDMs from NLRP3-deficient mice and Casp-1-deficient mice and their respective control mice. (I) ELISA analysis of IL-1β in supernatants from human PBMCs pretreated with various concentrations of GSK219 and then stimulated with MSU crystals (200 µg/mL). Statistical significance was determined using two-way ANOVA, followed by Bonferroni’s post hoc test (A), Tukey post hoc tests (multiple comparisons, (C, H, I), and Student’s t-test (B, D–G). ***P<0.001. n=6 per group. ANOVA, analysis of variance; IL-1β, interleukin 1β; LPS, lipopolysaccharide; MSU, monosodium urate; PBS, phosphate buffered saline; PMA, phorbol-12-myristate-13-acetate; THP, human myeloid leukemia mononuclear cells.

Interestingly, neither GSK101 nor GSK219 showed any significant effects on the release of TNF-α from the LPS-primed wt BMDMs (online supplemental figure 7a,b). Moreover, TNF-α production was not affected in the LPS-primed BMDMs isolated from either the Trpv4-/- mice or the Cre+ MΦ-specific Trpv4 cKOs when compared with their respective controls (online supplemental figure 7c,d), suggesting that TRPV4 is unlikely involved in the first event of NLRP3 inflammasome activation and TRPV4 function is not required for TNF-α production in our MSU crystal-induced gout model.

Distinct mechanisms are involved in TRPV4-mediated NLRP3 inflammasome activation induced by GSK101 and MSU Crystals

Although the precise mechanism of MSU crystal-induced NLRP3 activation is not completely understood, the involvement of several key events including lysosomal rupture, activation of reactive oxygen species (ROS), and intracellular Ca2+ signaling has been reported.35 As TRPV4 regulates numerous cellular functions through intracellular Ca2+ signalling, we first tested BAPTA-AM, a cell-permeant Ca2+ chelator, on GSK101- and MSU crystal-induced IL-1β production from the LPS-primed wt BMDMs. Surprisingly, BAPTA-AM effectively blocked GSK101-induced IL-1β release while it only suppressed about half of MSU crystal-induced IL-1β release (figure 5A,B). In contrast, cytochalasin D, an actin polymerisation inhibitor that blocks >90% of phagocytosis, nearly abolished MSU crystal-induced IL-1β production without a significant effect on the GSK101-induced IL-1β production (figure 5A,B). These results suggest that the effect of GSK101-induced IL-1β production is largely dependent on TRPV4-mediated intracellular Ca2+ signalling while MSU crystals can induce IL-1β production through phagocytosis without completely relying on TRPV4-dependent intracellular Ca2+ signalling.

Figure 5

GSK101 and MSU crystals use distinct mechanisms to drive TRPV4-dependent NLRP3 inflammasome activation. (A, B) GSK101- or MSU crystal-induced IL-1β production from LPS-primed BMDMs in the presence of various chemical inhibitors: Bapta-AM (a cell permeant Ca2+ chelator), cytochalasin D (Cyt D, a potent phagocytosis inhibitor), CA-074Me (a selective inhibitor of CTSB), or ROS scavengers GSH-EE (ethyl ester of glutathione) and NAC (N-acetylcysteine). PBS is the vehicle control for all chemicals. (C) Effect of GSK219, Cyt D on phagocytosis of pHrodo Red Escherichia coli BioParticles in LPS-primed BMDMs using a fluorescent microplate reader. (D) Phagocytic activity of BMDMs from the Trpv4+/+ and Trpv4-/- mice was measured using a fluorescence plate reader. (E) GSK101-induced and MSU crystal-induced ROS production in LPS-primed BMDMs from the Trpv4+/+ and Trpv4-/- mice. (F) GSK101- and MSU crystal-induced increase in CTSB levels in LPS-primed BMDMs isolated from the Trpv4+/+ and Trpv4-/- mice. (G) Effect of GSK219 on CTSB production induced by GSK101 or MSU crystals in LPS-primed wt BMDMs. (H) Schematic diagram of the working hypothesis. Statistical significance was determined using Student’s t-test (A–G). When compare to Control group, ##P<0.001. When compare to other groups, *P<0.05, **p<0.01, ***p<0.001. n=4–5 for each group. IL-1β, interleukin; MSU, monosodium urate; ROS, reactive oxygen species; wt, wild-type.

