Introduction

The blood-brain barrier (BBB) is a dynamic structure mainly composed of brain microvascular endothelial cells (BMECs), pericytes, astrocytes, and a non-cellular component---the basement membrane (BM) [7, 57, 77]. By tightly regulating substance exchange between the CNS and circulation system, the BBB functions to maintain CNS homeostasis. Accumulating evidence suggests that BBB disruption contributes to the pathogenesis and progression of various neurological disorders [48, 81, 82]. For example, BBB breakdown affects inflammatory cell infiltration and is associated with the development/progression of ischemia-reperfusion injury [15, 32, 75]. It should be noted that the majority of BBB studies focus on its cellular constituents, and the role of the BM in BBB regulation remains largely unknown.

The BM consists of highly organized extracellular matrix proteins synthesized by astrocytes, BMECs, and mural cells, which include both pericytes and vascular smooth muscle cells (vSMCs) [29, 51, 67, 76]. Laminin, the only protein that is absolutely required for BM formation, is a trimer composed of α, β, and γ subunits [20, 51, 76, 77]. Among all five genetic variants of the α subunits, laminin-α4 and -α5 are highly expressed in blood vessels throughout the body [29, 67, 76]. Unlike laminin-α4, which is ubiquitously distributed in the vasculature, laminin-α5 expression shows a patchy pattern at smaller vessels [73]. The major cell types that synthesize laminin-α5 in the vasculature are BMECs and mural cells [26, 29, 46, 62, 65, 67, 80]. Recent studies demonstrated that knockout of laminin-α5 in endothelial cells failed to affect BBB permeability under homeostatic conditions [25, 63]. In TNFα-induced inflammation, however, these mutants showed significantly enhanced neutrophil extravasation in cremaster muscle [63]. In collagenase-induced intracerebral hemorrhage (ICH) model, these mutants displayed exacerbated inflammatory cell infiltration [25]. In addition, in the experimental autoimmune encephalomyelitis (EAE) model, decreased T cell infiltration into the brain and reduced disease susceptibility & severity were observed in laminin-α4 null mice [73], which exhibited compensatory and ubiquitous expression of laminin-α5 along the vasculature [73]. These findings suggest that endothelial laminin-α5 plays an inhibitory role in inflammatory cell extravasation under pathological conditions, although it is dispensable for BBB maintenance under physiological conditions [25, 63]. Whether mural cell-derived laminin-α5 is involved in BBB regulation under physiological and pathological conditions, however, remains unknown. Given that mural cell-derived laminin-α5 is an important component of the BM at the BBB [51, 76], we hypothesize that mural cell-derived laminin-α5 may also contribute to BBB integrity. In this study, we investigated the functions of mural cell-derived laminin-α5 in BBB regulation under homeostatic conditions and in ischemic stroke.

Materials and methods

Mice

The experimental protocols were reviewed and approved by the Institutional Animal Care and Use Committee at the University of Georgia and were in accordance with the National Institute of Health Guide for Care and Use of Laboratory Animals. The Animal Research: Reporting In Vivo Experiments (ARRIVE) guidelines for reporting experiments involving animals were strictly followed. Laminin-α5flox/flox mice were generated as described previously [50]. Pdgfrβ-Cre+ mice were a generous gift from Dr. Volkhard Lindner. These two transgenic lines were crossed to generate laminin-α5flox/flox: Pdgfrβ-Cre+ (α5-PKO) mice. Their wildtype littermates were used as controls. In this study, 194 mice (102 control and 92 α5-PKO) were used. All mice were housed in the animal facility at the University of Georgia with free access to water and food.

Middle cerebral artery occlusion (MCAO)

Eight-week-old control and α5-PKO mice were subjected to 45 min of focal cerebral ischemia produced by transient intraluminal occlusion of the middle cerebral artery using a filament as described previously [49, 68]. Briefly, mice were anesthetized with 2,2,2-tribromoethyl alcohol (250 mg/kg, i.p.). A midline neck incision was made and the common carotid artery (CCA), external carotid artery (ECA), and internal carotid artery (ICA) on the right side were carefully isolated. The ECA and CCA were ligated distal to the carotid bifurcation. The ICA was clipped temporarily. A 6–0 silicone monofilament suture (Doccol) with a 0.21 mm diameter was introduced into the CCA via an incision, advanced 9 mm distal to the carotid bifurcation and secured in place. Successful occlusion of the middle cerebral artery was confirmed with the PeriCam PSI HR system (Perimed) based on laser speckle contrast analysis technology. Animals showing diminished blood flow of at least 80% during occlusion with at least 75% recovery of blood flow after reperfusion were used for experimentation. The body temperature was maintained at 37.0 ± 0.5 °C during the surgery using a heating pad. Animals had free access to food and water throughout the reperfusion period. This ischemic model led to ~ 30% and ~ 20% mortality rates for control and α5-PKO mice, respectively.

