- Email: [email protected]
Interleukin 8 (CXCL8)-CXC chemokine receptor 2 (CXCR2) axis contributes to MiR-4437-associated recruitment of granulocytes and natural killer cells in ischemic stroke
Molecular Immunology 101 (2018) 440–449
Contents lists available at ScienceDirect
Molecular Immunology journal homepage: www.elsevier.com/locate/molimm
Interleukin 8 (CXCL8)-CXC chemokine receptor 2 (CXCR2) axis contributes to MiR-4437-associated recruitment of granulocytes and natural killer cells in ischemic stroke
T
Qianyi Hea,b, Xiaojuan Shia,c, Bin Zhoua,d, Junfang Tengb, Chaoqi Zhanga,c, Shasha Liua,d, Jingyao Liana,d, Benyan Luoe, Guoqiang Zhaof, Hong Lub, Yuming Xub, Yajun Lianb, Yanjie Jiab, ⁎ Yi Zhanga,c,d,g, a
Biotherapy Center, the First Affiliated Hospital of Zhengzhou University, Zhengzhou 450052, Henan, China Department of Neurology, the First Affiliated Hospital of Zhengzhou University, Zhengzhou 450052, Henan, China Department of Oncology, the First Affiliated Hospital of Zhengzhou University, Zhengzhou 450052, Henan, China d School of Life Sciences, Zhengzhou University, Zhengzhou 450052, Henan, China e Department of Neurology, First Affiliated Hospital, Medical College of Zhejiang University, Hangzhou 310003, Zhejiang, China f School of Basic Medical Sciences, Zhengzhou University, Zhengzhou 450001, Henan, China g Key Laboratory for Tumor Immunology and Immunotherapy of Henan Province, Zhengzhou 450052, Henan, China b c
A R T I C LE I N FO
A B S T R A C T
Keywords: Ischemic stroke CXCR2-CXCL8 microRNA Granulocytes Natural killer cells Prognosis
Granulocytes and natural killer (NK) cells have been linked to brain injury in ischemic stroke. However, their recruitment from peripheral leucocytes in stroke patients is not well understood. Here, the expression of the interleukin 8 (CXCL8) in plasma, and CXC chemokine receptor 2 (CXCR2) in peripheral leucocytes of patients with ischemic stroke were evaluated. Based on the results, CXCR2 expression positively correlated with granulocytes and NK cells, which were in turn attracted by CXCL8. The results also indicated that CXCR2 was a direct target of microRNA (miR)-4437, a negative regulator of CXCR2, which was downregulated in peripheral leucocytes from patients with ischemic stroke. Furthermore, serum CXCL8 levels were associated with the infarct volume and functional outcomes in patients with ischemic stroke. The results of the receiver operating characteristic curve analysis with an optimal cut-off value of 34 pg/mL indicated serum CXCL8 levels could be a prognostic indicator for ischemic stroke. In conclusion, these data highlighted the involvement of the CXCL8CXCR2 chemotactic axis in the recruitment of granulocytes and NK cells in ischemic stroke. Furthermore, miR4437 was suggested as a novel target for treating ischemic stroke, while the serum CXCL8 level could be a prognostic factor for ischemic stroke.
1. Introduction Accumulating evidence has indicated that the immune system and inflammation play a leading role in ischemic stroke, particularly in brain tissue damage, progression of ischemic lesions, and tissue repair (Burrows et al., 2016; Ma et al., 2015; Picascia et al., 2015). Typically following ischemic stroke there is a rapid infiltration of granulocytes,
particularly neutrophils, which can result in blood-brain barrier disruption, cerebral edema, and brain injury (Jickling et al., 2015). In addition to granulocytes, natural killer (NK) cells, which are key members of the innate immune system, accumulate in the ischemic hemisphere (Gan et al., 2014; Zhang et al., 2014). The NK cells determine the size of the brain infarct (Gan et al., 2014) and are responsible for ischemia-related neuronal death by secreting interferon
Abbreviations: NK, natural killer; PBMCs, peripheral blood mononuclear cells; CXCL8, interleukin-8; CXCR2, CXC chemokine receptor 2; IFN, interferon; miR, microRNA; AHA/ASA, American Heart Association/American Stroke Association; DWI, diffusion-weighted imaging; mRS, modified Rankin Scale; DMEM, Dulbecco’s modified eagle medium; FBS, fetal bovine serum; NC, negative control; RT-PCR, real-time reverse transcription polymerase; ELISA, enzyme-linked immunosorbent assay; PBS, phosphate buffered saline; PerCP, peridinin-chlorophyll-protein; CD, cluster of differentiation; PE-CY7, phycoerythrin-cyanin 7; APC, allophycocyanin; FITC, fluorescein isothiocyanate; FACS, Fluorescence-Activated Cell Sorting; GEO, Gene Expression Omnibus; SD, standard deviation; ROC, receiver operating characteristic; WT, wildtype; AUC, area under the curve; TNF, tumor necrosis factor ⁎ Corresponding author at: 1st Jianshe East Road, the First Affiliated Hospital of Zhengzhou University, Zhengzhou, 450052, Henan, China. E-mail address: [email protected] (Y. Zhang). https://doi.org/10.1016/j.molimm.2018.08.002 Received 13 February 2018; Received in revised form 16 July 2018; Accepted 3 August 2018 0161-5890/ © 2018 Elsevier Ltd. All rights reserved.
Molecular Immunology 101 (2018) 440–449
Q. He et al.
completed a 90-day clinical follow-up. The functional outcome was assessed using the modified Rankin Scale (mRS) with the CXCL8 levels being concealed. The mRS is typically used to evaluate the degree of disability or dependence in the daily activities of patients following stroke. The favorable functional outcome was defined by an mRS score of 0–2 points (0, no symptoms; 1, no significant disability; and 2, slight disability), while the unfavorable outcome was defined by an mRS score of 3–6 points (3, moderate disability; 4, moderately severe disability; 5, severe disability; and 6, deceased). In this study, we analyzed the relationship between ischemic stroke and peripheral leucocytes. According to cellular morphology, leucocytes can be divided into two groups: peripheral blood mononuclear cells (PBMCs) and polymorphonuclear cells. Polymorphonuclear cells were isolated by singlestep centrifugation of whole blood using Polymorphprep (Axis-Shield) as described previously (Thomas et al., 2015). Furthermore, PBMCs were collected using Ficoll-Hypaque density gradient centrifugation (Beijing Chemical Reagent Company, China) as described previously (Liu et al., 2015). Finally, we put these two groups of cells together for further testing.
(IFN)-γ (Gan et al., 2014; Zhang et al., 2014). On the other hand, chemokines are a class of pro-inflammatory cytokines that have the ability to attract and activate leukocytes. In the highly inflammatory sites of acute ischemic stroke, chemokines are mainly generated by microglial cells and infiltrating immune cells, resulting in an exacerbated inflammatory cascade and brain injury (Lee et al., 2015; Remus et al., 2015). The CXC chemokine receptor 2 (CXCR2) is a regulator of neutrophil homeostasis (Veenstra and Ransohoff, 2012). In experimental mice with ischemic stroke, it has been demonstrated that a vast amount of CXCR2+ granulocytes are activated and released from the bone marrow (Denes et al., 2011). Furthermore, CXCR2 blockade mitigated the neurological deficits and reduced the infarct volume (Herz et al., 2015; Sousa et al., 2013). Thus, CXCR2-positive granulocytes play an important role in ischemic brain injury. However, no study to date has evaluated the expression of CXCR2 in patients with stroke. Additionally, in ischemic stroke the expression of interleukin-8 (CXCL8), one of the ligands of CXCR2, is controversial. In particular, several previous studies have indicated no change in CXCL8 levels in patients with stroke (Montaner et al., 2003; Pedersen et al., 2004), while others demonstrated increased CXCL8 levels (Al-Bahrani et al., 2007; Grau et al., 2001a, b). Furthermore, the involvement of the CXCL8-CXCR2 chemotactic axis in the recruitment of immune cells in ischemic stroke is not well understood. This study aimed to investigate the effect and underlying mechanism of the CXCL8-CXCR2 chemotactic axis in ischemic stroke.
2.2. Cell culture and transfection 293 T cells were cultured at 37 °C in a 5% CO2-humidified incubator with Dulbecco’s Modified Eagle Medium (DMEM) high-glucose supplemented with 10% fetal bovine serum (FBS). The microRNA (miR)4437 mimic and negative control (NC) were chemically synthesized (Shanghai GenePharma Company, Shanghai, China) and had the following sequences: miR-4437 mimic, 5ʹ-UGGGCUCAGGGUACAAAG GUU-3ʹ and 5ʹ−CCUUUGUACCCUGAGCCCAUU-3ʹ; NC, 5ʹ-UUCUCCG AACGUGUCACGUTT-3ʹ and 5ʹ-ACGUGACACGUUCGGAGAATT-3ʹ. Cells were seeded onto 24-well plates at a density of 4 × 104 cells/well and co-transfected with oligonucleotides (100 nM, miR-4437 mimics or NC) and plasmids (500 ng, pmirGLO-CXCR2-3ʹUTR-WT or pmirGLOCXCR2-3ʹUTR-MUT). All procedures were performed using the Lipofectamine 3000 reagent (Invitrogen, Carlsbad, USA) following the manufacturer’s instructions. Forty-eight hours after transfection, cells were harvested for the luciferase activity assay.
2. Materials and methods 2.1. Patients A total of 140 patients experiencing the first week of their first-ever acute ischemic stroke were recruited at the First Affiliated Hospital of Zhengzhou University from 25 June 2015 to 20 June 2018. All enrolled patients signed a written informed consent, and all experimental protocols were approved by the Ethics Committees of the First Affiliated Hospital of Zhengzhou University. Patients were excluded if they had infections, autoimmune diseases, cancers, or other conditions that could impact immune system homeostasis. During this study, the diagnosis and treatment of ischemic stroke were carried out under the guidelines of the American Heart Association/American Stroke Association (AHA/ASA) (Jauch et al., 2013). Table 1 summarizes the basic characteristics of the enrolled patients. As controls, a total of 70 age- and sex-matched healthy donors were also recruited. The infarct volume was measured by diffusion-weighted imaging (DWI) according to the validated ABC/2 method (Sims et al., 2009). Most of the patients
2.3. Plasmid construction and luciferase activity assay The 60-bp sequence of the CXCR2 3ʹUTR, containing the predicted miR-4437 binding sites, and its mutant, were synthesized as primers by Sangon Biotech (China, Shanghai). Subsequently, sense and antisense oligonucleotides were mixed with 5X annealing buffer (Beyotime, China). The mixture was boiled for 10 min and then cooled slowly to ambient temperature. The double-stranded oligonucleotides were digested with SacI and Xhol. The resulting oligonucleotides were subcloned into the SacI and Xhol sites of the pmirGLO vector. Luciferase activity was measured using the Dual-Luciferase Reporter Assay System (Promega) on a Cetro XS3 LB 960 microplate luminometer (Berthold, Germany), according to the manufacturer’s protocol.
Table 1 Patient Characteristics. Variables
Study patients (n = 140)
Age, years (mean ± SD) Male sex, % (n) Han race, % (n) Time to enrollment, days (mean ± SD) DWI volume, mL (mean ± SD) Hypertension, % (n) Diabetes, % (n) Atrial fibrillation, % (n) Hypercholesterolemia, % (n) Smoke, % (n) Drink, % (n) Triglyceride (mmol/L) Cholesterol (mmol/L) Homocysteine (μmol/L) Baseline NIHSS Score (mean ± SD)
58.47 ± 10.91 69.29 (97) 100 (140) 3.24 ± 2.57 17.43 ± 9.05 77.86 (109) 28.57 (40) 3.57 (5) 13.57 (19) 33.57 (47) 26.42 (37) 1.50 ± 0.95 4.26 ± 1.00 15.96 ± 8.89 4.19 ± 3.56
2.4. RNA extraction and real-time reverse transcription polymerase chain reaction (RT-PCR) Total RNA of PBMCs and polymorphonuclear cells were extracted using TRIzol reagent (Invitrogen Life Technologies, Carlsbad, CA, USA). To determine miRNA expression, the RNA was subsequently reversetranscribed using the miScriptII RT Kit (Qiagen) according to the manufacturer’s instructions. The expression levels of miR-1305 and miR-4437 were quantified with SYBR Green Master (Roche) using a miRNA-specific forward primer and a universal poly (T) adaptor reverse primer. The U6 small nuclear RNA served as the internal reference. Real-time PCR was performed on an Agilent Mx3005 P Real-time PCR system using FS Universal.
