m e c h a n i s m s o f d e v e l o p m e n t 1 3 7 ( 2 0 1 5 ) 1 –1 0
Available at www.sciencedirect.com
ScienceDirect j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / m o d o
The extracellular protease stl functions to inhibit migration of v’ch1 sensory neuron during Drosophila embryogenesis Tashi Lhamo, Afshan Ismat * Department of Biology, Franklin and Marshall College, P.O. Box 3003, Lancaster, PA 17604-3003, USA
A R T I C L E
I N F O
A B S T R A C T
Article history:
Proper migration of cells through the dense and complex extracellular matrix (ECM) re-
Received 22 September 2014
quires constant restructuring of the ECM to allow cells to move forward in a smooth manner.
Received in revised form
This restructuring can occur through the action of extracellular enzymes. Among these ex-
14 April 2015
tracellular enzymes is the ADAMTS (A Disintegrin And Metalloprotease with ThromboSpondin
Accepted 27 April 2015
repeats) family of secreted extracellular proteases. Drosophila stl encodes an ADAMTS pro-
Available online 4 May 2015
tease expressed in and around the peripheral nervous system (PNS) during embryogenesis. The absence of stl displayed one specific neuron, the v’ch1 sensory neuron, migrating to
Keywords:
its target sooner than in wild type. During normal development, the v’ch1 sensory neuron
Cell migration
migrates dorsally at the same time it is extending an axon ventrally toward the CNS. Sur-
ADAMTS
prisingly, in the absence of stl, the v’ch1 neuron migrated further dorsally as compared to
PNS
the wild type at stage 15, but did not migrate past its correct target at stage 16, suggesting
stl
a novel role for this extracellular protease in inhibiting migration of this neuron past a certain
Drosophila
point.
Sensory neurons
1.
Introduction
Cell migration is an essential process during development of multicellular organisms. Cells can migrate in different ways; as individuals like primordial germ cells (Kunwar et al., 2006), or as a multicellular collective like the salivary gland of Drosophila melanogaster or vertebrate mammary gland (Andrew and Ewald, 2010; Kerman et al., 2006). In addition, the entire cell can migrate, or only a part of the cell can migrate, as is the case with many neurons where there is axonal outgrowth (Aman and Piotrowski, 2010; Dickson and Gilestro, 2006; Evans and Bashaw, 2010; Jang et al., 2007). Despite all these differences in types of migration, all migrating cells, or parts of cells, have very specific targets. Two questions to think about are what
© 2015 Elsevier Ireland Ltd. All rights reserved.
types of signals are migrating cells receiving along their path in order to reach their target, and how do migrating cells know when to stop migrating? Although we have gained much insight into these questions, much more remains unknown. Therefore, using a simpler model system like Drosophila may be able to help us tackle these questions. Migrating cells require cues, both attractive and inhibitory, from their environment to guide them to the correct target (Bradley et al., 2003; Kunwar et al., 2006; Tessier-Lavigne and Goodman, 1996). Attractants have been well studied, including the growth factor PDGF that signals from the oocyte to guide border cells toward the oocyte (Duchek et al., 2001; McDonald et al., 2003), the FGF bnl expressed in the mesoderm that guides tracheal cells (Reichman-Fried et al., 1994; Sutherland et al., 1996), the FGFs pyr and ths guiding both early
* Corresponding author. Department of Biology, Franklin and Marshall College, P.O. Box 3003, Lancaster, PA 17604-3003, USA. Tel.: +1 717 358 6301; fax: +1 717 358 4548. E-mail address: [email protected] (A. Ismat). http://dx.doi.org/10.1016/j.mod.2015.04.004 0925-4773/© 2015 Elsevier Ireland Ltd. All rights reserved.
2
m e c h a n i s m s o f d e v e l o p m e n t 1 3 7 ( 2 0 1 5 ) 1 –1 0
mesoderm cells and caudal visceral mesoderm (CVM) cells (Kadam et al., 2012; Reim et al., 2012; Stathopoulos et al., 2004), and primordial germ cells in Zebrafish responding to the chemokine SDF-1a (Doitsidou et al., 2002; Reichman-Fried et al., 2004). As attractants function to guide cells toward a signal to keep cells on the correct path, inhibitory cues are also functioning to guide these migrating cells away from the incorrect path. Therefore, repelling cells from a certain area can be as important as guiding cells toward a default path. For example, the lipid phosphates wunen and wunen2 are required to repel germ cells away from the ventral midgut and the CNS so that the germ cells stay on a strict path toward the gonad (Kunwar et al., 2006; Renault et al., 2004; Starz-Gaiano et al., 2001; Zhang et al., 1997). During neural crest cell migration, class 3 Semaphorins and Eph/ephrin signaling are both used as avoidance cues for different streams of neural crest cells as they migrate toward their correct target (Mayor and Theveneau, 2013). The embryonic nervous system in Drosophila is a very wellstudied model for examining how attractive and inhibitory cues function during cell migration. During axon outgrowth, in both the central nervous system (CNS) and peripheral nervous system (PNS), the axon travels to its target to make synaptic connections. Just like other types of cell migration, axons use attractive and inhibitory cues to guide them along the right path. Moreover, these attractive and inhibitory cues can work both in short range and long range (Tessier-Lavigne and Goodman, 1996). Semaphorins function as repulsive cues during motor axon guidance (Bates and Whitington, 2007; Van Vactor and Lorenz, 1999; Yu et al., 1998). The ligand for Robo receptors, Slit, is secreted by midline glial cells and inhibits longitudinal axons in the Drosophila CNS (Kidd et al., 1999; Simpson et al., 2000). Slit-Robo signaling has been shown to play a role in the PNS where Robo inhibits lateral cluster sensory axons from exploring aberrant pathways (Parsons et al., 2003). The question we are interested in is how these extracellular factors are presented to the migrating cells to influence their migration in a spatiotemporal manner. Most cells migrate through a dense three-dimensional extracellular matrix (ECM) composed of a complex mixture of proteins and macromolecules. The only way cells can change their shape, rearrange themselves, and eventually migrate is for this dense, complex matrix to get restructured and remodeled by certain proteases that function within the ECM. ADAMTS (A Disintegrin And Metalloprotease with ThromboSpondin motifs) extracellular protease are one family of secreted proteases that function within the ECM to degrade certain ECM components, thereby allowing for various cellular functions, including cell migration (Apte, 2009; Nandadasa et al., 2014; Wagstaff et al., 2011). Traditionally, it has been thought that extracellular proteases functioning in cell migration have one job, to clear a path in the ECM at the leading edge of a migrating cell to allow the cell room to move forward. Recently, it has been shown that the function of these enzymes extends past this one role. The ADAM protease Kuzbanian cleaves the Robo receptor to activate repulsion of axons along the CNS midline (Coleman et al., 2010; Schimmelpfeng et al., 2001). Additionally, Drosophila AdamTS-A functions to detach cells from the trailing edge, thereby allowing cells to move forward in a smooth manner (Ismat et al., 2013).
In this study, we have shown that a second Drosophila ADAMTS extracellular protease stall (stl) functions in the PNS to inhibit migration of one sensory neuron. Previously, stl was shown to play a role in ovarian development, where loss of stl caused an overgrowth of ovarian tissue (Ozdowski et al., 2009; Willard et al., 2004). Here we show that stl mRNA is expressed in the vicinity of the v’ cluster neuron (v’ch1), and that loss of stl causes the v’ch1 neuron to migrate to its target dorsal to the lch5 chordotonal organ cluster sooner than its wildtype (WT) counterparts. Therefore, this study suggests a novel and unexpected role for extracellular proteases in presenting inhibitors for cell migration.
2.
Experimental methods
2.1.
Drosophila strains
The following Drosophila strains were used: Canton S (CS) (Bloomington Stock Center, Indiana University), stlph57, stlpa49, stlawk26 (Ozdowski et al., 2009) (from T. Schupbach, Princeton University, USA), stla16 (Ozdowski et al., 2009) (from C. Cronmiller, University of Virginia, USA). All embryos were collected at 25 °C. All stl alleles were balanced over the SM6 evelacZ balancer, and stl mutant embryos were identified by the absence of β-gal.
2.2.
Staining procedures
Antibody stainings, fluorescent double stainings for proteins and mRNA, and in situ hybridizations were performed as described (Azpiazu and Frasch, 1993; Knirr et al., 1999; Reuter and Scott, 1990). The following antibodies were used: rabbit polyclonal αnti-βgal (1:3000, MP Biochemical), mouse anti22C10 (1:10, Developmental Studies Hybridoma Bank [DSHB]), rat αnti-Elav (1:10, DSHB), mouse αnti-Repo (1:10, DSHB), mouse αnti-β-gal (1:500, Promega), sheep αnti-digoxigenin (DIG) (1:1000, Molecular Probes). The following digoxigenin-labeled RNA probe was used: stl (T7 RNA polymerase transcript of EST AT15733). Secondary biotinylated αnti-mouse, biotinylated αnti-rabbit, biotinylated αnti-sheep, goat αnti-mouse-488, goat αnti-rat488, goat α-rat568 were used. Brightfield DIC images were taken on a Leica DMRB microscope using a ProgRes C5 color camera and ProgRes image capture software (Franklin & Marshall College). Fluorescent images were taken with a Leica SP2 laser scanning confocal microscope (Penn State College of Medicine, Hershey, PA).
2.3.
Statistical analysis
A G-test of statistical independence was used to compare all three classes (dorsal, inside, ventral) of v’ch1 sensory neuron position at each stage (15, 16, and 17).
3.
Results
3.1. The v’ch1 neuron migrates dorsally during embryogenesis All the sensory neurons in the abdominal segments have been identified, and a majority have previously been mapped
m e c h a n i s m s o f d e v e l o p m e n t 1 3 7 ( 2 0 1 5 ) 1 –1 0
out spatiotemporally (Bates and Whitington, 2007; Hartenstein, 1988; Parsons et al., 2003). Most sensory neurons in the Drosophila PNS start out in a dorsal position in the embryo, and send out an axon ventrally to eventually meet the CNS. However, in the case of the v’ch1 sensory neuron, the cell body migrates dorsally at the same time it is extending an axon ventrally toward the CNS (Bates and Whitington, 2007; Hartenstein, 1988; Parsons et al., 2003). Although previous work has described v’ch1 migration (Bates and Whitington, 2007; Kaminker et al., 2001; Mrkusich et al., 2010; Parsons et al., 2003), we decided to focus on the dorsal migration, and position, of this v’ch1 neuron in relation to the lch5 chordotonal organ cluster between stages 15 and 17 of embryogenesis. Wild type (WT) embryos labeled with anti-22C10 antibody clearly show the position of the v’ch1 neuron (Fig. 1B–D, pseudo-colored in blue) in relation to the lch5 chordotonal organ cluster (Fig. 1B–D, white arrowhead). At stage 15, the v’ch1 neuron is in a position ventral
3
to the lch5 chordotonal organ cluster (Fig. 1B). At stage 16, the v’ch1 neuron has migrated to a position inside the lch5 organ cluster (Fig. 1C). By the end of embryogenesis at stage 17, the v’ch1 neuron has migrated to its final position dorsal to the lch5 organ cluster (Fig. 1D).
