Endothelial peroxiredoxin-4 is indispensable for blood–brain barrier integrity and long-term functional recovery after ischemic stroke
Edited by Jeremy Nathans, Johns Hopkins University School of Medicine, Baltimore, MD; received January 15, 2024; accepted January 16, 2024
Significance
Using endothelium-specific gene knockout or overexpression, we demonstrate that Prx4 protects animals against blood–brain barrier (BBB) damage induced by cerebral ischemia/reperfusion (I/R) injury. Endothelial Prx4 preserves the functional integrity of the BBB by inhibiting myosin light chain/stress fiber formation and tight-junction protein disassembly during the early stages of cerebral I/R. Prx4-afforded endothelial protection blunts endothelial inflammation and leukocyte infiltration, thereby tempering cerebral inflammation and improving long-term stroke outcomes. These results uncover a critical role for Prx4 as an essential safeguard of the BBB after ischemic/reperfusion brain injury.
Abstract
The endothelial lining of cerebral microvessels is damaged relatively early after cerebral ischemia/reperfusion (I/R) injury and mediates blood–brain barrier (BBB) disruption, neurovascular injury, and long-term neurological deficits. I/R induces BBB leakage within 1 h due to subtle structural alterations in endothelial cells (ECs), including reorganization of the actin cytoskeleton and subcellular redistribution of junctional proteins. Herein, we show that the protein peroxiredoxin-4 (Prx4) is an endogenous protectant against endothelial dysfunction and BBB damage in a murine I/R model. We observed a transient upregulation of Prx4 in brain ECs 6 h after I/R in wild-type (WT) mice, whereas tamoxifen-induced, selective knockout of Prx4 from endothelial cells (eKO) mice dramatically raised vulnerability to I/R. Specifically, eKO mice displayed more BBB damage than WT mice within 1 to 24 h after I/R and worse long-term neurological deficits and focal brain atrophy by 35 d. Conversely, endothelium-targeted transgenic (eTG) mice overexpressing Prx4 were resistant to I/R-induced early BBB damage and had better long-term functional outcomes. As demonstrated in cultures of human brain endothelial cells and in animal models of I/R, Prx4 suppresses actin polymerization and stress fiber formation in brain ECs, at least in part by inhibiting phosphorylation/activation of myosin light chain. The latter cascade prevents redistribution of junctional proteins and BBB leakage under conditions of Prx4 repletion. Prx4 also tempers microvascular inflammation and infiltration of destructive neutrophils and proinflammatory macrophages into the brain parenchyma after I/R. Thus, the evidence supports an indispensable role for endothelial Prx4 in safeguarding the BBB and promoting functional recovery after I/R brain injury.
Sign up for PNAS alerts.
Get alerts for new articles, or get an alert when an article is cited.
Loss of blood–brain barrier (BBB) integrity is a grave consequence of cerebral ischemia and can impede the natural recovery of neurological function. After BBB damage, plasma proteins, proinflammatory immune cells, and fluids leak out of blood vessels into the parenchymal extracellular space, resulting in cerebral inflammation, tissue injury, and chronic neurological deficits (1–5). Thus, it is imperative to understand the biological processes underlying BBB damage and to identify protective agents that reduce brain injury after ischemic stroke and improve neurological functions for the long term.
The structural and functional integrity of the BBB is maintained by a tight junction (TJ)-sealed capillary endothelial cells (ECs), astrocyte end-feet, pericytes, and the extracellular matrix (ECM) (1). TJ proteins and adherent proteins in EC membranes are normally anchored to the cytoskeleton of ECs (6, 7). However, after cerebral ischemia and reperfusion, the BBB is attacked from multiple angles. For example, matrix metalloproteinases (MMPs) derived from neutrophils cause degradation of TJ proteins and infiltration of immune cells and inflammatory cytokines into the brain parenchyma after stroke. MMPs can induce irreversible BBB damage by 24 to 48 h after ischemia (8–10). Prior to this relatively delayed disassembly of TJ complexes and disruption of the BBB, rapid and subtle changes are observed in the endothelium, including higher caveolin-1-mediated transcellular vesicular transport, leading to BBB hyperpermeability as early as 6 h after cerebral I/R (11–13). Our recent studies demonstrate that ischemia induces rearrangement of TJ proteins and actin filaments in ECs by activating the Rho-associated protein kinase (ROCK)/myosin light chain (MLC) signaling cascade, resulting in the opening of endothelial paracellular junctions as early as 30 min after cerebral I/R (1). Endogenous actin-depolymerizing factors that suppress actin polymerization and TJ protein rearrangement may reverse this BBB damage during the early stages of ischemic injury (5, 14). Therefore, identifying endogenous inhibitors of actin polymerization in ECs may reveal potential biological targets for stroke therapies.
Peroxiredoxins (Prxs) are an ancient and ubiquitous family of antioxidant enzymes (15), with six isoforms (Prx1-Prx6) in mammalian cells. Prxs mediate multiple functions, such as protection against oxidative stress, inflammation, atherosclerosis, heart disease, and metabolic disorders (16, 17). Prxs also moderate central nervous disorders, such as Parkinson’s disease, traumatic brain injury, and brain ischemia (18–20), mainly through their antioxidative functions. Prx4 is a ubiquitously expressed member of the peroxiredoxin family (18) and can reduce hydrogen peroxide to water via thio-dependent catalytic actions (4–6). Prx4 is unique in that it is expressed in the endoplasmic reticulum, where it plays roles in antioxidant defense, protein folding, cell signaling, and cellular protection (19, 20). Prx4 is also highly expressed in peripheral infiltrated mesenchymal stromal cells (MSCs) and protects the integrity of the BBB after ischemia (21, 22). In the current study, we identify that the expression of Prx4 is transiently elevated exclusively in ECs in the brain after experimental stroke. These findings suggest that Prx4 could serve as a biological target to protect the integrity of the cerebral vasculature after ischemic brain injury. To our knowledge, the potential role of EC-specific Prx4 in BBB protection following I/R injury had not been reported previously. Thus, we used cell-targeted, in vivo genetic manipulations to test the hypothesis that EC-specific Prx4 expression protects against I/R-induced BBB damage and sustains long-term neurological recovery.
