1. Developmental Biology

A population of adult satellite-like cells in Drosophila is maintained through a switch in RNA-isoforms

  1. Hadi Boukhatmi
  2. Sarah Bray  Is a corresponding author
  1. University of Cambridge, United Kingdom
https://doi.org/10.7554/eLife.35954
Share this article
Cite this article
  1. Hadi Boukhatmi
  2. Sarah Bray
(2018)
A population of adult satellite-like cells in Drosophila is maintained through a switch in RNA-isoforms
eLife 7:e35954.
https://doi.org/10.7554/eLife.35954
  2. Download BibTeX
  3. Download .RIS

Abstract

Adult stem cells are important for tissue maintenance and repair. One key question is how such cells are specified and then protected from differentiation for a prolonged period. Investigating the maintenance of Drosophila muscle progenitors (MPs) we demonstrate that it involves a switch in zfh1/ZEB1 RNA-isoforms. Differentiation into functional muscles is accompanied by expression of miR-8/miR-200, which targets the major zfh1-long RNA isoform and decreases Zfh1 protein. Through activity of the Notch pathway, a subset of MPs produce an alternate zfh1-short isoform, which lacks the miR-8 seed site. Zfh1 protein is thus maintained in these cells, enabling them to escape differentiation and persist as MPs in the adult. There, like mammalian satellite cells, they contribute to muscle homeostasis. Such preferential regulation of a specific RNA isoform, with differential sensitivity to miRs, is a powerful mechanism for maintaining a population of poised progenitors and may be of widespread significance.

https://doi.org/10.7554/eLife.35954.001

Introduction

Growth and regeneration of adult tissues depends on stem cells, which remain undifferentiated while retaining the potential to generate differentiated progeny. For example, muscle satellite cells (SCs) are a self-renewing population that provides the myogenic cells responsible for postnatal muscle growth and muscle repair (Chang and Rudnicki, 2014). One key question is how tissue specific stem cells, such as satellite cells, are able to escape from differentiation and remain undifferentiated during development, to retain their stem cell programme though-out the lifetime of the animal.

It has been argued that the progenitors of Drosophila adult muscles share similarities with satellite cells and thus provide a valuable model to investigate mechanisms that maintain stem cell capabilities (Aradhya et al., 2015; Figeac et al., 2007). After their specification during embryogenesis, these muscle progenitors (MPs) remain undifferentiated throughout larval life before differentiating during pupal stages. For example, one population of MPs is associated with the wing imaginal disc, which acts as a transient niche, and will ultimately contribute to the adult flight muscles. These MPs initially divide symmetrically to amplify the population. They then enter an asymmetric division mode in which they self-renew and generate large numbers of myoblasts that go on to form the adult muscles (Gunage et al., 2014). In common with vertebrates, activity of Notch pathway is important to maintain the MPs in an undifferentiated state (Gunage et al., 2014; Mourikis and Tajbakhsh, 2014; Mourikis et al., 2012; Bernard et al., 2010). To subsequently trigger the muscle differentiation program, levels of Myocyte Enhancer factor 2 (Mef2) are increased and Notch signalling is terminated (Elgar et al., 2008; Bernard et al., 2006). Until now it was thought that all MPs followed the same fate, differentiating into functional muscles. However, it now appears that a subset persist into adulthood forming a population of satellite-like cells (Chaturvedi et al., 2017 and see below)). This implies a mechanism that enables these cells to escape from differentiation, so that they retain their progenitor-cell properties.

The Drosophila homologue of ZEB1/ZEB2, Zfh1 (zinc-finger homeodoman 1), is a candidate for regulating the MPs because this family of transcription factors is known to repress Mef2, to counteract the myogenic program (Siles et al., 2013; Postigo et al., 1999). Furthermore, zfh1 is expressed in the MPs when they are specified in the embryo and was shown to be up-regulated by Notch activity in an MP-like cell line (DmD8) (Figeac et al., 2010; Krejcí et al., 2009). In addition, an important regulatory link has been established whereby microRNAs (miRs) are responsible for down-regulating ZEB/Zfh1 protein expression to promote differentiation or prevent metastasis in certain contexts (Zaravinos, 2015; Vandewalle et al., 2009). For example, the miR-200 family is significantly up-regulated during type II cell differentiation in fetal lungs, where it antagonizes ZEB1 (Benlhabib et al., 2015). Likewise, miR-8, a miR-200 relative, promotes timely terminal differentiation in progeny of Drosophila intestinal stem cells by antagonizing zfh1 and escargot (Antonello et al., 2015). Conversely, down-regulation of miR-200 drives epithelial mesenchymal transition (EMT) to promote metastasis in multiple epithelial derived tumours (Korpal et al., 2008; Park et al., 2008). Such observations have led to the proposal that the ZEB/miR-200 regulatory loop may be important in the maintenance of stemness, although examples are primarily limited to cancer contexts and others argue that the primary role is in regulating EMT (Antonello et al., 2015; Brabletz and Brabletz, 2010). The MPs are thus an interesting system to investigate whether this regulatory loop is a gatekeeper for the stem cell commitment to differentiation.

To investigate the concept that ZEB1/Zfh1 could be important in sustaining progenitor-type status, we examined the role and regulation of zfh1 in Drosophila MPs/SCs. Our results show that zfh1 plays a central role in the maintenance of undifferentiated MPs and, importantly, that is expression is sustained in a population of progenitors that persist in adults (pMPs) through the activity of Notch. Specifically these pMPs express an alternate short RNA isoform of zfh1 that cannot be targeted by miR-8. In contrast, the majority of larval precursors express a long isoform of zfh1, which is subject to regulation by miR-8 so that Zfh-1 protein levels are suppressed to enable differentiation of myocytes. Expression of alternate zfh1-short isoform is thus a critical part of the regulatory switch to maintain a pool of progenitor ‘satellite-like’ cells in the adult. This type of regulatory logic, utilizing RNA isoforms with differential sensitivity to miRs, may be of widespread relevance for adult stem cell maintenance in other tissues.

Results

Zfh1 is required for maintenance of muscle progenitors

As zfh1 was previously shown to antagonize myogenesis (Siles et al., 2013; Postigo et al., 1999) it is a plausible candidate to maintain the muscle progenitor (MP) cells in Drosophila and prevent their differentiation. Its expression is consistent with this hypothesis as Zfh1 is present throughout the large group of MPs associated with the wing disc, which can be distinguished by the expression of Cut (Ct) (Figure 1A–A’’). At early stages Zfh1 expression is uniform (Figure 1—figure supplement 1), but at later stages the levels become reduced in the cells with high Cut expression (Figure 1A’’). These cells give rise to the direct flight muscles (DFMs), whereas the remaining MPs, where Zfh1 expression is high, give rise to the indirect flight muscles (IFMs) (Figure 1A’’; Sudarsan et al., 2001). Zfh1 expression in MPs is therefore regulated in a manner that correlates with different differentiation programs.

Figure 1 with 2 supplements see all
Download asset Open asset
Zfh1 expression and function in MPs and in adult pMPs.

(A–A”) Zfh1 (Green) and Cut (Purple) expression in MPs associated with third instar wing discs, (A’–A’’) higher magnification (3X) of boxed region in A. Zfh1 is present in all MPs, but those with highest Cut expression have lower levels of Zfh1 (asterisk). Scale bars: 50 μM, (n > 30 wing discs from three biological replicates). (B–C) Down regulation of zfh1 induces premature differentiation of the MPs (arrowhead in C’-C’’). β3-Tubulin (β3-Tub, Red) and Tropomyosin (Tm, Green) expression in control (B, 1151-Gal4 > wRNAi) and Zfh1 depleted (C, 1151-Gal4 > zfh1 RNAi) third instar wing discs, (B’–C’’) higher magnification (3X) of boxed regions in B and C. (n > 20 wing discs; from three biological replicates). (D–D’) Zfh1 expression (red) indicates the existence of persistent muscle progenitors (pMPs; arrows) associated with the muscle fibres (Phalloidin (Green), DNA/Nuclei (Blue); n > 10 heminota; from three biological replicates). The immune cell marker P1 was included in the immunostaining and is absent from the pMPs (see Figure 1—figure supplement 2). Scale bars: 50 μM. (E–E’’’) Zfh1 (Red) expressing pMPs (e.g. arrows in E’’’) are closely embedded in the muscle lamina of the adult indirect flight muscles and express Mef2 (myogenic cells; Mef2-Gal4 >Src::GFP, green). Nuclei (Blue), Scale bars: 25 μM, (n > 10 heminota; from two biological replicates).

https://doi.org/10.7554/eLife.35954.002

To determine whether Zfh1 is required in the MPs to antagonize myogenic differentiation we tested the consequences from silencing zfh1 specifically in MPs, using 1151-Gal4 to drive expression of interfering RNAs (RNAi). Two independent RNAi lines led to the premature expression of Tropomyosin (Tm), a protein normally expressed in differentiated muscles, in the most severe (KK103205 line)~80% of zfh1-depleted wing discs exhibited Tm expression (Figure 1B–C and Figure 1—figure supplement 1D–G). Similarly, expression of a Myosin Heavy Chain (MHC) reporter was detected in ~20% of zfh1 (KK103205) depleted wing discs (Figure 1—figure supplement 1H–I) indicating that small muscle fibers had formed precociously. Consistent with the premature expression of these muscle differentiation markers, decreased zfh1 led to abnormal β−3Tubulin staining, showing that the residual cells had altered cell morphology in 90% of zfh1-depleted wing discs, (Figure 1B’–C’). These results demonstrate that reduced zfh1 expression causes MPs to initiate the muscle differentiation program indicating that Zfh1 is required to prevent MP differentiation.

Lineage tracing experiments suggest that a subset of wing disc MPs have characteristics of muscle stem cells and remain undifferentiated even in adult Drosophila. Recently, these have been shown to express Zfh1 (Chaturvedi et al., 2017; Gunage et al., 2014), which is compatible with our observation that Zfh1 is necessary to prevent differentiation in MPs. In agreement, adult IFM muscle fibers were associated with sparse nuclei that retained high levels of Zfh1 expression whereas the differentiated muscle nuclei exhibited no detectable expression (Figure 1D–D’). To better characterise these Zfh1 positive (+ve) adult cells, we expressed a membrane-tagged GFP (UAS-Src::GFP) under the control of a specific muscle driver Mef2-Gal4 (Mef2 >GFP). This confirmed that Zfh1 was expressed in myogenic Mef2 >GFP expressing cells, and revealed that these cells were closely embedded in the muscle lamina (Figure 1E–E’’’). Although many of the Zfh1 expressing cells were clearly co-expressing mef2 >GFP, Zfh1 was also detected in another population that lacked Mef2 expression. Often clustered, these cells were co-labelled with a plasmatocyte marker P1/Nimrod indicating that they are phagocytic immune cells (Figure 1—figure supplement 2). A subset of the Zfh1 +ve cells are therefore myogenic and have characteristics of persistent muscle progenitors that likely correspond to the so-called adult satellite cells recently identified by others (Chaturvedi et al., 2017) (Figure 1D–E).

The results demonstrate that Zfh1 is expressed in MPs, where it is required for their maintenance, and that its expression continues into adult-hood in a small subset of myogenic cells. If, as these data suggest, Zfh1 is important for sustaining a population of a persistent adult progenitors, there must be a mechanism that maintains Zfh1 expression in these cells while the remainder differentiate into functional flight muscles.

zfh1 enhancers conferring expression in MPs

To investigate whether the maintenance of Zfh1 expression in larval and adult MPs could be attributed to distinct enhancers, we screened enhancer-Gal4 collections (Jenett et al., 2012; Jory et al., 2012; Manning et al., 2012) to identify zfh1 enhancers that were active in larval MPs. From the fifteen enhancers across the zfh1 genomic locus that were tested, (Figure 2 and Figure 2—figure supplement 1A) three directed GFP expression in the Cut expressing MPs at larval stages (Figure 2B–D). These all correlated with regions bound by the myogenic factor Twist in MP-related cells (Figure 2—figure supplement 1A; Bernard et al., 2010). Enhancer 1 (Enh1; VT050105) conferred weak expression in scattered progenitors (Figure 2B). Enhancer 2 (Enh2; VT050115) was uniformly active in all MPs and also showed ectopic expression in some non-Cut expressing cells (Figure 2C). Finally, Enhancer 3 (Enh3; GMR35H09) conferred expression in several MPs with highest levels in a subset located in the posterior (Figure 2D). Enh3 encompasses a region that was previously shown to be bound by Su(H) in muscle progenitor related cells, hence may be regulated by Notch activity (Figure 2—figure supplement 1A; Bernard et al., 2010; Krejcí et al., 2009). These results demonstrate that several enhancers contribute to zfh1 expression in the MPs.

Figure 2 with 1 supplement see all
Download asset Open asset
Regulation of zfh1 in MPs and adult pMPs.

