Introduction

In diverse developmental contexts, directed cell migration constitutes a conserved and vital feature underlying the formation of a variety of cellular assemblies, including tissues and organs (Araujo 2015; Campanale and Montell 2023; Chen et al. 2025; Diaz et al. 2025; Roper 2025; Schick et al. 2025; Chowdhury et al. 2026). Migrating cells navigate across multiple cell layers and travel long distances, which is controlled by soluble factors, chemokines, and intercellular communication, including cell adhesion (Paluch et al. 2016; Barton et al. 2024). Intriguingly, at times, proper migration of a specific cell type is contingent upon secreted signaling ligands functioning either as attractive or repulsive cues. As these migratory events are essential for generating fully functional and fertile multicellular organisms, a wide range of model systems, spanning from Dictyostelium to higher vertebrates, including mammals, are employed to elucidate underlying mechanisms (Santos and Lehmann 2004a; Deshpande et al. 2025). Given the biological relevance of these processes, severe cell migration aberrations have detrimental consequences, including sterility, metastasis and lethality (Paluch et al. 2016; Barton et al. 2024; Roper 2025; Chowdhury et al. 2026).

Embryonic gonad coalescence in Drosophila has emerged as a formidable context to analyze the mechanistic underpinnings of cell migration (Kunwar et al. 2006; Chen et al. 2025; Deshpande et al. 2025). Several features render it an ideal system for uncovering and elucidating multimodal mechanisms underlying cell migration. First, the embryonic Drosophila gonad consists of just two cell types: Primordial Germ Cells (PGCs) and Somatic gonadal precursor cells (SGPs), which are specified at distant locations, with the PGCs undergoing directed migration towards the SGPs. PGCs are initially formed at the posterior of the embryo and at the onset of gastrulation are located next to a layer of somatic cells. During gastrulation they are internalized by the invagination of the midgut. At the germband extension (stage 9), PGCs localized at the blind end of the gut begin to segregate away from one another and individualize in a coordinated manner before traversing through the gut epithelium. After emerging from the gut around stage 10, the migrating PGCs split into two bilateral clusters, enter the mesoderm and begin migrating towards the SGPs. The SGPs are specified in para-segments 10-13 and are arranged in two bilateral clusters. Once the PGCs and SGPs come into contact with each other, they intermingle and eventually compact into a tight cluster to form the paired primitive embryonic gonad.

The migration of the PGCs from the blind end of the gut and through the mesoderm is orchestrated by guidance signals emanating from the SGPs. The first hint of the possibility came from the discovery that 3-hydroxy-3-methylglutaryl-CoA reductase (Hmgcr) is responsible for the production or transmission of an attractive guidance cue. hmgcr is the rate-limiting step in several biosynthetic pathways including that of cholesterol and isoprenoids. Loss of hmgcr results in the scattering or mis-migration of PGCs in the posterior of the embryo, while ectopic expression of hmgcr induces PGCs to migrate abnormally towards the cells/tissues expressing hmgcr. In addition, hmgcr is initially expressed broadly in the mesoderm; however, as the embryo develops, its expression becomes progressively restricted, first to the posterior mesoderm and by stage 12, almost exclusively in the SGPs.

Though these studies demonstrated that hmgcr plays a key role in orchestrating PGCs migration, its immediate biosynthetic product (mevalonate) does not appear to be a plausible candidate as a guidance cue for the migrating PGCs (Santos and Lehmann 2004b). However, since mis-expression of hmgcr in many different cell types/tissues in the embryo induced PGCs to migrate towards the site of hmgcr expression, it was apparent that the actual attractant must be broadly expressed in the embryo but somehow limited in its ability to function as an attractant in the absence of hmgcr. The identification of Hedgehog (Hh) as the likely PGC attractant (Deshpande et al. 2001) supported this notion. Like hmgcr, Hh is expressed in the SGPs and their progenitors. Likewise, misexpression of hh induces a subset of PGCs to migrate towards the ectopic source of Hh. Interestingly, the mis-migration of PGCs induced by hh ectopic expression is typically less severe than that induced by hmgcr expression in the same cell types/tissues (Deshpande et al. 2013). This observation suggested that hmgcr was a limiting factor in the transmission of the Hh ligand. Several findings are consistent with this idea. First, in the ectoderm of hmgcr mutant embryos, the Hh ligand is retained in hh-expressing cells instead of being transmitted to the neighboring cells (Deshpande and Schedl 2005). Second, co-ectopic expression of hh and hmgcr in the CNS leads not only to much more severe PGC mis-migration, but also substantially enhances the transmission of the Hh ligand from the ectoderm into the underlying mesoderm where it decorates the migrating PGCs (Deshpande et al. 2013). Third, epistasis experiments show that hh functions downstream of hmgcr in PGC migration: the effects of ectopic hmgcr expression in the CNS on PGC migration can be suppressed by simultaneous knockdown of hh (Deshpande et al. 2023). Fourth, two other genes, toutvelu and dispatched, known to be important for the cytoneme-based transmission of the Hh ligand are also epistatic to ectopically expressed hmgcr (Deshpande et al. 2023).

The long-distance transmission of the Hh ligand requires the proteolytic cleavage of the precursor protein and a concurrent cholesterol modification. While several studies indicate that the cholesterol modification is important for the transmission of the Hh ligand from the SGPs to the migrating PGCs (Deshpande et al. 2016; Bialistoky et al. 2019), flies lack the enzymes downstream of hmgcr for biosynthesis of cholesterol and thus this cannot be the function of hmgcr in the fly Hh signaling pathway. A solution to this puzzle came from the studies of (Santos and Lehmann 2004b) who showed that genes in the isoprenoid branch of the hmgcr biosynthetic pathway (Farnesyl diphosphate synthase: fpps; Geranyl-geranyl diphosphate synthase: qme) function downstream of hmgcr in PGC migration as does the isoprenoid transferase Geranyl-geranyl transferase (GGT1). The isoprenoid biosynthetic pathway is also required for the release/transmission of the Hh ligand, and at least one target in the Hh signaling pathway for geranylation by GGT1 is the gγ subunit of the heterotrimeric G protein complex (Deshpande et al. 2009).

