Human Aortic Endothelial Cell Labeling with Positive Contrast Gadolinium Oxide Nanoparticles for Cellular Magnetic Resonance Imaging at 7 Tesla

Abstract

Positive T₁ contrast using gadolinium (Gd) contrast agents can potentially improve detection of labeled cells on magnetic resonance imaging (MRI). Recently, gadolinium oxide (Gd₂O₃) nanoparticles have shown promise as a sensitive T₁ agent for cell labeling at clinical field strengths compared to conventional Gd chelates. The objective of this study was to investigate Gado CELLTrack, a commercially available Gd₂O₃ nanoparticle, for cell labeling and MRI at 7 T. Relaxivity measurements yielded r₁ = 4.7 s⁻¹ mM⁻¹ and r₂/r₁ = 6.2. Human aortic endothelial cells were labeled with Gd₂O₃ at various concentrations and underwent MRI from 1 to 7 days postlabeling. The magnetic resonance relaxation times T₁ and T₂ of labeled cell pellets were measured. Cellular contrast agent uptake was quantified by inductively coupled plasma–atomic emission spectroscopy, which showed very high uptake compared to conventional Gd compounds. MRI demonstrated significant positive T₁ contrast and stable labeling on cells. Enhancement was optimal at low Gd concentrations, attained in the 0.02 to 0.1 mM incubation concentration range (corresponding cell uptake was 7.26 to 34.1 pg Gd/cell). Cell viability and proliferation were unaffected at the concentrations tested and up to at least 3 days postlabeling. Gd₂O₃ is a promising sensitive and stable positive contrast agent for cellular MRI at 7 T.

MAGNETIC RESONANCE IMAGING (MRI) has emerged over the past decade as a promising modality for cellular imaging owing to its high spatial resolution and exquisite soft tissue contrast. The use of contrast agents has been instrumental in this development as contrast-labeled cells can be visualized, thus allowing their biodistribution to be tracked and monitored in vivo. To date, cellular MRI studies have been performed in various experimental systems, including neurodegenerative disorders and stroke, tumors, myocardial infarction, transplantation, and tissue engineering,^1–5 to improve our understanding of normal and pathologic processes at the cellular level and to develop new cellular therapies.^6,7 The majority of these studies use superparamagnetic iron oxide nanoparticles to achieve sensitive cell detection as iron oxides are readily taken up by cells and provide a strong negative contrast on T₂ or T*-weighted images. However, negative agents tend to obliterate surrounding anatomy because they exert a susceptibility effect that extends well beyond the physical boundaries of labeled cells. Specific distinction of labeled cells is also difficult owing to negative contrast arising from natural sources, such as hemorrhage, susceptibility gradients around normal biologic structures, and background tissue with intrinsically low signal intensity. Owing to these fundamental drawbacks of negative contrast agents, the development of better agents that provide high sensitivity is a critical area of current research for cellular MRI applications. A promising development is positive contrast agents that provide signal enhancement and potentially more specific identification and better anatomic delineation of labeled cells.

Positive contrast, or enhanced signal from labeled cells on T₁-weighted images, can be obtained from contrast agents containing the paramagnetic gadolinium (Gd) ion. This approach was initially investigated with the clinically approved chelated form (Gd-DTPA), but its main limitation was low sensitivity owing to poor uptake and retention in cells, with most reports showing modest T₁ enhancement to none or diminished T₁-reducing effect intracellularly. More recently, several Gd-based agents have demonstrated positive contrast visualization,^8–13 either through higher relaxivities (eg, Gd-fullerenol,⁸ Gadofluorine M⁹) or higher cell uptake of contrast agents (eg, Gd hexanedione with a threefold uptake over Gd-DTPA.¹⁰ A particularly promising class of Gd-based agents for cellular MRI is gadolinium oxide (Gd₂O₃) nanoparticles.^14–20 These small nanoparticles can be efficiently ingested by cells and contain a high fraction of superficial Gd, thus offering key advantages over Gd chelates. One advantage is better cell uptake and retention, which have been demonstrated in several cell lines, most of which are phagocytic,^15,16 and recently in a glioma cancer model.¹⁷ Another advantage is a considerably higher relaxivity demonstrated at clinical field strengths, which enables a greater positive contrast effect and more sensitive visualization of labeled cells.

At higher fields used for preclinical imaging studies, achieving positive contrast is even more challenging owing to a decline in relaxivity of Gd compounds with increasing field strengths.²¹ Very few Gd agents have been reported for high-field cellular MRI,^8,12 and the performance of Gd₂O₃ has not been evaluated. In this study, the potential of Gd₂O₃ for positive contrast cellular MRI was investigated at 7 T on endothelial cells. Gado CELLTrack (BioPhysics Assay Laboratory [BioPAL], Worcester, MA), a commercially available 30 nm Gd₂O₃ colloid, was used to label human aortic endothelial cells, which are less phagocytic compared to cell lines in other Gd₂O₃ studies. Endothelial cells play a central role in angiogenesis, atherosclerosis, and cancer and are, hence, an important element in new cellular and revascularization therapies.^22,23 To date, the few studies on labeling endothelial cells have employed exclusively iron oxides for their visualization on MRI.^24,25 To our knowledge, this is the first application of Gd₂O₃ for labeling endothelial cells and, more importantly, the first high-field MRI study of Gd₂O₃ at 7 T. Cell uptake and visualization on MRI were investigated, as well as the effects on MR relaxation times and cell function. The goal was to establish Gd₂O₃ for high-field positive contrast cellular MRI and evaluate its potential for labeling endothelial cells in cardiovascular applications.

