In Vivo Tracking of Murine Adipose Tissue-Derived Multipotent Adult Stem Cells and Ex Vivo Cross-Validation

Stem cells are characterized by the ability to renew themselves and to differentiate into specialized cell types, while stem cell therapy is believed to treat a number of different human diseases through either cell regeneration or paracrine effects. Herein, an in vivo and ex vivo near infrared time domain (NIR TD) optical imaging study was undertaken to evaluate the migratory ability of murine adipose tissue-derived multipotent adult stem cells [mAT-MASC] after intramuscular injection in mice. In vivo NIR TD optical imaging data analysis showed a migration of DiD-labelled mAT-MASC in the leg opposite the injection site, which was confirmed by a fibered confocal microendoscopy system. Ex vivo NIR TD optical imaging results showed a systemic distribution of labelled cells. Considering a potential microenvironmental contamination, a cross-validation study by multimodality approaches was followed: mAT-MASC were isolated from male mice expressing constitutively eGFP, which was detectable using techniques of immunofluorescence and qPCR. Y-chromosome positive cells, injected into wild-type female recipients, were detected by FISH. Cross-validation confirmed the data obtained by in vivo/ex vivo TD optical imaging analysis. In summary, our data demonstrates the usefulness of NIR TD optical imaging in tracking delivered cells, giving insights into the migratory properties of the injected cells.


Introduction
Optical imaging encompasses several desirable characteristics: it is rapid, noninvasive and nontoxic (it is not based on radiation). For these reasons, it is the optimal tool for performing long-term longitudinal studies in vivo [1], given the possibility to adapt experimental protocols to different �elds of investigation. Speci�cally, time domain (TD) optical imaging technology allows for whole-body near infrared (NIR) �uorescence lifetime analysis, based both on the speci�city of �uorescence probes and the sensitivity of their emission lifetime to environmental characteristics [2]. Different kinds of probes can be conjugated with �uorescence dyes: antibodies [3], polysaccharides [4], peptides [5], and also cells can be imaged to evaluate their in vivo biodistribution [6].
At present, clinical optical imaging is an emerging �eld, and its promising results are supported by preliminary investigations on sentinel lymph node tracers [7] and on peripheral tissue perfusion [8]. In both cases, indocyanine green was used as NIR �uorophore. Moreover, extensive studies have been conducted to con�rm sensitivity and tissue diagnostic imaging potentiality of nanoparticles-based NIR contrast agents, such as quantum dots, resonant gold nanoshells, and dye-encapsulating nanoparticles [9].
In this study, we evaluate the applicability of the TD preclinical optical imaging system Optix to follow in the mouse the biodistribution of NIR-labelled murine adipose-tissuederived stem cells (mAT-MASC). Stem cells are de�ned as undifferentiated cells able to both self-renew in the longterm and to differentiate into specialized cell types. Several studies have tried to dissect the mechanisms regulating the fate of stem cells residing in different tissues [10]. Skeletal muscle dystrophies comprise a heterogeneous group of neuromuscular disorders characterized by progressive muscle wasting [11], for which no satisfactory treatments exist [12]. Numerous different therapeutic strategies have been devised to correct the dystrophic phenotype [13]. In this regard, multiple stem cell populations, both of adult or embryonic origin, have been assayed for their myogenic ability. To date, many of these strategies have failed, thus, underlying the need to identify the mechanisms controlling myogenic potential, to avoid immune response, and to promote homing and engrament of donor population to the musculature. In particular, it would be important to target lifesaving muscles, such as heart and diaphragm, which are involved in many dystrophies but are extremely difficult to access.
Preliminary results showed that mAT-MASC were able, in vitro, to differentiate along myogenic lineages and, importantly, were characterized by a wide in vivo migratory ability, even when injected intramuscularly (i.m.). Nonetheless, in the present work we did not investigate how injected cells could undergo proliferation and differentiation into speci�c phenotypes, but we decided to optimize a multidisciplinary approach to evaluate the ability of TD optical imaging technology to monitor the in vivo distribution of i.m. injected stem cells. Speci�cally, male, DiD-labelled, and eGFP expressing mAT-MASC were delivered into a healthy, congenic, female recipient. Cell presence was evaluated, respectively, by in vivo NIR TD optical imaging and Cellvizio Lab dynamic �beroptic �uorescence microscopy, by ex vivo TD optical imaging, qPCR, immuno�uorescence associated with lambda-scan analysis, and FISH for Y-chromosome techniques. e results con�rmed the potential of the applied complement of optical imaging techniques to be a complete and useful tool for tracking cell biodistribution and homing in small animals.

