The Drosophila Importin-α3 Is Required for Nuclear Import of Notch In Vivo and It Displays Synergistic Effects with Notch Receptor on Cell Proliferation

The Notch signaling pathway controls diverse cell-fate specification events throughout development. The versatility of this pathway to influence different aspects of development comes from its multiple levels of regulation. Upon ligand-induced Notch activation, the Notch intracellular domain (Notch-ICD) is released from the membrane and translocates to the nucleus, where it transduces Notch signals by regulating the transcription of downstream target genes. But the exact mechanism of translocation of Notch-ICD into the nucleus is not clear. Here, we implicate Importin-α3 (also known as karyopherin-α3) in the nuclear translocation of Notch-ICD in Drosophila. Our present analyses reveal that Importin-α3 can directly bind to Notch-ICD and loss of Importin-α3 function results in cytoplasmic accumulation of the Notch receptor. Using MARCM (Mosaic Analysis with a Repressible Cell Marker) technique, we demonstrate that Importin-α3 is required for nuclear localization of Notch-ICD. These results reveal that the nuclear transport of Notch-ICD is mediated by the canonical Importin-α3/Importin-β transport pathway. In addition, co-expression of both Notch-ICD and Importin-α3 displays synergistic effects on cell proliferation. Taken together, our results suggest that Importin-α3 mediated nuclear import of Notch-ICD may play important role in regulation of Notch signaling.


Introduction
The Notch pathway is an evolutionarily conserved signaling system which has been shown to play major role in cell fate determination, differentiation, proliferation and apoptotic events as well as self-renewal processes of different tissues [1][2][3][4][5][6]. The same pathway can be deployed in numerous cellular contexts to play varied and critical roles for the development of an organism. The Notch receptor is synthesized as a single polypeptide precursor, which during maturation in the trans-Golgi network is first cleaved by a furin protease into a N-terminal extracellular subunit and a C-terminal transmembrane intracellular subunit [7]. This heterodimeric receptor is then transferred to the cell membrane where it interacts with its ligands, Delta and Serrate in Drosophila (Delta and Jagged in vertebrates). Binding of ligands to extracellular domain leads to a metalloprotease-dependent cleavage in the extracellular portion of transmembrane intracellular fragment [8], which is followed by an intramembrane cleavage by the Presenilin-dependent gamma-secretase activity resulting the release of Notch intracellular domain [9][10][11][12]. The Notch intracellular domain is translocated to the nucleus where it binds to and activates a transcription factor, Suppressor of Hairless in Drosophila (CBF1 in vertebrates) [13][14][15]. This complex also recruits Mastermind [16] and other transcriptional coactivators leading to activation of Notch target genes such as the Enhancer of Split [E(spl)] complex genes [17].
In an effort to identify novel components involved in Notch signaling and its regulation, a yeast two-hybrid screen was carried out using the portion of intracellular domain of Notch receptor as bait and we identified Drosophila Importin-a3 as binding partner of Notch. Drosophila Importin-a3 protein is known to play major role in nuclear trafficking of different Nuclear Localization Signal (NLS) containing proteins such as Germ Cell-less [18], the large subunit of DNA polymerase a [19], heat shock transcription factor (dHSF) [20], Daxx [21], Naked cuticle (Nkd) [22] etc. Since nuclear transport protein Importin-a3 directly binds to portion of Notch intracellular domain which contains NLS [7], we were prompted to examine if Notch intracellular domain translocates to nucleus using the canonical nuclear transport machinery.
In human, there are seven Importin a family members, whereas Drosophila has Importin-a1, Importin-a2 and Importin-a3 coding genes. Among them only Importin-a3 binds to NLS-containing proteins via its Armadillo (Arm) motifs and to Importin-b via its Nterminal Importin-b binding domain (IBB) [23]. Importin-b interacts with nuclear pore complex (NPC) and targets NLS protein/Importin-a3/Importin-b trimeric complex to the nuclear pore for translocation into the nucleus. RanGTP concentration in the nucleus is high and it interacts with Importin-b, resulting in disassembly of the import complex releasing both Importin-a3 and the NLS cargo into the nucleus. Subsequently, Importin-a3 is free and forms a trimeric complex with RanGTP and CAS (Cellular apoptosis susceptibility) proteins. This trimeric complex is then exported to the cytoplasm, recycling Importin-a3 for another round of import. Importin-b is also recycled back to cytoplasm by binding to RanGTP in the nucleus [24].
Our molecular and genetic analyses presented here clearly demonstrate that Importin-a3 plays important role in nuclear transport of Notch-ICD and co-expression of Importin-a3, together with Notch-ICD, displays synergistic effects on signaling activity of the Notch receptor.

