Cell polarity is essential for generating cell diversity and for the proper function of most differentiated cell types. In many organisms, cell polarity is regulated by the atypical protein kinase C (aPKC), Bazooka (Baz/Par3), and Par6 proteins. Here, we show that Drosophila aPKC zygotic null mutants survive to mid-larval stages, where they exhibit defects in neuroblast and epithelial cell polarity. Mutant neuroblasts lack apical localization of Par6 and Lgl, and fail to exclude Miranda from the apical cortex; yet, they show normal apical crescents of Baz/Par3, Pins, Inscuteable, and Discs large and normal spindle orientation. Mutant imaginal disc epithelia have defects in apical/basal cell polarity and tissue morphology. In addition, we show that aPKC mutants show reduced cell proliferation in both neuroblasts and epithelia, the opposite of the lethal giant larvae (lgl) tumor suppressor phenotype, and that reduced aPKC levels strongly suppress most lgl cell polarity and overproliferation phenotypes.
Cell polarity is essential for the function of many cell types, as well as for generating cell diversity through asymmetric cell division. Over the past few years, it has become clear that cell polarity in many cell types and organisms is regulated by the evolutionarily conserved protein complex containing Baz/Par3 (Par3 in Caenorhabditis elegans, Bazooka in Drosophila, ASIP in mammals), Par6, and atypical protein kinase C (PKC-3 in C. elegans, aPKC in Drosophila, and PKCζ and λ in mammals; Doe, 2001; Ohno, 2001; Wodarz, 2002). The Baz/Par3 and Par6 proteins contain PDZ protein interaction domains, whereas aPKC is a serine/threonine kinase (Etemad-Moghadam et al., 1995; Hung and Kemphues, 1999; Suzuki et al., 2001). Each of the Baz/Par3, Par6, and aPKC proteins is reported to bind the others in vitro (Tabuse et al., 1998; Joberty et al., 2000; Lin et al., 2000; Wodarz et al., 2000; Suzuki et al., 2001; Betschinger et al., 2003), and the three proteins are reported to be interdependent for their normal localization in C. elegans blastomeres (Etemad-Moghadam et al., 1995; Tabuse et al., 1998; Hung and Kemphues, 1999), mammalian epithelia (Joberty et al., 2000; Suzuki et al., 2001), and Drosophila epithelia and neuroblasts (Wodarz et al., 2000; Petronczki and Knoblich, 2001).
The Baz/Par3–Par6–aPKC complex was initially characterized for its role in the proper partitioning of cell fate determinants during the asymmetric cell division of early blastomeres in C. elegans. In this system, the Par3–Par6–aPKC complex is localized to the anterior cortex of the one-cell embryo and is required to restrict the Par1 and Par2 proteins to the opposite posterior cell cortex (Kemphues, 2000). Analyses of cell polarity in other systems and cell types have shown the Baz/Par3–Par6–aPKC complex to be a widely used and flexible set of proteins that functions in conjunction with many proteins other than Par1 and Par2.
In Drosophila, epithelial cells have an apical/basal polarity that is reflected in the establishment of an apical membrane domain, apical adherens junctions, slightly more basal septate junctions, and a basolateral membrane domain (Tepass et al., 2001). The Baz/Par3–Par6–aPKC protein complex is targeted to the apical membrane domain, and is enriched at the adherens junctions (Tepass et al., 1990; Kuchinke et al., 1998; Wodarz et al., 2000; Bachmann et al., 2001; Hong et al., 2001; Petronczki and Knoblich, 2001).
A modified form of apical/basal polarity is present in mitotic Drosophila neuroblasts, where it is used to generate cellular diversity by asymmetric cell division. Drosophila neuroblasts divide asymmetrically along their apical/basal axis to regenerate an apically positioned neuroblast and bud off a smaller basally positioned ganglion mother cell (GMC). The GMC undergoes one subsequent cell division to generate neurons or glia (Goodman and Doe, 1993). The neuroblast apical domain is derived from the epithelial apical domain during the process of neuroblast delamination, and shows apical localization of the Baz/Par3–Par6–aPKC proteins during mitosis. Unlike epithelial cells, neuroblasts contain the protein Inscuteable (Insc), which is required to maintain apical localization of the Baz/Par3–Par6–aPKC proteins (Schober et al., 1999; Wodarz et al., 1999, 2000; Petronczki and Knoblich, 2001; Albertson and Doe, 2003). These proteins are required to establish a basal domain in mitosis. Among others, the Miranda protein and its cargo Prospero transcription factor and Staufen RNA-binding protein are localized to this domain (Doe and Bowerman, 2001).
