DEPDC1B is a key regulator of myoblast proliferation in mouse and man

Abstract Objectives DISHEVELLED, EGL‐10, PLECKSTRIN (DEP) domain‐containing 1B (DEPDC1B) promotes dismantling of focal adhesions and coordinates detachment events during cell cycle progression. DEPDC1B is overexpressed in several cancers with expression inversely correlated with patient survival. Here, we analysed the role of DEPDC1B in the regulation of murine and human skeletal myogenesis. Materials and methods Expression dynamics of DEPDC1B were examined in murine and human myoblasts and rhabdomyosarcoma cells in vitro by RT‐qPCR and/or immunolabelling. DEPDC1B function was mainly tested via siRNA‐mediated gene knockdown. Results DEPDC1B was expressed in proliferating murine and human myoblasts, with expression then decreasing markedly during myogenic differentiation. SiRNA‐mediated knockdown of DEPDC1B reduced myoblast proliferation and induced entry into myogenic differentiation, with deregulation of key cell cycle regulators (cyclins, CDK, CDKi). DEPDC1B and β‐catenin co‐knockdown was unable to rescue proliferation in myoblasts, suggesting that DEPDC1B functions independently of canonical WNT signalling during myogenesis. DEPDC1B can also suppress RHOA activity in some cell types, but DEPDC1B and RHOA co‐knockdown actually had an additive effect by both further reducing proliferation and enhancing myogenic differentiation. DEPDC1B was expressed in human Rh30 rhabdomyosarcoma cells, where DEPDC1B or RHOA knockdown promoted myogenic differentiation, but without influencing proliferation. Conclusion DEPDC1B plays a central role in myoblasts by driving proliferation and preventing precocious myogenic differentiation during skeletal myogenesis in both mouse and human.


| INTRODUC TI ON
DISHEVELLED, EGL-10, PLECKSTRIN (DEP) domain-containing 1B (DEPDC1B) and its paralog DEPDC1A are cell cycle-regulating proteins. 1 The DEPDC1B gene, at human chromosome 5q12, encodes a 61 kDa protein of 529 amino acids. DEPDC1B contains an N-terminal DEP domain and a C-terminal RHO-GAP (GTPase-activating protein)-like domain. The DEP domain is a globular region discovered in DISHEVELLED, EGL-10 and PLECKSTRIN and plays a role in mediating membrane localization, 2 and DEPDC1B is usually membrane-associated, being highly expressed during G2/M phase of the cell cycle. 1,3 The RHO-GAP domain is involved in RHO GTPase signalling (eg RAC, CDC42 and RHO) that regulates cell motility, growth, differentiation, cytoskeleton reorganization and cell cycle progression. 4 Membrane association via the DEP domain enables DEPDC1B to interact with G protein-coupled receptors, as well as membrane phospholipids necessary for Wnt signalling. However, the GAP domain of DEPDC1B lacks the critical arginine residue required for GAP activity. 1 The GAP domain of DEPDC1B can also interact with the nucleotide-bound forms of RAC1 and can control their activation. 5,6 DEPDC1B can also indirectly suppress activation of RHOA. 1 The transmembrane protein tyrosine phosphatase receptor type F (PTPRF) and the guanine nucleotide exchange factor H1 (GEF-H1) are required for RHOA activation. DEPDC1B inactivates RHOA by competing for binding of PTPRF, so allowing cell de-adhesion and cell cycle progression. 1 DEPDC1B expression oscillates during cell cycle progression, accumulating at the G2 phase, similar to checkpoint proteins such as cyclin B, which correlates with its function as a regulator of cell cycle. 1 DEPDC1B knockdown induces a significant delay in transition to mitosis, due to impairment of the de-adhesion process. 1 RHOA is required for formation and integrity of focal adhesion points, and DEPDC1B, as an indirect inhibitor of RHOA, promotes dismantling of focal adhesions, necessary for morphological changes preceding mitosis.
RHO GTPases including RHOA, RAC1 and CDC42 are also crucial regulators of skeletal myogenesis, 7 and their precise temporal regulation is critical for efficient myotube formation. 7,8 RHOA is required for the initial induction of myogenesis by activating serum response factor (SRF) 9 which induces the myogenic transcription factor MyoD. [10][11][12] In myocytes however, RHOA perturbs localization of M-cadherin, a cell adhesion molecule required for myoblast fusion, 13 and so needs to be inactivated before myoblast fusion. 14 Such inactivation is mediated by RHOE and GRAF1. 15,16 Therefore, precise modulation of RHOA activity is required for differentiation to proceed. 17 While Rac1 and CdC42 are required for myoblast fusion in Drosophila in vivo, 18 overexpression of RAC1 or CDC42 inhibits myogenesis in rat myoblasts. 19

