Increased brain damage after ischaemic stroke in mice lacking the chemokine receptor CCR5

Abstract

Background and purpose:

The chemokine receptor CCR5 is well known for its function in immune cells; however, it is also expressed in the brain, where its specific role remains to be elucidated. Because genetic factors may influence the risk of developing cerebral ischaemia or affect its clinical outcome, we have analysed the role of CCR5 in experimental stroke.

Experimental approach:

Permanent cerebral ischaemia was performed by occlusion of the middle cerebral artery in wild-type and CCR5-deficient mice. Locomotor behaviour, infarct size and histochemical alterations were analysed at different time points after occlusion.

Key results:

The cerebral vasculature was comparable in wild-type and CCR5-deficient mice. However, the size of the infarct and the motor deficits after occlusion were markedly increased in CCR5-deficient mice as compared with wild type. No differences between wild-type and CCR5-deficient mice were elicited by occlusion with respect to the morphology and abundance of astrocytes and microglia. Seven days after occlusion the majority of CCR5-deficient mice displayed neutrophil invasion in the infarct region, which was not observed in wild type. As compared with wild type, the infarct regions of CCR5-deficient mice were characterized by increased neuronal death.

Conclusions and implications:

Lack of CCR5 increased the severity of brain injury following occlusion of the middle cerebral artery. This is of particular interest with respect to the relatively frequent occurrence of CCR5 deficiency in the human population (1–2% of the Caucasian population) and the advent of CCR5 inhibitors as novel drugs.

Introduction

Chemokines are small proteins that play a key role in the regulation of leukocyte maturation and trafficking (Rossi and Zlotnik, 2000). However, chemokines and their corresponding receptors are also expressed in the CNS and regulate glial and neuronal cell functions (Cartier et al., 2005; Rostene et al., 2007). Because of its involvement in HIV entry into cells (Alkhatib et al., 1996), the chemokine receptor CCR5 (nomenclature follows Alexander et al., 2009) has become one of the most studied chemokine receptors and represents a new pharmacological target for HIV therapy. The CCR5 G-protein-coupled receptor is stimulated by three β-chemokines: CCL3/ macrophage inflammatory protein-1α (MIP-1α), CCL4/ MIP-1β and CCL5/regulated on activation, T cell expressed and secreted (Oppermann, 2004). Although constitutive CCR5 expression has been detected in astrocytes, microglia and neurons (Rottman et al., 1997; Westmoreland et al., 2002; Torres-Munoz et al., 2004), its role in healthy brain is still poorly understood. On the other hand, it has been reported that CCR5 is up-regulated after neurological insults (Galasso et al., 1998; Simpson et al., 2000) and might be involved in pathogenesis, in particular with respect to HIV-associated dementia (Kaul and Lipton, 2006).

Brain damage after stroke is the result of complex biochemical and cellular mechanisms. Indeed, following ischaemia, hypoxia-dependent necrosis occurs in the infarct core with irreversible consequences. Subsequently, several molecular pathways are activated and can be either protective or detrimental to the neighbouring regions that are functionally impaired, but potentially salvageable (Mehta et al., 2007). The evolution of such mechanisms determines the final extent of the infarct. The genetic background may also influence the clinical outcome of strokes. Indeed, genetic factors may either increase the risk of developing a stroke (Dichgans, 2007) or modulate the outcome in response to a given ischaemic lesion (Mallolas et al., 2006). There is evidence for increased CCR5 expression following focal cerebral ischaemia (Spleiss et al., 1998; Cowell et al., 2002; Kremlev et al., 2007). So far, it is not clear whether this up-regulation has a functional importance; however, this appears to be an important clinical question, as about 1–2% of the Caucasian population is homozygous for a 32 base-pair deletion in the coding sequence of the CCR5 gene (Martinson et al., 1997; Libert et al., 1998). This deletion impairs CCR5 conformation and leads to its degradation (Liu et al., 1996). The absence of CCR5 is detrimental in certain types of infection (Barr et al., 2005; Hardison et al., 2006; Khan et al., 2006) and can lead to severe complications in the CNS (Huffnagle et al., 1999; Glass et al., 2005; 2006; Thapa et al., 2007). In addition to the 32 base-pair deletion, numerous polymorphisms, influencing the activity of the CCR5 receptor, have been identified (Blanpain et al., 2000). Also, the recent introduction of CCR5 antagonists for the treatment of HIV (Emmelkamp and Rockstroh, 2007) requires an improved understanding of the potential negative effects of the lack of functional CCR5 (Stephenson, 2007).