(A) Schematic view of zfh1 genomic region, zfh1 regulatory enhancers are represented by green rectangles and arrows indicate transcription start-sites. Coding exons and untranslated regions are represented in black and grey boxes, respectively. (B-D) Three different zfh1 enhancers are active in the MPs (labelled with Cut, purple). Enh1 (VT050105, B) drives GFP (Green) in a subset of scattered MPs; Enh2 (VT050115, C) drives GFP throughout the MPs, and in some non-MP cells (asterisk); Enh3 (GMR35H09, D) is highly expressed in a subset of MPs located in the posterior region of the notum. Scale bars: 50 μM. (n = 30 wing discs). (E-E’’) Enh3-GFP (Green) expression is maintained in adult pMPs (characterised by low levels of Mef2, red; arrows E’-E’’) but not in differentiated muscle nuclei (high Mef2, red; arrowheads G’-G’’). Phalloidin marks muscles (Cyan) and DAPI labels all nuclei (Blue). Insets: boxed regions magnified 12.5 X. Scale bars 25 μM. (n = 10 heminota; from two biological replicates). (F-G) Muscle (IFM) preparation isolated from Enh3-Gal4 > UAS GFP pupae at 18–22 hr APF (F) or at 30–36 hr APF (G). Enh3-GFP (Green) and Zfh1 (Red) are detected in MPs. (F) At 18–22 hr Enh3-GFP activity is higher (arrowheads in E’) in some undifferentiated MPs located between muscle templates (muscles are labeled with Phalloidin, Blue, asterisks). (G) At 30–36 hr APF, Enh3-GFP (Green) activity and Zfh1 (Red) are detected in pMPs (arrowheads) and not in differentiated muscle nuclei. A few Zfh1 +ve cells do not express Enh3-GFP (Arrows G’’). Note: anti-P1, an immune cell marker, was included in the staining to exclude plasmatocytes from the analysis (see Figure 1—figure supplement 2).

https://doi.org/10.7554/eLife.35954.005

To determine which enhancer(s) are also capable of conferring zfh-1 expression in adult pMPs we assessed their activity in adult muscle preparations. Only Enh3 exhibited any activity in these cells (Figure 2E), where it recapitulated well Zfh1 protein expression (Figure 2E–E’’). Thus, Enh3-GFP was clearly detectable in scattered cells, which were closely apposed to the muscle fibers and contained low levels of Mef-2 (Figure 2E’’), and was not expressed in differentiated muscle nuclei (Figure 2).

During pupal stages MPs migrate and surround a set of persistent larval muscles that act as scaffolds for the developing IFMs (Roy and VijayRaghavan, 1998; Fernandes et al., 1991). By 18–22 hr after puparium formation (APF), fusion of myoblasts is ongoing and by 30–36 hr APF, most myoblasts have fused and myogenesis is advanced. By this stage, Zfh1 expression is already restricted to single cells (Chaturvedi et al., 2017). We therefore examined Enh3 activity during the pupal period, using Enh3-Gal4 > UAS GFP, which yields a higher level of expression than the direct Enh3-GFP fusion. At 18–22 hr APF, Enh3 activity was detected in both differentiating myoblasts (inside muscle templates) and undifferentiated MPs (between and around muscle templates) (Figure 2F’–F’’). Importantly, a subset of MPs located between muscle templates exhibited higher levels of Enh3 expression and of Zfh1 levels (Figure 2F’), whereas lower levels were present in the differentiating myoblasts. By 30–36 hr APF Enh3 expression was restricted to pMPs, which expressed high level of Zfh1 and were closely apposed to the muscle fibers (Figure 2G–G’’). At the same stage, we consistently detected a small number of Zfh1 +ve cells that lack detectable Enh3 expression and we speculate that these are undergoing differentiation, since myogenesis is still ongoing at this stage (Figure 2G). In general however, there is a strong correlation between Enh3 expression and the establishment of the adult Zfh1 +ve pMPs, suggesting that Enh3 is responsible for maintaining zfh1 transcription in this progenitor population during the transitionary phase between 20 hr and 30 hr APF.

If Enh3 is indeed responsible for expression of zfh1 in MPs and pMPs, its removal should curtail zfh1 expression in those cells. To test this, Enh3 was deleted by Crispr/Cas9 genome editing (∆Enh3; see Materials and methods). ∆Enh3 homozygous flies survived until early pupal stages allowing us to analyze the phenotype at larval stages. As predicted, ∆Enh3 MPs exhibited greatly reduced Zfh1 protein expression (Figure 2—figure supplement 1B–D) that correlated with decreased zfh1 mRNA levels (Figure 2—figure supplement 1E). Although striking, the effects of ∆Enh3 did not phenocopy those of depleting zfh1 using RNAi, as no premature up-regulation of muscle differentiation markers (MHC, Tm) occurred in ∆Enh3 discs (data not shown). This is likely due to residual zfh1 mRNA/protein (Figure 2—figure supplement 1), brought about by the activity of other zfh1 enhancers (e.g. Enh1 and Enh2, Figure 2B–C). Nevertheless, it is evident that Enh3 has a key role in directing zfh1 expression in MPs/pMPs.

Adult Zfh1 +ve MP cells contribute to flight muscles

By recapitulating Zfh1 in adult MPs, Enh3 provides a powerful tool to investigate whether the persistent MPs are analogous to muscle satellite cells, which are able to divide and produce committed post-mitotic myogenic cells that participate in muscle growth and regeneration. To address this we used a genetic G-trace method, which involves two UAS reporters, an RFP reporter that directly monitors the current activity of the Gal4 and a GFP reporter that records the history of its expression to reveal the lineage (Evans et al., 2009). When Enh3-Gal4 was combined with the G-trace cassette RFP expression was present in the muscle-associated pMPs, which have low Mef2 expression (Figure 3A–A’’). Strikingly, most of the muscle nuclei expressed GFP (Figure 3A’) suggesting that they are derived from ancestral Enh3 expressing cells. Furthermore, close examination of the Enh3 driven RFP expression showed that it often persisted in two nearby muscle nuclei (Figure 3A–A’). This suggests that these nuclei are recent progeny of Enh3-expressing cells, indicating that these cells have retained the ability to divide, a characteristic of satellite cell populations (Figure 3). To further substantiate this conclusion, we verified that adult Zfh1 +ve cells were actively dividing cells, using the mitotic marker phosphohistone-3 (pH3) staining. Many Zfh1 +ve cells co-stained with pH3 indicating that these adult cells remain mitotically active (Figure 3B–B’’).

Figure 3
Download asset Open asset
Adult pMPs contribute to muscle homeostasis.

(A-A’’) Lineage tracing shows that adult pMPs contribute to muscles. Cross-section of indirect flight muscles from adult flies where Enh3-Gal4 drives expression of the G-Trace cassette; GFP (Green) indicates myoblasts that have expressed Enh3-Gal4, RFP (Red) indicates myoblasts where Gal4 is still active, Mef2 labels muscle nuclei (Blue). Note that the RFP (Red, detected with anti-RFP in A’) persists in the recently born myoblasts (Mef2), which are closely localized to the pMPs. Insets: boxed regions magnified 20 X (n = 15 heminota; from three biological replicates). (B-B’’) pMPs are mitoticaly active, indicated by anti-phosphH3 (White in B’’). (PH3 detected in 47% of pMPs, n = 80; from three biological replicates). (C-C’’) Cross-section of indirect flight muscles from adult flies (Enh3-Gal4; tubGal80ts > G Trace) where Enh3-Gal4 directed G-Trace activity was induced for 10 days after animal hatching. GFP (Green, Arrowheads, C’) indicates descendants of pMPs (Blue, Arrows); Mef2 labels muscle nuclei (Red, White). (D) Proportion of newly born myoblasts (marked by GFP; e.g. C’) relative to total number of myoblasts (marked by Mef2; e.g. C’’) in muscle preparations. (n = 18 heminota; light and dark shading indicates data points collected from two independent replicates replicates). Scale bar: 60 μM. (E-F) Prolonged zfh1 depletion in pMPs (10 days after adult hatching) leads to a ‘held out’ wings posture; dorsal view of (E) control (Enh3-Gal4; tubGal80ts > UAS wRNAi;) and (F) zfh1 depletion (Enh3-Gal4; tubGal80ts > UAS-zfh1RNAi (KK 103205)) adult flies. (G-H) Transverse sections of DLM4 muscle stained with Phalloidin (Green) and Mef2 (Red, White) from the indicated genotypes. Fewer Mef2 +ve nuclei are present in muscles when zfh1 is depleted. Scale bars: 50 μM. (I) Similar reductions in muscle nuclei occur following zfh1 depletion (Enh3-Gal4; tubGal80ts > UAS-zfh1RNAi) or following genetic ablation of pMPs, via expression of the pro-apoptotic gene reaper (rpr; Enh3-Gal4;tubGal80ts > UAS rpr). The number of nuclei per section in the indicated conditions was significantly different, light and dark shading indicates data points collected from two independent replicates (>zfh1 RNAi ***p=0.0013, n = 16;>Rpr ***p<0.0001, n = 12).

https://doi.org/10.7554/eLife.35954.007

If the mitotically active Zfh1 +ve cells are indeed important for muscle homoestasis, their progeny should become incorporated into the muscle fibres. We therefore quantified the proportion of muscle nuclei derived from the pMPs during the first 10 days of the adult life, by using a temperature sensitive Gal80 (tubGal80ts) to restrict Enh3-Gal4 until eclosion and combining it with the G-Trace cassette to mark the progeny (Figure 3C–D). Strikingly, the conditional activation of Enh3-Gal4 in adults resulted in GFP labeling of ~24% of muscle nuclei (Figure 3D) indicating a significant role of the pMPs in contributing to muscle maintenance. Indeed when we used a similar regime to deplete zfh1 in pMPs and examined flies at ten days (Figure 3E–I) we found that ~ 30% of adult flies had a ‘held out wing’ posture (n = 93) (Figure 3E–F), a phenotype often associated with flight muscle defects (Vigoreaux, 2001). The number of nuclei per muscle (DLM4) was also significantly reduced (~20% fewer nuclei) in the aged adults when zfh1 was specifically depleted in the pMPs (Figure 3G–H). Likewise, genetic ablation of pMPs (by expressing the pro-apoptotic gene reaper) led to a similar reduction in muscle nuclei (Figure 3I). No ‘held out wing’ phenotype or muscle defects were observed in adult flies within 24 hr of zfh1 knock-down, indicating that the phenotypes at 10 days are due to a defect in the homeostasis of the adult flight muscles. Taken together the results argue that the adult Zfh1 +ve myoblasts cells resemble mammalian satellite cells, retaining the capacity to divide and provide progeny that maintain the adult flight muscles.

Notch directly regulates zfh1 expression in muscle progenitors and adults pMPs

As mentioned above, zfh1 is regulated by Notch activity in Drosophila DmD8, MP-related, cells (Krejcí et al., 2009), where Enh3 was bound by Su(H) (Figure 2A and Figure 2—figure supplement 1). Furthermore, phenotypes from depletion of zfh1 in MPs, were reminiscent of those elicited by loss of Notch (N) signaling (Figure 1 and Krejcí et al., 2009). Notch activity is therefore a candidate to maintain Zfh1 expression in the adult pMPs, thereby preventing their premature differentiation. As a first step to test whether Notch activity contributes to zfh1 expression, we depleted Notch in muscle progenitors by driving Notch RNAi expression with 1151-Gal4 (Figure 4A–C). Under these conditions, Zfh1 levels were significantly reduced, consistent with Notch being required for zfh1 expression in MPs. Second, the consequences of perturbing Notch regulation by mutating the Su(H) binding motifs in Enh3 were analyzed. Two potential Su(H) binding sites are present in Enh3 and both are highly conserved across species (Figure 2A). Mutation of both motifs, Enh3[mut], resulted in a dramatic decrease of the enhancer activity in the MPs (Figure 4D–F). This supports the hypothesis that Notch directly controls zfh1 expression in MPs by regulating activity of Enh3.

Figure 4
Download asset Open asset
Notch directs Zfh1 expression in MPs and pMPs.

(A-C) Zfh1 level (White) is significantly reduced when Notch is down regulated. Expression of Zfh1 in MPs (A) is severely reduced in the presence of Notch RNAi (B, 1151-Gal4 > UAS NotchRNAi), Scale Bars: 50 μM. (C) Quantification of Zfh1 expression levels (*p<0.05, n = 12 wing discs in each condition, light and dark shading indicates data obtained from two independent replicates). (D-F) Enh3 (D, Enh3-GFP, Green) expression in MPs (Purple, Zfh1) is abolished when Su(H) motifs are mutated (E, Enh3[mut]-GFP). Scale bars: 50 μM. (F) Quantification of expression from Enh3 and Enh3[mut] (*p=0.022, n = 14 wing discs in each condition, light and dark shading indicates data obtained from two independent replicates). (G-H) Enh3 (G, Enh3-GFP, Green) expression in adult pMPs (red, Zfh1) is abolished when Su(H) motifs are mutated (H, Enh3[mut]-GFP, Green), DAPI (Blue) reveals all nuclei. (I) Notch[NRE]-GFP (Green) is co-expressed with Zfh1 (Red) in the pMPs associated with the indirect flight muscles; DAPI (Blue) detects all nuclei. (n = 12 heminota; from two independent replicates). In G-I anti-P1 was included to label immune cells and exclude them from the analysis. Scale bars: 25 μM.

https://doi.org/10.7554/eLife.35954.008

Since Enh3 activity persists in the adult pMPs (Figure 2), we next analyzed whether mutating the Su(H) motifs impacted expression in these adult pMPs. Similar to larval stage MPs, Enh3[mut] had lost the ability to direct expression of GFP in the adult pMPs (Figure 4G–H). Thus, the Su(H) motifs are essential for Enh3 to be active in the adult pMPs. These data support the model that persistence of Zfh1 expression in adult MPs is likely due to Notch input, acting through Enh3.

