L arginine and radiofrequency electromagnetic field attenuate inflammatory and apoptotic injury after experimental subarachnoid hemorrhage

Abstract

Subarachnoid hemorrhage (SAH) triggers intense neuroinflammatory and apoptotic cascades that contribute to secondary brain injury, delayed cerebral ischemia, and poor neurological outcomes. Inflammatory mediators such as interleukin-6 (IL-6) and signal transducer and activator of transcription 3 (STAT3) amplify oxidative stress and endothelial dysfunction, whereas suppression of the hypoxia-inducible factor-1α (HIF-1α)/endothelial nitric oxide synthase (eNOS)/vascular endothelial growth factor (VEGF) pathway impairs hypoxic adaptation and neuronal survival. L-arginine (L-ARG), a nitric oxide precursor, and radiofrequency electromagnetic field (RF-EMF) therapy each exhibit neuroprotective potential through anti-inflammatory and microcirculatory mechanisms. Adult male Wistar rats were assigned to Control, SAH, SAH + L-ARG, SAH + RF, and SAH + L-ARG + RF groups (n = 8 per group). SAH was induced by autologous arterial blood injection into the cisterna magna, followed by daily treatment with L-ARG (500 mg/kg, p.o.) and/or RF-EMF (27.12 MHz, 30 min/day) for seven days. SAH markedly increased NF-κB, TNF-α, Cas-3, IL-6, and STAT3 expression while suppressing HIF-1α, eNOS, and VEGF transcripts. Combined treatment was associated with the most pronounced improvements across measured molecular markers, including reduced inflammation and apoptosis and normalization of mitochondrial regulators (↓BAX, ↑BCL-2). These findings suggest that combined L-ARG and RF-EMF exposure is associated with attenuation of inflammation and apoptosis-related molecular injury after experimental SAH; however, functional and perfusion-based outcomes are required before therapeutic efficacy can be inferred.

Introduction

Subarachnoid hemorrhage (SAH) is one of the most catastrophic cerebrovascular emergencies, characterized by the abrupt rupture of intracranial vessels and extravasation of blood into the subarachnoid space. The ensuing rise in intracranial pressure, cerebral ischemia, and secondary neurovascular injury contribute to its high mortality and morbidity¹. Beyond the primary bleeding event, delayed brain injury driven by oxidative stress, neuroinflammation, endothelial dysfunction, and apoptosis plays a dominant role in longer-term neurological outcome^2,3. In addition early brain injury after SAH is also shaped by acute vascular events, including microvasospasm, microthrombosis, impaired autoregulation, blood–brain barrier disruption, and early reductions in cerebral perfusion^2,4. These vascular mechanisms interact with inflammatory and oxidative pathways and may further amplify secondary neuronal and glial injury.

A central molecular feature of SAH is the overactivation of the interleukin 6 (IL-6)/signal transducer and activator of transcription (STAT3) axis, which amplifies neuroinflammatory signaling in multiple neurovascular cell populations⁵. Hemoglobin degradation products and damage-associated molecular patterns trigger toll-like receptor (TLR) and cytokine receptor pathways in resident microglia, astrocytes, neurons, cerebrovascular endothelial cells, pericytes, and infiltrating leukocytes, leading to nuclear factor kappa beta (NF-κB) and STAT3 activation⁶. This upregulates IL-6, tumor necrosis factor alpha (TNF-α), and other pro-inflammatory mediators that further promote microglial activation, oxidative stress, and blood-brain barrier (BBB) breakdown⁷. The resulting neuroinflammatory milieu impairs endothelial nitric oxide (NO) signaling, aggravates vascular spasm, and perpetuates cerebral hypoperfusion.

The hypoxia-inducible factor 1-alpha (HIF-1α) pathway serves as a key regulator of oxygen homeostasis under ischemic or oxidative stress. Stabilization of HIF-1α promotes transcription of protective downstream targets such as endothelial nitric oxide synthase (eNOS) and vascular endothelial growth factor (VEGF), facilitating vasodilation, angiogenesis, and neurovascular repair⁸. However, in SAH, excessive inflammation and reactive oxygen species (ROS) destabilize HIF-1α, suppressing eNOS and VEGF expression and thereby hindering endothelial regeneration⁹.

Simultaneously, SAH triggers mitochondrial-dependent apoptosis, a major contributor to delayed neuronal death. The imbalance between anti-apoptotic B-Cell Lymphoma 2 (BCL-2), and pro-apoptotic BCL-2-Associated X protein (BAX) disrupts mitochondrial membrane integrity, enabling cytochrome-c (Cyt-C) release and caspase-3 (Cas-3) activation, which culminates in neuronal apoptosis¹⁰. Thus, therapeutic interventions capable of restoring mitochondrial stability and reactivating the HIF-1α/eNOS/VEGF axis hold promise for preventing secondary brain injury after SAH.

L-arginine (L-ARG), a natural substrate of NOS, enhances nitric oxide (NO) synthesis and exerts vasodilatory, antioxidant, and anti-apoptotic effects via the NO/cyclic guanosine monophosphate (cGMP)/eNOS pathway¹¹. In parallel, low-intensity radiofrequency electromagnetic field (RF-EMF, 27.12 MHz-10 V/m E-field value) exposure has emerged as a biophysical neuromodulation strategy shown to attenuate oxidative stress, improve mitochondrial respiration by modulating calcium-dependent signaling, and enhance eNOS enzyme activation¹².

