Melanin as an Endogenous MRI Contrast Agent for Stem Cell Delivery in Stroke

In vivo measures of survival, growth, and migration of stem cell transplants for stroke therapy remain inadequate. Since such biomarkers in the central nervous system do not exist, we sought to investigate the potential of melanin as an endogenous MR contrast medium for monitoring cell lineage transplants. Expression of a human gene, tyrosinase, induces melanogenesis, but unlike particle-based tracking agents, tyrosinase incorporated into the host genome will be replicated during mitosis. Tyrosinase and the gene for a co-enzyme, tyrosinase-related protein 1, were placed under the CMV promoter. The construct was inserted into 293 HEK (human embryonal kidney) and iPS NPC (induced pluripotent neural progenitor) cells via plasmid transfection and viral transduction, respectively. Stroked mice were injected with melanin-producing cells. Control mice were injected with native cell lines. In vitro expression was measured by uorescent microscopy, immunocytochemistry, spectroscopy, PCR, and MRI. Robust in vitro melanin production was achieved in both cell lineages demonstrated by signicant T1 shortening on in vivo MRI. Pathologic correlation demonstrated colocalization of pigmented regions in the injection sites with human antinuclear antibody staining. Through induction of melanogenesis that is reproducible across multiple cell divisions, MR-based imaging of clinically relevant cell lineage transplants is possible.


Introduction
Stem cell delivery is a promising therapy in stroke as well as in neurodegenerative disorders [1][2][3]. However, in vivo measures of survival, growth, and migration of cell transplants remain inadequate.
Unlike a stem cell therapy in other organ systems, there are substantial limitations to tracking stem cell treatments in the brain and spinal cord after transplant-prohibiting an in vivo biomarker of these therapies [4]. Image-based tracking approaches have focused on MRI given its superb soft tissue contrast and commonplace clinical use. Ferrous intracellular compounds such as iron nanoparticles, among the rst such agents, produce negative contrast that may be confused with susceptibility artifact from hemorrhage or surgery [5,6]. These techniques quantify loss of signal in the T2 sequence [7]. Positive contrast agents like gadolinium nanoparticles obviate these particular shortcomings, but as with any particles, suffer from signal dilution with every mitosis [8]. We therefore sought to develop a more durable positive contrast agent whose signal intensity correlates with cell number [9].
Melanin is an endogenous pigment synthesized in neurons throughout the brain [10] that scavenges heavy metals and free radicals. Its accumulation within the normal human substantia nigra is seen on 3.0 Tesla MRI as foci of T1 shortening [11][12][13] due to paramagnetic effects imparted by bound iron and other metallic cations. Visualization of melanin in physiologic concentrations suggests it may serve as a low toxicity contrast agent. Insertion of a single gene, tyrosinase (TYR), induces melanogenesis in several human cell lines [14][15][16][17]. Unlike particle-based tracking agents, TYR incorporated into the host genome will be replicated during each cell division.
We demonstrate the feasibility of expressing tyrosinase in induced pluripotent stem neural precursor (iPS-NPC) cells as MRI tracking agents in stroked mouse brain. These ndings may support further research into cell tracking for human clinical trials.

Methods
This pilot study aimed to demonstrate a quanti able MRI signal from melanin production, through gene delivery in a cell lineage of potential use in stroke therapy. Therefore, the gene construct was designed with a strong, constitutive promotor to drive expression over the course of the experiment. Initial studies were performed in 293 HEK cells given their accessibility and ease of use. Then iPS NPCs, a lineage with potential clinical relevance, were assessed through similar experiments. HEK cells were sourced from Carmichael Lab at University of California, Los Angeles, and iPS NPC cells were sourced from Lowry Lab at University of California, Los Angeles.

Statistical Analysis
Descriptive measures and summary statistics were performed using SPSS software version 20.0 [21]. In addition, the Mann-Whitney test and Students' t-test were performed and signi cance was considered as two-sided p-value < = 0.05. Immunohistochemistry Specimens through photothrombotic stroke demonstrated a sharply demarcated border between the infarct and astrocyte and neuronal cells. Slices through transfected 293-HEK and transduced iPS-NPC IC injections collected 1 day after initial MRI were HuNu positive ( Figure VI). These regions co-localize with dark melanic pigment under light microscopy.

Discussion
All newly developed human therapies need some element of post-delivery monitoring to assess appropriate bioavailability and establish causal dose-response relationship of treatment effects. For pharmacological agents, this is relatively straightforward in the monitoring of drug levels. However, once cell lines or products are delivered into the human brain, their migration and survival are unknown.
Without adequate tracking methods, the fate of transplanted cells, the degree of migration and the relationship of dose to survival and recovery cannot be determined. As these are core issues of any developing human therapy [22], the inability to track transplanted cells will impact the effectiveness of human trials.
The ideal tracking agent for monitoring cell therapy is nontoxic, simple to measure, and highly correlated with transplant population size, location, and health [23]. Here, we show preliminary evidence that transgenic expression of TYR and TYRP1 could satisfy many of these conditions. Neuromelanin is a scavenger produced endogenously in neurons throughout the human brain, where it is stored in autophagic vacuoles [10]. Iron bound melanin is paramagnetic, creating a strong T1 shortening effect that lends itself to visualization by MRI at physiologic concentrations [11][12][13].
TYR and TYRP1 transgene delivery represents the simplest method of inducible melanogenesis and allows for regulated expression at the promotor level to reduce toxicity in vivo [15][16][17]. We have demonstrated the feasibility of labeling human cells, including biologically relevant iPS-NPC's [28,29].
Moreover, these cells were transplanted in a similar population size and peri-infarct location as employed in recent human stroke trials [30]. The TYR-TYRP1 transgene achieves adequate levels of expression and melanin synthesis to produce readily identi able signal by in vivo MRI of stroked mice.
Nevertheless, this study has several shortcomings. The gene construct employed in this pilot study induced high, constitutive expression and was not designed for prolonged in vivo analysis. Previous studies have demonstrated that constitutive, high TYR-TRP expression reduces cell survival [17]. Further investigation is necessary to determine the appropriate level of transgene expression that produces measurable T1 shortening without altering transplant cell physiology. This may be accomplished by incorporating additional regulatory elements into the transgene construct, such as a Tet-On inducible promotor [31] or constitutively active promoters with lower protein expression such as ubiquitin and PGP [32]. Additionally, the results of a small sample size of ve iPS NPC transplanted mice must be demonstrated in a larger cohort to show reproducibility.
Despite limitations, this study shows the feasibility of imaging clinically relevant stem cells in the brain utilizing an endogenous, positive contrast agent that is not prone to signal dilution with cell division.
Additional research is warranted to investigate more nuanced approaches such as cell-type speci c promoters for in vivo phenotyping of astrocyte (GFP A and Aldh1l [33,34]) and neuronal (parvalbumin and CCK [35]) lineages. Further, inducible promoters [34,36] would allow control of a cell transplant biomarker at a speci c time point, such as during an early phase of post-transplant stress or later period of functional integration.

Conclusion
MR imaging of clinically relevant cell lineage transplants through transgenic induction of melanogenesis is feasible. Further studies are needed to calibrate melanin synthesis to achieve non-toxic T1 shortening long term. These ndings may prove useful to researchers investigating noninvasive treatment monitoring of cell based human therapies.    Agarose gel of control and transduced iPS NPC DNA following TYR TRP RT-PCR. In the leftmost column, a 600 bp fragment corresponds to the tyrosinase mRNA transcript.