Photothermally Enhanced Drug Delivery by Ultra-Small Multifunctional FeCo/Graphitic-Shell Nanocrystals

Abstract

FeCo/graphitic carbon shell (FeCo/GC) nanocrystals (~4-5 nm in diameter) with ultra-high magnetization are synthesized, functionalized and developed into multifunctional biocompatible materials. We demonstrate the ability of this material to serve as an integrated system for combined drug delivery, near-infrared (NIR) photothermal therapy and magnetic resonance imaging (MRI) in-vitro. We show highly efficient loading of doxorubicin (DOX) by π-stacking on the graphitic shell to afford FeCo/GC-DOX complexes and pH sensitive DOX release from the particles. We observe enhanced intracellular drug delivery by FeCo/GC-DOX under 20 minutes of near-infrared (NIR) laser (808 nm) induced hyperthermia to 43°C, resulting in a significant increase of FeCo/GC-DOX toxicity towards breast cancer cells. The synergistic cancer cell killing by FeCo/GC-DOX drug delivery under photothermal heating is due to a ~2-fold enhancement of cancer cell uptake of FeCo/GC-DOX complex and the increased DOX toxicity under the 43°C hyperthermic condition. The combination of synergistic NIR photothermally enhanced drug delivery and MRI with the FeCo/GC nanocrystals could lead to a powerful multimodal system for biomedical detection and therapy.

The use of nanomaterials in medicine has offered the capability of combining multiple functionalities into a single system. Materials such as inorganic nanoparticles, carbon nanotubes, dendrimers, and liposomes have all been investigated to combine multiple therapeutic approaches such as imaging, drug delivery, and photothermal therapy.^1-11 Of particular interest with this class of materials is the ability to gain control over the toxicity of the therapy through controlled release or enhanced therapeutic efficacy.

One means of enhancing therapeutic efficacy has been through combining chemotherapy drugs with hyperthermia. Previous studies have heated regions of the body to target temperatures up to ~43°C to enhance efficacy of drugs such as doxorubicin (DOX), an anthracycline antibiotic used to treat a wide variety of cancer.^{3, 12-18} A drawback of clinical hyperthermia methods, such as limb perfusion, radiowave irradiation or incubation chambers, is that non-specific heating can result in enhanced drug effects throughout the general anatomical region, resulting in undesired side effects to normal organs and tissues.^{12, 13} In an effort to improve the safety of thermal cancer therapies, near-infrared (NIR) absorbing photothermal agents have been explored in combination with laser irradiation to provide site-specific heating and treatment of diseased regions with a high uptake of the photothermal agents.^{3, 4, 15-17}

There has been a recent interest in developing cooperative nanoscale systems to enhance doxorubicin toxicity via NIR irradiation.^{3, 4, 15-17} Synergy between DOX and NIR irradiation has been demonstrated through the addition of NIR photothermal agents, such as indocyanine green or gold nanomaterials, to cells subjected to DOX or DOX liposome treatment.^14-16 In these systems, the drug and photothermal sensitizer, or heat inducing material, are separate agents. Systems capable of both drug delivery and photothermal therapy have been developed by creating gold half-shells around DOX-loaded polymer spheres.^{3, 4, 17} While these approaches have shown favorable results, these particles tend to be large (>70 nm) in size and lack the ability to monitor the particle biodistribution via a clinically relevant imaging modality. Addition of a magnetic layer under the gold half-shell has imparted MRI capabilities, however, the size of the particles are undesirably large (~100 nm) and therapeutic efficacy has yet to be determined.¹⁷

In this work, we create a highly integrated nanoparticle system uniquely capable of drug delivery, imaging and photothermal therapy. To achieve this, we use an ultra-small ~4 nm magnetic nanocrystal containing a single crystalline iron-cobalt core surrounded by a single- or few-layer graphitic shell (FeCo/GC).¹⁹ The graphitic carbon shell is capable of loading DOX via π-stacking and exhibits a useful absorbance of NIR light affording photothermal effects. The highly magnetic metal core makes FeCo/GC an advanced MRI contrast agent, as FeCo is known to have one of the highest saturation magnetizations among all materials.^{19, 20} Therefore, we are developing ultra-small nanocrystals into a highly multifunctional agent, capable of both drug delivery and photothermal therapy with the additional benefit of tracking the nanoparticle system with a clinically relevant imaging modality.