A second group of proteins—Discs large (Dlg), Scribble (Scrib), and Lethal giant larvae (Lgl)—has been shown to regulate apical/basal cell polarity in Drosophila epithelia and neuroblasts (Bilder et al., 2000; Bilder and Perrimon, 2000; Ohshiro et al., 2000; Peng et al., 2000; Albertson and Doe, 2003). Members of this complex are rich in protein–protein interaction domains. Dlg contains one SH3 and three PDZ domains (Woods and Bryant, 1991), Lgl contains four WD40 repeats (Strand et al., 1994), and Scrib contains four PDZ domains as well as sixteen leucine-rich repeats (Bilder and Perrimon, 2000). Dlg and Scrib interact with a common binding protein called Gukholder (Mathew et al., 2002); Lgl has not been shown to physically interact with any other member of the group. In Drosophila epithelia, Dlg, Lgl, and Scrib are localized to the septate junction and basolateral membrane domain, where they are required for restricting the Baz/Par3–Par6–aPKC complex to the apical membrane domain (Bilder and Perrimon, 2000). In mitotic neuroblasts, Dlg, Scrib, and Lgl proteins localize to the cortex with apical enrichment, and are required for targeting basal proteins to the cortex (Ohshiro et al., 2000; Peng et al., 2000; Albertson and Doe, 2003). Their functions in addition to basal protein targeting include limiting cell proliferation (Humbert et al., 2003) and promoting spindle asymmetry in neuroblasts (Albertson and Doe, 2003). Recently, a connection has been made between the Baz/Par3–Par6–aPKC complex and the Dlg/Scrib/Lgl group of proteins. In both Drosophila and mammals, it has been shown that aPKC directly binds and phosphorylates Lgl (Betschinger et al., 2003; Yamanaka et al., 2003). In mammals, this phosphorylation is proposed to regulate epithelial tight junction formation (Yamanaka et al., 2003), and in Drosophila to regulate neuroblast basal protein targeting (Betschinger et al., 2003). Here, we investigate the role of aPKC in regulating cell polarity and cell proliferation in Drosophila, and the genetic interactions between aPKC and lgl. We conclude that aPKC is required for proper apical/basal polarity in neuroblasts and epithelia; that Baz/Par3 and aPKC have distinct functions in neuroblasts; and that a major function of aPKC is to suppress Lgl activity.
aPKC zygotic mutant embryos have normal epithelial and neuroblast polarity
It was previously reported that embryos homozygous for the aPKC k06403 allele, which contains a P element inserted into the third intron of the aPKC gene, died before gastrulation (Wodarz et al., 2000). We used the same aPKCk06403 mutation, but found that homozygous mutants survived to mid-larval stages (Table I) . We verified the presence of the aPKCk06403 mutant allele on all single- and double-mutant chromosomes by PCR (Fig. 1 A). To further understand the aPKCk06403 mutant allele, we assayed aPKC mRNA and protein levels in homozygous mutant larvae. Northern blots of homozygous mutant second instar larval brain tissue revealed much reduced levels of aPKC mRNA (Fig. 1 B), and antibody staining showed no detectable full-length protein in late second instar brain or imaginal disc tissue (Fig. 1 C; unpublished data). Thus, the aPKCk06403 allele is close to a molecular null allele. To determine if aPKCk06403 behaves as a genetic null allele, we compared aPKCk06403/aPKCk06403 to aPKCk06403/Df(2R)JP1 (a deficiency that removes the entire aPKC locus) for lethal phase and neuroblast polarity phenotypes. We find indistinguishable phenotypes for both genotypes (Table I), indicating that aPKCk06403 acts as a genetic null allele.
