Cell therapy centered on IL-1Ra is neuroprotective in experimental stroke
- 2.3k Downloads
Cell-based therapies are emerging as new promising treatments in stroke. However, their functional mechanism and therapeutic potential during early infarct maturation has so far received little attention. Here, we asked if cell-based delivery of the interleukin-1 receptor antagonist (IL-1Ra), a known neuroprotectant in stroke, can promote neuroprotection, by modulating the detrimental inflammatory response in the tissue at risk. We show by the use of IL-1Ra-overexpressing and IL-1Ra-deficient mice that IL-1Ra is neuroprotective in stroke. Characterization of the cellular and spatiotemporal production of IL-1Ra and IL-1α/β identifies microglia, not infiltrating leukocytes, as the major sources of IL-1Ra after experimental stroke, and shows IL-1Ra and IL-1β to be produced by segregated subsets of microglia with a small proportion of these cells co-expressing IL-1α. Reconstitution of whole body irradiated mice with IL-1Ra-producing bone marrow cells is associated with neuroprotection and recruitment of IL-1Ra-producing leukocytes after stroke. Neuroprotection is also achieved by therapeutic injection of IL-1Ra-producing bone marrow cells 30 min after stroke onset, additionally improving the functional outcome in two different stroke models. The IL-1Ra-producing bone marrow cells increase the number of IL-1Ra-producing microglia, reduce the availability of IL-1β, and modulate mitogen-activated protein kinase (MAPK) signaling in the ischemic cortex. The importance of these results is underlined by demonstration of IL-1Ra-producing cells in the human cortex early after ischemic stroke. Taken together, our results attribute distinct neuroprotective or neurotoxic functions to segregated subsets of microglia and suggest that treatment strategies increasing the production of IL-1Ra by infiltrating leukocytes or microglia may also be neuroprotective if applied early after stroke onset in patients.
KeywordsIL-1α/β Microglia Macrophages Inflammation Focal cerebral ischemia Human brain
Novel cell-based therapies have emerged during the past decade, with several ongoing trials investigating the therapeutic effect of bone marrow (BM) cells [48, 57], including mesenchymal stem cells , being a cellular source of IL-1Ra , which has anti-inflammatory effects  and neuroprotective functions in experimental stroke [43, 46, 52]. An advantage of using cell-based therapies in stroke, next to or in combination with recanalization therapies , is that cells, which actively infiltrate the neural parenchyma at the site of injury [56, 59], may be able to modulate the inflammatory response in tissues at risk of being recruited into the expanding infarct. The inflammatory response elicited by stroke is orchestrated as a continuum, driven by both activated microglia and infiltrating leukocytes, both known producers of interleukin-1β (IL-1β) [8, 10, 15, 19]. IL-1Ra is a competitive inhibitor of IL-1(α/β) signaling [6, 20, 39, 60, 61] when mediated through the IL-1 receptor type 1 (IL-1R1) found in low numbers on nearly all cells including cortical neurons [13, 36]. The IL-1R2 acts as a decoy receptor , and the IL-1R3, which is also expressed on cortical neurons, lacks an IL-1Ra binding site . Although, it is well documented that IL-1Ra attenuates the pro-inflammatory activities of IL-1(α/β) [6, 20, 39, 60, 61], key information about the cellular production of IL-1Ra and IL-1(α/β), and its orchestration in the tissue at risk in ischemic stroke is still lacking, but pivotal to the development of new therapeutic strategies.
Here, we established the cellular sources of IL-1Ra and IL-1(α/β) in a mouse model of stroke, and examined the neuroprotective and anti-inflammatory effect of therapeutically injected IL-1Ra-producing BM cells. Microglia, not infiltrating leukocytes, were the major sources of IL-1Ra with segregated subsets of microglia producing either IL-1Ra or IL-1β, demonstrating functional heterogeneity of microglia at the site of injury. The use of a radiation BM chimeric approach documented that IL-1Ra-producing leukocytes can be neuroprotective. Therapeutic injection of IL-1Ra-producing BM cells post-stroke amplified microglial production of IL-1Ra and reduced brain levels of IL-1β, collectively leading to smaller infarcts and improved functional outcome. In summary, we found the therapeutic injection of cells with increased IL-1Ra production to be neuroprotective, by mechanisms that involve amplification of the endogenous IL-1Ra production. The relevance of these results for future stroke treatment is emphasized by the demonstration that IL-1Ra is also expressed in the infarcted cortex in autopsies from stroke victims.
Materials and methods
In this study, we used adult male IL-1Ra KO mice that lacked all isoforms of IL-1Ra and heterozygous IL-1Ra-Tg mice carrying a transgene encoding the secreted form of IL-1Ra (sIL-1Ra) under the control of its endogenous promoter (Table S1). Both mouse strains were kindly donated by Dr. Emmet Hirsh (Columbia University, USA) and have previously been described in detail [32, 36]. Wild-type littermate (LM) mice were used as controls. IL-1Ra-KO mice lacked IL-1Ra mRNA in the brain, and IL-1Ra-Tg mice expressed higher levels of IL-Ra in the brain than LM mice [IL-1Ra: 252 ± 170 pg/mg (Tg), and 8 ± 3 pg/mg (LM), (mean ± SD), n = 14/group]. Both mouse strains were maintained as colonies at the Laboratory of Biomedicine, University of Southern Denmark, Odense, Denmark. In addition, young adult, male and female homozygous C57BL/6-Tg(UBC-GFP)30Scha/J (Stock No. 004353) (GFP-TG) breeding pairs were purchased from the Jackson Laboratory (Bar Harbour, Maine, USA) and transferred to the Department of Biomedicine, University of Aarhus, where they were kept as a colony. IL-1α/β-KO mice  were kindly donated by Dr. Yoichiro Iwakura and used for control tissue. All C57BL/6 mice were purchased from Taconic Ltd. (Denmark). The mice were housed under diurnal lighting conditions and given free access to food and water.