Materials and Methods

Cell Culture and Immunofluorescence

Normal human aortic endothelial cells (FC-0014) were obtained from Lifeline Cell Technology (Walkersville, MD). They were maintained in basal medium complete with appropriate supplements (VascuLife EnGS Medium Complete, LL-0002, Lifeline Cell Technology) and 1% penicillin-streptomycin. Fresh medium (13 mL) was added every 2 days, and cells were passaged 1:3 at 80% confluency in T75 flasks (BD Biosciences, Mississauga, ON) and incubated at 37°C, 5% CO₂. Before passaging, cells were washed twice with 5 mL of sterile phosphate-buffered saline (PBS) and incubated with 3 mL 0.25% trypsin–0.1 mM ethylenediaminetetraacetic acid (EDTA) for 3 minutes. Then 7 mL of medium with fetal bovine serum (FBS) was added to the cells. The entire 10 mL was collected in 15 mL tubes and spun at 1,300 rpm for 5 minutes. The liquid above the cell pellet was aspirated, and the cells were resuspended in fresh medium, passaged, and maintained according to company instructions.

Cell phenotype was confirmed by visualizing the expression of endothelial cell–specific von Willebrand factor (vWF) on immunofluorescence. Cells were grown on sterile coverslips for 3 days to allow for adhesion and proliferation. Samples were then fixed in 5% formalin in PBS for 15 minutes at room temperature (RT). After washing twice with 1× PBS, cells were permeabilized with ice-cold methanol for 10 minutes at −20°C to target internalized protein markers. Excess methanol was washed off twice with 1× PBS and blocked with 4% bovine serum albumin and PBS + 0.05% Tween (PBST) at RT for 1 hour. For the primary antibody, endothelial cells were incubated with mouse monoclonal vWF IgG1 antibody (1:300; Abcam, ab68545, Cambridge, MA) for 2 hours at RT and then washed five times (5 minutes/wash) with 1× PBST. For the secondary antibody, cells were incubated with polyclonal goat-antimouse IgG-H&L (Texas red) (1:400; Abcam, ab6787) for 1 hour at RT, and excess antibody was washed off five times with 1× PBST (5 minutes/wash). Cells were counterstained with VECTASHIELD Mounting Medium with DAPI for 1 minute (H-1200, Vector Laboratories, Burlingame, CA) and mounted onto slides.

Contrast Agent and Cell Labeling

Gado CELLTrack (–) (CL-50P02-6) was purchased from BioPAL, Inc. for cell labeling. This nanoparticle has been highly dialyzed and consists of Gd₂O₃ and a polymer coating of dextran. The entire complex retains a negative charge and was mixed with a cationic amino acid polymer, poly-L-lysine (PLL) (CL-00-01, BioPAL), to enhance cell uptake, a procedure also applied to iron oxides.²⁶ Figure 1 is a schematic of the nanoparticle structure. Small volumes of Gd₂O₃ and PLL (2.5:1 ratio) were incubated in sterilized 1.5 mL Eppendorf tubes at RT for 1 hour to allow complete solubilization.

Figure 1.

Cartoon illustrating the structure of PL-Gado CELLTrack consisting of a Gd oxide core covered with a polymer coat of dextran. The Gd oxide dextran complex (Gado CELLTrack) has been treated by the manufacturer to retain a negative charge. The complex when mixed with poly-L-lysine (PLL) forms a series of ionic interactions that causes PLL to combine with Gado CELLTrack. The multiple points of attachment ensure the stability of the complex. Courtesy of BioPhysics Assay Laboratory, Inc.

Endothelial cells were grown in 12-well plates (BD Falcon, Mississauga, ON) for 2 days in a 37°C, 5% CO₂ humidified incubator. When 70 to 80% confluency was reached, 1 mL of fresh medium was added and cells were incubated with appropriate concentrations of Gd₂O₃ and returned to the incubator for 24 hours. The following Gd₂O₃ concentrations in the incubation medium were prepared: 0 (control), 0.002, 0.02, 0.1, 0.2, and 1 mM of Gd ion. Cells were washed three times with sterile 1× PBS to remove excess contrast agents. Fresh supplemented medium was then given to the cells (1 mL/well). The medium was changed every 2 to 3 days with subsequent washing. The last wash was carried out prior to MRI to ensure that extracellular Gd₂O₃ was eliminated.

