Regulation of muscle differentiation and survival by Acheron

MECHANISMS OF DEVELOPMENT

1 2 6 ( 2 0 0 9 ) 7 0 0 –7 0 9

available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/modo

Regulation of muscle differentiation and survival by Acheron Zhaohui Wanga, Honor Glennb, Christine Brownc, Christos Valavanisc,d, Jing-Xia Liuc, Anandita Setha,c, Jeanne E. Thomasc, Rolf O. Karlstroma,c, Lawrence M. Schwartza,b,c,* a

Molecular and Cellular Biology Program, Morrill Science Center, University of Massachusetts, Amherst, MA 01003, USA Pioneer Valley Life Sciences Institute, 3601 Main Street, Springfield, MA 01199, USA c Department of Biology, Morrill Science Center, University of Massachusetts, Amherst, MA 01003, USA d Molecular Pathology and Genetics Division, Department of Pathology, Metaxa Cancer Hospital, Botassi 51, Piraeus 185 37, Greece b

A R T I C L E I N F O

A B S T R A C T

Article history:

Acheron (Achn), a phylogenetically-conserved member of the Lupus antigen family of RNA

Received 9 February 2009

binding proteins, was initially identified as a novel cell death-associated gene from the

Received in revised form

intersegmental muscles of the tobacco hawkmoth Manduca sexta. C2C12 cells are a standard

15 May 2009

model for the study of myogenesis. When deprived of growth factors, these cells can be

Accepted 20 May 2009

induced to: form multinucleated myotubes, arrest as quiescent satellite-like reserve cells,

Available online 28 May 2009

or undergo apoptosis. Achn expression is induced in myoblasts that form myotubes and acts upstream of the muscle specific transcription factor MyoD. Forced expression of ecto-

Keywords:

pic Achn resulted in the formation of larger myotubes and massive reserve cell death rel-

Muscle

ative to controls. Conversely, dominant-negative or antisense Achn blocked myotube

Apoptosis

formation following loss of growth factors, suggesting that Achn plays an essential, per-

MyoD

missive role in myogenesis. Studies in zebrafish embryos support this hypothesis. Reduc-

Myf5

tion of Achn with antisense morpholinos led to muscle fiber loss and an increase in the

Stem cell

number of surviving cells in the somites, while ectopic Achn enhanced muscle fiber forma-

Manduca

tion and reduced cell numbers. These results display a crucial evolutionarily conserved role

Zebrafish

for Achn in myogenesis and suggest that it plays key roles in the processes of differentia-

Myogenesis

tion and self-renewal.  2009 Elsevier Ireland Ltd. All rights reserved.

1.

Introduction

Vastly more cells are produced during embryogenesis than are ultimately required for organogenesis (Domingos and Steller, 2007). However only limited information exists about the signaling molecules that regulate the decision to differentiate or die. To identify candidate proteins, we exploited the intersegmental muscles (ISMs) of the moth Manduca sexta (Schwartz, 2008). The ISMs are composed of giant syncitial cells that die during the 30-h period following adult emergence at the end of metamorphosis. Using a differential

cloning strategy, we and others have identified a number of genes that are dramatically up-regulated coincident with the commitment of the ISMs to die, including: polyubiquitin (Schwartz et al., 1990), proteasome subunits (Dawson et al., 1995; Jones et al., 1995; Sun et al., 1997), apolipophorin III (Sun et al., 1995), Small Cytoplasmic Leucine-Rich Repeat Protein (Kuelzer et al., 1999) and Death-Associated LIM-Only Protein (Hu et al., 1999). One of the novel genes isolated in this screen encodes Acheron (Achn). Achn is structurally similar to the Lupus Antigen (LA) RNA binding protein and defines a new

* Corresponding author. Address: Department of Biology, Morrill Science Center, University of Massachusetts, Amherst, MA 01003, USA. Tel.: +1 (413) 545 2435; fax: +1 (413) 545 3243. E-mail address: [email protected] (L.M. Schwartz). 0925-4773/$ - see front matter  2009 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.mod.2009.05.003

MECHANISMS OF DEVELOPMENT

3subfamily of LA proteins (Valavanis et al., 2007). Yeast twohybrid analysis demonstrated that Achn binds to CASK/Lin2, a protein that can shuttle between the membrane and nucleus to regulate developmental decisions (Weng et al., 2009). In mammals, Achn is expressed predominantly in neurons, skeletal muscle and cardiac muscle. Achn expression is also elevated in some basal-like breast cancers in humans, and ectopic expression of Achn can drive tumor growth both in vitro and in xenograph models in vivo (Shao et al., unpublished observations). Since Achn was originally isolated from skeletal muscle, we exploited C2C12 cells and zebrafish embryos for in vitro and in vivo analyses of myogenesis, respectively. C2C12 cells were originally derived from adult mouse muscle and behave like satellite cells (Yaffe and Saxel, 1977). These cells proliferate indefinitely in a trophic factor-rich growth medium (GM). Following transfer to a low serum differentiation medium (DM), C2C12 cells choose one of three independent developmental fates. One population of cells expresses the basic helix-loop-helix (bHLH) transcription factor MyoD, exits the cell cycle, expresses the bHLH protein myogenin, and terminally differentiates into multinucleated myotubes (Thayer et al., 1989; Tapscott, 2005). These cells also up-regulate the expression of retinoblastoma protein (Rb) and p21, both of which serve to enhance cell survival and promote differentiation by increasing MyoD stability (Delgado et al., 2003; Shen et al., 2003). A second population of cells expresses the bHLH protein Myf5, enters G0, and expresses the anti-apoptotic protein Bcl-2 (Dominov et al., 1998, 2001; Yoshida et al., 1998). These ‘‘reserve’’ cells appear to be the in vitro counterpart of satellite cells, a pool of lineage-restricted muscle stem cells that facilitate muscle repair following injury (Morgan and Partridge, 2003; Ehrhardt et al., 2007; Peault et al., 2007). The last population of cells in the cultures fails to activate programs for either survival or differentiation and initiates apoptosis (Dee et al., 2002). Our loss- and gain-of-function analyses in C2C12 cells suggest that Achn acts early in myogenesis at a point upstream of MyoD to control developmental fate. Expression of ectopic Achn results in the formation of larger muscle fibers and the almost complete loss of the reserve/satellite pool. In contrast, blockade of Achn prevents myotube formation and allows myoblasts to survive even in the absence of growth factors. Manipulating Achn expression in zebrafish embryos produced comparable results. Taken together, these studies suggest that Achn may be an important intrinsic factor controlling early myogenesis.

