Targeting triple‐negative breast cancer with an aptamer‐functionalized nanoformulation: a synergistic treatment that combines photodynamic and bioreductive therapies

Background Areas of hypoxia are often found in triple-negative breast cancer (TNBC), it is thus more difficult to treat than other types of breast cancer, and may require combination therapies. A new strategy that combined bioreductive therapy with photodynamic therapy (PDT) was developed herein to improve the efficacy of cancer treatment. Our design utilized the characteristics of protoporphyrin IX (PpIX) molecules that reacted and consumed O2 at the tumor site, which led to the production of cytotoxic reactive oxygen species (ROS). The low microenvironmental oxygen levels enabled activation of a bioreductive prodrug, tirapazamine (TPZ), to become a toxic radical. The TPZ radical not only eradicated hypoxic tumor cells, but it also promoted therapeutic efficacy of PDT. Results To achieve the co-delivery of PpIX and TPZ for advanced breast cancer therapy, thin-shell hollow mesoporous Ia3d silica nanoparticles, designated as MMT-2, was employed herein. This nanocarrier designed to target the human breast cancer cell MDA-MB-231 was functionalized with PpIX and DNA aptamer (LXL-1), and loaded with TPZ, resulting in the formation of TPZ@LXL-1-PpIX-MMT-2 nanoVector. A series of studies confirmed that our nanoVectors (TPZ@LXL-1-PpIX-MMT-2) facilitated in vitro and in vivo targeting, and significantly reduced tumor volume in a xenograft mouse model. Histological analysis also revealed that this nanoVector killed tumor cells in hypoxic regions efficiently. Conclusions Taken together, the synergism and efficacy of this new therapeutic design was confirmed. Therefore, we concluded that this new therapeutic strategy, which exploited a complementary combination of PpIX and TPZ, functioned well in both normoxia and hypoxia, and is a promising medical procedure for effective treatment of TNBC.

presence of high levels of specific reductases, to develop new therapeutic strategies. Rapidly growing tumor cells are often exposed to hypoxia, a common sign of stress. Tumor cells may multiply farther away from the blood supply, which results in a relatively low oxygen tension of <8 mm Hg (1 %) compared with a normal blood oxygen pressure of 70 mm Hg or 9.5 % [14].
Photodynamic therapy (PDT), one of the clinically approved treatments, is a minimally invasive approach for treating various cancers [15]. The mechanism of PDT is based on the activation of photosensitizers (PSs) to the excited singlet state after visible light absorption, followed by intersystem crossing to the excited triplet state [16]. The excited PSs undergo photochemical reactions with oxygen (O2) to form cytotoxic reactive oxygen species (ROS) [17], which not only eradicates cancer cells, but also normal cells such as blood vessel epithelial cells. PDT has been implemented as an anti-breast cancer strategy [18], but the hypoxic tumor microenvironment in tumorous breast tissues often impeded the utility of PSs, such as PpIX, and caused undesirable consequences, such as angiogenesis, low rate of cancer cures, and increased likelihood of tumor recurrence [19]. In other words, PDT alone may enable a blockade of nutrients and oxygen in the cancerous areas, resulting in the killing of cancerous cells, but the formation of hypoxic areas somehow alters cancer cell metabolism and may thereby contribute to therapy resistance [20]. Therefore, it is clear that tumor hypoxia remains as one of the greatest challenges in treating solid tumors because cancer cells in such regions are a potent barrier to effective radiation therapy and immunotherapy [21]. Since oxygen consumption is a limiting factor for PDT [22], several strategies have been developed to improve its therapeutic efficacy to increase radical formation. Xia and co-workers [23] introduced oxygen-independent free radicals produced by a polymerization initiator system to destroy hypoxic cancer cells. Additionally, BD or hypoxia-activated prodrugs (HAP) have also been alternatives to eliminate hypoxic cancer cells.
It is known that both BD and HAP are inactive but can be converted into potent toxins under conditions of either low oxygen tension or in the presence of high levels of specific reductases [24,25].
For example, the cytotoxicity of drugs RB-6145, SR-4233 (TPZ), and E09 (Aqaziquone) that were used in a hypoxic environment, was approximately 50200 fold higher than that in an aerobic environment [26]. TPZ is a class of cytotoxic drugs with selective toxicity towards hypoxic mammalian cells that can be catalyzed by NADPH: cytochrome c (P450) reductase to form toxic hydroxyl and benzotriazinyl radicals, followed by the generation of ROS to damage DNA when a cell was deprived of oxygen [27].
TPZ has been evaluated in clinical trials in non-small cell lung cancer, head and neck cancer, cervical cancer, and metastatic melanoma [25]. Regrettably, it did not show satisfactory results as originally anticipated in clinical trials due to low cellular uptake efficiency, unsatisfied pharmacokinetics, and adverse side effects [25]. Based on the hypoxic tendency of TNBC and the complementary functions of PS and BD, we were motivated to target normoxic and hypoxic tumor areas by adopting PDT in conjunction with bioreductive therapy to evaluate the synergistic antitumor effects of this new nanoVector-assisted therapeutic strategy for TNBC. Co-delivery of PpIX and TPZ can be realized readily using hollow mesoporous silica nanoparticles (HMSNs), which is an ideal type of drug carrier because of its biocompatibility, degradability, high loading capacity and versatile surface chemistry [28][29][30][31][32].
In this study, MMT-2, a novel type of thin-shell HMSNs with three-dimensionally interconnected mesopores we previously developed [33], was applied to integrate the therapeutic utilities of PpIX and TPZ and the targeting capability of the DNA aptamer LXL-1 for TNBC cell line MDA-MB-231. PpIX and LXL-1 were modified covalently on the mesopores and external surface of MMT-2, respectively, and TPZ was finally loaded largely into the hollow interior of the functionalized MMT-2, designated as LXL-1-PpIX-MMT-2, by impregnation (Scheme 1). In vitro and in vivo studies showed that the obtained nanoVector TPZ@LXL-1-PpIX-MMT-2 was accumulated selectively at the tumor site and demonstrated high efficacy in killing MDA-MB-231 cells in normoxic and hypoxic areas under 630 nm irradiation; the synergistic effect was manifested clearly in a complete tumor eradication with enhanced efficiency.
Our results confirmed that this reliable nanomedical platform offers a promising strategy for TNBC targeted therapy, and it provides a solution for limited therapeutic efficacy that is often associated with PDT due to the deprivation of oxygen level in cancer cells.

