Directed evolution of a magnetic resonance imaging contrast agent for noninvasive imaging of dopamine

Abstract

The development of molecular probes that allow in vivo imaging of neural signaling processes with high temporal and spatial resolution remains challenging. Here we applied directed evolution techniques to create magnetic resonance imaging (MRI) contrast agents sensitive to the neurotransmitter dopamine. The sensors were derived from the heme domain of the bacterial cytochrome P450-BM3 (BM3h). Ligand binding to a site near BM3h's paramagnetic heme iron led to a drop in MRI signal enhancement and a shift in optical absorbance. Using an absorbance-based screen, we evolved the specificity of BM3h away from its natural ligand and toward dopamine, producing sensors with dissociation constants for dopamine of 3.3-8.9 microM. These molecules were used to image depolarization-triggered neurotransmitter release from PC12 cells and in the brains of live animals. Our results demonstrate the feasibility of molecular-level functional MRI using neural activity-dependent sensors, and our protein engineering approach can be generalized to create probes for other targets.

Figure 1

Ligand binding to the BM3 heme domain changes MRI contrast and optical absorption in a concentration-dependent manner. (a) T₁ relaxivity (r₁) of BM3h in PBS solution and in the presence of 400 μM arachidonic acid (AA) or 1 mM dopamine (DA); inset shows T₁-weighted spin echo MRI image intensity (TE/TR = 10/477 ms) of microtiter plate wells containing 240 μM BM3h in PBS alone (left) or in the presence of 400 μM arachidonic acid (middle) or 1 mM dopamine (right). (b) T₁ relaxation rates (1/T₁) measured from solutions of 28.5 μM BM3h incubated with 0–250 μM arachidonic acid. (c) Optical absorbance spectra of 1 μM BM3h measured alone (blue) and after addition of 400 μM arachidonic acid (gray) or 1 mM dopamine (orange). OD, optical density. (d) Difference spectra showing the change in BM3h absorbance as a function of wavelength upon addition of 400 μM arachidonic acid (gray) or 1 mM dopamine (orange). (e) Normalized titration curves showing binding of BM3h to arachidonic acid (gray) or dopamine (orange). We computed the optical signals used for titration analysis by subtracting the minimum from the maximum of difference spectra (arrowheads in d) under each set of conditions. Error bars in a, b and e reflect s.e.m. of three independent measurements (errors in e were smaller than the symbols).

Figure 2

Screen-based isolation of BM3h mutants with enhanced dopamine affinity. (a) Schematic of the directed evolution approach, including (left to right) generation of a mutant DNA library, transformation into E. coli and growth in multiwell plate format, spectroscopic analysis of each mutant’s ligand binding affinities, and detailed MRI and optical characterization of selected mutant proteins. (b) Histograms of mutant dopamine dissociation constants determined during each round of directed evolution, comparing each mutant protein’s relative dopamine affinity (measured in plate format) to the K_d of the parent protein (measured in bulk). K_d distributions for screening rounds 1 (black), 2 (green), 3 (red), 4 (cyan) and 5 (purple) are labeled with numbers in circles. Color-coded arrowheads indicate the measured K_ds of parent proteins used to create the library of mutants at each round; yellow arrowhead indicates the K_d of the mutant protein selected after round 5. (c) Dissociation constants for dopamine (DA; orange) and arachidonic acid (AA; gray) for WT BM3h and mutant BM3h variants isolated at each round of screening; progressive increases in dopamine affinity and attenuation of arachidonic acid affinity are evident. Colored arrowheads indicate correspondence with data in b. Error bars denote s.e.m. of three independent measurements. (d) Titration analysis of dopamine binding to WT BM3h and to proteins selected after each round of directed evolution (colored as in b). Mutant proteins identified by rounds 4 (8C8) and 5 (B7) were considered to be end products of the screening procedure. (e) X-ray crystal structure of WT BM3h (gray; heme group shown in orange) bound to palmitoleic acid (black), indicating the locations of amino acid substitutions accumulated during directed evolution of enhanced dopamine binding affinity. Each mutation’s location is marked with a blue sphere and a label color-coded according to the parent protein in which the substitution was first identified (see legend for b). The previously characterized I366V mutation (asterisk) was incorporated between screening rounds 4 and 5 to improve the thermostability of the engineered proteins.

Figure 3

Selected sensor proteins produce strong and specific MRI signal changes in response to dopamine. (a) Relaxivity values measured from BM3h-B7 (yellow bars) and BM3h-8C8 (purple bars) in PBS alone or in the presence of 400 μM dopamine (DA). Inset, T₁-weighted MRI signal (TE/TR = 10/477 ms) obtained from 195 μM BM3h-B7 or BM3h-8C8, each incubated in microtiter plate wells with or without 400 μM dopamine (wells ordered left to right as in the bar graph). (b) MRI image showing signal amplitudes measured from wells containing 28.5 μM WT BM3h, BM3h-8C8 or BM3h-B7, each incubated with increasing dopamine concentrations (0–63 μM, left to right). The image was obtained using a T₁-weighted pulse sequence (TE/TR = 10/477 ms). (c) Relaxation rates (1/T₁ values) measured from solutions of 28.5 μM WT BM3h (black), BM3h-B7 (yellow) or BM3h-8C8 (purple), as a function of total dopamine concentration. Curves were fitted using a ligand-depleting bimolecular association model. (d) Changes in 1/T₁ relative to ligand-free protein for 28.5 μM BM3h-B7 (yellow) or BM3h-8C8 (purple) incubated with 30 μM dopamine, serotonin (5HT), norepinephrine (NE), DOPA, arachidonic acid (AA), acetylcholine (ACh), GABA, glutamate or glycine. Inset, spectroscopically determined affinities (K_a = 1/K_d) of BM3h-B7 and BM3h-8C8 for dopamine, serotonin and norepinephrine. Error bars in panels a, c and d denote s.e.m. of three independent measurements.