2.

Results

2.1. Biological effects of Acheron mis-expression on C2C12 cells and primary myoblasts Achn expression in C2C12 cells maintained in serum rich growth medium (GM) was below detectable levels by Western blotting (Fig. 1A) and close to the limit of detection via immunocytochemistry (Fig. 1B, control). Following transfer to a low serum differentiation medium (DM), the levels of Achn protein increased in parallel with the formation of multinucleated

1 2 6 ( 2 0 0 9 ) 7 0 0 –7 0 9

701

Fig. 1 – Achn expression in C2C12 cells. (A) Western blot analysis. Total protein was collected from C2C12 cells cultured in GM and after 1, 2, 4 and 6 days after transfer to DM. In separate experiments mild trypsinization was used to selectively detach myotubes from plates while leaving the more adherent mononucleated reserve cells attached. Proteins were subjected to Western analysis for expression of Achn, myosin heavy chain (MHC) a muscle differentiation marker, and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) (loading control). Molecular weights are in kDa. (B) Immunocytochemistry. C2C12 cells were fixed and immunolabeled with a polyclonal antibody against Achn. Signal was visualized by DAB staining. Vector control cells (control); Achn-engineered cells (Achn); truncated Achn (tAchn) expressing cells: Achn anti-sense expressing cells (As-Achn). Scale bar indicates 20 lm.

myotubes (Fig. 1A). To determine if Achn expression was predominantly localized to myotubes, we used mild trypsinization

702

MECHANISMS OF DEVELOPMENT

to partially separate the myotubes from the more adherent mononucleated reserve cells. The level of Achn in the myotube enriched fraction was substantially higher than in the reserve cell fraction, and paralleled the distribution of myosin heavy chain, a terminal differentiation marker for skeletal muscle (Fig. 1A). Note that the reserve cell enriched fraction did contain residual myotubes as evidenced by MHC expression, a protein that is not detected in undifferentiated myoblasts. To understand the role that Achn might play in myogenesis, we generated monoclonal C2C12 cell lines that were stably transfected with expression vectors encoding: FLAG-tagged full-length Achn (Achn), a FLAG-tagged Achn lacking the first 33 amino acids (truncated Achn; tAchn), antisense Achn (AsAchn) or empty vector as a control. The expression of ectopic protein was verified by both immunocytochemistry with an anti-Achn antiserum (Fig. 1B, Achn, tAchn) and by Western blotting with an anti-FLAG monoclonal antibody (data not shown). Expression of As-Achn mRNA was verified by reverse transcriptase PCR, but its impact on the levels of endogenous Achn protein could not be quantified because of the low basal

1 2 6 ( 2 0 0 9 ) 7 0 0 –7 0 9

levels of protein we detect via Western blotting of control cells (Fig. 1A and B As-Achn). C2C12 cells were maintained in GM until they reached about 80% confluency and then transferred to DM. After 3 days in DM, the cells were fixed and stained for myosin heavy chain (MHC), filamentous actin, and DNA. Control cultures contained numerous thin, multinucleated, MHC-positive myotubes (Fig. 2A, control). Mononucleated reserve cells rich in actin stress fibers occupied spaces between the myotubes. Cells expressing ectopic Achn produced myotubes that were substantially larger than those seen in control cultures (Fig. 2A, Achn). The interstitial mononucleated reserve cell population was largely absent from these cultures, resulting in large acellular spaces between the myotubes, a phenomenon that is never seen in control cultures. When C2C12 cells expressed either truncated or antisense Achn mRNA, a dramatically different set of phenotypes was observed (Fig. 2A, tAchn and As-Achn). In these cultures, myotubes were largely absent and the plates were composed of an almost confluent monolayer of reserve cells. Comparable data were obtained

Fig. 2 – Effects of Acheron mis-expression on myoblast differentiation. (A) C2C12 cells were cultured in differentiation medium for 3 days and then fixed and tripled labeled: actin (red); MHC (green); and DNA (blue). Vector control cells (control); Achn-engineered cells (Achn); truncated Achn expressing cells (tAchn); Achn anti-sense expressing cells (As-Achn); Achn cells re-transfected with tAchn (Achn + tAchn), Achn cells re-transfected with As-Achn (Achn + As-Achn). (B) Differentiation index. Control and Achn-engineered C2C12 were evaluated after 3 days of differentiation for the percentage of nuclei within a field that were contained within MHC-positive myotubes. Values represent the means of three independent experiments. Error bars indicate SEM. (C) Differentiating C2C12 were evaluated for the total number of nuclei per myotube. Values are the mean of three independent experiments +/ SEM.

MECHANISMS OF DEVELOPMENT

with Achn-engineered primary mouse myoblasts obtained from neonatal mice (data not shown). In order to quantify these observations, we calculated the differentiation index for each C2C12 cell line by determining the ratio of nuclei seen in MHC-positive myotubes relative to the total number of nuclei in each field. We found that ectopic expression of Achn increased differentiation by about 65% (Fig. 2B) relative to control. In contrast, the expression of truncated or antisense Achn decreased differentiation to 6% and 25% of the control values, respectively (Fig. 2B). We also observed differences in the number of nuclei recruited into individual myotubes. Expression of ectopic Achn resulted in an approximately 35% increase in the average number of nuclei per myotube (within each field of view), whereas expression of either tAchn or As-Achn reduced this value by about 50% (Fig. 2C). Since comparable results were obtained with both truncated and antisense Achn, we speculated that tAchn functions as a dominant-negative regulator of endogenous Achn function. To test this hypothesis, C2C12 cells expressing ectopic Achn were re-transfected with vectors encoding tAchn, As-Achn or nothing, and five monoclonal transfected lines were isolated for each construct. The introduction of either tAchn or As-Achn abrogated the effects of ectopic Achn expression, with tAchn being more effective than As-Achn (Fig. 2A, Achn + tAchn and Achn + As-Achn). These cultures displayed reduced myotube formation and the retention of a greater proportion of mononucleated cells relative to cells expressing exogenous Achn alone.

