Hairpin Oligonucleotide Can Functionalize Gold Nanorods for in Vivo Application Delivering Cytotoxic Nucleotides and Curcumin: A Comprehensive Study in Combination with Near-Infrared Laser.

We prepared a multimodality nanocomplex by functionalizing gold nanorods (GNRs) with a cytotoxic nucleoside, 5-fluoro-2'-deoxyuridine (FdU) containing a DNA hairpin, along with complexation of pleiotropic molecule curcumin. Conjugates were investigated for anti-tumor activity using an Ehrlich carcinoma model in combination with 808 nm laser irradiation. We demonstrated that hairpin-functionalized GNRs are suitable for intravenous administration, including delivery of cytotoxic nucleotides and curcumin. Curcumin binding with FdU-hairpin-functionalized GNRs displayed improved anti-tumor activity in part by inducing a lymphocyte-mediated immune response. The complex showed notable photothermal activity in vitro; however, 808 nm laser irradiation of the tumor following treatment with the complex did not increase the anti-tumor effect significantly. Biodistribution studies depicted that the nanoconjugates localized primarily in the sinusoidal structures of the liver and spleen with minimal tumor accumulation. Curcumin complexation alleviated the reduction in the RBC count that was observed for the conjugate without curcumin, especially in combination with laser irradiation. Localization of FdU-hairpin-GNR conjugates in the liver and spleen evoked an inflammatory response, which was mitigated by curcumin complexation. However, no functional abnormality was found in the liver in any case. Curcumin binding also notably decreased nanoconjugate accumulation in lungs and significantly reduced inflammation. Biodistribution studies were consistent with previous reports, suggesting that optimization of the GNR size and surface coating is required for more efficient tumor localization via the enhanced permeability and retention (EPR) effect. Our studies demonstrate that DNA/RNA hairpins are suitable for GNR surface functionalization and enable delivery of cytotoxic nucleotides as well as curcumin in vivo with potential for synergistic anti-cancer therapy.

Introduction

Selective multimodality treatment is an emerging technique for treating patients with advanced malignancies using a less invasive approach.[1−4] In advanced stages of cancer, significant numbers of malignant cells survive therapeutic measures due to drug resistance, contributing to metastasis.[5,6] Therefore, there is an urgent need for progressive strategies to treat advanced cancer by overcoming chemoresistance and simultaneously reducing systemic toxicities.[7] Combined application of multiple modalities could enhance killing of chemoresistant cells, potentially without increasing systemic toxicities.[2,6,8] Gold nanorods (GNRs) are well known for their photothermal activity and inherent tumor targeting property via the enhanced permeability and retention (EPR) effect.[9,10] GNRs are now being developed as a suitable tool for constructing multimodality therapeutics.[11,12]

Translational applications of GNRs remain limited due to challenges with its surface functionalization.[13,14] Extensive studies have investigated the exchange of the GNR surface coating with biocompatible molecules including polyethylene glycol (PEG), peptides, and oligonucleotides.[11,15,16] The PEG coating results in an incomplete ligand exchange, while peptide conjugates display limited stability and dispersity in biologically relevant conditions.[17,18] Functionalization of GNRs with DNA has been studied by several research groups,[13,19,20] and in contrast with other biocompatible molecules, oligonucleotides could efficiently exchange with the overall surface coating. Further, DNA bases may be substituted with cytotoxic nucleotides to enhance the anti-tumor activity and DNA can form various secondary structures that enable intercalation of anti-neoplastic agents.[21,22] In our previous studies, we demonstrated that DNA hairpins could optimally functionalize the GNR surface without reducing its near-infrared (NIR) absorbance, while single- as well as double-stranded DNAs caused severe aggregation, including a major decrease in NIR absorbance. The resulting nanoconjugates are suitable for intracellular uptake and remain monodispersed in the cellular environment.[21] In a separate study, we also reported that the DNA hairpin could form stable, non-covalent complexes with curcumin that displayed notable biological activity.[22]

While earlier studies reported several applications for DNA-conjugated GNRs in vitro, the in vivo efficacy of these complexes has never been studied for further translational evaluation.[14,23,24] In the present study, we developed a multimodality nanoconjugate via functionalizing the GNR surface with a cytotoxic DNA hairpin followed by hydrophobic complexation of curcumin (Figure S1, Supporting Information). We used a cytotoxic nucleoside, 5-fluoro-2′-deoxyuridine (FdU), containing a hairpin for the GNR surface coating via Au–S bonding, while curcumin was bound in the hydrophobic environment of the GNR surface as well as in the hairpin minor groove (Figure S2, Supporting Information). FdU-containing oligonucleotides are well-known cytotoxic modalities having metabolic advantages over their clinically used counterpart, 5-fluorouracil.[25,26] Moreover, FdU hairpins should play multiple roles in our strategy, including GNR surface functionalization, exerting cytotoxic activity and providing a suitable environment for curcumin binding. The pleiotropic molecule curcumin could reduce the requirement of cytotoxic agents as well as the magnitude of hyperthermia necessary to achieve a beneficial anti-tumor effect, enhancing the multimodality approach.

We further investigated the biodistribution, toxicity, and anti-cancer activity of the nanoconjugates in combination with NIR laser in an Ehrlich carcinoma model. The conjugate was well tolerated in vivo and displayed moderate anti-tumor activity; however, 808 nm laser irradiation showed no synergistic photothermal effect. Atomic absorption spectroscopy demonstrated maximum gold accumulation in the liver followed by the spleen and lungs with limited accumulation in the tumor, consistent with GNR-size and coating optimization being required for more efficient tumor localization. Curcumin complexation countered the decrease in hemoglobin levels and reduced the pulmonary inflammation that was induced by FdU-hairpin-GNR treatment. Our findings demonstrate that the hairpin oligonucleotide could functionalize GNRs suitably for in vivo applications, delivering cytotoxic nucleotides and chemosensitizing agents for synergistic cancer therapy; however, the size distribution and coating need optimization to achieve an EPR effect.

