**Deep Tumor Penetration of Doxorubicin-Loaded Glycol Chitosan Nanoparticles Using High-Intensity Focused Ultrasound**

**Yongwhan Choi 1,2, Hyounkoo Han 2,3, Sangmin Jeon <sup>2</sup> , Hong Yeol Yoon <sup>2</sup> , Hyuncheol Kim <sup>3</sup> , Ick Chan Kwon 1,2,\* and Kwangmeyung Kim 1,2,\***


Received: 8 September 2020; Accepted: 14 October 2020; Published: 15 October 2020

**Abstract:** The dense extracellular matrix (ECM) in heterogeneous tumor tissues can prevent the deep tumor penetration of drug-loaded nanoparticles, resulting in a limited therapeutic efficacy in cancer treatment. Herein, we suggest that the deep tumor penetration of doxorubicin (DOX)-loaded glycol chitosan nanoparticles (CNPs) can be improved using high-intensity focused ultrasound (HIFU) technology. Firstly, we prepared amphiphilic glycol chitosan-5β-cholanic acid conjugates that can self-assemble to form stable nanoparticles with an average of 283.7 ± 5.3 nm. Next, the anticancer drug DOX was simply loaded into the CNPs via a dialysis method. DOX-loaded CNPs (DOX-CNPs) had stable nanoparticle structures with an average size of 265.9 ± 35.5 nm in aqueous condition. In cultured cells, HIFU-treated DOX-CNPs showed rapid drug release and enhanced cellular uptake in A549 cells, resulting in increased cytotoxicity, compared to untreated DOX-CNPs. In ECM-rich A549 tumor-bearing mice, the tumor-targeting efficacy of intravenously injected DOX-CNPs with HIFU treatment was 1.84 times higher than that of untreated DOX-CNPs. Furthermore, the deep tumor penetration of HIFU-treated DOX-CNPs was clearly observed at targeted tumor tissues, due to the destruction of the ECM structure via HIFU treatment. Finally, HIFU-treated DOX-CNPs greatly increased the therapeutic efficacy at ECM-rich A549 tumor-bearing mice, compared to free DOX and untreated DOX-CNPs. This deep penetration of drug-loaded nanoparticles via HIFU treatment is a promising strategy to treat heterogeneous tumors with dense ECM structures.

**Keywords:** glycol chitosan nanoparticle; high-intensity focused ultrasound; deep tumor penetration; dense ECM; cancer treatment

#### **1. Introduction**

Anticancer drug-loaded nanoparticles have been used extensively in cancer treatment. This is because drug-loaded nanoparticles can be efficiently localized at targeted tumor tissues via nanoparticle-derived enhanced permeation and retention (EPR) effects in many pre-clinical tests [1–3]. The rapid growth of tumor tissues can cause a leaky vasculature as well as a suppression of lymphatic drainage, resulting in making different characteristics from those of the normal vasculature. In particular, since nanoparticles can extravasate into tumor tissues efficiently via the EPR effect, the EPR effect is regarded as the golden standard in designing nanoparticles for drug delivery [4,5]. Therefore, various nanosized materials, such as liposomes, polymeric nanoparticles, metal nanoparticles, and inorganic nanoparticles have been used in tumor-targeting delivery systems [6–9]. However, challenges remain to further improve the therapeutic efficacy of drug-loaded nanoparticles in heterogeneous tumors [10–12]. In particular, the delivery efficacy of drug-loaded nanoparticles is hampered greatly by limited deep tumor penetration in the complex tumor microenvironment [13–15]. It has been known that heterogeneous tumors differ in their vascular structure and perfusion rate [16,17]. Moreover, the thick extracellular matrix (ECM) which consists of collagen and hyaluronan (HA) in the tumor tissue can inhibit the deep tissue penetration of drug-loaded nanoparticles [18,19]. This is because the dense ECM can act as a physical barrier to the accumulation and deep tissue penetration of drug-loaded nanoparticles [15,20]. Thus, the development of a way to ensure a deep penetration of drug-loaded nanoparticles into tumor tissue is an essential challenge to improve the therapeutic efficacy of nanoparticle-based drug delivery systems in cancer treatment.

