Mechanistic Insights for Magnetic Imaging and Control of Cellular Function

The vast biomolecular toolkit for optical imaging and control of cellular function has revolutionized the study of in vitro samples and superficial tissues in living organisms but leaves deep tissue unexplored. To look deeper in tissue and observe system-level biological function in large organisms requires a modality that exploits a more penetrant form of energy than visible light. Magnetic imaging with MRI reveals the previously unseen, with endogenous tissue contrast and practically infinite penetration depth. While these clear advantages have made MRI a cornerstone of modern medical imaging, the sparse library of molecular agents for MRI have severely limited its utility for studies of cellular function in vivo. The development of new molecular agents for MRI has suffered from a lack of tools to study the connection between changes in the microscale cellular environment and the corresponding millimeter-scale MRI contrast. Bridging this gap requires revisiting the mechanistic underpinnings of MRI contrast, casting aside some of the simplifications that smooth over sub-voxel heterogeneity that is rich with information pertinent to the underlying cell state. Here, we will demonstrate theoretical, computational, and experimental connections between subtle changes in microscale cellular environment and resultant MRI contrast. After reviewing some foundational principles of MRI physics in the first chapter, the second chapter of the thesis will explore computational models that have significantly enhanced the development of genetically encoded agents for MRI, including the first genetically encoded contrast agent for diffusion weighted imaging. By improving the efficacy of these genetically encoded agents, we unlock MRI reporter genes for in vivo studies of cellular dynamics much in the same way that the engineering of Green Fluorescent Protein has dramatically improved in vitro studies of cellular function. In the third chapter, we introduce our study that maps microscale magnetic fields in cells and tissues and connects those magnetic fields to MRI contrast. Such a connection has previously been experimentally intractable due to the lack of methods to resolve small magnetic perturbations with microscale resolution. To overcome this challenge, we leverage nitrogen vacancy diamond magnetometry to optically probe magnetic fields in cells with sub-micron resolution and nanotesla sensitivity, together with iterative localization of field sources and Monte Carlo simulation of nuclear spins to predict the corresponding MRI contrast. We demonstrate the utility of this technology in an in vitro model of macrophage iron uptake and histological samples from a mouse model of hepatic iron overload. In addition, we show that this technique can follow dynamic changes in the magnetic field occurring during contrast agent endocytosis by living cells. This approach bridges a fundamental gap between an MRI voxel and its microscopic constituents and provides a new capability for noninvasive imaging of opaque tissues. In the fourth chapter, we focus on the use of magnetic fields to perturb, rather than image, biological function. Recent suggestions of nanoscale heat confinement on the surface of synthetic and biogenic magnetic nanoparticles during heating by radiofrequency alternating magnetic fields have generated intense interest due to the potential utility of this phenomenon in non-invasive control of biomolecular and cellular function. However, such confinement would represent a significant departure from classical heat transfer theory. We present an experimental investigation of nanoscale heat confinement on the surface of several types of iron oxide nanoparticles commonly used in biological research, using an all-optical method devoid of potential artifacts present in previous studies. By simultaneously measuring the fluorescence of distinct thermochromic dyes attached to the particle surface or dissolved in the surrounding fluid during radiofrequency magnetic stimulation, we found no measurable difference between the nanoparticle surface temperature and that of the surrounding fluid for three distinct nanoparticle types. Furthermore, the metalloprotein ferritin produced no temperature increase on the protein surface, nor in the surrounding fluid. Experiments mimicking the designs of previous studies revealed potential sources of artifacts. These findings inform the use of magnetic nanoparticle hyperthermia in engineered cellular and molecular systems and can help direct future resources towards tractable avenues of magnetic control of cellular function.