Dystrophin-deficient myoblasts do not achieve terminal differentiation
In vitro differentiation of myoblasts can be induced in culture through the use of a low mitogen medium (differentiation medium; DM) for a few days. In this study, both C2C12 (non-dystrophic) and dfd13 (dystrophin-deficient) myoblasts were cultured in low mitogen medium (differentiation medium; DM) for 10 days before determining their terminal differentiation capacity. Immunofluorescence analyses (Fig 1A-B) showed both non-differentiated and differentiated myoblasts. Multinucleated myotube formation can be seen on day 10 of differentiation in C2C12 (A) myoblasts but is hardly/rarely found in dfd13 myoblasts where less multinucleated cells were observed (Fig 1B). This indicates that dfd13 differentiation capacity is impaired; however, the differentiating dfd13 myoblasts can be seen to be aligned and seem to be ready for cell fusion to become myotubes.
Myotube formation was determined by counting the myonuclei present in a myotube that expressed MyHC (fast-type II, F-MyHC) and then normalized against total nuclei. F‑MyHC, also called ‘fast-twitch’ fibers, were chosen as the terminal differentiation marker based on the existence of a spectrum of fiber types, with type II being the most developed form of myosin: type 1 ↔ 1/2A ↔ 2A ↔ 2A/2X ↔ 2X ↔ 2X/2B ↔ 2B (24). The number of myonuclei in C2C12 myotubes was 9-fold higher compared to myonuclei in differentiating dfd13 myoblasts (p<0.01). The percentage of myonuclei per total nuclei was 80.2% in C2C12 myotubes and 8.2% 6 in differentiating dfd13 myoblasts (Fig 1C).
Immunoblotting for F-MyHC, pan-myosin, and desmin was performed on days 4, 7, and 10 of differentiation (Fig 1D). Generally, F-MyHC expression was increased upon differentiation in C2C12 myoblasts but none of the differentiating dfd13 myoblasts showed any expression (Fig 1E). However, pan-myosin was expressed at day 10 in dfd13 myoblasts (Fig 1F), and there was a significant difference in expression (p<0.01) when compared to C2C12 myoblasts. Desmin expression in both myoblasts was also examined and was found to be drastically increased upon differentiation, in C2C12 myoblasts. Desmin was used as an intermediate differentiation marker and is a muscle specific type II intermediate filament that integrates the sarcolemma, Z-disk and nuclear membrane in myoblasts. In differentiating dfd13 myoblasts, desmin expression gradually increased over the 10 days and was significantly higher (p<0.01) in comparison to levels in C2C12 myotubes (day 10) (Fig 1G).
PTEN-PI3K regulation is perturbed in differentiating dystrophin-deficient myoblasts
Generally, PTEN expression was increased upon differentiation in both types of myoblasts. However, the accumulation of PTEN was found to be higher in differentiating dfd13 myoblasts compared to C2C12 myoblasts (Fig 2A). The accumulation of PTEN throughout differentiation was not significant in C2C12 myoblasts but there were significant increases on day 7 (p<0.05) and day 10 (p<0.05) when compared to non-differentiated cells. Densitometry analysis showed that PTEN expression was significantly higher (p<0.05) in dfd13 myoblasts compared to C2C12 in the non-differentiated stage. The differentiating dfd13 myoblasts were also showed significantly higher compared to the differentiating C2C12 myoblasts on days 4, 7, and 10 (Fig 2B).
PTEN has a negative effect on PI3K, as it concomitantly affects PI3K regulation. PI3K expression and its activity in both types of myoblasts were therefore examined. Generally, total PI3K expression was increased in dfd13 myoblasts throughout the differentiation period. However, significant accumulation of PI3K could be seen by day 7 (p<0.01) and day 10 (p<0.01) when compared to the non-differentiated dfd13 myoblasts. There was no significant difference in total PI3K expression in C2C12 myoblasts throughout the differentiation period compared to non-differentiated myoblasts. Densitometry analysis showed that there were significant differences in total PI3K expression in dfd13 myoblasts on day 4 (p<0.05), day 7 (p<0.01), and day 10 when compared to differentiating C2C12 myoblasts (Fig 2C). As illustrated in Figure 2D, densitometry analysis on PI3K activity was higher in dfd13 myoblasts at the non-differentiated stage, day 4, and reduced at day 10 of differentiation when compared to C2C12 myoblasts.
