Serum Neural Cell Adhesion Molecule Is Hyposialylated in Hereditary Inclusion Body Myopathy

Abstract

Hereditary inclusion body myopathy (HIBM) is a young-adult onset autosomal recessive disorder caused by a hypomorphic rate limiting enzyme of sialic acid biosynthesis. The enzyme is UDP-N-acetylglucosamine 2-epimerase/N-acetylmannosamine kinase, and is encoded by the GNE gene. HIBM causes slowly progressive muscle weakness and atrophy. Patients are typically diagnosed at 20-30 years of age, and most patients are incapacitated and wheelchair-confined by 30-50 years of age. Some sialic acid containing glycoproteins, including neural cell adhesion molecule (NCAM), are hyposialylated in HIBM muscle biopsy samples. We developed a method to allow detection of serum NCAM sialylation using Western blot, and tested serum samples from several patients and a HIBM mouse model. Preliminary results showed a clear difference in polysialylated and hyposialylated forms of NCAM extracted from serum, and showed NCAM is hyposialylated in HIBM serum samples. This initial finding may prove useful in reducing the need for serial muscle biopsies in HIBM treatment trials. Additional studies are underway to further validate this finding and to evaluate the specificity, reliability, and robustness of this potential serum biomarker for HIBM.

Introduction/Background

Hereditary inclusion body myopathy (HIBM) is an autosomal recessive young-adult onset neuromuscular disorder, originally described in clusters of Japanese patients as Nonaka myopathy (Nonaka et al., 1981) and in Iranian Jewish patients as vacuolar myopathy sparing the quadriceps (Argov and Yarom, 1984). HIBM is also known as distal myopathy with rimmed vacuoles (Nishino et al., 2002). HIBM causes slowly progressive weakness of the proximal and distal muscles with complete or partial sparing of the quadriceps (Eisenberg et al., 2001; Darvish et al., 2002). The etiology of HIBM is associated with mutations in GNE, which encodes UDP-N-acetylglucosamine 2-epimerase/N-acetylmannosamine (ManNAc) kinase, the bifunctional enzyme that catalyzes the first two steps of sialic acid biosynthesis (Hinderlich et al., 1997; Keppler et al., 1999). Thus, GNE mutations found in HIBM patients have been associated with reduction of both epimerase and kinase activity of this enzyme (Sparks et al., 2005), and have been further linked to the subsequent hyposialylation of HIBM skeletal muscle glycoproteins that results from diminished sialic acid levels (Huizing et al., 2004; Noguchi et al., 2004; Tajima et al., 2005; Broccolini et al., 2006; Ricci et al., 2006).

The major carrier of polysialic acid (PSA) in mammals is the neural cell adhesion molecule (NCAM) which binds homopolymers of alpha-2,8-linked PSA chain to form PSA-NCAM (Cremer et al., 1994). Besides its importance in neuronal development (Brusés and Rutishauser, 2001), PSA-NCAM is also thought to play a significant role in myogenesis and neuromuscular development, and in the structure and function of the neuromuscular junction (Rafuse and Landmesser, 1996; Rafuse et al., 2000). Due to its hyposialylated state in HIBM muscle biopsy samples, PSA-NCAM has been proposed as a potential biomarker for HIBM in muscle biopsy specimens (Ricci et al., 2006; Broccolini et al., 2010). Hyposialylation of HIBM skeletal muscle glycoproteins, including NCAM and Neprilysin, has been previously described (Ricci et al., 2006; Broccolini et al., 2008). Prior attempts to detect sialic alteration of serum glycoproteins in HIBM have been unsuccessful as most general serum proteins are derived from hepatocyte or B cell secreted proteins (Savelkoul et al., 2006; Huizing and Krasnewich, 2009). Here we describe the specific hyposialylation of polysialylated NCAM (PSA-NCAM) in serum of both human patients affected by HIBM and an HIBM mouse model suggesting that targeted analysis of muscle-derived proteins in serum may detect the biochemical alterations of HIBM.

