Abstract

The high prevalence of breast cancer is a global health concern, compounded by the lack of safe or effective treatments for its advanced stages. These facts urge the development of novel treatment strategies. Annexin A5 (ANXA5) is a natural human protein that binds with high specificity to phosphatidylserine, a phospholipid tightly maintained in the inner leaflet of the cell membrane on most healthy cells but externalized in tumor cells and the tumor vasculature. Here, we have developed a targeted photosensitizer for photothermal therapy (PTT) of solid tumors through the functionalization of single-walled carbon nanotubes (SWCNTs) to ANXA5—the SWCNT-ANXA5 conjugate. The ablation of tumors through the SWCNT-ANXA5-mediated PTT synergizes with checkpoint inhibition, creating a systemic anticancer immune response. In vitro ablation of cells incubated with the conjugate promoted cell death in a dose-dependent and targeted manner. This treatment strategy was tested in vivo with the orthotopic EMT6 breast tumor model in female balb/cJ mice. Enhanced therapeutic effects were achieved by using intratumoral injection of the conjugate and treating tumors at a lower PTT temperature (45°C). Intratumoral injection prevented the accumulation of the SWCNTs in major clearance organs. When combined with checkpoint inhibition of anti-programmed cell death protein-1, SWCNT-ANXA5-mediated PTT increased survival and 80% of the mice survived for 100 days. Evidence of immune system activation by flow cytometry of splenic cells strengthens the hypothesis of an abscopal effect as a mechanism of prolonged survival.

SIGNIFICANCE STATEMENT This study demonstrated a relatively high survival rate (80% at 100 days) of mice with aggressive breast cancer when treated with photothermal therapy using the SWCNT-ANXA5 conjugate injected intratumorally and combined with immune stimulation using the anti-programmed cell death protein-1 checkpoint inhibitor. Photothermal therapy was accomplished by maintaining the tumor temperature at a relatively low level of 45°C and avoiding accumulation of the nanotubes in the clearance organs by using intratumoral administration.

Introduction

Cancer incidence has been growing worldwide as has its global mortality. Accordingly, cancer is predicted to be the primary cause of mortality and, thus, the single most critical barrier to improving life expectancy in every country in the 21st century, according to the World Health Organization (Bray et al., 2018). More specifically, one in every 20 women is affected by breast cancer globally, and this statistic rises to one in every eight women until age 85 who live in developed countries (Fitzmaurice and Global Burden of Disease Cancer Collaboration, 2018). The absence of targeting receptor expression characterizes triple-negative breast cancer (TNBC). It affects approximately 10–20% of the women diagnosed with breast cancer (Costa and Gradishar, 2017; Siegel et al., 2019), and its poor prognosis is strongly associated with the lack of targeted therapies. The limited treatment options also increase the incidence of cancer recurrence in TNBC patients (Yin et al., 2020).

Anti-programmed cell death protein 1 (PD-1) checkpoint inhibition has revolutionized cancer treatment by binding to the PD-1 receptor on T-cells, which disrupts its interaction with its ligands overexpressed in cancer cells, PD-L1, and PD-L2, thereby attenuating the inhibitory signal that normally downregulates T-cell activation. This blockade effectively unleashes the cytotoxic potential of T cells to eliminate tumor cells. The first approved breast cancer immunotherapy was pembrolizumab (anti-PD-1), which when combined with neoadjuvant chemotherapy for TNBC improved pathologic complete response rates and event-free survival. Subsequent clinical trials explored other such combinations (e.g., NCT03752723, NCT02730130, NCT02513472, NCT03106415). However, some tumors, including breast cancer, often lack responsiveness to anti-PD-1 (Linette and Carreno, 2019). Therefore, further research is required to improve safety and efficacy, especially when the effectiveness of combinations falls short of optimal results.

Although cancers evade immune destruction, immunogenic cell death (ICD) may shift the balance toward antitumor immunity (Chen and Mellman, 2013). Ablating solid tumors can trigger ICD, inducing immune responses by releasing danger signals and tumor antigens. The promising combination of an ablation method, such as photothermal therapy (PTT), with immunotherapy may produce a synergistic effect that treats the primary tumor and suppresses metastasis by fostering a systemic immune response. PTT promotes ICD-associated immune responses and potentially catalyzes cytotoxic T-cell activation and immunologic memory (Galluzzi et al., 2017).

Here, we combine photosensitizer-mediated PTT ablation with anti-PD-1 checkpoint inhibition for a systemic anti-cancer response. PTT-mediated hyperthermia causes tissue damage above 41°C (Xie et al., 2011). However, effective tumor ablation requires temperatures of 45–60°C due to heat transfer and laser penetration. (Naylor et al., 2006; Hsiao et al., 2015; Xiang et al., 2015; Zhao et al., 2022). This phenomenon triggers ICD that assists in tumor removal by activating immune cells and also induces damage-associated molecular patterns release, which enhances native tumor antigen immunogenicity, while pro-inflammatory cytokines attract and activate immune cells (Li et al., 2020; Maruoka et al., 2021). PTT’s dual selectivity—region-specific laser targeting and photosensitizer localization—creates a highly targeted treatment option. Invasive techniques can be avoided when treating deep-seated tumors with optical fiber-equipped endoscopy and interventional procedures (Pacella et al., 2011).

