A robust activity marking system for exploring active neuronal ensembles

  1. Andreas T Sørensen
  2. Yonatan A Cooper
  3. Michael V Baratta
  4. Feng-Ju Weng
  5. Yuxiang Zhang
  6. Kartik Ramamoorthi
  7. Robin Fropf
  8. Emily LaVerriere
  9. Jian Xue
  10. Andrew Young
  11. Colleen Schneider
  12. Casper René Gøtzsche
  13. Martin Hemberg
  14. Jerry CP Yin
  15. Steven F Maier
  16. Yingxi Lin  Is a corresponding author
  1. Massachusetts Institute of Technology, United States
  2. Institute for Behavioral Genetics, University of Colorado Boulder, United States
  3. University of Wisconsin-Madison, United States
  4. Wellcome Trust Sanger Institute, United Kingdom
  5. University of Colorado Boulder, United States
7 figures and 2 additional files

Figures

Figure 1 with 4 supplements
Design and characterization of the RAM promoter (PRAM) and AAV-based RAM system in vitro.

Unless indicated otherwise, cultured mouse hippocampal neurons were transfected with luciferase reporter constructs on DIV5 and stimulated with 35 mM KCl for 6 hr on DIV 7 or 8 (See Supplementary file 2 for details). The relative luciferase value is the absolute luciferase value normalized against the internal control. The fold induction is the ratio of the relative luciferase values in stimulated and unstimulated conditions. (a) PRAM has higher fold induction than promoters in which the NRE/AP-1 element is replaced by other elements found enriched in the 11,830 activity-regulated enhancers (See ‘Methods’). CME is not regulated by activity. Each construct contains four enhancer modules (EM) inserted upstream of the FOS minimal promoter (left). The relative luciferase values are shown in Figure 1—figure supplement 2b. n = 5–7 separate experiments per condition, one-way ANOVA, Tukey’s post-hoc test. (b) The robust activity response of PRAM is a result of the combination of the four repeated RAM EM and FOS minimal promoter, which alone are respectively weakly responsive or non-responsive to neuronal activity. n = 5–10 separate experiments per condition, one-way ANOVA, Tukey’s post-hoc test. (c) Comparison of PRAM to various activity-dependent promoters. Promoter size (Kb) is shown in brackets. The relative luciferase values are shown in Figure 1—figure supplement 3a. n = 4–10 separate experiments per condition, Student’s t-test. (d) Comparison of PRAM and ESARE. n = 6–8 separate experiments per condition, Student’s t-test. (e) PRAM can be driven by overexpression of FOS and NPAS4 individually or in combination. n = 9–10 separate experiments per condition, one-way ANOVA, Dunnett’s post-hoc test. (f) Top, schematic outline of Tet-OFF system with PRAM driving tTA expression. Binding of tTA protein to the TRE promoter is prevented by Dox administration (+Dox); withdrawal of Dox (-Dox) allows downstream effector gene transcription. Bottom, schematic diagram of the AAV-RAM construct with critical genetic elements outlined. (g) Comparison of PRAM-d2tTA and PRAM-tTA by assaying pTRE-luciferase activity with and without Dox (+Dox, −Dox). The relative luciferase values are shown in Figure 1—figure supplement 3e. n = 3 separate experiments per condition, Student’s t-test. (h) Representative images of hippocampal neurons grown on a glia monolayer and infected with AAV-RAM-tdTomato (red). Neurons are identified by MAP2 staining (green). Cultures were either left undisturbed (No Stim) or stimulated with bicuculline and 4AP (+Bic/4AP), either with (+Dox) or without (-Dox) doxycycline added. The scale bar is 150 μm and applies to all images. (i) Quantification of h. Percentages of neurons (MAP2+) that are RAM+ (%RAM+) are plotted. n = 3 separate experiments per condition, one-way ANOVA, Tukey’s post-hoc test. All data in ae, g and i are mean ± SEM. *p<0.05, **p<0.01, ***p<0.001.

https://doi.org/10.7554/eLife.13918.003
Figure 1—figure supplement 1
Position-weight matrix of the top-ranking motif (AP-1, TGANTCA) identified by a Weeder de novo motif search.
https://doi.org/10.7554/eLife.13918.004
Figure 1—figure supplement 2
Characterization of enhancers by luciferase assay.