SAH triggers inflammatory cascades involving IL-6 and STAT3, leading to endothelial dysfunction, suppression of HIF-1α, eNOS, and VEGF signaling, and activation of mitochondrial apoptosis (↓BCL-2, ↑BAX, ↑caspase-3), ultimately contributing to neurovascular injury and neuronal cell death. RF-EMF is proposed to enhance Ca²⁺/calmodulin-dependent protein kinase signaling linked to eNOS activation and NO production, whereas L-arginine is presented as a substrate that may support NO bioavailability rather than as a direct activator of eNOS. The combined treatment is hypothesized to reduce inflammation and apoptosis while improving neurovascular homeostasis. Abbreviations: SAH, subarachnoid hemorrhage; IL-6, interleukin-6; STAT3, signal transducer and activator of transcription 3; HIF-1α, hypoxia-inducible factor 1 alpha; eNOS, endothelial nitric oxide synthase; VEGF, vascular endothelial growth factor; BCL-2, B-cell lymphoma 2; BAX, Bcl-2-associated X protein; RF-EMF, radiofrequency electromagnetic field; NO, nitric oxide.

Given their complementary molecular mechanisms, the concurrent application of L-ARG and RF therapy may provide an additive mechanistic benefit after SAH by modulating vascular, inflammatory, and mitochondrial pathways. The present study aimed to investigate the molecular and histopathological effects of L-ARG and RF-EMF exposure, alone and in combination, on histological and molecular correlates of injury in an experimental rat model of SAH. The proposed mechanistic framework (Fig. 1) summarizes hypothesized pathways by which combined treatment may attenuate inflammation, support endothelial homeostasis, and reduce apoptosis-related markers, providing an integrated experimental approach to SAH-induced neurovascular injury.

Fig. 1.

Schematic illustration of the proposed mechanisms by which L-ARG and RF-EMF therapy mitigate neuronal injury following SAH.

Materials and methods

Animals

A total of forty adult male Wistar albino rats (300–350 g, 10–12 weeks old) were obtained from the Experimental Animal Production and Research Center of Burdur Mehmet Akif Ersoy University (MAKU-HUDAL, Burdur, Türkiye). Animals were acclimatized for one week prior to experimentation and housed in standard polypropylene cages under controlled environmental conditions (temperature 23 ± 2 °C, relative humidity 55 ± 5%, and 12 h light/dark cycle). Standard pellet chow and tap water were provided ad libitum. All experimental procedures were performed between 09:00 and 17:00 h to minimize circadian variation.

Ethical approval

All experimental procedures were approved by the Burdur Mehmet Akif Ersoy University Experimental Animals Local Ethics Committee (MAKU-HADYEK, Burdur, Türkiye; Approval date: 09.07.2025/137/1527). The study was conducted in accordance with the principles of the Guide for the Care and Use of Laboratory Animals¹³ complied with national regulations on the ethical use of laboratory animals and ARRIVE 2.0 guidelines. The investigators strictly adhered to the 3Rs (Replacement, Reduction, and Refinement) principles throughout the experimental process.

Experimental protocol

The rats were randomly divided into five groups (n = 8 each):

1. Control group: 0.3 mL sterile saline was injected into the cisterna magna under 90 mg/kg ketamine (90 mg/kg, Keta-Control, Doğa İlaç, Türkiye) + 8–10 mg/kg xylazine (10 mg/kg, Xylazinbio, Bioveta, Czech Republic) anesthesia. Animals received saline (0.5 mL/day) by oral gavage (po) for 7 days.

2. SAH group: SAH was induced by autologous arterial blood injection (0.3 mL) into the cisterna magna. Animals were given po saline for 7 days¹⁴.

3. SAH + RF group: Following SAH induction, rats were exposed to a 27.12 MHz RF-EMF for 30 min/day for 7 consecutive days using a device positioned above the cages¹⁵.

4. SAH + L-ARG group: After SAH induction, rats received po 500 mg/kg/day L-ARG (Exvitamin, Aydın, Türkiye) for 7 days¹⁶.

5. SAH + RF + L-ARG group: Following SAH induction, rats were treated with both modalities 30 min/day RF-EMF exposure (27.12 MHz) and po 500 mg/kg/day L-ARG for 7 days.

Animals were randomly allocated to experimental groups before SAH induction. Because RF-EMF exposure and oral gavage procedures were technically apparent to the investigators administering the interventions, treatment administration was not fully blinded. However, outcome assessment was performed using coded samples. Histopathological, immunohistochemical, and molecular analyses were conducted by different investigators who were blinded to group allocation until completion of scoring, image analysis, and qRT-PCR data processing.

During SAH induction, one animal died due to a suspected brainstem injury. To maintain the planned sample size, this animal was replaced, resulting in a procedural mortality rate of 1/32 (3.13%) among SAH-induced animals (excluding the control group). No additional mortality or exclusions occurred during the postoperative follow-up period.

Postoperative care included recovery on a warming pad, daily monitoring of activity, grooming, hydration, and feeding, and analgesia according to institutional protocol. At the end of the treatment period, all animals were anesthetized with 90 mg/kg ketamine (90 mg/kg, Keta-Control, Doğa İlaç, Türkiye) and 8–10 mg/kg xylazine (10 mg/kg, Xylazinbio, Bioveta, Czech Republic). Brain tissues were carefully removed and divided following euthanasia via surgical exsanguination. During necropsy, care was taken to avoid mechanical damage to the cranial cavity or neural tissues. The skull was delicately opened using fine microsurgical instruments, and the whole brain was gently excised and examined macroscopically for alterations in surface morphology, color, or evidence of edema, hemorrhage, or hyperemia.

