MRI-detectable pH nanosensors incorporated into hydrogels for in vivo sensing of transplanted cell viability
Kannie W.Y. Chan1,2, Guanshu Liu1,3, Xiaolei Song1,2, Heechul Kim1,2, Tao Yu4,8,
Dian R. Arifin1,2, Assaf A. Gilad1,2, Justin Hanes4-8, Piotr Walczak1,2,
Peter C.M. van Zijl1,3, Jeff W.M. Bulte1-6, Michael T. McMahon1,3*
1The Russell H. Morgan Department of Radiology and Radiological Science, Division of MR Research, The Johns Hopkins University School of Medicine, Baltimore, Maryland, USA.
2Cellular Imaging Section and Vascular Biology Program, Institute for Cell Engineering, The Johns Hopkins University School of Medicine, Baltimore, Maryland, USA.
3F.M. Kirby Research Center for Functional Brain Imaging, Kennedy Krieger Institute, Baltimore, USA.
4Department of Biomedical Engineering, The Johns Hopkins University School of Medicine, Baltimore, Maryland, USA.
5Department of Chemical & Biomolecular Engineering, The Johns Hopkins University School of Medicine, Baltimore, Maryland, USA.
6Department of Oncology, The Johns Hopkins University School of Medicine, Baltimore, Maryland, USA.
7Department of Ophthalmology, The Wilmer Eye Institute, The Johns Hopkins University School of Medicine, Baltimore, MD, USA
8Center for Nanomedicine, The Johns Hopkins University School of Medicine
*Corresponding author: M.T. McMahon:
S1. Preparation and characterization of L-arginine liposomes
All reagents were used as received from Sigma, Avanti and Invitrogen, unless otherwise stated. In brief, we mixed egg phosphatidylcholine (PC) and cholesterol in a 1:0.5, 1:1 or 1:3 molar ratio with 5 mol% 1,2-dimyristoyl-sn-glycero- 3-phosphoethanolamine-N-[methoxy-(polyethyleneglycol)-2000] (PEG2000-PE), and with or without 1.68 mol% 1,2-dioleoyl- sn-glycero-3-phosphoethanolamine- N-(7-nitro-2-1,3-benzoxadiazol-4-yl) (NBD-PE) in chloroform. The solvent was removed and the resulting thin film was hydrated with 100 mg/mL L-arginine. The suspension was annealed at 55 °C for 2 hours. Small liposomes were obtained by sequential extrusion through polycarbonate filters with pore sizes of 400 nm and 200 nm. Non-incorporated L-arginine was removed by dialysis with a MW cutoff of 3.5 kDa. The PC content was quantified using a colorimetric Stewart assay1.
This assay measures the complexation of phospholipids (PC) with ammonium ferrothiocyanate in solution. Twenty μL of liposome sample was vortexed with 2 ml of ammonium ferrothiocyanate (0.1 M) in 2 mL of chloroform. The optical density at 485 nm of the chloroform phase was measured, from which the phospholipid concentration ([PC] in mg/ml) was determined using a calibration curve of phospholipid standards. The phospholipid concentration was then calculated from the number of PC per liposome, using
,
where Mlipo and Mlipid is the molar concentration of the liposomes and PC, respectively, and Ntot is the number of lipid molecules per liposome estimated from
where Souter and Sinner is the surface area of the outer shell and inner shell, respectively; a is the lipid head group area estimated as 45 Å2, d is the diameter of the liposome and h is the thickness of the bilayer being approximating 5 nm. The liposome size (d) was measured in phosphate buffered saline (PBS) using a zetasizer (Zetasizer Nano ZS90, Malvern Instruments).
S2. In vitro assessment of LipoCEST contrast and pH sensitivity
Typical phantom acquisition parameters were: TR=6.0 sec, effective TE=21.6 ms, RARE factor=16, tsat=4 sec, B1=3.6 μT (150 Hz), slice thickness=1 mm, acquisition matrix size=128x64, FOV=15x15 mm, and NA=2. We incremented the saturation offset ± 1 ppm (0.1 ppm steps) with respect to water in WASSR2 acquisition and ± 5 ppm (0.2 ppm steps) in z-spectra acquisition. Region of interest (ROIs) were drawn manually based on the MTw images to cover the entire region of LipoCEST capsules; mean intensities were used for plotting MTRasym.
Data were processed using custom-written scripts in MATLAB (Mathworks, Waltham, MA). Mean Z-spectra were calculated from an ROI placed over each tube region after B0 correction on a per voxel basis using WASSR. CEST contrast was quantified using MTRasym at the particular offset of interest (i.e. Δω= +2 ppm) (Fig. S1).