Characterization of materials
The as-synthesized MMT-2 displayed a characteristic XRD (X-ray diffraction) pattern that corresponded to Ia3d symmetry (cf. Figure S1), and the hollow morphology and the thin shell with ordered mesostructure of each nanoparticle could be observed by TEM (transmission electron microscopy) (Figure 1a). The ordered mesostructure was retained after subsequent steps of functionalization, as evidenced by the TEM image of LXL-1-PpIX-MMT-2 (Figure 1b). Successful functionalization of maleimide groups on the external surface and PpIX on the mesopores were confirmed by TGA (thermogravimetric analysis) and FTIR (Fourier-transform infrared spectroscopy), and the final conjugation of the DNA aptamer LXL-1 with maleimide groups was supported by the change in surface potential. The relative organic content of M-PpIX-MMT-2 (maleimide functionalized, PpIX-anchored MMT-2) was higher than that of M-MMT-2 (maleimide functionalized MMT-2), which showed weight losses of 29.8 wt% for M-PpIX-MMT-2 and 12.7 wt% for M-MMT-2, as revealed by TGA (Figure 1c).
Characteristic IR signals confirmed the presence of the maleimide groups and PpIX (Figure 1d). The amount of PpIX in M-PpIX-MMT-2 was estimated to be ~0.343 mole per gram of the sample by analyzing the absorbance at 405 nm [34] in the UV-vis spectrum of the sample. After conjugation of the highly negatively charged DNA aptamer, the zeta potential measured in PBS changed from -19 mV for M-PpIX-MMT-2 to -38 mV for LXL-1-PpIX-MMT-2 (Figure 1e). The hydrodynamic size of LXL-1-PpIX-MMT-2 was around 345.0 ± 3.4 nm as measured by dynamic light scattering (DLS) (Figure 1e). In addition, the cell viability of MMT-2 and LXL-1-MMT-2 on MDA-MB-231 cells was shown in Figure S2.

