Increased permeability of blood–brain barrier is mediated by serine protease during Cryptococcus meningitis

Abstract

Objectives

To determine the role of serine protease in the disruption of the blood–brain barrier (BBB) during Cryptococcus neoformans meningitis.

Methods

Reverse transcription–polymerase chain reaction and immunohistochemistry were used to determine the production of serine protease by different strains of C. neoformans. BBB permeability in immunosuppressed rats inoculated with C. neoformans or C. neoformans plus aprotinin was examined via Evans blue staining. In vitro BBB permeability (transwell passage of horseradish peroxidase) was determined in human brain microvascular endothelial cells (BMECs) cultured with serine protease or serine protease plus aprotinin. Electron microscopy of rat brain tissue was used to visualise C. neoformans infection.

Results

Serine protease mRNA and protein were detected in all C. neoformans serotypes. C. neoformans infection increased BBB permeability in vivo, but this effect was ameliorated by aprotinin. Treatment of BMECs with serine protease increased permeability in vitro. This effect was reversed by aprotinin.

Conclusion

Serine protease secreted by C. neoformans leads to BBB disruption during Cryptococcus meningitis. Serine protease may be a novel treatment target for Cryptococcus meningitis.

Introduction

The encapsulated fungus Cryptococcus neoformans is commonly found in soil. It causes the third most frequent opportunistic infection of the central nervous system (CNS) in patients with AIDS.^1–3 Between 6% and 10% of patients with AIDS will develop Cryptococcus meningitis,^4,5 which remains a leading cause of death in cohorts of HIV-infected individuals in Africa and Asia. ⁶ Traversal of the blood–brain barrier (BBB) by C. neoformans plays a key role in its pathogenesis, ⁷ and this is mediated by systems including transcytosis and a phagocytosis-mediated mechanism.^8,9 C. neoformans infection may also contribute to disruption of the BBB itself, ¹⁰ but the mechanisms remain unclear.

Secreted proteases, such as serine protease, may have important roles in the pathogenesis of fungal diseases,^11–13 possibly via the activation of matrix metalloproteinase (MMP) and subsequent collagen type 1 proteolysis and capillary tubular network collapse.^14,15 Serine proteases (kallikrein, trypsin, cathepsin G, etc.) have also been shown to directly activate MMPs that are capable of degrading various basement membrane components. ¹⁶

The present study used reverse transcription–polymerase chain reaction (RT–PCR) to determine the expression level of serine protease in several isolates of C. neoformans, and examined the effects of serine protease on BBB permeability in vitro and in vivo.

Materials and methods

Cell lines and culture

The China Medical Mycology Culture Collection Centre (Nanjing, Jiangshu Province, China) provided serotype A, serotype B and serotype D/AD C. neoformans isolates. Cells were grown in yeast morphology broth (Difco Laboratories, Detroit, MI, USA) at 35℃ in a rotary shaker at 280 rpm. Human brain microvascular endothelial cells (BMECs) were provided by the Chinese Academy of Sciences (Shanghai, China) and cultured at 37℃ in RPMI 1640 medium containing 10% fetal bovine serum.

Animals

Male Sprague–Dawley rats (n = 23, weight 180–220 g, age 6–8 weeks for inoculation studies; n = 14, age 1–20 days for BMEC studies) were purchased from the animal centre of the Second Military Medical University, Shanghai, China. Animals were kept in controlled temperature (23–25℃) and lighting (12 h light–dark cycle, lights on at 08:00 hours) conditions, with free access to water and food. The study was approved by the ethics committee of the Second Military Medical University.

Immunohistochemistry

The presence of serine protease in C. neoformans strains was analysed by immunohistochemistry of cells cultured on coverslips. Briefly, slides were incubated for 30 min at room temperature with goat anti-rabbit serine protease antibody (1:200 dilution; Boshide Co., Wuhan, China), washed three times using phosphate buffered saline (PBS; pH 7.4) at room temperature, then incubated with peroxidase conjugated mouse anti-goat secondary antibody (1:1000 dilution; Boshide Co.) for 30 min at room temperature. Slides were then washed three times with PBS, and staining was visualized using EnVision™+ (Dako, Glostrup, Denmark). The slides were examined by two pathologists (H.-M.Z. and J.-H.W.).

