Overexpression of Drosophila NUAK or constitutively-active Formin-like promotes the formation of aberrant myofibrils

Abstract

Muscle development and maintenance is central to the normal functioning of animals. Muscle tissues exhibit high levels of activity and require the dynamic turnover of proteins. An actomyosin scaffold functions with additional proteins comprising the basic contractile subunit of striated muscle, known as the sarcomere. Drosophila muscles are similar to vertebrate muscles in composition and they share a similar mechanism of development. Drosophila NUAK (NUAK) is the homolog of NUAK1 and NUAK2 in vertebrates. NUAK belongs to the family of AMP-activated protein kinases (AMPKs), a group of proteins with broad and overlapping cellular targets. Here we confirm that NUAK dynamically modulates larval muscle sarcomere size as upregulation of NUAK produces longer sarcomeres, including increased thin filament lengths. Furthermore, NUAK overexpression results in aberrant myofibers above the nuclei plane, upregulation of Formin-like (Frl), and an increase in newly synthesized proteins at sites consistent with actin filament assembly. Expression of constitutively-active Frl also produces aberrant myofibers similar to NUAK overexpression. These results taken together strongly suggest a functional link between NUAK and Frl in myofibril formation in an in vivo setting.

Summary statement

NUAK regulates actin assembly via upregulation of Frl concomitant with de novo protein synthesis.

Introduction

Structures comprised of actin-myosin complexes are present in virtually all eukaryotes (Rivero & Cvrcková, 2007). These actomyosin structures are necessary for diverse cellular processes ranging from endocytosis, intracellular transport, and mechanosensing to cell division, cell migration, and wound healing (Chanet et al., 2017; Hui, Nakamura, Dubrulle, & Parkhurst, 2023; Kumari et al., 2019; Murrell, Oakes, Lenz, & Gardel, 2015; Salbreux, Charras, & Paluch, 2012). Within cells, actomyosin systems are present in the form of cytokinetic rings, stress fibers in non-muscle cells, and muscle myofibrils (Pollard & Cooper, 2009; Sanger et al., 2005; Tojkander, Gateva, & Lappalainen, 2012). Four different categories of stress fibers have been identified based on their location and function within cells - dorsal stress fibers, ventral stress fibers, transverse arches and perinuclear actin caps. Except for non-contractile dorsal stress fibers, all other stress fiber types have myosin associated with the actin (Naumanen, Lappalainen, & Hotulainen, 2008; Tojkander et al., 2012). In non-muscle cells, stress fibers can be assembled in response to strain or generated de novo from the actin cortex via myosin pulses (Chapin, Blankman, Smith, Shiu, & Beckerle, 2012; Lehtimaki, Rajakylä, Tojkander, & Lappalainen, 2021). Multiple models exist describing myofibril formation in muscle cells. There is evidence that myofibrils form from pre-myofibril structures (Sanger et al., 2005), while other data suggests that latent protein complexes form prior to sequential sarcomere assembly (Rui, Bai, & Perrimon, 2010). Supporting the prior model, muscle-specific stress fibers act as precursors for cardiomyocyte sarcomere assembly (Fenix et al., 2018). Force-dependent mechanisms are also important for striated muscle patterns in Drosophila abdominal muscles that form from actomyosin myofibrils by mechanical tension and twitching (Weitkunat, Brasse, Bausch, & Schnorrer, 2017).

Key molecules in actin filament assembly are actin nucleators, which promote either the branched or linear addition of G-actin to form a filament. Branched actin nucleation promotion factors include the Wiskott–Aldrich syndrome (WAS) family of proteins, consisting of Wiskott–Aldrich syndrome protein (WASP), Scar/ WASP-family verprolin-homologous protein (WAVE), and Wiskott-Aldrich syndrome protein and SCAR homologue (WASH), among others (Kramer, Piper, & Chen, 2022; Rodriguez-Mesa, Abreu-Blanco, Rosales-Nieves, & Parkhurst, 2012). Spire and formins are involved in generating long, unbranched actin filaments (Bosch et al., 2007; Kobielak, Pasolli, & Fuchs, 2004; Sagot, Rodal, Moseley, Goode, & Pellman, 2002). The Drosophila genome encodes for six formin homology proteins – diaphanous (dia), Dishevelled Associated Activator of Morphogenesis (DAAM), cappuccino (capu), formin 3 (form3), Formin homology 2 domain-containing (Fhos), and Formin-like (Frl) (Tóth, Földi, & Mihály, 2022). Mouse Dia assembles dorsal stress fibers (Hotulainen & Lappalainen, 2006; Watanabe, Kato, Fujita, Ishizaki, & Narumiya, 1999) and the Drosophila DAAM and Fhos proteins promote thin filament assembly in adult muscles (Molnár et al., 2014; Shwartz, Dhanyasi, Schejter, & Shilo, 2016). These and other diverse functions performed by formins are of wide interest to better understand normal biology during development and how mutations result in disease (Labat-de-Hoz & Alonso, 2021).

