Biosynthesized nanoparticles of Tibetan medicine mercuric sulfide preparation to promote endocytosis and realize drug crossing through blood brain barrier

Abstract

See More

β-HgS as the main component of Tibetan medicine Zuotai (ZT) is widely used in the treatment of central nervous system diseases. Although the synergism of HgS has been clinically verified for more than 2000 years, its synergetic mechanism is still an unsolved mystery and a huge challenge. Notably, we clearly and intuitively demonstrate that Zuotai or β-HgS is auto-synthesized by organisms into spherical HgS nanoparticles with protein corona, with an overall particle size of 30–195 nm and a core of HgS NPs of 5–7 nm. Further research showed that HgS NPs facilitated the transport of Oxiracetam (ORT), Memantine Hydrochloride (MH) and Entinostat (MS-275) across the blood-brain barrier (BBB), especially the brain accumulation of MS-275 increased by more than three times. Meanwhile, HgS NPs enhanced the effect of MS-275 on spatial learning and memory ability of APP/PS1 mice. Further studies confirmed that HgS NPs increased the permeability of the blood-brain barrier (BBB) by regulating endocytosis (Upregulation of caveolin, clathrin, and dynamin, downregulation P-gp) thereby facilitate the transport of drugs across BBB with more than threefold higher brain accumulation and long-lasting efficiency enhancement (4 days). Our findings reveal that these in vivo-generated HgS NPs provide an efficient, convenient, and biocompatible drug delivery platform for crossing the BBB. Importantly, our findings offer new hopes and ideas for in-depth research on highly valuable and mysterious metal preparations in classic Chinese and Tibetan medicine.

Graphical Abstract

Introduction

The development of drugs to treat central nervous system (CNS) diseases, such as Alzheimer’s disease, Parkinson’s disease, stroke, and glioblastoma, is exceedingly complex and challenging due to the presence of the BBB [1,2,3,4]. Arguably the most tightly regulated interface between the CNS and circulation, the BBB prevents foreign harmful substances from infiltrating the brain, playing a crucial role in maintaining brain homeostasis [5, 6]. However, it also creates a formidable challenge for the passage of most therapeutic drugs [7,8,9,10]. With the advent of nanotechnology, various metal nanoparticles have increasingly become utilized as drug carriers, aiding in the traversal of the BBB for the treatment of CNS diseases [11,12,13]. For example, chiral Au Nanoparticles (NPs) stabilized by glutathione, exhibiting no noticeable toxicity, have shown promising therapeutic effects against Alzheimer’s disease (AD) by inhibiting amyloid-β aggregation and enhancing BBB permeability [14]. Additionally, several metal-based NPs (such as Ag and ZnO NPs) have demonstrated the potential to cross the BBB either via the paracellular space by modulating BBB permeability or through transcytosis-mediated routes [15]. Nonetheless, there remains a high desirability for metal NP drug carriers with simpler and easily generated processes and high biocompatibility.

Metals in traditional Chinese medicine, used clinically for over 2000 years, predominantly exist in sulfide forms. For example, α-HgS and As₄S₄ are the main components of traditional Chinese medicines cinnabar and realgar, respectively, and β-HgS is the major component of Tibetan medicine metacinnabar [16]. Modern toxicological studies have also confirmed that the chemical forms of mercury and arsenic are major determinants of their toxicity [17,18,19]. Our institute and several research institutions have carried out long-term and comprehensive research on the toxicological concerns of ZT such as acute toxicity, long-term toxicity and neurotoxicity, as well as its metabolic characteristics in vivo, and confirmed that ZT has no obvious toxicity at clinical doses [19,20,21,22,23,24]. As a core adjuvant in Tibetan medicine, metacinnabar (ZuoTai/ZT, with β-HgS as the main component) is widely used in many well-known Tibetan prescriptions and clinical preparations, playing a unique synergistic role in the treatment of stroke, paralysis, hypertension, and nervous system diseases [25,26,27]. However, the precise mechanisms underlying the therapeutic actions of heavy metal preparations in traditional Chinese medicine is still a challenge to overcome. The synthesis of metal nanoparticles (NPs) by living organisms represents an emerging field in bio-nanotechnology, offering high biocompatibility and paving new paths for drug delivery, imaging, and diagnostics. For instance, the luminescent cadmium telluride (CdTe) NPs generated by earthworms has been successfully applied in live-cell imaging [28]. Biologically formed magnetite NPs in the human brain were detected over 30 years ago and exhibit a unique combination of redox activity, surface charge, and strong magnetic behavior [29]. Pt NPs generated biologically in human blood when patients were treated with cisplatin demonstrate good biocompatibility, accumulation in tumor regions, prolonged anticancer effects through controlled drug release, and reversal of tumor cell drug resistance [30, 31]. These studies inspire the exploration of whether metals in traditional Chinese medicine can also spontaneously form nanoparticles and play a significant role in disease treatment.

In this study, we report the detection of biologically auto-generated HgS NPs with a core diameter of 5–7 nm in the blood and tissue immediately after oral administration of ZT or its main component β-HgS in mice and humans. Unexpectedly, the HgS NPs and protein coronas on their surface, with an overall diameter of 30–195 nm, were successfully isolated and visualized using high-performance liquid chromatography (HPLC) and scanning electron microscopy (SEM). Importantly, we found that biologically formed HgS NPs can significantly enhance the distribution and retention time of drugs in the mouse brain and improve the therapeutic efficacy of drugs on learning and memory function in APP/PS1 transgenic mice. Mechanistic research further confirmed that HgS NPs could achieve drug delivery into the brain by upregulating the expression of endocytosis proteins and reducing the expression of efflux proteins in the BBB, rather than damaging the integrity of the BBB, as they had no impact on tight junctions and adhesive junctions. HgS NPs auto-synthesized in vivo by administering clinically used ZT and its main component β-HgS can be directly applied for drug delivery across the BBB without any modification.

Materials and methods

Materials

Chromatographic grade methanol, acetonitrile, and dimethyl sulfoxide (DMSO) were procured from Tianjin Baishi Company (China). ZT was obtained from Qinghai Jiumei Tibetan Medicine Co., Ltd (China). Mercurysulfide (β-HgS) was sourced from Pansine Chemical Co., Ltd. Oxiracetam (ORT) (#C10079332), Memantine Hydrochloride (MH) (#621206), Entinostat (MS-275) (#29050, 99.31%) and Genistein (GEN) (DR0002-0020) were acquired from MedChemExpress (MCE, USA). Chlorpromazine Hydrochloride (CPZ) (abs817925-50) was purchased from Shanghai Aibixin Biootechnoligy CO., Ltd. Evans blue dye (E10428) was purchased from Aladdin (Shanghai, China). The MTT kit, annexin V-FITC/PI cell apoptosis detection kit, and BCA protein quantification kit were obtained from Shanghai Biyuntian Biotechnology Company (China). Glycine tetramethylethylenediamine (TEMED), RIPA cell lysate, and protein gel preparation kit were acquired from Beijing Solebao Technology Co., Ltd. (China). Primary antibodies used in western blot experiments, including occludin (ab167161, 1: 2000), ZO-1 (ab96587, 1: 2000), VE cadherin (ab205336, 1: 2000), P-gp (1: 1000), and β-actin (ab8226, 1: 1000) were purchased from Abcam (Cambridge, UK). Clathrin (#4796, 1: 2000), caveolin-1 (#3267, 1: 2000), and dynamin I/II (#2342, 1: 2000) antibodies were procured from Cell Signaling Technology (Beverley, USA). The horseradish peroxidase-labeled secondary antibody (BA1054, 1: 2000) was purchased from Doctoral Biotechnology Co., Ltd. (Wuhan, China). Primary antibodies used in immune system assays, including ZO-1 (ab96587, 1:80), were purchased from Abcam (Cambridge, UK), clathrin (#4796, 1: 80) and caveolin-1 (#3267, 1: 400) were purchased from Cell Signaling Technology (Beverley, USA), dynamin II (BM5290, 1: 80), and ABCB1 (BM4508, 1: 300) were purchased from Doctoral Biotechnology Co., Ltd. (Wuhan, China). Occludin (bs-10011R, 1: 200), dynamin (1: 100), VE cadherin (bs-0878R, 1: 200), and P-gp (1: 300) were acquired from Boorsen Biotechnology Co., Ltd. (Beijing, China). The secondary antibody goat anti-rabbit LgG Fc AF-594 (ab150092, 1: 400) was obtained from Abcam (Cambridge, UK).

Instruments

The HgS NPs were extracted and separated using an ultrahigh-speed centrifuge (Optima X, Beckman, Germany) and a Hanbang semipreparative liquid chromatograph (5100 HPLC, Jiangsu Hanbang Technology Co., Ltd., China), respectively. Lattice stripe images were acquired with a TEM (JEM-F200, JEOL, Japan) and further analyzed using Gatan digital micrograph software. Energy-dispersive X-ray spectroscopy (EDS) was recorded on a D8-X-ray diffractometer (Bruker, Germany). The overall appearances of HgS NPs with protein coronas were detected using SEM (Ultra Plus, Carl Zeiss AG, Germany). LC-MS/MS analysis for identifying the biologically relevant corona components was performed on a timsTOF Pro mass spectrometer (Bruker, Bremen, Germany) coupled to a nanoElute nanoflow chromatographic system (Bruker Daltonics, Bremen, Germany). Immunofluorescence experiments were conducted on a confocal microscope (STELLARIS 5 SR, Leica, Germany) equipped with a cell microscope imaging system (DFC7000T, Leica, Germany). The real-time distribution of drug in living cells was observed by High content imaging analysis system (operetta CLS, Revvity (Shanghai) Co., Ltd.) The dynamic distribution of the drug in living tissue was imaged using a small animal living optical imaging system (IVIS Lumina LT, PerkinElmer, USA) and a Micro-CT small animal living scanning imaging system (InSyTe FLECT/CT, Trifoil Imaging, USA). Hg elements in different organs of mice were determined by an atomic fluorescence spectrometer (AFS-933, Beijing Jitian Instrument Co., Ltd, China).

