Subjects and In Vivo Fixation
All animal procedures and protocols were conducted in accordance with the National Institute of Health Guidelines for the Care and Use of Laboratory Animals. The transgenic mouse line on a C57BL/6 background strain had a knockin of SEP on the amino-terminus of D2Rs and was generated in the Transgenic Mouse Facility at the University of California at Irvine. The procedure for transgenic mouse generation and evidence supporting the normal functionality of SEP-tagged D2Rs has been published previously (Robinson et al. 2017). For these ultrastructural analyses, mice from the Williams laboratory were perfused by S. Aicher, and their brains were shipped to the University of Pittsburgh for examination.
SEP-D2R transgenic mice were anesthetized with 100 mg/kg sodium pentobarbital i.p. and then perfused with 5-10 ml of saline containing 1000 units/ml of heparin. The cardiovascular system was then flushed with a 0.1 M phosphate buffer (PB; pH 7.4) containing one of two fixatives: 12.5 ml of 3.75% acrolein with 2% paraformaldehyde, followed by 50-75 ml of 2% paraformaldehyde, or 50-75 ml of 0.2% glutaraldehyde with 4% paraformaldehyde. Brains were extracted from the skull and post-fixed in the last fixative for 30-60 min. Brains showing optimal fixation were then stored in PB for overnight shipment to the University of Pittsburgh. Three cohorts consisting of 3, 5, and 3 brains were shipped. The first two cohorts were used for quantitation of SEP-D2R in the SNc and striatum (8 brains/animals total). The third cohort was used for dual labeling of SEP-D2R and AIS markers for electron microscopy. In addition, experimental tests to determine the ideal conditions for localizing antibodies against ankyrin-G and beta IV-spectrin were conducted in 12 wildtype mice of the C57BL/6 strain at the University of Pittsburgh following equivalent procedures to those described above.
Immunogold-silver and Immunoperoxidase Labeling
Fixed brains were cut into blocks containing the SNc or striatum that were then sectioned at 50 µm on a vibratome. Sections were treated with PB containing 1% sodium borohydride followed by rinsing with PB. Tissue was then incubated for 30 min in a blocking solution made in 0.1 M tris-buffered saline (TBS; pH 7.6) that contained 3% normal goat serum, 1% bovine serum albumin, and 0.04% Triton-X100. Sections were then transferred to chicken anti-GFP primary antibody (Avēs Labs, Inc; GFP-1010) at a concentration of 1:500 or 1:1000 in blocking solution. According to the vendor, antibody specificity has been determined using Western blot and immunohistochemistry from transgenic mice expressing GFP. Our laboratory has also determined that brain sections from non-GFP expressing animals contain no detectable immunoreactive signal at light or electron microscopy. Tissue sections were incubated in the GFP primary antibody overnight at room temperature for 12-15 hours, then rinsed in 0.1 M TBS followed by 0.01 M phosphate-buffered saline (PBS; pH 7.4).
Tissue was then rinsed with a washing buffer containing 3% normal goat serum, 0.8% bovine serum albumin, and 0.1% cold fish gelatin (Aurion). The secondary IgG antibody was 0.8 nM gold-conjugated goat anti-chicken (Aurion). Tissue was exposed to the secondary at 1:50 in washing buffer and incubated overnight for 12-15 hours at room temperature. Excess secondary was removed by rinsing three times each in washing buffer and PBS. Tissue was then exposed to 2.5% glutaraldehyde in PBS for 10 min, followed by several rinses in PBS. Sections were then treated four times for 10 min each with Enhancement Conditioning Solution (Aurion) diluted 1:10 with ultrapure water. Sections were exposed to an RGENT-SEM proprietary silver enhancement solution (Aurion) for 120-180 min. Additional treatments in Enhancement Conditioning Solution followed at four times 10 min each, with a final rinse in 0.1 M PB.
