PEGylated Terbium-Based Nanorods as Multimodal Bioimaging Contrast Agents

: Diagnostic imaging strongly relies on the use of contrast agents (CAs). In general terms, current CAs present undesirable side e ﬀ ects that encourage researchers and pharmaceutical companies to continually search for safer and more versatile alternatives. Here, we describe the synthesis and characterization of terbium-based nanorods (TbNRs) as a potential alternative to traditional CAs for magnetic resonance imaging (MRI) and X-ray computed tomography (CT). The paramagnetism and high atomic number of Tb provide TbNRs with both magnetic relaxivity and X-ray attenuation capabilities. After surface functionalization with a polyethylene glycol (PEG)- derived ligand, TbNRs showed high colloidal stability in physiological media. Additionally, toxicity studies conducted in cell cultures and zebra ﬁ sh embryos demonstrated the safety of the as-synthesized TbNRs, thus supporting their potential use as CAs. Lastly, in vivo imaging experiments in mice demonstrated that TbNRs produce remarkable contrast enhancement on both MRI and CT.


■ INTRODUCTION
Imaging-based diagnosis has long since become an essential tool for daily clinical practice.Among the different in vivo imaging modalities, MRI and CT stand out for their ability to render detailed 3D anatomical information along with functional information, especially in the case of MRI. 1 Although the use of contrast agents (CAs) is often unnecessary, in certain situations, they are essential for the proper visualization of a particular region of interest or to get relevant functional information, such as perfusion or diffusion parameters. 2 The most commonly used intravenous CAs are gadolinium-and iodine-based compounds for MRI and CT, respectively. 3,4Gd chelates are routinely used as MRI CAs to generate positive contrast in T 1 -weighted images as a result of shortening the longitudinal relaxation time (T 1 ) produced by the Gd cations on the protons of the nearby water molecules.As for iodine derivatives, they are used as CT CAs owing to the high atomic number of iodine (Z = 53), which strongly correlates with the X-ray attenuation coefficient.Image contrast in CT is produced by differences in the X-ray attenuation of tissues and, therefore, the presence of iodine derivatives will potentiate image contrast by increasing the attenuation coefficient of the tissues where they accumulate. 5lthough Gd chelates and iodinated compounds are still widely used at present, they have some important limitations.
−11 On the other hand, both types of CAs exhibit short blood half-lives since they are rapidly excreted by renal filtration due to their small size and, therefore, their use is limited to applications in which long circulation times are not required.
Current advances in medical imaging aim to increase diagnostic accuracy through different strategies, such as multimodal imaging evaluations.In this scenario, nanomedicine plays a key role, thanks to the enormous possibilities it offers to produce new nanoscale materials with improved properties and functionalities. 12Thus, numerous new nanoscale CAs with great potential to improve the diagnostic accuracy of MRI and CT have been produced over the past few years. 13,14Among them, nanostructures based on rare earth (RE) elements have attracted a great deal of attention because of their unique physicochemical properties, which provide them with great potential for bioimaging applications. 15−24 Likewise, lanthanide-based nanostructures are also suitable for CT imaging, thanks to the high atomic number (Z) of these elements, which allows them to efficiently absorb Xrays. 25,26The vast majority of these studies use the lanthanide as a dopant agent and not as the main component of the nanostructured material.Yet, other RE-based structures, mostly lanthanide oxide nanoparticles (NPs), have also been explored as CAs for in vivo imaging.−30 Regarding Tb-oxide NPs, Marasini et al. 31 recently reported the synthesis of ultrasmall paramagnetic NPs based on several lanthanide oxides, including Tb oxide.This work shows the potential of these NPs as MRI CAs, although no in vivo evidence has been reported so far.Indeed, it is worth noting here that a key aspect lacking in many of these studies is detailed in vivo characterization that demonstrates the real potential of these nanostructures as CAs.In addition to high biocompatibility, NP-based CAs must be able to escape the mononuclear phagocyte system and remain in circulation long enough to efficiently reach their target, especially in tumor imaging.However, many of the RE-based CAs, particularly those that use RE elements as the doping agent, exhibit large hydrodynamic diameters (HDs), usually above 100 nm, which in vivo lead to rapid liver uptake, limiting their applicability as CAs. 23,29,32On the other hand, long circulation times are determined not only by the size of the NPs but also by their coating.In this regard, stealth polymers such as polyethylene glycol (PEG) have been shown to efficiently prolong blood circulation times. 33Here, we report the synthesis of terbiumbased nanorods (TbNRs) with sizes of 2 × 9 nm with high colloidal stability and excellent luminescent, magnetic, and Xray attenuation properties.These TbNRs were well dispersed in physiological medium by a simple ligand exchange with a catechol-derived PEG ligand.The cytotoxicity of the resulting PEGylated TbNRs (PEG−TbNRs) was assessed in vitro using different assays, including MTT, live-dead, and flow cytometry, and in vivo using zebrafish embryos.Then, in vivo imaging studies were conducted in mice to determine the suitability of PEG−TbNRs as CT and MRI CAs.
