Cells and labeling procedure
The study was approved by the committee of human research at our institution. Primary human mesenchymal stem cells (hMSC) were obtained from bone marrow (BM) of a patient with no known bone marrow pathology, who was admitted to our institution for trauma surgery and provided consent for intraoperative donation of hMSCs for research purposes. BM cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM) High Glucose media (Invitrogen, Carlsbad, CA, USA) supplemented with 10% fetal bovine serum (FMB, Hyclone, Logan, UT, USA) and 1% Penicillin-Streptomycin. All experiments were performed between passages 10–16 to avoid senescence.
The cells were labeled with DiD (C67H103CIN2O3S, Vibrant cell labeling solution, Molecular Probes, Oregon, USA), a lipophilic, cyanine near-infrared fluorochrome with a molecular weight of 1052Da and excitation and emission maxima of 644nm and 665nm respectively, as confirmed by spectrometry. The cells were incubated for 20 minutes with a labeling solution, consisting of 5μl of DiD and 1ml of serum free media per 1.0*106 MSC. The cells were washed, counted and viability tested by the trypan blue exclusion assay (Sigma Aldrich, St. Louis, MO, USA). Representative samples of DiD labeled hMSC and unlabeled controls were imaged using a Zeiss-LSM 510 confocal fluorescence microscope.
Optical Imaging System
All studies were performed using the IVIS 50 small animal scanner (Xenogen, Alemeda, CA) using the Cy5.5 filter set (excitation filter passband: 615–665 nm, emission filter passband: 695–770 nm, background filter passband: 580–610 nm) to match the absorption and emission characteristics of the labeling fluorophore. Specifically, while DiD exhibits its maximum emission intensity at about 665 nm, the emission spectrum extend to about 800 nm with about half of the emitted photon flux at wavelengths longer than 695 nm, which allows them to be captured by the imaging system using the Cy5.5 filter set. A detailed description of the imaging system is provided by Troy et al. [16
Identical illumination parameters (exposure time = 2 seconds, lamp voltage = high, f/stop = 2, field of view = 12, binning = 4) were selected for each acquisition. Gray scale reference images were also obtained under low-level illumination. For in vitro studies, cell samples were placed in a non-fluorescing black container. Whole-body real-time OI scans of anesthetized rats were acquired in prone and supine position pre-injection and, 4, 24, 48 and 72 hours post injection.
OI of DiD labeled stem cells
For all in vitro studies, the DiD-labeled hMSC were suspended in DMEM (isotonic solution) to preserve viability during imaging. Two sets of control experiments were performed to establish a basic understanding of signal localization and behavior. In the first experiment we evaluated the fluorescence intensity of a decreasing concentration of labeled cells (2 × 106, 1 × 106, 5 × 105, 2.5 × 105, 1.25 × 105 and 6.25 × 104) and 5 × 105 unlabeled controls in 0.5 ml of DMEM. In a second experiment, we evaluated the fluorescence intensity arising from a set of concentrations of cells (2 × 106, 1 × 106 and 5 × 105) and 5 × 105 unlabeled controls in 0.5 ml of DMEM at day 0, 2, 4, 6 and 8 after labeling.
