Abstract
Background/Aim: 18 kDa Translocator protein (TSPO) is a mitochondrial protein up-regulated in colorectal carcinoma (CRC). Our purpose was to develop a TSPO-targeted doxorubicin prodrug (Dox-TSPO) which can be loaded onto drug-eluting beads for transarterial chemoembolization. Furthermore, we evaluated its loading and release kinetics and effects on cell viability. Materials and Methods: N-Fmoc-DOX-14-O-hemiglutarate was coupled with a TSPO ligand, 6-TSPOmbb732, using classical N,N,N’,N’-tetramethyl-O-(1H-benzotriazol-1-yl)uranium hexafluorophosphate coupling to produce Dox-TSPO. Loading and elution studies were performed using DC beads™. Cell viability studies were performed using CellTiter-Glo® Luminescent Cell Viability Assay. Results: Dox-TSPO was successfully synthesized and readily loaded onto and eluted from DC beads™, albeit at a slower rate than free doxorubicin. CRC cell lines expressing TSPO were 2- to 4- fold more sensitive to Dox-TSPO compared to free doxorubicin at 72 h. Conclusion: Dox-TSPO is a promising candidate for targeted and directed cancer treatment of CRC liver metastases.
Colorectal cancer (CRC) is the third leading cause of cancer mortality in the US, with 50,000 deaths annually (1). CRC frequently metastasizes to the liver, with 15% of patients presenting with synchronous hepatic metastases and 29% developing metachronous metastases within 3 years of diagnosis (2). Hepatic metastases increase mortality. The 5-year survival rate for patients with CRC if synchronous hepatic metastases are present is 2% compared to 90% if the cancer is localized to the colon (3).
The cytotoxic nature of systemic cancer treatments results in high rates of adverse effects making directed drug delivery systems an area of great interest. Transarterial chemoembolization (TACE) is a locoregional cancer treatment in which a chemotherapeutic agent and embolic material are directly injected into arteries supplying a tumor. This results in a dual method of tumor damage with higher concentrations of the chemotherapy in the area of the tumor than would be tolerated systemically and ischemia of the tissues secondary to embolization. A phase II study showed high response rates when colorectal hepatic metastases were treated with TACE using a combination of doxorubicin, mitomycin C, and cisplatin (4). Drug-eluting bead TACE (DEB-TACE) is a variant in which micro-beads made from biocompatible polyvinyl alcohol saturated with chemotherapeutic agent are injected into the tumor vasculature. The use of micro-beads results in sustained release, increasing the time that tumor cells are exposed to the agent. The micro-beads are permanent and result in durable occlusion of arteries. DEB-TACE with irinotecan beads for colorectal metastases to the liver was demonstrated to improve overall and progression-free survival, and quality of life when compared to systemic chemotherapy in a phase III study (5). To further improve drug delivery to tumor cells, there has been increasing interest in targeted agents.
The 18-kDa translocator protein (TSPO), previously known as the peripheral benzodiazepine receptor, is a transmembrane protein found in the mitochondrial membrane of cells throughout the body (6). It plays an important role in steroidogenesis through regulation of cholesterol translocation across the mitochondrial membrane, in cell proliferation by mediating translocation of pre-proteins needed for energy into mitochondria, and in apoptosis through regulation and maintenance of the transmembrane potential (6, 7). TSPO is up-regulated in brain, colorectal, breast, oral cavity, and prostate tumor cell lines (8-13). Due to the up-regulation of TSPO in a number of cancer cell lines, it is a potential target in directed oncological therapy.
Samuelson et al. experimented with targeting dendrimers using a TSPO ligand in C6 rat glioma cells and MDA-MB-231 human breast cancer cells, both of which have high TSPO expression. The targeted dendrimers were successfully internalized into the cancer cells but remained extracellular in the control group (14). Wyatt et al. performed a similar study synthesizing a TSPO-targeted near-infrared probe and showed successful targeting of MDA-MB-231 breast cancer cells in an in vivo nude mouse model (15).
TSPO targeting has also been studied for drug delivery. Margiotta et al. synthesized a novel picoplatin analog conjugated with a TSPO ligand and tested the targeted chemotherapeutic agent on SF188 and SF126 human glioma cells and C6 and RG2 rat glioma cells. The glioma cells exposed to the compound had 10- to 100-fold improved uptake when compared to free cisplatin. The new compound also demonstrated selectivity with 10-fold less activity against ovarian cells, which have low TSPO expression (16). In addition, our group recently reported a TSPO-targeted photosensitizer, IR700DX-6T, which caused effective TSPO+ cancer cell death and TSPO+ tumor inhibition through photodynamic therapy (17).
