Ptbp2 Alleviates Neuroinflammation and Blood–brain Barrier Disruption via Modulating Microglial Polarization in Ischemic Stroke

Abstract

Damage following ischemic stroke is worsened by microglial activation and subsequent neuroinflammation. Polypyrimidine tract binding protein 2 (Ptbp2) can influence the chemotaxis and repolarization of cancer-related macrophages; however, its specific role in microglial polarization and the underlying mechanisms are not yet fully understood. This study aimed to elucidate the neuroprotective mechanisms of Ptbp2 and examine its effects on microglial activation, neuroinflammation, and glucose metabolism following cerebral ischemia. Mice model of ischemic stroke was developed using temporary middle cerebral artery occlusion (tMCAO). Adeno-associated viruses were used for overexpression and knockdown in C57 mice, and microglial polarization, blood–brain barrier (BBB) integrity, and glycolytic parameters in the peri-infarct cortex were evaluated. RNA sequencing (RNA-seq) was performed on mouse brain tissues. To investigate the underlying mechanisms, the mouse brain microvascular endothelial cell line bEnd.3 and BV2 microglial cell line were used. The protective effect of Ptbp2 on BBB integrity following stroke was evaluated by targeted overexpression and knockdown. We found that Ptbp2 overexpression reduced microglia-mediated neuroinflammation and BBB damage while inhibiting pathological glycolysis, according to findings from both in vitro and in vivo studies. Additionally, Ptbp2 level was significantly downregulated in patients with stroke compared to controls, and was inversely correlated with the severity of neural impairment. Our study unveils novel immunomodulatory mechanisms in stroke and highlights Ptbp2 and its regulatory network as potential therapeutic targets for stroke.

Graphical Abstract

Following ischemic stroke, Ptbp2 reduces neuroinflammatory symptoms and the breakdown of the blood–brain barrier by regulating microglial polarization both in vivo and in vitro.

Supplementary Information

The online version contains supplementary material available at 10.1007/s12035-026-05704-3.

Introduction

With high rates of impairment and recurrence, ischemic stroke (IS) is one of the main causes of disability in China [1, 2].While intravenous thrombolysis remains a cornerstone treatment, its effectiveness for acute large vessel occlusion strokes is constrained by a narrow time window and bleeding risks. Mechanical thrombectomy has emerged as the standard reperfusion therapy, significantly improving outcomes and reducing in-hospital mortality, either as a standalone treatment or in combination with thrombolytics [3–5]. Thus, there is an urgent need for innovative and safe treatment approaches for ischemic stroke.

During ischemic stroke, the inflammatory response greatly affects the severity and progression of the disease [6, 7]. Microglia are the resident immune cells of the central nervous system and are essential for the onset and progression of ischemic stroke. Microglia are quickly activated and polarized into two different phenotypes within hours of brain ischemia. Among them, the M1-polarized microglia release large amounts of pro-inflammatory mediators, such as tumor necrosis factor-alpha (TNF-α) and interleukin-1β (IL-1β), which disrupts the structural integrity of the blood–brain barrier (BBB) and dramatically increases its permeability, thereby accelerating disease progression [8–10].In contrast, the M2-polarized microglia secrete anti-inflammatory cytokines such as interleukin-10 (IL-10), which exerts neuroprotective effects. Thus, a key treatment approach for regulating post-stroke neuroinflammation is to prevent excessive microglial activation and encourage their polarization toward the M2 phenotype. Recent studies have shown that metabolic reprogramming occurs in conjunction with microglial polarization, where the pro-inflammatory M1 phenotype is facilitated by increased pathological glycolysis [11, 12]. However, the core molecular targets linking this metabolic switch to the inflammatory response remain unclear.

Ptbp2 is a crucial RNA-binding protein that plays a role in controlling axonal growth, neuronal metamorphosis during neural development, and the etiology of several neurological conditions [13, 14]. Lately, researchers have been paying more attention to this protein. Studies have shown that Ptbp2 can restore axonal development in motor neurons carrying survival motor neuron (SMN) gene mutations [15, 16]..Moreover, this molecule can induce macrophage chemotaxis in mediastinal neuroblastoma and maintain their pro-inflammatory phenotype [17], suggesting a potential role in the regulation of neuroinflammation. However, its role in modulating the neuroimmune microenvironment after cerebral ischemia, particularly in the microglial polarization, remains unclear.

Using transcriptome analysis, we demonstrated that Ptbp2 levels are significantly reduced in microglia after stroke, and this downregulation was significantly correlated with the NF-κB pathway activation. To investigate whether Ptbp2 affects microglial polarization by controlling signaling pathways like NF-κB, Ptbp2 overexpression systems were created in both in vitro and in vivo models. To clarify the regulatory role of Ptbp2 in subsequent neuroinflammation and uncover the mechanistic insights and possible therapeutic targets, the effects of Ptbp2 on NF-κB pathway activity and microglial phenotypic switching were assessed.

Methods

Patients

This study was approved by the Ethics Committee of the Second Hospital of Hebei Medical University (Ethics Approval No: 2024-R558). Thirty patients with stroke hospitalized in the Department of Neurology at the Second Hospital of Hebei Medical University between September 2023 and October 2024 were enrolled. Blood samples from all patients were collected within 48 h of symptom onset to capture the acute phase inflammatory response. Additionally, 25 age and sex-matched healthy controls were recruited from the same organization. At 3-month follow-up, functional outcomes were evaluated using the modified Rankin Scale (mRS); good outcome was defined as an mRS score between 0 and 2.

