Key Roles of CACNA1C/Cav1.2 and CALB1/Calbindin in Prefrontal Neurons Altered in Cognitive Disorders

Key Points Question Why do genetic risk studies of mental disorders find consistent associations with CACNA1C? Findings In this study, the dorsolateral prefrontal cortical pyramidal cells most affected in cognitive disorders express elevated calcium-related signaling, delineated by CALB1 (calcium-buffering protein calbindin), and high CACNA1C (L-type calcium channel [LTCC] Cav1.2), GRIN2B (NMDA receptor GluN2B), and KCNN3 (SK3 channel) expression. These neurons require LTCC actions to sustain memory-related firing, but excessive levels, such as during stress, reduce firing via SK channel opening and induce pathology, especially when calbindin is lost with age and/or inflammation. Meaning These data explain why both loss- and gain-of function variants in CACNA1C are associated with an increase in risk of cognitive disorders.


eMethods
All research was conducted according to USDA and NIH guidelines and approved by the Yale or University of Pittsburgh IACUC (for macaques) and the Partners Human Research Committee (for humans).For human dlPFC tissues, brain donors were recruited by the Harvard Brain Tissue Resource Center/NIH NeuroBioBank (HBTRC/NBB), in a community-based manner, across the USA.Human brain tissue was obtained from the HBTRC/NBB.The HBTRC procedures for informed consent by the donor's legal next-of-kin and distribution of de-identified post-mortem tissue samples and demographic and clinical data for research purposes are approved by the Mass General Brigham Institutional Review Board.Post-mortem tissue collection followed the provisions of the United States Uniform Anatomical Gift Act of 2006 described in the California Health and Safety Code section 7150 and other applicable state and federal laws and regulations.Federal regulation 45 CFR 46 and associated guidance indicates that the generation of data from de-identified post-mortem specimens does not constitute human participant research that requires institutional review board review.

Transcriptomics
The methods for transcriptomic analyses are summarized in graphic form in eFigure 1, below.Fresh-frozen postmortem human dlPFC tissue from 50 neurotypical donors from the Harvard Brain Tissue Resource Center/NIH NeuroBioBank was processed for single nucleus RNA-seq (10x 3' v3).These brains had been previously selected as matched controls for a separate study on brain differences in schizophrenia; this sample had the following demographics: 22% female, 78% male; ages 50-88 years, all Caucasian with the exception of one American Indian/Alaskan Native and one Asian/Pacific Islander.The generation of expression profiles will be described in detail in a forthcoming manuscript from Ling, Goldman, Berretta and McCarroll.In brief, we developed and optimized a workflow for creating and analyzing "brain nuclei villages" -nuclei sampled from the dlPFC (BA46) from 20 different donors per village.Fresh-frozen postmortem macaque dlPFC tissue was provided by the Lewis group at the University of Pittsburgh and provided to the Krienen lab at Princeton.Fresh-frozen brain tissue (dlPFC containing both banks of the principal sulcus area 46) was obtained from two experimentally-naïve, female, adult (6.6 and 8.2 years of age) Rhesus macaque monkeys.Animals were deeply anesthetized and perfused with artificial cerebral spinal as previously described 1 in accordance with USDA and NIH guidelines and with the approval of the University of Pittsburgh IACUC.dlPFC tissue was dissected from each donor, obtaining a similar mass of tissue from each specimen while being careful to represent all cortical layers.The frozen tissue samples are then immediately pooled for simultaneous isolation of their nuclei; all subsequent processing stepsincluding nuclear isolation, encapsulation in droplets, and preparation of snRNA-seq librariesinvolve all of the donors together.This "Dropulation" workflow allows us to minimize experimental variability, including any technical effects on mRNA ascertainment and any effects of cell-free ambient RNA.Each nucleus in these experiments was then re-assigned to its donorof-origin using combinations of hundreds of transcribed SNPs; though the individual SNP alleles are shared among many donors, the combinations of many SNPs are unique to each donor in the cohort.
Nuclei with fewer than 500 genes or 1000 transcripts were removed from the analysis.Independent component analysis (ICA, using the fastICA package in R) was performed on the excitatory neurons, using the filtered digital gene expression matrix (DGE) after normalization and variable gene selection following procedures described in Saunders et al. 2018 2 .Louvain clustering (resolution 0.1, nearest neighbors = 25) was performed using the top 60 ICs.This resulted in 20 clusters, including 5 major GABAergic clusters (identified by expression of GAD1 or GAD2, and 7 major glutamatergic (SLC17A7+) clusters, as well as astrocytes, endothelial cells, microglia, oligodendrocytes and polydendrocytes.
ICA and Louvain clustering was repeated on each of the 7 glutamatergic clusters, resulting in further subdivisions of the major glutamatergic types for a total of 12 subclusters.Cell and gene weights on independent components (ICs) were manually inspected for evidence of cell-cell doublets or non-excitatory neurons, both of which were removed from the DGE.ICs that reflected artifactual signals or batch effects were identified and excluded, and Louvain clustering was performed on the retained ICs.Within each cluster, transcripts for each gene were summed across cells of that cluster.The resulting "metacell" counts were normalized to the total number of transcripts, then scaled to counts per 100,000 and log10 transformed.All data from the human cohort will be deposited in NeMO under accession number nemo:dat-bmx7s1t in a forthcoming manuscript from Ling et al.
