Mesoporous silica nanoparticles boost aggressive cancer response to hydrophilic chlorin e6-mediated photodynamic therapy

Abstract

Background

Chlorin e6 trisodium salt (Ce6) is a newly developed hydrophilic photosensitizer designed to mediate anticancer photodynamic therapy (PDT). The response of different cancer types and strategies to boost anticancer efficiency of Ce6-PDT are poorly studied.

Objectives

This study aimed to investigate the response of different cancer types to Ce6-PDT, identify the unresponsive ones, and develop a nanosystem for response enhancement.

Methods

Sk-Br-3, MCF-7, U87, and HF-5 cells were tested in 2D cell cultures. Ce6 uptake, PDT-mediated phototoxicity, ROS production, caspase 3/7 levels, and cell death mode were examined. Furthermore, U87 spheroids were treated with Ce6-PDT. Mesoporous silica nanoparticles (MSN) were synthesized and loaded with Ce6. Cellular uptake and phototoxicity of MSN-Ce6 were compared to free Ce6 in vitro and in vivo.

Results

Ce6 was detectable in the cell cytoplasm within 15 min. U87 cells showed the highest Ce6 cellular uptake. Upon Ce6-PDT, U87 cells were the most responsive ones with an 11-fold increase in ROS production. Here, 5 µM Ce6 and 4 J/cm² were enough to reach IC50. Ce6-PDT induced both necrotic and caspase-dependent apoptotic cell death and 75% reduction of spheroids volume. Also, MCF-7 and HF-5 cells responded well to Ce6-PDT treatment. Sk-Br-3 breast cancer cells, on the other hand, were the least responsive ones with 80% viability after treatment (5 µM Ce6, 8 J/cm²). However, MSN-Ce6 conjugates increased Sk-Br-3 cellular uptake of Ce6 sevenfold decreasing the IC50 irradiation dose by an order of magnitude. In a very aggressive breast cancer rat model, MSN-Ce6-PDT treatment led to suppression of tumor volume by 50% and elevation of both Bax and caspase 3 by 90% compared to the control while the corresponding values for Ce6-PDT were 30% and 70%, respectively.

Conclusion

The newly developed hydrophilic chlorin and even more its MSN conjugate show high activities in anticancer PDT.

Introduction

Photodynamic therapy (PDT) is a local treatment that combines a photosensitizer (PS), cellular oxygen, and light. The illumination step triggers a sequence of photochemical reactions that generate deleterious reactive oxygen species (ROS). The latter may mediate apoptotic/necrotic/autophagic cell death, cessation of blood flow, and immunostimulation (Abdel Gaber 2017). Some PSs, such as 5- aminolevulinic acid (5-ALA), have gained approval for clinical applications (Peng et al. 1997). 5-ALA is a prodrug for the photosensitizer protoporphyrin IX (PPIX) and is approved for the treatment of actinic keratosis. Furthermore, it is intensively investigated for the treatment of different types of cancer such as glioblastoma (Müller et al. 2020), breast cancer (Guney Eskiler et al. 2020) and hepatocellular carcinoma (Ozten et al. 2022).

Chlorin e6 (Ce6) is a PS that is still in the pre-clinical and early clinical phases. It is one of the hydrophobic porphyrin derivatives with an absorption maximum of 665 nm granting deep tissue penetration. In its native form, Ce6 is rather hydrophobic and therefore has limited bioavailability. However, chemical modifications were applied to enhance its hydrophilicity. Conjugation with amino acid moieties, exemplified by aspartate lead to the formation of mono L-aspartyl Ce6, commercially known as Laserphyrin^®, which is approved in Japan for the treatment of glioma and esophageal cancer (Kobayashi et al. 2022; Yano et al. 2021). Furthermore, complexation with surfactants such as polyvinylpyrrolidone in a 1:1 (w/w) dilution was also conducted in order to diminish the aggregation and to enhance the solubility (Hädener et al. 2015; Isakau et al. 2008). A simpler approach was the formation of a trisodium salt of the pure Ce6, commercially known as Fotolon^®. It is readily soluble in aqueous medium, and has shown good PDT efficiency in the treatment of glioma (Abdel Gaber et al. 2018), onychomycosis (Yasin et al. 2022), keratitis (Wu et al. 2017) and Helicobacter pylori (Hüttenberger et al. 2017; Simon et al. 2014). However, the susceptibility of other cancer types to Ce6-PDT is not adequately studied.

In our previous study (Abdel Gaber et al. 2018), the relationship between serum protein concentration present at the in vitro level in fetal calf serum (FCS) and Ce6 cellular uptake was studied. The model was an engineered human glioblastoma multiforme cell line over-expressing the efflux pump ABCG2 upon induction to represent the stem-like sub-population of cancer cells. Ce6—in contrast to 5-ALA-induced PPIX—tended to bind to proteins after delivery. So, only after saturation of FCS-binding, efficient cellular uptake of Ce6 could take place. The role of ABCG2 was rather on the efflux side and modified PPIX more than Ce6. The relationship between FCS concentration and Ce6 cellular uptake using standard cancer cell lines was not investigated. Studying this relationship could optimize Ce6 protocol in terms of dosing regimen and administration route for diagnostic and therapeutic purposes.

Conjugating the PS with nanoparticles (NPs) is a common strategy to enhance the PDT therapeutic response (Mfouo-Tynga et al. 2021; Abdel Gaber and Fadel 2023). Silica-based NPs, including Stöber and mesoporous silica nanoparticles (MSN), are known for their facile synthesis, straightforward surface modification, and biocompatibility (Mamaeva et al. 2013), and have therefore attracted interest as PS carriers (Couleaud et al. 2010). A previous attempt with pure Ce6 reported a considerably low loading degree of 23 µg Ce6/mg silica NPs when Ce6 was covalently bound to amino-functionalized Stöber NPs. Although the conjugate led to an improvement in PDT efficiency in vitro when compared to free Ce6, irradiation intensities as high as 60 J/cm² were needed to reach IC50 for breast cancer MDA-MB-231 cells (Bharathiraja et al. 2017). In another in vitro study, Ce6 was covalently attached to amino-functionalized silica NPs (Youssef et al. 2018). The reported Ce6 loading degree was only 6.5 µg/mg. Notably, already the bare silica NPs killed 30% of the glioblastoma cells, indicating that these particles were cytotoxic.

