Drosophila Patj plays a supporting role in apical-basal polarity but is essential for viability

Polarity proteins play evolutionarily conserved roles in regulating apical-basal polarity in epithelial cells (Wang and Margolis, 2007). Pals1-associated tight junction protein (Patj) is a multi-PDZ domain protein that binds the apical polarity protein complexes Crumbs (Crb)-Stardust (Sdt; Pals1) (Roh et al., 2002; Roh et al., 2003) and Par-6–aPKC (Nam and Choi, 2003; Wang et al., 2004). In Drosophila, the homolog of Patj was first identified as the product of discs lost (dlt) (Bhat et al., 1999), but Pielage et al. later showed that dlt actually encodes a protein involved in cell cycle control and the original dlt was renamed Drosophila Patj (dPatj) (Pielage et al., 2003). RNAi knockdown results suggested that dPatj is an essential gene that is required for initiating and establishing apical-basal polarity, primarily through its regulation of Crb (Bhat et al., 1999). However, Pielage et al. generated a synthetic dPatj mutant by rescuing all of the deleted genes with the exception of dPatj in a small deficiency and found it to be viable, disputing essential roles for dPatj in polarity and development (Pielage et al., 2003). Such conflicting results were partially reconciled when it was shown that this synthetic dPatj mutant is in fact a hypomorphic allele (dPatj^synhypo) that expresses a truncated dPatj protein from one of the rescuing DNA fragments (Nam and Choi, 2006). They constructed a new synthetic null mutant, dPatj^synnull, that was free of dPatj coding sequences and found that it was early larval lethal. Unfortunately, polarity defects could not be assessed in dPatj^synnull embryos due to difficulties in removing dPatj maternal contributions (Nam and Choi, 2006).

Thus, owing to the lack of suitable genetic mutants, it remains to be determined whether Patj is indeed essential for apical-basal polarity and general development. Here, we generated multiple molecularly and genetically defined null mutants of dPatj to address this key issue.

y¹ w^67c23; P{GSV2}GS50262/TM3, Sb¹ Ser¹ (stock #204965) was obtained from the Kyoto Drosophila Genetic Resource Center. Stocks obtained from the Bloomington Drosophila Stock Center: w¹¹¹⁸; Df(3L)BSC123/TM6B, Tb¹ (BL#9143); w^*; FRT-2A (BL#1997); w^*; Dr¹/TMS, P{ry[+t7.2]=Δ2-3}99B (BL#1610); hsFLP¹ y¹ w¹¹¹⁸; Dr¹/TM3, Sb¹ (BL#26902); w¹¹¹⁸; Ubi-GFP.nls^3L1 Ubi-GFP.nls^3L2 FRT-2A (BL#5825); y¹ vas-ϕC31^ZH-2A w^*; attP-9A^VK00020 (BL#24867); w^*; ovoD^1-18 FRT-2A/st¹ βTub85D^D ss¹ e^s/TM3, Sb¹ (BL#2139); ovo-FLP, w^* (BL#8727); y^d2 w¹¹¹⁸ ey-FLP (BL#5580); and w¹¹¹⁸; Dr^Mio/TM3, P{GAL4-twi.G}2.3, P{UAS-2xEGFP}AH2.3, Sb1 Ser¹ (BL#6663; hereafter referred to as w; Dr/TM3 twi>GFP). The knockout null allele crb^GX24w[–] (crb^ko) was described previously (Huang et al., 2009).

P{GSV2}GS50262 excision mutagenesis was performed according to the standard method (Hummel and Klämbt, 2008). dPatj^Δ7, dPatj^Δ13 and dPatj^Δ14 were isolated from a total of 253 excision candidates screened by Df(3L)BSC123 complementation. PCR verifications and sequencing were performed using primers (5′-3′) CGAGACGCGGCCGCGGCAGACACCTTATTGTTTTTCC and CGAGACGCTAGCCTAGTTCCGCCAGTCGGGAA. dPatj-L and dPatj-S DNA fragments were PCR amplified from genomic DNA using: dPatj-S Forward, CGAGACGCGGCCGCGGCAGACACCTTATTGTTTTTCC; dPatj-L Forward, CGAGACGCGGCCGCATTCCGGGTGATATTATTTTCATTG; and dPatj-S (or -L) Reverse, CGAGACACGCGTACTCATTGACCGTACGGAAGGTA.

