Fig. 1: Microstrokes impair spatial memory and place cell stability.

A Experimental setup: Mice were trained to navigate in a linear virtual reality (VR) corridor (4 m length) during simultaneous unilateral two-photon calcium imaging in CA1 of the hippocampus. B Schematic illustration of microstroke induction by injection of fluorescent microspheres (20 µm diameter) into the common carotid artery, inducing microstrokes. C Representative two-photon images of the same field of view (FOV) in CA1 showing GCaMP6f fluorescence before (left) and after (middle) stroke induction. Damaged tissue overexposed the sensor and is shaded red. Individual neurons and their ΔF/F calcium fluorescence traces could be tracked over several weeks (right). Landmark blood vessels are marked in red. Scale bars = 25 µm. D Timeline showing the sequence of events of the experiment in days relative to stroke induction. Chronic two-photon imaging in CA1 was performed while animals were navigating in the VR corridor during three experimental phases: before stroke (“healthy”, ≥−5 to 0 days before stroke), early (0–7 days) and late ( > 7 days – 28 days) after stroke. N = 5 independent experiments were performed with stroke and sham animals. E Histograms depicting the profile of an example mouse to lick for a water reward in the VR corridor. Green areas indicate reward zone locations. F Number of microspheres in the brain plotted against the task performance (spatial information of lick profile, in bits) during the three phases. Inset shows Pearson’s correlation coefficients, data points are individual animals in each stroke phase. “Healthy” represent data points from mice before microsphere injection serving as controls. G Left: Like F, but task performance plotted against the percentage of all imaged cells identified as place cells (place cell ratio). Right: Pearson’s correlation coefficients of task performance with metrics of neural activity such as the place cell ratio, the stability of neural spatial activity maps across trials (within-session stability) and the mean firing rate for the three phases. Error bars represent 95% confidence intervals. H Spatial activity maps of three exemplary neurons being active at the same corridor ___location (stable place cell), different locations (unstable place cell), or at no specific ___location (noncoding cell) during four healthy sessions. I Spatial activity maps of neurons imaged on multiple days throughout the experiment and sorted into the three functional classes from H (each row represents an individual neuron tracked before and after stroke). Boxplots are drawn with the box extending from the 25th to 75th percentiles, with the centre line at the median. Whiskers reach to the minimum and maximum values of the distribution. J–L Percentages of all tracked cells that were classified as stable place cells (J), unstable place cells (K) and noncoding cells (L). In J–L, statistics were evaluated using two-way repeated-measures ANOVA with the Greenhouse-Geisser correction and Tukey-Kramer multiple comparisons test. Asterisks indicate significances: *p < 0.05, **p < 0.01, ***p < 0.001.