Fig. 8: PV and SST plasticity ensures place cell stability and fidelity across multiple environments.

a Simulation protocol. An animal explores environment A for ten laps. It is then moved to environment B and explores it for 15 laps. Finally, the animal is moved back to the first environment (A′) and is allowed to run for another ten laps. Throughout this protocol, two-compartment CA1 pyramidal cells are simulated receiving spatially tuned Schaffer collateral inputs, temporoammonic input and PV and SST inhibitory inputs. Schaffer collateral synapses follow a Hebbian-type excitatory plasticity dependent on the coactivation of dendritic and somatic compartments (see “Methods”). PV and SST synapses undergo rate-based iPlas (PV-iLTD and SST-iLTP). We simulate the switch from environment A to environment B by randomly shuffling the identity of the SC inputs to the CA1 pyramidal neuron. Schematic depictions of environments A and B indicate their cyclical nature. b Diagram of simulated CA1 pyramidal cell and examples of somatic activity during exploration for iPlas ON. With iPlas (PV-iLTD and SST-iLTP) ON, place field location formed in environment A (top and bottom panel) remains stable after exposure to novel environment B (middle panel). c Diagram of simulated CA1 pyramidal cell and examples of somatic activity during exploration for iPlas OFF. Location of place fields formed in environment A is not maintained after exposure to a novel environment. d Spatial correlation between environment A before and after exposure to novel environment is maintained when iPlas is present but is reduced without iPlas (p < 0.0001, Mann–Whitney test, n = 100). e When iPlas is ON, average somatic activity of recently formed place cell is significantly reduced in new environment B but returns to higher levels when the animal returns to environment A′ (p < 0.0001, Friedman test, n = 100). f When iPlas is OFF, average somatic activity remains high in both environments (A versus B, p > 0.999; A versus A′, p = 0.472; B versus A′, p = 0.143, Friedman test, n = 100). g When SST-iLTP is turned off, leaving just PV-iLTD, place cell locations are not retained after exposure to a new environment. h When PV-iLTD is turned off, leaving just SST-iLTP, place cell locations are maintained across multiple environments. i Spatial correlation between environment A before and after exposure to novel environment is maintained when only PV-iLTD is turned off but is reduced when only SST-iLTP is turned off (p < 0.0001, Mann–Whitney test, n = 100). j SST-iLTP is turned off, leaving just PV-iLTD, average somatic activity remains high in both environments (A versus B, p = 0.967; A versus A′, p = 0.774; B versus A′, p > 0.999, Friedman test, n = 100). k When PV-iLTD is turned off, leaving just SST-iLTP, average somatic activity is lower and thus less robust (p < 0.0001, Friedman test, n = 100). Data presented as mean values ± SEM. Statistical significance between groups was assessed via Dunn’s multiple comparisons test. See also Supplementary Fig. 7.