A major limitation in understanding the pathophysiology of GNE Myopathy is the availability of appropriate model system for the study. A complete GNE knockout in mice model was embryonically lethal and transgenic mice model did not represent the disease phenotype 28,29. Thus, understanding GNE function and identification of appropriate drug targets has been a big challenge. Here, we describe a rat skeletal muscle cell-based model system with GNE exon deletion depicting the phenotype similar to GNE mutant cells. This model system can potentially be used for understanding the pathophysiology and drug screening for GNE function.
In the present study, we aim to generate a GNE exon knockout in L6 rat skeletal muscle cell line using Adenovirus assisted SEPT vector based homology recombination technology as shown in Fig. 138. SEPT stands for Synthetic Exon Promoter Trap. It consists of IRES (internal ribosomal entry site) and a promoterless Neomycin resistance gene, flanked by loxP sites (Fig. 2a). SEPT uses the promoter of the target gene after successful recombination, making it more efficient and reducing the chance of off-target recombinations 38. The presence of loxP sites in SEPT also allows removal of SEPT component after recombination using Cre-recombinase, leaving the gene intact without the targeted exon (Fig. 2b). Exon 3 represents epimerase domain region where majority of pathologically relevant mutations have been identified in GNE Myopathy.
The Rat GNE gene (Gene ID: 114711) of L6 skeletal muscle cell line is used as target in the study. It encodes two isoforms hGNE1 (753 amino acids) and hGNE2 (722 amino acids). In rat GNE gene, exon 3 (19615 bp-20066 bp) represents the epimerase exon that has the highest frequency of pathologically relevant GNE mutations. Generation of exon 3 knockouts in L6 rat skeletal muscle cells required cloning of homology arms in pJET1.2 blunt vector. Homology arms were amplified from the intronic region upstream (left homology arm) and downstream (right homology arm) of exon 3 of rat GNE gene as shown in Fig. 2b. The desired PCR bands of approx. 1 Kb from either homology arms were cloned in pJET/1.2 vector and sequence verified (Figure S1, S2 and S3). Following cloning of homology arms, a shuttle vector for exon 3 was generated in pAAV-MCS vector (Agilent Technologies) with SEPT fragment flanked by both the arms. The epimerase shuttle vector or GNE-EP-SEPT vector was confirmed by restriction mapping (Fig. 2c). Then, an AAV-virus carrying shuttle vector was generated in AAV 293 cells by transfecting the cells with the shuttle vector, RC vector and Helper vector (Agilent Technologies). Visible plaques were obtained from AAV293 cells and virus particles were harvested post-infection to determine the titer (Figure S4). L6 rat skeletal muscle cells were infected with AAV-virus particle carrying shuttle vector (GNE-EP-SEPT) and screened with G418 drug (800 µg/ml, Ameresco Company). G418 resistant clones were, further, screened by PCR for SEPT insertion in GNE via homologous recombination. As shown in Fig. 2d, the desired band confirming SEPT inserting of molecular weight 3.3 Kb was observed in clone 3. This clone was further confirmed by sequencing using Sanger Method (Figure S5). Further, the positive cells with SEPT insertion were infected with Ad-Cre virus (Vector Laboratory) harboring Cre-recombinase enzyme that will remove SEPT region using loxP sites. The screening of SEPT negative cells was done by replica plating as loss of SEPT will cause the cells to be G418 sensitive. The positive clones from replica plates were screened for exon 3 knock out by PCR amplification (Fig. 2d). The PCR positive clones were further confirmed by sequencing for loss of exon 3 in one of the alleles (Figure S6). The sequencing data confirmed the presence of both the alleles, wild type GNE with exon 3 (2.2 Kb fragment) and truncated GNE with loss of exon 3 (1.8 Kb fragment). This indicated that we generated a heterozygous skeletal muscle cell line with one wild type GNE allele and other allele knocked out at exon 3 of GNE. This cell line was labeled as SKM-GNEHz (heterozygous for GNE).