This study aimed to assess the PDT efficiency of the water-soluble trisodium salt of Ce6 in treating breast and glioma cancers compared to 5-ALA to better position it with respect to the FDA approved PS. Cancer types were ordered in terms of responsiveness and a third-generation PS developed by covalently conjugating Ce6 to MSN was tested for its ability to improve the therapeutic response in the least responsive cancer model. Bioassessments were conducted using human cell lines, 3D cellular spheroids, and an established cancer animal model.

Materials and methods

Materials

5-Aminolevulinic hydrochloride acid (5-ALA) was purchased from Medac GmbH, Hamburg, Germany. Chlorin e6 trisodium salt (Ce6) was a kind gift from Synverdis GmbH, Heidelberg, Germany. Basic fibroblast growth factor, Cell Titer Blue (CTB), and epidermal growth factor were purchased from Promega, Mannheim, Germany. 1-Ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC), N-hydroxysuccinimide (NHS), cetyltrimethylammonium bromide (CTAB), trimethylbenzene (TMB), tetraethylorthosilicate (TEOS), dimethyl sulfoxide (DMSO) and 3-aminopropyltriethoxysilane (APTES) were purchased from Sigma Aldrich, Germany, and used as received. Dulbecco’s modified Eagle’s medium (DMEM), DMEM F-12, McCoy’s medium, fetal calf serum (FCS), phosphate buffer saline (PBS), and penicillin/streptomycin mixture were purchased from Biochrom AG, Berlin, Germany.

Synthesis of mesoporous silica nanoparticles (MSN)

Amino-functionalized MSN were synthesized following our previously published protocol using cetyltrimethylammonium bromide (CTAB) as the porogen (Rosenholm et al. 2009) with the exception that 10 mol-% of the silica source tetraethylorthosilicate (TEOS) was substituted by 3-aminopropyltriethoxysilane (APTES). Furthermore, trimethylbenzene (TMB) was added as a pore swelling agent (n_TMB/n_CTAB = 2.5). After the synthesis, the surfactant was removed by extraction in acidic ethanol, and the particles were dried under vacuum at 60 °C.

Synthesis of MSN-Ce6 conjugate

Ce6 was activated by 24-h reaction with 1.5 molar equivalent of 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC), N-hydroxysuccinimide (NHS), in the presence of 1.5 mL anhydrous dimethyl sulfoxide (DMSO). MSN in phosphate buffer saline (PBS) was mixed with Ce6 at 1:5 wt/wt for 24 h. For conjugate collection, the mixture was centrifuged for 20 min at 7,000 g, and NPs were washed 3 times in PBS. Supernatants were collected and the concentration of Ce6 was quantified spectrophotometrically at 405 nm (Bharathiraja et al. 2017). The entrapment efficiency was calculated using Eq. (1) and the loading efficiency was calculated using Eq. (2).

$${\text{Entrapment}}\,{\text{ efficiency }}\left( \% \right) \, \, = \, \,({\text{Ce6}}_{{{\text{total}}}} - {\text{Ce6}}_{{{\text{unbound}}}}) /{\text{Ce6}}_{{{\text{total}}}} \times \,{1}00$$

(1)

$${\text{Loading}}\,{\text{ efficiency }}\left( \% \right) \, \, = \, \,({\text{Ce6}}_{{{\text{total}}}} - {\text{Ce6}}_{{{\text{unbound}}}}) /{\text{weight of NPs}} \times \,{1}00$$

(2)

Characterization of the NPs and synthesized conjugate

The morphology of the synthesized NPs was studied by transmission electron microscopy (Jeol 1200 (Jeol, Germany) using HT voltage of 120 kV and beam current of 65 µA), and the pore size was determined by nitrogen sorption at − 196 °C (Quantachrome Autosorb-1) and calculated using the NLDFT kernel developed for cylindrical pores, zeta-potential measurements were performed using a Zetasizer NanoZS Zen3600 setup (Malvern Panalytical, Germany) in aqueous KCl solutions (1 mM, particle concentration: 0.1 mg/mL). Fourier transmission infrared (FTIR) spectroscopy was conducted (Model no. 4000, JASCO, Japan) on Ce6, NPs, and conjugate with a range of 400–4000 cm⁻¹ using KBr pellets.

Ce6 release kinetics from MSN

MSN-Ce6 (1 mg of MSN-Ce6 conjugate/mL, at room temperature) was sonicated in PBS at a pH of 7.4 or pH of 5.5 for 10 min. Mixtures were left stirring at room temperature at 600 rpm. At different time intervals over 24 h, aliquots of the mixture were withdrawn and replaced with fresh PBS of the same pH to maintain the volume. The concentration of Ce6 was measured spectrophotometrically at 405 nm and data were presented as cumulative Ce6 release % against the different time points (Couleaud et al. 2010).

Cell culture

Human breast cancer cell line (MCF-7), human glioblastoma cell line (U87), and human normal fibroblasts (HF-5) were cultured in Dulbecco’s modified Eagle’s medium (DMEM) medium. The human breast cancer cell line (Sk-Br-3) was cultured in McCoy's medium. All media were supplied with 10% fetal calf serum (FCS), 100 μg/mL non-essential amino acids, and 1 mM sodium pyruvate. The antibiotic mixture of 100 U/mL penicillin and 100 μg/mL streptomycin was added to assure having an aseptic medium. Cells were cultured at 37 °C in a humidified atmosphere supplied with 5% CO₂.

Quantification of cellular uptake by flow cytometry

A FACS Calibur flow cytometer (BD Biosciences) was used, and data were collected for 10,000 cells and analyzed using Flowing Software 2.5.1 (Turku Center for Biotechnology, Turku University, Finland). The Geometrical means (Geomean) of the respective histograms were automatically calculated after gating the desired population to exclude dead cells. PPIX was excited using 488 nm and emission was detected at 620 nm while Ce6 was excited at 633 nm and emission was detected at 660 nm (Abdel Gaber et al. 2018). To study the effect of FCS concentration, Sk-Br-3, MCF-7, U87, and HF-5 cells were incubated for 4 h with either Ce6 concentration series (0.1–100 µM) or 5-ALA concentration series (0.075–4.8 mM). The incubation medium was either FCS-free or supplied with 10% FCS or 2% FCS. To study the effect of the incubation period, U87 cells were incubated with either Ce6 series (1–100 µM) for 15 min to 24 h or with 5-ALA concentration series (0.075 to 4.8 mM) for 1 or 8 h. To study the effect of Ce6 conjugate on Ce6 cellular uptake, Sk-Br-3, and HF-5 cells were incubated for 4 h in a medium supplied with 2% FCS with 5 µM Ce6 of either free Ce6 or MSN-Ce6.