Cell lines and cell culture

The primary human brain microvascular endothelial (HBMEC) cell lines were obtained from the American Typical Culture Collection (ATCC, Manassas, VA, USA). The cells were cultured in endothelial cell medium (Sciencell, USA) supplemented with 10% fetal bovine serum (Zhejiang Tianhang Biotechnology Co., Ltd., China), 100 U/mL penicillin (HyClone, USA), and 100 µg/mL streptomycin (HyClone, USA) at 37 °C in a humid atmosphere containing 5% CO₂. The Materials and Methods section should provide sufficient information to allow replication of the results. Begin with a section titled Experimental Design describing the objectives and design of the study as well as prespecified components.

Animals

C57BL/6J, BALB/C, APP/PS1 mice and SD rats were purchased from Beijing Huafukang Biotechnology Co., Ltd (China). The mice were housed in a clean-level environment with controlled temperature and humidity, a 12 h light and dark cycle, and ad libitum access to food and water. All animal experiments were approved by the Experimental Animal Ethics Committee of the Northwest Plateau Institute of the Chinese Academy of Sciences (CAS) and conducted following the Guidelines for Care and Use of Laboratory Animals of CAS.

In vivo biosynthesis and purification of HgS NPs

C57BL/6 mice were orally administered with ZT (3 mg/kg)β-HgS (1.5 mg/kg), after which they were anesthetized and perfused with PBS. Blood, liver, brain, and kidney samples were collected at various time points (30 min, 1 h, 4 h, 12 h, 24 h, and 48 h) post-administration. Tissues were minced and homogenized, followed by centrifugation at 10³ g to remove insoluble material. The resulting supernatants were then subjected to further centrifugation at varying speeds (from 10⁴ g for 40 min to 10⁵ g for 180 min) to isolate extracts containing different-sized HgS NPs. These extracts were suspended in deionized water for TEM analysis. Additionally, HgS NP extracts underwent purification via HPLC using a C18 column (5 μm, 300 Å, 4.6 × 250 mm) with a mobile phase composed of methanol and water (15:85, v/v) at a flow rate of 0.1 mL/min. The relatively pure HgS NPs collected were then analyzed using EDS, UV-VIS. The nanoparticles solution purified by HPLC was freeze-dried, and an appropriate amount of it was dissolved in ultrapure water. It was dripped on the silicon wafer and dried naturally at room temperature for SEM detection. The images were captured using Zeiss Ultra Plus, and the parameters were set to mag = 30–100 KX, ETH = 2.0KV, WD = 5.8 mm. The content of HgS NPs in various tissues was determined using an atomic fluorescence spectrometer.

HgS NPs auto-generated in human

Human blood samples were collected from four patients (two men and two women) at JiuMei Tibetan Hospital of Qinghai Province, 24 h after ingesting a prescription containing ZT. HgS NPs were isolated and purified using centrifugation methods similar to those described for isolating HgS NPs in mice. The samples were then analyzed using a 120 kV biological TEM. All human experiments were conducted with approval from the Ethics Committee of the Northwest Plateau Institute of Biology, Chinese Academy of Sciences.

Auto-synthesis of HgS NPs from ZT/β-HgS in vitro

ZT/β-HgS at a concentration of 50 ng/mL was introduced into the cell culture medium containing HMBECs. After incubation for durations of 6, 12, 24, 48, and 72 h, the cells and the culture medium were centrifuged at 10³ g for 5 min to separate them. The collected cells were lysed in lysis buffer for 5 min, and then dissolved in 5 mL of deionized water. Subsequently, the cell extracts and culture medium were subjected to separation and purification procedures as described previously for isolating HgS NPs from mice.

Protein Corona separation and identification

SDT buffer (4% SDS, 100 mM Tris-HCl, 1 mM DTT, pH 7.6) was utilized for lysing samples and extracting proteins. The protein content was determined using the BCA Protein Assay Kit (Bio-Rad, USA). Nanoparticles samples were obtained from mice and purified by gradient centrifugation combined with HPLC. All protein samples were individually aliquoted and stored at − 80 °C. For each sample, 20 µg of protein was combined with 5X loading buffer and boiled for 5 min. The proteins were separated on a 12.5% SDS-PAGE gel (constant current 14 mA, 90 min). Protein bands were visualized by Coomassie Blue R-250 staining. Protein digestion was performed using trypsin. The resulting peptide digests from each sample were desalted on C18 Cartridges (Empore™ SPE Cartridges C18, bed I.D. 7 mm, volume 3 mL, Sigma, USA), concentrated via vacuum centrifugation, and reconstituted in 40 µL of 0.1% (v/v) formic acid prior to LC-MS/MS analysis.

The peptides were analyzed using a timsTOF Pro mass spectrometer (Bruker) coupled to Nanoelute (Bruker Daltonics). Samples in buffer A (0.1% formic acid) were loaded onto a C18-reversed phase analytical column (75 μm × 25 cm, 1.9 μm, homemade) and separated with a linear gradient of buffer B (99.9% acetonitrile and 0.1% formic acid) at a flow rate of 300 nL/min. The mass spectrometer was operated in positive ion mode, collecting ion mobility MS spectra over a mass range of m/z 100–1700 and 1/k0 of 0.75 to 1.35. Subsequently, 10 cycles of PASEF MS/MS were performed with a target intensity of 1.5 k and a threshold of 2500. Active exclusion was enabled with a release time of 0.4 min. MS raw data for each sample were combined and searched using MaxQuant software for identification and quantitation analysis.

Quantitative detection of brain drugs

BALB/c mice were divided into nine groups, each consisting of 5 mice: MS-275 (20 mg/kg), ZT (3 mg/kg) + MS-275 (20 mg/kg), and β-HgS (1.5 mg/kg) + MS-275 (20 mg/kg). ORT (10 mg/kg), ZT (3 mg/kg) + ORT (10 mg/kg), and β-HgS (1.5 mg/kg) + ORT (10 mg/kg). MH (10 mg/kg), ZT (3 mg/kg) + MH (10 mg/kg), and β-HgS (1.5 mg/kg) + MH (10 mg/kg) [32,33,34]. Mice in the combined administration groups received oral administration of ZT/β-HgS water suspension after 3 h, by gavage of ORT and MH or intraperitoneal injection of MS-275. After anesthesia, mice in each group were perfused at different time points (1, 2, 3, and 4 h) post-administration, and their brains were harvested and frozen at -20 °C. Brain tissues were homogenized and ultrasonically extracted by adding DMSO with a ratio of 3:1 (V/M) three times for 30 min each at 30 °C. The extracts were then centrifuged at 3000 rpm for 5 min at 4 °C, and the supernatants were collected and concentrated to dryness. The dried samples were dissolved by ultrasonication in 2 mL of methanol for subsequent LC-MS analysis.

Drug concentration was determined using the Triple Quad 5500 LC-MS/MS system (AB SCIEX, USA). Chromatographic separation was performed on a RP-C18 column (ZORBAX SB-C18, 3.5 μm, 2.1 × 100 mm) with a mobile phase consisting of methanol and water (40:60, v/v) at a flow rate of 0.30 mL/min and a column temperature of 40 °C. The mass spectrometry conditions were as follows: electrospray ionization source, spray voltage 5500 V, ionization temperature 550 °C, curtain gas 30 psi, collision gas 9 psi, nebulizer gas 55 psi, auxiliary gas 55 psi, ion declustering voltage 75 V, entrance potential 10 V, collision energy 27 V, and collision cell exit potential 13 V. The monitored ion pairs for MS-275 were m/z 377.4/359.2 and 377.4/269.2.

NIR images of brain fluorescent drugs

Mice in each group (n = 3) received intravenous injections of CY5.5-conjugated MS-275 (20.0 mg/kg) after 4 h of oral administration of ZT (3.0 mg/kg)/β-HgS (1.5 mg/kg). Subsequently, mice were anesthetized at 1, 2, 3, and 4 h post-injection, and the fluorescence signal intensity in the brain was measured using an in vivo imaging system (IVIS Spectrum, Lumina LT, PerkinElmer, USA) with excitation at 640 nm and emission at 740 nm. 3D images were obtained through spectrum surface topology reconstruction and spectrum 3D signal reconstruction. Additionally, mice were anesthetized at 1 and 2 h post-injection of the fluorescent drug via tail vein, and brain tissues from each group were collected following cardiac perfusion. NIR images were captured and analyzed using IVIS Spectrum. The administration doses of SD rats were CY5.5-conjugated MS-275 (13.0 mg/kg), ZT (2.0 mg/kg), β-HgS (1.0 mg/kg), and the methods of administration and NIR detection were the same as those of mice. All fluorescence intensity statistical analyses in the experiment were performed using the instrument specific software Living image.

CT images of brain fluorescent drugs

The administration method for mice was consistent with that used in NIR imaging. Two h post-administration, fluorescence imaging was conducted using the Insyte FLECT/CT system (TriFoil Imaging, Chatsworth, CA, USA), with the animals anesthetized using isoflurane throughout the imaging procedure. CT image acquisition parameters were configured to a 35 kV lamp voltage and 950 µA current, with an exposure time of 180 ms and a total scanning duration of 15 min. For FLCT image acquisition, parameters were set to a 730 nm LED laser and 813 nm filter, with an exposure time of 17 ms and a total scanning duration of 40 min. Subsequently, images were reconstructed using TriFoil imaging software (Chatsworth, USA), and the reconstructed 3D CT and FLCT images were merged using Vivo Quant V3.0 software (invicro, USA).

Zebrafish in vivo imaging

AB line zebrafish at 30 dpf were selected for the experiment in dishes (5 tails/dish), with 3 replicates for each treatment. Zebrafish in ZT + MS-275-FITC group and β-HgS + MS-275-FITC group were pretreated with ZT (30 µg/mL) and β-HgS (15 µg/mL) for 48 h. Then each group was exposed to water containing 0.25 µg/mL MS-275- FITC. The fluorescence intensity of the drug in the brain was observed using stereomicroscope and laser confocal microscope at 12, 24, 48, and 72 h after the zebrafish were anesthetized.

