Transient Receptor Potential Melastatin Subfamily Member 2 Cation Channel Regulates Detrimental Immune Cell Invasion in Ischemic Stroke
Background and Purpose—Brain injury during stroke results in oxidative stress and the release of factors that include extracellular Ca2+, hydrogen peroxide, adenosine diphosphate ribose, and nicotinic acid adenine dinucleotide phosphate. These alterations of the extracellular milieu change the activity of transient receptor potential melastatin subfamily member 2 (TRPM2), a nonselective cation channel expressed in the central nervous system and the immune system. Our goal was to evaluate the contribution of TRPM2 to the tissue damage after stroke.
Methods—In accordance with current quality guidelines, we independently characterized Trpm2 in a murine ischemic stroke model in 2 different laboratories.
Results—Gene deficiency of Trpm2 resulted in significantly improved neurological outcome and decreased infarct size. Besides an already known moderate neuroprotective effect of Trpm2 deficiency in vitro, ischemic brain invasion by neutrophils and macrophages was particularly reduced in Trpm2-deficient mice. Bone marrow chimeric mice revealed that Trpm2 deficiency in the peripheral immune system is responsible for the protective phenotype. Furthermore, experiments with mixed bone marrow chimeras demonstrated that Trpm2 is essential for the migration of neutrophils and, to a lesser extent, also of macrophages into ischemic hemispheres. Notably, the pharmacological TRPM2 inhibitor, N-(p-amylcinnamoyl)anthranilic acid, was equally protective in the stroke model.
Conclusions—Although a neuroprotective effect of TRPM2 in vitro is well known, we can show for the first time that the detrimental role of TRPM2 in stroke primarily depends on its role in activating peripheral immune cells. Targeting TRPM2 systemically represents a promising therapeutic approach for ischemic stroke.
Ischemic stroke is the second most common cause of death worldwide. Tissue damage is thought to follow a biphasic course. The initial hypoxic damage is determined by immediate neuronal cell death leading to the formation of the infarct core, whereas secondary infarct growth is considered to be a consequence of systemic and local sterile inflammation.1
Ischemia in the central nervous system is characterized by oxidative stress and the release of a manifold of stress mediators, among them adenosine diphosphate ribose (ADPR) that is produced by poly-ADPR polymerase in response to oxidative stress, cyclic ADPR, calcium and nicotinic acid adenine dinucleotide phosphate.2 Because these factors modulate the open-probability of the calcium-permeable transient receptor potential melastatin subfamily member 2 (TRPM2) cation channel, this channel has been implicated in stroke pathophysiology. The highest expression levels of TRPM2 are found throughout the nervous system, such as neurons and microglial cells,3 but it can also be detected in a variety of other tissues including cells of the peripheral immune system,4 such as polymorphonuclear neutrophils and monocytes. Therefore, TRPM2 might be involved in early ischemic neuronal cell death but also in the subsequent detrimental sterile inflammation.
Involvement of TRPM2 in cerebral ischemic injury has recently been investigated,5–8 showing a pathogenetic contribution of TRPM2 to ischemic stroke. However, these analyses focused on the role of TRPM2 in neuronal injury during ischemia. Of note, Trpm2 deficiency in ischemic stroke has been suggested to preferentially protect neurons of male mice because of TRPM2 regulation by androgen signaling.5,7,8
Although in vitro experiments point toward a central role of TRPM2 in neuronal injury,6,9 its relative contribution to ischemic tissue injury in vivo by controlling immune cell activation has not been investigated. Notably, TRPM2 signaling controls specific functions in immune cells including production of cytokines and chemokines, chemotaxis of immune cells, and inflammasome activation.10–13 Activation of microglia through TRPM2 was observed after H2O2 and ADPR stimulation in vitro14 and was associated with nitric oxide synthesis and chemokine production.10
Here we show at 2 different experimental sites that TRPM2 detrimentally contributes to ischemic brain injury after stroke, which primarily depends on its role in activating peripheral immune cells. Although Trpm2-deficient neurons are protected against hypoxic stimuli in vitro, TRPM2 regulates neutrophil and macrophage infiltration in vivo that primarily determines its injurious role in stroke. Using separate and mixed bone marrow chimeric mice, we are able to unequivocally demonstrate that the Trpm2 deficiency on neutrophils and macrophages is key for the outcome after stroke, whereas Trpm2 deficiency in central nervous system–resident cells does not contribute to stroke outcome.
