1. Introduction
The primary mechanism of common cytotoxic chemotherapeutic drugs involves enhancing cellular DNA damage and/or depleting DNA-repair mechanisms to prevent further cell proliferation, thereby rendering eventual necrosis of the tumor [
1,
2]. Following cytotoxic damage, extensive cellular repair mechanisms are activated, which lead to drug resistance and thereby further complicate the determination of a chemotherapy dosage over time. Currently, doses and dosing schedules are derived by prospective and retrospective experimental studies, which determine the putative efficacious dose, while being less than the threshold for toxicity [
3,
4]. Starting doses for clinical trials are determined by the dose that results in 10% lethality in animal models rather than in the direct measures of the drug’s mechanism (e.g., extent of DNA damage) [
5]. Apart from the blood biomarkers or blood concentrations of the drug, there is a lack of significant biomarker metrics to assess the efficacy of treatment in an acute setting [
6,
7,
8,
9]. Thus, most chemotherapeutic treatment schedules are based on the toxicity tolerance and the lack of adverse symptoms/events. As such, methods to directly quantify and monitor chemotherapeutic action and cellular response are of great clinical need.
By leveraging specific molecular probes, imaging contrast and tissue properties, molecular imaging has the potential to provide quantitative data regarding the anatomical deposition of a drug, dynamic cellular response, tumor microenvironment, and the extent of treatment [
10]. Molecular imaging can be performed repeatedly with standard controls and is non-invasive and cost-effective when compared to biopsy or radiological techniques [
11]. Both in preclinical and clinical applications, molecular imaging has been used to evaluate the temporal and spatial distribution of molecular activities, including drug interactions, in biologically intact subjects. Recently developed methods enable imaging of metabolic processes, apoptosis and angiogenesis, but these provide only morphological data [
12,
13]. Quantitative measures of cytotoxicity are required to monitor cytotoxic drugs and titrate their doses to optimize the chemotherapeutic efficacy.
Direct monitoring of DNA lesions is one such approach: After DNA damage by cytotoxic agents to apurinic/pyrimidinic (AP) sites, repair occurs through DNA-repair pathways, such as the base excision repair (BER), which is the most common pathway for single-base damage [
14]. Therefore, monitoring AP sites provides a direct measure of the efficacy of a chemotherapeutic drug as well as the prognosis of the tumor cells. AP sites are currently measured in vitro using an aldehyde reactive probe in a chemoluminescent assay [
15], however no in vivo or clinical methods exist. Previous work has used positron emission tomography (PET) imaging in consort with [
11C]MX to bind and quantify AP sites [
15]. However, the clinical usage of PET for long term and repeated monitoring may be limited by cost, resource availability, and the limits of anatomical specificity [
16,
17,
18].
Recent advances have resulted in the development of numerous clinical applications and devices which leverage cellular and tissue imaging in the near infrared (NIR) range (700–900 nm) [
19,
20]. This is advantageous for in vivo applications, given the deeper penetration of NIR light into biological tissues that results from the decreased absorption and autofluorescence of biological tissues in the NIR region [
20]. NIR fluorescence imaging has proven to be a useful diagnostic technique for clinical applications ranging from angiography and tumor detection to lymphatic mapping and revascularization delineation [
21]. As compared to PET, NIR dyes can be administered repeatedly with little toxicity, are less expensive and resource-intensive, and provide high signal resolution [
22]. Having strong tissue penetration, high stability, biodegradability, and minimal nonspecific absorbance, the commercially available IR-780/Cy7.5 NIR fluorescence dye has been used as a theranostic and imaging agent for tumor cells [
23,
24]. For example, indo-cyanine green dye-loaded nanoparticles were previously reported to localize in folate-overexpressed tumors in vivo and in vitro. IR-780 has been used as highly selective tumor cell imaging and targeting agent in the NIR region [
25,
26].
