Vascular Signal Transducer and Activator of Transcription-3 Promotes Angiogenesis and Neuroplasticity Long-Term After StrokeCLINICAL PERSPECTIVE
Background—Poststroke angiogenesis contributes to long-term recovery after stroke. Signal transducer and activator of transcription-3 (Stat3) is a key regulator for various inflammatory signals and angiogenesis. It was the aim of this study to determine its function in poststroke outcome.
Methods and Results—We generated a tamoxifen-inducible and endothelial-specific Stat3 knockout mouse model by crossbreeding Stat3floxed/KO and Tie2-CreERT2 mice. Cerebral ischemia was induced by 30 minutes of middle cerebral artery occlusion. We demonstrated that endothelial Stat3 ablation did not alter lesion size 2 days after ischemia but did worsen functional outcome at 14 days and increase lesion size at 28 days. At this late time point vascular Stat3 expression and phosphorylation were still increased in wild-type mice. Gene array analysis of a CD31-enriched cell population of the neurovascular niche showed that endothelial Stat3 ablation led to a shift toward an antiangiogenic and axon growth-inhibiting micromilieu after stroke, with an increased expression of Adamts9. Remodeling and glycosylation of the extracellular matrix and microglia proliferation were increased, whereas angiogenesis was reduced.
Conclusions—Endothelial Stat3 regulates angiogenesis, axon growth, and extracellular matrix remodeling and is essential for long-term recovery after stroke. It might serve as a potent target for stroke treatment after the acute phase by fostering angiogenesis and neuroregeneration.
Long-term recovery after brain ischemia is linked to angiogenesis. How newly formed vessels can improve functional outcome is unclear. Angiogenesis restores oxygen and nutrient supply to the affected brain tissue, but poststroke angiogenesis starts too late to interact with ischemic cell death cascades. The newly formed vessels seem to be more relevant for regenerative mechanisms, such as neurogenesis, axonal growth, and synaptic plasticity.1 Angiogenesis and neurogenesis are intricately linked.2 Axonal growth cones and endothelial tip cells of sprouting vessels share many features and respond to the same signaling pathways.3 Angiogenic vessels provide neurotrophic support to newly generated neurons and facilitate synaptogenesis.4
Clinical Perspective on p 1782
Treatments are needed that foster regeneration at later time points. Strategies focusing on angiogenesis are ambivalent. Most angiogenic factors, like vascular endothelial growth factor (Vegf), may increase vascular permeability and thus increase edema formation. In addition, angiogenesis is strongly related to inflammatory pathways. For instance, Vegf and interleukin 6 (Il6) converge on the signal transducer and activator of transcription-3 (Stat3) signaling pathway; on the other hand, activation of Stat3 by Il6 increases Vegf production.5 In endothelial cells, Stat3 activation promotes angiogenesis by regulating endothelial cell migration and proliferation6 and is involved in vascular diseases such as pulmonary hypertension.7
Stat3 shows neuroprotective properties in the acute phase of stroke, inhibiting apoptosis and scavenging reactive oxygen species in neurons, thereby reducing infarct size.8–13 It might, however, induce detrimental effects in the subacute phase after stroke. It has been shown that Stat3 is mainly expressed and activated in microglia and macrophages 24 to 72 hours after cerebral ischemia.14 Blocking Stat3 phosphorylation within the first 72 hours reduced lesion size; this effect is mediated by an inactivation of Stat3 in microglia and macrophages.
