Membrane Targeting and Asymmetric Localization of Drosophila Partner of Inscuteable Are Discrete Steps Controlled by Distinct Regions of the Protein

Abstract

Asymmetric division of neural progenitors is a key mechanism by which neuronal diversity in the Drosophila central nervous system is generated. The distinct fates of the daughter cells derived from these divisions are achieved through preferential segregation of the cell fate determinants Prospero and Numb to one of the two daughters. This is achieved by coordinating apical and basal mitotic spindle orientation with the basal cortical localization of the cell fate determinants during mitosis. A complex of apically localized proteins, including Inscuteable (Insc), Partner of Inscuteable (Pins), Bazooka (Baz), DmPar-6, DaPKC, and Gαi, is required to mediate and coordinate basal protein localization with mitotic spindle orientation. Pins, a molecule which directly interacts with Insc, is a key component required for the integrity of this complex; in the absence of Pins, other components become mislocalized or destabilized, and basal protein localization and mitotic spindle orientation are defective. Here we define the functional domains of Pins. We show that the C-terminal region containing the Gαi binding GoLoco motifs is necessary and sufficient for targeting to the neuroblast cortex, which appears to be a prerequisite for apical localization of Pins. The N-terminal tetratricopeptide repeat-containing region of Pins is required for two processes; TPR repeats 1 to 3 plus the C-terminal region are required for apical localization but are insufficient to recruit Insc to the apical cortex, whereas TPR repeats 1 to 7 plus C-terminal Pins can perform both functions. Hence, the abilities of Pins to cortically localize, to apically localize, and to restore Insc apical localization are all separable, and all three capabilities are necessary to mediate asymmetric division. Moreover, the need for N-terminal Pins can be obviated by fusing a minimal Insc functional domain with the C-terminal region of Pins; this chimeric molecule is apically localized and can fulfill the functions of both Insc and Pins.

Most Drosophila central nervous system (CNS) neurons are derived from neural progenitors, neuroblasts (NBs) (1, 30). NBs undergo repeated asymmetric divisions, budding off a series of smaller ganglion mother cells (GMCs) from their dorsal and lateral sides. GMCs divide terminally to produce two progeny neurons or glia (9). Both NB and GMC divisions are asymmetric but appear to utilize two distinct intrinsic cell fate determinants. In mitotic NBs, two potential cell fate determinants, the homeodomain protein Prospero (Pros) (5, 21, 42) and Numb (41), as well as pros RNA, are localized as crescents to the basal cell cortex and segregate preferentially to the basal progeny (2, 10, 13, 18, 27, 35). Pros acts as a cell fate determinant for GMCs, promoting the expression of GMC-specific genes and repressing the expression of NB-specific genes (5, 42). However, for GMC cell divisions (and MP2, another CNS precursor undergoing terminal cell division) which generate daughter neurons with distinct identities, Numb is localized as a basal crescent and acts as a cell fate determinant (3, 36, 37), possibly by directly interacting with the intracellular region of Notch (N) (7) and thereby inhibiting N signaling in the progeny neuron which preferentially inherits the asymmetrically localized Numb.

The localization of the basal components is facilitated by a set of molecules which have been referred to as adaptors (33). The localization of pros RNA requires Staufen (2, 18, 32), a molecule capable of binding double-stranded RNA in vitro (39), which is implicated in the transport and anchoring of oskar and bicoid RNAs in the oocyte and early egg (38) and which also interacts with Insc. Localization of Pros and Staufen requires Miranda (Mira) (11, 22, 32, 33). Normal localization of Numb in NBs requires partner of numb (pon) (20). The adaptors themselves are localized in the same way as the molecules they help to localize.

A common theme for both the NB and GMC asymmetric division is the preferential segregation of localized molecular components to just one progeny cell. In order for this to occur, the orientation of the mitotic spindle has to be coordinated with the site of localization of the basal components. Wild-type (WT) NBs and (some) GMCs divide with their mitotic spindle oriented along the apical or basal (A/B) axis, and both A/B spindle orientation and asymmetric localization of the basal components depend on the formation and maintenance of a protein complex which is localized to the apical cortex of NBs, starting at late interphase. This apical complex includes the following: the novel protein Inscuteable (Insc) (15, 16); two PDZ domain-containing proteins, Bazooka (Baz, a Drosophila homologue of the nematode Par-3) (17, 31, 44) and DmPar-6 (26); the atypical protein kinase C DaPKC (43); and a protein with multiple TPR repeats, Partner of Insc (Pins) (23, 29, 45). Insc cannot be detected in the neuroectoderm, and its expression is first seen in the apical membrane of delaminating NBs. Among the apical components it is unique in that not only is it necessary for the A/B orientation of the mitotic spindle of NBs but its ectopic expression in ectodermal cells can cause their mitotic spindles to undergo an extra 90° rotation, resulting in an A/B orientation (12, 16).