In this work, we have developed a targeted photosensitizer for PTT for the efficient ablation of solid tumors based on the functionalization of single-walled carbon nanotubes (SWCNTs) to annexin A5 (ANXA5): SWCNT-ANXA5 conjugate. SWCNTs strongly absorb near-infrared light in the range at which water and hemoglobin have minimum absorption, making them a suitable photosensitizer for PTT (Braun and Smirnov, 1993; Kastner et al., 1994; Weissleder, 2001). More specifically, the (6,5)SWCNT, characterized by a chiral angle 6,5 of the hypothetic graphene sheet, has a peak absorbance at 980 nm, the wavelength used for the laser irradiation. ANXA5 is a protein from the family of annexins, which can bind to membranes expressing phosphatidylserine (PS) in a calcium-dependent manner (Lizarbe et al., 2013). It has specificity for cancer due to the strong binding to externalized PS on the surface of tumoral cells and tumor vasculature (Ran and Thorpe, 2002; Ran et al., 2002; Riedl et al., 2011). PS is a biomarker of various types of cancers, including TNBC, but is absent on the surface of healthy, non-apoptotic cells. As shown by Ran and Thorpe (2002) and Ran et al. (2002), PS expression on the luminal surface of vascular endothelial cells was absent in all of the ten measured organs, while PS expression was detected in six types of tumor tissues. Thus, the SWCNT-ANXA5 conjugate efficiently captures near-infrared (NIR) light and selectively elevates the temperature of tumors when bound specifically to cancer cells. The conjugate was produced using a hetero-bifunctional linker 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-[maleimide(polyethylene glycol)-3400] (known as DSPE-PEG-MAL). The DSPE component in the linker attaches to the SWCNT surface because of its hydrocarbon chains, given the surface’s strong hydrophobic nature. Meanwhile, the MAL part specifically reacts with the one cysteine of ANXA5 due to a thiol-maleimide interaction, creating succinimidyl thioether. PEG is incorporated to prevent proteins from clustering on the nanotube’s surface, which enhances its longevity and compatibility with biologic systems. In previous work, the systemic administration of SWCNT-ANXA5 led to the accumulation and lingering clearance of SWCNT in major organs, mainly the liver, kidney, and spleen (McKernan et al., 2021); thus, in this work, we used intratumoral injection of the conjugate to avoid that drawback. In this paper, we determined the optimized temperature for the activation of the immune system from PTT, and we assessed the combination of the SWCNT-ANXA5-mediated PTT with checkpoint inhibition of anti-PD-1, evaluating the mechanisms of anticancer immunity in the combination therapy.

Materials and MethodsMaterials.

For ANXA5 protein production and purification, the plasmid encoding ANXA5, pET-30 Ek/LIC/ANX, was previously constructed in this laboratory (Neves et al., 2013). Sodium hydroxide, sodium chloride, tryptone, kanamycin, yeast extract, isopropyl-beta-D-thiogalactopyranoside, N-p-tosyl-L-phenylalanine chloromethyl ketone, phenylmethylsulfonyl fluoride, β-mercaptoethanol, and sodium phosphate dibasic (Na2HPO4) were from Sigma Aldrich (St. Louis, MO). HRV 3C protease was from Acro Biosystems (Newark, DE). The HisTrap chromatography column (5 ml) was from Cytiva (Marlborough, MA). The 12–14 kDa regenerated cellulose dialysis membrane was from Fisher Scientific (Pittsburgh, PA). Laemmli sample buffer, ladder (marker), 4–20% mini-protean TGX stain-free pre-casted polyacrylate gels, Bio-Rad 10x Tris-glycine-SDS buffer, and Bradford dye reagent were from Bio-Rad Laboratories (Hercules, CA). Imperial protein stain was from Thermo Scientific (Waltham, MA). For conjugation of ANXA5 to SWCNTs, purified and freeze-dried (6,5) CoMoCAT SG65i SWCNTs were obtained from Chasm Advanced Material Company (Norman, Oklahoma), with a tubular carbon purity of ≥97% as determined by thermogravimetric analysis, Raman Q factor ≥0.97, an average length of 1 µm, an outer diameter of 0.78 ± 2 nm, and a (6,5) chirality composition of ≥40%, as determined by NIR fluorescence spectroscopy. DSPE-PEG-MAL (3.4 kDa) was from Creative PEGWorks (Winston Salem, NC). Spectra-Por dialysis membranes (2 and 100 kDa) were from Spectrum Laboratories, Inc. (Rancho Dominguez, CA). L-cysteine amino acid was from Sigma Aldrich. Bradford protein reagent was from Bio-Rad (Hercules, CA). All cell lines and cell media were from American Type Culture Collection (Manassas, VA, USA). Waymouth’s MB 752/1 medium), was from Gibco (Thermo Fisher, Waltham, MA). Antibiotic-antimycotic (10,000 IU penicillin, 10,000 μg/ml streptomycin, and 25 μg/ml amphotericin B) was from Corning (Kennebunk, MA, USA). Vascular cell basal medium and endothelial cell growth kit BBE (0.2% bovine brain extract, 5 ng/ml EGF, 10 mM L-glutamine, 0.75 units/ml heparin sulfate, 1 μg/ml hydrocortisone, 50 μg/ml ascorbic acid, 2% fetal bovine serum (FBS)) were from the American Type Culture Collection. Trypsin/EDTA, Nunc Laboratory-Tek chamber slides, and Alexa Fluor 488 were from Thermo Fisher Scientific. Flourogel was from Electron Microscopy Sciences (Hatfield, PA). For in vivo studies, BALB/cJ mice were from Jackson Laboratory (Bar Harbor, ME). ELISA kits and flow cytometry staining antibodies were from Biolegend (San Diego, CA). CD16/CD32 Fc Block was from eBioscience, San Diego, CA, USA. Anti-PD-1 was from Bio X Cell (Lebanon, NH).

