Author + information
- Received July 27, 2012
- Revision received November 1, 2012
- Accepted November 9, 2012
- Published online July 1, 2013.
- Peter J. Psaltis, MBBS, PhD∗,
- Karen M. Peterson, BSc∗,
- Rende Xu, MD∗,
- Federico Franchi, PhD∗,
- Tyra Witt, CVT∗,
- Ian Y. Chen, MD, PhD†,
- Amir Lerman, MD∗,
- Robert D. Simari, MD∗,
- Sanjiv S. Gambhir, MD, PhD‡ and
- Martin Rodriguez-Porcel, MD∗∗ ()
- ∗Division of Cardiovascular Diseases, Department of Internal Medicine, Mayo Clinic, Rochester, Minnesota
- †Division of Cardiovascular Medicine, Department of Medicine, Stanford University School of Medicine, Stanford, California
- ‡Department of Radiology and Molecular Imaging Program at Stanford, Stanford University School of Medicine, Stanford, California
- ↵∗Reprint requests and correspondence:
Dr. Martin Rodriguez-Porcel, Division of Cardiovascular Diseases, Department of Internal Medicine, Mayo Clinic, 200 First Street SW, Rochester, Minnesota 55905.
Objectives The goal of this study was to validate a pathway-specific reporter gene that could be used to noninvasively image the oxidative status of progenitor cells.
Background In cell therapy studies, reporter gene imaging plays a valuable role in the assessment of cell fate in living subjects. After myocardial injury, noxious stimuli in the host tissue confer oxidative stress to transplanted cells that may influence their survival and reparative function.
Methods Rat mesenchymal stromal cells (MSCs) were studied for phenotypic evidence of increased oxidative stress under in vitro stress. On the basis of their up-regulation of the pro-oxidant enzyme p67phox subunit of nicotinamide adenine dinucleotide phosphate (NAD[P]H oxidase p67phox), an oxidative stress sensor was constructed, comprising the firefly luciferase (Fluc) reporter gene driven by the NAD(P)H p67phox promoter. MSCs cotransfected with NAD(P)H p67phox–Fluc and a cell viability reporter gene (cytomegalovirus–Renilla luciferase) were studied under in vitro and in vivo pro-oxidant conditions.
Results After in vitro validation of the sensor during low-serum culture, transfected MSCs were transplanted into a rat model of myocardial ischemia/reperfusion (IR) and monitored by using bioluminescence imaging. Compared with sham controls (no IR), cardiac Fluc intensity was significantly higher in IR rats (3.5-fold at 6 h, 2.6-fold at 24 h, 5.4-fold at 48 h; p < 0.01), indicating increased cellular oxidative stress. This finding was corroborated by ex vivo luminometry after correcting for Renilla luciferase activity as a measure of viable MSC number (Fluc:Renilla luciferase ratio 0.011 ± 0.003 for sham vs. 0.026 ± 0.004 for IR at 48 h; p < 0.05). Furthermore, in IR animals that received MSCs preconditioned with an antioxidant agent (tempol), Fluc signal was strongly attenuated, substantiating the specificity of the oxidative stress sensor.
Conclusions Pathway-specific reporter gene imaging allows assessment of changes in the oxidative status of MSCs after delivery to ischemic myocardium, providing a template to monitor key biological interactions between transplanted cells and their host environment in living subjects.
Progenitor cell therapies are being developed as a therapeutic alternative for myocardial repair in coronary artery disease (1). Although studies have shown an overall benefit with improvements in left ventricular (LV) cardiac function after myocardial injury, the beneficial effect of cell-based intervention after myocardial injury has been modest, likely due to poor survival of transplanted cells (2). Until recently, progenitor cell studies were limited in their capacity to assess cell survival, both in cell culture and living subjects. Assessment of the fate and biology of cells after transfer to recipient myocardium relied on traditional ex vivo assays and molecular techniques (e.g., histology, western blotting), which are both invasive and restricted in their capacity to monitor temporal changes in a given subject. However, developments in molecular imaging techniques, such as reporter gene technology, have increasingly enabled the noninvasive surveillance of cell fate after cardiovascular application (3). Bioluminescent imaging (BLI) has provided helpful information in small animal models regarding the kinetics of cell viability by detecting transgene expression under constitutional promoters in living cells (4). Recent studies have demonstrated that reporter gene imaging can also be used to monitor biological processes, in addition to cellular viability, by using conditional transgenes whose expression is regulated by pathway-specific promoters (5,6).
