A temperature-ultrasound sensitive nanoparticle delivery system for exploring central neuroinflammation mechanism in stroke-heart syndrome

Abstract

Background

Cardiovascular events secondary to stroke–collectively classified as stroke-heart syndrome–greatly impair the patient’s prognosis, however its underlying mechanism has yet to be determined. To investigate the mechanism of central neuroinflammation and its effects on stroke-heart syndrome, a temperature-ultrasound responsive brain-targeted drug delivery system, DATS/MION-LPE, was synthesized to specifically study neuroinflammation in the mouse middle cerebral artery occlusion (MCAO) model.

Results

The specific polymer of DATS/MION-LPE can close the nanoparticle pores at 37 °C, restricting drug release in the circulation. After the nanoparticles were targeted to brains, the polymer can be cleaved under external ultrasound irradiation, reopening the nanoparticle pores and allowing drug release, therefore directly managing the neuroinflammation. After a stroke, a significant cerebral inflammation occurred, with elevated IL-1β and pyrin domain-containing 3 (NLRP3) inflammasome. Accordingly, significantly increased histone deacetylase 6 (HDAC6) and decreased sirtuin 1 (SIRT1) were observed. An antagonistic relationship between HDAC6 and SIRT1 was found, which can jointly regulate the cerebral NLRP3 expression. The systemic IL-1β and ATP levels were increased after the stroke, accompanied by a significant heart injury including contractile dysfunction, elevated IL-1β levels, and oxidative stress. Meanwhile, neuroinflammation can trigger sympathetic nervous overexcitation with associated heart damage. DATS/MION-LPE can targetedly effect on ischemic brain, exhibiting cerebral and cardiac protective effects including downregulated cerebral NLRP3 and HDAC6 expressions, upregulated SIRT1 expressions in brain, reduced IL-1β and ATP in circulation, and alleviated cardiac impairment.

Conclusion

This study introduced the key role of neuroinflammation in stroke-heart syndrome and first investigated the crucial HDAC6/SIRT1-NLRP3 circuit in this process. Heart injury secondary to stroke is mediated by neuroinflammation induced systemic inflammatory responses and sympathoexcitation. DATS/MION-LPE is a unique tool and effective therapeutic agent, which provides new insights into combinational heart and cardiac protection.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12951-024-02961-z.

Introduction

As a leading cause of global deaths, stroke is frequently complicated by secondary cardiac injuries including cardiac dysfunction, arrhythmia, and heart failure, which is also known as the stroke-heart syndrome [1–4]. To date, the underlying mechanism of the stroke-heart syndrome has not been fully investigated, while neuroinflammation has been recognized as the most important mechanism [5–7]. Some researchers have reported that cerebral ischemia can locally produce an abundance of cerebral inflammatory cytokines including NLRP3 and proinflammatory cytokine interleukin-1β (IL-1β), which could compromise the blood-brain barrier (BBB) and penetrate IL-1β and adenosine triphosphate (ATP) into circulation, resulting in systemic inflammation and secondary cardiac injury [8–10]. Elevated cerebral cytokine can also activate the hypothalamic-pituitary-adrenal axis, leading to an overactive sympathetic nervous system and stress-induced cardiac damage [11, 12]. Therefore, resulting in secondary systemic inflammation and sympathetic overactivity, neuroinflammation plays a critical role to stroke-induced heart damage. However, the underlying mechanism of neuroinflammation post -stroke is yet to be investigated.

Recent research has indicated that the nucleotide-binding domain, leucine-rich–containing family, NLRP3 inflammasome, is essential in cerebral ischemic neuroinflammation, which can produce the proinflammatory cytokine interleukin-1β (IL-1β) by elevating Caspase-1 [13]. HDAC6 can facilitate the microtubule polymerization process in the assembly of the NLRP3, while the genetic suppression of HDAC6 can effectively reduce inflammatory responses [14]. On the contrary, an anti-inflammatory molecule SIRT1, which belongs to the sirtuin family of deacetylases, can mitigate NLRP3-mediated inflammatory responses by regulating NF-κB and Akt [15, 16]. Furthermore, SIRT1 and HDAC6 can exhibit reciprocal regulatory effects by deacetylation and can also both serve as regulating molecules for NLRP3 expression [17]. However, the role of the “SIRT1-HDAC6-NLRP3 circuit” in either neuroinflammation or the subsequent stroke-heart syndrome has yet to be determined.

As a reductive gaseous signaling molecule, H₂S can suppress inflammation after ischemia and effectively reduce IL-1β production [18], presenting protective effects in both the ischemic brain and myocardium [19–21]. Therefore, H₂S could exert clear anti-inflammatory effects after cerebral ischemia and can be used as an ideal tool to study the neuroinflammation mechanism. However, conventional H₂S donors release H₂S rapidly in circulation without targeted delivery, which makes it difficult to attain therapeutic concentrations in specific organs. Therefore, there is still a high demand for a novel H₂S donor that can specifically release H₂S in targeted organs such as the brain, which can be utilized in the study of neuroinflammation mechanisms and secondary heart injury.

In the present study, we developed a brain-targeted drug delivery system to deliver H₂S specifically to the ischemic brain, which selects diallyl trisulfide (DATS) as the H₂S donor which could generate H₂S through enzymatic breakdown by reduced glutathione (GSH) and cystathionine γ-lyase (CSE) [22]. This system utilized mesoporous iron oxide nanoparticles (MIONs) as the frameworks, with temperature and ultrasound-sensitive copolymer p-(MEO₂MA-co-THPMA) grafted as goalkeepers to restrict systemic drug release. The lower critical solution temperature (LCST) of p-MEO₂MA) part of the copolymer allows it to respond to temperature changes. Below LCST, it exhibits hydrophilic properties, allowing the drug loading at 4 °C. When the temperature exceeds the LCST such as at 37 °C, the copolymer collapses to block the nanopores and restrict drug release. While the p-(THPMA) part can make the copolymer to be cleaved by external ultrasonic stimulation, reopening the nanopores to enable drug release. The nanoparticles were then encapsulated within the erythrocyte membranes (EM) to enhance the circulation time. And lactoferrin (LF) was attached to MION as the brain-targeting ligand which has been widely utilized in previous studies [23–25]. After loading of DATS, this delivery system, DATS/MION-LF-p(MEO₂MA-co-THPMA)-EM, referred to as DATS/MION-LPE, can reach and accumulate in the brain, releasing H₂S without systemic side effects. This novel brain-targeted H₂S delivery system was used to determine the central role of the HDAC6/SIRT1-NLRP3 circuit in the neuroinflammation mechanism and secondary heart injury, bearing the potential to be both an ideal study tool and a therapeutic agent in stroke-induced heart damage (Fig. 1).