Materials and Methods
Animals and In Vivo Stroke Model
All animal experiments were approved by local animal care committees (Behörde für Lebensmittelsicherheit und Veterinärwesen Hamburg and Regierung von Unterfranken). We conducted the experiments according to the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 83-123, revised 1996) and performed all procedures in accordance with the ARRIVE guidelines (http://www.nc3rs.org/ARRIVE). Male C57BL/6J Trpm2–/– mice12 were kindly provided by Dr Y. Mori, Kyoto University, Japan. We randomized all mice and conducted transient middle cerebral artery occlusion (tMCAO) for 1 hour as previously described15 using the intraluminal filament method (6-0 nylon) in a blinded fashion. The detailed experimental description can be found in the online-only Data Supplement. We dissolved the TRPM2 channel blocker, N-(p-amylcinnamoyl)anthranilic acid (ACA; Sigma, St Louis, MO), in dimethylsulfoxide and then further diluted it in PBS and injected it intraperitoneally 2 hours after tMCAO in wild-type (WT) littermate controls and Trpm2–/– mice at 5 or 25 mg/kg body weight.16 We administered PBS-diluted dimethylsulfoxide by intraperitoneal injections in sham-treated mice. Sample size calculation was performed (stroke size from pilot experiments, significance level 0.05, power 90%) and resulted in 9 animals per group to see a difference of 23% in stroke size.
Analysis of Infarct Size by TCC Staining
We analyzed infarct size by harvested brains and cutting them into 1 mm slices (Braintree Scientific, 1 mm) followed by vital staining using 2% (wt/vol) TTC in phosphate buffer. We determined infarct volumes in a blinded fashion using NIH ImageJ software.
Bone Marrow Chimeras
For generation of mixed bone marrow chimeras, we irradiated 10-week-old mice using a cesium-137 gamma irradiator (BIOBEAM 2000, Leipzig, Germany). After 24 hours we reconstituted them with bone marrow cells. The detailed experimental description can be found in the online-only Data Supplement.
Antibodies and Flow Cytometry
We performed flow cytometry for the analysis of cell types as previously described.15 The detailed experimental description can be found in the online-only Data Supplement.
Cell Sorting and RNA Isolation
We isolated immune cells according to the protocol for flow cytometry and stained with antibodies (all eBioscience) against CD11b (M170), CD11c (N418), CD45 (30-F11), and Ly6G (1A8). We sorted cells using a BD FACSAria and collected cells in Dulbecco modified Eagle medium with 30% fetal calf serum. We isolated total RNA from cells using QIA-Shredder spin columns and the RNeasy Micro Kit (Qiagen, Hilden, Germany) and transcribed complementary DNA using Maxima First Strand cDNA Synthesis Kit for reverse transcription polymerase chain reaction (Fermentas, Waltham, MA).
Quantitative Real-Time Polymerase Chain Reaction
We obtained real-time PCR primers from Applied Biosystems (Carlsbad, CA): Actinb Mm00607939_s1; Cxcl1 Mm00433859_m1; Ccl2 Mm00441242_m1; Csf1 (granulocyte-macrophage colony-stimulating factor) Mm00438334_m1; Cxcl2 Mm00436450_m1; Il6 Mm00446190_m1; Il10 Mm00439616_m1; Nos2 (iNOS) Mm00440502_m1; Il1b Mm00434228_m1; Tnf Mm00443258_m1; Trpm2 Mm00663098_m1 and Mm01177249_g1; TATA box binding protein Mm00446971_m1. We purchased probe mixtures from Fermentas (Waltham, MA). The detailed experimental description can be found in the online-only Data Supplement.
