Hypoxia-triggered single molecule probe for high-contrast NIR II/PA tumor imaging and robust photothermal therapy.

Hypoxia is a common characteristic of solid tumors. This important feature is associated with resistance to radio-chemotherapy, which results in poor prognosis and probability of tumor recurrence. Taking advantage of background-free NIR II fluorescence imaging and deeper-penetrating photoacoustic (PA) imaging, we developed a hypoxia-triggered and nitroreductase (NTR) enzyme-responsive single molecule probe for high-contrast NIR II/PA tumor imaging and hypoxia-activated photothermal therapy (PTT), which will overcome cellular resistance during hypoxia.
Methods: The single molecule probe IR1048-MZ was synthesized by conjugating a nitro imidazole group as a specific hypoxia trigger with an IR-1048 dye as a NIR II/PA signal reporter. We investigated the NIR II fluorescence, NIR absorbance and photothermal effect in different hypoxia conditions in vitro, and performed NIR II/PA tumor imaging and hypoxia-activated photothermal therapy in mice.
Results: This versatile molecular probe IR1048-MZ not only realized high-contrast tumor visualization with a clear boundary by NIR II fluorescence imaging, but also afforded deep-tissue penetration at the centimeter level by 3D PA imaging. Moreover, after being activated by NTR that is overexpressed in hypoxic tumors, the probe exhibited a significant photothermal effect for curative tumor ablation with no recurrence. Conclusions: We have developed the first hypoxia-triggered and NTR enzyme-responsive single molecule probe for high-contrast NIR II/PA tumor imaging and hypoxia-activated photothermal therapy. By tracing the activity of NTR, IR1048-MZ may be a promising contrast agent and theranostic formulation for other hypoxia-related diseases (such as cancer, inflammation, stroke, and cardiac ischemia).

Introduction

Hypoxia is a common characteristic of solid tumors 1-3. This important feature is associated with resistance to radio-chemotherapy, which results in poor prognosis and probability of tumor recurrence 4-12. Therefore, there is a need to monitor tumor hypoxia for assessing the aggressiveness of tumors and predicting the outcome of therapy. Since hypoxia can cause overexpression of nitroreductase (NTR) and NTR level is directly related to the degree of hypoxia 13-14, NTR is considered an indicative biomarker of hypoxia and its level can be used to evaluate the hypoxic degree of a tumor 14-16. Under hypoxic conditions, NTR can catalyze the reduction of nitro compounds to the corresponding amine compounds in the presence of NADH as an electron donor 17-20. Such an enzyme-catalyzed reduction has been recognized to be a useful reaction for the design of hypoxia probes.

There are currently several fluorescence probes for tumor hypoxia detection 21-25. However, all of these fluorescence probes operate in the visible or NIR I region (700-900 nm). Recent studies have demonstrated that the new NIR II window (1000-1700 nm) is advantageous for biological imaging, with improved spatial resolution, a higher signal-to-noise ratio, and deeper tissue penetration depth, owing to reduced photon scattering and autofluorescence in the NIR II range 26-29. But, most studies in the NIR II window have focused on developing new image contrast agents for bioimaging, such as inorganic nanomaterials (single-walled carbon nanotubes, metal chalcogenide QDs, rare-earth nanoparticles), polymer nanoparticles (pDA), and organic small-molecule dyes (e.g., IR1061, CH1055, IR-FE, IR-FEP) 30. To date, only a few studies have focused on NIR II diagnostic sensors, and there are no available NIR II probes for tumor hypoxia detection 31.

Hypoxia has been widely reported to cause chemo- or radiation-resistance, which significantly undermine the efficacy of cancer therapies 4-6, 32-33. So, therapies overcoming cellular resistance during hypoxia are of great importance. Phototherapies, including photodynamic therapy (PDT) and photothermal therapy (PTT), have been introduced into clinical trials for cancer therapy because of their advantages for improving the anti-tumor effect by overcoming multidrug resistance 34-36, with reduced side effects. Actually, several PDT photosensitizers have been approved by the US Food and Drug Administration (FDA) for the treatment of some specific indications 37. Nevertheless, PDT is highly dependent on oxygen, thus it is not effective in the hypoxic environment of cancer. In contrast, PTT, as a new therapeutic technology, is not affected by the hypoxia in tumors 38-40. However, the currently available PTT photosensitizers generally lack selectivity, which leads to indiscriminate heating of both tumor tissue and normal tissues during treatment, possibly causing burn injuries in normal tissues. Therefore, exploring PTT agents that can be activated by the hypoxic tumor microenvironment would be much more attractive.