In the setting of cell delivery after myocardial infarction or ischemia/reperfusion (IR), the recipient microenvironment may confer a heightened state of oxidative stress to transplanted cells. Depending on both cell type and degree of oxidative stress, this may have adaptive or maladaptive effects on cell survival and biological function, ultimately influencing capacity for repair and/or regeneration (7–9). The ability to assess the biological interactions of transplanted cells with their host milieu in vivo, including their pro-oxidant status, could provide invaluable insights to assist with efforts to optimize stem cell functionality and therapeutic efficacy (10). Here we report on the novel use of conditional reporter gene labeling and imaging to monitor oxidative stress in mesenchymal stromal cells (MSCs) after their in vivo delivery to a rodent model of myocardial IR.
Detailed methods are provided in the Online Appendix. Bone marrow MSCs from rats were phenotypically characterized under in vitro conditions of increased oxidative stress. DNA reporter gene plasmids were designed, constructed, and used to transfect MSCs to enable assessment of oxidative stress signal. This involved using firefly luciferase (Fluc) driven by the promoter for the p67phox subunit of nicotinamide adenine dinucleotide phosphate (NAD[P]H p67phox–Fluc). To overcome the low expression levels typically inherent with transcriptional targeting of reporter genes, signal amplification was achieved by a recently validated 2-step transcriptional amplification (TSTA) strategy (11). The oxidant signal was normalized to the number of viable cells, assessed by using Renilla luciferase under the regulation of a constitutive cytomegalovirus promoter (CMV-Rluc). Transfected cells were transplanted to animals after IR (n = 7) and compared with sham controls (n = 8). Furthermore, the oxidative stress sensor was also evaluated in MSCs that had been pretreated with an antioxidant (n = 5).
Statistical comparisons were performed with parametric or nonparametric 2-sample Student t tests or 1-way analysis of variance (with Bonferroni posttest comparison), as appropriate. Results are expressed as mean ± SEM of multiple experiments. In all cases, statistical significance was established at 2-tailed p < 0.05.
Reporter gene detection of cellular oxidative stress in vitro
Initially, MSCs were assessed for their susceptibility to increased oxidative stress under in vitro stressors. After exposure to hypoxia (12) or low-serum culture for 24 h (2% fetal bovine serum [FBS]), MSCs displayed increased oxidative stress conversion of 2′7′-dichlorodihydrofluorescein diacetate (H2DCFDA) to 2′,7′-dichlorofluorescein, compared with control conditions (10% FBS) (2.9-fold difference; p < 0.05) (Fig. 1A). Both conditions stimulated differences in MSC expression of various proteins involved in producing or scavenging reactive oxygen species (ROS), including up-regulation of the p67phox regulatory subunit of NAD(P)H oxidase (Fig. 1B) (12). On this basis, a transfection plasmid was designed for an oxidative stress sensor that comprised the NAD(P)H p67phox promoter (353bp, PubMed nucleotide DQ662934) to regulate expression of the Fluc reporter gene (Fig. 1C). Cells were cotransfected with the CMV-Rluc and NAD(P)H p67phox–Fluc plasmids. After 24 hours of culture in low-serum media, MSC Fluc signal, corrected for cell number and transfection efficiency, was significantly increased compared with control conditions, as shown independently by BLI using a cooled charge-coupled device camera (1.6-fold difference; p < 0.05) (Fig. 1D) and luminometry (1.8-fold difference; p < 0.05) (Fig. 1E). In further experiments, MSC Fluc activity correlated strongly with oxidative stress readout in response to varying concentrations of FBS (R2 = 0.95, p < 0.01) (Fig. 1F). Importantly, MSCs pretreated with tempol, a superoxide dismutase mimetic (4-hydroxy-2,2,6,6-tetramethyl piperidinoxyl; 5 mmol/l) displayed a markedly lower Fluc:Rluc signal ratio after low-serum culture, compared with untreated cells (9.5 ± 2.1 vs. 49.6 ± 6.4, respectively; p < 0.001) (Fig. 1E). These in vitro results demonstrated that reporter genes, under the regulation of the NAD(P)H p67phox promoter, can be used to assess the oxidant status of MSCs across a range of oxidative stress levels.