Control of actin fiber dynamics involves kinases, such as Calcium/calmodulin-dependent protein kinase kinase 2 (CAMKK2) and Rho-associated protein kinase (ROCK) (Nishimura, Honda, & Takeichi, 2012; Tojkander, Ciuba, & Lappalainen, 2018). ROCK-mediated phosphorylation of myosin light-chain phosphatase (MLCP) and LIM domain kinase (LIMK) is involved in stress fiber assembly and actin polymerization in mammalian cell lines (Gorovoy et al., 2005; Kassianidou, Hughes, & Kumar, 2017; Sumi, Matsumoto, Takai, & Nakamura, 1999). An understudied protein implicated in actomyosin dynamics is NUAK, a member of the AMP-activated protein kinase (AMPK) family with broad and overlapping cellular targets (Sun, Gao, Chien, Li, & Zhao, 2013). Mammalian NUAK1 phosphorylates Myosin phosphatase target subunit 1 (MYPT1) leading to a decrease in phosphatase activity, thus increasing phosphorylated Myosin light chain-2 (MLC2) for actin stress fiber assembly (Zagórska et al., 2010). NUAK2 is targeted to stress fibers via myosin phosphatase Rho-interacting protein (MRIP) to induce actin stress fiber formation through MYPT1-mediated MLC dephosphorylation (Vallenius et al., 2011). This actomyosin network localization of NUAK2 is also found in the apical region of neural tube cells (Bonnard et al., 2020). Several studies in vertebrates implicate a role for NUAK in various cancers, possibly through actomyosin contractility, whereby NUAK1 knockdown reduces actin polymerization in a glioblastoma cell line (Lu et al., 2013). While these data together imply that NUAK proteins may be important regulators of actomyosin structures in vastly different biological and cellular contexts, the underlying mechanisms remain unclear.

Drosophila NUAK (NUAK) is the homolog of NUAK1 and NUAK2 (Hirano et al., 2006). Two pieces of evidence suggest a link between NUAK and the regulation of actin structures. First, NUAK knockdown affects actin cone maturation in spermatid formation (Couderc, Richard, Vachias, & Mirouse, 2017). Second, we previously demonstrated that sarcomere length, including that of the thin filament, increases upon overexpression of NUAK (Brooks et al., 2022). Further evidence for NUAK-mediated muscle formation emerged from studies in C. elegans whereby the homolog Unc-82 is required for thick filament assembly (Schiller et al., 2017). However, the mechanism of NUAK function is largely unexplored in relation to actin fiber assembly in muscles. Since actomyosin patterns underlie sarcomere structure and function, these minimal contractile units of muscle provide an elegant system to study the role of NUAK in myofibril actin structure formation leading to mature sarcomeres. Here we leverage an overexpression system that produces excess actin-containing myofibrils on the surface of muscle and for the first time show a unique ring-like structure juxtaposed next to actin fibers, possibly enhancing their formation.

Results and Discussion

Increased NUAK levels cause aberrant myofiber formation

The lengthening of thin filaments upon muscle-specific NUAK overexpression suggested an increase in filament assembly or a decrease in filament disassembly (Brooks et al., 2022). For further analysis, we expressed NUAK in muscle using the Gal4-UAS system (Brand & Perrimon, 1993) and focused on the ventral longitudinal muscles 3 (VL3) and 4 (VL4) in the abdominal segments of third-instar larvae (L3) for sarcomere measurements. Properly organized muscle striations were observed below nuclei in w¹¹¹⁸ control muscles (Fig. 1A–A’’). While organized myofibers were also present in distal planes below the nuclei in Mef2>NUAK muscles (Fig. 1B’, B’’), the sarcomeres generated here were longer than controls (Fig. 1C,D). We also observed numerous aberrant fibers in the superficial layer of muscle above the nuclei plane (Fig. 1B, B’’). Frequently these excess fibers resulted in the ‘joining’ of myofibrils from one muscle to another (Fig. 1B, asterisk). These observations suggest that optimum levels of NUAK expression and/or NUAK activity are necessary for the coordinated patterning of sarcomeres and myofibrils.

Figure 1. Sarcomere length and patterning is altered upon muscle-specific NUAK overexpression.

