Regulation of fiber-specific actin expression by the Drosophila SRF ortholog Blistered

Abstract

Serum response factor (SRF) has an established role in controlling actin homeostasis in mammalian cells, yet its role in non-vertebrate muscle development has remained enigmatic. Here, we demonstrate that the single Drosophila SRF ortholog, termed Blistered (Bs), is expressed in all adult muscles, but Bs is required for muscle organization only in the adult indirect flight muscles. Bs is a direct activator of the flight muscle actin gene Act88F, via a conserved promoter-proximal binding site. However, Bs only activates Act88F expression in the context of the flight muscle regulatory program provided by the Pbx and Meis orthologs Extradenticle and Homothorax, and appears to function in a similar manner to mammalian SRF in muscle maturation. These studies place Bs in a regulatory framework where it functions to sustain the flight muscle phenotype in Drosophila Our studies uncover an evolutionarily ancient role for SRF in regulating muscle actin expression, and provide a model for how SRF might function to sustain muscle fate downstream of pioneer factors.

Keywords: Actin; Drosophila; Flight muscle; SRF; Serum response factor.

Fig. 1.

Downregulation of blistered (bs) affects indirect fight muscle development. (A) Schematics of RNAi screening for fiber-specific factors. Colored circles represent myoblasts (smaller size) and founder cells (larger size) that contain single nuclei (black circles). Multinucleate ovals represent IFM (red) and TDT (green), two muscles arising from different differentiation programs. Lightning bolts indicate RNAi-based knockdown (KD) of candidate genes within all myoblasts and founder cells (empty circles). Corrupted shape signifies affected muscle fiber phenotype. (B) blistered locus with chromosomal coordinates (top), gene boundary (gray box) and bs transcript isoforms (RA, RB and RC) with introns (thin lines) and exons (boxes) with coding (red) and noncoding (blue) regions. Orange bars show target sites for the RNAi constructs used in this study. On the bottom, two lines represent two genetic deletions with retained (solid line) and removed (dashed) genomic sequences. (C) Cryosectioned thoraces of control and bs knockdown (KD) flies with muscles stained for polymerized actin (F-actin, green). There is a strong, selective reduction in the green signal in IFMs upon bs KD. (D) qPCR-based quantification of bs transcripts in IFM and TDT muscles dissected from control and bs KD young adults. Data are mean±s.d. **P<0.01 (t-test). (E) Thoracic muscles in bs trans-heterozygous mutant (exel6082/BSC603) and genetically rescued bs KD flies. The restoration of F-actin staining in IFMs of the rescued fly (1151>bs^KK; UAS-bs). IFM, indirect flight muscle; TDT, jump muscle.

Fig. 2.

bs affects myofibril diameter in flight muscles. (A) Cross-sectional (upper panels) and longitudinal (lower panels) views of IFM myofibrils in pupae (48 h apf) and adult flies, under control and bs knockdown (KD) conditions. Sarcomeres (brackets) are revealed by alpha-actinin immunostaining for Z-lines (blue) and counterstained using phalloidin for F-actin (green). Arrowheads show excess of alpha-actinin staining, concentrated outside the Z-lines. (B,C) Distribution of myofibril diameters (B) and sarcomere lengths (C) from 48 h apf pupae and adults under three genetic conditions: control (1151/+); bs KD (1151>bs^KK) and genetic rescue of bs KD (1151>bs^KK; UAS-bs). Median results from individual flies (20-100 myofibrils analyzed per fly) are plotted (four to seven per group). Gray horizontal lines show calculated median positions within each group.

Fig. 3.

Expression of bs isoforms in flight and jump muscles. (A) Schematic representation of structural differences of the two annotated Bs protein isoforms. The MADS box is a DNA-binding domain. (B) Activity of the bs locus driving expression of nuclear LacZ reporter (nucLacZ) in the enhancer-trap line BDSC25753. Reporter-produced nuclear β-galactosidase was detected by immunofluorescence (red) in addition to F-actin (green) and nuclear (blue) counterstaining that was added to the lower panel. Jump muscle is outlined; arrowheads indicate tracheal cell nuclei outside muscles. (C) Expression of bs isoforms in isolated IFM and TDT, as detected by RT-PCR. Fiber-specific markers (Act88F and Act79B) and ribosomal RNA (18S rRNA) validate sample purity and equal loading. (D) Expression of bs isoforms between IFM and TDT muscles, as determined by qPCR. Data are mean±s.d. (three biological repeats).

Fig. 4.

Changes in the flight muscle transcriptome in response to bs knockdown. (A) Distribution and characterization of transcriptionally active IFM genes grouped into expression tiers. (B) Identification and brief characterization of genes with significantly altered expression levels. Dotted lines delineate the confidence interval (±3 standard deviations from the regression line). r, Pearson's correlation coefficient. The box indicates the position of Act88F in the expression coordinates.