Baz, DmPar-6, and DaPKC, unlike Insc, are expressed and apically localized in neuroectodermal cells where they are required to maintain epithelial polarity. They remain localized as apical cortical crescents and appear to be required to recruit Insc to the apical membrane in delaminating NBs. Pins, in contrast, is not apically localized in the ectoderm and is not required for the initiation of apical Insc localization in delaminating NBs. However, it is required for the integrity of the apical complex in mitotic NBs. In the absence of pins function, Insc is located in the cytoplasm, and the stability of Baz is compromised. Interestingly, apical Insc and Pins crescents are also formed in dividing GMCs, and insc function is required for the asymmetric localization and segregation of Numb. In insc mutants, Numb is distributed to both GMC progeny, and (some of) the resultant sibling neurons are unable to resolve distinct fates (3). In pins mutants, similar failures to resolve distinct sibling neuronal fates are also observed (45).

We have previously shown, using an overexpression paradigm, that a central region of Insc (amino acids [aa] 252 to 578) can act as a minimal functional domain sufficient for all aspects of insc function (40). Here we delineate the coding regions of Pins with respect to its known functions in the asymmetric cell division of NBs and attempt to understand how the Pins coding regions act to facilitate Insc apical cortical localization. We show that distinct regions of the Pins coding region are required for Insc cortical and apical localization. Moreover, the ability of Pins to localize apically can be separated from its ability to maintain apical localization of Insc. Our results further demonstrate that the C-terminal region of Pins, when ectopically expressed and mislocalized to the NB cortex, can cause severe defects in asymmetric divisions of NBs, which are not seen in NBs lacking pins function. Finally, we show that a chimeric protein containing the Insc minimal functional domain linked to the Pins C-terminal region can fulfill both insc and pins function. Our results suggest that the N-terminal region of Pins harboring the TPR repeats is required for its own apical cortical localization and for recruitment of Insc to the apical cortex. However, the C-terminal region containing the GoLoco repeats, when localized to the apical cortex (by fusing it with the Insc minimal functional domain), can fulfill all other aspects of its function in asymmetric divisions of NBs.

MATERIALS AND METHODS

Constructs for germ line transformation.

The full-length pins cDNA lacking the 3′ untranslated region (UTR) and tagged with a sequence of two tandem FLAG epitopes was inserted into pBluescript KS(+) (pKS-pins-FG-FG). Various Pins internal deletion constructions (Δ1, Δ2, Δ3, Δ4, and Δ5) were obtained with the ExSite PCR-based site-directed mutagenesis kit (Stratagene). The deletion constructs T13, T37, T47, C-Pins, and Ct-Pins were generated by the PCR method. For Insc/Pins chimeric protein, the cDNA fragment containing the sequence encoding the central region of Insc (aa 252 to 583) was amplified by PCR and ligated with the cDNA fragment encoding the C-terminal region of Pins (aa 378 to 658) and two FLAG epitopes. As a result, all the deletion constructs of Pins are tagged by two FLAG epitopes at the C terminus. The constructs were cloned into the hs-Casper and pUAST vectors for germ line transformation.

Drosophila genetics.

Transgenic flies carrying the various heat shock (hs) or upstream activation sequence (UAS) transgenes were crossed to the pins^p89 mutant background by standard genetic methods. Pins⁻ embryos and the transgene products can be identified by the lack of β-galactosidase staining (cyo Ubx-lacZ) and anti-FLAG staining, respectively.

Generation of a new anti-Pins antibody against the C-terminal region of Pins.

A fusion protein containing the C-terminal region (aa 378 to 658) of Pins was expressed as a glutathione S-transferase fusion protein and purified by using glutathione-coupled Sepharose (Amersham). Mice were immunized and given booster doses at 2-week intervals by standard methods. Specificity of sera was confirmed by the lack of staining in null mutant embryos.

Embryo collection, heat shock treatment, and immunohistochemistry.

Embryos were collected at 8-h intervals and dechorionated in 50% bleach. After several washes in a solution of phosphate-buffered saline plus 0.01% Triton X-100, embryos were incubated in a 34°C water bath for 10 min to induce ectopic Pins expression at the appropriate level and then allowed to recover for 1 h in a moist chamber at 25°C prior to fixation and immunocytochemistry. For the analysis of RP2 rescue, the pins^p89/Tm3 sb Ubx-lacZ male flies with UAS transgenes were crossed with homozygote sca-Gal4/sca-Gal4; pins^p89/pins^p89 female flies. Embryos were collected for Even-Skipped (Eve) staining. Standard fixing protocol was used for normal histochemistry assay (40). A quick fixing method was used for tubulin staining (40). Embryos were mounted in Vectashield (Vector Labs) and analyzed by laser scanning confocal microscopy (Bio-Rad MRC 1024). The following primary antibodies were used in our study: rabbit and mouse anti-C-terminal Pins antibodies (against the region from aa 378 to 658); rabbit anti-Insc, rabbit anti-Mira (from F. Matsuzaki), rabbit anti-Numb (from Y.-N. Jan), 2B8 (mouse anti-Eve, from Kai Zinn), MR1A (mouse anti-Pros, from C. Q. Doe), E7 (β-tubulin, The Developmental Studies Hybridoma Bank), mouse anti-FLAG M2 (Kodak), and mouse (Promega) and rabbit (Cappel) anti-β-galactosidase. Cy3- and fluorescein isothiocyanate-conjugated secondary antibodies were from Jackson Laboratory. The DNA dye To-Pro 3 (Molecular Probes) was used to visualize the chromosomes.

