Neurons and glia show mutual reliance in many functional and developmental aspects of biology. Glia migrate collectively and over long distances to establish an intricate relationship with neurons. Defective glia migration is associated with several human diseases including glial brain tumors and defective regeneration following injury in the nervous system (Klämbt, 2009; Kocsis and Waxman, 2007; Oudega and Xu, 2006). Hence, a thorough understanding of the molecules involved in the process of glia migration may contribute to the development of therapeutics for these pathologies. Research progress in recent years has revealed the involvement of chemotropic cues in glia migration (von Hilchen et al., 2010; Chen et al., 2010; Spassky et al., 2002; Kinrade et al., 2001; Liu et al., 2012). Netrins, a class of secreted laminin-related extracellular proteins, have been described as chemotropic guidance cues for axons and migrating cells during neural development (Lai Wing Sun et al., 2011; Harris et al., 1996; Ishii et al., 1992; Kennedy et al., 1994; Mitchell et al., 1996; Serafini et al., 1994). In vertebrates, Netrins secreted by the floor plate cells act differentially on the migrating oligodendrocyte precursor cells and this differential outcome is dependant upon the type of receptors expressed in the migrating cells (Spassky et al., 2002; Wolf and Friedl, 2009; Jarjour, 2003; Tsai et al., 2003). Netrins act as chemoattractants through the DCC/Frazzled family of receptors (von Hilchen et al., 2010; Lai Wing Sun et al., 2011; Chan et al., 1996; Keino-Masu et al., 1996; Kolodziej et al., 1996; Timofeev et al., 2012) and chemorepellants through the Unc5 receptor family (Keleman and Dickson, 2001; Labrador et al., 2005). In Drosophila, the two Netrins (NetA and NetB [Harris et al., 1996; Mitchell et al., 1996; Keleman and Dickson, 2001]) and their receptor Frazzled (Fra) mediate the attraction of embryonic longitudinal glia toward the midline (von Hilchen et al., 2010). Despite the extensive knowledge on these ligands and receptors, the transcriptional control underlying chemoattraction and the impact of this pathway remain largely unknown. For example, what is the specific role of the receptors and how do they contribute to the different steps of collective migration, such as initiation, maintenance and arrest? Related to this issue, which transcription factors regulate the cell-specific and timely expression of the receptors? We here investigate the chemoattraction cascade that controls cell migration using the chain of glial cells across the L1 nerve in the developing Drosophila wing (Aigouy et al., 2004, 2008; Berzsenyi et al., 2011; Kumar et al., 2015).

We show that only one of the two Drosophila Netrins, NetB, serves as a chemoattractant for collective glia migration. The role of Fra is to control the time of initiation of glia migration in a dosage-dependent manner, whereas Unc5 acts as the repellant receptor that controls glial arrest. Finally, we identify the transcription factor that controls the expression of Fra at the appropriate time and levels: Glial cell deficient/Glial cell missing (Glide/Gcm or Gcm, for the sake of simplicity), the fate determinant that is expressed early and transiently in the glial lineages (Hosoya et al., 1995; Vincent et al., 1996; Jones et al., 1995). Thus, we find that an early gene, which regulates the expression of transcription factors that execute a specific cell fate, also regulates effector genes that controls late developmental events. To our knowledge, this is the first report showing that a fate determinant directly controls collective cell migration, prompting us to revisit the role and mode of action of these types of molecules during development.

Collective migration is a complex biological process that allows cells to leave their place of birth and reach their destination in a coordinated and timely manner. Here we dissect the signaling pathway underlying the collective migration of glia on the developing Drosophila wing. The careful coordination between extrinsic cues and intrinsic, cell autonomous mechanisms ensure that timely migration of the glia occurs on the wing disc. The chemoattractant receptor Fra controls glia migration in response to a long-distance signal, thechemoattractant NetB. Fra is expressed before glia start to move and triggers migration initiation in a dose-dependent manner. Such tight control of Fra expression depends upon the Gcm transcription factor, which directly induces the expression of Fra at threshold levels. Thus, the glial determinant factor also affects a specific step of collective migration by regulating the key effector gene.

Collective migration comes in different flavors: streams, chains, sheets and clusters, which all imply tight coordination and cell–cell interactions (Klämbt, 2009; Gilmour et al., 2002; Gupta and Giangrande, 2014; Marín et al., 2010; Rørth, 2003). The small cluster of Drosophila border cells migrates through nurse cells towards the oocyte in response to growth factors (Montell, 2003; Rørth, 2009). The stream of hundreds of proliferating cells of the fish lateral line migrate directionally upon expressing different chemokine receptors within the collective (Dambly-Chaudiere et al., 2007; David et al., 2002; Ghysen and Dambly-Chaudière, 2004; Haas and Gilmour, 2006). The current challenge is to dissect the role of the signaling pathways in the different steps (initiation, maintenance and arrest) and features of collective migration (adhesion, overall velocity and timing) and to analyze this in vivo. The second challenge is to understand how those signaling pathways are regulated, an issue that has been addressed at the post-transcriptional level (Yu and Bargmann, 2001) but much less so at the transcriptional level.

By focusing on the Netrin signaling pathway, which has been extensively studied in the context of axonal navigation and cell proliferation, we have addressed the above questions using the migrating wing glia of Drosophila. The analysis of the mutant phenotypes and the time-lapse approach show that the Fra receptor controls glia migration along the L1 nerve. Reducing the amount of Fra delays migration, whereas excessive Fra in the glial cells triggers their precocious migration. Thus, the Fra receptor plays an instructive role in the first step of collective migration, initiation. These data highlight the importance of quantitative regulation: large cohorts of cells are likely to need strong forces to switch from an immotile to a motile phenotype; therefore only the strong expression of the receptor allows migration toward the chemoattractant. Other migratory collectives also depend on quantitative regulation. Typically, epithelial cells migrating in groups exert much stronger forces than an individual cell before and after epithelial-mesenchymal transition (du Roure et al., 2005). The cell adhesion molecule E-cadherin plays an important role in supporting the directional and efficient collective migration of epithelial cells by mediating adhesion and force generation (Li et al., 2012). Additionally, the collectively migrating border cell cluster cannot exert enough adhesive/pulling force to move between the nurse cells in the absence of E-cadherin (Niewiadomska et al., 1999).

The timely regulation of the Fra receptor and its early role in collective migration allowed us to show that the Gcm fate determinant affects migration through this receptor. Thus, early genes not only trigger the expression of transcription factors that in turn implement a specific developmental program, but also directly contribute to the acquisition of specific phenotypes, such as the migratory potential. Similarly, the Lim-homeodomain transcription factor Islet was shown to specify the electrical properties of motor neurons by repressing the expression of the ion channel Shaker during development, suggesting that the regulation of late genes by early transcription factors might also be a common phenomenon (Wolfram et al., 2012). These observations prompt us to revisit the role of the so-called ‘master regulators’, which indeed control processes beyond fate choices.

Netrins are thought to act either in a gradient at long range, as secreted molecules, or at short range, as membrane-tethered molecules (Lai Wing Sun et al., 2011; Brankatschk and Dickson, 2006). In contrast with studies in the Drosophila embryo and visual system (von Hilchen et al., 2010; Timofeev et al., 2012; Brankatschk and Dickson, 2006), we observed that secreted NetB in the developing wing acts at a long range: glia migration is delayed when solely membrane-tethered NetB is available at near-endogenous levels. While we still do not know whether the distinction between long- and short-range signaling depends on the complement of surface receptors and the associated transduction pathways, it seems that different strategies are used in specific processes. Future studies will reveal whether long-range signaling is specifically dedicated to migration over many cell diameters and/or to large collectives.

NetA and NetB have been suggested to act on the embryonic longitudinal glia through Fra (von Hilchen et al., 2010) and both Netrins are reported to act as chemoattractants, even though in one case, only NetB was proposed to mediate dendritic targeting via Fra (Matthews and Grueber, 2011). Clearly, downregulating or overexpressing NetA has no impact on the migrating glial cells of the wing and NetA cannot rescue the NetB phenotype. While it is possible that NetA may affect other aspects of wing glia biology, our data indicate that NetB alone serves as a chemoattractant ligand to guide L1 glia migration upon signaling to Fra. Studies in other systems will assess whether it is only NetB that works as a long-range chemoattractant. If this were the case, it would be interesting to assess whether the different behavior of NetA and NetB relies on the intrinsic potentials of the two ligands (e.g. the affinity with which NetA and NetB bind to their receptors) or on extrinsic cues (e.g. cell-specific cofactors modulating NetA and/or NetB activity).

The chemoattractant receptor DCC/Fra and the chemorepellant receptor Unc5 have been mostly studied in different cell types, so it is unclear whether the same process and cells require the counteracting activities of these molecules. While mammalian migrating oligodendrocyte precursor cells express both DCC and Unc5, the precise function of these receptors has not been assessed (Tsai et al., 2003). We found that the two receptors Fra and Unc5 are expressed by the same set of PNS glial cells, with Fra appearing earlier than Unc5. The two receptors seem to have a different impact on wing glia migration and their stoichiometry is an important factor, as reducing Unc5 levels does have an effect on migration when fra is downregulated (Figure 8—figure supplement 1b last columns). The fact that overexpressing Unc5 and Fra have opposite effects suggests that Unc5 converts the Netrin-mediated signal from attraction to repulsion in the Drosophila wing glia, in agreement with published data on chimeric molecules. For example, when fused to the cytoplasmic domain of Unc5, the extracellular domain of DCC is as effective in mediating the chemorepellant response to Netrin as a wild type Unc5 (Hong et al., 1999). Additionally, a study performed on Drosophila showed that fusing the cytoplasmic domain of Fra with the extracellular domain of Unc5 signals chemoattraction (Keleman and Dickson, 2001). This may explain how the same cues such as Netrins can work differently on various aspects of collective migratory behavior.