Blood Sample Processing

Non-coagulated whole blood samples were drawn from the forearm of each participant. One aliquot of the sample was centrifuged at 2000 × g for 20 min to isolate the serum, and TNF-α levels were quantified using a specific ELISA kit (BOSTER, China, Cat. EK0525). The second aliquot was carefully layered on the top of lymphocyte separation medium (Ficoll) after diluting with phosphate-buffered saline (PBS). Peripheral blood mononuclear cell (PBMC) layers were collected by centrifugation at 400 × g for 30 min. Total RNA was extracted from PBMCs using the TRIzol LS reagent (Invitrogen, Cat# 10296028CN), followed by quantitative analysis of Ptbp2 expression.

Animals

Eighty 12-week-old specific pathogen-free (SPF) male C57BL/6N mice (weighing 22–26 g) were obtained from the Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China). The animals were housed under SPF conditions in a controlled environment (temperature: 22 ± 3 °C; humidity: 60 ± 5%; 12/12 h light/dark cycle) with ad libitum access to a standard chow and water. Following one-week acclimatization, the mice were randomly allocated to either the control or experimental group. All procedures performed in this study were approved by the Animal Care and Use Committee of the Second Hospital of Hebei Medical University and adhered to the guidelines outlined in the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Transient Middle Cerebral Artery Occlusion (tMCAO) Model

The tMCAO model was developed according to a previously established protocol [18]. Briefly, mice were anesthetized while maintaining spontaneous respiration. The right common carotid artery (CCA) and external carotid artery (ECA) were carefully exposed through a midline neck incision. A silicon-coated monofilament (diameter: 0.22 ± 0.01 mm; Yushun Bio, China) was then introduced through the ECA lumen and advanced into the internal carotid artery (ICA) until it reached the origin of the middle cerebral artery (MCA), achieving occlusion. After occlusion for 60 min, the filament was withdrawn to initiate reperfusion. Throughout the procedure, body temperature was maintained at 37 ± 0.5 °C using a heating pad. The sham-operated mice underwent the same surgical procedure, except for filament insertion.

Stereotactic Injection of AAV Vectors

Stereotaxic delivery of the AAV overexpression vector (HBAAV2/9-m-Ptbp2-3xflag-ZsGreen) was performed by injecting into the right lateral ventricle(Zhou, Tao, Yu, Wu, Hui, Xu et al., 2023). Following cranial exposure, the injection site was targeted 1.55 mm right and 1.05 mm posterior to the bregma. A microinjector (Hamilton, Shanghai, China) was implanted at a depth of 3.5 mm, and the viral suspension was infused at a controlled rate over 10 min.

Measurement of Cerebral Infarct Volume

Following extraction, the collected brains were frozen at −20 °C to measure the volume of the infarct. The brain slices were preserved in 4% paraformaldehyde and stained with TTC (2%, 37 °C, 20 min; Sigma T8877). A stereomicroscope (ZEISS Axio Zoom V16) was used to image the sections (Shi, Zou, Jia, Shi, Yang, Liu et al., 2021) and ImageJ software (NIH) was used to measure the infarct area.

Infarct volume was assessed histologically using Nissl staining. After fixation in 10% formaldehyde for 48 h, brain samples were dehydrated, embedded in paraffin, and cut into 3.5-μm sections. Sections were deparaffinized, rehydrated, and incubated with 1% toluidine blue (Solarbio, G1434) according to the manufacturer’s instructions. Following staining, slides were mounted with a neutral resin and examined under the ZEISS Axio Zoom V16 stereomicroscope.

Behavioral Analysis

Neurological functions in mice after tMCAO were assessed using three behavioral tests: the modified Neurological Severity Score (mNSS) [19], rotarod test [20, 21], and the adhesive removal test [22]. Mice were trained for three consecutive days prior to tMCAO surgery and were re-evaluated on days 1, 2, and 3 after tMCAO. All tests were conducted by investigators blinded to the experimental groups.

Quantitative Real-Time PCR

Reverse transcription was performed using the SureScript™ First-Strand cDNA Synthesis Kit (GeneCopoeia). The resulting cDNA was amplified using the real-time PCR system (Agilent, Santa Clara, CA, USA) with BlazeTaq SYBR Green qPCR Mix (GeneCopoeia) as the fluorescent dye. The following primer sequences were used:

Ptbp2:

• Forward: 5’-TCAGGCAGTGTTCTCAGCAG-3’

• Reverse: 5’-GAGGGAGCCCCATCCATTTT-3’

CD86:

• Forward: 5’-TGTTTCCGTGGAGACGCAAG-3’

• Reverse: 5’-TTGAGCCTTTGTAAATGGGCA-3’

Arg-1:

• Forward: 5’-GCTTGCGAGACGTAGACCCT-3’

• Reverse: 5’-CCATCACCTTGCCAATCCC-3’

Immunofluorescence Staining

After anesthetizing the mice, transcardial perfusion was performed sequentially using 0.9% saline and 4% paraformaldehyde. The brains were post-fixed and dehydrated in 30% sucrose for 48 h and sectioned (15 μm thick) on a cryostat (Thermo Scientific, USA). The sections were then processed for staining by permeabilization with 0.5% Triton X-100 (25 min), blocking with 10% donkey serum (1 h, 37 °C), and finally incubation with primary antibodies overnight at 4 °C. The antibodies used included rabbit anti-CD86 (Boster, Cat. BM4121), rabbit anti-Iba-1 (Wako, Cat. 019–19741), rabbit anti-Arg-1 (Proteintech, Cat. 16,001–1-AP), rabbit anti-Occludin (Invitrogen, Cat. 40–4700), rabbit anti-ZO-1 (Proteintech, Cat. 21773–1-AP), and rat anti-CD31 (BD Biosciences, Cat. 550–274). After washing, the sections were incubated with species-appropriate secondary antibodies (Alexa Fluor 488 or 594, Jackson ImmunoResearch, USA) for 1 h, mounted with DAPI Fluoromount-G (Southern Biotech, Cat. 0100–20), and imaged using a laser scanning confocal microscope (ZEISS LSM880, Germany).