Confocal images were acquired using a Leica TCS SP8 Gated STED 3X super resolution microscope, with the HC PL APO 100X/1.40oil white objective (Leica) and HCX PL APO CS 63X/1.40 oil white objective (Leica), or a Zeiss LSM 880 Airyscan with the Plan-Apochromat 20x/0.8M27 objective.Z-stacks were obtained with ~0.3-m steps under laser excitation at 407nm, 488nm, 543nm, 561nm, or 633nm, as determined by the antibody combination being imaged.Emission filter bandwidths and sequential scanning acquisition were set up in order to avoid possible spectral overlap between fluorophores.Confocal images were deconvoluted with Hugyens Professional version 22.04 (Scientific Volume Imaging, The Netherlands, http://svi.nl) using the theoretical PSF.The confocal Z-stacks were processed into maximum intensity Zprojections using Fiji and background subtraction with a rolling ball radius of 50 pixels and 100 pixels was applied to all 63X and 100X channels, respectively (applied to the entire panel).Images were adjusted for brightness and contrast, labeled, and assembled into a figure using Adobe Photoshop CS5 Extended (version 12.0.4×64, Adobe Systems Incorporated) and Adobe Illustrator 25.0.1 (Adobe Systems Incorporated version 27.7).
To quantify the percent of CB cells that were positive for Cav1.2, β1-AR, NMDAR-GluN2B, or SK3, we sampled multiple fields of view in layer III of each section.We used the isolated CB channel to trace lightly labeled CB+ neurons with characteristic pyramidal-like morphology, and measured the mean grey value of these traces in the isolated channel containing the second protein (e.g., NMDAR-GluN2B).We compared the mean grey value in each CB pyramidal neuron trace to the average mean grey value sampled from the neuropil.CB pyramidal cells were deemed colocalized with the second protein if the mean grey value exceeded the average from the sampled neuropil.

Single label IHC-
The expression pattern of the transcription factor, RORB, in macaque dlPFC was examined due to its correlation with Alzheimer's pathology (see eFig.11 below).This analysis utilized immunoperoxidase immunohistochemistry, where sections of macaque dlPFC (26-31y) were transferred for 1 hr to Tris-buffered saline (TBS) containing 5% bovine serum albumin, plus 0.05% Triton X-100 to block non-specific reactivity, and incubated in primary antibody for RORB (1:250 dilution; rabbit, HPA008393, Millipore Sigma) in TBS for 48 hr at 4°C.The tissue sections were incubated in goat anti-rabbit biotinylated antibody (Vector Laboratories) at 1:300 in TBS for 2 hr, and developed using the Elite ABC kit (Vector Laboratories) and diaminobenzidine (DAB) as a chromogen.
Unlike the other antibodies, the β1-AR antibody (rabbit anti-1 adrenergic receptor at 1:100 [cat#AAR023, RRID:AB_2340886, Alomone]) was not knockout-validated, and thus we performed an additional control, utilizing the manufacturer's bespoke blocking peptide (Alomone, cat # BLP-AR023).Tissue sections were processed for β1-AR immunofluorescence, following the antigen retrieval and blocking steps outlined above, and test sections were incubated with the primary antibody for β1-AR and the blocking peptide (at 10x the concentration of the primary antibody, as instructed).The sections were then paired with goat antirabbit AF488 (Invitrogen).As shown here, neuronal labeling was evident when the blocking peptide was omitted (A); negligible labeling was observed in the test section that had been processed with the blocking peptide (B).
In contrast to the β1-AR antibody, the Cav1.2,SK3 and Cav1.3 antibodies have all been validated in conditional knockout animals [4][5][6] .Normal sera and IgG-free BSA were purchased from Jackson Immunoresearch.All chemicals and supplies for electron microscopy were purchased from Sigma Aldrich and Electron Microscopy Sciences, respectively.