The key element in this study is time: Gcm accumulates and triggers the onset of glial migration through fra but subsequently fades away. The fact that the expression of fra stays on till the end of migration suggests that another player, possibly a direct target of Gcm, maintains its expression. Being a major Gcm target, repo represents a potential candidate. Our in vivo and cell transfection data using a Repo expression vector are in line with this hypothesis and suggest that Repo may play a regulatory role in the maintenance of fra at late migratory stages (Figure 9a). It is known that Repo expression first depends on Gcm but subsequently becomes independent of it through autoregulation (Flici et al., 2014), indicating a transition from early to late events in gliogenesis (see model in Figure 9b). We propose that the initial Fra expression in wing glia strictly depends on Gcm and triggers the initiation of migration (see the graph in Figure 3—figure supplement 1e). Accordingly, the overexpression of Fra using the repo promoter does not induce migratory defects (Figure 9c,d) and repo>fra KD wings show only modest defects in migration initiation (Figure 9—figure supplement 1b). Later on, however, when Gcm fades away, Repo takes over in controlling Fra expression, hence affecting later phases of migration (Figure 9—figure supplement 1b, see model in Figure 9b). Conversely, the unc5 repellant is a weak target of Gcm thatmay require Repo and controls migration completion.

With respect to what has previously been described in embryonic glia migration, we propose a different role for Netrins and their receptors in the migration of L1 glia. The migration of embryonic longitudinal glial cells seems to rely on very early expression of Fra, which is detected in the longitudinal glial precursor, but no longer in the longitudinal glial cells (von Hilchen et al., 2010; Kolodziej et al., 1996). In our work, we do detect Fra expression in the wing glia. However, while in the case of the longitudinal glia both Netrins guide the glia toward the midline, the L1 glia seem to require only NetB. Moreover, ectopic expression of Netrin affects the migration of the longitudinal glia, whereas it does not do so in wing glia, hence suggesting that Netrin has a role in migration efficiency rather than on directionality. And finally, we propose that Fra and Unc5 act in different migratory phases of the same glial population, the first phase being under the direct control of the Gcm glial determinant. Overall, this paper displays a model in which the ligand NetB is expressed early during wing development (Figure 7—figure supplement 2), whereas its receptor Fra is expressed right before migration onset and cell autonomously decides the time of migration initiation of the glial chain. Unc5 appears later and contributes to migration termination. The acquisition of the Fra levels that trigger migration is part of the cell-specification program dictated by the Gcm glial determinant, which not only induces the expression of downstream transcription factors but also directly implements a specific developmental program by triggering the expression of effector genes.

Together, these findings in the Drosophila wing suggest that the dynamic and coordinated action of fate determinants and chemotropic cues contribute to the timely and efficient migration of the glial cells. A similar molecular mechanism, relying on Netrins or other localized attractive cues and their receptors, may be used in other cases of collective migration.

1. 1. Aigouy B
   2. Lepelletier L
   3. Giangrande A

   (2008) Glial chain migration requires pioneer cells

   Journal of Neuroscience 28:11635–11641.

   https://doi.org/10.1523/JNEUROSCI.1361-08.2008

   • PubMed
   • Google Scholar
2. 1. Aigouy B
   2. Van de Bor V
   3. Boeglin M
   4. Giangrande A

   (2004) Time-lapse and cell ablation reveal the role of cell interactions in fly glia migration and proliferation

   Development 131:5127–5138.

   https://doi.org/10.1242/dev.01398

   • PubMed
   • Google Scholar
3. 1. Asha H
   2. Nagy I
   3. Kovacs G
   4. Stetson D
   5. Ando I
   6. Dearolf CR

   (2003)

   Analysis of Ras-induced overproliferation in Drosophila hemocytes

   Genetics 163:203–215.

   • PubMed
   • Google Scholar
4. 1. Bashaw GJ
   2. Goodman CS

   (1999) Chimeric axon guidance receptors: the cytoplasmic domains of slit and netrin receptors specify attraction versus repulsion

   Cell 97:917–926.

   https://doi.org/10.1016/s0092-8674(00)80803-x

   • PubMed
   • Google Scholar
5. 1. Bataillé L
   2. Augé B
   3. Ferjoux G
   4. Haenlin M
   5. Waltzer L

   (2005) Resolving embryonic blood cell fate choice in Drosophila: interplay of GCM and RUNX factors

   Development 132:4635–4644.

   https://doi.org/10.1242/dev.02034

   • PubMed
   • Google Scholar
6. 1. Bernardoni R
   2. Miller AA
   3. Giangrande A

   (1998)

   Glial differentiation does not require a neural ground state

   Development 125:3189–3200.

   • PubMed
   • Google Scholar
7. 1. Bernardoni R
   2. Vivancos V
   3. Giangrande A

   (1997) glide/gcm is expressed and required in the scavenger cell lineage

   Developmental Biology 191:118–130.

   https://doi.org/10.1006/dbio.1997.8702

   • PubMed
   • Google Scholar
8. 1. Berzsenyi S
   2. Kumar A
   3. Giangrande A

   (2011) Homeostatic interactions at the front of migration control the integrity and the efficiency of a migratory glial chain

   Journal of Neuroscience 31:13722–13727.

   https://doi.org/10.1523/JNEUROSCI.2473-11.2011

   • PubMed
   • Google Scholar
9. 1. Bosch JA
   2. Tran NH
   3. Hariharan IK

   (2015) CoinFLP: a system for efficient mosaic screening and for visualizing clonal boundaries in Drosophila

   Development 142:597–606.

   https://doi.org/10.1242/dev.114603

   • PubMed
   • Google Scholar
10. 1. Brankatschk M
    2. Dickson BJ

    (2006) Netrins guide Drosophila commissural axons at short range

    Nature Neuroscience 9:188–194.

    https://doi.org/10.1038/nn1625

    • PubMed
    • Google Scholar
11. 1. Cattenoz PB
    2. Popkova A
    3. Southall TD
    4. Aiello G
    5. Brand AH
    6. Giangrande A

    (2016) Functional conservation of the Glide/Gcm regulatory network controlling glia, hemocyte, and tendon cell differentiation in Drosophila

    Genetics 202:191–219.

    https://doi.org/10.1534/genetics.115.182154

    • PubMed
    • Google Scholar
12. 1. Chan SS
    2. Zheng H
    3. Su MW
    4. Wilk R
    5. Killeen MT
    6. Hedgecock EM
    7. Culotti JG

    (1996) UNC-40, a C. elegans homolog of DCC (Deleted in Colorectal Cancer), is required in motile cells responding to UNC-6 netrin cues

    Cell 87:187–195.

    https://doi.org/10.1016/S0092-8674(00)81337-9

    • PubMed
    • Google Scholar
13. 1. Chen H
    2. Wei Q
    3. Zhang J
    4. Xu C
    5. Tan T
    6. Ji W

    (2010) Netrin-1 signaling mediates NO-induced glial precursor migration and accumulation

    Cell Research 20:238–241.

    https://doi.org/10.1038/cr.2010.7

    • PubMed
    • Google Scholar
14. 1. Dambly-Chaudière C
    2. Cubedo N
    3. Ghysen A

    (2007) Control of cell migration in the development of the posterior lateral line: antagonistic interactions between the chemokine receptors CXCR4 and CXCR7/RDC1

    BMC Developmental Biology 7:23.

    https://doi.org/10.1186/1471-213X-7-23

    • PubMed
    • Google Scholar
15. 1. David NB
    2. Sapede D
    3. Saint-Etienne L
    4. Thisse C
    5. Thisse B
    6. Dambly-Chaudiere C
    7. Rosa FM
    8. Ghysen A

    (2002) Molecular basis of cell migration in the fish lateral line: role of the chemokine receptor CXCR4 and of its ligand, SDF1

    PNAS 99:16297–16302.

    https://doi.org/10.1073/pnas.252339399

    • Google Scholar
16. 1. de Nooij JC
    2. Letendre MA
    3. Hariharan IK

    (1996) A cyclin-dependent kinase inhibitor, Dacapo, is necessary for timely exit from the cell cycle during Drosophila embryogenesis

    Cell 87:1237–1247.

    https://doi.org/10.1016/S0092-8674(00)81819-X

    • PubMed
    • Google Scholar
17. 1. du Roure O
    2. Saez A
    3. Buguin A
    4. Austin RH
    5. Chavrier P
    6. Silberzan P
    7. Ladoux B
    8. Siberzan P

    (2005) Force mapping in epithelial cell migration

    PNAS 102:2390–2395.

    https://doi.org/10.1073/pnas.0408482102

    • Google Scholar
18. 1. Edgar BA
    2. Lehman DA
    3. O'Farrell PH

    (1994)

    Transcriptional regulation of string (cdc25): a link between developmental programming and the cell cycle

    Development 120:3131–3143.