We show environmentally sensitive loading and release of DOX from the FeCo/GC system, demonstrating faster drug release in acidic conditions that mimic the extracellular tumor environment. The highly controllable loading and release of DOX are attributed to the supramolecular loading of DOX on the graphitic shell of FeCo/GC.^{21, 22} When used to deliver DOX to MCF-7 cells, a human breast cancer line, FeCo/GC-DOX is slightly less toxic than free DOX, however, when combined with 20 minutes of NIR photothermal heating to 43°C a drastic increase in toxicity is observed for cancer cells treated with FeCo/GC-DOX. Using MRI, fluorescence measurements and flow cytometry, we reveal a ~2-fold increase in cellular uptake of FeCo/GC-DOX when incubated for 20 minutes at 43°C over 37°C. The significant increase in cellular uptake of FeCo/GC-DOX under photothermal laser irradiation combined with an increase in DOX efficacy at elevated temperatures could make FeCo/GC-DOX a powerful system for photothermally enhanced drug delivery and MRI guided cancer therapy.

Results and Discussion

To obtain highly biocompatible and water soluble FeCo/GC, nanocrystals synthesized by a chemical vapor deposition method¹⁹ were sonicated with a phospholipid-branched-polyethylene glycol carboxylate (PL-brPEG) and centrifuged to removed aggregates (see Methods).²⁰ The nanocrystals, as depicted in Figure 1a, were well dispersed and stable in buffer solutions and serum,¹⁹ with an average diameter of ~4 nm revealed by transmission electron microscopy TEM (Fig. 1b). The nanocrystals are known to have superior chemical stability in acid and air, and show no evidence of oxidation or degradation over a monitoring period of 1 month.¹⁹ While the shell does show some disordered nature, as evidenced by both a Raman G band and D band, the stability of the nanocrystal in acid demonstrates the non-porous nature of the carbon shell.¹⁹ Doxorubicin was loaded non-covalently onto water soluble FeCo/GC by simply mixing solutions under controlled pH and DOX concentrations for ~14 hours. After removal of free DOX by centrifuge filtration, evidence of DOX loading on FeCo/GC was visible by eye (Fig. 1c inset) due to the deep red appearance of the FeCo/GC-DOX suspension. DOX content was quantified using UV-vis absorbance, with a prominent DOX peak appearing around 490 nm over the FeCo/GC background (Fig. 1c). Solutions of FeCo/GC-DOX were stable in water and PBS buffer solutions over multiple weeks without significant aggregation.

Figure 1. Structure of FeCo/GC-DOX, drug loading and release.

(a) Schematic of DOX π-stacking on FeCo/GC. The FeCo core (shown in green) is surrounded by a single layer of graphite. The nanocrystal is made water soluble by a non-covalent interaction with phospholipid-branched-PEG. Doxorubicin, shown in white, loads non-covalently on the graphitic surface of FeCo/GC. (b) TEM images of FeCo/GC-DOX conjugates. The average diameter is ~4 nm for the FeCo/GC-DOX complex. A high-resolution TEM image of a single FeCo/GC-DOX nanocrystal is shown as the inset. (c) UV-visible absorbance spectra of free DOX, FeCo/GC-DOX (FeCo-DOX) or FeCo/GC (FeCo). The DOX concentration is 450 μM for free DOX and FeCo/GC-DOX samples. The FeCo/GC concentration is ~270 nM for both the FeCo/GC and FeCo/GC-DOX samples. Both free DOX and FeCo/GC-DOX show a characteristic peak around 490 nm. Suspensions of DOX, FeCo/GC-DOX and FeCo/GC are shown in the inset. (d) Loading of DOX on FeCo/GC shows a strong dependence on pH and the DOX concentration during the incubation. The z-axis is the number of DOX molecules loaded on a single FeCo/GC nanocrystal (see methods). (e) The release of DOX from the surface of FeCo/GC is accelerated at lower pH. (f) Concentration-dependent cell survival curves of MCF-7 cells incubated with free DOX or FeCo/GC-DOX for two days. IC₅₀ values of 1.2 μM and 2.9 μM were calculated for free DOX and FeCo/GC-DOX respectively.