Maximum intensity projections of VL3 and VL4 L3 muscles stained to label Actin (cyan), NUAK (magenta) or nuclei (yellow). (A-A’’) w¹¹¹⁸ (wild type) larval muscle preparation showing the top, or superficial layer in A, and the middle layer in A’. A’’ shows the Z series where nuclei are located on the top surface. (B-B’’) In Mef2>NUAK larval muscle preparations, aberrant muscle fibers are evident in the superficial layer (B). Though the sarcomere pattern is not altered in the middle layer, sarcomeres are longer (B’). B’’ shows aberrant muscle fibers are present above the nuclei. In a given length of VL4 muscle Mef2>NUAK has 19 sarcomeres whereas w¹¹¹⁸ has 27 sarcomeres (compare A’’ and B’’). Scale bar is 40 μm for A-B’. (C-D) Box and whisker plots showing sarcomere length changes in the VL3 and VL4 muscles upon NUAK overexpression. Laser power and gain were adjusted in Mef2>NUAK to avoid any saturation of the signal. N=8 for Mef2>lacZ.NLS and N=15 (C) or N=16 (D) for Mef2>NUAK. Two-tailed student t-test was performed. Mean +/− S.D. (*, p < 0.05; ***, p < 0.005)

NUAK expression pattern is dynamic during development

To explore the role of NUAK in actin assembly, we first assessed NUAK localization in muscles using antibody staining. To validate antibody specificity, we immunostained muscles from Mef2>NUAK RNAi larvae to confirm a decrease in NUAK levels upon knockdown (Fig. S1A–C). This result was substantiated via Western blotting whereby NUAK protein levels were reduced in NUAK RNAi muscles compared to controls (Fig. S1D, D’). Finally, we further confirmed NUAK protein localization in muscles with an HA-tagged construct of NUAK-miniTurbo under control of its endogenous promoter. We found that there is strong colocalization of HA and NUAK antibody staining (white arrowheads in Fig. S1E–E’’’). These data together validate the use of this antibody as a tool to assess NUAK protein localization and/or levels in larval muscle tissue.

Since modulation of NUAK protein levels alters sarcomere number and size (Fig. 1), we first assessed the pattern of NUAK expression during larval muscle growth. In wandering L3 body wall muscles, NUAK was weakly and broadly detected at the Z-disc (Fig. 2B, white arrows in i) and thick filament regions (Fig. 2B, white arrowheads in i). However, we occasionally, but consistently, observed enrichment of NUAK in regions that coincided with F-actin (Fig. 2A, A’ and Fig. 2B, white arrows in ii). Since animals at the wandering L3 stage have largely ceased muscle growth to undergo pupariation, we hypothesized that NUAK accumulation may be more frequent in early L3 larvae undergoing continual sarcomere addition (Balakrishnan et al., 2020). Even though the normal distribution of NUAK was prevalent (Fig. 2D, white arrow and arrowheads in iii), all larvae analyzed (n=8) showed enrichment of NUAK in regions that overlap with F-actin (Fig. 2C, C’ and Fig. 2D, white arrows in iv).

Figure 2. NUAK is enriched in regions corresponding to the thin filament.

(A-D’) w¹¹¹⁸ larval muscle preparation shows localized enrichment of NUAK in wandering L3 (wL3) (A-B’) and early L3 (eL3) stage (C-D’) larva. Actin is labeled in cyan, NUAK in magenta, and nuclei in yellow. Enrichment of NUAK is seen throughout the Z-plane (A’, B’ in wL3 and C’, D’ in eL3). i) and iii) NUAK localizes in a broad sarcomeric pattern that corresponds to the thick filament (white arrowheads) and the Z-disc (white arrows). ii) and iv) Representative confocal images show enrichment of NUAK (white arrows) in regions that overlap with F-actin staining in A and C. (E-F’’) Examples of NUAK enrichment in Mef2>lacZ wL3 (E-E’’) and eL3 muscles (F-F’’). Actin is labeled in cyan, NUAK in magenta, and z-disc (Kettin/Sls) in yellow. E-E’’) NUAK overlaps with Z-disc staining (white arrow) and is enriched at borders of the thin filament (edge of F-actin). F-F’’) In eL3 muscles NUAK accumulates at Z-discs (white arrows) and regions bordering F-actin (yellow arrows). Scale bar in A is 40 μm for A-D and 20 μm for i-iv, E-F’’.

