Genetic ablation of dynactin p150^Glued in postnatal neurons causes preferential degeneration of spinal motor neurons in aged mice

Genomic DNA fragments containing Dctn1 gene locus were isolated from a mouse genomic DNA library (Stratagene). A 9.3 kb KpnI/AflII fragment carrying exons 2–4 of Dctn1 was subcloned into the pBluescript vector for later modifications. To construct the targeting vector, the Dctn1 clone was modified by inserting the first loxP site and a neomycin (Neo) selection cassette flanked with two Frt sites in intron 1, and the second loxP site in intron 4 (Fig. 1a). The targeting vector was linearized at a unique NotI site and transfected into 129/SvJ ES cells. After neomycin resistance (Neo) positive selection with G418 for 7 days, ES clones that had undergone homologous recombination were picked and screened by PCR, Southern blot analysis and partial genome sequencing (Fig. 1b). Two positive ES clones were expanded and injected into blastocysts. The resulting male chimera mice were bred with wild-type C57BL/6 J female mice to obtain Dctn1^{LoxP-Frt-Neo-Frt} mice. The Dctn1^{LoxP-Frt-Neo-Frt} mice were then crossbred with FLPe knock-in mice [15] (JAX, stock number: 003946) to remove Neo selection marker and get Dctn1^LoxP/ mice. Finally, the Dctn1^LoxP/ mice were crossed with Thy1-Cre [16] (JAX, Stock Number: 006143) or Cre/Esr1 [17] (JAX, Stock Number: 004682) mice to delete exons 2 to 4 and thereby the MTBD-containing p150^Glued from the Cre-expressing cells. The mice were housed in a 12 h light/dark cycle and fed regular diet ad libitum. All mouse work followed the guidelines approved by the Institutional Animal Care and Use Committees of the National Institute on Aging, NIH.

Genomic DNA was prepared from tail biopsy using DirectPCR Lysis Reagent (Viagen Biotech) and subjected to PCR amplification using specific sets of PCR primers for each genotype, including Dctn1^LoxP/ knock-in mice (mDCTN1-Ex4-F, CAG CTG CAA AGA CCA GCA AA; mDCTN1-Ex5-R: CAC ACC ACC TTC TTA GGC TTC A), Cre transgenic mice (CRE-F, CAT TTG GGC CAG CTA AAC AT; CRE-R, TGC ATG ATC TCC GGT ATT GA).

Neuronal cells or neural tissues were homogenized in 50 mM Tris-HCl, pH 7.5, containing 150 mM NaC1, 1 mM EDTA and Protease Inhibitor Cocktail (Thermo Scientific). The cytosolic extracts were prepared by centrifugation at 16,000 × g for 15 min at 4 °C, and further clarified at 100,000 × g for 15 min. The resulting supernatant was layered over a 12 ml, 5–20% sucrose density gradient and centrifuged at 120,000 × g for 18 h at 4 °C. 1-ml gradient fractions were collected and equal volumes of all fractions analyzed by SDS-PAGE followed by Western blotting.

Open-field test. As described previously [18], the ambulatory and rearing activities of male mice were measured by the Flex-Field Activity System (San Diego Instruments). Flex-Field software was used to trace and quantify mouse movement in the unit as the number of beam breaks per 30 min.

Rotarod test. As described previously [19], male mice were placed onto a rotating rod with auto-acceleration from 0 rpm to 40 rpm for 1 min (San Diego Instruments). The length of time the mouse stayed on the rotating rod was recorded. Three measurements were taken for each animal during each test.

Hindlimb clasping test. In the hindlimb clasping test, male mice were suspended by the tail and observed for 10 s. The severity of hindlimb clasping phenotype was scored as follows: 0 = both hindlimbs consistently spread outward and away from the abdomen, 1 = one hindlimb retracted toward the abdomen, 2 = both hindlimbs partially retracted toward the abdomen, 3 = both hindlimbs completely retracted toward the abdomen. For each mouse, three separate tests were taken over 3 h, and the averaged hindlimb clasping severity score of the three tests was calculated.