    • PubMed
    • Google Scholar
19. 1. Edgar BA
    2. O'Farrell PH

    (1989) Genetic control of cell division patterns in the Drosophila embryo

    Cell 57:177–187.

    https://doi.org/10.1016/0092-8674(89)90183-9

    • PubMed
    • Google Scholar
20. 1. Flici H
    2. Cattenoz PB
    3. Komonyi O
    4. Laneve P
    5. Erkosar B
    6. Karatas OF
    7. Reichert H
    8. Berzsenyi S
    9. Giangrande A

    (2014) Interlocked loops trigger lineage specification and stable fates in the Drosophila nervous system

    Nature Communications 5:4484.

    https://doi.org/10.1038/ncomms5484

    • PubMed
    • Google Scholar
21. 1. Freeman MR
    2. Delrow J
    3. Kim J
    4. Johnson E
    5. Doe CQ

    (2003) Unwrapping glial biology: Gcm target genes regulating glial development, diversification, and function

    Neuron 38:567–580.

    https://doi.org/10.3410/f.1013753.191693

    • PubMed
    • Google Scholar
22. 1. Garbe DS
    2. O'Donnell M
    3. Bashaw GJ

    (2007) Cytoplasmic domain requirements for Frazzled-mediated attractive axon turning at the Drosophila midline

    Development 134:4325–4334.

    https://doi.org/10.1242/dev.012872

    • PubMed
    • Google Scholar
23. 1. Ghysen A
    2. Dambly-Chaudière C

    (2004) Development of the zebrafish lateral line

    Current Opinion in Neurobiology 14:67–73.

    https://doi.org/10.1016/j.conb.2004.01.012

    • PubMed
    • Google Scholar
24. 1. Giangrande A
    2. Murray MA
    3. Palka J

    (1993)

    Development and organization of glial cells in the peripheral nervous system of Drosophila melanogaster

    Development 117:895–904.

    • PubMed
    • Google Scholar
25. 1. Gilmour DT
    2. Maischein HM
    3. Nüsslein-Volhard C

    (2002) Migration and function of a glial subtype in the vertebrate peripheral nervous system

    Neuron 34:577–588.

    https://doi.org/10.1016/S0896-6273(02)00683-9

    • PubMed
    • Google Scholar
26. 1. Gupta T
    2. Giangrande A

    (2014) Collective cell migration: "all for one and one for all"

    Journal of Neurogenetics 28:190–198.

    https://doi.org/10.3109/01677063.2014.896911

    • PubMed
    • Google Scholar
27. 1. Haas P
    2. Gilmour D

    (2006) Chemokine signaling mediates self-organizing tissue migration in the zebrafish lateral line

    Developmental Cell 10:673–680.

    https://doi.org/10.1016/j.devcel.2006.02.019

    • PubMed
    • Google Scholar
28. 1. Harris R
    2. Sabatelli LM
    3. Seeger MA

    (1996) Guidance cues at the Drosophila CNS midline: identification and characterization of two Drosophila Netrin/UNC-6 homologs

    Neuron 17:217–228.

    https://doi.org/10.1016/S0896-6273(00)80154-3

    • PubMed
    • Google Scholar
29. 1. Hayashi S
    2. Ito K
    3. Sado Y
    4. Taniguchi M
    5. Akimoto A
    6. Takeuchi H
    7. Aigaki T
    8. Matsuzaki F
    9. Nakagoshi H
    10. Tanimura T
    11. Ueda R
    12. Uemura T
    13. Yoshihara M
    14. Goto S

    (2002) GETDB, a database compiling expression patterns and molecular locations of a collection of Gal4 enhancer traps

    Genesis 34:58–61.

    https://doi.org/10.1002/gene.10137

    • PubMed
    • Google Scholar
30. 1. Hidalgo A

    (1994) Three distinct roles for the engrailed gene in Drosophila wing development

    Current Biology 4:1087–1098.

    https://doi.org/10.1016/S0960-9822(00)00247-5

    • PubMed
    • Google Scholar
31. 1. Hong K
    2. Hinck L
    3. Nishiyama M
    4. Poo MM
    5. Tessier-Lavigne M
    6. Stein E

    (1999) A ligand-gated association between cytoplasmic domains of UNC5 and DCC family receptors converts netrin-induced growth cone attraction to repulsion

    Cell 97:927–941.

    https://doi.org/10.1016/S0092-8674(00)80804-1

    • PubMed
    • Google Scholar
32. 1. Hosoya T
    2. Takizawa K
    3. Nitta K
    4. Hotta Y

    (1995) glial cells missing: a binary switch between neuronal and glial determination in Drosophila

    Cell 82:1025–1036.

    https://doi.org/10.1016/0092-8674(95)90281-3

    • PubMed
    • Google Scholar
33. 1. Inaba M
    2. Yuan H
    3. Yamashita YM

    (2011) String (Cdc25) regulates stem cell maintenance, proliferation and aging in Drosophila testis

    Development 138:5079–5086.

    https://doi.org/10.1242/dev.072579

    • PubMed
    • Google Scholar
34. 1. Ishii N
    2. Wadsworth WG
    3. Stern BD
    4. Culotti JG
    5. Hedgecock EM

    (1992) UNC-6, a laminin-related protein, guides cell and pioneer axon migrations in C. elegans

    Neuron 9:873–881.

    https://doi.org/10.1016/0896-6273(92)90240-E

    • PubMed
    • Google Scholar
35. 1. Jacques C
    2. Soustelle L
    3. Nagy I
    4. Diebold C
    5. Giangrande A

    (2009) A novel role of the glial fate determinant glial cells missing in hematopoiesis

    The International Journal of Developmental Biology 53:1013–1022.

    https://doi.org/10.1387/ijdb.082726cj

    • PubMed
    • Google Scholar
36. 1. Jarjour AA
    2. Manitt C
    3. Moore SW
    4. Thompson KM
    5. Yuh SJ
    6. Kennedy TE

    (2003)

    Netrin-1 is a chemorepellent for oligodendrocyte precursor cells in the embryonic spinal cord

    Journal of Neuroscience 23:3735–3744.

    • PubMed
    • Google Scholar
37. 1. Jones BW
    2. Fetter RD
    3. Tear G
    4. Goodman CS

    (1995) Glial cells missing: a genetic switch that controls glial versus neuronal fate

    Cell 82:1013–1023.

    https://doi.org/10.1016/0092-8674(95)90280-5

    • PubMed
    • Google Scholar
38. 1. Keino-Masu K
    2. Masu M
    3. Hinck L
    4. Leonardo ED
    5. Chan SS
    6. Culotti JG
    7. Tessier-Lavigne M

    (1996) Deleted in Colorectal Cancer (DCC) encodes a netrin receptor

    Cell 87:175–185.

    https://doi.org/10.1016/S0092-8674(00)81336-7

    • PubMed
    • Google Scholar
39. 1. Keleman K
    2. Dickson BJ

    (2001) Short- and long-range repulsion by the Drosophila Unc5 netrin receptor

    Neuron 32:605–617.

    https://doi.org/10.1016/S0896-6273(01)00505-0

    • PubMed
    • Google Scholar
40. 1. Kennedy TE
    2. Serafini T
    3. de la Torre JR
    4. Tessier-Lavigne M

    (1994) Netrins are diffusible chemotropic factors for commissural axons in the embryonic spinal cord

    Cell 78:425–435.

    https://doi.org/10.1016/0092-8674(94)90421-9

    • PubMed
    • Google Scholar
41. 1. Kinrade EF
    2. Brates T
    3. Tear G
    4. Hidalgo A

    (2001)

    Roundabout signalling, cell contact and trophic support confine longitudinal glia and axons in the Drosophila CNS

    Development 128:207–216.

    • PubMed
    • Google Scholar
42. 1. Klämbt C

    (2009) Modes and regulation of glial migration in vertebrates and invertebrates

    Nature Reviews Neuroscience 10:769–779.

    https://doi.org/10.1038/nrn2720

    • PubMed
    • Google Scholar
43. 1. Kocsis JD
    2. Waxman SG

    (2007) Schwann cells and their precursors for repair of central nervous system myelin

    Brain 130:1978–1980.

    https://doi.org/10.1093/brain/awm161

    • PubMed
    • Google Scholar
44. 1. Kolodziej PA
    2. Timpe LC
    3. Mitchell KJ
    4. Fried SR
    5. Goodman CS
    6. Jan LY
    7. Jan YN

    (1996) frazzled encodes a Drosophila member of the DCC immunoglobulin subfamily and is required for CNS and motor axon guidance

    Cell 87:197–204.

    https://doi.org/10.1016/S0092-8674(00)81338-0

    • PubMed
    • Google Scholar
45. 1. Kumar A
    2. Gupta T
    3. Berzsenyi S
    4. Giangrande A

    (2015) N-cadherin negatively regulates collective Drosophila glial migration through actin cytoskeleton remodeling

    Journal of Cell Science 128:900–912.

    https://doi.org/10.1242/jcs.157974

    • PubMed
    • Google Scholar
46. 1. Labrador JP
    2. O'keefe D
    3. Yoshikawa S
    4. McKinnon RD
    5. Thomas JB
    6. Bashaw GJ

    (2005) The homeobox transcription factor even-skipped regulates netrin-receptor expression to control dorsal motor-axon projections in Drosophila

    Current Biology 15:1413–1419.

    https://doi.org/10.1016/j.cub.2005.06.058

    • PubMed
    • Google Scholar
47. 1. Lai Wing Sun K
    2. Correia JP
    3. Kennedy TE

    (2011) Netrins: versatile extracellular cues with diverse functions

    Development 138:2153–2169.

    https://doi.org/10.1242/dev.044529

    • PubMed
    • Google Scholar
48. 1. Lane ME
    2. Sauer K
    3. Wallace K
    4. Jan YN
    5. Lehner CF
    6. Vaessin H

    (1996) Dacapo, a cyclin-dependent kinase inhibitor, stops cell proliferation during Drosophila development