The purpose of this study was to develop and evaluate a novel drug delivery system including a TSPO-targeted doxorubicin prodrug (Dox-TSPO) and DEB-TACE. Doxorubicin is an anthracycline antibiotic that is an effective antineoplastic agent in many types of malignancy and has been successfully used in DEB-TACE for CRC metastasis to the liver (18); however, it is known to cause dose-limiting cardiotoxicity along with severe nausea, vomiting, and alopecia when used systemically (19). Evaluation included study of loading and release kinetics in beads used in DEB-TACE and in vitro cell viability studies. Through TSPO-targeting and DEB-TACE delivery, we expect improved drug efficacy and reduced adverse effects.
Materials and Methods
Chemicals. Pd[P(t-Bu)3]2, hexamethylenediamine, potassium hydroxide, cetyltrimethylammonium bromide, toluene, N,N,N’,N’-tetramethyl-O-(1H-benzotriazol-1-yl)uranium hexafluorophosphate (HBTU), N,N-diisopropylethylamine (DIPEA), dimethylformamide (DMF), piperidine, and trifluoroacetic acid (TFA) were purchased from Sigma Aldrich (Saint Louis, MO, USA). Doxorubicin hydrochloride was purchased from A Chemtek, Inc (Worcester, MA, USA). N-(9-Fluorenylmethoxycarbonyloxy)succinimide (Fmoc-OSu) was obtained from Oakwood Chemical (Estill, SC, USA). Glutaric anhydride was purchased from Alfa Aesar (Tewksbury, MA, USA).
Prodrug synthesis. N-(2-Bromo-5-methoxybenzyl)-N-(5-fluoro-2 phenoxyphenyl)acetamide, the TSPO ligand, was prepared according to Bai et al. (20). N-Fmoc-DOX-14-O-hemiglutarate, the doxorubicin prodrug precursor, was prepared according to Nagy et al. (21). Dox-TSPO was synthesized by coupling of N-Fmoc-DOX-14-O-hemiglutarate and 6-TSPOmbb732 using classical HBTU coupling followed by fluorenylmethoxycarbonyl protecting group (Fmoc) deprotection with piperidine according to Figure 1 (detailed in Supplementary Material). The final chemical structure was fully characterized by proton nuclear magnetic resonance spectroscopy (1H NMR), Carbon-13 nuclear magnetic resonance spectroscopy (13C NMR) and mass spectrometry (MS) analysis.
Loading doxorubicin and Dox-TSPO into beads. DC beads™ were purchased from BTG (London, UK). DC beads™ are biocompatible polyvinyl alcohol hydrogel microspheres that are chemically modified to allow for controlled loading and elution of chemotherapeutic drugs. They are used clinically for TACE procedures. A solution of DC beads™ (100-300 μm; 200 μl) were added to a cuvette that was maintained at 37°C. Sodium phosphate solution was removed from the bead solution and the remaining beads were rinsed with 200 μl of water three times before the cuvette was charged with 990 μl of water. Doxorubicin or Dox-TSPO solution (10 μl) in dimethyl sulfoxide (DMSO; 17 mM) were added to the cuvette and thoroughly mixed. The absorption of the solution at 480 nm was recorded every 30 seconds for the first 200 minutes and every minute thereafter. These loading studies were repeated using larger DC beads™ (300-500 μm; 200 μl).
Releasing doxorubicin and Dox-TSPO from beads. After DC beads™ were loaded with doxorubicin or Dox-TSPO, the excess loading solution was removed and the bead slurry was rinsed with 200 μl of water three times before being diluted with 200 μl of water. The loaded bead suspension (100 μl) was transferred to a cuvette and the excess water was removed. Dulbecco's minimum essential cell medium (DMEM; 1 ml) from Sigma Aldrich was added to the cuvette to initiate the release. The cuvette was incubated at 37°C for 7 days. The absorption of the solution at 480 nm was recorded every day. This was performed for solutions of both the 100-300 μm and 300-500 μm DC beads™.