Annexin A5 Production and Purification.

Recombinant ANXA5 was produced as previously described (Zang et al., 2006). In brief, BL21(DE3) Escherichia coli harboring the plasmid containing pET-30 Ek/LIC/ANXA5 were initially inoculated and grown in Terrific broth medium with kanamycin (35 μg/ml) and then induced to produce ANXA5. The bacteria were collected and sonicated. The debris-free supernatant from cell lysate was loaded into a nickel HisTrap column. After washing, an endotoxin removal step was added using a 1% Triton X-100 wash (Van Rite et al., 2013). After another wash, ANXA5 protein (with an N-terminal six histidine tail) was eluted. After dialysis against sodium phosphate buffer, the (His)6-tagged protein was cleaved with the HRV 3C protease. Final column purification was performed by loading cleaved protein into the affinity column and collecting the first flow-through with ANXA5 without the (His)6 tag. The protein solution was dialyzed against a 20 mM sodium phosphate buffer containing 100 mM NaCl (pH 7.4) a final time before being aliquoted in cryogenic vials and flash-frozen in liquid nitrogen. Aliquots were placed in a -80°C freezer for long-term storage. The purified protein was quantified using the Bradford assay and analyzed with SDS-PAGE electrophoresis for purity.

Conjugation of SWCNTs to Annexin A5.

The conjugation was performed using the linker DSPE-PEG-MAL (3.4 kD) as previously described (Neves et al., 2013). Briefly, SWCNTs (6 mg) were added to 5 ml of a 1% SDS solution and sonicated for 1 hour at 19.8 W of power (E = 35,640 J) (VirSonic 100 ultrasonic cell disruptor, VirTis). Sonicated SWCNTs were centrifuged in an ultracentrifuge (16,000 g) for 1 hour. The DSPE-PEG-MAL linker (1.5 mg) dissolved in 1% SDS solution (1 ml) was added to 5 ml of the SWCNT suspension and then mixed at room temperature for 30 minutes with gentle shaking. The suspension was dialyzed in a 2-kDa dialysis membrane for 8 hours against 3 L of deionized water. The dialysate was changed after the first 4 hours. SWCNT-linker suspension (2 ml) was mixed with the concentrated ANXA5 (1 ml of 5 mg/ml totaling 5 mg of ANXA5) at room temperature for 2 hours with gentle shaking. Any unreacted linker sites were blocked with L-cysteine, added to the suspension and allowed to react for 1 hour at room temperature with gentle shaking. The suspension was then dialyzed using a 100 kDa dialysis membrane to remove any unbound ANXA5 or L-cysteine for 8 hours against 3 L of 20 mM sodium phosphate buffer (pH 7.4) with a dialysate change after 4 hours. The final suspension was centrifuged at 16,000 g for 1 hour to remove any aggregates. The conjugate was stored in a glass vial at 4°C until use. The conjugate was used within 1 week of the end of conjugation.

Cell Culture.

EMT6 murine breast carcinoma cells were cultivated with Waymouth’s MB 752/1 medium supplemented with 2 mM glutamine, 15% FBS, and 1% antibiotic-antimycotic (10,000 IU penicillin, 10,000 μg/ml streptomycin, and 25 μg/ml amphotericin B). Human umbilical vein endothelial cells (HUVECs) were cultured in vascular cell basal medium supplemented with endothelial cell growth kit BBE (0.2% bovine brain extract, 5 ng/ml EGF, 10 mM L-glutamine, 0.75 units/ml heparin sulfate, 1 μg/ml hydrocortisone, 50 μg/ml ascorbic acid, and 2% FBS) and 1% antibiotic-antimycotic. Cells were passaged using 0.25% (w/v) trypsin in 0.53 mM EDTA before reaching 85% confluency. All cell lines were cultured under a 5% CO2-supplemented atmosphere at 37°C. The medium was refreshed every 48 hours.

In Vitro Studies.