Results
Prx4 Is Specifically Up-regulated in ECs after Stroke In Vivo and Prevents Breach of the Endothelial Lining after Oxygen–glucose Deprivation In Vitro.
Prx4 protein expression was analyzed by immunostaining coronal brain sections of sham-operated mice and stroke mice at 1, 3, 6, and 24 h after transient focal cerebral ischemia (tFCI/R) (Fig. 1 A–D). Prx4 immunofluorescence was undetectable in noninjured CD31+ ECs, despite western blot analyses demonstrating Prx4 expression in nonischemic microvessel lysates (SI Appendix, Fig. S1D) and total brain lysates (SI Appendix, Fig. S9 C and D). Prx4 immunofluorescence was up-regulated in CD31+ ECs in the infarct core area at 6 (P < 0.0001) and 24 h (P = 0.0013) after tFCI/R (Fig. 1 C and D), notwithstanding the decrease of CD31+ vessels (SI Appendix, Fig. S1A). Prx4 immunofluorescence displayed little overlap with NeuN (marker of mature neurons), GFAP (marker of astrocytes), Iba1 (marker of microglia and monocyte-derived macrophages), or Olig2 (the oligodendroglial lineage marker) (SI Appendix, Fig. S1B). Western blot analysis was performed using isolated microvessels from nonischemic brains and brains collected 3, 6, and 24 h after tFCI/R (SI Appendix, Fig. S1D). Prx4 protein was detected in all samples but was up-regulated by fourfold at 6 h after tFCI/R (P < 0.0001).
Fig. 1.
To understand the function of Prx4 in ECs subjected to tFCI/R injury, an in vitro BBB model was utilized. A monolayer of human brain microvascular endothelial cells (HBMECs) was cultured on top of a porous membrane in transwells (SI Appendix, Fig. S1E) and was transduced with lentiviral particles carrying Prx4 to force Prx4 overexpression (SI Appendix, Fig. S1F). At 48 h after transduction, confluent HBMECs were subjected to oxygen–glucose deprivation/reperfusion (OGD/R) for 1 h followed by reperfusion. No differences in cell viability were observed before or after OGD/R (SI Appendix, Fig. S1G). In this in vitro BBB model, the fluorescence intensities of 0.95-kDa cadaverine and 4.4-kDa dextran, but not 70-kDa dextran, were increased in the media of the abluminal chamber at 1 and 3 h after OGD/R. The appearance of fluorescence in this abluminal chamber models the extravasation or leakage of tracer from the interior of the vessel (luminal chamber) into the brain parenchyma. Lenti-Prx4 transfection inhibited OGD/R-induced BBB leakage of cadaverine and 4.4-kDa dextran (SI Appendix, Fig. S1H). In contrast, lentivirus-mediated transfection of catalase, another hydrogen peroxide-scavenger like Prx4, failed to reduce OGD/R-induced BBB leakage of cadaverine or 4.4-kDa dextran (SI Appendix, Fig. S1I) Thus, endothelial Prx4 overexpression preserves BBB integrity under OGD/R conditions, and the underlying mechanism is independent of its antioxidant properties.
Specific Deletion of Prx4 from ECs In Vivo Exacerbates BBB Damage by Enhancing Redistribution of Junctional Proteins in Brain Vessels after Stroke.
Despite the lack of Prx4 detection in nonischemic ECs using immunohistochemistry, Prx4 is detectable by western blots in uninjured ECs in the absence of stroke (SI Appendix, Fig. S1D). We therefore sought to test whether endothelial Prx4 protects against BBB damage induced by tFCI/R by targeted knockout. We constructed a tamoxifen-inducible conditional knockout mouse strain that selectively deletes Prx4 from ECs (Prx4 eKO) (SI Appendix, Fig. S2 A and B). To confirm that TEK-CreER-driven gene knockout occurred exclusively in ECs, the TEKCreERT2 mice were crossed with Ai14 mice to obtain EC-RFP reporter mice (SI Appendix, Fig. S2C). After tamoxifen induction, RFP expression was confirmed as exclusively colocalized with EC marker CD31 (SI Appendix, Fig. S2 D and E). The level of endothelial Prx4 protein was reduced by 74.3% in Prx4 eKO mice compared to the wild type 24 h after tFCI/R (P = 0.0301) (SI Appendix, Fig. S2 F and G).
The morphology and density of cerebral microvessels (SI Appendix, Fig. S3 A–C) and neurovascular coupling responses (SI Appendix, Fig. S3 D–F) were then examined in nonstroke mice, which revealed no differences between Prx4 eKO mice and WT littermates. Thus, genetic knockout of Prx4 within endothelia does not significantly affect the gross anatomy or physiology of brain microvasculature under baseline (homeostatic) conditions.
We then induced tFCI/R in Prx4-eKO and WT mice to evaluate the effects of Prx4 eKO on ischemic brain injury and neurological outcomes. In pilot studies, we found that the Prx4-eKO mice exhibited elevated vulnerability to tFCI/R compared to WT mice. Thus, we employed a shorter duration of tFCI (45 min rather than 60 min). Prx4-eKO and WT mice showed comparable reductions in cortical cerebral blood flow (SI Appendix, Fig. S4 A–D) and body weight (SI Appendix, Fig. S4E), mortality rate (SI Appendix, Fig. S4F), blood gas, and other hematological parameters (SI Appendix, Table S2). These results suggest that comparable levels of brain ischemia were induced in Prx4-eKO and WT mice. Thus, differences across eKO and WT groups cannot be attributed to gross morphological changes in brain vasculature or different extents of the original ischemic insult.