APP/PS1 mice behavioral tests

Male APP/PS1 mice were randomLy assigned to one of six groups, each comprising 8 mice: Control (wild type, C57BL/6), MS-275 (20 mg/kg), ZT (3 mg/kg), ZT (3 mg/kg) + MS-275 (20 mg/kg), β-HgS (1.5 mg/kg), and β-HgS (1.5 mg/kg) + MS-275 (20 mg/kg). The administration regimen involved oral gavage of ZT/β-HgS and intraperitoneal injection of MS-275 once every 2 days. After 7 days of treatment, the mice underwent a battery of behavioral tests including open field, Morris water maze (MWM), and Y-maze assessments.

Open field test

The spontaneous exploration ability of mice was assessed using the open field test. Mice were individually placed in a 90 cm × 90 cm open field box, ensuring a low-light environment and ambient noise levels below 50 dB. Their exploratory behavior within the central region of the open field was recorded for 5 min using Ethovision 14 analysis software (Noldus Information Technology, Wageningen, Netherlands).

MWM test

The spatial learning and memory abilities of the mice were evaluated using the Morris water maze (MWM) test. The test was conducted in a water tank measuring 120 cm in diameter and 80 cm in height, filled with water maintained at 22–24 °C to two-thirds of its height. A circular platform with an 8 cm diameter was fixed in one quadrant, submerged 2 cm below the water surface. All trials were conducted by researchers blinded to the experimental groups, with the mice’s performance recorded using a video camera system and analyzed using Ethovision 14 analysis software. The MWM test comprised two stages: acquisition training and probe trial. During the acquisition training phase, conducted over five consecutive days with four trials per day, mice were trained to locate the hidden platform. The trial ended when a mouse found the platform within 60 s and remained on it for 15 s. The time taken from entering the water to finding the platform (escape latency) was recorded. If a mouse failed to find the platform within 60 s, it was guided to the platform and allowed to stay for 15 s, with an escape latency of 60 s recorded. On the sixth day, the platform was removed, and the spatial memory of the mice was assessed in a probe trial. Each trained mouse was placed in the water at the opposite quadrant of the original platform location and allowed to swim for 60 s. The number of platform crossings, time spent in the target quadrant, and swimming distance was recorded to evaluate spatial memory abilities of the mice.

Y-maze test

The Y-maze test, which includes the free alternation experiment and exploration of a novelty arm were used to test the mice’s willingness to explore new environments and their short-term working memory abilities. The Y-maze apparatus consisted of three white background cuboids measuring 30 cm × 10 cm × 8 cm (L × W × H), with a 120° angle between two arms. In the free alternation experiment, mice were positioned at the center of the Y-maze and permitted to shuttle through the three arms. The sequence of mice entering and exiting each arm within 5 min was recorded and analyzed using Ethovision 14 analysis software. A correct selection was defined as entering three different arms consecutively. The correct selection times (n) of each mouse and the total arm entry times (N) were recorded to calculate the correct rate for each mouse. For the novelty arm exploration experiment, one arm of the Y-maze was enclosed as a novelty arm. During the training phase, mice were allowed to freely explore the open arms for 5 min, before being returned to their cage for a 1-h interval. In the testing phase, the closed novelty arm was opened, and mice were given 5 min to freely explore. The exploration time of each mouse in the novelty arm was recorded.

Intracellular drug uptake in HBMECs

HBMECs were cultured in 24-well plates with collagen-coated round coverslips at a cell density of 5 × 10⁴/mL. Seven different cell administration groups were established: MS-275-FITC (2 ng/mL) alone incubation, ZT (50, 100, 300 ng/mL) combined with MS-275, and β-HgS (25, 50, 150 ng/mL) combined with MS-275 coincubation. Cells in each group were treated with the corresponding drugs for 6, 12, 24, 48, and 72 h, respectively. At the end of the incubation period, cells were rinsed with PBS three times, fixed with 4% PFA for 15 min, and stained with DAPI (1:100). Intracellular drug uptake was visualized and evaluated by laser confocal microscopy at an emission wavelength of 488 nm and the fluorescence intensity was analyzed using Las X life science software. To verify the correlation between drug uptake and HgS NPs, cells in the combined administration groups were preincubated with ZT/β-HgS for 36 and 42 h, followed by the addition of MS-275 to continue incubation for 12 and 6 h, respectively (ensuring that ZT/β-HgS incubates cells for 48 h to synthesize NPs). Intracellular accumulation of MS was then evaluated. The cell processing and image acquisition procedures were consistent with those described above. For real-time imaging of live cells, HBMEC cells were seeded at 1 × 10⁴ cells/well into PerkinElmer CellCarrier Ultra 96 well plates and cultured for 24 h. The medium was discarded and fixed with tissue cell fixative (0201A23) for 15 min. Then the nuclei were stained with 5 µmol/L Hoechst 33,342 live cell DNA dye for 1 h and the cytoplasm was stained with 1 µmol/L Calcein Am for 40 min in the incubator. After washing with PBS, 2 ng/mL MS-275-Cy5 was added to each group, and 300 ng/mL ZT or 150 ng/mL β-HgS was added to the combined administration groups. The cells continued to be cultured online in a high-content imaging analysis system, and real-time imaging was performed at 6, 12, 24 and 48 h. The instrument specific software Harmony was used to acquire images and analyze fluorescence signals. At least 5 non overlapping fields of each sample were randomly selected and analyzed.

Proteomics analysis

HBMECs were cultured in ECM at a cell density of 5 × 10⁴/mL. The cells of the two treatment groups were incubated with 300 ng/mL ZT and 150 ng/mL HgS for 48 h respectively, and the normal cells were used as the control. The treated cells were suspended in 200 µL lysis buffer (4% SDS, 100 mM DTT, 150 mM Tris-HCl pH 8.0) on ice. Cells were disrupted using a homogenizer and boiled for 5 min, further sonicated, and boiled again for 5 min. After the sample was centrifuged at 16,000 rpm for 15 min, the supernatant was taken and quantified using a BCA protein assay kit (Bio-Rad, USA). Subsequently, each 200 µg protein sample was digested according to the FASP procedure described in the literature. In short, the sample was centrifuged repeatedly on an ultrafiltration filter (Microcon units, 30 kD) using 200 µL UA buffer (8 M Urea, 150 mM Tris-HCl pH 8.0) to remove low molecular weight impurities. Then 100 µL of 0.05 M iodoacetamide was added to UA buffer to block the reduced cysteine residues, and the sample was incubated in the dark for 20 min. The filter was washed 3 times with 100 µL UA buffer and then washed 2 times with 100 µL 25 mM NH₄HCO₃. Then protein suspension was digested overnight at 37 °C using 4 µg trypsin (Promega) in 40 µL of 25 mM NH₄HCO₃. The collected filtrate was measured for peptide concentration using a Nanodrop device with OD280. Further, peptides were labeled with TMT reagents according to the manufacturer’s instructions (Thermo Fisher Scientific). 100 µg of peptide equivalent dissolved in 100 µL of 0.05 M TEAB solution at pH 8.5 was mixed with TMT reagent dissolved in 41 µL of anhydrous acetonitrile and incubated at room temperature for 1 h. Then, 8 µL of 5% hydroxylamine was added to the sample and incubated for 15 min to quench the reaction, followed by freeze-drying for testing.

The peptides mixture labeled with TMT was fractionated using a Waters XBridge BEH130 column (C18, 3.5 μm, 2.1 × 150 mm) on Agilent 1290 HPLC at a flow rate of 0.3 mL/min. Buffer A is composed of 10 mM ammonium formate, while buffer B is composed of 10 mM ammonium formate containing 90% acetonitrile. Both buffers are adjusted to pH 10 using ammonium hydroxide. A total of 30 fractions were collected from each sample and concatenated to 15 fractions according equal interval RPLC. Then fractions were dried and analyzed by nano LC-MS/MS. Q exactive mass spectrometer coupled with easy NLC (Thermo Fisher Scientific) was used for LC-MS/MS analysis. Each fraction was separated using a C18 reverse phase column (12 cm long, 75 μm ID, 3 μm), following a linear gradient with a flow rate at 300 nL/min for 90 min. Buffer A is composed of 2% acetonitrile and 1% formic acid, while buffer B is composed of 90% acetonitrile and 1% formic acid. The setting of linear gradient buffer B is as follows: 0–2 min, 2–5%; 2–62 min, 5–20%; 62–80 min, 20–35%; 80–83 min, 35–90%; 83–90 min, 90%. MS data acquisition uses the data dependent top15 method to dynamically choose the most abundant precursor ions from survey scan (m/z 300–1800) for HCD fragmentation. The target value is determined based on predictive Automatic Gain Control (pAGC). Full MS uses an AGC target value of 1e6 and a maximum injection time of 50 ms and MS2 uses a target AGC value of 1e5 and a maximum injection time of 100 ms. Dynamic exclusion duration was 30 s. Survey scans were acquired at a resolution of 70,000 at m/z 200 and resolution for HCD spectra was set to 35,000 at m/z 200. Normalized collision energy was 30. The instrument was run with peptide recognition mode enabled.

LC-MS/MS raw data was imported into Proteome Discoverer 2.4 software (version 1.6.0.16), according to Uniprot Database for data interpretation and protein identification. An initial search was set at a precursor mass window of 10 ppm. The search followed an enzymatic cleavage rule of Trypsin/P and allowed maximal two missed cleavage sites and a mass tolerance of 20 ppm for fragment ions. Fixed modifications are set to Carbamidomethyl (C), TMT10 plex (K), TMT10 plex (N-term), and variable modifications are set to Oxidation (M) and Acetyl (Protein N-term). The false discovery rate (FDR) for peptide and protein identification was set to 1% and TMT reporter intensity was used for quantification.