Cellular uptake and in vivo targeting efficiency of LXL-1-PpIX-MMT-2
The cell uptake and targeting efficiency of LXL-1-PpIX-MMT-2 toward MDA-MB-231 breast cancer cells (a TNBC cell line) were investigated. After treating with either free PpIX or LXL-1-PpIX-MMT-2 for 5 h, cells were harvested and lysed to determine the uptake of PpIX by a fluorospectrometer. In addition, a Laser Confocal Microscope was used to monitor targeting efficiency. (Figure 2a). The LXL-1-PpIX-MMT-2-treated group demonstrated four times more PpIX accumulation than that of the cell group treated with PpIX alone. The confocal microscopy imaging also indicated that cells treated with LXL-1-PpIX-MMT-2 (for 5 h) showed greater enhanced fluorescence than that in the free PpIX-treated group (Figure 2b). These results suggested that LXL-1-PpIX-MMT-2 was able to be taken up and accumulated in MDA-MB-231 cells (Figure 2a, b).
Next, the targeting efficiency of LXL-1-PpIX-MMT-2 was studied. Three breast cancer cells were Due to structural characteristics, the hydrophobic PpIX is supposedly to deposit in hepatic rather than renal excretion [35]. Our LXL-1-PpIX-MMT-2 was designed to increase PpIX accumulation in tumor, but to reduce in the other organs. In vivo targeting was revealed by an IVIS (Figure 2e). Aside from the tumor site, free PpIX accumulated significantly in the liver, lung and kidney. On the other hand, LXL-1-PpIX-MMT-2 was apparently retained in the tumor, and somewhat less in the others. There was no doubt that our LXL-1-PpIX-MMT-2 helped to deliver the cargo drug to the targeted region, which enhanced the accumulation of the PS in the tumor. These results showed that LXL-1-PpIX-MMT-2 was able to target MDA-MB-231 xenografts without significant residuals in other organs, which promoted safety and therapeutic efficacy.

Effect of oxygen level on photodynamic cytotoxicity
Considering the basic principles of PDT, factors including oxygen level, irradiation time, and the concentration of photosensitizer are presumably key elements that decide the efficacy of PDT. Optimal conditions to maximize the effectiveness of PDT were, therefore, investigated. Figure S3 showed the efficiency of PpIX for generating singlet oxygen upon irradiation under different oxygen level. Figure   S4 revealed the effect of oxygen level on in vitro production of ROS in cancer cells. In addition, MDA-MB-231 cells were cultured at either a normal oxygen level or under hypoxic conditions (at 5%, 2%, and 1% oxygen) to examine the minimum oxygen level to obtain an acceptable photodynamic therapeutic outcome. Significant photodynamic cytotoxicity of PpIX at oxygen level of 21% and 5% was observed, whereas reduced cytotoxicity was exhibited under hypoxic conditions (2% and 1% oxygen level) (Figure 3). With sufficient oxygen supply, cytotoxicity increased with the elevated amount of photosensitizer and irradiation time. In contrast, under hypoxic conditions, decreased cytotoxicity occurred at a relatively high photosensitizer concentration (0.8 μM of PpIX, equivalent to 0.46 μg/mL) and long irradiation period (4 min). Our results agreed with previous studies [36] indicating that satisfactorily high oxygen level was required to photoactivate PpIX to induce photodynamic cytotoxicity.
Based on our results, 0.4 μM PpIX at 2% oxygen was able to eradicate ~50% of treated cells, thus, 0.4 μM PpIX was selected for further study (Figure 3b).

Effect of oxygen level on TPZ cytotoxicity
We know that high levels of oxygen caused cytotoxic TPZ radicals to become less harmful TPZ molecules. The low oxygen level ranging from 0.3% to 4.2% in tumor microenvironments has been discussed extensively [14], which encouraged us to examine TPZ cytotoxicity under various hypoxic conditions (oxygen levels: 1%, 2%, 5%) and under normoxia (oxygen level: ~21%). As anticipated, the observed cytotoxicity was enhanced with the increase in TPZ concentration and lower oxygen level (Figure 3e). TPZ displayed low toxicity (cell viability ~80% at 60 μM) at a high oxygen level (~21%). With a limited supply of oxygen, improved cytotoxicity was revealed even at a low TPZ concentration (cell viability ~50% at 20 μM). In addition, significant cytotoxicity (<50% cell viability) was found at a higher TPZ concentration (60 μM, equivalent to 11 μg/mL) with low oxygen levels (such as, 5%, 2%, and 1%).
Therefore, a TPZ concentration of 60 μM was identified as the optimum effective dosage for further studies. Our observations agreed with the results as reported in previous studies [24,37,38], in which the cytotoxicity of TPZ was inversely associated with oxygen level.