3.2.
stl encodes an ADAMTS extracellular protease
The human genome encodes 19 members of the ADAMTS family of metalloproteases (Apte, 2009). The Drosophila genome encodes three ADAMTS metalloproteases, AdamTS-A (Ismat et al., 2013), CG4096, and stall (stl) (Ozdowski et al., 2009; Willard et al., 2004). Like all other ADAMTSs, stl is secreted, and contains a signal peptide, a prodomain, a Zn-dependent metalloprotease domain, a disintegrin-like domain, thrombospondin-type 1 repeats, a cysteine-rich domain, and a C-terminal spacer. However, stl only contains one thrombospondin-type 1 repeat (Fig. 2). Phylogenetic analysis reveals that stl is most closely related to human ADAMTS13, a protease required for the cleavage of von Willebrand factor (vWF) in blood coagulation (Cal et al., 2002; Ismat et al., 2013; Lee et al., 2012; Zheng, 2013). Three different loss-of-function alleles of stl were used in this study: stlph57, where a point mutation causes a stop codon just C-terminal to the metalloprotease domain, thereby deleting the DI, TS, Cys and spacer domains; stlpa49, where a point mutation only deletes the C-terminal spacer; and stlawk26, the most severe truncation as only the signal peptide and part of the prodomain remain (Fig. 2) (Ozdowski et al., 2009). All three alleles have previously been used as complete loss-of-function mutants of stl (Ozdowski et al., 2009; Willard et al., 2004). Although the truncated proteins generated by point mutations in stlph57 and stlpa49 only delete domains C-terminal to the metalloprotease domain, it is possible that these truncated proteins do not have much protease activity, as compared to the full length Stl protein. The metalloprotease domain alone of ADAMTS13 has no proteolytic activity toward the vWF substrate, and ADAMTS13 missing the Cysteine-rich and C-terminal spacer regions seem to have very minimal proteolytic activity (Ai et al., 2005; Soejima et al., 2003), suggesting that any deletion of protein domains in Stl may also abolish any proteolytic activity.
3.3. stl mRNA is expressed in and around the PNS during embryogenesis
Fig. 1 – The v’ch1 sensory neuron cell body migrates dorsally during embryogenesis. A. Stage 16 WT embryo stained with anti-22C10 to visualize the sensory neurons. B–D. Blown-up images of one abdominal segment showing the lch5 chordotonal organ (white arrowhead) and the v’ch1 sensory neuron (pseudo-colored in blue). B. At stage 15, the cell body of the v’ch1 sensory neuron (outlined in blue in B) is positioned ventral to the lch5 chordotonal organ (white arrowhead in B). C. At stage 16, v’ch1 neuron is inside the lch5 chordotonal organ. D. By stage 17, the v’ch1 neuron is now dorsal to the lch5 chordotonal organ.
Extracellular proteases, including matrix metalloproteases (MMPs), ADAM proteases, and ADAMTS proteases have all been shown to play significant roles in cell migration, during both development and disease states (Blelloch and Kimble, 1999; Blelloch et al., 1999; Coleman et al., 2010; Vu and Werb, 2000). Recently, it was shown that one of the ADAMTS proteases in Drosophila, AdamTS-A, is expressed in several migratory cell types during embryogenesis, and is required for their proper migration (Ismat et al., 2013). Previous reports showed a requirement for stl in ovarian growth control (Ozdowski et al., 2009; Willard et al., 2004). Therefore, we were curious whether stl was expressed during embryogenesis, and if so, does it also play a role in cell migration. stl mRNA is expressed in the region of the peripheral nervous system (PNS) of thoracic (t) and abdominal (a) segments of late stage Drosophila embryos,
4
m e c h a n i s m s o f d e v e l o p m e n t 1 3 7 ( 2 0 1 5 ) 1 –1 0
Fig. 2 – stl encodes an ADAMTS extracellular protease. WT stl encodes a protein with a signal peptide (SP), a prodomain, a Zn-dependent metalloprotease domain, a disintegrin-like domain (DI), a Thrombospondin-type I repeat (TS), a Cysteinerich domain (Cys), and a C-terminal spacer (spacer). The three loss-of-function alleles of stl, and the truncated proteins they make are: stlph57 which has the DI, TS, Cys, and ADAM domains deleted; stlpa49 which contains every domain except the ADAM spacer; and stlawk26 which deletes everything except the SP, and part of the prodomain (pro). Figure modified from Ozdowski et al., 2009.
starting at stage 15 (Fig. 3A,B). A ventral view shows no stl mRNA expression in the central nervous system (CNS) (Fig. 3B). In order to find out which cells within and/or around the PNS express stl mRNA, we co-stained WT embryos with stl mRNA and various molecular markers of the nervous system, including a marker for all neuronal nuclei, Elav, a sensory neuronspecific marker, 22C10, and a pan-glial marker Repo. stl mRNA (red) seemed to co-localize with Elav (green) in specific lateral and ventral sensory neurons (white arrowheads in Fig. 3E). Additionally, it is clear that stl mRNA (red) co-localized with 22C10 (green) in dorsal sensory neurons (white arrowheads in Fig. 3F). stl mRNA is also expressed just ventral and dorsal to the v’ch1 sensory neuron (white arrows in Fig. 3E,F). Since Stl is a secreted protease, we wanted to know if the stl mRNA that did not co-localize with neurons could be expressed in glial cells that surround certain axons on the PNS (Sepp and Auld, 2003). It is possible that Stl gets secreted by the glial cells to influence migration of nearby neurons. Using a panglial marker Repo, we did not observe any co-localization between stl mRNA and Repo (data not shown). From these data, stl seemed to be expressed within and around PNS neurons during late embryogenesis, specifically in cells dorsal and ventral to the v’ch1 sensory neurons in the abdominal segments.
The absence of stl function displays the v’ch1 3.4. sensory neuron migrating to its target sooner than in wild type Traditionally, when it comes to cell migration, it is thought that the role of extracellular proteases is to clear a path for the migratory cell by cleaving specific ECM components that would otherwise block the migratory cell from being able to move forward. However, it has recently been shown that extracellular proteases are required for more than just clearing a path at the leading edge of migrating cells. For example, the
function of Drosophila AdamTS-A is to detach migrating cells from the trailing end to allow them to move forward (Ismat et al., 2013). The expression of stl mRNA in and around the sensory neurons of the PNS prompted us to examine migration of these neurons more carefully. As displayed in Fig. 1, we mapped the dorsal migration of the v’ch1 sensory neuron during stages 15, 16, and 17, relative to the lch5 chordotonal organ cluster. In WT embryos at stage 15, the v’ch1 sensory neuron (black arrowhead) is normally ventral to the lch5 chordotonal organ cluster (white arrow) (Fig. 4A,C). In the absence of stl at stage 15, we observed the v’ch1 neuron in a more dorsal position relative to the lch5 chordotonal organ cluster (Fig. 4B,D). In order to quantify this observation, we categorized every v’ch1 neuron as ventral (pink cell, Fig. 4G), inside (yellow cell, Fig. 4F), or dorsal (blue cell, Fig. 4E) to the lch5 chordotonal organ cluster (gray cells, Fig. 4E–G), and graphed the percentages of the total v’ch1 neurons scored (Fig. 4H). Stage 15 WT v’ch1 neurons were ventral to the lch5 chordotonal organ cluster more than 70% of the time (Fig. 4H, pink bar). In contrast, only about 50% of v’ch1 neurons in stl mutants were ventral to the lch5 chordotonal organ cluster (Fig. 4H, pink bars). This significant difference was seen with three different single stl homozygous alleles (stlph57, stlpa49, stlawk26), as well as two transalleleic combinations (stlph57/pa49, stlph57/awk26). Moreover, whereas only 20% of v’ch1 neurons were inside the lch5 organ cluster in WT (Fig. 4H, yellow bar), approximately 40% of v’ch1 neurons were inside the lch5 organ cluster in the absence of stl (Fig. 4H, yellow bars). Clearly, in the absence of stl at stage 15, the v’ch1 sensory neuron has migrated further dorsal than in WT. These data suggest that this extracellular protease, expressed in cells dorsal to the v’ch1 sensory neuron (Fig. 3E,F), may normally function to inhibit migration of this sensory neuron. The next question is whether this neuron continues to migrate further dorsal and migrate past its correct target, or does this neuron reach its target and stop migrating.
m e c h a n i s m s o f d e v e l o p m e n t 1 3 7 ( 2 0 1 5 ) 1 –1 0
5
Fig. 3 – stl mRNA is expressed in and around the sensory neurons during embryogenesis. A. Lateral view of a stage 16 embryo showing stl mRNA in both the thoracic (t), and abdominal (a) segments. B. Ventral view of a stage 15 embryo showing no expression in the central nervous system (CNS). C. Stage 16 WT Embryo stained with stl mRNA (red) and a panneuronal nuclear marker Elav (green). D. Stage 16 WT Embryo stained with stl mRNA (red) and a PNS marker 22C10 (green). E. Blown-up image of boxed area in C showing co-localization of stl mRNA and Elav in ventral sensory neurons (white arrowheads). F. Blown-up image of boxed area in D showing co-localization of stl mRNA and 22C10 in dorsal sensory neurons (white arrowheads). stl mRNA also expressed in cells just ventral and dorsal to v’ch1 sensory neuron (white arrows in E, F).
We next wanted to know whether the v’ch1 sensory neuron migrates past its correct target, or reaches its target faster. In order to answer this question, we scored the position of the v’ch1 neuron in relation to the lch5 chordotonal organ cluster at stage 16 and stage 17 (Fig. 5). At stage 16, the v’ch1 neuron is inside the lch5 organ cluster (Fig. 1C), and at stage 17 the v’ch1 neuron is dorsal to the lch5 organ cluster (Fig. 1D). At stage 16, the v’ch1 sensory neurons were not in a more dorsal position in embryos missing stl, as compared to WT (Fig. 5A–D, I, compare black arrowheads in C,D). Although the single homozygous alleles stlph57 and stlpa49 do show a significant increase in the percentage of v’ch1 neurons in a more dorsal position than WT, when these two alleles are in trans to each other (stlph57/pa49), there is no significant difference in the v’ch1 position compared to WT (Fig. 5I). At stage 17, the v’ch1 neuron has reached its correct target dorsal to the lch5 chordotonal organ cluster, and has not migrated any further dorsal in the
absence of stl (Fig. 5E–H, I, compare black arrowheads in G,H). Quantification of this result shows no significant difference between WT and stl mutants (Fig. 5I). Clearly, these data suggest that, in the absence of stl, the v’ch1 neuron migrates to its target sooner than its WT counterpart, but does not over-migrate past its target.
4.