For neuropathology study, mice were perfused with 4% paraformaldehyde (PFA) in cold phosphate buffered saline (PBS), brains and spinal cords were collected, post-fixed overnight, submerged in 30% sucrose in PBS for at least 72 h, and sectioned at 40–50 μm thickness using CM1950 cryostat (Leica). For muscle pathology analysis, mice were euthanized, gastrocnemius muscles were dissected, snap frozen in 2-methylbutane (Sigma-Aldrich) cooled in dry ice, sectioned at 20 or 40 μm thickness, and fixed in 4% PFA in PBS for 15 min. For Nissl staining, sections were stained with 0.1% Cresyl Violet (Sigma-Aldrich). For HE staining, hematoxylin and eosin stain kit (Vector Laboratories) were used as suggested by manufacturers. For immunostaining, antibodies specific to p150^Glued N-terminus (amino acid 3–202, 1:200, BD Biosciences), p150^Glued C-terminus (amino acid 1266–1278, 1:500, Abcam), neuron-specific nuclear protein (NeuN, 1:1000, Millipore), CTIP2 (1:200, Abcam), choline acetyltransferase (CHAT, 1:500, Millipore), Glial Fibrillary Acidic Protein (GFAP, 1:1000, Abcam), Ionized calcium-Binding Adaptor Molecule-1 (IBA1, 1:1000, Wako Chemicals), acetylated α-tubulin (ac-TUBA, 1:500, Abcam), Lysosomal Associated Membrane Protein 2 (LAMP2, 1:200, Abcam), Tyrosine Hydroxylase (TH, 1:500, Pel-Freeze) and Calbindin D28K (1:200, Abcam) were used as suggested by manufacturers. For immunofluorescence study, Alexa Fluor 488-, 546- or 647-conjugated secondary antibody (1:500, Invitrogen) was used to visualize the staining. Alexa Fluor 488-conjugated α-bungarotoxin (1:500, Invitrogen) was applied to label postsynaptic acetylcholine receptors of NMJs. Fluorescence images were captured using LSM 780 laser-scanning confocal microscope (Zeiss). The paired images in all the figures were collected at the same gain and offset settings. Post collection processing was applied uniformly to all paired images. The images were presented as either a single optic layer after acquisition in z-series stack scans at 1.0 μm intervals from individual fields or displayed as maximum-intensity projections to represent confocal stacks. For immunohistochemistry study, Vectastain Elite ABC Kit and DAB Kit (Vector Laboratories) were used to visualize the staining. Bright field images were captured by Axio microscope Imager A1 (Zeiss).

For the quantitative assessment of various marker protein distributions, images were taken using identical settings and exported to ImageJ (NIH) for imaging analysis. Images were converted to an 8-bit color scale (fluorescence intensity from 0 to 255) using ImageJ. The areas of interest were first selected with Polygon or Freehand selection tools and then subjected to measurement by mean optical intensities or area fractions. The mean intensity for the background area was subtracted from the selected area to determine the net mean intensity.

