Decreased progranulin levels in patients and rats with subarachnoid hemorrhage: a potential role in inhibiting inflammation by suppressing neutrophil recruitment

All procedures in the human study were approved by Jinling Hospital’s medical institutional review board and were all performed in accordance with the Declaration of Helsinki.

All experimental protocols including animal use and surgical procedures were approved by the Animal Care and Use Committee of Jinling Hospital and conformed to the Guide for the Care and Use of Laboratory Animals by the National Institutes of Health.

CSF was obtained from 43 SAH patients and 4 joint replacement patients (control group). The patients were divided into a control group (n = 4) and three SAH groups based on the time after SAH: 1–3 days (n = 16), 4–7 days (n = 18), and ≥8 days (n = 9). The injury severity scores including Hunt&Hess grade, Modified Fisher grades, World Federation of Neurological Surgeons (WFNS) Grading System, and Glasgow Coma Scale (GCS) were immediately recorded for SAH patients after admission to the hospital. The CSF was obtained through lumbar puncture and during surgery. After centrifugation (3000 g, 5 min), the supernatant of CSF was collected and stored at −80 °C. The clinical outcomes of these patients were evaluated at discharge according to Glasgow Outcome Scale (GOS) and Modified Rankin Scale (MRS). Detailed information was presented in Table 1.

Male Sprague–Dawley (SD) rats (280–320 g) were purchased from the Animal Center of Jinling Hospital, Jiangsu, China. The rats were raised under a 12-h light/dark cycle (25 ± 1 °C) and had free access to food and water at the Animal Center of Jinling Hospital.

Rats were randomly divided into six groups: sham, 12 h, and 1, 3, 5, and 7 days SAH groups, respectively. Rats were sacrificed for sample collection at the corresponding time points after SAH.

Based on PGRN results, we increased the levels of PGRN and sacrificed all animals at 24 h after SAH. Recombinant human PGRN (r-PGRN) (R&D Systems, Inc., Minneapolis, MN, USA) (r-PGRN in 5 μL of phosphate-buffered saline (PBS)) or an equal volume of PBS were administrated at 30 min after SAH through a single intracerebroventricular (i.c.v.) injection [24]. Animals were randomly divided into seven groups: sham, SAH, SAH + PBS, and SAH + PGRN (1, 3, 10, and 15 ng per rat). Initially, four concentrations were used to test the neuroprotective effects of r-PGRN on EBI after SAH. We chose these doses based on an ischemic stroke study [19]. Neurologic scores and brain water content were measured at 24 h after SAH. In the second set of experiments, which were based on our initial study, rats were treated with a dose of r-PGRN that gave the most marked effect for further studies. Six rats were used in each group.

Experimental SAH models were produced as reported previously [25]. In brief, the rats were anesthetized with chloral hydrate (0.4 mg/kg, IP, Jinling Hospital). The head and inguinal region was carefully shaved and disinfected. Rats were then fixed in a stereotaxic apparatus. A midline scalp incision was made and a 1-mm hole was drilled 8.0 mm anterior to Bregma in the midline. To prevent loss of CSF and bleeding from the midline vessels, we used bone wax to plug the burr hole. Rats were placed in the supine position and an insulin syringe (BD Science) was used to draw 300 μL of nonheparinized blood from the femoral artery. The needle was advanced 11 mm into the prechiasmatic cistern through the burr hole, at a 45° angle to the vertical plane, and then 300 μL of blood was injected into the prechiasmatic cistern over 20 s. An equal volume of normal saline was injected into the prechiasmatic cistern of rats in the sham group (Fig. 1). The burr hole was sealed with bone wax, and the incision was surgically sutured. Rats were kept at 30 °C, with their head placed in a downward position for 20 min. After recovery from anesthesia, the rats were returned to their cages and housed at 25 ± 1 °C. Rats that died during surgery or surgical recovery were excluded from the study, and the procedure was repeated until the final group size reached the planned experimental number.

The tail artery was cannulated to measure blood pressure (BP). For intracranial pressure (ICP) monitoring, a butterfly needle was percutaneously placed into cisterna magna through foramen magnum. And then the needle was connected to the tube of ICP monitoring device and transferred the pressure to the transducer [25, 26]. BP and ICP were recorded during blood injection and before the animals were killed.