RESULTS

The C-terminal 199 aa of Pins are necessary and sufficient for targeting to the NB cortex.

The 658-aa deduced Pins protein contains seven TPR repeats (8) in its N-terminal region and three GoLoco repeats (34) in its C-terminal regions. Ten different deletion variants of Pins and a fusion protein containing the functional central domain of Insc (40) with the C-terminal region of Pins containing the GoLoco repeats were constructed, and all these proteins had two FLAG tags at their respective C termini (Fig. 1); these mutant versions of pins were placed under the control of the hsp70 promoter or UAS enhancer. Multiple independent transgenic animals carrying each of these mutant variants were generated, and the localization and function of these variant proteins were assessed in WT and pins mutant backgrounds.

FIG. 1.

Summary diagram of the Pins dissection analysis. (Left) Schematic representations of the deduced WT Pins protein and deletion variants. The first six constructs possess the pins 5′ UTR and lack the 3′ UTR. The last six constructs lack both UTRs. All constructs carry a double-FLAG tag at their extreme C termini. TPR repeats (green boxes) and GoLoco motifs (blue boxes) are indicated. (Right) Subcellular localization of the various Pins proteins induced from each of the constructs and the endogenous Insc protein were assessed in both NBs and the cells of mitotic domain 9. Abbreviations: A, apical cortical localization; Cyto, cytoplasmic localization; and Cor, cortical localization. The ability of the various constructs to reorient the mitotic spindle was analyzed in mitotic domain 9 of Pins⁻ embryos (//, spindle reorientation parallel to the ectoderm; ⊥⊥, spindle orientation perpendicular to the ectoderm). The capacity of the various constructs to correctly localize Miranda (Mir) and resolve RP2 and RP2sib fates were assessed in Pins⁻ mutants (+, WT localization; −, abnormal localization). The percentages of hemisegments with RP2 duplication (RP2 dup.) are shown (see Materials and Methods). nd, not determined.

To ascertain how the various mutant Pins proteins localize in mitotic NBs, transgene expression was confirmed by Western blot analyses on heat-shocked embryo extracts with a commercial anti-FLAG antibody (Fig. 2A). The localization of the ectopically expressed proteins was assessed in stage 10-11 embryos by performing immunofluorescence with the anti-FLAG antibody (see Materials and Methods). Several conclusions can be derived from the results shown in Fig. 2 and summarized in Fig. 1. All deletion variants which lack the C-terminal region containing the three GoLoco repeats show cytoplasmic localization. Conversely, all variants that possess the C-terminal region localize to the cell cortex. The minimal region sufficient to specify cortical localization is the extreme C-terminal 199-residue region (aa 460 to 658) containing the GoLoco repeats. However, for asymmetric apical localization, the C-terminal region is, by itself, insufficient (see next paragraph). These observations led us to conclude that while the C-terminal region of Pins (199 residues contained in Ct-Ps) is necessary and sufficient for targeting to the NB cortex, additional sequences from the N-terminal region of the protein are required for asymmetric localization to take place.

FIG. 2.

Expression and localization of the various modified Pins. (A) Western blot analyses of protein extracts from WT embryos in which expression of each of the Pins constructs was induced by heat shock. Anti-FLAG (M2) was used to detect the mutant Pins proteins and the Insc/Pins chimeric protein (Fig. 7). The numbers to the left of the gels indicate molecular mass (in kilodaltons). WT embryos ectopically expressing the various modified Pins are stained with anti-Pins (green) and a DNA dye (cyan). Ectopically expressed full-length (FL) Pins (B) and Pins deletion variants Δ2 (D) and Δ4 (F) are asymmetrically localized as apical crescents in NBs. Pins Δ1 (C), Pins Δ3 (E), and Ct-Pins (I) localize throughout the cell cortex in the majority of mitotic NBs. Pins Δ5 (G) and T17 (H) lacking the C-terminal region localize to the cytoplasm in both NBs and epithelium. The NB apical cortex is at the top.

Apical localization requires sequences from the N-terminal region of Pins.