2.2. Myogenic pathways affected by mis-expression of Acheron The effect of manipulating Achn expression on the formation of myotubes was also mirrored at the biochemical level. Coincident with the formation of multinucleated myotubes in control cultures, there was a dramatic increase in the expression of MHC (Fig. 3A). MHC levels were higher still in cells expressing ectopic Achn. Conversely, cells expressing tAchn failed to induce significant MHC expression, and AsAchn cells had greatly reduced MHC. In addition, expression of tAchn or As-Achn was able to inhibit the enhanced expression of MHC induced by ectopic Achn. In these experiments and the others we performed, tAchn was a more potent inhibitor of differentiation than As-Achn. Cycling C2C12 cells express both MyoD and Myf5 but at different phases of the cell cycle (Kitzmann et al., 1998; Lindon et al., 1998). When trophic factors are removed, expression of MyoD and Myf5 becomes restricted to myotubes and reserve cells, respectively. Our morphological and biochemical analyses suggest that Achn pushes myoblasts towards either differentiation or death at the expense of reserve cells. To determine if this effect might be mediated by these bHLH transcription factors, we examined the expression of MyoD and Myf5 in our Achn-engineered C2C12 cells. A dramatic increase in MyoD was observed in control cells 24 h after transfer to DM (Fig. 3B). Achn engineered cells expressed elevated levels of MyoD even in GM, supporting the hypothesis that Achn pushes the cells towards differentiation. In contrast, tAchn blocked MyoD induction in DM. To determine if Achn functions upstream of MyoD, tAchn-expressing C2C12 cells

1 2 6 ( 2 0 0 9 ) 7 0 0 –7 0 9

703

Fig. 3 – Changes in muscle-specific gene expression Acheron engineered cells. (A) Total protein was collected from cells cultured in GM or after 3 days in DM for all six lines described in Fig. 1 above and subjected to Western analysis for MHC expression. (B)Western blot analysis of MyoD and Myf5 proteins in: control, Achn, and tAchn cells. Total proteins were collected in GM (G) and DM for 1, 2 and 3 days. Three days after transfer to DM, floating apoptotic cells were removed prior to processing in lane ‘‘A’’. (C) MyoD expression in tAchn cells without (left) and with (right) ectopic MyoD. DM sample was taken at day 3. (D) MHC expression in tAchn cells without (left) and with (right) ectopic MyoD expression. Samples were examined 3 days after transfer to DM.

were stably re-transfected with a MyoD expression construct and the expression of ectopic MyoD was confirmed by Western blotting (Fig. 3C). Following transfer to DM, these cells produced large numbers of MHC-positive myotubes (Fig. 3D), supporting the hypothesis that Achn acts upstream of MyoD and is required for normal MyoD expression in myoblasts. Ectopic Achn blocked the expression of Myf5 protein, resulting in a 40% decrease relative to controls (Fig. 3B), consistent with the loss of reserve cells from the population. When the floating apoptotic cells were removed from these cultures, Myf5 was almost undetectable (Lane A; Fig. 3B), suggesting that surviving Myf5-positive reserve cells were essentially absent. In contrast, tAchn expression resulted in a substantial increase in endogenous Myf5 protein expression relative to controls.

2.3.

Acheron regulation of cell survival

One of the dramatic effects of expressing ectopic Achn in C2C12 cells was the absence of the reserve cell population

704

MECHANISMS OF DEVELOPMENT

(Fig. 2A, Achn). The failure to accumulate reserve cells could have resulted from the selective death of mononuclear cells or instead their recruitment into myotubes. To discriminate between these two possibilities, we used vital dye staining to quantify the levels of cell death within the cultures. In control cultures, cell death was low in GM and then increased about threefold to about 15% by 48 h after transfer to DM (Fig. 4A). Cells expressing ectopic Achn had a higher basal level of cell death in GM than control cells. By 48 h after transfer to DM, the level of cell death was about 40%, almost three times that seen in control cells. Caspase-3 assays were performed with these cultures as a direct assay of apoptosis (Fig. 4B). As predicted from the data presented above, caspase-3 activity was low in control cells cultured in GM, and then increased over time following transfer to DM. The absolute levels of caspase-3 activity were far greater in the Achn-expressing cells under all conditions examined, while expression of tAchn reduced caspase-3 activity. Taken together, these data support the hypothesis that Achn drives reserve cell death and that cell loss occurs via apoptosis.

1 2 6 ( 2 0 0 9 ) 7 0 0 –7 0 9

achn expression and function during somitogenesis. Achn is expressed strongly in the central nervous system in a regionally restricted pattern (L.M.S. and R.O.K., unpublished data) and more weakly in pre-somitic mesoderm and forming somites, primarily in mesenchymal cells (Fig. 5). To determine if Achn is required for normal myogenesis, we injected 2-cell zebrafish embryos with either in vitro transcribed zebrafish achn mRNA or achn translation blocking morpholinos (MOs) and then assayed development at 24 and 48 hours post-fertilization (hpf). Two independent translation blocking MOs produced comparable defects in myogenesis, and a mixture containing both of the MOs at concentrations 2–3-fold lower than the effective dose for each individual MO produced identical phenotypic defects. These data strongly support the hypothesis that the actions of the MOs resulted from specific effects on achn mRNA rather than as a result of some non-specific toxicity. While achn mRNA injection had little effect on the gross

2.4. Acheron is required for muscle development during zebrafish embryogenesis zTo determine if these data were an artifact of in vitro manipulation, we turned to zebrafish embryos to examine

Fig. 4 – Effects of Acheron mis-expression on cell death. (A) Trypan blue staining was used to quantify cell death in control and Achn cells in GM, and days 1 and 2 in DM. (B) Caspase-3 activity was measured in C2C12 cells in GM, and 12 or 24 h after transfer to DM. Open bars = control; black bars = Achn; and hatched bars = tAchn. Results were normalized to total cell number. One of three representative experiments is presented. Mean +/ SEM.