Previous studies also showed that crystals can use receptor-independent membrane-based immune sensing to initiate the activation of NLRP3 inflammasome, which is blunted by a non-selective cation channel blocker ruthenium red, also a TRPV4 channel blocker.36 37 We thus speculated that MΦ-expressed TRPV4 might be involved in the cell surface contact with various crystals. Indeed, both pharmacological blockade and genetic ablation of TRPV4 function significantly reduced MΦ phagocytosis (figure 5C,D and online supplemental figure 8a,b). Phagocytosis of crystal structures is known to produce large amounts of ROS.38 Interesting, application of MSU crystals but not GSK101 promoted ROS production in the LPS-primed BMDMs in a TRPV4-dependent manner (figure 5E). Consistent with these results, IL-1β production induced by MSU crystals but not GSK101 was markedly reduced by two ROS scavengers: ethyl ester of glutathione (GSH-EE) and N-acetylcysteine (figure 5A,B).

Increased phagocytosis triggered by crystal structures also leads to lysosomal damage resulting in the cellular release of lysosomal contents such as cathepsin B (CTSB) that can be sensed by NLRP3 inflammasome.39 40 Both MSU crystals and GSK101 significantly increased CTSB release from LPS-primed BMDMs, which was suppressed by either genetic or pharmacological inhibition of TRPV4 function (figure 5F,G). Further, IL-1β production induced by both GSK101 and MSU crystals was blunted by CA-074Me, a selective inhibitor of CTSB (figure 5A,B). Collectively, these findings suggest that TRPV4 is critically involved in the phagocytosis of MSU crystals by MΦs, which leads to the production of ROS and lysosomal leakage and ultimately NLRP3 inflammasome activation. On the other hand, the GSK101-induced NLRP3 inflammasome activation is accomplished by lysosomal leakage following TRPV4-dependent intracellular Ca2+ signalling (figure 5H).

TRPV4 is involved in NLRP3 inflammasome activation produced by crystalline but not non-crystalline inflammasome activators

Besides MSU crystals, other medically relevant crystals, such as calcium pyrophosphate dihydrate (CPPD) which causes pseudogout, SiO2, and alum adjuvant, as well as classic non-crystalline NLRP3 inflammasome activators such as ATP and pore-forming toxins nigericin and gramicidin also activate NLRP3 inflammasome and increase IL-1β production. We, thus, tested if TRPV4 is involved in the signalling pathways shared by these activators. Surprisingly, GSK219 pretreatment or genetic ablation of TRPV4 function from LPS-primed MΦs markedly reduced IL-1β production induced by CPPD, alum, or silica crystals (figure 6a), whereas IL-1β production induced by ATP, nigericin or gramicidin was unaffected (online supplemental figure 9a–c), suggesting that TRPV4 function is essential for the activation of NLRP3 inflammasome by commonly used crystalline NLRP3 activators but not the noncrystalline activators ATP and pore-forming toxins in vitro.

Figure 6

TRPV4 mediates both in vitro and in vivo inflammatory responses produced by crystalline but not non-crystalline NLRP3 inflammasome activators. (A) Pharmacological inhibition (GSK219, 1 µM) and genetic ablation of TRPV4 function using both global and MΦ-specific cKOs severely reduced IL-1β release from LPS-primed BMDMs stimulated with CPPD (100 ng/mL), alum (200 ng/mL) or SiO2 (100 ng/mL). (B) Orotracheal instilled silica crystals increased neutrophil infiltration in the bronchoalveolar lavage fluid in the Trpv4+/+ and Cre- Cx3cr1CreERT; Trpv4f/f mice, which was significantly reduced in the Trpv4-/- and Cre+ Cx3cr1CreERT; Trpv4f/f mice, respectively. (C–E) Intraperitoneal injections of MSU crystals (C), CPPD (D), or alum (E) crystals markedly promoted neutrophil infiltration in the Trpv4+/+ and Cre- Cx3cr1CreERT; Trpv4f/f mice, which was significantly reduced in the Trpv4-/- and Cre+ Cx3cr1CreERT; Trpv4f/f mice, respectively. Statistical significance was determined using Student’s t-test (A–E). *P<0.05, **p<0.01, ***p<0.001. n=5–6 for each group. IL-1β, interleukin 1 β; MSU, monosodium urate.

To further investigate the role of TRPV4 in crystal-induced inflammation in vivo, we transorally instilled silica crystals into the Trpv4-/- mice or Cre+ Cx3cr1CreERT; Trpv4f/f Trpv4 cKOs and their respective controls. Flow cytometry detected a robust neutrophil infiltration in bronchoalveolar lavage fluids from both the Trpv4+/+ mice and Cre- Cx3cr1CreERT; Trpv4f/f Trpv4 cKOs, which was significantly reduced in the Trpv4-/- mice and Cre+ Cx3cr1CreERT; Trpv4f/f Trpv4 cKOs, respectively (figure 6B). Similar effect was also observed for MSU, CPPD and aluminium hydroxide crystals in a mouse model of peritonitis in which the recruitment of neutrophils induced by i.p. injections of diverse crystals was markedly reduced by genetic ablation of TRPV4 function (figure 6C–E), confirming that TRPV4 is required for crystal-induced inflammation in vivo.