RF-EMF exposure set-up

RF-EMF exposure was applied at 27.12 MHz. A custom-built generator capable of producing a stable and homogeneous 27.12 MHz frequency at 0.8 W output was used to generate the RF-EMF. The exposure protocol consisted of a single daily 30-minute session. The electromagnetic field was configured to cover the Euro type-4 cage uniformly, positioned approximately 30 mm above the cage floor, and validated using CST Studio Suite 3D simulations (Oxfordshire, England). The RF applicator generated an electric field intensity of 10 V/m, delivering 0.8 W of RF power through a specially designed PCB antenna. Field intensity measurements were performed using an ELF meter (England), and spectrum analyses were conducted at each experimental stage using ROHDE-SCHWARZ FSH6 and PROMAX AE-566 spectrum analyzers (USA). All procedures took place in an electromagnetically shielded room located in the Medical Faculty Animal Research Laboratory, with a shielding effectiveness of 80 dB at 27.12 MHz. Specific absorption rate and intracranial temperature were not directly measured in this study; therefore, the contribution of thermal effects cannot be excluded. Accordingly, any discussion of temperature-independent mechanisms is presented as a hypothesis that requires dedicated dosimetry and in situ temperature monitoring in future work.

Histopathological method

Following gross evaluation, the brains were fixed in 10% neutral buffered formalin to preserve tissue morphology. After fixation, samples were routinely processed, dehydrated in graded ethanol, cleared in xylene, and embedded in paraffin. Serial sections of 5 μm thickness were prepared with an automated microtome and stained with hematoxylin and eosin for histopathological examination under light microscopy.

Histopathological findings were evaluated and semiquantitatively scored based on the criteria adapted from Mielke et al. (ESAS Score)¹⁷. Scores ranged from 0 (normal histology) to 4 (severe damage), enabling standardized comparison of tissue alterations among experimental groups (Table 1).

Table 1.

Histopathological and immunohistochemical scores of subarachnoid hemorrhages¹⁷.

	Histopathology	Immunohistochemistry
0	Normal meningeal and parenchymal structure	No staining
1	Blood in the subarachnoid space, ventricles, or brain parenchyma.	Focal/weak
2	Localized or diffuse thin subarachnoid hemorrhage, no intraventricular, or intraparenchymal hemorrhage.	Diffuse/weak
3	Diffuse or localized thick subarachnoid blood layers, no intraventricular, or intraparenchymal hemorrhage.	Diffuse/strong staining.
4	Intraventricular or intraparenchymal hemorrhage in association with subarachnoid hemorrhage, regardless of thickness or location.

Immunohistochemical method

For immunohistochemical evaluation, paraffin-embedded brain sections were mounted on poly-L-lysine–coated slides to ensure optimal tissue adherence and stained using the streptavidin–biotin peroxidase method. After deparaffinization in xylene and rehydration through graded ethanol, antigen retrieval was achieved in citrate buffer (pH 6.0) by microwave heating. Endogenous peroxidase activity was quenched with 3% hydrogen peroxide for 10 min, followed by blocking of nonspecific binding using a protein-blocking solution.

Sections were incubated overnight at 4 °C with primary antibodies against Cas-3 (EPR16888, ab179517), NF-κB p65 (ab16502), and TNF-α (RM1005, ab307164) (all from Abcam, Cambridge, UK), diluted 1:100 according to manufacturer instructions. Detection was carried out using the Mouse and Rabbit Specific HRP/DAB Micropolymer Detection Kit (ab236466, Abcam), with diaminobenzidine as the chromogen and hematoxylin for counterstaining. Negative controls were processed identically but without the primary antibody. Immunoreactivity was semi-quantitatively scored on a four-point scale (Table 1).

Images were captured using an Olympus CX41 microscope equipped with a ×40 objective lens under identical illumination settings. Regions of interest (ROI) were selected from anatomically matched cortical areas adjacent to the subarachnoid space, and consistent imaging fields were maintained across samples by referencing standardized anatomical landmarks. Images were acquired using Cell Sens software, and staining intensity and area were analyzed quantitatively with ImageJ (version 1.48, NIH, Bethesda, MD, USA). All evaluations were performed in a blinded manner by an experienced pathologist to ensure objectivity and reproducibility.

Reverse transcription-polymerase chain reaction (RT-qPCR)

Using the manufacturer’s protocol, RNA was isolated from homogenized brain tissues with the GeneAll RiboEx (TM) RNA Isolation Kit (GeneAll Biotechnology, Seoul, Korea). The amount and purity of the RNAs obtained were measured with the BioSpec-nano nanodrop (Shimadzu Ltd. Kyoto, Japan) device. 1 µg RNA was used for cDNA synthesis. cDNA synthesis, A.B.T. ™ cDNA Synthesis Kit (Atlas Biotechnology, Türkiye) was carried out in a thermal cycler according to the protocol. Primer designs were made by detecting specific mRNA sequences and testing possible primer sequences using the NCBI website. Primer specificity was confirmed by single-peak melting curve analysis for all target genes. The sequences of the primer sequences used are shown in Table 2. Amplicon lengths were 209 bp, slightly above the typical qRT-PCR range (70–200 bp), to ensure accurate coverage of the target region.

Table 2.

Primary sequences, product size and accession numbers of genes.

