Dysregulated Wnt and NFAT signaling in a Parkinson's disease LRRK2 G2019S knock-in model

Abstract

Parkinson's disease (PD) is a progressive late-onset neurodegenerative disease leading to physical and cognitive decline. Mutations of leucine-rich repeat kinase 2 (LRRK2) are the most common genetic cause of PD. LRRK2 is a complex scaffolding protein with known regulatory roles in multiple molecular pathways. Two prominent examples of LRRK2-modulated pathways are Wingless/Int (Wnt) and nuclear factor of activated T-cells (NFAT) signaling. Both are well described key regulators of immune and nervous system development as well as maturation. The aim of this study was to establish the physiological and pathogenic role of LRRK2 in Wnt and NFAT signaling in the brain, as well as the potential contribution of the non-canonical Wnt/Calcium pathway. In vivo cerebral Wnt and NFATc1 signaling activity was quantified in LRRK2 G2019S mutant knock-in (KI) and LRRK2 knockout (KO) male and female mice with repeated measures over 28 weeks, employing lentiviral luciferase biosensors, and analyzed using a mixed-effect model. To establish spatial resolution, we investigated tissues, and primary neuronal cell cultures from different brain regions combining luciferase signaling activity, immunohistochemistry, qPCR and western blot assays. Results were analyzed by unpaired t-test with Welch's correction or 2-way ANOVA with post hoc corrections. In vivo Wnt signaling activity in LRRK2 KO and LRRK2 G2019S KI mice was increased significantly ~ threefold, with a more pronounced effect in males (~ fourfold) than females (~ twofold). NFATc1 signaling was reduced ~ 0.5-fold in LRRK2 G2019S KI mice. Brain tissue analysis showed region-specific expression changes in Wnt and NFAT signaling components. These effects were predominantly observed at the protein level in the striatum and cerebral cortex of LRRK2 KI mice. Primary neuronal cell culture analysis showed significant genotype-dependent alterations in Wnt and NFATc1 signaling under basal and stimulated conditions. Wnt and NFATc1 signaling was primarily dysregulated in cortical and hippocampal neurons respectively. Our study further built on knowledge of LRRK2 as a Wnt and NFAT signaling protein. We identified complex changes in neuronal models of LRRK2 PD, suggesting a role for mutant LRRK2 in the dysregulation of NFAT, and canonical and non-canonical Wnt signaling.

Keywords: Immune system; LRRK2; LRRK2 G2019S; LRRK2 KO; Mouse models; NFAT signaling; Parkinson’s diseases; Primary cultures; Sex differences; Wnt signaling.

Figure 1

Experimental design for in vivo analyisis of canonical Wnt and NFATc1 signaling changes in wild type, LRRK2 KO and LRRK2 G2019S KI mice over time. Mice were injected intracranially at P0 with lentivirus containing a construct composed of a RE site for either TCF/LEF or NFATc1, a gene encoding luciferase and a second gene encoding for GFP. Expressed luciferase is able to cleave luciferin as a substrate, producing a detectable bioluminescence signal (A). This bioluminescence was measured regularly over 28 weeks in wild type (WT), LRRK2 knockout (KO) and LRRK2 G2019S knock-in (KI) mice in vivo (B). After 28 weeks, post-mortem analysis revealed luciferase activity prominently in the brain, excluding the other organs (C). Luciferase kinetics was determined with an ideal measurement time frame, highlighted in dark blue, n = 3 (D). Relative luciferase activity is shown in WT mice over time with signal correction by the signal generated SFFV construct, used as positive control (E).

Figure 2

Canonical Wnt signaling changes in wild type, LRRK2 KO and LRRK2 G2019S KI mice in vivo over time. Mice were injected intracranially at P0 with a lentiviral biosensor for TCF/LEF to detect Wnt signaling activities. Repeated measures of the resulting bioluminescence signal were detected over a period of 28 weeks in wild type (WT), LRRK2 knockout (KO) and LRRK2 G2019S knock-in (KI) mice. Data is shown as mean relative luciferase activity ± SEM, normalized to WT set to one (A, B, and C). Sex separation reveals differences between LRRK2 genotypes for only male (B) and only female mice (C). The effect of time is shown as mean relative luciferase activity ± SEM, normalized to week one set to one for each individual LRRK2 genotype (D, E, and F). Sex related differences are represented as mean raw luciferase signal ± SEM in photons per seconds (G). Wnt signaling changes in WT, KO and KI half brain protein samples are shown as relative protein level for active β-catenin (ABC) in relation to total β-catenin (H–M). Data is shown as fold-change to WT and β-actin served as loading control. Statistical significance, indicated as ****P < 0.0001 and *P < 0.05 was tested for genotype differences and effect of time by mixed-effects analysis, and the impact of sex on basal in vivo Wnt signaling was analyzed by unpaired t-test with Welch’s correction. Immunoblot experiments were tested via unpaired t-test with Welch’s correction with n = 16, 8 males and 8 females. The original Western Blot images corresponding to Fig. 2 H-M are shown in the Supplementary Material 1 page 5.