We used three behavioral activity examinations (Table 2) to evaluate the neurologic function of rats at 24 h after SAH as previously described [27‐29]. Grading of neurologic deficits was as follows: severe neurologic deficit (scores = 4–6), moderate neurologic deficit (scores = 2–3), mild neurologic deficit (scores = 1), and no neurologic deficit (scores = 0).

Rat brains were removed at 24 h after SAH. The brains were weighed immediately to determine the wet weight and dried ar 80 °C for 72 h to determine the dry weight. The percentage of brain water content was calculated as follows: brain water content (percentage) = ((wet weight − dry weight) / wet weight) × 100 %.

We used Evans blue (EB) extravasation to assess BBB permeability at 24 h after SAH. As described previously [30], EB dye (2 %, 4 ml/kg) was injected intraperitoneally with a 3-h circulation time. Rats were anesthetized and perfused transcardially with 0.9 % normal saline solution (4 °C) to remove intravascular EB dye. Next, brains were removed and weighed. Brain samples were homogenized in physiological PBS (PH = 7.4) and centrifuged (15,000 g, 30 min). The resulting supernatant (0.5 mL) was added to an equal volume of 50 % trichloroacetic acid. After incubation overnight and centrifugation (15,000 g, 30 min, 4 °C), the supernatant was measured at 610 nm for spectrophotometric quantification. EB content was calculated as microgram per gram of protein.

Brains were harvested from rats after intracardiac perfusion with 0.9 % normal saline solution (4 °C) under anesthesia at 24 h after SAH. Brains that had obvious clots in the prechiasmatic cistern were selected for further analysis. After blood clots on the brain tissue were cleared, the temporal lobe tissue was harvested on ice and stored in −80 °C for protein extraction. For immunohistochemistry (IHC) and terminal deoxynucleotidyl transferase-mediated uridine 5’-triphosphate-biotin nick end-labeling (TUNEL), rats were perfused with 4 % buffered paraformaldehyde (4 °C) after 0.9 % normal saline solution (4 °C) perfusion, followed by immersion in 4 % buffered paraformaldehyde (4 °C).

To extract total protein, proper size of tissue was mechanically homogenized in 20 mM Tris (PH 7.6, 0.2 % SDS, 1 % Triton X-100, 1 % deoxycholate, 1 mM phenylmethylsulfonyl fluoride (PMSF), and 0.11 IU/ml aprotinin) (all from Sigma, St.Louis, MO, USA). Homogenate was centrifuged at 12,000 g for 15 min at 4 °C. The supernatant was collected and stored at −80 °C.

Equal amounts of protein were loaded in each lane of sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to a polyvinylidene-difluoride (PVDF) membrane. The membrane was blocked for nonspecific binding with 5 % defatted milk for 2 h at room temperature and incubated overnight at 4 °C with primary antibodies. For primary antibodies, we used goat anti-PGRN (1:200; sc-11342, Santa Cruz, CA, USA), rabbit anti-Myeloperoxidase (MPO) (1:200; sc-16128, Santa Cruz, CA, USA), rabbit anti-zonula occludens 1 (ZO-1) (1:200; sc-10804, Santa Cruz, CA, USA), rabbit anti-matrix metalloproteinase-9 (MMP-9) (1:5,000; ab38898, Abcam), rabbit anti-bcl-2 (1:200; sc-492, Santa Cruz, CA, USA), rabbit anti-cleaved caspase-3 (1:5,000; 9661, Cell Signaling, Beverly, MA, USA), and rabbit anti-β-actin (1:5,000; AP0060, Bioworld Technology, Minneapolis, MN, USA). After being washed with TBST (3 × 10 min), the membrane was incubated with rabbit anti-goat or goat anti-rabbit horseradish peroxidase (HRP)-conjugated IgG (1:5,000, BS30503 and BS13278, Bioworld Technology, Minneapolis, MN, USA) for 2 h at room temperature. Using the enhanced chemiluminescence reagent kit (Millipore Corporation, Billerica, MA, USA), bands were visualized. Quantification was performed by optical density methods using ImageJ software (NIH), and the data was normalized to β-actin.

