Spatiotemporal activation of unfolded protein response by combined sonodynamic therapy and proteasome inhibition with bortezomib-conjugated TiN nanoparticles

Abstract

Activation of unfolded protein response (UPR) has recently been identified as a new target for cancer therapy. However, pharmacological agents activating UPR have not been approved in clinical practice. In this study, spatiotemporally controllable UPR activation was realized by the combination of ultrasound (US)-triggered sonodynamic therapy (SDT) with titanium nitride (TiN) NPs and proteasome inhibitor bortezomib (BTZ). BTZ-conjugated TiN NPs were prepared by using a linker of carminic acid (CA)-functionalized bovine serum albumin (BSA). CA interacts with BTZ to form dynamic boronate ester bond enabling pH- and ultrasound-responsive drug release and thus overcoming the challenge of poor drug release for the conventional delivery systems. Compared with traditional sonosensitizer TiO₂, TiN has superior SDT efficiency. TiN-mediated SDT inhibits not only the β1/β5 active sites of proteasomes, but also the β2 proteasome activity not targeted by BTZ, thus synergistically triggering endoplasmic reticulum (ER) stress. Moreover, SDT and BTZ complement each other perfectly to activate all three pro-apoptotic UPR pathways including PERK/eIF2α/ATF4, IRE1/JNK and ATF6, thereby leading to dramatic upregulation of CHOP expression, calcium overload and mitochondrial dysfunction. Importantly, this intensified ER stress effectively triggers immunogenic cell death (ICD), characterized by the massive release of damage-associated molecular patterns (DAMPs), thereby further recruiting and activating cytotoxic T lymphocytes and reshaping the immunosuppressive tumor microenvironment. In summary, making up the current lack of effective pharmacological agents activating UPR, this study provides a new strategy to spatiotemporally activate UPR through the ingenious combination of SDT and proteasome inhibition with BTZ-conjugated TiN NPs.

Graphical abstract

Highlights

• •CA was found to bind BTZ through dynamic-covalent boronate bond after screening.
• •BTZ release from CA is pH- and ultrasound-responsive.
• •CA-functionalized BSA was used as a bridge to prepare BTZ-conjugated TiN NPs.
• •TiN-based SDT and BTZ synergistically inhibit all three subunits of proteasome.
• •TiN-based SDT and BTZ synergistically activate proapoptotic UPR pathways.

1. Introduction

Unfolded protein response (UPR) has recently been identified as a new target for treating several diseases [[1], [2], [3]]. In response to intrinsic and extrinsic stimuli such as metabolic insult and oxidative stress, unfolded protein accumulation in the endoplasmic reticulum (ER) leads to ER stress. ER stress can trigger an adaptive UPR to restore protein homeostasis by initiating the activation of three pathways [1]. However, when ER stress is too severe or prolonged, an excessive activation of UPR (terminal UPR) can also redirect signaling from cytoprotective adaptation to cell death [2,4]. Activation of UPR has recently been identified as a new target for cancer therapy. However, pharmacological agents activating terminal UPR directly have not been approved in clinical practice.

Ubiquitin-proteasome system (UPS) is responsible for the degradation of resident or abnormal proteins accumulated in the ER [5]. Proteasome inhibitors can inhibit proteolytic activity of proteasomes to induce ER stress. Bortezomib (BTZ), a dipeptide boronic acid, is the first proteasome inhibitor clinically approved for treating multiple myeloma [6]. Despite its success as treatments for hematologic malignancies, BTZ is ineffective against solid tumors partially due to cytoprotective adaptive UPR [7]. Proteasome inhibitors alone cannot induce sufficiently severe UPR activation [1]. While under the ER stress condition induced by proteasome inhibition, converting adaptive UPR into terminal UPR leading to cancer cell death can be potentially achieved by applying UPR activators [8].

Ultrasound (US)-triggered sonodynamic therapy (SDT) has recently emerged as a powerful and promising approach for the treatment of deep-seated tumors due to noninvasiveness, high spatiotemporal selectiveness, and deep penetration capacity [9]. US controllably activates sonosensitizers to produce massive reactive oxygen species (ROS, mainly singlet oxygen (¹O₂)) and consequent oxidative cell damage. Integration with SDT has recently been considered as an effective strategy to improve the efficacies of cancer treatments including chemotherapy [10,11] and immunotherapy [9,12]. In the preliminary study, we found that SDT can induce UPR activation, especially if it is combined with the proteasome inhibitors. As ROS generation can interact with proteasome inhibitors to synergistically trigger ER stress [13], we hypothesized that combination of proteasome inhibition and SDT can achieve a spatiotemporal and synergistic “two-hit” assault on cancer cell proteostasis. BTZ first triggers ER stress through protein accumulation, which is then amplified through oxidative damage by SDT, leading to overactivation of the terminal UPR. Accumulating evidence suggests that such severe ER stress can also trigger immunogenic cell death (ICD) through the release of damage-associated molecular patterns (DAMPs) [14]. However, a sophisticated nanoplatform capable of combining proteasome inhibitors and sonosensitizers to achieve tumor microenvironment-responsive drug release and external US-triggered protein disruption remains lacking.

In this study, a dynamic covalent boronate ester-based assembly strategy was developed to combinatorially deliver titanium nitride (TiN) nanoparticles (NPs) and BTZ for combined proteasome inhibition and SDT (Scheme 1). As titanium nanostructures have a high binding affinity to the biomolecules such as serum proteins [15], bovine serum albumin (BSA) was used as the surface modification for the TiN NPs. The protein corona can reduce the non-specific cellular uptake of nanoparticles, thereby prolonging blood circulation [16]. BSA was then functionalized with carminic acid (CA) for effective loading with BTZ. pH- and ultrasound-responsive drug release was achieved by this delivery system. Detailed design and novelty of this study were illustrated as follows.

Scheme 1.

Schematic illustration of BTZ-conjugated TiN NPs via reversible boronate ester bonds for synergistic unfolded protein response (UPR) activation through combination of proteasome inhibition and sonodynamic therapy (SDT). a) Discovery of carminic acid (CA) for BTZ loading though pH- and US-responsive dynamic covalent boronate ester bonds. After screening from polyphenol compound library, CA was found to interact to BTZ through reversible boronate ester bonds, and controlled drug release can be realized by the acidic environment and US irradiation. b) Preparation of BSA-CA coated TiN NPs and BTZ loading based on the intermolecular coordination between BTZ and CA. c) ROS and additional inhibition of proteasomal β2 active site by SDT enable more accumulation of ubiquitinated proteins triggering enhanced ER stress after the combination treatment with BTZ and SDT. ER stress-induced calcium overload and ROS elevation lead to mitochondrial dysfunction. More importantly, the severe ER stress induced by the combination therapy triggers the activation of terminal UPR pathways including PERK/eIF2α/ATF4, ATF6, IRE1/JNK leading to cell death, with the evidence of upregulation of CHOP and caspase-12 activation.

First, TiN NPs with higher sonosensitizing efficiency were synthesized for SDT. For the SDT study, the inorganic sonosensitizers have the advantages of lower phototoxicity and higher chemical stability, compared with organic ones. Among them, titanium-based nanostructures have generated considerable scientific interest due to their tunable physicochemical properties, excellent catalytic efficiency, and good biocompatibility [17]. However, the representative sonosensitizer titanium dioxide (TiO₂) has a relatively low ROS quantum yield activated by US owing to its wide bandgap (3.2 eV) and fast integration of the electron (e⁻) and hole (h⁺) [18]. Recently, TiN nanostructures as a type of transition metal nitride material with highly effective electrocatalytic properties have attracted much attention. In the TiN nanostructures, the tetravalent Ti bonds with trivalent N, resulting in free-electron surplus. At the same time, there are vacancies in the lattice structure of TiN nanostructures. The electron cloud is unevenly distributed around the vacancy, resulting in deficient electric dipoles for enhancing microwave excited thermoacoustic therapy. Due to the existence of surface deficient structure of TiN which enhances the separation efficiency of the e⁻-h⁺ pairs [19], the sonosensitizing efficiency of TiN was assumed to be better than traditional TiO₂ nanostructures under US irradiation. In this study, we synthesized the TiN NPs, and found that TiN NPs have a higher ROS quantum yield than TiO₂ under US. TiN was thus used as a new sonosensitizer in this work.

Second, CA was identified for BTZ loading based on the dynamic covalent binding bond which is ultrasound- and pH-responsive. BTZ can chelate with polyphenols such as epigallocatechin gallate (EGCG) and quercetin (QUE) due to boronate-catechol complexation chemically blocking the boronic acid active site of BTZ. Therefore, polyphenols with strong interactions with BTZ due to too stable bonds should be avoided for BTZ delivery. In the preliminary screening study, we found that most polyphenols strongly bind BTZ. However, CA, an anthraquinone compound with keto-enol moieties, can bind to BTZ reversibly and does not block the cytotoxic effect of BTZ. CA is classified as a colorant exempt from certification by FDA and is also approved as a food additive in Europe. Inspired by the dynamic covalent chemistry for drug delivery, CA was utilized for delivery of BTZ in this study. Detailed proton nuclear magnetic resonance (¹H NMR), and Raman spectroscopy were conducted to confirm the reversible dynamic binding of CA and BTZ. Moreover, controlled BTZ release can be triggered by ultrasound and acidic environment.

Third, spatiotemporally controllable UPR activation through combined SDT and proteasome inhibition with BTZ-conjugated TiN NPs was applied to achieve a superior anticancer effect. The synergistic mechanism for UPR activation, the effect of the combination of SDT and BTZ on proteasome subunits and UPR pathways were finally demonstrated.

The current proteasome inhibitors inhibit the three proteolytically active β-ring subunits of the proteasome with varying degrees of efficacy, being most potent towards β5 subunit, less potent for β1-subunit and with little activity towards β2 subunit. In this study, it was firstly reported that SDT inhibits not only the proteasomal β1/β5 active sites, but also the β2 proteasome activity not targeted by BTZ. As a result, BTZ and SDT complement each other perfectly to induce more accumulation of ubiquitinated proteins triggering ER stress.

Activation of UPR is initiated by the activation of three ER transmembrane proteins including protein kinase R-like ER kinase (PERK), inositol requiring enzyme 1 (IRE1), and activating transcription factor 6 (ATF6) [1]. This combination therapy synergistically activates pro-apoptotic UPR pathways including PERK/eIF2α/ATF4, IRE1/JNK, and ATF6, leading to dramatic upregulation of CHOP expression, calcium overload and mitochondrial dysfunction. RNA sequencing study confirmed the above results.

In summary, this study designed BTZ-conjugated TiN NPs based on a new dynamic covalent boronate ester-based assembly method. Making up the current lack of effective pharmacological agents activating UPR, this study provides a new strategy to spatiotemporally activate UPR through the rational combination of SDT and proteasome inhibitor. Moreover, in the past years, overcoming proteasome inhibitor resistance and expanding the application in the treatment of solid tumors have become core goals for the development of proteasome inhibitors [7]. Several new drug delivery systems [20,21] and combination therapies [22,23] were developed to solve the above problems, albeit without success in clinical translation. The results from this study also indicate that SDT shows great potential in sensitizing the anticancer effects of proteasome inhibitors and reshaping the tumor immune microenvironment via induction of ICD.