The probable detection of SEP-D2R in axon initial segments necessitated confirmation using antibodies specific for this neuronal compartment, namely beta IV-spectrin (Berghs et al. 2000) (Lacas-Gervais et al. 2004) and ankyrin-G (Hedstrom et al. 2008; Le Bras et al. 2014). An affinity purified, rabbit polyclonal antibody against beta IV-spectrin was obtained as a gift from Dr. Matthew Rasband at Baylor College of Medicine. The specificity of this antibody was previously demonstrated by preadsorption with the antigen, Western blot analysis, and complete loss of immunolabeling in mice with a knock-out of the beta IV-spectrin Σ1 isoform (Lacas-Gervais et al. 2004) (Lysakowski et al. 2011). A mouse monoclonal antibody against ankyrin-G was purchased from the UC Davis/NIH NeuroMab Facility. Specificity was previously demonstrated by preadsorption with the peptide antigen and Western blot analysis (Le Bras et al. 2014). The two antibodies were used at concentrations from 1:50 to 1:1000 in procedures similar to those described above, except that secondary antibodies were biotinylated goat anti-mouse for ankyrin-G and goat anti-rabbit for beta IV-spectrin, both from Vector Laboratories. Tissue was incubated with secondary antibodies at 1:400 for 30 min and rinsed three times in TBS for 10 min each. Antibody-bound tissue was then exposed to a 30 min incubation in avidin-biotin peroxidase complex (ABC solution, Vector Laboratories) before being rinsed again in TBS three times for five min each. Tissue was then exposed for 3.5 min to a solution containing 0.022% diaminobenzidine and 0.003% hydrogen peroxide, followed by a three-step rinse in TBS for 5 min each and one rinse in PB.
For dual labeling of both SEP-D2R and one of the AIS markers, mouse brains were fixed with acrolein and sectioned, and then the sections were incubated in a mixture of primary antibodies. From pilot testing, the best results were obtained with anti-GFP at 1:1000 and anti-beta IV-spectrin at 1:100. Immunoperoxidase processing occurred before immunogold labeling, and then silver enhancement was carried out for 60-70 min.
Immunofluorescence Imaging of SEP-D2R in the AIS
All procedures were approved and performed in compliance with the appropriate guidelines set forth by the Institutional Animal Care and Use Committee at Oregon Health & Science University. Mice were group-housed in vented plastic cages on a 12-hour light/dark cycle with food and water available ad libitum. Male transgenic mice homozygous for the SEP-D2R knockin (21-25 days old) were used for this part of the study.
Animals were anesthetized with isoflurane and rapidly decapitated into modified Krebs buffer at 34°C containing (in mM): 126 NaCl, 2.5 KCl, 1.2 MgCl2, 2.4 CaCl2, 1.4 NaH2PO4, 25 NaHCO3, and 11 D-Glucose. Krebs buffer used for decapitation, slicing, and slice recovery contained 10 µM MK-801. Horizontal slices were cut at a thickness of 222 µm using a Leica VT1000S vibratome and allowed to recover for 30 min, both in Krebs bubbled with 95/5% O2/CO2. To amplify the SEP-D2R signal, slices were incubated for 45 min in bubbled Krebs with an anti-GFP antibody (1:400) conjugated to Alexa Fluor 488 (ThermoFisher cat#: A-21311). Excess antibody was washed with a 15 min incubation in Krebs following labeling. Fixation was achieved by incubating slices in 4% PFA in PBS for 30 min at RT. Slices were blocked and permeabilized in PBS containing 0.5% fish skin gelatin (FSG) and 0.5% Tween-20 for 1 hr at RT. Primary antibody incubation was conducted overnight at 4°C in PBS containing 0.1% Tween-20 and 0.5% FSG. Anti-tyrosine hydroxylase was purchased from Sigma Aldrich (cat#: T1299); anti-Ankyrin-G was a generous gift from Paul Jenkins (University of Michigan). Both primary antibodies were used at 1:500. Slices were washed 3X for 15 min in PBS at RT before being labeled with secondary antibody at 1:500 for 1 hr at RT in the same solution used for primary labeling. Secondaries used were anti-mouse IgG conjugated to Alexa Fluor 555 (ThermoFisher cat# A-21422) and anti-rabbit IgG conjugated to Alexa Fluor 647 (ThermoFisher cat# A-21245). Slices were washed 4X for 15 min in PBS before being mounted and coverslipped using Fluoromount aqueous mounting medium (Sigma-Aldrich cat#: F4680) and nail polish to seal.
Slides were imaged on a Zeiss LSM980 with AiryScan2 using a 63X oil immersion objective. Z-stacks of labeled cells were acquired in Airyscan SR mode with a step size of 0.15 µm and a pixel size of 0.14 µm. Airyscan2 processing was carried out in Zen Blue, and images were presented as a maximum intensity projection of the compressed z-stack.