It is apparent that residues 378 to 459 located between the TPR and GoLoco repeats are not required for Pins apical localization, since their deletion (e.g., Δ4 in Fig. 1) does not prevent normal Pins apical cortical localization (Fig. 2F). What role does the N-terminal region of Pins containing the TPR repeats play in the localization process? All of the mutant proteins containing only sequences from the N-terminal region of Pins show cytoplasmic localization, indicating an inability to target to the cortex (T17, T37, and T47 in Fig. 1 and Fig. 2G and H). However, when N-terminal sequences including the seven TPR repeats are provided in conjunction with the C-terminal sequences (e.g., Δ4 in Fig. 1 and Fig. 2F), apical localization can take place. Results with constructs in which TPR repeats 1 to 3 (Δ1) or TPR repeats 4 to 7 (Δ2) were deleted further delineated the region of Pins required for apical crescent formation. Removal of TPR repeats 1 to 3 drastically reduces the frequency of apical crescent formation (Fig. 2C), with apical crescents detected in only 10% (n = 71) of mitotic NBs. In contrast, removal of TPR repeats 4 to 7 (Δ2) does not appear to greatly affect apical localization, since 80% (n = 50) of mitotic NBs can apically localize the mutant protein (Fig. 2D). Hence, it appears that the N-terminal portion of Pins containing TPR repeats 1 to 3 is necessary for Pins apical localization. These results indicate that while the C-terminal 199 aa are sufficient for cortical localization, the N-terminal 211 aa including TPR repeats 1 to 3 are required to localize Pins apically.

Distinct requirements for apical Pins localization and for recruitment of Insc to the apical cortex.

In the absence of pins (both maternal and zygotic), Insc does not apically localize but rather takes on a cytoplasmic localization. The various mutant versions of Pins were ectopically expressed in embryos lacking both maternal and zygotic Pins (henceforth referred to as Pins⁻ embryos) to assess their capacity for restoring Insc apical localization. These embryos were double labeled with anti-FLAG and anti-Insc to simultaneously detect both mutant Pins and Insc. Not surprisingly, ectopic expression of mutant forms of Pins which show cytoplasmic or cortical localization does not alter the cytoplasmic localization of Insc in Pins⁻ embryos (Fig. 3B to D and F). However, even in Pins⁻ mitotic NBs in which the mutant Δ2 protein was apically localized, Insc remains in the cytoplasm (Fig. 3C). These results demonstrate that apically localized Δ2 mutant protein was unable to restore apical Insc localization. In contrast, ectopic expression of Δ4 which contains all seven TPR repeats along with the C-terminal 199 aa does restore apical Insc localization (Fig. 3E). Hence, apical localization of mutant Pins per se is not sufficient for apical Insc localization. The region containing TPR repeats 1 to 7 can interact with Insc (45), whereas neither Δ2 nor Δ1 mutant proteins can in yeast two-hybrid assays (data not shown); these results are consistent with the notion that the apical localization ability of Pins does not depend on its ability to interact with Insc, whereas its ability to recruit Insc to the apical cortex does.

FIG. 3.

Requirements for recruitment of endogenous Insc to the apical cortex. The positions of Insc (red) and DNA (cyan) are indicated. Ectopically expressed full-length (FL) Pins (A) and Pins Δ4 (E) in Pins⁻ mutants restore Insc apical localization. None of the other mutant Pins can restore Insc apical crescents; four examples are shown for Δ1 (B), Δ2 (C), Δ3 (D), and Δ5 (F). Note that NBs in panels B and C are identical to those in Fig. 2C and D; therefore, in pins⁻ NBs expressing Pins Δ2 (C), Insc is located in the cytoplasm, although Pins Δ2 is able to localize to the apical cortex, suggesting that Pins apical crescent formation and localization can be independent of Insc at metaphase. (G to J) Reexamination of the dependence of Pins apical localization on insc function. (G) Pins (green) form intense apical crescents in WT metaphase NBs (100% [n = 31]). In insc mutants, the majority of prophase NBs show Pins cortical localization (92% [n = 40]; data not shown); however, at metaphase, 61% of NBs (n = 60) localize Pins as crescents (H), 18% NBs show a weak crescent (I), and 21% NBs show weak cortical localization (J). The intensity of Pins is not affected in epithelial cells but is strongly reduced in 39% of insc mitotic NBs (I and J). Note that images were taken at the same gain and processed in parallel. The NB apical cortex is at the top.

We also wondered whether the ectopically expressed and apically localized Δ2 mutant form of Pins in Pins⁻ embryos might be able to restore the A/B orientation of mitotic spindles. In the cells of mitotic domain 9 (Fig. 4) as well as mitotic NBs (data not shown), mitotic spindle orientation determined by anti-tubulin staining remains defective and often fails to adopt an A/B orientation. In fact, the only two constructs that can restore A/B spindle orientation, Flag-tagged full-length Pins and Δ4, encode products which not only localize apically themselves but can also restore apical Insc localization in Pins⁻ NBs (Fig. 1 and 4). These observations are consistent with the interpretation that apical Insc is necessary to specify A/B mitotic spindle orientation and that apical Pins alone cannot replace Insc function.