2. Materials and methods

2.1. Synthesis of TiO₂ NPs

TiO₂ NPs were prepared according to a previously reported method with a slight modification [24]. Typically, TiOSO₄ powder (1 g) was dissolved in DI H₂O (120 mL) and stirred for 10 min. NH₃·H₂O (16 mL, 35 wt%) and H₂O₂ (8 mL, 50 wt%) were added sequentially to a reaction flask. The mixture was stirred for 30 min until a transparent yellow solution was formed. Subsequently, absolute ethanol (39 mL) was introduced dropwise into the solution under continuous stirring. The resulting mixture was then heated and maintained at 80 °C for 12 h.

2.2. Synthesis of TiN NPs and TiN@BSA NPs

The TiN NPs were obtained through the nitridation treatment of TiO₂ NPs in an ammonia-flowing environment (ammonia flow rate was 0.1 L min⁻¹, 800 °C, 5 h). BSA-coated TiN NPs were prepared with slight modifications according to previous reports [25,26]. BSA (5 mg) was dissolved in KH₂PO₄ buffer (1 mL, 10 mM, pH 5.2). BSA solution (1 mL) was added to TiN NPs (1 mL, 10 mg mL⁻¹) and sonicated (480 W, 8 kHz, 1s on and 1s off) in an ice bath for 5 min, and the mixture was then stirred for 4 h. The BSA-coated TiN NPs were then collected by centrifugation (5000 rpm, 5 min) and washed with DI H₂O.

2.3. Characterization of TiN and TiN@BSA NPs

The sizes and zeta potential values of TiO₂ and TiN NPs were determined by dynamic light scattering (DLS) using a Malvern Zetasizer Nano ZS90 analyzer. The transmission electron microscopy (TEM) imaging and elemental mapping were performed using an FEI HRTEM TF20 equipped with energy-dispersive X-ray spectroscopy (EDX). Scanning electron microscopy (SEM) imaging was performed on a JEOL JSM7800F microscope operated at 20 kV. The absorption spectra of the samples were measured by a SpectraMax Paradigm multimode microplate reader (Molecular Devices, USA). The BSA concentration was measured using a BCA protein assay. Briefly, the TiN NPs were reacted with the reagent, centrifuged at 5000 rpm for 10 min, and the BSA concentration in the supernatant was measured. A solution of free BSA, prepared at a concentration equivalent to that reacted with the nanoparticle samples, served as a control. Changes in the characteristic protein absorbance peak were then recorded.

2.4. ¹O₂ generation activity of SDT

The production of ¹O₂ was detected with 1,3-diphenylisobenzofuran (DPBF) [27]. ¹O₂ leads to DPBF being oxidized to 1,2-dibenzoylbenzene (DBB). The oxidation of DPBF can be monitored by the decrease in absorbance. This makes DPBF a useful probe for detecting and quantifying the presence of ¹O₂. First, TiN (10 μL, 4 mg mL⁻¹) was mixed with DPBF (100 μL, 1 mg mL⁻¹) and DI H₂O (890 μL). The mixture was irradiated with ultrasound (50 kHz, 2 W cm⁻², 50% duty cycle) in the dark for predetermined time intervals using a handheld ultrasound machine. The sonodynamic generation of ¹O₂ was quantified by monitoring the decay in the characteristic absorbance of DPBF. The UV−vis−NIR spectra were recorded using the microplate reader.

2.5. ^•OH generation activity of SDT

The generation of hydroxyl radicals (^•OH) was quantified using a chromogenic assay based on the oxidation of 3,3′,5,5′-tetramethylbenzidine (TMB) [28]. The production of ^•OH was determined by measuring the increase in the characteristic absorbance of oxidized TMB (oxTMB) at 652 nm. Typically, the working test solution was prepared by mixing TMB (1 mM), TiN NPs (40 μg mL⁻¹) in a sodium acetate buffer (0.5 M, pH 4.5), and the mixture was exposed to ultrasound (50 kHz, 2 W cm⁻², 50% duty cycle) for prolonged durations in the dark. Afterward, the changes in TMB were characterized by recording the absorption intensity in UV−vis−NIR spectra by the microplate reader.

2.6. Cell viability assay

To screen polyphenol-BTZ conjugates with tunable stability of boronate ester bonds, 2500 4T1 cells were seeded into each well in a 96-well plate. After incubation for 12 h, the cells were treated with BTZ (0–40 nM BTZ) or polyphenol-BTZ conjugates (0–40 nM BTZ). CA-BTZ conjugate was obtained by mixing 100 μL CA (10 mM, in DI H₂O) and 100 μL BTZ (10 mM, 20 mM, 30 mM, in DMSO), followed by stirring for 30 min at room temperature. The EGCG-BTZ, QUE-BTZ, DA-BTZ, TA-BTZ, CFA-BTZ, GA-BTZ, ED-BTZ, RHE-BTZ, PP-BTZ, AA-BTZ, CAS-BTZ, AR-BTZ, and CUR-BTZ were also prepared. After 48 h of treatment, cell viability was determined using a standard 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. The culture medium was replaced with a fresh medium containing 0.5 mg mL⁻¹ MTT and incubated for 4 h. The MTT solution was then aspirated, and the resulting purple formazan crystals were solubilized in 150 μL of DMSO. After shaking for 10 min, the absorbance at 540 nm was recorded using the microplate reader. Cell viability was calculated as a percentage relative to the untreated control group.

To verify the successful loading of BTZ, BTZ (100 μL, 150 μM) was incubated with TiN@BSA or TiN@BSA-CA (400 μL, 50 mg mL⁻¹ TiN) for 12 h. After centrifugation (12,000 rpm, 10 min), the supernatant and precipitate drug-loaded NPs were collected, and the precipitate NPs were resuspended in 500 μL DI H₂O. The cytotoxicity of the precipitate drug-loaded NPs and the supernatant was evaluated by the MTT assay. The mixture of TiN and BTZ without centrifugation was also used as a positive control. The same volume of each sample was taken and diluted 1000 times with culture medium for cell treatment. The final concentrations of BTZ and TiN for the mixture of TiN and BTZ without centrifugation are 30 nM and 40 μg mL⁻¹, respectively.

To investigate the effect of TiN NPs on the viability of normal cells, the RAW 264.7 murine macrophage cells were seeded in a 96-well plate and incubated overnight. The cells were then treated with TiN or TiN@BSA-CA (0–200 μg mL⁻¹ TiN) for 48 h. The cell viability was determined by the MTT assay.

For BTZ and SDT combination treatment, the cells were then treated with TiN (20 μg mL⁻¹ TiN), BTZ (20 nM BTZ), BTZ-TiN (20 μg mL⁻¹ TiN, 20 nM BTZ) for 8 h. For SDT treatment, the designated cells were treated with US (50 kHz, 2 W cm⁻², 50% duty cycle) for 4 min. To study the effect of ROS scavenger on SDT, the cells were treated with TiN for 8 h in the presence of N-acetylcysteine (NAC, 2 mM) and subjected to a 4-min US irradiation. After 48 h of treatment, cell viability was determined by the MTT assay. Calcein AM/PI staining was also used to detect living and dead cells. After the treatments, the medium was removed, and a fresh medium (100 μL) containing calcein AM (2 μM) and PI (8 μM) was added and incubated for 30 min. The cells were observed under an Olympus IX71 fluorescence microscope (USA).

2.7. Characterization of CA-BTZ

The UV−vis absorption spectra analysis of CA-BTZ was performed on the SpectraMax Paradigm multimode microplate reader. CA, BTZ, and CA-BTZ were vacuum-dried overnight and Raman scattering measurement was performed using a Renishaw inVia Qontor micro-Raman spectrometer. A 785 nm laser with a power output of 10 mW and an integration time of 5 s was utilized for each measurement. Deuterated methanol was used as the solvent, and ¹H NMR spectroscopy analysis of CA-BTZ was performed using a Bruker AVANCE III 600 MHz NMR spectrometer equipped with a 5 mm Cryoprobe (CPDCH600S3 C/HD-05 Z) (Germany).

2.8. Drug loading and encapsulation efficiency assay

The conjugation of CA to BSA was accomplished via a standard EDC/NHS-mediated amide coupling reaction. Briefly, CA (6.4 mg, in 5 mL DI H₂O) was activated by the sequential addition of NHS (45.0 mg, in 5 mL DI H₂O) and EDC (74.0 mg, in 5 mL DI H₂O) under stirring for 4 h at room temperature. Subsequently, a solution of BSA (50.0 mg, in 5 mL DI H₂O) was added dropwise to the activated CA mixture, and the mixture was stirred for 24 h at room temperature. The final product was purified by dialysis (MWCO 3000 Da) against deionized water to remove unreacted reagents, and the CA-BSA conjugate was collected. CA-BSA was added to TiN (volume ratio 1:1) and then sonicated (480 W, 8 kHz, 1s on and 1s off) in an ice bath for 5 min. After stirring for 4 h, TiN@BSA-CA was obtained. TiN-BTZ was prepared by adding BTZ to TiN@BSA-CA and stirring overnight, followed by centrifugation (12,000 rpm, 30 min). The supernatant was collected, and the BTZ concentration of the supernatant was determined using an Agilent 6230 time-of-flight liquid chromatography-mass spectrometry (LC-MS) system (USA). The encapsulation efficiency (EE) and drug loading content (LC) of BTZ were calculated.

EE (%) = (mass of BTZ in the TiN-BTZ/total mass of BTZ added) × 100% (1)

LC (%) = (mass of BTZ in the TiN-BTZ/total mass of TiN-BTZ) × 100% (2)

2.9. BTZ release assay

The above-obtained BTZ‐loaded TiN NPs (100 μg) were added into phosphate-buffered saline (PBS, pH 7.4, 1 mL) or citrate buffer (pH 4.5, 20 mM, 1 mL). The mixed solution was stirred for 0–48 h at 37 °C. An aliquot of medium (100 μL) was withdrawn for analysis at different time points. An equal volume of PBS or citrate buffer was supplemented to the solution. The collected aliquots were centrifuged at 12,000 rpm for 30 min, and the concentration of the drug in the resulting supernatant was quantified using the LC-MS system.

To investigate the effect of ultrasound on drug release, BTZ‐loaded TiN NPs (100 μg) were added into PBS (pH 7.4, 1 mL) or citrate buffer (pH 4.5, 20 mM, 1 mL), and stirred at 37 °C. The mixture was exposed to ultrasound (50 kHz, 2 W cm⁻², 50% duty cycle, 4 min), and further incubated for 30 min. After centrifugation (12,000 rpm, 30 min), the drug concentration of the supernatant was determined by the LC-MS system. All release experiments were conducted in triplicate.

2.10. Intracellular GSH content assay

One hundred thousand 4T1 cells per well were seeded into a 6-well plate. After overnight incubation, the cells were treated with TiN (20 μg mL⁻¹ TiN), BTZ (20 nM BTZ), or TiN-BTZ (20 μg mL⁻¹ TiN, 20 nM BTZ) for 8 h. For SDT treatment, the designated cells were treated with ultrasound (50 kHz, 2 W cm⁻², 50% duty cycle) for 4 min. Following 24-h incubation, the total cellular proteins were extracted by lysing the cells in ice-cold RIPA buffer containing protease and phosphatase inhibitors for 30 min. The protein concentration of the resulting lysates was then quantified using a Bradford protein assay. The intracellular GSH content was measured using a GSH/GSSG assay kit and performed according to the manufacturer's protocol.

2.11. Cell apoptosis assay

One hundred thousand 4T1 cells per well were seeded into a 6-well plate. After overnight incubation, the cells were treated with TiN (20 μg mL⁻¹ TiN), BTZ (20 nM BTZ), or TiN-BTZ (20 μg mL⁻¹ TiN, 20 nM BTZ) for 8 h. For SDT, the designated cells were exposed to ultrasound (50 kHz, 2 W cm⁻², 50% duty cycle) for 4 min. To measure apoptosis, the cells were harvested after 24-h incubation (including the culture supernatant to collect non-adherent cells), stained using a FITC Annexin V apoptosis detection kit, and immediately analyzed on a BD FACSAria III flow cytometer (USA).