Fig. 1.

Schematic representation of the synthesis and working mechanism of the novel brain-targeted H₂S controlled-release system, DATS/MION-LPE. The therapeutic agent is loaded into the MION-LPE for a targeted delivery and accumulation in the brain, while minimizing systemic release. External ultrasound stimulation facilitates the disintegration of particles and the subsequent release of H₂S. This system is utilized to investigate the mechanisms of stroke-heart syndrome and the therapeutic effects of DATS/MION-LPE. EM: erythrocyte membranes; LF: lactoferrin; DATS: diallyl trisulfide

Materials and methods

Materials and reagents

Diallyl trisulfide (DATS) and lactoferrin (LF) were purchased from Aladdin Chemical (Shanghai, China). MEO₂MA, THPMA, maleimide-Polyethylene Glycol-N-Hydroxysuccinimide (Mal-PEG-NHS), and N-succinimidyl-S-acetylthioacetate (SATA) were obtained from Sigma-Aldrich (St. Louis, MO). The N₂ atmosphere was used to deaerate the solutions before the experiments. The Supplementary Materials show the detailed information regarding the materials, reagents, apparatus, and characterization.

Synthesis of MION-LPE

A mixture of styrene, methacrylic acid, potassium persulfate, and other components was dispersed in deionized water and heated to generate polystyrene nanoparticles. Potassium nitrate, ferrous chloride, and methylbenzylamine were then introduced to synthesize MIONs. Using 3-aminopropyltriethoxysilane (APTES) and Mal-PEG-NHS, the intermediate Mal-PEG-MION was synthesized. Moreover, brain targeted ligand LF and SATA were combined with Mal-PEG-MION to obtain MION-LF. The copolymer p-(MEO₂MA-co-THPMA) was then synthesized from THPMA, MEO₂MA, and other reagents through heating, cooling, rewarming, co-precipitation with petroleum ether, and lyophilization under vacuum. MION-LF was mixed with butenoic acid to bind vinyl groups and then combined with the copolymer p-(MEO₂MA-co-THPMA) and heated under nitrogen. The product MION-LF-p(MEO₂MA-co-THPMA) was obtained by lyophilization. The EM were coated on the particle surface to produce MION-LF-p(MEO₂MA-co-THPMA)-EM, which is referred to as MION-LPE. MION-LPE was stored in the 37 °C incubator to maintain its stability and function. The Supplementary Materials present the detailed protocol.

Characterization

Transmission electron microscopy (TEM) imaging was performed using a JEM-2100 instrument (Hitachi, Tokyo, Japan). The fluorescent characteristics were analyzed using the Biotek Synergy 4 system. The particle size of the synthesized nanoparticles (NPs) was determined via dynamic light scattering (DLS) on a Mastersizer2000 (Malvern Instruments Inc, UK). The surface area and pore size distributions were characterized via N₂ adsorption using the Micromeritics ASAP 2010 instrument. The surface areas were calculated with Brunauer–Emmett–Teller method, whereas the pore sizes were determined with Barrett–Joyner–Halenda (BJH) technique. Infrared spectra were acquired with a Nicolet Nexus spectrometer. Proton NMR (¹H-NMR) spectra for p-(MEO₂MA-co-THPMA) were recorded on a Bruker spectrometer.

In vitro drug release study

The Supplementary Material showed the detailed protocols for loading DATS into MION or MION-LPE. DATS/MION-LPE was stored and transported in a 37 °C incubator to maintain their stability and function. For the drug release assessment, DATS/MION or DATS/MION-LPE was dispersed in phosphate-buffered saline (PBS) at 50 µg/ml and a pH of 7.4. The samples were collected at 1, 2, 4, 8, 12, 24, and 48 h after exposure to various temperatures and ultrasound stimuli. We evaluated the safety of ultrasound stimulation using different ultrasound frequencies (25 kHz, 50 kHz, 1 MHz, and 2 MHz) for 10 min at mice brain, with the spatial and temporal power densities at approximately 0.5 W cm⁻². The detailed description was described in the Supplementary Materials. The H₂S concentrations were quantified using a high-performance liquid chromatography system (Agilent Technologies, 1260 Infinity, CA, USA) with fluorescence detection at excitation and emission wavelengths of 390 nm and 475 nm, respectively [26, 27]. Chromatographic separation was achieved using an Eclipse XDB-C18 column (150 × 4.6 mm, 5 μm). Furthermore, the entire analysis was completed within 48 h.

In vitro cytotoxicity assessment and protection effects of DATS/MION-LPE

The Supplementary Materials present the detailed methods of cell acquisition and culture. The H9C2 and PC12 cells were treated with DATS/MION-LPE in their culture media. The compound was added to 96-well plates, each containing 1 × 10⁴ cells in a 100 µL medium, with DATS/MION-LPE concentrations adjusted to 0, 5, 10, 25, 50, 80, and 100 µg/mL. After 24 h of incubation, the cells were washed with PBS and treated with CCK-8 solution (10 µL) per well, and the absorbance of per well was measured at 450 nm. The control group received an equivalent volume of deionized water. Cell viability was reported as a percentage of the control group. Moreover, lactate dehydrogenase (LDH) levels were assessed with LDH assay kit (Sangon, Shanghai, China).

