Immunoreactivity of valosin-containing protein in sporadic amyotrophic lateral sclerosis and in a case of its novel mutant

VCP immunohistochemistry in ALS

For VCP immunohistochemistry, formalin-fixed, paraffin-embedded 6-μm-thick sections were deparaffinized and immunostained for VCP. We applied 2 distinct primary antibodies for VCP(mouse monoclonal [Cat. No. ab11433], Abcam plc, Cambridge, UK; 1:500; and rabbit polyclonal [Cat. No. AP6920b], Abgent, San Diego, California, USA; 1:75). Bound antibodies were detected with the appropriate VECTASTAIN Elite ABC Kit (Vector Laboratories, Burlingame, California, USA). We assessed staining specificity by replacing the primary antibodies with an appropriate amount of phosphate-buffered saline solution containing 3% bovine serum albumin. No deposits of reaction products were seen in the sections thus treated. Procedures involving the use of human material were performed in accordance with ethical guidelines set by Kyoto University.

For quantitative study of VCP-positive nuclei of the motor neurons, we obtained 3 distinct sections from each case, taken at an interval of 30 μm, to avoid counting the same nucleus twice. All observations were made by 2 examiners who had no information regarding the clinical history and condition of the individual providing the spinal specimen. Only anterior horn cells (AHCs) with a distinct nucleus and nucleolus were counted. The frequency of VCP-positive motor neurons in the lumbar cord was compared between control patients and SALS patients by the nonparametric Mann–Whitney U test using Prism software (GraphPad, La Jolla, USA).

Clinical features of the ALS patient with a VCP mutation

Our patient was a 41-year-old Japanese man without any family history of ALS, IBM, Paget disease or FTD (Figure 1a). He had developed right arm and leg weakness at age 36. Based on the results of elevated alkaline phosphatase up to 2,968 IU/l, bone radiographs, and electromyography, he was diagnosed at age 37 as having ALS and Paget disease of bone. He gradually developed weakness of arms and legs, dysarthria, dysphagia, and respiratory failure, which required percutaneous endoscopic gastrostomy and noninvasive positive-pressure ventilation at age 38. He was referred to our hospital at age 39. Examinations disclosed weakness in all extremities, muscle atrophy in the distal arms, and fasciculation in the right leg. Atrophy or fasciculation of the tongue was not evident. Patellar tendon reflexes were brisk, but Achilles tendon reflexes and those in the upper limbs were decreased. Neither pathologic reflexes nor clonus was observed in the extremities. Neuropsychological tests showed no evidence of FTD. A computed tomography (CT) scan showed systemic osteolytic change, but no atrophy of the brain. Needle electromyography (EMG) showed active and chronic denervation potential including spontaneous discharges and late recruitment in the left arm and leg. A frozen muscle specimen from the right quadriceps muscle at this time showed neurogenic changes without evidence of IBM. Base on the revised El Escorial criteria [32], the patient was diagnosed as clinically possible ALS.

His extremities gradually became weaker, resulting in a bedridden condition. Shortly after he developed a fracture of the right femoral bone, he died of respiratory failure at the age of 41. Throughout the clinical course he did not develop any cognitive decline or mood disorder, Parkinsonism, spastic paraplegia, cerebellar ataxia or decubitus ulcer. Also, the patient did not have any abnormalities in cardiac, hepatic, visual, auditory, sensory or autonomic systems. No frozen brain tissue was available.

Genetic analysis

Genetic analysis of VCP, Cu/Zn superoxide dismutase 1 (SOD1), TDP-43, fused in sarcoma (FUS), Charged multivesicular body protein 2b (CHMP2B), angiogenin, Sequestosome 1 (SQSTM1), and chromosome 9 open reading frame 72 (C9ORF72) was approved by either the Kyoto University Graduate School of Medicine Ethical Committee or Hiroshima University Ethical Committee, and written informed consent was obtained from the above patient prior to his demise. DNA was extracted from a blood sample; and by using previously reported primers, we amplified the exons of each gene and sequenced them on anABI310 sequencer (Applied Biosystem, Foster City, California, USA). To detect C9ORF72 expansion, we performed a repeat-primed polymerase chain reaction, as reported previously [33]. As a result we identified a novel heterozygous M158V mutation in the VCP gene (Figure 1b). This mutation was not present in the 1000 Genomes project (http://www.1000genomes.org/), dbSNP (http://www.ncbi.nlm.nih.gov/SNP/) or in the Human Genetic Variation Browser (includes genetic variations determined by exome sequencing of 1,208 Japanese individuals and genotyping data of common variations obtained from a cohort of 3,248 individuals [http://www.genome.med.kyoto-u.ac.jp/SnpDB/]). The other genes were normal. Conservation analysis of the mutated amino acid by using National Center for Biotechnology (NCBI) HomoloGene demonstrated conservation among species at the mutation site (Figure 1c). The mutation site in this case (amino acid 158 in VCP) was located near the reported mutation sites (amino acids 155 [26] and 159 [28] in VCP) within the same domain (Figure 1d). In Figure 1d, mutation sites in VCP gene associated with ALS and IBMPFD are also listed from the literature [8],[9],[13]-[15],[26],[28],[31],[34]-[39].