Hippocampal Neuron and Microglia Cell Culture and Brain Slice Preparation
We obtained neuronal cell cultures from mouse embryos (embryonic day 18), microglia cell cultures from mice on postnatal days 1 to 5, and acute coronal brain slices from 6- to 10-week-old Trpm2-deficient mice and their WT littermates as previously described.17 The detailed experimental description can be found in the online-only Data Supplement.
We stained brains after standard immunohistochemistry procedures with antibodies against GFAP (1:200; DAKO, Hamburg, Germany), Iba-1 (1:200; Wako, Osaka, Japan), Ly-6G (1:1000; Biolegend, San Diego, CA), NeuN (1:1000; Chemicon, Billerca, MA), cleaved caspase 3 (1:400; CellSignal, Danvers, MA), CD16/32 (1:100; BD Bioscience, Franklin Lakes, NJ), CD206 (1:100; RnDSystems, Minneapolis, MN). The detailed experimental description can be found in the online-only Data Supplement.
Data are reported as mean±SD. Statistical analyses were performed using the appropriate test indicated in the figure legends. Briefly, Student t test was used to compare infarct volumes, Mann–Whitney U test for the comparison of clinical scores, and 1-way ANOVA for multiple comparisons with Bonferroni post hoc test, after validating the normal distribution of these data sets (Kolmogorov–Smirnov test). P values <0.05 were considered statistically significant.
TRPM2 Deficiency Is Protective in Stroke
We assessed infarct size and neurological scores after 1 hour of transient tMCAO. At 2 independent experimental sites (Würzburg and Hamburg), Trpm2–/– mice showed a significantly reduced infarct size and milder disability scores compared with littermates at day 1 (P<0.0001) and day 3 (P=0.01) after tMCAO (Figure 1A and 1C). All Trpm2–/– and littermate control mice included in the tMCAO experiment for infarct volume survived until the analysis of the infarct volume. Notably, both genotypes showed the same reduction in regional cerebral blood flow in the tMCAO model assessed by laser Doppler and were not different in physiological parameters before, during, and after tMCAO (Figure 1B; Figure I in the online-only Data Supplement).
Trpm2 Expression in Stroke
We analyzed TRPM2 expression after 24 hours of cerebral ischemia in the tMCAO model and observed a significant upregulation of Trpm2 in whole-brain mRNA (Figure 2A). By immunocytofluorescence we detected Trpm2 expression in MAP-2–positive neurons and CD11b-positive microglia cells in vitro (Figure 2B). In cells of the peripheral immune system, we found Trpm2 to be highly expressed in bone marrow–derived neutrophils (Figure 2C) and macrophages (Figure 2D), 2 major contributors of poststroke inflammation.18 Next, we analyzed whether Trpm2 is regulated in infiltrating macrophages and resident microglia after stroke in vivo or in neuronal in vitro cultures after challenge with oxidative stress or proinflammatory cytokines. Neither macrophages/microglia exhibited a change in Trpm2 mRNA levels 24 hours after tMCAO compared with cells from sham-operated mice (Figure IIA in the online-only Data Supplement), nor did neurons show alterations in Trpm2 mRNA levels in vitro (Figure IIC and IID in the online-only Data Supplement).
Neuronal Protection in Trpm2–/– Mice After Ischemia
To address whether TRPM2 may be directly involved in neuronal injury, we analyzed markers of neuronal apoptosis by immunohistochemistry. Indeed, we found a significant decrease (P=0.01) of apoptotic (cleaved caspase-3-positive) neurons in ischemic hemispheres of Trpm2–/– mice compared with WT mice at day 1 after ischemic stroke (Figure 3A). To delineate whether the neuroprotective effect in the Trpm2–/– mice is because of a direct increase in resistance toward energy shortage, we subjected cortical slice cultures to oxygen glucose deprivation. After 6 hours of ischemia in vitro, we detected significantly decreased levels of apoptotic neurons (P=0.008) in brain slices of Trpm2–/– mice compared with littermate controls (Figure 3B).
