C9orf72 BAC Transgenic Mice Display Typical Pathologic Features of ALS/FTD

Abstract

Noncoding expansions of a hexanucleotide repeat (GGGGCC) in the C9orf72 gene are the most common cause of familial amyotrophic lateral sclerosis and frontotemporal dementia. Here we report transgenic mice carrying a bacterial artificial chromosome (BAC) containing the full human C9orf72 gene with either a normal allele (15 repeats) or disease-associated expansion (∼100-1,000 repeats; C9-BACexp). C9-BACexp mice displayed pathologic features seen in C9orf72 expansion patients, including widespread RNA foci and repeat-associated non-ATG (RAN) translated dipeptides, which were suppressed by antisense oligonucleotides targeting human C9orf72. Nucleolin distribution was altered, supporting that either C9orf72 transcripts or RAN dipeptides promote nucleolar dysfunction. Despite early and widespread production of RNA foci and RAN dipeptides in C9-BACexp mice, behavioral abnormalities and neurodegeneration were not observed even at advanced ages, supporting the hypothesis that RNA foci and RAN dipeptides occur presymptomatically and are not sufficient to drive neurodegeneration in mice at levels seen in patients.

Figure 1. Generation of C9orf72 BAC transgenic mice

(A) Schematic diagram of C9-BAC construct. The pCC1 BAC backbone (8.128 kb) was used to isolate a region contained the entire human C9orf72 gene (~36 kb) from a patient with C9-ALS, including ~110 kb upstream and ~20 kb downstream. >30 different C9-BAC founder lines were generated with either the contracted control BAC (15 repeats) or expansions of various sizes. (B) Southern blot analysis of the repeat length in C9-BAC founder lines. A contracted subclone was used to generate the control (F08-CTR) line with 15 GGGGCC repeats. Lines F08-CTR, F112 and F113 showed comparable expression, and lines F112 and F113 had a variety of repeat sizes ranging from ~100-1000. (C) Relative expression of human C9orf72 mRNA in cortex from C9-BAC transgenic lines measured by qPCR. (D) Western blot analysis of C9orf72 protein (48kD) expressed in C9-BAC mouse cortex (left). Anti-C9orf72 (top), and anti-Actin (bottom). Western blot quantitation (right) normalized to nontransgenic (NTg) control. (E) Sequence alignments of exon 1a and exon 1b utilization from 5’RACE analysis of C9orf72 transcripts. iPSC-derived motor neurons are shown on the left (from BAC donor patient) to demonstrate exon 1a and 1b utilization in human cells. (F) C9orf72 promoter methylation assay as assessed by MSRE-qPCR. Fibroblasts (Fibro), induced pluripotent stem cells (iPSC) and iPSC-derived motor neurons (MN) are shown for BAC donor patient (~800 repeats), which revealed variable methylation from ~10% in fibroblasts to ~50% in iPSC and MNs. Methylation levels were low (~5%) in C9-BAC transgenic lines. Lymphoblastoid cell lines (LCL) from a non-expanded patient (CTR) and C9orf72 expansion carrier (C9+) are shown for comparison. *All data are shown as mean ± SEM. See also Figure S1.

Figure 2. C9-BAC transgenic mice develop RNA foci and RAN peptides similar to human C9orf72 expansion carriers

(A) Fluorescence in situ hybridization (FISH) of sense and antisense RNA foci in F08-CTR (red box) and F112 C9-BAC mice. CER-GL: cerebellar granular layer; CER-PC: cerebellar purkinje cells, hippo-DG: hippocampus dentate gyrus; FC: frontal cortex; PMC: primary motor cortex; SC-MN: spinal cord motor neuron. Scale bar = 10μm. (B) Quantification of sense and antisense foci in nontransgenic (NTg), F08-CTR and F112 C9-BACexp transgenic mice across brain regions. Lower graph shows the percent of cells with foci across different brain regions at 3, 6, and 8 months of age. (C) Immunostaining of poly(GP) in 20 month old NTg or F112 C9-BACexp mice showing inclusions in brain regions including cortex, hippocampus dentate gyrus (Hippo-DG), and cerebellar granular layer. (D) Soluble poly(GP) is present in the cortex of 6 month-old C9-BACexp transgenic mice (F112 and F113), but not in NTg mice, nor F08-CTR mice, as assessed using a poly(GP) immunoassay. (n=2 for F112, F113, NTG, and F08-CTR) (E) Insoluble poly(GP) levels are lower than soluble poly(GP) levels (C) in the cortex of C9-BACexp transgenic mice, as determined by immunoassay of the insoluble fraction. (n=2 for F112, F113, NTG, and F08-CTR) (F) Immunoassay of Poly(GP) levels in different brain regions. Cerebellum showed the highest levels, while levels were lowest in the spinal cord. (n=3) (G) Comparison of poly(GP) levels in the cortex from C9-BAC mouse lines and C9orf72-positive FTLD cases. Lines F112 and F113 produce poly(GP) at similar levels to those detected in two different human C9orf72+ FTD frontal cortex samples (FTD #1 and FTD #2). *All data are shown as mean ± SEM. See also Figure S2.