We evaluated the effects of Prx4 eKO on acute BBB leakage 1 to 24 h after tFCI/R. The degrees of extravasation of fluorescent tracer Alexa 555 cadaverine (0.95 kDa) and endogenous plasma IgG (∼150 kDa) into the brain parenchyma were quantified 1, 3, and 24 h after tFCI/R (Fig. 1 E and F and SI Appendix, Fig. S5 A–D). tFCI/R induced cadaverine leakage across the BBB of WT mice within 1 h after tFCI/R, showing fluorescence leakage predominantly into the striatum of the ischemic hemisphere (Fig. 1E and SI Appendix, Fig. S5A). At 3 and 24 h after tFCI/R, BBB leakage of cadaverine and plasma IgG became more obvious and encompassed both striatum and cortex (Fig. 1E and SI Appendix, Fig. S5A). Compared to WT mice, Prx4-eKO mice exhibited greater BBB leakage volumes for cadaverine and plasma IgG at 3 and 24 h after tFCI/R (Fig. 1F). IgG leakage was predominant in the external capsule (ExC) of stroke mice, consistent with the high vulnerability of white matter to stroke. Prx4-eKO mice also had increased brain infarct volumes (P = 0.0490) 24 h after tFCI/R compared to WT mice (Fig. 1H). Since no gross leakage of cadaverine was detected in the cortex 1 h after tFCI/R (Fig. 1E and SI Appendix, Fig. S5A), we evaluated subtle cadaverine leakage into parenchymal cells of the cortex (CTX), striatum (STR), and ExC (SI Appendix, Fig. S6 A and B). The number of cadaverine+ cells in the CTX and STR rose further in Prx4-eKO mice, at 1 h and 3 h after I/R (SI Appendix, Fig. S6C).
We further evaluated subtle leakage of the BBB by intravenously (i.v.) injecting a smaller tracer sodium fluorescein (NaFI, molecular weight: 376.3 Da) into stroke mice at the onset of reperfusion. NaFl fluorescence intensity was measured in homogenates of the ipsilateral cortex 1 h after tFCI/R. The fluorescence intensity of NaFI was higher in Prx4-eKO compared to WT mice (P = 0.0397) (Fig. 1G). Taken together, the results suggest that EC-specific Prx4 knockout aggravates BBB leakage 1 to 24 h after tFCI/R.
Our previous study has shown that actin polymerization and formation of F-actin stress fibers in ECs cause intracellular translocation of junctional proteins and increased BBB permeability within 1 h after stroke (1). Accordingly, we evaluated actin polymerization in purified cerebral microvessels from ipsilateral hemisphere of Prx4 eKO and WT stroke mice. As expected, actin polymerization, measured as a rise in F-actin/G-actin ratios (P = 0.0117), was induced in Prx4 eKO mice, but not in WT mice, 1 h after 45 min tFCI/R (Fig. 1I). The phalloidin staining for F-actin on brain sections showed a dramatic induction of stress fibers in CD31+ ECs across vessels of different sizes in Prx4-eKO mice 1 h after tFCI/R (Fig. 1J).
Excessive stress fibers lead to the disassembly of junctional proteins and membrane-to-cytosol translocation in ECs (1). Thus, we examined whether Prx4-eKO mice displayed greater junctional protein translocation. Subcellular protein fractionation was performed using purified cerebral microvessels 3 h after tFCI/R, and each fraction was probed with antibodies against zona occludens-1 (ZO-1) and occludin. The levels of ZO-1 and occludin remained unchanged in whole-cell lysates after tFCI/R in Prx4-eKO or WT mice (Fig. 1K). However, the levels of ZO-1 and occludin were increased in the actin cytoskeleton fraction (ACF) (Fig. 1L), and simultaneously decreased in the cell membrane fraction (Fig. 1M), in Prx4-eKO mice after tFCI/R. These results demonstrate that endothelial Prx4 protects against actin polymerization, redistribution of endothelial junctional proteins, and acute BBB leakage after tFCI/R.
Endothelial Prx4 Alleviates Stress Fiber Formation by Inhibiting MLC Phosphorylation after Stroke.
We then explore the mechanism underlying F-actin stress fiber formation after tFCI/R in Prx4 eKO mice, focusing on the activation (phosphorylation) of MLC (Fig. 2A). To test this possibility, triple-label immunofluorescent staining (p-MLC, lectin, and DAPI) was performed in brain sections from WT and Prx4 eKO mice 1 and 24 h after tFCI/R (Fig. 2B). The p-MLC immunofluorescence was not present in sham-operated brains but was markedly induced in all stroke mice 1 or 24 h after tFCI/R (Fig. 2C). The percentage of p-MLC+ vessels was increased in the ischemic core of Prx4-eKO mice 1 h after tFCI/R compared to WT mice (P = 0.0062, Fig. 2D), suggesting that Prx4 inhibits MLC phosphorylation within 1 h after ischemia.
Fig. 2.
To determine whether Prx4 regulates F-actin stress fiber formation via p-MLC after tFCI/R, ML-7, a selective MLC kinase inhibitor, was administered (1 mg/kg, i.p.) 5 min after reperfusion. ML-7 reduced the percentage of p-MLC+ microvessels in the ischemic areas of WT (P = 0.001) and Prx4 eKO (P < 0.0001) mice, respectively, 1 h after tFCI/R (Fig. 2C). ML-7 also significantly reduced p-MLC in WT (P = 0.0356) and Prx4-eKO (P = 0.0018) mice 24 h after tFCI/R (Fig. 2D). Consistent with a role of p-MLC in facilitating stress fiber formation, ML-7-treated Prx4-eKO mice showed reduced stress fibers 1 h (P = 0.0059, Fig. 2E and SI Appendix, Fig. S7A) and 24 h (P = 0.0029, SI Appendix, Fig. S7B) after tFCI/R.
