Gene trapping in reversibly arrested myoblasts 27 A gene-trap strategy identifi es quiescence-induced genes in synchronized myoblasts RAMKUMAR SAMBASIVAN1,3, GRACE K PAVLATH2 and JYOTSNA DHAWAN1 1Centre for Cellular and Molecular Biology, Hyderabad 500 007, India 2Department of Pharmacology, Emory University, Atlanta, GA 30322, USA 3Present address: Department of Developmental Biology, Pasteur Institute, 75724 Cedex 15 Paris, France *Corresponding author (Fax, 91-40-2716-0591; Email, [email protected]) Cellular quiescence is characterized not only by reduced mitotic and metabolic activity but also by altered gene expression. Growing evidence suggests that quiescence is not merely a basal state but is regulated by active mechanisms. To understand the molecular programme that governs reversible cell cycle exit, we focused on quiescence-related gene expression in a culture model of myogenic cell arrest and activation. Here we report the identifi cation of quiescence- induced genes using a gene-trap strategy. Using a retroviral vector, we generated a library of gene traps in C2C12 myoblasts that were screened for arrest-induced insertions by live cell sorting (FACS-gal). Several independent gene- trap lines revealed arrest-dependent induction of βgal activity, confi rming the effi cacy of the FACS screen. The locus of integration was identifi ed in 15 lines. In three lines, insertion occurred in genes previously implicated in the control of quiescence, i.e. EMSY – a BRCA2-interacting protein, p8/com1– a p300HAT-binding protein and MLL5 – a SET domain protein. Our results demonstrate that expression of chromatin modulatory genes is induced in G0, providing support to the notion that this reversibly arrested state is actively regulated. [Sambasivan R, Pavlath G K and Dhawan J 2008 A gene-trap strategy identifi es quiescence-induced genes in synchronized myoblasts; J. Biosci. 33 27–44] 1. Introduction cycle regulators and muscle-specifi c transcription factors of the basic helix–loop–helix family is known to initiate Mounting evidence suggests that cell cycle exit into G0 differentiation (Wei and Paterson 2001) and the mechanisms is under active transcriptional control, and not merely a by which irreversible arrest is maintained in differentiated basal state resulting from an absence of growth-promoting muscle cells are well understood (Olson 1992; Lassar et al signals (Yusuf and Fruman 2003). Myogenic differentiation 1994). By contrast, little is known about the mechanism involves irreversible withdrawal of committed myoblasts by which reversible growth arrest occurs in satellite cells, from the cell cycle followed by the sequential activation the quiescent myogenic progenitor cells found in adult of tissue-specifi c genes (Nadal-Ginard 1978; Andres muscle (Seale and Rudnicki 2000; Dhawan and Rando and Walsh 1996). The balance between proliferation and 2005), but must involve an uncoupling of cell cycle exit differentiation is controlled by events in early G1 (Clegg and differentiation. Genes induced specifi cally during et al 1987). Molecular cross-talk between ubiquitous cell reversible growth arrest may control the process by which Keywords. Gene-trap; MLL5; myoblast; p8; quiescence. Abbreviations used: βgal, βgalactosidase; CFU, colony forming units; ENT, EMSY N-terminus; FACS, fl uorescent activated cell sorting; FDG, fl uorescein di-β-D-galactopyranoside; LTR, long terminal repeat; MLL5, mixed lineage leukaemia gene 5; MOI, multiplicity of infec- tion; MRF, myogenic regulatory factor; PETG, phenyl ethyl thiogalactoside; PHD, plant homoeo domain; PR, polymerase chain reaction; RNAi, RNA interference; RT, reverse transcriptase; SA, splice acceptor; SET, conserved in Su(var), Enhancer of Zeste, Trithorax http://www.ias.ac.in/jbiosci J. Biosci. 33(1), March 2008, 27–44, © IndianJ. Academy Biosci. 33 of(1), Sciences March 2008 27 28 Ramkumar Sambasivan, Grace K Pavlath and Jyotsna Dhawan this separation of the two programmes occurs. Moreover, 1992; Voss et al 1998; Hansen et al 2003). Gene trapping increasing evidence suggests that quiescence is not merely a has been applied to identify differentially expressed genes basal metabolic state, but is actively regulated. Therefore, we in cell lines (Gogos et al 1996; Lih et al 1996), and has sought to identify genes that are induced during reversible some advantages when compared with other approaches growth arrest in muscle cells. such as differential display-polymerase chain reaction Cultured mouse myoblasts of the C2C12 cell line (Yaffe (PCR) or microarrays. Since insertion occurs essentially and Saxel 1977; Blau et al 1983) have been widely used at random, genes expressed at very low levels can also be to model myogenic differentiation in culture (Olson 1992; identifi