Figure 3. C9-BAC transgenic mice do not show evidence of neurodegeneration or functional deficits in the nervous system

(A) Body weight and (B) grip strength normalized to body weight at 18 months of age. For this and all behavioral studies below, male mice (NTg, n=8; F112, n=10) were used unless otherwise indicated. (C) Rotarod testing of cohorts of nontransgenic (NTg) and C9-BACexp mice (F112). No differences in sensorimotor coordination, or motor learning across trials, were observed in either young (3 months; NTg, n=8; F112, n=8) or aged (18 months; NTg, n=8; F112, n=10) animals. (D) Open field test at 18 months comparing total activity (left) and time spent in the center (right) for NTg mice and C9-BACexp (F112) mice. (E) 3 chamber test for sociability and social novelty at 16 months. F112 mice prefered spending time with the stranger and the novel object more than in the neutral zone, no different from NTg mice. (F) Y-Maze assessment of memory and novelty seeking at 18 months was not different between C9-BACexp mice and NTg controls. (G) Neuromuscular junction staining (bungarotoxin = red, synaptophysin/neurofilament = green) of the tibialis anterior muscle in NTg and C9-BACexp (F112) mice at 20 months showed no differences in occupancy, fragmentation or area (n=6). (H) Femoral motor and sensory nerve plastic sections from NTg, F08-CTR, and C9-BACexp (F112) mice. (I) Axon counts of femoral motor and sensory nerves (n=6) in C9-BAC (F112) and control mice (NTg, F08-CTR) at 20 months of age. See also Figure S2.

Figure 4. C9-BACexp mice show evidence of nucleolar stress, and decreased RNA foci and DPR proteins after antisense oligonucleotide treatment

(A) Immunofluorescence staining of nucleolin (red) costained with DAPI (blue) in nontransgenic (NTg) and C9-BACexp (F112) mice in the cerebellar purkinje cells (CER-PC), frontal cortex (FC), and primary motor cortex (PMC). C9-BACexp mice showed partial displacement of nucleolin from the nucleolus. (scale bar = 20 μm) (B) Quantitation of nucleolin staining distribution revealed a shift in the ratio of the amount of staining in the nucleus vs. the nucleolus (n=4 animals; two sided t-test, NTg vs. F112; *p<0.05; **p<0.01). (C) RNA FISH of sense foci in primary cortical cultures from neonatal NTg and F113 C9-BACexp mice. RNA foci were observed in 73% (±9%) of neurons and astrocytes. (D) Representative calcium recording from culture cortical neurons (NTg, top and F112, bottom). (E) Frequency of calcium transients observed quantitated using Fluo-4 live imaging (n=27, NTg and n=68, F112). (F) Glutamate toxicity assay in primary cortical neuron cultures showing Tuj1 staining after 24 hours of glutamate treatment at the indicated dose. (scale bar = 400 μm) (G) The number of neurons (TuJ1 positive) were divided by total number of cells (DAPI, not shown) and then normalized to the controls (no treatment) for both F113 and NTg. (H) RNA FISH of sense foci in primary cortical cultures treated with either control antisense oligonucleotides (Scr ASO) or an ASO targeting exon 2 of the human C9orf72 transcript (C9 ASO). (I) Quantification of RNA foci reduction after C9 ASO treatment (**p=0.004, two sided t-test). (J) Immunoassay showing reduction of poly(GP) DPR proteins in cortical cultures from C9-BACexp (F112 line) mice treated with C9 ASO compared to Scr ASO (**p=0.004,one way ANOVA). See also Figure S3.