Sections were incubated for 72 h at 4 °C with primary antibodies in TBS, and transferred for 2 h at room temperature to species-specific biotinylated Fab' or F(ab')2 fragments in TBS.In order to reveal immunoperoxidase labeling, sections were incubated with the avidin-biotin peroxidase complex (ABC) (1:300; Vector Laboratories, Burlingame, CA, United States of America) and then visualized in 0.025% Ni-intensified 3,3-diaminobenzidine tetrahydrochloride (DAB; Sigma Aldrich, St. Louis, MO, United States of America) as a chromogen in 100mM PB with the addition of 0.005% hydrogen peroxide for 6-12 minutes, depending on the target of interest.After the DAB reaction, sections were exposed to osmification (concentration 1%), dehydration through a series of increasing ethanol concentrations (70-100%), and infiltrated with propylene oxide.Tissue sections were counterstained with 1% uranyl acetate in 70% ethanol.Standard epoxy resin embedding followed typical immunoEM procedures followed by polymerization at 60°C for 48 h.Omission of primary antibodies or substitution with non-immune serum resulted in complete lack of immunoperoxidase labeling.Similarly, labeling was nullified when blocking the biotinylated probes with avidin/biotin.Blocks containing dlPFC layer III were sampled and mounted onto resin blocks.The specimens were cut into 50 nm sections using an ultramicrotome (Leica) and analyzed under a Talos L120C transmission electron microscope (Thermo Fisher Scientific).Several plastic blocks of each brain were examined using the 4 th to 12 th surface-most sections of each block (i.e., 200-600 nm), in order to sample the superficial component of sections, avoiding penetration artifacts.Structures were digitally captured at x25,000-x100,000 magnification with a Ceta CMOS camera and individual panels were adjusted for brightness and contrast using Adobe Photoshop and Illustrator CC.2020.01image editing software (Adobe Systems Inc.).Approximately, 2000 micrographs of selected areas of neuropil with immunopositive profiles were used for analyses with well-defined criteria.
For quantification of the distance from the SER to membrane-bound Cav1.2,SK3, or β1-AR, we selected photomicrographs of spines with synaptic or extrasynaptic membrane-associated immunoperoxidase label and a discernable spine apparatus.Using Reconstruct 12 , we calibrated the pixel size of each image and measured the shortest distance from the center of the membrane bearing the immunoperoxidase label to the spine apparatus.Data were analyzed and plotted using Prism 9 (Graphpad Software, San Diego CA).Means are reported with standard error of the mean.
The criteria summarized in Peters et al. (A.Peters, S. L. Palay, H. D. Webster, The Fine Structure of the Nervous System: Neurons and Their Supporting Cells.Oxford Univ.Press, New York, NY, 1991) were utilized for profile identification.Dendritic spines in the PFC are typically long and thin, devoid of mitochondria, with the presence of a noticeable postsynaptic density at asymmetric synapses, and often containing an elaborated spine apparatus, the extension of the smooth endoplasmic reticulum into the spine.Dendritic shafts were typically round in perpendicular planes or irregular shaped when assessed in horizontal planes, usually containing mitochondria and numerous tubular and pleomorphic cellular organelles.Depending on the proximity to axon terminals, various dendritic shafts received synaptic inputs.Axon terminals contained accumulations of synaptic vesicles and the axoplasm of these terminals usually contained neurofilaments and mitochondria.The synaptic innervations made by these axon terminals were either asymmetric, containing spherical vesicles, or symmetric, containing pleomorphic vesicles, with typical differences in postsynaptic density, respectively.Unmyelinated axons were small, round processes with a predominantly even and regular shape, traversing the neuropil in a straight orientation, often forming bundles in perpendicular planes.Astrocytic processes were typically of irregular morphology, forming contours that filled the empty space around neuronal elements.

Physiology-
The methods for single unit recordings from the cognitively-engaged macaque dlPFC, coupled with iontophoresis of drug locally onto the neuron being recorded, are summarized in eFigure 7. Two rhesus monkeys (Macaca mulatta; monkey T, 24 years old, female; monkey C, 15 years old, male) were trained on the ODR task (Fig. 4A), where the monkey fixates on a central point for 0.5sec to initiate a trial; a cue then comes on for 0.5sec in one of eight locations, randomly selected by the computer.The monkey must remember the spatial location over a delay period of 2.5sec while maintaining fixation centrally.At the end of the delay, the fixation point is extinguished, and the monkey can move its eyes to the remembered location within 1sec for juice reward.The monkey can then fixate on the central spot to initiate a new trial.Thus, the contents of working memory must be constantly updated on each trial.Iontophoretic, single unit recordings were made from the dlPFC subregion near the caudal principal sulcus, anterior to the frontal eye fields.This subregion of the dlPFC is known to be essential for the accurate performance of this task, as temporary lesions to this area markedly impair performance [13][14][15][16] .The iontophoretic electrode contains a central carbon fiber for neuronal recording, surrounded by 7 glass micropipettes for ejection of drug solutions using minute amounts of electric current (+10-60nA).Note that drugs must contain an electric charge to be amenable to this technique.
Each compound was dissolved at 0.01 M concentration in either sterile water or DMSO.Neural signals were digitized (15kHz; micro 1401, Cambridge Electronics Design, Cambridge, UK) and acquired using the Spike2 software (CED, Cambridge, UK).