After brain ischemia, endothelial cells seem to express and phosphorylate Stat3 later than neurons and microglia, beginning at 48 hours and continuing for at least 7 days.8 This is the time window in which expression and phosphorylation of Stat3 by endothelial cells reach their highest levels compared with those of other cell types.8 This delayed activation of the angiogenic Stat3 pathway in endothelial cells after cerebral ischemia may indicate an important function of endothelial Stat3 in regenerative mechanisms. We demonstrated previously that lack of Il6 increased lesion size and worsened functional outcome 28 days after ischemia. The underlying mechanism seemed to be a reduction of angiogenesis.15
It was therefore the aim of this study to search for the specific role of endothelial Stat3 in the pathophysiology of cerebral ischemia and in particular its effects on poststroke angiogenesis and long-term outcome. To obtain a knockout of Stat3 that is restricted to endothelial cells in vivo, we used an inducible endothelial-specific Stat3 knockout mouse model by crossbreeding Stat3floxed/KO mice16 with Tie2-CreERT2 mice17 (endothelial-Tie2-CreERT2;Stat3floxed/KO). Tie2-CreERT2 littermates treated with tamoxifen served as controls. Cerebral ischemia was induced by filamentous occlusion of the middle cerebral artery (MCAo) for 30 minutes. This study shows for the first time that endothelial Stat3 activation is necessary for poststroke angiogenesis, alters long-term lesion size, and plays a beneficial role in functional improvement after stroke. Endothelial loss of Stat3 leads to an antiangiogenic and axon growth-inhibiting micromilieu within the neurovascular niche with induced Adamts9 expression and, consequently, an altered remodeling of the extracellular matrix.
A detailed Methods section is provided in the online-only Data Supplement.
Animals and Treatments
All of the animal experiments were approved by the local governmental authorities (Landesamt für Gesundheit und Soziales, G0354/11). Stat3floxed/KO mice16 were crossbred with Tie2-CreERT2 mice.17 Endothelial-specific knockout of Stat3 was induced by a 5-day intraperitoneal application of tamoxifen (1 mg/d) 7 days before surgery. Tie2-CreERT2 littermates were treated with tamoxifen and served as a control group. In a total of 3 animals, 2 wild-type mice and 1 knockout animal, no lesions developed. These 3 animals were excluded from the study.
Model of MCAo
Filamentous middle cerebral artery occlusion of the left side was performed for 30 minutes with indicated reperfusion time points according to the published standard operating procedure in our laboratory18 in the DIN9001-certified laboratories at the Charité Department of Experimental Neurology.
Real-Time Reverse-Transcription Polymerase Chain Reaction
RNA from sections of ischemic and contralateral hemispheres was extracted using TRIzol (Invitrogen, Carlsbad, CA). A total of 1 μg of RNA was transcribed with murine leukemia virus reverse transcriptase (Promega). Real-time polymerase chain reaction was performed with intron spanning primers for tyrosine 3-monooxygenase/tryptophan 5-monooxygenase activation protein and succinate dehydrogenase complex, subunit A, flavoprotein as internal controls19 and primers for Stat3 using Sybr green (Qiagen, Limburg, the Netherlands). Fold amount of mRNA compared with mean of internal controls was calculated using the following equation: 2–(Ct[gene of interest]–Ct[internal control]).
Histology and Immunofluorescence
Mice were placed under deep anesthesia, cardially perfused with NaCl solution, and decapitated. The brains were snap frozen in –40°C isopentane for cryostat sectioning 2 and 28 days after MCAo. Cryosections (20 μm) were fixed with acetone/methanol and stained for caveolin-1, pStat3 (Tyr705), bromodeoxyuridine, NeuN, isolectin-B4, and Adamts9. Cytospins after CD31-MACS were stained for caveolin-1. Orthogonal projections of vascular networks in the peri-infarct region were stained with isolectin-B4 (see the online-only Data Supplement for details).
Lesion Volume Determination and Functional Outcome
We quantified cerebral lesion volume with ImageJ analysis software (National Institutes of Health, Bethesda, MD) on 20-μm NeuN-3,3'-diaminobenzidine–stained cryostat sections and calculated volume by summing up the infarct sizes of each section as described previously18 or by T2-weighted magnetic resonance imaging. Functional outcome was determined with behavioral tests (Rotarod test and the extended neuroscore). To determine lesion size, we used a total of 16 endothelial-Tie2-CreERT2;Stat3floxed/KO and 23 Tie2-CreERT2 littermates. All underwent magnetic resonance imaging at 2 days after ischemia. A subset of 9 endothelial-Tie2-CreERT2;Stat3floxed/KO and 12 Tie2-CreERT2 littermates was killed at 2 days after ischemia and a subset of 7 endothelial-Tie2-CreERT2;Stat3floxed/KO and 11 Tie2-CreERT2 littermates at 28 days for histological lesion size determination. For behavioral tests, 12 endothelial-Tie2-CreERT2;Stat3floxed/KO and 12 Tie2-CreERT2 littermates were used.