    Cell 87:1225–1235.

    https://doi.org/10.1016/S0092-8674(00)81818-8

    • PubMed
    • Google Scholar
49. 1. Lasko P

    (2013) Development: new wrinkles on genetic control of the MBT

    Current Biology 23:R65–67.

    https://doi.org/10.1016/j.cub.2012.12.012

    • PubMed
    • Google Scholar
50. 1. Lawrence PA
    2. Morata G

    (1976) Compartments in the wing of Drosophila: a study of the engrailed gene

    Developmental Biology 50:321–337.

    https://doi.org/10.1016/0012-1606(76)90155-X

    • PubMed
    • Google Scholar
51. 1. Li L
    2. Hartley R
    3. Reiss B
    4. Sun Y
    5. Pu J
    6. Wu D
    7. Lin F
    8. Hoang T
    9. Yamada S
    10. Jiang J
    11. Zhao M

    (2012) E-cadherin plays an essential role in collective directional migration of large epithelial sheets

    Cellular and Molecular Life Sciences 69:2779–2789.

    https://doi.org/10.1007/s00018-012-0951-3

    • PubMed
    • Google Scholar
52. 1. Liu X
    2. Lu Y
    3. Zhang Y
    4. Li Y
    5. Zhou J
    6. Yuan Y
    7. Gao X
    8. Su Z
    9. He C

    (2012) Slit2 regulates the dispersal of oligodendrocyte precursor cells via Fyn/RhoA signaling

    Journal of Biological Chemistry 287:17503–17516.

    https://doi.org/10.1074/jbc.M111.317610

    • PubMed
    • Google Scholar
53. 1. Manhire-Heath R
    2. Golenkina S
    3. Saint R
    4. Murray MJ

    (2013) Netrin-dependent downregulation of Frazzled/DCC is required for the dissociation of the peripodial epithelium in Drosophila

    Nature Communications 4:2790.

    https://doi.org/10.1038/ncomms3790

    • PubMed
    • Google Scholar
54. 1. Marín O
    2. Valiente M
    3. Ge X
    4. Tsai LH

    (2010) Guiding neuronal cell migrations

    Cold Spring Harbor Perspectives in Biology 2:a001834.

    https://doi.org/10.1101/cshperspect.a001834

    • PubMed
    • Google Scholar
55. 1. Matthews BJ
    2. Grueber WB

    (2011) Dscam1-mediated self-avoidance counters netrin-dependent targeting of dendrites in Drosophila

    Current Biology 21:1480–1487.

    https://doi.org/10.1016/j.cub.2011.07.040

    • PubMed
    • Google Scholar
56. 1. Ming GL
    2. Song HJ
    3. Berninger B
    4. Holt CE
    5. Tessier-Lavigne M
    6. Poo MM

    (1997) cAMP-dependent growth cone guidance by netrin-1

    Neuron 19:1225–1235.

    https://doi.org/10.1016/S0896-6273(00)80414-6

    • PubMed
    • Google Scholar
57. 1. Mitchell KJ
    2. Doyle JL
    3. Serafini T
    4. Kennedy TE
    5. Tessier-Lavigne M
    6. Goodman CS
    7. Dickson BJ

    (1996) Genetic analysis of Netrin genes in Drosophila: Netrins guide CNS commissural axons and peripheral motor axons

    Neuron 17:203–215.

    https://doi.org/10.1016/S0896-6273(00)80153-1

    • PubMed
    • Google Scholar
58. 1. Montell DJ

    (2003) Border-cell migration: the race is on

    Nature Reviews Molecular Cell Biology 4:13–24.

    https://doi.org/10.1038/nrm1006

    • PubMed
    • Google Scholar
59. 1. Newquist G
    2. Drennan JM
    3. Lamanuzzi M
    4. Walker K
    5. Clemens JC
    6. Kidd T

    (2013a) Blocking apoptotic signaling rescues axon guidance in Netrin mutants

    Cell Reports 3:595–606.

    https://doi.org/10.1016/j.celrep.2013.02.017

    • PubMed
    • Google Scholar
60. 1. Newquist G
    2. Hogan J
    3. Walker K
    4. Lamanuzzi M
    5. Bowser M
    6. Kidd T

    (2013b) Control of male and female fertility by the netrin axon guidance genes

    PLoS One 8:e72524.

    https://doi.org/10.1371/journal.pone.0072524

    • PubMed
    • Google Scholar
61. 1. Niewiadomska P
    2. Godt D
    3. Tepass U

    (1999) DE-Cadherin is required for intercellular motility during Drosophila oogenesis

    The Journal of Cell Biology 144:533–547.

    https://doi.org/10.1083/jcb.144.3.533

    • PubMed
    • Google Scholar
62. 1. Oudega M
    2. Xu XM

    (2006) Schwann cell transplantation for repair of the adult spinal cord

    Journal of Neurotrauma 23:453–467.

    https://doi.org/10.1089/neu.2006.23.453

    • PubMed
    • Google Scholar
63. 1. Pauli A
    2. Althoff F
    3. Oliveira RA
    4. Heidmann S
    5. Schuldiner O
    6. Lehner CF
    7. Dickson BJ
    8. Nasmyth K

    (2008) Cell-type-specific TEV protease cleavage reveals cohesin functions in Drosophila neurons

    Developmental Cell 14:239–251.

    https://doi.org/10.1016/j.devcel.2007.12.009

    • PubMed
    • Google Scholar
64. 1. Pfeiffer BD
    2. Jenett A
    3. Hammonds AS
    4. Ngo T-TB
    5. Misra S
    6. Murphy C
    7. Scully A
    8. Carlson JW
    9. Wan KH
    10. Laverty TR
    11. Mungall C
    12. Svirskas R
    13. Kadonaga JT
    14. Doe CQ
    15. Eisen MB
    16. Celniker SE
    17. Rubin GM

    (2008) Tools for neuroanatomy and neurogenetics in Drosophila

    PNAS 105:9715–9720.

    https://doi.org/10.1073/pnas.0803697105

    • Google Scholar
65. 1. Popkova A
    2. Bernardoni R
    3. Diebold C
    4. Van de Bor V
    5. Schuettengruber B
    6. González I
    7. Busturia A
    8. Cavalli G
    9. Giangrande A

    (2012) Polycomb controls gliogenesis by regulating the transient expression of the Gcm/Glide fate determinant

    PLoS Genetics 8:e1003159.

    https://doi.org/10.1371/journal.pgen.1003159

    • PubMed
    • Google Scholar
66. 1. Riabinina O
    2. Luginbuhl D
    3. Marr E
    4. Liu S
    5. Wu MN
    6. Luo L
    7. Potter CJ

    (2015) Improved and expanded Q-system reagents for genetic manipulations

    Nature Methods 12:219–222.

    https://doi.org/10.1038/nmeth.3250

    • PubMed
    • Google Scholar
67. 1. Rørth P

    (2003) Communication by touch: role of cellular extensions in complex animals

    Cell 112:595–598.

    https://doi.org/10.1016/s0092-8674(03)00156-9

    • PubMed
    • Google Scholar
68. 1. Rørth P

    (2009) Collective cell migration

    Annual Review of Cell and Developmental Biology 25:407–429.

    https://doi.org/10.1146/annurev.cellbio.042308.113231

    • PubMed
    • Google Scholar
69. 1. Serafini T
    2. Kennedy TE
    3. Galko MJ
    4. Mirzayan C
    5. Jessell TM
    6. Tessier-Lavigne M

    (1994) The netrins define a family of axon outgrowth-promoting proteins homologous to C. elegans UNC-6

    Cell 78:409–424.

    https://doi.org/10.1016/0092-8674(94)90420-0

    • PubMed
    • Google Scholar
70. 1. Soustelle L
    2. Aigouy B
    3. Asensio ML
    4. Giangrande A

    (2008) UV laser mediated cell selective destruction by confocal microscopy

    Neural Development 3:11.

    https://doi.org/10.1186/1749-8104-3-11

    • PubMed
    • Google Scholar
71. 1. Soustelle L
    2. Trousse F
    3. Jacques C
    4. Ceron J
    5. Cochard P
    6. Soula C
    7. Giangrande A

    (2007) Neurogenic role of Gcm transcription factors is conserved in chicken spinal cord

    Development 134:625–634.

    https://doi.org/10.1242/dev.02750

    • PubMed
    • Google Scholar
72. 1. Spassky N
    2. de Castro F
    3. Le Bras B
    4. Heydon K
    5. Quéraud-LeSaux F
    6. Bloch-Gallego E
    7. Chédotal A
    8. Zalc B
    9. Thomas JL

    (2002)

    Directional guidance of oligodendroglial migration by class 3 semaphorins and netrin-1

    Journal of Neuroscience 22:5992–6004.

    • PubMed
    • Google Scholar
73. 1. Timofeev K
    2. Joly W
    3. Hadjieconomou D
    4. Salecker I

    (2012) Localized netrins act as positional cues to control layer-specific targeting of photoreceptor axons in Drosophila

    Neuron 75:80–93.

    https://doi.org/10.1016/j.neuron.2012.04.037

    • PubMed
    • Google Scholar
74. 1. Tsai HH
    2. Tessier-Lavigne M
    3. Miller RH

    (2003) Netrin 1 mediates spinal cord oligodendrocyte precursor dispersal

    Development 130:2095–2105.

    https://doi.org/10.1242/dev.00424

    • PubMed
    • Google Scholar
75. 1. Vincent S
    2. Vonesch JL
    3. Giangrande A

    (1996)

    Glide directs glial fate commitment and cell fate switch between neurones and glia

    Development 122:131–139.