Body weight loss and neurological function

The body weight loss was evaluated daily from days 1 to 7 after surgery. Neurological function was assessed using the modified neurological severity scores (mNSS) system, which evaluates motor, sensory, reflex and balance functions, as described previously [16, 38, 58]. Briefly, mice were scored based on their performances in a variety of tests as described in Additional file 1: Table S1. The sum of these scores (0–14) was used to reflect their neurological function after MCAO. Higher scores indicate worse neurological function. Animals were habituated to the testing environment prior to experiments and the investigator who scored the animals was blinded to the genotypes.

Brain sectioning

Serial sectioning was used in this study. Briefly, 20 μm-thick serial sections were cut with Cryostat (Micro HM 550, Thermo Scientific). Eight sections evenly distributed along the rostral-to-caudal axis were collected from each brain.

Infarct volume and neuronal death

Brain infarct volume was quantified as infarct volume percentage (%) as described previously [43, 53, 56]. Briefly, cresyl violet-stained brain sections were imaged using the Nikon Eclipse Ti microscope. The areas of the contralateral hemisphere (Ci), ipsilateral hemisphere (Ii), and ipsilateral non-ischemic region (Ni) were determined using the Image J software (NIH), and the infarct volume (%) was calculated as:\( \mathrm{Infarctvolume}\left(\%\right)=\left(\frac{\sum \limits_i\left(\left(\frac{I_i-{N}_i}{I_i}\right){C}_i\right)}{2{\sum}_i{C}_i}\right)x100 \).

Neuronal death was assessed using Fluoro-Jade C (FJC) staining as described previously [59, 79]. Specifically, the number of FJC+ cells was counted in each field. At least 3 random fields from each section, 8 serial sections per brain, and 4 animals were used for quantification.

BBB permeability

Evans blue (EB) and/or FITC-Dextran (4kD) were used to assess BBB permeability as described previously [15]. Briefly, control and α5-PKO mice were injected retro-orbitally with 80 μl EB (2%, Sigma E2129) and/or 50 μl FITC-Dextran (25 mg/ml, Sigma FD4). For non-ischemic study, FITC-Dextran was allowed to circulate for 12 h. After transcardial perfusion, the brains were collected, homogenized in formamide, and centrifuged at 20,000 rpm for 20 min. The fluorescence intensity of the supernatant was measured using a SpectraMax M2 plate reader (Molecular Devices) at 450/550 nm. Mice without FITC-Dextran injection were used to determine baseline reading, which was subtracted from raw reading to obtain FITC-Dextran leakage. Leakage in α5-PKO mice was normalized to that in controls. For ischemic study, both tracers were injected 4 h before mice were transcardially perfused at each time point after injury. Each brain hemisphere was homogenized in formamide and centrifuged at 20,000 rpm for 20 min. The absorbance and fluorescence intensity of the supernatant were measured using a SpectraMax M2 plate reader at 620 nm and 450/550 nm, respectively. EB or FITC-Dextran leakage was defined as the difference of absorbance or fluorescence intensity between contralateral and ipsilateral hemispheres. Leakage in α5-PKO mice was normalized to that in controls.

Brain edema

Brain edema was assessed using both brain water content [79] and brain swelling [52, 69] and -α5 [46, 64, 65]. In addition, laminin-α2 has also been found in vSMCs from large vessels [45], such as the aorta and carotid arteries. There is also evidence showing that laminin-β1 is expressed in vSMCs in develo** vessels, whereas laminin-β2 is found in mature vasculature [28], suggesting a switch from β1- to β2-containing laminins during vessel maturation. Collectively, these results suggest that mural cells mainly express laminin-411, − 511, and possibly − 211 during development; and laminin-421, − 521, and possibly − 221 in adulthood under homeostatic conditions.