We were intrigued by these patterns of NUAK enrichment that corresponded to different regions in the thin filament marked by F-actin. To further understand NUAK localization, we immunostained both wandering and early L3 larval muscles with the Z-disc marker Kettin/Sls (Lakey et al., 1990). Indeed, we observed NUAK enrichment at the Z-disc (white arrows in Fig. 2E–E’’ and Fig. 2F’F’’) or brightly stained regions that corresponded to the edge of the thin filament (yellow arrows in Fig. 2F’F’’). The presence of aberrant myofibrils, increase in sarcomere length upon NUAK overexpression, and enrichment of NUAK to thin filaments, suggest a role for NUAK in actin-mediated sarcomere and/or myofibril assembly.

Frl and NUAK colocalize at actin assembly sites

Since actin nucleators are known regulators of actin filament assembly, we looked for changes in actin nucleator levels using available antibodies. We found that Mef2>NUAK muscles (Figs. 3C–C’’) showed an obvious increase in Frl protein levels compared to control (Mef2>lacZ.NLS) muscles (Figs. 3A–A’’). Moreover, in NUAK overexpression muscles, there was a clear enrichment and colocalization of Frl with the thin filament marked by actin not present in Mef2>lacZ.NLS muscles (dashed boxes in Figs. 3B,D). Validation of this anti-Frl antibody was previously demonstrated in Frl mutant embryos (Dollar et al., 2016). Since we were assessing Frl in larval muscles, we further confirmed antibody specificity in this tissue. We first used Western blotting to demonstrate a reduction in Frl protein levels in Actin5C (Act5C)>Frl RNAi larvae compared to controls (Fig. S2A). Next, we showed that expression of wild-type (WT) Frl in a single muscle (5053>Frl WT) was detected at higher levels by the Frl antibody than in surrounding muscles expressing Frl at endogenous levels (Fig. S2B).

Figure 3. Coordinate increase of NUAK and Frl in NUAK overexpression muscles.

(A-G) Dissections and immunostainings of L3 muscles with Actin labeled in cyan, NUAK in magenta, and Frl in yellow. (A-B) Frl is expressed at low levels in Mef2>lacZ.NLS (control) muscles (see white line in A’ and corresponding line plot in B). (C-D) In Mef2>NUAK overexpression muscles, the levels of Frl are substantially increased, enriched in nuclei (C’) and partially overlap with Actin (see white line in C’ and corresponding line plot in D where white dashed rectangle marks the Actin-enriched thin filament). (E) In Mef2>lacZ.NLS muscles there is localized accumulation of NUAK and Frl (white dashed box). (F-F’’) This NUAK enrichment colocalizes with Actin and Frl (white arrowheads). (G) In Mef2>NUAK muscles, ring-like NUAK structures (white arrowheads) partially overlap with Frl staining. This is a maximum intensity projection of the superficial layer of a larval muscle where extra myofibers are observed. (H-H’’’) High magnification images of a NUAK ring showing colocalization with Frl and Actin. An actin fiber is seen coming out from the NUAK ring (cyan asterisk in H’’). Scale bar is 40 μm in A, C, E, 10 μm in G, and 2 μm in H.

To test if the increase in Frl coincided with NUAK-enriched sites, we co-labeled Frl with NUAK and actin in early L3 control muscles undergoing continual sarcomere addition. Indeed, NUAK accumulated at the Z-disc coinciding with Frl and actin (dotted area in Fig. 3E and white arrowheads in Fig. 3F), further supporting the interplay between NUAK and actin assembly. In contrast, the actin nucleator Scar was not altered in control or NUAK-overexpressing muscles (Fig. S3A–D). We also examined the distribution of two other formins, Formin3 and Fhos. Formin3 was not altered upon NUAK overexpression whereas Fhos showed a mild increase upon NUAK overexpression (Fig. S3E–H). We next wanted to explore molecular mechanisms that may contribute to elevated Frl protein levels. Real-time PCR on control or NUAK-overexpressing muscles revealed no alteration in Frl mRNA levels (Fig. S2D), thus ruling out transcriptional regulation. Next we mimicked NUAK kinase activity to determine the impact on Frl expression and/or localization through overexpressing an activated version of NUAK (Mef2>NUAK T226E) in muscles (Zhao, Brooks, Guo, & Geisbrecht, 2023). Indeed, increased kinase activity phenocopied the increase in Frl levels similar to that of wild type NUAK overexpression (Fig. S4A–D). These results together suggest that the accumulation of Frl may be due to increased protein stabilization or decreased protein degradation, possibly relying on NUAK kinase activity.