Total RNA isolation

For RNA isolation, cells (9 × 10⁹) were harvested by centrifugation at 1000  g for 5 min, washed in RNase-free ice-cold water, and pelleted by centrifugation at 1000  g for 5 min. The supernatant fraction was discarded and cell walls were mechanically disrupted by vortexing for several minutes with 0.45–0.55-mm glass beads. Total RNA was extracted using a Yeast RNA extraction Kit (Omega, Carlsbad, CA, USA) according to the manufacturer’s instructions. RNA concentration and purity were determined spectrophotometrically, and RNA integrity was visualized by electrophoresis through a denaturing gel.

First-strand cDNA was generated using a reverse transcription kit (Qiagen, Hamburg, Germany) and PCR was performed using a Taq DNA polymerase kit (Invitrogen, Carlsbad, CA, USA). The cycling programme involved preliminary denaturation at 95℃ for 3 min, followed by 40 cycles of denaturation at 95℃ for 5 s, annealing at 55℃ for 10 s, and elongation at 72℃ for 15 s, followed by a final elongation step at 76℃ for 5 min. Primer sequences were: C. neoformans serine protease, forward, 5′-AACTCCACCACAAACACCA-3′ and reverse, 5′-GGAAAGACTCAGTCCCGTAA-3′; β-actin (internal control), forward, 5′-GCCCTTGCTCCTTCTTCTAT-3′ and reverse, 5′-GACGATTGAGGGACCAGACT-3′ (all Sheng Gong Co., Shanghai, China). The specificity of the PCR was verified by subjecting amplification products to agarose gel electrophoresis.

Rat BMEC isolation

Rat brain capillaries were isolated as described, ¹⁷ with modifications. Briefly, fresh brains were obtained from 10- to 20-day-old rats, and the cerebellum, brain stem, choroid plexus and meninges were carefully dissected. Cortices were cut into 1–2 mm³ fragments in D-Hank’s solution containing 5% fetal bovine serum (FBS; Hyclone, UT, USA), then centrifuged at 1000  g for 5 min. The pellet containing crude microvessels was further digested in a solution containing 0.25% trypsin for 1 h at 37℃. BMECs were isolated by centrifugation at 2000  g for 15 min at 4℃ in 15% dextran. The supernatant fractions were then removed, the pellets were resuspended in 33% Percoll® (Sigma, St Louis, MO, USA) solution and centrifuged at 1000  g for 10 min at 4℃. BMECs (1 × 10⁹/ml) were cultured in 24-well plates in Dulbecco’s modified Eagle’s medium (DMEM) containing1 ng/µl platelet-derived endothelial growth factor (Sigma-Aldrich, St Louis, MO, USA), 20% FBS, 12.5 mg/ml heparin, 100 U/ml penicillin and 100 µg/ml streptomycin. Cultures were incubated at 37℃ in a humid atmosphere (100%) of 5% carbon dioxide in air.

BMEC identification

Morphological and immunocytochemical analyses of BMEC monolayers were performed with cells grown to confluency on collagen-coated coverslips. Morphology was examined using an Olympus IX 70 microscope (Tokyo, Japan) with phase contrast. Immunocytochemical analysis of factor VIII was used as a BMEC-specific marker. ¹⁸ Coverslips were washed with Hank’s balanced salt solution (pH 7.2; Beyotime, Hangzhou, China), fixed in cold acetone–methanol (1:1 vol/vol) for 15 min, then air dried, sealed, and stored at −20℃. Cells were rehydrated, washed with PBS (pH 7.4) containing 0.1% bovine serum albumin (BSA) and 0.01% Tween and blocked in 10% normal goat serum for 15 min at room temperature. Coverslips were incubated with rabbit anti-rat factor VIII (1:200 dilution; Sigma-Aldrich) for 1 h at room temperature, washed three times with PBS–BSA–Tween at room temperature (5 min each wash), then incubated with peroxidase-labelled goat anti-rabbit secondary antibody (1:500 dilution; Sigma-Aldrich) for 30 min at room temperature. Coverslips were then washed with PBS–BSA–Tween for 5 min at room temperature and mounted on slides with glycerol or aquamount. Specimens were viewed via Diaphot fluorescence microscope (Nikon, Tokyo, Japan) equipped with a standard fluorescein isothiocyanate filter combination. Primary and/or secondary antibodies were omitted for controls. Immunopositivity was analysed using the Allred scoring system and examined by two pathologists (H.-M.Z. and J.-H.W.). ¹⁹

In vitro experiments

In order to evaluate the response of rat BMECs to serine protease, cells were cultured to confluency and then incubated for 30 min with (i) DMEM (as above), (ii) DMEM with 0.2 µg/ml serine protease, or (iii) DMEM with 0.2 µg/ml serine protease plus 10 nM aprotinin (Sigma-Aldrich).