18 kDa Translocator protein (TSPO)-targeted doxorubicin prodrug (Dox-TSPO) synthesis. Classical N,N,N’,N’-tetramethyl-O-(1H-benzotriazol-1-yl) uranium hexafluorophosphate (HBTU) coupling of doxorubicin and the TSPO ligand followed by fluorenylmethoxycarbonyl protecting group (Fmoc) deprotection, with a yield of 82%. DIPEA: N,N-Di-isopropylethylamine; DMF: dimethylformamide; TFA: trifluoroacetic acid.
Comparison of loading efficacy of doxorubicin (Dox) and the 18 kDa translocator protein (TSPO)-targeted doxorubicin prodrug (Dox-TSPO) into 100-300 μm (small) and 300-500 μm (large) DC beads™.
Cells. HT-29, and HCT-116 CRC cell lines and Hep G2 and Hep 3B hepatocellular carcinoma (HCC) cell lines were purchased from the American Type Culture Collection (ATCC; Manassas, VA, USA). The HT29 cells were maintained in McCoy's 5A medium with L-glutamine also purchased from the ATCC. The HCT-116, Hep 3B, and Hep G2 cells were maintained in DMEM from Sigma–Aldrich. Both media were supplemented with 10% fetal bovine serum (Sigma–Aldrich) and 100 U/ml penicillin. Cells were incubated at 37°C with 95% air and 5% CO2.
In-vitro cell-viability assays. Stock solutions of 1-2.5 mM were made with doxorubicin and Dox-TSPO using DMSO as a solvent. Cells were cultured in 55 cm2 flasks. When the culture flasks were saturated, cells were resuspended using trypsin and seeded into 96-well plates, 100 μl per well of the cells in medium, yielding approximately 5.7×104 cells per well (assuming 1×105 cells/cm2 in a 100% confluent culture). The plates were allowed to incubate for 24 h. The medium was then removed and replaced with fresh medium with either 0, 10, or 50 μM of Dox-TSPO or doxorubicin. Dilutions were performed immediately before the addition of the drugs to cells using distilled water. The number of viable cells was then measured at 24, 48, 60, and 72 h using the CellTiter-Glo® Luminescent Cell Viability Assay (Promega, Madison, MI, USA) according to the manufacturer's instructions. Luminescence of plates was read using a Synergy H4 Hybrid Multi-Mode Microplate Reader (BioTek, Winooski, VT, USA) 10 minutes after addition of the solution. Two independent experiments were performed in triplicate.
Statistical analysis. Data from cell viability studies are expressed as mean values with standard error (SE). Data were collated from two separate experiments performed in triplicate. Statistical comparisons of Dox-TSPO versus doxorubicin for each time point and drug concentration was performed using two-sided t-tests. All calculations were performed using PRISM 6 (GraphPad Software Inc., La Jolla, CA, USA). p-Values of less than 0.05 were considered statistically significant.
Comparison of elution kinetics of doxorubicin (Dox) and the 18 kDa translocator protein (TSPO)-targeted doxorubicin prodrug (Dox-TSPO) from DC beads™ in Dulbecco's minimum essential cell medium. Results are shown as the mean±standard error of independent trials using 100-300 μm and 300-500 μm DC beads™.
Results
Prodrug synthesis. The prodrug was synthesized as per Figure 1. 1H NMR, 13C NMR and MS analysis confirmed correct chemical structure. For greater detail regarding prodrug synthesis and NMR spectra (please see the Supplementary Material).
Loading doxorubicin and Dox-TSPO into beads. Dox-TSPO readily loaded into DC beads™ in a fashion similar to doxorubicin (22). When DC beads™ were charged with the orange-colored Dox-TSPO solution, the color of the drug solution slowly diminished. Simultaneously, as the colored drug molecules moved inside the bead cavities, the blue-colored beads turned red. Eighty-eight percent of Dox-TSPO was loaded versus over 99% of the same amount of doxorubicin into a 200 μl solution of DC beads™. Free doxorubicin required a 3-h loading time and Dox-TSPO a 5-h loading time in order to reach equilibrium (Figure 2). There was no significant difference in loading rates between the 100-300 μm and 300-500 μm DC beads™.