Herein, we aimed to develop a novel NTR enzyme-responsive single molecule probe for high-contrast NIR II/PA tumor imaging and hypoxia-activated photothermal therapy (Figure ). A simple and well-designed small-molecule IR1048-MZ was synthesized by conjugating a nitro imidazole group (2-(2-nitroimidazolyl)ethylamine, MZ) with an IR-1048 dye in the NIR II window. The electron-withdrawing MZ group induces electron-transfer, which subsequently results in fluorescence quenching and decreases the absorbance of IR-1048 41-43. When the MZ group was reduced by NTR 44-50 to the corresponding amine group, the NIR II fluorescence and photoacoustic PA signal of IR1048-MZ were recovered. Thus, as a NIR II/PA sensor, IR1048-MZ enables the selective detection of local NTR concentrations. Furthermore, the NIR II/PA probe can be used to pinpoint a tumor and monitor tumor hypoxia. Notably, strongly amplified photothermal efficacy was also activated simultaneously by the enhanced absorption in the NIR I region that was triggered by NTR. The imaging and PTT results in mice bearing tumor xenografts demonstrated that IR1048-MZ could be used to visualize tumors with a clear boundary and deep tissue penetration and function as an efficient NTR-activated photosensitizer with minimum damage to normal tissues. This successful hypoxia-triggered NIR II fluorescence/NIR-PA/NIR-PTT turn-on system provides a new strategy for the further design of theranostic probes for the simultaneous activation of multimodality imaging and phototherapy.

Results and Discussion

Synthesis and characterization of IR1048-MZ molecular probe

IR1048-MZ was synthesized according to the routes depicted in Figure . The nuclear magnetic resonance (NMR) spectroscopy and high-resolution mass spectrometry (HRMS) data on the chemical characterization of the intermediates and the desired product are available in Supplementary Material. To demonstrate IR1048-MZH is the product of NTR-catalyzed IR1048-MZ in hypoxic conditions, HRMS and HPLC analyses were performed. Under hypoxia (0% pO2), the HRMS spectrum of the reaction solution of IR1048-MZ with NTR showed a characteristic peak for IR1048-MZH at m/z = 741.3216 [M + H]+ (Figure ). Whereas, under normoxia (20% pO2), the HRMS spectrum of the reaction solution showed no characteristic peak for IR1048-MZH (Figure ). These results verified that NTR catalyzed the reduction of IR1048-MZ to IR1048-MZH under hypoxia. Meanwhile, IR1048-MZH was further verified as a major final product by HPLC analysis. As shown in Figure , standard samples of NADH, NTR, IR1048-MZ, and IR1048-MZH all showed different peaks at 5.36, 8.39, 12.82, and 15.81 min distinctly. When IR1048-MZ was mixed with NTR in the presence of NADH under hypoxia (0% pO2), a new peak at 15.81 min corresponding to IR1048-MZH was also observed, concomitant with a reduction in the peak at 12.82 min representing IR1048-MZ in the HPLC trace (Figure ). While under normoxia (20% pO2), the HPLC spectrum of the reaction solution showed no characteristic peak for IR1048-MZH (Figure ). All of these results confirmed the deduced mechanism of the conversion of IR1048-MZ into IR1048-MZH in the presence of NTR under hypoxia.

Theoretical calculations in Figure demonstrated that the bandgap between the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO) of IR1048-MZ was 2.572. But, the HOMO-LUMO bandgap of IR1048-MZH was 2.236, which was lower than that of IR1048-MZ. These data suggested the electron-transfer process in IR1048-MZH was suppressed by the reduction of IR1048-MZ and resulted in enhanced emission intensity. Subsequently, to verify the sensitivity of IR1048-MZ to NTR, we first analyzed the absorption and fluorescence spectra of IR1048-MZ in the presence and absence of NTR in vitro. The probe showed a very weak absorption in the range of 900-1060 nm. However, after being reduced by NTR, it produced a maximum absorption peak at about 980 nm (Figure ). Moreover, as shown in Figure , probe IR1048-MZ had almost no fluorescence emission at 1046 nm with a low fluorescence quantum yield (ΦF) of 1.88×10-5, but its reaction solution with NTR exhibited as much as a 106.9-fold enhancement at 1046 nm (ΦF = 0.006). The maximum excitation and emission wavelengths of probe IR1048-MZ were 980 nm and 1046 nm, respectively. The calculated Stokes shift between the excitation and emission reached 66 nm. Such a large shift can improve the sensitivity of probe IR1048-MZ, which is a desired characteristic in probe design. Importantly, the introduction of a hydrophilic 2-nitroimidazole into the probe also greatly improved the probe's aqueous solubility, thus make it suitable for biological systems (Figure and Figure ).