The degree of loading of DOX onto FeCo/GC showed a strong dependence on both pH and DOX concentration during the DOX loading incubation. By altering the loading conditions, we were able to load between ~200 and ~2000 DOX molecules onto each FeCo/GC nanocrystal (see Methods) as shown in Figure 1d. The highest DOX loadings were achieved at basic conditions where the daunosamine group of DOX was deprotonated and the solubility in water reduced. Conversely, at lower pH conditions, DOX solubility was increased and loading on FeCo/GC is minimized. This suggested hydrophobic interactions and π-stacking as the driving force for DOX loading on FeCo/GC. The aromatic character of DOX enabled π-stacking onto the graphitic sidewall of FeCo/GC in a similar manner to DOX loading on carbon nanotubes and graphene.^21-23 The π-stacking interaction was confirmed by fluorescence quenching of DOX when loaded onto FeCo/GC (see Supporting Information Figure S1).

The release of DOX from the FeCo/GC surface showed a slow, continuous process over the course of multiple days (Fig. 1e). This release was accelerated at acidic conditions as the daunosamine group of DOX was protonated and water solubility increases.²⁴ Samples of FeCo/GC-DOX conjugates dialyzed over 8 days showed that the release of DOX from FeCo/GC is ~3 times greater at pH 5.5 than pH 7.4. This activated release in acidic conditions makes our material appealing for drug delivery applications as both the extracellular environment of tumors and the intracellular lysosomes and endosomes are acidic.

The 4 nm nanocrystals showed a slower DOX release profile than single-walled carbon nanotubes (~1 nm diameter) possibly due to a stronger interaction between the DOX and the graphitic sidewall of FeCo/GC.²¹ The larger diameter of FeCo/GC has less curvature and may provide a more planar surface for DOX loading, leading to higher drug loading and slower release. The diameter dependent release was a driving force to use 4 nm nanocrystals over the 7 nm nanocrystals used in previous work. ^{19, 20}

Initial tests on the toxicity of FeCo/GC-DOX conjugates were performed on the human breast adenocarcinoma cell line MCF-7. FeCo/GC-DOX and free DOX were incubated with cells for 2 days prior to viability measurement. Both FeCo/GC-DOX and free DOX induced significant cell death above 10 μM, with less than 20% viable cells after 2 days (Fig. 1f). At lower concentrations, free DOX proved to be slightly more toxic than FeCo/GC-bound DOX with IC₅₀ values of 1.2 μM (0.7 μg DOX/mL) and 2.9 μM (1.7 μg DOX/mL) respectively (Fig. 1f). The decreased toxicity of FeCo/GC-DOX could be attributed to the slow, continual release of DOX from the FeCo/GC surface over the two day incubation period. This resulted in lower concentrations of released DOX available for cell killing during the incubation period. It should be noted that the toxicity was only due to DOX as FeCo/GC without drug attached showed no significant toxicity to MCF-7 cells (see Supporting Information Figure S2).

Graphitic materials such as carbon nanotubes,^25,10 graphite-oxide²⁶ and larger diameter FeCo/GC¹⁹ have all been shown to generate heat under NIR irradiation. The optical absorbance of the graphitic shell of FeCo/GC-DOX conjugates at ~800 nm makes them an ideal candidate for use with biological systems due to the minimal absorbance of cellular and tissue components at this wavelength.²⁷ The heating capabilities of FeCo/GC, FeCo/GC-DOX, and DOX were investigated during 808 nm laser irradiation. When samples of 450 μM DOX (260 μg DOX/mL) and FeCo/GC suspensions with or without 450 μM loaded DOX were irradiated with an 808 nm laser at 0.3 W/cm², the FeCo/GC containing samples showed selective graphitic carbon shell-mediated heating over the free DOX solution (Fig. 2a&b). Thermal imaging with an infrared camera confirmed a temperature increase of ~35°C for solutions containing FeCo/GC during the 10 minutes of laser irradiation (Fig. 2c). The ability for efficient photothermal heating of the FeCo/GC-DOX system compared to free drug suggested the multifunctional therapeutic potential of this ultra-small, highly integrated system for combined MRI, drug delivery and NIR photothermal therapy.

Figure 2. Photothermal heating of DOX and FeCo/GC Suspensions.