Fig. 5.

bs knockdown affects Act88F expression. (A) Semi-quantitative western blots prepared with serially diluted protein lysates from isolated IFMs and TDTs, dissected from control (1151/+) and bs KD (1151>bs^KK) flies. Gray step-slopes show the dilution rate (factor of 2). The membranes were stained using anti-actin antibody (upper panels) as well as general protein dye (lower panels). Intensities of the upper panel bands were normalized to intensities of eight most prominent bands from the lower panels and are presented on the graphs below. (B) Abundance of Act88F transcripts in isolated IFMs, as determined by qPCR. (C) Activity of the full-length, cloned Act88F enhancer driving LacZ expression. LacZ expression levels were quantitatively determined in whole-fly lysates by liquid β-galactosidase assay. All data are an average of three or four independent measurements or assays±s.d. Student's t-test was used to compare values between control and bs KD samples. **P<0.01.

Fig. 6.

Bs directly binds to the Act88F enhancer. (A) The two most conserved putative Bs-binding sequences within the Act88F enhancer. Sites 1 and 2 share sequence similarities with the canonical CArG motif (framed, W=A/T). Yellow highlight indicates nucleotides with absolute conservation across eight distantly related Drosophila species, including: D. melanogaster (mel), D. pseudoobscura (pse), D. virilis (vir), D. mojavensis (moj) and D. willistoni (wil). Blue highlight indicates positions with significant conservation that contained mismatches in not more than two species. (B) Schematics of the full-length cloned Act88F enhancer (Act88F-FL) and its truncated version (Act88F-AB), with locations of putative Bs-binding sites. (C) Electrophoretic mobility shift assay (EMSA). Nuclear extracts from S2 cells transfected with a Bs-PA expression plasmid were examined for protein/DNA complex formation (marked with arrowhead), using radiolabeled probes representing site 1, site 2 or classical CarG sequences. Essential components of binding reactions are indicated, including Bs protein (Myc-tagged Bs-PA isoform), a 50-fold excess of non-labeled probes (intact or mutated) and anti-Myc antibody, which were used to validate binding specificity. DNA/protein complex formation is seen only with site 1 and CarG probes, and it is specific to Bs protein presence. (D) In vivo activities of two Act88F-LacZ reporters, bearing the enhancers depicted in B. Data were obtained in whole-fly lysates and represent averaged β-galactosidase activity from three independent assays (±s.d.). t-test results: **P<0.01; n.s., P>0.05.

Fig. 7.

Expression of Act88F during adult myogenesis depends on exd, hth and bs, but at temporally distinct stages. (A) Schematics of pupal developmental timeline, indicating important phases of IFM development (double arrows) and activity windows for various genetic drivers (blue ribbons) used for knocking down hth and bs genes. The Mef2^TS driver is temperature sensitive, which allows deliberate activation at different times, as indicated. Exact timing of each driver is approximate, but the chronological order of different drivers is maintained. (B) Effects of hth (blue) and bs (orange) knockdowns, induced at various timepoints of development, on the expression of Act88F. Quantification of Act88F transcripts was performed by qPCR in pharate adults and is expressed as percentage of similarly treated genetically matched controls. Each data point is an average from three or four biological repeats±s.d. Solid lines show the approximated trend. Act88F expression depends on hth only at very early stages of IFM development, whereas bs retains control over Act88F for the entire period.

Fig. 8.

Flight muscle identity factors hth and exd rely on bs for sustained Act88F expression. (A) Exd expression in thoracic muscles. Cryosections of pharate adult thoraces were immunolabeled for Exd protein (red) and counterstained for actin myofibrils (green) and DNA (blue). The boxed area is shown at higher magnification in B. Exd protein (representing Exd/Hth complex) is not detectable in the jump muscle (TDT, outlined), but is well-expressed in nuclei of flight muscles (IFM). (B) Experimental identity conversion of jump muscle fibers. Wild type (WT), normal condition of small jump muscle cells (outlined). Expression of UAS-exd and UAS-hth transgenes (Act79B>exd;hth) was achieved using a jump muscle-specific Act79B-Gal4 driver and confirmed by nuclear Exd accumulation (red). Exd-expressing fibers demonstrate altered morphology of myofibrils (green) that indicates fiber type conversion. When bs knockdown was introduced (Act79B>exd;hth;bs^KK), it did not affect Exd nuclear localization or fiber type transformation. (C) Quantification of bs-RA and Act88F expression upon experimental fiber fate conversion. Transcript abundance was determined by qPCR in samples obtained by microsampling from the regions outlined by the dotted line in B. Myosin heavy-chain (Mhc) was used for transcript normalization. Each bar represents the average of three measurements±s.d. **P<0.01, using Student's t-test.

Fig. 9.

Model for Act88F transcriptional initiation and maintenance. In adult muscle progenitor cells, no expression of the Act88F gene (blue box) is evident, as indicated by the condensed chromatin state of its promoter/enhancer. In muscle precursors, upon specification as flight muscles, the pioneer factors Exd and Hth bind to the Act88F enhancer and loosen the chromatin to provide initial activation of Act88F transcription (thin line with arrow). In nascent flight muscles, Bs binds to the CarG box in the proximal promoter and boosts Act88F transcription (thick line with arrow). Later, in differentiating fibers, the Exd/Hth complex may no longer be present at the enhancer, but Bs maintains a high level of transcription from Act88F. In contrast, in developing jump muscles, Bs is unable to access the CarG box without Exd and Hth, so the Act88F gene remains inactive.