Electron Microscopy Tissue Preparation and Image Analysis
Lipids in brain sections were fixed with 1% osmium tetroxide in 0.1 M PB for 1 hour. After rinsing in PB, tissue was exposed in five min increments to increasing concentrations of ethanol in water: 30%, 50%, 70%, and 95%, followed by two rounds of 100% ethanol. This was followed by two five min rounds of propylene oxide incubation. Samples were then placed overnight in a 1:1 mixture of propylene oxide and epoxy resin (EM-Bed 812, Electron Microscopy Sciences) before being transferred to 100% epoxy for 2-3 hours. Sections were then flat-embedded between two clear sheets of commercial plastic, and the resin was cured at 60°C for 72 hours until solid. Plasticized sections were cut to include the areas of interest, either SNc or dorsolateral striatum. These cut pieces were then glued onto blocks of solid epoxy resin. The sections were trimmed into trapezoidal shapes that contained the area of interest identified by landmarks such as blood vessels and major white matter tracts. After trimming, ultrathin sections (60 nm) were cut from the resin blocks using an ultramicrotome and collected onto 400 mesh copper grids. The grid-bound tissue was then counterstained with uranyl acetate and lead citrate to enhance contrast.
Tissue was observed using a FEI Morgagni transmission electron microscope. Only one ultrathin section per grid was selected for initial analysis to avoid double counting of immunopositive profiles in serial sections. The area analyzed was near the interface between tissue and plastic resin where antibody penetration is maximal. Roughly 3,000 to 12,000 µm2 of tissue was examined per animal, and approximately 15 micrographs were captured from each tissue section. Most micrographs contained multiple immunolabeled profiles.
The MicroBrightField Neurolucida program v.9 was used to analyze immunopositive neuronal profiles, defined as containing at least three gold particles, regardless of location, i.e. plasma membrane or intracellular. Many profiles contained more than three immunogold particles. Neurolucida was used to trace and quantify the perimeter and area of labeled profiles, while the short-axis diameter was measured manually on the micrographs with a ruler.
To compare the prevalence and density of SEP-D2R in different cellular compartments, immunopositive structures were identified based on established morphological characteristics (Peters et al. 1991). Dendrites were recognized by their larger size, smoothly contoured membranes, organelle content, and receipt of synapses; organelles consisted of mitochondria and smooth endoplasmic reticulum in the smallest dendrites, as well as rough endoplasmic reticulum and Golgi bodies in the largest. Infrequent dendritic spines in the SNc were identified by their thorny or bulbous shape and clear protrusion from dendrites. In the striatum, more numerous dendritic spines were recognized by their small size, round or cup shape, absence of mitochondria, and tendency to be postsynaptic to axon terminals. Axon varicosities were identified by their round structure and concentrations of small clear vesicles, often clustered near synapses. If axons formed a synapse, the type of synapse was recorded as either asymmetric – having parallel membranes, a widened space between membranes spanned by filaments, and a thick postsynaptic density, or symmetric – having the same features but for a thin or absent postsynaptic density (Peters et al. 1991) (Carlin et al. 1980). Astrocytes were identified by their vacuous cytoplasm and serpentine shape, often becoming markedly narrow to fill in spaces between other cellular profiles. Structures were considered unidentified if they lacked substantially distinguishing traits in single sections, which was more common with smaller profiles.
The location of immunogold-silver particles within neuronal profiles was also determined using Neurolucida. SEP-D2Rs were considered to be associated with an organelle or the plasma membrane if the corresponding immunogold particle was within 20 nm of the visible structure, as determined using a 20 nm cursor in Neurolucida. This distance was chosen based on the approximate size of antibodies and the corresponding length of the immunogold antibody complex (Mathiisen et al. 2006). In further analysis of membrane-associated SEP-D2Rs, the neighboring cellular profile immediately apposing each gold particle was also characterized and recorded: dendrite, axon, astrocyte, or unidentified.
For each type of immunoreactive profile in each animal, the number of membrane-associated immunogold particles were totaled and used to determine an average density of immunolabeling per perimeter. Similarly, the total number of immunogold particles were summed and used to calculate the average density of immunolabeling per area. Finally, the proportion of gold particles that were associated with the plasma membrane was expressed as a percentage of the total for each profile. Regional comparisons of immunogold density and membrane distribution of immunolabeling utilized Student t-tests assuming unequal variance, and reported p values were two-tailed. Categorical comparison of synapse type formed by immunolabeled axons between the SNc and striatum was performed using Fisher’s Exact Test.