(a) Image of solutions containing 450 μM DOX either as free DOX or FeCo/GC-DOX (FeCo-DOX). The FeCo/GC (FeCo) solution shown has the same FeCo/GC content as the FeCo/GC-DOX solution. (b) Thermal image after 10 minutes of 808 nm laser irradiation at 0.3 W/cm². Both the FeCo/GC-DOX and FeCo/GC samples show significant photothermal heating, while the free DOX solution is not heated by the laser. (c) Temperature measurements from thermal images acquired during the 10 minute laser irradiation show a temperature increase of ~35°C for FeCo/GC containing samples. The free DOX sample showed almost no heating effect from laser irradiation.

To test the photothermal effect of FeCo/GC-DOX in drug delivery, we compared 808 nm laser irradiated versus non-irradiated MCF-7 cells mixed with either FeCo/GC-DOX, free DOX, FeCo/GC or a PBS control. For the laser irradiated samples, a target temperature of 43°C for FeCo/GC and FeCo/GC-DOX samples was maintained for 20 minutes. Cells incubated with either DOX or PBS were irradiated concurrently, although no significant heating was observed. After the 20 minute irradiation/incubation period, cells were washed to fully remove the drug and allowed to proliferate for 2 days before viability assessment.

Cell viability measurements demonstrated a significant increase in toxicity for FeCo/GC-DOX when combined with 808 nm photothermal treatment (Fig. 3a&b). For non-irradiated MCF-7 cells, FeCo/GC-DOX inhibited cell growth by 27-28% at concentrations of 30 or 60 μg DOX/mL, respectively (Fig. 3a). Inclusion of laser irradiation for 20 minutes afforded a photothermal heating of the solution to 43°C during the FeCo/GC-DOX incubation period, resulting in cell growth inhibition of 64% and 69% for the 30 and 60 μg DOX/mL cases respectively (Fig. 3a&b). The viability of all other treatment groups (free DOX, FeCo/GC or PBS treated cells) remained similar to their non-irradiated controls (Fig. 3a). Addition of photothermal therapy more than doubled the toxicity of FeCo/GC-DOX, causing more cell death than DOX in its free form with or without 808 nm laser irradiation. Importantly, there was no significant reduction in viability for cells incubated with FeCo/GC with or without photothermal heating to 43°C. This indicated that the FeCo/GC nanocrystals were non-toxic, and photothermal heating of cells to 43°C for 20 minutes alone was not sufficient to induce cell death. Additionally, cell viability was not affected in samples subjected to laser irradiation in the absence of a photothermal agent, such as in the PBS and free DOX incubated samples. The target temperature of 43° C was chosen to enhance DOX toxicity and minimize damage to cells not exposed to the drug.

Figure 3. Cellular toxicity enhancement of DOX conjugates when combined with photothermal heating to 43°C.

(a) Cell viability of MCF-7 cells after incubation with free DOX, FeCo/GC-DOX (FeCo-DOX), FeCo/GC (FeCo). DOX concentrations were either 30 or 60 μg DOX/mL. Samples were incubated for 20 minutes and combined with heating to 43°C through photothermal laser heating (red bars) or thermal heating in a water bath (blue bars). Non-heated samples are shown in black. FeCo/GC, free DOX or untreated control cells showed no viability dependence on the incubation conditions (heated vs. non-heated). FeCo/GC-DOX efficacy was significantly enhanced when incubations were performed at 43°C. The viability of control cells or FeCo/GC (NP) incubated cells were not affected by laser irradiation or thermal heating to 43°C for 20 minutes. (b) MCF-7 viability plots demonstrate the large increase in toxicity resulting from FeCo/GC-DOX (FeCo-DOX) at 43°C over 37°C, while cells incubated with FeCo/GC (FeCo) are not affected by thermal heating to 43°C for 20 minutes. (c) Cell toxicity of DOX at 37°C and 43°C reveals thermal enhancement of DOX toxicity at low doses of DOX (2 μg/mL). DOX efficacy of highly toxic doses (30 μg/mL) is not enhanced at 43°C due to the sufficiently high efficacy of high-dose DOX at 37°C.

