Association of RanGAP to Nuclear Pore Complex Component, RanBP2/Nup358, is Required for Pupal Development in Drosophila

Summary:

Ran’s GTPase activating protein (RanGAP) is tethered to the nuclear envelope (NE) in multicellular organisms. We investigated the consequences of RanGAP localization in human tissue culture cells and Drosophila. In tissue culture cells, disruption of RanGAP1 NE localization surprisingly has neither obvious impacts on viability nor nucleocytoplasmic transport of a model substrate. In Drosophila, we identified a region within nucleoporin dmRanBP2 required for direct tethering of dmRanGAP to the NE. A dmRanBP2 mutant lacking this region shows no apparent growth defects during larval stages, but arrests at the early pupal stage. A direct fusion of dmRanGAP to the dmRanBP2 mutant rescues this arrest, indicating that dmRanGAP recruitment to dmRanBP2 per se is necessary for the pupal ecdysis sequence. Our results indicate that while the NE localization of RanGAP is widely conserved in multicellular organisms, the targeting mechanisms are not. Further, we find a requirement for this localization during pupal development.

Graphical Abstract

Ran’s GTPase activating protein (RanGAP) is tethered to the nuclear envelope in multicellular organisms. Chen et al. find that although this localization is dispensable for many aspects of cellular function in mammals, it is essential for pupation in Drosophila, suggesting RanGAP localization at the nuclear envelope contributes to developmental processes.

Introduction:

The Ran GTPase is critical for nuclear-cytoplasmic transport, nuclear envelope (NE) assembly, and mitotic chromosome segregation (Macara, 2001, Hetzer et al., 2002, Gruss and Vernos, 2004, Harel and Forbes, 2004). These processes are driven by gradients of GTP-bound Ran (Ran-GTP) and GDP-bound Ran (Ran-GDP). During interphase, Ran’s chromatin-bound nucleotide exchange factor, RCC1, maintains high levels of Ran-GTP within nuclei. Conversely, Ran’s GTPase activating protein, RanGAP, resides in the cytoplasm and promotes Ran-GTP hydrolysis, so that Ran-GDP is predominant in the cytosol. The RanGTP/RanGDP gradient is essential for the transport of cargoes across the NE: nuclear import complexes disassemble in the presence of RanGTP, while nuclear export complexes disassemble after RanGTP hydrolysis (Gorlich and Kutay, 1999).

Fungal RanGAPs consist largely of a conserved catalytic domain (Figure 1A), and they are diffusely distributed throughout the cytosol (Matynia et al., 1996, Hopper et al., 1990). By contrast, multicellular organisms, including humans, plants, frogs and flies, localize RanGAP to the NE through additional targeting domains, which are structurally divergent (Mahajan et al., 1997, Matunis et al., 1998, Kusano et al., 2001, Rose and Meier, 2001, Jeong et al., 2005). For example, vertebrate RanGAP1 is SUMOylated in its C-terminus, facilitating its association to the nucleoporin RanBP2/Nup358 at the nuclear pore complex (NPC) via a remarkably complex set of interactions, while Arabidopsis RanGAP (AtRanGAP1) has an N-terminal WPP domain that directly tethers it to the NE (Matunis et al., 1998, Mahajan et al., 1997, Xu et al., 2007). These distinct targeting mechanisms suggest convergent evolution of RanGAP localization at the NE in multicellular organisms, but the function of NE targeting remains to be elucidated.

Figure 1. Displacing RanGAP1 from nuclear envelope does not cause apparent defects in human tissue culture cells.

A. Schematic comparison of the domain architecture of RanGAPs from S. pombe (SpRna1p), Drosophila melanogaster (DmRanGAP), Homo sapiens (HsRanGAP1), and Arabidopsis thaliana (AtRanGAP1). Catalytic domain, green; acidic domain, pink; and nuclear envelope targeting domain, yellow.

B. Western blot analysis of whole cell lysate from DLD1 parental, RanGAP1^WT-Flag and RanGAP1^KR-Flag cells using RanGAP1 antibody. A SUMOylated form and an unmodified form of RanGAP1 were shown in DLD1 parental and RanGAP1^WT-Flag cells, while only unmodified RanGAP1 can be observed in RanGAP1^KR-Flag cells.

C. Live imaging of DLD1 cells harboring mCherry tagged RanGAP1^WT and RanGAP1^KR. RanGAP1^WT-mCherry localizes primarily at the nuclear envelop whereas RanGAP1^KR-mCherry is dispersed in cytoplasm.

D. Cell growth analysis of HCT116 with RanGAP1^WT-Flag and RanGAP1^KR-Flag. (N=3, p= 0.7769).

E. Analysis of time from nuclear envelope break down (NEBD) to anaphase. No statistically significant difference in mitotic timing was observed between HCT116 cells harboring RanGAP1^WT-Flag versus RanGAP1^KR-Flag (N=34 and N=48, respectively, p=0.0521).

F. Protein transport assay in DLD1 cells harboring RanGAP1^WT-Flag versus RanGAP1^KR-Flag using a 46 kDa mCherry-tagged model substrate carrying classic NLS and NES domains (N= 4 and 5 for RanGAP1^WT- and RanGAP1^KR- expressing cells, respectively) (Niopek et al., 2016).

G. First order kinetic fitting curves for nuclear export (upper panel) and import (lower panel) with apparent import/export rates (k), derived from (F).

Three hypotheses for the purpose of NE targeting of RanGAP in multicellular organisms have been proposed, which are not mutually exclusive. First, anchorage of RanGAP at the NE might facilitate more efficient nuclear import and export (Mahajan et al., 1997, Kehlenbach et al., 1999). While smaller yeast cells grow well without RanGAP recruitment to the NE, it has been hypothesized that RanGAP targeting becomes important to boost transport efficiency in multicellular organisms whose cells are larger in volume. Second, RanGAP targeting may facilitate some other feature of multicellular eukaryotes that is not shared with fungi, such as open mitosis (Joseph et al., 2002, Joseph et al., 2004). Gradients of Ran-GTP convey important spatial information during open mitosis (Arnaoutov and Dasso, 2003, Arnaoutov and Dasso, 2005), and these gradients are shaped in vertebrate cells by recruitment of the RanGAP/RanBP2 complex to mitotic spindles and kinetochores (Joseph et al., 2002, Joseph et al., 2004). Third, there may be a requirement for RanGAP recruitment to the NE during complex developmental processes such as specialization of tissues and organs in multicellular organisms. We decided to re-evaluate the function of the RanGAP-NE interaction in mammalian tissue culture cells and in Drosophila in order to test these potential roles of RanGAP localization.

To evaluate cellular roles of RanGAP1 targeting, we created a cytosolic version of RanGAP1; Using CRISPR gene editing in human colorectal adenocarcinoma tissue culture cells (DLD1 and HCT116), we substituted lysine-524 with arginine in RanGAP1, preventing its SUMOylation and association to RanBP2 (Mahajan et al., 1997, Matunis et al., 1998). To our surprise, nuclear transport of model protein cargos, cell growth rates and cell viability were not affected, arguing against an essential role of RanGAP1 localization in facilitating interphase or mitotic cellular functions. We thus turned to Drosophila to address the role of RanGAP localization in the context of a multicellular developing organism. To start, we investigated the mechanism of dmRanGAP recruitment to the NE, which had not been previously characterized. As in mammals, we found that dmRanGAP is recruited to the NE through association to dmRanBP2, albeit through direct and SUMO-independent binding to a 23-amino-acid domain within dmRanBP2. To eliminate this interaction, we removed the dmRanGAP binding sequence within dmRanBP2 using CRISPR gene editing. Reminiscent of what we observed in tissue culture cells, uncoupling RanGAP from the NE in Drosophila leads to no apparent cell cycle or growth defects through the end of the 3^rd larval instar. However, these flies arrest early in pupal development prior to eversion of the cephalic complex. This developmental arrest was rescued by a direct fusion of dmRanGAP to the dmRanBP2 mutant, indicating that recruitment of dmRanGAP to dmRanBP2 per se was necessary for the pupal ecdysis sequence during development. Collectively, our results indicate that while the localization of dmRanGAP to the NE is widely conserved in multicellular organisms, the targeting mechanisms are not. Further, while we did not observe strong evidence that RanGAP localization promotes nuclear transport, cell growth or cell cycle progression, we find a previously unreported requirement for this localization in critical pupal developmental.