Human postmortem brain tissue
This study was performed using postmortem brain tissue from 21 stroke cases. Sex, age, infarcted brain area and the age of the infarct are given in Table S2. Four cases are part of a previous study on surfactant protein-D in human ischemic brain tissue . Control tissues included brain, kidney, pancreas, lung, placenta and tonsil (not shown). The study was approved by the Regional Committee on Health Research Ethics from the Region of Southern Denmark (Project-ID S-20080042) and performed at the Department of Pathology, Odense University Hospital, Denmark.
All studies were conducted using young age-matched male mice. Focal cerebral ischemia was induced by either permanent (p) or transient (t) middle cerebral artery occlusion (MCAo). pMCAo permanent occlusion of the distal part of the left MCA was performed as detailed in Clausen et al. . tMCAo transient occlusion (40 min) of the right MCA was performed using the intraluminal filament placement technique, as detailed in Inacio et al. , under 1.5 % isoflurane anesthesia. All studies were conducted as randomized, double-blinded studies, with inclusion of sham-operated mice and unlesioned controls (Ctl). Blood flow and body temperature were monitored using the optical fiber probe (VP10 M200ST, UK) and endoscopic temperature probe (T6a, UK) from Moor Instruments. The study was approved by the Danish Veterinary and Food Administration (J. no. 2011/561-1950). An overview of the different mouse groups used for assessment of the effect of IL-1Ra and IL-1Ra-producing cells on infarct volume is provided in Table S1.
Irradiation BM chimeric mice
Irradiation BM chimeric mice were generated as detailed in Clausen et al. . All recipient mice were irradiated with a single dose of 9.5 Gy from a 137Cs source and BM transplantation into the tail vein was performed within 2 h of whole body irradiation. Recipient mice were allowed to reconstitute for 6 weeks prior to pMCAo. The extent of chimerism was assessed in blood samples taken from C57BL/6 mice reconstituted with BM cells from GFP mice at the time of termination (n = 5). Flow cytometry showed that 91.8 ± 0.9 % (mean ± SD) of the circulating leukocytes were CD11b+CD45high GFP+ leukocytes, indicating a successful reconstitution of the mice.
BM cells for post-stroke treatment
Harvesting and transplantation BM cells were harvested in RPMI medium (Gibco, Life Technologies), and approximately 107 cells were injected into the tail vein of recipient mice 30 min after MCAo . Cell labelling A Vybrant® CFDA SE Cell (CFSE) Tracer Kit (V12883, Invitrogen) was used for cell labeling and visualization in situ or by flow cytometry. Characterization BM cells harvested from IL-1Ra-Tg and LM mice (n = 5/group) were analyzed with regard to cell numbers and cell size using the Scepter 2.0 automated Cell Counter from Millipore. To assure compatibility with the Scepter correlation range (50 × 105–1.5 × 106 cells/ml), 10 µl of BM cell suspension was first counted using a Bürker-Türk counting chamber. Next, 100 µl of a BM cell suspension from IL-1Ra-Tg and LM mice were counted using the Scepter automated cell counter employing 40-μm Scepter sensors (Millipore). Histograms and overlays were computed utilizing the Scepter™ 2.0 Cell Counter and Software Pro system.
Blood gas analysis Venous blood samples  collected 30 min after BM injection (n = 6–8/group). were analyzed for pO2/pCO2, pH, electrolytes, glucose and lactate. Temperature Rectal temperature was measured on all mice before surgery (0 h), before BM injection (30 min) and 1 and 3 h after pMCAo using a temperature probe with a Model Bat 12 unit (Physiotemp, New Jersey, USA). Body weight All mice were weighed at 0 h and 1, 3 and 5 days after pMCAo.
Grip strength The grip strength test was performed as previously detailed [4, 40] with the principal investigator blinded to the treatment. The highest force score of the front paws was recorded pre- and post-surgically, using the grip strength meter from BIOSEP (BIO-GT-3, France). Hargreaves test Thermal nociception was tested as detailed [5, 49] using the plantar test device (Ugo Basile, Model 37370, Italy). The investigator was blinded to the treatment. Five measurements were registered for the hind paws pre- and post-surgically, the lowest and highest value was excluded and the mean calculated. The mice were allowed to recover for 15 min between trials.
Open field test Locomotor and anxiety-related activities were investigated using the open field test [Box: 45 (W) × 45 (D) × 40 (H) cm] with the investigator blinded to the treatment. This test was conducted as previously detailed [41, 45] over a 10 min period. Movement was recorded automatically, while time to first rear, digging, grooming, jumping, urination and droppings were recorded manually and documented as the number of events.
Fresh (CO 2 ) frozen tissue Brains were processed into six parallel series of sections (30 μm). Separate series were collected on glass slides and used for infarct size analysis, in situ hybridization (ISH) and immunohistochemistry (IHC) or in Eppendorf tubes and used for qPCR, Western blotting, and ELISA. qPCR The tissue was processed as detailed by Clausen et al. . Flow cytometry The tissue was processed as detailed by Clausen et al. . Paraformaldehyde fixed tissue Brains were fixed and processed into 12 series of sections (16 μm) for histological analysis was performed as detailed by Clausen et al. .
Infarct size analysis was performed according to Cavalieri´s principle for infarct volume estimation (IFV) as previously described on one series of sections spanning the infarct [4, 27] with blinded assessment of the outcome.
