Dissection of Drosophila melanogaster Flight Muscles for Omics Approaches

The dissection protocols presented above are useful to generate IFM-enriched samples from 16 h after puparium formation (APF) until the adult stage. Dissected flight muscle samples can be used for multiple applications, and have so far been successfully applied for RT-PCR^4,17, RNA-Seq^16,32, ChIP^36,37, Western blotting^14,41 and mass spectrometry experiments (see below). To help potential users dissecting for RNA-based applications, we first present our results highlighting important considerations specifically for isolation of RNA from IFMs. To more broadly demonstrate the utility of our dissection protocols, we then illustrate some of the possible –omics applications using our data on the RNA-binding protein Bruno1.

IFM dissection protocol yields high quality RNA

It is important to determine the number of flies to be dissected in advance, as coding mRNA is estimated to constitute only 1–5% of total RNA⁴². We obtained on average 24 ± 9 ng of total RNA per fly from IFM dissected from 1 d adults (Figure 4F and Supplemental Figure 1A), with yields typically increasing with experience. This yield of total RNA per fly is relatively constant, fluctuating around 25 ng for IFM dissected at 16 h APF, 24 h APF, 30 h APF, 48 h APF, 72 h APF and 90 h APF (Figure 4F and Supplemental Figure 1B,D,E). These observations also reflect any RNA isolated from contaminating fat, tendon, trachea or other cell types, which may be higher in samples isolated from earlier timepoints. Thus, we obtained >1 μg of total RNA from IFM from 50 flies and typically dissect IFM from 100−150 flies to generate >3 μg of total RNA for RNA-Seq samples.

The method of RNA isolation affects the quantity and quality of recovered RNA, and we encourage users to validate their isolation approach. For example, while isolation using method 1 produces on average 1143 ± 465 ng of total RNA from IFM from 50 1 d adult flies, isolation with various commercial kits yields anywhere from 186 ± 8 ng to 1261 ± 355 ng of total RNA (Figure 4G and Supplemental Figure 1C). RNA isolated from commercial kits is generally of good quality (Figure 4H and Supplemental Figure 1F), but low recoveries suggest that RNA may not be efficiently eluted from the columns. RNA integrity can also be compromised by use of a kit as done in method 2 (Figure 4H, second plot), likely due to buffer constitution and heat treatments, leading to severe fragmentation that can impact downstream experiments.

It is also important to observe proper RNase-free technique when isolating and handling RNA samples. Although freeze-thaw cycles and a 4 h room temperature incubation do not dramatically impact RNA integrity profiles, even small amounts of RNase lead to rapid RNA degradation (Figure 4I and Supplemental Methods). Users are still encouraged to work on ice and limit freeze-thaw to prevent RNA hydrolysis and fragmentation. This was not detected here but preventing RNase contamination by using filter tips and DEPC-treated buffers is absolutely essential.

The efficiency of reverse transcription also impacts the success of downstream applications. We obtained reliable results with two of three commercial RT kits we tested, which both amplify strong RT-PCR bands for ribosomal gene rp49 (Figure 4J). However, RT Kit #2 may be more sensitive for the detection of low-expressed transcripts, as we obtained stronger bands for the RNA-binding protein bru1 for all three biological replicates (Figure 4J). Taken together, these results illustrate that high-quality RNA can be isolated from IFMs dissected with this procedure.

Dissected IFMs produce high quality mRNA-Seq and proteomics data

Using IFM dissected according to the above protocol at 30 h APF, 72 h APF and from 1 d adult flies, we previously showed that the RNA-binding protein and CELF1-homologue Bruno1 (Bru1, Arrest, Aret) controls an IFM-specific splicing pathway downstream of the transcription factor Spalt major (Salm)¹⁶. IFMs from null mutants as well as flies with muscle-specific bruno1 RNAi (bru1-IR) display sarcomere growth defects, misregulation of myosin activity and ultimately hypercontraction and loss of muscle fibers^16,17. Below we demonstrate the utility of dissected IFMs for whole proteome mass spectrometry and show that several of the expression changes we observed on the RNA level are also evident on the protein level. We further highlight a specific developmental splice event in Mhc that was found to be regulated by Bruno1, illustrating that mRNA-Seq and RT-PCR from dissected IFMs can be used to demonstrate the regulation of alternative splice events.

Depending on library quality and depth, mRNA-Seq data can be analyzed on the level of gene units (averaging read counts over all exons of a gene), individual exons, or splice junctions. mRNA-Seq data from bru1-IR IFMs compared to wildtype shows weak changes in expression on the gene unit level¹⁶ (Figure 5A). At 72 h APF, there is already a trend for sarcomere genes such as muscle LIM protein at 60A [Mlp60A], actin 57B [Act57B], muscle-specific protein 300 kDa [Msp300], or Stretchin-Mlck [Strn-Mlck]) that are important for proper muscle development to be downregulated in bru1-IR muscle (Figure 5A and Supplemental Table 1). However, we have shown previously that on the level of individual exons, there is a much stronger downregulation of specific sarcomere gene isoforms¹⁶, suggesting the major function of Bruno1 is to control alternative splicing (Supplemental Table 1).