Trpm2 Deficiency Attenuates Immune Cell Infiltration After Experimental Stroke
Because TRPM2 was shown to fulfill important functions in immune cells, the observed reduction in neuronal injury after tMCAO could also be the result of less secondary injury as a consequence of altered poststroke inflammation. Indeed, 3 days after tMCAO, we observed a significant decrease in neutrophil (P<0.0001) and macrophage (P=0.005) infiltration in Trpm2–/– mice compared with littermate control mice in ischemic hemispheres (Figure 4A), whereas infiltration of dendritic cells and CD45+ lymphocytes (Figure 4A) as well as T-cell subsets (NK cells; CD4+, CD8+, and γδ T cells; Figure IIIA in the online-only Data Supplement) did not differ between genotypes. Notably, the observed differences were not because of an alteration of the cell composition in the peripheral blood (Figure IIIB in the online-only Data Supplement). We detected no difference between both genotypes in microglial proliferation, morphological activation, and expression of M1/M2 lineage markers (CD16/32 and CD206) in microglia and macrophages 3 days after tMCAO (Figure 4A–4C; Figures IIIC, IIID, and IVD in the online-only Data Supplement). In flow cytometry analysis, differentiation of infiltrating macrophages and microglia was performed by CD45 expression as previously reported.18 We next analyzed whether Trpm2 deficiency impacts cytokine and chemokine secretion in the ischemic brain. After tMCAO, neither cytokine profiles as measured by intracellular cytokine staining of interleukin-17A (IL-17A) and interferon-γ (lymphocytes; Figure IVA and IVB in the online-only Data Supplement) and tumor necrosis factor-α (macrophages, neutrophils, dendritic cells, and microglia; Figure IVC in the online-only Data Supplement) nor chemokine and cytokine levels in whole-brain mRNA (CXCL-1, CXCL-2, G-CSF, IL-6, CCL-2, IL-10, iNOS, TNF-α, IL-1β) showed significant differences between Trpm2–/– and control animals (Figure 4D).
Trpm2 Regulates Neutrophil and Macrophage Migration Into Ischemic Brains
To further interrogate whether Trpm2 deletion exerts its beneficial effects in stroke by regulating immune cell invasion into ischemic brains, we made use of bone marrow chimeric mice. After lethal irradiation we reconstituted WT mice with Trpm2–/– bone marrow and Trpm2–/– mice with WT bone marrow (Figure 5A). The radiation protocol did not change Trpm2 expression in the cortex (Figure IIB in the online-only Data Supplement). Notably, WT animals reconstituted with Trpm2–/– bone marrow showed significantly smaller lesions in the tMCAO model compared with Trpm2–/– animals reconstituted with bone marrow from WT controls (P=0.04; 17.3±8.1 versus 25.6±10.2 mm3; Figure 5B). This underscores the important role of TRPM2 in peripheral immune cells for the extent of ischemic tissue damage. However, this did not translate into the rather insensitive measurement of neurological impairment (Figure 5B). To verify that TRPM2 regulates migration of immune cells during stroke, we next created mixed bone marrow chimeric mice. We reconstituted irradiated CD45 congenic C57BL/6J Ly5.1 mice (CD45.1) with equal amounts of bone marrow cells derived from Trpm2–/– CD45.2 and WT CD45.1 mice (Figure 5C). Six weeks after bone marrow transplantation, we analyzed the reconstitution in the peripheral blood by flow cytometry and observed an approximately equal distribution of CD45.1+ and CD45.2+ immune cells in each animal (data not shown). Three days after tMCAO, we detected a significantly reduced infiltration of Trpm2–/– neutrophils in the ischemic hemisphere compared with WT neutrophils (P=0.002; Figure 5D). Consistent with the finding in Trpm2–/– mice, we also found reduced numbers of Trpm2-deficient macrophages (P=0.2) albeit to a lesser extent than neutrophils (Figure 5D). By contrast, dendritic cell and lymphocyte migration was not affected by the deficiency of Trpm2 (Figure 5D).
