Bifocal and PP1 interaction regulates targeting of the R-cell growth cone in Drosophila

Developmental Biology 288 (2005) 372 – 386 www.elsevier.com/locate/ydbio

Bifocal and PP1 interaction regulates targeting of the R-cell growth cone in Drosophila Kavita Babu a,*,1, Sami Bahri b, Luke Alphey c, William Chia a,2 a

Temasek Life Science Laboratory and Department of Biological Sciences, 1 Research Link, National University of Singapore, 117604, Singapore b Institute of Molecular and Cell Biology, 61 Biopolis Drive, Proteos 138673, Singapore c Department of Zoology, Oxford University, South Parks Road, Oxford, UK Received for publication 16 February 2005, revised 7 September 2005, accepted 10 September 2005 Available online 8 November 2005

Abstract Bifocal is a putative cytoskeletal regulator and a Protein phosphatase-1 (PP1) interacting protein that mediates normal photoreceptor morphology in Drosophila. We show here that Bif and PP1-87B as well as their ability to interact with each other are required for photoreceptor growth cone targeting in the larval visual system. Single mutants for bif or PP1-87B show defects in axonal projections in which the axons of the outer photoreceptors bypass the lamina, where they normally terminate. The data show that the functions of bif and PP1-87B in either stabilizing R-cell morphology (for Bif) or regulating the cell cycle (for PP1-87B) can be uncoupled from their function in visual axon targeting. Interestingly, the axon targeting phenotypes are observed in both PP1-87B mutants and PP1-87B overexpression studies, suggesting that an optimal PP1 activity may be required for normal axon targeting. bif mutants also display strong genetic interactions with receptor tyrosine phosphatases, dptp10d and dptp69d, and biochemical studies demonstrate that Bif interacts directly with F-actin in vitro. We propose that, as a downstream component of axon signaling pathways, Bif regulates PP1 activity, and both proteins influence cytoskeleton dynamics in the growth cone of R cells to allow proper axon targeting. D 2005 Elsevier Inc. All rights reserved. Keywords: Bifocal; Protein phosphatase-1 (PP1); Axon guidance; Photoreceptors; Optic lobe; Cytoskeleton; F-actin

Introduction Precise axon targeting is achieved by integrating signaling events involving guidance receptors and their downstream components (Tessier-Lavigne and Goodman, 1996). Some of the molecules involved in axonal guidance in the Drosophila embryo also play a role in retinotopic axonal targeting, suggesting that a similar set of molecules governs axon guidance and normal topographic mapping of the axons in the embryo and the fly visual system. These include Dock (Desai et al., 1999; Garrity et al., 1996; Hing et al., 1999; Rao and Zipursky, 1998), Trio (Awasaki et al., 2000; Bateman et al., 2000; Luo, 2000;

* Corresponding author. E-mail address: [email protected] (K. Babu). 1 Present address: Department of Molecular Biology, Massachusetts General Hospital, Richard B. Simches Research Building, 185 Cambridge Street, CPZN 7250, Boston, MA 02114, USA. 2 Requests for reagents should be made to [email protected]. 0012-1606/$ - see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.ydbio.2005.09.017

Newsome et al., 2000b), Lar (Clandinin et al., 2001; Krueger et al., 1996; Maurel-Zaffran et al., 2001), Cadherins (Iwai et al., 1997, 2002; Lee et al., 2001), Ephrins and their receptors (Bossing and Brand, 2002; Dearborn et al., 2002; Palmer and Klein, 2003; Schmucker and Zipursky, 2001; Yu and Bargmann, 2001), Protocadherins (Lee et al., 2003; Senti et al., 2003) and Dptp69D (Desai and Purdy, 2003; Desai et al., 1996; Garrity et al., 1999; Newsome et al., 2000a; Sun et al., 2000). The Drosophila eye has 750– 800 repeats of a basic unit called an ommatidium and comprised of 8 photoreceptor cells (R1– R8). Overlying these photoreceptor cells are four cone cells and two primary pigment cells surrounding the cone cells. The eight axons from the photoreceptors of an ommatidial unit in the eye travel via the optic stalk into the optic lobe region of the brain as part of an ommatidium-specific fascicle with the axons of R1 – R7 surrounding the R8 axon (Hanson, 1993). Incoming R-cell axons induce the production and differentiation of glia in the optic lobe lamina (Huang and Kunes, 1998; Selleck and Steller, 1991; Winberg et al., 1992). Lamina

K. Babu et al. / Developmental Biology 288 (2005) 372 – 386

differentiation by R-cell-derived signals plays a crucial role in matching the number of afferents to their targets. Different R cells in the eye disc extend axons that terminate at different levels in the optic lobe. The 6 outer photoreceptors, R1 – R6, of each ommatidium terminate their axons at the outer layer of the optic lobe or the lamina, while the inner two photoreceptors R7 and R8 send their axons beyond the lamina into the second optic ganglion, the medulla (Hanson, 1993). The growth cone from an initial R8 neuron can be thought to pioneer the path to retinotopic target destination and enters the optic lobe first, while the growth cones of R2 and R5, R3 and R4, R1 and R6 and R7 follow in sequence along the expanding ommatidial fiber. This process of axonal targeting is apparent by the thirdinstar larval stage of Drosophila development, such that at this stage a very precise pattern of connectivity is apparent, with R1 to R6 axons ending at the lamina and R7 and R8 axons entering the medulla and sending their axons into two distinct neuropiles in the medulla (Kunes and Steller, 1993; reviewed by Kunes, 1999; Kunes et al., 1993). Targeting errors for R-cell axons have been described for several mutations including those of the receptor tyrosine phosphatase receptors such as dlar, dptp69d and off-track (Cafferty et al., 2004; Clandinin et al., 2001; Garrity et al., 1999). Downstream components of the phosphotyrosine signaling system, of which dock is a component, also play important roles in axon guidance in the visual system (Garrity et al., 1996). Targeting defects in these mutant backgrounds include irregular and uneven lamina layer. In addition, R1 – R6 axons that normally synapse at the lamina can bypass the lamina and enter the medulla or stop at the wrong points in the lamina or medulla (Clandinin et al., 2001; Desai et al., 1999; Garrity et al., 1996). Here, we show that Bifocal and PP1-87B are components of a genetic axon guidance pathway in the visual system involving the phosphatase receptors, Dptp69D and Dptp10D. The Bifocal protein is expressed in neuronal cells in both the embryonic nervous system and the larval visual system of the fly (Bahri et al., 1997; Helps et al., 2001). Although Bif is expressed on embryonic CNS axons, bif mutant embryos do not show any obvious defects in axon guidance. However, Bif has been shown to be required for the normal shape of the R-cell rhabdomeres in the fly eye (Bahri et al., 1997) and for anchoring of proteins to the posterior of the oocyte along with the F-actin-binding protein, Homer (Babu et al., 2004). In the visual system, we have previously shown that Bif is one of the regulators of phosphatase activity, and it directly interacts with PP1-87B via its consensus phosphatase-binding motif, RVQF (Helps et al., 2001). We have also shown that the Bif – PP1-87B complex is required for normal photoreceptor rhabdomere morphogenesis (Helps et al., 2001). Furthermore, Ruan et al. (2002) have shown that the loss of bif function can give rise to axon guidance phenotypes in the larval optic lobe and that in this system bif interacts with misshapen (msn), a gene which encodes for a protein kinase that regulates normal photoreceptor axon targeting (Ruan et al., 2002). Although Bif has been shown to function in R-cell guidance, several questions remained unanswered. For examples: (1) are the guidance defects an indirect consequence of a