Only Delay cells were selected for subsequent drug testing.Delay cells show persistent firing across the delay period if a cue had been presented at its "preferred direction", e.g.180º, but not other, "nonpreferred" directions.This spatially tuned, persistent firing generates a mental representation of visual space that is foundational to visuospatial working memory 14 .For the sake of space, figures only show the neuronal firing for a neuron's preferred direction and just one nonpreferred direction in the opposing direction.The spatial tuning of a Delay cell was analyzed using a d' measure of spatial selectivity during the delay epoch [d' = meanprefmeannonpref/√ (sd 2 pref + sd 2 nonpref)].Delay cells often fire to the cue and to the response as well as across the delay epoch, and their persistent firing is coincident with gamma bursts in local field potentials (LFPs) that average the electrical activity of a larger expanse of neurons 17,18 .However, as iontophoresis ejects a very small amount of drug in a very limited area, LFPs are usually only altered by very large doses of lipophilic compounds.Please note that there are other types of task-related neurons in the primate dlPFC, e.g.Fixation, Cue, Response and Response Feedback cells, but these have very different molecular dependencies than Delay cells, e.g.relying more on AMPAR than NMDAR-GluN2B neurotransmission.Given the time-intensive nature of this research, the current study focused on Delay cells due to their importance to working memory 13 , their localization in dlPFC layer III 13 and their known dependency on NMDAR-GluN2B (GRIN2B) neurotransmission 19 .Statistical analyses utilized repeated measures ANOVA with two-tailed significance.

Cognitive behavior-
Rhesus monkeys (n=6, 10-22 years, 5 male, 1 female) were pretrained on the delayed response task, a test of visuospatial working memory performed with a manual response in a Wisconsin General Test Apparatus.Monkeys test for highly palatable food rewards, limiting the need for food regulation.This version of the task allows free movement that is more appropriate for studies of the effects of mild stress on cognitive performance.Monkeys were tested by an experimenter who was highly familiar with the normative behavior of each animal, but unaware of drug treatment conditions.Animals were rated for changes in sedation/agitation and aggression using 9 point rating scales.
For each monkey, a dose of the anxiogenic pharmacological stressor, FG7142 (0.5-1.0 mg/kg, i.m. 30 min.before testing), was found that reliably impaired accuracy of performance, but was mild enough that the monkey would work for the full 30 trials that made up a daily test session.Please note that FG7142 creates a classic stress response in humans 20 , monkeys 21,22 and rodents 23 , including cortisol/corticosterone release 20,22,24 and increased NE release in the PFC e.g. 25 , and discriminates similar to other stressors 23 .However, unlike many environmental stressors, the dose can be titrated to find a humane paradigm, where the amount that impairs accuracy, but retains motivation to perform the task for all 30 trials, and allows direct comparison to previous research using this compound.A dose of the β1-AR antagonist, betaxolol (between 0.001-0.1 mg/kg, p.o., one hour before testing) and of nimodipine (0.1-1.0 mg/kg, p.o., one hour before testing) was identified for each animal that did not improve performance on its own, in order that subsequent challenge of the stress response could not be explained by simple additive effects of the two treatments.Following these initial dose-finding experiments, the effects of systemic pretreatment (p.o., one hour before testing) with the β1-AR antagonist, betaxolol, or the LTCC blocker nimodipine were used to challenge the effects of FG7142 or vehicle (veh) i.m. 30 minutes before testing on spatial working memory performance.Statistical analyses employed repeated-measures 2-ANOVA-R with Tukey's multiple comparisons test for post-hoc comparisons with two-tailed significance.2), and GRIN2B (encoding the NMDA receptors with a GluN2B subunits that flux high levels of calcium).The superficial layer CUX2-expressing cell groups are shown in red, pink and purple; the PLD5 expressing cells in light blue; other excitatory neuron subgroups in dark blue.Population sizes are represented in the key, expressed as number of nuclei out of 20,000.The CUX2-expressing cells have higher CALB1 than other excitatory cells, consistent with their expressing high levels of calcium-related signaling proteins, where CACNA1C (Cav1.2) and GRIN2B (NMDAR-GluN2B) would flux high levels of calcium into the neuron.As shown in the main text, this would reduce neuronal firing when calcium opens SK potassium channels, which are also highly expressed in these neurons (KCNN3).C. Same as (B) but in macaque dlPFC.Note that there is a very group of excitatory cells in both species, highlit in light blue, that express the highest levels of CACNA1C that express PLD5 but not CUX2, and have less CALB1.shows co-expression of β1-AR and calbindin in layer III pyramidal cells in macaque dlPFC; biotin-streptavidin was used to amplify the β1-AR signal.Note that ADRB2 was undetectable in CALB1-expressing pyramidal cells and thus this receptor was not examined here.