(A) Real-time RT-PCR quantification of Grin2b, Grin2a, and Grin1 mRNAs in CK-Setdb1 and wildtype hippocampus, as indicated. N=6–8/group, *p<0.05, Mann Whitney test. (B) Western blot analysis of Grin2b and Grin2a proteins from hippocampus (Hipp) and prefrontal cortex (PFC) of adult CK-Setdb1 and wildtype littermates. Relative protein levels were normalized to GAPDH and histone H3, respectively. N=7–15/group. **p<0.01, Mann Whitney test. All data shown as mean ± S.E.M.

(A) Binding patterns of Setdb1 methyltransferase (black dots) and, for comparison, KRAB-associated protein 1 (KAP-1) (white dots) in adult mouse forebrain (wildtype) within 60Kb from Grin2b transcription start site (TSS) as indicated. Data expressed as chip-to-input ratio (mean ± S.E.M), N=3 chip experiments/antibody. Notice that Setdb1 occupancy is highest at the target site in intron III (marked by red line), corresponding to same target site shown in Fig. 3 and Table 1. Notice further that KAP-1 occupancy is highest within first 2–3Kb from Grin2b TSS. Notice further that ChIP-seq H3K4me3 markings (shown in gray) at Grin2b are almost exclusively limited to this site. This portion of Grin2b is part of its most proximal Hind III fragment (marked by blue box) and harbors a CpG island as indicated. (B) Chromatin conformation capture (3C) profiles from adult mouse forebrain (wildtype) for proximal 35Kb of Grin2b, colored bars each represent a specific HindIII restriction fragment. Position and orientation of 3C primers 1–12 as indicated. Each 3C profile (Ba-c) is anchored on specific restriction fragment/primer as indicated by colored bar in interaction map. Y-axis shows the normalized pixel intensities (from agarose gel) for specific PCR product from anchoring primer and each of the remaining 11 primers as indicated by color code. Data shown as mean ± S.E.M from two 3C libraries processed in triplicate. Notice in (Ba) and (Bb) that there are comparatively high levels of interaction between 5Kb of 5′ end of Grin2b (marked in blue) and fragments 30Kb further downstream from the TSS. Setdb1 target site shown as red line. Notice further (Bc) that this interaction is not detected with anchoring fragment immediately upstream of TSS. (C) Model for Grin2b chromatin loop, based on data shown in (A) and (B). The loop tethers the 5′ end of the gene to intronic sequences more than 30Kb downstream from the TSS, including the Setdb1-target site. Notice while the majority of Setdb1 is bound to the 3′ part of the loop, a smaller fraction is found at the 5′ part (see also A).

(A) Current-voltage (I–V) relationship of representative nucleus accumbens/ventral striatum neurons from wildtype and CK-Setdb1 mice recorded in whole-cell patch-clamp in a slice of tissue. Notice the strong rectification in the hyperpolarizing range, the large action potential overshoot, presence of a ramp, and hyperpolarized resting potential (~ −87 mV) that identify these neurons as MSN. (B) Representative NMDA-EPSPs from wildtype and transgenic neurons, recorded at resting membrane potential before (10 μM CNQX) and during ifenprodil perfusion (10 μM CNQX + 100 μM ifenprodil). Notice that NMDA-EPSP amplitude remained unchanged in CK-Setdb1 neuron after ifenprodil exposure. (C) Mean amplitudes of NMDA-EPSPs evoked every 20 sec before (CNQX) and during ifenprodil exposure (CNQX + ifenprodil) in slices from wildtype (open circles) and CK-Setdb1 mice (filled circles). EPSPs were evoked in an Mg²⁺-free ACSF and their amplitudes were expressed as percent of baseline (CNQX). Data (mean ± S.E.M.) are shown for 4 transgenic and 5 wildtype neurons (animals). A 5^th CK-Setdb1 neuron was removed as an outlier (Grubbs test). Two way ANOVA, genotype F_1,6 = 105.2, p < 0.001. * p < 0.05, Newman Keuls posthoc. (D) Same recording as in (C), but now from CK-Setdb1 mice treated with Setdb1-siRNA (filled triangles). N =4 neurons. Notice that in (C), wildtype but not CK-Setdb1 neurons show reduced NMDA-EPSP from baseline (dotted line) in presence of ifenprodil. Notice further that in (D), siRNA treatment restores NMDA-EPSP sensitivity to ifenprodil in CK-Setdb1 neurons. (E) Representative examples of (left) anti-myc immunohistochemistry and (right) anti-Setdb1 immunoblots from ventral striatum of untreated and Setdb1-siRNA treated CK-Setdb1 mice. Bar, 100 μm. Notice robust decrease of myc and Setdb1 immunoreactivity after siRNA treatment. Data in (D, E) from brains harvested 60 hours after siRNA delivery into striatum.

(A) Sucrose consumption in mice treated with Ro25-6981 or saline once daily for 10 days. First dose was given at day 0. N=10/treatment. Two-way ANOVA, day F_9,153=6.38, p<0.01; treatment F_1,17=4.28, p<0.05; *p<0.05, **p<0.01, Newman-Keuls post hoc. (B) Representative hippocampal immunoblots showing NR2B/Grin2b and histone H3 (pan) immunoreactivity in animals exposed to Grin2b-siRNA (siR) or vehicle (V).