RAC1
and CDC42 can have this dual role by activating the C-Jun N-terminal kinase (JNK), a negative regulator of myogenesis, but also activating the stress-activated protein kinase (SAPK) and p38: pathways necessary for myogenesis. 20 Moreover, RAC1 inhibits myogenic differentiation by preventing complete withdrawal of myoblasts from the cell cycle 21 and exogenous expression of RAC1 and CDC42 impair cell cycle exit and induce loss of cell contact inhibition. 22 This suggests a function of RAC1 and CDC42 during proliferation, rather than during the differentiation process. DEPDC1B expression is repressed by PITX2, a bicoid-related homeobox transcription factor implicated in regulating the left-right patterning and organogenesis. 6,23,24 The first intron of the human and mouse DEPDC1B gene contains multiple consensus DNA-binding sites for PITX2, and knockdown of PITX2 in murine C2C12 myoblasts promotes an increase in DEPDC1b at the protein level. PITX2 particularly, but also PITX3, are additionally involved in regulation of muscle development and adult muscle stem (satellite) cell function. [25][26][27][28][29][30][31] Finally, DEPDC1B is overexpressed in various cancers including breast, oral, non-small-cell lung, melanoma and prostate, and represents a potential biomarker and therapeutic target. 3,5,[32][33][34][35][36] Interestingly, in many human cancer cells, DEPDC1B also has a nuclear location. 37 DEPDC1B overexpression in breast cancer cells can increase phosphorylation of ERK and promote cell proliferation and delay cell death. 3 DEPDC1B is also highly expressed in oral cancer and overexpression induces augmentation of ERK1/2 activity by RAC1 GTP, promoting cell migration and invasion in cancer cell lines. 5 DEPDC1B up-regulation in non-small-cell lung cancer has a reverse correlation with patient survival, and its overexpression promotes tumour cell migration and invasion through activating Wnt/β-catenin signalling. 33 Increased DEPDC1B expression in prostate cancer is associated with an advanced clinical stage of disease. 32 Deregulation of RHO GTPases is implicated in tumours that have characteristics of skeletal muscle (rhabdomyosarcomas), 22,38,39 but the role of DEPDC1B in rhabdomyosarcomas is unknown.
Here, we examined the role of DEPDC1B in both mouse and human skeletal myogenesis. DEPDC1B was expressed in both mouse and human proliferating myoblasts, but expression fell precipitously during myogenic differentiation. SiRNA-mediated knockdown of DEPDC1B drastically reduced proliferation and induced precocious myogenic differentiation in both species. In myogenic cells, DEPDC1B function was unaffected by manipulation of β-catenin, while DEPDC1B expression was unaltered by PITX2 levels. Given that DEPDC1B can suppress RHOA activation, it was unexpected that DEPDC1B/RHOA co-knockdown had an additive effect in myoblasts, reducing proliferation and enhancing differentiation more than DEPDC1B knockdown alone. DEPDC1B was expressed in human Rh30 rhabdomyosarcoma cells at a higher level than in myoblasts, and while DEPDC1B down-regulation did not affect proliferation, it did promote myogenic differentiation. Thus, DEPDC1B plays a central role in driving proliferation and preventing differentiation during skeletal myogenesis.