373

lack of Bif function in rhabdomere morphogenesis? (2) Is binding to PP1 essential for Bif’s axon guidance role? (3) Does Bif form a complex with actin directly or indirectly? Here, we demonstrate that the axon guidance defects in bif mutants are not mere consequences of its effects on cell morphology. We also show that the PP1-binding site of Bif is essential for its function and that PP187B itself is also required for proper axon guidance. We further demonstrate a dominant genetic interaction between bif and PP1-87B, indicating that these two molecules may function in the same signaling pathway during normal photoreceptor axon targeting. Moreover, bif genetically interacts with the receptor tyrosine phosphatases Dptp10D and Dptp69D in the larval optic lobe, and it can directly bind to F-actin, suggesting that Bif may contribute to the propagation of signaling via these receptors to the actin cytoskeleton during the process of axon guidance. Materials and methods Genetics The protein phosphatase-1 mutants used in this study, pp1-87B hs46 , pp1-87B e211 and pp1-87B e078 , have been previously described by Axton et al. (1990). We recombined the mutants onto an FRT chromosome and examined the eye discs of transallelic larvae. These transallelic animals developed fairly normal eye discs and were used for further examinations. The transgenic fly lines carrying PP1 inhibitors and UAS-PP1-87B have been previously described (Parker et al., 2002). The bif mutants and bif transgenic lines, UAS-bif + and UAS-bif F995A , used in this study were previously described by Bahri et al. (1997) and Helps et al. (2001). The cDNA for the shorter Bif isoform called bif 10DA was cloned into the pUAST vector, and fly transformants were generated using standard genetic techniques. Ectopic expression of the various transgenes in the fly eye was performed using GMR-GAL4 (Hay et al., 1997). The Ro tau -lacZ marker was used to mark R2R5 axons in the larval brain, and the Rh1-slacZ and Rh4-slacZ markers were used for labeling R1 – R6 and R7 axons respectively in the adult brain (Garrity et al., 1999; Newsome et al., 2000a). The dptp10d alleles, dptp10d 1 and deletion Df(1)59 which deletes both bif and dptp10d, were previously described (Bahri et al., 1997; Sun et al., 2000). The dptp69d 1 allele and a deletion uncovering dptp69d were obtained from Bloomington Stock Center, and other mutant lines were obtained either from Bloomington or as gifts from various laboratories.

Immunohistochemistry and microscopy Larval eye discs/brain complexes were dissected and fixed in accordance with previously published protocols with minor modifications (Wolff and Ready, 1991). Wandering third-instar larval and 55 h pupal eye discs/brains were dissected in phosphate-buffered saline and fixed in 4% paraformaldehyde for 30 min on ice. The dissected brains or eye discs were washed, blocked with BSA or goat serum and stained with primary antibody overnight. After several washings, secondary antibodies were added, and the specimens were mounted for microscopic observation. Several primary antibodies were used: rabbit antiPP1-87B (1:1500), sheep anti-Bif (1:1000), mAb 24B10 (1:3), rabbit anti-hGal (1:3000 from Cappel), mAb anti-Dac (1:500) or rabbit anti-Repo (1:500). Either fluorescent-conjugated secondary antibodies (FITC 1:100 dilution or cy3 1:750 dilution) or peroxidase-conjugated secondary antibodies (from Jackson laboratories used at 1:150 dilution) were used for signal detection. To enhance the HRP signal in the adult brains, the Vectashield ABC kit was used. Fluorescent-stained specimens were mounted in Vectashield and visualized using confocal microscopy. F-actin was visualized with tetramethyl rhodamine isothiocyanate – phalloidin (Sigma) or Alexa Fluor 546 Phalloidin (Molecular Probes). Fluorescent signals were detected using confocal microscopy. Transmission electron microscopy was used to visualize rhabdomeres of the adult eye as previously described (Helps et al., 2001).

374

K. Babu et al. / Developmental Biology 288 (2005) 372 – 386

Actin co-sedimentation assay The actin co-sedimentation assay was performed as previously described (Hohaus et al., 2002) with minor changes. Full-length and different deletions of bif were cloned into the pGADT7 vector that was used to in vitro translate Bif products. For making the constructs containing the N-terminal 500 amino acids of Bif, a 500 bp EcoRI/ClaI fragment was isolated from bif cDNA and inserted into the pGADT7 vector. To make other constructs, different regions of bif were amplified by PCR, and the PCR products were then digested with EcoRI/XhoI (these sites were engineered in the PCR products) and cloned into the pGADT7 vector. The actin co-sedimentation assay was performed using G-actin from Sigma. The G-actin was allowed to polymerize using salt and ATP, and the in vitro translated and radiolabeled polypeptides (Bif or luciferase control) were then added to the polymerized F-actin or the control solution (buffer without F-actin) and incubated for 60 min at room temperature. The mixtures were then spun at 100 K in an airfuge for 40 min, and both supernatants and pellets were then boiled in 2 SDS buffer and run on an SDS-PAGE gel. In each experiment, the samples were run onto two gels, one for autoradiography to detect the labeled protein and another for Coomassie staining to detect the actin band.

Results The role of Bif in axon guidance and rhabdomere shape can be uncoupled The involvement of Bif in axon guidance of R cells and its expression and localization in the rhabdomeres and axonal growth cones of these cells have been previously described (Bahri et al., 1997; Ruan et al., 2002). The results of these studies suggested a role for Bif in rhabdomere morphogenesis and a cell autonomous requirement in the outer R cells for their axons to be targeted to the lamina. In bif R47 (an antigen-minus allele) animals, most R2 – 5 axons bypass the lamina and terminate in the medulla, clumping of axons as well as breaks in the lamina surface are visible in these mutants (Figs. 1A–D; Ruan et al., 2002), but no gross abnormalities in the optic lobe can be detected (Figs. 3E – J). However, it is not clear how the axonal defects might relate to the morphogenesis defects in bif mutants. To uncouple the function of Bif in axon guidance from its function in rhabdomere morphogenesis, we made use of the two protein isoforms of Bif that have been previously described (Bahri et al., 1997; Helps et al., 2001). These Bif isoforms are generated by alternative splicing events, with the larger (1196 amino acids) bif + isoform produced from five exons and the smaller (1063 amino acids) bif 10Da isoform produced from six exons. We have shown previously that the bif + isoform driven by GMR-GAL4 rescues the rhabdomere as well as the actin localization defects associated with the bif mutant discs (Helps et al., 2001). A similar rescue using the

bif 10Da isoform of bif was also obtained (Figs. 2A, B, I, J and data not shown), suggesting that either isoform is sufficient to produce normal rhabdomere morphology. To ascertain whether these bif isoforms rescue the axon guidance defects, GMR-GAL4, which is expressed in all post-mitotic photoreceptor cells of the Drosophila eye (Hay et al., 1997), was used to drive the expression of UASbif + or UAS-bif 10Da transgenes specifically in the eye disc but not the optic lobe in bif mutant backgrounds. In these experiments, only bif + and not bif 10Da was able to rescue the axonal defects of bif eyes (Figs. 1E, F, I, J). These data revealed differential requirements for the two bif isoforms, with bif 10Da being largely required for normal R-cell rhabdomere morphology but not having a predominant role in normal axonal projections onto the optic lobe, while bif + is involved in both R-cell morphology and axon guidance. The uncoupling of Bif functions in axon guidance and rhabdomere morphology also suggested that the axon guidance phenotype of bif mutants is not merely a secondary effect of the R-cell morphological abnormalities. Overexpression of bif 10Da in a wild type background did not yield any obvious axonal phenotype as observed with mAb24B10 staining; furthermore, the photoreceptor rhabdomeres as observed under transmission electron microscopy were largely normal in shape and organization (Supplementary Fig. 1). It was shown by Ruan et al. (2002) that Bif + overexpression causes premature stop of the larval axons before they reach the target lamina. We find that the overexpression of the Bif10Da isoform does not cause such a premature stop phenotype, and therefore it is unlikely that the lack of rescue of the bif bypass phenotype is due to such an effect. Adult brains of bif mutants were also examined by 24B10 staining (Figs. 2C, D) and markers specific for either R1 – R6 projections (Rh1-slacZ; Figs. 2E, F) or R7 projections (Rh4slacZ; Figs. 2G, H). Surprisingly, the bif mutant adult brains showed largely normal arrays of axons, even though the rhabdomeres of the mutant adult retina have defective morphology (Figs. 2A, B; Bahri et al., 1997), indicating that the axonal phenotypes observed in the mutant larval optic lobe are largely corrected in the mutant adult brain and further demonstrating that the axon guidance phenotype and morphological defects of R cells can be uncoupled in bif mutants. The PP1-binding site in Bif is required for its function in photoreceptor axon guidance We have previously shown that Bif can associate with PP1-87B in vivo and can directly bind and inhibit PP1-87B