In order to investigate whether CCR5 can influence the extent of brain damage induced by stroke, we analysed the effects of permanent middle cerebral artery occlusion (MCAo) in wild-type and CCR5 knockout mice. Our results show an enhancement of the post-ischaemic injury in CCR5-deficient mice and suggest a neuroprotective role of CCR5.

Methods

Mice

All animal care and experimental procedures were approved by the ethical committee of the University of Geneva and the Cantonal Veterinary Office. The CCR5-deficient mice in the C57BL/6 background (Kuziel et al., 2003) were kindly provided by Professor SS Ahuja (University of Texas at Austin, USA). Food and water were supplied ad libitum in a quiet room at 25°C with a 12 h light/dark cycle. Male mice of 8–10 weeks of age were used for MCAo.

MCA_o

Surgical operations were performed as previously described (De Bilbao et al., 2000). Briefly, the right temporo-parietal region of the head was shaved and a 2 mm incision was made vertically between the orbit and the ear. Under an operating microscope, a small burr hole (1 mm²) was made with a high-speed micro-drill to expose temporal vessels. After identification, the MCA was occluded by electrocoagulation using a small cauterizer without damaging the brain surface. The duration of the surgery did not exceed 15 min. Arterial blood pressure and blood parameters (pH, glucose, O₂ and CO₂ pressure) were analysed in randomly selected mice (n = 6 per group) before and after the MCAo surgery.

Wild-type and CCR5-deficient mice were killed at 2 and 7 days after MCAo and processed for histological analysis. After perfusion with a solution of 4% paraformaldehyde in phosphate-buffered saline (PBS) (pH 7.35), brains were removed, paraffin-embedded and cut in 10 µm thick sections.

Infarct analysis

Quantification of the infarct area was performed on cresyl violet-stained sections throughout the rostro-caudal extent of the lesion. For each brain slice, the infarcted area and the area of the whole section (total brain area in the right and left hemispheres) were measured using the Metamorph software (Molecular Devices, Sunnyvale, CA, USA). Infarct area was calculated as a percentage of the whole brain section analysed. Rostro-caudal extension of the infarct was measured from rostral pole, at the lateral septum level, to caudal pole, at the posterior thalamus/mesencephal level, corresponding to the brain territory covered by the MCA. Slides were coded so that the analysis was made blindly.

Grip test

Motor deficits were assessed by a grip test as previously described (Haddad et al., 2006) from 1 to 7 days after MCAo. Mice were picked up by the tail and placed on a taut string 60 cm long suspended 40 cm above a table. Grip score was measured as the length of time that mice remained on the string in some manner (using one or more paws, tail, tail plus paws) for a maximum of 30 s. Each experiment was conducted randomly and blindly in order to avoid uncontrolled influences.

Analyses of cerebral vasculature

Mice were anaesthetized and perfused transcardially with carbon black to visualize cerebral vasculature. After 10 min, animals were decapitated and heads stored in 4% paraformaldehyde for 48 h at 4°C. Brains were carefully removed and photographs were taken of the dorsal brain surfaces. Cerebral vasculature analysis was performed as previously described (Maeda et al., 1998). The distance from the inter-hemispheric fissure and the anterior cerebral artery and MCA anastomoses was measured at coronal planes 2, 5 and 7 mm from the frontal pole.

Semi-quantitative end-point PCR

Total RNA was isolated from homogenized tissues using RNeasy mini kit (Qiagen, Hombrechtikon, Switzerland) according to the manufacturer's instructions. Residual genomic DNA was removed using RNase-free DNase (deoxyribonuclease) set (Qiagen). Total RNA (1 µg) was reverse transcribed using the superscript II kit according to the manufacturer's instructions (Invitrogen, Basel, Switzerland). Semi-quantitative end-point PCR was performed by determining the suitable number of PCR cycles giving linear cDNA amplification of each gene of interest using Taq DNA polymerase (Qiagen). Amplification of the housekeeping gene encoding the L32 ribosomal protein was used as control. The list of the primers used is given in Table 1. PCR bands were scanned, and optical densities were analysed by ImageJ software (http://rsb.info.nih.gov/ij/).

Table 1.