Results:

RanGAP1 localization at the NE is not essential for nuclear transport or cell growth.

SUMOylation of human RanGAP1 at lysine 524 is required for its association with RanBP2, a large nucleoporin associated to the cytosolic face of the NPC (Mahajan et al., 1997, Matunis et al., 1998) (Figure 1A). To abolish RanGAP1 interaction with RanBP2, we biallelically mutated lysine 524 to arginine within the endogenous RanGAP1 gene locus in two human colorectal adenocarcinoma tissue culture cells (DLD1 or HCT116), simultaneously adding either FLAG- or mCherry-tags to the RanGAP1 protein (Figure 1B, C, Supplemental Figure 1A, B). The resulting mutant protein (RanGAP1^KR) showed only the unSUMOylated form (Figure 1B) and lost NE association (Figure 1C), as predicted. By contrast, similarly tagged wildtype RanGAP1 (RanGAP1^WT) retained SUMOylation and NE targeting (Figure 1B, C). These data are consistent with previous reports that SUMOylation is required for anchorage of RanGAP1 at the NE (Mahajan et al., 1997, Matunis et al., 1998).

Displacement of RanGAP1 from the NPC did not cause significant defects in cell viability or proliferation (Figure 1D, Supplemental Figure 1C). It has previously been shown that the association of RanGAP1 to RanBP2 controls its localization to mitotic spindles during cell division (Joseph et al., 2002). Nevertheless, the time required for RanGAP1^KR- and RanGAP1^WT-expressing cells to progress from nuclear envelope breakdown to anaphase was similar (Figure 1E). Although we cannot rule out that the binding of RanGAP1 to RanBP2 may have some important role(s) during mitosis, our results suggest that RanGAP1-RanBP2 interaction is not essential for mitotic progression. To test the role of RanGAP1 localization in nuclear import and export, we used a 46 kDa model nuclear transport substrate with classical NLS and NES domains (Niopek et al., 2016). We observed that the rates of both import and export were similar in cells expressing RanGAP1^KR or RanGAP1^WT (Figure 1F, G). These data suggest that RanGAP1 localization within the cytoplasm has a marginal impact on nuclear trafficking of proteins containing classical NLS or NES, at least in DLD1 cells, thus arguing against models in which NE localization of RanGAP increases the efficiency of bulk nuclear import and export in large cells.

Taken together, our data did not support a critical role of RanGAP1 targeting for cell growth, mitotic progression or nuclear transport within cultured DLD1 cells. We therefore wondered whether RanGAP localization might be important in the complex developmental pathways of multicellular eukaryotes and decided to test this idea using Drosophila melanogaster as a genetically tractable model organism.

Nuclear envelope anchorage of dmRanGAP is SUMO independent.

Similar to mammalian RanGAP1, the Drosophila RanGAP1 homolog, dmRanGAP, localizes to the nuclear envelope and can be co-precipitated with the Drosophila RanBP2 homolog, dmRanBP2 (Supplemental Figure 2A, B) (Kusano et al., 2001). However, two findings indicated that the mechanism of association between dmRanGAP and dmRanBP2 differs from the SUMO1-dependent mechanism observed in mammalian cells (Mahajan et al., 1997, Matunis et al., 1998). First, the primary sequence of dmRanBP2 lacks the internal repeat motif (IR) that forms the binding site between SUMO-conjugated RanGAP1 and RanBP2 in mammalian cells (Figure 2A) (Reverter and Lima, 2005, Werner et al., 2012). Second, dmRanGAP in flies migrated on SDS-PAGE with the same mobility as recombinant dmRanGAP expressed in E. coli (Supplemental Figure 2C), suggesting that it is not SUMOylated.

Figure 2. dmRanGAP directly interacts with dmRanBP2.

A. Schematic representations of human and fly RanBP2 homologs and two endogenous splicing isoforms of fly RanBP2. Note that although many functional domains of RanBP2s are highly conserved between two species, dmRanBP2 lacks the internal repeat (IR) domain. R, Ran binding domain; ZnF, Zinc finger domain; IR, internal repeat domain; and CHD, Cyclophilin homology domain.

B. Coomassie blue staining of a pull-down assay using anti-HA antibody. HA-dmRanGAP pulls down only dmRanBP2B-Frag but not dmRanBP2A-Frag.

C. Coomassie blue staining of a pull-down assay using streptavidin beads. The streptavidin beads were incubated with either biotinylated peptide containing aa 2197-2219 of dmRanBP2 (BP2 peptide) or randomized sequence with the same composition (Scrambled). HA-dmRanGAP can be pulled down by BP2 peptide but not by scrambled peptide.

D. Imaging of NeonGreen tagged dmRanGAP in enterocytes from 2^rd instar midgut. Compared to the nuclear envelope localization in control flies (ranbp2A/TM6B), dmRanGAP is displaced from the nuclear envelope in homozygous ranbp2A flies.

We performed a truncation analysis to understand dmRanGAP targeting, expressing dmRanGAP fragments in Drosophila S2 cells followed by immunoprecipitation of dmRanBP2. We found that the C-terminal domain of dmRanGAP is required for dmRanBP2 interaction (Supplemental Figure 2D). Reciprocally, we generated fragments of dmRanBP2 to map the dmRanGAP binding domain. We found a 450 amino acid fragment of dmRanBP2 (aa 1801-2250) that binds dmRanGAP (Supplemental Figure 2E). Previous reports (Smith et al., 2007) indicated that alternative splicing generates two isoforms of dmRanBP2 that differ by 23 amino acids within this region, which we will call dmRanBP2A and dmRanBP2B (Figure 2A). By analyzing Drosophila melanogaster RNA-Seq data from the Sequence Read Archive and RT-PCR (Fear and Oliver, 2018), we found that dmRanBP2B is the major isoform in most tissue types and dmRanBP2A is primarily testis-specific (Supplemental Table 2, Supplemental Figure 3). To examine whether these 23 amino acids are required for binding to dmRanGAP, we expressed and purified corresponding protein fragments using a Baculovirus expression system, one of which included the variant 23 amino acids plus flanking sequences (dmRanBP2B-frag) while the other contained only the flanking sequences (dmRanBP2A-frag). Using purified dmRanGAP as bait, we found that dmRanBP2B-frag bound dmRanGAP, but dmRanBP2A-frag did not, indicating that the 23 amino acids are necessary for dmRanGAP binding (Figure 2B). To test whether this small domain is sufficient for dmRanGAP binding, we synthesized the 23-amino-acid peptide with a N-terminal biotin and found that it precipitated purified dmRanGAP protein while a scrambled polypeptide of the same amino acid composition did not (Figure 2C).