Stereology for CSMNs, striatal neurons, SMNs and midbrain dopaminergic neurons was performed as described previously [19, 20]. To examine the number of CSMNs and striatal neurons, series of sagittal brain sections (40 μm per section thickness, every ninth section, nine sections per case) were processed for Nissl staining or CTIP2 immunohistochemistry. The motor cortex was designated as layer V neurons in the anteromedial cortex, and outlined according to the mouse brain in stereotaxic coordinates. To examine the number of SMNs, series of coronal spinal cord sections (50 μm per section, every tenth section from L3 to L5, ten sections per case) were processed for Nissl staining or CHAT immunohistochemistry. To examine the number of midbrain dopaminergic neurons, series of coronal sections across the midbrain (40 μm per section, every fourth section from Bregma − 2.54 to − 4.24 mm, ten sections per case) were processed for TH immunohistochemistry. For the unbiased stereological estimation of CSMNs, striatal neurons, SMNs and midbrain dopaminergic neurons, the number of CTIP2 positive or Nissl-stained large (>15 μm in diameter) cortical neurons in layer V, Nissl-stained neurons in stratum, CHAT positive or Nissl-stained large (>35 μm in diameter) spinal neurons in ventral horn, and TH-positive neurons in substantia nigra pars compacta (SNpc) were assessed using the Optical Fractionator function of Stereo Investigator 10 (MicroBrightField). Six mice were used per genotype at each time point. Counters were blinded to the genotypes of the samples. The sampling scheme was designed to have a coefficient of error less than 10% in order to obtain reliable results.

PSD fractions were prepared from mouse brains as described previously [18]. All procedures were performed at 4 °C. Mouse brains were isolated and homogenized in 10 volumes of ice-cold Buffer A (0.32 M sucrose, 5 mM HEPES, pH 7.4, 1 mM MgCl₂, 0.5 mM CaCl₂, Protease and Phosphatase Inhibitor Cocktail) with Teflon homogenizer (12 strokes). The homogenate was spun at 1400 × g for 10 min. Supernatant (S1’) was saved and pellet (P1’) was homogenized again with Teflon homogenizer in another 10 volumes of Buffer A (5 strokes). After centrifugation at 700 × g for 10 min, the supernatant (S1’) was collected and pooled with S1, and the pellet (P1’) were saved as crude nuclear fraction. Pooled S1 and S1’ were centrifuged at 13,800 × g for 10 min to collect the crude synaptosomal pellet (P2) and the supernatant (S2), which contains the cytosol and light membranes. S2 was centrifuged at 100,000 × g for 15 min to separate the cytosolic fraction (Cyt) and the light membrane fraction (LM). P2 was resuspended in 10 volumes of Buffer B (0.32 M sucrose, 6 mM Tris, pH 8.0, supplemented with Protease and Phosphatase Inhibitor Cocktail) with Teflon homogenizer (5 strokes). The P2 suspension was loaded onto a discontinuous sucrose gradient (0.85 M/1 M/1.15 M sucrose solution in 6 mM Tris, pH 8.0), and centrifuged at 82,500 × g for 2 h. Synaptosome fraction (Syn) was collected from the layer between 1 M and 1.15 M sucrose, supplemented with 0.5% Triton X-100 and mixed for 15 min. The suspension was spun at 32,800 × g for 20 min, and the resulting pellet was saved as PSD fraction (PSD). P1’, LM and PSD fractions were lysed in 1% SDS buffer. SDS were added into Cyt and Syn fraction to 1% final concentration. Equal amount of total protein from each fraction were resolved in SDS-PAGE and applied to western blot analysis.

Mouse primary neuron cultures were prepared from the cortex or spinal cord of newborn (postnatal day 0, P0) pups, as described previously [21, 22]. Briefly, individual cortex or spinal cord was dissected and subjected to papain digestion (5 U/ml, Worthington Biochemicals) for 40 min at 37 °C. The digested tissue was carefully triturated into single cells using increasingly smaller pipette tips. The cells were then centrifuged at 250 × g for 5 min and resuspended in warm Basal Medium Eagle (BME, Sigma-Aldrich) supplemented with 5% heat-inactivated fetal bovine serum (FBS; Invitrogen), 1× N2/B27 supplement (100× stock, Invitrogen), 1× GlutaMax (100× stock, Invitrogen), 0.45% D-glucose (Sigma-Aldrich), 10 U/ml penicillin (Invitrogen), and 10 μg/ml streptomycin (Invitrogen). The dissociated neurons were seeded in plated in Biocoat Poly-D-Lysine Cellware plate (BD Biosciences), and maintained at 37 °C in a 95% O₂ and 5% CO₂ humidified incubator. Twenty-four hours after seeding, the cultures were switched to serum-free medium supplemented with 1 μM cytosine β-D-arabinofuranoside (Sigma-Aldrich) to suppress the proliferation of glia and 1 μM 4-hydroxytamoxifen (4-OHT, Sigma-Aldrich) to induce CRE recombinase activity. Starting from 5 days in vitro (DIV), culture medium was changed twice every week.