Figure 3

NFATc1 signaling changes in wild type, LRRK2 KO and LRRK2 G2019S KI mice in vivo over time. Mice were injected intracranially at P0 with a lentiviral biosensor for NFATc1 to detect NFATc1 signaling activities. Repeated measures of the resulting bioluminescence signal were detected over a period of 28 weeks in wild type (WT), LRRK2 knockout (KO) and LRRK2 G2019S knock-in (G2019S) mice. The data is shown as mean relative luciferase activity ± SEM, normalized to WT set to one (A, B, and C). Sex separation reveals differences between LRRK2 genotypes for only male (B) and only female mice (C). The effect of time is shown as mean relative luciferase activity ± SEM, normalized to week one set to one for each individual LRRK2 genotype (D, E, and F). Sex related differences are represented as mean raw luciferase signal ± SEM in photons per seconds (G). Statistical significance, indicated as ****P < 0.0001, ***P < 0.0005, and **P < 0.005 was tested for genotype differences and effect of time by mixed-effects analysis, and the impact of sex on basal in vivo NFATc1 signaling was analyzed by unpaired t-test with Welch’s correction.

Figure 4

GFP staining of injected and non-injected mice with the TCF/LEF biosensor. Mice were injected intracranially at P0 with a lentiviral construct, containing a RE site for TCF/LEF and a second gene for GFP. Six months later, brains were collected from injected (1 and 2) and non-injected control mice (3 and 4). Brains were fixed in PFA, coronally cryosectioned into 40µm thick slices, which were stained via immunohistochemistry for GFP. Positive cells for GFP expression are visible in dark brown indicated by the green arrows. Positive cells were detectable in olfactory bulb (A), cortex (B), hippocampus (C), superior colliculus (D), striatum (E) and thalamus (F and G). No positive cells were measurable in corresponding regions of non-injected control brains. Images in 2 and 4 represent the zoom of indicated squares in 1 and 3. Scale bar for all images is 50µm.

Figure 5

GFP staining of injected and non-injected mice with the NFATc1 biosensor. Mice were injected intracranially at P0 with a lentiviral construct, containing a RE site for NFATc1 and a second gene for GFP. Six months later, brains were collected from injected (1 and 2) and non-injected control mice (3 and 4). Brains were fixed in PFA, coronally cryosectioned into 40µm thick slices, which were stained via immunohistochemistry for GFP. Positive cells for GFP expression are visible in dark brown indicated by the green arrows. Positive cells were detectable in olfactory bulb (A), cortex (B), hippocampus (C), superior colliculus (D) and thalamus (E and F). No positive cells were measurable in corresponding regions of non-injected control brains. Images in 2 and 4 represent the zoom of indicated squares in 1 and 3. Scale bar for all images is 50µm.

Figure 6

Signaling pathway component changes in the half brain of LRRK2 KO and LRRK2 G2019S KI mice compared to WT mice. Wnt and NFAT signaling component changes in half brain samples between LRRK2 KO and WT (A1, B1), and LRRK2 G2019S KI and WT (A2, B2) are shown on a transcriptional (A) and protein level (B). Relative expression of relevant gene candidates was detected via quantitative real time PCR. Data is shown as log₂fold-change to WT expression. Gapdh and Hprt served as housekeeping genes (A₁, A₂). Statistical significance shown as **P < 0.005 and *P < 0.05 was tested via unpaired t-test with Welch’s correction with n = 6, 3 males and 3 females (B₁, B₂). Representative plots are displayed, and data is shown as log₂fold-change to WT protein levels. β-actin served as loading control. Statistical significance indicated as ***P < 0.0005, **P < 0.005, and *P < 0.05 was tested via unpaired t-test with Welch’s correction with n = 6, 3 males and 3 females. Western blots in this figure were cut to display all results in a concise format. The original uncut Western Blot images corresponding to this figure are shown in the Supplementary Material 2.

