Background
Glycyrrhizin (Gly), a glycoconjugated triterpene extracted from the licorice root,
Glycyrrhiza glabra [
1], is a natural anti-inflammatory product. It has been used clinically to treat hepatitis B and C in Japan [
2]. Gly inhibits pro-inflammatory cytokine and chemokine expression, such as CC-chemokine ligand 2 [
3] and TNF-α [
4]. More importantly, recent studies have shown that Gly protects brain tissue in global ischemia [
5], intracerebral hemorrhage-induced brain injury [
6], and focal ischemia [
7]. These studies have shown that Gly inhibits high-mobility group box 1 (HMGB1) release, neuroinflammation, and apoptotic cell signaling in the ischemic brain. Despite the solid protective effects of Gly against brain injury, the underlying protective mechanisms remain elusive.
In this current study, we hypothesized that Gly protects against brain injury via mediating T cell activity. This hypothesis is based on two observations. First, T cells have recently been shown to contribute to brain injury, as smaller infarction is found in severe combined immunodeficient (SCID) mice containing no T or B cells [
8]. Recently, we have reported that a T cell deficit in nude rats also resulted in smaller infarction induced by middle cerebral artery (MCA) suture occlusion [
9]. Second, previous studies have shown that Gly enhances anti-inflammatory cytokines, such as IL-10, and inhibits neutrophil infiltration in lipopolysaccharide-induced acute lung injury and T cell infiltration in liver in concanavalin A (ConA)-induced hepatitis [
4,
10]. However, whether T cells are important or not for the protective effect of Gly against stroke has not been studied.
As discussed above, although T cells are well known to contribute to brain injury induced by stroke, how T cells are recruited in the brain and activated after stroke remains unknown. We further hypothesized that T cell activity is modulated by HMGB1, and that Gly inhibits T cell activity via inhibition of HMGB1 release after stroke, based on the following reasons. First, HMGB1 is released in the brain from necrotic neurons as early as 1 h after stroke [
11] and is released into the cerebral spinal fluid (CSF) and bloodstream after stroke [
12,
13]. Second, once HMGB1 is released into the blood, it may regulate T cell function [
14,
15]. Third, Gly has been identified as an HMGB1 inhibitor [
16], and Kim et al. reported that Gly attenuated brain injury by blocking inflammation mediated by microglia/macrophages as an HMGB1 inhibitor [
7]. Nevertheless, whether HMGB1 regulates T cell activity and whether Gly modulates neuroinflammation via inhibiting HMGB1-mediated T cell activity have not been studied.
In this report, we first examined Gly’s protection in both wild-type rat and mouse stroke models, and then we used T cell-deficient nude rats and SCID mice to study whether T cells are the targets for the protective effects of Gly against stroke. The inhibitive effects of Gly on T cell activities were evaluated using in vitro splenocyte cultures, in which T cell proliferation is induced by ConA stimulation. We also studied the effect of Gly on CD4 T cell functional subsets Th1, Th2, and Treg, by examining gene expressions of IFNγ, IL-4, and IL-10, respectively. We then further examined whether Gly administration inhibited HMGB1 release after stroke and whether HMGB1 promoted T cell activity.
Methods
Focal cerebral ischemia
Anesthesia was induced with 5 % isoflurane and maintained by 1.5 to 3 % isoflurane in medical air and balanced with 1 to 20 % O
2 using a facemask, in Sprague-Dawley (SD) rats (Charles River, Wilmington, MA, USA) and T cell-deficient nude rats (NCI, Frederick, MD, USA) (280 to 320 g) or C57BL/6J wild-type mice (Jackson Laboratory, Bar Harbor, ME, USA) and SCID mice (#001913; B6.CB17-Prkdcscid/SzJ) (25 to 30 g). Rectal temperature was maintained at 37 ± 0.5 °C with a heating pad (Harvard Apparatus, Holliston, MA, USA). Focal ischemia was induced by 100 min of transient MCA suture occlusion in rats or 60 min of occlusion in mice, as described [
9,
17,
18]. In brief, we introduced a silicone-coated 4-0 (for rats) or 6-0 (for mice) monofilament (Doccol Corp., Redlands, CA, USA) into the left external carotid artery and advanced through the carotid bifurcation to occlude the MCA. The anesthesia was discontinued after suture insertion. The animals were briefly re-anesthetized after the desired occlusion time, and the filament was withdrawn. Sham-operated animals underwent an identical procedure, except that the suture was not inserted. Physiological measurements were performed as described before [
17,
19]. Heart rate, oxygen saturation, and respiratory rate were continuously monitored (STARR Life Sciences Corp., Allison Park, PA, USA). The left femoral artery was cannulated with a polyethylene tube for arterial blood pressure monitoring (Puritan-Bennett Corp., Wilmington, MA, USA). Arterial blood pH, PaO
2, and PaCO
2 were monitored with an i-STAT Portable Clinical Analyzer (East Windsor, NJ, USA) on blood drawn from either the femoral or carotid artery.
