Dynamic changes in the hippocampal neuronal circuits activity following acute stress revealed by miniature fluorescence microscopy imaging

Abstract

Coordinated activity of neuronal ensembles is a basis for information processing in the brain. Recent development of miniscope imaging technology enabled recordings of neuronal circuits activity in vivo in freely behaving animals. Acute stress is believed to affect various hippocampal functions, especially memory. In the current study, we utilized miniscope imaging to investigate the hippocampal neuronal circuits properties in a mouse as function of time and immediately in response to an acute stress, induced by passive restraint, 3 h and 10 days after. Comprehensive quantitative analysis of network activity changes at the neuronal ensembles level revealed highly stable neuronal activity parameters, which exhibited a rapid and robust shift in response to acute stress stimulation. This shift was accompanied by the restructuring of the pairwise-correlated neuronal pairs. Remarkably, we discovered that ensembles activity characteristics returned to the initial state following recovery period, demonstrating hippocampal homeostatic stability at the neuronal circuits level. Obtained results provide an evidence about hippocampal neuronal ensembles activity in response to acute stress over time.

Keywords: Acute stress; Calcium imaging; Hippocampus; Homeostatic stability; Miniature fluorescence microscopy; Miniscope; Neuronal circuits; Quantitative analysis.

Fig. 1

Experimental pipeline schematic presentation. a Freely behaving mouse with an attached miniscope v3. b Fixed sagittal brain slice with hole above the hippocampus after surgery, 4× magnification (in the right corner fluorescent image of the GCaMP6f fluorescence, 10× magnification). By dotted line GRIN lens boarders are drawn. c CA1 hippocampal neurons activity recorded by the miniscope. Illustration represents sum of 1000 frames from a single recording. d Timescale for neuronal activity visualization by miniscope in the Open-field behavioral test with acute stress modeling as a great external stimulus. e Extracted single neuron activity as active and inactive state using spike method with values “warm” of 50 and “cold” of 0 in “NeuroActivityToolkit”. f Recorded neuronal ensembles activity representation in a binarized form for a single recording (mouse 1, baseline day 1, 182 neurons). White lines indicate single neuron activation

Fig. 2

Shift in the activity properties at the single neuronal level as well as in the entire neuronal ensembles after acute stress modeling. a–c Calcium events number per minute for individual mouse for single-cell activity comparison. d Mean value of calcium events as total neuronal circuits characteristic for all states (baseline (n = 15) vs stress (n = 3), p = 0.0002; baseline (n = 15) vs 3 h (n = 3), p = 0.0009; baseline (n = 15) vs 10 days (n = 3), p = 0.9690; stress (n = 3) vs 3 h (n = 3), p = 0.9467; stress (n = 3) vs 10 days (n = 3), p = 0.0066; 3 h (n = 3) vs 10 days (n = 3), p = 0.0212; Ordinary One-way ANOVA with Tukey test for multiple comparison, F(3,20) = 14.29). e Individual response from the shared neurons between sessions for all states. All comparisons are significantly different with p < 0.0001, except: stress (n = 73 neurons) vs 3 h (n = 64 neurons), p > 0.9999 and baseline (73 neurons) vs 10 days (64 neurons), p > 0.9999. f Percentage of neurons that responded to stress or not for stress, 3 h and 10 days states. (g) Left: representative calcium traces of the same stress activated neuron on day 5 (baseline), stress, 3 h and 10 days states. Middle: representative calcium traces of the same stress inhibited neuron on day 5 (baseline), stress, 3 h and 10 days states Right: representative calcium traces of the same stress non-responsive neuron on day 5 (baseline), stress, 3 h and 10 days states. Scale bars corresponds to 25 s. Kruskal–Wallis test with Dunn’s test for multiple comparisons was used for comparisons in a–e, ns: no significant difference, #: p < 0.05; ##: p < 0.01; ###: p < 0.001; ####: p < 0.0001. Ordinary one-way ANOVA following Tukey post-hoc test was implemented for comparison in d, *: p < 0.05; **: p < 0.01; ***: p < 0.001; ****: p < 0.0001. All the data presented as mean ± SEM