His parents were still alive and not affected with ALS or any other neurodegenerative diseases. His father was 73 years of age and had been diagnosed with prostate cancer, whereas his mother was 71 years old and healthy. Neither parent carried the VCP mutation identified in the proband. Haplotype analysis using 15 short tandem repeat markers supported their genetic kinship, confirming that the mutation occurred de novo in the present case. No sample from the patient’s brother was available.

Neuropathologic examinations

For neuropathologic examinations, formalin-fixed, paraffin-embedded 6-μm-thick sections were deparaffinized and stained with hematoxylin and eosin (H&E) or used for Klüver-Barrera (KB) staining. For immunohistochemistry, after antigen retrieval by heat/autoclaving (10 min at 121°C in 10 mM anhydrous citric acid buffer, pH 6.0), the sections were immunostained as described above.

The antigens recognized by the primary antibodies used in this study were the following: TDP-43(rabbit polyclonal [Cat No. 10782-2-AP], ProteinTec Group Inc., Chicago, Illinois, USA; 1:200), phosphorylated TDP-43 (mouse monoclonal [Cat No. TIP-PDT-M01], Cosmo Bio Co., Ltd., Tokyo, Japan; 1:2000), trans-Golgi-network (sheep polyclonal [Cat No. 610832], Novus Biologicals, Inc, Littleton, Colorado, USA; 1:250), VCP(mouse monoclonal [Cat No. ab11433], Abcam plc, Cambridge, UK; 1:500; rabbit polyclonal [Cat No. AP6920b], Abgent, San Diego, California, USA; 1:75), FUS(rabbit polyclonal [Cat No. HPA008784], Sigma-Aldrich Inc., St. Louis, Missouri, USA; 1:100), CD68 (mouse monoclonal [Cat No. M0814], DAKO, Glostrup, Denmark; 1:100), p62(mouse monoclonal [Cat No. 610832], Becton Dickinson and Company, Franklin Lakes, New Jersey, USA; 1:700), ubiquitin(rabbit polyclonal [Cat No. U5379], Sigma-Aldrich Inc., St. Louis, Missouri, USA; 1:100), and optineurin (rabbit polyclonal [Cat No. 100000], Cayman Chemical Company, Ann Arbor, Michigan, USA; 1:200).

Electron microscopy

In the case with the VCP mutation, the sample from formalin-fixed medulla was fixed in 2% glutaraldehyde with phosphate buffer (pH 7.4). After fixation, sections were cut into pieces about 1 mm thick, postfixed with 1% osmium tetroxide for 2 h, dehydrated, and embedded in epoxy resin. Each block was then cut into semithin sections about 1 μm in thickness and stained with toluidine blue. Appropriate regions were subsequently cut into ultrathin sections and stained with uranyl acetate and lead citrate for electron microscopy.

Proteomics analysis

The effect of the newly detected missense mutation (M158V) and previously reported mutations of 2 autopsied ALS-VCP cases (R155H and R159H) was analyzed with Mutation Taster (http://www.mutationtaster.org), Sorting Intolerant From Tolerant (SIFT, http://sift.bii.a-star.edu.sg/), and PolyPhen-2 (http://genetics.bwh.harvard.edu/pph/).

Vectors for in vitro analysis

To assess the functional importance of the novel mutation identified in this study, we created vectors (pcDNA3) expressing wild-type (WT) VCP and mutant VCPs (R155H, M158V, R159H and A232E) tagged with FLAG at their N-terminus.