FIG. 4.

Requirements for A/B mitotic spindle orientation. Expression of full-length (FL) Pins (A) and Pins Δ4 (E) can restore WT reorientation of mitotic spindle in mitotic domain 9 cells of Pins⁻ embryos. None of the other deletion variants of Pins, Δ1 (B), Δ2 (C), Δ3 (D), and Δ5 (F), can restore mitotic spindle reorientation along the A/B axis. Microtubule was visualized with E7, an anti-β-tubulin monoclonal antibody from The Developmental Studies Hybridoma Bank. Note that although Δ2 is localized apically in the cells of mitotic domain 9 (data not shown), spindle orientation remain parallel to the surface (C). Mitotic spindle orientation of NBs is also defective in all constructs except full-length Pins and Δ4, as deduced by the visualization of DNA (Fig. 5; also data not shown).

The observation that in NBs in which Insc is cytoplasmic, mutant forms of Pins (e.g., Δ2) can nevertheless take on an apical localization in a significant portion of the NBs was surprising and led us to reexamine the dependence of Pins apical localization on insc function. By using an antibody raised against a C-terminal Pins fusion protein, Pins localization in insc22 (an amorphic allele) NBs was assessed throughout mitosis. Our results indicate that while Pins localization is frequently cortical at prophase (92% [n = 40]) as previously shown (45), by metaphase, 61% (n = 60) of insc NBs show prominent crescents of Pins (Fig. 3H) (which was not observed in our previous study [45]),18% (n = 60) of NBs show Pins crescents with reduced intensity (Fig. 3I), and only 21% (n = 60) insc NBs exhibit cortical localization or drastically reduced intensity in Pins staining (Fig. 3J). The discrepancy with our previous report that Pins show primarily cortical distribution in insc mitotic NBs is due to the improved quality of the antibodies used in this study (see Materials and Methods). These data indicate that at metaphase, Pins retains significant capacity to form crescents, even in the absence of Insc, and the capacity for this insc-independent localization does not require TPR repeats 4 to 7.

Localization of the basal components and resolution of distinct sibling cell fates.

We assessed the effects of expression of the WT Pins and the mutant Δ1 to Δ5 Pins proteins on the localization of Mira and Pon, two basally localized proteins, in Pins⁻ NBs. Not surprisingly, only WT Pins and Δ4, which had the ability to apically localize the endogenous Insc when they were ectopically expressed, were able to restore WT basal localization of Mira (Fig. 5) and Pon (not shown) in Pins⁻ mitotic NBs.

FIG. 5.

Asymmetric localization of Miranda. Miranda (red) is correctly localized at metaphase and asymmetrically segregated at telophase in Pins⁻ NBs ectopically expressing full-length Pins (FL-Pins) and Pins deletion variant Δ4. Ectopic expression of Pins Δ1 and Δ3 do not restore basal localization of Miranda in Pins⁻ NBs; in these two cases, Mira crescent formation is not only disrupted at metaphase but telophase rescue also fails to occur. In contrast, 100% of telophase NBs (n = 30) in Pins⁻ embryos show proper telophase rescue of Mira. In Pins⁻ NBs ectopically expressing Pins Δ2 and Δ5, Mira localization is identical to that observed in Pins⁻ NBs; there is no restoration of basal Mira localization at metaphase, but telophase rescue occurs normally. The NB apical cortex is at the top. DNA staining is shown in cyan.

In pins, baz, and insc loss-of-function NBs, proteins which normally show basal localization are often mislocalized at metaphase. However, by late anaphase or telophase, there appears to be a compensatory mechanism which aligns a basal crescent with one of the spindle poles, leading to asymmetric segregation of the basal proteins to the smaller of the two daughter cells in essentially all of the NB divisions (31), a phenomenon called telophase rescue (4, 24). The versions of Pins with various deletions which fail to restore normal localization of the basal proteins in Pins⁻ NBs can be subdivided into types with respect to telophase rescue. At one extreme is Δ5 in which telophase rescue and asymmetric segregation of basal proteins occur normally following its ectopic expression in Pins⁻ NBs (Fig. 5K and L), despite the fact that spindle orientation and basal protein localization are abnormal at metaphase (Fig. 1 and 3). However, when Δ1 or Δ3 are expressed in Pins⁻ NBs, not only are the events associated with asymmetric divisions defective at metaphase but telophase rescue is also partially disrupted; we observe a high frequency (33% [n = 63]) of divisions in which both daughter cells inherit Mira and the orientation of mitotic spindle is fully randomized (Fig. 5C, D, G, and H); many of the NB divisions also produce daughters of equal sizes (91% [n = 90]).