RNA extraction, cDNA synthesis and qPCR were performed with published primers for IL-1β mRNA and HPRT1 mRNA as described by Clausen et al. . In addition, the following primer sequences were used, all spanning exon–exon junctions with sequence specificity checked using BLAST: IL-1Ra mRNA (sense: AAC CAG CTC ATT GCT GGG TAC TTA; antisense: GCC CAA GAA CAC ACT ATG AAG GTC), IL-1α mRNA (Sense: GTC GGC AAA GAA ATC AAG ATG; antisense GTC TTC GTT TTC ACT GTA ACA G), IL-1R1 mRNA (sense: ACA ACG TGA GCT TCT TCG GAG TA; antisense: GGT CTG TCC CTC TTG TTT TCA TCT ATT), and IL-1R2 mRNA (sense: AGG AAT ACA ACA TCA CTA GGA ATA TTG AAC; antisense: TCA GTC TTG ACC CCA ATG ATG CT). Cytokine mRNA levels were normalized to a pool of non-lesioned brains. Quantitative PCR investigating temporal changes were conducted using C57BL/6 mice with 1, 2, 4, 6, 12 and 24 h survival, including shams and non-lesioned controls (n = 10–12/group).
In situ hybridization
Alkaline phosphatase (AP)-conjugated oligodeoxynucleotide probe sequences for IL-1β mRNA and GAPDH mRNA were used, as described by Clausen et al. [11, 12]. In addition, the following probes were synthesised: IL-1Ra mRNA (5′TGC CCC CGT GGA TGC CCA AGA ACA C) and IL-1α mRNA (probe 1: 5′GTG CTG ATC TGG GTT GGA TGG TCT CTT C; probe 2: 5′GTT TCT GGC AAC TCC TTC AGC AAC ACG G). Probe specificity was validated as detailed by Clausen et al. , with the inclusion of sections from IL-1Ra-KO mice (Figure S1). Temporal profiles for IL-1Ra mRNA, IL-1α mRNA and IL-1β mRNA were obtained from one series of sections from C57BL/6 mice with 30 min, and 1, 4, 6, 12 and 24 h survival, including sections from non-lesioned controls (n = 6/group). IL-1Ra mRNA was also analyzed in one series of sections from mice with 5 days of survival (n = 10/group).
Flow cytometry was performed using the FACSCalibur flow cytometer and CellQuest Pro Software (BD Bioscience) [10, 40]. IL-1α+ and IL-1β+ microglia [CD11b+CD45dim] and leukocytes [CD11+CD45high] were identified as detailed by Clausen et al. . Samples analyzed for IL-1Ra protein were fixed and permeabilized in Cytofix/Cytoperm for 20 min at 4 °C. Cells were washed in 1 ml 1 × PermWash™ buffer (BD Bioscience), blocked using 10 % fetal calf serum (FCS) with 50 μg/ml Syrian Hamster IgG (11.3 mg/ml) (Jackson Immunoresearch) and incubated with anti-IL-1Ra prior to surface marker labeling of microglia and macrophages in mice with 6, 12 and 24 h survival in addition to non-lesioned controls (Ctl) (n = 4–8/group) . BM cells were quantified based on the CFSE signal, and cell numbers were estimated in mice with 6 h survival as described previously [3, 64]. All antibodies used are listed in Table S3.
Mouse brain tissue CD41, IL-1α and IL-1RII were visualized using the streptavidin/horseradish peroxidase (HRP) technique, IL-1β, using the peroxidase labeled “Ready to use” EnVision+ polymer (Dakocytomation, Denmark) [10, 12] and IL-1Ra and IL-1RI, using the Vectastain ABC kit (PK4005 and PK6104; Vector Laboratories) on series of sections from mice with 4, 6, 12 and 24 h survival in addition to non-lesioned controls (n = 6/group). IHC double-fluorescence staining This was performed as detailed in Clausen et al.  on 16 sections spanning the anterior commissure including areas of both frontal and parietal cortices (n = 5–6 mice/staining). The specificity of staining for IL-1α, IL-1β and IL-1Ra was verified by the absence of staining in IL-1α/β-KO and IL-1Ra-KO brains, and the specificity of all other antibodies was verified by the substitution of the primary antibody with isotype or serum controls (Table S3), which were devoid of signal (data not shown). Human brain tissue Formalin-fixed paraffin-embedded  hematoxylin and eosin (HE)-stained tissue samples were evaluated by independent pathologists before serial sections were stained for IL-1Ra, rIgG2a, Iba1 CD45, CD68 and GFAP (Table S3). Slides were stained on the AutostainerPlus platform (Dako, Denmark) as previously described by Hermansen et al. . The PowerVision+Poly_HRP IHC detection System (PV6107, Leica Biosystems) was used for the detection of all antibodies . Sections stained with IgG or sections where primary antibodies were omitted were devoid of staining. IHC double staining was performed using the BenchMark fully automated staining instrument and the Vectastain Universal Elite ABC kit (Vector Labs, USA) . Sections were counterstained with Mayer’s hematoxylin (Bie & Berntsen, Herlev, Denmark). All antibodies used are listed in Table S3.
Scoring of IL-1Ra staining intensity
The infarcts were categorized according to age as follows at 1–2 days (n = 8) and 5 to ≥7 days (n = 13). The infarct core, the peri-infarct and normal-appearing tissue were identified in hematoxylin-stained sections. The regions were transferred onto parallel sections stained for IL-1Ra and scored as follows: Score 0: no staining; Score 1: very weak staining; Score 2: weak staining; Score 3: moderate staining; Score 4: strong staining; Score 5: very strong staining. The scoring was performed by two observers blinded to the age of the infarcts.
IL-1Ra (IL-1Ra Kit: MRA00, R&D) and IL-1β (IL-1β Kit: CMC0813, Invitrogen) ELISA were performed on one series of sections and conducted as specified in the ELISA kits. Measurements were normalized to the total protein content of the sample as measured using the method by Bradford .
Western blotting was performed on one series of sections as described in Clausen et al.  with minor modifications. Western blotting for phosphorylated (p)-SAPK/JNK (Thr183/Tyr185) (1:1000), p-p44/p42 MAPK (ERK1/2)(1:2000) and p-p38 MAPK (Tyr180/Tyr182) (1:1000) (S9910 Kit, Cell Signaling) was performed using 4–12 % gels (Life Technologies) using transcription factor IIB (TFIIB) (Cell Signaling, 1:1000) as a loading control and SeeBlue Plus2 standard (Invitrogen) as an indicator of size. Bands were amplified using a SuperSignal®Extended duration substrate (34075, Thermo) and visualized by chemiluminescence using the FUSION 7X camera. Densitometry was performed using Image J (version 1.47v) (NIH) following recommendations of the Image J developers. Analysis was performed on two independent gels with n = 2/group and data were normalized to TFIIB and represented as percentages relative to naive LM mice.