| Retroviral expression vectors
The retroviral backbone pMSCV-puro (Clontech) was modified to replace the puromycin selection gene with an IRES-eGFP to create pMSCV-IRES-eGFP, which served as control. 44 Mouse Depdc1b-V5 (Depdc1b-201 transcript encoding for a protein of 529 amino acids with a V5-Polyhistidine region at the C-terminus) and the human PITX2C cDNA (transcript variant 3, coding for a protein of 324 amino acids) were amplified by RT-PCR and cloned into pMSCV-IRES-eGFP.
Retrovirus was then packaged in 293T or Phoenix cells using standard methods. Proliferating satellite cells and immortalized human myoblasts

| Western blot
Primary murine satellite cells were transfected with siRNA against DEPDC1B (Si Depdc1B Mix or Si Depdc1B N2) and allowed to proliferate for 72 h and then collected for Western blot analysis. Blots were probed against DEPDC1B (Sigma, HPA072558-100UL, 1:500), and total protein was visualized using the Bio-Rad TGX Stain-Free Technology.
For quantification, pixel intensity of either the whole lane (total protein image) or the band corresponding to DEPDC1B (DEPDC1B immunolabelled image) was quantified using Fiji, and knockdown efficiency presented as DEPDC1B intensity normalized to whole protein intensity.

| RT-qPCR
Mouse or human myoblasts were cultured in 6-well plates in proliferation medium or switched to differentiation medium for the stated duration. Total RNA was extracted using the RNeasy Kit (Qiagen) and cDNA prepared from 500 ng to 1 μg of RNA with the QuantiTect Reverse Transcription Kit with genomic DNA wipeout (QIAGEN). qPCR was performed on a Mx3005PQPCR system (Stratagene) with Brilliant II SYBR Green reagents and ROX reference dye (Stratagene) or QuantiNova SYBR Green PCR Kit (QIAGEN) using primers shown in Table S2.

Relative expression between proliferating and differentiated myoblasts
and between control and test samples was measured in 3 replicates and significance tested using a two-tailed Student's t test in Microsoft Excel.

| Immunolabelling of cells
EdU incorporation was assayed using a Click-iT EdU Imaging Kit (Life Technologies) as per manufacturer's instructions. Satellite cell-derived myoblasts were immunolabelled as previously described. 45 Briefly, fixed myoblasts were permeabilized with 0.5% Triton X-100/PBS for 10 min at room temperature and blocked with 5% goat serum/PBS for 60 min at room temperature. Primary antibodies (Table S3)

| Image acquisition
Images of plated myoblasts were acquired on a Zeiss Axiovert 200 M microscope using a Zeiss AxioCam HRm and AxioVision software version 4.4 (Zeiss). For colocalization analysis, images were acquired on a confocal microscope (Zeiss). Images were adjusted globally for brightness and contrast and analysed with ImageJ and RStudio 46 software.

| Statistical testing
For analysis of immunolabelling following siRNA treatment, cells in multiple unit areas per well per experimental condition were counted, and data from each mouse were expressed as a single mean ± SD. Such mean ± SD from at least three mice were used for each condition tested.
Significant difference (P < .05) between control and a test sample was determined using a paired two-tailed t test in Excel (Microsoft). For clonal cell lines, cells in multiple unit areas per well per experimental condition were counted and data from each well expressed as a single mean ± SD. Such mean ± SD from at least three independent wells were used for each condition tested and significant difference (P < .05) between control and a test sample determined using an unpaired twotailed t test in Excel (Microsoft). For RT-qPCR, at least three independently siRNA-transfected wells per condition were assessed.

| DEPDC1B is located in the nucleus of murine satellite cell-derived myoblasts
Endogenous DEPDC1B protein was investigated in mouse primary satellite cells using a commercially available DEPDC1B antibody (HPA038255). DEPDC1B was clearly located in the nuclei of proliferating satellite cell-derived myoblasts ex vivo ( Figure 1A). After

| DEPDC1B regulates proliferation and prevents precocious differentiation in murine myoblasts
DEPDC1B function was investigated using siRNA-mediated knockdown in murine satellite cells. Depdc1B expression was decreased by ~72% by two different Silencer selected pre-designed siRNA targeting Depdc1b (N1 and N2) compared to transfection with control siRNA ( Figure 1D). Western blot analysis ( Figure S3A) confirmed that siRNA against Depdc1B also reduced DEPDC1B at the protein level ( Figure S3B).