Stage 15 embryos homozygous for aPKCk06403 (subsequently called aPKC mutants) have readily detectable maternal aPKC protein (unpublished data) and normal epithelial and neuroblast polarity, as assayed by multiple apical and basal markers (unpublished data). We conclude that zygotic aPKC is dispensable for embryonic epithelial polarity, neuroblast polarity, and embryonic viability.
aPKC zygotic mutants have defects in larval neuroblast cell polarity
aPKC zygotic mutants died as late second instar larvae, which allowed us to analyze neuroblast cell polarity in mid-second instar larval brains. In wild-type mitotic neuroblasts, we detected apical cortical localization of Baz/Par3, Pins, Lgl, and Par6 throughout mitosis (Fig. 2). aPKC mutant larvae were identified by the lack of GFP expression from the balancer chromosome. In these mutants, Par6 and Lgl proteins were delocalized into the cytoplasm in all metaphase neuroblasts; the only enrichment of either protein was faint ectopic Par6 at the cleavage furrow in telophase (Fig. 2). In contrast, all mutant neuroblasts showed normal apical localization of Baz/Par3, Pins, and Dlg proteins (Fig. 2). We conclude that aPKC is essential for Lgl and Par6 apical localization, but is not required for Baz/Par3, Insc, Pins, or Dlg apical localization.
Next, we investigated the role of aPKC in basal protein localization in second instar larval neuroblasts. In the wild type, Miranda protein is localized to the basal cortex from prometaphase through telophase, ultimately being partitioned specifically into the basal GMC (Fig. 2). In aPKC mutants at late second instar (when the larvae become uncoordinated), Miranda is localized around the entire cortex from metaphase through telophase in all neuroblasts examined (Fig. 2); earlier in second instar (when larvae are moving normally), about half the neuroblasts still showed weak basal enrichment at telophase (Table I). We conclude that aPKC function is required to exclude Miranda from the apical cortex of the neuroblast, even during telophase when basal localization becomes Baz/Par3-independent (Schober et al., 1999; Peng et al., 2000; Albertson and Doe, 2003; Cai et al., 2003).
It is commonly accepted that the Baz/Par3–Par6–aPKC complex is required for orienting the neuroblast mitotic spindle orthogonal to the overlying ectodermal cell layer, thus reliably placing the GMCs toward the inside of the embryo (Doe and Bowerman, 2001; Jan and Jan, 2001; Ahringer, 2003). We could not assay the role of aPKC in embryonic neuroblast spindle orientation due to the large pool of maternal aPKC protein, but we did assay spindle orientation in aPKC mutant larval neuroblasts. We observed that the metaphase spindle was always centered on the apical Baz/Par3 crescent in these mutant neuroblasts (Fig. 2, insets). Thus, aPKC is not required for aligning the mitotic spindle along the intrinsic axis of apical/basal polarity within larval neuroblasts, and it is likely that Par6 is also dispensable, as it is completely delocalized in aPKC mutants.
To determine the cell-autonomous aPKC phenotype, and to determine whether a more severe neuroblast phenotype could be observed when neuroblast life was not curtailed by larval death, we generated aPKC mutant clones. The MARCM system (Lee and Luo, 2001) was used to generate positively marked mutant clones in 0–12-h aPKC/+ first instar larvae, and the phenotype was assayed in third larval instar brains. By this time, aPKC was completely undetectable in neuroblasts within the clone (Fig. 3 A, arrowhead). Wild-type neuroblasts outside the clone (aPKC/+ or +/+) showed normal apical Insc, apical Baz/Par3, and basal Miranda localization (Fig. 3, B and D). As in mutant brains, Baz/Par3 and Insc proteins were localized normally to the apical neuroblast cortex (Fig. 3, C and E), and Miranda lost its normal restriction to the basal cortex (Fig. 3 C). We conclude that aPKC is essential for Par6 and Lgl apical localization and for excluding Miranda from the apical cortex, but it has no role in apical localization of Baz/Par3, Insc, Pins, Dlg, or in spindle orientation.