DWI, diffusion weighted imaging; NIHSS, National Institutes of Health Stroke Scale; SD, standard deviation. 441
Molecular Immunology 101 (2018) 440–449
Q. He et al.
2.5. In vitro transwell migration assay The Transwell migration assay, which is typically used to measure cell motility and invasiveness toward a chemo-attractant gradient, was performed using 24-well Transwell plates (pore size 5 μm; Corning Incorporated, USA) as described previously (Brennan et al., 2013; Moriconi et al., 2007). Briefly, 50 u L of a peripheral leucocyte suspension (1.5 × 106 cells/mL) from the stroke patients was pre-incubated at 37 °C for 15 min in the presence or absence of Reprarixin (20 nM, MedChemExpress) or SB225002 (10 nM, MedChemExpress) and then the cells were seeded in the upper compartment of the Transwell plate. Recombinant CXCL8 (600 μL of media, final concentration 50 ng/mL) was added to the lower chamber in certain experimental conditions. Subsequently, the plates were incubated for 2 h at 37 °C in a 5% CO2-humidified incubator. Finally, the proportions of the various subsets of cells in the lower chamber were analyzed.
2.6. Enzyme-linked immunosorbent assay (ELISA) for plasma CXCL8 The serum concentration of CXCL8 was measured using a sandwich ELISA kit according to the manufacturer’s protocol (BioLegend, San Diego, CA) as described previously (Li et al., 2016). All samples were measured in duplicate.
2.7. Flow cytometry analysis
Fig. 1. CXC Chemokine Receptor 2 (CXCR2) and interleukin 8 (CXCL8) were upregulated in patients with ischemic stroke. (A) Data from the GSE16561 dataset of the CXCR2 expression Z score in peripheral blood mononuclear cells (PBMCs) of patients with ischemic stroke. (B) Flow cytometric analysis of CXCR2+ cells in the peripheral leucocytes of patients (N = 20) and healthy donors (N = 10). (C) The data from the GSE16561 dataset of the CXCL8 expression Z score in PBMCs from patients with ischemic stroke. (D) The protein level of CXCL8 in plasma of patients (N = 120) and healthy donors (N = 60), measured by enzyme-linked immunosorbent assay (ELISA).
For the flow cytometry analysis, 20 patients with stroke and 10 healthy donors were randomly selected from the total participants included in this study. Peripheral leucocytes were washed twice with phosphate buffered saline (PBS), resuspended in flow buffer (PBS containing 2% FBS) and stained with anti-cluster of differentiation 3 (CD3), anti-CD8, anti-CD4, anti-CXCR2, anti-CD11b, anti-CD33, and anti-CD56 antibodies (Biolegend, San Diego, CA) for 30 min on ice. The detailed antibody information used in this study is presented in Supplementary Table 1. In addition, nuclear staining with 7-amino-actinomycin D (7AAD, BioLegend) was employed as a standard technique to exclude dead cells following manufacturer’s instructions. After adding 7-AAD for 5 to 10 min in the dark, stained cells were analyzed using a 6-color Fluorescence-Activated Cell Sorting (FACS) CantoII flow cytometer (Becton Dickinson, San Jose, CA). To detect non-specific signals, concentration- and isotype-matched non-specific antibodies were used. Supplementary Fig. 1 demonstrated the flow cytometry gating strategy.
3. Results 3.1. CXCL8-CXCR2 chemotactic axis is upregulated in the peripheral blood of patients with ischemic stroke To analyze the expression of CXCR2 mRNA in patients with ischemic stroke, the GEO database was explored. The GSE16561 cluster analysis dataset (Barr et al., 2010) demonstrated that CXCR2 expression was increased in patients with stroke (Fig. 1A), while the GSE22255 dataset (Krug et al., 2012) indicated no statistical difference between patients and healthy donors (data not show). Peripheral leucocytes were subsequently collected from 20 patients with ischemic stroke and 10 matched healthy controls for flow cytometric analysis. The results indicated a significantly higher percentage of CXCR2+ cells in peripheral leucocytes of patients compared to the peripheral leucocytes of healthy donors (Fig. 1B). Both the GSE16561 (Barr et al., 2010) (data not show) and GSE22255 (Krug et al., 2012) (Fig. 1C) datasets provided evidence of increased CXCL8 mRNA expression in patients. However, only the GSE22255 dataset (Krug et al., 2012) indicated statistical difference. The results of the plasma ELISA assay indicated a significantly elevated CXCL8 protein level in patients (N = 120) compared with healthy controls (N = 60; Fig. 1D).
2.8. Bioinformatics and statistical analysis The accession numbers for all datasets reported in this study were: GSE16561 (Barr et al., 2010) (RNA microarray), GSE22255 (Krug et al., 2012) (RNA microarray), and GSE55937 (Jickling et al., 2014) (noncoding RNA microarray). The TargetScan program (http://www. targetscan.org/) was used to identify miRNAs that could regulate CXCR2. The processed gene or miRNA relative expression data were downloaded from these datasets using Gene Expression Omnibus (GEO, https://www.ncbi.nlm.nih.gov/geo/). The relative expression data were presented as the Z score (Ellis et al., 2001). Stata 12 (Stata-Corp LP, College Station, TX, USA) was used for all statistical analyses. Experimental data were presented as mean ± standard deviation (SD). Student’s t-test and Mann-Whitney nonparametric U test were used for group comparisons. Spearman correlation analysis was performed between CXCL8 or CXCR2 expression and other factors. Chi-square test was also used to evaluate statistically significant differences between categorical variables. Statistical significance was set at a P value < 0.05. Receiver operating characteristic (ROC) curve analysis was performed to evaluate if CXCL8 was a prognostic factor for ischemic stroke. The optimal cut-off level of CXCL8 was ascertained by Youden’s index.
3.2. CXCL8-CXCR2 chemotactic axis promotes the migration of granulocytes and NK cells To investigate which cell subtypes are recruited by the CXCL8CXCR2 chemotactic axis, we first analyzed the correlation between 442
Molecular Immunology 101 (2018) 440–449
Q. He et al.
Fig. 2. Interleukin 8 (CXCL8)-CXC Chemokine Receptor 2 (CXCR2) chemotactic axis was correlated with granulocytes and natural killer (NK) cells. (A) CXCR2+ cells in the peripheral leucocytes positively correlated with plasma CXCL8 concentration in patients with ischemic stroke (N = 20). (B) Analysis of GSE16561 data indicated that CXCR2 was positively correlated with CD11b and CD56, but negatively correlated with CD33. (C) Analysis of data from GSE22255 indicated that CXCR2 positively correlated with CD11b and CD33.
with the percentage of CD56 (data not shown). Subsequently, we performed a Transwell migration assay to further study the effect of the CXCL8-CXCR2 chemotactic axis on granulocytes and NK cells. CXCL8 has two main receptors, CXCR1 and CXCR2. We used the CXCR2 selective antagonist SB225002 and the CXCL8 receptor inhibitor Reparixin, which weakly inhibits CXCR2-mediated cell migration and strongly blocks CXCR1-mediated chemotaxis, to investigate the solo effect of CXCR2 on granulocyte and NK cell attraction. We conducted the Transwell migration assay in four groups. Group one was without antagonist pre-treatment and without CXCL8 in the lower chamber. Group two only had CXCL8 in the lower chamber. Groups three and four were pre-treated with SB225002 and Reparixin, respectively, and then CXCL8 was added in the lower chamber. Fig. 4A–C showed the flow cytometric analysis results of CD11b+, CD33+, and CD56+ cells in the lower chamber, respectively. When analyzing the absolute number of cells, there were statistical differences between each of the groups (Fig. 4D). The percentages of CD11b+, CD33+ and CD56+ cells in the lower chamber were shown in Fig. 4E–G, respectively. The results indicated that CXCL8 could attract CD11b+, CD33+
CXCL8 and CXCR2 expression. As shown in Fig. 2A, CXCR2 in peripheral leucocytes were positively correlated with the serum concentration of CXCL8 (R = 0.74, P < 0.001). To label various cell subtypes, unique cell markers were used, such as CD8 and CD4 for T cells, CD19 and CD20 for B cells, CD11b and CD33 for granulocytes, CD56 for NK cells, and CD14 for monocytes. The GSE16561 dataset (Barr et al., 2010) provided evidence that CXCR2 positively correlated with CD11b and CD56, and negatively correlated with CD33, indicating that CXCR2 was mainly expressed by mature granulocytes and NK cells (Fig. 2B). Analysis of the GSE22255 dataset (Krug et al., 2012) demonstrated that CXCR2 was expressed by both mature and early-stage granulocytes (Fig. 2C). We further confirmed these results at the protein level in 20 patients. Fig. 3 shows the coexpression of CXCR2 with CD11b (Fig. 3A), CD33 (Fig. 3B), and CD56 (Fig. 3C). Analyzing the coexpression of CXCR2 with CD11b or CD33, we found that the percentage of doublenegative and double-positive cells was higher compared to the other two groups. For CD56, the percentage of double-negative cells was much higher than that in the other three groups. Through correlation analysis, we found the percentage of CXCR2 was positively correlated 443
Molecular Immunology 101 (2018) 440–449
Q. He et al.
Fig. 3. Co-expressions of CXCR2 with CD11b, CD33 and CD56 (N = 20) in peripheral leucocytes. (A), Co-expression of CXCR2 with CD11b. (B), Co-expression of CXCR2 with CD33. (C), Co-expression of CXCR2 with CD56.
were subsequently transfected with miR-1305 mimics, miR-4437 mimics, or NC and evaluated for the expression level of CXCR2 mRNA in the cells. The results indicated a significant downregulation in the miR4437 transfected group (Fig. 5D). Furthermore, miR-4437 overexpression also resulted in CXCR2 downregulation at the protein level by flow cytometry (Fig. 5E). The CXCR2 3ʹUTR has two binding sites for miR-4437, namely from 228 to 250 and 1224 to 1246 (Fig. 6A, C). Firefly and Renilla Luciferase Assay kits were used to verify whether miR-4437 can directly recognize these two binding sites. The expression ratio of firefly to renilla was decreased when co-transfected with miR-4437 and wildtype (WT) reconstructed plasmids. This result indicated that miR-4437 could directly recognize both the upstream and downstream binding sites (Fig. 6B, D). Taken together, these findings strongly indicated the possibility that the increased CXCR2 expression was the direct result of miR-4437 downregulation.
and CD56+ cells, as the percentages of these three kinds of cells were increased in the CXCL8 group. For CD11b + cells, CXCR1 and CXCR2 had an equal effect, since there was no difference between CXCL8 plus SB225002 and CXCL8 plus Reparixin (Fig. 4E). For CD33+ cells, CXCR2 played a dominant role because CD33+ cells were more inhibited in the CXCL8 plus SB225002 group (Fig. 4F). For CD56+ cells, CXCR1 played a dominant role (Fig. 4G). Taken together, these findings strongly indicated the involvement of the CXCL8-CXCR2 chemotactic axis in the recruitment of granulocytes and NK cells in ischemic stroke. 3.3. CXCR2 is regulated by MiR-4437 It has been previously indicated that CXCR2+ granulocytes were rapidly activated and released from the bone marrow in a mouse model of ischemic stroke (Denes et al., 2011). The question remains as to whether there are other post-transcriptional modifications that lead to the upregulation of CXCR2 expression. Therefore, a search was performed for ischemic stroke-related decreased microRNAs that could downregulate CXCR2 in 20 patients and 20 healthy controls (Fig. 5). The strategy to search for appropriate microRNAs is shown in Fig. 5A. Two potential microRNAs, miR-1305 and miR-4437, were discovered. Real-time results indicated that these two microRNAs are down-regulated in both PBMCs (Fig. 5B) and granulocytes (Fig. 5C). 293 T cells
3.4. Serum CXCL8 level is a potential prognostic factor for ischemic stroke To investigate the effect of the CXCL8-CXCR2 chemotactic axis in ischemic stroke, the potential prognostic role of CXCL8 was assessed by evaluating the mRS. There was a positive correlation between the serum CXCL8 levels and mRS at day 90 post-ischemia (Fig. 7A). 444
Molecular Immunology 101 (2018) 440–449
Q. He et al.
Fig. 4. Results of the Transwell migration assay (N = 5 independent experiments). S stands for CXCR2 selective antagonist SB225002. R stands for CXCL8 receptors inhibitor Reparixin. (A), Flow cytometry chart of CD11b+ cells in the lower chamber (B), Flow cytometry chart of CD33+ cells in the lower chamber (C) Flow cytometry chart of CD56+ cells in the lower chamber. (D), Absolute cell numbers in the lower chamber. (E), Statistical chart of CD11b+ cells percentage in the lower chamber. (F), Statistical chart of CD33+ cells percentage in the lower chamber. (G), Statistical chart of CD56+ cells percentage in the lower chamber.