The results imply that Notch should be expressed and active in the adult pMPs. To investigate this, we made use of a Notch[NRE]-GFP reporter line. Notch[NRE] is an enhancer from the Notch gene, and itself regulated by Notch activity, such that it is a read out both of Notch expression and of Notch activity (Simón et al., 2014). Robust expression of Notch[NRE]-GFP reporter was detected in Zfh1 +ve adult pMPs, confirming that Notch is active in these cells (Figure 4I) but not in the differentiated muscles. Together, the results show that zfh1 expression in the adult pMPs requires Notch activity acting through Enh3.

zfh1 is silenced by the conserved microRNA miR-8/miR-200 in MPs

Although transcriptional control of zfh1 by Notch explains one aspect of its regulation, since all larval MPs express Zfh1 it remained unclear how a subset maintain this expression and escape from differentiation to give rise to adult pMPs. A candidate to confer an additional level of regulation on zfh1 expression is the micro RNA miR-8/miR-200, which is important for silencing zfh1 (and its mammalian homologue ZEB1) in several contexts. The regulatory loop between miR-8/miR-200 and zfh1/ZEB has been extensively studied in both Drosophila and vertebrates and is mediated by a miR-8/miR-200 seed site located in the 3’ untranslated region (3’UTR) (Antonello et al., 2015; Vallejo et al., 2011; Brabletz and Brabletz, 2010). Moreover, miR-8 was previously reported to be involved in flight muscle development (Fulga et al., 2015).

To determine whether miR-8 could down-regulate zfh1 in muscle progenitors to promote their differentiation into muscles, we first examined the spatiotemporal expression pattern of miR-8 at larval, pupal and adult stages (Figure 5) using miR-8-Gal4, whose activity reflects the expression of the endogenous miR-8 promoter (Karres et al., 2007). Expression of miR-8-Gal4 was almost undetectable in the larval MPs, although it was highly expressed in the wing pouch (Figure 5). A low level of miR-8-Gal4 expression was also detected in a subset of the MPs where Zfh1 levels are slightly reduced (high Ct expressing DFM precursors; Figure 5). Thus, expression of miR-8 is inversely correlated with Zfh1; its overall expression is low in larval MPs where Zfh1 expression is important to prevent their differentiation (Figure 5B–B and Figure 1A).

Figure 5
Download asset Open asset
Expression dynamics of miR-8 and Zfh1 in MPs and pMPs during indirect flight muscle development.

(A-C) miR-8 (Green, miR-8-Gal4 > UAS-mCD8::GFP) is not highly expressed in MPs (Cut, Purple) but is prevalent in the wing disc pouch (Arrow), notum (Arrowhead) and air sac (Asterisk). Higher magnification shows that low level of miR-8 expression can be detected in the subset of MPs where Zfh1 is normally low (Arrowhead in C) but not in other MPs. Scale bars: 50 μM. (n > 20 wing discs; from three biological replicates). (D-D’’) IFM preparation isolated from miR-8-Gal4 > UAS-nlsGFP pupae at 18–22 hr APF. At this stage miR-8 (Green; D and D’) is co-expressed with Zfh1 (Red; D and D’’) in MPs but miR-8 is more highly expressed in differentiated myoblasts (Asterisks in D’), found within the muscles (labeled with Phalloidin, Blue), compared to the undifferentiated MPs, found between the muscles (Arrowheads in D’). In contrast, Zfh1 (D, D’) is detected at lower levels in the differentiated myoblasts compared to the undifferentiated MPs (Asterisks in D’). Scale bar: 25 μM. (E-E’’) IFM preparation isolated from miR-8-Gal4 > UAS-nlsGFP pupae at 30–36 hr APF. At this stage, mir-8 is highly detected in the differentiated muscle nuclei. The majority of the Zfh1 +ve pMPs (Red) do not express mir-8 (Green) (Arrowheads). Few Zfh1 +ve pMPs express mir-8 (Arrows). Scale bar: 50 μM. (F-F’’) In adult IFMs, mir-8 (Green) expression is absent from Zfh1 +ve (red) pMPs (Arrows, D-D’) but is present at uniformly high levels in IFMs, (Phalloidin, Blue). Scale bars: 50 μM. (n = 20 heminota; from three biological replicates).

https://doi.org/10.7554/eLife.35954.009

We subsequently compared miR8-Gal4 and Zfh1 expression in 18–22 hr APF pupae (Figure 5D–D’’). At this stage, miR-8-Gal4 and Zfh1 expression overlapped in most, if not all, myogenic nuclei (Figure 5D). However, miR-8-Gal4 expression level was elevated in the differentiated myoblasts, which are located inside the muscle templates (Figure 5D–D’). Conversely, Zfh1 expression level was slightly lower in this population and higher in the undifferentiated MPs (Figure 5D–D’’). Thus miR-8 and Zfh1 have reciprocal low and high expression patterns in the MPs at this stage of myogenesis.

By 30–36 hr APF, Zfh1 expression was restricted to pMPs while miR-8-Gal4 was predominantly expressed throughout the differentiated myoblasts (Figure 5E). Notably, a few of the Zfh1 +ve cells at this stage retained miR-8-Gal4 expression (Figure 5E–E’). Similar to 30–36 hr APF, adult muscles had uniform and high levels of miR-8-Gal4 (Figure 5F–F’’). However, at this time, miR-8-Gal4 expression was totally absent from all Zfh1 +ve adult pMPs (Figure 5F–F’’). These data show that, during muscle formation, miR-8 expression level is inversely correlated to Zfh1; supporting the model that miR-8 negatively regulates zfh1.

Given their complementary expression patterns we next tested the impact of miR-8 overexpression on Zfh1 protein levels in larval MPs (Figure 6A–B). Zfh1 protein levels were significantly diminished under these conditions, in agreement with miR-8 regulating zfh1 post-transcriptionally (Figure 6C). Although this manipulation was not sufficient to cause premature up-regulation of muscle differentiation markers or associated morphological changes in MPs, the few surviving adults all displayed a held-out wing phenotype, which is often associated with defective flight muscles (Vigoreaux, 2001). Next we assessed whether miR-8 activity/expression changes in response to Mef2 levels, a critical determinant of muscle differentiation, using a miR-8 sensor (containing two miR-8 binding sites in its 3’UTR [Kennell et al., 2012]). Expression of the miR-8 sensor was specifically decreased when Mef2 was overexpressed in MPs, suggesting that miR-8 expression responds to high level of Mef2 (Figure 6—figure supplement 1).

Figure 6 with 1 supplement see all
Download asset Open asset
The conserved microRNA miR-8/miR-200 antagonizes zfh1 to promote muscle differentiation.

(A-C) Effect of miR-8 overexpression (1151-Gal4 > UAS-miR-8) on Zfh1 (White) protein level in MPs. Scale Bars: 50 μM. (C) Zfh1 expression is significantly reduced by miR-8 over-expression. (***p=0.0009, n = 12 wing discs in each condition, light and dark shading indicates data points from two independent replicates). (D-E) Sagittal sections of adult IFMs stained for Phalloidin (Blue), Zfh1 (Green) and P1 (Red). Down regulating miR-8 during muscle differentiation (Mef2-Gal4 > UAS-miR-8-Sp) increases the final number of adult pMPs. (F) The number of pMPs in adult IFMs in the indicated conditions was significantly different. (****p<0.0001, n = 18 adults for each genotype; light, dark and intermediate shading indicates data points from three independent replicates). Scale bars: 100 μM.

https://doi.org/10.7554/eLife.35954.010

If down-regulation of zfh1 by miR-8 is important to allow muscle differentiation, selective depletion of miR-8 should allow more MPs to escape differentiation. To achieve this, a miR-8 sponge construct (UAS-miR-8Sp; Fulga et al., 2015) was expressed using Mef2-Gal4, so that it would decrease miR-8 activity in differentiating myoblasts. Adult muscles were still formed under these conditions. However the final number of pMPs was significantly increased (Figure 6D–F). Conversely ectopic expression of miR-8 in adult MPs (using Enh3-Gal4) led to a reduction in the number of muscle nuclei similar to that seen with zfh1 down-regulation (Figure 6—figure supplement 1D). Together, these data argue that miR-8 up-regulation during muscle differentiation blocks Zfh1 production to allow MPs to differentiate and if its expression is compromised, more MPs are permitted to escape differentiation. It is also possible that miR-8 has additional targets, besides Zfh1, that are involved in maintenance/differentiation of MPs.

An alternate short zfh1 isoform is transcribed in adult pMPs

To retain their undifferentiated state, the adult pMPs must evade miR-8 regulation and maintain Zfh1 expression. The zfh1 gene gives rise to three different mRNA isoforms; two long zfh1 isoforms (zfh1-long; zfh1-RE/RB) and one short zfh1 isoform (zfh1-short; zfh1-RA) (Figure 7A). Although zfh1-long isoforms have two additional N-terminal zinc fingers, all three RNA-isoforms produce proteins containing the core zinc finger and homeodomains needed for Zfh1 DNA-binding activity (Figure 7—figure supplement 1A; Postigo et al., 1999). Importantly, zfh1-short isoform has a shorter 3’UTR, which lacks the target site for miR-8 (Figure 7A; Antonello et al., 2015), as well as differing in its transcription start site (TSS) (Figure 7A; Flybase FBgn0004606). The lack of a seed site makes zfh1-short insensitive to miR-8 mediated down-regulation. This means that zfh1-short expression would enable cells to retain high level of Zfh1 protein, even in the context of miR-8 expression and, if present in a subset of MPs, could explaining how they can escape differentiation.

Figure 7 with 1 supplement see all
Download asset Open asset
Transcriptional dynamics of zfh1 isoforms in MPs and pMPs.

(A) Schematic representation of zfh1 isoforms. zfh1-short (zfh1-RA) is initiated from a different transcription start site and has shorter 3’UTR that lacks the target site for miR-8 (Green; Antonello et al., 2015) present in zfh1-long isoforms (zfh1-RB, zfh1-RE); the position of the miR-8 seed sites in zfh1-long 3’ UTR are depicted. Non-coding exons and coding exons are depicted by grey and black boxes respectively, red lines indicate the probes used for FISH experiments in B-E. (B-C) zfh1-long is present uniformly in MPs. (n > 10 wing discs from two replicates; B, Green and B’, White) whereas zfh1-short is only detected in a few MPs (n > 15 wing discs from three replicates; C, Green and C’, White), detected by in situ hybridisation in wild type third instar wing discs stained for Zfh1 (Purple). Scale bars: 10 μM. (B’ B’’, C’ C’’) Higher magnifications of boxed regions (Scale bars: 50 μM). (D-E) In adult IFMs zfh1-long is detected in the pMPs (n = 11 pMPs from two replicates, arrow in D’) and in some differentiated nuclei located in their vicinity (arrowheads in D’) whereas zfh1-short is only present in pMPs (n = 15 pMPs from two replicates, arrow in E’). Enh3 expression (Green, Enh3-Gal4 > UAS-mCD8GFP) labels adult pMPs and Mef2 (Blue) labels all muscle nuclei. Scale bars: 20 μM.

https://doi.org/10.7554/eLife.35954.012

To determine whether zfh1-short is indeed expressed in MPs, we designed fluorescent probes specific for the zfh1-long and zfh1-short isoforms and used them for in situ hybridization (FISH) at larval and adult stages (Figure 7A). In larval stages (L3), zfh1-long isoforms were present at uniformly high levels in the MPs (Figure 7B–B’’’) whereas zfh1-short was expressed at much lower levels and only detected in a few MPs in each disc (Figure 7C–C’’’). In adult muscles, where pMPs were marked by Enh3-Gal4 > GFP and low Mef2 expression (Figure 7D–E), high levels of zfh1-short and much lower levels of zfh1-long (Figure 7D–E’’’) were present in the pMPs. Indeed, zfh1-short was only present in the pMPs whereas dots of zfh1-long hybridization were also detected in some differentiated nuclei (with high level of Mef2) (Figure 7). Thus, zfh1-short is expressed in a few larval MPs and is then detected at highest levels specifically in the adult pMPs but is not transcribed in adult muscle nuclei. Since zfh1-short is not susceptible to regulation by miR-8, its specific transcription may therefore be determinant for maintaining high levels of Zfh1 in a subset of progenitors and enable them to escape differentiation.