GFP was inserted before the dPatj stop codon in dPatj-S to produce dPatj::GFP-S. dPatj-L and dPatj::GFP-S were integrated into attP-9A^VK00020 [VK20 at the cytologic location 99F8 (Venken et al., 2006)] via ϕC31-mediated DNA integration (Groth et al., 2004; Bischof et al., 2007) to generate dPatj-L and dPatj::GFP-S, which were then recombined with dPatj^Δ7.

dPatj^Δ7 germline clone embryos were collected from crosses of ovo-FLP; dPatj^Δ7 FRT-2A/ovoD FRT-2A × dPatj^Δ7 [or dPatj^Δ13 or Df(3L)BS123 or dPatj^Δ7 VK20-dPatj-L]/TM3, twi>GFP (Xu and Rubin, 1993; Chou and Perrimon, 1996). To generate follicular clones, young females of hs-FLP/+; dPatj^Δ7 FRT-2A/ubi-GFP FRT-2A were heat shocked for 1 hour at 38°C and then aged for 2-4 days prior to ovary dissection. To generate larval imaginal disc clones, embryos from the hs-FLP; ubi-GFP FRT-2A × dPatj^Δ7 FRT-2A/TM3, twi>GFPr crosses were collected for 24 hours, heat shocked for 1 hour at 38°C 1 day later, and maintained at 25°C until the third instar larvae emerged. To generate dPatj photoreceptor clones, white pupae (less than 2-3 hours into pupation) were collected from the ey-FLP; dPatj^Δ7 FRT-2A/TM3 twi>GFP × ubi-GFP FRT-2A cross and aged at 18°C until the desired pupal stages (Hong et al., 2003). Pupae were genotyped by the absence of twi>GFP.

Immunostaining of embryos, larval imaginal discs and adult ovaries (Tanentzapf et al., 2000; Hong et al., 2001; Huang et al., 2009) and pupal retina dissection and immunostaining (Hong et al., 2003) were performed as previously described. Primary antibodies were: mouse anti-dPatj (full-length) and rabbit anti-dPatj-N (the first 517 amino acids) (Huang et al., 2009) 1:500; rabbit anti-GFP (Huang et al., 2009) 1:1500; guinea pig anti-Baz (Huang et al., 2009) 1:500; rabbit and rat anti-Par-6 (Huang et al., 2009) 1:500; rabbit anti-Sdt (Hong et al., 2001) 1:100-200; mouse anti-Crb (Cq4; DSHB) 1:10-100; mouse anti-Arm (N2 7A1; DSHB) 1:50-100; rat anti-DE-Cad (DCAD2; DSHB) 1:100; mouse anti-Dlg (4F3; DSHB) 1:50; and rabbit anti-aPKC (Santa Cruz) 1:1000. Secondary antibodies: Cy2-, Cy3- or Cy5-conjugated goat anti-rabbit IgG, anti-rat IgG, anti-mouse IgG and anti-guinea pig IgG (Jackson ImmunoResearch), all at 1:400. Images were taken on an Olympus FV1000 confocal microscope (Center for Biologic Imaging, University of Pittsburgh Medical School) and processed in Adobe Photoshop for compositions.

P{GSV2}GS50262 is a viable P-element insertion located 303 bp upstream of the first ATG of the dPatj coding sequence (Fig. 1A). By mobilizing GS50262 with transposase (Hummel and Klämbt, 2008), we recovered several imprecise excision mutants that failed to complement Df(3L)BSC123, which is a 11,312 bp deficiency encompassing the dPatj locus (Cook et al., 2012) (Fig. 1A-C and supplementary material Table S1). dPatj^Δ7, dPatj^Δ13 and dPatj^Δ14 contained deletions that all began at the GS50262 insertion site and removed the 5′UTR plus the conserved L27 motif required for the binding of Sdt or Pals1 (Roh et al., 2002; Li et al., 2004; Feng et al., 2005) (Fig. 1A-C). The 2,509 bp deletion in dPatj^Δ7 further removed the first three PDZ domains that are required for interaction with Crb (Bhat et al., 1999) and Par-6 (Nam and Choi, 2003), whereas deletions in dPatj^Δ13 and dPatj^Δ14 also removed the first two and the first PDZ domain, respectively (Fig. 1B). Antibodies against the N-terminal and full-length dPatj proteins (Huang et al., 2009) revealed maternal contributions of dPatj in dPatj^Δ7 zygotic mutant embryos (Fig. 2A). However, in dPatj^Δ7 germline clone (dPatj^Δ7GLC) mutant embryos, which removed dPatj maternal contributions, dPatj staining was absent, confirming that dPatj^Δ7 is a genetic null mutant (Fig. 2B-F).