Neuroimmune interaction between the TRPV1-expressing sensory nociceptors and TRPV4-expressing MΦs contributes to MSU crystal-induced gouty arthritis

Prior studies have demonstrated that cytokines released from activated immune cells contribute to MSU crystal-induced pain and inflammation through sensitising TRPV1-expressing nociceptive sensory neurons, suggesting that the TRPV1-expressing primary nociceptors are the downstream mediator of MSU crystal-induced sensory hypersensitivity.41 42 Moreover, besides pain-related responses, MSU crystal-induced inflammation was also inhibited in mice deficient in nociceptor-expressed TRPA1 or TRPV1 channels,41–47 which prompted us to hypothesise that activation of the TRPV1-expressing primary nociceptors during MSU crystal-induced acute gout flares can further enhance joint inflammation by promoting the function of the TRPV4-expressing MΦs, forming a positive feedback loop. To test this hypothesis, we selectively ablated the TRPV1-expressing primary nociceptors with a super potent TRPV1 agonist resiniferatoxin (RTX).48 As expected, RTX treatment effectively reduced the reflexive mechanical (online supplemental figure 10a) and thermal (online supplemental figure 10b) pain-related responses. Strikingly, joint swelling (online supplemental figure 10c), IL-1β expression (online supplemental figure 10d), the number of neutrophils and MΦs (online supplemental figure 10e,f), the MPO activity (online supplemental figure 10g), and the TRPV4 mRNA expression in synovial resident MΦs (online supplemental figure 10h) in the MSU crystal-induced gout model were all markedly attenuated, supporting our hypothesis that the MSU crystal-induced MΦ-dependent inflammation requires the presence of the TRPV1-expressing primary nociceptors and a neuroimmune interaction between the TRPV1-expressing primary nociceptors and the TRPV4-expressing synovial MΦs is essential for the generation of MSU crystal-induced acute gout flares.

Discussion

In this study, we showed that MΦ-expressed TRPV4 is selectively involved in the inflammatory responses induced by diverse crystals. The conclusion was supported by multiple experimental measures. First, we showed that TRPV4 is functionally expressed by MΦs and TRPV4 expression is upregulated in both MSU crystal-stimulated mouse synovial MΦs and PBMCs from human patients with acute gout flares. Second, we provided evidence that TRPV4 function is required for the activation of the NLRP3 inflammasome and subsequent IL-1β production induced by diverse crystals including MSU, CPPD and SiO2 but not non-crystalline NLRP3 inflammasome activators ATP, nigericin or gramicidin. Third, the reflexive pain-related responses and inflammation induced by IA injections of MSU crystals were markedly attenuated by both pharmacological inhibition and genetic ablation of TRPV4 function. Last, the TRPV1-expressing nociceptors are required for MSU crystal-induced gouty arthritis through increasing TRPV4 expression, the number of synovial resident MΦs and the release of MΦ-derived cytokines. which supports the hypothesis that a neuroimmune axis of the TRPV4-expressing resident MΦs-TRPV1-expressing primary nociceptors plays a critical role in the generation of crystal-induced gouty arthritis in mice. Of note, our findings might explain why the crystals are in patients for years without causing overt gout flares. It is well known that acute gout flares are tightly associated with the levels of uric acid in patients. For example, the accumulation of high level of uric acid caused by a diet rich in red meat, seafood and beverages sweetened with fructose49 might activate TRPV4-expressing synovial resident MΦs, resulting in release of inflammatory cytokines which subsequently provoke action potential firing in nociceptors, thereby transmitting pain to the brain and driving the neurogenic inflammation to release neurotransmitters such as substance P and calcitonin gene related peptide (CGRP) in the affected joints and promote joint inflammation through regulating macrophage functions,50 51 forming a positive feedback loop and driving the production of acute gout flares.

Our findings also challenge the long-held assumption that TRPV4 is expressed by primary nociceptors mediating neurogenic inflammation and nociception52–56 because we did not find any evidence supporting a functional expression of TRPV4 in the primary nociceptors. First, we did not detect TRPV4 expression in DRG neurons using RNAscope in situ hybridisation which otherwise revealed a robust expression of TRPV1 in many DRG neurons. Second, GSK101 did not elicit TRPV4-dependent Ca2+ response in DRG neurons where capsaicin induced a robust Ca2+ response. Third, the MSU crystal-induced gouty arthritis was not significantly affected in sensory neuron-specific TRPV4 cKOs using the Trpv1Cre line which covers ~80% of the nociceptors.57 These results demonstrate that TRPV4 unlikely engages in crystal-induced joint pain and inflammation through a direct neurogenic action, like that mediated by nociceptor-expressed TRPV1 and TRPA1 channels.43 58 It should be noted that nerve injury and inflammation can induce macrophage expansion in the DRG,59 60 leading to heightened pain responses. TRPV4 expression in these MΦs but not sensory neurons might contribute to the generation of pain-related responses. Future studies are required to test this possibility.