(A) Setdb1 expression in postnatal day (P) 7 and adult wildtype cerebral cortex. (B) CK-Setdb1 transgene construct. (C) Anti-myc immunoreactivity in brain sections of 8 week old CK-Setdb1 and wildtype mice in cerebral cortex (Cx), striatum (Str), hippocampus (Hipp) and cerebellar cortex (Cbc). Scale bars: Cx, Str, 100 μm; Hipp 200 μm, Cbc 150 μm. (D) Anti-myc (red) immunofluorescence in neuronal nuclei (anti-NeuN, green) (arrows), but not in non-neuronal nucleus (dotted line); blue, DAPI. Scale bar, 10 μm. (E) Immunoblots show (top) 180 kDa anti-myc signal in prefrontal cortex (PFC), Str and Hipp of CK-Setdb1 (Set), but not wildtype (w) mice; (bottom) increased 180 kDa anti-Setdb1 signal with two different antibodies, Ab-1 and Ab-2 (see Methods) in transgenic cerebral cortex (Ab-1) and hippocampus (Ab-2), each lane loaded with 20 μg protein. (F) Representative slot-blot shows anti-H3K9me3 and IgG control chromatin immunoprecipitates (ChIP) hybridized with major satellite DNA probe. (G) Representative blot shows mycSetdb1 immunoactivity in “Input” and anti-myc (myc) ChIP from CK-Setdb1 (Set), but not wildtype (w) forebrains. Protein extracted from Cx and anti-histone H3 (H3) ChIP material were processed as positive and negative controls, respectively.

(A,B) Setdb1 target sites at Grin2b (mouse) and GRIN2B (human). (A) (Top) Distribution of scores calculated with MAT software in 1.3 Mb region of mouse chromosome 6 enclosing the Grin2b gene. Scores are calculated over a 600 bp interval and assigned to the tiling probe at the center of the interval. Scores are plotted as a custom track in the UCSC genome browser (http://genome.ucsc.edu/). Significance threshold is 3.66 for p < 1e-5. (Bottom) Sequence alignment between mouse and human GRIN2B gene loci at Setdb1 target sites. Conserved sequences are indicated with *. (B) Schematic overview shows Setdb1 target sites in 5′-untranslated regions from mouse (solid arrow) and human (dashed arrow) Grin2b/GRIN2B loci. (C) Representative blot shows mycSetdb1 immunoreactivity in human glioma (U87-MG) culture transfected with mycSetdb1 cDNA (Set+), but not control (C). Histone H3 was used as loading control. (D) Representative immunoblots show Setdb1 and GRIN2B immunoreactivity in Set+ and control U87-MG cultures.

(A) Representative images show co-expression of myc (red) immunoreactivity and GFP (green) in pyramidal neuron layer of the hippocampal CA1 field from CK-Setdb1/CK-H2BeGFP mice. Neither transgene is expressed in stratum oriens (SO) that harbors primarily GABAergic neurons; blue = DAPI counterstain. Scale bars 40 μm. (B) Example of an acutely dissociated GFP positive neuron from hippocampus of CK-Setdb1/CK-H2BeGFP mice at postnatal day 18. Scale bar 10 μm. (C) (left) NMDA currents evoked by NMDA from 1 μM to 3 mM for different hippocampal neurons. Horizontal bar represents timing of drug exposure. (right) Dose response plot of peak NMDA currents from CK-Setdb1 (CK-Setdb1^+/0/CK-H2BeGFP^+/0) (filled circle) and controls (CK-Setdb1^0/0/CK-H2BeGFP^+/0) (open circle), respectively N=8–10/genotype. Notice that dose-response curves are indistinguishable between CK-Setdb1 and control mice (D) (Left) Representative individual currents of 3 hippocampal neurons in response to 500 μM NMDA. Red lines represent single exponential fits of current desensitization; values (sec) are indicated above the traces. (Right) Scatter plot shows individual NMDA current desensitization rate (open circle) and averaged rate (black dots) for GFP positive neurons from CK-Setdb1 and control mice. Notice significantly faster desensitization in GFP positive CK-Setdb1 neurons in comparison to control neurons. N=8-19/group; *p<0.05, one-tailed Mann Whitney test. Data shown as mean ± S.E.M.

(A) Sucrose consumption (transgenic line #1), expressed as % of total fluid intake, in CK-Setdb1 and wildtype mice over the course of 10 days; N=10–11/group. Two-way ANOVA, day × genotype, F_9,162 =2.16, p < 0.05, *p<0.05, Newman-Keuls post hoc. (B) Accelerated extinction of learned helplessness (transgenic line #1 and #2) in 2-10 month old male CK-Setdb1 mice as compared to wildtype littermates. (Left) Time line shows shock training and escape behavior testing. (Right) Bar graph shows averaged escape latencies 24 hr and 7 days after the end of training. Notice significant reduction in escape latency in CK-Setdb1 mice on day 7; N = 8–9 per group. *p ≤ 0.05, Newman-Keuls post hoc after two-way ANOVA.

(A) Scatter plot shows shock reactivity scores (flinching, vocalization, running and jumping) in CK-Setdb1 and littermate control mice exposed to an escalating current (0.05–0.7 mA, 2 sec each), as indicated. (B) Passive avoidance: bar graph show s latency (sec) to enter dim chamber before (pre), and 24 hr or 7 days after (post) single foot shock. (A, B) N=8–11/genotype. (C) Fear conditioning: (Top) experimental design, including time lines for context and tone. (Bottom) Bar graphs show (Left) context and (Right) tone induced total immobility/freezing in CK-Setdb1 and wildtype littermates over the course of 180 and 20 sec for context and tone, respectively, 24 hr and 7 days after shock or sham exposure, as indicated. Notice no significant differences between transgenic and control mice. All data shown as mean ± S.E.M.