Measurement of cerebral infarction area
Infarct size was measured 72 h after stroke. Animals were anesthetized with isoflurane and decapitated, and the brains were removed and cut into five coronal slices with a 2-mm thickness with a rodent brain slicer matrix (Zivic Instruments, Pittsburgh, PA, USA). The sections were stained with 2 % 2,3,5-triphenyltetrazolium chloride (TTC) in saline for 20 min at 37 °C and transferred to a solution of 4 % paraformaldehyde in phosphate-buffered saline (PBS) overnight. The infarct area was analyzed by an observer, blinded to the experimental conditions, using the NIH Image program (ImageJ 1.37v, Wayne Rasband available through the National Institutes of Health) as described previously [
9,
17,
18]. To correct for the effects of cerebral edema, the infarction in each section was normalized to the non-ischemic contralateral side and expressed as percentage of the contralateral hemisphere.
Drug injection
Gly solution (30 mg/ml, #3567800; Calbiochem, San Diego, CA, USA) was prepared in Tris buffer (pH 7.5), and was injected intraperitoneally (i.p.) into animals. To prove the protective effect of Gly against stroke in rats, in the first group, Gly (200 mg/kg) was injected immediately before MCA occlusion and 2 h after reperfusion. In the second group, Gly was injected immediately post-reperfusion and 3.5 h after reperfusion. In the third group, Gly was injected only immediately post-reperfusion. In the control group, the vehicle solution containing no Gly was injected.
Confocal immunofluorescent staining
Ischemic or sham-operated rats were euthanized with an overdose of isoflurane and perfused with icy PBS, followed by 4 % paraformaldehyde in PBS (pH 7.4), as previously described [
9,
17,
18]. The ischemic brains were taken and post-fixed for 48 h in 4 % paraformaldehyde in PBS (pH 7.4) and cut into 50-μm sections. Immunofluorescent staining was carried out on free-floating sections under moderate shaking. All washes and incubations were done in 0.1 M PBS (pH 7.4) containing 0.3 % Triton X-100. The sections were incubated for 1 h with blocking solution (0.1 M PBS, 0.3 % Triton X-100, and 5 % equine serum). After washing, the sections were incubated overnight at 4 °C with primary antibodies.
For double staining of HMGB1 and NeuN proteins, the sections were incubated with a mixture of primary anti-HMGB1 antibody (1:100, ab18256; Abcam Inc, Cambridge, MA, USA) and anti-NeuN antibody (diluted 1:100, MAB377; Millipore, Hayward, CA, USA). Activated microglia/macrophages were stained with mouse anti-rat CD68 antibody (1:200, MCA341GA; AbD Serotec, Oxford, England), and neutrophils were stained with a rabbit anti-human myeloperoxidase (MPO) antibody (diluted 1:50, #A0398; DAKO, Carpinteria, CA, USA). The sections were then washed and incubated for 2 h at room temperature with an Alexa 594-conjugated goat anti-rabbit antibody (1:200 for HMGB1 or MPO) or an Alexa 488-conjugated goat anti-mouse antibody (for CD68 or NeuN, 1:200; Invitrogen, Carlsbad, CA, USA). After washing, the sections were mounted on glass slides using Vectashield mounting medium with 4′,6-diamidino-2-phenylindole (DAPI; Vector Laboratories, Burlingame, CA, USA). A negative control without primary antibodies was performed in parallel. Images were taken with a Zeiss Axiovert inverted fluorescent microscope (Jena, Germany).
To quantify the number of CD68- or MPO-positive cells, three sections (0.20 to 0.70 mm rostral to the bregma) in each animal were randomly chosen, from which the number of CD68- or MPO-positive cells in predefined areas were counted using ImageJ software (NIH) and an average number for each animal was calculated. All counts were performed by an investigator blinded to the coded sections.