Further evidence that the Hh signaling pathway orchestrates PGCs migration came from the experiments demonstrating that the reciprocal activities of the two receptors of the Hh ligand, namely Patched (Ptc) and Smoothened (Smo). Germline line clones and ectopic expression experiments show that the Hh receptor Patched (Ptc) is required in the PGCs for proper migration (Deshpande et al. 2001). When ptc function is compromised in PGCs, they tend to clump together as would be expected if they constitutively “receive” a signal. Reception of the Hh ligand by the PGCs results in the internalization of Hh and Ptc into large intercellular puncta (Kim et al. 2021) whereas the Smoothened (Smo) protein associates with the membrane in an actin-rich protrusion at the leading edge of the migrating PGCs (Kim et al. 2021). Contrary to Ptc, when Smo activity is disrupted, the PGCs scatter in the mesoderm (Deshpande et al. 2001). The non-canonical target for Smo activation is a G-protein coupled receptor, Trapped in Endoderm (Tre1). Like Smo, Tre1 localizes to the protrusions at the migration front together with dPIP5k, the enzyme responsible for generating PI(4, 5)P₂ (Kim et al. 2021). The phosphatidylinositol 4-phosphate 5-kinase (PIP5k) targeted to the migration front by Tre1 induces the local accumulation of PI(4, 5)P₂ which, via WASP, activates Arp2/3 dependent actin polarization at the leading edge. In tre1 mutants PIP5K does not localize to the migration front and as a result the production of PI(4, 5)P₂ does not become polarized. Likewise, actin is not preferentially polymerized at the migration front, thus linking the downstream events to the initial targeting of Smo to the membrane at the migration front in response to the reception of the Hh ligand (Kim et al. 2021).

While Hh expressed by the SGPs induces PGC migration towards the SGPs by functioning as an attractant, there are two related lipid phosphate phosphatases (LPPs) wunen (wun) and wunen-2 (wun2) that guide PGC migration by generating, in some as yet unknown manner, the repulsive signals (Starz-Gaiano et al. 2001; Hanyu-Nakamura et al. 2004; Renault and Lehmann 2006). Both wun and wun2 are maternally deposited and the maternal products are found in both the soma and the PGCs. They are also zygotically expressed in the epidermis, gut primordium, and nervous system. Intriguingly, wun or wun2 specific transcripts are missing from the blind end of the gut primordium, a site which coincides with the region where the PGCs undergo transepithelial migration across the gut. Both ‘loss’ and ‘gain’ of function experiments have indicated that Wun(s) in the soma (of maternal and zygotic origin) function to repel migrating PGCs, and they avoid invading the ‘wrong’ tissues because of the presence of Wun(s). For instance, PGCs bypass the nervous system due to endogenous expression of wun(s) in the ventral nerve chord (Zhang et al. 1997; Hanyu-Nakamura et al. 2004; Renault et al. 2004; Renault et al. 2010; Leblanc and Lehmann 2017). Conversely, when either wun or wun2 is expressed in the mesoderm, PGCs are unable to enter the mesoderm. In addition to this non-autonomous repulsive function, the Wun(s) are also required in migrating PGCs where they contribute to their individualization i.e. PGC-PGC repulsion and help facilitate their proper migration (Renault et al. 2004; Renault et al. 2010).

Though the wun(s) have been studied extensively since their discovery in the ‘90s (Zhang et al. 1997; Hanyu-Nakamura et al. 2004; Renault et al. 2004; Renault et al. 2010; Leblanc and Lehmann 2017) the mechanistic basis for their repulsive activity in the soma and their cell autonomous function(s) in PGC has remained a mystery. One candidate target that could potentially connect both their non-autonomous function in the soma and their cell autonomous function in PGCs is the Hh signaling pathway. Here we show that the wun(s) function to suppress Hh signalling both in the soma and in migrating PGCs. In the soma, their activity counteracts hmgcr-dependent potentiation of Hh transmission. In PGCs, the wun(s) appear to function to suppress the response to the reception of the Hh ligand. In addition, we demonstrate that, Suppressor of actin 1 (Sac1), a lipid phosphatase, plays a similar role in PGC migration by downregulating Hh signalling. To better understand the activities of the two enzymes involved in lipid metabolism, Wun(s) and Sac1, we compared unbiased lipidomes from the respective embryonic extracts. Intriguingly, these lipid metabolism enzymes appear to exert their effects on the Hh signaling pathway by targeting different phospholipids (Burnett and Howard 2003; Burnett et al. 2004).

Results

Embryos derived from mothers compromised for wun display enhanced levels of Hh targets

In stage 8-11 embryos, there is a positive autoregulatory loop in the embryonic ectoderm in which hh activates wingless (wg) expression in Hh-receiving cells, and wg in turn upregulates the expression of engrailed (en) in hh-expressing cells (Fig. 1A-B). We previously showed that mutations in hmgcr disrupt this autoregulatory loop, leading to significant downregulation of En expression. Thus, if the Wun(s) function to suppress the hh signaling pathway, we might expect to observe an opposite phenotype—namely that En protein expression will be upregulated when their activity is compromised.

Figure 1.

As a test of this idea, wun was maternally (M-) depleted, by driving the expression of UAS-wunRNAi in the adult female germline using nos-Gal4 (Fig. 1 Fig. 1C-E). We then examined En protein expression in the progeny (F2) of these females (mated to wild-type males). At stage 10, En expression in these wun^M- embryos is appreciably increased compared to wild-type (Fig.1, C1 versus D1, E1). In addition to a higher level of expression in individual cells, there is a readily discernable increase in the number of En-positive cells per stripe (Fig. 1, C1’ vs D1’, E1’). Upregulation of En expression together with a broadening of En stripe is also observed in wun2^M- embryos (generated using the same strategy) although the phenotype is weaker compared to wun^M- (Suppl. Fig.1). Thus, compromising the activity of either wun or wun2 maternally is sufficient to upregulate Hh signaling in the ectoderm.

In addition to being maternally deposited, the wun(s) are zygotically expressed in the gut endoderm and in a segmentally repeating pattern in the embryonic ectoderm. To determine whether zygotically expressed wun also functions to suppress Hh signaling, we mated females carrying a maternal-αtubulin Gal4 transgene (Mat Gal4) to UAS-wunRNAi males. Suppl. Fig. 2 shows that En expression is upregulated in these F1 embryos as compared to the control.

These findings indicate that both maternal and zygotic components of wun function to suppress hh signaling. If this is correct, then ectopic expression of wun in the zygote should suppress hh signaling and result in a decrease in En expression. To test this prediction, we mated females carrying Mat Gal4 to males carrying a UAS transgene encoding a wunen-GFP fusion protein. As shown in Suppl. Fig. 2C, zygotic overexpression of wun (Wun^OE) in the embryo results in a decrease in En protein expression.