Fig. 5 – Expression of achn in developing zebrafish somites. (A) Dorsal view of the trunk region at the 5 somite stage (12 h post fertilization (hpf)), anterior left. achn mRNA is expressed at low levels in pre-somitic mesoderm (psm), with more expression medially near the notochord (nc) than laterally. In the first somites that form, achn is expressed in mesenchymal cells (arrowhead), with little expression in epithelialized border cells (arrow). (B) Lateral view of the trunk at the 10 somite stage (15 hpf), anterior left. achn expression is maintained in a mosaic fashion in the mesenchyme (arrowhead), with little expression at somite borders (arrow). Black lines outline somites. Scale bars: 50 lm.

MECHANISMS OF DEVELOPMENT

morphology of embryos, MO-injected embryos were smaller than controls and had visibly disrupted somites (Fig. 6). Despite these defects, MO injected embryos were relatively well formed, indicating early patterning and organogenesis were not generally disrupted. MO injection resulted in a substantial reduction in somite size and affected the development of both slow and fast fibers (Fig. 6). We first examined the effects of achn mRNA injection on the development of embryonic slow twitch muscle fibers at 24 hpf. This superficial group of muscles contains only a small number of fibers, thus allowing detection of structural variations of individual fibers in intact whole-mount embryos. Expression of ectopic achn did not appreciably change the number of fibers generated (control = 9.71 ± 0.34, n = 20; achn mRNA injected = 9.59 ± 0.83, n = 21; P = 0.54) (Fig. 6B), although

1 2 6 ( 2 0 0 9 ) 7 0 0 –7 0 9

705

it did result in a slight increase in muscle fiber diameter (Fig. 6J). In contrast, injection of achn MOs resulted in a reduction in the number of slow muscle fibers (6.75 ± 0.49, n = 14) in the majority of injected embryos (Fig. 6C). Approximately 15% of the MO-treated embryos had no discernible slow-twitch muscle fibers and this group was excluded from our fiber analysis, suggesting that the effects of achn MOs were even more significant than reported. The slow fibers that did form in achn MO injected embryos were atrophic and had less than half the diameter of the fibers seen in the controls (Fig. 6J). In addition, this small population of fibers often displayed indistinct Z bands, suggesting defects in sarcomere structure, and perhaps contractility. These data support the hypothesis that Achn plays an essential role in myogenesis in zebrafish embryos.

Fig. 6 – Acheron function in zebrafish myogenesis. (A–C) Lateral views of 24 hours postfertilization (hpf) zebrafish embryos labeled with the F59 antibody to reveal slow twitch muscle fibers, white lines show extent of one somite (S). achn mRNA injected embryos (B) had well formed slow fibers (red) and slightly larger somites than seen in uninjected control embryos (A). achn MO injected embryos (C) had reduced somites with less organized slow fibers. (D,F and H) Trunk cross-sections of 48 hpf embryos. achn mRNA injected embryos (F) had larger somites than uninjected embryos (D) with well formed fast fibers throughout the trunk. In contrast, achn MO injected embryos (H) had highly reduced somites with poorly formed fibers. (E,G and I) Confocal images of To-Pro3 stained 48 hpf embryo cross-sections. Left somite is outlined. achn mRNA injection (G) led to a decrease in nuclei. In contrast, ach MO injected embryos (I) had more nuclei despite having smaller somites. (J) Mean diameter (lm) of slow twitch fibers in uninjected, achn mRNA injected, and achn MO-injected embryos. Error bars indicate SEM. (K) Mean number of To-Pro3 labeled nuclei in somite regions of 12 lm sections. Error bars indicate SEM. In A-C scale bar = 20 lm, D–I scale bar = 50 lm. DIC, differential interference contrast microscopy; S, somite.

706

MECHANISMS OF DEVELOPMENT

Since embryonic slow twitch fibers are mononuclear (Roy et al., 2001) we examined the role of Achn in somite formation at 48 hpf, when multinucleated fast twitch muscles comprise the bulk of somite volume. Injection of achn mRNA resulted in embryos with slightly larger somites containing well developed, robust muscle fibers (Fig. 6F). In contrast, achn MO injection led to a major reduction in somite size, which is accounted for by a loss of organized fast fibers (Fig. 6H). Thus, Achn appears to regulate the formation of both fast and slow twitch fiber formation in vivo. Based on our studies of C2C12 cells, we predicted that loss of Achn function would not only reduce muscle fiber formation but also result in the enhanced survival of mononucleated myoblasts within the somites. To begin to test this hypothesis we counted the number of nuclei present in the somites of embryos with enhanced or reduced Achn expression. Ectopic expression of achn led to a decrease in the number of nuclei persisting in each segment (Fig. 6G and K). In contrast, embryos injected with achn MOs had approximately 30% more persisting cells within each somite than did control animals (Fig. 6I and K). This increase in cell number is in stark contrast to the smaller somite size found in these injected animals and further supports the hypothesis that the defects observed in achn MO-injected embryos specifically result from the loss of Achn function. Taken together, the results of our in vitro and in vivo analysis strongly support the hypothesis that Achn plays an early and essential role in myogenesis and may play a role in the regulation of muscle-restricted stem cell survival.

3.