To verify the protective effects of DATS/MION-LPE on in vitro cellular hypoxia/reoxygenation injury, the cardiomyocytes and neurons were treated with DATS/MION-LPE at concentrations ranging from 4 to 100 µg/ml, in combination with 2 mM glutathione (GSH). The control group was established with an equivalent volume of PBS containing only 2 mM GSH. After ultrasound stimulation (1 MHz, 10 min), the cells were incubated for an additional 4 h, followed by a 4 h reoxygenation process. Moreover, the cell viability and LDH levels were measured.

Establishment of the animal model

All experiments were performed in compliance with the relevant laws and institutional guidelines. All animal experiments were conducted in accordance with the ethical guidelines set forth by the Institutional Animal Care and Use Committee of the Department of Laboratory Animal Science at Fudan University (Grant No. 20160780A176) and adhered to the principles outlined in the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH publication no. 85 − 23, revised 1996). The detailed number of experimental animals is provided in the Supplementary Material (Table. S1).

C57BL/6 N male mice (20 g, 8 weeks old) were obtained from Vital River Laboratory Animal Technology Co., Ltd. (Beijing). Transient focal cerebral ischemia was induced in mice through 60 min of MCAO on the left side. The Supplementary Materials present the detailed methods. The sham operation group underwent the same surgical process without inducing an occlusion. Laser Speckle Imaging (LSI) was used to evaluate cerebral blood flow changes using the speckle technique in the MCAO and Sham groups. 20 mg/kg of DATS/MION-LPE and same volume of PBS were administrated in MCAO models, and ultrasound stimulation on brain was performed at 4 h after the treatments. At 14 d after surgery, brains and hearts were harvest for following experiments.

To ablate the primary afferent neurons, 50 mg/ml of capsaicin in olive oil (Aladdin, Shanghai, China) was administered via subepicardial injection 2 w before MCAO induction. Olive oil solvent injection was applied in the Control (n = 6). To stimulate the sympathetic nervous system, another group of mice (n = 6) received a daily oral gavage of pseudoephedrine at 20 mg/kg/d (Aladdin, Shanghai, China).

In vivo safety assessment and pharmacokinetic experiments of DATS/MION-LPE

The mice were administered with DATS/MION-LPE intravenously at 20 mg/kg through the tail vein. Euthanasia was performed at 24 h, 7 d, 14 d, and 30 d after drug administration, and major organs were collected including the brain, heart, liver, spleen, and kidneys (n = 3). The tissue sections were prepared at a thickness of 5 μm and stained with hematoxylin and eosin. Numerous high-magnification fields were randomly selected for microscopic examination. In a parallel experiment, a group of mice (n = 6) were administered with either DATS/MION-LPE (20 mg/kg) or an equivalent volume of PBS via tail vein injection. Following the injection, the mice were anesthetized with ketamine hydrochloride (50 mg/kg) to facilitate femoral artery cannulation. The heart rates and blood pressures were continuously monitored using a SMUP-E4 biosignal processing system over 16 h. The blood samples for hematological and serological analysis were collected at 2 h, 24 h, and 7 d after drug injection. Moreover, DATS/MION (20 mg/kg) and DATS/MION-LPE (20 mg/kg) were administered for H₂S measurement. The blood samples were then collected at 1, 2, 4, 8, 12, 16, 24, 36, and 48 h after drug injection, whereas the plasma H₂S concentrations were measured after centrifugation. The Supplementary Materials present the detailed methods.

In vivo targeting ability of DATS/MION-LPE to the brain

The fluorescent dye DiR was separately loaded into MION-LPE or MION, and the C57 mice (n = 15) were intravenously injected with DiR@MION-LPE (20 mg/kg) or DiR@MION (20 mg/kg). At 1, 2, 4, 8, and 24 h after drug injection, the mice were euthanized, and major organs were collected for the quantitative measurement of the fluorescence signal. After the administration of DATS/MION-LPE, ultrasound stimulation (1 MHz, 10 min) was applied to the heads of the mice at 2, 4, and 8 h after drug injection. Immediately after ultrasound stimulation (1 MHz, 10 min), their brain tissues were collected for the determination of H₂S concentration. The Supplementary Materials present the detailed methods.

Histological analysis of the brain and heart

DiR@MION-LPE or DiR@MION (20 mg/kg) was injected into the mice via the tail vein, followed by euthanasia after 24 h. Brain tissues were weighed, fixed in 4% formaldehyde, and embedded in paraffin for section preparation. The neurons were detected using a neurofilament (NF) protein NF200 antibody (Sigma, 1:500), and the nuclei were stained with 4’,6-diamidino-2-phenylindole (DAPI). The DiR signals (excitation at 750 nm, emission at 782 nm) were visualized using fluorescence microscopy (Leica Camera Co., Wetzlar, Germany). Following the capsaicin subepicardial injection and MCAO surgery, the hearts were fixed in formaldehyde. The paraffin-embedded sections were stained with an antiNF antibody (1:500), antityrosine hydroxylase (TH) antibody (1:500), and anticalcitonin gene-related peptide (CGRP) antibody (1:200). The quantification of the TH-positive and CGRP-positive nerve fibers on the epicardium was performed using the ImageJ software.

Echocardiography and hemodynamic measurement

To confirm the protective effects of DATS/MION-LPE, the mice were divided into three groups (n = 8): the DATS/MION-LPE group (MCAO with 20 mg/kg DATS/MION-LPE administration), vehicle-treated group (MCAO with 20 mg/kg PBS administration), and sham operation group. Ultrasound stimulation on brain was performed at 4 h after the treatments. The echocardiographic assessments were conducted using Vevo 3100 echocardiography, capturing the M-mode images 2 w after surgery. The measurements covered the left ventricular end-diastolic volume (LVED), ejection fraction (EF), and fractional shortening (FS). The survival rates were tracked for 30 d, and the heart weights (HW) were measured after euthanasia. The brain and heart tissues from each group were collected for western blot analysis.