Spinal neurons were cultured at the density of 1.2 × 10⁶ per well in Poly-D-Lysine coated 6-Well plate (BD Biosciences) and treated with 0 or 10 μM glutamate (Sigma-Aldrich) on 20 DIV. After 24-h treatment, neurons were washed with PBS containing 0.1 mM CaCl₂ and 1 mM MgCl₂ (PBS/CM), and then incubated with 1 ml biotin solution (0.5 mg/ml Sulfo-NHS-SS-Biotin in cold PBS/CM, Thermo Fisher Scientific) for 20 min at 4 °C and then washed subsequently with PBS/CM, 0.1 M glycine solution, and TBS (25 mM Tris-HCl, pH 7.4, containing 137 mM NaCl) buffer. Cells were harvested in 250 μl RIPA buffer (Sigma) supplemented with protease inhibitor and phosphatase inhibitor cocktail, lysed on ice for 30 min, and centrifuged at 16,000 × g for 15 min at 4 °C. Resulting supernatant containing equal amount of total protein was incubated with 50 μl neutral-avidin agarose (Pierce) at 4 °C for 2 h with gently rotating. After washing in RIPA buffer for 5 times, the biotin-labeled surface protein was eluted with SDS sample buffer by heating at 70 °C for 10 min. Total proteins and isolated biotinylated surface proteins were analyzed by western blotting.

Spinal neurons were cultured at the density of 1.0 × 10⁵ per well in Poly-D-Lysine coated 96-Well plate (BD Biosciences) and treated with 0 or 10 μM glutamate (Sigma-Aldrich) on 20 DIV. After 24-h treatment, the cells were incubated with 0.5 mg/ml MTT (Sigma-Aldrich) at 37 °C for 4 h. After the media were removed, dimethyl sulfoxide (DMSO, 100 μl) was added to each well to solubilize the formazan crystals generated by viable mitochondrial succinate dehydrogenase from MTT. The absorbance at 570 nm was measured using a SpectraMax M5 Multi-Mode Microplate Reader (Molecular Devices) as the MTT reducing activity of the cells.

Neurons or tissues were homogenized and sonicated in RIPA buffer (Sigma-Aldrich) or 1% SDS lysis buffer (50 mM Tris–HCl, 150 mM NaCl, 2 mM EDTA, pH 7.5, and 1% SDS) supplemented with Protease Inhibitor Cocktails (Thermo Scientific). Lysates were clarified by centrifugation at 15000 × g for 15 min at 4 °C. The supernatants were quantified for protein content using the bicinchoninic acid (BCA) assay kit (Thermo Fisher Scientific) and separated by 4–12% NuPage BisTris-PAGE (Invitrogen) using MES or MOPS running buffer (Invitrogen). The separated proteins were then transferred to nitrocellulose membranes using the iBlot Dry Blotting system (Invitrogen) and incubated with specific primary antibodies. The antibodies used for western blot analysis included p150^Glued N-terminus (amino acid 3–202, 1:1000, BD Biosciences), p150^Glued C-terminus (amino acid 1266–1278, 1:1000, Abcam), dynactin subunit p62 (1:1000, Abcam), dynactin subunit p50 (1:5000, BD Biosciences), dynactin subunit Actin Related Protein 1 (ARP1, 1:1000, Sigma-Aldrich), dynein heavy chain (DHC, 1:500, Santa Cruz Biotechnology), dynein intermediate chain (DIC, 1:1000, Sigma-Aldrich), dynein light chain (DLC, 1:500, Santa Cruz Biotechnology), α-tubulin (TUBA, 1:10000, Abcam), acetylated α-tubulin (ac-TUBA, 1:10000, Abcam), tyrosinated α-tubulin (tyro-TUBA, 1:10000, Abcam), detyrosinated α-tubulin (detyro-TUBA, 1:10000, Abcam), Autophagy Related Protein ATG3 (ATG3, 1:1000, Cell Signaling), ATG5 (1:1000, Cell Signaling), ATG7 (1:1000, Cell Signaling), autophagy related protein LC3 (1:1000, Abcam), Lysosomal Associated Membrane Protein 1 (LAMP1, 1:1000, BD Biosciences), LAMP2 (1:1000, Abcam), synaptophysin (SYP, 1:10000, Millipore), Postsynaptic Density Protein 95 (PSD95, 1:1000, Sigma-Aldrich), AMPA receptor subunit GLUR1 (1:1000, Abcam), GLUR2 (1:1000, BD Biosciences), NMDA receptor subunit NR2A (1:250, BD Biosciences), NR2B ((1:500, BD Biosciences), and β-actin (1:5000, Sigma-Aldrich). Protein signals were visualized by IRDye secondary antibodies and Odyssey system (LI-COR Biosciences), and quantified with NIH ImageJ software.