Determination of Gd Content in Cells

Samples were assayed for Gd using inductively coupled plasma–atomic emission spectroscopy (ICP-AES) at the ANALEST facility (Toronto, ON). The analysis was performed on a PerkinElmer spectrometer (Model Optima 7300DV ICP AEOS, PerkinElmer, Waltham, MA) and certified against an independent source traceable to the National Institute of Standards and Technology Standard Reference Material. The detection threshold for Gd was 0.01 mg/mL. The cellular Gd concentration was determined by dividing the total Gd content by the number of cells.

Cell Viability and Proliferation Assays

To assess cell viability and proliferation, 96-well assay plates containing endothelial cells (≈ 8,000 cells in 250 μL of culture medium per well) were set up. Following a 24-hour period of cell adhesion onto the well plates, Gd₂O₃ was added to achieve the same concentrations as in the 12-well plate preparations for MRI. Cells were incubated with Gd₂O₃ for 24 hours, washed three times with 1× PBS, and either assayed immediately (1 day group) or incubated in 250 μL fresh culture medium for an additional 2 days (3-day group) or 6 days (7-day group). The medium was replaced every 48 hours. Fresh medium (100 μL) was replenished prior to performing the following assays.

CellTiter-Blue Cell Viability Assay (G8080, Promega, Madison, WI) was used to assess cell cytotoxicity following Gd₂O₃ incubation. Twenty microliters of CellTiter-Blue reagent was added to each well, shaken for 10 seconds, and incubated at 37°C, 5% CO₂ for 2 hours. Fluorescence was recorded at 560/590 nm (Spectra MAX Gemini Microplate Spectrofluorometer, Molecular Devices, Sunnyvale, CA).

CellTiter 96 Aqueous One Solution Cell Proliferation Assay (MTS [3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxy-methoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, inner salt; MTS]) (G3582, Promega) was used to determine the number of viable cells undergoing proliferation. Twenty microliters of combined MTS-PBS solution was dispensed into each well. Cells were incubated for 2 hours at 37°C, 5% CO₂. Absorbance was recorded at 490 nm.

For all assays, samples were set up as triplicates. Blank wells (media with reagent alone) and controls (media with unlabeled endothelial cells) were included. Fluorescence versus concentration of test samples was plotted using GraphPad Prism (GraphPad, LaJolla, CA).

Magnetic Resonance Imaging

Imaging was performed on a 7 T preclinical MRI magnet (BioSpec 70/30, Bruker, Ettlingen, Germany) using Bruker commercial hardware, namely, a B-GA12 gradient coil, a 7.2 cm cylindrical linear transmit coil, and a four-channel murine phased-array receiver coil. Prior to MRI, cells were prepared as follows. From the 12-well plates used to cultivate cells for MRI, approximately 250,000 cells per well were harvested using 100 μL 0.25% trypsin and 0.1 mM EDTA and subsequent 100 μL 10% FBS medium. Cells were transferred to 200 μL pipette tips, with bottoms sealed through melting and the tips capped with polymerase chain reaction tube caps. Two pipette tips were placed in a larger 15 mL tube, which was spun for 10 minutes at 1,500 rpm to create a cell pellet. Subsequently, the old medium was aspirated from the tips, and 200 μL of fresh human aortic endothelial cell medium (Lifeline Cell Technology) was added on top of the cells. There were approximately 250,000 cells per pipette tip.

Cell pellets in the pipette tips were kept on ice during the transfer to MRI. The tips were then secured firmly in an upright position in plasticine and placed on the phased-array coil. Quantitative T₁ relaxation times were measured using a two-dimensional (2D) saturation-recovery RARE sequence: RARE factor = 2; repetition time (TR) = 5,000, 3,000, 2,000, 1,500, 1,000, 750, 500, 250, 150, 125 ms; echo time (TE) = 7.43, 22.3, 37.1, 52.0, 66.8, 81.7, 96.6, 111.4 ms; field of view = 2.56 cm; slice thickness = 4 mm; in-plane resolution = 100 × 100 μm; scan time = 22 minutes. Quantitative T₂ relaxation times were measured using a spin-echo Carr-Purcell-Meiboom-Gill sequence: TR = 4,000 ms; 128 echoes with TE = 4.19, 8.38, · 536.5 ms; slice thickness 5 = mm; in-plane resolution = 200 × 200 μm; number of averages = 2; scan time = 13 minutes.

T₁ and T₂ relaxation times were also measured for Gd₂O₃ in water over the same concentration range used for cell incubation (0.002–1 mM) to determine the r₁ and r₂ relaxivity constants at 7 T.

Data Analysis

MRI raw data were transferred to an independent workstation for quantitative data analysis. All data processing was performed using in-house software developed in Matlab version 7.8 (MathWorks, Natick, MA). Quantitative maps of T₁ relaxation times were computed by nonlinear least-squares fitting to the signal intensity on the saturation-recovery curve at the minimum TE. Quantitative T₂ relaxation times were computed by fitting signal intensity to a monoexponential decay function added to a constant offset to account for noise. Relaxivity constants r₁ and r₂ of Gd₂O₃ in solution were determined by linear regression analysis of the change in relaxation rates (1/T₁ and 1/T₂) versus Gd concentration. All data were expressed as mean value ± standard deviation. All significance testing was based on the Student t-test, with a probability value (p) < .05 considered significant.