To assess SWCNT-ANXA5-mediated PTT in vitro, EMT6 cells at 70% confluency and HUVEC cells at 100% confluency to mimic healthy cells were previously grown in T-75 flasks and were seeded at 2 × 105 into sterile surface-treated 24-well plates. At 24 hours after seeding, cell media was aspirated, and fresh media enriched with 6 mg/L SWCNT-ANXA5 suspension and 2 mM of Ca2+ (CaCl2) was added to the wells, due to calcium-dependent binding of ANXA5 to PS. Incubation was performed for 2 hours at 37°C. The wells were washed three times with sterile phosphate buffered saline (PBS) to remove unbound SWCNT-ANXA5. Each group was evaluated at least in triplicate. Different controls were studied to guarantee the specificity of the assay. Here, cells were treated similarly to previous studies by Neves et al. (Neves et al., 2013).

For PTT treatment, each well subjected to the laser treatment was irradiated for different times at 980 nm-irradiation at a power density of 1 W/cm2 using a Diodevet-50 NIR laser (B&W Tek Inc., Newark, DE). The irradiation was always performed by having the plate sitting on top of a platform and aiming the optical fiber at the bottom of the plate, positioned a few centimeters away from the bottom to obtain the desired size beam. Each well was carefully adjusted to be centered with the laser beam. The beam size was adjusted using a laser-sensitive paper (Zap-it paper), which burns when exposed to a laser beam. A power meter (Ophir Optronics, Ltd., Israel) determined the power attenuation from the fiber and plates. The power meter probe was placed on top of a microtiter plate (without the lid), and the power was determined to reach a power level of 1 W/m2. The irradiation was performed at room temperature or inside an incubator at 37°C to mimic physiologic temperatures.

The Alamar Blue assay was used to determine the cell viability after PTT. The cells were incubated with the reagent for 2–4 hours at 37°C until the coloration of the media in untreated control wells turned pink. Changes in fluorescence were then quantified with a microplate reader (excitation wavelength at 530 nm and emission at 590 nm). The viability was determined by normalizing the results to the average fluorescence of the untreated control wells, which is set to 100%, giving each well the relative viability. The average background fluorescence (no cell wells) was also subtracted from sample fluorescence. Temperature measurement of the cell media bulk temperature was performed with a type J thermocouple inserted into the bottom of the well.

For cell staining, the cells were then seeded on a chamber microscope slide (Nunc Laboratory-Tek) and were allowed to grow for 24 hours. The cells were then incubated for 2 hours at 37°C with Alexa Fluor 488-labeled SWNT-ANXA5 at a concentration of 6 mg/l. This was followed by washing four times with Waymouth’s MB 752/1 medium supplemented with 2 mM glutamine, 15% FBS, and 1% antibiotic-antimycotic (10,000 IU penicillin, 10,000 μg/ml streptomycin, and 25 μg/ml amphotericin B) and 2 mM Ca 2+. The cells were then fixed with 0.25% glutaraldehyde. A drop of an antifade reagent (Fluoro-gel with TES buffer) was placed on a microscope cover slip, which was then placed on top of the sample to preserve the fluorescence. Fluorescent images were taken by a Leica SP8 confocal laser scanning microscope with a 63x oil objective (1.40 NA). Annexin A5-AlexaFluor 488 conjugates were detected by an argon laser with the excitation peak at 488 nm and the emission window between 500 nm and 570 nm. Transmission light images were taken simultaneously to show the overall cell morphologies. Z-series images were collected with Z-step size at system optimized size. Fluorescence pictures were taken using a Leica SP8 Upright CLS/multiphoton/FLIM microscope. The images were processed using Fiji ImageJ.

Animal Handling Procedures.

All animal studies were performed following the protocols approved by the Institutional Animal Care and Use Committee of the University of Oklahoma and conducted by staff with proper training. Animals were housed in a pathogen-free facility at the University of Oklahoma and monitored daily.

In Vivo Studies.

For tumor induction, six-week-old female BALB/cJ mice were injected with 1x106 EMT6 cells suspended in 100 μl PBS using a 30-gauge needle. The injection was performed subcutaneously in the fourth mammary fat pad, close to the nipple, which characterizes an orthotopic tumor. Mice weight and tumor volumes were measured every 3–5 days during each study. Tumor volumes were measured using a caliper, and volume (V) was determined through the modified ellipsoid formula V = (1/2) × (L×W2), where length (L) is the longest diameter of the tumor and the perpendicular diameter is the width (W). For all the studies, mouse health was assessed every 3–5 days for signs of distress. Mice were euthanized when the tumor size was greater than 15 mm in any of the measurement directions or when the mice were sick, characterized by dehydration, weight loss greater than 20%, recumbent posture, breathing difficulty, or loss of leg function due to tumor proximity. Mice were fed a standard chow diet.

During photothermal irradiation of tumors with an NIR laser light, mice were anesthetized with 3% isoflurane and 97% oxygen using a nose cone. Treatments were performed when tumors reached a size of ∼5 mm (approximately 60 mm3). SWCNT-ANXA5 injections were performed at a 1.2 mg/kg dose of SWCNT. Breast tumor-bearing mice were treated with PTT similarly to previous studies by Neves et al. (Neves et al., 2013) and McKernan et al.(McKernan et al., 2021), using a previously established safe laser irradiation level of 1 W/cm2.