Oligonucleotide probes and conditions used for semi-quantitative end-point RT-PCR

Gene	Primer sequences	Product size	Tm	# cycles
Ccr5	F: GCT GCC TAA ACC CTG TCA TC	199	60	32
	R: GTT CTC CTG TGG ATC GGG TA
Ccl3	F: ATG AAG GTC TCC ACC ACT GC	279	60	32
	R: CCC AGG TCT CTT TGG AGT CA
Ccl4	F: GCC CTC TCT CTC CTC TTG CT	196	60	32
	R: GTC TGC CTC TTT TGG TCA GG
Ccl5	F: CCC TCA CCA TCA TCC TCA CT	185	60	32
	R: CCT TCG AGT GAC AAA CAC GA
L32	F: GTG AAG CCC AAG ATC GTC AA	349	58	26
	R: TTG GTG ACT CTG ATG GCC AG

Immunohistochemistry

Brain sections were deparaffinized through graded alcohols, subjected to heat-induced epitope retrieval for 15 min in 0.01 M citrate buffer, pH 6.0, and incubated overnight at 4°C with primary antibodies diluted in PBS/0.3% Triton-X100 buffer at the following concentrations: myeloperoxidase 1:500 (MPO; Dako, Baar, Switzerland), glial fibrillary acidic protein 1:2000 (GFAP, Millipore, Billerica, MA, USA), ionized calcium binding adapter molecule-1 1:500 (IBA-1, Dako) and neuronal nuclei 1:1000 (NeuN, Millipore). Sections were then incubated for 1 h at room temperature with a biotinylated secondary antibody (Vector laboratories, Peterborough, UK). After several washes in PBS, sections were incubated with the ABC complex solution (Vector laboratories). Tissue-bound peroxidase was visualized using 3,3′-diaminobenzidine (DAB, Sigma-Aldrich, St. Louis, MO, USA) and H₂O₂. Counterstaining with cresyl violet was performed to visualize cell nuclei. Negative controls in the absence of the primary antibodies were performed for each experiment (data not shown). The surface occupied by GFAP, IBA-1 or MPO stained cells was measured at different coronal levels, using the Metamorph software (Molecular Devices): regions of interest (cortical infarct, neighbouring cortex and striatum, where cell density was increased for GFAP and IBA-1 immunoreactivity, and infarct core for MPO immunoreactivity) were designed according to stereotaxic coordinates (Franklin and Paxinos, 1997) and the pixels stained above threshold determined. Values are expressed as the ratio between the stained regions respect to the total area measured. The number of NeuN-positive cells was calculated as previously described (Schroeter et al., 2006) and using the Metamorph software (Molecular Devices): the number of neuronal cells (identified by NeuN staining) in the infarct core was automatically counted on digitally stored images and defined as the number of NeuN-stained cells for tissue area (neuronal density ipsi). For each brain slide, the neuronal density was similarly calculated in the corresponding brain area of the contralateral hemisphere (neuronal density contra), defined following stereotaxic coordinates (Franklin and Paxinos, 1997). Finally, the number of neurons in the infarct core was calculated as the ratio between neuronal density ipsi and neuronal density contra, and expressed as a percentage. In addition, a double labelling for terminal deoxyribonucleotidyl transferase (TdT)-mediated dUTP nick end labelling (TUNEL) staining and NeuN immunoreactivity was performed (described below). Neuronal cells in the infarct core identified by both markers were counted off-line on digitally stored high-power images. All the quantifications were made blindly.

In situ hybridization

In order to combine fluorescent in situ hybridization with fluorescent immunohistochemical detection, brain sections of mice after permanent MCAo were subjected to deparaffinization and heat-mediated epitope retrieval as described above, and hybridized according to the protocol from Roppolo et al. (2007). Digoxigenin-labelled Ccr5 RNA probe was obtained from Eurexpress Consortium (http://www.eurexpress.org) and prepared using the DIG RNA labelling kit (Roche, Basel, Switzerland) following the manufacturer's instructions. After detection of the Ccr5 mRNA Fast red fluorescent signal, sections were incubated overnight at 4°C with primary antibodies and then for 1 h at room temperature with secondary antibody AlexaFluor 488 (Molecular Probes, Basel, Switzerland). Cell nuclei were stained with DAPI. Imaging was performed on a LSM 510 Meta confocal laser scanner mounted on an Axio Imager Z1 microscope (Carl Zeiss, Feldbach, Switzerland). We used brain sections of CCR5-deficient mice also subjected to permanent MCAo as negative control.

TUNEL

TUNEL staining was done as previously described (De Bilbao et al., 2000) using recombinant terminal transferase, following the manufacturer's instructions (Roche). Single staining: TUNEL signal was detected by using a horseradish peroxidase HRP–ABC complex solution (1:100, Vector laboratories) visualized with DAB and H₂O₂. Double staining with NeuN: sections were deparaffinized, subjected to heat-induced epitope retrieval and incubated with NeuN antibody (as described above). After NeuN detection with an Alexa 555-labelled secondary antibody, sections were processed for TUNEL staining. In this case, TUNEL signal was detected with Alexa 488-labelled secondary antibody for streptavidin. Cell nuclei were stained with DAPI. Negative controls of TUNEL staining were performed by omitting TdT. The number of apoptotic/necrotic neurons was determined by counting the NeuN/TUNEL double-stained cells in the infarct region.