445
Molecular Immunology 101 (2018) 440–449
Q. He et al.
Fig. 5. MicroRNA (Mir)-4437 downregulates CXC Chemokine Receptor 2 (CXCR2) expression. (A) The strategy used to search for the appropriate microRNAs that regulated CXCR2 in ischemic stroke. (B) Compared with controls (N = 20), miR-1305 and miR-4437 were decreased in patients with ischemic stroke (N = 46) in PBMCs. (C), Compared with controls (N = 10), miR-1305 and miR-4437 were decreased in patients with ischemic stroke (N = 20) in granulocytes.(D) When transfected with miR-1305 mimics, miR-4437 mimics, and scrambled miRNA sequence (NC), the CXCR2 mRNA expression in 293 T cells was downregulated only following miR-4437 mimic transfection (N = 3 independent experiments). (E) When miR-4437 was overexpressed, CXCR2 was downregulated at the protein level as evidenced by flow cytometric assay (N = 3 independent experiments).
among the first responders following infection or injury, including stroke, which is indicated by their rapid migration to the affected tissues (Kim and Luster, 2015). The recruitment and activation of granulocytes play an important role in asthma (Bruijnzeel et al., 2015), chronic obstructive pulmonary disease (Silva et al., 2015), gout (Maueroder et al., 2015), and ischemic heart disease (Nahrendorf et al., 2015). In ischemic stroke, the increase of granulocytes and especially neutrophils is associated with stroke severity, infarct volume, and worse functional outcomes (Jickling et al., 2015). In this study, both CXCL8 and CXCR2 were increased in ischemic stroke, indicating an involvement of the CXCL8-CXCR2 axis in the recruitment of immunocytes. Results of the flow cytometric analysis indicated that CXCR2 was mainly expressed by granulocytes and NK cells. Additionally, the Transwell migration assay demonstrated that CXCL8 could recruit both granulocytes and NK cells. In conclusion, the CXCL8CXCR2 chemotactic axis plays an important role in the recruitment of granulocytes and NK cells. Furthermore, our results demonstrated that CXCR2 was negatively regulated by miR-4437, indicating that the
Furthermore, the serum CXCL8 levels were significantly higher in patients with unfavorable functional outcome (N = 47) compared with those with a favorable outcome (N = 73; Fig. 7B). Given that the volume of the cerebral infarction in the first week post-stroke is a useful indicator of the outcome, we compared CXCL8 levels with infarct volume detected by DWI. The results indicated a positive correlation between the two measures (Fig. 7C). Furthermore, based on the ROC curve (Fig. 7D), the area under the curve (AUC) that reflects the association of CXCL8 level and mRS was 0.78, and the optimal cut-off value of CXCL8 levels as a prognostic indicator of functional outcome was 34 pg/mL. Additionally, the sensitivity and specificity of the optimal cut-off value were 65.22% and 78.08%, respectively. 4. Discussion Accumulated evidence has suggested that ischemic stroke is a systemic inflammatory disease involving various immunocytes and cytokines. More specifically, granulocytes, especially neutrophils, are 446
Molecular Immunology 101 (2018) 440–449
Q. He et al.
Fig. 6. CXC Chemokine Receptor 2 (CXCR2) is a direct target of microRNA (miR)-4473. (A) Sequence of the wild-type (WT) miR-4437 site in the 3ʹUTR of CXCR2 (280–250) and mutation (MUT) of the core binding site. (B) The 3ʹUTR (280–250) with the WT, but not the MUT miR-4437 site, resulted in the miR-4437induced regulation of the luciferase reporter construct (N = 3 independent experiments). (C) Sequence of the WT miR-4437 site in the 3ʹUTR of CXCR2 (1224–1246) and MUT of the core binding site. (D) The 3ʹUTR (1224–1246) with the WT, but not the MUT miR-4437 site, resulted in the miR-4437-induced regulation of the luciferase reporter construct (N = 3 independent experiments).
onset was 51.7 years (Krug et al., 2012, 2010), suggesting that blood was drawn approximately 1 year from symptom onset in these patients. In the present study, patients were recruited within 1 week from symptom onset; thus, our results are more comparable to those of the GSE16561 dataset (Barr et al., 2010). Furthermore, both GSE16561 (Barr et al., 2010) and GSE22255 (Krug et al., 2012) used RNA microarrays to detect mRNA gene expression. Numerous post-transcriptional regulations such as RNA mis-splicing (Scotti and Swanson, 2016), RNA N-6-methyladenosine methylation (Yue et al., 2015), and regulatory elements targeting the 3ʹ-UTR including microRNAs, RNAbinding proteins, or long noncoding RNAs (Schwerk and Savan, 2015) can impede the translation efficiency. Therefore, for more accurate results, we evaluated the protein expression by flow cytometry and ELISA. While the initial aim of this study was to demonstrate the involvement of the CXCL8-CXCR2 chemotactic axis in granulocyte recruitment in ischemic stroke, our results also indicated a positive correlation between the CXCL8-CXCR2 chemotactic axis and the neural cell adhesion molecule CD56. It is well established that CD56 is expressed on both NK and NKT cells. The NKT cells are a unique lymphoid population of cells that express components of both the innate and adaptive arms of the immune system (Godfrey et al., 2004), and account for approximately 5% of the peripheral blood lymphocytes (O’Keeffe et al., 2015). NKT cells are CD1d-restricted, lipid antigen-reactive, immunoregulatory T lymphocytes that can promote cell-mediated immunity to tumors and infectious organisms, yet paradoxically can also suppress cell-mediated immunity associated with autoimmune disease and allograft rejection
increase of CXCR2 in ischemic stroke was partially caused by posttranscriptional modifications. Brain ischemia is typically followed by an acute systemic inflammatory reaction, which correlates with the stroke severity. However, this inflammatory reaction could be extended by stroke-related infections (e.g., stroke-associated pneumonia), inflammatory diseases (e.g., deep-vein thrombosis), and stroke risk factors (e.g., atherosclerosis, diabetes mellitus, and coronary artery disease) (Dziedzic, 2015). Therefore, systemic inflammation persists after the acute phase of ischemia stroke (days 0–3). Indeed, greater differences in several inflammatory cytokines between patients and healthy controls occur 4–7 days following ischemic stroke (Dziedzic, 2015; Emsley et al., 2003; Siniscalchi et al., 2016). Therefore, we specifically collected blood samples from the participants not only in the acute phase (poststroke days 1–3) but also in the subacute phase (post-stroke days 4–7). According to our results, the serum CXCL8 levels were not statistically different between the acute phase and subacute phase (data not show). Several discrepancies are worth noting between the GSE16561 (Barr et al., 2010) and GSE22255 (Krug et al., 2012) datasets. For example, while the GSE16561 dataset (Barr et al., 2010) indicated an upregulation of CXCR2 in ischemic stroke, GSE22255 (Krug et al., 2012) showed no difference between patients and healthy controls. The blood collection timing may account for these conflicts. Indeed, blood was drawn from patients with acute ischemic cerebrovascular syndrome at the early stage (within 24 h from symptom onset) in GSE16561 (Barr et al., 2010). In contrast, in GSE22255 (Krug et al., 2012), the mean age of patients at examination was 52.4 years, while the mean age at symptom 447
Molecular Immunology 101 (2018) 440–449
Q. He et al.
Fig. 7. Correlation of interleukin 8 (CXCL8) expression with the functional outcome. (A) The CXCL8 serum level was positively correlated with the modified Rankin Scale (mRS) at day 90 post-stroke (N = 120). (B) The CXCL8 serum levels between patients with favorable outcomes and unfavorable outcomes. (C) The CXCL8 serum level was positively correlated with infarct volumes. (D) Receiver operation curve (ROC) analysis indicated that the CXCL8 level may be used as a prognostic factor for the functional outcomes within 90 days following stroke.
ischemic stroke (Al-Bahrani et al., 2007; Grau et al., 2001a, b), it remains unknown whether CXCL8 could be a predictor for the functional outcome of ischemic stroke. We first confirmed that CXCL8 was a strong and independent prognostic marker of short-term functional outcome. However, there are several limitations in our study. First, we only evaluated the level of CXCL8 in plasma, not in the cerebral spinal fluid where it reflects inflammation in the central nervous system. Second, to obtain more definitive conclusions, a larger study and additional control samples are required to test the association of CXCL8 with functional outcome. These issues will be investigated in further studies.
(Godfrey et al., 2010). NK1.1 (CD161) is the most common NK cell receptor in NKT cells (Godfrey et al., 2010), suggesting that CD3+CD56+ NKT cells account for only a small group of NKT cells. Furthermore, after ischemic stroke, the proportion of NKT-like T cells in the peripheral blood was not significantly different from that in health controls (Peterfalvi et al., 2009). In contrast, NK cells account for approximately 5% to 10% of PBMCs in humans. Therefore, NK cells are the major source of CD56 in PBMCs. Therefore, we speculated that the CXCL8-CXCR2 chemotactic axis has an impact on NK cell invasion. However, further analysis will be warranted to identify the group of CD56+ cells that express CXCR2 and is attracted by CXCL8. Indeed, both NK and CD3+CD56+ NKT cells are pro-inflammatory cells (Rujkijyanont et al., 2013). Therefore, miR-4437, which is a negative regulator of CXCR2, might block the inflammatory reaction through at least those two types of immune cells. Furthermore, miR-4437 might be a new target for ischemic stroke treatment. CXCL8, which is a small soluble chemotactic factor of the CXC chemokine family, is produced by various types of cells including activated monocytes and macrophages, fibroblasts, lymphocytes, neutrophils, and endothelial cells (Singh et al., 2013). Various stimuli are involved in the CXCL8 induction, including IL-1, tumor necrosis factor (TNF), bacterial or viral products, and cellular stress (Rotondi et al., 2013). Moreover, CXCL8 has been advanced as a valuable prognostic predictor for a vast range of diseases such as acute dyspnea (Wiklund et al., 2016), dengue infection (Patra et al., 2015), acute myocardial infarction (Prondzinsky et al., 2012), and severe traumatic brain injury (Gopcevic et al., 2007). Although the level of CXCL8 is elevated in
5. Conclusions In conclusion, the CXCL8-CXCR2 chemotactic axis played an important role in granulocyte and NK cell recruitment. MiR-4437, which is a negative regulator of CXCR2, might be a new target for ischemic stroke treatment. Serum CXCL8 levels could also be a potential prognostic factor for ischemic stroke. Acknowledgments This research did not receive any specificgrant from funding agencies in the public, commercial, or not-for-profit sectors. Appendix A. Supplementary data Supplementary material related to this article can be found, in the 448
Molecular Immunology 101 (2018) 440–449
Q. He et al.
online version, at doi:https://doi.org/10.1016/j.molimm.2018.08.002.