The model predicts that Zfh1-short will counteract the myogenic program in a similar manner as previously described for Zfh1-long, which antagonizes Mef2 function (Siles et al., 2013). Premature expression of Mef2 (using 1151 Gal4) induces precocious differentiation of larval MPs, evident by ectopic expression of MHC-LacZ and a reduction of MPs (Figure 7—figure supplement 1 and ref). Expression of Zfh1-short was able to counteract the effects of Mef2, suppressing the precocious muscle differentiation phenotype and restoring the normal morphology of MPs. Thus, Zfh1-short retains the capacity to block Mef2 induced muscle differentiation (Figure 7—figure supplement 1).

zfh1-short isoform transcription requires Notch activity in adult pMPs

Expression of zfh1-short (zfh1-RA) is specifically retained in adult MPs (Figure 7E) where it may be critical for maintaining their progenitor status. Mechanisms that ensure this isoform is appropriately transcribed could involve Notch signaling, which is necessary for normal levels of Zfh1 expression in MPs (Figure 4). If this is the case, expression of a constitutively active Notch (Notch∆ECD) should up-regulate zfh1-short transcripts in the MPs at larval stages when their expression is normally low. In agreement, expression of active Notch in the progenitors (1151-Gal4 > Notch∆ECD) significantly increased the proportions of cells transcribing zfh1-short (Figure 8A–B’ and C).

Figure 8 with 1 supplement see all
Download asset Open asset
zfh1-short regulation by Notch is important to maintain muscle homeostasis.

(A-C) Expression of an activated Notch (1151-Gal4 > UAS-NΔECD) in MPs induces ectopic zfh1-short transcription. In situ hybridisation detecting zfh1-short (Green) in MPs (Zfh1, Purple) with wild type (A-A’) or elevated Notch activity (B-B’). Scale bars: 50 μM. (C) Quantification showing significant increase in zfh1-short transcriptional dots upon Notch up regulation, relative to total number of MPs (**p<0.01, Student t-test; n = 14 wing discs for each genotype, light and dark shading indicates data points from two independent replicates). (D-H) Notch depletion leads to a severe decrease in zfh1-short (Red) (E’-F’ and D) in pMPs (Green; Enh3-Gal4; UAS-mCD8GFP > UAS Notch-RNAi; tubGal80ts) and to an increase in Mef2 levels (Blue) (E-F’ and H). (G-G’ and H) zfh1 depletion in the pMPs (Enh3-Gal4; UAS-mCD8GFP > UASzfh1 RNAi; tubGal80ts) leads to an increase in Mef2 levels (Blue). Scale bars: 25 μM. Quantifications of zfh1-short (D) and Mef2 (H) in the indicated conditions show that the levels are significantly different (D, n = 21 pMPs for each genotype (**p<0.01); H, n = 15 pMPs for Control RNAi and n = 13 pMPs for Notch RNAi (*p<0.05) and n = 14 pMPs for zfh1 RNAi (**p<0.01)). In each condition light, dark and intermediate shading indicates data points from three independent replicates. (I). Prolonged Notch depletion in the pMPs (Enh3-Gal4; Gal80ts > UAS Notch RNAi) affects the muscle homeostasis. (**p<0.01, n = 14 for each genotype, light and dark shading indicates data points from two independent replicates).

https://doi.org/10.7554/eLife.35954.014

To address whether Notch is necessary for zfh1-short transcription in adult pMPs, we specifically depleted Notch levels after eclosion (using Enh3-Gal4 in combination with tubGal80ts to drive Notch RNAi; Figure 8E–F). Consistent with expression of zfh1-short being dependent on Notch activity, the levels of zfh1-short were significantly reduced in adult pMPs when Notch was down-regulated in this way (Figure 8E–F and D). Conversely, expression of zfh1-long isoform was less affected (Figure 8—figure supplement 1), suggesting other inputs besides Notch help to sustain zfh1-long in pMPs. Nevertheless, Mef2 accumulated to higher levels than normal in the adult pMPs, under these Notch RNAi conditions, indicating that Notch activity helps prevent their differentiation (Figure 8E’, F’ and H), most likely through sustaining a higher level of Zfh1 expression via its regulation of zfh1-short.

The increased level of Mef2 in the pMPs following Notch-depletion suggests they are losing their progenitor status and becoming differentiated. This forced differentiation of the pMPs would deplete the progenitor population and so should compromise muscle maintenance and repair. In agreement there was a significant reduction of the number of nuclei per muscle after ten days of Notch down-regulation in the pMPs (Figure 8I). These results are reminiscent of the targeted zfh1 down-regulation in the adult pMPs (Figure 3G–H and I), where high level of Mef2 was also prematurely detected (Figure 8G–G’ and H). Taken together, the data indicate that persistent Notch activity is required to maintain zfh1-short expression in pMPs, which protects the pMPs from differentiating by ensuring that sufficient Zfh1 protein is present to prevent Mef2 accumulating.

Discussion

A key property of adult stem cells is their ability to remain in a quiescent state for a prolonged period of time (Li and Clevers, 2010). Investigating the maintenance of Drosophila MPs we have uncovered an important new regulatory logic, in which a switch in RNA isoforms enables a sub-population of cells to escape miRNA regulation and so avoid the differentiation program. At the same time, analyzing expression of a pivotal player in this regulatory loop, zfh1, has revealed how this mechanism sustains a population of persistent progenitors associated with adult muscles in Drosophila, that appear analogous to mammalian satellite cells (Figure 9).

Figure 9
Download asset Open asset
Model summarizing the role of alternate zfh1 isoforms in the maintenance of adult pMPs.

zfh1-long (Grey) is expressed in all MPs at larval stage. Silencing of zfh1-long by miR-8 (Red) facilitates the MPs differentiation. zfh1-short (Green) transcription is driven and maintained in pMPs by a Notch responsive element (Enh3, Green rectangle), which may also contribute to zfh1-long regulation. Because zfh1-short is insensitive to miR-8, Zfh1 protein is maintained in pMPs, enabling them to escape differentiation and persist as MPs in the adult.

https://doi.org/10.7554/eLife.35954.016

Zfh1 maintains a population of satellite-like cells in adult drosophila

Until recently, the fly has been thought to lack a persistent muscle stem cell population, leading to speculation about how its muscles could withstand the wear and tear of its active lifestyle. Now it emerges that expression of Zfh-1 identifies a population of muscle-associated cells in the adult that retain progenitor-like properties (Figure 9 and Chaturvedi et al., 2017). Indeed we have found that Zfh-1 is critical to prevent these progenitor cells from differentiating. Its expression in the persistent adult ‘satellite-like’ cells is dependent on a specific Zfh1 enhancer, which is directly regulated by Notch. Activity of Notch is important for maintaining Zfh1 expression and hence is required to sustain the progenitor status of these cells, similar to the situation in mammalian satellite cells, which require Notch activity for their maintenance (Mourikis and Tajbakhsh, 2014; Bjornson et al., 2012).

Using lineage-tracing method we showed that adult Zfh1 +ve cells, in normal conditions, provide new myoblasts to the fibers. Furthermore, conditional down regulation of zfh1 led adult pMPs to enter differentiation and resulted in flight defects, evident by a held out wing posture. These results demonstrate that Zfh-1 is necessary to maintain these progenitors and that, similar to vertebrate satellite cells, the Zfh1 +ve progenitor cells contribute to the adult muscles homeostasis. Others have recently shown that the pMPs are expanded in conditions of muscle injury where they are likely to contribute to repair (Chaturvedi et al., 2017). Thus the retention of a pool of progenitor cells may be critical to maintain the physiological function of all muscles in all organism types, as also highlighted by their identification in another arthropod Parahyle (Alwes et al., 2016; Konstantinides and Averof, 2014). Drosophila notably differs because their satellite-like cells do not express the Pax3/Pax7 homologue (gooseberry; data not shown), considered a canonical marker in mammals and some other organisms (Chang and Rudnicki, 2014). Nor do they express the promyogenic bHLH protein Twist (data not shown), which is present in the muscle progenitors in the embryo (Bate et al., 1991). Instead, Zfh1 appears to fulfill an analogous function and it will be interesting to discover how widespread this alternate Zfh1 pathway is for precursor maintenance. Notably, the loss of ZEB1 in mice accelerates the temporal expression of muscle differentiation genes (e.g. MHC) suggesting that there is indeed an evolutionary conserved function of Zfh1/ZEB in regulating the muscle differentiation process (Siles et al., 2013). This lends further credence to the model that Zfh1 could have a fundamental role in preventing differentiation that may be harnessed in multiple contexts.

Switching 3' UTR to protect progenitors from differentiation

Another key feature of zfh-1 regulation that is conserved between mammals and flies is its sensitivity to the miR-200/miR-8 family of miRNAs (Antonello et al., 2015; Brabletz and Brabletz, 2010). This has major significance in many cancers, where loss of miR-200 results in elevated levels of ZEB1 promoting the expansion of cancer stem cells, and has led to a widely accepted model in which the downregulation of Zfh1 family is necessary to curb stem-ness (Brabletz and Brabletz, 2010). This fits with our observations, as we find that miR-8 is upregulated during differentiation of the MPs and suppresses Zfh1 protein expression. Critically however, only some RNA isoforms, zfh1-long, contain seed sites necessary for miR-8 regulation (Antonello et al., 2015). The alternate, zfh1-short, isoform has a truncated 3’UTR that lacks the miR-8 recognition sequences and will thus be insensitive to miR-8 regulation. Significantly, this zfh1-short isoform is specifically expressed in MPs that persist into adulthood and hence can help protect them from miR-8 induced differentiation during the pupal phases when both are co-expressed. However, the pMPs remain sensitive to forced miR-8 expression in the adult, suggesting the levels of Zfh1 are finely tuned by the expression of both zfh1-long and zfh1-short. This could be important to enable the differentiation of the MP progeny. Furthermore, the fact that Notch activity strongly promotes zfh1-short expression could explain how an elevated level of Notch signaling promotes expansion of pMPs following injury, as observed by others (Chaturvedi et al., 2017).

Together the data suggest a novel molecular logic to explain the maintenance of Drosophila satellite-like cells. This relies on the expression of zfh1-short, which, by being insensitive to miR-8 regulation, can sustain Zfh1 protein production to protect pMPs from differentiation (Figure 9). It also implies that Notch preferentially promotes the expression of a specific RNA isoform, most likely through the use of an alternate promoter in zfh1. Both of these concepts have widespread implications.

Alternate use of 3’UTRs, to escape miRNA regulation, is potentially an important mechanism to tune developmental decisions. Some tissues have a global tendency to favor certain isoform types, for example, distal polyadenylation sites are preferred in neuronal tissues (Zhang et al., 2005). Furthermore, the occurrence of alternate 3’UTR RNA isoforms is widespread (>50% human genes generate alternate 3’UTR isoforms) and many conserved miR target sites are contained in such alternate 3’UTRs (Tian and Manley, 2013; Sandberg et al., 2008). Thus, similar isoform switching may underpin many instances of progenitor regulation and cell fate determination. Indeed an isoform switch appears to underlie variations in Pax3 expression levels between two different populations of muscle satellite cells in mice, where the use of alternative polyadenylation sites resulted in transcripts with shorter 3’UTRs that are resistant to regulation by miR-206 (Boutet et al., 2012). The selection of alternate 3’-UTRs could ensure that protein levels do not fall below a critical level (Yatsenko et al., 2014), and in this way prevent differentiation from being triggered.

The switch in zfh1 RNA isoforms is associated with Notch-dependent maintenance of the persistent adult MPs. Notably, zfh1-short is generated from an alternate promoter, as well as having a truncated 3’UTR, which may be one factor underlying this switch. Studies in yeasts demonstrate that looping occurs between promoters and polyadenylation sites, and that specific factors recruited at promoters can influence poly-A site selection (Lamas-Maceiras et al., 2016; Tian and Manley, 2013). The levels and speed of transcription also appear to influence polyA site selection (Proudfoot, 2016; Tian and Manley, 2013; Pinto et al., 2011). If Notch mediated activation via Enh3 favors initiation at the zfh1-short promoter, this could in turn influence the selection of the proximal adenylation site to generate the truncated miR-8 insensitive UTR. The concept that signaling can differentially regulate RNA sub-types has so far been little explored but our results suggest that is potentially of considerable significance. In future it will be important to investigate the extent that this mechanism is deployed in other contexts where signaling coordinates cell fate choices and stem cell maintenance.