Enzyme-Linked Immunosorbent Assay
Enzyme-linked immunosorbent assays for determining Vegf (R&D systems) and Il6 (Invitrogen) levels were performed according to manufacturer’s instructions.
Isolation of CD31-Enriched Cells
Mice were killed 4 days after cerebral ischemia. The ipsilateral and contralateral hemispheres were divided and processed separately. Hemispheres from 7 animals were pooled. The tissue was dissociated using the Papain Neural Tissue Dissociation kit (Miltenyi Biotec, Bergisch Gladbach, Germany). Myelin was removed following a protocol published elsewhere.20 Cells were labeled with anti-CD31 microbeads (Miltenyi Biotec). We performed MACS isolation according to the manufacturer’s instructions. Total RNA from MACS was isolated using the RNeasy Plus Mini kit (Qiagen). A total of 21 endothelial-Tie2-CreERT2;Stat3floxed/KO and 21 Tie2-CreERT2 littermates were used and divided to 3 groups each before isolation.
GeneChip Microarray Assay
Sample preparation for microarray hybridization was carried out as described in the Ambion WT Expression kit protocol (Life Technologies, Frederick, MD) and the Affymetrix WT Terminal Labeling and Hybridization user manual (Affymetrix, Inc, Santa Clara, CA). Sample processing was performed at an Affymetrix Service Provider and Core Facility, KFB Center of Excellence for Fluorescent Bioanalytics (Regensburg, Germany).
Microarray Data Analysis
Summarized probe set signals were calculated by using the robust multiarray average21 algorithm with the Affymetrix GeneChip Expression Console Software. Probe sets with a change >2.0-fold and a Student t test P<0.05 were considered as significantly regulated with n=3 independent experiments.
Fluorescence-Activated Cell Sorting
Coexpression of antimouse CD11b-APC and CD45-efluor450 are shown in density plots. CD11b+CD45hi cells infiltrate monocytes/macrophages, whereas CD11b+CD45low cells are the intrinsic microglial population.
All of the data are presented as scatter dot plots with mean±SD. Detailed description of statistical analysis is provided in the figure legends and in the online-only Data Supplement.
Conditional Ablation of Endothelial Stat3
An endothelial-specific conditional knockout mouse for Stat3 was generated by crossbreeding endothelial specific CreERT2 mice (Tie2-CreERT2 mice) with Stat3floxed/KO mice, which have the exon 22 of Stat3 flanked by LoxP sites. A phosphorylation of the tyrosine residue at position 705 in the protein domain encoded by the exon 22 is indispensable for activation of Stat3. We used the Tie2-CreERT2 mouse generated by Forde et al17 (MGI:2450312), which showed highly specific endothelial expression with negligible expression in hematopoietic cells. To test efficacy and specificity of endothelial-Tie2-CreERT2;Stat3floxed/KO after tamoxifen administration, we performed immunofluorescence costaining of pStat3(Y705) and caveolin-1. Phosphorylation of Stat3 was induced by ischemia. We observed a significant decrease of phosphorylated Stat3+/caveolin-1+ endothelial cells of Tie2-CreERT2;Stat3floxed/KO mice compared with their Tie2-CreERT2 littermates (Figure 1A and 1B). As expected, we observed no reduction of phosphorylated Stat3 in caveolin-1 negative or nonendothelial cells in endothelial-Tie2-CreERT2;Stat3floxed/KO mice (Figure 1A and 1C). However, we did observe pStat3+/caveolin-1+ cells in endothelial Tie2-CreERT2;Stat3floxed/KO mice with tamoxifen treatment and ischemia. In endothelial cells, the loss of pStat3(Y705) was still evident at 28 days.