(a) Fold induction with PRAM and PNRE+AP-1 constructs following 6 hr of KCl stimulation. n = 4–10 separate experiments per condition, Student’s t-test, *p<0.05. (b) Relative luciferase activity with PRAM, enriched enhancers E1-E3 and CME constructs following 6 hr of KCl stimulation. Note log10 scale. Baseline activity is similar for all constructs (non-significant difference). For KCl conditions PRAM displays significantly higher activity than CME, E2 and E3 (*p<0.05, ***p<0.001) and E1 displays significant higher activity than CME and E2 (♯p<0.05, ♯♯p<0.01). CME construct expression is not induced by depolarization. n = 5–7 separate experiments per condition, one-way ANOVA, Tukey’s post-hoc test. The fold induction is shown in Figure 1c. (c) Fold induction with constructs containing four RAM enhancer modules attached to the following minimal promoters: CMV: cytomegalovirus; hBG: human beta-globin; Arc: activity-dependent cytoskeleton associated-protein. n = 5–12 separate experiments per condition, one-way ANOVA, Dunnett’s post-hoc test. (d) Fold induction with PRAM constructs containing different numbers of RAM enhancer modules. n = 5–12 separate experiments per condition, one-way ANOVA, Dunnett’s post-hoc test. Data in ad are mean ± SEM. *p<0.05, **p<0.01, ***p<0.001, n.s., non-significant.

https://doi.org/10.7554/eLife.13918.005
Figure 1—figure supplement 3
Characterization of the PRAM promoter in vitro.

(a) Relative luciferase activity of activity-dependent promoters. Note log10 scale. Statistically significant differences between baseline conditions (No Stim) are indicated by * and between KCl conditions by #. All other comparisons are non-significant. n = 4–10 separate experiments per condition, Student’s t-test. The fold induction is shown in Figure 1c. (b) Relative luciferase activity of PRAM following bicuculline (Bic) stimulation and blocking with Nimodipine and/or APV. n = 4 separate experiments per condition, one-way ANOVA, Tukey’s post-hoc test. (c) Relative luciferase activity of PRAM after 6 hr of stimulation with various growth factors and pharmacological agents. n = 4 separate experiments per condition, Student’s t-test. (d) Relative luciferase activity of PRAM in pure glial cultures following 4 hr of KCl stimulation. n = 2 separate experiments per condition. (e) Relative luciferase activity of PRAM-tTA- and PRAM-d2tTA-mediated pTRE-luciferase transcription with and without KCl stimulation, and with and without Dox (+Dox, -Dox). The fold induction is shown in Figure 1g . n = 3 separate experiments per condition, Student’s t-test. Data in ae are mean ± SEM. *p<0.05, **p<0.01, ***p<0.001, n.s., non-significant.

https://doi.org/10.7554/eLife.13918.006
Figure 1—figure supplement 4
Quantification of AAV-RAM-tdTomato infection of neuron-glia co-cultures determined by immunocytochemistry.

(a) Percentage of MAP2+ cells expressing tdTomato (RAM+) at various amounts of virus applied per well. Experiments in b and Figure 1h–i used 0.1 μl/well (red bar). n = 3 separate experiments per condition. (b) Co-labeling of MAP2 or GFAP with tdTomato (RAM+). GFAP did not co-label with tdTomato. N = 3 separate experiments per condition. All data in ab are mean ± SEM.

https://doi.org/10.7554/eLife.13918.007
Figure 2 with 1 supplement
Optimizing the working parameters for the AAV-RAM system in vivo.