Quantitative data are presented as mean ± SD. Comparison between two groups was performed using Mann–Whitney or paired t test. Multiple comparisons were performed using one-way ANOVA and repeated-measures ANOVA, followed by appropriate post hoc tests. Correlations were established using the non-parametric Spearman test. Statistical analysis was performed using the Prism 5 software for Windows (GraphPad). Statistical significance was established for P < 0.05.
Effect of IL-1Ra on infarct size
Cellular location of IL-1Ra and IL-1α/β mRNA and protein
The level of IL-1α mRNA was significantly up-regulated at 6, 12 and 24 h (Fig. 2h), with IL-1α mRNA and protein peaking in microglial-like cells located in the peri-infarct 12 and 24 h after pMCAo (Fig. 2i–m). IL-1β mRNA was increased at 1, 12 and 24 h (Fig. 2n) and showed larger increases than IL-1α mRNA (Fig. 2h). IL-1β mRNA+ cells were detected from 30 min after pMCAo and onward, increasing in number until 24 h, at which time IL-1β mRNA+ cells also occurred perivascularly (Fig. 2o–q). IL-1β was expressed by microglial-like cells in the peri-infarct at 4 and 24 h and by leukocyte-like cells 24 h after pMCAo (Fig. 2r, s). When performing Spearman correlation analysis, we observed a positive correlation between IL-1Ra and IL-1β mRNA at 1, 2, 6, 12 and 24 h after pMCAo and between IL-1Ra and IL-1α mRNA 24 h after pMCAo (Table S4). No correlation was observed between IL-1α and IL-1β mRNA (Table S4). IHC double-fluorescence staining showed the expression of IL-1Ra, IL-1α and IL-1β in CD11b+ cells (Fig. 2t–v), but not in glial fibrillary acidic protein-expressing astrocytes (data not shown), suggesting microglia and/or leukocytes as the major source of these cytokines. In addition, we found IL-1α in CD41+ platelets at 4–6 h, but not 24 h, after pMCAo (Fig. 2w, x), which adds to in vitro findings of IL-1α in platelets .
Since IL-1(α/β) and IL-1Ra are known to impact on infarct development, we additionally analyzed the mRNA expression of IL-1R1 and -1R2 by qPCR. We found small fluctuations in IL-R1 mRNA until 6 h and small increases at 12 and 24 h after pMCAo compared to non-lesioned and sham controls (Figure S2a, c). IL-1R1 immunoreactivity was associated with vascular cells at 6 h and glial-like cells at 24 h after pMCAo (Figure S2b). In comparison, IL-1R2 mRNA was significantly increased at 6, 12 and 24 h compared to non-lesioned and sham controls (Figure S2c), significantly correlating to the ischemia-induced increase in IL-1α mRNA (Table S4) and the occurrence of IL-1R2+ leukocyte-like cells in the peri-infarct (Figure S2d).
IL-1Ra is predominantly expressed by microglia
The number of IL-1β+ microglia exceeded the number of IL-1α+ microglia after pMCAo at all times of observation (Fig. 3b). Surprisingly, few CD11b+CD45high leukocytes were IL-1Ra+ apart from 6 h after pMCAo, when 16 ± 7 % (n = 4) were IL-1Ra+ (Fig. 3c). By 24 h, less than 0.3 ± 0.05 % (n = 4) of the leukocytes were either IL-1Ra+ or IL-1α+, whereas 17 ± 9 % (n = 4) expressed IL-1β (Fig. 3c). A comparison of the number of microglia and leukocytes expressing IL-1Ra, IL-1α and IL-1β showed that microglia was the major source of IL-1Ra and IL-1α, whereas both microglia and leukocytes were a source for IL-1β 24 h after pMCAo (Fig. 3b, c; Table S5). These findings were confirmed by the use of GFP irradiation BM chimeric mice (Figure S3a–c) showing that the majority of the IL-1Ra+ and IL-1α+ cells were GFP− microglia (Figure S3a, b), while IL-1β was expressed by both GFP− microglia and GFP+ leukocytes (Figure S3c). These results clearly show microglia as the major source of neuroprotective IL-1Ra after pMCAo, and additionally suggest that increasing IL-1Ra production in recruited cells could be of relevance in stroke therapy.
IL-1Ra and IL-1β are expressed by segregated subsets of microglia
Based on our previous findings showing that IL-1β and TNF are expressed by largely different subsets of microglia , we assessed the proportion of CD11b+CD45dim microglia co-expressing IL-1Ra, IL-1α and/or IL-1β 24 h after pMCAo. We found IL-Ra and IL-1β to be produced by segregated subsets of CD11b+CD45dim microglia with no co-expression (Fig. 3d, e). Approximately, 5 % of the gated IL-1Ra+ microglia co-expressed IL-1α (5.1 ± 1.9 % (n = 4)), while significantly higher proportions, approximately 50 % of the gated IL-1α+ cells co-expressed IL-1Ra (50.4 ± 6.5 % (n = 4)) (Figure S3d). Further, we found that approximately 8 % of the gated IL-1β+ microglia co-expressed IL-1α (8.0 ± 4.1 % (n = 4)) (Fig. 3e), with higher proportions, and approximately 17 % of the gated IL-1α+ microglia co-expressing IL-1β (17.3 ± 6.1 % (n = 4)) (Figs. 3e, S3d). Double-immunofluorescence staining confirmed the absence of co-expression between IL-1Ra and IL-1β 24 h after pMCAo, showing spatially segregated cells producing either IL-1Ra or IL-1β (Fig. 3f, i). We confirmed the co-expression of IL-1Ra and IL-1α (Fig. 3g, j) and IL-1α and IL-1β; the latter however was only observed in a few number of cells 24 h after pMCAo (Fig. 3h, k). Taken together, these results demonstrate a functional diversity among microglia at the site of injury.