Total protein was determined using a bicinchoninic acid assay kit (Pierce Biochemicals). The inflammatory cytokine levels of brain tissue were quantified using enzyme-linked immunosorbent assay (ELISA) kits specific for rats according to the manufacturer’s instructions (PGRN, P28799, from Raybiotech, USA; MPO, ab119605, from Abcam, USA; tumor necrosis factor-α (TNF-α), 950.090.096 and 865.000.096, from Diaclone Research, France; IL-1β, 850.006.096 and 670.040.096, from Diaclone Research, France). The detailed method is shown as follows: wash microwell strips twice with wash buffer. Standard dilution on the microwell plate is as follows: add 100 μl sample diluents, in duplicate, to all standard wells. Pipette 100 μl prepared standard into the first wells and create standard dilutions by transferring 100 μl from well to well. Discard 100 μl from the last wells. Alternatively, external standard dilution in tubes is as follows: pipette 100 μl of these standard dilutions in the microwell strips. Add 100 μl sample diluents, in duplicate, to the blank wells. Add 50 μl sample diluents to sample wells. Add 50 μl sample in duplicate, to designate sample wells. Cover microwell strips and incubate 2 h at room temperature. Empty and wash microwell strips five times with wash buffer. Add 100 μl Biotin—conjugate to all wells. Cover microwell strips and incubate 1 h at room temperature. Empty and wash microwell strips five times with wash buffer. Add 100 μl diluted streptavidin—HRP to all wells. Cover microwell strips and incubate 1 h at room temperature. Empty and wash microwell strips five times with wash buffer. Add 100 μl of TMB substrate solution to all wells. Incubate the microwell strips for about 10 min at room temperature. Add 100 μl stop solution to all wells. Blank microwell reader and measure color intensity at 450 nm. Inflammatory cytokine levels in the CSF of patients were calculated as pictogram per milliliter, and inflammatory cytokine levels in the brain cortex of rats were calculated as pictogram per milligram.

IHC was performed as previously described [31]. Rat brains, which had been fixed with 4 % buffered paraformaldehyde, were embedded in paraffin and cut into 10-μm slices. The sections were deparaffinized and incubated with 3 % H₂O₂ in PBS for 10 min. The sections were blocked with 5 % normal fetal bovine serum in PBS for 2 h followed by incubation with primary antibodies overnight at 4 °C. For primary antibodies, we used goat anti-PGRN antibody (1:200; sc-11342, Santa Cruz, CA, USA) and rabbit anti-MPO antibody (1:200; sc-16128, Santa Cruz, CA, USA). Each section was then incubated with rabbit anti-goat or goat anti-rabbit horseradish peroxidase (HRP)-conjugated IgG (1:500, BS30503 and BS13278, Bioworld Technology, Minneapolis, MN, USA) at room temperature for 60 min. Diaminobenzidine (DAB) and counterstaining were performed with hematoxylin. Positive cells were identified, counted, and analyzed under the light microscope by an investigator blinded to the groupings. Six random high power fields (400×) in each coronary section were selected, and the mean percentage of positive cells in the six fields was used for final analysis. A total of 10 sections from each sample were used for quantification.

According to our previous study [28], TUNEL staining was performed using an in situ cell death detection kit (Roche, Indianapolis, IN, USA). The severity of brain damage was evaluated by the apoptotic index, which was defined as the average percentage of TUNEL-positive cells. The TUNEL-positive cells were identified, counted, and analyzed under the light microscope by an investigator blinded to the groupings. Six random high power fields (400×) in each coronary section were selected, and the mean percentage of apoptotic neurons in the six fields was used for final analysis. A total of 10 sections from each sample were used for quantification.

PGRN levels as assessed by ELISA were summarized in Fig. 2a. High levels of PGRN were detected in the control group, while decreased levels were observed in the CSF of SAH patients (1–3 days, p < 0.05; 4–7 days and ≥8 days, p > 0.05). In addition, the levels of MPO, IL-1β and TNF-α were low in the control group and increased in the SAH groups (1–3 days, p < 0.05; 4–7 days and ≥8 days, p > 0.05) (Fig. 2b–d). The results showed that PGRN levels decreased after SAH, which may be associated with the levels of MPO, IL-1β, and TNF-α.