Increased Gene Expression and Stat3 Phosphorylation 28 Days After Ischemia
We determined the expression of Stat3 mRNA after a 30-minute left MCAo and differing reperfusion times up to 28 days after ischemia. In the ischemic hemisphere, there was a strong increase in Stat3 expression. The mRNA level was highest 2 days after reperfusion (4-fold induction) and then declined again at 3 days after reperfusion, but levels were still elevated at 28 days (Figure 1D). No induction took place at any time in the nonischemic hemisphere. The amount of activated Stat3 in endothelial cells was determined by immunofluorescence costainings for pStat3 and caveolin-1 at 2 and 28 days after cerebral ischemia. In endothelial cells of the peri-infarct area, Stat3 was still being phosphorylated at 28 days. Furthermore, at this time point, significantly more endothelial cells were immunoreactive for pStat3 than had been the case 2 days after ischemia (Figure 1E).
Conditional Ablation of Endothelial Stat3 Causes Worse Long-Term Stroke Outcome
Endothelial-Tie2-CreERT2;Stat3floxed/KO and Tie2-CreERT2 littermates received tamoxifen for 5 consecutive days and were subjected 2 days later to 30 minutes of MCAo (Figure 2A). In magnetic resonance imaging 2 days after ischemia, no difference was detected between the genotypes (Figure 2B and 2C). A subset of both genotypes (n=12 Tie2-CreERT2 littermates and n=9 Tie2-CreERT2;Stat3floxed/KO mice) was killed for histological evaluation of lesion size at 2 days after ischemia. NeuN immunohistochemistry and quantification of the lesion size consistently failed to reveal any difference in the degree of neuronal loss between endothelial-Tie2-CreERT2;Stat3floxed/KO and Tie2-CreERT2 littermates (Figure 2D and 2E). However, at 28 days after ischemia, we observed that genomic ablation of Stat3 had led to lesion sizes that were larger than those in Tie2-CreERT2 littermates (Figure 3A and 3B). At 14 days after 30-minute MCAo, functional outcome as tested by behavioral analysis was worse. In the Rotarod test, the time-to-drop was significantly lower (Figure 3C) and the extended neuroscore in endothelial-Tie2-CreERT2;Stat3floxed/KO mice was lower (Figure 3D). Taken together, ablation of endothelial Stat3 had no effect on short-term poststroke outcome. In fact, at 14 days after stroke, the functional outcome was worse, and at 28 days after stroke, lesion volume was greater.
Endothelial Stat3 Promotes Angiogenesis After Cerebral Ischemia
Next, we analyzed angiogenesis after cerebral ischemia and endothelial Stat3 ablation. Vessel density was determined by measuring the caveolin-1 positive area (see Methods section for details). We observed at 28 days after reperfusion that endothelial ablation of Stat3 had reduced vessel density (Figure 4A and 4C). The number of caveolin-1/bromodeoxyuridine double-positive cells was significantly reduced by 60% (P<0.005; Figure 4B and 4D), indicating a reduced proliferation of endothelial cells. Vessel density in peri-infarct striatal regions was reduced as visualized after 3-dimensional rendering of confocal z-stacks (Figure 4E and 4F and Movie I in the online-only Data Supplement). Il6 induces the expression of Vegf by activating the Stat3 signaling pathway. At 2 and 28 days after MCAo, we measured the serum levels of Il6 and Vegf using enzyme linked immunosorbent assay. There was no difference between the 2 genotypes, either at 2 or 28 days, for either Il6 or Vegf (Figure IA and IB in the online-only Data Supplement).
Endothelial Stat3 Ablation Is Followed By Increased Percentage of CD11b+/CD45low Cells and of CD11b+/CD45high Cells
With fluorescence-activated cell sorting analysis at 4 days after MCAo, we determined the percentage of CD11b+/CD45low and CD11b+/CD45high cells in the total cell population. It is generally assumed that CD11b+/CD45low cells are microglial cells and that CD11b+/CD45high cells represent invading myeloid cells. We found an increase of CD11b+/CD45low cells in the ischemic hemisphere compared with the contralateral hemisphere, and this increase was significantly higher after endothelial Stat3 ablation (Figure 4G and 4H). In Tie2-CreERT2 littermates, we found a slight trend toward increased invasion of myeloid cells in the ischemic hemisphere compared with the contralateral side, whereas a significant increase of invading myeloid cells was evident in endothelial-Tie2-CreERT2;Stat3floxed/KO. However, the number of myeloid cells in the ischemic hemisphere did not differ significantly between endothelial-Tie2-CreERT2;Stat3floxed/KO and Tie2-CreERT2 littermates.