Liu, J., Li, F., Ping, Y., Wang, L., Chen, X., Wang, D., Cao, L., Zhao, S., Li, B., Kalinski, P., et al., 2015. Local production of the chemokines CCL5 and CXCL10 attracts CD8+ T lymphocytes into esophageal squamous cell carcinoma. Oncotarget 6, 24978–24989. Ma, S., Zhao, H., Ji, X., Luo, Y., 2015. Peripheral to central: organ interactions in stroke pathophysiology. Exp. Neurol. 272, 41–49. Maueroder, C., Kienhofer, D., Hahn, J., Schauer, C., Manger, B., Schett, G., Herrmann, M., Hoffmann, M.H., 2015. How neutrophil extracellular traps orchestrate the local immune response in gout. J. Mol. Med. 93, 727–734. Montaner, J., Rovira, A., Molina, C.A., Arenillas, J.F., Ribo, M., Chacon, P., Monasterio, J., Alvarez-Sabin, J., 2003. Plasmatic level of neuroinflammatory markers predict the extent of diffusion-weighted image lesions in hyperacute stroke. J. Cereb. Blood Flow Metab. 23, 1403–1407. Moriconi, A., Cesta, M., Cervellera, M., Aramini, A., Coniglio, S., Colagioia, S., Beccari, A., Bizzarri, C., Cavicchia, M., Locati, M., et al., 2007. Design of noncompetitive interleukin-8 inhibitors acting on CXCR1 and CXCR2. J. Med. Chem. 50, 3984–4002. Nahrendorf, M., Frantz, S., Swirski, F.K., Mulder, W.J., Randolph, G., Ertl, G., Ntziachristos, V., Piek, J.J., Stroes, E.S., Schwaiger, M., et al., 2015. Imaging systemic inflammatory networks in ischemic heart disease. J. Am. Coll. Cardiol. 65, 1583–1591. O’Keeffe, J., Podbielska, M., Hogan, E.L., 2015. Invariant natural killer T cells and their ligands: focus on multiple sclerosis. Immunology 145, 468–475. Patra, G., Ghosh, M., Modak, D., Bandopadhyay, B., Saha, B., Mukhopadhyay, S., 2015. Status of circulating immune complexes, IL8 titers and cryoglobulins in patients with dengue infection. Indian J. Exp. Biol. 53, 719–725. Pedersen, E.D., Waje-Andreassen, U., Vedeler, C.A., Aamodt, G., Mollnes, T.E., 2004. Systemic complement activation following human acute ischaemic stroke. Clin. Exp. Immunol. 137, 117–122. Peterfalvi, A., Molnar, T., Banati, M., Pusch, G., Miko, E., Bogar, L., Pal, J., Szereday, L., Illes, Z., 2009. Impaired function of innate T lymphocytes and NK cells in the acute phase of ischemic stroke. Cerebrovasc. Dis. 28, 490–498. Picascia, A., Grimaldi, V., Iannone, C., Soricelli, A., Napoli, C., 2015. Innate and adaptive immune response in stroke: focus on epigenetic regulation. J. Neuroimmunol. 289, 111–120. Prondzinsky, R., Unverzagt, S., Lemm, H., Wegener, N.A., Schlitt, A., Heinroth, K.M., Dietz, S., Buerke, U., Kellner, P., Loppnow, H., et al., 2012. Interleukin-6, -7, -8 and -10 predict outcome in acute myocardial infarction complicated by cardiogenic shock. Clin. Res. Cardiol. 101, 375–384. Remus, E.W., Sayeed, I., Won, S., Lyle, A.N., Stein, D.G., 2015. Progesterone protects endothelial cells after cerebrovascular occlusion by decreasing MCP-1- and CXCL1mediated macrophage infiltration. Exp. Neurol. 271, 401–408. Rotondi, M., Coperchini, F., Chiovato, L., 2013. CXCL8 in thyroid disease: from basic notions to potential applications in clinical practice. Cytokine Growth Factor Rev. 24, 539–546. Rujkijyanont, P., Chan, W.K., Eldridge, P.W., Lockey, T., Holladay, M., Rooney, B., Davidoff, A.M., Leung, W., Vong, Q., 2013. Ex vivo activation of CD56(+) immune cells that eradicate neuroblastoma. Cancer Res. 73, 2608–2618. Schwerk, J., Savan, R., 2015. Translating the untranslated region. J. Immunol. 195, 2963–2971. Scotti, M.M., Swanson, M.S., 2016. RNA mis-splicing in disease. Nat. Rev. Genet. 17, 19–32. Silva, O.R., Montes, J.F., Garcia-Valero, J., Olloquequi, J., 2015. [Cellular effectors of the inflammatory response in chronic obstructive pulmonary disease (COPD)]. Rev. Med. Chil. 143, 1162–1171. Sims, J.R., Gharai, L.R., Schaefer, P.W., Vangel, M., Rosenthal, E.S., Lev, M.H., Schwamm, L.H., 2009. ABC/2 for rapid clinical estimate of infarct, perfusion, and mismatch volumes. Neurology 72, 2104–2110. Singh, J.K., Simoes, B.M., Howell, S.J., Farnie, G., Clarke, R.B., 2013. Recent advances reveal IL-8 signaling as a potential key to targeting breast cancer stem cells. Breast Cancer Res. 15, 210. Siniscalchi, A., Iannacchero, R., Anticoli, S., Pezzella, F.R., De Sarro, G., Gallelli, L., 2016. Anti-inflammatory strategies in stroke: a potential therapeutic target. Curr. Vasc. Pharmacol. 14, 98–105. Sousa, L.F., Coelho, F.M., Rodrigues, D.H., Campos, A.C., Barcelos Lda, S., Teixeira, M.M., Rachid, M.A., Teixeira, A.L., 2013. Blockade of CXCR1/2 chemokine receptors protects against brain damage in ischemic stroke in mice. Clinics Sao Paulo (Sao Paulo) 68, 391–394. Thomas, H.B., Moots, R.J., Edwards, S.W., Wright, H.L., 2015. Whose gene is it anyway? The effect of preparation purity on neutrophil transcriptome studies. PLoS One 10, e0138982. Veenstra, M., Ransohoff, R.M., 2012. Chemokine receptor CXCR2: physiology regulator and neuroinflammation controller? J. Neuroimmunol. 246, 1–9. Wiklund, K., Gransbo, K., Lund, N., Peyman, M., Tegner, L., Toni-Bengtsson, M., Wieloch, M., Melander, O., 2016. Inflammatory biomarkers predicting prognosis in patients with acute dyspnea. Am. J. Emerg. Med. 34, 370–374. Yue, Y., Liu, J., He, C., 2015. RNA N6-methyladenosine methylation in post-transcriptional gene expression regulation. Genes Dev. 29, 1343–1355. Zhang, Y., Gao, Z., Wang, D., Zhang, T., Sun, B., Mu, L., Wang, J., Liu, Y., Kong, Q., Liu, X., et al., 2014. Accumulation of natural killer cells in ischemic brain tissues and the chemotactic effect of IP-10. J. Neuroinflammation 11, 79.
References Al-Bahrani, A., Taha, S., Shaath, H., Bakhiet, M., 2007. TNF-alpha and IL-8 in acute stroke and the modulation of these cytokines by antiplatelet agents. Curr. Neurovasc. Res. 4, 31–37. Barr, T.L., Conley, Y., Ding, J., Dillman, A., Warach, S., Singleton, A., Matarin, M., 2010. Genomic biomarkers and cellular pathways of ischemic stroke by RNA gene expression profiling. Neurology 75, 1009–1014. Brennan, P.J., Brigl, M., Brenner, M.B., 2013. Invariant natural killer T cells: an innate activation scheme linked to diverse effector functions. Nat. Rev. Immunol. 13, 101–117. Bruijnzeel, P.L., Uddin, M., Koenderman, L., 2015. Targeting neutrophilic inflammation in severe neutrophilic asthma: can we target the disease-relevant neutrophil phenotype? J. Leukoc. Biol. 98, 549–556. Burrows, F., Haley, M.J., Scott, E., Coutts, G., Lawrence, C.B., Allan, S.M., Schiessl, I., 2016. Systemic inflammation affects reperfusion following transient cerebral ischaemia. Exp. Neurol. 277, 252–260. Denes, A., McColl, B.W., Leow-Dyke, S.F., Chapman, K.Z., Humphreys, N.E., Grencis, R.K., Allan, S.M., Rothwell, N.J., 2011. Experimental stroke-induced changes in the bone marrow reveal complex regulation of leukocyte responses. J. Cereb. Blood Flow Metab. 31, 1036–1050. Dziedzic, T., 2015. Systemic inflammation as a therapeutic target in acute ischemic stroke. Expert Rev. Neurother. 15, 523–531. Ellis, K.J., Shypailo, R.J., Hardin, D.S., Perez, M.D.C.G., Motil, K.J., Wong, W.W., Abrams, S.A., 2001. Z score prediction model for assessment of bone mineral content in pediatric diseases. J. Bone Miner. Res. 16, 1658–1664. Emsley, H.C., Smith, C.J., Gavin, C.M., Georgiou, R.F., Vail, A., Barberan, E.M., Hallenbeck, J.M., del Zoppo, G.J., Rothwell, N.J., Tyrrell, P.J., et al., 2003. An early and sustained peripheral inflammatory response in acute ischaemic stroke: relationships with infection and atherosclerosis. J. Neuroimmunol. 139, 93–101. Gan, Y., Liu, Q., Wu, W., Yin, J.X., Bai, X.F., Shen, R., Wang, Y., Chen, J., La Cava, A., Poursine-Laurent, J., et al., 2014. Ischemic neurons recruit natural killer cells that accelerate brain infarction. Proc. Natl. Acad. Sci. U.S.A. 111, 2704–2709. Godfrey, D.I., MacDonald, H.R., Kronenberg, M., Smyth, M.J., Van Kaer, L., 2004. NKT cells: what’s in a name? Nature reviews. Immunology 4, 231–237. Godfrey, D.I., Stankovic, S., Baxter, A.G., 2010. Raising the NKT cell family. Nat. Immunol. 11, 197–206. Gopcevic, A., Mazul-Sunko, B., Marout, J., Sekulic, A., Antoljak, N., Siranovic, M., Ivanec, Z., Margaritoni, M., Bekavac-Beslin, M., Zarkovic, N., 2007. Plasma interleukin-8 as a potential predictor of mortality in adult patients with severe traumatic brain injury. Tohoku J. Exp. Med. 211, 387–393. Grau, A.J., Aulmann, M., Lichy, C., Meiser, H., Buggle, F., Brandt, T., Grond-Ginsbach, C., 2001a. Increased cytokine release by leucocytes in survivors of stroke at young age. Eur. J. Clin. Invest. 31, 999–1006. Grau, A.J., Reis, A., Buggle, F., Al-Khalaf, A., Werle, E., Valois, N., Bertram, M., Becher, H., Grond-Ginsbach, C., 2001b. Monocyte function and plasma levels of interleukin-8 in acute ischemic stroke. J. Neurol. Sci. 192, 41–47. Herz, J., Sabellek, P., Lane, T.E., Gunzer, M., Hermann, D.M., Doeppner, T.R., 2015. Role of neutrophils in exacerbation of brain injury after focal cerebral ischemia in hyperlipidemic mice. Stroke 46, 2916–2925. Jauch, E.C., Saver, J.L., Adams Jr., H.P., Bruno, A., Connors, J.J., Demaerschalk, B.M., Khatri, P., McMullan Jr., P.W., Qureshi, A.I., Rosenfield, K., et al., 2013. Guidelines for the early management of patients with acute ischemic stroke: a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 44, 870–947. Jickling, G.C., Ander, B.P., Zhan, X., Noblett, D., Stamova, B., Liu, D., 2014. microRNA expression in peripheral blood cells following acute ischemic stroke and their predicted gene targets. PLoS One 9, e99283. Jickling, G.C., Liu, D., Ander, B.P., Stamova, B., Zhan, X., Sharp, F.R., 2015. Targeting neutrophils in ischemic stroke: translational insights from experimental studies. J. Cereb. Blood Flow Metab. 35, 888–901. Kim, N.D., Luster, A.D., 2015. The role of tissue resident cells in neutrophil recruitment. Trends Immunol. 36, 547–555. Krug, T., Manso, H., Gouveia, L., Sobral, J., Xavier, J.M., Albergaria, I., Gaspar, G., Correia, M., Viana-Baptista, M., Simoes, R.M., et al., 2010. Kalirin: a novel genetic risk factor for ischemic stroke. Hum. Genet. 127, 513–523. Krug, T., Gabriel, J.P., Taipa, R., Fonseca, B.V., Domingues-Montanari, S., FernandezCadenas, I., Manso, H., Gouveia, L.O., Sobral, J., Albergaria, I., et al., 2012. TTC7B emerges as a novel risk factor for ischemic stroke through the convergence of several genome-wide approaches. J. Cereb. Blood Flow Metab. 32, 1061–1072. Lee, S., Chu, H.X., Kim, H.A., Real, N.C., Sharif, S., Fleming, S.B., Mercer, A.A., Wise, L.M., Drummond, G.R., Sobey, C.G., 2015. Effect of a broad-specificity chemokinebinding protein on brain leukocyte infiltration and infarct development. Stroke 46, 537–544. Li, L., Yang, L., Wang, L., Wang, F., Zhang, Z., Li, J., Yue, D., Chen, X., Ping, Y., Huang, L., et al., 2016. Impaired T cell function in malignant pleural effusion is caused by TGFbeta derived predominantly from macrophages. Int. J. Cancer 139, 2261–2269.