Enzyme-Linked Immunosorbent Assay (ELISA)

At 3 days post-tMCAO, blood was obtained via cardiac puncture and serum was isolated by centrifugation (3000 × g, 15 min, RT). Moreover, supernatants were prepared from homogenates of peri-infarct cortical tissue in buffer, which were subsequently sonicated on ice and centrifuged (1500 rpm, 10 min). TNF-α and IL-10 concentrations were determined using the quick ELISA kits (TNF-α, BOSTER EK0527; IL-10, BOSTER EK0417) following the supplier's instructions.

Western Blot

Proteins were isolated from the brain tissues or cultured cells on ice. Equal amounts of proteins were resolved by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) and subsequently transferred onto PVDF membranes (Millipore). Membranes were then blocked with rapid blocking buffer (Boster, Cat. no. AR0041) and incubated with primary antibodies overnight at 4 °C, followed by incubation with horseradish peroxidase (HRP)-conjugated secondary antibodies (goat anti-rabbit IgG; Abcam, Cat. 423,920) for 1 h. Protein bands were visualized using the imaging system (LI-COR Biosciences). The primary antibodies used were rabbit anti-Ptbp2 (Abcam, Cat. ab154787), rabbit anti-CD86 (Boster, Cat. BM4121), rabbit anti-Arg-1 (Proteintech, Cat. 16,001–1-AP), rabbit anti-Occludin (Invitrogen, Cat. 40–4700), rabbit anti-ZO-1 (Proteintech, Cat. 21,773–1-AP), rabbit anti-TNF-α (Boster, Cat. BA0131), rabbit anti-PFKFB3 (Proteintech, Cat. 13,763–1-AP), rabbit anti-LDHA (Proteintech, Cat. 19,987–1-AP), rabbit anti-HK2 (Proteintech, Cat. no.22029–1-AP), rabbit anti-GLUT1 (Proteintech, Cat. 21,829–1-AP), rabbit anti-NF-κB p65 (Abcam, Cat. ab32356), and rabbit anti-NF-κB p-p65 (Proteintech, Cat. 82,335–1-RR).

Lactate Measurement

Lactate concentration in the brain tissues was determined using the lactate assay kit according to manufacturer’s instructions (Beyotime, Cat. P0393S). Briefly, homogenized tissue samples in the lysis buffer were sonicated on ice and centrifuged at 12,000 rpm for 3 min. The supernatant was incubated with the reaction mixture for 25 min. Finally, the absorbance at 450 nm was measured using a microplate reader, and lactate concentration was determined using a standard curve.

RNA Sequencing (RNA-seq) of Brain Tissue

Transcriptomic analysis was conducted on ischemic brain tissues collected 3 days post-stroke. After RNA extraction and sequencing at Majorbio (Shanghai, China), bioinformatic analyses, including Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses were performed to uncover differential gene expression and pathways linked to microglial polarization. Sequencing depth: the average sequencing depth was 15.08 GB of reads per sample. Normalization: gene read counts were normalized using the DESeq2 package. Differentially expressed genes were identified using the following criteria: |Fold Change|> 2 and adjusted P-value < 0.05.The lists of differentially expressed mRNAs and lncRNAs are provided in Supplementary Files 1, 2, and 3, respectively.

ceRNA Network Construction and Analysis

The ceRNA network was constructed based on RNA-seq data from the ischemic penumbra. We first identified putative ceRNA pairs by selecting RNA molecules that showed a significant positive correlation in expression (Pearson's |r|> 0.9, p < 0.01) and were supported by experimental evidence in the NPInter v5.0 database. CeRNA pairs were further required to share ≥ 2 common miRNAs.Subsequently, miRNA-target interactions for these pairs were identified by querying three experimentally validated databases (miRecords, miRTarBase, and TarBase) using the 'multiMiR' R package to confirm shared miRNA response elements (MREs). The final network was assembled by integrating these co-expression and miRNA interaction data. Hub genes within the network were objectively defined and ranked by their degree centrality, a measure of connection density.The complete list of the detailed LncRNA-mRNA and the detailed miRNA-mRNA interactions have been provided in Supplementary File 4 and Supplementary File 5, respectively.

Cell Culture

BV2 microglial cells (immortalized; RRID: CVCL_0182) were cultured in complete DMEM (supplemented with 10% FBS and 1% penicillin–streptomycin) at 37 °C with 5% CO₂ and passaged every two days. To induce activation, cells were exposed to LPS (2 μg/mL) in fresh DMEM for 24 h.

Establishment of the Oxygen–Glucose Deprivation (OGD) Model

bEnd.3 cells were plated in 384-well plates (Cellvis, Cat. p96-1.59) and maintained in high-glucose complete medium for 48 h. Prior to oxygen–glucose deprivation (OGD), the cells were rinsed with PBS and switched to glucose-free medium. OGD injury was induced by placing the cells in a tri-gas incubator (95% N₂, 5% CO₂) at 37 °C for 4 h. Following OGD, the cells were exposed to conditioned medium collected from differentially treated BV2 cells for an additional 24 h.