Results

Phenotype Expression in Cultured Endothelial Cells

Cultured normal human aortic endothelial cells were examined on phase-contrast microscopy (Figure 2A), and endothelial cell phenotype expression was confirmed by monitoring the expression of endothelial cell–derived vWF (previously known as factor VIII) via immunofluorescence (Figure 2B).

Figure 2.

Normal human aortic endothelial cells (passage 2). A, Phase-contrast microscopy (10× magnification). B, von Willebrand factor (vWF; factor VIII) immunofluorescence. Mouse monoclonal primary antibody IgG1 for vWF was detected with polyclonal goat-antimouse IgG-H&L (Texas red). Arrows indicate vWF staining (red). DAPI (blue) staining refers to fixed cell nuclei (20× magnification).

MRI Relaxivity Measurements of Gd₂O₃ in Solution

The relaxivity constants for Gd₂O₃ in water at 7 T were measured to be r₁ = 4.7 s⁻¹ mM⁻¹ and r₂ = 29.3 s⁻¹ mM⁻¹. These are summarized in Table 1 and compared against relaxivities of other Gd compounds measured at 7 T.^27–29 Figure 3 shows the plots of relaxation rates versus Gd concentration from which r₁ and r₂ were estimated.

Table 1.

Relaxivity Constants r₁ and r₂ of Gd Compounds Measured at 7 T

Contrast Agent	r₁	r₂	r₂/r₁	Reference
Gd₂O₃	4.7	29.3	6.2	This study
Gd-DOTA	2.6	4.9	1.89	27
Gd-DTPA	3.7	4.8	1.30	28
Gd-HP-DO₃A	3.3			29

Units for r₁ and r₂ (S⁻¹ mM⁻¹)

Figure 3.

Relaxivity measurements of Gd₂O₃ in water at 7 T. A, Relaxation rates 1/T₁ and 1/T₂ versus Gd₂O₃ concentration and relaxivities r₁ and r₂ calculated from linear regression slopes (R² = .997 for r₁; R² = .971 for r₂). B, MRI signal intensity on two-dimensional saturation recovery RARE sequence (TR = 1000 ms, TE = minimum).

Shortening of T₁ and T₂ Relaxation Times in Gd₂O₃-Labeled Cells

Figure 4 shows representative T₁- and T₂-weighted images from the 2D RARE sequence of pelleted Gd₂O₃-labeled human aortic endothelial cells. A variation in signal intensity in pellets labeled with different incubation medium concentrations ([Gd] = 0.002–1.0 mM) was observed in comparison with unlabeled controls. On T₁-weighted images, there is progressive signal enhancement with increasing incubation medium [Gd] over the range 0.002 to 0.1 mM. At the upper limit of [Gd] = 0.1 mM, competing T₂ effects begin to offset T₁ enhancement and are evident on T₂-weighted images as cell pellets begin to take on a hypointense appearance. At yet higher incubation concentrations, T₂ signal loss dominates, and signal voids are present on both T₁- and T₂-weighted images.

Figure 4.

Representative (A) T₁-weighted and (B) T₂-weighted MRIs of Gd₂O₃-labeled endothelial cell pellets at 7 T. Cells were incubated in medium at various [Gd] concentrations. Scan parameters were (A) TR/TE = 1,000/7.4 ms and (B) TR/TE = 5,000/22.3 ms. Note that from 0.1 mM and higher, cell pellets appear hypointense owing to competing effects from T₂ shortening.

Figure 5 demonstrates a decrease in both T₁ and T₂ relaxation times in labeled endothelial cells with increasing incubation medium [Gd]. Note that in the transition from [Gd] = 0.02 to 0.1 mM, T₂ drops significantly (from 50 ms to 15 ms), thus explaining why signal enhancement on a conventional T₁-weighted sequence can be gleaned from Gd₂O₃-induced T₁ shortening only at low concentrations.

Figure 5.

T₁ and T₂ relaxation times in Gd₂O₃-labeled endothelial cell pellets decrease with increasing incubation concentration [Gd].

Stability of Gd₂O₃ Labeling over Time

Figure 6 displays the stability of Gd₂O₃ inside labeled cells at various intervals after a 24-hour incubation period. A trend toward slowly increasing T₁ and T₂ relaxation times was observed at longer intervals, particularly at day 7, which could be due to signal dilution with cell proliferation. Overall, however, positive signal enhancement was maintained over 7 days post–cell labeling.

Figure 6.

T₁ and T₂ relaxation times in Gd₂O₃-labeled endothelial cell pellets over 7 days post–cell labeling. Shown are results for [Gd] = 0.02 mM.