449
Contents lists available at ScienceDirect
Molecular Immunology journal homepage: www.elsevier.com/locate/molimm
Interleukin 8 (CXCL8)-CXC chemokine receptor 2 (CXCR2) axis contributes to MiR-4437-associated recruitment of granulocytes and natural killer cells in ischemic stroke
T
Qianyi Hea,b, Xiaojuan Shia,c, Bin Zhoua,d, Junfang Tengb, Chaoqi Zhanga,c, Shasha Liua,d, Jingyao Liana,d, Benyan Luoe, Guoqiang Zhaof, Hong Lub, Yuming Xub, Yajun Lianb, Yanjie Jiab, ⁎ Yi Zhanga,c,d,g, a
Biotherapy Center, the First Affiliated Hospital of Zhengzhou University, Zhengzhou 450052, Henan, China Department of Neurology, the First Affiliated Hospital of Zhengzhou University, Zhengzhou 450052, Henan, China Department of Oncology, the First Affiliated Hospital of Zhengzhou University, Zhengzhou 450052, Henan, China d School of Life Sciences, Zhengzhou University, Zhengzhou 450052, Henan, China e Department of Neurology, First Affiliated Hospital, Medical College of Zhejiang University, Hangzhou 310003, Zhejiang, China f School of Basic Medical Sciences, Zhengzhou University, Zhengzhou 450001, Henan, China g Key Laboratory for Tumor Immunology and Immunotherapy of Henan Province, Zhengzhou 450052, Henan, China b c
A R T I C LE I N FO
A B S T R A C T
Keywords: Ischemic stroke CXCR2-CXCL8 microRNA Granulocytes Natural killer cells Prognosis
Granulocytes and natural killer (NK) cells have been linked to brain injury in ischemic stroke. However, their recruitment from peripheral leucocytes in stroke patients is not well understood. Here, the expression of the interleukin 8 (CXCL8) in plasma, and CXC chemokine receptor 2 (CXCR2) in peripheral leucocytes of patients with ischemic stroke were evaluated. Based on the results, CXCR2 expression positively correlated with granulocytes and NK cells, which were in turn attracted by CXCL8. The results also indicated that CXCR2 was a direct target of microRNA (miR)-4437, a negative regulator of CXCR2, which was downregulated in peripheral leucocytes from patients with ischemic stroke. Furthermore, serum CXCL8 levels were associated with the infarct volume and functional outcomes in patients with ischemic stroke. The results of the receiver operating characteristic curve analysis with an optimal cut-off value of 34 pg/mL indicated serum CXCL8 levels could be a prognostic indicator for ischemic stroke. In conclusion, these data highlighted the involvement of the CXCL8CXCR2 chemotactic axis in the recruitment of granulocytes and NK cells in ischemic stroke. Furthermore, miR4437 was suggested as a novel target for treating ischemic stroke, while the serum CXCL8 level could be a prognostic factor for ischemic stroke.
1. Introduction Accumulating evidence has indicated that the immune system and inflammation play a leading role in ischemic stroke, particularly in brain tissue damage, progression of ischemic lesions, and tissue repair (Burrows et al., 2016; Ma et al., 2015; Picascia et al., 2015). Typically following ischemic stroke there is a rapid infiltration of granulocytes,
particularly neutrophils, which can result in blood-brain barrier disruption, cerebral edema, and brain injury (Jickling et al., 2015). In addition to granulocytes, natural killer (NK) cells, which are key members of the innate immune system, accumulate in the ischemic hemisphere (Gan et al., 2014; Zhang et al., 2014). The NK cells determine the size of the brain infarct (Gan et al., 2014) and are responsible for ischemia-related neuronal death by secreting interferon
Abbreviations: NK, natural killer; PBMCs, peripheral blood mononuclear cells; CXCL8, interleukin-8; CXCR2, CXC chemokine receptor 2; IFN, interferon; miR, microRNA; AHA/ASA, American Heart Association/American Stroke Association; DWI, diffusion-weighted imaging; mRS, modified Rankin Scale; DMEM, Dulbecco’s modified eagle medium; FBS, fetal bovine serum; NC, negative control; RT-PCR, real-time reverse transcription polymerase; ELISA, enzyme-linked immunosorbent assay; PBS, phosphate buffered saline; PerCP, peridinin-chlorophyll-protein; CD, cluster of differentiation; PE-CY7, phycoerythrin-cyanin 7; APC, allophycocyanin; FITC, fluorescein isothiocyanate; FACS, Fluorescence-Activated Cell Sorting; GEO, Gene Expression Omnibus; SD, standard deviation; ROC, receiver operating characteristic; WT, wildtype; AUC, area under the curve; TNF, tumor necrosis factor ⁎ Corresponding author at: 1st Jianshe East Road, the First Affiliated Hospital of Zhengzhou University, Zhengzhou, 450052, Henan, China. E-mail address: [email protected] (Y. Zhang). https://doi.org/10.1016/j.molimm.2018.08.002 Received 13 February 2018; Received in revised form 16 July 2018; Accepted 3 August 2018 0161-5890/ © 2018 Elsevier Ltd. All rights reserved.
Molecular Immunology 101 (2018) 440–449
Q. He et al.
completed a 90-day clinical follow-up. The functional outcome was assessed using the modified Rankin Scale (mRS) with the CXCL8 levels being concealed. The mRS is typically used to evaluate the degree of disability or dependence in the daily activities of patients following stroke. The favorable functional outcome was defined by an mRS score of 0–2 points (0, no symptoms; 1, no significant disability; and 2, slight disability), while the unfavorable outcome was defined by an mRS score of 3–6 points (3, moderate disability; 4, moderately severe disability; 5, severe disability; and 6, deceased). In this study, we analyzed the relationship between ischemic stroke and peripheral leucocytes. According to cellular morphology, leucocytes can be divided into two groups: peripheral blood mononuclear cells (PBMCs) and polymorphonuclear cells. Polymorphonuclear cells were isolated by singlestep centrifugation of whole blood using Polymorphprep (Axis-Shield) as described previously (Thomas et al., 2015). Furthermore, PBMCs were collected using Ficoll-Hypaque density gradient centrifugation (Beijing Chemical Reagent Company, China) as described previously (Liu et al., 2015). Finally, we put these two groups of cells together for further testing.
(IFN)-γ (Gan et al., 2014; Zhang et al., 2014). On the other hand, chemokines are a class of pro-inflammatory cytokines that have the ability to attract and activate leukocytes. In the highly inflammatory sites of acute ischemic stroke, chemokines are mainly generated by microglial cells and infiltrating immune cells, resulting in an exacerbated inflammatory cascade and brain injury (Lee et al., 2015; Remus et al., 2015). The CXC chemokine receptor 2 (CXCR2) is a regulator of neutrophil homeostasis (Veenstra and Ransohoff, 2012). In experimental mice with ischemic stroke, it has been demonstrated that a vast amount of CXCR2+ granulocytes are activated and released from the bone marrow (Denes et al., 2011). Furthermore, CXCR2 blockade mitigated the neurological deficits and reduced the infarct volume (Herz et al., 2015; Sousa et al., 2013). Thus, CXCR2-positive granulocytes play an important role in ischemic brain injury. However, no study to date has evaluated the expression of CXCR2 in patients with stroke. Additionally, in ischemic stroke the expression of interleukin-8 (CXCL8), one of the ligands of CXCR2, is controversial. In particular, several previous studies have indicated no change in CXCL8 levels in patients with stroke (Montaner et al., 2003; Pedersen et al., 2004), while others demonstrated increased CXCL8 levels (Al-Bahrani et al., 2007; Grau et al., 2001a, b). Furthermore, the involvement of the CXCL8-CXCR2 chemotactic axis in the recruitment of immune cells in ischemic stroke is not well understood. This study aimed to investigate the effect and underlying mechanism of the CXCL8-CXCR2 chemotactic axis in ischemic stroke.
2.2. Cell culture and transfection 293 T cells were cultured at 37 °C in a 5% CO2-humidified incubator with Dulbecco’s Modified Eagle Medium (DMEM) high-glucose supplemented with 10% fetal bovine serum (FBS). The microRNA (miR)4437 mimic and negative control (NC) were chemically synthesized (Shanghai GenePharma Company, Shanghai, China) and had the following sequences: miR-4437 mimic, 5ʹ-UGGGCUCAGGGUACAAAG GUU-3ʹ and 5ʹ−CCUUUGUACCCUGAGCCCAUU-3ʹ; NC, 5ʹ-UUCUCCG AACGUGUCACGUTT-3ʹ and 5ʹ-ACGUGACACGUUCGGAGAATT-3ʹ. Cells were seeded onto 24-well plates at a density of 4 × 104 cells/well and co-transfected with oligonucleotides (100 nM, miR-4437 mimics or NC) and plasmids (500 ng, pmirGLO-CXCR2-3ʹUTR-WT or pmirGLOCXCR2-3ʹUTR-MUT). All procedures were performed using the Lipofectamine 3000 reagent (Invitrogen, Carlsbad, USA) following the manufacturer’s instructions. Forty-eight hours after transfection, cells were harvested for the luciferase activity assay.
2. Materials and methods 2.1. Patients A total of 140 patients experiencing the first week of their first-ever acute ischemic stroke were recruited at the First Affiliated Hospital of Zhengzhou University from 25 June 2015 to 20 June 2018. All enrolled patients signed a written informed consent, and all experimental protocols were approved by the Ethics Committees of the First Affiliated Hospital of Zhengzhou University. Patients were excluded if they had infections, autoimmune diseases, cancers, or other conditions that could impact immune system homeostasis. During this study, the diagnosis and treatment of ischemic stroke were carried out under the guidelines of the American Heart Association/American Stroke Association (AHA/ASA) (Jauch et al., 2013). Table 1 summarizes the basic characteristics of the enrolled patients. As controls, a total of 70 age- and sex-matched healthy donors were also recruited. The infarct volume was measured by diffusion-weighted imaging (DWI) according to the validated ABC/2 method (Sims et al., 2009). Most of the patients
2.3. Plasmid construction and luciferase activity assay The 60-bp sequence of the CXCR2 3ʹUTR, containing the predicted miR-4437 binding sites, and its mutant, were synthesized as primers by Sangon Biotech (China, Shanghai). Subsequently, sense and antisense oligonucleotides were mixed with 5X annealing buffer (Beyotime, China). The mixture was boiled for 10 min and then cooled slowly to ambient temperature. The double-stranded oligonucleotides were digested with SacI and Xhol. The resulting oligonucleotides were subcloned into the SacI and Xhol sites of the pmirGLO vector. Luciferase activity was measured using the Dual-Luciferase Reporter Assay System (Promega) on a Cetro XS3 LB 960 microplate luminometer (Berthold, Germany), according to the manufacturer’s protocol.
Table 1 Patient Characteristics. Variables
Study patients (n = 140)
Age, years (mean ± SD) Male sex, % (n) Han race, % (n) Time to enrollment, days (mean ± SD) DWI volume, mL (mean ± SD) Hypertension, % (n) Diabetes, % (n) Atrial fibrillation, % (n) Hypercholesterolemia, % (n) Smoke, % (n) Drink, % (n) Triglyceride (mmol/L) Cholesterol (mmol/L) Homocysteine (μmol/L) Baseline NIHSS Score (mean ± SD)
58.47 ± 10.91 69.29 (97) 100 (140) 3.24 ± 2.57 17.43 ± 9.05 77.86 (109) 28.57 (40) 3.57 (5) 13.57 (19) 33.57 (47) 26.42 (37) 1.50 ± 0.95 4.26 ± 1.00 15.96 ± 8.89 4.19 ± 3.56
2.4. RNA extraction and real-time reverse transcription polymerase chain reaction (RT-PCR) Total RNA of PBMCs and polymorphonuclear cells were extracted using TRIzol reagent (Invitrogen Life Technologies, Carlsbad, CA, USA). To determine miRNA expression, the RNA was subsequently reversetranscribed using the miScriptII RT Kit (Qiagen) according to the manufacturer’s instructions. The expression levels of miR-1305 and miR-4437 were quantified with SYBR Green Master (Roche) using a miRNA-specific forward primer and a universal poly (T) adaptor reverse primer. The U6 small nuclear RNA served as the internal reference. Real-time PCR was performed on an Agilent Mx3005 P Real-time PCR system using FS Universal.