aPKC zygotic mutant larvae have defects in epithelial cell polarity
We assayed epithelial cell polarity in wild-type and aPKC zygotic mutant eye imaginal discs, which were identified based on their characteristic connection to the larval brain via the optic nerve (Fig. 4 C, brackets). Wild-type (aPKC/+) second larval instar eye disc epithelia formed a smoothly folded cellular monolayer, and showed the expected distribution of apical/basal cell polarity markers; aPKC and Discs lost were localized to the apical plasma membrane, E-cadherin (Ecad) was localized to the subapical adherens junctions, and Dlg was localized to the even more basal septate junction and basolateral membrane domain (Fig. 4 A; unpublished data). In contrast, aPKC mutant second larval instar eye imaginal discs were not organized into an epithelial monolayer (Fig. 4, B–D). There was no detectable aPKC protein, and Ecad/Dlg staining was less intense than in the wild type (Fig. 4 B), either because the proteins were more widely distributed or because there was less protein present. When the gain was increased to visualize the low levels of Ecad and Dlg proteins, we observed Ecad and Dlg proteins distributed around the cell cortex rather than confined to their normal apical/basal domain (Fig. 4, C and D; unpublished data). Thus, the aPKC mutant eye imaginal disc epithelia appear to have a loss of epithelial polarity characterized by unpolarized cortical distribution of an apical adherens junction marker (Ecad) and a basolateral marker (Dlg).
aPKC is required for cell proliferation in neuroblasts and epithelia
During the course of analyzing the aPKC mutant clones, we observed a dramatic decrease in the number of neuronal progeny in the positively marked aPKC mutant clones compared with positively marked wild-type control clones. We focused our analysis on the well-characterized mushroom body neuroblasts: each neuroblast first generates neurons that send axons into the γ lobe of the adult mushroom body, and subsequently switches to producing neurons that have axon projections into the α and β lobes (Lee et al., 2000). We found that wild-type clones contained 300–400 neurons (n = 2), consisting of early-born γ neurons as well as later-born α and β neurons (Fig. 4 E). In contrast, aPKC clones contained 75–125 neurons (n = 3), and consisted primarily of early-born γ neurons (Fig. 4 F). Because few α and β neurons were present in the mutant clones, the most likely explanation for reduced clone size is the cessation of neuroblast division just after the early-born γ neurons were produced. A similar defect in cell proliferation was observed in aPKC mutant eye imaginal discs: aPKC mutant discs were always much smaller than normal second instar eye discs (Fig. 4, C and D), although the animals remained as second instar larvae for at least twice as long as normal. Thus, aPKC is required to maintain cell proliferation of larval neuroblasts and imaginal disc epithelial cells.
aPKC and lgl show dosage-sensitive negative genetic interactions
aPKC binds and phosphorylates Lgl in both mammals and Drosophila (Betschinger et al., 2003; Yamanaka et al., 2003), so we investigated possible genetic interaction between aPKC and lgl in regulating cell polarity and cell proliferation. First, we examined whether reduced aPKC could suppress an lgl neuroblast cell polarity phenotype in stage 15 embryos. In wild-type neuroblasts, Miranda was localized to the basal cortex, but in lgl4 mutants (which retain low levels of maternal Lgl protein; unpublished data), neuroblasts were observed with Miranda delocalized from the cortex into the cytoplasm and onto the mitotic spindle (Fig. 5, A and B; Table II). In contrast, lgl4 aPKC double mutant embryos showed a clear suppression of the lgl4 polarity phenotype, with more Miranda basal crescents than lgl4 single mutants (Fig. 5 C; Table II).
To test whether reduced aPKC levels could suppress the lgl brain tumor phenotype, we examined brain size and neuroblast numbers in early third instar larvae 74 h after larval hatching. Wild-type brains contained an average of 90 neuroblasts, but similarly staged lgl mutant brains were larger and possessed ∼135 neuroblasts (Fig. 5, D and E; Table II). When aPKC levels were reduced by 50% (lgl aPKC/lgl+), we observed a clear reduction in brain size and neuroblast number (Fig. 5 F; Table II). Thus, reduced aPKC levels suppressed the lgl brain tumor phenotype as well as the lgl neuroblast cell polarity phenotype.