(a) Schematic of the hippocampus showing the CA1, CA3 and DG regions with the injection site (DG) highlighted in blue. (bd) Experiments to determine the time required for Dox clearance to allow maximal effector gene expression. (b) Experimental scheme. Animals were co-infected with AAV-RAM-NLS-mKate2 and AAV-Ef1α-EGFP and kept on Dox diet (+Dox) for a minimum of 7 days. The animals were taken off Dox diet (−Dox) 12, 24, or 48 hr before kainic acid treatment to induce seizures and sacrificed 24 hr after this treatment. (c) Percentage of RAM+ cells among total EGFP+ cells in the DG stratum granulosum when seized 12, 24 or 48 hr after Dox removal. n = 3–4 animals per group, one-way ANOVA, Tukey’s post-hoc test. (d) Representative images from the data quantified in c. The scale bar is 150 μm and applies to all images. (eg) Experiments to determine the maturation of effector gene expression. (e) Experimental scheme. Animals were treated similarly to a, and 48 hr following Dox removal (-Dox), animals were seized then sacrificed 3, 6, 12, or 24 hr later. (f) Percentage of RAM+ cells among total EGFP+ cells in the DG stratum granulosum 3, 6, 12, or 24 hr after seizure. RAM+ cells are detected by fluorescence (mKate2, red) or immuno-staining (α-mKate2, cyan). n = 2–4 animals per group, two-way ANOVA, Bonferroni post-hoc test. (g) Representative images from the data quantified in f. Upper row: AAV-mKate2 (red), middle row: mKate2 detected by antibody (α-mKate2, cyan), lower row: EGFP (green). The scale bar is 150 μm and applies to all images. Data in c and f are mean ± SEM. **p<0.01, ***p<0.001.

https://doi.org/10.7554/eLife.13918.008
Figure 2—figure supplement 1
Rapid suppression of RAM expression on returning to Dox diet.

(a) Schematic timeline of the experimental procedure. Animals were co-injected with AAV-RAM-NLS-mKate2 and AAV-Ef1α-EGFP into DG and kept on Dox diet (+Dox) for a minimum of 7 days. Dox was withdrawn (−Dox) for 48 hr then resumed for 0, 24 or 48 hr prior to kainic acid treatment to induce seizures. Animals were sacrificed 24 hr after seizure induction. (b) Schematic drawing of the hippocampus showing the viral injection site. (c) Percentage of RAM+ cells among total EGFP+ cells in the DG with 0, 24 or 48 hr Dox blockade before seizure. n = 3–4 animals per condition, one-way ANOVA, Tukey’s post-hoc test. (d) Representative images from the data quantified in c showing mKate2 (red) and EGFP (green) labeling. The scale bar is 150 μm and applies to all images.

https://doi.org/10.7554/eLife.13918.009
Figure 3 with 4 supplements
The RAM system labels active neuronal ensembles in the DG of the hippocampus.

(a) Schematic timeline of the experimental procedure. Animals were kept on Dox diet (+Dox) from 24 hr before injection of AAV-RAM-NLS-mKate2 and AAV-Ef1α-EGFP into DG to a minimum of 7 days after the injection. Dox was then withdrawn (−Dox) 48 hr before either CFC, exposure to the novel context without shock (Context Only), receiving immediate shocks (Shock Only) or kainic acid treatment to induce seizures. Animals were sacrificed 24 hr later. A control cohort of similarly injected and treated animals was left undisturbed in their home cages (HC) for the entire period before sacrifice. (b) Schematic drawing of the hippocampus with the viral injection site (DG) highlighted in blue. (c) Percentage of RAM+ cells among total EGFP+ cells in the DG stratum granulosum for HC, CFC, Context Only, Shock Only and seized animals. The data for seized animals are replotted from Figure 2c (48 hr group). n = 6–9 animals per condition (with the exception for the Shock Only group, which consisted of 3 animals), one-way ANOVA, Tukey’s post-hoc test. (d) Representative images of the DG region showing mKate2 (red) and EGFP (green) labeling for each of the HC, CFC, Context Only, Shock Only and seizure conditions. The scale bar is 300 μm for the left images and 50 μm for the three right columns of images. These images are enlarged from the areas marked by purple squares. All data in c are mean ± SEM. ***p<0.001.

https://doi.org/10.7554/eLife.13918.010
Figure 3—figure supplement 1
RAM labeling in the hippocampus following sensory experience is prevented by Dox diet.

(a) Schematic timeline of the experimental procedure. Following injection of AAV-RAM-NLS-mKate2 and AAV-Ef1α-EGFP into DG, animals were left on Dox diet (+Dox) throughout. Animals were subjected to CFC and sacrificed 24 hr later. (b) Representative images of the DG from animals given CFC treatment. Left: mKate2 (red). Right: EGFP (green), DAPI (blue) and mKate (red) merged. No RAM+ positive cells were detected. The scale bar is 150 μm and applies to both images. n = 4 animals.

https://doi.org/10.7554/eLife.13918.011
Figure 3—figure supplement 2
RAM labeling of CA3 pyramidal cells following contextual fear conditioning (CFC).

(a) Schematic timeline of the experimental procedure. Animals were kept on Dox diet (+Dox) from 24 hr before injection of AAV-RAM-NLS-mKate2 and AAV-Ef1α-EGFP vectors into CA3 for a minimum of 7 days after the injection. Dox was then withdrawn (−Dox) 48 hr before CFC or kainic acid treatment to induce seizures. Animals were sacrificed 24 hr after CFC or seizure induction. A control cohort of similarly injected and treated animals was left undisturbed in their home cages (HC) for the entire period before sacrifice. (b) Schematic drawing of the hippocampus with the injection site (CA3) highlighted in blue. (c) Representative images of the CA3 region showing mKate2 (red) and EGFP (green) and DAPI (blue) labeling for HC, CFC and seizure conditions. Red arrows indicate RAM and EGFP double-labeled cells in CA3. The scale bar is 75 μm for the upper and middle (small) images and 50 μm for the lower images. The lower images are enlarged from the areas marked by purple squares. (d) Percentage of RAM+ cells among total EGFP+ cells in the CA3 stratum pyramidale for HC, CFC and seizure conditions. n = 3–4 animals per condition, Student’s t-test. All data are mean ± SEM. ***p<0.001.

https://doi.org/10.7554/eLife.13918.012
Figure 3—figure supplement 3
RAM labeling following CFC persists for at least a week.

(a) Schematic timeline of the experimental procedure. Following infection of AAV-RAM-NLS-mKate2 and AAV-Ef1α-EGFP into DG, animals were kept on Dox diet (+Dox) for a minimum of 7 days. Dox was withdrawn (−Dox) 48 hr before CFC. Animals were either sacrificed 24 hr later, or placed back on Dox diet for 1, 2 or 4 weeks before sacrifice. (b) Schematic diagram of the hippocampus showing the viral injection site. (c) Percentage of RAM+ cells among total EGFP+ cells in the DG stratum granulosum for CFC animals sacrificed 24 hr or 1, 2 or 4 weeks after CFC. Data are mean ± SEM. n = 3–4 animals per condition, one-way ANOVA, Tukey’s post-hoc test, *p<0.05, **p<0.01, ***p<0.001, n.s., non-significant. (d) Representative images of mKate2 (red) and EGFP (green) expressing cells 24 hr and 1, 2 and 4 weeks after CFC. The scale bar is 150 μm and applies to all images.

https://doi.org/10.7554/eLife.13918.013
Figure 3—figure supplement 4
Infection with RAM-channelrhodopsin (ChR2) or -archaerhodopsin (ArchT) prior to CFC enables light-activation or silencing, respectively, of repeated action potentials in RAM+ hippocampal DG granule cells.

(a) Schematic timeline of the experimental procedure. Animals were co-injected with AAV-RAM-ChR2:EYFP and AAV-TRE-mCherry or AAV-RAM-ArchT:EGFP and AAV-TRE-mCherry into the DG. mCherry was used for visual location of RAM+ cells during ex vivo electrophysiology. Animals were kept on Dox diet (+Dox) for a minimum of 7 days. Dox was withdrawn (−Dox) 48 hr before CFC, and animals were sacrificed 24 hr after CFC for electrophysiology. (b) Schematic diagram of the experimental procedure. 455 nm blue light for ChR2 activation was applied through an optical lens (40x objective). Only cells in the DG stratum granulosum were used for recording. (c) Photocurrent generated in a representative ChR2-RAM+ cell held at −70 mV in a voltage clamp when illuminated for 1s with blue light. (d) Quantification of the instant maximum (max) and constant photocurrents generated in ChR2-RAM+ cells. n = 9 cells from 3 animals. Data are mean ± SEM. (e) Repeated action potential evoked by 10 ms light pulses applied at 5 Hz to a ChR2-RAM+ cell held at its resting membrane potential, −75 mV, in current clamp mode. (f) Examples of RAM+ cells expressing ChR2-EYFP (green) and mCherry (red) in the granule cells layer and their merged image (together with DAPI, blue) as seen in a 300 μm slice used for electrophysiology. The scale bar is 20 μm and applies to all images. (g) As for b, but with 565 nm green light for ArchT activation. (h) Example of a photocurrent generated in ArchT-RAM+ cells held at −70 mV. (i) Quantification of constant photocurrents generated in ArchT-RAM+ cells. Data are mean ± SEM. n = 10 cells from 3 animals. (j) Light-mediated silencing of action potentials in an RAM-ArchT+ cell depolarized by an 80pA square current pulse from its resting membrane potential of −71 mV.

https://doi.org/10.7554/eLife.13918.014
Contextual memory recall preferentially reactivates cells initially labeled with RAM during memory encoding.

(a) Schematic drawing of the hippocampus with the viral injection site (DG) highlighted in blue. (b) Timeline of the experimental procedure. Animals were injected with AAV-RAM-NLS-mKate2 and kept on a Dox diet (+Dox) for at least 7 days after surgery. Two days after Dox removal (−Dox), animals were exposed to context A (for RAM labeling) and shocked, and then exposed to either context A again or a new context B 24 hr later (for IEG labeling). Animals were sacrificed 1.5 hr after the second context exposure. (c, d) Representative images of DG after A−A (c) and A−B (d) exposure showing DAPI staining (cyan), FOS staining (green), RAM labeling (red), and the merged image. (e, f) As panels c and d, except staining with NPAS4 (white) instead of FOS. (g) Freezing behavior observed during re-exposure to context A or B. n = 4–5 animals per condition, Student’s t-test. (h) Percentage of all DAPI-labeled cells (i.e. all neurons) in the DG stratum granulosum labeled with RAM (red) and FOS (green; n = 9 animals per condition), and the percentage overlap between the RAM and FOS labeled cells following A-A exposure (n = 4 animals) and A-B exposure (n = 5 animals). Student’s t-test. (i) As h, except with NPAS4 (white) instead of FOS. The scale bar is 50 μm for all images. Data in gi are mean ± SEM. *p<0.05, ***p<0.001.

https://doi.org/10.7554/eLife.13918.015
The RAM system labels active neuronal ensembles in the amygdala.

(a) Schematic timeline of the experimental procedure. While on a Dox diet (+Dox), AAV-RAM-NLS-mKate2 and AAV-Ef1α-EGFP vectors were injected into the lateral amygdala (LA, b), basal nucleus (BA, d), or central amygdala (CeA, g). After at least 7 days, Dox was removed (-Dox) for 48 hr, the animals were exposed to tone-fear conditioning (TFC; consisting of Tone and Shock, T+S), Tone only, or Shock only and sacrificed 24 hr later. A control cohort of similarly injected and treated animals was left undisturbed in their home cages (HC) for the entire period before being sacrificed. (b) Schematic drawing of the amygdala with the injection and the quantification site (LA) highlighted in blue. (c) Percentage of RAM+ cells among total EGFP+ cells in LA for HC, T+S, Tone, and Shock animals. n = 3–6 animals per condition, one-way ANOVA, Tukey’s post-hoc test. (d) As panel b, but for the BA region. (e) Percentage of RAM+ cells among total EGFP+ cells in BA for HC, T+S, Tone and Shock animals. n = 3–4 animals per condition, one-way ANOVA, Tukey’s post-hoc test. (f) Representative images of neurons labeled in LA and BA for HC, T+S, Tone and Shock conditions. Neurons labeled by AAV-RAM-mKate2 are red and cells labeled by AAV-Ef1α-EGFP are green. The merged left image is enlarged in the two right images. Red arrows indicate RAM and EGFP double-labeled cells. The scale bar is 300 μm and 50 μm for left and right images, respectively. (g) As panel b, but for the CeA region. (h) Percentage of RAM+ cells among total EGFP+ cells in CeA for HC and T+S animals. n = 3 animals per condition, Student’s t-test. (i) Representative images of neurons labeled in CeA region. The scale bar is 150 μm for all images. All data in c, and h are mean ± SEM. *p<0.05, **p<0.01, ***p<0.001.

https://doi.org/10.7554/eLife.13918.016
The RAM system in rats and flies.

(ad) RAM labels active neuronal ensembles in rats exposed to inescapable stress (IS). (a) Schematic timeline of the experimental procedure. Rats were injected with AAV-RAM-NLS-mKate2 and AAV-Ef1α-EGFP vectors in the medial prefrontal cortex (mPFC). After 10 days they were either subjected to IS or left undisturbed in their home cages (HC). (b) Schematic drawing of the rat brain with the red region indicating the target area (PL: prelimbic cortex of the mPFC) for virus infection and quantification of RAM+ cells. (c) Percentage of RAM+ cells among total EGFP+ cells in the prelimbic cortex in HC and IS animals. Data are mean ± SEM, n = 3–4 animals per group, Student’s t-test, **p<0.01. (d) Representative images of prefrontal cortex showing mKate2 (red) and EGFP (green) fluorescence in rats subjected to IS or HC conditions. Areas in purple squares are enlarged in the right image for each condition. CC: corpus callosum. Scale bars are 500 and 100 μm for the left and right images, respectively. (eh) The Drosophila RAM reporter system. (e) Schematic diagram of the Drosophila RAM reporter system. The RAM-luc transgene can be turned on in specific cell types by the targeted expression of Flp recombinase using the GAL4-UAS system (left) or a cell-type specific driver (right). (f) Drosophila RAM reporter activity has low baseline levels and high fold induction. The specificity of the RAM-luc to Flp recombinase was tested using a UAS-Gal4 system, in which Flp recombinase expression is under the control of a heat-shock HS promoter. Flies in the no heat-shock (No HS) condition were maintained at 20°C throughout development and experimental conditions. For the heat-shock (HS) condition, flies were exposed to a 37°C heat shock for 30 min and allowed to recover for a full day at 20°C before measuring reporter expression. To ensure that results were not due to insertional effects, the UAS-flp transgene was combined with fly lines with the reporter transgene on either chromosome II (RAM-luc;UAS-flp) or chromosome III (UAS-flp;RAM-luc). n = 40–47 flies per group, Student’s t-test, ***p<0.001. (g) Pan-neuronal RAM-luc reporter expression displays circadian rhythm. The RAM-luc reporter transgene was combined with a transgene expressing FLP recombinase in all adult neurons and luciferase activity measured in live flies over time. Bars under plots indicate day (light) and night (dark). (h) Pan-neuronal RAM-luc reporter expression is sensitive to memory formation in Drosophila. Flies as described in g were trained in an olfactory memory task. 24 hr after training, flies exposed to Forward Spaced (FS) training showed significantly higher RAM-luc expression than control flies exposed to Backward Spaced (BS) training. Bars under plots indicate day (light) and night (dark). n = 23–24 flies per group, Student’s t-test, **p<0.01.

https://doi.org/10.7554/eLife.13918.017
Figure 7 with 4 supplements
RAM labeling of transcriptionally active interneurons and application of Cre-dependent RAM (CRAM).

(ad) RAM labels active GAD67+ neurons in rats exposed to inescapable stress (IS). (a) Schematic timeline of the experimental procedure. Rats were injected with AAV-RAM-NLS-mKate2 and AAV-Ef1α-EGFP in the mPFC. After 10 days they were either subjected to IS or left undisturbed in their home cage (HC). (b) Schematic drawing of the rat brain with the red region indicating the target area (PL: prelimbic cortex of the mPFC) for viral infection and quantification of RAM+ cells. (c) Representative images of prefrontal cortex showing EGFP (green), mKate2 (red) and GAD67 (blue) fluorescence in rats subjected to IS or HC conditions. White arrows indicate RAM+ and GAD67+ double-labeled cells. The scale bar is 100 μm and applies to all images. (d) Percentage of RAM+ cells co-labeled with GAD67+ in the prelimbic cortex of HC and IS animals. Data are mean ± SEM, n = 3–4 animals per group, Student’s t-test. (eg) Neuronal activity-dependence of the PRAM promoter in GABAergic neurons. (e) Design of AAV vectors used for luciferase assays in dissociated neuronal cultures from Gad2-Cre transgenic mice. The reading frames of luciferase (luc2p) and renilla are double inverted and flanked by double loxP sites (FLEX) and inserted downstream of the PRAM and thymidine kinase promoter (PTK) respectively. (f) Relative luciferase activity of PRAM in Gad2-Cre cells following KCl stimulation, with application of Nimodipine and/or APV. n = 3 separate experiments per condition, one-way ANOVA, Tukey’s post-hoc test. (g) Relative luciferase activity of PRAM in Gad2-Cre cells after application of various neurotrophic factors and drugs. n = 3 separate experiments per condition, Student’s t-test. (h) Schematic diagram of the AAV-CRAM construct. The effector gene is flanked by double loxP sites (FLEX). (il) CRAM labels IS-activated mPFC neurons projecting to the dorsomedial striatum (DMS). (i) Experimental procedure. Rats were injected with CAV2-Cre in the DMS and AAV-CRAM-tdT in the PL (j), then either subjected to IS or left undisturbed in their home cages (HC). (k) Number of CRAM+ cells in the prelimbic cortex of HC and IS animals. Data are mean ± SEM, n = 2 animals per group. (l) Representative images of the prefrontal cortex showing tdT fluorescence in rats subjected to IS or HC. Areas in purple squares are enlarged in the lower images. CC: corpus callosum. Scale bars are 500 and 100 μm for the upper and lower images, respectively. Data in d, f, g and k are mean ± SEM. **p<0.01, ***p<0.001.

https://doi.org/10.7554/eLife.13918.018
Figure 7—figure supplement 1
RAM labeling of transcriptionally active parvalbumin (PV) cells after inescapable stress (IS).

(a) Schematic timeline of the experimental procedure. Rats were injected with AAV-RAM-NLS-mKate2 and AAV-Ef1α-EGFP vectors in the mPFC. After 10 days they were either subjected to IS or left undisturbed in their home cage (HC). (b) Schematic drawing of the rat brain with the red region indicating the target area (PL: prelimbic cortex of the mPFC) for virus infection and quantification of RAM+. (c) Percentage of RAM+ cells co-labeled with PV+ in the prelimbic cortex in HC and IS animals. Data are mean ± SEM, n = 3–4 animals per group, Student’s t-test, *p<0.05. (d) Representative images of prefrontal cortex showing EGFP (green), mKate2 (red) and PV (blue) fluorescence in rats subjected to IS or HC conditions. White arrow shows a RAM+ and PV+ double-labeled cell. The scale bar is 100 μm and applies to all images.

https://doi.org/10.7554/eLife.13918.019
Figure 7—figure supplement 2
RAM labeling in the DG hilus.

