Regulation of energy balance and feeding behavior has been studied extensively with a broad variety of research techniques. A currently popular approach is functional magnetic resonance imaging (fMRI) of brain activation in subjects who are looking at food-related images. A drawback of this technique however, is that the functional relevance of specific MRI signals in relation to food-related behavior remains unknown. Our goal was to develop a novel MRI-based approach for detection of activation in neuronal networks associated with feeding behavior in a rat model.To evaluate the potential of pharmacological MRI (phMRI) to detect specifically activated neuronal networks relevant to regulation of feeding behavior with the use of DREADD (Designer Receptor Exclusively Activated by Designer Drug) technology in rats.

To target specific neuronal projections in vivo, rats received stereotactic injections of two viral vectors. First, an adeno-associated viral vector containing a floxed inverted sequence of the designer receptor which is exclusively activated by the designer drug clozapine-N-oxide (CNO) was injected in the ventral tegmental area (VTA). Second, a Cre-recombinase expressing canine adenovirus-2 (CAV2-Cre) was injected in specific target areas (see below and Figure 1). CAV2-Cre infects nerve terminals, travels retrogradely, and recombines the floxed DREADD-receptor so that this receptor becomes expressed on the cell surface. Only neurons that are infected with both vectors, thus only those neurons that project from the VTA to the specific target area, express the DREADD-receptor [1]. The DREADD-ligand CNO is injected intraperitoneally, crosses the blood-brain barrier and upon binding to the DREADD-receptor, leads to increased neuronal activation of the pathway targeted. We aimed to detect DREADD-evoked neuronal activity in specifically targeted projections and connected network regions. In three groups of adult male Wistar rats different projections were targeted: the VTA–nucleus accumbens (NAcc) projection (n=13), the VTA–medial prefrontal cortex (mPFC) projection (n=11), and the VTA–basolateral amygdala (BLA) projection (n=10). Two control groups were either injected with saline during phMRI (see below) (n=11), or received sham vectors (n=8).

MRI was done at 9.4T during which rats were anesthetized with 1.5% isoflurane in O₂/air (1:4). Blood oxygenation level-dependent (BOLD) phMRI was conducted in combination with injection of CNO. Pharmacological MRI data were acquired using a gradient echo multi slice (GEMS) sequence, with 25 slices acquired in coronal orientation, TR/TE = 500/15 ms, flip angle = 50^o, FOV = 32x32 mm, matrix size = 128x128 voxels, resolution = 250x250x500 μm, and 2 averages.

To calculate whole-brain activation maps, we applied a generalized linear model (GLM)-based analysis with the mean BOLD signal time-course in the VTA as a regressor for each specific group [2]. After MRI, brains were perfusion-fixated for immunohistochemical staining of cFos.

We detected a 1.5-2% increase in BOLD signal in target areas starting within 15 minutes after injection of CNO, which was not observed in the control groups. Strongest activation was measured in the VTA-NAcc group, with clear BOLD signal increases in the VTA (mean signal increase: 1.66 ± 0.13%) and NAcc (1.09 ± 0.06%), and also in remote connected brain regions, such as the mPFC (1.51 ± 0.08%) (Figure 2). Minor negative responses were measured in regions outside the targeted network, which may be explained by hemodynamic steal or neuronal deactivation. Pharmacological MRI findings were confirmed by cFos immunohistochemistry, which revealed neuronal activation in DREADD-targeted areas on a histological level. Furthermore, a positive correlation was found between increase in BOLD signal and number of cFos positive neurons in the VTA (r=0.38; p=0.003).Our study shows that specific DREADD-evoked activity can be detected with phMRI, which provides exciting opportunities to assess neuronal network activity in association with specific (feeding-related) behavioral phenotypes.