Results and Discussion

Biophysical Characterization of the Conjugates

Analysis of UV–visible spectra (Figure a) showed that the NIR band of the cetyltrimethylammonium bromide (CTAB)-coated GNR is red-shifted from 791 to 810 nm upon DNA conjugation. A characteristic DNA band was observed in all FdU-hairpin-functionalized GNR samples. Curcumin complex formation changed the color of the suspension from reddish pink to yellowish brown (Figure S2a, Supporting Information) and significantly reduced the NIR band intensity by ∼18%. Curcumin complexes displayed a characteristic peak near 425 nm, indicating successful complexation. A red shift in the NIR band could have resulted from lower interferences of FdU-hairpin bases with the GNR surface, enhancing its surface plasmon resonance. Curcumin is complexed with FdU-hairpin-GNRs via a noncovalent interaction, where it has the opportunity to interact with the hairpin minor groove as well as the hydrophobic GNR surface. The close interaction between curcumin and the GNR surface leads to damping of plasmon resonance, resulting in NIR peak broadening and reduced band intensity. Fluorescence spectroscopy showed no band in the region of 450–650 nm in the FdU-hairpin-GNR sample, while curcumin binding resulted in a small peak at ∼490 nm. The characteristic fluorescence peak of curcumin at 535 nm is not observed due to the quenching effect of the GNR, which absorbs at 500–600 nm because of its transverse plasmon resonance (Figure b). Dynamic light scattering (DLS) spectroscopy showed that majority of the FdU-hairpin functionalized GNRs have hydrodynamic diameters in the range of 37.84–78.82 nm, while curcumin complexation shifted the hydrodynamic diameter to some extent toward the higher range, 50.75–91.28 nm (Figure c). Figure S3a, Supporting Information, also depicts a similar pattern of size distribution based on the scattering intensity. The presence of a single distribution in both cases with majority of the particles suggests no significant agglomeration.[27,28] The magnitude of the shift in the hydrodynamic diameter distribution is not enough to indicate aggregation; rather, it could be due to a size increase because of curcumin intercalation.[29−31] The zeta potential changed from +39.3 to −28.5 mV for the FdU-hairpin conjugation (Figure S3b, Supporting Information), suggesting a successful exchange of the GNR coating and removal of CTAB. In the case of curcumin complexation, the zeta potential was measured at −22.2 mV. A higher-magnitude zeta potential of the FdU-hairpin-GNRs suggested better stability in suspension, while curcumin binding reduced the repulsion between GNR conjugates to some extent. However, the resulting suspension should be sufficiently stable for further applications.[16,32,33] Dispersion of the nanoconjugates was further investigated using transmission electron microscopy (TEM). The micrographs depicted well-dispersed FdU-hairpin-GNR conjugates, while curcumin binding did not show significant changes in dispersion (Figure d,e).

Figure 1

Biophysical characterization of GNR conjugates. (a) UV–vis spectra showing the longitudinal SPR band of CTAB-capped GNRs around 791 nm, which has been shifted to 810 nm for FdU-hairpin functionalization. The band at 260 nm confirmed the DNA coating in both cases with and without curcumin complexation. The characteristic peak of curcumin at 425 nm was observed only in the case of the FdU-hairpin-GNR: curcumin conjugate. Curcumin complexation decreased the NIR band intensity by approximately 18% compared to the DNA-coated GNR. (b) Fluorescence spectra of curcumin, FdU-hairpin-GNRs, and FdU-hairpin-GNR:curcumin with 425 nm excitation. The characteristic fluorescence peak of curcumin was evident at ∼530 nm. FdU-hairpin-GNR:curcumin showed a small fluorescence band around 490 nm, while FdU-hairpin-GNR conjugates showed no fluorescence. (c) Representative histograms showing the distribution of hydrodynamic diameters of nanoconjugates with and without curcumin complexation. Majority of the hydrodynamic diameters of FdU-hairpin-GNR with and without curcumin ranged from 50.75 to 91.28 nm and 37.84 to 78.82 nm, respectively. (d, e) Transmission electron microscopic images of FdU-hairpin-GNR and FdU-hairpin-GNR:curcumin conjugates, respectively. The micrographs exhibit monodispersed nanoconjugates in both cases as curcumin complexation induced no significant change in dispersity (scale bar: 20 nm).

Evaluation of Photothermal Effect in Vitro

The heating efficiency of nanoconjugates was evaluated to determine the role of FdU-hairpin functionalization and curcumin complexation on the photothermal effect of GNR. Aqueous suspensions of 150, 300, and 600 μg/mL of concentration were irradiated by an 808 nm continuous laser beam with power of 1.0, 1.5, and 2.0 W for 30, 60, and 120 s. Representative graphs displaying the concentration, time, and laser power-dependent heating are exhibited in Figure a–c, respectively. All comparisons had p values less than 0.05. The concentration dependence of the photothermal effect of GNR conjugates followed a nonlinear pattern (Figure a) when irradiated at 1.5 W for 60 s. Compared to the CTAB coating, the FdU-hairpin conjugation caused a minute increase while curcumin complexation resulted in an ∼20% decrease at a 300 μg/mL concentration.[21] FdU-hairpin-functionalized GNRs should achieve biologically relevant hyperthermia with application of optimum laser power and time of irradiation. For example, 300 μg of CTAB- and FdU-hairpin-coated GNRs can increase the temperature of 1 mL of water by ∼8 °C when irradiated with a 1.5 W laser for 60 s, while curcumin-complexed GNR conjugates cause an increase of 6.2 °C with identical parameters. Previous studies reported that hyperthermia-induced malignant-cell death could be achieved above 42 °C.[34,35] The efficiency of heat-induced cell death could be further enhanced with simultaneous application of suitable cytotoxic agents, such as FdU and curcumin.[22]Figure b depicts the linear, time-dependent heating of GNR conjugates for a given concentration and laser power, which is informative to verify the attainment of therapeutically relevant hyperthermia. For example, 90 s irradiation of 300 μg of CTAB- as well as FdU-hairpin-coated GNRs can increase the temperature of 1 mL of water by 11 °C with 1.5 W laser power, whereas curcumin-complexed FdU-hairpin-GNRs caused a 5 °C temperature increase with similar parameters. However, combining multiple cytotoxic agents may reduce the extent of heating required to induce cell death. The laser power-dependent heating shows an exponential pattern (Figure c). While 1.5 W laser irradiation for 60 s can increase the temperature of 1 mL of water by 7.5 and 9.0 °C for 300 μg of CTAB- and FdU-hairpin-coated GNRs, respectively, a 6.2 °C temperature increase was observed with curcumin complexation in similar conditions. Thus, FdU-hairpin-functionalized GNRs with and without curcumin binding should be able to achieve biologically effective photothermal activity, utilizing available parameters. Moreover, elevated hyperthermia could be obtained in the tumor microenvironment, compared to in vitro conditions.[36,37]