Akt is less activated in dystrophin-deficient myoblasts
As PTEN is highly expressed in dfd13 myoblasts, it is predicted that Akt is not/less activated, as PTEN has previously been reported to modulate Akt activation in rhabdomyosarcomas cells; skeletal muscle cancer (25). In this study phosphorylation of Akt was not detected at Ser473 or Thr308 in dfd13 myoblasts during differentiation (Fig 3A). It has been reported that phosphorylation of Akt by PDK1 at Thr308 partially activates Akt while full activation requires phosphorylation of Ser473 which can be catalyzed by multiple proteins including Rictor-mTORC2 (26). This activation is responsible for myoblast proliferation as well as differentiation.
As depicted in Figure 3B, total Akt expression increased in both types of myoblast during differentiation, with slightly higher levels present in dfd13 myoblasts (no significant difference). Densitometry analysis of Akt activation via phosphorylation at Ser473 (Fig 3C) showed a significant difference when compared to C2C12 myoblast at all stages (non-differentiated. There was also a significant difference in Akt activation at Ser473 in C2C12 myoblasts when compared to the non-differentiated stage.
Rictor-mTORC2 is less activated in dystrophin-deficient myoblasts
A previous study reported that mTORC2 is essential for terminal myogenic differentiation (27). It was also reported that it participates in actin cytoskeleton arrangements (26), which might be involved in dystrophin functionality in myoblasts. Therefore, it was hypothesized that Akt inactivation is affected by rictor, a subunit of mTORC2, upstream of Akt, and responsible for Ser473 phosphorylation. However, the upstream protein that regulates rictor-mTORC2 remains unclear. In this study, both total rictor and mTOR expression was examined as well as its activation. However, it was impossible to classify the specific complexes of mTOR activity as phosphorylation at Ser2448 can cause the binding to both raptor and rictor (28).
As depicted in Figure 4A, immunoblot analysis of total rictor, mTOR and their phosphorylated forms was performed. Phosphorylated-rictor at Thr1135 was virtually not detected in dfd13 myoblasts throughout the differentiation period. However, rictor expression was increased upon differentiation in both types of myoblasts. Densitometry analysis showed that there was no significant difference in its accumulation in the two types of myoblasts Fig 4B). Rictor expression only showed a significant increase (p<0.05) in dfd13 myoblasts at day 10 compared to non-differentiated dfd13 myoblasts. Densitometry analysis of rictor activation by phosphorylation at Thr1135 (Fig 4C) showed a significant difference in C2C12 myoblasts compared to dfd13 myoblasts at all stages.
Figure 4D shows that total mTOR expression was increased in dfd13 myoblasts during differentiation but remained the same in C2C12 myoblasts. A significant accumulation was seen in dfd13 myoblasts on days 4, 7, and 10 when compared to C2C12 myoblasts. mTOR activity was significantly higher in non-differentiated dfd13 myoblasts when compared to non-differentiated C2C12 myoblasts. The activity showed a significant reduction upon differentiation in dfd13 myoblasts while C2C12 myoblasts showed a significant accumulation when compared to dfd13 myoblasts on days 4, 7, and 10 (Fig 4E). Expression and activation of mTOR were considered to represent endogenous levels of mTOR within both complexes i.e. mTORC1 or mTORC2, where distinct activation occurred. Phosphorylation of mTOR at Ser2448 makes it a major target for p70S6 kinase activation and is also an important event for raptor and rictor binding (28).
Foxo3 expression is highly increased and predominantly localized in the nucleus of differentiating dystrophin-deficient myoblasts
Inactivation of Akt affected another downstream protein, FoxO3 (Forkhead box O3), which is a target protein of Akt. FoxO3 is a transcription factor responsible for the activation of autophagy genes involved in the autophagy machinery. A previous study showed that FoxO3 controls autophagy in skeletal muscle in vivo and induced multiple autophagy genes, including LC3B transcription in skeletal muscle (29). In this study, FoxO3 expression in myoblasts was examined.
