(A, B) Immunostaining in spinal cord sections of a 1-month old non-transgenic (NT) and symptomatic homozygous TDP-43_PrP mice using an antibody to hTDP-43 shows hTDP-43 in nuclei of TDP-43_PrP mice (B & B′), with occasional cytoplasmic staining (arrows). hTDP-43 was not observed in NT mice (A & A′). (C–F) IHC analysis of spinal cord (C,D & E) or cortical (F) neurons using an antibody for the detection of TDP-43 phosphorylated at serines 403/404 and hematoxylin (C) or eosin (D,E & F) counterstain. Shown in panel C are nuclear bodies immunoreactive for pTDP-43 within a spinal cord motor neuron, while cytoplasmic pTDP-43-immunoreactive inclusions are shown in panels D,E and F. Scale bars in panels A and B represents 100 μm; in panels A′ and B′, bars = 50 μm; and in panels C–F, bars = 10 μm.

Silver staining with a method specific for neurodegeneration of neurites and neuronal cell bodies revealed argyrophilic degenerating neurites and neurons in spinal cord of symptomatic TDP-43_PrP mice (B) compared to NT controls (A). Toluidine blue stains show myelin vacuolization, with myelin ovoids (arrows and inset) in anterolateral funiculi of spinal cords of symptomatic TDP-43_PrP mice (C, D) but not in NT mice (E, F). Scale bars in panels A and B represent 20 μm, bars within panels of C and E are 40 μm, and bars in panels D and F are 20 μm.

(A–B) Immunohistochemistry shows hTDP-43 distributed throughout the gray matter of the spinal cord (A) and brain (B) in homozygous TDP-43_PrP mice. (C) Western blots of brain lysates from non-transgenic (NT), as well as hemizygous and homozygous TDP-43_PrP mice using antibodies that detect either both mouse and human TDP-43 (total TDP-43) or human TDP-43 (hTDP-43) only. GAPDH immunoreactivity was included to ensure equal loading. (D) Compared to NT and hemizygous mice, homozygous TDP-43_PrP mice had significant deficits in body weight. By 28 days, the average body weight of homozygous TDP-43_PrP mice was approximately half that of controls. Data shown is the mean ± SEM of 8 mice per group, *p<0.05, ***p<0.001, as assessed by Two-Way ANOVA followed by Bonferroni post-test. (E) At 1 month, brain weight of homozygous TDP-43_PrP mice was significantly lower than that of age-matched NT and hemizygous mice. Data shown is the mean of 7–14 mice per group. ***p<0.001, as assessed by One-Way ANOVA. (F–G) Upon tail elevation, homozygous mice (G) held their hindlimbs close to their body and failed to show proper escape extension while NT mice (F) showed normal escape response by splaying their hindlimbs. (H) Immunoblot of spinal cord lysates from NT, hemizygous and homozygous TDP-43_PrP mice using an antibody that detects total TDP-43. Note that TDP-43 fragments (~35 kDa and 25 kDa) in spinal cord co-migrate with C-terminal fragments generated by staurosporine (STP)-induced caspase activation in human neuroglioma cells expressing hTDP-43. NT= non-transgenic, hemi= hemizygous, homo=homozygous.

(A, F) Hematoxylin and eosin staining in spinal cord sections of 1 month old non-transgenic (NT) and symptomatic homozygous TDP-43 _PrP mice. Eosinophilic aggregates in spinal cord motor neurons from TDP-43_PrP mice (F, arrows and insert) are not observed in NT mice (A). (B, G) IHC analysis of spinal cord neurons in TDP-43_PrP mice (G) and NT mice (B) using an antibody for the detection of TDP-43 phosphorylated at serines 403/404 and eosin counterstain. (C, H) Abnormal ubiquitin immunoreactivity was present in the cytoplasm and nucleus of neurons in TDP-43_PrP mice (H) but not in NT mice (C). Enhanced IBA-1 (I, D) and GFAP (J, E) immunoreactivity indicative of activated microglia and reactive astrogliosis, respectively, were observed in TDP-43_PrP (I, J), but not NT mice (D, E). Scale bars in panels A, C–F, H–J represent 50 μm, bars within inserts of panels A and F are 25 μm, and bars in panels B and G are 10 μm.

The upper left inset of panel A depicts a low power image of a motor neuron in the anterior horn of a 1-month old symptomatic homozygous TDP-43_PrP mouse containing a cytoplasmic aggregate (arrow) and a peripherally located nucleus (N; Bar represents 5μm). Enlargement of the aggregate (panel A, proper) reveals clustered mitochondria (Bar represents 2μm). A large, abnormal mitochondrion with disorganized inner cristae (arrow) is shown in the right upper inset. Also observed are mitochondria with paucity of cristae and vacuoles within the mitochondrial matrix (arrowheads; Bar represents 0.2 μm). (B) The accumulation of mitochondria of various shapes and sizes, and small and large vesicles is observed in a swollen dendrite (Bar represents 1 μm). (C) Abnormally shaped (arrowheads) and degenerated (arrow) mitochondria, as well as autophagic vacuoles (*) are present within a swollen axon of a spinal cord neuron (Bar represents 0.25 μm). (D) An axon with a vacuolated myelin sheath (*), containing degenerating mitochondria (arrowheads), many vesicles/vacuoles and tightly packed neurofilaments (arrow) is shown (Bar represents 2 μm). (E) Immunoblot analysis of mitofusin 1 (MFN1), Fis1, DLP1 and Ser616-phosphorylated DLP1 expression in brain lysates of non-transgenic, hemizygous and homozygous TDP-43_PrP mice. Densitometric analysis of Western blots is shown. While total DLP1 levels did not change, phosphorylation of DLP1 at Ser616 was significantly increased in homozygous TDP-43_PrP mice compared to non-transgenic mice. Similarly, expression of Fis1, another component of the fission machinery was significantly upregulated in TDP-43_PrP mice. In contrast, mitofusin 1 (MFN1) expression was significantly decreased in TDP-43_PrP mice. * p < 0.05 compared to levels in non-transgenic mice, as assessed by one-way ANOVA (n=3). NT= non-transgenic, hemi=hemizygous and homo=homozygous. (F, J) Following IHC against the mitochondrial marker, COX-IV, spinal cord sections were counterstained with eosin (G, K). Notice the COX-IV-positive aggregates, which are also eosinophilic, in TDP-43_PrP mice (J, K) but not non-transgenic mice (F, G). Likewise, COX-IV-positive aggregates (L) stained blue following staining with toluidine blue (M) in TDP-43_PrP mice, supporting the presence of high phospholipid levels associated with mitochondria. No similar staining was observed in non-transgenic mice (H, I). Scale bar represent 24 μm.