Lentiviral Transfection of Ptbp2

BV2 cells were plated in 6-well plates and maintained in DMEM containing 10% FBS and 1% penicillin–streptomycin. After 10 h of culturing, a solution containing the virus and polybrene was added to the medium. To establish stable cell lines, transduced cells were subjected to selection using puromycin (Yeasen, Cat.no.60209ES) for a minimum of seven days until only stably transduced cells survived.

Statistical Analysis

All data are presented as the mean ± standard deviation (SD). The independent samples t-test was used for comparing two groups and one-way analysis of variance (ANOVA), followed by Tukey’s test was used for multi-group comparisons after assessing data distribution and variance equality. The association of blood Ptbp2 levels with TNF-α levels, NIHSS, and mRS was evaluated using Pearson's correlation test. A p-value < 0.05 was considered statistically significant.

Results

Ptbp2 is Downregulated While NF-κB Pathway is Activated Following Ischemic Stroke

To identify important regulators involved in microglia-mediated neuroinflammation following stroke, we isolated ischemic ipsilateral tissues for RNA-seq analysis from each mice group three days after ischemic stroke. The groups showed different gene expression patterns (Fig. 1a). Ptbp2, a gene not previously identified in stroke situations, was found to be highly downregulated by volcano plot analysis (logFC = −0.2633, p = 0.032), but was not among the most significantly changed genes (Fig. 1b). KEGG pathway analysis revealed significant abundance of the NF-κB signaling pathway genes (Fig. 1c), indicating that this pathway plays an essential role in microglial inflammation following stroke. Further support was provided by our RNA-seq data, which revealed a substantial increase in the mRNA levels of RelA (p65), an essential component of the NF-κB pathway, indicating elevated pathway activity.

Fig. 1.

Ptbp2 is downregulated and the NF-κB pathway is activated following ischemic stroke (a) Heatmap showing differential gene expression. (b) Volcano plot showing differentially expressed genes. (c) KEGG pathway analysis demonstrating NF-κB signaling pathway enrichment. (d) Quantitative analysis of Ptbp2 protein levels relative to GAPDH. (e) Representative western blots showing Ptbp2 expression (n = 3). (f) Relative Ptbp2 mRNA levels in mouse brain tissues (n = 3). Comparison between the two groups, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

Ptbp2 protein and mRNA levels in the tMCAO group were considerably lower than in the sham group (Fig. 1d–f), confirming Ptbp2 downregulating following ischemic stroke. Together, these results suggest that Ptbp2 participates in post-stroke neuroinflammatory processes by regulating the NF-κB pathway.

Ptbp2 Enhances Neurological Performance and Lowers Infarction Volume in Mice Following Stroke

Figure 2a displays the schematic of the experimental method. Mice in the tMCAO + oe-P2 group showed significantly lower cerebral infarct volume compared to the tMCAO group, according to the results of TTC and Nissl staining (Fig. 2b,d). As revealed by Nissl staining,Ptbp2 therapy dramatically reduced the infarct core from 43 to 13%, according to this investigation. Crucially, this decrease is ascribed to the ischemia penumbra's successful rescue, which prevented its development into the core by reducing it from 20 to 6% (Fig. 2e-g). The mNSS, rotarod test, and adhesive removal test were used to evaluate neurological functions on days 1, 2, and 3 following stroke. On day 1 after stroke, there was no discernible difference in the mNSS scores between tMCAO and tMCAO + oe-P2 groups; the tMCAO + oe-P2 group scores decreased dramatically by day 3 (Fig. 2h). Baseline sensorimotor functions were evaluated before rotarod and adhesive removal tests to remove ineligible mice. In contrast to the tMCAO group, tMCAO + oe-P2 group showed a noticeably higher latency to fall on days 2 and 3 of the rotarod test (Fig. 2i). The time to contact and time to remove the adhesive on days 2 and 3 after stroke was significantly shorter in the tMCAO + oe-P2 group than in the tMCAO group, according to the adhesive removal test (Fig. 2j–k). These findings demonstrate that upregulation of Ptbp2 significantly enhances neurological performance and reduces the number of infarcts in tMCAO model mice.

Fig. 2.

Ptbp2 enhances neurological performance and lowers infarction volume in mice following stroke (a) Schematic of experimental design. (b-g) TTC and Nissl staining of the brain sections from the SHAM, tMCAO, and tMCAO + oe-P2 groups on day three following stroke (n = 4). (h–k) Three days following tMCAO, behavioral tests were conducted on ten animals: (h) mNSS, (i) rotarod test, and (j-k) adhesive removal test. Comparison between two groups, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

Ptbp2 Overexpression Attenuates M1 Microglial Polarization In Vivo

We first validated the efficacy of our Ptbp2 genetically modified models. Compared to the tMCAO control, Ptbp2 mRNA levels were increased 2.4-fold in the overexpression (OE) group and decreased by 44% in the knockdown (KD) group (Fig. 3h, u). Accordingly, Western blot showed a robust 58% upregulation and a 50% downregulation of Ptbp2 protein, respectively (Fig. 3g, t), establishing reliable models for functional studies.

Fig. 3.