Cultured cells

SH-SY5Y and human embryonic kidney 293 T (HEK293T) cells were grown in Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% fetal bovine serum, 1% non-essential amino acids, and 1% penicillin-streptomycin-L glutamine sodium.

Plasmids

To generate FLAG-tagged VCP, we obtained WT VCP complementary deoxyribonucleic acid (cDNA) from a cDNA library and generated sequence variants, including R155H (464G > A), M158V (472A > G), R159H (476 G > A) and A232E (695C > A). WT and mutant VCPs were generated by PCR with primer pairs used to change the nucleotides and then inserted into the Not I/BamH I cloning site of the pcDNA3 vector. All constructs were verified by using the 3130xl Genetic Analyzer (Life Technologies, Carlsbad, California, USA. Cultured cells were transfected with the vector by using a FuGene HD transfection kit (Roche, Basel, Switzerland) according to the manufacturer’s protocol.

Immunoblot analysis

Cells were lysed in lysis buffer containing 10 mM Tris–HCl (pH 7.6), 150 mM NaCl, 1% Triton X-100, 1% sodium deoxycholate (Na-DOC), 0.1% sodium dodecyl sulfate (SDS), and a protease inhibitor cocktail (Complete EDTA-free protease inhibitor; Nacalai Tesque). Nuclei and membrane fractions were removed by centrifugation. Lysates were separated by SDS-poly-acrylamide gel electrophoresis (SDS-PAGE), and proteins were then transferred to a polyvinylidene difluoride (PVDF) membrane. The membrane was incubated with the appropriate primary antibody, followed by incubation with horseradish peroxidase (HRP)-conjugated anti-rabbit or anti-mouse IgG (Santa Cruz Biotechnology Inc, Dallas, Texas, USA) secondary antibody. Immunoreactive proteins were visualized by using the Pierce ECL Western Blotting Substrate (Thermo Scientific Pierce, Kanagawa, Japan) and an LAS3000 scanning system (Fuji Film, Tokyo, Japan). The following primary antibodies were used in the immunoblot analysis: anti-VCP (rabbit polyclonal [Cat No. AP6920b], Abgent, San Diego, California, USA; 1:200) and anti-β-actin (mouse monoclonal [Cat No. A5441], Sigma-Aldrich Inc., St. Louis, Missouri, USA; 1:2,000).

Immunocytochemistry

SH-SY5Y and HEK293T cells transfected with FLAG-conjugated WT or mutant VCP were fixed and immunostained with anti-TDP-43 or anti-FLAG, and stained with DAPI, at 48 h post-transfection. For immunofluorescence analysis of cell cultures, cultured cells were fixed with 4% PFA for 20 min, washed 3 times with PBS (2 min each), and then rendered permeable and blocked with 0.2% Triton X-100/5% goat serum in PBS for 15 min. The transfected cells were incubated with primary antibodies (anti-TDP-43 antibody, rabbit polyclonal, ProteinTech, 1:1000 and anti-FLAG M2, mouse monoclonal, Sigma-Aldrich Inc., St. Louis, Missouri, USA; 1:500, diluted in PBS containing 0.2% Triton X-100/5% goat serum) overnight and washed 3 times with PBS (5 min each). After the final wash, the cells were incubated with secondary antibodies (Alexa Fluor 488 donkey anti-rabbit IgG (H + L) or Alexa Fluor 546 donkey anti-mouse IgG (H + L), diluted in PBS containing 0.2% Triton X-100/5% goat serum) for 1 h, washed 3 times with PBS (5 min each), and mounted with Vectashield plus DAPI (Vector Laboratories, Burlingame, California, USA). Digital imaging was performed with an OLYMPUS FV-100 IX microscope (Olympus, Tokyo, Japan) using FV-10-ASW 3.1 software (Olympus). A blinded examiner counted more than 100 FLAG-tagged VCP-expressing cells from separate cultures for the presence of cytoplasmic TDP-43-positive cells. Results were analyzed by One-way ANOVA of Bonferroni’s multiple-comparison test using Prism software (GraphPad, La Jolla, USA).

Immunohistochemistry for VCP in control subjects, SALS patients, and the VCP mutant case

Two distinct anti-VCP antibodies applied in the present study gave the same results.