Results and Discussion
Importin-a3 is an Interacting Partner of Notch In a yeast two-hybrid screen, we identified Importin-a3 as an interacting partner of Notch. In the same screen, multiple positive clones of a well established binding partner of Notch-ICD, Suppressor of Hairless, were also identified, which validates our approach. The yeast two-hybrid screen of 6610 6 cDNAs from a Drosophila 0-24 h embryonic library was carried out using amino terminus of Notch intracellular domain (amino acids 1765-1895) as bait. Twenty one positive clones (His + ) were isolated and found to encode overlapping imp-a3 cDNAs. Sequence analysis of these clones revealed that the carboxy-terminal part of Importin-a3 (amino acids 240-502) is necessary and sufficient for binding Notch ( Figure 1A). This particular domain of Importin-a3 was shown earlier to interact with NLS containing proteins [25].
GST-pull down experiments using purified GST-Importin-a3 confirmed the interaction between Notch and Importin-a3. Different GST-Importin-a3 fusion proteins (full-length 1-514, amino terminus 1-224 and carboxy terminus 225-514) were expressed in bacteria and fusion products were isolated on Glutathione Sepharose beads. After extensive washing, the beads were incubated with extracts from third instar larval salivary glands in which Notch-ICD was overexpressed using ey-GAL4 driver. Deletion analysis of Importin-a3 protein demonstrated that carboxy-terminus portion of Importin-a3 is required for binding to Notch-ICD ( Figure 1B). Furthermore, co-immunoprecipitation experiment was carried out in which Notch-ICD was immunoprecipitated with HA-Importin-a3 from larval salivary glands when both proteins were co-expressed ( Figure 1C). Taken together, these results suggest that the Importin-a3 directly interacts with Notch and that the Importin-a3 binds with Notch through its C-terminus that is known to bind with NLS-containing proteins. To further analyze interactions between Importin-a3 and Notch, we investigated the subcellular localization of these proteins when UAS-HA-imp-a3 and UAS-Notch-ICD were co-expressed in larval salivary glands and eye imaginal discs using ey-GAL4 driver. Immunocytochemical analysis revealed that Importin-a3 and Notch-ICD indeed co-localized in cell nuclei ( Figure 1D1-1F4).

Genetic Interactions between imp-a3 and Notch Pathway Components
To address functional implications of the physical interaction between the Importin-a3 and Notch proteins, we investigated whether mutations in imp-a3 and Notch or other components involved in Notch signaling pathway display genetic interactions in transheterozygous combinations. We used two independent lossof-function imp-a3 alleles: imp a3 D93 and imp a3 D165 and one hypomorphic allele, imp a3 1(R59) [26]. A transheterozygous combination of Notch null allele, N 1 or a hemizygous Notch hypomorphic allele, N nd-3 and any one of the three imp-a3 alleles resulted in enhancement of wing nicking phenotype, indicating further reduction of the Notch function ( Figure 2A1-2B4). On the contrary when we used gain-of-function Notch allele, the Abruptex mutation (N Ax-16172 ) which displays a shortened longitudinal vein V (L5), we noticed an extension of L5 vein upto the wing margin in transheterozygous combination with imp-a3 mutations ( Figure 2C1-2C4). The wing vein thickening phenotype of Delta (Dl) and the wing notching phenotype of the dominant negative mutation of Serrate (Ser Bd-G ) were also enhanced by reducing the dose of imp-a3 ( Figure 2D1-2F4). A transheterozygous combination of dx, which is a cytoplasmic modulator of Notch activity, and imp-a3 alleles resulted in normal wings (data not shown) whereas dx in hemizygous combination with imp-a3 alleles showed wing phenotype that consist of extra vein material at the distal ends of wing veins ( Figure 2G1-2G4). These observations confirm a functional relationship between imp-a3 and Notch consistent with their molecular interactions.