Discussion

Achn was identified as a novel cell-death associated gene in a screen for transcripts that are induced when the ISMs of moth become committed to die (Valavanis et al., 2007). Like other genes isolated in this screen (Schwartz, 2008), the induction of Achn is tightly coupled to the commitment of the cells to die, and endocrine manipulations that delay death also block the expression of this transcript. However, Achn was also expressed in several tissues that were not fated to die, suggesting that it is not a component of the execution phase of the cell death program but instead serves as a signal transduction molecule that can regulate a number of distinct developmental and homeostatic processes, one of which is death. In the case of Manduca, up-regulation of Achn may enhance the ability of the muscles to overcome survival programs and initiate cell death as is seen in C2C12 cells (Fig. 4). Since Manduca is not amenable to molecular and genetic manipulations, we turned to C2C12 cells to analyze the role of mammalian Achn in muscle development and death. Achn expression was essentially undetectable in cells cultured in the presence of growth factors (GM) but then accumulated in myotubes following transfer to DM (Fig. 1A). In an earlier report we noted that there was no change in Achn expression in C2C12 cells following transfer to DM (Valavanis et al., 2007). The data presented in the current manuscript were generated with a new anti-Achn antiserum that appears much more specific than the one we originally developed, and several lines of evidence support the conclusion that Achn expression is developmentally regulated in C2C12 cells.

1 2 6 ( 2 0 0 9 ) 7 0 0 –7 0 9

When sub-confluent cultures were provided with adequate levels of growth factors, genetic alterations of Achn expression failed to induce significant changes in either cell growth or survival. However, expression of ectopic Achn resulted in greatly elevated levels of MyoD relative to control cells cultured in GM (Fig. 3). Interestingly, this enhanced expression of MyoD did not induce these cells to prematurely differentiate into myotubes in GM, suggesting that other growth factorregulated proteins serve to repress myogenesis (Kitzmann and Fernandez, 2001; Tapscott, 2005). However, when growth factors were withdrawn, the Achn-expressing cells were primed to differentiate; they had a greater index of differentiation and generated myotubes that were substantially larger than those normally produced in culture (Fig. 2A and B). The increase in myotube diameter may be due in part to an increase in the number of nuclei in each myotube (Fig. 2C), however an increase in biosynthesis per nucleus cannot be ruled out as an additional contributing factor. This enhanced formation of myotubes is surprising given that the number of neighboring cells available to align with myotubes and be recruited to fuse with them was greatly diminished in these cultures due to the increased apoptotic death of the mononucleated cells (Fig. 4). Expression of Myf5 in C2C12 cells is restricted to reserve cells, which like their in vivo counterpart satellite cells, arrest as self-renewing muscle-restricted stem cells. Ectopic Achn repressed the expression of Myf5 which presumably played a role in the selective loss of this cell population in vitro. These cells died by apoptosis as demonstrated by both their morphology and the activation of caspase-3. This hypothesis is further supported by the demonstration that Myf5 was preferentially expressed in the floating pre-apoptotic cells (Fig. 3). The loss of this subpopulation of cells was readily apparent in the cultures and resulted in bare patches on the substrate, a phenomenon that is never seen in normal cultures. While enhanced expression of Achn drove both myogenesis and the apoptotic death of the reserve cells, interference of Achn function with antisense, siRNA (data not shown) or tAchn had the opposite effect and inhibited both myotube formation and cell death. Instead, these cells arrested as mononucleated cells and retained the capacity to cycle when placed in GM. The failure to form myotubes appears to be an early developmental decision upstream of MyoD, since these engineered cells failed to upregulate MyoD following transfer to DM. This hypothesis is supported by the observation that expression of ectopic MyoD could drive the formation of MHC-positive multinucleated myotubes in cultures that constitutively express tAchn (Fig. 3C). Antisense and tAchn most likely reduce Achn function by different mechanisms. Truncated Achn appears to behave as a dominant-negative regulator. The only difference between wild type and tAchn is the loss of the first 33 amino acids at the N-terminus. This sequence lacks any discernable functional motifs but is just upstream of six putative serine residues with high (Procite) probability of phosphorylation (S51, S56, S58, S63, S67 and S72). We speculate that Achn may function in a regulatory complex that requires phosphorylation in order to facilitate the release and/or activation of MyoD or other key downstream regulatory proteins.

MECHANISMS OF DEVELOPMENT

To extend our analysis in vivo, we turned to somitogenesis in zebrafish embryos. Achn is expressed in the somites of zebrafish embryos (Fig. 5) and ectopic expression of achn led to the formation of fast and slow muscle fibers that were both larger and better developed than those produced in nonmanipulated embryos (Fig. 6). Ectopic achn also reduced the number of nuclei resident within the somites, consistent with a role in driving myoblast cell death. In contrast, blockade of endogenous achn expression with either of two antisenseMOs led to a reduction in muscle fiber formation in both the fast and slow twitch muscles and a concomitant increase in the number of somite nuclei. This retention of supernumerary cells is striking and suggests that the failure to produce robust muscle fibers was not an artifact of injection. Indeed, we were able to fully replicate all of the phenotypic effects of either MO by injecting them together at a concentration that was sub-phenotypic for either one alone. Taken together, these data support the hypothesis that Achn acts early in myogenesis as a key regulator of both survival and differentiation. In the presence of Achn, myoblasts acquire the capacity to either differentiate or die when growth factors become limiting. When endogenous Achn is reduced, cells lose the ability to activate either of these two programs and instead arrest as mitotically competent reserve/satellite cells. Thus, Achn appears to force myoblasts to lose their stem cell-like properties. Given its role in the differentiation of muscle, it is possible that mis-regulation of Achn expression or function could result in pathology of mesodermally derived cells. For example, rhabdomyosarcomas, the most prevalent childhood sarcoma (Gallego Melcon and Sanchez de Toledo Codina, 2005) arise in part from the apparent failure of myoblasts to differentiate or die during development. It also seems likely that Achn will play a role in developmental decisions in other lineages as well given that achn MO injection results in focal cell death within the developing zebrafish forebrain (data not shown). In addition, Achn expression is elevated in some human basal-like breast cancers and manipulations of Achn expression in breast cancer cells has dramatic effects on their ability to grow and initiate angiogenesis (Shao et al., unpublished observations). Thus, Achn may be responsible for controlling diverse developmental, homeostatic and pathological processes.