2.12. Caspase-3 activity assay

One hundred thousand 4T1 cells per well were seeded into a 6-well plate. After 24 h-treatments, the cells were washed with PBS and lysed on ice for 1 h in the lysis buffer (20 mM 4-(2-hydroxyethyl)piperazine-1-ethanesulfonic acid (HEPES), 2.5 mM MgCl₂, 10 mM KCl, 1 mM ethylenediaminetetraacetic acid (EDTA), 1 mM dithiothreitol (DTT), 0.5% NP-40, pH 7.4). After centrifugation (14,000 rpm, 15 min) at 4 °C, the supernatants were collected. For the activity assay, 15 μg of protein was incubated with a caspase-3 substrate Ac-DEVD-AMC (10 μg mL⁻¹) in 100 μL of reaction buffer (lysis buffer without NP-40) for 1 h at 37 °C in the dark. The cleavage of the substrate was quantified by measuring the fluorescence intensity (Ex/Em = 355/460 nm) using the microplate reader. The activity was calculated as a percentage relative to the control group.

2.13. Imaging of GFP-ubiquitin puncta

Three thousand U-87 MG cells stably expressing green fluorescent protein-tagged ubiquitin (GFP-ubiquitin) per well were seeded into a 96-well plate. After 12-h incubation, the cells were treated with TiN (20 μg mL⁻¹ TiN), BTZ (20 nM BTZ), TiN-BTZ (20 μg mL⁻¹ TiN, 20 nM BTZ) for 8 h. For SDT treatment, the designated cells were treated with US (50 kHz, 2 W cm⁻², 50% duty cycle) for 4 min. After further incubation for 24 or 48 h, the intracellular GFP-ubiquitin puncta were observed under the fluorescence microscope.

2.14. Proteasome activity assay

One hundred thousand 4T1 cells per well were seeded into a 6-well plate. Following 24 h-treatments, the cells were lysed on ice in a lysis buffer containing 50 mM HEPES, 10 mM NaCl, 1.5 mM KCl, 1 mM EDTA, 1 mM DTT, 1 mM ATP, 250 mM sucrose, and 0.5% NP-40 (pH 7.4). Protein concentration was determined, and the lysates were used for the activity assay. 15 μg of protein lysate was added to 100 μL of assay buffer (50 mM HEPES, 10 mM NaCl, 1.5 mM KCl, 1 mM EDTA, 1 mM DTT, 1 mM ATP, 250 mM sucrose, pH 7.4) containing one of the following substrates at a final concentration of 25 μM: Suc-LLVY-AMC (chymotrypsin-like), Z-LLE-AMC (caspase-like), or Boc-LRR-AMC (trypsin-like). After 30-min incubation at 37 °C, the fluorescence (Ex/Em = 380/460 nm) generated by the cleavage of AMC was measured using the plate reader. The proteasome activity was expressed as a percentage relative to the untreated control.

2.15. Western blotting assay

One hundred thousand 4T1 cells per well were seeded into a 6-well plate. Following the indicated treatments for 24 h, the cells were harvested and washed with PBS. The cells were lysed in ice-cold RIPA buffer containing protease and phosphatase inhibitors for 1 h with intermittent vortex every 15 min to extract total cellular protein. The lysates were subjected to centrifugation (12,000 rpm, 15 min, 4 °C), and the protein concentration of the supernatant was determined using Bradford protein assay. For western blotting, an equal amount of protein (20 μg) from each sample was resolved by SDS-PAGE and subsequently transferred onto a PVDF membrane. The membrane was blocked with 5% non-fat milk in TBST for 1 h at room temperature and then incubated with the appropriate primary antibodies overnight at 4 °C. The following day, the membrane was incubated with the corresponding HRP-conjugated secondary antibodies for 1 h at room temperature and visualized using an ECL detection kit.

2.16. In vitro ROS detection and JC-1 staining

4T1 cells were seeded at 150,000 cells per well in a confocal dish and incubated overnight. The cells were then treated with TiN (20 μg mL⁻¹ TiN), BTZ (20 nM BTZ), BTZ-TiN (20 μg mL⁻¹ TiN, 20 nM BTZ) for 8 h. For SDT treatment, the designated cells were treated with ultrasound (50 kHz, 2 W cm⁻², 50% duty cycle) for 4 min. After 4-h incubation, the cells were washed with PBS and then incubated with 2′, 7′-dichlorofluorescin diacetate (DCFH-DA, 10 μM) in serum-free medium for 20 min at 37 °C. For mitochondrial membrane potential detection, after 24-h incubation, the cells were stained with JC-1 (4 μM) for 20 min. Subsequently, the cells’ nuclei were stained with Hoechst 33342 (100 ng mL⁻¹) for 15 min. After washing with PBS, the cells were immediately imaged using a Leica TCS SP8 confocal microscope (Germany) or a fluorescence microscope.

2.17. Intracellular calcium staining

The 4T1 cells were seeded at 150,000 cells per well in a confocal dish and incubated overnight. The cells were then treated with TiN (20 μg mL⁻¹ TiN), BTZ (20 nM BTZ), or TiN-BTZ (20 μg mL⁻¹ TiN, 20 nM BTZ) for 8 h. For SDT treatment, the designated cells were treated with ultrasound (50 kHz, 2 W cm⁻², 50% duty cycle) for 4 min. After 24-h incubation, cells were rinsed with HBSS. The cells were stained with Fluo-4 AM (4 μM), Mag-Fluo-4 AM (5 μM) [29] together with ER-Tracker Red (1 μM), and Rhod-2 AM (4.5 μM) [30] for 20 min in HBSS buffer containing Pluronic F-127 (0.02%, w/v), respectively. Afterward, the cells were stained with Hoechst 33342 (100 ng mL⁻¹) for 15 min. After washing with HBSS, the cells were observed by the fluorescence microscope and the confocal microscope.

2.18. RNA sequencing study

4T1 cells were seeded at 100,000 cells per well in a 6-well plate and incubated overnight. The cells were then treated with BTZ (20 nM), TiN (20 μg mL⁻¹), TiN-BTZ (20 μg mL⁻¹ TiN, 20 nM BTZ) for 8 h. For SDT treatment, the cells were treated with US (50 kHz, 2 W cm⁻², 50% duty cycle) for 4 min after being treated with TiN or TiN-BTZ. After 24-h incubation, the cells were washed with PBS, RNA was then extracted using TRIzol™ reagent. The RNA samples were then sequenced on the Illumina sequencing platform by Metware Biotechnology Co., Ltd. (Wuhan, China). The DESeq2 analysis results were used to identify genes with significantly differential expression, based on a fold change >1.5 and a P value < 0.05. Enrichment analysis was performed based on the hypergeometric test for Kyoto Encyclopedia of Genes and Genomes (KEGG) and gene ontology (GO) analysis.

2.19. In vivo biodistribution study

To obtain Cy7-conjugated TiN NPs, TiN-NH₂ was first prepared. Specifically, TiN (5 mg) was added into H₂O₂ (2 mL, 30%) and stirred for 30 min to generate hydroxyl groups on the surface. The mixture was then centrifuged, washed, and resuspended in ethanol. Subsequently, the silane coupling agent 3-aminopropyltriethoxysilane (APTES, 5 μL) was added and reacted overnight to coat the surface with a thin layer of amino groups. The TiN-NH₂ and TiN@BSA NPs were then conjugated with the fluorescent dye Cy7. The 4T1 tumor-bearing mice were injected with the Cy7-conjugated TiN NPs and Cy7-conjugated TiN@BSA NPs (20 mg kg⁻¹ TiN) via the tail vein. In vivo fluorescence imaging was performed at 0, 2, 4, 6, 8, and 24 h on an IVIS imaging system (AniView100, BLT, China). At 24 h, the mice were sacrificed, and ex vivo imaging of the major organs and tumors was carried out.

2.20. In vivo pharmacokinetics study

Cy7-loaded TiN@BSA were injected into mice via the tail vein. At 0, 0.08, 0.5, 1, 2, 4, 8, 12, and 24 h post-administration, 0.05 mL of blood was collected from the eye vein. The samples were then diluted four-fold with DI H₂O and fluorescence intensity (Ex = 740 nm, Em = 780 nm) was measured by the microplate reader. A standard Cy7 solution was also added to blood from untreated animals in order to generate a standard curve of fluorescence intensity (F) versus blood concentration (C, μg mL⁻¹), which was used to calculate the Cy7 concentration in treated animals at different time points.

2.21. In vivo antitumor study

The 4T1 tumor-bearing mice were randomly divided into seven groups, with five mice in each group. They were then injected through tail vein with normal saline, BTZ (0.5 mg kg⁻¹), TiN (40 mg kg⁻¹), TiN-BTZ (40 mg kg⁻¹ TiN, 0.5 mg kg⁻¹ BTZ), US (50 kHz, 2 W cm⁻², 50% duty cycle, 8 min), TiN (40 mg kg⁻¹) plus US, and TiN-BTZ (40 mg kg⁻¹ TiN, 0.5 mg kg⁻¹ BTZ) plus US, respectively. The day when the mice received the first injection was defined as day 0. The injection was given every three days for a total of four times. US treatment was performed 24 h after injection. The tumor volume and body weight were monitored every two days. On day 12, all mice were sacrificed, and the tumor tissues were collected. The tissue samples were fixed in paraformaldehyde (PFA, 4%) for 24 h. The fixed tissues were then embedded in paraffin for sectioning and hematoxylin and eosin (H&E) staining. Subsequently, the tumor tissue sections were observed and photographed under a Leica DM2500 fluorescence microscope (Germany). The tumor tissues were also dehydrated with gradient sucrose solution and then embedded in OCT compound. Frozen sections (5 μm) were prepared using a Leica CM3050S cryostat (Germany) and stained with DCFH-DA (10 μM) for 30 min, and the nuclei were counterstained with Hoechst 33342 (10 μg mL⁻¹) for 20 min. The sections were imaged by the confocal microscope. In addition, fresh tumor tissues were also collected, and proteins were extracted for western blotting assay.

2.22. Immunofluorescence assay

For immunofluorescence staining, the frozen sections were first permeabilized with 0.1% Triton X-100 for 15 min and subsequently blocked with 10% goat serum for 1 h at room temperature. The sections were then incubated overnight at 4 °C with the primary antibody (1:100 dilution). Sections were washed and incubated for 1 h at room temperature with an Alexa Fluor 488-conjugated goat anti-rabbit secondary antibody (1:500 dilution). The nuclei were counterstained with Hoechst 33342 (10 μg mL⁻¹) for 20 min, and finally mounted and visualized under the confocal microscope.

2.23. TUNEL assay of tumor tissues

Apoptosis in tumor frozen sections was detected using a terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay kit. Briefly, the frozen tumor sections were first permeabilized with 0.1% Triton X-100 for 15 min at room temperature. After washing with PBS, the sections were incubated with the TUNEL reaction mixture for 1 h at 37 °C in a humidified chamber. Subsequently, the nuclei were counterstained with Hoechst 33342 (10 μg mL⁻¹) for 20 min. The sections were then mounted and visualized under the confocal microscope.

