Human LRRK2 G2019S mutation represses post-synaptic protein PSD95 and causes cognitive impairment in transgenic mice

Abstract

Background: LRRK2 G2019S mutation is associated with increased kinase activity and is the most common mutation associated with late-onset PD. However, the transgenic mouse model has not recapitulated cardinal PD-related motor phenotypes. Non-motor symptoms of PD including cognitive impairments are very common and may appear earlier than the motor symptoms. The objective of this study was to determine whether human LRRK2 with G2019S mutation causes hippocampus-dependent cognitive deficits in mice.

Results: Male (LRRK2-G2019S) LRRK2-Tg mice showed impairments in the early portion of the Two-day radial arm water maze acquisition trial as well as in the reversal learning on the third day. However, their performance was similar to Non-Tg controls in the probe trial. LRRK2-Tg mice also displayed impairments in the novel arm discrimination test but not in the spontaneous alternation test in Y-maze. Interestingly, there was no statistically significant locomotor impairment during any of these cognitive test, nor in the locomotor tests including open field, accelerating rotarod and pole tests. Expression of the postsynaptic protein PSD-95 but not the presynaptic protein synaptophysin was lower in hippocampal homogenates of LRRK2-Tg mice.

Conclusion: Consistent with previous reports in human LRRK2 G2019S carriers, the current data suggests that cognitive dysfunctions are present in LRRK2-Tg mice even in the absence of locomotor impairment. LRRK2 G2019S mutation represses the postsynaptic protein PSD-95 but not the presynaptic protein synaptophysin. This study also suggests that mild cognitive impairment may appear earlier than motor dysfunctions in LRRK2-G2019S mutation carriers.

Keywords: Cognitive impairment; Hippocampus; LRRK2; Learning and memory; Parkinson’s disease.

Fig. 1

Impaired spatial and reversal learning in G2019S LRRK2-TG mice. (A) Average latency, or Time-to-target (escape platform) in seconds and (B) average number of errors made during the two-day radial-arm water maze (RAWM) training. In a probe trial (test without the platform) conducted 24 h after the last training trial of day 2, both Non-Tg and (G2019S) LRRK2-Tg mice showed preference for the target location which was significantly higher than chance level (dotted line; 20%) in both (C) Percent time spent in the target location and (D) Percent number of entry into the target location in the probe trial done 24 h after the last training trial. (E) On day 3, the location of the target was changed as shown i.e., target previously at D was moved to the opposite side of the tank to either A or B; or target previously at C moved to A or E, etc. (F) Time to the new target and (G) errors made before finding the new target in the course of training the mice after changing the target location (each trial set vs. 1st trial set). Data expressed as Mean ± SEM. N = 6 for both groups. Sidak’s post hoc test, One sample or Independent T-test p-values ^*p < 0.05; ^**p < 0.01; ^***p < 0.005; Sidak’s post hoc test Day 1 vs. Day 2 within each genotype p-value ^#p < 0.05; ^##p < 0.01; ^###p < 0.005.

Fig. 2

Impaired spatial recognition but intact working memory in G2019S LRRK2-Tg mice. (A) Percentage of time spent in, and (B) percentage of number of entry (frequency) into both familiar and novel arms during the 3 min retention trial. (C) Percent alternation scores in Y-maze during two 5-min periods of the test. Dotted line represent chance level (at 50%). Paired and One-sample T-test p-values ^*p < 0.05; ^**p < 0.01; ^***p < 0.005; Independent sample T-test (Non-Tg vs. (G2019S) LRRK2-Tg) ^#p < 0.05.

Fig. 3

No evidence of motor dysfunctions in G2019S LRRK2-Tg mice in open field and Rotarod tasks (A) Total distance travelled during each of the five 2-min blocks of monitoring in an open field box. (B) Percentage of time spent in, and (B) Ridding time of Non-Tg and (G2019S) LRRK2-Tg mice on the rotarod at different acceleration settings. Data point (each acceleration setting) shows an average of the two consecutive trials at the same setting.

Fig. 4

Reduced expression of postsynaptic-but not presynaptic protein in G2019S LRRK2-Tg mice hippocampus. Immunoblots and relative optical density (ROD) quantifications of (A) PSD-95 and (B) Synaptophysin expression in hippocampal homogenates of (G2019S) LRRK2-Tg mice. Protein expression levels were determined by western blot using PSD-95 and Synaptophysin specific antibodies respectively. Protein expression levels expressed as ROD relative to β-Actin which was the loading control. Data expressed as Mean ± SEM. N = 6 for both groups. Independent T-test p-values ^*p < 0.05; ns = not significant. (C) Representative immunofluorescence images of hippocampal dentate gyri of Non-Tg (i–iv) and LRRK2-Tg (v–viii) mice. The dotted lines in (C) iii, iv, vii and viii represent the boundaries of the hilus bordered by the granular cell layers (GCL) on either side. Note the punctate staining of PSD-95 in iii and vii which are absent within the nuclei and very rare in the cell bodies of the granular cell layer (NeuN-positive) neurons (ii and vi). Also note the fairly even distribution of the PSD-95 punctate staining throughout the dentate gyrus (Hilus and around the cell bodies in the GCL) in Non-Tg mice sections (iii). This is not so in the LRRK2-Tg sections where the PSD-95 punctate staining is fewer especially in the GCL and the adjacent molecular layer. Scale bar 50 mm. DAPI in Blue, NeuN in Green, PSD-95 in Red.