In NUAK-overexpressing muscles that contain excess myofibrils in the superficial muscle layer, we noticed NUAK-enriched structures that resemble actin rings or barrels in both L2 or L3 muscles (white arrowheads in Figs. 3G and S5A,A’). We next immunostained for Glutactin (Glt) to label the basement membrane and confirmed that these NUAK structures are beneath the Glt layer and associated with myofibers (Fig. S5B–B’’). Closer examination revealed that the NUAK rings showed Frl puncta associated with actin staining, including actin fibers emanating from these structures (Fig. 3H–H’’’ and movie S1). These data show that Frl enrichment with actin may be a common feature associated with NUAK overexpression, both in sarcomere assembly during normal muscle growth and at unique ring-like structures present just above the muscle plane that may give rise to aberrant myofibers.

Protein synthesis increases upon NUAK overexpression

An increase in actin assembly would require newly synthesized proteins to accumulate at the site of thin filament synthesis. To test this, we used non-canonical amino acid tagging (NCAT) to label nascent proteins specifically in muscles. Expression of a mutant version of methionyl-tRNA synthetase tagged with EGFP (Mef2>dMetRS^L262G-EGFP) can charge methionyl-tRNA with the methionine surrogate azidonorleucine (ANL) (Erdmann et al., 2015). Larvae reared on cornmeal agar without ANL were used as a negative control. There was a clear difference between the Strep/ANL signal and negative control muscles (Fig. 4A–B’’), especially in the top layer of the muscle (Fig. 4C, D). Compared to the actin and EGFP signals, we found that the ANL signal decreased in the distal planes of the muscle (Fig. 4C’, D’). This suggests that newly synthesized proteins accumulate preferentially in the superficial layer of muscle near nuclei. We next performed NCAT labeling in control or NUAK-overexpressing muscles and observed elevated tetramethylrhodamine (TAMRA)/ANL signal corresponding to increased protein synthesis (Fig. 4E–G). The ring-like structures that accumulated in NUAK overexpression muscles showed a strong ANL signal that colocalized with NUAK and puncta labeled for Frl puncta (Figs. 4H and S5C–C’’’). These data further validate a link between NUAK, Frl and new protein synthesis and further suggest that these NUAK rings could be the site of new filament assembly.

Figure 4. NUAK overexpression causes an increase in nascent protein expression.

(A-B’’) L3 larval muscle preparations expressing Mef2>MetRS^L262G-EGFP where EGFP is yellow, Actin is cyan, and newly synthesized proteins are marked by Strep in magenta. Mef2>MetRS^L262G-EGFP L3 larva reared on normal food (negative control) show no ANL labeling (A’’), while ANL-fed larva shows specific labeling (B’’). (C-D’) The change in the signal of ANL labeling (Strep) plotted in C’ and D’ are based on a muscle image from Mef2>MetRS^L262G-EGFP L3 larva fed on normal (C) or ANL food (D). Graphs depicts the drop in ANL labeling (in magenta) in deeper muscle planes. The EGFP and Actin signals are plotted in yellow and cyan for comparison. (E-G) ANL labeling (TAMRA) shows increased signal in L3 larval muscles of Mef2>MetRSL262G-myc, UAS-NUAK (F) in comparison to Mef2>MetRSL262G-myc, UAS-lacZ.NLS (E). Quantitation of the TAMRA signal plotted in the bar and whisker plot shown in G. Two-tailed student t-test. Mean +/− S.D. (**, p < 0.01). N=9 for Mef2>lacZ.NLS and N=11 for Mef2>NUAK. (H) NUAK rings show colocalization with TAMRA (new protein synthesis) and Frl puncta. The NUAK ring and Frl puncta are juxtaposed next to the actin fiber. Scale bar is 40 μm for A-F and 2 μm for H.

To determine if Frl activity contributes to aberrant myofiber formation similar to that of NUAK overexpression, we expressed a form of Frl lacking the GTPase binding domain (GBD) and the diaphanous inhibitory domain (DID) (Dollar et al., 2016). Expressing two copies of this constitutively active Frl (Frl CA) in all larval muscles (Mef2>Frl CA) (Fig. 5A) or only in muscle VL1 (5053>Frl CA) (Fig. 5B) produced ectopic myofibers similar to Mef2>NUAK muscles. This excess myofiber phenotype was not observed in muscles expressing Frl WT (Fig. S2B’), even though a single copy of this transgene produced higher protein levels than two copies of Frl CA (Fig. S2C). We also observed rings decorated with Frl in Mef2>Frl CA muscles (Fig. 5C,C’), often in close proximity to F-actin structures (Fig. 5C,C’’), raising the intriguing possibility that this active form of Frl is promoting F-actin assembly.