For PTT, a 980 nm NIR laser was vertically mounted atop the stage where the mice were to be irradiated, its tip precisely targeted at the tumor site (see Fig. 2A). Prior to irradiation, the power attenuation of the laser fiber was measured using a power meter (Ophir Optronics, Ltd., Israel). The emitted signal was calibrated to a power output of 1.7 W by adjusting the settings on the generator. The laser tip was inserted into an opaque, hollow tube with open ends, with an internal surface area of 1.7 cm2 to control the irradiation area. This setup achieved a power density of 1 W/cm2. The anesthetized mice were carefully positioned on their backs to expose the orthotopic tumor in the abdominal region’s lower left quadrant. Each mouse was aligned so that the tumor was centered within the designated irradiation area. Tumor temperature was monitored using a FLIR E5-XT thermal camera, with the emissivity setting adjusted to 0.97 to accurately reflect the thermal properties of mouse skin (Ratko et al., 2020).

To evaluate the cytokine release related to tumor surface temperature, EMT6-tumor-bearing mice were injected intratumorally with SWCNT-ANXA5 when the tumors were approximately 5 mm in diameter (∼60 mm3) 12 days after tumor injection. The tumors were treated by a 980 nm NIR laser at 1 W/cm2. The tumor surface temperature was monitored during the irradiation by a thermal camera (FLIR E5-XT). Irradiation ceased immediately after the tumor surface reached the temperature assigned to the groups of 45, 50, 55, or 60°C. Mice were euthanized either 1 or 7 days after PTT for blood collection (respectively 24 hours and 168 hours). The concentration of four pro-inflammatory cytokines was measured in the serum by ELISA. The control group was intratumorally injected with SWCNT-ANXA5 but not treated with PTT.

To assess the effects of SWCNT-ANXA5-mediated PTT at 45°C in combination with anti-PD-1, a long-term survival test was performed. Mice were inoculated with EMT6 tumors on day 0. On day 7, when tumors were around 3 mm in diameter, treatment started for the assigned groups with an intraperitoneal injection of anti-PD-1 checkpoint inhibitor at 10 mg/kg. The same injection was performed two more times on days 10 and 15. On day 11, mice were injected intratumorally with SWCNT-ANXA5 when the tumors were around 5 mm in diameter (∼60 mm3) and treated with PTT. The tumors were treated by the 980 nm NIR laser at 1 W/cm2. The tumor surface temperature was monitored during the irradiation by a thermal camera (FLIR e-5). Irradiation immediately ceased when the tumor surface temperature reached 45°C. Mice’s health was monitored for 100 days after tumor induction, when the study was terminated.

For the mechanistic analysis of the combination therapy, mice were treated following the same protocol for the long-term survival test. However, the mice were euthanized by CO2 asphyxiation 14 days after PTT, when the spleens were harvested to quantify antitumoral immune effector cells. For that, spleens were kept in Dulbecco’s modified Eagle’s medium media on ice until use. Spleens were mechanically dissociated with a syringe plunger and strained through a 70 μm filter with 3 ml of flow cytometry staining buffer. Strained cells were centrifuged at 1000 g for 5 minutes, and the supernatant was discarded. Another 3 ml of flow cytometry buffer was added to resuspend and centrifuge the cells. The total cell count of each spleen was determined by a hemocytometer. Cells were counted and resuspended in flow cytometry buffer to a final concentration of 10 × 106 cells ml−1. Cells (100 μl) were blocked with CD16/CD32 Fc Block for 30 minutes at 4°C and stained for PE-Cy7-CD3, APC-CD8, and fluorescein isothiocyanate-CD4 with 50 μl antibody mixture for 1 hour at 4°C, protected from light. Cells were washed three times with 1 ml of permeabilization buffer followed by 5 minutes of centrifuging at 1000 g; after the last wash, cells were suspended in 500 μl of flow cytometry staining buffer and analyzed via a BD Accuri C6 flow cytometer (San Jose, CA). For flow cytometry analysis, lymphocytes were distinguished from other populations by plotting side scatter height versus forward scatter height and selecting the distinguished population around the area between 500,000 and 1,0000,000. Gated cells were plotted in forward scatter area versus forward scatter height to exclude multiplets (non-single cells). CD3+ single cells were gated positive by excluding ∼98% PE-Cy7-anti-CD3 non-stained (control) cells. Then, cells positive for either CD4 or CD8 were selected by excluding the 98% of unstained cells for fluorescein isothiocyanate-anti-CD4 and APC-anti-CD8, respectively. For each individual mouse, the percentage of CD3+CD4+ and CD3+CD8+ from total cells was recorded and multiplied by the total cell count obtained at the beginning of the experiment to obtain the total amount of helper T-cells and cytotoxic T-cells, respectively.

To improve the efficiency of the PTT at 45°C, a tumor control study was performed with prolonged irradiation times. Tumor-bearing mice were injected intratumorally with SWCNT-ANXA5 when the tumors were around ∼100 mm3. The tumors were treated by the 980 nm NIR laser at 1 W/cm2. Tumors were irradiated until the surface temperature reached 45°C and then held the temperature at 45°C ± 3°C by repeatedly turning the laser off and on. The turning off and on cycles were repeated for the total time of the irradiation session of 0, 1, 2 and, 5 minutes. After irradiation sessions, tumor volumes and body weights were monitored daily. The experiment was finalized 15 days after PTT when the remaining mice were euthanized.

