Progressive functional impairments of hippocampal neurons in a tauopathy mouse model

Abstract

The age-dependent progression of tau pathology is a major characteristic of tauopathies, including Alzheimer's disease (AD), and plays an important role in the behavioral phenotypes of AD, including memory deficits. Despite extensive molecular and cellular studies on tau pathology, it remains to be determined how it alters the neural circuit functions underlying learning and memory in vivo. In rTg4510 mice, a Tau-P301L tauopathy model, hippocampal place fields that support spatial memories are abnormal at old age (7-9 months) when tau tangles and neurodegeneration are extensive. However, it is unclear how the abnormality in the hippocampal circuit function arises and progresses with the age-dependent progression of tau pathology. Here we show that in young (2-4 months of age) rTg4510 mice, place fields of hippocampal CA1 cells are largely normal, with only subtle differences from those of age-matched wild-type control mice. Second, high-frequency ripple oscillations of local field potentials in the hippocampal CA1 area are significantly reduced in young rTg4510 mice, and even further deteriorated in old rTg4510 mice. The ripple reduction is associated with less bursty firing and altered synchrony of CA1 cells. Together, the data indicate that deficits in ripples and neuronal synchronization occur before overt deficits in place fields in these mice. The results reveal a tau-pathology-induced progression of hippocampal functional changes in vivo.

Keywords: Alzheimer's; learning and memory; neurodegeneration; place cells; tau.

Figure 1.

Tau pathology in young and old Tau mice. A, cresyl violet staining of coronal sections from a 4-month-old WT mouse, 4-month-old Tau mouse, 8-month-old WT mouse, and 8-month-old Tau mouse. Arrows indicate recording sites. Note the thinner than normal CA1 pyramidal cell layer in the 8-month-old Tau mouse, but the nearly normal CA1 layer thickness in the 4-month-old Tau mouse. B, C, MC-1 and PHF-1 stained coronal sections of 4-month-old (B) and 8-month-old (C) WT and Tau mice. For each column, the boxed area in CA1 of the top picture is expanded at the bottom. Note the extensive staining (brownish black) of the CA1 pyramidal cells for both MC-1 and PHF-1 in the 8-month-old Tau mouse (C), but only MC-1 staining in the 4-month-old Tau mouse (B). Scale bars: black, −0.5 mm; red, −25 μm.

Figure 2.

Place fields on familiar and novel tracks in young Tau mice were largely normal. A, B, Example place cells recorded from a WT (A) and a Tau (B) mouse during two sessions in which the mice ran along a trajectory on a familiar track. In each panel, the top spike raster plot shows the linearized position of the animal at the time of each spike during each lap of a session, and the bottom curve shows the average firing rate of the cell across all the laps of the session. Arrows indicate running direction. Asterisks indicate place fields. C–E, Distributions of spatial information (C), place field length (D), and stability (E) of WT and Tau place cells on the familiar track. Counts are normalized by total number of cells. F–H, Same as C–E, but for place cells on a novel track.

Figure 3.

Theta power was normal in young Tau mice during track running. A, B, Raw and theta-filtered (pass band, 6–12 Hz) LFP traces of a young WT mouse (A) and a young Tau (B) mouse. Calibrations apply to traces in both A and B. C, Average (mean and SE) PSD of LFPs recorded from young WT (blue) and Tau (red) mice. D, Same as C, but with PSDs normalized by total power of LFPs. E, Average (mean and SE) theta power of WT and Tau mice. F, Same as E, but normalized by total power. G, Spike theta phases versus spike positions for a place cell recorded from a young WT mouse and a cell recorded from a young Tau mouse. The phases (top) are plotted with three cycles to show phase precession. The firing rate curve of each cell (bottom) is also shown. Lines indicate linear regression results. H–J, Mean circular correlation (H), mean linear correlation (I), and mean slope of regression (J) between spike phases and spike positions across all cells recorded from young WT and Tau mice. ***p < 0.0001.

Figure 4.

Ripple power during SWS was reduced in young and old Tau mice. A–D, Raw and ripple-filtered (pass band, 100–250 Hz) LFP traces of a young WT mouse (A), a young Tau mouse (B), an old WT mouse (C), and an old Tau mouse (D). Scale bars apply to traces from all four mice. E, F, Average (mean and SE, which is half of the curve thickness) PSD of LFPs recorded from young (E) and old (F) WT (blue) and Tau (red) mice. G, H, Same as E and F, but with the PSD normalized by the power of the baseline frequency band of 80–100 Hz. I, Average (mean and SE) ripple power within the frequency band of 120–180 Hz in young and old WT and Tau mice. J, Same as E, but with ripple power normalized by the power of baseline frequency band of 80–100 Hz. ***p < 0.001.

Figure 5.

Identified ripple events during SWS were smaller in young and old Tau mice. A–C, Average amplitude (A), average duration (B), and average occurrence rate (C) of ripples during slow-wave sleep sessions for young and old WT (blue) and Tau (red) mice. Each dot is the average value of all ripples recorded in a session. Black lines are median values. D, Average ripple amplitudes of all sessions (blue, WT mice; red, Tau mice) are plotted against animal age at time of recording. Each dot represents a session. Red and blue lines are linear regressions for the red and blue dots, respectively, within the young age group (<5 months). *p < 0.05; **p < 0.01; ***p < 0.001. E, Ripple-triggered LFP averages of two example SWS sessions, one from a young WT mouse and one from a young Tau mouse. F, Ripple amplitudes plotted versus the peak amplitudes of ripple-triggered LFP averages for individual SWS sessions recorded from WT and Tau mice. Solid dots, Young mice; open circles, old mice; lines, linear regressions for the young WT and young Tau mice.

Figure 6.

A, B, Distributions of burst index for neurons recorded from young (A) and old (B) WT and Tau mice during SWS. Each plot is a histogram normalized by the total number of neurons.

Figure 7.

Firing synchrony among CA1 neurons during SWS was reduced in young and old Tau mice. A, B, Average pairwise cross-correlograms during SWS for young (A) and old (B) WT and Tau mice. Thin lines are the mean ± SE. Note the smaller peaks in the average cross-correlograms of Tau mice, compared to those of age-matched WT controls. C, Mean pairwise correlation, defined as the average cross-correlation value within the [−20, 20] ms time lag, for cell pairs recorded from young and old WT and Tau mice. ***p < 0.00001.