Synthesis of TbNRs.Synthesis of Tb Oleate.For the synthesis of Tb oleate, we used an adapted version of the protocol reported by Liu et al. 24 In brief, a mixture of 1.87 g of terbium chloride (7.05 mmol) and 4.75 g of sodium oleate (15.60 mmol) was dissolved in 7 mL of distilled water, 10 mL of ethanol, and 17 mL of hexane.This solution was refluxed for 4 h under an inert atmosphere, cooled to room temperature and transferred into a separation funnel.The organic layer, which contains the desired terbium oleate, was washed several times using distilled water.The hexane was evaporated using a rotavapor.
Synthesis of TbNRs.A total of 0.511 g (0.51 mmol) of terbium oleate, along with 0.250 g (0.89 mmol) of oleic acid and 3.5 mL of 1octadecene was heated at 320 °C for 1 h in an inert atmosphere (Scheme 1).Then, the mixture was cooled down to room temperature and washed several times with a mixture of ethanol and acetone (1:1) as precipitating agents, and subsequently centrifuged (10 min at 5.000 rpm).TbNRs were then resuspended in toluene.
Functionalization of TbNRs.The detailed procedure concerning the ligand synthesis is described in the Supporting Information.In brief, a solution containing 1.0 mL of TbNRs (10 g/L), 1.0 mL of the ligand GA-PEG3000-OH (0.1 M in CHCl 3 ), and 50 μL of triethylamine was put in a separating funnel.This mixture was shaken gently and then diluted by adding 5 mL of toluene, 10 mL of acetone, and 5 mL of Milli-Q water.Subsequently, this mixture was shaken, the aqueous layer collected in a round-bottom flask, and the remaining organic solvents rotaevaporated.The resulting watersoluble PEG−TbNRs were purified using 100 kDa cut-off centrifuge filters at 450 rcf until the free ligand was completely removed.Then, the PEG−TbNRs were diluted in the corresponding media and centrifuged during 5 min at 150 rcf.
Characterization.Transmission Electron Microscopy.Transmission electron microscopy (TEM) and scanning-TEM (STEM) images were obtained on a FEI TALOS F200 (FEI, Hillsboro, OR, USA) operating at 200 kV accelerating voltage and equipped with Super-X energy-dispersive X-ray spectrometry (EDX).Compositional analyses of the samples were performed by the combination of highangle annular dark-field images (HAADF) and EDX acquisitions in TEM and STEM modes (0.2 s dwell time and spatial drift correction).Sample preparation consisted of dropwise adding TbNR solution (∼1 g/L of Tb) onto a carbon-coated copper grid and letting it dry.The length and diameter of TbNRs were calculated as the average of a hundred TbNRs measured.

ACS Applied Nano Materials
UV−vis Spectroscopy.The UV−vis spectra were registered on a Cary 100 (Agilent, Waltham, MA, U.SA) spectrometer using a quartz tray with 1 cm light path.