Materials and Methods

The study described was performed by HIBM Research Group, a nonprofit molecular genetics laboratory specializing in performing high complexity clinical testing, regulated under the Clinical Laboratory Improvement Amendments of 1988, California Laboratory Field Services, and accredited by the College of American Pathologists.

All reagents were purchased from commercial sources and each batch validated in compliance with clinical diagnostic quality standards. Reagents and containers were sterile and nuclease free unless otherwise noted. Optical density measurements were performed using the Nanodrop ND-1000 Spectrophotometer^® (Nanodrop Technologies).

Specimen

Western blot analysis of PSA-NCAM was performed on human samples of control serum (nonmyopathic), myopathic serum (non-HIBM), and HIBM serum samples. Serum from mice bearing Gne^M712T mutation was also tested, using surviving homozygous progenies in FVB; B6 background strain.

Protein extraction

Protein extraction was performed using the Mem-PER Eukaryotic Membrane Protein Extraction Reagent Kit (Thermo Scientific). Fifty microliters of the serum sample was mixed with 150 μL of Reagent A (detergent) and vortexed for a homogeneous suspension. Samples were incubated for 10 min at room temperature, and occasionally vortexed (e.g., 5 s every 3-4 min). The lysed cells were placed on ice. Each sample was then mixed with a 450 μL 2:1 dilution of Reagent C and Reagent B respectively (Reagents B and C must be kept at 4°C or on ice at all times). Samples were then incubated on ice for 30 min and vortexed every 5 min. Tubes were centrifuged at 10,000 g for 3 min and the supernatant was collected in new tubes and incubated for 30 min in a 37°C water bath to separate the membrane protein fraction. Samples were centrifuged at room temperature for 2 min at 10,000 g to isolate the hydrophilic fraction from the hydrophobic fraction. The hydrophilic phase (top layer) was carefully removed from the hydrophobic phase (bottom layer) and collected. Phase separations were performed as efficiently and quickly as possible to avoid disappearance of the separate phase interface between layers that occurs at room temperature.

Protein sample purification

SDS-PAGE Sample Prep Kit (Thermo Scientific) was used for protein sample purification.

Resin preparation and sample addition

Twenty-five microliters of PAGE-prep Protein Binding Resin was added to the protein sample using a cut or large-orifice pipette tip. Ten microliters of sample with protein concentration of 30-45 μg/μL was added to resin protein and vortexed. One volume of 100% DMSO equal to the sample volume (10 μL) was added. Tubes were briefly vortexed. Samples were then incubated for 5 min at room temperature and occasionally vortexed to ensure maximum protein adhesion to the resin. Samples were then centrifuged at 2000 g for 2 min. Flow-through was discarded and the collection tube was blotted on paper towel to remove remaining solution.

Wash

Samples were washed using a 300 μL mixture consisting of a 1:1 ratio of DMSO and water. Three hundred microliters of Wash Solution was added to the resin and vortexed until a homogeneous suspension was obtained. Samples were centrifuged at 2000 g for 2 min. Flow-through was discarded and collection tube was blotted on paper towel to remove the remaining solution. Wash was repeated one additional time for a total of two washes.

Elution

Spin cups were transferred to a new collection tube and 50 μL of Elution Buffer was added to the resin. Tubes were capped and briefly vortexed to obtain a homogeneous suspension. Samples were then incubated at 60°C for 5 min. After incubation tubes were capped and vortexed. Samples were centrifuged at 2000 g for 2 min. Eluate in collection tube was retained and stored.

Desialylation using neuroaminidase

Serum samples from healthy nonmyopathic individuals were desialylated by treatment with Vibrio cholerae neuroaminidase (Sigma). Twenty microliters of normal nonmyopathy serum was treated with 0.5 μL of enzyme and incubated 4 h in a 37°C water bath, along with a parallel tube without enzyme.