4.

Experimental procedures

4.1.

Achn constructs

Full length and N-terminally truncated (tAchn) human Achn lacking the first 33 amino acids were cloned in-frame in pFLAG-CMV2-neo vector and then transferred to the pBabe retroviral vector. Both sense and antisense constructs were isolated. Zebrafish achn was identified by BLAST analysis of the Zebrafish Sequencing Project. Both 5 0 -UTR (AGGCGCCT GGGGAGTTTGTGTTTTC) and 3 0 -UTR (TGCGGCGATGCCATCAT TCAGTC) PCR primers were used to amplify the full-length cDNA from RNA isolated from 72 hours post fertilization (hpf) embryos using a hot start RT-PCR kit (Qiagen). The cDNA was sequenced and sub-cloned into pCR2.1 vector using a TA cloning kit (Invitrogen).

1 2 6 ( 2 0 0 9 ) 7 0 0 –7 0 9

4.2.

707

Cell culture

C2C12 cells were cultured in growth medium (GM) consisting of Dulbecco’s modified Eagle medium (DMEM) supplemented with 15% (V/V) calf serum (CS), 5% fetal bovine serum (FBS, Atlanta Biologicals) and 100 units/ml of penicillin–streptomycin (Gibco). To induce differentiation, subconfluent cultures were shifted to differentiation medium (DM) consisting of DMEM supplemented with 2% horse serum and 100 units/ml of penicillin–streptomycin. LipofectAmine (Gibco) was used to transfect C2C12 cells with Achn, tAchn, As-Achn or empty vector. The antibiotic– resistant stable transformants were selected in GM with G418 (500 lg/ml) or puromycin (3 lg/ml). About 10 monoclonal lines were randomly chosen from each transfection and further analyses were performed with typical individual lines. In some experiments, transfected cells were re-transfected with pBabe-hygro-MyoD or empty pBabe-hygro vector and stable cells selected with 250 lg of hygromycin B/ml (Sigma). Similarly, pBabe-puro-tAchn or pBabe-puro-AsAchn were stably transfected into neomycin-resistant full-length Achn over-expressing cells and selected with puromycin (3 lg/ml).

4.3.

Immunocytochemistry and immunofluorescence

C2C12 cells were grown on glass coverslips and then fixed for 3 min in cold acetone, air-dried, rehydrated in TBS (0.14 M NaCl, 2.7 mM KCl, 24.7 mM Tris-base, pH 7.4) and then incubated for 20 min in 3% hydrogen peroxide. Cells were labeled manually with Dakocytomation LSAB2 System-HRP kit according to the manufacturer’s instructions. Fixed cells were blocked for 20 min (Dako protein block), then incubated with a polyclonal rabbit anti-Achn antiserum (Kane et al., unpublished) at 1:25 dilution in Dako antibody diluent and then incubated with Dako anti-rabbit HRP polymer, followed by diaminobenzidine (Dako DAB+) for 27 min. Cells were rinsed 3· in TBS and the coverslips mounted on slides in 90% glycerol and imaged at 40· magnification with a Spot Insight digital color camera mounted on an Olympus BX1 microscope. For immunofluorescent labeling, C2C12 cells were grown in polystyrene slide flasks until approximately 90% confluent and then switched to DM and cultured for 3 more days. Cells were fixed for 2 min with 2% paraformaldehyde in PBS, permeabilized for 3 min with Karsenti’s lysis buffer (0.5% Triton X-100, 80 mM PIPES, 1.0 mM MgSO4, 5.0 mM EGTA, pH 7.0), then fixed for an additional two minutes in 2% paraformaldehyde. Cells were rinsed three times in PBST (PBS plus 0.05% Tween 20, pH 7.4) between each processing step. Actin was labeled with rhodamine-conjugated phalloidin (Invitrogen) in PBST. Cells were incubated with anti-muscle-specific myosin heavy chain (MHC) antibody (MF20, Developmental Studies Hybridoma Bank) diluted 1:50 followed by FITC-conjugated secondary antibody (1:100, Jackson ImmunoResearch Laboratories). Nuclei were stained with To-Pro 3 (Invitrogen) (1:1000). Cells were imaged with a Nikon C1 scanning confocal system with a Nikon TE 2000 inverted microscope using a 40· oil immersion objective lens. Images were processed with the C1 EZ-Viewer software.

708

4.4.

MECHANISMS OF DEVELOPMENT

Differentiation index

C2C12 cell cultures were stained for MHC and DNA as described above, and then the percentage of nuclei contained within MHC-positive myotubes was determined. Five fields were counted per slide for a total of 400–1000 nuclei per slide. Average values for three independent experiments were pooled and standard error from the mean (SEM) calculated. These same images were also analyzed to determine the average number of nuclei per myotube by dividing the number of nuclei present in MHC positive myotubes by the total number of myotubes in each field.

4.5.

Western blot analysis

Cells were extracted in Laemmli buffer and the protein concentration determined using a BCA kit (Pierce). Proteins (30 lg) were fractionated by SDS/10% polyacrylamide gel electrophoresis and then transferred to Immobilon P membrane (Millipore). Membranes were blocked in 5% nonfat dry milk in PBST and reacted with a primary antisera directed against the following proteins: Acheron (1:25), MHC (1:50, Developmental Studies Hybridoma Bank), GAPDH (1:5000, Alpha Diagnostics), MyoD (1:500, Pharmingen, or 1:200, sc-304, Santa Cruz), and Myf5 (1:200, sc-302, Santa Cruz). Antigens were detected with a horseradish peroxidase-conjugated goat anti-mouse or anti-rabbit antibody (1:2000, Vector Laboratories or 1:1000 Jackson ImmunoResearch Laboratories) followed by ECL chemoluminescence (NEN) and X-ray film autoradiography. For some experiments, myotubes were selectively collected from cultures by incubating the cells in a 1:1 ratio of 0.05% trypsin/EDTA:PBS for about 2 min, or until most myotubes detached. The myotubes were collected, and the remaining adherent mononucleate reserve cell-enriched fraction was then scraped and processed separately.