Fluorescence Spectroscopy.The fluorescence spectra were recorded with an Edinburgh Instruments FLS920 (Edinburgh, U.K) spectrometer with a quartz tray (light path = 1 cm).
Vibrating Sample Magnetometer.The analysis of the magnetization behavior was performed on a vibrating sample magnetometer (VSM) (MicroMag 3900, Princeton Measurements Corp., U.S.A) using magnetic fields ranging from −1 to 1 T at 5 and 300 K.
Dynamic Light Scattering.PEG−TbNR size distribution was analyzed using a Zetasizer Nano ZS90 (Malvern Instruments, Malvern, Worcestershire, U.K). PEG−TbNRs were dispersed in PBS or culture medium at a final concentration of 50 mg/L(Tb).ZEN0118-low-volume disposable sizing cuvettes were used for these measurements; 2.420 was set as the refractive index with an angle of detection of 173°backscatter (NIBS default).The measurement time was determined automatically.Three measurements were performed per sample.The size distribution was determined using the number mean.The analysis was done according to the general-purpose (normal resolution) model.
Fourier Transform Infrared Spectroscopy.Fourier transform infrared spectroscopy (FTIR) spectroscopy was performed on a JASCO FTIR-4100 equipped with an ATR accessory (MIRacle ATR, PIKE Technologies) coupled to a mercury cadmium telluride (MCT) detector cooled with liquid nitrogen.Spectra acquisition was performed from 4000 to 800 cm −1 (4 cm −1 resolution, 50 scans).
Nuclear Magnetic Resonance Spectroscopy.The 1 H NMR spectrum of the ligand prepared in CDCl 3 was acquired on a Bruker 400 MHz nuclear magnetic resonance (NMR) system (Bruker BioSpin, Rheinstetten, Germany).
Transversal Relaxivities (r 2 ).r 2 was measured in PBS at 1. Transverse relaxivity, r 2 , was determined as the slope of the linear fit of 1/T 2 over [Tb].
Inductively Coupled Plasma High-Resolution Mass Spectroscopy.Tb concentration was determined on a NexION ICP−HRMS (PerkinElmer, Waltham, MA, U.SA).In brief, 25 μL of the TbNR solution was diluted with 2.5 mM aqua regia.This mixture was left overnight and then diluted up to 25 mL with water.
Cytotoxicity.The cytotoxicity of PEG−TbNRs was assessed in N13 cells by the MTT assay (details are given in the Supporting Information).
Cell Morphology Studies.These studies were performed using a PerkinElmer operetta high content imaging system, which allowed us to evaluate cell morphology and calculate the percentage of cell death.Further details are given in the Supporting Information.
Teratogenicity Assay.Wild-type zebrafish were used for these experiments.At day 0, they were outcrossed, and then, the embryos were collected and incubated in E3 medium for 4 h at 28 °C.Unfertilized eggs were discarded.Those embryos that developed properly were exposed to different concentrations of PEG−TbNRs in E3 medium.Eggs were deposited in eight-well Petri dishes (∼30 eggs in 4 mL).The hatching rate, malformations, and survival were analyzed at 24, 48, and 96 h of PEG−TbNRs exposure.
Animals.BALB/c mice (male, ca.anesthetized (1% isoflurane), and a catheter was introduced into their tail vein for intravenous administration of the PEG−TbNRs.Then, the animals were transferred to the imaging system (CT or MRI).The PEG−TbNRs were administered intravenously (10 mg of Tb per kg).
In Vitro and In Vivo CT.CT images were obtained on a Bruker Albira system (Bruker Biospec, Bruker BioSpin, Ettlingen, Germany) using the following parameters: 35 μm X-ray focal spot size (nominal), 45 kVp energy, working at 400 μA.Attenuation values were expressed in Hounsfield units, with the scale adjusted to better visualize soft tissues, overexposing the bones.Liver and kidney pharmacokinetics were calculated from the average values within the same regions of interest (ROIs) at different experimental times.