S3. In vitro assessment of capsule permeability and mechanical strength
Membrane permeability was determined using dextran-FITC samples of 10, 40, 150, 250, and 500 kDa3. For each separate MW of dextran-FITC the measurement was performed in triplicate. Approximately 10 capsules were placed in a 180-mm-thick custom-made well on a glass slide and exposed to 20 μL of 0.05% dextran-FITC in saline, which was immediately sealed. The slides containing capsules were kept at room temperature for over 24 h and imaged using a confocal laser scanning microscopy (Zeiss LSM 500) (Fig. S2a). Once sealed, samples were also examined using fluorescence microscopy (Zeiss Axio Observer) to study the kinetics of lateral diffusion of dextran-FITC into the capsules. Images were taken every minute for 15 minutes, and the fluorescence intensity of the center of the capsules (n=3) as well as the surrounding solution was measured using ImageJ. Values are reported as intensity ratios of the two to compensate for variation of FITC concentration among the different dextran-FITC samples (Fig. 2b-d).
In order to determine mechanical strength, an osmotic stability test was performed. 500-600 capsules were washed with picopure water and incubated in 5 mL of picopure water at 37oC for one hour. After washing with saline, the numbers of fractured, swollen, and intact capsules were counted using an inverted light microscope (Olympus IX71). For a bead agitation test of mechanical strength, capsules were suspended in 1 mL of saline with 10 mM HEPES and 0.05 g of inert glass beads (diameter 0.425-0.6 mm, Sigma-Aldrich, St. Louis, MO) added. Samples were centrifuged for 3 hr at 13,200 rpm at room temperature using an Eppendorf Minispin microcentrifuge (Brinkmann, VWR LabShop, Batavia, IL). Ruptured, fractured, and intact capsules were then counted using the inverted light microscope.
S4. Luciferase transduction of hepatocytes and in vitro BLI
Human HepG2 cells were transduced with luciferase (Luc) using a cytomegalovirus (CMV)-based constitutive promoter, pLenti-4-CMV-fLuc2, with over 95% transduction efficiency for luciferase expression as determined by rabbit anti-Firefly Luc polyconal antibody (Abcam Inc.) staining with Hoechst33324 nuclear counterstaining (Fig. S3a). For using the BLI radiance to determine the amount of viable cells, a calibration curve was obtained from a known number of cells (counted using a hemocytometer) pipetted into a black, 96-well plate (Radiance=68.4*, estimated cell number-19855, R2=0.9993) (Fig. S3b). This standard curve was used to estimate the number of viable cells before and after CEST imaging in vitro (Fig. 5). Both APLLA and Lipo70-APSA microcapsules loaded with Luc-expressing HepG2 cells were kept in culture media and followed with BLI over a period of 1 month (Fig. S3c,d). As the cells had a >95% transduction efficiency and were transplanted s.c. directly under the skin of white mice, we used this calibration curve as an approximation to estimate the number of viable cells in vivo (Fig. 6f,g). Based on this, we estimate there were 0.34 and 0.19 million cells which died in the +Cells/-IS and +Cells/+IS groups respectively after two weeks, with a corresponding decrease in CEST contrast of 33% and 13%, respectively. Taken together with Fig. S6, which shows a correlation between BLI radiance and MTRasym in the +Cells/-IS group.
S.5 CEST MR and BL imaging in vivo
MT-weighted images (Fig. S7) were acquired using a MT module with a saturation pulse at an offset frequency of -10 ppm TR=5.0 sec, effective TE=21.6 ms, RARE factor=8, tsat=3 sec, B1=3.6 μT (150 Hz), slice thickness=1 mm, acquisition matrix size=128x64, FOV=23x18 mm, and NA=2). ROIs were drawn manually over the entire region containing LipoCEST microcapsules, and the mean intensities were used to plot the MTRasym values. The CEST/MT-weighted overlay (Fig. 6a-c) was based on an ROI placed over the entire s.c. region that contained capsules in the MT-weighted images.
Bioluminescence tomography (BLT) was performed using an IVIS 200 optical imaging system (Caliper Life Sciences, USA). Spectral images were obtained with an exposure time of 1 min using wavelengths from 580 to 660 nm and five 20-nm bandpass filters. These images were then used to reconstruct the bioluminescent source location and intensity based on a diffusion model of light propagation and absorption4. A structured light image was taken to reconstruct the surface topography of the mouse. Images were reconstructed and analyzed using the optical properties of muscle tissue included in the Diffuse Luminescence Image Tomography, DLIT4 package of Living image 4.0 software (Fig. S8).
S6. Histology
Two mice from each three groups were sacrificed on day 14 after transplantation. S.c. tissue containing transplanted microcapsules were fixed in 10% formalin and embedded in paraffin. Five mm thick tissue sections were deparaffinated and hydrated with a xylene/alcohol series and then stained with haematoxylin and eosin.
References
1 Stewart, J. C. Colorimetric determination of phospholipids with ammonium ferrothiocyanate. Anal Biochem 104, 10-14, (1980).
2 Kim, M., Gillen, J., Landman, B. A., Zhou, J. & van Zijl, P. C. Water saturation shift referencing (WASSR) for chemical exchange saturation transfer (CEST) experiments. Magn Reson Med 61, 1441-1450, (2009).