Western blot analysis of serum NCAM

Western blot analysis was conducted using the Mini Trans-Blot Electrophoretic Transfer Cell (Bio-Rad). Final protein concentration was measured by BradfordUltra (Expedeon), and 390 μg protein (15 μL) loaded per lane for electrophoresis separation. Purified protein sample was mixed with loading dye in a 1:1 ratio. Fifteen microliters of each sample was loaded. Loading dye was prepared as follows: Laemmli Sample Buffer (Bio-Rad) was mixed with 2-Mercaptoethanol (Bio-Rad) in a 20:1 ratio. Samples were then denatured at 95°C for 3 min using the ThermoHybaid PCR Express, model number HBPX110 (Hybaid Corporation), or the Techne Model TC-412 (Barloworld Scientific). Denatured samples were loaded on a Mini-Protean^® Precast Gel (Bio-Rad) and run at 100 V for 1 h and 20 min in running buffer (1:10 diluted solution of 10× Tris/Glycine/SDS buffer [ISC BioExpress]). After the run, the gel was equilibrated for 15 min in transfer buffer (1:10 diluted solution of 10× Tris/Glycine Buffer (ISC BioExpress) with one added milliliter of 10% SDS solution (Bio-Rad)). Samples were then electrophoretically transferred using fiber pads, filter paper, and Immun-Blot™PVDF Membrane (Bio-Rad), all of which had been previously soaked in transfer buffer for at least 30 min at 4°C. The transfer was completed at 30 V overnight at 4°C. The next morning, the membrane was placed in 1:10 diluted solution of Ponceau S Stain (amresco) to visualize the bands before proceeding. The stain was washed off completely using triple distilled water or TBST (1:20 diluted solution of 20× Tris-Buffered Saline (ISC BioExpress) with one added milliliter of Tween 20 (Bio-Rad)). The membrane was then incubated in blocking buffer (5% Blotting Grade Blocker Non-Fat Dry Milk TBST solution) for 1 h at room temperature, while shaking at 25 rpm. The membrane was next incubated using 200 μg/mL NCAM (RNL-1) mouse monoclonal IgG (Santa Cruz Biotechnology) diluted in blocking buffer at 1/10,000 for 2 h at room temperature, shaking at 25 rpm. The membrane was washed in TBST for 25 min at room temperature at 25 rpm. The TBST was switched out for fresh TBST every five minutes at which point the membrane container was rotated 90°. After being washed, the membrane was incubated using 2 mg/mL goat anti-mouse IgG (abcam) at 1/10,000 and Precision Protein StrepTactin-HRP Conjugate (Bio-Rad) at 1/100,000 in blocking buffer for 1 h at room temperature at 25 rpm, and washed again as described above. The membrane was incubated in 5 mL of SuperSignal^® West Dura Extended Duration Substrate (Thermo Scientific) for 5 min and subsequently dried and stored in a cassette to be developed following 30 s film exposure. Band images were measured for density using the image processing program ImageJ (NIH).

Results

Western blot analysis of serum NCAM sialylation was performed on serum samples obtained from several patients, as follows: three healthy volunteers with no myopathy, two patients with HIBM who had received no treatment, one patient with HIBM who had been consuming ManNAc of his own volition (anecdotally reported by the patient), and one patient with quadriceps sparing nonHIBM myopathy without GNE mutation and showing no hyposialylation on muscle biopsy NCAM (Broccolini et al., 2010). In all control serum samples (from unaffected individuals without muscle disease), sialylated NCAM migrated as one or more bands at approximate size of 150 kD. This signal weakens considerably when serum is treated with neuraminidase (sialidase). The density of the 150 kD bands of noncontrol human samples was measured relative to the average density of control human samples. Each human sample was compared to the standard individually. Serum samples from HIBM patients show a distinctly weaker signal at 150 kD similar to neuraminidase treated control serum. Serum samples from treated HIBM patients show a signal at 150 kD that is stronger than that of nontreated HIBM patients, but slightly weaker than control serum samples. In NonHIBM myopathy samples, the density of the 150 kD band is slightly weaker than normal control serum samples but noticeably stronger than HIBM samples (Fig. 1).

FIG. 1.