Satellite cells knocked down for DEPDC1B were pulsed with
EdU for 2 hours in proliferation medium, fixed, EdU incorporation visualized, immunolabelled for MYOGENIN and counterstained with F I G U R E 1 DEPDC1B controls proliferation and differentiation in mouse satellite cell-derived myoblasts. (A and B) Mouse satellite cellderived myoblasts were cultured in proliferation medium (A) or switched to differentiation medium for 2 days (B), fixed and immunolabelled for DEPDC1B (HPA072558) and β-catenin. DEPDC1B was found in the nuclei of both proliferating satellite cell-derived myoblasts and newly formed myotubes and unfused myocytes, but not at cell-cell-junctions. (C) Satellite cells were maintained in proliferation conditions or differentiation medium for 1 or 2 days and Depdc1b and Pax7 expression measured by RT-qPCR. Depdc1b was expressed in proliferating myoblasts, but then decreased during differentiation, mirroring expression of Pax7. (D) DEPDC1B was effectively knocked down via siRNA transfection (N1 and N2) compared to control siRNA (Si control), as assessed by RT-qPCR. (E) Satellite cells transfected with control siRNA (Si control) or siRNA against DEPDC1B were maintained in proliferation conditions, and either pulsed for EdU and immunolabelled for MYOGENIN, or (G) immunolabelled for KI67 and β-tubulin. (F) DEPDC1B knockdown reduced proliferation, but (H) promoted entry into differentiation. Data are mean ± SEM, where an asterisk denotes a significant difference (P < .05) between control and a test sample using a paired two-tailed t test, with multiple fields examined from each of 3 mice. Scale bar represents 100 µm | 7 of 16 FIGEAC Et Al. DAPI ( Figure 1E). Knockdown of DEPDC1B decreased the rate of cell proliferation (EdU + ve) and the total number of cells (DAPIstained) ( Figure 1F) compared to control siRNA. Reduction of cell proliferation following DEPDC1B knockdown was corroborated by a decreased proportion of KI67 + ve cells ( Figure 1G,H).
Since proliferation was reduced, we focused more on the entry into myogenic differentiation, rather than fusion index, which is affected by cell number. DEPDC1B knockdown caused an increased entry into the myogenic differentiation programme, as shown by the higher proportion of myoblasts with MYOGENIN immunolabelling compared to siRNA control ( Figure 1E,F).
We confirmed that siRNA-mediated knockdown of DEPDC1B reduced the proliferation rate by also using a mix of 4 siRNAs targeting DEPDC1B (Qiagen). RT-qPCR and Western blot analysis confirmed that these siRNAs also reduced Depdc1b mRNA ( Figure S4A) and DEPDC1B at the protein level ( Figure S3A) in murine primary myoblasts compared to control siRNA. DEPDC1B knockdown using these 4 separate siRNAs also significantly reduced the proliferation rate of satellite cell-derived myoblasts, as shown by the decreased proportion that had incorporated EdU ( Figure S4B,C).

| DEPDC1B is required for proliferation in human myoblasts
Examining the expression dynamics of DEPDC1B during myogenesis in human primary myoblasts from three individuals, we found that DEPDC1B expression was significantly higher in proliferating myoblasts than in myocytes/myotubes at differentiation days 1, 2 or 5, targeting human DEPDC1B mRNA, with high knockdown efficiency compared to control siRNA, as confirmed by RT-qPCR ( Figure 2C).
Following DEPDC1B knockdown, myoblasts were pulsed with EdU, which revealed a significantly reduced proliferation rate compared to control siRNA ( Figure 2D). After 2 days in differentiation medium, immunolabelling for MyHC showed that DEPDC1B knockdown resulted in a higher proportion of differentiated myocytes compared to control siRNA ( Figure 2E) suggesting that DEPDC1B normally operates to inhibit myogenic differentiation. Moreover, co-knockdown of both DEPDC1A and DEPDC1B together did not have an additive effect on proliferation rate or differentiation, compared to DEPDC1B knockdown alone ( Figure 3C,D).

| DEPDC1A does not affect proliferation in mouse or human myoblasts
Similarly, DEPDC1A knockdown in human C25Cl48 myoblasts did not affect the proliferation rate as assessed by the proportion of EdU or Ki67 + ve cells ( Figure 3H). DEPDC1A knockdown in human myoblasts though did decrease the proportion of MYOGENIN + ve cells, with more myoblasts (data not shown) and myocytes per unit area ( Figure 3I).