Immunoblot analyses (Fig 5A) showed that FoxO3 expression was increased in both types of myoblasts during differentiation. However, there was only a slight increase in FoxO3 in C2C12 myoblasts on day 4, and it remained at the same level until day 10. In non-differentiated dfd13 myoblasts FoxO3 levels were found to be significantly lower (p<0.01) compared to C2C12 myoblasts. FoxO3 then increased throughout the differentiation period and there was a significant difference in dfd13 myoblasts on day 10 compared to C2C12 myoblasts (Fig 5B).
Akt plays a role in inhibiting FoxO3 through phosphorylation at Thr24, Ser256, and Ser319, which leads to nuclear exclusion and activation. As Akt is inactivated in dfd13 myoblasts, it is thought that FoxO3 is not phosphorylated. Therefore, the non-phosphorylated FoxO3 is translocated to the nucleus and activates the gene involved in autophagy. Akt represses FoxO3 via phosphorylation resulting in nuclear exclusion. As Akt is not activated in dfd13 myoblasts, it was speculated that unphosphorylated-FoxO3 translocates into the nucleus and binds to the promoter, thus up-regulating autophagy-related genes such as LC3B, Atg5, and Atg7. Therefore, FoxO3 expression and localization needed to be examined within subcellular fractions, i.e. the nucleus and cytoplasm.
FoxO3 was found to be localized more to the nucleus of dfd13 myoblasts (Fig 5C), with approximately 81.6% of FoxO3 present in the nucleus of dfd13 myoblasts compared to only ~47.6% in C2C12 myoblasts during the undifferentiated stage. Surprisingly, on day 10 of differentiation, levels in the nucleus had accumulated in both types of myoblast; C2C12 ~78.2%, and dfd13 ~94.4% (Fig 5D).
Autophagy related proteins are highly increased in differentiating dystrophin-deficient myoblasts
It has been shown that FoxO3 is highly expressed and primarily localized to the nucleus of differentiating dfd13 myoblasts. Within the nucleus, FoxO3 acts as a transcriptional activator that can recognize and bind to DNA sequences resulting in the activation of genes involved in autophagy, such as those for Beclin1 and Atgs (29). It was hypothesized that autophagy is highly activated in differentiating dystrophin-deficient myoblasts.
Figure 6 presents immunoblot and densitometry analyses of autophagy-related proteins involved in autophagosome formation. Beclin1 expression was slightly increased in C2C12 myoblasts and highly increased in dfd13 myoblasts upon differentiation (Fig 6A). Densitometry analysis of Beclin1 expression showed significant accumulation on days 4, 7, and 10 in dfd13 myoblasts compared to C2C12 myoblasts (Fig 6B).
Atg5 was found to be not/less expressed in C2C12 myoblasts (C). It can be seen that Atg5 expression was increased upon differentiation in dfd13 myoblasts, and expression was significantly increased on days 4, 7, and 10 compared to the non-differentiated stage. Densitometry analysis also showed significant differences (p<0.01) for dfd13 myoblasts at every time point compared to C2C12 myoblasts (Fig 6C).
Densitometry analysis of Atg7 expression in dfd13 myoblasts showed its accumulation throughout the differentiation period; however, there was no significant difference in expression when compared to the non-differentiated stage in C2C12 myoblasts. Expression showed a significant increase in dfd13 myoblasts when compared to the non-differentiated stage and also significant on days 4, 7, and 10 compared to C2C12 myoblasts (Fig 6D).
Microtubule-associated light chain-3b expression is increased but autophagic flux is decreased during dystrophin-deficient myoblast differentiation
Autophagy related genes were highly activated, and the next step was to determine whether a double-membrane vesicle, known as autophagosome, had been formed. The conversion of Light Chain-3B (LC3B-I) to LC3B-II can be considered to represent total autophagosome formation in myoblasts during differentiation.
As shown in figure 7A, expression of LC3B-I and LC3B-II is increased in both differentiating C2C12 and dfd13 myoblasts when compared to the respective non-differentiated myoblasts. LC3B-I expression in differentiated dfd13 myoblasts was significantly higher (p<0.01) at day 10 when compared to differentiated C2C12 myoblasts (Fig 7B). LC3B-II was found to accumulate upon differentiation in both types of myoblasts, but was higher in dfd13 myoblasts, with a significant difference on day 10 (p<0.05) when compared with to C2C12 myoblasts (Fig 7C). However, the LC3B-II/LC3B-I ratio showed a reduction in dfd13 myoblasts compared to C2C12 myoblasts upon differentiation and was significantly different (p<0.05) (Fig 7D).