Many researchers have tried to improve the delivery efficiency of drug-loaded nanoparticles through the remodeling of the ECM in the tumor microenvironment [21,22]. Enzyme-conjugated nanoparticles have been used to increase the deep tissue penetration of nanoparticles through deconstructing the ECM structure, resulting in improvements in the therapeutic efficacy of the tumor. For example, matrix metalloprotease (MMP)-conjugated nanoparticles could break down the ECM structure, resulting in an improved delivery efficiency into deep tumor tissue and the therapeutic efficacy of drug-loaded nanoparticles [23]. Nevertheless, the applications of enzyme-conjugated nanoparticles is still limited due to the complex chemical reactions that bind the enzyme to the nanoparticle surface [24,25]. More practically, high-intensity focused ultrasound (HIFU) technology has been used to break down physically the dense ECM structure in the tumor microenvironment without any toxicity in normal organs. HIFU-mediated drug delivery systems could improve the delivery of high-molecular-weight antibodies and nanoparticles to tumor tissue, due to the successful destruction of the ECM barrier in tumor tissues [26,27]. In our previous report, we reported the exact mechanism of the HIFU-mediated deep tumor penetration of nanoparticles in heterogeneous tumor models [28,29]. Many human solid tumors express high levels of collagen and hyaluronan matrixes that can acts as physical barriers for inhibiting the deep tumor penetration of antibodies and high molecular anticancer drugs [30]. In particular, these EMC-rich tumor tissues composed of highly expressed collagen and hyaluronan matrixes can affect the accessibility and deep tumor penetration of nanosized drug delivery systems in pre-clinical tests [31]. Interestingly, the dense ECM structure of tumor tissues was successfully destroyed by non-invasive pulsed-HIFU exposure. Furthermore, the interstitial flow pressure (IFP) in the tumor tissue was reduced by normalizing the tumor vessels in ECM-rich tumors. Surprisingly, intravenously injected nanosized nanoparticles could be successfully accumulated at in ECM-rich tumors exposed to non-invasive HIFU treatments. These overall results demonstrate that ECM remodeling by HIFU treatment is a promising strategy to enhance the deep tumor penetration and enhanced tumor targeting of drug-loaded nanoparticles in solid tumors.

Glycol chitosan is a natural polysaccharide which is derived from chitosan, it has biocompatibility, non-toxicity, biodegradability, and easy fabrication properties [32]. Notably, a large number of reactive functional groups (primary amine and hydroxyl group) on the glycol chitosan backbone can be modified with cholanic acids, hydrotropic oligomer, photosensitizers, and fullerene, resulting in the formulation of nanomedicines for chemotherapy, gene therapy, and photodynamic therapy [33]. Among them, hydrophobically modified glycol chitosan can form self-assembled nanoparticles due to its amphiphilic structure. In particular, the hydrophobic inner cores of glycol chitosan nanoparticles can be used to deliver theranostic agents such as paclitaxel, docetaxel, and iron oxide nanoparticles via the EPR effect in tumor tissue, resulting in an improved drug delivery efficiency as well as tumor-specific imaging in pre-clinical mice tumor models [34–36].