| DEPDC1B affects key regulators of the cell cycle
To characterize the effects of knocking down DEPDC1B in proliferating human C25Cl48 myoblasts, we analysed expression of many key F I G U R E 2 DEPDC1B controls proliferation and differentiation in human myoblasts. (A) cDNA from primary human myoblasts obtained from 3 individuals collected in proliferation, and at differentiation day 1, 2 or 3, was analysed for DEPDC1B, CD1 and MYOGENIN expression by RT-qPCR. DEPDC1B was expressed at a higher level in proliferating myoblasts than during differentiation, mirroring expression of CD1, but opposite to that of MYOGENIN. (B) Immortalized C25Cl48 human myoblasts were maintained in proliferation conditions or differentiation medium for 3 or 5 days; again, DEPDC1B was expressed at a higher level in proliferating myoblasts than differentiation, also emulating CD1, but contrary to the MYOGENIN, expression profile. (C) DEPDC1B knockdown efficiency via siRNA transfection was effective compared to control siRNA (Si control), as quantified using RT-qPCR. (D) Control or DEPDC1B siRNA-transfected C25Cl48 human myoblasts were maintained in proliferation medium, pulsed with EdU and immunolabelled for β-tubulin. DEPDC1B knockdown decreased the proliferation rate. (E) DEPDC1B knockdown C25Cl48 myoblasts enter differentiation earlier, with an increased proportion of MyHC + ve cells after 2 days of differentiation. Data are mean ± SEM, where an asterisk denotes a significant difference (P < .05) between control and a test sample using either a paired (A) or an unpaired (B-E) two-tailed t test, with cDNA from 3 individuals or N = 3-5 independently siRNAtransfected wells analysed per condition. Scale bar represents 100 µm genes involved in regulation of cell cycle progression, including cyclins, cyclin-dependent kinases and cyclin-dependent kinase inhibitors ( Figure 4). Proliferating human myoblasts were transfected with siRNA control or siRNA against DEPDC1B and maintained in proliferation conditions for 48 h. DEPDC1B knockdown was confirmed, although DEPDC1A was also significantly affected ( Figure 4A).
Cyclin-dependent kinases CDK1, CDK2 and CDK6 were decreased ( Figure 4D), while the cyclin-dependent kinase inhibitors P15, P16, F I G U R E 3 DEPDC1A does not regulate proliferation of myoblasts. (A) Murine satellite cells were maintained in proliferation medium or differentiation medium for 12 h, 1 or 2 days, and Depdc1a expression measured by RT-qPCR. Depdc1a was expressed in proliferating satellite cell-derived myoblasts, but rapidly down-regulated during differentiation. (B) Depdc1a and Depdc1b were effectively knocked down individually or together via siRNA transfection compared to control siRNA (Si control), as shown by RT-qPCR. (C, D) SiRNA-transfected satellite cells were maintained in proliferation medium, pulsed with EdU and immunolabelled for MYOGENIN. (C) DEPDC1A knockdown did not affect cell proliferation or (D) differentiation, while DEPDC1A/DEPDC1B co-knockdown did not have any additive effects. (E) cDNA from primary human myoblasts obtained from 3 individuals collected in proliferation, and at differentiation day 1, 2 or 3, or (F) immortalized human myoblasts (C25Cl48) in proliferation or at differentiation day 3 or 5 were analysed for DEPDC1A expression by RT-qPCR. DEPDC1A was expressed at significantly higher levels in proliferating human myoblasts than during differentiation. (G) DEPDC1A knockdown via siRNA transfection was efficient, as shown by RT-qPCR. (H) Control (Si control) and DEPDC1A siRNA-mediated knocked down proliferating human myoblasts were pulsed with EdU and immunolabelled for Ki67. DEPDC1A knockdown did not affect myoblast proliferation. (I) SiRNAtransfected human myoblasts were maintained in differentiation medium for 1 day, fixed and immunolabelled for MYOGENIN. DEPDC1A knockdown inhibited the entry into myogenic differentiation, and increased total cells. Data are mean ± SEM, where an asterisk denotes a significant difference (P < .05) between control and a test sample using a paired (A-E) or unpaired (F-I) two-tailed t test, with multiple fields examined from either N = 3 mice, N = 3 individuals or 3 or more independent wells. Scale bar represents 100 µm P19, P21, P27, P53 and P57 were all significantly increased compared to control siRNA ( Figure 4E).