In Vivo MRI.MRI studies were performed on the 9.4 T Bruker Biospec system described in the Transversal relaxivities section by following the protocol described by us elsewhere. 34In brief, highresolution T 2 -weighted images and a sequence of fast T 2 -weighted images were acquired and analyzed to characterize the short-term pharmacokinetics (first 30 min).Then, quantitative T 2 maps were acquired at 0, 1, 24, 48, and 168 h to evaluate long-term biodistribution and pharmacokinetics.
Histology.At 168 h postinjection of the PEG−TbNRs, mice were euthanized and the kidneys, spleen, and liver were collected.Tissue architecture was assessed by light microscopy of hematoxylin and eosin (H&E)-stained tissue sections.Details are provided in the Supporting Information.
Statistical Analysis.The SPSS package (SPSS Inc., Chicago, Illinois) was used for the statistical analysis.Differences between different experimental conditions were determined by student's t-test or one-way analysis of variance (ANOVA), with a significance level of 0.05.

■ RESULTS AND DISCUSSION
Synthesis, Characterization, and Functionalization of TbNRs.TbNRs were synthesized by thermal decomposition of terbium−oleate in the presence of oleic acid in 1-octadecene.The morphology and size of the synthesized TbNRs were examined by TEM and EDX.The as-prepared particles showed an elongated shape of around 9.0 nm in length and 2.1 nm in width.Interestingly, these elongated particles were organized in elongated superstructures of around 80.2 nm in length and 9.2 nm in width (Figures 1a and S1).The formation of these assemblies of NPs occurred during the synthesis process, likely due to the different reactivity along the rods.These assemblies of oleic acid-capped NPs were ligand exchanged by catecholderived PEG ligands, resulting in water-soluble assemblies with a structure similar to that of the as-synthesized TbNRs, as shown in the TEM images.Therefore, the gentle ligand exchange procedure was not sufficient to disassemble the NPs, probably due to the high energy of interaction between them.
The as-prepared TbNRs were capped by oleic acid surfactants, giving a hydrophobic character to the nanomaterial.Thus, to make them biocompatible, TbNRs were transferred to aqueous media by following a previously reported ligand exchange method, 35 using a catechol-derived PEG−OH ligand.As previously shown, catechol anchors strongly interact with the iron oxide NP surface via their phenolic hydroxyl groups. 36These anchors form a stable 5-membered ring complex with the Fe cation.Similarly, lanthanide catecholate complexes are likely formed from trivalent lanthanide cations and catechol-derived ligands.Then, the bi-dentate catechol anchor of the ligand would react with the Tb at the surface of the NRs forming a stable complex similar to the terbium− catecholate complex [Tb 2 (cat) 3 ] produced in solution. 37,38haracterization of PEGylated TbNRs.PEG−TbNRs were thoroughly characterized by different physicochemical techniques to determine their morphology, size, optical properties, magnetic properties, and colloidal stability.The original morphology and sizes of the pristine NRs were preserved after ligand exchange, as confirmed by TEM and STEM (Figures 1c,d and S1).Moreover, the attachment of the PEGylated ligand to the surface of the TbNRs was confirmed by FTIR spectroscopy (Figure S2).Additionally, these functionalized TbNRs were very stable in water without any sign of precipitation over time, further confirming the success in incorporating the PEGylated ligands on the surface of the TbNRs.Further analysis by dynamic light scattering (DLS) showed the high stability of PEG−TbNRs in physiological media, specifically in PBS and cell culture medium.PEG− TbNRs showed HDs around 75 nm in PBS, and slightly lower (∼60 nm) in cell culture medium.After 24 h, the HDs of PEG−TbNRs increased by around 30% and then remained stable throughout the measurement time (1 week).This behavior demonstrates the stability of PEG−TbNRs in physiological media and also that no aggregates were formed during ligand exchange (Figure S3).