Herein, we evaluate the drug delivery efficacy and therapeutic efficacy of HIFU-triggered drug-loaded nanoparticles at ECM-rich tumor models, wherein the ECM-rich tumor tissues were treated with HIFU to destroy the dense ECM structure at A549 tumor tissues (Figure 1a). First, we prepared doxorubicin-loaded glycol chitosan nanoparticles as model drug-loaded nanoparticles. We expect that

doxorubicin (DOX)-chitosan nanoparticles (CNPs) are very suitable as model drug delivery systems, due to their high-tumor-targeting ability and low systemic toxicity in vivo [37]. The biodegradable and hydrophilic glycol chitosan polymers were modified hydrophobic 5β-cholanic acid and the conjugates were self-assembled to form glycol chitosan nanoparticles (CNPs). Next, the anticancer drug doxorubicin (DOX) was loaded into CNPs via a simple dialysis method, resulting in DOX-loaded CNPs (DOX-CNPs). The in vitro drug release, cellular uptake and cytotoxicity of HIFU-triggered DOX-CNPs were characterized in cultured cells. Finally, the deep tumor penetration and therapeutic efficacy of HIFU-triggered DOX-CNPs were carefully examined in an ECM-rich A549 tumor animal model, compared to free DOX and untreated DOX-CNPs. that doxorubicin (DOX)-chitosan nanoparticles (CNPs) are very suitable as model drug delivery systems, due to their high-tumor-targeting ability and low systemic toxicity in vivo [37]. The biodegradable and hydrophilic glycol chitosan polymers were modified hydrophobic 5β-cholanic acid and the conjugates were self-assembled to form glycol chitosan nanoparticles (CNPs). Next, the anticancer drug doxorubicin (DOX) was loaded into CNPs via a simple dialysis method, resulting in DOX-loaded CNPs (DOX-CNPs). The in vitro drug release, cellular uptake and cytotoxicity of HIFUtriggered DOX-CNPs were characterized in cultured cells. Finally, the deep tumor penetration and therapeutic efficacy of HIFU-triggered DOX-CNPs were carefully examined in an ECM-rich A549 tumor animal model, compared to free DOX and untreated DOX-CNPs.

**Figure 1.** (**a**) High-intensity focused ultrasound (HIFU) treatment of doxorubicin (DOX)-loaded glycol chitosan nanoparticles (CNPs) (DOX-CNPs) to increase their deep tumor penetration in extracellular matrix (ECM)-rich tumor models. (**b**) Schematic illustration of glycol chitosan-5β-cholanic acid conjugate. The glycol chitosan-5β-cholanic acid conjugates form self-assembled nanoparticles in aqueous condition. The anticancer drug of DOX can be loaded into CNPs via a dialysis method, **Figure 1.** (**a**) High-intensity focused ultrasound (HIFU) treatment of doxorubicin (DOX)-loaded glycol chitosan nanoparticles (CNPs) (DOX-CNPs) to increase their deep tumor penetration in extracellular matrix (ECM)-rich tumor models. (**b**) Schematic illustration of glycol chitosan-5β-cholanic acid conjugate. The glycol chitosan-5β-cholanic acid conjugates form self-assembled nanoparticles in aqueous condition. The anticancer drug of DOX can be loaded into CNPs via a dialysis method, resulting in DOX-CNPs.

#### resulting in DOX-CNPs. **2. Materials and Methods**

purification or modification.

#### **2. Materials and Methods**  *2.1. Materials*

*2.1. Materials*  Glycol chitosan (MW = 250 kDa; degree of deacetylation > 60%), doxorubicin hydrochloride (DOX-HCl), 5β-cholanic acid, 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide hydrochloride (EDC), N-hydroxysuccinimide (NHS), triethylamine (TEA), anhydrous methanol and anhydrous dimethyl sulfoxide (DMSO) were purchased from Sigma-Aldrich (Merck, Darmstadt, Germany). A fluorescent molecule, Cy5.5-NHS ester, was purchased from Lumiprobe Corporation (Hunt Valley, Glycol chitosan (MW = 250 kDa; degree of deacetylation > 60%), doxorubicin hydrochloride (DOX-HCl), 5β-cholanic acid, 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide hydrochloride (EDC), N-hydroxysuccinimide (NHS), triethylamine (TEA), anhydrous methanol and anhydrous dimethyl sulfoxide (DMSO) were purchased from Sigma-Aldrich (Merck, Darmstadt, Germany). A fluorescent molecule, Cy5.5-NHS ester, was purchased from Lumiprobe Corporation (Hunt Valley, MD, USA). All other chemicals were purchased as reagent grade and used without further purification or modification.