Figure 2

Evaluation of photothermal effects in vitro. Representative plots demonstrating photothermal effects of different GNR conjugates. Data sets are represented as means ± SEM, n = 3. Statistical analysis was done using one-way ANOVA with Tukey’s multiple comparison method; p value < 0.05 was considered statistically significant. (a) Concentration-dependent heating showed nonlinear relations in all cases when irradiated with 1.5 W of laser power for 60 s. (b) All GNR conjugates of 300 μg/mL of concentration demonstrated a linear, time-dependent temperature increase when irradiated with a 1.5 W laser. (c) Laser power-dependent photothermal effects suggested exponential relations in all cases of GNR conjugates with a 300 μg/mL concentration and 60 s irradiation time.

Evaluation of Anti-tumor Activity

Eradication of tumor tissue using a multimodality approach is an important objective in cancer research, especially in chemoresistant diseases. To investigate the potential synergistic anti-tumor activity of our GNR conjugates, we developed solid tumors in the neck region of Swiss albino mice (Figure S4a, Supporting Information). Initial tumor volumes were ∼100 mm3. Animals were treated by tail-vein injection of either of the different nanoconjugates, such as FdU-hairpin-GNR (FG), FdU-hairpin-GNR:curcumin (FGC), or sterile water (NT).[38] Each nanoconjugate-treated group was further divided into laser-treated and mock-treated groups to create six groups as mentioned in the Methods section. Laser treatment was done by continuously irradiating the tumor for 90 s with an 808 nm laser beam of 1.5 W/cm2, as shown in the video clip of the associated content. The time-dependent tumor growth is shown in Figure a. Beginning around day 7, tumors treated with FGC and its laser combination (FGCL) displayed slower tumor growth, while no-treatment (NT) and laser-only (L) groups continued the rapid growth rate; the difference in growth rates was statistically significant (p < 0.05; Table S1, Supporting Information) until conclusion of the study. FG with and without laser irradiation also showed higher growth rates comparing to FGC and FGCL, starting around day 14, and they remained statistically significant until the end of the study (p < 0.05). There was no statistically significant difference in tumor growth for FG-treated animals in the presence or absence of the laser with the no-treatment or laser-only group. The results demonstrate that curcumin-complexed FdU-hairpin-GNRs showed moderate anti-tumor activity, while a GNR-mediated photothermal effect was not observed with laser irradiation. Figure b depicts a graphical representation of the tumor weight after completion of the treatment and follow-up period. The results are consistent with the change in the tumor volume. Compared to the no-treatment group, only FGC and FGCL showed statistically significant reduction in tumor weight (Table S2, Supporting Information). FdU-hairpin-GNR treatment and its laser combination obtained no significant reduction in the tumor weight when compared to control groups. Laser irradiation showed no additional change in the tumor weight in any group. In Figure c, photographic images of representative tumors isolated from different experimental groups showed longer dimensions with the help of a caliper. Body weights of all animals were measured weekly, beginning from the day of tumor cell inoculation. Animals did not show any significant reduction in body weight due to nanoconjugate treatment (Figure S4b, Supporting Information) until completion of the study.

Figure 3

Anti-tumor activity of nanoconjugates. (a) Graphical representation showing time-dependent changes in the relative tumor volume (V/Vo) during the treatment and follow-up period. There was no significant change up to day 7. Starting day 14, the rate of tumor growth gradually slowed down in curcumin-complexed FdU-hairpin-GNRs (FGC) along with their laser combination (FGCL), while no-treatment (NT) and laser-only (L) groups remained rapidly growing. Results are depicted as means ± SEM, n = 6, in all cases. Statistical analysis was done using one-way ANOVA by Tukey’s multiple comparison method; p < 0.05 was considered statistically significant. (b) Representative bar diagram of tumor weight from different experimental groups. FdU-hairpin-GNR:curcumin (FGC) and its laser combination (FGCL) showed lower tumor weights compared to the no-treatment group. Data sets are represented as means ± SEM, n = 6. Statistical analysis was done using one-way ANOVA with Tukey’s multiple comparison method. “*” indicates p < 0.05 for NT vs FGC and NT vs FGCL. (c) Photographic images of representative tumors from different treatment groups, taken at the end of the experiment after sacrifice.

The cytotoxicity of FdU-hairpin-functionalized GNRs was also investigated in HeLa cells with and without curcumin complexation and in the presence or absence of 808 nm laser irradiation (Figure S5, Supporting Information). It is evident from the data that FdU-hairpin-coated GNRs demonstrated cytotoxicity in a dose-dependent manner at concentrations ranging from 93.75 to 3000 ng/mL. Curcumin complexation further enhanced the cytotoxic effect in all cases. However, 808 nm laser irradiation caused no additional cytotoxic activity.