Cellular localization assessment

Sk-Br-3 cells were incubated with either 10 µM Ce6 or 1.2 mM 5-ALA. U87 cells and spheroids were incubated with 5 µM Ce6 in the form MSN-Ce6 and free Ce6, respectively. All incubations were for 4 h in a medium supplied with 2% FCS. Nuclei were stained with Hoechst 33258 (Invitrogen, USA) as instructed by the manufacturer. Fluorescence was examined by Leica DM5000B fluorescent microscope (Leica, Germany).

Spheroid formation

Spheroids are 3-dimensional cellular structures characterized by a decreased gradient of oxygen and nutrient from the periphery to the center, similar to what is expected in an in vivo tumor mass (Pinto et al. 2020). Sk-Br-3, MCF-7, and U87 cells were cultured in FCS-free DMEM F-12 medium supplied with B27, basic fibroblast growth factor, and epidermal growth factor. Cells were cultured in ultra-low attachment T-flasks and left for 7 days in the incubator with minimal shaking (Abdel Gaber et al. 2018). To evaluate the growth kinetics, on the third day of culture, 3 spheres were transferred from the T-flasks to ultra-low attachment 96-well plates and they were imaged every 2 days for 2 weeks using the bright field of Leica DM5000B fluorescent microscope (Leica, Germany) at 10x magnification. The projected surface area of the formed spheres was estimated using Image J software downloaded from (www.NIH.org). Results were expressed as a fold increase in the projected area to the projected area of the respective spheres on the third day of culture.

Assessment of PDT efficiency

A group of dark control was included in which incubation with the highest tested PS concentration took place without irradiation. A group of laser control was included in which irradiation took place at the highest tested irradiation dose without prior PS incubation. A control group was included in which cells were neither incubated with PS nor irradiated.

For Ce6-PDT, Sk-Br-3, MCF-7, U87, and HF-5 cells were incubated for 4 h with either 5 or 10 µM Ce6 supplied in a 2% FCS-containing medium. For 5-ALA-PDT, Sk-Br-3, MCF-7, U87, and HF-5 cells were incubated with either 0.15 or 0.6 mM 5-ALA. For illumination, the PS-containing medium was removed and replaced with a phenol-red-free medium. Ce6 irradiation was conducted using a 665-nm diode laser (100 mW/cm²) and 5-ALA irradiation was conducted using a 635 nm diode laser (100 mW/cm²). The irradiation time varied to have an irradiation dose from 0.5 to 8 J/cm². Cells were cultured for another 4 h, and viability was measured using Cell titer blue assay (CTB) as previously described (Kammerer et al. 2011). Results were expressed as viability % relative to the control. Conditions needed to kill 50% of the cell population (IC50) were estimated with a 95% confidence interval (El-Gogary et al. 2019).

For spheroid Ce6-PDT, U87 spheroids were incubated with 10 µM Ce6 for 4 h. Ce6 was removed, and spheroids were irradiated with 2.4 J/cm². A fresh medium was placed, and spheroids were imaged to monitor their projected surface area for 5 days after treatment. Results were expressed as the projected surface area of spheroids over time.

For Ce6 conjugate-PDT, Sk-Br-3, and HF-5 cells were incubated for 4 h in a medium supplied with 2% FCS with 5 µM Ce6 of either free Ce6 or MSN-Ce6. The Ce6-containing medium was removed and cells were irradiated with an irradiation dose of up to 8 J/cm². Cells were cultured for a further 4 h and viability was measured by CTB assay. Results were expressed as viability % relative to the controls.

Measurement of reactive oxygen species (ROS) production

ROS was measured using the 2′, 7′–dichlorodihydrofluorescein diacetate (DCFDA) kit (Abcam, Germany). Sk-Br-3, MCF-7, U87, and HF-5 cells were treated with IC50 conditions of 10 µM free Ce6-PDT. Before illumination by 45 min, cells were incubated with 25 µM of the supplied working solution. Directly after irradiation, the fluorescence was measured (Ex: 495 nm and Em: 522 nm) using a microplate reader (BMG LABTECH, Germany). Results were expressed as fold differences relative to the control (El-Gogary et al. 2019).

Assessment of cell death mode and caspase 3/7 level

Sk-Br-3, HF-5, and U87 cells were treated with IC50 conditions of 10 µM free Ce6-PDT or 0.6 mM 5-ALA-PDT. Cells were collected and stained with Annexin V FITC and propidium iodide following the manufacturer´s instructions (BD Pharmingen FITC Annexin V apoptosis detection kit) for 15 min at room temperature. Fluorescence was quantified using flow cytometry. Annexin V (green color; Ex = 488 nm, Em = 500–550 nm) and propidium iodide (red color; Ex = 514 nm, Em = 650–730 nm) (El-Gogary et al. 2019). For the caspase 3/7 quantification assay, Sk-Br-3, MCF-7, U87, and HF-5 cells were treated with IC50 conditions of 10 µM free Ce6-PDT or 0.6 mM 5-ALA-PDT. At the end of the 4-h post-irradiation period, cells were incubated for 1 h at room temperature with the caspase 3/7 substrate provided in the kit (Apo-ONE Homogenous Caspase 3/7 Assay, Promega). The fluorescence of the produced rhodamine 110 was measured (Excitation at 495 nm and emission at 522 nm). Caspase 3/7 activity was expressed as fold elevation compared to the control untreated group (El-Gogary et al. 2019).

Breast cancer rat model

The study was approved by the institutional animal care and use committee of Kafrelsheikh University. Adult female Wister rats (weight 100–150 g) were purchased from the Medical Experimental Research Center of Tanta University. Rats were acclimatized for 15 days with free access to food and water. The light cycle of 12 h darkness was maintained. For breast cancer induction, rats were administered 7,12-dimethylbenz(a)anthracene (DMBA) dissolved in sesame oil as a single oral dose (20 mg/kg). Treatment was conducted 14 weeks after DMBA administration (Barros et al. 2004).