Embryos zygotically compromised for wun display increased levels of β-galactosidase driven by patched-Gal4

The interaction between Hh and Ptc results in the translocation of the Smo protein to the membrane and signal transduction (Hu and Song 2019; Ingham 2022; Zhang and Beachy 2023) (Fig. 1A, B). In the canonical pathway, this results in the transcriptional upregulation of ptc via Cubitus interruptus (Ci). Ptc upregulation provides a negative feedback loop that limits Hh activity and spread.(Chen and Struhl 1998; Briscoe et al. 2001; Casali and Struhl 2004) If wun functions to suppress this branch of the canonical hh pathway, compromising wun activity zygotically (Z-) should result in a detectable increase in ptc expression. We mated females carrying a ptc-gal4 UAS-β-gal (Johnson et al. 2000) recombinant to either control males or males carrying the UAS-wun RNAi transgene and stained the resulting (F1) embryos with anti-β-galactosidase (anti-β-gal) antibodies. The control embryos displayed a segmental pattern that mimics ptc expression (Fig. 1H). By contrast, the ptc-gal4 UAS-β-gal/UAS-wun RNAi embryos (Fig. 1I-J) displayed a considerable increase in the anti-β-gal specific signal in nearly all the embryos.

Simultaneous overexpression of wun in the nervous system can mitigate PGC migration defects induced by elav-gal4 UAS-hmgcr

While the experiments involving RNAi-based embryonic knockdown suggested that wun(s) function to down regulate the hh signaling pathway in embryos, they do not provide insights into the underlying mechanism. Interestingly, similar to wun(s), another phospholipid phosphatase, sac1, functions to attenuate the hh signaling pathway (Yavari et al. 2010). Studies in wing and eye discs have argued that sac1 downregulates the hh pathway, in the receiving cells, after reception of the Hh ligand. While a similar cell-autonomous mechanism downstream of reception could be deployed by the wun(s), their ability to inhibit PGC migration into endodermal and ectodermal tissues suggests that they could also suppress the transmission of the Hh ligand. (The cell-autonomous and non-cell autonomous functions of Wun(s) need not be mutually exclusive). If this model is correct, then wun(s) would counteract the potentiation of Hh transmission by hmgcr.

Expression of UAS-hmgcr in the embryonic CNS using an elav-Gal4 driver induces PGC migration defects. These PGC migration defects can be substantially enhanced by co-expressing UAS-hh in the CNS while they are suppressed by co-expressing UAS-hhRNAi (Deshpande et al. 2023). Likewise, RNAi knockdown in the CNS of two genes required for Hh transmission, ttv and disp also suppressed the PGC migration defects induced by ectopic hmgcr (Deshpande et al. 2023). To explore the idea that wun functions to attenuate the transmission of the Hh ligand, we examined if overexpression of wun (UAS-wun) is also able to mitigate the PGC migration defects induced by ectopic expression of hmgcr in the embryonic nervous system using the elav-Gal4 driver (embryos derived from the elavGal4UAS-hmgcr recombinant strain).

Stage 13-14 embryos were classified based on the number of mis-migrated/scattered PGCs: 0-4, 5-9, 10-14 and 15+ mis-migrated PGCs. Compared to control embryos (Fig. 2A and 2D), about 65% of elav-gal4 UAS-hmgcr/+ embryo had 10 or more PGCs scattered/mis-migrated (Fig. 2B and 2E). As observed for hh, ttv and disp RNAi knockdowns, co-expression of Wun suppresses the PGC migration defects induced by hmgcr. Only about 20% of elav-gal4 UAS-hmgcr/UAS-wunGFP embryos had 10 or more mis-migrated PGCs (Fig. 2C and 2F). The suppression of mis-migration induced by elav-gal4UAS-hmgcr by simultaneous overexpression of wun is quantified in Fig. 2G.

Figure 2.

wun^M- embryos display elevated levels of endogenous Hh-GFP in the PGCs and surrounding mesodermal cells

When hmgcr is co-expressed with hh-GFP in the embryonic CNS using an elav-Gal4 driver, it substantially enhances the transmission of the Hh-GFP ligand encoded by the transgene into and through the underlying mesoderm, where it decorates the migrating PGCs (Deshpande et al. 2013). Thus, a clear implication of the hmgcr-wun epistasis experiments in the previous section is that the wun(s) are able to negatively regulate the hh signaling pathway because they block the efficient transmission of the Hh ligand. If this inference is correct, the transmission of endogenously expressed Hh should be substantially increased when wun(s) activity is compromised. To test this idea, we used the nos-Gal 4/UAS-wun RNAi combination to deplete maternal deposition of wun. Females carrying this transgene combination were then mated to males carrying a BAC hh-GFP rescue construct (henceforth referred to as Hh-GFP) which recapitulates the pattern of expression of the endogenous hh gene (Chen et al. 2017). As shown in Fig. 3A-B, there is substantial increase in the transmission of the Hh ligand in wun^M- embryos, as suggested by increased Hh accumulation in the somatic mesoderm. As previously reported by (Kim et al. 2021) for wild type hh-GFP embryos, a substantial fraction of embryonic PGCs, display detectable levels of Hh-GFP from stage 9/10 onwards (Fig. 3D). Moreover, as predicted, there appeared to be a significant elevation in the number of Hh-GFP-puncta per PGC in the wun^M- embryos (Fig. 3E-3F). To better analyze the change, we quantitated the total number of readily discernible Hh-GFP puncta per PGC in the control (UAS-lexA RNAi) and wun^M- embryos. In control embryos, about 90% of the Hh-GFP signal-positive PGCs have 5 or less puncta and of these about 50% have only 1 or 2 Hh-GFP puncta. Since PGCs do not express Hh, the increase in Hh-GFP puncta/PGC in wun^M- embryos must, at least in part, be due to the upregulation of Hh-GFP transmission from the surrounding mesodermal and ectodermal tissues. To confirm this idea, we measured the frequency of Hh-GFP puncta in a region (ROI) dorsal to the gut that is devoid of PGCs. As shown in Fig.3C, the number of Hh-GFP puncta in this region of the mesoderm is greatly enhanced in wun^M- embryos. Taken together with the results in the previous sections, these findings indicate that wun(s) function to downregulate or block the transmission of Hh from cells that synthesize the ligand.

