This article describes an optimized sequence of events for multimodal imaging of cellular grafts in rodent brain using: (i) in vivo bioluminescence and magnetic resonance imaging, and (ii) post mortem histological analysis. Combining these imaging modalities on a single animal allows cellular graft evaluation with high resolution, sensitivity and specificity.
During the past decade, stem cell transplantation has gained increasing interest as primary or secondary therapeutic modality for a variety of diseases, both in preclinical and clinical studies. However, to date results regarding functional outcome and/or tissue regeneration following stem cell transplantation are quite diverse. Generally, a clinical benefit is observed without profound understanding of the underlying mechanism(s)1. Therefore, multiple efforts have led to the development of different molecular imaging modalities to monitor stem cell grafting with the ultimate aim to accurately evaluate survival, fate and physiology of grafted stem cells and/or their micro-environment. Changes observed in one or more parameters determined by molecular imaging might be related to the observed clinical effect. In this context, our studies focus on the combined use of bioluminescence imaging (BLI), magnetic resonance imaging (MRI) and histological analysis to evaluate stem cell grafting.
BLI is commonly used to non-invasively perform cell tracking and monitor cell survival in time following transplantation2-7, based on a biochemical reaction where cells expressing the Luciferase-reporter gene are able to emit light following interaction with its substrate (e.g. D-luciferin)8, 9. MRI on the other hand is a non-invasive technique which is clinically applicable10 and can be used to precisely locate cellular grafts with very high resolution11-15, although its sensitivity highly depends on the contrast generated after cell labeling with an MRI contrast agent. Finally, post-mortem histological analysis is the method of choice to validate research results obtained with non-invasive techniques with highest resolution and sensitivity. Moreover end-point histological analysis allows us to perform detailed phenotypic analysis of grafted cells and/or the surrounding tissue, based on the use of fluorescent reporter proteins and/or direct cell labeling with specific antibodies.
In summary, we here visually demonstrate the complementarities of BLI, MRI and histology to unravel different stem cell- and/or environment-associated characteristics following stem cell grafting in the CNS of mice. As an example, bone marrow-derived stromal cells, genetically engineered to express the enhanced Green Fluorescent Protein (eGFP) and firefly Luciferase (fLuc), and labeled with blue fluorescent micron-sized iron oxide particles (MPIOs), will be grafted in the CNS of immune-competent mice and outcome will be monitored by BLI, MRI and histology (Figure 1).
1. Cell Preparation
2. Cell Implantation in the CNS of Mice
3. In vivo Bioluminescence
4. In vivo Magnetic Resonance Imaging
5. Post Mortem Histology
6. Representative Results
We here visually presented an optimized sequence of events for successful multimodal imaging of Luciferase/eGFP-expressing (stem) cell populations in the CNS of mice. At first, the described GB MPIO labeling procedure results in highly efficient labeling of BMSC-Luc/eGFP cells, which can easily be validated using fluorescence microscopy (Figure 2A). Next, upon grafting of GB MPIO labeled BMSC-Luc/eGFP in the CNS of mice, survival of the cellular graft can be monitored by in vivo BLI based on luciferase activity (Figure 2B). Additionally, the exact localization of grafted cells can be monitored by in vivo MRI based on the iron content of the GB MPIO (Figure 2C). Finally, histological analysis allows validation of the results obtained by BLI and MRI. Stable survival was confirmed by a stable eGFP expression in time (Figure 2D). For this, colocalization of eGFP-expressing cells with blue fluorescent MPIO allows for exact determination of localization and survival of grafted cells. In addition, this dual fluorescent labeling of the stem cells reduces the possibility to observe false positive results based on background fluorescence created by inflammatory cells5.
Figure 1. Overview of the handling sequence
Figure 2. Multimodal imaging of stem cell grafts
In this report, we describe an optimized protocol for the combination of three complementary imaging modalities (BLI, MRI and histology) for detailed characterization of cellular implants in the CNS of immune competent mice. A combination of reporter gene labeling of cells, based on genetic modification with the reporter genes firefly Luciferase and eGFP, and a direct cell labeling with GB MPIO, leads to an accurate assessment of stem cell grafts in vivo.
