Human embryonic stem cells (hESCs) are pluripotent, holding the potential to differentiate into mature adult cells (1). Several animal studies have reported that the hESCs significantly restore damaged organs (2–4). However, cellular engraftment is complicated by teratoma formation. Therefore, an in vivo imaging system to monitor stem cell survival while also detecting and treating teratoma without molecular manipulation of the stem cells has been developed in this study.
Manganese-enhanced MRI (MEMRI) uses unique property of manganese (Mn2+), which enters only the metabolically active cells through voltage-gated calcium channels. Consequently, intracellular Mn2+ can induce T1-shortening effect generating positive contrast in viable cells (5–7). Although Mn2+ has been recognized as the first MRI contrast agent, known toxicity has hindered widespread use. Recently, sensitive in vivo detection of various cancers using MEMRI has been reported (8, 9). This study represents the first feasibility study to assess the theranostic potential of MEMRI by in vivo longitudinal evaluation and therapy of hESC-derived teratoma cells.
MATERIALS AND METHODS
Optimization of Therapeutic Dosage of MnCl2
All the following animal protocols were approved by the administrative panel on laboratory animal care at Stanford University. The therapeutic dosage of MnCl2 was optimized by administering tolerable maximal dosage of MnCl2 in severe combined immunodeficient (SCID) mice.
Culture of Undifferentiated hESC
The H9-line of hESCs (Wicell, Madison, WI) was trandscuded with firefly luciferase (hESC-Lucs) and maintained in an undifferentiated and pluripotent state by culturing on irradiated mEF feeder layer using the hESC culture medium (10).
hESC-Derived Teratoma Formation and MnCl2 Chemotherapy
Under general anesthesia with 1% isoflurane at 1 L/min oxygen, 1 × 106 of hESC-Luc were directly injected into SCID mouse hindlimbs by an insulin syringe (n = 12). These mice were divided into two groups: chemotherapy and control groups. The chemotherapy group (n = 6) was injected with 500 μL of 5 mM MnCl2 intraperitoneally (IP) 3 days a week. The control group (n = 6) was given 250 μL of 5 mM MnCl2 IP only prior to image acquisition.
Serial In Vivo MEMRI and Bioluminescence Imaging
Longitudinal in vivo evaluation by MEMRI and bioluminescence imaging (BLI) was performed on post-transplant weeks 2, 4, 6, and 8. For MEMRI, all the mice were injected with 250 μL of 5 mM MnCl2 IP 15 min prior to imaging (7). Under general anesthesia by 1% isoflurane at 1 L/min oxygen, each mouse was placed on top of a customized small animal surface coil in a prone position. Both hindlimbs were scanned using Signa 3.0 T Excite HD scanner (GE Healthcare system, Milwaukee, WI) with spin echo inversion recovery sequence: TR 600 ms, TE 5 ms, TI 120 ms, NEX 1, FOV 3, slice thickness 0.8 mm, space 0 mm, and matrix 128 × 128. MEMRI data were analyzed for contrast-to-noise-ratio (CNR) and 3D volume using image J (NIH, Bethesda, MD). CNR was calculated as follows: CNR = (signal intensity (SI)teratoma − SImuscle)/standard deviation of the image noise. Under general anesthesia, concurrent BLI was performed using the charged-coupled device camera (IVIS spectrum, Caliper, Mt. view, CA) following IP injection of D-luciferin at 375 mg/kg body weight. Bioluminescence images were acquired for 30 min at a 3 min interval and quantified in units of photons/(s·cm2·steridian) using Living image 2.5 (Caliper, Mt. view, CA; Ref.11).
At post-transplantation weeks 6 and 8, three mice in each group were sacrificed for histology. All the histological data were assessed by a veterinary pathologist, who was blinded to the study. Teratoma was processed for a paraffin block followed by H&E staining of 5 μm slice. For caspase 3 detection, immunohistochemistry procedure was performed according to the manufacturer's manual (Vector Laboratories, Burlingame, CA).
Descriptive statistics include mean and standard deviation. Comparison between two groups was performed using student's t-test. Comparison among multiple groups was conducted using analysis of variance with post hoc test. All data analyses were done by SPSS 16.0 (SPSS, Chicago, IL). Significance was assumed when P < 0.05.
Optimization of MnCl2 Administration
The chemotherapy protocol was optimized at 500 μL of 5 mM IP three times a week. All SCID mice tolerated the chemotherapy protocol for 8 weeks. But all the mice in the chemotherapy group developed feet swelling.
In Vivo MEMRI and BLI of Cell Survival and Teratoma Formation
CNR of the hESCs was compared with concurrent luciferase activity (Fig. 1a). With the hESC proliferation and subsequent teratoma formation, CNR and luciferase activity increased accordingly at each time point (Fig. 1b). These mice did not receive multiple doses of Mn2+ (chemotherapy group) but only prior to undergoing MEMRI (control group). Significant correlation was revealed between CNR and luciferase activity by Pearson correlation (P < 0.01). Significant increase in CNR was noted at 6 weeks post-transplantation, corresponding with teratoma formation on the H&E staining (Fig. 1c).