Figure 3.

Since the wun(s) are also required in PGCs (Renault et al. 2004; Renault et al. 2010), we cannot exclude a complementary cell autonomous effect in which the loss of maternally derived wun in migrating PGCs also contributes to the increase in PGC associated Hh-GFP puncta.

Aberrant localization of Smo in PGCs of wun^M- embryos

The reception of the Hh ligand results in the translocation of Smo from its cytoplasmic localization to the cell membrane. In PGCs Smo localizes to membrane protrusions at the leading edge of the migrating cells (Kim et al. 2021) and mediates the recruitment of the Tre1 GPCR receptor. Tre1 induces localized actin polymerization in response to a transduction pathway the produce PI(4,5)P2. Since Hh signaling to the migrating PGCs is polarized, there is typically only a single sharp focus of Smo protein which is localized at the leading edge (Fig. 4, A1-A4). Typically, wild-type PGCs display a single sharp focus with the residual protein being associated with the rest of the membrane. This pattern of localization is disrupted in wun^M- (Fig. 4, B1-B4; C1-C4). To better understand the aberrant behavior, we classified individual PGCs based on the pattern of Smo localization. The classes included (i) PGCs showing no Smo-specific signal, (ii) PGCs showing a single focused crescent of Smo at the leading edge (iii) PGCs displaying diffuse localization of Smo across the membrane (iv) PGCs having a combination of a leading edge and a diffuse localisation of Smo (v) PGCs with multiple foci enriched for Smo-specific staining. In control PGCs, most PGCs have a single focus or a single focus surrounded by a more diffuse pattern of localization (∼60%). In contrast, less than 40% of the PGCs in wun^M- embryos have a single focus or a single focus surrounded by a more pattern of localization. About 35% show strong staining across the entire cell, while over 10% have multiple Smo foci. The changes in Smo localization are quantitatively significant (Fig. 4D).

Figure 4.

PGC migration defects in wun^M- embryos mimic those in ptc^M-Z+ embryo

In earlier studies we found that PGCs in stage 11 embryos generated by mating ptc^- germline clone females to wild type males tended to get stuck in the gut and/or clump together (Deshpande et al. 2001). Since loss of ptc would result in the constitutive activation of the hh signaling pathway, one might imagine that the excessive multi-directional hh signaling evident in wun^M- embryos could result in the PGC phenotypes observed in the absence of ptc. To test this idea directly, we maternally depleted ptc and the wun(s) using the same RNAi strategy described above (nos-Gal4/UAS-ptcRNAi, nos-Gal4/UAS-wunRNAi and nos-Gal4/UAS-wun2RNAi females mated to wild type males) Suppl. Fig. 3 and Fig. 5. As controls, we generated embryos from mothers expressing lexARNAi (nos-gal4/UAS-lexARNAi).

Figure 5.

As shown in Suppl Fig. 3 we find that individualization of the PGCs in nearly half of the ptc^M- embryos (45%) is hindered, resulting in an inefficient exit across the midgut primordium. Consequently, PGCs lag behind in the gut forming small clumps. In a small but significant proportion of the ptc^M-embryos PGCs form a large clump inside the gut (Suppl. Fig. 3B). Similar phenotypes are observed in stage 10-13 embryos derived from nosGal4/UAS-wun RNAi females. In >55% of wun^M- embryos the individualization of PGCs was hindered, resulting in a delay in exiting the gut (Fig. 5C-G). In about 15% of embryos, the PGCs clump together within the midgut (Fig. 5B, G). Maternal depletion of wun2 results in similar PGCs phenotypes to those observed in wun^M-embryos; however, they are substantially more penetrant. Over 60% wun2^M- embryos have large clumps of PGCs either within the gut or along the dorsal surface of the gut (Fig. 5F, H). In another ∼15% of the wun2^M- embryos, several PGCs are delayed in exiting the gut.

wun or wun2 are required in PGCs during the initial phase of migration

The results in the previous section indicate that the wun(s) play a key role in directing PGC migration at a very early step when the cells first begin to exit the midgut. Since the maternal RNAi knockdowns results in a substantial upregulation in the transmission of the Hh ligand throughout the embryonic soma, this early PGC defect could simply be a consequence of excessive somatic hh signaling. On the other hand, maternal knockdown of the wun(s) will also reduce the activity of the wuns in the PGCs themselves. In this case, the early phenotypes could also reflect a PGC-autonomous requirement of the wun(s). To test this possibility, we used nos-Gal4 to drive UAS-wunRNAi (Fig. 5D) or UAS-wun2RNAi (Fig. 5G) in migrating PGCs and stained F1 embryos with anti-Vasa. In both cases we observe phenotypes similar to those in the maternal knockdowns, though they are not as severe. For the wun knockdown, PGC clumping was observed in about 5% of the embryos, while lagging PGCs were observed in about 20% of the embryos (Fig. 5E). The knockdown of wun2 was more effective: Premature clustering of the PGCs was observed in nearly 10% of the embryos, while 50% of the embryos had lagging PGCs (Fig. 5H). These findings confirm the idea that the wun(s) are required in PGCs for proper migration and suggest that this function is likely related to the proper functioning of the hh signaling pathway.

Sac1, a known lipid phosphatase, downregulates Hh signaling in the embryonic contexts including in the PGCs

As mentioned above, the wun(s) are not the only phospholipid phosphatase that function to downregulate hh signaling. The lipid phosphatase, Sac1, was found to negatively regulate the hh pathway in wing and eye discs (Yavari et al. 2010; Del BEL et al. 2018). As Sac1 is also expressed in the embryonic gut primordium and the ectoderm we decided to test if maternal loss of sac1 results in embryonic phenotypes similar to those observed for the wuns. We first examined the expression En in the ectoderm. Suppl Fig. 4 shows that as in the case of knockdown of the wun(s), both the width of the En stripes and the nuclear signal intensity are enhanced in sac1^M- embryos.