We assessed the ability of these mutant Pins proteins to restore the resolution of distinct sibling cell fates. In WT embryos, the first GMC derived from NB4-2 expresses Eve (6) and divides to produce two neurons with distinct fates, RP2 and RP2sib. RP2 maintains Eve expression, while RP2sib extinguishes Eve expression. In insc and Pins⁻ mutants, Numb fails to be asymmetrically segregated and a RP2sib→RP2 cell fate transformation is frequently observed (3, 45), leading to the duplication of the RP2 neuron. When expressed in Pins⁻ embryos, WT Pins and Δ4 can rescue this phenotype and reduce the frequency of RP2 duplication seen in Pins⁻ embryos, as judged by anti-Eve staining (Fig. 1 and 6F; also data not shown). Δ2 and Δ5 cannot rescue the Pins⁻ RP2 duplication phenotype (Fig. 6D and G and Fig. 1), but the anti-Eve staining pattern is well organized and does not show gross alterations from the WT pattern. In contrast, ectopic expression of Δ1 and Δ3 in Pins⁻ embryos causes gross disorganization at stage 14 in the Eve staining pattern (Fig. 6C and E). These embryos fail to survive after stage 15. In view of the fact that the expression of these proteins in Pins⁻ NBs fails to restore Insc to the apical cortex or asymmetric localization of basal components and disrupts telophase rescue, it is likely that distinct sibling cell fates are not resolved. It is also interesting that the severe disruption in the Eve staining pattern caused by the expression of Δ1 and Δ3 correlates with the disruption of telophase rescue (see Discussion).

FIG. 6.

Resolution of distinct RP2 and RP2sib sibling cell fates. Ventral views of stage 14 or 15 embryos stained with anti-Eve. There is only one Eve⁺ RP2 neuron per hemisegment in WT embryos (A). In Pins⁻ embryos, RP2 neuron is frequently duplicated due to an RP2sib→RP2 cell fate change (B). Ectopic expression of Δ1 (C) and Δ3 (E) in Pins⁻ embryos causes gross disorganization in the Eve staining pattern but not in the overall morphology of embryos. Expression of Δ2 (D) and Δ5 (G) does not rescue the RP2 defects seen in Pins⁻ embryos. Δ4 (F) and full-length Pins (not shown) expressed in Pins⁻ embryos can rescue the RP2 duplication phenotype caused by loss of pins function (quantitated in Fig. 1). Overexpression of the Insc/Pins chimera in Pins⁻ embryo significantly rescues the RP2 defects (H) (see text). Arrowheads point to RP2 neurons. The anterior of the embryo is to the left.

Characteristics of an Insc/Pins chimeric protein.

It is known that the GoLoco motifs in the Pins C-terminal region can bind the heterotrimeric G-protein subunit, Gαi (25). A recent study (28) has shown that Pins can associate with Gαi, causing release of the Gβγ subunits, which might act via a novel receptor-independent mechanism to cause asymmetric activation of heterotrimeric G-protein signaling within NBs (see Discussion). Hence, it can be argued that the Pins C-terminal region is critical for the generation of an asymmetric signal necessary for asymmetric divisions of NBs. Would it be possible to generate a molecule containing only the C-terminal region of Pins which is able to be apically localized and fulfill Pins function for asymmetric cell divisions of NBs? To test this idea, we produced transgenic animals carrying a chimeric protein in which the C-terminal 200-aa sequence of Pins is fused to the central region of Insc (aa 252 to 583) (Fig. 7A). We have previously shown that this central region of Insc can form apical cortical crescents and is fully functional (40). We ectopically expressed this chimeric protein in the WT and in insc and Pins⁻ mutants to assess its function and its ability to asymmetrically localize. This chimeric protein localizes as an apical cortical crescents in mitotic NBs when expressed in the WT (not shown) and insc (Fig. 7B) and Pins⁻ (Fig. 7C) mutants, indicating that its apical localization does not require insc or pins. Ectopic expression of the chimeric protein in insc or Pins⁻ mutant restores not only basal crescent formation of Mira (Fig. 7B and C) and Pon (data not shown) but also A/B orientation of the mitotic spindle in domain 9 cells (Fig. 7I and I′). In epithelial cells of Pins⁻ embryos, ectopically expressed chimeric protein is also apically localized (Fig. 7E), causing dividing epithelial cells to reorient their mitotic spindles along the A/B axis (Fig. 7I and I′′). Resolution of distinct sibling cell fates (as judged by the resolution of RP2 and RP2sib fates following anti-Eve staining) can be significantly restored by the ectopic expression of the chimeric protein; 78% (n = 209) of the hemisegments are WT (Fig. 6H), as opposed to 38% WT hemisegments (n = 210) in Pins⁻ embryos. These results indicate that the chimeric protein can simultaneously compensate for both insc and pins functions. Hence, when the requirement of the TPR repeats of Pins (for mediating its own apical localization and for recruitment of Insc to the apical cortex) is obviated, as is the case with the chimeric protein, apical localization of only the C-terminal portion of Pins is sufficient to provide Pins function in asymmetric divisions of NBs (see Discussion).