Taken together, our findings show that dmRanGAP targets to the NE through dmRanBP2, as in mammals, but that this association relies on direct binding. There may be biological regulation of this association through expression of different isoforms of dmRanBP2 produced by alternative splicing.

dmRanGAP anchoring to dmRanBP2 is essential during pupation.

To study the function of dmRanGAP at the nuclear envelope, we generated a mutant fly line (ranbp2A) that bears a precise 69bp (23-amino acid) deletion at the ranbp2 (Nup358) gene locus and confirmed that ranbp2A flies express only the dmRanBP2A isoform and lacks the dmRanBP2B isoform (Supplemental Figure 3). We also fused a fluorescent marker, NeonGreen, to the C-terminus of the dmRanGAP at its genomic locus to its monitor cellular localization. dmRanGAP protein was displaced from the NE in intestinal cells at the 2^nd instar stage, indicating that the maternal pool of dmRanBP2B is already depleted at this stage (Figure 2D). These data strongly indicate that the 23-amino-acid motif present exclusively in the dmRanBP2B isoform is required for dmRanGAP binding in vivo.

The ranbp2A larvae appeared relatively normal at the end of the late 3^rd instar stage. To assess growth differences, we analyzed their eye and wing imaginal discs (Figure 3A, 3D). Mitotic furrow formation in eye imaginal discs of ranbp2A flies was similar to control flies (Figure 3A) and the size of eye and wing discs was not significantly different from those in controls, indicating that displacement of dmRanGAP had little effect on the tissue growth rates up to this stage of development. However, ranbp2A flies ceased to develop further in pupal stages (Figure 3B, C). A self-cross of heterozygous ranbp2A flies (balanced with TM6B, a 3^rd chromosome balancer with a dominant tubby phenotype) yields a near perfect 2:1 ratio of heterozygous to homozygous phenotypes in late 3^rd instar larvae (309 viable heterozygous larvae and 154 viable homozygous larvae), indicating the lethality occurs in pupal stages of ranbp2A flies. ranbp2A flies transited prepupal stage (white pupae), but showed disruption of their ecdysis sequences, including head eversion (Figure 3B, C, video 1).

Figure 3. Homozygous ranbp2A mutant demonstrates developmental arrest at early pupal stages.

A. Size analysis of eye imaginal discs. 3^rd instar eye imaginal discs from heterozygous or homozygous ranbp2A flies were collected and stained with DAPI. Mitotic furrow is marked by a white arrowhead. No statistically significant difference in size was observed (p= 0.2300). An unpaired T-test was used for the statistical analysis.

B. Cumulative step histogram of head eversion event during pupal stage. Late 3^rd instar larvae (5-day old) were collected as day 0. While head eversion occurred robustly in controls (both w1118 and heterozygous ranbp2A flies), no head eversion event was observed in homozygous ranbp2A flies.

C. Behavior analysis of pupal ecdysis. Heterozygous or homozygous ranbp2A flies were collected at the white pupal stage (APF: after puparium formation). The homozygous ranbp2A fly failed to undergo head eversion (white arrowhead) and air bubble translocation after 24 hours and the air bubble occupies most of the pupa (black arrowhead) after 96 hours.

D. Size analysis of wing imaginal discs. Third instar wing imaginal discs from heterozygous or homozygous ranbp2A flies were collected and stained with DAPI. No statistically significant difference in size was observed (p= 0.5929). An unpaired T-test was used for the statistical analysis.

Taken together, our data confirm the 23-amino-acid sequence omitted in the dmRanBP2A isoform is essential for dmRanGAP association to the NE in vivo. ranbp2A flies grow normally as larvae, indicating that soluble dmRanGAP is sufficient to sustain cellular functions up to this point in development. Remarkably, ranbp2A flies consistently fail to complete metamorphosis and arrest prior to head eversion. While their normal larval growth circumstantially argues against the idea that this phenotype arises from a lack of bulk nuclear transport, it is impossible to exclude the possibility that it reflects from mis-localization of some nuclear transport cargo(s). However, overexpression of dmRanGAP using a universal driver (tub-Gal4) and 10x UAS-RanGAP transgene did not rescue ranbp2A flies, indicating that defects in the developmental program do not simply arise from insufficient RanGAP activity, but rather that association to the NE per se is essential during development (data not shown). It thus remains plausible that dmRanGAP localization to the NE is required for some non-transport function that is essential at this point in development.

Restoration of dmRanGAP localization rescues the development of ranbp2A flies.

To assess whether the developmental arrest in ranbp2A flies reflects solely the displacement of dmRanGAP, we tested whether the lethality of ranbp2A flies is rescued by the following transgenes: P{dmRanBP2^WT}, P{dmRanBP2A}, P{dmRanBP2A-69bp} (dmRanBP2 A isoform with the 23-amino-acid motif fused at the C-terminus), and P{dmRanBP2A-dmRanGAP} (dmRanBP2 A isoform with dmRanGAP fused at the C-terminus) (Figure 4A). Pupal developmental arrest of ranbp2A flies was fully rescued by P{dmRanBP2^WT}, P{dmRanBP2A-69bp}, and P{dmRanBP2A-dmRanGAP} but not P{dmRanBP2A} (Figure 4B, C, and video 2). The rescue of P{dmRanBP2A-69bp} suggests that the 23-amino-acid motif itself is capable of recruiting dmRanGAP regardless of the flanking sequence of dmRanBP2, consistent with our in vitro binding results. Importantly, P{dmRanBP2A-dmRanGAP}, a direct fusion between dmRanGAP and dmRanBP2 lacking the 23-amino-acid motif, is still able to rescue ranbp2A flies. These results indicate that nuclear envelope anchorage of dmRanGAP but not other binding partners at the 23-amino-acid motif is required for ecdysis sequences at the pupal stages during metamorphosis. Notably, a direct fusion between dmRanBP2A and catalytically inactive dmRanGAP failed to rescue ranbp2A flies (Supplemental Figure 4), indicating that the Ran-GTP hydrolysis function of dmRanGAP is required for pupal development.

Figure 4. Anchorage of dmRanGAP to dmRanBP2 is required for the developmental processes.

A. Schematic designs of rescue constructs. dmRanGAP binding motif of dmRanBP2: red.

B. Cumulative step histogram of head eversion event during pupal stage in homozygous ranbp2A flies rescued by different constructs. Late 3^rd instar larvae (5-day old) were collected as day 0. Note that transgenes, P{dmRanBP2WT}, P{dmRanBP2A-69bp}, and P{dmRanBP2A-dmRanGAP}, rescued the developmental arrest of ranbp2A flies.

C. Behavior analysis of pupal ecdysis. All flies were collected at the white pupal stage (APF: after puparium formation). Note that the transgene, P{dmRanBP2A}, failed to rescue the ranbp2A flies and showed no events of head eversion (white arrowhead) after 24 hours and the air bubble occupies most of the pupa (black arrowhead) after 96 hours.

Discussion:

The development of CRISPR-based gene editing has provided straightforward strategies to assess the functional consequences of RanGAP localization to the NE, a feature that is shared among multicellular eukaryotes but not observed in fungi. We applied these tools in both human tissue culture cells and Drosophila melanogaster. Specifically, we disrupted RanGAP1 localization in DLD1 or HCT116 cells and found that RanGAP1 recruitment to the NE was dispensable for cell growth and proliferation in this context, as well as for nuclear import and export of a model protein substrate (Figure 1). We further utilized Drosophila to address the importance of RanGAP localization in a developmental context. Displacement of dmRanGAP from the NE did not cause growth defects of imaginal discs during larval stages, but resulted in developmental arrest during pupal stages (Figure 3B, and 4B). Our findings support the idea that RanGAP localization to the NE of multicellular organisms is largely dispensable for efficient function of the Ran pathway in individual cells but is important for some role(s) in development or organogenesis.