MD, USA). All other chemicals were purchased as reagent grade and used without further

EDC and NHS [15,38]. Briefly, 150 mg of 5β-cholanic acid was dissolved in 120 mL of methanol and mixed with EDC (120 mg) and NHS (72 mg). A total of 500 mg of glycol chitosan was dissolved in

To prepare glycol chitosan nanoparticles (CNPs), hydrophobic 5β-cholanic acid was chemically

*2.2. Synthesis and Characterization of Doxorubicin-Loaded Glycol Chitosan Nanoparticles (DOX-CNPs)* 

#### *2.2. Synthesis and Characterization of Doxorubicin-Loaded Glycol Chitosan Nanoparticles (DOX-CNPs)*

To prepare glycol chitosan nanoparticles (CNPs), hydrophobic 5β-cholanic acid was chemically conjugated to the hydrophilic glycol chitosan through amide linkage formation in the presence of EDC and NHS [15,38]. Briefly, 150 mg of 5β-cholanic acid was dissolved in 120 mL of methanol and mixed with EDC (120 mg) and NHS (72 mg). A total of 500 mg of glycol chitosan was dissolved in 120 mL of methanol/deionized distilled water solution (1:1 *v*/*v*), followed by slow mixing with 5β-cholanic acid solution. The mixture was vigorously stirred at room temperature for 24 h, and then purified by dialysis against distilled water/methanol (1:1 and 1:0 *v*/*v*) using a Spectra/Por®4 dialysis membrane (MWCO = 12–14 kDa, Repligen Corporation, Waltham, MA, USA). The resulting solution was lyophilized to obtain a white powder of CNPs. For the in vitro and in vivo fluorescence monitoring of CNPs, CNPs were chemically modified with near-infrared fluorescence (NIRF) dye, Cy5.5. In brief, 100 mg of CNPs were dissolved in 40 mL of DMSO, followed by mixing with 1 mL of DMSO containing 1 mg of Cy5.5-NHS. The mixture was stirred for 12 h at room temperature. Subsequently, the mixture was purified by dialysis against distilled water for 2 days using a Spectra/Por® 4 dialysis membrane (MWCO = 12–14 kDa, Repligen Corporation, Waltham, MA, USA). The resulting solution was lyophilized to obtain Cy5.5-conjugated CNPs.

To prepare DOX-encapsulated CNPs (DOX-CNPs), DOX-HCl was physically encapsulated into CNPs using a simple dialysis method. In brief, 50 mg of CNPs was dissolved in 10 mL of DMSO/distilled water (1:1 *v*/*v*). A total of 21.4 mg of DOX-HCl was dissolved in 2 mL of DMSO/distilled water (1:1 *v*/*v*). Then, the DOX-HCl solution was treated 10.8 µL of TEA for desalting, followed by mixing with CNP solution. The mixture was purified by a dialysis against distilled water for 12 h using a Spectra/Por® 4 dialysis membrane (MWCO = 12–14 kDa, Repligen Corporation, Waltham, MA, USA). The resulting solution was filtered using 0.8 µm syringe filter, followed by lyophilizing to obtain DOX-CNPs.