    • PubMed
    • Google Scholar
76. 1. von Hilchen CM
    2. Hein I
    3. Technau GM
    4. Altenhein B

    (2010) Netrins guide migration of distinct glial cells in the Drosophila embryo

    Development 137:1251–1262.

    https://doi.org/10.1242/dev.042853

    • PubMed
    • Google Scholar
77. 1. Waltzer L
    2. Gobert V
    3. Osman D
    4. Haenlin M

    (2010) Transcription factor interplay during Drosophila haematopoiesis

    The International Journal of Developmental Biology 54:1107–1115.

    https://doi.org/10.1387/ijdb.093054lw

    • PubMed
    • Google Scholar
78. 1. West RJ
    2. Lu Y
    3. Marie B
    4. Gao FB
    5. Sweeney ST

    (2015) Rab8, POSH, and TAK1 regulate synaptic growth in a Drosophila model of frontotemporal dementia

    Journal of Cell Biology 208:931–947.

    https://doi.org/10.1083/jcb.201404066

    • PubMed
    • Google Scholar
79. 1. Wolf K
    2. Friedl P

    (2009) Mapping proteolytic cancer cell-extracellular matrix interfaces

    Clinical & Experimental Metastasis 26:289–298.

    https://doi.org/10.1007/s10585-008-9190-2

    • PubMed
    • Google Scholar
80. 1. Wolfram V
    2. Southall TD
    3. Brand AH
    4. Baines RA

    (2012) The LIM-homeodomain protein islet dictates motor neuron electrical properties by regulating K(+) channel expression

    Neuron 75:663–674.

    https://doi.org/10.1016/j.neuron.2012.06.015

    • PubMed
    • Google Scholar
81. 1. Yu TW
    2. Bargmann CI

    (2001) Dynamic regulation of axon guidance

    Nature Neuroscience 4 Suppl:1169–1176.

    https://doi.org/10.1038/nn748

    • PubMed
    • Google Scholar
82. 1. Zelhof AC
    2. Hardy RW

    (2004) WASp is required for the correct temporal morphogenesis of rhabdomere microvilli

    Journal of Cell Biology 164:417–426.

    https://doi.org/10.1083/jcb.200307048

    • PubMed
    • Google Scholar

Thank you for submitting your article "The Glide/Gcm fate determinant controls initiation of collective cell migration by regulating Frazzled" for consideration by eLife. Your article has been reviewed by three reviewers who are really familiar with the topic. They all three are positive about your work and we would be glad to consider a revised manuscript for publication in eLife. The evaluation has been overseen by a Reviewing Editor (Hugo Bellen) and a Senior Editor.

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

Summary:

Gupta et al. investigate here the function of netrins and their frazzled (Fra) and Unc5 receptors in glial cell chain migration in the Drosophila wing. Their findings show that frazzled is the major Netrin receptor facilitating initiation of glial migration, and that an early glial fate determinant TF Gcm directly activates frazzled expression. Furthermore, Netrin B but not Netrin A, apparently forms a long-range gradient that guides glia in their migration toward the CNS along peripheral nerves., and Unc5 serves as a repulsive receptor for migration of these wing glia. Transcriptional regulation of axon guidance cues and their receptors remains to be studied in detail. Therefore, this manuscript provides evidence extending our understanding of how these cues and receptors are utilized by glia and how they are transcriptionally regulated. In general, this paper appears to be one that could be of interest to the field and appropriate for eLife following revisions that address the concerns raised below.

1) The lab has previously demonstrated that Pioneer cells trigger chain migration. It would greatly improve this initial part of the paper, if the data on expression of Frazzled on migrating glial cells would be shown more convincingly. Instead of showing one confocal image in Figure 1E-G show nerves of wild type, and nerves of animals with glial or neuronal knockdown of Frazzled. There is a MiMIC insertion line available which would allow easy imaging of Frazzled protein expression (Nagarkar-Jaiswal et al., eLife, 2015).

Fra and Unc5 staining is not easy to see and compare across different conditions; in Figure 2D'-H', Figure 5C'-E', Figure 7A'-D', etc. GFP does not seem to match between different panels, such as in Figure 5C compared to C', or Figure 5D compared to D'. The white used here makes it difficult to separate individual RGB colors when these images are analyzed. It would be far easier for the reader to appreciate these results if there were separate panels for each staining in these figures, especially for Fra and UNC5 ICC. It would be even better to quantify the change in Fra staining in glia vs. neurons in the Fra-KD and GOF experiments using confocal optical sections.

2) It would be quite interesting and relevant in the light that collective migration is triggered by a few pioneers if a clonal analysis could be performed using the existing frazzled null allele. Do all cells have to express Fra? Provided that the wild type twin spot is allowed to stay, do these glial cells overtake mutant cell in front of them?

3) The observed genetic interaction between GCM and FRA is interesting. The fact that the fra³ phenotype is much worse in the background of a heterozygous rA87 insertion – which is a viable enhancer trap in the gcm promoter region and thus a very weak hypomorph – makes me wonder how the fra³ phenotype would look in trans to a gcm null allele? Is it possible to quantify the reduction of gcm mRNA in the different genetic backgrounds?

The authors use a migratory index to characterize the different genotypes. This is clear and straightforward but I wonder given the dramatic differences between 28APF and 24APF whether general developmental problems caused by the expression of the different transgenes also in many cells outside the nervous system (e.g. gcm is active also in the hemocyte lineages?) contribute to the phenotype. This could be controlled by for example comparing the time it takes to form the puparium and to eclose from the pupal cases.

4) Fra is also expressed at high levels (Figure 1F) in the axons that glia migrate along, raising the question of whether neuronal Fra influences glia migration, either directly or indirectly. What happens if Fra is knocked down specifically in neurons?

The authors claim that changes in Gcm expression lead to changes of Fra in glia. However, changes of Gcm expression in glia (using gcm-GAL4) seem to have profound effects on neuronal Fra expression. For example, in Figure 5D'-D', knockdown of Gcm using gcm-GAL4 leads to a decrease of Fra in L1, and even in L3 which is not wrapped by GFP+ processes in these images. Again, in Fra GOF (Figure 5E',E'), the Fra signal co-localizes more with red 22C10 staining than with green GFP. Is this due to leaky expression of gcm-GAL4 in neurons at earlier stages? Or, does this suggest non-cell autonomous functions of glial Gcm on neuronal Fra expression?

5) Pioneer glia vs. follower glia: during glia chain migration, pioneer glia at the tip of the chain are suggested to actively lead followers. Fra, however, appears to be detected all along the chain (Figure 1G and Figure 1—figure supplement 1), raising the question as to whether follower glia also use Fra to initiate and direct their migration independent of pioneer glia. It is technically challenging to use GAL4 drivers to differentially target pioneer or follower glia. However, have the authors considered using MARCM (Aigouy et al., 2008) to randomly target different glia in the wing and monitor their migratory behavior in either WT or Fra mutant clones? This could also address the cell autonomous vs cell-type autonomous (RNAi experiment) requirement for Fra in wing glia migration.

6) It is surprising that the NetB mutant (NetBΔ/Y, Figure 6A) has migration defects of similar severity to the Fra/+ heterozygote, and has milder effects than many other conditions, such as gcm>fra-KD and gcm>gcm-KD. This could be explained by NetB signaling through Unc5 to inhibit glial migration in Fra LOF. If this is true, the NetB mutant should enhance the Fra/+ heterozygote phenotype. Is this the case? Also, does ectopic expression of NetB in a different wing sub-region using drivers such as en-GAL4 or ap-Gal4 mis-direct glial migration?

7) In the subsection “fra plays an instructive role in L1 glia migration”, the authors claim 'Finally, we measured the migration speed at the time of initiation and it was not higher in Fra GOF glia (10 μm/h) than in control glia (40 μm/h)'. This is confusing since the numbers do appear to be significantly different. What is the actual speed of glial migration in Fra GOF vs. control? The authors should also include this important information in graphs using either the speed or duration of glial migration from initiation to completion of migration.

8) The authors try to rescue RNAi phenotypes by co-expression of RNAi and UAS-GOF constructs, stating that this 'excludes the possibility of off target effects'. This is not accurate since this rescue effect could be due to titration of siRNA away from the endogenous mRNA by overexpression of the exogenous mRNA. To directly rule out off target effects the rescue experiments should be done by using either a UAS construct bearing mutants that are insensitive to RNAi or perhaps an RNAi line that targets non coding regions.

9) Throughout the study, there are no details or explanations on the methods of quantification. For instance, Results: subsection “fra plays an instructive role in L1 glia migration”, third paragraph – this is very unclear! how exactly the authors define "initiation? and how was the quantification done? Figure 2J, how does author calculate the hAPF for initiation, costa reach, and complete migration? what is the n number? how early or late in migration do the authors consider a defect in initiation? These all have to be explained fully in the text (the Methods section or the legends). Without these knowledge, the readers will have a hard time to follow and do not understand how the conclusion was reached. Also, for the time-lapse figure, it is not clear what does 22:48 hAPF mean since there are different genotypes shown, and I cannot find any information on this in the legends nor methods.

10) Panel organization: In Figures 2, 3, 6, 7, in which the authors characterize glia migration phenotypes (MI), genotypes are assigned in different panels with the same or different time points. These arrangements make it difficult to compare different genotypes. For example, Figure 2A and B could be combined, and the same is true for Figures 3A and B and Figures 7E and F. In some panels, more genotypes should be included. For example, gcm-GAL4>gcm (GOF) is needed in Figure 3B, and NetB-GAL4> NetB needs to be added to Figure 6C. In Figure 6F and G, all three genotypes should be compared in both time points.