Pharmacological Inhibition of Trpm2 Is Protective in Stroke
ACA has been reported to effectively inhibit TRPM2-mediated currents at concentrations between 10 and 100 mmol/L.16 Indeed, although 5 mg/kg body weight ACA was not effective, administration of a single-dose 25 mg/kg body weight ACA 2 hours after stroke induction resulted in a significant reduction (P<0.001) in infarct size and improvement in clinical outcome compared with the injection of vehicle control (Figure 6A–6C). Furthermore, ACA treatment significantly suppressed neutrophil infiltration into the ischemic hemisphere (Figure 6D). Whereas ACA (25 mg/kg) ameliorated clinical outcome in the tMCAO model in WT mice, it caused no additional improvement in Trpm2–/– mice in comparison to vehicle-treated Trpm2–/– mice, indicating that ACA probably exerts its protective properties via targeting TRPM2 (Figure 6A–6C).
Our study demonstrates a key role for TRPM2 in cerebral ischemia. We show that TRPM2 in neutrophils and macrophages regulates their migratory capacity to ischemic brain thereby secondarily perpetuating brain injury. Trpm2–/– mice are protected from ischemic stroke and show an improved neurological outcome compared with WT mice. Although TRPM2 contributes to neuronal cell death in in vitro conditions, activation of TRPM2 in neurons and microglia seems to have minor effects on the outcome in our murine model of cerebral ischemia. This is implicated by our in vivo experiments with bone marrow chimeric mice that show that TRPM2 directly contributes to the migration of neutrophils and to a lesser extent of macrophages into the ischemic hemispheres and that TRPM2 in these cell types determines tissue damage. Equally, pharmacological inhibition of TRPM2 is able to suppress neutrophil migration and to ameliorate disability after stroke in a clinically relevant setting and therefore commends as treatment strategy in stroke.
Multiple studies have shown that the activation of the immune system has detrimental consequences in stroke.1 In this context, immunologic signaling cascades, which are involved in neutrophil infiltration and activation, seem to have a major impact on the tissue damage. We and others have shown that disruption of IL-1β or inflammasome signaling as well as neutralization of IL-17 and CXCR2 are protective in stroke.15,19
Because TRPM2 is expressed at high levels in neutrophils and has been shown to be involved in the chemotactic response of neutrophils,20 it represents a particular attractive target. The importance of TRPM2 signaling in brain ischemia is underlined by our finding that Cd38-deficient mice show an attenuated infiltration of immune cells and improved neurological outcome in experimental stroke.21 The ectoenzyme CD38 is expressed by different leukocyte subsets and generates ADPR and cyclic ADPR from its substrate NAD+, which in turn are able to activate TRPM2.20
The detrimental role of TRPM2 in other inflammatory conditions, in which reactive oxygen species acts as a main trigger of inflammation, was reported in a disease model of dextran sulfate sodium–induced colitis.12 The role of TRPM2 as a sensor for reactive oxygen species with subsequent chemokine production in macrophages was also demonstrated in a model of inflammatory and neuropathic pain.10 Notably, in the context of ischemia/reperfusion injury of the brain, chemokine production in the ischemic organ seems to be independent of TRPM2, because we could not detect an altered chemokine production in Trpm2–/– mice. Mixed bone marrow chimeras rather suggested a migratory defect for Trpm2–/– neutrophils and possibly macrophages. In support of our finding it has recently been shown in a model of myocardial infarction that neutrophil adhesion, a prerequisite for migration, depends on TRPM2.22
Regarding a possible pharmacological inhibition of TRPM2, one has to consider not only local anti-inflammatory properties but also exacerbated systemic immune alterations in patients with stroke.23 However, short-term selective inhibition of neutrophil migration should be beneficial without aggravating the stroke-induced immune suppression, thereby commending TRPM2 as an attractive target molecule in stroke treatment.