Western blot
The rats were euthanized 72 h after reperfusion with an overdose of isoflurane. CSF was collected and Western blot was performed as described in our previous study [
20]. In each lane, 12 μl CSF mixed with 4 μl loading buffer (#NP0007; Invitrogen) was subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis using NuPAGE 4 to 12 % Bistris Gel (#NP0323; Invitrogen) for 1.5 h. Protein bands were transferred from the gel to polyvinylidene fluoride membranes (Millipore) for 1 h. After the membranes were blocked with 5 % non-fat dry milk (Bio-Rad Laboratories, Hercules, CA, USA) in PBS/0.05 % Tween-20, a primary anti-HMGB1 antibody was added (diluted 1:500, ab18256; Abcam) and incubated overnight at 4 °C followed by the addition of a horseradish peroxidase-conjugated secondary antibody (1:2000; Cell Signaling Technology, Danvers, MA, USA) for 1 h. The membranes were scanned using Typhoon trio (GE Healthcare, Pittsburgh, PA, USA).
Serum collection and HMGB1 ELISA
Blood was collected, without anticoagulant, from the rats by cardiac puncture through the right atrium with a syringe and 18-gauge needle. The blood was placed into a 1.5-ml conical tube and left at room temperature for 20 min to allow the blood to clot. Next, the tubes were centrifuged at 5000×g for 10 min at room temperature. Afterwards, the serum was transferred to another 1.5-ml conical tube and stored at −80 °C. The serum samples were examined for HMGB1 content using an ELISA kit (# ST51011; IBL International, Hamburg, Germany).
The effect of Gly on splenocyte proliferation
The [3H]-thymidine incorporation assay was used to determine splenocyte proliferation. A total of 2 × 105 cells were seeded in each well of standard 96-well flat-bottom plates with or without 1 μg/ml ConA (Sigma-Aldrich, St. Louis, MO, USA). The cells were incubated for 66 h at 37 °C with 5 % CO2. Then, 0.5 μCi [3H]-thymidine (Perkin-Elmer, Waltham, MA, USA) was added to each well and incubated for another 18 h. The cells were harvested onto a glass fiber filter. The filter membrane was dried and the amount of radioactivity that corresponds to the number of divided cells during incubation was counted in a β-plate scintillation reader.
To study whether Gly inhibits splenocyte proliferation in vitro, splenocytes were harvested from naïve animals and cultured. A total of 2 × 105 cells were seeded in each well of standard 96-well flat-bottom plates with or without 1 μg/ml ConA. Gly was added to the wells to make final concentrations of 0, 12.5, 25, 50, 100, 200, 400, and 800 μM. The cells were incubated for 66 h at 37 °C with 5 % CO2. Then, 0.5 μCi [3H]-thymidine was added to each well and incubated for another 18 h. The cells were harvested onto a glass fiber filter. The filter membrane was dried, and the amount of radioactivity that corresponds to the number of divided cells during incubation was counted in a β-plate scintillation reader.
An ex vivo assay was also conducted to analyze the effects of Gly on splenocyte proliferation. Gly (200 mg/kg) was injected twice at 3.5-h intervals into SD rats, who did not receive stroke. The animals were euthanized at 0, 5, 12, 24, 48, and 72 h after the last injection, and splenocytes were prepared for culture. The cells were stimulated with or without ConA, and cell proliferation was measured as described above.
Cytokine gene expression in functional subsets of CD4 T cells
To detect the functional CD4 subsets (Th1, Th2, and Treg) that are affected by Gly treatment, mice were i.p. injected with Gly (200 mg/kg) at 0 and 3.5 h after reperfusion with 60 min of MCAo. Leukocytes were extracted from the ischemic brain 3 days after stroke, as described above, and then stimulated with Cell Stimulation Cocktail (500X, eBioscience, Cat# 00-4970-93) for 2 1/2 h, to produce IFN-γ, IL-4, and IL-10 gene expressions, which represent the activities of Th1, Th2, and Treg cells, respectively. The cells were washed with FACS buffer and blocked by Fc for 20 min on ice, and then stained by live/dead Aqua (Thermofisher, Cat# L34965), Ly6G (clone: 1A8, Biolegend, Cat#1272639), CD4 (clone: GK1.5, Biolegend, Cat#100421), and CD3 (clone: 17A2, Biolegend, Cat#100205) for 30 min on ice. CD4 T cells were then sorted by FACS sorter into tubes. From these purified CD4 T cells, messenger RNA (mRNA) were extracted using the RNeasy mini kit (QIAGEN, Cat#74104) and reversely transcripted into complementary DNA (cDNA) (Takara, Cat# RR047A). SYBR green master mix (Invitrogen, Cat# 11762-100) was used in real-time PCR. The following primers were used to determine IL-4, IL-10, and IFN-g gene expressions: GAPDH, forward (F): 5′ TTCACCACCATGGAGAAGGC 3′, reverse (R):5′ GGCATGGACTGTGGTCATGA 3′; IL-4, F: 5′ ACAGGAGAAGGGACGCCAT 3′; R: 5′ GAAGCCCTACAGACGAGCTCA 3′; IFNγ, F: 5′ TCAAGTGGCATAGATGTGGAAGAA 3′; R: 5′ TGGCTCTGCAGGATTTTCATG 3′; IL-10, F: 5′ GGTTGCCAAGCCTTATCGGA 3′; R: 5′ ACCTGCTCCACTGCCTTGCT 3′. All the primers used in this study have been validated in a previous study [
21].