To compare with NIR photothermal heating, cells were incubated with FeCo/GC, PBS, FeCo/GC-DOX or free DOX at 30 μg DOX/mL and heated thermally in a water bath to 43°C for 20 minutes. Cells were then washed to remove the drug and allowed to proliferate for 2 days. Cell viability proved to be similar to the laser irradiated case. Compared to cell samples incubated at 37°C, thermal heating of cells to 43°C for 20 minutes afforded much enhanced toxicity resulting from FeCo/GC-DOX incubation but not for cells incubated with DOX (at a high 30 μg/mL concentration), FeCo/GC or PBS (Fig. 3a&b). However, at a lower dose of free DOX (2 μg/mL), obvious enhancement of DOX toxicity was observed when the cellular incubation was carried out at 43°C vs. 37°C incubation (Fig. 3c).

To understand the cell toxicity results we investigated the effects of hyperthermia on MCF-7 cellular uptake of FeCo/GC-DOX. Cellular uptake of FeCo/GC-DOX at 37°C versus 43°C was assessed by using MRI. The high relaxivities of FeCo/GC nanocrystals (r₁= 31 mM^-1s^-1 and r₂= 185 mM^-1s^-1 at 1.5 T), attributed to the FeCo alloy core, makes this material a highly sensitive positive (T₁) and negative (T₂) contrast agent.¹⁹ In a 1 T field, the cells incubated with FeCo/GC-DOX at 43°C appeared much brighter than either the untreated cells, or the cells incubated with FeCo/GC-DOX at 37°C (Fig.4a). The increase in MRI signal was quantified by region of interest (ROI) measurements of the cell pellets in Figure 4b. The cells incubated at 43°C had a higher mean intensity than cells incubated at 37°C or non- treated cells, due to T₁ contrast enhancement by the magnetic nanocrystals. The results indicate higher cellular uptake of FeCo/GC-DOX at 43°C compared to 37°C.

Figure 4. Enhanced Cellular Uptake of FeCo/GC-DOX at 43°C.

(a) Magnetic resonance image of pelletted, untreated MCF-7 cells or pelletted MCF-7 cells treated with FeCo/GC-DOX (FeCo-DOX) for 20 minutes at 43°C and 37°C. Brightening of samples in the T₁-weighted images indicates FeCo/GC-DOX uptake. Cells incubated at 43°C with FeCo/GC-DOX appeared the brightest due to higher cellular uptake of FeCo/GC-DOX. (b) Region-of-interest (ROI) signal measurements of the cell pellets in part (a). MCF-7 cells treated with FeCo/GC-DOX had the highest signal intensity due to the higher uptake during the 20 minute incubation. (c) DOX fluorescence signal after being extracted from cells incubated with FeCo/GC-DOX (FeCo-DOX) for 20 minutes at 37°C or 43°C. (d) Flow cytometry measurements confirming higher DOX fluorescence in cells incubated with FeCo/GC-DOX at 43°C than at 37°C. Untreated MCF-7 cell signal is shown for comparison. The mean signal of cells treated with FeCo/GC-DOX for 20 minutes at 37°C versus 43°C is plotted in the inset. The ratio of the 43°C treatment versus the 37°C treatment is similar to the signal ratio in part (c).

Fluorescence and flow cytometry measurements of DOX fluorescence in cells that were incubated with FeCo/GC-DOX at 37°C and 43°C confirmed the MRI results. In Figure 4c, the DOX fluorescence intensity from lysed cells following a 20 minute incubation with FeCo/GC-DOX at 43°C was almost double that of the fluorescence intensity measured from cell lysate following incubation at 37°C for 20 minutes. DOX fluorescence from cells measured by flow cytometry confirmed the increased cellular uptake of FeCo/GC-DOX at hyperthermic conditions (Fig. 4d). While all cells incubated with FeCo/GC-DOX showed increased fluorescence over MCF-7 cells without FeCo/GC-DOX, the fluorescence of cells incubated at 43°C for 20 minutes was 1.9 times greater than that of cells incubated at 37°C for 20 minutes. These results confirm the ~2-fold enhanced cellular uptake of FeCo/GC-DOX at 43°C versus 37°C. The agreement in MRI and fluorescence results suggest that the hyperthermic incubation enhanced the uptake of the complete FeCo/GC-DOX complex, rather than increasing the uptake of only a single component. In addition, flow cytometry measurements did not show a significant increase in free DOX uptake when incubated at elevated temperatures (see Supporting Information Figure S3).