Genes	Primer sequence	Product size	Accession number
Rn18S (HouseKeeping)	F: CTCTAGATAACCTCGGGCCG	209 bp	NR_046237.2
	R: GTCGGGAGTGGGTAATTTGC
STAT3	F: ATCGCTGAGGTACAATCCCG	412 bp	NM_012747.2
	R: TCTTCGCAGGTTGTGCTGAT
BAX	F. CACGTCTGCGGGGAGTCAC	419 bp	NM_017059.2
	R: TAGAAAAGGGCAACCACCCG
BCL-2	F: CATCTCATGCCAAGGGGGAA	284 bp	NM_016993.2
	R: TATCCCACTCGTAGCCCCTC
eNOS	F: GGTTGACCAAGGCAAACCAC	247 bp	NM_021838.2
	R: CCTAATACCACAGCCGGAGG
HIF-1α	F: GCAACTAGGAACCCGAACCA	251 bp	NM_024359.2
	R: TCGACGTTCGGAACTCATCC
IL-6	F: CACAAGTCCGGAGAGGAGAC	168 bp	NM_012589.2
	R: ACAGTGCATCATCGCTGTTC
VEGF	F: TTCGTCCAACTTCTGGGCTC	482 bp	NM_001287111.1
	R: GCTTTCTGCTCCCCTTCTGT

F, Forward; R, Reverse; Rn18S, 18S Ribosomal RNA; STAT3, signal transducer and activator of transcription 3; BCL-2, B-Cell Lymphoma 2; BAX, BCL-2-Associated X protein; eNOS, endothelial nitric oxide synthase; HIF1α, Hypoxia-Inducible Factor 1-alpha; IL-6, Interleukin-6; VEGF, Vascular Endothelial Growth Factor.

Expression levels of genes were measured in a Biorad CFX96 (California, USA) real-time PCR instrument using 2 X SYBR green master mix (Nepenthe, Türkiye). In the study, the Rn18s was used as the housekeeping gene or normalization of gene expression levels. The reaction mixture was prepared according to the manufacturer’s protocol to a final volume of 20 µl. The resulting reaction mixture was placed in a real-time qPCR device determined according to the kit manufacturer’s protocol, and each sample was studied in 3 replications. PCR conditions, initial denaturation 94 °C 10 min. 1 cycle, denaturation 95 °C 15 s. and annealing/extension 58 °C 30 s. was applied as 40 cycles.PCR efficiencies for all primer sets were determined using standard curves and were within the acceptable range (90–105%). Relative mRNA levels were calculated by applying the 2^{− ΔΔCt} formula to the normalized results.

Statistical analysis

The Shapiro–Wilk test was first used to assess the normality of all quantitative data. For variables with a normal (parametric) distribution (p > 0.05), one-way analysis of variance (ANOVA) was applied to evaluate differences among groups, followed by Tukey’s post hoc test for pairwise comparisons when significant results were obtained. When data were not normally distributed, non-parametric comparisons were performed using the Kruskal–Wallis test followed by Dunn’s post hoc test. All statistical procedures were performed using GraphPad Prism (version 10.1; GraphPad Software, San Diego, CA, USA). Results were expressed as mean ± standard deviation, and a p-value < 0.05 was considered statistically significant.

Declaration of generative AI and AI-Assisted technologies in the writing process

During the preparation of this work, the authors used ChatGPT-5 to improve language and readability. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.

Results

Histopathological findings

Histopathological evaluation revealed marked morphological alterations in the SAH group compared with the control group. Extensive meningeal and SAH, cortical disorganization, hyperemia, and neuronal degeneration were evident, accompanied by widespread erythrocyte extravasation and tissue edema (Fig. 2B). In contrast, the control group displayed preserved cortical cytoarchitecture and minimal vascular dilation (Fig. 2A).

Fig. 2.

Histopathological alterations in the cerebral cortex following SAH and the effects of L-ARG and RF-EMF treatments. Representative H&E stained sections: (A) Control group displaying normal cortical cytoarchitecture and intact vascular integrity (arrow). (B) SAH group showing widespread neuronal necrosis, perivascular edema, and hemorrhage (arrow). (C) SAH + L-ARG group with partial preservation of neuronal morphology and reduced edema (arrow). (D) SAH + RF group demonstrating moderate histological recovery and fewer necrotic cells (arrow). (E) SAH + L-ARG + RF group exhibiting nearly normal cortical architecture and minimal inflammatory infiltration (arrow). Scale bar = 50 μm. The accompanying bar graph presents mean ± SD histopathological scores analyzed via one-way ANOVA followed by Tukey’s post hoc test (*p < 0.05; **p < 0.01; ***p < 0.001). The scatter plot shows individual data points for each animal (n = 8 per group). Abbreviations: SAH, subarachnoid hemorrhage; L-ARG, L-arginine; RF-EMF, radiofrequency electromagnetic field.

Treatment with L-ARG or RF-EMF notably attenuated these lesions, as evidenced by reduced hemorrhagic density and improved neuronal morphology (Fig. 2C–D). Among treated groups, RF therapy produced a greater reduction in cortical congestion and perivascular edema than L-ARG alone. The combined L-ARG + RF group demonstrated nearly normal cortical and meningeal organization with minimal vascular leakage (Fig. 2E). Semi-quantitative scoring confirmed a marked reduction in histopathological severity in SAH + RF and SAH + L-ARG + RF groups relative to SAH (p < 0.01 and p < 0.001 respectively), with the lowest scores observed in the combined therapy group (Fig. 2).

NF-κB immunohistochemical findings

Immunohistochemical analysis revealed a marked increase in NF-κB immunoreactivity within the cortical and subarachnoid regions of rats subjected to SAH compared with the control group. In the SAH group (Fig. 3B), intense nuclear and cytoplasmic brown immunopositivity was observed particularly in neurons and glial cells adjacent to the hemorrhagic area, indicating robust activation of inflammatory transcriptional signaling. This was accompanied by prominent vascular congestion and perivascular inflammatory cell infiltration. In contrast, control brains exhibited only minimal, focal NF-κB expression (Fig. 3A), consistent with basal physiological activity.

Fig. 3.