REs also show interesting optical properties, which are governed by the electron configuration and can be characterized by UV−vis and photoluminescence (PL) spectroscopy.Thus, the optical absorption in the UV region of REs can originate from transitions 4f → 5d.Therefore, the peak observed at 277 nm in the UV−vis spectrum could be assigned to the 4f 8 → 4f 7 5d transition, which normally appears at more than 200 nm in the case of Tb ions (Figure 2). 39The PL emission spectrum of the TbNRs dispersed in toluene at room temperature is shown (Figure 2).The emission spectrum of TbNRs under 325 nm excitation displayed four emission peaks, which could be assigned to the electron transitions from 5D 4 → 7F 6 (490 nm), 5D 4 → 7F 5 (546 nm), 5D 4 → 7F 4 (583 nm), and 5D 4 → 7F 3 (620 nm).Among these transitions, the green emission (546 nm) appeared as the strongest, which is in agreement with the Judd−Ofelt theory. 24dditionally, PEG−TbNRs were evaluated as potential CAs for in vivo imaging, specifically for CT and MRI.The ability of PEG−TbNRs to be used as CT CA was determined by measuring the X-ray attenuation.Interestingly, the slope of the linear fit of the X-ray attenuation versus the PEG−TbNRs concentration was 21.4 HU mM −1 [Tb], which is 10-fold higher than that obtained for the commercial CA iohexol (2.69 HU mM −1 ) (Figure 3b). 40On the other hand, the magnetic properties of PEG−TbNRs were studied to evaluate their potential as MRI CAs.First, the magnetic hysteresis loop measurements by the VSM showed that PEG−TbNRs exhibit superparamagnetic behavior at 5 K and paramagnetic at 300 K (Figure 3a).Then, their magnetic transverse relaxivity, r 2 , was determined at 1.44 and 9.4 T. PEG−TbNRs showed r 2 values of 10.4 mM −1 •s −1 and 48.5 mM −1 •s −1 (Figure 3c) at low and high magnetic fields, respectively.Therefore, we can conclude that our PEG−TbNRs behave as multimodal imaging nanosystems for optical imaging, CT, and MRI.However, the emission wavelength of these PEG−TbNRs, around 540 nm, limits their applicability in vivo in optical imaging due to the high tissue autofluorescence at that wavelength, 41 together with weak tissue penetration.On the contrary, their excellent X-ray attenuation and magnetic relaxation properties make them very promising CT and MRI CAs for bioimaging.
In Vitro Cytotoxicity Assessment.The evaluation of the cytotoxicity of new nanomaterials designed for biological applications is a crucial step that requires an exhaustive analysis.Therefore, we carried out a detailed analysis including multiple assays, namely, cell morphology, live-dead, and MTT assays.These assays were performed on the mouse microglial cell line, N13.After exposure to growing concentrations of PEG−TbNRs, from 0.1 to 100 μg/mL of Tb, N13 cells did not show evidence of morphological changes, even for the highest concentration evaluated (Figures 4a−c and S4).Furthermore, the total number of cells per well remained unaltered for all the concentrations tested (Figures 4d and S5).Indeed, the percentage of dead cells remained around 0%, without a statistically significant increase in all cases (Figures 4e and S5).Finally, no statistically significant effect on mitochondrial activity was observed with the MTT assay (Figure 4f).These results demonstrate that PEG−TbNRs exhibit excellent cellular biocompatibility, supporting their potential for in vivo applications.However, in vivo toxicity depends on many different factors that cannot be evaluated in cell cultures.Therefore, before proceeding with in vivo experiments on mice, toxicity was also assessed in vivo using zebrafish embryos.Teratogenicity Assessment on Zebrafish Embryos.Teratogenicity of PEG−TbNRs was evaluated in zebrafish embryos, as previously described. 42The survival and hatching rates of zebrafish embryos exposed to growing concentrations of PEG−TbNRs (0.1, 1, 10, 50, and 100 μg/mL of Tb) were assessed at different time points after fertilization.After 24 h postexposure, PEG−TbNRs were clearly observed to be interacting with the chorion of the embryos (Figure S6a).Regarding the hatching process, both control embryos and those exposed for 48 h to PEG−TbNRs showed 100% hatching (Figure S6b).These results agree with hatching rates reported for normal zebrafish embryos. 43In addition, neither mortality nor malformations were observed at any time point in the embryos exposed to PEG−TbNRs (Figures 5 and S7).