Figure 5. Frl CA produces aberrant myofibrils and ring-like structures.

(A-C’’) L3 muscles stained for F-actin (cyan) or Frl (yellow). (A,B) Unorganized myofibrils are present in muscles VL3 and VL4 (Mef2>Frl CA) (A) or muscle VL1 (5053A>Frl CA) (B). Images are maximum intensity projections. (C-C’’) Single plane images that show Frl is present in punctate structures (C) that resemble rings (white arrowheads in C’) and are often associated with F-actin filaments (C’’). Scale bar is 20 μm for A and B and 5 μm for C-C’.

We have provided a possible mechanism for actin fiber assembly that may be relevant to muscle and non-muscle cells, whereby NUAK increases Frl localization and possibly activity. We do not yet know if this requires kinase activity, as NUAK1 and NUAK2 can regulate actin stress fiber assembly in kinase-dependent and kinase-independent manners, respectively (Vallenius et al., 2011; Zagórska et al., 2010). Human Formin-like 1 (FMNL1), the Frl homolog, is expressed at low levels in human skeletal muscles (Gardberg et al., 2014) and is required for actomyosin assembly on lipid droplets (Pfisterer et al., 2017). However, there is scant information about functions of FMNL1 in actin-mediated muscle development. Altogether, we propose that excess NUAK forms a specialized structure, possibly through increasing protein synthesis, which recruits actin assembly machinery. This upregulation allows us to capture actin assembly events that are not possible during normal development and may describe a new mechanism for actin assembly involving Frl.

Materials and methods

Drosophila Genetics

The following stocks were obtained from the Bloomington Drosophila Stock Center (BDSC): w¹¹¹⁸ (BDSC:3605, RRID:FBal0018186), Mef2-Gal4 (BDSC:27390, RRID:FBti0115746), Act5C-Gal4 (BDSC:3954, RRID:FBti0012292), 5053A-Gal4 (BDSC:2702, RRID:FBti0001257), UAS-lacZ.NLS (BDSC:3956, RRID:FBti0012289), UAS-NUAK RNAi (BDSC:35194, RRID:FBti0144169), and UAS-Frl RNAi (BDSC:32447, RRID:FBti0132141). Stocks generated in the Geisbrecht lab: UAS-NUAK (RRID:FBal0360165) (Brooks et al., 2020) and UAS-NUAK T226E (RRID:FBal0397186) (Zhao et al., 2023). The NUAK-miniTurbo-HA line was generated by WellGenetics Inc. The UAS-dMetRS^L262G-EGFP/TM6b,Tb Hu (RRID:FBal0314188) and UAS-dMetRS^L262G-3xmyc/CyO (RRID:FBal0314189) stocks were kindly provided by Ulrich Thomas (Erdmann et al., 2015). UAS-Frl WT (RRID:FBal0318781), (on II) and UAS-Frl CA (on III) (RRID:FBal0318782) were generously provided by Andreas Jenny (Dollar et al., 2016). The individual UAS-Frl CA stocks were combined together to produce UAS-Frl CA, UAS-Frl CA (II,III) for the experiments in Fig. 5. Fly stocks were maintained on cornmeal agar medium at 25°C while genetic crosses were reared at 31°C unless otherwise indicated.

Dissection & Immunostaining

Either L2 larvae (~72 hrs AEL), early feeding (~96 hrs AEL) or wandering (~110 hrs AEL) L3 larvae were heat killed and dissected in 1X PBS on a sylgard plate and fixed in 4% (v/v) formaldehyde followed by three washes with 0.5% PBT (0.5% Tween-20 in 1X PBS) and blocked in PBST + 5% (v/v) normal goat serum (Vector laboratories, S-1000). Primary antibody incubation was performed overnight at 4°C followed by three washes with 0.5% PBT. Secondary antibodies were incubated for two hours at room temperature and washed three times with 0.5% PBT. The following antibodies were used: rabbit anti-NUAK (1:200, catalog # DZ41105, Boster Bio, Pleasanton, CA), mouse anti-sls 1B8–3D9 (1:200, Developmental Studies Hybridoma Bank, Iowa City, IA), rat anti-Frl (1:500, Jozsef Mihaly) (Tóth et al., 2022), rabbit anti-Glt #1 (peptide sequence SVGLRPDYNDYSDE, 1:200, Genscript, Piscataway, NJ), rat anti-Fhos (1:200, Eyal Schejter), rabbit anti-Form3 (1:100, Daniel N. Cox) and guinea pig anti-Scar (1:50, Jennifer Zallen). F-actin was labeled with either phalloidin 405, 488, or 647 (catalog #A30104, #12379, #A22287, ThermoFisher, Waltham, MA). Nuclei were labeled with Hoechst dye (1:10000, catalog #H3570, ThermoFisher, Waltham, MA). The following secondary antibodies (1:200 or 1:400) were from Thermofisher, Waltham, MA: donkey anti-rat IgG 488 #A21208, donkey anti-rabbit IgG 594 #A21207, donkey anti-mouse IgG 647 #31571, goat anti-rabbit IgG 488 #A11008, goat anti-rabbit IgG 647 #A21244, goat anti-guinea pig IgG 488 #A11073. Other labeling reagents included Streptavidin 594 (catalog #S11227, ThermoFisher, Waltham, MA) and TAMRA (catalog #900932, Sigma Aldrich, St. Louis, MO). Samples were imaged on Zeiss 700 confocal microscope with Zen black software.