(Q-T) MLIF shows co-expression of SK3 and calbindin in layer III pyramidal cells in macaque dlPFC.Note that calbindin expression is much higher in inhibitory interneurons than pyramidal cells, (e.g.F, R).Where present, blue labeling is the Hoechst nuclear counter stain.(B-C) Cav1.2 immunolabeling (teal arrowheads) can also be found on the plasma membrane of dendritic shafts with a pyramidal cell-like morphology, as well as on microtubules within the dendrite.In B, the dendrite is cross-sectioned, and the labeling on microtubules thus appear as puncta, while C shows a longitudinal view of a dendrite, where the labeling on microtubules can be seen in parallel bundles, likely undergoing intracellular trafficking.In C, the pyramidal cell dendrite receives a symmetric Type II synapse from a putative GABAergic interneuron (pseudocolored purple), consistent with pyramidal cell dendrites receiving inhibitory synapses.Labeling of astrocytic processes can also be seen (B,C).In contrast to dendrites and spines, there was little or no labeling of axons.As, astrocyte; Ax, axon; Mit, mitochondria; Den, dendrite.Scale bars, 200nm.eFigure 6 eFig.6-SK3 channels are expressed on the plasma membrane of macaque layer III dlPFC spines and dendrites, and in astrocytes, but not in axons.SK3 channels (orange arrowheads) are expressed on the plasma membrane of spines and dendrites (pseudocolored yellow), and in astrocytes (pseudocolored green), but not in axons.SK3 immunolabeling is seen in spines subjacent to the asymmetric post-synaptic density (A, C), and in extrasynaptic locations (B), often near the calcium-storing spine apparatus (A, B, pseudocolored pink).Although it is not technically feasible to test whether Cav1.2induced calcium signaling in dlPFC spines drives internal calcium release from the SER to open SK3 channels, the proximity of both Cav1.2 and SK3 channels to the SER in spines would support this hypothesis, with sufficient calcium release into the small spine cytosol to open large numbers of SK channels, based on data from other neurons showing that SK channels opening is detectable at ~0.2M of calcium 26 , which is within the range of calcium concentrations when released from the SER (~0.5-4M) 27,28 .SK3 labeling was also seen in pyramidal-like dendritic shafts along the plasma membrane and trafficking on microtubules (D-G, where D-E are longitudinal, and F-G cross-sections of dendrites), but showed limited axonal labeling.In G, the dendrite receives a symmetric Type II synapse from a putative GABAergic interneuron axon terminal (pseudo-colored purple).SK3 labeling is also seen in astrocytic processes (C, F, G; pseudocolored green).Synapses are between arrows.Ax, axon; As, astroglial; Mit, mitochondria; Sp, dendritic spine; Den, dendrite; End, endosome.Scale bars, 200 nm.A cue then appears for 0.5sec, followed by a delay period in which the monkey must remember the location of the cue for 2.5sec.At the end of the delay, the fixation point disappears and the monkey makes an eye movement to the remembered location for juice reward.The monkey can then fixate to initiate a new trial, where the cue appears randomly at one of the eight locations, and the contents of working memory must be updated with the new location.(B) The dlPFC recording site (PS=principal sulcus; AS=arcuate sulcus).(C) The iontophoretic electrode: a carbon fiber and surrounding glass micropipettes for drug delivery.(D) An example of a Delay cell, with sustained delay-related firing for its preferred direction, but not for nonpreferred directions.The degree of spatial tuning (firing for preferred vs. nonpreferred locations during the delay period) is quantified by signal detection using d'.eFigure 8 eFig.8-Estimate of drug concentration from iontophoresis at 20nA at 100 m from electrode tip.Iontophoresis delivers a very small amount of drug 29 , insufficient to alter behavior, influencing a small region of cortex.The effects of iontophoresis are very localized, but the actual degree of spread is highly dependent on a large number of both known and unknown variables, including the electric charge of the compound, its lipophilicity, and the microenvironment of the electrode (e.g.degree of nearby myelin).Some attempts at estimates ex vivo have shown a diameter of about 30 μm, similar to a cortical minicolumn 30 .A rare in vivo experiment by Gerhardt and Palmer 29 performed an iontophoresis experiment where they locally applied dopamine from the iontophoretic electrode, and measured the dopamine levels 100μm away using electrochemical detection.The barrel concentration of dopamine was 0.5M, and a 20nA current produced an increase of ~17μM from baseline after 40 sec.The concentrations of drug used in the current study were lower (0.005-0.01M); thus, based on this previous work we would estimate that a 20nA current would produce a concentration of drug about 0.17-0.34μMat 100μm away from the electrode tip.However, the actual concentration of any specific compound cannot be known..0071;Tukey's multiple comparisons: preferred direction: control vs S-Bay@25nA, p=0.3409; control vs S-Bay@75nA, p=0.0009;S-Bay@25nA vs S-Bay@75nA, p=0.0316; nonpreferred direction, control vs S-Bay@25nA, p=0.9884; control vs S-Bay@75nA, p=0.7846;S-Bay@25nA vs S-Bay@75nA, p=0.8664)..2805;Tukey's multiple comparisons: preferred direction: control vs dil@5nA, p=0.0294; control vs dil@30nA, p=0.9411; dil@5nA vs dil@30nA, p=0.0145; dil@30nA vs recovery, p=0.0719; non-preferred direction: control vs dil@5nA, p=0.9653; control vs dil@30nA, p=0.6205; dil@5nA vs dil@30nA, p=0.3905; dil@30nA vs recovery, p=0.619)Bottom right: Iontophoresis of diltiazem at 5nA increased the spatial tuning by increasing d', while iontophoresis of diltiazem at 30nA decreased the spatial tuning by decreasing d'.