To investigate the role of the sympathetic nervous system in cardiac injury after stroke, the mice were divided into five groups: the DATS/MION-LPE group (MCAO with 20 mg/kg DATS/MION-LPE administration), vehicle-treated group (MCAO with 20 mg/kg PBS administration), pseudoephedrine group, capsaicin group, and sham operation group (n = 8). For the pseudoephedrine and capsaicin groups, the MCAO surgery was performed at 14 d after drug administration, and the same echocardiographic assessments were conducted at 14 d after surgery. Furthermore, the brain and heart tissues from each group were collected for western blot analysis.

Stereotactic virus injection for the overexpression of proteins in the brain

The pAAV-hSyn-SIRT1-3xFLAG-P2A-EGFP-tWPA (5.4 × 10¹² vg/ml), pAAV-hSyn-HDAC6-3xFLAG-tWPA (5.4 × 10¹² vg/ml), and their respective vectors pAAV-hSyn-MCS-3xFLAG-P2A-EGFP-tWPA (1.64 × 10¹³ vg/ml) and pAAV-hSyn-MCS-3xFLAG-tWPA (1.64 × 10¹³ vg/ml) were all produced by OBio (Shanghai, China). The mice were injected into CA1 with the viral solution or vectors and then subjected to MCAO surgery. Seven days after MCAO surgery, the mice were euthanized, and fresh brain tissues were harvested for the western blot analysis. To investigate the role of HDAC6/SIRT1-NLRP3 circuit, mice brain tissues were collected for Western blot analysis at 7 d after injection before the peak of adenovirus-induced gene expression declines. In a parallel experiment, each group of mice underwent echocardiographic assessments at 7 d after injection. Blood samples were also collected at the same time to measure plasma ATP and IL-1β levels. The Supplementary Materials present the detailed methods.

Western blot assay

The brain or heart tissue samples were harvest at 14 d after surgery, and homogenized in precooled RIPA buffer using a rotor-stator homogenizer until the tissues were uniformly processed. The proteins were separated using 8% tris-glycine SDS-PAGE gels and subsequently transferred to the PVDF membranes. The antibodies used are listed as follows: NLRP3 (1:2000), HDAC6 (1:2000), SIRT1 (1:2000), Caspase-1 (1:2000), IL-1β (1:2000), β-tubulin (1:2000), β-actin (1:2000), and goat antirabbit secondary antibody (1:10,000). The bands were visualized using the Super Signal West Pico chemiluminescent substrate (Pierce, Pittsburgh, PA, USA) and imaged using the FluorChem Image System for densitometric analysis, normalizing for β-tubulin and β-actin.

Inflammation and antioxidant enzyme assessments for the blood and hearts

Twenty-four hours after surgery, the plasma ATP levels were measured by enzyme-linked immunosorbent assay (ELISA) kits (Nanjing Jiancheng Bioengineering Institute, Nanjing, China). The plasma IL-1β concentrations were determined using the IL-1β determination kits (Jianglai Bio). Fourteen days after surgery, the heart tissues were homogenized in ice-cold potassium phosphate buffer (50 mM, PH 6.98). The activities of the superoxide dismutase (SOD), catalase (CAT), and malondialdehyde (MDA) levels were quantified using the ELISA kit in accordance with the previously reported standard protocols [28]. The SOD and CAT activities, as well as the MDA levels, were standardized by protein content with the bicinchoninic acid assay.

Statistical methods

All data in this study were analyzed with SPSS Statistics Base 17.0. The continuous data were expressed as mean ± standard error of the mean. The statistical comparisons between the groups were assessed using one-way analysis of variance. The significant difference between the two groups was analyzed using Student’s t-test. Survival condition was analyzed using the Kaplan–Meier method. Statistical significance was set at a P value of < 0.05. The significance levels were denoted using asterisks: (*) for P < 0.05, (**) for P < 0.01, and (***) for P < 0.001.

Results

Characterization

The TEM images of the MIONs were presented in Fig. 2a, showing uniform geometric shapes and mesoporous structures. As presented in Fig. 2b, the morphology assessment showed the consistent attachments of LF and p(MEO₂MA-co-THPMA) onto the MIONs’ surfaces. The encapsulation of MION-LPE by the EM and the collapse of the copolymers p(MEO₂MA-co-THPMA) under ultrasound stimulation were further confirmed (Fig. 2c-f). The pore sizes of typical mesoporous of MIONs range from 1 to 9 nm (Fig. 2g), with the presence of a Type IV isotherm hysteresis loop (Fig. 2h). The average weight of MION-LPE is 1.43 times to that of MION-LF (Fig. 2i); the drug loading content and drug loading efficiency for DATS/MION-LPE were 43.66% ± 2.19% and 49.11% ± 3.36% (Fig. 2j); and the size of the DATS/MION-LPE particles ranged from 210 to 230 nm (average size of 217 ± 9.9 nm and a PDI of 0.265 ± 0.15) (Fig. 2k). When dispersed in saline, the DATS/MION-LPE maintained their size stability for up to 7 d (Fig. 2l).

Fig. 2.

Characterization of DATS/MION-LPE during synthesis and the drug loading process. The TEM images of MION (a), MION-LP (b), and DATS/MION-LPE (c). (d-f) The TEM images of the gradual disintegration of polymers encapsulated by the EM in DATS/MION-LPE under ultrasonic stimulation. (g) Pore size distribution of MION. (h) N₂ adsorption/desorption isotherms of MION. (i) Average weight of MION, MION-LF, and MION-LPE (n = 6). (j) Average weight of MION and MION-LPE before and after drug loading (n = 6). (k) The DLS results of DATS/MION and DATS/MION-LPE. (l) Stability of DATS/MION and DATS/MION-LPE in saline (n = 3). (m) Infrared spectra of MION and MION-LF in the range of 400–4000 cm^− 1, with the observation of C–O vibration of PEG alcohol molecules at 1112 cm^− 1, which is the functional group linking MION and LF. (n) Infrared spectra of MION and MION-LF in the range of 400–800 cm^− 1, showing typical C–S bond evidence at 667 cm^− 1 due to the binding of thiolated LF. (o) Infrared spectrum of the copolymer p(MEO₂MA-co-THPMA) showing the characteristic peaks of THPMA at 1104 cm^− 1 and those of MEO₂MA at 1723 cm^− 1. (p) Comparative infrared spectra of MION-LF and MION-LF-p(MEO₂MA-co-THPMA) in the range of 400–4000 cm^− 1, observing the characteristic peaks of the PEG alcohol molecules and THPMA at 1112 cm^− 1 and the vinyl functional group on MION at 1650 cm^− 1. (q) ¹H NMR spectra of p(MEO₂MA-co-THPMA) before and after 10 min of ultrasonic treatment (1 MHz)