Statistical analysis

Results are expressed as means ± standard error of the mean. Statistical analyses were performed using the appropriate test as indicated in the figure legends (Student's t-test, one-way or two-way analysis of variance followed by post hoc Tukey's test). For all tests, a P-value < 0.05 was taken as statistically significant.

Results

Enhanced stroke-related brain damage in CCR5-deficient mice

In order to assess the role of CCR5 in cerebral stroke outcome, we compared brain lesions 2 and 7 days after permanent MCAo in wild-type and CCR5-deficient mice. Two days after ischaemia, the rostro-caudal extension of the infarct was similar in CCR5-deficient mice, as compared with wild type (Figure 1F), while the area of the infarct core was significantly larger, mainly on its latero-medial extension from cortex to sub-cortical structures (39.7% more than wild type; Figure 1A vs. B and E). Seven days after MCAo, in both wild-type and CCR5-deficient mice the area of the infarct core was reduced (Figure 1E), in agreement with previous reports (De Bilbao et al., 2000; Schroeter et al., 2006). However, the length of the infarct remained unchanged in wild-type mice, extending from the level of anterior lateral septum [bregma 1.54 mm; (Franklin and Paxinos, 1997)] to the level of the posterior thalamus–anterior mesencephalon junction and ventral hippocampus [bregma −2.54 mm; (Franklin and Paxinos, 1997)]. The infarct core encompassed mainly the sensory cortex and was generally limited by the external capsule with few extensions into the striatum, but none into the thalamus (Figure 1C). Conversely, in CCR5-deficient mice, the extent of cortical brain damage was markedly increased along the anteroposterior axis, reaching up to the level of the superior colliculus, in the mesencephalon. As compared with wild type, the infarct was also more developed: (i) on the dorsoventral axis, in the cortex, reaching the motor cortex; and (ii) on the lateromedial axis, beyond the external capsule into the striatum, namely in the caudate–putamen complex (Figure 1D). Quantification of the size of the lesions indicated that CCR5 deficiency caused a significantly increased infarct core (88.6% more than wild type; Figure 1E) and rostro-caudal extension (39.7% more than wild type; Figure 1F), namely by reaching part of the auditory and visual cortex at the level of the superior colliculus [bregma −3.64 mm; (Franklin and Paxinos, 1997)]. Therefore, the presence of functional CCR5 appeared to limit cerebral injury after MCAo.

Figure 1.

Increased infarct size in CCR5-deficient (Ccr5^−/−) mice. Male mice were subjected to permanent middle cerebral artery occlusion (MCAo) and killed at 2 and 7 days after stroke. (A–D) Representative brain sections (10 µm) stained with cresyl violet of wild-type (Ccr5^+/+; A,C) and CCR5-deficient (B,D) mice brains 2 days (A,B) and 7 days (C,D) after stroke. Scale bar: 1 mm. (E) Infarcted areas were calculated as a percentage of the whole brain region examined. **P < 0.01 and ***P < 0.001 Ccr5^−/− versus Ccr5^+/+ mice 2 and 7 days after MCAo, °°°P < 0.001 Ccr5^+/+ mice 2 versus 7 days, ^§§§P < 0.001 Ccr5^−/− mice 2 versus 7 days using two-way anova followed by Tukey's post hoc test. (F) The rostro-caudal extension was measured between the rostral pole, at the lateral septum level, and the caudal pole, at the posterior thalamus/mesencephal level. ***P < 0.001 Ccr5^−/− versus Ccr5^+/+ mice 7 days after MCAo, ^##P < 0.01 Ccr5^−/− 7 versus 2 days using two-way anova followed by Tukey's post hoc test (n = 5–8).

Absence of CCR5 does not influence cerebral vasculature or physiological parameters

Several studies have demonstrated that individual and strain-related differences in the cerebrovascular architecture influence MCAo-induced brain lesion size (Connolly et al., 1996; Maeda et al., 1998). Hence, we performed carbon black labelling of cerebral vasculature in wild-type and CCR5-deficient mice. We measured the distances between the inter-hemispheric fissure and the anterior cerebral artery and MCA anastomoses. As shown in Figure 2A, we did not observe any difference in the superficial cerebrovascular architecture that could account for the different infarct sizes between wild-type and CCR5-deficient mice. Alterations of systemic physiological parameters can also modify outcome after cerebral stroke. We therefore analysed these parameters before and after MCAo. We did not observe any significant difference in arterial blood pressure and blood parameters (pH, glucose, O₂ and CO₂ pressure) before and after experimental stroke between wild-type and CCR5-deficient mice (Table 2).