Next, we tested whether reduced aPKC activity could suppress the previously described lgl epithelial cell polarity phenotype (Bilder et al., 2000). In wild-type larvae, late third instar leg and wing imaginal discs have a well-defined morphology (Fig. 6 A) with a smooth monolayer of columnar cells that have a clear apical/basal polarity (Fig. 6, D and D′). In contrast, lgl334 mutant third instar imaginal discs are highly disorganized and often fused with each other or nearby tissues (Fig. 6 B); at the cellular level, mutant epithelial cells are round instead of columnar, with an expanded apical membrane domain (Fig. 6, E and E′). Strikingly, lgl334 mutants that are heterozygous for an aPKC mutation have a nearly normal late third instar leg and wing imaginal disc morphology (Fig. 6 C), including columnar cell shape and an approximately normal apical membrane domain (Fig. 6, F and F′; Table II). In addition, we characterized an earlier stage of imaginal disc development (early third instar; 74 h after larval hatching at 25°C). We find that lgl mutants already show morphological disorganization of imaginal discs (Fig. 6 H), but both wild-type and lgl aPKC/lgl+ mutants show essentially normal leg and wing imaginal disc morphology (Fig. 6, G and I; unpublished data). We conclude that reduced aPKC levels can suppress the lgl epithelial cell polarity phenotype.
Are all lgl phenotypes suppressed by reduced aPKC levels? Recently, we have shown that ∼12% of the neuroblasts in lgl4 homozygotes have a reduced apical cortical domain and apical spindle poles, leading to an inverted asymmetric division and a smaller neuroblast size (Albertson and Doe, 2003). This lgl phenotype was not suppressed by reducing aPKC levels, and in fact may have been slightly enhanced (Table II, row 3). The inability of aPKC mutation to suppress the lgl “small neuroblast” or “inverted division” phenotype is consistent with a recent observation that aPKC positively regulates neuroblast cell size (Cai et al., 2003). Thus, aPKC and Lgl appear to cooperate to positively regulate neuroblast cell size, despite their negative interaction in regulating other aspects of cell polarity.
It was previously reported that aPKCk06403 zygotic mutant embryos died before gastrulation and that all epithelial and neuroblast polarity was lost (Wodarz et al., 2000). We find a much later onset phenotype: survival of aPKCk06403 homozygotes to second larval instar with no significant embryonic phenotype. Although surprising, we feel our results reflect the authentic zygotic aPKCk06403 mutant phenotype for several reasons. First, recent work from the Wodarz lab confirms that aPKCk06403 homozygotes survive to at least late embryonic stages without significant epithelial or neuroblast defects (Wodarz, A., personal communication), presumably due to the absence of deleterious second site mutations that were on the original aPKCk06403 chromosome. Second, we used PCR to verify the presence of the aPKCk06403 allele in all of our stocks, and used dominantly marked balancer chromosomes to independently confirm the identity of every aPKCk06403 homozygote that we analyzed. Third, we showed that aPKCk06403 homozygous second instar larvae had very low levels of aPKC mRNA and no detectable full-length protein; thus, we analyzed a very strong or null aPKC mutant phenotype. Fourth, we showed that aPKCk06403/aPKCk06403 and aPKCk06403/Df(2R)JP1 larvae have indistinguishable lethal phases and quantitative neuroblast phenotypes; thus, aPKCk06403 behaves as a genetic null allele. Finally, we showed that aPKCk06403 homozygous embryos had persistent maternal aPKC protein, explaining the lack of an embryonic phenotype in these mutants.
