The Him Gene Reveals a Balance of Inputs Controlling Muscle Differentiation in Drosophila

Summary Tissue development requires the controlled regulation of cell-differentiation programs. In muscle, the Mef2 transcription factor binds to and activates the expression of many genes and has a major positive role in the orchestration of differentiation [1–4]. However, little is known about how Mef2 activity is regulated in vivo during development. Here, we characterize a gene, Holes in muscle (Him), which our results indicate is part of this control in Drosophila. Him expression rapidly declines as embryonic muscle differentiates, and consistent with this, Him overexpression inhibits muscle differentiation. This inhibitory effect is suppressed by mef2, implicating Him in the mef2 pathway. We then found that Him downregulates the transcriptional activity of Mef2 in both cell culture and in vivo. Furthermore, Him protein binds Groucho, a conserved, transcriptional corepressor, through a WRPW motif and requires this motif and groucho function to inhibit both muscle differentiation and Mef2 activity during development. Together, our results identify a mechanism that can inhibit muscle differentiation in vivo. We conclude that a balance of positive and negative inputs, including Mef2, Him, and Groucho, controls muscle differentiation during Drosophila development and suggest that one outcome is to hold developing muscle cells in a state with differentiation genes poised to be expressed.


Supplemental Experimental Procedures
In Situ Hybridization, Immunohistochemistry, and Acridine Orange Assay In situ hybridizations with Digoxigenin-labeled RNA probes were as described previously [S1]. The DNA template for the Him probe was a cDNA in pBluescript II KS (2) isolated from a subtracted library [S1], and the DNA template for b3-tubulin was a b3-tubulin cDNA (a gift from R.Renkawitz-Pohl) subcloned into pBluescript II KS (+). Embryos were mounted in 80% glycerol. Acridine orange staining was essentially as described in Sullivan et al. [S2]. Single-antibody staining was as described in Ruiz-Gomez and Ghysen [S3] with the following primary antibodies: rabbit anti-Mhc (1:500, gift from D.Kiehart), guinea pig anti-Kruppel (1:1000, gift from D.Kosman), rabbit anti-Mef2 (1:1000, gift from B.Paterson), and guinea pig anti-Him (1:1500). Embryos were mounted either in an acetone: araldite mixture (1:1 ratio) or in 80% glycerol. In each case, embryos were viewed with a Zeiss Axioskop microscope, and images were captured with Axiovision software. Fluorescent double stainings were essentially as in Kaltschmidt et al. [S4] with the following primary antibodies: mouse anti-GFP (1:3000, Sigma) plus one of rabbit anti-Mhc (1:500), rabbit anti-Mef2 (1:1000), or rabbit anti-Twist (1:5000, gift from S. Roth). Embryos were mounted in Vectashield fluorescent mounting medium (Vector laboratories) and visualized on a Leica confocal microscope; images were processed with the Leica software. The indicated times of different stages of embryogenesis are for development at 25 C.

Generation of Transgenic Lines
We made UAS-Him and UAS-Him-DWRPW by subcloning from Him cDNA in pBluescript II KS (2) [S1] into pUAST [S5] as a restriction fragment and as a PCR-amplified fragment for removal of the Cterminal WRPW, respectively. All PCR-generated DNA fragments used in this project were sequence verified, and the primer sequences are available on request. We made UAS-Him-RNAi by cloning the complete PCR-amplified Him coding sequence into pGEM-T (Promega) and subcloning the sequence into pJM1084 (a spliceactivated UAS hairpin vector [S6]) on both sides of the intronic spacer (antisense 5 0 of the intron, sense 3 0 of the intron). The Him-GFP fusion gene construct comprised approximately 3.8 kb of upstream genomic sequence, the mGFP6 variant [S7] fused via a -SSSS-linker to the N-terminus of the Him coding sequence, and the Him 3 0 UTR. The GFP sequence was inserted by standard PCR procedures. Transgenic lines for all constructs were generated by injection of yw embryos according to published procedures [S8], and stocks homozygous for each construct were made. For each construct, a minimum of three independent transgenic lines was tested.

UAS-Him RNAi Analysis
The knockdown of Him expression was assessed first by in situ hybridization. Figure S2 shows a representative example. The knockdown was then quantitated by qRT-PCR of late stage 12/early stage 13 embryos (9 hr 10 min-9 hr 40 min AEL) as follows. RNA was isolated from three independent collections of >150 individually staged embryos from both wild-type and UAS-Him RNAi. cDNA generated with oligo dT was used to program a standard SYBR green q-PCR reaction. The (+) Him PCR primer spanned the single Him intron; the (2) strand primer was in the 3 0 UTR (sequences available on request). Each of the three biological replicates was assayed in triplicate for Him and the reference gene rp49. The expression of Him mRNA in Him RNAi relative to the wild-type was calculated with the 2 -DDC(T) method [S14]. The average reduction in the Him RNAi condition was to 44% of the wild-type Him RNA levels.

Hatching and Survival Assay
Three hundred developing embryos were aligned on apple-juice agar plates at 18 C, and the number of newly hatched larvae was scored. Surviving third instar larvae were put in tubes and allowed to develop until adulthood when the number of eclosing flies was scored.

Muscle Phenotype Analysis
The somatic-muscle phenotype was scored through examination of the muscle pattern. Each of the 30 muscles per abdominal hemisegment was analyzed systematically in three hemisegments (A2 to A4) for any differences compared to the wild-type. For description of the RNAi phenotype, 101 embryos were analyzed. For assessment of the genetic interaction between UAS-Him and gro or UAS-mef2, at least 45 embryos of each genotype were analyzed. The number of wild-type muscles was recorded for each embryo, and the mean was determined and then compared between UAS-Him alone and UAS-Him plus either gro or UAS-mef2. The significance was assessed with a two-sample t test.

In Vitro Protein-Protein Interactions
The GST-Him and GST-HimDWRPW constructs were generated with the pET-3a vector (Novagen), in which we removed the T7-Tag and replaced it with GST to create GST-pET-3a. Him and Him-DWRPW DNA fragments were generated by PCR from Him cDNA, cloned into pGEM-T (Promega), and subcloned into GST-pET-3a. In vitrotranslated [ 35 S]-Met-labeled Gro was produced, GST-fusion proteins were expressed and purified, and in vitro binding assays were undertaken as described [S15]. Analysis was by SDS PAGE, and radiolabeled Gro was visualized by autoradiography.

S2 Expression Assay
Mef2 activity was assayed with a mIR-1 luciferase reporter after transfection into Drosophila S2 cells as described previously [S16].
Supplemental References S1. Taylor, M.V. (2000). A novel Drosophila, mef2-regulated muscle gene isolated in a subtractive hybridisation-based molecular screen using small amounts of zygotic mutant RNA. Dev. Biol. 220, 37-52. Figure S1. Him Affects the Differentiation Phase of Muscle Development UAS-Him expression was driven in the developing mesoderm by twi-Gal4;twi-Gal4 at 25 C (B, D, F, and H) and compared with the wild-type (A, C, E, and G). When Him is overexpressed, muscle development proceeds similarly to the wild-type up to stage 13, as shown by immunostaining for the founder cell marker Kruppel (A and B) and for Mef2 (C and D). However, immunostaining for Mef2 shows that differentiation is dramatically affected by Him at stage 15 (E and F). The finding that acridine orange (AO) reveals an increase in cell death at stage 16 (G and H) suggests that cells that fail to differentiate into muscle die. At stage 17, Him overexpression results in a dramatic reduction in Myosin-expressing cells (Figure 1). In each case, representative examples of the phenotype are shown.