Antibody Verification by Western Blotting

Western blots were used to verify the NUAK or Frl antibodies. Genotypes for the RNAi experiments (Fig. S1D or S2A) were w¹¹¹⁸ (control), Mef2>NUAK RNAi and Act5C>Frl RNAi reared at 31°C (n = 6 for each genotype). Genotypes for Frl WT or Frl CA verification: Mef2>lacZ.NLS, Mef2>Frl WT, Mef2>Frl CA reared at 22°C. For each biological replicate, three whole larvae were rinsed with 0.7% (w/v) NaCl/0.04% (v/v) Triton X-100 and then rinsed with ultrapure water. The larvae were placed in 100μL of 1X Protein Loading Buffer (LI-COR, Lincoln, NE), boiled for 3 min, homogenized, and boiled for an additional 10 min. The samples were centrifuged for 2 min at 21,000 xg and loaded on a 6% polyacrylamide gel prepared from Invitrogen SureCast reagents (ThermoFisher, Waltham, MA). Gels were run with standard Tris-Glycine SDS running buffer in a BioRad Mini-PROTEAN Tetra Cell (Hercules, CA) and transfers were performed using a BioRad Transblot Turbo System (Hercules, CA) with Bjerrum Schafer-Nielsen (1x Tris-Glycine) SDS transfer buffer to a Millipore Immobilon-F 0.45μm polyvinylidene fluoride (PVDF) membrane (ThermoFisher, Waltham, MA). Total protein was visualized using the Revert Total Protein Stain (LI-COR, Lincoln, NE). Blots were blocked in BioRad EveryBlot Blocking buffer (Hercules, CA). The blots were incubated overnight in either rabbit anti-NUAK (1:1000) or rat anti-Frl (1:1000) primary antibodies diluted in blocking buffer. Blots were washed with standard Tris-buffered saline (TBS) + 0.1% SDS and incubated with the following secondary antibodies for 2 hrs at room temperature (~22°C): IRDye 800 CW Goat anti-Rabbit (LI-COR, Lincoln, NE) or anti-Rat HRP (ThermoFisher, Waltham, MA). After washing, the NUAK blot was imaged using the LI-COR Odyssy XF (LI-COR, Lincoln, NE). Prior to imaging with the same LI-COR instrument and software, the Frl blot was incubated with SuperSignal West Femto Maximum Sensitivity substrate (ThermoScientific, Waltham, MA). For the NUAK blot, normalized fold change was determined using the Analyze Gels function in ImageJ with total protein as the loading control. LI-COR Empiria software (LI-COR, Lincoln, NE) was used to calculate normalized fold change for Frl with total protein as the loading control. Graphs and statistics were generated with GraphPad Prism 10 software. Mann-Whitney t-tests were used to calculate significance.