11) Since Gcm is a very important factor in controlling neuronal and glial cell fates, it is crucial to address clearly whether this conversion function is in any part related to the migration phenotype the authors are trying to assessing in the study. However, there is only one experiment provided which wasn't able to give a clear answer. For instance, Figure 4—figure supplement 1: it is not common to analyze the co-positive cells of gcm>GFP and Elav as Elav stained for nucleus and GFP labels the membrane. Can't the author do the costaining of Repo and Elav? this might be more clear and definitive. Also, what is the percentage of gcm KD wings not having extra Elav positive cells? Is it high or low? Isn't it plausible that a high percentage of the gcm KD wings should have extra Elav cells? The gcm KD wings with normal Elav positive cells might as well mean that gcm expression wasn't knock-down and in that case why is the MI still defective? How do the authors reconcile the discrepancies? This section has to be explained clearly since glia-neuron switch is a very important function for Gcm and whether the migration function depends on this needs to be clearly addressed with clear, precise experimental results.

[Editors' note: further revisions were requested prior to acceptance, as described below.]

Thank you for resubmitting your work entitled "The Glide/Gcm fate determinant controls initiation of collective cell migration by regulating Frazzled" for further consideration at eLife. Your revised article has been evaluated by a Senior editor, a Reviewing editor, Hugo Bellen, and three reviewers.

The manuscript has been improved but there are some remaining issues that need to be addressed. The reviewers discussed their reviews on line and came to a consensus that they all agreed with reviewer 3. The reviewers feel that some of the points that they raised where not addressed and Reviewer #3 summarized these concerns.

Reviewer #3:

I read the revised paper and cannot support publication at this stage due to the poor demonstration that fra is expressed by the glia and the lack of a clear demonstration that the phenotype is due to loss of glial fra function. I admit there are some hints supporting this suggestion but I do not think the data are conclusive.

Expression of fra in glia. The authors added a new Figure 1H-J which shows fra expression upon glial knockdown. For me this is not showing that fra is expressed in glia. There is as much background as in the wild type and the control experiment, neuronal knockdown, is not shown based on the argument that elav is also expressed by glia. Please, use nsybGal4 which is specific – and/or generate clones! The quality of the antibody stainings is not consistent and for example in Figure 2E glial knockdown is also removing the neuronal Fra expression. For me the quality of the data is still simply not good enough.

The phenotypes are based on somewhat odd genetic backgrounds. May be there is a fra haploinsufficiency and heterozygous fra/+ animals show a phenotype. But to really determine the Fra phenotype, a true loss of function analysis is required. The argument that MARCM clones are hard to get is weird as the same lab has generated these clones before in another study. Glia-specific fra knockdown animals show roughly the same phenotype as heterozygous fra3/+ animals (MI 58 vs.61). Still only the RNAi induced phenotype can be completely rescued. The authors conclude that this may be due to the reduction of neuronal Fra expression (which was not tested). However, one could also argue that the authors are looking at a genetic background effect. Thus, a MARCM analysis is mandatory.

We had asked for the effect of gcm on the migratory phenotype of fra3/+ heterozygous animals. Although the authors provide arguments for an allelic series of the different gcm alleles they do not show the effect of a true gcm loss of function allele on the migratory index. Provided that the presence of one very weak, homozygous viable, gcm allele has such a profound effect on the migratory index, what would a true null allele do? The allelic series shown in the rebuttal letter indicates that homozygous gcm>GFP animals have the same effect as gcm>GFP gcmN7-4. Since gcmN7-4 is a null this implies that gcm>GFP is a null as well – but why do they survive then? Something is odd here.

[…]

1) The lab has previously demonstrated that Pioneer cells trigger chain migration. It would greatly improve this initial part of the paper, if the data on expression of Frazzled on migrating glial cells would be shown more convincingly. Instead of showing one confocal image in Figure 1E-G show nerves of wild type, and nerves of animals with glial or neuronal knockdown of Frazzled. There is a MiMIC insertion line available which would allow easy imaging of Frazzled protein expression (Nagarkar-Jaiswal et al., eLife, 2015).

Fra and Unc5 staining is not easy to see and compare across different conditions; in Figure 2D'-H', Figure 5C'-E', Figure 7A'-D', etc. GFP does not seem to match between different panels, such as in Figure 5C compared to C', or Figure 5D compared to D'. The white used here makes it difficult to separate individual RGB colors when these images are analyzed. It would be far easier for the reader to appreciate these results if there were separate panels for each staining in these figures, especially for Fra and UNC5 ICC. It would be even better to quantify the change in Fra staining in glia vs. neurons in the Fra-KD and GOF experiments using confocal optical sections.

As suggested by the referees we made sure that the labeling was indeed less prominent in the KD and more intense in the overexpressing animals. We have now included the data that also shows the glial knockdown of fra in a 18 hAPF old wing in Figure 1H-J and we have included panels that only shows Fra or Unc5 labeling in Figure 2C'-G'”, Figure 5C‐E”' and Figure 7A‐D”'.

The difference in GFP intensity in the merged panels is due to the strong red signal. Specifically, at late stages these differences are due to the fact that glia wrap neurons (which are in red) completely. In addition, we have quantified Fra staining in Figure 1E-J between control and fra KD L1 glia by including heat maps. Note that in the KD the neuronal (but not the glial) Fra signal is as high if not higher than in the wt.

Concerning the MiMIC insertion, we tried and stay consistent, using the same antibody that was used in previous papers, in particular in the paper on glial migration from Altenhein and collaborators (References on the used antibody: Von Hilchen et al., 2010; Timofeeev et al., 2012; Joly et al., 2007; Kolodziej et al., 1996; Orr et al., 2014; Gong et al., 1999; Pert et al., 2015).

2) It would be quite interesting and relevant in the light that collective migration is triggered by a few pioneers if a clonal analysis could be performed using the existing frazzled null allele. Do all cells have to express Fra? Provided that the wild type twin spot is allowed to stay, do these glial cells overtake mutant cell in front of them?

We see Fra expression in all L1 glia, which is shown in Figure 1—figure supplement 1(subsection “Frazzled expression in the glia of the developing Drosophila wing”, second paragraph). The clonal analysis is a valuable point, however we think that the data from this experiment will not be easy to interpret. Fra is also expressed in neurons, therefore getting clones that only affect glia will be difficult (low division rate) and even more to get clones that specifically affect the pioneer cells. We did in the past get MARCM clones in pioneer cells, but this is a very rare event. This was useful to analyze the morphology of pioneers, but the analysis of migration phenotypes requires high number of events in order to perform the statistics. More importantly, we have previously shown that several pioneers are at work, therefore mutating one or two of them may not be sufficient to see a phenotype.

3) The observed genetic interaction between GCM and FRA is interesting. The fact that the fra³ phenotype is much worse in the background of a heterozygous rA87 insertion – which is a viable enhancer trap in the gcm promoter region and thus a very weak hypomorph – makes me wonder how the fra³ phenotype would look in trans to a gcm null allele? Is it possible to quantify the reduction of gcm mRNA in the different genetic backgrounds?

The authors use a migratory index to characterize the different genotypes. This is clear and straightforward but I wonder given the dramatic differences between 28APF and 24APF whether general developmental problems caused by the expression of the different transgenes also in many cells outside the nervous system (e.g. gcm is active also in the hemocyte lineages?) contribute to the phenotype. This could be controlled by for example comparing the time it takes to form the puparium and to eclose from the pupal cases.

Concerning the interaction, gcm^rA87is a weaker allele than the gcm‐Gal4 line and, accordingly gcm^rA87/fra³ has a weaker migratory phenotype that gcm‐Gal4/ fra³. These data are now shown together in Figure 3, to comply with reviewers’ suggestion. For the benefit of the reviewers, we also show here below the rate of conversion and/or the number of repo positive cells in different allelic combinations (gcm-Gal4/ gcm^rA87, gcm-Gal4/ gcm-Gal4, gcm^rA87/ gcmN7-4 and gcm‐Gal4>gcm RNAi). The inverse correlation between the two phenotypes in the different backgrounds further sustains our hypothesis. We believe that the phenotypic analysis of the allelic series is more sensitive than RNA quantification, given the low number of gcm expressing cells. For the sake of simplicity, we do not think that the data here below should be included. Should the referee feel it necessary, however, we are eager to include them in Figure 5—figure supplement 1.

We did not observe any general developmental defect, such as delay in eclosion from the pupal case.

Author response image 1

Download asset Open asset

To rule out a possible hemocyte contribution in glia migration we downregulated gcm using a hemocyte driver collagen‐Gal4 and found no effect on glia migration. Further, we also knocked down fra using a temperature sensitive Gal80 construct of gcm driver and specifically knocked down fra post embryonically. Since gcm is only expressed in blood cells embryonically, this eliminates any possible contribution of hemocyte Gcm expression to the observed phenotypes. These analyses confirm that gcm and fra directly affect glia migration as in these conditions too glia migration is significantly delayed (Figure 5—figure supplement 1D and E).

4) Fra is also expressed at a high levels (Figure 1F) in the axons that glia migrate along, raising the question of whether neuronal Fra influences glia migration, either directly or indirectly. What happens if Fra is knocked down specifically in neurons?