| PITX2 does not control DEPDC1B in human myoblasts
DEPDC1B is repressed by PITX2, with PITX2 knockdown increasing DEPDC1B at the protein level in mouse C2C12 myoblasts. 6 To test whether this regulatory mechanism is conserved in man, we first analysed PITX2 expression. PITX2 expression in immortalized human C25Cl48 myoblasts increased during differentiation ( Figure   S6A). As a repressor of DEPDC1B, such an increase in PITX2 could explain the reduction of DEPDC1B in both human myocytes and myotubes (Figure 2A,B). To test this hypothesis, human immortalized C25Cl48 myoblasts were transduced with retrovirus encoding PITX2C, and overexpression was validated by RT-qPCR ( Figure   S6B). However, retroviral-mediated overexpression of PITX2C did not repress DEPDC1B expression in proliferating human myoblasts ( Figure S6B), and neither did overexpression of PITX1 or PITX3 (data not shown). We also performed the complementary experiment of knocking down PITX2 expression via siRNA and analysing DEPDC1B expression after 2 days of differentiation ( Figure S6C). Knockdown efficiency of PITX2 siRNA was validated (although it also affected Pitx1 and Pitx3), but no increased DEPDC1B expression was measured ( Figure S6C).

| DEPDC1B functions independently of WNT/βcatenin signalling in myogenic cells
In cancer cells, DEPDC1B affects WNT/β-catenin signalling 33 a pathway also involved in satellite cell regulation. 47,48 Therefore, we performed a double DEPDC1B/β-catenin siRNA-mediated knockdown rescue experiment in murine satellite cells, to test whether the reduced proliferation/precocious differentiation observed by knocking down DEPDC1B alone could be attributed to activation of β-catenin.
Efficient knockdown of DEPDC1B and/or β-catenin compared to control siRNA was confirmed by RT-qPCR ( Figure S7A). DEPDC1B knockdown again reduced proliferation (KI67 + ve myoblasts) and enhanced differentiation (MYOGENIN + ve myoblasts), but knockdown of β-catenin did not affect either proliferation or differentiation compared to control siRNA. Knockdown of both DEPDC1B and β-catenin did not rescue the effects on proliferation or differentiation compared to DEPDC1B knockdown alone ( Figure S7B).  Simultaneous DEPDC1B/RHOA knockdown caused a significantly more substantial increase in differentiation of human myoblasts compared to knockdown of either DEPDC1B or RHOA alone, as did double DEPDC1B/PTPRF compared to knockdown of either DEPDC1B or PTRF alone ( Figure 5E). or MYOGENIN expression ( Figure 6B).

| DEPDC1B knockdown increases myogenic differentiation in Rhabdomyosarcoma cells
Next, siRNA-transfected Rh30 cells were maintained in proliferation conditions for 2 days and immunolabelled for KI67 and F I G U R E 5 DEPDC1B and RHOA synergize in suppressing proliferation but enhancing differentiation in human myoblasts. (A) Immortalized human C25Cl48 myoblasts were transfected with control siRNA (Si control) or siRNA against DEPDC1B, RHOA, PTPRF, DEPDC1B/RHO, DEPDC1B/PTPRF or PTPRF/RHOA. Transfected cells were maintained in proliferation medium, mRNA was extracted, and expression of DEPDC1B, RHOA, PTPRF, RB1, MYOD and MYOGENIN was analysed by RT-qPCR. DEPDC1B knockdown cells had increased RB1, MYOD and MYOGENIN expression. RHOA knockdown caused an increase in MYOD and MYOGENIN. However, DEPDC1B/RHOA double knockdown caused a strong additive increase in RB1, MYOD and MYOGENIN expression. (B) SiRNA-transfected human C25Cl48 myoblasts were maintained in proliferation, pulsed with EdU, fixed and immunolabelled for β-TUBULIN. (C) A reduction of the proliferation rate was observed after DEPDC1B knockdown, but not with RHOA knockdown. Double DEPDC1B/RHOA or DEPDC1B/PTPRF knockdown had a synergistic effect in suppressing proliferation. (D) SiRNA-transfected human C25Cl48 myoblasts were maintained in differentiation medium for 2 days, fixed and immunolabelled for MyHC. Differentiation was induced in DEPDC1B knockdown and RHOA knockdown myoblasts, with an additive effect in double DEPDC1B/RHOA or DEPDC1B/PTPRF knockdown cells. Data are mean ± SEM, where an asterisk denotes a significant difference (P < .05) between control and a test sample, or as indicated with a bar, using an unpaired two-tailed t test, with 3 separately siRNA-transfected wells (RT-qPCR) or multiple fields analysed from 3 separate wells (immunolabelling) per condition. Scale bar equals 100 µm

