Double safety guarantees: Food-grade photothermal complex with a pH-triggered NIR absorption from zero to one

Abstract

Photothermal therapy has aroused great attention and showed promising potential in minimally invasive tumor ablation, but the clinical translation is still stifled by the concerns of unwanted injury to normal tissues. The safety concerns might be completely solved only when the two security obstacles of “material-toxicity” and “photo-toxicity” were overcome simultaneously. Herein, a completely non-toxic food-grade photothermal transduction agent (PTA) with double safety guarantees was invented, which shows an absolute transformation of the photothermal effect from “0” to “1” after being triggered by an acidic tumor microenvironment. Inspired by the classical starch-iodine test, a preprogrammed [starch-KI-KIO₃] complex was prepared in large quantities through a modified wet-milling procedure. It's demonstrated that a macroscopic consecutive reaction could be triggered by low pH to produce the starch-iodine complex which can generate lethal temperature under the near-infrared light irradiation. Meanwhile, the PTA shows excellent biocompatibility with no “material-toxicity” owing to the raw materials drawn from our daily food. Animal experiments reveal that the tumor microenvironment can activate the switch of photothermal effect from “0” to “1” successfully, which is thus responsible for the discriminative photo-damage to the tumor region while no “photo-toxicity” to normal tissue. The good treatment efficacy confirms the feasibility of such photothermal transduction agents with double safety guarantees in clinical applications.

Graphical abstracts

1. Introduction

Photothermal therapy (PTT) powered by near-infrared (NIR) light is arousing intensive investigations in tumor treatments [1], [2], [3]. The photothermal transduction agent (PTA) has been considered the soul of PTT, which can effectively tailor a hyperthermia turn-on response to NIR light so as to further induce tumor cell necrosis and apoptosis [4,5]. Over the past decade or so, numerous studies have been devoted to improving the photothermal conversion efficiency and exploring new PTA for efficient PTT, mainly including small dye molecules [6,7], metal-based nanoparticles [8,9], conjugated polymers [10], [11], [12], [13], carbon nanomaterials [14,15], and two-dimensional materials [16,17]. Although great achievements have been made in establishing PTT-related materials and methodologies, the clinical translation of PTT is still slow and hindered by the safety of PTA, mainly the concerns of unwished injury to normal tissues. In addition to high photothermal conversion efficiency, more importantly, an ideal PTA must be safe.

From the viewpoint of clinical applications, there are two inevitable security obstacles that need to be overcome. One is the “material-toxicity” of PTA, and the other is the “photo-toxicity”. “Material-toxicity” of PTA is a collective term including the cytotoxicity of PTA, and the toxicity caused by degradation products or long-term accumulation [18,19]. All of those issues may give rise to severe inflammatory responses and thus increase the risk of systemic toxicity. Amongst those PTAs in existence, most examples still suffer from great concerns regarding their long-term toxicity, accounting for the hindrance of their clinical implementations. For example, some studies have contradictorily presented that the presence of gold nanoparticles may destroy actin stress, induce apoptosis or cause inflammation in mouse livers [20,21]. It was also found that reduced graphene oxide nanoribbons exhibited genotoxic effects on the stem cells through DNA fragmentation and chromosomal aberrations, even at a low concentration of 0.1 μg/mL [22,23]. Therefore, it is highly desired to develop a completely non-toxic PTA for potential clinical applications of PTT.

The “photo-toxicity” of PTA is another security challenge that refers to non-specific thermal damage to normal tissues during irradiation [24]. Although PTT shows great potential in minimally invasive tumor ablation owing to the less attenuation and deeper penetration of NIR light throughout human tissue, the extraneous photo-damage along the laser path to the tumor is still unavoidable. The inevitable and uncontrollable distribution of PTA in healthy tissues may cause undesired photo-injury, referring to the side effect of photothermal conversion. Besides, it's unpractical and technically difficult to localize photoirradiation specifically on tumor cells without affecting other normal cells [25]. Hence, in order to comprehensively improve the safety of PTA, highly sensitive PTA with little “photo-toxicity” in normal tissues while producing lethal effects over the targeting tumors has always been among the top-priority choices. The recent advances in exploring the differences between tumor and normal tissues are inspiring the design of smart PTA in response to the tumor microenvironment [26,27]. One well-known feature of tumor physiology is acidosis. It's generally believed that the deregulated energy metabolism, insufficient blood perfusion, and uncontrolled proliferation collectively cause the acidic tumor microenvironment with a pH range of 6.4 - 7.0 [28,29]. Owing to the heterogeneity of tumor acidosis, the region with high proliferation but far from the vasculature, or in a late-developing stage would liberate and accumulate more H⁺ or lactate, resulting in a lower pH value even below 6.0 [29,30]. Besides, it's documented that many nanoparticles could be captured by the endocytic vesicles in tumor tissue (pH = 5.5 - 6.5 in the early endosome and pH = 5.0 - 5.5 in the late endosome/lysosome) [31,32]. Utilizing the difference of pH, a few pH-responsive PTAs have been exploited and investigated [33], [34], [35], [36]. They can achieve higher photothermal conversion along with redshift absorption once extravasated into tumor tissues (photothermal effect “ON”), while not activated in normal tissues (photothermal effect “OFF”). However, none of them can completely eliminate the photo-damage to normal tissues owing to their low ON/OFF ratio, which could be defined as the ratio of the absorption in the acidic state to that in the neutral state. Most of them still maintain a relatively high photothermal effect under the neutral condition, which can still cause undesired injury to normal cells. Thus, designing a real pH-triggered PTA with discriminative photo damage changing from zero to one will be a great breakthrough in eliminating their “photo-toxicity” and further improving their safety for clinical applications.