To confirm the hydrodynamic diameter of CNPs and DOX-CNPs, 1 mg of CNPs and DOX-CNPs were dispersed into 1 mL of PBS (pH 7.4) using a probe-type sonicator (Amp 21%, 1 min, VCX-750, Sonics and Materials, Newtown, CT, USA). The volume-weighted size distribution and zeta potential of CNPs and DOX-CNPs were measured using dynamic light scattering (DLS, Nano ZS, Malvern Panalytical Ltd., Grovewood Road, Malvern, UK) at 25 ◦C. The size stability and volume-weighted size distribution of CNPs and DOX-CNPs were monitored in both PBS (pH 7.4) and 1% FBS-containing PBS (pH 7.4) conditions using DLS. The morphologies of CNPs and DOX-CNPs were observed using transmission electron microscopy (TEM, CM30 Electron Microscope, Philips, CA, USA) at 200 eKV of an accelerating voltage. For TEM images, CNPs and DOX-CNPs were dispersed in distilled water and 5 µL of CNPs or DOX-CNPs solution was dropped on a 200 mesh carbon-coated copper grid, followed by negative staining using 2% uranyl acetate solution. Based on TEM images (*n* = 10), size distribution of CNPs and DOX-CNPs were determined using Image Pro Plus4.5 5 software (Media Cybernetics, Bethesda, Rockville, MD, USA).

### *2.3. In Vitro Release Profile of HIFU-Triggered DOX-CNPs*

An in vitro DOX release from DOX-CNPs was observed in 37 ◦C PBS (pH 7.4) containing 0.1% of Tween 80. To observe the HIFU-triggered DOX-release, 10 mg of DOX-CNPs was dispersed in PBS (2 mL) and placed in a dialysis membrane (MWCO = 100 kDa, Repligen Corporation, Waltham, MA, USA) (*n* = 3). The membrane was transferred into a 50 mL conical tube filled with 10 mL of PBS (pH 7.4) containing 0.1% of Tween 80. The conical tubes were placed in a 37 ◦C water bath then shaken horizontally (100 rpm). The dialysis membrane was treated with HIFU in destruction mode for 5 min (power: 10 MHz, mechanical index: 0.235) using a High-Resolution Micro-Imaging System (Vevo 770, Visualsonics, Toronto, ON, Canada). The amount of DOX released from CNPs at a pre-determined time point was determined using an absorbance at 490 nm measured by the UV–Vis spectrometer (G1103A, Agilent, Santa Clara, CA, USA).

#### *2.4. Cellular Uptake Behavior of HIFU-Triggered DOX-CNPs*

Human non-small cell lung tumor cells, A549, was cultured in Roswell Park Memorial Institute (RPMI) 1640 media (Welgene, Daegu, Korea) containing 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin at 37 ◦C in a 5% CO<sup>2</sup> incubator. To demonstrate the cellular uptake of DOX-CNPs, A549 cancer cells (1 <sup>×</sup> <sup>10</sup><sup>4</sup> cells) were seeded onto 35 mm glass bottom dish and incubated for 24 h. After 24 h post-incubation, the medium was replaced with 1 mL of FBS-free RPMI-containing Cy5.5-labeled DOX-CNPs (1 µg/mL of DOX, 1 mL). To characterize the cellular uptake mechanism of the HIFU-triggered cellular uptake mechanism of DOX-CNPs, DOX-CNP-treated A549 cells were treated with HIFU in destruction mode (power: 10 MHz, mechanical index: 0.235) for 5 min and then incubated for pre-determined time. DOX-CNP-treated A549 cells were washed twice using Dulbecco's Phosphate Buffered Saline (DPBS) and fixed with a 4% paraformaldehyde solution for 10 min. The nuclei of A549 cells was stained with DAPI for 5 min at room temperature. The fluorescence from A549 cells was visualized using confocal laser scanning microscopy (Leica TCS SP8, Wetzlar, Land Hessen, Germany) equipped with 405-diode (405 nm), Ar (488 nm), and HeNe-Red (633 nm) lasers. Tumor tissue fluorescence images were acquired using LAS X software (Leica Microsystems, Wetzlar, Land Hessen, Germany). The fluorescence of DOX and CNP was measured using Image Pro Plus 4.5 5 image analysis software (Media Cybernetics, Bethesda, Rockville, MD, USA).