(A) HIBM patient serum NCAM Western blot, (B) Densitometry plot. Lanes (1) molecular ladder, (2-4) healthy nonmyopathic control, (5) HIBM patient with GNE mutation, (6) myopathic patient without GNE mutation, (7) HIBM patient on ManNAc self-treatment, (8) desialylated healthy control, (9) molecular ladder, (10-11) healthy control, (12) HIBM patient with GNE mutation. Each lane represents single serum sample obtained by single blood draw. The two HIBM patient serum samples (lanes 5 and 12) show an intermediate signal between desialylated (lane 8) and control serum (lanes 2-4 and 10-11) at ∼150 kD marked by arrow. ManNAc self-treated HIBM patient (lane 7) shows improved signal at ∼150 kD. HIBM, hereditary inclusion body myopathy; NCAM, neural cell adhesion molecule; ManNAc, N-acetylmannosamine.

To verify that the difference in the Western blot image for HIBM serum samples is due to hyposialylation, control serum samples were desialylated (treated with sialidase). In the desialylated human controls, NCAM showed as a weak signal at 150 kD that resembled the NCAM band at 150 kD in HIBM serum samples. When compared with normal serum samples, the desialylated lanes show an extra clearly visible yet hazy band between 37 and 25 kD that is not present in normal serum samples. This hazy band is absent in HIBM serum samples and may represent a sialic acid rich glycan that has been detached as a result of the desialylation process.

Western blot analysis of serum NCAM sialylation was also performed on several groups of mice: M712T homozygous without treatment (water alone), M712T homozygous with N-acetylneuraminic acid (NeuAc) treatment, M712T homozygous with ManNAc treatment, wild-type control, and desialylated wild-type control. Wild-type serum samples show signals at 300 and 200 kD, and a weaker signal at 150 kD. The average density of the 200 kD bands of noncontrol mouse samples was measured relative to the average density of control mouse samples. The signal strength of wild-type samples is stronger than any other sample. Homozygous HIBM mice serum samples show a very weak but visible signal at 200 kD. Homozygous HIBM mice treated with NeuAc, the most common sialic acid found in mammalian cells, or ManNAc for just 1 month show a stronger signal at 200 kD than untreated HIBM mice. Both treated HIBM mice serum samples however, still show a weaker signal than seen in wild-type serum (Fig. 2). In desialylated wild-type mouse controls, the sample shows very weak signals above 88 kD, closely resembling the NCAM signal in HIBM mouse serum samples.

FIG. 2.

Mouse Serum NCAM Western blot. (A) shows the membrane after removal from the transfer apparatus and staining with Ponceau S to demonstrate the amount, uniformity, and effectiveness of protein transfer. (B) shows the developed image as final result of blotting procedure (30 s exposure). (C) shows densitometry plot. Lanes (1) Molecular ladder, (2) desialylated wild-type control, (3-4) homozygous Gne^M712T, (5-6) Sia treated homozygous, (7-8) ManNAc treated homozygous, (9-10) wild-type control. Untreated, treated, and control groups (lanes 3-10) represent pooled samples of n=5 per group, loaded in duplicate lanes. Homozygous HIBM mice (lanes 3-4) shows an intermediate signal between desialylated (lane 2) and control serum (lanes 9-10) at band sizes shown by arrows as 88, 150, 200, and 300 kD. Treated mice (lanes 4-8) show an improved signal compared with nontreated mice (lanes 3-4).

To confirm that what looks to be hyposialylation in the desialylated normal serum samples is not a result of protein breakdown due to their incubation at 37°C, normal serum samples without desialylation enzyme were incubated along with the normal serum samples with desialylation enzyme and analyzed by Western blot. The normal serum samples that had been incubated at 37°C showed the same band makeup as normal control serum samples and showed no evidence of protein degradation. Thus, the band makeup of the normal serum samples that were incubated with desialylation enzyme was likely due to desialylation and not incubation induced protein degradation.

