Microscopy

Laser confocal scanning microscopy was performed with a Zeiss LSM 780 microscope at room temperature. Images were taken with a 10 × objective with a numerical aperture (NA) of 0.45; a 20 × objective with a NA of 0.8; a 63 × oil objective with a NA of 1.46 or a 100 × oil objective with a NA of 1.46 (all objectives from Zeiss). The 63 × and the 100 × objective were used with fluorescence-free Immersol immersion oil 518F from Zeiss. An overview of the excitation and emission peak of used fluorochromes, the light source and filter set used for excitation, and the emission range used are provided in Table ​Table22.

Used fluorochromes

Overview of the excitation and emission peak of used fluorochromes, the light source and filter set used for excitation, and the emission range used

All tiled overview images were acquired with 10% overlap between tiles. Image stacks acquired for systematic random sampling consisted of either 1 × 18 or 2 × 18 tiles. Image stacks covered the entire uppermost 40 µm of prelimbic cortex in each slice imaged. Boundaries of prelimbic cortex were determined in accordance with those presented in a rat brain atlas (Paxinos and Watson 2004) and location of prelimbic cortex in each section was determined based on histological markers and distance from pia mater. Each image stack spanned five single focal plane images with 2.32 µm between single focal plane images. Image stacks with 1 × 18 tiles were a total of 134.95 × 2199.70 × 11.60 μm and image stacks with 2 × 18 tiles were a total of 256.41 × 2199.70 × 11.60 μm. The size of each voxel was 0.1318 × 0.1318 × 2.32 μm. However, even though voxels were 2.32 µm long, according to an equation to determine the full-width-of-half-maximum for an Airy disk provided by Zeiss (Toomre et al. 2019) the actual span of each single focal plane image was approximately 420–460 nm. Superimpositions between ventral motor thalamic axon terminals in the uppermost 40 µm of prelimbic cortex and postsynaptic structures were quantified in each of the five single focal plane images in the image stack using a technique known as systematic random sampling (for details see “Systematic Random Sampling”). Single focal plane images were not combined into a maximum intensity projection.

Expression of VGluT1 and VGluT2, but absence of GAD67 in VM axon terminals in prelimbic cortex was confirmed in image stacks that consisted of 10 or 11 single focal plane images with 0.58 µm between single focal plane images. The number of single focal plane images as well as distance between single focal plane images were adapted from (Lei et al. 2013), who were examining presence and absence of neurotransmitters in thalamostriatal axon terminals. Image stacks with 10 single focal plane images had a total size of 134.9511 × 134.9511 × 5.22 μm and those with 11 single focal plane images had a total size of 134.9511 × 134.9511 × 5.80 μm. The size of each voxel was 0.1318 × 0.1318 × 0.58 μm. However and as pointed out above even though voxels were 0.58 µm long, according to an equation to determine the full-width-of-half-maximum for an Airy disk provided by Zeiss (Toomre et al. 2019) the actual span of each single focal plane image was approximately 420–460 nm. Images cropped from maximum intensity projections performed across each image stack (methods adapted from (Lei et al. 2013) as well as, to examine absence of neurotransmitters, images cropped from single focal plane images spanning 0.58 μm are shown.

To determine whether RGS14 and 5HT3aR are expressed pre- or postsynaptic, we took two image stacks for each animal consisting of 10 single focal plane images with 0.14 μm between single focal plane images in prelimbic cortical layer I. The total size of each image stack was 134.9511 × 134.9511 × 1.40 μm and the size of each voxel was 0.066 × 0.066 × 0.14 μm. To determine whether RGS14 is expressed in dendrites of prelimbic pyramidal neurons that were marked with MAP2, for each animal one image stacks consisting of 10 single focal plane images with 0.13 μm between single focal plane images was taken in prelimbic cortical layers II/III. The total size of each image stack was 134.9511 × 134.9511 × 1.30 μm and the size of each voxel was 0.066 × 0.066 × 0.13 μm. All other image stacks span 20 single focal plane images with 0.29 μm between single focal plane images. The total size of each of these image stacks was 134.9511 × 134.9511 × 5.86 μm and the size of each voxel was 0.1318 × 0.1318 × 0.2893 μm. To determine if VM axon terminals in the uppermost 40 μm of prelimbic cortex superimpose with dendrites of pyramidal neurons, CS neurons or 5HT3aR-positive inhibitory interneurons in cortical layer I we looked for superimpositions in these single focal plane images that each were 0.29 µm thick. Single focal plane images were not combined into a maximum intensity projection. Images shown below were cropped from all the above mentioned single focal plane images spanning 0.29 μm, 0.13 μm or 0.14 μm.

Images were acquired using the ZEN 2011 SP7 FP1 software (Zeiss, black edition, version 14.0.8.201). Brightness and contrast were adjusted using ZEN 2011 SP7 FP1 software (Zeiss, black edition, version 14.0.8.201) and the “Enhance Contrast” option in Fiji (ImageJ version 1.51n or 1.52p). When necessary, brightness and contrast were adjusted separately for each channel using the mentioned software. 3D-reconstructions from these brightness- and contrast-adjusted images were performed using Imaris (version 9.3.0 64x). For final publication, brightness and contrast of images were additionally adjusted using Adobe Photoshop CS5 Extended (version 12.0.4 × 64). Adjustments to brightness and contrast were always applied equally across the entire image.

In addition, histological verification of AAV5-CAG-ArchT-GFP and AAV5-CAG-GFP injections into ventral motor thalamic nuclei was performed with a fixed stage BX51WI upright microscope from Olympus. A 100 W mercury lamp was used as light source. Alexa Fluor 594 was visualized using an excitation filter with a peak at 555 nm and a bandwidth of 25 nm, and an emission filter with a peak at 605 nm and a bandwidth of 25 nm. GFP was visualized using an excitation filter with a peak at 484 nm and a bandwidth of 15 nm, and an emission filter with a peak at 517 nm and a bandwidth of 30 nm. Images were acquired with a 4 × objective with a NA of 0.16 using the Neurolucida software (MBF Bioscience).