Synergistic effect of PDT-and TPZ-based combination therapy
The antitumor effects of PDT highly depend on the tumor oxygen level, but are hindered by hypoxic tumor microenvironments. To improve poor effectiveness of PDT associated with tumor hypoxia, we established a new therapeutic approach that combined two cancer drugs that work in a complementary fashion. PDT requires sufficient oxygen to generate toxic radicals that are harmful to tumor cells, so bioreductive prodrugs that can be activated to be highly toxic under low-oxygen conditions were a perfect match. Therefore, the combination treatment of PpIX and TPZ was conducted in vitro. With the elevated oxygen level, the cytotoxicity of PpIX and TPZ showed opposite trends (Figure 4a). Cell viability increased from 31% to 88% for the PDT-only group with the lower oxygen level (5% to 1%), but cell viability in the TPZ-only group decreased from 42% to 35% (oxygen level from 5% to 1%). Once we combined free PDT with free BD, elevated cytotoxicity was observed for all groups, in general, at different oxygen levels. However, cell viability increased from 4% to 22% with the decrease in oxygen level from 5% to 1%, which indicated that PDT played a dominant role in determining therapeutic efficacy. Moreover, the synergistic effect provided by this new combination treatment was observed because CDI (coefficient of drug interaction) values of 0.3, 0.49, and 0.7 were obtained at oxygen levels of 5%, 2%, and 1%, respectively, whereas CDI values that were more than or equal to one indicated antagonistic or additive effects, respectively [39] (Figure 4b). It was also claimed previously [37,40] that the combination of PDT and HAP prodrugs increased cell cytotoxicity synergistically.
Furthermore, it is noteworthy that the combination treatment with two free drugs exhibited less cytotoxicity at a low oxygen level of 1% compared with higher oxygen levels (5% and 2% O2); however, the combination treatment with our nanoVector, TPZ@LXL-1-PpIX-MMT-2, further decreased cell viability not only at the 1% oxygen level, but also at 2% and 5%, which holds promising potential to be used in hypoxic environments for tumors. We believe that this was due to the effective targeted delivery of PpIX and TPZ to MDA-MB-231 cells.
Although numerous reports have demonstrated evidences on the utilities of nanoDrug Delivery Systems in vitro and/or in vivo, limited research was conducted to evaluate therapeutic efficacy of nanotherapy on hypoxia formation and cytotoxicity in hypoxic regions. The use of nanoVectormediated combination therapy based on the complementarity of PDT and BD to enhance therapeutic efficacy against cancer, especially for tumor hypoxia, was addressed herein. We again confirmed that low oxygen level impaired PDT cytotoxicity, but promoted the activity of TPZ (cf. Figure 3, 4), which was in agreement with previous findings [25,38,40,41].
TNBC is aggressive with high mortality and difficult to treat [42]. The unsatisfactory therapeutic outcomes of conventional chemotherapy and therapeutic agents, primarily poly(ADP-ribose) polymerase inhibitors and EGFR inhibitors, argue for development of an effective targeted therapy for this ER/PR/HER2 receptor expression-lacking tumor. A genetic mutation in p53 has been revealed recently in TNBC that could be a therapeutic target [43]. Interestingly, the cytotoxicity of TPZ was observed previously in p53-dysfunctional epidermoid carcinoma (A431) cells [41]. In fact, there are a number of studies that utilized TPZ in combination with cisplatin to treat head and neck cancer, lung cancer, and breast cancer [44]. The utility of our nanoVector, together with findings obtained from previous studies [40,41], validated the effectiveness of PDT/BD combination therapy to eradicate cancer cells with the TP53 mutation, which offers an alternative approach for TNBC treatment.