One of our more unexpected findings is that aPKC mutant neuroblasts show normal Baz/Par3 apical localization. The Baz/Par3–Par6–aPKC complex has been suggested to form a functional unit that is interdependent for localization in C. elegans, mammals, and Drosophila (Doe and Bowerman, 2001; Ohno, 2001). We find that Baz/Par3 shows normal apical localization in aPKC mutant neuroblasts, showing that normal Baz/Par3 localization can occur without being part of the Par3–Par6–aPKC complex. In addition, we show that neuroblasts lacking apical aPKC and Par6 still form a molecularly defined apical cortical domain containing Baz/Par3, Insc, Pins, and Dlg. Our results allow us to propose a hierarchy for apical protein localization in neuroblasts: Baz/Par3–Insc–Pins–Dlg→aPKC→Par6–Lgl (Schober et al., 1999; Wodarz et al., 1999; Albertson and Doe, 2003). Our hierarchy is consistent with recent biochemical analyses in which a protein complex was isolated containing Par6–aPKC–Lgl, but not Baz/Par3 (Betschinger et al., 2003; Yamanaka et al., 2003). We suggest that aPKC may be required to anchor the Par6–aPKC–Lgl complex at the apical cortex of the neuroblast.
Both aPKC and Lgl are required for Miranda basal localization in neuroblasts, and all available data support a model in which Lgl is required for targeting Miranda to the neuroblast cortex, whereas aPKC blocks Lgl function on the apical side of the neuroblast. First, lgl mutants have little or no Miranda at the cortex (Ohshiro et al., 2000; Peng et al., 2000; Albertson and Doe, 2003). Second, aPKC mutants show uniform cortical Miranda localization (this paper). Third, a weak lgl phenotype can be suppressed by reducing aPKC levels, showing that aPKC activity antagonizes Lgl activity (this paper). Fourth, aPKC and Lgl physically interact (Betschinger et al., 2003). Finally, an overexpressed nonphosphorylatable Lgl protein is uniformly cortical and able to induce uniform cortical Miranda localization, whereas phospho-Lgl is preferentially released from the cell cortex (Kalmes et al., 1996; Betschinger et al., 2003). This has led to a model in which Lgl acts as an anchor for Miranda at the basal cortex, but is absent from the apical cortex due to aPKC-mediated phosphorylation (Betschinger et al., 2003). Although this simple model is attractive, we note that Lgl has never been observed colocalized with Miranda in a basal cortical crescent (Ohshiro et al., 2000; Peng et al., 2000; Albertson and Doe, 2003; Betschinger et al., 2003), and a role for cytoplasmic Lgl in Miranda localization has not been definitively ruled out.
The Baz/Par3–Par6–aPKC complex has a well-characterized role in regulating neuroblast spindle orientation (Doe and Bowerman, 2001; Jan and Jan, 2001; Knoblich, 2001; Ahringer, 2003). Spindle orientation can be measured relative to extrinsic landmarks around the neuroblast (e.g., perpendicular to the overlying ectoderm) or relative to intrinsic cues within each neuroblast (e.g., perpendicular to the Baz/Par3–Par6–aPKC apical crescent). Mutations in baz or par6 genes randomize embryonic neuroblast spindle orientation relative to the overlying ectoderm (Kuchinke et al., 1998; Wodarz et al., 1999; Petronczki and Knoblich, 2001), but it is not clear whether these phenotypes are due to disruption of the ectodermal layer or to a cell-autonomous defect in the neuroblast. We cannot assay the function of aPKC in embryonic neuroblast spindle orientation due to high levels of maternal aPKC protein present in aPKC zygotic mutant embryos. However, we do show that aPKC is not required for intrinsic spindle orientation in larval neuroblasts; the mitotic spindle is always perpendicular to the Baz/Par3–Insc–Pins apical crescent.
The Baz/Par3 and Par6 proteins are required to establish epithelial polarity in Drosophila (Muller and Wieschaus, 1996; Schober et al., 1999; Wodarz et al., 1999; Petronczki and Knoblich, 2001), and we show here that aPKC is also required for normal apical/basal epithelial cell polarity. The similar phenotype in baz, par6, and aPKC mutants may indicate that these proteins function together as a complex to provide a single function in epithelia, despite the evidence that they have independent functions in neuroblasts. One primary function may be the inhibition of Lgl activity because we find that the lgl epithelial polarity defects can be strongly suppressed by reducing aPKC levels. Lgl–Dlg–Scrib activity can also antagonize Baz/Par3–Par6–aPKC activity (Bilder et al., 2003), and it is tempting to speculate that aPKC inactivates Lgl by phosphorylation, whereas Lgl can inactivate aPKC by sequestering it into an Lgl–Par6–aPKC complex and out of the Baz/Par3–Par6–aPKC complex (Betschinger et al., 2003; Yamanaka et al., 2003).