In this context, this work proposes a new strategy to “cook” a safe PTA with double safety guarantees (no “material-toxicity” and no “photo-toxicity”) which we call SAFEs (pH-switched and food-grade ensured PTAs). The classical starch-iodine test is widely utilized in agriculture for maturity tests and in dermatology for detecting sweat glands. It's demonstrated that the amylose and amylopectin in starch can be complexed with iodine and display a broad absorption window from ultraviolet to NIR region [37]. Inspired by this, the starch-iodine complex is employed for tumor PTT here and innovatively endowed with a smart switch that can be triggered by low pH in the tumor environment. As described in Fig. 1a, those raw materials can be extracted from our daily foods, such as starch from potatoes, potassium iodide (KI) from seaweed, and potassium iodate (KIO₃) from iodized salt. Potato starch is totally non-toxic, and both KI and KIO₃ are general iodine sources for daily and clinical use to maintain the proper function of the vertebrate endocrine system [38]. Food-grade raw materials guarantee that the SAFEs is of no “material-toxicity” and easy to metabolize in the human body when at a proper dose. In principle, with the help of the redox reaction between KI and KIO₃ in acidic tumor microenvironments, a certain amount of iodine is produced and complexed with starch quickly. This process will change the absorption of so-called SAFEs from “0” (non-absorption) to “1” (NIR windows). Therefore, the photothermal effect can only be activated in acidic tumor tissues while non-absorption and no photo-damage to normal tissues, where the “photo-toxicity” can be almost completely eliminated. To the best of our knowledge, it is the first time that pH-responsive PTA can realize the shift of photothermal effect from “0” to “1” after being triggered by the acidic tumor microenvironment. The success of tumor-selective ablation according to the animal experiments convinces the availability of such SAFEs with double safety guarantees in clinical applications.

Fig. 1.

The preparation and characterization of SAFEs. (a) Schematic illustration of the cooked SAFEs for tumor-selective photothermal ablation. (b) Schematic illustration of freeze-drying and modified wet-milling for the preparation of SAFEs. (c) Dynamic light scattering (DLS) analysis of SAFEs dispersed in PBS solution (pH 7.4). The inserted photographs depict the Tyndall Effect of SAFEs nanoparticles. (d) SEM images of SAFEs nanoparticles. (e) The UV-Vis absorption spectra and representative photographs of the SAFEs suspension before (“0”) and after (“1”) addition of a small aliquot of acetic acid. The curve of state “1” is standardized based on the maximum absorption.

2. Materials and methods/experiment

2.1. Materials

Potato starch and agarose powder were purchased from Aoboxing Bio-tech Co. Ltd., Beijing. Potassium iodide (KI) and potassium iodate (KIO₃) were supplied by Aladdin Bio-Chem Technology Co. Ltd., Shanghai. Ethanol and formalin were obtained from Tong Guang Fine Chemical Co. Ltd., Beijing. Dulbecco's modified Eagle medium (DMEM) was supplied by HyClone, and 0.25% Trypsin-EDTA was obtained from Life Technologies. Penicillin-streptomycin solution (100 ×) and fetal bovine serum (FBS) were obtained from Cellgro. Phosphate buffer solution (PBS) was purchased from Biotopped, and the zirconia milling balls were purchased from Miqi Instruments Co. Ltd., Changsha.

2.2. Preparation of SAFEs

SAFEs/EtOH suspension (50 mg/mL) were prepared through a modified ball milling procedure. Three kinds of zirconia milling balls with different diameters (10.0 mm, 5.0 mm, and 0.5 mm) were selected as the grinding media. Typically, as shown in Fig. 1b, [starch+KI+KIO₃] aqueous solution was prepared first. Then the potato starch was dissolved in deionized water (20 mg/mL) via continuous stirring at 95 °C for 3 h. Next, a certain amount of KI and KIO₃ was added to the solution successively and mixed thoroughly. During this process, removing the carbon dioxide or adding a small dose of NaHCO₃ to maintain a neutral condition is necessary, which can avoid the reaction between KI and KIO₃. Finally, the [starch+KI+KIO₃] solid complex was obtained via freeze-drying. This complex (1.0 g) was put into a 50.0 mL milling jar, followed by adding the preselected milling balls. The first milling was dry-milling that can prepare SAFEs particles. But for obtaining nanoparticles with a smaller size, wet-milling was used for the second milling after adding 20.0 mL of ethanol into the jar. All the millings were carried out for 6 h with the milling cycle program that includes 5.0 min clockwise milling, 5.0 min standing, and 5.0 min anticlockwise milling. The specific feeding formula for preparing SAFEs with different mass fractions of iodized salts was provided in Table S1.