Discussion
In this study, we have shown an unexpected function for the ADAMTS extracellular protease stl in sensory neuron migration; stl functions to inhibit the v’ch1 sensory neuron from migrating dorsally too soon. The mRNA expression of stl just dorsal to the v’ch1 and lch5 chordotonal organ cluster (Fig. 3) was our first clue that stl could play a role in the proper
6
m e c h a n i s m s o f d e v e l o p m e n t 1 3 7 ( 2 0 1 5 ) 1 –1 0
Fig. 4 – Loss of stl displays over-migration of the v’ch1 sensory neuron at stage 15. A–D. Stage 15 WT (A, C), and stlph57 (B, D) embryos stained with anti-22C10. A. Stage 15 WT embryo. C. Blown-up image of boxed area in A shows v’ch1 neuron (black arrowhead) ventral to the lch5 chordotonal organ (white arrow). B. Stage 15 stlph57 embryo. D. Blown-up image of boxed area in B shows v’ch1 neuron (black arrowhead) inside the lch5 chordotonal organ (white arrow). E–G. Cartoon diagrams of v’ch1 position (pink, yellow, blue) relative to lch5 chordotonal organ (gray). H. Graph of percent v’ch1 cell body position (V, I, D) relative to the lch5 chordotonal organ at stage 15 for WT, single stl homozygous alleles (stlph57, stlpa49, stlawk26), and two transalleleic combinations (stlph57/pa49, stlph57/awk26). The number of neurons counted for each stage and genotype is in brackets. ** p < 0.001 based on a G-test of statistical independence.
migration of this sensory neuron. Surprisingly, embryos absent of stl function displayed the v’ch1 sensory neuron migrating to its target sooner than in WT. More specifically, at stage 15, the v’ch1 neuron was inside the lch5 chordotonal organ cluster as much or more often that it was ventral to this organ cluster in the absence of stl, as compared to the WT (Fig. 4). However, this over-migration of the v’ch1 neuron at stage 15 does not continue, and this neuron ends up at the same target position as in WT starting at stage 16 (Fig. 5). By stage 17, when this v’ch1 neuron is in its final position dorsal to the lch5 organ cluster (Fig. 1), there is no significant difference between stl mutants and WT in the position of the v’ch1 neuron (Fig. 5). These results
point to one of two models for how Stl, an extracellular protease, functions to inhibit migration of the v’ch1 sensory neuron. One model indicates Stl releasing an inhibitory cue to the v’ch1 sensory neuron early, at stage 15, to prevent it from reaching its target position too soon. In this case, Stl could release an inhibitor directly, such as Robo (Coleman et al., 2010), or indirectly by cleaving an ECM component that is sequestering the inhibitor. The second model indicates Stl destroying an attractive cue that would allow further migration of the v’ch1 sensory neuron past its target. Therefore, in the absence of stl function, this attractive cue has not been destroyed, and provides an unwanted early attractant for the v’ch1 neuron.
m e c h a n i s m s o f d e v e l o p m e n t 1 3 7 ( 2 0 1 5 ) 1 –1 0
7
Fig. 5 – The v’ch1 sensory neuron does not migrate past its target at stages 16 and 17. A–H. Stages 16 and 17 WT (A, C, E, G), and stlph57 (B, D, F, H) embryos stained with anti-22C10. A. Stage 16 WT embryo. C. Blown-up image of boxed area in A shows v’ch1 neuron (black arrowhead) inside the lch5 chordotonal organ (white arrow). B. Stage 16 stlph57 embryo. D. Blownup image of boxed area in B shows v’ch1 neuron (black arrowhead) inside the lch5 chordotonal organ (white arrow). E. Stage 17 WT embryo. G. Blown-up image of boxed area in E shows v’ch1 neuron (black arrowhead) dorsal to the lch5 chordotonal organ (white arrow). F. Stage 17 stlph57 embryo. H. Blown-up image of boxed area in F shows v’ch1 neuron (black arrowhead) dorsal to the lch5 chordotonal organ (white arrow). I. Graph of percent v’ch1 cell body position (V, I, D) relative to the lch5 chordotonal organ at stages 16 and 17 for WT, single stl homozygous alleles (stlph57, stlpa49, stlawk26), and two transalleleic combinations (stlph57/pa49, stlph57/awk26). The number of neurons counted for each stage and genotype is in brackets. * p < 0.05 based on a G-test of statistical independence. ** p < 0.001 based on a G-test of statistical independence.
4.1. Traditional and non-traditional roles for extracellular proteases in cell migration The vast majority of migrating cells go through the same cyclical steps in order to reach their target; they sense their environment, polarize in the direction of an attractive guidance cue, extend cellular protrusions in the direction of migration, release adhesions to the substrate from the trailing edge of the migrating cell, and propel themselves forward (Carey et al., 2011; Etienne-Manneville, 2008; Horwitz and Webb, 2003). Most of these migrating cells must perform these steps through a densely packed extracellular environment, the ECM, in order to reach their target. The ECM, therefore, is constantly being remodeled or restructured by extracellular proteases that target and cleave certain ECM components around migrating cells to allow forward migration (Vu and Werb, 2000). However, remodeling or restructuring of the surrounding ECM does not only take place at the leading edge of migrating cells, as was traditionally thought, but also occurs at the trailing edge to allow cells to detach from the ECM. Recently, the Drosophila extracellular protease AdamTS-A was shown to function in detaching the trailing end of salivary gland cells from the apical ECM in the salivary gland lumen. Without this detachment, the cells could not move forward in a smooth manner, stretching out as a result, and preventing proper migration of the salivary gland (Ismat et al., 2013). The human homologue of Drosophila Stl is ADAMTS13, known for cleaving a large polymeric adhesion protein von Willebrand factor (vWF) (Cal et al., 2002; Ismat et al., 2013; Lee et al., 2012; Zheng, 2013). Human ADAMTSs are grouped into four categories, based on structural and functional similari-
ties (Ismat et al., 2013). Adamts13 and its Drosophila homologue stl are the only members of one of the four groups, suggesting possibly strong similarities between these two proteins. Like other extracellular proteases, stl has multiple, even opposing functions during both development and homeostasis (Dong et al., 2003; Lee et al., 2012; Zheng, 2013). vWF, being an adhesion molecule, binds platelets, which then stops bleeding at the site of injury (Dong et al., 2003). The cleavage of vWF by ADAMTS13 prevents excessive blood coagulation, or thrombus from occurring (Dong et al., 2003; Zheng, 2013). Moreover, mice completely lacking Adamts13 display an increase in macrophage migration into an atherosclerotic lesion causing an increase in inflammation (Gandhi et al., 2011; Jin et al., 2012). In contrast, ADAMTS13 has been shown to have both antiangiogenic and pro-angiogenic properties, depending on the external environment of the migrating cells (Lee et al., 2012). This study demonstrates the complexity in assigning one specific role to this particular extracellular protease.
4.2. The v’ch1 sensory neuron migrates dorsally during embryogenesis In the Drosophila embryo, most neurons in the PNS start out in a more dorsal position and send out an axon ventrally to enter the CNS. The v’ch1 sensory neuron sends an axon ventrally that joins the segmental nerve (SN) while the lch5 chordotonal organ cluster sends an axon ventrally to meet the intersegmental nerve (ISN) (Bates and Whitington, 2007; Hartenstein, 1988; Parsons et al., 2003). Interestingly, the v’ch1 sensory neuron and the lch5 chordotonal organ cluster migrate during normal development of the PNS. Both the v’ch1 and lch5
8
m e c h a n i s m s o f d e v e l o p m e n t 1 3 7 ( 2 0 1 5 ) 1 –1 0
extend axons ventrally toward the CNS, and also extend one dendrite dorsally. It has recently been shown that the v’ch1 neuron-associated cap cell gets an attractive Netrin cue and pulls the neuron dorsally, suggesting that the v’ch1 neuron migrates passively (Mrkusich et al., 2010). Whether the v’ch1 neuron is passive and gets pulled dorsally by its cap cell or migrates actively, the fact remains that there are cues from the extracellular environment that are guiding these neurons. Therefore, it is possible for a secreted protease, such as Stl, to release a guidance cue that may be sequestered by the ECM environment it is in. Furthermore, guidance cues can be either attractive or inhibitory. The v’ch1 sensory neuron, like all migratory cells, has a very specific target, dorsal to the lch5 chordotonal organ cluster at stage 17 (Fig. 1).
4.3.
Axon guidance signaling pathways
Axon guidance of both longitudinal axons in the CNS and motor neurons in the PNS of Drosophila have been great systems with which to work out signaling pathways required for attraction and repulsion of axons. The Robo-Slit pathway is a well known pathway involved in repulsion of longitudinal axons across the CNS midline (Kidd et al., 1999; Simpson et al., 2000). More recently, the ADAM protease Kuz has also been shown to cleave the Robo receptor to allow for axon repulsion in the CNS (Coleman et al., 2010; Schimmelpfeng et al., 2001). In the PNS, it has been shown that the Robo receptor working with and without the secreted ligand Slit is important for proper axon guidance of lateral cluster sensory neurons (Parsons et al., 2003). It is also well established that PlexinSemaphorin signaling functions in repulsive guidance of motor neurons (Bates and Whitington, 2007; Van Vactor and Lorenz, 1999; Yu et al., 1998). Additionally, the ECM proteoglycan perlecan is secreted by motor neurons and required for Semaphorin-mediated repulsion (Cho et al., 2012). This particular study shows how important the right composition of ECM is for presentation of guidance cues for migration, as other ECM components like syndecan and dally-like could not substitute for perlecan (Cho et al., 2012). The ECM changes composition through restructuring, which is the function of extracellular proteases such as stl. Therefore, it would be interesting to see if there is any link between stl and Semaphorin-mediated repulsion. In this study, we have found that the extracellular protease stl seems to function to inhibit dorsal migration of the v’ch1 sensory neuron at stage 15 of embryogenesis. Our data support a model where Stl restructures the ECM around the v’ch1 neuron to present an inhibitory cue to stop the v’ch1 neuron from migrating too far too soon. This novel role for this protease is just the beginning where the major question still remains of what is the inhibitory signal this protease is presenting to the v’ch1 sensory neuron. Previous work has identified the secreted proteins Slit and Sema2a as inhibitors of neuronal migration (Bates and Whitington, 2007; Parsons et al., 2003). This work, along with several others, display a much more diverse array of functions for extracellular proteases than was previously thought. For example, matrix metalloproteinases (MMPs) are known to function in multiple cellular processes, including cell proliferation, apoptosis, and presenting growth factors to cells (Vu and Werb, 2000). In addition, ADAMTS20 is
required for melanoblast survival in mice (Silver et al., 2008). Through these new functions, we can gain more insight into how ECM restructuring influences basic cellular processes during both normal development as well as disease states like cancer.
Acknowledgements We would like to thank Trudi Schupbach (Princeton University, Princeton, NJ) and Claire Cronmiller (University of Virginia, Charlottesville, VA) for stl mutant alleles; Gregory Berry for intellectual discussions and reviewing of the manuscript; Pablo Jenik (Franklin & Marshall College, Lancaster, PA) for use of his Leica microscope for DIC images; and Wade Edris (Penn State College of Medicine, Hershey, PA) for technical support on the Leica SP2 confocal microscope. This project is funded, in part, under a grant with the Pennsylvania Department of Health using Tobacco CURE Funds. The Department specifically disclaims responsibility for any analyses, interpretations or conclusions. This work was supported by start-up funds from Franklin & Marshall College (106870.157540.51003.10) (to A.I.), and a Summer Hackman Research Fellowship (106045.157540.51003.21) (to T.L.).