Statistical analysis was performed using GraphPad Prism 5 (GraphPad Software). Data were presented as mean ± SEM. Statistical significance was determined by comparing means of different groups using unpaired t-test, one-way or two-way ANOVA followed by the post hoc test. ^*p < 0.05, ^**p < 0.01, ^***p < 0.001, ^****p < 0.0001.

To selectively disrupt the expression of p150^Glued, we initially generated Dctn1^LoxP knock-in mice by inserting two LoxP sites in the first and fourth introns of mouse Dctn1 gene locus, respectively, resulting in a complete loss CAP-Gly domain and partial loss of basic domain (Fig. 1a-c). The Dctn1^LoxP knock-in mice were then crossbred with Cre/Esr1 mice to generate Dctn1^LoxP/LoxP; Cre/Esr1 pups and littermate controls for neuronal cultures. The administration of tamoxifen activated the estrogen receptor (ESR) tagged-Cre and removed the floxed exons, resulting in a selective ablation of p150^Glued expression in cortical neurons cultured from postnatal day 0 (P0) Dctn1^LoxP/LoxP; Cre/Esr1 pups (Fig. 1d, e). P150^Glued proteins were identified by an antibody against the N-terminal 200 amino acids of p150^Glued. By contrast, the levels of p135 and other N-terminal truncated p150^Glued, called collectively as p135+, were substantially increased (Fig. 1d, e). The p135+ proteins were recognized by an antibody against the C-terminus of p150^Glued. On the other hand, the other dynactin subunits–p62, p50, and ARP1 showed comparable expression levels in Dctn1^LoxP/LoxP; Cre/Esr1 and control neuronal cultures (Fig. 1d, e). Therefore, genetic depletion of p150^Glued increases p135+ expression, but does not affect the other dynactin subunit levels in neurons.

To evaluate the impact of p150^Glued on the formation of dynein/dynactin complexes, we fractionated the cell lysates from Dctn1^LoxP/LoxP; Cre/Esr1 and control Dctn1^LoxP/LoxP cortical neuron cultures in a sucrose gradient column and compared the distribution patterns of dynein and dynactin subunits in different fractions. In control cell lysates, all the dynactin subunits and dynein DIC subunit concentrated in fractions 3 to 6 (Fig. 2a). The same distribution pattern of dynein and dynactin subunits was also observed in the p150^Glued-lacking cell lysates (Fig. 2b). Thus, a lack of p150^Glued does not affect the formation of dynein/dynactin complexes.