In this study, we failed to detect any changes in BBB permeability and CBF between control and α5-PKO mice under physiological conditions, suggesting that mural cell-derived α5-containing laminins are dispensable for BBB maintenance and CBF regulation under homeostatic conditions. Unlike these α5-PKO mice, mutants with laminin-γ1 deficiency (all γ1-containing laminins) in mural cells showed BBB breakdown and hydrocephalus in C57Bl6-FVB mixed background [26], suggesting an important role of mural cell-derived γ1-containing laminins in BBB maintenance and hydrocephalus pathogenesis, although we cannot exclude the possibility that BBB disruption is secondary to hydrocephalus. Together, these findings suggest the existence of compensation between mural cell-derived α5-containing laminins and α4/α2-containing laminins. In addition, it is also possible that the lack of phenotype in α5-PKO mice under homeostatic conditions is due to compensation by laminin isoforms from endothelial cells and/or astrocytes, which are in close proximity of mural cells [67]. For example, mural cell-derived α5-containing laminins and endothelial cell-derived laminin-511 may be able to compensate for each other’s loss. In support of this possibility, mice with laminin-α5 deficiency in endothelial cells are grossly normal and fail to show any defects under homeostatic conditions [25, 63].

After ischemia-reperfusion injury, α5-PKO mice demonstrated alleviated BBB disruption at days 1, 2, and 7 after injury. Consistent with the reduced BBB leakage, TJP (ZO-1 and claudin-5) levels were less severely reduced in the mutants at days 1 and 2 after injury. By day 7 after injury, however, BBB leakage but not TJP expression showed a significant difference between genotypes. This finding suggests that TJPs are not responsible for the difference in BBB integrity between genotypes at this time point, highlighting a possible role of transcytosis in BBB integrity maintenance. Echoed with this observation, tight junction-independent BBB disruption and the important role of transcytosis in BBB regulation have been reported in recent studies [2, 10, 17, 36, 71].

In addition, α5-PKO mice also displayed diminished inflammatory cell (neutrophil, lymphocyte, and mononuclear cell) infiltration, suggesting a “pro-infiltration” role of mural cell-derived laminin-α5 after ischemic injury. This is in contrast with a previously reported “anti-infiltration” role of endothelial laminin-α5. It has been demonstrated that loss of laminin-α5 in endothelial cells increased immune cell extravasation in cremaster muscle after inflammation [63] and in the brain after ICH [25]. In addition, in the EAE model, reduced infiltration of T lymphocytes in the brain was found in laminin-α4 null mice, which demonstrated compensatory & ubiquitous expression of laminin-α5 in the vasculature [73]. One explanation for this discrepancy is that mural cells and endothelial cells make different α5-containing laminins, which exert distinct functions to regulate immune cell extravasation. It should be noted, however, that we cannot exclude the possibility that mural cells make “new” laminin isoforms after ischemic injury, which are responsible for the observed “pro-infiltration” effect. Another possibility is that different injury/animal models and time points are responsible for this discrepancy. The “anti-infiltration” role of endothelial laminin-α5 is mainly supported by studies using a muscle inflammation model [63], an ICH model [25], and an EAE model [73], whereas the “pro-infiltration” role of mural cell-derived laminin-α5 is obtained from ischemia-reperfusion study. Unlike ischemia-reperfusion injury, the muscle inflammation model does not damage the BBB or the CNS. Additionally, loss of endothelial laminin-α5-induced increase of immune cell infiltration only occurs at a specific time point (1.5 h after TNFα injection) in this muscle inflammation model [63]. Although BBB disruption is replicated in the ICH model, brain pathology in ICH is completely different from that in ischemia-reperfusion injury. For example, blood vessel wall and BM are disrupted in the collagenase-induced ICH model, which causes immediate leakage of inflammatory cells into the brain, whereas such vascular damage is absent in the ischemia-reperfusion model. Therefore, it is unclear whether the increased accumulation of inflammatory cells in mutant brains is due to a direct “anti-infiltration” effect of endothelium-derived laminin-α5. In the EAE model, laminin-α4 global knockouts that showed compensatory up-regulation of laminin-α5 rather than endothelium-specific laminin-α5 knockouts were used [73]. Since mural cells also synthesize α4-containing laminins [30, 52, 69], both endothelium- and mural cell-derived laminin-α4 is ablated in these laminin-α4 knockouts. It is thus unclear whether the enhanced laminin-α5 is from endothelial cells or mural cells, which makes data interpretation difficult. We are currently investigating the role of endothelium-derived laminin-α5 in ischemic stroke using endothelium-specific laminin-α5 conditional knockout mice. Results from this study will contribute to our understanding of the biological function of endothelial laminin-α5.