Fig. 1. The R-cell axonal phenotype of bifocal mutant larvae and rescue experiments. Panels A (control bif R47 /+ or wt) and B (bif R47 /bif R47 ) show optic lobes stained with anti-chaoptin antibody 24B10; clumping of axons at the lamina and laminal breaks are visible in the mutant (arrowhead in B) as compared to the control optic lobe (A; inset in A is another control bif LH114 which does not delete into the bif coding sequence). This phenotype was observed in 100% (n = 50) of the mutant optic lobes. Panels C (control Ro tau -lacZ) and D (bif R47 /bif R47 ; Ro tau -lacZ) show targeting of the axons of photoreceptors R2 – R5 as visualized with antihgal staining; R2 – R5 axons terminate at the lamina in the control optic lobe (C; arrowhead; arrow in panel C shows the Bolwigs nerve which enters the medulla), but most of the R2 – R5 axons bypass the lamina and terminate at the medulla in the mutant (D; arrowhead). The R2 – R5 axons bypass phenotype was observed in 100% of the mutant optic lobes (n = 17) with 65 – 75% of axons bypassing the lamina. Panels E and G are representative images of the majority of preparations (78%, n = 36) showing rescue of the bif mutant axon clumping phenotype using UAS-bif + driven by GMR-GAL4 and visualized with 24B10 staining. Panels F and H are representative images of the majority of preparations (89%, n = 44) showing that neither UAS-bif 10Da nor UAS-bif F995A expression under GMR-GAL4 rescues the axon clumping phenotype associated with the bif mutations (arrowheads in panels F and H) as visualized with 24B10 staining. The R2 – R5 axons bypass phenotype of bif mutants is rescued by UAS-bif + (I) but not UAS-bif 10Da (J; arrowhead) or UAS-bif F995A (K; arrowhead) as visualized with anti-hgal staining, which detects the LacZ expression derived from the Ro tau -lacZ marker. Scale bar is 20 Am.

K. Babu et al. / Developmental Biology 288 (2005) 372 – 386

activity in vitro. Bif contains a consensus four amino acid PP1-binding motif, RVQF, towards the C-terminus. This RVQF motif is present in both Bif + and Bif 10Da isoforms

375

and spans amino acids 992 –995. This motif is required for Bif to bind PP1-87B in vitro and for bif function in vivo, and a single amino acid mutation in this motif of Bif +,

376

K. Babu et al. / Developmental Biology 288 (2005) 372 – 386

Fig. 2. bif mutants show normal axon targeting in adult optic lobes. Panels A (wt) and B (bif R47 /bif R47 ) are EM micrographs showing defective rhabdomeres shape of bif mutants (B) as compared to wt ommatidia (A). Panels C – H show the optic lobes of WT (C, E and G) and bif mutant (D, F and H) adults stained with markers for all or specific subsets of photoreceptor axons; no obvious defects are observed with 24B10 staining (D; bif mutants as compared to wt in C), with the Rh1-lacZ R1 – R6 axons marker (F; bif mutants as compared to wt in E) or with the Rh4-lacZ R7 axons marker (H; bif mutants as compared to wt in G). Panels I and J are EM micrographs showing rescue of the adult rhabdomere phenotypes of bif mutants using either UAS-bif + (I) or UAS-bif 10Da (J); rescue of abnormal pattern of rhabdomere arrangement was observed in 90% of ommatidia preparations (n = 30), and rescue of rhabdomere shape was observed in 81% of rhabdomeres (n = 210). Compare the rescue images in panels I and J to the bif mutants in panel B and wt in panel A. Scale bar is 500 nm.

Bif F995A, can abolish this interaction (Helps et al., 2001). In addition, the mutant Bif F995A form cannot rescue the bif Rcell morphological defects in vivo (Helps et al., 2001). To test whether an intact PP1-binding site in Bif is also required for axon guidance in the larval optic lobe, the mutant UASbif F995A was driven by GMR-GAL4 in bif mutant backgrounds (Figs. 1H, K). When compared with UAS-bif + , very little rescue of the axonal defects was observed with this

mutant UAS-bif F995A construct as visualized with 24B10 (Figs. 1G – H) and the Ro tau -lacZ marker (Figs. 1I, K). This indicated that the PP1-binding site of Bif is essential for its function in axon guidance, and it raised the possibility that PP1 may also play a role in R-cell axon guidance (see below). Consistent with this view, double heterozygous larvae (bif/+; pp1-87B/+) showed clumping of axons at the lamina as well as defects in axon targeting, with R2 – 5

K. Babu et al. / Developmental Biology 288 (2005) 372 – 386

377

Fig. 3. Genetic interaction between bif and pp1-87B and Dac and Repo staining in larval brains. Panels A – D show larval optic lobes stained with 24B10 (A, B) or anti-hgal (C, D). Panels A and C are representative images showing the normal 24B10 staining pattern (A) and the normal R2 – R5 axonal pattern (C; as visualized with the Ro tau -lacZ marker) observed in either bif R47 /+ or pp1-87B/+ heterozygotes. Transheterozygous bif/+; pp1-87B/+ animals show axon clumping in the lamina as visualized with 24B10 (B, arrow; phenotype observed in 91.67% of brains, n = 36) and mistargeting of axons to the medulla as visualized with the Ro tau lacZ marker (D, arrow; phenotype observed in 27 – 30% of axons, n = 5) is mistargeted into the medulla. Panels E – J are confocal images of anti-Dac (red) and antiRepo (green) stainings in wt (E – G) and bif R47 mutant (H – J) optic lobes; no obvious defects are observed with either Dac or Repo staining in bif mutants as compared to wt. Panels G and J are merged images of panels E and F and H and I respectively. Scale bar is 20 Am.

378

K. Babu et al. / Developmental Biology 288 (2005) 372 – 386

Fig. 4. Inhibition of PP1 activity causes axonal targeting defects in the larval eye disc. Overexpression of the NIPP1 (A) or the I-2PP1 (B) inhibitors by GMR-GAL4 causes no obvious abnormality in photoreceptor formation as can be visualized with mAb 22C10, but staining with mAb 24B10 reveals laminal breaks and clumping of axons (C, NIPP1; D, I-2PP1); the axon phenotype caused by NIPP1 overexpression was observed in 100% (n = 20) of preparations, while, for the weaker I-2PP1 inhibitor, ¨67% (n = 15) showed the phenotype. Using the Ro tau -lacZ axonal marker, a severe defect in R2 – R5 targeting is observed with NIPP1 overexpression (E; ¨55% of the axons are mistargeted to the medulla, and this phenotype is observed in 100% of optic lobes, n = 16), and a relatively milder phenotype is observed with I-2PP1 overexpression (F; ¨5% of the axons are mistargeted to the medulla, and this phenotype is observed in 61% of optic lobes, n = 18). Scale bar is 20 Am.

axons bypassing the lamina and entering the medulla in these optic lobes (Figs. 3C, D), whereas neither bif nor pp187B single heterozygotes show a phenotype with 24B10 or the Ro tau -lacZ marker (Figs. 3A, B; data not shown). These data further support the view that Bif and PP1 act together for normal photoreceptor axon targeting in the larval optic lobe.