Immunohistochemical detection of NF-κB in cerebral cortex tissues following SAH and the modulatory effects of L-ARG and RF-EMF treatment. Representative micrographs show: (A) Control group with negligible NF-κB immunostaining and preserved cortical structure. (B) SAH group exhibiting intense NF-κB immunopositivity (arrow) in neuronal and glial cells around hemorrhagic and edematous areas. (C) SAH + L-ARG group displaying reduced NF-κB staining and partial tissue recovery. (D) SAH + RF group showing moderate suppression of NF-κB immunoreactivity and improved cellular integrity. (E) SAH + L-ARG + RF group with minimal NF-κB expression and near-normal cortical morphology. Scale bar = 50 μm. Quantitative analysis of immunoexpression levels (mean ± SD) was performed by one-way ANOVA followed by Tukey’s post hoc test (*p < 0.05; ***p < 0.001). Combined treatment produced the strongest down-regulation of NF-κB, confirming the synergistic anti-inflammatory and neuroprotective action of L-ARG and RF-EMF therapy. Abbreviations: SAH, subarachnoid hemorrhage; L-ARG, L-arginine; RF-EMF, radiofrequency electromagnetic field; NF-κB, nuclear factor kappa B.

Treatment with L-ARG markedly attenuated NF-κB immunostaining, showing reduced nuclear translocation and weaker cytoplasmic positivity (Fig. 3C). Similarly, RF-EMF exposure led to a pronounced decline in NF-κB expression intensity, with a more uniform cortical architecture and diminished glial reactivity (Fig. 3D). Notably, the combined L-ARG + RF therapy produced the most striking reduction in NF-κB activity, where the immunolabeling pattern was comparable to that of the control group and inflammatory cell infiltration was nearly absent (Fig. 3E).

Semi-quantitative immunoscoring confirmed these observations: NF-κB expression was significantly elevated in the SAH group compared to controls (p < 0.001), whereas all treatment groups exhibited marked reductions (p < 0.05 SAH + L-ARG vs. SAH; p < 0.001 SAH + RF vs. SAH; p < 0.001 SAH + L-ARG + RF vs. SAH). The SAH + L-ARG + RF group showed the lowest NF-κB scores among treatments, demonstrating a synergistic anti-inflammatory effect (Fig. 3).

TNF-α immunohistochemical findings

In the control group, TNF-α immunoreactivity was weak and limited to a few glial and endothelial cells, consistent with basal physiological cytokine expression (Fig. 4A). In contrast, the SAH group exhibited strong and diffuse TNF-α positivity, particularly around the hemorrhagic loci, neuronal somata, and perivascular regions (Fig. 4B). Intense cytoplasmic brown staining in neurons and microglial cells indicated widespread inflammatory activation and cytokine overproduction. These findings align with the typical pro-inflammatory response following subarachnoid hemorrhage, which contributes to secondary neuronal injury and BBB disruption.

Fig. 4.

Immunohistochemical expression of TNF-α in the cerebral cortex following SAH and treatment with L-ARG and RF-EMF. Representative micrographs: (A) Control group showing negligible TNF-α immunoreactivity and preserved cortical architecture. (B) SAH group exhibiting intense TNF-α positivity (arrow) in neurons and perivascular inflammatory cells surrounding the hemorrhagic site. (C) SAH + L-ARG group displaying reduced TNF-α staining and partial morphological recovery. (D) SAH + RF group showing moderate attenuation of TNF-α expression and fewer inflammatory foci. (E) SAH + L-ARG + RF group demonstrating minimal TNF-α immunopositivity and nearly normal cortical organization. Scale bar = 50 μm. Bar graphs represent mean ± SD immunoexpression scores analyzed by one-way ANOVA followed by Tukey’s post-hoc test (*p < 0.05; **p < 0.01; ***p < 0.001). Scatter plots show individual values (n = 8 per group). Combined therapy produced the greatest inhibition of TNF-α, highlighting a synergistic anti-inflammatory and neuroprotective effect. Abbreviations: SAH, subarachnoid hemorrhage; L-ARG, L-arginine; RF-EMF, radiofrequency electromagnetic field; TNF-α, tumor necrosis factor alpha.

Administration of L-ARG markedly reduced TNF-α immunolabeling compared with the SAH group (Fig. 4C). The staining was weaker, more focal, and confined mainly to perivascular areas. Similarly, RF-EMF exposure substantially suppressed TNF-α expression (Fig. 4D), resulting in reduced cytoplasmic intensity and more preserved cortical architecture. The combined L-ARG + RF group demonstrated the most striking improvement, showing nearly complete absence of TNF-α positivity and restoration of normal cytoarchitecture (Fig. 4E).

Quantitative immunoscoring analysis supported these histological findings. TNF-α expression was significantly elevated in the SAH group compared with controls (p < 0.001), while all treatment groups showed significant attenuation (p < 0.05 SAH + L-ARG vs. SAH; p < 0.001 SAH + RF vs. SAH; p < 0.001 SAH + L-ARG + RF vs. SAH). The L-ARG + RF combination produced the lowest TNF-α scores among treated groups (Fig. 4).

Cas-3 immunohistochemical findings

In the control group, Cas-3 immunoreactivity was almost absent or limited to scattered neurons showing faint cytoplasmic staining, consistent with physiological baseline apoptotic turnover (Fig. 5A). In sharp contrast, the SAH group exhibited intense, widespread Cas-3 positivity within both neuronal cytoplasm and nuclei, particularly in cortical and subpial regions adjacent to hemorrhagic loci (Fig. 5B). The dense brown staining indicated extensive activation of the intrinsic apoptotic pathway following subarachnoid bleeding.

Fig. 5.