Figure 2.

Cerebral vasculature and stroke-induced motor deficits in wild-type (Ccr5^+/+) and CCR5-deficient (Ccr5^−/−) mice. (A) Cerebrovascular architecture in wild-type and CCR5-deficient mice was compared by measuring the distance from the inter-hemispheric fissure to the anastomotic line between the middle cerebral artery and the anterior cerebral artery in carbon black-perfused brains. Data are shown as mean ± standard error, n = 3. (B,C) Motor deficits were evaluated by performing a grip test before and after middle cerebral artery occlusion (MCAo), at different time points before sacrifice. (B) Grip score before (training) and at different time points following MCAo. *P < 0.05, **P < 0.01 Ccr5^−/− mice training versus 3 days and 24 h after MCAo; ^#P < 0.05 Ccr5^+/+ versus Ccr5^−/− mice 3 days after MCAo; ^##P < 0.01 Ccr5^+/+ versus Ccr5^−/− mice 24 h after MCAo using two-way anova followed by Tukey's post hoc test (n = 6). (C) Representative images of motor behaviour displayed by Ccr5^+/+ and Ccr5^−/− mice during the grip test 7 days after MCAo.

Table 2.

Physiological parameters before and after permanent middle cerebral artery occlusion (MCAo) in wild-type (Ccr5^+/+) and CCR5-deficient (Ccr5^−/−) mice

	Before MCAo		After MCAo
	Ccr5+/+	Ccr5−/−	Ccr5+/+	Ccr5−/−
Glucose (mg·mL−1)	1.73 ± 0.09	1.78 ± 0.09	1.64 ± 0.09	1.51 ± 0.09
pH	7.39 ± 0.01	7.35 ± 0.02	7.39 ± 0.02	7.35 ± 0.02
pCO2 (kPa)	5.14 ± 0.19	5.18 ± 0.09	5.26 ± 0.17	5.56 ± 0.08
pO2 (kPa)	9.59 ± 0.43	9.94 ± 0.22	9.02 ± 0.44	9.27 ± 0.31
Arterial blood pressure (mmHg)	105 ± 2	101 ± 2	102 ± 4	100 ± 3

Blood parameters (glucose, pH, CO₂ and O₂ pressure) and arterial blood pressure were analysed in wild-type (Ccr5^+/+) and CCR5-deficient (Ccr5^−/−) mice before and after permanent MCAo. No statistical significant differences were found using two-way anova (n = 6).

Increased motor deficits in CCR5-deficient mice

In order to examine post-ischaemic neurological deficits, we performed a grip test, which allows the detection of motor function impairment. In this behavioural test, wild-type mice did not exhibit significant motor deficit following cerebral ischaemia (Figure 2B and Video S1). In contrast, the motor performance of CCR5-deficient mice was significantly diminished at 24 h and 3 days after MCAo (Figure 2B and Video S1). At later time points, CCR5-deficient mice showed spontaneous functional recovery and were able to grip until they reached the threshold time. Spontaneous functional recovery in rodents has been previously described (Hunter et al., 2000; Zausinger et al., 2000). However, the pattern of behaviour during the grip test remained distinct. While wild-type mice gripped with four paws and the tail, CCR5-deficient mice gripped with only one or two paws, barely finding a position of equilibrium on the string (Figure 2C).

Thus, increased post-ischaemic lesions in CCR5-deficient mice led to impaired motor functions.

Expression of Ccr5 after MCAo

We analysed the transcriptional expression of Ccr5 in the brain of wild-type mice following permanent MCAo. Twenty-four hours and 7 days after stroke, the mRNA levels of Ccr5 and its ligands Ccl3, Ccl4 and Ccl5 were increased in the hemisphere ipsilateral to the lesion as compared with the contralateral hemisphere (Figure 3A, G). The concomitant expression of CCR5 ligands suggested that CCR5 receptor may also be activated. We also performed in situ hybridization experiments in order to define in which cells Ccr5 was expressed. In wild-type mice the Ccr5 mRNA signal was punctiform (Figure 3B), a distribution observed for other genes (Daberkow et al., 2007; Sonomura et al., 2007). The Ccr5 mRNA probe did not yield any signal in CCR5-deficient mouse brain sections (Figure 3C). In agreement with previous reports (Rottman et al., 1997; Westmoreland et al., 2002; Torres-Munoz et al., 2004), we observed that Ccr5 was present in neurons, astrocytes and microglia (Figure 3D–F).