We observed that untethering dmRanGAP from RanBP2 caused no obvious defects in overall growth of Drosophila larvae before the onset of pupation (Figure 3A). As with mammalian tissue culture cells, these findings suggest that mis-localization of dmRanGAP does not decrease the efficiency of nuclear transport to such an extent that it becomes limiting for cell growth or function through the 3^rd instar stage of development. Thus, experiments in both systems argue against the idea that RanGAP localization is essential to support basic cellular function in multicellular eukaryotes.

By contrast, ranbp2A flies dramatically ceased development after puparium formation and arrested prior to head eversion (Figure 3B, C). This defect clearly arose from a deficit of dmRanGAP recruitment, as opposed to some other function of dmRanBP2B, because the phenotype was rescued through fusion of catalytically active dmRanGAP to dmRanBP2A (Figure 4). We envision two possible reasons why dmRanGAP becomes essential at this point in development. First, while our data in human colon cancer DLD1 cells does not suggest that RanGAP localization to the NE affects the efficiency of transport for cargoes bearing classical NLS and NES sequences, it is possible that in Drosophila, transport of certain cargo(s) with different classes of NLS/NES may be essential for pupal development. Alternatively, failure to undergo pupation may reflect some non-canonical function of dmRanGAP. For example, yeast RanGAP participates in silencing heterochromatin (Kusano et al., 2004, Nishijima et al., 2006) and disruption of dmRanGAP localization is associated with altered organization of satellite DNA sequences with the Responder (Rsp) element of the Segregation Distorter pathway in flies (Wu et al., 1988, Merrill et al., 1999). This RanGAP function appears to become important in the context of development, organogenesis or the expression of complex developmental pathways, aspects of RanGAP activity that could be less important in isolated cells or unicellular organisms. Interestingly, prevention of RanGAP tethering to the NE has no obvious phenotype in A. thaliana and these plants develop normally (Boruc et al., 2015), suggesting that localization of RanGAP to the NE might be important either for only a very specific developmental event or for an execution of the yet unknown function that is not manifested in laboratory conditions.

Finally, it is notable that the mechanisms through which plants (Rose and Meier, 2001, Jeong et al., 2005), mammals (Mahajan et al., 1997, Matunis et al., 1998) and flies (Figure 2) recruit RanGAP to the NE share only limited similarity. Arabidopsis RanGAP (AtRanGAP1) is anchored to the outer nuclear envelope through association to WPP domain-interacting proteins (AtWIPs), which are plant-specific Klarsicht/ANC-1/Syne homology (KASH) family members (Zhou et al., 2012). By contrast, mammalian cells form a multi-protein complex that contains SUMOylated RanGAP1 as wells as RanBP2 and the SUMO E2 enzyme, Ubc9 (Reverter and Lima, 2005). Like mammalian RanGAP1, we found that dmRanGAP binds to the NE through dmRanBP2, but it does not require SUMOylation of dmRanGAP or recruitment of the SUMO E2 enzyme, Ubc9 (Supplemental Figure 2C). Rather, binding of dmRanGAP to dmRanBP2 is mediated through direct protein-protein interactions and appears to be regulated through expression of dmRanBP2 isoforms selectively including or excluding a 23-amino acid sequence that is necessary and sufficient for dmRanGAP recruitment. It is an open question as to whether these distinct targeting mechanisms reflect an evolutionary convergence for RanGAP recruitment to the NE or possible divergence over time to allow this association to be regulated by alternative means, such as splicing or SUMOylation.

Collectively, our results indicate that while the localization of RanGAP to the NE is widely conserved in multicellular organisms, the targeting mechanisms are not. Our data further suggest that this localization may not be essential at the level of individual cells but promotes critical processes necessary for development. It will be interesting to explore the regulation of RanGAP in different tissues in the future, in order to understand the interplay of these components in regulating development.

Limitations of study

Although we found that both RanGAP-RanBP2 interaction and the catalytic function of RanGAP are critical for pupal development, the question of how the tissue development connects with the enzymatic reaction of RanGAP remains unresolved. A thorough interrogation of tissue- and developmental stage-specific rescue would be necessary to uncover the process(es) that is driven by the RanGAP-RanBP2 interaction. We also do not know whether RanBP2-RanGAP1 interaction in vertebrates is similarly critical for embryogenesis. Both these are important questions that we plan to address in our future studies.

STAR Methods:

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Dr. Mary Dasso (dassom@mail.nih.gov).

Materials availability

All unique/stable reagents generated in this study are available from the Lead Contact with a completed Materials Transfer Agreement.

Data and Code Availability

The published article includes all datasets generated or analyzed during this study.

This paper does not report original code.

Any additional information required to reanalyze the data reported in this work paper is available from the Lead Contact upon request.

Experimental model and subject details

Human colorectal carcinoma tissue culture cells HCT116 were cultured in McCoy’s 5A (ATCC) supplemented with heat-inactivated 10% FBS (Atlanta Biologicals) and antibiotics (100 IU/ml penicillin and 100 μg/ml streptomycin) in 5% CO2 atmosphere at 37°C.

Human colorectal carcinoma tissue culture cells DLD1 were cultured in DMEM (Life Technologies) supplemented with heat-inactivated 10% FBS (Atlanta Biologicals), antibiotics (100 IU/ml penicillin and 100 μg/ml streptomycin) and 2 mM GlutaMAX (Life Technologies) in 5% CO₂ atmosphere at 37°C.

Flies were maintained on standard Drosophila food (Bloomington formulation, LabExpress) at 25°C unless otherwise specified. The Drosophila lines used include: W1118 (wild-type control; Bloomington Drosophila Stock Center), w1118; ranbp2A/TM6B,Tb¹,Hu¹, w1118 ;RanGAP-NeonGreen, w1118; P{dmRanBP2WT}, w1118; P{dmRanBP2A}, w1118; P{dmRanBP2A-69bp}, w1118; P{dmRanBP2A-dmRanGAP}, w1118; P{dmRanBP2A-dmRanGAPΔ}, w1118; P{dmRanBP2A-dmRanGAPΔ^DA}, and w1118; P{dmRanBP2A-dmRanGAPΔ^RA}. Transgenic and CRISPR/Cas9 gene edited fly lines were generated by BestGene Inc.

S2 Schneider cells were cultured in Shield’s and Sang’s M3 media (Sigma), 1X Insect Media Supplement (IMS), 1X Pen/Strep, 2% FBS at 26°C and 3% CO₂.

Method details

Cell culture and gene targeting

For transfection, 4 x 10³ cells/well were plated in 24-well plates a day before transfection. Plasmids for transfection were prepared using the NucleoSpin buffer set (Clontech) and VitaScientific columns. Cells for gene targeting were transfected with 250 ng of donor and 125ng of 2 gRNA plasmids using ViaFect (Promega) transfection reagent according to the manufacturer’s instruction. Seventy-two hours after transfection, cells were seeded on 10-cm dishes with the selective antibiotics (puromycin 1 μg/ml) until clones were formed on a plate. Analysis of localization and expression of targeted proteins in clones were performed after 14- to 20-day period post transfection on a 24-well lumox (Sarstedt) plates. Clones with proper protein localization were propagated for genomic PCR and western blot analysis in regular complete media without selective antibiotics.