Treatment groups

Rats were grouped into five groups (n = 4) as listed in Table 1. The biosafety and biodegradation of MSN were reported before (Mamaeva et al. 2011) and thus a group of “MSN only” was not included to minimize the number of used rats. Free Ce6 and MSN-Ce6 were intratumorally injected. For the PDT groups, irradiation with 180 J/cm² from a non-thermal LED source (660 nm, 100 mW/cm², 30 min) was conducted 24 h after Ce6 administration either in the free or the conjugate form. Rats were weighed and euthanized by cervical dislocation 2 weeks after applying PDT. Tissues were collected, and their weight and volume were recorded (Jeong et al. 2011). Volume was calculated using Eq. (3):

$${\text{Tumor }}\,{\text{volume }}\, = \, \,{\text{width}}^{{2}} \times {\text{length }} \times \, 0.{5}{\text{}}$$

(3)

Table 1 Treatment groups to assess Ce6-PDT either in the free form or conjugated as MSN-Ce6 in treating breast cancer

Histopathological assessment

Tissues were thoroughly washed with saline. They were fixed in paraformaldehyde and gradual dehydration protocol using elevating ethanol concentration (70–100%) was applied. They were embedded in paraffin and sliced into a 5-µm-thick section mounted on glass slides. Sections were stained using hematoxylin and eosin (Hematoxylin and Eosin Stain Kit, Vector Laboratories, CA, USA). Slides were examined using 40 × magnification of a light microscope where signs of inflammation, necrosis, and fibrosis were investigated (Abdel Khalek et al. 2021).

Immunohistochemical assessment

Deparaffinated sections were incubated overnight at 4 °C with 1:50 dilutions of rabbit-anti-cleaved caspase 3 or Bax antibodies (Cell Signaling Technology, USA). Sections were then incubated with biotinylated goat-anti-rabbit-IgG and streptavidin/alkaline phosphatase complex at 1:200 dilution. Slides were counterstained with hematoxylin and examined by 40 × objective. The expression level of caspase 3 and Bax was reported as average optical density collected from random 200 points (Cabral et al. 2021).

Statistical analysis

At least, each experiment was repeated 3 times, and presented are the average and standard deviation. Statistical analysis was applied using GraphPad Instat software using ANOVA and Tukey post-test (P < 0.05). P value was depicted as * when it was less than 0.05, as ** when it was less than 0.01, and as *** when it was less than 0.001.

Results

Ce6 cellular uptake, and intracellular localization

Cellular uptake of Ce6 was studied in relation to serum concentration (Fig. 1a) and compared to 5-ALA-PPIX (Fig. 1b). Results depicted are for Sk-Br-3 cells incubated for 4 h with a series of concentrations of either PS. The lack of serum was associated with a significantly higher cellular fluorescence of both Ce6 and PPIX at all tested concentrations when compared to the respective values when the medium was supplied with 10% serum. The findings for all cell lines are summarized in (Additional File 1:  Table 1) . As seen, the highest fold difference in measured PPIX fluorescence was 10.4 fold and it was detected when MCF-7 cells were incubated with 0.3 mM 5-ALA. In the case of Ce6, the highest fold difference was 34.6 times, and it was measured when Sk-Br-3 cells were incubated with 5 µM Ce6.

Fig. 1

a Cellular uptake of Sk-Br-3 cells incubated for 4 h with a series of Ce6 concentrations in the absence of fetal calf serum (FCS) and in the presence of 10% FCS. b Cellular uptake of Sk-Br-3 cells incubated for 4 h with a series of 5-ALA concentrations in the absence of FCS and in the presence of 10% FCS. c Cellular uptake of U87 cells incubated with a series of Ce6 concentrations for 1–24 h in the presence of 2% FCS. d Cellular uptake of U87 cells incubated with a series of 5-ALA concentrations for 1 and 8 h in the presence of 2% FCS. e Ce6 cellular accumulation in Sk-Br3, MCF-7, U87 and HF-5 cells after 4 h incubation with either 5 or 10 µM Ce6 in the presence of 2% FCS. f 5-ALA-PPIX cellular accumulation in Sk-Br3, MCF-7, U87 and HF-5 cells after 4 h incubation with either 0.3 or 1.2 mM 5-ALA in the presence of 2% FCS. g Representative fluorescence image of Sk-Br-3 incubated for 4 h with 10 µM Ce6 supplied in 2% FCS. h Representative fluorescence image of Sk-Br-3 cells incubated for 4 h with 1.2 mM 5-ALA supplied in 2% FCS

The kinetics of Ce6 cellular uptake was measured over 24 h using U87 cells while fixing the serum concentration to 2% (Fig. 1c). A concentration-dependent increase in the cellular uptake was observed at all tested incubation periods. Ce6 was detected in the cells after 15 min (Additional file 1: Fig. S1). For a given concentration, PPIX cellular fluorescence increased with the increase of the 5-ALA-incubation time but reached a plateau at both tested incubation periods (Fig. 1d).

The cellular accumulation was studied with respect to the cell type as well. Cellular uptake of 5 and 10 µM Ce6 after 4 h incubation supplied in 2% FCS with all tested 4 cell lines (Fig. 1e) showed that U87 had the highest accumulation level and Sk-Br-3 cells the lowest. It is worth noting that the non-cancerous cell line HF-5 used as a model for healthy cells showed a high level of Ce6 uptake that was not statistically different from the uptake by U87 and MCF-7 cells when the concentration was 5 µM and was indifferent from MCF-7 when the concentration was 10 µM.

The cellular accumulation of PPIX after 4 h incubation with 5-ALA in a 2% FCS-containing medium (Fig. 1f) revealed that at 0.3 mM 5-ALA, MCF-7 showed the lowest PPIX cellular accumulation of the tested cell lines. At 1.2 mM 5-ALA, both MCF-7 and Sk-Br-3 cells showed a statistically insignificant difference in PPIX accumulation, and the same was observed for U87 and HF-5 cells. As shown, the PPIX accumulation levels in the tested breast cancer cells were 2 times higher than the respective values for U87 and HF-5 cells. Importantly, the increase in 5-ALA concentration was not associated with an increase in the PPIX accumulation for U87 and HF-5.