Release of doxorubicin and Dox-TSPO from beads. The majority of Dox-TSPO was released in the first 3 days from the DC beads™, after which the process became progressively slower as it approached equilibrium. Dox-TSPO reached maximum release in 1 day and greater than 30% of the drug was eluted after 3 days (Figure 3). There were similar rates of release with the 100-300 μm and 300-500 μm DC beads™. Dox-TSPO was not completely released, and the beads did not lose the characteristic dark red color indicating the presence of residual Dox-TSPO at equilibrium. Doxorubicin was also partially retained in the beads due to its strong ionic interaction with the sulfates, however, it demonstrated a higher proportion of drug elution, with 50% of the drug having been released at 3 days (23, 24).
Comparison of loss of viable cells following incubation with the 18 kDa translocator protein (TSPO)-targeted doxorubicin prodrug (Dox-TSPO) versus doxorubicin at different time points and concentrations in colorectal cancer (HT-29 and HCT-116) and hepatocellular carcinoma (Hep G2 and Hep 3B) cell lines.
In vitro cell-viability assays. Table I and Figure 4 summarize the cell viability results of Dox-TSPO and free doxorubicin against HT-29, and HCT-116 CRC cell lines and the Hep G2 and Hep 3B HCC cell lines. The HT-29 cell line overexpresses TSPO as documented in the NCI-60 Cancer Microarray database. HCT-116 expresses the median amount of TSPO when compared to cancer cell lines in the NCI-60 database and Hep G2 and Hep 3B both underexpress TSPO (25). A significant time- and concentration-dependent reduction in viable cells in all cell lines with both Dox-TSPO and doxorubicin versus controls was observed (p<0.05). In the HT-29, Hep G2, and Hep 3B cell lines, Dox-TSPO caused a significant loss of viable cells at earlier time points than did doxorubicin. In these three cell lines, there was a significant loss of viable cells when incubated with 10 μM of Dox-TSPO at 24 h, with doxorubicin not causing a significant loss until 48 h (p<0.05 for all). No difference in onset of significant loss in number of viable cells was seen at the other concentrations. In the HCT-116 cell line, both Dox-TSPO and doxorubicin caused a significant reduction in number of viable cells at all concentrations at 24 h (p<0.05).
In the HT-29, HCT-116, and Hep G2 cell lines, at the majority of time and concentration points, Dox-TSPO performed superiorly to free doxorubicin (Table I). In the HCT-116 cell line at 48 and 72 h and in the HT-29 cell line at 24 and 72 h, Dox-TSPO resulted in significantly higher loss of viable cells versus doxorubicin at all concentrations (p<0.05 for all). In HT-29 cells, 10 μM of Dox-TSPO also led to significantly superior loss of viable cells compared to doxorubicin at 48 and 60 h and 50 μM at 60 h (p<0.05). Of the cell lines with a low expression of TSPO, a consistent increased loss in viable cells with Dox-TSPO was also found at 24, 48, and 60 h, with a plateau effect in the Hep G2 cell line at 72 h for all concentrations (p<0.05 for all). A consistent difference in number of viable cells between the two drugs was not observed in the Hep 3B line.
When cell lines expressing TSPO were directly compared to those with low TSPO expression, they were found to be more sensitive to Dox-TSPO with 2- to 4- fold more viable cells in the doxorubicin-treated groups compared to the Dox-TSPO-treated groups at 72 h. There was not a significant difference in number of viable cells between the Dox-TSPO- and doxorubicin-treated Hep G2 and Hep 3B cell lines at this time point.
Discussion
Dox-TSPO was successfully synthesized as confirmed by NMR and MS analysis. It readily loaded into DEB-TACE beads, albeit at a slower rate than free doxorubicin, 5 versus 3 h. Doxorubicin and Dox-TSPO are hydrophobic molecules. When the molecules enter the DEB-TACE beads, water molecules are displaced. The larger Dox-TSPO requires more water molecules to be released from the bead cavity. This adds an energy burden, likely contributing to the longer drug loading time. In addition, the ionic interaction (22) between the protonated primary amine on doxorubicin and sulfonate groups inside the bead cavity is the main driving force when loading doxorubicin into DEB-TACE beads. The added secondary amine on the TSPO targeting moiety is a p-methoxyaniline derivative, which is not protonated at the pH level of the loading solution. Therefore, the loading process of Dox-TSPO did not kinetically benefit from the additional amine group, also likely contributing to the slower loading. Clinically, this difference in loading time may not be significant given beads are preloaded with chemotherapy prior to DEB-TACE procedures.