Figure 3.

Expression of Ccr5 and its ligands (Ccl3, Ccl4 and Ccl5) after permanent middle cerebral artery occlusion (MCAo). Wild-type mice (Ccr5^+/+) were subjected to experimental stroke through permanent MCAo. (A) The mRNA extracted from contralateral and ipsilateral parts of the brain 24 h and 7 days after stroke was analysed by RT-PCR. The ribosomal L32 house keeping gene was used as control. Similar results were obtained from four animals. (B,C) In situ hybridization was performed on Ccr5^+/+ and Ccr5^−/− brain sections after permanent MCAo. Representative images of Ccr5^+/+ (B) and Ccr5^−/− (C) brain sections (infarct border) showing Ccr5 mRNA signal (red dots) and DAPI staining (blue). Scale bar: 25 µm. (D–F) Representative confocal images of the infarct border showing neurons (D), astrocytes (E) and microglia (F) stained, respectively, with neuronal nuclei, glial fibrillary acidic protein and ionized calcium binding adapter molecule-1 antibodies (green), Ccr5 probe (red dots) and DAPI (blue). Scale bar: 25 µm. (G) Semi-quantitative analyses obtained through densitometric scanning of Ccr5, Ccl3, Ccl4, Ccl4 RT-PCR products, normalized with respect to the ribosomal protein L32. ***P < 0.001 using two-way anova followed by Tukey's post hoc test (n = 4).

Astrocyte and microglia response after MCAo in wild-type and CCR5-deficient mice

Given that inflammatory processes may contribute to differences in the extent of post-ischaemic lesions (Wang et al., 2007) and that Ccr5 was expressed in astrocytes and microglia (Figure 3E–F), we compared the inflammatory cell response to ischaemia in wild-type and CCR5-deficient mice. Astrocytes were detected with a GFAP antibody and activated microglia/macrophages with an IBA-1 antibody. Two days after MCAo, only relatively low amounts of astrocytes and activated microglia were detected. Conversely, 7 days after permanent MCAo, GFAP- and IBA-1-immunoreactive cells were accumulated in structures anatomically connected with the infarct (Figure 4A–L), such as the surrounding cortex, the striatum, the ipsilateral thalamus nuclei (posterior and ventro-posterior), the corpus callosum, as well as the contralateral cortex and thalamus. Considering the respective infarct area, there were no differences in the amount of GFAP- or IBA-1-immunoreactive cells between wild-type and CCR5-deficient mice (Figure 4Q–V). Also, both astrocytes and microglia did not show altered morphology in CCR5-deficient mice (Figure 4M–P). Thus, absence of CCR5 does not influence glia recruitment or morphological appearance.

Figure 4.

Astrocyte and microglia response after stroke in wild-type (Ccr5^+/+) and CCR5-deficient (Ccr5^−/−) mice. Astrocytes and microglia/macrophages were detected by immunohistochemistry with glial fibrillary acidic protein (GFAP) or ionized calcium binding adapter molecule-1 (IBA-1) antibodies, respectively, in Ccr5^+/+ and Ccr5^−/− mouse brains 7 days after middle cerebral artery occlusion (MCAo). (n = 8). Representative images of (A–F) GFAP or (G–L) IBA-1 staining at different coronal levels of Ccr5^+/+ (A–C and G–I) and Ccr5^−/− (D–F and J–L) brain sections. Scale bar: 1 mm. (M–P). High-power images showing astrocytes (M,N) and microglia (O,P) in brains of Ccr5^+/+ (M,O) and Ccr5^−/− (N,P) mice 7 days after MCAo. Scale bar: for (M, N) 90 µm, for (O, P) 45 µm. (Q–V) Quantification of GFAP (Q–S) and IBA-1 (T–V) staining: the surface occupied by GFAP- or IBA-1-immunoreactive cells (y-axis) was quantified as described in the Method section in brain sections of Ccr5^+/+ and Ccr5^−/− mice 2 and 7 days after MCAo. ***P < 0.001 using two-way anova followed by Tukey's post hoc test. No significant differences were detected between Ccr5^+/+ and Ccr5^−/− mice (n = 4–8).

Neutrophil recruitment after stroke in wild-type and CCR5-deficient mice

We analysed neutrophil recruitment in ischaemic lesions by staining with an antibody against MPO. Neutrophils were barely detected in wild-type mice, at both 2 and 7 days after stroke (Figure 5A,C). In contrast, CCR5-deficient mice displayed a massive invasion of the infarct core by neutrophils 7 days after ischaemia (Figure 5D,F). However, no such invasion was observed at the earlier time point (2 days, Figure 5B, E), suggesting that neutrophil accumulation at injured regions is a late event.