Proteins data were analyzed using Perseus software, Microsoft Excel and R statistical computing software. The significantly differentially expressed proteins were screened by the cut-off value of ratio fold change > 1.20 or < 0.83 and P value < 0.05. Hierarchical clustering was used to group proteins according to protein expression levels. Then protein sequences were annotated from uniprotkb/Swiss prot, Kyoto Encyclopedia of genes and genomes (KEGG) and Gene Ontology (GO). Fisher’s exact test was used with FDR correction for multiple testing in GO and KEGG enrichment analyses.

Immunofluorescence assays

HBMECs were cultured in 24-well plates, each equipped with collagen-coated round coverslips. Following a regimen of incubation with ZT (300 ng/mL)/β-HgS (150 ng/mL) for 24 and 48 h, respectively, cells from each group underwent fixation with 4% PFA for 15 min. Post-fixation, the cells were meticulously blocked using 150 µL of blocking solution at room temperature for 30 min. Subsequently, the cells were incubated overnight at 4 °C with primary antibodies that were appropriately diluted in blocking solution: rabbit anti-ZO-1 (1:1000; Abcam), rabbit anti-Occludin (1:200; Aobosen), rabbit anti-VE-cadherin (1:400; Aobosen), rabbit anti-Cavelin (1:400; Cell Signaling), rabbit anti-Dynamin (1:100; Aobosen), rabbit anti-Clathrin (1:80; Cell Signaling) and rabbit anti-P-gp (1:300; Boster). After a thorough rinse with PBS, cells were further incubated with a secondary antibody conjugated with Alexa 594 (1:400) and DAPI (1:1000) for 1 h at room temperature in the dark. In parallel, BALB/c mice were randomLy apportioned into three groups: Control, ZT (3 mg/kg), and β-HgS (1.5 mg/kg), with 6 mice allocated to each group. Over a span of three consecutive days, the mice received continuous administration of the designated substances. Two hours after the final administration, the mice were anesthetized, and initially perfused with PBS (pH 7.4), followed by 4% PFA. Subsequently, the brains were harvested and fixed with 4% PFA for 24 h. After fixation, the brains underwent dehydration in 30% sucrose for 48 h, followed by embedding in an embedding medium and freezing at -80 °C. Coronal brain Sect. (12 μm slices) were prepared using a low-temperature thermostat to assess the expression of caveolin, dynamin, clathrin, and P-gp in vivo. Following a thorough wash with PBS, the sections were blocked with a 5% FBS-containing blocking solution for 1 h, and then incubated overnight at 4 °C with the primary antibodies. Similarly, the expression of ZO-1, VE-cadherin, and occludin in brain tissue was evaluated using paraffin Sect. (5 μm slices). These sections were prepared and stained as per established protocols with slight modifications. Post-staining, the sections were incubated with secondary antibodies linked with DyLight 594 (1:400), DAPI (1:1000), and isolectin (1:200) at room temperature for 1 h. Finally, images of the cells and brain sections were acquired and analyzed using a laser confocal microscope equipped with a cell microscope imaging system.

Western blotting

Protein extraction from brain tissues or cultured HBMECs commenced with the addition of lysis buffer for 30 min, followed by centrifugation at 12,000 rpm for 10 min to obtain the supernatant for subsequent protein quantification and immunoblotting. A total of 20 µg of protein was subjected to standard sodium dodecyl sulfate polyacrylamide gel electrophoresis (PAGE) at a constant voltage of 80 V. The separated proteins were then transferred onto polyvinylidene fluoride (PVDF) membranes (Merck Millipore, Germany) at a constant current of 200 A. After sealing the PVDF membranes for 1 h with 8% powdered nonfat milk in tris buffered saline with tween-20 (TBST), they were incubated overnight at 4 °C with primary antibodies at different dilution proportions. Following three rinses, the PVDF membranes were subjected to a 1:2000 dilution of goat anti-rabbit secondary antibody at 37 °C for 1 h. Immunoreactive bands were visualized using an enhanced chemiluminescence kit and imaged by the Tanon 5200 fully automatic chemiluminescence imaging analysis system (Beijing, China). Band density was quantified using Image J software.

Statistical analysis

All data are presented as mean ± Standard Error of Mean (SEM). The comparison between multiple groups was conducted using Prism 9.0 software and a one-way ANOVA with Dunnett’s multiple comparisons test. The comparison between the two groups was conducted using an unpaired t-test conducted by two students. The number of repetitions and samples are clearly marked in each legend. P < 0.05 was defined as significant difference between groups.

Results

Structural characterization of HgS NPs synthesized in higher organisms

Tibetan medicine ZT is a blue-black powder made from mercury and sulfur through a tedious process similar to Bhasmas in Ayurvedic medicine. Notably, only sulfide forms of mercury are utilized in oral traditional medicines, and the chemical forms of mercury significantly influence their disposition, efficacy, and toxicity. Recent research has confirmed the good biocompatibility of ZnO NPs and Pt NPs auto-synthesized in animals and humans, demonstrating their potential in cancer-targeted therapy and imaging [35]. Given the propensity of metals or metal compounds to generate metal NPs in physiological environments owing to their physicochemical properties, we hypothesized that β-HgS in ZT might undergo instantaneous biosynthesis into HgS NPs in the blood, facilitating drug delivery and enhancing efficacy.

To test this hypothesis, blood samples from patients administered ZT (0.6 mg/kg) and mice administered ZT (3.0 mg/kg) or β-HgS (1.5 mg/kg) were collected 24 h post-oral ingestion and isolated by differential centrifugation ranging from 10³ to 10⁵ g at 4 °C. Encouragingly, transmission electron microscopy (TEM) revealed NPs with a diameter of 5–7 nm (Fig. 1A). Further confirmation of the presence of these NPs was obtained in the brain, liver, and kidney tissues of mice following cardiac perfusion, indicating their presence not only in the blood but also within parenchymal cells across various tissues (Fig. 1B). Notably, NPs were observed to traverse the BBB and enter the brain, potentially facilitating drug delivery into the brain. Importantly, these NPs were absent in ZT or β-HgS water solutions using the aforementioned isolation method, suggesting their biosynthesis by organisms rather than pre-existence in ZT or β-HgS (Fig. S1A, B). In addition, NPs in mice were purified further using HPLC (Fig. S1C-H), revealing UV spectroscopy absorption peaks at approximately 220 and 265 nm (Fig. 1C). The above results preliminarily confirmed that ZT/β-HgS were synthesized into uniform NPs by higher organism, but the elemental composition of nanoparticles still needed to be further clarified. Energy dispersive X-ray spectroscopy (EDS) confirmed the presence of Hg, Pb, Au, and S elements in NPs generated from ZT and Hg and S elements in nanoparticles generated from β-HgS (Fig. 1D), suggesting the possible synthesis of other metal salts in ZT into nanoparticles. High-resolution TEM analysis further confirmed the particle size distribution of captured nanoparticles ranging between 5 and 7 nm, with clear lattice fringes indicating a crystal plane spacing of 0.279 nm consistent with HgS nanoparticles. Selected area electron diffraction also revealed diffusion rings characteristic of nanoparticle crystals (Fig. 1E, F). These results confirmed that these nanoparticles should be HgS NPs.

As expected, HgS NPs could be rapidly formed and detected in various organs in mice within 30 min after the administration of ZT or β-HgS (Fig. S2). However, the synthesis speed of HgS NPs in vitro is significantly slower. When cells were treated with ZT (50 ng/mL) or β-HgS (25 ng/mL) for 6, 12, 24, 36, and 48 h, only minimal HgS NPs were detected at 24 h, with abundant and stable HgS NPs observed in both cells and culture medium after 36 h (Fig. S3). The lattice structure of HgS NPs biologically formed in vitro was more susceptible to disappearing under laser during TEM detection compared to that formed in vivo. We speculate that the protein coronas formed on the surface of HgS NPs in vivo and in vitro are different, playing a crucial role in the formation and stability of HgS NPs.

Fig. 1

Purification and structural characterization of HgS nanoparticles (NPs) in vivo. (A) TEM images of unpurified NPs in the blood of patients and mice receiving oral administration of ZT/HgS (5–7 nm). (B) HgS nanoparticles in various tissues of mice after cardiac perfusion (Brain, Liver, Kidney). (C) UV absorption spectra of nanoparticles in different tissues purified by HPLC. (D) EDS spectrum of ZT and HgS NPs synthesized in mice. ZT NPs: Hg, S, Au, Ag, Pb; HgS NPs: Hg, S. (E) High-resolution TEM images of lattice fringes of HgS NPs (Interplanar spacing: 0.279 nm). (F) Selected area electron diffraction reveals diffusion rings in nanoparticle crystals

Morphology and composition of HgS NPs protein Coronas

In biological environments, the protein corona can be formed around the surface of nanomaterials instantaneously owing to the high surface free energy of nanomaterials [36, 37]. This biologically formed protein corona, in turn, can influence the biodistribution and therapeutic effects of the nanoparticles. Although the composition of the protein corona has been widely identified using HPLC-MS, its morphology remains a fundamental challenge, especially the morphology of autosynthesized nanoparticles in animals has not been reported so far [38, 39]. Recently, Sheibani and Kokkinopoulou have reported the morphology of the protein corona in vitro as an undefined and loose network of proteins using TEM [38, 39]. By contrast, we clearly detected the overall appearance of HgS NPs with protein coronas using field emission scanning electron microscopy (SEM) after HPLC purification from mice. The protein corona, as commonly supposed, forms a dense and layered shell coating the nanoparticle (Fig. 2, Fig. S4, S5). The HgS NPs with protein corona are spherical and roughly distributed in the range of 30–195 nm, considerably larger than the size of naked HgS NPs (Fig. 2A and B). Importantly, purification of large-sized nanoparticles from brain also challenges the conventional notion that nanoparticles within the size range of 5–100 nm are critical for crossing the BBB [40].