In vivo monitoring of oxidative stress in MSCs
We then tested the oxidative stress sensor in a rat model of anterior wall IR (30-min ischemia). Over the first week after IR surgery, ischemic myocardium displayed histological changes of progressive inflammatory cell infiltration and collagen deposition (Figs. 2A and 2B), with incremental expression of 8-hydroxy-2’-deoxyguanosine, a marker of oxidative stress, that was prominent by 48 h (p < 0.05 vs. sham) (Figs. 2C and 2D). Immunoblotting of myocardial lysates revealed up-regulation of markers/mediators of inflammatory (nuclear factor kappa-light-chain-enhancer of activated B cells), apoptotic (bcl-2 associated X-protein), and oxidative stress pathways (NAD[P]H p67phox, xanthine dehydrogenase, heme oxygenase) in the ischemic and peri-ischemic tissue compared with remote and sham myocardium, especially from day 2 after surgery (Fig. 2E, Online Fig. 1). At this time point, there was also echocardiographic evidence of reduced systolic function (ejection fraction 49.3 ± 4.9% for IR vs. 75.4 ± 1.1% for sham; p < 0.05) (Fig. 2F). Therefore, this model of IR produced early injurious changes that would be expected to create a stressful milieu for cells transplanted into the anterior LV myocardium.
Next, allogeneic MSCs (7.5 × 105) cotransfected with CMV-Rluc and NAD(P)H p67phox–Fluc were administered transepicardially into the ischemic territory of rats, 10 min after reperfusion, or sham surgery. Bioluminescence imaging was performed at 6, 24, and 48 h to detect Fluc activity after administration of its substrate, D-luciferin. Fluc signal was present at low levels over the heart and lungs in sham controls at 6 h, indicating low-grade stress experienced by MSCs early after injection into nonischemic myocardium and prompt redistribution of cells to the lungs (Online Fig. 2A). By comparison, an intense Fluc signal was detected above background noise in the region of the heart, but not in other organs, throughout the first 48 h in ischemic animals and was significantly higher than for sham rats at all time-points (p < 0.01) (Online Fig. 2B). Although absolute values for total Fluc radiance diminished over time (IR group: 71.1 ± 29.0 × 104 photons/s at 6 h, 19.3 ± 6.0 × 104 photons/s at 24 h, 11.9 ± 2.7 × 104 photons/s at 48 h; p < 0.001), the relative difference in cardiac Fluc signal was highest between the ischemic and sham groups at 48 h (11.9 ± 2.7 × 104 photons/s vs. 2.2 ± 0.5 × 104 photons/s; p < 0.01) (Figs. 3A and 3B, Online Fig. 2B). This increase in Fluc in the IR setting was confirmed after the signal was corrected for number of cells with Rluc, which was measured by using ex vivo luminometry of LV homogenates, because an in vivo signal could not be reliably imaged due to tissue attenuation (Fluc:Rluc ratio 0.026 ± 0.004 for IR vs. 0.011 ± 0.003 for sham at 48 h; p < 0.05) (Fig. 3C, Online Fig. 2B). Tissue immunofluorescent staining was also used to verify that transplanted MSCs (transfected to constitutively express green fluorescent protein) did in fact express NAD(P)H p67phox 48 h after administration to ischemic (but not sham) myocardium (Fig. 3D).
MSCs transfected with a Null-Fluc vector displayed very low levels of Fluc activity (1.5 ± 0.3 × 104 photons/s by BLI at 48 h) (Figs. 3A to 3C), underscoring the specificity of the detected oxidative stress signal. Furthermore, treatment of MSCs with 5 mmol/l of tempol for 24 h before delivery to IR animals markedly attenuated the Fluc signal to levels similar to those in shams (3.7 ± 0.7 × 104 photons/s at 48 h). This finding corroborated the in vitro observation that expression of Fluc could be used not only to detect different levels of oxidative stress but also to monitor changes in the biology of transplanted cells (e.g., after an antioxidant intervention).
In this proof-of-principle study, we validated a pathway-specific reporter gene approach for detecting oxidative stress in cells after their transfer to ischemic myocardium. Using the NAD(P)H p67phox promoter expressing Fluc, we were able to demonstrate Fluc signal in MSCs transplanted to ischemic myocardium. Moreover, this study provides preliminary evidence that this novel oxidative stress sensor may be used to monitor the effect of adjuvant interventions targeted to improve the biology of transplanted progenitor cells.