Endothelial Stat3 Ablation Results in an Antiangiogenic, Axon Growth-Inhibiting and Extracellular Matrix–Degrading Micromilieu
We analyzed the transcriptional profile of CD31-enriched cells at 4 days after cerebral ischemia and endothelial Stat3 ablation. Our isolation and enrichment method maintains the interaction between endothelial cells and adjacent cells. We determined the cellular composition of the isolated cells by immunostaining cytospins after CD31-MACS and found endothelial cells to be the dominant fraction (Figure 5A). We found marker genes for endothelial cells, neurons, astrocytes, and oligodendrocytes to be present in the isolated RNA (Figure 5B). In summary, the isolated cell population reflects the cellular composition of the neurovascular niche.
The expression of 255 genes differed significantly (by >2-fold) between the genotypes (Table I in the online-only Data Supplement). Gene arrays were performed from 3 independent experiments with 7 animals per experiment and genotype (in total, 21 Tie2-CreERT2;Stat3floxed/KO and 21 Tie2-CreERT2 littermates). Overall, we found a complex network response with an increase of antiangiogenic factors and a decrease of angiogenic factors after endothelial Stat3 ablation (Figure 5C and Table 1). Moreover, we observed an increase of factors that inhibit axonal growth and a decrease of factors that promote axonal growth (Figure 5D and Table 2). We also identified a decreased expression of several components of the extracellular matrix (ECM) and an induction of ECM-degrading enzymes. Interestingly, Adamts9, a potent antiangiogenic and ECM-degrading enzyme, is strongly induced after endothelial Stat3 ablation. In the rat it is reported to be expressed exclusively in neurons and astrocytes directly after cerebral ischemia.22 In immune staining we found Adamts9 to be expressed mainly in endothelial cells of the peri-ischemic area at 4 days after ischemia (Figure 6A). In addition, we found a slight but significant increase of reactive oxygen species–producing genes like iNOS and NOX4 and induced expression of CD14 (Table I in the online-only Data Supplement). These changes point to an increased inflammatory activation after endothelial Stat3 ablation.
In summary, endothelial Stat3 ablation leads to a shift of the transcriptional profile within the neurovascular niche toward an antiangiogenic and axon growth-inhibiting micromilieu. Moreover, we found a shift toward an ECM-degrading micromilieu.
Endothelial Stat3 Ablation Changes the Glycosylation Pattern of the ECM After Cerebral Ischemia
The composition of the ECM changes after cerebral ischemia.23 Prominently, the expression of proteoglycans is increased. This type of ECM composition hinders growing axons from entering the lesion volume and thereby impairs axonal plasticity and functional recovery. The axon growth-inhibiting properties of the proteoglycans are conditioned by the glycosylation pattern of the glycosaminoglycan side chains.23 Isolectin-B4 is a lectin derived from Griffonia simplicifolia that binds terminal α-d-galactosyl residues, which are present mainly on the surface of endothelial and monocytic cells. It has been shown that isolectin-B4 binds the proteoglycan versican.24 We stained the brains with isolectin-B4 at 28 days after cerebral ischemia to identify functional changes in ECM in the genotypes. With isolectin-B4 we found an increased staining of the ECM in the ischemic area (Figure 6B). Moreover, the volume of the isolectin-B4 positive ECM area increased after endothelial Stat3 ablation (Figure 6B and 6C).
The major finding of this study is that activation of endothelial Stat3 is crucial for long-term outcome after stroke. We generated an endothelial-specific and inducible knockout mouse model by crossbreeding mice (Tie2-CreERT2;Stat3floxed/KO) to analyze the function of endothelial Stat3 in stroke. Endothelial Stat3 ablation caused increased lesion size in the long term after stroke but not in the short term. Angiogenesis was reduced, whereas the percentage of microglia increased, indicating microglia proliferation. The expression profile within the neurovascular niche was changed to an antiangiogenic, axon growth inhibiting micromilieu with ECM degradation. Most prominently, expression of the antiangiogenic and proteoglycan-cleaving Adamts9 was induced after endothelial Stat3 ablation and cerebral ischemia. We found Adamts9 to be expressed mainly in endothelial cells in the peri-ischemic area. Moreover, after endothelial Stat3 ablation, we found altered glycosylation patterns of the ECM. In the end, functional outcome was worse after endothelial Stat3 ablation and cerebral ischemia.
Short-term analysis of lesion volumes 2 days after reperfusion revealed no differences between the genotypes, either in magnetic resonance imaging or in histology, after 30 minutes of MCAo (Figure 2A through 2D). This confirms previous data on the kinetics of Stat3 phosphorylation: Stat3 is highly induced and activated in neurons in the ischemic core and peri-infarct region within 8 hours after stroke.8 Stat3 exerts its antiapoptotic effects in neurons.11,25,26 Stat3 phosphorylation in contrast does not start in endothelial cells before 48 hours after ischemia.8 An endothelial-specific knockout of Stat3 should therefore not impact the acute outcome after stroke. Our results thus fit nicely together with the known kinetics of endothelial Stat3. We compared genotype lesion sizes at 28 days after ischemia and detected yet upregulated and phosphorylated Stat3 in endothelial cells (Figure 1D and 1E). Endothelial Stat3 ablation increased the lesion volume at this later time point (Figure 3A and 3B).
At 28 days after ischemia, vessel density (Figure 4A and 4C) and endothelial cell proliferation (Figure 4B and 4D) were reduced after endothelial Stat3 ablation. After cerebral ischemia, Stat3 is activated by several inflammatory factors, including Il6, Lif, and Cntf.27–29 Activation of Stat3 by Il6 is followed by increased angiogenesis, which is in part mediated by transcriptional induction of Vegf.30 This led us to seek differences in Il6 and Vegf serum levels after endothelial knockout of Stat3. Contrary to our expectations, however, endothelial ablation of Stat3 had no impact on systemic Il6 or Vegf levels (Figure IA and IB in the online-only Data Supplement). Endothelial Stat3 is apparently not involved in elevated Il6 serum levels after cerebral ischemia, and the induction of angiogenesis after stroke by Stat3 activation is independent of Vegf.
We established a method to isolate endothelial cells and adjacent cells of the neurovascular niche after endothelial Stat3 ablation and cerebral ischemia and we analyzed the transcriptional profile of this cell population by gene array analysis. Capillary endothelial cells are embedded in a firm basal membrane consisting mainly of collagens. The triple helix structure of collagens can be destroyed by collagenase digestion but is resistant to cleavage by trypsin, pepsin, or papain.31 Therefore, we digested the brain matrix with papain to maintain the interaction between the basal membrane of the vasculature and the adherent cells of the neurovascular niche. Endothelial cells and all adherent cells were isolated by CD31-MACS. We chose a time point 4 days after ischemia, when angiogenesis and axonal growth are still going on but mechanisms of the acute phase after ischemia have already ceased.
In the CD31-enriched population 4 days after cerebral ischemia, we identified 255 genes that showed differing expression levels between genotypes (Table I in the online-only Data Supplement). Overall, endothelial Stat3 ablation led to an antiangiogenic and axon growth–inhibiting micromilieu (Figure 5A and 5B). Interestingly, endothelial Tie2-CreERT2;Stat3floxed/KO mice showed higher expression of apelin, apelin receptor, and placental growth factor, all potent angiogenic factors, than did their Tie2-CreERT2 littermates. This might indicate a compensatory activation of alternative angiogenic pathways that, after activation, however, did not change angiogenesis. It is known that, after ischemia, the outgrowing vessels support neuroblast migration in a scaffold-like manner, directing migration to the damaged areas.2 The reduction of angiogenesis after endothelial Stat3 ablation precludes this process and might thus contribute to impaired neuroplasticity.
Most prominent was an increased expression of Adamts9 with strong antiangiogenic activity.32 It is reported that Adamts9 in the rat is mainly expressed in neurons 24 hours after cerebral ischemia.22 We found its expression mainly in endothelial cells in the peri-ischemic tissue 4 days after ischemia. Adamts proteases are classified in 3 subfamilies based on their preference for cleaving specific ECM components.33 Adamts9 belongs to the proteoglycanases and degrades chondroitin sulfate proteoglycans. Chondroitin sulfate proteoglycans are major components of the ECM and are characterized by a core protein with at least 1 covalently bound glycosaminoglycan side chain and a variable number of N- and O-linked oligosaccharides.
After cerebral ischemia, the composition of the ECM is changed to a more juvenile matrix type.34 Therefore, the ECM first has to be degraded. In proteoglycans this is performed predominantly by Adamts proteases. Finally, the expression of all proteoglycans is much higher after ischemia, and they contribute to a large extent to the development of the ECM component in the lesion volume. The chronic lesion volume on the one hand improves the physical stability of the ischemic area; on the other hand it strongly impairs neuroplasticity.23 Thus, digestion of proteoglycans by Adamts proteases is thought to have beneficial effects.33 Endothelial Stat3 ablation led to an intense upregulation of Adamts9, which quite specifically cleaves the long variants of the proteoglycans, aggrecan and versican, but not the short proteoglycan brevican. However, the expression of the short proteoglycans neurocan and brevican was reduced, as was the expression of tenascin-R. Also reduced was the expression of phosphacan and its cellular binding partners Ncam, Nrcam, and contactin-2.
In general, the expression pattern of the matrix within the neurovascular niche was shifted toward the juvenile matrix type with a looser mesh-size and thus axon growth-promoting ECM environment, with less intercellular ECM binding after endothelial Stat3 ablation. Nonetheless, the expression of other axon growth-inhibiting genes and the suppression of growth-promoting genes seemed to outweigh this effect after endothelial Stat3 ablation. Adamts9 expression was not able to reduce chronic lesion size or improve functional outcome. Furthermore, in isolectin-B4 staining we found an increase of α-d-galactosyl residues in the chronically damaged tissue. The axon growth-inhibiting activity of proteoglycans is partly mediated by their glycosylation patterns. Endothelial Stat3 might be highly relevant for regulation of the glycosylation pattern of the ECM in the chronic lesion and might thereby influence axon growth and neuroplasticity.
Beyond that, we found a slightly increased invasion of monocytes 4 days after cerebral ischemia and endothelial Stat3 ablation. Proliferation of microglia was increased. The expression of reactive oxygen species–building genes, such as iNOS and NOX4, as well as CD14, was higher, which might contribute to delayed neuronal cell death and increased lesion size in the long term after ischemia.
As a consequence, treatment strategies that focus on Stat3 signaling of endothelial cells offer the following 3 advantages: 1) accessibility for intravenous treatment, 2) treatment options in the acute phase of stroke beyond the narrow time window for thrombolysis, and 3) pleiotropic neuroregenerative action with widespread consequences for brain plasticity through changes in the microenvironment. Although this study provides substantial insight into the consequences of endothelial Stat3 signaling on long-term outcome after cerebral ischemia, further studies need to address the beneficial effects of endothelial-specific Stat3 signaling as a treatment option. We suggest using drugs that do not pass the blood–brain barrier to achieve an endothelial-specific stimulation of Stat3 and avoid activation of microglial Stat3, which might have detrimental consequences.
Expert technical assistance by Monika Dopatka, Janet Lips, and Nadine Weser is greatly acknowledged. We are grateful to Catherine Aubel for editing assistance.
Sources of Funding
This work was supported by grants from the SFB TR43 (The Brain as a Target of Inflammation) to M.E., U.D., H.K., S.A.W. and C.H. from the Deutsche Forschungsgemeinschaft. This work was supported by the German Research Foundation (Exc 257) and the Federal Ministry of Education and Research (01 EO 08 01, Center for Stroke Research Berlin, to M.E., U.D., and C.H.).