2.24. In vivo biocompatibility study

The blood and major organs were collected from treated mice. Plasma levels of alanine aminotransferase (ALT), aspartate aminotransferase (AST), creatinine (CRE), blood urea nitrogen (BUN), interleukin-6 (IL-6), tumor necrosis factor-α (TNF-α), and serum levels of immunoglobulin G (IgG) and immunoglobulin M (IgM) were measured using the kits. After fixation with 4% PFA for 24 h, the tissue samples were embedded in paraffin wax for section. The sections were stained with H&E and then observed under the microscope.

2.25. Statistical analysis

The data were presented as mean ± standard errors of the mean (S.E.M.). Statistical analysis was performed by using GraphPad Prism 8 (GraphPad Software Inc. USA). The statistical differences were evaluated using one-way analysis of variance (ANOVA) followed by Tukey's post hoc test. A P < 0.05 was considered as statistically significant.

3. Results and discussion

3.1. Synthesis and characterization of TiN NPs

Titanium-based nanostructures exhibit tunable physicochemical properties, catalytic versatility, and good biocompatibility, positioning them as multifaceted platforms for biomedical applications [17]. The synthesis of TiN NPs was schematically illustrated in Fig. 1a according to a previously reported method with a minor modification [24]. The TiO₂ NPs were first prepared by oxidative hydrolysis of titanium oxysulfate in the basic solution. The SEM image (Fig. S1) demonstrates the obtained TiO₂ NPs have a relatively uniform size. The TiN NPs were then obtained through the nitridation treatment of TiO₂ NPs in an ammonia-flowing environment (flow rate: 0.1 L min⁻¹, 800 °C, 5 h). The SEM (Fig. 1b) and TEM (Fig. 1c) images show that the prepared TiN NPs have a regular spherical shape and a narrow size distribution. Fig. 1d–f shows the high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) image of the prepared TiN NPs and the corresponding elemental distribution. The uniform spatial distribution of Ti and N, as revealed by elemental mapping, provides solid evidence for the successful synthesis of TiN NPs.

Fig. 1.

Morphology and composition of TiN and TiN@BSA NPs and sonodynamic performance of TiN NPs. a) Preparation of BSA-coated TiN NPs. b) SEM image of TiN NPs. c) TEM image of TiN NPs. d) HAADF-STEM image of TiN NPs. Corresponding Ti e) and N f) distribution maps, respectively. Hydrodynamic sizes of TiN g) and TiN@BSA h) NPs. i) BSA characterization of the TiN@BSA NPs. The solutions of H₂O, BSA, TiN, and TiN@BSA NPs were separately reacted with reagents in the BCA protein assay kit, and the UV–vis absorption spectra of the supernatants after centrifugation were measured. j) UV–vis absorption spectra of BSA, TiO₂, TiN, and TiN@BSA NPs. k–m) Comparison of the sonodynamic performance of TiO₂ and TiN NPs using DPBF probe. The TiO₂ or TiN NPs were mixed with the DPBF solution, and treated with US irradiation (50 kHz, 2 W cm⁻², 50% duty cycle) for the duration of 0–6 min. UV–vis absorption spectra of the DPBF supernatants k, l) after centrifugation were measured. m) Normalized absorption (A_t/A₀) of the DPBF supernatant after the US irradiation in the presence of TiO₂ or TiN NPs, where A_t and A₀ represent the absorbance at the end and initial time, respectively. n–o) Comparison of the sonodynamic performance of TiO₂ and TiN NPs using TMB probe.

BSA can improve the pharmacokinetic properties of the nanotherapeutics and accumulation in tumors, resulting in enhanced therapeutic efficacies [31]. As the titanium nanostructures have a high binding affinity to biomolecules including serum proteins due to van der Waals interaction, TiN NPs were subsequently coated with BSA by simple stirring. The BSA density was optimized to sufficiently mask the TiN core, conferring “stealth” properties, while effectively presenting the albumin motifs for albumin receptor-mediated tumor targeting. The hydrodynamic diameters of the TiN and TiN@BSA NPs in water determined by the dynamic light scattering (DLS) analysis are 166.5 ± 18.1 nm (polydispersity index (PDI): 0.26 ± 0.24, Fig. 1g) and 141.9 ± 0.1 nm (PDI: 0.19 ± 0.01, Fig. 1h), respectively. The increase in the dispersibility of TiN in water by BSA may cause a decrease in the hydrodynamic size of TiN@BSA. After the coating of BSA, the zeta potential of TiN NPs decreases from −14.1 ± 0.9 mV to −30.9 ± 0.7 mV (Fig. S2). The presence of BSA in the TiN@BSA NPs was also confirmed by the BCA protein assay. A linear relationship between the absorbance and the concentration of BSA is shown in Fig. S3. Compared with the standard BSA solution, the colored complex in the TiN@BSA supernatant was significantly weakened (Fig. 1i). TiN NPs exhibit a typical broad absorption in the UV–vis region as the absorption spectrum shown in Fig. 1j. The TiN@BSA NPs have an additional absorption peak at 278 nm which is due to the presence of BSA (Fig. 1j).

3.2. Sonodynamic performance of TiN NPs

The typical sonosensitizer TiO₂ exhibits limited ROS generation efficiency under US irradiation due to its wide bandgap and rapid electron-hole (e⁻-h⁺) recombination dynamics. In contrast, transition metal nitrides have emerged as promising alternatives, demonstrating exceptional electrocatalytic performance due to their unique physicochemical properties, including metallic conductivity and tunable d-band electronic structures [32]. The biomedical applications of TiN including photothermal therapy (PTT) in the second near-infrared (NIR-II) window [33], and in vivo bioimaging [34] have emerged in recent years. The inherent surface-deficient structure of TiN facilitates enhanced e⁻-h⁺ pair separation efficiency [19], the sonosensitizing efficiency of TiN was therefore assumed to be better than traditional TiO₂ nanostructures under US irradiation. To validate the sonodynamic performance of TiN NPs, DPBF was employed to detect ¹O₂. In the presence of 40 μg mL⁻¹ TiO₂ (Fig. 1k) or TiN NPs (Fig. 1l), US induces a progressive decrease in DPBF absorption, with increasing US exposure time. Under the identical US irradiation condition, TiN NPs induce a faster DPBF oxidation rate than the TiO₂ NPs (Fig. 1m), indicating that TiN has superior sonosensitizing activity. In addition, ^•OH production was quantitatively assessed using TMB as a colorimetric probe. The UV–vis spectroscopy reveals time-dependent ^•OH generation by TiO₂ under US irradiation, as evidenced by increasing TMB oxidation (Fig. 1n). Under US irradiation, TiN also induces more TMB oxidation compared to TiO₂ (Fig. 1o), indicating that TiN can produce more ^•OH generation under the SDT condition.

3.3. BTZ loading and controlled release

BTZ (chemical structure shown in Fig. S4) was the first proteasome inhibitor approved by the U.S. FDA for treating multiple myeloma. However, BTZ exhibits suboptimal pharmacokinetic properties, characterized by limited tissue biodistribution and rapid systemic clearance, and long-term treatment with BTZ in clinical practice is also hampered by its several severe side effects and acquired drug resistance. Nanocarrier-based delivery systems are engineered to enhance the preferential accumulation of BTZ at the tumor site by exploiting the enhanced permeability and retention (EPR) effect.

Especially, engineering nanoplatforms incorporating dynamic chemical interactions [35] that enable on-demand drug release can potentially lead to enhanced tumor accumulation, achieve precisely controlled drug release, and mitigate off-target toxicity. The dynamic covalent chemistry between boronic acid and catechol [36], diol [37], or keto-enol moieties [38], particularly through reversible condensation reactions, enables the formation of pH-responsive boronate ester linkages with tunable stability.

Different polyphenol-derived compounds have varying binding affinities to boronic acid [39]. In this boronate ester bond-based delivery strategy, a polyphenol with moderate affinity to boronic acid is essential to ensure the optimal loading and release efficiency of BTZ. The polyphenols with strong interactions with BTZ leading to too stable bonds should be avoided, as they usually compromise the anticancer effect of BTZ [40,41]. In the preliminary study, various polyphenol-derived compounds (chemical structures shown in Table 1) were incubated with BTZ, and then the cytotoxicity of polyphenol-BTZ mixtures was evaluated. After screening, we found that most polyphenols including caffeic acid (CFA), gallic acid (GA), dopamine (DA), QUE, EGCG, and tannic acid (TA) inhibit the cytotoxicity of BTZ (Table 1 and Fig. S5). These results also confirm that the polyphenols with catechol groups can chelate with the boronic acid group of BTZ to form a thermodynamically stable five-membered boronate ester ring, thereby neutralizing the cytotoxicity of BTZ.

Table 1.

Polyphenol-BTZ interaction screening and prediction of binding affinity.

Compound	Structure	Reduction of BTZ's cytotoxicity after co-incubation	BTZ binding affinity prediction
CFA		Significant	+++
GA		Significant	+++
EGCG		Significant	+++
DA		Significant	+++
QUE		Significant	+++
TA		Significant	+++
CUR		Significant	+++
AA		Significant	+++
ED		Slight	++
RHE		Slight	++
QZ		Slight	++
CAS		No/minor	+/−
PP		No/minor	+/−
CA		No/minor	+/−
AR		No/minor	+/−

Remark: The polyphenols with strong interactions with BTZ leading to too stable bonds should be avoided, as they usually compromise the anticancer effect of BTZ. “+++“: strongest binding; “++“: stronger binding; “+“: weak binding; “–“: no binding.

Considering that the keto-enol moiety can dynamically bind with BTZ [38,42], various keto-enol derivatives (chemical structures shown in Table 1) were further incubated with BTZ and the cytotoxicity of the mixtures was evaluated. Curcumin (CUR) was also found to inhibit the cytotoxicity of BTZ in this study (Fig. S6 and Fig. S7). The anthraquinone compounds with keto-enol moieties including emodin (ED), rhein (RHE), and anthrarufin (AA) were found to inhibit the cytotoxicity of BTZ to varying degrees. However, the anthraquinone compounds casanthranol (CAS), 1,4-dihydroxyanthraquinone (quinizarin, QZ), purpurin (PP), carminic acid (CA), and alizarin red (AR) have no significant effect on the cytotoxicity of BTZ (Fig. S6 and Fig. S7). The above results suggest that the attenuation of BTZ cytotoxicity by these compounds is dictated by their molecular structure. This effect depends on the presence of a spatially accessible and geometrically favorable chelating site. Compared to CUR, the anthraquinone compounds with 1–2 keto-enol moieties possessing planar and rigid structures show weaker interaction with the boronic acid group of BTZ. Interestingly, the steric hindrance between the two keto-enol moieties is vital for the affinity to the boronic acid. The 1,5-dihydroxyanthraquinone structure of AA, for instance, provides two highly effective peri-keto-enol binding sites. Unlike AA, the two keto-enol moieties of QZ are in the para position, and the steric hindrance may prevent the formation of a stable chelate ring. Compared with QZ, PP has an additional hydroxyl group at position 2. This additional hydroxyl group may increase steric hindrance around the chelate site, making the resulting boronate complex less stable than AA. Compounds such as CAS, CA, and AR have high water solubility due to the introduction of bulky substituents such as glucose or sulfonic acid group. This may also increase the steric hindrance to the keto-enol moieties, weakening their binding affinity to BTZ. Therefore, an unhindered catechol or keto-enol moiety capable of forming a five- or six-membered ring may be a key structural determinant of the ability of these phenolic compounds to antagonize BTZ activity.