α5-PKO mice exhibited milder vascular damage, such as less severe BBB disruption and decreased inflammatory cell infiltration, and attenuated neurological injury, including reduced ischemic volume, diminished neuronal death, and improved neurological function. Given that inflammatory cells actively contribute to secondary brain injury after stroke [1], we speculate that the attenuated neurological injury is due to milder vascular damage. In support of this possibility, extravasated neutrophils have been demonstrated to contribute to neuronal injury and brain edema in ischemic injury [12, 34, 37, 55, 60]. Similarly, lymphocytes are found to be responsible for delayed post-ischemic injury [39, 40]. In addition, monocytes have been shown to play a detrimental role in the acute phase (up to 3 days) after ischemic injury, although a beneficial role is reported in the chronic phase (after day 3) [21, 22]. Consistent with these reports, reduced numbers of neutrophils, lymphocytes, and mononuclear cells were observed in α5-PKO mice after ischemic injury, especially at early time points. It should be noted, however, that we are unable to exclude the possibility that attenuated neurological injury leads to milder vascular damage.

α5-PKO mice demonstrate a better outcome after ischemia-reperfusion injury, suggest a detrimental role of mural cell-derived laminin-α5 in ischemic injury. Similar to our α5-PKO mutants, mice with endothelium-specific deletion of integrin-α5 demonstrated substantially reduced infarct size, increased BBB integrity and improved neurological function after stroke [54], highlighting an adverse effect of endothelial integrin-α5 in ischemic stroke. Together, these findings suggest that mural cell-derived α5-containing laminins and endothelial integrin-α5 may use a converging signaling pathway to modulate the development/progression of ischemic stroke, although integrin-α5 is not a classical laminin receptor [6, 74]. Identifying the receptors and downstream signaling pathways may provide innovative molecular targets with therapeutic potential in ischemic stroke.

Due to the multiphasic nature of ischemic stroke, this study has a few limitations. First, only the transient ischemic model was used in this study. The transient ischemic model involves both ischemia and reperfusion. However, most strokes found in human patients only involve ischemia without reperfusion [19, 23, 42, 61, 70]. Thus, it is important to test the biological function of mural cell-derived laminin-α5 in the permanent ischemic model. Second, only one ischemic duration (45 min) was used in this study. It is known that longer occlusion causes more severe injury [44, 49]. Currently, various ischemic durations ranging from 30 to 120 min have been used in rodent MCAO studies [13, 14, 35, 49]. We chose 45-min ischemia for two reasons: (1) compared to other durations, 45-min ischemia consistently induced significant ischemic injury with less mortality in our hands, and (2) significant differences in stroke outcomes between control and α5-PKO mice were observed with 45-min ischemia. Other ischemic durations should be tested in future studies. Third, only young mice were used in this study. Aging is a risk factor for ischemic stroke and actively influences stroke outcomes [19, 23, 42]. Therefore, it is important to examine the biological function of mural cell-derived laminin-α5 in ischemic stroke using aged mice in the future. Fourth, unlike previous studies reporting improved outcomes in young female mice [19, 23, 42], we failed to observe gender differences in infarct volume, neurological severity score, and body weight loss. It should be noted that, although not statistically significant, a trend toward attenuated injury was observed in female mice independent of genotype. This discrepancy may be explained by the relatively small animal number used in each group and/or other factors, such as the severity of injury and sensitivity of assays. Future research is needed to address these limitations.

Conclusions

Collectively, our results suggest that mural cell-derived laminin-α5 is dispensable for BBB maintenance and CBF regulation under homeostatic condition. In ischemic stroke, however, loss of mural cell-derived laminin-α5 attenuates vascular damage and improves stroke outcomes, indicating a detrimental role of mural cell-derived laminin-α5 in ischemic stroke. These findings identify mural cell-derived laminin-α5 as a molecular target with therapeutic potential in ischemic stroke.