Investigation of the Biodistribution and Blood Biochemistry

Figure a depicts the distribution of gold in the tumor tissues and vital organs, such as the liver, kidney, lung, heart, and spleen. The biodistribution was investigated by atomic absorption spectroscopy after 4 weeks of treatment and 2 weeks of a follow-up period. The liver and spleen showed maximum accumulation of gold in the range of 1.0–1.4 μg/g followed by the lungs and heart with 0.3–0.7 and 0.2–0.4 μg/g, respectively. In all treatment groups, the tumor showed a low concentration of gold, ∼0.1 μg/g, while the kidneys also demonstrated gold accumulation at a similar level. Maximum gold accumulation in the liver and spleen could be attributed to their fenestrated structure of capillaries.[39] The endothelium of the kidney glomeruli is also known to be fenestrated. However, our data show very low concentrations of gold in the kidneys in all treatment groups, possibly because the basal membrane of the glomeruli acts as a barrier for the GNR.[39] Very low concentrations of gold in tumor tissue suggested that the nanoconjugate size distribution and surface coating was non-optimal for indirect tumor targeting via EPR. Curcumin complexation possibly changed the GNR surface character that decreased the distribution of gold in the lungs significantly.

Figure 4

Biodistribution of gold and blood biochemistry. (a) Graphical representation depicting the biodistribution of Au using atomic absorption spectroscopy. Maximum accumulation was observed in the cases of the liver and spleen followed by that of the lungs, whereas low levels of gold were found in the tumor, kidney, and heart. FdU-hairpin-GNR-treated animals (FG and FGL) showed greater accumulation of gold in the lungs compared to groups with curcumin complexes (FGC and FGCL). Results are expressed as means ± SEM, n = 6. Statistical analysis was done using a one-way ANOVA test by Tukey’s multiple comparison method; p < 0.05 was considered statistically significant in all cases. “*”, “#”, “Δ”, and “φ” are designated to FG vs FGC, FG vs FGCL, FGL vs FGC, and FGL vs FGCL, respectively. (b) Graphical representations of aspartate aminotransferase (AST) activity and alakaline phopsphatase (ALP) activity in serum and serum creatinine and urea levels. Results showed no significant change in any case. Results are expressed as means ± SEM, n = 6. Statistical analysis is done using a one-way ANOVA test by Tukey’s multiple comparison method (p ≤ 0.05).

Liver function test results (Figure b, top row, and Table S3, Supporting Information) showed no significant change in aspartate aminotransferase (AST) activity. Alkaline phosphatase (ALP) activity decreased slightly in cases of FdU-hairpin-GNR treatment and its laser combination compared to no treatment; however, ALP activities of all groups remained within the normal range (Table S4, Supporting Information). Overall, AST as well as ALP activity suggested no significant alteration in hepatocellular function due to any experimental treatments. Blood urea and creatinine levels demonstrated mild decreases in all treatment groups compared to no treatment (Figure b, bottom row); however, all groups remained within normal limits (Tables S5 and S6, Supporting Information). Therefore, urea and creatinine levels did not indicate any treatment-related renal dysfunction in the presence or absence of laser irradiation.[40]

Evaluation of Hematological Parameters

Hematological parameters of all experimental groups were evaluated after the completion of the study, and the results are depicted in Table . Laser irradiation caused significant reduction in hemoglobin (Hb) levels, 24.4%, compared to the no-treatment group (Table S7, Supporting Information). The maximum Hb decrease, 28.89%, occurred in the FGL group. Curcumin complex formation countered the Hb decrease, which was induced by FGL. In the case of curcumin-bound conjugates, the deficiency of Hb was 4%, while its laser combination showed Hb reduction by 7.4% compared to no treatment. The RBC count also confirmed the hemolytic effect of FGL, while curcumin binding reduced the magnitude of hemolysis.[41−44] The RBC count was maximum, 6.8 × 106/mm3, in no treatment followed by curcumin-bound nanoconjugate treatment, 6.4 × 106/mm3 and its laser combination, 5.7 × 106/mm3. The RBC count was significantly lower in FG and 808 nm laser irradiation as well as in their combination, 5.4 × 106, 5.4 × 106, and 5.3 × 106/mm3, respectively (Table S8, Supporting Information).

Table 1

Tabular Representation of Hematological Parameters of Different Treatment Groupsa

group	Hb (g/dL)	RBC (106/mm3)	WBC (103/mm3)	neutrophil (%)	lymphocyte (%)
NT	13.5 ± 0.1	6.8 ± 0.06	14.2 ± 0.17	50.55 ± 0.8	41.4 ± 0.93
L	10.2 ± 0.1	5.4 ± 0.06	12.5 ± 0.27	50.4 ± 1.06	42.22 ± 0.63
FG	12.5 ± 0.13	5.4 ± 0.17	11.9 ± 0.13	55.4 ± 0.83	40.3 ± 0.83
FGL	9.6 ± 0.06	5.3 ± 0.06	12.4 ± 0.13	52.4 ± 0.77	42.8 ± 0.8
FGC	12.9 ± 0.06	6.4 ± 0.17	13.8 ± 0.13	48.4 ± 0.63	48.33 ± 0.73
FGCL	12.4 ± 0.1	5.7 ± 0.06	13.7 ± 0.2	45.3 ± 0.46	46.45 ± 0.7

Results are expressed as means ± SEM, n = 6, in all cases. Statistical analysis was done using a one-way ANOVA test by Tukey’s multiple comparison method; p < 0.05 was considered statistically significant. Laser irradiation reduced the hemoglobin count in both laser-only (L) and FdU-hairpin-GNR + laser (FGL) groups. Curcumin complexation improved the hemoglobin levels both in the presence and absence of laser irradiation (FGC and FGCL). Changes of the RBC count followed similar pattern to hemoglobin. The WBC count also increased in the presence of curcumin both with and without a laser, compared to FdU-hairpin-GNR and FdU-hairpin-GNR + laser. The neutrophil percent was higher in the FdU-hairpin-GNR treated group and its laser combination, compared to the no-treatment and laser-only group; the percent of neutrophils further decreased in the case of curcumin complexation. The reverse pattern was observed for the lymphocyte percentage.