The authors claim that changes in Gcm expression lead to changes of Fra in glia. However, changes of Gcm expression in glia (using gcm-GAL4) seem to have profound effects on neuronal Fra expression. For example, in Figure 5D'-D', knockdown of Gcm using gcm-GAL4 leads to a decrease of Fra in L1, and even in L3 which is not wrapped by GFP+ processes in these images. Again, in Fra GOF (Figure 5E', E'), the Fra signal co-localizes more with red 22C10 staining than with green GFP. Is this due to leaky expression of gcm-GAL4 in neurons at earlier stages? Or, does this suggest non-cell autonomous functions of glial Gcm on neuronal Fra expression?

The neuronal expression of Fra is an important issue and we thank the reviewer(s) for the comment. We analyzed L1 axons of 17 hAPF fra³ wings and found that about 30% of them show a transient phenotype of delayed axonal growth (see now Figure 3—figure supplement 2B-D). We also analyzed wings where fra is knocked down using an Elav driver and got a very strong migratory phenotype (Data not shown in the manuscript). However, the available Elav drivers (we tested two of them) are also expressed in wing glia and epithelial cells (Author response image 2). Other known drivers are either expressed in neuronal subsets (motoneurons) or too late for our purpose. Thus, in the wing, we were unable to use a driver that specifically activates or knocks down gene expression in neurons. These analyses suggested that downregulating fra in neurons may indirectly affect glia migration and prompted us to use a driver that specifically affects glia.

Author response image 2

Download asset Open asset

We think there is no change in the neuronal expression of Fra upon gcm KD or GOF. Figure 5D‐D’” is only one example and the decrease might be due to the fact that these images are few sections and not the whole projection. Concerning the role of Gcm on neuronal expression of Fra. We never observed the Gcm driver being expressed in neurons. For all the above reasons, we do not think there is a non-cell autonomous effect.

5) Pioneer glia vs. follower glia: during glia chain migration, pioneer glia at the tip of the chain are suggested to actively lead followers. Fra, however, appears to be detected all along the chain (Figure 1G and Figure 1—figure supplement 1), raising the question as to whether follower glia also use Fra to initiate and direct their migration independent of pioneer glia. It is technically challenging to use GAL4 drivers to differentially target pioneer or follower glia. However, have the authors considered using MARCM (Aigouy et al., 2008) to randomly target different glia in the wing and monitor their migratory behavior in either WT or Fra mutant clones? This could also address the cell autonomous vs cell-type autonomous (RNAi experiment) requirement for Fra in wing glia migration.

The reviewers raise an interesting point, i.e. what is the role of Fra, since it is all along the L1 chain. We believe that the suggested MARCM clonal analysis is beyond the scope of this manuscript. We indeed used MARCM to follow the morphology of single pioneers but mutant clones on single pioneer glial cells are very rare, which makes it hard to draw conclusions that are statistically significant. If we go for big clones, they will also affect neurons, which will not help assessing the cell autonomous effect. See also answer to Q2.

To comply with the comment, we asked whether the ligand Net B accumulates on specific glia, based on the recent paper from Pert et al., (Biology Open 2015) showing Net B accumulation in Fra expressing cells. Immunolabeling assays on young wings indeed show preferential Net B accumulation in some cells at the front of the chain. We speculate that the receptor is all over the glia but pioneer glia are closer to the ligand than the other glia and some pioneers accumulate it, perhaps due to receptor‐mediated endocytosis. This may in turn triggers the ‘promigratory’ downstream pathway in those cells.

Author response image 3

Download asset Open asset

6) It is surprising that the NetB mutant (NetBΔ/Y, Figure 6A) has migration defects of similar severity to the Fra/+ heterozygote, and has milder effects than many other conditions, such as gcm>fra-KD and gcm>gcm-KD. This could be explained by NetB signaling through Unc5 to inhibit glial migration in Fra LOF. If this is true, the NetB mutant should enhance the Fra/+ heterozygote phenotype. Is this the case? Also, does ectopic expression of NetB in a different wing sub-region using drivers such as en-GAL4 or ap-Gal4 mis-direct glial migration?

We thank the reviewer(s) for the thoughtful comments. The reviewer is correct, the NetB phenotype might be milder than that observed in fra3/+ because of the effects of NetB on unc5. Along this line, Figure 8—figure supplement 1C shows that UAS‐RNAi for both unc5 and fra has a weaker phenotype than UAS‐fra‐RNAi alone. As per the double mutant NetBΔ/Y; fra³/+, this genotype would not inform on the role of Unc5, because in a hemizygous NetBΔ/Y animal both Fra and Unc5 should be absent anyway, therefore removing the fra receptor would not make any difference. We thought that this experiment may nevertheless be informative and tell whether fra is also submitted to other signals, as suggested by recent studies (Neuhaus-Follini et al; 2015). We hence performed the suggested experiment and it does seem that fra also depends on other signals, see data on the right graph. We would not include this data in the main text, however, as they go beyond the scope of the manuscript, which is to show the impact of a fate determinant on late aspects of cell specification. As requested, we show the data on the ectopic expression of NetB using en‐Gal4 in the posterior wing compartment and using GMR 29F05-Gal4 that is expressed in the distal part of the wing. These experiments show that ectopic NetB does not affect glia migration (Now Figure 8—figure supplement 1A). Thus, NetB does not act as a guidance molecule triggering directionality. This is also discussed in the manuscript on page 23.

Author response image 4

Download asset Open asset

7) In the subsection “fra plays an instructive role in L1 glia migration”, the authors claim 'Finally, we measured the migration speed at the time of initiation and it was not higher in Fra GOF glia (10 μm/h) than in control glia (40 μm/h)'. This is confusing since the numbers do appear to be significantly different. What is the actual speed of glial migration in Fra GOF vs. control? The authors should also include this important information in graphs using either the speed or duration of glial migration from initiation to completion of migration.

To comply with the comments and clarify our statements, we performed further analyses that are now incorporated in Figure 3—figure supplement 2. The time-lapse analysis shows that fra GOF glia start moving earlier than control glia (Figure 2I, Figure 3—figure supplement 1E). The gcm>fra GOF glia phenotype is solely due to altered initiation rather than to acceleration and we have added a speed graph of control and fra GOF glia in Figure 3—figure supplement 2A. The speed analysis was done on the most proximal glial cell of the chain (μm/h, y-axis; hAPF, x‐axis). The distance covered by the front cell was measured by analyzing the position of the glial soma every hour, as already shown in Kumar et al., 2015. We followed the cell starting from the point of initiation till complete migration. The migratory speed increases both in control and fra GOF glia until they reach the level of costa and then it decreases, however, fra GOF glia display a more uniform speed than that of control glia: they accelerate less in the early phases and slow down less in the late phases.

8) The authors try to rescue RNAi phenotypes by co-expression of RNAi and UAS-GOF constructs, stating that this 'excludes the possibility of off target effects'. This is not accurate since this rescue effect could be due to titration of siRNA away from the endogenous mRNA by overexpression of the exogenous mRNA. To directly rule out off target effects the rescue experiments should be done by using either a UAS construct bearing mutants that are insensitive to RNAi or perhaps an RNAi line that targets non coding regions.

We take the point made by the reviewer. However, we would like to emphasize that the RNAi data are further supported by the opposite phenotype obtained in gain of function experiments and by the fact that we could rescue the fra³mutant phenotype by reintroducing a UAS‐fra transgene in glia (Figure 3—figure supplement 1A). We nevertheless played down the interpretation of the RNAi data in the main text, to comply with the comment.

9) Throughout the study, there are no details or explanations on the methods of quantification. For instance, Results: subsection “fra plays an instructive role in L1 glia migration”, third paragraph – this is very unclear! how exactly the authors define "initiation? and how was the quantification done? Figure 2J, how does author calculate the hAPF for initiation, costa reach, and complete migration? what is the n number? how early or late in migration do the authors consider a defect in initiation? These all have to be explained fully in the text (the Methods section or the legends). Without these knowledge, the readers will have a hard time to follow and do not understand how the conclusion was reached. Also, for the time-lapse figure, it is not clear what does 22:48 hAPF mean since there are different genotypes shown, and I cannot find any information on this in the legends nor methods.

As the referee suggests, we included a detailed description of how the time-laps analysis was performed and how the quantification was made in the Methods section of the manuscript (in vivoimaging paragraph). These methods were already defined in our previous publications (Berzseny et al., 2011, Kumar et al., 2015).

“Based on the time-lapse videos, the migratory process has been subdivided in three phases: the earliest one describes migration initiation around 18 hAPF; this time point defines the movement of the soma of the first cell at the front of the chain. […] In order to quantify the migratory defects, the differences in time between the control and mutant wings were compared and the graphs were plotted using the standard Student's t test method.”

We have used the above phases to select the images from the time-lapsevideos. Typically, Figure 2J, Kshows a snap shot at 21h48, which corresponds to the time by which wild type glia have reached the Costa, in average. Figure 3D-F,shows a snap shot at 22h48, which corresponds to the time by which GOF glia have completed migration, in average. We have clarified this in the text.

10) Panel organization: In Figures 2, 3, 6, 7, in which the authors characterize glia migration phenotypes (MI), genotypes are assigned in different panels with the same or different time points. These arrangements make it difficult to compare different genotypes. For example, Figure 2A and B could be combined, and the same is true for Figures 3A and B and Figures 7E and F. In some panels, more genotypes should be included. For example, gcm-GAL4>gcm (GOF) is needed in Figure 3B, and NetB-GAL4> NetB needs to be added to Figure 6C. In Figure 6F and G, all three genotypes should be compared in both time points.