For particle labeling of BMSC, the GB MPIO particle size (1.63μm) in combination with the anionic surface (carboxyl substitutes) is suggested to facilitate endocytotic uptake of these particles by non-phagocytic cells16-18. However, it needs to be mentioned that the incubation time to obtain high labeling efficiency is cell type dependent. For example, phagocytic cells (like macrophages and dendritic cells) are able to internalize these particles more rapidly (incubation time of 0.5 – 1 hr) as compared to non-phaghocytic cell types (like BMSC or neural stem cells), which need an incubation time of minimum 16 hr for efficient labeling. This needs to be taken into account while performing cell labeling experiments and efficiency of labeling should always be determined in advance using fluorescence microscopy and/or flow cytometric analysis. As labeling of cell populations with GB MPIOs does not influence cell viability, cell proliferation or cell phenotype5, the combination of iron oxide and a fluorophore in these particles therefore creates an ideal opportunity to visualize them both by MRI and optical imaging. Moreover, due to the high iron content and large particle size, MPIO particles can be used for single cell12,19,20 and single particle imaging21 by MRI. This is extremely interesting when it comes to study cell migration as single cells can be observed. However, the high iron content of the used GB MPIO particles has also certain disadvantages when aiming to delineate the exact size of the graft site by MRI. The very high iron content of MPIOs not only creates magnetic inhomogeneities at the implant site, but also in the surrounding of the cellular implant, resulting in an overestimation of the implant site volume and a less accurate discrimination between implanted cells and the surrounding tissue. Consequently, the type of particle, needed for cell labeling, will depend on the study design. When aiming to determine stem cell migration, a very high sensitivity and thus high iron containing MPIOs will be needed. On the other hand, when aiming for the accurate localization of stem cell grafts, smaller SPIOs with a lower iron content might be a preferable alternative22. Moreover, iron particles generate negative contrast introducing another problem concerning the discrimination between implanted cells and bleeding following cell grafting. Therefore a lot of research is going on towards the use of positive contrast agents for cell labeling. Next to the particle type, the type of sequence used for the MRI acquisition also depends on the aim of the study. When MRI is used to delineate the implant to obtain accurate information on the location of the implant and the implant volume, the T2 sequence (spin echo) is favourable as it provides anatomical information with high accuracy. On the other hand, the T2* sequence (gradient echo) is a sequence used to study possible migration of grafted cells as this sequence takes into account all inhomogenities of the magnetic field rendering a higher sensitivity towards compounds influencing the magnetic field. This sequence is absolutely necessary when it comes to the detection of a small amount of cells that might have migrated out of the implanted region. Using this sequence we were able to conclude that BMSC-Luc/eGFP cells do not migrate after striatal implantation in the CNS of mice.
While MRI provides data regarding cell implant localization, additional BLI analysis provides data regarding cell implant survival2,4,11. As the enzymatic reaction between the luciferase enzyme and the substrate luciferin needs the presence of O2 and ATP, it is a reliable technique to study cell survival. Necrotic cells will not express the luciferase enzyme and no O2 and ATP will be present. Moreover, only cells expressing the luciferase enzyme are able to catalyze the reaction, resulting in the production of light, which resembles the high sensitivity of the technique. BLI can be performed in a quantitative manner because the amount of produced light is proportional with the amount of luciferase-expressing cells. However, important considerations need to be taken into account while performing BLI quantitatively as several factors other than cell amounts can influence the expression level of luciferase or the amount of detected light. For example, anesthesia level can influence the luciferase enzyme and/or the luciferin availability23, different factors and compounds can influence the substrate availability and/or uptake24-26 or promoter activity27-30, resulting in differences in signal output. Moreover the use of BLI is limited to small animals as the technique is restricted to low tissue penetration of light. Consequently, when aiming to perform quantitative BLI to follow up cell survival after transplantation, all parameters possibly influencing signal variations need to be kept as standard as possible. For example, implantation depth of cellular grafts, anaesthesia level and administration route / amount of the substrate administered before BLI acquisition, proper shaving of the animal’s skin, etc.
Taking into account the above described advantages and disadvantages of molecular imaging, the obtained results always need to be validated by histological analysis. Hereby various cellular phenotypes and cellular interactions between different cell types can be visualized with highest sensitivity post mortem. Despite the fact that this technique is the validation tool of choice, the presence of background fluorescence created by inflammatory cells surrounding and/or invading the graft site needs to be considered, as this might lead to false positive results. However, the use of a dual labeling with a fluorescent reporter gene, such as eGFP, and a fluorescent probe, such as GB MPIO, reduces the possibility of misidentification (i.e. grafted cells should display both specific fluorescences, without displaying background fluorescence in irrelevant light channels).
Summarizing, while BLI is a unique technique to monitor cell survival in time in a quantitative manner with very high sensitivity but low resolution, MRI is the technique of choice to overcome the low resolution obtained with BLI. Combination of these in vivo techniques renders sufficient information regarding survival and localization of grafted cells in vivo in a non-invasive manner. Finally, cellular graft interactions with surrounding tissue and/or endogenous inflammatory cells can easily be monitored post mortem using histology. Although, at the moment, the latter still needs to be done by histological analysis, the main challenges for the future will be to improve existing in vivo molecular imaging modalities to allow real-time assessment of the parameters with similar resolution and sensitivity as obtained using histological analysis.
The authors have nothing to disclose.
The authors work was supported by research grant ID-BOF 2006 of the University of Antwerp (granted to PPo and AVdL), by research grant G.0136.11 and G.0130.11 (granted to AVdL, ZB and PPo) and 1.5.021.09.N.00 (granted to PPo) of the Fund for Scientific Research-Flanders (FWO-Vlaanderen, Belgium), by SBO research grant IWT-60838: BRAINSTIM of the Flemish Institute for Science and Technology (granted to ZB and AVDL), in part by a Methusalem research grant from the Flemish government (granted to ZB), in part by EC-FP6-NoE DiMI (LSHB-CT-2005-512146), EC-FP6-NoE EMIL (LSHC-CT-2004-503569), and by the Inter University Attraction Poles IUAP-NIMI-P6/38 (granted to AVDL). Nathalie De Vocht holds a PhD-studentship from the FWO-Vlaanderen. Peter Ponsaerts is a post-doctoral fellow of the FWO-Vlaanderen.
Name of the reagent | Company | Catalogue number | Comments |
IMDM | Lonza | BE12-722F | Component of the cell growth medium CEM |
Fetal bovine serum | Gibco | 10270-106 | Component of the cell growth medium CEM |
Horse serum | Gibco | 1605-122 | Component of the cell growth medium CEM |
Penicillin-streptomycin | Gibco | 15140 | Component of the cell growth medium CEM |
Fungizone | Gibco | 15290-018 | Component of the cell growth medium CEM |
PBS | Gibco | 14190 | |
Puromycine | Invivogen | ant-pr-1 | |
trypsin | Gibco | 25300 | |
GB MPIO | Bangs Laboratories | ME04F/7833 | |
D-luciferin | Promega | E1601 | |
Ketamine (Ketalar) | Pfizer | ||
Xylazine (Rompun) | Bayer Health care | ||
Isoflurane | Isoflo | 05260-05 | |
0.9% NaCl solution | Baxter | ||
paraformaldehyde | Merck | 1.04005.1000 | |
sucrose | Applichem | A1125 | |
Micro-injection pump | KD scientific | KDS100 | |
Photon imager | Biospace Lab | ||
9.4T MR scanner | Bruker Biospin | Biospec 94/20 USR | |
BX51 microscope | Olympus | BX51 | |
Mycrom HM cryostat | Prosan | HM525 | |
syringe | Hamilton | 7635-01 | |
30 gauge needle | Hamilton | 7762-03 | |
Photo Vision software | Biospace Lab | ||
M3vision software | Biospace Lab | ||
Paravision 5.1 software | Bruker Biospin | ||
Amira 4.0 software | Visage Imaging |