Fig. 2

Overall structure of HgS nanoparticles. (A) SEM images of the HPLC-purified NPs. NP size: 30–195 nm. (B) Schematic diagram of the overall structure

We further employed liquid chromatography mass spectrometry (LC-MS) to qualitatively and quantitatively identify the biologically relevant corona components. Over 200 different plasma corona proteins for HgS NPs were detected and quantified. Previous studies have mostly indicated that the protein corona is composed of only dozens of proteins, even in highly complex biological environments [12, 36, 41]. Tenzer recently reported that the protein coronas rapidly formed (< 0.5 min) on silica and polystyrene NPs in human plasma comprised nearly 300 different proteins [42]. Serum albumin is a major component of serum and has been found to be the main protein on the surface of some metal NPs such as Au NPs [43] and Pt NPs [12]. However, in this study, the main proteins on the surface of HgS NPs included hemoglobin subunit, transcriptional repressor, and superoxide dismutase, and albumin was rare (Table S1). Albumin is not the main protein of HgS NPs protein coronas, which may be related to the purification method and purity. The relatively thorough removal of miscellaneous proteins from tissues may be an important reason why we can clearly observe the complete structure of nanoparticles. The protein coronas formed on the surface of NPs might play essential roles in maintaining the drug delivery and targeting properties of the NPs.

ZT/β-HgS facilitate the penetration of drugs across BBB

The increase of drug concentration in the brain is a compelling evidence for the capability of HgS NPs to facilitate drug penetration across the BBB. Hence, a series of qualitative and quantitative experiments were conducted to elucidate the role of HgS NPs in promoting drug entry into the brain in healthy mice (Fig. 3A). Initially, the contents of ORT, MH and MS-275 in the brain tissue of healthy mice at different time points were quantified using LC-MS. The contents of the three drugs in the brain of the combined administration groups were increased to varying degrees. After oral administration for 4 h and 6 h, the drugs contents in the brain of ORT combined administration groups and MH + ZT group were significantly higher than that of the single administration group. In particular, the combination of ZT/β-HgS with MS-275 resulted in significantly higher drug concentrations in brain tissue from 1 to 4 h compared to the administration of MS-275 alone, with the amount of MS-275 in the brain facilitated by ZT or β-HgS increasing by 1.5 times or 1.4 times, respectively, at 1 h, 2.4 times or 2.2 times, respectively, at 2 h, 4.4 times or 4.2 times, respectively, at 3 h, and 3.6 times or 4.5 times, respectively, at 4 h (Fig. 3A). These results confirm that ZT/β-HgS can promote the entry of drugs with different structures into the brain.

ORT and MH, as commonly used drugs in clinic, can readily pass through the BBB. However, as an inhibitor of HDAC3, a new target for Alzheimer’s disease, MS-275 has a brain entry rate of less than 10% [44]. Therefore, in order to better reflect the promoting effect of ZT/β-HgS on drug into the brain, MS-275 was selected as a control drug for further study [45,46,47]. The mice were orally administered with ZT (3 mg/kg)/β-HgS (1.5 mg/kg), followed by the intravenous injection of MS-275 at a dose of 20 mg/kg labeled with CY5.5 fluorophore after 4 h. Free MS-275 or MS-275 and ZT/β-HgS combination were administered to healthy mice, and the dynamic distribution of the drug in living tissue was imaged using an in vivo imaging system (IVIS) at intervals of 1, 2, 3, and 4 h. As anticipated, mice administered with MS-275 and ZT/β-HgS exhibited significantly higher fluorescence signals than those treated with MS-275 alone at 1–4 h throughout the entire body, including the head, suggesting that HgS NPs facilitate drug absorption in vivo, particularly across the BBB (Fig. 3B). Three-dimensional IVIS imaging and FLECT/CT preliminarily confirmed that although most of the fluorescence signals in the combined administration groups were still concentrated in the periphery of the brain, the signals entering into the brain had different degrees of enhancement compared with the administration of MS-275 alone. (Fig. 3C, D, E). To mitigate interference from blood or scalp in the brain, mice were anesthetized at 1 and 2 h post-administration, and brains were harvested after cardiac perfusion for IVIS imaging (Fig. 3F and G). Notably, combined administration of MS-275 and ZT/β-HgS exhibited significantly higher fluorescence signals in the brain compared to free MS-275 at 1 and 2 h, further confirming the ability of HgS NPs to facilitate drug transport across the BBB. These qualitative results further confirm that HgS NPs significantly promote drug accumulation in brain tissue. Similarly, using in vivo imaging, we also confirmed that ZT/β-HgS can promote drug delivery into the brain in zebrafish and rats (Fig. S6). In clinical Tibetan medicine treatment, ZT is typically administered once every 2–3 days. Therefore, its sustainability for synergistic drug action was evaluated. Mice were administered with ZT/β-HgS by gavage, followed by MS-275 administration after 2, 4, and 8 days, respectively, via intravenous injection. The drug concentration in the mice brain tissue was assessed 1 h post-dosing. As anticipated, the effect of promoting drug accumulation in the brain lasted for 4 days, with no significant effect observed on the 8th day, indicating that ZT/β-HgS exhibits a sustained effect in promoting drug transport across the BBB (Fig. 3H).

Fig. 3

ZT/β-HgS promotes the drug to cross the BBB and enter the brain parenchyma. (A) Quantitative analysis of ORT, MH and MS-275 in the brain at different times after ZT/β-HgS combined with MS-275 administration by LC-MS (n = 5). (B) Bioluminescence images of whole mice treated with ZT or β-HgS combined with MS-275 at different time points show the accumulation of drugs in the brain (n = 3). (C) 3D reconstructed images displayed the distribution of MS-275 in the brain under different administration methods. In the combined administration groups, the fluorescence signals of the drug localized in the brain were significantly enhanced, not just in the muscles or skin. (D) Quantitative analysis of fluorescence signals of the drug in the whole-mouse brains. (E) CT images of the mouse brains at 2 h after ZT/β-HgS combined with MS-275 administration. (F, G) Bioluminescence images and quantitative analysis of brain tissues at 1 and 2 h after combined administration (n = 3). (H) The long-term effectiveness of ZT/β-HgS promotes the drug to cross the BBB. Mice were intravenously injected with MS-275 (20 mg/kg) on days 2, 4, and 8 after oral administration of ZT/β-HgS, and the concentrations of MS-275 in the brains after 2 h were measured by LC-MS. Data were presented as means ± SEM. compared with MS group, * p < 0.05, ** p < 0.01, *** p < 0.001

Synergistic effect of ZT/β-HgS on improving the learning and memory function in APP/PS1 mice

All the aforementioned findings demonstrate that ZT/β-HgS significantly facilitates drug entry into the brain across the BBB. These results prompted us to investigate the in vivo synergistic therapeutic effects on the CNS. The APP/PS1 double-transgenic mice model, characterized by memory and cognitive impairment, serves as an appropriate model for assessing the synergistic therapeutic effects of HgS NPs on the CNS. Although MS-275, a histone deacetylase (HDAC1 and HDAC3) inhibitor targeting a novel therapeutic pathway associated with memory and cognitive impairment, demonstrates some efficacy in improving the learning and memory abilities of APP/PS1 mice, its brain penetration rate is less than 10% [48]. In the experimental setup, model mice were administered with ZT (3.0 mg/kg)/β-HgS (1.5 mg/kg) via gavage and/or MS-275 (20 mg/kg) via intraperitoneal injection for a duration of two weeks. The learning and memory functions of the mice were assessed through three distinct behavioral experiments, including the open field test, Morris water maze (MWM), and Y-maze test.

The open field test was employed to evaluate the exploratory behavior of mice across different administration groups on the 7th day of experimentation. In comparison to wild-type (WT) mice, the APP/PS1 mice exhibited markedly reduced spontaneous exploration behavior, as evidenced by shorter exploration times and paths in the central region (Fig. 4A). Conversely, mice treated with MS-275 alone or in combination with ZT/β-HgS displayed significant increases in exploration time and path length in the central region compared to model mice. Specifically, when compared to the group receiving MS-275 alone, mice administered with MS-275 and ZT demonstrated significantly greater exploration time or distance in the center, whereas no significant changes were observed in the ZT/β-HgS administration group. Although the combined administration of MS-275 and β-HgS also significantly increased the exploration time, there was no significant difference in the exploration distance in the center, which may be related to the running speed of the mice in this group (Fig. 4B, C). These findings suggest that ZT/β-HgS notably enhances the efficacy of the drug on the spontaneous exploration behavior of APP/PS1 mice.

The MWM test was conducted following 8 days of administration to assess the spatial learning and memory abilities of the mice. The spatial navigation task, involving the search for a hidden platform, was conducted from the 9th to the 13th day, followed by a probe trial on the 14th day after the removal of the hidden platform. Compared to WT mice, the APP/PS1 mice exhibited evident deficiencies in spatial learning and memory, as indicated by prolonged latency times to reach the platform, reduced activity, shorter time spent in the target quadrant post-platform removal, fewer platform crossings, and longer random swimming distances (Fig. 4D-H). Remarkably, ZT/β-HgS alone did not exert a noticeable effect on the spatial cognition and memory of APP/PS1 mice. However, following MS-275 treatment, the spatial cognition and memory abilities of the APP/PS1 mice were modestly improved. Specifically, mice treated with MS-275 in combination with ZT/β-HgS exhibited progressively shorter latency times over the course of training days (Fig. 4D), more focused search trajectories and longer durations spent in the target quadrant (Fig. 4E, F), increased platform crossings with an average of 4 times (Fig. 5G), and significantly shorter random swimming distances compared to APP/PS1 mice (Fig. 4H). These findings suggest that ZT/β-HgS significantly enhances the efficacy of the drug in improving the spatial learning and memory abilities of APP/PS1 mice.

Furthermore, Y-maze tests, comprising the free alternation experiment and exploration of the novel arm, were conducted to assess the mice’s propensity to explore new environments and their short-term working memory abilities. Mice treated with MS-275 in combination with ZT/β-HgS exhibited higher alternation rates and spent more time in the novel arm compared to the group treated with MS-275 alone, whereas no significant changes were observed in the ZT/β-HgS administration group (Fig. 4I).