After SAH, no statistical difference in physiological parameters (BP and ICP) was observed among the groups (data not shown). There were 221 rats used in the present study and the mortality rates during and after surgery were as follows: (1) the sham group (0 of 6 rats), the SAH groups on 12 h (1 of 7 rats), 1 days (2 of 8 rats), 3 days (1 of 7 rats), 5 days (2 of 8 rats), and 7 days (2 of 8 rats); (2) the sham group (0 of 30 rats), the SAH group (6 of 36 rats), the SAH + PBS group (5 of 35 rats), the SAH + PGRN (1 ng per rat) (2 of 14 rats), the SAH + PGRN (3 ng per rat) (3 of 13 rats), the SAH + PGRN (10 ng per rat) (5 of 35 rats), and the SAH + PGRN (15 ng per rat) (2 of 14 rats).

The time course of PGRN protein levels were assessed by western blotting. PGRN protein levels decreased at 12 h after SAH with the lowest level observed at 24 h (Fig. 3a). According to the western blotting results, we selected the sham group and the 24-h group (lowest level of PGRN) for IHC. PGRN-positive cells markedly decreased at 24 h after SAH compared with the sham group (Fig. 3b).

To test the neuroprotective effects of r-PGRN on EBI, we evaluated the neurologic function and brain water content at 24 h after SAH. In this study, r-PGRN was given at four different concentrations to SAH rats. As shown in Fig. 4, treatment with r-PGRN at 3, 10, and 15 ng/rat dramatically ameliorated neurologic function (p < 0.01, p < 0.001) and brain edema (p < 0.05, p < 0.01). However, r-PGRN at 1 ng/rat had no significant effect. In addition, r-PGRN at a higher dose of 15 ng/rat did not exhibit a better neuroprotective effect than that at 10 ng/rat. These results suggested that r-PGRN could alleviate EBI and that r-PGRN at 10 ng/rat exhibited the most beneficial effect. Therefore, we selected 10 ng/rat for further experiments.

Western blotting (Fig. 5a) showed that the protein levels of MPO significantly increased in the SAH and SAH + PBS groups (p < 0.01) and decreased in the SAH + PGRN group (p < 0.05). IHC (Fig. 5b) revealed that MPO-positive cells in the SAH and SAH + PBS groups were significantly increased relative to that of the sham group (p < 0.001). Following r-PGRN treatment, the numbers of MPO-positive cells decreased (p < 0.01).

ELISA was used to measure the production of pro-inflammatory cytokines (TNF-α and IL-1β). As shown in Fig. 6, the concentrations of cytokines were significantly increased at 24 h after SAH (p < 0.001) and significantly reduced after treatment with r-PGRN (p < 0.01).

To demonstrate the effects of PGRN on brain edema, BBB permeability and protein levels of MMP-9 and ZO-1 were tested at 24 h after SAH. MMP-9 levels and ZO-1 degradation significantly increased in the SAH and SAH + PBS groups (p < 0.01; Fig. 7a, b) and were largely reduced after r-PGRN treatment (p < 0.05; Fig. 7a, b). As shown in Fig. 7c, EB indicated that BBB permeability increased after SAH (p < 0.001) and dramatically reduced after PGRN treatment (p < 0.01).

To illustrate the effects of PGRN on neural cell apoptosis, the levels of apoptosis-related proteins (bcl-2 and cleaved caspase-3) and extent of TUNEL were monitored at 24 h after SAH in our study. As shown in Fig. 8a, b, Bcl-2 levels decreased after SAH (p < 0.01) and were upregulated by r-PGRN treatment (p < 0.05). In contrast, cleaved caspase-3 levels increased after SAH (p < 0.01) and reduced after treatment with r-PGRN (p < 0.05). In addition, TUNEL (Fig. 8c) revealed that there were few TUNEL-positive cells in the sham group, whereas the apoptotic index in the SAH group was significantly higher than that of rats in the sham group (p < 0.001). The percentage of TUNEL-positive cells reduced after r-PGRN treatment (p < 0.01).