Our previous study suggests the G59S missense mutation in the MTBDs of p150^Glued might cause the autosomal dominant motor neuron disease through a dominant-negative mechanism [7]. On the one hand, the G59S substitution disrupts the normal folding and destabilizes the mutant p150^Glued proteins. On the other hand, the residual mutant proteins might dimerize with the wild-type proteins and thereby compromises the overall function of p150^Glued. However, the exact pathogenic mechanisms of G59S mutation remain controversial [1]. To test the hypothesis that p150^Glued is required for the survival of motor neurons, we decided to remove p150^Glued from both CSMNs in the frontal cortex and SMNs in the ventral spinal cords by generating Dctn1^LoxP/LoxP; Thy1-Cre cKO mice. In line with the early study [16], Cre was widely expressed by neurons in the forebrain regions (Additional file 1: Figure S1A), including olfactory bulb, frontal cortex, striatum, and hippocampus, as well as in the spinal cord (Additional file 1: Figure S1B). Although the expression of Cre was detectable in the brain extracts of P0 pups, a substantial reduction of p150^Glued expression became apparent in 2-month-old Dctn1^LoxP/LoxP; Thy1-Cre mice (Additional file 1: Figure S1C). The depletion of p150^Glued restrictive to adult neurons may avoid any potential developmental defects in the p150^Glued cKO mice.

Immuno-staining showed endogenous p150^Glued proteins were highly expressed by neurons in the olfactory bulb, cerebral cortex, hippocampus, midbrain, and hindbrain regions of control Dctn1^LoxP/LoxP mice (Fig. 3a). Following the expression pattern of Cre (Additional file 1: Figure S1A), the levels of p150^Glued were substantially depleted in the olfactory bulb, frontal cortex, and hippocampus of Dctn1^LoxP/LoxP; Thy1-Cre cKO mice (Fig. 3a). More importantly, p150^Glued was absent in the CSMNs of cKO mouse brains, while the p135+ remained (Fig. 3b). Here the CSMNs were indicated by staining with CSMN-specific marker protein CTIP2 [23]. P150^Glued was also abundantly expressed by spinal neurons, including the SMNs in the ventral horns, but disappeared in the cKO spinal cords (Fig. 3c, d). In contrast, the p135+ proteins were still present in the SMNs of cKO mice (Fig. 3d). Therefore, we generated a line of p150^Glued cKO mice that depleted the expression of p150^Glued in the adult CSMNs and SMNs, while kept p135 and other MTBD-lacking variants.

The Dctn1^LoxP/LoxP; Thy1-Cre p150^Glued cKO mice were born at a normal Mendelian ratio, developed normally, and had similar body weights as littermate controls (Fig. 4a). The cKO mice also displayed normal ambulatory movement in the open-field tests (Fig. 4b), but showed markedly reduced rearing at 18 months of age (Fig. 4c), and stayed less time on the rotating rods compared to the controls starting at 12 months of age (Fig. 4d). These data suggest that a lack of p150^Glued affects the motor control of aged animals.

To investigate whether p150^Glued depletion causes motor neuron loss, we used the unbiased stereology approach to count the numbers of lumbar SMNs and CSMNs in 9 and 18-month-old control Dctn1^LoxP/LoxP mice and Dctn1^LoxP/LoxP; Thy1-Cre cKO mice. Series of evenly sampled lumbar spinal sections were stained with SMN marker CHAT or histological dye Nissl for neuron counting (Fig. 5a). There was modest but statistically significant reduction of CHAT-positive SMN numbers in the lumbar spinal cord of 18-month-old cKO mice (Fig. 5b). The decrease of lumbar SMN numbers in 18-month-old cKO mice was also observed in the Nissl-stained sections (Fig. 5b). On the other hand, no significant alterations of lumbar SMN numbers were found in 9-month-old cKO mice (Fig. 5b). Additionally, no apparent changes of CTIP2-positive or Nissl-stained CSMN numbers were detected in the motor cortex of 18-month-old cKO mice (Fig. 5c, d). Given that endogenous p150^Glued are ubiquitously expressed throughout central neuronal system, we investigated whether p150^Glued depletion led to neurodegeneration in other motor function-related brain regions, including midbrain, striatum and cerebellum. Unbiased stereology revealed that no obvious loss of TH-positive dopaminergic (DA) neurons in substantia nigra pars compacta (SNpc) (Additional file 1: Figure S2A, B) and striatal neurons in 18-month-old cKO mice (Additional file 1: Figure S2C-E). Furthermore, neither the thickness of molecular layer (ML) and granule cell layer (GCL), or the density of Calbindin D28K-positive Purkinje cells in the cerebellum of 18-month-old cKO mice was obviously altered (Additional file 1: Figure S2C, F-I). These results demonstrate that p150^Glued depletion in adult neurons causes preferentially SMN loss in aged animals.