In order to confirm the binding between keto-enol moieties on the anthraquinone compounds and BTZ, UV–vis absorption spectra of anthraquinone-BTZ conjugates with different molar ratios were subsequently measured. CA, AA, QZ, PP, and AR have a characteristic absorption peak in the region of 400–500 nm. Compared with each free anthraquinone compound, the absorption peak height of AA-BTZ, QZ-BTZ, PP-BTZ, or AR-BTZ conjugate is altered to varying degrees, though no obvious shift of the absorption peak was observed (Fig. S8). However, the characteristic absorption peak of CA-BTZ is dramatically red-shifted compared with that of CA (Fig. 2a), indicating the interaction between CA and BTZ. Considering the carboxyl group on the CA molecule can be utilized for the conjugation to BSA, CA was selected for further study.

Fig. 2.

BTZ loading and release. a) UV–vis absorption spectra of CA-BTZ at various BTZ/CA molar ratios. b) Raman spectra of CA, BTZ, and CA-BTZ. c) Mass spectrum of CA-BTZ detected in anion MS screening mode. d) Schematic diagram of BTZ loading and cytotoxicity evaluation. BTZ (100 μL, 150 μM) was incubated with TiN@BSA or TiN@BSA-CA (400 μL, 50 mg mL⁻¹ TiN) for 12 h. After centrifugation (12,000 rpm, 10 min), the supernatant and pellet were collected, and the precipitated drug-loaded NPs were resuspended in H₂O (500 μL). In addition, a mixture of BTZ (100 μL, 150 μM) and TiN (400 μL, 50 mg mL⁻¹ TiN) without centrifugation was prepared for subsequent cell viability study. e) Viability of the cells after incubation with the supernatant or drug-loaded NP suspension. The supernatant or drug-loaded NP suspension (d) was diluted 1000 times with culture medium for treating cells (30 nM BTZ and 40 μg mL⁻¹ TiN for the cells treated with the mixture of TiN and BTZ without centrifugation). After 48-h incubation, the cell viability was determined by MTT assay. f) LC-MS chromatograms of initial BTZ solution and supernatant after drug loading. g) Time- and pH-dependent BTZ release from TiN-BTZ NPs in PBS (pH 7.4) and citrate buffer (pH 4.5) at 37 °C. h) US-triggered BTZ release from TiN-BTZ NPs in PBS (pH 7.4) or citrate buffer (pH 4.5). Data represent the mean ± S.E.M., n = 3, ∗∗∗P < 0.001.

The binding between BTZ and CA was investigated via Raman spectroscopy (Fig. 2b). The Raman spectrum of BTZ exhibits two distinct bands at 881 and 1003 cm⁻¹, characteristic of the symmetric B–O stretching mode of the boronic acid group [–B(OH)₂] [43,44]. These characteristic peaks are reduced in the CA-BTZ spectrum, providing direct evidence of reaction to the boronic acid group. Meanwhile, the formation of the new boronate ester linkage is evidenced by the emergence of a distinct band at 1310 cm⁻¹, which is assigned to the coupled B–O and C–O asymmetric stretching vibrations within the newly formed cyclic ester [45]. The spectroscopic evidence confirms the formation of a cyclic boronate ester product between BTZ and CA. Subsequently, this conjugation process was also investigated using ¹H NMR spectroscopy in deuterated methanol (CD₃OD) (Fig. S9). Compared to BTZ before conjugation, the hydrogen peaks at δ 0.74–1.44 (d, H1′–4′ in the methyl group of BTZ) and δ 8.69–9.18 (s or d, H10′–12′ on the aromatic ring of BTZ) change significantly when it is conjugated to CA (Fig. S9a). This result indicates that BTZ reacts with CA to form a new compound. In addition, ¹H NMR spectrum of CA reveals that the chemical shifts of hydrogen peaks at δ 3.42–3.50 (m, H1′–4′ on the glucose moiety of CA) do not change after it reacts with BTZ (Fig. S9b), suggesting that vicinal diol groups in the glucose moiety of CA do not participate in the formation of the boronate ester bond. In order to determine whethercatechol or keto-enol is involved in the reaction of CA-BTZ, mass spectrometry analysis was performed. The mass spectrometry of BTZ and CA are shown in Fig. S10. The m/z peaks appearing at 365.18, 383.19, and 419.17 are assigned to [M − H₂O − H]⁻, [M − H]⁻, [M + Cl] of BTZ. The major m/z peak appearing at 491.09 is assigned to [M − H]⁻ of CA. The major m/z peaks appearing at 839.27 and 857.28 are assigned to [M − H₂O − H]⁻ and [M − H]⁻ of CA-BTZ in Fig. 2c, which is attributed to the dynamic binding of keto-enol moiety functionalities on CA with BTZ. These above results indicate that CA can bind with BTZ reversibly, suggesting that CA can be utilized for the delivery of BTZ. BSA was first functionalized with CA through EDC/NHS coupling chemistry. The obtained BSA-CA was then coated onto the TiN NPs to prepare TiN@BSA-CA NPs. To evaluate its potential for systemic administration, the colloidal stability of the as-prepared NPs was systematically monitored. The particle size and PDI of TiN@BSA-CA NPs remain remarkably stable in both PBS and serum-containing medium over a 14-day period (Fig. S11). In contrast, the particle size and PDI of unmodified TiN NPs increase slightly in serum, indicating that the BSA coating significantly enhances the physiological stability of the NPs. TiN@BSA-CA NPs were subsequently incubated with BTZ, and the BTZ-loaded TiN@BSA-CA NPs denoted as TiN-BTZ were finally obtained after centrifugation. The hydrodynamic diameter of the TiN-BTZ NPs is 155.0 ± 20.9 nm (Fig. S12).

In order to verify that BTZ was successfully loaded, the TiN@BSA-CA or TiN@BSA NPs were incubated with BTZ, and after centrifugation, the cytotoxicity of the precipitate drug-loaded NPs and the supernatant was evaluated and compared with the mixture of TiN and BTZ without centrifugation (Fig. 2d). Cell viability was determined using standard MTT assay. The pellet BTZ-loaded TiN NPs functionalized with CA have equal cytotoxicity as the standard BTZ sample, but the supernatant after drug loading is almost non-cytotoxic. In contrast, the pellet obtained after centrifugation of the mixture of TiN@BSA and BTZ shows no obvious cytotoxicity, while the supernatant shows toxic effect as the standard BTZ sample (Fig. 2e and Fig. S13). These results also strongly prove the successful loading of BTZ on the TiN@BSA-CA NPs. LC-MS analysis (Fig. S14) reveals this TiN-BTZ formulation achieves a drug loading capacity (LC) of 27.5 ± 2.1 wt% and an encapsulation efficiency (EE) of 88.7 ± 2.1% (Fig. 2f). The tumor microenvironment (TME) plays a crucial role in cancer progression, and pH-sensitive NPs release drugs within the acidic tumor microenvironment [46]. The boronate ester bonds are cleavable under the slightly acidic endosomes/lysosomes (pH 4.5–5.5) and the tumor interstitial space. The BTZ release behavior was studied in both neutral phosphate-buffered saline (PBS, pH 7.4) and citrate buffer (pH 4.5, mimicking endosome/lysosome microenvironment) [47]. The pH-responsive drug release profile of TiN-BTZ is evident shown in Fig. 2g. After 48-h incubation, 91.6 ± 6.8% of BTZ is released in the acidic citrate buffer, while 59.0 ± 2.6% of BTZ is released in the PBS.

In addition, the SDT-based nanoplatform has the potential to trigger local drug release and improve the therapeutic effect [48]. Although the acidic environment of the tumors and lysosomes can initiate the cleavage of the boronate ester bonds, this chemical hydrolysis is typically a sustained and slow process. However, for the synergistic “ER stress amplification” strategy to be effective, a high intracellular concentration of BTZ is required precisely during the SDT process. The ultrasound responsiveness of the system provides this critical “on-demand” capability. The acoustic cavitation generated by US irradiation acts as a physical trigger to accelerate drug release and disrupt the nanocarrier structure, resulting in on-demand release of BTZ. This spatiotemporal control ensures that the highest drug release coincides with the burst of ROS generation, thereby maximizing the synergistic damage to cancer cells while minimizing premature leakage in normal tissues. Therefore, controlled BTZ release behavior from TiN-BTZ was further investigated under both passive diffusion condition and US-triggered activation. The US can promote the release of BTZ from TiN-BTZ in both neutral and acidic environments. In PBS and citrate buffer, the release rates of BTZ are only 14.3 ± 3.1% and 24.5 ± 0.7%, respectively. However, under the condition of US irradiation, the drug release increases to 44.6 ± 1.3 % and 59.0 ± 2.6 %, respectively (Fig. 2h). This result suggests in addition to the pH-responsive activation, the reversible boronate ester bond is also responsive to US. Local US can further trigger the BTZ release in the target tumors.

3.4. Cytotoxicity induced by combination therapy of BTZ and SDT in 4T1 cells

Massive ROS generated by SDT can induce severe damage of macromolecules, triggering tumor apoptosis and necrosis. SDT has emerged as a promising strategy to potentiate the therapeutic effects of chemotherapy. As monotherapy with proteasome inhibitors for treating solid tumors shows inadequate efficacy, combination therapy with proteasome inhibitors and SDT was considered as an alternative strategy to achieve tumor eradication more effectively, leading to significantly improved synergistic therapeutic effects with reduced systemic toxicity. The 4T1 murine breast cancer cell line was selected as a representative model of breast cancer due to its high malignancy. First, its rapid proliferation rate correlates with high ER stress, rendering these cells hypersensitive to the proteotoxic stress amplified by the TiN-mediated SDT and BTZ combination. Second, as a syngeneic model, it possesses an intact immune environment, which is essential for evaluating SDT-triggered ICD. Herein, we systematically investigated the synergistic anticancer efficacy of the combination of chemotherapy and SDT using TiN-BTZ in 4T1 cells.

The TiN NPs exhibit good biocompatibility in both 4T1 and RAW 264.7 cells, and cell viabilities remain about 90% even at the high concentration of 200 μg mL⁻¹ (Fig. S15a and b). CA and CA-BSA also exhibit good biocompatibility and have no obvious effect on the viability of 4T1 cells (Fig. S15c). Free BTZ (20 nM) alone reduces the viability of 4T1 cells to 70.4 ± 3.1%. The cytotoxicity of BTZ can be enhanced by the TiN-mediated delivery. BTZ-loaded TiN@BSA-CA NPs at 20 and 40 μg mL⁻¹ TiN further decrease the viability to 64.8 ± 3.9% and 56.9 ± 3.5%, respectively (Fig. S15d).

Due to the high SDT activity of TiN NPs as shown above, the effect of SDT on cell viability was then studied. US irradiation in the absence of TiN has no effect on the viability of 4T1 cells. However, in the presence of TiN NPs, TiN-mediated SDT reduces the cell viability time-dependently (Fig. S15e). The ROS scavenger NAC can antagonize SDT-induced cell death (Fig. S15e), providing the direct evidence that ROS is the dominant executor to induce cell death during SDT.

The anticancer effect of SDT combined with BTZ was first investigated in the 4T1 cells. In the presence of 40 μg mL⁻¹ TiN, 4-min US reduces the cell viability to 66.5 ± 6.2%. This SDT parameter was chosen for the combined therapy of BTZ and TiN-mediated SDT (TiN-BTZ + US). Treatment with TiN-BTZ NPs (20 nM BTZ) reduces the viability of 4T1 cells to 56.4 ± 4.7%, while the combined therapy of BTZ and TiN-mediated SDT dramatically reduces the viability to 30.5 ± 6.7% (Fig. 3a). This result indicates that combined treatment shows enhanced anticancer effect compared to the monotherapy with either BTZ or SDT.