Using whole-proteome mass spectrometry on dissected IFMs, we can show similar regulation on the protein level (Figure 5B and Supplemental Table 2). Of the 1,895 peptide groups detected, 524 (28%) of them are misregulated in Bru1^M2 mutant IFM in 1 d adults (Supplemental Table 2). Downregulation of both Strn-Mlck and Mlp60A protein is also observed, matching observations at the transcript level in our mRNA-Seq data. Despite the limited number of database peptides that map to specific protein isoforms (see Supplemental Methods for analysis details), for sarcomere proteins Tropomyosin 1 (Tm1), upheld (up/TnT), Mhc, bent (bt/projectin) and Paramyosin (Prm) we observe upregulation of peptides from one isoform and downregulation of another (Figure 5B), confirming our previous observations of similar regulation on the RNA level¹⁶. This demonstrates that dissected IFMs are useful for both mRNA-Seq and proteomics applications.

As a further example of how omics data can complement traditional approaches to enhance and extend biological insight, we chose to focus on splicing at the C-terminus of Mhc. A previously characterized protein trap line called weeP26 is inserted in the final intron of Mhc^43,44 (see Supplemental Methods for exact location). weeP26 contains a strong splice acceptor and is incorporated into presumably all Mhc transcripts (Figure 5C). However, the GFP labeled protein in IFM is incorporated into two "dots" on either side of the M-line, while in leg muscle, it incorporates uniformly across the M-line and weakly across the thick filaments (Figure 5E). Orfanos and Sparrow showed these "dots" in IFM form due to a developmental Mhc isoform switch: the Mhc isoform expressed before 48 h APF is GFP-labeled as the weeP26 exon inserts in the open read frame, while the Mhc isoform expressed after 48 h APF is unlabeled, as the weeP26 exon is included downstream of the stop codon in the 3'-UTR⁴⁴.

Our mRNA-Seq data allowed us to characterize C-terminal Mhc isoform expression in greater detail. While two different Mhc terminations have been reported^43,44, our mRNA-Seq data and current Flybase annotation (FB2019_02) suggest that there are actually three possible alternative splice events at the Mhc C-terminus (Exon 34-35, 34-36, or 34-37) (Figure 5C), which is confirmed by RT-PCR (Figure 5D). weeP26 GFP is inserted in the intron between Exon 36 and 37; thus, as both Exon 34-35 and Exon 34-36 isoforms contain stop codons, GFP can only translated in the Exon 34-37 isoform (resulting in Exon 34-GFP-37). We further could see both temporal and spatial regulation of all Mhc isoforms. In IFM, we observe an Mhc isoform switch from Exon 34-37 to Exon 34-35 between 30 h APF and 48 h APF (Figure 5C,D,F) at 27 °C, even though this is not yet visible by immunofluorescence at 48 h APF (Figure 5E). Legs already express a mixture of Exon 34-37 and Exon 34-35 at 30 h APF, and by 72 h APF express all three Mhc isoforms (Figure 5D,F). Adult jump muscle (TDT) also expresses all three Mhc isoforms (Figure 5F), suggesting this is generally true for tubular somatic muscles. Thus, our mRNA-Seq data allow extension of previous findings by narrowing the timeframe for the Mhc isoform switch in IFM and characterizing Mhc isoform use in tubular muscles.

Mhc isoform regulation in salm and bru1 mutant IFM were then examined. In both cases, we saw misregulation of weeP26. Salm mutant IFMs fail to complete the developmental switch in Mhc isoform expression and phenocopy leg splicing patterns at later stages, including gain of the Exon 34-36 event (Figure 5F). This agrees with previous findings that loss of Salm results in a near-complete fate transformation of IFM to tubular muscle¹⁶. Bru1-IR and bru1 mutant IFM, similar to salm^-/- IFM, retains the Exon 34-37 splice event through adult stages (Figure 5E,F), resulting in a weeP26 GFP labeling pattern resembling leg muscle, but it does not gain the Exon 34-36 event. This suggests that Bruno1 is necessary in IFM to at least partially control the developmental switch in Mhc alternative splicing, but it indicates that additional splicing factors are also misregulated in the salm^-/- context. Furthermore, this example illustrates how RT-PCR and mRNA-Seq data from dissected IFM can be valuable in gaining a deeper understanding of developmental splicing mechanisms and observed morphological defects.