Overexpression of PP1-87B in the eye causes axon guidance defects Since Bif can inhibit the activity of PP1 (Helps et al., 2001), we wondered whether PP1-87B overexpression in wild type backgrounds might mimic the effects of bif loss-of-function on axon guidance. To test this, a UAS-PP1-87B construct was

Fig. 5. Axonal defects in pp1 mutants. Two alleles of pp1 (pp1 e211 and pp1 e078 ) were studied in trans with the strong allele of pp1, pp1 hs46 . The larval eye discs of pp1 e078 /+ (A; control) and the transheterozygotes pp1 e078 /pp1 hs46 or pp1 e211 /pp1 hs46 (B) do not show any obvious defects with 22C10 staining. Both transheterozygous combinations show an uneven lamina as visualized with mAb 24B10 staining with the pp1 e078 /pp1 hs46 (D; arrowhead) showing a more severe phenotype (97.2%; n = 36) than pp1 e211 /pp1 hs46 (C). Using the Ro tau -lacZ marker, both pp1 transheterozygous combinations show ¨50% of R2 – R5 axons mistargeted into the medulla (panel E is pp1 e211 /pp1 hs46 with 91.6% of preps, n = 24, showing the phenotype; F is pp1 e078 /pp1 hs46 with 100% of preps, n = 17, showing the phenotype). Panel G is a larval brain overexpressing PP1-87B and showing an uneven lamina (arrowhead) as visualized with 24B10 (this phenotype is observed in 100% of optic lobes, n = 28); overexpression of PP1 also gives rise to mistargeted R2 – R5 axons with the majority of preps showing mild bypass phenotype (panel I is a representative image, with 30 – 35% of the axons bypass the lamina into the medulla) and the remainder of the preps showing severe targeting defects (panel J is a representative image, with more than 70% of the axons entering the medulla). Panel H is a control pp1 e078 /+ optic lobe showing normal targeting of R2 – R5 axons to the lamina as visualized with the Ro tau -lacZ marker. Scale bar is 20 Am.

K. Babu et al. / Developmental Biology 288 (2005) 372 – 386

driven with GMR-GAL4 in the eye disc, and the optic lobes from these animals were stained with 24B10. Axonal breaks in the lamina similar to those observed in bif mutants were observed (Fig. 5G). The R2 – R5 axon-specific marker also showed mistargeting of these axons into the medulla (Figs. 5I, J).

379

Inhibition of PP1 activity or mutations in PP1 cause R-cell axon targeting defects We wanted to determine whether PP1-87B mutants themselves exhibit any axonal phenotype in the eye, but it

380

K. Babu et al. / Developmental Biology 288 (2005) 372 – 386

was not possible to use homozygotes of the available strong PP1-87B mutant alleles as this gene is required for normal cell cycle progression in larval brain neuroblasts and in the eye imaginal disc; hence, most PP1-87B mutants die at larval or early pupal stages (Axton et al., 1990). More importantly, these mutants have severely deformed small eye discs due to cell division defects, making the defective axonal projections they exhibit difficult to interpret (data not shown). As a first step in determining PP1 function in axon guidance, specific PP1 inhibitors, NIPP1 (Bennett et al., 2003; Beullens et al., 1992; Parker et al., 2002; Van Eynde et al., 1995) and I-2PP1 (Bennett et al., 1999, 2003; Helps and Cohen, 1999; Huang and Glinsmann, 1976; reviewed in Cohen, 2002), were driven post-mitotically in the eye disc using pGMR-GAL4. The rationale being that by inhibiting PP1 activity in post-mitotic cells the problems associated with cell cycle progression should be circumvented, making it possible to examine the effects of PP1 loss on axonal targeting. PP1 has been shown to associate with endogenous inhibitors, such as Inhibitor 2 (I-2) and Nuclear Inhibitor of PP1 (NIPP1), which may help to prevent inappropriate dephosphorylation of nonphysiological targets. It has also been shown that ectopic expression of the inhibitors specifically reduces PP1 activity in vivo and results in phenotypes resembling those of PP1c mutants (Bennett et al., 2003). Upon overexpression of the inhibitors under GMR-GAL4, the eye disc morphology appeared essentially normal, but the photoreceptor axons showed defects (Figs. 4A – D). The lamina formed by the axons expressing either NIPP1 (Fig. 4C) or I-2PP1 (Fig. 4D) was discontinuous with breaks, and the R2 – R5-specific Ro tau -lacZ marker showed axon mistargeting to the medulla mainly with the stronger NIPP1 inhibitor (Fig. 4E); the weaker I-2PP1 inhibitor showed mild or no mistargeting defect in the optic lobes as visualized with the R2 – R5-specific marker (Fig. 4F). These results suggest that PP1 activity is required for photoreceptor axon guidance in the fly visual system. Since NIPP1 and I-2PP1 are general PP1 inhibitors and not specific for PP1-87B, we next used different allelic combinations to bypass the early cell cycle defects associated with the pp1-87B mutations. Three alleles of pp1-87B, pp1 e211 , pp1 e078 and pp1 hs46 (Axton et al., 1990) were used to examine the pp1-87B phenotype in the eye disc. Using the strong pp1 hs46 allele in trans with the other 2 weaker alleles allowed larval development to proceed; in these particular transheterozygous combinations, the larval eye discs formed fairly normally as can be visualized with 22C10 (Figs. 5A, B), presumably because there is sufficient residual PP1 activity to enable cell divisions to occur at a near normal level. Staining of the pp1 transheterozygotes with 24B10 showed axons clumping at the lamina and laminal breaks (Figs. 5C, D), and some of the R2 – R5 axons in these animals were also mistargeted to the medulla (Figs. 5E, F). Taken together, the axonal defects in the PP1-87B mutant combinations in which cell cycle defects appear to be minimal further strengthen the view that PP1-87B is required for normal photoreceptor axon guidance.

Like Bif, PP1-87B is expressed in R-cell axons and in the optic lobe Bif expression and localization in the rhabdomeres and axonal growth cones have been previously described (Bahri et al., 1997; Ruan et al., 2002). However, a close examination of Bif staining revealed widespread expression in the optic lobe (Figs. 6, 7) which was not previously reported. Double labeling with anti-Bif and anti-Dachshund (Garrity et al., 1999; Mardon et al., 1994) showed that Bif localizes in the outer region of the lamina (Figs. 6A – C), where the photoreceptor axons initially send their projections into the optic lobe (Kunes et al., 1993). Bif also showed co-localization with F-actin in the optic lobe (Figs. 6J –L), similar to what has been reported in the eye disc (Bahri et al., 1997). In these experiments, Bif staining was stronger in the optic lobe cortex and weaker in the cortical axons (Figs. 6M– O). Bifocal staining was weak/absent in the eye disc and optic lobe of bif R47 mutants (Figs. 6P – R), indicating that anti-Bif staining in these tissues is specific. On the other hand, F-actin staining was still visible in these mutants. Although Bif is expressed in the optic lobe, no obvious defects were detected in bif mutants with either Dachshund (Mardon et al., 1994), which marks the proliferating cells of the lamina, or with Reversed polarity (Halter et al., 1995), which marks the glia (Figs. 3E –J). This indicated that the formation of the lamina and glia appears normal in bif mutants. Double labeling with anti-Bif and anti PP1-87B antibodies of wild type preparations indicated that both proteins have overlapping expression patterns in most regions of the optic lobe (Figs. 7A – C). Similar to the staining pattern of Bif, antiPP1-87B staining in the brain also showed enhanced signals in the brain cortex and the outer layer of the lamina, while it was considerably weaker in the axons. In addition, double staining with Bif and PP1-87B or Bif and phalloidin showed that both Bif and PP1-87B are expressed in the axons of the optic lobe (Figs. 7D – I). These experiments indicated that, like Bif, PP187B is expressed in R-cell axons consistent with a role in targeting of these axons. It should be noted here that the possibility of the PP1-87B antibody cross-reacting with any of the other 3 isoforms of PP1 present in Drosophila cannot be excluded. We carried out staining on pp1 mutants and found that the mutants still show positive staining; this is most probably due to the fact that the tissues retain residual PP1-87B activity because the transheterozygous combination we used to circumvent cell division defects associated with loss of PP187B is not null for PP1-87B. We therefore cannot rule out the possibility of cross-reaction of the PP1-87B antibody we used with different PP1 isoforms. Bif interacts genetically with receptor tyrosine phosphatases and can directly bind F-actin The interaction of Bif with intracellular signaling molecules like misshapen (Ruan et al., 2002) and PP1 (this work) seems to be important for normal R-cells axon guidance. In order to identify additional, more upstream, components, which interact