In Vivo CT and MRI Studies.CT imaging studies showed an increment of 5.5 ± 2.1 HU in the liver 1 h postadministration of PEG−TbNRs (10 mg of Tb per kg), which increased up to 12.1 ± 3.1 after 24 h (Figure 6).Then, the ΔHU values decreased, being 10.4 at 24 h and 6.1 at 168 h.Therefore, the maximal accumulation was detected at 24 h, and then, PEG−TbNRs were slowly cleared from the liver.Importantly, these results were obtained with an injection dose approximately 50 times lower than that recommended for iohexol, which is 430 mg per kilogram for body imaging applications. 44n vivo biodistribution and pharmacokinetics of the PEG− TbNRs were also evaluated by MRI.Short-term pharmacokinetics by dynamic T 2 imaging showed that the PEG−TbNRs were rapidly taken up by the liver, reaching a value of 50% of relative enhancement during the first 30 min after the intravenous administration of PEG−TbNRs (10 mg of Tb) (Figure 7a,b).
No significant accumulation was observed in the rest of the organs.Moreover, long-term pharmacokinetics showed a maximal T 2 decrease (ΔT 2 of 7.1 ms) at 24 h after administration, in agreement with the CT results shown above.A similar trend was observed in the kidneys, where the maximum T 2 decrease (ΔT 2 of 5.9 ms) was reached at 24 h.After 48 h, the kidneys recovered their initial T 2 values, whereas the liver only showed a slight T 2 recovery, in agreement with the slow liver clearance described in the CT studies (Figure 7c).
Quantitative Evaluation of Circulating Lifetime of PEG−TbNRs in the Bloodstream.Tb content in blood and the liver at 1 and 24 h was quantified by ICP−MS (Figure 8).The concentration of Tb in blood was 2.3 and 1.4 mg/L at 1 and 24 h postadministration, which corresponds to 10 and 6% of the injected dose, respectively.These results confirmed the long blood half-life of the PEG−TbNRs.As expected, an opposite trend was observed in the liver, showing 16.2 mg/kg of tissue at 1 h and 34.5 mg/kg of tissue at 24 h, respectively.Thus, the ICP results are in very good agreement with the CT and MRI results, demonstrating that PEG−TbNRs have long blood residence time, being still present at 24 h, and that they undergo slow hepatic clearance.Blood residence time is of utmost importance for some in vivo applications, in particular for tumor targeting.Efficient tumor targeting, whether passive, via enhanced permeability and retention (EPR) effect, or active, has been shown to require long circulation times. 45herefore, the ICP−MS results, which are in good agreement with the MRI findings and with the long-circulating nature of our PEG−TbNRs, further support the great potential of these nanosystems for in vivo applications, specifically for cancer imaging.
In Vivo Toxicity Assessment.Finally, potential harm caused by the intravenous administration of PEG−TbNRs was assessed histologically and also by following the weight profile of the animals.H&E staining was carried out on tissue samples of the liver, kidneys, and spleen at 168 h postadministration of PEG−TbNRs.The liver sections of the animals injected with PEG−TbNRs resembled those injected with PBS.No evidence of acute and subacute liver injury, such as vacuolated swelling of the hepatocytes' cytoplasm, was observed in the liver sections (Figure S8).Likewise, no substantial changes were observed in the kidneys of the animals injected with PEG− TbNRs when compared to control animals.Kidneys exhibited regular intact brush borders and glomeruli around the Bowman's capsule (Figure S8). 46Finally, the spleen also presented normal histological features with unaltered red and white pulp, which is characteristic of their normal architecture.Regarding the weight profile, a transient decrease in body weight was observed, with a maximum difference between groups at 24 h after the administration of PEG−TbNRs.However, after 168 h, the animals regained their normal weight, indicating the absence of toxic effects (Figure S9).