Our experimental results demonstrated that the WBC count followed a similar pattern to RBC. However, in all cases, the WBC count remained within the normal range of 10.2 × 103 to 14.0 × 103/mm3 (Table S9, Supporting Information).[40] The neutrophil population was maximum in the FG treatment, 55.4%, followed by its laser combination, 52.4%. Curcumin complexation reduced the neutrophil population both in the presence and absence of a laser (48.4 and 45.3%, respectively). The increased neutrophil percent is consistent with an inflammatory response to FG, while the curcumin complex formation appeared to counter the inflammatory response.[45,46] The lymphocyte population followed the reverse pattern compared to neutrophils with FG lowering lymphocyte levels and the curcumin complex formation countering the reduction. Statistical analysis of changes in neutrophil and lymphocyte populations is depicted in Tables S10 and S11, Supporting Information, respectively. FACS analysis of blood samples, probing CD4 and CD8 markers, also suggested an increase in the T-lymphocyte population for curcumin-bound FdU-hairpin-GNR treatment and its laser combination. Further, flow cytometry data revealed elevated levels of NK cells with curcumin complexation (Figure S6, Supporting Information).

Histopathological Investigation

Histopathological investigation of tumors along with the vital organs from all experimental groups was done by analyzing hematoxylin and eosin (H&E)-stained tissue sections (Figure ). The NT group showed a normal tumor morphology characterized by hyperchromatic nuclei with irregular nuclear membranes, whereas the cytoplasm was scanty. Laser-only treatment showed a similar morphology to that of the no-treatment group, depicting discrete necrotic foci with nuclear disintegration. Apoptotic bodies with highly condensed pyknotic, ink-dot-like nuclei and an intensely eosinophilic cytoplasm, as marked by black dashed circles, were detected in all nanoconjugate-treated groups in the presence and absence of a laser. Patches of necrotic regions were also detected in these treatment groups, marked with yellow arrows. Histological analysis indicates both FdU-hairpin-GNR and FdU-hairpin-GNR:curcumin could induce apoptosis to some extent with and without laser irradiation. The observations were further reconfirmed by immunohistochemical staining.

Figure 5

Histological investigation of tumor sections from different treatment groups. Tumor tissues were processed, sectioned, and stained with hematoxylin–eosin (H&E). The nuclei were stained with the dark purple color of hematoxylin. Necrotic tumor foci (marked with yellow arrows) were homogeneously pink with dark purple dotted structures of degraded nuclei. All of the nanoconjugate treatment groups showed the presence of apoptotic bodies with a bright eosinophilic cytoplasm and condensed nucleus, marked with black dashed circles. The magnification is 400× with a scale bar of 50 μm.

Figure represents microscopic images of H&E-stained liver tissue sections from all experimental groups with 200× magnification. The NT and L treatment groups showed normal features of hepatic parenchyma. Images from the FG treatment group depicted massive infiltration of mononuclear cells, marked using black arrows, around a portal triad along with signs of fibrosis. These infiltrates majorly comprise focal aggregates of lymphocytes, plasma cells, and macrophages, accumulated due to inflammation resulting from nanoconjugate buildup.[47] Meanwhile, fibrosis represents a reparative process to replace injured cells, reflecting moderate levels of damage in hepatic parenchyma.[48] A closer view of the lymphocytic infiltration around the central vein is depicted in Figure S7 of the Supporting Information with 400× magnification. However, a higher magnification showed no morphological distortion of hepatocytes in any case. The degree of inflammatory infiltration was significantly reduced for FGC and FGCL treatments with very few lymphocytic cells around the central veins. No sign of fibrosis was observed in any case of curcumin-bound nanoconjugate treatments. Evident from the observations, curcumin complexation significantly mitigated the inflammatory response in hepatic tissue.

Figure 6

Histopathological investigation of H&E-stained liver tissue sections. The no-treatment group (NT) demonstrated normal hepatic parenchyma. The laser-only group (L) shows similar histology to NT. In the FdU-hairpin-GNR-treated group (FG), severe inflammation with lymphocytic infiltration (indicated with black arrows) and fibrosis was observed around the portal triad. A similar observation was found in the FdU-hairpin-GNR + laser group (FGL) without any sign of fibrosis. The curcumin-complexed groups with and without a laser (FGCL and FGC) showed lesser lymphocytic infiltration (shown with black arrows). Magnification: 200×, scale bar: 50 μm.

Figure exhibits H&E-stained lung tissue sections of all studied groups with 400× magnifications (Figure S8, Supporting Information, 100× magnification covering a larger area). The NT group showed a typical alveolar structure with normal thickness of the septa and bronchial walls. The bronchial lumen is designated with the letter “B” in the image. The alveolar septa slightly thickened in the L treatment group. In the FG treatment group, the alveolar septa were severely thickened due to infiltration of inflammatory cells along with erythrocyte deposition, resulting in remarkable shrinkage of alveolar space. Congestion of alveolar space might lead to impaired gas exchange and pulmonary dysfunction, resulting from GNR accumulation.[49] Laser irradiation did not change the conditions significantly; peribronchial thickening and congestion of bronchial space with erythrocytic deposition (Figure S8, Supporting Information) were observed. The degree of infiltration was notably reduced in curcumin-complexed nanoconjugates, both in the presence and absence of a laser, possibly due to the anti-inflammatory effect of curcumin.[50] Moreover, curcumin complexation probably altered the GNR surface character, reducing nanoconjugate localization in the lungs that might also decrease pulmonary inflammation.

Figure 7

Histopathological investigation of H&E-stained lung tissue sections. The bronchial lumen is marked with “B” in all relevant images. The no-treatment group (NT) showed histological features of normal lung tissue with thin alveolar septa and regular thickness of the bronchial wall. The laser-only group (L) showed more thickening of the perivascular space as compared to NT. FdU-hairpin-GNR-treated groups in the presence and absence of a laser (FGL and FG) showed severe interstitial inflammatory cell infiltration with alveolar wall thickening and erythrocyte deposition in the alveolar space. FdU-hairpin-GNR + laser also showed a similar inflammatory response. Inflammatory infiltration was significantly reduced in both of the curcumin-bound nanoconjugate-treated groups (FGC and FGCL) with little thickening of alveolar septa. Magnification of 400× with a 50 μm scale bar.