It has long been established that endocytosis is an energy-dependent process showing a dependence on temperature and adenosine triphosphate (ATP) levels.^28-30 Cellular endocytosis proceeds at 37°C, yet is severely hindered at lower temperatures (e.g. 4°C).^{28, 29} Previous studies have indicated clathrin-dependent endocytosis as the mechanism of carbon nanotube uptake into mammalian cells.^{30, 31} It is expected that FeCo/GC is endocytosed in a similar manner to carbon nanotubes due to similar graphitic surface with phospholipid coating, with cellular uptake dependant upon temperature, thereby leading to greater uptake of FeCo/GC-DOX conjugates at 43°C over 37°C. Notably, 20 minutes at 43°C is well tolerated by the MCF-7 cells as we observed in this work.

While the photothermally enhanced cellular uptake of FeCo/GC-DOX mostly explains the enhanced cellular toxicity observed herein (Fig. 3a), many previous reports suggest enhanced toxicity of DOX at temperatures around 43°C.^{14, 18, 32, 33} To investigate if this was a potential mechanism for enhancement of FeCo/GC-DOX toxicity under hyperthermic conditions we tested the toxicity of free DOX on cells when incubated for 20 minutes at 37°C or 43°C. At high free DOX concentrations, e.g. 30 μg DOX/mL, the drug is very effective at inhibiting cell growth. At these high doses, cell toxicity appears to be independent of temperature during the 20 minute incubation period. This is in good agreement with previous observations that drug doses most effective at normal temperatures do not necessarily demonstrate higher efficacy at elevated temperatures.^{16, 33} At a much lower dose of 2 μg/mL free DOX, however, there is a significant enhancement of toxicity when the incubation is done at 43°C as seen in Figure 3c. This indicates that heating to 43°C increases the toxicity of DOX when the drug alone is not capable of killing the majority of cells. For cells incubated for 20 minutes at 37°C, the lower dose of free DOX (2 μg/mL) more closely resembles the FeCo/GC-DOX incubations in Figure 3a&b, where cell growth is inhibited by less than 30%. At moderately toxic concentrations of either free DOX or FeCo/GC-DOX, exposure of cells to 43°C for 20 minutes could enhance the cellular toxicity of DOX delivered via FeCo/GC. Therefore, the photothermally enhanced toxicity of FeCo/GC-DOX is likely due to the synergistic effects of increased cellular uptake of FeCo/GC-DOX at 43°C and an increase in DOX toxicity at elevated temperatures.

The photothermal enhancement of FeCo/GC-DOX toxicity, synergistically improved by drug delivery through increased cellular uptake and enhanced drug toxicity, suggest that photothermally enhanced delivery of DOX via highly magnetic FeCo/GC under NIR irradiation could be a powerful technique for cancer therapy and imaging. The NIR laser irradiation conditions used in this work afford selective heat generation only in the presence of FeCo/GC, without which no heating of cells or tissues was observed.

While the use of FeCo/GC-DOX has only been evaluated in vitro this system could readily be translated to an in vivo treatment method. Previous studies using DOX-loaded single-walled carbon nanotubes (SWNTs) for in vivo cancer treatment resulted in higher DOX levels in tumors due to high SWNT accumulation via the enhanced permeability and retention effect. ²² In addition, DOX delivery via SWNTs reduced systemic drug-related toxicity and resulted in longer survival times of SWNT-DOX treated mice.²² It is predicted that FeCo/GC-DOX may accumulate in tumors with similar or even higher efficiency than SWNTs due to their ultra-small size. Uptake of FeCo/GC-DOX could be further increased through the addition of a tumor targeting ligand on the terminus of the PEG chain. Accumulation of FeCo/GC-DOX in a tumor would enable selective photothermal heating of the tumor region by 808 nm laser irradiation. Photothermally enhanced drug delivery by FeCo/GC and NIR laser irradiation, in combination with enhanced drug efficacy at 43°C, could lead to a dramatic improvement in chemotherapeutic response in vivo. The enhanced susceptibility of malignant cells, over non-malignant normal cells, to temperatures between 41°C and 43°C could further enhance the efficacy of this treatment while minimizing toxicity to healthy cells.³⁴ In addition, the reduced toxicity of FeCo/GC-DOX over free DOX at 37°C could lower the systemic toxicity associated with chemotherapy.²² Furthermore, MRI contrast enhancement from FeCo/GC could be used as a tool to guide therapy and monitor the tumor response. These features warrant future investigations into the use of FeCo/GC-DOX as an integrated nanoparticle system for combined cancer imaging, drug delivery and photothermal therapy.