K. Babu et al. / Developmental Biology 288 (2005) 372 – 386

381

Fig. 6. Bif expression pattern in the optic lobe. Panels A – O are confocal images of wt third-instar larval optic lobes stained with anti-Bif (green) and various other markers. Double labeling with anti-Dac (A – C; red) shows Bif expression in the cortex of the optic lobe, and a high level of expression is seen on the outermost region of the lamina (arrowhead in panel A, merged panel C) where Dac is expressed (B). Double labeling with 24B10 (D – I) shows a high level of Bif expression in the axons that pass through the optic stalk (arrowhead in D, merged panel F); Bif expression in the axons when they enter the optic lobe becomes reduced when compared to its expression in the optic stalk (D and G), whereas 24B10 staining is still seen at high levels in the optic lobe axons (H and merged panel I). Double labeling with phalloidin (J – O) shows high expression for Bif (J) and co-localization with F-actin (K, merged panel L) in the cortex. Expression of Bif in a deeper plane of focus (M and merged image O) in the optic lobe is less than in the cortex (compare M to J). No expression of Bif is seen in bif R47 mutant optic lobes (P), while phalloidin staining is still seen in these mutants (Q). Panel R is a merged image of panels P and Q. Scale bar is 20 Am.

with bif, genetic interaction experiments between bif and other molecules involved in photoreceptor axon guidance were performed. bif R47 /+ individuals are wild type (Fig. 3A) as are dptp69d/+ or dptp10d/+ (Fig. 8A; data not shown), however, axonal defects were observed in double transheterozygous optic lobes of bif/dptp10d and bif/+; dptp69d/+ combinations (Figs. 8D, E). This phenotype was also seen in optic lobes from females heterozygous for a deletion removing bif and dptp10d (Fig. 8F). There was no phenotype in double heterozygotes of bif R47 with dock, pak, dlar, ena or abl (Fig. 8B; data not shown). These genetic interaction experiments

suggested that bif preferentially interacts with the upstream dptp10d and dtptp69d and that bif might be a component of axon guidance pathways involving these signaling receptors. We have used the Ro tau -lacZ axonal marker on the bif/+, dptp69d/+ and bif/dptp10d/+ double heterozygotes, but we did not detect any mistargeting phenotype. This may indicate that Bif does not cooperate with the receptors to influence the ‘‘stop decision’’ of growth cones at the lamina. However, our results with 24B10 staining showed uneven lamina only with the bif mutant in combination with the 10D and 69D mutants, but we did not see a similar effect on the lamina when we used genetic

382

K. Babu et al. / Developmental Biology 288 (2005) 372 – 386

Fig. 7. Bifocal and PP1 are expressed in axons. Panels A – C show co-expression of Bif (A, green) and PP1 (B, red) in the wt optic lobe; panel C is a merged image of panels A and B. Panels D – F show double labeling of Anti-Bif (D, green) and phalloidin (E, red) in wt optic lobes; note that Bif is expressed in the axons in a manner similar to F-actin; the arrow in panel E indicates axons stained with phalloidin, and panel F is a merged image of panels D and E. Panels G – I show double staining of wt optic lobes with phalloidin (G, red) and anti-PP1 (H, green); note PP1 staining in axons (H, arrows). Scale bar is 20 Am.

combinations between bif and several other mutants including dpak, dock, dlar, ena and abl. Hence, we have concluded that bif genetically interacts with the receptor phosphatases. It has previously been shown that Bif from embryonic fly extracts can be pulled down using F-actin columns (Sisson et al., 2000), indicating that Bif may be a component of an actinassociated complex of proteins in vivo. However, the mode of interaction between Bif and F-actin is not clear since Bif does not have any obvious actin-binding domains (Bahri et al., 1997). In order to determine whether the interaction of Bif with F-actin is direct, a co-sedimentation assay was performed using F-actin and in vitro translated Bif. In this experiment, Bif was pulled down by F-actin, indicating that Bif and F-actin can bind directly to each other (Fig. 9A). Further dissection of the Bif protein mapped the F-actin-binding domain to 2 novel regions in the last 400 amino acids of the Bif + protein (schematic in Fig. 9B). Only one of these domains is shared by both Bif + and Bif10Da isoforms, while the other actin-binding domain is specific to Bif + (Fig. 9B). These experiments indicated that Bif/F-actin complexes may form through direct binding and that Bif may link various axon guidance signaling pathways to the actin cytoskeleton.

Discussion Bifocal is a protein required for normal photoreceptor rhabdomere morphology in Drosophila (Bahri et al., 1997), and, in this process, it functions together with PP1 (Helps et al., 2001). Bif is also a target of the Ser/Thr kinase Misshapen, and, together, these molecules regulate photoreceptor axon guidance and targeting during the larval stages (Ruan et al., 2002). In this report, we have shown that Bif function in R-cell axons can be uncoupled from its function in maintaining R-cell shape and morphology and that the PP1-binding site in Bif is required for its function in axon guidance. In addition, we have shown that the Bif-interacting partner, PP1-87B, is required for axon guidance in the visual system and that Bif is likely to function as a downstream component of signaling pathway(s) that involves the receptor tyrosine phosphatases, Dptp10D and Dptp69D. Finally, we have shown that Bif can directly bind Factin and thus might provide a link between axon guidance pathways and the underlying actin cytoskeleton. The rescue experiments utilizing the two different isoforms of bif demonstrate that the function of Bif in photoreceptor axon guidance can be uncoupled from its function in rhabdomere

K. Babu et al. / Developmental Biology 288 (2005) 372 – 386

383

Fig. 8. Genetic interaction between bif and receptor tyrosine phosphatases. The receptor tyrosine phosphatases dptp69d and dptp10d genetically interact with bif. Control optic lobes from ptp10d/+ (A) and ptp69d/+ (C) do not show any defects when stained with mAb 24B10. Panel B is a bif/+; pak/+ optic lobe showing normal axonal pattern as visualized with 24B10 staining. Staining of bif/+; ptp69d/+ optic lobes with mAb 24B10 shows clumps and breaks in the lamina (D, arrowhead; 84% or preps show a phenotype, n = 37). A similar phenotype is seen when one copy of bif and one copy of ptp10d are simultaneously removed (E, arrowhead; 85% of preps show a phenotype, n = 34) or in animals heterozygous for a small deletion that removes both bif and ptp10d (F, arrowhead; 89% of preps show a phenotype, n = 38). Scale bar is 20 Am.

morphology in photoreceptor cells, and this indicates that Bif has a dual role in the fly visual system, normal axon connectivity in the larval stages and the formation of normal rhabdomeres in the adult eye. It should however be noted that the rhabdomere rescue obtained using both the bif + and the bif10Da isoforms of bif was not complete, possibly because the expression pattern and/or level of Bif expression in the rescue experiments are somehow different from the wild type expression of Bif. The larger Bif + isoform is 1196 amino acids long, whereas the shorter Bif 10DA isoform is 1063 and is formed by an alternative splice site between exon 4 and 5. It is somewhat surprising to find that the two isoforms of Bif behave differently with respect to axon guidance since both isoforms contain the PP1 and actinbinding sites (see Fig. 9). One possible explanation is that the additional C-terminal sequences present in Bif + are required for

its function in axon guidance. Interestingly, this region comprises one of the actin-binding domains. The precise role of these sequences in vivo is not clear at present, but it is possible to envisage roles for them in protein folding or mediating binding to other partners. In this regard, one plausible scenario is that the two Bif isoforms may assume distinct patterns of subcellular localization due to their different C-termini. It is possible for future approaches to address this question by generating specific antibodies that recognize the different isoforms of Bif or expressing tagged versions of the two Bif products. Mutations in the conserved phosphatase-binding motif of Bif, which abolish binding of Bif to PP1-87B, render the protein less effective in axon guidance. We have previously reported similar effects on Bif activity for these mutations in