REFERENCES
Ai, J., Smith, P., Wang, S., Zhang, P., Zheng, X.L., 2005. The proximal carboxyl-terminal domains of ADAMTS13 determine substrate specificity and are all required for cleavage of von Willebrand factor. J. Biol. Chem. 280, 29428– 29434. Aman, A., Piotrowski, T., 2010. Cell migration during morphogenesis. Dev. Biol. 341, 20–33. Andrew, D.J., Ewald, A.J., 2010. Morphogenesis of epithelial tubes: insights into tube formation, elongation, and elaboration. Dev. Biol. 341, 34–55. Apte, S.S., 2009. A disintegrin-like and metalloprotease (reprolysin-type) with thrombospondin type 1 motif (ADAMTS) superfamily: functions and mechanisms. J. Biol. Chem. 284, 31493–31497. Azpiazu, N., Frasch, M., 1993. tinman and bagpipe: two homeo box genes that determine cell fates in the dorsal mesoderm of Drosophila. Genes Dev. 7, 1325–1340. Bates, K.E., Whitington, P.M., 2007. Semaphorin 2a secreted by oenocytes signals through plexin B and plexin A to guide sensory axons in the Drosophila embryo. Dev. Biol. 302, 522– 535. Blelloch, R., Kimble, J., 1999. Control of organ shape by a secreted metalloprotease in the nematode Caenorhabditis elegans. Nature 399, 586–590. Blelloch, R., Anna-Arriola, S.S., Gao, D., Li, Y., Hodgkin, J., Kimble, J., 1999. The gon-1 gene is required for gonadal morphogenesis in Caenorhabditis elegans. Dev. Biol. 216, 382– 393. Bradley, P.L., Myat, M.M., Comeaux, C.A., Andrew, D.J., 2003. Posterior migration of the salivary gland requires an intact visceral mesoderm and integrin function. Dev. Biol. 257, 249– 262. Cal, S., Obaya, A.J., Llamazares, M., Garabaya, C., Quesada, V., Lopez-Otin, C., 2002. Cloning, expression analysis, and structural characterization of seven novel human ADAMTSs,
m e c h a n i s m s o f d e v e l o p m e n t 1 3 7 ( 2 0 1 5 ) 1 –1 0
a family of metalloproteinases with disintegrin and thrombospondin-1 domains. Gene 283, 49–62. Carey, S.P., Charest, J.M., Reinhart-King, C.A., 2011. Forces during cell adhesion and spreading: implications for cellular homeostasis. Stud. Mechanobiol. Tissue Eng. Biomater. 4, 29–69. Cho, J.Y., Chak, K., Andreone, B.J., Wooley, J.R., Kolodkin, A.L., 2012. The extracellular matrix proteoglycan perlecan facilitates transmembrane semaphorin-mediated repulsive guidance. Genes Dev. 26, 2222–2235. Coleman, H.A., Labrador, J.P., Chance, R.K., Bashaw, G.J., 2010. The Adam family metalloprotease Kuzbanian regulates the cleavage of the roundabout receptor to control axon repulsion at the midline. Development 137, 2417–2426. Dickson, B.J., Gilestro, G.F., 2006. Regulation of commissural axon pathfinding by slit and its Robo receptors. Annu. Rev. Cell Dev. Biol. 22, 651–675. Doitsidou, M., Reichman-Fried, M., Stebler, J., Koprunner, M., Dorries, J., Meyer, D., et al., 2002. Guidance of primordial germ cell migration by the chemokine SDF-1. Cell 111, 647–659. Dong, J., Moake, J.L., Bernardo, A., Fujikawa, K., Ball, C., Nolasco, L., et al., 2003. ADAMTS-13 metalloprotease interacts with the endothelial cell-derived ultra-large von Willebrand factor. J. Biol. Chem. 278, 29633–29639. Duchek, P., Somogyi, K., Jekely, G., Beccari, S., Rorth, P., 2001. Guidance of cell migration by the Drosophila PDGF/VEGF receptor. Cell 107, 17–26. Etienne-Manneville, S., 2008. Polarity proteins in migration and invasion. Oncogene 27, 6970–6980. Evans, T.A., Bashaw, G.J., 2010. Axon guidance at the midline: of mice and flies. Curr. Opin. Neurobiol. 20, 79–85. Gandhi, C., Khan, M.M., Lentz, S.R., Chauhan, A.K., 2011. ADAMTS13 reduces vascular inflammation and the development of early atherosclerosis in mice. Blood 119, 2385–2391. Hartenstein, V., 1988. Development of Drosophila larval sensory organs: spatiotemporal pattern of sensory neurons, peripheral axon pathways and sensilla differentiation. Development 102, 869–886. Horwitz, R., Webb, D., 2003. Cell migration. Curr. Biol. 13, R756– R759. Ismat, A., Cheshire, A.M., Andrew, D.J., 2013. The secreted AdamTS-A metalloprotease is require for collective cell migration. Development 140, 1981–1993. Jang, A.C., Starz-Gaiano, M., Montell, D.J., 2007. Modeling migration and metastasis in Drosophila. J. Mammary Gland Biol. Neoplasia 12, 103–114. Jin, S.Y., Tohyama, J., Bauer, R.C., Cao, N.N., Rader, D.J., Zheng, X.L., 2012. Genetic ablation of Adamts 13 gene dramatically accelerates the formation of early atherosclerosis in a murine model. Arterioscler. Thromb. Vasc. Biol. 32, 1817–1823. Kadam, S., Ghosh, S., Stathopoulos, A., 2012. Synchronous and symmetric migration of Drosophila caudal visceral mesoderm cells requires dual input by two FGF ligands. Development 139, 699–708. Kaminker, J.S., Singh, R., Lebestky, T., Yan, H., Banerjee, U., 2001. Redundant function of Runt Domain binding partners, Big brother and Brother, during Drosophila development. Development 128, 2639–2648. Kerman, B.E., Cheshire, A.M., Andrew, D.J., 2006. From fate to function: the Drosophila trachea and salivary gland as models for tubulogenesis. Differentiation 74, 326–348. Kidd, T., Bland, K.S., Goodman, C.S., 1999. Slit is the midline repellant for the robo receptor in Drosophila. Cell 96, 785–794. Knirr, S., Azpiazu, N., Frasch, M., 1999. The role of the NK homeobox gene slouch (S59) in somatic muscle patterning. Development 126, 4525–4535.
9
Kunwar, P.S., Siekhaus, D.E., Lehmann, R., 2006. In vivo migration: a germ cell perspective. Annu. Rev. Cell Dev. Biol. 22, 237–265. Lee, M., Rodansky, E.S., Smith, J.K., Rodgers, G.M., 2012. ADAMTS13 promotes angiogenesis and modulates VEGFinduced angiogenesis. Microvasc. Res. 84, 109–115. Mayor, R., Theveneau, E., 2013. The neural crest. Development 140, 2247–2251. McDonald, J.A., Pinheiro, E.M., Montell, D.J., 2003. PVF1, a PDGF/ VEGF homolog, is sufficient to guide border cells and interacts genetically with Taiman. Development 130, 3469–3478. Mrkusich, E.M., Osman, Z.B., Bates, K.E., Marchingo, J.M., DumanScheel, M., Whitington, P.M., 2010. Netrin-guided accessory cell morphogenesis dictates the dendrite orientation and migration of a Drosophila sensory neuron. Development 137, 2227–2235. Nandadasa, S., Foulcer, S., Apte, S.S., 2014. The multiple, complex roles of versicam and its proteolytic turnover by ADAMTS proteases during embryogenesis. Matrix Biol. 35, 34–41. Ozdowski, E.F., Mowery, Y.M., Cronmiller, C., 2009. Stall encodes and ADAMTS metalloprotease and interacts genetically with Delta in Drosophila ovarian follicle formation. Genetics 183, 1027–1040. Parsons, L., Harris, K.-L., Turner, K., Whitington, P.M., 2003. roundabout gene family functions during sensory axon guidance in the Drosophila embryo are mediated by both Slitdependent and Slit-independent mechanisms. Dev. Biol. 264, 363–375. Reichman-Fried, M., Dickson, B., Hafen, E., Shilo, B.Z., 1994. Elucidation of the role of breathless, a Drosophila FGF receptor homolog, in tracheal cell migration. Genes Dev. 8, 428–439. Reichman-Fried, M., Minina, S., Raz, E., 2004. Autonomous modes of behavior in primordial germ cell migration. Dev. Cell 6, 589– 596. Reim, I., Hollfelder, D., Ismat, A., Frasch, M., 2012. The FGF8related signals Pyramus and Thisbe promote pathfinding, substrate adhesion, and survival of migrating longitudinal gut muscle founder cells. Dev. Biol. 368, 28–43. Renault, A.D., Sigal, Y.J., Morris, A.J., Lehmann, R., 2004. Soma-germ line competition for lipid phosphate uptake regulates germ cell migration and survival. Science 305, 1963–1966. Reuter, R., Scott, M.P., 1990. Expression and function of the homeotic genes Antennapedia and Sex combs reduced in the embryonic midgut of Drosophila. Development 109, 289–303. Schimmelpfeng, K., Gogel, S., Klambt, C., 2001. The function of leak and kuzbanian during growth cone and cell migration. Mech. Dev. 106, 25–36. Sepp, K.J., Auld, V.J., 2003. Reciprocal interactions between neurons and glia are required for Drosophila peripheral nervous system development. J. Neurosci. 23, 8221–8230. Silver, D.L., Hou, L., Somerville, R., Young, M.E., Apte, S.S., Pavan, W.J., 2008. The secreted metalloprotease ADAMTS20 is required for melanoblast survival. PLoS Genet. 4, e1000003. Simpson, J.H., Bland, K.S., Fetter, R.D., Goodman, C.S., 2000. Shortrange and long-range guidance by Slit and its Robo receptors: a combinatorial code of Robo receptors controls lateral position. Cell 103, 1019–1032. Soejima, K., Matsumoto, M., Kokame, K., Yagi, H., Ishizashi, H., Maeda, H., et al., 2003. ADAMTS-13 cysteine-rich/spacer domains are functionally essential for von Willebrand factor cleavage. Blood 102, 3232–3237. Starz-Gaiano, M., Cho, N.K., Forbes, A., Lehmann, R., 2001. Spatially restricted activity of a Drosophila lipid phosphatase guides migrating germ cells. Development 128, 983–991. Stathopoulos, A., Tam, B., Ronshaugen, M., Frasch, M., Levine, M., 2004. Pyramus and thisbe: FGF genes that pattern the mesoderm of Drosophila embryos. Genes Dev. 18, 687–699.
10
m e c h a n i s m s o f d e v e l o p m e n t 1 3 7 ( 2 0 1 5 ) 1 –1 0
Sutherland, D., Samakovlis, C., Krasnow, M.A., 1996. Branchless encodes a Drosophila FGF homolog that controls tracheal cell migration and the pattern of branching. Cell 87, 1091–1101. Tessier-Lavigne, M., Goodman, C.S., 1996. The molecular biology of axon guidance. Science 274, 1123–1133. Van Vactor, D.V., Lorenz, L.J., 1999. Neural development: the semantics of axon guidance. Curr. Biol. 9, R201–R204. Vu, T.H., Werb, Z., 2000. Matrix metalloproteinases: effectors of development and normal physiology. Genes Dev. 14, 2123– 2133. Wagstaff, L., Kelwick, R., Decock, J., Edwards, D.R., 2011. The roles of ADAMTS metalloproteinases in tumorigenesis and metastasis. Front. Biosci. 16, 1861–1872.
Willard, S.S., Ozdowski, E.F., Jones, N.A., Cronmiller, C., 2004. stallmediated extrinsic control of ovarian follicle formation in Drosophila. Genetics 168, 191–198. Yu, H.H., Araj, H.H., Ralls, S.A., Kolodkin, A.L., 1998. The transmembrane Semaphorin Sema I is required in Drosophila for embryonic motor and CNS axon guidance. Neuron 20, 207– 220. Zhang, N., Zhang, J., Purcell, K.J., Cheng, Y., Howard, K., 1997. The Drosophila protein Wunen repels migrating germ cells. Nature 385, 64–67. Zheng, X.L., 2013. Structure-function and regulation of ADAMTS13 protease. J. Thromb. Haemost. 11, 11–23.