Fig. 3.

Cytotoxicity induced by the combination therapy of BTZ and SDT. a) Viabilities of 4T1 cells after 48-h treatment with PBS, BTZ (20 nM), TiN (40 μg mL⁻¹), or TiN-BTZ (20 nM BTZ, 40 μg mL⁻¹ TiN) in the presence or absence of US irradiation (50 kHz, 2 W cm⁻², 50% duty cycle, 4 min). b) Effect of antioxidant NAC (2 mM) on the cytotoxicity induced by the various treatments. c) Intracellular GSH content, d) caspase-3 activity, and e) apoptosis level after the various treatments for 24 h. Data represent the mean ± S.E.M., n = 3, ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001.

In the presence of NAC, the viability of the cells treated with TiN-BTZ NPs (40 μg mL⁻¹ TiN, 20 nM BTZ) recovers from 63.7 ± 4.8% to 75.0 ± 2.8% (Fig. 3b). Furthermore, under the US treatment with the TiN-BTZ NPs, NAC pre-treatment recovers the cell viability from 43.9 ± 6.5% to 63.5 ± 6.2% (Fig. 3b). This result indicates that the enhanced anticancer effect by the combination of BTZ and SDT is mainly due to the elevated oxidative stress. It was reported that elevated intracellular GSH level can neutralize the cytotoxic effect of BTZ [49]. The combination treatment induces the most pronounced depletion of intracellular GSH pools (Fig. 3c), with the GSH/GSSG ratio reaching the lowest level (Fig. S16) among all treatment groups.

Cell apoptosis was assessed by the measurement of caspase-3 activity and flow cytometry. During apoptosis, mitochondrial cytochrome c release initiates apoptosome formation through complexation with apoptotic protease-activating factor 1 (Apaf-1) and procaspase-9, which subsequently activate the effector caspases-3/7 via proteolytic cleavage, ultimately executing the intrinsic apoptotic pathway. Caspase-3 enzymatic activity was quantitatively measured using the fluorogenic substrate Ac-DEVD-AMC [50]. Neither TiN NPs (40 μg mL⁻¹) nor BTZ (20 nM) activates caspase-3 activity after the incubation for 24 h. Compared with free BTZ, TiN-BTZ NPs slightly elevate the activity of caspase-3. Moreover, the combination of BTZ and TiN-mediated SDT dramatically enhances the caspase-3 activity (Fig. 3d). These results indicate that the combined treatment with BTZ and SDT synergistically induces cell apoptosis. The apoptosis rates in TiN-BTZ (20 nM BTZ) and TiN-mediated SDT (40 μg mL⁻¹ TiN) groups are 25.0 ± 0.8% and 26.1 ± 2.1%, respectively. Importantly, the combination treatment with BTZ and TiN-mediated SDT induces a significantly higher proportion (56.0 ± 1.3%) of cells undergoing apoptosis compared with each monotherapy (Fig. 3e and Fig. S17). The cell viability was also evaluated by calcein-AM/PI staining, and the live cells were stained with green fluorescence by calcein-AM, whereas dead cells were stained with red fluorescence by PI. As shown in Fig. S18, treatment with BTZ alone induces negligible cell death, and slight cell death was observed in the TiN-BTZ group, suggesting that TiN-based delivery system enhances the efficacy of BTZ. TiN-mediated SDT also induces a noticeable cell death. A dramatic shift from live to dead cells is observed in the combination of BTZ and TiN-mediated SDT group. The majority of cells were stained with bright red fluorescence by PI, and only a sparse population of living cells were green-stained. This result provides direct visual evidence that the combination of BTZ with TiN-mediated SDT irradiation leads to overwhelming cell death.

3.5. Combination therapy inhibits cellular proteolytic activity

UPS is a major route for more than 80% intracellular proteins, selectively eliminating aberrant (damaged/unfolded) and short-lived proteins. This process involves two stages including ubiquitination of target substrates and subsequent ATP-dependent degradation by the proteasomes [51]. UPS regulates numerous biological processes such as DNA repair, cell survival, and apoptosis [52].

Autophagy is another important route for the degradation of intracellular proteins [53], and protective autophagy can occur after SDT. Recently, it was reported that autophagy inhibition is able to enhance the anticancer effect of SDT [54]. However, UPS regulation has not been utilized to improve the effect of SDT, although UPS was reported to perform like autophagy to protect the cancer cells from the damage to photodynamic therapy [55]. Since UPS is the vital route for the elimination of ROS-induced damaged proteins under SDT, a synergistic interaction between proteasome inhibitors and SDT was hypothesized. Effect of the combination of BTZ and SDT on the intracellular ubiquitinated protein level was first studied in the U-87 MG cells stably expressing GFP-ubiquitin. As anticipated, BTZ (20 nM) induces pronounced accumulation of GFP-ubiquitin-positive puncta (Fig. 4a) and high expression of ubiquitinated proteins (Fig. 4b) at 24 h. Exposure to the TiN NPs (40 μg mL⁻¹) has minimal effect on the intracellular ubiquitinated protein level, regardless of whether US is used or not. TiN-BTZ induces significant accumulation of both GFP-ubiquitin puncta (Fig. 4a) and total ubiquitinated proteins (Fig. 4b and Fig. S19a). More importantly, the level of ubiquitinated proteins is further increased and even maintained for a longer period of time when US is simultaneously conducted. While the accumulation of GFP-ubiquitin puncta is reduced after 48 h of treatment with BTZ (20 nM) alone, treatment with TiN-BTZ or TiN-BTZ plus US still induces a significant accumulation of GFP-ubiquitin puncta in the cells (Fig. 4a).

Fig. 4.

ER stress and terminal UPR activation induced by the combination of BTZ and SDT. a) Imaging of GFP-ubiquitin puncta in U-87 MG cells after the various treatments for 24 or 48 h. b) Expression of ubiquitinated proteins in the 4T1 cells after the treatments for 24 h. c) SDT interacts synergistically with BTZ to inhibit caspase-like, chymotrypsin-like, and trypsin-like proteasomal activities. The proteasomal activities of the cell lysates were determined by using fluorogenic substrates after 24-h treatments. d) Intracellular ROS staining by DCFH-DA, and immunofluorescence staining of e) γ-H2A.X and f) CHOP after the 24-h treatments. Nuclei were stained with Hoechst 33342. g, h) Expression of BiP, γ-H2A.X, p-IRE1α, XBP-1s, p-JNK, caspase-12, ATF6, p-eIF2α, and ATF4 proteins after 24-h treatments. The 4T1 cells were treated with PBS, BTZ (20 nM), TiN (40 μg mL⁻¹), or TiN-BTZ NPs (20 nM BTZ, 40 μg mL⁻¹ TiN) in the presence or absence of US irradiation (50 kHz, 2 W cm⁻², 50% duty cycle, 4 min) for 24 h. Data represent the mean ± S.E.M., n = 3, ∗∗P < 0.01, ∗∗∗P < 0.001.

The 26S proteasome is a multi-subunit protease complex comprising a cylindrical 20S core particle capped by one 19S regulatory particle. The 20S proteasome core contains three different substrate preferences of catalytic β subunits (β1, β2, and β5), exerting caspase-like, trypsin-like, and chymotrypsin-like proteolytic activities, respectively. The current proteasome inhibitors inhibit the three proteolytically active β-ring subunits of the proteasome with varying degrees of efficacy, being most potent towards β5-subunit, less potent for β1-subunit and with little activity towards β2-subunit [56]. As shown in Fig. 4c, BTZ (20 nM) significantly inhibits intracellular chymotrypsin-like and caspase-like activities of the proteasomes, while it does not affect the trypsin-like activity.

It was reported that inhibiting the β2 site of the proteasome could strongly sensitize the myeloma and solid tumor cells to proteasome inhibitors. In our previous study, we found that TiO₂ sensitizes the anticancer effect of BTZ through the selective inhibition of trypsin-like proteolytic activity [57]. Therefore, nanotechnology which additionally inhibits proteasome activity shows great potential for sensitization of proteasome inhibitors. Similar to TiO₂, TiN NPs (40 μg mL⁻¹) selectively suppress the trypsin-like activity, without affecting the other two catalytic activities (Fig. 4c). Moreover, SDT with the TiN NPs further inhibits the chymotrypsin-like and trypsin-like activities of the proteasomes. We hypothesized that the concomitant inhibition of the β2 subunit in 4T1 cells by TiN-mediated SDT would synergistically enhance the proteasomal suppression achieved by BTZ. When SDT is combined with BTZ, the combination treatment results in a more pronounced inhibition of all three proteolytic activities of the proteasomes compared with either monotherapy. Therefore, the above results suggest that TiN-mediated SDT has unique proteasome inhibitory activity, especially the trypsin-like (β2) activity that is insensitive to BTZ and thus has a synergistic inhibitory effect with BTZ on the proteasome activity.

3.6. Combination therapy causes ER stress and terminal UPR overactivation

ER is the primary site for protein synthesis, post-translational modification, and Ca²⁺ storage in the eukaryotes. Pathological accumulation of unfolded and misfolded proteins in the ER lumen induces proteotoxic stress, known as ER stress. US-triggered SDT generates a large amount of ROS, leading to an increase in the burden of unfolded proteins, thereby disrupting ER homeostasis and inducing ER stress [58]. The intracellular ROS generated were detected by DCFH-DA. As shown in Fig. 4d and Fig. S19b, compared with the control group, TiN-mediated SDT enhances the intracellular ROS level. As reported, the proteasome inhibitor BTZ mildly elevates the intracellular level of ROS. The combined treatment with BTZ and TiN-mediated SDT (20 nM BTZ, 40 μg mL⁻¹ TiN) results in a synergistic amplification of intracellular ROS production. ROS-mediated oxidative stress usually causes DNA damage [59], and γ-H2A.X is a sensitive molecular marker of DNA double-strand breaks [60]. After the treatment, immunofluorescence staining of γ-H2A.X was performed to assess DNA damage (Fig. 4e and Fig. S19c). Treatment with TiN NPs (40 μg mL⁻¹) has no effect on the expression of γ-H2A.X, while TiN-mediated SDT significantly elevates the expression of γ-H2A.X. Notably, treatment with TiN-BTZ (40 μg mL⁻¹ TiN, 20 nM BTZ) induces more γ-H2A.X expression than BTZ (20 nM) monotherapy. Among these treatment groups, the combination of BTZ and TiN-mediated SDT induces the highest γ-H2A.X expression in the nuclei (Fig. 4e and Fig. S19c), indicating severe DNA damage. Western blotting analysis of γ-H2A.X expression confirms this synergistic effect (Fig. 4g and Fig. S19d). These results are consistent with the above cell viability data. Therefore, ROS generated by the combination of BTZ and SDT results in a higher level of ROS in the cells than BTZ or TiN monotherapy, causing the severest DNA damage and finally cell apoptosis.

The misfolded proteins are first linked with ubiquitin molecules, which are then recognized and degraded by the 26S proteasome ultimately [5]. Inhibition of the proteasome by BTZ leads to excessive accumulation of unfolded proteins, causing ER stress, such as the accumulation of ROS. The adaptive UPR mitigates proteotoxic stress within the ER by attenuating mRNA translation, enhancing protein folding, and promoting clearance of terminally misfolded proteins through the proteasome-mediated ER-associated degradation (ERAD) pathway. UPR can be initiated by the activation of three ER transmembrane proteins including PERK, IRE1α, and ATF6 [1]. A central feature of this response is the marked upregulation of the ER chaperone immunoglobulin heavy chain-binding protein (BiP), which identifies and translocates terminally misfolded proteins to the cytoplasm. These substrates are tagged by ubiquitin for destruction by the proteasome. With the proteasome inhibited, the clearance pathway remains obstructed. However, when ER stress is too severe or prolonged, an excessive activation of UPR (terminal UPR) can also redirect signaling from cytoprotective adaptation to cell death [2].