Flow
Cytometry. Cells at the third passage in culture (P3) were detached and stained with the following properly conjugated antibodies: CD90, CD34, CD105, CD117, MHC-Ia/Ie, and MHC-1b (all from Santa Cruz Biotechnology), CD133, SCA-1, and CD9 (from eBioscience), CD44 and CD45 (from BD). Properly conjugated isotype matched antibodies were used as negative controls. e analysis was performed by CyAn (Beckman Coulter), and both, fraction of positive cells and intensity of expression, were evaluated for all antigens.

mAT-MASC Labelling with
DiD. mAT-MASC cells were collected aer trypsinization and suspended at a concentration of 1 × 10 6 /mL in serum-free culture medium. Following, 5 L of Vybrant DiD cell-labeling solution (Molecular Probes) was supplied per mL of cell suspension. DiD is a lipophilic carbocyanines tracer with markedly red-shied �uorescence excitation and emission spectra which can be used for cellular adhesion studies and migration applications [16]. Once incorporated into membranes, DiD is highly �uorescent and photostable. e supplier reported high extinction coefficients (EC > 125,000 cm − M − at their longest-wavelength absorption maximum) and short excitedstate lifetimes (∼1 nanosecond) in lipid environments.
Cells were incubated for 1 hour at 37 ∘ C under agitation in the dark. Aerwards, the suspension was centrifuged at 900 g and washed three times with PBS to remove the free dye. At that time, DiD-labelled cells (1 × 10 6 /200 L serum-free culture medium) were seeded on an optical 96-wells-plate and analyzed with Optix in order to verify that the labelling process had been successfully succeeded.

Animal Model and Treatment for Optical Imaging Scan.
Female C57/BL6 mice of 6-8 weeks old ( were purchased from Harlan (San Pietro al Natisone, Italy) and maintained under pathogen-free conditions. Mice were anesthetized using a gaseous anaesthesia system (Biological Instruments, Italy), based on iso�urane mixed to oxygen and nitrogen protoxide. Anaesthesia was �rst induced in a preanaesthesia chamber (2% iso�urane), and then the mouse was placed on the heated imaging bed of the Optix under 1% iso�urane. Moreover, mice were shaved in the regions of interest in order to avoid fur laser scattering. e experimental mice were injected in the le tibialis anterior muscle with 3 × 10 6 mAT-MASC eGFP -DiD cells.
All the experimental procedures were conducted in compliance with the guidelines of European (86/609/EEC) and Italian (D.L.116/92) laws and were approved by the Italian Ministry of University and Research. In all imaging experiments, a 670 nm pulsed laser diode with a repetition frequency of 80 MHz and a time resolution of 12 ps light pulse was used for excitation. Fluorescence emission was collected at 700 nm and detected through a fast photomultiplier tube (PMT) coupled to a time-correlated single-photon counting system. Two-dimensional scanning regions of interest (ROI) were selected, and laser power, integration time, and scan step were optimized according to the emitted signal. e data were recorded as temporal point-spread functions, and the images were reconstructed as �uorescence intensity and �uorescence lifetime maps. All the in vivo analyses were preceded by native scans of the mice prior to injection of the labelled cells in order to provide a baseline for later analyses. Animals were followed for 48 hours ( or 7 days ( aer the injection.