Antitumor activity of LXL-1-PpIX-MMT-2 in a MDA-MB-231 xenograft tumor model
Conventionally, chemotherapy is often given after surgery because information collected from postsurgical pathology is necessary to determine the optimum regimen for cancer treatment. Today, given the increasing interest in local/regional therapy, localization of the tumor is feasible [45]. Numerous molecular approaches for diagnosis and characterization of breast tumors are available to provide detailed information to predict chemotherapy outcomes before surgery [46]. With the precise localization of tumors, we believe that the direct injection of chemotherapeutic drugs at the site of the tumor will enable the relief of serious systematic toxicity caused by the drugs themselves. Accordingly, intratumoral administration was performed in our in vivo study, which attempted to further improve the survival and quality of life for patients.
To evaluate further the therapeutic effectiveness of this novel nanotherapeutic strategy, we used NU/NU female mice (4 wk old) that carried human breast tumor xenografts in two thighs. NanoVectors and free drugs were administrated i.t. as described previously (Figure 5a). Treatment with TPZ@LXL-1-PpIX-MMT-2 demonstrated the best therapeutic efficacy among all experimental animal groups (Figure 5b, c). Additionally, no significant body weight loss was observed during the study period ( Figure 5d). Moreover, as evidenced by H&E staining (Figure 5e), tumors treated with our nanoVectors showed reduced cell density compared with those groups treated with single free drugs (PpIX or TPZ), or a combination of free drugs (PpIX + TPZ). The tumor hypoxic area was also examined by immunohistochemical staining of pimonidazole-protein adducts in hypoxic areas (Figure 5e). The hypoxic zone in the PpIX-treated group was larger than that of the PBS-treated and TPZ-treated groups.
TPZ@LXL-1-PpIX-MMT-2 not only restrained the formation of notable hypoxia, but also promoted cell death in the same region as observed by reduced cell density compared with the PBS group. PDT increased hypoxia due to its inherent cytotoxic mechanism, where photosensitizers interacted with oxygen to form ROS that led to the formation of a hypoxic tumor microenvironment.
In summary, MMT-2 comprising thin-shell hollow mesoporous silica nanoparticles was selected as the drug vector for PDT/BD combination therapy. The material featured large hollow interior, thin mesoporous shell and uniform particle size, and was promising for the development of drug delivery systems. The interstitial hollow cavities served as depots to accommodate various therapeutic agents, and mesopores enabled therapeutic agents to diffuse through the shell. Furthermore, the surface silanol groups on the mesopores and external surface enabled versatile and selective functionalization for anchoring targeting (e.g. DNA aptamer LXL-1) or functional (e.g. photosensitizer PpIX) moieties. In short, we developed a novel nano combination therapeutic approach that targeted TNBC. The combination of PDT and TPZ eradicated cancer cells synergistically and effectively in both normoxic and hypoxic regions of tumor tissues. This nanotherapy enhanced the retainment of chemotherapy drugs in tumors, yet decreased drug accumulation in the other non-target organs, which suggested it is a promising strategy for treating TNBC. Our study not only verified the feasibility of PDT/BD combination therapy in cancer treatment, but also paved the way for the development of a therapeutic strategy for malignant neoplasm in hypoxic regions.

Conclusion
Given the lack of effective treatments for TNBC, numerous efforts have been devoted in the past to augment therapeutic opportunities for TNBC patients. The phase III IMpassion130 trial using chemotherapy plus atezolizumab (a fully humanized, engineered monoclonal antibody of IgG1 against the programmed cell death ligand 1, PD-L1) compared with chemotherapy plus placebo brought breast cancer into the era of antibody-based therapeutic approaches; however, limitations of the therapeutic antibody approach included high medical cost, poor tissue accessibility, insufficient pharmacokinetics, and imperfect interactions with the immune system. Previous studies have been reported on the applicability of nanoDrug Delivery Systems; however, the effectiveness of PDT/BD combination nanotherapy in tumor hypoxia was less frequently discussed. Herein, we successfully developed a synergistic approach to target TNBC under both normoxia and hypoxic conditions. The use of HMSNs modified with the aptamer, LXL-1, was confirmed to target TNBC and release TPZ to eradicate tumors under hypoxic conditions. On the other hand, a photosensitizer that was fixed inside HMSNs generated a sufficient level of radicals to shrink tumors under normoxic conditions with PDT.
This design employed the mechanism of action using a combination of two medicines, which demonstrated promising potential for TNBC therapy. These observations encourage us to conduct further investigations of our nanoVector to treat hypoxia-associated diseases because hypoxia-induce heterogeneous environments promote tumor invasiveness, angiogenesis, drug resistance, and metastasis, and impair therapeutic efficacy.