Figure 1: IFM development and staging of pupae. (A) Schematic of IFM development at 24 h APF, 32 h APF, 48 h APF, 72 h APF, and 1 d adults showing compaction of flight muscles (green) at ~32 h APF and subsequent fiber growth to fill the thorax. Tendons are in dark grey. (B) Confocal images of fixed IFMs from open book dissections (24 h, 32 h, 48 h)¹⁹ or thorax hemisections (72 h, 1 day) stained for actin (rhodamine phalloidin, magenta) and GFP (green). (C,D) Images of GFP fluorescence in live pupae illustrating intact IFM morphology of the dissection fly line in the dorsal (C) or lateral (D) plane. Asterisks mark IFM location. (E) To prepare for dissections, fly stocks should be flipped or crosses set 3–4 days in advance. (F) Prepupae are selected by their white color (yellow arrowheads) and isolated using a wetted paintbrush (F',F''). (G) Prepupae should be sexed to separate females from males based on the presence of testes which appear as posteriorly located translucent balls (yellow asterisks). (H) Pupae are aged on wetted filter paper in 60 mm dishes. Scale bars = 100 µm (B), 1 cm (C,D,E,H), 1 mm (F,F'',G). Please click here to view a larger version of this figure.

Figure 2: Dissection of IFMs before 48 h APF. (A) Addition of 1x PBS buffer to a black dissecting dish with a transfer pipette. (B) Transfer of staged pupae using a paintbrush. (C) Under a fluorescent dissecting microscope to visualize GFP, gentle pushing of the pupa to the bottom of a dissecting dish using #5 forceps (outlined in grey). The "X" in a circle denotes motion into the image. (D,E) Grasping of the pupae anteriorly (D), then poking of the pupae just behind the thorax (E). Dash in a circle denotes no motion. (F,G) Pulling with the anterior forceps (arrow) to remove the pupal case (F), then removal of the abdomen (G). (H) Repetition of C-G for several pupa. Yellow dotted lines are numbered denoting contributing pupae. (I, J) Use of the forceps (I) to isolate IFMs from surrounding tissue (J). Dot in a circle denotes motion out of the page. (K,L) Removal of contaminants including fat and jump (TDT) muscles (K) to generate a clean IFM sample (L). TDT has lower GFP expression and a different shape than IFM fibers (K'). (M,N,O) Use of a clipped pipette tip (M) to collect dissected IFMs (N) and its transfer to a microcentrifuge tube (O). Scale bars = 1 cm (A,B,M,O), 1 mm (C-G), 500 µm (H-L,N). Please click here to view a larger version of this figure.

Figure 3: Dissection of IFMs after 48 h APF. (A) Aligning of pupae on double-stick tape. (B) Removal of pupae from the pupal case by opening anteriorly (B), cutting the case dorsally (B'), and lifting out the pupa (B''). Circle symbols represented the same as Figure 2. (C) Transfer of pupae to buffer. (D) Removal of the abdomen by cutting with scissors (yellow double arrows) and separation from thoraxes (D'). (E, F) Addition of clean buffer (E), then cutting of thoraxes in half longitudinally (F,F'). (G,H) Dissections can be performed under white light (G) or fluorescence to visualize the GFP (H); cutting of the IFMs on one side (G'), then the other side (G''); lifting out of the thorax with forceps (outlined in grey) (G'''). (I,J,K) Collection of IFMs in buffer (I) and removal of contaminating ventral nerve cord (VNC), gut, and jump muscle (TDT) (J) to generate a clean IFM sample (K). TDT has lower GFP expression and a different shape than IFM fibers (J'', K'). (L,M) Use of forceps to transfer IFMs (L) to a microcentrifuge tube (M). Scale bars = 1 cm (A,E,M), 1 mm (B-D',F-L). Please click here to view a larger version of this figure.

Figure 4: IFM preservation and RNA isolation details. (A) IFMs are pelleted by centrifugation for 5 min at 2000 x g. (B) IFM pellet (arrow) and pellet under fluorescence (B'). (C) Removal of all buffer with a pipette tip. (D) For RNA extraction, resuspension of pellet in isolation buffer. This step can be skipped to dry-freeze dissected IFMs. (E) Freezing of sample in liquid nitrogen or on dry ice and storage at -80 °C. Scale bars = 10 cm (A), 1 mm (B,B’), 1 cm (C,D,E). (F) Nanograms (ng) of total RNA from dissected IFM obtained per fly at 16 h APF, 24 h APF, 30 h APF, 48 h APF, 72 h APF, 90 h APF, and 1 d adult. Error bars = SD. (G) Total RNA isolated from IFM dissected from 50 1 d adult flies using different extraction methods. Error bars = SD. (H) Representative traces to assay RNA integrity after different extraction methods. The ribosomal bands run just below 2000 nucleotides (nt) and the marker band at 25 nt. Additional traces available in Supplemental Figure 1. (I) Representative traces of a freshly isolated RNA sample (top), a sample freeze-thawed 25x on dry ice (second plot), a sample left for 4 h on the bench (third plot), and a sample treated with RNase A (bottom plot). Note complete degradation of RNA upon addition of RNase A. (J) RT-PCR gel from kits as labeled for bru1 and rp49. The relative intensity of the bru1 band normalized against rp49 is plotted below. Error bars = SEM (unpaired t-test, p = 0.0119). Please click here to view a larger version of this figure.