Cell Viability and Proliferation

Figure 7 illustrates the effect of increasing [Gd] in the incubation medium on cell viability and proliferation. Viability was not affected significantly at any of the Gd₂O₃ concentrations relative to unlabeled cells. Proliferation was also unaffected at the lower Gd₂O₃ concentrations; however, a significant decrease was observed at [Gd] = 0.2 mM and higher (p < .05).

Figure 7.

Viability and proliferation assays show no change in Gd₂O₃-labeled endothelial cells relative to control (0 mM) at any of the concentrations except where indicated by an asterisk (*p < .05).

Figure 8 examines cell viability and proliferation at different time points post–cell labeling. Compared to unlabeled cells, neither viability nor proliferation was significantly altered in labeled cells at 1 and 3 days postlabeling. However, a latent effect was observed on day 7, where both viability and proliferative capacity were significantly reduced (p < .05).

Figure 8.

Viability and proliferative capacity in Gd₂O₃-labeled endothelial cells (filled circles) are unaffected relative to unlabeled controls (open circles) up to at least 3 days post–cell labeling. A significant decline is observed on day 7 (*p < .05). Shown are results for [Gd] = 0.02 mM.

Intracellular Gd Content

The mean cellular uptake of Gd₂O₃, expressed as pg Gd/cell, was assessed by ICP-AES and summarized in Table 2. The uptake of Gd₂O₃ into human aortic endothelial cells increased with increasing [Gd] in the incubation medium. No saturation effect was detected, and the amount of Gd₂O₃ taken up was linearly proportional to its concentration in the incubation medium (r = .995, p < 1e–4). Approximately half of the Gd₂O₃ in solution was taken up by cells.

Table 2.

Cellular Uptake of Gd₂O₃ Measured on ICP-AES

Incubation Concentration [Gd] (mM)	Mass of Gd (pg/cell)
0	0
0.002	0.71
0.02	7.26
0.1	34.1
0.2	84.9
1.0	289

Gd₂O₃ = gadolinium oxide; ICP-AES = inductively coupled plasma–atomic emission spectroscopy.

A linear correlation exists between [Gd] and cellular uptake (r = .995, p < 1e–4).

Discussion

Positive T₁ contrast agents based on Gd provide an important advantage over negative T₂ iron oxides that are traditionally used in cell labeling and cellular MRI applications: they generate a positive signal change to enhance detection of labeled cells, provide better anatomic delineation, and reduce confusion between iron-labeled cell signal voids and negative contrast artifacts. Gd-based agents in the conventional chelated form are suboptimal for cell labeling and positive contrast visualization because of their small size and high hydrophilicity,³⁰ which hinder spontaneous cell uptake and retention. Toward creating a positive T₁ agent with high sensitivity, different Gd formulations have been devised to improve cell uptake and optimize relaxation effects. Of these, Gd₂O₃ nanoparticles have shown the most promise.^14–20 Enhanced relaxivity for different Gd₂O₃ nanoparticles has been demonstrated at clinical field strengths, and their uptake efficiency has been investigated, albeit predominantly in phagocytic cell lines. However, it is unknown if Gd₂O₃ remains a suitable T₁ agent at higher field strengths and if it can be ingested efficiently by less phagocytic cells. The goal of this study was to determine the potential of a readily available Gd₂O₃ nanoparticle for labeling endothelial cells and providing positive T₁ contrast for high-field MRI at 7 T.

Gado CELLTrack, a commercially available 30 nm Gd₂O₃ colloid, was investigated for cell labeling and cellular MRI at 7 T. Structurally, it is analogous to iron oxide and is much larger than conventional Gd chelates; it is expected to be more readily taken up by cells and retained for longer periods. Our results at 7 T demonstrated the potential of Gd₂O₃ as a positive T₁ contrast agent for high-field cellular MRI. The major findings were as follows: (1) Gd₂O₃-labeled cells showed positive T₁ contrast at 7 T; (2) Gd₂O₃ uptake in endothelial cells was high, much higher than for most Gd compounds and comparable to Gd₂O₃ uptake reported in glioma cells; (3) uptake in cells was linear over the incubation concentration range studied (up to [Gd] = 1.0 mM); (4) optimal positive T₁ contrast was achieved at incubation concentrations ([Gd] < 0.1 mM) much lower than typical levels used for Gd agents; (5) T₁ and T₂ relaxation times increased slightly with time after cell labeling, but positive contrast was maintained up to 7 days; and (6) there was no deleterious effect of Gd₂O₃ on cell viability or proliferation for at least 3 days post–cell labeling.