To analyze the combination therapy of anti-PD-1 and PTT at 45°C for 5 minutes, tumor-bearing mice were treated with three injections of anti-PD-1 (intraperitoneal injection on days 7, 10, and 15 at 10 mg/kg). The tumors were treated by 980 nm NIR laser at 1 W/cm2 after intratumoral injection of SWCNT-ANXA5 (1.2 mg/kg) on day 11. The turning off and on cycles were repeated for the irradiation session of 5 minutes. The group that received irradiation for 5 minutes was compared with an untreated control group that did not receive any treatment and another group treated with anti-PD-1 and PTT, but the irradiation stopped immediately when the tumor surface temperature reached 45°C.

Ex Vivo SWCNT Detection for Biodistribution Analysis.

BALBc/J mice were inoculated with EMT6 tumors and were injected intratumorally with SWCNT-ANXA5 (1.2 mg/kg of SWCNTs) when the tumors were around 5 mm in diameter (∼60 mm3). The tumors were treated by the 980 nm NIR laser at 1 W/cm2. Tumors were irradiated until the surface temperature reached 45°C. Mice were euthanized by CO2 asphyxiation 1 week after PTT. Major organs were harvested and weighed. Organs were frozen and stored in a −80°C freezer. Tissue lysates for analysis of SWCNT concentration were prepared according to the protocol from Liu et al.(Liu et al., 2008). Briefly, after thawing samples, organs were individually ground in a tissue grinder added to a lysing buffer composed of 1% SDS, 1% Triton X-100, 10 nM dithiothreitol, and 40 mM tris-acetate-EDTA buffer to complete volume of 10 ml. Tissue was digested overnight in an incubator at 70°C. Each sample was sonicated for 1 hour to resuspend SWCNTs from digested tissue. Samples were analyzed using intrinsic SWCNT fluorescence by single excitation fluorescence measurement using an NS MiniTracer (Applied Nanofluorescence, Houston, TX).

Statistical Analysis.

The statistical significance of in vitro results was assessed using an unpaired t test with Welch’s correction for assumed unequal variances for experiments with only two groups. One-way ANOVA with Dunnett’s multiple comparisons was used for tests with more than two groups, comparing treated groups to the control group. The statistical significance of cytokine levels in serum, tumor volumes, and splenic cell counts in the in vivo studies was assessed by one-way ANOVA with Dunnett’s multiple comparisons comparing treated groups to the control group. In survival experiments with only three groups, one-way ANOVA with Tukey’s multiple comparisons comparing every pair of groups was used to assess statistical analysis between tumor volumes. The Mantel-Haenszel log-rank test determined the statistical significance of survival curves by comparing treated groups to the untreated control. All analyses were carried out in GraphPad Prism 8 software.

ResultsSWCNT-ANXA5 Conjugation.

Recombinant annexin A5 was successfully produced in Escherichia coli, purified by affinity chromatography, and analyzed by SDS-PAGE electrophoresis. Conjugation of SWCNT to ANXA5 was performed using DSPE-PEG-MAL (3.4 kDa). After purification, the mass ratio of ANXA5 to SWCNT in conjugate was determined by spectrophotometry. The ANXA5 concentration was determined by the Bradford assay (absorbance at 595 nm) and compared with a bovine serum albumin protein standard curve. As a control, the absorbance of PEGylated SWCNT (addition of the DSPE-PEG-maleimide linker to suspended SWCNT and 8-hour dialysis against deionized water) at 595 nm was measured and showed minimal absorption due to the lack of protein. The concentration of SWCNT was determined by measuring the endpoint absorbance at 800 nm and compared with the standard curve. After measuring absorbance of ANXA5 at 800 nm, no significant absorbance was seen at that wavelength. The average ANXA5/SWCNT weight ratio of the different batches of conjugates used for the in vitro and in vivo experiments presented in this study was 28 ± 6.6. Atomic force microscopy of this conjugate was obtained previously (Neves et al., 2013).

In Vitro Studies.

EMT6 and HUVEC cell viabilities were used to assess the efficiency and specificity of SWCNT-ANXA5-mediated PTT in vitro. In Fig. 1A, the NIR-laser treatment at room temperature induced a significant decrease in cell viability after 3 minutes of irradiation when the EMT6 cells were incubated for a short period (2 hours) and the unbound conjugate was washed off, compared with the control group where the cells were not incubated with SWCNT-ANXA5. The decline in viability was not observed in the control group consisting of EMT6 cells irradiated with the laser and absent of the conjugate, even in higher irradiation times. In a parallel experiment performed under the same conditions, the bulk temperature of the cell media after laser irradiation was measured (Fig. 1B). The cell media of EMT6 treated with SWCNT-ANXA5-mediated PTT showed an increased macroscale temperature elevation compared with the control treated with laser irradiation alone caused by the presence of SWCNT bound to the surface of the microscale cell monolayer. By comparing the results from Fig. 1, A and B and assuming a conservative room temperature of 21°C, the higher macroscale temperature reached in those experiments was around 32°C, which induced cell death in 83% of cells.