A recurring obstacle for cell therapy studies continues to be the significant loss of viable cells from target myocardium in the early aftermath of transplantation, ultimately limiting the scope of therapeutic benefit (13). Although the immediate retention of cells is adversely influenced by mechanical forces during the delivery process, ongoing attrition and failure of engraftment are also contributed to by noxious stimuli in the foreign environment of the recipient heart. This has been shown to apply for cell delivery into healthy, intact myocardium but is further accentuated in the setting of ischemia/infarction, in which a combination of reduced tissue perfusion and reperfusion injury, inflammatory cell infiltration, and pro-oxidant factors incite local release of oxidative free radicals and production of ROS, imparting increased oxidative stress to recently administered cells (9,14). Although the precise effects of increased oxidative stress on cell survival, proliferation, differentiation, maturation, and paracrine function are largely undetermined, they are thought to depend on variables such as cell type, myocardial distribution of cell delivery, and severity of the myocardial pro-oxidant state (15–17). Recently, transcriptomic analysis has revealed that the overwhelming influence of ischemic and inflamed myocardium on bone marrow mononuclear cells early after transplantation is to trigger survival responses (9). In the case of MSCs, exposure to ROS (likely depending on level of ROS) may cause either deleterious or favorable effects on viability, senescence, and biological function (7,8,12,18). Furthermore, the importance of the oxidative status of MSCs has also been highlighted by showing that its manipulation (e.g. by gene transfection, hypoxic preconditioning, pharmacological intervention) may protect against apoptosis and enhance reparative capacity (10,12,19–22).
With our goal of monitoring oxidative stress in MSCs, we elected to use the Fluc reporter gene system, which has been a cornerstone of BLI, to evaluate cell fate in small animal studies (4,15). Previously, the implementation of reporter gene labeling has mainly been in the context of using constitutional promoters, to provide surrogate information about the number of viable cells retained in myocardium over different time intervals. Recent developments in pathway-specific reporter gene imaging have begun to allow surveillance of other biological processes in transplanted cells in vivo, notably their differentiation along cardiomyocyte or endothelial lineages by using conditional promoters, such as cardiac sodium–calcium exchanger-1 (Ncx-1) or Tie2, respectively (5,6).
In selecting an appropriate promoter that would trigger Fluc expression in “stressed” MSCs, our choice of NAD(P)H p67phox was based on its reliable up-regulation when MSCs were exposed to hypoxic (12) or low-serum conditions, accompanied by increased production of ROS. Although the use of NAD(P)H p67phox to regulate reporter gene expression may not necessarily reflect activation of alternative ROS pathways (e.g., xanthine oxidase, mitochondrial P450 cytochromes), its validity was supported here by the ability to detect increased Fluc signal in cells exposed to pro-oxidant conditions both in vitro and in vivo. Compared with IR, Fluc intensity was considerably lower in sham animals, in keeping with the expectation that their myocardium would provide a less hostile milieu for transplanted cells, based on lower tissue expression of pro-oxidant, pro-inflammatory, and apoptotic markers. Importantly, immunostaining revealed that MSCs retained within ischemic myocardium did indeed express NAD(P)H p67phox, confirming the biological changes of MSCs in the in vivo setting of IR. Furthermore, the correction of Fluc signal for viable cell number by using a second reporter gene (CMV-Rluc) clearly demonstrated that the up-regulation of NAD(P)H p67phox–Fluc in IR animals compared with sham animals was not merely due to retention of higher cell numbers in ischemic tissue. It is worth noting that measurement of Rluc did require ex vivo luminometry, as in vivo detection by BLI was complicated by the fact that Rluc emits blue wavelength photons (λpeak = 480 nm) when coelenterazine is used as its substrate, and these can be strongly attenuated by the subject's chest wall. Novel developments of red-shifted Rluc mutants may allow better in vivo signal detection in future studies (23).
One of the main challenges for pathway-specific reporter gene labeling is the relatively weak activity inherent in transcriptional targeting with most conditional promoters. Although the use of an amplification strategy (TSTA) necessitated complex vector design and testing due to the inclusion of multiple elements in addition to the promoter and reporter gene (Fig. 1C), it was a critical component in this study to achieve detectable Fluc signal in rat myocardium. Although previous work has demonstrated that TSTA vectors may achieve profound amplification of weak promoters, including those that are cardiac specific (e.g. troponin T, alpha-myosin heavy chain) (11,24), the current study provides early in vivo evidence of their utility for imaging transfected cells in recipient myocardium.
One of the potentially important applications for a new imaging sensor for cellular oxidative stress will be to enable surveillance of its fluctuation over time. This could contribute insights regarding the optimal therapeutic window for cell delivery after myocardial injury and reveal critical intervals during which adjuvant interventions may help to modify the oxidative status of retained cells. Unfortunately, a major confounder of the ability to interpret temporal tends in absolute biological signal is the rapid and progressive diminution of cell number from recipient myocardium (3,15). We have also observed this to be the case in our model system during separate experiments, in which CMV-Fluc MSCs were used to follow the time course of their retention (data not shown). This finding, along with the transient nature of the plasmid transfection strategy used here, is likely to have contributed to the sharp decline in absolute NAD(P)H p67phox–Fluc signal observed after cell injection in both the IR and sham groups (Online Fig. 2). Although we found that reporter gene expression is maintained in vitro for at least 2 weeks after transfection, peak signal typically occurs within the first 4 days. In future studies, the use of stable genomic integration (e.g., by retroviral vectors, TALE nucleases) of the NAD(P)H p67phox–Fluc sensor in transplanted cells will be warranted.