Results were comparable in both cell lines although DLD-1 cell lines were preferentially used in the imaging-based experiments due to their flatter cellular morphology.

Construction of human cell lines

The CRISPR/Cas9 system was used for endogenous gene targeting. For human cell lines, sgRNA plasmids were generated with primers listed in Supplemental Table 1 and integrated into pX330 vector using Zhang Lab General Cloning Protocol (Cong et al., 2013). For sgRNA selection we used CRISPR Design Tools from http://crispr.mit.edu:8079 and https://figshare.com/articles/CRISPR_Design_Tool/1117899. The donor vector, pCassette-RanGAP1-wildtype-mCherry-P2A-Puro, was built from pEGFP-N1 vector (Clontech). Briefly, homology left arm (synthesized by Eurofin USA), mCherry-P2A-Puromycin resistant gene, and homology right arm were cloned into pEGFP-N1 vector between NdeI and NotI restriction sites using Gibson assembly method to replace CMV promoter and EGFP sequences. For generating the donor vector, pCassette-RanGAP1-KR-mCherry-P2A-Puro, PCR amplification using a pair of primers flanking the region encoding lysine 524 was used to introduce an A to G conversion. The PCR product was then cloned into pCassette-RanGAP1-WT-mCherry-P2A-Puro between NdeI and FseI sites. The donor vectors for RanGAP1 targeting with FLAG tag fusion, pCassette-RanGAP1-wildtype-FLAG-P2A-Puro and pCassette-RanGAP1-KR-FLAG-P2A-Puro, were generated by directly inserting short dsDNA encoding FLAG tag sequence into pCassette-RanGAP1-WT-mCherry-P2A-Puro and pCassette-RanGAP1-KR-mCherry-P2A-Puro plasmids, respectively, between FseI and BamI restriction sites.

All plasmid constructions were generated with Platinum SuperFi DNA Polymerase (Invitrogen) with primers/synthesized oligos listed in Supplemental Table 1.

Genotyping human cell lines

Genomic DNA of targeted DLD1 or HCT116 cells was extracted with Wizard® Genomic DNA Purification Kit (Promega). Clones were genotyped by PCR for biallic editing with Platinum SuperFi DNA Polymerase (Invitrogen) with primers listed in Supplemental Table 1. The PCR products were analyzed by electrophoresis in 1% agarose gel with ethidium bromide.

Sample preparation from whole cell lysate and western blot

RanGAP1^WT or RanGAP1^KR DLD1 cells were grown in a 12-well plate. Cells were harvested at 80% confluency and centrifuged at 500 g for 5 min. Pellets of the cells were lysed in 2x Laemmli sample buffer, boiled for 15 min at 100°C, and ultra-centrifuged at 117K for 10 min at 20°C. The protein samples were separated by SDS-PAGE for 1 hr and blotted onto PVDF membrane. The membrane was blocked in 5% non-fat milk for 1 hr with anti-RanGAP1 or anti-tubulin primary antibodies overnight at 4°C. The membrane was then washed 3 times with 1x TNT buffer (150 mM NaCl, 1 mM Tris-HCl, pH 7.5, 0.1% tween-20) and probed for 2 hr with the secondary anti-mouse or anti-rabbit antibodies conjugated with HRP in a dilution of 1:5000. Detection of the signal was performed using SuperSignal WestPico PLUS Chemiluminescent Substrate (ThermoScientific).

Mitotic timing measurement by time-lapse fluorescence microscopy

To visualize DNA signal in real time, we used CRISPR/Cas9 to fuse infra-red fluorescent Protein (IFP) to the C-terminus of RCC1 protein (Regulator of Chromosome Condensation 1). RCC1-IFP transfected HCT116 cells were grown on a 4-well glass bottom chamber (ibidi), imaged on the Eclipse Ti2 inverted microscope (Nikon), equipped with an Ultraview spinning disk confocal system (Ultraview Vox Rapid Confocal Imager; PerkinElmer) and controlled by Volocity software (PerkinElmer) utilizing 40x/1.3 oil Nikon PlanFluor immersion objective lens. Cells were imaged once every 5 min in FluoroBrite DMEM (ThermoFisher) media. The microscope was equipped with temperature-, CO₂- and humidity-controlled chamber that maintained a 5% CO₂ atmosphere and 37°C.

Protein transport assay

The nuclear transport assay was adapted from the protocol outlined in (Niopek et al., 2016). RanGAP1^WT and RanGAP1^KR DLD-1 cell lines were seeded on a 4-well glass bottom chamber (ibidi) in complete DMEM media. Cells were then transfected with 500 ng of NLS-mCherry-LEXY plasmid (pDN122) using ViaFect (Promega) transfection reagent according to the manufacturer’s protocol and changed to FluoroBrite DMEM (ThermoFisher) media. Post-transfection, cells were kept in the dark for 24 hr prior to imaging. Cells expressing NLS-mCherry-LEXY were excited with a 561 nm laser line with 30 ms exposure. The cells were then exposed with 405 nm laser line with 1 s exposure to induce nuclear export of the model substrate for 15 minutes. The 405 nm laser was then shut off to induce nuclear import of the model substrate for 20 minutes. The cells were then exposed to 405 nm laser again to induce another around of nuclear export of the model substrate for 15 min. During the course of experiment, cells were imaged every 30 seconds using 561 nm laser to follow mCherry signal of the model substrate and microscope was maintained 5% CO₂ atmosphere and 37°C. Cells were imaged using the Eclipse Ti2 inverted microscope (Nikon), equipped with an Ultraview spinning disk confocal system (Ultraview Vox Rapid Confocal Imager; PerkinElmer) and controlled by Volocity software (PerkinElmer) utilizing Nikon CFI60 Plan Apochromat Lambda 60x/1.4 oil immersion objective lens with D-C DIC slider 60x II.

Recombinant proteins and HA-tag pull down assay

cDNA for dmRanGAP and gene fragment of dmRanBP2B isoform (aa 2151-2250, dmRanBP2B-Frag) were amplified by PCR from cDNA library generated from mRNA isolated from whole fly lysate [yw]. The PCR product of HA-tagged dmRanGAP1 were inserted to the pFastBac vector between BamHI and NheI sites. To generate gene fragment of dmRanBP2A isoform (dmRanBP2A-Frag), a 69bp internal deletion was introduced by PCR amplification using a pair of primers (Supplemental Table 1) and dmRanBP2B-Frag sequence as the template. The PCR products of dmRanBP2A-Frag and dmRanBP2B-Frag were then cloned into pFastBac-HT vectors between EcoRI and NotI sites and expressed as 6His-tagged versions in Sf9 insect cells using Bac-to-Bac system (Invitrogen). The 6His-tagged proteins were prepared in baculovirus-expressed Sf9 insect cells as described previously (Arnaoutov and Dasso, 2014).

For HA-tag pull down assay, 5 μg of either dmRanBP2A-Frag or dmRanBP2B-Frag was mixed with 6 μg of HA-RanGAP in 200 μL of binding buffer (PBS, 1mM DTT, and 0.1% tween-20). Next, 20 μL of anti-HA beads (Roche) was added to the mixture and incubated for 2 hr at 4°C under constant rotation. Beads were pelleted down at 500 g for 10 s, washed 3 times with 500 μL binding buffer. 2X Laemmli sample buffer was then added to the pelleted beads and boiled for 5 min before analyzed by SDS-PAGE with Coomassie blue staining.