Conclusion

In this work we developed an ultra-small (~4 nm core size) magnetic nanocrystal system for combined drug delivery, imaging and NIR-induced hyperthermia. In addition to unique supramolecular drug loading and desirable pH sensitive release, we are able to take advantage of the photothermal heating of FeCo/GC to significantly enhance the efficacy of the FeCo/GC-DOX system when heated to 43°C for 20 minutes. Investigations into the cause of the enhanced efficacy of FeCo/GC-DOX when combined with 20 minutes of photothermal therapy revealed a 2-fold increase in cellular uptake of FeCo/GC-DOX at 43°C over 37°C. This increased cellular uptake, combined with increased DOX toxicity at 43°C at the relevant dosage, explains the significant enhancement of FeCo/GC-DOX when combined with 808 nm. laser irradiation. Laser irradiation did not affect cells incubated with free DOX, FeCo/GC or with PBS control cells. In addition, exposure of cells to 43°C for 20 minutes without drug present did not significantly affect the cell viability. Photothermally enhanced drug delivery enabled by the highly integrated FeCo/GC-DOX system could offer a new method of highly effective drug delivery and cancer imaging, while reducing systemic drug related toxicity.

Methods

FeCo/GC Synthesis

FeCo/GC nanocrystals were prepared as previously described.^{19, 20} Briefly, iron nitrate and cobalt nitrate were dissolved in methanol and combined with silica powder to make a growth catalyst. Nanocrystals were formed using methane chemical vapor deposition (CVD) for 5 minutes at 800°C. Silica was removed by HF etching, followed by washing with 10% ethanol and finally pure water. Samples were solubilized by sonicating with 1.8 mg/mL phospholipid-branched polyethylene glycol-carboxylate (PL-brPEG) for 1 hour, followed by centrifugation at 15,000 rpm for 6 hours to remove any aggregates. PL-brPEG was synthesized as previously described.²⁰

Drug Loading on FeCo/GC

Prior to drug loading, excess PL-brPEG was removed by washing twice through a 100 kDa centrifuge filter (Millipore). DOX loading was achieved by simple mixing of aqueous DOX solution and ~ 600 nM FeCo/GC. Under standard loading conditions, the loading was done in 1x PBS and 1.7 mM DOX, with the pH adjusted to 8.5- 9 using 0.5 M carbonate buffer. For DOX loading tests the DOX concentration was varied from 0.5-2 mM during loading and the pH was adjusted from pH 5-9 using dilute HCl or carbonate buffer. Samples were incubated for ~14 hours at room temperature with gentle agitation. After the incubation period, DOX that was not loaded onto the surface of FeCo/GC was removed via filtration through a 100 kDa centrifuge filter. Samples were washed with PBS then water a minimum of 6 times until no DOX was detected in the filtrate. To remove any aggregates, samples were centrifuged for 5 minutes at 10,000 g.

The concentration of DOX or FeCo/GC was determined by UV-vis using a DOX extinction coefficient of 10,500 M^-1 cm^-1 at 490 nm. All UV-vis was done using a Cary 300 spectrophotometer (Varian). The nanocrystal concentration of FeCo/GC was determined by measuring the absorbance at 808 nm, using an extinction coefficient of 3.5*10⁶ M^-1cm^-1. The extinction coefficient was calculated by relating the graphite absorption at 808 nm to the metal content of a nanocrystal solution and the calculated metal content of a single 4 nm diameter nanocrystal.¹⁹ When both DOX and FeCo/GC were present, the DOX absorbance was extracted by subtracting the background FeCo/GC signal normalized at 808 nm. Concentrations of FeCo/GC-DOX used during experiments are reported as DOX concentration in solution.

DOX Release Tests

FeCo/GC-DOX samples were filtered thoroughly and placed in 3500 MWCO dialysis membranes (Fisher) and placed in PBS at either pH 7.4 or pH 5.5 after adjusting with HCl. Three separate bags were dialyzed at each condition with aliquots taken at different intervals for determination of DOX loading.