The cellular localization of Ce6 was investigated by fluorescence microscopy after incubating Sk-Br-3 cells for 4 h with 10 µM Ce6 supplied in 2% FCS. As shown in Fig. 1g, this water-soluble form of Ce6 succeeded to permeate the cell membrane and was homogeneously distributed within the cytoplasm of the cells. This finding illustrated that the quantified fluorescence signal detected by flow cytometry was mainly originating from intracellular Ce6. The same cell line was incubated with 1.2 mM 5-ALA for 4 h in 2% FCS-containing medium, and as seen in Fig. 1h, PPIX showed a distinct tendency to accumulate at the cell membrane.

In vitro Ce6-PDT

Based on the previous findings, an incubation period of 4 h in a medium supplied with 2% FCS was selected for the in vitro PDT evaluations. Ce6-PDT was conducted on all tested cell lines using either 5 or 10 µM Ce6, and irradiation doses up to 8 J/cm². The viability results of Sk-Br-3, MCF-7, U87, and HF-5 cells are presented in Fig. 2 a, c, e, and g, respectively. Neither the dark control in the presence of 10 µM Ce6 nor the irradiation of the cells at 8 J/cm² without Ce6 exerted any cytotoxicity on any of the tested cell lines.

Fig. 2

a Sk-Br-3, c MCF-7, e U87, g HF-5 cells viability results after 4 h incubation with either 5 or 10 µM Ce6 followed by irradiation with 665 nm for various durations. In the dark control group, cells were incubated with 10 µM Ce6 for 4 h without irradiation, and in the light control group, cells were irradiated with 8 J/cm² without prior Ce6 incubation: b Sk-Br-3, d MCF-7, f U87, h HF-5 cells viability results after 4 h incubation with either 0.15 or 0.6 mM 5-ALA followed by irradiation using 635 nm. In the dark control group, cells were incubated with 0.6 mM 5-ALA for 4 h without irradiation and in the light control group, cells were irradiated with 8 J/cm² without prior 5-ALA incubation. Experiments were repeated at least 3 times and presented are mean and standard deviation

There was a significant decline in the cell viability for a given Ce6 concentration with increasing irradiation dose. Moreover, there was a statistically significant, concentration-dependent decrease in viability at a given irradiation dose (Fig. 2 a, c, e, g). The cells varied in their responsiveness to Ce6-PDT, as summarized in Table 2. At 5 µM Ce6, HF-5, and Sk-Br-3 were the least responsive cell lines and did not reach 50% cell death (IC50) within the tested irradiation dose range. At a Ce6 concentration of 10 µM, there was a 4 -times decrease in the irradiation dose needed to reach IC50 for both MCF-7 and U87 cells. The breast cancer cell line Sk-Br-3 required a 3.6 times higher irradiation dose than the other three cell lines (HF-5, U87, and MCF-7) to reach IC50.

Table 2 Summary of conditions needed to reach IC50 of Sk-Br-3, MCF-7, U87, and HF-5 after Ce6-PDT and 5-ALA-PDT

For comparison, the 4 cell lines were treated with 0.15 and 0.6 mM 5-ALA-PDT at 4 h incubation in a medium supplied with 2% FCS and with an irradiation dose of up to 8 J/cm². The results are shown in Fig. 2b, d, f, and h, respectively. As for Ce6, the control experiments showed no cytotoxicity. As summarized in Table 2, at 0.15 mM 5-ALA, none of the cell lines reached IC50 except U87 cells, which reached IC50 at an irradiation dose of 7 J/cm². When the concentration was 0.6 mM 5-ALA, the most responsive cell line was Sk-Br-3 as it required only 3 J/cm² to reach IC50, while both U87 and MCF-7 required 4 J/cm². The least responsive cell line was HF-5, in line with the known selectivity of 5-ALA towards cancer cells.

In vitro, Ce6-PDT triggered ROS production and mode of cell death

ROS produced in the different cell lines were measured directly after PDT (10 µM Ce6-PDT, varying irradiation dose) and the results are presented in Fig. 3a. A significant increase in the ROS levels as compared to the control was observed in all cell lines. At all irradiation doses, U87 cells were the highest ROS-producing cells, while Sk-Br-3 and HF-5 cells were the lowest ROS-producing ones.

Fig. 3

a ROS production level of Sk-Br3, MCF-7, U87, and HF-5 after 4 h incubation with 10 µM Ce6 followed by irradiation using 665 nm for various durations expressed as fold difference relative to the control. b Apoptotic and necrotic % of Sk-Br-3, U87, and HF-5 cells after treatment with 10 µM Ce6-PDT conditions corresponding to IC50 conditions. c Apoptotic and necrotic % of Sk-Br-3, U87, and HF-5 cells after treatment with 0.6 mM 5-ALA-PDT conditions corresponding to IC50 conditions. d Caspase 3/7 level of Sk-Br-3, MCF-7, and U87 cells after treatment with 10 µM Ce6-PDT conditions corresponding to IC50 conditions. e Caspase 3/7 level of Sk-Br-3, MCF-7, and U87 cells after treatment with 0.6 mM 5-ALA-PDT conditions corresponding to IC50 conditions. Treatment conditions are summarized in Table 3. All experiments were repeated at least 3 times and presented are mean and standard deviation

Irradiation doses corresponding to IC50 at 10 µM Ce6-PDT were applied to the different cell lines, and the mode of cell death was investigated (Fig. 3b). Cells exhibited both apoptotic and necrotic cell death modes. On the contrary, corresponding measurements using 0.6 mM 5-ALA-PDT showed a predominant apoptotic cell death mode (Fig. 3c). The observed apoptotic death mode was caspase-dependent for both Ce6 (Fig. 3d) and 5-ALA-PDT (Fig. 3e), as evidenced by the significant elevation in the caspase 3/7 level exemplifying the executioner caspases in all tested cell lines compared to the control. SK-Br-3 showed the lowest level of caspase 3/7 following Ce6-PDT which matched having the lowest percentage of apoptotic cells observed for this cell line.