2.3. Characterizations

Dynamic light scattering (DLS, Zetasizer NANO 590, Malvern Instruments Co. Ltd.) was used to measure the hydrodynamic diameter of SAFEs nanoparticles. Nanoparticles’ morphology was characterized on a scanning electron microscope (SEM, SU8010, HITACHI). Element mapping images of potassium and iodine were carried out on an energy dispersive spectrometer (EDS, Bruker XFlash 6|60). The UV-vis absorption of SAFEs suspension after being triggered by acid was measured by a spectrometer (SHIMADZU UV-3600).

2.4. Photothermal conversion tests

A photothermal testing device (Fig. 2a) was set up to monitor the real-time core temperature changes, including a NIR light source with a wavelength of 808 nm, a cuvette loading photothermal fluid with a light path of 1 mm, and a thermometer (UNIT UT325) to record the real-time temperature. As shown in Fig. 2, SAFEs with different mass fractions of iodized salts were dispersed into PBS with different pH values. After incubating for 24 h at 37 °C, 150 μL of SAFEs suspension was transferred to the thin cuvette. The digital thermometer and infrared thermal camera were used to measure the temperature changes when receiving NIR light irradiation with a certain power density. The UV-vis absorption of SAFEs suspension after incubation was also measured by a spectrometer (SHIMADZU UV-3600).

Fig. 2.

Photothermal tests and calculation of photothermal conversion efficiency. (a) Schematic illustration of the photothermal testing device. (b) The temperature changes and (c) the IR images of SAFEs loaded with different mass fractions of iodized salts in PBS buffer (pH 5.0) under light irradiation with a power density of 1.0 W/cm². (d) The temperature curve, (e) temperature changes, and (f) the IR images of SAFEs in PBS solution (pH 5.0) under light irradiation with different power densities. (g, h) The temperature changes against on/off light irradiation of PBS buffer (pH 5.0) with and without SAFEs. (i, j) The linear fitted correlation between time (t) and -lnθ from the corresponding cooling period in Fig. 2g and 2h. θ is defined as the ratio of $\Delta T$ to $\Delta T_{max}$. The power density of laser is 1.0 W/cm², and the irradiation time is 10 min.

2.5. Kinetics tests of SAFEs

The kinetics of SAFEs in weak acidic buffer was investigated following the method mentioned in 2.4. 25 mg/mL of SAFEs suspensions were prepared via dispersing SAFEs into PBS with different pH values (5.0, 6.0, 6.5, 7.4). During the incubation, SHIMADZU UV-3600 was used to detect their UV-vis absorption at different times. Similarly, with the fixed pH of 6.0, SAFEs suspensions with different concentrations (5.0, 10, 25 mg/mL) were prepared. The change of their absorption spectrum was also measured by a UV-vis spectrometer. The details of kinetics calculation could be seen in the Supporting Information.

2.6. Cytotoxicity tests

The cytotoxicity of the SAFEs was evaluated by the Cell Counting Kit-8 (CCK-8) assay [39]. HeLa cells were chosen as tumor cells and L02 cells (human normal liver cells) were selected as the normal cells here. Typically, cells were seeded in 96-well plates and incubated for 24 h for cell attachment at 37 °C with 5% CO₂. Meanwhile, the SAFEs extract was obtained by immersing the SAFEs particles in DMEM for 24 h at 37 °C. After attenuation, those extracts at different concentrations (0.02, 0.05, 0.10, 0.20, 0.50, 1.00, 2.00 g/mL) were added into the cells, which would be incubated for another 24 h. To perform the CCK-8 assay, the culture medium was removed, and then the cells were washed three times using PBS. The cells were incubated at 37 °C for 2 h to allow the formation of formazan dye after adding the culture medium with 10 μL of CCK-8 reagents. At last, their optical absorbance of the formazan dye at 450 nm was measured via a microplate spectrophotometer (Epoch 2, Biotek).

2.7. Investigations of biocompatibility

The biocompatibility of the SAFEs was evaluated by detecting the inflammation of mice. All of the animal experiments were approved by the Animal Ethics Committee of the Institute of Process Engineering in Beijing, and all experimental procedures complied with the ethical regulations. The nude mice (Female, BALB) were supplied by the Vital Laboratory Animal Center (Beijing, China). Normal mice were divided into 3 groups including [Saline], [SAFEs/Saline], and [SAFEs/EtOH]. The agent of [SAFEs/Saline] was prepared by dispersing the SAFEs particles into the saline solution. It should be noted that the agent of [SAFEs/EtOH] was obtained from the wet-milling utilizing EtOH as the liquid medium. Those mice were subcutaneously injected with 50 µL of Saline, SAFEs/Saline (50 mg/mL), and SAFEs/EtOH (50 mg/mL), respectively. The mice skin tissues of injection were photographed and collected at 24 h and 48 h. The histopathological tests of H&E staining were carried out according to a standard procedure to evaluate their inflammation