To determine whether p-MLC mediates BBB leakage after tFCI/R, ML-7 was administered to WT and Prx4-eKO mice, and leakage volumes for cadaverine or IgG were measured after tFCI/R. ML-7 reduced the cadaverine leakage volume at 1 h after tFCI/R in both WT (P = 0.0345) and Prx4-eKO (P = 0.0436) mice, and at 24 h after tFCI/R in Prx4-eKO mice only (P = 0.0002) (Fig. 2 F and G). ML-7 significantly reduced IgG leakage volume in WT (P = 0.0251) and Prx4-eKO (P = 0.0011) mice 24 h after tFCI/R (Fig. 2 F and H). Pearson correlation coefficient analyses demonstrated a strong positive association between F-actin+ microvessels and cadaverine leakage 1 h (r = 0.5991, P = 0.0016, Fig. 2G) and 24 h (r = 0.6703, P = 0.0002, SI Appendix, Fig. S7C) after tFCI/R; there was a trend toward positive correlation between F-actin+ microvessels and IgG leakage volume 1 h after tFCI/R (r = 0.3885, P = 0.0550, Fig. 2H), but a strong positive correlation between F-actin+ microvessels and IgG leakage volume 24 h after tFCI/R (r = 0.5344, P = 0.0059, SI Appendix, Fig. S7C).
To further elucidate the regulatory role of Prx4 in MLC phosphorylation and BBB leakage after cerebral ischemia, the HBMEC in vitro BBB model was either transduced with lenti-Prx4 for 48 h to overexpress Prx4 (the empty vector served as control), with or without ML-7 (5 µM) treatment for 30 min prior to 1-h OGD (SI Appendix, Fig. S8 A and B). At 3 h after OGD, both Prx4 overexpression and ML-7 treatment abolished OGD-induced p-MLC expression (P < 0.0001) as well as leakage of the 4.4-kDa dextran through the BBB (P < 0.0001) (SI Appendix, Fig. S8 A and B). The next set of experiments examined the effect of MLC knockdown (lentivirus vector carrying MLC-targeting shRNA, MLCt), or Prx4 overexpression, or the combination of both, on ODG-induced stress fiber formation and BBB leakage. MLC knockdown and Prx4 overexpression alone could both suppress OGD-induced formation of F-actin+ stress fibers (P < 0.0001, SI Appendix, Fig. S8 C and E) and reduce leakage of 4.4-kDa dextran through the BBB (P < 0.0001, Fig. 3E), whereas the combination of both stimuli did not exert additive effects on stress fiber formation or BBB leakage (SI Appendix, Fig. S8 C and E).
Fig. 3.
Contractile stress fibers increase tensions on cell–cell junctions that are anchored to the actin cytoskeleton and, thus, disassemble the junctional complexes (23). We have shown previously that at 1 to 3 h after OGD, the adherens junction protein vascular endothelial (VE)-cadherin partially shifted from the membrane fraction to the actin cytoskeleton fraction, which could be prevented by inhibiting actin polymerization and stress fiber formation (1, 7). Our immunocytochemical results confirmed the decreased levels of VE-cadherin at its normal expression sites, the cell–cell contact zones, at 3 h after OGD (SI Appendix, Fig. S8 D and E), consistent with its redistribution after OGD. This change was markedly attenuated in HBMECs overexpressing Prx4 or with MLC knockdown (SI Appendix, Fig. S8 D and E).
Taken together, these results support the notion that Prx4 alleviates ischemia-induced BBB leakage by inhibiting MLC phosphorylation and stress fiber formation and the resulting junctional protein redistribution.
EC-targeted Prx4 Overexpression Attenuates BBB Leakage after Stroke.
To test the generalizability of our in vitro results showing the protective effects of endothelial Prx4 against BBB damage in HBMECs, we constructed a transgenic mouse strain with EC-targeted overexpression of Prx4 tagged by HA (Prx4-eTG mice, SI Appendix, Fig. S9 A and B). As shown by western blotting probing Prx4 as well as the HA tag, Prx4 was robustly expressed in the Prx4-eTG brain (SI Appendix, Fig. S9 C and D).
We then induced 60 min of tFCI/R in Prx4-eTG and WT mice to evaluate the effects of Prx4-eTG on ischemic brain injury and outcomes. Prx4-eTG and WT mice showed comparable reductions in cortical cerebral blood flow (SI Appendix, Fig. S10 A–C) and body weight (SI Appendix, Fig. S10D), and long-term mortality rates (SI Appendix, Fig. S10E). These results suggest that comparable levels of brain injury were induced in both Prx4-eTG and WT mice.
We evaluated the effects of Prx4-eTG on acute BBB leakage 1 to 24 h after tFCI/R. The extravasation of fluorescent tracer Alexa 555 cadaverine and endogenous plasma IgG into the brain parenchyma was quantified 1, 3, and 24 h after tFCI/R (Fig. 3 A–C). Prx4-eTG mice exhibited less leakage of cadaverine at 1 h (P = 0.0142), 3 h (P = 0.0016), and 24 h (P < 0.0001) after tFCI/R compared to WT mice (Fig. 3B). Prx4-eTG mice also showed lower IgG leakage volumes (P < 0.0001) (Fig. 3C) and brain infarct volumes (P = 0.0087) (Fig. 3D) 24 h after tFCI/R compared to WT mice. We also determined the in vivo effects of Prx4 eTG on MLC phosphorylation and F-actin+ stress fiber formation in brain microvessels after tFCI/R. Prx4-eTG mice showed lower p-MLC at 1 and 24 h (P < 0.0001) after tFCI/R compared to WT tFCI/R mice (Fig. 3 E and F). Prx4-eTG mice also showed less stress fiber formation (P = 0.0168) after tFCI/R, compared to WT mice (Fig. 3 G and H).
Prx4 Impedes the Infiltration of Peripheral Immune Cells into the Brain Parenchyma and Abolishes Neuroinflammation after Stroke.