Glioma spheroids Ce6-PDT

Sk-Br-3 and MCF-7 cells formed irregular spheroids, and tended to form bulky spheroid clusters, while U87 cells produced regularly spherical shaped ones (Additional file 1: Fig. S2). Thus, U87 spheroids were selected for further studies. Incubating U87 spheroids (Fig. 4a) with Ce6 resulted in an evident fluorescence, confirming Ce6 cellular uptake (Fig. 4b). Upon illumination, the periphery of the spheroids became loose, and cell dissociation progressed with time (Fig. 4c, d). Incubation of the spheroids without illumination (dark control group) resulted in a 5 times increase in the spheroid projected surface area on the fifth day of treatment. One session of Ce6-PDT suppressed the growth, resulted in a significant decline in the projected area by 60% (Fig. 4e) and the corresponding volume by 75% at day 5 after PDT compared to dark control spheroids.

Fig. 4

a Representative image of U87 spheroid using a light microscope. b Representative image of U87 spheroid using fluorescence microscope after incubation for 4 h with 10 µM Ce6 and staining the nuclei with Hoechst 33258. c Representative light microscope image of U87 spheroid treated with Ce6-PDT on day 0. d Representative light microscope image of U87 spheroid treated with Ce6-PDT on day 3. e Fold increase of U87 spheroid projected area monitored over 5 days after incubation for 4 h with 10 µM Ce6 followed by 2.4 J/cm² irradiation using 665 nm compared to the dark control group in which spheroids were only incubated with 10 µM Ce6 without irradiation. Spheroids were imaged using 10 × magnification. Experiments were repeated at least 3 times and presented are mean and standard deviation

MSN-Ce6-conjugate characterization and release

As the PDT efficiency of Ce6 was evidently directly correlated with the extent of cellular uptake, we hypothesized formulating Ce6 together with cationic MSN would lead to higher apoptosis levels for Sk-Br-3 cells. We, therefore, synthesized highly amino silane-functionalized MSN through co-condensation, resulting in particles having an amino silane content of 16.5 wt%, a particle diameter of 313 ± 21 nm (Fig. 5a), an average pore size of 5 nm, and a specific surface area of 915 m²/g. The zeta-potential value measured in 1 mM KCl was + 28 m, in agreement with the high amino silane content. Ce6 was then covalently linked to the particles using EDC/NHS coupling, and the coupling efficiency was 99.9% resulting in a Ce6 content of 20% wt/wt.

Fig. 5

a Transmission electron microscope image of amine-functionalized mesoporous silica nanoparticles. b Ce6 release profile of MSN-Ce6 conjugate sonicated in phosphate buffer saline at room temperature in pH: 5.5 or pH: 7.5 over 24 h

FT-IR spectroscopy (Additional file 1: Fig. S3) confirmed a successful covalent Ce6 conjugation to the MSN, as a band appeared at 1636 cm⁻¹ after conjugation, which could be assigned to an amide carbonyl bond (Aboelmaati et al. 2021). Ce6 release from MSN-Ce6 conjugate was quantified spectrophotometrically over 24 h either in pH of 7.4 or pH 5.5 buffers, to mimic the pH-conditions in the bloodstream and the tumor stroma acidic environment (Fig. 5b). At both pHs, 40% of Ce6 was released within the first 6 h, while the corresponding values measured after a release time of 24 h was 97% at pH = 5.5 and 65% at pH = 7.4.

MSN-Ce6 cellular uptake, localization, and PDT

The cellular uptake of MSN-Ce6 by Sk-Br-3 cells was determined by FACS. The use of the MSN-Ce6 conjugate significantly increased Ce6 cellular uptake as compared to free Ce6, as shown in Fig. 6a. For example, at a Ce6 concentration of 5 μM, internalized Ce6 was sevenfold to that of free Ce6. At higher concentrations, the uptake of Ce6 was also clearly enhanced when conjugated to MSN, but the relative difference in uptake when compared to free Ce6 decreased with increasing Ce6 concentration and was about threefold at a Ce6 concentration of 10 μM. The increased Ce6 uptake was also reflected in the PDT efficiency, as shown in Fig. 6b. The viability of the Sk-Br-3 cells irradiated once for 40–80 s with an intensity in the range of 4–8 J/cm² decreased from about 75% in case of free Ce6 to less than 20% in the case of MSN-Ce6-PDT. Ce6 administered using the MSN-Ce6 conjugate was mainly localized in the cytoplasm of U87 cells that had been incubated for 4 h with 5 µM Ce6 concentration (Fig. 6c). The distribution closely resembled that of free Ce6 (Fig. 1g).

Fig. 6

a Cellular uptake of Sk-Br-3 after 4 h incubation with either 5 µM free Ce6 or the same concentration of Ce6 but in the MSN-Ce6 conjugate form. b Viability results of Sk-Br-3 cells incubated for 4 h with 5 µM free Ce6 or the same concentration of Ce6, but in the MSN-Ce6 conjugate form followed by irradiation using 665 nm for various durations. c Representative fluorescence images of U87 control cells and cells incubated for 4 h with 5 µM Ce6 in the MSN-Ce6 conjugate form showing the nuclei stained with Hoechst 33258, Ce6 fluorescence and the merge of both. d Cellular uptake of HF-5 after 4 h incubation with either 5 µM free Ce6 or same concentration of Ce6 but in the MSN-Ce6 conjugate form. e Viability results of HF-5 cells incubated for 4 h with 5 µM free Ce6 or the same concentration of Ce6, but in the MSN-Ce6 conjugate form followed by irradiation using 665 nm for various durations. Experiments were repeated at least 3 times and presented are mean and standard deviation

Similar findings were seen in HF-5 cells (Fig. 6d). We note that the high PDT efficiency was also seen for the non-cancer cell line HF-5 (Fig. 6e).

In vivo PDT response

Inspired by the promising in vitro results obtained by the Ce6-MSN conjugate, we pursued an in vivo study focusing on a very aggressive type of breast cancer chemically induced by the administration of the immunosuppressor DMBA to rats. Based on the low in vitro selectivity of the Ce6-MSN particles, intratumoral administration was used, and the results were compared with corresponding results obtained for free Ce6. Either free Ce6 or MSN-Ce6 was injected, followed by 30 min irradiation using a non-thermal LED (100 mW/cm²) 24 h after Ce6 administration. Images of the tumors collected at the end of the experiments, i.e., 2 weeks after a single PDT application, showed a visually evident treatment success in the case of both MSN-Ce6 and free Ce6 groups compared to the dark control or the untreated control groups (Fig. 7a). The reduction in the tumor volume compared to the untreated control was 30% in the case of free Ce6-PDT and 50% in the case of Ce6-MSN-PDT (Fig. 7b). A significant decline in the collected tumor mass was also observed; 1.6 ± 0.13 g in the case of the untreated control, 1.12 ± 0.12 g in the case of free Ce6-PDT (30% reduction), and 0.67 ± 0.09 g in the case of MSN-Ce6-PDT (60% reduction) (Fig. 7c). Those findings showed that the observed enhancement of PDT response in vitro (Fig. 6b) was valid also in vivo.