2.8. Leakage simulation of SAFEs in PBS

50 mg of SAFEs particles was loaded in a dialysis bag (500D) and then put in a beaker with 25 mL of deionized water. 500 μL of the leaked water solution taken out at 0 h, 6 h, 12 h, and 24 h was mixed intensively with 500 μL of HCl (2 mol/L). After reacting completely, 300 μL of the mixing solution was added into 3 mL of soluble starch standard solution (200 μg/mL) and the absorption of the mixed solution was measured by UV-Vis spectrometer. After determining the concentration of iodine in the mixed solution, the leakage ratio of iodine could be calculated (See the details in the Supporting Information).

2.9. In vivo photothermal-induced tumor ablation test

Tumor-bearing mice were prepared by implanting 100 μL of MCF-7 cell suspension (10⁷ cells/mL) at the right hind hip and raised for about 2 weeks until the tumor volume reached about 100 mm³. All the mice bearing MCF-7 tumors were divided into three groups including [Saline+NIR], [SAFEs], and [SAFEs+NIR] when the tumor size reached about 100 mm³.

The treatment procedure is shown in Fig. 6a. On day 0, the mice in [Saline+NIR] were given an intra-tumor injection with 50 µL of saline, and the other two groups were injected with 50 µL of SAFEs/ethanol (50 mg/mL). Notably, the single dose of injection has 0.782 mg of iodine content, which is still within the tolerable upper iodine uptake level of 1.1 mg/day [40]. On day 1 and day 3, all the mice in [Saline+NIR] and [SAFEs+NIR] received 808 nm laser irradiation with the power density of 1.0 W/cm² for 10 min, while no irradiation to those in [SAFEs]. On day 7, all the mice received the second drug administration with decreased dosage and the mice in [Saline+NIR] and [SAFEs+NIR] received the last irradiation. The tumor condition, the weight of mice, the lengths and widths of tumors were checked every other day. On day 21, all the mice were sacrificed, and the tumors were collected, weighed, photographed, and saved in formalin. At last, the histopathological tests of H&E staining were performed according to a standard procedure.

Fig. 6.

In vivo photothermal-induced tumor ablation test. (a) Schematic diagram of the tumor treatment procedure. (b) The relative tumor volume after various treatments. (c) Photographs and (d) the weight of the resected tumors from each group obtained after 21 days of treatments. *P < 0.05. P-values were calculated by independent samples t-test. (e) Histological images of H&E stained tumor sections collected from mice after various treatments. All data are presented as mean ± SD (n = 4).

In order to demonstrate the protection to normal tissue, a group of normal mice also received the same injection of SAFEs subcutaneously in the right hind hip as a control group. During the treatment, an infrared thermal camera was used to record their temperature changes (Fig. 5). After the first irradiation, the skin tissues that received the injection of SAFEs were collected and performed histopathological tests via H&E staining. All animal experiments were conducted in accordance with the protocols authorized by the Institutional Animal Care and Use Committee (IACUC) and in compliance with the legal requirements for laboratory animals in China.

Fig. 5.

In vivo photothermal test. (a) Time-dependent IR thermal images of MCF-7 tumor-bearing mice with Saline or SAFEs injection under NIR light irradiation. (b) Real-time IR thermal images of no tumor-bearing mice with SAFEs injection under NIR light irradiation. The power density of 808 nm laser was set as 1.0 W/cm². (c) The specific temperature changing curve of the above three groups. The photographs of (d, f) MCF-7 tumor-bearing mice and (e, g) no tumor-bearing mice after being injected SAFEs for 1 day. Histological images of H&E staining skin sections receiving SAFEs injection from (h) MCF-7 tumor-bearing mice and (i) no tumor-bearing mice.

