Background: The development of robust reporter molecules for imaging gene expression in the context of living, optically opaque animals is critical for facilitating research in the emerging fields of genetic and cell‒based therapeutics. Despite their wide prevalence in cell biology, fluorescent and luminescent reporter genes have limited utility in this respect due to the poor penetration of light into deep tissues. In contrast to optical techniques, magnetic resonance imaging (MRI) offers excellent depth penetration and high spatial and temporal resolution. However, existing reporter genes for MRI are primarily based on metal-binding proteins, which are limited by their reliance on metal ion bioavailability, or derived from acidic and basic proteins with exchangeable protons, which tend to have relatively low sensitivity. In this work, we introduce an entirely new class of MRI reporter genes that works by altering water diffusivity in cells. Specifically, we show that the human water channel aquaporin 1 (AQP1) is a sensitive, nontoxic, and metal-free reporter that produces robust MRI contrast by increasing diffusion of water molecules across the cell membrane.
Methods: We used lentiviral transfection to construct stable CHO, U87, and N2A cell lines expressing AQP1 and control cells expressing GFP. AQP1 and GFP-expressing cells were imaged using diffusion-weighted MRI to map effective water diffusivity across a broad range of gene expression levels and in mixed cell populations. Finally, AQP1 was used to image gene expression in a murine intracranial tumor model. All animal protocols were approved by the Institutional Animal Care and Use Committee of the California Institute of Technology.
Results: AQP1-expression was found to enhance water diffusivity in several mammalian cell lines by 82‒187%, resulting in enhanced contrast compared to GFP-expressing control cells imaged using diffusion-weighted MRI. Next, using doxycycline-dependent AQP1 expression, we demonstrated that AQP1-based increase in molecular diffusion of water can be used to reliably image gene expression across a broad, dynamic range spanning 0.01 to 10 μg/mL inducer concentration. Notably, AQP1 expression at submicromolar concentrations (~ 457 μM estimated) was found to be sufficient for producing significant MRI contrast, thereby placing AQP1 among the most sensitive MRI reporters. Furthermore, we developed Monte Carlo models for water diffusion in mixed cell populations, which predicted that AQP1-based increase in water diffusion should be evident even in heterogeneous populations comprising only a fraction of AQP1-labeled cells. Consistent with these predictions, our experiments revealed that as few as 10% AQP1 expressing cells is sufficient to produce MRI contrast in diffusion weighted images. Finally, we demonstrated that AQP1 can be used for imaging induced gene expression in mouse brain tumor xenografts, with the average intensity in AQP1 tumors decreasing by 41% following doxycycline administration.
Conclusions: Our work lays the foundation for AQP1 as the first in a class of MRI reporter genes that works by altering water diffusion in cells without impacting cell size or viability. As a sensitive, metal-free, and nontoxic reporter, AQP1 is well suited for dynamic imaging of gene expression in vivo. We anticipate that the wide prevalence of diffusion weighted MRI, together with the high performance, biocompatibility, and engineering capacity of aquaporins will enable the pervasive application of diffusion altering reporter genes in basic biological research as well as support the clinical development of genetic and cell-based therapies.