Article Figures & Data
Figures
- Figure 1.
Broad alteration in mRNA translation in the G2019S LRRK2 mouse brain. A, A schematic of ribosome profiling workflow with mouse brain tissue. B–D, TE was calculated to estimate translational activity. Global TE distributions between (B) GS LRRK2 TG and non-TG control, (C) LRRK2 KO and WT, and (D) GS/DA LRRK2 TG and non-TG control were compared. All values are in log2, and each data point represents a single transcript. In scatterplots, centerline is a guideline with slope of 1, meaning that the dots on the line do not have TE value differences between the genotypes. SD of TE differences: 0.226 (GS LRRK2 vs control), 0.179 (GS/DA LRRK2 vs control), 0.273 (LRRK2 KO vs WT). Standard z score was calculated, and ±1.5 cutoff was used to select TE up and TE down genes. Triplet periodicity is normal across the results (Extended Data Fig. 1-1). E, F, Histogram of TE differences (δTE, ΔTE) between (E) GS LRRK2 TG and non-TG control or (F) LRRK2 KO and WT. Z score ±1.5 cutoff was used, and TE values are in log2. Each ribosome profiling experiment was firstly analyzed independently to ensure reproducibility. Two independent results were analyzed together by DESeq (Anders and Huber, 2010; n = 2). Expression analysis results including TE values were compiled (Extended Data Fig. 1-2).
- Figure 2.
5′UTR secondary structure mediates translational effects of G2019S LRRK2. A, B, Correlation between estimated 5′UTR folding energy and TE changes in (A) GS LRRK2 TG or (B) LRRK2 KO. Box plot overlaid with violin plot visualizes the median, the first and the third quartile along with the data distribution pattern. 5′UTR folding energy for transcripts was retrieved from UCSC genome database (mm9). The same z score ±1.5 cutoff was used. Group sizes: GS TG (TE up: 687, TE down: 335), KO (TE up: 596, TE down: 576). Statistical significance was determined by one-way ANOVA with Bonferroni correction. C, D, Genes with complex 5′UTR secondary structure (estimated folding energy: <–250 kcal/mol, 1145 genes) or simple 5′UTR secondary structure (>–20 kcal/mol, 1036 genes) were selected, and the TE differences between (C): GS LRRK2 TG mice and control mice (D): LRRK2 KO mice and WT mice were plotted. Statistical significance was tested with Wilcoxon signed-rank test [C: p < 0.001 (simple), p = 0.03533 (complex); D: p = 0.6007 (simple), p < 0.001 (complex)]. 3′UTR structures do not show correlation (Extended Data Fig. 2-1). E, F, Differential icSHAPE reactivity profiles between TE up and TE down genes. The same TE up and TE down genes with z score ±1.5 were used; (E) GS TG (TE up: 687, TE down: 335), (F) LRRK2 KO (TE up: 596, TE down: 576). icSHAPE data from mouse ES cells were extracted (Spitale et al., 2015), and a window of −100–0 nt 5′ of start codon (CDS start) was used. Average icSHAPE reactivity values: all genes: 0.236, TE up (GS): 0.229, TE down (GS): 0.240, TE up (KO): 0.237, TE down (KO): 0.219. Statistical significance (compared with all genes) was measured by non-parametric Mann–Whitney test. Error bars indicate SEM, *p < 0.05, **p < 0.01, ***p < 0.001.
- Figure 3.
G2019S LRRK2 increases mRNA translation independent of initiation factors. A, B, Western blotting and quantification of T136 S15 phosphorylation in the mouse brain. LRRK2 KO (A) and G2019S LRRK2 transgenic (B) mice. Whole-brain lysate was used. n = 3, biological replicates. Statistical significance was determined by (A) unpaired t test (B) one-way ANOVA with Bonferroni correction. C, D, Schematics of HCV-IRES and CrPV-IRES reporters. E–G, HCV-IRES reporter assays. C, n = 4; D, n = 3, respectively. H–J, CrPV IRES reporter assays. F, n = 4; G, n = 3, respectively. Reporter assays were performed in primary mouse cortical neurons with transient transfection, and each experiment is an average of triplicates. All values were divided by the average of control values. Reporter mRNA levels were controlled (Extended Data Fig. 3-1). WT, wild type; Fluc, firefly luciferase; RLuc, Renilla luciferase; RLU, relative light units. Statistical significance was determined by one-way ANOVA with Bonferroni correction. Error bars indicate SEM, *p < 0.05, **p < 0.01, ***p < 0.001, ns = no significance.
- Figure 4.
Ribosome footprint distributions on Atf4 uORFs in the LRRK2 KO brain. A, B, Ribosome footprints distribution in the 5′UTR of Atf4 gene (visualized: chr15:80,086,569–80,086,862). Red box indicates the region that ribosomes are depleted in the LRRK2 KO brain. C, RNA structure prediction of the Atf4 uORF sequences by ViennaRNA RNAfold (Lorenz et al., 2011). The regions of depleted ribosome footprints have high probability to form secondary structure. In addition, relationship between G2019S LRRK2 and eIF2ɑ was addressed (Extended Data Fig. 4-1).
- Figure 5.