Available at www.sciencedirect.com
ScienceDirect j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / m o d o
The extracellular protease stl functions to inhibit migration of v’ch1 sensory neuron during Drosophila embryogenesis Tashi Lhamo, Afshan Ismat * Department of Biology, Franklin and Marshall College, P.O. Box 3003, Lancaster, PA 17604-3003, USA
A R T I C L E
I N F O
A B S T R A C T
Article history:
Proper migration of cells through the dense and complex extracellular matrix (ECM) re-
Received 22 September 2014
quires constant restructuring of the ECM to allow cells to move forward in a smooth manner.
Received in revised form
This restructuring can occur through the action of extracellular enzymes. Among these ex-
14 April 2015
tracellular enzymes is the ADAMTS (A Disintegrin And Metalloprotease with ThromboSpondin
Accepted 27 April 2015
repeats) family of secreted extracellular proteases. Drosophila stl encodes an ADAMTS pro-
Available online 4 May 2015
tease expressed in and around the peripheral nervous system (PNS) during embryogenesis. The absence of stl displayed one specific neuron, the v’ch1 sensory neuron, migrating to
Keywords:
its target sooner than in wild type. During normal development, the v’ch1 sensory neuron
Cell migration
migrates dorsally at the same time it is extending an axon ventrally toward the CNS. Sur-
ADAMTS
prisingly, in the absence of stl, the v’ch1 neuron migrated further dorsally as compared to
PNS
the wild type at stage 15, but did not migrate past its correct target at stage 16, suggesting
stl
a novel role for this extracellular protease in inhibiting migration of this neuron past a certain
Drosophila
point.
Sensory neurons
1.
Introduction
Cell migration is an essential process during development of multicellular organisms. Cells can migrate in different ways; as individuals like primordial germ cells (Kunwar et al., 2006), or as a multicellular collective like the salivary gland of Drosophila melanogaster or vertebrate mammary gland (Andrew and Ewald, 2010; Kerman et al., 2006). In addition, the entire cell can migrate, or only a part of the cell can migrate, as is the case with many neurons where there is axonal outgrowth (Aman and Piotrowski, 2010; Dickson and Gilestro, 2006; Evans and Bashaw, 2010; Jang et al., 2007). Despite all these differences in types of migration, all migrating cells, or parts of cells, have very specific targets. Two questions to think about are what
© 2015 Elsevier Ireland Ltd. All rights reserved.
types of signals are migrating cells receiving along their path in order to reach their target, and how do migrating cells know when to stop migrating? Although we have gained much insight into these questions, much more remains unknown. Therefore, using a simpler model system like Drosophila may be able to help us tackle these questions. Migrating cells require cues, both attractive and inhibitory, from their environment to guide them to the correct target (Bradley et al., 2003; Kunwar et al., 2006; Tessier-Lavigne and Goodman, 1996). Attractants have been well studied, including the growth factor PDGF that signals from the oocyte to guide border cells toward the oocyte (Duchek et al., 2001; McDonald et al., 2003), the FGF bnl expressed in the mesoderm that guides tracheal cells (Reichman-Fried et al., 1994; Sutherland et al., 1996), the FGFs pyr and ths guiding both early
* Corresponding author. Department of Biology, Franklin and Marshall College, P.O. Box 3003, Lancaster, PA 17604-3003, USA. Tel.: +1 717 358 6301; fax: +1 717 358 4548. E-mail address: [email protected] (A. Ismat). http://dx.doi.org/10.1016/j.mod.2015.04.004 0925-4773/© 2015 Elsevier Ireland Ltd. All rights reserved.
2
m e c h a n i s m s o f d e v e l o p m e n t 1 3 7 ( 2 0 1 5 ) 1 –1 0
mesoderm cells and caudal visceral mesoderm (CVM) cells (Kadam et al., 2012; Reim et al., 2012; Stathopoulos et al., 2004), and primordial germ cells in Zebrafish responding to the chemokine SDF-1a (Doitsidou et al., 2002; Reichman-Fried et al., 2004). As attractants function to guide cells toward a signal to keep cells on the correct path, inhibitory cues are also functioning to guide these migrating cells away from the incorrect path. Therefore, repelling cells from a certain area can be as important as guiding cells toward a default path. For example, the lipid phosphates wunen and wunen2 are required to repel germ cells away from the ventral midgut and the CNS so that the germ cells stay on a strict path toward the gonad (Kunwar et al., 2006; Renault et al., 2004; Starz-Gaiano et al., 2001; Zhang et al., 1997). During neural crest cell migration, class 3 Semaphorins and Eph/ephrin signaling are both used as avoidance cues for different streams of neural crest cells as they migrate toward their correct target (Mayor and Theveneau, 2013). The embryonic nervous system in Drosophila is a very wellstudied model for examining how attractive and inhibitory cues function during cell migration. During axon outgrowth, in both the central nervous system (CNS) and peripheral nervous system (PNS), the axon travels to its target to make synaptic connections. Just like other types of cell migration, axons use attractive and inhibitory cues to guide them along the right path. Moreover, these attractive and inhibitory cues can work both in short range and long range (Tessier-Lavigne and Goodman, 1996). Semaphorins function as repulsive cues during motor axon guidance (Bates and Whitington, 2007; Van Vactor and Lorenz, 1999; Yu et al., 1998). The ligand for Robo receptors, Slit, is secreted by midline glial cells and inhibits longitudinal axons in the Drosophila CNS (Kidd et al., 1999; Simpson et al., 2000). Slit-Robo signaling has been shown to play a role in the PNS where Robo inhibits lateral cluster sensory axons from exploring aberrant pathways (Parsons et al., 2003). The question we are interested in is how these extracellular factors are presented to the migrating cells to influence their migration in a spatiotemporal manner. Most cells migrate through a dense three-dimensional extracellular matrix (ECM) composed of a complex mixture of proteins and macromolecules. The only way cells can change their shape, rearrange themselves, and eventually migrate is for this dense, complex matrix to get restructured and remodeled by certain proteases that function within the ECM. ADAMTS (A Disintegrin And Metalloprotease with ThromboSpondin motifs) extracellular protease are one family of secreted proteases that function within the ECM to degrade certain ECM components, thereby allowing for various cellular functions, including cell migration (Apte, 2009; Nandadasa et al., 2014; Wagstaff et al., 2011). Traditionally, it has been thought that extracellular proteases functioning in cell migration have one job, to clear a path in the ECM at the leading edge of a migrating cell to allow the cell room to move forward. Recently, it has been shown that the function of these enzymes extends past this one role. The ADAM protease Kuzbanian cleaves the Robo receptor to activate repulsion of axons along the CNS midline (Coleman et al., 2010; Schimmelpfeng et al., 2001). Additionally, Drosophila AdamTS-A functions to detach cells from the trailing edge, thereby allowing cells to move forward in a smooth manner (Ismat et al., 2013).
In this study, we have shown that a second Drosophila ADAMTS extracellular protease stall (stl) functions in the PNS to inhibit migration of one sensory neuron. Previously, stl was shown to play a role in ovarian development, where loss of stl caused an overgrowth of ovarian tissue (Ozdowski et al., 2009; Willard et al., 2004). Here we show that stl mRNA is expressed in the vicinity of the v’ cluster neuron (v’ch1), and that loss of stl causes the v’ch1 neuron to migrate to its target dorsal to the lch5 chordotonal organ cluster sooner than its wildtype (WT) counterparts. Therefore, this study suggests a novel and unexpected role for extracellular proteases in presenting inhibitors for cell migration.
2.
Experimental methods
2.1.
Drosophila strains
The following Drosophila strains were used: Canton S (CS) (Bloomington Stock Center, Indiana University), stlph57, stlpa49, stlawk26 (Ozdowski et al., 2009) (from T. Schupbach, Princeton University, USA), stla16 (Ozdowski et al., 2009) (from C. Cronmiller, University of Virginia, USA). All embryos were collected at 25 °C. All stl alleles were balanced over the SM6 evelacZ balancer, and stl mutant embryos were identified by the absence of β-gal.
2.2.
Staining procedures
Antibody stainings, fluorescent double stainings for proteins and mRNA, and in situ hybridizations were performed as described (Azpiazu and Frasch, 1993; Knirr et al., 1999; Reuter and Scott, 1990). The following antibodies were used: rabbit polyclonal αnti-βgal (1:3000, MP Biochemical), mouse anti22C10 (1:10, Developmental Studies Hybridoma Bank [DSHB]), rat αnti-Elav (1:10, DSHB), mouse αnti-Repo (1:10, DSHB), mouse αnti-β-gal (1:500, Promega), sheep αnti-digoxigenin (DIG) (1:1000, Molecular Probes). The following digoxigenin-labeled RNA probe was used: stl (T7 RNA polymerase transcript of EST AT15733). Secondary biotinylated αnti-mouse, biotinylated αnti-rabbit, biotinylated αnti-sheep, goat αnti-mouse-488, goat αnti-rat488, goat α-rat568 were used. Brightfield DIC images were taken on a Leica DMRB microscope using a ProgRes C5 color camera and ProgRes image capture software (Franklin & Marshall College). Fluorescent images were taken with a Leica SP2 laser scanning confocal microscope (Penn State College of Medicine, Hershey, PA).
2.3.
Statistical analysis
A G-test of statistical independence was used to compare all three classes (dorsal, inside, ventral) of v’ch1 sensory neuron position at each stage (15, 16, and 17).
3.
Results
3.1. The v’ch1 neuron migrates dorsally during embryogenesis All the sensory neurons in the abdominal segments have been identified, and a majority have previously been mapped
m e c h a n i s m s o f d e v e l o p m e n t 1 3 7 ( 2 0 1 5 ) 1 –1 0
out spatiotemporally (Bates and Whitington, 2007; Hartenstein, 1988; Parsons et al., 2003). Most sensory neurons in the Drosophila PNS start out in a dorsal position in the embryo, and send out an axon ventrally to eventually meet the CNS. However, in the case of the v’ch1 sensory neuron, the cell body migrates dorsally at the same time it is extending an axon ventrally toward the CNS (Bates and Whitington, 2007; Hartenstein, 1988; Parsons et al., 2003). Although previous work has described v’ch1 migration (Bates and Whitington, 2007; Kaminker et al., 2001; Mrkusich et al., 2010; Parsons et al., 2003), we decided to focus on the dorsal migration, and position, of this v’ch1 neuron in relation to the lch5 chordotonal organ cluster between stages 15 and 17 of embryogenesis. Wild type (WT) embryos labeled with anti-22C10 antibody clearly show the position of the v’ch1 neuron (Fig. 1B–D, pseudo-colored in blue) in relation to the lch5 chordotonal organ cluster (Fig. 1B–D, white arrowhead). At stage 15, the v’ch1 neuron is in a position ventral
3
to the lch5 chordotonal organ cluster (Fig. 1B). At stage 16, the v’ch1 neuron has migrated to a position inside the lch5 organ cluster (Fig. 1C). By the end of embryogenesis at stage 17, the v’ch1 neuron has migrated to its final position dorsal to the lch5 organ cluster (Fig. 1D).