In Vivo MEMRI and BLI of Teratoma Reduction
The chemotherapy group showed significant reduction in the volume of teratoma compared with the control group (9.4 ± 4.2 vs. 16.8 ± 1 at week 6 and 64 ± 49 vs. 123 ± 70 at week 8, P < 0.05, Fig. 2a). CNR was comparable between the chemotherapy and control groups throughout the 8 week period. At weeks 4 and 6, a trend toward higher CNR was demonstrated in the chemotherapy group. However, at week 8, CNR demonstrated a higher trend in the control group (P > 0.05, Fig. 2b). BLI demonstrated a significantly decreased luciferase activity in the chemotherapy group compared with the control group at weeks 6 and 8 (P < 0.05, Fig. 2c).
H&E staining of teratoma in both groups revealed hESCs-derived mature teratoma at post-transplantation period, week 6. Development of three germ layers, consistent with teratoma formation, was similar between the two groups, suggesting Mn2+ does not affect the pluripotent property of hESCs (Fig. 3a). The chemotherapy group showed increased proportion of dead cells within the teratoma compared with the control group (Fig. 3b). Finally, immunohistochemistry demonstrated positivity for caspase 3 only at week 8 in the chemotherapy group, indicating apoptotic cell death while the control group did not show any caspase 3 signal (Fig. 3c).
Although several animal studies have demonstrated the enormous potential of hESCs to restore damaged organs, the formation of teratoma has been a major obstacle to clinical implementation of hESCs (12). To address this critical issue, we previously demonstrated sensitive and reliable in vivo detection of hESC-induced teratoma by MEMRI (7).
Interestingly, Mn2+ was reported as the first MRI contrast agent, which increased R1 in the liver, kidney, and heart (13). As Mn2+ is a competitive inhibitor of calcium channel, it blocks calcium influx through the voltage-gated calcium channel in the heart and the brain, possibly compromising cardiac and neurologic function (14). Although little is known about Mn2+ interaction within the hESCs, it is bound to intracellular or nuclear calcium binding proteins and mostly taken up by mitochondria (15, 16). The intracellular accumulation of Mn2+ by viable hESCs induces T1 shortening effect, generating positive MRI signal (6). Thus, MEMRI localizes viable hESCs, allowing accurate in vivo evaluation of the anatomical and biological properties following cell transplantation. However, cellular toxicity of Mn2+ has been known to cause cardiac toxicity by acute exposure and Parkinson-like symptoms following chronic exposure. Our data also showed higher concentration of Mn2+ caused extensive mineralization of myocardium, resultant fibrosis, and death in SCID mice. The increased Mn2+ uptake by the hESC-derived teratoma cells may be due to their enhanced metabolism and immature voltage-gated calcium channels, resulting in excessive intracellular accumulation, and subsequent, cell death (17–20). Future studies are planned to test this hypothesis by comparing the differential uptake of Mn2+ and the toxicity between the undifferentiated embryonic stem cells and embryonic stem cell derived cardiac cells.
Longitudinal assessment of the biokinetics of hESC-derived teratoma has demonstrated that Mn2+ could eliminate a substantial portion of the teratoma burden. Significant reduction in the 3D volume and luciferase activity of the teratoma is generated at weeks 6 and 8 following serial systemic Mn2+ administration. This finding suggests that the cytotoxic effects could be reached by week 6. This finding is also corroborated with positive caspase 3 expression in the chemotherapy group only at week 8 and not week 6 in either group. Our laboratory has reported significant annexin expression by hESCs when cells are labeled with higher dose of manganese (7). In addition, several studies have shown manganese chloride induced apoptosis and cell death in various cell types (20–23). Thus, this study demonstrates that the increased intracellular accumulation of Mn2+ within the teratoma cells induces apoptosis, and subsequent, cell death.
The limitations of our study include the nonselective manganese uptake. Although the increased cellular metabolism and the immature calcium channels of the differentiating hESCs may underlie the mechanism for the MEMRI-mediated theranostic approach, a targeted approach will be desirable to avoid systemic toxicity (24). A novel method using antibodies specific to the teratoma cells to deliver the Mn2+ nanoparticles is currently underway (25). Furthermore, complete eradication of the teratoma is not achieved. Differential uptake of manganese depending on the specific cell types within the teratoma may account for this finding (19, 20). In addition, we do not compare the therapeutic efficacy of MEMRI with cisplatin-based regimen. However, to the best of our knowledge, no report is available on the chemotherapeutic option against hESC-derived teratoma. Finally, we do not establish a correlation between the MEMRI signal and the histology of the early, proliferative stage of hESC growth. Comparison between MEMRI and traditional MRI methods at the early proliferative stage was also not performed.
In conclusion, this study demonstrates the feasibility of the theranostic capability of MEMRI to track the viable transplanted teratoma cells and attenuate the growth of the hypermetabolic teratoma cells.