Construction of fly lines

For Drosophila gene targeting, the designs of sgRNA and donor plasmids were based on flyCRISPR (https://flycrispr.org) with the following modifications. sgRNA plasmids were generated using flyCRISPR Target Finder (Supplemental Table 1) (Gratz et al., 2014, Iseli et al., 2007) and integrated into pU6-BbsI-chiRNA vector. The donor vector for targeting RanGAP, pSHD-DsRed-DmRanGAP-NG, was built from pScarlessHD-DsRed by first cloning homology left arm and NeonGreen sequence between two AarI sites and then cloning homology right arm between two SapI sites using Gibson assembly method. The donor vector for targeting RanBP2 (Nup358), pSHD-DsRed-DmRanBP2Δintron was built from pScarlessHD-DsRed by sequentially cloning homology left and right arms between 2 AarI sites and 2 SapI sites, respectively, using Gibson assembly method.

Constructs for P-element insertion of RanBP2 rescue genes were all derived from an intermediate plasmid, pCaSpeR4-dmRanBP2Step1. The plasmid was generated by inserting the majority of 5’ dmRanBP2 genomic fragment sequence (from 500bp upstream of dmRanBP2 gene to the donor site of dmRanBP2 alternative splicing intron) into pCaSpeR4 between EcoRI and XhoI restriction sites. AvrII restriction site was introduced by the reverse primer sequence (Supplemental Table 1) for insertion of wildtype dmRanBP2 or mutant dmRanBP2 fragments using Gibson assembly method.

P{dmRanBP2^WT} is a transgene consist of wildtype dmRanBP2 gene. The injecting construct, pCaSpeR4-dmRanBP2^WT was generated by cloning the remaining 3’ dmRanBP2 genomic fragment into AvrII restriction site of pCaSpeR4-dmRanBP2Step1. P{dmRanBP2A} is a transgene consist of dmRanBP2 gene without a 69bp internal sequence that is absent in dmRanBP2 A isoform. The injecting construct, pCaSpeR4-dmRanBP2A, was generated by cloning 3’ dmRanBP2 genomic fragment omitting the 69bp internal sequence into the AvrII restriction site of pCaSpeR4-dmRanBP2Step1. P{dmRanBP2A-69bp} is a transgene consist of dmRanBP2 gene with the 69bp internal sequence translocated to the 3’ end of the dmRanBP2 gene before the stop codon. The injecting construct, pCaSpeR4-dmRanBP2A-69bp, was generated by cloning 1) 3’ dmRanBP2 gene fragment omitting the 69bp internal sequence, 2) the 69bp internal sequence, and 3) 3’UTR into AvrII restriction site of pCaSpeR4-dmRanBP2Step1. P{dmRanBP2A-dmRanGAP} is a transgene consist of dmRanBP2 gene with a precise 69bp internal deletion and dmRanGAP cDNA fused to the 3’ end of dmRanBP2 gene before the stop codon. The injecting construct, pCaSpeR4-dmRanBP2A-dmRanGAP, was generated by cloning 1) 3’ dmRanBP2 gene fragment omitting the 69bp internal sequence, 2) dmRanGAP cDNA, and 3) 3’UTR into AvrII restriction site of pCaSpeR4-dmRanBP2Step1. P{dmRanBP2A-dmRanGAPΔ} is a transgene consist of dmRanBP2 gene with a precise 69bp internal deletion and cDNA of dmRanGAP N-terminal catalytic domain fused to the 3’ end of dmRanBP2 gene before the stop codon. The injecting construct, pCaSpeR4-dmRanBP2A-dmRanGAPΔ, was generated by cloning 1) 3’ RanBP2 gene fragment omitting the 69bp internal sequence, 2) cDNA of RanGAP N-terminal catalytic domain, and 3) 3’UTR into AvrII restriction site of pCaSpeR4-dmRanBP2Step1. P{dmRanBP2A-dmRanGAPΔ^DA} and P{dmRanBP2A-dmRanGAPΔ^RA} are the same as P{dmRanBP2A-dmRanGAPΔ} except an alanine-substituted mutation at asparagine 241 (DA) or arginine 87 (RA) to abolish the Ran-GTP hydrolysis activity (Kusano et al., 2001). The injecting construct, pCaSpeR4-dmRanBP2A-dmRanGAPΔ^DA and pCaSpeR4-dmRanBP2A-dmRanGAPΔ^RA were generated by cloning 1) 3’ RanBP2 gene fragment omitting the 69bp internal sequence, 2) synthesized gene strands encoding N-terminal catalytic domain of dmRanGAP with either D241A or R87A mutation, and 3) 3’UTR into AvrII restriction site of pCaSpeR4-dmRanBP2Step1.

All plasmid constructions were generated with Platinum SuperFi DNA Polymerase (Invitrogen) with primers listed in Supplemental Table 1.

RT-PCR

Total RNA of P{dmRanBP2A}; ranbp2A (ranbp2A isoform only flies) larvae (3^rd instar), w1118 3rd instar larvae, early pupae and adult, and w1118 testis (10 each) were extracted by TRIzol (Invitrogen). Genomic DNA was removed by TURBO DNA-free kit (Ambion) and 4ug of RNA was used for cDNA synthesis with oligo(dT) primers by AMV reverse transcriptase (NEB). For RT-PCR amplification of ranbp2, an amplification using “General forward and reverse” or “Isoform specific forward and reverse” primer pairs were done with a denaturation step at 95°C for 5 min, followed by 40 cycles of denaturation at 98°C for 30s, primer annealing at 62°C for 30s, and primer extension at 72°C for 30s. Upon completion of the cycling steps, a final extension at 72°C for 10 min was done and then the reaction was stored at 4°C. The PCR products were analyzed by electrophoresis in 1.2 % agarose gel with ethidium bromide.

Genotyping fly lines

The CRISPR/Cas9 gene edited fly stocks were genotyped by non-lethal genotyping using adult wings with the following modifications (Carvalho et al., 2009). In brief, adult wings were collected and placed at the bottom of a 0.2 mL PCR tube. The wings were then directly covered by 20 μL of the PCR mix with Platinum SuperFi DNA Polymerase and primers (Supplemental Table 1) for PCR amplification. The PCR products were analyzed by electrophoresis in 1% agarose gel with ethidium bromide.

Antibodies

Anti-dmRanGAP rabbit polyclonal antibody was raised against full length Drosophila RanGAP.

Imaging of Drosophila larval tissue

Second instar larval guts were dissected in PBS, transferred to an Eppendorf tube containing 4 % paraformaldehyde in PBS, and fixed for 5 min at RT on a rotator. The samples were pelleted at 500 g for 10 s, washed once for 5 min with TBS-T, containing 1% TritonX-100, and stained with Hoechst 33342 (Invitrogen) for 5 min. The samples were then washed once with TBS-T and mounted with VECTASHIELD® antifade mounting medium (Vector Laboratories).

Images were captured with Eclipse Ti2 inverted microscope (Nikon), equipped with an Ultraview spinning disk confocal system (Ultraview Vox Rapid Confocal Imager; PerkinElmer) and controlled by Volocity software (PerkinElmer) utilizing Nikon CFI60 Plan Apochromat Lambda 60x/1.4 oil immersion objective lens with D-C DIC slider 60x II. NeonGreen and Hoechst 33342 signals were excited with 488 nm and 408 nm laser lines, respectively. A series of 0.5 μm optical sections were acquired. Images were captured and analyzed using Volocity (PerkinElmer) and Image J (National Institutes of Health) software, respectively. Images represent maximum intensity projections of entire z-stacks.