#### *2.5. In Vitro Cytotoxicity Test of HIFU-Triggered DOX-CNPs*

The A549 tumor cells (5 <sup>×</sup> <sup>10</sup><sup>3</sup> cells/well) were seeded onto 96-well plates and stabilized for 12 h. The A549 cells were then incubated with various concentrations (0, 0.1, 1, 10, 100 and 500 µg/mL of DOX) of free DOX, CNPs and DOX-CNPs for 24 h. To measure cell viability, 10% (*v*/*v*) CCK-8 solution was added to each well, followed by further incubation for 1 h at 37 ◦C. The absorbance at 450 nm was measured using a microplate reader (VERSAmax™, Molecular Devices Corp., Sunnyvale, CA, USA).

To analyze the cell viability after HIFU treatment, A549 cells were incubated with CNPs and DOX-CNPs (100 µg/mL of DOX) for 24 h. These A549 cells were exposed to HFU in destruction mode (power: 10 MHz, mechanical index: 0.235) for 5 min, followed by further incubation for 24 h at 37 ◦C in a 5% CO<sup>2</sup> incubator. Finally, the A549 cells were washed twice with DPBS. Subsequently, 10% (*v*/*v*) CCK-8 solution was added to each well, followed by further incubating for 1 h at 37 ◦C. The absorbance at 450 nm was measured using a microplate reader (VERSAmax™, Molecular Devices Corp., Sunnyvale, CA, USA).

#### *2.6. In Vivo Biodistribution of HIFU-Triggered DOX-CNPs in A549 Tumor-Bearing Mice*

All animal experiments were performed in compliance with the guidelines of the Institutional Animal Care and Use Committee (IACUC) in the Research Animal Resource Center of Korea Institute of Science and Technology (approved number: 2017-109). To establish A549 tumor-bearing mice, <sup>1</sup> <sup>×</sup> <sup>10</sup><sup>7</sup> cells of A549 cells were inoculated in the left flank of male Balb-c/nude mice (4 weeks old, ORIENT BIO Inc., Gyeonggi-do, Korea). When the tumor volume reached approximately <sup>250</sup> <sup>±</sup> 50 mm<sup>3</sup> , Cy5.5-labeled DOX-CNPs (20 mg/kg, 100 µL) was injected through the tail vein. HIFU (VIFU 2000, ALPINION, Gyeonggi-do, Korea) was applied at the tumor site for 5 min simultaneously with an intravenous injection of Cy5.5-DOX-CNPs as pre-set conditions (intensity: 5 W/cm<sup>2</sup> , frequency: 1.5 MHz, duty cycle: 10%, pulse repetition frequency: 1 Hz, time per spot: 30 s, interval: 2 mm). Near infrared fluorescence (NIRF) images of mice animal models were carried out through IVIS SPECTRUM (Xenogen, Alameda, CA, USA). To compare the tumor and organ distributions of Cy5.5-labeled DOX-CNPs, the mice were sacrificed 24 h post-injection. Tumor, liver, lung, spleen, kidney, and heart were dissected from mice, and then NIRF images were obtained via IVIS SPECTRUM.

To observe the deep tissue penetration of Cy5.5-labeled DOX-CNPs, tumor tissues were excised 24 h post-injection of Cy5.5-labeled DOX-CNPs. Excised tumor tissues were embedded in an optimum cutting temperature tissue compound (OCT compound, Sakura Finetek, Chuo-ku, Tokyo, Japan), followed by a transfer to a refrigerator at under −20 ◦C for 24 h. The tumor tissue blocks were sectioned with a 10 µm thickness with a cryostat (Leica, Bannockburn, IL, USA). The tumor tissue slides were washed with distilled water twice to remove the OCT compound, followed by nuclei staining using DAPI solution for about 10 min. After being washed three times with DPBS, tumor tissue slides were fixed using mounting solution (Vectashield, Vector Laboratories Inc., Burlingame, CA, USA). The fluorescence of the tumor tissue was observed using a fluorescence microscopy (OLYMPUS, Tokyo, Japan).