Next, we tested whether the PGCs from sac1^M- embryos display PGC migration defects. We find that the PGCs in sac1^M- embryos also exhibit migration phenotypes, though they are less severe than those in wun^M- or wun2^M-. PGCs lagging in the midgut are observed in stage 10 sac1^M- embryos, and small clumps of PGCs are seen in the midgut of some stage 11 sac1^M-embryos (Fig. 6; compare A, B with D, E respectively). In stage 13 embryos, there are scattered PGCs, and the number of PGCs coalescing into the embryonic gonad is reduced (Fig. 6C versus 6F). The same phenotypes are observed when nos-Gal4 is used to drive paternally derived UAS-sac1RNAi in the migrating PGCs (Fig. 6; G, H vs I, J respectively). Interestingly zygotic knock down of reciprocally active Kinase, stt4, also leads to PGC migration defects (Yavari et al. 2010) (Fig. 6 K-M; See Suppl. Fig. 5 for quantitation). Furthermore, supporting a genetic interaction between wun and sac1, a higher proportion of embryos displayed precocious clustering of PGCs (Fig. 6 P) and Smo mislocalization at germ band extension stage (Suppl. Fig. 6).

Figure 6.

Endogenous Hh-GFP specific puncta in the mesoderm and the adjacent PGCs are enhanced in sac1^M-embryos

The analysis of sac1 function in the eye and wing discs suggested that it functions in receiving cells to inhibit the membrane association of Smo. When sac1 activity was compromised in receiving cells, Smo association with the membrane was substantially enhanced. Moreover, in this developmental context the phenotypes induced by the loss of sac1 activity were found to be independent of the Hh ligand. Consistent with a similar cell autonomous role, we found that knockdown sac1 activity in PGCs induced migration defects. However, since we also observed PGC migration phenotypes when sac1 activity is maternally depleted, it seemed possible that sac1 might also have a cell non-autonomous function in the hh signaling pathway analogous to wun(s). To explore this possibility, we mated Mat Gal4/UAS-sac1-RNAi females with hh-GFP males. The experiments in Fig. 7 show that somatic loss of sac1 results in a substantial upregulation in the transmission of the Hh ligand. In panels A-C (Fig. 7) we examined the accumulation of Hh-GFP puncta in the mesoderm. As indicated by the quantitation in panel C, the number of Hh-puncta is increased ∼2.5 fold. Accompanying this increase in mesodermal Hh-GFP puncta, we find that there is a significant increase in the number of Hh-GFP puncta per PGC in sac1^M-. (Fig. 7F). About 55% control PGCs had 1-3 Hh-GFP puncta as opposed to ∼20% of the sac1^M- PGCs. ∼ 80% of sac1^M- PGC had 4 or more Hh-GFP puncta compared to about 53% in control. More importantly, none of the PGCS in control embryos had 10+ Hh-GFP puncta. Altogether, our data shows that maternally compromising sac1 activity results in excess accumulation of Hh-GFP in the mesoderm and PGCs.

Figure 7.

Comparative unbiased lipidomics of wun^M- and sac1^M- embryonic extracts

While depletion of the wun(s) and sac1 give a similar range of hh signaling pathway phenotypes, the substrates for these enzymes involved in lipid metabolism are likely different. The preferred substrate for sac1 is phosphatidylinositol 4-phosphate (PI4P) which it dephosphorylates to phosphatidylinositol (PI) (Yavari et al. 2010).

While the wun(s) share homology with a mammalian family of lipid phosphate phosphatases (LPP1, LPP2 and LPP3), biochemical experiments suggest that their substrate preferences are likely different. The mammalian LPP family proteins dephosphorylate a broad range of phospholipids, including among others lysophosphatidic acid (LPA), phosphatidic acid (PA) and ceramide-1-phosphate C(1)P (Dillon et al. 1997; Jasinska et al. 1999). Unlike the mammalian LPP family members, (Burnett and Howard 2003) found that wun is unable to dephosphorylate either PA or C(1)P. However, wun was able to dephosphorylate LPA with a similar efficiency to that of the mammalian LPP1 and LPP3 enzymes. To identify possible targets for the wun and sac1 we performed unbiased lipidomics on embryonic extracts derived from wun^M-and sac1^M-4-8 hr embryos (for details, see Materials and Methods).

Figure 8.

Because compromising sac1 has been reported to increase PI4P levels, we initially focused on phosphoinositide levels. Notably, embryos maternally compromised for sac1 showed increased levels of various phosphoinositide species (estimated and collectively represented as PI) compared with control samples. Evidently, PI levels are not appreciably altered in wun^M- embryos. Surprisingly, however, DAG levels are considerably elevated in wun^M- extracts as compared to either the control or the sac1^M-samples. Membrane-bound DAG subsequently participates in a variety of physiological alterations (Nilssen et al. 2005; Zhong et al. 2022; Falzone and Mackinnon 2023)

Since DAG is not phosphorylated, its accumulation in wun^M- extracts cannot be attributed directly to the loss of wun activity. Instead, the accumulation of elevated levels of DAG must be an indirect consequence of the loss of wun function. Elevated levels of DAG are typically generated by the activation of phospholipase C, which breaks down PIP2 into DAG and IP3. Typically, signaling pathways that utilise GPCRs can turn on phospholipase C by promoting the dissociation of Gα from the Gβγ dimer. Phospholipase C can also be activated by excessive levels of lysophosphatidic acid (LPA) which is the only known wun substrate (Burnett and Howard 2003). Thus, the specific and robust increase in DAG levels upon compromising Wun activity dovetails nicely with aberrant localization of Smo, a GPCR involved in mediating Hh signal. Together these data support the notion that Wunen, likely employs Hh and the downstream components of the signaling cascade to engineer germ cell repulsion.

Discussion

Migrating cells in different developmental contexts are exposed to a plethora of signals in their cellular environment. Only a select few of these signals has the ability to guide motility of these cells and can function as either attractive or repulsive cues. Conversely, the migrating cells are also endowed with a restricted and well-calibrated ability to respond selectively to either move away from or towards the signal. For instance, axonal migration in vertebrate embryos is guided by attractive influence of Netrins and Sonic-hedgehog whereas they are repelled by BMP ligands (Tessier-Lavigne 1994; Butler and Dodd 2003; Charron et al. 2003; Wu et al. 2019). PGC migration in Drosophila is engineered by Hmgcr and Wun dependent attractive and repulsive influences respectively. While Hh acts as an attractive ligand for migrating PGCs, no specific repulsive signal has been identified thus far. It has been proposed that components of the Hmgcr dependent enzymatic pathway harness attractive potential of SGP derived Hh signal. Data presented here document that repulsive influence mediated by wun also targets Hh, however, in a qualitatively opposite, antithetical manner. In the following, we will discuss how a competitive reciprocal balance between hmgcr and wun modulates Hh activity to carve the precise path of migrating PGCs.