FIG. 7.

A chimeric protein containing the Insc minimal functional domain and the Pins C-terminal region possesses both insc and pins function. (A) Schematic representation of the Insc/Pins chimeric protein. The ectopic expressed Insc/Pins chimeric protein (green) is localized to the apical cortex of insc (B) and Pins⁻ (C) NBs; note that the orientation of the mitotic spindle in these cells is along the A/B axis as deduced from DNA staining (cyan). In Pins⁻ embryos, the Insc/Pins chimera is also able to form an apical crescent and orientate mitotic spindles along the A/B axis in both mitotic domain 9 cells (D and I′) and dividing epithelial cells (E and I"). Double label of Insc/Pins chimeric protein (green) (F) and endogenous Insc (red) (G) in Pins⁻ NBs indicates that the chimeric protein is unable to recruit Insc to the apical cortex (H [the superimposed image of panels F and G]). In the presence of the chimeric protein, Mira (red) (B to E) is correctly localized to the basal cortex in insc (B) and Pins⁻ NBs (C). The NB apical cortex is at the top.

We have also examined the ability of ectopic chimeric protein to recruit endogenous Insc in Pins⁻ NBs. The anti-Insc antibody raised in our lab recognizes only epitope(s) from the N terminus of Insc (40) and does not bind the chimeric Insc/Pins protein. In Pins⁻ NBs, although the ectopic chimeric protein is apically localized and functional, the endogenous Insc remains cytoplasmic, as indicated by anti-Flag and anti-Insc double immunostaining (Fig. 7F to H), suggesting that the Insc/Pins chimeric protein is unable to recruit endogenous Insc to the apical cortex of dividing NBs. These findings are consistent with our earlier observation (Fig. 3) that the N-terminal TPR repeats are necessary for recruitment of Insc to the apical cortex.

Since both WT Insc and the chimeric protein possess the Insc sequences which specify its apical localization, why can ectopic chimeric protein localize asymmetrically to the apical cortex while endogenous Insc remains in the cytoplasm in Pins⁻ NBs? One obvious difference between WT Insc and the chimeric protein is the cortical targeting signal in the C-terminal region of Pins which can function in the absence of WT Pins and is part of the chimeric protein but absent in Insc. One likely explanation for the different behaviors of the two proteins is that in NBs, the Insc apical localization signal can work only if the protein is already localized to the cortex but does not work if the protein is localized in the cytoplasm; it may be that Insc has to be cortically localized before its apical localization can take place (see Discussion).

DISCUSSION

Our results demonstrate that both the C-terminal region containing the GoLoco motif (aa 460 to 658) and the N-terminal region containing TPR repeats (aa 1 to 377) are required to localize Pins to the apical cortex of NBs. The C-terminal region is necessary and sufficient to specify protein localization to the NB cortex. Cortical localization appears to be a prerequisite for Pins apical localization, since all mutant Pins proteins lacking the GoLoco motifs are cytoplasmic. In addition to the C-terminal GoLoco motifs, the minimal N-terminal region required for truncated Pins apical cortical localization contains TPR repeats 1 to 3. In Pins⁻ NBs, although the TPR repeats 1 to 3 plus C-terminal Pins variant is apically localized per se, it cannot recruit Insc from the cytosol and restore Insc apical localization; whereas TPR repeats 1 to 7 plus C-terminal Pins does have the capacity of restoring Insc apical localization. This ability to restore Insc apical localization correlates with the ability to interact with Insc, since Pins variants containing TPR repeats 1 to 7 will interact with Insc, whereas Pins variants containing only TPR repeats 1 to 3 or TPR repeats 4 to 7 will not. Although the abilities of Pins to cortically localize, to apically localize, and to restore Insc apical localization are all separable, only molecules with all three capabilities are able to restore A/B spindle orientation, basal localization of Mira and Pon, and the specification of distinct sibling cell fates when ectopically expressed in Pins⁻ embryos.

C-terminal GoLoco-containing region of Pins.

One striking feature associated with the ectopic expression of the C-terminal GoLoco-containing portion of Pins in Pins⁻ embryos is severe morphological phenotypes which are not associated with Pins⁻ embryos or embryos overexpressing WT Pins (which do not produce phenotypic defects in NBs). These defects include failure to rescue at telophase, gross misorganization of Eve-expressing neurons, and NB divisions in which the sizes of normally asymmetric daughter cells become equal. These defects are seen only when Pins variants containing the C-terminal GoLoco repeats are localized to the cortex of Pins⁻ NBs but not if they are apically localized. What might be causing these defects?