#### *2.7. Antitumor E*ffi*cacy of DOX-CNPs with HIFU Treatment*

To evaluate antitumor efficacy in animal models, 1 <sup>×</sup> <sup>10</sup><sup>7</sup> cells of A549 cells were inoculated in the left flank of male Balb-c/nude mice (4 weeks old, ORIENT BIO Inc., Gyeonggi-do, Korea). When the tumor volume reached approximately 60 <sup>±</sup> 5 mm<sup>3</sup> , saline, DOX (2 mg/kg), DOX-CNPs (2 mg/kg of DOX) were injected into the A549 tumor-bearing mice through the tail vein. At 1, 3, 5 and 7 days post-injection, the tumor tissues were treated with HIFU for 5 min as pre-set conditions (intensity: 5 W/cm<sup>2</sup> , frequency: 1.5 MHz, duty cycle: 10%, pulse repetition frequency: 1 Hz, time per spot: 30 s, interval: 2 mm). Tumor volume and survival rate were monitored for 22 days to evaluate the antitumor efficacy of each group.

#### *2.8. Histological Analysis*

To observe the ECM-rich structure of A549 tumor tissue, collagen in murine squamous cell carcinoma (SCC7) and A549 tumor tissues were stained using Masson's trichrome staining method [28].

In brief, 1 <sup>×</sup> <sup>10</sup><sup>6</sup> cells of SCC7 cells and 1 <sup>×</sup> <sup>10</sup><sup>7</sup> cells of A549 cells were inoculated in the left flank of male Balb-c/nude mice (4 weeks old, ORIENT BIO Inc., Gyeonggi-do, Korea). When SCC7 and A549 tumor tissues grew to 250 <sup>±</sup> 50 mm<sup>3</sup> , both tumors were excised and fixed in 4% formaldehyde solution. The tumor tissues were then embedded in paraffin after dehydration. Paraffin-embedded tumor tissues were cut to 6 µm thick and tissue slides were stained with Masson's trichrome staining solution.

Furthermore, ECM structure changes of A549 tumor tissues after HIFU treatment were observed using Masson's trichrome staining method. In brief, 1 <sup>×</sup> <sup>10</sup><sup>7</sup> cells of A549 cells were inoculated in the left flank of male Balb-c/nude mice (4 weeks old, ORIENT BIO Inc., Gyeonggi-do, Korea). When A549 tumor tissues grew to 250 <sup>±</sup> 50 mm<sup>3</sup> , the tumor tissues were treated with HIFU for 5 min as pre-set conditions (intensity: 5 W/cm<sup>2</sup> , frequency: 1.5 MHz, duty cycle: 10%, pulse repetition frequency: 1 Hz, time per spot: 30 s, interval: 2 mm). The tumor tissues were then excised and embedded in paraffin after dehydration. Paraffin-embedded tumor tissues were cut to 6 µm thick and tissue slides were stained with Masson's trichrome staining solution. Blue-stained collagen in the tumor tissue slides were observed using a light microscope (Olympus, Tokyo, Japan). For the quantification of blue-stained collagen areas, they were measured using Image Pro Plus 4.5 5 image analysis software (Media Cybernetics, Bethesda, Rockville, MD, USA).

To observe the organ toxicity of DOX-CNPs according to HIFU exposure, the liver and kidneys were excised from the A549 tumor-bearing mice at 24 h post-injection of DOX-CNPs with HIFU exposure. The excised liver and kidneys were fixed using 4% formaldehyde solution, followed by embedding in paraffin after dehydration. The paraffin-embedded tissues were cut into 6 µm-thick sections and were stained with hematoxylin and eosin (H&E).

To observe morphological changes in the tumor tissues, tumor tissues were excised from the A549 tumor-bearing mice at 22 days post-treatment. The excised tumor tissues were fixed using 4% formaldehyde solution, followed by embedding in paraffin after dehydration. The paraffin-embedded tissues were cut into 6 µm-thick sections and were stained with hematoxylin and eosin (H&E). All histological analysis images were acquired through a light microscope (Olympus, Tokyo, Japan).