Figure 5.

Neutrophil infiltration after middle cerebral artery occlusion (MCAo) in wild-type (Ccr5^+/+) and CCR5-deficient (Ccr5^−/−) mice. Neutrophils were detected by immunohistochemistry using an antibody against myeloperoxidase (MPO). (A–D) Representative brain sections of Ccr5^+/+ (A,C) and Ccr5^−/− (B,D) mice stained with MPO antibody 2 days (A,B, n = 5) and 7 days (C,D, n = 8) after MCAo. Scale bar: 1 mm. (E,F) Quantification of the MPO staining in the infarct core 2 (E) and 7 (F) days after MCAo. The surface occupied by MPO-immunoreactive cells within the infarct core (y-axis) was quantified as described in the Method section. *P < 0.05 using Student's t-test (n = 8).

Increased neuronal death in CCR5-deficient mice

The increased infarct size in CCR5-deficient mice might be associated with a decreased neuronal survival. We therefore analysed the neurons in the infarct core that were still immunoreactive for NeuN antibody, as well as the signs of cell death using TUNEL labelling. By counting NeuN-immunoreactive cells in the infarct region, we observed a significant decrease in the number of neurons in CCR5-deficient mice (Figure 6G). In support of this observation, the number of TUNEL-positive cells was markedly increased in CCR5-deficient mice as compared with wild type (Figure 6A vs. B and H). Double staining with NeuN antibody and TUNEL labelling suggested that the majority of dying cells in CCR5-deficient infarcts were neurons (Figure 6C–F). Thus, the greater brain damage in CCR5-deficient mice was associated with increased neuronal death in the infarct region.

Figure 6.

Increased neuronal death in CCR5-deficient (Ccr5^−/−) infarct brains. Neuronal death was analysed by immunohistochemistry on brain sections 7 days after middle cerebral artery occlusion. (A,B) Representative images of terminal deoxyribonucleotidyl transferase-mediated dUTP nick end labelling (TUNEL) staining in wild-type (Ccr5^+/+; A) and Ccr5^−/− (B) ischaemic brains. (C–F) Representative image showing co-staining of TUNEL (green, E), NeuN (red, F) and DAPI (blue). Picture in (D) highlighted by dotted box in (C). Scale bars: (A,B) 340 µm; (C) 250 µm, (D–F) 80 µm. (G) Number of neurons in infarct regions normalized to the number of neurons in the corresponding contralateral parts. ***P < 0.001 Ccr5^+/+ versus Ccr5^−/− using Student's t-test (n = 6). (H) Numbers of TUNEL-positive neurons in infarct regions. ***P < 0.001 using Student's t-test (n = 6).

Discussion and conclusions

In this study, we show that CCR5 deficiency in mice exacerbates the severity of cerebral infarction after MCAo. In ischaemic brains, CCR5 was expressed in neurons, astrocytes and microglia. The infarct region of CCR5-deficient mice was characterized by strikingly increased neuronal injury, without detectable changes in astrocytes and microglia. Thus, CCR5 expression is neuroprotective in this mouse stroke model.

The fact that CCR5 contributes to protective mechanisms in stroke points towards a unique role of this chemokine receptor. Indeed, the interaction of chemokines with their receptors promotes leukocyte trafficking and inflammation and is therefore expected to aggravate cerebral infarction. In support of this concept, mice deficient for CCR2, CCL2 (=CCR2 ligand), or CX3CL1 (=CX₃CR1 ligand) showed decreased infarct size after ischaemic stroke (Hughes et al., 2002; Soriano et al., 2002; Dimitrijevic et al., 2007). Similarly, several chemokine inhibitors or antagonists of chemokine receptors reduced brain damage after stroke: (i) TAK779, a CCR2 antagonist in mice (Takami et al., 2002; Saita et al., 2007); (ii) NR58-3.14.3 and vMIP-2, broad-spectrum antagonists of chemokine receptors (Beech et al., 2001; Takami et al., 2001); (iii) repertaxin, a CXCL8 inhibitor (Garau et al., 2005); and (iv) DF2156A, a dual inhibitor of CXCR1 and CXCR2 (Garau et al., 2006). However, here we report that CCR5 deficiency aggravates brain damage and behavioural outcome after cerebral stroke.