Materials and methods

Key resources table
Reagent type (species)
or resource		DesignationSource or referenceIdentifiersAdditional information
Gene (D. melanogaster)zfh1NAFLYB:FBgn0004606
Gene (D. melanogaster)NotchNAFLYB:FBgn0004647
Gene (D. melanogaster)miR-8NAFLYB:FBgn0262432
Genetic reagent
(D. melanogaster)		Enh3-Gal4Janelia Research CampusBDSC: 49924, FLYB: FBtp0059625FlyBase symbol: P{GMR35H09-GAL4}attP2
Genetic reagent
(D. melanogaster)		zfh1 RNAi (kk 103205)Vienna Drosophila RNAi CenterVDRC: 103205
Genetic reagent
(D. melanogaster)		miR-8-Gal4Kyoto Stock CenterDGRC: 104917Genotype: y[*] w[*]; P{w[+mW.hs]=GawB}NP5247/CyO, P{w[-]=UAS lacZ.UW14}UW14
Genetic reagent
(D. melanogaster)		UAS-miR-8-SpBloomington Stock CenterBDSC: 61374, FLYB: FBst0061374Genotype: P{UAS-mCherry.mir-8.sponge.V2}attP40/CyO; P{UAS-mCherry.mir-8.sponge.V2}attP2
Genetic reagent
(D. melanogaster)		Enh3-GFPThis paper
Genetic reagent
(D. melanogaster)		∆Enh3This paper
Genetic reagent
(D. melanogaster)		UAS-zfh1-shortThis paperUAS-Zfh1-short construct provided by BDGP, Clone # UF5607
Genetic reagent (D. melanogaster)Notch[NRE]-GFPSarah Bray (Cambridge, UK)
Antibodyanti-Zfh1Ruth Lehmann (New York, USA)
Antibodyanti-Mef2Eileen Furlong (Heidelberg, Germany)
Recombinant DNA reagentpCFD4AddgeneAddgene # 49411
Recombinant DNA reagentpDsRed-attPAddgeneAddgene # 51019

Drosophila genetics

Request a detailed protocol

All Drosophila melanogaster stocks were grown on standard medium at 25°C. The following stains were used: w118 as wild type (wt), UAS-white-RNAi as control for RNAi experiments (BL35573), UAS-zfh1-RNAi (VDRC: KK103205, TRiP: BL29347), zfh1 deficiency (BL7917), Mef2-Gal4 (Ranganayakulu et al., 1996), UAS-Mef2 (Cripps et al., 2004), UAS-G-Trace (BL28281), UAS-Notch-RNAi (BL7078), Notch[NRE]-GFP (Simón et al., 2014), UAS-Notch∆ECD (Chanet et al., 2009; Fortini and Artavanis-Tsakonas, 1993; Rebay et al., 1993), miR-8-Gal4 (Karres et al., 2007), UAS-miR-8 (Vallejo et al., 2011), UAS-miR-8-Sp (BL61374 and Fulga et al., 2015), UAS-Scramble-SP (BL61501), UAS-mCD8::GFP (BL5137), UAS-GFPnls (BL65402), UAS-Src::GFP (Kaltschmidt et al., 2000), MHC-lacZ (Hess et al., 2007), 1151-Gal4 (Anant et al., 1998), miR-8-sensor-EGFP (Kennell et al., 2012), CG9650-LacZ (Ahmad et al., 2014), UAS-Reaper (BL5824). Enhancer-Gal4 lines described in Figure 2 and Figure 2—figure supplement 1 are either from Janelia FlyLight (http://flweb.janelia.org) or Vienna Tiles Library (http://stockcenter.vdrc.at/control/main).

RNAi experiments were conducted at 29°C. For adult specific manipulations in pMPs tubGal80ts (McGuire et al., 2003) was used to limit Enh3-Gal4 expression to a defined period of time. Crosses were kept at 18°C and eclosed adults were shifted to 29°C until dissection.

Immunohistochemistry and in situ hybridization

Request a detailed protocol

Immunofluorescence stainings of wing discs were performed using standard techniques. Dissection and staining of the pupal muscles was performed according to (Weitkunat and Schnorrer, 2014). Adult muscles were prepared and stained as described in (Hunt and Demontis, 2013). The following primary antibody were used: Rabbit anti-Zfh1 (1:5000, a gift from Ruth Lehmann, New York, USA), Mouse anti-Cut (1:20, DSHB), Rabbit anti-β3-Tubulin (1:5000, a gift from Renate Renkawitz-Pohl, Marburg, Germany), Rat anti-Tropomyosin (1:1000, Abcam, ab50567), Goat anti-GFP (1:200, Abcam, ab6673), Rabbit anti-Ds-Red (1:25; Clontech, 6324496), Rabbit anti-Mef2 (1:200, a gift from Eileen Furlong, Heidelberg, Germany), Mouse anti-P1 (1:20, a gift from István Andó, Szeged, Hugary), Mouse anti-pH3 (1:100, Cell Signaling Technology, #9706), Mouse anti-β-Gal (1:1000, Promega, Z378A), Alexa-conjugated Phalloidin (1:200, Thermo fisher, Waltham/Massachusetts), Rat anti-Dcad2 (1:200, DSHB). In situ experiments were carried out according to Stellaris-protocols (https://www.biosearchtech.com/assets/bti_custom_stellaris_drosophila_protocol.pdf). Antibodies were included to the overnight hybridization step (together with the probes). zfh1 probes were generated by Bioresearch Technologies. The sequence used for zfh1-short probe span 393 bp of the first zfh1-RA exon, for zfh1-long probe, the sequence of the third exon (711 bp) common to both zfh1-RB and zfh1-RE was used (see Figure 5).

Construction of transgenic lines and mutagenesis

Request a detailed protocol

For Enh3-GFP reporter line, the genomic region chr3R: 30774595..30778415 (Enh3/GMR35H09) according to Flybase genome release r6.03 was amplified using yw genomic DNA as template. Enh3 fragment was then cloned into the pGreenRabbit vector (Housden et al., 2012). For Enh3[mut]-GFP line, two Su(H) biding sites were predicted within Enh3 sequence using Patser (Hertz and Stormo, 1999) and mutated by PCR based approach with primers overlapping the Su(H) sites to be mutated with the following sequence modifications: Su(H)1 AGTGGGAA to AGGTGTGA and Su(H) 2 TTCTCACA to TGTTTGCA. Both constructs were inserted into an AttP located at 68A4 on chromosome III by injection into nos-phiC31-NLS; attP2 embryos (Bischof et al., 2007). The UAS-zfh1-short transgenic line (Figure 7—figure supplement 1) was similarly generated using phiC mediated integration of an AttB plasmid carrying UAS-zfh1-short (BDGP Clone # UF5607) into an AttP site at position 25C7. The transgene produced detectable nuclear Zfh1 protein when crossed to Gal4 driver lines (data not shown). All constructs were fully sequenced and analyzed prior to injection.

CRISPR/Cas9 genome editing

Request a detailed protocol

CRISPR mediated deletion of Enh3 was performed according to (Port et al., 2014). For generating guide RNAs, two protospacers were selected (sgRNA1 GCATTCCGCAGGTTTAGTCAC and sgRNA2 GCGATAACCCGGCGACCTCC) flanking 5’ and 3’ Enhancer-3 regions, (http://www.flyrnai.org/crispr/). The protospacers were cloned into the tandem guide RNA expression vector pCFD4 (Addgene #49411) (http://www.crisprflydesign.org/wp-content/uploads/2014/06/Cloning-with-pCFD4.pdf). For the homology directed repair step, two homology arms were amplified using yw genomic DNA as template with the following primers (Homology arm1: Fwd. 5’ GCGCGAATTCGGGCTAAACGCCAGATAAGCG 3’ Rev. 5’ TTCCGCGGCCGCCACTGGATTCCACGGCTTTTCG 3’– Homology arm 2: Fwd. 5’ GGTAGCTCTTCTTATATAACCCGGCGACCTCCTCG3’- Rev. 5’GGTAGCTCTTCTGACC GGACGAAAAACTAGCGACC) and cloned into the pDsRed-attP (Addgene #51019) vector (http://flycrispr.molbio.wisc.edu/protocols/pHD-DsRed-attP). Both constructs were injected into nos-Cas9 (BL54591) embryos. Flies with ∆Enh3 were identified via the expression of the Ds-Red in the eyes and confirmed with sequencing of PCR fragment spanning the deletion. ∆Enh3 flies were then crossed to strains carrying a deletion (BL7917), which removes zfh1. None of the tranheterozygote animals survived to adults confirming that ∆Enh3 lethality maps to the zfh1 locus.

Microscopy and data analysis

Request a detailed protocol

Samples were imaged on Leica SP2 or TCS SP8 microscopes (CAIC, University of Cambridge) at 20X or 40X magnification and 1024/1024 pixel resolution. Images were processed with Image J and assembled with Adobe Illustrator. Quantification of fluorescence signal intensities was performed with Image J software. In each case the n refers to the number of individual specimens analyzed, which were from two or more independent experiments. For experiments to compare and measure expression levels, samples were prepared and analyzed in parallel, with identical conditions and the same laser parameters used for image acquisition. For each confocal stack a Sum slices projections was generated. Signal intensities were obtained by manually outlining the regions of interest, based on expression of markers, and measuring the average within each region. The values were then normalized to similar background measurements for each sample. In Figure 8 the number of transcriptional zfh1-short dots was counted manually with Image J and normalized to the total number of nuclei (Zfh1 staining), which was determined by a Matlab homemade script. Graphs and statistical analysis were performed with Prism seven using unpaired t-test. Error bars indicate standard error of the mean. Further statistical details of experiments can be found in the figure legends.

Quantitative RT PCR

Request a detailed protocol

30 Wing Imaginal discs from each genotype were dissected and RNA isolated using TRIzol (Life technologies). Quantitative PCR were performed as described (Krejcí and Bray, 2007). Values were normalized to the level of Rpl32. The following primers were used. Rpl32, Fwd 5’-ATGCTAAGCTGTCGCACAAATG-3’ and Rev 5’-GTTCGATCCGTAACCGATGT-3’. zfh1 Fwd 5’- GTTCAAGCACCACCTCAAGGAG-3’ and Rev 5’- CTTCTTGGAGGTCATGTGGGAGG-3’. (Product common to all three zfh1 isoforms).

Data availability

All data generated or analysed during this study are included in the manuscript and supporting files.

References

    1. Anant S
    2. Roy S
    3. VijayRaghavan K
    (1998)
    Twist and Notch negatively regulate adult muscle differentiation in Drosophila
    Development 125:1361–1369.
    1. Bate M
    2. Rushton E
    3. Currie DA
    (1991)
    Cells with persistent twist expression are the embryonic precursors of adult muscles in Drosophila
    Development 113:79–89.
    1. Fernandes J
    2. Bate M
    3. Vijayraghavan K
    (1991)
    Development of the indirect flight muscles of Drosophila
    Development 113:67–77.

Decision letter

  1. Manfred Frasch
    Reviewing Editor; Friedrich-Alexander-Universität Erlangen-Nürnberg, Germany

In the interests of transparency, eLife includes the editorial decision letter and accompanying author responses. A lightly edited version of the letter sent to the authors after peer review is shown, indicating the most substantive concerns; minor comments are not usually included.

[Editors’ note: a previous version of this study was rejected after peer review, but the authors submitted for reconsideration. The first decision letter after peer review is shown below.]

Thank you for submitting your work entitled "A population of adult satellite-like cells in Drosophila is maintained through a switch in RNA-isoforms" for consideration by eLife. Your article has been favorably evaluated by a Senior Editor and three reviewers, one of whom is a member of our Board of Reviewing Editors. The reviewers have opted to remain anonymous.

Our decision has been reached after consultation between the reviewers. Based on these discussions and the individual reviews below, we regret to inform you that your work will not be considered further for publication in eLife.

The reviewers agree that this is a very interesting study proposing a conceptually novel regulatory logic in muscle stem cell development. However, all three reviewers conclude that some of the key pieces of evidence are still too preliminary and do not yet firmly support the conclusions made. As you will see from the attached reviews, the main weaknesses include a) the absence of clonal evidence showing that the satellite-like cells are contributing to muscle homeostasis, b) uncertainties regarding the role of Zfh1 short in these cells and their precursors, c) the lack of data demonstrating the developmental expression dynamics of miR-8, Enh3-GFP, and Zfh1 within the satellite-like cells and their precursors, and d) the insufficient characterization of mir-8 function within these cells that would support the biological relevance of this regulatory logic. As the reviewers expect that the work to address these issues would take significantly longer than the two months requested for a resubmission, we are returning the manuscript to you and hope that the reviews will provide helpful guidance for further strengthening the study.

Reviewer #1:

Boukhatmi and Bray present intriguing data suggesting the existence of muscle stem cells in adult Drosophila that may function like the satellite cells in vertebrates during muscle repair upon injury and aging. The authors propose a major role of the transcription factor Zfh1 in keeping these cells undifferentiated. They extend the regulatory pathway further by findings indicating that a short isoform of zfh1 is being induced by Notch/Su(H) in these satellite-like cells, which in contrast to the long form is not targeted by the microRNA miR-8 that could otherwise prevent expression of Zfh1 in them. Through this mechanism, Notch could stabilize the undifferentiated state of these satellite-like cells.

In principle, these are very important and novel findings that would have broad relevance in the muscle stem cell field. However, as discussed below, there are still some major gaps and inconsistencies in the data and models, which means that the evidence for functional muscle satellite cells in Drosophila in this study is largely circumstantial but without solid proof, and some aspects of the proposed regulatory pathway are questionable. However, if the authors were able to fill in these gaps convincingly the manuscript would be highly appropriate for publication in eLife.