The effect of HMGB1 on splenocyte proliferation
To test whether HMGB1 has direct effects on splenocyte proliferation, a total of 2 × 105 cells were seeded in each well of standard 96-well flat-bottom plates. We then added recombined human HMGB1 (rhHMGB1 0, 62.5, 125, 250, 500, 1000 ng/ml) to cultured splenocytes and stimulated them with or without 1 μg/ml ConA. The cells were incubated for 66 h at 37 °C with 5 % CO2. The effects of rhHMGB1 on splenocyte proliferation were measured using a Cell Counting Kit-8 (CK-04-11; Dojindo Molecular Technologies, Rockville, MD, USA; n = 5), and repeated twice.
To further study the effects of Gly on splenocyte proliferation in the presence of rhHMGB1, splenocytes were cultured and stimulated with ConA, with or without adding rhHMGB1 (500 ng/ml). A series of Gly concentrations of 0, 6.25, 12.5, 25, 50, 100, 200, and 400 μM was added, and then cell proliferations were determined as described above.
FACS analyses of immune cells in the brains
Mice were deeply anesthetized with isoflurane 72 h after reperfusion. Ischemic ipsilateral brains were dissected after transcardial perfusion with 100 ml of PBS. Brain leukocytes were collected as described previously [
22]. In brief, ipsilateral hemispheres were homogenized and filtered through a 70-μm cell strainer. After centrifugation, the cells were resuspended in 7 ml FACS buffer, completely mixed with 3 ml of 90 % Percoll (GE Healthcare), and then 1 ml of 70 % Percoll was loaded under cell suspension. The cell suspension was then centrifuged at 500 g for 30 min at 4 °C. Leukocytes at the interphase were collected for immunostaining and FACS analysis. These cells were stained with differentially fluorochrome-labeled mAbs against TCR, CD4, and CD8 (AbD Serotec) on ice for 30 min to identify T cells, CD4+ T cells, and CD8+ T cells. Data on stained samples were acquired on a BD LSR II flow cytometer using Diva software (v6.1.2; Becton Dickinson, San Jose, CA, USA) and analyzed using FlowJo software (v7.6.2; Tree Star, Ashland, OR, USA).
Statistical analysis
Data are expressed as mean ± SEM. All data tested met the normal distribution with one-sample Kolmogorov-Smirnov test by using SPSS 16.0. Asymptotic significance for P > 0.05 was considered to meet normal distribution. Differences were considered statistically significant for P values < 0.05. Student t test was used when two-group comparisons were made. Two-way ANOVA was used when both genotype and Gly treatment were taken into account, followed by Bonferroni post-tests using Prism 5 (Software for Science, San Diego, CA, USA).
Discussion
Although a few studies have shown that Gly has solid neuroprotective effects against cerebral ischemia, the underlying protective mechanisms remain elusive. In this study, after confirming the protective effects of Gly in both rats and mice, we report for the first time that Gly reduces brain infarction by inhibiting T cell activity associated with HMGB1 release. This is supported by several lines of evidence. First, Gly inhibited infarction in both wild-type rats and mice, but not in nude rats and SCID mice. This was further confirmed by the protective effects of Gly in an in vitro co-culture of splenocytes and neurons, suggesting that T cells are the targets of Gly protection against stroke. Second, T cell proliferation was inhibited when Gly was injected in vivo and when Gly was added to cultured splenocytes, as evaluated by an in vitro assay. Third, Gly administration inhibited gene expressions of IFNγ, but not IL-4 and IL-10 in purified CD4 T cells. Fourth, Gly injection in vivo inhibited HMGB1 release in CSF and serum after stroke, and an in vitro assay showed that Gly inhibited T cell proliferation stimulated by HMGB1. Taken together, it is plausible that Gly inhibits brain injury by inhibiting T cell activity, which is at least partly mediated by HMGB1 release after stroke.