Discussion

We developed and tested a method to allow Western blot detection of NCAM sialylation in samples of human and mouse serum. Previous examination of PSA-NCAM in normal muscle samples showed that it migrates as a broad band at 150 kD (Ricci et al., 2006). In this study, Western blot analysis of normal control mouse serum showed PSA-NCAM migrating at 300, 200, and 150 kD, whereas normal control human serum NCAM migrated at 150 kD. In both mouse and human serum NCAM, a clear difference is seen between control bands with strong signals and HIBM serum bands that show much weaker signals and resemble desialylated control serum samples. These initial findings provide early evidence that serum biomarkers targeting PSA may be developed for HIBM. If soluble PSA-NCAM or other sialylated glycoproteins (sialoprotein) found in serum partly originate from muscle, then improved sialylation of a serum sialoprotein marker may reflect the restoration of sialylation in the muscles of an HIBM patient as they are treated; thus, potentially reducing the need for more invasive serial muscle biopsies during interventional clinical trials. ManNAc or NeuAc treatment prevents muscle disease in the HIBM mouse model (Malicdan et al., 2009). Since ManNAc or NeuAc also improve sialylation of NCAM, our findings suggest that biomarkers targeting PSA may ultimately prove useful as a surrogate outcome measure in HIBM clinical trials.

Further studies, perhaps with more specific monoclonal antibodies and an increased number of subjects, are necessary to validate the use of serum NCAM as a reliable and robust biomarker for HIBM. Other antibodies that preferentially bind to the NCAM protein backbone or PSA can be used to develop a sensitive and quantitative ELISA based assay.

Conclusion

Western blot analysis of PSA-NCAM was performed on human control serum (nonmyopathic), human myopathic serum (non-HIBM), and human HIBM serum samples. Serum from mice bearing GNE^M712T mutation was also tested. There is a marked difference in the signal strength between the control samples and all other sample groups, with control samples showing the strongest signal. This difference seems to be reduced by ManNAc treatment in both patients and the HIBM mouse model. Future work includes development of such a quantitative assay and determining if improvement of PSA-NCAM sialylation level correlates with functional benefit in human HIBM patients.

Footnotes

Acknowledgment

The authors extend their gratitude to Advancement of Research for Myopathies (ARM) who supported this study.

Disclosure Statement

E.K. is an employee, stockholder and executive of Ultragenyx, a company developing a treatment for HIBM. D.D. and Y.V.-A. are collaborating with Ultragenyx, and may gain rights to exercise Ultragenyx stock options.

References

Argov

, Yarom

. 1984. “Rimmed vacuole myopathy” sparing the quadriceps. A unique disorder in Iranian Jews. J Neurol Sci, 64:33-43. www.ncbi.nlm.nih.gov/pubmed/6737002. 2010 December 22.

Broccolini

et al. 2006. Neprilysin participates in skeletal muscle regeneration and is accumulated in abnormal muscle fibres of inclusion body myositis. J Neurochem, 96:777-789. www.ncbi.nlm.nih.gov/pubmed/16405511. 2010 December 23.

Broccolini

et al. 2008. Hyposialylation of neprilysin possibly affects its expression and enzymatic activity in hereditary inclusion-body myopathy muscle. J Neurochem, 105:971-981.

Broccolini

et al. 2010. Analysis of NCAM helps identify unusual phenotypes of hereditary inclusion-body myopathy. Neurology, 75:265-272. www.ncbi.nlm.nih.gov/pubmed/20644153. 2010 December 13.

Brusés

, Rutishauser

. 2001. Roles, regulation, and mechanism of polysialic acid function during neural development. Biochimie, 83:635-643. www.ncbi.nlm.nih.gov/pubmed/11522392. 2011 March 22.

Cremer

et al. 1994. Inactivation of the N-CAM gene in mice results in size reduction of the olfactory bulb and deficits in spatial learning. Nature, 367:455-459. www.ncbi.nlm.nih.gov/pubmed/8107803. 2011 March 22.

Darvish

, Vahedifar

, Huo

. 2002. Four novel mutations associated with autosomal recessive inclusion body myopathy (MIM: 600737) Mol Genet Metab, 77:252-256. www.ncbi.nlm.nih.gov/pubmed/12409274. 2010 December 23.

