Molecular imaging with engineered physiology

Abstract

Using molecular imaging in vivo, biomolecular and cellular phenomena can be investigated within their relevant physiological context, addressing a central challenge for 21st century biomedicine and basic research. To advance neuroscience in particular, molecular-level measurements across the brain inside the intact organism are required. However, existing imaging strategies and available probes have been limited by serious constraints. Magnetic resonance imaging (MRI) provides deeper tissue penetration depth than optical imaging and better spatial resolution and greater versatility in sensor design than radioactive probes. The most important drawback for MRI probes has been the need for high concentrations in the micromolar to millimolar range, leading to analyte sequestration, complications for noninvasive brain delivery, and toxicity. Efforts to address the sensitivity problem, such as nuclear hyperpolarization, introduce their own technical constraints and so far lack generality. Here, we introduce a conceptually novel molecular imaging technique based on artificially induced physiological perturbations, enabling molecular MRI with nanomolar sensitivity. In this imaging strategy, we take advantage of blood as an abundant endogenous source of contrast compatible with multiple imaging modalities including MRI and optical imaging to decouple the concentration requirement for molecular sensing from the concentration requirement for imaging contrast. Highly potent vasoactive peptides are engineered to respond to specific biomolecular phenomena of interest at nanomolar concentrations by inducing dilation of the microvasculature, increased local bloodflow, and consequently, large changes in T₂*-weighted MRI contrast. This principle is exploited to design activatable probes for protease activity based on the calcitonin gene-related peptide (CGRP) and validate them for brain imaging in live rats; to use CGRP as a genetic reporter for cell tracking; and to create fusions of a vasoactive peptide from flies to previously characterized antibodies capable of crossing the blood-brain barrier (BBB), suggesting the possibility of minimally invasive brain delivery of such probes. We demonstrate the feasibility of highly sensitive molecular MRI with vasoactive probes at concentrations compatible with in situ expression of probes and delivery across the BBB, and show that vasoactive peptides are a versatile platform for MRI probe design which promises unprecedented in vivo molecular insights for biomedicine and neuroscience.