Images representing histological sections of spleen are exhibited in Figure . Both NT and L groups showed normal morphological features of splenic tissue with discrete white and red pulp, designated as “WP” and “RP”, respectively. Negligible megakaryocytes were observed in both cases. All other treatment groups showed an increased volume of white pulp and numerous megakaryocytes. Additionally, FG and FGL showed a presence of fibrous tissue in the red pulp, marked with the letter “F”; this was absent in the case of curcumin-complexed treatment groups. The red pulps of the spleen are rich in cords of Billroth and splenic sinusoids, whereas white pulps are composed of lymphocytes.[51] The increase in the volume of white pulp suggests greater infiltration of lymphocytes as a chronic inflammatory or immune response due to nanoparticle accumulation.[52,53] A sign of fibrosis in the red pulp regions also supports the response to inflammation in spleen. An increase in the number of splenic megakaryocytes signifies hematopoetic stress.[54]Figures S9 and S10 in the Supporting Information show H&E-stained sections of myocardial and renal tissues, respectively. None of the experimental animals demonstrated any morphological abnormality in myocardial or renal tissues.

Figure 8

Histopathological investigation of H&E-stained splenic tissues. The no-treatment group (NT) showed normal morphological features with distinct white and red pulps and a limited number of megakaryocytes. The laser-only group (L) showed similar morphology to the no-treatment group. An increase in the volume of white pulp along with numerous megakaryocytes (enlarged in the insets) was observed in the case of FdU-hairpin-GNR with and without a laser (FGL and FG) as well as curcumin-bound nanoconjugates with and without a laser (FGCL and FGC). In all images, the white pulp is marked as “WP”, red pulp as “RP”, and fibrous tissue as “F”. Magnification of 100× and a scale bar of 50 μm.

Immunohistochemical Analysis

Immunohistochemical analysis was done against the active caspase 3 protein to detect apoptotic cells in tumor and vital organs such as the liver, spleen, kidney, heart, and lung. Figure depicts the microscopic images of immunohistochemically stained tumor tissues along with the graphical representation of the apoptotic index. NT and L groups showed minimal presence of cleaved caspase 3 positive cells with percentages of 1.48 and 2.14, respectively. FG treatment with and without a laser showed increased caspase 3+ cells, 5.37 and 5.23%, respectively. The maximum number of apoptotic cells was found in the FGC treatment with 12.98% active caspase 3 expressing cells, while a combination with a laser showed a similar pattern of apoptosis with a value of 10.52%. These increased apoptotic cell percentages for the curcumin complex are consistent with increased anti-tumor activity for these treatment groups (Figure ). A similar magnitude of apoptosis in the presence as well as absence of a laser suggests no significant photothermal effect, possibly because of very little GNR accumulation in tumor tissue.

Figure 9

Immunohistochemical analysis of tumor tissues. (a) Immunohistochemical investigation of an apoptotic marker, cleaved caspase 3, in tumor tissues from different experimental groups. Hematoxylin was used as a counter stain, and 3,3′-diaminobenzidine (DAB) was used as a chromogen. Active caspase 3 positive cells appeared brown in color. The no-treatment and laser-only groups (NT and L) showed very few active caspase 3 positive cells. FdU-hairpin-GNR with and without a laser (FGL and FG) showed a moderate number of apoptotic cells. The maximum number of apoptotic cells was found in curcumin-complexed FdU-hairpin-GNR (FGC) and its laser combination (FGCL). In all cases, square regions depicting caspase 3 positive cells are magnified in the insets. Magnification is 400× with a scale bar of 50 μm. (b) Graphical representation depicting the percentage of cleaved caspase 3 positive cells in the tumoral tissues. Maximum caspase 3 positive cells were found in curcumin-complexed FdU-hairpin-GNR group (FGC) having a value of 12.98% followed by its laser combination (FGCL) with 10.52%. FdU-hairpin-GNRs with and without a laser (FGL and FG) exhibited percentages of 5.37 and 5.23, respectively. Minimum percentages of caspase 3 positive cell were observed in no-treatment (NT) and laser-only (L) group with values of 1.48 and 2.14, respectively. Results are expressed as means ± SEM, n = 10. Statistical analysis is done using a one-way ANOVA test by Tukey’s multiple comparison method. ***p < 0.001, **p < 0.01 when compared to the NT group; < 0.001, < 0.05 when compared to group L; ΔΔΔp < 0.001 when compared to group FG; and φφφp < 0.001 when compared to group FGL.

Figure S11 in the Supporting Information represents photomicrographs of immunohistochemically stained liver tissue sections. Apoptotic cells with positively stained condensed nuclei were detected in the sinusoidal lining for all treatment groups, i.e., FdU-hairpin-GNRs as well as curcumin-bound FdU-hairpin-GNRs, both in the presence and absence of a laser, suggesting inflammatory stress due to high accumulation of nanoparticles in the liver. On the other hand, the absence of active caspase 3 expression in the hepatocytes is in compliance with no significant negative effect on the liver functioning due to nanoconjugate treatment. In Figure S12, Supporting Information, immunohistochemical analysis of lung sections showed active caspase 3 positive cells in the case of FdU-hairpin-GNR treatment, both with and without a laser, while curcumin binding demonstrated no expression of any apoptotic marker in any case. The no-treatment and laser-only groups also showed no active caspase 3 positive cell. The immunohistochemical investigation of splenic, myocardial, and renal tissues is represented in Figures S13– S15 in the Supporting Information, respectively; the results showed no active-caspase 3 positive cells in any experimental group for all cases.