Given the recent advances in molecular probe development and NIR imaging, we assembled a molecular probe system which is active in the NIR spectrum: methoxyamine-modified cyanine-7 (Cy7MX). [
14C]methoxyamine [MX] binds directly to AP sites in proportion to the extent of the DNA damage (via BER or other DNA-repair mechanisms) and can be used as a quantitative indicator for the damage induced by a chemotherapy [
23,
27]. In order to deliver the nonpolar Cy7MX probe in the blood, we need a highly stable carrier system that can carry the molecular probe effectively in the blood and deliver selectively to tumors. Among others, gold nanoparticles (AuNP) offer an ideal carrier system for loading and delivering the molecular probes, as the shape, size, and surface structure are optimal for small molecule delivery [
28]. When coated with polyethylene glycol (PEG), AuNPs provide a stable amphiphilic surface for lipophilic chemotherapeutic agents [
29]. PEGylated AuNPs have demonstrated highly efficient drug delivery into tumors by avoiding clearance via the reticuloendothelial system (RES), and deposit within minutes of injection [
30,
31]. Exploiting the enhanced permeability and retention (EPR) effect, AuNPs are able to selectively deposit cargo in tumors and remain bioinert until they are excreted [
32,
33]. We therefore employed a PEGylated AuNP carrier system to facilitate the systemic transport of the molecular probe in the blood environment and deliver the probe to tumor cells. Imaging of this system in vivo enables a real-time analysis of cytotoxicity.
In this paper, we present a molecular imaging technique to directly measure the cytotoxicity induced by chemotherapeutic agents as a function of AP sites. We have designed a Cy7MX-loaded PEGylated AuNPs-based drug-carrier platform to selectively deliver and release a unique NIR molecular probe that binds directly to AP sites of DNA damage. In vitro DLD1 colon cancer cell line experiments explore the specificity of the molecular probe by profiling cellular uptake during chemotherapy. Further, in vivo mouse models are used to determine the selective delivery and fluorescent readout of our probe. This study aims to lay the foundation for a Cy7MX-loaded PEGylated AuNP system that can provide a quantitative measure of the chemotherapeutic drug’s effect, uptake kinetics and cytotoxic chemotherapy efficacy.
3. Discussion
Since AP sites are key intermediates in the BER pathway, their quantitative and dynamic measurement in cellular DNA is crucial for the efficacy evaluation of therapeutic treatments. Currently, the assessment of AP sites can only be achieved in vitro using extracted DNA from tumor tissues. However, this is a relative measure based on chemiluminescence using an aldehyde reactive probe (ARP). ARP is an invasive assay and cannot be used to directly monitor AP-site content in tumors. The direct imaging and quantitative assessment of AP sites in vivo in real-time will provide a platform technology for efficacy evaluation of a variety of DNA-targeted chemotherapies—any that produce AP sites and invoke BER. Understanding the dynamic of AP site formation and repair will allow physicians and researchers to determine optimal dose strategies of single and combination treatment schedules. Furthermore, the direct imaging of AP sites will help to determine the optimal dose schedule to potentiate drug administration based on persistence of AP sites.
In this study, we developed a NIR-based imaging technique that can directly measure the cytotoxicity induced by chemotherapeutic agents as a function of AP sites. Using a Cy7MX-loaded PEGylated AuNPs-based drug-carrier platform, we selectively delivered and released a unique NIR molecular probe in the animal models that bound directly to AP sites of DNA damage in vivo. NIR imaging is cost-efficient, easy to operate, and has the advantage of multichannel imaging. Further, heptamethine cyanine dyes have been used extensively in NIR fluorescent imaging as contrast agents for tumor imaging. For these reasons, we developed a cyanine-based NIR probe, Cy7MX, which exhibits promising properties of binding to AP sites. Because inherent NIR fluorescence of Cy7MX can penetrate the skin, the subsequent NIR fluorescent scan in flank xerograph tumor models would thus allow the detection and quantification of AP site formation in real-time in the tumor tissues directly under the skin.
The salient objectives of this study were to (1) design and prepare a molecular probe that specifically bound to AP sites and (2) study the efficacy when delivered using a PEGylated AuNPs carrier platform tuned for delivery through blood [
28,
36]. We applied a simple, but elegant mechanism relying on the interactions of the hydrophobic molecular probe with the biological milieu: We loaded Cy7MX onto AuNPs through successful shielding of the hydrophobic Cy7MX by the amphiphilic PEG. The NPs circulated the blood stream and selectively delivered Cy7MX to tumorous tissues, when the AuNPs achieved contact with cellular membranes. This mechanism works because of the combination of the mechanical properties of the Au NPs and the polarity effects between Cy7MX, the PEG coating, and the cellular surface. Spectroscopic studies were used to characterize the CY7MX dye and AuNPs platform and study the loading and release behavior in nonpolar and polar media, which forms the basis for the probe-carrier delivery mechanism. Further, in vitro cell culture tests demonstrated targeted binding of Cy7MX to AP sites and specific localization in the nuclei. Finally, delivery to xenografted tumors in a mouse model exhibited highly targeted delivery to tumor sites and signal intensities correlating to chemotherapeutic treatment. It should be noted that once Cy7MX binds to AP sites, it prevents further repair of DNA, which leads to eventual cell death and tissue necrosis, an extra feature of the system. This study establishes Cy7MX-AuNPs to be a highly viable platform for a molecular imaging technique to quantify cytotoxic chemotherapy through fluorescence imaging in the NIR region. To date, there have been few similar combined approaches that strategically leverage the properties of nanoparticles with chemo-diagnostic molecular probes.