Given that astrogliosis and microgliosis are often in accompany with motor neurodegeneration in the spinal cord [24], we stained the lumbar spinal sections with antibodies against astrogliosis marker GFAP and microgliosis marker IBA1. In line with previous findings [24], we observed a substantial induction of astrogliosis and microgliosis in the spinal cord of 18-month-old cKO mice (Fig. 5e, f).

To investigate whether p150^glued depletion impairs the connections between SMN axon terminals and the innervated muscles, we examined the neuromuscular junctions (NMJs) of gastrocnemius muscles of 18-month-old control and p150^Glued cKO mice. α-Bungarotoxin (BTX), which specifically labels acetylcholine receptors at the postsynaptic sites, was applied to label the NMJs (Fig. 6a). Meanwhile, acetylated α-tubulin (ac-TUBA), which is enriched in axons, was used to mark the axon terminals (Fig. 6a). While the NMJs in control mice remained mostly intact, the majority of NMJs in p150^Glued cKO mice were either partially or fully fragmented, in company with severe denervation (Fig. 6a, b). The hindlimb clasping test has been used to rate the extent of neuromuscular function impairment in rodent models of motor neuron diseases [25]. 18-month-old male control and p150^Glued cKO mice were suspended by the tail for 10 s and the severity of hindlimb clasping was scored. Control mice splayed their hindlimbs outward and away from the abdomen with no clasping reflex, while p150^Glued cKO mice retracted their hindlimbs toward the abdomen with substantially higher hindlimb clasping severity score than control mice (Fig. 6c, d). Given that denervation often induce skeletal muscle atrophy in motor neuron diseases, we measured muscle weight and examined gross muscle pathology of control and p150^Glued cKO mice. We observed a substantial decrease of gastrocnemius muscle weight in 24-month-old male p150^Glued cKO mice (Fig. 6e). Hematoxylin and eosin (HE) staining of gastrocnemius cross sections revealed that 24-month-old male control mice had a polygonal shaped muscle fibers with homogenous fiber size, whereas p150^Glued cKO mice exhibited significantly smaller size of muscle fiber in accompany with pathological hallmarks of myofiber de−/regenerating and inflammatory cells infiltration (Fig. 6f, g). These results demonstrate that a loss of p150^glued caused substantial NMJ disruption and muscle atrophy in aged mice.

Microtubules are formed by the polymerization of α- and β-tubulin dimers [26]. Various post-translational modifications (PTMs) of α-tubulin affect the microtubule dynamics [27]. To assess the impact of p150^Glued depletion on α-tubulin PTMs, we examined the levels of α-tubulin acetylation, tyrosination, and detyrosination by western blots in the spinal cord homogenates of 9-month-old Dctn1^LoxP/LoxP; Thy1-Cre cKO mice and littermate Dctn1^LoxP/LoxP mice (Fig. 7a). We observed a substantial increase of acetylated α-tubulin levels in the p150^Glued cKO samples (Fig. 7a, b). We further confirm this finding by immuno-staining the lumbar spinal sections of 9-month-old mice. There were a few spots in the ventral root SMN axon bundles positive to acetylated α-tubulin staining in the control spinal cords (Fig. 7c). By contrast, a drastic upregulation of acetylated α-tubulin was found in the SMN axons of cKO mice (Fig. 7d). These data demonstrate that p150^Glued depletion leads to substantial increase of acetylated α-tubulin levels in the SMN axons, suggesting potential alterations of microtubule stability in p150^Glued-lacking SMNs.