Conclusions

In the present study, we successfully developed nanoconjugates by functionalizing GNRs with a cytotoxic FdU-hairpin along with curcumin complexation for a multimodal, therapeutic approach. We further investigated the biodistribution, anti-tumor activity, and toxicity of FdU-hairpin-functionalized GNRs in combination with 808 nm laser irradiation in an Ehrlich carcinoma model. This study demonstrated for the first time that hairpin-structured oligonucleotides can suitably functionalize GNRs for in vivo applications. Our studies showed no acute toxicity due to nanoconjugate aggregation and blockage in the circulatory system upon intravenous delivery. Most of the nanoconjugates accumulated in the liver and spleen, while a very small fraction was found in the tumor, both in the presence and absence of curcumin. Curcumin-complexed nanoconjugates showed moderate anti-tumor activity, while the absence of curcumin caused higher tumor growth. Accordingly, the maximum degree of apoptosis in tumoral tissue was observed in the curcumin-bound nanoconjugate treatment. The 808 nm laser irradiation did not contribute any additional reduction in the tumor volume for any group. FdU-hairpin-GNR treatment in combination with a laser showed a hemolytic effect, while curcumin significantly improved the hemoglobin and RBC counts. FdU-hairpin-GNR treatment also induced an inflammatory response in the liver, spleen, and lung. In contrast, curcumin complexation reduced inflammation, especially in the liver and lung. The larger accumulation of nanoconjugates in the liver did not result in any hepatocellular dysfunction. Finally, hairpin-structured oligonucleotides could be suitable for GNR functionalization as well as cytotoxic nucleotide and curcumin delivery for synergistic cancer therapy; however, the GNR size and surface coating need to be optimized for passing through the liver and spleen to achieve the EPR effect.

Methods

Gold Nanorod Functionalization and Curcumin Complexation

Gold nanorod synthesis and surface functionalization were performed following the previously published work of Das et al.[21] The synthesized GNRs had an average length of 30.0(±5) nm and diameter of 10.0(±2) nm, including absorbance maxima in the range of 790–810 nm. For GNR functionalization, we used a thiolated DNA hairpin with 10 adenine bases at the 3′ end followed by a CGAAG loop and 10 FdU bases at the 5′ end (Figure S1, Supporting Information). FdU containing a thiolated DNA hairpin was purchased from the Wake Forest School of Medicine, United States. Before conjugation, all thiolated DNA sequences were purified by gel-filtration chromatography using Sephadex G-25. The CTAB-coated GNR suspension was washed to remove excess CTAB. DNA stock solutions were prepared in HPLC-grade water at a concentration of 100 μM. The monomeric DNA hairpin was prepared by heating DNA solutions at 85 °C for 5 min followed by snap-cooling in ice. Thiolated DNA molecules were conjugated on the GNR surface by adding ice-cold DNA solutions in the pre-chilled GNR suspensions. Unattached DNA molecules were removed by centrifuging the conjugation mixture at 4 °C for 1 h at 10,000g; nanoconjugates were collected in a pellet, and free DNA molecules were separated in the supernatant. The final FdU-hairpin-functionalized GNR pellet was resuspended in an equal volume of HPLC-grade water.

For curcumin complexation, a 100 μM stock was prepared by dissolving curcumin powder in absolute ethanol (99% v/v), while mixing on a magnetic stirrer overnight at room temperature (25–30 °C), protected from light. The calculated volume of curcumin was first taken in a microcentrifuge tube, and the FdU-hairpin-GNR conjugate was added then mixed thoroughly to make a final DNA:curcumin molar ratio of ∼1:1 and final ethanol concentration of 25% (v/v). The color of the suspension turned yellowish brown from reddish pink. The complexation mixture was then incubated at room temperature (25–30 °C) for 30 min followed by 30 min at 4 °C. The entire process was done protected from light. Following incubation, the reaction mixture was centrifuged at 10,000g for 1 h at 4 °C to remove unbound curcumin. The supernatant was discarded, and the pellet was re-suspended in an equal volume of HPLC-grade water. Curcumin complexation turned the color of the GNR conjugate from reddish pink to yellowish brown (Figure S2a, Supporting Information). DNA attachment and curcumin complexation were further confirmed by UV–visible and fluorescence spectroscopy.

Spectroscopic Characterization of the Conjugates

UV–visible spectra were acquired under ambient conditions using a TECAN infinite M 200 PRO spectrophotometer. All conjugate suspensions were diluted three times with HPLC-grade water and scanned at a rate of 1 nm/s over the range of 230–1000 nm using a quartz cuvette of 1 cm in path length. For fluorescence spectroscopy, FdU-hairpin-GNR conjugates with and without curcumin complexation were scanned in a CARY 100 fluorescence spectrophotometer. Samples were excited at 425 nm, and emission spectra were recorded for the region of 450 to 650 nm at a scan rate 1 nm/s. Dynamic light scattering spectroscopy of the GNR conjugates was performed using a Zetasizer (Nano S ZEN -1600, Malvern) instrument. One milliliter of the aqueous suspension of each sample was placed in a clear disposable cuvette, and the hydrodynamic diameter was measured at 25 °C using an incident light of 632.8 nm with a detection angle of 173O and the attenuator setting at 7.

Microscopic Characterization of the Conjugates

Any change in the dispersion of GNR conjugates due to curcumin complexation was investigated by transmission electron microscopy. Conjugate suspensions were dropped on 300 mesh carbon-coated copper grids and incubated for 30 min at room temperature (25–30 °C). Excess liquid was removed and air-dried under similar conditions. Samples were visualized with 200 kV of accelerating voltage at ambient temperature using a TECNAI TF 200 transmission electron microscope.

Evaluation of Photothermal Effect in Vitro

The heat emitted by GNR conjugates upon NIR irradiation was evaluated by measuring the change in temperature of their aqueous suspensions with a mercury thermometer. One milliliter of each sample was placed in sealed NIR-transparent glass cuvettes equipped with a thermometer. Aqueous suspensions of varying concentrations, 150, 300, and 600 μg/mL, were then irradiated by an 808 nm laser beam using a CNI-MDL-III 808 (FC) laser system. Samples were irradiated with power levels of 1.0, 1.5, and 2.0 W for 30, 60, or 120 s. The initial temperature of each solution was recorded prior to irradiation, and the maximum temperature was recorded post irradiation.[55] The net temperature increase was estimated by subtracting the heating of pure water in similar conditions.