Increasing IL-1Ra production in infiltrating leukocytes can decrease infarct size
To track IL-1Ra synthesizing BM-derived leukocytes in the ischemic brain, we also subjected IL-1Ra-KO mice reconstituted with BM cells from IL-1Ra-Tg mice (Tg–KO*) to pMCAo. ISH of sections from these mice showed multiple IL-1Ra mRNA+ cells in the peri-infarct of irradiated IL-1Ra-KO mice reconstituted with BM cells from IL-1Ra-Tg mice (Tg–KO*), compared to no cells in IL-1Ra-KO mice, 24 h after pMCAo (Fig. 4b, c). By analyzing RNA and protein isolated from parallel sections, we, as expected, observed a robust expression of IL-1Ra mRNA and protein in Tg–KO* mice, and not in IL-1Ra-KO mice, 24 h after pMCAo (Fig. 4d, e). Interestingly, the increase in IL-1Ra synthesis was associated with a lower ischemia-induced increase in IL-1β mRNA levels in Tg–KO* compared to IL-1Ra-KO mice (qPCR, IL-1Ra-KO: 96.6 ± 67.5, n = 12; Tg–KO*: 16.6 ± 11.3, n = 14; P < 0.001; Mann–Whitney test).
In combination, these results provide first evidence that increasing IL-1Ra expression by infiltrating leukocytes, controlled by endogenous kinetics can promote and support neuroprotective signaling pathways in the ischemic cortex.
IL-1Ra-overproducing BM cells infiltrate the ischemic cortex and induce IL-1Ra synthesis in microglia
Prior to the therapeutic injections of BM cells obtained from either IL-1Ra-Tg or LM mice, we labeled them with the fluorescent marker CFSE (carboxyfluorescein diacetate succinimidyl ester), allowing the cellular detection by flow cytometry 6 h after pMCAo (Fig. 5c). We found that approximately 6 % of the recruited CD11b+CD45high cells in both LM–LM and Tg–LM mice were CFSE+ BM cells, and that the numbers of recruited CFSE− CD11b+CD45high leukocytes were comparable 6 h after pMCAo (Fig. 5d). We analyzed the mean fluorescence intensity (MFI) of IL-1Ra in the CFSE+ BM recruits uncovering significantly higher IL-1Ra synthesis in BM cells originating from IL-1Ra-Tg (Tg) compared to LM mice 6 h after pMCAo (Fig. 5e).
Because the major source of IL-1Ra in the ischemic brain is microglia, supplemented by few recruited leukocytes (Fig. 3b, c), we asked whether the CFSE+ BM recruits could impact IL-1Ra synthesis by these cells. Flow cytometry analysis showed that the number of IL-1Ra+CD11b+CD45high leukocytes was comparable in Tg–LM, LM–LM and LM mice 6 h after pMCAo, as well as in their non-lesioned control mice (Fig. 5f, g). In contrast, we observed a significant increase in the number of IL-1Ra+CD11b+CD45dim microglia in both Tg–LM and LM–LM mice compared to non-treated LM mice 6 h after pMCAo (Fig. 5h, i). The finding of significantly higher numbers of IL-1Ra+CD11b+CD45dim microglia in Tg–LM compared to LM–LM mice stressed the source-benefit of the cells, revealing BM from IL-1Ra-Tg mice as potent inducers of IL-1Ra synthesis by microglia after stroke (Fig. 5h, i). These data were supported by ELISA data showing significantly more IL-1Ra protein in Tg–LM mice compared to LM mice 6 h after pMCAo (Tg–LM: 10.7 ± 6.5, LM–LM: 2.7 ± 2.6; LM: 2.7 ± 3.2 pg/mg; n = 5 mice/group). Thus, microglia are highly plastic cells which can be induced to increase their production of neuroprotective IL-1Ra.
Since mitogen-activated protein kinase (MAPK) signaling pathways play an important role in the early inflammatory response after pMCAo (25), the effect of the BM treatment on MAPKs signaling pathways, including extracellular signal-regulated kinase (ERK), Jun N-terminal kinase (JNK) as well as p38, was explored in the tissue from the same mice (Figure S4a–c). The level of phosphorylated (p) JNK was elevated by 60–80 % 6 h after pMCAo in mice treated with IL-1Ra-overproducing BM cells, compared to pMCAo LM mice (Figure S4b). These findings suggest that the increased availability of IL-1Ra in the BM-treated mice modulate the ischemia-induced activation of the JNK pathway.
IL-1Ra-producing BM cells are neuroprotective, reduce brain IL-1β levels and improve functional outcome
We next tested the neuroprotective effect of IL-1Ra-overproducing BM cells, in the model of tMCAo (Table S1), which gives rise to a large lesion comprising both cortical and subcortical areas (Fig. 6c). We observed a significant reduction in infarct size in Tg–LM mice and LM–LM mice compared to LM controls, 24 h after tMCAo (Fig. 6c) (survival rate/group >95 %), thereby supporting the findings in mice subjected to pMCAo, additionally showing an infarct-reducing effect of LM BM cells 24 h after tMCAo.