3. Results and discussion

3.1. Preparation and characterization of the SAFEs

A scalable off-shoot of nanoparticle fabrication method termed ball-milling [41,42] was used to prepare SAFEs in large quantities. This top-down method minimizes the need for lengthy laboratory synthesis and tedious separation processes, owning attractive advantages in elegant simplicity and generality. The SAFEs was fabricated through a modified ball-milling procedure, which included freeze-drying, dry-milling, and wet-milling processes. As shown in Fig. 1b, [potato starch+KI+KIO₃] aqueous solution was typically prepared in deionized water which had removed the dissolved CO₂ to guarantee a low concentration of H⁺ in the solution. The two kinds of iodized salts could be distributed uniformly in the solid starch after lyophilization. Finally, SAFEs could be fabricated via successive dry- and wet-milling of those lyophilized starch. During the wet-milling, the liquid medium can facilitate the motion of solid starch among zirconia balls to induce more collision among them, thus leading to the SAFEs with a smaller size. Notably, ethanol was chosen as an ideal liquid medium here because both KI and KIO₃ cannot dissolve in ethanol ensuring the preservation of salts within starch. Besides, as a classical agent in percutaneous chemical ablation, ethanol could be also used in local intratumor injection to improve the delivery efficiency and accelerate the necrosis of tumor tissue [43,44]. After fabrication, the existence of potassium and iodine in SAFEs was detected by energy dispersive spectroscopy (EDS, Fig. S1), confirming the successful loading of salts within starch. The morphologies and average diameters of SAFEs nanoparticles were measured by dynamic light scattering (DLS, Fig. 1c) and scanning electron microscope (SEM, Fig. 1d). Noting that the data measured by DLS was the hydrodynamic diameter when SAFEs dispersed in PBS buffer solution, so the diameter in DLS is larger than that under SEM observation owing to the expansion and movement of partially hydrated starch chains. Consistently, there is an obvious Tyndall Effect in Fig. 1c depicting the existence of nanoparticles. What's more, the transformation of light absorption from “0” to “1” could be remarkably visualized by the color change when the system was added with a small aliquot of acetic acid (Fig. 1e, Movie S1). With the trigger of acid, SAFEs displayed an absolute transformation from non-absorption to a broad absorption window from visible light to the NIR region. This pH-triggered transformation of NIR absorption lays the groundwork for tumor-selective photothermal effect and further elimination of “photo-toxicity”.

3.2. Photothermal test of the SAFEs

The acidified SAFEs exhibited evident absorption in the visible light region with a wide shoulder extending to the NIR window (Fig. 1e), which renders the possibility to serve as PTA. To evaluate the photothermal effect of SAFEs, the SAFEs suspensions were subjected to NIR laser irradiation by which deeper tissue penetration is predicted. Accordingly, the real-time core temperature of solutions was recorded via an electronic thermometer and an infrared thermal camera (Fig. 2a). Keeping the molar ratio between KI and KIO₃ as 5: 1, a series of SAFEs consisting of starch and iodized salts with different mass fractions of iodized salts (${\omega }_{KI+KI{O}_3}$) were prepared (Table S1). According to the maximum steady-state temperature of SAFEs solutions relative to the ambient temperature (Fig. 2b), which was accomplished by placing SAFEs (5.0 mg/mL, pH 5.0) under 808 nm laser irradiation for 10 min, it is revealed that better photothermal effect is obtained when the mass fraction increases from 10% to 30%, but the photothermal conversion capability decreases upon further increasing the iodine content. The SAFEs with the mass fraction of 30 wt.% exhibits the best photothermal effect with the maximum temperature difference $\Delta T$ of 42.1 °C under the illumination of NIR light (808 nm, 1.0 W/cm²). Such a formula (30 wt.% of KI and KIO₃, and 70 wt.% starch) would be mainly focused for the further in vitro and in vivo investigations unless otherwise indicated. Previous studies suggested that starch can only combine with about 32 wt.% iodine for valid color development [45]. And when the mass fraction ${\omega }_{KI+KI{O}_3}$ equal 30 wt.%, the real iodine content ${\omega }_{I_2}$ is about 31.28 wt.% (See the calculation in Table S1). It's the reason why the photothermal effect begins worse when the ${\omega }_{KI+KI{O}_3}$ exceeds 30 wt.% in this SAFEs system because the color development is mainly determined by starch rather than iodine. The synchronous infrared thermal imaging provides a direct view of the temperature changes (Fig. 2c) which is consistent with the color changes upon increasing the mass fraction of iodized salts (Fig. S2). The temperature change led by the photothermal effect could be controlled via adjusting the power density of irradiation, where the heating is fastest as well as the largest maximum temperature at 1.0 W/cm² (Fig. 2d). And the $\Delta T$ is nearly proportional to the laser power density, which increases from 9.4 to 42.1 °C, when the laser power density is improved from 0.2 to 1.0 W/cm² (Fig. 2e,f). Noting that the temperature change at a fixed irradiation time remained almost unchanged after 6 cycles of repeated laser irradiation (Fig. S3), SAFEs possesses desirable photostability during the course of repeated photothermal therapy. For further exploring the photothermal effect of SAFEs, the photothermal conversion efficiency ($\eta$) of SAFEs dispersed in PBS buffer (pH 5.0) is assessed by Eq. 1 [46, 47]:

$$\begin{array}{c}\eta =\frac{hA\Delta T_{max}-P_s}{I\left(1-{10}^{-A_{\lambda }}\right)}\end{array}$$ (1)

Where $h$ is the heat transfer coefficient; $A$ is the surface area under laser irradiation; $\Delta T_{max}$ is the temperature difference of SAFEs solution between the maximum steady-state temperature and the ambient temperature; $P_s$ is the photothermal conversion power induced by the light absorbance of pure PBA solution;$I$ is the laser power and $A_{\lambda }$ is the absorbance of the SAFEs solution at the wavelength of 808 nm. The temperature increase and decrease were analyzed for a typical system of SAFEs in PBS buffer (Fig. 2g,i,f) and a contrast system of pure PBS buffer (Fig. 2h,j) under NIR irradiation at a power density of 1.0 W/cm². According to the cooling period, the photothermal conversion efficiency of this SAFEs solution was calculated as 29.07%, which is among the average level of many NIR PTAs [48] (See calculation details in the Supporting Information).