The effect of the combination treatment on the three UPR signaling pathways was further studied. The expression of BiP protein is slightly increased by BTZ or TiN-mediated SDT treatment, while it is significantly increased after the combined treatment with BTZ and SDT (Fig. 4g and Fig. S19e). Altered UPR activation status can cause tumor cells to become resistant to proteasome inhibitors. For example, elimination of IRE1α has been demonstrated to be sufficient to induce BTZ resistance [61]. In this study, the combination therapy with BTZ and SDT using TiN-BTZ significantly increases the protein level of phosphorylated IRE1α (Fig. 4h and Fig. S19f), suggesting that SDT can sensitize the BTZ treatment via IRE1α activation and induction of a terminal UPR. During this ER stress, IRE1α activation leads to XBP-1 activation. Simultaneously, IRE1α can also activate the JNK pathway potentially leading to the downstream pro-apoptotic signals [62]. Western blotting analysis indicates that XBP-1s expression (Fig. 4h and Fig. S19g) is not altered, but phosphorylated JNK expression (Fig. 4h and Fig. S19h) is significantly elevated after the combination treatment. In addition, IRE1α can bind tumor-necrosis factor receptor-associated factor 2 (TRAF2) to form a complex can activate caspase-12, which serves as the executioner of ER stress-induced apoptosis [63]. We found that combination treatment increases the level of cleaved caspase-12 (Fig. 4h and Fig. S19i). ATF6 transports from the ER to the Golgi, upregulating ER chaperone genes and ERAD components [64]. Similarly, the ATF6 protein level is significantly increased after the combination treatment (Fig. 4h and Fig. S19j). Moreover, both phosphorylated eIF2α and ATF4 which are the markers of the arms of the ER stress response are upregulated in the cells after the combination treatment than either monotherapy alone (Fig. 4h and Fig. S19k–l).

All these three UPR branches can induce the upregulation of CHOP [65], and a pro-apoptotic pathway is triggered via CHOP upon excessive UPR activation. Immunofluorescence staining result indicates that SDT can induce the expression of CHOP in the nuclei of the 4T1 cells (Fig. 4f and Fig. S19m). When the cells were treated with a combination of BTZ and SDT, a significant increase in CHOP expression was observed in the nuclei. The above results suggest that SDT can sensitize the BTZ treatment through excessive activation of terminal UPR, inducing cell apoptosis.

3.7. Calcium overload induced by the combination treatment

Mitochondria are also the downstream target organelles for ER stress and are involved in cell proliferation and death through regulation of Ca²⁺ and ROS homeostasis [66]. In order to elucidate the mechanism for the cell death induced by the combined treatment of BTZ and SDT, intracellular Ca²⁺ level and mitochondrial function were then studied.

Proteasome inhibition and ROS burst generated by SDT lead to the accumulation of misfolded proteins in the ER, triggering ER stress and UPR. This not only promotes the Ca²⁺release from ER but also leads to mitochondrial Ca²⁺ overload, thereby generating more mitochondrial ROS and exacerbating ER stress [67]. The effect of combination therapy on intracellular Ca²⁺ was then studied by fluorescence imaging. Fluo-4 AM was used to monitor Ca²⁺ level in the cytoplasm. As visualized by Fluo-4 AM staining in Fig. 5a and Fig. S21a, the control cells show minimal green fluorescence, representing a low basal level of cytosolic Ca²⁺. Conversely, the combination of TiN-BTZ and US treatment induces a dramatic and widespread increase in green fluorescence intensity, demonstrating a substantial elevation in cytosolic Ca²⁺ concentration. To identify the source of this cytosolic Ca²⁺ surge, Mag-Fluo-4 AM was used to stain the Ca²⁺ within the ER which is the primary intracellular store of Ca²⁺ [68]. ER was stained with red fluorescence by ER-Tracker Red. As shown in Fig. 5b and Fig. S20a, the control cells display calcium signal with bright green fluorescence which is overlapped with that of ER, indicating the high Ca²⁺ content in the ER. This fluorescence signal of Ca²⁺ is significantly attenuated after the combined treatment with TiN-BTZ and SDT, indicating significant depletion of Ca²⁺ in the ER stores due to ER stress (Fig. 5b, Fig. S20a, and Fig. S21b). Given the increase in cytosolic Ca²⁺, a mitochondria-specific fluorescent probe Rhod-2 AM was used to stain the Ca²⁺ to confirm the Ca²⁺ enters the mitochondria from the ER [69]. As shown in Fig. 5c, Fig. S20b, and Fig. S21c, the control cells show a negligible Rhod-2 AM signal. However, a robust increase in red fluorescence in the mitochondria was observed in the cells after the combination treatment, indicating a massive accumulation of Ca²⁺ within the mitochondrial matrix. Mitochondria can act as a temporary buffer, sequestering excess cytosolic Ca²⁺ to prevent cell damage. However, the massive and sustained release of Ca²⁺ from the ER appears to overwhelm this buffering capacity, leading to mitochondrial calcium overload.

Fig. 5.

Effect of the combination of BTZ and SDT on MMP and intracellular calcium level. Imaging of a) Ca²⁺ in whole cells by Fluo-4 AM (green), b) Ca²⁺ in ER by Mag-Fluo-4 AM (green), and c) Ca²⁺ in mitochondria by Rhod-2 AM (green). ER was stained with ER-Tracker Red. d) MMP assay by JC-1 (green and red) staining. Red fluorescence indicates JC-1 aggregates in the mitochondrial matrix (high MMP), while green fluorescence indicates JC-1 monomers in the cytoplasm (low or disrupted MMP). The nuclei were stained with Hoechst 33342 (blue). The 4T1 cells were treated with PBS, BTZ (20 nM), TiN (40 μg mL⁻¹), or TiN-BTZ NPs (20 nM BTZ, 40 μg mL⁻¹ TiN) in the presence or absence of US irradiation (50 kHz, 2 W cm⁻², 50% duty cycle, 4 min) for 24 h.

Excessive influx of Ca²⁺ into the mitochondrial matrix can have serious consequences. It directly leads to reduced mitochondrial membrane potential [70]. The mitochondrial membrane potential (MMP) was assessed using the potential-dependent fluorescent probe JC-1. JC-1 probe distribution shifts from J-aggregates (red, polarized mitochondria) to monomers (green, depolarized) when the mitochondrial membrane potential is lowered [13]. As shown in Fig. 5d, the red fluorescence signal is reduced by either TiN-BTZ (40 μg mL⁻¹ TiN, 20 nM BTZ) or TiN (40 μg mL⁻¹)-mediated SDT treatment, while the green fluorescence signal is enhanced, demonstrating mitochondrial dysfunction. In contrast, compared to either monotherapy, treatment with TiN-BTZ plus SDT induces a more significant shift from red to green fluorescence. The combination therapy also induces maximal mitochondrial depolarization as the dramatic accumulation of JC-1 monomer with green fluorescence. Beyond mitochondrial depolarization, the impairment of mitochondrial integrity and cellular homeostasis often leads to the massive release of DAMPs, such as ATP [71]. As a crucial biochemical indicator of immunogenic cell death, the extracellular ATP levels were further quantified. Consistent with the MMP results, treatment with TiN-BTZ plus SDT triggers a significant elevation in ATP release compared to the control and mono-treatment groups (Fig. S21d), indicating severe irreversible cellular damage and metabolic exhaustion resulting from the synergistic effect of proteasome inhibition and SDT.

These results suggest that TiN-BTZ combined with ultrasound treatment triggers a series of ER stress events. Proteasome inhibition and ROS generated by TiN-mediated SDT lead to mitochondrial calcium overload, triggering a burst of mitochondrial ROS, creating a vicious cycle, and ultimately inducing apoptosis through mitochondrial damage.

3.8. RNA sequencing for the mechanism prediction

A comprehensive RNA sequencing analysis was performed to elucidate the mechanism of the synergistic effect of the combination treatment. An initial global assessment of the data using principal component analysis (PCA) (Fig. S22a) and volcano plot (Fig. S22b) were used to visualize the sample variation and expression difference statistics. And the heat map can also reflect that the treatment group has a prominent difference in gene expression, as shown in Fig. 6a. Examination of individual genes provided specific evidence for the activation of pathways. Notably, the expression of key pro-apoptotic and stress-response genes is significantly upregulated, including the ER stress-specific caspase Casp12, the master stress sensor ATF3, and the UPR effector Ppp1r15a (Fig. 6b). Concurrently, genes crucial for cellular defense and survival are markedly downregulated (Fig. 6c). They include Gpx1 which is a key antioxidant enzyme, and Ucp2 which mitigates mitochondrial ROS, suggesting a state of severe oxidative stress. Critically, the downregulation of Lig1 impairs DNA repair capacity. Moreover, the reduced expression of Atp2b4, a plasma membrane calcium pump, provides direct molecular explanation for the impaired calcium efflux and cytosolic overload observed previously. In order to understand the biological significance of these differentially expressed genes (DEGs), Kyoto Encyclopedia of Genes and Genomes (KEGG) and GO enrichment analysis were conducted. KEGG analysis indicates that the combination treatment reveals a significant enrichment of pathways directly involved in programmed cell death, including apoptosis and necroptosis. It correlates with the protein processing in ER and ROS (Fig. 6d). Consistent with these results, GO enrichment analysis shows that several pathways regulating apoptosis are prominent in the combined treatment (Fig. 6e), including oxidative stress, proteasome-involved mechanism, UPR pathway, calcium ion influx and DNA damage. These results suggest that BTZ combined with TiN-mediated SDT treatment exerts superior anticancer effect via synergistic proteasome inhibition, ROS-mediated oxidative damage, UPR activation, and calcium overload, inducing the necrosis and apoptosis of tumor cells.

Fig. 6.

RNA-seq analysis. a) Heat map displaying the expression of the differentially expressed genes (DEGs) for each sample in the RNA-seq dataset (P < 0.05, fold change ≥1.5). Red and blue colors indicate upregulated and downregulated gene expression, respectively. Line plots showing the representative upregulated b) and downregulated c) genes across different treatment groups. d) KEGG pathway enrichment analysis of DEGs. e) GO enrichment analysis of biological processes for DEGs. The bubble size represents the number of enriched genes (count), and the color scale indicates the P value. The 4T1 cells were treated with PBS, BTZ (20 nM), TiN (40 μg mL⁻¹), or TiN-BTZ NPs (20 nM BTZ, 40 μg mL⁻¹ TiN) in the presence or absence of US irradiation (50 kHz, 2 W cm⁻², 50% duty cycle, 4 min) for 24 h. RNA-seq was then performed.