The functioning of the hh signaling pathway in guiding the migration of the PGCs towards the SGPs is now reasonably well understood. Hh expressed by the SGPs is received by the migrating PGCs and its reception is transduced into their directed motility towards the source of the Hh. The response to the reception of the Hh ligand is highly polarized. Consequently, Smo is translocated from the cytoplasm to the membrane where it accumulates in a single sharp focus at the leading edge of the migrating PGCs. Smo subsequently recruits the GPCR Tre1 to the membrane and Tre1 activates F-actin assembly at the leading edge through the production of the signaling molecule PI(4,5)P₂ (Kim et al. 2021).

To ensure that Hh ligands produced by the SGPs are preferentially received by the PGCs is especially critical to the deployment of hh as a guidance molecule in this context. Previous studies identified one such mechanism, namely the potentiation of Hh transmission. Though hh is widely expressed in ectodermal and mesodermal tissues throughout much of the embryo, its ability to attract migrating PGCs is primarily limited to the Hh produced by the SGPs (and their progenitors). One critical factor that distinguishes Hh produced by the SGPs from Hh expressed elsewhere in the embryo is hmgcr. Hmgcr and the downstream isoprenoid biosynthetic pathway function to greatly enhance the long-distance transmission of the Hh ligand. When hmgcr is expressed in other cell types or tissues, the endogenous Hh ligand expressed in these cells (or tissues) is able to compete with the Hh ligand produced by the SGP and induce the mismigration of the PGCs. Thus, key to the potentiation activity of hmgcr is the fact its expression becomes progressively restricted to the SGPs during PGC migration. In addition to hmgcr several other genes important for the transmission are expressed in the mesoderm or posterior mesoderm (Ricardo and Lehmann 2009; Deshpande et al. 2013; Deshpande et al. 2016). However, the potentiation of Hh transmission from the SGPs (and their mesodermal progenitors) by hmgcr is not in itself sufficient. We show here that the long-distance transmission of hh expressed in other cell types and tissues in the embryo must also be specifically suppressed in order to ensure that the migrating PGCs make their way to the SGPs. The suppression of hh signaling is mediated by the two Wun(s), Wun and Wun2. It was initially discovered that Wun functioned to repel migrating PGCs (Zhang et al. 1996; Zhang et al. 1997). Subsequent studies identified the neighboring gene wun2 (Starz-Gaiano et al. 2001; Hanyu-Nakamura et al. 2004) and showed that it also functions to repel migrating PGCs. The two wun(s) encode lipid phosphate phosphatases and exhibit similar patterns of expression. Both gene products are maternally deposited and are present in the soma as well as the PGCs. In addition, both genes are zygotically expressed in the epidermis, gut primordium, and nervous system, tissues that migrating PGCs usually avoid.

While the repulsive activity of the wun(s) has been well documented, the mechanism of their action was not known. Using both the ‘loss’ and ‘gain-of-function’ experiments here we demonstrate that Wun(s) “repel” migrating PGCs because they are able to suppress the transmission of the Hh ligand—and this suppression of Hh transmission from other cell types and tissues enhances and refines the ability of Hh produced by the SGPs to attract the migrating PGCs. Perhaps the epistasis experiment presented in the manuscript constitutes the most direct demonstration of this mechanism of action involving antithetical activities of Hmgcr and Wun on Hh transmission. When hmgcr is ectopically expressed in the CNS, it induces PGC mismigration because it substantially enhances the transmission of the Hh ligand that is normally produced in the CNS. For this reason, the migration defects induced by ectopic hmgcr can be suppressed by co-expressing hh-RNAi with hmgcr. A similar suppression is observed for wun—in this case when wun is co-expressed with hmgcr using the same CNS driver.

Consistent with an ability to antagonize the hh signaling pathway, we find that when either wun or wun2 is maternally depleted the hh signaling pathway is substantially upregulated. This enhanced activity is evident in the pattern of En expression in the embryonic ectoderm during mid-embryogenesis: there is an increase in the level of En expression and in the number of cells expressing En. Visualization of Hh produced by a BAC:hh-GFP rescue construct indicates that the transmission of Hh is upregulated when maternal wun is depleted. In a wild-type background Hh-GFP puncta decorate a subset of the migrating PGCs. In wun^M- embryos, the number of Hh-GFP puncta/PGCs increases. Excess Hh-GFP puncta are also seen in the mesoderm, indicating that there is a general enhancement in Hh transmission. The reception of the Hh ligand in migrating PGCs is also substantially altered. Instead of there being one primary membrane associated Smo focus, nearly half of the PGCs in wun^M- embryos have high levels of Smo all around their membranes or multiple Smo foci.

The enhanced transmission of the Hh ligand when wun is maternally depleted together with the hmgcr/wun epistasis experiments suggest that the Wun(s) function to mitigate the hh signaling pathway in hh expressing cells. On the other hand, the Wun(s) are known to be required not only in the soma but also in the PGCs themselves. In this case, the mislocalization of Smo in PGCs of wun^M- embryos could be due at least in part to an inappropriate cell autonomous activation of the hh signaling pathway in the receiving cells. As cytonemes extend from both Hh producing and Hh receiving cells in the imaginal discs, it is also conceivable that the loss of maternal wun in PGCs contributes to the increase in their labeling by Hh-GFP puncta. Two lines of evidence provide additional support for the idea that the hh pathway is upregulated in both the sending and receiving cells. The first would be the finding that ptc-like PGC migration defects are observed in both maternal wun/wun2 RNAi knockdowns and in wun/wun2 RNAi knockdowns in migrating PGCs. The second is the finding that a loss of another lipid phosphatase, sac1, also induces a similar spectrum of PGC migration defects. Consistently, as in the case of wun, embryos compromised for sac1 display PGC migration defects between stages 10-13 including retention of PGCs in the gut and reduction of PGC count in the coalesced gonads in a subset of embryos. Furthermore, as observed in wun^M- embryos, sac1^M-embryos also display an increased Hh association with migrating PGCs and an overall increase in the level of Hh in the surrounding mesoderm.

Analogous phenotypic consequences of the embryonic knockdowns of wunen(s) and sac1 and their respective influence on Hh signaling on the embryonic PGCs are interesting in their own right. These observations coaxed us to revisit the reportedly crucial involvement of lipid metabolism and function vis-a-vis Hh signaling, PGC migration and their functional relationship. Wunen(s) have been characterized as lipid phosphate phosphatases (LPPs) based on their homology to the vertebrate LPPs. Substantial genetic characterization including rescue experiments has further supported the notion that LPP3 (human) and Wunen(s) likely share targets. By comparison, Sac1 has a relatively narrower and specific target range. Especially in the context of Hh signaling, it balances the Stt4 kinase activity to maintain the levels of PI4, which in turn, regulates membrane localization of Smo to modulate signaling downstream of Hh. Consequently, the phenotypic similarities between wun and sac1, in the context of PGC migration are intriguing.