The Fourier-transform infrared spectroscopy (FT-IR) for MION, MION-LF, and MION-LPE all indicated the characteristic Fe-O bond vibration around 550 cm^− 1, confirming the core iron oxide composition (Fig. 2m, n, p). The presence of the C–O bond at 1112 cm^− 1 for the MION-LF signals confirmed the attachment of PEG as a linker between MION and LF (Fig. 2m). The sulfuration of LF was confirmed by the appearance of a C–S bond peak at 667 cm^− 1 (Fig. 2n). Moreover, the temperature- and ultrasound responsive functionalities p(MEO₂MA-co-THPMA) were confirmed by the characteristic peaks of THPMA at 1104 cm^− 1 and MEO₂MA at 1723 cm^− 1 (Fig. 2o) based on the corresponding peaks detected in the MION-LPE spectrum (Fig. 2p). The fracture of THPMA in p(MEO₂MA-co-THPMA) under external ultrasound stimulation was confirmed by the ¹H NMR spectral images (Fig. 2q).

In vitro and in vivo drug release and biosafety assessments

The in vitro drug release assessment of DATS/MION-LPE under different temperatures and ultrasound stimulations were monitored over 48 h. The DATS/MION-LPE showed minimal drug release under 37 °C (Fig. 3a), which was triggered by subsequent ultrasound stimulation (Fig. 3b). In the in vitro drug release experiments, the stimulation effects of 1 MHz and 2 MHz on DATS/MION-LPE drug release were quite close. For better penetration depth, the optimal frequency for triggering drug release was determined to be 1 MHz (Fig. 3c). The DATS/MION-LPE presented negligible cytotoxicity in cardiomyocytes and neurons (37 °C, 1 MHz, 10 min). The survival rates for both cardiomyocytes and neurons remained above 90% at concentrations of 0–25 µg/ml and above 80% at 100 µg/ml (Fig. 3d). The in vitro biosafety was further confirmed via the LDH assay (Fig. 3e). In the hypoxia-reoxygenation injury model, the DATS/MION-LPE significantly protected the cardiomyocytes and neurons, especially at 8 µg/ml for cardiomyocytes and 10 µg/ml for neurons (Fig. 3f).

Fig. 3.

Drug release characteristics and toxicity assessment of DATS/MION-LPE in vivo and in vitro under different external stimuli. (a) The in vitro drug release rates of DATS/MION-LPE and DATS/MION at different temperatures (n = 3). (b) Drug release rates under ultrasonic stimulation (n = 3). (c) Release efficiency of DATS/MION-LPE at different ultrasonic frequencies (n = 3). (d) (e) Cell viability and LDH activity of cardiomyocytes and neuronal cells after the administration of various concentrations of DATS/MION-LPE (n = 6). (f) Cell viability of cardiomyocytes and neuronal cells after treatment with various concentrations of DATS/MION-LPE, following 4 h of hypoxia and 4 h of reoxygenation (n = 6). (g, h) Heart rate changes and arterial pressure changes in mice within 16 h after drug injection of the drug (n = 6). (i) Temporal changes in plasma H₂S concentrations after the injection of DATS/MION and DATS/MION-LPE (n = 3)

The in vivo safety assessment of ultrasound stimulation at different frequencies all revealed no obvious histopathological changes in the brain (Fig. S1). And the in vivo safety assessment performed after DATS/MION-LPE administration revealed no obvious histopathological changes in the major organs (Fig. S2). Furthermore, the heart rates and blood pressures were not significantly affected, indicating hemodynamic stability after drug administration (Fig. 3g, h). The hematological and biochemical analyses also revealed no obvious systemic toxicity associated with the DATS/MION-LPE administration (Table. S2). Compared with the rapid release features of DATS/MION with peak plasma H₂S concentrations at 8 h after drug administration, the DATS/MION-LPE showed minimal H₂S release in circulation (Fig. 3i).

In vivo cerebral targeted ability of DATS/MION-LPE

The in vitro imaging experiments confirmed that DiR@MION-LPE and DiR@MION can exhibit fluorescence at specific wavelengths (Fig. S3). The in vivo imaging showed the significant brain-targeted ability of DiR@MION-LPE; fluorescent signal accumulation was also detected in the livers, where nanoparticles are primarily metabolized (Fig. 4a). The immunofluorescence imaging of the brain tissues indicated that DiR@MION-LPE was mainly localized in the neuronal cells (Fig. 4b). The quantitative fluorescence analysis further confirmed a markedly increased brain accumulation of DiR@MION-LPE compared with DiR@MION (Fig. 4c, d). Accordingly, after loading DATS instead of DiR, the external ultrasound stimulation (1 MHz, 10 min) can significantly increase the cerebral H₂S content in the DATS/MION-LPE group compared with the DATS/MION group (Fig. 4e).

Fig. 4.