Figure 7

Signaling pathway component changes in the cortex of LRRK2 KO and LRRK2 G2019S KI mice compared to WT mice. Wnt and NFAT signaling component changes in in cortical samples between LRRK2 KO and WT (A1, B1), and LRRK2 G2019S KI and WT (A2, B2) are shown on a transcriptional (A) and protein level (B). Relative expression of relevant gene candidates was detected via quantitative real time PCR. Data is shown as log₂fold-change to WT expression. Gapdh and Hprt served as housekeeping genes (A₁, A₂). Statistical significance shown as *P < 0.05 was tested via unpaired t-test with Welch’s correction with n = 6, 3 males and 3 females (B₁, B₂). Representative plots are displayed, and data is shown as log₂fold-change to WT protein levels. β-actin served as loading control. Statistical significance indicated as ***P < 0.0005, **P < 0.005, and *P < 0.05 was tested via unpaired t-test with Welch’s correction with n = 6, 3 males and 3 females. Western blots in this figure were cut to display all results in a concise format. The original uncut Western Blot images corresponding to this figure are shown in the Supplementary Material 3.

Figure 8

Signaling pathway component changes in the hippocampus of LRRK2 KO and LRRK2 G2019S KI mice compared to WT mice. Wnt and NFAT signaling component changes in hippocampal samples between LRRK2 KO and WT (A1, B1), and LRRK2 G2019S KI and WT (A2, B2) are shown on a transcriptional (A) and protein level (B). Relative expression of relevant gene candidates was detected via quantitative real time PCR. Data is shown as log₂fold-change to WT expression. Gapdh and Hprt served as housekeeping genes (A₁, A₂). Statistical significance was tested via unpaired t-test with Welch’s correction with n = 6, 3 males and 3 females (B₁, B₂). Representative plots are displayed, and data is shown as log₂fold-change to WT protein levels. β-actin served as loading control. Statistical significance indicated as **P < 0.005, and *P < 0.05 was tested via unpaired t-test with Welch’s correction with n = 6, 3 males and 3 females. Western blots in this figure were cut to display all results in a concise format. The original uncut Western Blot images corresponding to this figure are shown in the Supplementary Material 4.

Figure 9

Signaling pathway component changes in the striatum of LRRK2 KO and LRRK2 G2019S KI mice compared to WT mice . Wnt and NFAT signaling component changes in striatal samples between LRRK2 KO and WT (A1, B1), and LRRK2 G2019S KI and WT (A2, B2) are shown on a transcriptional (A) and protein level (B). Relative expression of relevant gene candidates was detected via quantitative real time PCR. Data is shown as log₂fold-change to WT expression. Gapdh and Hprt served as housekeeping genes (A₁, A₂). Statistical significance shown as **P < 0.005 and *P < 0.05 was tested via unpaired t-test with Welch’s correction with n = 6, 3 males and 3 females (B₁, B₂). Representative plots are displayed, and data is shown as log₂fold-change to WT protein levels. β-actin served as loading control. Statistical significance indicated as ***P < 0.0005, **P < 0.005, and *P < 0.05 was tested via unpaired t-test with Welch’s correction with n = 6, 3 males and 3 females. Western blots in this figure were cut to display all results in a concise format. The original uncut Western Blot images corresponding to this figure are shown in the Supplementary Material 5.

Figure 10

Stimulation of primary neuronal cultures from LRRK2 KO, LRRK2 G2019S KI and WT mice. Wnt (TCF/LEF) and NFATc1 signaling changes are presented under treated and untreated conditions in primary cultured cortical (A) and hippocampal cells (B) from LRRK2 KO (1), LRRK2 G2019S KI (2), and WT mice. Cells were transduced with a lentiviral biosensor displaying Wnt or NFATc1 signaling activity. Data is shown as mean relative luciferase activity ± SEM, normalized to untreated WT set to 100%. Bioluminescence was measured 24h after treatment. Statistical significance shown as ###P < 0.0005, ##P < 0.005, and #P < 0.05 was tested for the effect of LRRK2 genotype via 2-way ANOVA and Bonferroni’s multiple comparison test and the effect of treatment shown as ****P < 0.0001, ***P < 0.0005, **P < 0.005, and *P < 0.05 was tested via 2-way ANOVA and Turkey’s multiple comparison test with n = 5 independent cultures.