imp-a3 Mutant Cells have Elevated Notch Protein Levels
To further analyze the molecular implications of Importin-a3 and Notch interaction in vivo, we examined the effects of imp-a3 loss-of-function on the endogenous Notch protein. We generated imp-a3 loss-of-function clones in two different larval tissues, eyeantennal imaginal discs and salivary glands, using imp a3 D93 null mutant [26] and the FLP/FRT system [27]. The FLP activity that is under the control of eyeless promoter (ey-FLP) was used to induce somatic recombination events in eye discs, while hsFLP was used to generate somatic clones in salivary glands. Salivary glands begin to develop at 4.5 hours of development and complete by 10 hours of embryonic development. Thus, to generate somatic clones in salivary glands, 4-8 hours old embryos were subjected to a single heat shock (37uC for 45 min). The imp-a3 mutant clones in either tissues were identified by the absence of GFP expression. Notch protein levels in the cytoplasm were strongly elevated in imp-a3 mutant cells compared with the surrounding wild-type cells in both salivary glands and eye-antennal discs ( Figure 3A1-3C4). These elevated Notch protein levels in imp-a3 mutant cells may be due to the lack of nuclear transport of Notch-ICD in the absence of Importin-a3.

Gain-of-function Effect of imp-a3 on Notch Protein
In parallel with the imp-a3 loss-of-function analysis, we also determined the gain-of-function effect of imp-a3 by monitoring the effects of its ectopic expression on Notch localization. We examined the distribution of endogenous Notch protein in wing imaginal discs in which imp-a3 expression was driven by en-GAL4 driver. Expression of UAS-HA-imp-a3 with en-GAL4 driver resulted in cytoplasmic aggregates of Notch protein in posterior compartment cells of wing discs ( Figure 3D1-3D3, also see Figure S1F1-S1F3). Since en-GAL4 is a posterior compartment specific driver, cells from anterior compartment serve as an internal control ( Figure 3D1-3D3, and S1E1-S1E3). This effect is specific for imp-a3 since when the other two importins, imp-a1 and imp-a2, were overexpressed in wing discs using en-GAL4 driver, cytoplasmic aggregates of Notch were never observed (see Figure S1A1-S1D3). Thus, this effect on Notch accumulation is a specific consequence of imp-a3 expression. Our analysis does not exclude the possibility that these aggregates may be some kind of vesicular structures which remains to be determined.