1) Currently, the authors do not have any direct evidence that the observed cells act as muscle stem cells that can fuse with muscles for repairing damage. The Enh3-Gal4>G-TRACE lineage data with GFP (Figure 2) only show that both the muscles and the satellite-like cells are derived from adepithelial cells of the wing discs, as would be expected. The demonstration of proliferation is interesting, but not proof for their potential to achieve muscle repair. The crucial experiment would be to perform this experiment conditionally, such that GFP becomes only expressed in the satellite-like cells and their descendants after hatching, and then look for GFP-positive nuclei within muscles upon aging or injury. I realize there may be some diffusion of GFP in the syncytia but nevertheless, the bulk of GFP should be imported into the nuclei that express GFP RNA.

Related to this point, in Figure 7 the authors show that, upon knockdown of zfh1 after hatching and 8 days of further aging of the flies, their IFMs contain ca. 20% fewer nuclei, which they take to mean that normally that many satellite cells fuse to repair the aging muscles during this time period. This would be a surprisingly large percentage and an important result. If true, the conditional lineage tracing experiment requested above should show ca. 20% GFP-positive nuclei in IFM cross sections of aged flies similar to those shown in Figure 7. Alternatively, could the reduction in muscle nuclei be caused by leaky earlier dsRNA expression during muscle development?

2) Given the recurring off-target and other artifacts with RNAi, the differences in phenotypes with different zfh1 RNAi lines are a bit worrisome. Such effects could also explain the different phenotypes with the DeltaEnh3 allele and with forced miR-8 expression. In particular, whereas with the KK line there is some premature differentiation while the undifferentiated bulk of adepithelial cells look normal, with the GD line all of the adepithelial cells look morphologically abnormal but none are differentiated. First, the authors should verify by antibody stainings that Zfh1 levels are indeed strongly reduced in these experiments. Second, they should rule out that the effect with the KK line is caused the known second site insertion that activates the Hippo pathway, which is present in a significant portion of the KK lines (Vissers et al., Nature Comm. 2016). Third, they should use the Valium10 and -20 lines from TRiP to check for consistent phenotypes. In this context, it is not always clear which one of the lines was used in a particular experiment (e.g., in Figure 7. And does the other line give the same phenotype?).

3) –In the subsection “An alternate short zfh1 isoform is transcribed in adult MPs” and elsewhere the authors argue repeatedly that Zfh1 expression and thus stem cell maintenance must "evade" regulation by miR-8, which they explain by the expression of the short isoform that lacks miR-8 target sequences. However, in Figure 4 they show that mir-8 is not even expressed in the satellite-like cells, which means there is actually no need for evasion in these cells. Currently, there is only evidence that miR-8 is expressed in fully differentiated muscles, which could be downstream of differentiation regulators such as Mef2 (in line with data from Figure 6—figure supplement 1). So the proposed evasion mechanism is highly speculative and cannot be presented as a fact without any supporting data. Such data could for instance come from G-TRACE experiments with miR-8-GAL4 that show that miR-8 is indeed transiently expressed in the precursors of the satellite-like cells.

4) –Subsection “zfh1 isoform transcription requires Notch activity in adult SCs”, second paragraph, Figure 6, Figure 8—figure supplement 1: The authors ascribe the high Mef2 levels (i.e., possible entry into differentiation) in satellite cells with depleted Notch to reduced zfh1-short. However, as they show, zfh1-long is unaffected by Notch depletion and because mir-8 is absent in these cells (see above), zfh1 could still suppress differentiation. This would argue for zfh1-independent effects of Notch.

Reviewer #2:

Bokhatmi and Bray report the presence of a population of satellite-like stem cells in the adult of Drosophila that are maintained through a switch of zfh1 RNA -isoforms that enable escape from microRNA miR-8/miR-200 control. They authors show that differentiation of MP is associated with expression of miR-8, which targets zfh1-long RNA thus down-regulating zfh1 protein. They also map an enhancer (Enh3) region that drives expression of zfh1 gene in some MPs in the larval wing imaginal discs and the adult MPs. Interestingly, this enhancer is regulated by Su (H) and these authors show very clearly that Notch-Su(H) activates an alternate zfh1 short isoform in adult MPs, where Notch is also normally activated. zfh1 short RNA lacks the miR-8 seed sequence and thus cannot be silenced by miR-8. Altogether the data suggest that Notch-driven zfh1 short allows the zfh1 to escape down-regulation of miR-8 and thus terminal differentiation in a subset of larval muscle progenitor cells. The authors also postulate that expression of the alternate zfh1-short isoform is a critical part of the regulatory switch to maintain a pool of progenitor "satellite-like" cells in the adult.

I have some critics and specific comments that need clarification:

1) Role of Zfh1 short RNA isoform in the maintenance of muscle progenitors. Some additional experiments, such as gain-of-expression of zfh1 isoforms, specifically the short form to substantiate the model.

2) Loss of mir-8 assays is paramount to the model. The mir-8 sponge should produce phenotypes that substantiate the biological relevance of the regulatory switch to maintain a subset of MPs retaining zfh1. mir-8 loss would be expected to increase the number of pMPs.

3) Epistatic studies to link mir-8 overexpression to zfh1 (gain of mir-8 along with gain of zfh1 and/or loss of mir-8 and loss of zfh1). These experiments seem important for substantiating the biological relevance of the regulatory loci discovered.

4) The conclusion that adult Zfh1+ve cells retain capacity to divide and produce progeny should be reinforced with direct data. G-TRACE is not a cell lineage method – it traces the dynamics of a Gal4 expressing cell. It is very likely the authors' conclusion is right but a clonal cell lineage method is more appropriate for this particular issue.

In summary, this is a very interesting study with a conceptually novel regulatory logic that may have broader implications for understanding how other stem cells and cancer stem-like cells may escape terminal differentiation. Nonetheless, some additional data would be required to substantiate that the zfh1 short isoform contributes (or is capable) to the maintenance of satellite-like progenitor cells.

Reviewer #3:

In the submitted manuscript H. Boukhatmi and S. Bray analyse an intriguing issue related to the mechanisms of maintenance of persistent Muscle Progenitors (pMPs) in adult Drosophila muscles focusing on flight muscles. Considering that identification and functional characterisation of Drosophila muscle satellite-like cells that express Zfh-1 has already been reported (open access BioRxiv Chaturvedi et al., manuscript posted 25 Jan 2016 from VijayRaghavan lab) the Boukhatmi and Bray's manuscript carries two new findings: i) pMPs preferentially express a short isoform of zfh1, which is devoid of miR8 seed site and ii) zfh1 is not only present in pMPs but also plays an important role in their maintenance. Authors also identify and apply a new tool, zfh1-Enh3-Gal4 that allows them to target/visualize pMPs.

My comments focus on these two aspects, which make this manuscript of potential interest for eLife if documented more rigorously.

1) Differential post-transcriptional regulation of zfh1 isoforms by miR8 could indeed represent a major regulatory mechanism in setting pMPs population. Authors have tools in their hands to document when this isoform switch takes place. In Figure 2D they show that in 3rd instar larvae Enh3>GFP is largely expressed in Zfh1/Cut positive MPs in the notum region and then they jump to adult stage to show Enh3-GFP restricted to pMPs only. It would be interesting to see when in development this restriction is effective. In other words when the satellite-like pMPs could be identified and visualized for the first time. Time lapse imaging of IFM development in Enh3-GFP pupa (approach developed in Schnorrer lab) could represent an interesting view point as well. The developmental restriction of Enh3-GFP expression should be correlated with miR8 sensor and/or mir8-Gal4 expression to evaluate whether these two regulatory events are coordinated. In relation to this it would be important to better correlate Enh3 activity with transcriptional regulation of zfh1-short isoform. It is unclear whether Enh3 indeed regulates preferentially transcription of zfh1-short. Levels of short and long transcripts could be quantified in ΔdeltaEnh3 context in wing discs (instead of Zfh1 protein shown in Figure 2G) and in adult pMPs.

2) Regarding second major finding that zfh1 is required for maintenance of pMPs it could be important to test functional relevance of the two isoforms. Enh3>miR8 could be applied to deplete pMPs of zfh1-long and Enh3>NRNAi to see if effects of N and zfh1RNAi are similar (with experimental setup like in Figure 7). Generation of CRISPRs mutation of zfh1-short promoter/TSS site could help in defining function of this isoform.

3) Referring to the role of pMPs in muscle homeostasis one complementary experiment could be to test effects of ablation of these cells in adult flies (Enh3>Rpr).

[Editors’ note: what now follows is the decision letter after the authors submitted for further consideration.]

Thank you for resubmitting your work entitled "A population of adult satellite-like cells in Drosophila is maintained through a switch in RNA-isoforms" for further consideration at eLife. Your article has been favorably evaluated by Diethard Tautz (Senior Editor) and three reviewers, one of whom is a member of our Board of Reviewing Editors.

The manuscript has been improved but there are some remaining issues that need to be addressed before acceptance, as outlined below:

The main issue concerns the evaluation of phenotypes based on staining intensities. It is still unclear whether each of these is derived from a single batch of staining or from several experimental replicates. In case of the former, this would lower the level of confidence because of inevitable variabilities in the fixation and staining procedures.

So we request either to provide information on the numbers of independent experiments in each case or, if the data came only from single experiments, to perform additional ones to obtain more meaningful statistics.

The comments of the reviewers are attached and you may respond to them as well where you find it necessary. But the main point to be dealt with is the one above.

Reviewer #1:

The revised resubmission by Boukhatmi and Bray has addressed the majority of the criticisms of the reviewers. The most important additions include:

1) Experimental evidence by fate mapping and ablation experiments that support the notion that the pMPs serve as satellite-like cells to promote adult muscle homeostasis. The recently published eLife paper by Chaturvedi et al. on the same topic is now incorporated and referenced.

2) Immunohistochemical analyses of the time course of miR-8 vs. Zfh1 expression from pupal to adult stages. Although miR-8 expression was monitored indirectly via miR-8-GAL4 driving UAS-nGFP, which due to Gal4 and GFP perdurance likely produces longer-lasting signals than miR-8 RNA proper, these data do help to strengthen the points made by the authors. The dynamic activity of Enh3 from zfh1 has been documented in more detail as well.

3) Conditional depletion experiments for Zfh1 in adults, which show a role of zfh1 in maintaining the pMPs and, presumably as a consequence, in adult muscle homeostasis.

4) Conditional depletion experiments for miR-8 using forced expression of a miR-8 sponge and, conversely, of miR-8, results of which support the mechanistic model proposed by the authors.

Altogether, these data significantly strengthen the story presented in the manuscript.

Reviewer #2:

Authors generated impressive set of new data that support and further reinforce their initial findings. In the revised version of the manuscript they also answer all issues raised in my comments.

I would like to congratulate them for this very nicely documented work that provides an example of a how differentiation versus non-differentiation of a cell could be regulated.

The images they provide are convincing but regarding the intensity it would be good to have further information in how they had performed the analysis. Usually, it is a good idea to do flip-out clones so that there is internal control and one can appreciate the different intensity between the mutant cells and the wt surrounding within the same staining. However this might be difficult for some experiment here.

Reviewer #3:

This is a much improved manuscript which clarified all my previous concerns and specific questions. The authors make a strong case and provide a mechanistic understanding on how a population of satellite-like stem cells in the adult of Drosophila maintain their undifferentiated state through a switch of zfh1 RNA -isoforms that enable the escape from the miR-8 microRNA inhibition. The regulatory logic is further sustained in the current revised version and now the authors provide new functional evidence for the requirement of mir-8 and the zfh1 short-isoform. Along with the previous results, the new data consolidate the model that differentiation of MP is linked to miR-8-mediated inhibition of zfh1 (zfh1-long RNA) and that scape from miR-8 (zfh1-short) maintains a pool of MPs that sustain muscle homeostasis. The time course analyses of expression of the genes, and the data that zfh1-positive cells contribute to differentiated adult muscle are also strong and the images clear and of high quality. In their study, they authors also map the regulation of the zfh1-short isoform to Notch, and identified an enhancer (Enh3) region that drives expression of zfh1 gene in some MPs in the larval wing imaginal discs and the adult MPs. This Enh3-Gal4 has been, and will be a valuable tool for further studies of these cells. The data are consistent with Notch-driving expression of the alternate zfh1 short isoform at the right time to allow this important stem cell factor to escape down-regulation by miR-8 in a subset of larval muscle progenitor cells. Some conclusions are derived from indirect data, but the authors explain the technical limitations and provide alternative experimental data that form a consistent model. In summary, the authors provide support for the model that expression of the alternate zfh1-short isoform is a critical part of the regulatory switch to select a pool of progenitor "satellite-like" cells that are capable to sustain muscle repair in the adult fly.

https://doi.org/10.7554/eLife.35954.019

Author response

[Editors’ note: the author responses to the first round of peer review follow.]

Reviewer #1:

[…] 1) Currently, the authors do not have any direct evidence that the observed cells act as muscle stem cells that can fuse with muscles for repairing damage. The Enh3-Gal4>G-TRACE lineage data with GFP (Figure 2) only show that both the muscles and the satellite-like cells are derived from adepithelial cells of the wing discs, as would be expected. The demonstration of proliferation is interesting, but not proof for their potential to achieve muscle repair. The crucial experiment would be to perform this experiment conditionally, such that GFP becomes only expressed in the satellite-like cells and their descendants after hatching, and then look for GFP-positive nuclei within muscles upon aging or injury. I realize there may be some diffusion of GFP in the syncytia but nevertheless, the bulk of GFP should be imported into the nuclei that express GFP RNA.

