Profiling Serum Antibodies Against Muscle Antigens in Facioscapulohumeral Muscular Dystrophy Finds No Disease-Specific Autoantibodies

Genetically confirmed FSHD patients (N = 138) were retrospectively recruited from the FSHD–FOCUS cohort study (Nijmegen, The Netherlands). Serum samples from all patients were provided by Radboud Biobank, Nijmegen, the Netherlands. Healthy control sera (N = 20) were either obtained from the Sanquin Blood Supply Foundation (Nijmegen, the Netherlands) or collected at the neurology out-patient clinic (Radboud university medical center, Nijmegen, The Netherlands) from healthy volunteers participating in the FSHD-Inflammation study (Nijmegen, The Netherlands). Gender and age at examination were recorded for both groups. Additionally, disease duration, Medical Research Council (MRC) sum score, Ricci Score, and 'Lamperti Score were recorded for the patient group. Disease duration was determined from age at disease onset. The MRC sum score ranges from 0 (complete paralysis) to 60 (normal muscle strength) and was obtained by summing the manual muscle testing scores of three upper limb muscles (shoulder abductors, elbow flexors, and wrist extensors) and three lower limb muscles (hip flexors, knee extensors, and foot dorsiflexors). The Ricci Score (0–10 scale) [19] and and the Lamperti Score (0–15 scale) [20] take into account the extent of weakness in various body regions and considers the descending spread of symptoms from face and shoulders to pelvic and leg muscles typical of FSHD. A subset of 109 patients out of 138 underwent a full leg magnetic resonance imaging (MRI) examination as described by Mul et al [21]. Briefly, transversal Dixon and short tau inversion recovery (STIR) sequences were acquired from the upper and lower leg to evaluate the degree of muscle fatty infiltration and inflammation, respectively. Finally, FSHD genetics characteristics (FSHD type and D4Z4 repeat length) were recorded as well from all FSHD patients. See Table 1 for additional details.

This study was conducted according to the principles of the Declaration of Helsinki (version October 2013) and in accordance with the Medical Research Involving Human Subjects Act (WMO). The local Institutional Review Board (CMO region Arnhem-Nijmegen, The Netherlands) approved the study under file number CMO 2017–3119. All subjects provided written informed consent.

The reactivity of FSHD sera and healthy control sera was tested with different classes of muscle tissue and myoblast/myotube cell line extracts (as a source of muscle antigens) following a stepwise workflow: 1) healthy human muscle, 2) FSHD muscle, 3) healthy control myotubes, 4) FSHD myotubes, and 5) inducible DUX4 myoblasts (Fig. 2).

Human healthy muscle specimens obtained from knee surgery leftovers and FSHD quadriceps needle muscle biopsies (see Table 2 for additional information) were snap frozen in liquid nitrogen and stored at –80 °C. Frozen healthy muscle pieces and FSHD muscle pieces were pulverized with a Microdismembrator (Braun Biotech International, Melsungen, Germany), resuspended in 10 volumes of RIPA buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 0.1% SDS, 1% Nonidet P-40, 0.5% sodium deoxycholate (DOC), and Complete protease inhibitor cocktail), sonicated at 4 °C, and centrifuged for 15 minutes at 13.000 g. Healthy and FSHD muscle tissue protein extracts were separated by SDS-PAGE followed by western blotting.

Human healthy control (mHC), FSHD1 (mF1), and FSHD2 (mF2) myotubes (see Table 2 for additional information) were washed three times in cold PBS, resuspended in LDS buffer (NuPAGE™ LDS Sample Buffer, Invitrogen) (mHC and mF1) or RIPA buffer (mF2) for 30 minutes to ensure complete cell lysis, and centrifugated for 15 minutes at 13,000 g. mHC, mF1, and mF2 protein extracts were separated by SDS-PAGE followed by western blotting.