With increasing NTR concentration from 0 to 10 μg/mL, the absorption peak (980 nm) intensity of IR1048-MZ increased linearly (Figure ). The fluorescence intensity of IR1048-MZ after reacting with different NTR concentrations is also illustrated in Figure . The increase in NTR concentration resulted in an increase in the fluorescence intensity of the probe at 1046 nm (λex = 980 nm). Moreover, there was a good positive linear relationship between NTR (0-10 μg/mL) and the fluorescence intensity of the probe (F = 18675 + 20631 × [NTR] μg/mL, R = 0.9939). The detection limit value of NTR was found to be 43 ng/mL. The fluorescence kinetic curve of IR1048-MZ and NTR depicted in Figure indicated that higher concentration of NTR resulted in a stronger fluorescence intensity and faster response. The increases in NIR absorption of the probe after reaction with NTR lead to a strong PA signal and high thermal efficacy. The NTR-activated PA imaging and PT imaging in vitro are presented in Figure and Figure , respectively. As the NTR concentration was increased from 0 to 10 μg/mL, the PA signal increased linearly from 121 to 936, and the temperature of the solutions after 980 nm NIR laser irradiation (0.1 W/cm2) for 2 min also increased from 30.4 °C to 57.6 °C. The photothermal conversion efficiency was measured to be 20.2 %, approximately 5.5-fold higher than the value reported for ICG (∼3.1%) 37. Having excellent responsiveness to NTR, 1048-MZ also works steadily in pH 6.0-8.0 (Figure ) or in the presence of various bioactive molecules (Figure ). The high selectivity expands the applications of IR1048-MZ in complex intracellular environments. These in vitro experiments indicated that the NIR II fluorescence, PA, and PT were activated by NTR simultaneously and that IR1048-MZ is a promising theranostic agent that can be used for hypoxia-activated imaging and therapy in vivo.

NIR II fluorescence imaging of tumor hypoxia with IR1048-MZ in vivo

Fluorescence imaging of animals in the NIR II window dramatically benefits from deeper penetration due to minimal autofluorescence and reduced scattering of photons that is inversely proportional to wavelength as ~ λ (w = 0.22-1.68) in biological systems 51-54. In the next step, the feasibility of NIR II fluorescence imaging of tumors with IR1048-MZ was investigated by intravenous injection of IR1048-MZ into BALB/c nude mice bearing subcutaneous A549 tumor, which is known to express NTR under hypoxic conditions 55-56. The NIR II fluorescence signal in the tumor tissue was clearly visible at 10 h post-injection (p.i.) and reached its maximum at about 14 h p.i. Notably, unlike with most of the reported probes, the in vivo NIR II fluorescence imaging of tumor hypoxia with IR1048-MZ did not result in an observable background signal (Figure ). The biodistribution of the probe was studied by measuring the probe content (total content of IR1048-MZ and IR1048-MZH) in different organs at various time points after intravenous injection of IR1048-MZ by inductively coupled plasma mass spectrometry (Figure ). We found that the highest accumulation of the probe occurred in the kidney at 5 h post-injection, indicating that the probe eliminated by renal excretion pathways from the animal body. Ex vivo NIR II fluorescence imaging of dissected tumor and organs obtained from the different groups of A549 tumor-bearing mice was performed after 14 h p.i. to further assess the characteristics of the probe (Figure ). High NIR II fluorescence intensity for IR1048-MZ was observed in the tumor, whereas the organs (heart, lung, liver, spleen, and kidneys) showed no detectable signal. The tumor-to-background ratio at 14 h p.i. was as much as 30 (Figure ), about 3.8 times that of the newly reported NTR-sensitive probe in the NIR I region 56. The specific response to hypoxia and depression of autofluorescence makes IR1048-MZ an excellent background-free hypoxia probe for imaging tumors in mice with high contrast, thus made it superior to most of the reported probes.

3D PA imaging of tumor hypoxia with IR1048-MZ in vivo

PA tomography offers dramatic benefits for imaging in live animals, providing high-resolution images at depths in the cm range, which has been recognized as a promising alternative to X-ray scans and MRI for clinical diagnosis 57-59. It deserves to be mentioned that, for the first time, we realized hypoxic tumor detection by monitoring the 3D PA signal of IR1048-MZ with deep-tissue penetration and imaging in vivo. Hypoxia-activated PA imaging in A549 tumor-bearing BALB/c nude mice was carried out by tail vein injection of IR1048-MZ, and 3D PA signals at the tumor site at the indicated time points were recorded (Figure ). The 3D PA signals in the tumor regions were distinguishable at 10 h post-injection, reached their maximum value at about 14 h, and then began to decrease. This behavior agreed well with the NIR II fluorescence imaging results. Moreover, the penetration depth of the photoacoustic (PA) imaging in the longitudinal section (L) 14 h post-injection was up to 14.6 ± 0.2 mm (Figure ), which demonstrated that hypoxia-activated PA imaging provided deep tissue penetration and high spatial resolution. Considering the results from NIR II fluorescence and PA imaging in tumor-bearing mice, IR1048-MZ definitely serves as a specific hypoxia-activated NIR II/ PA imaging probe for accurate detection of the boundary and depth of tumors.