Ex Vivo Time Domain
Optical Imaging Analysis. Aer the last imaging session was performed, mice were sacri�ced by cervical dislocation in deep anaesthesia. e tissues of interest, such as the injected and the contralateral muscle, adipose tissues, brain, heart, diaphragm, liver, �exor digitorum brevis (FDB), spleen, and lung, were collected, washed with PBS twice, and imaged with the Optix system.

Data Processing.
To estimate �uorescence intensity and lifetime, the background signal intensity recorded with the baseline image for each animal before the injection of the labelled cells was subtracted from each postcontrast image using Optix OptiView soware [17].

Fluorescence Intensity Analysis.
Fluorescence intensity values are reported in normalized counts (NC) representing the photon count for unit excitation laser power and unit exposure time to allow comparison between different images. To perform the comparison, an identical ROI on each image was positioned to encompass the area under investigation.
2.9. Fluorescence Lifetime Analysis. e �uorescence lifetime results are obtained by �tting every �uorescence decay curve corresponding to each pixel measured by Optix using the Levenberg Marquet least squares method [18].

In Vivo Dynamic Fluorescence Microscopy Cellvizio Lab.
A probe-based confocal �uorescence microscope Cellvizio Lab (VisualSonics Inc. Toronto, Canada) was used. In all experiments the �eld of view was between 300 and 650 m, resolution from 1.4 to 3.5 m, the frame rate from 8 to 200 fr/sec, and excitation laser 680 nm. Two healthy animals were treated i.m. with 3 × 10 6 mAT-MASC eGFP -DiD in the le tibialis anterior muscle and analyzed by Cellvizio Lab 48 hours aer the injection. Speci�cally, the day of the analysis mice were anaesthetized with 70 L of a solution 12% Zoletil 100 plus 7.5% Rompun 2% in Physiological saline solution i.m. to induce deep surgical anaesthesia and analgesia. Eyes were treated with an ophthalmic gel to prevent drying. Legs were shaved, and the area was disinfected with an iodopovidone solution. A small incision was made on the skin in the region of the thigh muscle with a 18G needle under sterile conditions, and the optical probe was inserted to record the images. At the end of the acquisition process, the incision was sutured with surgical and sterile silk.
Tissue specimens were collected either from mice injected with mAT-MASC eGFP -DiD or PBS. Excised tissues were partially formalin �xed and para�n embedded and, in addition, part snap frozen. Immuno�uorescence staining was performed on 5 m thick tissue sections. Primary antibodies, antigen retrieval, and staining protocols are listed in Table  1. Nuclei were stained by DAPI (Vector Laboratories, Inc), and Vectashield (Vector) was used as mounting medium. Epi�uorescence and phase contrast images were obtained with a live cell imaging dedicated system consisting of a Leica DMI 6000B microscope connected to a Leica DFC350FX camera (Leica Microsystems); 10X (numerical aperture: 0.25), 40X oil immersion (numerical aperture: 1.25), and 63X oil immersion (numerical aperture: 1.40) objectives were employed. Confocal images were collected using Confocal Laser Microscope (Leica TCS-SP2, Leica Microsystems, Wetzlar, Germany), utilizing a 63X oil immersion objective (numerical aperture: 1.40) or a 40X oil immersion objective (numerical aperture: 1.25). Adobe Photoshop soware was utilized to compose and overlay the images and to adjust contrast (Adobe, USA).

Fluorescence In Situ Hybridization (FISH). Cambio's
StarFISH Mouse Whole Chromosome-Speci�c paint kit was used to detect donor Y-chromosome in frozen sections. e probe was Cy3 labelled. Detection was performed following the protocol provided by the company. Sections obtained from female mice injected with PBS were employed as negative controls, while sensitivity was determined on sections obtained from untreated male mice.
2.13. Lambda Scan Analysis. is assay was performed using a Leica TCS-SP2 confocal microscope. Analyses were performed on murine sections (aer appropriate immunolabelling) of mice injected with eGFP + DiD − cells, eGFP − DiD + cells, or eGFP + DiD + cells. A555-labelled secondary antibody was utilized to recognize anti-eGFP primary antibody. e emission signal for Alexa555 was excited at 543 nm with an argon laser, and its �uorescence intensity was recorded generating a lambda stack ranging from 563 to 798 nm at 5 nm intervals. e emission signal for DiD was excited at 633 nm with a helium/neon laser, and its �uorescence intensity was recorded generating a lambda stack ranging from 653 to 798 nm at 6 nm intervals. e lens and corresponding numerical aperture were 63X and 1.4, respectively. Sampling consisted of 30 eGFP + and/or DiD + cells. Additionally, 10 cells negative for these markers and present in the same samples were used as control to discriminate background auto�uorescence from speci�c labelling. Each determination was restricted to a region of interest (ROI) comprised within each DiD and/or eGFP-positive cell. For each ROI, a graph plotting mean pixel intensity and the emission wavelength of the lambda stack was generated.
2.14. DNA Extraction and qPCR Analysis. Genomic DNA for PCR analysis was puri�ed from frozen tissues using QIAmp DNA Mini Kit (QIAGEN). Weighed biopsies obtained from 2 mice injected with mAT-MASC eGFP -DiD, 2 untreated mice, and two C57BL/6-Tg[CAG-eGFP]1Osb/J[Jackson Laboratory] were incubated for 96 h at 56 ∘ C in lysis buffer and processed to purify the DNA, as recommended by the manufacturer. DNA concentration was estimated by using NanoDrop 2000 (ermo Scienti�c), and quantitative realtime PCRs were performed using LightCycler 480 realtime PCR system (Roche). Ampli�cation reactions were performed using manufacturer-provided reagents following the standard recommended ampli�cation conditions (LightCycler 480 Probes Master). 50 ng of puri�ed DNA from various tissues were ampli�ed. Roche's Universal ProbeLibrary Assay platform was used to design primers and �nd suitable internal probes. e primers and probes for the target gene eGFP were forward primer 5 ′ -GCATCGACTTCAAGGAGGAC-3 ′ and reverse primer 5 ′ -GTTGATGTTGTCGGTGTTGCAG-3 ′ ; the probe labelled with �uorescent reporter and quencher was UPL PROBE#78 (Roche's Universal ProbeLibrary). As control, mouse beta-actin was ampli�ed: forward primer 5 ′ -CCATCTTGTCTTGCTTTCTTCA-3 ′ and reverse primer 5 ′ -ATGAGACACACCTAGCCACC-3 ′ ; the probe was UPL PROBE#63 (Roche's Universal ProbeLibrary). To determine ampli�cation e�ciency and to estimate, for each specimen, the absolute concentration of both eGFP gene and betaactin gene, calibration curves for eGFP gene and betaactin gene, respectively, were constructed. Negative control for the target gene was isolated from mice that did not undergo transplantation. Positive control DNA was isolated from C57BL/6-Tg[CAG-eGFP]1Osb/J[Jackson Laboratory]. To quantify the number of transplanted eGFP cells in mouse tissues, we calculated the ratio between beta-actin and eGFP copy numbers and multiplied this value per 100000. erefore, we could express the ratio of eGFP DNA copy numbers per 100000 murine cells.

Multipotent Adult Stem Cells Were Isolated from Mouse-Derived Adipose Tissue and Characterized.
In order to obtain mAT-MASC (i.e., multipotent adult stem cells from murine adipose tissue samples) we applied, with minor modi�cations, the method previously described for the isolation of human MASC from liver, heart, bone marrow [14], and peripheral blood [15]. e isolation protocol allowed establishment and in vitro expansion of cell lines, named mAT-MASCs, from adipose tissue fragments obtained both from . mAT-MASC were obtained from all samples, con�rming the high reproducibility and efficiency of the optimized method [14,15]. All cell lines were grown on �bronectin coated dishes, in an expansion medium supplemented with mPDGF-BB, mEGF, and containing 2% FBS. Primary cultures reached 80% of con�uence within one week, and, at the third passage in culture (P3, 20th-25th generation), mAT-MASC displayed a homogeneous �broblast-like morphology (Figure 1(a)) and had a population doubling time of ± 6 hours. As expected, mAT-MASC obtained from transgenic mice (mAT-MASC eGFP ) showed a variable expression of eGFP protein (Figure 1(a)). e surface immunophenotype of wild-type and mAT-MASC eGFP -DiD was assessed by �ow cytometry at the third passage in culture ( . All cell lines shared a similar mesenchymal immunophenotype characterized by a larger fraction of cells expressing CD90, CD9, CD13, and Sca-1, while CD45, CD117, CD105, CD34, MHC-I, MHC-II, and CD44 were expressed in a minority of the cells (Figure 1(b)). is antigenic pattern was very similar to the one described by Verfaillie's group for murine MAPCs [19]. To evaluate if adipose-derived cell lines displayed stem cell properties, we tested the expression of transcription factors known to be associated with a pluripotent state [20]. As shown in Figures  1(c)-1(f), Oct-4, Nanog, and Sox-2 proteins were detected, by immuno�uorescence, in a large fraction of mAT-MASC at the third passage in culture, con�rming the undifferentiated state of the cultured cells. In order to evaluate whether mAT-MASC were characterized by the ability to differentiate into multiple mature cell types of mesodermic origin, cells were cultured under appropriate differentiation inducing conditions (see Section 2), and the acquisition of functional as well as molecular evidence of differentiation was assessed. Cells cultured in an osteogenic medium exhibited calcium deposits (Von Kossa staining, Figure 1(g)), while the capability of mAT-MASC to differentiate into adipocytes was demonstrated by their ability to store neutral triglycerides and lipids (Oil Red-O staining, Figure 1(h)). e ability of mAT-MASC to differentiate into muscle cells was tested in a medium containing VEGF, bFGF, and IGF-1. Aer the differentiation period, a fraction of cells, lower than 30%, expressed organized �laments of smooth-muscle actin (SMA) (Figure 1(i)), while almost all cells showed organized �laments of alpha-sarcomeric actin (ASA) (data not shown). e presence of functional competent receptors involved in calcium handling was demonstrated by spontaneous intracellular calcium transients, as displayed by Fluo-4 assays (Figures 1(j)-1(k)). Altogether, the accumulated evidence shows that mAT-MASC represent a population of primitive, multipotent cells easy to obtain and expand in vivo and, therefore, they are a suitable cell source for in vivo regenerative study. labelling with TD optical imaging was applied. Cells labelled with DiD were injected i.m. into the le tibialis anterior muscle of healthy mice. Whole body scans showed the highest �uorescence signal in the region of injection, while intensity decreased exponentially over time according to the cells migration (Figure 2(a)). In order to optimize �uorescence acquisition parameters, the cells injection region was maintained outside the scanning area, and a manually selected region of interest (ROI), encompassing the contralateral leg, was investigated. e �uorescence signal increased slowly and was signi�cantly higher than precontrast from 24 hours until 7 days aer injection (Figure 2 (Figure 2(c), D, and E). is result provided the possibility to discriminate between the labelled cells and the free dye, in order to con�rm that the signal under investigation in vivo was not amenable to DiD itself. e summary of lifetime is reported in Table 2: to compare the data, all the values were acquired in a normalized range between 3 and 9 ns.

Ex vivo TD Optical Imaging Con�rmed the In Vivo Res�lts.
At the end of the experiments, mice were sacri�ced by cervical dislocation, and the organs were excised and washed in PBS. en, legs muscle, brain, heart, diaphragm, liver, FDBs, spleen and lung were analyzed by Optix optical imaging. is way, any unspeci�c contribution due to auto�uorescence from other organs and absorption and scattering within the body and fur could be avoided, thereby increasing both speci�city and sensitivity of the probe detection. Ex vivo evaluation of organs at 48 hours aer injection clearly showed that the highest �uorescence emission was detected in the liver (Figure 3 Table 2). (c) In vitro lifetime analysis of mAT-MASC eGFP -DiD (D) and free DiD (E): the different trend of the curves indicates the possibility to discriminate by a lifetime analysis between the labelled cells and the pure dye (for the values see Table 2). and eliminated bac�ground signal. �ifetime gating con�rmed the speci�city of the signals with a common mean value of 1,6 ns (Figure 3(b)), and lifetime curves analysis reported comparable trends for all the excised tissues (Figure 3(c)), e analysis performed in the animals sacri�ced one wee� aer the cell injection showed the presence of a speci�c signal only in the injected muscle and in the surrounding fat, but not in the other analyzed tissues (data not shown). e result  e graph representing lifetime trends con�rms the comparability between the signals recorded from the di�erent organs ex vivo, which are represented by the same slope of the curves. e colors, respectively, represent yellow, liver; green, heart; red, brain; white, lung; grey, diaphragm; blue, spleen; pink, contralateral muscle.
suggested the hypothesis that, in this interval, mAT-MASC-DiD are probably spread in the body in a concentration too low to be detectable by Optix. Only in the region of injection, which was characterized by a high starting concentration, there is a weak but detectable signal. Finally, the in vivo and ex vivo optical imaging procedures here previously described, which were optimized to monitor cells migration, are purely qualitative and not quantitative, and the �gures reported are representative. For this reason, no intersubject variation relative to the migration of labelled cells has been performed.

mAT-MASC-DiD Cells Migration Can Be Visualized by In Vivo Dynamic Fluorescence Microscopy Cellvizio Lab.
e application of Cellvizio �ab �beroptic �uorescence microscopy allowed us to collect longitudinal high resolution data with a low invasiveness in living animals. It was possible to study cell homing and migration in realtime in vivo 48 hours aer cells injection. e results obtained con�rm that in the le injected leg there was an abundance of mAT-MASC eGFP -DiD cells (Figure 4(a)) (see Video 1, Supplementary Material available online at http://dx.doi.org/10.1155/2013/426961) and that in the right contralateral leg it was feasible to image individual cells which had migrated from the injection site (Figure 4(b)) (Video 2, Supplementary Material). Finally, while individual cells could not be imaged by means of Optix system, in vivo �beroptic �uorescence microscopy allowed us to track single cells that were heterogeneously dispersed, con�rming their ability to migrate in vivo. PCR was performed. Genomic DNA was extracted from frozen tissues of mice injected with 3 × 10 6 mAT-MASC eGFP -DiD, not injected mice (negative control), and transgenic mice constitutively expressing eGFP (positive control). To evaluate the efficiency and to calculate the regression r value, serial dilutions of genomic DNA extracted from eGFP + and wild type mice were used to create distinct standard curves (Figures 5(a) and 5(b)). Absolute concentration of both, eGFP and beta actin gene, was extrapolated from each standard curve ( Figure 5(c)). e relative amount of eGFP expressing cells in each of the evaluated tissues was estimated as a ratio of eGFP and beta actin gene copies and normalized to 100,000 nuclei ( Figure 5(c)). As reported in the graph, only a small number of mAT-MASC eGFP -DiD cells could be detected 7 days aer cell injection into the major �lter organs like the spleen and the lung (between 1/and 10/100,000 total cells), but never in the liver. As expected, the injected muscle showed the highest mAT-MASC eGFP -DiD cell content (10,000 ± 1,361 eGFP + cells/100,000 total cells). Moreover, we con�rmed the presence of eGFP + cells in tissues far from the injection site, like the contralateral tibialis anterior muscle, indicating that cells could migrate and persist in tissues other than the injection site for at least 7 days, even if no pathological conditions were induced in the recipient animal. Of great interest, both diaphragm and heart contained eGFP + cells ( Figure 5(c)). is is of paramount importance, considering that both organs can be severely involved in muscular dystrophies [21] and they are notoriously difficult to reach for cell-based therapy techniques [13].