To address a possible mechanism responsible for the beneficial effect of increased IL-1Ra, we quantified IL-1β mRNA and protein in parallel sections by qPCR and ELISA (Fig. 6d–f). The ischemia-induced increase in IL-1β mRNA was significantly smaller in Tg–LM mice compared to LM–LM mice 24 h after pMCAo (Fig. 6d). Consistent with this observation, we also observed less IL-1β protein in Tg–LM mice compared to LM–LM mice 24 h after pMCAo (Fig. 6d). By day 5, both IL-1β mRNA and protein in LM, LM–LM and Tg–LM were back at baseline levels (Fig. 6e), as measured in non-lesioned B6 mice (see Fig. 2c for mRNA) and non-lesioned LM mice [IL-1β: 137 ± 72 pg/mg (mean ± SD), n = 3]. Similar to the pMCAo mice, we also observed a smaller ischemia-induced increase in IL-1β mRNA and protein in Tg–LM mice compared to LM mice 24 h after tMCAo (Fig. 6f), collectively suggesting an effect of increased IL-1Ra production on IL-1β expression after both pMCAo and tMCAo.
We used a battery of behavioral tests to assess the functional recovery in mice subjected to 5 days pMCAo or mice subjected to tMCAo. As the MCA supplies cortical areas controlling the contralateral front and hind limb, we measured the neuromuscular function of the left and right forelimbs (Fig. 6g–i). We detected a significant contralateral weakness in the right (R) front paw of LM mice 1, 2 and 5 days after pMCAo compared to normal baseline strength (Fig. 6g). A weakness first observed on day 2 and 5 after pMCAo in LM–LM mice was accordingly not observed in Tg–LM mice (Fig. 6g). By day 5, both LM and LM–LM mice showed pronounced grip asymmetry in the left and right forelimbs, whereas the performance of Tg–LM mice was comparable in the two forelimbs (Fig. 6h). A similar grip strength outcome was observed in LM–LM and Tg–LM mice 24 h after tMCAo (Fig. 6i). In addition, performing the Hargreaves and open field test on mice subjected to tMCAo, we observed an improved functional recovery in Tg–LM-treated mice (Table S6). Hargreaves test revealed similar nociception in the left and right hind paw prior to surgery with LM and LM–LM mice showing a significantly lower sensitivity compared to Tg–LM mice 24 h after tMCAo (Table S6). Using the open field test, we observed that the Tg–LM mice exhibited a more explorative activity, revealing that LM–LM and Tg–LM mice traveled longer (cm) and faster (cm/s) than LM mice 24 h after tMCAo. However, only Tg–LM mice showed reduced anxiety-related behavior spending more time in the center of the maze (Table S7).
Unlike LM and LM–LM mice, the Tg–LM pMCAo mice did not succumb to a significant 3 % weight loss (P < 0.01) on day 5, suggesting improved feed efficiency and physical well-being of the mice (weight loss/group after 5 days: LM 3.7 %; LM–LM 3.2 %; Tg–LM 0.2 %) (Figure S5j).
IL-1Ra in human stroke patients
In this study, we also report on the changes in the cellular IL-1(α/β) content, identifying both microglia and platelets as the source of IL-1α and microglia and leukocytes as the source of IL-1β after pMCAo. Our data contradicts findings arguing that IL-1α is the predominant form of IL-1 following stroke , but supports the view that IL-1α plays an important role in stroke pathology, by strengthening in vitro findings of IL-1α in platelets . The presence of IL-1α-carrying platelets likely affects the balance between IL-1/IL-1Ra in the infarcted cortex. By 24 h, we observed a few IL-1(α/β) co-producing microglia, expected to be linked to IL-1α secretion as IL-1β has been identified as a functional shuttle for IL-1α release . In parallel, we observed multiple IL-1α/IL-1Ra co-producing microglia supporting the view that intracellular IL-1Ra could regulate the action of intracellular IL-1α . Overall, these results can explain the shift toward early IL-1-induced neuroinflammation and illustrate the complexity of cellular cross talk following stroke. Imbalance between IL-1 and IL-1Ra has been shown to determine the degree of cardiac remodeling after myocardial infarction , and a high IL-1Ra/IL-1β ratio is associated with a better outcome in patients after traumatic brain injury .
Microglia and not infiltrating leukocytes attenuate the effect of IL-1 in stroke pathology, serving as an endogenous pool of IL-1Ra, as also suggested from our analysis of autopsy material from stroke patients. Our results demonstrate that microglia become effective on demand, by increasing their IL-1Ra production when encountering IL-1Ra-producing BM cells, leading to attenuation of IL-1-induced inflammation and neurodegeneration. However, to substantiate this treatment approach as neuroprotective, studies will have to be performed in female mice, aged mice and in mice with post-stroke end points greater than 5 days. Furthermore, to be of therapeutic value in the treatment of acute stroke, this treatment strategy shall also be proven to be beneficial using prolonged therapeutic time windows and ideally enhance the outcome when and if combined with thrombolysis therapy.
We suggest that BM cells, through the release of IL-1Ra, promote a neuroprotective microglial phenotype in the stroke-injured cortex, a view which is supported by evidence showing that mesenchymal stem cells can shape microglial effector function in vitro and promote a neuroprotective microglial phenotype . We further provide insights into the functional effect of IL-1Ra-producing BM cells by showing their modulation of JNK/SAPK signaling, a key regulatory pathway for neurite outgrowth , neuronal differentiation and plasticity  and cell survival . As mesenchymal stem cells are a recognized source of IL-1Ra , our finding that BM cells modulate microglial IL-1Ra production after stroke in mice may also shed light on an unresolved mechanism underlying the efficacy of mesenchymal stem cell therapy in stroke [29, 37].
In conclusion, our finding that increasing the IL-1Ra expression by infiltrating leukocytes promotes neuroprotection, and that post-stroke injection of IL-1Ra-producing BM cells amplifies endogenous IL-1Ra production, suggests that identifying a way to increase the IL-1Ra expression in leukocytes may be beneficial in future stroke therapy.