Although SAFEs displays a competitive photothermal efficiency in the buffer of pH 5.0, it's found that 5 mg/mL SAFEs cannot be well triggered to generate enough the I₂-starch complex as well as PTA with the desirable photothermal ability (Fig. S4). To further explore the pH response of SAFEs, the kinetics of SAFEs in weak acidic buffers were investigated (See the details in the Supporting Information). It's demonstrated that the reaction of SAFEs in an aqueous environment displays a reaction kinetics like a 3^rd order reaction:

$$\begin{array}{c}r=k_3\left[{\mathrm{H}}^{+}\right]{\left[\mathrm{I}{{\mathrm{O}}_3}^{-}\right]}^2\end{array}$$ (2)

Herein, $k_3$ (3.6 $\times$ 10⁷ L²·mol⁻²·h⁻¹, Figs. S5 and S6) represents the reaction rate constant of this 3^rd order reaction. It's notable that the composition of SAFEs is pre-designed with the fixed molar ratio of I⁻ and IO₃⁻ ([I⁻]: [IO₃⁻] = 5: 1).

For this system, whether SAFEs can be pH-triggered to generate enough I₂-starch complex as well as present obvious color change mainly depends on the amount of produced I₂ during the incubation time. Hence, the reaction rate is the most critical factor in determining the performance of SAFEs in an acidic environment. Through Eq. 2, the amount of the I₂-starch complex or the reaction rate of generating I₂ mainly depends on the actual concentration of H⁺, I⁻, and IO₃⁻ in the environment. In other words, the pH response of SAFEs can be regulated and controlled by adjusting the concentration of I⁻ and IO₃⁻ to adapt to various acidic environments with different concentrations of H⁺.

It's clear from the intuitive photographs in Fig. 3a that the generation of I₂-starch complex is highly dependent on the concentration of SAFEs. At a low concentration (5.0 mg/mL), SAFEs cannot result in an effective color change in the pH 6.0 buffer, but it can generate enough I₂-starch complex along with an obvious color change when rising the concentration of SAFEs to 2-folds or even 5-folds. When the concentration was set as 25 mg/mL, SAFEs can achieve a complete transformation of the light absorption from “0” to “1” (Fig. 3b), referring that it could not only result in a significant color change in pH 6.5, but also maintain the colorless condition in the neutral environment of pH 7.4. Their photothermal performance and the calculation of photothermal conversion efficiency at such a weak acidic buffer are also recorded in Figs. 3c and S7. By controlling the concentration of SAFEs, the photothermal ability can be adjusted to adapt to different acidic environments. As for the dose of 25 mg/mL, the maximum temperature can reach 55.0 °C even at pH 6.5, which has reached the temperature range of photothermal therapy to induce tumor cells apoptosis. Notably, the maximum temperature for the group of 25 mg/mL SAFEs in pH 6.5 buffer (55.0 °C) is similar to that in the group of 5.0 mg/mL SAFEs in pH 5.0 buffer (55.9 °C, Fig. 2), which is consistent with their similar reaction rate calculated by Eq. 2. It's demonstrated that the SAFEs with suitable design can achieve entirely distinct performances of light absorption and photothermal ability in different pH environments. In terms of the pH difference between normal tissues and tumor microenvironments, it's anticipated that the SAFEs would cause ignorable photo-damage to normal tissues under NIR laser irradiation, while in a tumor microenvironment, the photothermal effect can be activated and cause hyperthermia to kill tumor cells.

Fig. 3.

The photothermal test under weakly acidic buffer environment. (a) Representative photographs of SAFEs dispersed in PBS buffer with different pH values after incubation at 37 °C for 24 h. (b) UV-Vis absorption spectra of the SAFEs suspension in Fig. 3a with the concentration of 25 mg/mL. The suspensions under pH 5.0 and pH 6.5 were diluted 100 and 20 times, respectively. (c) IR images of SAFEs in the buffers (pH 6.0 and pH 6.5) under the laser irradiation with a power density of 1.0 W/cm² for 10 min.

3.3. Investigations of cytotoxicity and biocompatibility

To explore the potential application of SAFEs in biomedicine, the biosafety of SAFEs needs to be testified. Typically, the cytotoxicity of SAFEs was evaluated through the relative viabilities of Hela cells and human normal liver cells (L02) via the Cell Counting Kit-8 (CCK-8) assay [39]. All the cells were incubated with SAFEs at various concentrations for 24 h in 96-well plates. No cytotoxicity of SAFEs was found to cancer cells or normal cells as their cell viabilities were still maintained at about 100% in the presence of SAFEs with a high concentration of milligram level (2 mg/mL, Fig. 4a,b), confirming the biological safety of SAFEs which made of food-grade materials.

Fig. 4.