LC3B-II levels correlate with the number of autophagosomes formed; however, this could indicate either the up-regulation of autophagosome formation or a blockage in autophagic degradation. Besides, it does not conclusively indicate autophagic degradation. Therefore, further analysis was performed to determine autophagic flux, with a lysosome inhibitor used as a positive control.
To accurately determine autophagic activity, measuring the increase in the number of autophagosomes is required. Previous data showed that LC3B conversion was increased in C2C12 but reduced in dfd13 myoblasts upon differentiation; however, the results from measuring LC3B conversion alone could be inappropriately interpreted. LC3B-II itself has been reported to be degraded by autophagy and also tends to be more sensitive than LC3B-I during immunoblotting analysis (30). Autophagosome formation is an intermediate stage of autophagy, and there could be either the generation of autophagosomes or the blocked conversion of autolysosomes. Therefore, an accurate measurement of autophagic flux is required. In this study a Cyto-ID autophagy detection kit was utilized which selectively labeled autophagic vacuoles independent of the LC3B protein, thus eliminating the need for transfection. Chloroquine was used as a control as it passively diffuses into lysosomes and increases the pH, thus inhibiting lysosome function and blocking fusion with autophagosomes to become an autolysosome.
Figure 7 presents the density plots (E-H) and histograms (I-L) from flow cytometry-based profiling for both C2C12 and dfd13 myoblasts in the non-differentiated and differentiated state. From the density plot images, the size and complexity of each myoblast can be determined. Based on the density plot images, it can be seen that the dfd13 myoblasts (Fig 7F) are smaller and less complex than C2C12 myoblasts (Fig 7E), although the differences between these cells are not obvious. After differentiation, it can be seen that C2C12 cells become more complex (Fig 7G) while dfd13 cells remain the same (Fig 7H). A Cyto-ID autophagy detection kit and FITC were utilized to label/stain autophagosomes. The collected data were gated and analyses were performed based on the myoblasts in the gated region. Histograms were then plotted, as shown in Fig 7I-L; the black line indicates the unstained myoblasts (negative control), the blue line stained myoblasts, and the red line stained myoblasts treated with chloroquine (positive control). From the histograms, data were gated based on the intensity (101-103) (X-axis) and then the percentage obtained.
Generally, both treated and non-treated myoblasts showed a reduction in autophagosome counts after 10 days of differentiation. Non-treated myoblasts had fewer autophagosomes when compared to treated myoblasts, which is due to chloroquine acting as a lysosome inhibitor, preventing autophagosome from fusing with lysosomes and causing the accumulation of autophagosomes (Figure I-L). As illustrated in Figure 7M, the number of autophagosomes was significantly decreased after 10 days of differentiation in both non-treated C2C12 myoblasts (p<0.05) and dfd13 myoblasts (p<0.05) when compared to the non-differentiated stage. It was also significantly decreased in treated C2C12 myoblasts (p<0.005) and dfd13 myoblasts (p<0.005) after 10 days of differentiation. There was a significant reduction in autophagosomes in dfd13 myoblasts when compared to C2C12 myoblasts at the equivalent stages and treatment.
From these results, autophagic flux can be determined. Autophagic flux can be calculated by subtracting the chloroquine-treated from the untreated myoblasts, which enables the total number of non-fused autophagosomes to be measured and also the total number of autophagosomes formed. As depicted in Figure 7N, autophagic flux was increased in C2C12 myoblasts after 10 days of differentiation, whereas it was reduced in dfd13 myoblasts. When compared to non-differentiated C2C12 myoblasts, the number of autophagosomes in dfd13 myoblasts was significantly higher (p<0.05) and was slightly higher in differentiated dfd13 myoblasts compared to differentiated C2C12 myoblasts.
Overall, autophagy activity was decreased upon differentiation in dystrophin-deficient myoblasts. Based on the flow cytometry analysis, autophagic flux was decreased, as the total number of autophagosomes detected was higher in dfd13 myoblasts (both non-differentiated and differentiated) when compared to C2C12 myoblasts. Besides, autophagy was 2-fold higher in non-differentiated myoblasts (proliferation state) compared to the differentiated state.