(B) An example neuron (the same cell as Fig. 5B, left panel) is shown where iontophoresis of the β1-AR antagonist, betaxolol, markedly enhanced delay firing for the neuron's preferred direction, and the firing was reduced to the control level when betaxolol application was terminated (recovery).Top: Raster and histogram data shown for the preferred and nonpreferred directions.Bottom left: the mean±SEM firing rate of this cell during the Delay period of the task.Betaxolol significantly enhanced the delay-related firing for the preferred direction (repeated measures two-way ANOVA, Fdirectionxdrug(2, 38) = 4.365, p=0.0197;Fdirection(1, 38) = 99.55,p<0.0001;Fdrug(2, 38) = 10.05,p=0.0003;Tukey's multiple comparisons: preferred direction: control vs beta@25nA, p=0.0005; beta@25nA vs recovery, p=0.0004; control vs recovery, p=0.9634; nonpreferred direction, control vs beta@25nA, p=0.1746; beta@25nA vs recovery, p=0.9979; control vs recovery, p=0.3042).Bottom right: Iontophoresis of betaxolol increased the spatial tuning by increasing d'.The mean±SEM firing rate of this cell during the Delay period of the task.Co-application of ZD7288 with S-Bay prevented the delay firing reduction caused by S-Bay (repeated measures two-way ANOVA, Fdirectionxdrug(2, 34) = 0.8820, p=0.4232;Fdirection(1, 34) = 10.39,p=0.0028;Fdrug(2, 34) = 14.68, p<0.0001;Tukey's multiple comparisons: preferred direction: control vs ZD+S-Bay, p=0.1654; control vs S-Bay, p=0.0001;ZD+S-Bay vs S-Bay, p=0.0065; non-preferred direction: control vs ZD+S-Bay, p=0.5167; control vs S-Bay, p=0.0194;ZD+S-Bay vs S-Bay, p=0.1666).Bottom right.Co-application of ZD7288 with S-Bay prevented the spatial tuning reduction caused by S-Bay.

eFigure 1 eFigure 1 -
Schematic of methods used to perform single nucleus RNAseq from human dlPFC and a 3D portrayal of CALB1, CACNA1C and GRIN2B expression in the excitatory neurons of human and macaque dlPFC.A. Excitatory neurons were identified by expression of the canonical markers SLC17A7 and VGLUT1.B. Transcriptomic analyses of human dlPFC pyramidal cells, showing expression levels of CALB1 (encoding the calcium binding protein, calbindin), CACNA1C (encoding the LTCC Cav1.

eFigure 3 eFig. 3 - 4 eFig. 4 -
Additional examples of immunoEM ultrastructural localization of Cav1.2 channels, SK3 channels and 1-AR in layer III macaque dlPFC.Cav1.2 (A, teal arrowheads), β1-AR (B, red arrowheads) and SK3 channels (C, orange arrowheads) can all be seen in dendritic spines receiving asymmetric (presumed glutamatergic) synapses, often localized on the plasma membrane near the calcium-storing/releasing smooth endoplasmic reticulum (SER; termed the spine apparatus when it is elaborated in the dendritic spine).The SER spine apparatus is highlit with pink pseudocoloring.Note that the Cav1.2 labeling can be seen near the post-synaptic density (PSD) and near the SER.Spines (Sp) are pseudocolored yellow; axon terminals (Ax) are pseudocolored blue; Mit=mitochondrion.Scale bars: 200nm.eFigure Cav1.2 LTCCs and β1-ARs are expressed in dendritic spines near the SER spine apparatus in layer III macaque dlPFC.(A-E) Cav1.2 immunolabeling (indicated by teal arrowheads) is prominently expressed in dendritic spines (pseudocolored yellow), frequently near the calcium-storing smooth endoplasmic reticulum (SER) spine apparatus (pseudocolored pink).Cav1.2 are seen on the plasma membrane at perisynaptic locations (A, B, D), and flanking the excitatory synapse, as well as within the post-synaptic density (C, D).In A1, a dendritic spine emanaties from a dendritic shaft of a pyramidal cell in dlPFC layer III neuropil, with Cav1.2 immunolabeling observed perisynaptically near the glutamatergic-like synapse (higher magnification image of same dendritic spine profile in A2).Cav1.2 immunolabeling can also be seen amongst the membranes of the SER spine apparatus (E).Astrocytic labeling (pseudocolored in green) for Cav1.2 can also be seen in B. (F-G) Additional examples of β1-AR immunolabeling (indicated by red arrowheads) on the plasma membrane of a layer III dendritic spine, near the synapse and the SER spine apparatus.β1-AR immunolabeling is also observed on the astrocytic glial leaflets (pseudocolored in green) near the dendritic spine.All dendritic spines receive axospinous Type I asymmetric glutamatergic-like synapses, which are between arrows.As, astrocyte; Ax, axon; Mit, mitochondria; Sp, dendritic spine; Den, dendrite.Scale bars, 200 nm.