3.2.
stl encodes an ADAMTS extracellular protease
The human genome encodes 19 members of the ADAMTS family of metalloproteases (Apte, 2009). The Drosophila genome encodes three ADAMTS metalloproteases, AdamTS-A (Ismat et al., 2013), CG4096, and stall (stl) (Ozdowski et al., 2009; Willard et al., 2004). Like all other ADAMTSs, stl is secreted, and contains a signal peptide, a prodomain, a Zn-dependent metalloprotease domain, a disintegrin-like domain, thrombospondin-type 1 repeats, a cysteine-rich domain, and a C-terminal spacer. However, stl only contains one thrombospondin-type 1 repeat (Fig. 2). Phylogenetic analysis reveals that stl is most closely related to human ADAMTS13, a protease required for the cleavage of von Willebrand factor (vWF) in blood coagulation (Cal et al., 2002; Ismat et al., 2013; Lee et al., 2012; Zheng, 2013). Three different loss-of-function alleles of stl were used in this study: stlph57, where a point mutation causes a stop codon just C-terminal to the metalloprotease domain, thereby deleting the DI, TS, Cys and spacer domains; stlpa49, where a point mutation only deletes the C-terminal spacer; and stlawk26, the most severe truncation as only the signal peptide and part of the prodomain remain (Fig. 2) (Ozdowski et al., 2009). All three alleles have previously been used as complete loss-of-function mutants of stl (Ozdowski et al., 2009; Willard et al., 2004). Although the truncated proteins generated by point mutations in stlph57 and stlpa49 only delete domains C-terminal to the metalloprotease domain, it is possible that these truncated proteins do not have much protease activity, as compared to the full length Stl protein. The metalloprotease domain alone of ADAMTS13 has no proteolytic activity toward the vWF substrate, and ADAMTS13 missing the Cysteine-rich and C-terminal spacer regions seem to have very minimal proteolytic activity (Ai et al., 2005; Soejima et al., 2003), suggesting that any deletion of protein domains in Stl may also abolish any proteolytic activity.
3.3. stl mRNA is expressed in and around the PNS during embryogenesis
Fig. 1 – The v’ch1 sensory neuron cell body migrates dorsally during embryogenesis. A. Stage 16 WT embryo stained with anti-22C10 to visualize the sensory neurons. B–D. Blown-up images of one abdominal segment showing the lch5 chordotonal organ (white arrowhead) and the v’ch1 sensory neuron (pseudo-colored in blue). B. At stage 15, the cell body of the v’ch1 sensory neuron (outlined in blue in B) is positioned ventral to the lch5 chordotonal organ (white arrowhead in B). C. At stage 16, v’ch1 neuron is inside the lch5 chordotonal organ. D. By stage 17, the v’ch1 neuron is now dorsal to the lch5 chordotonal organ.
Extracellular proteases, including matrix metalloproteases (MMPs), ADAM proteases, and ADAMTS proteases have all been shown to play significant roles in cell migration, during both development and disease states (Blelloch and Kimble, 1999; Blelloch et al., 1999; Coleman et al., 2010; Vu and Werb, 2000). Recently, it was shown that one of the ADAMTS proteases in Drosophila, AdamTS-A, is expressed in several migratory cell types during embryogenesis, and is required for their proper migration (Ismat et al., 2013). Previous reports showed a requirement for stl in ovarian growth control (Ozdowski et al., 2009; Willard et al., 2004). Therefore, we were curious whether stl was expressed during embryogenesis, and if so, does it also play a role in cell migration. stl mRNA is expressed in the region of the peripheral nervous system (PNS) of thoracic (t) and abdominal (a) segments of late stage Drosophila embryos,
4
m e c h a n i s m s o f d e v e l o p m e n t 1 3 7 ( 2 0 1 5 ) 1 –1 0
Fig. 2 – stl encodes an ADAMTS extracellular protease. WT stl encodes a protein with a signal peptide (SP), a prodomain, a Zn-dependent metalloprotease domain, a disintegrin-like domain (DI), a Thrombospondin-type I repeat (TS), a Cysteinerich domain (Cys), and a C-terminal spacer (spacer). The three loss-of-function alleles of stl, and the truncated proteins they make are: stlph57 which has the DI, TS, Cys, and ADAM domains deleted; stlpa49 which contains every domain except the ADAM spacer; and stlawk26 which deletes everything except the SP, and part of the prodomain (pro). Figure modified from Ozdowski et al., 2009.
starting at stage 15 (Fig. 3A,B). A ventral view shows no stl mRNA expression in the central nervous system (CNS) (Fig. 3B). In order to find out which cells within and/or around the PNS express stl mRNA, we co-stained WT embryos with stl mRNA and various molecular markers of the nervous system, including a marker for all neuronal nuclei, Elav, a sensory neuronspecific marker, 22C10, and a pan-glial marker Repo. stl mRNA (red) seemed to co-localize with Elav (green) in specific lateral and ventral sensory neurons (white arrowheads in Fig. 3E). Additionally, it is clear that stl mRNA (red) co-localized with 22C10 (green) in dorsal sensory neurons (white arrowheads in Fig. 3F). stl mRNA is also expressed just ventral and dorsal to the v’ch1 sensory neuron (white arrows in Fig. 3E,F). Since Stl is a secreted protease, we wanted to know if the stl mRNA that did not co-localize with neurons could be expressed in glial cells that surround certain axons on the PNS (Sepp and Auld, 2003). It is possible that Stl gets secreted by the glial cells to influence migration of nearby neurons. Using a panglial marker Repo, we did not observe any co-localization between stl mRNA and Repo (data not shown). From these data, stl seemed to be expressed within and around PNS neurons during late embryogenesis, specifically in cells dorsal and ventral to the v’ch1 sensory neurons in the abdominal segments.
The absence of stl function displays the v’ch1 3.4. sensory neuron migrating to its target sooner than in wild type Traditionally, when it comes to cell migration, it is thought that the role of extracellular proteases is to clear a path for the migratory cell by cleaving specific ECM components that would otherwise block the migratory cell from being able to move forward. However, it has recently been shown that extracellular proteases are required for more than just clearing a path at the leading edge of migrating cells. For example, the
function of Drosophila AdamTS-A is to detach migrating cells from the trailing end to allow them to move forward (Ismat et al., 2013). The expression of stl mRNA in and around the sensory neurons of the PNS prompted us to examine migration of these neurons more carefully. As displayed in Fig. 1, we mapped the dorsal migration of the v’ch1 sensory neuron during stages 15, 16, and 17, relative to the lch5 chordotonal organ cluster. In WT embryos at stage 15, the v’ch1 sensory neuron (black arrowhead) is normally ventral to the lch5 chordotonal organ cluster (white arrow) (Fig. 4A,C). In the absence of stl at stage 15, we observed the v’ch1 neuron in a more dorsal position relative to the lch5 chordotonal organ cluster (Fig. 4B,D). In order to quantify this observation, we categorized every v’ch1 neuron as ventral (pink cell, Fig. 4G), inside (yellow cell, Fig. 4F), or dorsal (blue cell, Fig. 4E) to the lch5 chordotonal organ cluster (gray cells, Fig. 4E–G), and graphed the percentages of the total v’ch1 neurons scored (Fig. 4H). Stage 15 WT v’ch1 neurons were ventral to the lch5 chordotonal organ cluster more than 70% of the time (Fig. 4H, pink bar). In contrast, only about 50% of v’ch1 neurons in stl mutants were ventral to the lch5 chordotonal organ cluster (Fig. 4H, pink bars). This significant difference was seen with three different single stl homozygous alleles (stlph57, stlpa49, stlawk26), as well as two transalleleic combinations (stlph57/pa49, stlph57/awk26). Moreover, whereas only 20% of v’ch1 neurons were inside the lch5 organ cluster in WT (Fig. 4H, yellow bar), approximately 40% of v’ch1 neurons were inside the lch5 organ cluster in the absence of stl (Fig. 4H, yellow bars). Clearly, in the absence of stl at stage 15, the v’ch1 sensory neuron has migrated further dorsal than in WT. These data suggest that this extracellular protease, expressed in cells dorsal to the v’ch1 sensory neuron (Fig. 3E,F), may normally function to inhibit migration of this sensory neuron. The next question is whether this neuron continues to migrate further dorsal and migrate past its correct target, or does this neuron reach its target and stop migrating.
m e c h a n i s m s o f d e v e l o p m e n t 1 3 7 ( 2 0 1 5 ) 1 –1 0
5
Fig. 3 – stl mRNA is expressed in and around the sensory neurons during embryogenesis. A. Lateral view of a stage 16 embryo showing stl mRNA in both the thoracic (t), and abdominal (a) segments. B. Ventral view of a stage 15 embryo showing no expression in the central nervous system (CNS). C. Stage 16 WT Embryo stained with stl mRNA (red) and a panneuronal nuclear marker Elav (green). D. Stage 16 WT Embryo stained with stl mRNA (red) and a PNS marker 22C10 (green). E. Blown-up image of boxed area in C showing co-localization of stl mRNA and Elav in ventral sensory neurons (white arrowheads). F. Blown-up image of boxed area in D showing co-localization of stl mRNA and 22C10 in dorsal sensory neurons (white arrowheads). stl mRNA also expressed in cells just ventral and dorsal to v’ch1 sensory neuron (white arrows in E, F).
We next wanted to know whether the v’ch1 sensory neuron migrates past its correct target, or reaches its target faster. In order to answer this question, we scored the position of the v’ch1 neuron in relation to the lch5 chordotonal organ cluster at stage 16 and stage 17 (Fig. 5). At stage 16, the v’ch1 neuron is inside the lch5 organ cluster (Fig. 1C), and at stage 17 the v’ch1 neuron is dorsal to the lch5 organ cluster (Fig. 1D). At stage 16, the v’ch1 sensory neurons were not in a more dorsal position in embryos missing stl, as compared to WT (Fig. 5A–D, I, compare black arrowheads in C,D). Although the single homozygous alleles stlph57 and stlpa49 do show a significant increase in the percentage of v’ch1 neurons in a more dorsal position than WT, when these two alleles are in trans to each other (stlph57/pa49), there is no significant difference in the v’ch1 position compared to WT (Fig. 5I). At stage 17, the v’ch1 neuron has reached its correct target dorsal to the lch5 chordotonal organ cluster, and has not migrated any further dorsal in the
absence of stl (Fig. 5E–H, I, compare black arrowheads in G,H). Quantification of this result shows no significant difference between WT and stl mutants (Fig. 5I). Clearly, these data suggest that, in the absence of stl, the v’ch1 neuron migrates to its target sooner than its WT counterpart, but does not over-migrate past its target.
4.