Investigations of cytotoxicity and biocompatibility. Cytotoxicity evaluations of (a) HeLa cells and (b) L02 human normal liver cells by Cell Counting Kit-8 (CCK-8) assay after incubation with various concentrations of SAFEs. (c) The representative pictures of no tumor-bearing mice and their histological images of H&E staining skin sections after receiving the subcutaneous injection of Saline, SAFEs/Saline and SAFEs/EtOH for different times. Black dot circles: injection area; *: the residual parts of SAFEs; black arrow: slight skin lesions; red solid circles: randomly chosen fields for inflammatory infiltration evaluation.

Furthermore, the biocompatibility of SAFEs was also evaluated by the inflammation tests. As shown in Fig. 4c, the inflammation of those nude mice with the injection of different agents were checked after the incubation of 24 h and 48 h. As expected, those mice in the group of SAFEs/Saline showed negligible lesions and inflammation during the 48 h of observation, which is the same as that in the Saline group. It proves again that the SAFEs itself made of food-grade materials is biologically safe. In order to obtain SAFEs nanoparticles with a smaller size for better endocytosis, ethanol was chosen as the liquid medium of wet-milling in the following in vivo tumor ablation tests. It shows that the ethanol would induce slight skin lesions and tolerable inflammation (Figs. 4c and S8), but this concern does not arise from the “material-toxicity” of SAFEs itself. The damage from ethanol could be mitigated via decreasing the dose or overcome by utilizing suitable wrapping or coating techniques. Combining the good biocompatibility of SAFEs without “material-toxicity” and “photo-toxicity” to normal tissue, it's expected that the SAFEs could be a safe PTA for efficient and selective tumor ablation in vivo.

3.4. In vivo photothermal-induced tumor ablation

Coupling the properties of excellent biocompatibility and smart pH-triggered photothermal performance, SAFEs is expected as superior performers of tumor ablation in vivo. Before in vivo test, the leakage of iodine in SAFEs is evaluated by a simulation experiment (see the Supporting Information for more details). It's demonstrated that it still maintains more than 70% of iodine content in SAFEs even after 24-hours of incubation time, which guarantees enough transformation of SAFEs in tumor tissues (Fig. S9). Besides, by incubating the tumor cells with fluorescein isothiocyanate (FITC) modified potato starch nanoparticles, it's demonstrated that the SAFEs nanoparticles could be ingested by tumor cells (Fig. S10). The cell membranes can engulf the nanoparticles and deliver them across the cell membrane into those endosomal vesicles with lower pH values, which further guarantees the success of pH-triggered photothermal effect.

Furthermore, a MCF-7 tumor-bearing mouse model was built to investigate the anti-tumor property of SAFEs [49,50]. To demonstrate the photothermal effect of SAFEs in response to the tumor microenvironment, female BALBc nude mice bearing MCF-7 tumors were intratumorally injected with 50 μL of saline and SAFEs while the normal mice also received the same injection of SAFEs subcutaneously in the right hind hip as a control group. According to IR thermal imaging results, the local temperature of tumor tissues where SAFEs was injected rapidly increases to 50.4 °C under 808 nm laser irradiation with the power density of 1.0 W/cm² for 10 min, displaying 29.6 °C higher than mice shell temperature without receiving laser irradiation (Fig. 5a). This high temperature is sufficient to destroy tumor cells in vivo. By contrast, in the saline group, the local temperature of tumor tissues injected with saline only reaches 36.6 °C, merely 8.8 °C higher than mice shell temperature. This slight warming phenomenon is mainly attributed to the inherent light absorption by skin tissue and blood vessel, which is unavoidable and of limited harm to healthy tissue under standard light power density. In the non-tumor group, the temperature change of normal mice injected with SAFEs under laser irradiation is the same as that of the saline group, suggesting SAFEs is not converted as a photothermal agent in homeostatic conditions (Fig. 5b,c). These in vivo studies present that SAFEs has no photothermal effect in healthy tissues (“0”) without the ability of photo-damage to normal cells, while it can be activated at an acidic tumor microenvironment and accordingly converts NIR light into hyperthermia (“1”), which causes lethal injury to tumor cells. In principle, this successful transition of the photothermal effect from “0” to “1” triggered by the tumor acidosis can minimize the “photo-toxicity” of PTA to normal tissue. More intuitive insights to confirm the “0-to-1” transition of SAFEs are provided in Fig. 5d-g. There is a notable color change in the tumor-bearing mice after the injection of SAFEs into the tumor for 24 h, indicating the successful generation of starch-iodine complex. However, no color change appears in the subcutaneous tissues of normal mice (Fig. 5e,g). To further rule out the possibility of inflammation-induced acidity, H&E staining skin sections were collected and investigated from the subcutaneous tissues near the injection site of SAFEs (Fig. 5h,i). Rarely inflammation could be observed in either normal tissue or tumor tissue, well agreeing with the cytotoxicity studies in regard to the good biocompatibility of SAFEs. It's envisioned that this particular class of pH-triggered SAFEs with double safety guarantees would give the maximum protection to normal cells while killing tumor cells via smart photothermal conversion.