■ CONCLUSIONS
We report the synthesis of PEGylated TbNRs with potential application as multimodal imaging CAs.TbNRs were synthesized by thermal decomposition of the as-prepared Tb−oleate.Then, TbNRs were successfully functionalized using a catechol-derived PEG ligand to render water-stable TbNRs.These PEG−TbNRs showed high colloidal stability in different media, absence of toxicity in vitro in cell cultures and in vivo in zebrafish embryos, and excellent properties as CT and MRI CAs.On the other hand, after intravenous injection in mice, the PEG−TbNRs showed excellent in vivo imaging contrast both in CT and MRI, long blood residence time, being still present after 24 h, and effective hepatic clearance.Altogether, these results demonstrate the high potential of PEG−TbNRs as in vivo multimodal CAs.
22 g weight, n = 3), provided by Janvier Labs (Le Genest-Saint-Isle, France), were used in all in vivo imaging experiments.The experiments were carried out in accordance with our local ethics committee guidelines and were consistent with the national regulations for the use and care of experimental animals (R.D. 53/2013).Prior to imaging experiments, mice were

Figure 1 .
Figure 1.TEM (a) and EDX (b) images of TbNRs before functionalization.TEM (c) and STEM (d) images of PEGylated TbNRs.Inset shows a higher-magnification image (scale bar corresponds to 10 nm).

Figure 3 .
Figure 3. (a) hysteresis loop of PEG-TbNRs measured at 5 K (black) and 300 K (blue); (b) X-ray attenuation measured as the linear relationship between the intensity of the CT images and the concentration of Tb in the PEG−TbNR solutions; (b); (c) relaxation rate (1/T 2 ) vs the concentration of Tb in the PEG−TbNR solutions measured at 1.44 T (black dots) and 9.4 T (red dots), and the linear fits (black and red lines) whose slopes correspond to the transverse relaxivities (r 2 ) of PEG−TbNRs at both magnetic fields.

Figure 4 .
Figure 4. Optical microscopy images of N13 cells: (a) negative control, (b) positive control, (c) cells exposed PEG−TbNRs (100 μg/mL).Images correspond to the merging of TO-PRO-3 iodine (red), DAPI (blue), and bright field (gray).Scale bars correspond to 100 μm.Bar charts correspond to (d) total number of cells/well, (e) percentage of dead cells, and (f) cell viability (MTT assay) after exposure to increasing concentration of PEG−TbNRs for 24 h.

Figure 5 .
Figure 5. Morphological analysis of zebrafish embryos exposed to growing concentrations of PEG−TbNRs at different time points (24, 48, and 96 h).

Figure 6 .
Figure 6.Representative in vivo CT images at different time points (0, 1, 24, 48, and 168 h) after the injection of PEG−TbNRs.The liver has been highlighted for better visualization.Values are in Hounsfield units and the scale adjusted for better visualization of soft tissues, overexposing the bones.

Figure 7 .
Figure 7.In vivo pharmacokinetics of intravenously injected PEG−TbNRs in BALB/c mice: (a) short-term pharmacokinetics in the liver; (b) short-term pharmacokinetics in the kidneys; (c) T 2 -weighted MR images at different times postadministration of PEG-TbNRs; (d) T 2 variation in different tissues determined from in vivo T 2 maps (values correspond to average T 2 and ΔT 2 from 3 animals).

Figure 8 .
Figure 8. ICP−MS quantification of Tb content at 1 and 24 h post intravenous injection of PEG−TbNRs.(a) Absolute Tb concentration, (b) relative Tb concentration (% to the injected dose).Error bars correspond to standard deviation (n = 3).