Size measurement of Drosophila imaginal discs

Late 3^rd instar eye or wing imaginal discs were dissected and dissected in PBS, transferred to an Eppendorf tube containing 4 % paraformaldehyde in PBS, and fixed for 1 hr at RT on a rotator. The samples were pelleted at 500 g for 10 s, washed three times for 10 min with TBS-T, containing 1 % TritonX-100, then blocked for 1 hr at RT in TBS-T, containing 10 % normal goat serum. The samples were then stained with Hoechst 33342 for 10 min at RT on a rotator. The specimen were washed as above, then mounted with VECTASHIELD® antifade mounting medium (Vector Laboratories).

Images were captured on the Eclipse Ti2 inverted microscope (Nikon), equipped with an Ultraview spinning disk confocal system (Ultraview Vox Rapid Confocal Imager; PerkinElmer) and controlled by Volocity software (PerkinElmer) utilizing 40x/1.3 oil Nikon PlanFluor immersion objective lens. Hoechst 33342 signals was excited with a 408 nm laser line.

Behavior analysis

White pupae were collected and coated with mineral oil (Sigma-Aldrich) to increase their transparency and placed ventral side down on a cover glass, which was then attached to a glass slide with molding putty to form a small chamber. Behavior was recorded from the ventral side for 24 hr with Teslong USB Microscope (Teslong). Video were recorded and processed by Photo Booth (Apple Inc., Cupertino, CA) software and ImageJ (National Institutes of Health) software, respectively.

Immunofluorescence staining of Drosophila S2 cells

Immunofluorescence staining of dmRanBP2 and dmRanGAP was performed using S2 Schneider Drosophila melanogaster cell lines, cultivated on coverslips coated with Concavalin A (MP Biomedical) in Shield’s and Sang’s M3 media (Sigma), 1X Insect Media Supplement (IMS), 1X Pen/Strep, 2% FBS at 26°C and 3% CO₂. After reaching 80% confluency, the cells were washed once with PBS and fixed with cold methanol during 15 min at −20°C, and then rehydrated with TBST, containing 50 mM Tris-HCl pH 7.5, 150mM NaCl, and 0.05 % TX-100. The samples were then blocked for 1 hr at RT in TBST with 10 % normal goat serum and incubated overnight at 4°C with rabbit anti-dmRanGAP (1:300) and chicken anti-dmRanBP2 (1:300) primary antibodies. After removing primary antibodies, the sample were washed 3 times 5 min each in TBST and incubated with the corresponding secondary antibodies (1:500; Alexa-labeled goat anti-rabbit, anti-chicken (Invitrogen)) and Hoechst 33342 for 2 hr at RT. The specimen were washed as above, then mounted in ProLong Diamond Antifade Mountant (ThermoFisher Scientific).

Immunoprecipitation assay

Drosophila S2 cells were scraped and pelleted down at 500 g for 1 min and the pellet was resuspended by passing through 22 g needle in ice-cold IP buffer, 20mM Hepes pH 7.0; 150 mM NaCl; 2.5mM MgCl2; 0.1% NP-40; 0.2% Empigen BB, with addition of protease and phosphatase inhibitors cocktails (Roche). All further procedures were performed at 4°C. After 10 min extraction, insoluble materials were removed by centrifugation (14 000 g, 5 min) and the supernatant was used for immunoprecipitation. For each immunoprecipitation, a lysate from 4x10⁷ S2 cells, 25 ug of Rabbit anti-RanGAP1 or chicken ant-RanBP2 antibodies and 70 ul of Protein A Dynabeads ®, (Life Technologies) were used. After 1.5 hr incubation, beads were washed with IP buffer 3 times, and then immunoprecipitation products were eluted with 30 μL of 0.2 M glycine pH 2.4. The eluent was neutralized with 5 μL Tris pH 8.0 and boiled in 2X Laemmli sample buffer. Protein samples were resolved by SDS-PAGE, blotted onto PVDF membrane. The membrane was blocked in 5% non-fat milk for 1 hr with anti-dmRanGAP or anti-dmRanBP2 primary antibodies overnight at 4°C. The membrane was then washed 3 times with 1x TNT buffer (150 mM NaCl, 1 mM Tris-HCl, pH 7.5, 0.1% tween-20) and probed for 2 hr with the secondary anti-rabbit or anti-chicken antibodies conjugated with HRP in a dilution of 1:5000. Detection of the signal was performed using SuperSignal WestPico PLUS Chemiluminescent Substrate (ThermoScientific).

Molecular weight comparison between endogenous and recombinant dmRanGAP by SDS-PAGE

cDNA for dmRanGAP was amplified by PCR from cDNA library generated from mRNA isolated from whole fly lysate [yw]. The PCR products were then cloned into pET28a vector (Novagen) between NcoI and XhoI and expressed in BL21-CodonPlus(DE3-RIL cells (Agilent Technologies).

To harvest the recombinant dmRanGAP expressed in E. coli, 20 μL of overnight cultured E. coli were inoculated in 1 mL fresh lysogeny broth (LB) and incubate for 2.5 hr at 37°C. The cells were cultured for another 4 hours with or without 1mM IPTG. The pelleted samples were then resuspended with 2X Laemmli sample buffer and boiled for 15 min at 100°C. For protein preparation from the whole fly lysate, 10 3^rd instar larvae were homogenized in 2X Laemmli sample buffer, and boiled for 15 min at 100°C. The whole fly lysate and bacterial lysate were then analyzed by SDS-PAGE and western blot with anti-dmRanGAP antibody.

Analysis of RNA-Seq data from the Sequence Read Archive

To identify tissues that express the dmRanBP2 A isoform, we utilized an RNA-Seq dataset that remapped and combined over 8,000 Drosophila RNA-Seq results from NCBI Sequence Read Archive (GEO Accession ID: GSE117217). From the raw sequencing results, we searched and counted reads that contain dmRanBP2 A-specific splicing junction (Chromosome 3 right arm, from 25,256,856 to 25,256,924 bp in BDGP Drosophila reference genome release 6).

Quantification and statistical analysis

To quantify the cell growth, HCT116 cells were seeded at 5 x 10³ using a 12-well plate and DLD-1 cells were seeded at 2 x 10⁴ using 6-well plates. Cell density was measured directly by cell counter (TC10 automated cell counter, BioRad) after culturing for 1, 2, 3, and 4 days. The experiment was performed in triplicate for each time point. An unpaired T-test was used for the statistical analysis. GraphPad Prism software was used to generate graphs and calculate statistical values. N in each graph (Figure 1D, and Supplemental Figure 1C.) represents the number of replicates.

To quantify mitotic timing, IFP signals were excited by a 640 nm laser with 300 ms exposure at each time point. A series of 0.5 μm optical sections were acquired every 5 min. Images were captured and analyzed using Volocity (PerkinElmer) and Image J (National Institutes of Health) software, respectively. Images represent maximum intensity projections of entire z-stacks. An unpaired T-test was used for the statistical analysis. GraphPad Prism software was used to generate graphs and calculate statistical values. N in Figure 1E represents the number of cells in which the mitotic timing were measured.