Ptbp2 overexpression attenuates M1 microglial polarization in vivo (a) Schematic illustration of ischemic penumbra in filament-induced ischemic stroke model. (b-e) Quantitative analysis and representative immunofluorescence images of CD86⁺/Iba-1⁺ and Arg-1⁺/Iba-1⁺ cells in mice (n = 3). (f) Quantitative analysis of protein levels relative to GAPDH (g, i, j) and representative western blots showing CD86, Arg-1, and Ptbp2 expression (n = 3). (k) Quantitative analysis of TNF-α levels normalized to GAPDH and (l) representative western blots showing TNF-α expression (n = 3). (m) LDHA levels in the brain tissue. (n-o, h, u) Relative CD86, Arg-1, and PtbP2 mRNA levels determined by RT-qPCR (n = 3). (p-q) Representative TTC-stained brain sections from the tMCAO and tMCAO + KD-P2 groups at three days post-surgery (n = 6). (r) The modified Neurological Severity Score (mNSS) was assessed three days after tMCAO induction (n = 10). (s) Quantitative analysis of protein levels relative to GAPDH (t, v, w) and representative western blots depicting CD86, Arg-1, and Ptbp2 bands (n = 3). Scale bar = 50 μm; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

Microglial polarization status was assessed using the microglial markers Iba-1, CD86, and CD206. The tMCAO + oe-Ptbp2 group showed a reduction in CD86⁺/Iba-1⁺ cell ratio and a significant rise in the proportion of Arg-1⁺/Iba-1⁺ cells (Fig. 3d,e) compared to the tMCAO group, suggesting that Ptbp2 overexpression attenuates tMCAO-induced M1 polarization. Western blotting and RT-qPCR analyses supported these findings, demonstrating that the Ptbp2-overexpressing group had considerably higher Arg-1 expression and downregulated CD86 expression (Fig. 3f, g, i, j).

Ptbp2 knockdown significantly enhanced the cerebral infarct volume and worsened neurological deficits after ischemic stroke (Fig. 3p–r). Concomitantly, Ptbp2 knockdown enhanced the pro-inflammatory microglial polarization, as evidenced by a marked increase in CD86 levels without altering Arg-1 expression (Fig. 3s). This finding indicates that the absence of Ptbp2 specifically augments the pro-inflammatory microglial responses.

We assessed LDHA expression in mouse brain tissues to determine the impact of Ptbp2 overexpression on cerebral metabolic status. Following Ptbp2 overexpression, LDHA levels significantly decreased after a transient increase in the tMCAO group (Fig. 3m). Interestingly, alterations in LDHA expression closely correlated with alterations in inflammatory molecules TNF-α and CD86 expression (Fig. 3k). These findings suggest that dysregulated glycolytic reprogramming in the brain tissue is linked to neuroinflammation after ischemic stroke. Importantly, both pathogenic processes were concurrently alleviated by Ptbp2 overexpression. These results support the hypothesis that Ptbp2 is important for controlling immunometabolic reprogramming.

Ptbp2 Overexpression Attenuates M1 Microglial Polarization In Vitro

We introduced empty viral vector controls (oe-NC and kd-NC) to investigate the effect of Ptbp2 on microglial polarization.Validation of our Ptbp2 manipulation models demonstrated a 75% increase and a 40% decrease in protein levels in the OE and KD groups, compared to their respective LPS + NC controls (Fig. 4b, f), thereby establishing reliable models for subsequent functional investigations in an inflammatory setting. LPS treatment induced a robust pro-inflammatory activation, as conventionally defined by the significant upregulation of CD86. Notably, this inflammatory challenge also led to an increase in Arg-1 expression, which we interpret as part of a context-dependent stress-adaptive response rather than a canonical anti-inflammatory program. (Fig. 4a).The involvement of Ptbp2 was demonstrated by a considerable downregulation of CD86 following Ptbp2 overexpression in contrast to results of the LPS group. This effect was not observed in the oe-NC group. Ptbp2 knockdown induced a significant upregulation of CD86, with no corresponding change in Arg-1 expression (Fig. 4e).Using immunofluorescence labeling, the involvement of the NF-κB pathway in this process was examined. Results showed that Ptbp2 overexpression reduced CD86⁺/Iba-1⁺ cell ratio while increasing the Arg-1⁺/Iba-1⁺ cell fraction. Western blotting and RT-qPCR results supported these findings, demonstrating that Phorbol 12-myristate 13-acetate (PMA) treatment reversed the regulatory effect of Ptbp2 on CD86 and Arg-1 expression (Fig. 4g–j). These results demonstrate that Ptbp2 reduces inflammation by modifying the NF-κB signaling pathway.

Fig. 4.

Ptbp2 overexpression attenuates M1 microglial polarization in vitro (a, e) Quantitative evaluation of protein levels relative to GAPDH (b-d, f-h) (n = 3) and representative western blots showing CD86 and Arg-1 expression in BV-2 cells. (i-k) Quantitative analysis and representative immunofluorescence images of Arg-1⁺/Iba-1⁺ and CD86⁺/Iba-1⁺ BV-2 cells (n = 3). (l, m) CD86 and Arg-1 mRNA levels in BV-2 cells. (n-p) Quantitative analysis of protein levels normalized to GAPDH (n = 3) and representative western blots showing CD86 and Arg-1 expression in BV-2 cells. Scale bar = 100 μm; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, compared to the control group; ns, not significant

Ptbp2 reduce Blood–Brain Barrier Disruption in Cerebral Infarction

We used the Evans blue dye to evaluate BBB integrity three days following stroke. Additionally, we measured the levels of the tight junction proteins ZO-1 and Occludin. Evans blue leakage was prevented by Ptbp2 overexpression in tMCAO model mice (Fig. 5a). Immunofluorescence labeling revealed that Ptbp2 overexpression increased the levels of ZO-1 and Occludin in blood vessels surrounding the infarct within cortex region (Fig. 5c, d). ZO-1 and Occludin protein levels were considerably higher in the peri-infarct cortex of the Ptbp2 overexpression group than in the tMCAO group (Fig. 5g). These results demonstrate that Ptbp2 reduces BBB damage caused by cerebral infarction.