In the 8 controls, VCP-positive nuclei or cytoplasm was rarely observed in the AHCs investigated (Figure 2a); only a few nuclei of AHCs showed faint immunoreactivity against VCP (Figure 2a inset). Glial nuclei were virtually not stained in these subjects (Figure 2b). In contrast, AHCs from 9 SALS cases frequently showed positive VCP immunoreactivity in the cytoplasm and in the nucleus (Figure 2c). The staining intensity of neuronal nuclei of the SALS cases was much greater than that in the control subjects. Within the cytoplasm, the deposits of immunoreaction product were diffusely distributed, but this antibody recognized no intranuclear inclusion, skein-like inclusion, round hyaline inclusion or Bunina body. Moreover, the glial nuclei were also stained in the cases with SALS (Figure 2d). However, VCP-positive glial cytoplasmic inclusions (GCIs) were not observed.

We also identified some neuronal (Figure 2e) and glial (Figure 2f) nuclei that were immunoreactive for VCP in the ALS case with the heterozygous M158V VCP mutation (case 18 in the Table). In the ALS-VCP case, VCP immunoreactivity in the AHCs was less robust than that in the SALS cases. No VCP-positive inclusions were detected in any of the cases investigated.

In the motor cortex and hippocampus, there was no significant immunoreactivity for VCP except for faint immunoreactivity of glial nuclei.

The frequency of VCP-positive nuclei in the lumbar cord of patients with SALS (mean ± standard deviation, 26.1 ± 31.5) was significantly higher (p < 0.05, Mann–Whitney U test) than that for the control subjects (1.8 ± 3.2%; Figure 2g). In the ALS-VCP case 7.7% of the AHC nuclei were VCP positive. On serial section analysis, there was no apparent relationship between VCP immunoreactivity and intracytoplasmic TDP-43 accumulation.

Neuropathologic examinations of the patient with M158V ALS-VCP

Frozen muscle specimens from the right quadriceps muscle at age 39 showed neurogenic changes without evidence of IBM (Figure 3a). In the immunohistochemical investigation, TDP-43 was normally detected in the muscle nuclei, which showed no abnormal inclusions (Figure 3b). Immunostaining for neither VCP nor OPTN revealed any abnormal inclusions.

At autopsy of the M158V ALS-VCP patient, pneumonia, fatty liver, and atheromatous plaques of the aorta were recognized. Microscopic examination of the vertebra revealed mixed osteoclastic-osteoblastic activity and a chaotic picture of trabecular bone (“mosaic” pattern) that was compatible with Paget disease of bone(Figure 3c).

Neuropathologically, the formalin-fixed brain weighed 1,435 g. Macroscopic examination showed no evidence of cerebral atrophy (Figure 3d, e). In H&E- and KB-stained sections, the spinal anterior horns and the corticospinal tracts showed evidence of degeneration, especially in the lumbar cord (Figure 3f). Cervical and lumbar cord showed cell loss with gliosis in the anterior horns (Figure 3g). Onuf’s nucleus and Clarke’s column were preserved. H&E-stained sections revealed the presence of Bunina bodies (Figure 3h) and spheroids in the anterior horns of the cervical and lumbar cords.

Immunohistochemical investigation demonstrated TDP-43-positive intracytoplasmic inclusions in the AHCs (Figure 4a). GCIs positive for TDP-43 were sparsely scattered in the spinal cord (Figure 4b). The TDP-43-positive intracytoplasmic inclusions were also reactive with the anti-ubiquitin antibody (Figure 4c,d). The nucleus of these inclusion-bearing neurons was invariably immunonegative for TDP-43 (Figure 4c). Analysis of consecutive sections revealed that these TDP-43-positive inclusions were also reactive with anti-p62 antibody (Figure 4e,f) and anti-OPTN antibody (Figure 4g,h). However, TDP-43-positive inclusions (Figure 4i) were indiscernible on the VCP-immunostained sections (Figure 4j). Careful examination of 164 AHCs in 8 VCP-stained sections from the cervical, thoracic, and lumbar spinal cords revealed no immunopositive inclusions with any of the VCP antibodies. Analysis of serial sections disclosed no apparent correlation between the presence of TDP-43-positive inclusions and nuclear and cytoplasmic distribution of VCP immunoreactivity in AHCs. In the superior frontal cortex, temporal cortex, motor cortex, and hippocampus, there was no significant immunoreactivity with VCP antibody.