Fig. 3

Hippocampal neuronal ensembles activation properties in normal and perturbed conditions. Distributions of neurons percent with given number of activation in normal condition and a right after stress (baseline (n = 15) vs stress (n = 3): 0.66–1.00: p = 0.0403; 1.00–1.33: p = 0.0059; 1.33–1.66: p = 0.0029; 3.00–3.33: p = 0.0206; 3.33–3.66: p = 0.0029; 3.66–4.00: p = 0.0250; 4.33–4.66: p = 0.0235; 4.66–5.00: p = 0.0029; 5.00–5.33: p = 0.0103; 5.33–5.66: p = 0.0029; 5.66–6.00: p = 0.0015, Mann–Whitney test for all). b 3 h post acute stress (baseline (n = 15) vs 3 h (n = 3): 2.33–2.66: p = 0.0029; 2.66–3.00: p = 0.0029; 3.00–3.33: p = 0.0029; 3.33–3.66: p = 0.0029; 3.66–4.00: p = 0.0029; 4.00–4.33: p = 0.0015; 4.33–4.66: p = 0.0235; 4.66–5.00: p = 0.0029; 5.00–5.33: p = 0.0103; 5.33–5.66: p = 0.0059, Mann–Whitney test for all). c 10 days post acute stress (baseline (n = 15) vs 10 days (n = 3): 0.33–0.66: p = 0.0208; others are non-significant, Mann–Whitney test for all). Distributions of network spike rate metric in normal condition and d right after stress (baseline (n = 15) vs stress (n = 3): 0.0–2.5: p = 0.0018; 2.5–5.0: p = 0.0018; 10.0–12.5: p = 0.0044; 12.5–15.0: p = 0.0026; 15.0–17.5: p = 0.0044; 17.5–20.0: p = 0.0184, Mann–Whitney test for all). e 3 h post acute stress (baseline (n = 15) vs 3 h (n = 3): 0.0–2.5: p = 0.0018; 2.5–5.0: p = 0.0026; 7.5–10.0: p = 0.0035; 10.0–12.5: p = 0.0009; 12.5–15.0: p = 0.0026; 15.0–17.5: p = 0.0219, Mann–Whitney test for all). f 10 days post acute stress (baseline (n = 15) vs 10 days (n = 3): 7.5–10.0: p = 0.0228, others are not significantly differing, Mann–Whitney test for all). Distributions of active cells percentage above threshold in normal condition and g right after stress (baseline (n = 15) vs stress (n = 3): 2.5%: p = 0.0025; 5.0%: p = 0.0012; 10.0%: p = 0.0257; 15.0%: p = 0.0282; 20.0%: p = 0.0221; 25.0%: p = 0.0025, Mann–Whitney test for all). h 3 h post acute stress (2.5%: p = 0.0025; 5.0%: p = 0.00025; 10.0%: p = 0.0025; 15.0%: p = 0.051; 20.0%: p = 0.0147; 25.0%: p = 0.051, Mann–Whitney test for all). i 10 days post acute stress (all the values are not significantly differing, Mann–Whitney test for all). #: p < 0.05; ##: p < 0.01; ###: p < 0.001. All the data presented as mean ± SEM

Fig. 4

Pairwise correlation properties stay stable while percent of co-active neurons elevates after acute stress. a Pearson’s correlation coefficient value for all states with preset threshold values calculated by “spike” method. b Pearson’s correlation coefficient value for all states with preset threshold values calculated by “full” method. c Number of connected pairs of neurons in relation to all the pairs with preset threshold (threshold = 0; baseline (n = 15) vs 3 h (n = 3): p = 0.012, Kruskal–Wallis test following Dunn’s test for multiple comparison). d Number of connected pairs of neurons as a fraction from all pairs with preset threshold. Network degree (e) right after stress (baseline (n = 15) vs stress (n = 3), threshold: 0.00: p = 0.0473; 0.05: p = 0.0156; 0.10: p = 0.0091; 0.15: p = 0.0098; 0.20: p = 0.0118; 0.25: p = 0.0158; 0.30: p = 0.0175; Student’s t-test for all). f 3 h post acute stress (baseline (n = 15) vs 3 h (n = 3), threshold: 0.00: p = 0.0155; 0.05: p = 0.0132; 0.10: p = 0.0167; 0.15: p = 0.0277; 0.20: p = 0.0460; 0.25: p = 0.0688; 0.30: p = 0.100, Student’s t-test for all). g 10 days post acute stress (all differences are non-significant, Student’s t-test). ns-non-significant; #: p < 0.05; #: *: p < 0.05; **: p < 0.01. All the data presented as mean ± SEM. For e, f and g dotted line represents threshold level for “strongly” correlated units

Fig. 5

Redistribution of strongly correlated hippocampal neuronal pairs in space in a response to acute stress. a Mean Euclidian distance for weakly correlated neurons (p = 0.9040, F(3,20) = 0.1869, One-way ANOVA test). b Mean radial distance for weakly correlated neurons (p = 0.2875, F(3,20) = 1.347, One-way ANOVA test). c Mean Euclidian distance for strongly correlated neurons (baseline (n = 15) vs stress (n = 3): p = 0.0190, F(3,20) = 4.178, One-way ANOVA with Dunnett’s post-hoc test). d Mean radial distance between strongly correlated neuronal pairs (baseline (n = 15) vs stress (n = 3): p = 0.0026; baseline (n = 15) vs 3 h (n = 3): p = 0.0479, F(3,20) = 6.936, One-way ANOVA with Dunnett’s post-hoc test). ns-non-significant; *: p < 0.05; **: p < 0.01. All the data presented as mean ± SEM

Fig. 6

Hippocampal neuronal ensembles properties representation in 2D coordinates in normal conditions and at different time points after acute stress. a Total representation of the neuronal ensembles state in different experimental time points. PCA coordinates for baseline and b right after acute stress state (baseline (n = 15) vs stress (n = 3): X: p = 0.0054; Y: p = 0.3495, Student’s t-test). c 3 h post acute stress modeling (baseline (n = 15) vs 3 h (n = 3): X: p = 0.0447; Y: p = 0.0139, Student’s t-test for both). d 10 days after stress modeling (baseline (n = 15) vs 10 days (n = 3): X: p = 0.9020; Y: p = 0.5656, Student’s t-test). ns-non-significant; *: p < 0.05; **: p < 0.01. All the data presented as mean ± SEM