DWI, diffusion weighted imaging; NIHSS, National Institutes of Health Stroke Scale; SD, standard deviation. 441
Molecular Immunology 101 (2018) 440–449
Q. He et al.
2.5. In vitro transwell migration assay The Transwell migration assay, which is typically used to measure cell motility and invasiveness toward a chemo-attractant gradient, was performed using 24-well Transwell plates (pore size 5 μm; Corning Incorporated, USA) as described previously (Brennan et al., 2013; Moriconi et al., 2007). Briefly, 50 u L of a peripheral leucocyte suspension (1.5 × 106 cells/mL) from the stroke patients was pre-incubated at 37 °C for 15 min in the presence or absence of Reprarixin (20 nM, MedChemExpress) or SB225002 (10 nM, MedChemExpress) and then the cells were seeded in the upper compartment of the Transwell plate. Recombinant CXCL8 (600 μL of media, final concentration 50 ng/mL) was added to the lower chamber in certain experimental conditions. Subsequently, the plates were incubated for 2 h at 37 °C in a 5% CO2-humidified incubator. Finally, the proportions of the various subsets of cells in the lower chamber were analyzed.
2.6. Enzyme-linked immunosorbent assay (ELISA) for plasma CXCL8 The serum concentration of CXCL8 was measured using a sandwich ELISA kit according to the manufacturer’s protocol (BioLegend, San Diego, CA) as described previously (Li et al., 2016). All samples were measured in duplicate.
2.7. Flow cytometry analysis
Fig. 1. CXC Chemokine Receptor 2 (CXCR2) and interleukin 8 (CXCL8) were upregulated in patients with ischemic stroke. (A) Data from the GSE16561 dataset of the CXCR2 expression Z score in peripheral blood mononuclear cells (PBMCs) of patients with ischemic stroke. (B) Flow cytometric analysis of CXCR2+ cells in the peripheral leucocytes of patients (N = 20) and healthy donors (N = 10). (C) The data from the GSE16561 dataset of the CXCL8 expression Z score in PBMCs from patients with ischemic stroke. (D) The protein level of CXCL8 in plasma of patients (N = 120) and healthy donors (N = 60), measured by enzyme-linked immunosorbent assay (ELISA).
For the flow cytometry analysis, 20 patients with stroke and 10 healthy donors were randomly selected from the total participants included in this study. Peripheral leucocytes were washed twice with phosphate buffered saline (PBS), resuspended in flow buffer (PBS containing 2% FBS) and stained with anti-cluster of differentiation 3 (CD3), anti-CD8, anti-CD4, anti-CXCR2, anti-CD11b, anti-CD33, and anti-CD56 antibodies (Biolegend, San Diego, CA) for 30 min on ice. The detailed antibody information used in this study is presented in Supplementary Table 1. In addition, nuclear staining with 7-amino-actinomycin D (7AAD, BioLegend) was employed as a standard technique to exclude dead cells following manufacturer’s instructions. After adding 7-AAD for 5 to 10 min in the dark, stained cells were analyzed using a 6-color Fluorescence-Activated Cell Sorting (FACS) CantoII flow cytometer (Becton Dickinson, San Jose, CA). To detect non-specific signals, concentration- and isotype-matched non-specific antibodies were used. Supplementary Fig. 1 demonstrated the flow cytometry gating strategy.
3. Results 3.1. CXCL8-CXCR2 chemotactic axis is upregulated in the peripheral blood of patients with ischemic stroke To analyze the expression of CXCR2 mRNA in patients with ischemic stroke, the GEO database was explored. The GSE16561 cluster analysis dataset (Barr et al., 2010) demonstrated that CXCR2 expression was increased in patients with stroke (Fig. 1A), while the GSE22255 dataset (Krug et al., 2012) indicated no statistical difference between patients and healthy donors (data not show). Peripheral leucocytes were subsequently collected from 20 patients with ischemic stroke and 10 matched healthy controls for flow cytometric analysis. The results indicated a significantly higher percentage of CXCR2+ cells in peripheral leucocytes of patients compared to the peripheral leucocytes of healthy donors (Fig. 1B). Both the GSE16561 (Barr et al., 2010) (data not show) and GSE22255 (Krug et al., 2012) (Fig. 1C) datasets provided evidence of increased CXCL8 mRNA expression in patients. However, only the GSE22255 dataset (Krug et al., 2012) indicated statistical difference. The results of the plasma ELISA assay indicated a significantly elevated CXCL8 protein level in patients (N = 120) compared with healthy controls (N = 60; Fig. 1D).
2.8. Bioinformatics and statistical analysis The accession numbers for all datasets reported in this study were: GSE16561 (Barr et al., 2010) (RNA microarray), GSE22255 (Krug et al., 2012) (RNA microarray), and GSE55937 (Jickling et al., 2014) (noncoding RNA microarray). The TargetScan program (http://www. targetscan.org/) was used to identify miRNAs that could regulate CXCR2. The processed gene or miRNA relative expression data were downloaded from these datasets using Gene Expression Omnibus (GEO, https://www.ncbi.nlm.nih.gov/geo/). The relative expression data were presented as the Z score (Ellis et al., 2001). Stata 12 (Stata-Corp LP, College Station, TX, USA) was used for all statistical analyses. Experimental data were presented as mean ± standard deviation (SD). Student’s t-test and Mann-Whitney nonparametric U test were used for group comparisons. Spearman correlation analysis was performed between CXCL8 or CXCR2 expression and other factors. Chi-square test was also used to evaluate statistically significant differences between categorical variables. Statistical significance was set at a P value < 0.05. Receiver operating characteristic (ROC) curve analysis was performed to evaluate if CXCL8 was a prognostic factor for ischemic stroke. The optimal cut-off level of CXCL8 was ascertained by Youden’s index.
3.2. CXCL8-CXCR2 chemotactic axis promotes the migration of granulocytes and NK cells To investigate which cell subtypes are recruited by the CXCL8CXCR2 chemotactic axis, we first analyzed the correlation between 442
Molecular Immunology 101 (2018) 440–449
Q. He et al.
Fig. 2. Interleukin 8 (CXCL8)-CXC Chemokine Receptor 2 (CXCR2) chemotactic axis was correlated with granulocytes and natural killer (NK) cells. (A) CXCR2+ cells in the peripheral leucocytes positively correlated with plasma CXCL8 concentration in patients with ischemic stroke (N = 20). (B) Analysis of GSE16561 data indicated that CXCR2 was positively correlated with CD11b and CD56, but negatively correlated with CD33. (C) Analysis of data from GSE22255 indicated that CXCR2 positively correlated with CD11b and CD33.
with the percentage of CD56 (data not shown). Subsequently, we performed a Transwell migration assay to further study the effect of the CXCL8-CXCR2 chemotactic axis on granulocytes and NK cells. CXCL8 has two main receptors, CXCR1 and CXCR2. We used the CXCR2 selective antagonist SB225002 and the CXCL8 receptor inhibitor Reparixin, which weakly inhibits CXCR2-mediated cell migration and strongly blocks CXCR1-mediated chemotaxis, to investigate the solo effect of CXCR2 on granulocyte and NK cell attraction. We conducted the Transwell migration assay in four groups. Group one was without antagonist pre-treatment and without CXCL8 in the lower chamber. Group two only had CXCL8 in the lower chamber. Groups three and four were pre-treated with SB225002 and Reparixin, respectively, and then CXCL8 was added in the lower chamber. Fig. 4A–C showed the flow cytometric analysis results of CD11b+, CD33+, and CD56+ cells in the lower chamber, respectively. When analyzing the absolute number of cells, there were statistical differences between each of the groups (Fig. 4D). The percentages of CD11b+, CD33+ and CD56+ cells in the lower chamber were shown in Fig. 4E–G, respectively. The results indicated that CXCL8 could attract CD11b+, CD33+
CXCL8 and CXCR2 expression. As shown in Fig. 2A, CXCR2 in peripheral leucocytes were positively correlated with the serum concentration of CXCL8 (R = 0.74, P < 0.001). To label various cell subtypes, unique cell markers were used, such as CD8 and CD4 for T cells, CD19 and CD20 for B cells, CD11b and CD33 for granulocytes, CD56 for NK cells, and CD14 for monocytes. The GSE16561 dataset (Barr et al., 2010) provided evidence that CXCR2 positively correlated with CD11b and CD56, and negatively correlated with CD33, indicating that CXCR2 was mainly expressed by mature granulocytes and NK cells (Fig. 2B). Analysis of the GSE22255 dataset (Krug et al., 2012) demonstrated that CXCR2 was expressed by both mature and early-stage granulocytes (Fig. 2C). We further confirmed these results at the protein level in 20 patients. Fig. 3 shows the coexpression of CXCR2 with CD11b (Fig. 3A), CD33 (Fig. 3B), and CD56 (Fig. 3C). Analyzing the coexpression of CXCR2 with CD11b or CD33, we found that the percentage of doublenegative and double-positive cells was higher compared to the other two groups. For CD56, the percentage of double-negative cells was much higher than that in the other three groups. Through correlation analysis, we found the percentage of CXCR2 was positively correlated 443
Molecular Immunology 101 (2018) 440–449
Q. He et al.
Fig. 3. Co-expressions of CXCR2 with CD11b, CD33 and CD56 (N = 20) in peripheral leucocytes. (A), Co-expression of CXCR2 with CD11b. (B), Co-expression of CXCR2 with CD33. (C), Co-expression of CXCR2 with CD56.
were subsequently transfected with miR-1305 mimics, miR-4437 mimics, or NC and evaluated for the expression level of CXCR2 mRNA in the cells. The results indicated a significant downregulation in the miR4437 transfected group (Fig. 5D). Furthermore, miR-4437 overexpression also resulted in CXCR2 downregulation at the protein level by flow cytometry (Fig. 5E). The CXCR2 3ʹUTR has two binding sites for miR-4437, namely from 228 to 250 and 1224 to 1246 (Fig. 6A, C). Firefly and Renilla Luciferase Assay kits were used to verify whether miR-4437 can directly recognize these two binding sites. The expression ratio of firefly to renilla was decreased when co-transfected with miR-4437 and wildtype (WT) reconstructed plasmids. This result indicated that miR-4437 could directly recognize both the upstream and downstream binding sites (Fig. 6B, D). Taken together, these findings strongly indicated the possibility that the increased CXCR2 expression was the direct result of miR-4437 downregulation.
and CD56+ cells, as the percentages of these three kinds of cells were increased in the CXCL8 group. For CD11b + cells, CXCR1 and CXCR2 had an equal effect, since there was no difference between CXCL8 plus SB225002 and CXCL8 plus Reparixin (Fig. 4E). For CD33+ cells, CXCR2 played a dominant role because CD33+ cells were more inhibited in the CXCL8 plus SB225002 group (Fig. 4F). For CD56+ cells, CXCR1 played a dominant role (Fig. 4G). Taken together, these findings strongly indicated the involvement of the CXCL8-CXCR2 chemotactic axis in the recruitment of granulocytes and NK cells in ischemic stroke. 3.3. CXCR2 is regulated by MiR-4437 It has been previously indicated that CXCR2+ granulocytes were rapidly activated and released from the bone marrow in a mouse model of ischemic stroke (Denes et al., 2011). The question remains as to whether there are other post-transcriptional modifications that lead to the upregulation of CXCR2 expression. Therefore, a search was performed for ischemic stroke-related decreased microRNAs that could downregulate CXCR2 in 20 patients and 20 healthy controls (Fig. 5). The strategy to search for appropriate microRNAs is shown in Fig. 5A. Two potential microRNAs, miR-1305 and miR-4437, were discovered. Real-time results indicated that these two microRNAs are down-regulated in both PBMCs (Fig. 5B) and granulocytes (Fig. 5C). 293 T cells
3.4. Serum CXCL8 level is a potential prognostic factor for ischemic stroke To investigate the effect of the CXCL8-CXCR2 chemotactic axis in ischemic stroke, the potential prognostic role of CXCL8 was assessed by evaluating the mRS. There was a positive correlation between the serum CXCL8 levels and mRS at day 90 post-ischemia (Fig. 7A). 444
Molecular Immunology 101 (2018) 440–449
Q. He et al.
Fig. 4. Results of the Transwell migration assay (N = 5 independent experiments). S stands for CXCR2 selective antagonist SB225002. R stands for CXCL8 receptors inhibitor Reparixin. (A), Flow cytometry chart of CD11b+ cells in the lower chamber (B), Flow cytometry chart of CD33+ cells in the lower chamber (C) Flow cytometry chart of CD56+ cells in the lower chamber. (D), Absolute cell numbers in the lower chamber. (E), Statistical chart of CD11b+ cells percentage in the lower chamber. (F), Statistical chart of CD33+ cells percentage in the lower chamber. (G), Statistical chart of CD56+ cells percentage in the lower chamber.