Fig. 5.

Ptbp2 prevents blood–brain barrier disruption during cerebral infarction (a-b) Representative images and quantification of Evans blue dye extravasation in the whole brain three days following stroke (n = 4). (c-f) Quantitative analysis of CD31 expression using Occludin and ZO-1, along with representative immunofluorescence images (n = 4). (g) Quantitative analysis of protein levels relative to GAPDH (h, i) and representative western blots showing ZO-1 and Occludin expression in the peri-infarct cortex three days post-stroke (n = 4). Scale bar = 50 μm; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

Ptbp2 Modulates Microglial Polarization in Vitro to Prevent Blood–Brain Barrier Damage

To investigate whether Ptbp2 modifies microglial polarization to maintain BBB integrity, we used an in vitro coculture system. Immunofluorescence staining revealed that the levels of ZO-1 and Occludin were significantly decreased in the endothelial cells following OGD (Fig. 6d). Endothelial damage was synergistically worsened further by LPS stimulation, suggesting that inflammatory mediators from activated microglia exacerbate BBB damage during ischemia.

Fig. 6.

Ptbp2 modulates microglial polarization in vitro to prevent blood–brain barrier damage (a) bEnd.3 cells in culture plates were exposed to OGD for 6 h, reperfused for 4 h using conditioned media from BV2 cells that had received different treatments. (b-d) Quantitative analysis of ZO-1 and Occludin expression and representative immunofluorescence images (n = 3). (e, h) Quantitative analysis of protein levels relative to GAPDH and representative western blots showing ZO-1 and Occluding expression (f-g, i-j) (n = 3). (k) Quantitative analysis of protein levels relative to GAPDH and representative western blots showing TNF-α expression (l) (n = 3). (m) ELISA for quantifying TNF-α levels in BV-2 cells (n = 6). (n-p) Quantitative analysis of ZO-1 and Occludin expression and representative immunofluorescence images (n = 3). Scale bar = 50 μm; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

Interestingly, this damage was significantly reduced and ZO-1 and Occludin expression was restored following exposure to conditioned media from Ptbp2-overexpressing BV2 cells (Fig. 6e). To determine whether this protective effect is dependent on the NF-κB pathway, we conducted a rescue experiment using the NF-κB activator PMA. Results showed that the protective effects of Ptbp2 overexpression were largely eliminated by co-treatment with PMA.

Since TNF-α is an effector molecule of NF-κB signaling, we assessed whether TNF-α alone is sufficient to disrupt BBB functions. ZO-1 and Occludin levels were dramatically lowered when recombinant TNF-α protein was directly added to bEnd.3 cells (Fig. 6h). This detrimental effect was substantially reversed by cotreatment with the TNF-α neutralizing antibody. Immunofluorescence staining demonstrated that the TNF-α neutralizing antibody restored endothelial continuity and the belt-like locations of tight junctions (Fig. 6n). Next, we investigated whether Ptbp2 directly regulates TNF-α in the absence of inflammatory stimuli. Ptbp2 knockdown elevated basal TNF-α levels, whereas its overexpression suppressed TNF-α expression (Fig. 6k–m), establishing TNF-α as a downstream Ptbp2 target. Altogether, these results demonstrated that Ptbp2 inhibits the release of TNF-α, which is mediated by NF-κB, consequently protecting BBB against ischemia disturbances.

Ptbp2 Modifies Microglial Polarization by Suppressing the NF-κB Pathway

We used both in vitro and in vivo assays to assess the expression of inflammatory molecules and the NF-κB signaling pathway proteins to investigate the involvement of Ptbp2 in neuroinflammation. Western blotting revealed significantly higher levels of phospho-p65 (pp65) and TNF-α in the tMCAO group compared to the oe-Ptbp2 group (Fig. 7a, b, e). Additionally, the group with Ptbp2 overexpression showed a notable decrease in CD86 levels and a significant increase in Arg-1 expression compared to the tMCAO group (Fig. 4g). We used an NF-κB pathway activator in the rescue trial. The increase in pp65, CD86, and TNF-α levels indicated that this treatment significantly reversed the anti-inflammatory effects of Ptbp2 overexpression. Further, ELISA performed on mice serum, tissue homogenates, and cell culture supernatants confirmed that Ptbp2 overexpression increased the levels of IL-10 while lowering TNF-α levels. Co-treatment with the NF-κB activator reversed these effects (Fig. 7g–l). Taken together, these findings imply that Ptbp2 overexpression suppresses microglia-mediated neuroinflammation following ischemic stroke, possibly by blocking the NF-κB pathway.

Fig. 7.