Cell viability of HT-29 (A) and HCT-116 (B) colorectal cancer cell lines and Hep G2 (C) and Hep 3B (D) hepatocellular carcinoma cell lines. The number of viable cells following incubation with the 18 kDa translocator protein (TSPO)-targeted doxorubicin prodrug (Dox-TSPO) or free doxorubicin (Dox) is presented relative to those of the control (mean±standard error). Two independent trials were performed in triplicate.
After Dox-TSPO entered the bead cavity, sulfonate groups were expected to surround the molecule. In this significantly different solvation environment, the acid dissociation constant (Ka) of the secondary amine may have shifted (23), resulting in protonation and therefore additional ionic interactions with nearby sulfonates. The thermodynamically favored interaction was demonstrated when attempts to release the molecule from the beads using saline solution were found to be unsuccessful. Unlike doxorubicin, which was promptly released from DEB-TACE bead in saline, little to no Dox-TSPO was released under the same condition. The ionic strength of the saline solution was insufficient to overcome the fortified force and displace the Dox-TSPO from inside beads. We later found that DMEM is capable of releasing Dox-TSPO from the DC beads™, albeit more slowly when compared to free doxorubicin. This slower elution may actually be of benefit in cancer treatment as this can allow for longer exposure times of tumor tissues to the chemotherapeutic agent.
We hypothesized that Dox-TSPO would be more effective in causing a loss of viabiIity in cell lines with high TSPO expression secondary to cell targeting and enhanced uptake. This was supported by our data as our two cell lines that expressed TSPO showed increased sensitivity to Dox-TSPO when compared to free doxorubicin, with the doxorubicin-treated cells having a 2- to 4-fold higher number of viable cells following incubation for 72 h. The cell lines that underexpress TSPO did not show a significant difference in viable cells following this length of incubation. This finding suggests specificity of the prodrug.
Loss of viable HT-29, Hep G2, and Hep 3B cells occurred at a significantly faster rate with Dox-TSPO than with free doxorubicin at identical concentrations. It is unclear whether the same would hold true for the HCT-116 cell line as both agents caused significant loss of viable cells at the first time point. This is likely explained by the targeting moiety accelerating drug uptake into cells. A similar study using CD19-targeted doxorubicin in Burkitt's lymphoma cells showed increased efficacy and specificity of the targeted drug when compared to free doxorubicin alone. Subsequent flow cytometry demonstrated that the targeted drug accumulated more rapidly within the cells, thus explaining its enhanced effect (26). Future internalization studies would be important to perform to confirm this mechanism for Dox-TSPO. The presumed accelerated uptake of Dox-TSPO, is a benefit as it can reduce the time that doxorubicin stays in the systemic circulation.
Further investigation needs to be performed to elucidate the mechanism of action of Dox-TSPO, its in vivo effects, localization, efficacy, and adverse effects. If further studies continue to support the improved targeting of doxorubicin to tumor cells using TSPO as a ligand, there is potential for the ligand to be conjugated with other chemotherapeutic agents to expand its utility.
We were able to successfully synthesize, load, and release our novel mitochondria-targeted doxorubicin prodrug from DC beads™. In addition, the prodrug resulted in superior loss of viable cells when compared to doxorubicin in human HCT-116 and HT-29 CRC cancer cell lines and similar rates of loss of viable cells to doxorubicin in the human Hep G2 and Hep 3B HCC cell lines. Because of its equivalent to superior effectiveness in reducing viable cells when compared to free doxorubicin and its capacity to be used with DEB-TACE beads, the targeted prodrug is a promising candidate for targeted and directed cancer treatment in CRC.
Footnotes
↵* These Authors contributed equally to this study.
Authors' Contributions
Study Conception and study design: HSK and MB. Collected the data: JBJ, XL, MX and JML. Contributed data and analysis tools: JBJ, XL, MX and JML. Performed analysis: JBJ, XL, MX and JML. Initial draft of the article: JBJ and XL. Editing: JBJ, XL, MX, JML, MH and HSK. Final approval: JBJ, XL, MX, JML, MH and HSK. Supervision of the study: HSK and MH.
This article is freely accessible online.
Conflicts of Interest
The Authors have no relevant conflicts of interest.
Supplementary Material
Available at: https://sites.google.com/view/dox-tspo-supplemental-material/home
- Received July 30, 2020.
- Revision received August 16, 2020.
- Accepted August 23, 2020.
- Copyright© 2020, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved
