Dorsal-CA1 Hippocampal Neuronal Ensembles Encode Nicotine-Reward Contextual Associations

Abstract

Natural and drug rewards increase the motivational valence of stimuli in the environment that, through Pavlovian learning mechanisms, become conditioned stimuli that directly motivate behavior in the absence of the original unconditioned stimulus. While the hippocampus has received extensive attention for its role in learning and memory processes, less is known regarding its role in drug-reward associations. We used in vivo Ca²⁺ imaging in freely moving mice during the formation of nicotine preference behavior to examine the role of the dorsal-CA1 region of the hippocampus in encoding contextual reward-seeking behavior. We show the development of specific neuronal ensembles whose activity encodes nicotine-reward contextual memories and that are necessary for the expression of place preference. Our findings increase our understanding of CA1 hippocampal function in general and as it relates to reward processing by identifying a critical role for CA1 neuronal ensembles in nicotine place preference.

Keywords: dorsal CA1 hippocampus; in vivo calcium imaging; nicotine; place preference.

Figure 1. Chemogenetic inhibition of dorsal-CA1 neurons blocks nicotine-CPP acquisition and expression

(A) Diagram of bilateral AAV5-CaMKIIα-HA-hM4D(Gi)-IRES-mCitrine DREADD viral injections into the dorsal-CA1 region of the hippocampus (top). Confocal (10×) image of AAV5-CaMKIIa-HA-hM4D(Gi)-IRES-mCitrine DREADD expression in CA1 (bottom). Scale bar =250 microns. (B) Zoomed in confocal (10×) image of AAV5-CaMKIIa-HA-hM4D(Gi)-IRES-mCitrine DREADD expression (20× inset), and (C) AAV5-CaMKIIα-mCitrine (control virus) expression in the CA1 (20× inset), mCitrine (green), See also Figure S2F. (D) Timeline of behavioral testing for CPP acquisition experiment. (E) Activation of hM4D(Gi) DREADDs in the CA1 by CNO 30 minutes before each nicotine (PM) conditioning session attenuates the acquisition of nicotine CPP and nicotine-primed CPP (n=9 mCitrine mice (controls), n=13 hM4D(Gi) mice, Two-Way ANOVA, F_(1,20)=23.31,p<0.05, bonferonni post hoc post-test (left two bars): ^*p<0.01, prime (right two bars): ^***p<0.0001), (F) without changing locomotor responses during post testing (Two-Way ANOVA F_(1,20)=0.052, p=0.82). (G) CNO before nicotine conditioning sessions does not change locomotor responses to nicotine (Two-way ANOVA, F_(1,20)=0.24,p=0.63). (H) Representative behavior tracks from hM4D(Gi) and control mice during CPP post testing and nicotine prime test. (I) Timeline of behavioral testing for CPP expression experiment. (J) Activation of hM4D(Gi) DREADDs in the CA1 by CNO 30 minutes before the CPP post-test blocks nicotine CPP expression, but does not affect the expression of nicotine-primed CPP 24 hours later (n=11 mCitrine mice (controls), n=12 hM4D(Gi) mice, Two-way ANOVA, F_(1,21)=10.63,p=0.004, bonferonni post hoc post-test (left two bars): p=0.02, prime (right two bars): p=0.44), (K) without changing locomotor response to during testing (Two-Way ANOVA, F_(1,21)=3.26,p=0.78). (L) No difference in conditioning day locomotor behavior was seen between control and hM4D(Gi) mice used in the CPP expression experiment (Two-way ANOVA, F_(1,21)=0.012,p=0.91). All data are presented as mean ± SEM.

Figure 2. In vivo Ca²⁺ imaging of dorsal-CA1 hippocampal neuronal activity during the acquisition of nicotine conditioned place preference

(A) Cartoon of virus injection (left) with 10× confocal image of AAV5-CaMKIIα-GCaMP6f expression in the CA1 HIP (middle) and cartoon of microendoscope on top of confocal image of CA1 region with GRIN lens implant (right), GCaMP6f (green), Nissl (blue). (B) Raw in vivo epifluorescence image of AAV5-CaMKIIα-GCaMP6f expression in the CA1 taken with mini-scope from (A) during CPP behavior. (C) Transformed image of (B) to show relative change in fluorescence (ΔF/F0). White arrows indicate cell bodies. (D) Photograph of mouse with mini-epifluorescence microscope for Ca²⁺ imaging during behavior.(E) CPP behavioral testing/Ca²⁺ imaging timeline and cartoon of CPP apparatus. See also Figure S1A–C. (F) Mice conditioned with nicotine (Nicotine-paired CPP expressing mice) had significantly higher CPP scores (time spent in the nicotine-paired side during post-test minus pre-test) than saline control mice (Unpaired t-test, n=6 per group, t₍₁₀₎=3.72,p=0.004). See also Figure S1D. (G) Correlation between Ca²⁺ event frequency per cell during nicotine (PM) conditioning sessions and subsequent CPP scores in nicotine-paired mice (CPP expressing and non-CPP expressing, n=8, Pearson’s correlation coefficient r=0.74,p=0.03). See also Figure S1E–G. (H) Neuron map and pie chart indicating that >90% of imaged neurons were successfully tracked across all 6 behavior sessions (white cells). Blue cells indicate neurons missed during the pre-test and red cells indicate neurons missed during the post-test (<5%). (I) Representative Ca²⁺ temporal traces on corresponding neuron maps from 30 seconds of activity recorded during a saline-paired (AM) (left) and nicotine-paired (PM) (right) conditioning session. Purple traces represent neurons that only show Ca²⁺ activity during nicotine-paired sessions. Green traces indicate neurons that only show Ca²⁺ activity during the saline-paired session. Corresponding neuronal egg map labels neurons based on whether they displayed Ca²⁺ activity during the saline-paired session (green neurons), during the nicotine-paired session (purple neurons) or during both sessions (grey neurons). See also Figure S1H–K. (J) In CPP expressing mice, the frequency of Ca²⁺ activity (events/minute/cell) of neurons active during the nicotine-paired session is significantly greater than during the saline-paired session (Paired t-test, n=5 mice, t₍₄₎=9.21,p=0.0008). (K) Saline-controls do not show a difference in Ca²⁺ frequency (events/minute/cell) between AM and PM conditioning sessions (Paired t-test, n=4 mice, t₍₃₎=1.27,p=0.29). See also Figure S1L–M. (L) Normalized Ca²⁺ response intensity distribution for nicotine CPP-expressing mice showing the percentage of cells that have higher intensity in the nicotine-paired side compared to the saline-paired side (n=5 mice, Paired t-test, t₍₄₎=5.44,p=0.01). (M) Normalized Ca²⁺ response intensity distribution for saline control mice showing the percentage of cells have higher intensity in the PM side compared to the AM side (n=4 mice, Paired t-test, t₍₃₎=0.13,p=0.90). See also Figure S1N–O. All scale bars equal 100 microns. Data are mean ±SEM. ^*p<0.05, ^**p<0.01.

Figure 3. In vivo Ca²⁺ imaging of CA1 hippocampal neuronal activity during the acquisition and expression of an operant reward task

(A) Timeline of imaging and Fixed Ratio 1 (FR1) schedule of operant lickometer training and illustration of mouse in operant chamber. (B) Representative neuron map and temporal traces for 10 neurons during a 1 hour FR1 training session. Corresponding colors represent neurons and traces. (C) Representative behavior from one mouse during the first day of FR1 training (day 33), no difference is seen between the number of active and inactive nosepokes during the first training session. (D–E) By day 37 (FR1 training day 5), mice poke the active nosepoke significantly more than the inactive nosepoke (F) Representative raster and histogram showing licking behavior aligned with Ca²⁺ events during the 5 seconds before and 20 seconds after each cue presentation. (G–I) Neurons are distributed into a two-dimensional field. The x-axis shows Ca²⁺ event frequency after each active nosepoke and the y-axis shows Ca²⁺ event frequency before each active nosepoke or reward (I). (J–L) Normalized Ca²⁺ response intensity distribution before and after each active nosepoke, before FR1 criteria met (day 33) (J) and after (day 37) (K) and before and after a reward (L). (M) Cellular spatial map and corresponding bar graph categorizing neurons based on whether they displayed Ca²⁺ activity for more than 20% of the total Ca²⁺ activity of that neuron time locked to one of two actions: An active nosepoke (blue), or during the first sucrose lick following sucrose delivery (green). Pink represents neurons that were active during both actions. Bar graph shows the number of neurons (%) in each group. No statistically significant differences were seen. (N) Normalized Ca²⁺ activity raster showing Ca²⁺ activity for neurons active during both behaviors (M, Pink neurons), during each active nosepoke (top) and during the first lick after sucrose delivery (bottom). See also Figure S2A–E. (O) Normalized Ca²⁺ activity raster plot showing Ca²⁺ activity for neurons active during each active nosepoke (M, Blue neurons). (P) Normalized Ca²⁺ activity raster plot showing Ca²⁺ activity for neurons active during the first lick after sucrose delivery (M, green neurons). Note: there is a brief period of inhibition in this group just before the onset of the first lick. All traces are mean with SD.

Figure 4. Total Ca²⁺ activity of all imaged CA1 cells mapped onto spatial locations within the CPP apparatus together with “Place Cell” analysis

(A–D) Gaussian-smoothed (σ=3.5cm) heat maps represent the total normalized Ca²⁺ transient activity (corrected by time on each pixel) for all cells corresponding to the location of the mouse within the CPP chamber during the pre-test and post-test in (A) CPP expressing mice, (B) saline controls, (C) non-CPP expressing, and (D) nicotine unpaired control mice. (E–F) Accumulated Ca²⁺ activity is plotted versus accumulated times for each 3cm² region in each side of CPP chamber for all mice by different colors in CPP expressing group (E) and saline group (F). Data from CPP-expressing mice in nicotine side during posttest (purple) is fitted with a linear polynomial regression curve with corresponding color. (G) Average percentage of place cells remapped from pretest to post-test (25cm²). CPP-expressing mice had a higher percentage of consistent place cells (with saline controls: Paired t-test, t₍₅₎=6.871,p=0.0010; with non-CPP-expressing: Paired t-test, t₍₅₎=6.165,p=0.0016; with nicotine-unpaired: Paired t-test, t₍₅₎=6.224,p=0.0016). (H) Place cell analysis showed that of the cells that responded to a specific location in the saline-paired side during the pre-test in Gaussian-smoothed (σ=3.5cm) density maps, were remapped after conditioning to encode a different spatial location in nicotine paired side during post-test (Representative cells 1–4). Red dots mark its position during Ca²⁺ events. (I) Some cells maintained place cell activity of their location during the pre-test across all testing sessions (Representative cells 5–8). All scale bars equal 10cm

Figure 5. In vivo Ca²⁺ imaging of dorsal-CA1 Hippocampal neuronal activity during the expression of nicotine conditioned place preference

(A) Timeline of CPP behavior and cartoon of CPP apparatus with arrows representing side transitions during the pre-test and post-test (Day 1 and Day 4). Purple arrows represent ‘nicotine-paired transitions’ from the saline-paired side to the nicotine-paired side (30 seconds before and after the mouse enters the nicotine-paired side). Green arrows represent ‘saline-paired transitions’ from the nicotine-paired side to saline-paired side (30 seconds before and after the mouse enters the saline-paired side). (B) CPP-expressing mice show a significantly larger change in frequency of Ca²⁺ activity (events/minute/cell) during nicotine-paired transitions during the post-test than was seen prior to conditioning during the pre-test (Paired t-test, n=5 mice, t₍₄₎=2.84,p=0.04). C–E. The frequency of Ca²⁺ activity did not change during ‘saline-paired’ or ‘nicotine-paired’ transition events in (C) saline control mice (n=4 mice, Paired t-test t₍₃₎=0.56,p=0.62), (D) non-CPP expressing mice (n=3 mice, Paired t-test t₍₂₎=0.15,p=0.89), or (E) nicotine-unpaired mice (n=3 mice, Paired t-test t₍₂₎=0.148,p= 0.89). (F–G) Neurons (n=1313 neurons, 5 mice) activated during nicotine-paired and saline-paired transitions the (F) pre-test and (G) post-test displayed in a 2-dimensional field where the x-axis represents the change in Ca²⁺ frequency during nicotine-paired transitions and the y-axis represents the change in Ca²⁺ frequency during saline-paired transitions. Purple points represent neurons with a higher frequency of Ca²⁺ transients (events/min/cell) during nicotine-paired transitions. Green points represent neurons with a higher frequency of Ca²⁺ transients during saline-paired transitions. See also Figure S3A–C. (H) CPP expressing mice have significantly more nicotine-paired neurons than saline-paired neurons (n=5 mice, paired t-test, t₍₄₎=5.32,p=0.006). (I–K) Control groups show no difference in the number of neurons paired with either transition context: (I) saline controls (n=4, Paired t-test, t₍₃₎=0.05,p=0.97), (J) non-CPP expressing (n=3, Paired t-test, t₍₂₎=0.26,p=0.83), or (K) nicotine-unpaired controls (n=3, Paired t-test, t₍₂₎=0.55,p=0.64). (L) Representative locomotor track from a nicotine CPP-expressing mouse during the CPP post-test and corresponding traces from neurons with higher Ca²⁺ activity during nicotine-paired transitions (purple traces) and neurons with higher activity during saline-paired transitions (green traces) displayed on egg maps. (M) Left panel: Representative neuron egg map from the pre-test showing neurons with higher frequencies of Ca²⁺ transients during either nicotine-paired (purple) or saline-paired (green) transitions. Right panel: Representative neuron egg map from a nicotine CPP-expressing mouse during the post-test (i.e. during CPP expression). More neurons have higher frequencies of Ca²⁺ transients during nicotine-paired transitions than during saline-paired transitions (n=701 nicotine-paired neurons (purple) and 419 saline-paired (green) neurons). (N–O) Raster represents total Ca²⁺ responses in CPP-expressing mice during nicotine-paired transitions and (O) during saline-paired transitions during the post-test. Each row represents one transition (top). Red line indicates actual time of side entry. The corresponding aligned trace represents the total normalized Ca²⁺ activity during each transition. See also Figure S3A–G and Figure 4A–E.