Based on our experiments with bone marrow chimeras, we can attribute the detrimental effects of TRPM2 ion channels mainly to the peripheral immune system. Nevertheless, as seen in our in vitro oxygen glucose deprivation experiments, a direct contribution of TRPM2 to neuronal cell death has to be considered. TRPM2 is highly expressed in the brain3 and is implicated in calcium-dependent cell death in response to oxidative stress.9 Beside these in vitro studies, several in vivo studies indicate that TRPM2 contributes to ischemic neuronal cell death.5,7,8 Mechanistically, deleterious TRPM2 effects in neurons are not only related to calcium-dependent cell death but also to a shift of the expression ratio of N-methyl-d-aspartate receptor GluN2A/GluN2B subunits6 toward a decreased expression of the prosurvival N-methyl-d-aspartate receptor GluN2A.
Overall, the current study was performed in a network of experimental laboratories (Hamburg, Muenster, Würzburg) and the key findings were reproduced in independent laboratories. Regarding the current discussion on preclinical phase III stroke trials,24 our multicenter approach represents a first step to further validate preclinical animal studies with the goal to overcome the translational roadblock.
In conclusion, the present study revealed the role of TRPM2 in postischemic central nervous system inflammation. TRPM2 activation in peripheral immune cells leads to exacerbation of ischemic brain damage, and TRPM2 might therefore display a promising treatment target of the detrimental postischemic inflammatory response after stroke. The general feasibility of this approach has been demonstrated by our experiments with the TRPM2 blocker ACA, which was able to inhibit central nervous system tissue damage in tMCAO even when applied 2 hours after the onset of stroke. Therefore, further investigations for specific, safe, and well-tolerable TRPM2 inhibitors are warranted.
We thank Ellen Orthey (Hamburg), Daniela Urlaub (Wuerzburg), and Heike Menzel (Wuerzburg) for excellent technical assistance. We thank Y. Mori, Kyoto University, for providing Trpm2 knockout mice.
Sources of Funding
This work was supported by grants from the Werner Otto Stiftung (1/81; to M.A. Friese and T. Magnus), Landesexzellenzinitiative Hamburg (to M. Glatzel, T. Magnus, and C. Gerloff), ERANET/NANOSTROKE (to T. Magnus), the Innovative Medizinische Forschung of the University of Münster (N. Melzer), the Deutsche Forschungsgemeinschaft (SFB 688, TP A13; to C. Kleinschnitz), and the Deutsche Stiftung für Neurologie (C. Kleinschnitz and S.G. Meuth).
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.114.005836/-/DC1.
- Received April 15, 2014.
- Revision received July 29, 2014.
- Accepted August 15, 2014.
- © 2014 American Heart Association, Inc.
- Alim I,
- Teves L,
- Li R,
- Mori Y,
- Tymianski M
- Haraguchi K,
- Kawamoto A,
- Isami K,
- Maeda S,
- Kusano A,
- Asakura K,
- et al
- Sumoza-Toledo A,
- Lange I,
- Cortado H,
- Bhagat H,
- Mori Y,
- Fleig A,
- et al
- Kraft R,
- Grimm C,
- Grosse K,
- Hoffmann A,
- Sauerbruch S,
- Kettenmann H,
- et al
- Gelderblom M,
- Weymar A,
- Bernreuther C,
- Velden J,
- Arunachalam P,
- Steinbach K,
- et al
- Meuth SG,
- Budde T,
- Kanyshkova T,
- Broicher T,
- Munsch T,
- Pape HC
- Gelderblom M,
- Leypoldt F,
- Steinbach K,
- Behrens D,
- Choe CU,
- Siler DA,
- et al
- Partida-Sanchez S,
- Gasser A,
- Fliegert R,
- Siebrands CC,
- Dammermann W,
- Shi G,
- et al
- Hiroi T,
- Wajima T,
- Negoro T,
- Ishii M,
- Nakano Y,
- Kiuchi Y,
- et al
- Dirnagl U,
- Hakim A,
- Macleod M,
- Fisher M,
- Howells D,
- Alan SM,
- et al