Microfluidic device
A customized microfluidic device was designed to label cells with high throughput using mechanoporation. The device comprises one inlet, 3 or 5 mechanoporation channels in which chevron ridges (9.6 µm gap size) were embedded, and one outlet14. The devices were fabricated using standard polydimethylsiloxane (PDMS) molding procedures. Then a plasma bonder (PDC-32G Harrick) was used to bond to microscope glass slides with PDMS.
Cell culture
All experimental procedures involving animals comply with the ARRIVE guidelines. In addition, our study protocol was approved by the Administrative Panel on Laboratory Animal Care of Stanford University. All experiments involving animals were performed according to the approved protocol. Adipose tissue-derived stem cells (ADSCs) were isolated and harvested from the knee joints of Goettingen minipigs (Marshall Farms, North Rose, NY). According to our previously established techniques14, tissue samples were collected from the infrapatellar fat pad, dissociated with type I collagenase, and the isolated cells were characterized based on specific stem cell markers. The cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM) medium supplemented with 1% antibiotic–antimycotic mix and 10% fetal bovine serum (FBS). For experiments, ADSCs were expanded up to passage 6.
Characterization of MSNs by TEM
Propylamine-functionalized mesoporous silica nanoparticles (200 nm particle diameter and 4 nm pore size, Sigma-Aldrich, St Louis, MO, USA) were deposited on carbon/formvar coated copper grids and the size of particles were measured by TEM (Transmission Electron Microscope, JEM-1400 series 120 kV, JEOL USA Inc., Pleasanton, CA, USA). Cells labeled thorugh mechanoporation were fixed with 2% glutaraldehyde and imaged using the same procedure.
Radiolabeling of MSNs with 68Ga
To label MSNs with 68Ga, a chelator-free reaction method was used10,22. This approach results in the formation of a stable coordination complex between the radiometals and deprotonated Si–O–groups on the MSN surface First, 4.5 µg of MSNs, activated in ethanol overnight, were mixed with 1.5 ml of 68GaCl3 solution (~ 600 MBq) eluted with HCl (0.1 N), 30 µl of amonium hydroxide solution (Sigma-Aldrich, St Louis, MO, USA), and 100 µl of MES buffer (0.1 M, pH 7.3) to maintain a reaction pH of 7.3. The mixture was incubated at 75 °C for 15 min, and then centrifuged at 14,000 rpm to remove residual 68Ga. Radiolabeling purity was analyzed using an AR-2000 radio-TLC plate reader (BioScan Inc., Washington, DC, USA). Before cell treatment, the MSNs were coated with a lipid bilayer using cationic liposome transfection agent (Lipofectamine 2000, Invitrogen, California, USA), as previsouly described10.
Fluorescence imaging and flow cytometry
For in vitro validation, MSNs were labeled with FITC (fluorescein isothiocyanate isomer I; Sigma-Aldrich, St Louis, MO, USA) by mixing 1 mg of MSNs with 0.1 mg of FITC in 1 ml volume. After incubation at room temperature for 1 h, the mixture was washed with PBS to remove excess FITC. For cell treatment, a cationic liposome transfection agent (Lipofectamine 2000, Invitrogen, California, USA) was used to coat MSNs with a lipid layer. For mechanoporation cell labeling, ADSCs (2 × 105 cells) were mixed with 1 ml lipid-coated FITC-MSNs (4.5 µg/ml) in 2 ml FACS buffer [3x] and 3 ml PBS and passed through the microfluidic device (flow rate 0.5 ml/min). For microscopic visualization, cells were labeled with a fluorescent membrane tracer, DiI (Invitrogen, Carlsbad, CA, USA; red fluorescence; Ex: 565 nm, Em: 594 nm; incubation at 37 °C for 15 min), and a nucleus stain, Hoechst 33,342 (NucBlue Live, ReadyProbes; Thermo Fisher Scientific, Waltham, MA, USA; incubation at 37 °C for 5 min). To confirm the uptake of FITC-MSNs, cells were visualized by fluorescence microscopy (EVOS FL, ThermoFisher Scientific, Santa Clara, CA, USA) and quantified by flow cytometry (BD FACSAria Fusion sorter, BD Bioscience, San Jose, CA, USA).
In vitro PET imaging and gamma counting
ADSCs (2 × 105 cells) were radiolaeled by mixing them with 1 ml 68Ga-MSNs (~ 25 MBq/ml) in 2 ml FACS buffer [3x] and 3 ml PBS and passing them through the mechanoporation device (5 channels, flow rate 0.5 ml/min). After mechanoporation, the cells washed three times with PBS to remove residual 68Ga-MSNs, then the labeled cells (3–48 × 103 cells) were seeded in 24 well plate for PET imaging. For comparison, ADSCs were also passively incubated with 68Ga-MSNs (~ 25 MBq/ml for 10, 30, and 60 min) and seeded in the same plate. The well plate was imaged using PET (10 min scan; Inveon D-PET, Siemens Preclinical Solutions, Knoxville, TN). For quantitation, region of interest (ROI) analysis was performed on PET images using the Inveon Research Workplace (IRW) software. After PET imaging, the absolute radioactivity of cells was confirmed by gamma counting (AMG, Hidex, Turku, Finland). We also assessed radiolabel efflux by labeling ADSCs (1 × 104 cells/well) with 68Ga-MSNs using the microfluidics device, then incubating the cells for 30, 60, 90, or 120 min in 24 well plate. After centrifugation and washing in PBS, the remaining radioactivity was measured by gamma counting.