Immunohistochemical detection of Cas-3 in cerebral cortex tissues following SAH and the therapeutic effects of L-ARG and RF-EMF. Representative micrographs show: (A) Control group with negligible Cas-3 immunoreactivity and normal cortical architecture. (B) SAH group exhibiting strong Cas-3 positivity (arrow) in neuronal cytoplasm and nuclei around hemorrhagic foci. (C) SAH + L-ARG group showing reduced Cas-3 staining and partial tissue preservation. (D) SAH + RF group demonstrating moderate suppression of Cas-3 expression and improved neuronal morphology. (E) SAH + L-ARG + RF group with minimal Cas-3 positivity and preserved cytoarchitecture. Scale bar = 50 μm. The accompanying bar graph presents mean ± SD immunoexpression scores analyzed by one-way ANOVA followed by Tukey’s post hoc test (**p < 0.01; ***p < 0.001). Combined treatment significantly reduced Cas-3 activation, highlighting the synergistic anti-apoptotic potential of NO support and RF-EMF-mediated neuroprotection. Abbreviations: SAH, subarachnoid hemorrhage; L-ARG, L-arginine; RF-EMF, radiofrequency electromagnetic field; Cas-3, caspase-3.

Administration of L-ARG resulted in a marked reduction of Cas-3 immunostaining intensity (Fig. 5C). The positive cells were fewer and mainly localized near the perivascular zones, suggesting partial inhibition of apoptosis. Likewise, RF-EMF exposure substantially diminished Cas-3 expression (Fig. 5D), reflecting decreased neuronal apoptotic activity and restoration of normal cortical morphology. The combined L-ARG + RF group displayed the most profound attenuation of Cas-3 immunoreactivity, approaching control levels (Fig. 5E), with preserved neuronal morphology and minimal cytoplasmic staining.

Semi-quantitative analysis confirmed these observations. Cas-3 scores were significantly elevated in the SAH group compared with controls (p < 0.001), whereas all treatment groups exhibited significant reductions (p < 0.01 SAH + L-ARG vs. SAH; p < 0.001 SAH + RF vs. SAH; p < 0.001 SAH + L-ARG + RF vs. SAH). Among them, the L-ARG + RF combination achieved the lowest Cas-3 expression, indicating a synergistic inhibition of neuronal apoptosis (Fig. 5).

HIF-1α signaling pathway gene expression analysis

Quantitative RT-PCR analysis demonstrated significant alterations in the HIF-1α signaling pathway following SAH. Expression levels of IL-6 and STAT3 were markedly elevated in the SAH group compared with controls (p < 0.001 for both). Conversely, HIF-1α, eNOS, and VEGF mRNA levels were significantly down-regulated in the SAH group (p < 0.001 for all).

Treatment with L-ARG alone produced limited molecular effects compared to RF-EMF and combined treatment groups. On the other side, RF-EMF exposure alone resulted in decreased inflammatory gene expression and partial recovery of endothelial and hypoxia-related markers relative to the SAH group (p < 0.001 for STAT3 and IL-6; p = 0.006 for eNOS; not significant for VEGF and HIF-1α). Notably, the combined L-ARG + RF therapy provided the most pronounced effect compared to the injury group, restoring HIF-1α, eNOS, and VEGF expression toward baseline levels (p < 0.001 for all) while markedly suppressing IL-6 and STAT3 (p < 0.001 for both) (Fig. 6).

Fig. 6.

Effects of L-ARG and RF-EMF on HIF-1α-related gene expression in the cerebral cortex after SAH. Relative mRNA expression levels of IL-6, STAT3, HIF-1α, eNOS, and VEGF were quantified by qRT-PCR and normalized to Rn18S. The SAH group exhibited a significant increase in IL-6 and STAT3 expression and a concomitant reduction in HIF-1α, eNOS, and VEGF compared to controls. The combined L-ARG + RF therapy produced the strongest normalization, indicating synergistic modulation of inflammatory and hypoxia-responsive genes. Data are expressed as mean ± SD (n = 8). One-way ANOVA followed by Tukey’s post hoc test was used for statistical comparison (*p < 0.05; **p < 0.01; ***p < 0.001). Scatter plots display individual experimental data points. Abbreviations: SAH, subarachnoid hemorrhage; L-ARG, L-arginine; RF-EMF, radiofrequency electromagnetic field; STAT3, signal transducer and activator of transcription 3; eNOS, endothelial nitric oxide synthase; HIF1α, Hypoxia-Inducible Factor 1-alpha; IL-6, Interleukin-6; VEGF, Vascular Endothelial Growth Factor.

Mitochondrial apoptosis pathway analysis

Quantitative mRNA profiling demonstrated a significant dysregulation of mitochondrial apoptotic signaling following SAH. In the SAH group, pro-apoptotic BAX expression was markedly upregulated (p < 0.001) compared with controls. Concurrently, anti-apoptotic BCL-2 expression was substantially downregulated (p < 0.001).

Treatment with L-ARG only reduced BAX (p < 0.05 vs. SAH) mRNA levels. RF-EMF exposure alone decreased BAX transcription (p < 0.001) and moderately increased BCL-2 levels, but this was insignificant. Importantly, the combined L-ARG + RF therapy exerted the most pronounced effect, significantly suppressing BAX overexpression and reinstating near-normal BCL-2 transcription compared to the SAH group (p < 0.001 and p < 0.05 respectively).

The BAX/BCL-2 ratio was significantly elevated in the SAH and SAH + L-ARG groups compared with controls (p < 0.001 for all). In contrast, this ratio was markedly reduced in the control, SAH + RF, and SAH + L-ARG + RF groups relative to SAH + L-ARG, reflecting restoration of anti-apoptotic balance (p < 0.001 for all). Notably, both RF-EMF treated groups showed a pronounced decrease in the BAX/BCL-2 ratio versus SAH, confirming that RF-EMF exposure-alone or combined with L-ARG effectively suppressed mitochondrial apoptosis (p < 0.001 for all) (Fig. 7).

Fig. 7.