ed. Differential display, though sensitive, is generally Lassar et al 1994) and for molecular analysis, which led to biased towards abundant transcripts. In addition, gene-trap the identifi cation of key muscle regulators such as MyoD insertions may result in loss-of-function phenotypes for (Davis et al 1987). Mitogen deprivation of asynchronous haplo-insuffi cient genes. Reporter expression in gene-trap C2C12 cultures triggers irreversible cell cycle arrest, lines also allows the activity of the locus to be followed in fusion and differentiation into multinucleated myotubes different conditions. (Andres and Walsh 1996; Blau et al 1983). By contrast, Here we describe a gene-trap screen using live cell sorting during suspension culture, myoblasts exit the cell cycle in to identify quiescence-induced genes in synchronized C2C12 an undifferentiated state despite the presence of saturating myoblasts. Our fi ndings suggest that as in ES cells, most concentrations of growth factors (Milasincic et al 1996; insertional events occur in the 5′ regions of transcriptional Sachidanandan et al 2002). Importantly, this anchorage- units. In three gene-trap myoblast lines, expression of the dependent arrest is synchronously reversed upon restoration reporter transcript and of the endogenous mRNA were found of the surface contacts. Reversibly arrested C2C12 myoblasts to be induced during synchronization in G0. All three genes model several aspects of satellite cells (Milasincic et al (MLL5, p8/com1, EMSY) have been previously implicated 1996; Sachidanandan et al 2002). First, myoblasts arrest in in growth control. Thus, quiescence-induced genes with G0, as evidenced by the absence of DNA synthesis, a 2C potential growth regulatory activities were identifi ed using DNA content and suppression of growth-associated genes. the gene-trap strategy. Second, expression of MyoD and Myf5 is suppressed and Our results suggest that reversible G0 arrest in myoblasts differentiation-dependent genes such as myogenin, myosin is associated with the induction of gene expression, lending and muscle creatine kinase are not induced. Third, arrested support to the notion that quiescence is an actively regulated myoblasts are synchronously activated out of G0; they state. sequentially express MyoD and Myf5 in G1 and enter the S phase. Finally, in addition to the myogenic regulators, several 2. Materials and methods genes implicated in satellite cell (SC) arrest (Beauchamp et al 2000), commitment (Seale et al 2000) and activation 2.1 Cell culture (Cornelison and Wold 1997; Dhawan and Rando 2005) are regulated appropriately during reversible arrest in culture. C2C12 myoblasts (Yaffe and Saxel 1977; Blau et al 1983) Taken together, these fi ndings indicate that reversible were obtained from H Blau (Stanford University) and a arrest in culture involves the regulation of some key genes subclone A2 generated in our laboratory (Sachidanandan et implicated in satellite cell function in vivo (Sachidanandan et al 2002) was used in all the experiments. Myoblasts were al 2002; Dhawan and Rando 2005). maintained in growth medium (GM; DMEM supplemented To understand the control of entry into and exit from G0, with 20% FBS and antibiotics). we employed a gene-trap screen using a retroviral vector Differentiation was induced in cultures at ~80% to identify genes that may be induced during arrest. Gene confl uence after washing with PBS and incubating in traps are created by random insertions of a retroviral vector differentiation medium (DM: DMEM with 2% horse serum) containing a promoterless reporter gene into the target replaced daily for 3 days. Multinucleated myotubes appear genome, and allow effi cient detection of transcriptionally after 24 h in DM and peak at 3–5 days. The fusion index active loci because of splice acceptor (SA) sequences was determined on day 3 by counting the percentage of total engineered upstream of the reporter (Gossler et al 1989; nuclei contained in myotubes of >2 nuclei. Friedrich and Soriano 1991; Reddy et al 1991; Skarnes et al Suspension culture of myoblasts was as described 1992; Voss et al 1998). Reporter expression can occur even (Milasincic et al 1996; Sachidanandan et al 2002). Briefl y, if the insertions occur within introns, by splicing to upstream subconfl uent cultures were harvested using trypsin and exons. Gene traps are also mutagenic and, therefore, when cultured as a single cell suspension at a density of 105 cells/ applied to embryonic stem (ES) cells, this strategy has been ml in DMEM medium containing 1.3% methyl cellulose, successfully used for large-scale functional analysis of the 20% FBS, 10 mM HEPES and antibiotics. After 48 h, when mouse genome (Friedrich and Soriano 1991; Skarnes et al >98% of cells entered G0, suspended cells