In its envisioned clinical implementation, an oncologist would administer a chemotherapeutic drug, followed by our molecular probe. Periodic imaging in the hours and days following treatment would enable the clinician to ascertain the efficacy and potential development of resistance to the chemotherapy. Using this data, the clinician could accordingly titrate the dosage in order to minimize toxicity and increase efficacy in a data-driven fashion. Currently, the clinical correlate is a blood test for a biomarker that is an indirect readout of the chemotherapeutic effect. Blood tests not only require invasive, repetitive draws, but also provide a readout only once a pathology lab completes the appropriate assay—which could take hours or days. In contrast, this system provides a molecular-scale real-time reporter of the success of chemotherapy in a format that is non-invasive, repetitive, non-destructive, cost-effective, and in real-time [
11].
4. Materials and Methods
4.1. Synthesis of DDA-Coated AuNPs
AuNPs were selected to be the carrier platform to load the Cy7MX probe in this study. AuNPs were coated with dodecylamine (DDA) through the Brust–Schiffrin method to provide a monodisperse solution and prevent aggregation according to a published synthesis approach [
30]. Briefly, 0.25 mM of tetraoctylammonium bromide (294136, Sigma–Aldrich, St. Louis, MO, USA) and 0.6 mM DDA (325163, Sigma–Aldrich) were dissolved in 5 mL toluene (244511, Sigma–Aldrich). Then, 0.53 mmol gold(III) chloride trihydrate; HAuCl
4·3H
2O (520918, Sigma–Aldrich) solution (30% in HCl solution) was added into the above mixture and allowed to stir for 2 h at 25 °C. Next, 0.25 mM of aqueous NaBH
4 (452904, Sigma–Aldrich) was added to the organic phase and stirred for 2 h at 45 °C. The obtained nanoparticles were precipitated in ethanol (24194, Sigma–Aldrich) and redispersed in 3 mL of CHCl
3 (C298-500, Fisher Scientific, Pittsburgh, PA, USA).
4.2. PEGylation of DDA Coated AuNPs
A layer of PEG was added to the AuNPs to confer desirable stability, shielding, and delivery benefits in biological media [
30]. Ligand exchange was performed with alpha-methoxy-omega-mercapto poly(ethylene glycol) (PEG, MW = 5000) (Meo-PEG-SH) (Laysan Bio, Arab, AL, USA) ligands. UV-visible spectroscopy (Cary Bio50, Varian spectrophotometer, Santa Clara, CA, USA) was used to determine the concentration of the AuNPs. A 1:500 ratio of AuNPs:PEG ligands was stirred in CHCl
3 for 48 h at 25 °C. This ratio was determined to be optimal for the nanoparticles as it provides maximum stability and dispersity [
28]. The CHCl
3 was then evaporated and the PEGylated AuNPs were dissolved in distilled water. Centrifugation (14,000 rpm, 5 min) was performed twice to remove unattached PEG ligands. Dynamic laser scattering (DLS) was performed to determine the size, morphology, and dispersity of the particles. Particles were analyzed in water. Transmission electron microscopy (TEM) was also performed to confirm ligand exchange and visualize the dispersion of particles.