Fig. 7

a Representative photo images of collected tumor tissues of DMBA-induced breast cancer rats from the untreated group, dark control group ( administered 5 mg/kg Ce6 with no irradiation), free Ce6-PDT group (administered 5 mg/kg of free Ce6 followed by 180 J/cm² irradiation using 665 nm) and MSN-Ce6-PDT group (administered dose equivalent to 5 mg/kg of Ce6 followed by 180 J/cm² irradiation using 665 nm) bar 9 mm. b Average tumor volume. c Average tumor weight. d Histopathological images of hematoxylin and eosin stained non-lactating mammary gland tissues collected from the aforementioned treatment groups. Blue arrows marked the lumen of the duct filled with atypical proliferative cells and red arrows marked the malignant epithelial cells invading the stroma forming small ductal structures (upper row X:100 bar 100 µm, lower row X: 400 bar 50 µm)

Histological examination of collected tissues (Fig. 7d) revealed that a very invasive ductal carcinoma in situ (DCI) was developed in the untreated control, where the epithelial cells filled and expanded the ducts forming glandular spaces. Necrotic and dysplastic ducts were surrounded by necrotic and inflamed stroma mainly due to neutrophil infiltration that caused extensive comedo necrosis (blue arrows). A similar histological picture was seen in the case of the dark control group where malignant epithelial cells invading the stroma forming small ductal structures (red arrows), and large areas of necrosis with the infiltration of some inflammatory cells were seen. The severity of the DCI decreased in the free Ce6-PDT group and was the mildest in the case of the MSN-Ce6-PDT group.

In vivo MSN-Ce6-PDT relation with caspase 3 and Bax levels

The expression of caspase 3 (Fig. 8a) and Bax (Fig. 8b) in the tissues of dark control, free Ce6 (5 mg/kg) and MSN-Ce6 (5 mg/kg) were studied by immunohistochemical assays. As demonstrated in the optical density percentage of caspase 3 (Additional file 1: Fig. S4) and Bax staining (Additional file 1: Fig. S5), the expression of both caspase 3 and Bax was significantly elevated in the treated groups compared with the dark control and were significantly higher in the case of MSN-Ce6-PDT compared to the group of Ce6-PDT.

Fig. 8

a Microscopic images of immunostained breast sections against caspase 3 counterstained with Mayer's hematoxylin showing positively stained infiltrating tumor cells (black arrows) around affected lactating duct. b Microscopic images of immunostained breast sections against Bax counterstained with Mayer's hematoxylin showing positively stained infiltrating tumor cells (black arrows) around the affected lactating duct. The dark control group was DMBA-induced breast cancer rats administered 5 mg/kg Ce6 with no irradiation, the free Ce6-PDT group was DMBA-induced breast cancer rats administered 5 mg/kg of free Ce6 followed by 180 J/cm² irradiation using 665 nm and MSN-Ce6-PDT group was DMBA-induced breast cancer rats administered dose equivalent to 5 mg/kg of Ce6 followed by 180 J/cm² irradiation using 665 nm (X: 400 bar 50 µm)

Discussion

The anticancer PDT efficiency of a water-soluble form of Ce6 was investigated. Its cellular uptake was greatly reduced in the presence of serum proteins as compared to the benchmark used in this study, 5-ALA-PPIX (Table 1 in supplementary material). This could indicate that trisodium salt of Ce6 is adsorbed to serum proteins, which in turn can reduce cellular uptake. From a translational point of view, such adsorption could be attractive (Jeong et al. 2011), but the formation of Ce6–protein complexes remains speculative at this stage. Those findings were in agreement with a previous study using Sn(IV) chlorine e6 dichloride trisodium salt, (Al-Khaza’leh et al. 2011). The major protein components of serum are albumin and low-density lipoproteins (LDL). It was reported that Ce6 (acid form) binds to both types of proteins, with an association constant to albumin of 1.8 ± 0.2 × 10⁸ M⁻¹ and 6.9 ± 1.0 × 10⁷ M⁻¹ for LDL. The complexation was found to be pH dependent, increased for LDL and decreased for albumin (Mojzisova et al. 2007a) at lower pH values. It is a desirable feature that the PS is carried in the bloodstream by abundantly present proteins such as albumin. For the delivery to acidic tumor stroma, it is of particular interest that the binding constant is lowered at acidic pH to allow for the detachment of the PS from the protein carrier.

Ce6 was intracellularly detected after 15 min (Additional file 1: Fig. S1) which is in line with previous reports that could detect Ce6 in human fibroblasts 15 min after incubation (Mojzisova et al. 2007b). In a previous study, this form of Ce6 was used for PDT application against multidrug-resistant Staphylococcus aureus after 30 min of incubation (Winkler et al. 2016). The demonstrated fast cellular uptake and concentration-dependent increase in the fluorescence signal (Fig. 1c) implied the potential suitability of the used trisodium salt of Ce6 for outpatient time-sparing diagnosis sessions. In the case of 5-ALA, the increase in concentration and incubation period resulted in a plateau of accumulated PPIX (Fig. 1d). This plateau can be explained by the need for the enzymatic conversion of 5-ALA into PPIX which takes place in the mitochondria. The availability of the enzyme ferrochelatase is negatively affecting the accumulation of PPIX as confirmed by the significant increase in the accumulated PPIX upon administration of ferrochelatase inhibitors (Fukuhara et al. 2013; Teng et al. 2011).