Transheterozygotes among dPatj^Δ7, dPatj^Δ13, dPatj^Δ14 and Df(3L)BSC123 suggested that the zygotic mutants of dPatj were pupal lethal, but dPatj^Δ7, dPatj^Δ13 and dPatj^Δ14 appeared to carry additional background mutations that caused prepupal lethality (supplementary material Table S1). To confirm that the loss of dPatj specifically caused pupal lethality, we developed two independent genomic rescue lines, VK20-dPatj-L and VK20-dPatj::GFP-S, that each only contained dPatj as the single intact gene (Fig. 1B). dPatj^Δ7 VK20-dPatj-L or dPatj^Δ7 VK20-dPatj::GFP-S recombinants fully rescued dPatj^Δ7, dPatj^Δ13 and Df(3L)BSC123 (supplementary material Table S1). dPatj^Δ7 VK20-dPatj-L also fully rescued the lethality of dPatj^Δ7GLC embryos (supplementary material Table S1), indicating that the zygotic expression of dPatj is sufficient for development. It is likely that the early larval lethality observed in the dPatj^synnull mutants (Nam and Choi, 2006) was due to additional background mutations or incomplete rescue. The expression pattern of GFP-tagged dPatj from VK20-dPatj::GFP-S was indistinguishable from the dPatj immunostaining pattern (data not shown).

During Drosophila embryogenesis, cellularization is the earliest event in the formation of the polarized embryonic epithelia, and one striking feature of dPatj is its dynamic association with the leading edge of the invaginating membrane in cellularizing cells (Fig. 2B) (Bhat et al., 1999). It was reported that dPatj RNAi knockdown impairs cellularization and results in shortened epithelial cells (Bhat et al., 1999), suggesting that dPatj is required in membrane invagination and the initiation of apical-basal polarity. Surprisingly, in stage 5 dPatj^Δ7GLC embryos (n>10), we found no discernible defects in cellularizing cells in the absence of dPatj (Fig. 2B). Because dPatj^Δ7GLC embryos completed their development through the embryonic and larval stages (supplementary material Table S1), even if the loss of dPatj caused cellularization defects that were too subtle to be observed by immunostaining, such defects were not detrimental to normal development. Our data confirm that dPatj is dispensable for cellularization and initiating apical-basal polarity.

After cellularization, establishing apical-basal polarity in polarizing embryonic epithelia requires the Crb-Sdt complex (Tepass and Knust, 1993; Bachmann et al., 2001; Hong et al., 2001). Again, RNAi knockdown of dPatj severely disrupted the apical localization of Crb and apical-basal polarization (Bhat et al., 1999). Similarly, in polarizing MDCK cells, RNAi knockdown of Patj caused the loss of Pals1 from tight junctions and delayed tight junction formation (Straight et al., 2004; Shin et al., 2005). These data support an essential role of dPatj in establishing apical-basal polarity. Nonetheless, we found that in stage 10 and 11 dPatj^Δ7GLC embryos, both Crb and Sdt showed normal subcellular localization (Fig. 2C,D). Crb expression was slightly reduced in dPatj^Δ7GLC embryos, but the overall pattern was clearly undisturbed (Fig. 2C). We further examined the subcellular localization of several representative polarity proteins in dPatj^Δ7GLC embryos, including: DE-Cadherin (DE-Cad; Shotgun – FlyBase), which marks the adherens junction; Bazooka (Baz, or Par-3), which is required for the early stage of apical-basal polarization; Par-6 and atypical PKC (aPKC), which form an evolutionarily conserved apical complex that regulates multiple polarity proteins through aPKC-mediated phosphorylation (Suzuki et al., 2001; Betschinger et al., 2003; Plant et al., 2003; Sotillos et al., 2004; Krahn et al., 2010; Morais-de-Sá et al., 2010); and Discs large (Dlg, or Dlg1), which is required for specifying the basolateral membrane domain and junctions (e.g. septate junctions). As shown in Fig. 2E-H, in dPatj^Δ7GLC embryos none of these polarity proteins showed any discernible changes in their expression level or subcellular localization. We conclude that Drosophila Patj is not essential for initiating and establishing apical-basal polarity.

The reduction of Crb expression in the absence of dPatj has also been reported previously in polarized epithelial cells, such as the follicular cells of ovaries (Tanentzapf et al., 2000). Because such studies also used a deficiency-based dPatj^MY10 allele with multiple gene deletions (Bhat et al., 1999), we generated dPatj^Δ7 clones in larval imaginal discs and adult follicular cells to investigate their phenotypes in polarized epithelial cells (Fig. 3A-E). The dPatj^Δ7 mutant clones in the imaginal epithelia showed mild reductions in the expression of Crb but not Baz (Fig. 3A), whereas clones in the follicular cells showed much stronger reductions in Crb expression (Fig. 3B). Therefore, dPatj appears to be specifically required to maintain high levels of Crb expression in polarized epithelial cells. Consistent with the observations of Tanentzapf et al. (Tanentzapf et al., 2000), reduced Crb expression in dPatj mutant cells did not disrupt their apical-basal polarity: other polarity markers, such as Baz, DE-Cad, Par-6 and aPKC, all appeared to be unaffected (Fig. 3A-E).