Cell viability assays
To assess the potential toxicity of the labeling procedure, labeled and unlabeled ADSCs (2.5 × 103 cells/well) were seeded in 96-well plates and incubated for 48 h. The cells were then incubated with CCK-8 solution (Sigma-Aldrich, St Louis, MO, USA) for 1 h. The mean optical density (OD) of the samples was measured at 450 nm using a GloMax multi-detection system (Promega, Madison, WI, USA). In addition, DNA damage was assessed using the γH2AX assay. ADSCs (1 × 104 cells/well; either unlabeled, or labeled using mechanoporation or 60 min passive incubation) were seeded and cultured for 1 h in Lab-Tek II chamber slides (ThermoFisher Scientific, Santa Clara, CA, USA). After one hour, these cells were fixed in 4% formaldehyde for 10 min and stained using primary anti-phospho-histone H2A.X (Ser139) antibody (1:100; cat# 05-636, Sigma-Aldrich, St Louis, MO, US). After overnight staining at 4 °C, secondary staining was conducted using anti-mouse Alexa Fluor 488 antibody (1:100; ThermoFisher Scientific, Santa Clara, CA, USA). Then, the stained cells were visualized by fluorescence microscopy (EVOS FL, ThermoFisher Scientific, Santa Clara, CA, USA) and the images were quantified using Image J software. Additionally, apoptosis was detected using an Annexin-V staining kit (Abcam, Cambridge, England). ADSCs (1 × 105 cells/well) were seeded and cultured for 48 h in 24-well plates post labeling. Then, the stained cells were visualized by fluorescence microscopy and the images were quantified using Image J software. Finally, to assess proliferation, ADSCs (1 × 104 cells/well) were seeded and cultured for 48 h post-labeling in Lab-Tek II Chamber SlideTM (1:100; ThermoFisher Scientific, Santa Clara, CA, USA). After 48 h, these cells were fixed in 4% formaldehyde for 10 min, and stained using primary anti-Ki67 antibody (1:100; cat# ab15580, Abcam, Cambridge, UK). After overnight staining at 4 °C, secondary staining was conducted using using anti-rabbit Alexa Fluor 594 antibody (1:100; ThermoFisher Scientific, Santa Clara, CA, USA). The cells were then visualized by fluorescence microscopy and the images were quantified using Image J software.
PET/MR imaging of dual-labeled stem cells in pig knee joints
To evaluate the approach in a clinical environment, we used a previously established model of cell transplantation based on artificially created cartilage defects in pig knee joints22. ADSCs were dual-labeled with 68Ga-MSNs and Ferumoxytol (AMAG Pharmaceuticals Inc., Cambridge, MA, USA), used here as an MRI contrast. For Ferumoxytol, a FDA-approved iron supplement, we followed the labeling protocol from our previous study14. First, ADSCs (1 × 107 cells) were mixed with Ferumoxytol (10 mg/ml) and 68Ga-MSNs (~ 100 MBq/ml) in FACS buffer and PBS, then passed through the microfluidic device (5 channels, flow rate 0.5 ml/min). After mechanoporation, the cells were washed 3 times with PBS to remove residual Ferumoxytol and 68Ga-MSN. Dual-labeled (n = 4) and unlabeled (n = 4) cells were implanted into 8 cartilage defects in the femoral end of pig knee specimens. The cells were implanted using fibrin glue (Ethicon, Somerville, NJ) and then the joint capsule, muscles and skin were sutured. PET/MRI images were obtained using a clinical 3 T Signa PET/MR scanner (GE Healthcare, Milwaukee, Wisconsin). PET images were acquired using a 30 min acquisition time, simultaneously with the MRI acquisition. MRI included proton density (PD) weighted fast spin echo (FSE) with fat saturation [acquisition time (TA) = 16 min, field of view (FOV) = 14 cm, repetition time (TR) = 2700 ms/echo time (TE) = 32 ms, flip angle (FA) = 110°, matrix size 192 × 192, slice thickness (SL) = 1 mm] and multi-echo spin echo sequences (TR = 1200, TE = 10,20,30,40,50,60,70,80, FA = 90, matrix size 192 × 192, slice thickness 1.1, FOV = 14, TA = 13 min). To quantify the PET images, the scanner specific software (Image QC, GE Healthcare, Milwaukee, Wisconsin) was used. To quantify the MRI images, T2-relaxation time maps were generated and T2-relaxation times of each implant were measured.
Statistical analysis
All data are presented as mean ± standard deviation. Statistical significance was considered attained for P values < 0.05 based on Student’s t-test (two-tailed, unpaired samples).