Importin-a3 is Required for Notch Nuclear Localization
Endogenous Notch-ICD is not easily detectable in nucleus by immunostaining using antibody specific for intracellular domain of Notch, since very little amount of the cleaved product is translocated to nucleus for carrying out its downstream function [28,29]. Recently it has been reported that endogenous Notch-ICD is detectable in the nucleus of pIIa cells derived by  19,20]. A region of Importin-a3 (amino acids 240-502) that was sufficient for binding to Notch, based on yeast two-hybrid analysis, is shown below the full-length protein. (B) GST-pulldown assay was performed with lysate of salivary glands in which Notch-ICD was overexpressed using salivary gland specific GAL4 driver (sgs-GAL4) and purified recombinant GST-Importin-a3 full-length (amino acids 1-514), amino-terminal (amino acids 1-224), carboxy-terminal (amino acids 225-514) and other controls as indicated. GST pulled down proteins were analyzed by western blotting with anti-Notch (C17.9C6) antibodies. GST-Importin-a3 fulllength and GST-Importin-a3 carboxy-terminus pulled down Notch-ICD. (C) Co-immunoprecipitation of HA-Importin-a3 and Notch-ICD. HA-Importin-a3 and Notch-ICD were co-expressed in larval salivary glands and immunoprecipitated with anti-HA agarose.  Importin-a3 Mediates Nuclear Import of Notch PLOS ONE | www.plosone.org in transheterozygous combination with different alleles of imp-a3 (F2-F4). Wing phenotype of dx hemizygotes (G1) with imp-a3 mutants showed enhanced phenotype that consist of extra vein material at the distal ends of wing veins (G2-G4). n = 100 wings for each genotype. The expressivity and penetrance of the phenotype for each genotype were 100% except for N 1 heterozygotes (A1) and N 1 heterozygotes with different alleles of imp-a3 (A2-A4) in which penetrance of the phenotype was 45%. Scale bar, 250 mm. doi:10.1371/journal.pone.0068247.g002 Importin-a3 Mediates Nuclear Import of Notch PLOS ONE | www.plosone.org asymmetric division of Sensory organ precursor cells (SOPs) [30]. It was not straightforward to test whether transport of endogenous Notch-ICD was inhibited in the absence of Importin-a3. When Notch-ICD is overexpressed, it is readily detectable in the nucleus [31][32][33] (also see Figure 1D1-1F4). To determine the role of Importin-a3 in nuclear localization of Notch we overexpressed Notch-ICD in imp-a3 mutant clonal cells using the MARCM technique [34]. Although Notch-ICD was under the control of UAS and the Tub-GAL4 driver was used to globally drive Notch-ICD expression, the presence of GAL80 inhibited GAL4-induced expression of Notch-ICD in all cells except in those cells in which GAL80 was eliminated due to FLP-FRT mediated somatic recombination events. At the same time imp-a3 gene function was also eliminated in same cells. As a result, these imp-a3 mutant clonal cells, which were also marked with GFP, overexpressed Notch-ICD under Tub-GAL4 driver. In parallel, MARCM analysis of wild-type clones without imp-a3 mutation was also carried out and subcellular localization of Notch protein was determined. In GFP-marked imp-a3 mutant clonal cells, Notch-ICD was completely excluded from the nucleus ( Figure 4A1-4B3), whereas Notch-ICD was readily detectable in the nucleus of wild-type clonal cells without imp-a3 mutation ( Figure 4C1-4D3).

Importin-a3 Displays Synergistic Effects with Notch Signals on Cell Proliferation
Earlier studies have shown that overexpression of activated form of Notch (Notch-ICD) in eye discs results in roughening of the eye with fused or missing ommatidia and bristle irregularities [33]. We have also observed that overexpression of Notch-ICD using ey-GAL4 driver exhibits rough eye phenotype with fused or abnormally sized ommatidia together with extra and missing bristles ( Figure 5B5). Immunostaining of Elav in eye-discs revealed that overexpression of Notch-ICD results in fusion of ommatidia and defective ommatidial spacing ( Figure 5B1-5B4). Interestingly, co-expression of Notch-ICD and Importin-a3 using the same ey-GAL4 driver results in a considerable enhancement of the adult eye phenotype with more frequent fusion of ommatidia and appearance of abnormally sized ommatidia with extra bristles ( Figure 5C5). Similarly, Elav staining of larval eye discs in which both Notch-ICD and Importin-a3 were overexpressed also showed enhanced defects in ommatidial spacing and misrotated ommatidia ( Figure 5C1-5C4). We have also noticed that larval eye discs as well as wing discs, in which both Notch-ICD and Importin-a3 were overexpressed, are considerably larger than only Notch-ICD overexpressing eye or wing discs and these discs are thicker, wrinkled and highly distorted as compared to only Notch-ICD overexpressing discs ( Figure 5A1, 5B1, 5C1, and S2). These results reveal a synergistic effect between Importin-a3 and Notch-ICD on cell proliferation in eye and wing discs.
We have observed that 58% ey-GAL4 driven Notch-ICD overexpressing pupae (n = 450) emerged as adult flies and this was reduced to 23% in which both Notch-ICD and Importin-a3 were overexpressed (n = 450) using the same ey-GAL4 driver ( Figure 5D). Moreover, flies that survived in which both Notch-ICD and Importin-a3 were overexpressed have significantly reduced life span as compare to only Notch-ICD overexpressing flies ( Figure 5E).
It has previously been shown that mammalian Notch1-ICD must accumulate in the nucleus to induce neoplastic transformation of baby rat kidney cells (RKE) [35]. Another report has described that downregulation of importins a3, a5, a7 and Importin b strongly inhibits HeLa cell proliferation and on the basis of these findings it was proposed that import pathways of various substrates that are essential for cell proliferation were blocked, resulting in proliferation inhibition [36]. Interestingly, our data also suggest that overexpression of Importin-a3 along with Notch-ICD displays a synergistic effect between Importin-a3 and Notch activation on cell proliferation in the eye and wing imaginal discs which is likely to be caused by stronger activation of Notch signals due to the greater import of Notch-ICD into the nucleus by overexpressed Importin-a3.
Our results establish that the nuclear transport of Notch-ICD is mediated by the canonical Importin-a3/Importin-b transport pathway and co-expression of both Notch-ICD and Importin-a3 displays synergistic effects on cell proliferation. Earlier a genomewide loss-of-function analysis by transgenic RNAi in Drosophila has been carried out to study the Notch signaling pathway during external sensory organ development and many novel components including nuclear import pathway components and nuclear pore components have been identified as Notch regulators [37]. Although it has been documented through cell culture based experiments that Importins a3, a4 and a7 mediate nuclear import of Notch-ICD in mouse myoblast and human HeLa cells [38], our present report is the first in vivo study showing role of Importin-a3 in nuclear import of Notch-ICD in Drosophila and its synergistic effects with Notch signals on cell proliferation.
Notch signaling is known to affect a broad spectrum of cell-fate decisions throughout development. To allow the Notch signal to be deployed in numerous cellular contexts, many different mechanisms have evolved to regulate the level, duration, and spatial distribution of Notch activity [2,4,39]. It is known that the transport of proteins and RNAs in and out of nucleus plays important role in the regulation of gene expression during every stage of development and tissue differentiation [40]. Nuclear import of Notch-ICD may play important role in regulation of Notch signaling activity and it remains possible that this mode of regulation may be involved in many other signaling pathways.