Comparison with other cell labeling studies based on Gd contrast agents underscores several distinct advantages and features of Gd₂O₃. First, Gd₂O₃ is shown for the first time to exert an appreciable T₁ relaxation effect at 7 T. Very few Gd-based compounds, all non-Gd₂O₃ formulations,^8,12 have demonstrated this capability. Second, T₁ enhancement in labeled cells was optimal at low incubation concentrations (< 0.1 mM), which is a range similar to other Gd₂O₃ nanoparticles^16,17 and is much lower than that (< 25 mM) of conventional Gd agents.^12,13 Analysis of intracellular Gd content revealed that contrast uptake was linear with incubation concentration up to 1.0 mM, which is consistent with linearity of Gd₂O₃ uptake observed in another study in human monocytic leukemia cells.¹⁶ More importantly, positive contrast was achieved at low incubation concentrations owing to efficient internalization of Gd₂O₃: approximately half of the Gd was taken up by cells. Over the incubation concentration range of 0.02 to 0.1 mM, the intracellular Gd content was 7.26 to 34.1 pg Gd/cell, which is comparable to another report where 12 pg Gd/cell was measured in glioma cells incubated in 3.8 mM of Gd₂O₃.¹⁷ Our results and those of Faucher and colleagues¹⁷ suggest that the internalization of Gd₂O₃ is comparable to that of iron oxides^31,32 and is much more efficient relative to other Gd compounds, for which intracellular uptake is generally very low (0.119 pg/cell¹¹ and 0.3 pg/cell at 25 mM incubation concentration).¹³ An estimate of the intracellular molar concentration may be obtained by assuming a mean cell volume of 1,200 μm³, which would yield 38.5 and 181 mM of Gd corresponding to 0.02 and 0.1 pg/cell, respectively. This efficient cell uptake of Gd₂O₃, even in endothelial cells that have limited phagocytic capacity, suggests an active mechanism and may be further attributed to the use of PLL, which transforms the surface of the nanoparticle to a positive charge and thereby assists its attraction to cells (plasma membranes have a negative surface charge).

In addition to a high cellular accumulation of Gd₂O₃, relaxivity plays an integral role in determining the degree of positive contrast that can be obtained. A range of relaxivities have been reported for Gd₂O₃ with varying particle diameters: from r₁ = 8.8 to 9.9 s⁻¹mM⁻¹ and r₂/r₁ = 1.3 for 1 to 2.5 nm particles,^18,19 to r₁ = 3.6 s⁻¹mM⁻ and r₂/r₁ = 3.5 for 5 nm particles,¹⁶ to r₁ = 0.1 s⁻¹mM⁻¹ and r₂/r₁ = 81.6 for 30 nm particles.²⁰ Although these relaxivity measurements in aqueous solutions were performed at clinical field strengths and cannot be directly compared to our results at 7 T, the measured relaxivities of our nanoparticle (r₁ = 4.7 s⁻¹mM⁻¹ and r₂/r₁ = 6.2) fall within the range reported in the literature. It is important to note that since T₁ and T₂ relaxation effects coexist (T₁ effects predominate at low Gd concentrations, but competing T₂ effects dominate at higher concentrations³³), optimal positive contrast is obtained at intermediate concentrations. In our study, labeled cells were best visualized when incubated with Gd₂O₃ at concentrations less than 0.1 mM, a threshold that is slightly lower than 0.6 mM for a different Gd₂O₃ nanoparticle at 1.5 T.^14,16 These numbers are consistent with a lower concentration requirement for maximal signal contrast from Gd₂O₃ nanoparticles compared to Gd-DTPA.¹⁴

Further considerations for a suitable cell labeling contrast agent include its retention in cells and its effects on cell function. We found that Gd₂O₃ induced positive contrast changes that were retained for at least 7 days following cell labeling. Relaxation times T₁ and T₂ slowly increased with time but remained well within the range expected for signal dilutions owing to cell proliferation. For human aortic endothelial cells, we measured the doubling time to be approximately 42 hours, which is significantly longer than other cell types studied with Gd₂O₃. Therefore, by days 3 and 7, we would expect one and two doublings, respectively, which would account for the slight increase in T₁ and T₂. Few reports exist for comparison on T₁ and T₂ stability, none of which is based on Gd₂O₃. The one study that showed retention of contrast changes was performed on Gadofluorine M,¹³ which noted that 1/T₁ decreased 3 days post–cell labeling but then remained relatively stable up to day 7. The stability of Gd₂O₃ contrast compared to Gd chelates soon after cell labeling suggests enhanced retention in cells that is most likely due to the larger size of Gd₂O₃. Analysis of cell function showed that Gd₂O₃ does not alter viability and proliferation at the concentrations tested. However, a significant decrease in both viability and proliferation was observed on day 7. The reason is unclear, and assessment of cell function in the days following Gd₂O₃ labeling is not available from other literature investigations for comparison.

Future work will involve more comprehensive studies into the mechanisms of action and applications of Gd₂O₃. Stem cells and other phenotypes that are useful in cell tracking applications will be investigated. For different cell types, we need to gain insight into how Gd₂O₃ affects cell metabolism and function and characterize details related to its action (eg, optimal incubation time, cell uptake mechanism, intracellular distribution, and retention period). For example, although a 24-hour incubation period was used in this study to ensure contrast uptake in endothelial cells, shorter intervals of 4 or 8 hours may also suffice. Several observations from this study also deserve further investigation. One observation is a larger r₂/r₁ compared to other studies, which translates to reaching the maximum positive contrast at a lower Gd concentration. This may be related to different field strengths, and this hypothesis will need to be validated at lower fields. Another observation is a hyperintense signal in the supernatant above the cell pellets. Given that cells were washed immediately prior to MRI and the fresh medium added was free of Gd₂O₃, there may have been slow lysis of cells as a result of dense cell packing in a pellet formation, which would disperse some contrast into the overlying supernatant. Finally, beyond these in vitro characterizations, the in vivo efficacy of Gd₂O₃ for cell labeling and tracking at 7 T will need to be demonstrated in animal studies.