| D ISCUSS I ON
DEPDC1B is a cell cycle-regulated gene that is highly expressed during the G2/M phase of the cell cycle. 1 DEPDC1B plays a role in cell cycle progression and is overexpressed in many cancers. 34 Here, we analysed DEPDC1B function during skeletal myogenesis.
DEPDC1B is expressed in proliferating mouse and human myoblasts, but levels rapidly decrease during differentiation, in agreement with DEPDC1B being almost undetectable in human skeletal muscle. 5 At the protein level, DEPDC1B was present in the nucleus in both proliferating murine myoblasts and differentiated multi- and P57 that inhibit cell cycle progression at different stages. Similar observations have been made in other cells types, for example DEPDC1B knockdown also inhibits proliferation in HeLa 1 and human malignant melanoma cells. 34 In parallel with inhibiting proliferation, DEPDC1B knockdown myoblasts also undergo precocious myogenic differentiation.
DEPDC1A has a similar structure to DEPDC1B and is also regulated during the cell cycle, with DEPDC1A knockdown in HeLa cells causing a similar phenotype to DEPDC1B knockdown, with simultaneous depletion of both DEPDC1A and DEPDC1B having additive effects. 1 We found that DEPDC1A has a similar expression profile to DEPDC1B in mouse and man, with expression clearly falling in differentiation compared to levels during cell cycle. However, DEPDC1A knockdown had no effect on murine myoblast proliferation or entry into differentiation and did not act synergistically with DEPDC1B.
It is of note that in human myoblasts, while DEPDC1A knockdown did not affect EdU incorporation or Ki67 expression, it did lead to an increase in cell number per unit area and fewer cells containing MYOGENIN. Thus, proliferation in myoblasts is supported by DEPDC1B, with DEPDC1A not appearing to have a major role.
To investigate the transcriptional regulation of DEPDC1B, we examined PITX2, which has been reported to be a repressor of DEPDC1B in murine immortalized C2 myoblasts. 6 To test whether this regulatory mechanism was conserved in human myoblasts, we overexpressed PITX2C or knocked down PITX2, but neither F I G U R E 6 DEPDC1B or RHOA knockdown enhance myogenic differentiation in Rhabdomyosarcoma cells. (A) Control human C25Cl48 myoblasts (HCM25), Rh30 and RMS-YM cells were maintained in proliferation medium, mRNA was extracted, and expression of PAX3, PAX7, MYF5, MYOD, MYOGENIN, DEPDC1A and DEPDC1B was analysed by RT-qPCR. DEPDC1A or DEPDC1B were expressed at higher levels in Rh30 cells compared to control human myoblasts. (B) Rh30 cells were transfected with either control siRNA (Si control) or siRNA against DEPDC1B and/or RHOA, mRNA was extracted, and expression of DEPDC1B, RHOA, MYOGENIN and RB1 was analysed by RT-qPCR. DEPDC1B and/or RHOA knockdown were validated, and MYOGENIN was only induced in RHOA knockdown cells and RB1 only in DEPDC1B knockdown cells. No additive effects were observed with DEPDC1B/RHOA double knockdown. (C) SiRNA-transfected Rh30 cells were maintained in proliferation, fixed and immunolabelled for KI67 and MYOGENIN. (D) No reduction in proliferating cells was observed with DEPDC1B knockdown, RHOA knockdown or double DEPDC1B/RHOA knockdown. However, the number of MYOGENIN-positive cells was increased only by RHOA knockdown. (E) SiRNA-transfected Rh30 cells were maintained in differentiation medium for 5 days, fixed and stained with Phalloidin and immunolabelled for MyHC. (F) Cell differentiation was slightly but significantly enhanced by DEPDC1B knockdown, but more robustly increased by RHOA knockdown, with a trend (P = .06) towards an additive effect with double DEPDC1B/ RHOA knockdown compared to RHOA knockdown alone. Data are mean ± SEM, where an