The authors acknowledge Dr. Emmet Hirsh for donating IL-1Ra-KO and IL-1Ra-Tg mice and Dr. Yoichiro Iwakura for donating IL-1α/β-KO mice. Acknowledgments go to Jens Leander Johansen (Millipore) for donating Scepter sensors and to Robert Graham Quinton Leslie for his valuable input on flow cytometry and Tadeusz Wieloch for passing on the model of tMCAo in mice. The authors acknowledge the technical assistance provided by Lene Jørgensen, Sussanne Petersen, and Signe Marie Andersen at the Department of Neurobiology, University of Southern Denmark for laboratory assistance and Helle Wohlleben at the Department of Pathology, Odense University Hospital, Denmark. We thank Kerstin Beirup at the Institute for Experimental Brain Research, Wallenberg Neuroscience Center, Lund University, Lund Sweden for her technical assistance with the tMCAo mouse model.
This work was supported by Grants from the Lundbeck Foundation, Hørslev Fonden, Arvid Nielsen Fonden, Brødrene Hartmann Fonden, Legat til Yngre Kvindelige Forskere på Sundhedsvidenskab, Fonden til Lægevidenskabens Fremme, Aase og Ejnar Danielsens Fond, Familien Hede Nielsens Fond, Snedkermester Sophus Jacobsen og Hustru Astrid Jacobsens Fond, Kong Christian d. 10 Fond, Frimodt-Heineke Fonden, Direktør Kurt Bønnelycke og Hustru Grethe Bønnelyckes Fond, Mogens Svarre Mogensens Fond, Carla Cornelia Storch Møllers Legat, Else Poulsens Mindelegat to B.H Clausen; the Danish MRC, Carlsbergfondet, the Lundbeck Foundation, Savværksejer Jeppe Juhl og hustru Ovita Juhls Mindelegat, Aase og Ejnar Danielsens Fond, P.A. Messerschmidt og Hustrus Fond and Overlægerådets Legatudvalg to K.L. Lambertsen; the Swedish Research Council Grant No. 2012-2229, the Gyllenstiernska Krapperup, the Royal Physiographic Society, Crafoord, A. E. Berger, Wiberg, Bergvall, G & J Kock Foundations to T. Deierborg; the Lundbeck Foundation and the Danish MRC to A.A. Babcock; and the Novo Nordic Foundation to B. Finsen.
- 1.Abbate A, Salloum FN, Van Tassell BW, Vecile E, Toldo S, Seropian I, Mezzaroma E, Dobrina A (2011) Alterations in the interleukin-1/interleukin-1 receptor antagonist balance modulate cardiac remodeling following myocardial infarction in the mouse. PLoS One 6:e27923. doi: 10.1371/journal.pone.0027923 CrossRefPubMedPubMedCentralGoogle Scholar
- 4.Bach A, Clausen BH, Moller M, Vestergaard B, Chi CN, Round A, Sorensen PL, Nissen KB, Kastrup JS, Gajhede M et al (2012) A high-affinity, dimeric inhibitor of PSD-95 bivalently interacts with PDZ1-2 and protects against ischemic brain damage. Proc Natl Acad Sci USA 109:3317–3322. doi: 10.1073/pnas.1113761109 CrossRefPubMedPubMedCentralGoogle Scholar
- 9.Clausen B, Degn M, Martin N, Couch Y, Karimi L, Ormhoj M, Mortensen ML, Gredal H, Gardiner C, Sargent II et al (2014) Systemically administered anti-TNF therapy ameliorates functional outcomes after focal cerebral ischemia. J Neuroinflammation 11:203. doi: 10.1186/PREACCEPT-2982253041347736 CrossRefPubMedPubMedCentralGoogle Scholar
- 10.Clausen BH, Lambertsen KL, Babcock AA, Holm TH, Dagnaes-Hansen F, Finsen B (2008) Interleukin-1beta and tumor necrosis factor-alpha are expressed by different subsets of microglia and macrophages after ischemic stroke in mice. J Neuroinflammation 5:46. doi: 10.1186/1742-2094-5-46 CrossRefPubMedPubMedCentralGoogle Scholar
- 12.Clausen BH, Lambertsen KL, Meldgaard M, Finsen B (2005) A quantitative in situ hybridization and polymerase chain reaction study of microglial-macrophage expression of interleukin-1beta mRNA following permanent middle cerebral artery occlusion in mice. Neuroscience 132:879–892. doi: 10.1016/j.neuroscience.2005.01.031 CrossRefPubMedGoogle Scholar
- 22.Emberson J, Lees KR, Lyden P, Blackwell L, Albers G, Bluhmki E, Brott T, Cohen G, Davis S, Donnan G et al (2014) Effect of treatment delay, age, and stroke severity on the effects of intravenous thrombolysis with alteplase for acute ischaemic stroke: a meta-analysis of individual patient data from randomised trials. Lancet 384:1929–1935. doi: 10.1016/S0140-6736(14)60584-5 CrossRefPubMedPubMedCentralGoogle Scholar
- 23.Fettelschoss A, Kistowska M, LeibundGut-Landmann S, Beer HD, Johansen P, Senti G, Contassot E, Bachmann MF, French LE, Oxenius A et al (2011) Inflammasome activation and IL-1beta target IL-1alpha for secretion as opposed to surface expression. Proc Natl Acad Sci USA 108:18055–18060. doi: 10.1073/pnas.1109176108 CrossRefPubMedPubMedCentralGoogle Scholar
- 24.Galea J, Ogungbenro K, Hulme S, Greenhalgh A, Aarons L, Scarth S, Hutchinson P, Grainger S, King A, Hopkins SJ et al (2011) Intravenous anakinra can achieve experimentally effective concentrations in the central nervous system within a therapeutic time window: results of a dose-ranging study. J Cereb Blood Flow Metab 31:439–447. doi: 10.1038/jcbfm.2010.103 CrossRefPubMedPubMedCentralGoogle Scholar