Compromised BBB integrity in the acute stages of ischemic stroke facilitates the infiltration of leukocytes from the peripheral circulation into the brain parenchyma and aggravates the secondary brain injury (1). Thus far, we have established that endothelial Prx4 plays an essential role in moderating BBB permeability after tFCI/R. Therefore, we hypothesized that endothelial Prx4 mitigates immune cell infiltration by reducing BBB disruption after brain ischemia. To this end, flow cytometry was performed to quantify infiltrated leukocytes and lymphocytes in the ischemic brain hemisphere of Prx4-eKO and WT mice 3 d after FCI/R (Fig. 4A and SI Appendix, Fig. S11). Peripheral immune cells were minimally detectable in sham-operated WT and Prx4 eKO mice and not increased (P > 0.05) in WT mice after tFCI/R (Fig. 4A). However, several types of brain-infiltrating immune cells, including macrophages (P = 0.0044), neutrophils (P = 0.0016), T cells (P = 0.0072), and dendritic cells (P = 0.0050), were significantly higher in numbers in Prx4-eKO mice after tFCI/R compared to sham injury (Fig. 4A). Increased brain infiltration of macrophages and neutrophils in Prx4-eKO mice was confirmed by immunofluorescence staining of F4/80 (marker of macrophages) and Ly6G (marker of neutrophils) 24 h after 45 min-tFCI/R. The numbers of infiltrated macrophages (P = 0.0028) and neutrophils (P = 0.0076) were significantly higher in Prx4-eKO mice after tFCI/R compared to sham groups (Fig. 4 B and C).
Fig. 4.
We further tested the hypothesis using Prx4-eTG and WT mice subjected to 60 min-tFCI/R rather than the 45-min tFCI/R in earlier figures. Flow cytometry was performed to quantify the number of infiltrated immune cells in the ischemic brain 5 d after tFCI/R (Fig. 4D), which showed that under the standard 60-min tFCI/R paradigm, there was increased infiltration of macrophages, neutrophils, dendritic cells, and lymphocytes in both WT and Prx4 eTG mice after tFCI/R (Fig. 4D). The number of infiltrated macrophages (P < 0.0001) and neutrophils (P = 0.0002), but not dendritic cells (P = 0.8613) or lymphocytes, was significantly decreased in the ischemic hemisphere of Prx4-eTG mice compared to the WT mice (Fig. 4D). Immunofluorescence staining for F4/80 and Ly6G at 24 h after 60-min tFCI/R (Fig. 4E) confirmed that the early infiltration of macrophages (P = 0.0458) and neutrophils (P = 0.0103) was significantly decreased in Prx4-eTG mice compared to WT mice (Fig. 4F).
To determine whether Prx4-eTG-afforded attenuation of brain infiltration of peripheral immune cells after stroke lowers the inflammatory burden in the brain, inflammatory factor-protein microarray analysis from the ipsilateral brain tissue lysates was performed 5 d after tFCI/R. Stroke-induced elevation of several proinflammatory chemokines and cytokines, including CCL11, platelet factor-4 (PF4), IFN-γ, TNFRII, TNF-α, and CXCL13, was abolished in Prx4-eTG mice after tFCI/R compared to WT mice (Fig. 4G and SI Appendix, Fig. S12).
Prx4 Suppresses Inflammation of the Endothelial Lining in Brain Vasculature.
Inflammation of the microvascular wall and local production of proinflammatory cytokines and chemokines can contribute to secondary (and possibly permanent) BBB disruption after stroke. To explore the role of endothelial Prx4 in inflammation of the vasculature, we performed the NanoString assay for 770 immunity and/or inflammation-related genes using isolated cerebral microvessels from Prx4-eKO and WT mice, 24 h following sham operation or tFCI/R. Prx4 eKO exerted profound effects on gene expression profiles in both sham-operated, nonischemic mice and WT mice subjected to tFCI/R (SI Appendix, Fig. S13), increasing ratios between proinflammatory and anti-inflammatory gene expression levels. In particular, Prx4-eKO mice showed downregulation of anti-inflammatory genes (e.g., Creb1, Sumo1, Il15ra, Mavs, Rpl28, Rpl36al, Maff, and Tmem100) and upregulation of proinflammatory genes (e.g., Ms4a2, Tcl1, Ncaph, Sox10, Casp3, and Ltc4s) after tFCI/R compared to WT mice (SI Appendix, Fig. S13C). These results suggest that endothelial Prx4 is essential for prevention of proinflammatory responses under physiological and pathological conditions.
We also performed confirmational immunofluorescent staining for Tcl1, one of the proinflammatory factors found to be up-regulated in Prx4-eKO stroke mice by the NanoString assay. Tcl1+ cells were nondetectable in nonischemic mice but were increased in the abluminal side of blood vessels in the ipsilateral hemisphere of Prx4-eKO mice (SI Appendix, Fig. S13 D and E).
Endothelial Prx4 Critically Impacts Stroke Outcomes for the Long Term.
We hypothesized that the sequential actions by endothelial Prx4 on BBB leakage, immune cell infiltration, cerebral inflammation, and brain tissue protection during acute stages of ischemic injury might exert a substantial impact on long-term functional stroke outcomes. To test this hypothesis, Prx4-eKO and WT mice were subjected to 45-min tFCI/R. Sensorimotor deficits were analyzed using the adhesive removal and foot fault tests up to 35 d after tFCI/R; spatial cognitive functions were evaluated in the same cohort by the Morris water maze (MWM) test between 22 to 27 d after tFCI/R. As shown in Fig. 5 A and B, Prx4-eKO mice had worse sensorimotor deficits and poorer spontaneous recovery than WT mice up to 35 d after tFCI/R. In the MWM test, the 45-min tFCI/R did not impair spatial learning abilities of WT mice. However, Prx4-eKO mice had worse spatial learning abilities compared to WT mice after tFCI/R (P = 0.0421) (Fig. 5C). Neither spatial memory nor swimming speed was impaired in WT or Prx4-eKO mice after 45-min tFCI/R (Fig. 5 D and E). Chronic brain atrophy measured by MAP2 staining was larger in Prx4-eKO mice compared to WT mice (P = 0.0110) (Fig. 5F).
Fig. 5.
We also evaluated the behavior performance of Prx4-eTG and WT mice for up to 35 d after 60-min tFCI/R. As expected, WT mice exhibited robust deficits in the adhesive removal test, the foot fault test, and the MWM test after tFCI/R (Fig. 5 G–J). In contrast, Prx4-eTG mice performed better than WT mice in all neurobehavioral tests (Fig. 5 G–J). The chronic brain atrophy volume was approximately halved in Prx4-eTG mice compared to WT mice 35 d after tFCI/R (P < 0.0001; Fig. 5L).