Relative mRNA expression of mitochondrial apoptotic markers (BAX and BCL-2) in cerebral cortex tissues. SAH induced a significant increase in pro-apoptotic BAX and a concurrent decrease in anti-apoptotic BCL-2 expression, resulting in a pro-apoptotic shift. The combined L-ARG + RF group exhibited the greatest normalization of both genes, indicating a synergistic inhibition of mitochondrial apoptosis. Data are presented as mean ± SD (n = 8). Statistical analysis was performed using one-way ANOVA followed by Tukey’s post hoc test (*p < 0.05; ***p < 0.001). Scatter plots represent individual experimental data points. Abbreviations: SAH, subarachnoid hemorrhage; L-ARG, L-arginine; BCL-2, B-Cell Lymphoma 2; BAX, BCL-2-Associated X proteins.

Discussion

The present study demonstrates that combined L-ARG and RF-EMF therapy attenuates molecular and histopathological markers associated with neurovascular injury following SAH by concurrently targeting inflammation, endothelial dysfunction, and mitochondrial apoptosis. The findings demonstrate that SAH provokes intense neuroinflammatory activation (↑NF-κB, ↑TNF-α, ↑IL-6/STAT3), suppression of the HIF-1α/eNOS/VEGF pathway and mitochondrial apoptosis (↑BAX, ↑Cas-3, ↓BCL-2), all of which were significantly reversed by L-ARG and RF-EMF co-treatment.

Following hemorrhage, blood degradation products and damage-associated molecular patterns activate TLR4/NF-κB and IL-6/STAT3 signaling, initiating a cytokine storm that perpetuates vascular inflammation and oxidative stress. As demonstrated by Liu et al.¹⁸, the IL-6/STAT3 axis amplifies microglial reactivity and increased BBB permeability. In the present model, the marked up-regulation of IL-6 and STAT3 mRNA and the strong TNF-α and NF-κB immunopositivity observed in the SAH group confirm this feed-forward inflammatory amplification. Conversely, both L-ARG and RF-EMF treatments significantly attenuated these markers, suggesting that suppression of the IL-6/STAT3–NF-κB inflammatory loop is a key protective mechanism.

Inflammation and endothelial dysfunction are closely interrelated. According to Kagoya et al.¹⁹, TNF-α and NF-κB form a self-sustaining positive feedback circuit, in which TNF-α stimulates NF-κB nuclear translocation and subsequent cytokine release. This reciprocal activation was evident in our SAH model, where intense perivascular TNF-α staining co-localized with NF-κB-positive nuclei, showing exacerbation of BBB breakdown. The decline of both signals following treatment indicates that RF-EMF and L-ARG jointly disrupt this pro-inflammatory cascade, possibly through reduced ROS accumulation and NO-mediated inhibition of NF-κB transcriptional activity.

The down-regulation of HIF-1α, eNOS, and VEGF in the SAH group indicates suppression of hypoxia-adaptive vascular signaling. Semenza et al.⁸ highlighted that HIF-1α serves as a master regulator of oxygen homeostasis, directly inducing eNOS and VEGF to promote vasodilation and angiogenesis. However, Liu et al.²⁰ reported that prolonged IL-6/STAT3 activation impairs HIF-1α stability via ROS-dependent hydroxylation, a mechanism consistent with our findings. Thus, the inflammatory milieu secondary to SAH likely interferes with HIF-1α stabilization, contributing to reduced eNOS-derived NO and endothelial collapse. The normalization of HIF-1α and its downstream targets by combined L-ARG + RF supports a reactivation of this protective hypoxic response.

Endothelial dysfunction not only impairs cerebral perfusion but also contributes to mitochondrial oxidative injury. Doughan et al.²¹ described a bidirectional crosstalk between vascular endothelium and mitochondria, whereby increased mitochondrial ROS generation results in decreased NO bioavailability causing mitochondrial dysfunction through a self-sustaining positive feedback loop, sensitizing cells to apoptosis. The present results support this interconnection: as eNOS and VEGF expression recovered, the pro-apoptotic markers BAX and Cas-3 declined, and anti-apoptotic BCL-2 increased, indicating restoration of mitochondrial integrity through improved endothelial function.

The shift in the BAX/BCL-2 ratio and activation of Cas-3 observed in the SAH group reflect classical mitochondrial-mediated apoptosis. Spierings et al.²² emphasized that mitochondrial permeability transition driven by oxidative stress results in Cyt-C release and Cas-3 activation, leading to cell death. Furthermore, Zhu et al.²³ demonstrated that similar pathways play a role in neuronal loss induced by neonatal cerebral hypoxia-ischemia. In light of our findings in SAH group, Cyt-C release and Cas-3 activation is likely responsible for delayed neuronal death after SAH. The substantial decrease in Cas-3 immunoreactivity and normalization of BAX/BCL-2 expression in combined treated groups suggest that L-ARG and RF-EMF together inhibit this cascade by maintaining redox balance and mitochondrial potential.

L-ARG confers protection primarily through its role as a substrate for NOS. Yamada et al.²⁴ showed that L-ARG enhances endothelial NO production via eNOS phosphorylation, which improves cerebral blood flow. Hamidi et al.²⁵ proposed that L-ARG also exerts anti-inflammatory and anti-apoptotic actions through cGMP-dependent signaling. This aligns with our observation that L-ARG suppressed Cas-3 and NF-κB levels, implying that NO-mediated signaling contributes to both mitochondrial preservation and anti-inflammatory response. Notably, because L-ARG alone did not significantly restore eNOS mRNA expression in our model, its beneficial effect should not be interpreted as direct transcriptional activation of eNOS, but rather as support of nitric oxide availability or downstream endothelial signaling.