To measure the rate of protein import and export, we measured fluorescent intensity at each time point from three locations in both cytoplasm and nucleus of each cell to obtain average intensity. Adjusted nuclear and cytoplasmic values were then calculated by subtracting their respective average values by the value of the background (a point near each measured cell without fluorescent protein signal). Percentage of fluorescence intensity was then calculated by the quotient of the adjusted fluorescence intensity divided by the sum of the adjusted nuclear and cytoplasmic values. N represents the number of recorded protein shuttling events. Image analysis was performed on Volocity (PerkinElmer) and ImageJ (National Institutes of Health) software with Time Series Analyzer V3 plugin and ROI Manager dialog box. The graphs, first-order kinetic model and the apparent rate were generated by GraphPad Prism software.

To quantify the size of imaginal discs, the areas of eye or wing imaginal discs were measured by Image J (National Institutes of Health) software. An unpaired T-test was used for the statistical analysis. GraphPad Prism software was used to generate graphs and calculate statistical values. N in graphs (Figure 3A, 3D) represents the number of imaginal discs being measured.

To obstain the cumulative frequency of head eversion, fertilized eggs were collected on the same day to synchronize the growth, and 58-70 late 3rd instar larvae (5-day old) were collected in vials containing food (Bloomington formulation, LabExpress) as day 0. Number of head eversion events was recorded daily for 8 days. GraphPad Prism software was used to generate graphs. N in graphs (Figure 4B and Supplemental Figure 6) represents the number of flies being recorded.

Supplementary Material

1

Supplemental video 1: Behavior analysis of control and ranbp2A flies during pupal development. Related to Figure 3.

2

Supplemental video 2: Behavior analysis of ranbp2A flies rescued by different transgenes during pupal development. Related to Figure 4.

4

Supplemental table 1: Sequences of oligonucleotides used in this study. Related to STAR Methods.

5

Supplemental table 2: Splcing events of dmRanBP2 gene from RNA-Seq results deposited in Sequence Read Archive. Related to Figure 2.

Key Resources Table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Rabbit polyclonal anti-human RanGAP1	(Joseph et al., 2004)
Mouse monoclonal anti-alpha-tubulin (Clone DM1A)	Sigma-Aldrich	T6199
Chicken polyclonal anti-Drosophila RanBP2	(Arnaoutov et al., 2020)
Rabbit polyclonal anti-Drosophila RanGAP	This manuscript
HRP conjugated sheep anti-mouse IgG antibody	GE Healthcare Life Sciences	NA931V
HRP conjugated donkey anti-rabbit IgG antibody	GE Healthcare Life Sciences	NA934V
Bacterial and Virus Strains
NEB® 5-alpha Competent E. coli	NEB	C2987H
MAX Efficiency® DH10Bac™ Competent E. coli	Invitrogen	10359-016
Chemicals, Peptides, and Recombinant Proteins
Biotinylated BP2 peptide (VRPTTHEVIPPLPMTLPLLTLPQ)	LifeTein
Biotinylated scrambled peptide (VPRMTPTLTPHLELVTPILPPQL)	LifeTein
Anti-HA Affinity Matrix (3F10, monoclonal)	Roche	11815016001
Dynabeads M-280 Streptavidin	ThermoFisher Scientific	11205D
Protein A Dynabeads	Life Technologies	10002D
Hoechst 33342	Invitrogen™	H3570
Platinum™ SuperFi™ DNA Polymerase	Invitrogen™	12351010
Mineral oil	Sigma-Aldrich	M5904
Experimental Models: Cell Lines
Human: DLD-1	ATCC	CCL-221
Human: HCT116	ATCC	CCL-247
Experimental Models: Organisms/Strains
D. melanogaster. w[1118];;ranbp2A/TM6B,Tb[1],Hu[1]	This manuscript
D. melanogaster. w[1118];RanGAP-NeonGreen	This manuscript
D. melanogaster. w[1118];P{dmRanBP2WT}	This manuscript
D. melanogaster. w[1118];P{dmRanBP2A}	This manuscript
D. melanogaster. w[1118];P{dmRanBP2A-69bp}	This manuscript
D. melanogaster. w[1118];P{dmRanBP2A-dmRanGAP}	This manuscript
D. melanogaster. w[1118];P{dmRanBP2A-dmRanGAPΔ}	This manuscript
D. melanogaster. w[1118];P{dmRanBP2A-dmRanGAPΔDA}	This manuscript
D. melanogaster. w[1118];P{dmRanBP2A-dmRanGAPΔRA}	This manuscript
Oligonucleotides
Primers for cloning and genotyping, see Supplemental Table 1.	Eurofins USA
Oligo synthesis of left homology arm for RanGAP1 (wildtype)	Eurofins USA
Oligo synthesis of left homology arm for RanGAP1 (K524R)	Eurofins USA
Recombinant DNA
pCassette-RanGAP1IC-wildtype-FLAG-P2A-Puro	This manuscript
pCassette-RanGAP1IC-KR-FLAG-P2A-Puro	This manuscript
pCassette-RanGAP1IC-wildtype-mCherry-P2A-Puro	This manuscript
pCassette-RanGAP1IC-KR-mCherry-P2A-Puro	This manuscript
pX330	(Cong et al., 2013)	Addgene plasmid #42230
pX330-RanGAP1sgRNA.3	This manuscript
pX330-RanGAP1sgRNA.4	This manuscript
pScarlessHD-DsRed	a gift from Kate O’Connor-Giles	Addgene plasmid #64703
pSHD-DsRed-DmRanGAP-NG	This manuscript
pSHD-DsRed-DmRanBP2Δintron	This manuscript
pU6-BbsI-chiRNA	(Gratz et al., 2013)	Addgene plasmid #45946
pU6-BbsI-sgGAP-1	This manuscript
pU6-BbsI-sgGAP-2	This manuscript
pU6-BbsI-sgBP2-1	This manuscript
pU6-BbsI-sgBP2-2	This manuscript
pU6-BbsI-sgBP2-3	This manuscript
pCaSpeR4	(Thummel et al., 1988)
pCaSpeR4-dmRanBP2Step1AvrII	This manuscript
pCaSpeR4-dmRanBP2WT	This manuscript
pCaSpeR4-dmRanBP2A	This manuscript
pCaSpeR4-dmRanBP2A-69bp	This manuscript
pCaSpeR4-dmRanBP2A-dmRanGAP	This manuscript
P{dmRanBP2A-dmRanGAPΔ}	This manuscript
P{dmRanBP2A-dmRanGAPΔDA}	This manuscript
P{dmRanBP2A-dmRanGAPΔRA}	This manuscript
NLS-mCherry-LEXY (pDN122)	(Niopek et al., 2016)	Addgene plasmid #72655
Software and Algorithms
Photobooth (MacOS)	Apple Inc., Cupertino, CA
ImageJ Fiji v2.0.0-rc-68/1.52h	(Schindelin et al. 2012)	https://fiji.sc/; RRID: SCR_002285
GraphPad Prism v8	GraphPad Inc.	http://graphpad.com; RRID: SCR_002798
Other
Teslong USB Microscope	Teslong www.teslong.com	MS100

Highlights:

1. RanGAP1-RanBP2 interaction is not essential for cell growth in human cell lines.

2. Displacing RanGAP1 has no effect on model substrate transport with classic NLS/NES.

3. RanGAP-RanBP2 interaction is functionally conserved in Drosophila.

4. Loss of RanGAP-RanBP2 interaction in flies leads to pupal developmental arrest.

Abbreviations:

NE

nuclear envelope

Ran-GTP

GTP-bound Ran

Ran-GDP

GDP-bound Ran

NPC

nuclear pore complex

SUMO

Small Ubiquitin-related Modifier

RanBP2

Ran Binding Protein 2 also known as Nup358

IR

RanBP2 internal repeat motif

RanGAP

Ran GTPase activating protein