Representative human cell lines of glioblastoma (U87) and breast cancer (Sk-Br-3 and MCF-7) in addition to normal skin fibroblasts (HF-5) were studied for their ability to accumulate Ce6 after 4 h of incubation and their corresponding phototoxic activity was studied. Phototoxicity results (Table 2) were in line with the cellular accumulation findings (Fig. 1e) and indicated that Sk-Br-3 was the least responsive cell line to Ce6-PDT. We hypothesized that a nanosystem capable of increasing the cellular uptake of Ce6 would overcome the observed Sk-Br-3 resistance to Ce6-PDT. Cationic nanoparticles were selected since they have repeatedly been shown to be internalized more efficiently by cancer cells as compared to anionic nanoparticles (Heuts et al. 2021; Jambhrunkar et al. 2014; Lin et al. 2019; Panariti et al. 2012), and lead to enhanced therapeutic outcomes (Wang et al. 2016; Wittig et al. 2013). Furthermore, the large surface area and pore volume of MSN should allow for high Ce6 loading.

The use of cationic MSN as carriers for Ce6 led to a higher accumulation of Ce6 in the Sk-Br-3 cells (Fig. 6a), for which otherwise a low cellular uptake and/or high efflux was observed, and the higher intracellular Ce6 concentrations achievable with MSN led to a pronounced increase in the PDT activity (Fig. 6b) (an up to sevenfold increase in cellular uptake as compared to free Ce6, and a tenfold increase in therapeutic response in vitro). This finding suggested that the poor internalization of free Ce6 by the Sk-Br-3 cells indeed was the main reason for the poor PDT efficiency of free Ce6 in this case. An efficient endosomal escape of Ce6 from the MSN preferentially residing in endosomes and lysosomes was confirmed by the observed cytoplasm localization of Ce6 after MSN-Ce6 administration (Fig. 6c).

The higher PDT activity achieved with MSN-Ce6 as compared to that of free Ce6 was also mirrored in vivo in a model for aggressive breast cancer. Compared to the allograft model developed by murine cell inoculation, the DMBA model, particularly when using rats instead of mice, is considered more aggressive and challenging to treat (Zeng et al. 2020). Here, a 1.7-fold increase in therapeutic outcome was observed for the intratumorally injected MSN-Ce6 as compared to free Ce6 (Fig. 7b) and correlated with an increase in caspase 3 expression (Fig. 8a). Those results are consistent with the obtained caspase-dependent (Fig. 3d) apoptotic death mode (Fig. 3b) in vitro findings and agreed with the elevation of caspase 3 and Bax levels reported for Ce6-PDT of osteosarcoma (Yu et al. 2020). Necrosis was also reported in our case which differed from the reported purely apoptotic cell death modes for the acidic form of Ce6 (Wawrzyńska et al. 2010), Ce6 polyvinyl pyrrolidone complex (Ali-Seyed et al. 2011; Chin et al. 2008) or Ce6-C-15-ethyl ester (Wang et al. 2022). One possible explanation for the observed necrotic cell death mode is a shift from an excessive apoptotic mode (Nicotera and Melino 2004).

The synthesized MSN-Ce6 showed a loading efficiency of 20% which is high in comparison to most literature reports on comparable systems (Table 2 in Additional file 1), and just slightly lower than the highest Ce6 loading degree reported in the literature, 22 wt%. The latter was obtained through the physisorption of acidic Ce6 to dendritic MSN from water (Xu et al. 2019). However, the latter case had a lower encapsulation efficiency. In our system, almost 40% of Ce6 was released in the first 4 h (Fig. 5b). The fast-release kinetics can be related to a detachment of aminosilanes from the silica network, in combination with the hydrolysis of the silica network, a process that can be catalyzed by the presence of basic amino-functions (Etienne and Walcarius 2003). The fast particle dissolution can also be connected to the fact that the silica network in co-condensed MSN has a lower degree of condensation than often used calcined MSN, and the degree of silica condensation is directly reflected in the dissolution kinetics (Möller and Bein 2019). However, we note that the silica dissolution-triggered release of Ce6 will probably be slightly lower in the presence of serum proteins (Lin et al. 2022), but the results highlight the suitability of the MSN as carriers for Ce6.

Our synthesized MSN-Ce6 significantly enhanced the therapeutic response of free Ce6 more than many reported nanosystems such as (Kim et al. (2020), Song et al. (2022) and Su et al. (2017) (Table 2 in Additional file 1) as indicated by the significantly mild conditions needed to reach IC50 at the in vitro level (lowest reported irradiation of 0.88 J/cm², among the shortest incubation periods of only 4 h and the lowest concentrations of only 5 µM). As listed, some nanosystems were less phototoxic than free Ce6 (Gaio et al. 2019; Lee et al. 2017) and others decreased cancer cell viabilities from 60 to 65% in the case of free Ce6 to 50% (Feng et al. 2019; Kumari et al. 2019; Xiao et al. 2012). Spheroids simulate allografts to a great extent (Pinto et al. 2020) and our hydrophilic Ce6 suppressed the spheroid volume by 75% compared to the control after one treatment session (Fig. 4e). Those findings are superior to an earlier study where higher fluence was applied (30 J/cm²), but to reach only 50% reduction in spheroid radius %, 10 days post-PDT were required (Gaio et al. 2019). At the in vivo level, our study is the only one using aggressive DMBA-induced rats (Zeng et al. 2020) as a breast cancer animal model (Table 2 in Additional file 1). A humane endpoint of the experiment was after 15 days, and the measured control tumor volume reached 3200 mm³ which clearly demonstrates the severity of the disease. Our synthesized MSN-Ce6 managed to suppress the tumor volume by 50% compared to the control. As listed, the obtained results are satisfying when compared to other nanosystems. It should be noted that the results of the used hydrophilic Ce6 were superior to the reported ones in which the hydrophobic form was used.

A low selectivity for cancer cells was observed for both free Ce6 (Table 2) and the MSN-Ce6 conjugate (Fig. 6d, e), although HF-5 cells produced the lowest ROS level after illumination (Fig. 3a). Molecular and biochemical reasons for differential cellular responses should be studied to fully understand the obtained results. Those findings suggest that further developments should be directed towards enhancing the uptake kinetics in the cancer cells by active targeting approaches.

Conclusion

This study has demonstrated the PDT efficiency of the water-soluble trisodium salt of Ce6 in treating breast and glioma cancers delineated by rapid cellular accumulation, ROS production upon illumination, and the induction of caspase-dependent apoptotic cell death mode in addition to a necrotic one. Conjugating Ce6 to MSN has significantly enhanced the cellular uptake and the PDT response of Ce6 at both the in vitro and the in vivo levels. Further developments should focus on improving the selectivity towards cancer cells.