Fig. 1.

Cytotoxicity of SWCNT-ANXA5-mediated PTT in vitro. Cells were incubated for 2 hours with the SWCNT-ANXA5 or not incubated with it as a control prior to laser irradiation. After the cells were washed four times with PBS, laser irradiation (980 nm) was carried out at 1.0 W/cm2. (A) Relative cell viability of EMT6 cells after laser treatment of EMT6 cells at room temperature. After incubation, cells were left at room temperature for 30 minutes prior to irradiation. Laser irradiation was applied continuously to each well up to the specific time indicated for each data point. (B) Temperature of cell media after irradiation, in parallel experiment to (A), the endpoint temperature of the cell media was measured with a type K thermocouple immersed in the cell media. (C) Cell viability of EMT6 cells after NIR-laser treatment at 37°C in the incubator. Another control group was studied incubating cells with PEGylated SWCNTs (untargeted) for 2 hours. (D) Cell viability of confluent HUVEC cells (no PS externalization) after NIR-laser treatment at 37°C in the incubator. Data are presented as mean ± S.D. Statistical analysis: unpaired t test with Welch’s correction for assumed unequal variances for A, B, and D (n = 2–3, *P < 0.05 and **P < 0.01). One-way ANOVA with Dunnett’s multiple comparisons comparing treated groups with the control group for C (n = 3, *P < 0.05, **P < 0.01, and ***P < 0.001 of SWCNT-ANXA5 group compared with the control group; •P < 0.05, ••P < 0.01, and •••P < 0.001 of SWCNT-ANXA5 compared with PEGylated SWCNT group). (E) SWCNT-ANXA5 conjugated to Alexa FluorTM 488 was observed to bind to EMT6 cells (A) but not to HUVEC cells (C). Transmission images for the same cells are shown in A’,B’ and D,D’. The confluency level for both cell lines grown on microscope slides was approximately 70%. Control groups without the addition of SWCNT-ANXA5 conjugated to Alexa Fluro488 are shown in B and B’ for the EMT6 cells and C’ and D’ for the HUVEC cells. Average fluorescent intensities and autofluorescence were measured for the EMT6 cells (F). Cells incubated with SWCNT-ANXA5 conjugated to Alexa Fluro488 had significantly higher signals than the autofluorescence signals from the EMT6 cells. ***P < 0.001. Scale bar: 50 μm.

The previous in vitro experiments were performed at room temperature (Fig. 1, A and B), while experiments in Fig. 1, C and D were performed at 37°C to better mimic physiologic conditions for cell ablation. Fig. 1C shows that EMT6 cells treated with SWCNT-ANXA5-mediated PTT at 37°C had a significant cell viability decrease after 1 minute of irradiation at 1 W/cm2. Comparatively, cells treated with laser alone (control) or PEGylated SWCNT-mediated PTT (lacking targeting capabilities) did not have a significant decrease in cell viability even after 7 minutes of laser irradiation. For all those experiments, cells were incubated for only 2 hours with either SWCNT-ANXA5 or PEGylated SWCNT, followed by four washing steps with PBS to remove any unbound SWCNT. In Fig. 1D, confluent HUVECs were treated with the same conditions as EMT6 cells in Fig. 1C. Confluent HUVEC cells mimic healthy endothelial cells, therefore lacking PS expression on the outer leaflet of the phospholipid bilayer. The control group of HUVEC cells (no SWCNT-ANXA5 incubation) did not present a decrease in cell viability from the laser irradiation, as expected when compared with the EMT6 cells that were only treated with laser irradiation. In the other group, confluent HUVECs were incubated with SWCNT-ANXA5 for 2 hours, followed by four washing steps, identically to previous experiments. Even after 7 minutes of treatment, corresponding to a power density of 420 J/cm2, SWCNT-ANXA5 did not have a significant cell viability decrease.

The labeling of the SWCNT-ANXA5 conjugate with Alexa FluorTM 488 was successful, resulting in a fluorescent signal when visualized with a fluorescence microscope. The stability of the dye allows for the detection of fluorescence for at least a week. For EMT6 cancer cells, the fluorescence increased significantly (Fig. 1F) when the cells were incubated with the SWCNT-ANXA5 conjugate (Fig. 1E,A) compared to the control with no SWCNT-ANXA5 conjugate was present (Fig. 1E,B). Transmission images for EMT6 cells with and without SWCNT-ANXA5 conjugate incubation are shown in Fig. 1E,A’ and Fig. 1E,B’, respectively. For the healthy HUVEC cells, no fluorescence from Alexa FluorTM 488 was observed, indicating no binding of SWCNT-ANXA5 (Fig. 1E,C); and no fluorescence was observed in the control with no SWCNT-ANXA5 conjugate present (Fig. 1E,C’). Transmission images for HUVEC cells with and without SWCNT-ANXA5 conjugate incubation are shown in Fig. 1E,C’ and Fig. 1E,D’, respectively.

Effect of Tumor Surface Temperature In Vivo on Immune System Activation.