Hypoxia-activated PTT of tumor with IR1048-MZ in mice

Considering that the NIR absorption of the probe increased significantly in the presence of the overexpressed NTR in hypoxic tumors, we supposed that this single molecule probe, apart from serving as a NIR II/PA imaging probe, could be used as a hypoxia-activated photothermal agent for tumor photothermal ablation in vivo. First, to verify that the photothermal effect was selectively caused by hypoxia in vivo, an inhibition test of hypoxia with dicoumarin (an inhibitor of endogenous nitroreductase) was performed 56, 60. 100 µL dicoumarin solution (0.2 mM) was injected into tumor 1 on the left hind leg while tumor 2 on the right hind leg was untreated, and the mouse was then injected with 200 μL of IR1048-MZ (40 μg/mL) via the tail vein (Figure ). 14 h after the injection, both tumor 1 and tumor 2 were irradiated under 980 nm NIR laser (0.1 W/cm2) for 2 min. Then infrared thermal imaging was performed. As demonstrated in Figure , remarkably, the temperature of tumor 2 increased from 30 °C to 58 °C, which was sufficient to kill cancer cells, but the temperature of tumor 1 was still at 30 °C. Subsequently, to demonstrate that the NTR level in the tumor was down-regulated by the injection of the inhibitor (dicoumarin), a commercial hypoxyprobe (pimonidazole antibody, PIMO) immunofluorescence assay was carried out on tumor slices 61-63. As shown in Figure , the untreated group exhibited strong green fluorescence, which indicated that the untreated tumor was hypoxic with high levels of NTR. But, the group treated with the inhibitor exhibited weak fluorescence, which suggested that the NTR level was successfully down-regulated by the inhibitor. These results verified that the photothernal effect of the probe was selectively activated by hypoxia in vivo, indicating that the hypoxia-triggered photothermal sensitizer, which is specifically activated by tumor, renders the sensitizer maximal therapeutic efficacy and minimal side effects to normal tissues.

Next, the treatment outcome of hypoxia-activated PTT with IR1048-MZ was evaluated in A549 tumor-bearing mice. As shown in the infrared thermal imaging results (Figure ), for the mice treated with IR1048-MZ and 980 nm laser irradiation at a low irradiance of 0.1 W/cm2, their surface temperature increased quickly from ∼30 °C to ∼57 °C within 2 min. Conversely, the temperatures of tumors in the other groups without the hypoxia-dependent NIR absorbance under the same laser irradiation displayed little change. At 5 d after the PTT treatment, the tumors in the IR1048-MZ plus NIR laser irradiation group disappeared, and no tumor relapse was found from day 5 to day 30. In contrast, the tumors in the other groups showed rapid growth (Figure ). In addition, the survival rate of the mice treated with IR1048-MZ with NIR laser irradiation was 100% after 30 d, which was a significant improvement compared with those of the other groups (Figure ). The body weights of these experimental groups were not significantly different from that of the control group, indicating that the treatments were reasonably well-tolerated (Figure ). Additionally, the major blood indexes and liver-function indexes of mice injected with IR1048-MZ (Table ) show no obvious abnormity or difference when compared with those of mice in the control group, further demonstrating the high biocompatibility of IR1048-MZ. Furthermore, tumor sections from the different groups were examined with a hematoxylin and eosin (H&E) staining assay. In Figure , a morphology characteristic of apoptotic cells was observed in the H&E-stained tumors treated with IR1048-MZ and NIR laser irradiation, but this morphology was not observed in the other three groups. These results demonstrated that IR1048-MZ is safe and has great potential for use in hypoxia-activated PTT of tumors with high efficiency in vivo.

Conclusions

In summary, we have developed the first hypoxia-triggered and NTR enzyme-responsive single molecule probe for high-contrast NIR II/PA tumor imaging and hypoxia-activated photothermal therapy. The IR1048-MZ molecular probe itself has a very weak NIR II and PA signal. However, after activation in hypoxic tumors, the probe emits intense NIR II fluorescence and PA signal, which provided more accurate and deeper tissue imaging of the position and boundary of tumors. Importantly, the hypoxia-activated photothermal therapy of IR1048-MZ exhibited outstanding phototherapy efficacy with rapid increases in temperature and resulted in ablation of tumors with no recurrence. Therefore, the versatile IR1048-MZ probe was not only an activable NIR II/PA imaging probe with rapid, selective, and sensitive hypoxia response, but also a robust small-molecule photosensitizer for photothermal therapy. By tracing the activity of NTR, IR1048-MZ may be a promising contrast agent and theranostic formulation for hypoxia-related diseases (such as cancer, inflammation, stroke, and cardiac ischemia).

Supplementary figures and tables.

Click here for additional data file.