4.3. Cy7MX Dye Loading on PEGylated AuNPs
First, Cy7MX was synthesized as previously reported in Condie et al., 2015 [
37]. Briefly, the cyanine-7 dye (425311, Sigma–Aldrich) was coupled with tert-butyl 3-(2-(4-hydroxyphenyl) acetamido)propoxycarbamate (prepared according to Salisbury et al., 2002 [
38]) in the presence of sodium hydride (223441, Sigma–Aldrich) to give the substitution product, which was subsequently deprotected with trifluoroacetic acid (302031, Sigma–Aldrich) to yield Cy7MX with a 19% overall yield. Cy7MX dye was then loaded onto the AuNPs to provide a shielding effect from hydrophilic media and an eventual release into hydrophobic media. UV-visible spectroscopy was used to determine the concentration of the PEGylated AuNPs. A 1:10 ratio of AuNPs:Cy7MX was stirred in CHCl
3 for two days at 25 °C. The fluorescence (emission at 780 nm) and absorption spectra of the sample were measured at 1, 2, 4, 8 and 24 h. Absorbance and fluorescence spectra (excitation at 780 nm) of the dye in CHCl
3 (C607SK-1, Fischer Scientific) were obtained as a control for comparison against the dye when loaded on PEGylated AuNPs. The organic phase was evaporated and the Cy7MX dye-loaded PEGylated AuNPs were redissolved in DI H
2O under sonication for 30 min.
4.4. DLS Characterization
Because the size of the AuNPs platform can greatly influence the delivery properties of the system, DLS was used to determine the hydrodynamic size of the AuNPs and ensure a normal distribution. Particles were analyzed in water. A BI-200SM laser light scattering goniometer with a BI 9000AT autocorrelator from Brookhaven Instruments Corp was used to perform DLS. Each sample was measured in triplicate for one minute each at a detection angle of 90° from a 200 μm pinhole. The dust cutoff was set to 5000. The output measurement was the lognormal mean number averaged diameter.
4.5. Loading and Stability of Cy7MX
UV-visible spectroscopy was performed to detect a monodisperse sample of AuNPs, which is indicated by an absorbance spectrum at 532 nm, whereas aggregation in the sample is indicated by a red shift above 532 nm. For each measurement, 2 mL of sample was placed in a quartz cuvette and placed in the spectrophotometer at room temperature for both types of spectroscopy. The scans were performed at the medium speed setting. The spectra of AuNPs before, during, and after loading of Cy7MX were obtained and analyzed to yield information regarding the loading behavior. Additionally, the spectra of Cy7MX in aqueous (DI H2O) and organic (CHCl3) media were obtained. A Varian Eclipse Fluorescence Spectrophotometer was used for fluorescence measurements at an excitation of 780 nm with a slit size of 2.5 mm, scanned at medium speed. The spectra of Cy7MX-AuNPs in different solvents were compared by absorbance and fluorescence to check the stability of the particles in conditions that were isotonic, acidic, and isosmotic relative to that of biological systems.
4.6. Biphasic Release Study
To gauge the time profile and release characteristics of the Cy7MX dye from the surface of the AuNPs, a biphasic release study was carried out. 1 mL of Cy7MX-AuNPs in DI H2O was placed in a cuvette. 2 mL of CHCl3 were added to the cuvette and the phase-separated system was stirred for 24 h. Absorbance of the layers was measured from 200 to 800 nm periodically for 24 h.
4.7. AuNPs Uptake
To quantify the AuNPs transport into cells, a cellular uptake study was carried out on DLD1 colon cancer cells (ATCC #CCL-221) obtained from the laboratory of Sanford Markowitz at Case Western Reserve University. DLD1 WT cells [
37] were seeded into 6-well plates at a seeding density of 0.3 × 10
6 cells/mL and grown to 70% confluence in cell culture growth media. Cells were treated with 2 mL of either 0.01 µM PEGylated AuNPs or PBS (control group) in media. An identical group was treated with 5% SDS in the media to perforate the cell membrane and serve as a negative control. At 2, 4, 12, and 24 h, the medium was removed, and the cells were washed with 2 mL of PBS three times. The cells were then harvested from the substrate with 2% typsin-EDTA. The absorbance of the lysate was measured and the wavelengths of 510–550 nm were integrated to determine the concentration. The lysates were further analyzed by graphite furnace atomic absorption spectroscopy (GFAAS) in a GTA-110 with a programmable auto-sampler (Varian, Inc., Palo Alto, CA, USA). The wavelength of the hollow cathode gold (Au) lamp was 242.8 nm and values for the concentration of Au in the samples were calibrated by a series of Au standard solutions.