Taken together, these results suggest a critical role for endothelial Prx4 expression in long-term stroke outcomes. Whereas cell-targeted deletion of endogenous Prx4 from ECs chronically exacerbates stroke outcomes and brain tissue loss, forced overexpression of the Prx4 transgene in ECs improves stroke outcomes and preserves brain tissue for the long term.
Discussion
In the present study, we used cell-targeted genetic manipulations to investigate the role of endothelial Prx4 in the maintenance of BBB integrity in the acute reperfusion period after ischemic injury and in the sustained protection of neurological function during the chronic reperfusion phase. Endothelial Prx4 diminishes BBB leakage as expected, as cell-targeted deletion of Prx4 within ECs exacerbated BBB damage within 24 h following I/R. Conversely, when Prx4 was specifically overexpressed within ECs, acute BBB damage was diminished. Second, Prx4 lowers BBB leakage by interfering with a cascade of detrimental events mediated by MLC activation, stress fiber formation, and TJ protein disassembly. Third, endothelial Prx4 prevents microvasculature inflammation and impedes extravasation of peripheral leukocytes into the brain parenchyma, which otherwise negatively impacts long-term stroke outcomes (Fig. 6). Thus, Prx4 may be an important therapeutic target for stroke with reperfusion.
Fig. 6.
During the early stages of cerebral reperfusion after stroke, BBB disruption is followed by the extravasation of blood components into the brain, thereby compromising not only neurovascular function but also raising the probability of permanent brain damage (11). While a wide array of molecules is believed to regulate the structural integrity and functional properties of the BBB in health and disease, the evidence presented here confirms an essential role for Prx4 in acute BBB protection and brain tissue preservation. PRX4 expression is up-regulated in cerebral ECs by 6 h after ischemic injury and reperfusion, but it is also present under homeostatic or baseline conditions. Due to its basal expression in the endothelia, Prx4 knockout worsens BBB leakage of small (<1 kDa) tracers as early as 1h after the ischemic insult. After reperfusion, BBB breakdown continues to progress and, within 3 to 24 h, large (>150 kDa) proteins can pass through the injured BBB. We report evidence that PRX4 expression impedes both types of leakage. Thus, endothelial PRX4 is critical for maintenance of a functional BBB during the early stages of reperfusion after stroke.
At least three mechanisms may mediate BBB leakage during the early stages (≤24h) of reperfusion injury—the caveolin, paracellular, and MMP pathways (11). Caveolin-dependent transcellular transport increases BBB permeability 6 to 24 after tFCI/R (13); however, a later study based on caveolin-1 knockout or knockdown experimental approaches failed to confirm the role of caveolin-1-dependent transcytosis in tFCI/R-induced BBB leakage (1). The paracellular pathway likely contributes to early stages of BBB leakage, between 0.5 and 24 h after I/R (1), wherein rapid actin polymerization and F-actin stress fiber formation is followed by aberrant subcellular redistribution of TJ proteins and loss of the endothelial seal (11). The MMP pathways depend on activation of MMP-9 and MMP-2, usually ~24 h after stroke or later, and cause irreversible BBB disruption (11). We have found that endothelial deletion of Prx4 hyperactivates the paracellular BBB leakage pathway within 1 to 3 h after experimental stroke, culminating in F-actin stress fiber formation and translocation of occludin and ZO-1 from the plasma membrane to cytoskeleton fractions. Conversely, overexpression of Prx4 attenuates the paracellular pathway for BBB leakage in vitro and in vivo.
After reperfusion injury, the protein that suppresses actin polymerization, ADF/cofilin, is quickly phosphorylated and disabled, resulting in conversion of monomeric G-actin to short-chain F-actin (1). Upon activation (phosphorylation), MLC promotes formation of dense stress fibers from short F-actin polymers and elicits actomyosin contraction and subcellular translocation of tight junction proteins—leading to BBB opening (14, 16). Our results support a role for the MLC pathway in Prx4-afforded BBB protection. First, EC-specific Prx4 knockout enhanced MLC phosphorylation/activation and F-actin stress fiber after tFCI/R (Fig. 2). In contrast, endothelial overexpression of Prx4 inhibited MLC phosphorylation, reduced F-actin stress fiber formation, and attenuated BBB leakage in cellular and animal models of I/R (Fig. 3). Second, shRNA-mediated MLC knockdown in ECs prevented F-actin stress fiber formation and subcellular translocation of TJ proteins, and reduced BBB leakage in vitro after OGD (SI Appendix, Fig. S8). The lack of additive effects of MLC knockdown and simultaneous Prx4 overexpression suggests that these factors converge on the same pathway when controlling stress fiber formation and BBB leakage. Third, the selective MLC inhibitor ML-7 decreased F-actin stress fiber formation and BBB leakage in I/R insult. The ability of the MLC kinase inhibitor to interfere with stress fiber formation and increased BBB leakage under conditions of endothelial Prx4 knockout suggests that the regulation of MLC lies downstream of endothelial Prx4 in vivo.
Endothelial Prx4 substantially reduces brain infiltration of peripheral immune cells after experimental stroke, especially proinflammatory macrophages and neutrophils (Fig. 4), perhaps through two interrelated mechanisms. Rapid opening of the BBB in the acute stage of reperfusion promotes immune cell infiltration into the ischemic parenchyma (1, 7). The infiltrating neutrophils and macrophages release MMP-9 and MMP-2, degrade the extracellular matrix and basal lamina, and amplify proinflammatory responses of the blood vessels (1). Endothelial Prx4 suppresses microvascular inflammation and production of chemokines, which are responsible for immune cell infiltration. Prior work also suggests that Prx4 directly suppresses nuclear factor-kappa B and inhibits inflammation in a diabetes model (24). Our NanoString results further demonstrate that Prx4-deficient microvessels shift toward a proinflammatory transcriptome. Accordingly, the protein array data confirmed that Prx4 suppressed proinflammatory cytokines and chemokines (e.g., CCL11, PF4, IFN-γ, TNFRII, TNF-α, and CXCL13) in brain microvessels that had been subjected to ischemia. Thus, endothelial PRX4 works through dual mechanisms—suppression of peripheral immune cell invasion and abolishment of proinflammatory responses within the cerebral vasculature.
One of the limitations of the present study is the emphasis on male animals, as sex dimorphisms are often observed in ischemic brain injury and recovery (25–27). Pilot work of ours suggests that healthy female mice are highly resistant to tFCI/R-induced infarction and BBB disruption compared to age-matched males. However, future studies to explore potential sex differences in PRX4-afforded protection against BBB damage are still warranted. The impact of aging on stroke-induced BBB disruption (28) was also not addressed here. We have also not yet tested delivery of membrane-permeable, recombinant PRX4 protein into ECs but we believe that such a translationally relevant approach may offer hope for stroke treatment in conjunction with recanalization therapy.
Conclusions
The results of this study support a role for endothelial PRX4 in alleviating early-onset BBB leakage after ischemic/reperfusion brain injury. PRX4 reduces phosphorylation and activation of myosin light chain, thereby suppressing actin polymerization, F-actin stress fiber formation, and translocation of junctional proteins in the cerebral vasculature. PRX4 also limits endothelial inflammation after ischemic/reperfusion injury and brain invasion of proinflammatory immune cells. These results reveal that endothelial PRX4 serves as a natural brake against BBB leakage, invasion of proinflammatory immune cells, inflammation of the cerebral vasculature, permanent loss of brain tissue, and chronic behavior deficits. Thus, endothelial PRX4 is a valid candidate for continued testing of stroke therapies.
Materials and Methods
All animals were housed in a temperature- and humidity-controlled facility with a 12-h light/dark cycle. Food and water were available ad libitum. All experimental procedures were in accordance with the NIH Guide for the Care and Use of Laboratory Animal and the ARRIVE guidelines (Animal Research: Reporting in Vivo Experiments) and approved by the University of Pittsburgh Institutional Animal Care and Use Committee. All efforts were made to minimize animal suffering and the number of animals used. Surgeries and all outcome assessments were performed by investigators blinded to mouse genotype and experimental group assignments.
For further key material and methodological details, see SI Appendix.
Statistical Analyses.
The Shapiro–Wilk test was performed to determine the normality of all data. Data are presented as mean ± SD for normal distribution and median (Max, Min) for non-normal distribution. Prism software (Version 9.3.1) was used for statistical analyses. The Student’s t test (parametric) or Mann–Whitney U test (nonparametric) was used for the comparison between two groups. The differences among three or more groups were analyzed with one-way ANOVA (parametric) followed by Bonferroni post hoc or Kruskal–Wallis test (nonparametric) followed by Dunn post hoc test of equal variances. Welch’s ANOVA followed by Dunnett’s post hoc was used for unequal variance. The correlation between F-actin and cadaverine or IgG leakage volume was analyzed by Pearson correlation analysis (normal distribution) or Spearman correlation analysis (non-normal distribution). Differences in means across groups with repeated measurements over time were analyzed by the repeated-measures two-way ANOVA followed by Bonferroni post hoc tests. A P value of ≤0.05 was considered statistically significant.
Data, Materials, and Software Availability
All study data are included in the article and/or SI Appendix.
Acknowledgments
This project was supported by the US NIH grant NS089534 (to J.C.), the US Department of Veterans Affairs (VA) Merit Review BX002495 (to J.C.), and the University of Pittsburgh School of Medicine. J.C. is the Richard King Mellon Professor of Neurology and a recipient of a VA Senior Research Career Scientist Award. M.V.L.B. is the Sylvia and Robert S. Olnick Professor of Neuroscience. We thank Patricia Strickler for administrative support.
Author contributions
M.V.L.B. and J.C. designed research; N.X., X.J., W.Z., Y.S., Q.Y., T.Y., and S.L. performed research; N.X., X.J., W.Z., X.H., and J.C. analyzed data; and N.X., X.J., W.Z., R.K.L., R.F.K., R.A.S., M.V.L.B., and J.C. wrote the paper.
Competing interests
The authors declare no competing interest.
Supporting Information
Appendix 01 (PDF)
- Download
- 8.74 MB
References
1
Y. Shi et al., Rapid endothelial cytoskeletal reorganization enables early blood-brain barrier disruption and long-term ischaemic reperfusion brain injury. Nat. Commun. 7, 10523 (2016).
2
A. Shindo et al., Biphasic roles of pentraxin 3 in cerebrovascular function after white matter stroke. CNS Neurosci. Ther 27, 60–70 (2021).
3
J. H. Seo et al., AKAP12 Supports Blood-Brain Barrier Integrity against Ischemic Stroke. Int. J. Mol. Sci. 21, 9078 (2020).
4
A. Nirwane, Y. Yao, Cell-specific expression and function of laminin at the neurovascular unit. J. Cereb. Blood Flow. Metab. 42, 1979–1999 (2022).
5
M. Kang, Y. Yao, Basement membrane changes in Ischemic stroke. Stroke 51, 1344–1352 (2020).
6
L. Zhu et al., High glucose upregulates myosin light chain kinase to induce microfilament cytoskeleton rearrangement in hippocampal neurons. Mol. Med. Rep. 18, 216–222 (2018).
7
Y. Shi et al., Endothelium-targeted overexpression of heat shock protein 27 ameliorates blood-brain barrier disruption after ischemic brain injury. Proc. Natl. Acad. Sci. U.S.A. 114, E1243–E1252 (2017).
8
J. Fujii, Y. Ikeda, T. Kurahashi, T. Homma, Physiological and pathological views of peroxiredoxin 4. Free Radic. Biol. Med. 83, 373–379 (2015).
9
Z. A. Wood, E. Schroder, J. Robin Harris, L. B. Poole, Structure, mechanism and regulation of peroxiredoxins. Trends Biochem. Sci. 28, 32–40 (2003).
10
Y. Ikeda et al., Different consequences of reactions with hydrogen peroxide and t-butyl hydroperoxide in the hyperoxidative inactivation of rat peroxiredoxin-4. J. Biochem. 149, 443–453 (2011).
11
X. Jiang et al., Blood-brain barrier dysfunction and recovery after ischemic stroke. Prog. Neurobiol. 163–164, 144–171 (2018).
12
S. Nag, R. Venugopalan, D. J. Stewart, Increased caveolin-1 expression precedes decreased expression of occludin and claudin-5 during blood-brain barrier breakdown. Acta Neuropathol. 114, 459–469 (2007).
13
D. Knowland et al., Stepwise recruitment of transcellular and paracellular pathways underlies blood-brain barrier breakdown in stroke. Neuron 82, 603–617 (2014).
14
C. Allen, K. Srivastava, U. Bayraktutan, Small GTPase RhoA and its effector rho kinase mediate oxygen glucose deprivation-evoked in vitro cerebral barrier dysfunction. Stroke 41, 2056–2063 (2010).
15
S. G. Rhee, H. A. Woo, I. S. Kil, S. H. Bae, Peroxiredoxin functions as a peroxidase and a regulator and sensor of local peroxides. J Biol Chem 287, 4403–4410 (2012).
16
E. Vandenbroucke, D. Mehta, R. Minshall, A. B. Malik, Regulation of endothelial junctional permeability. Ann. N.Y. Acad. Sci. 1123, 134–145 (2008).
17
K. Stark et al., Capillary and arteriolar pericytes attract innate leukocytes exiting through venules and “instruct” them with pattern-recognition and motility programs. Nat. Immunol. 14, 41–51 (2013).
18
M. Schnoor, C. A. Parkos, Disassembly of endothelial and epithelial junctions during leukocyte transmigration. Front. Biosci. 13, 6638–6652 (2008).
19
G. C. Jickling et al., Targeting neutrophils in ischemic stroke: Translational insights from experimental studies. J. Cereb. Blood Flow. Metab. 35, 888–901 (2015).
20
J. M. Gidday et al., Leukocyte-derived matrix metalloproteinase-9 mediates blood-brain barrier breakdown and is proinflammatory after transient focal cerebral ischemia. Am. J. Physiol. Heart Circ. Physiol. 289, H558–H568 (2005).
21
Y. Liu et al., Mesenchymal stem cells attenuated blood-brain barrier disruption via downregulation of aquaporin-4 expression in EAE mice. Mol. Neurobiol. 57, 3891–3901 (2020).
22
Y. Huang et al., Targeted homing of CCR2-overexpressing mesenchymal stromal cells to ischemic brain enhances post-stroke recovery partially through PRDX4-mediated blood-brain barrier preservation. Theranostics 8, 5929–5944 (2018).
23
K. Burridge, E. S. Wittchen, The tension mounts: Stress fibers as force-generating mechanotransducers. J. Cell Biol. 200, 9–19 (2013).
24
A. Abbasi et al., Circulating peroxiredoxin 4 and type 2 diabetes risk: The Prevention of Renal and Vascular Endstage Disease (PREVEND) study. Diabetologia 57, 1842–1849 (2014).
25
C. Cirillo et al., Post-stroke remodeling processes in animal models and humans. J. Cereb. Blood Flow Metab. 40, 3–22 (2020).
26
P. Sun et al., Genetic deletion of endothelial microRNA-15a/16-1 promotes cerebral angiogenesis and neurological recovery in ischemic stroke through Src signaling pathway. J. Cereb. Blood Flow Metab. 41, 2725–2742 (2021).
27
P. K. Chandra et al., Transcriptome analysis reveals sexual disparities in gene expression in rat brain microvessels. J. Cereb. Blood Flow Metab. 41, 2311–2328 (2021).
28
P. T. Ronaldson, T. P. Davis, Regulation of blood-brain barrier integrity by microglia in health and disease: A therapeutic opportunity. J. Cereb. Blood Flow Metab. 40, S6–S24 (2020).
Information & Authors
Information
Published in
Classifications
Copyright
Copyright © 2024 the Author(s). Published by PNAS. This article is distributed under Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND).
Data, Materials, and Software Availability
All study data are included in the article and/or SI Appendix.
Submission history
Received: January 15, 2024
Accepted: January 16, 2024
Published online: March 4, 2024
Published in issue: March 12, 2024
Keywords
Acknowledgments
This project was supported by the US NIH grant NS089534 (to J.C.), the US Department of Veterans Affairs (VA) Merit Review BX002495 (to J.C.), and the University of Pittsburgh School of Medicine. J.C. is the Richard King Mellon Professor of Neurology and a recipient of a VA Senior Research Career Scientist Award. M.V.L.B. is the Sylvia and Robert S. Olnick Professor of Neuroscience. We thank Patricia Strickler for administrative support.
Author contributions
M.V.L.B. and J.C. designed research; N.X., X.J., W.Z., Y.S., Q.Y., T.Y., and S.L. performed research; N.X., X.J., W.Z., X.H., and J.C. analyzed data; and N.X., X.J., W.Z., R.K.L., R.F.K., R.A.S., M.V.L.B., and J.C. wrote the paper.
Competing interests
The authors declare no competing interest.
Notes
This article is a PNAS Direct Submission.
Authors
Metrics & Citations
Metrics
Altmetrics
Citations
Cite this article
Endothelial peroxiredoxin-4 is indispensable for blood–brain barrier integrity and long-term functional recovery after ischemic stroke, Proc. Natl. Acad. Sci. U.S.A.
121 (11) e2400272121,
https://doi.org/10.1073/pnas.2400272121
(2024).
Copied!
Copying failed.
Export the article citation data by selecting a format from the list below and clicking Export.
Cited by
Loading...
View Options
View options
PDF format
Download this article as a PDF file
DOWNLOAD PDFLogin options
Check if you have access through your login credentials or your institution to get full access on this article.
Personal login Institutional LoginRecommend to a librarian
Recommend PNAS to a LibrarianPurchase options
Purchase this article to access the full text.