In vivo studies were used to model immunity interactions. In our previous study of PTT using SWCNTs targeted by conjugation to ANXA5, the conjugate was injected intravenously, which resulted in the accumulation of SWCNTs in several organs, including the liver and kidneys, with slow clearance from these organs (McKernan et al., 2021). Therefore, in the current study, the SWCNT-ANXA5 conjugate was injected intratumorally at the same dosage used for intravenous injection in the previous study. In Fig. 2, EMT6-tumor-bearing mice were treated with SWCNT-ANXA5-mediated PTT, while the irradiation stopped immediately when the tumor surface temperature reached the temperature assigned to the groups. The levels of the pro-inflammatory cytokines tumor necrosis factor (TNF)-α and interleukin (IL)-12 were measured in serum either one day (24 hours) or seven days (168 hours) after the irradiation as evidence of immune activation by reaching various endpoint tumor surface temperatures. The group where PTT stopped at 45°C had an elevation of TNF-α at 24 hours (Fig. 2B) and at 168 hour (Fig. 2C) after PTT and IL-12 p70 at 24 hours (Fig. 2D). Comparatively, higher temperatures did not significantly increase the level of TNF-α at 168 hours and IL-12 at p70 at 24 hours compared with control (untreated). The mild temperature of 45°C was therefore used for the following experiments for a more efficient immune activation during SWCNT-ANXA5-mediated PTT.

Fig. 2.

SWCNT-ANXA5-mediated PTT-associated immune response as determined by the concentration of pro-inflammatory cytokines for different final tumor surface temperatures. Tumor-bearing mice were injected intratumorally with SWCNT-ANXA5 (1.2 mg/kg) and treated with PTT at 1.0 W/cm2. PTT stopped immediately when the tumor surface reached the final temperatures of the assigned groups (45, 50, 55, or 60°C). Blood samples were collected 1 day (24 hours) and 7 days (168 hours) after PTT, and the levels of TNF-α and IL-12 p70 in the serum were analyzed by ELISA compared with untreated control (injected intratumorally with SWCNT-ANXA5 but not treated with PTT). (A) Experimental setup for laser irradiation: Mice were anesthetized with 3% isoflurane via nose cone, while the irradiation generated by laser tip had area controlled by an opaque hollow cylinder. Thermal imaging was used to assess tumor surface temperature continuously throughout irradiation (figure made in BioRender.com). (B, C, D,) Levels of proinflammatory cytokines after irradiation: (B) TNF-α (24 h), (C) TNF-α (168 h), (D) IL-12 (24 h). Significant increase in serum cytokine levels is seen for 45°C, which is indication of immune activation, while it is hypothesized that higher temperatures might have induced very rapidly, possibly necrotic, cell death, inhibiting signaling pathways for immune response. Data are presented as mean ± S.D. (n = 4–5). Statistical significance was analyzed for the treated groups compared with the control group by one-way ANOVA with Dunnett’s multiple comparisons test. Statistical significance is indicated by *P < 0.05 and **P < 0.01.

Combination of SWCNT-ANXA5-Mediated PTT at 45°C and Anti-PD-1 Checkpoint Inhibition.

Fig. 3 shows a long-term survival study that was performed with the combination therapy of PTT and checkpoint inhibition with anti-PD-1. EMT6-tumor-bearing mice were injected intratumorally with SWCNT-ANXA5 when tumor sizes were ∼5 mm. The NIR irradiation was immediately stopped when the tumor surface temperature reached 45°C. Three doses of anti-PD-1 were injected intraperitoneally in the assigned groups. The combination therapy followed the schedule shown in Fig. 3A. After the treatment, mice in all treated groups did not present behavioral changes, being active during their nocturnal hours, which includes regular grooming, walking, running, and climbing in the cage, and normal food and water intake, which is an indication that the treatment was not toxic for the animals. Normal behavior was disrupted by the advancement of the disease stage, which was observed by dehydration, recumbent posture, breathing difficulty, or loss of leg function due to tumor proximity. The survival curve (Fig. 3B) shows that only the combination therapy was able to increase survival compared with the untreated control group (**P < 0.01). Parallelly, Fig. 3C shows the average tumor volume of the animals assigned to each group from the survival experiment in Fig. 3B. Sudden decreases in average tumor volumes are due to the death of mice inside of a group. Only the anti-PD-1 + PTT group mice had undergone complete tumor recession within a week after the PTT, encompassing 50% of animals in that group. All these animals survived for 100 days after tumor induction. Those mice did not undergo tumor regrowth and survived until the end of the study (100 days after tumor inoculation), yielding a 50% survival at 100 days (Fig. 3B). Disruption from normal behavior was not observed in the animals that had complete tumor recession in group anti-PD-1 + PTT during their entire lifespan.

Fig. 3.

Antitumor effect of the combination therapy of SWCNT-ANXA5 mediated PTT and checkpoint inhibition (anti-PD-1). (A) Treatment schedule for the combination therapy of EMT6 tumor model with SWCNT-ANXA5 mediated PTT and anti-PD-1. Mice with well-developed orthotopic syngeneic tumors (d ≥ 5 mm) were injected intratumorally with SWCNT-ANXA5 (1.2 mg/kg). Tumors were irradiated with a 980-nm laser at a power density of 1 W/cm2 until tumor surface t