Yeast Two-hybrid
A 393 bp Drosophila Notch cDNA (accession number M11664) fragment which encodes amino acids 1765-1895 containing NLS was amplified by polymerase chain reaction (PCR) and cloned in frame with the sequence encoding the LexA DNA-binding domain of bait vector. This construct was used as bait to screen oligo(dT)primed D. melanogaster 0-24 h embryo cDNA libraries cloned in pGAD prey vectors containing GAL4 activation domains. A yeast two-hybrid screen was carried out as described previously [39]. Finally, all positive pGAD plasmids from His+ colonies were isolated and sequenced to identify interactors.
Intact reading frames for all constructs were verified by DNA sequence analysis. GST and GST fusion proteins were expressed in E. coli BL21 cells at 37uC with 2 mM isopropyl-1-thio-b-Dgalactopyranoside (IPTG) induction. Bacteria were lysed in solution of Cell Lytic TM express tablet (Sigma) with complete protease inhibitor (Roche). Glutathione Sepharose (GE Healthcare Bio-Sciences) beads were washed in cold phosphate buffered saline (PBS), 3 times 30 minutes each. 50% slurry was made in PBS.
Intracellular domain of Notch was overexpressed in salivary glands by salivary gland specific GAL4 driver (sgs-GAL4) and third instar larval salivary glands were dissected and washed in PBS, then lysed in lysis buffer (50 mM Tris pH8.0, 0.1% TritonX-100, 10% Glycerol, 200 mg/ml lysozyme and 1 mM PMSF) for 3 hrs at 4uC.The supernatant was collected after centrifugation for 20 min at 12,000 rpm. Glutathione Sepharose beads alone or incubated with GST fusion proteins mixed with salivary gland lysate and rotated for 3 hrs at 4uC followed by three times washing with PBST (1X PBS, 1% Triton-X-100), 15 min each. Beads were boiled in 2X Laemmli buffer for 5 min and samples were loaded in 12% denaturing gel with Spectra multicolor broad range protein ladder used as a marker (Fermentas). Proteins were separated on nonreducing SDS-PAGE (without b-mercaptoethanol) and transferred onto PVDF membrane (Bio-Rad). Blot was probed with mouse anti-Notch C17.9C6 in 1:3000 dilution (Developmental Studies Hybridoma Bank) and secondary antibody goat anti-mouse IgG-AP conjugate in 1:2000 dilution (Molecular Probes) in blocking solution (4% skimmed milk in TBST-50 mM Tris, pH 7.5, 150 mM Nacl, 0.1% Tween-20). Then after washing in TBST thrice, colour was detected by Sigma FAST TM BCIP/NBT (Sigma).
Immunoprecipitation from larval salivary glands was carried out as described previously [39]. HA-Importin-a3 and Notch-ICD proteins were expressed in larval salivary glands under the control