Chemicals and reagents
All chemicals and reagents were of analytical grade and were used as received without further purification. Benzylcetyldimethyl-ammonium chloride (BCDAC, 97%), bovine serum albumin (BSA),

Preparation of TPZ@LXL-1-PpIX-MMT-2
MMT-2 was synthesized following the procedures reported previously [33]. In a typical synthesis, TEOS that contained 5% CO2 at 37 o C. The growth medium was changed every 48 h, and cells were trypsinized (using 0.1% trypsin) and subcultured when they grew to about 90% confluence.

In vitro effectiveness of PpIX, TPZ, and TPZ@LXL-1-PpIX-MMT-2
To investigate the in vitro effectiveness of PpIX, TPZ, and nanoVectors, cell viability studies at both normoxia and hypoxia condition were conducted. For normoxia effectiveness, MDA-MB-231 cells were first seeded in a 96-well plate at a density of 10 4 per well/well and then placed directly in a cell incubator for 18 h after addition of designated drugs. To test for hypoxia effectiveness, home-made double-layered atmosphere bags (atmobag) filled with the desired gas composition, which was N2 5% CO2, and oxygen levels of 5%, 2%, or 1%, were used to mimic the hypoxic condition experimentally. We treated tested cells with TPZ@LXL-1-PpIX-MMT-2 that was suspended in PBS originally and mixed with culture medium at an appropriate ratio prior to use, or appropriate concentration of free PpIX, which was dissolved in DMSO originally and mixed with culture medium at a ratio of 1:99 prior to use. The

Histological analysis of tumors
For histological studies, tumor tissue was fixed in 10% formalin for one week and embedded in paraffin.
Tumor tissue was sectioned (3 m) before being fixed on glass slides and allowed to dehydrate overnight. Sections were subjected to the dewaxing and rehydration through the use of xylene and a series of decreasing alcohol concentrations (100%, 95%, 90%, 80% ethanol/ddH2O, and finally ddH2O).
For hematoxylin and eosin stain (H&E) analysis, sections were stained with hematoxylin and eosin to confirm the cell density and to observe the details of cellular and tissue structures. To visualize hypoxic areas immunohistochemically, a commercially available hypoxyprobe kit (Hypoxyprobe™-1 Omni kit, Hypoxyprobe, USA) was used according to the manufacturer's protocol. In brief, each group of animals 20 was i.p. administered with 60 mg/kg Hypoxyprobe™-1 solution (Pimonidazole) 1 h before sacrifice; the tumor tissue that was intended to be analyzed for the amount of hypoxia was prepared as described above. Next, the deparaffinized tissue sections were treated with 3% H2O2 to block endogenous peroxidase activity, followed by incubation with FBS to reduce non-specific binding. The primary antibody (PAb2627A) (1:200, Hypoxyprobe, Inc, USA) was added to the tissue section-mounted slide and allowed to react overnight at 4 °C. After washing three times with Tris-buffered Saline (TBS) with tween-20, the slide was subsequently incubated with the secondary antibody for 1 h to complete tissue preparation for immunostaining.

Statistical analysis
Experiments were performed in triplicate and repeated at least three times. Data were presented as means ± standard deviation (SD). The t-test was used to evaluate whether there was any statistical significance between the means of two independent groups. In this study, p-values of <0.05 represented results that were statistically significant, and p-values of <0.01 were considered to be highly statistically significant.

Supplementary information
Supplementary information accompanies this paper at https :// Additional file 1: Figure S1 XRD pattern of MMT-2. Figure S2.

Competing interests
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Ethics approval and consent to participate
All animal experiments conducted in current study were performed in compliance with the NHMRC Taiwan Code of Practice for the care and use of animals for scientific purposes, and approved by the institutional animal care and use committee (IACUC) of National Taiwan University.

Consent for publication
Not applicable.

Availability of data and material
All data generated or analyzed during this study are included in this manuscript.

Funding
The authors gratefully acknowledge the financial support from the Ministry of Education of Taiwan