The passage of horseradish peroxidase (HRP) through confluent human BMEC monolayers was assessed via transwell cell culture chambers (polycarbonate filters, pore size 3.0 µm; Costar, Cambridge, MA, USA). ²⁰ Human BMECs were incubated with (i) RPMI, (ii) RPMI with 0.2 µg/ml serine protease or (iii) RPMI with 0.2 µg/ml serine protease plus 10 nM aprotinin. Human BMECs were seeded at 2 ×10⁴ cells/filter in 200 µl RPMI 1640 medium, and the lower chamber was filled with the same medium (800 µl). Cells were cultured for 3–4 days to attain confluence; the medium in the upper compartment was then carefully removed and replaced with 200 µl of fresh medium containing 0.1 µg HRP. At this point, the lower compartment was also refilled with fresh medium. At 1, 2, 3 and 4 h, aliquots of 5 µl were taken from the lower compartment and HRP concentration was determined spectrophotometrically at 470 nm.

In vivo experiments

Male Sprague–Dawley rats (n = 23) were immunosuppressed with a subcutaneous injection of triamcinolone acetonide 20 mg/kg per day for 3 days (Steris Laboratories, Inc., Phoenix, AZ, USA). After 3 days, neutropenia was induced with intraperitoneal injection of cyclophosphamide 300 mg/kg (Pharmacia Inc., Kalamazoo, MI, USA). After a further 24 h, rats were sedated with 44 mg/kg diazepam and ketamine HCl (Fort Dodge Laboratories Inc., Fort Dodge, IA, USA) and 0.04 mg/kg atropine (Elkin-Sinn, Inc., Cherry Hill, NJ, USA) via intramuscular injection. Rats were then injected with 1.2 ml 3% Evan’s Blue dye via the right femoral vein. After 10 min, animals were divided into three groups: (i) inoculated with 2 × 10⁶C. neoformans serotype B cells (total volume 100 µl) via intravenous injection; (ii) inoculated with C. neoformans plus aprotinin (0.8 mU); or (iii) control animals (n = 7 per group). The remaining two animals were used for electron microscopy studies (below). Animals were killed 12 h after infection, perfused with 500 ml normal saline, and brains were removed and frozen at −20℃ for 4–5 h. Cryostat sections (30 µm) were placed on gelatine-coated glass slides and dried in containers containing silica gel. The presence of Evan’s blue in the CNS was visualized using a Nikon Eclipse 800 fluorescence microscope with an excitation wavelength of 620 nm. Slides were examined by two pathologists (H.-M.Z. and J.-H.W.).

Electron microscopy

Rats infected with C. neoformans serotype B (as above; n = 2) were killed 12 h after infection and perfused intracardially with 200 ml Sörensen buffer followed by 300 ml 3% paraformaldehyde/1.5% glutaraldehyde solution. Brain tissue close to the meninges was dissected and fixed in 1.5% glutaraldehyde in PBS for several days at 4℃. After rinsing with PBS, the blocks were osmicated in 2% osmic acid for 1 h at 4℃, dehydrated, impregnated with isoamyl acetate, and embedded in a mixture of araldite and epoxyresin (Serva, Heidelberg, Germany).

Ultrathin sections (4 µm) were cut with an OMU III ultramicrotome (Leica, Mannheim, Germany), stained with uranyl acetate and lead citrate, and examined with a CM100 B10 twin Electron microscope (Philips, Eindhoven, The Netherlands). Sections were examined by two pathologists (H.-M.Z. and J.-H.W.).