MARCM Clonal Analysis
The MARCM system was used to generate GFP-marked imp-a3 mutant clones overexpressing Notch-ICD. First UAS-Notch-ICD; FRT82B imp a3 D93 /TM6B stock was generated by appropriate genetic crosses. To generate GFP-marked clones, females of UAS-Notch-ICD; FRT82B imp a3 D93 /TM6B were crossed to male y w hsFLP Tub-GAL4 UAS-GFP; FRT82B Tub-GAL80 flies. In parallel, a control experiment was carried out in which females of UAS-Notch-ICD; FRT82B/TM6B were crossed to male flies of y w hsFLP Tub-GAL4 UAS-GFP; FRT82B Tub-GAL80 genotype. Heat shock was given at 37uC for 45 min at 24 hrs AEL and third instar female larvae were analyzed for GFP-marked clones.

Immunocytochemistry and Confocal Microscopy
Drosophila third instar larval imaginal discs, brain and salivary glands were dissected in cold PBS. Immunostaining of these tissues was performed as described previously [39]. Primary antibodies, mouse anti-Notch (C17.9C6) at 1:300 dilution, Rat anti-Elav at 1:200 (Developmental Studies Hybridoma Bank), Rabbit anti-HA at 1:100 (Sigma) and secondary antibodies, goat anti-mouse antibodies Alexa Flour 555 at a 1:200 dilution (Invitrogen), goatanti-rat antibodies conjugated to FITC at a 1:200 dilution (Sigma), goat-anti-rabbit antibodies conjugated to FITC at a 1:200 dilution (Jackson ImmunoResearch Laboratories, West Grove, PA) were used for immunostaining. After washing secondary antibodies, DAPI (49, 6-diamidino-2-phenylindole dihydrochloride) was used to detect the nucleus and washed twice in PBS for 10 min each. Then larval tissues were mounted in FluoroGuard Antifade Reagent (Bio-Rad) and images were captured with a Zeiss LSM510 Meta laser confocal microscope. Figure S1 Overexpression of Importin-a3 specifically results in the formation of cytoplasmic aggregates of endogenous Notch protein. (A1-F3) UAS-imp a1, UAS-imp a2, and UAS-imp a3 transgenes were expressed under the control of en-GAL4 driver, which is expressed in posterior compartment cells of wing discs. Localization of endogenous Notch protein in anterior compartment (A1-A3) and posterior compartment (B1-B3) of a wing disc in which UAS-imp a1 expression was driven by en-GAL4. Similarly, localization of Notch protein in anterior compartment (C1-C3) and posterior compartment (D1-D3) of a wing disc in which UAS-imp a2 was overexpressed and distribution of Notch protein in anterior compartment (E1-E3) and posterior compartment (F1-F3) of a wing disc in which UAS-imp a3 was overexpressed. Note that there is no difference in Notch localization in anterior and posterior compartment in case of UAS-imp a1 and UAS-imp a2 overexpression (A1-D3) while presence of more number of Notch aggregates in cytoplasm of posterior compartment cells (F1-F3) compare to anterior compartment cells (E1-E3) in UAS-imp a3 overexpressed wing disc. Images in A3, B3, C3, D3, E3, and F3 are merges of those in A1 and A2, B1 and B2, C1 and C2, D1 and D2, E1 and E2, and F1 and F2, respectively. Insets in A3, B3, C3, D3, E3, and F3 show higher magnification images of a single cell in the corresponding disc. Scale bar, 10 mm.