Dynein/dynactin complexes mediate the transport of autophagosomes and lysosomes in cells [28]. Early studies show impairments of autophagosome and lysosome trafficking in the SMNs of transgenic mice overexpressing disease-related p150^Glued G59S mutation, resulting in cytoplasmic accumulation of autophagosome and lysosome-enriched proteins [29, 30]. Accordingly, we examined the levels of autophagosome and lysosome marker proteins in the spinal cord lysates of 9-month-old p150^Glued control and cKO mice by western blots. Substantial increases of autophagosome protein LC3II [31] and lysosome proteins LAMP1 and LAMP2 [32] were found in the cKO spinal cords (Fig. 8a, b). However, no apparent alterations of other autophagosome proteins ATG3, ATG5, ATG7, and LC3I were detected (Fig. 8a, b). Since the LAMP2 antibody also works well in immunostaining of brain tissues, we further checked the distribution pattern of LAMP2-postive signals in the SMNs of 9-month-old control and cKO mice. We observed markedly increased LAMP2 signals in the perinuclear regions of cKO SMNs (Fig. 8c). Together, our results reveal the impairments of autophagosome and lysosome transportation in the p150^Glued-depleted SMNs.

Neurons communicate with each other through synapses, where the transmitters released from the presynaptic axon terminals bind to the receptors located in the opposite postsynaptic sites in dendritic spines, small protrusions from dendritic shaft [33]. P150^Glued is enriched in the axon terminals and facilitate the loading of cargos to dynein-mediated axonal retrograde transport in developing neurons [34]. To investigate whether p150^Glued plays any role in the postsynaptic sites, we purified postsynaptic density (PSD)-enriched fractions from control mouse brains. The purity of extracted PSD fractions was verified by the presence of high levels of PSD95 proteins, but not the presynaptic protein synaptophysin (Fig. 9). Interestingly, p150^Glued, p135, and other dynactin subunits were found in the PSD fractions, whereas dynein heavy chain (DHC), DIC, and tubulins were undetectable (Fig. 9). On the other hand, low levels of dynein light chain (DLC) were present in the PSD fraction (Fig. 9). These data suggest that p150^Glued and dynactin may mediate the transport of postsynaptic cargos from the postsynaptic sites to dynein motor complex associated with microtubules in the dendrites.

Excitotoxicity mediated by glutamate and its receptors has been implicated in the pathogenesis of motor neuron diseases [35]. Using western blots, we found the levels of glutamate receptors GLUR1, GLUR2, NR2A, and NR2B, as well as postsynaptic density protein 95 (PSD95) were all substantially elevated in the spinal cord lysates of 9-month-old Dctn1^LoxP/LoxP; Thy1-Cre cKO mice (Fig. 10a, b). By contrast, we did not observe any obvious changes of presynaptic protein synaptophysin expression (Fig. 10a, b). To investigate whether more glutamate receptors are transported to the synapses, we isolated the biotin-labeled cell surface proteins from cultured spinal neurons, and found substantial elevation of glutamate receptor GLUR1, GLUR2, NR2A and NR2B present at the membrane fractions (Fig. 10c, d). In correlation with the increased cell surface targeting of glutamate receptors, p150^Glued cKO neurons showed more severe cell death after treated with 10 μm glutamate (Fig. 10e). Therefore, p150^Glued depletion causes more postsynaptic accumulation of glutamate receptors, making p150^Glued-lacking neurons more susceptible to excitotoxicity.