Statistical analyses

Data were expressed as mean ± SEM and compared using one-way analysis of variance followed by Dunnett’s T3 post hoc test, or the χ²-test. Statistical analyses were performed with SPSS® version 19.0 (SPSS Inc., Chicago, IL, USA) for Windows®. P-values < 0.05 were considered statistically significant.

Results

Representative photomicrographs of immunohistochemistry for serine protease in C. neoformans serotypes are shown in Figure 1A. Serine protease mRNA levels were significantly higher in serotype B than serotype A or D/AD (P < 0.05 for each comparison; Figure 1). Strong expression of serine protease was higher in serotype B than serotype A or D/AD (Table 1).

Figure 1.

A: Representative light photomicrograph of immunocytochemistry for serine protease (brown staining) in Cryptococcus neoformans monolayers. Left panel: weak staining (serotype A). Right panel: strong staining (serotype B). Original magnification × 1000. B: Relative levels of serine protease mRNA in different C. neoformans serotypes, as assessed by reverse transcription–polymerase chain reaction. Data are mean ± SE of three independent experiments. *P < 0.05 versus A and D/AD.

Table 1.

Expression of serine protease in different serotypes.

Serotype	n	Weak expression	Strong expression
A	13	7 (53.8%)	6 (46.2%)
B	13	1 (7.7%)	12 (92.3%)
D and AD	6	2 (33.3%)	4 (66.7%)

Differences between groups were evaluated with the χ²-test. The percentage of strong expression of serine protease in serotype B was significantly higher than that in serotype A (Fisher’s exact significance P = 0.03). There was no significant difference between serotype B and serotype D and AD (Fisher’s exact significance P = 0.22).

C. neoformans-infected rats showed leakage of Evans blue dye across the BBB (Figure 2A, centre panel) that was ameliorated by injection of aprotinin simultaneous to infection (Figure 2A, right panel). There was no dye leakage in noninfected control animals (Figure 2A, left panel).

Figure 2.

A: Representative immunofluorescence photomicrographs of rat brain tissue stained with Evans blue. Left panel: brain tissue from control animals showing no dye leakage. Centre panel: brain tissue from Cryptococcus neoformans-infected animals showing obvious dye leakage. Right panel: brain tissue from C. neoformans-infected rats coinoculated with aprotinin showing little dye leakage. Original magnification × 180. B, C: Horseradish peroxidase (HRP) activity in the lower chamber of a transwell culture system containing human BMECs cultured alone, with C. neoformans/serine protease, or with C. neoformans plus aprotinin/serine protease plus aprotinin. Data are mean ± SEM of triplicate experiments; *P < 0.05 versus both other groups.

Treatment of human BMEC monolayers with the culture supernatant fraction from C. neoformans culture or serine protease significantly increased their permeability to HRP compared with control cultures at 2, 3 and 4 h after treatment (P < 0.05 for each comparison; Figure 2B, 2C). Co-culture with aprotinin reversed this effect, with no significant difference in HRP clearance from control cultures at any timepoint (Figure 2B, 2C).

Discussion

It has been shown that C. neoformans secretes proteases that alter BBB integrity. For example, C. neoformans-secreted urease promotes microvascular sequestration, thereby enhancing CNS invasion. ²¹ In addition, extracellular phospholipases were found to play an important role in cryptococcosis dissemination in a murine model. ²² The integrity of the BBB can be perturbed by serine proteases such as fibrinolytic enzymes. ²³ It is therefore possible that BBB breakdown during Cryptococcus meningitis may be mediated by C. neoformans-secreted serine protease.

The present study used RT–PCR and immunohistochemistry to verify serine protease production by C. neoformans. Elevated BBB permeability (Evans blue dye leakage) was found in immunosuppressed rats infected with C. neoformans; this effect was largely ameliorated by simultaneous injection with aprotinin – a serine protease inhibitor. Meanwhile, the present study showed increased HRP clearance across human microvascular endothelial cell monolayers cultured with C. neoformans/serine protease compared with control cultures and those incubated with C. neoformans/serine protease and aprotinin. Taken together, these findings suggest a specific role for serine protease in Cryptococcus meningitis, but the exact mechanism remains to be elucidated.

This study has established a clear connection between serine protease and BBB disruption. The increase in BBB permeability may be mediated via contraction of microvascular endothelial cells. Serine protease may be a novel treatment target for Cryptococcus meningitis.