Figure 4

BM3h-based sensors measure dopamine release in cell culture. (a) PC12 cells depolarized by addition of 54 mM K⁺ were stimulated to release dopamine (DA) into supernatants containing a BM3h-based sensor; cells did not release dopamine after addition of 54 mM Na⁺. (b) T₁-weighted spin echo MRI signal amplitudes (TE/TR = 10/477 ms) measured from the supernatants of PC12 cells incubated with 32 μM BM3h-B7 in the presence of K⁺ (stimulus) or Na⁺ (control). Inset, MRI image of microtiter wells under corresponding conditions. (c) Relaxation rates measured from the samples in b, minus the relaxation rate of buffer not containing BM3h-based sensors. Given the approximate concentration of BM3h variants in these samples, the Δ(1/T₁) values presented here can be converted to apparent relaxivities of 0.23 and 0.50 mM⁻¹ s⁻¹ in K⁺ and Na⁺ incubation conditions, respectively. (d) Data from c were used to estimate the concentrations of dopamine present in samples treated with K⁺ and Na⁺ (dark bars). We independently measured the concentrations of dopamine under equivalent conditions using ELISA (light bars).

Figure 5

BM3h-8C8 reports dopamine in injected rat brains. (a) Top, coronal MRI image (0.7 mm anterior to bregma, averaged over the injection period) from a single rat injected with 500 μM BM3h-8C8 in the presence (orange dashed circle) or absence (blue dashed circle) of equimolar dopamine; the image contrast was linearly adjusted for display. MRI hyperintensity is noticeable near the tip of the dopamine-free cannula. The circles indicate approximate ROIs (~1.5 mm around cannula tips) over which image intensity was averaged for quantitative analyses. Bottom, map of percent signal change (%Δ) for the same animal, computed by comparing pre- and post-injection MRI signal. Areas corresponding to both high- and low-dopamine co-injections (+DA and −DA) are delineated by apparent signal changes, but the strong difference between the two conditions is clear. (b) Time courses of relative signal change observed during injection of BM3h-8C8 −DA (blue) or +DA (orange), averaged over multiple animals (n = 7) in ROIs denoted in a. Gray shading denotes the 20-min injection period. (c) Corresponding time courses of a control injection in which WT BM3h was introduced instead of the dopamine sensor (n = 5). (d) Statistical parametric map of t-test significance values (color scale) for correlation of MRI intensity with low- and high-K⁺ conditions in an individual rat, overlaid on a corresponding T₁-weighted coronal slice (grayscale) showing injection cannulae used for BM3h-8C8 infusion (left, purple dashed circle) and WT BM3h control infusion (right, black dashed circle). (e) Maps of percent signal difference (SD) between high- and low-K⁺ conditions observed in 2.7-mm-diameter ROIs centered around BM3h-8C8 sensor (left) and WT BM3h control (right) injection sites, after spatial coregistration and averaging across multiple animals (n = 6); ROIs correspond approximately to the color-coded circles in d. Voxels outlined in green are those that showed the most significant correlation with the K⁺ stimulus regressor in the group analysis (Student’s t-test, P < 0.01); these generally showed ~1% mean signal change. Gray cross-hatching indicates approximate locations of the infusion cannulae. (f) Mean MRI signal change from baseline observed during high-K⁺ (dark bars) and low-K⁺ (light bars) periods in ROIs centered around infusion sites for BM3h-8C8 (purple) and WT BM3h (gray) proteins. ROIs were cylinders 2.7 mm in diameter and extending over three 1-mm-thick slices registered around the infusion sites; signal was averaged in unbiased fashion over all voxels, regardless of correlation with the stimulus. The signal difference in the presence of BM3h-8C8 was statistically significant (P = 0.0008, asterisk). (g) Graph shows the mean time course of MRI signal in voxels within the BM3h-8C8–infused ROI and identified as correlated (P < 0.05) with the stimulus, averaged over animals and binned over 1.5-min intervals (shaded area denotes s.e.m., n = 6; individual traces are shown in Supplementary Fig. 6 online). Gray vertical bars denote periods associated with highest K⁺ stimulation, accounting for delays due to convective spreading of K⁺ from the cannulae tips and the dead time of the injection apparatus. Arrowheads indicate the timing of pump switches associated with transitions from low to high (up) and from high to low (down) K⁺ infusion conditions. Panels above the graph depict ‘snapshots’ of signal change spaced throughout the first K⁺ stimulation cycle, as indicated by the dotted lines. The ROI corresponds to the left side of e, and the color scale denotes 0% (black) to 3% (yellow) signal change from baseline at each voxel and time point.