Related to this point, in Figure 7 the authors show that, upon knockdown of zfh1 after hatching and 8 days of further aging of the flies, their IFMs contain ca. 20% fewer nuclei, which they take to mean that normally that many satellite cells fuse to repair the aging muscles during this time period. This would be a surprisingly large percentage and an important result. If true, the conditional lineage tracing experiment requested above should show ca. 20% GFP-positive nuclei in IFM cross sections of aged flies similar to those shown in Figure 7. Alternatively, could the reduction in muscle nuclei be caused by leaky earlier dsRNA expression during muscle development?

We thank the reviewer for their suggestion and comments. We note however that in our initial experiments we had focused on the perdurance of the RFP signal in the newly born myoblasts to trace their origin, because of the caveats mentioned by the reviewer. However we agree that the question of whether these cells contribute to normal homeostatic muscle repair is an important one, and has not been addressed before. We have taken two strategies to investigate. First as suggested by the reviewer we have used thermosensitive Gal80 (Gal80ts) to conditionally inactivate the Enh3-Gal4 until the adult flies emerge, then raised the temperature to allow Gal4 activity and traced the progeny from this adult-only expression with G-Trace. Specifically we quantified the GFP-positive nuclei present within adult muscles 10 days after they had emerged, and found that around 24% of the nuclei contained GFP (new Figure 3C-D, subsection “Adult Zfh1+ve MP cells contribute to flight muscles”, last paragraph).These GFPpositive nuclei were most concentrated in proximity to the satellite-like cells supporting the model that their progeny contribute to muscle homeostasis. As the reviewer rightly points out, the proportion of cells that we label fits well with the predictions based on the effects from depleting zfh1 in these cells.

Second, to provide further evidence that the satellite cells are important for normal muscle repair, we performed a genetic ablation by expressing a pro-apoptotic gene, reaper. Strikingly, this led to a similar reduction in muscle nuclei to that from zfh1 depletion (new Figure 3I, see the aforementioned paragraph). Altogether these data support the conclusion that the pMPs are contributing to muscle maintenance by providing new myoblasts.

2) Given the recurring off-target and other artifacts with RNAi, the differences in phenotypes with different zfh1 RNAi lines are a bit worrisome. Such effects could also explain the different phenotypes with the DeltaEnh3 allele and with forced miR-8 expression. In particular, whereas with the KK line there is some premature differentiation while the undifferentiated bulk of adepithelial cells look normal, with the GD line all of the adepithelial cells look morphologically abnormal but none are differentiated. First, the authors should verify by antibody stainings that Zfh1 levels are indeed strongly reduced in these experiments. Second, they should rule out that the effect with the KK line is caused the known second site insertion that activates the Hippo pathway, which is present in a significant portion of the KK lines (Vissers et al., Nature Comm. 2016). Third, they should use the Valium10 and -20 lines from TRiP to check for consistent phenotypes. In this context, it is not always clear which one of the lines was used in a particular experiment (e.g., in Figure 7. And does the other line give the same phenotype?).

We appreciate the reviewer’s concerns about the specificity of the zfh1 RNAi KK line that was the primary tool used in our analysis. Following the reviewer’s request we have verified that the KK RNAi line strongly reduces Zfh1 protein levels (new Figure 1—figure supplement 1D, E). We have also taken on board their important suggestion to confirm the results using further independent RNAi lines. We have new data demonstrating that the TRiP Valium10 zfh1 shRNA gives the same premature differentiation phenotype as zfh1 RNAi KK (new Figure 1—figure supplement 1F-H). We have clarified in the figure legends which specific RNAi line was used for a given experiment.

With respect to the “differences in phenotypes” we note that there is a correlation between the severity of the phenotype, as measured by ectopic expression of differentiation markers, and the morphology of any residual MP/adepithelial cells. In some discs there is strong Tropomyosin expression in all the cells and there are few with any undifferentiated morphology, in others there are fewer Tropomyosin expressing cells and many of the MP/adepithelial cells retain a more normal morphology. In the relatively mild conditions from ∆Enh3 or from miR-8 overexpression, there is residual Zfh1 present and most of the cells retain their normal morphology rather than becoming differentiated.

Thus the differences resemble those commonly seen with alleles of varying severity (so-called “allelic series”) as expected if the experiments result in different degrees of zfh1 depletion.

3) –In the subsection “An alternate short zfh1 isoform is transcribed in adult MPs” and elsewhere the authors argue repeatedly that Zfh1 expression and thus stem cell maintenance must "evade" regulation by miR-8, which they explain by the expression of the short isoform that lacks miR-8 target sequences. However, in Figure 4 they show that mir-8 is not even expressed in the satellite-like cells, which means there is actually no need for evasion in these cells. Currently, there is only evidence that miR-8 is expressed in fully differentiated muscles, which could be downstream of differentiation regulators such as Mef2 (in line with data from Figure 6—figure supplement 1). So the proposed evasion mechanism is highly speculative and cannot be presented as a fact without any supporting data. Such data could for instance come from G-TRACE experiments with miR-8-GAL4 that show that miR-8 is indeed transiently expressed in the precursors of the satellite-like cells.

We agree that the “evasion” model implies that mIR-8 expression should overlap with Zfh1 during the stages where pMPs are selected and we appreciate the reviewer’s suggestions. To address this point we have examined the expression of mIR-8 and Zfh1 during pupal stages (at circa 20hr and 30hr after pupa formation) when the pMPs are specified. These data, which are shown in new Figure 5D-F(subsection “zfh1 is silenced by the conserved microRNA miR-8/miR-200 in MPs”, third paragraph) demonstrate that there is a phase where the two overlap in their expression. At 20hr the differentiating myoblasts have higher levels of mIR-8 than the remaining MPs while the converse is seen for Zfh-1, which is at highest levels in the MPs. By 30 hours, few Zfh1 expressing cells remain and a small subset of these retain miR-8 expression (new Figure 5D-F). Together these new data show that miR-8 is initially expressed in the precursors and that there is an important window where its expression overlaps with that of Zfh1 in the pMPs

4) Subsection “zfh1 isoform transcription requires Notch activity in adult SCs”, second paragraph, Figure 6, Figure 8—figure supplement 1: The authors ascribe the high Mef2 levels (i.e., possible entry into differentiation) in satellite cells with depleted Notch to reduced zfh1-short. However, as they show, zfh1-long is unaffected by Notch depletion and because mir-8 is absent in these cells (see above), zfh1 could still suppress differentiation. This would argue for zfh1-independent effects of Notch.

The reviewer is correct that our data argue that the main effect of perturbing Notch will be on zfh1short and that residual zfh1-long in the pMPs would, in principle, be able to prevent differentiation. However the levels of zfh1-long at this stage are comparatively low, so it is likely that the amounts of Zfh1 in the Notch-depleted pMPs are not sufficiently high to robustly inhibit the differentiation in all of the pMPs. We have modified the text (Discussion) to explain these points and included the caveat that there could also be zfh1 independent effects of Notch as highlighted by the reviewer.

Reviewer #2:

[…] I have some critics and specific comments that need clarification:

1) Role of Zfh1 short RNA isoform in the maintenance of muscle progenitors. Some additional experiments, such as gain-of-expression of zfh1 isoforms, specifically the short form to substantiate the model.

To show that Zfh1-short can counteract muscle differentiation, we have generated a UAS-Zfh1Short transgenic line and shown that its expression can prevent the premature differentiation phenotype caused by over-expressing Mef2 in the muscle progenitors. These new data, which have been added to new Figure 7—figure supplement 1 (subsection “An alternate short zfh1 isoform is transcribed in adult pMPs”, last paragraph), show that Zfh1-short protein isoform retains the ability to antagonize Mef2, as shown originally for Zfh1-long (Siles et al.,2013), and support the model that Zfh1-short protects the MPs from differentiation.

2) Loss of mir-8 assays is paramount to the model. The mir-8 sponge should produce phenotypes that substantiate the biological relevance of the regulatory switch to maintain a subset of MPs retaining zfh1. mir-8 loss would be expected to increase the number of pMPs.

We thank the reviewer for this helpful suggestion. We have now tested the consequences from expressing the miR-8 sponge specifically in the muscle progenitors during the differentiation phase, (using Mef2-Gal4). As predicted by the reviewer, this “mopping up” of miR-8 led to a significant increase in the number of adult pMPs. These new data, which have been added to new Figure 6 and discussed in the main text (subsection “zfh1 is silenced by the conserved microRNA miR-8/miR-200 in MPs”, last two paragraphs), further substantiate the model that mIR-8 counteracts the maintenance of pMPs.

3) Epistatic studies to link mir-8 overexpression to zfh1 (gain of mir-8 along with gain of zfh1 and/or loss of mir-8 and loss of zfh1). These experiments seem important for substantiating the biological relevance of the regulatory loci discovered.

Because mIR-8 and Zfh1 are ultimately required in complementary cells, it is difficult to design ideal experiments combining their misexpression/knockdown along the lines proposed by the reviewer, especially because over-expression of Zfh1-long causes lethality. We have however performed several additional experiments to test the different aspects of the model individually. For example, we have over-expressed miR-8 in the pMPs, where it leads to a reduction in pMPs consistent with miR-8 shifting the balance towards differentiation (Figure 6—figure supplement 1 and subsection “zfh1 is silenced by the conserved microRNA miR-8/miR-200 in MPs”, last two paragraphs). We have expressed the mIR-8 sponge, as summarized above, to deplete mIR-8 and shown that this increases the number of MPs as predicted (new Figure 6, see the aforementioned paragraphs). We have demonstrated that zfh1-short expression can prevent muscle differentiation induced by Mef2 (new Figure 7—figure supplement 1; subsection “An alternate short zfh1 isoform is transcribed in adult pMPs”, last paragraph). We believe that these additional experiments greatly substantiate the regulatory logic that we propose.

4) The conclusion that adult Zfh1+ve cells retain capacity to divide and produce progeny should be reinforced with direct data. G-TRACE is not a cell lineage method – it traces the dynamics of a Gal4 expressing cell. It is very likely the authors' conclusion is right but a clonal cell lineage method is more appropriate for this particular issue.

We appreciate the concerns of the reviewer, which were also raised by reviewer 1 (point 1). Due to the nature of the tissue, it is very difficult to perform a classic clonal lineage tracing (we have tried extensively without real success). We have therefore taken the strategy suggested by reviewer 1, namely to induce G-TRACE expression only in the adult by restricting it with Gal80ts until after eclosion. When we use Gal80ts to conditionally activate the Enh3-Gal4 in adult flies and trace the progeny with G-Trace, we detect a limited number of GFP-positive nuclei present within adult muscles 10 days after they had emerged (new Figure 3C-D, subsection “Adult Zfh1+ve MP cells contribute to flight muscles”, last paragraph). These GFP positive nuclei were most concentrated in proximity to the satellite-like cells supporting the model that they are derived from the MPs. Furthermore, in our initial experiments we focused on the perdurance of the RFP signal in the newly born myoblasts to link them directly to the cells of origin (Figure 3A, subsection “Adult Zfh1+ve MP cells contribute to flight muscles”, first paragraph), because of the caveats mentioned by the reviewer. Finally we have performed new experiments ablating the pMP cells and have shown that this reduces the number of muscle nuclei (new Figure 3I, subsection “Adult Zfh1+ve MP cells contribute to flight muscles”, last paragraph), supporting the conclusion that the pMPs give rise to progeny that contribute to the muscles. We believe that together these data provide clear evidence that the pMPs retain capacity to divide and produce progeny that contribute to the adult muscles.

In summary, this is a very interesting study with a conceptually novel regulatory logic that may have broader implications for understanding how other stem cells and cancer stem-like cells may escape terminal differentiation. Nonetheless, some additional data would be required to substantiate that the zfh1 short isoform contributes (or is capable) to the maintenance of satellite-like progenitor cells.

We are glad that the reviewer finds the work conceptually novel and important. We believe that the additional data we have provided in the revised version substantiate our model that Zfh1-short contributes to progenitor maintenance.

Reviewer #3:

In the submitted manuscript H. Boukhatmi and S. Bray analyse an intriguing issue related to the mechanisms of maintenance of persistent Muscle Progenitors (pMPs) in adult Drosophila muscles focusing on flight muscles. Considering that identification and functional characterisation of Drosophila muscle satellite-like cells that express Zfh-1 has already been reported (open access BioRxiv Chaturvedi et al., manuscript posted 25 Jan 2016 from VijayRaghavan lab) the Boukhatmi and Bray's manuscript carries two new findings: i) pMPs preferentially express a short isoform of zfh1, which is devoid of miR8 seed site and ii) zfh1 is not only present in pMPs but also plays an important role in their maintenance. Authors also identify and apply a new tool, zfh1-Enh3-Gal4 that allows them to target/visualize pMPs.

My comments focus on these two aspects, which make this manuscript of potential interest for eLife if documented more rigorously.