eFigure 5 eFig. 5 -
Cav1.2 LTCCs and β1-ARs are expressed in pyramidal cell-like dendritic shafts in layer III macaque dlPFC.(A) An example of β1-AR immunolabeling (red arrowheads) on the plasma membrane of a pyramidal-like dendrite (pseudocolored yellow).The β1-AR labeling also appears to cluster along microtubules, suggestive of receptor trafficking within the dendrite.

eFigure 7 eFig. 7 -
Schematic of methods used for single unit recording from macaque dlPFC during a spatial working memory task, with iontophoretic delivery of drug to alter the local neurochemical environment.(A) Schematic of the ODR spatial working memory task, where the macaque initiates a trial by fixating on a central spot.

eFig. 9 - 10 eFig. 10 -
Cav1.3 LTCCs have a more generalized expression pattern in macaque layer III dlPFC, including axonal labeling.(A-C) Cav1.3 immunolabeling (dark teal arrowheads) can be seen in layer III dendritic spines, e.g. on the extrasynaptic plasma membrane (A, B) and under the post-synaptic density (C).All dendritic spines receive axospinous Type I asymmetric glutamatergic-like synapses.(D) Cav1.3 immunolabeling can be seen in layer III dendritic shafts with a pyramidal-like phenotype.This example of a longitudinal profile shows Cav1.3 immunolabeling along microtubules, likely trafficking within the neuron.The localization in dendrites and spines is similar to Cav1.2.(E-F) In contrast to Cav1.2, Cav1.3 immunolabeling is prominent in axons, including axon terminals (E), and within myelinated axonal segments (F).(G) Cav1.3 immunolabeling is also seen in astrocytic leaflets ensheathing axospinous, asymmetric synapses.Panel G shows an example of Cav1.3 immunolabeling in a likely astrocytic leaflet (green) near a glutamate-like synapse on a dendritic spine.The localization of Cav1.3 (encoded by CACNA1D) on presynaptic axon terminals is consistent with gain-of-function mutations in CACNA1D, but not CACNA1C, being linked to increased risk of seizure disorders 31 , as increased Cav1.3 on axon terminals would increase glutamate release.Synapses are between arrows.Ax, axon; As, astroglial; Mit, mitochondria; Sp, dendritic spine; Den, dendrite.Scale bars, 200 nm.eFigure Iontophoresis of the LTCC agonist, S-Bay K 6844, or the β1-AR agonist, xamoterol, markedly reduced delay-related firing and spatial tuning of dlPFC Delay cells.(A) An example neuron (the same cell as Fig. 4E, left panel) is shown where iontophoresis of the L-type calcium channel (LTCC) activator, S-Bay K 6844, markedly reduced delay firing for the neuron's preferred direction.Top: Raster and histogram data shown for the preferred and nonpreferred directions.Bottom left: the mean±SEM firing rate of this cell during the Delay period of the task.S-Bay significantly reduced the delay-related firing for the preferred direction (repeated measures two-way ANOVA, Fdirectionxdrug(2, 35) = 2.946, p=0.0657;Fdirection(1, 35) = 30.44,p<0.0001;Fdrug(2, 35) = 5.710, p=0 Bottom right: iontophoresis of S-Bay decreased the spatial tuning by decreasing d'.(B) An example neuron (the same cell as Fig. 5A left panel) showing that iontophoresis of the β1-AR agonist, xamoterol, markedly reduces delay firing for the neuron's preferred direction, and the firing was partially restored when xamoterol application was terminated (recovery).Top: Raster and histogram data shown for the preferred and nonpreferred directions.Bottom left: the mean±SEM firing rate of this cell during the Delay period of the task.Xamoterol significantly reduced the delay-related firing for the preferred direction (repeated measures two-way ANOVA, Fdirectionxdrug(2, 35) = 5.854, p=0.0064;Fdirection(1, 35) = 14.72, p=0.0005;Fdrug(2, 35) = 2.093, p=0.1385;Tukey's multiple comparisons: preferred direction: control vs xamo@40nA, p=0.0055; xamo@40nA vs recovery, p=0.002; control vs recovery, p=0.6591; nonpreferred direction, control vs xamo@40nA, p=0.9855; xamo@40nA vs recovery, p=0.446; control vs recovery, p=0.4342).Bottom right: Iontophoresis of xamoterol decreased the spatial tuning by decreasing d'.