Hence, we decided to assess whether they shared targets that may explain the similar phenotypes. The comparison of unbiased Lipidomics confirmed a few expectations but also threw up some unanticipated surprises. Foreseeably levels of different phosphoinositide species were substantially elevated upon the embryonic loss of sac1. Surprisingly however, a similar loss of wun did not alter PI or PA levels rendering them unlikely to be biologically relevant entities. Our current evidence supports the notion that Wunen(s) may function more broadly as a lipid metabolic or lipid-modifying enzyme(s), rather than lipid phosphatases per se. In this regard it is noteworthy that a specific and significant increase in DAG levels was observed upon loss of wun. The possible mechanistic underpinnings of this increase and its relationship with the phenotypic consequences are tantalizing as DAG accumulation is one of the indicators of activation of GPCR signalling (Nilssen et al. 2005; Zhong et al. 2022; Falzone and Mackinnon 2023). Of note, Smo and Tre1, the two receptors demonstrated to act downstream of Hh signaling during PGC migration, are members of the GPCR family. Precocious clumping observed upon loss of Wun(s) could thus be an outcome of premature activation of one (or both) of these receptors.

Drosophila PGCs migrate in a calibrated manner across a complex cellular landscape. To arrive at their destination, germ cells ought to possess the ability to respond to the ‘right’ signal and ignore the ‘wrong’ ones. Moreover, PGCs must distinguish between the right signal emanating from the wrong source. Wunen(s) contribute to this ability as they constitute the other side of the coin. They hinder the ability of Hh to function as an attractive force if it is emanating from a ‘wrong source’. The antithetical abilities of Hmgcr and Wun thus allow to fine tune the Hh signaling by potentiating the SGP derived ‘right’ signal and squelching the one emanating from the ‘wrong’ sources like ectoderm, nervous system etc. Moreover, this is unlikely to be an isolated instance and thus harbors tremendous potential as various disease conditions including metastasis involve long distance migration and inappropriate adhesion and lodging. Future experimental strategies will help elucidate mechanistic underpinnings of these events in the context of PGC migration and may also pave the way to understand its importance in other contexts where directed cell migration is a critical determinant of either normal or aberrant development.

Materials & Methods

Fly husbandry and stocks

Flies were raised on standard cornmeal agar at 25 °C unless stated otherwise. The Drosophila melanogaster lines, with IDs, used are: CS (BDSC_64349), UAS-lexA RNAi (BDSC_67947), UAS-eGFP (BDSC_41552), UAS-ptc RNAi (BDSC_55686), UAS-smo RNAi V20 (BDSC_53348), UAS-smo RNAi V22 (BDSC_43134), nos-gal4 VP16 (BDSC_4937), UAS-sac1 RNAi (BDSC_56013), UAS-sac1 RNAi ( VDRC #’s 44376 and 37216), maternal-α-tubulin-gal4-VP16 (Mat-gal4 on both II and III, Gift from Daniel St. Johnston), ptc-gal4 UAS-lacz (Gift from Matthew P Scott), Hh:GFP Bac (Gift from Thomas B. Kornberg), UAS-wun-GFP (Gift from Ken Howard),

For maternal knockdown experiments, we crossed UAS-X RNAi males, where ‘X’ is the gene to be knocked down, to maternal gal4 driver females (either mat-gal4 VP16 or nos-gal4 VP16). F2 embryos were collected for study, and since these are maternally compromised, their genotype is denoted as X^M-. For zygotic knockdowns, denoted as X^Z-, UAS-X RNAi males were crossed to maternal gal4 drivers (mat-gal4 or nos-gal4) females, and F1 embryos were used for study. nos-gal4 VP16, unlike mat-gal4 V16 is enriched specifically in PGCs. For zygotic overexpression of wun, we used F1 embryos (denoted as wun^OE) derived from mat-gal4 mothers crossed to males carrying the UAS-wun-GFP transgene.

Embryo Fixation

Embryos were collected, typically aged 3-8 hours (with a focus on stages 10-11) or 0-12 hours (for overnight collections of later stages), in a sieve and dechorionated in 4% sodium hypochlorite for 90 seconds. After thorough washing with distilled water, the embryos were fixed in a 1:1 heptane:4% Paraformaldehyde (PFA) solution for 20 minutes, followed by removal of the PFA and a fresh heptane wash for 20 minutes. Furthermore, chilled methanol was added, and the embryos were devitellinized by vigorous shaking in a 1:1 heptane: methanol mixture. Devitellinized embryos settle at the bottom and are stored in 100% methanol at –20 °C until immunostaining.

Immunostaining and Imaging

For immunostaining, embryos were rehydrated by four washes with 0.3% PBS-Triton X-100 (PBST), followed by blocking in 2% bovine serum albumin (BSA) in 0.3% PBS-T (blocking solution) for 1 hour at room temperature. Embryos were incubated overnight at 4°C with primary antibodies diluted in fresh blocking solution at the appropriate antibody-specific dilutions. The following primary antibodies in specified dilutions were used: Rabbit anti-Vasa (1:1000, Gift from Trudi Schüpbach), Mouse anti-En (1:50, DSHB), Mouse anti-Wg (1:20, DSHB), Chicken anti-GFP (1:500, Invitrogen), Mouse anti-Smo (1:50, DSHB), Mouse (1:20), and Chicken anti-βgal (1:1000, Abcam). Following four washes with 0.3% PBS-T, the embryos were incubated in the appropriate secondary antibodies for 1 hour at room temperature. The following secondary antibodies were used in a 1:1000 dilution: goat anti-mouse AlexaFluor 488, goat anti-chicken AlexaFluor 568, goat anti-rabbit AlexaFluor 647 (1:1000, Invitrogen). Embryos were washed four times with 0.3% PBS-T, and Hoechst (1:1000 from a stock of 10 mg/ml) was added to the penultimate wash. All washes mentioned in the protocol are 10-15 mins each. Embryos were mounted in 20mM Tris pH=8, 0.5% n-propyl gallate, 90% glycerol and observed under a Nikon AX confocal microscope or an Evident inverted microscope (FV4000) using 40X and 20X objectives, respectively. Image processing was performed using the ImageJ program.