4.6.

Cell death assays

Trypan blue assays were used as previously (Hu et al., 1999). The percentage of cell death was determined by counting each cell line in triplicate. For caspase assays, cells were grown in 96-well plates until approximately 90% confluent, lysed and aliquots transferred to a separate plate for assay using Caspase-Glo (Promega) according to the manufacturer’s directions. A separate aliquot was used for fluorometric lactate dehydrogenase (LDH) assays with CytoTox-One (Promega). To correct for differences in cell number within each sample, caspase-3 luminescence values were normalized to the LDH values. Cell death was evaluated in GM and DM (12 and 24 h after transfer). Each experiment was conducted on triplicate wells.

1 2 6 ( 2 0 0 9 ) 7 0 0 –7 0 9

previously described (Westerfield, 2002). Embryos were dechorionated using 0.2 mg/ml pronase (Sigma). Embryos were fixed in 4% paraformaldehyde for 2 h at room temperature and further processed as previously described (Karlstrom et al., 1999). Embryos were cleared and mounted in 75% glycerol and examined using Differential Interference Contrast optics on a Zeiss compound microscope or via confocal microscopy as described above. The anti-MHC monoclonal F59 antibody (1:10; Developmental Studies Hybridoma Bank) (Crow and Stockdale, 1986) was detected with a TRITC-conjugated goat anti-mouse antiserum (1:300; Sigma). E-Z C1 confocal software was used to measure the diameter of slow twitch muscle fibers. The number of muscle fibers was determined by counting fibers in the dorsal half of each embryo. Some embryos were also cryostat sectioned at 12 lm and stained with the nuclear stain To-Pro 3 as above. Confocal images were generated and somite nuclei were quantified by manual counting.

4.8.

Sense and anti-sense RNA injections

Zebrafish achn sense mRNA was synthesized using the mMessage mMachine kit (Ambion) and 90 pg was injected. Two anti-sense oligonucleotide morpholinos (MOs) were synthesized (GeneTools LLC) against the achn 5 0 -UTR (AchnUTRMO; AGTCTGCCGCGCTGCACGGAGAAGC) and the translational start site (AchnMO; CGGCTGCTCGCTGC TCATGGTGTCC). A 1 mM MO stock was diluted in injection buffer (58 mM NaCl, 0.7 mM KCl, 0.4 mM MgSO4, 0.6 mM Ca(NO3)2) and injected in 1–2 cell stage embryos. AchnUTRMO was used at a dilution of 1:3 with a total of 1.1 ng MO injected, and AchnMO was used at a dilution of 1:5 with a total of 0.68 ng injected. We also used a mixture of the two MOs, each diluted 1:10, with a total of 0.66 ng injected. Individually, the MOs had no phenotypic effect at this concentration. All of the MO data presented is from embryos injected with this mixture. After injection, embryos were grown at 28.5 C until desired time points and were fixed in 4% paraformaldehyde for further processing.

4.9.

In situ hybridizations

Whole mount in situ hybridization was performed as described (Karlstrom et al., 1999). The Achn cDNA was inserted into pCS2 + vector (RZPD, Berlin, Germany), then the SP6/T7 DIG RNA Labeling Kit (Roche) was used to synthesize the both sense and antisense probes. Hybridization was detected using an alkaline phosphatase tagged antibody to digoxigenin (Roche), followed by overnight incubation in NBT/BCIP colorimetric buffer.

Acknowledgements 4.7.

Zebrafish

Fish lines were maintained at the UMass Amherst zebrafish facility as described (Westerfield, 2002). Wild type embryos were obtained from crossing Leopard/long fin (TL strain) fish. For all experiments, embryos were grown at 28.5 C and were staged in hours post fertilization (hpf) as

We thank Jeffery Kane for the isolation of zebrafish achn cDNA, John Nambu and Stephen Devoto for a critical reading of the manuscript, Brooke Bentley for help sectioning embryos, and Meng-Chieh Shen for the some of the achn in situ hybridization analyses in zebrafish. Supported by NIH Grants to LMS (NS042898) and ROK (NS039994), and Grants from the

MECHANISMS OF DEVELOPMENT

Rays of Hope Foundation, the Collaborative Biomedical Research Program, and the Center of Excellence in Apoptosis research (CEAR) to LMS.