The role of aPKC in cell proliferation has not been previously investigated in Drosophila. There are three lines of evidence showing that aPKC promotes cell proliferation in neuroblasts and epithelia. We find that the number of cells in aPKC mutant mushroom body neuroblast clones is significantly lower than the number in wild-type clones. There appears to be a normal number of early-born γ neurons in these clones, followed by only a few later-born α and β neurons. The normal number of early-born γ neurons suggests that loss of aPKC does not lead to cell death in this population; moreover, we see no decrease in the number of neuroblasts per brain lobe in aPKC mutant larvae (unpublished data). We conclude that cell death is not contributing to the reduction in neurons observed in the clones, but rather, that the neuroblast stops dividing near the time the neuroblast switches over to generating α and β neurons. The neuroblast may become arrested at some point in the cell cycle, or it may undergo a terminal division to generate a pair of GMCs (perhaps due to both daughter cells inheriting Miranda and Prospero GMC determinants). A second indication that aPKC promotes cell proliferation is that we observe far fewer epithelial cells in aPKC mutant eye imaginal discs compared with the wild type, even with an additional day of growth as second instar larvae. Finally, we observe that a 50% reduction in aPKC levels (aPKC/+) can strongly suppress the epithelial and brain overproliferation phenotypes of lgl mutants. Together, our data show that aPKC positively regulates cell proliferation in epithelia and neuroblasts. Interestingly, reduction in the function of the mammalian atypical PKCζ (using overexpression of a dominant-negative kinase) can suppress Rac1/cdc42-induced overproliferation (Qiu et al., 2000). Thus, aPKC may have an evolutionarily conserved role in promoting cell proliferation, as well as in the establishment of cell polarity.
Although aPKC and Lgl act antagonistically to regulate many aspects of epithelial and neuroblast cell polarity and cell proliferation, they may share a common positive function in regulating neuroblast apical cell size. Previously, we showed that lgl zygotic mutants have some embryonic telophase neuroblasts with an abnormally small apical cortical domain, apical spindle pole, and neuroblast size (Albertson and Doe, 2003). These defects are not suppressed by reducing aPKC levels, and in fact may be enhanced (Table II). Thus, we propose that Lgl and aPKC both act positively to promote large apical cell and spindle pole size. It has been reported that Baz/Par3–Par6–aPKC and Pins–Gαi act in parallel pathways to promote large apical cell size and apical spindle size (Cai et al., 2003); Lgl could be acting as part of the Pins–Gαi pathway, or as a third input promoting apical cell and spindle size.
Materials And Methods
Fly genetics and morphological analysis
The aPKCk06403, lgl4, lgl334, and Df(2R)JP1 stocks were obtained from the Bloomington Stock Center (Bloomington, IN), and rebalanced over CyO actin::GFP. All alleles are described in Flybase (http://flybase.bio.indiana.edu:82/).
Genomic DNA was isolated as described by the Berkeley Drosophila Genome Project (http://www.fruitfly.org/about/methods/inverse.pcr.html). PCR primers were designed to detect the presence of the PlacW transposon in the third intron of aPKC as it had been described previously (Wodarz et al., 2000). One primer was complementary to the 3′ end of the transposon (5′-CACTCGCAGTTATTGCAAGC-3′), and the other was complementary to the third exon of aPKC (5′-TCGCCATCGCAAGACAATCCCG-3′). PCR was performed using the Expand system (Roche).
RNA was isolated from second larval instar brains 48 h after hatching. Brains were dissected in Schneider's medium and immediately transferred to TRIzol® reagent (Invitrogen); RNA was isolated according to the manufacturer's directions. RNA was separated on a formaldehyde gel and transferred to zetabind membrane (Cuno, Inc.). Random-primed probe was prepared using the entire aPKC coding region, and the SV40 poly(A)+ region from pUAST, as a template.