eFigure 11 eFig. 11 -
Iontophoresis of the LTCC antagonist, diltiazem, produced an inverted-U dose response, while the β1-AR antagonist, betaxolol, markedly enhanced delay-related firing and spatial tuning of dlPFC Delay cells.(A) An example neuron (the same cell as Fig. 4F, left panel) is shown where iontophoresis of the LTCC blocker, diltiazem, produced an inverted U effect on the delay firing.Low dose of diltiazem markedly enhanced delay firing for the neuron's preferred direction, while higher dose of diltiazem reduced the firing.Top: Raster and histogram data shown for the preferred and nonpreferred directions.Bottom left: the mean±SEM firing rate of this cell during the Delay period of the task.Iontophoresis of diltiazem, produced an inverted U effect on the delay firing (repeated measures two-way ANOVA, Fdirectionxdrug((3, 52) = 4.579, p=0.0064;Fdirection(1, 52) = 14.11, p=0.0004;Fdrug((3, 52) = 1.312, p=0

eFigure 13 eFig. 13 -
Iontophoresis of the SK channel blocker, NS8593, markedly enhanced the delayrelated firing and spatial tuning of dlPFC Delay cells, while blockade of either SK or HCN channels blocked the effects of LTCC opening.(A) An example neuron (the same cell as Fig. 6A, left panel) is shown where iontophoresis of the SK channel blocker, NS8593, markedly enhanced delay firing for the neuron's preferred direction, and the firing was reduced to control levels when NS8593 application was terminated (recovery).Top: Raster and histogram data shown for the preferred and nonpreferred directions.Bottom left.The mean±SEM firing rate of this cell during the Delay period of the task.NS8593 significantly enhanced the delay-related firing for the preferred direction (repeated measures two-way ANOVA, Fdirectionxdrug(2, 40) = 1.773, p=0.1829;Fdirection(1, 40) = 32.14, p<0.0001;Fdrug(2, 40) = 9.993, p=0.0003;Tukey's multiple comparisons: preferred direction: control vs NS@25nA, p<0.0001;NS@25nA vs recovery, p=0.0668; control vs recovery, p=0.0855; nonpreferred direction, control vs NS @25nA, p=0.1959;NS @25nA vs recovery, p=0.2306; control vs recovery, p=0.978).Bottom right.Iontophoresis of NS8593 increased the spatial tuning as indicated by an increase in d'.(B) An example neuron (the same cell as Fig. 6B, left panel) is shown where iontophoresis of the SK blocker, NS8593, prevented the reduction in firing caused by the LTCC activator, S-Bay, when these two agents were co-applied.Top: Raster and histogram data shown for the preferred and nonpreferred directions.Bottom left.The mean±SEM firing rate of this cell during the Delay period of the task.Co-application of NS8593 with S-Bay prevented the delay firing reduction caused by S-Bay (repeated measures two-way ANOVA, Fdirectionxdrug(2, 39) = 1.091, p=0.3459;Fdirection(1, 39) = 13.81,p=0.0006;Fdrug(2, 39) = 9.020, p=0.0006;Tukey's multiple comparisons: preferred direction: control vs NS+S-Bay, p=0.979; control vs S-Bay, p=0.0082;NS+S-Bay vs S-Bay, p=0.0021; non-preferred direction: control vs NS+S-Bay, p=0.9111; control vs S-Bay, p=0.4288;NS+S-Bay vs S-Bay, p=0.1456).Bottom right.Co-application of NS8593 with S-Bay prevented the spatial tuning reduction caused by S-Bay.(C) An example neuron (the same cell as Fig. 6C, left panel) is shown where iontophoresis of the HCN blocker, ZD7288, prevented the reduction in firing caused by the LTCC activator, S-Bay, when these two agents were co-applied.Top: Raster and histogram data shown for the preferred and nonpreferred directions.Bottom left.

eFigure 14 eFig. 14 - 15 eFig. 15 -
Behavioral data showing performance of the spatial delayed response test of spatial working memory under control vs. stress conditions, with pretreatment with vehicle or test drug.A dose of the β1-AR antagonist, betaxolol (between 0.001-0.1 mg/kg, p.o., one hour before testing) (A) or of the LTCC blocker, nimodipine (0.1-1.0 mg/kg, p.o., (B) one hour before testing) was identified for each animal that did not improve performance on its own, to challenge the stress response without concerns of additive effects.The nimodipine dose was similar to that used in humans.Pretreatment with systemic administration of a β1-AR antagonist or an LTCC antagonist protected working memory performance from the detrimental effects of a mild stressor.Vehicle (Veh) pretreatment in combination with the pharmacological stressor, FG7142 (Str) significantly impaired performance compared to vehicle+vehicle control performance, while pretreatment with the β1-AR antagonist betaxolol (Beta) or the LTCC antagonist, nimodipine (Nim), prevented working memory deficits in FG7142-treated animals.Results represent mean percent correct on the delayed response task with all individual data points shown (n=6).*** p<0.001 compared to veh+veh; ⌘ p<0.001 compared to veh+stress.eFigure RORB is widely expressed in dlPFC pyramidal cells, including in the CUX2A and CUX2C subgroups.(A) Violin plot of RORB expression in human dlPFC pyramidal cells.(B) Violin plot of RORB expression in macaque dlPFC pyramidal cells.The CUX2A cell group is Cluster 0, the CUX2B cell group is Cluster 1, the CUX2C cell group is Cluster 3, and the PLD5A cell group is Cluster 11. (C) Widespread expression of RORB in macaque dlPFC across layers, including high expression in many layer III pyramidal cells (D).These protein results for RORB are corroborated by recent RNA-seq results in macaque 32 and human 33 dlPFC layer III pyramidal cells using laser-capture microdissection methods, providing laminar-and cell-type specific resolution.