Development of Tumors in Mouse Models and Evaluation of Anti-cancer Effects

All animal experiments were performed under a protocol approved by the institutional animal ethics committee of the Chittaranjan National Cancer Institute (CNCI), Kolkata. The tumor model was developed by subcutaneously injecting 1 × 107 Ehrlich ascites carcinoma (EAC) cells in the neck region of six week old female Swiss Albino mice. EAC cells were gifted by Dr. Subhadip Hajra (CNCI, Kolkata). Animals were used for the experimental procedure ∼14 days following inoculation of the tumor cells when the tumor volume reached ∼100 mm3. A total of 36 mice were divided in 6 groups randomly, according to their treatment procedure: (1) NT (no-treatment group), (2) L (laser-only-treated group), (3) FG (FdU-hairpin-GNR-treated group), (4) FGL (FdU-hairpin-GNR and laser-treated group), (5) FGC (FdU-hairpin-GNR:curcumin-treated group), and (6) FGCL (FdU-hairpin-GNR:curcumin and laser-treated group). Each of the 24 mice from FG, FGL, FGC, and FGCL groups was injected with 80 μL of the corresponding conjugate suspension (∼150 μg/mL) via their tail veins once a week for four weeks. The total injected dose of gold was ∼0.05 mg/animal (∼1.6 mg/kg body weight).[11] In the case of the animals from NT and L groups, 80 μL of sterile water was injected in similar conditions. After 4 h of injection, mice from L, FGL, and FGCL groups were irradiated using a CNI-MDL-III-808 (FC) laser with continuous exposure for 90 s at a power level of 1.5 W/cm2.[55] Animals were anesthetized for the laser treatment procedure. Anesthesia was administered by intraperitoneal injection of ketamine HCl (80 mg/kg) along with xylazine (10 mg/kg). The tumor size was measured once a week for 6 weeks following treatment initiation using a Mitutoyo Absolute AOS Digimatic digital caliper. The tumor volume was calculated as V = (tumor length × tumor width2)/2. The relative tumor volume was calculated as V/Vo (Vo is the tumor volume when the treatment was initiated). At the end of 6 weeks, the blood samples were collected from each animal by retro-orbital bleeding. Then, the animals were sacrificed and the vital organs such as the liver, kidneys, spleen, heart, and lungs along with the tumor were isolated. All superficial burned patches on tumors were cut off with significant margins to avoid any artifact for histological and immunohistological analysis.

Investigation of the Biodistribution of Au and Blood Biochemistry

Tumor sections along with the liver, kidney, spleen, heart, and lung were isolated from the treated animals and washed with phosphate-buffered saline, PBS, followed by freezing at −20 °C for later use. Frozen tissue samples were thawed at room temperature and weighed. For each 1 g of the sample, 4 mL of nitric acid and 1 mL of 30% H2O2 were added in a closed Teflon vessel followed by incubation at room temperature for 45 min.[56,57] Tissues were further digested in a CEM MARS Press microwave digestion system at 160 °C and a power level of 400 W, 100% efficiency, with 20 min of ramping and 15 min of holding time.[58] After digestion, each of the digested tissue samples were evaporated completely to remove the acid and resuspended in 2 mL of 0.4% HCL (v/v). Atomic absorption spectra were taken at 242.2 nm using a graphite tube atomizer (Varian, GTA120) coupled with a Zeeman atomic absorption spectrometer (Varian, AA240Z) and a hollow cathode gold lamp (Agilent Technologies India Pvt. Ltd). Data were analyzed using SeptraAA 5.0 and Microsoft Excel 2010 software.

For the liver and kidney function test, blood samples were collected in non-heparinized tubes and centrifuged at 3000 rpm for 10 min followed by separation of serum fractions. Analysis of aspartate aminotransferase (AST) and alkaline phosphatase (ALP) activity was done following the 2,4-DNPH method and Kind and King’s assay, respectively. Serum creatinine and urea levels were determined following alkaline picrate and diacetyl monoxime (DAM) assays, respectively. All serum analyses were done using commercial kits as mentioned in the Supporting Information.

Investigation of Hematological Parameters

For the investigation of haematological parameters, blood samples from all animals were collected in tubes containing heparin (20 IU/mL).[59] Sahli’s method was employed to determine the blood hemoglobin level. Hematological parameters including RBC, WBC, and differential WBC counts were performed by standard procedures.[60]

Histological and Immunohistological Analysis

For histological and immunohistological analysis of tumor and vital organs, tissue sections were fixed in 10% formalin. Tissue embedding, sectioning, and mounting on glass slides were done following standard protocols.[61,62] For histological analysis, tissue sections were stained with hematoxylin and eosin.[63] For immunohistochemical study, sections were deparaffinized and hydrated following the Abcam standard protocol.[64] The sections were pre-treated with a heat-mediated antigen retrieval method in citric acid buffer (pH 6.0) at 70 °C for 20 min. After antigen retrieval, the sections were completely cooled to room temperature while keeping them in the citric acid buffer. Then, the sections were washed with PBS. The areas around the tissue sections were blotted, and a hydrophobic barrier was drawn around each section using a PAP pen (Sigma-Aldrich, cat. no. Z377821-1EA). Blocking was done using a commercially available immunoperoxidase secondary detection system (Millipore, cat no. DAB150). One drop of 20 μg/mL anti-cleaved caspase 3 antibody (Abcam, cat no. ab2302) was added on each of the sections and incubated overnight at 4 °C in a humid chamber. The biotinylated secondary antibody and streptavidin–HRP conjugate system along with the 3,3′-di-amino-benzidine (DAB) chromogen were used for detection, utilizing the reagents and protocol provided in the kit, Millipore, cat. no. DAB150. Both H&E-stained and immuno-stained slides were visualized using a bright-field DM 100 microscope (Leica MikrosystemeVertrieb GmbH, Germany). Images were processed using Image J 1.46 software.[65]

Statistical Analysis

Graphical representations of experimental results were prepared using Microsoft Excel 2010 and Graphpad Prism (version 7). All the experimental data sets are expressed as means ± SEM. Data sets were statistically analyzed using one-way analysis of variance (ANOVA) followed by Tukey’s multiple comparison test using Graphpad Prism (version 7) software. In all cases, p value < 0.05 was considered statistically significant.