Conclusion

This study demonstrates that Gd₂O₃ nanoparticles are a promising contrast agent for cell labeling and positive T₁ contrast cellular MRI at 7 T. Significant signal enhancement was demonstrated in endothelial cells labeled with Gd₂O₃ owing to efficient cell uptake of contrast agent at low incubation concentrations. The advantages of the Gd₂O₃ used in this study include its availability, ease of labeling without complicated assisted internalization into cells, optimal contrast agent uptake at low concentrations, positive contrast at high field strengths, and stable labeling for at least 7 days. These results suggest that Gd₂O₃ is suitable for time-course cell tracking studies and may overcome limitations associated with traditional Gd compounds or iron oxides.

Footnotes

Acknowledgment

Financial disclosure of authors: This work was supported by a grant from the Canadian Institutes of Health Research.

Financial disclosure of reviewers: None reported.

References

Kraitchman

Heldman

Atalar

. In vivo magnetic resonance imaging of mesenchymal stem cells in myocardial infarction. Circulation 2003;107:2290–3, doi:10.1161/01.CIR.0000070931.62772.4E.

Hoehn

Küstermann

Blunk

. Monitoring of implanted stem cell migration in vivo: a highly resolved in vivo magnetic resonance imaging investigation of experimental stroke in rat. Proc Natl Acad Sci U S A 2002;99:16267–72, doi:10.1073/pnas.242435499.

Bernas

Foster

Rutt

. Imaging iron-loaded mouse glioma tumors with bSSFP at 3 T. Magn Reson Med 2010;64:23–31, doi:10.1002/mrm.22210.

Shen

Zhong

Duan

. Magnetic resonance imaging of mesenchymal stem cells labeled with dual (MR and fluorescence) agents in rat spinal cord injury. Acad Radiol 2009;16:1142–54, doi:10.1016/j.acra.2009.03.016.

Saldanha

Doan

Ainslie

. Micrometer-sized iron oxide particle labeling of mesenchymal stem cells for magnetic resonance imaging-based monitoring of cartilage tissue engineering. Magn Reson Imaging 2011;29:40–9.

Modo

Hoehn

Bulte

. Cellular MR imaging. Mol Imaging 2005;4:143–64.

Liu

Frank

. Detection and quantification of magnetically labeled cells by cellular MRI. Eur J Radiol 2009;70:258–64, doi:10.1016/j.ejrad.2008.09.021.

Anderson

Lee

Frank

. Gadolinium-fullerenol as a paramagnetic contrast agent for cellular imaging. Invest Radiol 2006;41:332–8, doi:10.1097/01.rli.0000192420.94038.9e.

Henning

Saborowski

Golovko

. Cell labeling with the positive MR contrast agent Gadofluorine M. Eur Radiol 2007; 17:1226–34, doi:10.1007/s00330-006-0522-9.

10.

Tseng

Shih

Stobinski

. Gadolinium hexanedione nanoparticles for stem cell labeling and tracking via magnetic resonance imaging. Biomaterials 2010;31:5427–35, doi:10.1016/j.biomaterials.2010.03.049.

11.

Hsiao

Tsai

Chung

. Mesoporous silica nanoparticles as a delivery system of gadolinium for effective human stem cell tracking. Small 2008;4:1445–52, doi:10.1002/smll.200701316.

12.

Crich

Biancone

Cantaluppi

. Improved route for the visualization of stem cells labeled with a Gd-/Eu-chelate as dual (MRI and fluorescence) agent. Magn Reson Med 2004;51:938–44, doi:10.1002/mrm.20072.

13.

Daldrup-Link

Rudelius

Oostendorp

. Targeting of hematopoietic progenitor cells with MR contrast agents. Radiology 2003;228:760–7, doi:10.1148/radiol.2283020322.

14.

Engstroüm

Klasson

Pedersen

. High proton relaxivity for gadolinium oxide nanoparticles. MAGMA 2006;19:180–6, doi:10.1007/s10334-006-0039-x.

15.

Ahrén

Selegárd

Klasson

. Synthesis and characterization of PEGylated Gd2O3 nanoparticles for MRI contrast enhancement. Langmuir 2010;26:5753–62, doi:10.1021/la903566y.

16.

Klasson

Ahrén

Hellqvist

. Positive MRI contrast enhancement in THP-1 cells with Gd2O3 nanoparticles. Contrast Media Mol Imaging 2008;3:106–11, doi:10.1002/cmmi.236.

17.