384

K. Babu et al. / Developmental Biology 288 (2005) 372 – 386

Fig. 9. Bif binding to F-actin in vitro. Panel A shows a series of gels from cosedimentation assays of F-actin with the various radiolabeled Bif + products; the supernatant fractions in the presence (+As) or absence ( As) of F-actin and the pellet fractions in the presence (+Ap) or absence ( Ap) of F-actin are indicated on top. The first row is co-sedimentation of an in vitro translated luciferase control showing the labeled luciferase remains in the supernatant (+As), and it does not pellet by F-actin (+Ap). The second row is co-sedimentation of an in vitro translated full-length Bif + (FL Bif) showing the labeled Bif + protein precipitated with the F-actin pellet (+Ap); the Bif + band is absent from the pellet fraction in the absence of F-actin ( Ap). The third row is co-sedimentation of an in vitro translated N-terminal aa500 of Bif + (N-ter Bif); this N-terminal product is not pelleted by F-actin (+Ap). The fourth row is co-sedimentation of an in vitro translated C-terminal aa396 of Bif + (C-ter Bif); the C-terminal product is pelleted by F-actin (+Ap); The last row is a control gel stained with Commassie blue to show the F-actin band. Panel B shows schematic diagrams of the two Bif isoforms, Bif + and Bif 10Da, and the various protein regions used in the F-actin co-sedimentation assays; (+) indicates binding to F-actin and ( ) indicates lack of binding to F-actin. The first aa1000 are common between Bif + and Bif 10Da, but the two protein isoforms differ in their C-termini; two broad domains of Bif + spanning aa800 – 950 and aa1000 – 1196 (in red) bind actin, while Bif 10Da contains only the aa800 – 950 actin-binding domain; the last aa63 of Bif 10Da (which are not found in Bif +) do not bind F-actin. The RVQF PP1-binding motif in Bif + and Bif 10Da is indicated by an arrow.

the morphogenesis of rhabdomeres (Helps et al., 2001). Another interesting aspect is that Bif can form a complex with the cytoskeletal F-actin both in vivo and in vitro (Sisson et al., 2000; this work). Changes in the actin cytoskeleton are known to be essential for remodeling of the axon growth cone. These data suggest that, in addition to its inhibitory regulatory effect on the phosphatase activity, Bif might serve as a direct link between PP1-87B and the actin cytoskeleton. Similar observations were reported for neurabins, which are the inhibitors of PP1 in mammals. It has been shown that neurabins bind actin and serve as linkers between the actin cytoskeleton and the plasma membrane at the cadherin-based cell –cell adhesion sites. Neurabin I is highly concentrated at the synapse of

developed neurons and in the lamellapodia of growth cones during the development of neurons, suggesting that it could be required for synapse formation or function (MacMillan et al., 1999; McAvoy et al., 1999; Oliver et al., 2002; Terry-Lorenzo et al., 2002). Although it does not share any sequence homology with mammalian neurabins, Bif could be a functional homologue of neurabins in the Drosophila larval visual system by associating with PP1-87B and the actin cytoskeleton. Such functions of Bif might consequently affect the phosphorylation status of various actin-binding proteins like myosin II (reviewed by Tan et al., 1992) and cofilin (reviewed by Lawler, 1999), which are known to undergo phosphorylation and dephosphorylation. A role for PP1 itself in axonal connectivity in the larval optic lobe is supported by two observations: firstly, expressing PP1 inhibitors in post-mitotic cells in the eye imaginal disc results in axon defects; secondly, transallelic combinations of pp1-87B mutants, which are able to bypass the earlier defects of cell cycle arrest that are usually seen in homozygotes of strong pp1-87B mutants, reveal axon targeting defects in the larval lamina. Although the PP1 inhibitors are not specific to PP1-87B, they nevertheless do support the notion that the phosphatase activity of PP1 is required for proper axon guidance. In the transallelic combinations of pp1 mutations, formation of the R cells of the larval eye disc is not adversely affected, suggesting that the axon guidance phenotype of pp1-87B mutants is not simply due to the lack of cells in the developing eye disc. Interestingly, both pp187B mutants as well as PP1 overexpression give rise to axons bypassing the lamina and entering the medulla of the optic lobe. This suggests that an optimal level of PP1 activity in the eye disc and optic lobe may be required, and either increase or decrease of PP1 levels could lead to changes in the phosphorylation status of target proteins and subsequent defects in the growth cone cytoskeleton and axon guidance. PP1 is one of the most abundant eukaryotic protein phosphatases that dephosphorylate serine and threonine residues of target proteins. The activities of various PP1 catalytic (PP1c) subunits are extensively regulated in various organisms and tissue types within a single organism. PP1c binds various regulatory molecules, which could bring the phosphatase to its site of action (reviewed by (Cohen, 2002) or act as adaptors for the phosphatase at its site of function. There are 4 related PP1c subunits in Drosophila (Dombradi et al., 1990b, 1993). Drosophila PP1s are very similar proteins encoded by different loci in the fly genome, and they are variably regulated at different points in development (Alphey et al., 1997; Asztalos et al., 1993; Axton et al., 1990; Baksa et al., 1993; Bennett et al., 1999; Carvalho et al., 2001; Dombradi and Cohen, 1992; Dombradi et al., 1990a,b, 1993; Helps et al., 1998, 2001; Raghavan et al., 2000; reviewed by Cohen, 2002). It is conceivable that Bif acts as one of the regulatory subunits of PP1 in the fly eye, and it is required either as a complex with PP1 and the actin cytoskeleton or for recruiting PP1 to subcellular sites where the PP1 activity is required to modulate the actin cytoskeleton. The former hypothesis seems more plausible as bif mutant larval and pupal eye discs do not show any obvious changes in PP1 staining (our unpublished data).

K. Babu et al. / Developmental Biology 288 (2005) 372 – 386

Based on the genetic interactions of bif and the receptor tyrosine phosphatase genes, Dptp10D and Dptp69D, and on the data of Bif functional and biochemical interactions with PP1-87B and F-actin, a possible model for Bif function in axon guidance could be put forward: activation of receptor molecules leads to signaling events (perhaps also involving misshapen) that activate Bif in the growth cone which in turn leads to inhibition of PP1 activity and changes in the actin cytoskeleton to support axon outgrowth and guidance. In this scenario, PP1 must have a threshold level of activity as decreased levels of PP1-87B and general inhibition of PP1 activity also cause axon guidance defects. Bif might participate in lowering PP1 activity to the threshold level but must not abolish all PP1 activity in order to allow normal photoreceptor axon guidance. Interestingly, overexpression of Bif as four copies has been shown to induce axonal targeting defects (Ruan et al., 2002). Finally, bif mutants affect axon targeting in the larval optic lobe, but these defects seem to be corrected in the adult. How this axon defect is rectified in the adult stages remains unclear, and it probably occurs during the extensive remodeling that takes place during the pupal stages of fly development. In future studies, it would be interesting to determine if any of the other molecules involved in Drosophila photoreceptor axon guidance behave in a similar manner, and it would perhaps also be informative to classify phenotypes into those that can be repaired and those which cannot. Acknowledgments We thank the following people for providing flystocks, antibodies and other reagents: Barry Dickson, David Van Vactor, Graeme Marden, Larry Zipursky, Kai Zinn, Yang Rao, Nick Harden, Nick Helps, Tricia Cohen and Ulrike Heberlein. We also thank the Bloomington stock center for providing fly stocks. Our special thanks to Richard Tuxworth for help with the image collection, Rachna Kaushik for critical comments on the manuscript and the other laboratory members for discussion and help. The Wellcome Trust UK, Temasek Life Science Laboratory (W. Chia and K. Babu) and Institute of Molecular and Cell Biology (S. Bahri) for financial support. This work was supported in part by a Wellcome Trust Principal Research Fellowship and Programme Grant to WC. Appendix A. Supplementary data Supplementary data associated with this article can be found in the online version at doi:10.1016/j.ydbio.2005.09.017. References Alphey, L., Parker, L., Hawcroft, G., Guo, Y., Kaiser, K., Morgan, G., 1997. KLP38B: a mitotic kinesin-related protein that binds PP1. J. Cell Biol. 138, 395 – 409. Asztalos, Z., von Wegerer, J., Wustmann, G., Dombradi, V., Gausz, J., Spatz, H.C., Friedrich, P., 1993. Protein phosphatase 1-deficient mutant Drosophila is affected in habituation and associative learning. J. Neurosci. 13, 924 – 930. Awasaki, T., Saito, M., Sone, M., Suzuki, E., Sakai, R., Ito, K., Hama, C., 2000. The Drosophila trio plays an essential role in patterning of axons by regulating their directional extension. Neuron 26, 119 – 131.