Primary myoblasts from a healthy control individual (see Table 2 for additional information) were first immortalized with hTERT and CDK4 transductions, followed by random insertion of codon-altered DUX4 cDNA sequence as described in detail by Jagannathan et al. [22]. In short, cells were transduced with lentivirus containing the plasmid pCW57.1-DUX4-CA (Addgene, #99281). Cells were selected and maintained in 1μg/ml puromycin containing medium. Individual clones were generated by single cell dilution cultures in 96-well format and clones were individually validated for DUX4-induction upon doxycycline treatment and DUX4-induced cytotoxicity within 24 hours. Inducible DUX4 myoblasts were next cultured in proliferation medium F-10, supplemented with 20% fetal bovine serum (FBS), 1% penicillin/streptomycin (GIBCO), 1μM dexamethasone (DEX), and 10 ng/ml FGF-basic and 1μg/ml puromycin at 37 °C in a 5% CO2 atmosphere. DUX4 induction was achieved by a 17 hours incubation of cultured myoblasts in above described proliferation medium, supplemented with 2μg/ml doxycycline (DOX). Protein extract from myoblasts with induced DUX4 expression (DOX+) and without DUX4 induction taken as negative control (DOX-) were used for immunoblotting as described above.

Muscle, myotubes, and inducible DUX4 myoblasts protein extracts prepared as described above, were separated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE), followed by transfer to nitrocellulose membranes. Three millimeter wide strips were cut and blocked in phosphate-buffered saline, 0.1% Tween-20 (PBS-T), 5% non-fat dry milk, and incubated with FSHD patient or healthy control sera, 500-fold diluted in 5% milk in PBS-T. Subsequently, strips were incubated with the secondary antibody goat–anti-human immunoglobulin conjugated to IRDye800 (Rockland, Gilbertsville, PA) and visualized with the Odyssey system (LI-COR Biosciences, Lincoln, NE).

A possible relationship between DUX4 expression and serum reactivity was investigated by immunofluorescence staining of DUX4 expressing myoblasts incubated with either FSHD patient or healthy control sera. Immunofluorescence staining was performed on an inducible DUX4 cell line (1926 iDUX4 cell line). Cells were fixed in 2% paraformaldehyde-PBS solution for 15 min at room temperature and washed twice with PBS. Subsequently, cells were permeabilized in 1% Triton X-100, PBS solution, washed with PBS, and incubated for 1 hour at room temperature with rabbit anti-DUX4 antibody (E5.5, ad124699, Abcam, 2000-fold diluted) and FSHD patient or healthy control sera (10-, 100-, or 500-fold diluted). Cells were then washed three times with PBS for 10 min at room temperature, followed by incubation with goat-anti-rabbit immunoglobulin conjugated to AlexaFluor 488 (400-diluted) and goat-anti-human antibody conjugated to AlexaFluor 568 (400-fold diluted). After 1 hour of incubation, cells were washed three times with PBS and incubated with DAPI for 10 minutes. Images were captured with a widefield fluorescence microscope (Leica LAS X).

The sample size of this study was not based on a power statistical analysis. Immunoblots and immunofluorescent staining were qualitatively analyzed. Subject characteristics, observed reactivities frequencies, staining pattern and staining intensities were summarized through a descriptive statistical analysis conducted using GraphPad prism 5 (GraphPad software, La Jolla, CA).

Up to one hundred and thirty eight FSHD patient sera (48±16 years, 48% male) and twenty healthy control sera (47±14 years, 50% male) were included for antibody screening against skeletal muscular antigens. The total number of tested sera in each set of experiments is specified in the following sub-sections. Demographic, clinical, and genetics characteristics of patients and healthy controls are summarized in Table 1.

In total, 138 sera of FSHD patients (FSHD1 = 131; FSHD2 = 7) and 11 sera of healthy controls were tested for autoantibody reactivities with healthy skeletal muscle protein extract by immunoblotting (Supplementary Figure 1). The results showed 5 relatively frequently occurring bands. Reactivity with a polypeptide of approximately 160 kDa was observed with 20 FSHD sera; reactivity with a polypeptide of approximately 98 kDa with 10 FSHD sera; reactivity with a polypeptide of approximately 73 kDa with 20 FSHD sera; reactivity with a polypeptide of approximately 48 kDa with 7 FSHD sera; and reactivity with a polypeptide of approximately 36 kDa with 26 FSHD sera. Additional immunoblots, in which the reactive FSHD sera (Fig. 3 A-E) and healthy control sera (Fig. 3F) were analyzed on flanking strips from the same blot revealed that similar reactivities were also seen with some healthy control samples. Additionally, we did not observe differences between FSHD1 and FSHD2 sera reactivities. We did not observe significant differences as well between the reactivity of FSHD patient sera reporting STIR hyperintensity on leg MRI and those without STIR hyperintensity. The frequencies by which these reactivities were observed in each group are summarized in Table 3.