We are glad the reviewer considers that our results are of potential interest for eLife. We would like to emphasise the fact that our data address the mechanism of how muscle satellite cells are maintained and demonstrate their role in normal muscle homeostasis. As noted by the reviewer this is a substantial advance on the recent work from VijayRaghavan’s lab, which very nicely documented the existence of the persistent muscle precursor/satellite cell population and showed their expansion during muscle injury (Chaturvedi et al., 2017). The reviewer is correct that they also reported that Zfh-1 is expressed in these cells, although these data were only added to their manuscript during the final round of revisions, July 2017, at the same time ours was initially submitted. We have modified our text to take on board the additional features in their final published version. However we stress that they have not demonstrated that Zfh1 is functionally important nor shed any light on its regulation, which are key the aspects of our work and that lead to a novel model for stem cell maintenance.

1) Differential post-transcriptional regulation of zfh1 isoforms by miR8 could indeed represent a major regulatory mechanism in setting pMPs population. Authors have tools in their hands to document when this isoform switch takes place. In Figure 2D they show that in 3rd instar larvae Enh3>GFP is largely expressed in Zfh1/Cut positive MPs in the notum region and then they jump to adult stage to show Enh3-GFP restricted to pMPs only. It would be interesting to see when in development this restriction is effective. In other words when the satellite-like pMPs could be identified and visualized for the first time. Time lapse imaging of IFM development in Enh3-GFP pupa (approach developed in Schnorrer lab) could represent an interesting view point as well. The developmental restriction of Enh3-GFP expression should be correlated with miR8 sensor and/or mir8-Gal4 expression to evaluate whether these two regulatory events are coordinated. In relation to this it would be important to better correlate Enh3 activity with transcriptional regulation of zfh1-short isoform.

We agree with the reviewer that there was a gap in our analysis of Enh3-GFP and thank them for encouraging us to investigate the timing of its developmental restriction. Following their suggestions we have carried out two further sets of experiments. In the first, we have followed Enh3_GFP and Zfh1 expression at two stages during pupal development, 20h APF and 30h APF, see new Figure 2F, G and subsection “zfh1 enhancers conferring expression in MPs”, third paragraph. Both Enh3_GFP and Zfh1 are broadly expressed in myoblasts and in MPs at 20h, where a subset of unfused MPs express higher levels of both. By 30hrs Enh3_GFP and Zfh1 are restricted to a few scattered cells, the presumptive pMPs, which are closely apposed to the muscle fibers. At the same stage, a small number of Zfh1+ve cells lack detectable Enh3 expression and we speculate that these are undergoing differentiation. These data suggest that expression from Enh3 correlates with the restriction/selection of the pMPs, which begins around 20 hrs and is largely completed by 30 hrs.

In the second experiments we have investigated mIR-8 expression at similar stages (new Figure 5, subsection “zfh1 is silenced by the conserved microRNA miR-8/miR-200 in MPs”). At 18-22hr APF miR-8-Gal4 and Zfh1 expression overlap in most if not all muscle nuclei (new Figure 5D). However miR-8-Gal4 expression level was elevated in the differentiated myoblasts, whereas Zfh1 expression level was slightly lower in this population and higher in the undifferentiated MPs. By 30h APF, their expression becomes mutually exclusive, except in a few rare MPs where both are co-expressed (new Figure 5E). Thus mIR-8 and Zfh1 develop reciprocal expression profiles between 20 and 30hr APF, as predicted if mIR-8 down-regulates Zfh1. The timing of this restriction is similar to that seen for Enh3, supporting the model that Enh3 is responsible for the persistence of Zfh1 in MPs.

It is unclear whether Enh3 indeed regulates preferentially transcription of zfh1-short. Levels of short and long transcripts could be quantified in ΔdeltaEnh3 context in wing discs (instead of Zfh1 protein shown in Figure 2G) and in adult pMPs.

We agree with the reviewer that we have not fully ruled out the possibility that Enh3 could also regulate zfh-1-long. We are hampered by the fact that it is difficult to quantify the transcript levels unless we can reliably measure the individual foci in the FISH, so we could not reliably assess changes in zfh-1-long in the wing disc in ∆Enh3. In addition the early lethality of ∆Enh3 means that it is not plausible to monitor the isoforms in the adult pMPs. What we have shown is that Enh3 is the only tested zfh1 enhancer that is activated and maintained in the pMPs, where zfh1-short is the predominant isoform (Figure 2 and 7) and that depletion or up-regulation of Notch, which acts through Enh3, specifically affects zfh1-short expression (Figure 8). We cannot exclude that Enh3 also regulates zfh1-long in some circumstances, especially as the region defined is 3kb and may encompass more than one regulatory element. We have commented specifically on the possibility in the text (subsection “zfh1-short isoform transcription requires Notch activity in adult pMPs” and Figure legend 9).

2) Regarding second major finding that zfh1 is required for maintenance of pMPs it could be important to test functional relevance of the two isoforms. Enh3>miR8 could be applied to deplete pMPs of zfh1-long and Enh3>NRNAi to see if effects of N and zfh1RNAi are similar (with experimental setup like in Figure 7). Generation of CRISPRs mutation of zfh1-short promoter/TSS site could help in defining function of this isoform.

We have taken on board the thoughtful suggestions of the reviewer to explore further the role of Zfh1 short by comparing the effects from expressing N-RNAi and zfh1-RNAi in the adult pMPs using Gal80ts in combination with Enh3-Gal4 so the knock down occurs only in the adults (new Figure 8I, subsection “zfh1-short isoform transcription requires Notch activity in adult pMPs”, last paragraph). Both have similar effects on the muscles, reducing the number of nuclei in the same way as when the pMPs are ablated genetically (new Figure 3). As Notch primarily promotes expression of the zfh1-short, this supports the model that this isoform is important in preventing MPs from differentiating. While the generation of a CRISPR mutation specific for zfh1-short is potentially a powerful way to further test the role of this isoform, unfortunately it is lethal at an early stage (preliminary unpublished data from a colleague), due to expression of this isoform in embryos, so it is not possible to analyze the adult muscle phenotype.

As suggested by the reviewer, we have also now tested the consequences of forced miR-8 expression in the adult (new Figure 6—figure supplement 1D) and found that it causes a reduction in pMPs. The fact that the pMPs remain sensitive to miR-8, which would only target zfh1-long, suggests that the levels of Zfh1 are finely tuned by the expression of both zfh1-long and zfh1-short in these cells. We have explicitly discussed this possibility in the Discussion.

3) Referring to the role of pMPs in muscle homeostasis one complementary experiment could be to test effects of ablation of these cells in adult flies (Enh3>Rpr).

We thank the reviewer for this thoughtful suggestion. We have, as they proposed, performed a genetic ablation of the pMPs in adult MPs by expressing the proapoptotic gene reaper. As predicted this has led to a decrease in muscle nuclei number (new Figure 3, subsection “Adult Zfh1+ve MP cells contribute to flight muscles”, last paragraph). These results are consistent with those from zfh1 RNAi assays and provide a complementary experiment showing that the pMPs have an important role in muscle homeostasis.

[Editors' note: the author responses to the re-review follow.]

Reviewer #2:

[…] The images they provide are convincing but regarding the intensity it would be good to have further information in how they had performed the analysis. Usually, it is a good idea to do flip-out clones so that there is internal control and one can appreciate the different intensity between the mutant cells and the wt surrounding within the same staining. However this might be difficult for some experiment here.

We are glad the reviewer is positive about all the additional data we have included. We appreciate their concerns about analysis of staining intensities and agree that in tissues where flip-out clones are effective this is a powerful way to have internal controls. Unfortunately, because the cells become dispersed, the method does not work well in this tissue. However, in all cases the experiments have been replicated and we now present the data from those replicates by plotting all of the data points and shading them accordingly. We hope that the more transparent and informative plotting of all our data points demonstrates clearly the robustness of our replicate data sets and addresses these concerns.

https://doi.org/10.7554/eLife.35954.020

Article and author information

Author details

  1. Hadi Boukhatmi

    Department of Physiology, Development and Neuroscience, University of Cambridge, Cambridge, United Kingdom
    Contribution
    Conceptualization, Supervision, Funding acquisition, Validation, Investigation, Visualization, Methodology, Writing—original draft, Project administration, Writing—review and editing
    Competing interests
    No competing interests declared

  2. Sarah Bray

    Department of Physiology, Development and Neuroscience, University of Cambridge, Cambridge, United Kingdom
    Contribution
    Conceptualization, Supervision, Funding acquisition, Validation, Investigation, Visualization, Methodology, Writing—original draft, Project administration, Writing—review and editing
    For correspondence
    sjb32@cam.ac.uk
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-1642-599X

Funding

Medical Research Council (MRL007177/1)

  • Sarah Bray

European Molecular Biology Organization (ALTF-325-2013)

  • Hadi Boukhatmi

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

We acknowledge the Bloomington Stock Center, the VDRC Stock Center and the Developmental Studies Hybridoma Bank for Drosophila strains and antibodies. We thank Ruth Lehmann, Renate Renkawitz-Pohl, Eileen Furlong and István Andó for antibodies. We thank Eva Zacharioudaki, Alain Vincent and Michalis Averof for critical reading of the manuscript and other members of SJB lab for valuable discussion; and we are grateful to Hannah Green and Juanjo Perez-Moreno for advice on muscle preparations. This work was funded by a program grant from the MRC to SJB and by an EMBO Long Term Fellowship for HB.

Reviewing Editor

  1. Manfred Frasch, Friedrich-Alexander-Universität Erlangen-Nürnberg, Germany

Publication history

  1. Received: February 15, 2018
  2. Accepted: April 7, 2018
  3. Accepted Manuscript published: April 9, 2018 (version 1)
  4. Version of Record published: April 26, 2018 (version 2)

Copyright

© 2018, Boukhatmi et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 3,346
    Page views
  • 507
    Downloads
  • 23
    Citations

Article citation count generated by polling the highest count across the following sources: Crossref, Scopus, PubMed Central.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Hadi Boukhatmi
  2. Sarah Bray
(2018)
A population of adult satellite-like cells in Drosophila is maintained through a switch in RNA-isoforms
eLife 7:e35954.
https://doi.org/10.7554/eLife.35954

Categories and tags

Research organism

Further reading

    1. Developmental Biology

    Npr3 regulates neural crest and cranial placode progenitors formation through its dual function as clearance and signaling receptor

    Arun Devotta, Hugo Juraver-Geslin ... Jean-Pierre Saint-Jeannet
    Research Article Updated

    Natriuretic peptide signaling has been implicated in a broad range of physiological processes, regulating blood volume and pressure, ventricular hypertrophy, fat metabolism, and long bone growth. Here, we describe a completely novel role for natriuretic peptide signaling in the control of neural crest (NC) and cranial placode (CP) progenitors formation. Among the components of this signaling pathway, we show that natriuretic peptide receptor 3 (Npr3) plays a pivotal role by differentially regulating two developmental programs through its dual function as clearance and signaling receptor. Using a combination of MO-based knockdowns, pharmacological inhibitors and rescue assays we demonstrate that Npr3 cooperate with guanylate cyclase natriuretic peptide receptor 1 (Npr1) and natriuretic peptides (Nppa/Nppc) to regulate NC and CP formation, pointing at a broad requirement of this signaling pathway in early embryogenesis. We propose that Npr3 acts as a clearance receptor to regulate local concentrations of natriuretic peptides for optimal cGMP production through Npr1 activation, and as a signaling receptor to control cAMP levels through inhibition of adenylyl cyclase. The intracellular modulation of these second messengers therefore participates in the segregation of NC and CP cell populations.

    1. Developmental Biology
    2. Evolutionary Biology

    Mutation of vsx genes in zebrafish highlights the robustness of the retinal specification network

    Joaquín Letelier, Lorena Buono ... Juan R Martínez-Morales
    Research Article

    Genetic studies in human and mice have established a dual role for Vsx genes in retina development: an early function in progenitors’ specification, and a later requirement for bipolar-cells fate determination. Despite their conserved expression patterns, it is currently unclear to which extent Vsx functions are also conserved across vertebrates, as mutant models are available only in mammals. To gain insight into vsx function in teleosts, we have generated vsx1 and vsx2 CRISPR/Cas9 double knockouts (vsxKO) in zebrafish. Our electrophysiological and histological analyses indicate severe visual impairment and bipolar cells depletion in vsxKO larvae, with retinal precursors being rerouted toward photoreceptor or Müller glia fates. Surprisingly, neural retina is properly specified and maintained in mutant embryos, which do not display microphthalmia. We show that although important cis-regulatory remodelling occurs in vsxKO retinas during early specification, this has little impact at a transcriptomic level. Our observations point to genetic redundancy as an important mechanism sustaining the integrity of the retinal specification network, and to Vsx genes regulatory weight varying substantially among vertebrate species.

    1. Developmental Biology

    Development: The hidden depths of zebrafish skin

    Yue Rong Tan, Megan Liaw, Chen-Hui Chen
    Insight

    Single-cell transcriptome analysis of zebrafish cells clarifies the signalling pathways controlling skin formation and reveals that some cells produce proteins required for human teeth to acquire their enamel.