Such a protective role of CCR5 could not have been predicted from inflammatory pathologies outside the CNS. For example, there is evidence for a detrimental role of CCR5 in experimental atherosclerosis (Braunersreuther et al., 2007), in development of human cardiovascular disease (Afzal et al., 2008), and in experimental pulmonary emphysema (Ma et al., 2005). In contrast, there are indications for a protective effect of CCR5 in the CNS. Indeed, CCR5-deficient mice infected with Cryptococcus neoformans, herpes simplex virus type 2 or West Nile virus, displayed increased severity of CNS damage (Huffnagle et al., 1999; Glass et al., 2005; Thapa et al., 2007). In these cases, the protection by CCR5 was attributed to its capacity to increase host-defence through CCR5-dependent regulation of leukocyte trafficking. The fact that such a protective role is also found in a non-infectious condition, such as stroke, might hint towards a more global role of CCR5 in neuroprotection. Several lines of argument support this concept. Thus, stimulation of endogenous CCR5 by its natural ligands triggered neuroprotective mechanisms and prevented apoptosis in response to the HIV envelope protein gp120 (Meucci et al., 1998; Kaul and Lipton, 1999;Catani et al., 2000; Kaul et al., 2007). Further, the CCR5 ligands CCL3 and CCL4 prevent excitotoxicity induced by NMDA (Bruno et al., 2000) and nerve injury-induced motor neuron death was accelerated in CCR5-deficient mice (Gamo et al., 2008).

The mechanism underlying the protective effect of CCR5 after stroke could be due to its capacity to regulate inflammatory response. Glia abundance and morphology is not altered in CCR5-deficient mice after stroke (this study), hippocampal damage (Babcock et al., 2003) or nerve injury-induced motoneuron death (Gamo et al., 2008). Nevertheless, the last study reported evidence for a functional difference of CCR5-deficient microglia, which, in vitro, released increased amounts of potentially neurotoxic cytokines after exposure to LPS (Gamo et al., 2008). Thus, it is possible that CCR5 limits microglia activation and neuroinflammation, thereby decreasing brain damage. This might also explain the important neutrophil invasion in CCR5-deficient brains observed at late time points in our study. Theoretically, CCR5 could be also involved in inflammatory monocyte migration into the brain after ischaemic damage (Getts et al., 2008); however, the unchanged number of IBA-1-positive cells makes this possibility less likely. As CCR5 is also expressed in neurons, the possibility that it might directly regulate neuronal survival should also be considered. Indeed, there is an emerging consensus that chemokine receptors are expressed in neurons, where they modulate various neuronal functions, including neurotransmission, neuronal migration and neuronal survival (Rostene et al., 2007).

An important issue raised by the protective role of CCR5 in experimental cerebral ischaemia is its potential clinical relevance for cerebrovascular pathologies in humans, from stroke to vascular dementia. In other types of pathologies, there is a good correlation between mice and humans with respect to the impact of CCR5 deficiency. This is true for the decreased peripheral cardiovascular risk (see above), but also for the increased severity of infection with West Nile virus (Glass et al., 2005; 2006; Lim et al., 2008). Thus, is CCR5 deficiency, found in 1–2% of the Caucasian population, a risk factor for the severity of stroke in humans? And if yes, what would be the clinical presentation? Indeed, CCR5 deficiency could lead to an increased number of mild strokes (transient ischaemic attacks presenting as stroke), or to an increased severity of stroke. This question is also interesting in the context of HIV, where the CCR5 antagonist maraviroc has been recently introduced as a treatment, which prevents HIV from entering CD4 cells via the CCR5 co-receptor (Emmelkamp and Rockstroh, 2007; Sayana and Khanlou, 2009). As HIV-infected individuals have an increased risk of developing cerebrovascular disease (Cole et al., 2004; Tipping et al., 2007), the question whether maraviroc might increase frequency or severity of stroke needs to be addressed.

In summary, our results reveal a novel function of CCR5 in the CNS and suggest that the pharmacological interventions enhancing or decreasing CCR5 function could modify neuronal survival after ischaemia.

Glossary

Abbreviations:

DNase

deoxyribonuclease

GFAP

glial fibrillary acidic protein

IBA-1

ionized calcium binding adapter molecule-1

MCA

middle cerebral artery

MCAo

MCA occlusion

MIP

macrophage inflammatory protein

MPO

myeloperoxidase

NeuN

neuronal nuclei

RANTES

regulated on activation, T cell expressed and secreted

TdT

terminal deoxyribonucleotidyl transferase

TUNEL

TdT-mediated dUTP nick end labelling

Conflicts of interest

The authors have no conflicting financial interests.

Supporting information

Additional Supporting Information may be found in the online version of this article:

Video S1 Representative videos of Ccr5^+/+ and Ccr5^−/− mice during the grip test. (Video01) Ccr5^+/+ and (Video02) Ccr5^−/− mice 24 h after permanent middle cerebral artery occlusion.

Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.