Discussion
In this study, we have shown an unexpected function for the ADAMTS extracellular protease stl in sensory neuron migration; stl functions to inhibit the v’ch1 sensory neuron from migrating dorsally too soon. The mRNA expression of stl just dorsal to the v’ch1 and lch5 chordotonal organ cluster (Fig. 3) was our first clue that stl could play a role in the proper
6
m e c h a n i s m s o f d e v e l o p m e n t 1 3 7 ( 2 0 1 5 ) 1 –1 0
Fig. 4 – Loss of stl displays over-migration of the v’ch1 sensory neuron at stage 15. A–D. Stage 15 WT (A, C), and stlph57 (B, D) embryos stained with anti-22C10. A. Stage 15 WT embryo. C. Blown-up image of boxed area in A shows v’ch1 neuron (black arrowhead) ventral to the lch5 chordotonal organ (white arrow). B. Stage 15 stlph57 embryo. D. Blown-up image of boxed area in B shows v’ch1 neuron (black arrowhead) inside the lch5 chordotonal organ (white arrow). E–G. Cartoon diagrams of v’ch1 position (pink, yellow, blue) relative to lch5 chordotonal organ (gray). H. Graph of percent v’ch1 cell body position (V, I, D) relative to the lch5 chordotonal organ at stage 15 for WT, single stl homozygous alleles (stlph57, stlpa49, stlawk26), and two transalleleic combinations (stlph57/pa49, stlph57/awk26). The number of neurons counted for each stage and genotype is in brackets. ** p < 0.001 based on a G-test of statistical independence.
migration of this sensory neuron. Surprisingly, embryos absent of stl function displayed the v’ch1 sensory neuron migrating to its target sooner than in WT. More specifically, at stage 15, the v’ch1 neuron was inside the lch5 chordotonal organ cluster as much or more often that it was ventral to this organ cluster in the absence of stl, as compared to the WT (Fig. 4). However, this over-migration of the v’ch1 neuron at stage 15 does not continue, and this neuron ends up at the same target position as in WT starting at stage 16 (Fig. 5). By stage 17, when this v’ch1 neuron is in its final position dorsal to the lch5 organ cluster (Fig. 1), there is no significant difference between stl mutants and WT in the position of the v’ch1 neuron (Fig. 5). These results
point to one of two models for how Stl, an extracellular protease, functions to inhibit migration of the v’ch1 sensory neuron. One model indicates Stl releasing an inhibitory cue to the v’ch1 sensory neuron early, at stage 15, to prevent it from reaching its target position too soon. In this case, Stl could release an inhibitor directly, such as Robo (Coleman et al., 2010), or indirectly by cleaving an ECM component that is sequestering the inhibitor. The second model indicates Stl destroying an attractive cue that would allow further migration of the v’ch1 sensory neuron past its target. Therefore, in the absence of stl function, this attractive cue has not been destroyed, and provides an unwanted early attractant for the v’ch1 neuron.
m e c h a n i s m s o f d e v e l o p m e n t 1 3 7 ( 2 0 1 5 ) 1 –1 0
7
Fig. 5 – The v’ch1 sensory neuron does not migrate past its target at stages 16 and 17. A–H. Stages 16 and 17 WT (A, C, E, G), and stlph57 (B, D, F, H) embryos stained with anti-22C10. A. Stage 16 WT embryo. C. Blown-up image of boxed area in A shows v’ch1 neuron (black arrowhead) inside the lch5 chordotonal organ (white arrow). B. Stage 16 stlph57 embryo. D. Blownup image of boxed area in B shows v’ch1 neuron (black arrowhead) inside the lch5 chordotonal organ (white arrow). E. Stage 17 WT embryo. G. Blown-up image of boxed area in E shows v’ch1 neuron (black arrowhead) dorsal to the lch5 chordotonal organ (white arrow). F. Stage 17 stlph57 embryo. H. Blown-up image of boxed area in F shows v’ch1 neuron (black arrowhead) dorsal to the lch5 chordotonal organ (white arrow). I. Graph of percent v’ch1 cell body position (V, I, D) relative to the lch5 chordotonal organ at stages 16 and 17 for WT, single stl homozygous alleles (stlph57, stlpa49, stlawk26), and two transalleleic combinations (stlph57/pa49, stlph57/awk26). The number of neurons counted for each stage and genotype is in brackets. * p < 0.05 based on a G-test of statistical independence. ** p < 0.001 based on a G-test of statistical independence.
4.1. Traditional and non-traditional roles for extracellular proteases in cell migration The vast majority of migrating cells go through the same cyclical steps in order to reach their target; they sense their environment, polarize in the direction of an attractive guidance cue, extend cellular protrusions in the direction of migration, release adhesions to the substrate from the trailing edge of the migrating cell, and propel themselves forward (Carey et al., 2011; Etienne-Manneville, 2008; Horwitz and Webb, 2003). Most of these migrating cells must perform these steps through a densely packed extracellular environment, the ECM, in order to reach their target. The ECM, therefore, is constantly being remodeled or restructured by extracellular proteases that target and cleave certain ECM components around migrating cells to allow forward migration (Vu and Werb, 2000). However, remodeling or restructuring of the surrounding ECM does not only take place at the leading edge of migrating cells, as was traditionally thought, but also occurs at the trailing edge to allow cells to detach from the ECM. Recently, the Drosophila extracellular protease AdamTS-A was shown to function in detaching the trailing end of salivary gland cells from the apical ECM in the salivary gland lumen. Without this detachment, the cells could not move forward in a smooth manner, stretching out as a result, and preventing proper migration of the salivary gland (Ismat et al., 2013). The human homologue of Drosophila Stl is ADAMTS13, known for cleaving a large polymeric adhesion protein von Willebrand factor (vWF) (Cal et al., 2002; Ismat et al., 2013; Lee et al., 2012; Zheng, 2013). Human ADAMTSs are grouped into four categories, based on structural and functional similari-
ties (Ismat et al., 2013). Adamts13 and its Drosophila homologue stl are the only members of one of the four groups, suggesting possibly strong similarities between these two proteins. Like other extracellular proteases, stl has multiple, even opposing functions during both development and homeostasis (Dong et al., 2003; Lee et al., 2012; Zheng, 2013). vWF, being an adhesion molecule, binds platelets, which then stops bleeding at the site of injury (Dong et al., 2003). The cleavage of vWF by ADAMTS13 prevents excessive blood coagulation, or thrombus from occurring (Dong et al., 2003; Zheng, 2013). Moreover, mice completely lacking Adamts13 display an increase in macrophage migration into an atherosclerotic lesion causing an increase in inflammation (Gandhi et al., 2011; Jin et al., 2012). In contrast, ADAMTS13 has been shown to have both antiangiogenic and pro-angiogenic properties, depending on the external environment of the migrating cells (Lee et al., 2012). This study demonstrates the complexity in assigning one specific role to this particular extracellular protease.
4.2. The v’ch1 sensory neuron migrates dorsally during embryogenesis In the Drosophila embryo, most neurons in the PNS start out in a more dorsal position and send out an axon ventrally to enter the CNS. The v’ch1 sensory neuron sends an axon ventrally that joins the segmental nerve (SN) while the lch5 chordotonal organ cluster sends an axon ventrally to meet the intersegmental nerve (ISN) (Bates and Whitington, 2007; Hartenstein, 1988; Parsons et al., 2003). Interestingly, the v’ch1 sensory neuron and the lch5 chordotonal organ cluster migrate during normal development of the PNS. Both the v’ch1 and lch5
8
m e c h a n i s m s o f d e v e l o p m e n t 1 3 7 ( 2 0 1 5 ) 1 –1 0
extend axons ventrally toward the CNS, and also extend one dendrite dorsally. It has recently been shown that the v’ch1 neuron-associated cap cell gets an attractive Netrin cue and pulls the neuron dorsally, suggesting that the v’ch1 neuron migrates passively (Mrkusich et al., 2010). Whether the v’ch1 neuron is passive and gets pulled dorsally by its cap cell or migrates actively, the fact remains that there are cues from the extracellular environment that are guiding these neurons. Therefore, it is possible for a secreted protease, such as Stl, to release a guidance cue that may be sequestered by the ECM environment it is in. Furthermore, guidance cues can be either attractive or inhibitory. The v’ch1 sensory neuron, like all migratory cells, has a very specific target, dorsal to the lch5 chordotonal organ cluster at stage 17 (Fig. 1).
4.3.
Axon guidance signaling pathways
Axon guidance of both longitudinal axons in the CNS and motor neurons in the PNS of Drosophila have been great systems with which to work out signaling pathways required for attraction and repulsion of axons. The Robo-Slit pathway is a well known pathway involved in repulsion of longitudinal axons across the CNS midline (Kidd et al., 1999; Simpson et al., 2000). More recently, the ADAM protease Kuz has also been shown to cleave the Robo receptor to allow for axon repulsion in the CNS (Coleman et al., 2010; Schimmelpfeng et al., 2001). In the PNS, it has been shown that the Robo receptor working with and without the secreted ligand Slit is important for proper axon guidance of lateral cluster sensory neurons (Parsons et al., 2003). It is also well established that PlexinSemaphorin signaling functions in repulsive guidance of motor neurons (Bates and Whitington, 2007; Van Vactor and Lorenz, 1999; Yu et al., 1998). Additionally, the ECM proteoglycan perlecan is secreted by motor neurons and required for Semaphorin-mediated repulsion (Cho et al., 2012). This particular study shows how important the right composition of ECM is for presentation of guidance cues for migration, as other ECM components like syndecan and dally-like could not substitute for perlecan (Cho et al., 2012). The ECM changes composition through restructuring, which is the function of extracellular proteases such as stl. Therefore, it would be interesting to see if there is any link between stl and Semaphorin-mediated repulsion. In this study, we have found that the extracellular protease stl seems to function to inhibit dorsal migration of the v’ch1 sensory neuron at stage 15 of embryogenesis. Our data support a model where Stl restructures the ECM around the v’ch1 neuron to present an inhibitory cue to stop the v’ch1 neuron from migrating too far too soon. This novel role for this protease is just the beginning where the major question still remains of what is the inhibitory signal this protease is presenting to the v’ch1 sensory neuron. Previous work has identified the secreted proteins Slit and Sema2a as inhibitors of neuronal migration (Bates and Whitington, 2007; Parsons et al., 2003). This work, along with several others, display a much more diverse array of functions for extracellular proteases than was previously thought. For example, matrix metalloproteinases (MMPs) are known to function in multiple cellular processes, including cell proliferation, apoptosis, and presenting growth factors to cells (Vu and Werb, 2000). In addition, ADAMTS20 is
required for melanoblast survival in mice (Silver et al., 2008). Through these new functions, we can gain more insight into how ECM restructuring influences basic cellular processes during both normal development as well as disease states like cancer.
Acknowledgements We would like to thank Trudi Schupbach (Princeton University, Princeton, NJ) and Claire Cronmiller (University of Virginia, Charlottesville, VA) for stl mutant alleles; Gregory Berry for intellectual discussions and reviewing of the manuscript; Pablo Jenik (Franklin & Marshall College, Lancaster, PA) for use of his Leica microscope for DIC images; and Wade Edris (Penn State College of Medicine, Hershey, PA) for technical support on the Leica SP2 confocal microscope. This project is funded, in part, under a grant with the Pennsylvania Department of Health using Tobacco CURE Funds. The Department specifically disclaims responsibility for any analyses, interpretations or conclusions. This work was supported by start-up funds from Franklin & Marshall College (106870.157540.51003.10) (to A.I.), and a Summer Hackman Research Fellowship (106045.157540.51003.21) (to T.L.).