ANL Labeling

A mutant form of Methionyl tRNA synthetase (MetRS) tagged with either EGFP or Myc (dMetRS^L262G-EGFP /dMetRS^L262G-myc) was expressed in the larval musculature. The mutant MetRS incorporates the non-canonical amino acid azidonorleucine (ANL) (catalog # HAA9390, Iris Biotech GmbH, Marktredwitz, Germany) in place of methionine in a ratio of 1:10 (Erdmann et al., 2015). ANL-labeled experimental flies for were fed on 4 μm ANL food (final concentration) for 18–20 hrs before dissection in HL-3 solution (70 mM NaCl, 5 mM KCl, 20 mM MgCl₂, 10 mM NaHCO₃, 115 mM Sucrose, 5 mM Trehalose, 5 mM HEPES, 0.1 mM CaCl₂) and prefixed in two drops of 4% paraformaldehyde (PFA) in 1X PB (pH 7.2) for 1 minute. Samples were then fixed for 30 minutes in 4%PFA in 1X PB (pH 7.2). After fixation, samples were washed 3 times with 0.2%Triton X-100 in 1X PBS (pH 7.2) followed by 3 washes in 1X PBS (pH 7.8) for 15 mins each at room temperature on a shaker platform with gentle shaking. To tag the ANL labelled proteins, the labeling mixture of TBTA (200 μM), Biotin-Alkyne (25 μM) or Tamra-Alkyne (1 μM), TCEP (500 μM), and CuSO4 (200 μM) in 1X PBS (pH 7.8) was prepared. Reagents were added in the sequence as mentioned above, after addition of each reagent the solution was vortexed for 10 seconds. The larval body wall muscles were incubated in 200 μl of the labeling mixture at 4°C overnight on a shaker platform with gentle shaking. Samples were then washed with 1X PBS-Tween20 (pH 7.4) and 1X PBT three times each for 15 mins at room temperature before incubation with primary and secondary antibodies. For fluorescent labeling, either Strep-594 or TAMRA dye was used. Immunostaining for NUAK or Frl were performed after NCAT labeling as described above.

qPCR

Total RNA was collected from eight wandering L3 larvae which were shifted from 25°C to 31°C at the late embryo/early L1 stage. RNA was prepared from dissected muscle carcasses and purified using the RNeasy Mini Kit (Qiagen, Hilden, Germany). Three RNA samples for each genotype were prepared. cDNA was synthesized from 100ng of RNA using qScript XLT cDNA SuperMix kit (Quanta Biosciences, Beverly, MA). For the qPCR reactions, 1:10 dilutions of the cDNA were combined with PowerUp SYBR Green Master Mix (ThermoFisher, Waltham, MA). The following primers were used (final concentration = 0.5μM): Frl forward 5’-CATGCCCACCACAGATGAG, reverse 5’-ATCATGTCCCATTTCTTCTCGTC

rp49 forward 5′-GCCC¬AAGGGTATCGACAACA, reverse 5′-GCGCTTGTTCGATCCGTAAC. All primers were synthesized by Integrated DNA Technologies (IDT, Stokie, IL). Quantitative transcript levels were obtained using a QuantStudio 3 instrument with QuantStudio Design and Analysis software (TheroFisher, Waltham, MA). Normalize fold changes were calculated using the 2-ΔΔCt method and graphed as Mean ± SEM using GraphPad 9.5.0.

Image Analysis

Images were cropped and assembled using ImageJ and Adobe Photoshop. All image analysis was performed using ImageJ. For sarcomere length measurements (Figs. 1C and D), total muscle length was divided by the number of sarcomeres. Muscle length was measured by drawing a line along the muscle length in ImageJ by using Analyze>Measure function. The number of sarcomeres were counted manually (VL3 muscles: n = 8 for w¹¹¹⁸ and n = 15 for Mef2>NUAK, VL4 muscles: n= 8 for w¹¹¹⁸ and n = 16 for Mef2>NUAK). For intensity plot profiles (Figs. 3B,D) a line was drawn along the muscle lengths in Mef2>lacZ.NLS (fig. 3A’) and Mef2>NUAK (Fig. 3C’) and the plot was made in ImageJ using Analyze>Plot Profile function. The same procedure was followed for Fig. S3B,D,F,H and Fig. S4B,D. For intensity measurements in Fig. 4C’ and D,’ a single representative image of the Mef2>MetRSL262G-EGFP larvae fed on normal food (Fig. 4 C,C’) or ANL food (Fig. 4 D,D’) was used. A box was drawn in the VL4 muscle region and mean intensity was measured using Analyze>Measure function. Intensities were measured individually in each plane for every channel. Intensity vs z-plane was plotted in a different color for each channel. For the intensity measurement in Fig. 4G, a box was drawn in the region with strong TAMRA signal enrichment in Mef2>MetRSL262G-myc,NUAK larval muscles (n = 11) and compared with a similar region in Mef2>MetRSL262G-myc,lacZ.NLS larval muscles (n = 9). Intensity measurement was performed in ImageJ with Analyze>Measure function. For NUAK intensity measurements in Fig. S1C, five samples of each control and NUAK RNAi were used. For each, three random boxes of same size were drawn in the VL3/VL4 muscle in ImageJ and the intensity was measured with the with Analyze>Measure function. To normalize the intensity, the average mean intensity from the control sample was used. All graphs were plotted in GraphPad 9.5.0.

Data availability

All relevant data can be found within the article and its supplementary information.