R E F E R E N C E S

Crow, M.T., Stockdale, F.E., 1986. Myosin expression and specialization among the earliest muscle fibers of the developing avian limb. Dev. Biol. 113, 238–254. Dawson, S.P., Arnold, J.E., Mayer, N.J., Reynolds, S.E., Billett, M.A., Gordon, C., Colleaux, L., Kloetzel, P.M., Tanaka, K., Mayer, R.J., 1995. Developmental changes of the 26 S proteasome in abdominal intersegmental muscles of Manduca sexta during programmed cell death. J. Biol. Chem. 270, 1850–1858. Dee, K., Freer, M., Mei, Y., Weyman, C.M., 2002. Apoptosis coincident with the differentiation of skeletal myoblasts is delayed by caspase 3 inhibition and abrogated by MEKindependent constitutive Ras signaling. Cell Death Differ. 9, 209–218. Delgado, I., Huang, X., Jones, S., Zhang, L., Hatcher, R., Gao, B., Zhang, P., 2003. Dynamic gene expression during the onset of myoblast differentiation in vitro. Genomics 82, 109–121. Domingos, P.M., Steller, H., 2007. Pathways regulating apoptosis during patterning and development. Curr. Opin. Genet. Dev. 17, 294–299. Dominov, J.A., Houlihan-Kawamoto, C.A., Swap, C.J., Miller, J.B., 2001. Pro- and anti-apoptotic members of the Bcl-2 family in skeletal muscle: a distinct role for Bcl-2 in later stages of myogenesis. Dev. Dyn. 220, 18–26. Dominov, J.A., Dunn, J.J., Miller, J.B., 1998. Bcl-2 expression identifies an early stage of myogenesis and promotes clonal expansion of muscle cells. J. Cell Biol. 142, 537–544. Ehrhardt, J., Brimah, K., Adkin, C., Partridge, T., Morgan, J., 2007. Human muscle precursor cells give rise to functional satellite cells in vivo. Neuromuscul. Disord. 17, 631–638. Gallego Melcon, S., Sanchez de Toledo Codina, J., 2005. Rhabdomyosarcoma: present and future perspectives in diagnosis and treatment. Clin. Transl. Oncol. 7, 35–41. Hu, Y., Cascone, P.J., Cheng, L., Sun, D., Nambu, J.R., Schwartz, L.M., 1999. Lepidopteran DALP, and its mammalian ortholog HIC-5, function as negative regulators of muscle differentiation. Proc. Natl. Acad. Sci. USA 96, 10218–10223. Jones, M.E., Haire, M.F., Kloetzel, P.M., Mykles, D.L., Schwartz, L.M., 1995. Changes in the structure and function of the multicatalytic proteinase (proteasome) during programmed cell death in the intersegmental muscles of the hawkmoth, Manduca sexta. Dev. Biol. 169, 436–447. Karlstrom, R.O., Talbot, W.S., Schier, A.F., 1999. Comparative synteny cloning of zebrafish you-too: mutations in the hedgehog target gli2 affect ventral forebrain patterning. Genes Dev. 13, 388–393. Kitzmann, M., Fernandez, A., 2001. Crosstalk between cell cycle regulators and the myogenic factor MyoD in skeletal myoblasts. Cell. Mol. Life Sci. 58, 571–579. Kitzmann, M., Carnac, G., Vandromme, M., Primig, M., Lamb, N.J., Fernandez, A., 1998. The muscle regulatory factors MyoD and Myf-5 undergo distinct cell cycle-specific expression in muscle cells. J. Cell Biol. 142, 1447–1459.

1 2 6 ( 2 0 0 9 ) 7 0 0 –7 0 9

709

Kuelzer, F., Kuah, P., Bishoff, S.T., Cheng, L., Nambu, J.R., Schwartz, L.M., 1999. Cloning and analysis of Small Cytoplasmic Leucine-Rich Repeat Protein (SCLP), a novel, phylogeneticallyconserved protein that is dramatically up-regulated during the programmed death of moth skeletal muscle. J. Neurobiol. 41, 482–494. Lindon, C., Montarras, D., Pinset, C., 1998. Cell cycle-regulated expression of the muscle determination factor Myf5 in proliferating myoblasts. J. Cell Biol. 140, 111–118. Morgan, J.E., Partridge, T.A., 2003. Muscle satellite cells. Int. J. Biochem. Cell Biol. 35, 1151–1156. Peault, B., Rudnicki, M., Torrente, Y., Cossu, G., Tremblay, J.P., Partridge, T., Gussoni, E., Kunkel, L.M., Huard, J., 2007. Stem and progenitor cells in skeletal muscle development, maintenance, and therapy. Mol. Ther. 15, 867–877. Roy, S., Wolff, C., Ingham, P.W., 2001. The u-boot mutation identifies a hedgehog-regulated myogenic switch for fiber-type diversification in the zebrafish embryo. Genes Dev. 15, 1563– 1576. Schwartz, L.M., 2008. Atrophy and programmed cell death of skeletal muscle. Cell Death Differ. 15, 1163–1169. Schwartz, L.M., Myer, A., Kosz, L., Engelstein, M., Maier, C., 1990. Activation of polyubiquitin gene expression during developmentally programmed cell death. Neuron 5, 411–419. Shen, X., Collier, J.M., Hlaing, M., Zhang, L., Delshad, E.H., Bristow, J., Bernstein, H.S., 2003. Genome-wide examination of myoblast cell cycle withdrawal during differentiation. Dev. Dyn. 226, 128–138. Sun, D., Swaffield, J.C., Johnston, S.A., Milligan, C.E., Zoeller, R.T., Schwartz, L.M., 1997. Identification of a phylogenetically conserved sug1 CAD family member that is differentially expressed in the mouse nervous system. J. Neurobiol. 33, 877– 890. Sun, D., Ziegler, R., Milligan, C.E., Fahrbach, S., Schwartz, L.M., 1995. Apolipophorin III is dramatically up-regulated during the programmed death of insect skeletal muscle and neurons. J. Neurobiol. 26, 119–129. Tapscott, S.J., 2005. The circuitry of a master switch: Myod and the regulation of skeletal muscle gene transcription. Development 132, 2685–2695. Thayer, M.J., Tapscott, S.J., Davis, R.L., Wright, W.E., Lassar, A.B., Weintraub, H., 1989. Positive autoregulation of the myogenic determination gene MyoD1. Cell 58, 241–248. Valavanis, C., Wang, Z., Sun, D., Vaine, M., Schwartz, L.M., 2007. Acheron, a novel member of the lupus antigen family, is induced during the programmed cell death of skeletal muscles in the moth Manduca sexta. Gene 393, 101–109. Weng, H., Kim, C., Valavanis, C., Wang, Z., Schwartz, L.M., 2009. Acheron, an novel LA antigen family member, binds to cask and forms a complex with id transcription factors. Cell. Mol. Biol. Lett. 14, 273–287. Westerfield, M., 2002. The Zebrafish Book. A Guide for the Laboratory use of Zebrafish (Danio rerio). University of Oregon Press, Eugene, OR. Yaffe, D., Saxel, O., 1977. A myogenic cell line with altered serum requirements for differentiation. Differentiation 7, 159–166. Yoshida, N., Yoshida, S., Koishi, K., Masuda, K., Nabeshima, Y., 1998. Cell heterogeneity upon myogenic differentiation: downregulation of MyoD and Myf-5 generates ’reserve cells’. J. Cell Sci. 111 (Pt 6), 769–779.