4.8. Cy7MX Uptake
To identify the quantities of the Cy7MX dye in PEGylated AuNPs and the time profiles of uptake by cells, an uptake study was carried out. The cultured cells were treated with 2 mL of either 0.1 µM Cy7MX-AuNPs or 0.1 µM Cy7MX or PBS in media. At 30 min, 2, 4, and 24 h, the media was removed and the cells were washed with 2 mL of 1× PBS three times. The cells were harvested from the substrate using cell scrapers. The absorbance of the lysate was measured and integrated from 750–800 nm (Cy7MX absorbance peaks at 780 nm).
4.9. Cy7MX Localization
To ensure the localization of Cy7MX in nuclei, 20,000 DLD1 cells were plated on glass coverslips in 6-well cell culture plates. DLD1 cells were incubated in 2 mL of media for 24 h. 2 mL of 1 mM 5-florouracil (F6627, Sigma–Aldrich) diluted in media or 2 mL of DMEM media were added to two groups of cells. At 96 h, media was removed, and the cells were washed with 1× PBS twice. Cells were then fixed with 2 mL of methanol at −20 °C for 10 min. Uracil DNA glycosylase (UDG) treatments in UDG reaction buffer—either 100 or 250 units—were added to each sample and incubated overnight. Cells were then washed with 1× PBS and incubated with a 25 µM Cy7MX probe for one hour while stirring. The samples were then rinsed with 1x PBS, stained with 4′,6-diamidino-2-phenylindole (DAPI), and soft mounted onto slides. Cells were imaged with a Leica STP 600 confocal microscope under 80x magnification. Excitation was performed at 405 nm and 633 nm with detectors measuring DAPI from 410–480 nm (32% gain) and Cy7MX from 650–680 nm (41% gain). Images were analyzed with ImageJ.
4.10. In Vivo Testing of Cy7MX Uptake
In vivo imaging experiments on colon cancer xenografted NCR nude mice were carried out at Case Western Reserve University in the Case Center for Imaging Research in Cleveland, Ohio under the IUCAC protocol (ID# 2014-0067) approved by the Case Western Reserve Institutional Animal Care and Use Committee Office on 16 March 2016. Twenty 8-week old NRC nude mice (6–8 weeks of age) were fed a Teklad 2018 S special alfalfa-free diet for two weeks to reduce auto fluorescence. DLD1 colon cancer cells and T98G glioma cells were cultured in cell growth media and were used at passage 12 and 8, respectively. At 90% confluence, cells were harvested, and each mouse was injected with 3 million cells per tumor site—glioma on the right flank and colon cancer on the left. After two weeks, the tumors had grown to approximately a centimeter in width. Half of the mice were treated with the chemotherapeutic drug 5-Fluorodeoxyuridine (5-Fdu) at a dosage of 0.2 mg of drug per gram of mouse every three days. PBS (806544, Sigma–Aldrich, St. Louis, MO, USA) was administered to the other half as placebo. After 3 doses, the mice were injected intravenously with either 100 µM Cy7MX or 100 µM Cy7MX–10 µM AuNPs. The Cy7MX concentration was 5 nmol per 25 g of mouse. Chemotherapy continued during the duration of imaging and the mice were kept in a normal ambient light environment.
Using a Maestro in vivo fluorescence imaging system (Cambridge Research and Instrumentation, Inc., Woburn, MA, USA), black and white photographs as well as fluorescence images were obtained before injection, at injection, and 1, 24, 48 h, and 1 week after injection. Deep red (750–950 nm) excitation and emission filters were used. Images were obtained with 3 s exposure. The multispectral images were separated into their component spectra (Cy7MX, auto fluorescence, and background). After removing the background and auto fluorescence, the signal intensity in tumor regions was quantified. The tumor regions were identified to be regions of interest from the black and white images and overlaid onto the fluorescence images to quantify signal.
4.11. Biodistribution
After one week, the mice were sacrificed and organs including the heart, liver, brain, tumors, and kidneys were isolated and harvested. All tissues were digested with 70% nitric acid (HNO3) (438073, Sigma–Aldrich) at 35 °C for 48 h. Digested samples were diluted in 2.5 mL of HNO3 and analyzed with a GFAAS in a GTA-110 with a programmable auto-sampler (Varian, Inc., Palo Alto, CA, USA). The cathode Au lamp had a wavelength of 242.8 nm. A set of standards was used to create a calibration curve of known values. Using this calibration as a guideline, the concentrations of Au in the organ samples were found.