On the other hand, RF-EMF exposure may exert predominantly non-thermal bioeffects, although thermal contributions cannot be completely excluded. Therefore, conclusions regarding strictly non-thermal mechanisms should be interpreted with caution. Pall et al.²⁶ demonstrated that low-intensity RF fields activate calcium/calmodulin-dependent protein kinase pathways, thereby enhancing eNOS activity and mitochondrial respiratory efficiency. In this study, RF-EMF alone produced substantial down-regulation of NF-κB and Cas-3, supporting the hypothesis that electromagnetic stimulation restores calcium-dependent NO synthesis and reduces oxidative apoptotic stress.

When administered together, L-ARG and RF-EMF produced the most consistent improvement across histological and molecular injury associated markers in this model. Gul et al.²⁷ reported that dual metabolic and electromagnetic modulation exerts greater anti-oxidative and anti-apoptotic effects than either intervention alone by synchronizing mitochondrial redox and endothelial NO signaling. Consistent with this, our combined treatment group exhibited the lowest histopathological damage, minimal TNF-α and Cas-3 expression, and near-normal HIF-1α/eNOS/VEGF levels. The convergence of metabolic (NO-driven) and biophysical (RF-EMF-induced) modulation appears to harmonize vascular tone, suppress inflammation, and stabilize mitochondrial membranes, providing comprehensive protection against secondary brain injury. However, formal interaction testing was not performed for all endpoints, and these findings should be interpreted as enhanced combined effect rather than definitive synergy.

These findings highlight the intricate cross-talk between inflammation, hypoxia signaling, and mitochondrial apoptosis. The IL-6/STAT3-NF-κB axis drives inflammatory gene expression²⁸ which in turn suppresses HIF-1α and eNOS activity, compromising vascular function^29[,30. Reduced NO exacerbates oxidative mitochondrial stress, triggering BAX/Cas-3-dependent apoptosis, which further amplifies inflammation through damage-associated molecular pattern release effectively interrupted by combined therapy.

The translational implication of these results lies in targeting multiple converging pathways rather than isolated molecular events. Fujii et al.³¹ emphasized that multimodal interventions that simultaneously restore endothelial function and suppress neuroinflammation yield superior outcomes in experimental SAH. Our findings support this notion, suggesting that metabolic–biophysical co-therapy could represent a new paradigm in neurocritical management. Given the non-invasive and biocompatible nature of RF-EMF exposure and the clinical availability of L-arginine, this dual approach may offer a feasible adjunctive strategy for mitigating secondary injury and preventing delayed cerebral ischemia following SAH.

Despite its mechanistic depth, this study has several limitations. First, it is an experimental animal model; therefore, extrapolation to human SAH should be approached cautiously. Second, the study was designed to assess molecular and histopathological injury markers rather than functional recovery. Behavioral testing, neurological severity scoring, cerebral blood flow assessment, angiographic evaluation of vasospasm, and direct perfusion measurements were not included. Therefore, the present findings should not be interpreted as evidence of functional therapeutic efficacy, but rather as evidence of treatment-associated modulation of inflammatory, endothelial, and apoptotic injury pathways. Thirdly, the single-time-point assessment limit temporal resolution of molecular dynamics. Moreover, while RF-EMF parameters were standardized, quantitative dosimetry and field mapping within brain tissue were not performed, therefore, thermal contributions cannot be excluded. Although primer specificity was verified, primer efficiency analysis was not performed, which represents a methodological limitation of the present study. Another limitation of this study is the reliance on semi-quantitative histopathological and immunohistochemical scoring, which may introduce observer-dependent variability despite blinded evaluation. Future studies incorporating fully quantitative image-based morphometric or automated digital pathology analyses would strengthen objectivity. Circulating inflammatory cytokines were not directly assessed postoperatively, which limits the interpretation of systemic inflammatory responses following surgery. Finally, the sample size was modest, and formal interaction testing was not performed for all endpoints; therefore, the apparent greater effect of combined L-ARG and RF-EMF treatment should not be interpreted as definitive biological synergy. Future studies should prospectively incorporate behavioral outcomes, perfusion endpoints, vasospasm assessment, dose–response analyses, and dedicated dosimetry before translational conclusions are drawn.

In conclusion, this study shows that combined L-ARG and RF-EMF treatment was associated with reduced histopathological injury and normalization of inflammatory, endothelial, and mitochondrial apoptosis related markers after experimental SAH. By down regulating IL-6 STAT3 NF-κB TNF-α signaling, supporting HIF-1α eNOS VEGF related endothelial pathways, and improving BAX BCL-2 Cas-3 associated mitochondrial responses, the combined treatment produced the most favorable mechanistic profile among groups. These findings support a promising experimental co therapy that requires confirmation using functional behavioral outcomes, perfusion endpoints, and dedicated dosimetry and temperature monitoring before translational conclusions can be drawn.

Author contributions

ATB, MC, RT, HA, and SC contributed to the investigation. ATB, HA, and SC contributed to the methodology. ATB, HA, MYT, OK, AM, and OO were responsible for data curation. ATB, MC, HA, and SC performed validation. OK, HA, AM, and OO were involved in visualization. ATB, MC, and HA prepared the original draft. MYT conducted the molecular analysis. MYT, AM, and OO performed the formal analysis. AM and OO carried out the histopathological and immunohistochemical evaluations. HA contributed to the conceptualization, project administration, and supervision. All authors contributed to the review and editing of the manuscript. All authors read and approved the final version of the manuscript.

Funding

The study was funded by the corresponding author.

Data availability

The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

Ethics approval

All animal experiments were approved by the Mehmet Akif Ersoy University Local Animal Experiments Ethics Committee (MAKU-HADYEK) with the decision dated 09.07.2025, no: 137/1527. All procedures were performed in accordance with the ARRIVE 2.0 guidelines and internationally accepted standards for animal experimentation.