Antibody production, staining, and imaging
The Par6 antibody was generated in rats against aa 130–255. The protein fragment was isolated with a His tag that was subsequently removed. The Lgl antibody was generated against an Lgl COOH-terminal 21-aa peptide (Albertson and Doe, 2003); using our fixation protocol, this antisera detects only apical cortical Lgl protein, and not the uniform cortical or cytoplasmic protein reported for other Lgl antisera (Ohshiro et al., 2000; Peng et al., 2000).
Larvae were raised on standard food. Brains and imaginal discs were dissected in Schneider's Medium (Sigma-Aldrich), fixed for 20 min in 4% formaldehyde, and stained as described previously (Cox et al., 1996). Embryos were fixed and stained as described previously (Albertson and Doe, 2003). The primary antibodies used were mouse anti-Dlg 4F3E3 (1:100; Parnas et al., 2001), rat anti-Scribble (1:2,500; Albertson and Doe, 2003), rat anti-DE-cadherin (1:100; Oda et al., 1994), rabbit anti-aPKC (1:1,000; Santa Cruz Biotechnology, Inc.), mouse anti-α-tubulin (1:5,000; Sigma-Aldrich), rat anti-Lgl (1:150; Albertson and Doe, 2003), rabbit anti-Miranda (1:1,000; Albertson and Doe, 2003), rat anti-Par6 (1:200), rabbit anti-Baz/Par3 (1:500; Wodarz et al., 2000), rabbit anti-phosphohistone H3 (1:1,000; Upstate Biotechnology), rabbit anti-Insc (1:1,000; Kraut et al., 1996), and rabbit anti-Pins (1:1,000; Yu et al., 2000). The secondary antibodies Cy5- and Alexa® Fluor 488–conjugated goat anti–rabbit IgG, Cy5- and Alexa® Fluor 488–conjugated goat anti–rat IgG, and Cy3-conjugated goat anti–mouse IgG (Jackson ImmunoResearch Laboratories or Molecular Probes, Inc.) were used at 1:400. Samples were mounted in AquaPolymount (Polysciences, Inc.) and imaged on a confocal microscope (Radiance; Bio-Rad Laboratories). Low magnification transmitted light images of larval brains and imaginal discs were collected on a microscope (Axioplan; Carl Zeiss MicroImaging, Inc.). Confocal images were processed with ImageJ (National Institutes of Health, Bethesda, MD), and figures were assembled in Adobe Photoshop® v6.0.
Generating positively marked aPKC mutant clones
We used the MARCM system to generate clones (Lee et al., 2000), using stocks available from the Bloomington Stock Center. To generate mushroom body aPKC clones, elavGAL4 UASmCD8 hsFlp;FRTG13 tubPGAL80 flies were crossed to FRTG13 aPKCk06403/CyO flies. Heat shock of first instar larvae from this cross at 37°C for 1 h induced Flp-mediated recombination at FRT sites that generated homozygous aPKC clones that lacked GAL80. In these clones, GAL4-mediated transcription of mCD8-GFP was derepressed, thereby positively marking the mutant clones. Clones were examined in third larval instar brains and 2–5-d-old adults.
We thank the Bloomington stock center for fly stocks; Rhiannon Penkert for generating the Par6 fragment; Mike Marusich at the University of Oregon Hybridoma Lab for assistance in antibody generation; Andreas Wodarz for communicating unpublished results; and Greg Kothe for help with Northern blots. We thank Karsten Siller and Sarah Siegrist for comments on the manuscript.
This work was supported by Damon Runyon postdoctoral fellowships (to M. Rolls and C.-Y. Lee), a National Institutes of Health graduate student training grant (to R. Albertson), and by the Howard Hughes Medical Institute, where C.Q. Doe is an Investigator.
M. Rolls and R. Albertson contributed equally to this paper.
Abbreviations used in this paper: aPKC, atypical protein kinase C; Dlg, Discs large; Ecad, E-cadherin; GMC, ganglion mother cell; Insc, Inscuteable; Lgl, Lethal giant larvae; Scrib, Scribble.