385

Axton, J.M., Dombradi, V., Cohen, P.T., Glover, D.M., 1990. One of the protein phosphatase 1 isoenzymes in Drosophila is essential for mitosis. Cell 63, 33 – 46. Babu, K., Cai, Y., Bahri, S., Yang, X., Chia, W., 2004. Roles of Bifocal, Homer, and F-actin in anchoring Oskar to the posterior cortex of Drosophila oocytes. Genes Dev. 18, 138 – 143. Bahri, S.M., Yang, X., Chia, W., 1997. The Drosophila bifocal gene encodes a novel protein which colocalises with actin and is necessary for photoreceptor morphogenesis. Mol. Cell. Biol. 17, 5521 – 5529. Baksa, K., Morawietz, H., Dombradi, V., Axton, M., Taubert, H., Szabo, G., Torok, I., Udvardy, A., Gyurkovics, H., Szoor, B., et al., 1993. Mutations in the protein phosphatase 1 gene at 87B can differentially affect suppression of position-effect variegation and mitosis in Drosophila melanogaster. Genetics 135, 117 – 125. Bateman, J., Shu, H., Van Vactor, D., 2000. The guanine nucleotide exchange factor trio mediates axonal development in the Drosophila embryo. Neuron 26, 93 – 106. Bennett, D., Szoor, B., Alphey, L., 1999. The chaperone-like properties of mammalian inhibitor-2 are conserved in a Drosophila homologue. Biochemistry 38, 16276 – 16282. Bennett, D., Szoor, B., Gross, S., Vereshchagina, N., Alphey, L., 2003. Ectopic expression of inhibitors of protein phosphatase type 1 (PP1) can be used to analyze roles of PP1 in Drosophila development. Genetics 164, 235 – 245. Beullens, M., Van Eynde, A., Stalmans, W., Bollen, M., 1992. The isolation of novel inhibitory polypeptides of protein phosphatase 1 from bovine thymus nuclei. J. Biol. Chem. 267, 16538 – 16544. Bossing, T., Brand, A.H., 2002. Dephrin, a transmembrane ephrin with a unique structure, prevents interneuronal axons from exiting the Drosophila embryonic CNS. Development 129, 4205 – 4218. Cafferty, P., Yu, L., Rao, Y., 2004. The receptor tyrosine kinase Off-track is required for layer-specific neuronal connectivity in Drosophila. Development 131, 5287 – 5295. Carvalho, A.B., Dobo, B.A., Vibranovski, M.D., Clark, A.G., 2001. Identification of five new genes on the Y chromosome of Drosophila melanogaster. Proc. Natl. Acad. Sci. U. S. A. 98, 13225 – 13230. Clandinin, T.R., Lee, C.H., Herman, T., Lee, R.C., Yang, A.Y., Ovasapyan, S., Zipursky, S.L., 2001. Drosophila LAR regulates R1 – R6 and R7 target specificity in the visual system. Neuron 32, 237 – 248. Cohen, P.T., 2002. Protein phosphatase 1—Targeted in many directions. J. Cell Sci. 115, 241 – 256. Dearborn Jr., R., He, Q., Kunes, S., Dai, Y., 2002. Eph receptor tyrosine kinasemediated formation of a topographic map in the Drosophila visual system. J. Neurosci. 22, 1338 – 1349. Desai, C., Purdy, J., 2003. The neural receptor protein tyrosine phosphatase DPTP69D is required during periods of axon outgrowth in Drosophila. Genetics 164, 575 – 588. Desai, C.J., Gindhart Jr., J.G., Goldstein, L.S., Zinn, K., 1996. Receptor tyrosine phosphatases are required for motor axon guidance in the Drosophila embryo. Cell 84, 599 – 609. Desai, C.J., Garrity, P.A., Keshishian, H., Zipursky, S.L., Zinn, K., 1999. The Drosophila SH2 – SH3 adapter protein Dock is expressed in embryonic axons and facilitates synapse formation by the RP3 motoneuron. Development 126, 1527 – 1535. Dombradi, V., Cohen, P.T., 1992. Protein phosphorylation is involved in the regulation of chromatin condensation during interphase. FEBS Lett. 312, 21 – 26. Dombradi, V., Axton, J.M., Barker, H.M., Cohen, P.T., 1990. Protein phosphatase 1 activity in Drosophila mutants with abnormalities in mitosis and chromosome condensation. FEBS Lett. 275, 39 – 43. Dombradi, V., Axton, J.M., Brewis, N.D., da Cruz e Silva, E.F., Alphey, L., Cohen, P.T., 1990. Drosophila contains three genes that encode distinct isoforms of protein phosphatase 1. Eur. J. Biochem. 194, 739 – 745. Dombradi, V., Mann, D.J., Saunders, R.D., Cohen, P.T., 1993. Cloning of the fourth functional gene for protein phosphatase 1 in Drosophila melanogaster from its chromosomal location. Eur. J. Biochem. 212, 177 – 183. Garrity, P.A., Lee, C.H., Salecker, I., Robertson, H.C., Desai, C.J., Zinn, K., Zipursky, S.L., 1999. Retinal axon target selection in Drosophila is regulated by a receptor protein tyrosine phosphatase. Neuron 22, 707 – 717.