445
Molecular Immunology 101 (2018) 440–449
Q. He et al.
Fig. 5. MicroRNA (Mir)-4437 downregulates CXC Chemokine Receptor 2 (CXCR2) expression. (A) The strategy used to search for the appropriate microRNAs that regulated CXCR2 in ischemic stroke. (B) Compared with controls (N = 20), miR-1305 and miR-4437 were decreased in patients with ischemic stroke (N = 46) in PBMCs. (C), Compared with controls (N = 10), miR-1305 and miR-4437 were decreased in patients with ischemic stroke (N = 20) in granulocytes.(D) When transfected with miR-1305 mimics, miR-4437 mimics, and scrambled miRNA sequence (NC), the CXCR2 mRNA expression in 293 T cells was downregulated only following miR-4437 mimic transfection (N = 3 independent experiments). (E) When miR-4437 was overexpressed, CXCR2 was downregulated at the protein level as evidenced by flow cytometric assay (N = 3 independent experiments).
among the first responders following infection or injury, including stroke, which is indicated by their rapid migration to the affected tissues (Kim and Luster, 2015). The recruitment and activation of granulocytes play an important role in asthma (Bruijnzeel et al., 2015), chronic obstructive pulmonary disease (Silva et al., 2015), gout (Maueroder et al., 2015), and ischemic heart disease (Nahrendorf et al., 2015). In ischemic stroke, the increase of granulocytes and especially neutrophils is associated with stroke severity, infarct volume, and worse functional outcomes (Jickling et al., 2015). In this study, both CXCL8 and CXCR2 were increased in ischemic stroke, indicating an involvement of the CXCL8-CXCR2 axis in the recruitment of immunocytes. Results of the flow cytometric analysis indicated that CXCR2 was mainly expressed by granulocytes and NK cells. Additionally, the Transwell migration assay demonstrated that CXCL8 could recruit both granulocytes and NK cells. In conclusion, the CXCL8CXCR2 chemotactic axis plays an important role in the recruitment of granulocytes and NK cells. Furthermore, our results demonstrated that CXCR2 was negatively regulated by miR-4437, indicating that the
Furthermore, the serum CXCL8 levels were significantly higher in patients with unfavorable functional outcome (N = 47) compared with those with a favorable outcome (N = 73; Fig. 7B). Given that the volume of the cerebral infarction in the first week post-stroke is a useful indicator of the outcome, we compared CXCL8 levels with infarct volume detected by DWI. The results indicated a positive correlation between the two measures (Fig. 7C). Furthermore, based on the ROC curve (Fig. 7D), the area under the curve (AUC) that reflects the association of CXCL8 level and mRS was 0.78, and the optimal cut-off value of CXCL8 levels as a prognostic indicator of functional outcome was 34 pg/mL. Additionally, the sensitivity and specificity of the optimal cut-off value were 65.22% and 78.08%, respectively. 4. Discussion Accumulated evidence has suggested that ischemic stroke is a systemic inflammatory disease involving various immunocytes and cytokines. More specifically, granulocytes, especially neutrophils, are 446
Molecular Immunology 101 (2018) 440–449
Q. He et al.
Fig. 6. CXC Chemokine Receptor 2 (CXCR2) is a direct target of microRNA (miR)-4473. (A) Sequence of the wild-type (WT) miR-4437 site in the 3ʹUTR of CXCR2 (280–250) and mutation (MUT) of the core binding site. (B) The 3ʹUTR (280–250) with the WT, but not the MUT miR-4437 site, resulted in the miR-4437induced regulation of the luciferase reporter construct (N = 3 independent experiments). (C) Sequence of the WT miR-4437 site in the 3ʹUTR of CXCR2 (1224–1246) and MUT of the core binding site. (D) The 3ʹUTR (1224–1246) with the WT, but not the MUT miR-4437 site, resulted in the miR-4437-induced regulation of the luciferase reporter construct (N = 3 independent experiments).
onset was 51.7 years (Krug et al., 2012, 2010), suggesting that blood was drawn approximately 1 year from symptom onset in these patients. In the present study, patients were recruited within 1 week from symptom onset; thus, our results are more comparable to those of the GSE16561 dataset (Barr et al., 2010). Furthermore, both GSE16561 (Barr et al., 2010) and GSE22255 (Krug et al., 2012) used RNA microarrays to detect mRNA gene expression. Numerous post-transcriptional regulations such as RNA mis-splicing (Scotti and Swanson, 2016), RNA N-6-methyladenosine methylation (Yue et al., 2015), and regulatory elements targeting the 3ʹ-UTR including microRNAs, RNAbinding proteins, or long noncoding RNAs (Schwerk and Savan, 2015) can impede the translation efficiency. Therefore, for more accurate results, we evaluated the protein expression by flow cytometry and ELISA. While the initial aim of this study was to demonstrate the involvement of the CXCL8-CXCR2 chemotactic axis in granulocyte recruitment in ischemic stroke, our results also indicated a positive correlation between the CXCL8-CXCR2 chemotactic axis and the neural cell adhesion molecule CD56. It is well established that CD56 is expressed on both NK and NKT cells. The NKT cells are a unique lymphoid population of cells that express components of both the innate and adaptive arms of the immune system (Godfrey et al., 2004), and account for approximately 5% of the peripheral blood lymphocytes (O’Keeffe et al., 2015). NKT cells are CD1d-restricted, lipid antigen-reactive, immunoregulatory T lymphocytes that can promote cell-mediated immunity to tumors and infectious organisms, yet paradoxically can also suppress cell-mediated immunity associated with autoimmune disease and allograft rejection
increase of CXCR2 in ischemic stroke was partially caused by posttranscriptional modifications. Brain ischemia is typically followed by an acute systemic inflammatory reaction, which correlates with the stroke severity. However, this inflammatory reaction could be extended by stroke-related infections (e.g., stroke-associated pneumonia), inflammatory diseases (e.g., deep-vein thrombosis), and stroke risk factors (e.g., atherosclerosis, diabetes mellitus, and coronary artery disease) (Dziedzic, 2015). Therefore, systemic inflammation persists after the acute phase of ischemia stroke (days 0–3). Indeed, greater differences in several inflammatory cytokines between patients and healthy controls occur 4–7 days following ischemic stroke (Dziedzic, 2015; Emsley et al., 2003; Siniscalchi et al., 2016). Therefore, we specifically collected blood samples from the participants not only in the acute phase (poststroke days 1–3) but also in the subacute phase (post-stroke days 4–7). According to our results, the serum CXCL8 levels were not statistically different between the acute phase and subacute phase (data not show). Several discrepancies are worth noting between the GSE16561 (Barr et al., 2010) and GSE22255 (Krug et al., 2012) datasets. For example, while the GSE16561 dataset (Barr et al., 2010) indicated an upregulation of CXCR2 in ischemic stroke, GSE22255 (Krug et al., 2012) showed no difference between patients and healthy controls. The blood collection timing may account for these conflicts. Indeed, blood was drawn from patients with acute ischemic cerebrovascular syndrome at the early stage (within 24 h from symptom onset) in GSE16561 (Barr et al., 2010). In contrast, in GSE22255 (Krug et al., 2012), the mean age of patients at examination was 52.4 years, while the mean age at symptom 447
Molecular Immunology 101 (2018) 440–449
Q. He et al.
Fig. 7. Correlation of interleukin 8 (CXCL8) expression with the functional outcome. (A) The CXCL8 serum level was positively correlated with the modified Rankin Scale (mRS) at day 90 post-stroke (N = 120). (B) The CXCL8 serum levels between patients with favorable outcomes and unfavorable outcomes. (C) The CXCL8 serum level was positively correlated with infarct volumes. (D) Receiver operation curve (ROC) analysis indicated that the CXCL8 level may be used as a prognostic factor for the functional outcomes within 90 days following stroke.
ischemic stroke (Al-Bahrani et al., 2007; Grau et al., 2001a, b), it remains unknown whether CXCL8 could be a predictor for the functional outcome of ischemic stroke. We first confirmed that CXCL8 was a strong and independent prognostic marker of short-term functional outcome. However, there are several limitations in our study. First, we only evaluated the level of CXCL8 in plasma, not in the cerebral spinal fluid where it reflects inflammation in the central nervous system. Second, to obtain more definitive conclusions, a larger study and additional control samples are required to test the association of CXCL8 with functional outcome. These issues will be investigated in further studies.
(Godfrey et al., 2010). NK1.1 (CD161) is the most common NK cell receptor in NKT cells (Godfrey et al., 2010), suggesting that CD3+CD56+ NKT cells account for only a small group of NKT cells. Furthermore, after ischemic stroke, the proportion of NKT-like T cells in the peripheral blood was not significantly different from that in health controls (Peterfalvi et al., 2009). In contrast, NK cells account for approximately 5% to 10% of PBMCs in humans. Therefore, NK cells are the major source of CD56 in PBMCs. Therefore, we speculated that the CXCL8-CXCR2 chemotactic axis has an impact on NK cell invasion. However, further analysis will be warranted to identify the group of CD56+ cells that express CXCR2 and is attracted by CXCL8. Indeed, both NK and CD3+CD56+ NKT cells are pro-inflammatory cells (Rujkijyanont et al., 2013). Therefore, miR-4437, which is a negative regulator of CXCR2, might block the inflammatory reaction through at least those two types of immune cells. Furthermore, miR-4437 might be a new target for ischemic stroke treatment. CXCL8, which is a small soluble chemotactic factor of the CXC chemokine family, is produced by various types of cells including activated monocytes and macrophages, fibroblasts, lymphocytes, neutrophils, and endothelial cells (Singh et al., 2013). Various stimuli are involved in the CXCL8 induction, including IL-1, tumor necrosis factor (TNF), bacterial or viral products, and cellular stress (Rotondi et al., 2013). Moreover, CXCL8 has been advanced as a valuable prognostic predictor for a vast range of diseases such as acute dyspnea (Wiklund et al., 2016), dengue infection (Patra et al., 2015), acute myocardial infarction (Prondzinsky et al., 2012), and severe traumatic brain injury (Gopcevic et al., 2007). Although the level of CXCL8 is elevated in
5. Conclusions In conclusion, the CXCL8-CXCR2 chemotactic axis played an important role in granulocyte and NK cell recruitment. MiR-4437, which is a negative regulator of CXCR2, might be a new target for ischemic stroke treatment. Serum CXCL8 levels could also be a potential prognostic factor for ischemic stroke. Acknowledgments This research did not receive any specificgrant from funding agencies in the public, commercial, or not-for-profit sectors. Appendix A. Supplementary data Supplementary material related to this article can be found, in the 448
Molecular Immunology 101 (2018) 440–449
Q. He et al.
online version, at doi:https://doi.org/10.1016/j.molimm.2018.08.002.