3.9. In vivo biodistribution and pharmacokinetics of TiN@BSA NPs

4T1 cells were implanted subcutaneously into the flank of female BALB/c mice to establish the xenograft tumor model. Cy7-labeled TiN@BSA NPs were injected into mice, and Cy7-labeled TiN@BSA NPs were tracked ex vivo at 2, 4, 6, 8, and 24 h post-injection to assess biodistribution. As shown in Fig. 7a and b, the tumors exhibit sustained accumulation of Cy7-labeled TiN@BSA NPs, with fluorescence intensity of Cy7 peaking at 8 h post-injection. Persistent tumor retention of TiN@BSA NPs was observed at 24 h post-injection, with minimal off-target accumulation in the major organs. To confirm the accumulation of TiN@BSA NPs in the tumor, infrared thermography was employed to monitor tumor temperature under 808-nm laser irradiation, because TiN has a strong absorption in the NIR region enabling photothermal therapy. As shown in Fig. S23, the tumor temperature in mice treated with TiN@BSA NPs is significantly higher than that in the TiN group, reaching approximately 50 °C. This superior photothermal effect is attributed to the enhanced tumor accumulation of TiN@BSA NPs. These results suggest that TiN@BSA NPs can effectively accumulate at the tumor site due to the active targeting ability of BSA.

Fig. 7.

In vivo biodistribution and pharmacokinetics study. a) In vivo fluorescence imaging of tumor-bearing mice at different time points (2, 4, 6, 8, 24 h) after intravenous injection of Cy7-labeled TiN or Cy7-labeled TiN@BSA NPs. b) The major organs and tumors were harvested for ex vivo fluorescence imaging for 24 h after intravenous injection. c) Blood drug concentration-time curve. The 4T1 tumor-bearing mice were intravenously injected with Cy7-labeled TiN or Cy7-labeled TiN@BSA NPs (20 mg kg⁻¹ TiN), the fluorescence intensity of Cy7 in the blood was measured. Data represent the mean ± S.E.M., n = 3.

Protein corona formation alters biological and pharmacokinetic properties of NPs. In order to determine the pharmacokinetics parameters of the prepared NPs, mice were intravenously injected with Cy7-labeled TiN and TiN@BSA (Cy7, 200 μg kg⁻¹), and the plasma concentration of Cy7 was monitored over time (Fig. 7c). Compared to unmodified TiN, TiN@BSA significantly prolongs the terminal elimination half-life (t_1/2z), which is consistent with the observed higher area under the curve (AUC) and levels in blood (C_max), as well as a decrease in clearance (CL) (Table S1). These results suggest that TiN@BSA can improve the pharmacokinetics of TiN.

It was also reported that secreted protein acidic and rich in cysteine (SPARC) which is the receptor for albumin exhibits a high expression level in the tumors [72]. The above results indicate that BSA coating is beneficial to improve the tumor targeting ability and blood circulation time of the TiN@BSA NPs.

3.10. In vivo antitumor effect of combination treatment with TiN-BTZ and SDT

The antitumor efficacy of the combination with BTZ and SDT therapy was assessed in 4T1 tumor-bearing BALB/c mice. As shown in Fig. 8a, when the tumor volumes reach approximately 60–100 mm³ (day 0), the mice were treated with the following regimens: saline control, BTZ (0.5 mg kg⁻¹ BTZ), TiN (40 mg kg⁻¹ TiN), TiN-BTZ (40 mg kg⁻¹ TiN, 0.5 mg kg⁻¹ BTZ), US (50 kHz, 2 W cm⁻², 50% duty cycle, 4 min), TiN (40 mg kg⁻¹ TiN) plus US, and TiN-BTZ (40 mg kg⁻¹ TiN, 0.5 mg kg⁻¹ BTZ) plus US on day 0, 3, 6, and 9, respectively. The US group received the US treatment 12 h after injection. Tumor volumes (Fig. 8b) and body weights (Fig. 8c) were recorded every two days. The control group shows sustained tumor growth, and treatment with BTZ only slightly inhibits tumor growth. Although the TiN NPs do not affect the growth of the tumor, TiN-BTZ moderately delays the tumor growth, indicating that delivery of BTZ with TiN effectively increases the therapeutic effect of BTZ in vivo. The US alone in the absence of TiN NPs has no effect on the tumor growth, but TiN-mediated SDT shows comparable tumor growth inhibition to TiN-BTZ therapy. Compared with each monotherapy, the combination treatment with BTZ and TiN-mediated SDT shows the strongest tumor suppression (Fig. 8b–d). The tumor weight data (Fig. 8e) also confirms the above result. At the endpoint, the combination treatment with BTZ and TiN-mediated SDT leads to the highest growth inhibition rate (Fig. 8f). Meanwhile, body weights remain stable in all treatment groups (Fig. 8c), indicating minimal treatment-associated toxicity.

Fig. 8.

In vivo antitumor effect. a) Schematic illustration for the design of experiments in vivo. The xenograft tumor model was established by subcutaneous injection of 4T1 cells in the BALB/c mice. The mice were intravenously injected with normal saline, BTZ (0.5 mg kg⁻¹), TiN NPs (40 mg kg⁻¹), or TiN-BTZ NPs (40 mg kg⁻¹ TiN, 0.5 mg kg⁻¹ BTZ) on day 0, 3, 6, and 9, respectively. For SDT treatment, the mice were exposed to US irradiation (50 kHz, 2 W cm⁻², 50% duty cycle, 8 min). b) Tumor volume growth profiles and c) body weights of 4T1 tumor-bearing mice at various time points. d) Photographs, e) weights, and f) calculated inhibition rates of the tumors at the end of treatment. g) Expression of ubiquitinated proteins, BiP, and γ-H2A.X in the tumors by western blotting analysis. Data represent the mean ± S.E.M. (n = 5), ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001.

Consistent with the in vitro results, BTZ and TiN-mediated SDT combination therapy induces the maximal ubiquitinated protein accumulation in the tumor tissue from western blotting analysis (Fig. 8g and Fig. S24a) and immunofluorescence staining study (Fig. 9). Both BiP and γ-H2A.X protein levels are also the highest in the tumors after the combination treatment (Fig. 8g and Fig. S24b–c). H&E analysis of the tumor tissue sections shows that BTZ, TiN-BTZ, or TiN-mediated SDT produces only a slight damaging effect on tumors, while the combination with BTZ and TiN-mediated SDT induces substantial tumor destruction with the signs of nuclear pyknosis, karyolysis, and karyorrhexis (Fig. 9). The cell proliferation and apoptosis levels in tumors were further investigated by immunofluorescence staining of the proliferation marker Ki67 and TUNEL assay, respectively. In the control group, the tumors exhibit robust Ki67 expression and a negligible number of TUNEL-positive cells (Fig. 9 and Fig. S25). Compared with the treatment with BTZ, TiN-BTZ, or TiN-mediated SDT, BTZ and TiN-mediated SDT combination therapy synergistically induces a superior antitumor effect, as evidenced by the highest ROS level, reduced expression of Ki67 and increase of TUNEL-positive cells. These observations further confirm the SDT enhances efficacy of BTZ (Fig. 9 and Fig. S25).

Fig. 9.

H&E, TUNEL assay (red), ROS staining by DCFH-DA (green), and immunofluorescence staining of ubiquitinated proteins (green), Ki67 (green), HMGB1 (red), CRT (green) and CD8 (red) of the tumor sections. The nuclei were stained with Hoechst 33342 (blue).

SDT has also been demonstrated to elicit robust antitumor immunity by triggering ICD [73]. Notably, SDT-based synergistic platforms can significantly amplify the ICD effect compared to the monotherapies, thereby optimizing immune activation. During ICD, tumor cells typically produce DAMPs, characterized by exposure of the calreticulin (CRT) on the surface of dying cells, followed by the release of high-mobility group box 1 (HMGB1) from the nucleus and ATP secretion. These DAMPs serve as potent immunological adjuvants that recruit cytotoxic T lymphocytes (CTLs) into the tumor. To evaluate the ICD-inducing capacity in vivo [74,75], immunofluorescence staining was performed on tumor sections. As shown in Fig. 9 and Fig. S25, the TiN-mediated SDT group exhibits distinct CRT surface exposure (green) and HMGB1 translocation (red) from the nucleus to the cytoplasm/extracellular space. More pronounced CRT exposure on the cell membrane and HMGB1 release were observed in the BTZ and SDT combination therapy group. In contrast, the control group shows only very weak ICD marker signals. Notably, with the massive production of DAMPs, the proportion of CD8⁺ effector T cells (red) infiltrating the tumor tissue in the combination therapy group increases significantly. This finding confirms that the “dual-hit” strategy can transform “cold” tumors into “hot” tumors by inducing intense ER stress, thereby enhancing the attack of CTLs on tumor tissue.

3.11. Safety evaluation of the combination therapy

Biosafety of the therapeutic agents is the first requisite to their further clinical translation for cancer therapy. Thus, the biocompatibility was determined, mice were sacrificed after completion of treatment, blood samples and the major organs were collected for detecting biochemical indicators and organ toxicity. To assess the in vivo safety of the TiN and SDT treatment, key indicators of liver, kidney function, and inflammatory factors were evaluated. No statistically significant differences in the levels of the liver enzymes AST and ALT, the renal function markers BUN and CRE, as well as IL-6 and TNF-α between the treated and control groups was observed (Fig. S26 a–f). In addition, we evaluated the immunotoxicity of the NPs and investigated whether the treatment would induce an immunogenic response in the body. The results show no significant changes in serum immunoglobulin G (IgG) and immunoglobulin M (IgM) before and after the treatment (Fig. S26 g and h), proving that the treatment does not induce an immunogenic response in the body. These results suggest that the therapeutic intervention does not induce any discernible hepatotoxicity, nephrotoxicity or immunogenic response. The H&E staining shows there is no harm to the organs after the treatment (Fig. S26i). Taken together, the combination therapy using TiN-BTZ shows promising therapeutic outcomes and favorable biocompatibility.

4. Conclusions

In summary, an “all-in-one” strategy that integrates TiN-augmented SDT with proteasome inhibitor BTZ was developed for cancer therapy based on spatiotemporal activation of UPR. CA-functionalized BSA was used as a bridge to connect TiN NPs and BTZ. BTZ can coordinate with CA via non-covalent intermolecular interactions and thus be loaded on the surface of BSA-coated TiN NPs via self-assembly. BTZ and SDT complement each other perfectly to inhibit all three proteolytic sites of proteasome, β1, β2, and β5. The combination therapy can exacerbate oxidative damage, leading to a more severe disruption to proteostasis, and synergistically upregulate the activation of ER stress and UPR. In addition, this “dual-hit” approach can effectively trigger strong ICD in tumors. Both in vitro and in vivo studies demonstrated that the combination therapy exhibits a superior anticancer effect. Taken together, combination therapy with SDT and proteasome inhibitors shows great potential for solid tumor therapy.

CRediT authorship contribution statement

Dan Yin: Writing – original draft, Validation, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Xuan Wu: Methodology, Investigation. Jian-Li Chen: Methodology, Investigation. Ke An: Methodology, Investigation. Yini Fang: Methodology, Investigation. Hao-Ming Xiong: Methodology, Investigation. Li Wang: Methodology, Investigation. Chen Huang: Methodology, Investigation. Guo-Yuan Zhu: Methodology, Investigation. Liang Cheng: Methodology, Investigation. Jianfang Wang: Writing – review & editing, Validation, Supervision, Resources, Project administration, Conceptualization. Xiao-Ming Zhu: Writing – review & editing, Validation, Supervision, Resources, Project administration, Funding acquisition, Conceptualization.

Ethics approval and consent to participate

All the procedures for animal use in this study were conducted in strict accordance with the Guidelines for the Care and Use of Laboratory Animals. All animal experiments were approved by the Animal Ethics Committee, Macau University of Science and Technology (No. AL011/DICV/DIS/2025).

Declaration of competing interest

The authors declare no conflict of interests.

Appendix A. Supplementary data

The following is the Supplementary data to this article: