Biological Trace Element Research

, Volume 126, Issue 1, pp 83–91

MRI-based Visualization of Iron-labeled CD133+ Human Endothelial Progenitor Cells


    • Department of Oncology, Guang An Men HospitalChina Academy of Chinese Medical Sciences
  • Baoting Chao
    • Shandong Medical Imaging Research InstituteShandong University

DOI: 10.1007/s12011-008-8192-x

Cite this article as:
Liu, J. & Chao, B. Biol Trace Elem Res (2008) 126: 83. doi:10.1007/s12011-008-8192-x


The objective of this study is to build up a kind of effective approach to multiply CD133+ endothelial progenitor cells (EPCs) and visualize cells by labeling with two FDA-approved agents based on MRI technique. CD133+ cells were separated by immunomagnetic microbeads selection and grew with serum-free medium. Seven days later, CD133+ cell production was collected and co-incubated with iron complex for 24 h for labeling. The iron-labeled cells were suspended into agarose gel and scanned by MRI for visualization. Labeled cells were also analyzed for cell viability. Iron can be effectively introduced into CD133+ EPCs plasma in culture and visualized by changing the MRI signal intensity. Iron had no influence on cell viability. Conclusion: Iron substance can be applied to label CD133+ cells without cytotoxicity and iron-labeled cells can be visualized by MRI image. Due to the non-invasive property and repeatability of MRI technology and this kind of method could be used for tracing in vivo stem cells in the future.


Endothelial progenitor cellsIronMRIStem cell visualization


In 1997, Asahara et al. [1] indentified the putative precursor of endothelial cells (ECs)—endothelial progenitor cells (EPCs). This special cell population can also participate in blood vessel formation, a procedure also known as vasculogenesis which used to be thought as only taking place in the development of the embryo. This kind of vessel formation is basically different from classic angiogenesis, which denotes the formation of new blood vessels from pre-existing ones by in situ division and proliferation of mature ECs. Although still controversial, accumulating evidence have demonstrated that EPCs in the bone marrow (BM) could be correspondingly released, mobilized, and proliferated towards mature ECs by which, i.e., vasculogenesis, EPCs perform vitally in newly formed blood vessels, revascularization, and vascular homeostasis [2]. Since blood supply is certainly involved in several serious pathophysiological situations including tumor and ischemia wherein the blood vessel are activated or interrupted undesirably, EPCs and EPCs-mediated vasculogenesis attract more and more interests recently. Yet the in-depth investigation is limited especially in in vivo dynamic observation on EPCs owing to imperfectness of technique in cell labeling and tracking. In the present study, we effectively developed a method by which EPCs are directly multiplied in serum-free culture and then labeled with iron substance which can be visualized by non-invasive MRI image.

Materials and Methods

Immunoseparation of CD133+ EPCs

CD133+ cells were separated using AC133 cell isolation kit (Miltenyi Biotec Inc., Germany) and a magnetic LS+ separation column according to the manufacturer’s instructions. Mononuclear cells (MNCs) were isolated from either human buffy coat or umbilical cord blood by density gradient centrifugation with Ficoll-Paque (GE healthcare, UK). The MNCs were resuspended in 300 ml MACS Buffer (PBS containing BSA, EDTA, pH 7.2) at 108 total cells. One hundred microliters of CD133 blocking reagent and an equal volume of CD133 immunomagnetic microbeads were added per 108 MNCs and incubated for 30 min at 4°C. After incubation, the cells were washed twice in MACS buffer. The pellet was resuspended in 500 ml MACS buffer and applied to the separation column. The column was washed four times with 3 ml MACS buffer in the magnetic field. The column was removed from the magnet and the cells eluted with 5 ml of MACS buffer. The mothers and the volunteers were given informed consent based on guidance from the regional ethics committee.

Multiplication of CD133+ EPCs

At the density of 5 × 105 cells/ml, separated CD133+ cells were cultured in Stemline® II Hematopoietic Stem Cell Expansion medium (Sigma Aldrich, USA). Cells were counted everyday and new medium was applied every 3 or 4 days. Stemline® II Hematopoietic Stem Cell Expansion medium contains an appropriate amount of Stemline™II Cell Expansion basic liquid medium (Sigma Aldrich, USA), l-glutamine (Invitrogen, USA) 2 mM/L, Recombinant Human Stem Cell Factor (R&D, Europe) 40 ng/ml, FLT3 (Sigma Aldrich, USA) 40 ng/ml, TPO (thrombopoietin) (Sigma Aldrich, USA) 10 ng/ml, penicillin/streptomycin 10 μl/ml and no serum ingredient inside.

Labeled by Iron Complex

Resovist (Schering, German) and protamine sulfate (Sigma Aldrich, USA) complexes were prepared freshly at 50 μg/ml and 5 μg/ml final concentrations, respectively. At the 7th day, cell production was collected and co-cultured with complexes for 24 h. After fixing with 1% formaldehyde, CD133+ cells were resuspended evenly in agarose gel and readied for MRI scanning. In order to assess the influence on labeled cells viability, cell production was continuously co-cultured with iron complex at various concentrations (Resovist: 10, 20, 50, 100, 200 μg/ml and protamine sulfate: 1, 2, 5, 10, 20 μg/ml, respectively) as well as cells without iron for 6 days. Cell numbers were counted every day and trypan blue exclusion test was used to determine the number of viable cells present in a cell suspension. One way ANOVA statistics was performed to compare each concentration with control.

Prussian Blue Staining

Using a 200-μl volume of aspirated suspension from cell culture, cell slices were performed by Cytospin cell centrifugation (Thermo Scientific, USA) according to the manufacturer’s instructions. Prussian blue stain was performed according to the protocol offered by manufacturer. Briefly, place cell slices in Working Iron Staining Solution for 10 min and rinse in deionized water. Counterstain 5 min in Working Pararosaniline Solution. Rinse in deionized water and air dry (Accustain Iron Stain, Sigma Aldrich, USA). Observe under microscope after staining.

MRI Scanning and Data Analysis

The expanded cells were collected and resuspended in agarose gel at 0.15% concentration in NUNC tubes at two different concentrations of 3,500 or 10,000 cells per 2 ml. Cell suspensions in agarose gel should be preserved below room temperature before MRI was performed to get transversal sections of each tube. Seven-tesla high power MRI (Oxford Instruments, Oxford, UK) with horizontal magnet is equipped with 12.5 Gauss/cm gradients system (Tesla Engineering Limited, West Sussex, UK) and Gradient echo. Parameters setting are as follows: TR = 500 ms, TE =4.3 ms, NEX = 8, Matrix = 256 × 256, FOV = 60 mm × 60 mm, slice = 10, gap = 0.1 mm, thickness = 0.8 mm, acquisition time = 17 min 4 s. The agarose gel solution without cell inside was also scanned as control. The change of MRI signal intensity and areas of region of interest (ROI) in percentage were analyzed. For each sample, three sections were scanned and percentages of ROI in each section were analyzed and calculated by Photoshop. The data were averaged and paired t-test was performed to compare the statistical difference between two groups with different cell numbers.


Cell Multiplication

CD133+ cells grew slightly larger after they were cultured for 24 h in Stemline™II serum-free medium. Cell counting showed decreasing amount during the first few days followed by gradually increasing afterwards. After being cultured for 7 days, the total amount of peripheral blood originated cells can reach 10 ± 0.89 × 106 cells per milliliter which has been expanded by fivefold, even eightfold over the starting number. Surprisingly, under the same condition, CD133+ cell population originating from umbilical cord blood expanded much faster than their peripheral blood counterpart. By day 7, the final amount of umbilical blood cells could reach 50 ± 5.43 × 106 cells/ml.
Fig. 1

Immediately after separation by immune-microbeads cell production was identified by flow cytometry which told 89 ± 4.48% CD133+ cells. Seven days later, multiplied cells were collected and assessed again when 85 ± 7.43% of them still maintained CD133 surface makers

Surface expression of CD133 was also assessed by flow cytometry using a FACS Calibur. The purity of cell population separated by CD133 immunomagnetic microbeads can reach percentage of 89 ± 4.48%. Even after 7 days still 85 ± 7.43% of cells maintained CD133 surface “stem” marker (Fig. 1). A minority of cells in culture had changed initial round into spindle shape at day 7, which indicates that cell progenitors proliferated towards downstream mature ECs generation. As expected, spindle-shaped mature ECs occupied gradually increased proportion over time.

Iron Complex Labeling and Assessment of Cell Viability

After co-culture with Resovist and protamine sulfate complex for 24 h, iron-labeled cells were stained by Prussian blue by which cell nucleus appeared red and cytoplasm was pink. Under a light microscope, iron complex was shown being taken up and well distributed inside cytoplasm as blue particles. No blue particles were seen in unlabeled cells. Comparison within groups with various concentrations showed that iron complex had no influences on CD133+ cell survival. Statistically, there are no differences between control group and other concentration groups. The dividing cell can even be seen on Cytospin slice with Prussian blue stain which is also supportive to the non-cytotoxicity of iron complex on cell viability (Fig. 2).
Fig. 2

CD133+ cells are in round shape and unattached in culture (upper left picture). When co-cultured with iron complex with various concentrations the cell growth is not influenced as showed in upper right graph (from Con.1 to Con.5 the concentration of each group is: Resovist: 10, 20, 50, 100, 200 μg/ml and protamine sulfate: 1, 2, 5, 10, 20 μg/ml, respectively. The columns in each concentration group represent from day 1 to day 6, respectively). The lower pictures show that iron-labeled cells were stained with Prussian blue (left) and dividing cell can be even seen (right, black arrow). Con. concentration

MRI Visualization

Cells with iron particle inside can be effectively visualized with T2*-weighted MRI imaging. Labeled cells induced the decrease in signal intensity which were shown as scattered low-signal spots on T2*weighting images. Delta signal intensity (ΔSI) of 3,500 labeled EPCs is −38.03 ± 4.75% and that of 10,000 labeled EPCs is −51.37 ± 9.32%, respectively, of the background signal intensity. The region of interest (ROI) is also analyzed by calculating the proportional area in percentage (percentage of area of ROI in proportion with that of transversal section of agarose tube). Higher cell numbers (10,000 cells) stimulated larger area (37.00 ± 4.63%) of signal intensity change than that (27.58 ± 4.34%) of lower numbers (3,500 cells; Fig. 3).
Fig. 3

Labeled cells, 3,500 and 10,000, were respectively suspended in agarose gel in tube which were then scanned by 7-T MRI to get T2*-weighting images. Two methods were used to analyze the influence of iron substance inside EPCs on MRI signal. The significant difference for either delta SI or area of ROI demonstrated that higher labeled cell number can induce greater change of MRI signal. ROI region of interest, SI signal intensity


BM-derived EPCs have been proved that they can critically participate in post-natal vasculogenesis which is related to tumor, stroke even diabetes [36]. For instance, Id−/− mutant mice cannot be supportive to tumor growth and metastasis due to defective vascularization induced by inhibited mobilization of EPCs from BM, which can be restored completely by co-inoculation of wild-type BM cells [7]. Shaked Y et al. [8] found that EPCs in peripheral blood were unregulated remarkably 4 h after being treated with an anti-vascular drug and contributed to neo-angiogenesis; whereas, normally, EPCs is a relatively rare population existing in blood circulation. The study of Cho H. J. [9] found that introduction of exogenous EPCs could promote revascularization in ischemic heart. Accumulating evidence have indicated that EPCs play a very important part in the post-natal new formation of blood vessels. One can hopefully foresee that involvements of EPCs population in vasculogenesis could hold a very good promise for future clinical application by, most likely, promoting formation of blood vessel in stroke or synergistically enhancing effect of tumor vascular disrupting treatment. The versatility of EPCs could allow us to treat diseases that are closely correlated with blood supply in a new field. But the investigation on mechanism of EPCs participating in vasculogenesis remains largely unclear—including chemotactic movement, localization and environment of cell proliferation and maturation, relationship between angiogenesis and vasculogenesis as well as effect of anti-vascular drugs against EPCs-mediated tumor vascularization. For this consideration, we established an effective method in the present study to visualize EPCs by using two FDA-approved agents—Resovist and protamine sulfate.

Resovist is a kind of super-paramagnetic contrast media, the main ingredient of which is Fe3O4 with unique crystal structure. It has been proved that Resovist can be excreted naturally by liver and spleen without body hazard. Resovist was originally developed as an organ-specific MRI contrast agent, used for the detection and characterization of especially small lesions. It consists of iron nanoparticles coated with carboxydextran which can be accumulated by phagocytosis inside cells and shorten T2 relaxation time. With this special property, Resovist now has clinically served as a detector for liver diseases and tumors [10, 11]. Protamine sulfate is usually administered to reverse the large dose of heparin administered during certain surgeries and also used in gene transfer. In the present study protamine sulfate acted as a transfection agent to facilitate the endosomal incorporation of iron particles into human EPCs. In previous studies, there are several transfection agents were used to complex to dextran-coated iron to label cells such as Superfect, Lipofectamine Plus and poly-l-lysine (PLL) [12, 13]. Compared to these transfection agents, protamine sulfate has been approved for clinical use by FDA as well. Apparently, the method of cell labeling by use of two FDA-approved agents would be more desirable and acceptable in translating ex vivo labeling method to clinical research or application related to EPCs investigation. Theoretically, this kind of approach which was developed in the present study can be used to trace other kinds of stem cell population bedside EPCs.

Labeled CD133+ cells can change the magnetic field and then lower MRI signal which make cells visualized in an MRI image. A co-incubation study demonstrated that the existence of iron oxide had no influence on cell viability when combined with protamine sulfate at different concentrations. Furthermore, since MRI technique is non-invasive and repeatable, this method could be applied for trace of in vivo pre-labeled EPCs. Through labeling cells with magnetic stuff before injection, it will be possible to detect what time and how many cells have mobilized to lesions according to the change of MRI signal density. Expectedly, the change of MRI signal intensity is associated with labeled cells numbers at the same incubation time and iron concentration in which case, a higher cell number leads to greater signal intensity change. Considering the magnetic area stimulated by several, even a single cell is spatially larger than the cell volume, this established method could be used to visualize an even single labeled stem cell. Considering basic principle of contrast-sensitive MRI image, the region of declined magnetic signal was supposed to be larger than the exact size of the iron particle itself and even than the whole cell spatial volume. So the scattered spots with low signal intensity on MRI image could not be regarded as a real cell or a real cell cluster. They are supposed to alter magnetic fields around MR contrast substance contained in a single or several cells.

Since EPCs will lose their “stem” property over time, as proven by flow cytometry in the present study, we incubated CD133+ cells with iron and protamine sulfate for 24 h. Otherwise, if longer incubation time would be applied, EPCs will proliferate towards mature ECs and not be “stem cell” anymore even if they can take up additional iron substance and are more detectable on MRI. This issue should be considered seriously when labeling stem cell population. But there is another aspect to this approach which deserves serious consideration and future investigation: since EPCs are labeled in a one-off way by co-culturing with contrast, the iron oxide content inside cytoplasm will be divided along with division of mother cell into two offspring cells and therefore diluted somehow, but not exactly half dilution, which lessen the change of MRI signal intensity even down below detection threshold after several generations. Actually, we observed the gradually decreased iron content after few generations over time which caused gradually decreased MRI signal and it needs to be further investigated to catch living and continuously dividing cell by MRI visualization before possible application to assess cell growth and division. Taken together, we think effective stem cell labeling can be provided by incubating EPCs with Resovist and protamine sulfate complex for 24 h.

We used Stemline™II liquid medium without serum in case complicated ingredients in serum could promote stem cell to differentiate and mature. Compared to EGM-2-containing serum and other components, we found almost 90% of cells in culture still maintain a round shape and possess CD133 surface marker after 7 days even they started to grow bigger; whereas, most cells in EGM-2 medium have proliferated into spindle-shaped mature ECs and lose their stem cell markers by same period (data not shown). Besides the promotion of complicated protein components in serum on stem cell proliferation, the commonly used vascular endothelial growth factor (VEGF) in EGM-2 medium, which has been well known as a pro-angiogenesis factor, might also stimulate the accelerated lost of stem cell marker. We also found that the reproductive speed of umbilical blood cells was much faster than their counterpart originated from peripheral blood though they were isolated and treated by exactly the same procedure and condition. It is still an unexplained but interesting issue why umbilical CD133+ cells have hyper-reproductivity over peripheral blood. But at least we can say umbilical blood is the better reservoir with more robust stem cells inside.

Currently, the pre-clinical trial has proven that intravenous introduction of EPCs can partly recover the myocardium function of ischemic heart [14]. Though there is no research yet involved in the possibility of EPCs acting as a potential approach to treat tumor, we can still foresee that further investigation of reciprocal relationship of this kind of cell population and tumor vasculogenesis could open up new vistas for vascular disrupting tumor treatment. In that case, the method we developed in the present study will provide an effective way that might be useful for in-depth mechanism of EPCs participating in blood vessel formation and monitoring EPCs-based treatment of a variety of vasculogenesis-related diseases.

Copyright information

© Humana Press Inc. 2008