Moreover, the expression levels of HDAC3, the target protein of MS-275, in the brain tissue of mice from each group were assessed via Western blot analysis. As depicted in Fig. 4J, the expression level of HDAC3 in the brains of APP/PS1 mice was significantly higher compared to that in WT mice. Notably, treatment with ZT/β-HgS alone did not induce any discernible change in HDAC3 expression. Conversely, mice administered with either the combined MS-275 and ZT/β-HgS or MS-275 alone exhibited a significant inhibitory effect on HDAC3 expression compared with APP/PS1. Remarkably, the combined administration demonstrated a stronger inhibitory effect on HDAC3 compared to administration of MS-275 alone, indicating that ZT/β-HgS could facilitate drug entry into the brain and enhance the inhibitory effect on the target protein. Overall, the results of several in vivo behavioral experiments collectively confirm that ZT/β-HgS can enhance the efficacy to some extent by promoting drug penetration across the BBB.

Fig. 4

Synergistic effects of ZT/β-HgS on MS-275. (A-C) Open field test. The movement track of mice in the open field (A), the time (B) and distance (C) spent in the center of open field. (D-H), MWM tests spatial learning and memory abilities of mice. Escape latency (D), thermal images of mice swimming trajectories (E), time spent in the target quadrant (F), times of crossing the original location of the platform (G) and swimming distance of mice in the water tank (H). (I) Alternation rate and time spent in the new arm in Y-maze. (J) ZT/β-HgS combined treatment significantly reduced the expression of MS-275 target protein HDAC3 in mouse brain. n = 8 mice per group. Data were presented as means ± SEM. compared with APP/PS1 group, *p < 0.05, **p < 0.01, ***p < 0.001. Compared with MS group, # p < 0.05

ZT/β-HgS facilitate the uptake of drugs in HBMEC cells

Brain microvascular endothelial cells (BMECs) constitute the main components of the BBB, responsible for transporting essential nutrients into the brain parenchyma and removing waste products from the brain. However, the majority of CNS drugs exhibit insignificant brain penetration (1-4%) due to low BBB permeability and/or rapid elimination [49]. Firstly, the cytotoxicity of MS-275, ZT, and β-HgS on human brain microvascular endothelial cells (HBMECs) was assessed, and the optimal dose for in vitro administration was determined (Fig. S7). Subsequently, we evaluated the effect of ZT/β-HgS on promoting drug entry into HBMECs. ZT (50, 100, and 300 ng/mL)/β-HgS (25, 50, and 150 ng/mL) was loaded with MS-275 (2 ng/mL) labeled with FITC fluorophore to image cellular uptake by confocal laser scanning microscopy (CLSM).

As expected, after incubating the cells with ZT/β-HgS and MS-275 for 48 h, the fluorescence intensity in cells treated with different doses of ZT/β-HgS was significantly enhanced compared with the group administered MS-275 alone, exhibiting a certain dose dependence (Fig. 5A-C, Fig. S8, S9). Results after 72 h of administration further confirmed that the intracellular fluorescence intensity of the combined administration group remained significantly higher than that of the single administration group (Fig. 5D and Fig. S10). Furthermore, drug accumulation at 48 h in the ZT/β-HgS high-dose groups was observed using a High-content screening system, with a substantial amount of drug accumulation and significant fluorescence signal enhancement clearly visible in the cytoplasm of live cells (Fig. 5E). These findings suggest that ZT/β-HgS could facilitate the transport of MS-275 into HBMECs.

Despite ZT/β-HgS has been confirmed to promote drug transport both in vitro and in vivo these findings, doubts persist regarding whether HgS NPs generated instantaneously in vivo can bind to drugs and facilitate drug delivery. To address this, after combined administration of ZT (3 mg/kg)/β-HgS (1.5 mg/kg) with MS-275 (20 mg/kg) in healthy mice for one day, the HgS NPs were separated from collected blood using differential centrifugation, varying from 10³ to 10⁵ g, at 4 °C, and the drug was extracted from the centrifugation product using dimethyl sulfoxide (DMSO) for subsequent detection by HPLC. However, no drugs were detected, indicating that the drug did not bind to the HgS NPs (Fig. S11). This suggests that the HgS NPs biologically formed in vivo may facilitate drug delivery in a stimulatory rather than carrier role.

Fig. 5

ZT/β-HgS promotes the entry of MS-275 into HBMEC. (A-B) After 48 h of treatment with different drugs (MS (2 ng/mL), ZT (50, 100, 300 ng/mL) + MS (2 ng /mL), β-HgS (25, 50, 150 ng/mL) + MS (2 ng/mL)), the fluorescence intensity of the drugs in the HBMECs was observed by a confocal laser scanning microscope. MS-275 was labeled with FITC (green signal). Nuclei were stained with DAPI (blue signal). Scale bars, 50 μm. (C-D) Evaluation of intracellular drug fluorescence intensity in each treatment group at 48 and 72 h. (E) ZT (300 ng/mL)/β-HgS (150 ng/mL) combined with MS-275 incubation of HBMECs for 48 h, real-time observation of drug distribution and enrichment within live cells by High-content screening system (Operetta CLS, PerkinElmer, US). Data represent five independent experiments. * p < 0.05, ** p < 0.01, *** p < 0.001 versus MS

Proteomics suggests that ZT/β-HgS promotes vesicle transport

The above results confirm that ZT/β-HgS can be synthesized into nanoparticles and promote the drug into the brain to achieve synergistic effect, but deny the assumption that nanoparticles may be carriers. It is speculated that their mode of action should be to temporarily open the BBB or promote its active transport. To further clarify the possible mechanisms, we used proteomics analysis to investigate their regulatory effects on proteins in HBMECs. The results showed that compared with the control group, ZT significantly upregulated 29 proteins and downregulated 6 proteins, β-HgS significantly upregulated 20 proteins and downregulated 10 proteins, and 16 proteins were jointly regulated by both (Fig. 6A). From the analysis results of PCA and heatmap (Fig. 6B, C), it can be seen that β-HgS, as the main component of ZT, has a clear consistency with ZT in regulating proteins in cells. First of all, we noticed that there were no significant changes in intercellular tight junction and adhesion junction proteins, indicating that ZT/β-HgS may not promote drug delivery to the brain by temporarily interrupting BBB. Interestingly, most of these proteins are associated with intracellular and intercellular material transport. The results of subcellular organelle localization showed (Fig. 6D) that 24 proteins were localized on the cell membrane, and a large number of proteins were localized on the endoplasmic reticulum, lysosome and mitochondria. These organelles are involved in the synthesis and transport of intracellular substances. Among them, WDR91 and RALGAPB, which may be involved in regulating endocytosis and exocytosis [50,51,52,53,54,55,56,57], were significantly upregulated, while AGAP3, which inhibits the expression of caveolin [58], was significantly downregulated, and EPHA2, which promotes the expression of P-gp [59,60,61], was also significantly downregulated (Fig. 6E). The expression regulation of these proteins promotes the expression of key endocytosis proteins clathrin and caveolin, while reducing the expression of efflux protein P-gp, which may be related to ZT/β-HgS promoting drug entry into the brain. In addition, CD63, SLC7A6, and TSPAN4 were significantly downregulated, and they served as specific marker proteins for solute carrier, exosome, and migrasome [62,63,64,65,66,67,68], suggesting that ZT/β-HgS may also affect intracellular and intercellular vesicle transport. Meanwhile, the results also showed that MACF1, ATF6 and DCTN2 were also significantly regulated. These proteins are related to the structure and function of microtubules, which are the key supporting structures for vesicle transport [69,70,71]. We further verified the expression of these proteins (WDR91, RALGAP, AGAGP3, EPHA2, SLC7A6, TSPAN4, ATF6, DCTN2) by Western blotting, and confirmed that their expression changes in cells were basically consistent with the proteomic results (Fig. 6F). In general, ZT/β-HgS may not temporarily interrupt the tight junction of HBMECs, but have an important impact on the vesicular transport process of cells.

Fig. 6

ZT/β-HgS affects the differential expression of proteome in HBMECs. (A) Volcano plots present the differentially expressed proteins of ZT and β-HgS compared with HBMECs in the control group (n = 3, FC > 1.2, P value < 0.05). Statistical analysis was performed with a two-tailed unpaired Student’s t test. Proteins with significantly increased and decreased expression are indicated by red and blue dots, respectively. (B) Heat map of cluster analysis of differentially expressed proteins in each group based on one-way ANOVA. Red and green represent proteins up-regulation and down-regulation, respectively. (C) Principal component analysis of all proteomes in different treatment groups. (D) Subcellular localization of 51 differentially expressed proteins based on GO functional annotation. The number and proportion of proteins on different subcellular organelles were annotated. (E) Effect of ZT and β-HgS on the expression of intracellular and intercellular transport related proteins. (F) Key differential proteins in proteomics were validated in vitro by Western blot and quantitative analysis. One way ANOVA was used to compare the differences between the treatment groups and the control group. n = 3, * P < 0.05, **P < 0.01

Mechanism of ZT/β-HgS on improving BBB permeability to drug

In order to further verify the results of proteomics, we studied the expression changes of BBB junction and endocytosis proteins in vivo and in vitro. The brain microvascular endothelial cells (BMECs) comprising the BBB are interconnected by intercellular junctions, notably tight junctions (TJs) and adherens junctions (AJs), which are crucial for maintaining BBB integrity and regulating its permeability. Degradation of TJs proteins such as occludin, AJs protein vascular endothelial (VE)-cadherin, and TJs accessory protein zonula occludens-1 (ZO-1) within BMECs can lead to capillary leakage, thereby increasing BBB permeability [72,73,74]. To explore whether HgS NPs facilitate the degradation of intercellular junctions within the BBB, we conducted immunohistochemistry and Western blot analyses for occludin, ZO-1, and VE-cadherin in vitro using human brain microvascular endothelial cells (HBMECs) treated with ZT/β-HgS for 48 h (Fig. 7A, B, Fig. S12, S13). Remarkably, no alterations in the distribution or expression levels of these proteins were observed in cells treated with ZT/β-HgS compared to normal cells. Consistent with the in vitro findings, in vivo experiments also demonstrated that the expression levels of occludin, ZO-1, and VE-cadherin in the brains of mice treated with ZT/β-HgS were comparable to those in normal mouse brains (Fig. 7C, D, Fig. S14). Both in vivo and in vitro results indicated that HgS NPs did not induce increased BBB permeability by damaging tight junctions and adherens junctions.