NIR Laser Photothermal Heating

Samples were irradiated using an 808 nm fiber coupled laser (RMPC Lasers). For solution heating, samples containing 450 μM DOX (260 μg DOX/mL) or equivalent FeCo/GC concentration were irradiated with the laser at a power density of 0.3 W/cm² for 10 minutes. The temperature of all samples was monitored using a thermal camera (Mikron Infrared, Oakland, NJ) and periodically confirmed using a thermocouple. The heating curve was determined by measuring the images and plotting the average temperature ± the standard deviation.

Cell Culture and Viability Tests

MCF-7 cells were cultured using Dulbecco’s Modified Eagle Medium containing 4.5 g/L D-Glucose, 110 mg/L Sodium Pyruvate, and L-Glutamine supplemented with 10% fetal bovine serum and 1% penicillin and streptomycin. Cells were cultured at 37°C in a humidified atmosphere containing 5% CO₂. For IC₅₀ measurements, MCF-7 cells were plated into 96 well plates and allowed to adhere prior to addition of varying concentrations of DOX or FeCo/GC-DOX in PBS. Samples were incubated for 2 days before washing the wells to remove the drugs and replacing the medium. Viability was measured by adding CellTiter 96 (Promega) into the medium. The color change was quantified by measuring the absorbance at 490 nm using a plate reader. Viability is plotted as the fraction of the non-treated control wells.

For laser irradiated and thermally heated cell viability tests, MCF-7 cell were allowed to adhere to a 96 well plate before adding in drug or FeCo/GC samples. Samples were incubated with cells during the 20 minute laser irradiation/heating period. During the laser irradiation period temperatures were closely monitored using a thermal camera and thermoprobe and the laser power was adjusted to maintain the temperature of FeCo/GC containing wells at ~43°C. There was little to no non-specific heating of the DOX or PBS control wells during this laser irradiation. For thermal heating, 96 wells plates were heated in a water bath connected to a temperature controller allowing for the temperature to be regulated at either 37°C or 43°C.

Following the 20 minutes drug incubation, cells were washed three times, replaced with fresh medium and allowed to grow for 2 days before replacing the medium once more and measuring the viability as discussed above. All samples are plotted as the fraction of non-irradiated/non-heated cells mixed with PBS.

MRI Determination of FeCo/GC-DOX Cell Uptake

For cellular uptake tests, vials of MCF-7 cells were incubated with 100 μM FeCo/GC-DOX for 20 minutes in a thermally regulated water bath. Following the incubation cells were washed thoroughly and put in fresh vials mounted in an agar phantom. MRI was done using a 1 T small animal scanner (M2, Aspect, Isreal) using a steady state gradient recalled echo sequence (T_r = 9 ms, T_e= 3 ms). Lightening of the cell pellets was indicative of FeCo/GC-DOX uptake and was confirmed by taking ROI measurements of the pellet regions using ImageJ software (NIH). Error bars represent the average of five measurements between two vials incubated at each condition.

Fluorescence Determination of FeCo/GC-DOX Cell Uptake

MCF-7 cells were plated at equal cell densities into 24 wells plates and allowed to adhere. After 1 day cell medium was replaced with Hank’s Balanced Saline Solution (HBSS) containing FeCo/GC-DOX at a DOX concentration of 30 μg/mL. Cells were incubated for 20 minutes in a water bath maintained at either 37°C or 43°C. Following the incubation cells were washed with cold HBSS and then lysed using 0.75% Triton-X 100 in water. DOX was removed from the surface of FeCo/GC by adding acidified isopropanol (0.3 N HCl) and extracting at -20°C overnight. Equal volumes of extracts and blanks were added to a multi-well plate and DOX fluorescence was measured using a plate reader (ex. 485 nm, em. 580 nm). Background subtracted DOX signal was plotted in Figure 4c. The error bars represent the standard deviation of 4 samples at each temperature. Flow cytometry samples were prepared by incubating cells with 575 nM FeCo/GC-DOX or PBS for 20 minutes at 37°C or 43°C. Cells were washed thoroughly and measurements were performed on a LSR II flow cytometer. DOX fluorescence from gated, singlet MCF-7 cells was monitored using the phycoerythrin (PE) channel (532 nm excitation, 575/25 nm emission bandpass filter).