For cells that must undergo dramatic remodeling of their apical-basal polarity, the supporting role of dPatj in maintaining the expression and subcellular localization of Crb might become crucial. In Drosophila, the highly polarized subcellular structures in the photoreceptor soma are generated through a complex morphogenetic process requiring many polarity proteins, including Baz, Crb, Sdt, Par-6 and aPKC (Izaddoost et al., 2002; Pellikka et al., 2002; Hong et al., 2003). It has been reported previously that in dPatj^synnull photoreceptors, the apparent loss of dPatj causes the disruption of apical-basal polarity in the soma (Nam and Choi, 2003; Nam and Choi, 2006). To conclusively show that such phenotypes are indeed specific to dPatj, we generated dPatj^Δ7 mutant photoreceptor clones and examined their apical-basal polarity in early pupal retina at ∼37% pupal development (pd) (Fig. 3F-I). The dPatj^Δ7 photoreceptors showed no dPatj expression (Fig. 3G) but did exhibit frequent mislocalization of major apical polarity proteins. Armadillo (Arm; Drosophila β-Catenin homolog) and Baz, which marks adherens junctions, were often disrupted in the dPatj^Δ7 photoreceptors (Fig. 3G,H), in addition to the basolateral mislocalization of Par-6 and Crb (Fig. 3I). Therefore, although it is dispensable in normal epithelia, dPatj might be crucial to maintain apical-basal polarity under more complex and stringent developmental contexts.

We did not observe any obvious morphological abnormalities, such as the loss of imaginal discs, in dPatj^Δ7/Df(3L)BSC123 (hereafter dPatj^Δ7/Df) larvae (data not shown). However, within several hours of pupal formation, the dPatj^Δ7/Df pupae exhibited deformities in shape, with a flattened anterior end and rounded posterior (Fig. 4A). At 16% pd, the head region was underdeveloped, although the imaginal discs appeared to have successfully everted (Fig. 4B), suggesting that even without dPatj the imaginal disc cells could still modulate their cell junctions to accomplish disc eversion (Pastor-Pareja et al., 2004). Older dPatj^Δ7/Df pupae failed to develop head and thoracic regions and growth of the everted imaginal discs stalled (Fig. 4C). The pupal defects and lethality were completely rescued in VK20-dPatj-L or dPatj^Δ7 VK20-dPatj::GFP-S (Fig. 4D and supplementary material Table S1). The mild reduction in Crb expression observed in the dPatj mutants was unlikely to be responsible for the pupal lethality because the additional removal of one copy of crb in dPatj^Δ7/Df(3L)BSC123 crb^ko or dPatj^Δ7 crb^ko/dPatj^Δ13 did not enhance pupal lethality (supplementary material Table S1). The viability of dPatj^synhypo (Pielage et al., 2003; Nam and Choi, 2006) suggests that the N-terminal 260 amino acids of dPatj might be sufficient for pupal development, but at present the specific function of dPatj in pupal development remains to be identified.

In summary, our molecularly defined dPatj null alleles conclusively show that dPatj is not essential for establishing or maintaining apical-basal polarity. The mild reduction in Crb expression in the absence of dPatj suggests that dPatj plays a supporting role in maintaining the Crb-Sdt complex in polarized epithelial cells. Such a scenario exhibits similarities with the situation for p120catenin (Adherens junction protein p120 – FlyBase) mutants: despite RNAi results suggesting that p120catenin is essential for cell adhesion, genetic mutants revealed that it only plays a supporting and non-essential role in cell adhesion in Drosophila (Myster et al., 2003). It remains possible, however, that mammalian Patj might have evolved increasingly important roles in apical-basal polarity, which might explain the differences between our dPatj null mutant phenotypes and those of mammalian cells with RNAi knocked-down Patj levels (Straight et al., 2004; Wang et al., 2004; Shin et al., 2005).

We thank the Bloomington Drosophila Stock Center and Kyoto Drosophila Genetic Resources Center for fly stocks; the Developmental Studies Hybridoma Bank (DSHB) for antibodies; and the University of Pittsburgh Medical School Center for Biologic Imaging for providing access to confocal microscopes.