Eisenberg

et al. 2001. The UDP-N-acetylglucosamine 2-epimerase/N-acetylmannosamine kinase gene is mutated in recessive hereditary inclusion body myopathy. Nat Genet, 29:83-87. www.ncbi.nlm.nih.gov/pubmed/11528398. 2010 December 23.

Hinderlich

, Stasche

, Zeitler

et al. 1997. A bifunctional enzyme catalyzes the first two steps in N-acetylneuraminic acid biosynthesis of rat liver. Purification and characterization of UDP-N-acetylglucosamine 2-epimerase=N-acetylmannosamine kinase. J Biol Chem, 272:24313-24318.

10.

Huizing

et al. 2004. Hypoglycosylation of alpha-dystroglycan in patients with hereditary IBM due to GNE mutations. Mol Genet Metab, 81:196-202. www.ncbi.nlm.nih.gov/pubmed/14972325. 2010 December 23.

11.

Huizing

, Krasnewich

. 2009. Hereditary inclusion body myopathy: a decade of progress. Biochim Biophys Acta, 1792:881-887.

12.

Keppler

, Hinderlich

, Langner

et al. 1999. UDP-GlcNAc 2-epimerase: a regulator of cell surface sialylation. Science, 284:1372-1376.

13.

Malicdan

et al. 2009. Prophylactic treatment with sialic acid metabolites precludes the development of the myopathic phenotype in the DMRV-hIBM mouse model. Nat Med, 15:690-695.

14.

Nishino

et al. 2002. Distal myopathy with rimmed vacuoles is allelic to hereditary inclusion body myopathy. Neurology, 59:1689-1693. www.ncbi.nlm.nih.gov/pubmed/12473753. 2010 December 23.

15.

Noguchi

et al. 2004. Reduction of UDP-N-acetylglucosamine 2-epimerase/N-acetylmannosamine kinase activity and sialylation in distal myopathy with rimmed vacuoles. J Biol Chem, 279:11402-11407. www.ncbi.nlm.nih.gov/pubmed/14707127. 2010 December 26.

16.

Nonaka

et al. 1981. Familial distal myopathy with rimmed vacuole and lamellar (myeloid) body formation. J Neurol Sci, 51:141-155. www.ncbi.nlm.nih.gov/pubmed/7252518. 2010 December 24.

17.

Rafuse

, Landmesser

. 1996. Contractile activity regulates isoform expression and polysialylation of NCAM in cultured myotubes: involvement of Ca2+ and protein kinase C. J Cell Biol, 132:969-983. www.ncbi.nlm.nih.gov/pubmed/8603927. 2011 March 22.

18.

Rafuse

, Polo-Parada

, Landmesser

. 2000. Structural and functional alterations of neuromuscular junctions in NCAM-deficient mice. J Neurosci, 20:6529-6539. www.ncbi.nlm.nih.gov/pubmed/10964958. 2011 March 22.

19.

Ricci

et al. 2006. NCAM is hyposialylated in hereditary inclusion body myopathy due to GNE mutations. Neurology, 66:755-758. www.ncbi.nlm.nih.gov/pubmed/16534119. 2010 December 23.

20.

Savelkoul

et al. 2006. Normal sialylation of serum N-linked and O-GalNAc-linked glycans in hereditary inclusion-body myopathy. Mol Genet Metab, 88:389-390.

21.

Sparks

et al. 2005. Use of a cell-free system to determine UDP-N-acetylglucosamine 2-epimerase and N-acetylmannosamine kinase activities in human hereditary inclusion body myopathy. Glycobiology, 15:1102-1110. www.ncbi.nlm.nih.gov/pubmed/15987957. 2010 December 23.

22.

Tajima

et al. 2005. Distal myopathy with rimmed vacuoles: impaired O-glycan formation in muscular glycoproteins. Am J Pathol, 166:1121-1130. www.ncbi.nlm.nih.gov/pubmed/15793292. 2010 December 13.