A few recent studies have reported that Gly protects against brain injury induced by cerebral ischemia [
23‐
27]. Gly inhibits pro-inflammatory responses, including microglial activation [
7], TLR cell signaling pathways [
23], pro-inflammatory cytokines, free radical enzymes, and transcription factors, such as IL-1, TNF-α, COX-2, iNOS, NF-kB, and HMGB1 release [
23‐
27]. Despite these studies, whether T cells are directly involved in the protective effect of Gly against stroke has not been studied. T cells have been demonstrated to contribute to brain injury induced by stroke in mice [
8,
28‐
30]. T cells, like macrophages and neutrophils, were found to infiltrate the ischemic brain after stroke [
31], suggesting that T cells may form physical contact with brain tissue or secrete inflammatory factors to destroy neurons. In addition, infarct sizes were robustly reduced in immunodeficient SCID mice lacking both T cells and B cells and that with the reconstitution of T cells, but not of B cells, increased infarction [
8]. More recent studies suggest that deficiency in either CD4 or CD8 T cells results in smaller infarct size [
28,
29], while deficiency in regulatory T cells (Treg) leads to enlarged, delayed infarct sizes [
30]. Our laboratory not only confirmed the effects of T cells and CD4 and CD8 T cells, but also further showed the distinctive roles of CD4 T cell subtypes in mice, with reduced infarction in Th1-deficient mice and increased infarction in Th2-deficient mice, while Treg deficiency had no effects on acute infarction [
18]. Despite different T cell subsets having distinct effects, total T cells contribute to brain injury.
For the first time, we provide solid evidence that T cells are the target of the protective effect of Gly against stroke. Initially, we found that the robust protective effect of Gly on infarction and inflammation is related to the reduced number of infiltrating T cells and CD4 and CD8 T cells, suggesting that T cells may be one of the targets of Gly administration. Indeed, we found that Gly had no protective effect in nude rats and SCID mice who are T cell-deficient, but that Gly gained a protective effect in these animals when T cells were reconstituted. Similar to these in vivo results, wild-type splenocytes co-cultured with neurons resulted in more neuronal death than SCID splenocytes; Gly attenuated neuronal death in this co-culture, but it had no protection in neurons co-cultured with T cell-deficient splenocytes. Thus, both in vivo and in vitro data indicate that the presence of T cells is essential for the protective effect of Gly. Furthermore, we provided direct evidence that Gly blocks T cell proliferation. An ex vivo assay showed that T cell proliferation in cultured splenocytes harvested from animals treated with Gly was inhibited compared with vehicle injection. In vitro assays showed that the addition of Gly to splenocyte cultures also blocked T cell proliferation. Lastly, we showed that Gly administration inhibited IFNγ gene expression, but not IL-4, and IL-10 gene expression in purified CD4 T cells. Thus, it is likely that Gly inhibited Th1 functional subsets, but not Th2 and Treg cells. Taken together, it is plausible that Gly blocks brain injury by inhibiting T cell functions.
Although T cells are involved in brain injury, how they are recruited and activated after stroke remains unknown. We provide evidence that T cell activity can be induced by HMGB1, as rhHMGB1 promoted T cell proliferation in cultured splenocytes. We further prove that Gly inhibited HMGB1-modulated T cell proliferation in cultured splenocytes, suggesting the protective effects of Gly are at least partly dependent on their inhibitive effects on HMGB1-mediated T cell functions. Our in vivo studies further show that Gly attenuated HMGB1 release from the ischemic brain and attenuated its levels in CSF and serum. Taken together, inhibition of HMGB1 release and T cell function may be a major reason for the protective effects of Gly against stroke.
Our central finding in this study is that Gly reduces brain infarction by inhibiting T cell function, which has significant values for clinical translation. Most neuroprotectants in previous studies were designed to target pathological cascades occurring in the ischemic brain, such as glutamate antagonist, free radical scavengers, and apoptotic and necrotic cell signaling pathways. However, our current study suggests that T cells in the peripheral circulation are a therapeutic target for Gly administration. This provides a rationale to facilitate its clinical translation.
There are some limitations in this study. In Fig.
4, nude rats and SCID mice were reconstituted with splenocytes but not purified T cells. Thus, a question remains, whether other transferred cells, such as monocytes/macrophages, and NK cells, have contributed to the protective effect of Gly treatment. Indeed, we cannot exclude a possibility that Gly also inhibits brain injury by blocking macrophage activities, as our additional preliminary studies showed that Gly had a trend of inhibiting macrophage activities (data not shown). It is known that T cells cross talk with macrophages, whether Gly blocks brain injury by inhibiting both T cells and macrophages will be studied in the future.