3 Gardner, C. M., Burke, N. A. & Stover, H. D. Cross-linked microcapsules formed from self-deactivating reactive polyelectrolytes. Langmuir 26, 4916-4924, (2010).
4 Kuo, C., Coquoz, O., Troy, T. L., Xu, H. & Rice, B. W. Three-dimensional reconstruction of in vivo bioluminescent sources based on multispectral imaging. J Biomed Opt 12, 024007, (2007).
Figure S1 In vitro MRI contrast of microcapsules. a, Z-spectra of Lipo70-APSA, Lipo70-APLLA and APLLA. b, Corresponding MTRasym values, calculated using MTRasym = (S-Δω - S+Δω)/S-Δω. In this equation, SΔω and S-Δω are the MR signals with the saturation RF pulses applied at Δω (CEST offset of +2 ppm) and –Δω (the opposite offset in respect to water resonance, i.e. -2 ppm).
Figure S2 Permeability study using dextran-FITC. a, Confocal microscopy images of APLLA, Lipo70-APLLA, APSA and Lipo70-APSA (LipoCEST) microcapsules after 24 hrs of incubation with dextran-FITC (MW 10-500 kDa). b-e, Intensity ratios of capsule to background for 15 min of incubation with dextran-FITC for APLLA (b), Lipo70-APLLA (c), APSA (d), and Lipo70-APSA (e).
Figure S3 Human hepatocytes in culture. a, Microscopic images of transduced HepG2 cells stained for luciferase (green) and cell nuclei (blue). b, Radiance obtained from a known number of hepatocytes. c, Average BLI radiance vs time for encapsulated hepatocytes in vitro with the BLI radiance levels from ~150 microcapsules per well on day 0, 1, 7, 14, and 30 in culture. d, Light microscopic image of a Lipo70-APSA capsule on day 0 and 30 showing cell proliferation within the capsule. Scale bar = 200 μm.
Figure S4 LipoCEST capsules with varying numbers of cells. (a) MTw images and MTRasym map at 2 ppm of capsules containing live cells and apoptotic cells with increasing numbers of cells per capsules (i-iii. 0.5x103, 1.5x103 and 5.0x103 cells per capsule, respectively) at 0 hr and 12 hrs after the addition of staurosporine. (b) relative MTRasym (MTRasym12hrs/MTRasym0hr) and relative radiance (Radiance12hrs/Radiance0hr) (n=3) of the corresponding phantoms with live cells and apoptotic cells with increasing numbers of cells (n=3), which indicates the estimated 0.1, 0.3, 1 and 2 million of dead cells per phantom using BLI.
Figure S5 BLI of transplanted LipoCEST encapsulated hepatocytes (Encapsulated cells, +Cells/-IS) and hepatocytes without encapsulation (Cells only). (a) BL images of representative animal in encapsulated cells group on day 0, 1, 3, 7, 14 and 28, and cells only group at day 0, 1, 3 and 7 after transplantation; (b) radiance of the two groups (n=3) over 28 and 7 days respectively.
Figure S6 Correlation between CEST and BLI. a, Correlation between CEST (MTRasym) and relative radiance (BLI) for the +Cells/+IS (blue) and b, +Cells/-IS (red) group over two weeks (p<0.05 using Spearman test with 95% confidence interval, r=0.81).
Figure S7 MT-weighted images. Images correspond to the same mice shown in Figure 6a, which represent mice from the three groups over two weeks.
Figure S8 BLT of hepatocyte-containing LipoCEST microcapsules 14 days after s.c. transplantation. A reduced number of viable cells was found in the central capsule (red arrow) region for the mice that received no immunosuppression (+Cells/-IS).
Figure S9 In vivo CEST and BL imaging of LipoCEST capsules containing hepatocytes in +Cells/+IS, +Cells/-IS and +DeadCells/+IS groups (n=3). (a) MTRasym of the three groups of animals showing significant differences at day 7, 14 and 28; (b) MTRasym map at 2 ppm of the representative animals from each group; CEST contrast decreased at day 1 and day 28 in the +Cells/+IS group, and was constant in the +DeadCells/+IS group. (c) Relative radiance of the +Cells/+IS and the +Cells/-IS groups and (d) the corresponding BL images of the representative animals. (e) a comparison between the CEST contrast (solid bars) and relative radiance (open bars) of the +Cells/+IS group. (*, P<0.05; **, P<0.01; ***, P<0.001).
Fig. S10 The CEST data for all the groups (left panel) and BLI data for the cell containing groups (right panel). A two-factor repeated measures ANOVA was used for comparing the data at each time point with the corresponding significance levels (-Cells vs +Cells/-IS with *, P<0.05; ***, P<0.001 in black) and (+Cells/+IS vs +Cells/-IS with ***, P<0.001 in blue).
1