386

K. Babu et al. / Developmental Biology 288 (2005) 372 – 386

Garrity, P.A., Rao, Y., Salecker, I., McGlade, J., Pawson, T., Zipursky, S.L., 1996. Drosophila photoreceptor axon guidance and targeting requires the dreadlocks SH2/SH3 adapter protein. Cell 85, 639 – 650. Halter, D.A., Urban, J., Rickert, C., Ner, S.S., Ito, K., Travers, A.A., Technau, G.M., 1995. The homeobox gene repo is required for the differentiation and maintenance of glia function in the embryonic nervous system of Drosophila melanogaster. Development 121, 317 – 332. Hanson, M.a., 1993. The development of the optic lobe. In: Martinez-Arias, M.a. (Ed.), The Development of Drosophila melanogaster, vol. II. Cold Spring Harbour Laboratory Press. Hay, B.A., Maile, R., Rubin, G.M., 1997. P element insertion-dependent gene activation in the Drosophila eye. Proc. Natl. Acad. Sci. U. S. A. 94, 5195 – 5200. Helps, N.R., Cohen, P.T., 1999. Drosophila melanogaster protein phosphatase inhibitor-2: identification of a site important for PP1 inhibition. FEBS Lett. 463, 72 – 76. Helps, N.R., Vergidou, C., Gaskell, T., Cohen, P.T., 1998. Characterisation of a novel Drosophila melanogaster testis specific PP1 inhibitor related to mammalian inhibitor-2: identification of the site of interaction with PP1. FEBS Lett. 438, 131 – 136. Helps, N.R., Cohen, P.T., Bahri, S.M., Chia, W., Babu, K., 2001. Interaction with protein phosphatase 1. Is essential for bifocal function during the morphogenesis of the Drosophila compound eye. Mol. Cell. Biol. 21, 2154 – 2164. Hing, H., Xiao, J., Harden, N., Lim, L., Zipursky, S.L., 1999. Pak functions downstream of Dock to regulate photoreceptor axon guidance in Drosophila. Cell 97, 853 – 8563. Hohaus, A., Person, V., Behlke, J., Schaper, J., Morano, I., Haase, H., 2002. The carboxyl-terminal region of ahnak provides a link between cardiac L-type Ca2+ channels and the actin-based cytoskeleton. FASEB J. 16, 1205 – 1216. Huang, F.L., Glinsmann, W., 1976. A second heat-stable protein inhibitor of phosphorylase phosphatase from rabbit muscle. FEBS Lett. 62, 326 – 329. Huang, Z., Kunes, S., 1998. Signals transmitted along retinal axons in Drosophila: Hedgehog signal reception and the cell circuitry of lamina cartridge assembly. Development 125, 3753 – 3764. Iwai, Y., Usui, T., Hirano, S., Steward, R., Takeichi, M., Uemura, T., 1997. Axon patterning requires DN-cadherin, a novel neuronal adhesion receptor, in the Drosophila embryonic CNS. Neuron 19, 77 – 89. Iwai, Y., Hirota, Y., Ozaki, K., Okano, H., Takeichi, M., Uemura, T., 2002. DNcadherin is required for spatial arrangement of nerve terminals and ultrastructural organization of synapses. Mol. Cell. Neurosci. 19, 375 – 388. Krueger, N.X., Van Vactor, D., Wan, H.I., Gelbart, W.M., Goodman, C.S., Saito, H., 1996. The transmembrane tyrosine phosphatase DLAR controls motor axon guidance in Drosophila. Cell 84, 611 – 622. Kunes, S., 1999. Stop or go in the target zone. Neuron 22, 639 – 640. Kunes, S., Steller, H., 1993. Topography in the Drosophila visual system. Curr. Opin. Neurobiol. 3, 53 – 59. Kunes, S., Wilson, C., Steller, H., 1993. Independent guidance of retinal axons in the developing visual system of Drosophila. J. Neurosci. 13, 752 – 767. Lawler, S., 1999. Regulation of actin dynamics: the LIM kinase connection. Curr. Biol. 9, R800 – R822. Lee, C.H., Herman, T., Clandinin, T.R., Lee, R., Zipursky, S.L., 2001. Ncadherin regulates target specificity in the Drosophila visual system. Neuron 30, 437 – 450. Lee, R.C., Cladinin, C.R., Lee, C.H., Chen, P.L., Meinertzhagen, I.A., Zipursky, S.L., 2003. The protocadherin Flamingo is required for axon target selection in the Drosophila visual system. Nat. Neurosci. 6, 557 – 563. Luo, L., 2000. Trio quartet in D. melanogaster. Neuron 26, 1 – 2. MacMillan, L.B., Bass, M.A., Cheng, N., Howard, E.F., Tamura, M., Strack, S., Wadzinski, B.E., Colbran, R.J., 1999. Brain actin-associated protein phosphatase 1 holoenzymes containing spinophilin, neurabin, and selected catalytic subunit isoforms. J. Biol. Chem. 274, 35845 – 35854. Mardon, G., Solomon, N.M., Rubin, G.M., 1994. dachshund encodes a nuclear protein required for normal eye and leg development in Drosophila. Development 120, 3473 – 3486.

Maurel-Zaffran, C., Suzuki, T., Gahmon, G., Treisman, J.E., Dickson, B.J., 2001. Cell-autonomous and-nonautonomous functions of LAR in R7 photoreceptor axon targeting. Neuron 32, 225 – 235. McAvoy, T., Allen, P.B., Obaishi, H., Nakanishi, H., Takai, Y., Greengard, P., Nairn, A.C., Hemmings Jr., H.C., 1999. Regulation of neurabin I interaction with protein phosphatase 1 by phosphorylation. Biochemistry 38, 12943 – 12949. Newsome, T.P., Asling, B., Dickson, B.J., 2000a. Analysis of Drosophila photoreceptor axon guidance in eye-specific mosaics. Development 127, 851 – 860. Newsome, T.P., Schmidt, S., Dietzl, G., Keleman, K., Asling, B., Debant, A., Dickson, B.J., 2000b. Trio combines with dock to regulate Pak activity during photoreceptor axon pathfinding in Drosophila. Cell 101, 283 – 294. Oliver, C.J., Terry-Lorenzo, R.T., Elliott, E., Bloomer, W.A., Li, S., Brautigan, D.L., Colbran, R.J., Shenolikar, S., 2002. Targeting protein phosphatase 1 (PP1) to the actin cytoskeleton: the neurabin I/PP1 complex regulates cell morphology. Mol. Cell. Biol. 22, 4690 – 4701. Palmer, A., Klein, R., 2003. Multiple roles of ephrins in morphogenesis, neuronal networking, and brain function. Genes Dev. 17, 1429 – 1450. Parker, L., Gross, S., Beullens, M., Bollen, M., Bennett, D., Alphey, L., 2002. Functional interaction between NIPP1 and PP1 in Drosophila: consequences of over-expression of NIPP1 in flies and suppression by coexpression of PP1. Biochem. J. 368 (Pt. 3), 789 – 797. Raghavan, S., Williams, I., Aslam, H., Thomas, D., Szoor, B., Morgan, G., Gross, S., Turner, J., Fernandes, J., VijayRaghavan, K., Alphey, L., 2000. Protein phosphatase 1beta is required for the maintenance of muscle attachments. Curr. Biol. 10, 269 – 272. Rao, Y., Zipursky, S.L., 1998. Domain requirements for the Dock adapter protein in growth-cone signaling. Proc. Natl. Acad. Sci. U. S. A. 95, 2077 – 2082. Ruan, W., Long, H., Vuong, D.H., Rao, Y., 2002. Bifocal is a downstream target of the Ste20-like serine/threonine kinase misshapen in regulating photoreceptor growth cone targeting in Drosophila. Neuron 36, 831 – 842. Schmucker, D., Zipursky, S.L., 2001. Signaling downstream of Eph receptors and ephrin ligands. Cell 105, 701 – 704. Selleck, S.B., Steller, H., 1991. The influence of retinal innervation on neurogenesis in the first optic ganglion of Drosophila. Neuron 6, 83 – 99. Senti, K.A., Usui, T., Boucke, K., Greber, U., Uemura, T., Dickson, B.J., 2003. Flamingo regulates R8 axon – axon and axon – target interactions in the Drosophila visual system. Curr. Biol. 13, 828 – 832. Sisson, J.C., Field, C., Ventura, R., Royou, A., Sullivan, W., 2000. Lava lamp, a novel peripheral golgi protein, is required for Drosophila melanogaster cellularization. J. Cell Biol. 151, 905 – 918. Sun, Q., Bahri, S., Schmid, A., Chia, W., Zinn, K., 2000. Receptor tyrosine phosphatases regulate axon guidance across the midline of the Drosophila embryo. Development 127, 801 – 812. Tan, J.L., Ravid, S., Spudich, J.A., 1992. Control of nonmuscle myosins by phosphorylation. Annu. Rev. Biochem. 61, 721 – 759. Terry-Lorenzo, R.T., Carmody, L.C., Voltz, J.W., Connor, J.H., Li, S., Smith, F.D., Milgram, S.L., Colbran, R.J., Shenolikar, S., 2002. The neuronal actinbinding proteins, neurabin I and neurabin II, recruit specific isoforms of protein phosphatase-1 catalytic subunits. J. Biol. Chem. 277, 27716 – 27724. Tessier-Lavigne, M., Goodman, C.S., 1996. The molecular biology of axon guidance. Science 274, 1123 – 1133. Van Eynde, A., Wera, S., Beullens, M., Torrekens, S., Van Leuven, F., Stalmans, W., Bollen, M., 1995. Molecular cloning of NIPP-1, a nuclear inhibitor of protein phosphatase-1, reveals homology with polypeptides involved in RNA processing. J. Biol. Chem. 270, 28068 – 28074. Winberg, M.L., Perez, S.E., Steller, H., 1992. Generation and early differentiation of glial cells in the first optic ganglion of Drosophila melanogaster. Development 115, 903 – 911. Wolff, T., Ready, D.F., 1991. The beginning of pattern formation in the Drosophila compound eye: the morphogenetic furrow and the second mitotic wave. Development 113, 841 – 850. Yu, T.W., Bargmann, C.I., 2001. Dynamic regulation of axon guidance. Nat. Neurosci. 4, 1169 – 1176.