Faucher

Guay-Bégin

Lagueux

. Ultra-small gadolinium oxide nanoparticles to image brain cancer cells in vivo with MRI. Contrast Media Mol Imaging 2010 Dec 1. [Epub ahead of print]

18.

Park

Baek

Choi

. Paramagnetic ultrasmall gadolinium oxide nanoparticles as advanced T1 MRI contrast agent: account for large longitudinal relaxivity, optimal particle diameter, and in vivo T1 MR images. ACS Nano 2009;3:3663–9, doi:10.1021/nn900761s.

19.

Bridot

Faure

Laurent

. Hybrid gadolinium oxide nanoparticles: multimodal contrast agents for in vivo imaging. J Am Chem Soc 2007;129:5076–84, doi:10.1021/ja068356j.

20.

Fortin

Petoral

Soderlind

. Polyethylene glycol-covered ultra-small Gd2O3 nanoparticles for positive contrast at 1.5 T magnetic resonance clinical scanning. Nanotechnology 2007; 18:395501, doi:10.1088/0957-4484/18/39/395501.

21.

Caravan

. Strategies for increasing the sensitivity of gadolinium based MRI contrast agents. Chem Soc Rev 2006;35:512–23, doi:10.1039/b510982p.

22.

Rafii

Lyden

. Therapeutic stem and progenitor cell transplantation for organ vascularization and regeneration. Nat Med 2003;9:702–72, doi:10.1038/nm0603-702.

23.

Montesano

Orci

Vassalli

. In vitro rapid organization of endothelial cells into capillary-like networks is promoted by collagen matrices. J Cell Biol 1983;97:1648–52, doi:10.1083/jcb.97.5.1648.

24.

van Tiel

Wielopolski

Houston

. Variations in labeling protocol influence incorporation, distribution and retention of iron oxide nanoparticles into human umbilical vein endothelial cells. Contrast Media Mol Imaging 2010;5:247–57, doi:10.1002/cmmi.379.

25.

Lee

Shuter

Chan

. The use of microgel iron oxide nanoparticles in studies of magnetic resonance relaxation and endothelial progenitor cell labelling. Biomaterials 2010;31:3296–306, doi:10.1016/j.biomaterials.2010.01.037.

26.

Unger

. How can superparamagnetic iron oxides be used to monitor disease and treatment? Radiology 2003;229:615–6, doi:10.1148/radiol.2293031017.

27.

Chapon

Lemaire

Franconi

. Assessment of myocardial viability in rats: evaluation of a new method using superparamagnetic iron oxide nanoparticles and Gd-DOTA at high magnetic field. Magn Reson Med 2004;52:932–6, doi:10.1002/mrm.20210.

28.

Shahbazi-Gahrouei

. Gadolinium-porphyrins: new potential magnetic resonance imaging contrast agents for melanoma detection. J Res Med Sci 2006;11:217–23.

29.

Noebauer-Huhmann

Szomolanyi

Juras

. Gadolinium-based magnetic resonance contrast agents at 7 Tesla: in vitro T1 relaxivities in human blood plasma. Invest Radiol 2010; 45:554–8, doi:10.1097/RLI.0b013e3181ebd4e3.

30.

Merbach

Toth

. The chemistry of contrast agents in medical magnetic resonance imaging. Chichester (NY): Wiley; 2001.

31.

Hsiao

Tai

Chu

. Magnetic nanoparticle labeling of mesenchymal stem cells without transfection agent: cellular behavior and capability of detection with clinical 1.5 T magnetic resonance at the single cell level. Magn Reson Med 2007;58:717–24, doi:10.1002/mrm.21377.

32.

Heyn

Bowen

Rutt

. Detection threshold of single SPIO-labeled cells with FIESTA. Magn Reson Med 2005;53:312–20, doi:10.1002/mrm.20356.

33.

Modo

Cash

Mellodew

. Tracking transplanted stem cell migration using bifunctional, contrast agent-enhanced, magnetic resonance imaging. Neuroimage 2002;17:803–11, doi:10.1006/nimg.2002.1194.

Human Aortic Endothelial Cell Labeling with Positive Contrast Gadolinium Oxide Nanoparticles for Cellular Magnetic Resonance Imaging at 7 Tesla

Abstract

Materials and Methods

Cell Culture and Immunofluorescence

Contrast Agent and Cell Labeling

Determination of Gd Content in Cells

Cell Viability and Proliferation Assays

Magnetic Resonance Imaging

Data Analysis

Results

Phenotype Expression in Cultured Endothelial Cells

MRI Relaxivity Measurements of Gd2O3 in Solution

Shortening of T1 and T2 Relaxation Times in Gd2O3-Labeled Cells

Stability of Gd2O3 Labeling over Time

Cell Viability and Proliferation

Intracellular Gd Content

Discussion

Conclusion

Footnotes

Acknowledgment

References

MRI Relaxivity Measurements of Gd₂O₃ in Solution

Shortening of T₁ and T₂ Relaxation Times in Gd₂O₃-Labeled Cells

Stability of Gd₂O₃ Labeling over Time