Characterization of GNE Exon-3 knockout cell line (SKM-GNEHz)
The effect of exon 3 deletion on GNE transcript and protein production was analyzed in the heterozygous cell line (SKM-GNEHz). For transcript analysis, mRNA from wild type GNE and SKM-GNEHz cell lines was isolated and subjected to RT-PCR using primers 5’ GAGAAGAACGGGAATAACCGGAAG 3’ and 3’ GATGAGCGTCACAAAGTTCTCCTG 5’ spanning from exon 2 and exon 9 of wild type GNE. As shown in Fig. 3a, the mRNA from wild type GNE gene would yield a product of 1.6 Kb while from exon-3 knocked out GNE gene would yield a product of 1.1 Kb. The amplified PCR fragments were analyzed on 1% agarose gel. As shown in Fig. 3b, only one fragment corresponding to 1.6 Kb was observed in SKM-GNEHz similar to wild type GNE. This indicates that transcript corresponding to knock out allele may not be synthesized, probably due to improper splicing. However, the densitometry analysis of the mRNA band indicates that there is approx. 45% reduction in the transcript level of GNE (Fig. 3b) in SKM-GNEHz as compared to wild type L6 suggesting that knockout of GNE from an allele resulted in reduction in GNE transcript level. In order to determine the copy number of GNE gene in SKM-GNEHz, real time PCR analysis was done using primers indicated in Materials and Methods. The Ct value comparison showed 2-fold reduction in GNE levels for SKM-GNEHz compared to wild type GNE cell line (Fig. 3c). Our study suggests a 2-fold reduction in GNE copy number in SKM-GNEHz, thereby indicating production from only one allele. The protein expression levels of GNE were analyzed in SKM-GNEHz compared to wild type GNE by immunoblot analysis using specific antibody. As shown in Fig. 3d, only single protein band corresponding to wild type GNE (approximately, 79 kDa) was observed in SKM-GNEHz. This suggests that no protein was, probably, translated from knock out allele. Whether these changes in GNE gene, transcript and protein levels are significant to affect GNE function needs to be investigated further.
Recently, Chakravorty et al reported that a patient with a deletion in the promoter of one allele GNE and having a single mutation in the other allele showed GNE myopathy phenotype6. This shows that heterozygous knockout or non-functional one allele may lead to GNE Myopathy. Even though there were no changes in the transcript and protein size of GNE in the heterozygous knockout SKM-GNEHz, we further investigated whether GNE knockout in one allele could result in a phenotype. We determined the sialic acid content of SKM-GNEHz as compared to wildtype L6 rat muscle cells. Cells were seeded in 100 mm dishes and grown in DCCM media supplemented with 4 mM of L-glutamine for 24 h. The sialic acid content was determined by the resorcinol-periodate method as described before. Approximately, 20% reductions in the conjugated sialic acid and 10% reduction in total sialic acid content was observed in SKM-GNEHz cells as compared to the wild type L6 skeletal muscle cells (Fig. 4a). Further, sialic acid supplementation of cells restored total sialic acid content in the SKM-GNEHz cells (Fig. 4b). This study suggests that heterozygous GNE knock out cells do not show drastic reduction in sialic acid content. Whether one wild type allele is sufficient for sialic acid production in the cell needs to be explored further.
Although no significant reduction in the sialic acid content of GNE heterozygous knockout was observed, it was important to assess the effect of GNE knock out on its activity. For this purpose, the epimerase activity of GNE was determined in SKM-GNEHz cells. Approximately, 60% reduction in epimerase enzyme activity of SKM-GNEHz was observed as compared to wildtype L6 cells (Fig. 4c). Significant reduction in GNE enzyme activity could lead to altered cellular functions mediated by GNE. This study would help us identify the potential role of GNE in the muscle cell.
Effect of GNE exon 3 knockout on actin dynamics in rat skeletal muscle cell
Role of actin, being a key cytoskeletal component, is significant in muscle contraction, regeneration and migration. While sliding of thin actin filament over thick myosin filament is important for muscle contraction, assembly and disassembly of actin filament provides driving force for cell migration 39. Maintenance of monomer G-actin and filamentous F-actin pool is tightly regulated inside the cell 40,41. In striated muscle, the sarcomeric actin filaments undergo dynamic turnover without altering the sarcomeric structure 42. Further, actin dynamic is critical for myoblast cell migration, and myoblast fusion vital for muscle regeneration 43,44. In GNE myopathy, muscle regeneration is severely affected and actin may have a role in regulating this phenomenon. Therefore, we investigated the change in actin dynamics level, by studying the levels of G-actin and F-actin in SKM-GNEHz cells. G-actin and F-actin pools were isolated from L6 skeletal muscle cells and SKM-GNEHz and subjected to SDS-PAGE analysis followed by immunoblotting with specific antibodies. As shown in Fig. 5a, G-actin levels were found to be higher in SKM-GNEHz compared to wild type muscle cells. Also F/ G-actin ratio was significantly reduced in SKM-GNEHz cell line compared to control (Fig. 5a). This study suggests that polymerization of actin monomers (G-actin) to F-actin is inhibited in GNE-SKMHz cells.