Immunohistochemical investigation of the Golgi apparatus revealed characteristic fragmentation of it in the AHCs (Figure 4k), whereas non-motor neurons in the posterior horn had preserved Golgi apparatuses (Figure 4l). Using the same method reported previously [39], we found that in our case 61.5% (16/26) of the AHCs from 3 distinct spinal cord segments had a fragmented Golgi apparatus. In the corticospinal tracts, immunostaining for CD68 showed infiltration of CD68-positive microglia.

In the hypoglossal nuclei, the motoneurons were depleted in number. Within the cytoplasm of the residual neurons Bunina bodies and TDP-43- and p62-positive inclusions were identified. The Bunina bodies were immunopositive for cystatin C (Figure 4m). Electron microscopy demonstrated Bunina bodies consisting of irregularly shaped electron-dense material containing vesicles (Figure 5a-c).

Betz cells were not apparently depleted. Hippocampal sclerosis was not found. In the primary motor, superior frontal, and temporal cortices, putamen, thalamus, cerebellum, and hippocampus, there were no neuronal nuclear or intracytoplasmic inclusions or GCIs found by any of the staining procedures, including p62 immunohistochemistry.

By amyloid β and AT8-immunohistochemistry, this case showed no senile plaque, but only a few isolated neurofibrillary tangles in the hippocampus and amygdala (CERAD 0, Braak-Braak stage II). Immunostaining for α-synuclein and FUS revealed no pathologies.

Bioinformatics and in vitro studies

The effect of the newly detected missense mutation on VCP protein structure was analyzed by in silico analysis using different software. In Mutation Taster (http://www.mutationtaster.org), M158V was estimated to be disease causing (probability > 0.9999), and no Single Nucleotide Polymorphism (SNP) was found in the site of mutation. The known disease-causing VCP mutations R155H and R159H were also diagnosed as disease causing (probability > 0.9999) in Mutation Taster.

In SIFT (http://sift.bii.a-star.edu.sg/) the M158V mutation, as well as the known mutation R155H, was not tolerated (p = 0.00), meaning that the mutation is deleterious. On the other hand, the known mutation R159H [28],[34],[38] was tolerated (p = 0.07, in tolerant threshold <0.05), although the p-value was near the threshold. Polyphen-2 (http://genetics.bwh.harvard.edu/pph/) analysis showed possible damage to the structure by M158V, as indicated by a Position-Specific Independent Counts (PSIC) value of 0.896 (sensitivity: 0.82; specificity: 0.94), R155H was analyzed to be possibly damaging with a score with PSIC value 0.849 (sensitivity: 0.83; specificity: 0.93); although R159H ,which is indeed a disease-causing mutation, was predicted to be benign with a score of PSIC value 0.1 (sensitivity: 0.93; specificity: 0.85). Taken together, in silico analysis strongly suggested that out novel mutation M158V is pathogenic.

To assess the functional importance of the novel mutation identified in our patient, we created mammalian expressing vectors for WT VCP and mutant VCPs (R155H, M158V, R159H, and A232E) tagged with FLAG at their N-terminus. Immunoblotting detected overexpression of WT and mutant VCP in the transfected SH-SY5Y cells (Figure 6a) and HEK293T (Additional file 1: Figure S1a), compared with the expression in the mock-transfected cells, at 48 hours after transfection. TDP-43 is a predominantly a nuclear protein, and translocation of it from the nucleus to the cytoplasm and aggregation therein are features of previously reported types of mutant VCP- related ALS [26],[28]. Mutations in VCP were also reported to cause abnormal TDP-43 translocation in cultured cells including SH-SY5Y [40] and HEK293T [41]. So we investigated whether our novel mutation (M158V) caused the translocation of TDP-43 from the nucleus to the cytoplasm. Immunocytochemistry showed TDP-43 to be located in the nuclei of mock-transfected cells and in those transfected with WT-VCP (Figure 6b). In the mutant VCP-expressing cells, immunoreactivity of TDP-43 was often observed in the cytoplasm by 48 hours after transfection (Figure 6b). However, no evident aggregation of TDP-43 or VCP was observed in the transfected cells. Quantitative evaluation of cytoplasmic TDP-43-positive cells demonstrated that the expression of mutant VCP was significantly associated with the presence of cytoplasmic TDP-43 (Figure 6c). There was no statistical difference in the number of cells with cytoplasmic TDP-43 between the cultured cells transfected with different VCP mutations. In the study using HEK293T, an increase in cytoplasmic TDP-43 translocation was also observed in cells transfected with mutant VCP (Additional file 1: Figure S1b, c).