When the tumor size reached about 100 mm³, three groups of MCF-7 tumor-bearing BALBc nude mice were treated with different therapy protocols in order to identify the in vivo anti-tumor therapeutic efficacy of SAFEs (Fig. 6a). One group of mice were intratumorally injected with saline that called [Saline+NIR] and the other two groups injected with SAFEs were termed as [SAFEs] and [SAFEs+NIR]. The two groups of [Saline+NIR] and [SAFEs+NIR] both received the 808 nm laser irradiation at 1.0 W/cm² while no irradiation was applied to the [SAFEs] group. After the treatment, the tumor size and the weight of mice in three different groups were checked every other day. As expected, the temperature of tumor in [SAFEs+NIR] group reached to above 50 °C, much higher than that in [Saline+NIR] group (37.1 °C, Fig. S11). It's found that the tumor treated with the combination of SAFEs and laser irradiation were effectively ablated after 21 days of treatment. Particularly, as the scars at the site of irradiation minimized gradually, the tumor size shrank or even disappeared without recurrence during the 21 days of treatment course (Figs. 6b and S12). By contrast, tumors in the two control groups showed similar growth speed, suggesting irradiation or SAFEs injection by itself cannot inhibit the tumor development (Fig. 6b).

Furthermore, all the tumors were collected after the finish of treatment and observation. As shown in Fig. 6c,d, irradiation or SAFEs injection by itself exhibited similar suppression of tumor growth, and the tumor weights collected from these two groups are very close. The best in vivo therapeutic efficacy was observed in the group of [SAFEs+NIR], and tumors in some cases are completely destroyed that cannot be found after dissection. H&E staining of tumor sections were performed on day 21 to investigate the fate of tumor cells after experiencing treatment in the three groups (Fig. 6e). In the group of [Saline+NIR] and [SAFEs], there is no obvious cell necrosis or apoptosis, and the tumor cells have high cell density with normal morphology. While in [SAFEs+NIR] group, histological sections show an obvious decrease in tumor cells density. Moreover, the body weights of mice in the treatment group kept at a steady state after receiving the last irradiation, demonstrating the good biocompatibility and low toxicity of photothermal therapy to mice (Fig. S13). These results convinced the excellent therapeutic efficacy of SAFEs, offering great potential in serving SAFEs as a safe PTA for selective tumor ablation in clinic.

4. Conclusion

A safe PTA with double safety guarantees (no “material-toxicity” and no “photo-toxicity”) was invented via formulating foods into intelligent photothermal transduction agents in response to the tumor environment. In this work, a macroscopic consecutive reaction was designed through the [starch-KI-KIO₃] complex, which can generate I₂ after being triggered by tumor acidosis. And then the iodine will be complexed by starch, showing a transformation of NIR absorption and photothermal effect from “0” to “1”. A scalable fabrication method of modified ball-milling was put forward to prepare the SAFEs in large quantities that can get SAFEs particles or SAFEs/ethanol nanoparticles suspension with smaller size after wet-milling. The pH-triggered photothermal effect was verified at both in vitro test and in vivo tumor ablation. Specifically, the local temperature of tumor tissues injected with SAFEs can dramatically reach 50.4 °C within 10 min while no obvious heating existed in normal tissues (36.0 °C). From cytotoxicity studies and histological staining, no “material-toxicity” was found in this system, indicating the good biocompatibility of this food-grade PTA. Further in vivo studies validated the discriminative photoinduced damage to tumor regions with excellent tumor treatment efficacy while no “photo-toxicity” to normal tissues. PTA with double safety guarantees proposed in this work is still at an infancy stage and has some problems before clinical trials on human bodies, but it opens up a new way to develop safe PTA via utilizing food-grade materials and designing consecutive reactions triggered by the tumor environment. Future efforts will be devoted to solving the leakage of iodized salts from the starch matrix via suitable wrapping or improving the targeting capacity to ensure PTA could be accumulated precisely in tumor areas via intravenous injection. It is envisioned that this tumor microenvironment-triggered photothermal strategy based on food materials can minimize the “material-toxicity” and “photo-toxicity” to normal tissues, which might lead to a new class of safer and more selective anti-tumor therapy, thus further promoting the clinical translation of PTT.

Declaration of competing interest

The authors declare that they have no conflicts of interest in this work.

Biographies

Hongguang Liao received his B.S. in chemistry and M.S. in chemistry and physics of polymers from Renmin University of China under the supervision of Prof. Yapei Wang, in 2017 and 2020 respectively. He is currently a Ph.D. student in the Graduate School of Life Science, Hokkaido University. He is dedicated to developing novel photothermal conversion materials and their clinical applications.

Yapei Wang(BRID: 08359.00.61662) is a full professor at the School of Chemistry and Life Resources, Renmin University of China. He received his B.S. from Jilin University in 2004 and Ph.D. from Tsinghua University under the supervision of Prof. Xi Zhang in 2009, both in chemistry. After graduation, he spent 2 years as a postdoctoral fellow in Prof. Joseph M. DeSimone's laboratory at the University of North Carolina at Chapel Hill. In 2012, he joined the Department of Chemistry, Renmin University of China. His research interests are mainly focused on photothermal conversion materials, liquid electronics, and interface science.