- 28.Gueorguieva I, Clark SR, McMahon CJ, Scarth S, Rothwell NJ, Tyrrell PJ, Hopkins SJ, Rowland M (2008) Pharmacokinetic modelling of interleukin-1 receptor antagonist in plasma and cerebrospinal fluid of patients following subarachnoid haemorrhage. Br J Clin Pharmacol 65:317–325. doi: 10.1111/j.1365-2125.2007.03026.x CrossRefPubMedPubMedCentralGoogle Scholar
- 29.Gutierrez-Fernandez M, Fuentes B, Rodriguez-Frutos B, Ramos-Cejudo J, Vallejo-Cremades MT, Diez-Tejedor E (2012) Trophic factors and cell therapy to stimulate brain repair after ischaemic stroke. J Cell Mol Med 16:2280–2290. doi: 10.1111/j.1582-4934.2012.01575.x CrossRefPubMedPubMedCentralGoogle Scholar
- 33.Horai R, Asano M, Sudo K, Kanuka H, Suzuki M, Nishihara M, Takahashi M, Iwakura Y (1998) Production of mice deficient in genes for interleukin (IL)-1alpha, IL-1beta, IL-1alpha/beta, and IL-1 receptor antagonist shows that IL-1beta is crucial in turpentine-induced fever development and glucocorticoid secretion. J Exp Med 187:1463–1475CrossRefPubMedPubMedCentralGoogle Scholar
- 34.Hutchinson PJ, O’Connell MT, Rothwell NJ, Hopkins SJ, Nortje J, Carpenter KL, Timofeev I, Al-Rawi PG, Menon DK, Pickard JD (2007) Inflammation in human brain injury: intracerebral concentrations of IL-1alpha, IL-1beta, and their endogenous inhibitor IL-1ra. J Neurotrauma 24:1545–1557. doi: 10.1089/neu.2007.0295 CrossRefPubMedGoogle Scholar
- 38.Klementiev B, Li S, Korshunova I, Dmytriyeva O, Pankratova S, Walmod PS, Kjaer LK, Dahllof MS, Lundh M, Christensen DP et al (2014) Anti-inflammatory properties of a novel peptide interleukin 1 receptor antagonist. J Neuroinflammation 11:27. doi: 10.1186/1742-2094-11-27 CrossRefPubMedPubMedCentralGoogle Scholar
- 40.Lambertsen KL, Clausen BH, Babcock AA, Gregersen R, Fenger C, Nielsen HH, Haugaard LS, Wirenfeldt M, Nielsen M, Dagnaes-Hansen F et al (2009) Microglia protect neurons against ischemia by synthesis of tumor necrosis factor. J Neurosci 29:1319–1330. doi: 10.1523/JNEUROSCI.5505-08.2009 CrossRefPubMedGoogle Scholar
- 41.Lambertsen KL, Gramsbergen JB, Sivasaravanaparan M, Ditzel N, Sevelsted-Moller LM, Olivan-Viguera A, Rabjerg M, Wulff H, Kohler R (2012) Genetic KCa3.1-deficiency produces locomotor hyperactivity and alterations in cerebral monoamine levels. PLoS One 7:e47744. doi: 10.1371/journal.pone.0047744 CrossRefPubMedPubMedCentralGoogle Scholar
- 42.Lambertsen KL, Ostergaard K, Clausen BH, Hansen S, Stenvang J, Thorsen SB, Meldgaard M, Kristensen BW, Hansen PB, Sorensen GL et al (2014) No effect of ablation of surfactant protein-D on acute cerebral infarction in mice. J Neuroinflammation 11:123. doi: 10.1186/1742-2094-11-123 CrossRefPubMedPubMedCentralGoogle Scholar
- 44.Luheshi NM, Kovacs KJ, Lopez-Castejon G, Brough D, Denes A (2011) Interleukin-1alpha expression precedes IL-1beta after ischemic brain injury and is localised to areas of focal neuronal loss and penumbral tissues. J Neuroinflammation 8:186. doi: 10.1186/1742-2094-8-186 CrossRefPubMedPubMedCentralGoogle Scholar
- 45.Madsen PM, Clausen BH, Degn M, Thyssen S, Kristensen LK, Svensson M, Ditzel N, Finsen B, Deierborg T, Brambilla Ret al (2015) Genetic ablation of soluble tumor necrosis factor with preservation of membrane tumor necrosis factor is associated with neuroprotection after focal cerebral ischemia. J Cereb Blood Flow Metab. doi: 10.1177/0271678X15610339
- 46.Maysami S, Wong R, Pradillo JM, Denes A, Dhungana H, Malm T, Koistinaho J, Orset C, Rahman M, Rubio Met al (2015) A cross-laboratory preclinical study on the effectiveness of interleukin-1 receptor antagonist in stroke. J Cereb Blood Flow Metab. doi: 10.1177/0271678X15606714
- 48.Moniche F, Montaner J, Gonzalez-Marcos JR, Carmona M, Pinero P, Espigado I, Cayuela A, Escudero I, de la Torre-Laviana FJ, Boada C et al (2014) Intra-arterial bone marrow mononuclear cell transplantation correlates with GM-CSF, PDGF-BB, and MMP-2 serum levels in stroke patients: results from a clinical trial. Cell Transplant 23(Suppl 1):S57–S64. doi: 10.3727/096368914X684934 CrossRefPubMedGoogle Scholar
- 49.Novrup HG, Bracchi-Ricard V, Ellman DG, Ricard J, Jain A, Runko E, Lyck L, Yli-Karjanmaa M, Szymkowski DE, Pearse DD et al (2014) Central but not systemic administration of XPro1595 is therapeutic following moderate spinal cord injury in mice. J Neuroinflammation 11:159. doi: 10.1186/s12974-014-0159-6 CrossRefPubMedPubMedCentralGoogle Scholar
- 50.Ortiz LA, Dutreil M, Fattman C, Pandey AC, Torres G, Go K, Phinney DG (2007) Interleukin 1 receptor antagonist mediates the antiinflammatory and antifibrotic effect of mesenchymal stem cells during lung injury. Proc Natl Acad Sci USA 104:11002–11007. doi: 10.1073/pnas.0704421104 CrossRefPubMedPubMedCentralGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.