Figure 5: Application of IFM dissections to investigate Bruno1 function in alternative splicing. (A) Volcano plot of mRNA-Seq data (gene unit) from IFMs dissected at 72 h APF. Genes that are significantly differentially regulated between bru1-IR and wildtype IFM (padj < 0.05, abs(log₂FC) >1.5) are shown in blue, and non-significant genes in grey. Sarcomere proteins are highlighted in red, and select genes are labeled. (B) Volcano plot of whole proteome mass spectrometry results from 1 d adult IFMs. Proteins significantly different between bru^M2 mutants and wildtype (FDR < 0.05) are shown in blue, nonsignificant proteins in grey. Sarcomeric proteins are highlighted in red. Peptides corresponding to genes in (A) are labeled in red. Sets of peptides mapping to different isoforms of the same protein are labeled in the same color. (C) Scheme of the C-terminus of Mhc illustrating distinct transcript isoforms and insertion location of the weeP26 gene trap (see Supplemental Methods for insertion point). RT-PCR primers are denoted as black lines above transcripts. Read counts per kilobase per million bases (RPKM) from mRNA-Seq are shown for IFMs dissected from wildtype at 30 h APF (orange) and 72 h APF (red), from bru1-IR (blue) and salm^-/- (cyan) at 72 h APF and from whole leg (green) at 72 h APF. (D) RT-PCR with primers against Mhc showing the isoform switch in IFM between 30 h APF and later timepoints. The Exon 34-35 splice event is only weakly observed in bru^M3 mutant IFM or in the adult leg. (E) Confocal images of weeP26 GFP localization in wildtype IFM sarcomeres at 48 h APF and 90 h APF compared to 90 h APF leg muscle. Scale bars = 1 µm. (F) Splice junction quantification from mRNA-Seq data for genotypes and timepoints as labeled. Junction reads are presented as the ratio of a specific splice event (Exon 34 to 35 in grey, 34 to 36 in purple, and 34 to 37 in green) to the total number events sharing the exon 34 splice donor. Please click here to view a larger version of this figure.

Supplemental Figure 1: (A,B,C) RNA yields from samples of the same genotype dissected by the same researcher in the same week. After all samples were dissected, RNA was isolated and measured the same day. (A) Nanograms (ng) of total RNA obtained from IFM dissections per 1 d adult fly. Error bars = SEM. (B) Total RNA obtained from dissected IFM per fly at 30 h APF, 48 h APF, 72 h APF and 1 d adult. (C) Total RNA isolated from IFM dissected from 50 1 d adult flies using different extraction methods. (D) Total RNA concentrations per fly from dissected legs, jump muscle (TDT) and IFM. More RNA is obtained from the larger IFMs. Error bars = SD. (E) Total RNA concentrations per fly of IFM dissected from controls compared to RNAi or mutant samples at 30 h APF, 72 h APF and 1 d adult. For mutants, w¹¹¹⁸ was used as wildtype control. Mutant data are compiled from bru1-IR, salm^-/- and another RNA-binding protein mutant. Note that for these manipulations, RNA yields are decreased in 1 d adult due to muscle atrophy and loss, so more flies need to be dissected to obtain sufficient quantities for omics approaches. Errors bars = SD. (F) Additional traces showing RNA integrity for the RNA isolation methods shown in Figure 4G and in Supplemental Figure 1C. Please click here to view a larger version of this figure.

Supplemental Methods: A detailed description of the methods and reagents used throughout the text and, in particular, to generate the data shown in Figure 1A-D, Figure 4F-K, Figure 5, Supplemental Table 1, and Supplementary Table 2. These data motivate the dissection protocol and demonstrate its utility for RNA isolation, mRNA-Seq, RT-PCR, and proteomics. Please click here to download this file.

Supplemental Table 1: Table of 72 h APF mRNA-Seq data for sarcomere proteins identifying differentially expressed genes (via DESeq2) and exons (via DEXSeq) in bru1-IR vs. wildtype IFMs.

Supplemental Table 2: Table of whole proteome mass-spectrometry data from 1 d adult identifying differentially expressed proteins and protein isoforms in bru^M2 mutant vs. wildtype IFMs.