Since the protein composition of skeletal muscle tissue of FSHD patients may be partially different from that of healthy individuals, 85 FSHD sera (FSHD1 = 84; FSHD2 = 1) and 3 healthy control sera were subsequently analyzed with an extract of FSHD muscle biopsies (see Table 2 for additional details about FSHD muscle biopsies). The most frequently occurring reactivities on immunoblots were observed with polypeptides of approximately 60 kDa (6 FSHD sera; 1 healthy control serum), 45 kDa (3 FSHD sera; none of the healthy control sera) and 37 kDa (3 FSHD sera; 1 healthy control serum) (Supplementary Figure 2). A representative set of tested patient and control immunoblots are shown in Fig. 4. The frequencies by which these reactivities were observed are summarized in Table 4.

Since skeletal muscle biopsies not only contain muscle fibers but also other cell types, we decided to use also extracts from cultured myotubes as a source of antigens. The reactivity of 49 FSHD sera (FSHD1 = 48; FSHD2 = 1) and 3 healthy control sera was analyzed with extracts from (i) myotubes derived from a healthy individual (HC), (ii) myotubes derived from a FSHD1 patient (F1), and (iii) myotubes derived from a FSHD2 patient (F2), again by immunoblotting (Supplementary Figure 3). In spite of the appearance of reactive bands on both patient and control blot strips, a specific reactivity pattern was not observed. A polypeptide of approximately 60 kDa was relatively frequently detected with the FSHD1 and FSHD2 extracts (reactive with 6 FSHD sera), but not with the HC extract (Fig. 5). However, a 60 kDa polypeptide was also recognized by one of the three healthy control sera, although this was observed for all types of myotube extracts (Fig. 5). The frequencies of these reactivities are summarized in Table 5.

Although no potentially FSHD-specific autoantibody reactivities were detected with the FSHD myotubes protein extracts, we further explored the possibility that DUX4 induced the expression of autoantigenic proteins by using extracts from a myoblast cell line in which DUX4 was expressed. To investigate the presence of autoantibodies to these proteins in FSHD patient sera, 86 sera (FSHD1 = 84; FSHD2 = 2) and 3 healthy control sera were incubated with blot strips containing extracts from these myoblasts, in which DUX4 expression was induced by doxycycline (DOX+) (Supplementary Figure 4). Note that in these cultures DUX4 was detectably expressed in only a subset of nuclei (Fig. 6A).

The reactivity with extracts from non-induced cells was analyzed in parallel. Reactivity with a 60 kDa band in the DUX4-deficient extract was observed in 7 FSHD sera, whereas this reactivity was not detected for the DUX4-containing extract. Conversely, 3 FSHD sera showed reactivity at 60 kDa with the DUX4-expressing myoblast extract, but not with the non-induced myoblast extract. Similarly, a 100 kDa band was detected with 2 FSHD sera only in the extract from DUX4-expressing myoblasts. Finally, a 75 kDa polypeptide was recognized by one FSHD sample in the DUX4-expressing myoblast extract and by two FSHD samples in the absence of DUX4. The frequencies of these reactivities are summarized in Table 6.

The DUX4 inducible myoblast cell line allowed the analysis of FSHD sera by immunofluorescence and to explore the relationship between DUX4 expression and autoreactivity. Fixed cells were incubated with 85 FSHD and 14 healthy control sera (Supplementary Figure 5). DUX4 expression was visualized with an anti-DUX4 antibody, which showed that DUX4 was detectably expressed in only a subset of the nuclei, in agreement with previous observations (Fig. 6A). This phenomenon facilitated a comparison of the staining by patient autoantibodies in DUX4-expressing and non-expressing cells on the same specimen. After staining with a fluorescent secondary antibody, we observed cytoplasmic and (peri)nuclear staining in all samples (Fig. 7). However, the fluorescence intensity differed among samples and was graded as weak, medium, and strong. Besides, among and within the same samples we observed a variation in nuclear staining patterns comprising the following patterns: homogeneous, nuclear speckles, nucleolar, and peripheral. Nevertheless, staining intensity and staining pattern frequencies did not differ between patient and control sera (Fig. 7 A-B). Importantly, FSHD and healthy control sera were reactive in DUX4 expressing as well as non DUX4 expressing cells (Fig. 7 C-F) excluding in the tested cohort a DUX4 dependent relationship.