Additionally, the potential impact of HgS NPs on BBB disruption was evaluated through an Evans blue (EB) penetration assay. Mice were administered EB dye (2% w/v, 4 mL/kg) via tail vein injection at 4 h and 2 days after ZT/β-HgS administration (3.0 mg/kg, 1.5 mg/kg). The mice were euthanized 2 h after EB injection, and brain tissues were photographed and quantitatively analyzed following cardiac perfusion. As anticipated, no blue EB dye was detected in the brains of both normal mice and those administered with ZT/β-HgS (Fig. 7E). Quantitative analysis further confirmed that the absorbance values across all groups were similar, indicating that HgS NPs did not induce BBB disruption (Fig. 7E). These findings collectively suggest that HgS NPs facilitate drug accumulation in the brain without affecting the distribution and expression levels of intercellular junctions, thereby not leading to increased BBB permeability.

Then, whether ZT/β-HgS can enhance the permeability by regulating the transporters within the BBB? Caveolin-mediated or clathrin-mediated endocytosis are the primary pathways through which free drugs or nanoparticles can traverse the BBB and enter the central nervous system (CNS). Conversely, the efflux transporter P-glycoprotein (P-gp), expressed on both sides of the BBB, plays a crucial role in impeding potential therapeutics from crossing the BBB.

To examine whether HgS NPs induce increased BBB permeability by modulating transporters, we first assessed the impact of HgS NPs on key transcytosis proteins, including caveolin, clathrin, and dynamin, as well as the efflux protein P-glycoprotein (P-gp), in vitro using human brain microvascular endothelial cells (HBMECs). Following a 48-h incubation period with ZT/β-HgS, immunofluorescence imaging was conducted to evaluate the expression levels of caveolin, clathrin, dynamin, and P-gp in HBMECs using confocal laser scanning microscopy (CLSM). Compared to normal cells, those treated with ZT/β-HgS exhibited significantly enhanced fluorescence signals of caveolin, clathrin, and dynamin, indicating that HgS NPs could upregulate the expression of endocytosis proteins. Conversely, the fluorescence signals of the efflux protein P-gp in cells treated with ZT/β-HgS were notably reduced compared to normal cells, suggesting that HgS NPs could inhibit the expression of efflux proteins (Fig. 7F, Fig. S15). Moreover, the promotional and inhibitory effects of HgS NPs on endocytosis proteins and the efflux protein were confirmed by Western blot analysis (Fig. 7G, Fig. S16). Specifically, compared to normal cells, the expression levels of all three endocytosis proteins in cells treated with ZT/β-HgS were significantly upregulated, with approximately 68%/61%, 68%/34%, and 27%/12% increases in caveolin, clathrin, and dynamin, respectively, while the efflux protein level was markedly downregulated, with approximately 14%/22% decreases in P-gp.

The influence of HgS NPs on endocytosis proteins and the efflux protein of the BBB in healthy mice was also assessed. The mice received intragastric administration of ZT/β-HgS once every two days for three doses. Subsequently, the brains were collected for protein expression analysis after cardiac perfusion, and the cortex and hippocampus were isolated and processed for Western blotting analysis of protein levels. Western blotting analysis confirmed that the expression levels of caveolin, clathrin, and dynamin in the cerebral cortex and hippocampus were all significantly increased by approximately 34%/37%, 88%/73%, and 70%/42%, respectively, whereas the expression levels of P-gp were markedly reduced by 20%/36% (Fig. 7H). Moreover, immunohistochemical images revealed a significant increase in the distribution and fluorescence intensity of these three endocytosis proteins in brain microvessels, while the distribution and fluorescence intensity of the efflux protein P-gp in microvessels were notably decreased (Fig. 7I, J, Fig. S17). Furthermore, the caveolin and clathrin inhibitors genistein and chlorpromazine hydrochloride have been proven to eliminate the effect of ZT/β-HgS in promoting drug entry into HBMECs cells (Fig. S18). Overall, these findings are preliminarily confirmed that HgS NPs may facilitate drug penetration of the BBB into the brain by upregulating key endocytosis proteins and downregulating the efflux protein, rather than through temporary disruption of intercellular junctions in the BBB.

Fig. 7

Effects of ZT/β-HgS on BBB permeability. (A) Effect of ZT/β-HgS on HBMECs intercellular connexins (ZO-1, occludin, and VE-cadherin) at 48 h. (B) Western blots did not show that the expressions of connexins in HBMECs were affected by ZT/β-HgS. (C) The expressions of connexins in brain were quantitatively analyzed by confocal laser scanning microscopy. (D) Western blots and quantitative analysis found that ZT/β-HgS had no effect on the expressions of connexins in mouse brain. (E) Physical disruption of the BBB was characterized by Evans blue (EB) penetration assay. (F) Effects of ZT/β-HgS on transporters in HBMECs were evaluated by immunocytochemistry (n = 3). (G-H) The expression of transporters in cells and brain were evaluated by Western blot (n = 3). (I-J) Effects of ZT/β-HgS on transporters in microvessels of BBB in mice (n = 6). Double-labeled immunostaining of caveolin, clathrin, dynamin and P-gp in mouse brain frozen sections. Target proteins were labeled with DyLight 594(Red). Nuclei were stained with DAPI (blue signal), and blood vessels were labeled with fluorescein isothiocyanate lectin (green signal). Data were presented as means ± SEM. Experiment performed at least three times. One-way ANOVA was used to analyze the differences between groups. * P < 0.05, **P < 0.01, ***P < 0.001 compared to Con

ZT/β-HgS facilitate the drug across BBB in the form of HgS NPs

The preceding results raise two distinct points. Firstly, ZT and β-HgS can be synthesized into uniform-sized HgS NPs both in vitro and in vivo. Secondly, ZT and β-HgS can facilitate drug penetration across the BBB to achieve synergy by modulating the expression of active transporters in the BBB. However, these processes are inherently separate, and the enhancement of drug transport across the BBB does not necessarily imply a direct correlation with HgS NPs. How then can we confirm whether this effect is indeed linked to HgS NPs or their synthesis process? One indirect approach involves examining whether the onset time of the observed effect aligns with the in vivo distribution of HgS and the synthesis time of HgS NPs.

Therefore, the quantification of Hg content in different tissues was assessed using atomic fluorescence spectrometer (AFS) analysis. The results revealed that the concentration of Hg in the brain continuously increased over time after oral administration of ZT/β-HgS, peaking at 20 ng/g at 24 h, and then gradually decreasing to 13.9 ng/g at 48 h, 5.65 ng/g at 96 h, and 4.41 ng/g at 192 h (Fig. S19A). The changing trend of Hg content in the brain mirrored the incremental trend of drugs in the brain, indicating a potential relationship between Hg or HgS NPs and the promotion of drug penetration across the BBB. On the 8th day, post-administration of ZT/β-HgS, the Hg content in the mouse brain rapidly decreased to near-normal levels, explaining the lack of drug increment after 8 days of ZT/β-HgS administration (Fig. S19B, C). Additionally, the changing trends of Hg content in the lung and blood mirrored those observed in the brain. However, in the liver and kidneys, the Hg concentration remained elevated on the 8th day, possibly due to metabolism (Fig. S19D, E). These results collectively demonstrate that promoting drug accumulation in the brain is positively correlated with the temporal distribution of Hg or HgS NPs in the brain.

Further, the onset time of the increased cellular drug uptake capacity following combined administration was initially evaluated. Cells incubated with ZT/β-HgS and MS-275 exhibited nearly identical fluorescence signals at 6, 12, and 24 h compared to those treated with free MS-275 alone, indicating that ZT/β-HgS did not enhance the interaction between MS-275 and HBMEC cells (Fig. S20). Quantitative analysis revealed that the fluorescence intensity of the combined administration groups was slightly weaker than that of the single administration group, although not significantly different (Fig. 8A). These findings diverged markedly from the previous intracellular drug uptake results at 48 h, suggesting that ZT/β-HgS did not enhance drug entry into cells within the initial 24 h post-combined administration. Considering the biosynthesis speed of HgS NPs in cells discussed earlier, it is notable that the timeframe of ZT/β-HgS facilitating MS-275 entry into cells closely mirrored the time of HgS NPs formation. This observation indirectly suggests that ZT/β-HgS exerts a synergistic effect subsequent to the generation of HgS NPs. To further substantiate this inference, cells were pre-incubated with ZT/β-HgS alone for 36 and 42 h to ensure HgS NPs formation, followed by the addition of MS-275 labeled with FITC fluorophore to the pre-treated cells. As anticipated, cells pre-incubated with ZT/β-HgS exhibited significantly higher fluorescence signals at 6 and 12 h post-incubation with MS-275 compared to those without prior cultivation with ZT/β-HgS (Fig. 8B, C, Fig. S21).