Ptbp2 modifies microglial polarization by downregulating the NF-κB pathway (a- e) Representative western blots of p-p65 and p65 and quantitative analysis of protein levels normalized to GAPDH (n = 3). (e-f) Quantitative analysis of TNF-α levels normalized to GAPDH and representative western blots (n = 3). (g-l) ELISA for evaluating TNF-α and IL-10 levels in the mouse blood, brain tissue, and BV-2 cells (n = 6). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

Ptbp2 Regulates Glycolytic Reprogramming by Suppressing the NF-κB Pathway

To determine whether Ptbp2 overexpression influences metabolic state besides neuroinflammation, we assessed the expression of glycolytic markers. Both the tMCAO model group and LPS-stimulated group showed a considerable increase in LDHA, GLUT1, HK2, and PFKFB3 protein levels (Fig. 8a,b), indicating glycolytic reprogramming after stroke. Notably, Ptbp2 overexpression drastically decreased the expression of LDHA, GLUT1, HK2, and PFKFB3. These findings demonstrate that Ptbp2 overexpression inhibits abnormal glycolytic activation while inducing anti-inflammatory response. We used an NF-κB activator to clarify the mechanism underlying this synergistic regulatory effect. Further investigation revealed that the expression of LDHA, GLUT1, HK2, and PFKFB3 was significantly upregulated. Taken together, our findings provide evidence that Ptbp2 is essential for reducing microglial inflammatory polarization and glycolytic metabolism by inhibiting the NF-κB signaling pathway.

Fig. 8.

Ptbp2 regulates glycolytic reprogramming in microglia by suppressing the NF-κB pathway (a-e) Quantitative analysis of protein levels relative to β-actin (n = 3) and representative western blots showing LDHA, HK2, PFKFB3, and GLUT1 levels in mouse brain tissue. (f-j) Quantitative analysis of protein levels normalized to β-actin and representative western blots showing LDHA, HK2, PFKFB3, and GLUT1 expression in BV-2 cells (n = 3); *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

Correlation Between Ptbp2 Levels and TNF-α, NIHSS, and MRS in Patients with Stroke

To demonstrate the clinical relevance of our findings, we evaluated Ptbp2 and TNF-α levels in blood samples from ischemic stroke patients with blood collected within 48 h of symptom onset and controls using RT-qPCR and ELISA. Stroke patients had significantly higher levels of TNF-α and lower Ptbp2 levels compared to controls (Fig. 9a, b). Ptbp2 expression was significantly lower in patients with moderate-to-serious stroke (NIHSS score > 5) than in those with mild stroke, according to the additional subgroup analysis (Fig. 9c). This suggests that Ptbp2 expression is correlated with severe neurological damage. A strong negative correlation was observed between Ptbp2 expression and TNF-α levels, 90-day functional outcome (mRS score), and entrance neurological deficit (NIHSS score) (Fig. 9d–f). These results provide clinical support for preventive and neuroprotective actions of Ptbp2 observed in our initial investigations, showing that lower Ptbp2 levels are strongly associated with increased systemic inflammation, severe brain injury, and worse long-term prognosis.

Fig. 9.

Correlation between Ptbp2 and TNF-a, NIHSS and MRS. (a-b) Ptbp2 and TNF-α levels in human blood samples (30 controls and 30 stroke patients) analyzed using RT-qPCR and ELISA. (c) Blood samples from stroke patients with varying NIHSS scores (0–5 vs. 6–20) were subjected to RT-qPCR for evaluating Ptbp2 levels. (d-f) The association between Ptbp2 and NIHHSS/MRS/TNF-α was demonstrated using the linear regression analysis. (g) Ptbp2 levels in human blood following stroke had moderate diagnostic value for determining outcome during the 3-month follow-up according to ROC

Finally, the predictive value of Ptbp2 in stroke patients with unsatisfactory functional outcomes was assessed using ROC curve analysis. Adverse prognosis was significantly predicted by Ptbp2 (AUC = 0.77, P = 0.01). At a cutoff value of 36.56, Ptbp2 predicted poor outcomes with high specificity (93.75%) and sensitivity (71.43%) (Fig. 9g), indicating its potential as a highly specific prognostic biomarker.

Suppmental Fig. 1 Cellular Source of Ptbp2 Expression in the Post-ischemic Mouse Brain

To definitively identify the cellular source of Ptbp2 in the ischemic brain, we employed immunofluorescence co-staining. The Ptbp2 signal in the ischemic penumbra mainly co-localized with the microglial marker Iba1, but not with the astrocytic marker GFAP or the endothelial marker CD31, as shown in Supplementary Fig. 1. Quantitative analysis revealed that over 90% of Iba1-positive microglia were co-positive for Ptbp2; this percentage was much higher than that of GFAP-positive astrocytes or CD31-positive endothelial cells (Supplementary Fig. 1).Collectively,, these results definitively identify microglia as the primary cellular source of Ptbp2 in the ischemic brain.

Supplementary Fig. 2 Ptbp2 is Identified as a Central Hub in the ceRNA Network of Ischemic Stroke

Systems-level analysis via a ceRNA network model identified Ptbp2 as a major hub based on its high degree centrality (Supplementary Fig. 2). This objective prediction, indicating its potential role as a master regulator, provided the rationale for prioritizing Ptbp2 in our experiments.

Discussion

During the period immediately following ischemic stroke, neuroinflammation is a major mediator of subsequent injury [23, 24]. Therapeutic strategies targeting the early inflammatory response are essential for better outcomes [25, 26]. Our study highlights the potential of the RNA-binding protein Ptbp2 as a novel therapeutic target. This study is the first showing the complex regulatory role of Ptbp2 in microglial polarization, neuroinflammatory responses, and metabolic reprogramming after ischemic stroke.

We observed Ptbp2 downregulation following ischemic stroke and found that reduced Ptbp2 expression facilitates M1 microglial polarization, which worsens neuroinflammation and compromises the BBB integrity. Although Ptbp2 was not the most differentially expressed gene in the volcano plot analysis, it was selected for further study due to its consistent and significant downregulation at mRNA and protein levels, and its possible involvement in key inflammatory pathways. The anti-inflammatory effect of Ptbp2 was abolished by concurrent administration of the NF-κB activator PMA. Collectively, our findings revealed that Ptbp2 exerts anti-inflammatory effects via the NF-κB signaling pathway.