�.�. In Situ I�enti�cation an� Characteri�ation o� �A�-MASC -DiD into the Injection Site.
With the aim of con�rming the presence of and to localize mAT-MASC eGFP -DiD in situ, two independent techniques were adopted: (1) eGFP detection by immuno�uorescence in association with lambda scan analysis and (2) �uorescence in situ hybridization (FISH) speci�c to the murine Y chromosome. e in situ presence of mAT-MASC eGFP was further demonstrated by using an antibody recognizing the endogenous expression of the eGFP protein ( Figures 5(d)-5(e)). To validate the speci-�city of the recorded signals, spectral analysis was performed [22]. e emission spectra for GFP and DiD were clearly distinct from the emission spectra for tissue auto�uorescence, con�rming the accuracy of the immunolabelling protocol ( Figures 5(g), 5(i), and 5(k)).
As shown in panel k and l, cells simultaneously displaying spectra resembling the single emissions of DiD or Alexa 555 (�uorophore conjugated to the secondary antibody developed against speci�c eGFP primary antibody) were speci�cally found in injected muscle (red lines) and not in tissue sections obtained from not-transplanted mice (grey lines).
Since male mAT-MASC eGFP -DiD were i.m. injected into female recipients, the presence, in injected female muscle sections, of Y-chromosome positive cells supported the evidence that viable donor cells persisted in tibialis anterior muscle (Figures 5(m)-5(o)).
Vitality and proliferative rate acquired by donor cells in vivo were further assessed by costaining cells for eGFP and MCM5 (Figures 5(d)-5(f)), a protein involved in the control of DNA replication. We estimated that 51 ± 5.7% of eGFP-positive cells were positive for the proliferation marker suggesting that engraed cells retain the ability to duplicate in vivo.