GoLoco repeats have been shown to interact with Gαi, a subunit of heterotrimeric G proteins (25). A recent study (28) has provided compelling evidence that heterotrimeric G-protein signaling plays a key role in NB asymmetric divisions. Gαi is apically localized in Drosophila mitotic NBs; moreover, it directly binds Pins in vitro and associates with Pins in vivo. Pins can act to release the βγ subunits, and biochemical experiments indicate that it is the GoLoco domains of Pins which cause the dissociation of Gαi from Gβγ. Disrupting G-protein function causes defects in mitotic spindle orientation and asymmetric localization of the normally basally localized cell fate determinants during NB divisions. These findings (28) suggest that Pins may act as a receptor-independent G-protein activator which would presumably initiate G-protein signaling from the apical cortex of NBs. It is unclear whether it is the Gαi-GDP-Pins complex or the free Gβγ which is the effector involved in maintaining NB asymmetry. Nevertheless, one possible interpretation of the phenotypes induced by the cortical localization of C-terminal Pins is that they result from gain of function due to deregulated G-protein signaling from throughout the cortex rather than from a restricted apical region of the NB cortex.

Fusion of the Insc minimal functional domain to C-terminal Pins.

If the C-terminal region of Pins is responsible and sufficient for generating the signal which maintain NB asymmetry, is it possible to create a situation in which apical cortical localization of just this C-terminal moiety (without the N-terminal region) can functionally substitute for the full-length Pins? The difficulty, of course, is that under normal circumstances the N-terminal TPR repeats are required both for recruitment of Insc and for localization of Pins to the apical cortex. We can circumvent both of these problems and obviate the need for N-terminal Pins by making a chimeric protein composed of the Insc minimal functional domain and the C-terminal region of Pins. The apical localization signal contained within the Insc minimal functional domain can localize the chimeric protein to the apical cortex and since this chimeric protein can provide Insc function, there is no longer the need to recruit the endogenous Insc to the apical cortex. Hence, this chimeric protein provides for Insc function as well as apically localizing C-terminal Pins without the need for N-terminal Pins. We have shown that when this chimeric protein is ectopically expressed in Pins⁻ embryos, it can restore normal NB (and GMC) asymmetric division (Fig. 7). These results therefore demonstrate that when the N-terminal region of Pins is no longer necessary to provide a functional Insc moiety at the apical cortex or to apically localize Pins, the apically localized C-terminal Pins can functionally replace full-length Pins and mediate all aspects of NB asymmetric division. Hence, the chimeric protein can restore the A/B orientation of the mitotic spindles, the asymmetric localization of the basal components, and the specification of distinct sibling neuronal cell fates.

Cortical and apical localization.

It is interesting that many of the proteins which are localized either as apical or basal cortical crescents in NBs can show cytoplasmic localization either at certain stages in the cell cycle or in particular mutant backgrounds. The only protein whose dynamics of localization has been studied using real-time analysis in vivo has been Pon (19). It is apparent from this study using a functional Pon/GFP fusion that the initial event shortly following entry to mitosis is the recruitment of Pon/GFP from the cytoplasm to the NB cortex and the restriction of Pon/GFP to the basal cortex occurs in an insc- and myosin-dependent manner later in the cell cycle. Although similar real-time studies have not been performed for members of the protein complex that localize to the apical cortex, a number of observations suggest that a similar series of events (cytoplasm to cortex to apical cortex) may in fact be operating to restrict these molecules to the apical cortex.

For both Insc (14) and Pins (this study), it is possible to generate deletion variants which localize either to the cytosol or to the cortex rather than to the apical cortex. In the case of Pins, it is clear that N-terminal TPR repeats 1 to 3 are required for apical localization; however, this sequence by itself shows cytoplasmic localization without the C-terminal region of Pins which is necessary and sufficient for cortical localization. These observations are consistent with the notion that for both molecules, cortical localization may be a prerequisite for apical localization. Further support for this view comes from a comparison of the localization of endogenous full-length Insc and the Insc/Pins chimera (Fig. 7) in Pins⁻ NBs. Although both molecules possess the Insc apical localization signal, the endogenous full-length Insc is cytoplasmic, whereas the chimeric molecule takes on an apical cortical localization. The simplest interpretation of the different localizations of the two molecules is that whereas the full-length Insc cannot reach the NB cortex in the absence of endogenous Pins, the C-terminal region of Pins can take the chimeric protein to the cortex; once cortical localization is achieved, the apical localization signal can then take the chimeric protein to the apical cortex. Although circumstantial, these observations suggest that the ability to reach the cell cortex may be a prerequisite for the later, more restricted localization of both apical and basal component proteins involved in asymmetric cell divisions of NBs.