Liu, J., Li, F., Ping, Y., Wang, L., Chen, X., Wang, D., Cao, L., Zhao, S., Li, B., Kalinski, P., et al., 2015. Local production of the chemokines CCL5 and CXCL10 attracts CD8+ T lymphocytes into esophageal squamous cell carcinoma. Oncotarget 6, 24978–24989. Ma, S., Zhao, H., Ji, X., Luo, Y., 2015. Peripheral to central: organ interactions in stroke pathophysiology. Exp. Neurol. 272, 41–49. Maueroder, C., Kienhofer, D., Hahn, J., Schauer, C., Manger, B., Schett, G., Herrmann, M., Hoffmann, M.H., 2015. How neutrophil extracellular traps orchestrate the local immune response in gout. J. Mol. Med. 93, 727–734. Montaner, J., Rovira, A., Molina, C.A., Arenillas, J.F., Ribo, M., Chacon, P., Monasterio, J., Alvarez-Sabin, J., 2003. Plasmatic level of neuroinflammatory markers predict the extent of diffusion-weighted image lesions in hyperacute stroke. J. Cereb. Blood Flow Metab. 23, 1403–1407. Moriconi, A., Cesta, M., Cervellera, M., Aramini, A., Coniglio, S., Colagioia, S., Beccari, A., Bizzarri, C., Cavicchia, M., Locati, M., et al., 2007. Design of noncompetitive interleukin-8 inhibitors acting on CXCR1 and CXCR2. J. Med. Chem. 50, 3984–4002. Nahrendorf, M., Frantz, S., Swirski, F.K., Mulder, W.J., Randolph, G., Ertl, G., Ntziachristos, V., Piek, J.J., Stroes, E.S., Schwaiger, M., et al., 2015. Imaging systemic inflammatory networks in ischemic heart disease. J. Am. Coll. Cardiol. 65, 1583–1591. O’Keeffe, J., Podbielska, M., Hogan, E.L., 2015. Invariant natural killer T cells and their ligands: focus on multiple sclerosis. Immunology 145, 468–475. Patra, G., Ghosh, M., Modak, D., Bandopadhyay, B., Saha, B., Mukhopadhyay, S., 2015. Status of circulating immune complexes, IL8 titers and cryoglobulins in patients with dengue infection. Indian J. Exp. Biol. 53, 719–725. Pedersen, E.D., Waje-Andreassen, U., Vedeler, C.A., Aamodt, G., Mollnes, T.E., 2004. Systemic complement activation following human acute ischaemic stroke. Clin. Exp. Immunol. 137, 117–122. Peterfalvi, A., Molnar, T., Banati, M., Pusch, G., Miko, E., Bogar, L., Pal, J., Szereday, L., Illes, Z., 2009. Impaired function of innate T lymphocytes and NK cells in the acute phase of ischemic stroke. Cerebrovasc. Dis. 28, 490–498. Picascia, A., Grimaldi, V., Iannone, C., Soricelli, A., Napoli, C., 2015. Innate and adaptive immune response in stroke: focus on epigenetic regulation. J. Neuroimmunol. 289, 111–120. Prondzinsky, R., Unverzagt, S., Lemm, H., Wegener, N.A., Schlitt, A., Heinroth, K.M., Dietz, S., Buerke, U., Kellner, P., Loppnow, H., et al., 2012. Interleukin-6, -7, -8 and -10 predict outcome in acute myocardial infarction complicated by cardiogenic shock. Clin. Res. Cardiol. 101, 375–384. Remus, E.W., Sayeed, I., Won, S., Lyle, A.N., Stein, D.G., 2015. Progesterone protects endothelial cells after cerebrovascular occlusion by decreasing MCP-1- and CXCL1mediated macrophage infiltration. Exp. Neurol. 271, 401–408. Rotondi, M., Coperchini, F., Chiovato, L., 2013. CXCL8 in thyroid disease: from basic notions to potential applications in clinical practice. Cytokine Growth Factor Rev. 24, 539–546. Rujkijyanont, P., Chan, W.K., Eldridge, P.W., Lockey, T., Holladay, M., Rooney, B., Davidoff, A.M., Leung, W., Vong, Q., 2013. Ex vivo activation of CD56(+) immune cells that eradicate neuroblastoma. Cancer Res. 73, 2608–2618. Schwerk, J., Savan, R., 2015. Translating the untranslated region. J. Immunol. 195, 2963–2971. Scotti, M.M., Swanson, M.S., 2016. RNA mis-splicing in disease. Nat. Rev. Genet. 17, 19–32. Silva, O.R., Montes, J.F., Garcia-Valero, J., Olloquequi, J., 2015. [Cellular effectors of the inflammatory response in chronic obstructive pulmonary disease (COPD)]. Rev. Med. Chil. 143, 1162–1171. Sims, J.R., Gharai, L.R., Schaefer, P.W., Vangel, M., Rosenthal, E.S., Lev, M.H., Schwamm, L.H., 2009. ABC/2 for rapid clinical estimate of infarct, perfusion, and mismatch volumes. Neurology 72, 2104–2110. Singh, J.K., Simoes, B.M., Howell, S.J., Farnie, G., Clarke, R.B., 2013. Recent advances reveal IL-8 signaling as a potential key to targeting breast cancer stem cells. Breast Cancer Res. 15, 210. Siniscalchi, A., Iannacchero, R., Anticoli, S., Pezzella, F.R., De Sarro, G., Gallelli, L., 2016. Anti-inflammatory strategies in stroke: a potential therapeutic target. Curr. Vasc. Pharmacol. 14, 98–105. Sousa, L.F., Coelho, F.M., Rodrigues, D.H., Campos, A.C., Barcelos Lda, S., Teixeira, M.M., Rachid, M.A., Teixeira, A.L., 2013. Blockade of CXCR1/2 chemokine receptors protects against brain damage in ischemic stroke in mice. Clinics Sao Paulo (Sao Paulo) 68, 391–394. Thomas, H.B., Moots, R.J., Edwards, S.W., Wright, H.L., 2015. Whose gene is it anyway? The effect of preparation purity on neutrophil transcriptome studies. PLoS One 10, e0138982. Veenstra, M., Ransohoff, R.M., 2012. Chemokine receptor CXCR2: physiology regulator and neuroinflammation controller? J. Neuroimmunol. 246, 1–9. Wiklund, K., Gransbo, K., Lund, N., Peyman, M., Tegner, L., Toni-Bengtsson, M., Wieloch, M., Melander, O., 2016. Inflammatory biomarkers predicting prognosis in patients with acute dyspnea. Am. J. Emerg. Med. 34, 370–374. Yue, Y., Liu, J., He, C., 2015. RNA N6-methyladenosine methylation in post-transcriptional gene expression regulation. Genes Dev. 29, 1343–1355. Zhang, Y., Gao, Z., Wang, D., Zhang, T., Sun, B., Mu, L., Wang, J., Liu, Y., Kong, Q., Liu, X., et al., 2014. Accumulation of natural killer cells in ischemic brain tissues and the chemotactic effect of IP-10. J. Neuroinflammation 11, 79.
References Al-Bahrani, A., Taha, S., Shaath, H., Bakhiet, M., 2007. TNF-alpha and IL-8 in acute stroke and the modulation of these cytokines by antiplatelet agents. Curr. Neurovasc. Res. 4, 31–37. Barr, T.L., Conley, Y., Ding, J., Dillman, A., Warach, S., Singleton, A., Matarin, M., 2010. Genomic biomarkers and cellular pathways of ischemic stroke by RNA gene expression profiling. Neurology 75, 1009–1014. Brennan, P.J., Brigl, M., Brenner, M.B., 2013. Invariant natural killer T cells: an innate activation scheme linked to diverse effector functions. Nat. Rev. Immunol. 13, 101–117. Bruijnzeel, P.L., Uddin, M., Koenderman, L., 2015. Targeting neutrophilic inflammation in severe neutrophilic asthma: can we target the disease-relevant neutrophil phenotype? J. Leukoc. Biol. 98, 549–556. Burrows, F., Haley, M.J., Scott, E., Coutts, G., Lawrence, C.B., Allan, S.M., Schiessl, I., 2016. Systemic inflammation affects reperfusion following transient cerebral ischaemia. Exp. Neurol. 277, 252–260. Denes, A., McColl, B.W., Leow-Dyke, S.F., Chapman, K.Z., Humphreys, N.E., Grencis, R.K., Allan, S.M., Rothwell, N.J., 2011. Experimental stroke-induced changes in the bone marrow reveal complex regulation of leukocyte responses. J. Cereb. Blood Flow Metab. 31, 1036–1050. Dziedzic, T., 2015. Systemic inflammation as a therapeutic target in acute ischemic stroke. Expert Rev. Neurother. 15, 523–531. Ellis, K.J., Shypailo, R.J., Hardin, D.S., Perez, M.D.C.G., Motil, K.J., Wong, W.W., Abrams, S.A., 2001. Z score prediction model for assessment of bone mineral content in pediatric diseases. J. Bone Miner. Res. 16, 1658–1664. Emsley, H.C., Smith, C.J., Gavin, C.M., Georgiou, R.F., Vail, A., Barberan, E.M., Hallenbeck, J.M., del Zoppo, G.J., Rothwell, N.J., Tyrrell, P.J., et al., 2003. An early and sustained peripheral inflammatory response in acute ischaemic stroke: relationships with infection and atherosclerosis. J. Neuroimmunol. 139, 93–101. Gan, Y., Liu, Q., Wu, W., Yin, J.X., Bai, X.F., Shen, R., Wang, Y., Chen, J., La Cava, A., Poursine-Laurent, J., et al., 2014. Ischemic neurons recruit natural killer cells that accelerate brain infarction. Proc. Natl. Acad. Sci. U.S.A. 111, 2704–2709. Godfrey, D.I., MacDonald, H.R., Kronenberg, M., Smyth, M.J., Van Kaer, L., 2004. NKT cells: what’s in a name? Nature reviews. Immunology 4, 231–237. Godfrey, D.I., Stankovic, S., Baxter, A.G., 2010. Raising the NKT cell family. Nat. Immunol. 11, 197–206. Gopcevic, A., Mazul-Sunko, B., Marout, J., Sekulic, A., Antoljak, N., Siranovic, M., Ivanec, Z., Margaritoni, M., Bekavac-Beslin, M., Zarkovic, N., 2007. Plasma interleukin-8 as a potential predictor of mortality in adult patients with severe traumatic brain injury. Tohoku J. Exp. Med. 211, 387–393. Grau, A.J., Aulmann, M., Lichy, C., Meiser, H., Buggle, F., Brandt, T., Grond-Ginsbach, C., 2001a. Increased cytokine release by leucocytes in survivors of stroke at young age. Eur. J. Clin. Invest. 31, 999–1006. Grau, A.J., Reis, A., Buggle, F., Al-Khalaf, A., Werle, E., Valois, N., Bertram, M., Becher, H., Grond-Ginsbach, C., 2001b. Monocyte function and plasma levels of interleukin-8 in acute ischemic stroke. J. Neurol. Sci. 192, 41–47. Herz, J., Sabellek, P., Lane, T.E., Gunzer, M., Hermann, D.M., Doeppner, T.R., 2015. Role of neutrophils in exacerbation of brain injury after focal cerebral ischemia in hyperlipidemic mice. Stroke 46, 2916–2925. Jauch, E.C., Saver, J.L., Adams Jr., H.P., Bruno, A., Connors, J.J., Demaerschalk, B.M., Khatri, P., McMullan Jr., P.W., Qureshi, A.I., Rosenfield, K., et al., 2013. Guidelines for the early management of patients with acute ischemic stroke: a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 44, 870–947. Jickling, G.C., Ander, B.P., Zhan, X., Noblett, D., Stamova, B., Liu, D., 2014. microRNA expression in peripheral blood cells following acute ischemic stroke and their predicted gene targets. PLoS One 9, e99283. Jickling, G.C., Liu, D., Ander, B.P., Stamova, B., Zhan, X., Sharp, F.R., 2015. Targeting neutrophils in ischemic stroke: translational insights from experimental studies. J. Cereb. Blood Flow Metab. 35, 888–901. Kim, N.D., Luster, A.D., 2015. The role of tissue resident cells in neutrophil recruitment. Trends Immunol. 36, 547–555. Krug, T., Manso, H., Gouveia, L., Sobral, J., Xavier, J.M., Albergaria, I., Gaspar, G., Correia, M., Viana-Baptista, M., Simoes, R.M., et al., 2010. Kalirin: a novel genetic risk factor for ischemic stroke. Hum. Genet. 127, 513–523. Krug, T., Gabriel, J.P., Taipa, R., Fonseca, B.V., Domingues-Montanari, S., FernandezCadenas, I., Manso, H., Gouveia, L.O., Sobral, J., Albergaria, I., et al., 2012. TTC7B emerges as a novel risk factor for ischemic stroke through the convergence of several genome-wide approaches. J. Cereb. Blood Flow Metab. 32, 1061–1072. Lee, S., Chu, H.X., Kim, H.A., Real, N.C., Sharif, S., Fleming, S.B., Mercer, A.A., Wise, L.M., Drummond, G.R., Sobey, C.G., 2015. Effect of a broad-specificity chemokinebinding protein on brain leukocyte infiltration and infarct development. Stroke 46, 537–544. Li, L., Yang, L., Wang, L., Wang, F., Zhang, Z., Li, J., Yue, D., Chen, X., Ping, Y., Huang, L., et al., 2016. Impaired T cell function in malignant pleural effusion is caused by TGFbeta derived predominantly from macrophages. Int. J. Cancer 139, 2261–2269.
449