Lipidomics

Mat-Gal4 females were crossed with each of UAS-mcherry RNAi, UAS-wun RNAi, UAS-Sac1 RNAi (two lines L1 & L2). F2 embryos were collected for 0-4 hours and then aged at 25 °C for 4 hours. Embryos were washed in 1x phosphate buffered saline (PBS) and snap frozen until further processing. For lipid extraction, the embryos were thawed on ice, suspended in 550 μL of cold, sterile 1x PBS, pH 7.4 and homogenized using a tissue homogenizer (Bullet Blender24, Next Advance) with one scoop of glass beads (0.5-mm diameter; Next Advance) at a speed setting of 10 for 5 min at 4 °C. The protein concentration of each lysate was estimated using the BCA protein assay kit (Pierce) according to the manufacturer’s protocol. 500 μL of the lysates were used for lipid extraction using established methods as described earlier (Kumar et al. 2024). Briefly, for each sample, 500 μL lysate was transferred to a glass vial and vigorously mixed with 1.5 mL of 2:1 chloroform (CHCl₃): methanol (MeOH) containing an internal standard mix [1 nmol each of pentadecanoic acid (Sigma-Aldrich, Catalog No. P6125, for negative ion mode analysis) and C17:0/20:4 phosphatidylcho line (Avanti Polar Lipids, Catalog No. LM-1002, for positive ion mode analysis)]. The resulting mixture was centrifuged at 1500g for 15 min to separate the aqueous (top) and organic (bottom) layers, and the organic layer (bottom) was gently pipetted into a new glass vial. To enrich the extraction of phospholipids, the aqueous layer was acidified using 20 μL formic acid (Merck, Catalog No. 5.33002) and further vigorously mixed with 1 mL of CHCl_3. The resulting mixture was again separated by centrifugation at 1500g for 15 min, and the organic layer (bottom) was separated and pooled with the organic extract from the previous extraction. The pooled organic phase was dried under a stream of nitrogen gas, resolubilized with 500 μL of CHCl₃, and transferred to another glass vial to remove any trace aqueous contamination. The organic layer was dried again under a stream of nitrogen gas and resolubilized in 200 μL of 2:1 CHCl₃/MeOH, and 10 μL of this was injected into an Agilent 6545 Quadrupole Time-of Flight (QTOF) LC-MS instrument for untargeted lipidomics measurements in each ion mode (positive and negative). A standard untargeted lipidomics run in positive or negative ion mode lasted 60 min, and the LC columns, solvents, and solvent gradients were identical to those previously reported by us (Kumar et al. 2024). All LC-MS/MS experiments were performed using an electrospray ionization (ESI) source with the following parameters: drying and sheath gas temperature = 320 °C; drying and sheath gas flow rate = 10 L/min; fragmentor voltage = 150 V; capillary voltage = 4 kV; nebulizer pressure = 45 psi; and nozzle voltage = 1 kV. For the analysis of different lipids from the embryos, a Personal Compound Database Library (PCDL) curated for lipids was generated, and peaks corresponding to lipids were verified based on relative retention times, mass error threshold (20 ppm or less at MS1), and MS/MS fragments obtained. All lipid species were quantified by measuring the area under the curve relative to the respective internal standard and then normalizing this to the protein concentration of the respective sample.

Quantification & Statistical Analysis

The statistical analysis are detailed in figure legends. n in each graph represents the number of embryos, and N represents the number of replicates. To quantify PGC migration defects, embryos were stained with α-Vasa antibody and imaged. In embryos older than stage 13, PGCs away from the somatic gonads were counted as mis-migrating cells. For maximum-intensity projections, equivalent slices were used across genotypes to ensure the best representation.

Figure supplements

Supplementary Figure 1.

Supplementary Figure 2.

Supplementary Figure 3.

Supplementary Figure 4.

Supplementary Figure 5.

Supplementary Figure 6.

Data availability

All data generated or analyzed during this study are included in the manuscript and supporting files; source data files have been provided for all figures.

Acknowledgements

We thank: Bloomington Drosophila Stock Centre (BDSC), supported by NIH grant P40OD018537, for fly stocks; Snehal Patil and Yashwant Pawar for fly media and stock maintenance; Microscopy facility at IISER Pune, managed by Dr Santosh Podder and Vijay Vitthal; Saddam Shekh for technical assistance and maintenance of the biological mass spectrometry facility at IISER Pune.

Additional information

Funding

GD’s visits to IISER Pune (2023-2025) are supported by the Ministry of Education (MoE) Scheme for the promotion of Academic & Research Collaboration (SPARC), Grant ID SPARC-1587, managed by IIT Kharagpur, which facilitates collaboration between IISER Pune and Princeton University. Pratiksha Trust Extra-Mural Support for Transformational Aging Brain Research grant EMSTAR/2023/SL03 to GR, facilitated by the Centre for Brain Research (CBR), Indian Institute of Science, Bangalore. SSK acknowledges financial support from the Anusudhan National Research Foundation (ANRF), Government of India (Grants: Swarna Jayanti Fellowship SB/SJF/2021-22/01). SSK & GSR receive Intramural support from IISER Pune. The IISER Drosophila media and Stock Centre are supported by the National Facility for Gene Function in Health and Disease (NFGFHD) at IISER Pune, which in turn is supported by an infrastructure grant from the DBT, Govt. of India (BT/INF/22/SP17358/2016). AR is a graduate student supported by a research fellowship from the Department of Biotechnology (DBT), Govt. of India. AER is an undergraduate student supported by a KVPY/INSPIRE fellowship, Department of Science & Technology, Govt. of India.

Contributions

GD conceptualized the project, with inputs from PS, SK and GSR. AR executed the experiments, with contributions from AER, JD, AI and GD. KK conducted the Lipidomics experiments under SK’s supervision. GD wrote the initial draft, with input from AR, AER, KK, SSK, and GSR. All authors edited and refined the manuscript to its final form. GSR, GD, PS & SSK were responsible for funding acquisition, supervision and project management.

Funding

Pratiksha Trust EMSTAR (EMSTAR/2023/5L.03)

• Girish S Ratnaparkhi

SPARC (1587)

• Girish Deshpande

Anusudhan National Research Foundation (SB/SJF/2021-22/01)

• Siddhesh S Kamat