The alteration in actin pattern was further confirmed by staining with phalloidin using confocal microscopy. The cells were grown in DCCM media for 24 h followed by TRITC-phalloidin staining as described in Material & Methods. Confocal images showed disrupted F-actin pattern in SKM-GNEHz cell line compared to control (Fig. 5b). F-actin filament was shorter and thinner in SKM-GNEHz while wild type L6 cells showed longer and striated actin pattern. This is the first report to demonstrate the disruption of actin cytoskeletal organization in skeletal muscle cell with altered GNE function.
GNE exon-3 knockout affects RhoA signaling and its effector molecule
Rho GTPases are small molecules that play central role in regulating actin assembly 45. The activity of RhoA is upregulated in heterotopic ossified muscle cells in DMD (Duchenne Muscular Dystrophy) leading to weakness and degeneration of muscle cells 46. RhoA is also localized in the focal adhesion involved in triggering downstream integrin signaling leading to actin assembly and disassembly 47,48. We also investigated the levels of RhoA in SKM-GNEHz as compared to wild type control cells. Immunoblot analysis with anti-RhoA antibody showed that expression level of RhoA protein was upregulated in SKM-GNEHz cells as compared to wild type L6 cells (Fig. 6a). Densitometric analysis of immunoblots revealed statistically significant increase in RhoA levels in SKM-GNEHz cells compared to control. Downstream cascade of RhoA upregulation relate to upregulation of ROCK that further activates LIMK which in turn phosphorylates Cofilin. Phosphorylation of Cofilin inactivates its actin severing function, thereby inhibiting F-actin depolymerization. In our study, we found increased phosphorylation of Cofilin in SKM-GNEHz cell lines compared to control cells (Fig. 6b) suggesting inactivation of Cofilin. This inhibits actin severing and may cause slow turn-over of actin pool.
The signaling molecules for actin formation upstream of RhoA involve Src Tyrosine kinases and FAK (focal adhesion kinases). Src is involved in stimulation of initial actin filament nucleation and regulates the rate of turnover of actin filament 49. We observed approximately 28% reduction in Src activation in SKM-GNEHz cells line compared to control cells (Fig. 6c). This suggests down regulation of Src activity that may have significant effects in actin assembly in SKM-GNEHz cells.
Expression and phosphorylation of FAK regulates the myoblast differentiation, muscle fiber formation and muscle size 50,51. FAK is required for muscle cell migration and F-actin formation 52. Reduced p-FAK/FAK ratio was observed in GNE heterozygous cell line compared to control suggesting implications to FAK activation (Fig. 6d). Our study indicates that GNE may have a role in regulating signaling molecules of actin assembly pathway and alteration in GNE activity may significantly disrupt the cytoskeletal organization via actin.
Effect of GNE exon − 3 knockout in L6 myoblast cell migration
Myoblast cell migration is critical for muscle cell regeneration and myoblast fusion critical for muscle differentiation. Migration of cell is directly regulated by actin dynamics. During the process of cell migration, actin polymerization provides the required force for cell to move forward. Migration of cells involves four different steps- protrusion of lamellipodia in the front, focal adhesion protein recruitment, retraction of the rare and disassembly of the focal adhesion from the rare 53,54. In the present study, we found that F/G actin is reduced in SKM-GNEHz as compared to L6 wild type cells. To investigate whether alteration in actin dynamics in SKM-GNEHz affects the cell migration, we conducted wound healing assay using cell culture inserts and followed wound closure for 24 h as mentioned in Material and Methods. As shown in Fig. 7, cell migration was drastically reduced in SKM-GNEHz compared to L6 cells. There was approx. 60% reduction in cell migration. This finding suggests that GNE may play a critical role in regulating cell migration property of muscle cell and may contribute to slow regeneration of muscle cell in GNE myopathy.