(a) Schematic timeline of the experimental procedure. Following injection of AAV-RAM-NLS-mKate2 and AAV-Ef1α-EGFP into DG, animals were left on Dox diet (+Dox) for at least 7 days. Dox diet was withdrawn (-Dox) 48 hr before CFC and animals were sacrificed 24 hr afterwards. A control cohort of animals was left undisturbed in their home cages (HC) and received the same Dox treatment. (b) Schematic diagram of the hippocampus showing the targeted viral injection site (blue). RAM+ and EGFP+ double-labeled cells located exclusively in the hilus of the DG (red area) are quantified in panel c. (c) Percentage of RAM+ cells among total EGFP+ cells in the DG hilus for HC and CFC conditions. Data are mean ± SEM. n = 4 animals per condition, Student’s t-test, n.s., non-significant. (d) Percentage of RAM+ cells in the DG hilus region that are mossy cells (GluR2/3+, 49 out of 86 cells) and somatostatin positive (SST+, 101 out of 254 cells) after CFC. n = 3 animals per condition. (e) Representative image of RAM+/GluR2/3+ cells (indicated by arrows) in the DG hilus after CFC. The scale bar is 50 μm. (f) Representative image of RAM+/SST+ cells (indicated by arrows) in the DG hilus after CFC. The scale bar is 20 μm.

https://doi.org/10.7554/eLife.13918.020
Figure 7—figure supplement 3
Validation of the AAV-CRAM system.

(a) CRAM-tdT and Ef1α-EGFP viruses were injected into wild-type or Gad2-Cre animals, and after at least 7 days kainic acid treatment was given to induce seizures. Dox was not administered throughout the experiment. (b) Schematic drawing of the hippocampus with the injection site (DG) highlighted in blue. (c) In wild-type animals, CRAM-labeled cells (tdT) were not detectable 24 hr after kainic acid treatment. The scale bar is 200 μm for the left image and 50 μm for the zoomed-in images on the right. (d) In Gad2-Cre animals, 24 hr after kainic acid treatment, CRAM labeled only GABAergic (predominately somatostatin –expressing) neurons while sparing the granule cells. The scale bar is 200 μm for the image on the top and 50 μm for the zoomed-in images on the bottom.

https://doi.org/10.7554/eLife.13918.021
Figure 7—figure supplement 4
FOS expression in parvalbumin (PV) and somatostatin (SST) positive cells in primary visual cortex (V1) and dentate gyrus (DG) of the hippocampus.

(a) Percentage of FOS+ cells among total PV+ and SST+ cells in the V1 region and DG for saline and seized conditions. Behavioral seizures were induced by PTZ or KA for maximal FOS expression. Data are mean ± SEM. n = 4–8 animals per condition, two-way ANOVA, Tukey’s post-hoc test. (b, c) Representative images of V1 (b) and DG (c) showing FOS (green), SST (red), and PV (red) immuno-staining. The scale bar is 100 μm for all large images. For the zoomed-in images the scale bar is 50 μm for Saline and PTZ conditions in V1, and 25 μm and 50 μm for PV and SST conditions in DG, respectively. These images are enlarged from the areas marked by purple squares. Red arrows indicate FOS+ and PV+ or SST+ double-labeled cells.

https://doi.org/10.7554/eLife.13918.022

Additional files

Supplementary file 1

A full list of AP-1 containing 10mers (128 in total) ranked by their enrichment factor.

See ‘Methods’ for further details. 10mers used in PRAM and enhancer elements E1-E3 (Figure 1a, and Figure 1—figure supplement 2b) are highlighted in bold. 10mers with corrected p-value≥0.05 are marked in grey color.

https://doi.org/10.7554/eLife.13918.023
Supplementary file 2

Experimental conditions and statistics.

https://doi.org/10.7554/eLife.13918.024

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Andreas T Sørensen
  2. Yonatan A Cooper
  3. Michael V Baratta
  4. Feng-Ju Weng
  5. Yuxiang Zhang
  6. Kartik Ramamoorthi
  7. Robin Fropf
  8. Emily LaVerriere
  9. Jian Xue
  10. Andrew Young
  11. Colleen Schneider
  12. Casper René Gøtzsche
  13. Martin Hemberg
  14. Jerry CP Yin
  15. Steven F Maier
  16. Yingxi Lin
(2016)
A robust activity marking system for exploring active neuronal ensembles
eLife 5:e13918.
https://doi.org/10.7554/eLife.13918