Brain-targeted drug release characteristics of DATS/MION-LPE. (a) Fluorescence signals in the major organs at 1 h, 2 h, 4 h, 8 h, 24 h, and 48 h after the injection of DiR@MION-LPE and DiR@MION. (b) Immunofluorescence images of frozen brain tissue 24 h after the injection of DiR@MION-LPE and DiR@MION. (c) Comparison of fluorescence signal intensities in the liver. (d) Comparison of fluorescence signal intensities in the brain (n = 3). (e) H₂S content in the brain tissue at 2 h, 4 h, and 8 h after drug administration of DATS/MION-LPE, followed by ultrasonic stimulation (n = 3)

DATS/MION-LPE effectively alleviates post-stroke Cardiac Injury

DATS/MION-LPE was administered at 1 h after MCAO. Ultrasound stimulation (1 MHz, 10 min) was then applied at 4 h after drug administration. The vehicle-treated group and the sham group served as the control (Fig. 5a). The successful establishment of the stroke model was confirmed by brain perfusion imaging and the cerebral blood flow assessment (Fig. 5b, c). The reprehensive echocardiography images were recorded at 14 d after MCAO (Fig. 5d). Although all stroke mice showed an increased LVED volume and impaired EF and FS values, the DATS/MION-LPE administration significantly alleviated these cardiac injuries (Fig. 5e-g), with reduced cardiac atrophy (Fig. 5h). The survival rates in the DATS/MION-LPE group were higher than that in the vehicle-treated group (Fig. 5i). Interestingly, Ex Vivo heart perfusion experiments demonstrated that the ultrasound stimulation (1 MHz, 10 min) on hearts treated with DATS/MION-LPE also alleviated the cardiac oxidative stress damage (Fig. S4).

Fig. 5.

H₂S in the brain alleviates poststroke cardiac dysfunction. (a) Schematic diagram of the experimental procedure. (b) Representative brain laser speckle perfusion images of the sham operation group and after MCAO surgery. (c) Comparison of cerebral blood flow values (n = 7 for sham and n = 14 for MCAO). (d) Representative m-mode echocardiogram images of mice 14 d after surgery from the sham operation group, vehicle-treated group (PBS) after MCAO surgery, and DATS/MION-LPE–treated group after MCAO surgery. (e-g) The measurements of LVED, EF, and FS from echocardiography (n = 3). (h) Quantification of HW to tibia length ratio in each group 30 d after surgery (n = 6). (i) Kaplan–Meier survival rates of each group after 30 d after surgery (n = 8)

HDAC6/SIRT1-NLRP3 signaling plays a central role in the mechanism of neuroinflammation

Figure 6a shows the experimental protocols used to investigate the effecting mechanisms of DATS/MION-LPE on post-stroke-neuroinflammation. Compared with the vehicle group, DATS/MION-LPE can significantly limit the increase of cardiac LVED post-stroke and preserve the EF and FS values (Fig. 6b-e). Compared with the sham operation group, the NLRP3 expression was significantly increased after stroke, and H₂S release from DATS/MION-LPE was shown to effectively limit the NLRP3 activation at 14 d after MCAO (Fig. 6f, g). Meanwhile, the upregulated HDAC6 and decreased SIRT1 expression were detected in the ischemic brain, while DATS/MION-LPE can suppress HDAC6 and enhance SIRT1 expression at 14 d after MCAO (Fig. 6f, h, i). Moreover, DATS/MION-LPE apparently reduced the elevated levels of cerebral Caspase-1 and IL-1β 14d after stroke (Fig. 6f, j, k). These results suggest that DATS/MION-LPE can reduce post-stroke-neuroinflammation by modulating the crosstalk of HDAC6, SIRT1, and NLRP3. The elevated systemic levels of the inflammatory mediators ATP and IL-1β after stroke were also subsequently suppressed by the inhibition of central neuroinflammation (Fig. 6l, m).

Fig. 6.

H₂S targeted to the brain can alleviate poststroke cerebral inflammation by regulating the HDAC6, SIRT1, and NLRP3. (a) Schematic diagram of the experimental procedure. (b) Representative m-mode echocardiogram images of different groups at 14 d after surgery, along with measurements of LVED (c), EF (d), and FS (e) from echocardiography (n = 3). (f-k) Evaluation of the expression levels of NLRP3, HDAC6, SIRT1, procaspase-1, and IL-1β in the brain tissues 14 d after surgery via western blot analysis, along with the densitometric values of the bands normalized to β-tubulin signals (n = 3). Measurement of the ATP (l) and IL-1β (m) levels in the plasma of different groups 24 h after surgery (n = 3)

The precise relationships between NLRP3, SIRT1, and HDAC6 in central neuroinflammation were also investigated. At 7 d after MCAO, SIRT1 overexpression can significantly limit the expression of NLRP3 and HDAC6, as well as cerebral Caspase-1 and IL-1β (Fig. 7a-f), which was related to alleviated cardiac dysfunction, and decreased plasma ATP and IL-1β levels (Fig. S5a-f). In contrast, HDAC6 overexpression can inhibit the SIRT1 expression, with elevated levels of cerebral NLRP3, Caspase-1, and IL-1β (Fig. 7g-l), with exacerbated cardiac damage and increased plasma ATP and IL-1β levels (Fig. S5g-l). Furthermore, in the in vitro experiments which co-cultured cardiomyocytes with conditioned media from primary neuronal cells, the overexpression of SIRT1 or HDAC6 revealed mutual regulation effects, and both of SIRT1 and HDAC6 can directly regulate co-cultured cardiomyocyte viability and LDH activity (Figure. S6). These findings suggest an antagonistic interaction between SIRT1 and HDAC6, which regulates the NLRP3 production and associated inflammatory mediators, emphasizing the role of the HDAC6/SIRT1-NLRP3 circuit in the central neuroinflammation and cardiac function injury after stroke.

Fig. 7.