The online-only Data Supplement is available with this article at http://circ.ahajournals.org/lookup/suppl/doi:10.1161/CIRCULATIONAHA.114.013003/-/DC1.
- Received September 5, 2014.
- Accepted March 13, 2015.
- © 2015 American Heart Association, Inc.
- Adams RH,
- Eichmann A.
- Ohab JJ,
- Fleming S,
- Blesch A,
- Carmichael ST.
- Funamoto M,
- Fujio Y,
- Kunisada K,
- Negoro S,
- Tone E,
- Osugi T,
- Hirota H,
- Izumi M,
- Yoshizaki K,
- Walsh K,
- Kishimoto T,
- Yamauchi-Takihara K.
- Yahata Y,
- Shirakata Y,
- Tokumaru S,
- Yamasaki K,
- Sayama K,
- Hanakawa Y,
- Detmar M,
- Hashimoto K.
- Jung JE,
- Kim GS,
- Narasimhan P,
- Song YS,
- Chan PH.
- Yamashita T,
- Sawamoto K,
- Suzuki S,
- Suzuki N,
- Adachi K,
- Kawase T,
- Mihara M,
- Ohsugi Y,
- Abe K,
- Okano H.
- Dziennis S,
- Jia T,
- Rønnekleiv OK,
- Hurn PD,
- Alkayed NJ.
- Gertz K,
- Kronenberg G,
- Kälin RE,
- Baldinger T,
- Werner C,
- Balkaya M,
- Eom GD,
- Hellmann-Regen J,
- Kröber J,
- Miller KR,
- Lindauer U,
- Laufs U,
- Dirnagl U,
- Heppner FL,
- Endres M.
- Takeda K,
- Kaisho T,
- Yoshida N,
- Takeda J,
- Kishimoto T,
- Akira S.
- Irizarry RA,
- Hobbs B,
- Collin F,
- Beazer-Barclay YD,
- Antonellis KJ,
- Scherf U,
- Speed TP.
- Jung JE,
- Kim GS,
- Chan PH.
- Kinouchi T,
- Kitazato KT,
- Shimada K,
- Yagi K,
- Tada Y,
- Matsushita N,
- Sumiyoshi M,
- Satomi J,
- Kageji T,
- Nagahiro S.
- Ali C,
- Nicole O,
- Docagne F,
- Lesne S,
- MacKenzie ET,
- Nouvelot A,
- Buisson A,
- Vivien D.
- Kadler KE,
- Baldock C,
- Bella J,
- Boot-Handford RP.
The number of patients experiencing ischemic stroke is increasing because of an aging population. Acute treatment is limited to thrombolysis, which has a short therapeutic window (4.5 hours) and poor accessibility. Poststroke angiogenesis correlates with recovery and might be causally related to functional brain recovery. Signal transducer and activator of transcription-3 ablation, which in mice was restricted to endothelial cells and induced only a few days before focal brain ischemia, increased lesion size and worsened long-term functional outcome after stroke but had no effect in the short term. Signal transducer and activator of transcription-3 ablation reduced angiogenesis and changed the transcriptional profile within the neurovascular niche to an antiangiogenic and axon growth–inhibiting micromilieu with increased remodeling of the extracellular matrix and altered glycosylation patterns of the extracellular matrix. This is the first study that provides a link between a pure endothelial intervention and neuroplasticity. We identified endothelial signal transducer and activator of transcription-3 as a key factor for angiogenesis and modification of the microenvironment of the neurovascular niche. Drugs targeting endothelial signal transducer and activator of transcription-3 have the following potential advantages: extended treatment window (for up to 48 hours after stroke onset) and endoluminal access (oral or intravenous); thus, there is no need for penetration of the blood–brain barrier. Our study improves our understanding of the molecular mechanisms underlying poststroke recovery in mice. In particular, this novel mechanism provides evidence for a causal role of angiogenesis in neuronal recovery and supports the concept of endothelial targeting of drugs for the treatment of cerebral ischemia.