Calcium currents recorded in SNpc DA neurons. A, Comparison of spontaneous AP firing pattern of DA neurons between wild-type and GS LRRK2 mouse slices. B, Calcium currents were measured in mouse SNpc DA neurons using whole-cell patch clamp recordings. C, Quantification of calcium peak currents. Data are expressed as means ± SEM, n = 12 slices from 12 animals for each group. Intrinsic properties were measured (Extended Data Fig. 5-1).
Tables
Sample Type Mapped reads Mouse control 1 Ribo 11,528,964 mRNA 35,167,826 Mouse control 2 Ribo 14,236,436 mRNA 53,086,785 Mouse control 3 Ribo 28,913,143 mRNA 37,446,599 G2019S TG 1 Ribo 10,832,574 mRNA 37,329,310 G2019S TG 2 Ribo 11,636,313 mRNA 42,813,645 G2019S/D1994A TG 1 Ribo 11,391,779 mRNA 62,883,025 G2019S/D1994A TG 2 Ribo 30,967,177 mRNA 33,887,256 Mouse WT (vs KO) 1 Ribo 6,069,632 mRNA 44,668,380 Mouse WT (vs KO) 2 Ribo 6,331,204 mRNA 8,025,070 LRRK2 KO 1 Ribo 5,552,451 mRNA 65,571,861 LRRK2 KO 2 Ribo 7,322,616 mRNA 18,455,674 LRRK2 WT 3 (STR) Ribo 22,256,190 LRRK2 WT 3 (VMB) Ribo 8,773,494 LRRK2 KO 3 (STR) Ribo 22,069,910 LRRK2 KO 3 (VMB) Ribo 6,492,191 Ribo: ribosome profiling; mRNA: RNA-Seq; TG: transgenic mice; WT: wild type, KO: knock-out mice; STR: striatum; VMB: ventral midbrain.
Extended Data Figure 1-1
Triple periodicity of ribosome profiling data. A, B, Triplet periodicity of ribosome profiling datasets were visualized to ensure the quality of the libraries. Transcript coordinates were re-aligned based on the rounded half point of the ribosome footprint [5′ end + (footprint length/2)]. Conserved triplet periodicity indicates that the libraries are faithfully representing translating ribosomes, ensuring the quality of the RPF libraries. There was no significant change found in ribosome footprint length, periodicity, and distribution in any LRRK2 mouse models (data not shown). Download Figure 1-1, TIF file.
Extended Data Figure 1-2
Ribosome profiling expression analysis results. Download Figure 1-2, XLS file.
Extended Data Figure 2-1
3′UTR secondary structure is not related to translational effects of G2019S LRRK2. A, B, 3′UTR secondary structure folding energy differences between TE up and TE down genes (standard z score ±1.5 was used). Unlike the 5′UTR folding energy comparison, 3′UTR folding energy did not show opposing directions of effects between G2019S (GS) LRRK2 transgenic (TG) and LRRK2 KO mice. Statistical significance was determined using Wilcoxon signed-rank test [A, p = 0.002338 (TE up), p = 0.02327 (TE down); B, p = 0.01194 (TE up), p = 0.0254 (TE down)]. C, D, TE differences of 5′ TOP mRNAs in LRRK2 mouse models. Wilcoxon signed-rank test; C, p = 0.2112; D, p = 0.09034. background signal. Error bars indicate SEM, *p < 0.05, **p < 0.01, ns = no significance. Download Figure 2-1, TIF file.
Extended Data Figure 3-1
Reporter transcript levels for IRES reporter assays. A, B, qPCR measurement of luciferase transcript levels in IRES reporter assays. One-way ANOVA with Bonferroni correction was used, and there were no significant changes in the reporter transcript levels detected; **p < 0.01, ***p < 0.001, ns = no significance. Download Figure 3-1, TIF file.
Extended Data Figure 4-1
Delayed ISR recovery from G2019S LRRK2-expressing neurons. A, Phosphorylation of eIF2ɑ in the G2019S LRRK2 transgenic brains. Dissected striatal tissues, age three to four months, n = 3, biological replicates. B, Mouse cortical neurons were prepared from pregnant transgenic breeders at E15. Pups were separated and individually genotyped. Control: wild type or single transgenic (CaMKII-tTA or tet-G2019S LRRK2), G2019S LRRK2: double transgenic. Tg: thapsigargin (1 μm). *, background signal. Download Figure 4-1, TIF file.
Extended Data Figure 5-1
Intrinsic properties of mouse brain DA neurons. A, Summary of electrophysiological characteristics of DA neurons in SNpc during recordings, including pipette resistance (Rp), input resistance (Rin), series resistance (Rseries), leak currents (Leak), and resting membrane potential. B, Spontaneous AP firing pattern in DA neurons. C, A representative single AP wave with a half width of 2 ms. D, Evoked APs. The presence of a sag (arrow) in the membrane potential and APs were detected in current-clamp immediately after rupturing the membrane. E, Immunofluorescence image showing recorded neurons are TH-positive. Alexa Fluor 568 was injected to label recorded neurons. Scale bar: 50 μm. Data are expressed as means ± SEM, WT, n = 6 slices from 6 mice; GS LRRK2, n = 6 slices from 6 mice. Download Figure 5-1, TIF file.
In this issue