Upregulation of SIRT1 and HDAC6 in the brain to validate the HDAC6/SIRT1-NLRP3 pathway. (a-f) Representative images and quantitative western blot analysis revealed that the overexpression of SIRT1 in the brain reduced the expression levels of NLRP3, HDAC6, procaspase-1, and IL-1β in the brain of mice after MCAO (n = 3). (g-l) Representative images and quantitative western blot analysis revealed that the overexpression of brain HDAC6 decreased the expression of SIRT1 and increased the expression levels of NLRP3, HDAC6, procaspase-1, and IL-1β in the brain of mice after MCAO (n = 3)

Sympathoexcitation induced by central neuroinflammation exacerbates post-stroke cardiac injury

To further investigate the role of the sympathetic nervous system on the stroke-heart syndrome, pseudoephedrine and capsaicin were separately utilized on the heart’s surface before MCAO surgery to activate or inhibit the sympathetic nervous system effects (Fig. 8a). The results suggested that the MCAO surgery procedure did not affect the integrity of the afferent nerves on the heart’s surface compared with the sham operation group (Fig. S7). The immunohistochemical staining confirmed that capsaicin effectively ablated the cardiac primary sensory nerve fibers stained with the CGRP antibody (Fig. 8b). Compared with the vehicle-treated group, DATS/MION-LPE can significantly limit the increase of cardiac LVED post-stroke and preserve the EF and FS values. This protective effect was reduced by pseudoephedrine application and enhanced by capsaicin utilization (Fig. 8c-f). The results of the cardiac oxidative stress assays also revealed that the cardiac SOD, CAT, and MDA levels were reduced by DATS/MION-LPE and the protective effects of which were impaired by pseudoephedrine and elevated by capsaicin utilization (Fig. 8g-i). Accordingly, the cardiac Caspase-1 and IL-1β levels also showed similar results (Fig. 8j-l).

Fig. 8.

H₂S in the brain can alleviate poststroke cardiac injury influenced by the sympathetic nerve. (a) Schematic diagram of the experimental procedure. (b) Immunohistochemical staining for NF, catecholaminergic nerve fibers, and CGRP in the heart tissue sections of mice treated with either olive oil or capsaicin, where the catecholaminergic nerve fibers were stained with TH antibody. (c) Representative m-mode echocardiogram images of different groups 14 d after surgery, along with measurements of LVED (d), EF (e), and FS (f) from echocardiography (n = 3). (g-i) Expression levels of SOD, CAT, and MDA in cardiac tissues 14 d after surgery (n = 3). (j-l) Assessment via western blot analysis of the expression levels of procaspase-1 and IL-1β in the cardiac tissues 14 d after surgery, along with the densitometric values of the bands normalized to β-actin signals (n = 3)

Discussion

Post-stroke cardiovascular events, recently recognized as the stroke-heart syndrome, can present various manifestations including acute myocardial infarction, nonischemic myocardial injury, left ventricular dysfunction, arrhythmias, and heart failure [29–31], which can be detected as early as three days after stroke [32, 33]. The present study has shown that the mice exhibited significant cardiac contractile dysfunction, heart atrophy, and adverse hemodynamic changes after stroke, with impaired overall survival. Recently, neuroinflammation has been recognized as the central pathological mechanism of stroke, which can induce microglia to produce abundant inflammatory factors and recruit immune cells [34, 35]. Moreover, inflammation-induced mitochondrial stress can further activate astrocytes to produce proinflammatory cytokines, enhancing the inflammatory microenvironment in the central nervous system [36]. These inflammatory factors can increase the permeability of the BBB, which allows peripheral immune cells to penetrate the damaged BBB and infiltrate into ischemic tissues, further exacerbating inflammation in the central nervous system [37]. Ultimately, central neuroinflammation can trigger a systemic inflammatory response, and inflammatory factors in circulation such as IL-1β can induce severe cardiac damage [38]. Dying neurons can also release ATP as a proinflammatory damage-associated molecular pattern into circulation to impair certain organs [39, 40]. Our results revealed increased systemic IL-1β and ATP levels after stroke [38]. Accordingly, Caspase-1 and IL-1β were significantly increased in the hearts after stroke, accompanied by oxidative stress-induced myocardial damage.

Recently, the NLRP3 inflammasome is critical in the mechanisms of neuroinflammation [41, 42], and its activation can induce the IL-1β release in a Caspase-1-dependent manner [43]. After stroke, the increased mitochondrial permeability in the neurons can promote the assembly of NLRP3 [44], and increased mitochondrial stress can lead to the release of mitochondrial reactive oxygen species (mtROS) facilitating NLRP3 deubiquitination [45]. Several studies have also reported that NLRP3 activation is closely related to neurological deficits and neuronal death during stroke [46]. HDAC6 is another important factor in cerebral inflammatory damage and is associated with the dysfunction of regulatory T cells in the ischemic brain [47, 48]. Inhibitors targeting HDAC6 can also exhibit neuroprotective effects after stroke [49, 50]. Furthermore, HDAC6 can also increase ROS production by deacetylating redox regulatory proteins, such as peroxiredoxin I (Prx I) and peroxiredoxin II (Prx II), which may promote NLRP3 activation [51]. As a NAD⁺-dependent protein deacetylase, SIRT1 exhibits effective protective effects in cerebral ischemic injury [52, 53]. SIRT1 can downregulate the NF-κB expression by deacetylating p65 and can upregulate the nuclear factor erythroid 2-related factor 2 (Nrf2), which can induce definite antioxidant effects [54, 55]. Interestingly, a previous study by Hernández-Jiménez et al. presented upregulated SIRT1 at 48 h after MCAO, exerting a protective role during cerebral ischemia as a possible compensatory response to inflammation [56]. However, many researches have confirmed that the prolonged ischemic damage after stroke can lead to persistent neuroinflammation, which may ultimately result in the downregulation of SIRT1 [57–59]. Our results also revealed that the expressions of NLRP3 and HDAC6 were significantly upregulated in the ischemic brain 14 d after MCAO, while the expression of SIRT1 was significantly downregulated, which may be responding to the long-term pathophysiological changes in ischemic brains.