REFERENCES
Ai, J., Smith, P., Wang, S., Zhang, P., Zheng, X.L., 2005. The proximal carboxyl-terminal domains of ADAMTS13 determine substrate specificity and are all required for cleavage of von Willebrand factor. J. Biol. Chem. 280, 29428– 29434. Aman, A., Piotrowski, T., 2010. Cell migration during morphogenesis. Dev. Biol. 341, 20–33. Andrew, D.J., Ewald, A.J., 2010. Morphogenesis of epithelial tubes: insights into tube formation, elongation, and elaboration. Dev. Biol. 341, 34–55. Apte, S.S., 2009. A disintegrin-like and metalloprotease (reprolysin-type) with thrombospondin type 1 motif (ADAMTS) superfamily: functions and mechanisms. J. Biol. Chem. 284, 31493–31497. Azpiazu, N., Frasch, M., 1993. tinman and bagpipe: two homeo box genes that determine cell fates in the dorsal mesoderm of Drosophila. Genes Dev. 7, 1325–1340. Bates, K.E., Whitington, P.M., 2007. Semaphorin 2a secreted by oenocytes signals through plexin B and plexin A to guide sensory axons in the Drosophila embryo. Dev. Biol. 302, 522– 535. Blelloch, R., Kimble, J., 1999. Control of organ shape by a secreted metalloprotease in the nematode Caenorhabditis elegans. Nature 399, 586–590. Blelloch, R., Anna-Arriola, S.S., Gao, D., Li, Y., Hodgkin, J., Kimble, J., 1999. The gon-1 gene is required for gonadal morphogenesis in Caenorhabditis elegans. Dev. Biol. 216, 382– 393. Bradley, P.L., Myat, M.M., Comeaux, C.A., Andrew, D.J., 2003. Posterior migration of the salivary gland requires an intact visceral mesoderm and integrin function. Dev. Biol. 257, 249– 262. Cal, S., Obaya, A.J., Llamazares, M., Garabaya, C., Quesada, V., Lopez-Otin, C., 2002. Cloning, expression analysis, and structural characterization of seven novel human ADAMTSs,
m e c h a n i s m s o f d e v e l o p m e n t 1 3 7 ( 2 0 1 5 ) 1 –1 0
a family of metalloproteinases with disintegrin and thrombospondin-1 domains. Gene 283, 49–62. Carey, S.P., Charest, J.M., Reinhart-King, C.A., 2011. Forces during cell adhesion and spreading: implications for cellular homeostasis. Stud. Mechanobiol. Tissue Eng. Biomater. 4, 29–69. Cho, J.Y., Chak, K., Andreone, B.J., Wooley, J.R., Kolodkin, A.L., 2012. The extracellular matrix proteoglycan perlecan facilitates transmembrane semaphorin-mediated repulsive guidance. Genes Dev. 26, 2222–2235. Coleman, H.A., Labrador, J.P., Chance, R.K., Bashaw, G.J., 2010. The Adam family metalloprotease Kuzbanian regulates the cleavage of the roundabout receptor to control axon repulsion at the midline. Development 137, 2417–2426. Dickson, B.J., Gilestro, G.F., 2006. Regulation of commissural axon pathfinding by slit and its Robo receptors. Annu. Rev. Cell Dev. Biol. 22, 651–675. Doitsidou, M., Reichman-Fried, M., Stebler, J., Koprunner, M., Dorries, J., Meyer, D., et al., 2002. Guidance of primordial germ cell migration by the chemokine SDF-1. Cell 111, 647–659. Dong, J., Moake, J.L., Bernardo, A., Fujikawa, K., Ball, C., Nolasco, L., et al., 2003. ADAMTS-13 metalloprotease interacts with the endothelial cell-derived ultra-large von Willebrand factor. J. Biol. Chem. 278, 29633–29639. Duchek, P., Somogyi, K., Jekely, G., Beccari, S., Rorth, P., 2001. Guidance of cell migration by the Drosophila PDGF/VEGF receptor. Cell 107, 17–26. Etienne-Manneville, S., 2008. Polarity proteins in migration and invasion. Oncogene 27, 6970–6980. Evans, T.A., Bashaw, G.J., 2010. Axon guidance at the midline: of mice and flies. Curr. Opin. Neurobiol. 20, 79–85. Gandhi, C., Khan, M.M., Lentz, S.R., Chauhan, A.K., 2011. ADAMTS13 reduces vascular inflammation and the development of early atherosclerosis in mice. Blood 119, 2385–2391. Hartenstein, V., 1988. Development of Drosophila larval sensory organs: spatiotemporal pattern of sensory neurons, peripheral axon pathways and sensilla differentiation. Development 102, 869–886. Horwitz, R., Webb, D., 2003. Cell migration. Curr. Biol. 13, R756– R759. Ismat, A., Cheshire, A.M., Andrew, D.J., 2013. The secreted AdamTS-A metalloprotease is require for collective cell migration. Development 140, 1981–1993. Jang, A.C., Starz-Gaiano, M., Montell, D.J., 2007. Modeling migration and metastasis in Drosophila. J. Mammary Gland Biol. Neoplasia 12, 103–114. Jin, S.Y., Tohyama, J., Bauer, R.C., Cao, N.N., Rader, D.J., Zheng, X.L., 2012. Genetic ablation of Adamts 13 gene dramatically accelerates the formation of early atherosclerosis in a murine model. Arterioscler. Thromb. Vasc. Biol. 32, 1817–1823. Kadam, S., Ghosh, S., Stathopoulos, A., 2012. Synchronous and symmetric migration of Drosophila caudal visceral mesoderm cells requires dual input by two FGF ligands. Development 139, 699–708. Kaminker, J.S., Singh, R., Lebestky, T., Yan, H., Banerjee, U., 2001. Redundant function of Runt Domain binding partners, Big brother and Brother, during Drosophila development. Development 128, 2639–2648. Kerman, B.E., Cheshire, A.M., Andrew, D.J., 2006. From fate to function: the Drosophila trachea and salivary gland as models for tubulogenesis. Differentiation 74, 326–348. Kidd, T., Bland, K.S., Goodman, C.S., 1999. Slit is the midline repellant for the robo receptor in Drosophila. Cell 96, 785–794. Knirr, S., Azpiazu, N., Frasch, M., 1999. The role of the NK homeobox gene slouch (S59) in somatic muscle patterning. Development 126, 4525–4535.
9
Kunwar, P.S., Siekhaus, D.E., Lehmann, R., 2006. In vivo migration: a germ cell perspective. Annu. Rev. Cell Dev. Biol. 22, 237–265. Lee, M., Rodansky, E.S., Smith, J.K., Rodgers, G.M., 2012. ADAMTS13 promotes angiogenesis and modulates VEGFinduced angiogenesis. Microvasc. Res. 84, 109–115. Mayor, R., Theveneau, E., 2013. The neural crest. Development 140, 2247–2251. McDonald, J.A., Pinheiro, E.M., Montell, D.J., 2003. PVF1, a PDGF/ VEGF homolog, is sufficient to guide border cells and interacts genetically with Taiman. Development 130, 3469–3478. Mrkusich, E.M., Osman, Z.B., Bates, K.E., Marchingo, J.M., DumanScheel, M., Whitington, P.M., 2010. Netrin-guided accessory cell morphogenesis dictates the dendrite orientation and migration of a Drosophila sensory neuron. Development 137, 2227–2235. Nandadasa, S., Foulcer, S., Apte, S.S., 2014. The multiple, complex roles of versicam and its proteolytic turnover by ADAMTS proteases during embryogenesis. Matrix Biol. 35, 34–41. Ozdowski, E.F., Mowery, Y.M., Cronmiller, C., 2009. Stall encodes and ADAMTS metalloprotease and interacts genetically with Delta in Drosophila ovarian follicle formation. Genetics 183, 1027–1040. Parsons, L., Harris, K.-L., Turner, K., Whitington, P.M., 2003. roundabout gene family functions during sensory axon guidance in the Drosophila embryo are mediated by both Slitdependent and Slit-independent mechanisms. Dev. Biol. 264, 363–375. Reichman-Fried, M., Dickson, B., Hafen, E., Shilo, B.Z., 1994. Elucidation of the role of breathless, a Drosophila FGF receptor homolog, in tracheal cell migration. Genes Dev. 8, 428–439. Reichman-Fried, M., Minina, S., Raz, E., 2004. Autonomous modes of behavior in primordial germ cell migration. Dev. Cell 6, 589– 596. Reim, I., Hollfelder, D., Ismat, A., Frasch, M., 2012. The FGF8related signals Pyramus and Thisbe promote pathfinding, substrate adhesion, and survival of migrating longitudinal gut muscle founder cells. Dev. Biol. 368, 28–43. Renault, A.D., Sigal, Y.J., Morris, A.J., Lehmann, R., 2004. Soma-germ line competition for lipid phosphate uptake regulates germ cell migration and survival. Science 305, 1963–1966. Reuter, R., Scott, M.P., 1990. Expression and function of the homeotic genes Antennapedia and Sex combs reduced in the embryonic midgut of Drosophila. Development 109, 289–303. Schimmelpfeng, K., Gogel, S., Klambt, C., 2001. The function of leak and kuzbanian during growth cone and cell migration. Mech. Dev. 106, 25–36. Sepp, K.J., Auld, V.J., 2003. Reciprocal interactions between neurons and glia are required for Drosophila peripheral nervous system development. J. Neurosci. 23, 8221–8230. Silver, D.L., Hou, L., Somerville, R., Young, M.E., Apte, S.S., Pavan, W.J., 2008. The secreted metalloprotease ADAMTS20 is required for melanoblast survival. PLoS Genet. 4, e1000003. Simpson, J.H., Bland, K.S., Fetter, R.D., Goodman, C.S., 2000. Shortrange and long-range guidance by Slit and its Robo receptors: a combinatorial code of Robo receptors controls lateral position. Cell 103, 1019–1032. Soejima, K., Matsumoto, M., Kokame, K., Yagi, H., Ishizashi, H., Maeda, H., et al., 2003. ADAMTS-13 cysteine-rich/spacer domains are functionally essential for von Willebrand factor cleavage. Blood 102, 3232–3237. Starz-Gaiano, M., Cho, N.K., Forbes, A., Lehmann, R., 2001. Spatially restricted activity of a Drosophila lipid phosphatase guides migrating germ cells. Development 128, 983–991. Stathopoulos, A., Tam, B., Ronshaugen, M., Frasch, M., Levine, M., 2004. Pyramus and thisbe: FGF genes that pattern the mesoderm of Drosophila embryos. Genes Dev. 18, 687–699.
10
m e c h a n i s m s o f d e v e l o p m e n t 1 3 7 ( 2 0 1 5 ) 1 –1 0
Sutherland, D., Samakovlis, C., Krasnow, M.A., 1996. Branchless encodes a Drosophila FGF homolog that controls tracheal cell migration and the pattern of branching. Cell 87, 1091–1101. Tessier-Lavigne, M., Goodman, C.S., 1996. The molecular biology of axon guidance. Science 274, 1123–1133. Van Vactor, D.V., Lorenz, L.J., 1999. Neural development: the semantics of axon guidance. Curr. Biol. 9, R201–R204. Vu, T.H., Werb, Z., 2000. Matrix metalloproteinases: effectors of development and normal physiology. Genes Dev. 14, 2123– 2133. Wagstaff, L., Kelwick, R., Decock, J., Edwards, D.R., 2011. The roles of ADAMTS metalloproteinases in tumorigenesis and metastasis. Front. Biosci. 16, 1861–1872.
Willard, S.S., Ozdowski, E.F., Jones, N.A., Cronmiller, C., 2004. stallmediated extrinsic control of ovarian follicle formation in Drosophila. Genetics 168, 191–198. Yu, H.H., Araj, H.H., Ralls, S.A., Kolodkin, A.L., 1998. The transmembrane Semaphorin Sema I is required in Drosophila for embryonic motor and CNS axon guidance. Neuron 20, 207– 220. Zhang, N., Zhang, J., Purcell, K.J., Cheng, Y., Howard, K., 1997. The Drosophila protein Wunen repels migrating germ cells. Nature 385, 64–67. Zheng, X.L., 2013. Structure-function and regulation of ADAMTS13 protease. J. Thromb. Haemost. 11, 11–23.