Discussion
Here, NIR TD optical imaging technology for in vivo longitudinal studies based on NIR dye labelled-cells and �uorescent lifetime analysis was applied, and these characteristics enable both high speci�city and sensitivity for biodistribution studies due to the dyes' spectral characteristics [1,2]. Speci�cally, we evaluated the migratory ability of mAT-MASC injected i.m. in mice. mAT-MASC were cultured utilizing methods previously optimized to expand multipotent adult stem cells [14,15]. To investigate whether TD optical imaging could track in vivo mAT-MASC, 3 × 10 6 DiD-labelled cells were injected into the le tibialis anterior muscle of a congenic wild-type female recipient. Animals were monitored for a short (48 hours) time and also for an extended period (7 days) to evaluate cells biodistribution over time. Injected animals that were analyzed within the �rst 48 hours aer injection by TD optical imaging showed a speci�c signal in the region of the injection that decreased with time. Conversely, the �uorescence intensity analysis performed in the contralateral leg revealed a slow increase with a peak signal aer 48 hours. To assess the cell migration and to see if the signal can be discriminated from the free dye, we performed lifetime analysis, based on the comparison between lifetime values obtained for the injection site and the contralateral leg. e relative curves obtained showed similar trends and comparable values (1.50 and 1.56 ns, resp.). Moreover, it can be assumed that the signal is not due to the free dye because it was demonstrated that labelled cells and DiD have different speci�c lifetime values (1.55 and 1.10 ns resp.).
Ex vivo analysis performed 48 hours aer the injection con�rmed the data obtained in vivo, and in addition it revealed the presence of a speci�c signal in �lter organs such as the lung, spleen, and liver, as well as the brain, adipose tissue, le and right �exor digitorum brevis, diaphragm and heart. is homogeneous presence of mAT-MASC in all the organs analyzed could be traced to the nonspeci�c biodistribution of the cells in the short period that had elapsed aer treatment. Ex vivo analysis and in vivo TPSF curve evaluation suggested the presence of mAT-MASC eGFP -DiD in the leg opposite the injection site, supporting the hypothesis that the cells reached, possibly through systemic distribution, distant sites.
To further investigate and demonstrate the speci�city of the signal cell migration was analysed 48 hours following injection by a �bered confocal microendoscopy system, which provided a direct, rapid, and accurate visualization of labelled cells at the muscle in both legs.
In order to evaluate whether mAT-MASC eGFP -DiD could persist longer, a new group of animals was analysed daily by Optix preclinical optical imager for 7 days aer the injection. A progressive decrease in �uorescence signal in the region of the treatment was observed, and at one week, no speci�c signal could be detected in the leg opposite the injection site. It could be supposed that one week aer the treatment the cells are too dispersed throughout the body or that the dye is too diluted to be detected by Optix. is result gave a complete overview of how much the experiment can be projected into a longitudinal scheme.
At the same time point of one week aer cells injection, the in vivo results have been con�rmed by ex vivo optical imaging analysis where no speci�c signal had been detected neither analysing the organs singly, except for the injected muscle.
Considering a potential microenvironmental contamination by cell-free DiD, tracking of labelled cells using nongenetically encoded markers should always be accompanied by cross-validation using multimodality approaches [21]; therefore, we used cells that could be recognized by different in vivo and ex vivo techniques. Speci�cally, mAT-MASC were isolated from male mice expressing constitutively eGFP as marker gene product, which was readily detectable using techniques of immuno�uorescence (eGFP protein) and qPCR (eGFP gene). Moreover, male cells injected into wildtype female recipients were identi�ed by FISH detection of the Y-chromosome. An assay aimed at detecting both eGFP and beta-actin genes was optimised, thus allowing quanti�cation of the relative number of eGFP + cells into wild type tissues. As expected, the highest number of eGFP + cells was detected at the injection site, where 10% of the nuclei were estimated to harbour the transgene. In all the other analyzed tissues the estimated number of eGFP + nuclei was extremely low, between 1 and 10 cells per 10 5 total nuclei. Engraed cells found at the site of implantation were viable and proliferating, and almost half of them were differentiated into myocytes and CD146 positive vascular cells (data not shown). Despite the fact that lifetime analysis is able to distinguish DiD-labelled cells with high speci�city, the sensitivity of the technique does not allow detection of very rare cells (1-10 eGFP + /10 5 nuclei); therefore additional techniques must be applied with this purpose.

Conclusion
In the present work, the results showed the utility of using an imaging technologies panel to track delivered cells, especially during the �rst days aer injection, giving insights into their migratory properties. Cells biology and differentiation capability were not the speci�c goals of this paper: the experiments have been originally planned to de�ne an appropriate tool to obtain a cross-validated study, both in vivo, for analyzing cells biodistribution and their speci�c localization, and ex vivo, for evaluating their accumulation and properties. In particular, knowledge about the biodistribution of mAT-MASC was acquired: such as con�rmation of migratory properties and, at the same time, monitoring them using different imaging applications, long and short times aer injection for evaluation.
It can be concluded that the multimodality approach we applied could be of paramount importance in the study of systemic diseases, where it could be utilized to monitor the effect of immunomodulatory or other pharmacological interventions on engraed cell number and distribution. Finally, the proven ability of mAT-MASC eGFP -DiD to migrate to distant targets suggests that they could represent a suitable cell source to be utilized in systemic diseases, such as the skeletal muscle disorders, and that the future perspectives would be to further investigate mAT-MASC behaviour, proliferation, and differentiation in vivo.