To date, the regulatory interactions among NLRP3, HDAC6, and SIRT1 in neuroinflammation have not been investigated. Our results revealed the upregulated HDAC6 and downregulated SIRT1 levels in the ischemic brain, along with the elevated expressions of NLRP3. Using DATS/MION-LPE to inhibit the central neuroinflammation, the SIRT1 was significantly upregulated, while the HDAC6 and NLRP3 were downregulated. Previous studies have indicated that HDAC6 can deacetylate various nonhistone proteins and regulate the formation and degradation of autophagosomes, thereby modulating the initiation and activation of NLRP3 [60]. Moreover, HDAC6 can also enhance the NF-κB activities by deacetylating microtubules, thereby promoting the transcription of NLRP3 [61, 62]. On the contrary, the SIRT1 could inhibit the transcriptional activity of NLRP3 by deacetylating the p65 subunit of NF-κB and reduce ROS production by activating the Nrf2 antioxidant signaling pathway, further suppressing NLRP3 activation [63, 64]. Previous studies have also shown that HDAC6 and SIRT1 can jointly regulate the deacetylation of K-renin-angiotensin system K-(RAS) and β-catenin, both of which play a synergistic role in regulating the function of p62 protein and Hsp90 protein [65–67]. While in inflammatory responses, HDAC6 and SIRT1 can jointly regulate the activity of NF-κB [68]. Therefore, we investigated the interactions among HDAC6, SIRT1, and NLRP3 by independently upregulating HDAC6 or SIRT1 expressions in ischemic brain tissues, indicating that HDAC6 and SIRT1 can antagonize each other and jointly regulate the NLRP3 activity in the post-stroke brain. The present study investigated the existence of the HDAC6/SIRT1-NLRP3 circuit in regulating neuroinflammation after stroke, suggesting that targeting the HDAC6/SIRT1-NLRP3 circuit can effectively alleviate the stroke-heart syndrome.

The overexcitation of the sympathetic nerve is another potential mechanism in secondary heart injury after stroke [69]. Our study demonstrated that both stimulating sympathetic nerves and ablating cardiac surface sympathetic receptors can influence the severity of secondary cardiac damage. Evidence has confirmed that central neuroinflammation can activate the RAS, leading to increased sympathetic nervous activities through the stimulation of the primary effector peptide angiotensin II (Ang II) and angiotensin type 1 receptor (AT1R) [70–72]. The activation of Ang II can increase the IL-1β expressions in brainstem astrocytes and activate microglia, which induces ROS production that further disrupts the BBB [73, 74], exacerbating neuroinflammation [75]. Previous studies have reported that sympathetic receptor blockers can alleviate common cardiac dysfunction [40]; however, the specific impact of sympathetic nervous system activities on stroke-heart syndrome has yet to be determined. Our results suggest that the activation of the sympathetic nerve activity can significantly exacerbate cardiac injury and heart dysfunction after stroke and can also increase the myocardial inflammatory cytokine levels and oxidative stress. These adverse effects can be alleviated by ablating the cardiac surface sympathetic receptors. These findings confirm the important role of sympathetic nerve function in the development of stroke-heart syndrome.

Currently, the lack of ideal drugs to treat both brain and secondary heart damage limits the investigation of the underlying mechanisms of stroke-heart syndrome. Existing study methods are used to mainly manage injured hearts by systemic administration, which cannot prove the essential role of neuroinflammation in post-stroke cardiac damage. Therefore, we developed a novel drug delivery system DATS/MION-LPE, which combines the temperature-and ultrasound-sensitive copolymer p-(MEO₂MA-co-THPMA), the brain-targeting ligand LF, and the H₂S donor DATS. DATS/MION-LPE can effectively restrict the drug release in the circulation and the targeted release of H₂S in the ischemic brain, directly managing the neuroinflammation. Thus, we first determined the key role of central neuroinflammation and the underlying HDAC6/SIRT1-NLRP3 circuit in the stroke-heart syndrome. The new delivery system can also act as a therapeutic agent to reduce central proinflammatory cytokine production, alleviate cardiac oxidative stress damage, and preserve heart function. There are still some limitations in this study. First, although we validated the involvement of the HDAC6/SIRT1-NLRP3 circuit in the central neuroinflammation, and demonstrated that both increased systemic inflammatory factors and activated sympathetic nervous system can contribute to the development of post-stroke cardiac injury, the connection between the brain and heart in this syndrome is complex and warrants further investigation. Secondly, our research still limits in exploring how cerebral ischemia impacts on hearts, the potential application of DATS/MION-LPE in other brain injury-related diseases should be explored in further studies. Finally, the ultrasound stimulation conditions could be further optimized. We selected the frequency of 1 MHz based on its penetration depth and safety, but further “medium” frequency range (0.7–3 MHz) should be evaluated as possible therapeutic frequency in the future [76], however, our results presented that the ultrasound spatial and temporal energy density was approximately 0.5 W/cm², which is safe at 1 MHz for a 10 min duration [77].

Conclusion

Overall, this study introduced the key role of neuroinflammation in stroke-heart syndrome and investigated the HDAC6/SIRT1-NLRP3 circuit in this mechanism. Regulating the HDAC6/SIRT1-NLRP3 circuit can effectively reduce neuroinflammation, thereby alleviating subsequent heart damage. We also confirmed that secondary heart injury is mediated by both systemic inflammatory responses and sympathetic nerve overactivation. Furthermore, the novel temperature- and ultrasound-sensitive nanoscale controlled-release drug delivery system DATS/MION-LPE is a unique tool for studying the mechanisms of stroke-heart syndrome, which also can work as an effective therapeutic agent, providing new insights into the application of the nanoscale drug delivery systems in cardiac protection.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

X. Sun, L. Pang and Y. Wang conceived and designed this study. M. Jiang, Z. Zhu, Z. Zhou, Z. Yan, K. Huang, M. Jieensi, R. Jiang and X. Fan contributed to materials construction, data analysis, histological experiments, and the original draft. X. Sun, M. Jiang, and Y. Wang participated in writing- review and editing. All authors reviewed the final manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (82370377, 81601663), Natural Science Foundation of Shanghai (23ZR1408800), and Clinical Research Project Supported by Huashan Hospital Fudan University.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

All animal experiments conducted in this study were approved by the Institutional Animal Care and Use Committee of the Department of Laboratory Animal Science at Fudan University (Grant No. 20160780A176).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.