asterisk denotes a significant difference (P < .05) between control and a test sample, or as indicated by a bar, using an unpaired two-tailed t test, with 3 separately siRNA-transfected wells (RT-qPCR) or multiple fields per 3 wells (immunolabelling) per condition. Scale bar equals 100 µm procedure affected DEPDC1B expression. Thus, PITX2 does not appear to have a role in direct transcriptional regulation of DEPDC1B in human myoblasts. Further investigations will be required to identify protein/DNA-binding elements for other factors involved in myogenesis located in DEPDC1B regulatory regions.
To better understand how DEPDC1B regulates myoblast proliferation, we investigated pathways described in other systems.
DEPDC1B is overexpressed in non-small-cell lung carcinoma, where it can enhance both cell migration and invasion, an effect partially due to activation of canonical WNT signalling through β-catenin. 33 In mouse, while Wnt/β-catenin signalling is transiently active in myoblasts during muscle regeneration in adult, genetic inactivation of β-catenin in satellite cells does not affect regeneration. 49 Interestingly though, constitutive activation of β-catenin in satellite cells extends the myoblast phase of regeneration to ultimately generate smaller fibres and increase muscle fibrosis. 49 We previously reported that inhibiting Axin1/2 in satellite cells increases signalling via β-catenin to cause inhibited proliferation and precocious entry into differentiation. 47 Thus, DEPDC1B knockdown caused a phenotype similar to activation of β-catenin. 47 We therefore tried to rescue DEPDC1B knockdown by also down-regulating β-catenin, but DEPDC1B/β-catenin double knockdown did not rescue proliferation or precocious differentiation. Therefore, DEPDC1B function is not directly associated with β-catenin levels in skeletal myogenesis. Knocking down both genes therefore allows a synergetic induction of differentiation, with withdrawal from cell cycle being a prerequisite for myogenic differentiation.
RHOA promotes myogenic differentiation but needs to be down-regulated to then allow fusion. 8,11,13,[15][16][17] In human myoblasts, we found that RHOA knockdown does not affect proliferation, but causes precocious entry to differentiation, suggesting an inhibitory effect of RHOA on induction of myogenic differentiation. The active form of RHOA is high in proliferation and at the later stages of differentiation, but not at the initiation of differentiation, 50 supporting a potential inhibitory role of RHOA on differentiation. However, in proliferating murine C2C12 myoblasts, overexpression of a dominant-negative RHOA (N19-RhoA) 11 or RHO-specific inhibitor (tat-C3) 17 inhibits myogenic differentiation. These contradictory results could be due to cell line differences (immortalized murine C2C12 versus immortalized human myoblasts) or in the strategy to either inactivate RHOA or reduce its levels via siRNA. A specific level and temporal activation of RHOA seem to be important to allow coordinated myogenic progression. 17 DEPDC1B is overexpressed in many cancers, 3 There are other pathways that DEPDC1B may also interact with. RAC1 is regulated by DEPDC1B in some cell lines 5,6 and in muscle cells, and RAC1 inhibits myogenic differentiation and induces loss of cell contact inhibition. Moreover, RAC1 is activated in rhabdomyosarcoma cell lines, and overexpression of dominant-negative forms inhibits cell proliferation of RMS. 22 Therefore, a potential interaction of DEPDC1B and RAC1 could occur in muscle. DEPDC1B overexpression also increases phosphorylation of ERK, 3,5

CO N FLI C T O F I NTE R E S T
The authors declare no potential conflicts of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.