It should also be recognized that alternative oxidative stress pathways may be more relevant than NAD(P)H oxidase for other progenitor cell types currently under investigation in cardiovascular studies. Further evaluation and refinement of oxidative stress imaging will therefore need to determine optimal stress-related promoters for individual cell types, as well as for different disease contexts, including chronic myocardial ischemia and nonischemic cardiac disease.
Implications and clinical translation
Previous rodent studies have used constitutive reporter gene imaging by bioluminescence to address key questions pertaining to optimal cell type (15), timing interval (17), and site of administration (16). Their findings have provided novel insights to assist in the planning and design of clinical stem cell studies. Similarly, in the case of an oxidative stress sensor, we anticipate that BLI will enable rapid and high throughput interrogation of biological interactions between transplanted cells and their recipient environment to help screen and hone cell and tissue engineering strategies in preclinical models, before guiding their translation to the clinical realm. In the context of MSCs, such optimization strategies may include the use of immunoenriched, genetically engineered, preconditioned, or cardiopoiesis-directed MSCs (10). Furthermore, conditional promoters, such as NAD(P)H p67phox, may even be implemented in a “theranostics approach” (11), whereby imaging and therapeutic applications are combined by using bidirectional systems to regulate the selective expression of linked transgenes (e.g., reporter and therapeutic genes) in transfected cells.
Despite its many advantages for cell tracking in rodents, the lack of spatial resolution and tissue depth penetration of BLI precludes its use in human subjects. Thus, the direct implementation of oxidative stress reporter gene imaging in large animal and clinical studies will necessitate the replacement of Fluc with reporter genes specifically designed for clinical imaging modalities (3). Owing to its superior cell detection sensitivity (femtomolar range) and radiochemical flexibility, it is likely that positron emission tomography (PET) will be better placed for initial clinical translation than either single-photon emission computer tomography or magnetic resonance imaging, although hybrid PET–magnetic resonance imaging or PET–computed tomography will be useful for achieving optimal balance of oxidative stress detection and anatomic resolution. To this end, porcine studies have already demonstrated the feasibility and validity of PET-based (e.g., 18F-FHBG PET) detection of constitutive reporter gene expression (e.g., herpes simplex thymidine kinase) in retrovirally transfected MSCs after myocardial delivery (25–27).
This study demonstrates the feasibility of using pathway-specific reporter gene labeling to monitor the oxidative stress of transplanted progenitor cells, providing a novel platform to help optimize the use of cell therapies.
For the full protocol and supplemental figures, please see the online version of this article.
The luminometer used was obtained through a grant from Turner BioSystems. This work was supported in part by the National Institutes of Health (HL 88048, Dr. Rodriguez-Porcel) and the Mayo Foundation (Dr. Rodriguez-Porcel). Dr. Psaltis has received funding from the National Health and Medical Research Council of Australia. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- bioluminescence imaging
- fetal bovine serum
- firefly luciferase
- left ventricular
- mesenchymal stromal cell
- nicotinamide adenine dinucleotide phosphate
- positron emission tomography
- reactive oxygen species
- 2-step transcriptional amplification
- Received July 27, 2012.
- Revision received November 1, 2012.
- Accepted November 9, 2012.
- American College of Cardiology Foundation
- Terrovitis J.V.,
- Smith R.R.,
- Marban E.
- Wu J.C.,
- Chen I.Y.,
- Sundaresan G.,
- et al.
- Sheikh A.Y.,
- Huber B.C.,
- Narsinh K.H.,
- et al.
- van der Bogt K.E.,
- Sheikh A.Y.,
- Schrepfer S.,
- et al.
- Hung T.C.,
- Suzuki Y.,
- Urashima T.,
- et al.
- Swijnenburg R.J.,
- Govaert J.A.,
- van der Bogt K.E.,
- et al.
- Tsubokawa T.,
- Yagi K.,
- Nakanishi C.,
- et al.
- Iyer M.,
- Wu L.,
- Carey M.,
- Wang Y.,
- Smallwood A.,
- Gambhir S.S.
- Gyongyosi M.,
- Blanco J.,
- Marian T.,
- et al.