Microglia, the resident immune cells of the central nervous system, can polarize into two main phenotypes within hours of ischemic injury. The M1 type releases pro-inflammatory cytokines TNF-α and IL-1β while the M2 type secretes anti-inflammatory factors IL-10 and IL-4 [27–30]. Our study showed that Ptbp2 overexpression considerably improved the inflammatory response, decreased M1 polarization, increased M2 polarization, and suppressed microglial activation both in vitro and in vivo.

Activated microglia release large amounts of cytokines (TNF-α, IL-6) and chemokines, which can disrupt the BBB [31–33]. To replicate ischemia pathology followed by secondary inflammation, we created an in vitro model in which bEnd.3 endothelial cells cultured in OGD were exposed to conditioned media from BV2 microglial cells. We observed that Ptbp2 overexpression dramatically reduced endothelial cell damage in inflammatory settings. Subsequent mechanistic analyses demonstrated that TNF-α directly undermines the integrity of the BBB, and the administration of a neutralizing antibody specific to TNF-α significantly restored barrier functions, highlighting the crucial and distinct role of TNF-α in this process.

NF-κB, a classical transcription factor regulating inflammation, controls neuroinflammation and other processes, causing cytokine storm and advancing ischemic injury [34–36]. It has been shown that M1 polarization is successfully decreased by blocking the NF-κB signaling pathway [37–39]. Western blot analysis verified that Ptbp2 overexpression reduces pp65 levels in both in vitro and in vivo models. The anti-inflammatory benefits of Ptbp2 were reversed by the NF-κB activator PMA, demonstrating the involvement of this pathway.

Previous studies showed that increased glycolytic flow coincides with M1 microglial activation to fulfill bioenergetic and biosynthetic requirements related to the production of pro-inflammatory mediators [30, 40–42]. Following stroke, the inflammatory response and brain damage are worsened by lactate build-up due to abnormal glycolysis activation [43–45]. Our results demonstrated that glycolytic activity is markedly elevated in mouse brain tissue after ischemic stroke, while Ptbp2 overexpression attenuated both M1 polarization and glycolytic activation. Considering its role in controlling microglial inflammation, we speculated that Ptbp2 acts on the microglia to modulate cerebral glycolytic reprogramming following stroke. Crucially, the anti-inflammatory and glycolytic-inhibitory effects of Ptbp2 were abolished by the pharmacological activation of NF-κB, suggesting that the pathway serves as a key regulatory node that can orchestrate immunometabolic reprogramming. Beyond transcriptional regulation, Ptbp2, as an RNA-binding protein, may additionally fine-tune metabolic reprogramming at the post-transcriptional level by directly controlling the splicing or stability of mRNAs encoding enzymes, such as LDHA and HK2. While our study primarily revealed the protective effects of Ptbp2 via microglial regulation, its established role in neuronal development and axonal guidance prompted us to speculate that the beneficial effects of Ptbp2 on neurological scores and infarct volume may not solely due to improvement in the inflammatory microenvironment that facilitated neuronal growth. A partial contribution from direct neuroprotective or repair-promoting mechanisms is plausible. Future studies should delineate the cell type-specific functions of Ptbp2 in immune cells and neurons to fully understand its therapeutic potential. Nevertheless, our study has several limitations. Although our rescue experiments strongly support the central role of the NF-κB pathway, the possibility that Ptbp2 functions through other unknown or compensatory pathways cannot be ruled out. While our systems biology analysis identified Ptbp2 as a central hub within a ceRNA-coexpression network, the precise mechanistic contributions of the competing miRNA partners to this regulatory circuit remain to be fully elucidated and represent a promising direction for future investigation.

Conclusion

Ptbp2 ameliorates ischemic brain injury by attenuating microglial M1 polarization. Additionally, Ptbp2 suppress the NF-κB signaling pathway, thereby lowering the post-ischemic glycolytic metabolism. Our study identifies Ptbp2 as a possible molecular target for stroke treatment and emphasizes the necessity of comprehensive therapeutic strategies that can concurrently modify the interconnected networks of inflammation, metabolism, and immunity.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

Ptbp2

Polypyrimidine tract-binding protein 2

tMCAO

Transient middle cerebral artery occlusion

LPS

Lipopolysaccharide

BBB

Blood-brain barrier

bEnd.3

Mouse brain microvascular endothelial cell line

Author Contributions

Wen Ting Xu: Writing-Original draft, Methodology, Formal analysis, Data curation.  Lin Lin Li: Writing-Review and editing, Visualization, Methodology, Data curation. Cong Zhang: Visualization, Software. Meng Jia Zhou: Supervision, Formal analysis. XiangJian Zhang: Writing-review and editing, project administration, funding acquisition. All authors have read and approved the final manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (Grant Nos. 81974184 and 82271366), the Hebei Provincial Health Commission (Grant No. 20190061), and the Hebei Provincial Key Laboratory of Vascular Homeostasis (Grant No. 20567630H).

Data Availability

The data used to support the findings of this study are available from the corresponding author upon request.

Declarations

Ethics Approval and Consent to Participate

The study protocol was approved by the Ethics Committee of The Second Hospital of Hebei Medical University.All efforts were made to minimize animal suffering and to reduce the number of animals used in the experiments.

Consent for Publication

All the authors have read and approved the final draft. However, this is not applicable to individual data.

Competing interests

The authors declare no competing interests.

Declaration of Generative AI in Scientific Writing

No artificial intelligence tools were used in this work.