In particular, the expression levels of transcytosis proteins and efflux protein also exhibited a close correlation with the generation time of HgS NPs. As depicted in Fig. 8D and Fig. S22, the expression levels of caveolin, clathrin, and dynamin proteins in cells treated with ZT/β-HgS showed no significant changes, and P-gp was not downregulated compared to normal cells at 24 h. However, significant changes in the expression levels of these proteins were observed at 48 h. These findings further indirectly support the notion that the changes in protein expression may be triggered by the generation of HgS NPs in the organism. Unlike the synthesis speed observed in vitro, the synthesis speed of HgS nanoparticles in mice is notably rapid, making it challenging to precisely track the onset time of their synergistic effects. However, their clearance speed is slow, and the presence or absence of nanoparticles in the brain may be correlated with their efficacy. The long-term efficacy of ZT and β-HgS has been previously confirmed, with the effect of facilitating drug transport across the BBB lasting for up to 4 days in vivo. If the effect is indeed related to HgS NPs, a substantial number of nanoparticles should be detected in the brain within the effective period. Conversely, when the effect diminishes, these nanoparticles should diminish or become insufficient to exert their function. To examine this hypothesis, brain tissues from mice were collected at 2, 4, and 8 days after administration of ZT or β-HgS, and the distribution of nanoparticles in the brain was assessed using TEM. The results revealed a substantial presence of nanoparticles in the brain tissue of mice on day 2 and day 4, whereas it was challenging to observe nanoparticles or only a few nanoparticles in the brain of mice on day 8 (Fig. 8E). The duration of existence of HgS NPs in the mouse brain closely aligns with the duration of the effect of ZT/β-HgS in promoting drug entry into the brain (as shown in Fig. 4H), further indirectly suggesting that the synergistic effect may indeed be related to HgS NPs or, at the very least, their synthesis process. These findings confirm that after ZT/β-HgS is auto-synthesized into HgS NPs by the organism, these nanoparticles promote enhanced transcytosis, thereby facilitating drug delivery across the BBB.

Fig. 8

HgS NPs were the key to the synergistic effect of ZT/β-HgS. (A) ZT/β-HgS has no effect on drug entry into cells when no HgS NPs were synthesized within 6 to 24 h. (B) HBMEC were pre-incubated with ZT/β-HgS for 36 h and 42 h to ensure the formation of HgS NPs, and then after incubation with MS-275 for 6 h and 12 h respectively, the amount of drug entering the cells and the fluorescence intensity increased significantly. MS-275 was labeled with FITC (green signal). Nuclei were stained with DAPI (blue signal). Scale bar, 50 μm. (C-D) The expression levels of transcytosis proteins in cells incubated with ZT/β-HgS for 24 h. (E) Estimation of the number of HgS NPs in the brain. Data represent five independent experiments. * P < 0.05, ***P < 0.001 compared to Ms

Discussion

The use of heavy metal preparations in traditional Chinese medicine has persisted despite ongoing concerns, primarily due to their unique therapeutic effects, particularly in treating CNS disorders. Safety concerns surrounding heavy metal preparations, such as ZT and cinnabar, have been extensively studied, with numerous studies affirming their safety and nontoxic nature [19,20,21,22,23,24]. However, the precise mechanisms underlying the therapeutic actions of heavy metal preparations in traditional Chinese medicine have remained elusive. Recent advancements in the field of nanomedicine have revealed a novel pathway for drug delivery through the spontaneous synthesis of metal nanoparticles (NPs) by living organisms, showcasing their high biocompatibility [28,29,30,31]. This discovery has sparked a groundbreaking hypothesis suggesting that HgS nanoparticles are spontaneously generated in the human body following the oral administration of metal preparations, subsequently facilitating drug delivery. Remarkably, the presence of HgS nanoparticles with diameters ranging from 5 to 7 nm has been confirmed in the blood of patients and mice treated with ZT/β-HgS, indicating the in vivo generation of HgS NPs. Notably, the morphology of HgS NPs synthesized in mice were clearly and completely presented, with 5–7 nm nanoparticles wrapped in a dense layered protein coronas, and an overall particle size of 30–195 nm. This result intuitively confirms the speculation about the morphology of selfsynthesized metal nanoparticles in animals, and hard protein coronas may be an important factor determining the distribution of nanoparticles within the body. But this is different from the loose protein coranas structure formed by nanoparticles in vitro, which may be related to the difficulty of in vitro serum culture to replicate the complex biological processes in vivo. Purification of large-sized nanoparticles from brain also challenges the conventional notion that nanoparticles within the size range of 5–100 nm are critical for crossing the BBB [73]. From the recent studies on the synthesis of metal nanoparticles in animals, researchers have reported the in vivo synthesis and structural characterization of CdTe, magnetite, ZnO and Pt. The particle sizes of these nanoparticles ranged from 2 nm to 40 nm, indicating that the particle sizes of nanoparticles produced by different metal materials in vivo were significantly different due to their chemical properties [28,29,30,31]. Furthermore, to the best of our knowledge, the commonly anticipated dense and layered protein corona morphology has not been verified in any previous study [38, 39]. In this study, more than 200 different plasma corona proteins were detected for HgS NPs and quantified by LC-MS. Unlike protein coronas on other metal NPs such as Au NPs [13, 14] and Pt NPs [12], the protein coronas on the surface of HgS NPs are mainly consist of hemoglobin subunits, transcriptional repressors, and superoxide dismutase, rather than albumin. More studies on the surface modification of nanoparticles are necessary, especially on the loading of some functional proteins, which can significantly improve the performance of nanoparticles as carriers. The potential implications of these protein coronas formed on the HgS NP surface might be closely related to the delivery of drugs. For example, hemoglobin, in addition to its typical functions of penetrating and transporting oxygen, is also involved in many basic physiological processes such as respiratory reactions and energy transfer processes. Hemoglobin loaded on a variety of nano polymeric materials has been widely studied and applied as a carrier for oxygen delivery in vivo [75, 76]. Compared with free hemoglobin, hemoglobin polymerized by nanomaterials can prolong the circulation time in vivo, reduce nephrotoxicity and prevent vasoconstriction [77, 78]. In addition, hemoglobin can also exchange electrons with metal nanoparticles to form stable biological conjugates. The formation of metal nanoparticles bio conjugates not only increases the stability and biological function, but also may change the distribution and toxicity of metal nanoparticles in vivo [79]. Similarly, whether the combination of HgS NPs and transcriptional receptors can make it have the function of regulating the expression of specific genes is still worth further study [80, 81]. However, owing to the complexity of protein coronas, understanding how they affect HgS NPs across the BBB for drug delivery remains a tremendous challenge and warrants further studies. While several metal NP platforms have been recently explored for drug delivery across the BBB, challenges related to complex design, potential toxicity, and ethical considerations have hindered large-scale manufacturing and clinical translation. Hence, it is still highly desirable to explore an efficient, safe, biocompatible, and easily accessible metal NP drug delivery system across the BBB for the treatment of CNS diseases. We have for the first time confirmed that metal preparations like ZT/β-HgS, clinically used for more than 2000 years in China, can auto-form HgS NPs in the organism for drug delivery across the BBB. Oral administration of ZT/β-HgS in healthy mice has demonstrated that HgS NPs can transit from the bloodstream into the brain, exhibiting high distribution and prolonged retention time. Moreover, HgS NPs significantly enhance drug transport across the BBB, resulting in more than a threefold increase in brain accumulation compared to drug administration alone. Importantly, HgS NPs have been shown to improve the therapeutic efficacy of drugs in enhancing spatial learning and memory abilities in APP/PS1 mice. Mechanistic investigations have revealed that HgS NPs primarily facilitate drug transport across the BBB via caveolin and clathrin-mediated endocytosis pathways, while also downregulating the expression of the efflux protein P-glycoprotein (P-gp) to increase drug absorption in the brain. Importantly, HgS NPs do not induce structural damage to the BBB, as they do not affect tight junctions and adherens junctions. These findings collectively underscore the potential of HgS NPs as a promising drug delivery system for CNS diseases, paving the way for future research and clinical applications in this field. Elucidating the synthesis mechanism is the basis for the development and application of functional metal nanomaterials. At present, it is believed that the synthesis process of metal nanoparticles includes nucleation and growth, but the chemical properties of different metals vary greatly, and different synthesis systems may lead to completely different mechanisms [82]. In addition, the formation mechanism of protein corona shows that the interfacial and electrochemical properties of metal materials are the key factors to determine the types of binding biomolecules [37, 43]. The proteins composition and formation mechanism of protein coronas of different metal materials may be quite different. Therefore, elucidating the synthesis mechanism of HgS NPs in higher organisms is of great significance for the directional biosynthesis of nanomaterials with drug delivery function.

Regardless of synergistic effects, HgS NPs with protein coronas auto-generated in the organism offer distinct advantages over existing NP systems. Specifically, the longstanding clinical use of ZT/β-HgS in Tibetan medicine establishes a solid foundation and safety profile for employing HgS NPs in the drug delivery for CNS diseases. Moreover, the spontaneous formation of HgS NPs in the organism facilitates the drug transport across the BBB without the need for in vitro drug loading or chemical modification, streamlining the clinical application process. Another notable advantage is that HgS NPs promote drug accumulation in the brain through caveolin and clathrin-mediated endocytosis pathways, suggesting broad applicability across different drugs. Furthermore, different from chemically synthesized metal nanomaterials, HgS NPs synthesized in higher organisms are loaded with a huge protein coronas on the surface, which not only helps to increase the size of nanoparticles and reduce their high reactivity, thereby reducing biological toxicity [42, 83], but also the protein coronas with different components and functions give them good biocompatibility and new functions [37]. Therefore, the excellent drug delivery function and favorable biosafety of HgS NPs enhance their attractiveness as a drug delivery platform. In conclusion, HgS NPs auto-generated in the organism represent a novel and potentially transformative drug delivery platform for CNS therapeutics.

Conclusions

This study revealed for the first time that β-HgS, the main component of Zuotai, the core metal preparation in Tibetan medicine, can be synthesized into nanoparticles with a diameter of 30–195 nm in the organism, and promote different structural types of drugs into the brain by regulating endocytosis in the blood-brain barrier, thus achieving synergy. It is particularly noteworthy that we have not only observed the biosynthesis of HgS nanoparticles in vivo and in vitro, but also obtained the most clear and complete morphological characteristics of HgS nanoparticles synthesized by higher living organisms currently available. Metal preparations have always been a very unique existence in Tibetan medicine and traditional Chinese medicine. Although they have been criticized, their unique efficacy in clinical application has never been replaced. This study may open a new idea for further revealing the synergistic mechanism of metal preparations. Nanoparticles with a particle size of 30–195 nm can be synthesized in vivo.