As per referee’s suggestion we have reorganized Figure 2A, B; Figure 3A, B and Figure 7E, Finto one panel. We have also included the migratory index of gcm>gcm GOF in Figure 3Aand the migratory index of NetB>NetB GOF in Figure 6C. Finally, as requested, we have also included all the genotypes in Figure 6F and G.

11) Since Gcm is a very important factor in controlling neuronal and glial cell fates, it is crucial to address clearly whether this conversion function is in any part related to the migration phenotype the authors are trying to assessing in the study. However, there is only one experiment provided which wasn't able to give a clear answer. For instance, Figure 4—figure supplement 1: it is not common to analyze the co-positive cells of gcm>GFP and Elav as Elav stained for nucleus and GFP labels the membrane. Can't the author do the costaining of Repo and Elav? this might be more clear and definitive. Also, what is the percentage of gcm KD wings not having extra Elav positive cells? Is it high or low? Isn't it plausible that a high percentage of the gcm KD wings should have extra Elav cells? The gcm KD wings with normal Elav positive cells might as well mean that gcm expression wasn't knock-down and in that case why is the MI still defective? How do the authors reconcile the discrepancies? This section has to be explained clearly since glia-neuron switch is a very important function for Gcm and whether the migration function depends on this needs to be clearly addressed with clear, precise experimental results.

Author response image 5

Download asset Open asset

To address this issue we included the co-‐labeling of Repo and Elav in (Figure 5—figure supplement 1A-B”), as requested. Concerning the possible indirect effects of glia to neuron conversion on migration, we find that only 14% of the gcm KD wings show conversion, whereas 75% of the KD wings show delayed migration. Further, to make sure that glia to neuron transformation in gcm knockdown wings does not affect glia migration we analyzed two sets of KD wings. The first set comprises of wings with fate transformation and the migratory index of these wings is very low, 25%. The second set includes wings without fate transformation; in this case, the migratory index is 45%, which is still significantly different from the control (Figure 5—figure supplement 1C). Last but not the least, the number of converted cells is very low (10). We also analyzed an allelic combination that is weaker than the KD and that only shows fate conversion in 10% of the wings. We still find a migratory delay in these wings (gcm‐Gal4/gcm^rA87animals, See Figure 3A). This analysis confirms that Gcm affects glia migration irrespective of the change in the number of glial cells due to the fate transformation, which altogether fits perfectly with the finding that Gcm directly induces the expression of the Fra receptor.

[Editors' note: further revisions were requested prior to acceptance, as described below.]

[…]

Reviewer #3:

I read the revised paper and cannot support publication at this stage due to the poor demonstration that fra is expressed by the glia and the lack of a clear demonstration that the phenotype is due to loss of glial fra function. I admit there are some hints supporting this suggestion but I do not think the data are conclusive.

Concerning the detailed points described in, we addressed the three issues, as previously agreed:

1) Better evidence for Fra expression/requirement in glia

2) Migratory assay in fra4/gcm null wings

3) Mitotic clones

Expression of fra in glia. The authors added a new Figure 1H-Jwhich shows fra expression upon glial knockdown. For me this is not showing that fra is expressed in glia. There is as much background as in the wild type and the control experiment, neuronal knockdown, is not shown based on the argument that elav is also expressed by glia. Please, use nsybGal4 which is specific – and/or generate clones! The quality of the antibody stainings is not consistent and for example in Figure 2E glial knockdown is also removing the neuronal Fra expression. For me the quality of the data is still simply not good enough.

We followed the suggestion and used the nsybGal4 line as a neuronal driver. We first assessed expression of this driver in wing neurons, next we used the nsybGal4 line to knock down Fra in neurons and analyze glial migration. We found no defects in the migratory index, except in wings that carry an axonal navigation defect. These data are now presented in Figure 3—figure supplement 1. The wings carrying axon navigation defects were of course eliminated from the graph, as it has already been published that axonal defects result in glial migration defects.

We add Fra immunolabelling at 18 hAPF in combination with a second glial reporter (Repo>GFP). See Figure 1.

We removed panels H-J from Figure 1, as requested.

We improved the quality of Figure 2 showing glial Fra expression at later stages, as requested.

The phenotypes are based on somewhat odd genetic backgrounds. May be there is a fra haploinsufficiency and heterozygous fra/+ animals show a phenotype. But to really determine the Fra phenotype, a true loss of function analysis is required. The argument that MARCM clones are hard to get is weird as the same lab has generated these clones before in another study. Glia-specific fra knockdown animals show roughly the same phenotype as heterozygous fra3/+ animals (MI 58 vs.61). Still only the RNAi induced phenotype can be completely rescued. The authors conclude that this may be due to the reduction of neuronal Fra expression (which was not tested). However, one could also argue that the authors are looking at a genetic background effect. Thus, a MARCM analysis is mandatory.

We complied with the request of the referee and analyzed the migratory phenotype in transheterozygous animals fra3/gcm null (N7-4), which further confirms the data obtained with the hypomorphs and the KD line. These data are now presented in Figure 3—figure supplement 2, second last column.

We would like to clarify that the allelic series showing the reduction in the number of Repo+ cells in the last rebuttal letter shows the effects of gcm^rA87/gcm^N7-4and not gcm>GFP/ gcm^N7-4, as claimed by the referee (see the same figure panel). We have already mentioned in the manuscript that gcm^rA87 is a weak gcm hypomorphic allele. Therefore, the effect it produces on the number of repo cells when combined with null gcm^N7-4 allele is similar to gcm>GFP/gcm>GFP.

We had asked for the effect of gcm on the migratory phenotype of fra3/+ heterozygous animals. Although the authors provide arguments for an allelic series of the different gcm alleles they do not show the effect of a true gcm loss of function allele on the migratory index. Provided that the presence of one very weak, homozygous viable, gcm allele has such a profound effect on the migratory index, what would a true null allele do? The allelic series shown in the rebuttal letter indicates that homozygous gcm>GFP animals have the same effect as gcm>GFP gcmN7-4. Since gcmN7-4 is a null this implies that gcm>GFP is a null as well – but why do they survive then? Something is odd here.

We performed mitotic clone analyses. We ordered the strain to produce MARCM clones in WT or in Fra KD cells. We also ordered strains to perform a more refined clonal analysis using the recently developed coinFLP technique in a wild type and in a Fra KD background. The crosses from this second approach went faster. Given the time frame we had agreed upon and the fact that the PhD student could only come for one month, we pursued the analyses of the coinFLP clones, which was indeed very successful. We could identify wild type and mutant clones in the same wing using two membrane reporters (GFP- green cells: WT cells; RFP – red cells: Fra DK cells). In this way, we could confirm that red (Fra KD) cells contain much less Fra than green (WT) cells. Furthermore, we could specifically identify WT and Fra KD glia using the Repo marker that is detected in glial nuclei. All these data are now shown in Figure 1.

Author details

1. Tripti Gupta

   1. Department of Functional Genomics and Cancer, Institut de Génétique et de Biologie Moléculaire et Cellulaire, Illkirch, France
   2. Centre National de la Recherche Scientifique, UMR7104, Illkirch, France
   3. Institut National de la Santé et de la Recherche Médicale, U964, Illkirch, France
   4. Université de Strasbourg, Illkirch, France

   Contribution

   TG, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article, Contributed unpublished essential data or reagents

   Competing interests

   No competing interests declared.

2. Arun Kumar

   1. Department of Functional Genomics and Cancer, Institut de Génétique et de Biologie Moléculaire et Cellulaire, Illkirch, France
   2. Centre National de la Recherche Scientifique, UMR7104, Illkirch, France
   3. Institut National de la Santé et de la Recherche Médicale, U964, Illkirch, France
   4. Université de Strasbourg, Illkirch, France

   Present address

   Department of Entomology, University of California, Riverside, United States

   Contribution

   AK, Conception and design, Acquisition of data, Contributed unpublished essential data or reagents

   Competing interests

   No competing interests declared.

3. Pierre B. Cattenoz

   1. Department of Functional Genomics and Cancer, Institut de Génétique et de Biologie Moléculaire et Cellulaire, Illkirch, France
   2. Centre National de la Recherche Scientifique, UMR7104, Illkirch, France
   3. Institut National de la Santé et de la Recherche Médicale, U964, Illkirch, France
   4. Université de Strasbourg, Illkirch, France

   Contribution

   PBC, Helped finalizing the revised version of the manuscript, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

   Competing interests

   No competing interests declared.

4. K VijayRaghavan

   1. Department of Developmental Biology and Genetics, Tata Institute for Fundamental Research, Bangalore, India
   2. National Centre for Biological Sciences, Tata Institute for Fundamental Research, Bangalore, India

   Contribution

   KV, Hepled analyzing the adat anad suggested several experiments for the project, Analysis and interpretation of data

   Competing interests

   KV: Senior editor, eLife

   "This ORCID iD identifies the author of this article:" 0000-0002-4705-5629

5. Angela Giangrande

   1. Department of Functional Genomics and Cancer, Institut de Génétique et de Biologie Moléculaire et Cellulaire, Illkirch, France
   2. Centre National de la Recherche Scientifique, UMR7104, Illkirch, France
   3. Institut National de la Santé et de la Recherche Médicale, U964, Illkirch, France
   4. Université de Strasbourg, Illkirch, France

   Contribution

   AG, Conception and design, Analysis and interpretation of data, Drafting or revising the article

   For correspondence

   angela@igbmc.fr

   Competing interests

   No competing interests declared.

   "This ORCID iD identifies the author of this article:" 0000-0001-6278-5120

• 1,076

  Page views

• 300

  Downloads

• 4

  Citations

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