TRPV4 is functionally expressed in articular chondrocytes in multiple species,61–63 and promotes distinct mechanoelectrical transduction pathways to regulate the metabolic response of chondrocytes to dynamic loading.64 65 Interestingly, although chondrocyte-expressed TRPV4 was reported to contribute to age-related OA,32 knockout of TRPV4 channels was shown to promote the development of OA in male mice.66 Moreover, IA injections of a TRPV4 agonist stimulated chondrocyte anabolic changes and decreased the length of the damaged area in a surgically induced rat model of OA, suggesting that TRPV4 activation protects rats from the development of OA.67 Surprisingly, our results using cartilage-specific Trpv4 cKOs showed that the chondrocyte-expressed TRPV4 does not contribute to MSU crystal-induced gouty arthritis. Therefore, chondrocyte-expressed TRPV4 plays distinct roles in the settings of different types of joint inflammation.

The classic NLRP3 inflammasome activators including ATP, nigericin, gramicidin and diverse crystals (MSU, CPPD, alum and SiO2) use distinct mechanisms to activate the NLRP3 inflammasome.68 For instance, ATP acts on the P2×7 receptors while nigericin and gramicidin are pore forming toxins, although membrane permeability is engaged in both pathways.69–71 On the other hand, crystal signalling involves phagolysosome and lysosome rupture.7 Our data showed that TRPV4 is involved in the NLRP3 inflammasome activation induced by diverse crystals but not ATP and pore-forming toxins. Therefore, TRPV4 likely engages in the crystal-induced signalling in MΦs which is separate from these classics signalling pathways. Although the precise mechanism how TRPV4 channel is activated in the gout model remains unclear, our findings suggest that mechanosensitive TRPV4 channel might be critically involved in the cell surface contact by various crystals and be directly activated by swelling-related cell volume change (online supplemental figure 11) associated with the phagocytosis of MSU crystals, which subsequently causes lysosomal leakage and drives ROS production, leading to NLRP3 inflammasome activation in the joint MΦs.39 72 73 On the other hand, TRPV4-mediated NRLP3 inflammasome activation induced by GSK101 relies primarily on intracellular Ca2+ signaling and subsequent lysosomal leakage although the relationship between the TRPV4-mediated intracellular Ca2+ signaling and lysosomal leakage requires further investigation. Considering TRPV4 is activated by many forms of physical and chemical stimuli,74 75 it might play a critical role in inflammation through converging many signalling pathways driving NLRP3 inflammasome activation.

In summary, we demonstrated that the TRPV4-expressing resident MΦs are the key mediator of MSU crystal-induced gouty arthritis in mice. Mechanistically, the mechanosensitive TRPV4 channel can be selectively activated by crystal-induced cell volume change, which leads to NLRP3 inflammasome activation and subsequent production and release of inflammatory cytokines that drive the joint inflammation and pain-related responses. We also showed that this process requires the presence of the TRPV1-expressing nociceptors and demonstrated that a neuroimmune interaction between the TRPV1-expressing primary nociceptors and the TRPV4-expressing synovial MΦs is a critical cellular mechanism underlying acute gout flares. Identification of the function of TRPV4 in crystal-induced inflammation should facilitate the development of new therapeutic interventions to treat inflammatory conditions associated with crystal disposition.

Ethics statements

Patient consent for publication

References

Supplementary materials

Footnotes

  • ZL and LC are joint first authors.

  • Handling editor Josef S Smolen

  • ZL and LC contributed equally.

  • Contributors HH, LC and ZL conceived the experiments. LC and ZL performed the experiments and analyzed the data. JF, ZX, JL and XH. assisted with Ca2+ imaging and whole-cell patch-clamp recording. FW and BK assisted with flow cytometry. JL and PY assisted with RAN scope. YZ, XY, LD, ZZ, LC and PL assisted with behavioural assays. ZL assisted with collection of human blood samples. LD, PL, PY and ZL assisted with Western blot and in situ hybridisation. BK, RS and KL assisted with manuscript preparation. LC, JF and HH supervised the project and wrote the manuscript. All authors discussed the manuscript, commented on the project and contributed to manuscript preparation.

  • Funding This work was supported by grants from the National Institutes of Health R01DK103901, R01AR077183, and R01AA027065 to HH.

  • Competing interests BK has served as a consultant for AbbVie, ABRAX Japan, Almirall, Cara Therapeutics, Maruho, Menlo Therapeutics, Pfizer, and Third Rock Ventures. BK is also founder, chief scientific officer, and stockholder of Nuogen Pharma and stockholder of Locus Biosciences.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.