Animals and arthritis induction
The Committee on Animal Research at our institution approved this study. Fourteen 4- to 6- week-old female homozygous athymic nude rats (Harlan, Indianapolis, Indiana, USA; 150–250g) were used in this study as they permitted the administration of allogenic MSC. Standard rodent chow caused significant production of autofluorescence thus the rats were fed a manganese-free diet (ssniff R/M-H, ssniff Spezialdiaeten GmbH, Soest, Germany) throughout the study. An immune mediated polyarthritis was induced under isofluorane anesthesia by an intra-peritoneal (960μl) injection of 1.0ml (5.2mg) of Peptidoglycan-Polysaccharide (PG PS 10S) (Fischer Scientific, Pittsburg, PA, USA), a compound composed of fragments of streptococcal cell walls. Although the animals have a well-known impaired immune response, previous studies have shown that the rats nonetheless develop an immune mediated arthritis with this agent [17
]. The animals were observed daily for clinical signs of arthritis (joint swelling and limping) and underwent cell injection and OI when a polyarthritis of the ankle joints had developed, which occurred on day three post-injection (p.i.). The medio-lateral diameter of the ankle joints was measured with a caliper under anesthesia before PGPS10S injection and directly before the optical imaging studies, when clinical signs of arthritis had appeared. At this point, n = 11 anesthetized rats received an intra-peritoneal injection of 3×106
DiD-labeled hMSC in 0.5ml of serum free DMEM. Three additional control animals received an intra-peritoneal injection of 15μl of DiD in 0.5ml of serum free DMEM (n = 1), PGPS arthritis induction but no cell or dye injection (n = 1) and no arthritis induction and no dye injection (n = 1).
OI Image analysis
Images were acquired and analyzed using Living Image 2.5 software (Xenogen, Alameda, CA, USA) integrated with Igorpro (Wavemetrics, Lake Oswego, OR, USA). The digitized image intensity is expressed in arbitrary units as the fluorescent image is divided by a reference image (image of a reference object) to account for the spatial distribution of the excitation light. It must be recognized that the as recorded images contain three main image components arising from a) the fluorescence of DiD, b) the native fluorescence (autofluorescence) of the cells or experimental animals and c) a scattering image of the object arising from a small leakage of light through the excitation filters, an inherent limitation of the imaging system [16
For the in vitro experiments, image analysis involved the designation of regions-of-interest (ROI) as the circular area of the well containing the cell concentrations to extract the average intensity (sum of the intensity of all pixels within this ROI divided by the number of pixels).
For the in vivo experiments, image analysis involved the definition of the ROI in each location of enhanced fluorescence as the area having intensity larger than 50% of the peak signal intensity. This ROI was automatically selected by the imaging software. The operator also defined regions of interest where there is no detectable fluorescence from DiD and served as reference representing the background (autofluorescence and leakage) components. The average normalized intensities from these ROIs of the background were also recorded using a second set of excitation filters, referred to as background filter in the Optical Imaging System
section. This second filter provides excitation of the target with light that is out of resonance with the absorption spectrum of DiD, thus generating an image where the autofluorescence and leakage components are the dominant contributors. The ratio of the average intensity from the background ROIs under the two excitations provide the means to monitor the stability of the background components in time and can be used to further normalize the data against such changes [16
After the last imaging procedure, the animals were sacrificed and the ankle joints, liver, lymph nodes, heart and lungs were harvested. Additionally, the ankle joints were decalcified. The samples were then bisected parasagitally, dehydrated, paraffin embedded and sectioned into 5μm transverse slices. These sections were stained with hematoxylin and eosin (H&E) for evaluation of cell morphology, with diamidino-2-phenylinole (DAPI) for core-counterstaining and localization of DiD fluorescence, and with CD44 immunostains for detection of hMSCs.
The statistical analysis was performed using SAS software (SAS Institute Inc, V6.9, Cary, NC, USA). Measured raw fluorescence signal intensity was described as means and standard deviations for the various experimental groups. A two-way analysis of variance (ANOVA) was used for in-vitro experiment with the dependent variable of the optical intensity and the two categorical dependent variables of days and the concentrations. A linear mixed random effects model was used for the in-vivo experiments and applied to ROI of each location separately. The animal identity was the random factor. The signal intensity was the dependent variable. The observation time was treated as a categorical variable. All three controlled animals were labeled as one control group. Interaction between treatment group and observation time were used to assess the treatment difference at different time points. Linear contrasts were used to determine the significance of change from baseline due to treatment effect at a specific observation time. P-values were adjusted for multiple comparisons using Tukey-Kramer’s method. Statistical significance was assigned for a p < 0.05.